Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), June 2017 Revija za mikroelektroniko, elektronske sestavne dele in materiale letnik 47, številka 2(2017), Junij 2017 ISSN 0352-9045 UDK 621.3:(53+54+621+66)(05)(497.1)=00 ISSN 0352-9045 Informacije MIDEM 2-2017 Journal of Microelectronics, Electronic Components and Materials VOLUME 47, NO. 2(162), LJUBLJANA, JUNE 2017 | LETNIK 47, NO. 2(162), LJUBLJANA, JUNIJ 2017 Published quarterly (March, June, September, December) by Society for Microelectronics, Electronic Components and Materials - MIDEM. Copyright © 2016. All rights reserved. | Revija izhaja trimesečno (marec, junij, september, december). Izdaja Strokovno društvo za mikroelektroniko, elektronske sestavne dele in materiale – Društvo MIDEM. Copyright © 2016. Vse pravice pridržane. Editor in Chief | Glavni in odgovorni urednik Marko Topič, University of Ljubljana (UL), Faculty of Electrical Engineering, Slovenia Editor of Electronic Edition | Urednik elektronske izdaje Kristijan Brecl, UL, Faculty of Electrical Engineering, Slovenia Associate Editors | Odgovorni področni uredniki Vanja Ambrožič, UL, Faculty of Electrical Engineering, Slovenia Arpad Bürmen, UL, Faculty of Electrical Engineering, Slovenia Danjela Kuščer Hrovatin, Jožef Stefan Institute, Slovenia Matija Pirc, UL, Faculty of Electrical Engineering, Slovenia Matjaž Vidmar, UL, Faculty of Electrical Engineering, Slovenia Editorial Board | Uredniški odbor Mohamed Akil, ESIEE PARIS, France Giuseppe Buja, University of Padova, Italy Gian-Franco Dalla Betta, University of Trento, Italy Martyn Fice, University College London, United Kingdom Ciprian Iliescu, Institute of Bioengineering and Nanotechnology, A*STAR, Singapore Malgorzata Jakubowska, Warsaw University of Technology, Poland Marc Lethiecq, University of Tours, France Teresa Orlowska-Kowalska, Wroclaw University of Technology, Poland Luca Palmieri, University of Padova, Italy International Advisory Board | Časopisni svet Janez Trontelj, UL, Faculty of Electrical Engineering, Slovenia - Chairman Cor Claeys, IMEC, Leuven, Belgium Denis Đonlagić, University of Maribor, Faculty of Elec. Eng. and Computer Science, Slovenia Zvonko Fazarinc, CIS, Stanford University, Stanford, USA Leszek J. Golonka, Technical University Wroclaw, Wroclaw, Poland Jean-Marie Haussonne, EIC-LUSAC, Octeville, France Barbara Malič, Jožef Stefan Institute, Slovenia Miran Mozetič, Jožef Stefan Institute, Slovenia Stane Pejovnik, UL, Faculty of Chemistry and Chemical Technology, Slovenia Giorgio Pignatel, University of Perugia, Italy Giovanni Soncini, University of Trento, Trento, Italy Iztok Šorli, MIKROIKS d.o.o., Ljubljana, Slovenia Hong Wang, Xi´an Jiaotong University, China Headquarters | Naslov uredništva Uredništvo Informacije MIDEM MIDEM pri MIKROIKS Stegne 11, 1521 Ljubljana, Slovenia T. +386 (0)1 513 37 68 F. + 386 (0)1 513 37 71 E. info@midem-drustvo.si www.midem-drustvo.si Annual subscription rate is 160 EUR, separate issue is 40 EUR. MIDEM members and Society sponsors receive current issues for free. Scientific Council for Technical Sciences of Slovenian Research Agency has recognized Informacije MIDEM as scientific Journal for microelectronics, electronic components and materials. Publishing of the Journal is cofi- nanced by Slovenian Research Agency and by Society sponsors. Scientific and professional papers published in the journal are indexed and abstracted in COBISS and INSPEC databases. The Journal is indexed by ISI® for Sci Search®, Research Alert® and Material Science Citation Index™. | Letna naročnina je 160 EUR, cena posamezne številke pa 40 EUR. Člani in sponzorji MIDEM prejemajo posamezne številke brezplačno. Znanstveni svet za tehnične vede je podal pozitivno mnenje o reviji kot znanstveno-strokovni reviji za mikroelektroniko, elektronske sestavne dele in materiale. Izdajo revije sofinancirajo ARRS in sponzorji društva. Znanstveno-strokovne prispevke objavljene v Informacijah MIDEM zajemamo v podatkovne baze COBISS in INSPEC. Prispevke iz revije zajema ISI® v naslednje svoje produkte: Sci Search®, Research Alert® in Materials Science Citation Index™. Design | Oblikovanje: Snežana Madić Lešnik; Printed by | tisk: Biro M, Ljubljana; Circulation | Naklada: 1000 issues | izvodov; Slovenia Taxe Percue | Poštnina plačana pri pošti 1102 Ljubljana 57 Content | Vsebina 59 71 79 91 101 113 129 139 143 Journal of Microelectronics, Electronic Components and Materials vol. 47, No. 2(2017) Izvirni znanstveni članki A. Pajkanović, M. Videnović-Misić, G. M. Stojanović: Načrtovanje in karakterizacija 130 nm CMOS širokopasovnega ojačevalnika z nizkim šumom H. Uršič, A. Benčan, E. Khomyakova, S. Drnovšek, I. F. Mercioniu, K. Makarovič, D. Belavič, C. Schreiner, R. Ciobanu, P. Fanjul Bolado, B. Malič: Debele plasti Pb(Mg,Nb)O3–PbTiO3 na podlagah iz keramike z nizko temperature žganja D. Šekuljica, S. Badessi, M. Ferrante, M. Vidmar: Ugotovitve o uporabi Lune pri G/T meritvah antenskih sistemov frekvenčnega pasu X. A.Amsaveni, P.T Vanathi: Reverzibilno skrivanje podatkov na osnovi Radonove in diskretne valčne transformacije S. M. Kahar, V. C. Hong, L. C. Chuan, S. C B Gopinath, M. K. Md Arshad, L. B. Ying, F. K. Loong, U. Hashim, Y. Al-Douri: Sinteza nanodlačic iz silicijevega karbida z mikrov- alovnim segrevanjem: Vpliv temperature gretja T. Saje, M. Vidmar: Radioteleskop za 21 cm vodikovo črto M. Lešnik, D. Verhovšek, N. Veronovski, S. Gatarić, M. Drofenik, J. Kovač: Hidrotermalno sintentiziran TiO2 dopiran z N z visoko fotokatalitsko aktivnostjo v vidnem delu svetlobnega spektra Doktorske disertacije na področju mikroelektronike, elektronskih sestavnih delov in materialov v Sloveniji v letu 2016 Napoved in vabilo k udeležbi: 53. Mednarodna konferenca o mikroelektroniki, napravah in materialih z delavnico o materialih za pretvorbo energije in njihovih aplikacijah Naslovnica: Lokalni piezoelektrični odziv debele plasti 0.65Pb(Mg1/3Nb2/3)O3–0.35PbTiO3 določen z mikroskopom na atomsko silo s piezoelektričnim modulom (H. Uršič et al.) Original scientific paper A. Pajkanović, M. Videnović-Misić, G. M. Stojanović: Design and Characterization of a 130 nm CMOS Ultra-Wideband Low-Noise Amplifier H. Uršič, A. Benčan, E. Khomyakova, S. Drnovšek, I. F. Mercioniu, K. Makarovič, D. Belavič, C. Schreiner, R. Ciobanu, P. Fanjul Bolado, B. Malič: Pb(Mg,Nb)O3–PbTiO3 Thick Films on Metalized Low-temperature Co-fired Ceramic Substrates D. Šekuljica, S. Badessi, M. Ferrante, M. Vidmar: Considerations about the use of the Moon in X-band antenna G/T measurements A.Amsaveni, P.T Vanathi: Reversible Data Hiding Based on Radon and Integer Lifting Wavelet Transform S. M. Kahar, V. C. Hong, L. C. Chuan, S. C B Gopinath, M. K. Md Arshad, L. B. Ying, F. K. Loong, U. Hashim, Y. Al-Douri: Synthesis of Silicon Carbide Nanowhiskers by Mi- crowave Heating: Effect of Heating Temperature T. Saje, M. Vidmar: A Compact Radio Telescope for the 21 cm Neutral-Hydrogen Line M. Lešnik, D. Verhovšek, N. Veronovski, S. Gatarić, M. Drofenik, J. Kovač: Highly Efficient Photocatalytic Activity in the Visible Region in Hydrothermally Synthesized N-doped TiO2 Doctoral theses on Microelectronics, Electronic Components and Materials in Slovenia in 2016 Announcement and Call for Papers: 53nd International Conference on Microelectronics, Devices and Materials With the Workshop on Materials for Energy Conversion and Their Applications Front page: Local piezoelectric response of the 0.65Pb(Mg1/3Nb2/3)O3–0.35PbTiO3 thick film determined by piezo-response force microscope (H. Uršič et al.) 58 59 Original scientific paper  MIDEM Society Design and Characterization of a 130 nm CMOS Ultra-Wideband Low-Noise Amplifier Aleksandar Pajkanović1,2, Mirjana Videnović-Misić2, Goran M. Stojanović2 1Faculty of Electrical Engineering, University of Banja Luka, Bosnia and Herzegovina 2Faculty of Technical Sciences, University of Novi Sad, Serbia Abstract: The design of an ultra-wideband low noise amplifier is presented in this paper. Schematic level design is described, as well as integrated circuit layout techniques applied and post-layout simulation results. After fabrication using the standard 130 nm CMOS process node, on-chip characterization has been performed. The simulation and characterization results are presented analyzed and discussed in detail. Keywords: CMOS integrated circuits (IC); analog/radio-frequency (RF); ultra-wideband (UWB); low-noise amplifier (LNA); on-chip characterization Načrtovanje in karakterizacija 130 nm CMOS širokopasovnega ojačevalnika z nizkim šumom Izvleček: Članek obravnava ultra širokopasoven ojačevalnik z nizkim šumom. Predstavljena je shema, uporabljene tehnike integracijskega vezja in rezultati simulacij. Po izdelavi v standardni 130 nm CMOS tehnologiji je bila opravljena karakterizacija na nivoju čipa. Predstavljeni so simulacijski in karakterizacijski rezultati. Ključne besede: CMOS integrirana vezja; analogna radio frekvenca (RF); ultra široki pas (UWB); ojačevalnik z nizkim šumom (LNA); karakterizacija na čipu * Corresponding Author’s e-mail: aleksandar.pajkanovic@etf.unibl.org Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 59 – 70 1 Introduction Different applications employing ultra-wideband (UWB) systems are under investigation in the lower frequency range of radio-frequencies (RF), which is around 1-10 GHz [1]. Such applications of interest in- clude high resolution radars [2], medical imaging [3], communication systems [4] and many more. Physical layer in common for all these implementations employs UWB signals which are characterized with high relative bandwidths, wider than in any other standard commer- cialized until now [5]. This possesses plenty of new chal- lenges for the RF integrated circuit (IC) designers in an already complex engineering environment [6, 7]. UWB signal is defined in [1] as either a signal of abso- lute bandwidth (B) larger than 500 MHz, or a signal of relative bandwidth larger than 20 %, where relative bandwidth (Br) is calculated as follows: u d r c f fB f − = (1) where fu, fd and fc represent upper and lower band limit, and a central frequency, respectively. Documents defin- ing frequency ranges, emissions and other UWB regula- tions were released in the United States first in 2002 [8] and in EU, Japan, Korea, Singapore and China since. UWB technology may, thus, utilize a frequency range of up to of 3.1-10.6 GHz. The whole range of 7.5 GHz is used only in the USA. In EU, the UWB band is divided in two sub-bands: lower (3.168-4.752 GHz) and higher (6.336-8.976 GHz). In Japan the sub-bands are given as: lower (3.696-4.752 GHz) and higher (7.392-10.032 GHz), while Korea and China have their own specifica- tions. Other important UWB technology regulations 60 A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 include its applications’ definitions, such as indoor, out- door, portable, fixed installed; data speeds of up to 480 Mb/s; and maximum emission levels, i.e. power spectral density (PSD) measured in terms of equivalent isotro- pically radiated power (EIRP). Moreover, in some of the mentioned sub-bands, e.g. lower EU sub-band, inter- ference mitigation techniques are obligatory. As a con- sequence of such stringent regulations, total emitted power allowed is very low, and equal to -41.3 dBm/MHz. In the case of the full allowed specter (3.1-10.6 GHz), this means that the total transmitted power may not be greater than 0.56 mW. Therefore, commercial UWB transmission is limited to short range applications [1, 5]. To exploit these frequency ranges, there are two ap- proaches to the design of UWB communication sys- tems: impulse radio (IR), which is shown in Figure1a, and orthogonal frequency division multiplexing (OFDM), which is shown in Figure 1b. In the former case, the transmission is based on ultra-short pulses, thus cover- ing the whole available band or sub-band. In the latter case, the available UWB bandwidth is divided into a set of wideband OFDM channels [1]. IR-UWB technique is more appropriate for applications where simple modu- lation schemes, such as on-off keying, provide enough signal integrity and offer higher energy efficiency and lower cost [9]. Regardless of the system architecture, the front-end wideband low noise amplifier (LNA) is obligatory as the first stage of the receiver [5], Fugure 1. Such amplifier must meet several stringent requirements, e.g. broad- band input matching, sufficient gain with wide band- width, low noise figure, etc. [5, 10]. For the past decade CMOS represents standard technology in the RF IC do- main [11, 12, 13]. For the design presented in this paper, a standard eight-metal layer, 130 nm CMOS technology node was chosen. A variety of MOS devices are available, includ- ing low- and high-threshold versions, devices operat- ing at higher supply voltages (3.3 V) and devices in- tended for RF application. For circuit implementation presented within this paper RF MOS transistors are cho- sen, their nominal supply voltage (VDD) and transition frequency (ft) being 1.2 V and 105 GHz, respectively. In general, the transistors are capable for appropri- ate performance at frequencies up to 10 % of ft [14]. This means that UWB applications are feasible utilizing the MOS devices available within the selected process node. Additional advantage of this particular process node, in the context of RF IC design, is the availability of standard inductors. These are implemented in metal layer 8 and are all of spiral topology, either circular or rectangular, their inductance ranging from 100 pH up to 10 nH. Figure 1: Two UWB communication system architec- tures [5]: a) IR and b) OFDM In section 2 we provide a short introduction on figures of merit utilized within this paper for the LNA perfor- mance characterization. Then, in section 3 a brief over- view of the related work is given. In the sections that follow, we present an LNA designed to operate in the EU UWB upper sub-band (69 GHz). The schematic level design procedure is described in section 4. The char- acterization procedure and the results obtained are presented in section 5. In section 6 the results char- acterized and simulated are analyzed and discussed, whereas in section 7 a conclusion follows. 2 Figures of Merit In order to specify design requirements of an UWB am- plifier, one is to use similar notions to those used when specifying a narrowband amplifier [14], such as gain, noise figure and input matching. However, the main dif- ference is that these features must be achieved over a bandwidth of up to 10 GHz [1]. For example, according to Bode-Fano criterion [15], it is not possible to achieve arbitrary low reflection coefficient Γ(ω) in the arbitrary wide bandwidth, if there is a reactive component in the load. That is the reason why wideband amplifiers must show higher reflection coefficient than their narrow band counterparts with the same transistor dimen- sions. This means that the information on any of those specifications in the context of RF IC is complete only if given over a range of frequencies. Furthermore, since these frequencies in the case of UWB applications ex- tend well into microwave spectrum, we employ some figures of merit used in microwave engineering [11, 16] to precisely specify and, later on, characterize the LNA performance. Of course, different parameter values represent a standard in UWB case [17]. 2.1 Scattering parameters For a high frequency and broad bandwidth character- ization of a two port network a two-by-two scattering (S) parameters matrix is used [11, 16]: BDF T/R LNA Mixer LPF VGA ADC Antenna b) LO LNA Antenna Mixer Pulse generator LPF Demodulator a) 61 11 12 21 22 . S S S S S   =     (2) Each of the matrix members in equation (2) has physi- cal meaning: S11 – input reflection coefficient, S12 – reverse transmission, S21 – a sort of gain [16], as it relates output wave to in- put wave, S22 – output reflection coefficient. Normally, in the case of a LNA, the reflection coeffi- cients and the reverse transmission coefficient should be as low as possible, whereas the gain should be as high as possible. 2.2 Noise factor Three main sources of electrical devices noise are ther- mal, Schottky and flicker noise [16]. As opposed to the well known signal-to-noise ratio, S/N in the domain of RF IC design a parameter mostly used to present the in- formation on internal noise are the noise factor, F, and the noise figure, NF. Noise factor represents the ratio of signal-to-noise ratio at the input and signal-to-noise ratio at the output: ( ) ( ) / . / input output S N F S N = (3) Noise figure is a dB representation of noise factor, ob- tained as follows: 10log .NF F= (4) 2.3 Linearity Two parameters are used to characterize an amplifier in the aspect of linearity: 1-dB compression point, P1dB, and input-referred third-order intermodulation (IM) intercept point, IIP3. The former represents the upper limit of the input signal power for which the LNA pro- vides the expected output. This limit is defined as the input signal power which causes the real output to be less than expected output by exactly 1 dB [11, 16]. The latter figure of merit, IIP3, is required to take into account the influence of IM products; namely, the exis- tence of two signals of frequencies close to each other at the input, gives rise to IM products. Second-order IM products can be easily filtered out, but third-order products can rise at frequencies within the information signal bandwidth and, thus, cause linearity issues [11]. The number associated with IIP3 is obtained by bring- ing two signals of close frequencies and of equal am- plitudes to the circuit input. Then, both output signal power and output third-order IM product power are plotted versus input power signal. Extrapolation of those two curves yields an intercept point. The Pin at which the extrapolated intercept point appears is actu- ally the IIP3. These two figures of merit are related as follows [11]: 1dB 3 9.6 dB,IIP P− = (5) under the condition that all nonlinearities of the order higher than 3 can be neglected. 2.4 Stability Another working mode which amplification circuit may not enter during normal operation is oscillation. A figure of merit that needs attention in this context is circuit stability. It is possible to maintain the circuit stability at arbitrary input signal magnitudes (uncondi- tional stability). There are multiple parameters defined as stability factors, but those used within this paper are the µ and µ’ factor. The former represents the distance from the Smith chart center point to the area where in- stability occurs caused by the load. It is calculated as follows [15]: 2 11 * 22 11 12 21 1 , S S S S S µ − = − ∆ + (6) where: 11 22 12 21 .S S S S∆ = − (7) The latter is the distance from the center point to the area where instability occurs caused by the source. It is obtained in similar fashion: 2 22 * 11 22 12 21 1 . S S S S S µ − = − ∆ + ′ (8) A two-port network is unconditionally stable if µ >1 and µ’ >1. 3 Related Work Achieving broadband gain is a fundamental require- ment in an UWB receiver, which means that this is also necessary for any LNA – as it is the first stage of a receiv- A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 62 er in any of the cases mentioned in section 1. Depend- ing on the system architecture, the approximate band covered by the LNA in most cases is either of the three frequency ranges: (i) from 3.1 to 5 GHz (low band), (ii) 6 to 10.6 GHz (high band) or (iii) 3.1 to 10.6 GHz (full band). LNAs present in literature can also be classified accord- ing to the circuit topology applied to meet the require- ments for each of the figures presented in section 2. Those can be broadly categorized into four types, as follows [18]: - distributed amplifier, - input reactive networks, - resistive-feedback, and - common-gate circuits. These possible implementations are shown in Figure 2 at the highest level of abstraction. Figure 2: The standard wide bandwidth input match- ing techniques: a) distributed amplifiers, b) input re- active network, c) resistive-feedback and d) common- gate circuit Distributed amplifiers, 0a, provide wide bandwidth characteristics, but tend to consume large DC currents due to the distribution of multiple amplifying stages which makes them unsuitable for low-power applica- tions. Besides, such implementations contain a number of on-chip inductivities, so the whole circuit demands a larger area. In 0b, a topology which adopts a band- pass LC filter at the input of the LNA for wideband input matching is shown. The bandpass-filter-based topol- ogy incorporates the input impedance of the amplifier as a part of the filter, and shows good performances while dissipating small amounts of DC power. However, the inclusion of LC filter at the input demands a number of reactive elements, which introduce additional noise and increase the chip area needed. In 0c and 0d, resis- tive-feedback and common-gate topologies principles are shown, respectively. The resistive feedback based amplifiers provide good wideband matching and flat gain, but the noise figure deteriorates due to additional resistive element and power dissipation increases. The common-gate input characteristic depends on the transistors geometry and the inductance in the source circuit. These parameters can be set in such a way that the circuit provides wideband input matching [18]. In [18] an LNA is designed applying the RC feedback to- pology, employing a gain enhancement technique and containing only one inductor. A frequency selective broadband LNA is presented in [19], where a topology of either a global or local feedback or the combination of both is investigated. In [20] a two-stage common- source (CS) LNA that utilizes forward-body-bias (FBB) technique in n-type MOS devices is presented. The au- thors in [21] also employed the FBB technique along with the current-reuse scheme and active shunt-feed- back towards their goal of ultra-low power consump- tion. In [22] an UWB LNA with operating frequency range from 50 MHz to 10 GHz with resistive feedback and π-matching network is presented. From the papers mentioned in the previous paragraph and keeping in mind the dates of those publications (2014-2017), we can conclude that UWB is an active re- search area interesting from the design of RF IC point of view. Designers [18-22] are utilizing different tech- nologies, topologies, techniques and approaches while trying to optimize performance over a large number of, often opposing each other, requirements. Those requirements differ from case to case, thus no general way of comparing LNA performance is possible. There- fore, no figure of merit can be used on its own, rather the whole design must be considered within the con- text of specific application. 4 Low Noise Amplifier Design In the following subsections, we present a UWB LNA, designed using the Cadence Design Systems® tool- chain and fabricated using the standard 130 nm CMOS process. First the topology choice is presented, where each stage is thoroughly discussed. Then physical de- sign details are presented, describing the circuit lay- out. Finally, simulation results after parasitic extraction (postlayout simulation) are presented. 4.1 Topology Considerations UWB circuits and systems must deal with numerous trade-offs [11]. For example, to design a highly linear A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 ... ... a) ... ... b) ... ... c) ... d) M1 M2 M1 M1M1 ... R1 L11 L21 L22L12 L1 L2 C1 C2 L1 R1 VBIAS 63 amplifier, large values of transistor overdrive voltages (VOD=VGSVT) are required; which causes the increase in drain currents and, consequently, in power consump- tion. This means that high linearity and low power con- sumption are opposed design goals. Analogous to this conclusion, when other LNA design goals mentioned in sections 1 and 2 are considered, similar facts can be derived; i.e. it is a matter of trade-off between figures of merit how well the circuit will perform overall. In that context, the most interesting topologies out of those discussed in section 3 are resistive-feedback (Figure 2c) and common-gate (Figure 2d). Both of them satisfy input matching across a wide frequency range, and offer a compromise between the numerous de- mands. For high gain conditions, the noise and gain performance of a resistive-feedback and of a common- gate is virtually the same. A key difference arises at high frequencies, where the load capacitance CL has a very significant impact on the input impedance in the case of the resistive-feedback amplifier, while this is not so in the common-gate case [5]. Derivations thoroughly presented and discussed in literature [17], show that the source impedance of a common-source topology yielding minimum noise factor must be inductive in nature. As the input impedance of a MOSFET in such configuration is capacitive, providing a good match to a 50 Ω source is a difficult task. Nevertheless, for an LNA, presenting a resistive impedance of this value to the external circuits and sub-circuits is a critical require- ment – therefore, the LNA topology and the elements it comprises of, must be selected accordingly. The sim- plest approach would be to connect a 50 Ω resistor between the gate and source terminals of a common- source connected MOSFET. However, the resistor adds thermal noise of its own and, as it creates a voltage di- vider, it attenuates the signal by a factor of two. It turns out, as it is further explained in subsection 4.2, that a common-gate topology realizes resistive input imped- ance, as shown in equation (9). However, common-gate amplifier topology cannot typically be used directly in UWB front-ends, as a con- sequence of its inadequate noise performance over the frequency range of interest, as well as potential failure to meet gain-bandwidth product requirement. This sin- gle-transistor topology thus needs to be enhanced to achieve the desired noise, gain and bandwidth specifi- cations [5]. For this reason, the second stage consisting of a common-source amplifier employing the shunt- peaking technique [16] is cascaded to the first stage. The proposed solution schematic is shown in Figure 3. This LNA circuit can be divided in three sub-circuits: common-gate (first stage), bandpass filter and com- mon-source (second stage). All transistors operate in the strong inversion region. Two circuit nodes directly controlling transistor bias- ing are accessible from outside through the pads, thus making the LNA operating region adaptable even after fabrication. These connections are omitted from Figure 3 for simplicity, but allow fine tuning of M1 and M2 op- erating points through VB1 and VB2 values setup. This is done with the idea to enable compensation of the po- tential process variations. Finally, substrate of each transistor is grounded with a high resistivity resistor (body floating). In this way sub- strate current noise referred to drain node is reduced, resulting in overall NF reduction of about 0.5 dB [17]. Figure 3: Proposed circuit (biasing, substrate contacts, pads and body floating resistors omitted) 4.2 Common-gate Stage The first stage consists of a transistor M1 in a common- gate configuration with a coil LS1 in the source and an RLC resonant circuit in the drain. In its first approxima- tion, its input impedance is: 1 ,in m Z g ≈ (9) where gm is transistor’s transconductance. This rela- tion is quite straightforward and, thus, M1, along with inductor LS1, is used to set input impedance towards the goal of 50 Ω, i.e. input return loss (S11) below -10 dB. The source input matching is needed in order to avoid signal reflections at the input of the LNA or the altera- tions of the characteristics of the RF filter preceding the LNA, such as pass-band ripple and stop-band attenua- tion [7]. A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 VDD VB1 vIN vOUT C L CD1 LD1 LD2 LS1 M1 M2 VB2 64 Voltage gain of the common-gate stage is given as [1]: 1 1 , 2 1 out m out in out D V g r V r R + =   +   (10) where rout represents the M1 output resistance and RD1 is the resistance in the drain of M1. The resistor RD1 is not shown in Figure 3, as it is actually composed of resistive parasitics contained in LD1, CD1 and interconnects. This common-gate configuration also acts as a tuned amplifier; namely, the resonant circuit consisting of LD1, CD1 and RD1 enables this subcircuit to amplify the signal within the band around the resonant frequency. The resonant circuit is not decoupled from the rest of the amplifier, so in all considerations other elements also must be included. Concretely, it is influenced by the M1 parasitic output impedance and the bandpass filter in- put impedance. Including the additional parasitics, the resonant circuit is tuned to 5.8 GHz. For a MOSFET transistor operating in saturation, the most dominant noise source is channel thermal noise. Power spectral density of a saturated MOSFET is calcu- lated in the following way [16]: 2 04 Δnd di kT g fγ= ⋅ ⋅ ⋅ ⋅ (11) is assumed the dominant source of noise, where gd0 is the drain-source conductance at zero VDS and γ is the correction factor named excess-noise factor. For a sub- micron MOSFET, we assume: gd0/gm > 1 and γ = 2/3 for a long-channel saturated transistor in strong inversion. Value of γ can be larger, γ > 1, in the case of a short- channel transistor, as it strongly depends on the chan- nel length modulation effect [16]. The noise factor of a common-gate device at low frequencies, when the input impedance is matched to the source, is given by: 1 4 . S L RF R γ= + + (12) which, indirectly, yields NF, also. Thus, as the gain is in- creased by increasing the value of RL, the noise factor similarly asymptotically assumes a value of 1+γ. This re- sult also assumes that the common-gate amplifier uses an RF choke, which in this case is LS1. The inductor is necessary, as the usage of a resistor or a current source instead would increase the noise factor [15, 16]. There- fore, the main purpose of LS1 is the reduction of noise factor. To achieve this, its value must be carefully select- ed. This is done first by preliminary calculations, based on the fact that this inductor, to enable noise figure reduction, must resonate with the total capacitance in its proximity, which includes: capacitance of the input signal pad (Cpad), the parasitic of the transistor M1 (CSB1 and CGS1) and its own parasitic capacitance (CLS1). A first order approximation yields: 1 1 1 1 2 res S pad SB GS LS fL C C C C π⋅ ⋅ = + + + (13) where fres is the frequency at which the resonance oc- curs, in this case being equal to the frequency the RLC circuit in the drain of M1 is tuned to (5.8 GHz). As these capacitive parasitics cannot be known a priori, the calculation according to equation (13) is only the first step; namely, the final value of LS1 is yielded through simulations in several iterations. 4.3 Common-source Stage The cascaded second stage is a common-source cir- cuit consisting of the transistor M2 loaded by the coil LD2. Just as in the previous stage, a resonant circuit was used to set the working frequency range, in this stage it is done by a single coil in the drain circuit. This ap- proach is known as shunt-peaking technique [16]. At higher frequencies, as the impedance of the induc- tance increases, that of the load capacitance decreases. By properly controlling the relative value of the load inductance in relation to the parasitic capacitance, a flat gain can be achieved over a wider bandwidth. In fact, a bandwidth extension of as much as 70% can be achieved by use of a single inductor, in comparison to a simple shunt RC load. In the case of wideband ampli- fiers, the inductor does not require a high-quality fac- tor, since the bandwidth is supposed to be the widest possible. Besides its influence on the gain characteristic, LD2 di- rectly determines the output return loss, S22. 4.4 Bandpass filter Impedance matching between the amplifying stages is achieved employing the bandpass filter composed of an inductor L and a capacitor C. The capacitance C is also used to decouple the first and the second stage, thus enabling the M2 transistor biasing. Keeping in mind this other purpose of the capacitive element and including the influence of the rest of the circuit, the bandpass filter is tuned to 9.5 GHz. 4.5 Layout In Figure 4 circuit’s floorplan is presented. Elements occupying the largest area are four inductors, twelve A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 65 pads and a large decoupling capacitor formed as a ver- tically interdigitated structure encircling the LNA core, formed in metal layers 1 and 8. During schematic level design, pad influence on LNA performance (especially on Γ(ω)) was taken into ac- count (even though omitted from Figure 3). Groups of three pads on the left and right represent input and output ports in the constellation ground-signal- ground (GSG), where the middle pad is input and out- put, respectively. Top and bottom pad groups are in power-ground-logic form, where “logic” contacts are used as the inputs for transistor biasing control. Both power supply and ground connections are present on two sides (top and bottom) of the design in order to secure equal voltage levels of VDD over the whole die. The transistors M1 and M2 are each implemented as multiple transistors in parallel. Thus, more fingers are available to reduce effective gate resistance [17, 23]. Contrary to analog circuits where components and in- terconnects can be placed in the vicinity of each other, in the case of RF circuits that is not always possible. To ensure inductor operation without crosstalk, they must be safely distanced from other circuit components. The same consideration is applied to interconnects, as their behavior significantly changes at high frequencies (HF). For this reason plenty of empty space can be seen in Figure 4. However, that may not be fabricated as such, because metal density limitations are present in every CMOS technology node [24]. Therefore, these areas are filled with metal islands in order to satisfy the demand for metal density while degrading circuit performance as minimum as possible. The circuit occupies silicon area of 0.89 mm2, whereas the LNA core (LNA design without the pads and the in- terdigitated capacitor) occupies the area of 0.66 mm2. 4.6 Post-layout Simulation Results After the parasitic extraction and prior to fabrication, final scattering parameters and noise figure results are shown in Figure 5. For this nominal case, M1 biasing is set at VB1=570 mV and M2 biasing is done through a cur- rent mirror – the reference branch of which is biased at VB2=1.2 V. The maximum gain, S21, is 15.48 dB, whereas the 3 dB frequencies are at 6.31 and 9.07 GHz. Input matching, measured by S11, is better than -10 dB over the whole range. Output reflection coefficient is some- what higher than -10 dB. However, such values for S22 are acceptable [17]. Minimum value of NF within the frequency range of in- terest is 3.8 dB at 7.10 GHz. Power consumption is 18.41 mW from the supply volt- age of 1.2 V. Figure 5: Post-layout scattering parameter simulation results interdigitated capacitor structure VB1 control VB2 control Figure 4: LNA layout screenshot as designed A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 Figure 6: Post-layout noise figure simulation results 66 Linearity figures of merit of the designed LNA are sum- marized in 0, as defined in section 0. IIP3 is simulated in two cases, hence the designations @ 50 MHz and @ 200 MHz. In the former case, the second signal is a 50 MHz offset relative to the main signal, whereas in the latter case the second signal is a 200 MHz offset relative to the main signal. Table 1: Linearity figures of merit simulation results f [GHz] 6.4 7 7.6 8.2 8.8 IIP3 @ 50 MHz [dBm] 0.38 1.33 3.18 2.38 0.95 IIP3 @ 200 MHz [dBm] 0.92 1.27 3.25 2.59 1.00 P1dB [dBm] - 8.68 - 8.33 - 6.35 - 7.01 - 8.60 5 Characterization The on-chip characterization set-up is shown in Figure 7, consisting of VNA (N5240A from Keysight Technolo- gies ®), RF probe station, RF cables, two GSG probes and two DC PGL probes (all from Cascade Microtech ®). First the influence of the equipment is canceled through VNA calibration process – short, open, load and thru (SOLT) procedure in this case – and then the characteri- zation is performed. Figure 7: Measurement set-up In Figure 8 fabricated circuit microphotograph is shown, while probe contact with pads is secured. Con- trary to Figure 4, in this image metal filings are obvious. In Figure 9-11 characterization results are shown at nominal biasing as given in subsection 4.6, witnessing scattering parameters behavior close to simulated val- ues. A frequency shift of less than 10 % is present in all characteristics. Maximum gain is 12.33 dB, whereas the 3 dB band ranges from 5.74 GHz to 8.14 GHz, as shown in Figure 9. Input matching raises above -10 dB at mid- dle frequencies (around 7.54 GHz) but remains below for the rest of the 3 dB range, which is shown in Figure 10. Output reflection coefficient also deteriorates com- pared to post-layout simulation results shown in Figure 5, but within acceptable limits, as it is shown in Figure 11. Figure 9: S21 characterization results (red) compared against the post-layout simulation results (blue) Figure 10: S11 characterization results (red) compared against the post-layout simulation results (blue) A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 Figure 8: Die microphotograph 67 Figure 11: S22 characterization results (red) compared against the post-layout simulation results (blue) In Figure 12 linearity characterization results are pre- sented, showing a P1dB point at -4.5 dBm of input power. Figure 12: 1-dB compression point (P1db) characteriza- tion results 6 Discussion Characterization results deviate from the postlayout simulation as the frequency shift of 10 % is noticed in Figure 9 when compared to Figure 5. Therefore, 3 dB bandwidth of this circuit is from 5.74 to 8.14 GHz. Fur- thermore, LNA gain is deteriorated by 3 dB, as the maxi- mum in postlayout is 15.48 dB, whereas the character- ization yielded a maximum value of S21 as 12.33 dB. This is the reason why P1dB is somewhat improved (-4.5 dBm) compared to expected results (Table 1); namely, since the gain is smaller, the amplifier will operate in the lin- ear region at higher input signal power. Measured S11 is above -10 dB for a segment of the 3 dB bandwidth, in the vicinity of 7.54 GHz. Finally, S22 also departs from the predicted curve, but it does remain less than -5 dB over the frequency range of interest. Frequency shift to lower frequencies and decrease in S21 are both signs of increased resistive parasitic com- ponents within the circuit interconnects [24]. Current density within a conductor of a circular cross-section is given as [25]: ( ) ( ) ( )( ) ( )( ){ }0 00 Re jIm ,J u J J u J u= + (14) Where ju s= − , 's k r= , k ωµσ′ = , r distance to the conductor axis, J(0) current density along the conduc- tor axis and J0 Bessel’s function of order zero. At high frequencies, equation (14) can be approximated as fol- lows: ( ) x SJ x J e δ − = (15) where JS is current surface density, x distance to the conductor surface and δ penetration depth, given by: 1 , f δ πµσ = (16) where µ and σ represent permeability and conductiv- ity, respectively. Equation (14) is valid only for conductor of circular cross-section, whereas approximation (15) is also valid for conductors of rectangular cross-section. This effect is known as skin effect and it actually means that at low frequencies current flows through the whole cross-sec- tion uniformly; while, as the operating frequency rises, current flow is retreating towards the conductor sur- face. If the skin effect is dominant, current flows almost completely on the surface of the conductor. This further means that the cross-section of the part of the conduc- tor used for current flow reduces as the frequency rises. Reduction of cross-section increases conductor surface resistance, which is directly proportional to f [25], which theoretically justifies the decrease in gain mag- nitude and its shift to the lower frequencies, visible in Figures 9-11. In Figure 13 and Figure 14 the influence of transistor M1 biasing to circuit operation is shown – S11 and S21, respectively. Two resonant pairs of frequencies can be seen in Fig- ure 13. At values of 476, 526 and 576 mV for VB1, one pair of resonant frequencies and at values of 810 mV, 1 and 1.2 V, another pair of resonant frequencies is no- ticed for S11. The reason for such fundamental change in behavior is a consequence of different modes of operation of transistor M1. For values of 476, 526 and A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 68 576 mV it operates in the weak inversion operation re- gion and for values of 810 mV, 1 and 1.2 V it operates in the strong inversion. All parasitic capacitances (ex- cept for gate-substrate capacitance, denoted as Cgb) are zero, whereas they (gate-source and substrate source, for example, denoted as Cgs and Cbs) rise to significant values in the strong inversion saturation [26]. Thus, as gate-source voltage, VB1, of M1 raises and it passes intro strong inversion, the resonant frequencies shift to left. Even though S11 deteriorates above -10 dB at nominal value of VB1 (Figure 9), it Figure 13 it can be seen that VB1 may be used to eliminate this variation, e.g. by setting VB1=476 mV. In Figure 14 it is seen that variation of S21 may not be remedied as easily, but characteristics that are easily influenced to the extent are 3 dB range and S21 variation over that range. During characterization, VB2 was also varied. However, its influence to S22 was negligible, as that parameter is primarily determined by LD2. In Table 2, a summary of the circuit performance pre- sented in this paper is given, along with several other works cited. The purpose of this table in no way is a claim which of the circuits performance is better, since each of them was designed to optimize a different figure of merit; for example, in [21] the main goal was low power consumption, whereas in [22] the authors achieved very high linearity. Therefore, Table 2 is given here in order to point out that the characterization re- sults obtained within this paper are of the same order like the results found in relevant and up-to-date litera- ture. Table 2: This work result summary and comparison to related work This work [2] [3] [4] [5] technology [nm] 130 180 90 130 130 S21MAX [dB] 12.33 10.15 15 14 13.28 3 dB range [GHz] 5.74-8.14 1.1-5 3.5-9.25 0.6-4.2 0.05-10 S11 [dB] < -10 < -10 < -10 < -10 < -10 S22 [dB] < -5 < -10 < -10 < -10 < -10 NFmin [dB] 3.8* 4.05* 2.4 4 3.29 P1dB [dBm] -4.5 -9.5 -17.25 -19.6** 3.6** VDD [V] 1.2 1.8 0.8 0.5 1.2 PDD [mW] 18.41* 28.54 9.6 0.25 31.2 area [mm2] 0.66 0.35 0.56 0.39 0.77 * simulated ** estimated according to equation (5) This work represents the continuation of research pre- sented in [27]. In the next iteration of circuit redesign, electromagnetic (EM) properties [12, 24], such as skin effect, and PVT compensation techniques [28, 29] will be included. 7 Conclusion Successful characterization of a fabricated UWB LNA using a standard 130 nm CMOS technology node is presented in this paper. The characterization results show that the techniques applied during the design phase of the circuit successfully fulfill its task: amplifica- tion over a wide frequency range with low noise factor. To prove this, a comparison with several state-of-the- art LNA designs found in literature is given in Table 2. The designed LNA provides 12.33 dB gain within the upper EU UWB band, its input reflection coefficient be- ing less than -10 dB over the whole range. Minimum noise figure is shown to be 3.8 dB, while the circuit con- sumes 18.41 mW of power from a 1.2 V supply voltage. The amplifier remains linear for the input power levels up to -4.5 dBm and its area on chip is 0.66 mm2. A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 Figure 13: Transistor M1 biasing influence to S11 Figure 14: Transistor M1 biasing influence to S21 69 The in-depth discussions of the design procedure, the figures of merit and, especially, characterization ap- proach provide detailed insight in the steps performed to achieve the obtained results. The characterization results do deviate less than 10 % of the post-layout simulation results, as a consequence of the skin effect; namely, due to the fact that the current is flowing on the surface of the conductor, resistance of the signal line increases proportionally to the square root of the operating frequency. However, techniques to tackle these effects are recognized and will be implemented in future work. 8 Acknowledgement This research is done within the project: SENSEIVER-ITN – Low-cost and energy-efficient LTCC sensor/IR-UWB transceiver solutions for sustainable healthy environ- ment, 2012-2015 a part of the FP7 Marie Curie Initial Training Network funded by the European Commis- sion, contract number 289481 and partly within the project TR32016. 9 References 1. B. Razavi, RF Microelectronics, Prentice Hall, 2011. 2. A. Djugova, J. Radic, M. Videnovic-Misic, B. Goll and H. Zimmermann, “A Compact 3.1-5 GHz RC Feedback Low-Noise Amplifier Employing a Gain Enhancemenet Technique,” Informacije MIDEM, Journal of Microelectronics, Electronic Components and Materials, vol. 44, no. 3, pp. 201-211, 2014. 3. S. Bagga, A. L. Mansano, W. A. Serdijn, J. R. Long, K. Van Hartingsveldt and K. Philips, “A Frequency- Selective Broadband Low-noise Amplifier with Double-Loop Transformer Feedback,” IEEE Trans- actions on Circuits and Systems I: Regular Papers, vol. 61, no. 6, pp. 1883-1891, June 2014. 4. M. Parvizi, K. Allidina and M. N. El-Gamal, “Short Channel Output Conductance Enhancement Through Forward Body Biasing to Realize a 0.5 V 250 up uW 0.6–4.2 GHz Current-Reuse CMOS LNA,” IEEE Journal of Solid-State Circuits, vol. 51, no. 3, pp. 574-586, 2016. 5. Y.-Y. Tey, H. Ramiah, N. M. Noh and U. R. Jaged- heswaran, “A 50 MHz-10 GHz, 3.3 dB NF, +6 dBm IIP3 Resistive Feedback Common Source Amplifi- er for Cognitive Radio Application,” Microelectron- ics Journal, vol. 61, no. 3, pp. 89-94, March 2017. 6. A. Pajkanovic and M. Videnovic-Misic, “An Ultra Wideband, 6-9 GHz, 130 nm CMOS Low Noise Amplifier,” in 21st Telecommunications forum TEL- FOR 2013, Belgrade, 2013. 7. B. Milinkovic, M. Milicevic, D. Simic, G. Stojanovic and R. Djuric, “Low-pass Filter for UWB System with the Circuit for Compensation of Process In- duced On-chip Capacitor Variation,” Informacije MIDEM, Journal of Microelectronics, Electronic Components and Materials, vol. 45, no. 4, pp. 266- 276, 2015. 8. M. Milicevic, B. Milicevic, D. Simic, D. Grujic and L. Saranovac, “Temperature and Process Com- pensated RF Power Detector,” Informacije MIDEM, Journal of Microelectronics, Electronic Components and Materials, vol. 46, no. 1, pp. 24-28, 2016. 9. J. Radic, A. Djugova, L. Nagy and M. Videnovic- Misic, “A Low-Complexity and Energy-Efficient IR-UWB Pulse Generator in 0.18 um Technology,” Informacije MIDEM, Journal of Microelectronics, Electronic Components and Materials, vol. 43, no. 3, pp. 179-184, 2013. 10. B. Razavi, Design of Analog CMOS Integrated Cir- cuits, McGrawHill Education, 2016. 11. P. Allen and D. Holberg, CMOS Analog Circuit De- sign, Oxford University Press, 2011. 12. Niknejad and Ali, Electromagnetics for High- Speed Analog and Digital Communication Cir- cuits, Cambridge, 2007. 13. R. Gharpurey and P. Kinget, “Ultra Wideband: Cir- cuits, Tranceivers and Systems,” in Ultra Wideband: Circuits, Tranceivers and Systems, Springer, 2008, pp. 1-23. 14. A. Hastings, Art of Analog Layout, 2005. 15. T. Zwick, W. Wiesbeck, J. Timmermann and G. Ad- amiuk, Ultra-wideband RF System Engineering, Cambridge: Cambridge University Press, 2013. 16. O. Taheri, A. Maunder and P. Mousavi, “Correlation- Based UWB Radar for Thin Layer Resolution,” IEEE Antennas and Wireless Propagation Letters, vol. 15, pp. 901-904, 2016. 17. A. T. Mobashsher and A. Abbosh, “Performance of Directional and Omnidirectional Antennas in Wideband Head Imaging,” IEEE Antennas and Wireless Propagation Letters, vol. 15, pp. 1618- 1621, 2016. 18. H. Bahrami, S. A. Mirbozorgi, A. T. Nguyen, B. Gos- selin and L. A. Rusch, “System-Level Design of a Full-Duplex Wireless Transceiver for Brain–Ma- chine Interfaces,” IEEE Transactions on Microwave Theory and Techniques, vol. 64, no. 10, pp. 3332- 3341, Oct 2016. 19. S. Voinigescu, High-Frequency Integrated Cir- cuits, Cambridge University Press, 2013. 20. C. F. Liu and L. S. I., “A Broadband Noise-Canceling CMOS LNA for 3.1-10.6 GHz UWB Receivers,” IEEE Journal of Solid-State Circuits, vol. 42, no. 2, pp. 1-16, 2007. 21. T. Lee, The RF CMOS Integrated Circuit Design, Prentice Hall, 1998. A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 70 22. D. M. Pozar, Microwave Engineering, 2nd Ed., John Wiley & Sons Ltd., 1998. 23. J. Li, S. Song, X. Chen, H. Nian and W. Shi, “Design and Implementation of a Novel Directional Cou- pler for UHF RFID Reader,” Electronics Journal, vol. 20, no. 1, pp. 22-26, June 2016. 24. S. Pandey and J. Singh, “A 0.6 V, Low-power and High-gain Ultra-wideband Low-noise Amplifier with Forward-body-bias Technique for Low-volt- age Operations,” IET Microwaves, Antennas and Propagation, vol. 9, no. 8, pp. 728-734, 2015. 25. Y. T. Lo and J. F. Kiang, “Design of Wideband LNAs using parallel-to-series resonant matching net- work between common-gate and common- source stages,” IEEE Transactions on Microwave Theory and Techniques, vol. 59, no. 9, pp. 2285- 2294, 2011. 26. D. Grujic, “Design of Monolithic Microwave Inte- grated Circuits for 60 GHz Band - PhD thesis, in serbian,” University of Belgrade, School of Electri- cal Engineering, Belgrade, 2013. 27. FCC, “First report and order: revision of part 15 of the Commission’s rules regarding ultra-wide- bandtransmission systems,” Et Docket, 98-153, 2002. 28. B. Popovic, Elektromagnetika, Belgrade: Grad- jevinska knjiga, in Serbian, 1985. 29. Y. Tsividis, Operation and Modeling of the MOS Transistor, New York, Oxford: Oxford University Press, 2011. 30. M. C. Schneider and Galup-Montoro, CMOS Ana- log Design using All-Region MOSFET Modeling, Cambridge: Cambridge University Press, 2010. Arrived: 14. 03. 2017 Accepted: 16. 06. 2017 A. Pajkanović et al; Informacije Midem, Vol. 47, No. 2(2017), 59 – 70 71 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 71 – 78 Pb(Mg,Nb)O 3 –PbTiO 3 thick films on metalized low-temperature co-fired ceramic substrates H. Uršič1, A. Benčan1,2, E. Khomyakova1, S. Drnovšek1, I. F. Mercioniu3, K. Makarovič1,2,4, D. Belavič5, C. Schreiner6, R. Ciobanu6, P. Fanjul Bolado7, B. Malič1 1Jožef Stefan Institute, Ljubljana, Slovenia 2Centre of Excellence NAMASTE, Ljubljana, Slovenia 3Institutul National Cercetare-Dezvoltare pentru Fizica Materialelor, Bucharest, Romaina 4Keko Equipment Ltd., Žužemberk, Slovenia 5HIPOT-RR d.o.o., Otočec, Slovenia 6Technical University of Iasi, Iasi, Romania 7DROPSENS, Parque Tecnológico de Asturias - Llanera, Spain Abstract: Compatibility of screen-printed piezoelectric 0.65Pb(Mg1/3Nb2/3)O3–0.35PbTiO3 thick films with metalized low-temperature co-fired (LTCC) ceramic substrates is examined. Such substrates are interesting for micro-electro mechanical systems, for example for piezoelectric sensors and actuators, where functional layers are usually Pb-based piezoelectrics. In this study the special attention is given to the influence of the Au, Ag and Ag/Pd electrode materials coated over the LTCC on the functional properties of the films. The best phase purity, dielectric and piezoelectric properties were obtained in the films on gilded substrates. No secondary phases were observed at the film/Au interface by scanning electron microscope. The piezoelectric coefficient d33 of the films on gilded substrates is equal to 120 pC/N. Keywords: PMN-PT; piezoelectric; thick film; domain structure; low-temperature co-fired ceramic; LTCC Debele plasti Pb(Mg,Nb)O 3 –PbTiO 3 na podlagah iz keramike z nizko temperature žganja Izvleček: V članku poročamo o pripravi piezoelektričnih 0,65Pb(Mg1/3Nb2/3)O3–0,35PbTiO3 debelih plasti na podlagah iz keramike z nizko temperature žganja. Omenjene podlage so uporabne v mikro elektro mehanskih sistemih, kot so piezoelektrični senzorji in aktuatorji, zato je njihova kompatibilnost s funkcijskimi plastmi velikega pomena. Preučevali smo vpliv spodnje elektrode na lastnosti funkcijskih plasti. Pripravili smo vzorce s tremi različnimi elektrodami, z Ag, Ag/Pd in Au. Ugotovili smo, da plasti na različnih elektrodah izkazujejo različno fazno sestavo, dielektrične in piezoelektrične lastnosti. Plasti na Au elektrodah izkazujejo veliko boljše dielektrične in piezoelektrične lastnosti ter vsebujejo manj sekundarnih faz kot plasti na Ag in Ag/Pd elektrodah. Piezoelektrični koeficient d33 plasti na podlagah iz keramike z nizko temperaturo žganja z zlatimi elektrodami je 120 pC/N. Ključne besede: PMN-PT; piezoelektrik; debela plast; domenska struktura; keramika z nizko temperaturo žganja, LTCC * Corresponding Author’s e-mail: hana.ursic@ijs.si 1 Introduction The thick-film technology is attractive for the produc- tion of simple functional structures and also quite com- plex systems. Screen-printing is one of the most widely used thick-film deposition techniques for producing up to a hundred micrometres thick piezoelectric films. Commercially available piezoelectric thick films are mostly based on Pb(Zr,Ti)O3 solid solution [1]. The most commonly used substrate materials for piezoelectric ceramic thick films are silicon and alumina. However, low-temperature co-fired ceramics (LTCC) are promis- ing for the fabrication of three-dimensional ceramic 72 H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 structures or ceramic micro-electro-mechanical sys- tems (c-MEMS) and for this reason considered as desir- able substrates for functional layers. The LTCC are based on glass-ceramic composites or crystallizing glass and densify at relatively low temperature (around 850 °C), which enables the use of low cost conductive materi- als and fast firing profiles [2, 3]. Therefore the LTCC and thick-film technologies enable fast and easy fabrication of electronic devices and systems, which could reduce the cost of devices and shorten the time of fabrication [2, 4]. The LTCC materials are especially interesting for actuator and sensor applications due to lower Young’s modulus (90–110 GPa) than alumina (210–410 GPa), which enables high actuation sensitivity [1]. An alternative to Pb(Zr,Ti)O3-based piezoelectric ma- terial is (1-x)Pb(Mg1/3Nb2/3)O3–xPbTiO3 (PMN–100xPT) solid solution with the morphotropic phase boundary at x = 0.35 [5, 6]. The PMN–35PT thick films prepared on silicon and alumina substrates possesses high dielec- tric permittivity (more than 3000 at 1 kHz and room temperature [7-9]) and high piezoelectric coefficients (d33 of 150 to 200 pC/N [9, 10]), and are therefore prom- ising for sensor and actuator applications [11]. The aim of this work was to examine the compatibility of PMN–35PT piezoelectric thick films with metalized LTCC substrates. Until recently, only the Pb(Zr,Ti)O3- based piezoelectric thick films were processed on LTCC substrates [12-15]. But, the glass phase from LTCC may interact with PbZr0.53Ti0.47O3 (PZT) thick film and PbO from the thick film may diffuse into the LTCC during annealing [12, 13], which leads to changes in the thick film’s phase composition and consequently functional properties. Hence the processing of piezoceramic films on LTCC materials is still challenging. In order to pre- vent the film-substrate interactions different solutions were reported, such as interposing of the barrier layer [12] or use of a very dense Au bottom electrode [15]. In this work the PMN–35PT thick films were prepared on metalized LTCC substrates with the interposed 15 mm thick PZT barrier layer to minimize the film-substrate interactions. 2 Experimental methods For the synthesis of the PMN–35PT powder, PbO (99.9 %, Aldrich), MgO (98 %, Aldrich), TiO2 (99.8 %, Alfa Ae- sar) and Nb2O5 (99.9 %, Aldrich) were used. A mixture of these oxides in the molar ratio corresponding to the PMN–35PT stoichiometry with an excess of 2 mol% PbO was high-energy milled for 72 h at 300 rpm in a Retsch Model PM 400 planetary mill, and additionally in an attritor mill for 4 h at 800 rpm in isopropanol. The powder was then heated at 700 °C for 1 h. The size of the particles after heating was determined by a light- scattering technique using a Microtrac S3500 Series Particle Size Analyzer instrument. The results were de- rived from the area particle size distribution. The most common parameter used is the median particle size d50, where the area of all particles smaller than d50 ac- counts for 50 % of the area of all the particles. The pow- der morphology was analysed using the field-emission scanning electron microscopy (FE-SEM) JSM 7600F (Jeol, Tokyo, Japan). The powders were dispersed in ac- etone under ultrasound, and a few drops were spread on highly oriented pyrolytic graphite substrates. The PMN–35PT thick-film paste was prepared from a mix- ture of the PMN–35PT powder and an organic vehicle consisting of alpha-terpineol, [2-(2-butoxy-ethoxy)- ethyl]-acetate and ethyl cellulose in the ratio 60/25/15. For more details on PMN–35PT powder and thick-film paste processing and characterization see [7, 16]. The LTCC substrates were prepared by laminating three layers of the LTCC tape (DuPont 951) at 50 °C and 20 MPa. The laminated tapes were than heated at 450 °C for 1 h to burn-out the organic binder and densified at 875 °C for 15 min. In order to minimize the chemi- cal interactions between the PMN–35PT thick film and the LTCC substrate, a PZT barrier layer was interposed between substrate and bottom electrode as suggest- ed in [7, 17]. The PZT thick-film paste was printed on substrate and sintered 870 °C for 1 h. For more details regarding PZT powder and thick-film paste processing see [7]. The PZT/LTCC structure is further-on referred as the substrate. Thick-film silver (Ag ESL 9912MM), silver/ palladium (80Ag/20Pd FERRO EL44-001) and gold (Au ESL 8884G) conductors were printed on the barrier lay- er/substrate structures and fired at 850 °C for 1 h. The thicknesses of the sintered electrodes were 50, 25 and 25 mm, respectively. The PMN–35PT paste was printed five times on the electrode substrate with intermediate drying at 150 °C after each printing step. The samples were pressed at 50 MPa, heated at 500 °C for 1 h to decompose the or- ganic vehicle [16], sintered at 850 °C for 2 h in covered alumina crucibles in the presence of PbZrO3 packing powder with an excess of 2 mol% PbO and then cooled to room temperature with a rate of 2 °C/min. The X-ray diffraction (XRD) patterns of the films were recorded using a PANalytical X’Pert PRO MPD (X’Pert PRO MPD, PANalytical, Almeo, The Netherlands) diffrac- tometer with Cu–Kα1 radiation. The XRD patterns were recorded in the 2q region from 10° to 70° using a detec- tor with a capture angle of 2.122°. The exposure time for each step was 100 s and the interval between the obtained data points was 0.034°. 73 For microstructural analysis the samples were mounted in epoxy, ground and polished using standard metal- lographic techniques. The microstructures of the sam- ples were investigated with a FE-SEM JSM 7600F (Jeol, Tokyo, Japan) equipped with an Inca Energy Detector. The energy-dispersive X-ray spectroscopy (EDXS) anal- yses were performed at 15 keV. The average grain size (GS) was estimated from the SEM images obtained by backscattered electrons (BE) of the film’s surface. For the stereological analysis more than 400 grains per sample were measured with the Image Tool Software. The GS is expressed as the Feret’s diameter. The topography and piezoresponse images were re- corded with an atomic force microscope (AFM; Asylum Research, Molecular Force Probe 3D, Santa Barbara, CA, USA) equipped with a piezoresponse force mode (PFM). A tetrahedral Si tip coated with Ti/Ir (Asyelec-01, Atomic- Force F&E GmbH, Mannheim, Germany) with the curva- ture diameter of 30 nm ± 10 nm was applied for scanning the sample surface. The out-of-plane amplitude PFM im- ages were measured in the Single mode at ac amplitude signal of 20 V and frequency of ~300 kHz. The local am- plitude and phase hysteresis responses were measured in the Dual AC Resonance Tracking Switching Spectros- copy (DART-SS) mode with the waveform parameters: increasing step signal with maximum amplitude of 60 V and frequency 0.1 Hz; overlapping sinusoidal signal of amplitude 5 V and frequency 20 Hz, off-loop mode. For dielectric and piezoelectric measurements, Cr/Au electrodes with a 1.5-mm diameter were deposited on the top surface of the films using RF-magnetron sputtering (5 Pascal). The dielectric permittivity (ε`) and dielectric losses (tan δ) versus temperature were measured with a HP 4284 A Precision LCR Meter by ac amplitude of 1 V and frequencies of 1, 10, and 100 kHz during cooling from 300 °C to 25 °C. The films were poled at 160 °C with a dc electric field from 2.5 to 7.5 kV/mm for 5 min and field-cooled to 25 °C. After pol- ing the samples were aged for 24 h. The piezoelectric constants d33 were measured by Berlincourt piezome- ter (Take Control PM10, Birmingham, UK) at alternating stress frequency of 100 Hz. 3 Results and discussion 3.1 Characterisation of PMN–35PT powder The particle size distribution of the PMN–35PT powder used for preparation of the thick-film paste is shown in Figure 1(a). The d50 was 0.32 mm. Around 10 % of the particles are larger than 1 mm, while the majority of them has a sub-micrometre size. Such distribution was confirmed by the FE-SEM analysis, where both fine sub- micrometre and coarse particles were observed (Figure 1(b)). The XRD pattern of PMN–35PT mechanochemi- cally synthesized powder after calcination at 700 °C is shown in Figure 1(c). All reflections correspond to the perovskite phase. Figure 1: (a) The area particle size distribution, (b) FE- SEM micrograph and (c) XRD pattern of the PMN–35PT powder after heating at 700 °C for 1 h. A coarse par- ticle and smaller, submicron sized particles are marked by an arrow and a circle in panel (b), respectively. The indexed peaks of the pseudo-cubic perovskite phase (JCPDS 81-0861) are shown in brackets in panel (c). H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 74 3.2 The influence of the bottom electrode material on the phase composition and functional properties of PMN–35PT films In order to study the influence of the bottom electrode layer on the properties of functional thick-film struc- tures three different electrode materials were used; Ag, Ag/Pd and Au. Corresponding XRD patterns of ~60 mm thick PMN–35PT films on the metalized substrates are shown in Figure 2. Figure 2: (a) XRD patterns of PMN–35PT thick films on Ag-, Ag/Pd- and Au-metalized substrates. The indexed peaks of the pseudo-cubic perovskite phase are shown in brackets (JCPDS 81-0861). (b) An enlarged 2θ-region from 23 to 65°. The peaks corresponding to PbO (JCPDS 78-1664) and pyrochlore Pb1.83Nb1.71Mg0.29O6.39 (JCPDS 33-0769) are marked by o and x, respectively. The re- flection, which cannot be clearly identified by crystallo- graphic cards, is marked by a question mark in Fig. (b). In addition to the perovskite reflections, also low-in- tensity PbO and pyrochlore reflections are observed in XRD patterns of the films on Ag- and Ag/Pd-metal- ized substrates. In the latter, also one reflection exists, which cannot be clearly identified by crystallographic cards in JCPDS base (marked by a question mark in Fig- ure 2(b)). Note that the penetration depth of X-rays in PMN−35PT thick films is around 10 μm [10] and there- fore the results show only the composition of the up- per part of the thick films. On the other hand in the XRD pattern of PMN–35PT thick film on Au-metalized substrate the perovskite reflections and only a trace of two reflections corresponding to PbO phase can be observed. These low-intensity diffraction peaks are most probably related to the initial excess of PbO in the starting powder mixture (see Experimental methods). The secondary phases determined from the XRD pat- terns together with the dielectric properties of PMN– 35PT thick films on metalized substrates are collected in Table 1. Table 1: The secondary phases (SP) determined from the XRD patterns, the room-temperature dielectric and piezoelectric properties of PMN–35PT thick films on metalized substrates. electrode SP ε` at 10 kHz tan δ at 10 kHz d33 at 5.5 kV/mm Ag PbO, Py 250 /* / Ag/Pd PbO, Py, UP 300 0.04 / Au traces of PbO 1050 0.04 120 /not possible to pole /*not possible to measure at 10 kHz, tan δ is 0.06 at 100 kHz UP-unidentified phase, Py-pyrochlore The dielectric permittivity of the films on Au-metal- ized substrates is 1050 at 10 kHz, which is more than three times higher than in the films on Ag- and Ag/ Pd-metalized substrates. The tan δ of films on Au- and Ag/Pd-metalized substrates are in the same range. For films on Ag-metalized substrates the tan δ at 10 kHz was not possible to measure. Low values of ε` (i.e. 235) were previously reported also for PZT films on Ag/ Pd-metalized LTCC (with no barrier layer between the LTCC and the electrode layer), which was attributed to intense interactions between the functional film and the substrate [14]. Furthermore, we were able to pole only the films on Au-metalized substrates (Table 1). The measured d33 coefficient of such films is 120 pC/N. Due to the superior phase purity, dielectric properties and ability of poling, we further analysed only the PMN– 35PT films on Au-metalized (gilded) substrate. The mi- crostructural, dielectric properties as a function of tem- perature and piezoelectric properties of these films are further discussed below. 3.3 The PMN–35PT thick films on gilded substrates The photo of a PMN–35PT thick film on the gilded sub- strate is shown in Figure 3(a). The thick film surface is crack-free on cm scale. The BE-SEM images of the sur- face and polished cross-section of the film are shown H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 75 in Figure 3(b) and (c), respectively. Fig. 3(b) reveals a porous, but uniform microstructure. The estimated av- erage grain size is ~0.8 μm. Some PbO, seen as a white inclusion in Figure 3(b), is also present in the film. The results are in accordance with the XRD analysis, where in addition to the perovskite phase, a trace amount of PbO phase was also detected (see Fig. 2(b)). As seen in Figure 3(c) the thicknesses of the PZT barrier layer, Au and PMN–35PT film are ~15, ~25 and ~60 mm, respec- tively. The thickness of the PMN–35PT film is quite ho- mogeneous across the substrate. No secondary phases are observed at the PMN–35PT/Au interface (inset in Figure 3(c)). Below the PZT layer into the LTCC sub- strate, a ~50 mm thick, brighter layer is formed (marked with an arrow). According to the EDXS analysis, this layer is PbO-rich and can be attributed to diffusion of PbO from PZT into the LTCC. Such layer was previ- ously reported also in PZT/Au/LTCC structures, where the PbO diffused from the PZT functional layer into the substrate [13]. In Figure 4 the ε` and tan δ as a function of tempera- ture are shown for the PMN–35PT thick film on gilded substrate. At 30 °C and 1 kHz the ε` and tan δ are 1150 and 0.06, respectively. The temperature of ε`max (∼4800) at 1 kHz is 166 °C ± 1 °C, which is in agreement with the previously published temperature for PMN–35PT bulk ceramics [18, 19]. The ε` at room temperature is approxi- mately three times lower than the one of the films on platinized alumina [7, 8] and approximately four times lower than PMN–35PT bulk ceramics [20], which could be attributed to the higher porosity and smaller average grain size of studied films on gilded LTCC substrates. On the other hand, the prepared films exhibit at least two times higher ε` at room temperature than previously published PZT films on LTCC substrates [12-14]. Figure 4: (a) ε’ and (b) tan δ of the PMN–35PT film on gilded substrate as a function of temperature at 1, 10 and 100 kHz. Figure 3: (a) The photo and BE-SEM images of (b) thick film surface and (c) polished cross-section of PMN– 35PT thick film on gilded substrate. The thick arrow in panel (b) marks the PbO phase. In panel (c) the exten- sion of a Pb rich layer in the LTCC is marked by an ar- row and the inset shows an enlarged view of the PMN– 35PT/Au interface. H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 76 Further the local piezoelectric response and ferro- electric domain configuration were investigated. The topography and out-of-plane PFM amplitude images are shown in Figure 5 (a) and (b), respectively. In the topographical image, the grain boundaries can be clearly identified, while in the amplitude image the grain boundaries are visible as dark non-active bound- aries. The enhanced local piezoelectric activity is evi- dent as brighter areas within the grains. An example of such region is marked by no. 1 in Figure 5 (b). The most frequently observed domains in the film are ir- regularly shaped domains found also in Pb(Sc0.5Nb0.5)O3 [21]. The size of the domains varies from a few hundred nanometres to a micrometre. Examples of such irreg- ularly shaped domains are marked by thick arrows in Figure 5 (b). In addition rear lamellar-like domains can be found (marked by a thin arrow). Irregularly shaped and lamellar domains were previously observed also in PMN–35PT films on alumina substrates [22]. The local PFM amplitude and phase hysteresis loops are shown in Figure 5 (c) confirming a typical hysteretic response of the ferroelectric material and indicating the local do- main switching. The PMN–35PT thick films were poled with a dc elec- tric field of 2.5–7.5 kV/mm as described in Experimen- tal section. The d33 values are collected in Table 2. For the films poled at 2.5 kV/mm, the d33 was 65 pC/N, but it increased with the increasing poling field until 5.5 kV/mm. The highest d33 of 120 pC/N was measured for the samples poled at electric field amplitudes be- tween 5.5 and 7.5 kV/mm. The d33 coefficient is only 20 % lower than the one reported for PMN–35PT films on platinized alumina, i.e., ~150 pC/N for similar film thick- ness [10]. We note that the reduction of d33 in PZT thick films deposited on LTCC or on alumina was much more pronounced than in our PMN-PT films; i.e., 75 pC/N ver- sus 125 pC/N, which is about 40 %. Such strong dete- rioration of the piezoelectric response was attributed to the interaction between the PZT film and the LTCC substrate [13]. It is also important to mentioned that the piezoelectric coefficient of PMN–35PT bulk ceram- ics prepared from the mechanochemically synthesized powder, but sintered at 1200 °C is ~640 pC/N (at similar poling conditions; electric field of 4.5 kV/mm) [20]. Table 2: The piezoelectric coefficient d33 measured for the PMN–35PT films poled at different electric field am- plitudes. Epoling (kV/mm) d33 (pC/N) 2.5 65 3.5 85 4.5 100 5.5 120 6.5 120 7.5 120 Figure 5: (a) Topography and (b) out-of-plane PFM amplitude image of the PMN–35PT thick film on gild- ed substrate. (c) PFM amplitude and phase hysteresis loops obtained from the areas marked by no. 1, 2 in the panel (b). H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 77 The degradation of dielectric and piezoelectric prop- erties of the PMN–35PT films on metalized LTCC sub- strates in comparison to the properties of the films on metalized alumina substrates could be related to the higher porosity of the films deposited on the former substrates. Further work is thus needed to improve the densification of the films on LTCC substrates. 4 Summary and conclusions The compatibility of piezoelectric PMN–35PT thick-films with metalized LTCC substrates (DuPont 951) was stud- ied. Three different bottom electrodes were used; Ag, Ag/Pd and Au. The XRD analysis revealed the presence of a large amount of secondary phases in the films pre- pared on Ag- and Ag/Pd-metalized substrates, while in the film deposited on the Au-metalized substrate only a trace of PbO phase was detected. Furthermore, about three times higher room temperature dielectric permit- tivity (1050 at 10 kHz) was measured in the films on Au- in comparison to the ones on Ag- and Ag/Pd-metalized substrates. The films on Ag- and Ag/Pd-metalized sub- strates could not be poled, but the ones on gilded sub- strates exhibit the piezoelectric coefficient d33 as high as 120 pC/N. Irregularly-shaped and lamellar-like domains with the size from a few hundred nanometres to micro- metres were observed and the domain switching was confirmed by piezo-response force microscopy. In conclusion, we succeeded to prepare ~60 mm thick PMN–35PT films on gilded substrates, where no film- substrate interactions were observed by SEM. The films exhibit high piezoelectric properties, much higher than the ones previously published for lead-based thick films on LTCCs. However, further work is needed to improve the densification of these films aiming to further enhance their functional properties. 5 Acknowledgements This work was funded by M-ERA.NET project “Integrated sensors with microfluidic features using LTCC Technology” INTCERSEN (PR-06211) and the Slovenian Research Agency (research core funding No. P2-0105). Techni- cal support by Mitja Jerlah, Jena Cilenšek, Brigita Kmet, Aneja Tuljak and Mateo Markov (Erasmus+ program) is gratefully acknowledged. 6 References 1. Zarnik M S, Ursic H, Kosec M. Recent progress in thick-film piezoelectric actuators prepared by screen-printing, in: Piezoelectric Actuators (open access), J. E. Segel, Ed., Nova Science Publishers Inc., 2011. https://www.novapublishers.com/cat- alog/product_info.php?products_id=30015 2. Imanaka Y. Multilayered low temperature co- fired ceramics (LTCC) technology. New York, USA, Springer, 2005. http://link.springer.com/ book/10.1007%2Fb101196 3. Makarovic K, Meden A, Hrovat M, Holc J, Bencan A, Dakskobler A, Kosec M. The effect of process- ing conditions on the properties of LTCC mate- rial, J. Am. Ceram. Soc., 2012; 95: 760–767. http:// onlinelibrary.wiley.com/doi/10.1111/j.1551- 2916.2011.05027.x/abstract 4. Belavic D, Hrovat M, Zarnik MS, Holc J, Kosec M. An investigation of thick PZT films for sensor ap- plications: A case study with different electrode materials, J. Electroceram., 2009; 23: 1–5. http:// link.springer.com/article/10.1007/s10832-008- 9495-1 5. Singh A K, Pandey D. Evidence for MB and MC phases in the morphotropic phase boundary region of (1-x)[Pb(Mg1/3Nb2/3)O3]–xPbTiO3: A riet- veld study, Phys. Rev. B, 2003; 67, 064102 1–12. https://journals.aps.org/prb/pdf/10.1103/Phys- RevB.67.064102 6. Handbook of Advanced Dielectric, Piezoelectric and Ferroelectric Materials, Z.-G. Ye, Ed. Cam- bridge, England, Woodhead Publishing Ltd, 1st ed., 2008. http://www.sciencedirect.com/sci- ence/book/9781845691868 7. Ursic H, Hrovat M, Holc J, Tellier J, Drnovsek S, Guib- lin N, Dkhil B, Kosec M. Influence of the substrate on the phase composition and electrical proper- ties of 0.65PMN–0.35PT thick films, J. Eur. Ceram. Soc., 2010; 30: 2081–2092. http://www.sciencedi- rect.com/science/article/pii/S0955221910001615 8. Gentil S, Damjanovic D, Setter N. Development of relaxor ferroelectric materials for screen-printing on alumina and silicon substrates. J. Eur. Ceram. Soc., 2005; 25: 2125–2128. http://www.sciencedi- rect.com/science/article/pii/S095522190500169X 9. Gentil S, Damjanovic D, Setter N. Pb(Mg1/3Nb2/3)O3 and (1-x)Pb(Mg1/3Nb2/3)O3–xPbTiO3 relaxor fer- roelectric thick films: processing and electrical characterization. J. Electroceram., 2004; 12: 151– 161. http://link.springer.com/article/10.1023/B:- JECR.0000037720.39443.e3 10. Ursic H, Zarnik, MS, Tellier J, Hrovat M, Holc J, Kosec M. The influence of thermal stresses on the phase com- position of 0.65Pb(Mg1/3Nb2/3)O3–0.35PbTiO3 thick films, J. Appl. Phys., 2011; 109: 014101 1–5. http:// aip.scitation.org/doi/full/10.1063/1.3526971 11. Ursic H, Zarnik MS, Kosec M. Pb(Mg1/3Nb2/3)O3– PbTiO3 (PMN–PT) material for actuator applica- tions, Smart Materials Research (open access), H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 78 2011; 2011: 452901 1–6. https://www.hindawi. com/journals/smr/2011/452901/ 12. Hrovat M, Holc J, Drnovsek S, Belavic D, Cilensek J, Kosec M. PZT thick films on LTCC substrates with an interposed alumina barrier layer, J. Eur. Ceram. Soc., 2006; 26: 897–900. http://www.sciencedi- rect.com/science/article/pii/S0955221905000634 13. Ursic H, Hrovat M, Belavic D, Cilensek J, Drnovsek S, Holc J, Zarnik MS, Kosec M. Microstructural and electrical characterisation of PZT thick films on LTCC substrates, J. Eur. Ceram. Soc., 2008; 28: 1839–1844. http://www.sciencedirect.com/sci- ence/article/pii/S0955221908000253 14. Golonka L J, Buczek M, Hrovat M, Belavic D, Dziedzic A, Roguszczak H, Zawada T, Properties of PZT thick films made on LTCC, Microelectronics Int., 2005; 22: 13–16. http://www.emeraldinsight. com/doi/abs/10.1108/13565360510592171 15. Gebhardt S, Seffner L, Schlenkrich F, Schonecker A. PZT thick films for sensor and actuator appli- cations, J. Eur. Ceram. Soc., 2007; 27: 4177–4180. http://www.sciencedirect.com/science/article/ pii/S095522190700163X 16. Ursic H, Tchernychova E, Bencan A, Jouin J, Holc J, Drnovsek S, Hrovat M, Malic B. The influence of the platinum substrate roughness on the fer- roelectric properties of 0.65Pb(Mg1/3Nb2/3)O3– 0.35PbTiO3 thick films, Informacije MIDEM, Jour- nal of Microelectronics, Electronic Components and Materials (open access), 2014; 44: 12–18, http:// www.midem-drustvo.si/Journal%20papers/ MIDEM_44(2014)1p12.pdf 17. Holc J, Hrovat M, Kosec M. Interactions between alumina and PLZT thick films. Mater. Res. Bull., 1999; 34: 2271–8. http://www.sciencedirect.com/ science/article/pii/S0025540899002275 18. Alguero M, Ricote J, Jimenez R., Ramos P, Carreaud J, Dkhil B, Kiat J M, Holc J, Kosec M. Size effect in morphotropic phase boundary Pb (Mg1/3Nb2/3)O3–PbTiO3, Appl. Phys. Lett., 2007; 91: 112905 1–3. http://aip.scitation.org/doi/ abs/10.1063/1.2778471 19. Leite E R, Scotch A M, Khan A, Li T, Chan H M, Harmer M P, Liu S F and Park S E. Chemical he- terogeneity in PMN–35PT ceramics and effects on dielectric and piezoelectric properties, J. Am. Ceram. Soc., 2002; 85: 3018–3024. http://online- library.wiley.com/doi/10.1111/j.1151-2916.2002. tb00572.x/abstract 20. Ursic H, Tellier J, Hrovat M, Holc J, Drnovsek S, Bobnar V, Alguero M, Kosec M. The Effect of Po- ling on the Properties of 0.65Pb(Mg1/3Nb2/3)O3– –0.35PbTiO3 Ceramics, Jpn. J. Appl. Phys., 2011; 50: 035801-6. 21. Ursic H, Drnovsek S, Malic B. Complex doma- in structure in polycrystalline Pb(Sc0.5Nb0.5)O3, J. Phys .D: Appl. Phys., 2016; 49: 115304 1–4. http://iopscience.iop.org/article/10.1088/0022 -3727/49/11/115304 22. Ursic H, Ricote J, Amorin H, Holc J, Algue- ro M. Ferroelectric domain configurations in 0.65Pb(Mg1/3Nb2/3)O3–0.35PbTiO3 thick films determined by piezoresponse force mi- croscopy, J. Phys. D: Appl. Phys., 2012; 45: 265402 1–11. http://iopscience.iop.org/arti- cle/10.1088/0022-3727/45/26/265402 Arrived: 17. 04. 2017 Accepted: 13. 07. 2017 H. Uršič et al; Informacije Midem, Vol. 47, No. 2(2017), 71 – 78 79 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 79 – 89 Considerations about the use of the Moon in X-band antenna G/T measurements Darko Šekuljica1, Stefano Badessi1, Massimiliano Ferrante2, Matjaž Vidmar3 1European Space Agency 2Vitrociset S.p.A. 3University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia Abstract: The most common G/T quality factor measurement methods applicable for operational X-band (8 GHz - 12 GHz) parabolic antennas with a reflector aperture diameter between 7 and 13m are reviewed. Analyses have shown that the most adequate astronomical source for the G/T measurement of the antennas with the size of interest is the Moon. Since the Moon’s angular diameter is wider than the antenna’s Half Power Beam Width (HPBW), a thorough analysis of the extended source size correction factor is performed. As a result, an efficient correction factor approximation method which is more accurate in case of efficiency losses is identified, and a best-fit estimation method for the HPBW verification is introduced. The proposed G/T quality factor measurement procedure is verified on an operational X-band 11 meter Cassegrain antenna, used for Low Earth Orbit (LEO) satellite data acquisition. The results obtained with the proposed method, have provided accurate G/T factor es- timations that are consistent and in line with the expectations. As further confirmation, the validity of the measurement method results is also verified against a G/T measurement done with the Cassiopeia A radio star as an RF source. Keywords: X-band, G/T measurements, Moon, Satellite communications, Extended source size correction factor Ugotovitve o uporabi Lune pri G/T meritvah antenskih sistemov frekvenčnega pasu X. Izvleček: Najbolj pogoste metode za merjenje kakovostnega faktorja G/T pri operativnih paraboličnih antenah frekvenčnega pasu X (8 GHz – 12 GHz) s premerom zaslonke zrcala med 7 in 13 metrov so bile pregledane v tej študiji. Analiza je pokazala, da je Luna najbolj ustrezen astronomski vir za G/T meritve omenjenih anten. Ker je Lunin zorni kot širši od -3dB širine snopa antene (HPBW), je narejena celovita analiza korekcijskega faktorja za uporabo porazdeljen- ega vira. Na podlagi tega je identificirana najbolj učinkovita metoda približka korekcijskega faktorja za omenjene antene. Uvedena je tudi najbolj prilegajoča metoda ocene HPBW z ciljem preverjanja izmerjenih vrednosti. Predlagani postopek merjenja G/T kakovostnega faktorja je bil preverjen na operativni 11 metrski Cassegrain anteni v frekvenčnem pasu X, ki se uporablja za pridobivanje podatkov s satelitov LEO (nizka Zemljina orbita). Rezultati dobljeni z uporabo omenjene predlagane metode so dali konsistentne in točne ocene G/T faktorja v skladu s pričakovanji. Dodatna potrditev veljavnosti postopka je dana s primerjanjem dobljenih rezultatov in izmerjenega G/T faktorja z uporabo radio zvezde Cassiopeia A kot RF izvora. Ključne besede: G/T, frekvenčni pas X, Luna, satelitske komunikacije, korekcijski faktor razširjene velikosti vira * Corresponding Author’s e-mail: darko.sekuljica@gmail.com 1 Introduction The antenna power Gain over system noise Tempera- ture ratio (G/T), sometimes also labelled as “figure of merit”, can be regarded as a quality factor used to in- dicate the ground station capability to discriminate be- tween signal and noise in a radio communication sys- 80 D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 tem [1]. The higher the ratio, more efficient the ground station is in receiving a weak signal. At the European Space Agency, as a part of the standard requirements for Earth Observation Satellites acquisi- tion service contracts, a minimum G/T ratio is always specified to ensure the service is capable of properly acquiring uncorrupted data. Periodic measurements of the G/T are therefore required to ensure that the an- tennas selected to deliver the service maintain the re- quired performance. Typical antennas, for the purposes of Earth Observa- tion (EO) satellite data acquisition, are receiving X- band with a main reflector aperture in the range of 7 to 13 meters. These ground station antennas perform routinely full elevation and azimuth rotations and ac- celerations which, beside the wind and gravity effects, add to the possibility of antenna distortions. Because of their intense use, the antennas are constant- ly operational with small allotments of time available for performing the G/T measurements. Short pauses between satellite passes and frequent performance verifications require an accurate, efficient and time sav- ing G/T measurement method. Considering the antenna size range of interest, and the corresponding far field region distance, the optimal RF source has to be found in the skies. Direct G/T measure- ment using the Moon as an RF source, for the antenna size range of interest, has been selected and elaborat- ed in this paper, providing a simple, fast and highly ac- curate G/T quality factor estimation method. This paper is organized as follows. In Section II, the analysis of the available celestial sources has been described. Direct G/T calculation procedure, correc- tion factor analysis and concerns from practical point of view are given in Section III. Section IV presents detailed G/T measurement procedure and the quality factor measurement results obtained with an opera- tional 11 meter X-band antenna located at the e-GEOS station in Southern Italy. Finally, Section V draws some conclusions. 2 Analysis of the adequate available sources For the antenna size range of interest in the X-band, the far-field region determined by the Fraunhoffer’s distance, results always more than 2.5km from the an- tenna location. In that case, terrestrial measurements with a distant antenna set as a source are highly im- practical due to the terrain variety, high multi-path and various man-made or natural sources. Therefore, celes- tial sources were considered. As the measurement pro- cedure must be valid and applicable for any antenna location, the selected source has to be well-defined and frequently visible. The most obvious celestial sources are geostationary satellites, and natural sources like Sun, Moon and radio stars. Unfortunately, commercial geostationary satel- lites operating in the X-band, with well-known charac- teristics are uncommon. Hence, geostationary satellites could not have been taken as a suitable RF source. The strongest celestial natural source observed from Earth is the Sun. However, the Sun is subject of a possible and unpredictable variations with time due to Sun storms and Sun bursts [2, 3]. These variations, along with the extended source size of the Sun in respect to the an- tenna’s HPBW, can introduce a significant uncertainty in the final quality factor estimation. Radio star flux densities have been well-defined in [4, 5, 6]. For the antenna size range of interest, 7-13 meters, a radio star can be considered a point-like source. The latter means that using a radio star as an RF source in G/T measurement minimizes the error introduced by using G/T correction factors. However, the brightness of radio stars is much lower than those of the Sun or the Moon. For X-band antennas with reflector diameter smaller than 18m [7], using a direct G/T measurement method with radio stars may introduce significant un- certainties. These uncertainties, and consequently the errors in G/T estimation, are caused by the very small Y- factor, i.e. the ratio between the received noise power while pointing at the radio star, and the received noise power while pointing at the cold sky. The Moon’s radiation can be modelled with the black- body radiation that changes with the lunar phases and the Earth-Moon geometry. Y-factor readings using the Moon as a source are well above 1dB for the antennas of interest. Considered as a black-body, the Moon’s ra- diation can be efficiently approximated as the radiation from a uniform brightness disk, introducing minimal errors [8, 9] in G/T estimation. Uniform brightness disk flux density approximation depends on the Moon’s sol- id angle that changes with Earth-Moon geometry, and on the average brightness temperature of the Moon. For the X-band, it can be approximated using the Ray- leigh-Jeans law: 2 2 0 2 ΩB MoonMoon Moon f k TS c = (1) where SMoon is black-body flux density given in [W m -2 Hz-1], f is the frequency in [Hz], kB is Boltzmann constant 81 given with kB ≅ 1.38 . 10-23 [m2 kg s-2 K-1], c0 [m s -1] is the speed of light, and ΩMoon is the Moon’s solid angle in [sr]. The Moon average brightness temperature MoonT [K], is a function of the frequency, lunar phase and solar mean anomaly. In this paper, the yearly variation in the solar irradiation at the lunar surface due to the Earth’s eccentric orbit has been neglected. Approximation of the Moon’s average brightness temperature has been given by [8, 9]: ( )10 0 1 cosMoon TT T T φ ψ   = − −    (2) In the above equation, 0 T represents the constant brightness temperature term expressed in [K], 1 T is the first harmonic of the brightness temperature given in [K], φ is the lunar phase angle in degrees and ψ is the phase lag in degrees [°]. In case that the lunar phase angle is in the decreasing cycle, a value of 360φ φ′ = − should be used in the Equation (2) calculation. Accord- ing to [8] the error introduced using the approximation of the Moon’s average brightness temperature given by the Equation (1) is less than 0.18%. The values 0 T [K], 1 T [K], ψ [°] and the ratio 1 0/T T , were determined from accurate radio measurements at few selected frequencies and have been interpolated in be- tween [8, 9]. Given parameters can be estimated using the following interpolation equations depending on the frequency of interest fGHz given in [GHz]: 0 24.43 207.7 GHz T f = + 1.2241 0 0.004212· GHz T f T = 43.83 1 0.0109 GHzf ψ = + (3) The Moon ephemeris, as are lunar angular diameter, lunar phase and other parameters depending on the observer location and on the current orbital positions of the Earth, Moon and Sun are provided by NASA web interface in [10]. The Moon’s flux density dependence on the lunar angular diameter and on the lunar phase are shown on the Figure 1 and Figure 2. Both results were calculated for the fre- quency of interest f = 8.1775 [GHz], and individually pre- sented with the fixed lunar phase angle of φ = 240o and the fixed angular diameter of qMoon = 0.5 o respectively. Figure 1: Lunar flux density change with angular diam- eter Figure 2: Lunar flux density change with phase angle 2.1 Uncertainties and proposed RF source The overall G/T measurement uncertainty is given “in quadrature” and presented for the Moon, the Sun and the radio star Cassiopeia A [5, 7, 8, 11]: Table 1: G/T measurement uncertainties Moon Sun Cassiopeia A S 0.33 dB 0.3 dB * 0.1 dB Y-factor 0.1 dB 0.1 dB * 1 dB K1 0.05 dB 0.05 dB 0.05 dB K2 0.2 dB 0.2 dB 0.01 dB G/T 0.4 dB 0.38 dB * 1 dB It must be noted that values in the Table 1 represent the G/T uncertainties due to the flux density S, the Y-factor and correction factors K1 and K2, not their uncertainties themselves. Also, the (*) sign marks that the possible D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 �Moon [◦] S M oo n [ W m 2 H z ] 0.48 0.5 0.52 0.54 0.56 ×10−22 2.4 2.6 2.8 3 3.2 3.4 φ � [◦] S M oo n [ W m 2 H z ] 0 100 200 300 400 ×10−22 2.4 2.6 2.8 3 3.2 3.4 φ 82 sunspot contribution is not considered because of its complicated computability. Considering the Moon’s good stability, high flux densi- ty and low uncertainty, it has been selected as the most adequate source for an accurate G/T measurement, for parabolic antennas with reflector sizes in the range of 7-13 meters. 3 G/T direct measurement method G/T direct calculation method requires a measure- ment of two well-defined signal sources [12]. Hence, the measurement is obtained by received noise power readings when the antenna is pointing at i) the Moon - PMoon [W Hz -1], and ii) the cold sky - Psky [W Hz -1]. Deri- vation of the G/T direct calculation method equation can be found in [13], while the final expression for the direct G/T quality factor calculation, expressed in [K-1], is given with the Equation (4): ( ) ( )( ) 1 22 8 · · 1 · · · B i i Moon k YG K K T S π θ θ λ − = (4) where kB [m 2 kg s-2 K-1] is Boltzmann constant, SMoon [W m-2 Hz-1] is the Moon’s flux density, λ [m] is the wave- length of interest, and Y represents the well known Y- factor noise power ratio given in linear scale. K1 and K2 are correction factors due to atmospheric attenuation and extended source size respectively. Finally, qi is the elevation angle at which the measurement was per- formed. Atmospheric attenuation for the X-band is very low, and is mostly composed of the attenuations due to gases, water vapour and scintillation. Those attenu- ations can be efficiently estimated using advanced mathematical models given in [15, 16, 17, 18]. Howev- er, the estimation of attenuations caused by fog, rainy clouds, and precipitation is based on empirical and static models. Therefore, the measurements should be performed in clear sky conditions. The correction factor for the atmospheric attenuation is expressed in linear scale and given as a sum of the attenuation contribu- tors. Impact of the atmospheric attenuation correction on the final G/T quality factor calculation can be within few tenths of decibel, whereas the impact of the ex- tended source size correction can result in few-decibel change of the final G/T quality factor calculation. The extended source size correction factor K2 is therefore a factor of great importance when using the Moon as an RF source in G/T measurements, and will be further dis- cussed in a separate subsection. 3.1 Elevation angle adjustment The measurement of G/T quality factor, using the Moon as an RF source, is performed at different elevation an- gles, depending on the Moon’s ephemeris. Therefore, it was important to establish a reference elevation angle, to which all measurement results will be adjusted. This was done in order to provide more realistic results and to allow mutual comparison of the results. A reference value for the G/T measurement elevation angle adjustment was chosen to be the elevation of qREF = 5° measured from the local horizon [13]. This ref- erence value adjusts a G/T quality factor for the worst case scenario, i.e. minimum elevation angle for satellite data acquisition, providing the most relevant G/T qual- ity factor result. The antenna power gain G is a property of the antenna and is constant with elevation angle variation. Therefore, the G/T quality factor values can be adjusted by noting that the variation of G/T factor with elevation angle is equal to the variation of the cold sky noise power with elevation angle: ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) i sys i sys REF n REF sys i n i REF sys REF GG T T PT G G T P T T θ θ θ θ θ θθ θ = = = (5) Where G/T(qi) is the quality factor value obtained for the elevation angle at the moment of measurement, and G/T(qREF) is the quality factor value adjusted to the reference value Tsys and Pn are the system noise tem- perature and the received noise power respectively, given for both measurement and reference elevation angles. The ratio of Pn(qi) and Pn(qREF) can be labelled as KEA and can be expressed both in linear and logarithmic scale, yielding a shortened expression of the adjusted G/T quality factor: ( ) ( ) n i EA n REF P K P θ θ = ( ) ( )·REF i EAG G KT Tθ θ= ( ) ( ) [ ] REF i EA dB dB dB G G K T T θ θ   = +       (6) 3.2 Extended source size correction factor The radio source whose angular diameter exceeds one fifth of the antenna’s half power beamwidth (HPBW) is considered an extended radio source. Each portion of D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 83 the RF source contributes to the received noise power according to the antenna radiation pattern. In the case of extended source, the peak of the antenna beam is assumed to be aligned with the center of the source, while the edges of the source are received by parts of the antenna beam with lower gain. This can result in the measured noise power smaller than what would be expected for the antenna’s effective collecting area and the aperture illumination [19]. Therefore it is necessary to correct the measurement by the extended source size correction factor K2 given by [13]: ( ) ( ) ( ) ( ) ( ) 2 2 0 0 2 2 2 0 0 , sin , , sin Moon Moon B d d K B g d d θπ θπ θ φ θ θ φ θ φ θ φ θ θ φ = ∫ ∫ ∫ ∫ (7) Where qMoon [°] is the Moon’s angular diameter, B(q,φ)[W m-2 sr-1 Hz-1] is the Moon’s brightness distribution, and g(q,φ) is the antenna radiation pattern normalized to the maximum directivity value, given in the linear scale. Considering the Moon’s radiation as that from a uni- form brightness disk, the expression of the K2 correc- tion factor can be simplified as: ( ) ( ) ( ) 2 2 0 0 2 2 2 0 0 sin , sin Moon Moon d d K g d d θπ θπ θ θ φ θ φ θ θ φ = ∫ ∫ ∫ ∫ (8) This approximation restricts measurement window to during the near full Moon phase. However, the inte- gration over the antenna’s radiation pattern is a rather complicated task and some alternative methods will be presented. Using the Moon as an RF source, the ratio of angular source size to the antenna’s HPBW is less than 3, which makes it possible to obtain a good approxima- tion of the normalized antenna radiation pattern using the normalized Gaussian far-field radiation pattern. The normalized Gaussian approximation of a radiation pattern, and the derivation of the K2 correction factor expression using the Gaussian pattern, are presented in [20, 13]: ( ) ( ) 2 2 2 ln 2 · ln 2 · 1 Moon HPBW Moon HPBWK e θ θ θ θ   −         = − (9) The approximated K2 correction factor depends on the Moon’s angular diameter qMoon [°] and the antenna’s HPBW qHPBW [°]. While the Moon’s angular diameter can be easily obtained on [10], the antenna’s qHPBW has to be properly measured. For the purpose of the HPBW verification, the well known estimation expression was used [21]: ·HPBW k d λθ = (10) where λ is wavelength in [m], d is the antenna reflector diameter in [m] and k is the antenna beamwidth factor. In the published literature, it is possible to find different definitions of the beamwidth factor, like k = 70 in [22], and k = 65 in [21]. However, the beamwidth factor is dependent on the feed’s edge-tapering and cannot be approximated with a constant value. For that purpose, using the reflector antenna analysis software GRASP, numerous radiation patterns for the antenna range of interest were produced. Beside changing the antenna reflector size, the edge tapering was also varied. The feed used in the simulations was a Gaussian beam pat- tern feed. Results were then analysed and extrapolated in order to find a best fitting expression for bandwidth factor definition: ( )58.96 1 0.0107· ek T= + (11) Where Te represents the absolute value of the edge ta- per given in the logarithmic scale [dB]. Usually, the extended source size correction factor es- timation method is provided by the antenna vendor in polynomial expression form. It represents the best fit for the specific antenna reflector sizes, and is given with relation to the frequency, fGHz [GHz], and the Moon’s an- gular diameter, qMoon. An example of the polynomial K2 expression and corresponding coefficients can be found in [13]. The comparison of the K2 polynomial expression, with K2 correction factor expression using the Gaussian pattern and HPBW estimation equation, is presented in Figure 3 - Figure 5. The K2 factor values were cal- culated for Cassegrain antennas designed with the reflector diameters of: 5.4, 7.3, 9.1, 10.26 and 11.28 meters, and have been interpolated in between using the polynomial interpolation and then adjusted to the uniform circular aperture illumination (k = 58.96 [14]). For the reference value, simulated radiation patterns were used in the general expression for the K2 correc- tion factor, given by Equation (8). The purpose of this comparison has been to analyse the K2 approximation method’s fit to the reference values for different edge tapering and antenna reflector sizes. Results presented in Figure 3 - Figure 5 were calculated for the frequency D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 84 of fGHz = 8.1775 [GHz], including blockage effect, with edge tapering of both Te = - 10 dB and Te = - 15 dB. On the figures, red curve presents the K2 reference value, blue curve presents the polynomial K2 expression, ma- genta curve presents the K2 values obtained with the Equations (9, 10) considering the beamwidth factor k = 70. Finally, the black curve presents the K2 values ob- tained with the Equations (9, 10) considering the novel beamwidth factor given with Equation (11). Figure 3: K2 approximation method comparison for Cassegrain antenna with Te = -10 dB and qMoon = 0.56°. Figure 4: K2 approximation method comparison for Cassegrain antenna with Te = -10 dB and qMoon = 0.5° From the figures given above it can be observed that the polynomial approximation method diverges by the largest factor from the reference level. The divergence of the polynomial approximation is assumed to be due to possible double-shaping of the antenna reflectors. To estimate beamwidth factor in order to use Gaussian approximation for double-shaped antennas, the value of uniform illuminated aperture k =58.96 can be taken for a rough estimation. Also, it can be seen that the Gaussian approximation when beamwidth factor is given as k = 70 yields signifi- cantly lower values than the reference. That beamwidth factor is probably suitable just for one specific edge ta- pering, usually used for the transceiver antennas. It can be concluded that the Gaussian pattern approxi- mation method for the K2 estimation is the best fitting K2 approximation method. Also, the HPBW approxima- tion expression using the proposed bandwidth factor provides a good estimation, and therefore the verifica- tion value, for the antenna of interest HPBW. Finally, the polynomial approximation method values can be too optimistic resulting in significant error in G/T estima- tion. Another important observation is that the polynomial approximation values, are constant for the antenna of interest, calculated for the optimal geometry an- tenna configuration. In case of antenna distortion the electrical properties of the antenna change. As a con- sequence, the extended source size factor value is re- duced, and use of a constant K2 can introduce signifi- cant errors in the G/T quality factor estimation. Analysis of the possible discrepancies in G/T estimation using the constant K2 factor in case of antenna distortions has been made (Figure 6 - Figure 8). For this purpose, every antenna was firstly designed and simulated for optimal performance. The K2 was in this case calculated with the numerical integration of the simulated antenna radiation pattern. This was done to neglect the uncertainties, as the scope of this test Figure 5: K2 approximation method comparison for Cassegrain antenna with Te = -15 dB and qMoon = 0.5° D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 Antenna reflector diameter [m] K 2 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 Polynomial Gauss(k = 70) Gauss(k = 58.96(1 + 0.0107 �Te)) Reference Antenna reflector diameter [m] K 2 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 Polynomial Gauss(k = 70) Gauss(k = 58.96(1 + 0.0107 �Te)) Reference Antenna reflector diameter [m] K 2 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 Polynomial Gauss(k = 70) Gauss(k = 58.96(1 + 0.0107 �Te)) Reference 85 is to provide an insight in problems occurring when using the constant correction factor values. Then, the radiation pattern simulation was performed for few axial displacements of the secondary reflector, in order to simulate antenna distortions. Therefore, the figures (Figure 6 - Figure 8) present the G/T quality factor re- sults of: i) Optimal geometry antenna and optimal geometry antenna’s K2 – green curve, ii) Distorted an- tenna keeping the optimal geometry antenna’s K2 – red curve, and iii) Distorted antenna with K2 calculated for the distorted antenna – blue curve. Figure 6: Secondary reflector displacement of 0.5 cm Figure 7: Secondary reflector displacement of 2 cm It can be concluded that G/T estimation using the con- stant K2 factor value, for the antennas with efficiency degradation due to distortions, is too optimistic. Also, the real degradation of distorted antenna G/T quality factor, expressed in logarithmic scale, can be more than twice the G/T quality factor degradation using the con- stant K2 factor value. 4 Measurements The measurement procedure is a result of detailed study and analysis, while the proposed settings for the spectrum analyser have been traded-off and selected in order to provide the best compromise between sta- bility and measurement error. The measurements were performed on the operational 11m Cassegrain anten- na, using the proposed procedure. Finally, obtained results were compared with well-defined G/T measure- ments using the radio star as a source. 4.1 Measurement procedure To ensure the maximum possible Y-factor readings, it is advised performing the measurements in the days between waxing and waning Moon phase. The Moon’s elevation should be higher than 30° to ensure that the sidelobes looking at the ground are 40 dB below the maximum directivity. Also, it should be verified that no other RF source is in the near vicinity of the Moon. Before performing the set of measurements for G/T fac- tor calculation, the antenna’s horizontal and vertical radiation pattern cut should be measured. This is done to verify the antenna’s HPBW, nulls and sidelobe behav- iour. Measurement can be roughly performed letting the Moon pass over the antenna boresight. Then, the final check of the measured HPBW can be performed following the Equation (4). Proposed spectrum analyser settings for the appropri- ate noise power readings are as follows: Figure 8: Secondary reflector displacement of 4 cm D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 Feed displacement of 0.5cm Antenna reflector diameter [m] G / T [d B / K ] 7 8 9 10 11 12 13 26 28 30 32 34 36 38 Opt. geo. antenna, opt. geo. K2 Dist. antenna keeping opt. geo. K2 Dist. antenna, dist. K2 Feed displacement of 2cm Antenna reflector diameter [m] G / T [d B / K ] 7 8 9 10 11 12 13 26 28 30 32 34 36 38 Opt. geo. antenna, opt. geo. K2 Dist. antenna keeping opt. geo. K2 Dist. antenna, dist. K2 Feed displacement of 4cm Antenna reflector diameter [m] G / T [d B / K ] 7 8 9 10 11 12 13 26 28 30 32 34 36 38 Opt. geo. antenna, opt. geo. K2 Dist. antenna keeping opt. geo. K2 Dist. antenna, dist. K2 86 Table 2: Proposed spectrum analyser settings Center frequency IF (typically 750 MHz) Frequency span 0 Hz dB/div 1 RBW 100 kHz VBW 10 Hz Marker ON Sweep time 100 ms Average 10 Each measurement consists of three different values: i) On-source – received noise power when pointing the Moon, ii) Off-source – received noise power pointing the cold sky, at the same elevation and 5° tilt in azimuth, and iii) 5°elevation – received noise power pointing the cold sky at the elevation of 5°, and same azimuth as the Off-source measurement. The measurements are repeated several times to mini- mize the Y-factor reading uncertainty. Correction fac- tors and flux density are calculated using the methods described in this paper, and finally, the average values in linear scale of Y-factor and KEA values are used to esti- mate the G/T quality factor of measured antenna. 4.2 Performed measurements - Moon Measurements were performed following the pro- posed procedure, first using the Moon and then using the Cassiopeia A radio star as a source, with the pur- pose of proposed method validation. The antenna un- der measurement is shown in Figure 9, and its charac- teristics are given in Table 3. Table 3: Antenna characteristics Type Cassegrain Frequency f = 8.1775 GHz Reflector diameter d = 11.28 m Approx. gain G ≈ 57.5 dBi The Moon ephemeris were obtained from [10] in order to estimate its flux density, which is given in Table 4: Table 4: Lunar ephemeris and flux density Parameter Value 0 T 210.687 K 1 0 /T T 0.05515 MoonT 214.0622 K ψ (lunar phase leg) 40.243 ° φ (average lunar phase angle) 147.114 ° Lunar phase cycle Increasing Moonθ (lunar angular diameter) 0.549 ° SMoon (lunar flux density) 3.1710-22 W m-2 Hz-1 Table 6: Measurement results with Moon as a source Time (UTC) [hh:mm] Elevation [°] On-source [dBm] Off-source [dBm] 5°elevation [dBm] Set 1 19 :23 40.30 -79.26 -82.87 -81.48 Set 2 19 :28 40.66 -79.38 -82.90 -81.50 Set 3 19 :31 40.75 -79.28 -82.87 -81.48 Set 4 19 :33 40.91 -79.38 -82.90 -81.50 Set 5 19 :36 40.98 -79.34 -82.87 -81.59 Set 6 19 :38 41.04 -79.30 -82.85 -81.60 Figure 9: X-band ground station with d = 11.28 m D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 87 The atmospheric attenuation correction factor was calculated using the local weather data at the time of the measurement, according to the latest ITU-R rec- ommendation. Extended source size correction factor, however, is calculated using the measured antenna’s HPBW, and the Gaussian pattern K2 approximation ex- pression. Measured HPBW has been verified with the proposed beamwidth approximation equation. Both correction factors are given in Table 5. Table 5: Local weather information and correction fac- tors Parameter Value Weather Cloudy Date 13.10.2016 Local Temperature 14.9 °C Local Humidity 91 % Local Pressure 960.9 hPa K1 1.033 HPBWθ 0.19 ° K2 5.71 The next step was to perform the measurements of the Y-factor following the proposed procedure. Measure- ment was repeated six times and the results are given in Table 6. Finally it was possible to estimate the G/T quality factor for the antenna under measurement. Table 7: Quality factor estimation using the Moon as a source Parameter Value Y-factor average 2.266 KEA average 0.733 G/T 4457 K-1 [G/T]dB 36.5 dBK-1 G/T uncertainty 0.4 dB Polynomial expression for K2 correction factor depending on frequency and Moon’s angular diameter provided the result of K2 = 6.35. Including the polynomial expression K2 value to the final G/T calculation using the Moon as a source yields the result [G/T]dB = 37 dBK -1. Using the polynomial ex- pression, the K2 measurement uncertainty is not known. 4.3 Performed measurements – Cassiopeia A Measurements using Cassiopeia A as a source were performed following the same measurement proce- dure using the same spectrum analyser. Cassiopeia A has very-well defined flux density and decay factor de- scribed in [5, 23]. The weather information at the time of measurement, Cassiopeia A flux density, and corre- sponding correction factors are presented in Table 8. Table 8: Local weather information and correction factors Parameter Value Weather Clear Date 19.10.2016 Local Temperature 14 °C Local Humidity 90 % Local Pressure 960 hPa K1 1.012 Cas Aθ 0.0767 ° HPBWθ 0.19 ° K2 1.0565 SCasiopeia A 401.56.10-26 W m-2 Hz-1 The measurements using the Cassiopeia A as a source are presented in Table 9, while the G/T quality factor estimation is presented in Table 10. Table 10: Quality factor estimation using the Cassio- peia A as a source Parameter Value Y-factor average 1.086 KEA average 0.811 G/T 4794 K-1 [G/T]dB 36.8 dBK-1 G/T uncertainty 1 dB Table 9: Measurement results with Cassiopeia A as a source Time (UTC) [hh:mm] Elevation [°] On-source [dBm] Off-source [dBm] 5°elevation [dBm] Set 1 17 : 28 57.30 -83.25 -83.63 -82.75 Set 2 17 : 31 57.60 -83.32 -83.66 -82.76 Set 3 17 : 33 57.90 -83.29 -83.65 -82.74 Set 4 17 : 36 58.30 -83.28 -83.64 -82.71 Set 5 17 : 38 58.60 -83.26 -83.61 -82.69 Set 6 17 : 42 58.90 -83.28 -83.64 -82.71 D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 88 From the presented results it can be observed that the estimation of G/T quality factor value using Cassiopeia A as a source, provides G/T value higher than when us- ing the Moon as a source for 0.3 dB. However, also the uncertainty of G/T calculated value is much higher than when using the Moon as a source. For the small angular diameters as is the case with the Cassiopeia A radio star, the polynomial expression is not valid. 5 Conclusions In this paper, an accurate and time effective method for G/T quality factor measurement using the Moon as an RF source has been described. The Moon was selected as the most adequate RF source for the antennas of in- terest, because of its stabile radiation flux density and low introduced uncertainty. The proposed method is considered valid for parabolic antennas with the reflec- tor apertures ranging from 7 to 13 meters. In case of smaller size antennas, the Y-factor readings using the Moon fall under 1dB introducing larger uncertainties. For antennas of larger sizes, however, the K2 factor un- certainty becomes dominant resulting in significant possible G/T estimation error and in that case a star ra- dio source becomes the preferred option. As the Moon results to be an extended source for the antennas of interest, the important achievement was to propose an improved method to compute the ex- tended source size correction factor. The improvement of a well-known K2 expression is given with polynomial expression for more accurate estimation of measured HPBW. This aspect turned out to be of key importance because a small error in the correction factor can result in an error of several tenths of a decibel in the final G/T value. Results have shown that the best correction factor es- timation can be obtained using a Gaussian pattern ap- proximation. Also, it turned out that other estimation methods, such as the use of a fixed polynomial expan- sion, are often too optimistic. Besides, by providing a constant correction factor val- ue for the antenna of interest, these methods are hid- ing possible degradations of the G/T quality factor in case of antenna efficiency degradations. Results have shown that the real degradation of distorted antenna quality factor expressed in dB can be more than twice the degradation using a constant correction factor val- ue. Proposed method uses the measured HPBW value confronted with the proposed polynomial expression, both for verification and for precision value rounding to two digits. The proposed direct measurement method is of great interest for the G/T measurements of typical X-band LEO satellite reception ground stations that have small pauses between satellite acquisitions. Some measurements taken on an operational X-band Cassegrain antenna 11 m antenna have been presented to confirm the model. Measurements were performed on site on a tight schedule and following the procedure given in this paper. Measurement results are consistent and in line with expectations, and have shown a good agreement with measurements made on the ground station using the Cassiopeia A as an RF source. Based on the obtained results, the European Space Agency (ESA/ESRIN) has updated the procedure used for periodic G/T measurement, using the method de- scribed in this paper. 6 Acknowledgments The authors would like to express their gratitude to Paolo Rutigliano and Michele Paradiso of e-GEOS for the opportunity they provided to participate in their periodic G/T measurement campaign, and for the prac- tical knowledge they shared during the process. 7 References 1. D. F. Wait, M. Kanda, W. Daywitt in C. Miller, “A study of the measurement of G/T using Cassio- peia A,” tech. rep., DTIC Document, 1974. 2. W. C. Daywitt, “On 10-60 GHz G/T measurements using the Sun as a source: A preliminary study (Report, 1985-1986).” (1986). 3. D. A. Giudice, and J. P. Castelli, “The use of extrater- restrial radio sources in the measurement of an- tenna parameters.” IEEE Transactions on Aerospace and Electronic Systems 2 (1971): 226-234. 4. M. Ott, et al. “An updated list of radio flux den- sity calibrators.” Astronomy and Astrophysics 284 (1994): 331-339. 5. J. W. M. Baars, et al. “The absolute spectrum of CAS A-an accurate flux density scale and a set of sec- ondary calibrators.” Astronomy and Astrophysics 61 (1977): 99-106. 6. E. Ekelman and C. Abler, “Antenna gain mea- surements using improved radio star flux den- sity expressions,” in Antennas and Propagation Society International Symposium, 1996. AP-S. D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 89 Digest, volume=1, pages=172-175, year=1996, organization=IEEE. 7. D. F. Wait, M. Kanda, W. Daywitt and C. Miller, “A study of the measurement of G/T using Cassio- peia A,” tech. rep., DTIC Document, 1974. 8. W. C. Daywitt, “An error analysis for the use of pres- ently available lunar radio flux data in broadbeam antenna-system measurements,” Error analysis for the use of presently available lunar radio flux data in broadbeam antenna-system measurements.. Re- port NBS-TN-1073, Natl. Bur. Stand., Washington, DC, USA, 30 pp., vol. 1, 1984. 9. Z. Kopal, Advances in astronomy and astrophysics, vol. 3. Academic Press, 2013. 10. JPL-NASA, “HORIZONS Web-Interface.” Available at: http://ssd.jpl.nasa.gov/horizons.cgi. Accessed: 24. 10. 2016]. 11. W. C. Daywitt, “On 10-60 GHz G/T measurements using the Sun as a source: A preliminary study.” Report, 1985-1986 National Bureau of Standards, Boulder, CO. Electromagnetic Fields Div. (1986). 12. W. Welch and D. DeBoer, “Expected properties of the ATA antennas,” ATA Memo Series, vol. 66, 2004. 13. D. Šekuljica, “Using the Moon as a calibrated noise source to measure the G/T figure-of-merit of an X-band satellite receiving station with a large an- tenna 200...400 wavelengths in diameter” : mas- ter’s thesis. Ljubljana: [D. Šekuljica], 2017. XVI, 116 str., ilustr. https://repozitorij.uni-lj.si/IzpisGradiva. php?id=88902. [COBISS.SI-ID 11686996]. 14. T. A. Milligan, Modern antenna design. John Wiley & Sons, 2005. 15. ITU-R recommendation, “Attenuation by atmo- spheric gases,” ITU-R P. 676. 16. ITU-R recommendation, “Reference Standard At- mospheres,” ITU-R P. 835. 17. ITU-R recommendation, “Propagation data and prediction methods required for the design of Earth-space telecommunication systems,” ITU-R P. 618. 18. ITU-R recommendation, “Attenuation due to clouds and fog,” ITU-R P. 840. 19. A. Solovey in R. Mittra, “Extended source size correction factor in antenna gain measure- ments,” in Microwave Conference, 2008. EuMC 2008. 38th European, pages=983-986, year=2008, organization=IEEE. 20. H. Ko, “On the determination of the disk tempera- ture and the flux density of a radio source using high-gain antennas,” IRE Transactions on Antennas and Propagation, vol. 9, no. 5, str. 500-501, 1961. 21. ITU, Handbook on Satellite Communications (FSS). International telecommunication union, 2002. 22. R. A. Nelson, “Antennas: the interface with space,” Via Satellite, Sep, 1999. 23. J. Baars, P. Mezger in H. Wendker, “The Spectra of the Strongest Non-Thermal Radio Sources in the Centimeter Wavelength Range.,” The Astrophysical Journal, vol. 142, str. 122, 1965. Arrived: 06. 04. 2017 Accepted: 26. 06. 2017 D. Šekuljica et al; Informacije Midem, Vol. 47, No. 2(2017), 79 – 89 90 91 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 91 – 99 Reversible Data Hiding Based on Radon and Integer Lifting Wavelet Transform A.Amsaveni1, P.T Vanathi2 1Kumaraguru College of Technology, Department of Electronics and Communication Engineering, Coimbatore, India 2PSG College of Technology, Department of Electronics and Communication Engineering, Coimbatore, India Abstract: This paper presents a reversible data hiding technique based on radon and integer lifting wavelet transform to secure the data transmitted over communication network. The technique focuses on three optimality criteria, namely imperceptibility, robustness, and reversibility. The frequency domain strategy is applied due to its superior performance over the spatial domain techniques in certain important aspects like robustness and reversibility towards signal processing and image processing operations. The cover image is first transformed from spatial domain to radon domain and then, this radon image is applied with integer lifting wavelet transform. As the Radon transform performs rotation, scaling and translation operations on the cover image, it changes the locations of the secret bits. Hence, it is very difficult to detect the embedded data without taking the inverse Radon transform and subsequently, it increases the security of the embedded payload. The integer lifting wavelet transform guarantees complete reversibility as they produce integer wavelet coefficients. Then, the middle bit planes of high frequency lifting coefficients are compressed using arithmetic coding to provide space for embedding secret payload. As the proposed framework embeds data in red, green, and blue channels, it can work well for a variety of images with different distribution of colors. Keywords: reversible data hiding; radon transform; integer lifting transform;bit plane coding;arithmetic coding Reverzibilno skrivanje podatkov na osnovi Radonove in diskretne valčne transformacije Izvleček: Članek opisuje tehnike reverznega skrivanaj podatkov na osnovi Radonove in diskretne valčne transformacije za varovanje podatkov preko omrežja. Tehnika sloni na trhe kriterijih: neopaznost, robustnost in reverzibilnost. Uporabljena je strategija na osnovi frekvence saj prednjači pred prostorsko tehniko v ključnih točkah robustnosti in reverzibilnosti obdelave signalov in slik. Slika je najprej pretvorjena v radon proctor in nato še z diskretno valčno transformacijo. Radonova transformacija opravi rotacijo, skaliranje in translacijo in spremeni lokacijo skrivnih bitov. Brez inverzne transformacije je skoraj nemogoče odkriti skrite podatke. Valčna transformacija zagotavlja popolno reverzibilnost in določi valčnbe koeficiente, ki so stisnjeni z aritmetičnim kodiranjem. Predlagan postopek vsebije morer, zelen in rdeč kanal, tako da je uporaben za številne slike. Ključne besede: reverzibilno skrivanej podatkov; radonova transformacija; diskretna valčna transformacija; koditanje bitne ravnine; aritmetično kodiranje * Corresponding Author’s e-mail: amsaveni.a.ece@kct.ac.in 1 Introduction Securing data transmitted over the internet has be- come a challenging issue caused by the advancement in data digitization and communication networking over the past decade. Therefore, it is necessary to de- vise strategies to secure information during the  pro- cess of information exchange. Reversible data hiding has emerged as a major research area due to the phe- nomenal growth in internet and multimedia technolo- gies. It involves concealing confidential data within an- other seemingly innocuous cover media such as text, video, audio, images, and compression coding [1]. 92 A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 Basic terminologies used in data hiding are as follows: Secret Payload – Message to be embedded in the cover image; Cover image – Image that carries the secret mes- sage; Stego image / Embedded image – Cover image after embedding the secret payload (Cover image + Secret data); Imperceptibility – Measure of distortion which is caused by embedding the secret message in the original cover image. It is the inability of the human eye to differentiate between the cover image and the embedded image. Generally, embedded images with higher imperceptibility are preferred in data hiding; Robustness – Robustness is the ability of the secret pay- load to withstand various intentional attacks such as image processing operations and unintentional attacks such as addition of noise; Embedding Capacity – Data embedding rate or number of bits embedded per pixel, measured in terms of bits per pixel; Security / Undetect- ability – Any data hiding system may be considered as secure if the possibility of knowing the presence of a secret message in any cover medium is very difficult; Attacks – Process of revealing the hidden data from the embedded image by attacking with various signal pro- cessing and image processing techniques. While embedding secret information in a cover image, the emphasis is on two key problems. The first one is to produce an embedded image with a tolerable level of quality so that the distortion produced due to data embedding is imperceptible. The second is to pro- duce the embedded image that is distortion tolerant (robust), i.e., even if the embedded image is attacked during communication, the hidden data can be recov- erable. Robustness of data hiding techniques can be enhanced if the properties of the cover image could be properly utilized. By considering these aspects, embed- ding data in frequency domain becomes more popu- lar compared to spatial domain techniques [2]. The frequency domain techniques modify the frequency coefficients of the cover image by applying a specific transformation function on the cover image. They are designed to be imperceptible and robust against vari- ous geometrical transformations and external attacks. Frequency domain schemes use transformation meth- ods, such as integer cosine transform [3] or integer wavelet transform [4] to compute the transform coef- ficients of the cover image. Then, these coefficient val- ues are modified to embed the secret data. In revers- ible data hiding technique based on DCT, one secret bit was embedded into two neighboring DCT coefficients in an image block [5].The secret payload is embedded in the high frequency coefficients of discrete wavelet transform (DWT) by exploiting the statistical properties of the cover image [6]. The DWT works well against var- ious image processing attacks. As it produces floating coefficients, embedded data is potentially lost while re- constructing the cover image by inverse wavelet trans- form. This drawback is overcome by integer wavelet transform (IWT). The secret data is embedded into the middle frequency of the integer lifting wavelet domain by modifying the histogram of the cover image [7]. Data embedding is also performed in radon domain, which shows a considerable improvement in bit er- ror rate. A radon-based approach was introduced to incorporate translation invariance properties to the payload [8]. RST invariant watermarking technique has been proposed by utilizing the Fourier transform and transforming them to log polar coordinates, which are quite flexible towards rotation, scaling and translation attacks [9]. Hybrid transform has been proposed based on the unique features of the transforms in the hybrid combi- nation, so that it is able to address the robustness and reversibility criteria. Accordingly, the rotation, scaling and translation properties of the Radon transform and reversibility property of integer lifting transform have been joined together in a hybrid formation. The paper investigates a combinatorial data hiding approach us- ing radon transform and integer lifting wavelet trans- form (ILWT). Radon transform ensures robustness and ILWT makes the algorithm reversible by using the lift- ing scheme on orthogonal and bi-orthogonal wave- lets. The superiority of this combination is also tested and compared with the other existing works in litera- ture. The proposed method improves the quality of embedded image as well as the robustness of embed- ded payload against various attacks compared to the existing methods. This paper is organized as follows. Section 2 describes the Radon transform. Section 3 explains Integer Lifting Transform. Section 4 discusses the proposed algorithm. Section 5 presents the experimental results and, Sec- tion 6 concludes the paper. 2 Radon transform Radon transform is a linear transform which is an effec- tive method to analyze signal between the spatial do- main and its projection space. It represents the image as a collection of projections along various directions [10]. It computes the projection of the image intensity along a radial line oriented at a specific angle. For each angle q and at each distance ρ, the intensity of a ray perpendicular to the ρ axis is summed up at R (ρ,q). Ra- don transform converts rectangular coordinates (x, y) into polar coordinates (ρ,q). The simplest form of dis- crete Radon transform is to select finite number of the 93 angular variable of projection, then to take the summa- tion on the discrete image along the projection line. As shown in Fig.1, the radon transform of a two dimen- sional function f (x, y) is the integral of function f along a straight line parallel to the y-axis, which is given by, ( )'R xθ = ( )' ' , ' ' 'f x cos y sin x sin y cos dyθ θ θ θ − − −∫ (1) Where ' ' x y       cos sin x sin cos y θ θ θ θ        −    = (2) Figure 1: Geometry of the Radon transform An efficient reversible data hiding method must be ro- bust against a wide range of image processing opera- tions such as image enhancement, cropping, rotation, scaling, compression, and signal processing operation such as addition of noise. However, conventional data hiding algorithms are more sensitive to geometric distortions. Hence, radon transform is introduced to perform rotation, scaling, and translation operations on the cover image. These operations change the posi- tioning of the secret bits. Without taking inverse Radon Transform, it is very difficult to detect the embedded data and subsequently, this increases the security of embedded payload. 3 Integer lifting wavelet transform There are many researches that have been explored us- ing wavelets in the field of image processing and image steganography. The main advantage of wavelet is that they offer multi-resolution capability, which is similar to the operation of the human visual system. Wavelets provide an optimal representation of signals. Normally, wavelet-based data hiding gives better performance compared to other methods. As the conventional wavelet transform performs a convolution of the input image and the wavelet basis, it requires large memory space for the computation process. The time taken and the large memory required for the conventional wave- let transform is reduced in the lifting method. Lifting transform is a technique used in constructing second generation wavelets entirely in the frequency domain. It is fast compared to the first generation wavelets, as it requires only addition and subtraction. Conventional wavelet transform is not suitable for re- versible data hiding scheme as reversibility property is not guaranteed. Wavelet transform operates on a floating-point arithmetic basis. An image that has inte- ger intensity values in the spatial domain is converted into decimal wavelet coefficients. The wavelet coeffi- cients are modified appropriately during data hiding operation and inverse wavelet transform is carried out to reconstruct the image in the spatial domain. A se- rious note here is that practically wavelet coefficients are truncated or rounded as it is not viable to represent the coefficients to its full accuracy. Information is po- tentially lost while reconstructing the image by inverse wavelet transform. But reversible data hiding schemes have to recover the host image without distortion along with the secret payload. Eventually, this makes the discrete wavelet transform a poor choice for revers- ible data hiding [11]. To address this specific issue, an invertible integer lifting wavelet transform is used in the proposed scheme. The system operates on integer arithmetic and alleviates the loss of any information via forward and inverse transforms [12]. The lifting wavelet transform decomposes the image into frequency subbands, which contain approxima- tion and detail coefficients. The system reserves the detail coefficients, which have texture, edges, and re- gion boundary for data hiding. It is an insensible region for human visual system. An advantage of the Lifting Scheme is that it can be converted easily into a trans- form that maps integers to integers while retaining the perfect reconstruction property. Thus, embedding data in integer lifting wavelet domain satisfies the proper- ties like security, imperceptibility, and robustness of the proposed technique. 4 Bit plane embedding using binary arithmetic coding The Binary Arithmetic Coding can be exploited for compressing the bit planes of grayscale/colour images. As arithmetic coding is a lossless compression method, it guarantees the recovery of original payload. In bit plane embedding, the most significant bits for each A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 94 pixel are grouped into one bit plane, the next most significant bits into another bit plane and so on till the least significant bit plane. Mostly, the five highest order planes contain visually significant data. The other lower bit planes contain fine details in the image. Lesser the bit plane number, lesser is its contribution to the final stage. Statistically, there is an equal distribution of zeroes and ones in the lower planes of the image than in the high- er planes. This leads to lower compression ratio and lower embedding capacity in the lower bit planes than in the higher planes. This is because binary sequences of length L having higher probability may be encoded more compactly than another one of the same length with a lower probability. But the signal to noise ratio falls down as higher bit planes are altered for embed- ding [13]. The most significant bit plane contains the most critical approximation values of the image. Hence, modifica- tions made in higher bit-plane may degrade the quality of the cover image. In order to have the embedded im- age visually as same as the cover image, data is hidden in one or more middle bit planes. The bits in one or more bit-planes can be compressed to provide space to hide data like text or image due to the existence of redundant information. The ap- proximate coefficients in the LL sub-band contribute to visual perception. Hence, the secret bits are embed- ded in LH, HL or HH subbands (Detail Coefficients). The original bits in the selected bit plane of LH, HL or HH subbands are compressed using arithmetic coding to provide space for embedding the payload bits. The structure of the embedded bit plane is shown in Fig.2. CHH, CHL, and CLH headers represent the header information of the compressed HH, HL, and LH sub bands. They describe the bit distribution required for arithmetic encoder and decoder. CHH, CHL, and CLH lengths denote the length of the compressed bit stream in the chosen bit plane of the LH, HL, and HH subbands. 5 Proposed methodology Majority of the methods discussed in the literature addressed only a few of the desired characteristics, namely lossless/reversible, imperceptible, high pay- load capacity and robustness, and not all. The pro- posed method of reversible data hiding is based on bit plane embedding in radon and integer lifting wavelet domain. It aims to meet all the desired characteristics to an optimal level. The block diagram of the proposed data embedding algorithm is shown in Fig.3. Apply Radon Transform to red/ green/ blue chanel Integer Liing Wavelet Transform Select LH, HL and HH bands Choose middle bit plane(s) Apply arithmec coding Embedding algorithm Take Inverse ILWT and Inverse radon transform Combine all the three channels Approximaon Co-efficients Modified Detail Co-efficients Stego Image Message to be embedded Compressed LH, HL&HH Cover Image Figure 3: Block diagram of embedding process CLH Header (16 bits) CHL Header (16 bits) CHH Header (16 bits) CLH Length (16bits) CHL Length (16bits) CHH Length (16 bits) Length of embed- ding data (32 bits) Secret Payload Figure 2: Structure of the embedded bit plane A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 95 5.1 Data Embedding Algorithm Input: Cover image with M-rows and N-columns, Se- cret payload bits. Output: Stego Image Step 1: Read the cover image of size M x N. Step 2: Separate color channels (Red, Green, and Blue) of the cover image. Step 3: Apply radon transform on any one of the channels. Step 4: Notice that Radon image undergoes a single level integer lifting wavelet transform which results in 4 subbands (LL, LH, HL, HH) of size M/2 x N/2, each. Step 5: Construct binary images from the chosen bit planes of LH, HL, and HH bands. Step 6: Compress the original bits in the chosen bit plane of these bands using arithmetic coding and obtain the header information required for the arithmetic encoder and decoder. Step 7: Read the secret payload and convert it into a bit string. Step 8: Concatenate the header length, header infor- mation, and compressed bit streams of CLH, CHL, and CHH and the secret payload to get a single bit stream. Step 9: Embed bit stream into chosen bit plane of LH band. If not enough, embed in HL band and then in HH band and observe that it results in embedded LH, HL, and HH components. Step 10: Apply inverse integer lifting transform on LL coefficients and the modified LH, HL, and HH coefficients. Step 11: Compute inverse radon transform of the im- age obtained from Step 10. Step 12: Combine all the three color channels to get stego image. 5.2 Data Extraction Algorithm Step 1: Read the stego image of size M x N. Step 2: Separate color channels (Red, Green, and Blue) of stego image. Step 3: Apply radon transform to the channel in which the data is embedded. Step 4: Notice that Radon image undergoes single level integer lifting wavelet transform which results in 4 subbands (LL, LH, HL, HH) of size M/2 x N/2, each. Step 5: Construct binary images from chosen bit planes of LH, HL, and HH bands. Step 6: Derive the header information and header length needed for arithmetic decoding. Step 7: Extract the compressed bits from the chosen bit plane of these bands using arithmetic de- coder and decompress the subbands to get the reconstructed subbands. Step 8: Apply inverse integer lifting transform on the reconstructed subbands LH, HL, and HH along with LL subband. Step 9: Compute inverse radon transform of the im- age obtained from Step 8. Step 10: Combine all the three channels to get the original cover image. 6 Expermental results In order to investigate the performance of the pro- posed data hiding algorithm, several experiments are carried out in a computer system equipped with Intel core 2 duo processor with 2 GB memory and a clock speed of 2 GHz. Matlab 8 (R2013a) platform is used for the digital simulation of the algorithm. Five standard 512 x 512 color images such as (a) Airplane,(b) Baboon, (c) Boat, (e) Lena and (e) Pepper, obtained from USC-SIPI (Image database), have been used as cover images. The performance of the algorithm is investigated in terms of imperceptibility, robustness, and reversibility. 6.1 Imperceptibility The metrics used to test the imperceptibility property of the proposed algorithm are PSNR (Peak Signal to Noise Ratio) and SSIM (Structural Similarity Index Meas- ure). The PSNR for an image of size M x N is calculated by, PSNR = 10 log10 (255 2 / MSE) dB (3) where, MSE= ( ) ( )( ) 1 1 , ' , M N x y p x y p x y = = −∑∑ 2 (4) where p(x, y) stands for the pixel value in the cover im- age and p’(x, y) is the pixel value at position (x, y) in the stego image after embedding the secret message. M and N denote the number of rows and columns of the image and (x, y) denotes the pixel coordinates. The quality of the stego image is also calculated using SSIM as follows: SSIM = ( ) ( ) ( )( ) 2μμy c1 2σxy c2 μ2x μ2y c1 σ2x σ2y c2+ + + + ++ (5) where x and y are same size windows of the cover and stego images and μx and μy are corresponding x and y averages. σ2x and σ 2 y are the variances of x and y and A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 96 σxy is the covariance of x and y. The positive constants c1 and c2 are included to avoid a null denominator. Typically c1 = (k1L) 2 ; c2 = (k2L) 2; L= (2no.of bits/pixel) -1; [k1, k2]=[0.01,0.03] by default. Table 1 shows the PSNR and SSIM of cover images after embedding a payload of 10,000 bits using the wavelet cdf2.4 (Cohen-Daubechies-Feauveau 2.4) in the fourth bit plane of Red, Green and Blue channels. As the PSNR obtained from the stego images is greater than 42 dB, the embedded payload is highly imperceptible to the human eye, i.e., the perceptual quality of the result- ant stego images is good. The red channel offers bet- ter PSNR and SSIM compared to green and blue chan- nels. The red channel gives an improvement in PSNR of about 2.0 to 7.0 dB over green and blue channels. Among all cover images, the Airplane image yields bet- ter PSNR for the same payload. Table 1: Quality metrics Cover image PSNR (dB) and SSIM Red Channel Green Channel Blue Channel PSNR SSIM PSNR SSIM PSNR SSIM Airplane 47.70 0.9771 47.65 0.9867 46.70 0.9519 Baboon 43.29 0.9900 42.91 0.9704 42.49 0.9894 Boats 46.96 0.9921 43.23 0.9740 43.69 0.9693 Lena 45.87 0.9820 42.46 0.9780 43.46 0.9409 Pepper 46.86 0.9784 44.34 0.9840 42.39 0.9656 The Fig.4. shows the embedded images of size 512 x 512, after embedding a payload of 10,000 bits using cdf2.4 in the fourth bitplane of red channel of standard images. Table 2 summarizes the quality of the Lena image un- der varying payload, after embedding in the bit plane 4, 5, and 6 of red channel using wavelet cdf2.4. Table 2: Embedding capacity Vs. PSNR (dB) Secret payload (bits) Embedding Rate (bpp) PSNR (dB) Bit Plane 4 Bit Plane 5 Bit Plane 6 1000 0.004 45.27 43.08 41.42 3000 0.011 45.26 43.07 41.40 6000 0.023 45.14 43.04 41.20 10,000 0.038 45.10 43.02 41.06 20,000 0.076 44.68 42.54 40.83 40,000 0.153 42.21 40.41 38.33 50,000 0.191 42.20 39.80 38.33 70,000 0.267 41.49 39.52 37.44 80,000 0.305 41.28 38.79 37.17 86,000 0.328 41.16 38.72 37.13 90,000 0.343 Insufficient 38.64 37.11 95,000 0.362 Insufficient Insufficient 37.04 96,000 0.366 Insufficient Insufficient Insufficient The perceptual quality of cover image will get reduced if the data is embedded in higher bit planes and also PSNR drops down as more number of bits embedded in that plane. The bit plane 4 can accommodate only 90,000 bits as the bit plane 4 provides less space for data embedding compared to other bit planes. The maximum embedding capacity of bit planes 5 and 6 is 95,000 bits and 96,000 bits respectively. Beyond 95,000 bits, there is no space to accommodate the secret pay- load in both the bit planes. The embedding capac- ity completely depends upon cover image and is also based on the bit distribution of the chosen bit plane. The PSNR varies from 41.12 dB at the embedding rate of 0.004 bits per pixel to 37.04 dB at 0.362 bits per pixel for bit plane 6. The experimental results of the proposed scheme are compared with the various schemes discussed in Tsai & (a) (b) (c) (d) (e) Figure 4: Standard color images of size 512 x 512 after embedding a payload of 10,000 bits (a) Airplane; (b) Ba- boon; (c) Boat; (d) Lena; (e)Pepper A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 97 Sun [14], Fu & Shen [15], and Niu et al [16] and summa- rized in Table 3. The PSNR of the cover images is meas- ured after embedding a payload of 10,000 bits in Red channel using the wavelet cdf2.4.The PSNR offered by the proposed scheme is about 15% greater than Niu et al scheme. The proposed scheme gives better PSNR for Lena image compared to Baboon and Boat images. Table 3: Comparison of the proposed scheme with the schemes in the literature Host image PSNR (dB) Tsai and Sun Scheme Fu and Shen Scheme Niu et al Scheme Proposed Scheme Lena 39.43 38.10 40.57 45.87 Baboon 41.76 38.90 41.67 43.29 Boat 42.49 39.72 40.71 44.03 6.2 Robustness Robustness is measured using Bit Error Rate (BER) and is defined as: k k 0 b b n k= = ∑BER ’ / N (6) where b and b’ are embedded and extracted bits re- spectively, N is the total number of secret bits embed- ded and represents the XOR operation. The value of BER ranges between 0 and 1. If BER is closer to 1 then, it means that the error value of extracted data is higher. The value of BER is calculated after retrieving the se- cret data from the embedded block. Lower the BER%, higher is the accuracy of the extracted secret data. The stego images are added with Gaussian noise with a variance of 0.2 and 0.4, Poisson Noise, Impulse Noise with a variance of 0.05 and 0.1 and Speckle Noise with a variance of 0.05 and 0.1. Generally, addition of noise is responsible for the degradation of the image. The im- age processing operations such as Rotation (5 and 10 degrees), Scaling (200% and 400%), Blurring and Crop- ping (10% and 25%) are performed on embedded im- ages. After subjecting to the attacks, the original cover image is extracted and the bit error rate of extracted payload over secret payload is measured. Table 4 summarizes the experimental results for the proposed data hiding scheme against various attacks. As the BER is about 0.15 to 0.35 % of embedded payload, the algorithm is robust against various intentional and unintentional attacks. 6.3 Reversibility In order to ensure the reversibility, the extracted cover image and the original cover image must be similar.The metric used to measure the similarity between the two images is Normalized Correlation Coefficient (NCC). The value of 0 represents no correlation. NCC will ap- proach to one if the extracted cover image resembles the original cover image. The Normalized Correlation Coefficient between cover image and extracted cover image is defined as, ( ) ( )( ) ( )( ) ( )( ) , 2 2 , , ( , ) , , , − − = − − ∑ ∑ ∑ mean meanx y mean meanx y x y f x y f g x y g f x y f g x y g NCC (7) Table 4: Effect of various attacks on BER Attacks Bit Error Rate Airplane Baboon Boats Lena Pepper Gaussian Noise (σ2 =0.2 ) 0.0032 0.0034 0.0035 0.0032 0.0034 Gaussian Noise (σ2 =0.4) 0.0036 0.0036 0.0037 0.0037 0.0038 Poisson Noise 0.0028 0.0030 0.0029 0.0030 0.0031 Impulse Noise (σ2=0.05) 0.0017 0.0020 0.0021 0.0018 0.0022 Impulse Noise (σ2=0.10) 0.0019 0.0022 0.0024 0.0020 0.0024 Speckle Noise (σ2=0.05) 0.0018 0.0021 0.0025 0.0026 0.0024 Speckle Noise (σ2=0.10) 0.0019 0.0024 0.0027 0.0027 0.0026 Rotation (5°) 0.0018 0.0021 0.0020 0.0022 0.0024 Rotation (10°) 0.0020 0.0023 0.0023 0.0024 0.0026 Scaling (200%) 0.0016 0.0017 0.0018 0.0019 0.0022 Scaling (400%) 0.0018 0.0019 0.0020 0.0020 0.0024 Blurring (5) 0.0024 0.0023 0.0020 0.0019 0.0023 Blurring (10) 0.0027 0.0026 0.0022 0.0021 0.0025 Cropping (10%) 0.0026 0.0028 0.0030 0.0029 0.0031 Cropping (25%) 0.0031 0.0034 0.0033 0.0031 0.0033 A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 98 where f (x, y) is the original cover image, g (x, y) is the extracted cover image. Table 5 summarizes the experimental results for the proposed data hiding scheme against various attacks. As the NCC values are greater than 0.98, it is concluded that the algorithm restores the original cover image ex- actly at the destination. 7 Conclusion In this paper, a reversible data hiding technique based on radon and integer lifting wavelet transform is pre- sented. Hybrid transform has been proposed based on the unique features of the transforms, so that it is able to address the robustness and reversibility criteria. As the Radon transform performs rotation, scaling and translation operations on the cover image, it changes the positioning of the secret bits. The integer lifting wavelet transform guarantees complete reversibility. The original bits in the selected bit planes of LH, HL or HH subbands are compressed using arithmetic cod- ing to provide space for embedding the payload bits. Generally, middle bit planes are used for embedding as they provide a balanced trade-off between embedding capacity and visual quality of that stego image. Data is embedded in red, green, and blue channels of the color image independently. As the PSNR obtained for the stego images is greater than 42 dB, the embedded payload is imperceptible to the human eye. The results have been compared with the existing works in the literature and the proposed method gives 10 to 15% improvement in PSNR. As the BER is about 0.15 – 0.35 % of embedded payload, the algorithm is robust to at- tacks. From the simulation results, it is inferred that the proposed algorithm exhibits reversibility due to high NCC values. 8 References 1. Amsaveni, A. and Vanathi, P.T. (2015) ‘A compre- hensive study on image steganography and steganalysis techniques’, Int. J. Information and Communication Technology, Vol. 7, Nos. 4/5, pp.406–424. 2. Amsaveni, A. and Vanathi, P.T. (2015) ‘An efficient reversible data hiding approach for colour imag- es based on Gaussian weighted prediction error expansion and genetic algorithm’, Int. J. Advanced Intelligence Paradigms, Vol. 7, No. 2,pp.156–171. 3. Yang, B., Schmucker, M., Funk, W., Busch, C & Sun, S (2004), ‘Integer DCT-based reversible watermark- ing for images using companding technique’, Proceedings of Electronic Imaging, Science and Technology, vol. 5306, pp. 405-415. 4. Xuan, G., Zhu, J., Chen, J., Shi, Y.Q., Ni, Z., & Su, W (2002), ‘Distortion less data hiding based on inte- ger wavelet transform’, IEEE Electronics Letters., vol. 38, no. 25, pp. 1646-1648. 5. Lin, S.D., Shie, S.C., & Guo, J.Y., (2010), ‘Improving the robustness of DCT based image watermark- ing against JPEG compression’, International Jour- nal of Computer Standards and Interfaces, vol. 32, no. 1, pp. 54-60. Table 5: Effect of various attacks on NCC Attacks Normalized Correlation Co-efficient Airplane Baboon Boats Lena Pepper Gaussian Noise (σ2 =0.2 ) 0.9817 0.9843 0.9825 0.9819 0.9824 Gaussian Noise (σ2 =0.4 ) 0.9850 0.9842 0.9860 0.9805 0.9814 Poisson Noise 0.9829 0.9844 0.9832 0.9822 0.9814 Impulse Noise (σ2=0.05) 0.9839 0.9843 0.9816 0.9811 0.9813 Impulse Noise (σ2=0.10) 0.9810 0.9816 0.9804 0.9802 0.9807 Speckle Noise (σ2=0.05) 0.9833 0.9822 0.9871 0.9864 0.9846 Speckle Noise (σ2=0.10) 0.9811 0.9804 0.9810 0.9820 0.9825 Rotation (5°) 0.9880 0.9806 0.9802 0.9824 0.9816 Rotation (10°) 0.9863 0.9794 0.9788 0.9804 0.9810 Scaling (200%) 0.9835 0.9820 0.9869 0.9846 0.9863 Scaling (400%) 0.9825 0.9806 0.9832 0.9828 0.9822 Blurring (5) 0.9854 0.9810 0.9809 0.9817 0.9826 Blurring (10) 0.9825 0.9806 0.9789 0.9738 0.9805 Cropping (10%) 0.9732 0.9726 0.9727 0.9728 0.9716 Cropping (25%) 0.9702 0.9706 0.9701 0.9706 0.9678 A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 99 6. Xiang, S., Kim, H.J., & Huang, J (2008), ‘Invariant Image watermarking based on statistical features in the low frequency domain’, IEEE Transactions on Image Processing, vol. 14, no. 12, pp. 2140- 2150. 7. Xuan, G., Zhu, J., Chen, J., Shi, Y.Q., Ni, Z., & Su, W., (2002), ‘Distortion less data hiding based on inte- ger wavelet transform’, IEEE Electronics Letters., vol. 38, no. 25, pp. 1646-1648. 8. Stankovic, S, Djurovic, I & Pitas, I 2001,  ‘Water- marking In The Space/Spatial-Frequency Domain Using Two-Dimensional Radon-Wigner Distribu- tion’, IEEE Transactions on Image Processing, vol. 10, no. 4, pp. 650-658. 9. Lin, C, Wu, M, Bloom, J, Cox, I, Miller, M & Lui, Y 2001, ‘Rotation, scale, and translation resilient wa- termarking for images’, IEEE Transactions on Im- age Processing, vol. 10, no. 5, pp. 767-782. 10. Kim, H, Baek, Y, Lee, H & Suh, Y 2003, ‘Robust image watermark using Radon transform and bispectrum invariants’, Lecture Notes in Comput- er Science, pp. 145-159. 11. Lee, S, Yoo, CD & Kalker, T 2007, ‘Reversible image watermarking based on integer-to-integer wave- let transform’, IEEE Transactions on Information Forensics and Security, vol. 2, no. 3, pp. 321-330. 12. Calderbank, AR, Daubechies, I, Sweldens, W & Yeo, B 1998, ‘Wavelet transforms that map integers to integers’, Applied and Computational Harmonic Analysis, vol. 5, no. 3, pp. 332-369. 13. Xuan, G, Zhu, J, Chen, J, Shi, YQ, Ni, Z & Su, W 2002, ‘Distortion less data hiding based on integer wavelet transform’, IEEE Electronics Letters., vol. 38, no. 25, pp. 1646-1648. 14. Tsai, HH & Sun, DW 2007, ‘Color image watermark extraction based on support vector machines’ Journal of Information Sciences, vol. 177, no. 2, pp. 550-569. 15. Fu, YG & Shen, RM 2008, ‘Color image watermark- ing scheme based on linear discriminant analysis. Computer Standard & Interfaces’,vol. 30, no. 3, pp. 115-120. 16. Niu, PP, Wang, XY, Yang, YP & Ming, Y 2011, ‘A novel color image watermarking scheme in non- sampled contourlet-domain’, Journal of Expert Systems with Applications, vol. 38, no. 3, pp. 2081- 2098 . Arrived: 31. 03. 2017 Accepted: 12. 06. 2017 A.Amsaveni et al; Informacije Midem, Vol. 47, No. 2(2017), 91 – 99 100 101 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 101 – 111 Synthesis of Silicon Carbide Nanowhiskers by Microwave Heating: Effect of Heating Temperature Suhaimi Mat Kahar1, Voon Chun Hong1, Lee Chang Chuan2, Subash C B Gopinath1,3, Mohd Khairuddin Md Arshad1, Lim Bee Ying4, Foo Kai Loong1, Uda Hashim1, Yarub Al-Douri5 1Institude of Nanoelectronic Engineering, Universiti Malaysia Perlis, Perlis, Malaysia 2School of Manufacturing Engineering, Universiti Malaysia Perlis, Perlis, Malaysia 3School of Bioprocess Engineering, Universiti Malaysia Perlis, Perlis, Malaysia 4School of Materials Engineering, Universiti Malaysia Perlis, Perlis, Malaysia. 5Physics Department, Faculty of Sciences, University Sidi-Bel-Abbes, 22000, Algeria. Abstract: Silicon carbide (SiC) is an attractive material for its excellent properties such as wide band gap, high chemical stability and thermal conductivity. The conventional method for the preparation of SiC is Acheson process, a time and energy consuming process. In this article, demonstration of SiC nanowhiskers synthesis has been done by using microwave heating. Silica and graphite in the ratio 1:3 were mixed in ultrasonic bath, dried on hot plate and cold pressed uniaxially into a pellet die. The pellets were heated by using laboratory microwaves furnace at 1350ºC, 1400ºC and 1450ºC with heating rate of 20oC/min and soaked for 40 minutes. Different characterizations and testing were done to study the effect of heating temperature on the synthesis of SiC nanowhiskers. 1400oC is proved to be the most suitable tempearture for the synthesis of SiC nanowhiskers. β-SiC appeared as the only phase in the x-ray diffraction pattern of SiC nanowhiskers formed at 1400˚C with no traces of raw materials. Field emission scanning electron microscopy confirmed the presence of only a negligible amount of graphite or silica in SiC nanowhiskers synthesized at 1400oC. Furthermore, energy dispersive x-ray spectroscopy analysis revealed that only elemental C and Si were present in SiC nanowhiskers synthesized at 1400oC. Meanwhile, photoluminescence spectrum indicated the presence of single β-SiC peak at 440 nm which is associated with band gap of 2.8 eV. Single absorption bands of Si-C bond were detected at 803.5 cm-1 in fourier transform infrared spectrum. SiCNWs produced in this study at 1400oC has good thermal stability with 6% of weight loss, indicates its potentiality for high temperature electronics. Keywords: Microwave heating; Silicon carbide nanowhiskers; Synthesis; Graphite; Silica Sinteza nanodlačic iz silicijevega karbida z mikrovalovnim segrevanjem: Vpliv temperature gretja Izvleček: Silicijev karbid (SiC) je zelo zanimiv material zaradi svojih odličnih lastnosti, kot je široka energijska reža, kemijska stabilnost in termična prevodnost. Konvencionalen je SiC pridobiva z Acheson-ovim procesom, ki p aje energijsko in časovno izredno potraten. V članku predstavljamo sintezo nanodlačic SiC s pomočjo gretja z mikovalovi. V ultrasonični kopeli je bil apripravljena mešanica silicija in grafita v razmerju 1:3, nato posušena na vroči plošči in hladno stisnjena matrico peleta. Peleti so bili nato segrevani z laboratorisko mikrovalovno pečiso pri 1350 oC, 1400 oC in 1450 oC s hitrostjo gretja 20 °C/min in trajanjem 40 min. Opravljene so bile različne karakterizacije nanidlačic. Izkazalo se je, da je temperature 1400 °C najprimernejša za izdelavo nanodlačic. V vzorcu sipanja x žarkov se izkazalo, da nanodlačice vsebujejo le β-SiC brez ostankov surovega materiala. Elektronska mikroskopija je potrdila prisotnost le zanemarljivega dela silicija in grafita. Fotoluminiscenca je nakazala le eden vrh signala pri 440 nm kar je v skladu z energijsko režo 2.8 eV. Absorpcijski pas je bil zaznan pri 803.5 cm-1 v fourierjevi transformaciji infrardečega spektra. Izdelane nanodlačice so pokazale dobro termično stabilnost z 6 % izgubo teže, kar pomeni, da predstavljajo zanimiv material za visokotemperaturno elektroniko. Ključne besede: mikrovalovno gretje; nanodlačice iz silicijevega karbida; sinteza; grafit; silicij * Corresponding Author’s e-mail: chvoon@unimap.edu.my 102 S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 1 Introduction Silicon carbide (SiC) is a very attractive semiconductor due to its excellent properties such as high hardness, good flexibility, high thermal conductivity, high ther- mal stability, excellent chemical stability and large band gap. Because of these attractive properties, it possesses great potential for industrial and engineering applica- tions such as abrasives [1, 2], high power electronics [3, 4], harsh environment electronics and composite reinforcements [5, 6]. SiC nanomaterials such as SiC nanocrystals and nanowhiskers have many potential electronic applications. For example, SiC nanocrystals (NCs) exhibit photoluminescence in the near-UV to the visible blue spectral region and making them attractive candidates for the fabrication of light-emitting devices [7]. Moreover, several field emission measurements on the SiCNWs suggested that SiCNWs are potential candidates for the cold cathode field emission device (FED) because of their unique electrical, chemical, and mechanical properties [1]. SiC has been produced by several methods, however, most SiC are produced now-a-days using the Acheson process [2]. This process produces SiC by heating mixture of  quartz  sand and powdered coke (carbon-based material) in an iron bowl using voltages 50,000 V for 20 hours at temperatures around 2200–2400˚C. The drawbacks of this industrial production process include high energy consumption and the product has low purity. Moreover, this process is time consuming; therefore several alternative meth- ods have been previously reported for SiC synthesis. Most commonly used methods for SiC synthesis are carbon thermal reduction [9], physical evaporation, sol-gel process [10, 11] and chemical vapor deposition [12]. However, there are still some drawbacks that limit the wide applications of these methods, such as high energy consumption, long processing time and exten- sive chemical usage, although these processes can suc- cessfully synthesize SiC. Recently, researchers have applied microwave heating for the synthesis of inorganic materials [13-16]. From mid-1980s until 2007, hundreds of papers have been published regarding the applications of microwaves in chemical synthesis [16-21]. Development of new routes for the synthesis of inorganic material is an in- tegral aspect of materials chemistry. The development of alternative synthesis methods is a continuing need for fast and energy-efficient techniques. Microwaves are electromagnetic radiation, whose wavelengths lie in the range of 1 mm to 1 m [16]. Microwave syn- thesis has emerged in recent years as a new method to synthesize a variety of materials that has shown significant advantages against conventional synthesis procedures. Microwaves can volumetrically heat ma- terials and give sudden increase in the temperature of the material comparing to conventional heating processes that rely on external radiant energy to heat materials by mode of conduction, convection and ra- diation. Microwave heating is a process in which the materials couple with microwave, absorb the electro- magnetic energy volumetrically, and transform into heat [15]. The carbonaceous materials are among the most sensitive to microwaves irradiation [23]. This is due to the fact that carbon based materials generate heat from the motion of electron through joule heating within the grain of carbon when exposed to microwave irradiation, although carbon based materials have no freely-rotatable dipoles [23]. Materials scientists have identified several advantages of microwave processing of ceramics such as economical, rapid heating, large scale production and reduced cracking and thermal stress [14]. Other than that, Mingos et al. [12] proposed that synthesis of inorganic material using microwave heating can enhance the mechanical properties of the material since the sintering time is generally shortened and thus reduced the possibility of secondary crystal- lization. Silicon carbide nanowhisker (SiCNW) is a silicon car- bide 1-D nanostructure in whisker/needle form. One- dimensional silicon carbide (1D SiC) nanomaterials have shown unusual properties such as extremely high strength, good flexibility and fracture toughness, lead to many potential applications such as sensors, field emitting diodes and solar cells [1, 24]. 1-D nanostruc- ture is also expected to play an important role as both interconnect and as functional units in fabrication of electronics, optoelectronics, electrochemical and elec- tromechanical devices at nanoscale dimensions [1, 25, 26]. In particular, β-SiC nanowhiskers, with an energy band gap of 2.39 eV and relatively high electron mo- bility would be a suitable material for applications in nanoelectronic devices. In this study, microwave heating was used to synthe- size SiCNWs from the mixture of graphite and silica since it is generally faster, cleaner, and more economi- cal than the conventional methods. The effect of heat- ing temperature was studied to determine the most suitable temperature for the synthesis of SiCNWs from silica and graphite. Previously, several researches have studied the effect of heating temperature on the syn- thesis of silicon carbide from silica and carbon-based starting materials. For examples, Wang et al. [27] stud- ied the synthesis of SiC whiskers on graphitic layers us- ing expanded graphite (EG) by silicon vapor deposition without catalyst at temperature ranged from 1000 to 1400˚C. Wang et al. [27] found that the amount of β-SiC on graphite increases with the temperature and the largest amount of β-SiC formed at 1400˚C. Other than that, Jin Li et al. [28] have synthesized nanostructured 103 SiC particles and whiskers from rice husk by microwave heating at temperature ranged from 1100˚C to 1500˚C. They found that 1500 ˚C is the most ideal temperature for the synthesis of β-SiC. Therefore, heating tempera- ture is believed to have significant effect on the quality and purity of the end products during the synthesis of SiCNWs. To the best of our knowledge, no study on the effect of heating temperature on the microwave syn- thesis of SiCNWs from graphite and silica was reported. Thus, in this study, the effect of heating temperature on the morphology, composition, optical properties and purity was studied and presented. 2 Material and Methods 2.1 Sample preparation Silica (particle size ≤ 50 µm) and extra pure fine graphi- te powder (particle size ≤ 50 µm) were used as starting material. Mixture of silica and graphite in molar ratio of 1:3 with total of 1 gram was acquired. Ethanol was used as liquid medium to mix the raw material. Ultra- sonic mixing bath was used as the external mean to generate vibration in the ethanol for the homogene- ous mixing of the raw materials. The mixtures were then dried using hot plate to vaporize the ethanol. Be- fore subjecting to microwave heating, the mixture was compressed to become pellet. The process of making pellet is essential to separate the mixture of SiO2 and graphite from the graphite powder placed around the pellet inside the crucible . The pressure that applied to the mixture during the compression process was 312.4 MPa to ensure mixtures were fully compressed. 2.2 Synthesis of SiCNWs by microwave heating Microwave heating was performed in Synotherm mi- crowave sintering furnace (MW-L0316V) with multi- mode cavity in which 2.45 GHz microwave radiation was bring out through a waveguide. The pellet was placed in silica crucible and it was placed in microwave cavity as shown in Fig. 1. Silica sand was used as heat insulator to prevent heat lost. SiC suscep- tor functioned as microwave absorber to absorb and convert electromagnetic energy to heat because SiC susceptor is a good microwave absorbing material. The pellets were heated to different temperatures of 1350˚C, 1400˚C and 1450˚C with heating rate of 20°C/ min and soaked for 40 minutes. The synthesis was per- formed under argon atmosphere. 2.3 Characterization of SiC nanowhiskers After the microwave heating was conducted, samples were characterized by using x-ray diffraction (XRD), field emission scanning electron microscopy (FESEM), energy-dispersive x-ray spectroscopy (EDX), photolu- minescence spectroscopy (PL), fourier transform infra- red spectroscopy (FTIR) and thermo-gravimetric analy- sis (TGA). The morphologies of samples were observed by using FESEM model Nova Nano 450 at magnification 200K and accelerating voltage of 5 kV while EDX (EDX OX- FORD FM29142) was used to determine the elemen- tal composition of the specimens. The samples in the powder form were added into ethanol and ultrasoni- cated for homogeneous dispersion. The dispersions were then dropped on the substrate and the substrate was heated by using hot plate to evaporate the etha- nol. The samples on the substrate were then subjected to characterization using FESEM and EDX. The built in software for OXFORD FM29142 enables automatic cor- rection and robust spectrum processing that works in non-flat sample measurement with no need for any background fitting adjustment. Meanwhile, XRD Sie- mens Diffractometer Model D-5000 using Cu Kα radia- tion source in θ/2θ mode was used to investigate the composition of specimens. Measurements were made with fast duration scan (1s) and small step size (0.02°). Optical properties of SiCNWs synthesized from the mix- tures were identified by using the photoluminescence spectroscopy (PL FL3-11 J81040) with xenon lamp at 400 watt and excitation wavelength at 360 nm and re- corded from wavelength of 300 nm to 650 nm while FTIR (FTIR MAGNA550 kBr) was used to scan the sam- ples from 500 to 4000 nm-1 with spectrum resolution of 4 cm-1. Purity of SiCNWs was evaluated indirectly by using Perkin-Elmer Pyris 6 TGA analyzer. Samples about 10 mg were heated from 30 to 1300°C with the heating rate of 10 °C/min in atmospheric air to investigate the purity of the as synthesized SiCNWs.Figure 1: Setup for sample preparation inside the mi- crowave cavity. S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 104 3 Result and Discussions 3.1 Characterization of SiCNWs using FESEM Fig. 2 shows the FESEM images of SiO2 and graphite subjecting to microwaves heating at different tempera- tures. It can be seen that the heating temperature of SiO2 and graphite significantly influenced the synthesis of SiCNWs. Fig. 2 (a) shows the mixture SiO2 and graph- ite after subjecting to microwaves heating at 1350˚C. It can be observed that only a small amount of nanow- hiskers were formed around particles that is believed to be unreacted SiO2 and graphite. Besides, the nano- whiskers that formed from the reaction between silica and graphite were not fully grown as indicated by red circles in Fig. 2 (a). It is believed that this might be due to the fact that heating temperature at 1350˚C was not sufficiently high to enable the full reaction between graphite and SiO2 for the complete formation of SiC na- nowhisker. Similar observation was reported by Wang et.al that hybridized silicon carbide (SiC) whiskers on graphitic layers in expanded graphite (EG) by silicon va- por deposition without catalyst. They reported that for the synthesis conducted at low heating temperature (1100˚C to 1300˚C), small amount of SiC was produced due to incomplete reactions [27]. Fig. 2 (b) shows the FESEM image of SiCNWs formed from the mixture of SiO2 and graphite that was subjec- ted to microwaves heating at 1400˚C. SiC in the form of nanowhiskers can be observed clearly. The diame- ters of SiCNWs are uniform along the length of the na- nowhiskers. Only small amount particles of graphite or silica were observed, such that almost all graphite and silica were converted to SiCNWs. Wang et.al [27] have also reported similar result in which large amount of SiC in the form of nanowhiskers were formed at 1400˚C. The diameters of the nanowhisker were measured by using ImageJ version 1.48 and they were ranged bet- ween 70 nm and 100 nm. Fig. 2 (c) shows the SiC whiskers formed by microwaves heating mixture of SiO2 and graphite at 1450˚C. It can be observed that the amount SiC whiskers are similar comparing to those formed at 1400˚C in Fig. 2 (b). Lar- ge amount of SiCNWs can be observed. The diameter of the SiCNWs formed were measured and are ranged from 70 nm to 120 nm which are similar to SiCNWs in Fig. 2 (b). The diameter for the SiCNWs formed at 1450 ˚C are slightly larger and this might due to the increase of heating temperature. Similar result were also obtai- ned by Wang et al [27]. They reported that the diameter of β-SiC nanowhiskers in the specimen treated at 1400 ˚C were larger than those treated below 1400 ˚C [27]. With the increasing heating temperature, the rate of re- action of SiC was increased. This caused higher SiC for- mation rate on the intially formed SiCNWs during the heating process, and led to larger diameter of SiCNWs at higher temperature. 3.2 Characterization of SiCNWs using XRD XRD patterns of SiCNWs synthesized from mixture of SiO2 and graphite at different temperatures are shown in Fig. 3. For SiCNWs synthesized at 1350˚C as shown in Fig. 3 (a), small peaks corresponding to SiO2 at 2θ of 22.3˚ associated with plane (100) of SiO2 (JCPDS card 01-089-3434) was observed. A peak of carbon phase was also observed at 27˚ corresponding to plane (002) of graphite (JCPDS card 03-065-6212). Figure 2: FESEM images of SiCNWs synthesized by microwave heating of mixture of SiO2 and graphite at heating temperatures of a) 1350˚C b) 1400˚C and c) 1450˚C. S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 105 Generally, the presence of carbon and SiO2 is due to the presence of unreacted graphite and silica. Peaks of ß- SiC (111) and (220) were also observed at 2θ of 36˚and 61˚. This indicated that reaction of SiO2 and graphite to form SiCNWs at 1350˚C was incomplete. This obser- vation is in good agreement with the corresponding FESEM image in Fig. 2 (a) where only small amount of SiCNWs along with graphite and silica were observed. Fig. 3 (b) shows the XRD pattern of SiCNWs synthesized from mixture of SiO2 and graphite at 1400˚C. Three peaks that were corresponded to (111), (220) and (311) crystal planes of cubic ß-SiC (JCPDS card 074-2307) were observed at 2θ of 36˚, 61˚ and 72.5˚. Bin li et.al. [10] also reported similar result in which diffraction peaks of β-SiC at 2θ of 35.8˚, 60˚ and 71.8˚ corresponding to (111), (220) and (311) cubic reflections were obtained. No signal of either SiO2 or carbon was deteced in this XRD pattern. It can be concluded that mixture of SiO2 and graphite were converted completely to SiCNWs when 1400˚C was used such that the amounts of the raw materials were too small to be detected by XRD. This result is in good agreement with the FESEM im- ages in Fig. 2 (b) in which only SiCNWs were observed. For SiCNWs synthesized from mixture of SiO2 and graphite at 1450˚C, as in Fig. 3 (c), peaks corresponded to ß-SiC as major phase appeared at 2θ values of 36˚, 61˚ and 72.5˚, respectively. The relative intensities of these peaks were similar compared to Fig. 3 (b). In good agreement with FESEM image of Fig 2 (c), SiCNWs were observed with only small amount of graphite or silica particles. Figure 3: XRD patterns of SiCNW synthesized by micro- wave heating of mixture of SiO2 and graphite at heat- ing temperatures of a) 1350˚C, b) 1400˚C and c) 1450˚C. 3.3 Characterization of SiCNWs using EDX Fig. 4 shows the EDX spectra of the SiCNWs synthesized by microwave heating at 1350˚C, 1400˚C and 1450˚C. Qualitative analysis was conducted to identify the ele- ments that are present in the as synthesized SiCNWs. EDX spectra with high accuracy after subjecting to au- tomatic correction and robust spectrum processing us- ing the built in software were obtained. Fig. 4 (a) shows the EDX peak of SiCNWs that synthesized from mixture of silica and graphite at 1350˚C. From the peak, 3 ele- ments were detected which are Si, C and O. O element is corresponded to the presence of silica in the end pro- duct. This indicated that silica was not fully reacted in this process, and this is in good agreement with XRD result in Fig. 3 and FESEM images in Fig. 2 (a). Similar observation was reported by Quah et al. [29] and they attributed the presence of O element in the EDX spec- trum to the presence of unreacted SiO2 in final product. For EDX spectra of SiCNWs synthesized at 1400˚C and 1450˚C, peaks corresponded to Si and C elements were observed. This indicated that mixture of SiO2 and graphite reacted completely to form SiCNWs at 1400˚C and 1450˚C. Figure 4: EDX spectra of SiCNWs synthesized by micro- wave heating of mixture of SiO2 and graphite at heating temperatures of a) 1350˚C, b) 1400˚C and c) 1450˚C. 3.4 Characterization of SiCNWs using FTIR FTIR transmission spectra of SiCNWs synthesized from mixture of graphite and SiO2 at different heating tem- peratures are shown in Fig. 5. From the spectra, it can be concluded that SiCNWs were successfully synthe- sized at all temperatures since FTIR peaks correspond- ed to Si-C stretching bond were present at around 800 cm-1 in all FTIR spectra of SiCNWs. However, as in Fig. 5 a), it can be observed that SICNWs at heating tempera- ture of 1350˚C has absorption band of relatively low intensity at 801.7 cm-1 that indicates only small amount of SiCNWs were formed. FTIR peak corresponded to S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 106 stretching bond of Si-O bonding group was also ob- served at 1097.28 cm-1. The presence of this absorption band indicated the presence of unreacted SiO2. Similar absorption bands were reported by Zhao et al. [30] and Rajarao et al. [31]. Zhao et al. [30] obtained such ab- sorption peak at 1080cm-1 and they suggested that this peak was associated with the Si–O–Si bond of mesopo- rous silica. Rajarao et al. [31] also reported absorption band at 1045cm-1 and this peak was attributed to Si- O-Si bond. Absorption bands at 1640.2 cm-1 were also observed in Fig. 5 (a) due to the presence of C=C bonds of graphite [32]. The presence of absorption bands of both SiO2 and graphite indicate that 1350˚C was insuf- ficient to enable complete reaction between SiO2 and graphite and thus some of the SiO2 and graphite were left unreacted. For FTIR spectrum of SiCNWs formed at 1400˚C, only one peak was observed at 803.5cm-1 corresponding to the presence of Si–C bond that indicated the successful synthesis of SiCNWs. This result is in good agreement with the XRD pattern of SiCNWs synthesized at 1400˚C in Fig. 3 (b) which indicates the presence of single phase β-SiC and thus denotes complete conversion of graphite to SiCNWs. FTIR spectrum of SiCNWs synthe- sized at 1450˚C as shown in Fig. 5 (c) also revealed the presence of single phase β-SiC due to the presence of absorption peak of Si-C stretching bonds centered at 804.6 cm-1. Figure 5: FTIR spectra of SiCNWs synthesized by micro- wave heating of mixture of SiO2 and graphite at heat- ing temperatures of a) 1350˚C, b) 1400˚C and c) 1450˚C. 3.5 Characterization of SiCNWs using PL PL spectra of SiCNWs synthesized at different heating temperatures were showed in Fig. 6. Fig. 6 shows peaks of SiCNWs at 440 nm (2.8 eV) in all spectra. The peaks are obviously blue-shifted in comparison with the band gap of 3C-SiC (2.39 eV). The blue shift of the PL peak of 3C-SiC nanomaterials has been reported by se- veral researchers [33,34,35]. For example, the peak at 418 nm for 3C-SiC nanobelts has been reported by Wu et al [36]. They proposed that the location of this peak depends on the nanostructure, morphology and size of 3C-SiC materials. The collective influence of size con- finement effect and defects lead to the blue shift of the peak. Thus, the peak emission appeared around 440 nm may be due to size confinement effect and defects. In Fig. 6 (a), PL spectrum of SICNWs synthesized from blend of SiO2 and graphite at 1350˚C shows the pres- ence of PL peak attributed to oxygen discrepancy in SiO2 and carbon at wavelength about 380 and 620 nm which corresponded to band gap of 3.2 eV and 2.0 eV, respectively. This PL spectrum result is in good agree- ment with the XRD result of SICNWs synthesized at 1350˚C which shows the presence of XRD peak corre- sponded to SiO2 and carbon. Nandanwar et al [37] re- ported the characterization of SiO2 nanoparticles and also reported PL peak of pure SiO2 at 381.8 nm. Fig. 6 (b) and (c) shows that in the PL spectra of SiC- NWs synthesized at 1400˚C and 1450˚C, only one peak appeared at 425nm and this peak is corresponded to β-SiC. This indicated that only SiC is present in the SiC- NWs synthesized at heating temperature 1400˚C and 1450˚C. This result is in good consistent with the XRD result in Fig. 3 (b) and (c) that graphite and SiO2 react completely to form single phase SiCNWs. Figure 6: PL spectrum of SiCNWs synthesized by micro- wave heating of mixture of SiO2 and graphite at heat- ing temperatures of a) 1350˚C, b) 1400˚C and c) 1450˚C. 3.6 Thermal Gravimetric Analysis of SiCNWs Thermal Gravimetric Analysis (TGA) curves of SiCNWs synthesized at different heating temperatures are pre- sented in Fig. 7. TGA was conducted to evaluate indi- rectly the quantity of SiCNWs. For SiCNWs synthesized S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 107 at 1450˚C in Fig. 7 (b), the weight loss started at 700ºC with a total of 7 wt %. Similar weight loss occurred for SiCNWs synthesized at 1400˚C with 6 wt% as shown in Fig. 7 (a). These small weight losses of SiCNWs can be attributed to the oxidation of small amount of un- reacted carbon and loss of moisture. The presence of moisture may happen during the handling of sample since sample is powder which can easily absorb mois- ture. Corriu et al. [38] proposed that the weight loss which occurred at 450 ºC to 750 ºC domain was attrib- uted to the air oxidation of the carbon. This indicated graphite was almost fully converted to SiCNWs with only very small amount of unreacted carbon for both SiCNWs synthesized at 1400 ˚C and 1450 ˚C. This result is in good agreement with XRD result in Fig. 3 a) and b) in which carbon and silica were too little to be detect- ed. This high resistance toward oxidation for SiCNWs synthesized at 1400˚C and 1450 ˚C is attributed to the formation of pure SiCNWs. TGA curves show no weight loss at temperature higher than 800 ºC for both SiCNWs formed at 1400 and 1450 ˚C, indicating the remaining residue were SiCNWs. Fig. 7 c) shows that for SiCNWs synthesized at 1350˚C, a total of 45% of weight loss is observed starting at 700 to 950 ºC and this weight loss was attributed to the oxidation of unreacted carbon in SiCNWs. These results are in good agreement with the XRD result displayed in Fig. 3 which showed the presence of peak of unre- acted carbon in SiCNWs synthesized at 1350˚C. This result demonstrated that SiCNWs produced at 1400˚C and above has relatively high purity and good thermal stability. Figure 7: TGA curves of SiCNWs synthesized by micro- wave heating of mixture of SiO2 and graphite at heat- ing temperatures of a) 1400˚C, b) 1450˚C and c) 1350˚C. 3.7 Mechanism of Synthesis of SiCNWs by Microwaves Heating For this research, the interactions between carbon- based material (graphite) and microwave irradiation are important to generate heat thus give many advan- tages in many aspects to synthesize SiCNWs. Since the quart materials are not sensitive to microwave, in this study we proposed that the heat from graphite (car- bon based material) are transferred to silica via external means such as conduction, convection and radiation to assist the heating of silica. Lin He et al. [22] proposed that silica is an inorganic material that almost cannot react to microwave and the reaction are not as effective as carbon based material from the calorimetric study based on these materials. The homogeneous mixture between silica and graphite therefore significantly af- fects the uniformity of temperature increase for both materials. For this reason, ultrasonic mixing of graphite and silica using ethanol as medium provided homoge- neous mixing. The mechanism of microwave heating varies according to the interaction between the microwaves and target materials. Dielectric heating occurs when dielectric materials such as graphite interact with microwaves. Electric field component of electromagnetic interact with charged particles (electrons) of carbon causes the material to generate heat. Graphite is known as carbon-based material that contains charged particles which are free to move in a delimited region of the ma- terial [23, 39]. When electromagnetic field is subjected to the material such as graphite, current traveling in phase with the electromagnetic field is induced. The electron from the carbon material cannot couple to the changes of phase in the electric field and causes energy to dissipate in the form of heat. Fig. 8 shows the mechanism of dielectric heating that based on motion of electrons from carbon material to generate heat. Motion of electron from carbon through joule heating within the grain generates heat. This reaction is called Maxwell-Wagner effect and it is significantly different from the reaction between electromagnetic wave and Figure 8: Interaction of microwave with graphite leads to dielectric heating of graphite. S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 108 polar liquid such as water that heat up due to vibration of molecules [40,23]. The overall reaction of formation for SiC through carbo- thermal reduction is generally written as [41]: SiO2(s) + 3C(s) = SiC(s) + 2CO(g) (1) AG° = 598.18 - 0.3278 T (kJ) There are multiple reactions between silica and graph- ite before the formation of SiCNWs. First reaction is the solid-solid reaction between silica and graphite caus- ing the carbothermal reduction of silica by graphite to form SiO and CO gases by following reaction [41]: SiO2(s) + C(s) → SiO(g) + CO(g) (2) AG° = 668.07 - 0.3288 T (kJ) The vapour-solid (VS) mechanism was suggested to ex- plain the formation of SiCNWs. From reaction (3), SiO gas reacts with C to produce SiC nuclei as follow [42]: SiO(s) + 2C(s) → SiC(g) + CO(g) (3) AG° = -78.89 + 0.0010 T (kJ) Cetinkaya et.al stated for VS mechanism, Si-containing vapors such as Si gas or SiO gas are believed to react with CO gas or C solid to form SiC nuclei [43]. Thus, SiC particles from reaction (3) are believed to serve as nu- cleation sites for VS mechanism to occur. SiO(g) + 3CO(g)→SiCNWs (s) + 2CO2(g) (4) From reaction (4), the VS mechanism occurred when SiO vapour and CO vapour deposited at the tip of SiC nuclei that formed from reaction (3). Fig. 9 summarizes the overall reaction between graphite and silica for the formation of SiCNWs. The nanowhisker grows along the directions of the least stable plane and forms SiCNWs, as in Fig. 10. J. Wei et al. [44] and Dehghanzadeh [45] et al. have proposed that the nanowire growth might be at- tributed to the reaction between SiO and carbon gases. The effect of temperature for synthesis of SiCNWs has been studied by thermodynamic calculation. The Gibbs free energy for overall reaction in reaction (2) de- creases with temperature thus denoted that the reac- tion is non-spontaneous reaction. Based on Gibbs free energy, reaction (2) is highly endothermic and the reac- tion is favorable to occur as the temperature increases [46]. Synthesis of SiC is basically dependent on the for- mation of SiO gas. Lee et al. also proposed that SiC is synthesized through the formation of intermediate SiO [47]. Figure 10: Schematic of SiCNWs growth from graphite and silica by microwaves heating a) Mixture of graph- ite and silica b) Exposing mixture of SiO2 and graphite to microwave irradiation until 1400˚C c) Formation of SiO gas, Co gas and SiC nucleus after exposed to mi- crowave irradiation at high temperature d) Formation of SiCNW. The Gibbs free energy for the reaction between SiO gas and carbon as in reaction (3) is negative and thus the reaction is spontaneous, regardless of the tempera- ture. Thus, the overall SiC formation is defined by SiO formation in reaction (2), since the Gibbs free energy decreases significantly on temperature. Furthermore, formation of SiO gas is the rate determining step for the overall reaction of SiC formation [48, 49]. Some re- searchers have studied the rate of reaction to synthesis SiC based on Arrhenius equation [49, 50]. k = Ae–Ea/RT (5) Rate constant and activation energy are calculated based on the Arrhenius equation in equation (5). Kavi- tha et al. [50] studied the synthesis of nano silicon car- bide powder from agricultural waste and calculated the activation energy during the synthesis of SiC using the Arrhenius equation. They reported that with the in- creasing of heating temperature, the activation energy decreased and caused the rate of reaction for the syn- thesis of SiC to increase. This explained the significant effect of temperature on the rate of reaction of the syn- thesis of SiC. Furthermore, for temperature at 1350˚C, it is believed that the partial pressure of SiO gases pro- duced was lower than those produced at temperature above 1400 ˚C. Y. Li et al. [51] reported that the partial Figure 9: Schematic of SiCNWs growth from graphite and silica by microwaves heating. S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 109 pressure of SiO as predominant gas increased with temperature which was originated from the oxidation of silica. Thus, the amount of nucleation sites produced from reaction of SiO gas and carbon (solid) at 1350˚C was expected to be lower comparing to those formed at 1400 ˚C and 1450 ˚C. SiO and CO gases as the prod- uct from reaction at (3) and (4), respectively, react to form SiCNWs but the reaction was incomplete due to the lack of reactant (SiO) at low temperature. This ex- plained the incomplete formation of SiCNWs at 1350˚C when the heating duration was set to 40 minutes. For temperature above 1400 ˚C, as shown by the results, heating duration of 40 minutes was sufficient for com- plete synthesis. 4Conclusions SiC nanowhiskers have been successfully synthesized by microwaves heating of mixture of SiO2 and graph- ite in the ratio of 1:3. The effect of heating temperature during microwave heating was studied. SiCNWs were characterized by using X-ray diffraction (XRD), field emission scanning electron microscopy (FESEM), en- ergy dispersive x-ray spectroscopy (EDX), photolumi- nescence spectroscopy (PL), fourier transform infrared (FTIR) and thermo-gravimetric analysis (TGA). 1400˚C is the most suitable temperature for the synthesis of SiCNWs because of complete reaction between silicon dioxide and graphite resulted in the formation of single phase β-SiC nanowhiskers in nanoscales as proven by the results obtained from characterization and testing. By using 1350 ˚C for the synthesis of SiCNWs, traces of unreacted graphite and SiO2 were detected that indi- cated incomplete conversions of graphite and silica to SiCNWs while synthesis of SiCNWs at 1450 ˚C resulted in SiCNWs with diameter higher than 100 nm. 5 Acknowledgments The authors are grateful to the Department of Higher Education, Ministry of Higher Education, Malaysia for funding this research through the Fundamental Re- search Grant Scheme (FRGS) with the grant number [9003-00441]. The author also would like to acknowl- edge all the team members in Institute of Nano Elec- tronic Engineering (INEE), Universiti Malaysia Perlis (UniMAP) for their guidance and help. 6 References 1. Prakash, J., Venugopalan, R., Tripathi, B. M., Ghosh, S. K., Chakravartty, J. K., & Tyagi, A. K., Chemistry of one dimensional silicon carbide materials: Princi- ple, production, application and future prospects, Progress in Solid State Chemistry, 43(3), 2015, pp. 98–122. 2. Yang, Y., Yang, K., Lin, Z. M., & Li, J. T.. Mechanical- activation-assisted combustion synthesis of SiC, Materials Letters, 61(3), 2007, pp. 671–676. 3. Elyassi, B., Kim, T. W., Sahimi, M., & Tsotsis, T. T., Effect of polystyrene on the morphology and physical properties of silicon carbide nanofibers. Materials Chemistry and Physics, 118(1), 2009, pp. 259–263. 4. Jin, H. B., Li, J. T., Cao, M. S., & Agathopoulos, S. Influence of mechanical activation on combus- tion synthesis of fine silicon carbide (SiC) powder. Powder Technology, 196(2), 2009, pp. 229–232. 5. Krzysztof Górecki, J. Z, The Influence of Diodes and Transistors Made of Silicon and Silicon Car- bide on the Nonisothermal Characteristics of Boost Converters. Informacije MIDEM, 42(3), 2012, pp. 176–184. 6. Sviličić, B., Mastropaolo, E., & Cheung, R. (2013). Two-port piezoelectric silicon carbide MEMS can- tilever resonator. Informacije MIDEM, 43(1), 2013, pp. 22–26. 7. Zhu, J., Wu, H., Chen, H. T., Wu, X. L., & Xiong, X. Tunable violet–blue emission from 3C-SiC na- nowires. Physics Letters A, 373(18–19), 2009, pp. 1697–1700. 8. Li, B., Song, Y.-C., Zhang, C.-R., & Yu, J.-S. Synthe- sis and characterization of nanostructured silicon carbide crystal whiskers by sol–gel process and carbothermal reduction. Ceramics International, 40(8), 2014, pp. 12613–12616. 9. Raman, V., Bahl, O. P., & Dhawan, U. Synthesis of silicon carbide through the sol-gel process from different precursors. Journal of Materials Science, 30, 1995, pp. 2686–2693 10. Lin, M., Loh, K. P., Boothroyd, C., & Du, A. Y. Nano- cantilevers made of bent silicon carbide nanow- ire-in-silicon oxide nanocones. Applied Physics Letters, 85(22), 2004, pp. 5388–5390. 11. Agrawal, D. K. Microwave processing of ceramics. Current Opinion in Solid State and Materials Sci- ence, 3(5), 1998, pp. 480–485. 12. Mingos, D. M. P. Microwave syntheses of inorgan- ic materials. Advanced Materials, 5(11), 1993, pp. 857–859. 13. Oghbaei, M., & Mirzaee, O. Microwave versus con- ventional sintering: A review of fundamentals, ad- vantages and applications. Journal of Alloys and Compounds, 494(1–2), 2010, pp. 175–189. S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 110 14. Rao, K. J., Vaidhyanathan, B., Ganguli, M., & Ram- akrishnan, P. a. Synthesis of Inorganic Solids Using Microwaves. Chemistry of Materials, 11, 1999, pp. 882–895. 15. Millos, C. J., Gavin Whittaker, A., & Brechin, E. K. Micro- wave heating - A new synthetic tool for cluster syn- thesis. Polyhedron, 26(9–11), 2007, pp. 1927–1933. 16. Belmonte, T., Henrion, G., Cardoso, R. P., Noel, C., Arnoult, G., & Kosior, F. (2008). Microwave Plasmas at Atmospheric Pressure: New Theoretical Devel- opments and Applications in Surface Science. In- formacije MIDEM, 38(4), 2008, pp. 272–276. 17. Vesel, A., Mozetic, M., & Balat-Pichelin, M. (2011). Absorption of microwave power in nitrogen plasma at moderately low pressure. Informacije MIDEM, 41(1), 2011, pp. 18–21. 18. Kepenienė, V., Tamašiūnaitė, L. T., Ienė, J. J., Vaičiūnienė, J., Kondrotas, R., & As, V. P. Platinum- Niobium (V) Oxide / Carbon Nanocomposites Pre- pared By Microwave Synthesis For Ethanol Oxida- tion. Materials Science, 22(2), 2016: pp 243–248. 19. Meng, F., Wang, X., Jiang, D., & Ma, R. Microwave Absorbing and Magnetic Properties of the Polyani- line-. Materials Science, 22(3), 2016: pp 354–357. 20. Zong, B., Wu, Y., Phuoc, N. N., Ho, P., Qiu, J., Yang, Y., Han, G. Quick fabrication of appropriate morphology and composition CoFe films with desirable microwave properties. Int. J. Mater. Res, (110), pp. 1077–1085. 21. Xing, R., Guo, J., Miao, C., Liu, S., & Pan, H. Fabri- cation of protein-coated CdS nanocrystals via microwave-assisted hydrothermal method. Jour- nal of Experimental Nanoscience, 9(6), 2014, pp. 582–587. 22. He, C. L., & Chen, Y. Q. Calorimetry Study of Micro- wave Absorption of Some Solid Materials. Journal of Microwave Power and Electromagnetic Energy, 47(4), 2013, pp. 251–261. 23. Kim, T., Lee, J., & Lee, K.-H. Microwave heating of carbon-based solid materials. Carbon letters, 15(1), 2014, pp. 15–24. 24. Koguchi, M., Kakibayashi, H., Yazawa, M., Hiruma, K., & Katsuyama, T. Crystal Structure Change of GaAs and InAs Whiskers from Zinc-Blende to Wurtzite Type. Japanese Journal of Applied Phys- ics, 31(7R), 1992, pp. 2061. 25. Martin, A., Schopf, C., Pescaglini, A., O’Riordan, A., & Iacopino, D. Synthesis, optical properties and self-assembly of gold nanorods. Journal of Experi- mental Nanoscience, 7(6), 2012, pp. 688–702. 26. Fung, M. K., Ng, A. M. C., Djurišić, A. B., Chan, W. K., & Wang, H. Preparation of 8-hydroxyquinoline wires by decomposition of tris(8-hydroxyquino- line) aluminium. Journal of Experimental Nano- science, 7(5), 2012, pp. 578–585. 27. Wang, Q., Li, Y., Jin, S., & Sang, S. Catalyst-free hybridization of silicon carbide whiskers and ex- panded graphite by vapor deposition method. Ceramics International, 41(10), 2015, pp. 14359– 14366. 28. Li, J., Shirai, T., & Fuji, M. Rapid carbothermal syn- thesis of nanostructured silicon carbide particles and whiskers from rice husk by microwave heat- ing method. Advanced Powder Technology, 24(5), 2013, pp. 838–843. 29. Quah, H.-J., Cheong, K. Y., & Lockman, Z. Stimula- tion of silicon carbide nanotubes formation using different ratios of carbon nanotubes to silicon di- oxide nanopowders. Journal of Alloys and Com- pounds, 475(1–2), 2008: pp 565–568. 30. Zhao, B., Zhang, H., Tao, H., Tan, Z., Jiao, Z., & Wu, M. Low temperature synthesis of mesoporous sili- con carbide via magnesiothermic reduction. Ma- terials Letters, 65(11), 2011: pp 1552–1555. 31. Rajarao, R., Ferreira, R., Sadi, S. H. F., Khanna, R., & Sahajwalla, V. Synthesis of silicon carbide na- noparticles by using electronic waste as a carbon source. Materials Letters, 120, 2014: pp 65–68. 32. Coates, J. Interpretation of Infrared Spectra, A Practical Approach. Encyclopedia of Analytical Chemistry, 2000: pp 10815–10837. 33. Chiu, S.-C., & Li, Y.-Y. SiC nanowires in large quan- tities: Synthesis, band gap characterization, and photoluminescence properties. Journal of Crystal Growth, 311(4), 2009: pp 1036–1041. 34. Li, Y., Chen, C., Li, J.-T., Yang, Y., & Lin, Z.-M. Surface charges and optical characteristic of colloidal cu- bic SiC nanocrystals. Nanoscale research letters, 6(1), 2011: pp 454. 35. y. Verd Francisco Arredondo. Cerámica y vidrio, 279(c), 1969: pp 279–284 36. RenbingWu, LinglingWu, Guangyi Yang, Yi Pan, Ji- anjun Chen, R. Z. and J. L. Fabrication and photo- luminescence of bicrystalline SiC nanobelts. Jour- nal of Physics D: Applied Physics, 40(12), 2007: pp 3697. 37. Nandawar, R., Purnima, S., & Fozia, Z. H. Synthesis and characterization of SiO2 nanoparticles by sol- gel process. Indian Journal of Science, 1(1), 2012: pp 6–10. 38. Corriu, R. J. P., Gerbier, P., Guérin, C., & Henner, B. Poly[(silylene)diacetylene]/fine metal oxide pow- der dispersions: use as precursors to silicon-based composite ceramics. Journal of Materials Chemis- try, 10(9), 2000: pp 2173–2182. 39. Menéndez, J. a, Abdul Rahman, R., Arenillas, B. F., Y. Fernández, L. Z., E.G.Calvo, & J.M.Bermúdez. Mi- crowave heating processes involving carbon ma- terials. Fuel Processing Technology, 91(1), 2010, pp. 1–8. 40. Stuerga, D. Microwave-Material Interactions and Dielectric Properties, Key Ingredients for Mastery of Chemical Microwave Processes. In Microwaves S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 111 in Organic Synthesis (Vol. 1, pp. 1–61). 2008, pp. Weinheim, Germany: Wiley-VCH Verlag GmbH. 41. Li, X., Zhang, G., Tang, K., Ostrovski, O., & Tronstad, R. synthesis of silicon carbide by carbothermal re- duction of quartz in h 2 -ar gas mixtures, 2015, pp. 548–554. 42. Li, X., Zhang, G., Tang, K., Ostrovski, O., & Tronstad, R. Carbothermal Reduction of Quartz in Different Gas Atmospheres. Metallurgical and Materials Transactions B: Process Metallurgy and Materials Processing Science, 46(3), 2015, pp. 1343–1352. 43. Cetinkaya, S., & Eroglu, S. Chemical vapor deposi- tion of C on SiO2 and subsequent carbothermal reduction for the synthesis of nanocrystalline SiC particles/whiskers. Journal of the European Ce- ramic Society, 31(5), 2011, pp. 869–876. 44. Fu, Q.-G., Li, H.-J., Shi, X.-H., Li, K.-Z., Wei, J., & Hu, Z.-B. Synthesis of silicon carbide nanowires by CVD without using a metallic catalyst. Materials Chemistry and Physics, 100(1), 2006, pp. 108–111. 45. Dehghanzadeh, M., Ataie & Heshmati-Manesh, S. synthesis of nanosize silicon carbide powder by carbothermal reduction of sio2. International Journal of Modern Physics: Conference Series, 5, 2012, pp. 263–269. 46. Weimer, A. W., Nilsen, K. J., Cochran, G. A., & Roach, R. P. Kinetics of carbothermal reduction synthesis of beta silicon carbide. AIChE Journal, 39(3), 1991, pp. 493–503. 47. Lee, J. G., Miller, P. D., & Cutler, I. B. carbothermal re- duction of silica Department of Materials Science and Engineering University of Utah, Salt Lake City, Utah 84112 (U.S.A.), 84112(1), pp. 707–711. 48. Changhong, D., Xianpeng, Z., Jinsong, Z., Yongjin, Y., Lihua, C., & Fei, X. The synthesis of ultrafine SiC powder by the microwave heating technique. Journal of Materials Science, 32(9), 1997, pp. 2469–2472. 49. Dijen, F. K. Van. The Chemistry of the Carbother- mal Synthesis of / 3-SIC : Reaction Mechanism , Reaction Rate and Grain Growth, 7(199 l), 1991, pp. 177–184. 50. Kavitha, N., Balasubramanian, M., & Vashistha, Y. D. Synthesis and Characterization of Nano Silicon Carbide Powder from Agricultural Waste. Transac- tions of the Indian Ceramic Society, 70(3), 2011, pp. 115–118. 51. Li, Y., Wang, Q., Fan, H., Sang, S., Li, Y., & Zhao, L. Synthesis of silicon carbide whiskers using reac- tive graphite as template. Ceramics International, 40(1), 2014, pp. 1481–1488. Arrived: 15. 03. 2017 Accepted: 26. 07. 2017 S. M. Kahar et al; Informacije Midem, Vol. 47, No. 2(2017), 101 – 111 112 113 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 113 – 128 A Compact Radio Telescope for the 21 cm Neutral- Hydrogen Line Tadeja Saje, Matjaž Vidmar Abstract: Thanks to the many independent technological achievements in the recent years, serious radio astronomy is again within reach of amateur astronomers. Observation of the λ=21cm neutral-hydrogen line is relatively straightforward, since the strongest hydrogen clouds in our galaxy achieve an equivalent back-body brightness temperature of around T ≈100K. Yet such observation provides important information about the structure and velocity of our galaxy. In this article the design, construction and calibration of a suitable radio telescope is discussed in detail making efficient use of available hardware and software. Important practical details like radio- interference mitigation in an urban environment are also discussed. Finally, the results obtained with the prototype radio telescope are presented both as the hydrogen spectra in selected directions as well as a 3D map (galactic longitude, galactic latitude and velocity profile) of our Milky Way galaxy visible from our latitude 46° north. Keywords: radio telescope; LNA; band-pass filter; hydrogen line; radio astronomy; feed horn. Radioteleskop za 21 cm vodikovo črto Izvleček: Številni med sabo neodvisni tehnološki dosežki zadnjih nekaj let omogočajo, da je resna radioastronomija ponovno dostopna amaterskim astronomom. Opazovanje spektralne črte atomarnega vodika λ=21cm je razmeroma preprosto, saj spektralna svetlost najmočnejših oblakov vodika v naši galaksiji dosega enakovredno temperaturo sevanja črnega telesa okoli T ≈100K. Hkrati takšno opazovanje daje pomembne informacije o sestavi in hitrosti naše galaksije. V tem prispevku so podrobno opisani načrtovanje, izgradnja in umerjanje primernega radioteleskopa ob učinkoviti uporabi razpoložljive strojne in programske opreme. Opisane so tudi pomembne praktične podrobnosti, kot je izogibanje radijskim motnjam v mestnem okolju. Končno, rezultati meritev z izdelanim prototipom radioteleskopa so prikazani kot spektri atomarnega vodika v izbranih smereh in kot 3D zemljevid (galaktična dolžina, galaktična širina in profil hitrosti) naše galaksije Rimske ceste, kot jo vidimo iz naše zemljepisne širine 46° severno. Ključne besede: radioteleskop; LNA; pasovno prepustno frekvenčno sito; vodikova črta; radioastronomija, žarilec. * Corresponding Author’s e-mail: tadeja.saje@gmail.com 1 Amateur radio astronomy For many thousand years, astronomy was limited to optical observations in the visible part of the electro- magnetic spectrum. Finally, in middle of the 20th cen- tury the first successful observations of celestial radio and microwave sources were made. Today astronomical observations are made over all of the electromagnetic spectrum. Space-based telescopes may be required at wavelengths where the Earth’s atmosphere is opaque to electromagnetic waves. Visible-light observations remain the most popular even today. Large and extremely expensive profession- al instruments are located at carefully-selected remote sites with clear skies and low levels of light pollution. Besides these large professional instruments, amateur optical observations with much smaller instruments lo- cated at non-perfect sites still play an important role in the science of astronomy. For example, most new aster- oids are discovered by amateur observers. Although the very first radio map of the sky was plot- ted by an amateur astronomer, Grote Reber in 1938- 1946 [1], radio astronomy quickly evolved into using extremely large antennas and sophisticated signal pro- cessing, all of them out of reach of amateurs. On the other side, it is worth noting that the most important discoveries of early radio astronomy like the non-ther- mal radiation of many celestial sources including the Sun, the λ =21 cm neutral-hydrogen line or the cosmic background radiation were made with relatively simple equipment originally designed for a completely differ- ent purpose. 114 T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Things started changing in favor of amateur radio as- tronomy after about 1990. The widespread deployment of satellite television made medium-size parabolic an- tennas and corresponding positioners available to amateurs. The semiconductor industry mastered III-V semiconductors, in particular GaAs based HEMTs al- lowing exceptionally low-noise receivers operating at room temperature well into the microwave region. Both allowed successful amateur observations of the λ=21cm neutral-hydrogen line [2]. At the turn of the century, television broadcasting shifted from analog to digital. Several analog micro- wave point-to-point link and analog satellite-television antennas were decommissioned since large anten- nas are no longer required by the new, more efficient digital modulations. The computing power of personal computers and corresponding interfaces increased to allow the comprehensive signal processing required for digital-television reception. Both allowed very suc- cessful radio-astronomical observations with amateur means [3, 4]. Radio astronomy excels in interferometry, especially VLBI. While smaller baseline interferometers could work with dedicated microwave point-to-point links, VLBI usually required sophisticated synchronization with atomic clocks and physical transfer of high-capacity magnetic- tape recordings for further processing. Today global navigation satellite systems (GPS, GLONASS and similar) offer precise worldwide synchronization of lo- cal clocks as well as local correction for unwanted prop- agation effects like ionospheric or tropospheric delays. The fiber-optic network offers high-capacity worldwide internet connectivity. Since both inexpensive synchro- nization and inexpensive data transfer is available to amateurs, the major achievements of amateur radio astronomy are yet to come. In this article the design, construction and calibration of a compact radio telescope is discussed in detail. The latter can produce 3D maps of the hydrogen distribu- tion in our galaxy. Much of the described hardware and software can be used to observe other celestial sourc- es, like point and distributed continuum sources, a few neighbor galaxies, some of the strongest pulsars etc. Fi- nally, several such radio telescopes could be combined into a powerful interferometer. 2 Radio-telescope design 2.1 System requirements The λ=21cm neutral-hydrogen line is not emitted by dense celestial objects like stars, but by lone hydrogen atoms in the interstellar medium. The density of the in- terstellar medium in a galaxy may be very low, just a few thousand particles in a cubic meter of space. Most of this interstellar matter is atomic hydrogen. This hy- drogen has two energy levels in its electronic ground state with an energy difference corresponding to a frequency of f0=1420405751.7667Hz. The upper level can be excited by the collision of two atoms. This ex- cited level then decays after an average lifetime of 11 million years by emitting a photon at the frequency mentioned. Yet our Milky Way and other galaxies are very large ob- jects. Although the interstellar space can be considered rather high vacuum, the few lone hydrogen atoms rep- resent a significant fraction of the total mass of a galaxy or other celestial body. The mass ratio varies: there are celestial objects with lots of sparse hydrogen atoms and few visible stars and the opposite is also possible. Due to the huge size of interstellar space, the radiation at λ =21cm sums up to significant and detectable values. Since relative velocities between different parts of our Milky Way galaxy span up to Δv≈±200km/ s, the expect- ed Doppler shift is in the range of Δf ≈±1MHz around a central frequency of f0 ≈1420.4MHz. Other galaxies are moving at even higher velocities with respect to us re- sulting in much larger Doppler shifts. Most important of all, the Doppler shift of the hydrogen line allows precise velocity measurements revealing the motions of celes- tial objects. The frequency band from fMIN =1400MHz to fMAX =1427MHz is therefore protected worldwide for radio-astronomy use. The λ=21cm neutral-hydrogen radiation from the Milky Way is coming from distributed sources with an- gular diameters of several degrees and an equivalent back-body brightness temperature of up to T ≈100K. Such sources can be reliably detected with rather small antennas with a diameter above d >1m≈5 λ. Using a slightly larger antenna d ≈3m ≈15 λ already reveals the spiral structure of the Milky Way. The receiving system of any radio telescope should add the lowest possible own noise to the weak celestial sig- nals. The receiving system noise includes the antenna noise and the receiver electronics noise. The antenna noise mainly comes from unwanted side lobes of its radiation pattern picking up thermal radiation of the warm neighborhood. Modern semiconductor devices produce a similar amount of noise when operated at room temperature. The sum of both is in the range TS =50...100K for relatively small antennas d ≈15 λ and available low-noise amplifiers. 115 2.2 Block diagram The design of the described compact radio telescope is based on available components: a 3.1m (10 feet) diameter surplus parabolic mesh reflector originally manufactured by EchoStar and intended for satellite- television reception in the 4GHz band. The parabolic dish is installed on an EGIS azimuth/elevation rotor EPR-203. Its corresponding control unit EPS-103 allows computer control through an RS-232 interface. A con- venient solution is to run Python scripts on a personal computer to track a selected object in the sky and/or scan the galactic plane as shown on Figure 1. Figure 1: Block diagram of a compact hydrogen-line radio telescope. Other parts of the radio telescope have to be custom designed and built. The VE4MA antenna-feed design is a good compromise among parabolic-mirror illumina- tion efficiency, unwanted side lobes picking up noise and aperture blocking. Of course, the best available HEMTs have to be used in the low-noise amplifier. Since 1.42GHz is not a particularly high frequency for modern electronics, frequency conversions are not required. The custom signal processing only includes custom- designed band-pass filters for 1.42GHz and additional amplifiers to overcome the relatively high noise of available receivers. Further signal processing is split into two independ- ent branches. A quick look at the received signal as well as interference threats is provided by a standard, scanning-receiver type spectrum analyzer Rigol DSA- 815. Much more efficient signal processing can be per- formed with a FFT spectrum analyzer once the wide dynamic range of a scanning-receiver spectrum ana- lyzer is no longer required. FFT spectrum analysis can be performed efficiently on modern personal comput- ers using an inexpensive DVB-T dongle as the analog interface. 2.3 Antenna and positioner The antenna of a radio telescope requires an unob- structed view to a much larger part of the sky than a typical satellite-television antenna covering the geosta- tionary arc. In particular, on the northern hemisphere a radio telescope should have an unobstructed view in the direction south. It therefore makes sense to install a small radio telescope on the roof or other elevated place to avoid local obstacles as shown on Figure 2. Figure 2: Mesh reflector with feed, LNA and azimuth/ elevation positioner. The spatial resolution of any radio telescope depends on the antenna size. Assuming an uniform illumination of the antenna aperture, the spatial resolution of the described single-antenna radio telescope is estimated as: 0.21 1.22 1.22 0.083 4.7 3.1 om rd d m λα = ⋅ = ⋅ = ≈ (1) In the case of a radio telescope, the antenna requires two-axis rotation. An azimuth/elevation or X/Y rotator can be used to keep the antenna pointed to a particular source in the sky in spite of the rotation of the Earth. On the other hand, the rotation of the Earth is frequently used in many radio-telescope designs as an additional axis of rotation. In the described radio telescope, all three available axis of rotation (two-axis positioner and Earth) can be used depending on the desired scan op- eration. The EGIS azimuth/elevation rotor EPR-203 is equipped with gears for 360° azimuth range and 90° elevation range. 2.4 Feed design In the microwave frequency range, the sky noise is usu- ally much lower that the ground noise. The equivalent T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 116 black-body temperature of the sky may be less than TSKY <10K while the ground noise is close to TGROUND ≈290K. Large satellite-receiving antennas and radio telescopes are designed to avoid collecting ground noise through side lobes of their radiation patterns. Antennas larger than d >100 λ are frequently built as dual-reflector Cassegrain telescopes. In this way any spillover of the feed sees the cold sky over the edge of the secondary reflector. An alternative solution is to control the radiation pat- tern of the feed using a corrugated horn or a corrugat- ed flange. Neither solution is practical with small, rota- tional-symmetric parabolic mirrors due to the blockage of the relatively large secondary mirror or corrugated feed. An efficient solution for small d < 30 λ parabolic mirrors is a simple circular-waveguide horn with the Kumar choke [5], also known as the VE4MA feed [6] as shown on Figure 3. Figure 3: VE4MA feed design. Changing the diameter of the circular waveguide, the di- mensions of the Kumar choke as well as its position rela- tive to the horn aperture allows optimizing the VE4MA feed for deep, rotational-symmetric parabolic mirrors with focal-to-diameter ratios in the range f/d =0.3...0.5. The dimensions shown on Figure 3 should be close to optimal for our EchoStar mesh dish with a focal-to-diam- eter ratio of f/d =0.4 at a central frequency of f0 =1.42GHz. The described VE4MA feed was practically implement- ed from pieces of thin aluminum sheet bolted together with several M3 screws as shown on Figure 4. The posi- tion of the Kumar choke with respect to the horn aper- ture was adjusted experimentally for the lowest system noise temperature TS as shown on Figure 4. The efficiency of the Kumar choke was further checked by measuring the radiation pattern of the feed both with the Kumar collar and the bare waveguide horn with the choke removed. The measurements also show important side effects that might impair the perfor- mance of the radio- telescope antenna. For example, a circular waveguide only supports the propagation of its fundamental TE11 mode when its diameter exceeds dTE11 >0.586 λ. Using a simple, non- symmetric probe to excite the fundamental TE11 mode, higher order modes may also be excited. The next high- er mode is the TM01 mode that appears at dTM01 > 0.765 λ. Since the circular-waveguide diameter was selected as d =0.73 λ rather close to the TM01 cutoff, the latter does not receive much attenuation in a short waveguide horn. The final result is a squint of the waveguide-horn main radiation lobe in the E plane. Fortunately the Ku- mar choke in the VE4MA feed is able to correct this de- fect as shown on Figure 5. Figure 5: Measured feed E-plane radiation pattern. Little if any squint is visible in the measured H-plane radiation pattern of the bare waveguide horn as com- pared to the VE4MA feed on Figure 6. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Figure 4: Practical implementation of the VE4MA feed. 117 Figure 6: Measured feed H-plane radiation pattern. In both E and H planes, the Kumar choke provides a substantial improvement of the feed radiation pattern resulting in a more than 5dB decrease of its unwanted side lobes. Further, the VE4MA feed exhibits a much better rotational symmetry than the bare waveguide horn. The improved feed radiation pattern allows a sub- stantially lower system noise as measured during the calibration in section 3.2. of this article. The radio emissions of certain celestial sources are polarized, while radio emissions from other celestial sources are not polarized. The radio emission of neutral hydrogen atoms at λ=21cm is not polarized. Therefore the polarization of a simple radio telescope for the neu- tral- hydrogen line may be arbitrary. The simplest solu- tion is to use linear polarization as with the described linearly-polarized VE4MA feed. Both horizontal linear polarization and vertical linear polarization were ex- perimented with identical results. While using linearly-polarized feeds with relatively small d <30 λ, rotational-symmetric parabolic mirrors, a new problem appears. A significant part of the feed radiation is reflected by the parabolic mirror back into the same feed. Considering the described mesh dish with a focal f =1.23m and the described VE4MA feed with a gain of about G≈8.55dBi ≈7.15 at λ=21cm, the magnitude of the described reflection can not be ne- glected: Γ 0.097 20.3 4 G dB f λ π ⋅= ≈ ≈ − ⋅ (2) This additional reflection may either degrade or im- prove the feed impedance matching, since the overall antenna return loss is a phasor sum of the different re- flections involved. On the other hand, the noise perfor- mance of very low-noise microwave receivers is very sensitive to the source (antenna) impedance matching. Some low-noise receivers may even become unstable when connected to a badly mismatched antenna. The return loss of the described bare VE4MA feed in free space was measured as ∣Γ∣=−13.4dB using a direc- tional coupler with a measured directivity better than -30dB. After installing the feed in the focal point of the parabolic mirror, the return loss of the whole antenna improved to ∣Γ∣=−17.8dB at f0 =1.42GHz. This return loss was considered good enough so that any further modifications to the feed were considered unnecessary. The above discussion is no longer valid if a parabolic mirror with a different focal length from ours is used, since the phase of the reflection from the dish chang- es really quickly and is proportional to twice the focal length! Using an arbitrary parabolic mirror, the imped- ance matching of the complete antenna should al- ways be checked. In the worst case, the feed design, in particular the probe length and its position inside the waveguide horn may need to be changed. 2.5 LNA design The semiconductor devices of choice for microwave low-noise amplifiers (LNAs) are high- electron mobility transistors (HEMTs) based on GaAs. The best commer- cially-available devices are designed to operate in a ZK =50Ω characteristic-impedance environment in the fre- quency range around f0 =12GHz for satellite-television reception. Such devices are therefore not optimized for operation at the neutral-hydrogen frequency of f0 =1.42GHz. Similar HEMT structures with a much wider gate are re- quired for low-noise operation in a ZK =50 Ω environ- ment at L-band frequencies. Besides not being volume- production items, wider-gate devices operate at higher currents increasing the power dissipation. The latter increases the chip temperature impairing their noise performance. It therefore makes sense to use standard 12GHz HEMT devices even at much lower frequencies but in a higher-impedance environment as described in [7]. The circuit published in [7] was originally designed for GaAs MESFETs that operated at about VDS ≈ 4V. Modern low-noise GaAlAs/GaAs HEMTs operate at much lower voltages VDS ≈1.5V. Operation of a low-noise HEMT at higher voltages may affect its reliability and trigger T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 118 long-term degradation effects. In order to improve the reliability with HEMT devices, the simple source-bias re- sistor from [7] was replaced with an active bias circuit including a low-frequency silicon NPN transistor. When a Ku-band HEMT or MESFET is operated at L band, an efficient input-impedance match for lowest noise is a series inductor with the gate while no special output-impedance matching at the drain is required. A two-stage amplifier design offers around GLNA≈30dB of gain at reasonable stability. Both gain and stability can be adjusted with the inter-stage match. A GaAs MESFET can be used in the second stage as shown on Figure 7. Figure 7: LNA circuit diagram. In order to operate at higher impedances, the circuit of the LNA is not built on a printed- circuit board. The circuit of the LNA is supported in free air by several lead-less 470pF ceramic disc capacitors soldered to the bottom of a small brass box. The remaining com- ponents are traditional parts (not SMD) with wire leads. Although such a design is not suitable for volume pro- duction, it makes sense to achieve top performance of the radio telescope. The supply voltage +12V= is fed through the output SMA connector using a bias tee at the other end of some 15m of coaxial cable feeding the indoor equipment. The input connector is a male SMA to be screwed directly on the antenna feed to avoid any feed-line losses as shown on Figure 8. Figure 8: Practical implementation of the LNA. Many LNA prototypes were built using different GaAs devices either in the original circuit [7] or with the modified bias as shown above. Both MESFETs and HEMTs could readily achieve noise temperatures in the TLNA≈30K range as measured with a HP8970 noise- figure meter equipped with a HP364A ENR≈ 5dB cali- brated noise source. The best LNA samples could reach even TLNA=25K at a room temperature of 20°C (293K) after carefully adjusting the input impedance match. HEMTs typically provide much more gain than MESFETs, therefore using HEMTs in both stages may lead to in- stability. The LNA is followed by the indoor receiver with a noise temperature in the TRX ≈ 290K (room temperature) range. Considering an aCOAX =−10dB cable loss to the rooftop antenna and a LNA gain of GLNA ≈30dB, the in- door receiver adds about TRX´ ≈3K of noise to the whole radio telescope noise temperature. 2.6 Signal processing and distribution A carefully-designed LNA may contribute less than TLNA ≈ 30K to the total system noise of the radio telescope, but other components may be much noisier. For exam- ple, a good radio- frequency spectrum analyzer has a noise figure in the FSA ≈ 25dB range corresponding to a noise temperature of about TSA ≈10 5 K. A GLNA ≈30dB LNA followed by aCOAX =−10dB cable loss sets the con- tribution of the spectrum analyzer to an unacceptable TRX´ ≈1000K to the total system noise or in other words more than one order of magnitude larger than any oth- er noise source in the system. Additional low-noise amplification is therefore required in front of spectrum analyzers and other “deaf” instru- mentation. Additional amplification brings a new prob- lem: a high-gain amplifier chain may easily be driven into saturation by strong terrestrial radio transmitters operating at nearby frequencies. The whole signal pro- cessing and distribution chain includes a bias tee for the LNA, a first band-pass filter, a MMIC amplifier, another band-pass filter, another MMIC amplifier, a power split- ter and a SDR receiver front-end as shown on Figure 9. Microwave filters can be built in many different tech- nologies. For the radio-astronomy frequency band at Figure 9: Signal processing and distribution chain. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 119 1.42GHz any filters have to be custom designed and built therefore excluding high-volume technologies like SAW devices. Considering low-volume applications, cavity filters are simple to manufacture from standard aluminum tubing yet individual λ/ 4 resonators achieve an unloaded quality in excess of QU >1000. A comb ar- rangement of the individual λ/ 4 resonators allows a more compact band-pass filter than an interdigital de- sign. Considering the filtering requirements of the ra- dio telescope, two separate but almost identical comb filters with three λ/ 4 resonators each are required as shown on Figure 10. Figure 10: Comb-filter design. The input and output coupling is achieved with probes similar to the one used to excite the waveguide horn. The three resonators are made from 8mm diameter aluminum rod, while the cavity is made from standard aluminum tube of rectangular cross-section of 40mmX60 mm and 2.5mm thick walls. The rectangular tube extends 38mm beyond the probe positions so that the electromagnetic field already decays inside the tube. Covers on both sides of the tube are only required to keep dust and dirt out- side of the filter cavity as shown on Figure 11. The described simple all-aluminum construction with few junctions provides excellent electrical performanc- es that stay stable in time. The λ/ 4 resonators are tuned to the exact frequency with capacitive tuning screws in the opposite wall of the cavity. The bandwidth is selected to cover the whole 1.42GHz radio-astronomy frequency band. An order-of-magnitude narrower filter could be used to observe the neutral-hydrogen radia- tion only from our Milky Way galaxy. At a pass-band bandwidth of B≈ 25MHz the high un- loaded quality of the individual resonators allows a low insertion loss of the whole three-stage filter in the aBPF ≈−0.5dB range. The exact value of the latter is a function of the actual filter alignment as visible on the measured response on Figure 12. Figure 12: Measured comb-filter response. Unfortunately, other components of the radio tel- escope do not exhibit such a flat frequency response over the 1.42GHz radio-astronomy band. Since most of the ripple comes from the reflections at both ends of the relatively long coaxial cable between the outdoor LNA and the indoor processing chain, it makes sense to adjust the first cavity filter to flatten the overall fre- quency response of the radio telescope. In this way the first filter is also used as an equalizing filter. The additional signal gain is provided by two MMIC amplifiers manufactured by Mini-Circuits in InGaP HBT technology. The amplifier chips are packaged in SOT-89 plastic packages that provide good radio-frequency ant thermal performance at the same time. Both pack- aged amplifiers are installed on printed-circuit boards together with a few bias components. Finally both printed- circuit boards are installed in aluminum cases with SMA connectors as shown on Figure 13. A low-noise MMIC Gali-52+is used in the first stage pro- viding a measured gain of G52 =19.3dB with a measured noise figure of F52=2.8dB in the 1.42GHz band. A power MMIC Gali-5+ is used in the second stage providing a measured gain of G5 =19.1dB with a measured noise figure of F5 =4.1dB in the 1.42GHz band. Figure 11: Inside the comb-filter cavity. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 120 The gain of the whole chain including two band-pass filters and two MMIC amplifiers approaches G≈37dB. The latter is high enough to neglect the noise contribu- tion of a spectrum analyzer or software-defined radio (SDR) front end. A Wilkinson power divider with a split- ting loss of aW≈−3dB is used to supply two independ- ent instruments at the same time. 3 Instrument calibration 3.1 Positioner calibration No matter what directional antenna and correspond- ing positioner are used, after installation both need ac- curate mechanical calibration before they can be used for satellite communications and/or radio-astronomy work. In particular, the available EGIS azimuth/elevation rotor EPR-203 has a linear tooth-gear transmission for rotation in the azimuth plane and a non-linear worm- gear- plus-lever transmission for rotation in the eleva- tion plane. The azimuth calibration is simply finding the offset constant from the reference position due to installation and rotor errors. Once found, the offset con- stant may simply be inserted in the corresponding EGIS control unit EPS-103. The non-linear elevation transmission of the EGIS rotor EPR-203 is much more tricky. The look-up table correc- tion built in the EPS-103 control unit was found rather inaccurate. Adjusting the remaining elevation offset constant did not solve the problem. A working solu- tion was to derive our own elevation-correction table in 5-degree steps using a gravity inclinometer as shown on Figure 14. Finally, linear interpolation between two neighbor cor- rection-table points is implemented in the Python code controlling the antenna positioner. The inclinometer is further used to correct small instal- lation errors. The supporting mast of the antenna was found to deviate as much as one degree from vertical. Rather than performing mechanical adjustments on the antenna and/or its support structure, such small errors are easily corrected in Python code as vector rotations. The final check of the antenna/rotor alignment is track- ing the Sun. The position of the latter can be computed precisely at any time. The Sun is the strongest natural source of radio waves in the sky. Of course the Sun can be tracked optically, for example by checking the po- sition of the feed shade on the parabolic reflector. Us- ing rotational-symmetric parabolic reflectors, the feed shade should fall exactly in the center of the dish as shown on Figure 15. Figure 15: Checking the feed shade. 3.2 System noise temperature A very important parameter of any satellite-receiving station and/or radio telescope is the system noise temperature. The latter describes the amount of noise added by the station hardware to the received signal. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Figure 13: Packaged MMIC amplifiers. Figure 14: Using the inclinometer. 121 The simplest procedure to measure the system noise temperature is to turn the antenna into two different targets radiating as black bodies at well known temper- atures. The described hot/cold method yields the hot/ cold power ratio: ( ) ( ) B HOT SHOT HOT S COLD B COLD S COLD S B k T TP T TY P B k T T T T ⋅ + + = = = ⋅ + + (3) Both the receiver bandwidth B and the Boltzmann con- stant kB ≈1.38⋅10-23J/K cancel out in the hot/cold ratio. The radio-telescope antenna is first pointed to a cold part of the sky assuming TCOLD ≈10K. The antenna is af- terwards pointed to a forest (good microwave absorber if dry) assuming THOT ≈ 290K. The described radio-tele- scope with the VE4MA feed achieved a hot/cold ratio Y ≈7dB≈5. The measured system noise temperature is: 290 5 10 60 1 5 1 HOT COLD S T Y T K KT K Y − ⋅ − ⋅= = = − − (4) It is estimated that these 60K include about 32K com- ing from the LNA, about 3K coming from the remaining stages of indoor signal processing and about 25K from the antenna side lobes observing the warm Earth. Such a low antenna noise temperature of just 25K is an excel- lent result for the VE4MA feed and a relatively small d ≈20 λ parabolic dish obstructed by the feed itself and its four metal supporting struts. For comparison, the same measurement was also per- formed with the a simple waveguide- horn feed by re- moving the Kumar choke from the VE4MA feed. A hot/ cold ratio of only Y ≈5dB≈3.16 was obtained in the latter case. The corresponding measured system noise temperature doubled to TS ≈120K therefore halving the sensitivity of the radio telescope! Removing the Kumar choke therefore increases the antenna temperature up to 85K due to the much larger side lobes of a simple waveguide horn. 3.3 Antenna illumination efficiency Yet another important parameter to check is the anten- na aperture-illumination efficiency. The latter may be impaired both due to an incorrect illumination (usually under-illumination) and/or due to mechanical defects of the parabolic-reflector surface. Neither can be de- tected with distributed sources like those used in the system-noise-temperature measurement described in the previous chapter. A calibrated point source of radiation is required to measure the antenna aperture illumination efficiency. The Sun is a strong natural radio source. The Sun is small enough to be considered a point source since its angu- lar diameter αSUN ≈0.5 ° ≪ α≈4.7 ° is much smaller than the resolution of the described radio telescope. Unfor- tunately the radio radiation from the Sun is changing with the Sun activity. Changes are different at different frequencies. Fortunately for our measurement the activity of the Sun is monitored by several observatories around the world on many different frequencies. This data is regu- larly published on the internet [8]. In our case the most useful observation is at f0=1415MHz and the most use- ful observatory is located in San Vito dei Normanni near Brindisi in Italy. The latter provides Sun data close to our local noon with the Sun high on the sky. During our measurement San Vito reported a solar- flux spectral density of S/B =69SFU =69⋅10-22 W/m2/Hz at f0=1415MHz. At the same time we measured a hot/ cold ratio of Y ≈13.5dB≈ 22.4 while pointing the radio telescope both to the Sun and to a cold part of the sky. Considering that our antenna only receives half of the reported radiation on one polarization and that the cold sky radiates at TCOLD ≈10K we obtain the effective area of our antenna as: ( ) ( ) ( ) ( )23 21 2 2 2 1 2 1.38 10 / / 22.4 1 60 10 6.9 10 / 5.99 B S COLD eff k Y T T A S B W Hz K K K W m Hz m − − ⋅ − ⋅ + = ≈ ⋅ ⋅ ⋅ − ⋅ + ≈ ≈ ⋅ ≈   (5) Dividing the effective area with the physical area of our 3.1m parabolic dish we estimate the illumination effi- ciency as: ( ) ( ) 2 2 2 5.99 0.79 / 2 3.1 / 2 effA m d m η π π = ≈ ≈ (6) An aperture illumination efficiency of 79% is an excel- lent result for a relatively small d ≈20 λ parabolic dish obstructed by the VE4MA feed and its four metal sup- porting struts. This result also proves that the perfor- mance of our parabolic dish is not affected by me- chanical distortions, at least at the given wavelength of λ =21cm. 3.4 Radio-interference mitigation An amateur radio-telescope builder usually has little if any choice to select the place where to place his instru- ment. Wind is a primary concern for a large antenna. Snow and ice loading may also be important factors. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 122 While winds and precipitations have been observed and measured for many years, radio interference is a big unknown in a constantly changing world of elec- tromagnetic pollution of any kind. Traditional amateur astronomers usually prefer remote places to avoid optical pollution. Such places may have high winds and receive lots of precipitations in winter. An amateur radio telescope likely has to be placed in an urban environment where meteorological constraints are less severe. The level of radio interference in an urban environment may span over several orders of magnitude. Therefore at least some rough interference measurements are recommended in site before build- ing such an instrument. In our case the radio telescope is placed in a suburb of a town in a closed valley protected from winds. Inter- estingly, no radio interference from radio transmitters was found in the 1.42GHz protected astronomy band. The nearest radio transmission was found at about 1436MHz outside the protected band. It looks like certi- fied radio transmitters are shielded and filtered enough not to cause harmful radio interference. As a precau- tion, a radio-frequency spectrum analyzer covering a 50MHz wide span around 1.42GHz is always connect- ed to the radio telescope to check for interference as shown on Figure 16. Figure 16: Pass-band around 1420.4MHz and interfer- ence at 1436MHz. On the other hand, different kinds of radio interference were experienced from electronic and electrical equip- ment not intended to transmit any radio frequency at all. Live-video IP cameras from different manufacturers, cheap and expensive products alike, were identified as the worst source of radio interference, radiating har- monics of all clocks used by their internal digital cir- cuits. Unlike suggested in many articles and websites, video IP cameras can not be used to help aiming the antenna of a radio telescope! Most electronic equipment containing digital circuits causes narrow-band interference around 1420.000MHz +/- tolerances of their internal crystal oscillator. This in- terference may be a high-order harmonic of an internal clock operating at 10.000MHz or 20.000MHz. A simple countermeasure is to move all computers and network devices at least two concrete floors below the radio-tel- escope antenna. Such unintentional interference also decays quickly with distance. Of course all unneces- sary electronic equipment may simply be switched off when the radio telescope is operating... Finally, we could not identify the source of some kind of wide-band interference that appears periodically. It may be coming from a spark-ignition system, most like- ly natural-gas heater, since its occurrence and period was found correlated with low outside temperatures. It should be noted that both narrow-band and wide- band low level interference could only be detected us- ing software- defined-radio (SDR) techniques and very long integration times. 4 Hydrogen spectrum measurements 4.1 Spectrum analyzer Our first experiments were performed with a conven- tional, scanning-receiver type spectrum analyzer Rigol DSA-815. Such an instrument has a better dynamic range than its all-digital FFT- based counterparts. On the other hand, a scanning receiver uses the incoming signal rather inefficiently. Most of the signal energy is lost while the scanning receiver dwells on other frequencies. In the case of radio astronomy, the signal-to-noise ra- tio is low. The signal coming from celestial sources is random just like thermal noise. Measuring the average power 〈P〉 of random signals requires lots of averag- ing. The fluctuation (intended as standard deviation) of the result ΔP only decreases with the square root of the number of measurements N according to the Dicke equation [9]: Δ 1 1 ΔVIDEOBP T P B TN B τ = = = = ⋅ (7) In the case of a radio receiver, the number of inde- pendent measurements N = B⋅τ is the product of the receiver bandwidth and measurement time. Most scanning-receiver type spectrum analyzers have built- in averaging in the form of a low-pass video filter with a cutoff frequency BVIDEO ≪ B much smaller than their resolution bandwidth. To observe the neutral-hydrogen radiation at λ=21cm a sensible setting is a resolution bandwidth of B=10kHz. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 123 Setting the video bandwidth to BVIDEO =10Hz is equiva- lent to an averaging factor of N =1000. The fluctuation of the result is thus reduced by a factor of √1000≈31.6. The typical ~10dB noise “grass” of a spectrum analyzer is reduced to ~0.3dB. Considering the noise sources in our radio telescope T =TSKY+T S≈10K +60K =70K, the stan- dard deviation ΔT =〈T〉/eN =2.2K is a good estimate for the minimum observable signal coming from celestial sources. Unfortunately such a measurement becomes very slow due to the inefficient use of the incoming signal energy by a scanning receiver. Observing the hydrogen fre- quency band of our Milky Way galaxy requires scanning a band of Δf =2MHz. A single observation requires 20s although only 100ms are effectively used: Δ Δ 120 100 VIDEO VIDEO f ft s ms B B B B τ τ⋅= = = = = ⋅ (8) On the other hand, if the same measurement is per- formed with a FFT spectrum analyzer, the latter making full use of the incoming signal energy, the observation time reduces to t = τ=1/BVIDEO =100ms. Considering the rotation of the Earth and the beam width of the antenna used, a correction of the antenna direction is required about once per minute to track a se- lected deep-space radio source. A simple Python script “zvezdar.py” transforms the requested right ascension (celestial longitude) and declination (celestial latitude) into azimuth and elevation for a particular location of the observer. Corrections for mechanical imperfections of the antenna rotor EPR-203 are included. Since its cor- responding control unit EPS-103 only allows one com- mand at a time, elevation and azimuth commands are sent by the Python script alternatively every 25 seconds through an RS-232 interface at 9600bps. 4.2 Sample measured spectra A radio telescope for the neutral-hydrogen line does not see much radiation coming from dense, optically bright stars, but the radiation form sparse, lone hydro- gen atoms. In the direction of the star Deneb (constel- lation Cygnus) three (outer) arms of our Milky Way gal- axy can be seen due to their different Doppler shifts. Our Solar system is approaching all three (outer) arms (positive Doppler shift) at different relative velocities as shown on Figure 17. The equivalent black-body brightness temperature of a hydrogen cloud T(f)= THOT – TCOLD can be calculated by reversing equation (3) to obtain THOT from known values of TCOLD and TS as well as the measured ratio Y (f) with the spectrum analyzer. In particular, the measured ratios on Figure 17 are about Y (f)≈ 3dB for the first arm of the Milky Way and about Y (f)≈ 2dB for the second and third arm. On the other hand, the central part of our Milky Way galaxy is much denser resulting in a continuous neu- tral-hydrogen spectrum. Individual (inner) arms can hardly be resolved. Due to the rotation of our galaxy, the region in the direction of the constellation Scutum is moving away from our Solar system resulting in a negative Doppler shift as shown on Figure 18. Figure 18: Neutral-hydrogen spectrum from the direc- tion of the star α Scutum. 4.3 Measured hydrogen column density Integrating the neutral-hydrogen power spectrum represented as the equivalent black-body brightness temperature T (f), the total column density of neutral- hydrogen atoms can be derived: ( )1/2 0 8 ln2 MAX MIN f B H f k T f df h c π τη λ ⋅ ⋅ = ⋅ ⋅ ⋅ ⋅ ∫ (9) Inserting the physical constants in the above equation: the Boltzmann constant k B ≈1.38⋅10 -23 J/K, the Planck constant h≈6.63⋅10-34 Js, the speed of light c0≈3⋅10 8 m/s, T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Figure 17: Neutral-hydrogen spectrum from the direc- tion of the star Deneb. 124 the excited neutral-hydrogen half life τ 1/2 ≈3.3⋅1014s (~11 million years), the wavelength λ =21cm and referring the observed equivalent black-body brightness tem- perature T (v ) to the relative velocity in place of frequen- cy T (f), the above equation can be simplified into [10]: ( )18 2 atoms 1.82 10 cm km/s MAX MIN H T dv K ν ν ν η = ⋅ ⋅ ⋅ ⋅∫ (10) Figure 19 shows the equivalent black-body brightness temperature T (v ) as a function of velocity calculated from the observed frequency spectrum in the direction of the star Deneb. Figure 19: Neutral-hydrogen velocity and column den- sity. The integrated hydrogen-atom column density is about ηH ≈6.6⋅1021 atoms/cm2 spread over a distance of ap- proximately 35000 light − years≈3.3⋅1020 m =3.3⋅1022 cm. The average hydrogen density in the interstellar space is therefore very low, just N ≈0.2 atoms/cm3=2⋅105atoms/m3 or very good vacuum indeed... 5 Hydrogen mapping f our galaxy 5.1 Software signal processing Radio astronomy and in particular interferometry were one of the first large-scale applications of digital-signal processing. The reason is very simple: the signal-to- noise ratio in radio astronomy is always low. Therefore just a few bits of resolution are required to adequately represent and process the signals, simplifying the hard- ware. Rather than using exotic hardware in the described radio telescope, an inexpensive DVB-T USB dongle based on the RTL2832U chip was used for frequency down conversion and A/D conversion. Besides a dedi- cated DVB-T demodulator, the RTL2832U chip includes a straightforward 8-bit A/D converter for analog radio reception. The latter is well supported both with driv- ers, application software and development tools in all known operating systems on personal computers. The A/D converter can process signal bandwidths up to B≈ 2.4MHz. The latter is more than sufficient to observe the neutral-hydrogen spectrum of our Milky Way galaxy. The RTL2832U chip is frequently coupled with the R820T2 tuner (down converter) chip. Although the latter is only specified up to fMAX ≥1GHz, most chips exceed fMAX≥1.7GHz at room temperature. The R820T2+RTL2832U combination therefore allows direct processing of the amplified and filtered 1.42GHz hydro- gen-line signal provided that the tuner chip R820T2 is kept at a constant room temperature. An improved ver- sion of the DVB-T dongle is used in our radio telescope featuring a TCXO frequency reference in an aluminum case. The latter provides both shielding and cooling of all internal components as shown on Figure 20. Figure 20: Inexpensive SDR front-end. Although excellent open-source development tools exist for the R820T2+RTL2832U combination, ready- made software was tried first with excellent results. In particular we successfully experimented the freeware “HDSDR version 2.76” [11] running on “WindowsXP”. The DVB-T dongle requires its own low-level USB driver “zadig_xp.exe” and the corresponding interface “ExtIO_ RTL2832.dll” to HDSDR. The main advantage of HDSDR over other similar software is the ability to use large av- eraging factors and display small amplitude differenc- es, both of them very necessary for radio astronomy. We only used the radio-frequency FFT spectrum ana- lyzer included besides several other receiver functions in HDSDR. A center frequency of f0=1420.405MHz and an observation span of Δ f =1.44MHz are good choices for hydrogen-line observations of our Milky way gal- axy. The optimum FFT size was found experimentally T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 125 as 214 =16384 to separate easily hydrogen spectra from narrow-band interference. The spectrum is first averaged by a factor of 2048 in- side HDSDR. The averaged spectrum is displayed on a 1280-pixel wide LCD screen, therefore 16384 spectrum lines are further averaged by a factor of about 13 to 1280 display columns. All this averaging reduces the power (temperature) fluctuations to approximately 10dB/ e2048⋅13≈0.06dB. The “HDSDR” software can display the averaged spec- trum plot as a function of frequency or as a spectrum waterfall with different colors indicating signal inten- sity. Further, the waterfall can be made slow enough to allow mechanical scanning of the sky and its color scale can be selected to show intensity changes of less than 0.5dB. One of the first experiments with “HDSDR” run- ning independently on two personal computers show- ing 2D images (galactic longitude and relative velocity) of our Milky Way galaxy is shown on Figure 21. Figure 21: Waterfall display with HDSDR software. The A/D converter inside the RTL2832U chip includes an analog anti-aliasing filter on its input. The latter has a ripple of about +/-0.5dB over the central three quarters of the observed frequency span and rolls off sharply (about -3dB) at both span edges. This is an excellent re- sult for an analog filter and in most applications goes unnoticed. On the other hand, radio astronomy re- quires measuring very small differences of signal levels. Unfortunately the current version of the HDSDR free- ware does not include a normalization of the response nor other countermeasures to correct the above-men- tioned pass-band ripple. 5.2 Galactic-plane scanning One of the most important historical discoveries of hy- drogen-line observations is that our Milky Way is a spiral galaxy. The plane of the Milky Way is currently inclined by 62.8° with respect to the equatorial plane. This incli- nation is slowly increasing due to the precession of the Earth’s rotation axis. Most of the hydrogen-line radia- tion comes from the plane of our galaxy. Hydrogen- line radiation from other directions is much weaker. There- fore the plane of the Milky Way is the target of choice to begin with hydrogen-line observations. The whole galactic plane is not visible from our latitude 46° north. Considering the beam width of a 3.1m-diam- eter antenna at 1.42GHz, celestial objects can only be observed at elevations larger than EL≥10 °above the local horizon to avoid thermal noise radiated by the warm ground. In our case we were lucky enough not to have any obstacles nor interference sources in the direction south where the most difficult-to-see celestial objects appear at particular times during the day. Galactic coordinates, galactic longitude and galactic latitude, are the coordinates of choice for hydrogen- line observations. The galactic plane of our Milky Way galaxy is by definition galactic latitude zero GLAT =0. Galactic longitude zero GLON =0 is defined by radio ob- servations as the radio source Sagittarius A*, supposed to be a super-massive black hole in the center of our galaxy. Galactic longitude increases in the opposite di- rection of the rotation of our galaxy. The angular ve- locities of different parts of our galaxy are different resulting in Doppler shifts. Our scanning of the galactic plane is explained us- ing the artist’s impression of the Milky Way [12]. Sagittarius A has a declination of about d y-29°, therefore it reaches a elevation of just 15° above the southern horizon at our latitude 46° north. It there- fore makes sense to start the galactic-plane scan- ning just before Sagittarius A at a galactic longitude of about GLON =-10o. The observed galactic longitude is then increased at a carefully selected rate for about 12 hours to end our scanning around galactic longitude GLON=265° as shown on Figure 22: In this way the whole visible galactic arc is observed during one single scan with plenty of time for data averaging. Considering the rotation of the Earth such a scan both starts and ends at the southern horizon. The direction of the scan is important. Scanning by increasing the galactic longitude the galactic equator appears high above the local horizon at our latitude 46° north. On the other hand, scanning in the oppo- site direction, decreasing the galactic longitude from GLON=265o to GLON=-10o, the galactic equator ap- pears rather low above our northern horizon. Un- fortunately local obstructions at our micro location preclude such scanning. Besides steering the antenna positioner to scan the galactic plane, the Python script “mlekar.py” also records the times and corresponding galactic longitudes and latitudes in a text file. On the T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 126 other hand, the “HDSDR” software puts time tags on the waterfall display. By combining both, the waterfall can be referenced to the galactic longitude. Since the com- puter screen is not high enough, periodic screen-shots are taken with the “Chronolapse” software and the final result is stitched together with the “IrfanView” image- processing software as shown on Figure 23. Figure 23 shows unprocessed and uncorrected data. Be- sides the hydrogen spectrum, continuum (broadband) celestial sources are also visible as wide horizontal fea- tures like the center of our galaxy (always well visible), the star-forming region Cygnus X or the supernova rem- nant Cassiopeia A (barely visible on some scans). Some artifacts are caused by the pass-band ripple of the analog anti-aliasing filter of our receiver. Ground ther- mal noise is visible at the beginning and at the end of the scan. Narrow-band CW interference (clock harmon- ics from digital equipment) is visible as narrow spectral lines around 1420.000MHz. Periodic wide-band inter- ference (likely spark ignition) is visible towards the end of the scan. The interruption at galactic longitude 180° is due to our programming mistake of the rotor control unit EPS-103 causing a 360° rotation of our antenna. 5.3 3D map of the Milky way Performing several galactic-longitude scans at different galactic latitudes, a 3D map of our galaxy can be built. Considering the beam width of our antenna, separate scans were made in 5° galactic-latitude steps. A map composed of five such scans at galactic latitudes of -10°, -5°, 0°, +5° and +10° is shown on Figure 24. Observing southern galactic latitudes becomes increas- ingly more difficult from the northern hemisphere. Missing data at galactic latitude -10° (south) had to be replaced with a gray bar at the beginning of the scan. On the other hand, the time of year of the galactic-lon- Figure 23: Waterfall display with HDSDR software. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Figure 22: Galactic-plane scanning from our latitude. 127 gitude scan at galactic latitude +10° (north) was inten- tionally selected to cross the Sun. The most interesting scan is in the galactic plane (ga- lactic latitude 0°) showing even distant arms of our galaxy with large Doppler shifts of the hydrogen line. Moving away from the galactic plane either south or north the amount of visible hydrogen decreases. Even more important, moving away from the galactic plane only local hydrogen with a relatively small Doppler shift remains visible. 6 Conclusion In order to understand the difficulties in operating a radio telescope, it is fair to compare its performance a well-known device like a GSM phone operating in the nearby 1.8GHz frequency band. Both the hydrogen line of a single galactic arm and the GSM signal have a similar bandwidth of about B≈ 200kHz. While the GSM phone requires a signal with the equivalent black-body brightness temperature of TGSM ≈30000K, the arms of our galaxy on average reach TGALAXY ≈30K. GSM signals are therefore stronger by a factor of 1000=30dB than hydrogen- line signals. The presented compact radio telescope has an antenna with an effective area about 10000=40dB times larger than a GSM phone. Large professional radio telescopes have effective areas about 1000=30dB times larger than our radio telescope. In total, our compact radio telescope is about 70dB more sensitive than a GSM phone and about 30dB less sensitive than a large pro- fessional radio telescope. Nevertheless we were able to obtain a 3D map, inten- sity versus galactic longitude, galactic latitude and Doppler shift of our Milky Way galaxy with our compact radio telescope. Such a 3D map lead to the discovery of the spiral structure of our galaxy some 60 years ago using much larger instruments. A further limitation of our compact radio telescope is its location at 46° north- ern latitude while the most interesting features of our galaxy and neighbor celestial objects are located in the southern sky. Our results were obtained using signal-processing software that was not originally designed for radio- astronomical observations. Much improvement could be obtained by writing or own code to allow response normalization, arbitrary averaging factors and known radio-interference rejection. In this was the hydrogen spectra of less bright celestial objects could be ob- served like the Andromeda galaxy as well as many con- tinuum sources like some of the strongest pulsars. Of course, the most ambitious goal are radio interferom- eters with arbitrary baselines including several similar compact radio telescopes, synchronized using GNSS (global satellite navigation systems) and linked via in- ternet. T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 Figure 24: Measured 3D map of the Milky Way. 128 7 References 1. Grote Reber: “Galactic Radio Waves”, Sky and Tel- escope, Vol.8, No.6, April, 1949. 2. Mal Wilkinson, John Kennwell: “Hydrogen-line observations of the Galaxy and the Magellanic Clouds”, Australian Journal of Astronomy, Vol.5, No.4, pp 121-133, 1994. 3. Jean-Jacques Maintoux: “Radioastronomie”, http://f1ehn.pagesperso-orange.fr/fr/f_radioas- tro.htm 4. Nguyen Van Hiep, Pham Tuan Anh, Pham Ngoc Diep, Pham Ngoc Dong, Do Thi Hoai, Pham Thi Tuyet Nhung, Nguyen Thi Thao, Pierre Darriulat: “The VATLY Radio Telescope”, Communications in Physics, Vol.22, No.4, pp 365-374, 2012. 5. A. Kumar, “Reduce Cross-Polarization in Reflector- Type Antennas,” Microwaves, pp 48-51, March 1978. 6. B.W. Malowanchuk, VE4MA, “Use of Small TVRO Dishes for EME,” Proceedings of the 21st Confer- ence of the Central States VHF Society, ARRL, pp 68-77, 1987. 7. Matjaž Vidmar: “Ein sehr rauscharmer Antennen- verstarker fuer das L-band”, pp 163-169, UKW Ber- ichte, 3/1991, or “A Very Low Noise Aerial Ampli- fier”, pp 90-96, VHF Communications, 1992/2. 8. Solar Radio Data, U.S. Dept. of Commerce, NOAA, Space Weather Prediction Center, http://legacy- www.swpc.noaa.gov/ftpdir/lists/radio/7day_rad. txt 9. Robert H. Dicke: “The Measurement of Thermal Radiation at Microwave Frequencies.”, The Review of Scientific Instruments, Vol.17, No.7, pp 268- 275, 1946. 10. James J. Condon, Scott M. Ransom, NRAO: “The HI 21 cm Line”, http://www.cv.nrao.edu/course/ astr534/HILine.html 11. High Definition Software Defined Radio, http:// www.hdsdr.de/ [12] Artist’s impression of the Milky Way (updated - annotated), https://com- mons.wikimedia.org/wiki/File:Artist’s_impres- sion_of_the_Milky_Way_(updated_annotated). jpg Arriived: 13. 03.2017 Accepted: 01. 08. 2017 T. Saje et al; Informacije Midem, Vol. 47, No. 2(2017), 113 – 128 129 Original scientific paper  MIDEM Society Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 129 – 137 Highly Efficient Photocatalytic Activity in the Visible Region of Hydrothermally Synthesized N-doped TiO 2 Maja Lešnik1, Dejan Verhovšek1, Nika Veronovski1, Srđan Gatarić1, Mihael Drofenik2, Janez Kovač3 1Cinkarna Celje, d.d. Inc., Celje, Slovenia 2University of Maribor, Faculty of Chemistry and Chemical Engineering, Maribor, Slovenia 3Jozef Stefan Institute, Department of Surface Engineering and Optoelectronics, Ljubljana, Slovenia Abstract: Nanocrystalline rutile titanium dioxide (TiO2) samples doped with various amounts of nitrogen (N) atoms were prepared using a hydrothermal synthesis route and a polycrystalline TiO2 precursor. The doped rutile nanocrystallites were analysed with transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS) and UV–Vis spectroscopy. The Kubelka–Munk band-gap calculation were used to examine the UV–Vis reflectance spectra. The measurements of the photocatalytic activity were performed utilizing FT-IR. A remarkable increase in the photocatalytic activity of the doped rutile nanocrystallites was detected, when applying the isopropanol degradation method with UV–Vis light irradiation. Keywords: Nanoparticles, TiO2, Rutile, Visible photocatalyst Hidrotermalno sintentiziran TiO 2 dopiran z N z visoko fotokatalitsko aktivnostjo v vidnem delu svetlobnega spektra Izvleček: Z uporabo hidrotermalne sinteze in polikristalinične rutilne oblike nanodelcev TiO2 smo pripravili monokristalinične rutilne delce, dopirane z različnimi koncentracijami dopanta N. Za analizo dopiranih rutilnih TiO2 nanodelcev smo uporabili elektronsko mikroskopijo (TEM), fotoelektronsko spektroskopijo (XPS), rentgensko praškovno difrakcijo (XRD), UV-VIS spektroskopijo in kalkulacijo energijskih vrzeli s Kubelka Munk. Pri merjenju fotokatalitske aktivnosti z metodo degradacije izporopanola v alkohol smo ugotovili, da dopiranje rutilnega TiO2 z N povzroči izrazit premik delovanja fotokatalitske aktivnosti na vidni svetlobi svetlobnega spektra. Ključne besede: Nanodelci, TiO2, rutil, vidni fotokatalizator * Corresponding Author’s e-mail: maja.lesnik@cinkarna.si 1 Introduction Nanocrystalline titanium dioxide (TiO2) is the most widely investigated n-type semiconductor due to its high photocatalytic activity under UV light, which is important for numerous outdoor applications, such as wastewater treatment, air purification and self-clean- ing applications (walls, concretes). [1-8] Recently, new insights have been presented in the development of TiO2 photocatalysts that could efficiently utilize not only the UV light but also the visible light in the solar spectrum and could therefore be appropriate for inte- rior photochemical applications. [8-16] Chemical modi- fications of the TiO2 crystal lattice achieved by doping with cations or anions appear to be the most promis- ing approach to enhance the visible-light absorption power. [17] 130 The dopants can be interstitially or substitutionally incorporated into the TiO2 crystal lattice. Different concentration levels of these dopants might influ- ence the new electronic states localized in the gap or the electronic band edge narrowing, leading to an in- crease in the visible-light absorption efficiency. [14, 17] However, it is well known that the photocatalytic activ- ity under exposure to visible light, associated with the mobility of the excited electrons and holes and their recombination rate, differs, depending on the dopant type, its concentration and the lattice position that it occupies. [14, 17] On the other hand, anion doping has a tremendous effect on visible-light photocatalytically active TiO2. [18, 19] Among all the attempts at non- metal doping in TiO2, nitrogen doping has shown the greatest promise for achieving visible-light active pho- tocatalysts. The incorporation of nitrogen into the TiO2 crystal lattice is advantageous, due to it having a simi- lar atomic radius to oxygen and a lower electronega- tivity than oxygen. [20] The modification mechanism of N-doped TiO2, its ability to absorb visible light and visible-light photocatalysis is still under investigation. There are three different hypotheses that could explain the phenomena. Firstly, in N-doped TiO2 the energies of the N 2p and O 2p states are similar. The consequence of this is band-gap narrowing and the ability to absorb visible light. [22] Secondly, oxygen sites are substituted by nitrogen atoms and the intermediate energy level is formed below the conductive band edge. [21] Thirdly, doping with nitrogen forms oxygen-vacancy defect sites, which are the major factor in visible-light photo- catalytic activity. [23, 24] Rutile, as an n-type TiO2 semi- conductor, exhibits oxygen vacancies on the surface. Nitrogen doping introduces additional oxygen vacan- cies, which leads to an even more efficient photocata- lytic activity. [25] It is well known that the absorption properties of N- doped anatase and raw rutile TiO2 are distinguishable. The structures of anatase and rutile differ in the posi- tion of the octahedron, resulting in a tetragonal struc- ture for both modifications. [3] The other reason for the different absorption is the electron density of the N-doped anatase or rutile. [26] Doping with nitrogen provides N 2p states located above the O 2p valence band. Since rutile has a smaller band gap than anatase, this furthermore enhances the valence band. [27] The same findings were confirmed by Yang and co-authors in their DFT calculations on nitrogen-doped structures of rutile crystals. [15] Liu, in his work, demonstrated that nitrogen-doped TiO2 with more rutile phase has more defects than the nitrogen-doped TiO2 with less rutile phase, which enhances the photocatalytic effi- ciency. [25] The photocatalytic efficiencies of rutile and anatase are related to the formation of hydroxyl radi- cals that prevent electron–hole recombination during exposure to sunlight. Some studies have demonstrated that the rutile crystal phase exhibits enough hydroxyl groups, which are believed to act as light photocata- lysts, i.e., to accept the holes generated by UV illumina- tion and form hydroxyl radicals and thus prevent elec- tron–hole recombination. [28, 29] The selection of nitrogen doping for the rutile crystal structure was based on theoretical studies published recently in the open literature, particularly on the nu- merous advantages of visible-light absorption and en- hanced photocatalytic activity. [26-29] In the present study we report a new synthesis proce- dure of N-TiO2 visible-light photocatalyst based on the hydrothermal synthesis using the polycrystalline rutile TiO2 nanocrystallites. 2 Experimental 2.1 Preparation method The hydrothermal synthesis of N-doped rutile TiO2 na- nocrystallites was performed in a Teflon-lined, stain- less-steel autoclave with a volume of 80 mL. To prepare the TiO2-doped sample the reactor was loaded with a 50-mL aqueous suspension of polycrystalline rutile TiO2 nanocrystallites provided by Cinkarna Celje, Inc., having a mass concentration between 60-150 g/L (cal- culated as TiO2) and 1 mass % (based on TiO2 content) of urea ((NH2)2CO, 99% w/w, Merck). The mixture of polycrystalline rutile TiO2 nanocrystallites and dopant was then stirred for at least 15 minutes. The autoclave was put into a preheated oven and was hydrothermal treated at 180 °C for 24 hours. At the end of the heat- ing process the autoclave was taken out of the oven and left to cool to room temperature. The as-prepared product was diluted with distilled water, washed on a laboratory centrifuge (MPW 350 – Med. Instruments, High brushless centrifuge, 4000 rpm, 20 minutes). The washing was continued until the conductivity of the ef- fluent was less than 900 µS/cm. The final product was an aqueous suspension of doped rutile having 10 mass % of TiO2 nanocrystallites. Samples with urea/TiO2 ra- tios of 0.01, 0.02, 0.03, 0.06, 0.08 w/w, labelled as sam- ples: B, C and D, E, F, were then prepared using the same process by varying the content of added urea. The sam- ple A was prepared using the same process, but with- out the addition of urea as a dopant. 2.2 Characterization of samples The crystallinity of the particles was examined using X- ray diffraction (XRD) performed on a Cubi X PRO PW M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 131 3800 instrument (PANanalytical) (Cu-Kα radiation (λ = 1.5418Å)). In order to acquire the TiO2 powders for the X-ray powder-diffraction (XRD) measurement, the suspensions were dried at 80 °C, ground and the pow- der pressed into pellets that were used to perform the measurements. The average crystallite size was determined using dif- fraction-peak (100) broadening and Scherrer’s formula based on the FWHM (Full Width at Half Maximum) of the XRD peak. The specific surface areas (SBET) of particles were deter- mined using Tristar 3000, the automatic gas analyser (Micromeritics Instrument Co.). The morphology and the size of the particles were ex- amined with a transmission electron microscope (TEM, Jeol JEM-2100, Jeol Ltd.,Tokyo, Japan). The samples for the TEM specimens were ultrasonically dispersed and the suspensions were collected using carbon-support- ed copper grids. The UV–Vis diffuse reflectance spectra were collected on an Agilent-Cary 300 UV–Vis spectrophotometer equipped with an integrating sphere (Varian Inc., USA). The measurements of the photocatalytic activity were performed in a sealed gas-solid reactor at room tem- perature and a relative humidity of 60%, utilizing FT-IR spectroscopy (Spectrum BX model Perkin Elmer spec- trometer). The model pollutant was isopropanol in the gas phase. During the photocatalytic reaction the iso- propanol oxidizes to acetone and subsequently to car- bon dioxide and water under UV irradiation (Xe lamp, 300 W). The light imitates the solar spectrum and emits both ultraviolet (UV) and visible (VIS) light. The reac- tor is at a distance of 4 cm from the lamp. The samples were dried under ambient conditions and prepared by milling 50 mg of the material. To perform the measure- ments, 20 µL of isopropanol was injected into the sys- tem. This volume represents around 2000 ppm of gas phase for the isopropanol in the system. The amount and ratio of isopropanol and the formed acetone were monitored in real time. The evaluation of the photocat- alytic activity is based on the acetone-formation kinet- ics and is given in ppm/h. [33] The chemical composition of the surfaces was deter- mined by X-ray photoelectron spectroscopy (XPS). XPS analyses were performed with a TFA XPS spectrometer, produced by Physical Electronics Inc., equipped with a monochromated Al–Kα X-ray source (1486.6 eV), un- der ultra-high vacuum (10-7 Pa). Samples in the form of powders were deposited on the adhesive carbon tape. The analyzed area was 0.4 mm in diameter and the analyzed depth was 3–5 nm. The high-energy reso- lution spectra were acquired with an energy analyzer operating at a resolution of about 0.6 eV and a pass energy of 29 eV. The XPS spectra were processed with the software MultiPak. Prior to the spectra process- ing, the same spectra were referenced to the C-C/C-H peak in the C 1s core level at a binding energy of 284.8 eV. The accuracy of the binding energies was about ± 0.2 eV. Quantification of the surface composition was based on the XPS peak intensities, taking into account the relative sensitivity factors provided by the instru- ment manufacturer. [38] Three different places were analyzed on each sample and the data were averaged. 3 Results and discussion 3.1 The crystallite phase and size of rutile particles Figure 1: XRD patterns of undoped and N-doped TiO2 samples with different ratios of urea to TiO2. Figure 1 presents the XRD patterns of the samples pre- pared using a modified hydrothermal process and vari- ous ratios of urea to TiO2. The presence of specific peaks (2θ = 27.38°, 36.06°, 41.19°) was taken as an attributive indicator of rutile titania. [20, 31] However, no N-de- rived peak is detected for N-TiO2, even when the ratio of urea to TiO2 was 0.08. It can also be seen from the XRD patterns that the nitrogen-doped samples show more intensive diffraction peaks, indicating a more pronounced crystallinity for the N-doped crystallites. From Table 1 it is clear that the crystallite size increases with the amount of urea in the precursor suspension, i.e., the ratio of urea to polycrystalline TiO2 (w/w) pre- cursor changes from 0.01 to 0.08. The increase of the crystallite size, for identical hydrothermal conditions, due to a larger amount of urea, can be assigned to the vigorous thermally assisted decomposition reaction of the urea, which enhances the kinetics of mass transport during the dissolution, precipitation and growth of TiO2 nano-crystallites. Thus, a crystallite-size increase is straightforward and proportional to the amount of urea in the starting suspension. On the other hand, when the addition is a compound that is stable during the hydrothermal synthesis conditions it would, as ex- pected, hinder the crystallite growth and thus decrease M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 132 the final crystallite size, which is the case when the sup- pression of crystallites size is planned. As a consequence, the specific surface areas (SBET) show a steady decrease in parallel with the crystallite size increase. Here, an exception proves sample B with a much smaller specific surface regarding the general trend in the sequence, which might be a consequence of the exaggerated crystallite agglomeration. However, in general the morphology of the crystallites follows the general expectation. Table 1: Average crystallite size, specific surface area and band-gap energies for various N-doped TiO2 samples. Sample Urea/TiO2 Specific surface area (m2/g) Crystallite size (100) (nm) Band gap (eV) A - 70.1 16.7 3.01 B 0.01 42.5 23.6 3.04 C 0.02 68.1 27.6 3.03 D 0.03 63.7 33.0 3.03 E 0.06 56.9 42.6 3.04 F 0.08 51.0 48.7 3.02 3.2 The morphology of the doped TiO 2 particles The morphologies of the N-doped TiO2 are shown in the TEM micrographs of Figure 2. The hydrothermally synthesized, doped, rutile TiO2 nanocrystallites have an oval/spherical morphology and are uniform in size. The crystallite sizes observed with the TEM match with those obtained from the Scherrer estimation using the peak broadening of the XRD spectra, which has shown a com- parable crystallite size up to a urea/TiO2 ratio of 0.03. On the other hand, the morphologies of samples E and F, prepared at ratios of urea to TiO2 of 0.06 and 0.08, respec- tively, exhibit a larger crystallite size (TEM images not shown). 3.3 UV–Vis diffuse reflectance spectra Figure 3: UV–Vis reflectance spectra of the undoped TiO2 and the TiO2 doped with different urea-to-TiO2 ra- tios, indicated in the legend. For the examination of the effects of doping on TiO2, an evaluation of the optical properties is the most appro- priate method. UV–Vis spectroscopy and diffuse reflec- tance spectroscopy were chosen as the techniques for the optical studies of N-doped TiO2. In our work diffuse reflectance spectroscopy was used to examine the vis- ible-light sensitivity. The influence of nitrogen doping on the UV–Vis spectra properties for the rutile TiO2 is demonstrated in Figure 3. The reflectivity dependence of the wavelength of the pure TiO2 has a typical sharp edge of reflection at around 420–400 nm. Compared with the spectrum of undoped TiO2, the N-doped sam- M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 (a) (b) Figure 2: TEM micrographs of samples prepared with different ratios of urea to TiO2: (a) undoped TiO2 and (b) urea/TiO2 = 0.02. 133 ples exhibit a very similar curve progression; however, there is a small but distinguishable shift in the absorb- ance region of the visible range 400–550 nm. [30, 31, 34] The N-doped samples exhibit a slightly difference in the colour, which could provide a small absorbance in visible region. [41] An exception is observed for the 0.03-doped sample D, which shows a more notable red shift. So, based on the intensity of absorption for all the samples we can assume that the nitrogen entered the TiO2 crystal lattice under the reported hydrothermal condition. The same finding was reported by Huang. [31] It was reported that the visible-light absorption could be brought about by band-gap narrowing. However, it was also reported that the localized N states within the band gap and the Ti3+ defects could also provide the absorption red shift. [34, 35] In addition, Hu showed that the band gaps of the doped samples were the same, indicating that N doping did not change the band gap of the TiO2. [32] The doping of TiO2 with N atoms improved its visible-light absorption, increased the numbers of photons in the photocatalytic reaction and thus enhanced the photocatalytic activity in the visible region. The band-gap energies of the rutile nanocrystallites, estimated using Kubelka–Munk model are summarised in Table 1. [35, 36] The values of the band-gap energies of the doped samples were compared with a control sample (undoped rutile nanocrystallites), which was calculated to have a band gap of 3.01 eV. The calcu- lated value for the undoped rutile nanocrystallites is in agreement with the theoretical value of 3.0 eV for the rutile modification. [37] The results show that the band- gap energies of all the N-doped samples are practically the same as the control sample. A possible explanation is that the visible-light absorption occurs due to the colour centres formed by the N-doping process rather than by a narrowing of the band gap. The research was conducted on various N-doped metal-oxide nanoparti- cles. The band-gap narrowing does not occur, even for significantly high doping levels, such as 25 % doping. [32, 34]. It can be concluded that the main effect of N doping is a slightly improved absorption at long wave- lengths, which enhances the visible photocatalytic ac- tivity of these material. It could be concluded that the main effect of N doping is the improved absorption at long wavelengths due to the shallow trap states inside the TiO2 crystal lattice, which enhances the visible pho- tocatalytic activity of these materials. [22, 24, 25] 3.4 Investigation of chemical states of TiO 2 samples The surface chemical composition and the chemical states of the TiO2 samples were analyzed by means of XPS. The survey spectra (not shown) are similar and indi- cate the presence of Ti, O and C in all the samples, while N is visible only in the spectra from the urea-modified TiO2 samples and confirms a successful treatment. The surface chemical compositions are presented in Table 2. The carbon on the surface of the undoped sample can be related to the surface contamination and the synthesis conditions. For the TiO2 samples treated with urea a nitrogen signal appeared. The highest nitrogen concentration (0.8 at.%) was observed on the surface of the TiO2 sample D (urea/TiO2 ratio 0.03). On the sam- ple treated with a higher urea concentration, sample F (urea/TiO2 ratio 0.08), we observed less nitrogen (~ 0.3 at.%). The amount of nitrogen on the surfaces of the analyzed samples correlates with the photocatalytic activity. High-energy-resolution C 1s, O 1s, N 1s and Ti 2p XPS spectra were acquired to further understand the chemical bonding. In the high-energy-resolution O 1s spectra (not shown here) we were able to observe the presence of two different components by using a fit- ting procedure. The main contribution is attributed to the Ti-O in the TiO2 (529.9 eV) and the other minor peak can be ascribed to the surface hydroxyl Ti-OH (531.4 eV). [39] A comparison of the O 1s spectra from the undoped sample with the treated samples shows no major differences. Nitrogen was only detected in the urea-treated TiO2 samples. High-energy-resolution N 1s XPS spectra from the undoped TiO2 and the TiO2/urea ratio of 0.03 are shown on Figure 4. The maxima of the N 1s spectra, for all the treated samples, are located at 400 eV, which indicates interstitial nitrogen integrated into the TiO2 lattice. It is known that the peak at around 400 eV is related to the N-O, N-C or N-N type of bonds. [40] A comparison of the high-energy-resolution Ti 2p spec- tra from all the analysed samples is shown in Figure 5. In the acquired Ti 2p spectra a doublet peak is visible, containing both Ti 2p3/2 and Ti 2p1/2 components, which appear at 458.6 eV and 464.3 eV, respectively, with 5.7 eV spin-orbital splitting. This corresponds to a Ti4+ va- Table 2: Surface composition in at. % of the undoped TiO2 and TiO2 modified with urea using different con- centrations. Sample Urea/TiO2 C O Ti N A - 26.1 52.2 21.7 B 0.01 21.5 55.8 22.3 0.3 C 0.02 23.4 54.6 21.4 0.7 D 0.03 24.8 53.4 21.1 0.8 E 0.06 20.0 56.7 22.7 0.6 F 0.08 25.2 53.1 21.4 0.3 M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 134 lence state. The peaks are narrow and no significant dif- ferences, like shifting in the binding energy, between the undoped and treated samples were observed (Fig- ure 5). Figure 4: N 1s XPS spectrum of undoped TiO2 (a) and TiO2 modified with a TiO2/urea ratio = 0.03 (b). Figure 5: XPS spectra of Ti 2p from all the samples. 3.5 Photocatalytic activity measurements To evaluate the photocatalytic activity of the undoped and N-doped TiO2 in the visible range, the degradation of isopropanol under UV+VIS and Vis irradiation was in- vestigated. The results of the photocatalytic activities are presented in Figure 6, based on the acetone-for- mation kinetics, and are given in ppm/h. As illustrated in Figure 6, different N-doped TiO2 catalysts differ in the degradation of isopropanol under the same ex- perimental conditions. One can see that i) in general, N-doped TiO2 samples achieve a higher photocatalytic activity than the undoped TiO2 samples and ii) the pho- tocatalytic activity increases with the surface-nitrogen concentration. Among all of the investigated N-doped TiO2 samples, sample E, with a urea/TiO2 ratio of 0.06 and a corre- sponding surface-nitrogen concentration of 0.6 at %, displays the highest photocatalytic efficiency for iso- propanol degradation. Nearly the same photocatalytic efficiency was also detected with the sample C, hav- ing otherwise a lower urea/TiO2 ratio of 0.02; however, it exhibits a similar surface-nitrogen concentration of 0.7 at %. In addition, sample E exhibits a lower specific surface area than the sample C with a urea/TiO2 ratio of Figure 6: Photocatalytic activity of undoped TiO2 and TiO2 doped with different ratios of urea under a) UV+Vis irradiation and b) Vis irradiation. (a) (b) M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 135 0.02, Table 1. Therefore, the high photocatalytic activ- ity of sample C with a surface nitrogen concentration of 0.7 at % is not a consequence of a higher specific surface area, but the result of a high surface-nitrogen concentration. This is in accordance with the general trend that the N∙ centres enhanced the photocatalytic activity in the visible range. As a result, an increase in the surface area does not automatically produce an increase in the photocata- lytic activity, demonstrating that the higher activity is a consequence of a high surface-nitrogen concentration and not of the surface-regulated process. On the other hand, a greater surface area provides more active sites on the TiO2 surface for the degradation of the organic pollutant. [20, 32] 4 Conclusions N-doped rutile TiO2 nanocrystallites that exhibit a strong increase in their photocatalytic activity were successfully prepared using the hydrothermal method. The absorbance of N-TiO2 in the visible-light region is the most important when concerning the material’s application since it can be activated with solar light and thus exhibits an enhanced photocatalytic visible-light activity. The narrowing of the band gap does not occur, indicating that the major effects of N doping are an en- hanced absorption at long wavelengths and the hole- trapping sites, which retards the hole–electron recom- bination and might be useful in enhancing the visible photocatalytic activity of these materials. The maxima of the N 1s spectra, for all the treated samples, indicate that the interstitial nitrogen is integrated into the TiO2 lattice. The N-doped TiO2 samples achieved a higher photocatalytic activity in the UV and visible-light re- gions than the undoped sample. This higher photocat- alytic activity has a close correlation with the enhanced visible-light absorption of the doped samples. 5 Acknowledgment This work is partially financially supported by the Min- istry of Education, Science and Sport of the Republic of Slovenia under the contract 34/2013-SOF. The authors declare that there is no conflict of interest regarding the publication of this paper. 6 References 1. M. R. Hoffmann, S. T. Martin, W. Choi, D. W. Bah- nemann, Environmental applications of semicon- ductor photocatalysis, Chem. Rev., 1995, 95(1), 69-96 2. A. Fujishima, T. N. Rao, D. A. Tryk, Titanium dioxide photocatalysis, Journal of Photochemistry and Photobiology C: Photochemistry, 2000, 1, 1-21 3. M. Palaez, N. T. Nolan, S. C. Pillai, M. K. Seery, P. Falaras, A. G. Kontos, P. S. M. Dunlop, J. W. J. Ham- ilton, J. A. Byrne, K. O’Shea, M. H. Entezari, D. D. Dionysiou, A review on the visible light active ti- tanium dioxide photocatalysts for environmental applications, Applied Catalysts B: Environmental, 2012, 125, 331-349 4. X. Chen, S. S. Mao, Titanium dioxide nanomateri- als: synthesis, properties, modifications and appli- cations, Chem. Rev., 2007, 107, 2891-2959 5. D. P. Macwan, P. N. Dave, S. Chatuverdi, A review on nano-TiO2 sol gel synthesis and its applica- tions, J. Mater. Science, 2011, 46, 3669-3686 6. S. Banerjee, D. D. Dionysiou, S. C. Pillai, Self-clean- ing applications of TiO2 by photo-induced hydro- philicity and photocatalysis, Applied Catalysis B: Environmental, 2015, 176-177, 396-428 7. W. Shen, C. Zhang, Q. Li, W. Zhang, L. Cao, J. Ye, Preparation of titanium dioxide nano particle modified photo catalytic self-cleaning concrete, Journal of cleaner production, 2015, 87, 762-765 8. S. M. Gupta, M. Tripathi, A review of TiO2 nano- particles, Chinese Science Bulletin, 2011, 56 (16), 1639-1657 9. S. Banerjee, S. C. Pillai, P. Falaras, K. E. O’Shea, J. A. Byrne, D. D. Dionysiou, New insights into the mechanism of visible light photocatalysis, The Journal of Physical Chemistry Letters, 2014, 5, 2543-2554 10. H. Feng, M. Zhang, L. E Yu, Hydrothermal synthe- sis and photocatalytic performance of metal-ions doped TiO2, Applied Catalysis A: General, 2012, 413-414, 238-244 11. G. Liu, L. Wang, H. Gui Hang, H.-M. Cheng, G. Q. Lu, Titania-based photocatalysts-crystal growth, doping and heterostructuring, Journal of Materi- als Chemistry, 2010, 20, 831-843 12. A. Zaleska, Doped-TiO2: A review, Recent Patents on Engineering, 2008, 2, 157-164 13. S. G. Kumar, L. G. Devi, Review on modified TiO2 photocatalysis under UV/visible light selected results and related mechanisms on interfacial charge carrier transfer dynamics, The Journal of Physical Chemistry A, 2011, 115, 13211-13241 14. M. V. Dozzi, E. Selli, Doping TiO2 with p-block ele- ments: Effects on photo catalytic activity, Journal M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 136 of Photochemistry and Photobiology C: Photo- chemistry, 2013, 14, 13-28 15. Y. Dai, K. Yang, B. Huang, S. Han, Theoretical study of N-doped TiO2 rutile crystals, J. Phys, Chem B, 2006, 110, 24011-24014 16. S. Rehman, R. Ullah, A. M. Butt, N.D. Gohar, Strate- gies of making TiO2 and ZnO visible light active, Journal of Hazardous Materials, 2009, 170, 560- 569 17. M. Fernandez-Garcia, A. Martinez-Arias, J. C. Conesa, Visible light – responsive titanium oxide photocatalysts: preparation based on chemical method, Environmentally Benign Photocatalysts, Nanostructure Science and Technology, 2010 18. C. Burda, Y. Lou, X. Chen, A. C. S. Samia, J. Stout, J. L. Gole, Enhanced nitrogen doping in TiO2 nano- particles, Nano Letters, 2003, 3 (8), 1049-1051 19. M. Long, W. Cai, Visible light responsive TiO2 mod- ification with non-metal elements, Frontiers of Chemistry in China, 2011, DOI 10.1007/s11458- 011-0243-8 20. J. Ananpattarachai, P. Kajitvichyanukul, S. Seraphin, Visible light absorption ability and pho- to catalytic oxidation activity of various interstitial N-doped TiO2 prepared from different nitrogen dopants, Journal of Hazardous Materials, 2009, 168, 253-261 21. Z. L. Zeng, First principle study on the structural and electronic properties of N atoms doped rutile TiO2 of oxygen vacancies, Advances in Materials Science and Engineering, 2014, 2015, 10 pages 22. M. Batzill, E. H. Morales, U. Diebold, Influence of nitrogen doping on the defect formation and sur- face properties of TiO2 rutile and anatase, Physical Review Letters, 2006, 96 (2), 026103 (4) 23. B. Ohtani, Titania photocatalysis beyond recom- bination: A critical review, Catalysts, 2013, 3, 942- 953 24. T. Ihara, M. Miyoshi, Y. Iriyama, O. Matsumoto, S. Sugihara, Visible light active titanium dioxide photocatalyst realized by an oxygen-deficient structure and by nitrogen doping, Applied Cataly- sis B: Environmental, 2003, 42, 403-409 25. G. Liu, X. Wang, Z. Chen, H. M. Cheng, G. Q. Lu, The role of crystal phase in determining photo cata- lytic activity of nitrogen doped TiO2, Journal of Colloid and Interface Surface, 2009, 329, 331-338 26. C. Di Valentin, G. Pacchioni, A. Selloni, Origin of the different photocatalytic activity of N-doped anatase and rutile TiO2, Physical review B, 2004, 70, 085116 27. O. Diwald, T. L. Thompson, E. G. Goralski, S. D. Wal- ck, J. T. Yates, Jr., The effect of nitrogen ion implan- tation on the photo catalytic activity of TiO2 rutile single crystals, J. Phys. Chem B, 2004, 108, 52-57 28. J. Lin, J. C. Yu, D. Lo, S. K. Lam, Photo catalytic ac- tivity of rutile Ti1-xSnxO2 solid solutions, Journal of Catalysis, 1999, 183, 368-372 29. H. Liu, L. Gao, Codoped rutile TiO2 as a new pho- tocatalyst for visible light irradiation, Chemistry Letters, 2004, 33 (6), 730-731 30. F. Peng, L. Cai, L. Huang, H. Yu, H. Wang, Prepa- ration of nitrogen-doped titanium dioxide with visible light photo catalytic activity using a fac- ile hydrothermal method, Journal of Physics and Chemistry of Solids, 2008, 69, 1657-1664 31. D. Huang, S. Liao, S. Quan. L. Liu, Z. He, J. Wan, W. Zhou, Synthesis and characterization of visible light responsive N-TiO2 mixed crystal by a modi- fied hydrothermal process, Journal of Non-Crys- talline Solids, 2008, 354, 3965-3975 32. S. Hu, A. Wang, X. Li, H. Löwe, Hydrothermal syn- thesis of well-dispersed ultrafine N-doped TiO2 nanoparticles with enhanced photo catalytic ac- tivity under visible light, Journal of Physics and Chemistry of Solids, 2010, 71, 156-162 33. M. Tashibi, U. Lavrenčič Štangar, A. Sever Škapin, A. Ristić, V. Kaučič, N. Novak Tušar, Titania-con- taining mesoporous silica powders: Structural properties and photocatalytic activity towards isopropanol degradation, Journal of Photochem- istry and photobiology A: Chemistry, 2010, 216, 167-178 34. X. Qiu, C. Burda, Chemically synthesized nitrogen doped metal oxide nanoparticles, Chemical Phys- ics, 2007, 339, 1-10 35. M. D’Arienzo, R. Scotti, L. Wahba, C. Battocchio, E. Bemporad, A. Nale, F. Morazzoni, Hydrothermal N-doped TiO2: Explaining photo catalytic proper- ties by electronic and magnetic identification of N active sites, Applied Catalysis B: Environmental, 2009, 93, 149-155 36. S. Valencia, J.M. Marin, G. Restrepo, Study of the bandgap of synthesized titanium dioxide nanoparticles using the sol-gel method and a hydrothermal treatment, The Open Materials Sci- ence, 2010, 4, 9-14 37. M. Landmann, E. Rauls, W.S. Schmidt, The elec- tronic structure and optical response of rutile, anatase and brookite TiO2, J. Phys.: Condens. Mat- ter, 2012, 24 195503 38. J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomb- en, “Handbook of X-Ray Photoelectron Spectros- copy”, Physical Electronics Inc., Eden Prairie, Min- nesota, USA, 1995 39. J. Yu, H. Yu, B. Cheng, M. Zhou, X. Zhao, Enhanced photocatalytic activity of TiO2 powder (P 25) by hydrothermal treatment, Journal of Molecular Ca- talysis A:Chemical, 2006, 253, 112-118 40. Y. Nosaka, M. Matsushita, J. Nishino, A. Y. Nosaka, Nitrogen doped titanium dioxide photocatalysts M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 137 for visible response prepared by using organic compounds, Science and Technology of Ad- vanced Materials, 2005, 6, 143-148. 41. P. He, J. Tao, X. Hunag, J. Xue, Preparation and photocatalytic antibacterial property of nitrogen doped TiO2 nanoparticles, Sol-Gel Sci. Technol. 2013, 68, 213-218 Arrived: 23. 05. 2017 Accepted: 01. 08. 2017 M. Lešnik et al; Informacije Midem, Vol. 47, No. 2(2017), 129 – 137 138 139 Original scientific paper  MIDEM Society Doktorske disertacije na področju mikroelektronike, elektronskih sestavnih delov in materialov v Sloveniji v letu 2016 Doctoral theses on Microelectronics, Electronic Components and Materials in Slovenia in 2016 Univerza v Ljubljani / University of Ljubljana Fakulteta za elektrotehniko / Faculty of Electrical Engineering 1 Avtor Vladimir Furlan Author Naslov Frekvenčno nastavljiva antena z visokim izkoristkom na feroelektriku Title Frequency-agile, highly-efficient antenna on ferroelectric substrate Mentor prof. dr. Matjaž Vidmar Supervisor 2 Avtor Mario Trifković Author Naslov Zmanjšanje preklopnega šuma v sinhronih digitalnih vezjih na osnovi razporejanja signala ure Title Switching noise reduction in synchronous digital circuit based on clock skew scheduling Mentor prof. dr. Drago Strle Supervisor 3 Avtor Tine Dolžan Author Naslov Napredne membranske mikročrpalke s piezoelektričnim vzbujanjem na osnovi PDMS elastomera Title Advanced piezoelectrically actuated membrane micropumps based on PDMS elastomer Mentor dr. Danilo Vrtačnik Supervisor 4 Avtor Matija Podhraški Author Naslov Integrirani mikrosenzorski sistemi z mikrotuljavicami Title Integrated microsensor systems with microcoils Mentor prof. dr. Janez Trontelj Supervisor Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 139 – 142 140 5 Avtor Uroš Nahtigal Author Naslov Integrirani barvni senzor svetlobe z velikim dinamičnim območjem Title Integrated color light sensor with high dynamic range Mentor prof. dr. Drago Strle Supervisor 6 Avtor Luka Bogataj Author Naslov Zasnova in izvedba visokostabilnega optoelektronskega oscilatorja Title Design and implementation of a highly-stable optoelectronic oscillator Mentor prof. dr. Matjaž Vidmar Supervisor 7 Avtor Blaž Kirn Author Naslov Zmogljivost sončnih fotonapetostnih elektrarn skozi življenjski cikel Title Life cycle performance of solar photovoltaic power plants Mentor prof. dr. Marko Topič Supervisor Univerza v Mariboru / University of Maribor Fakulteta za elektrotehniko, računalništvo in informatiko/Faculty of Electrical Engineering and Computer Science 1 Avtor Janko Horvat Author Naslov Regulirana prožilna stopnja za močnostne tranzistorje Title Controlled gate driver for power transistors Mentor prof. dr. Miro Milanovič Supervisor 2 Avtor Aleksandar Dodić Author Naslov Korekcijski faktorji koeficientov digitalnih struktur pri spremembi frekvence vzorčenja Title Correctio factors of digital structure coefficients at sampling frequency change Mentor prof. dr. Zmago Brezočnik Supervisor Doktorske disertacije/Doctoral theses; Informacije Midem, Vol. 47, No. 2(2017), 139 – 142 141 3 Avtor Uroš Pešović Author Naslov Model verjetnosti pogreškov IEEE 802.15.4 brezžičnega prenosa pri so-kanalski interferenci in šumu ozadja Title Error probability model for IEEE 802.15.4 wireless transmission with co-channel interference and background noise Mentor prof. dr. Peter Planinšič Supervisor 4 Avtor Marijan Španer Author Naslov Superkondenzator in energijska izkoriščenost baterijsko napajanih vozil Title Supercapacitor and energy efficiency of battery powered vehicles Mentor prof. dr. Karel Jezernik Supervisor Mednarodna podiplomska šola Jožefa Stefana / Jožef Stefan International Postgraduate School 1 Avtor Rok Rudež Author Naslov Razvoj debeloplastne oksidne elektronske keramike Title Development of thick-film oxide-based electronic ceramics Mentor prof. dr. Slavko Bernik Supervisor 2 Avtor Mojca Presečnik Author Naslov Mikrostrukturne in termoelektrične lastnosti keramike tipa p v sistemu Ca-Co-O Title Microstructural and thermoelectric characteristics of p-type ceramics in the Ca-Co-O system Mentor prof. dr. Slavko Bernik Supervisor 3 Avtor Tina Bakarič Author Naslov Priprava porozne keramike Pb(Zr0.53Ti0.47)O3Pb(Zr0.53Ti0.47)O3 z načrtovano mikrostrukturo in oblikovanje debelih plasti z brizgalnim tiskanjem Title Processing of porous Pb(Zr0.53Ti0.47)O3Pb(Zr0.53Ti0.47)O3 ceramics with a designed microstructure and patterning of thick films by inkjet printing Mentor Dr. Danjela Kuščer Hrovatin Somentor Dr. Tadej Rojac Supervisor Co-Supervisor Doktorske disertacije/Doctoral theses; Informacije Midem, Vol. 47, No. 2(2017), 139 – 142 142 4 Avtor Jitka Hreščak Author Naslov Sinteza in karakterizacija nedopirane in s stroncijem dopirane keramike na osnovi kalijevega natrijevega niobata Title Synthesis and characterization of undoped and strontium-doped potassium sodium niobate ceramics Mentor Dr. Andreja Benčan Somentor Prof. dr. Barbara Malič Supervisor Co-Supervisor 5 Avtor Marko Vrabelj Author Naslov Raziskave elektrokaloričnega pojava v polikristaliničnem relaksorskem feroelektriku 0.9Pb(Mg1/3Nb2/3)O3−0,1PbTiO30.9Pb(Mg1/3Nb2/3)O3−0,1PbTiO3 Title Investigations of the electrocaloric effect in the polycrystalline 0.9Pb(Mg1/3Nb2/3)O3−0.1PbTiO30.9Pb(Mg1/3Nb2/3)O3−0.1PbTiO3 relaxor ferroelectric Mentor Prof. dr. Barbara Malič Supervisor 6 Avtor Tanja Pečnik Author Naslov Mikrostruktura in dielektrične lastnosti tankih plasti (Ba,Sr)TiO3(Ba,Sr)TiO3, pripravljenih s sintezo v raztopini Title Microstructure and dielectric properties of solution-derived (Ba,Sr)TiO3(Ba,Sr)TiO3 thin films Mentor Prof. dr. Barbara Malič Supervisor 7 Avtor Evgeniya Khomyakova Author Naslov Integracija debelih plasti bizmutovega ferita na keramične in kovinske podlage z metodo sitotiska Title Integration of screen-printed bismuth ferrite thick films onto ceramic and metal substrates Mentor Dr. Andreja Benčan Golob Somentor Dr. Tadej Rojac Supervisor Co-Supervisor Bold font is used for the original title of the thesis (and language in which the thesis is written), normal font for title translation. Doktorske disertacije/Doctoral theses; Informacije Midem, Vol. 47, No. 2(2017), 139 – 142 143 Journal of Microelectronics, Electronic Components and Materials Vol. 47, No. 2(2017), 143 – 143 Announcement and Call for Papers October 4th – 6th, 2017 Jožef Stefan Institute, Ljubljana, Slovenia ORGANIZER: MIDEM Society - Society for Microelec- tronics, Electronic Components and Materials, Ljublja- na, Slovenia CONFERENCE SPONSORS: Slovenian Research Agen- cy, Republic of Slovenia; IMAPS, Slovenia Chapter; IEEE, Slovenia Section; GENERAL INFORMATION The 53rd International Conference on Microelectronics, Electronic Components and Devices with the Work- shop on Materials for Energy Conversion and Their Applications continues a successful tradition of the annual international conferences organised by the MIDEM Society, the Society for Microelectronics, Elec- tronic Components and Materials. The conference will be held at Jožef Stefan Institute, Ljubljana, Slovenia, leading Slovenian scientific research institute, from OCTOBER 4th – 6th, 2017. Topics of interest include but are not limited to: - Workshop focus: Materials for Energy Conversion and Their Applications: - Electrocalorics and Thermoelectrics - Novel monolithic and hybrid circuit processing techniques, - New device and circuit design, - Process and device modelling, - Semiconductor physics, - Sensors and actuators, MIDEM 2017 53rd INTERNATIONAL CONFERENCE ON MICROELECTRONICS, DEVICES AND MATERIALS WITH THE WORKSHOP ON MATERIALS FOR ENERGY CONVERSION AND THEIR APPLICATIONS - Electromechanical devices, Microsystems and na- nosystems, - Nanoelectronics - Optoelectronics, - Photonics, - Photovoltaic devices, - New electronic materials and applications, - Electronic materials science and technology, - Materials characterization techniques, - Reliability and failure analysis, - Education in microelectronics, devices and mate- rials. ABSTRACT AND PAPER SUBMISSION: Prospective authors are cordially invited to submit up to 1 page abstract before May 1st, 2017. Please, iden- tify the contact author with complete mailing address, phone and fax numbers and e-mail address. After notification of acceptance (June 15th, 2017), the authors are asked to prepare a full paper version of six pages maximum. Papers should be in black and white. Full paper deadline in PDF and DOC electronic format is: August 31st, 2017. IMPORTANT DATES: Abstract deadline: May 1st, 2017 (1 page abstract or full paper) Notification of acceptance: June 15th, 2017 Deadline for final version of manuscript: August 31st, 2017 Invited and accepted papers will be published in the conference proceedings. Deatailed and updated information about the MIDEM Conferences is available at http://www.midem-drustvo.si/ under Conferences. Call for papers Boards of MIDEM Society | Organi društva MIDEM MIDEM Executive Board | Izvršilni odbor MIDEM President of the MIDEM Society | Predsednik društva MIDEM Prof. Dr. Marko Topič, University of Ljubljana, Faculty of Electrical Engineering, Slovenia Vice-presidents | Podpredsednika Prof. Dr. Barbara Malič, Jožef Stefan Institute, Ljubljana, Slovenia Dr. Iztok Šorli, MIKROIKS, d. o. o., Ljubljana, Slovenija Secretary | Tajnik Olga Zakrajšek, UL, Faculty of Electrical Engineering, Ljubljana, Slovenija MIDEM Executive Board Members | Člani izvršilnega odbora MIDEM Darko Belavič, In.Medica, d.o.o., Šentjernej, Slovenia Dr. Slavko Bernik, Jožef Stefan Institute, Ljubljana, Slovenia Dr. Miha Čekada, Jožef Stefan Institute, Ljubljana, Slovenia Prof. DDr. Denis Đonlagič, UM, Faculty of Electrical Engineering and Computer Science, Maribor, Slovenia Prof. Dr. Leszek J. Golonka, Technical University Wroclaw, Poland Dr. Vera Gradišnik, Tehnički fakultet Sveučilišta u Rijeci, Rijeka, Croatia Leopold Knez, Iskra TELA d.d., Ljubljana, Slovenia mag. Mitja Koprivšek, ETI Elektroelementi, Izlake, Slovenia Prof. Dr. Miran Mozetič, Jožef Stefan Institute, Ljubljana, Slovenia Prof. Dr. Janez Trontelj, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Dr. Danilo Vrtačnik, UL, Faculty of Electrical Engineering, Slovenia Supervisory Board | Nadzorni odbor Prof. Dr. Franc Smole, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia prof. dr. Drago Strle, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Igor Pompe, Ljubljana, Slovenia Court of honour | Častno razsodišče Emer. Prof. Dr. Jože Furlan, Slovenia Dr. Marko Hrovat, Slovenia Dr. Miloš Komac, Slovenia Informacije MIDEM Journal of Microelectronics, Electronic Components and Materials ISSN 0352-9045 Publisher / Založnik: MIDEM Society / Društvo MIDEM Society for Microelectronics, Electronic Components and Materials, Ljubljana, Slovenia Strokovno društvo za mikroelektroniko, elektronske sestavne dele in materiale, Ljubljana, Slovenija www.midem-drustvo.si