M. MA^EK, A. OBLAK: Ni-Fe ALLOY THIN FILMS FOR AMR SENSORS
499–501
Ni-Fe ALLOY THIN FILMS FOR AMR SENSORS
TANKE PLASTI ZLITINE Ni–Fe ZA IZDELAVO AMR-SENZORJEV
Marijan Ma~ek1, Alen Oblak2
1University of Ljubljana, Faculty of Electrical Engineering, Tr`a{ka 25, 1000 Ljubljana, Slovenia
2LEK, Verov{kova 57, 1000 Ljubljana, Slovenia
marijan.macek@fe.uni-lj.si
Prejem rokopisa – received: 2016-07-15; sprejem za objavo – accepted for publication: 2016-09-19
doi:10.17222/mit.2016.203
Ni-Fe alloy films deposited in sputtering apparatus and the subsequent AMR sensor fabrication were successfully developed.
The results on test devices are comparable to the commercially available device with AMR as measured on the standard
multipole magnetic tape.
Keywords: Ni-Fe alloy, AMR, sputtering
Razvit je bil postopek napr{evanja in kasnej{e obdelave tankih plasti zlitine Ni-Fe, potrebnih za izdelavo senzorjev AMR.
Rezultati testnih uporov, merjenih na standardnem, mnogopolnem magnetnem traku, so primerljivi s komercialnimi senzorji.
Klju~ne besede: zlitina Ni-Fe, AMR, napr{evanje
1 INTRODUCTION
Magnetoresistance (MR) is a phenomenon that
reflects the effect of an external magnetic field on the
resistance change of certain materials. Among known
MR the anisotropic magnetoresistance phenomenon
(AMR) is widely used in the field of position sensors and
encoders due to its robustness compared to optical ones.
The AMR effect arises from spin–orbital coupling and
therefore depends on the angle  between the magnetic
moment M and the current j.1 Therefore, resistivity  is a
function of  and can be written as in Equation (1),
where the symbols  and I denote the perpendicular and
parallel components:
    = + ⊥sin cos2 2 (1)
The response, i.e., the difference between both com-
ponents of resistivity, of AMR sensors strongly depends
on the soft ferromagnetic material properties. A very
common choice are NixFe1–x alloys, known under
commercial name Ni–Fe alloy with typical concentration
x close to 0.8. The properties of thin layers however de-
pend on the deposition parameters (pressure, tempera-
ture, deposition rate, applied magnetic field), subsequent
annealing and of course on the layer thickness.
Conditions should be set in such a way as to obtain a
highly oriented crystalline structure. In the absence of an
external field during the film’s deposition the correct
deposition condition leads usually to the (111) preferred
orientation of film with the easy axis of magnetisation
perpendicular to the surface. For other orientations of the
easy axis, an external magnetic field of 5–15 kA/m
during deposition or subsequent annealing is essential.
In our work we concentrate the efforts in order to
sputter deposit thin films of Ni–Fe from a Ni80Fe20 target
without an external magnetic field. Therefore, we are
limited in variation of deposition parameters (power P
and Ar pressure pAr), which influence the deposition rate
and the chemical constitution1 of the deposited films and
on to the effects of subsequent annealing at temperatures
from 150 °C to above 400 °C in forming gas (10 % H2 in
N2).
2 EXPERIMENTAL PART
Thin Ni–Fe films were deposited onto 500-nm-thick
thermally grown SiO2 on 100-mm Si test wafers. The
depositions were performed in Perkin Elmer Sputtering
Model 2400 apparatus working at a base pressure below
1.3×10–4 Pa from a Ni80Fe20 target. The pressure of the
Ar during the sputtering was set to 1.1 Pa as first results
suggest that higher pressures give unsatisfactory results.
The Ni–Fe alloy films were usually deposited onto a
buffer layer in order to promote preferential (111)
crystallite orientation. A very common buffer is Ta, but
(Ni0.81Fe0.19)0.66Cr0.34 layer2 gives better results. We tried
to simulate the NiFeCr depositing thin layer of
Cr/NiFe/Cr in the way to reach preferred ratio between
Cr and Ni–Fe alloy.
The resistors were defined by the so-called lift-off
process since with standard positive PR process and
subsequent etching in Ni etchant the minimum dimen-
sion could not be kept in the desired limits. For test
resistors devices with two different geometries LxW
(500×25 μm2, 150×3 μm2) from the laboratory test
vehicles. After the finished "lift-off" step test and device
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Materiali in tehnologije / Materials and technology 51 (2017) 3, 499–501 499
UDK 669.24:669.15:621.793 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 51(3)499(2017)
wafers were annealed in quartz tube in forming gas
(10 % H2 in N2) atmosphere at different temperatures
200–420 °C. The device wafers were finished by the
standard double metal process with the highest tempera-
ture 350 °C during inter-metal dielectric deposition.
The properties of the AMR layers were analyzed on
test samples by XRD analysis and magnetic measure-
ments by a Lake Shore 7307 vibrating-sample magneto-
meter. On the same test wafers sheet resistance and
surface roughness were measured after different anneal-
ing steps. The sheet resistance was measured by a
4–point probe and roughness by a laser scanning micro-
scope.
The AMR measurements of the resistors were made
in a measured magnetic field of permanent magnet or
multipole magnetic tape with periodic magnetization as
used for standard applications with 2 = 4 mm. The
results on the tape were also compared to the results of
single resistor from commercial AMR encoder bridge
structure.
3 RESULTS AND DISCUSSION
The AMR effect is inversely proportional to the
thickness, but unfortunately the coercivity Hc limits the
lower thickness to about 50 nm.1 The layer thickness t
can be calculated from the measured sheet resistance Rsh
and the known resistivity , t = /Rsh. Unfortunately, in
thin films the resistivity depends on its thickness and
structure as modeled by 3. The accepted values are
123±30 nm,4 but H. Faltin5 observed even 196 nm for
120-nm-thick films deposited onto Si. Not at least,
within the fixed geometry the resistivity of single resistor
of about 1 k is the optimum one. Therefore, an opti-
mum Rsh is between 5 /sq and 10 /sq.
To characterize the AMR films their sheet resistance
Rsh was measured after different processing steps, and
after isochronal (30 min) annealing in forming gas. The
results are shown in Figure 1. There is a rapid change in
Rsh for the samples deposited without a buffer layer. The
layers deposited onto the buffer exhibit a moderate
reduction in the resistivity up to the annealing at 350 °C,
which is also the highest temperature in the process.
Generally speaking, samples without buffer, and samples
deposited with lower power have a higher Rsh. Since the
sputtering energy P.t was set to 17 kJ the difference is in
the layer structure. The final Rsh was 8–12 /sq, close to
our demands for an AMR sensor. Annealing at 400 °C
and 420 °C does not significantly change the resistance,
but microscopic investigations show unwanted changes
in the structure of the film and an increase in the surface
roughness Sra from 10–20 nm to over 50 nm, especially
for films deposited without a buffer.
The reduction of Rsh after annealing at 350 °C
indicates strong changes in the crystal structure of the
Ni–Fe alloy layer. The XRD analysis in Figure 2 shows
changes in the preferred (111) orientation after annealing
up to 350 °C. Annealing at higher temperatures has no
effect on the orientation of the crystallites. In the XRD
pattern a weak (200) Cr peak is also observed as well a
strong (400) Si. Obviously our presumption for buffer
layers does not work. In any case, the Cr buffer has some
beneficial effects on the surface roughness and it signifi-
cantly reduces the sheet resistivity of at least the
annealed samples.
The magnetic properties of the deposited AMR layers
are shown in next Figure 3. From rather rough plot we
can see a distinctive difference between the plots for
easy axis of magnetization (full circles), perpendicular
orientation of magnetic field to the surface and hard axis
(open circles), lying parallel to the surface. From this
plot the coercivity Hc is estimated to be 5kA/m. Since
these are the very first experiments, the effects of pro-
M. MA^EK, A. OBLAK: Ni-Fe ALLOY THIN FILMS FOR AMR SENSORS
500 Materiali in tehnologije / Materials and technology 51 (2017) 3, 499–501
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 2: XRD pattern for a layer deposited onto buffer layer at 400
W and pressure 1.1 Pa after deposition and isochronal annealing at
(250, 350 and 420) °C
Slika 2: XRD-vzorec plasti, nanesene pri 400 W in tlaku 1.1 Pa na
vmesno plast po nanosu in izokronalnih popu{~anjih na (250, 350 in
420) °C
Figure 1: Rsh measured after the deposition and isochronal annealing
in forming gas; crosses denote sheet resistances for films deposited
without buffer layer
Slika 1: Plastna upornost Rpl po nanosu in izokronalnih popu{~anjih v
tvornem plinu; kri`i predstavljajo plastne upornosti filmov, nanesenih
brez vmesne plasti
cessing parameters on to magnetic properties and the
structure revealed by XRD of soft ferromagnetic layers
have not been studied yet.
The first experimental AMR resistors show promis-
ing results. In Figure 4 the responses R/Ravg of the two
test resistors LxW = 500×25 μm with slight differences in
the processing (W#10 without buffer, W#08 with buffer
layer) and a commercial AMR encoder are shown. The
measurements were performed on a commercially avail-
able multipole magnetic tape with 2 = 4 mm in steps of
0.1 mm and at a distance of 1.5 mm above the tape with
a manual manipulator in all cases. From Figure 4 it is
clear that there is no significant difference between both
test samples, but their responses (~1.7 %) are somewhat
lower than that on the commercial (~1.9 %) sensors.
4 CONCLUSIONS
The first AMR resistors were successfully fabricated
by sputtering of Ni–Fe alloy Ni80Fe20 in a Perkin Elmer
apparatus and the necessary process steps for the
subsequent annealing and fabrication of the resistors
with predefined geometries with minimum dimensions
W = 3 μm were established.
The results from tests of the AMR resistors were
comparable to the commercial one and encourages us to
proceed with the development of extrapolators of custom
design.
Acknowledment
The authors acknowledge RLS company and
especially director Mr. J. Novak and their employees
Mrs. D. Domanjko and Mr. M. Jane`i~ for support and
encouraging during the first steps of AMR material
process research and work on the design of the test
encoder. Authors acknowledge dr. S. Gyergyek and dr.
M. Spreitzer of the Jozef Stefan Institut, Ljubljana, for
magnetic properties and XRD measurements.
5 REFERENCES
1 A. V. Svalov, R. Aseguinolaza, A. Garcia-Arribas, I. Orue, J. M.
Barandiaran, J. Alonso, M. L. Fernández-Gubieda, G. V. Kurlyand-
skaya, Structure and Magnetic Properties of Thin Ni–Fe alloy Films
Near the Transcritical State, IEEE Trans Magn, 46 (2010) 2,
333–336
2 W. Y. Lee, M. F. Toney, D. Mauri, P. Tameerug, E. Allen, High Mag-
netoresistance Permalloy Films Trough Growth on Seed layers of
(Ni0.81Fe0.19)1–xCrx, IEEE Trans. Magn., 36 (2000) 1, 381–387
3 P. Wismann, K. Muller, Surface Physics, Springer-Verlag, Berlin
1975
4 A. C. Reilly, W. Park, R. Slater, B. Ouaglal, R. Loloee, W. P. Pratt Jr.,
J. Bass, Perpendicular Giant magnetoresistance of Co91Fe9/Cu
Exchange-Biased Spin-Valves: A Further Test of the Unified Picture,
JMMM, 195 (1999), L269–L274
5 H. A. Faltin, The Magnetic and Electrical Properties of Ni–Fe alloy-
Carbon Thin Film Multilayers, thesis College of William and Mary,
Virginia, USA 2000
M. MA^EK, A. OBLAK: Ni-Fe ALLOY THIN FILMS FOR AMR SENSORS
Materiali in tehnologije / Materials and technology 51 (2017) 3, 499–501 501
MATERIALI IN TEHNOLOGIJE/MATERIALS AND TECHNOLOGY (1967–2017) – 50 LET/50 YEARS
Figure 4: A typical response R/Ravg of 2 test resistors and commer-
cial reference to position over periodic multipole magnetic field with
2 = 4 mm at a distance of 1.5 mm
Slika 4: Tipi~na krivulja odziva R/Ravg 2 testnih uporov in komer-
cialne reference na pomik nad periodi~no spreminjajo~im magnetnim
Figure 3: Magnetization curve for sample deposited with 400 W at
1.1 Pa, directly onto thermally grown SiO2 and annealed at 420 °C for
30 min
Slika 3: Krivulja magnetizacije za vzorec nanesen pri 400 W in tlaku
1.1 Pa, direktno na termi~en SiO2 in popu{~ano 30 min na 420 °C