Določanje sledov kovin v jeklih in superzlitinah, po predhodni ločbi, s plamensko AAS
AAS Determination of Trace Metals in Steeis and Superalloys with Preliminary Separation
T. Lavrič*, A. Osojnik*, Z. Kristan*
Delo opisuje primerjalno študijo ločbe ionsko asoci-iranih kompleksov Ag, Bi, Cd, Pb, Se, Sn, Te in Zn s tri-oktiifosfinoksidom (TOPO) in trioktilaminom (TOA) v me-tilizobutilketon (MIBK). Raziskujemo možnost uporabe obeh skupinskih ločb naštetih kovin za določanje nečistoč v jeklih in superzlitinah na osnovi niklja, s končnim določanjem s plamensko AAS. Separiranje in določanje preverjamo s certificiranimi referenčnimi materiali. Zbrani so tudi podatki analize sledov za nikljeve zlitine iz proizvodnje in podatki o občutljivosti, točnosti, ponovljivosti določanja ter o praktični meji zaznavnosti za vsak element.
UVOD
Sledovi Ag, Bi, Cd, Pb, Se, Sn, Te in Zn ter še nekaterih elementov vplivajo kvarno na mehanske, fizikalne in tehnološke lastnosti iitega železa, nekaterih vrst jekla, predvsem pa so škodljivi v superzlitinah, ki se uporabljajo pri visokih temperaturah. Nezaželjene in škodljive so že primesi reda velikosti nekaj ppm ( = |ig g1) in manj. Pravilna ocena takih količin zahteva precizne metode določanja, ki so vezane izključno na instrumentalno tehniko. Vendar so redke in zelo drage tiste tehnike, ki imajo, brez predkoncentriranja, zadostno občutljivost za direktno analizo sledov kovin in da so, istočasno brez predhodne ločbe sledov cd matrice, proste motenj glavnih komponent. Plamenska AAS, neplamenska AAS in ICP-AES zahtevajo ločbo ter skoncentriranje ppm količin kovin od osnove z namenom, da se odstranijo motnje ali/in zaradi izboljšave parametrov določanja. Primerne tehnike so ekstrakcija, hidridna tehnika, soobarjanje, redukcija do elementarnega stanja. Na izbiro ločbe vplivajo fizikalne in kemične lastnosti določanega elementa in matrice, kar vpliva tudi na to, ali je ločba lahko eno-ali večele-mentna.
Namen predstavljene študije je bil zbrati podatke o možnosti skupinske ločbe in skoncentiranja ppm količin jodidov Ag, Bi, Cd, Pb, Se, Sn, Te in Zn iz večje zatehte jekla oziroma superzlitine na nikljevi osnovi v 5, 10 ali 20 ml organske faze TOPO-MIBK ali TOA-MIBK. Kontrola ločbe je potekala z analizo organskih ekstraktov s pla-
* SŽ — Metalurški inštitut
■• Originalno publicirano: ZZB22(1988)4 ■'' Rokopis prejet, avgust 1988
UDK: 543.064:669.14 ASM/SLA: S11d, S11g, 1—54
The present work describes a comparison study of the separation of Ag, Bi, Cd, Pb, Se, Sn, Te and Zn, as iodide complexes ion-associated vvith trioctyphosphin-oxid (TOPO) and vvith trioctylamine (TOA), into methyli-sobuthylketon (MIBK).
The efficiency of both multielement separations, fol-lovved by flame AAS determination, vvas investigated to determine the traces of analytes in different grade of steeis and superalloys. Standard reference materials vvere used to test separation and determination. The sensitivity, accuracy and repeatibility data, and also the practicai limit of detection for each element are gathered and assessed.
INTRODUCTION
Mechanicaj physical and technological properties of various types of čast iron, some low alloy steeis, and es-pecially superalloys for high temperature appiication strongly depended on trace metal contents such as Ag, Bi, Cd, Pb, Se, Sn, Te, Zn, and others. Because of their harmfull effect already at the ppm (=ugg~1) level and lovver a great attention has been paid to the knovviedge of ppm concentrations of trace elements. Ali that re-quires reliable, correct methods of determination, that are exclusively connected vvith instrumenta/ technics. Except a fevv and expensive technics have satisfactory sensitivity for direct determination of trace meta/s with-out preconcentration and that they are at the same tirne free of interferences of the matrix vvithout separation. The conventional flame /4/1 s, flameless AAS, ICP-AES demand the separation and preconcentration of ppm amounts of metals from the matrix in order to avoid interferences or/and to improve parameters of determination. The convenient separation methods are solvent ex-traction, hydride generation, coprecipitation, reduction to elementaI state. The possibility of appiication of these separation methods depends on physical and chemical properties of the determined trace and of the matrix, vvhat influences also to one-or multielement separation.
The aim of the present study vvas to gather anaiyticai data obtained by the use of preconcentration solvent ex-traction technic of iodide s of Ag, Bi, Cd, Pb, Se, Sn, Te, and Zn from large vveight from various kinds of steeis and superalloys on nickel base by the use of 5, 10, 20 ml TOPO-MIBK and TOA-MIBK organic phase, follovved by
mensko AAS. Uvedli smo TOPO-MIBK ločbo po Burke-ju1 2, ki jo uporablja avtor za določanje Bi, Pb, Sb in Sn v aluminiju, železu in nikljevih zlitinah. Janoušek3 in Staats4 uporabljata podobno ločbo za določanje mikroelementov v jeklih. Janoušek določa le Bi, Pb, Sb in Sn v nelegi-ranem jeklu, medtem ko je Staats prilagodil ločbo določanju Cu, Zn, Mo, Ag, Sn, Sb, Pb in Bi za nelegirana in malolegirana jekla. Ločba TOA-MIBK je manj v rabi na področju analize jekla5 6, bolj za posamezne -elemente-sledove v bakru, svincu in selenu7 8 9. Dodatno smo izvedli primerjalno študijo določanja Te z redukcijo do elementarnega Te. Opisane so separacijske metode in predstavljeni rezultati za certificirane referenčne materiale (CRM) ter nekaj nikljevih zlitin iz proizvodnje. V možnih primerih smo izvedli za CRM primerjavo z dosegljivimi podatki drugih avtorjev.
EKSPERIMENTALNI DEL
Aparatura
Meritve smo izvajali z AA spektrometrom Perkin El-mer model 2380, ki je bil opremljen z zunanjim izvorom za brezelektrodne žarnice. Za določanje Bi, Se, Sn in Te smo uporabljali kot primarni izvor svetlobe brezelektrodne žarnice, za določanje Ag, Cd, Pb in Zn pa votle kato-de. Signale smo beležili na printerju Perkin Elmer PRS-10. Za vsak element smo optimirali lego gorilca in parametre plamena na maksimalno absorbanco z razprševa-njem standarda v ustreznem ekstraktu. Za vse elemente smo uporabljali plamen zrak-acetilen in »impact bead«, le za Sn plamen didušikov oksid-acetilen in »vetrnico«. Čas integracije je bil vedno 1 sec pri hitrosti razprševa-nja organske faze ~ 5 ml min1. Meritve smo izvajali za vsak element na njegovi najbolj občutljivi-resonančni liniji. Meritve smo izvajali za vse elemente izključno samo v linearnem območju, kjer je zadoščalo umerjanje z enim standardom, čeprav je instrument sposoben sprejeti tri standarde za umerjanje in korekturo krivulje.
Reagenti in materiali
Vse kemikalije so bile nabavljene pri fa. Fluka. MIBK in kisline so bile p.a., TOA purum kvalitete. Uporabljali smo 5 % raztopino TOPO v MIBK in enako 5 % TOA v MIBK. Za pripravo vzorcev smo uporabljali sigradur ča-še. Uporabljali smo CRM naslednjih dobaviteljev: BAS, NBS in MBH. Večina jih je bila kupljena v obliki finih ostružkov, nekateri v obliki diskov (MBH), iz katerih smo pripravili ostružke z ene strani diska. Kemična sestava uporabljenih CRM je razvidna iz tabele 1.
Postopki za jekla in nikljeve zlitine
Opisana sta dva postopka ekstrakcije Ag, Bi, Cd, Pb, Se, Sn, Te in Zn ter en postopek redukcije Te. Vsem trem so skupne tri faze: raztapljanje s pripravo vzorca, ločba in zadovoljivo določanje. Vsaj en slepi vzorec pridružimo vsakemu postopku.
— TOPO-MIBK ekstrakcija: Osnova postopka je ek-strakcija jodidov enega ali več elementov, kot jo navaja Burke1, z umerjanjem z raztopinami, ki jih vodimo skozi celoten postopek2. Vzorec jekla ali nikljeve zlitine (1—5 g, odvisno od vsebnosti sledov) počasi in previdno raztopimo v 3 + 1 mešanici HCI + HN03. Raztopino iz-parimo na ~ 5 ml, dodamo 10 ml HCI, razkrojimo HN03 z mravljinčno kislino. Raztopino izparimo do goste sirupa-ste konsistence, da izženemo vse dušikove okside in da se razkroji mravljinčna kislina. Dodamo HCI, askorbinsko
flame AAS. We followed the TOPO-MIBK separation of Burke1 2, applied by author for Bi, Pb, Sb, and Sn determination in aluminium, iron and nicke base alloys. Janoušek3 and Staats4 performed similar separation for determination of trace elements in steels. Janoušek3 determined Bi, Pb, Sb, and Sn in unalloyed steel but Staats4 adapted separation for the determination of Cu, Zn, Mo, Ag, Sn, Sb, Pb, and Bi for unalloyed and low alloyed steels. The TOA-MIBK extraction is a little applied separation for steels5 6 but more for the determination of single elements in copper, lead, and seienium7 s 9. For Te also a comparison vvith reduction to elementaI Te was made. Methods of separation are shortly described and the results presented for the analytes of certified reference materials (CRMs) and elected nicke I base alloys from production. Some results for CRMs are compared vvith available literature data.
EXPERIMENTAL Instrumentation
Perkin Elmer atomic absorption spectrometer Model 2380, equipped vvith an externai electrodeless discharge povver supply vvas used. As primary light source electrodeless discharge lamps vvere employed for the determination of Bi, Se, Sn, Te, and for the determination of Ag, Cd, Pb, and Zn hollovv cathode lamps vvere used. The signals vvere printed on a Perkin Elmer PRS-10 printer. Other conditions vvere as follovvs: the burnerpo-sition and flame parameters for each element vvere opti-mized for maximum absorbance by aspirating standard organic solution. The impact bead and air-acetylen flame vvere used for ali test analytes, except for Sn nitrous ox-ide-acetylen and spoiler vvere used. Integration times of 1 sec vvere used in ali cases in conjunction vvith an aspi-ration time — 5 ml min'1 of organic solution. The most sensitive-resonance line was chosen for each element. The measurements vvere made for ali analytes exclusive-ly in the linear concentration range vvhere the calibration vvith one standard vvas sufficient, othervvise the instrument is capable of accepting up to three standards for calibration and curve correction.
Reagents and materials
Ali analytical vvork reagents vvere supplied from Fluka, MIBK and acids p. a. grade, TOPO puriss, and TOA purum grade. TOPO-5 % solution in MIBK, TOA-5 % solution in MIBK vvas applied. Sigradur vessels vvere used for sample preparation. Certified Referenc Materials (CRMs) supplied by the BAS, NBS and MBH vvere used. The most of them vvere supplied in fineiy divided form, but some of them vvere in disc form .(MBH). From discs fine drillings vvere cut on one side Table 1.
Procedures for steels and nickel base alloys
Two solvent extraction methods for determination of Ag, Bi, Cd, Pb, Se, Sn, Te, and Zn are described, and one procedure for redution of Te.
Ali three further described procedures have three steps in common: dissolution vvith pretreatment of the sample, separation, and satisfactory determination. A blank vvas included in each of the follovving digestion procedures:
— TOPO-MIBK ekstraction: Procedure based on the extraction of iodides of one or more elements re-
N
® kc
LJ "D h c - to
® r-" ^
S>
□a "cT
-O
D) M
N O)
Ž "J Rs
tO
"O > "O
O C .115
■S ®
ca v) E 3 .C 42 č S
"S Ž
<u ta
1	5 ® tb >- U
f| "O
M J) X> v C t}
ta to "O
fl a) tO
■O O
2	C
o o a.-c
3	9-.<2 b o. <o
o*
■O XI tO 10
C
.o
■G ta
to
j*
ta c
n i*, ta -s O ><0^
J)

O CD
V ~
c\j o
° s
v
O O Tt-
	LO
	o
	V
	CO
	p
	m
	o
	V
	cšT
	o
	V
	1 LO
	o
	o
	V
	C\J
	o
	V
	to
	O
	CO
o	1 oo
o	CO
	p
	
	+1
	CD
	
	LO
	O
	V
CD	1 C\J
	o
	V
	LO
	O
	V
	O,
	O
	CO
	o 1
	1 CM
cp	
00	O
	LO
	o
	o
	+1
	C\J
	o
	CD
	-T—
	o
	v
q q o z z z
o o
v - £i
,Q
a o.
+1
Csl
V
m cu
oi V
t. V

m
V
w n o
V
T-						
LO						
00	m	T—	tN	CO	LO	CD
	CD	CD	CD	CD	s	5
O	CO	CO	CO	CO	CO	CO
cr	(/)	CO	CO	CO	CO	CO
Z)	m	m	m	m	O	O
UJ	2	z	z	z	m	m
c\j
o o
l tO	T	1-	lf>
o		o	*—
o	cx>	V	V
V			
CM	+1		
O	O		
V			
co
I
CVI CM
q
O
LL
_ O	_
00 CO	00
O) O)	O)
<
o
S 5
I
g
Q) i
5) -Cl
I
C
>o o o
■O
I
tu £
.C £
.S
40 q>
a »i i i a o1
"S o
C
3 ^
O o 5
o) ta
o ct
c ®
- o
en ta E
c: n co i- CO
uS 00 (£) o n
"	S 0)
®	S
S	I
<D	^
>N	N
°	S
O
tO	^
l	^
S	g
I	3
S	g
CO
"ta	'O
B	§
2	t 0)
£	"
i	i
S	£
N	to
O	tO
O	0)
•-	J
O M W ^ I 3 m cd O m uizz®5
_ r o ® 00 o
Q>
.C g
a
O o
C
tu
>o
o §
Q 2:
kislino in jodidni reagent ter ekstrahiramo v 5, 10, 20 ali 30 ml 5 % TOPO v MIBK. Čas ekstrakcije je 30 sek. Organsko fazo razpršimo v plamen. Za določitev enega samega elementa zadostuje 5 ml ekstrakta. Izkušen analitik lahko določi vseh osem elementov iz 30 ml ekstrakta.
Podrobni eksperimentalni podatki kot tudi diskusija k TOPO-MIBK ekstrakciji so na voljo v literaturi12. Pri naših poskusih smo povečali količino zmesne kisline za raztapljanje 5 g vzorca iz 60 na 100 ml.
2— TOA-MIB ekstrakcija: Priprava vzorca in ločba sta povsem enaki kot za TOPO-MIBK ekstrakcijo, le da smo uporabljali kot ekstraktant 5 % TOA v MIBK, čas ekstrakcije je daljši 5 min.
— Redukcija Te do elementarnega stanja s hidra-zinhidrokloridom: Priprava vzorca je enaka kot opisujemo spredaj do vključno z dodatkom mravljinčne kisline. Nato dodamo ~ 25 ml HCI ter razredčimo na ~ 200 ml. Kislost raztopine naj bo 2—3 M. Dodamo dvojno količino hidrazinhidroklorida, glede na zatehto vzorca. Postavimo na vodno kopel pri 100 °C za 30 min, da se pojavi črna oborina izpadlega Te, nato pustimo čašo z oborino nadaljnih 30 min poleg vroče plošče. Koagulirano oborino zberemo na milipor filtru 0.2 (im, speremo z vodo in raztopimo v čimmanj hladne HN03, ohladimo, dodamo 5 ml HCI, razredčimo na 25 ml, razpršimo v plamen.
REZULTATI IN DISKUSIJA
Umerjanje in meritve
Za umerjanje aparature smo sprva v glavnem uporabljali CRM, ki smo jih pripravljali na enak način kot vzorce. Zaradi pomanjkanja CRM s ppm vsebnostjo iskanih elementov smo včasih naredili referenčne vzorce, že pred raztaplanjem, z združevanjem elektrolitskega železa (NBS 365) ali zelo čiste nikljeve zlitine (BCS 354) z različnimi CRM v različnih utežnih razmerjih.
Nadalje smo pripravljali raztopine za umerjanje iz standardnih raztopin, ki smo jih dodajali elektrolitskemu železu (NBS 365) ali superzlitini (BCS 345). Standardne raztopine smo vedno dodali šele po raztopitvi matrice. Tako pripravljene sintetske vzorce smo vodili skozi celoten postopek. Vse standardne raztopine smo pripravili iz čistih kovin ali p.a. soli (Ag). Z obema načinoma umerjanja dobimo skoraj identične rezultate za vse elemente, razen za Bi, kjer smo ugotovili, da dobimo pravilne rezultate le z umerjanjem s CRM; z umerjanjem s sintetskimi vzorci dobimo prenizke rezultate.
Pravilno umerjanje je zelo važen dejavnik celotnega postopka. Kritična točka je v pravilni izbiri ničle instrumenta. Kot ničlo instrumenta vstavimo vedno elektrolitsko železo (NBS 365) ali nikljevo zlitino (BCS 345), ki jo v nadaljnjem imenujemo delovna slepa raztopina in jo pripravimo enako kot vzorce. Dobljeni rezultati so namreč pokazali, da dobimo napačne rezultate, če vstavimo kot ničlo čisti MIBK, izmerimo koncentracijo delovne slepe raztopine in to vrednost naknadno odštevamo od celotne vrednosti za umerjalni standard ali preiskovani vzorec. Delovno slepo raztopino pripravimo za vsako skupino vzorcev in jo obvezno uporabljamo za izpiranje med posameznimi vzorci, posebno kadar merimo v bližini praktične meje zaznavnosti in pri ogrevanju gorilca.
Uporabnost obeh ekstrakcijskih sistemov
Meritve Ag, Bi, Cd, Pb, Se, Sn, in Te v TOPO-MIBK in TOA-MIBK ekstraktih so pokazale podobno učinkovitost ločbe v obeh sistemih, medtem ko se Zn ekstrahira dobro le v TOPO-MIBK, v TOA-MIBK največ 50 %, pri izbranih delovnih pogojih. Izkoristki ložb so za vse ostale ele-
ported by Burke1, vvith calibration solutions that have been taken through extraction procedure2, vvas applied. The sample of steel or nickel base alloy (1—5 g depend-ing on amount of traces) siowly and carefully vvas dis-solved to avoid losses by sputtering in 3+ 1 mixture of HCI+ HN03. The resulting solution vvas then evaporated to ~ 5 mi, then 10 ml HCI vvas added, and nitric acid de-co mpose d by formic acid. The solution was then evaporated nearly to dryness to sthick consistence drive off to ali nitrogen oxide fumes and finaly destroying of formic acid. After the add it i on of HCI, ascorbic acid, and iodide reagent, the extraction into 5, 10, 20 or 30 ml of 5% TOPO-MIBK follovved, extraction tirne 30 sec. The organic extract vvas aspirated into flame. For determination of one element only 5 ml of extract satisfy. An experienced analyst is able to determine aH eight analytes from 30 ml of extract. Data for experimental conditions as well as discussion on the TOPO-MIBK separation may be found in the literature1 2. In our research acid mixture was increased for 5 g sample from 60 to 100 ml.
—	TOA-MIBK extraction: The preparation of the sample and separation procedure are the same as under TOPO-MIBK extraction, except vve added 5 % TOA in MIBK instead of TOPO the extraction tirne vvas 5 min.
—	Reduction of Te to elementa! state by hydrazine hydrochloride: The dissolution procedure as described above, inclusive of formic acid addition. After that hydrochloric acid vvas added to adjust the acidity to 2—3 M and the volume to ~ 200 ml (~ 25 ml concen-trated hydrochloric acid/200 mi). Double amounts of hy-drazine hydrochloride vvas added vvith regard to the sample vveight. The mixture vvas incubated in a water bath at 100 °C for 30 min till black precipitate of Te ap-peared. The mixture vvas allovved to stand near the hot plate further 30 min. Coaguiated precipitate vvas collect-ed on a millipor filter 0.2 pm, vvashed vvith vvater and dis-solved in a few ml of cold concentrated nitric acid, cooled, 5 ml of hydrochloric acid vvas added, diluted to 25 ml and applied to AAS.
RESULTS AND DISCUSSION
Calibration and measurements
In the first stage calibration solutions from CRMs vvere used for calibration of the instrument, vvhich vvere prepared on the same way as the samples. To overcome the lack of standards vvith ppm amounts of analytes, reference materials vvere sometimes prepared vvith knovvn amounts of traces by spiking electrolytic iron (NBS 365) or very ciean nickel base alloy (BCS 345) vvith different CRMs in different proportions before the dissolution step.
Later also calibration solutions from varying amounts of a stock solution of respective element and electrolytic iron (NBS 365) or superalloy (BCS 345) vvere prepared. The stock solutions vvere added always after dissolution of the matrix. AH standard solutions vvere prepared from ciean metals or p.a. salts (Ag).
With both calibration made vve achieved nearly ident-ical results for ali elements. Except for Bi vve stated, that correct results vvere obtained vvith calibration vvith CRMs. Calibration vvith synthetic standards yields to low results.
The calibration is an important factor in ali proce-dures. A critical point in both liquid-extraction proce-dures vvas imposed by correct choice of the zero of the instrument. The vvorking blanc, eiectrolytic iron or very ciean nickel base alloy taken through a/l procedure,
mente v obeh sistemih med 85—110 %. Če upoštevamo nizke koncentracije, lahko ugotovimo, da so izkoristki dobri.
Oba ekstrakcijska sistema sta uporabna za analizo sledov v jeklih in nikljevih zlitinah.
Prednost ločbe s TOPO-MIBK so: je izrazita večele-mentna ločba sledov, ki omogoča tudi določanje Zn po opisanem postopku. Odlikuje jo kratek čas ekstrakcije (30 sek) v primerjavi s 5 min ekstrakcije s TOA-MIBK.
Prednost ločbe s TOA-MIBK so: je tudi večelement-na ločba, ki bi bila morda ob drugačnih pogojih preiskovane raztopine uspešna tudi za Zn. Nihanje ozadja je manj izrazito kot pri TOPO-MIBK, kar pride posebno do izraza pri Se, ki je težko določljiv s plamensko AAS, za katerega dobimo s TOA-MIBK stabilnejše odčit-ke za slepi vzorec in vzorec, s tem v zvezi boljši standardni odmik, kar je posebno važno, ko določamo koncentracije v bližini praktične meje zaznavnosti. Standardni odmik desetih meritev istega vzorca za 0,2 jig ml"1 Se je bil v TOPO-MIBK 15-25 %, v TOA-MIBK 10-20 %.
Meritve občutljivosti za TOPO-MIBK in TOA-MIBK ekstrakte dajejo približno enake vrednosti za vse elemente, z izjemo Sn, kjer je razlika med obema sistemoma občutnejša, tabela 2. Na splošno je občutljivost do-
Tabela 2: Vpliv TOPO- in TOA-MIBK ter H20 + HCL sistemov na občutljivost določanja.
Table 2: Effect of TOPO- or TOA-MIBK, and H20 + HCL sy-stems on the sensitivity of determination
Absorbanca I Absorbance
Element ugml-1 -
TOPO-MIBK TOA-MIBK H20 + HCL
Ag	0.5	0.166	0.178	0.062
Bi	3.0	0.178	0.168	0.058
Cd	0.5	0.238	0.238	0.120
Pb	3.0	0.202	0.204	0.094
Se	3.0	0.044	0.036	0.014
Sn	10.0	0.040	0.027	0.012
Te	3.0	0.126	0.114	0.046
Zn	0.2	0.258	/	0.072
Tabela 3: Praktične meje zaznavnosti in ugotovljena linearna območja umerjanja v organskem sistemu.
Table 3: Praktical limlts of detection and linear ranges of cali-bration in organic system.
Element	Praktična meja zaznavnosti Practical detection Umit		Območje linearnosti Range of linear response (ug mr1)b
	(|xg 9~1)a'6	(%)	
Ag	0.2	2 x 10-5	0.05-3.0
Bi	0.4	4 x 10~5	0.2-6.0
Cd	0.1	1 x10"5	0.1-0.5
Pb	0.4	4 x 10-6	0.2-4.5
Se	1.0	1 x10"4	0.5—3.0
Sn	1.0	1 x 10-4	1 -50
Te	0.4	4 x 10~5	0.2-6.0
Zn	0.05	5 x 10~6	0.2—0.6
a — Pri teži vzorca 5g, ekstrakt 5 ml Using 5g sample, extract 5 ml
13 — Vse vrednosti so približno enake za TOPO- in TOA-MIBK, z izjemo za Zn
AH values are approximafely equal for TOPO- and TOA-MIBK, exceptly for Zn
ločanja dobra, najslabša je za Sn. V primerjavi s klorovo-must be sef as zero of the instrument used in this study. The results obtained confirmed, that it is vvrong to mea-sure the concentration of the vvorking blank against ciean MIBK separately, and subsequent its substraction from the total signal for calibration sample and tested sample.
A vvorking blank should be prepared vvith each set of samples, and may be used obligatory for vvashing out betvveen each sample, particularly by measuring near the practical limit of detection, and for burner vvarm up.
Application of extraction systems
Measurements of Ag, Bi, Cd, Pb, Se, Sn, and Te in TOPO-MIBK and TOA-MIBK extracts yield approximately the same degree of separation; for Zn only the extrac-tion into TOPO-MIBK is convenient, the extraction in TOA-MIBK is less than 50 %, under chosen vvorking con-ditions. 85—110 % recoveries for aH analytes vvere af-forded, Considering the low concentrations the recoveries can be regarded as acceptable. The both extraction systems are available for the analysis of t race elements, mentioned above, in steels and nickel base alloys. Ad-vantages of TOPO-MIBK separation are: it is an expres-sive multielement separation technic, that hat is anabie to separate and preconcentrate also Zn, under described procedure. The separation is distinguished by less tirne consuming for shaking, 30 sec against 5 min for TOA-MIBK. Advantages of TOA-MIBK separation are: is also a multielement separation technic, vvhich could be useful also for Zn in other conditions of the vvorking soiution. The fluctuation of the signals of the back-ground is smalier than of TOPO-MIBK solutions vvhich contributes to slightly more stable signals. This /s espe-cially evident for Se, vvhich is difficult to determine in TOPO-MIBK. This contributes to stable signals for vvorking blank and sample soiution and lovver relative standard deviation, vvhich is especialiy noticeable when concentrations are near the practical detection limit. Relative standard deviation for ten measurements of one sample on 0.2 ug mi~1 level of Se vvas approximately 15—25 % in the TOPO-MIBK and 10-20 % in the TOA-MIBK soiution.
Measurements of sensitivity of determination for aH analytes in both solvent extraction procedures vvere nearly the same, vvith exception for Sn, vvhere the differ-ence is grater. Generaily the sensitivity of determination is good except for Sn. The effects of enhanced sensitivi-ty of the analytes in organic phases compared vvith HsO+ HCI solutions are shovvn in Table 2. Proportional improvements in sensitiviry reaiise better practical detection limits. Practical detection limits, the lovvest concentration vvhich can be determined under routine conditions, based on 5 g sample and 5 ml organic extract are shovvn in Table 3.
AH measurements vvere made for aH analytes exciu-sively in linear concentration range shovvn in Table 3.
RepeatibiHty test chiefly indicated a relative standard deviation of 2% at about 100 /ig g~ 1 level, 10 % at about 6 ug g~' level and 20—30 % at about 2 ug g~1 level vvith-iri different batehes on different days. Recomended procedures gave good reproducibility for the same organic extract from day to day for aH analytes.
Application of reduction procedure for tellurium
A s can be seen from Table 4, the Te is determined enaugh preciseiy by the reduction method to elementa/
<
CM OD
cn
X
m
i ? oiO
3 o
i ž O "J
c i
N Qi
S?
<D 01
«iS
jo
x m
5 <
O
o co o
X
m
co OT O m
co m
OT O m
co OT
m
fig
O1-*.-
s s" >0 ^
O Ol o^
^ O « £
-f-*
"5 3
n <0 <B Oi
DC OT
Tt rr W 41 « 01 JQ XI
o a
co (0 CD
OT
m
CD
o
DC LU
CM			
b			
co			
b			
+1			
to	CM		CO
b	b	b	b
	V	V	V
O O
5 +i
co m o b
CM CM
cm cvj b b
cm
v
r^ -t
^ cm +1 +1 cm r--
S' co co
o o +1 +1 co co
		CM
		+1
00	co	00
c\i	b	CT)
o o +1 +1. cm o 't "t **
O) <
a> cm b *-' +1 +1 cn
cri b
cm cm b b
V	V
cm O
V
CO
b o co +1 +1 . -"t T" ^t
iri r^ CD
<9 ^ o b +1 +1 •t CM CO •>* ^ tri
oo cn o o b b
		CM	CM
		V	v
a		CM	
z		V	
tO	m		
o	o	in	co
b	b	T"	b
+1	+1	"H	+1
00			CD
co	co	b	O
b	b	CM	CM
■>t			
b		CM	
		to	tO
b	b	b	b
V V
o o o
V V V
o o o
V V V
■O
O
+1 +1 CO CM
V v
CVIUIS 00 00 co
co b +1 . T- CO
tO rt
f-	T- T-	co	-fr CM
o	o b	b	b b
V	V V	V	V V
d		to	
		CM	
Z		b	
n o.
+1 N.
CM O +1
cn b
+1 to
Tf CO
b o
+1 +1 CM
CM CO '
r^. co b b +| -H-o o co
+1 +1 CT) tO
<1) OT
O O co cm +1 +1 co cm T~~~ O t
i- cn
-■t cm +1 +1 cm o ct)
CO CT)
co m b b +1 +1 CO T-tri tb
to
SE co b jj CT)
+1 ^ ft," (O CT) CT)
O CT> CO
CT) cn
h- cm O ■>-+1 +1 CO t^ O
co co
V V
m m b b
V	V
cm O
V
-H
tO CM < CT) CT) i
o
+1
CM h-
ob1
cm cm in ob*— -H +1 +1 O) tO co tri tri to
c OT
	o JC
	—
	
	V
	tn
cm	Tt
T-'	+1
+1	tn
co	cri
	cm
	cn
	cm
	
	b
	V
	m
	b
	V
co	
	to
+1	b +L
b _	cm tn
T"	rt
	
	J
LO	tO
b	T-'
+1	+1«
tr>	ct) co
oi a>	■tf iri
a»
. to
tO
co
-H.
00 co
N
C0
t
|
T3
"O «1
t tu o
to <0 3
0	—
c d. ^ Co9
S"? Al?
"cA tO £ -p C £L S m ti) S O
1	-S s i i S
s
5 5 §
2	"3 < o Al o)
1	sls
• "F — ° ^ c c a.
m = "P -
-f g -O =J
O n ^ «
to Ž
(D i. O
> B
C/) N O ti) C ■D 0)
_	tO
>	z
I	I
<	co
Q_ 01
o>< __
1 :1 "+I-D
Tabela 5: Rezultati analize elementov-sledov v raznih vzorcih iz proizvodnje. Table 5: Analytical results of trace eiements for various samples from production.
Vzorec			Vsebnost/Content(\ig g-		1 = ppm)			
Sampie	Ag	Bi	Cd	Pb	Se	Sn	Te	Zn
Aluminij Aluminium	<3	<1	0.24	160	1.2	17.4	0.6	39
Nikelj Nickel	0.8	<1	0.05	12.4	2.7	15.6	0.8	5
Kobalt Cobalt	1.5	<1	11.1	5.9	2.9	4.3	0.7	12
Krom Chromium	1.7	<1	0.04	20.5	2.1	2.0	0.9	2.4
Titan Titanium	6.9	<1	0.15	<1	18.7	220	5.6	3
Molibden Molybdenum	1.9	<1	0.13	5.1	4.5	20.0	2.4	0.08
NIMONIC 263/1a	0.24	0.2	0.12	2.9	2.3	14.0	1.3	2.4
NIMONIC 263/2®	0.20	<1	0.10	5.5	2.0	11.4	0.46	15.5
NIMONIC 263/3a	0.28	<1	0.10	7.2	1.8	11.2	0.70	32
INCONEL"	0.89	5.5±0.5°(7)	0.16	33.7 ±3.5(7)	3.9	96.8	2.1	13.6
a— Ni-zlitina: 50%Ni, 20%Co, 20%Cr, 5.9%Mo, 2.5%Ti, 0.36%AI, 0.65%Fe Ni-alloy
b - Ni-Fe zlitina: 55%Ni, 18%Cr, 0.3%Co, 16.6%Fe, 2.8%Mo, 5%Nb, 0.6%AI, 1.0%Ti Ni-Fe alloy
c— Srednja vrednost ± standardni odmik od sr. vr. (n = 7) Mean ± standard deviation of the mean (n= 7)
dikovo kislimi raztopinami se v organskih sistemih poveča občutljivost od 2.5 do 3-krat.
Praktične meje zaznavnosti, t.j. najnižja koncentracija, ki smo jo še lahko določili z rutinskim delom, pri uporabi 5 g zatehte vzorca in 10 ml organskega ekstrakta, so razvidne iz tabele 3.
Vse meritve smo izvajali za vse elemente samo v linearnih območjih, ki jih navajamo v tabeli 3.
Test ponovljivosti daje relativni standardni odmik 2 %. na nivoju 100 p.g g"1, 10 % na nivoju 6 p.g g 1 in 20— 30 % na nivoju 2 pg g \ in to glede na vzorce, ki so bili pripravljeni ob različnih dnevih. Postopka, ki ju priporočamo, dajeta za vse elemente dobro ponovljivost istega organskega ekstrata v več zaporednih dnevih.
Uporabnost postopka redukcije telurja
Kot je razvidno iz tabele 4, se da Te določati dovolj točno z metodo redukcije do elementarnega stanja, ki ji sledi določanje z AAS iz H20 + HCI raztopin.
Na splošno dosegamo dobra ujemanja s CRM v širokem koncentracijskem območju. Občutljivost in praktična meja zaznavnosti sta slabši v primerjavi z organskimi raztopinami. Meja zaznavnosti je 6 p.g g-1. Postopek je uporaben tako za analizo jekla kot zlitine na osnovi niklja.
Analiza certificiranih referenčnih materialov
Z namenom, da bi preverili pravilnost ločbe in skon-centiriranja, smo analizirali večje število CRM. Vrednosti, dobljene za vrsto standardov jekel in nikljevih zlitin, se zadovoljivo ujemajo s certificiranimi vrednostmi, kar potrjuje, da se da doseči z opisanima postopkoma ločbe
state, followed by AAS determination of HsO+ HCI soiu-tions. In general, there is a good agreement vvith CRMs ovver a vvide concentration range, but the sensitivity and the practical detection Umit are poorer, than for the or-ganic solutions. The Umit is 6 ug g~ '. The procedure is available for the anaiysis of steels an nickel base alloys.
Anaiysis of certified reference materials
To assess the accuracy of the separation and prec-oncentration methods many CRMs vvere appiied. The values obtained for a range of standard steels and nickel base alloys shovv close agreement vvith certified values for appiied procedures, confirm a agood enrichment factor and also prove that accurate results can be obtained at ppm level. The results are given in Table 4 together vvith those obtained by other investigators from literature. Recoveries are usually higher than 85 %. The determined values are for some CRMs higher and for some iovver when compared vvith the certificate. It is interesting to note that also the data in certificates are in a spread, and that some values are not certified but only given for information.
Analysis of samples from production
The established methods are appiied on the sampie from production, see Table 5.
There vvas problem vvith dissolution of nickel base al-loy vvith 5 %. Nb. In this čase the addition of H^Oj vvas not sufficient1, but the addition of 1 ml HF prior or after the dissolution of the sampie vvas successful.
dobro skoncentiranje sledov in pravilne rezultate na ppm nivoju. Rezultate navajamo v tabeli 4, skupaj s podatki drugih avtorjev, ki smo jih zasledili v literaturi. Izkoristki so običajno večji od 85 %. Nekatere naše vrednosti so včasih višje, drugič nižje v primerjavis certifikati. Opozoriti je treba, da se tudi v certifikatih gibljejo analizne vrednosti v določenih razponih, nekatere vrednosti celo niso certificirane, ampak dane le v vednost.
Analiza vzorcev iz proizvodnje
Vpeljane metode uporabljamo za analizo vzorcev iz proizvodnje, tabela 5.
Problem je bil z raztapljanjem nikljeve zlitine, ki vsebuje 5 % Nb. Dodatek H2S04 ni zadoščal1, popolno razto-pitev vzorca je omogočil dodatek 1 ml HF.
CONCLUSION
In this study it has been shovvn that the solvent ex-traction of Ag, Bi, Cd, Pb, Se, Sn, and Te-iodides can be used as vvith TOPO-MIBK as vvith TOA-MIBK procedures of separation, oniy Zn can be separated oniy vvith TOPO-MIBK. The proven technics are available for the analysis of steels and superalloys for trace elements of metallurgical interest. The methods offer satisfactory accuracy, good reproducibility, and excellent sensitivity for each element determined.
ZAKLJUČEK
Preštudirali smo ločbi jodidov Ag, Bi, Cd, Pb, Se, Sn in Te ter ugotovili, da sta obe ločbi uporabni za vse naštete elemente, medtem ko je za Zn uporabna le TOPO-MIBK ločba. Preverjeni tehniki sta uporabni za analizo tistih elementov sledov v jeklih in superzlitinah, ki so zanimivi za metalurgijo. Odlikuje ju dobra točnost, ponovljivost in odlična občutljivost.
LITERATUR A/fl£FEfl£AfC£S
1.	K. E. Burke, Analyst (London) 97, 1972, 19-28
2.	K. Thornton, K. E. Burke, Analyst (London) 99, 1974, 469-470
3.	I. Janovšek, Hutn. Listy 1982, 10, 737-738
4.	G. Staats, Fresenius Z. Anal. Chem. — je v tisku/is in print
5.	S. D. Shete, V. M. Shide, Michrochim. Acta I. 1984, 63—67
6.	B. J. Spivakovet. al.,Zh. Anal. Khim. 31, 1976,757—763
7.	K. Kasiura. Chem. Anal. (Warszawa) 31, 1986, 143—150, Abstracts Fresenius Z. Anal. Chem. 326, 1987,1, 69
8.	E. VVunderlich, W. Hadeler, Fresenius Z. Anal. Chem. 281, 1976, 300
9.	lwao Tsukahara, Toshimi Yamamoto, Talanta 28, 1981, 585-589
10.	B. Welz, M. Welcher, Spectrochim. Acta 36B (1981) 439-462.