V O L U M E 23, NO. 3, M A R C H 1 9 5 1 The range of nickel concentration chosen will depend upon the cell thickness to be used. The order of addition of the reagents is not critical, provided that the niosime is added last. DISCUSSION
The nickelous niosime method appears to have several advantages over the conventional dimcthylglyosime colorimetric procedure. The gum arabic suspensions are much more stable, although the maximum intensity ia not reached immediately. The drastic conditions of Rollet’s method cause most cations to precipitate, but the milder conditions of the new method permit the direct determination of nickel i n the presence of averal of these ions. Although comparison of curves 1 and 2 of Figure 5 viould indicate that Rollet’s method is much niore sensitive, the absorption of ferric citrate is usually so grrat that the smaller maximum at 530 nip is used instead of the intense irinximuni a t 443 mp. The absorbancy index of nickelous riiosinie a t 550 mp is of thr, s a ~ n e order as that of Rollet’s complex a t 530 mp. On this basis, the sensitivities of the two methods are comparable. The nickelous nioxiine method has the great advantage of simplicit,y. It, requires fcw prcliminary separations and eliminates the extractions required in some procedures. The accuracy and reproducibility are suffirient for the purpopes of trace analysis. LITERATURE CITED
(1) Abbey,
s., ANAL.CHBM.,20, 630
(1948).
Alexander, 0. R., Godar, E. M.,and Linde, S . J., ISIJ. KXG. CHEM.,ANAL.ED.,18, 206 (1946).
(2)
453 American Society for Testing Materials, Philadelphia, ‘’A.S. T.M.Methods of Chemical .4nalysis of Metals,” p. 161, 1943. Ayres, G . H., and Smith, F., IND. EX. CHEM.,ANAL.ED., 11, 365 (1939).
Brownlee, K. A., “Industrial Experimentation,” p. 59, Brooklyn, N. Y., Chemical Publishing Co., 1947. Diehl, H., “Applications of the Dioximes to Analytical C%emistry,” Columbus, Ohio, G . Frederick Smith Chemical Co., 1940.
Fairhall, L. T., J . I d . Hug., 8, 528 (1926). Fischer, J., and Cayard, M., 2. anal. Chem., 122, 251 (1941). Furman, N. H., and McDuffie, R., Atomic Energy Commission, Classified Report, AECM-4234 (1947). Jiirvinen, K. K., 2. Nahr. Genussm., 45, 183 (1923). Jaffe, E., Industria chimica, 9, 151 (1934). Johnson, W. C., and Simmons, M., Analyst, 71, 554 (1946). Lindt, V., 2. anal. Chem.. 53, 165 (1914). Makepeace, G. R., and Craft, C. H., IND.E m . CHEX.,. ~ N A I . . ED., 16, 375 (1944). Mandel, J., J. Chem. Education, 26, 534 (1949). hfitchell, .4. M., “Spectrophotometric Study of ColorinietJ.ic Methods for the Determination of Nickel.“ 11,s. thesis. Purdue University, 1945. Mitchell, A. M.,and Mellon, M. G.,ISD. ERG.CHEar., A N \ L ED., 17, 380 (1945). . Lloeller, T., Ihid., 15, 346 (1943). Murray, W. M.,Jr., and -4shley, 9. E. Q., Ibid., 10, 1 (1938). Perry, M. H., and Serfass, E. J., ANAL.CHEM.,22, 565 (1950). Ringbom, A , 2. anal. Chem., 115, 332 (1938). h l l e t , A. P., Compt. T a d . , 183, 212 (1926). Snedecor, G. W., “Statistical Methods,” 3rd ed., pp. 5 5 . 184, Ames, Iowa, Iowa State College Press, 1940. Voter, R. C., and Banks, C. V.,. ~ N . \ L .CHCM.,21, 1320 (1949). RECEIVED July 31, 1950. Contribution No. 114 from the Institute for Atomic Research and Department of Chemistry. Iowa State College, Ames, Iowa. W o r k performed in the Ames Laboratory of the Atomic Energy Commission.
Spectrophotometric Determination of Cerium(W) A. I. MEDhLI-4 A N D B. J . BYRNE Brookhasen National Laboratory, Upton, Long Island, IV. Y It was desired to study the behavior of solutinns of cerium at as low concentrations as could be conveniently determined. It appeared likely that the sensitivity of an existing colorimetric method, based on the yellow color of ceriurii(IV),could be improved by working in the ultraviolet. Cerium(1V) was found to show an absorption maximum at approximately 320 mp, with a molar extinction coefficient of -5.58 X IO8, in 1 N sulfuric acid. t photometric sensiti\it>
T
HE fanii1i:ir yellow color ot tile ceric iou in acid mrdiuni has
been made the basis for i~ colorirnetric procedure, in which cerous ion is oxidized with persulfate ill boiling 1 N sulfuric acid with silver ion as a catalyst (8). If the intensity of the yellow color is determined visually, the sensitivity is approximately 10 micrograms of cerium per square centimeter, while the photometric sensitivity with a blue filter is 0.5 microgram (8). .I spectrojihotomotric study in the vi6it)Ie range (2’) has established that the ahsorption of ceric sulfate illcreases continuously up to 480 mp (t.he lower limit of the study), so that as pointed out by Sandell ( S ) , increased sensitivity would be expected to result from use of monochromatic violet light. Since completion of the present, xvork, Freedman and Hurne ( 1 ) have shown that the maxirniini ahsorption occurs in the neighborhood of 315 I ~ I P , and have reported on work in which cerium was oxidized by a modification of the procedure of Sandell, and measurements were made a t 315 mp.
of 0.025 microgram can thus be realized. In the determination of cerium by oxidation with persulfate, it is important not to take too large an amount of persulfate or of ammonium ion, because both residual persulfate, and nitrate ion formed by oxidation of ammonium ion, absorb appreciably at 320 nip. 4 spectrophotometric procedure for the deterniination of traces of ceriii~iiimproves sensitivity twentyfold oter the pre+iouscolorimetric procedure.
ABSORWION SPECTRUM OF CERIC ION
The absorption spectrum of a solution of ceric sulfate (1.92 X 10-5 V) in 1 N sulfuric acid, obtained with a Cary recording spectrophotometer (10.0-cm. cell), is shown in Figure 1. T h r maximum, which is fairly broad, occurs a t approximately 320 mp. The rnolar extinction coefficient is 5.58 X IO3, permit,t,ing a photometric sensitivity of 0.025 microptam of cerium. The effect of v i d strength on the location and height of the ma?;iniuni is slight,, over the range 0.1 to 6 .V sulfuric acid. DEVELOPMENT O F PROCEDURE
I t was first attempted to determine cerium according to the procedure of Sandell (8), with the modification that the absorption was measured a t 320 mp, using a Beckman DU spectrophotometer with a 1-cm. silica cell. In this procedure, 10 ml. of solution are made 1 N in sulfuric acid, and 0.2 gram of ammonium
ANALYTICAL CHEMISTRY
454
persulfate and 0.5 mg. of silver nitrate are added; the solution is boiled for 5 minutes, cooled, and diluted to 10 ml., and the transmittance is then measured. This procedure gave erratic results when the measurements were made a t 320 mp. Further study revealed two principal sources of error: absorption of light a t 320 mp by any persulfate remaining after boiling, and absorption by the nitrate ion formed by oxidation of the ammonium ion by persulfate. ;In absorption spectrum of 0.0010 M ammonium persulfate in 1 -Ar sulfuric acid, measured in the 10-em. cell of the Cary spectrophotometer, with 1 Y sulfuric acid as the blank, is shown in Figure 2, A . The absorption increases continuously as the wave length is derreased; this is also seen from the curves of Figure 2 , B and C, which were obtained with 0.010 and 0.10 Ji persulfate, respectively. 4 n absorption spectrum of potassium nitrate is shown in Figure 3: the location of the maviniuni at 301.5 mp has been established previously ( 6 ,9).
' 2,
0
1
\
>2 0 0320361 220 230 240
250
260
270
WAVELENGTH (mp)
)20 400 1
360
Absorption Spectra of Ammonium Persulfate in 1 N Sulfuric Acid
Figure 2.
A . 0.001 M ammonium persulfate B . 0.01 M ammonium persulfate C. 0.1 M ammonium persulfate
I
I 240
260
280
300
320
340
360
380
400
WAVELENGTH I m p )
14-
Figure 1. -4bsorption Spectrum of Cerium(1V) in 1 Sulfuric Acid
N
Both of the above sources of error can be made negligible by using one tenth the amount of persulfate recommended by Sandell (8). While satisfactory results are obtained with this amount of ammonium persulfate, it is evident that potassium persulfate is to be preferred. The solutions after oxidation are stable for 10 to 20 minutes, if water of good quality is used throughout; with water containing traces of organic matter, rapid fading has been found.
12-
ldz
-
0
z5 c
6.4
-
2-
240
Table I.
Recommended Procedure with Varying Amounts of Silver Nitrate and Potassium Persulfate [ a l l with 155 y Ce(III), boiled 10 min.]
AgXOj, Mg. 0.50 0.12
None 2.50
0.50 0.50
KzSzOa, Mg. 24 24
24 24 6.0 96
Extinction (320 m d 0.640 0,648 0.598 0.660 0.640 0.666
Recommended Procedure. The amount of cerium taken should be between 0.04 and 0.20 mg. for maximum accuracy (although as little as 0.003 mg. can be detected), in 10 ml. of 1 K sulfuric acid in a 30-ml. beaker. To this solution are added 0.5 mg. of silver nitrate (0.2 ml. of a 0.25yGsolution) and 24 mg. of potassium persulfate (1.0 ml. of a 2.4y0 solution). A small silicon carbide chip is added to promote even boiling, the beaker is covered, and the solution is boiled for 5 to 10 minutes, with addition of water if necessary to maintain a volume of 6 to 10 ml. The beaker is then placed in cold (15 ") water for 5 minutes, and the solution is transferred to a IO-ml. volumetric flask and made up to volume. The extinction is measured a t 320 mp in a silica cell, and the concentration of cerium is determined from a calibration curve. RESULTS OBTAINED WITH RECOMMENDED PROCEDURE
Construction of Calibration Curve. To 5.00 ml. of ceric sulfate (0.0888 K ) 0.1 iV hydrogen peroxide was added until the solution was colorless. From this was prepared by successive
260
280
300
320
340
WAVELENGTH I m p )
Figure 3.
Absorption Spectrum of Potassium Nitrate (0.0101' M )
dilutions a solution 1.11 X M in cerous ion, in 1 X sulfuric acid; aliquots of this solution were treated by the above procedure, and the extinctions were measured in a 1-cm. cell with a Beckman DU spectrophotometer (hydrogen lamp, 0.34-mm. slit width). The results are shown in Figure 4. Beer's law is followed closely over the range studied; the extrapolated blank (0.003) is virtually negligible. Similar curves obtained a t wave lengths other than 320 mp are also shown in Figure 4; Beer's law is followed a t these wave lengths also, although the sensitivity is, of course, not as great. Variations in Recommended Procedure. TIMEOF BOILING. Measured extinctions of solutions initially 1.11 X lo-' M in cerous sulfate, oxidized according to the above procedure with various times of boiling, are shown in Figure 5 . Satisfactory results are found with between 5 and 35 minutes of boiling. AMOUNTS OF PERSULFATE AND SILVERNITRATE. The effects of variations in the amounts of these ingredients are shown in Table I. Satisfactory results are obtained with the recommended amounts of silver and persulfate, or with as little as one fourth these amounts.
V O L U M E 2 3 , NO. 3, M A R C H 1 9 5 1 Table 11.
455
Oxidation of Ammonium Ion to Nitrate Ion in Absence of Cerium
(Solutions boiled 10 minutes, then made up to 10 ml.)
KzSzOa, M x 102
NHd+ .M x 102
8.9b
AgNOa,
0.89
..
0:io 0.50
1i:i 17.8 17.8 1.78c 1.78 1.78 200. 20.
8.9 8.9 8.9 0.89C 0.89 0.89 0.89
HzSO4
AIg.
l'2b 1 AV
0150 0.50
1" 1 "vr 1N
0:bO 0.50
h'oa Extinction Formed, (320 mp) M X 109" 0.054 0.079 i.'~ 0.017 0.2 0.008
0.023 0.005 0.004 0.054 0.005
With the recommended amount of potassium persulfate, and 2 .M ammonium ion (added as ammonium sulfate), the amount of nitrate apparently formed is greater than would correspond stoichiometrically to the amount of persulfate taken. The significance of this is not clear at present.
0.0
0.4 0.1 0.0 1.2
lAV 0.1 Calculated from extinction at 301 mp, corrected for blank (boiled without ammonium sulfate). b Unboiled. extinction due t o persulfate. C These amounts correspond to 20 mg. of ammonium persulfate (equiralent to 24 mg. of potassium persulfate).
Table 111. Oxidation of Ammonium Ion to R'itrate Ion in Presence of Cerium (-411solutions 1 N in sulfuric acid, with 0.50 mg. of AgNO3, 24 mg. of KzSzOs, and 155 y of cerous or ceric cerium. Solutions boiled 10 minutes, then made up to 10 ml.) NO>SHI~, Extinction Formed. M x 102 Cerium (320 m r ) .II x 102" 0.640 Cerous .. 0.643 Ceric ,. 0.870 Cerous 4.2 200 0.665 0.4 Cerous 20 0.635 Cerous 2 0.0 0.837 Ceric 3.2 200 a Calculated from extinction at 301 m p , corrected for blank. -
EFFECT OF AMMONIUM ION
Oxidat,ion of ammonium ion to nitrogen and nitrate ion by persulfate in the presence of silver ion (at 25" C.) has been reported by Marshall and subsequent workers (3,4,6 , 7 , IO). Results obtained under the conditions of the present procedure are given in Tables I1 and 111. From the data of Table I1 it is seen that the persulfate-ammonia reaction leads to the formation of much more nitrate in initially neutral solution than in 1 *V sulfuric acid; this effect of acid is in agreement with the observations of Marshall and Inglis ( 7 ) . I n 1 N acid, the nitrate formed from ammonium persulfate in the amount used by Sandell (8)is sufficient to give a significant ext,inction at 320 m& even in the absence of cerium.
CONCENTRATION OF CERIUM T A K E N l M X 1 O 5 )
!':::I Figure 4. Extinction us. Concentration of Cerium Taken in Recommended Procedure
,640
5 630 W
620
~i~~~~ 5 .
5
~
10 15 20 25 T I M E OF BOILING ( M I N U T E S )
30
~at 320 mF~ us. ~i~~ i of ~
35
~ ~i
I n t,he presence of 155 micrograms of cerium(III), the extent of formation of nitrate is severalfold higher than in the absence of cerium. Thus, with the recommended amount of persulfate, the initial presence of 0.2 M ammonium ion causes a significant error in the cerium determination, although the corresponding blank (Table 11) is negligible. However, with 0.02 41 ammonium ion, which Table I\'. Interference by Various Metals in Determination of Cerium is slightly greater than the amount of am[All extinction coefficients are those of individual compounds; extinctions are those of monium ion present in 2o mg' Of ammonium indicated mixtures with 1.11 X 10-4 JM cerium(II1) carried through recommended procedure with 20 mg. of (NHdzSzOs] persulfate (corresponding to the recommended Molar Extinction Coefficient, Amount amount of potassium persulfate), no error is found. e , Sq. Cm./Millimole Taken, Extinction Compound 320 m p 340 mp 270 mp 320mp 340mp When cerium is present initially in the quadriTaken valent rather than the trivalent state, the amount Cw(8Oc)aalone 4 . 8 3 X 103 4 . 1 1 X 103 2 . 2 6 X 103 1 . 1 1 X lo-' 0 . 6 4 0 0 . 5 5 0 CuSOr.5HzO .. .. 2 . 8 X 10' 1.0 X 1 0 - 2 0.639 0,550 of nitrate formed (from 2 4f ammonium ion) is 1 . 0 X 10-1 0 . 6 7 6 0 . 5 5 6 considerably greater than in the absence of cerium; FeSO4'7HzO or Fez (S04) 3" 6 . 3 X 102 3 . 3 X 102 j.8 X 102 2.0 X 0.684 .. this indicates t h a t the cerium acts here principally 1.0 x 10-4 0.920 .. as a catalyst rather than by an induced reaction. KzCrzO7 6 . 3 X 102 1 . 2 X 103 2 . 0 X l o 3 6 . 9 X 10-8 0.642 .. c
_
_
_
_
x
KRInO4
1.5
NH4VOs
3 . 2 X lo2
103
1.2
x
103
1.9 X loz
UOzSO4.3HzO
1 . 2 X 102
2 . 3 X 101
Th(NOs)r,4HzO
1 . 6 X 102
1.2
KSOs
3.0
x
101
..
Ferric sulfate taken for determination of analytical procedure.
3.3 X 1.3 X 2 . 4 x io2 1.0 x 4.0 X 2.0 x 1.0 X 1.0 X 10s 1.0 X 1.0 X 1 . 0 X 102 1.0 X 1.0 X 1.9 1.0 X 1.0 X extinction coefficient:
..
0.653 0.707 0.634 0.641 0.639 0.661 0.653 0.758 10-5 0 . 6 3 6 10-4 0.660 10-2 0.675 10-1 0.980 ferrous sulfate 10-6 10-4 10-5 10-6 10-3 10-4 1O-I 10-4
.,
..
..
EFFECT OF METAL IONS
..
A number of metal ions, of such a nature as would be expected either to occur with cerium in nature or to interfere with the cerium determination, have been studied according to thefollowina d a n . First, the absorDtion spectra of solutions of C.P. compounds in 1 JiT sulfuric acid were determined in the ultraviolet, with the Cary instrument (10-cm. cell). The extinction coefficients a t 320 mM were calculated, and on this basis various
..
,.
0.556 0.574
0.552 0.563 0.656 0.562 used in
l
i ~~
~
456
ANALYTICAL CHEMISTRY
amounts of each substance were taken, toget,her with a known amount of cerium(III), for analysis according to the above procedure with ammonium persulfate (20 mg.) as reagent. \Yith many metals it n'm evident from the absorption curves that the int,erference could he greatly reduced, a t comparatively slight cost in sensitivity of cerium determinat.ion, by ivorking at 340 rather than 320 nip. The results are given in Table IY :inti Figures 6 and 7. D a h for potassium nitrate are also incluc!cvi.
-
,
!
I
240
260
Figure 7 .
280
\
320 340 WAVELENGTH ( m p )
300
360
380
400
,Absorption Spectra o f Varioiis Compounds iii
Ultraviolet
than expected on the b:tsis of its extinction coefficient, wit,h manganese, on the othrr hand, less interference was foiinii than expcrted.
\ V ~ Rhigher
,
240
Figure 6.
,
,
260
280
--- --___ ..., 300 320
-___
\..
-*
340
WAVELENGTH ( m p l
,~j
___----- --.-&\
360
aeo
Absorption Spectra of Various Compounds Ultraviolet
400
(1) Ireedman, A. J..
L I T E H T I I H E ClTED
anti Hump, D. N.. AXIL. CHEJI.. 22,
932
(1950).
iri
T h i . molar extinction coefficients a t 320 nip of the majority of t,hc compounds of Table IV arr of the order of 10% of that of oerium(1V). For this reason separations must be made from ii larger number of elements than if t.he color of thc ceriuni(1V) is mc:nsured in the violet, portion of the visible spectmm. On thc ot,her hand, the interference due to copper, particularly at 340 mp, is so slight that it is possihlr to determine cerium accuratel). i n a strong blue solution Ttw interfcrrnee found with iroii
(2) Kasline. c. T., and >Ie1loii, 11. (;., Isn. EN;. CHEA!... $ X A I Eo.. 8, 4(i3 (l9.'36i. V.,J . . i f f /( ,' h t u z . .So,., 49, 2659 (1927); SO, 2080 (11928). (4) Kiig. ('. V., and Giiawold. I . I,., Ibid., 32, 14Y3 (1:)3?). ( 6 ) Ilortiirn. G., Z . phusik. Chf,n?.,B43, 415 (1930). (ti) .\IarshiLll, H., Proc. Roy. ,Sot. Edinburgh, 23, 163 (19OUi. ( 7 ) 3Iarshall. H.. and Inglis, .I. li. H.. Ibid., 24, 58 (1902). is) &indell, E. R.,"CoIorinieti4c. Determination of Traces of .\letals," New York, I n Publishers, 1944. (91 ~chribe.G.. Ber., 59, '6). (10) Yost, D. &I., J . A V L .C'het~/.SOC..48, 374 (1026). ( 3 ) King, C.
RI.:CEI\-ED July 21. 1950. R e m a r r h carried oil: )under the aiisgicea of thr Atornic Energy Commission
Spectrophotometric Determination of 8-Quinolinol And Some of Its Halogenated Derivatives W. T. HASKINS AND G E O R G E W . LUTTERMOSEH National Institutes of Health, Public Health Service, Bethesda 14, .\Id.
I
N CONR'ECTIOS Lvith a study of the mode of action of t,litx
iodine-cont,aining 8-quinolinol compounds uwd for the trentmerit, of amebiasis, it was desired to cstimate quantities of the ortiel, of it fraction of a milligram of the drugs arid their partially o r romplt~tc~ly dehalogcnatcd derivatives. Other investigators ( I , 9,4 , 7, IO, 1 2 ) and t,he authors (9) have studied the abporption, diet,rihtion, i t r i d excretion of these iodinated compounds heretofore by tiekrmining the amount of iodine prrwnt in the biological matrrials by chemical or tracer twhniqutw. I-Iowevcr, it, may not al\vays he assumed that the iodine measured in this way is still bound t>othe 8-quinolinol riuclcus, for it, may have beer reniovcd hy some met,abolic process and subsequently transferred t o some othrr moiety. Thus it t)ecomc:s desirable to determiiic. riot only the iodin(: associated with thcb drug but also the quaxitit>. o f 8-qUinOliIKJl prcsent. Published mc)thods for the determination of 8-quinolinol depend 011 t.itration or gravimetric precipitation with bromine (6, Y, 8, 11) and hence are not applicable to dcrivittives which are alrriidy s u b s t i t u t ~ ~ind thcb 5 and 7 positions. Fresenius (6) has
dcxribed a spectrophotometric method for the determinatioii of 5-chloro-7-iodo-8-quinolinol by measuring the color developed ill glacial acetic acid in the prtscnce of ferric iron. However, this method suffers because of its cstn:mcL sensitivity to the water content of thc solvent. Grabbe ( ? ) stated that a colorimetric mctthod W M employed for the estimation of small amounts of 8-quinolinol using ferric chloride in aqueous solution, but gave no details. It was recognized that the intense grwn color of the complrs formed by 8-quinolinols with ferric iron offered a wnsitivr and n~lativclyspecific met,hod for their determination by spcmcatrophotometry, provided a suit,ahlr solvent n-ere available. It I V ~ R found that methyl Cellosolve (2-mcthosyethanol) was satisfnctory for the purpose, as it is a good eolvcnt, for &quinolinol, 5rhloro-8-quinolinol, S-chloro-7-iodo-8-quinolinol(vioform), 5,Tdiiodo-8-quinolinol (diiodoquin), and their iron complexes. The color produced by the iron complexes in this solvent, which is not hygroscopic, is insensitive to small changes in water content. RIet,hyl Cellosolve is not suitable for the determination of 8quinolinol-5-~ulfonicacid or T-iodo-8-quinolino1-5-sulfonic aciil.
