ANALYTICAL CHEMISTRY
726 however, are the results obtained with ether-soluble lac. The values reported in the literature for this portion of lac are generally tietween 90 and 110, compared with the low value of 70 obtttiiictl t)y the present method. Possibly the ether-soluhle portion of lac contains ester linkages highly susceptible to alkali; a structurr involving a lactone linkage has been suggested by Bhowmik and Srn (I). The saponification value of the ether-soluble lac is about the same as that mentioned in the literature. LITERATURE CITED ( I ) I3howmik, T., and Sen, H. K., J . Indian Chem. SOC.,I n d . Ed., 7, 160 (1944). (2) Diitish Standard Specification for Bleached Lac, S o . 1284 (1 946). (3) British Standard Specification for Lac, No. 954 (1941).
(4) Gardner, 1%’. H., and Whitmore, W.F., IND.ENQ.CHEM.,A N A L . ED.,1. 205 (1929). (5) Gidvani, B. S., and Dobbie, J. hl.. London Shellac Research Bureau, Bull. 4 (1940). (6) Knott, E. B., J . SOC.Chem. Ind. (Trans.). 59, 252 (1940). (7) Kolthoff, I. hT.. and Furman, N. H., “Volumetrio Analysis,” Vol. 11, pp. 350-89, New York, John Wiley Br Sons, 1929. (8) Ludtke, M.,Angew. Chrm., 47, 520, 537 (1934). (9) hlurty, N. N., \Veiiiberger, H., and Gardner, W.H., Indian Lac Research Institute, Bull. 25 (1936). (10) Xabar, G. XI., and Padmanabhan, C. V., private communication. (11) Rangaswami, hI., and Sen, H. K., “Handbook of Shellac Analysis,” p. 65, Indian Lac Research Institute 1942. (12) Rueiczka, W., Farben-Chem., 13, 25 (1942); Peinturea, pigments, nernis, 17, 610 (1942). (13) Reinberger, H., and Gardner, W.H., IND.ENG.CHEM.,h A L . ED.,5, 267 (1933). RECEIVEDApril 12, 1949.
Microdetermination of Iodine An Improvement in Reflux Distillation Apparatus and Technique J. W. THORIAS, L. A. SIIINN, 11. G. WISEM;IN, AND L. A. MOORE Bureau of Dairy Industry, Unitad S t a t e s Depa;tment of Agriculture, Washington, D .
IiILEE papers (2, 3,8) describe the determination of iodine in rsmall amounts of plasma by employing the catalytic effect of iodide on cerate reduction by arsenite. There are certain undesirable features of each proposed method. The Chaney micromethod (5)as modified by Taurog and Chaikoff (8)has given very low and variable recoveries (10 to SOYc)in the authors’ laboratory and i n others ( 2 , 5 ) . Barker’s modification ( 2 ) of the absorbing solution used by Talbot ( 7 ) is undesirable because i t greatly lengt hcns an already long procedure. The procedure outlined by Taurog and Chaikoff (8)and Chaney (S), consisting of [vet-ashing the sample with chromic-sulfuric acids folloived by distillation, has been adapted \vith two principal modifications, and used with good results. The addition of a sodium hydroxide in small amount of arsenitc to the 1.2 ml. of the alisorption trap made possible satisfactory (94 to 1OOC;) recovc8ric.s of a d t l d iodine. One or two drops of a sodium arsenite solution \Yere used, made by dissolving 2.43 grams of arsenic triositle in 15 nil. of lvater containing 1.7 grams of sodium hydroxide, diluting, neutralizing with sulfuric acid, adding 10 grams of sodium bicarbonate, and diluting to 500 ml. A solution of arsenic triositle alone in sodium hydroxide would probably be adequate for t h e absorbing liquid. In other halide distillation techniques ( I , 6 )arsenite has been used in the trap in order to ensure quantitative rrcoveries. The reducing medium in the absorption trap pro1)ably prevents oxidation of the distilled iodide. Iodide is then determined in the distillate diluted to the desired volume, using the catalytic action of iodide on the reduction of cerate by arsenite. A simplified rrflus still (Figure 1 ) permits the operator to digest and distill from the same flask having but one standard-taper ground-glass joint. Aktu;tlly a reflux still was found to be unnecessary for recovery of iodide, but the volume in the boiling Bask must be maintairied. Add 25 ml. of redistilled uater to the chromic acid-digested sample in a ,500-ml. flask. Put 1.2 ml. of 1% sodium hydroxide and 1 drop of 0.1 .V sodium arseiiitc solution into the absorption trap, \vliic,h is w l l insulated k). ivwpping it with asbestos or cork. h the fhsk t o the refliir still and heat it. When disirts atltl 3.0 nil. of 50% phosphorous arid to the flask through the droliping funiicl. It is not necessary to use pressure to fowe the phosilhoroii* : i v i t I through the dropping funnel, as it is when using r l l c s Rigas moilifc~:ttionof the C‘hancy still ( 2 , 5, 8). I>istill for 7 t o 8 millute., removc he:it from flask, and draw off distillate into :t 25-1111. volumc~tric,fla6k. K a s h the trap with sm;ill portionr of \v:ii-ni rc~lic~illcd \v:itrr and collect thcsc in the voliimrtric f h s k . It i- niuc,h rinipler :tnd emier to wash out this ati~i~rptiori tr:iri , i f : c - r . i ~ i c . 1 1~1istiIl:~tion than to wash the trap on
C.
the Riggs-Chancy still. The distillate volume is made up-to 25 ml. when aliquots are taken for the colorimetric determination of iodide by its catalytic effect on reduction of cerate in the presence of excess arsenite. The float valve is merely a thin-walled hollow glass sphere with a short tail. The capillary water return is 6 cm. long and is turned up 2 cm. Its inside diameter is 1 to 2 mm The entire still w a ~ made by one of the authors (H.G.W.). Details of other techniques and procedures have been sdequately described ( 2 , 3 ,8). LVith one exception all samples of C.P. chromic acid that were tried contained too much iodine for this procedure. However, a sample of Grasselli’s technical grade was found to be sufficiently low in iodine (approximately 0.05 microgram per gram). GreenTo W a t e r
Condenser
9
3
Absorption Trop
3
-
Funnel
I I 1 I 1
To D i g e s t i o n Distillation
Flosk
Figure 1.
Reflux Distillation Apparatus
727
V O L U M E 2 2 , N O . 5, M A Y 1 9 5 0 man (6) used a sample of Fisher's technical grade that contained nearly 0.03 microgram per gram. One sample of C.P. arsenic trioxide contained a substance that interfered with the iodidecatalyzed reduction of cerate by arsenite. With the authors' still, boiling for 7 minutes after addition of the phosphorous acid gave recoveries of 94 to 100% when arsenite was used in the trap. Using arsenite in the distillation trap Greenman ( 5 ) reports recoveries averaging 95% for iodide in diiodotyrosine compared with low variable amounts without the use of arsenite. Since this manuscript was prepared another article describing a suitable still for this purpose has come to the attention of the authors ( 4 ) .
LITERATURE CITED (1) (2) (3) (4) (5)
Alford, W. C., J . I d . H y g . Tozkol., 29, 3 9 G 9 (1947). Barker, 5. B., J . B i d . Chem., 173, 715-24 (1948). Chaney, A. L., IND. ENG.CHEM.,ANAL.E D . , 12, 179-81 (1940). Connor, A. C., et al., Surgery. 25, 510-17 (1949). Greenman, University of Pittsburgh, personal communication.
1948. ( 6 ) Martinek, M. J., and Marti, W. C., IND.ENQ.CHEM.,ANAL.EO., 3, 408 (1931). (7) Talbot, N. B., Butler, A. M., Salzman, A. H., and Rodriguez. P. M J . Biol. Chem., 153, 479-88 (1944). (8) Taurog, A., and Chaikoff, I . L., Zbid., 163, 313-22 (1946). RECEIVED June 8. 1949.
Polarographic Data on Zinc in Small Concentrations Deviations from the Ilkovic Equation LUDWIG S . CH.4MPA AND ABRAHAM WALLACH Division of Industrial Hygiene, Montana State Bonrd of Health, Helena, .Wont.
the development of a method for the determination D URING of small amounts of zinc in various concentrations of very
dilute sulfuric acid by polarographic means, a lack of proportionality between wave height and concentration was noted. A closer study verified these initial findings and the results that were obtained under varying conditions of drop time and concentration are reported below. APPARATUS
The determinations were made with a Sargent Model XI1 polarograph. Lithium chloride (0.1 N ) was w e d as the supporting electrolyte, because maxima effects-reported with other alkali chlorides-were absent. The galvanometer calibration factor, according to the manufacturer, was 0.0051 microampere per millimeter. The ratios of the sensitivity settings (Zto 5, 5 to 10, 10 t o 20) were checked in triplicate, using various concentrations of cadmium (with zinc there was a noticeable decrease in diffusion current with an increase in applied e.m.f. after 1.0 volt), in order to make sure that conversion from each of these four sensitivities was permissible on the instrument. To ensure more accurate measurements of the wave heights, the concentration of cadmium in each case was such that the wave height for the less sensitive reading was from 46 to 60 mm. and that for the corresponding more sensitive reading was 112 to 120 mm. The maximum deviation during each set of runs varied from less than 0.5% for the JO, 20 comparison to less than 1% for the 5, 10 and 2, 5 comparisons. The ratio of average wave heights for the 10, 20 comparison was exactly 2.00; for the 5, 10 sensitivity comparison was 1.99 (instead of 2.00); and for the 2, 5 comparison was 2.53 (instead of 2.50). EXPERIMENTAL
Kolthoff and Lingane (3)have stated that with a drop time between 3 and 6 seconds there is strict linear proportionality between the diffusion current and the concentration of metal ions analyzed. However, they mention that other investigators (1, 6) have claimed that at very small concentrations the diffusion current is greater than corresponds to strict linear proportionality with the concentration. Because the authors were dealing with small amounts of zinc, a series of experiments was run with dropping times varying from 2.77 to 5.90 seconds to determine whether or not there was strict proportionality between current and concentration. The first series of experiments was made with five different concentrations of zinc, varying from 0.10 to 0.005 mg. per ml. (1.53 X loe3to 0.77 X 10-4 M ) . A primary solution was prepared by dissolving zinc sulfate heptahydrate, and standardized by the potassium ferrocyanide method mentioned by Treadwell and Hall (7). From this solution a standard was prepared containing 0.1 mg. of zinc per ml. in 0.1 N lithium chloride. The other concentrations were obtained by taking aliquot portions (25 or 50 ml.) of the 0.1 N standard or weaker concentrations alread made, and diluting them in proper volumetric flasks with 0.1 N Ethium chloride.
The sensitivity used for 0.10 mg. per ml. was 20; for 0.05 mg. per ml., 10; for 0.02 mg. per ml., 5 ; and for 0.01 and 0.005 mg.oper ml., 2. The temperature of the water bath was kept a t 25 * 0.2" C. In measuring the wave heights for zinc, it waa found that reproducible results were best obtained by drawing the limiting current line parallel to the residual current whenever its angle with the horizontal was less than that formed by the residual current. (There is a decrease in diffusion current with an increase in e.m.f. above 1.0 volt.) Corrections for residual current were made, as indicated by Kolthoff and Lingane ( d ) , in measuring diffusion current. The polarogram obtained at the 3.Gsecond drop time were run in triplicate on each of two separate portions of solution. With the other drop times, determinations were mostly made in quadruplicate on the same portion of solution. Where the deviation on one of these four results was greater by four times the average deviation of the remaining results, this single result wm discarded. The data are shown in Table I. DISCUSSION
Ion Concentration. iiS can be seen from Table I, the ratio of diffusipn current to concentration is not constant within the range of zinc concentrations tested, but increases progressively with decreasing concentrations a t drop times less than 3.7 seconds, being more marked with lower drop times. The ratio of diffusion current to concentration also increases a t higher drop times (4.18 and 4.90 seconds). Xhether it is progressive or not was difficult to ascertain, because an experimental variation in measurement of the polaroprams of even 0.5 mm. a t these drop times can cause a relative difference of 1%. For example, if one were to use the 43.8 figure (drop time 4.18, concentration 0.005) the average wave height would be 43.5 and the per cent deviation 3.3 instead of3.1. Effective Pressure. In view of these observations, the data were analyzed to see whether (with all other factors constant) the diffusion current was Broportional to the square root of the effective pressure on the dropping mercury. According to data indicated by Kolthoff and Lingane ( 4 ) the back pressure due to interfacial tension a t the surface of the mercury drop varies, with usual capillaries used, from 2 to 3 cm. of mercury. In making this correction on the height of the mercury column, the figure 2.5 cm. was used. Any error due to a deviation of 0.5 cm. would make little difference in the square root of the effective heights used in the data in Table 11. The data in Table I1 indicate that although with a concentration of 0.1 mg. of zinc per ml., the ratio of id/h1/2does not vary to any greater degree with varying pressures of mercury columns than the data obtained by Maas ( 6 ) , this is not so with lower concentrations of zinc. At a concentration of 0.005 mg. of zinc per ml., the deviation is about 14.2y0 with a difference in pressure of only 257,; and for a range of pressure that varies twofold, there is a deviation in the ratio of i d / h l / *of about 17.9%.
