Absorptiometric Determination of Lead in Rubber Products and Compounding Materials K. E. KRESS The Firestone lire & Rubber
Co., Akron, Ohio
b A semimicro ultraviolet absorptiometric method that is rapid and sensitive has been developed for the quantitative determination of lead in rubber products and trace amounts of lead in rubber-grade zinc oxide. The ash of rubber products, or the inorganic material as is, is dissolved in a minimum amount of concentrated hydrochloric acid, then made up to contain a Soy0 excess of the acid. The absorbance is measured at 250, 270, and 289 mp to calculate lead content. The interference of iron and most other cations is eliminated mathematically as background absorbance. Copper content must be less than about one tenth of the lead present.
T
litharge (PbO) as a n inorganic accelerator of vulcanization of rubber products is no longer as important as it was prior to the discovery of organic accelerators (3). However, the use of lead in some form is still prevalent in a limited number of products (2, 4 ) . Table I illustrates applications of inorganic and organic lead compounds in rubber products. Lead compounds are still used primarily to activate or accelerate vulcanization, but metallic lead has found application as a filler in balance dough because of its great density. An analysis for type of accelerator in a n unknovvn rubber product (6) viould not be complete in some cases without a check for lead. The ASThI procedure for the chemical analysis of rubber products includes a conventional macro scale gravimetric method, where lead is determined as lead sulfate ( I ) . A sensitive absorptiometric procedure has been developed for microdetermination of sulfur in rubber products as lead sulfate ( 5 ) . A similar technique is here adapted to the rapid determination of lead as a minor--e.g., 0.1 to 5%-component in rubber products on a micro scale. Lead may be determined in rubber-grade lead compounds as a control test by the same micro method. The basis of this test is the absorbance of lead in 50% hydrochloric acid, in R hich it has a sharp symmetrical maximum a t 270 mp (Figure 1). HE USE OF
Measure the absorbance a t 250, 270, and 289 mp, using a blank of 50% acid. If absorbance is above 1.8, dilute with 50% acid as needed. Include all dilutions in calculating sample concentration, c . Substitute appropriate absorbance values in the equation below t o determine lead content of the sample.
METHOD
Equipment. Beckman Model D U 1.OO-cm. spectrophotometer with quartz cells. Borosilicate glass beakers, 5 ml., for dry ashing or test tubes, 16 X 100 mm., for wet ashing. Glass-stoppered graduated cylinder, 10 ml. Reagents. Hydrochloric acid, 50% by volume. Make a n exact 1 to 1 dilution of reagent grade concentrated hydrochloric acid (370j0). Instrument Calibration. Refer to the calibration procedure outlined by Kress (5). Procedure. Homogenize the sample in a rubber mill, then weight a 1to 10-mg. sample, or more if less than 1% lead is expected. ilsh in a 5-ml. beaker a t 550" C. for about 10 minutes, then in the front of a high-temperature muffle (900' C.) for 1 or 2 minutes until ashing is complete. [Alternately, wet ash with 3 to 5 drops of perchloric acid and about 1 ml. of concentrated nitric acid in small test tubes (6)1. Cool the beaker, then dissolve the ash in about 2 ml. of 50% hydrochloric acid, stirring with a glass rod if necessary. Transfer to a 10-ml. glass-stoppered cylinder using 50% acid wash. Dilute to volume with 50% acid and mix Gell. If wet ashing is used, pipet 10 ml. of 50% acid into the test tube and mix.
Table 1.
WAVE L€NGlH tl'i)A Figure 1. Interference of ferric iron in 50% hydrochloric acid
--- 500 mg. of zinc oxide with impurity - 0.2 mg. of lead alone
lead and iron
Applications of Lead in Rubber Compounding % Lead
% Lead
Compounding Ingredient Inorganic Litharge (PbO) Red lead (PbaOa) Metallic lead
Content (Theory)
Typical Application
93
1 to 3y0 as activator for
91 100
Organic Methyl ledate (lead dimethyl dithiocarbamate)
46
SPDX (lead phenyldimethyldithiocarbamate complex)
34
GF y ies d iue Expected in Ash
1t o 3 carbamate accelerators in natural rubber wire insulation 3 t o 5% as activator for 3 to 5 oxime accelerators in synthetic butyl tubes Filler in tire balance 50 dough to add weight
0.2% to 0.6% as acceler- 0.1 to 0.3 ator for natural rubber
electrical wire insulation 0.1 to 0.4% asaccelerator 0.03 to 0 . 1 5 for natural and GR-S synthetic rubber stocks
VOL. 29, NO. 5, MAY 1957
803
~~~
~
Determination of Lead in Rubber Products
Table II.
Theoretical Form and c;
Product Tire balance dough No. 1
Metallic lead, unknown
Methods of Anal\ sis Absorptiometric Gravimetric Ket ash Dry ash
Electrolytic 47 46 48 Av. 48
..
9 7 7 O k O 7
Tire balance dough KO.2
Metallic Lead, unknown
...
Butyl stock No. 1 (production)
PbrO4", 5 . 1 0
...
...
Butyl stock S o . 2 (three different pro- Pb304,5.10 duction samples, a , b, c)
...
...
Control stock No. 3
...
47 3
47.1 52.9 50.8 52.8 Av. 4 9 . O z k l . 8 5 2 . 8 z k 0 . 1
50.6 50.1 5 0 . 4 f0 . 3 5.28 5.42 Av. 5 . 3 5 4.17 4.52 4.00 4.34
... a 5.55 b 5.23
PbaOa, 5.10
...
c 5.49 5.11b
Reported as Pb304corrected for rubber-grade pigment purity of 95% as Pb304 by analysis. Higher recovery by wet ashing makes this method preferred, but dry ashing is usually satisfactory if care is taken to dissolve all of the ash residue. a
*
yo lead
=
A AS,,, X 100 Aap,b,, X c(mg./ml.)
(2)
The data of Table I1 illustrate the precision of the absorptionietric determination of lead in rubber products. Agreement lvith both electrolytic and gravimetric methods is good. With the absorptionietric method, there is no danger of loss through solubility of the lead sulfate, as n-ith the grayimetric procedure. RUBBER C O M P O U N D I N G MATERIALS
I n the manufacture of white rubber products, it is undesirable to have lead compounds present, because they form lead sulfide with sulfur added for vulcanization and cause a brown discoloration. Consequently, it is necessary to control the lead content of zinc oxide, which is a common ingredient in tire white d e n - a l l stocks. The lead content of rubber-grade zinc oxide is usually below 0.2%. The electrolytic procedure used, which normally requires a large sample-e.g., 20 gramsis not believed to be accurate a t lead concentrations beloiv about 0.01%. There is some danger of occlusion of zinc, and the weight of lead recovered is often only 1 or 2 mg., which cannot be weighed accurately enough with normal macro-sized electrodes when lead content is critical. The absorptiometric method offers a n unusually rapid procedure which is of acceptable accuracy for traces of lead as low as 0.001% in zinc oxide. Lead in Zinc Oxide. The sample size of zinc oxide is determined b y the normal lead content as follows: 804
ANALYTICAL CHEMISTRY
Normal Lead Sulfate Content, 7; Below 0 004
0 03 0.1 1
Approx. Sample Weight, Grams (G. Zn0/10 Rfl.) 1 to 2 0 4 0 1 0 01
The ITeights given are close to the maximum allowable in 10 ml. of 50% acid without further dilution. Weigh an appropriate amount of the zinc oxide and transfer to the 10-ml. cylinder. Add just enough concentrated hydrochloric acid to dissolve the sample with vigorous shaking. Add the last of the acid drop by drop. Dilute to 5.0 ml. lvith distilled water, then to 10.0 ml. with concentrated hydrochloric acid. Rlix n-ell and measure the absorbance at 250, 270, and 289 mp. Thereafter, folloir- the procedure for determination of lead in rubber products. The results are usually multiplied by 1.16 to convert to lead sulfate.
The absorbance curves of Figure 2 show why this is so. Apparently the zinc chloride formed has no direct effect on the lead absorbance, but it lowers the total absorbance indirectly by reducing the excess acid concentration. Consequently, all samples for determination of trace amounts of lead should first be just d k o l v e d in the minimum amount of concentrated or 50yG acid before diluting to a 50% excess by volume. Major Interferences. Absorbance in hydrochloric acid is not specific for lead. Other heavy metals, particularly mercury and bismuth, alsoabsorb selectively, but their absorbance
EXPERIMENTAL W O R K
Acid Concentration. The wave length and intensity of maximum absorbance of the lead chloride complex depend on the acid concentration (6,8). Where the lead content is about lYO or higher, as in the ash of rubber products or the analysis of lead compounds, the small amount of ash or the micro-sized sample can often be most conveniently dissolved directly in 50% hydrochloric acid. However, for traces of lead in zinc oxide, the large amount of sample to be dissolved greatly reduces the free or excess hydrochloric acid concentration. Lead sulfate determined in rubbergrade zinc oxide by direct solution in 5070 hydrochloric acid may run as low as 70% of the true values, as determined with SOYo free hydrochloric acid solvent or by electrolysis (Table 111).
0.2 0.1-
\ \
'.---------,
_.__ -.\.
.WAVE LENGTH
.-__.___ ._. Rl&
Figure 2. Effect of excess acid on absorbance maximum Zinc oxide solution containing 0.01% l e a d SUIf a t e impurity In 50% acid, equivalent to 20% excess by volume In 70y0 acid, equivalent to 50% excess by volume Zinc oxide solution containing no l e a d b y electrolysis In 50% acid
---
-.-
in thP 250- t o 280-nip region is 1011 a i d nonselective. -4s minor components, they n ould not cause a n y appreciable interference t h a t u-odd not be corrected for as background absorbance by the calculation developed to eliniinate iron interference. This is also true for the lighter elements: vanadium, chromium, tin. and antimony. which qhon interfering absorbance. Of these six elements, only antimony (as the sulfide) and hisiiirith (as a carbamate accelerator complex) have found any use in rubber compounding. The chlorideq of calcium, magnesium, aluminum, and zinc, n hich are the mo5t common elenients in the ash of rubber product^ hesides iron, have essentially no absorbance in the ultraviolet abore 230 mp. Of trace elements that niay be prerent in rubber products, such as ropper, manganese, cobalt, and nickel, the chloride of copper is the only one that absorbs appreciably above 240 mp (Figures 3 and 4). IROX.If iron interference is exces4ve, a liquid-liquid extraction of the 5O7c hydrochloric acid solution with ethyl ether removes the greater part of the iron without extracting lead chloride. Another chemical niethod for removing iron interference, when present in conrentrations comparable to the lead content, is u5e of titanous chloride to recluce the iron to the ferrous state (8). Normally, the iron content should be loner than lead for best results, but it can be tolerated to the extent of about tnice the lead content. As ferric chloride has a n intense yellow coloration in hydrochloric acid, absence of niore than a light yellon coloration indicates that iron interference is not excessive. K h e n inorganic lead compounds are added to rubber products, the lead cmtent is usually so much greater than Table Ill.
be eliminated mathematically by considering it t o be a linear background absorbance. A suitable equation for elimination of iron absorbance was derived as follows.
210
230 250 270 2bO WAVE LENGTH
30
330
m/u
Figure 3. Negative background interference in 50% hydrochloric acid
- 5.0 p.p.m. o f ferric iron
... . 50.0 p.p.m. o f manganese
---- Antimony
the trace of iron present that absorbance of iron is almost negligible. For rubber-grade zinc oxides the iron content is usually of the same order of, or less than, the lead content. Therefore, in rubber chemistry analyses where lead is present, the absorbance curve in 50% hydrochloric acid is usually one of combined iron and lead absorbance, with the lead maximum a t 270 m p distinct and usually dominant. Figure 1 illustrates the combined absorbance of lead and iron in a rubbergrade zinc oxide. Absorbance will be similar for the ash of rubber products c c Lcad = whose iron content approaches that of lead. Iron absorbance betTveen 250 and (L4;M"- L4;*9) - (-%+E) 289 mp is essentially linear n-ith a small negative curvature (Figure 3 ) , and can AaE!o X c n here
Absorptiometric vs. Electrolytic Determination of Lead Sulfate in Commercial Rubber-Grade Zinc Oxide
c2 Lead Sulfate by Electrolysis
5 Lead Sulfate by Direct Soln. in 50% HC1
2 27 2 35 A V . 2 31 0 58
0.53
0.10 0.11 Av. 0.105 0.099 0.100 Av. 0.10 0.04
0.026
...
1.92
7' Lead Sulfate Found by Adding
Excess of Found 2.20 2,32 2.22 2.32 0.56 0.57 0.132 0.127
50% HC1 after Soln. Av. Mean dev.
0.098
0.098
...
0.036 0.036 2.44
0,036
0.00
0.0004 0.0002
0.0003
2.26
f0.06
0.56
&0,01
0.130
10.002
10,0001
x 100
(3)
A = measured absorbance a t single specified n ave length a = absorptivity a t single n a i e length
For 1-cin. cell, a
=
,4/c
S14f,,= differential absorbance (so dcfined for convenience). ; I . defined mathematically in the tevt it represents an absorbance difference of the sample betneen 270 and 289 nip after correction for the slope of interfering background absorbance. =
Zinc Oxide, C.P.
0.0002 to none (Unreliable at this concentration)
I n Figure 1 the lead absorbance niaximum is symmetrical. I n the absence of iron, absorbance a t 289 nip is essentially identical to that a t 250 inp. The distances between these two wave lengths of equal absorbance and the wave length of the maximum (270 nip) are 20 and 19 mp. To siniplify the calculations of background absorhance, they are considered equidistant (20 nip) froin the wave length of the maximum. and also of equal absorbance. Interference due to intensity of background absorbance may be eliminated by basing calculations on difference in absorbance between 270 and 289 nip. Interference due to increaie in background absorbance betn-eeii 289 mp and 270 mp may be eliminated by determining the change in absorbance of the background over this range. I n Figure 1, z1 is assunietl equal to Q. Then, by similar triangles, interference due to the increase of the background absorbance betn een 289 nip and 270 mp (ul) is equal to W y2 or to % ( A z o - Azs9). Combining these terms, the corrected differential absorbance a t 270 nip due to lead in the sample, after both intensity and increase in intensity of iron absorbance hare been subtracted mathematically n ill be given by the Equation 1. Per cent lead in the sample niay then be calculated from the ratio of the specific absorbance diffcrence for the sample to that obtained for a lead standard ( 5 ) .
sample absorbance
A U E : ~= difference in absorptivity betneen 270 and 289 mp for lead alone in 50% hydrochloric acid. The value should be approximately 50 but varies somewhat Tvith different inqtruinents. c = concentration of sample in grams per liter or nig. per milliliter a t dilution at which absorbance is measured
COPPER. Copper promotes deterioration of natural rubber products and is VOL. 2 9 , NO. 5 , M A Y 1957
805
carefully controlled by actual analysis of the raw materials going into the final product. The normal rubber product contains copper a t concentrations below 0.002%1 which does not interfere with the determination of lead \\-hen lead is added as a compounding material. Most ruhber-grade zinc oxides, with normal lead content below 0.2% as lead sulfate, have lead contents ranging from 10 to 200 times the copper content. With an occasional sample the leadcopper ratio may fall to 6 or rise to over 500. Only with the purest grades of nearly lead-free zinc oxides will the copper content approach that of lead. The ultimate limit and accuracy of determination of lead in purer grades of zinc oxides and other materials depends on whether the copper content is well below that of lead. Figure 4 shows that possible interference due to copper will be below 2% of the lead content as long as copper concentration is below about one tenth of the lead present. If much copper is present, the absorbance a t 289 mp will be greater than a t 250 mp. With iron and most other interfering materials, the absorbance a t 250 mp will normally be higher than a t 289 mp. Therefore] if absorbance a t 289 mp is significantly greater than a t 250 mp, copper interference may be expected] and the method should not be used without some modification to eliminate copper interferenee. A possibility is the estimation of copper in lead by the ratio of the absorbance a t 270 mp to that a t 289 mp, followed by correction for the interference of this amount of copper. The specific absorbance of copper a t 270 mp is about the same as that of lead, but its maximum is so broad that absorbance a t 289 mp is about 7 times that of lead. Because of this, the interference of copper with the lead determination is greatly reduced when the absorbance difference method (A270 - -4289) is used to determine lead absorbance of a sample. For lead AaT&-o-289is about 50, while for copper, is only about 8.4 (Figure 4). Therefore] even pure copper calculated on the absorbance difference method would amount to only 17% as lead. When the copper to lead ratio is less than 1 to 10, the copper interference will be below 1.7% of the lead present. Copper content of zinc oxide for use in rubber products is below 0.003%. This maximum figure would calculate as only 0.0005% lead. It is helpful to remember that copper chloride in 50y0 hydrochloric acid has a rather strong yellow color if present a t concentrations much above 10 p.p.m. The ether extraction for ferric chloride removal will not affect this copper chloride coloration. Choice of Wave Length Interval. The wave lengths of 250 and 289 mp Tvere arbitrarily placed near the lead 806
ANALYTICAL CHEMISTRY
I
UO 250 270 290 WAVE
310
330
LENGTH
Figure 4. Positive background interference in 50% hydrochloric acid
- 25 p.p.m. of copper e--
20 p.p.m. of lead
-.-Sum of copper and lead absorbance absorption curve minimum to retain good sensitivity. An even narrower wave length interval might be chosen to make the background correction more accurate] because the assumption of a linear background holds more closely as the interval is reduced. The effect of narrowing the wave length interval to reduce interference from elements that absorb rather strongly] but nonselectively, over the 250- to 289-mp interval is shown in Figure 3 for the antimony curve. The absorbance of antimony, vanadium, chromium, tin, and molybdenum, all reported to interfere with lead absorbance a t 270 mp ( 8 ) ,is essentially nonselective (no maximum) over this spectral region. Through the 250- to 289-mp interval, each element exhibits a negative curvature resembling that of iron ( AAi in Figure 3) but varying in slope anddepression from the essentially linear and shallow slope of vanadium to the sharp drop and distinct depression of antimony absorbance. Of the elements mentioned] antimony mould cause the greatest interference] but all would tend to lower apparent lead content if present in appreciable amounts as background interference. Incidentally, this negative absorbance would also be assurance that no lead would be found where none was present, in the absence of positive copper interference. As the vave length interval is narrowed through 39, 30, and 20 mp, the Aa value for lead falls from 50 to 43 to 28, respectively. The drop from an absorptivity of 50 to 28 reduces the sensitivity for lead determination by about 44%. Calculations have shown that, by narrowing the interval to 20 mp the relative interference of all the above elements with negative sloping curves is reduced to about half its value a t the 40-mp interval.
The intensity of absorbance of interfering elements a t 270 mp in 50% acid is comparable t o lead for only iron and copper. The absorptivity lralue of antimony is about 50% that of lead, while the intensity of vanadium] chromium, mercury, bismuth, thallium, and tin absorbances is less than 8% of lead a t 270 mp. Consequently, as long as the concentration of interfering elements (other than copper and iron) is less than the lead present, interference due to negative curvature of background absorbance may be considered negligible. Of course, the closer the background absorbance approaches linearity] or the lower the concentration of nonlinear interference, the more accurate this calculation will be. Though background of possible elemental interference has been shown to be negative, the absorbance of indeterminate materials in commercial products (as dirt in zinc oxide) tends to make background absorbance approach linearity more closely than is indicated in the curve of pure iron and antimony in Figure 3. Therefore] the choice of a wide (39 mp) wave length interval to obtain high sensitivity is believed justified here, when traces of lead in zinc oxide and in the ash of rubber products are being determined. I n other applications the optimum conditions for interference elimination may be calculated. OTHER APPLICATIONS
The rapid micromethod for determining lead should prove satisfactory for many materials other than rubber compounding chemicals. Its possible application to determination of trace amounts of lead in reagent grade sodium, calcium, magnesium, zinc, cobalt, nickel, and manganese compounds appears promising. It should prove sensitive enough t o determine lead in toxicological studies on biological specimens, or for lead compounds in petroleum products. In rubber chemistry] the use of absorptiometric analysis n-ith 50% hydrochloric acid should prove ideal for determining antimony when antimony sulfide is used in rubber products (7). It would also be an excellent method for determining bismuth when the thiocarbamate salt of bismuth (Bismate) is used as an accelerator (8). The elements that might interfere ivith the determination of antimony and bismuth, such as mercury, chromium, vanadium, and thallium, are never used in rubber products. The 50% acid should also prove of value in many applications as a screening reagent in qualitative anal>& for several of the heavier metallic cations, once the more common iron is removed by ether extraction of the acid solution.
ADVANTAGES
The main advantages of the absorptiometric method over conventional electrolytic and gravimetric procedures are its improved simplicity, speed, sensitivity, and accuracy. Lead may be determined down to about 1 p.p.m., the lower limit depending on the amount of background interference. ACKNOWLEDGMENT
Thanks are due to Jack Z. Falcon and Borji Pavlovich, who did much of the
phia, Pa., “ASTM Standards. Chemical Analysis of Rubber Products”, Method D 297-43T, 1952. (2) Brooks, L. A , , Rowley, A. C., Vunder-
(4)India Rubber World, “Compounding Ingredients for Rubber,” Bill Brothers, New York, 1947. ( 5 ) Kress, K. E., ANAL.CHEM.27, 1618 (1955). (6) Kress, K. E., Mees, F. G. S., Zbid., 27, 528 (1955). (7) Memmler, D. K., “The Science of Rubber,” American ed. by R. F. Dunbrook, V. N. Morris, Reinhold, New York, 1934. (8) Merritt, C., Hershenson, H. M., Rogers, L. B., ANAL. CHEM.2 5 , 572 (1953).
(3) Davis, C. C., Blake, J. T., “Chemistry and Technology of Rubber,” Reinhold, Kern York, 1937.
RECEIVED for March 21, 1956. Accepted December 15, 1956.
analytical work in helping to prove the method. Permission of the Firestone Tire and Rubber Co. to publish this paper is gratefully acknowledged. LITERATURE CITED
(1) Am. Soc. Testing Materials, Philadel-
bill News 22, No. 1, p. 10 (1956).
Potentiometric Method for Determination of Carbonyl Sulfide in Petroleum Gases DOUGLAS B. BRUSS, GARRARD E. A. WYLD, and EDWARD D. PETERS Shell Development
Co., Emeryville, Calif.
b Carbonyl sulfide in petroleum gases can b e determined potentiometrically after absorption in alcoholic rnonoethanolamine solution by titrating it with standard silver nitrate solution. Mercaptans and hydrogen sulfide are first removed from the gas stream by absorption in alcoholic caustic solution. Concentration as low as 2 mg. per cubic meter of gas may b e conveniently determined.
D
of sulfur compounds in refinery gas streams is of considerable importance in the petroleum and petrochemical industries. Sulfur constituents such as hydrogen sulfide, mercaptans, carbonyl sulfide, and free sulfur may act, in certain processes, as catalyst poisoning agents and corrosive agents. A potentiometric method has been developed for the determination of carbonyl sulfide in gases in concentrations as low as 2 mg. per cubic meter. Soloveichik (IO) determined carbonyl sulfide in soil gases colorimetrically by absorbing the gas in alcoholic potassium hydroxide and converting the carbonyl sulfide to hydrogen sulfide. dvdeeva ( I ) described a method of determining carbonyl sulfide and carbon disulfide by absorption in selective reagents, oxidation, and measurement as the barium sulfate precipitate. A method of determining carbon disulfide and carbonyl sulfide in gases by adsorption in piperidine-chlorobenzene or alcoholic potassium hydroxide and resolution of the ETERMISATION
proportion colorimetrically or iodometrically was reported by Riesz and Wohlberg (9). Field and Oldach (3,4) used a catalytic method to convert organic sulfur to hydrogen sulfide which is absorbed in caustic, converted to a bismuth sulfide suspension, and read photometrically. They also developed a procedure for the determination and identification of sulfur compounds in gas mixtures based on differences in solubility in an inert solvent ( 7 ) . Haken-ill and Rueck ( 5 ) absorbed carbonyl sulfide in alcoholic potassium hydroxide and determined the amount iodometrically. The use of selective absorbents for analyzing manufactured gas for sulfur compounds is described by MacHattie and McNiven (6). Thiophene, carbonyl sulfide, and carbon disulfide in producer gas have been determined ( 2 ) by treating a gas sample with a piperidine-ethyl alcohol reagent and measuring in the ultraviolet region. Rapoport (8) used a series of absorbents and a combustion technique for the determination of carbon disulfide, carbonyl sulfide, and thiophene. The method reported here consists of the passage of the sample gas through two scrubbers, the first of which removes hydrogen sulfide and mercaptans by absorption in 30% caustic and the second of which absorbs carbonyl sulfide in alcoholic monoethanolamine. The contents of the monoethanolamine scrubber are titrated with silver nitrate solution in acidic alcoholic titration solvent, using the potential difference between a glass reference elec-
trode and a silver sulfide electrode t o indicate the end point. The precipitates from several titrations were filtered from the solutions, washed, and dried in a desiccator. They were then weighed, dissolved in nitric acid, and analyzed for silver content by potentiometric titration with potassium iodide. The results showed a silver content of approximately S6%, which corresponds closely to that of silver sulfide. The titration of the contents of the monoethanolamine scrubber requires 2 moles of silver nitrate per mole of carbonyl sulfide. The following equations are postulated as a possible explanation of the stoichiometry. Equation 4 describes the over-all reaction. COS
+ HOCHZCH~KH~-P HOCHzCH2NH
1
(1)
‘SH HOCHZCHZNH
+ +
HzS
8 + \
SH HOCHzCH2NHZ+ (HOCHzCH2NH)zC=O
(2)
+ +
H2S 2AgNOa+AgzS 2HNOs (3) COS 2HOCHzCHsNHs 2AgNOa+ AgzS (HOCHzCHzNH)zC=O 2”Os (4)
+
+
+
APPARATUS A N D REAGENTS
Absorbers.
Gas
washing
VOL. 29, NO. 5, M A Y 1957
bottles,
807
