ANALYTICAL CHEMISTRY
1328 Table IV.
Effect of Overheating on Reactivity of Bitumen-Rubber Mixture 95% bitumen 5 0 / 6 0 a n d 5% Pulvatex) Rubber Content, % S in Calcd. Insoluble, Insoluble, on % % Found Average CsHs 4.5 3.9 28.9-28.6 4.5-4.5 7.4-7.4 3.0 3.9 21.4-21.9 3.0-3.0 6.6-6.5 0.04 3.9 0.7-1.3 0.02-0.05 1.7-1.7
(Mixture. Preheating of Sample Temp., Time, O C. Hr. 230 1 260 2.5 295 2
impunity so far as this test is concerned, a few specimens were heated to a high temperature. One sample of bitumen-Pulvatex was heated to 230°, 260°, and 295' C.; a t 295' C. the bitumen hardened to such an extent that it would have been unusable for road making. The results of these experiments (see Table IV) show that after 2 hours of heating a t 295" C.. nearlv all traces Of the rubber vanished. rubber is knoTvn"to undergo cyclization and degradation at this temperature. Obviously then, these degradation products do not form insoluble rubber derivatives with sulfur.
Error,
%
2i: -99
Since synthetic rubber changes thermally in a different way, it may be that this, on the contrary, would be identifiable after being heated. This implies that the negative result of a rubber estimation in bitumen is conclusive only if the mixture has not been overheated in the course of the procedure. LITERATURE CITED
(1) , , Decker. H. C. J.. de. and Nijveld. H. A. W. World Petroleum Congr., Proc., 3rd Congr., Hague 1951, Section VII. (2) Heurn, F. C., van, and Begheyn, 11.A., I n d i a Rubber J., 81, 847
(1931). (3) Numajiri, S., Rubber Chem. and Technol., 10, 792 (1937). (4) Salomon, G., Chem. Weekblad, 48, 292 (1962). (5) Salomon, G., chapter on rubber derivatives in R. Houwink, "Chemie und Technolonie der Kunstoffe." 3rd ed.. Vol. 11. Leipzig C 1, Bkademische Verlagsgesellschaft Geest' & Portig K.-G., 1954. (6) Stevens, H. P., and Stevens, W. H., J . SOC.Chem. I n d . ( L o n d o n ) . 48, 55T (1929). R~~~~~~~ for review M~~ 9, 1953. ~ c c e p t e dApril 8, 1954. tion 223 of the Rubber-Stichting. Delft, Holland.
Communics-
Determination of Germanium in Coal, Coal Ash, and Flue Dust WILLIAM J. FREDERICK, JAXON A. WHITE, and
H. E. BIBER
Research and Development Laboratory, United States Steel Corp., Pittsburgh, Pa.
This investigation was undertaken to develop a rapid and accurate method for the determination of small amounts of germanium such as exist in coal and waste products. A chemical method is described in which the germanium is converted to CazGe04, separated as germanium tetrachloride by distillation, and determined as cinchonine germanomolybdate. A spectrochemical method for coal and coal ash is also described, in which bismuth is added as an internal standard and a direct current arc is used as the means of excitation. For small amounts of germanium, the spectrochemical method is more sensitive and also less time-consuming than the chemical method after synthetic standards have been prepared and working curves established.
treated with hydrochloric acid and the germanium was separated from interfering elements by distillation. The germanium was then determined in the distillate. In considering methods for determining the amount of germanium in the distillate, one of the methods recommended by Davies and Morgan ( 1 ) was investigated. It consisted of precipitating the germanium as cinchonine germanomolybdate and weighing it as such. In the present investigation, it was found necessary to change this procedure somewhat because the distillate amounted t o more than 40 ml. The nitric acid concentration was kept low to reduce the possibility of precipitating molybdic acid, and the solution was kept ice cold because a l o m r blank was obtained than when working a t room temperature. ANALYTICAL PROCEDURE
I
N CONSECTION with a survey of coal deposits and byproducts of coal combustion, a search was made for a method of determining germanium in these materials. It was found that a comprehensive study of the chemical reactions of both organic and inorganic compounds of germanium had been made by Johnson (3, 4 ) and Krause and Johnson ( 5 ) in this country; Davies and hlorgan ( 1 ) in Great Britain investigated a number of analytical procedures. Headlee and Hunter (8)and Rusanov and Bodunkov (6) described procedures for spectrographic determination of germanium. The present paper evaluates some of these methods and describes a new procedure for treating material so as to prevent volatilization of germanium compounds and to render them soluble in hydrochloric acid. CHEMICAL METHODS
The preliminary treatment for coal consisted of oxidizing the coal with concentrated sulfuric and nitric acids and also of igniting similar samples with calcium carbonate a t 1000" C. Samples of coal ash and of flue dust were fused with sodium carbonate and duplicates of these samples were ignited with calcium carbonate a t 1000" to 1050" C. In all cases, the ignited samples were
Mix 5 grams of the 100-mesh coal sample with fine reagent grade calcium carbonate and transfer the mixture to a porcelain dish or t o a wide-form crucible. If the amount of germanium in the sample is more than 3 mg., use a sample smaller than 5 grams. Above 3 mg. of germanium, the precipitate becomes bulky and difficult to filter; this increases the time required t o make the determination. Place the dish containing the mixture in a muffle furnace. Hold the furnace a t 480" C. until all the volatile matter has been removed from the coal, then heat t o 1000" C. and hold a t this temperature for 1 hour. Coal ash samples can be placed directly in the muffle a t 1000" C. Flue dust should be heated at 1050" C. for 1 hour. Remove the sample from the muffle, cool, and transfer the ignited residue to a distilling flask. Add 100 ml. of water and 2 ml. of potassium chromate solution. Place a 250-ml. tall-form beaker containing 100 ml. of ice-cold water under the condenser so that the tip of the condenser extends to within '/4 inch of the bottom of the beaker. Keep the water in the beaker ice cold by surrounding it with ice. Sext, add 200 ml. of concentrated hydrochloric acid to the distilling flask, bubble carbon dioxide through the acid, and start the distillation by applying heat. Continue the distillation until 50 ml. has distilled over. Transfer the distillate to a 600-ml. beaker and adjust the volume of the solution t o 250 ml. with distilled water. Cool in ice water and neutralize with ammonium hydroxide (methyl red indicator). To prevent local overheating and consequent loss of germanium, stir the solution continuously while the ammonium hydroxide is being added. Conduct the neutralization by
V O L U M E 2 6 , NO. 8, A U G U S T 1 9 5 4
1329
adding ammonium hydroxide until a distinct yellow is obtained, and then add 2 N nitric acid until the first pink color appears. Dilute the solution to 400 ml. with distilled water and cool to 10" to 15' C. in ice water. A4dd20 ml. of 25% ammonium nitrate solution, 16 ml. of 2% ammonium molybdate solution and 30 ml. of 2 N nitric acid, and stir until thoroughly mixed. Add 9 ml. of cinchonine solution and stir. Let stand for 3 hours in ice water, filter through a weighed, fine-texture, Selas crucible, and wash 5 times with ammonium nitrate solution. D r y the crucible and precipitate in an oven a t 160' C. for 2 hours, cool in a desiccator, and weigh. The increase in weight represents cinchonine germanomolybdate. Two blanks should be run with each lot of samples and the germanium found in the sample should be corrected by the amount of the average blank. Calculation: (Wt. of ppt. - wt. of blank) X 2.955 = Wt. of sample
B
ADO
% Ge
with distilled water. Each milliliter of solution 1 contained 0.069 mg. of germanium and each milliliter of solution 2 contained 0,091 mg. of germanium. The analysisof aliquots of these samples checked whether they were distilled through the apparatus in Figure 1 or not. A series of synthetic samples containing known amounts of germanium was analyzed by the procedure described. The calculations were based on the assumption that the cinchonine germanomolybdate molecule contained four cinchonine radicals. According to this assumption, the germanium content of the compound is 2.385% (factor 0.02385). The results obtained with this factor, Table I, were low by approximately 20%.
Table I. Use of Factors 0.02385 and 0.02955 for Calculating Germanium Content Ge Added, hlg.
nci
0.69 0.91 0.91 0.91 0.91 0.91 1.04 1.37
Ge Found hfg. (Factdr 0 02385) 0.52 0.78 0.72 0.70 0.75 0.74 0.83 1.14
Deviation, .Mg. -0.17 -0.13 -0.19 -0.21 -0.16 -0.17 -0.21 -0.23
Ge Found, Mg. (Factor 0.02955) 0.65 0.95 0.89 0.87 0.92 0.87 1.03 1.41
Deviation J4g. -0.04 +0.04 -0.02 -0.04 +0.01 -0.04
-0.01
+0.04
Table 11. Theoretical Composition and Actual Analysis of Compound Theoretical Composition, % C Mo Ge (CloH2*0Nz)4HP(Ge~~o,*O~)2 9 . 9 7 37.66 2.385 ( C ~ ~ H Z Z O N ~ ) Z H ~ ( G1 ~ 8 .J 5 8' ~ O4 ~ 6 .~ 8 7~ ~ ~2 ) ,955 Actual Analysis of Ppt., %, 20.85 44.02 ...
Figure
1.
Apparatus for Distilling Tetrachloride
Germanium
Special Reagents Required. AMMONIUM NITRATE,257,. Dissolve 250 grams of ammonium nitrate in water and dilute to 1000 ml. AMMONIUM MOLYBDATE, 2%. Dissolve 20 grams of (SH,), RIoi024.H20 in 1000 ml. of distilled water. KITRICACID,2AV. Dilute 127 ml. of nitric acid to 1000 ml. with distilled water. CISCHOSINE,2.5%. Dissolve 25 grams of cinchonine in 1000 ml. of 2.5N nitric acid. XITRICACID,2.5,V. Dilute 159 ml. of nitric acid to 1000 ml. with distilled water. AMMONIUM NITRATESOLUTION.Dissolve 25 grams of ammonium nitrate in 1000 ml. of water that contains 50 ml. of 2 N nitric acid. POTASSIUM CHROMATE SOLUTION.Dissolve 50 grams of potassium chromate in 100 ml. of distilled water. EXPERIMENTAL WORK
The accuracy and precision of the cinchonine germanomolybdate method was determined with solutions containing known amounts of germanium. Solutions were prepared from two separate lots of germanium(1V) oxide by dissolving a weighed amount of each lot in sodium hydroxide and diluting to 1000 ml.
~
~ Wt. 3046 2457
Since this factor gave low results, the germanium content was recalculated, assuming two cinchonine radicals in the molecule, This compound has a germanium content of 2.955% (factor 0.02965). The results obtained with this factor (Table I ) were correct within the limits of experimental error. Table I1 shows the theoretical percentages of carbon, molybdenum, and germanium for the formulas (ClaH220N2)4Ha(Ge Mo1204~)and ( C I ~ H ~ ~ ~GeMolt040). N ~ ) ~ H ~Since ( the factor derived from the latter formula gave correct results, it was decided to make a chemical analysis of the cinchonine germanomolybdate precipitated under the conditions outlined in the analytical procedure. The carbon and molybdenum contents were found to be 20.85 and 44.02%, respectively (Table 11). 9 higher carbon content and a lower molybdenum content were obtained than indicated by the formula. Until further work is done, such as determining the molecular weight of the compound, the factor 0.02955 must be considered empirical. SPECTROCHEMICAL METHODS
At the time this investigation was undertaken, only two quantitative spectrochemical methods for germanium in coal or coal ash were found in the literature One was a total energy method developed by Headlee and Hunter ( 2 ) and the other was an internal standard method for coal ash devised by Rusanov and Bodunkov (6). The methods described here are similar to that of Rusanov and Bodunkov (6) with respect to the internal standard element but the techniques of the methods are different. EXPERIMENTAL WORK
Because the composition of coal varies considerably, it was necessary to add as an internal standard an element that does not occur in coal and that has physical characteristics similar to
l
~
1330
ANALYTICAL CHEMISTRY
those of germanium. Bismuth was found t o be suitable for this purpose. The standards used for thecalibration of the methods were prepared synthetically. The coal standards were made by mixing a weighed amount of C . P . germanium(1V) oxide with a weighed amount of germanium-free coal. B portion of this first standard was then mixed with a portion of the coal to produce a second standard. This dilution procedure was repeated to obtain five standards covering the range 0.0005 to 0.0065% germanium. The coal ash standards were prepared in a similar manner, the matrix being a sample of germanium-free coal ash. Nine standayds were made covering the range of 0.0013 to 0.67% germanium. The procedure developed for the analysis of coal differed slightly from that for the analysis of coal ash. T o every 7 parts of a coal sample, 1 part of bismuth(II1) oxide was added. T o every 3 parts of a coal ash sample, 4 parts of pure graphite powder and 1 part of bismuth(II1) oxide were added. Each combination was thoroughly mixed and then packed into a sample electrode, which n a s a high purity graphite rod, 2 inches long and l / 4 inch in diameter, having a center post and a 1 mm.-deep crater. The samples were excited by a rectified direct current arc discharge; the counter electrodes being high purity graphite rods inch in diameter, having 120" conical tips. The excitation conditions are shown in Table 111. (The use of counter electrodes of '/*-inch diameter gave less reproducible results, despite the smoother burning arc produced.) A grating spectrograph having a first-ordrr plate factor of about 3.4.4 per mm. was used for these analyses. Each sample was analyzed in triplicate. RESULTS
The results of the analysis of four coal samples and eleven coal ash samples are shown in Tables IV and V. The reproducibility of the methods \vas investigatd by using one coal sample and four coal ash samples. The results are shown in Table 1-1. The reproducibility of thcx rcsulti: for germanium in coal ash is comparable to that gcnrrally obtained with direct current arc excitation at similar concentration levels. The reproducibility of t h e rwultq for germanium in c o d is poor.
Table 111, Excitation Conditions for Spectrochemical Determination of Germanium Discharge Conditions Capacitance, microfarads Inductance, microhenlies Resistance, ohms Output potential, rolts Current, amperes Time, seconds Sample electrode -4nalytical gap, m m .
Coal Ash Analysis Iin
400 15 340 12 90 Lower and uositive 3
Coal Analysis 60 400
15 340 12 90 Lower and negatives 3
Spectrograph Conditions Coal analysis (0.0005 to 16-Micron entrance slot 25$& Incident light transmitted to entrance slit 0.008% Ge) Firet-order spectrum 2170-3900 A . Coal ash analysis (0.001 IO-IIicron entrance slit 2 5 % Incident light transmitted to entrance slit t o 0.05% Ge) First-order spectrum 2170 t o 3900 -1. Coal ash analysis (0.03 14-1Iicron entrance slit 100% Incident light transmitted t o entrance slit to 0.4% Ge) Second-order spectrum 2525 t o 3350 A . a
Reversed polarity is used here t o increase line-to-background ratios
Table IV.
Determination of Germanium in Coal
(Comparison of Kjeldahl, calcium carbonate, and spectrochemical methods) Ge, 92 Sample Kjeldahl Calcium carbonate Spectrochemical NO.* method method method 0.0037 0,004 o.oni 6 0 . on5 0.0031 o.om 7 0,004 0,0040
