I N D U S T R IA L A N D E N GIN E ER IN G C H E M I ST R Y
110
series of four or more test tubes containing the same aliquot of standard or unknown. To obtain the best results, a 5-cc. aliquot should contain from 3 to 157 ( l y = 0.001 mg.) of copper. Too dilute or too concentrated solutions give colors which make it difficult to determine the end point. A preliminary titration will show the approximate amount of copper in the unknown after the titer has been determined for the chromotropic reagent, which is added to the aliquots in gradations of 0.1 cc. At one end of the series the color in the test tubes should he distinctly purple, a t the other end distinctly brown. The first aliquot which shows a brown tinge when examined in the colorimeter is taken as the end point. TABLE r. RECOVERY OF ADDEDCOPPER SAMPLE
COPPERIN SAMPLE P.p.m.
13 (sirup) 9 (sirup) 307 peas) 308 [pea@) 309 (peas) 310 (peas) 104 (sirup) 105 (sirup)
4.38 8.34 73.30 73.00 71.50 81.20 2.90 3.50
COPPER
ADDED P.p.m. 1.27 1.27 2.30 4.60 11.60 23.00 37.70 50.90
NOTE. After this manuscript was submitted for publication, two papers of interest with regard to copper methods were published (6, 9). Conn and collaborators (6) make the following statement with regard to the four copper methods which they employed: “Summarizing experience with these four methods, the chromotropic acid method can be eliminated because of the difficulty in distinguishing the end point. The Biazzo method is not sufficiently sensitive. The possibility of interference due to turbidity makes the xanthate method undesirable. The carbamate method appears to be by far the most desirable, since it is the most sensitive, and the disturbing factor of turbidity is not encountered.” The authors’ results as well as those of Ansbacher (l,,??, 4, 6) and Ohlson (9) have demonstrated that the chromotropic acid method is at least as sensitive as the carbamate method of Conn and collaborators, and apparently more reliable, judging by a comparison of the data on recovery of added copper for the two methods.
TOTAL ADDED COPPER COPPER COPPER FOUNDRECOVERED RECOVERED P.p.m. P.8.m. lol %,6 5.67 1.29 9.66 1.32 103.9 75.40 2.10 91.3 78.00 5.00 108.7 82.60 11.00 ”’’ 104.50 23.30 101.3 42.90 40.00 106.1 54.10 50.60 99.4
The authors were unable to distinguish differences in color where less than a drop of the chromotro~icreagent was used, If the reagent was to ‘”7 Of copper per One drop or 0.05 cc. would be equivalent to 0 . 3 3 5 ~of copper. Ansbacher found the reagent to be accurate to 0 . 5 of ~ copper. Table 1 shows the recovery of copper added to sorghum sirup and peas. ”*,
VOl. 7, No. 2
LITERATURE CITED
(1) Ansbaoher, “&tude de Chimiothbrapie de la Tuberculose,” pp. 26-35, Geneva, Imprimerie Kundig, 1933. (2) Ansbacher, Remington, and Culp, IND.ENG.CHEM.,Anal. Ed., 3,314 (1931). (3) Callan and Henderson, Analyst, 54,650 (1929). (4) Cherbuliez and Ansbacher, Arch. path. Anat. (Virchow’s),276, 365 (1930). (5) Cherbuliei and Ansbacher, H e h . Chirn. Acta, 13, 187 (1930). (6) Conn, L. W.,Johnson, A. H., Trebler, H. A., and Karpenko, V., IND.ENG.CHEM.,Anal. Ed., 7, 15 (1935). (7) Elvehjem and Lindow, J. Bioi. Chern., 435 (1929). ( 8 ) Moseley, H. W., Rohwer, A. G., and Moore, M. C., Science, 79, 507-8 (1934). (9) Ohlson and Daum, J. Nutrition, 9, 76 (1935). (10) SuPPlee and Bellis, J. Dairy SCi.9 5,455(1922). RECEIVDD December 10,1934.
Adsorption of Organic Liquids by Cellulose Products J. WIERTELAKAND I. GARBACZ~WNA, Institute of General Chemistry, University of Poznan, Poland
A
TTENTION has recently been called by Mease (6) to the fact that other liquids than water may be held tenaciously by fibrous materials. The same conclusion has been arrived a t independently by Booth (1) and Wiertelak (8) in connection with standard analyses performed on cellulose and wood products. In case the samples analyzed contained but minute a’mounts of extractives, they were found to weigh more after extraction with an organic solvent, such as a benzene-alcohol mixture, and subsequent drying to constant weight than before the extraction, although the solvent undoubtedly contained some extracted substances. This study was therefore undertaken in order to extend the scant information offered by Mease as well as to ascertain to what extent the adsorption of organic liquids may interfere with standard analyses of wood and cellulose products, as they are usually performed in technical laboratories.
EXPERIMENTAL *
Adsorption experiments have been performed on the following cellulosic substances: commercial adsorbent cotton; Cross and Bevan cellulose prepared from white birch, probably Betula papyrijera, Marshall; commercial viscose rayon and commercial nitrate rayon, both supplied by the Rayon Works a t Tomasz6w, Poland; and commercial cuprammonium rayon, supplied by the Bemberg Aktien Gesellschaft a t Barmen, Germany. The chemical composition of these
cellulose products is described in detail in a recent paper by Wiertelak (8). The liquids used for adsorption were: commercial 95 per cent ethyl alcohol; methyl alcohol Kahlbaum; n-butyl alcohol Kahlbaum, gereinigt; n-propyl alcohol, Merck; commercial diethyl ether; benzene, Merck or Kahlbaum, purum; commercial gasoline, boiling range 100’ to 150’ C.; pyridine, Polish Coke Association, purissimum; a benzenealcohol mixture, 1 to 2; and a benzene-alcohol mixture, 2 to 1. The two benzene-alcohol mixtures were prepared from commercial 95 per cent ethyl alcohol and Merck or Kahlbaum benzene. All liquids were redistilled repeatedly before use. Most of the above cellulose specimens were used in air-dry condition, their moisture content being shown by a separate determination to amount to from 7 to 10 per cent on the basis of the oven-dry sample. Approximately 1 gram of the airdry sample, accurately weighed, was placed in an Erlenmeyer flask and kept in a thermostat a t 25’ C.for 15 minutes in contact with 25 cc. of the liquid to be tested. Then it was filtered through an alundum crucible, dried in air in an electrical drying oven a t 105’ C. to constant weight, and weighed. In one series a sample of cotton was oven-dried a t first, and then treated with the adsorption liquid. The rest of the procedure was the same as above. The percentage increase in the oven-dry weight due to adsorption, calculated on the basis of the oven-dry weight of the fibers before treat-
March 15, 1935
ANALYTICAL EDITION
ment represents the adsorption value. The results are recorded in Tables I and 11. As is evident from Table I, if cotton is dried to the oven-dry state before immersion in the liquid, it adsorbs a higher amount of alcohol than in the air-dry condition; similarly, regenerated celluloses and Cross and Bevan cellulose obtained from wood adsorb more alcohol than cotton. Evidently the adsorption of alcohol by various cellulose products is a function of their state of degradation. Since it is known (4, 5 ) that prolonged heating of cellulose products at low temperature causes their chemical modification, it may well be that oven-dried cotton, being already slightly degraded, adsorbs a higher amount of alcohol. TABLEI. ADSORPTIONOF 95 PER CENTETHYLALCOHOL BY CELLULOSE PRODUCTS (Results are expressed in percentage of oven-dry material) EXPT.I EXPT.I1 EXPT.I11 AVERAQE Cotton 1.59 1.64 1.57 1.60 Cotton, oven-dry 1.94 1.93 1.93 1.93 Cross and Bevan cellulose 1.83 2.12 1.98 1.98 Viscose rayon 2.13 2.10 2.12 Nitrate rayop 1.74 1.64 1:67 1.68 Cuprammoniumrayon 2.15 2.14 2.11 2.13
TABLE11. ADSORPTION OF ORGANIC LIQUIDS BY COTTON AND NITRATE RAYON (Results are expressed in -COTTONExpt. Expt. I I1 Methylalcohol 0 . 8 0 0 80 Ethyl alcohol 1 . 5 9 1:64 Propyl alcohol 2 65 2 64 n-Butyl alcohol 1142 1:84 Diethyl ether 0.06 0 Benzene 0.21 0.28 Gasoline 0 0 Pvridine Binsene-ethyl doohol,2tol
..
..
2.22
2.37
percentage of oven-dry material) ~ N I T R A TRAYONB Expt. Expt. Expt. I I1 I11 Av. Av. 0.80 0.94 0.96 0.85 0.92 1:57 1 . 6 0 1 . 7 4 1 . 6 4 1 . 0 7 1.08
Expt. I11
2.65
.. ..,. ..
0 0 0.22 0.24 0 0
..
..
1.19 0.60 0.33 0.01 0.15 0.56
2.33
2.31
2 . 3 3 2.81
2.75 2.63
3.23
2.98
1189 1 . 7 2
Benzene-ethyl alcoho1,ltoZ 1 . 2 3 1.26 1 . 2 9 1 . 2 6 4 Loss in weight.
TABLE111. BENZENE-ALCOHOL, 2 TO 1, EXTRACTIVES IN SAWDUST (Expressed in percentage of oven-dry weight of wood) TRUEV A L U ~ . Low VALUE, BY WEIGHINQ BY WE~IQHINQ EXTRACTIVES S A W D U ~ T DIFFERENCE^ SPECIES
3.21
The alcohol adsorbed on both cotton and nitrate rayon may be removed by washing the samples with water. The samples, when dried carefully, return to their original ovendry weight only if air-dry samples have been used in the adsorption experiments. When oven-dry cotton has been used a stronger washing, preferably with warm water, is needed. The same removal of alcohol from the cellulose samples can be attained by treating them with diethyl ether and drying. The samples assume their original oven-dry weight. Attempts were made to isolate the adsorbed alcohol by treating 20 grams of cotton with alcohol, drying, and subsequently washing with 500 cc. of cold water. The washings were distilled and in the 250 cc. of collected distillate alcohol was determined by Rozman's titration method with standard permanganate (7). :In two determinations an almost quantitative yield was obtained. Alcohol adsorbed on 20 grama of cotton: By difference in weight By titration with potassium permanganate
applied, it appeared of value to determine in what way the adsorption of the mixture by the fibers influences the final results. Therefore extraction experiments with a benzene-alcohol mixture, 2 to 1, were performed on sawdust, which was a t hand in the laboratory: loblolly pine sapwood, 40 to 60 mesh sample; white fir, heartwood, 60 to 80 mesh; eastern white pine, 40 to 60 mesh; and white birch, 60 to 80 mesh, The extraction was performed in a Soxhlet apparatus, and the extracted sawdust in the alundum crucible was weighed after drying, as prescribed in the standard methods compiled by Bray and Wiertelak (9,s).On the other hand, the extraction liquid in the Soxhlet flask was evaporated, and the amount of extractives determined by weighing. On the assumption that drying the extractives did not induce any chemical change, this represented the true value for the extractable matter. If the benzene-alcohol mixture was adsorbed in the sawdust analyzed, the value for the extractable matter, obtained by the difference in weight of the oven-dry sawdust before and after extraction, would be smaller than the above true value-via., differing by the weight of the adsorbed benzene-alcohol mixture. The results, representing average values of a t least four extractions, are compiled in Table 111. The results in Table I11 prove that the benzene-alcohol mixture is indeed adsorbed by the sawdust, since all true values for the extractive content of sawdust are higher than those obtained by weighing the sawdust before and after extraction. The difference is in some instances as high as 20 per cent.
1.19 1.19 0.44 0.52 0.33 0.33 0.020 0 0 0.07 0 . 5 6 0149 0 . 5 4
3.42
Gram
%
0.21 0.2032 0.1995 0.2014
1.12
....
111
$t/l,.,I@pi?P Eastern,white pine
%
%
4.97 2.18 11.43 1.99
19 8.2 17.1
2.09
12.45 2.40
White birch 0
% 5.02
1
Based on the true value.
SUMMARY
It is shown that various cellulose preparations adsorb alcohols, pyridine, and especially benzene-alcohol mixtures. The adsorbed liquid is not removed from the fibers even by prolonged heating a t 105" C., but may be removed by washing with ether or water. Ether, benzene, and gasoline are not adsorbed. The adsorption causes a marked lowering of the value for the extractive content of sawdust, if that value is determined by weighing the sawdust. ACKNOWLEDGMENT The authors express their gratitude to S. Glixelli, director of the Institute of General Chemistry, for reviewing the paper and for permission to publish these results.
LITERATURE CITED
(1) Booth, J., British Forest Products Research Laboratory at Princes Risborough, Bucks, private communication, July, 1933. (2) Bray, M. W., Paper Trade J., 87, 59 (1928). EXTRACTION OF WOODSAMPLES WITH BENZENE-ALCOHOL (3) Bray, M. W., and Wiertelak, J., Polish Agr. Forestal Ann., 25, MIXTURE 203 (1931). L. F . , and Harris, E , E., IND. ENQ.CHEM.,24, 873 The above results, especially those with benzene-alcohol (4) Hawley, (1932). mixtures, indicate that cellulose products adsorb a quantity (5) Hawley, L. F., and Wiertelak, J., Ibid., 23, 184 (1931). of liquid, which even by prolonged heating cannot be removed (6) Mease, R.T., Ibid., Anal. Ed., 5, 317 (1933). from the samples. Since in standard analyses on wood and (7) Rozman, B . Y., Zhur. Prikladnoi Khim., 2, 191 (1929). (8) Wiertelak, J., Prremysl Chem., 18, 648 (1934). Av.
1.07
other cellulose products extraction with some organic solvent, especially with a benzene-alcohol mixture, is usually
RECEIYED November 14, 1934.
