MARCH 15. 1939
ANALYTICAL EDITION
Summary The turbidimetric method for determining carbon dioxide is simple and rapid. Only the sample and final suspension need to be measured accurately. The method can estimate carbon dioxide in amounts between 0.5 and 10 mg., but a smaller quantity than 0.5 mg. may also be determined. The average probable error is *2.6 per cent. Most of this error is associated with the precision of the turbidimeter which is about b 2 . 0 per cent.
Literature Cited
153
(3) Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” p. 623, New York, John Wiley & Sons, 1929. (4) Partridge, E. P., and Schroeder, W. C., IXD.ENQ.CHEM.,Anal. Ed., 4, 271-8 (1932). (5) Raymond, A. L., and Weingarden, H. M., J. Biol. Chem., 74, 189202 (1927). (6) Scott, W. W., “Standard Methods of Chemical Analysis,” p. 121 New York, D. Van Nostrand Co., 1925. (7) Shaw, J. A., J. IND. ENG.CHEM.,13, 1151-2 (1921). (8) Sheen, R. T., Kahler, H. L., and Ross, E. M., Ibid., Anal. Ed., 7, 262-5 (1935). (9) Snell, F. D., and Snell, C. T., “Colorimetric Analysis,” pp. 95, 121, New York, D. Van Nostrand Co., 1936.
(1) Cain, J. R., and Maxwell, L. C., J. IND. ENG.CHEM.,11, 852-60
(1919). (2) Greenberg, D. M., Moberg, E. G., and Allen, E. C., Ibid., Anal. Ed., 4,309-13 (1932).
RECEIVEDNovember 17, 1938. Published by permission of the Director of the Bureau of Mines, United States Department of the Interior. (Not subject t o copyright.)
Effect of Various Pre-extractions on the Lignin Determination of Wood ELWIN E. HARRIS AND R. L. MITCHELL, Forest Products Laboratory, Madison, Wis.
T
H E present methods for the determination of lignin in wood are subject to criticism because they cannot be applied alike to all types of material. The sulfuric acid method, which is favored by most analysts because it is simpler to perform and gives, under controlled conditions, more uniform results than other methods, is subject to the same criticism: Certain carbohydrates are converted into a water-insoluble material, which appears in the lignin residue if allowed to stand in contact with the sulfuric acid too long a time or a t too high a temperature (6, 7, 10, 11). To overcome this objection and to make control of temperature unnecessary, several investigators (6, 8) have suggested that these unstable carbohydrates be removed by a preliminary extrtiction with hot dilute acid. They assumed that the easily hydrolyzable carbohydrates as determined by Hawley and Fleck (4) were the same as those which were responsible for the formation of the insoluble ligninlike residue. Campbell and Bamford (1) contend that preliminary treatment
causes a polymerization of some carbohydrate substance with an increase in lignin yield. Work on this aspect of the analysis of lignin has been in progress a t the Forest Products Laboratory a t Madison, Wis., for several years. Preliminary extraction with organic solvents, cold water, hot water, dilute sodium hydroxide, barium hydroxide, sodium carbonate, sodium sulfite, and dilute hydrochloric, dilute sulfuric, and 3 per cent oxalic acid have been tried and, in addition, various concentrations of dilute sulfuric acid under pressure. The work with sodium hydroxide on sugar maple sawdust was reported by Harris (S), various treatments with hot and cold water and organic solvents by Ritter and Barbour (Q), and some of the work showing the effect of 3 per cent sulfuric acid on sugar maple sawdust by Cohen and Harris ( 2 ) . This report contains, in table form, the results of further work with sulfuric acid and of various other pretreatments on wood.
OF ACIDPRE-EXTRACTION ON SPRUCE AND MAPLEWOOD TABLE I. EFFECT
Wood Treatment
Total Time of Treatment Hours
*.3
1. 2. 3. 4. 5. 6.
6 9 12 18 3 6 9
12 18
..4
1 4 6
+ + ++ +
No. 12 3 hours’ boiling with No. 17 3 hours’ bojljng wjth No. 18 3 hours‘ boiling w t h No. 19 3 hours’ boiling wjth No. 20 6 hours’ boiling with a Percentages are oalculated from unextracted oven-dried wood. b Percentages based on isolated lignin. C Loss of lignin (due t o treatment) Original lignin (after alcohol-benzene extraction) ’
17. 18. 19. 20. 21.
3 6 9 12 18
AlcoholBenzeneWater Extractives“
Loss by Treatment“
% 4.3
......
.. .. .... .. .. 4:i7 .. .. .. *. .. .. .... ..
Lignin Contentm
Me0 in Woods
%
%
%
%
24: 0 28.0 30.2 33.9 37.0
27.47 26.4 25.5 24.9 23.7 22.9
6:i6 6.25 5.87 5.34 4.95
16.6 16.5 16.4 15.1 15.0 14.8
16.65 24.2 28.4 28.7 29.0 4.07 23.0 26.7 36.4 34.8
26.1 24.0 23.7 23.7 23.7 22.75 19.85 19.10 17.6 17.75
25.5 31.5 32.6 36.0 38.0
20.3 18.5 18.3 17.4 16.9
.... ..
ye0 in Ligninb
Calculated Losg of Lignin%
%
...
3.65 6.95 9.5 13.9 16.8
.... ..
16.5 16.4 16.4 16.0 15.9 20.4 20.5 20.3 20.2 20.0
4.7 11.5 13.9 12.2 16.0 22.5 22.0
6.60 5.95 5.48 5.40 5.40
20.3 20.4 20.3 20.5 20.0
10.5 18.5 19.5 23.5 25.5
...
VOL. 11, NO. 3
INDUSTRIAL AND ENGINEERING CHEMISTRY
154
TABLE11. EFFECT OF PRE-EXTRACTION WITH ACID,SALTS, AND ALKALION LIGNINDETERMINATION OF MAPLEWOOD (On boiling water bath)
Lignin
Loss b y Solvent Loss by Extrac- Treat-
Treatment
How8
... 4
Alcohol-benzene extraction
% 4.07 ,
I
:: ....
73
..
3
merit
Based on Original Wood
Me0 In Lignin
4.07 2.4
i:::
22.7 22.5
20.4 20.5
;;:: i:::
10.8 25.9 24.3
21.2 21.2 19.4
20.4 20.5 20.3
%
%
%
into a large volume of water, which caused the material to precipitate again. This material contained 16.8 per cent of methoxyl and had properties that indicated that it was lignin. Maple wood lost 22.5 per cent of the lignin when heated 4 hours with 3 per cent sulfuric acid. When the heating was continued for 6 hours, some of the lignin was again converted into an insoluble compound, since the loss is slightly less than with a 4-hour treatment. Oxalic acid, during six 3hour treatments, removed 25.5 per cent of the lignin, which was more than that removed by the action of sulfuric acid (6). Material precipitated from the sulfuric and oxalic acid solutions on standing or on increasing the acid concentration
TABLE 111. EFFECT OF HYDROLYSIS ON LIGNINAND CELLULOSE VALUES IN MAPLE WOOD 120 Pounds' Pressure, 30 Minutes
Treatment
EX; penment, No,
Meth-
Hydrolyzed wood
Cellulose"
Lignina
Ligninb
in lienin
%
%
%
%
%
OXY1
135 Pounds' Pressure, 30 Minutes
EX;
penment No.
Hydrolyzed wood
%
sulfurio acid 485 63.70 64.95 6 3 . 4 32.5 21.1 20.6 473 hydfolysis 486 61.83 62.30 61.9 36.2 2 2 . 5 20.3 474 2% acid 487 59.81 60.10 59.6 37.6 22.6 20.3 475 3% acid 2 3 . 4 5 7 . 1 3 9 . 8 488 2 0 . 0 5 8.74 5 8 . 9 5 476 4% acid 489 56.80 56.90 5 5 . 6 4 1 . 8 2 3 . 8 1 9 . 7 477 5% acid 490 5 5.72 . . . 6% acid 491 5 4 : i 8 5210 4 6 : 6 2 5 : 4 1 8 : 4 54.00 484 7% acid 27.95 2 2 . 9 2 0 . 7 8 2 . 5 0 6 2 . 5 500 Water 22.9 20.7 68.37 6 3 . 0 33.6 ... 499 3% oxalic acid 0 Per cent cellulose and lignin yields based on oven-dry residue after hydrolysis. b Per cent lignin based on weight of original wood. 1%
...
.,I
...
Experimental Procedure Sawdust obtained from sawing selected air-dried wood free from bark, knots, and compression wood, was cut to pass a 40-mesh screen in a Wiley mill and extracted successively with cold water, hot 95 per cent alcohol, hot alcohol-benzene mixture, followed by alcohol and then water, and finally dried in the air to 5.5 per cent moisture content. Table I gives the data for the effect of successive treatments on a boiling water bath with 3 per cent sulfuric and oxalic acid, and also other indiyidual treatments with sulfuric acid on spruce and maple sawdust. Table I1 shows the effect of various other materials a t the temperature of the water bath. The furfural determination was omitted because, as shown by Cohen and Harris (2), practically all the furfural-forming material was destroyed by contact with acid. Five hundred cubic centimeters of the acid solution were used for the extraction of each 2 grams of wood. All extractions were carried out in groups of six for each series of extractions and the average of these was taken. The acid extract was filtered from the sawdust while hot, since cooling caused a precipitation of some of the dissolved material. Table I11 gives the lignin and cellulose analytical values obtained upon maple wood after treatment a t 8.436, 9.49, and 10.545 kg. per sq. cm. (120, 135, and 150 pounds per square inch) steam pressure in sulfuric acid a t concentrations from 1 to 7 per cent, in water, and in 3 per cent oxalic acid. Discussion Examination of Tables I and I1 shows that pretreatment of either softwood or hardwood with dilute sulfuric or oxalic acid removed lignin as well as carbohydrate material. Spruce wood, which had been extracted by six 3-hour treatments with 3 per cent sulfuric acid, lost 16.8 per cent of its lignin. (No record was made of the loss a t the end of the fifth 3-hour treatment.) Three treatments with 3 per cent oxalic acid removed 13.9 per cent of the lignin from spruce. Filtrates from the extractions of spruce ~7oodgave a precipitate that was soluble in glacial acetic acid, alcohol, and chloroform. The glacial acetic acid solution of this material was poured
...
Cellulosea
MethOFYI in Liz- Ljg- lignin" ninb nin
%
%
%
%
66.0 63.0 61.0 58.5 55.0 54.4 49.9
33.3 37.1 39.0 41.1 45.0 45.4 50.2
21.2 22.9 23.3 24.1 25.5 25.6 27.3
20.5 20.2 20.0 19.5 19.2 18.9 17.5
. . . . . . . .
. . . . . . . .
150 Pounds' Pressure, 30 Minutes
Meth-
Ex: penment No.
Hydrolyzed wood
%
%
%
492 493 494 495 496 497 49s
62.87 60.52 58.24 57.20 56.55 54.03 52.00
64.6 61 6 58:5 56.4 52.5 49.4 43.0
35.0 38.5 41.5 44.6 47.2 50.0 57.0
...
...
Cellulosea
OXY1
Lignin"
Ligninb
in lignin
%
%
2 2 . 0 20.2 23.3 24.3 19:s 25.6 2 6 . 2 18:5 27 3 2916 1 6 : 9
. . . . . . . . . . . .
. . . . . . . . . . .
was dissolved in glacial acetic and reprecipitated by pouring into water. This material had the same methoxyl content (20.6 per cent) and other properties which indicated that it was lignin. Treatment with ammonium oxalate had about the same effect on the lignin content as did extraction with water (see Table 11). There was a slight loss of lignin. Sodium sulfite solution also removed lignin from wood. This may be accounted for by the alkalinity of the solution. Barium hydroxide removed less because the barium derivative of lignin is insoluble. Fifteen per cent sulfuric acid dissolved less Iignin from wood than did 3 per cent, perhaps because the higher concentration depressed the solubility of the lignin. Hydrochloric acid also had the property of removing lignin; the lignin content was lowered almost 15 per cent'by treatment with 2 per cent acid for 3 hours a t 90" C. When the hydrolysis took place under pressure, as shown in Table 111, a loss of lignin was observed a t the lower concentrations of acid and lower pressures. Increase in acid and increase in pressure caused a reprecipitation of some of the lignin and also the conversion of some of the carbohydrates into a ligninlike residue that was isolated with the lignin and, consequently, gave a residue with a low methoxyl content. This work shows lignin to be soluble in dilute acid solutions. It may be precipitated by long heating, heating under pressure, or increasing the acid concentration. Extraction of lignin-containing material with 3 per cent acid, sodium sulfite, and other salts or bases removes lignin and should be avoided if an accurate determination of the lignin content of a sample of wood is desired.
Literature Cited Campbell, W. G., and Bamford, K. F., Biochem. J., 30, 419
.
11936). ~~I
Cohen, W. E., a n d Harris, E. E., IND.ENG.CHEM., Anal. Ed., 9, 234 (1937'). Harris, E. E., Ibid.,5, 105 (1933). H a w l e y , L. F., and Fleck, L. C., IND. ENG.CHEM., 19, 850 (1927).
MARCH 15, 1939
ANALYTICAL EDITION
(5) Norman, A. G., and Jenkins, S.H., Nature, 131, 729 (1933). (6) Paloheimo, L., Biochem. Z., 165,463 (1925); 214, 161 (1929). (7) Peterson, C. J., Hixon, R. M., and Walde, A. W., ISD.ENG. CREM.,Anal. Ed., 4, 216-17 (1932). (8) Phillips, Max, and Goss, M. J., J. Assoc. Agr. Chem., 21, 140-5 (1938). (9) Ritter, G. J., and Barbour, J., IND. ENG.CHEM.,Anal. Ed., 7, 238 (1935).
155
(10) Ritter, G. J., Mitchell, R. L., and Seborg, R. M., J. Am. Chem. SOC.,55, 2989-91 (1933). (11) Sherrard, E. C., and Harris, E. E., IND. ENG.CHEM.,24, 103 (1932). RECEIVEDNovember 4, 1938. Presented before the Division of Cellulose Chemistry a t the 96th Meeting of the American Chemioal Society, Milwaukee, Wis., September 5 t o 9, 1938.
Testing Dentifrice Abrasives MERVYN L. SMITH Research Laboratory, John & E. Sturge, Ltd., 1 Wheeley’s Road, Birmingham, England
C
ONSIDERABLE emphasis is now placed on the physical properties of the insoluble materials that are almost universally used in dentifrices to assist the brush in cleaning the tooth surface. This close attention was stimulated because certain dentifrices were considered unduly abrasive, while a number of scattered observations on the effect of brushing extracted teeth with commercial dentifrices (6) indicated a danger of damage t o the various tooth structures resulting from daily use over a period of years. It has, moreover, been shown recently (2, 11) that although wear of the enamel is very slight with the majority of dentifrices in use today, the softer tissues exposed by gum recession are much more open to attack. Consequently it is important to have a means of testing the abrasiveness of fine powders. Such tests are necessary for control in choosing suitable types of powder bases and for testing finished dentifrices in order to exclude those having excessive abrasive effects. For these routine purposes biological surfaces that are variable and require tiresome repetition in the test are obviously unsuitable and some standard surface is essential. Several instruments have been described which measure the abrasiveness of fine powders in arbitrary units against such surfaces, but so far their readings have not been standardized in terms of powders of characteristic physical properties nor have the abrasivenesses of the powders tested been related to the amount of wear produced by them on tooth structures. A systematic study should aim first a t grading a series of defined powders on the abrasion apparatus and then attempting to establish limits of particle size and hardness-the two important characteristics involved-liable to cause serious damage to the several tooth structures involved. In this connection Ray and Chaden (6) have pointed out that different individuals probably require dentifrices of different abrasiveness. An experimental approach of this type is simple when one considers a single substance, since it is self-evident that the grading order of a series showing increasing particle size must be the same against all surfaces. (The actual range of abrasiveness of the series will, of course, vary with the different surfaces according to the degree of penetration of the particles.) Difficulty must, however, be anticipated in comparing chemically different powders. Thus a series of samples each containing the same number of identically sized particles could conceivably be graded in different orders by different surfaces because of the specific interaction of the physical factors involved, such as hardness, particle shape, ductility, etc. At the same time it should be possible to assess different powders closely enough for practical purposes, especially as the range of hardness and particle size likely to be used is small.
Grading of F i n e Abrasive Powders
SCRATCH versus ABRASIONTEST.The effect of the abrasive can be estimated either by examining the scratches left on the surface after a small amount of rubbing or by continuing the treatment until a measurable amount of the surface has been removed. The two methods are complementary, each having certain advantages. The scratch test is selective, different sized scratches being made by different particles (8), and is ideal for finding small amounts of added adulterant such as pumice or emery. Even the very simple form of test described in Federal Specification FFF.D.191 for dentifrices is fairly sensitive (IO). The abrasion test gives a quantitative figure for the amount of abrasion but it is not possible to distinguish scratches of different depth. For the purpose of grading powders, however, a quantitative abrasion test is essential, though discrimination must be used in assessing the results when heterodisperse powders are being tested. Experimentally considerable difficulty is involved in these tests, as may be appreciated by the experience of Souder and Schoonover (IO) who found with a rotating table test that “the addition of 10 per cent of fine emery to a paste of minimum abrasiveness did not increase appreciably the loss in weight of the disk.” The apparatus of Ray and Chaden (6) is, however, very sensitive, while the modification with glass bed (8) can distinguish an addition of 0.020 gram per cent of a coarse ingredient in a fine dentifrice. Incidentally, the explanation of Souder and Schoonover that “a film of soap or some other ingredient prevents the two surfaces coming together” is not tenable, since results have been obtained in substantial agreement in comparing powders alone and made up in commercial dentifrices. I n previous work on grading the powders have not been sufficiently characterized and the reason for the wide variation in tests with different samples does not appear. Thus Ray and Chaden using commercial dentifrices found figures for calcium carbonate from 1.5 to 10 (milligrams of weight loss from an antimony block after 5,000 revolutions) and for calcium phosphates from 0.7 to 17, and these differences are presumably due to differences in particle size distribution and the presence of impurities. For the results, however, to have more than a restricted value referring to the particular sample, each series should be characterized from this point of view. Wright and Fenske (11) gave ranges of a similar order (with the same apparatus) using powders, but again no physical description was attempted. An earlier communication (8) described tests on a series of experimentally precipitated chalk powders by which a relation between particle size and abrasiveness was established.
