July, 1925
I-IDUSTRIAL AND ENGINEERIXG CHEMISTRY
741
Determination of Copper Numbers of Cellulose Materialsls2 By C. J . Staud and H. LeB. Gray EASTMAN KODAKCo., ROCHESTER, N. Y.
A proportion of t h e copper sulfate and alkaline tartrate of Fehling solution, Soxhlet modification, has been found which gives no black precipitate on heating, a minimum blank, and a negligible autoreduction. Study of the effect of time and temperature of heating on the copper number shows in all cases an increase in copper number until heating has been continued for 30 minutes, and after t h a t a constant value. Rag pulp heated with Fehling solution for 30 minutes a t 100' C. and 2 hours a t 75' C. gives a copper number corresponding to t h e lower temperature. If t h e filtrate resulting from this treatment is heated for some time a t 100' C. a few milligrams of
cuprous oxide are precipitated. If the material is left in the solution and heating continued a t the higher temperature for 2 to 3 hours a value intermediate between those a t 100' and 75' C. is obtained. The cuprous oxide which dissolves when t h e temperature is lowered from 100" C. to 75' C. is shown to be in solution in t h e cuprous state. The results are roughly quantitative. The reduced copper is determined by the method of Knecht and Thompson, using ferric alum. A modification of t h e Schwalbe method for determining copper numbers is proposed, and the results of its application to a variety of materials are exhibited.
.. .. .. . . .. .. ..
T
HE Schwalbelj* method for the determination of the
copper numbw of cellulose as given in the report of the Committee on Cellulose of the Celluloqe Division of the -1UERICAN CHEMICAL SOCIETY* leads to erratic results. I n some cases negative numbers are obtained-i. e., the copper obtained frorn the sample is less than that found in the blank. To ascertain, if possible, the cause of these variations was the object of this inyestigation. Since the assumption that the adsorption of cupric compounds from hot Fehling solution iq the same as from the cold solution rests on very meager evidence, it mould seem advisable to eliminate the cellulose number if possible. A method that determines the reduced copper only was adapted by Knecht and Thompson3 to cellulose work froin the quantitative method given by Sutton4 for copper in ores. This consists in treating the reduced copper with ferric alum solution and titrating the ferrous ion with permanganate Clibbens and Geake5 have used this method and have obtained the reduced copper hy use of a carbonate-bicarbonate solution of copper, the Braidy method.6 Benesch7 adapted to cellulose the method of Andanti* for reducing sugar in which the cuprous oxide is determined by the use of ferric alum. This work appeared after the present investigation had been concluded. The ferric alum method has been used throughout this investigation. The Schwalbe method as given in the report of the comniittee2 gave variable amounts of cupric and cuprous oxide in the blank. The amount of these precipitates appears to be related to the ratio of the components and to the temperature a t which the solutions are heated. Schwalbeg attempts to obviate this difficulty by using a definite size of flame, a apecial type of burner, and asbestos shields. Quisumbing and Thomaslo have shown that by varying the ratio of alkaline tartrate to copper, the volume and other components being kept constant, the cuprous oxide obtained from a definite amount of dextrose varied. This suggested that by varying the concentration of alkaline tartrate a ratio might be found which would give no cupric and a minimum of cuprous oxide in the blank. Presented before the Division of Cellulose Chemistry a t the 69th Meeting of the American Chemical Society, Baltimore, hld , 4prd 6 t o 10, 1925 2 Communication No. 231 from the Research Laborator) of the Eastman Kodak Company * Xumbers In text refer to bibliography a t end of article
The Solution The alkaline tartrate and copper sulfate solution were made up as directed in the method of the Because of the nature of cellulose it was considered advisable to maintain the volume of 400 cc. The proportion of the copper solution in the total volume was changed from 50 cc. to 20 cc. because the amount of copper formerly used seemed excessive. Preliminary experiment showed that if the components were mixed in the cold and then heated much less cupric oxide ryas produced. By varying the amount of alkaline tartrate a proportion of 30 cc. alkaline tartrate to 20 cc. of copper sulfate solution when heated a t boiling for 15 minutes gave no precipitate of cupric oxide and less than 0.1 mg. of cuprous oxide. The solutions made up in accordance with the above concentration Tvere boiled for varying lengths of time and the cuprous oxide was determined by the ferric
Y
.Ooo'
;
l b
15
210
25
Time
30
35
4;
40
sf Heating
in
Minutes
Figure 1
alum method. The results are shown in Figure 1 and Table I. This solution could be heated for 2 hours a t 100' C. without any appreciable formation of cuprous oxide. Table I-Autoreduction (Heating a t boiling point) Time of heating Reduced copper Minutes Mg. 10 0.084 15 0.084 20 0.072 30 0,099 40 60
0.084 0.10
Time and Temperature of Heating
1
The time of heating necessary to obtain the maximum reduction was investigated. This was first done a t boiling temperature, but because of danger of superheating and because of the lapse of time between the start of heating and the
IXDCTSTRIAL A S D ENGINEERISG CHE*MISTRY
742
boiling point, during which time reduction is taking place, experiments were run a t 100" C. At this temperature for an equal time of heating the results were lower. Determinations were then made a t 75" and 100" C., and a t the boiling point of the solution with varying lengths of time. The results (Figure 2 and Table 11) show that the maximum reduction is practically reached a t 30 minutes. The curve falls off so abruptly below 30 minutes that it was considered advisable to use 45 minutes as the time of heating. Determinations made a t 75", %', 92.5", and 100" C., and a t the boiling point show that as the temperature increases the
19
18
17 I6
When the filtrate from the rag pulp, which had been heated with Fehling solution a t 100" C. for 30 minutes and a t 75" C. for 2 hours, was placed on a water bath for 45 minutes and heated, a precipitate of cuprous oxide was observed. The amount obtained in this way was 4.94 mg. Although far below the theoretical value, it seemed to indicate the precipitability of cuprous oxide from the filtrate with rise of temperature. Re-reversibility
I4 13
12 It 10
zfL 0 9 0
,poe
If, after bringing about reversal as described above, the solution containing the paper or pulp was again heated a t 100" C. for 2 to 3 hours, the value for the copper number increased. Star Linen paper treated in this way gave a copper number of 0.33, and rag pulp 3.52. The corresponding values for these materials a t 75" C. are 0.24 and 2.71. The increase in magnitude of the copper number indicates clearly a trend in the direction of the former value obtained for 100' C. Whether the cuprous oxide which dissolved by lowering the temperature from 100" to 75" C. could be reprecipitated quantitatively by prolonged heating has not been ascertained. However, it has been demonstrated, as described above, that the dissolved copper was in the cuprous condition.
0.7 06 05 04
03 02 01 00
amount of cuprous oxide increases. The temperature-copper number curves were made for two materials, a paper and a pulp. As shown in Figure 3 and Table 111, the curves for these two materials are not parallel. This precludes the possibility of applying a corrective factor for errors in temperature. In view of this it was decided to recommend 100" C. as the temperature for making the determination. Table 11-Change
paper was 0.24 and for rag pulp 2.70. Determinations on Star Linen paper and rag pulp were then made by heating a t 100" C. for 30 minutes, cooling to 75" C., and allowing to stand for 2 hours at this temperature before filtering. The determination of cuprous oxide was then made. Star Linen paper gave an average value of 0.18 and the rag pulp 2.66. The time necessary for equilibrium to be reached was not investigated, but 2 hours appeared to have been ample. That the cuprous oxide dissolved in the solution as a cuprous complex was shown (using Parr'sll method for determining cuprous copper) in the filtrate obtained when the Fehling solution was filtered from the cellulose. In the case of paper pulp there was a difference of 62 mg. between the weight of copper obtained a t 100" C. and that obtained by heating a t 100" C. and then allowing to stand for 2 hours a t 75' C. It was necessary to modify Parr's method by running a blank with fresh Fehling solution and subtracting the value obtained. Precipitation of Cuprous Oxide by Heating Filtrates
15
B
a
Vol. 17, S o . 7
of Copper Number w i t h Time of Heating
Time of heating 7 Minutes Boiling point 1.36 5 1.54 10 1.68 15 1.77 20 1.82 30 1.86 40 1.83 60 1.57 90 120
-----
TEMPERATURE 1000 c. 1.11 1,33 1.43
750 c. 0.107 0.18
1.52
0.24
1 . .51 1.52
0.214
0.214
Reversibility
The decrease in amount of cuprous oxide obtained as the temperature was lowered led to the belief that the cuprous oxide obtained a t 100" C. might be partially dissolved a t a lower temperature. To ascertain this, determinations were made using Star Linen paper and rag pulp. At 100" C . the average value for Star Linen paper was 1.52 and for rag pulp 4.73. At 75" C. the average value for Star Linen
Retention of Ferric Alum
In the determination of cuprous oxide by the ferric alum method, the iron ammonium sulfate is washed out of the material, with 100 ce. of 2 N sulfuric acid. That this amount affords adequate washing is shown by the fact that the value for the blank remained the same when the filter was washed with 200 cc. 2 N sulfuric acid. When Star Linen was used the following copper numbers were obtained: 3-gram sample, time of beating, 20 minutes a t boiling point VOLUME2 N HzSOa USED Sample 100 cc 200 cc I 1 71 1 68 ... I1 1.67 1.i2 1.71 I11 1.70 1.69 IV
Weight of Sample
Using the Braidy method, Clibbens and Geake observed the following variation in copper number with weight of sample : Modified cellulose preparation number 34 35 41 42
-COPPER 1 gram Sample 0 SO 2 21 3 09 3 92
NUMBER2 5 grams Sample 0 Si 2 33 3 48 ,
4 36
I S D C S T R I A L A-YD ESGINEERING CHEMISTRY
July, 1925
Table 111-Change of Copper Number with Temperature of Heating. Temperature Heating for 30 minutes c. Star Linen paper Rag pulp 102.5 100 92 8.5 _t 3
5.0 4 ,77 4.01 3.41 2.70
1,80 1.62 1.02
O.59f 0.24
743
( B ) 692 grams of potassium sodium tartrate, which have been recrystallized from warm water and dried at room temperature, are dissolved in approximately 1 liter of water. T o this are added 200 grams of sodium hydroxide and the solution made up to 2 liters. The sodium hydroxide is prepared by making a saturated solution and allowing it to settle for 10 days. A sample is then pipetted out and titrated with standard hydrochloric acid The yolume of the solution containing 200 grams of sodium hydroxide is then calculated and used Ferric Ammonium Alum Solzction. One hundred grams of ferric ammonium alum, (XH& SO4Fes(SOJ3.24H20, are placed in 700 cc. of water and 140 cc. of concentrated sulfuric acid added. When the solution is clear it is made up to I liter. PREPARATION OF Samm-\Then paper is used it is cut into pieces 2 to 5 mm. square. Linter5 require no treatment. Raw cotton, sliver, and standard cellulose made therefrom are first pulled apart and then cut finely. For pulp the following procedure has been found useful After weighing, the pulp is placed in a 150-cc, glass-stoppered, aide-mouthed bottle, about 50 or 75 glass beads are dropped in, and 75 cc of distilled Mater added The bottle is then qhaken vigorously for 10 to 15 minutes The contents are washed with 25 cc. of water, into the balloon flask in which the determination is made Shaking with glass beads reduces the pulp to a finely divided, "free" state which facilitates the determination S0te-A sample of rag pulp from the beater was dried to constant weight a t 40" One sample -as treated with the glass beads the other torn into small pieces and the copper number determmed on each Thehe g a \ e the folloning results 2 81 and 2 5 5 , respectively
Figure 3
t-siiig the modified Fehling solution and maiiitaining all factors as nearly constant as possible, weights of sample of Star Linen of 1, 2, 3, 4, and 6 grams and of oxidized cellulose of 0.375,0.75,and 1.5 grams were taken for the determinations. The following table and Figure 4 (Table IV) show that the value of the copper number in this procedure is not greatly affected by the weight of sample. Slav L i n e n P a b e v (Solution boiled for 20 minutes) COPPER
h-1MBER----
I---
n'eight of sample ... 1 gram 2 grams 3 grams 4 gram'i ti grams I 1.76 1.Si 1.70 1.85 1.i4 I1 1.68 1.77 1.67 1.62 ,.. I11 ... l.i3 1 71 Averape 1.72 1.79 1 69 1'53 1.74 U-eight of reduced copper, grams 0.01T2 0.0338 0.0509 0.073fi 0.1046 O s i d i z e d Cellzrlose'? (Time of heating, 1 5 minutes a t boiling point) 1. 5 0 75 0 3i5 \Veight of sample, grams Copper number J 85 5 so 5.49 \Teight of reduced copper, gramr 0 0878 0.0433 0.0206
Back Oxidation of Cuprous Oxide
The back oxidation of cuprous oxide to cupric oxide n-as also investigated. It was found that the cuprous oxide after filtration may stand a t room temperat'ure in cont'aclt with air for more than 20 hours with negligible oxidation to cupric oxide. Rag pulp was heated for 30 minutes a t 100" C. and the determination for copper number of cuprous oxide made immediately. The value for copper number so obtained was 4.77. STheii the cuprous oxide was allowed to stand on the filter for I5 hours after precipitation, the value of the copper number x a s 4.68, arid a duplicate experiment in which 21 hours elapsed hetn-een the precipitation and determination of the cuprous oxide gave a copper number of 4.7'6. Method
SOLUTIOSS--JA)Copper sulfate crystals are washed to remove extraneous matter and dried; 137 grams of CuBOi.5HzO are dissolved in distilled water anti made up to 2 liters.
10
PRo CEDURE-LSUally a 3-gram sample is 09 used. However, for substances having a high copper number, or nhen the supply is o, limited, smaller amounts may be satis- T W E factorily e m p l o y e d . g The sample is placed in o5 a 1.5-liter balloon flask : equipped with a short o4 reflux condenser and d stirrer. Into the top of a the reflux condenser is 'O serted a small, shortstemmed funnel, through n hich passes the shaft of the stirrer. The flask is then placed in a bath a t 100" C.
3
r
e
3
4
5
6
Weiqhf in Gms Figure 4
Table IV-Increase in Weight of Cuprous Oxide Precipitated with Increase in Weight of Sample (Heating 20 minutes a t boiling point) R'eight of sample gram? 1 2 3 4 6 W'eight reduced copper grams 0 0172 0 0358 0 0509 0 0736 0 1046 (1 sample)
I n another 1.5-liter balloon flask ale placed 350 cc of water nhen the copper number of paper. linter, raw cotton, drier, or standard celluloqe is being determined, for pulp 250 cc. of mater are used In all cases 20 cc of copper sulfate solution and 30 cc. of alkaline tartrate are added and the contents of the flask heated to 100" C when it is added to the qample in the other flask. This aclclition may be made in from 20 to 30 seconds The sample is heated with the d u t i o n on the nater bath for 45 minutes, with stirring sufficiently rapid to keep the con~
INDUSTRIAL AND ENGINEERING CHEMISTRY
744
tents in gentle motion. The flask is then removed and the solution immediately filtered on a Buchner funnel through two sheets of Whatman No. 5 filter DaDer. (In d a c e of filter paper, asbestos prepared for use In -Gooch chcibles was tried. The difference in the blank obtained was very small.) The material on the filter is washed, first with 1 liter of cold water and then with 750 cc. of boiling water. The Buchner funnel is then transferred to a 250-cc. filter flask. The contents are treated with 25 cc. of ferric alum solution, in the case of standard cellulose. Materials having a higher copper number, or fine texture, may require more ferric alum to remove the cuprous oxide. In such cases it is added in 25-cc. portions and a blank determined to correct for the additional ferric alum. The ferric alum retained by the material is washed out with 100 cc. 2 N sulfuric acid, applied in two 50-cc. portions. The Buchner funnel is then removed and the contents of the ,filter flask titrated to a light pink with 0.04 N potassium permanganate. Calculations Cc. KMn04 used X 0.00254 X F 100 - Copper number Weight of samde 1 cc. 0.04 N KknO4 = 6.00254 gram copper F = Factor of KMnOl
--
CODDer N u m b e r s of Various Cellulosic Materials I1 I11 Average 1 Standard cellulose from: ( a ) Arizona Egyptian 0.0040 0.0048 0.0057 sliver 0,0041 0.0025 0.0047 ( b ) Clarksdale Miss. 0.0065 0.0047 0.0045 0.0057 0.0032 (c) American ieeler 0,0049 0.0025 0,0038 ( d ) Peruvian sliver 0.0041 0.053 0.051 0.049 ( e ) Sulfite pulp 0.0057 0.0049 cf) Linter A 0.0041 American peeler, extracted alcohol and ether. No chemical treat0.0078 0.0082 0.0074 ment Sulfite pulp, after treatment 0.051 0.053 0.049 with 17.5 per cent alkali 0.0097 0.0098 0.0113 o.oos2 Raw linters A 0.055 0,053 Raw linters B 0.058 0.011 0.011 0.012 Raw linters C 0.0196 0.0189 0.0197 Bleached linters C 0.0202 Material precipitated from cuprammonium solution I 0.041 0.049 0 045 Raw linters D 0.020 0.018 0,019 Raw linters E 0.0098 0,0091 0.0082 0.0090 Material precipitated from cuprammonium B 0.017 Paper M 0.029 0.033 0.031 0.031 0.027 0,029 Bleached linters C 3.02 2.97 2.98 2.99 Hydrocellulose A Material precipitated from cuprammonium C 17.3 17.4 17.35 Star Linen paper 1.56 1.45 1.56 1.52 Rag pulp 4.77 4.76 4.68 4.73
Bibliography l--Schwalbe, Robinoff, and Schulze in Robinoff’s Dissertation, Darm. stadt, 1912, and Schulze’s Dissertation, Darmstadt, 1910. 2-Ind. Eng. Chem., 16, 748 (1923). 3-Knecht and Thompson, J . SOC.Dyers Colozwisls, 36, 255 (1920). 4-Sutton, “Volumetric Analysis,” 9th ed., p. 186. . 5-Clibbens and Geake, J. Textile Inst., 15, T31 (1924). ti-Braidy, Rev. gen. mal. color, 26, 35 (1921); J . Textile Inst., IS, T33 (1924). 7-Benesch, Chem. Zlg., 48, 861 (1924). 8-Andanti, Boll. chim.-farm., 65, 33 (1916). 9--Schwalbe, “Die chemische Betriebskontrolle in der Zellstoffund Papierindustrie,” Berlin, 1922. 10-Quisumbing and Thomas, J . Am. Chem. SOL.,43, 1512 (1921). 11-Parr, J . Am. Chem. SOC.,22. 685 (1900). J . SOL.Chem. Ind., 16, 908 (1897); Compt. rend., 1121 12-Vignon, 126, 448 (1897).
Size of Micrographs-Micrographs accompanying manuscripts submitted for publication usually have to be printed without reduction in order that the details may be clear. For this reason authors are asked to submit such figures, if possible, not more than 2 inches wide, and in no case more than 3l/a inches wide. If reduction of the micrograph is not objectionable this fact should be indicated. We wish to use a sufficient number of micrographs properly to interpret the article, but authors should bear in mind that a n unnecessarily large number detracts from its appearance.
Vol. 17, No. 7
Separation of Manganese in the Analysis of Limestone and Similar Materials”’ By Alice Whitson Epperson BUREAU OF STANDARDS, WASHINGTON, D. C.
I
S HIS text on the “Analysis of Silicate and Carbonate
Rocks,’’ Hillebrand3 cites the use of ammonia and ammonium persulfate for the purpose of precipitating manganese together with iron and aluminium, and states that the possibility of partial precipitation of calcium as sulfate has not been investigated. The following tests were therefore made with the idea of simplifying the subsequent determinations of lime and magnesia, both of which would be in error if manganese were left in solution. The elimination of manganese is especially desirable in the analysis of Portland cement and certain types of limestones, for most of the error falls on magnesia which is often one of the deciding factors in the acceptance of many lime-bearing materials. A good precipitation of alumina demands careful adjustment of the pH of the solution, and this Blum4 accomplished by the use of either methyl red or rosolic acid. I n preliminary tests it was found that persulfate destroys methyl red quickly, but not rosolic acid. Rosolic acid was therefore adopted as indicator, and the following modified Blum method was used in the analysis of the Bureau of Standards Argillaceous Limestone No. 1, a sample of hydrated lime, and one of Portland cement, to which had been added measured amounts of MnO in certain cases: To the cool filtrate from the silica determination (150 to 200 cc. in volume and containing 10 t o 16 cc. HCl) add 5 cc. of ammonium hydroxide (sp. gr. 0.90), 5 drops of rosolic acid indicator (0.5 gram dissolved in a mixture of 50 cc. of 95 per cent alcohol and 50 cc. water), and 1 gram of solid ammonium persulfate. Carefully neutralize with ammonium hydroxide until a pink color just appears. Heat the solution t o boiling and boil for 1 minute. Filter immediately, wash moderately with a hot 2 per cent solution of ammonium chloride, and reserve the filtrate. Dissolve the precipitate in hot dilute hydrochloric acid, cool, reprecipitate as before, filter, and combine the filtrate with the reserved one. Table I MNO PER CENT M c O ADDED Present Found Gram REMARKS 1.94 1.94 None 1.94 1.97 None 0.015 No M n in CaO or 1.94 1.96 MgzPaO7 No M n in CaO. less 4 37.65 37.62 1.94 1.93 0.015 than 0.1 mb. in MgiPzOi Less than 0.1 me. Mn 2.33 0.025 5 92.67 92.64 2.32 in CaO; trace“Mn in MgaPzOr M n as in Expt. 5 0.025 2.36 92.67 92.72 2.32 6 No Mn detected in 2.32 0.025 2.37 92.67 92.70 7 CaO or MgzPzO? Trace Mn in CaO and 2.23 63.20 2.26 0.05 8 63.13 MgzPzOT 9 63.13 63.10 2.26 2.23 0.05 Trace M n in CaO; none detected in MgzPzO7 10 63.13 63.18 2.26 2.25 0.05 Trace M n i n CaO and MgzPzO? Experiments 1 to 4 were made on argillaceous limestone; Expts. 5 to 7 were made on hydrated lime’ Expts. 8 to 10 were made on Portland cement. Results in Expts. 5 to 10 repdrted on nonvolatile basis. Trace = less than 0.03 mg. PER CENTCAO Expt. Present Found 1 37.65 37.60 2 37.65 37.70 3 37.65 37.66
-
After this treatment calcium was determined by double precipitation as oxalate and magnesium by double precipitation as phosphate, according to the procedures given in U . S. Geological Survey Bulletin 700. Received April 4, 1925. Published by permission of the Director of the Bureau of Standards. a U.S. Geol. Suruey, Bull. 700, 111. 4 Bur. Slandards, Sci. Paper 286 (August 10, 1916); J . Am. Chem. SOC.,38, 1282 (1916). 1
2
