RECOVERY OF PRODUCTS RESULTING FROM TREATMENT OF WOOD WITH CAUSTIC Donald F. Qthmer end Robert H. Royer' Polytechnic Institute, Brooklyn, N. Y.
Th
lized sodium oxalate was then redissolved in hot water and a calculated amount of milk of lime added. Calcium oxalate and sodium hydroxide were thus formed. The precipitate of calcium oxalate was allowed to settle, and the supernatant caustic solution drawn off and used for another cycle, along with that from the first centrifuging. The calcium oxalate was washed with water, and the first washings were also added to the sodium hydroxide for recyc!e. The calcium salt m-as next decomposed with sulfuric acid to form oxalic acid and calcium sulfate. The calcium sulfate was filtered from the solution of oxalic acid. The oxalic acid was then crystallized from solution by evaporation. The above process described by Hubbard (a), must have presented difficulty in operation with regard to the recycling of the caustic soda, aside from complications due to the salts of acetic and formic acids which, besides being wasted, must hare interfered with the separation of the oxalate salts. The humus would have accumulated with each recycle until the caustic soda was unduly contaminated, which would necessitate the discard of caustic periodically, t o decrease the humus content of the recycle caustic soda. Koller (6) suggested that the mother liquor from the 38" BB. solution a t the beginning of the process be evaporated to dryness, burned to get rid of organic material, and fused to sodium carbonate. This was then to be recausticized with lime t o caustic soda for recycling. Several plants did employ this treatment of the mother liquor, other plants did not. Evidently, then, it was merely a question of economics, based on the size of the plant, whether t o purge the recycled caustic periodically or to install fusion pots, a recausticizing plant, and evaporators to concentrate the weak caustic liquors to the proper concentration for re-use in fusing the sawdust. Sodium formate and sodium acetate were not considered in the early production of oxalic acid, since efforts were directed only to the yield of oxalic acid and the recovery of excess caustic soda for recycle. The additional revenue obtainable from the recovery of these sodium salts of the volatile acids would help to defray the expense of the production of the major product and might make the process commercially feasible in competition with the methods nom used. Little or no work has been done on the problem of separating the constituents existing in the mother liquor after crystallized sodium oxalate has been removed from the 38' B6. solution discussed above (6). Five salts are present in the fusion cake-sodium oxalate, acetate, formate, carbonate, and hydroxide. Few descriptions of separation methods for such complex systems of pure salts have been given. I n this case the separation is further complicated by the presence of organic material in the fused cake, which interferes with crystallization. The tools available for such a separation may be considered:
e separation and recovery of products from the treatment of w o o d w i t h caustic i s the step f o l l o w i n g the study of optimum conditions and the development of a method and equipment for large-scale production. This paper reports w o r k o n one of the methods applicable t o industrial use. Caustic fusion produces a cake containing humus materials, sodium oxalate, sodium acetate, sodium formate, sodium carbonate, and excess sodium hydroxide. A separation process for the five sodium compounds from aqueous solution w h i c h w o u l d b e applicable t o manufacturing operations was the first consideration; this complex system was first studied using a synthetic cake containing these materials, added i n the k n o w n ratio o f salts present i n the fused material. Of the several methods w h i c h might b e used t o o b tain salable products, successive steps of fractional crystallization and precipitation, f o l l o w i n g evaporation to the desired concentration, w e r e used t o give preliminary separation o f the oxalate, of the acetate and formate together, and of the carbonate and caustic. Double decomposition w i t h lime produced calcium salts,of the oxalic acid, w h i c h are more economical for treatment w i t h sulfuric acid For recovery of the acid; this treatment also recausticized the carbonate for re-use. The separation of the constituents of the actual cake was hampered b y the presence of humus materials. The use of activated carbon as a preliminary treatment i n obtaining the crystallizable salts was found helpful. Flow sheets and material balances of the separation and recovery steps are presented.
N T H E two previous papers the variables affecting caustiIdescribed. cization during the fusion of sawdust with alkali were The minimum ratio of caustic soda to sawdust
and the maximum yield of oxalic acid per pound of caustic soda used and not recoverable, and per pound of dry sawdust, were major factors (page 262) as well as the development of a continuous, workable, fusion process (page 268). The methods of separation and recovery of the products resulting from the treatment of wood with caustic were not satisfactory, however, and further study was desirable. The products formed as a result of the above treatment are sodium oxalate, formate, acetate, carbonate, volatile constituents such as methanol, noncondensables such as carbon dioxide, carbon monoxide, nitrogen, and oxygen, complex organic impurities identified here as humus, and excess alkali. Until about the time of the first World War, there was a commercial process based on the fusion of sawdust with caustic soda; oxalic acid was the product of chief interest. After fusion the mass was dissolved in the smallest possible quantity of water. The hot solution was then adjusted to 38" BB. and allowed to cool. The sodium oxalate crystallized out and was freed by a centrifugal extractor. The crystal-
asystem. gA;t ~ The new ~solid phase may ~ ~~ be filtered,
~
~~
$
2. The salts may be dissolved in a solvent and an agent added which precipitates a component or components, but does not react chemically.
Present address. E. I. du Pont de Nemours & Company, Inc.. Wilmington, Del. 1
274
~ ~~
~
INDUSTRIAL AND ENGINEERING CHEMISTRY
March, 1942
3. The salts may be leached or extracted with some liquid medium having a preferential solubility for one or more of the compounds. 4. The components may be separated by fractional crystallization, using heating, cooling, and evaporation of water as a means of influencing solubility.
Method 4 was used since it is the most direct. SEPARATION OF PURE SALTS BY FRACTIONAL CRYSTALLIZATION
After various preliminary studies, a synthetic cake was made up using pure salts in the composition obtained in the previous report (page 268), as follows: Sodium oxalate Sodium acetate Sodium formate Sodium hydroxide Sodium carbonate
Grams 53.00 27.25 5.90 209.00 GO. 00
% ' of Total Salts 14.9 7.7 1.7 58.8 10.9
This totaled 355.15 grams of salts, which were dissolved in 4000 grams of water, the amount required to dissolve the salts. The density of the hot solution was 9.5" BB. A covered, stainless steel kettle was used with a stirrer operating a t about 60 r. p. ni. and with the shaft through the cover. The cover prevented carbon dioxide in the air from combining with the caustic soda in solution to form more sodium carbonate (Figure 1).
m
-MOTOR
STIRRER
275
The percentages in Table I represent the amount of each salt in each fraction. A clearer picture of the distribution of salts between fractions appears in Table 11. The first column in each case gives the percentage of the total original amount of the respective salt which is crystallized in the several individual fractions. The second column in! each case represents the cumulative percentages of the respective salts up to the indicated fraction, crystallized of the total original amount (Figure23). FURTHER SEPARATION OF FIVE FRACTIONS
The next step was the separation of the components of the various fractions. The salts of fraction 1 were separated as follows: An amount of the cake was dissolved in water in the ratio of 20 of water to 1 of cake, to dissolve the salts completely; and a stoichiometric quantity of milk of lime was added, based on the amount of oxalate in solution. The calcium oxalate formed was filtered from the warm solution. This was washed with water until free of the caustic which had formed during the reaction. The washings were added to the filtrate from filtration of the calcium salt from the mother liquor. The mother liquor containing sodium acetate, sodium formate, and lye, resulting from the action of milk of lime on the sodium oxalate, was concentrated to 50" BB. and then allowed to cool to room temperature. Sodium acetate and sodium formate crystallized out. the mixture of acetate and formate crystals was filtered from the concentrated solution on a Buchner funnel. The acetate and formate crystals were not separated, although this can be done by fractional crystallization.
TIRREL-
BURNER
R INGSTAND
-
FUNNEL 1
1
Figure
JACKETEDBUCHNER
1.
i
SUCTION
With the stirrer operating and the stainless steel kettle heated by two Tirrel burners, suspended horizontally on opposite sides of the kettle so that the flames played across the bottom of the kettle parallel to each other, the solution was concentrated to 15" BB. The burners were then removed and a steam bath substituted, to keep the solution hot. A steamjacketed Buchner funnel and a receiver immersed in hot water were prepared, and the hot solution was filtered by suction (Figure 2). The purpose of keeping the filtrate hot was to prevent further crystallization and to facilitate removal of the solution from the flask. The salts were removed and dried in the oven a t 110" C. This constituted fraction 1. The hot solution was returned to the kettle and evaporation continued. The solution was next evaporated to 32' BB, (hot) and the above procedure followed. This apparatus simulated the salting evaporator used in a batch operation. Five fractions were obtained in this manner, the fifth being liquid caustic soda residue. The solid phase was removed from the solution each time by the method previously described, and analyzed as reported in the second paper. A small amount of iron oxide was usually evident and was attributed to attack on the stainless steel. The results are shown in Table I.
FLASK
STEAMEXHAUST
Apparatus for Evaporation and Crystallization Studies
FKETTLE
L/
\i
Figure 2. Apparatus for Hot tration of CrystaIlized Salts
Fil-
TABLEI. FRACTIONAT~ON OF SYNTHETIC SOLUTION OF FOUR SALTS AND CAUSTICSODA Fraction No. HnO evaporated Differential Q Total % of original mixt. Temp. a$ end of fractionation, C. Density at end of fractionation (hot), O Bb. Na oxalate, N a acetate Na format;, Na carbonate, % Na hydroxide, % ' Ferric oxide, % Water, %
94
a b
Room temperature. Liquid.
1
2
3
4
5
51.8
30.7
4.5
8.8
....
51.8
82.5
87.0
100 15 91.82 7.04 0.95 0 0 0.33
....
102 32 19 04 32.42 6.55 40.47 0
1.44
....
110 34 0 2.10 0.53 97.37 0 0
....
96.8 B a l a n c e as NaOH soln 147 50.5 0 80.39
19.61 0 0 0
.. ..
254 500 0
0 0 0 52.10b 1.81 46.09
INDUSTRIAL AND ENGINEERING CHEMISTRY
276
Vol. 34, No. 3
AND CUMULATIVE PERCENTAGES OF TOTAL SALTS IN SUCCESSIVE FRACTIONS OF SYNTHETIC SOLUTION TABLE11. INDIVIDUAL
Na Na Na Na
oxalate
acetate formate carbonate Na hydroxide
1Individual 89.37 13.42 8.48 0 0
Cumulative' individual 89.37 13.42 8.48 0 0
2
Cumulative individual
10.63 36.55 33.31 12.53 0
100 48.97 41.79 12.63 0
The second fraction was composed of sodium oxalate, sodium Carbonate, sodium acetate, and sodium formate; and a part of it was dissolved in fifteen times its weight of water. Two thirds of the water was evaporated t o obtain maximum oxalate yield. The solid phase was filtered from the solution, and both solid phase and mother liquor were analyzed. Results were as follows: Sodium oxalate Sodium acetate Sodium formate Sodium carbonate Water Iron oxide
Solid Phase
Mother Liquor
83.54% 13.15 2.72 0.59
0.29y* 2.73 0.55 3.71 92.57 0.15
... ...
3-
Cumulative 0 55.65 49.53 100 0
0 6.68 7.74 87.47 0
IndividualCurnulativ; 0
44.35 50.47 0 0
0 100 100 0 0
7 -
Individual 0 0 0 0
100
5
Cumulative 0 0 0 0 100
practically all of the caustic soda present as such in the original cake. It might seem desirable to attempt t o increase the oxalate content of fraction 1 (Table 11) by further concentration or evaporation so that fraction 2 would contain less oxalate. As already mentioned, however, further evaporation makes the carbonate begin to crystallize. Sodium acetate and sodium formate seem to persist throughout the crystallization process, with largest amounts crystallizing in the second and fourth fractions. Because the system is approaching saturation with respect to sodium hydroxide, all of the carbonate crystallizes out in fraction 3. On this basis, all of the carbonate can be removed from a caustic solution which has been evaporated to a certain concentration (4). While fractional crystallization of the components in the five-solid system seems the most direct method, other ways of separating the components may be more advantageous although more expensive as process operations. This will be discussed further.
There was a small amount of carbonate in the solid phase and a small amount of oxalate in the filtrate. The oxalate content in the filtrate could be reduced to a negligible amount at the expense of increasing the carbonate content in the solid phase (8). The oxalate, acetate, and formate fraction obtained from the solid phase in this manner could now be "limed',, a s was t h e SEPARATION OF first fraction, and MATERIALS IN the filtrate r e s u l t i n g ' FUSION CAKE from separation of the T h e a c t u a l fused calcium oxalate could material from causticbe treated in the same sawdust fusion was next manner as described i n v e s t i g a t e d . A run above. was carried out with The third fraction was maple sawdust in the almost all carbonate. fusion trough as previwith some acetate and ously described (page formate. Water was 268) t o produce fused added until the carbonm a t e r i a l i n sufficient ate content was 7 per q u a n t i t y f o r further cent by weight (7). study and for pilotThe carbonate was conplant operations. An verted to caustic soda attempt was made to by adding milk of lime. use the same technique T h e same condition and precautions as preexisted then as with the viously outlined, but the first fraction when it results were not so good was treated with milk as the best obtained by of lime. the previous operator T h e s o l u t i o n now who had had greater comprised sodium hye x p e r i e n c e with this droxide, sodium acetate, process and equipment. and sodium f o r m a t e . The fused cake was This solution was likeanalyzed by the method wise concentrated to reported (page 268), and 50" B6. and allowed to the results are shown in cool. A c e t a t e a n d P Table 111. formate crystallized out, 2 3 4 5 l 2 4.5 I 2 3 4 1 2 . 5 4 5 I 2 4 s ,TE HYDROXID The composition of XALATE CARE ACETATE FORI' E and were added to other the synthetic cake, or Figure 3. Relative Amounts of Each Constituent i n acetate-formate crystals five-solid system of pure from the treatment of Successive Fractions (I to 5) salts, was based on the other fractions. Thick lines indicate percentage i n individual fractions; thin lines, ratio of salts in the fused The fourth fraction recumulative percentages up t o indicated Fraction. cake reported in the sulted in acetate and formate crystals, after earlier paper. By developing greater skill in control, the higher proportions of oxalic concentrating and cooling the mother liquor, The U t h fraction was the residual mother liquor and contained acid (70 per cent of the original dry wood substance or even
1-
E
March, 1942
INDUSTRIAL AND ENGINEERING CHEMISTRY
higher) might be obtained. For the purpose of this investigation, however, the cake with above analysis was used as a basis for further study. The ratio of the salts in the fused cake was approximately the same as that of the synthetic cake studied, except that the sodium carbonate content was 16.9 per cent, based on total salts; in the actual fused cake there was only 2.55 per cent sodium carbonate which was an advantage in the later work since a large amount of carbonate was not desirable. This low content of carbonate in the fused cake did not affect the solubility relations of the other salts present. Further investigation on the fused cake, therefore, was continued. Efforts were made to apply the data obtained from the pure salt system to the fused cake as it was processed.
TABLZI 111. ANALYSISOF FUSEDCAEE % of Fused Cake 11.19 7.52 4.17 3.06 1.94 1.31 2.55
60.85 19.30
Ori inal Dry Wt. o f Sawdust 43.99 29.57 16.39 12.03 7.44 5.15 11.94
...
...
A quantity of fused cake was dissolved in ten times its weight of water, and the solution was evaporated to 15" BB. in the stainless steel pan. The Tirrel burners were removed, and a steam bath was substituted to keep the solution hot. A steam-jacketed Buchner funnel and a receiver immersed in hot water were prepared, and the hot solution was filtered by suction. During evaporation the solution foamed, but as the concentration increased, foaming decreased until it finally stopped. Foaming was less of a problem in handling the pure salts than in handling the fused cake. The extraneous organic material (probably wood fatty acid soap) dissolved in the caustic solution probably caused the greater foaming, as it does in the paper industry where "black liquor" containing organic material is concentrated, As filtration progressed, suction of the precipitate became increasingly difficult. Dark humuslike particles (possibly unreacted wood substance) were evident in the funnel, as well as a slimy, dark brown, semisolid mass. There were no crystals in the funnel. The filtrate, however, was further concentrated to 32" BB., and the same procedure followed. An additional slimy, organic, dark brown mass was evident after filtration of the concentrated solution. No crystals were embedded in the precipitate. Erdmann (1) showed that aromatic substances were present in the products of the alkali fusion of certain types of wood, but he found no such substances volatile in steam distillation. The presence of impurities in solution makes crystallization difficult, and if these impurities were removed, crystallization would take place readily as with the pure salts. Absorption of the impurities was tried. Fifty grams of fused cake were dissolved in 500 grams ff water; to this were added 50 grams of an 8-14 mesh activated carbon of the granular type used for water purification and selling for about 3.5 cents a pound. The solution was brought near the boiling point and held there for half a n hour. The density of the solution was 10" BB. This density was necessary to ensure complete solution of the salts. Then the solution was filtered on a Biichner funnel. The light orange filtrate was 8.5" BB. and weighed 541 grams. A portion of the filtrate was evaporated to 15" BB. Clean, white crystals resulted, corresponding to fraction 1of the earlier work.
217
The above experiment was repeated to obtain an orange-colored 8.5" BB. filtrate. This filtrate was treated with 15 grams of activated carbon. It was brought to and kept near the boiling point for 15 minutes. The solution was again filtered on a Buchner funnel. The filtrate was waterwhite. The density before and after filtration of the treated solution was still almost exactly 8.5" BB.; thus the impurities in solution did not appreciably affect its density. The previous methods were followed on the water-white solution, and operations went forward the same as for the synthetic cake. When concentrated from 34' to 50.5" BB. to give fraction 4, the solution became somewhat lemon-colored, but crystallization was in no way impeded. For each gram of fused cake a total of 1.3 grams of activated carbon was used to clarify the solution. This includes the carbon required by the second treatment; a better choice of activated carbon would probably eliminate the second treatment. Increase in contact time and other adjustments of the method would probably decrease the carbon per unit of fused cake to a more nearly economic amount. This is important because activated carbon is a comparatively expensive raw material. For each gram of fused cake 10 grams of water were used. This large amount of water gives a comparatively dilute solution in which the activated carbon acts efficiently. After the treatment of the fused cake solution with activated carbon, the density was 8.5" BB. This is explained by the fact that humus was included in the ratio of fused cake to water. The percentage of water evaporated was little more in concentrating the solution from 8.5" to 15" than it was in concentrating the synthetic cake from 9.5" to 15" BB. The differential percentages of water evaporated between fractions above 15" BB. were practically the same as in the case of the synthetic cake. A lower ratio of water to cake could theoretically have been utilized in such a manner as to extract only acetate and formate, leaving oxalate as the residue and giving the first fraction without evaporation. Complete solution of the various fractions gives a more direct starting point and prevents the necessity for the time-consuming extraction of salts; acetate and formate are dissolved; and oxalate is left. The extraction method might well be used in the ultimate setup of the process to minimize evaporation costs, if the probable increase in the amount of activated carbon required to purify the solution a t higher concentrations is overcome. Advantage was not taken in this separating process of differences of solubilities of the several salts, hot and cold. The study of this variable was left for a later report. PILOT-PLANT RUN
A quantity of fused cake (of the composition indicated in Table 111) was dissolved in ten times its weight of hot water in a gas-fired copper kettle. The solution 'was brought near the boiling point and then pumped to a ceramic filter. The tray of the filter contained granular, 8-14 mesh, activated carbon. Suction was applied to increase the rate of filtration. The filtrate was recirculated until clear. After this operation the clear solution was pumped to an outside calandria type, forced-circulation evaporator having a salt receiver. A coarse cloth filter was used on top of the regular screen because of the fineness of the crystals produced. The solution was concentrated to 15" BB. When this point was reached, steam was turned off and the solution was "dropped" into the
INDUSTRIAL AND ENGINEERING CHEMISTRY
278
salt receiver. The first fraction was left on the screen of the basket after the solution was sucked back into the body proper for further concentration. Further evaporation of the solution in the evaporator up to 50" BB. yielded the second and third fractions on the screen of the basket, By the time the solution was evaporated to 50" BB., the volume was not much more than enough to fill the calandria, pipes, and circulation pump. For convenience in handling, the liquor was drained from the evaporator and put in a covered copper kettle. The solution was allowed to cool to room temperature. A stainless steel screen placed over the outlet of the kettle separated the acetate-formate crystals which crystallized out from the caustic solution when it was drained. The first three fractions obtained on the screen of the evaporator basket were small enough in bulk to be conveniently handled in the stainless-steel kettle previously used. They were separated into their component parts according to the method previously described. SUGGESTED FLOW SHEET
Although this process has the obvious disadvantage of high cost for activated carbon, and other processes may be .developed with lower costs, nevertheless it will operate. Figure 4 shows a plant operation which may be used as a basis for further study. Three parts of caustia soda to one of sawdust are mixed together (see page 262) in a bin which leads to a screw-type conveyor. This conveyor feeds the mixture to a gas-fired conveyor, similar to that already developed and used in this work. Here the chemical process
Vol. 34, No. 3
takes place. The "oxalatok,'as it may be termed, serves to convert the sawdust into sodium oxalate, sodium acetate, sodium formate, and sodium carbonate by action of the caustic soda at high temperatures. Humus, fatty acids, and excess caustic soda are also present. The oxalator feeds into a tank where ten times its weight of water is added for solution of the fused cake in preparation for purification by the activated carbon. This is then pumped through the activated carbon to eliminate impurities and clarify the solution prior to fractional crystallization by the evaporators. Although the flowsheet indicates that steam is used for cleaning the activated carbon, this was not tried to any great extent in the laboratory or during the pilotplant run. The activated carbon used in the laboratory was boiled with hot water to determine whether i t could be recovered in this manner. Results seemed promising, but this was not studied further. It would be desirable to recover activated carbon to decrease costs. Also increase in contact time would help decrease the amount of activated carbon per unit of fused cake. This can be done by recycling solution through carbon, as indicated on the flow sheet. The separate evaporators are indicated to concentrate the purified solution to various concentrations for the several fractions. The first evaporator operates with the solution a t 15" BB., and 'the crystals are removed in the usual manner from the salt basket. The following fractions would be obtained similarly from the mother liquor successively circulated to the other evaporators operating a t their respective optimum densities and temperatures. By using three evaporators, a continuous process is available, and on a large enough
OX. OXALATE AC. ACETATE
F.
FORMATE
C. CARBONATE OXALIC A C I D
Figure
4.
Flow Sheet for Production of Oxalic A c i d and Recovery of Reaction Products
,
March, 1942
I N D U S T R I A L A N D E N G I N E ER I N G C H E-MIS T FLY
scale better heat economy would be secured by suitable multiple-effect operation. The various fractions are separated into their component parts. The first fraction, containing most of the oxalate, is treated with milk of lime to separate the oxalate as the insoluble calcium salt. As a result of this conversion sodium hydroxide is formed. The calcium salt is filtered or separated from the sodium hydroxide, acetate, and formate also present in the first fraction. The calcium oxalate is washed to remove adhering solution and is added to the filtrate from the filter. The solution is pumped to a vacuum caustic concentrator to be evaporated to 50” B6. and then allowed to cool. The acetate and formate crystallize out on cooling and are removed from the liquor by a centrifuge. The caustic soda is recycled for re-use in a subsequent fusion. The calcium salt is converted into calcium sulfate and oxalic acid by mixing with sulfuric acid. The calcium sulfate is removed by filtration and washed t o remove adhering oxalic acid. The weak oxalic acid solution is concentrated to 15” B6. in a lead pan and allowed to cool. Additional calcium sulfate comes down and is again removed by filtration. The 15” B6. liquor is pumped to a vacuum evaporator to be concentrated to 30” B6. and then allowed to cool. Crude crystals of oxalic acid come down and are separated by a centrifuge; the resulting liquor, which is high in sulfuric acid, is recycled to the acid converter for re-use. The crude crystals are redissolved in hot water and recrystallized, the oxalic acid crystals being separated in the same manner as the crude product. The second fraction is dumped into a dissolver, and fifteen times its weight of water is added. Two thirds of the water is evaporated and practically all of the oxalate comes out, plus some acetate and formate. This is centrifuged from solution and treated in the first liming step, as described above. The mother liquor is added to the third fraction. The third fraction, consisting of carbonate, acetate, and formate, is dumped into the first “carbonate limer” where the concentration of the carbonate is adjusted to be 7 per cent by weight (7). The resulting solution is limed with milk of lime, which converts the carbonate t o caustic soda. The reaction takes place in a thickener. Two stages of causticization are required to convert carbonate to caustic soda (7), or, preferably, other systems such as the Dorr may be used. The solution then consists of caustic soda, sodium acetate, and sodium formate. This solution is pumped to the caustic concentrator for treatment similar t o that for the first fraction. The 34” B&.liquor in the third evaporator is pumped t o the caustic concentrator after the crystals, which constitute the third fraction, have been removed and concentrated as previously described. The type of evltporator for this final evaporation would probably be different from that for the previous steps because of the greater density and viscosity. The acetate and formate are removed from this
219
evaporation in a mixture of crystals. As markets exist for a mixture of the two acids or a mixture of the two salts, they are merely washed, but the crystal mixture could be fractionally separated if necessary. Thus, a process has been described which appears applicable to the separation of the salts resulting from caustic fusion of sawdust. Perhaps the study of the time required for each unit operation would result in the elimination or simplification in the amount of equipment employed as indicated on the flow sheet. The several fractionating evaporation steps could be conducted in the same evaporator if the operations were on a small scale. The coordination of the time cycle of each unit operation would also permit the use of other pieces of equipment for more than one operation and thus cut down the cost of the plant. Methanol, not considered in this report, is also formed during the fusion of sawdust with caustic soda. Mahood and Cable (6) reported that they obtained 2.5 per cent methanol from maple, based on the dry weight of the sawdust; and a previous paper of this series indicates that 5.5 per cent was obtained. The recovery of methanol might prove advantageous economically but would require a closed unit for the fusion step. To secure large yields of oxalic acid, air would have to be recycled through such a closed system, and the methanol condensed and scrubbed out of the noncondensable gases. As the gases given off include carbon monoxide, an explosivemixture might result. An economic survey of this process would be required t o determine whether it could compete with present methods of producing oxalic acid. ACKNOWLEDGMENT
Appreciation is expressed to Randall D. Sheeline for his help throughout this work, to the Child’s Pulp Colors Company for use of laboratory space, and to the Gray Chemical Company for the activated carbon. LITERATURE CITED Erdmann, “Annalen der Chemic”, Vol. V, p. 223 (1867). Hawley, L. F., and Wise, L. E., “Chemistry of Wood”, A. C. S, Monograph, New York, Chemical Catalog Co., 1926. Hubbard, Ernst, “Utilization of Wood Wastes”, London, Scott, Greenwood & Son, 1902. Jones, R. O., U. S. Patent 1,500,994 (1924). Koller, Th., “Utilization of Waste Products”, London, Scott, Greenwood & Son, 3rd ed., 1918. Mahood and Cable, J. IND.ENQ.CHBM.,14, 1056 (1922). Partington, J., “Alkali Industry”, 2nd ed., London, BailliQre, Tindall & Cor, 1925. Scheibel, E., Brooklyn Polytechnic Inst., unpub. rept., 1938.
