INDUSTRIAL AND ENGINEERING CHEMISTRY
1114
The roof-type still is similar to the construction of a greenhouse, which has a much longer anticipated life than 10 years. The basis of amortization for the solar still may therefore be considerably longer than 10 years. The Las Salinas solar still was in continuous operation for 36 years. It may have operated for a longer time. Solar distillation appears to be a very promising method for producing fresh water in semitropical and tropical zones. The writer expects to be able to continue its development.
literature Cited (1) Abbot, C. G. Pub. No. 3530, Smithsonian Inst. iWisc. Coll., 98, No. 5 (1930); Vol. 2, Smithsonian I n s t . Ser. 1944; U. S.Patent 2,141,330 (Deo. 27, 1938).
( 2 ) Liultman,W. W., Eng. and Sci., 12, No. 5, 12-16 (1949). Aultman, W. W., J . Am. W a t e r Works Assoc., 42, 775-94 (August 1950). (3) Barnes, W. S.,U. S.Patent 2,383,234 (Aug. 21, 1945). (4) Bohmfalk, B. H., Ibid., 2,332,294 (Oct. 19, 1943). (5) Boutaric, A,, Chaleur & Ind., 11, 69-66, 147-155 (1930). (8) Boutaric, A., Recherches et Inventions, 8 , 205-15 (1927). (7) Campobasso, J. J., J . Am. W a t e r Works Assoc., 40, 547-52 (1948). (8) Chapman, 0. L., R e d a m a t i o n Era, 35, 162 (August 1949). (9) Delano, W. R. P., et al. (to Gallowhur Chemical Co.), U. S. Patents 2,398,291 and 2,398,292 (April 9, 1946); 2,402,737 (June25, 1946); 2,405,118 (Aug. 6, 1946); 2,405,877 (Aug. 13, 1946); 2,412,466 (Dec. 10, 1946); 2,413,101 (Dee. 24, 1946); 2,427,262 (Sept. 9 , 1947). (10) Dooley, G . W,, U. S.Patent 1,812,516 (June 30, 1931). (11) E n g . News-Record, 1 4 4 , 3 2 4 (1May 18,1950). (12) Fritz, S., Heating a n d Ventilating, 46, 69-74 (January 1949). (13) Ibid., 46,85-9 (July 1949). (14) Ginnings, D. C., U. S. Patent 2,445,350 (July 20, 1948) (granted
Vol. 45, No. 5
under the act of March 3, 1883 as amended April 30, 1928; 370 O.G. 757).
(15) (16) (17) (18) (19)
Hand, I. F., Hehting and VentiZati.ng, 47,92-4 (January 1950). Hand, I. F., M o n t h l y Weather Reo., 69, 95-125 (1941). Hand, I. F., U . S. W e a t h e r B u r . Tech. Paper, No. 11 (1949). Harding, J., Proc. Inst. Civil Eng., 73, 284-8 (1883). Hollingsworth, F. N., Heati?zg and Ventilating, 45, 99 (August
1948). (20) Hottel, H. C., and Woertz, B. B., Mech. Eng., 64, 91 (1942). (21) Kausch, O., “Die unmittelbare ausnutzung der Sonnenenergie,” Weimar, Carl Steinert, 1920. (22) Keenan, J. H., and Keyes, F. G., ‘‘Steam Tables,” New York, J. Wiley & Sons,Inc., 1944. (23) Latham A., Jr., Mech. Eng., 68, 221-4 (1946). (24) Lavoisier, A. L., “Oeuvres de Lavoisier,” Vol. 3, Table 9, Paris, (25) (26) (27) (28) (29)
(30)
Son Excellence le ministre de l’instruction publique et des cultes, 1882-1893. Mouohot, A., “La chaleur solaire et ses applications industrielles,” Paris, Gauthier-Villars, 1869. Pasteur, F., Compt. Rend., 30, 187, 1928. Richards, J., Recherches et Inventions, 8 , 474-5 (1927). Simpson, H. S., and Palmer, E. J. (to Allis Chalmers Co.), U. S. Patent 2,424,142 (July 15, 1947). Telkes, M., “Distilling water with solar energy,” unpublished report (January 1943). Telkes, AI., OSRD Rept., KO.5225, OT8, PB 21120 (May
1945). (31) Cshakoff, E. A. (by direct, and mesne assignments, of 35% t o S. A. Baron, Kew Orleans, La., for the benefit of himself and F. A. Middleton), U. S.Patents 2,455,834 and 2,455,835 (Dec. 7, 1948). (32) Wheeler, N. W., and Evans, IV. W., Ibid., 102, 633 (May 3 , 1870). RECEIVED for review April 17, 1961. A C C E P T E D February 24, 1953, Presented before the Division of Agriculture and Food Chemistry, 119th Meeting, AMERICAAC R E m c A L S O C I E T Y , Boston, AIass. Publication No. 22 of the Solar Energy Conversion Project.
0
0 0
Concurrent Production of Fatty Alcohols and Sodium Cyanate
0
FRED 0. BARRETT, J. D. FITZPATRICKI,
RICHARD G. KADESCH
AND
Emery Industries, Inc., Cincinnati, Ohio
T
HE original Bouveault-Blanc method ( 4 ) for the reduction of an ester to the corresponding alcohol employed an excess of sodium and ethyl alcohol. The procedure has been modified by using toluene (6) or petroleum ether ( 3 ) as a solvent. Yields are only 70 to 80% because of side reactions including acetoacetic ester condensation and evolution of hydrogen. Further improvements involving careful adjustment of the ratio and quantities of reactants ( 2 2 ) and the use of a secondary reducing alcohol ( 7 ) have given yields over 90%. This improved process is applied commercially to produce fatty alcohols from tallow, hydrogenated tallow, and coconut oil (9). Methyl isobutyl carbinol is used as reducing alcohol and toluene as solvent. When reduction is complete the mixture is quenched in water, and the solvent-fatty alcohol layer is separated and distilled. The aqueous glycerol-sodium hydroxide layer is conveniently utilized in soap manufacture. As four atoms of sodium are used in the reduction of each ester group, the amount of by-product sodium hydroxide is large:
0
/I
CHzOCR
I
f
i;!
CH-OCR
address, Applied Science Research Laboratory, University of Cincinnati, Cincinnati 21, Ohio.
1
--+
OH
CHzOCR CH20h-a
I I
CHONa
+ 3 RCHzONa + 6 CH3CHCHzCH(CH3)z 1
CHzONa CHzONa
I I
CHONa CHzO?ia CHzOH I
6HOH 1 Present
+ 12 Na + 6 CH3CHCHzCH(CH3)2
I
CHZOH
(1)
OXa
+ 3 RCHzONa + 6 CH&HCH2CH(CH& + 12 H20+ 1
ONa
+ 3 RCH,OH + 6 CH3CHCH2CH(CH& + 12 NaOH 1
OH
(2)
May 1953
INDUSTRIAL AND ENGINEERING CHEMISTRY
1115
esters of some mono- and dibasic acids are given in Table 11. Use of xylene in most cases Solvent- Reduction Reaction with Urea % Yield , NaOCN Ester Time, Time, T;mz.., Fatty Purity, as diluent in place of the usual Triglyceride" Solvent Ratio Hours hours alc. NaOCN Glycerol % toluene permits a higher reflux Coconut oil Toluene 1 . 3 :1 1.5 8 110-20 83.8 98.5 88.5 94 temperature and shortens the 1.3:l 1.5 . . . . 84.0 ..b 88 Xylene 1.3:l 1.5 4 135-40 84.5 .. 88 93:5 time required to complete the 2 . 1 :1 2 4 135-40 Glyceryl trioleate Xylene 85.0 98 89 95 Babassu oil Xylene 2.3:l 1.5 4 135-40 86.5 95 60 97 reaction between urea and the Palm kernel oil Xvlene 2 . 2 :1 1.5 6 135-40 85.0 96 60 97 alcoholates. Some rnsults by Hydrogenated tal1.5 low Xylene 2.2:l 6 135-40 87.0 95 85 93.5 the standard water quench Tallow Xylene 1.1:l 1 4 135-45 86.2 98 88 97.5 1.8:l 2 8 135-40 82.0 Soybean oil Xylene 95 93.5 method are included in Tables I a Triglyceride contained 0.2 t o 0.3% free fatty acid, calculated as oleic acid. and 11 for comparison. The b Run worked up by conventional method of water hydrolysm. Glycerol obtained by distilling sodjum.cyanate filtrate, separating lycerol layer from distil!ate. and washing yields of fatty alcohol and glycfatty alcohol layer m t h water. Low yield is probably due t o polyether formation during distille.tion. erol by the two methods are comparable. Sometimes when urea is used, emulsions prevent Table 11. Sodium Reduction of Methyl Esters" efficient separation when the 70FFA in Solvent-Ester Temp. of Urea % Yield NaQCN glycerol is washed from the Methyl Ester Esterb Ratio Reaction, C. F a t t y ala. NaOCN Punty, % sodium cyanate filtrate. Oleate 0 ..8 2.5~1 132-37 85.0 98 95 NI 3:1 132-40 85.0 Behenate 98 5 93 A disadvantage in using 0.25 5:l 135 91.2 Stearate 98C 95 0.25 3:l 135-40 77.0 Azelate 96 94 methyl esters is the higher visUndec ylenhte 0.25 2:1 .... 84.0 ..d .. cosity of the reaction mixture Reduction time 2 hours. Time of urea reaction 8 hours. Solvent xylene. which requires more diluent b Free f a t t y acid content calculated as oleic acid. Ammonium thiocyanate used t o decompose alcoholates, giving sodium thiocyanate as by-product. (7),because of the lower solud Run worked up by conventional method of water hydrolysis. bility of sodium methylate. The methanol complicates the solvent recovery. By bsing There is no convenient way of separating the small amount of methyl isobutyl carbinol esters both these objections are elimivaluable glycerol from its aqueous sodium hydroxide solution. nated. There is no special alcohol recovery involved as with glycThis has created a difficulty along with recent expansions of fatty erides or methyl esters, since the alcohol is the same as the reducalcohol production capacity by the sodium reduction process. ing alcohol used which is already being recovered in the normal solThe quantity of by-product glycerol-sodium hydroxide solution vent recovery operation of stripping. Data for the reduction of produced has become greater than the amount that can be utilized various methyl isobutyl carbinol-fatty acid esters are given in in soap manufacture. Table 111. Here also the two methods give comparable yields of fatty alcohol. A modification of the method of sodium reduction of esters Another device for avoiding complication of alcohol recovery is has been developed in this laboratory which overcomes these to use an ester of the alcohol to be produced-e.g., octadecyl steadifficulties and actually offers some advantages. This is the rate. These results are given in Table IV. The reaction mixture subject of a recent U. S. Patent (8).- The sodium reduction is viscosity, as well as fatty alcohol and sodium cyanate yields, run in the usual way, but the reaction mixture is not quenched in is about the same as with the methyl isobutyl carbinol esters. water as in step 2; instead, the alcoholates are decompoFed under The ease of filtration of the sodium cyanate varies considerably. anhydrous conditions by the addition of urea. Ammonia is Runs involving triglycerides appear to give an easier filtering evolved and sodium cyanate forms as in 2A: cyanate than do the other esters. Filtration is improved when CHzONa the original reduction is more complete and when the urea reacI tion is completed in a shorter time. hHONa 3 RCHzONa 6 CH&HCH2CH(CH& I 12 HgNCONHz + Several other compounds were studied as replacements for urea I CH20Na ONA in the decomposition of the alcoholates. Ammonium thiocyanate gave good results in producing sodium thiocyanate CHzOH
Table 1.
Sodium Reduction of Triglycerides
..
C
+
I
+
+
+ +
b 3 O H -!- 3 RCHzOH -I- 6 CH3CHCH2CH(CHa), 1 12 NaOCN 12 NH3 (2A) I OH CHzOH The sodium cyanate precipitates from the solvent medium and is filtered off. The filtrate is washed with water to remove glycerol and the fatty alcohol is distilled. Thus, a separation of glycerol is achieved and a valuable coproduct, sodium cyanate, produced. In this way the sodium is used twice-once in the reduction of step 1 and again in the alcoholate-urea reaction of step 2A. An additional advantage of the new process is the eliminationof the fire hazard(9) when the alcoholates are quenched in water. A disadvantage is the slower reaction with urea. The sodium cyanate is obtained in 95 to 98% yield of 93 to 97% purity. It is free of cyanide but contains 0.25% to 0.50% carbonate. The cyanate purity may be higher, since a sample of C.P. potassium cyanate assayed 95.8% by the analytical method used. The results of the application of the process t o various triglycerides are summarized in Table I. Equally satisfactory results can be obtained with other types of esters. Data for the methyl
Table 111.
Sodium Reduction of Methyl Isobutyl Carbinol Esfersa
% Yield Esterification Catalyst Fatty N?OCN Esterb Presentc alc. NaOCN Purity, yo Coconut No 88.5 d Coconut Yes 87.2 d Coconut NO 87.5 97 95 Coconut 87.6 Yes 97.5 94.5 Coconut No 88.5 Coconut NO 91.5 d Coconut 89.8 Yes d .. Stearate Yes 90.5 d Stearate 89.5 Yes 97 94:5 Stearate' 89.0 No 97 95 a Solvent xylene. X lene-ester ratio 1 t o 1. Reduction time 1 hour. Time of urea reaction 8 gours. Temperature of urea reaction 140-45' C. b First four coconut ester runs on one batch of ester. Other three were on diffoerent batch. p-Toluenesulfonio acid (0.75) was esterification oatalyst used in preparing esters. Sometimes this catalyst was removed by washing ester with water and drvinz and sometimes it was not. d Workgd up by conventional method of water hydrolysis. 8 Xylene-ester ratio 0.6 to 1. Time of urea reaction 3.5 hours, during which solvent wai removed by distillation, causing temperature t o increase from 140' to 180 Solvent added back prior t o filtering off cyanate.
.. .. .. .. .. ...
.
.. .. .. ..
1116
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Vol. 45, No. 5
understandable in view of the observations of Bader, Dupre, and Schutz ( 1 ) . These norkers obtained an almost quantitative vield of sodium cyanate of high purity, also low in carbonate and ~ \ T ~ O C N free of cyanide. by refluxing dry urea 1%-ithsodium butoxide in Purity, dry butanol. Anhydrous conditions are necessary for good % results. Thus, the reaction of alcoholic sodium hydroxide with urea produces sodium cyanate containing large amounts of car95 bonate, because water is liberated in the reaction which hydro95 lyzes a large pari of the cyanate preeent: 95 1s
Table IV.
Sodium Reduction of Higher Alcohol Estersa
y, 'FA Ester Stearyl stearate
in Esterb
Reaction with Urea % ' Yield Time, Tzmp., Fattoy hr. C. ale. KaOCN
0.28 4.5 131-43 95 97 0.28 4.5 138-45 95 97d Lauryllaurate 0.67 4.5 138-45 9.5 96 Tallow alcohol ester of tallow fatty acids 1.12 3 134-47 95. 96.5 92 n-Octyl ester of coconut fatty acids 0.76 3.5 133-44 95 95 9.5 Solvent xylene. Xylene-ester ratio 1 to 1. Reduction time 1 hour. b Free fatty acid content calculated as oleic acid. e This allows for the fact t h a t about half of the alcohol does not arise from the reduction b u t is already present in the initial ester. I n other words, a 95% yield is 1.90 moles of alcohol per mole of ester. d Prior t o filtration of cyanate xylene was added to bring xylene-ester ratio to 1 . 5 : l . This improved ease of filtration relative t o previous run. Q
HaNCONHz
+ NaOH --+ KaOCN + NH8 + HzO
Similarly, under experimental conditions when very small amounts of water were added to increase the solubility of urea and its rate of reaction, considerable carbonate was produced. Although the amount of ammonium cyanate in equilibrium with urea is small under anhydrous conditions, this is believed to be the first step in the formation of sodium cyanate ( 1 ) . As the ammonia and sodium cyanate are rapidly removed, a slow isomerization of all the urea to ammonium cyanate occurs, which controls the rate of the reaction:
H,SCOKH? (Table 111, run 3). Of course, here simple double decomposition is involved, a preliminary isomerization (see Discussion) being unnecessary. Reaction of a completed glyceryl trioleate reduction mixture with thiourea for 10 hours a t 138' C. gave only a 63% yield of sodium thiocyanate. Ammonia was evolved during the reaction of a completed tallow reduction mixture with biuret for 4.5 hours, but the yield of sodium cyanate was only about 60%. Procedure Basically, the reduction procedure used is that described by recent workers in this field (7, 9). Sodium and enough solvent (toluene or xylene) for cover are heated to reflux under nitrogen. When the sodium is sufficiently dispersed, a solution of the ester (about 200 grams) and methyl isobutyl carbinol in the remainder of the solvent is run in. The rate of addition is controlled so as to maintain good reflux. The sodium and methyl isobutyl carbinol are used in 5% excess of the calculated amounts based on Equation 1 and on the experimentally determined saponification equivalent of the ester. When addition is complete, stirring is continued for 15 minutes longer. When the reaction mixture is decomposed with water, the mixture is gradually added to the water under a protective cover of nitrogen or steam. The water is sufficient to give a 15% sodium hydroxide solution. The upper solvent-alcohol layer is separated from the lower aqueous sodium hydroxide or sodium hydroxideglycerol layer. Any emulsion is usually broken on standing and heating or by the addition of methanol. The solvent-alcohol layer is distilled. After removal of solvent the fatty alcohol is distilled a t 10-mm. pressure. The yield is calculated from the hydroxyl number, since a small amount of ester may be present. The modified method involves the following changes after the reduction is completed as described above. Dry, crystalline urea (Du Pont commercial grade) equivalent to the sodium used is added during about 15 minutes to the reaction mixture with continued stirring and refluxing. Ammonia is evolved and sodium cyanate slowly precipitates. Stirring and refluxing are continued and after several hours the reaction is completed as indicated by the cessation of ammonia evolution. The sodium cyanate i s filtered off at 80" to 100" C., washed with hot xylene-methyl isobutyl carbinol mixture, dried a t 100' C., weighed, and analyzed. Analysis involves either precipitation as silver cyanate and titration of excess silver nitrate remaining (12) or determination of the ammonia, which is liberated in acid solution (5, l a ) . Results from the two methods are in agreement. The combined filtrate and washings are washed with large volumes of water. Emulsion troubles are often encountered and acidification, followed by neutralization of the splvent-alcohol layer after separation, can be employed to minimize this. The glycerol yield is determined by titrating the aqueous solution USing a modified periodate method (IO). The washed solvent-alcohol mix is distilled as above.
Discussion The high purity of the sodium cyanate produced by the present process, involving reaction of urea with the mixture of alcoholates,
KH40CK
+ R O S a --+
NHdOCN SaOCN
+ ROH + NHs
Under these conditions it appears that the limited solubility of urea also contributes to its slow reaction. The time required for reaction TI ith urea varied considerably. With xylene as the solvent the triglyceride and higher alcohol ester runs required only 3 to 4 hours, whereas the methyl and methyl isobutyl carbinol ester runs needed 8 hours. This does not seem to be entirely explained by differences in refluxing temperature. Table I shows the advantage of increasing the reflux temperature. Xylene has about the maximum boiling point feasible when coconut oil is reduced, as a further increase prevents its separation from the Ion-er coconut alcohols during solvent recovery. High boiling solvents would probably be feasible for use with tallow, methyl stearate, etc. Use of less solvent or its removal prior to urea reaction is also helpful (Table 111, last run). Results of further studv of the factors influencing the rate of urea reaction are reported elsewhere ( 2 ) . The difficdties often encountered a i t h emulsions when glycerol is n-ashed out after the urea decomposition are probably due mainly to the presence of soap and the slight mutual solubility hetween the aqueous glycerol phase and the solvent-alcohol phase. Differences in by-product impurities and the nature of the distillation residue resulting from the two methods of decomposition should be expected. For example, small amounts of glyceryl monoesters are likely to result from the slightly incomplete reduction of a triglyceride. The strongly alkaline conditions of water hydrolysis would liberate the glycerol from this monoglyceride impurity and form soap n-hich would remain in the distillation residue. Urea decomposition vould probably leave the monoglyceride intact and some would distill with the product alcohol.
Literuture Cited (1) Bader, R., DuprB, D. J., and Schutz, F.. Biochim.et Biopiius. Acta, 2,543 (1948). (2) Barrett, F. O., Fitzpatrick, J. D., and Kadesch, R. G., J . Am. Oil Chemists' Sac., in press. (3) Bleyberg, W., and Ulrich, H., Be?., 64, 2504 (1931). (4) Bouveault, L., and Blanc, G., Compt. rend., 136, 1676 (1903); 137, 60, 328 (1903); Bull. soc. chim., (3) 31, 666, 1203 (1904). (5) Du Pont de Nemours & Co , E. I., Electrochemicals Dept., "iMolten Salt Baths," p. 62. (6) Ford, S. G., and Marvel, C. S., Org. Syntheses, 10,62 (1930). (7) Hansley, V. L., IND.Exc. CHEM.,39, 55 (1947); U. S. Patent 2,096,036 (Oct. 19,1937). (8) Kamlet, J., Ibid.,2,563,044 (Aug. 7, 1951) (9) Kastens, M. L., and Peddicord, H., IND.ENQ.CHEX.,41, 438 f 1949). (10)PoGe, W. D., and hlehlenbacher, V. C., J. Am. Oil Chemists' SOC., 27, 64 (1950). (11) Scott, N. D., and Hansley, V. L., U. 8. Patent 2,019,022 (Oct. 29,1935). (12) Williams, H. E., "Cyanogen Compounds," 2nd ed., p. 401, London, Edward Arnold and Co., 1948. RECEIVED for review September 29, 1962. ACCBIPTED January 1 9 , 1953. Presented before the Division of Industrial and Engineering Chemistry a t SOCIETY, Atlantic City, the 122nd Meeting of the AXERICAS CHEMICAL N. J.
