April 1948
,
INDUSTRIAL A N D E N G I N E E R I N G C H E M I S T R Y
A e r o ~ lplot was made to obtain the adsorption isotherms, which were plotted a t 20' C. intervals as adsorbate concentration versus the adsorption pressure. Only the isotherms of acetone on carbon A (Figure 4)are reproduced here. Another cross plot was made of the isosteres (lines of constant concentration of adsorbate on adsorbent) by plotting the logarithm of the adsorbate pressure a t constant concentration versus a temperature scale calibrated from the logarithm of the vapor pressure of acetone as a reference substance (4). The steps used were as follows: (1) A sheet of standard logarithmic paper had the temperatures indicated corresponding to ths vapor pressures of acetone taken along the X axis'. ( 2 ) Ordinates corresponding to these temperatures were drawn. (3) Values of the data of equilibrium adsorption pressures were plotted on these temperature ordinates according to the logarithmic calibration of the Y axis; and the lines were drawn connecting these points. These lines so obtained were straight for all systems (Figures 5, 7, 8), indicating the excellent mutual correlation of the data. The slopes of these isosteres were then plotted versus adsorbate concentration and the instantaneous heats of adsorption calculated from these slopes and plotted a t 50'. loo", 150",and 200" C. versus adsorbate concentra&ionin Figures 6 and 9. This series of graphs represents a complete picture of the adsorption systems investigated; a further correlation of these data is presented elsewhere ( 3 ) .
743
Examination of the data obtained and of that previously reported (4, 6) showed that the adsorption capacity as well as the heats of adsorption vary widely for different carbons as well as for different vapors. These differences are shown best b y studying instantaneous heats of adsorption curves, determined from the slopes of the isosteres (4). It is observed that the same adsorbate on different carbons may have either an increasing or a decreasing instantaneous heat of adsorption with increasing adsorbate concentration. A discussion of the adsorption mechanism to account for this phenomenon is beyond the scope of this paper. ACKNOWLEDGMENT
Thanks are due the Chemical Sales Corporation of Pittsburgh, Pa., for supporting this research and to Robert T. Weil of Manhattan College for supplying the cathetometer. LITERATURE CITED (1)
McBain, J. W., and Bakr, A. M., J . Am. Chem. Soc., 48,
690
(1926). (2) Maytin, W., Metals & Alloys, 17, 1203 (1943). (3) Othmer, D. F., and Josefowitz, Samuel, IND.ENG.CHEM.,40, 733 (1948). (4) Othmer, D. F., and Sawyer, F. G., Ibid., 35,1269 (1943). ( 5 ) Sawyer, F. G., andOthmer, D. F., Ibid., 3 6 , 8 9 4 (1944).
RECEIVED September 5, 1946. Presented in part before the Division of Industrial and Engineering Chemistry a t the 110th Meeting of,the AYERIC A N CHEMICAL SOCIETY, Chicago, Ill.
Partition Coefficients of Formaldehyde Solution d
HARRY G. JOHNSON1 AND EDGAR L. PIRET University of Minnesota, Minneapolis, Minn. D a t a for systems of formaldehyde, water, and various water-insoluble organic solvents indicate that the lower members of the series of aliphatic alcohols are the solvents best suited for the extraction of formaldehyde from solutions.. In systems involving the alcohols, a temperature investigation indicates that the partition coefficients are straight-line functions of temperature in. the range of 2' to 45" C. However, the temperature effect is not large. In the formaldehyde, water, and alcohol systems the addition of inorganic salts, soluble in water hut insoluble in the alcohols, caused a large increase in the value of the partition coefficients.
F
ORMALDEHYDE is today a most importanl commercial chemical, yet practically no data on partition coefficients for this compound have appeared in the literature in the past 40 years. The present work was carried out to provide such information. An application of these data could lead to the selection of a solvent for the extraction of formaldehyde from its aqueous solutions. These data should also find application among the many and increasing uses of formaldehyde in the synthetic resin and other indu~t~ries. In the selection of a solvent for formaldehyde the following distinction may be made between types of solvents. The solvents may be classified as inert solvents and as reaction solvents. Inert solvents are those which under normal conditions do not form chemical compounds with formaldehyde. Reaction sol1
Present address, General Mills, Inc., Minneapolis, Minn.
vents, on the other hand, are those in which some reversibIe chemical reaction can occur between the formaldehyde and t h e solvent. Included in this group are the alcohols and water. In the case of the alcohols, the formaldehyde and the alcohols combine chemically to yieId hemiacetals or acetals, depending on pH conditions. Walker in his monograph ( 7 ) points out that formaldehyde exists in aqueous solutions as an equilibrium mixture of methylene glycol, CHS(OH)Z, and several low-molecular weight polyoxymethylene glycols of the type HO(CH,O),H. Since the formaldehyde hydrates, or glycols, are relatively unstable and are formed by reversible reactions, the aqueous formaldehyde solutions may be regarded as solutions or mivtures of formaldehyde and water. Because of this compound formation, aqueous solutions of formaldehyde show solvent properties very similar to those of ethylene glycol and glycerol. Also, the solvent properties of formaldehyde solutions are somewhat similar to those of water (8). In the distribution of formaldehyde between water and a water-insoluble reaction solvent, the partition coefficient may favor the organic phase because of the formation of a chemical compound more stable than the formaldehyde hydrates. When reaction solvents are employed the extract probably consists of an equilibrium mixture of methylene glycols, water, and hemiacetals: HOCHzOH
+ C4H90H
CdH90CHzOH
+ HsO
(1)
On the other hand, the partition coefficients for formaldehyde between inert solvents and water favor the aqueous phase;
INDUSTRIAL AND ENGINEERING CHEMISTRY
744
a t ordinary temperatures formaldehyde is only slightly soluble in inert solvents such as diisopropyl ether and chloroform (6). It is probable t h a t methylene glycols rather than formaldehyde are extracted when inert solvents are employed. I n systems involving the inert solvents, one would expect, by the ordinary rules of solubility, that formaldehyde would be more soluble in the lower than in the higher members of the homologous series ( 5 ) . DETERMINATION OF EQUILIBRIUM DATA
MATERIALS.The formaldehyde used in all the experiments n as obtained from paraformaldehyde supplied by E. I. du Pont de Semours Br Company. This polymer analyzed as follon7s: Formaldehyde, % Acid as formic, % Methyl alcohol, % Ash, 7c Water, yo (by difference)
97.3 0,026 Undetectable 0 032 2.6
l l o s t of the solvents invest'igated were obtained from the Eastman Kodak Company. Three grades of chemicals T-iere usedEastman, practical, and technical. The Eastman and pract'ical grades were used as received, aft,er comparing their refractive indices with those given in the literature. The technical grade chemicals were purified by fractionat'ion a t atmospheric pressure. The boiling range of the collected fract,ions are given in Table I. The dibut,yl cellosolve was supplied by the Carbide and Carbon Chemicals Corporation. The trichloroethyl alcohol was prepared in the laboratory from a U.S.P. grade of chloral hydrate, chemically pure isopropyl alcohol, and aluminum turnings ( 1 ) . METHOD. Aqueous formaldehyde solutions were pfepared from paraformaldehyde and distilled water by refluxing unt,il the solids had completely dissolved. The 25% formaldehyde solutions prepared in this way were diluted to 2.5% for part, of t,he work. The systems .investigated were prepared by adding equal Galumes of a solvent to a st'andard formaldchydc solution in a precision Sormax glass-stoppered graduate. I n tmheexperiment,s involving the addition of inorganic salts or p H adjustment,s, a s to the graduate. The mixtures fourt,h component ~ ~ - 9added were thoroughly agitated for 48 hours by rotating the cylinders at 24 revolutions per minute in an end-over-end movement in a water bath a t a constant temperature. Preliminary experiments had indicated that t,his was sufficient time t o obtain equilibrium conditions. Except for the t'richloroet'hyl alcohol runs, all the experiments were run in duplicat,e. The results report,ed here are the average of the two tests. The equilibrium phases were analyzed for volume, density, and formaldehyde content. The volumes were read directly from the graduate, and the densities were obtained wit'h either a iT7est.phal specific gravity balance or a Sprengel pycnometer. The formaldehyde contents of the solutions were determined by the sodium sulfite method (10) except for the runs involving p-cresol and the aldehyde and ketones. The alkaline peroxide method (11) was used in the p-cresol experiment whereas the methone procedure (13) was used in the aldehyde and liet,one experiments. The formaldehyde contents were determined for both phases in the majority of the experiments. The accuracy of the formaldehyde material balances were 98.5y0 or greater, except in four experiments. I n the four exceptions side react'ions occurred; these prevented more accurat,e resulh. In those experiments in which the formaldehyde contents were found for t,he mater phases only, the formaldehyde in the organic phases were calculated by difference. RESULTS
I n order to facilitate interpretation, the equilibrium data were calculated on three bases: partition coefficient (K)defined as the ratio of grams of formaldehyde per liter of organic phase t o the grams of formaldehyde per liter oi water phase; weight % ( A ) of
Vol. 40, No. 4
the formaldehyde appearing in the organic phase, based on the amount of formaldehyde added t'o the system; and weight 70 of formaldehyde in each of the two equilibrium phases. TEIREE-COMPOYEST SYSTEMS AT 25 C. The experimental systems contained formaldehyde, water, and various solvents as t,he t'hird component. Results obtained with init,ial aqueous solut,ions containing about 25Y0 and 2.5y0 formaldehydc are given in Table I. Also included in Table I are dat,a for chloroform systems corresponding to initial aqueous solutions containing 7.15, 15.57, and 31.1270 formaldehyde. Of the compounds investigated, the alcohols appear t,o be the best solvents for the extraction of formaldehyde. Little difference can be foupd betl-ieen t,he loner aliphatic alcohols. Trichloroethyl alcohol T-ias prepared because formaldehyde is soluble in cthyl alcohol, and because the replacement, of hydrogen atoms by chlorine atoms in t,he alcohol, which renders the compound partly insoluble in wat,er, might not, great,ly reduce the solubility of formaldehyde in the chlorinated alcohol. This !vas not the case, hon-ever, and the partition cceficients for these systems correspond very nearly to the coefficient>sof the systems containing tert-amyl alcohol. Esters l-iere the nest best solvents for the extraction of formaldehyde from aqueous solutions. The n-butyl lactate contains, in addit.ion t.o the carbonyl group, a hydroxyl group. Hovever, the effect of the multiple functional groups was not, very great, compared to the monofunctional alcohols or esters. The p-hydroxy ethyl ethyl aniline, a tertiary aromatic a.mine, probably owes its solvent propert'ies to the hydroxy group. The remaining solvents investigated may be listed in the order of decreasing partition coefficients-ketones, ethers, tertiary aromatic amines, chlorinat,ed hydrocarbons, aldehydes, and aromatic nit,ro compounds. I n almost every system investigated, the partition coefficients n-ere greater for the more dilute formaldehvde solutions. This was especially noticeable in t,he alcohol series. Howcvcr, in general, t8heopposite appeared to be true in the. ketone and the chlorinated hydrocarbon systems. A distinct disadvantage of the aliphal ic alcohols as solvents for the ext,raction of formaldehyde is the amount. of T-iater which is compatible with the solvent,. The miscibility of water and the alcohols xas increased by the addition of formaldehyde. The fact was JT-ell illustrat,ed in the case of the see-butyl alcohol with t,he 25% aqueous formaldehyde solut'ion in which a single phase resulted after mixing. The solubility of sec-butyl alcohol in
30
40
50
TEMPERATURE,OC. Figure 1. Temperature Effect on Partition Coefficients for Systems of Formaldehyde, .ilcohols, and Water Concentrated formaldehyde mixtures, 0 = n-amyl alcohol, A = fusel oil; dilute formaldehyde mixtures, = n-amyl alcohol, 0 = fusel oil
+
'
April 1948
INDUSTRIAL AND ENGINEERING CHEMISTRY
74s
~~
DATAFOR FORMALDEHYDE SOLUTIONS TABLEI. EQUILIBRIUX Initial Aq. Soln., Wt. Analytical % FormalMethoda dehyde
Solvent and Gradea
K
Wt. '%, Formaldehyde Water Org. phase phase
A
Density, G./Cc. Water Org. phase phase
Error in Formaldehyde Material Balance, %
Concentrated Solutions, 25' C. I
Alcohols n-Butyl, P sec-Butyl, P n-Amyl, P Fusel oilc, T sec-Amyld, P tert-Amyl P n-Hexyl ' P 2-Ethyl'butyl, T (145O to 14G0 C.) n-Octyl E Trich1o;oethyl (148' to l55O C.) Cyclohexyl, P Benzyl E GlycerAl or-?-dichlorohydrin. P p-Cresol, P -4ldehydes and ketones n-Butyraldehyde, E Methyl ethyl ketone, E Methvl amyl ketone, E Amines' Dimethyl aniline, P &Hydroxy ethyl ethyl aniline, P Chlorinated hydrocarbons Chloroform, T (60' to 61' C.) Chloroform T (60' to 61' C.) 1,1,2-Trichlbroethane, T (111' to 1120 C.) Esters and acids n-Butyl formate, E n-Butyl lactate, E Oleic acid, P Ethers Diisopropyl ether, T (67' to 69' C.) Dibutyl Cellosolve Nitro compound Nitrobenzene, P
24.88 24,88 24.70 24.70 24.88 24.88 25.31 24.88 25.42 24.88 24.88 25.37 25.42 24.70
83.5. 1.88 Single phase 73.7 1.67 74.9 1.75 0,874 55.3 80.7 0.851 67.6 1.39 1.52 68.5 1.06 61.6 0.693 54.0 0.952 60.8 78.3 1.78 0.832 58.8 0.493 49.5
16.61
8.06
....
0.9207
....
1,0120
....
co.2
37:20. 17.62 14.12 13.68 17,Ol 17.20 16.07 8.24 13.31 15.06 9.61 9.394
9.07 8.80 13.81 15.16 10.60 9.85 12.51 15.31 13.33 8.86 13.44, 18.90
0.8983 0.8938 0.8858 0.9461 0,8900 0.8947 0.8743 1,4030 0,988 1.0713 1,2810 1.0480
1,0230 1.0214 1.0363 1.0010 1,0289 1,0280 1.0382 1.0897 1.0359 1.0266 1,1019 1.0559
-0.6 +0.2 -0.8 -0.9 -0.9 -0.6
.. 2
24,75 24.70 24.75
0,00441 0.387 Single phase 0.0338 3.38
0.132
24.19
0,8512
1,0508
+0.7
1.os
24: 23
0.8150
1.0719
-0.9
3 1
24.70 25.42
0.00655 1. O l
0,653 60.2
0.181e 12.79
24,61 13.02
0,9550 1.067
1,0733 1.0387
1 1
25.13 31.12
0,0126 0.0141
-0:e
1.23 1.41
0.227 0.321
24.70 30.73
1 ,4796 1.4780
1 ,0792 1.0969
+0.7 +0.6
1
24.88
0,0210
2.06
0.380
24.29
1.4428
1.0773
-0.5
24.70 25.42 25.42
0.327 0.638 0.0105
24.9 54.4 1.05
7.26 11.22 0.318e
18.87 17.00 25.20
0.9038 1.0184 0,8990
1.0634 1.0536 1.0746
-0.1 -0.2
1 1 3
1 3
24.88 24.70
0.00856 0,0220
0.817 2.15
0.305 0.6806
24.18 24.20
0.7286 0.8379
1,0717 1.0720
-0.6
1
24.70
0
0
0
24.70
1.1968
1,0748
1
.. 1 1
1 1 1 1 1 1
1 1 1
4
2
....
....
....
...
i-0.1 +0.6 +0.6 +0.9 -3.5
...
...
...
...
0.0
Dilute Solutions, 25' C. Alcohols 0.568 0,9906 2.060 -0.6 80.6 2.455 3.13 n-Butyl 1.150 0.9745 1.534 64.1 1.21 +0.7 2,455 sec-Butyl 0.598 2.234 0 . 9985 7 8 . 5 +0.3 2.533 3.15 n-Amyl 0.567 2.189 0.9960 78.4 -0.2 3.23 2.455 Fusel oil 1.207 0,9972 1.481 -1.2 51.9 2 455 1 . 0 3 sec-Amyl 1.517 1.184 44.7 0.9884 4-0.2 0.686 2 455 tert-Amyl 0.621 1.0019 2.101 76.8 4-0.4 2.84 2 455 n-Hexyl 0.593 2,167 1. 0000 77.0 3.09 1-0.4 2.455 2-Ethyl butyl 0.765 2,069 1.0020 71.4 ro.1 2.26 2 533 n-Octyl 1.586 0,641 41 . O 1.0417 0.555 $0.8 2.455 Trichloroethvl 1.203 1.0000 1,3368 54.3 1 . 0 6 2.455 Cyclohexyl 0.562 1,810 1.0028 80.4 +O:S 3.35 2.533 Benzyl 1.402 0.797 1.0526 43.7 -3.0 0.700 2.533 Glycerol a-y-dichlorc)hydrin 2.290 0.477~ 1.0058 23.4 0.213 ... 2.455 4 p-Cresol AJdehydes and ketones 2.329 0.8164 0,492 0.0164 1.0002 -0.8 0.00575 2.485 2 n-Butyraldehyde 0.396 1.781 0.8330 0.9645 -1.5 8.97 0.192 2 2.485 Methyl ethyl ketone 0,0891 2.420 0.8126 1.0049 -0.3 2.96 0.0298 2.485 2 Methvl . amyl . ketone Amines 2.533 0,0221 2.16 0.0576e 2.481 0.9546 1,0070 ... Dimethyl aniline 3 0,900 1.042 1,0012 -0.7 &Hydroxy ethyl ethyl aniline 1 2,533 1.78 65.8 1.533 Chlorinated hydrocarbons 0.0134 2,509 1.4780 1.0060 -1.2 0.768 0.00785 1 2.559 Chloroform 7.035 1.4780 0,806 0.0398 1.0210 -0.5 0,00819. 1 7.150 Chloroform 15.32 1.4806 0.965 0.107 1 ,0465 -0.6 0.00989 1 15.57 Chloroform 2.516 1.4432 0,115 0,00203 1.,0068 +0.2 0.00116 1 2.513 1,1,2-Tl?chloroethane Esters and acids 1.009 0.8894 1.0087 -0.1 57.9 1.644 1.44 1 2.455 n-Butyl formate 1.377 0.9894 52.9 1.211e 1,0059 ... 0.865 3 2.533 n-Butyl lactate 0.0182e 2.520 0.8980 1.0065 ... 0.640 ,0.00645 3 2.533 Oleic acid Ethers 2.387 0.7260 1.0066 ... 5.20 0.1816 0.0547 3 2.513 Diisopropyl ether 2.358 4.04 0.1196 0.8350 1.0059 ... 0.0419 3 2.455 Dibutyl Cellosolve Nitro compound Nitrobenzene * 1 2.533 0 0 0 2,556 1.1995 1.0066 -0.6 a E = Highest in purity; P = chemicals considered of sufficient purity for usual laboratory synthesis; and T = redistilled high grade commercial chemicals. b 1 = sodium sulfite method for both phases; 2 = 'methone method for both phases; 3 = sodium sulfite method for water phase only; 4 = alkaline peroxide method for water phase only. 0 Mixture of amyl alcohols consisting chiefly of isobutyl carbinol and sec-butyl carbinol. . d Mixture of sec-n-amyl alcohols. 8 By difference. 8
.
I
TABLE11. TEMPERATURE EFFECTO N DISTRIBUTION OF F O R M A L D E H Y D E B E T W E E N TvATER AND A M Y L ALCOHOLS'
Temp, O
c.
2
25 0 45.5
!g
Concn
24 70
n-Amyl Alcohol K Conon
yi
1 67 1 56
'
K
COncn.b
3 15 2 81
24 70 24 70
2;; 3" E! ig $;
2 533 2 456
Fusel Oil Concn
t!
1 76 1 60
253; 3"
2 455 2 455
3 24 2 86
a Determination of formaldehyde in both phases b y sodium sulfite method. Error in formaldehyde material balances &I1 less than 1%. b Weight % formaldehyde in initial aqueous solution.
mater at 20" C. is 12.5 grams per 100 cc. The phenomenon is also illustrated in the system of methyl ethyl ketone, water and formaldehyde. EFFECT OF TEMPERATURE AND pH. As temperature can have an important effect on solubility, the partition coefficients for the amyl alcohols were determined at 2', 8.5", 25', and 45.5" C. The data are shown in Table I1 and in Figure 1. Although a decrease in temperature favors a better extraction of formaldehyde from the aqueous solution, the temperature effect is shown t o be small. I n the reactions of aliphatic alcohols and formaldehyde the
INDUSTRIAL
746
AND ENGINEERING CHEMISTRY
Vol. 40, No. 4
The data show that,, as expected, the addition of salts had a markedly favorable infiucnce. Both salts caused a n increase in the partit,ion coefficients’ However, the amounts of formaldeError in Initial hyde found in the alcohol phases a t equilibrium were increased Aq. S o h , 1 3 H of Formaldehyde Material W t . yo Equ1l. A q . by only 5 to 10% over the amounts in the systems in which no Formaldehyde Phase K A Balance, % salt was present. Salts appear to push the equilibrium of rc1.66 68.8 -6.5 24.70 action (Equation 1)t o the right by removing water. The greatest 24.70 0 .’55 1,65 68.5 -5,s 3.5 1.67 73.9 -0.8 5 change appeared in t,he equilibrium volumes bemuse the presence 24.70 3 . 6 1 . 6 6 7 3 . 5 24.70 24.70 7.0 1.73 75.8 +0.5 of the salts decreased the mutual solubilit,jes of water and the 24.70 7.1 1.72 75.2 alcohols. 24.70 10.0 1.67 z3.3 -- 01. 1. 2 d3.0 -1.5 The addition of an increased amount of sodium sulfate had a 24.70 10.2 1.66 a Determination of formaldehyde in both phases by sodium sulfite net favorable effect even method. though the higher temperature, as shown in Figure 1, TABLEI F T . ADDITION O F S O D I C 1 1 CHLORIDE AND SODIUhl SULFATE T O SYSTEMS O F tends t o reduce the solubility FORMALDEHYDE, b‘TaTER, AND n-BuTYL .ALCOHOLa of formaldehyde in the solvent Dilute Formaldehyde Solutions Concentrated FormaIdehyde Solutions phase. 24.88 25.37 25.42 23.42 2.456 2,533 2.533 2.533 InitiFl aq. s o h . , % CHzO
O F FIORMALDETABLE 111. EFFECTOF p H 0 3 DISTRIBUTION HYDE BETWEEN WATERAXD n-dxur, ALCOHOL^
Equil. temp., C. Salt added Ratio wt. salt t o wt. CHz0
K
a
CHzO in org. phase, wt. yo C H 2 0in water phase, W’t. 74
25
23
25
None 0
KaC1
NapSOi
0.915 1.88 3.12 83.0 81.8 16.61 20.08 8.06 4.80 0.9207 0.8952 1.0120 1.2000
0,740 3.11 84.4 19.45 4.54 0,9039 1,2459
47 NanSOi 1.08 3.20 85.1 19.45 4.28 0.9027 1.2840
25
25
25
47
Xone 0
KaCl
NazS04
h-asS04
3.13 80.6 2.060 0.568 0.8552 0,9906
11.3 5.82 85.5 2.442 0.297 0,8326 1,1780
7.84 4.87 85.9 2.319 0,340 0,8410 1.1789
16.7 6.03 86.7 2.372 0.250
COMPARISON WITH PREVIOUS WORK
The partition coefficients for only three systems are re1,3190 corded in the International 2.68 1.47 1.82 1.73 1.29 1.04 1.23 1.11 Critical Tables. These are fO.2 t0.2 -0.6 f0.4 +0.6 -0.4 +0.2 -0.3 for the systems of formaldea Determination of formaldehyde in both phases by sodium sulfite method. hyde, water, and the solvents diethyl ether ( 2 , 4 ) , chloioform (3, 4 , and n-amyl alcoTABLE 1’. ADDITIOPi O F S O D I T 3 1 C H L O R I D E A K D SODIUM SLTLFATE T O SYSTEMS O F F O R M A L D I ; H Y D h , WATER,. ~ U Dn-XiuYL ALCOHOL, 25‘ C.“ hol (3, 4 ) . P a t a for five difDilute Formaldehyde ferent mixtures of chloroform, Concent Solution ~ a t e r and , formaldehyde and 2.533 2,533 2.533 24.70 Initial aq. s o h , wt. % CHzO none NaCl NanSOa. Sone for two different mixtures of Salt added 0 11.6 8.42 0 Ratio wt. salt to wt. CHz0 n-amy1 alcohol, water, and 3.15 4.87 4.25 1.67 K 78.5 8 3 . 1 8 3.3 7 3 . 7 A formaldehyde u ere determined 2.234 2.420 2.354 17.20 CHzO in org. phase, 1st. % 0.598 0.350 0.391 9.07 CHzO in water phase, wt. % in the present work. The 0.8407 0.8312 0.8355 0.8983 Density org. phase, g . , / C C . agreement of the data of this 0.9985 1.1814 1.1854 t w nhasp. e./cc. 1,0230 Density ma.-. 1.14 0.990 1.11 1.72 Val. ratio org. phase’t; ‘kater phase work with the data of Herz +0.3 f0.4 +1.0 -0.6 Error in formaldehyde material balance, % and Lewy for the systemb a Determination of formaldehyde in both phases by sodium sulfite method. including n-amyl alcohol i i satisfactory I n the chloroform systems, however, the agreement is poor. Disagreements with early data on formalpH of the solutions is a n important consideration. Mineral dehyde extraction are not surprising since the sohtions used acids favor acetal formation, whereas under neutral or alkaline in early physical studies often contained as much as 104.4, conditions hemiacetals are the reaction products (9). The p H methyl alcohol. of aqueous fofmaldehyde solutions is about 3.5. Sulfuric acid was added t o adjust the pH below this value, and sodium hydroxide was used t o obtain the neutral or alkaline solutions. ACKNOWLEDGXIENT The effect of the pH on the equilibrium conditions in n-amyl alcohol systems is s h o m in Table III .shere the p H values are The authors wish t o express their appreciation to J. F. Ka1kc.i given for the equilibrium water phases. for several helpful suggestions made in this york. At a p H in the range 3 to 7, a formaldehyde material balance with 99% accuracy could be obtained. Within this range no LITERATURE CITED appreciable effect of pH x a s noted, Outside this range good material balances were not obtained. At a lower p H the forma(1) Chalmers, W., Org. Synthesis, 15,80 (1935). (2) Hant,zsch and Vagt, 2 . physik. C h e m . , 38, 705 (1901). tion of acetals occurred and a t a higher p H the Canniszaro re(3) Herz, W., and Lewy, M.. Jahresber. Schlcs. Ges. vater. Kulful-, action took place. Naturw. Sekt., p. I (1906). ADDITIOW OF INORGANIC SALTS. T o determine the effect of (4) International Gitical Tables, T’ol. 111, p. 422, New York, Meinorganic salts on the equilibrium conditions, salting-out experiGraw-Hill Book Go., 3928. ( 5 ) Kamm, O., “Qualitative Organic Analysis,” p. 9, Ne-tv York, ments were conducted. The salts investigated were sodium John Wiley & Sons, 1932. chloride and sodium sulfate in solutions containing n-butyl and (6) Walker, J. F., “Formaldehyde,” p. 25, New York, Reinkold n-amyl alcohols. The effect oi a larger amount of sodium sulfate Publishing Corp., 1944. on the equilibrium was determined by an experiment carried out (7) I b i d . , p, 29. ( 8 ) Ibid., p. 56. a t 47” C. for a system including the components sodium sulfate (9) I b i d . , p. 135. and n-butyl alcohol. The amounts of salt used in these experi(10) I b i d . , p. 257. ments were the quantities t o give essentially saturated aqueous (11) Ibid.,p. 259. phases. The detailed data of the n-butyl alcohol systems are (12) Ibid., p. 263. given i n Table IV, and the data for the n-amvl alcohol systems are given in Table V. RECEIVED September 29, 1947 Density org. phase, g . / c c . Density water phase, g . / c o . vel. ratio org. phase to water phase Error in fornialdehyde inaterial balance, %
~
~
~
0.8381