2832 LEDBURY AND BLAIR THE PARTIAL FORMALDEHYDE CCCXC1.-The Partial Formaldehyde VUPOUT Pressures By WILFRID LEDBURY and ETHELBERT WILLIAM BLAIR. THE investigation of the partial formaldehyde vapour pressures of aqueous formaldehyde solutions described in Part I (this vol., p. 26) showed that following a prolonged initial exposure a t 15", equilibrium between the liquid and rapour phases was only gradually attained when the temperature of a solution was subsequently changed to 20" or 0". An explanation of this was based upon the assumption that a new equilibrium between complex (polymerised, hydrated etc.) and simple molecules in solution depending on the temperature was gradually approached as time elapsed. The presence in aqueous formaldehyde solution of both the polymerised and monomolecular forms is made evident by the work of Auerbach and Barschall (Arb.Kais. Ges.-A. 1905 22 584; Chm. Centr., 1905 II 1081) who carried out molecular-weight determinations by the cryoscopic method. These determinations indicated that in solutions containing 37-38% (by vol.) of formaldehyde there probably exist polymerides of complexity greater than trimolecular . Auerbach and Barschall concluded that the equilibrium between the different molecular forms of formaldehyde is reversible so that the condition of aqueous solutions of formaldehyde a short time after preparation depends onIy on the concentration and temperature. In order to study further the influence of temperature and solution concentration on the partial formaldehyde vapour pressures of aqueous formaldehyde solutions determinations have been carried out a t 35" and 45" and these new data correlated with those pre-viously obtained at 0" and 20".It has been previously pointed out by the authors that methyl alcohol when present in small amounts (compare the proportion of methyl alcohol to formaldehyde in commercial formalin) con-siderably enhances the partial formaldehyde vapour pressures of formaldehyde solutions a t 20". Formaldehyde in methyl-alcoholic solution and in a mixture of water and methyl alcohol containing a large proportion of the latter has since been investigated in this connexion and also the effect of the presence of small amounts of the alcohol in formaldehyde solutions a t 35" and 45". of Apww Solutions of Formaldehyde.Part 11. E x P E R I M E N T A L. Method and Appratus.-Aqueous solutions of formaldehyde free from methyl alcohol were prepared either by the method of con-tinuous refluxing previously described or by dissolving solid para VAPOUB PBESSUBES OF AQUEOUS SOLUTIONS ETW. 2833 foddehyde uncontaminsted with methyl alcohol in boiling water. The solutions 80 prepared were kept for 2 or 3 weeks at room temperature ; it was then assumed that the equilibrium between the difFerent molecular species of formaldehyde in solution had attained a value dependent only on the concentration and temperature of exposure irrespective of the method of preparation (Auerbach and BarscI.mIl loc. cit. ; Auerbach and Pliiddemann ibid. 1914,47,116). To ascertain the effect of the presence of small amounts of methyl alcohol on the partial formaldehyde vapour pressures of the solutions, the alcohol having an acetone content of less than O.lO% was added to a series of the solutions in such amounts as to provide a constant ratio CH40/CH,0.For the preparation of formaldehyde in methyl-alcoholic Solution a current of dry nitrogen was passed over para-formaldehyde in a silica tube at 170". Formaldehyde vapour, cam& forward by the gas-stream was absorbed in methyl alcohol, as free aa possible from acetone. The solution was filtered from any insoluble polyoxymethylenes and kept Beveral weeks before use. This solution has an exceptionally irritant action on the membranes of the eyes and nose. Unlike aqueous formaldehyde, it produces this effect after a minute or so.The formaldehyde content (p. 15%) of this solution was estimated both before and after refluxing with dilute sulphuric acid and the Uerence in the values showed that the equivalent of about 5 g. of formddehyde per 100 C.C. was combined as methylal. When a methyl-alcoholic solution of formaldehyde was prepared by heating an excess of pure paraformaldehyde with methyl alcohol in a seded tube for 6 hours at 120" the major portion of the aldehyde in solution was combined as methylal. The free formaldehyde content was 6.63 g. per 100 c.c. whilst the methylal present waa equivalent to 12-0 g. per 100 C.C. Refluxing paraformaldehyde with methyl alcohol at the ordinary pressure for 2 days yielded a solution con-taining. 5.16 g. of free formaldehyde and 144 g.of combined formaldehyde per 100 C.C. As in the corresponding determinations at 20" and 0" the " dynamic " or " flow " method was employed for the series of experiments at 35" and 45". The apparatus used was similar to that described in Part I except that a heating unit of 500 watts was substituted for the small carbon-filament lamp. Since accurate estimations of small amounts of formaldehyde are possible by the use of the iodometric method of Romijn (Anatyst 1897 22 221; see also Ckm.-Ztg. 1901 25 74; Ber. 1898 31 1979; 1901 34, 2817) it was necessary to pass volumes of air of only 2 or 3 litres a t the most for each vapour pressure determination. During the course of a run involving several successive determinations on th 2834 LEDBURY AND BLAIR THE PAR- FORMALDEHYDE same solution the alteration of the strength of the formaldehyde solution in the " carburettors " was thus made negligible even in the case of concentrated solutions at 45" (compare Part I).De Waal ( P k m . WeekbZad 1907,44,1207; J . Phcbrm. Chim. 1907,26,498) has shown that in presence of air increase of temperature promotes oxidation of formaldehyde in solution but that this is slight even at 50" after an exposure for 400 hours. In the present series of determinations air was passed at the rate of about 1 litre in 3 hours. All estimations of formaldehyde and methyl alcohol were carried out 88 previously described and a similzr procedure of experi-mentation was followed. F d b y & Vapour Pressures of Aqueous Fumxddehyde Solutions For solutions of formaldehyde not containing methyl alcohol, the vapour pressure values at 35" obtained after the passage of several litres of air (for successive determinations) were slightly hqgher than those obtained at the outset.This increase however, wit8 not nearly so marked as at 20° and the eventual steady maxi-mum was more rapidly reached. Solutions containing methyl alcohol each gave a series of practically identical values from the beginning and this was also the case with both sets of solutions at 45". It is evident that at the temperatures under consideration, the new equilibrium between the various molecular species in solution is attained much more rapidly than at 0" or 20". The equilibrium values of the formaldehyde vapour pressures for form-aldehyde solutions free from methyl alcohol at 35" are in Table I, and the corresponding values for solutions containing methyl alcohol (CH,O/CH,O = 0.13) in Table II.(In Tables I-V g = grams of formaldehyde in 100 C.C. of formalin solution mt = mg. of formaldehyde vapour in 1 litre of issuing air at to and p = partial pressure of formaldehyde vapour in mm. of Hg.). Curves A and B in Fig. I represent the relationship between concentration of solution and partial pressure at 35' for pure solutions and for solutions containing methyl alcohol respectively. at 35" and 45". TABLE I. g ............ 1.09 5.15 11.8 18.6 20.8 31.0 39.5 T I Z ~ ~ ~ ......... 0.27 1.13 2-06 2-87 3.17 4.27 4.58 p ............ 0.166 0.695 1-29 1-80 1-94 2.48 2.81 TABLE 11.g (CH,O/CH,O = 0.13) ......... 5-24 13.05 23.7 38-3 ms6o ................................. 1.200 2.52 3.97 5.39 p ....................................... 0-735 1-56 2.45 3-3 vmom PmssmEs OF AQUEOUS SOLUTIONS ETC. 2835 The data obtained from solutions at 45" are in Tables I11 a;d Iv, and the curves showing variation of formaldehyde vapour pressure wifh solution concentration are plotted in Fig. 2. TABLE 111. g ....................................... 10.8 20.4 28.75 m4; ................................. 3-77 6:17 7.70 p ....................................... 2.30 3.79 4-72 TABLE IV. g (CH,O/C&O = 0.13) ......... 9.76 19.15 27.0 m450 ................................. 3-88 6.62 8-62 p ....................................... 2.39 4.07 5.3 FIG.2. 4.0 :::E 1.0 0 1 1 j v i I 50 20-0 30.0 40.0 10.0 20.0 30.0 40.0 G m . of formaldehyde per 1OOc.c. of eoh4timt. A methyl &hot ob8mt. B CH401CH,0=0.13. Variation with Ternperdure of the Partial Formaldehyde Vapou~ Pressures of Aqueous F d d e h y d e Solutions. From data derived from the curves drawn to represent the variation of partial formaldehyde vapour pressure with solution concentration at particular temperatures vix. 0" 20" 35" and 45", the graphs of Fig. 3 have been plotted to illustrate the manner in which the partial pressures change with temperature. The latter curves are plotted for solutions containing respectively 5 10 20, 30 and 40 g. of formaldehyde in 100 C.C. It is apparent that the differences between the partial pressures of the more concentrate 2836 LEDBUBY BND B= THE P ~ T I A L FORMALDEHYDE solutions at a given temperature are less marked than in the case of the weaker solutions.The data from molecular-weight deter-mimtiom the low values of the partial formaldehyde vapour pressures at the temperatures under consideration and the gradually dimini8hing dplclg of the vapour pressure-concentration curves demonstrate the increasing preponderance of polymerides in formalin solutions with increasing concentration. The relatively high FIG. 3. Temperature. concentrations of polymerised formaldehyde in the stronger solu-tions explains the particular characteristic of the curves referred to above. When the logarithms of the partial formaldehyde vapour pressures (log p ) are plotted against the reciprocals of the absolute tem-perature ( l / T ) for solutions of given strengths a series of almost parallel straight lines is obtained.In Fig. 4 the curves indicate the straight-line relationships between log, p and 1/T for solutions having formaldehyde contents of 10 20 30 and 40 g. per 100 c.c. VAPOUB PBESSUBES OF AQUEOUS SOLUTIONS ETC. 2837 respectively. Thus the formaldehyde partid vapour pressure of my aque~us formaldehyde solution up to a 40% concentration can be expressed for the temperatures under consideration by 8 simpMed form of Rankine's equation (Edinburgh New Phil. J., July 1849) connecting vapour pressure md temperature : log, p = a - B/T, where ale% = 9-47 ~ ~ 2 0 % = 9-70 &30% = 9.81 a@% = 9.87 (approx.) and p (for all solutions up to a%) = 2905 (approx.).A efraight-line relationship is somewhat deviated from when loglop for any given solution strength is plotted against T. The formula log p = a + pT + yT2 + . . . . which Perman (J. 1903, FIG. 4. 0*00370 0.00360 0*00350 F; 0*00340 040330 040320 0.003 10 88 1168) successfully applied to aqueous solutions of ammonia, was found to be inapplicable t o aqueous solutions of formaldehyde. blethyl-alcoholic Solution of FormuZd&y&.-Prior to the determin-ation of the partial formaldehyde vapour pressure at 20" of a methyl-alcoholic solution (15.9 g. CE,O per 100 c.c.) blank determin-ations were carried out on a sample of the methyl alcohol employed, since although this alcohol was the purest obtainable it waa not quite free from acetone (less than 0.10%).At 20° the blank values were equivalent fo about 0-1 mm. of formaldehyde estima-tions being carried out iodometrically. Further vapour-pressure determinations have been carried out at 20" and 0" on formaldehyde in methyl alcohol-water solution the alcohol being present in preponderating amount. This latter solution was obtained b 2838 LEDBTJRY AND BLAIR THE PARTIAL FORMALDEHYDE diluting with water the original 15% methyl-alcoholic solution in order to obtain a liquor containing 10 g . of free formaldehyde per 100 C.C. Blanks were performed at 20" and 0" on a mixture of methyl alcohol and water in which the proportions of the two components were the same as in the formaldehyde solution examined. The data obtained are in Table V solution A contained 15.97 g .of free formaldehyde (equivalent of 5.0% formaldehyde as methylal) in 100 C.C. of methyl-alcoholic solution; solution B contained 10 g. of free formaldehyde 59 g. of methyl alcohol and 26.7 g. of water in 100 C.C. TABLE V. Solution. t . mp. P. A ................................. 20" 3-68 2-27 B ................................. 20 1-93 1.16 B ................................. 0 0.390 0.234 It will be seen that at 20" the partial formaldehyde vapour pressure of an approximately 15% solution of formaldehyde in methyl alcohol is almost five times the value of that for an aqueous solution of corresponding strength. For a solution containing about 60% of methyl alcohol there is an approximately threefold increase at both 20" and 0" over an aqueous solution containing a corresponding amount of formaldehyde (10 g.per 100 c.c.). In all cases practically constant vapour pressure values were obtained from the outset. Since methylal vapour carried forward from the methyl-alcoholic solution to the absorption worms does not affect the iodometric estimation of the formaldehyde fixed by the water (Bergstrom, J. Amer. Chem. Soc. 1923,45 2150) the figure obtained represents the partial vapour pressure of uncombined formaldehyde in methyl-alcoholic solution. If it be assumed that the methylal (which is present to the extent of the equivalent of 5 g. of formaldehyde per 100 C.C. of 15% methyl-alcoholic solution) does not exert a very pronounced influence in its solvent capacity on the partial vapour pressure of the free formaldehyde in solution then it is found that the application of the simple mixture rule to a 15% aqueous solution in which CH,O/CH,O = 0.13 by no means accounts for the enhanced value a t 20" brought about by the addition of the alcohol to the original aqueous formalin i.e.from 0.49 to 0.59 mm. Hg. It has been pointed out that the removal of methyl alcohol from commercial formalin (40% by vol.) increases the tendency of para-formaldehyde to separate from solution. Experiments have shown, however that paraformaldehyde is more readily soluble in water than in methyl alcohol ; thus the precipitation of paraformaldehyd VAPOUR PRESSURES OF AQUEOUS SOLUTIONS ETC. 2839 from formalin which has been freed from methyl alcohol cannot be attributed to the removal of a constituent having a greater solvent capacity for formaldehyde.It therefore appears that the presence of methyl alcohol in aqueous formaldehyde solutions must bring about an alteration of the equilibrium between simple and complex molecules in such a way as to enhance considerably the concentra-tion of mono-molecular formaldehyde. The consequent decrease in the concentration of the polymerides naturally decreases their liability to precipitation as paraformaldehyde. Such an effect on the above-mentioned equilibrium serves to explain why methyl alcohol present in small amount in an aqueous formaldehyde solution increases the partial formaldehyde vapour pressure to an extent beyond that anticipated from the application of the simple mixture rule.It is here assumed as hitherto that the formaldehyde vapour pressure is almost entirely dependent on the concentration of the simple molecular form of the aldehyde. Xummar y . The partial formaldehyde vapour pressures of aqueous solutions of formaldehyde free from methyl alcohol have been determined by the dynamic method at 35" and 45". Corresponding determinations, carried out on solutions of formaldehyde containing small amounts of methyl alcohol (CH,O/CH,O = 0.13) have shown that the influence of the alcohol in enhancing the vapour pressures is similar to that noted at 20". The effect of increased temperature in bringing about a more rapid adjustment of equilibrium conditions has also been shown. From the vapour pressure values a relationship connecting functions of the vapour pressure and the temperature has been derived. The partial formaldehyde vapour pressure of a methyl-alcoholic solution of formaldehyde has been determined and the fiuence of methyl alcohol on the formaldehyde vapour pressures of aqueous formaldehyde solutions considered. The authors desire to express their thanks to the Department of Scientific and Industrial Research and to the Admiralty for permission to publish these results. ROYAL NAVAL COBDITE FACTOBY, HOLTON HEATH. [Received S e w e r lab 1926.