Solubility of CO2 and N2O in Certain Solvents

512 WILLIAM Kl/NERTH. ‘ 3:23;: SOLUBILITY OF C02 AND N10 IN CERTAIN SOLVENTS. Ev WILLIAM KUNERTH. SYNOPSIS. Solubility 42/ car and Nro in Twelve satmir. is“ la 16" c.—since according to the Lewis-Langmuir theory these two gases have similar molecular structures, it is at interest to compare their solubilities in various liquids. In the method adopted. the air was thoroughly removed iroi-n the solvent by hoi ng and then the gas to be nested, having been asreiully purified with the help oi liquid air. was admitted and shalren up with the solvent until no further solution took place. ohservations accurate to better than one per cent. were made (or water. omm. antic acid. methyl azmhoi. pyridine. elhyl alcohol, benzuldehydz, aniline, nmyl atzlalt. zlhylenz bromide. isoamyl alcohol. and chloroform. Taken in this order. the ratio oi the solubility or C0. , to that or me decreases regularly iron. [.34 (2o°) ior water to 9.66 ior chloroiorrn. This range or variation is small. and moreover the ratio is nearly constant {or each solvent, changing less than one per cent. (or six solvents, and not more than three per cent. for the others exoesn chloroionn and acetone. Also. the oernperuzure eoenicient (ds/aim is in most cases nearly the saline ior the two gases. It is always negative. the solubility decreasing with increasing tempemture. Discussion of Snggzxlzd Solubility Reinnons, for cases in Liguziis.—Ruouzx's law does not hold ior the solubility at gases in liquids. It is also shown that there is lime, ii any, relation between solubility and the diiiereuoe hetween the internal or tuhesvan prersuvex oi solvent and solute. However, the ratio oi the solubilities oi cor and N20 varies regularly wins the dielzclric cunshml oi the solvent. and since this constant may be taken as an index or the polarity ol the solvent, and since co. is more active chemically and chereiore has stronger polarity than N10. this result suggests that polarity may he an important factor in determining the relative solubility oi gases in liquids. INTRODUCTION. N view of the great interest in molecular structure at this time and in view of the possible relation between it and solubility it has seemed of interest and importance to get more complete and more accurate data on certain types of solubility than have hitherto been available. In the work which is here to be described the solubility of carbon dioxide and of nitrous oxide in twelve different liquids was determined over a range of 18° C. These gases were used because according to recent theories of molecular structure‘ they have the same number and arrangement of electrons. They should therefore exhibit little if any difference in solubility, if the latter depends only on the number and arrangement of electrons. If they do show difierence then, according to these theories, it must be due to other properties and the magnitude of the dilference should be of interest for any theory of solubility. I Koseell, Ann. der Physik. 49. p. :29 (1916). Lewis. Jonr. oi Amer. chem. Sac.. 33. p. 762 (1916). Langmuir. Jour. oi Amer. Chem. soc. or. o. 568 (1929). Thomson. Phil. Mag.. 37.1:-419 (mp). ‘,jgf>s?“x-] SOLUBILITY op co, AND M0. 513 APPARATUS. The apparatus used for this work was very much like that used by McDaniel.‘ The chief part of it is shown in figure I. It consists of a . gas burette A (vol. 120 c.c.) which is graduated to o.5 c.c. and can be read to 0.1 c.c., and an absorption pipette B (vol. 31.3 c.c.) Connected by a glass capillary C. The whole forms one solid piece of glass and is clamped to a wooden frame (not shown in the figure) which is held in an upright position on an iron rod near the burette. With this rod as a pivot the apparatus can be very thoroughly shaken when solution of the gas in the solvent is to take place in the pipette. Eiiiay R 5/3%” Fig. r. At each end of the capillary there is a three-way stopcock (D, E). The burette is provided with a water jacket F. an electric heating coil G, and a compressed-air inlet H. The reservoir I connected with burette A by a rubber tube and containing mercury makes it possible to keep the gas in A at atmospheric pressure. The pipette is likewise provided with a water jacket J, an electric heating coil K, and a compressed-air inlet L. The coils G and K are of Chrome] A wire and either one may be used alone or they may be connected in series. They are so adjusted in length that the resistances are to each other as the respective radiating surfaces of the two water jackets. This insures an equal rate of temperature increase in the two vessels when the coils are in series. ‘McDaniel. Iour. oi Phys. Chem.. :5. p. 587 (rgu). 5 I 4 WILLIA M K UNERTH. §§§‘,‘;‘;f Two thermometers M and N can be read to 01° C. and are used to indicate the temperature of the gas in the burette and pipette respectively. 0 is a mercury»seal stopcock, and P is a boiler of pyrex glass. W is an exit for the air which has been bubbled through the water in J. Com- pressed air is also bubbled through F to stir the water and hence assure the same temperature at top and bottom. Stopcock E is carefully ground to prevent leakage, and graphite is used as a lubricant on it and also on the mercury-seal stopcock 0. Grease could be used on neither E nor 0 because of the organic solvents used in this work. ORDER OF PROCEDURE. In making any determination the burette A is filled with mercury except that a few drops of the solvent to be used are put in at the top. After the air has all been pumped out of the tube RE and it has been washed out with the pure gas which is to be used in this work, cock D is so adjusted as to allow the pure gas to displace the mercury in the burette. Part of the solvent placed in the burette evaporates and thus saturates the gas with the vapor of the solvent. This arrangement makes it unnecessary, as will appear later, to know the vapor pressure of the solvent at any temperature, a quantity which had to be allowed for in ]ust’s work.‘ It is probably for this reason that Just's results on the solubility of CO; in liquids do not check very accurately those obtained here. The air is completely removed from the solvent in P by boiling for five minutes under reduced pressure and then applying a strong aspirator pump usually for upwards of half an hour more. During this time the solvent in P decreases by one fourth or one fifth of its volume. While this is being done Q is connected to S. Then Q is closed by the use of a rubber tube and clip, and forthwith connected to V. When the air has next been pumped out of B with a good vacuum pump, the boiler P is raised above the level of B and suflficient air-free solvent is run into B by gravity to approximately half fill B. To make sure that the last trace of air has been removed from the capillary C, this tube is in each case washed out twice with pure gas before a solubility determination is made. A short stub of pressure tubing is used to connect V and Q. Through this the liquid runs from P into B. Since this takes only a moment and since the liquids here used have little if any affect on rubber, this proceeding is not objectionable. The solvent is thus at no time exposed to air after once the air has been removed. The air is then pumped out of the capillary C and after stopcock T has been closed and A, C, and ET allowed to communicate with each I Just. 2. 5. mr Phys. Chemie. 37. n. :42 (war). ,‘;g§«;“X SOLUBILITY or C0: AND M0. 515 other, the height of gas in A at atmospheric pressure is carefully read. The mercury reservoir I can be raised or lowered to secure atmospheric pressure in A. Next E is turned so as to bring B and A into communica- tion, B is then shaken to hasten solution and the mercury in A rises indicating the volume of gas dissolved. When further shaking causes no more decrease in the volume of the gas in A the solution is assumed to be saturated. This takes only a few minutes of thorough agitation of B. This apparatus can be.shaken more vigorously than that of other in- vestigators and hence solution is hastened. Since the burette is close to the pivot of the apparatus the mercury in it is little agitated when the pipette is vigorously shaken. In consequence of this there was no emulsion between the mercury and the layer of solvent in the burette above the mercury. Before solution takes place the temperature is made the same in the two jackets A and B either by the use of ice water or the electric heating coils. After the reading has again been taken on the burette the tem- perature is raised simultaneously in both burette and pipette by means of coils G and K. During this time the shaking is continued, and at the proper time readings are repeated on the burette at intervals of 2° up to 36“. At a temperature much below that of the room it is difficult to keep the two jackets constant and at the same temperature because of radiation. ' When the reading for the temperature at 36° has been taken the solvent in B is measured in a small graduate calibrated to 0.2 cc. Dividing the volume of gas which disappears into the solvent by the volume of solvent used gives the solubility at any one temperature. In order to determine the solubility at the other temperatures correction for temperature change has to be applied to the volume of the solvent and also for the expansion of the gas when saturated with the vapor of the solvent as it is in burette A. The expansion coeFficient of the latter is determined separately in the ordinary way either before or after a regular run. Every determination on solubility is repeated at least once, and recorded results are accurate to within about I per cent. The nitrous oxide for this work was obtained from a steel cylinder of the gas purchased from the S. S. White Dental Company. It was rated as 99.7 per cent. pure, and to further insure its purity it was frozen with ethyl alcohol and liquid air, and while it was in that state all the gases were pumped off. When the frozen nitrous oxide had been thus sub- limed it was passed over P205 and thus freed from possible traces of moisture. Its density was then determined and found to be 1.968 grams per liter at standard temperature and pressure. [swam WILLIAM KUNERTH. Sims 516 With the exception of a short stub of pressure tubing connecting the steel cylinder with the glass system, the entire apparatus containing the purified gas was one piece of solid glass. This insures the Continued purity of the gas when it has once been obtained in pure form, and as it is always under pressure a leak would mean a loss of gas rather than a contamination and would be registered on the attached manometer. The carbon dioxide was made by heating Na]-ICO3; and after the gas had passed through a long CaCl, tube to free it from the water, it was treated in the same way as was the nitrous oxide. The solvents were obtained from standard chemical companies, and tests for boiling point and density were found to be in accord with the tables. RESULTS. Table I. shows the solubilities of CO, and of N20 in each one of the twelve liquids used, at intervals of 2° from 18°—36°. The solubility is expressed in c.c. of the gas under existing barometric pressure and at the , temperature specified, per cc. of solvent under the same conditions. It will be noticed that the solubility decreases as the temperature increases, also that the solubilities of CO2 and of N;O in any one solvent at a given temperature are very much alike. Table II. represents the coeflicient of temperature change of solubility times 10’, i.e., (41:/x-dt) at the temperatures indicated. In each case the solubility decreases with incr ase of temperature. In Table III. the seconifgalumn contains the number of molecules of the solvent (M ,,,1,,,.“) for each molecule of CO, (M C0,) in a saturated solution at 20° C. The last oolumn contains the corresponding number of molecules of solvent for each molecule of N20. ATTEMPTS AT THE PREDICTION OF SOLUEILITY BY R.AoUL1"s LAW. There are two general principles which have been used in the endeavor to predict solubility and it will be of interest to see how these fare in the interpretation of the foregoing results. The first one to use Raoult's Law in an attempt to predict solubility seems to have been Dolezalek.‘ As used by him this law means that the partial vapor pressure of one component divided by its vapor pressure when pure and in the liquid state is equal to its number of molecules divided by the total number of molecules in the solution. It is an empi- rical law giving, for a very limited number of substances,’ a linear relation- ship between concentration and vapor pressure. - Doleuilek. z. s. fLlr Fliysik Chemie. 54. p. 727 (1903) and p. I919 (no). a Dolezalek, z. s. (Llr Physik Chemie. 54, p. 725 (1903). 517 ] SOLUBILITY OF C0: AND N10. vm.. xxx No. 5. 2.» mm: N3 I tzu £2 and 2 mi :4 am.” ova was “.3 M9». 2.» 84 «Na ans 84 S." 2." 3; S: 8.. and Ned :.:_ 3.. Ga 2: as .2 3.N 36 Ba «.3 Si 8.» 22. was 83. 0: «u.~ S." 93 3.~ Em; a