1730
T'ol, 06
KOTES
Fuoss-Onsager equation, The dielectric constant and viscosity of mater used in ihe computation were 78.54 and 0.008903 poise, respectively. The density of both salts is about 1.43. The maximum correction for solvent conductance was 1%. The results of the analysis are given in Table I1 where AC = limiting conductance, UJ = ion size parameter from J-terms in the equation, K A = association constant for the formation of ion pairs, and U A = standard deviation in A-units of the data points from the equation.
TIIE HEAT OB FORR.IATION OF GASEOUS
METHYL NITRITE' BY JAMES D. RAYAXD A. ARSOLDCERSIION % h o d of C l r r i r i r d i g , CIIroioiu Instilute of 'I'eLhnoloyy, ilfiuiitu 13 6'eoi
yzo
Itrcezted .Julg ii, 19Cl
Calculations which have been made recently by Gray and Pratt2 indicate that the value for tho heat of formation of methyl nitrite calculated from the equilibrium study of Leermakers and Ramsperger3 is somewhat in error. Although Gray TABLE I1 a r d Pratt2 quote unpublished data of Baldrey, DERIVEDCONSTANTS Lotsgesell, and Style as evidence for a revised Salt Ao dJ KA ah value for the heat of formation of methyl nitrite. DMD-BDS 134.35 f 0.15 4.30 f 0.13 40 f 5 0 04 this value was obtained from solution calorimetry DMD-BPDS 124.08 z!z 0 09 4.41 f .14 -5 z!z 4 0.03 ar,d depends ultimately on a small difference beThe association constant of the DMD-BDS is tween two large numbers. Geiseler and Thier40, compared to that of DMD-BPDS, which is felder4 have determined the heat of combustion of essentially zero. This difference can be explained methyl nitrite, but their value corresponds in by the clectrostatic interaction. The distances precision to an uncertainty in entropy of f 3 e.u.. between the charges on the D M D and BDS ions which is not sufficiently accurate to determine the are almost the same, so that when an ion pair forms, barrier to rotation of the methyl group. W e fell the charge sites of the two ions are very close to that the direct determination of the heat of reeach other. Thus, greater association would be action between methyl alcohol and nitrosyl chloexpected for the DMD-BDS than for DMD- ride to form methyl nitrite and hydrogen chloride, BPDS, where the distance between the two charges reaction 1, mould be a much more sensitive method on the anion is of the order of three times the for measurement of the heat of formation of methyl chhrge separation in the cation. CH,OH(g) NOCl(g) = CH,ONO(g) The values of the limiting conductances are not H C W (1) precise because the salts are hygroscopic, but this uncertainty has little effect upon the other parameters of the equation. The limiting conductance nitrite. In this reaction, the heats of formation of the DMD ion can be computed from Atkinson's of the other participants all have been determined values3J for the BDS and BPDS ions; the two with high accuracy: thus the precision of the values are given in Table 111. The difference of Joule expansion reactant mixing gas calorimeter of about 1% between them probably is due to a small Ogg and Ray5 would be adequate to determine the amount of water still in our sample of DMD- barrier to rotation of the methyl group in methyl RDS; this salt was unusually difficult to dehydrate. nitrite. Experimental Also given in Table I11 are the (uncorrected) Materials.-Methyl alcohol was purified by the method of Stokes radii calculated for the D M D ion.
+
TABLE I11 SINGLEION CONDUCTANCES Salt
DMD-BDS
DMD-BPDS
Xa
-
59.94 48.99
)io
+
74.4 75.1
108R-
lOsR+
3.07 3.76
2.47 2.45
The center-to-center ion pair distances computed from the hydrodynamic radii of DMD-BDS and DMD-BPDS are 5.54 and 6.21 A., respectively; these are higher than the values of UJ of 4.3 and 4.4, respectively. These hydrodynamic dimensions should be larger than the electrostatic because the charge sites are situated near the ends of the prolate ellipsoidal ions and thus the distance of closest approach of the charge sites might be expected to be less than the sum of the mean Stokes radii of the ions. Thus the Fuoss-Onsager eq. (1) can adequately represent data for 2-2 salts in aqueous solution up to concentrations of about 2.5 X 10-3 jl4 to O.O20j,. For the case of bolaform electrolytes, we note that the degree of association is sensitive to the ratios of the charge separations in the two ions, the association being greatly increased when this ratio is near unity.
+
Gillo.6 Nitrosyl chloride was prepared as described previously'by the reaction of nitric oxide with chlorine. Methyl nitrite was prepared as described previously.* Dry hydrogen chloride gas was prepared by the reaction of C.P. sulfuric acid with reagent grade potassium chloride in a vacuum system. Apparatus and Procedure.-The Joule expansion reactant mixing gas calorimeter employed has been described previously.* I n the present case the calorimeter contained a 269.8-m1. gas reaction bottle. The energy equivalent of the calorimeter including 275 ml. of chlorobenzene liquid was found to be 161.O cal./deg. by the heat of water vaporization method which has been described by Ray.$ The value -10,520 cal./mole was used for the heat of vaporization of water a t 25". Reactions were carried out a t 25". The thermochemical calorie exactly equal to 4.184 absolute joules was used in calculations. The calorimeter wa8 filled with gases from a vacuum system which was equipped with
(1) Presented in part a t the America.n Chemical Society Meeting, St. Louis, Missouri, March 21-30, 1961. (2) P. Gray and 31.W. T . Pratl, J . Chem. Soc., 3403 (1058). (3) J. A. Leermakers and H. C. Ramsperger, J. A m . Chem. Soc., 54, 1832 (1932). (4) G. Geiseler and W. Thierfelder, Z. phusilc. Chem. (Frankfurt), 29, 248 (1961). ( 5 ) R. A. Ogg, Jr., and J. D. Ray, J . Phys. Chem., 61, 1087 (1957). (6) L. Gillo, Ann. chim., [11] 12, 281 (1939). (7) J. U. Ray and R. A. Ogg, Jr., J. Chem. Phys., 26, 984 (1957). (8) J. D. Ray and R. A. Ogg, Jr., J . Phys. Chem., 63, 1522 (1959). (9) J. D,Ray, Rea. Sei. Instr., 27, 863 (1956).
175I
NOTES
glass Bourdon gages for measuremcnt of the pressures of the corrosive gases. The equilibrium constaiit for eq. 1 was found to be needed for interpretation of the calorimetric data. The old value
TABLE I1
DATAFOR REACT~ON OB 300 MM. PRESSURE OF CH,,OH IN THE
269.8-XL.CALORIMETER BOTTLEWITH NOCl
ADDED TO
GIVEVARIOUS FINALPREBBURE~
lor
K =
The observed temperature rises, A T calorimeter, are corrected for the Joule expansion of the NOCl into the calorimeter bottle (3rd column) to give the temperaturo change asqociated with the reaction (4th column).
(HCI) (CH3ONO) (CHdlH) (NOC1)
given by Lf,ermakrrs and Ramspergcr3 is 1.92 f 0.6. The value of K determined by Cox and Raylo is 0.73 based on quantitative infrared spectroscopy and 0.69 based on the rate constants of the forward and reverse reactions. The value 0.73 i s used in the present calculations. Since the equilibrium constant for the reaction is nearly unity, it is ineces~aryto employ a large excess of one reactant in order to obtain nearly complete reaction. Unfortunately, the reverse reaction of eq. 1 results in the formation of a liquid phase when a large excess of either reactant is used. The reaction of 100 mm. pressure of mcthyl alcohol in the 269.8-m1. calorimetcr bottle a t 25' with excess nitrosyl chloride was found t o yield a liquid phase only at low NOCl pressures, and was used to evaluate the heat of the reaction. In a typical run, the calorimeter bottle contained 100 mm. of MeOH. After measuring the initial temperature drift, YOCl was admitted to a final pressure of 244 mm. The observed temperature rise of the calorimeter was 0.0187'. From this was subtracted the temperature rise due to the Joule expansion of the nitrosyl chloride into the calorimeter bottle, 0.0079', to give 0.0108O change in temperature due to the reaction. The heat of reaction per mole of methyl nitrite formed is given by
AH
=
(calorimeter energy equivalent) (torr. temp. _______rise) (moles of CH30N0 formed a t equilib.) (161.0) (0.0108) _______ - -2210 calories (exothermic) (0.000788) T24B1,E 1
CALCUL.4TTH:I) PRESSURES, Poalcd. = nxociRT/V, AND OBSERVEDPRBSSURES OF WEIGHEDAMOUNTS OF NOCl ADNITTEDTO AN EVACUATED FLASK PLUS GAGE P L U S VOLUME 310.8 ML. LEADSOF TOTAL ,-----P,
mm.----
g.
T, OK.
Calcd.
Obsd.
0.489 0.6563 1,098 1.566 1,720
298.0 299.0 299.2 299.8 299.0
446.5 600.0 1008 1438 1576
444.5 593.5 984 1388 1511
NOCI,
AT calorimPfinal, eter mm. uneorr.
AT
AH
Joule expansion
AT Reaction
PCH,ONO, reaction 1, mm. a t cal./mole equilib. CHaONO
54.3 61.3 65.7 70.1 71.7 77.0 81.2 83.1 85.3 86.0 87.1 88.3
244 0.0187 0.0079 0.0108 ,0218 ,0106 .0112 294 ,0132 .0113 .O.M 342 ,0290 ,0158 ,0132 392 ,0300 .0173 ,0127 416 ,0217 .0165 .0382 507 ,0186 .0289 .0475 630 .0510 ,0342 ,0168 732 ,0580 .0423 ,0157 874 928 ,0592 .0452 .0140 ,0692 .0507 .0185 1027 1124 ,0730 ,0559 .0 171
Average
-2210 -2030 - 1910 -2090 - 1970 -2380 -2540 -2240 - 2040 - 1810 -2360 -2150 -2140 f 200
Separate experiments showed that in the pressure range used in the above reactions. there was no detectable heat of reaction between KC1 and SOC1, or between CH30NO and NOC1. Further experiments in which 100 mm. pressure of CH30H was admitted to a 310.8 ml. total volume system comprised of flask, gage, and leads, followed by admission of weighed amounts of XOC1, showed that no dimeric or trimeric compounds formed. The calculated pressures on mixing methyl alcohol and nitrosyl chloride agreed with thpse observed when allowance was made for the gas imperfection of NOCl given in Table I. The value for the heat of reaction 1a t 25" found is -2,140 f 200 cal. (standard deviation) when corrected to represent total reaction with the equilibrium constant 0.73 given by Cox and Ray.10 The value calculated for the standard heat of formation of methyl nitrite gas a t 25" is - 15.64 i 0.20 kcal./mole. The heat of formation of nitrosyl chloride was taken from Ray and Ogg.11 The heats of formation of methyl alcohol and hydrogen chloride were taken from Rossini, et
Discussion
The heat of formation AHo2g~CHaOXO(g) calculated from the equilibrium study of Leermakers and Ramsperger3 is - 16.28 kcal./mole; the value of Baldrey, Lotzgesell, and Style2 is -14.93 f Table I1 lists observed temperature rises associated with varying final pressures in the calorimeter bottle, and the 0.26; that of Geiseler and Thierfelder* -16.8 f 0.8; combination of the heat of combustion respective pressures of CHZONO calculated to be present at equilibrium(. A small correction was made by succemive of methyl nitrate gas of Whittaker, Wheeler, approximation to allow for the effect of the fugacity of and Pike,I3 -29.4 f 0.8, with the value of Ray NOCl on the equilibrium amount of methyl nitrite. This fugacity correction amounted to at most 30 small cal. in the and Oggs for the heat of reaction of nitrogen heat of reaction per mole of methyl nitrite formed. Tem- pentoxide gas with methyl nitrite gas gives perature rkes due to the heat of expansion of the NOCl into -14.36 f 0.9. that from the present study is the calorimeter bottle were allowed for. The heat.s of ex- -15.64 f 0.20. The average of all five results pansion include an almost negligible correction for the imperfection of the NOCl gas calculated from the data of Table I. is -15.62 f 0.94. When the value of the present The Joule-Thompson heat of expansion, .uLcp, was not study, - 15.64, is combined with the value of Cox present in -the experiments. The calorimeter was filled by and Raylo for equilibrium 1, the entropy of first opening the stopcock on the calorimeter bottle to the methyl nitrite gas is calculated to be 66.81 f vacuum system manifold which contained a pressure of The entropy calculated by Gray and NOCl in excess of that of the methyl alcohol in the bottle. 0.67. Then the stopcock to the NOC1 sOorage flask was opened to Pratt2 based on the geometry of the cis-trans
fill the calorimeter slowly. Thus, the throttling took place a t the NO01 storage Aa$k outlet stopcock and the manifold and leads to the calorimeter bottle acted as a heat exchanger tv eliminate the cooling effect of the JouleThompson expansion. .-
(10) J. R. Cox, Jr., and J. D. Ray, J . Chem. Phys.. 34, 1072 (1961).
(11) J. D. Ray and R. A. O g g , Jr., rbid., 31, 168 (1959). (12) r. D. Rossini. D. D. Wagmau, 1 ' . H. Evans, S.Levine, and 1. Jaffe, "Selected Values of Chemical Thermodynamic Properties," Circular 500, U. S.Ror. Standards, 1852. (13) H. H. Whittaker, W. H. Wheeler, and H. H. M. Pike, J . Inst. Fuel, 20, 137 (1947).
1752
isomers and their vibrational frequencies is 73.8, assuming free rotation. Gray and Reevcs'4 have found a barrier 10,500 f 2000 cal. ascribed to hindered rotation of the S O group. When an entropy deficit of 3.0 e.u. corresponding to this barrier is subtracted, there still remains an entropy deficit of 4 f 0.7 e.u. between the calculated and experimental value. Gray and Pratt2 calculated the maximum entropy due to free rotation of the methyl group to be 3.58 e.u. and that for the N O group to be 5.80. The entropy deficit 4 i 0.7 e.u. found in the present work thus corresponds to an essentially fixed methyl group. Such a high barrier to rotation of the methyl group is reasonable since in methyl nitrate the distance closest approach of a methyl group H to 0 is 1.8 A., whereas the structyral data given by R o g o ~ s k i lindicate ~ only 1.3 A. for this distance in methyl nitrite. Further evidence for a large barrier to methyl group rotation in methyl nitrite is given by Tarte's16 analysis of the C-D vibrational bands in CHzDONO and Wagner's17 interpretation of the intensity variation of the Raman -OS0 lines which was ascribed to hydrogen bonding. Acknowledgments.-This research was supported in part by a grant from the Research Corporation, and by a grant from the National Science Foundation. (14) (15) (16) (17)
Vol. 66
NOTES
P. Gray and L. W. Reeves, J . Chem. Phys., 32, 1878 (1960). F. Rogowski. Ber., B76, 244 (1942). P. Tarte, Bull. soc. chim. Belges, 62, 401 (1953). J. Wagner, J . phys. radium, 18, 526 (1954).
HAMMETT CORRELATIOKS FOR THE SOLUBILITY OF GASEOUS H7i7DROGEK CHLORIDE IK CERTAIN AROMATIC
SYSTEMS BY MORRISRAPOPORT, C. KINNEYHAS COCK,^ AND EDWARD A. MEYERS Department of Chemasfry. The A . and M. College of Texas, CoZZege Statton, Texas Received February 9, io62
I n the typical Hammett relation, a functional group property, such as an equilibrium or rate constant, is involved. In sharp contrast, there now are numerous quantitative relationships in the literature between a molecular property, such as the energy of electronic spectral excitations ( h ~ )and , Nammett's U-values. W e have therefore looked for Hammett relations in the realm of solubility phenomena, which are also characterized by the lack of a localized "reaction" site and where gross molecular structure similarly plays an important role. We have found that the data collected by Brown and Brady3 for the solubility of hydrogen chloride in solutions of some aromatic compounds follow a linear Hammett relation. Because of recent successful spectra-structure (1) Abstracted from a portion of the Ph D. DisRertation of 111 R , Mag, 1961. (2) To whom inquiries should be addressed, a t Department of Chemistry, The A. and &I. College of Texas, College Station, Texas. (3) H. 6. Brown pnd J. 33. R r s d y , J. Am. Chsm, S a o . , 74, 33713 (10$21
