174
OM N. BHATNAGAR AND CECILM. CI~ISS
Thermodynamic Properties of Nonaqueous Solutions. VI. Enthalpies of Solution of Some Tetraalkylammonium Iodides in Water and N,N-Dimethylformamide at 25 O by Om N. Bhatnagar and Cecil M. Crissl Department of Chemistry, University of Miami, Coral Gablea, Florida
$3124
(Received July 29, 1988)
Heats of solution of (CH&NI, (C*H&NI, and (C3H,),NI have been measured in N,N-dimethylformamide (DMF) and in water in the concentration range 3.5 X 10-4 to 5 x IO-Smand have been extrapolated to infinite dilution by the simple Debye-Huckel equation in order to obtain standard heats of solution in the two solvents. Enthalpies of transfer, AHt,", of the salts from water to DMF have been calculated and compared with other AHtrodata taken from the literature. Although in every case examined there is a11 unexpectedly large change in AHtrobetween CsI and (CHB)dNI,it is shown that this change cannot be fully explained in terms of the icelike structure of water molecules around the tetraalkylammonium ions. The modified Born equation is shown to be inadequate to explain, even qualitatively, the behavior of the enthalpies of transfer for these salts and the alkali metal iodides from water to DMF and other solvents, or between two nonaqueous solvents.
Introduction Tetraalkylammonium halides have attracted much attention in the past few years because of their abnormal behavior in aqueous s o l u t i o n ~ . Most ~ ~ ~ generally the abnormal properties have been attributed to the structural changes in water which occur in the vicinity of the R4N+ species. While these salts have been of considerable interest generally, only a few enthalpy-of-solution data exist for these compounds in either aqueous or nonaqueous systems a t 25". Latimer, Pitzer, and Coulter4 reported heats of solution of (CH&NI in water and Samoilov5 reported a heat of solution in water a t one concentration for (CH3)4NBr. More recently Arnett and McKelveya have measured heats of solution of the tetraethylammonium halides in water and in dimethyl sulfoxide (DnlrSO); Wu and Friedman' reported heat data for tetramethyl- and tetraethylammonium halides in water and propylene carbonate (PC); and Boyd and Wang8 have made similar measurements in water, methanol, and acetonitrile. I n all these latter investigations, heats of transfer from water to the nonaqueous solvent were evaluated. No heat-of-solution data have been reported for (C3H7)&T in water or for any tetraalkylammonium salts in dimethylformamide (DMF). I n continuation of our enthalpy measurements in nonaqueous ~olutions,g~~0 it appeared worthwhile to measure heats of solution of the tetraalkylanimonium iodides in water and DMF, to evaluate the enthalpies of transfer, a,nd to compare them with enthalpies of transfer for other solvent pairs.
Experimental Section Salts. Eastman Kodak's (CH3)dNI and (C3"I)dT\TI and Baker's (C2HJ4KI were used in the experiments. The Journal of Physical Chemistry
(CzH5)kNI and (CgH,)4NIwere recrystallized twice from a chloroform-ether mixture and (C&),NI was recrystallized twice from a methanol-ether mixture. The salts were dried a t 70" in a vacuum oven for 3-4 days, then in a vacuum desiccator for 3 days. Because of the hygroscopic nature of the salts, they were immediately transferred to a vacuum drybox under a dry nitrogen atmosphere. The iodide content for all three salts, determined gravimetrically by AgI, was 99.6% of the theoretical value. Solvents. Singly distilled water, which was passed through a deionizing column, was used for all the aqueous studios. Eastman's DMF was distilled at 85' and 15 mm of pressure through a distillation column filled with glass helices. The distillate was dried over freshly heated BaO for 8-10 hr, the contents being shaken intermittently. The liquid phase was removed and distilled again from BaO. This process was repeated until the conductance was 2 X 10-7 ohm-' cm-l or below. The amount of water in the DMF, as deter(1) To whom correspondence should be directed. (2) W. Y. Wen and S. Saito, J . Phys. Chem., 68, 2639 (1964). (3) R. H. Wood, H. L. Anderson, J, D. Beck, J. R. France, W.E. de Vry, and L. J. Boltzberg, (bid.,71, 2149 (1967). (4) W. M. Latimer, K. S. Pitaer, and L. V. Coulter, J . Amer. Chem. Boo., 62, 2845 (1940). ( 5 ) 0. Y. Samoilov, Dokl. Akad. Naulc SSSR,81, 641 (1951). (6) E. M.Arnett and D.R. McKelvey, J . Amer. Chem. SOC.,88,2598 (1966). (7) Y. C. \Vu and H. L. Friedmanr$. Phys. Chem., 70, 2020 (1966). (8) R. H. Boyd and P. S. Wang, Abstracts of Papers, 155th National
Meeting of the American Chemical Society, San Francisco, Calif., April 1968. See also P.S. Wang, M.S. Thesis, Utah State University, Logan, Utah, 1967. (9) R. P. Held and C. M. Criss, J. Phys. Chem., 69, 2611 (1965). (10) R. P, Held and C. M. Criss, $bid., 71, 2487 (1967).
175
THERMODYNAMIC PROPERTIES OF NONAQUEOUS SOLUTIONS mined by B Luft Karl Fischer titrator, never exceeded 0.01%. Apparatus and Procedure. The calorimeter, associated instrumentation, and procedure have been described in detail.9 The same precautions were taken to avoid contact of D M F with atmospheric moisture. Because of the low decomposition temperature of the tetraalkylammonium salts, the thin-walled sample bulbs were immersed in ice water to the base of the neck while being sealed by a torch under vacuum. Originally it was planned to measure heats of solution of (GHg)dNI, but this salt was observed to float on the surface of the solvent causing an extremely uncertain dissolution time.
Table I1 : Integral Heats of Solution of Tetraalkylammonium Iodides in Water a t 25" 104,~
mol/kg of HzO
The results are given in Tables I and 11. The data were extrapolated to infinite dilution in accordance with the simple Debye-Huckel theory, although in some cases it did not represent the best straight line through the data. The resultant standard heats of solution are listed in Table 111,along with the standard deviation of the data from the curve. A Debye-Huckel limiting slope of 1740 cal kg'/z rn01-*/~lo was used in extrapolating the D M F solutions. Within the limits of experimental error, the observed slopes in both solvents were not significantly different from theory. The tetraalkylammoniurn salts, like other electrolytes,1° exhibit much smaller heats of solution in D M F than in water. From the standard heats of solution, standard enthalpies of transfer of the electrolytes from water to D M F have been calculated and are listed in Table IV.
Table I : Integral Heats of Solution of Tetraalkylammonium Iodides in DMF a t 25'
8.33 15.1 21.7 24.1 38.9 54.8
9.92 9.92 10.00 9.98 9.99 10.05 10.03
(CZH:)~NI 1.87 2.12 2.39 3.36 4.02 4.47 5.43 (CaH7)4NI 1.93 3.03 4.20 4.38 5.77 7.06
3.50 4.49 5.71 11.3 16.2 20.0 29.5 3.72 9.20 17.6 19.2 33.3 49.8
6.70 6.62 6.71 6.81 6.75 6.78 6.76 2.79 2.74 2.79 2.76 2.78 2.80
Table I11 : Standard Heats of Solution of Tetraalkylammoiiium Iodides in DMF and Water a t 25'
-AHSO, kcal/mol-----
P
DMF
HaOa
HzOb
4.010 i 0.014 9.99 i 0.05 10.13 (10.050) (CHs)dNI 6.86 ( C ~ H G ) ~ N I 3.420 i 0.025 6.74 f 0.06 (CsH7)dNI 1.996 i 0.021. 2.765 f 0.025 5
This work.
b
Reference 8 except value in parentheses,
which is from ref 4.
ion,, mol/kg of DMF
(C€h)eNI 2.86 2.87 3.89 4.66 4.91 6.24 7.40
8.19
Results and Discussion
AHBI kcal/mol
102m'/z
AHs,
10arn1/z
koal/mol
2.77 3.01 4.20 8.78 23.3
(CHa)aNI 1.66 1.74 2.05 2.96 4.83
4.04 4.02 4.06 4.05 4.06
7.58 9.42 21.9 30.7 42.5 52.0
(CZHS)~NI 2.75 3.07 4.68 5.54 6.52 7.21
3.46 3.44 3.49 3.49 3.49 3,50
2.51 4.21 13.8 19.6 38.1 44.8
(GH7 )4NI 1.58 2.05 3.72 4.43 6.17 6.69
2.01 2.06 2.02 2.07 2.08 2.07
Table IV also lists standard enthalpies of transfer of alkali metal iodides from water to DMF, calculated from data in the literature,l0-'2 as well as heats of transfer from water to other solvents, also reported in the literatureqG --8 The derivative of the Born equation, as modified by Latimer, Pitzer, and Slansky,'3 gives for the enthalpy of solvation the expression10
(11) L. Weeda and G. Somsen, Rec. Trav. Chim.,86, 893 (1967). (12) V. B. Parker, "Thermal Properties of Aqueous Cni-univalent Electrolytes," National Bureau of Standards Report NSRDS-NBS 2 , U.S. Government Printing Office, W'ashington, D. C., April 1865. (13) W. M. Latimer, K. S. Pitzer, and C. XI. Slansky, J . Chem. Phya., 7, 108 (1939).
Volume 73, Number 1 January 1968
176
OM
Table IV: Standard EnthaIpies of Transfer from Water to Various Nonaqueous Solvents a t 25” -----AHtro
(solvent+-HzO), kcal/mol PCe DMSO’
Salt
DMFa
LiI NaI KI CSI (CH3)dNI (CzH3)4NI (C3H7)4NI
-3.97b (-11.8)’ - 1 1 . 3 j b (-12.15)d 12.90b -12 2Zb (-12.22)d - 5.98 -3.32 -0 . 7 6 9
-
-0.10 -3.22 -6.02 -7.18 -4.74 -0.59
CHaOH
-9.67 -11.36 -10.30 -1.97
-6.21’ -3.82‘ -O.3gh 1.2P
a Where only heats of solution in the nonaqueous solvent have been reported, the heats of solution in water, used in calculating the heats of transfer, have been taken from ref 12. Heat-of solution-data taken from ref 11. ’ Calculated from heat-ofsolution data of NaI, NaBr, and LiBr taken from ref 10. The standard heat of solution of LiBr was obtained from an extrapolation of the data from concentrations > 3 X m. The standard heats of formation of the crystalline salts used in the calculation are from W. M.Latimer, “The Oxidation States of the Elements and Their Potentials in Aqueous Solutions,” 2nd ed, Prentice-Hall, Inc., Englewood Cliffs, Pj. J., 1952. Heatof-solution data taken from ref 10. e Reference 7. Reference 6. O S . I. Drakin and C. Yu-min, Russ. J. Phys. Chem., 38, 1526 (1964). Reference 8.
‘
’
where ei and ri are the charge and crystal radius of the ion, respectively; E is the dielectric constant of the solvent; and 6 is a parameter which depends upon the sign of the charge on the ion and the solvent. The difference in the enthalpy of solvation for a given ionic species in two solvents, X and Y, gives the enthalpy of transfer of the corresponding ion between the two solvents. Consequently, if one assumes that &/dT is negligibly small, one can obtain for the enthalpy of transfer of ions from solvent X to solvent Y the expression
CRIW
Table V: Standard Enthalpies of Transfer between Various Nonaqueous Solvents a t 25”‘ kcal/mol-----DMSO
r___--AHtro.
PC
Solvent+DMF 3.87 (11.7) 8 . 1 3 (8.93) 1.68 (2.48) 6.88 1.54 5.04(5.04) 1.92 1.24 2.73 1.35 Solvent +PC -6.45 -5.34 -3.12
-1.38
(3)
CHaOH
5.14 (5.94) 8.40(8.40) 5.59 4.57 -2.99 3.36 4.35 1.84
Solvent-DMSO
3.46 6.48 3.22
+
The Journal of Physical Chemistry
nf.
CECIL
They explain this phenomenon in terms of the iceberg effect proposed by Frank and Evans116in which water molecules are more highly ordered in the vicinity of hydrocarbons. Presumably, the hydrocarbon qualities of the tetraalkylammonium ions give the extra stabilization for iceberg formation in aqueous solutions. An examination of enthalpies of transfer of salts from water to D M F and DMSO, listed in Table IV, shows that there is neither a continuous increase nor a decrease of AH*, with increasing radius as predicted by eq 3, even for the alkali metal ions. The lack of a trend consistent with eq 3 is also observed for enthalpies of transfer of the alkali metal ions from water to formamide and Nmethylformamide.17 Even though, in general, the alkali metal ions do not follow the trend predicted by eq 3, there is, neverthelcss, surprisingly large changes in the enthalpies of transfer between CsI and (CH3)4WIfor each of the solvents listed in Table IV. On first consideration one might conclude that these large changes are caused by the unique structure of water around the tetraalkylammonium ions as suggested by \Vu and Friedman7 for the H20-PC solvent pair. While this argument is appealing, it probably is not the complete explanation, since interruptions can also be observed between CsI and (CH3)dNI for solvent pairs not involving water. Table V lists enthalpies of transfer of electrolytes between nonaqueous solvents. Although the sparsity of data makes trends more difficult to discern, there neverthe-
where K X and K y are constants characteristic of the solvent and equal [l - ( l / e ) ] { l - [T(de/dT)/e(e l ) ] )where E is the dielectric constant of the respective solvent . For cations, 6 is approximately constant for the solvents HzO, DMF, and PC.10t14 If this generally true, eq 2 can be writ,ten as
where 6 = 6x = b y . Consequently, the enthalpies of 6) for transfer should be a linear function of l/(Ti cations. Wu and Friedman7 and Friedman16have shown that the enthalpies of transfer of ions between water and propylene carbonate decrease with increasing radius for the alkali metal ions but there is an interruption of this trend between Cs+ and the tetraalkylammonium ions.
N. BHATNAGAR AND
5
See Table IV for sources of data.
(14) Y . C.Wu and H. L. Friedman, J . Ph.ys. Chem., 70, 501 (1966). (15) H.L. Friedman, ibid., 71, 1723 (1967). (16) H.S. Frank and M. W. Evans, J . Chem. Phys., 13, 507 (1945). (17) L. Weeda and G. Somsen, Rec. Trav. Chim., 85, 159 (1966).
177
THEPYROLYSIS OF BIACETYL less appear to be interruptions between CsI and (CH& XI for the DMF-PC, CH30H-Di\4F, and CHSOHDRISO solvent pairs. The interruption for the CH3OH-PC solvent pair occurs between ( C H 3 ) N and (C2HB)N. At the present time it appears that no general statement can be made in terms of solvent structure to explain the AH,," data for the various solvent pairs. We conclude that (1) the modified Born equation, satisfactory as it may be in describing free energies of solvation,13J*as it is now expressed is not suitable for predicting heats of transfer of cations even qualit,atively, and (2) the large change observed in AHtpoas one proceeds from Cs+ to (CH&N+ is not explained by the water structure alone but is the result of special interactions of the tetraalliylammonium ions with solvents in general.
Weeda and Somsenl' have applied the van Eck theory'g to enthalpies of transfer of alkali metal iodides and have found AHt," to be relatively constant, as predicted by the theory, for sodium, potassium, rubidium, and cesium iodides between water and formamide but not constant, for the same salts upon transfer from water to N-methylformamide. Table I V shows that, in general, the van Eck theory is not consistent with AHt,' data.
Acknowledgment. The authors are grateful to the National Science Foundation for their financial support through Grant GP-7870.
(18) C. M. Criss and E. Luksha, J . Phys. Chenz., 72, 2966 (1968). (19) C. L. van P. van Eck, Thesis, University of Leiden, Leiden, The Netherlands, 1958.
The Pyrolysis of Biacetyl and the Third-Body Effect on
the Combination of Methyl Radicals by K. J. Hole and M. F. R. Mulcahy CSIRO Diaision of Mineral Chemistry, Coal Research Laboratory, Chatswood, NSTV, Australia
(ReceCed July 19,1968)
The kinetics of pyrolysis of biacetyl at 677-776°K and 0.6-45 torr have been investigated by the stirred-flow method. The decomposition (to carbon monoxide, methane, ketene, acetone, ethane, and 2,3-pentanedione) is shown to be a chain reaction initiated by reaction 1, propagated by reactions 3 and 8, and terminated mainly by reaction 6, which is strongly pressure dependent. The following kinetics data were obtained (k and A being in em8 mol-1 sec-I) (CHZC0)2 -+ 2CH3CO (3) CH3 (CH3CO)i CH4 CHzCOCOCHa (6) CH3 CH3 (+M) CzHa (3-M) (8) CH3 (C&CO)z-+ (CHa)zCO CH3CO (1)
+ + +
-+
+
Order
k (at 730°K)
log A
1 2
7.6 x 10-5
16 0
1.2
11.8
109
67.2 1 3 . 3 9.1
1 . 6 X lozo 16.3 w-13 2.6 X lo8 10.3 6.3 Pressures of biacetyl, at which the second-order rate constant for reaction 6 is half its limiting high-pressure value, increase from 14 torr at 677°K to 30 torr at 776°K and are in satisfactory agreement with values calculated by prcvious authors on the basis of the R R K M theory assuming a "loose" activated complex and strong deactivating collisions. The presence of a trace of 2-hydroxy-2-methylbutan-3-one in the products is regarded as evidence that reaction 8 proceeds through formation of the radical (CH&C(0.)COCH3. The heat of formation of the gaseous acetyl radical is estimated to be -5.1 f 2.0 kcal mol-' at 298°K. -+
+
Introduction R/Iuch of the recent progress in understanding the kinetics of pyrolytic and free-radical reactions has been achieved from studies of the same elementary reactions in different reaction systems. Translation of an elementary reaction from one environment to another may bring into prominence in the second system features which are vestigial or absent in the first;
3
x
E , kcal mol-'
2
important examples of this are the occurrence of pressure dependence in rec~mbinationsl-~ and the unimolecular decomposition^^-^ of radicals. The study of biacetY1 reported here TvaS carried out with a twofold Pur(1) R. E. Dodd and E. W. R. Steacie, Proc. Roy. SOC.,A223, 283 (1954)o (2) N. L. Arthur and T. N. Bell, Chem. Commun., 9 , 166 (1965). (3) L.F. Loucks, Can. J. Chem., 45, 2775 (1967).
Volume 73, Number 1 Janua,ry 1969