26
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
T. G. Ryan, T. E. M. Sambrook, and G. R. Freeman
acetylene concentration it seems possible that B adds to another acetylene molecule forming C. The structure of C is not known. However, the possible identification of C as the radical 1-hydroxyhexatrienyl-6 is supported by the fact that the second UV absorption maximum (Ama -260 nm) of C is found to be similar to that of hexatriene.1° The addition of an acetylene molecule to both A and B is also supported by the decrease in activation entropy found. Because of the small doses used dimerization or disproportionation reactions are not possible in the measured time interval. However, it is not excluded that species C is an intramolecular rearrangement produced from 1-hydroxyhexatrienyl-6 since it seems not to add further to acetylene. C decays by a bimolecular reaction to give unknown products. 11. Low [CzH2]and High Dose. Under these conditions the addition reaction of A with acetylene is depressed and the reactions of A could be studied. Hydroxyl ions convert A in a diffusion-controlled reaction to the formylmethyl radical (CH2CHO). This reaction can only be understood if it is assumed that A is deprotonated at every encounter with OH- and the resulting radical anion D reacts quickly with water (i-e., reactions 5 and 6). The bimolecular disappearance of A results in a species 240 nm (Figure 5). One possible which absorbs at, , ,A product which could absorb at this wavelength is the dimer F (reaction 7). Other possible products deriving from
disproportionation reactions should absorb at lower wavelengths. It is of interest to note that dimer F has a lifetime of several minutes. It rearranges probably to succinaldehyde. From the bimolecular disappearance of A it follows that the spontaneous rearrangement of A to E has a reaction rate constant smaller than lo5 s-l.
Acknowledgment. The authors thank Mr. F. Schworer and Dr. B. Parsons for helpful discussions and Messr. Klever, Teodoru, and Toepfer and especially Mrs. Brinkmann for technical assistance. References and Notes (1) P. 0. Clay, G. R. A. Johnson, and J. Weiss, J. Phys. Chem., 63, 862 (1959). (2) M. Fiti, Rev. Roum. Chim., 19, 377 (1965). (3) C. Walling and G. El-Taliawi, J . Am. Chem. SOC.,95, 848 (1973). (4) P. Neta and R. W. Fessenden, J . Phys. Chem., 76, 1957 (1972). (5) G. Grabner, N. Getoff, and F. Schworer, Int. J. Radiat. Phys. Chem., 5, 393 (1973). (6) H. Stephen and T. Stephen, Ed., "Solubilities of Inorganic and Organic Compounds", Pergamon Press, New York, N.Y., 1963. (7) E. Sawlcki and J. D. Pfaff, Chem.-Anal., 55, 6 (1966). (8) R. L. Willson, C. L. Greenstock, G. E. Adams, R. Wageman, and L. M. Dorfman, Int. J. Radiat. Phys. Chem., 3, 211 (1971). (9) E. J. Land and M. Ebert, Trans. Faraday Soc., 63, 1181 (1967). (10) "Handbook of Chemistry and Physics", 53rd ed,The Chemical Rubber Co., Cleveland, Ohio. (1 1) K. M. Bansal, M. Gratzel, A. Henglein, and E. Janata, J. Phys. Chem., 77, 16 (1973). (12) M. C. Sauer and 8. Ward, J . Phys. Chem., 71, 3971 (1967).
Radiation Sensitized Chain Reactions. Aqueous Nitrous Oxide and Methanol' T. G. Ryan, T. E.
M. Sambrook, and G. R.
Freeman"
Chemistry Department, University of Alberta, Edmonton, Alberta T6G 2G2, Canada (Received June 21, 1977) Publication costs assisted by the University of Alberta
The radiolysis of aqueous solutions containing methanol and nitrous oxide was examined in the temperature range 498-573 K. The high yields of nitrogen and formaldehyde are explained by a radiation induced chain process. The proposed mechanism involves the transfer of an oxygen atom from NzO to a CHzOH radical producing .0CH20H. The supporting evidence for the mechanism is a linear dependence of G(N2)and G(CH20) on the square root of the NzO concentration at large NzO concentrations, a linear dependence of the G values on D-'12, and the same activation energy for both G(Nz) on G(CH20).
Introduction In a previous paper2 the results obtained from irradiation, at high temperatures, of water containing 2-propanol and N 2 0 were given. A mechanism was proposed which involved the reaction of .0C(CH3)20Hradicals with 2propanol as one of the propagation steps. The present work involves the irradiation, a t elevated temperatures, of water containing methanol and NzO. The yields of nitrogen and formaldehyde were found to be independent of methanol concentration over the range 0.05-0.5 M. This behavior leads to the postulate that the aOCH20H radical reacts with the water rather than with methanol. Experimental Section Materials. The purification procedure used for water has been described previously.' Methanol (Spectro Grade) obtained from Fisher Scientific Co. was used as received. Chromotropic acid (Practical Grade) obtained from Eastman Organic Chemicals and formaldehyde (40 % 0022-385417812082-0026$0 1.OO/O
aqueous solution) from Fisher Scientific Co. were used as supplied. Sample Preparation and Irradiation. The procedure for preparation and irradiation of samples has been described earliern2The dose rate in the 6oCoGammacell was 1.6 X 1017 eV/g min and the dose 1.6 X 1017eV/g. The solutions were at their natural pH. Product Analysis. The procedure for analysis of nitrogen has been described previously.2 Formaldehyde was determined using the chromotropic acid m e t h ~ d .Cali~ brations were made using standard formaldehyde solution. The 40% formaldehyde solution was used as a standard and the exact concentration of formaldehyde in this solution was determined through its methone d e r i ~ a t i v e . ~
Results Nitrogen and formaldehyde were the principal products found from the radiolysis of water containing nitrous oxide and methanol at elevated temperatures. At 573 K, with
0 1978 American Chemical Society
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978 27
Radiation Sensitized Chain Reactions
O
I
I
I
'
i
I
/
I
i
'
I I
1 (3
.I
OO
01
02
03
I
I
04
47 Flgure 1. Product yields plotted against the square root of the nitrous oxide molarity at 573 K. [CH,OH] = 0.53 f 0.03 M: 0, G(N2); A , G(CH20). The full curve was calculated from eq 12 and 13 using the rate constant values given in Table I: 0 , G(N2) in the absence of methanol.
-
00
2
4
6
8
10
12
DR-$(10"ev/g min 1-4
Figure 3. Product yields plotted against (dose rate)-''2. [NpO] = 96 f 5 mM, [CH30H] = 0.53 f 0.03 M, T = 573 K: 0, G(N,); A, G(CHp0).
Discussion The following set of reactions is consistent with the results: H,O w *H,.OH .H, .OH + HCH,OH H,, IS,O + .CH,OH N,O + .CH,OH + N, + *OCH,OH .OCH,OH + H,O HOCH,OH + .OH
(3 100
-+
-+
*OCH,OH + HCH,OH HOCH,OII t *CH,OH HOCH,OH 3 CH,O .t H,O 2*CH,OH P, .OCH,OH + *CH,OH+ P, -+
[CH30H], M
Flgure 2. Product yields plotted against the methanol molarity at 573 K. [ N 2 0 ] = 96 f 5 mM: 0, G(N2); A , G(CH20).
[CH,OH] = 0.53 f 0.03 M and [N20] = 96 f 5 mM the yields were G(N2) = 188 and G(CH20) = 149. 1. Effect of Nitrous Oxide Concentration. The method of estimating the Ostwald solubility coefficient for N 2 0 in water a t 573 K has been described previously.2 Figure 1 displays the yields of nitrogen and formaldehyde for irradiations a t 573 K and 0.53 f 0.03 M methanol. The nitrous oxide concentration was varied from 4.3 to 149.3 mM and G(N2)values were in the range 9 to 230. The G(HCH0) values varied from 13 to 181. Blank sample yields of nitrogen and formaldehyde were equivalent to between 0 and 7 at 573 K and appropriate blank yields have been subtracted from all values reported in this work. 2. Effect of Methanol Concentration. Figure 2 contains results from the irradiation of samples containing 96 f 5 mM nitrous oxide and concentrations of methanol between 0.06 and 510 mM. The irradiation temperature was 573 K. The nitrogen yields range from 35 to 188 and the formaldehyde yields from 0 to 149. At 0.06 mM methanol concentration a G(N& of 35 represents complete consumption of the methanol at 1.6 X 1017eV/g. 3. Effect of Dose Rate and Temperature. The samples contained 96 f 5 mM nitrous oxide and 0.53 f 0.03 M methanol. Yields of nitrogen and formaldehyde as a function of (dose rate)-lI2 a t 573 K are given in Figure 3. The G values decreased as the dose rate was increased from 0.6 to 15.0 eV/g min). The total dose normally used was 16 X 10l6 eV/g but at the lower dose rates it was necessary to use doses down to 3 X 10l6eV/g to prevent excessively long irradiation times. As the temperature was increased from 498 to 573 K nitrogen and formaldehyde yields increased from 58 to 188 and 53 to 149, respectively.
-+
(1) (2) (3)
(4) (5) (6)
(7)
2*OCH,OH + P,
The reason for considering (H, OH) as initiators has been given previously.2 Reaction 2 is well established6 and reactions analogous to 3 and 5 have been discussed previously.2 Reaction 4 is proposed to explain the fact that the product yields are independent of methanol concentration for (0.05-0.5) M methanol (Figure 2). This would not be true if reaction 5 were the rate-determining process in this system. The kinetic analysis is not altered if reaction 4 is replaced by a decomposition: *OCH,OH + CH,O
+ *OH
(10)
but reaction 10 is not consistent with the results from the 2-propanol system2 where the .0(CHJ20H radical would be expected to undergo reaction 10 as well. The 2-propanol results show a dependence of the yields on the 2-propanol concentration that is consistent with reaction 5, not reaction 10, being important in the propagation.2 Reaction 4 is probably endothermic in the gas phase by about 18 kcal/mol, as the H-OCH20H bond dissociation energy is not likely to be greater than that in methanol, 100 kcal/mol.6 If the solvation energies of OH and HOCHzOH are greater than those of H 2 0 and .0CH20H, reaction 4 may be less unfavorable energetically in aqueous solution than in the gas phase. In water solution the HOCH20H species is stable, as the constant Kqq =
[ HOCH,OH]/[ CHZO]
(11)
has a value of 2 X lo3 at room t e m p e r a t ~ r e . ~ If it is assumed that reactions 3 and 4 are the important propagation processes, and that termination is by 7 and 8, then a steady state treatment leads to a third-order equation:
28
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978
k$8[*CH2OHl3 + (k&8[N20] + k&7)[.CHz0Hl2 - ksI[.CH20H] - k J = 0 (12) where I = (DG(int)/lOONA) is the rate of initiation of the chain process, G(int) = G(-HzO) is the G value for initiation, assumed equal to 5 for this work, D is the dose rate in eV/L s, and NA is Avogadro's number. Equation 12 can be solved for the C H 2 0 H concentration by an iterative method using a computer, if values can be assigned to the rate constants. The value of k7,the rate constant for the recombination and disproportionation of two -CH20H radicals, is 1.2 X lo9 L/mol s at room temperaturens For reasons given previously,2 1 X 1O'O L/mol s is a reasonable estimate for k7 a t 300 "C. The rate constant for the combination and disproportionation of CHzOH and .OCHZOH is assumed to have a value of k8 = 1 X 1O1O L/mol s. The magnitudes of k3 and h4 are estimated from the experimental data. The line drawn through the data in Figure 1 is calculated using k , = 2.5 X lo4 L/mol s and k4 = 38 L/mol s. The concentration of water at 300 "C was taken as 39 M in all the solution^.^ First the .CH20H concentration is determined using eq 12 and then the G(NJ and G(CH20) values are obtained from
The shape of the curve in Figure 1 can be understood by examining reactions 1-4 and 6-8. If one considers 8 to be the only mode of termination at higher nitrous oxide concentrations, -0.1 M, then use of the long chain approximation leads to G(CH20) = G(N2) =
Equation 14 predicts that G(N2)plotted against the square root of the nitrous oxide concentration should give a straight line through the origin. From Figure 1 it appears that one would have to study higher nitrous oxide concentrations to see such an effect. This conclusion is supported by the fact that the calculated curve obtained from 12 and 13 agrees with eq 14 for nitrous oxide 21 M. At low nitrous oxide concentration, -0.01 M, there is a curvature of the data back toward the origin. For small concentrations of N20 reaction 3 will occur at a lower rate, so that the .CH20H concentration is increased and termination by reaction 7 is more probable. Considering reaction 7 to be the only termination reaction leads to
05) The bending of the experimental data for [Nz0]1/2 0.1 M1/2in Figure 1indicates a first-order dependence on N 2 0 concentration in this region, in agreement with eq 15. Figure 2 shows a plot of G(Nz)and G(CHzO)against the methanol concentration. The plot demonstrates that the system is not sensitive to methanol concentration over a concentration range (0.5+0.05) M and this result is the reason why reaction 5 was considered to be unimportant in this system. The decrease in the yields below 0.05 M methanol, Figure 2, is probably because the methanol concentration is not large enough to scavenge all the radicals in the spurs. This conclusion is supported by the fact that the yields are again independent of methanol
T. G. Ryan, T. E. M. Sambrook, and G. R. Freeman I
200,
I
I
I
t
t
I
101'7
I I8
I
I 19
I
J
I 20
IOOO/T(K)
Figure 4. Arrhenius plot of chain product yields. [N20] = 95 f 5 mM, [CHsOH] = 0.53 f 0.03 M: 0, G(N2); A, G(CH2O).
concentration between 1 and 9 mM, Figure 2. The nitrous oxide concentration study is not complicated by the above effects because the methanol concentration used was 0.53 f 0.03 M. This is the region where the yields are independent of the methanol concentration, Figure 2. Figure 3 contains a plot of G(Nz) and G(CH20)against D-'/'for a methanol concentration of 0.53 ak 0.03 M and 96 f 5 mM nitrous oxide. As mentioned above these are conditions where eq 14 applies. The linear dependence of G(N2) and G(CH20) on D-lI2,predicted by eq 14, is born out by experiment, Figure 3. However, an intercept of 50 G units is obtained from the experimental data and eq 14 predicts an intercept of zero. This discrepancy is not understood. Arrhenius plots for nitrogen and formaldehyde are given in Figure 4. The conditions for the temperature study were 96 f 5 mM in nitrous oxide and 0.53 f 0.3 M methanol, so that eq 14 is applicable. The lines are somewhat curved, which indicates that the propagation steps have different activation energies. The average slope gives '/z(E3 + E4 - E,) i= 9 kcal/mol. Reaction 8 is assumed to be diffusion controlled so that its activation energy is about 3 kcal/mol. This means that E3 E4 z 21 kcal/mol. It is not possible from the present work to determine individual values for E, and E4. The yield of N2 under all conditions studied was higher than the yield of formaldehyde (Figures 1-4), while the mechanism predicts that the Nzand CHzO yields should be the same. Radiolysis of water containing nitrous oxide in the absence of methanol at 573 K yielded a G(Nz) of -9, Figure 1. This could account for some of the difference between Nz and CH20 yields from samples containing methanol. The remaining difference is most likely due to consumption of CHzO by reaction. A sample containing 9 X M CH2O and 96 f 5 mM N 2 0 was irradiated at 573 K. All the formaldehyde was consumed and 87 G units of Nz and 7 G units of CO were produced. The consumption of CHzO represents G(-CH20) = 35, so about two molecules of N2 result from the consumption of one C H P Omolecule. Formaldehyde is consumed by a chain reaction. To explain the difference between the Nz and formaldehyde yields requires that 8% of the initial CH20 formed reacts further to produce nitrogen. Comparison between the Methanol and 2-Propanol Systems. Table I contains a summary of the rate constants for the methanol and 2-propanol systems. The value of k3 is two orders of magnitude larger in the 2-propanol than in the methanol system. The value of k5 is probably also greater for 2-propanol than for methanol which means for methanol. However, reaction 5 was not k 4 / k 51 observed in the methanol system, because [H20]/ [CHSOH] 2 lo2. By contrast the variation of G(N2) with [2-
+
The Journal of Physical Chemistry, Vol. 82, No. 1, 1978 29
Hydrophobic Hydration of Tetraalkylammonium Bromides
TABLE I:
R a t e C o n s t a n t Values at 573
R=H
Reaction ( 3 ) N,O t .CR,OH + N, t *OCR,OH (4) --OCH,OH* t H,O + HOCH,OH t O H ( 5 ) +OCH,OH t HCR,OH HOCR,OH t .CR,OH (7) B.CR,OH + P, (8) .CR,OH t *OCR,OH (9) B.OCR,OH -+ P, a
2.5 X
lo4
+
R = CH, 2.9 X 10'
38
8.7 x
-+
-+
P,
1 x 10'Oa 1 x 10'O
io3
1x 1 x 1O'O
-
reactions in water a t 300 "C to be -1 X lo1" L/mol s, one obtains (AS*3 AS*5 - AS*8) = 0 in the propanol system and (AS*B-t &SI4 - ASS8) -32 cal/mol K in the methanol system. The values of AS* for reactions 3,5, and 8 are all probably near zero, which indicates a large negative value, --32 cal/mol K, for reaction 4. This supports the suggestion that solvent rearrangement (increase of solvation energy) makes a significant contribution to the driving force of reaction 4.
K
h(M-'s-')
a
References and Notes
Assumed.
p r ~ p a n o l ] l demonstrated /~ the occurrence of reaction 5 in solutions where [HzO]/[2-propanol] 2 lo4, hence we for 2-propanol. conclude that h4/k5 I Figure 4 in ref 2 and the present Figure 4 may be used to estimate the sums of the entropies of activation of the propagation reactions in the propanol and methanol systems. Using the reaction sequence in Table I, and taking the diffusion controlled limit of radical/molecule
(4) (5) (6)
(7) (8) (9)
Financially assisted by the National Research Council of Canada. T. G. Ryan and G. R. Freeman, J . Phys. Chem., 81, 1455 (1977). E. C. Bricker and H. R. Johnson, Ind. Eng. Chem., Anal. Ed., 17, 400 (1945). R. L. Shriner, R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds", Wiley, New York, N.Y., 1956. A. Kato and R. J. CvetanoviE, Can. J . Chem., 46, 235 (1968). J. G. Calvert and J. N. Pttts, "Photochemistry", Wiley, New York, N.Y., 1966, pp 824-826. Y. Ogata and A. Kawasaki in "Chemistry of the Carbonyl Group", Vol. 2, J. Zabicky, Ed., Interscience, Toronto, 1970, p 4. M.Sirnic, P. Neta, and E. Hayon, J. Phys. Chem., 73,3794 (1969). R. W. Gallant, "Physical Properties of Hydrocarbons", Vol. 2, Gulf Publishing Co., Houston, Tex., 1970, p 190.
Hydrophobic Hydration of Tetraalkylammonium Bromides in Mixtures of Water and Some Aprotic Solvents W. J. M. Heuvelsland, C. de Visser, and G. Somsen" Department of Chemistty, Free University, De Boelelaan 1083, Amsterdam, The Netherlands (Received July 26, 1977) Publication casts assisted by the Free University of Amsterdam
Enthalpies of solution of tetra-n-pentylammonium bromide in mixtures of water and N,N-dimethylformamide (DMF) and of tetra-n-butylammonium bromide in mixtures of water and dimethyl sulfoxide (Me2SO)and of water and N,N-dimethylacetamide (DMA) have been measured calorimetrically at 298.15 K over the whole mole fraction range. All profiles of the enthalpy of solution vs. solvent composition show endothermic maxima. The results are interpreted in terms of a simple hydration model with two parameters: the enthalpic effect of hydrophobic hydration in pure water, and the number of solvation sites of one alkyl group. After elimination of the influence of the Br- ion, both parameters are in good agreement with those found for some nonelectrolytes. Finally, the role of the cosolvent i s considered more systematically.
I. Introduction T h e hydrophobic hydration of larger tetraalkylammonium halides in water is strongly influenced by the addition of polar cosolvents. This is expressed among others by substantial changes of the enthalpies of dilution,l enthalpies of solution,24 and partial molar heat capacities5 of n-Bu4NBr when water is replaced as solvent by aqueous mixtures of N,N-dimethylformamide (DMF), dimethyl sulfoxide (MezSO), or dioxane. In earlier report^,^,^ we showed that the enthalpies of solution of n-Bu4NBr, nPr4NBr, and Et4NBr in mixtures of water and DMF can be interpreted in terms of a cooperative hydration model originated by Mastroianni, Pikal, and Lindenbaum.l From this cage model the experimental results can be described by two parameters, i.e., the number of solvation sites n surrounding an alkyl group and the enthalpic effect of hydrophobic hydration in pure water Hb(H20). In a recent study Lindenbaum, Stevenson, and Rytting7 have applied the same approach to some nonionic solutes. For the three tetraalkylammonium bromides mentioned before we found that both parameters increase with increasing number of 0022-3654/78/2082-0029$0 1.OO/O
C atoms. Since it might be expected that some leveling-off will occur a t larger chain lengths, we felt it desirable to extend our measurements to higher homologues. However, because of the slow rate of dissolution of n-Hex4NBr and n-Hep4Br in water we had to confine our measurements to n-Pen4NBr. As we have pointed out earlier4DMF is not essential for our model approach. According to the model the values of the parameters Hb(H20) and n ought to be independent of the choice of the cosolvent as long as the latter does not show specific interactions. In order to test the model on this particular point we have measured enthalpies of solution of n-Bu4NBr in mixtures of water and MezSO and water and N,N-dimethylacetamide (DMA). From the point of view of ion-solvent interactions the selected cosolvents MezSO and DMA are comparable with DMF. All three are dipolar aprotic solvents with Kirkwood correlation factors close to l,8v9 the same donor properties, and only slightly different acceptor pr0perties.l" This similarity in behavior is also expressed by the integral enthalpies of mixing with water. At 298.15 K DMF, 0 1978 American Chemical Society