8731
J . Phys. Chem. 1991,95, 8731-8737
-
evolution continues as j’increases, such that by (0,O) (0,9), the indirect estimates are in close agreement with the quantum ones, whereas the direct ones are systematically high. As we have discussed in connection with Figure 2, it appears empirically within the bin histogram method that when trajectories with low j ‘ (relative to the center of the bin) dominate the contributions to the direct upward cross section, we have a sufficient condition for QCT to overestimate the cross section near threshold. The other commonly used method of determining cross sections from quasiclassical trajectories is the smooth histogram method35of Truhlar et al., which in fact aggravates this type of overestimation of direct upward cross sections in the threshold region even further, since trajectories of even lower j’are allowed to contribute to the cross section to higher states. Where the barrier is not dynamically important, QCT estimation of upward cross sections indirectly appears to minimize the systematic effects due to binning. We have made the same observation for (0,O) (~’=l,j’=1,3,5).*~ Sometimes experimental and theoretical cross sections (or relative populations) are presented as a function ofj’for fixed values of E,,, and 0’. For example, see Figure 6 in Barg et al.,36 Figure 18 in Buchenau et al.,” or Figure 12 in Rinnen et which all compare upward direct QCT cross sections with quantum or experimental results. They show that the results from direct QCT are higher at (or extend to) higher j’. Examination of our sequence of Figure 1, parts a-e, in which j’increases from 1 to 9 will show the form of such a diagram for our results. Using E,,, = 1.1 eV as an example, we see that the direct QCT cross section is slightly lower than the quantum curve for the lower j’ but higher for larger j’, particularly for j’= 7,9. Thus our direct cross sections will also appear rotationally hot with respect to the quantum results. Note however that this can be alleviated by use of indirect cross sections which eliminates the overestimation at higher j ’ (fixed E,,,). DMBE Potential: QCT-Exact Quantum Comparison. Manolopoulos et aL2’ carried out the calculation of integral cross sections in the energy range 0.95-1.35 eV for (0,O) (0,j’= 1,3,5,7) on the DMBE surface. These are also plotted in Figure 1. For j ’ = 1, 3 their cross sections are consistently higher than
-
-
(35) Blais, N. C.; Truhlar, D.G. J . Chem. Phys. 1977, 65, 5335. (36) Barg, G.-D.;Mayne, H. R.; Toennies, J. P.J. Chem. Phys. 1981,74, 1017. (37) Buchenau, H.; Toennies, J. P.; Arnold, J.; Wolfrum, J. Ber. BunsenGes. Phys. Chem. 1990, 94, 1231. (38) Rinnen, K. D.; Kliner, D. A. V.; Zare, R. N. J . Chem. Phys. 1989, 91. 7514.
the corresponding quantum LSTH calculations. The interpretation is that since the collinear bamer on the DMBE surface is slightly lower than that on the LSTH surface, exchange occurs more easily. By j ’ l 5, their cross sections agree closely with the LSTH results. For these transitions the collinear barrier is less important dynamically and away from this barrier the two surfaces at low energy are very similar.39 The good agreement of exact quantum cross sections for (0,O) (v’=l, j’=l,3) bears this out.39 We have calculated QCT cross sections at a few energies using the DMBE surface (see Figure 1). These QCT results are shifted with respect to the corresponding QCT calculations using the LSTH surface by approximately the same amount that the two quantum results were. In this limited energy range somewhat above the threshold for exchange, it appears that tunneling contributions for the DMBE surface are not larger than those for the LSTH surface. Recommendations. Where the dynamic threshold is affected by the barrier height, both direct and indirect QCT appear to suffer from the effects of ZPE leak. For cases where the dynamic barrier is dominated by the energy difference between initial and final states, direct (upward) cross sections near threshold are overestimated by the histogram binning technique used, whereas the indirect ones are not. We therefore recommend that in the absence of quantum results the indirect cross sections should be adopted near threshold. This is consistent with the recommendations of Ashton et al.,27 arrived at through consideration of factors affecting reaction barriers, of Leasure et a1.28 in the consideration of dynamic thresholds, and of Blais et al.? arrived at through comparison of thermal rate constants with those from variational transition state theory. At higher energies where the indirect and direct cross sections agree within their statistical errors, a weighted average can be taken.
-
Acknowledgment. We thank J . Z. H. Zhang for kindly making available his unpublished quantum results, W. J. Keogh for help with the DMBE calculations, and P. W. Brumer and J. P. Valleau for useful discussions. We acknowledge the contributions of J. E. Dove, sadly now deceased, to the formative stages of this work. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada and the Connaught Fund of the University of Toronto. Registry NO. H,12385-13-6. (39) Auerbach, S. M.;Zhang, J. Faraday Tram. 1990,86, 1701.
Z. H.; Miller, W. H. J. Chem. Soc.,
Gas-Phase Aromatic Substitution. Reactivity of (Trifluoromethoxy)benzene toward Charged Eiecttophiies Fulvio Cacace,* Maria Elisa Crestoni, Annito Di Marzio, and Simonetta Fornarini University of Rome ‘La Sapienza”, P. le Aldo Moro, 5, 00185 Rome, Italy (Received: March 22, 1991: In Final Form: May 28, 1991) The reactivity of C6H50CF3toward gaseous cations, including C2H5+, i-C3H7+,(CH3)2F+, and CH30(H)N02+,has been investigated with mass spectrometric and radiolytic techniques. Two distinct kinetic patterns emerge, depending on whether formation of an early electrostatic adduct between the reactants is rate determining. In this case, which occurs in ethylation and isopropylation, the reactant displays positional but not substrate selectivity, e.g., i-C,H7+ is characterized by a kvpCF,/kw ratio of 1 .O and by a 77% ortho, 12% meta, and 11% para orientation. By contrast, methylation and nitration display both substrate and positional selectivity, e.g., CH30(H)N02+is characterized by a kCsH50CF3/kC6Hs ratio of 0.096 and by a 35% ortho, 5% meta, and 60% para orientation. Only the latter reactions lend themselves to construction of free-energy correlations, e.&, a up+value of +0.12 has been calculated for the nitration of CsHSOCFoby CH30(H)N02+in CH, at 37.5 OC, compared with the +0.07 value from the nitration by N02+BF4-in solution. The gas-phase reactions are compared with the solution-chemistry counterparts, and their distinctive features are discussed. The mechanism of electrophilic aromatic substitution has been the focus of sustained interest for over a century and a cornerstone
in the development of physical organic chemistry. Far from declining, the interest in this class of reactions has recently ’been
0022-3654/91/2095-873 1%02.50/0 0 1991 American Chemical Society
8732 The Journal of Physical Chemistry, Vol. 95, No. 22, 199I1 enhanced by new, or revitalized, mechanistic ideas, e.g., the concept of single-electron transfer has spurred the current upsurge of studies on aromatic nitration.' Among the "nonconventional" methodologies that have contributed to the recent advances, gas-phase approaches aimed at the mechanistic study of intrinsic, solvent-free reactivity patterns are gaining increased recognition, particularly in the case of charged electrophiles. Meaningful kinetic correlation between gas-phase and solution results presupposes however certain essential conditions, hardly met by purely mass spectrometric approaches. As a matter of fact, the only comprehensive set of data on the competitive rates and orientation of gas-phase ionic substitutions has been obtained with an integrated approach where mass spectrometric methods are complemented by radiolytic techniques, characterized by an extended pressure range and by the positive determination of the isomeric composition of the products.2 Whereas the systematic exploitation of the above techniques has provided a wealth of data on many aromatic substitutions occurring in the gas phase, currently regarded as useful comparison terms by the solution chemists,' perhaps the most significant result is a sharper definition of the conditions necessary to make the comparison meaningful. Of particular note is the strict requirement that the reacting species be thermally equilibrated with the bath gas, which is essential, inter alia, to a meaningful definition of the reaction temperature. Meeting such a requirement is often difficult, because fast reactions can occur before the excess internal energy imparted to the ion-molecule pair by the electrostatic interaction of the reagents is removed by thermalizing collision, even at the highest pressures of the bath gas experimentally accessible. This accounts for the interest attached to gas-phase aromatic substitutions of deactiuated substrates, characterized by a relatively large activation energy and hence by a relatively slow conversion into the products of the electrostatic ion-molecule complex, which allows collisional thermalization of the latter. Unfortunately, the mechanistic peculiarities of gas-phase ionic reactions have so far severely restricted the number of suitable substrates, in that many charged electrophiles react selectively with most deactivating groups, e.g., COOR, CN, NO, NO2, CHO, COCH3, etc., rather than with the rings4 In this paper we describe and discuss the reactivity of a deactivated substrate, C6HSOCF3,whose behavior is well characterized in solutionS*6and whose reactions with typical charged electrophiles, such as C2HS+,i-C3H7+,(CH3)2F+, and CH30(H)N02+ provide a veritable epitome of the alternative kinetic patterns of gas-phase aromatic substitutions.
(1) (a) Ebcrson, L. Electron Transfer Reactions in Or anic Chemistry, Springer: Berlin, 1987. (b) Eberson, L.; Radner, F. A&. them. Res. 1987, 20, 53. (c) Kochi, J. K. Angnv. Chem., Int. Ed. Engl. 1988, 27, 1227. (d) Olah, G.A.; Malhotra, R.; Narang, S.C. Nitrution;VCH: New York, 1989. (2) (a) Cacace, F. Acc. Chem. Res. 1988, 21. 215. (b) Cacace, F. StructurelReuctivity und Thermochemistryof Iom; Ausloos, P., Lias, S.G., Eds.; D. Riedel: Dordrecht, Holland, 1987; p 467. Related studies on gasphase aromatic substitution: (c) Cacace, F.; Cipollini, R.; Giacomello, p.; Pwsagno, E. Gurr. Chim. Itul. 1911104,977. (d) Attin&,M.; Cacace, F.; Yaiiez, M. J. Am. Chem. Soc. 1987, 109, 5092. (e) Fornarini, S.J . Org. Chem. 1988, 53, 1314. (0 AttinH, M.; Cacace, F.; de Petris, G.Angew. Chem., In?. Ed. Engl. 1987, 26, 1177. (8) AttinH, M.; Cacace. F.; de Petris, G.;Grandinetti, F. In?. J. Mass Spectrom. Ion Processes 1989,90, 263. (h) Attin&,M.; Cacace, F.; Crestoni, M. E.; de Petris, G.;Fornarini, S.; Grandinetti, F. Can. 1.Chem. 1 M , 66,3099. Recent studies on heteroaromatic gas-phase substitution: (i) Crestoni, M. E.; Fornarini, S.; Spaanza. M. J. Am. Chem. SOC.1990, II2,6929. (1) Angelini, G.; Sparapani, C.; Speranza, M. Ibid. 3060. (3) Taylor, R. Electrophilic Aromatic Substitution; Wiley: New York, 1990. (4) E.g., extensive mass spectrometric and radiolytic evidence shows that gaseous cations react with the substituent of strongly deactivated substrates such as C6HIN02 and C,H,CN; see: (a) Kruger, T. L.; Flammang, R.; Litton, J. F.;Cooks, R. G.Tetrahedron Letr. 1976, 50, 4555. (b) Cacace, F.; Ciranni, G.; Giacomello, P. J . Am. Chem. Soc. 1982, 104, 2258. (c) Burinsky. D. J.; Campana, J. E. Org. Mass Spectrom. 1988, 23, 613. (5) Olah, G, A.; Yamato, T.; Hashimoto, T.; Shih, J. G.; Trivedi, N.; Singh. B. P.; Rteau, M.; Olah, J. A. J . Am. Chem. Soc. 1987, 109, 3708. (6) Deebois, M. Bull. Soc. Chem. Fr. 1986,855.
Cacace et al. Experimental Section Materials. The gases used were research-grade samples from Matheson Gas Products, Inc., whose stated purity exceeded 99.9 mol %, and were used without purification. CH3F from commercial sources was invariably contaminated by appreciable amounts of (CH3)20,ranging from 0.2 to 1 mol %, as shown by analyses carried out by GC using a 2-m-long, 3-mm4.d. glass column packed with 60/80 mesh Carbopack B+1% SP-1000, a modified polyethylene glycol from Supelco Co., operated isothermally at 80 OC. Purification of CH3F by the freezing-thawing technique in the presence of SA molecular sieves in a vacuum line reduces appreciably but not entirely the (CH3)ZO content. It should be noted that, even at the level of 0.1 mol %, (CH3)20still competes efficiently with the aromatic substrates, especially if deactivated, owing to their correspondingly low concentration in the radiolytic experiments. (Trifluorometh0xy)benzenewas obtained from Aldrich Chemicals Co. with a stated purity of 99 mol % and was used without further purification. The other chemicals used were obtained from commercial sources or prepared according to standard procedures, checking their purity on the same columns used for the analysis of the irradiated systems. The isomeric (trifluoromethoxy)toluenes, required as reference standards for the identification of the products from the methylation of (trifluoromethoxy)benzene, were synthesized as described by Olah and co-worker~.~ The isomeric (trifluoromethoxy)cumenes were obtained directly, although in minute amounts (ca. 0.2 mg), from the radiolytic isopropylation of (trifluoromethoxy)benzene,run to a higher substrate conversion than in typical kinetically oriented experiments. The products mixture was recovered by freezing the irradiated vessel in liquid nitrogen and repeatedly washing its walls with ethyl acetate. The solution was then concentrated, and the products, after removal of the large excess of unreacted trifluoromethoxybenzene, were isolated and purified by preparative GLC, using a 4-m-long column, packed with Carbowax 20M+ KOH (2%) on Supelcoport. The isomers thus separated were dissolved into CD2C12and identified by 'H NMR spectroscopy, using a Varian XL 300 instrument, with the deuterium signal of the solvent as the lock and TMS as reference frequency. CI Mass Spectrometry. The spectra were recorded on a Hewlett-Packard Model 5892A quadrupole instrument, connected to a Model 5934A data system and whose CI ion source could be fitted with a Bourdon-type mechanical gauge to calibrate the readings of the ionization manometer of the spectrometer. Owing to the many sources of systematic errors that are known to affect absolute pressure readings, the values reported, albeit reproducible and internally consistent, are only indicative of the order of magnitude of the pressure in the ion source. The CI spectra were run at 80 "C ion-source temperature. Irradiation. The gaseous reaction mixtures were prepared by introducing the appropriate amounts of the bulk constituent (CH,, C3H8, CH'F) and of the required additives (CH30N02, 02, (CH3)2CO), together with fragile glass ampules containing the aromatic substrates, into evacuated and carefully outgassed Pyrex bulbs (125 mL) connected to a greaseless vacuum line. After cooling to 77 K, the bulbs were sealed off, the fragile ampules were broken, and sufficient time was allowed for complete mixing and thermal equilibrium of the gaseous systems before the radiolysis, performed at 37.5 OC in a 220 Gammacell from Nuclear Canada Ltd. to a total dose of 2 X lo4 Gy, at a dose rate of 5 x 103 Gy h-l. The products were analyzed by GC, using a Model 8700 instrument from Perkin-Elmer Co. and a Model 5700A instrument from Hewlett-Packard Co., or by GC/MS, using a Model 5890 gas chromatograph connected to a Model 5970B mass-selective detector from Hewlett-Packard Co. The following columns were used: (i) A 50-m-long, 0.2-mm4.d. fused-silica capillary column, coated with a 0.5-"-thick cross-linked methylsilicone film (PONA column from Hewlett-Packard), operated isothermally (2 min) at 50 OC and then heated at rates of 6 OC min-' to 150 OC and of 15 OC min-I to 160 OC;(ii) a 30-m-long, 0.25-i.d. polyethylene glycol (Supelcowax 10) bonded-phase column (0.25-pm film thickness) from Supelco Co., operated isothermally ( I min) at 60 OC, and then raising the temperature
(Trifluoromethoxy)benzene Reactivity
0
The Journal of Physical Chemistry, Vol. 95, No. 22, 1991 8133
100 mlz
0
200
Figure 1. CH, CI mass spectrum of C6HSOCF3(=S): 1, m / z = 29, C2Hs+;2, m / z = 41 C3HS+;3, m / z = 143 (S - F)+; 4, m / z = 163, (S + H)'; 5, m / z = 191, (S+ CzH?)+. The large CHS+peak at m / z = 17 and low-intensity isotopic satellites are not shown.
100
4
1
i%
2
0 0
100
m/z
200
300
Figure 2. C3H8CI mass spectrum of C6HSOCF3:I, m / z = 43, C3H,+; 2, m / z = 162, S+;3. m / z = 163, (S + H)+; 4, m / z = 205, (S + C3H7)+.
to 150 OC at a rate of 4 OC mi&; (iii) a 30-m-long, 0.53-mm4.d. Vocol bonded-phase column (3-pm film thickness) operated isothermally (3 min) at 50 OC, and then raising the temperature to 140 OC at a rate of 4 OC min-I. The preparative separation of (isopropy1fluoromethoxy)benzenes was carried out using a Model ATC/f, Series 400 gas chromatograph from C. Erba Co., equipped with a hot-wire detector. Results Chemical Ionization (CI) Mass Spectrometry. A typical CI mass spectrum of C6H50CF3(4) recorded in CH4 (92 OC, 1 Torr) is reproduced in Figure 1. The ions arising from the aromatic substrate include the (S- F)+ fragment, the protonated adduct (S+ H)+, and the alkylated adduct (S + C2H5)+. The C3H8CI spectrum, reported in Figure 2, is somewhat different, being characterized by a much higher abundance of the alkylated adduct, (S + C3H7)+,which predominates over the protonated substrate (S+ H)+ by nearly an order of magnitude. Finally, a spectrum obtained by using as the reactant gas a dilute solution of C H 3 0 N 0 2in CH,, whose composition was adjusted to maximize the abundance of the CH30(H)N02+nitrating cation, is reproduced in Figure 3. In addition to the (S- F)+, (S+ H)+, and (S+ C2Hs)+peaks from the reactions of the CH5+and C2H5+ ions from CH4,the spectrum displays the expected nitrated adduct, (S + NOz)+,at m / z = 208. It should be noted that the CI spectra depend appreciably on the experimental conditions, especially on the pressure and the temperature of the ion source. Furthermore, quantitative correlation of the C l spectral data with the results of the radiolytic
-
loo
m/z 200
300
Figure 3. CH4/CH3ONO2mass spectrum of C6H50CF3:1, m / z = 78, (CH30(H)N02)+;2, m / z = 123 (CH30NOz+ NO2)+;3, m / z = 143, (S - F)+; 4, m/z = 163, (S + H)+; 5, m/r = 191, (S + C2Hs)+;6, m / z = 208, (S NO2)+.
+
experiments is difficult, owing to the widely different conditions that characterize the two sets of experiments, particularly as concerns the much higher pressures typical of the irradiated systems. These qualifications do not detract, however, from the significance of the mass spectrometric results,which provide direcr evidence for the formation of charged species corresponding to the intermediates of all electrophilic substitutions investigated, i.e., the (S+ C2H5)+, (S + C3H7)+,and (S + NO2)+ adducts. Another feature of the CI spectra is of interest, Le., the predominance of alkylation over protonation noted in the case of i-C3H7+, which contrasts with the behavior of C2H5+ and suggests that the proton affinity (PA) of C6H50CF3does not considerably exceed that of C3H6, 179.5 kcal m01-I.~ Such inference is reasonable if one takes into account the PA of unsubstituted benzene, 181.3 kcal mol-',7 and the effect of a powerful electron-withdrawing group, such as CF30. Radiolytic Reactions. The reactions, except for a few experiments performed at 100 "C, have been carried out at 37.5 OC, at pressures ranging from 695 to 1450 Torr. The composition of the irradiated systems, the absolute yields, and the isomeric composition of the products as well as the substrate selectivity of the gaseous electrophiles deduced from competition experiments are reported in Table I. The absolute yields of the products are expressed by their G+M values, Le., the number of molecules formed/100 eV. The internal consistency of the G values is satisfactory (standard deviation below lo%), and their absolute accuracy, affected by dosimetric problems and other sources of systematic errors, is estimated around 30%. Nevertheless, consideration of the absolute yields is useful, showing in the first place that aromatic substitution is a major reaction pathway of the gaseous electrophiles investigated. The absolute yields depend, in general, on the composition of the irradiated systems, where the aromatic substrates compete for the electrophile and with other nucleophiles, either deliberately added to the system (vide infra) or formed from its radiolysis. In those systems that contain no added bases and where the concentration of the aromatic substrates is relatively high, their substitution products account for a large fraction of the ionic reactants. Consider, as an example, the competitive ethylation of C6H50CF3(4.5 Torr) in the presence of C6H6 (1.4 Torr), used as the reference substrate, yielding isomeric ethyltrifluoroanisoles and ethylbenzene, whose combined G values amount to ca. 0.21-0.25, whereas the G value for the formation of the C2H5+ reactant from the ionization of CH, in the pressure range of interest is 0.9 f 0.2.* It should be noted, in this connection, that alkylation is but one of the reaction (7) Unless specified otherwise, all thermochemical data are taken from: Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J . Phys. Chem. ReJ Dara 1988, 17, suppl. 1. (8) Ausloos, P.; Lias, S.G.; Gorden, R., Jr. J. Chem. Phys. 1%3,38,2207; 1963, 39, 3341.
8734 The Journal of Physical Chemistry, Vol. 95, No. 22, 1991
Cacace et al.
TABLE I: C.o-phse RcrctiOnr of C&pCF, wtth Cbuged EketropBites
Alkylation by C2H5+in CH4" substrate and positional selectivity
products yields (G+,,,)
-
system composition, Torr
PhOCFI
PhH
1.2 1.4 0.9 1 .o 4.5 1.7
1.7 1.5 1.3 6.7 1.4
kPbOCF~/ kphH
MeEO 0.15 0.05 0.083 0.21 0.079 0.12
0.063 (54:27:19) 0.032 (58:26:16) 0.045 (54:28:18) 0.022 (58:26:16) 0.136 (52:27:21) 0.145 (55:26:19)
2.1 0.8
1.o
0/2P 1.4 1.8 1.5 1.8 1.2 1.4
0.60 0.69 0.78 0.70 0.54 0.7 I
m/2p 0.7 1 0.81 0.78 0.81 0.64 0.68
Alkvlation bv I-C2H,+ in CIH,* I
PhOCF3
PhX
x=F
1.4c
1.2 1.3 1.9 X = H: 1.8 1.6 I .2 6.0 0.8
1 .5d
1.2 1 .o
1.3 1 .o
1.3 1.75
I
0.75 1.80
0.94 (7017:13) 0.85 (72:1612) 0.93 (721910) 0.61 (79:13:8) 0.53 (77:12:11) 0.41 (77:12:11) 0.49 (75:13:12) 0.60 (76:13:11)
0/2P 2.7
k m 1.10 0.90 0.98 1.10 1.00 0.91 1.05 0.91
0.73 (73:11:16) 0.82 (807:13) 1.51 (83:6:11) 1.o
0.65 0.54 2.18 0.30
3.0 3.7 4.9 3.5 3.5 3.1 3.4
m/2P 0.65 0.67 0.75 0.81 0.54 0.54 0.54 0.59
Alkylation by (CH3),F _ _ in CHIP PhOCF3
PhH
1.3 1.6
1.8
1 .o
2.2 I .3 3.8
1.o
Me2C0
CF.0-Q
0.035 (77:14:9) 0.038 (76:13:11) 0.031 (78:12:10) 0.066 (70:19:11) 0.037 (75:14:11) 0.11 (64:21:15)
4.3 0.45
1.3 1.1 8.1 0.97
kPWCFi/
(o:mp)
Me
PhMe
kPbH
0.37 0.20 0.31 0.37
0.13 0.12 0.13
f
f
0/2P 4.3 3.4 3.9 3.2 3.4 2.1
0.090
0.31
0.091
mI2P 0.78 0.59 0.60 0.86 0.64 0.70
Nitration by MeO(H)N02+in CH4* C F 3 0 a
PhOCF3
PhH
I .3
2.3
(on#
No2
0.087 (35560)
~,FJ
PhN02 1.6
kPLH 0.096
0/2P 0.29
m/2p 0.041
'The gaseous systems contained CHI (700 Torr) and O2(IO Torr), irradiation temperature 37.5 "C. *The gaseous systems contained C3H, (700 Torr) and O2 (IO Torr), at 37.5 "C. eIrradiation performed at 100 OC. dC3H8pressure 1440 Torr. 'The gascow systems contained CH3F (680 Torr) and O2 (IO Torr) at 37.5 "C. fNot measured. #The gaseous systems contained CH, (700 Torr), CH30N02(20 Torr),and O2 (10 Torr). channels of the gaseous C2HS+ions, since protonation of the aromatic substrates is energetically allowed as well and is found to occur efficiently under CI conditions. The same considerations apply to isopropylation, whose cumulative yield, measured in the absence of added bases in systems containing relatively high concentrations of the aromatic substrates, accounts for some 90% of the i-C3H7+ions formed from the ionization of p r ~ p a n e . ~ Whereas nitration by gaseous CH30(H)N02+ions is also quite efficient, methylation by (CH3)2F+is characterized by relatively low absolute yields, the combined G+" values of the methylated a r e n a approaching only 0.5. The singularity is explained by the presence as an impurity in the CH3Fgas of a gaseous nucleophile, (CH3)20,which competes with the aromatic substrate(s) for the (CH3),F+ reactant, converting very efficiently the latter into unreactive (CH3)20+ions, as shown by CI mass spectrometry of CH3F/(CH3)20mixtures. Even at the lowest concentrations of (CH3)20attainable in practice, the ether represents a significant sink of (CH3)2F+ions, which accounts for the low absolute yields of methylated aromatics measured in these systems. Discussion The formation of the charged electrophiles upon ionization of the gaseous precursors used in this work, Le., of C2H5+from CH4, (9)(a) Ausloos, P.; Lias, S . G. J. Chem. Phys. 1%2,36, 3163. (b) Lias, S. G.; Ausloos, P. J . Chem. Phys. 1%2 37.877. (c) Sandoval, 1. B.;Ausloos, P.J . Chem. Phys. 1%3,38, 2452. (d) Ausloos, P. Ion-MoleculeReactions; Franklin, J. L., Ed.;Plenum: New York, 1970.
i-C3H7+from C3HB,(CH3)zPfrom CH,F, and CH3O(H)NOZ+ from a dilute solution of C H 3 0 N 0 2in CH,, is well established as a result of extensive mass spectrometric and radiolytic studiesu9 and requires no further discussion here. It should be noted only that the charged electrophiles, E+,formed in the irradiated gases undergo many unreactive collisions with the parent molecules and are effectively thermalized before their reactive encounter with the aromatic substrate, S,which is highly diluted in the system. As to the mechanistic aspects of the gas-phase aromatic s u b stitution, the present results conform to the following reaction model. Under the conditions prevailing in the radiolytic systems, the ions of interest can be expected to be associated with the bath-gas molecules, e.g., extrapolation of equilibrium data from high-pressure low-temperature CI experiments indicates that isopropyl ions exist in propane gas a t 310 K, 720 Torr as (iC3H7+--C3H8)electrostatic complexes.l0 The aromatic substitution can be envisaged as a two-step sequance, involving preliminary displacement of the bulk-gas molecule M by the substrate giving an early adduct, stabilized by purely electrostatic interactions, and probably best described from the structural standpoint as a 7r complex: (E+.-M)
k + S& (E+-S) + M 1 k-i
(1)
(10) Sunner. J. A,: Hirao. K.; Kebarle, P. J. Phys. Chem. 1989,93,4010. The binding energy of f-C3Hf to C3Hs is 13.6 kcal mol-'. The adduct of (CH3)2F+with CH3F is probably even more stable.
The Journal of Physical Chemistry, Vol. 95, No. 22, 1991 8735
(Trifluoromethoxy)benzene Reactivity
I
I
k,
[M--E+]+ PhX a
M + IphX- -E+] k,
k.1
.
a-complex
Figure 4. Energy profile of gas-phase addition of cationic electrophiles to aromatic substrates following rtspectively kinetic pattern I (a) and I1
(b).
The adduct can undergo back dissociation upon collision with M, or evolve into a covalently bound arenium ion (a complex): H-E
1
A k @x
The above scheme applies strictly only to C2H5' and i-C3H7+, which undergo addition to the aromatic substrate, but it can easily be adapted to those cations, such as (CH3)2F+and CH30(H)NO2+,that undergo nucleophilic displacement of a neutral species, CH3For CH,OH, by the aromatic substrate. Depending on the reactivity of the electrophile and on the activation of the substrate, two limiting kinetic patterns can be identified: (1) In those cases where the barrier E' to the formation of 2 is lower than the Eo barrier to back dissociation of 1, which corresponds to the difference of the binding energy of E+ to the substrate and to the bulk-gas molecule, formation of 1 becomes rate determining and !he reaction displays positional but not substrate selectivity (Figure 4a). (11) In those cases where E' > Eo (Figure 4b), formation of any electrostatic adducts 1 is not rate determining, and the reaction displays both substrate and positional selectivity, obeys standard solution-chemistry kinetics, and lends itself to the construction of free-energy correlations. For a given substrate, the kinetic pattern depends on the reactivity of the charged electrophile as well as on the nature of the bath gas, the factors that determine the balance between E' and E". Such expectation is fully borne out by a comparative evaluation of the selectivity of different cations in their gas-phase reaction with C6H50CF3. The isomeric composition of the neutral end products, whose formation is traced to the deprotonation of the arenium ions 2 by a gaseous base, has been used to evaluate the positional selectivity of the electrophile, whose substrate selectivity has been deduced from competition experiments carried out using C6H6as the reference substrate. Such estimates are based on the following assumptions, whose validity has been demonstrated in a number of related studies2 first, the gas-phase substitutions are first order with respect to the aromatic substrates; second, intramolecular and/or intermolecular isomerization of ions 2 is not significant: third, deprotonation (eq 3)
is fast, and in any case not rate determining. The validity of the above crucial assumptions in the case of interest has been checked
within the experimental range accessible to radiolytic experiments. Thus, the kC6HsOCF,/kC6H6 ratios of the various substitutions have been found to be insensitive to systematic variations of the relative concentration of the substrates, changed by a 20-fold ratio in ethylation, by a 10-fold ratio in isopropylation, and by a 5-fold ratio in methylation. The isomeric composition of the products is not significantly affected by the presence of acetone, a strong gaseous base having a PA of 196.7 kcal mol-l.7 This speaks against the significance of intramolecular isomerization that should be depressed by acetone, acting as a base and/or as a nucleophile. The insensitivity of the kC6HsOCF,/kC6H6 ratios to the presence of added acetone suggests that the deprotonation step (eq 3) is not rate-determining, As to intermolecular isomerization of ions 2, e.g., by transalkylation processes, it should also be affected by the presence of the base as well as by changes of the [C6H,0CF3]/[C6H6] ratio, two factors that have been found ineffective in causing significant modifications in the isomeric composition of the products. In conclusion, the experimental evidence suggests that the selectivity of the gas-phase substitutions investigated is not appreciably affected by secondary isomerization. Having disposed of the preliminary aspects, we can now prcceed to examine the specific reactions investigated. Ethylation. Gaseous C2H5+,HOf= 21 5.6 kcal mol-',' is known as an extremely reactive and indiscriminant reactant, lacking both substrate selectivity, e.g., kC6H5CH,/kC6H6 = 0.8 in CH,, and positional selectivity, i.e., the 2 para/meta ratio is as low as 1.3 in the ethylation of toluenenk In the lack of the thermochemical data required to evaluate the energetics of the reaction of C6H50CF3with C2H5+,as, for that matter, with the other electrophiles of interest, it is useful, for comparison purposes, to mention that alkylation of C6H6 by free C2Hs+ions is exothermic by ca. 44 kcal mol-I.lI Consistent with the high reactivity of the electrophile, ethylation of C6H50CF3displays very low substrate and positional selectivity, reflected by a kCdsOCF/kC6H6ratio of 0.7 f 0.1 and by an extent of meta substitution as high as 27%. We do not attach any special significance to the small substrate discrimination apparently displayed by CZH5+,a reactant that has previously been found to alkylate benzene faster than toluene. We regard instead gas-phase ethylation as a process lacking substrate selectivity, tracing the small deviation from unity of the k , F3/kCbH6 ratio to factors other than the intrinsic reactivity of%e aromatics, especially to the different branching ratio of alkylation and protonation, the two competitive reaction channels open to C2H5+. The simultaneous lack of substrate selectivity and the small, yet well measurable positional discrimination of C2H5+allow one to assign ethylation the kinetic pattern I. The only feature where the specific influence of the gaseous reaction environment can be discerned is the relatively pronounced ortho orientation, measured by an ortho/2 para ratio of 1.5. As a comparison, alkylation of C6H50CF3by C2H5F/BF3in excess of the liquid substrate is also characterized in solution by a low positional selectivity (3 1.4% meta substitution), but its ortho/2 para ratio is as low as 0.28.5 As to the directing properties of the OCF3 substituent, it behaves in gas-phase ethylation as an ortho-para directing group, although of very low selectivity. Isopropylation. Gaseous i-C3H7+,Hof = 190.9 kcal mol-l,' is also a highly reactive and indiscriminate electrophile, lacking ratio ranges from 0.9 substrate selectivity, e.g., the kC6HsCH3/kC6H6 (1 1) The LUPvalues of the alkylation processes have been estimated from the data of ref 7 , assuming that the PA of the ipso carbon of the alkylated substrate is equal to that of an unsubstituted position of benzene, cf.: (a) Devlin, J. L., 111; Wolf, J. F.;Taft, R.W.; Hehre, W. J. J . Am. Chem. Soc. 1976, 98, 1990. In the case of the methylation, the AHo value has been calculated from the methyl cation affinity of CHIF reported by: (b) McMahon, J. B.; Heinis, T.; Nicol, G.; Hovey, J. K.;Kebarle, P. J . Am. Chem. Soc. 1988,110,7591 and of C6H6,taken from: Sen Sharma, D.K.;Kebarle, P. J . Am. Chem. Soc. 1982,104,19, corrected for the PA of the 'ipso" position of toluene, assumed equal to that of benzene; see ref 1la. Since the calculated values refer to free ions, they represent the upper limit to the exothermicity of the alkylation. Accurate calculations would require knowledge of the binding energies of both the electrophile and of the arenium ion to the bath-gas molecules, the latter term being certainly smaller but not negligible.
8736 The Journal of Physical Chemistry, Vol. 95, No. 22, 1991
Cacace et ai.
TABLE II: SeiectiUity of SabstiMfolr Rmctioos of C&@CF, by Different Ekcbopbiks rad Comp.risoa with H.lokunnes orientation substrate C6H5mF3 C6HJF
C6HsCI C~HSOCF, C6HJF
C6HsCI C6HJOCF3 C6H5F
C6HsCI C6H50CF3 C6H5F C~HSCI
reactant
medium
NOZ'BF4N02'BFL N02'BFL AIC13/i-C3H7Br AIC13/i-C3H7Br AIC13/i-C3H7Br
CH3N02 CH3N02 sulfolane CH3N02 CH3NO2 CH3NO2
CH,O(H)NOZ+ CH30(H)N02+ CH,O(+H)N02+ i-C H i-C:Hi+ i-C3H7+
CHI CHI CHI C3Hs C3Hs C3H8
kGHflIkGH6
Solution 0.19 0.45 0.14 0.03 0.23 0.10 Gas Phase" 0.096 0.15 0.19 b b b
ortho 11.5 8.5 22.1 28.5 41.8 49.8 35 14 36 77 80 77
mea%
para%
ref
0 0 0.7 8.4 1.9 7.9
88.5 88.5 76.6 63.1 56.3 42.3
5 5 5 5 5 5
5 13
60
IO
54
12 7 12
11
this work lb lb this work
13
17
11
17
73
@Reactionscarried out at 37.5 OC, 760 Torr. *The measured reactivity ratios are close to unity, since these reactions follow kinetic pattern I; see text.
to 1.1 in C3Hs,and displaying a positional selectivity slightly higher than that of C2H5+,e.g., the 2 para/meta ratio is 2.7 in the isopropylation of toluene.12 Isopropylation is considerably less exothermic than ethylation, e.g., AHo -24 kcal mol-' in the reaction of free i-C3H7+ions with C6H6.11 lsopropylation of C6H@CF3 follows kinetic pattern I, being characterized by the complete lack of substrate discrimination (kc,H,,FJk,H, = 1.0 f 0.1) and by a positional selectivity that is significant by gas-phase standards. In fact, ortho-para substitution amounts to ca. 86%, and there is a sharp bias in favor of ortho substitution. Both features are typical of the gaseous reaction environment, as revealed by comparison with the selectivity measured in solution, where alkylation of C6H50CF3by i-C3H7Br/AlC13in CH3N02is characterized by a much higher H ~ F = 0.03), and by an substrate discrimination ( ~ c ~ /kC6H6 ortho/2 para ratio as low as 0.22.$ Metbylation. The reactivity of (CH3)2F+, HOf = 147 kcal mol-'? toward toluene is comparable to that of i-C3H7+,as shown inter alia by the ratio of 0.7 measured in CH3F gas and by a 2 para/meta ratio of 2.5.13 The energetics of isopropylation and of methylation of aromatic substrates are also similar, e.g., the exothermicity of methyl transfer from free (CH3)2F+ to CsH6 is ca. 29 kcal mo1-l.l' Methylation of C6H50CF3by (CH3)*F+marks the transition to the kinetic pattern 11, being characterized by an appreciable ratio ~ of 0.1 1. substrate selectivity expressed by a k ~ H , w F , / k c ~ H The orientation is also fairly selective, ortho-para substitution amounting to nearly 90%. with a strong bias for ortho methylation, measured by an orth0/2 para ratio of 3.4. The positional selectivity of the reaction displays a remarkable similarity with that typical of isopropylation. Unfortunately, no solution-chemistry data are available for comparison, since the attempted A1CI3-catalyzed alkylation of C6H50CF3 with CH3CI leads to predominant chlorine exchange with the OCF3 group.5 Nitration. The nitrating cation, CH30(H)N02+,Hof = 161 kcal mol-', is obtained from the protonation of methyl nitrate by CH5+and C2Hs+ions from CH4, both in CI mass spectrometry and in radiolytic systems." It reacts with aromatics as a moderately selective electrophile, characterized by a p+ value of -3.9 in CH, at 37.5 0C,zd,15 Among the reactions investigated the gas-phase nitration of C6HSOCF3displays the highest substrate selectivity and the most
=
selective orientation, measured by a bpn /+ ratio of 0.096 and by a 2 para/meta ratio as high as 24. kitration provides a typical example of the kinetic pattern 11, and hence its selectivity parameters can legitimately be used in free-energy correlations. In this connection, from the p+ value of the gas-phase aromatic nitration by CH30(H)N02+one can derive a u value of +0.12, which compares well with the only value avaihble for nitration by N02+BF4-in solution, +0.07.5 The agreement is reasonable if one considers the profound differences existing between the reaction media, which have led to the definition of a special set of u+ values for gas-phase reactions.I6 +
Coneludii Remarks The study of C6H50CF3allows a useful test of current concepts on gas-phase aromatic substitution, a clearcut discrimination between the kinetic patterns that characterize the behavior of a single substrate toward charged electrophiles of different reactivity and points to the conditions required for a meaningful correlation with solution kinetics. The selectivity of the gas-phase reactions investigated characterizes OCF, as a substituent displaying a negative inductive and a positive conjugative effect, whose combination brings about a deactivation of the substrate, accompanied by predominant ortho-para orientation, consistent with the solution-chemistry trend.5.6 The data illustrated in Table I1 show that the close similarity existing in condensed-phase reactions between OCF, and the halogens, especially F and C1, is even more striking in the gas phase, the ohly significant difference being the considerably higher extent of ortho alkylation occurring in the gas phase. Such a bias in favor of the positions ortho to substituents with unshared electron pain has been noted in other substrates, i.e., the hal~benzenes'~ and phenol and anisole,'* and traced to the strong demand for solvation typical of gas-phase ionic reactions, leading to localized electrostatic interactions between the n-type electrons of the substituent and positively polarized H atoms of the electrophile.'' Such a feature is particularly notable in isopmpylation and in methylation by dimethylhafonium ions, whereas other gas-phase substitutions, e.g., acetylati~n,'~ benzoylation,m nitration,2d and of course terf-butylation?' are characterized by 2 para/ortho ratios close to, or larger than, unity. The widely different ortho/para ratios typical of i-C3H7+and (CH3)J=+on one hand and of CH30(H)N02+on the other hand noted in this study support the view that predominant substitution ortho to groups containing unshared electron pairs, far from being
~
( 12) AttinB, M.; Cacace, F.; Ciranni, G.: Giacomello, P. J. Am. Chem. Soc.
1977,99,261 1 and refennoes therein. (13) (a) Cipollini, R.; Pepe,N.; Speranza, M.; Lilla. G. Guzz. Chim. frul. lwS,108,33. (b) Speram, M.; Pepe,N.;Cipollini, R. J. Chem. Soc., Perkin Trans. 2 1979, 1179. (14) Reference 2d. See also: Attinil, M.; Cacace, F. Guzz. Chim. fral. 1988,118,241. (IS) (a) AttinB. M.;Cacace, F. J . Am. Chem. Soc. 1986,108, 318. (b) Attini, M.;Cacace, F.;de Petris, 0. Angmt. Chem., fnt. Ed. Engl. 1987,26, 1177. (c) Attid, M.: Cacaa, F.; Ricci, A. Gars. Chim. ftul. 1989,119,217.
(16)See ref 3, p 477. (17) AttinH, M.;Giacomello, P. J . Am. Chem. Soc. 1979, 101, 6040. (18) AttinH, M.; Cacace, F.; Ciranni, G.; Giacomello, P. J. Chem. Soc., Perkin Trans. 2 1979,891. (19)Giacomello, P.; Speranza, M. J . Am. Chem. Soc. 1977,99,7918. (20) Occhiucci, G.; Cacace, F.; Speranza, M. J . Am. Chem. Soc. 1986, 108, 872. (21)A t t d , M.;Cacace., F.; Ciranni, G.; Giamello, P. J. Am. Chem. Soc. 1977,99,4101.
8737
. I Phys. . Chem. 1991, 95, 8737-8741
a distinctive and general feature of gas-phase aromatic substitution, occurs only in those cases where the electrophile fulfils certain requirements. Paramount among the latter seem to be the presence of H atoms carrying a substantial fraction of the positive charge, capable of establishing localized H bonds with the unshared electrons of the substituent, as well as a suitable geometry of the cation, allowing simultaneous interaction with the n-type and the *-type nucleophilic centers of the bidentate substrate, and finally the lack of steric hindrance to ortho substitution. All such features characterize to a comparable extent i-C3H7+ and
(CH3)2F+,the two gaseous reactants that consistently give the highest extent of substitution ortho to groups containing unshared electron pairs. Acknowledgment. We acknowledge the financial support of Italian National Research Council (CNR) and of Minister0 della Ricerca Scientifica e Tecnologica (MURST). Registry No. (Trifluoromethoxy)benzene, 456-55-3; ethylium, 14936-94-8; isopropylium, 19252-53-0 dimethylfluoronium,64710-12-9; methyl nitrate conjugate monoacid, 99573-80-5.
Local Density Functional Electronic Structures of Three Stable Icosahedral Fullerenes B. I. Dunlap,* D. W. Brenner, J. W. Mintmire, R. C. Mowrey, and C. T. White Theoretical Chemistry Section, Code 61 19, Naval Research Laboratory, Washington, D.C. 20375-5000 (Receioed: April 2, 1991)
Local density functional (LDF) electronic structures of icosahedral C,, Cleo, and C240are compared. These molecules are remarkably similar, and nothing is found to suggest that the two larger molecules are less stable than Cso. Ionization potentials are calculated by using both the transition-state approximation and differences between self-consistent-field calculations. Comparing these with one-electron eigenvalues supports the interpretation of photoelectron line shapes using theoretical cross sections calculated from LDF one-electron states of these large, highly symmetric molecules.
1. Introduction
Icosahedral (Ih) Csohas been isolated in macroscopic quantities in several laboratories.'-' Buckminsterfullerene is the smallest in an infinite series of sp2 carbon fullerenes* that are invariant under the 120 operations of I,,, the largest point group. The number of carbon atoms in the I,, fullerenes is given by the expression 20(i2 ij + J2), for j = 0 or i? Other values of j in this equation corresponds to fullerenes that are invariant only under the 60-operation point group, I, and thus not centrosymmetric. The clusters corresponding to ( i j ) = (1,O), C m and to (ij? = (2,0), Cso, are not closed-shell molecules in a Hilckel molecular orbital treatment of the .rr electrons.10 The HUckel closed-shell molecules considered in this work correspond to (ij? = (1,1), (3,0),and (2,2) and are Cm C I mand C m respeCtively.'O In these molecules each atom is symmetry-equivalent to at least 59 others. Therefore, the strain d a t e d with closing the molecules into a roughly spherical shape is distributed over a t least 60 atoms. Cleoand C240have larger radial dimensions than C,; therefore, the curvature away
TABLE I: Coordinates and Radial Distance from the Origin (R) of the Symmetry-InequivalentCso,CIm and C, Atoms" x,A y, A z, A R,A
c60 CIS0
+
(1) KrHtschmer, W.; Fostiropoulos, K.; Huffman, D. R.Chem. fhys. Lett. 1990, 170, 167. (2) Kritschmer, W.; Lamb, L. D.; Fostiropolous, K.; Huffman, D. R. Nature 1990, 347, 354. (3) Taylor, R.; Hare, J. P.; Abdul-Sada, A. K.; Kroto, H. W. J. Chem. Soc., Chem. Commun. 1990, 1423. (4) Johnson, R. D.; Meijer, G.; Bethune, D. S. J. Am. Chem. SOC.1990, 112,8983. ( 5 ) Ajie, H.; Alvarez. M. M.; Anz, S. J.; Beck, R. D.; Diederich, F.; Fostiropoulos, K.; Huffman, D. R.; Kritschmer, W.; Rubin, Y.;Schriver, K. E.;Sensharma, D.; Whetten, R. L. J. fhys. Chem. 1990, 94, 8630. (6) Haufler, R. E.;Conceicao, J.; Chibente, L. P. F.; Chai, Y.;Byrne, N. E.;Flanagan, S.; Haley, M. M.; OBrien, S. C.; Pan, C.; Xiao, Z.; Billups, W. E.;Ciufolini, M. A.; Hauge, R. H.; Margrave, J. L.; Wilson, L. J.; Curl, R. F.; Smalley, R. E. J. fhys. Chem. 1990, 94, 8634. (7) Hawkins, J. M.; Lewis, T.A.; Loren, S. D.; Meyer, A.; Heath, J. R.; Saykally, R. J. J. Org. Chem. 1990, 55, 6250. (8) Kroto, H. W. J. Chem. Soc., Faraday Trans. 1990, 86, 2465. (9) Fowler, P. W.; Cremona, J. E.;Steer, J. 1. Theor. Chim. Acra 1988, 73, 1.
(IO) Klein, D. J.; Seitz, W. A.; Schmalz, T. G. Nature 1986, 323, 703.
C2u)
3.462 0.728 3.430 4.627 6.768 6.921 6.804
0.694 5.953 5.021 4.287 0.712 1.38 1 2.779
0.000 O.OO0 O.Oo0 0.000 1.256 O.OO0 0.000
3.531 5.997 6.08 1 6.308 6.938 7.057 7.350
"The Cmcoordinates are from LDF optimization, and the other sets of coordinates are optimized by using EP from perfectly planar sp2 bonding with its three neighbors that each Cleoand C2@y b o n atom sees could be less than that of each C, carbon atom. Group theory and Huckel theory suggest that Cleoand C240could be as stable as Cm, yet so far Cleoand C240have been neither seen as extraordinarily abundant clusters nor isolated chemically. If Clsoand Cw are as stable as Cso, then perhaps new techniques can be found to synthesize them in quantity. If they can be synthesized they may be more stable than c60.
Preliminary experimental data suggest that Csois unstable to attack by molecular despite the fact that it has no defects or edges that might readily initiate oxidation. Perhaps larger fullerenes with less surface curvature would be more oxidatively stable and thus have greater applicability as precursors in materials science. A key to the discovery of current techniques for creating macroscopic amounts of C, was infrared (IR) ( I I ) Arbogast, J. W.; Darmanyan, A. P.; Foote, C. S.; Rubin, Y.;Dicderich, F. N.; Alverez, M. M.; Anz, S. J.; Whetten, R. L. J. fhys. Chem. 1991, 95, 11. (12) Haufler, R. E.;Chai, Y.;Chibante, L. P. F.; Conceicao, J.; Jin, C.; Wang, L.-S.; Maruyama, S.; Smalley, R. E.Mater. Res. Soc. Symp. froc. 1991, 206, 627. (13) Milliken, J.; Keller, T. M.; Baronavski, A. P.; McElvany, S. W.; Callahan, J. H.; Nelson, H. H. Chem. Mater. 1991, 3, 386.
This article not subject to U S . Copyright. Published 1991 by the American Chemical Society