Article pubs.acs.org/JPCA
Computational Study of Isopropylbenzenium Ions Stein Kolboe* inGAP Center for Research Based Innovation, Department of Chemistry, University of Oslo, P.O. Box 1033, Blindern, N-0315 Oslo, Norway S Supporting Information *
ABSTRACT: A theoretical investigation of the isopropylbenzenium ion system has been carried out with structures determined with B3LYP. Energies are calculated with the high accuracy composite methods, G3 (G3B3) and CBS (CBS-QB3). The main goal has been to resolve the following issue: Are there any stable ion π-electron complexes of the type C6H6/C3H7+ or C6H7+/C3H6 in this system? Two minimum points on the potential energy surface (PES) corresponding to benzenium ion/propene complexes were found. Due to free internal rotation, they represent only one species. The barrier for forming an isopropylbenzenium ion from the complex is low, so the lifetime will be short. Computation of the IR spectrum of the complex shows that there is a very intense absorption line due to C−H stretching, in an otherwise empty region, that may be used to identify the complex, if present. No stable C6H6/C3H7+ complex was found, but a quasi-stable species structurally corresponding to the earlier described stable C6H6/C4H9+ complex was observed. A simplistic explanation why the benzene/ isopropyl cation, in contrast to benzene/tert-butyl cation, does not form a stable ion/π-electron complex is given.
■
INTRODUCTION We have recently published theoretical studies of the ethylbenzenium and the tert-butylbenzenium ion systems.1,2 Central issues in both cases included the following: Can ion/πelectron complexes of the type C6H7+/CnH2n or C6H6/ CnH2n+1+ exist, and what is the mechanism of the reaction taking place when the alkyl group is split off? In the present paper, the same question regarding the isopropylbenzenium ion system is raised. A reason for carrying out these studies is the large number of experimental works that have appeared where the existence of such complexes is suggested to explain parts of the experimental observations, but their existence has not been shown.3−14 Comprehensive reviews of alkylarenium ion systems have been published by Kuck.15,16 In essence, the outcome of the two preceding theoretical studies is that the species C6H7+/C2H4 and C6H6/C4H9+ represent clear minima on the potential energy surfaces, so these complexes may exist, but complexes of the type C6H6/ C2H5+ and C6H7+/C4H8 were found to be unstable and therefore non-existing. Prior to our preceding studies of the ethyl- and tert-butylbenzenium ion systems, two papers utilizing quantum chemical methods to explore the possible existence of ion π-electron complexes in alkylbenzenium ion systems (isopropylbenzenium and tert-butylbenzenium ions) had appeared.10,11,17 Berthomieu et al. concluded that two ion/πelectron complexes are possible in the isopropylbenzenium ion system, C6H7+/C3H6 and C6H6/C3H7+. The structures of the complexes were obtained from optimizations at the HF/3-21G level with start structures apparently based on semiempirical computations with frozen fragment structures.10,11 Heidrich later carried out corresponding computations at a higher level of theory, MP2/6-31+G(d,p).17 According to his computations, no formation of an ion/π-electron complex can take place in © 2012 American Chemical Society
the isopropylbenzenium ion system, and an isopropylbenzenium ion is formed from the fragments. Our, more recent, computational results regarding the ethyl- and tert-butylbenzenium ion systems do, however, suggest that complex formation might be possible.1,2 A further study of the isopropylbenzenium ion system therefore seems worthwhile.
■
COMPUTATIONAL DETAILS The computations which are presented have all been performed using the Gaussian 03 program package.18 Geometries of the stationary states that were found and which are presented have been optimized at the B3LYP/6-311+ +G(d,p) level of theory. Energies of all species have been calculated at the B3LYP/6-311++G(d,p) level of theory with zero point energy (ZPE) corrections and at the much more accurate complete basis set level (CBS-QB3).19 With the exception of tert-butyl-2H-benzene, tert-butyl-3H-benzene, and the transition states for proton shifts on the benzenium ring, energies were also calculated at the Gaussian-3 (G3B3) level.20 To obtain convergence in structure optimizations, it is in many cases necessary to employ the Ultrafine integration grid. The geometric structures used in the CBS-QB3 and G3B3 computations and the corresponding ZPE corrections are based on B3LYP optimizations with the CBS basis set CBSB7 (=6311G(d,p)), respectively 6-31G(d) for G3 computations. The choice of basis set did not seem to influence the predicted structures much, and single point computations indicate rather uninteresting energy differences between the structures. In their standard formulations, both composite methods search for an Received: December 6, 2011 Revised: March 13, 2012 Published: March 13, 2012 3710
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
benzene were so low that the alkyl group was essentially a free rotor so that all rotamers would be present at all but very low temperatures, and no particular structure could be assigned to the various species.2 Considerably higher barriers for alkyl group rotation is found in the isopropylbenzenium ion system. Computations (B3LYP/6-31G(d,p)) have shown that the barriers for rotating the isopropyl group are higher than 6.3 kJ/mol (ip-2h-b). The, by far, highest rotation barrier is found for ip-1h-b, where it is 26 kJ/mol. The lowest barrier is low enough that a rotation will take place very often, but high enough that the ion most of the time retains the calculated structure. Figure 1a,b therefore represents the structures of these species quite well. It is common chemical wisdom that ip-4h-b should be the more stable isomer, followed by ip-2h-b (in line with the known para- and ortho-directing property of the isopropyl group). This is indeed in agreement with the calculations. The energies of the isomers as obtained by B3LYP/6-311++G(d,p), Gaussian-3 (G3B3), and the complete basis set (CBS-QB3) methods are given in Table 1. Assuming no important entropy
energy minimum during the structural optimization. When the intention is to determine a transition state energy, it is therefore necessary to carry out a transition state optimization (preferably using a tight convergence criterion) with the appropriate basis set and use the structure thus determined as input for the CBS and G3 computations so that the optimization stops at the transition state geometry because the input structure is a stationary state. Stationary states corresponding to ion π-complexes were found. From earlier experience with the ethylbenzenium and tert-butylbenzenium ion systems, it was expected that Hartree− Fock and MP2-based computations might disagree. Parallel HF and MP2 computations with several basis sets have therefore been carried out. Computations on these stationary states have also been extended to other functionals.
■
RESULTS AND DISCUSSION Isopropylbenzenium Ion Isomers. Isopropylbenzene can be protonated at four different C atoms. The four isomers thus formed are isopropyl-1H-benzene, isopropyl-2H-benzene, isopropyl-3H-benzene, isopropyl-4H-benzene. In the following, these isomers will be denoted by the acronyms ip-1h-b, ip-2h-b, ip-3h-b, and ip-4h-b. Geometry optimizations of these species have been carried out. The optimized structure of isopropylbenzene is shown in Figure 1a. This molecule has Cs symmetry with the isopropyl
Table 1. Energies of Isopropylbenzenium Ions and Transition State Energies for Proton Shifts between the Benzene Ring Carbons B3LYP/6-311++G(d,p) species
E/Eh
ΔE/kJ mol−1
ip-1h-b −350.395614 35.9 ip-2h-b −350.407216 5.4 ip-3h-b −350.401345 20.8 ip-4h-b −350.409278 0.0 transition states for hydrogen ring walka TS-12 −350.385502 62.4 TS-23 −350.384172 65.9 TS-34 −350.384942 63.9 barriers for hydrogen shifts C1 to C2 C2 to C1 C2 to C3 C3 to C2 26.5b 57.0 60.5 45.1 14.5c 40.2 45.1 30.9
CBS-QB3 E(CBS)/Eh
ΔE/kJ mol−1
−349.757874 −349.767664 −349.762259 −349.769897
31.6 5.9 20.1 0.0
−349.752336 −349.750460 −349.750706
46.1 51.0 50.4
C3 to C4 43.1 30.3
C4 to C3 63.9 50.4
a
TS-MN: The transition state for proton shift between carbon M and N. bCalculated with B3LYP, kJ mol−1. cCalculated with CBS-QB3, kJ mol−1.
Figure 1. (a) Geometry optimized structure of isopropylbenzene. The figure also represents the structures of isopropylbenzenium ions when the protonation has taken place at the ortho-, meta-, or para-position when CH is replaced by CH2 at the appropriate positions. All of these species have Cs symmetry. The isopropyl C−H is in the benzene ring plane. There is no symmetry plane normal to the benzene ring so C atoms C2 and C6, and C3 and C5 are not equivalent. The energy differences resulting from protonation at C2 or C6, and C3 or C5 are, however, negligible (≤1.1 kJ/mol). (b) Geometry optimized structure of ipso-protonated isopropylbenzene (ip-1h-b). There is no symmetry element, but the two bonds Cring−Cisopropyl and Cisopropyl−Cmethyl (the one pointing upward) are nearly in a plane normal to the benzene ring plane.
differences,21 the data in the table show that at room temperature protonated isopropylbenzene will consist of approximately 90% ip-4h-b and 10% ip-2h-b. The content of the other tautomers will be very small. It is also generally known that the proton can, fairly easily, move around the benzene ring. The energies of the transition states for shifting between adjacent atoms are also given in Table 1 as are also the calculated barriers for jumping from one benzene ring carbon atom to the next. The results given in Table 1 are best summarized by Figure 2. Attention is drawn to the discrepancy between the B3LYP and the CBS-based computational results for the transition state energies. The latter are known to be the more trustworthy. The highest barrier (CBS-QB3 computations) is seen from Table 1 to be 50.4 kJ/mol. This value is still low enough that proton walk in a thermal system is expected to take place rapidly. On the basis of the Arrhenius equation k = A × exp(−E/RT) with A ≈ 1013 s−1 and room temperature, one gets k ≈ 104 s−1.
C−H bond in the benzene ring plane. This figure may also be used to represent ip-2h-b, ip-3h-b, and ip-4h-b when a benzene C−H group is replaced by a CH2 group at the appropriate place. The structure of bu-1h-b is different and is shown in Figure 1b. It has no symmetry plane or axis (i.e., symmetry C1). The C−C bond in the isopropyl group is nearly, but not fully, in a plane normal to the benzene ring plane. In our preceding study of the tert-butylbenzenium ion system, it was found that the rotation barriers in the precursor hydrocarbon and the protonated species except tert-butyl-1H3711
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
Figure 3. Structures of the three ion π-electron complexes (A2, B1, B2) found with B3LYP structure optimizations.
the preceding study of the tert-butylbenzenium ion system, it was found that, although four π-electron complexes, called A1, A2, B1, and B2, were predicted by B3LYP (minima on the B3LYP potential energy surface), further computations at higher levels of theory, where also zero point energy corrections are included, showed that three of them (A2, B1, and B2) would transform with no barrier into the fourth, A1, which therefore was predicted to be the only observable π-complex in the tert-butylbenzenium ion system.2 Is there a C6H6/C3H7+ complex similar to the C6H6/C4H9+ complex found previously, and then called A1, with the structure shown in Figure 4a?2 (That complex is in the following called A1-bu to distinguish it from the corresponding isopropyl cation complex to be called A1.) Replacing one of the two methyl groups which are not in the symmetry plane of the complex A1-bu in Figure 4a gives a corresponding isopropyl cation complex, A1. A considerable number of computations have been carried out seeking a stable complex of this type. The starting point was taken to be the structure resulting from the replacement of one of the methyl groups of A1-bu by hydrogen, but retaining all of the other atom positions, or slight variations of this structure. Optimizations of the structures were carried out with several basis sets, various functionals, and MP2. When no constraints were imposed, the structure always converged to the isopropylbenzenium ion ip-1h-b in Figure 1b. It is therefore concluded that there is no stable benzene/isopropyl cation complex. However, it should be mentioned that, when a geometry optimization is run with the starting structure obtained by replacing a methyl group in A1-bu with hydrogen (i.e., A1 start structure), an initial rapid, but quite small, fall in energy is soon succeeded by a series of steps where the energy hardly changes. During these steps, the structure is visually little changed from the start structure. The forces acting in the system are small, small enough that the force criterion for convergence is often met. Eventually, a rapid fall in energy starts and the structure turns into that of the isopropylbenzenium ion, ip-1h-b. This observation is valid for the tested functionals, B3LYP, PBE1PBE, MPW1PW91, as well as MP2 computations. When frequency computations are carried out with the B3LYP functional and the basis sets 6-31G(d), 6-311G(d,p) (=CBSB7), and 6-311++G(d,p) (where there is convergence) in the quasi-converged region, they all show a transition state with a low imaginary frequency (−24 cm−1) corresponding to a rotational vibration which involves the isopropyl cation as a whole and a rotation of the methyl group that interacts with the benzene ring (see Figure 4). Restarting the optimization after having nudged the transition state normal coordinate slightly out of position, in either direction, led to formation of ip-1hb.23 Even though the complex A1 does not exist as a true
Figure 2. Isopropylbenzenium ion energies and transition state energies for proton walk on the ring.
Ion π-Electron Complexes. Earlier computations have shown that an ethylbenzenium ion may be transformed into a stable C6H7+/C2H4 ion π-electron complex (a minimum on the potential energy surface), but no C6H6/C2H5+ complex can be formed.1 The tert-butylbenzenium ion, on the other hand, may be transformed into a stable C6H6/C4H9+ complex.2 Geometry optimizations with the B3LYP DFT functional showed very shallow potential energy surface minima corresponding to complexes C6H7+/C4H8, but when the ZPE was taken into account, they proved unstable relative to the tert-butyl cation complex. The very different behaviors of the ethylbenzenium and the tert-butylbenzenium ion systems are clearly connected with the proton affinity difference between benzene and ethene on the one hand and the difference between benzene and isobutene on the other hand. The proton affinity of isobutene is 802.1 kJ/ mol, that is, more than 50 kJ/mol above benzene, and the proton affinity of ethene is 680.5 kJ/mol, which is 70 kJ/mol below that of benzene. In the case of the isopropylbenzenium ion system, the proton affinities of benzene and propene are 750.4 and 751.6 kJ/mol; that is, they are almost identical.22 On this basis, and the somewhat conflicting literature reports, it is of interest to investigate the isopropylbenzenium ion system. Since the experimental proton affinities of benzene and propene are almost equal, one might expect about equal probabilities for formation of benzenium ion/alkene complexes and benzene/isopropyl cation complexes. Furthermore, because of the similarity of proton affinities, the reaction energies for forming benzene + isopropyl cation or benzenium ion + propene from the isopropylbenzenium ion are known to be not much different. Geometry optimizations in the isopropylbenzenium ion system at the B3LYP/6-311++G(d,p) level found three minima on the PES corresponding to ion π-electron complexes. Optimizations with the basis sets 6-31G(d) and CBSB7 (∼6311G(d,p)), which are used as the basis for G3 and CBS computations, gave the same structures. The complexes are termed A2, B1, and B2, where A2 is a C6H6/C3H7+ complex and B1 and B2 are C6H7+/C3H6 complexes. These complexes are directly corresponding to the previously equally named complexes that were found in the tert-butylbenzenium ion system when a methyl group in the tert-butyl cation is replaced by hydrogen.2 The structures of A2, B1, and B2 are shown in Figure 3. The energies of the complexes are given in Table 2. In 3712
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
Table 2. Energies of the Species Treated Here at Three Different Levels of Theory energies relative to ip-4h-b/kJ mol−1
absolute energies/Eh species
B3LYP
CBS-QB3
G3B3
B3LYP
CBS-QB3
G3B3
ip-4h-b ip-1h-b quasi-A1 A2 B1 B2 TS-A2 TS-B1 TS-B2 C6H7+ + C3H6 C6H6 + C3H7+
−350.40928 −350.39561 −350.38765 −350.38277 −350.37696 −350.37684 −350.38284 −350.37569 −350.37560 −350.36760 −350.36776
−349.76990 −349.75787 −349.74441 −349.73456 −349.72627 −349.72659 −349.73515 −349.72568 −349.72680 −349.71551 −349.71643
−350.17908 −350.16681 −350.15165 −350.14184 −350.13532 −350.13583 −350.14274 −350.13307 −350.13445 −350.12471 −350.12356
0.0 35.9 56.8 69.6 84.9 85.2 69.4 88.2 88.4 109.4 109.0
0.0 31.6 65.7 92.8 114.5 113.7 91.2 116.1 113.1 142.8 140.4
0.0 32.2 72.0 97.8 114.9 113.7 95.4 120.8 117.2 142.7 145.8
further apart. MP2 computations with the small basis set 631G(d) indicate stable structures, but the relative orientations of the benzenium and the propene moieties may be very different. When the basis set is increased to 6-311++G(d,p), the B1 and B2 structures are definitely not reproduced at the MP2 level. The proton is quickly transferred from the benzenium ion to the propene molecule, and the structures are gradually transformed into the isopropylbenzenium ion, ip-1h-b. The barrierless proton transfer to propene when the computation is based on MP2 is somewhat surprising, given that the proton affinities of benzene and propene are almost equal. The transition state structures, TS-A2, TS-B1, and TS-B2 are shown in Figure 5. TS-B1 and TS-B2 in Figure 5b,c show the
Figure 4. (a) Structure of the previously found C6H6/C4H9+ ion πelectron complex (A1-bu). (b) Structure of a quasi-stable C6H6/C3H7+ (quasi-A1) ion π-electron complex.
minimum on the PES, the structure(s) that occurs in the C6H6/ C3H7+ species during the “almost converged” part of the optimization runs may be taken to represent a hypothetical complex, which is here called quasi-A1. The energy of quasi-A1 can be calculated also at the G3B3 and CBS-QB3 levels. The values are given in Table 2. The main interest of the quasi-A1 energy is that it allows an evaluation of the interaction energy in the quasi-A1 complex relative to the separated benzene molecule and the isopropyl cation, whose energies are also given in Table 2. The structure is depicted in Figure 4b. As formation of ip-1h-b eventually starts during an optimization run, the dihedral angle determined by the three carbon atoms and the hydrogen atom that replaced a methyl group (the secondary carbon atom) starts changing from being nearly 180° as it is in A1-bu to nearly 110° as is expected in an isopropyl group, and the central carbon atom in the isopropyl cation approaches the benzene ring carbon atom to which it will finally become attached. The result may be summed up in the following simplistic way: The isopropyl cation is not stiff enough to prevent the formation of an isopropyl group which forms a bond to the benzene moiety. This is the reason for the non-existence of an isopropyl cation benzene complex, A1. The tert-butyl cation, on the other hand, is stiff enough to make A1-bu a stable cation complex. Are the complexes A2, B1, and B2 stable complexes also within the Hartree−Fock and MP2 computational schemes? Computations on A2 within the Hartree−Fock and MP2 schemes concur with the extended B3LYP computations. This complex is not a stable structure. B1 and B2, on the other hand, receive support from HF/6-31G(d) and HF/6-311++G(d,p) computations. By visual inspection, there is little difference from the B3LYP structures, though the two moieties are a bit
Figure 5. Transition state structures of the complexes A2, B1, and B2. Optimizations based on slightly nudged structures along the reaction coordinate result in isopropylbenzenium ion (ip-1h-b), as depicted in Figure 1b. On their way there, they all pass a region where the structure is well represented by Figure 4b, quasi-A1.
transfer of a proton between the benzenium ion and the propene molecule. The non-negligible transition state barriers for transferring a proton from the benzenium ion to propene is confirmed by the G3 and CBS computations, as is seen from Table 2. The barriers are, however, seen to be so low that the lifetimes of B1 and B2 may be expected to be short at all but extremely low temperatures. (The barrier TS-B1 (B3LYP) without ZPE correction is 10.1 kJ/mol, and the G3 barrier without ZPE correction is 14.4 kJ/mol.) It was observed that when geometry optimizations were carried out on TS-B1 and TS-B2, after having moved the transitional proton very slightly toward the propene moiety, the systems passed through a series of structures very similar to A1 before finally being converted to ip-1h-b. Because of the short lifetimes of B1 and B2, their concentration is expected to be very low. However, the IR spectrum which is obtained from the frequency calculations is calculated to have an intense line due to C−H stretching in an otherwise empty frequency range (≈2400 cm−1). The computed spectrum is shown in Figure 6. The frequencies 3713
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
or C9H13+ → C6H7+ + C3H6
are available. They can be calculated from the experimentally known reaction enthalpy for C9H12 → C6H6 + C3H6 and the proton affinities for the same species. Utilizing the data given in the NIST Web book, one obtains 139.4 and 140.6 kJ/mol for the two reactions in the same order.22 The data given in Table 2, which gives energies at 0 K, do not directly give the reaction enthalpies at 298 K; they allow calculation of reaction energies at 0 K. However, the Gaussian program package also computes enthalpies at 298 K (or any other chosen temperature), so although the enthalpies have not been tabulated here, they are available. When the enthalpies at 298 K are utilized, the theoretically computed reaction enthalpies are obtained. Table 3 gives the theoretically Figure 6. Computed IR spectrum of B1 and B2. The strong line slightly below 2500 cm−1 will, if observed, prove the existence of the complex.
Table 3. Enthalpy Changes (ΔH298K ° /kJ mol−1) for Dealkylation of Isopropylbenzenium Ions According to Different Computational Schemes
are not scaled and are likely to be 3−4% too high.24 The very intense line at about 2400 cm−1, where there are no other lines, gives hope that the presence of B1/B2 can be observed with IR spectroscopy even if the concentration might be low. Further computations show that the propene moiety is essentially freely rotating relative to the benzenium ion. B1 and B2 are, therefore, strictly speaking, not representing two different species. A summing up of the results are given in Figure 7. A pronounced difference between the energies of B1/B2 and
reaction
B3LYP
CBS-QB3
G3B3
experimental
C9H13+ → C3H6 + C6H7+ C9H13+ → C3H7+ + C6H6
111.1 111.2
146.9 144.9
144.5 148.0
140.6 139.4
computed reaction enthalpies as given by the methodologies DFT (B3LYP), CBS (CBS-QB3), and G3 (G3B3) together with the experimentally based data. It is seen from the table that the reaction enthalpies from G3 and CBS agree satisfactorily with experiment, but B3LYP fails. The reaction energies at 0 K given in Table 2 and the reaction enthalpies given in Table 3 are not much different. It is known since the early paper by Olah et al. that the protonation of an alkylbenzene leads overwhelmingly to paraprotonation, in agreement with the computations (Table 2).25 Farcasiu et al. have shown that in the case of ethylbenzene there is also some ortho-protonation.26 They were able to show that the enthalpy of the ortho-protonated species is about 4 kJ/mol higher than the para-protonated species. This enthalpy difference is very near the value found here for isopropylbenzene (Table 2). Experimental data on the energy difference between the ortho- and para-protonated isopropylbenzene do not seem to exist. Chiavarino et al. studied the D/H exchange in fully deuterated isopropylbenzenium ions interacting with Hcontaining compounds with suitable proton affinities.27 They concluded that in isopropylbenzene the kinetically preferred sites for protonation are the ortho- and para-positions. They found a slight excess of ortho-protonation, but taking into consideration that there are two ortho-positions and only one para-position, there is a slight preference for the para-site. There is, therefore, also here, agreement between experiment and theory, but it must be taken into account that the computational result only refers to energy, not kinetics. Aschi et al. showed that although the alkylation of arenes with alkenes usually is considered to proceed via a preliminary protonation of the alkene, followed by addition of the carbenium ion thus formed to the arene, isopropylbenzene may be formed by a primary addition of propene to a benzenium ion.28 This experimental result is in agreement with the calculated small barrier (Table 2) for converting the complexes B1 and B2 into an isopropylbenzenium ion (ip-1h-
Figure 7. Energies of the investigated ions and complexes.
quasi-A1, that is, the complexes C6H6/C3H7+ and C6/H7+/ C3H6, is clearly seen. Because the proton affinities of benzene and propene are almost equal, the two systems C6H6/C3H7+ and C6/H7+/C3H6 have almost equal energies when the C6 and C3 parts are widely separated. Therefore, the energy difference between the two complexes may be ascribed to different interaction energies between the fragments C6H6 and C3H7+ in one complex and C6/H7+ and C3H6 in the other complex. Experimental Data versus the Computational Results. There are not many published experimental data that can serve to test the validity of the theoretical results. However, the reaction enthalpies for the reactions C9H13+ → C6H6 + C3H7+ 3714
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
Notes
b), which may, via proton walk on the benzene ring, be converted to the more stable species ip-4h-b. Mass spectrometric measurements in the isopropylbenzenium system have shown that when the ions split up there is formation of C6H6/C3H7+ and C6H7+/C3H6.4,11 Matthias and Kuck have shown that also more complex isopropylarenium ion systems behave similarly when these ions undergo fragmentation.29 These results are to be expected because of the essentially equal proton affinities of propene and benzene and is also in accordance with the calculated energies found here (Table 2).
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS Helpful discussions with Associate Professor Stian Svelle are acknowledged.
■
CONCLUSIONS A theoretical study of the isopropylbenzenium ion system has been carried out. The dependence of the isopropylbenzenium ion energies on the protonation site and the barriers for proton walk along the benzene ring have been calculated. The main objective is to investigate if ion π-electron complexes of the type C6H6/C3H7+ and/or C6H7+/C3H6 can exist. The computations have shown that there are two minimum points on the potential energy surface of the type C6H7+/C3H6, termed B1 and B2. They have both been fully characterized but may probably be seen as two rotamers of an essentially free rotor. The barrier for transforming the complexes into an isopropylbenzenium ion is about 5 kJ/mol (by the composite G3 method), low enough that a non-negligible lifetime may be expected at low temperatures in a thermal system. Their existence might be verified by IR spctroscopy. In an earlier work on the tert-butylbenzenium ion system, it was found that the complex C6H6/C4H9+ is rather stable. All attempts to find a corresponding stable C6H6/C3H7+ complex failed. The system always finally converged to an ipsoprotonated isopropylbenzenium ion. The difference between the two alkylbenzenium ion systems has been given the following simplistic explanation: The flat tert-butyl cation is stiff enough that there is a substantial energy needed to bend it, enough that the complex can persist. For transforming it into a tert-butylbenzenium ion, a barrier must be crossed. The isopropyl cation, on the other hand, is just flexible enough that it can be bent and without a barrier form an isopropylbenzenium cation. In spite of the lack of a converged C6H6/C3H7+ structure, it is shown that the energy and structure of the somewhat hypothetic C6H6/C3H7+ complex can be computed. The calculated reaction enthalpies for splitting off the alkyl group are in satisfactory agreement with experimentally based values when the computations are based on Gaussian-3 and complete basis set methods, lending support for the computed energies. Reaction enthalpy computation based on density functional theory with the B3LYP functional is markedly inaccurate.
■
ASSOCIATED CONTENT
S Supporting Information *
Structural details of important species are given as XYZ coordinates of atomic positions. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
(1) Kolboe, S.; Svelle, S.; Arstad, B. J. Phys. Chem. A 2009, 113, 917− 923. (2) Kolboe, S. J. Phys. Chem. A 2011, 115, 3106−3115. (3) Leung, H. W.; Harrison, A. G. Org. Mass Spectrom. 1977, 1977, 582−586. (4) Herman, J. A.; Harrison, A. G. Org. Mass Spectrom. 1981, 16, 423−427. (5) Sen Sharma, D. K.; Ikuta, S.; Kebarle, P. Can. J. Chem. 1982, 60, 2325−2331. (6) Cacace, F.; Ciranni, G. J. Am. Chem. Soc. 1986, 108, 887−890. (7) Cacace, F. Acc. Chem. Res. 1988, 21, 216−222. (8) Audier, H. E.; Monteiro, C.; Robin, D. New J. Chem. 1989, 13, 621−624. (9) Berthomieu, D.; Audier, H. E.; Denhez, J. P.; Monteiro, C.; Mourgues, P. Org. Mass Spectrom. 1991, 26, 271−275. (10) Berthomieu, D.; Brenner, V.; Ohanessian, G.; Denhez, J. P.; Millié, P.; Audier, H. E. J. Am. Chem. Soc. 1993, 115, 2505−2507. (11) Berthomieu, D.; Brenner, V.; Ohanessian, G.; Denhez, J. P.; Millié, P.; Audier, H. E. J. Phys. Chem. 1995, 99, 712−720. (12) Harrison, A. G.; Wang, J. Y. Int. J. Mass Spectrom. 1997, 160, 157−166. (13) Holman, R. W.; Gross, M. L. J. Am. Chem. Soc. 1998, 111, 3560−3565. (14) Miklis, P. C.; Ditchfield, R.; Spencer, T. J. Am. Chem. Soc. 1998, 120, 10482−10489. (15) Kuck, D. Mass Spectrom. Rev. 1990, 9, 583−630. (16) Kuck, D. Int. J. Mass Spectrom. 2002, 213, 101−144. (17) Heidrich, D. Angew. Chem., Int. Ed. 2002, 41, 3208−3210. (18) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; et al. Gaussian 03, revision B.04; Gaussian, Inc.: Pittsburgh, PA, 2004. (19) Montgomery, J. A. Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. J. Chem. Phys. 1999, 110, 2822−2827. (20) Baboul, A. G.; Curtiss, L. A.; Redfern, P. C.; Raghavachari, P. C. J. Chem. Phys. 1999, 110, 7650−7657. (21) Entropies are calculated by the Gaussian program package, and the calculated entropies are quite similar, but important contributions to the calculated entropies are due to torsional vibrations that should be treated as partially free rotations. The calculated entropies are therefore not reliable. On the other hand, essentially equal torsional vibrations with very similar frequencies are observed for all isomers. Assuming similar entropies is therefore reasonable. (22) NIST Web book, http://webbook.nist.gov/chemistry/, accessed November 18, 2010. (23) The computations reported by Berthomieu et al.,10,11 who worked at the HF/3-21G level when searching for complexes, were repeated and also extended to HF/6-31G(d) and HF/6-311++G(d,p). At the Hartree−Fock level, a PES minimum corresponding to a complex C6H6/C3H7+ was found. The PES barrier for transforming the complex into ip-1h-b was, however, only 1.3 kJ/mol (HF/6-31G(d)). The ZPE corrected barrier was negative. (24) Andersson, M. P.; Uvdal, P. J. Phys. Chem. A 2005, 109, 2937− 2941. (25) Olah, G. A.; Schlosberg, R. H; Porter, R. D.; Mo, Y. K.; Kelly, D. P.; Mateescu, G. D. J. Am. Chem. Soc. 1972, 94, 2034−2043. (26) Farcasiu, D; Melchior, M. T.; Craine, L. Angew. Chem., Int. Ed. Engl. 1977, 16, 315−316. (27) Chiavarino, B; Crestoni, M. E.; Di Rienzo, B.; Fornarini, S. J. Am. Chem. Soc. 1998, 110, 10856−10862.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. 3715
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
The Journal of Physical Chemistry A
Article
(28) Aschi, M.; Attina, M.; Cacace, F. Angew. Chem., Int. Ed. Engl. 1995, 34, 1589−1591. (29) Matthias, C; Kuck, D. Croat. Chem. Acta 2009, 82, 7−19.
3716
dx.doi.org/10.1021/jp211731j | J. Phys. Chem. A 2012, 116, 3710−3716
