Computational and matrix isolation studies of tetra-tert-butylethane

Article abstract of DOI:10.1002/ejoc.200900378

Irradiation of cryogenic xenon matrices containing the sterically congested alkane tetra-tert-butylethane (1) with intense pulses from a KrF excimer laser (λ_exc = 248 nm) have resulted in the formation of methane and isobutene as the major products. Larger fragments derived from 1, such as 2,2-dimethylbutane (10), were also detected, whereas no evidence was obtained for the formation of the elusive tetra-tert-butylethylene (2). The results suggest that biphotonic irradiation of xenon matrices containing tert-butylated precursor molecules may generally result in the efficient formation of isobutene and smaller alkanes. An analysis of the decomposition reactions of 1 by various computational methods revealed that commonly used DFT methods, such as B3LYP or BLYP, and MP2 theory do not provide a good description of the energetics of the decay reactions of the sterically overloaded alkane 1. The recently developed M05-2X method was the only method among those tested to yield an activation enthalpy for the cleavage of 1 in agreement with the experimental value.

Full text of DOI:10.1002/ejoc.200900378

FULL PAPER
DOI: 10.1002/ejoc.200900378
Computational and Matrix Isolation Studies of Tetra-tert-butylethane
Götz Bucher*[a][‡]
Dedicated to Professor Dr. C. Rüchardt on the occasion of his 80th birthday
Keywords: Matrix isolation / Xenon / Alkanes / Thermochemistry / Strained molecules / Computer chemistry
Irradiation of cryogenic xenon matrices containing the steri-
cally congested alkane tetra-tert-butylethane (1) with intense
pulses from a KrF excimer laser (λexc = 248 nm) have resulted
in the formation of methane and isobutene as the major prod-
ucts. Larger fragments derived from 1, such as 2,2-dimeth-
ylbutane (10), were also detected, whereas no evidence was
obtained for the formation of the elusive tetra-tert-butyleth-
ylene (2). The results suggest that biphotonic irradiation of
xenon matrices containing tert-butylated precursor mole-
cules may generally result in the efficient formation of isobut-
ene and smaller alkanes. An analysis of the decomposition
reactions of 1 by various computational methods revealed
that commonly used DFT methods, such as B3LYP or BLYP,
and MP2 theory do not provide a good description of the
energetics of the decay reactions of the sterically overloaded
alkane 1. The recently developed M05-2X method was the
only method among those tested to yield an activation en-
thalpy for the cleavage of 1 in agreement with the experi-
mental value.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
butylethane (1),[8] a known molecule that already contains
two adjacent di-tert-butylmethyl groups, and eliminate the
two tertiary hydrogen atoms.
Introduction
Tetra-tert-butylethylene (2), one of the most sought-after
molecules in organic chemistry, so far has remained resilient
to all attempts at its synthesis.[1] The naming of a book
chapter on the synthesis of 2 as “... an exercise in prepara-
tive futility”[2] may serve to illustrate the difficulties encoun-
tered in the synthesis of 2, which is otherwise predicted to
be a reasonably stable molecule.[3] Unsuccessful approaches
to the preparation of 2 (Scheme 1) include the coupling of
precursors containing a di-tert-butylmethyl moiety, such as
di-tert-butyl selenoketone (3) and di-tert-butyldiazo-
methane (4),[4a] the dimerization of matrix-isolated di-tert-
butylcarbene[4b] or of radical ions of di-tert-butylcarbene[4c]
or the McMurry coupling of di-tert-butyl ketone (5).[5] In
other attempts, a “bent-back” approach was used.[6] Tet-
rakis(1-methylcyclopropyl)ethene (6) can be synthesized,
but the last step in the synthesis of 2, the hydrogenation
of the cyclopropane rings, could not be achieved.[1,2,7] An
alternative, as yet untried approach is to take tetra-tert-
Scheme 1. Possible approaches to the synthesis of 2.
This contribution describes the attempts at the dehydro-
genation of 1 by pyrolysis in cryogenic xenon matrices. The
seminal work of Maier et al. showed that irradiation of bro-
mine-doped Xe matrices with light from an Hg low-pres-
sure lamp (λexc = 254 nm) results in the formation of XeBr
exciplexes that are capable of transferring their excitation
energy to a matrix-isolated guest molecule.[9] The guest
molecule thus may be locally heated at temperatures as low
as 10 K, which results in a range of fragmentation reactions.
A number of reactive intermediates, among them trimethyl-
enemethane[10] or the allyl radical,[11] have been synthesized
[a] Lehrstuhl für Organische Chemie II, Ruhr-Universität
Bochum,
Universitätsstr. 150, 44801 Bochum, Germany
[‡] Current address: WestCHEM, Department of Chemistry, Uni-
versity of Glasgow, Joseph-Black-Building,
University Avenue, Glasgow G12 8QQ, United Kingdom
E-mail: goebu@chem.gla.ac.uk
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.200900378.
4340
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 4340–4345

Computational and Matrix Isolation Studies of Tetra-tert-butylethane
this way. That the method can be used to dehydrogenate tic weak double peak at ν = 990 and 980 cm–1, which only
˜
saturated species is evidenced by the fact that the Xe-matrix shows up in this form in the calculated spectrum of 10.
cryogenic pyrolysis of propane yields, among others, allene It is clear, however, that the product mixture also contains
as a product.[11] The use of bromine as a dopant can be components other than methane, isobutene and 10, as dem-
avoided if high-energy pulses from a KrF excimer laser (λexc onstrated, for example, by the broad and poorly resolved
= 248 nm) are employed. Two-photon absorption of neat absorption at ν ≈ 1450 cm–1. The formation of some frag-
˜
Xe results in the formation of Xe excimers, which release ments can be ruled out conclusively. Ethane, as a possible
their energy to any matrix-isolated guest molecule in the final product of deisobutenylation, was not formed.[15] No
form of thermal excitation.[9] A necessary condition for this trace of the IR absorption attributable to tetra-tert-butyl-
procedure is the absolute absence of any absorption of the ethylene (2) could be detected. According to the calculation,
matrix-isolated guest molecule at the excitation wavelength. 2 should show one prominent IR band in a region (calcd. ϫ
This is fulfilled for 1. For comparison, analogous irradia- 0.9614: ν = 1178 cm–1) in which none of the other possible
˜
tion experiments with 1 were also performed by using argon products should have a major IR band. The only candidate
matrices.
is a very weak IR band found at ν = 1179 cm–1. This band,
˜
however, belongs to traces of unreacted precursor 1.
Results
Energy Calculations
Matrix-Isolation Spectroscopy
Matrix isolation of 1 in Ar could be achieved at a sample
Quantum mechanical calculations were performed to
temperature of 20 °C. The matrix containing 1 was sub- gain an insight into the thermodynamic feasibility of the
sequently irradiated by using both 248 nm irradiation from fragmentation reactions that result from the thermal exci-
a KrF excimer laser and 193 nm irradiation from an ArF tation of 1 and the subsequently formed products. The
excimer laser. No photochemical conversion was observed: B3LYP hybrid functional[16] usually yields useful geometries
1 proved to be absolutely photostable. If xenon was used as and infrared frequencies, whereas the energies obtained may
the matrix material, irradiation with high-energy pulses be of questionable value, in particular if larger molecules
(λexc = 248 nm, ca. 200 mJ/pulse) from a KrF excimer laser are involved.[17] For this reason, full geometry optimizations
resulted in the slow disappearance of IR bands of 1 and the were also performed at the M05-2X/6-31G(d) and RIMP2/
formation of a new infrared spectrum. It took a large cc-pVTZ levels of theory. No frequency calculations were
number of laser pulses (65000) to achieve a nearly complete attempted by using the RIMP2 method. Scheme 2 displays
conversion of 1 into the new products.[12] Figure S1 of the the results of the DFT and RIMP2 calculations.
Supporting Information shows an infrared spectrum of 1
As expected, the calculations indicate that the central C–
recorded in Xe, along with a calculated spectrum [B3LYP/ C bond connecting the two di-tert-butylmethyl fragments
6-31G(d)] for comparison. Figure S2 shows the infrared should be the most prone to cleavage. Although this is con-
spectrum after the irradiation experiment.
sistent with the findings of Rüchardt and co-workers that
heating of 1 in mesitylene at T = 141 °C leads to the forma-
Based on a comparison with literature data,[13,14]
a
number of the newly formed IR bands can be assigned to tion of di-tert-butylmethane in good yield,[8a] the bond-
methane (ν = 1300 cm–1) and isobutene (ν = 1656, 1478, dissociation enthalpies, as calculated by the three methods,
˜
˜
891 cm–1). A comparison of the experimental and calcu- are in such disagreement with each other that a thorough
lated band integrals allows the ratio of methane/isobutene investigation into the choice of method seemed warranted.
formed to be estimated as 1:3. As logical primary byprod-
To resolve the pronounced disagreement between the cal-
ucts, tri-tert-butyl(isopropylidene)ethane 8 could be formed culations, the homolytic cleavage of 1 into 2 equiv. of 14
along with methane and tri-tert-butylethane 7 could be was systematically investigated by using a variety of meth-
formed along with isobutene. The experimental spectrum in ods, including DFT methods, RIMP2 and Hartree–Fock
the range ν = 750–1500 cm–1 is compared in Figure S3 with theory with different basis sets. Table 1 shows the results.
˜
the calculated spectra of the possible products of demethyl-
ation and deisobutenylation.
According to Table 1, commonly used DFT methods
such as B3LYP or BLYP vastly underestimate the activation
Figure S3 indicates that an unambiguous assignment of enthalpy for the cleavage of 1 into 2 equiv. of 14. Increasing
product bands to individual product molecules is difficult. the size of the basis set is counterproductive with B3LYP
The low intensity of the product bands in the range ν = or BLYP, as it results in a further decrease in the calculated
˜
1150–1250 cm–1 shows that highly tert-butylated products BDE and activation enthalpy. The MP2 theory (both full
such as 7, 8, 9 or 2 are not present in high concentrations MP2 and RIMP2), on the other hand, dramatically overes-
as they are predicted to have prominent bands in this re- timates the stability of the central C–C bond in 1. Although
gion. The agreement between the experimental product no attempt was made to calculate the activation enthalpy
spectrum and the spectrum of 2,2-dimethylbutane (mono- for the cleavage of 1 by MP2 methods, the calculated bond
tert-butylethane; 10) appears to be the best. This is evi- dissociation enthalpies alone are significantly larger than
denced by the experimental band pattern (4:5:10 intensity the experimental value of ∆H‡. Again, no improvement was
ratio) between ν = 1410 and 1360 cm–1 and the characteris- observed upon increasing the size of the basis set used. The
˜
Eur. J. Org. Chem. 2009, 4340–4345
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4341

G. Bucher
FULL PAPER
Scheme 2. Fragmentation pathways upon thermal excitation of 1 in Xe matrices. In italics: calculated enthalpies (RIMP2/cc-pVTZ, in
kcalmol–1) relative to 1 (0.0 kcalmol–1). In parentheses: calculated entropies [B3LYP/6-31G(d), in calmol–1 K–1]. In ordinary font: calcu-
lated enthalpies [B3LYP/6-31G(d) + ZPE, in kcalmol–1] relative to 1 (0.0 kcalmol–1). Underlined: calcd. enthalpies [M05-2X/6-31G(d) +
ZPE, in kcalmol–1] relative to 1 (0.0 kcalmol–1).
Table 1. Calculated activation enthalpies for the cleavage of 1 into 2 equiv. of 14, calculated BDEs of the central C–C bond in 1 and
calculated bond lengths of the central C–C bond in 1.
Method
Basis set
BDE [kcalmol–1]
∆H‡ [kcalmol–1]
RC–C [Å]
HF + ZPE
6-31G(d)
6-311++G(d,p)
6-31G(d)
6-311++G(d,p)
6-31G(d)
6-311++G(d,p)
6-31G(d)
6-311++G(d,p)
cc-pVDZ
25.2
26.5
11.4
8.3
10.2
5.8
30.3
28.4
48.8
61.9
53.5
48.8
n.k.[a]
n.c.[a]
n.c.
1.632
1.631
1.636
1.632
1.654
1.650
1.616
1.613
1.617
1.613
1.606
1.616
n.k.
B3LYP + ZPE
BLYP + ZPE
M05-2X + ZPE
RIMP2
27.6
26.9
22.4
21.4
37.6
36.6
n.c.
n.c.
n.c.
n.c.
36.4
aug-cc-pVDZ
cc-pVTZ
cc-pVDZ
MP2
Experiment[b]
–
[a] n.c.: not calculated; n.k.: not known. [b] Ref.[8a]
Hartree–Fock theory slightly underestimates the BDE, but for the activation enthalpy (∆H‡rec,exp Ն 22.9 kcalmol–1) re-
yields better results than B3LYP, BLYP or MP2.
ported by Ingold and co-workers.[19]
According to Table 1, the recently developed[18] M05-2X
The strain energy of 1 can be calculated by evaluating
method is the only method among those used in this study the isodesmic reaction given in Equation (1).
to correctly predict the activation enthalpy for the cleavage
of 1. In combination with the 6-311++G(d,p) basis set, the tBu2CH–CHtBu2 + 4 CH3CH3 Ǟ
calculated value of ∆H‡ is 36.6 kcalmol–1, which is in excel-
Me2CH–CHMe2 + 4 C(CH3)4
(1)
lent agreement with the experimental value of ∆H‡
=
exp
36.4 kcalmol–1.[8a] However, even by using the modest 6-
The strain energy obtained from Equation (1) [M05-2X/
31G(d) basis set, the result obtained with M05-2X is very 6-311++G(d,p)] is Hs = 65.3 kcalmol–1. This is in excellent
good. The calculated value of ∆H‡ = 36.6 kcalmol–1 corre- agreement with the experimental value (obtained by meas-
sponds to a BDE of the central C–C bond in 1 of BDE = uring the heat of combustion of 1), which has been deter-
28.4 kcalmol–1 plus an activation enthalpy for the recombi- mined as Hs,exp = 66.3 kcalmol–1.[8c] At the B3LYP/6-
nation of 2 equiv. of 14 of ∆H‡ = 8.2 kcalmol–1. Note 311++G(d,p) level of theory, the strain energy of 1 is calcu-
rec
that this latter value is significantly below the lower limit lated as Hs = 76.3 kcalmol–1, which is clearly too high.
4342
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 4340–4345

Computational and Matrix Isolation Studies of Tetra-tert-butylethane
The fact that B3LYP and BLYP describe the energetics tions shown in Scheme 2. Note that the experimental data
of 1 poorly suggests that these methods might also be un-
suitable for calculating the strain energy of tetra-tert-butyl-
ethene (2). The strain energy of 2 was therefore re-evaluated
by calculating the isodesmic Equation (2)[3] at the M05-2X/
6-311++G(d,p) level of theory.
do not provide conclusive evidence for the formation of
neopentane, which has a very simple IR spectrum (B3LYP/
6-31G*: ν
= 1487, 1371, 1239 cm–1).
˜
calcd.
The experimental data indicate that 2,2-dimethylbutane
(10) is the most likely product formed upon complete de-
composition of 1. This molecule is formed in a series of
(2) bond cleavage/radical-pair disproportionation reactions, re-
sulting in a three-fold deisobutenylation of 1. The fact that
ethane is not formed as a product is easily rationalized as
its formation from 10 is predicted to be significantly endo-
thermic (Scheme 2). Note, however, that the calculations
predict 2,2,5,5-tetramethylhexane (9) to be the product
formed, as deisobutenylation of 9 to yield 10 is also pre-
dicted to be endothermic. At present, this discrepancy be-
tween the computational and matrix isolation results can-
not be resolved.
tBu2C=CtBu2 + 4 CH3CH3 Ǟ Me2C=CMe2 + 4 C(CH3)4
The strain energy of 2 thus calculated amounts to Hs
= 81.2 kcalmol–1 for the singlet-spin manifold and Hs =
35.3 kcalmol–1 for the triplet-spin manifold, which is sim-
ilar to the BLYP/Dzd values reported by Schleyer and
Schaefer and co-workers.[3] At the M05-2X/6-311++G(d,p)
level of theory, the triplet enthalpy of 2 is obtained as ∆HT
= 11.7 kcalmol–1, which is also in agreement with the value
(12.6 kcalmol–1) reported previously, and the C=C distance
is 1.374 Å, which also fits well with the findings by Schleyer
and Schaefer and co-workers.[3]
The M05-2X method gives excellent thermochemical
data for 1, whereas other common DFT methods, such as
B3LYP or BLYP, and MP2 theory do not. Table 1 shows
that the stability of 1 and the length of its central C–C bond
are related as a weak C–C bond corresponds to a long C–
C bond. The length of this bond is a function of the interac-
tion between the four tert-butyl groups placed around the
two central carbon atoms. This interaction consists of both
an attractive (dispersive interactions) and a repulsive com-
ponent (strain). Commonly used DFT methods, such as
BLYP or B3LYP, cannot describe dispersive interactions.[20]
They therefore underestimate the attractive component of
the interaction and thus predict a C–C BDE and an acti-
vation enthalpy for cleavage of 1 that is too low. MP2
theory, on the other hand, in principle is capable of describ-
ing dispersive interactions, but overestimates their
strength.[21] It therefore overestimates the stability of 1.
Among the methods tested, only the M05-2X method ap-
pears to achieve the correct balance between attractive and
repulsive intramolecular interactions in this molecule. Al-
though the M05-2X results are in excellent agreement with
Discussion
According to the results obtained by matrix isolation
spectroscopy, the thermal decay of 1, matrix-isolated in Xe,
takes place mostly by deisobutenylation and to a lesser de-
gree by loss of methane. Literature evidence[8a] as well as
the calculations presented in this study suggest that the cen-
tral C–C bond in 1 should be by far the most labile bond
in 1. This reaction initially leads to a pair of di-tert-bu-
tylmethyl radicals 14. Unlike the radical pairs resulting
from the fragmentation reactions around the periphery of
1, this radical pair cannot disproportionate into an alkane
and an alkene. Subsequent reactions are therefore limited
to recombination (I), the elimination of methyl radicals (II)
or the disproportionation into di-tert-butylmethane and di-
tert-butylcarbene (III), both of which may undergo further
decomposition (Scheme 3).
Scheme 3 illustrates that the thermal cleavage of the cen-
tral C–C bond in 1 might yield the same stable products the thermochemical data for 1 published by Rüchardt and
(isobutene and methane) as the peripheral cleavage reac- co-workers,[8] they are clearly inconsistent with the lower
Scheme 3. Possible decay channels upon initial thermal cleavage of the central C–C bond in 1. ∆ refers to thermal excitation by interaction
with an Xe exciplex.
Eur. J. Org. Chem. 2009, 4340–4345
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4343

G. Bucher
FULL PAPER
ical Organic Chemistry and funded by the EPSRC. The author is
grateful for this support.
limit for the activation enthalpy for the dimerization of the
di-tert-butylmethyl radical 14, as measured by Ingold and
co-workers employing ESR spectroscopy.[19] At present, no
explanation for this discrepancy can be offered.[22]
[1] D. Lenoir, C. Wattenbach, J. F. Liebman, Struct. Chem. 2006,
17, 419–422.
[2] H. Hopf, Classics in Hydrocarbon Chemistry, Wiley-VCH,
Weinheim, 2000.
Conclusions
[3] H. M. Sulzbach, E. Bolton, D. Lenoir, P. v. R. Schleyer, H. F.
Schaefer III, J. Am. Chem. Soc. 1996, 118, 9908–9914.
[4] a) T. G. Back, D. H. R. Barton, M. R. Britten-Kelly, F. S. J.
Guziec, J. Chem. Soc. Perkin Trans. 1 1976, 2079–2089; b) J. E.
Gano, R. H. Wettach, M. S. Platz, V. P. Senthilnathan, J. Am.
Chem. Soc. 1982, 104, 2326–2327; c) H. Bock, B. Berkner, B.
Hierholzer, D. Jaculi, Helv. Chim. Acta 1992, 75, 1798–1815.
[5] a) D. Lenoir, H. Burghard, J. Chem. Res. 1980, 4715–4725; b)
C. Villiers, M. Ephritikine, Chem. Eur. J. 2001, 7, 3043–3051.
[6] a) A. Krebs, W. Born, B. Kaletta, W.-U. Nickel, W. Rüger, Tet-
rahedron Lett. 1983, 24, 4821–4823; b) A. Krebs, W.-U. Nickel,
L. Tikwe, J. Kopf, Tetrahedron Lett. 1985, 26, 1639–1642; c) O.
Klein, H. Hopf, J. Grunenberg, Eur. J. Org. Chem. 2009, 2141–
2148.
[7] a) T. Loerzer, R. Gerke, W. Lüttke, Tetrahedron Lett. 1983, 24,
5861–5864; b) J. Deuter, H. Rodewald, H. Irngartinger, T. Lo-
erzer, W. Lüttke, Tetrahedron Lett. 1985, 26, 1031–1034.
[8] a) H.-D. Beckhaus, G. Hellmann, C. Rüchardt, Chem. Ber.
1978, 111, 72–83; b) S. Hellmann, H.-D. Beckhaus, C.
Rüchardt, Chem. Ber. 1983, 116, 2219–2237; c) M. A. Flamm-
ter Meer, H. D. Beckhaus, C. Rüchardt, Thermochim. Acta
1984, 80, 81–89.
Cryogenic pyrolysis of tetra-tert-butylethane (1), matrix-
isolated in xenon at T = 10 K, results in the multiple deiso-
butenylation of 1 with the likely formation of 2,2-dimeth-
ylbutane (10) in addition to isobutene and methane. Tetra-
tert-butylethene (2), however, is not formed. The data sug-
gest that tert-butyl substituents may thus potentially be
used as a protective group for matrix isolation spectroscopy,
to be removed by thermal excitation in xenon matrices.
Computational work performed on 1 indicates that the
thermochemistry of this sterically highly overloaded alkane
is poorly described by commonly used DFT, such as BLYP
or B3LYP, and MP2 methods, whereas the recent M05-2X
method yields energies in excellent agreement with pub-
lished experimental data.
Experimental Section
Tetra-tert-butylethane (1): This compound was synthesized by re-
ductive coupling of 3-chloro-2,2,4,4-tetramethylpentane, as de-
scribed by Rüchardt and co-workers.[8a] In slight variation of the
literature procedure, a commercial suspension of sodium metal in
toluene was used as the sodium source. The yield thus achieved
was higher than previously reported (up to 60%).
[9] G. Maier, C. Lautz, H. P. Reisenauer, J. Inf. Recording 2000,
25, 25–38.
[10] G. Maier, D. Jürgen, R. Tross, H. P. Reisenauer, B. A. Hess Jr,
L. J. Schaad, Chem. Phys. 1994, 189, 383–399.
[11] G. Maier, C. Lautz, Angew. Chem. Int. Ed. 1999, 38, 2038–
2041.
[12] IR spectra recorded after a smaller number of laser pulses do
not reveal the presence of significant amounts of compounds
other than those observed upon complete conversion of 1. This
is likely due to the fact that two-photon excitation requires
crystalline matrices, whereas the degree of crystallinity is small
in typical xenon matrices (c.f. ref.[9]). Therefore, molecules iso-
lated in unreactive matrix sites (low degree of local crystal-
linity) will show a much smaller reactivity than molecules in
reactive matrix sites (high degree of local crystallinity). Any
intermediary product formed in a more reactive matrix site will
thus decompose more rapidly than the unreacted precursor 1
present in less reactive matrix sites, which makes the observa-
tion of such intermediates difficult.
Matrix-Isolation Experiments: The matrix-isolation setup used has
been described previously.[23] Matrix-isolated samples of 1 in xenon
or argon matrices were obtained by the slow-spray-on technique,
with matrix temperatures Tdepo = 20 (Xe) or 30 K (Ar) during de-
position. Alkane 1 was kept at ambient temperature to allow for
sufficient vapour pressure. The excitation of the matrices contain-
ing 1 was performed by using a Lambda-Physik Compex 100 exci-
mer laser operated at 248 (KrF) or 193 nm (ArF). The pulse energy
at λ = 248 nm (193 nm) was 200 mJ/pulse (100 mJ/pulse), with a
repetition rate of 1 Hz.
Computational Methods: All DFT, MP2[24] and Hartree–Fock cal-
culations were performed by using the Gaussian03[25] suite of pro-
grams. RIMP2[26] optimizations with the cc-pVTZ, cc-pVDZ and
aug-cc-pVDZ basis sets[27] were performed by using the TUR-
BOMOLE software.[28] The DFT methods used include the BLYP
functional,[16] the B3LYP hybrid functional,[16] and the M05-2X
method,[18] in combination with the 6-31G(d) and 6-311++G(d,p)
basis sets.[29] Transition-state structures for the dimerization of 14
were optimized by using the Gaussian keyword combination guess
= (mix,always). The scaling factor (0.9614 for B3LYP) for the cal-
culated infrared spectra was taken from the literature.[30]
[13] L. H. Jones, S. A. Ekberg, B. I. Swanson, J. Chem. Phys. 1986,
85, 3203–3210.
[14] a) A. J. Barnes, J. D. R. Howells, J. Chem. Soc. Faraday Trans.
2 1973, 532–539; b) F. Hipler, R. A. Fischer, J. Müller,
PhysChemChemPhys 2005, 7, 731–737.
[15] a) P. Roubin, R. Kakou, P. Verlaque, M. Monnier, J. Pourcin,
H. Bodot, Chem. Phys. Lett. 1989, 160, 345–349; b) R.
Kakou Yao, J. Pourcin, J. Soc. Ouest-Afr. Chim. 1998, 005–006,
1–10.
[16] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[17] P. R. Schreiner, Angew. Chem. Int. Ed. 2007, 46, 4217–4219.
[18] Y. Zhao, N. E. Schultz, D. G. Truhlar, J. Chem. Theory Com-
put. 2006, 2, 364–382.
[19] G. D. Mendenhall, D. Griller, D. Lindsay, T. T. Tidwell, K. U.
Ingold, J. Am. Chem. Soc. 1974, 96, 2441–2447.
Supporting Information (see footnote on the first page of this arti-
cle): Figures S1–S3, Cartesian coordinates and the energies of sta-
tionary points.
[20] a) S. N. Kristyán, P. Pulay, Chem. Phys. Lett. 1994, 229, 175–
ˇ
180; b) P. Hobza, J. Sponer, T. Reschel, J. Comput. Chem. 1995,
Acknowledgments
16, 1315–1325; c) J. M. Pérez-Jordá, A. D. Becke, Chem. Phys.
Lett. 1995, 233, 134–137.
[21] J. G. Hill, J. A. Platts, H.-J. Werner, PhysChemChemPhys 2006,
8, 4072–4078.
The author thanks H. Bettinger for helpful discussions. Work per-
formed in Glasgow was supported by the Glasgow Centre for Phys-
4344
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 4340–4345

Computational and Matrix Isolation Studies of Tetra-tert-butylethane
[22] It is noted that the significantly lower limit for the barrier for
dimerization of 14 reported in ref.[19] is a priori incompatible
with the experimental observation that 1 can be prepared, for
example, by reduction of 3-chloro-2,2,4,4-tetramethylpentane.
Although an alternative mechanism for the formation of 1 in-
volving the formation of di-tert-butylmethylsodium has been
suggested,[19] the value presented in this study would provide a
more straightforward rationalization of the formation of 1 by
simple dimerization of 14.
perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A.
Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dap-
prich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q.
Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayak-
kara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian03, Re-
vision E.01, Gaussian, Inc., Wallingford, CT, 2004.
[26] F. Weigend, M. Häser, Theor. Chem. Acc. 1997, 97, 331–340.
[27] R. A. Kendall, T. H. Dunning Jr, J. Chem. Phys. 1992, 96,
6796–6806.
[23] W. W. Sander, J. Org. Chem. 1989, 54, 333–339.
[24] M. Head-Gordon, J. A. Pople, M. J. Frisch, Chem. Phys. Lett.
1988, 153, 503–506.
[25] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratch-
ian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-
[28] TURBOMOLE V5-9-1, University of Karlsruhe, 2007.
[29] R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54,
724–728.
[30] A. P. Scott, L. Radom, J. Phys. Chem. 1996, 100, 16502–16513.
Received: April 6, 2009
Published Online: July 17, 2009
Eur. J. Org. Chem. 2009, 4340–4345
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4345