Experimental and theoretical host-guest photochemistry; control of reactivity with host variation and theoretical treatment with a stress shaped reaction cavity; mechanistic and exploratory organic photochemistry

Article abstract of DOI:10.1016/S0040-4020(00)00504-4

In our previous studies prediction of the course of photochemical reactions in crystalline media was accomplished using molecular mechanics with imbedded quantum mechanically generated transition structures. However, there was a need for subsequent geometry optimization of the transition structure within the surrounding 'mini-crystal lattice'. One new approach was devised in which the molecules surrounding the transition structure were replaced by a rigid shell of inert gas atoms with subsequent ab initio geometry optimization of the imbedded transition structure. The present study aimed at providing a more realistic environment surrounding the reacting species. This made use of the recently developed Oniom computations of GAUSSIAN98 which permit ab initio computations on the reacting species while performing molecular mechanical computations on the surrounding molecules. Included was determination of the effect of reaction cavity size. A new analysis, 'Pairs', was developed giving the specific important interactions of atom pairs in a large system. Additionally, the present study extended our solid-state photochemistry to inclusion compounds. This study was both experimental and theoretical. Experimentally, five host-guest inclusion compounds having 4-p-cyanophenyl-4-phenylcyclohexenone as the guest were prepared and studied. Three afforded photoproduct resulting from cyanophenyl migration paralleling the regiochemistry in solution; however, the minor products of the solution chemistry were lacking. Strikingly, a fourth inclusion compound gave mainly phenyl migration. The fifth inclusion compound led to 1:1 phenyl versus cyanophenyl migration. The chiral hosts permitted synthesis of enantiomerically pure photoproducts. The Oniom computations required modification but properly predicted migratory behavior in accord with the experimental observations. Several theoretical conclusions in the present study were: (a) the effect of cavity size (b) the role of crystal relaxation and long range stress effects; (c) the reliability of least motion in predicting reactivity. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00504-4

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 6821±6831
Experimental and Theoretical Host±Guest Photochemistry;
Control of Reactivity with Host Variation and Theoretical
Treatment With a Stress Shaped Reaction Cavity;
Mechanistic and Exploratory Organic Photochemistry1,2
*
Howard E. Zimmerman, Igor V. Alabugin and Valeriya N. Smolenskaya
Contribution from the Chemistry Department of the University of Wisconsin, Madison, WI 53706, USA
Received 24 July 1999; accepted 7 September 1999
AbstractÐIn our previous studies prediction of the course of photochemical reactions in crystalline media was accomplished using
molecular mechanics with imbedded quantum mechanically generated transition structures. However, there was a need for subsequent
geometry optimization of the transition structure within the surrounding `mini-crystal lattice'. One new approach was devised in which the
molecules surrounding the transition structure were replaced by a rigid shell of inert gas atoms with subsequent ab initio geometry
optimization of the imbedded transition structure. The present study aimed at providing a more realistic environment surrounding the
reacting species. This made use of the recently developed Oniom computations of Gaussian98 which permit ab initio computations on
the reacting species while performing molecular mechanical computations on the surrounding molecules. Included was determination of the
effect of reaction cavity size. A new analysis, `Pairs', was developed giving the speci®c important interactions of atom pairs in a large system.
Additionally, the present study extended our solid-state photochemistry to inclusion compounds. This study was both experimental and
theoretical. Experimentally, ®ve host±guest inclusion compounds having 4-p-cyanophenyl-4-phenylcyclohexenone as the guest were
prepared and studied. Three afforded photoproduct resulting from cyanophenyl migration paralleling the regiochemistry in solution;
however, the minor products of the solution chemistry were lacking. Strikingly, a fourth inclusion compound gave mainly phenyl migration.
The ®fth inclusion compound led to 1:1 phenyl versus cyanophenyl migration. The chiral hosts permitted synthesis of enantiomerically pure
photoproducts. The Oniom computations required modi®cation but properly predicted migratory behavior in accord with the experimental
observations. Several theoretical conclusions in the present study were: (a) the effect of cavity size; (b) the role of crystal relaxation and long
range stress effects; (c) the reliability of least motion in predicting reactivity. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
crystal-lattice. The lattice geometry was obtained from
X-ray data and the transition structures utilized were
approximated by the ®rst intermediates leading to alterna-
tive reaction products. These intermediates were generated
by both molecular and quantum mechanics. However, due
to the size of the mini-crystal-lattice, it was not possible to
quantum mechanically geometry optimize the entire system.
Molecular mechanics optimization is unsuitable for open
shell species such as triplets. Accordingly, we developed a
method6 wherein the atoms nearest the reacting species
were transformed into rigid inert gas atoms and the remain-
ing atoms of the mini-crystal-lattice were annihilated. This
approach thus generated a ®xed shell with the shape of the
original surrounding lattice. With a surrounding rigid inert
gas shell, the reacting species was then geometry optimized
using ab initio quantum mechanics. However, while solving
the problem of proper ab initio geometry optimization of
open shell species (e.g. triplets) within the host shell, this
approach had the weakness of using an arti®cial lattice.7 By
including only van der Waals interactions between the
reacting species and the host lattice, hydrogen bonding
and p±p effects were omitted.
Much of organic photochemistry has focused on solution
reactivity. However, solid-state photochemistry is interest-
ing for a number of reasons. First, the course of the reaction
often differs from that in solution and reactions not possible
in solution may be obtained. Secondly, the factors control-
ling the course of the solid-state reactions still are uncertain.
Our interest in solid-state chemistry goes back more than a
decade. In these previous studies in solid-state organic
photochemistry, we have correlated our experimental
results with theory based on a quantitative model with the
reacting species surrounded by its neighbors. We termed
this aggregate a `mini-crystal-lattice', this based on X-ray
data. Initially we used three properties of the intermediate
reacting species, overlap with the surrounding lattice, least
motion, and volume increase.3 In subsequent research4,5 we
used molecular mechanics to assess the energy of a tran-
sition structure imbedded in the center of a rigid mini-
Keywords: mini-crystal lattice; chiral; host±guest photochemistry.
* Corresponding author. E-mail: zimmerman@chem.wisc.edu
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00504-4

6822
H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
Table 1.
Inclusion compound
Enantiomer imbedded
Host:guest ratio
Space group
Seebach±Toda 3
Seebach±Toda 3
Octa±Ph±Tetraol 4
Cy±Hex 5
R
S
S
RS
RS
1:1
2:1
1:1
1:1
1:1
P212121
P212121
2
P21/c
Pbca
Benzopinacol 6
Thus, we needed an approach which would permit ab initio
optimization of the guest reacting species while still retain-
ing at least molecular mechanical computation of the entire
mini-crystal lattice and also include lattice±guest inter-
action energy. Recently, Morokuma has devised combined
QM/MM methodology incorporated in Gaussian98,8 and
we planned to utilize this approach.
compound 3, also known as Taddol, which has proved
useful in forming inclusion compounds.11 A second was
the octaphenyl-tetraol 4; this followed the philosophy of
the Seebach±Toda structure but designed to be larger and
more extended. The synthesis simply involved the reaction
of diethyl d- or l-tartrate with 1,4-cyclohexanedione
followed by reaction with phenylmagnesium bromide. A
third host, cis-1,4-di(diphenyl-hydroxymethyl)cyclohexane
5, was synthesized and proved especially useful. A last host
used was benzopinacol 6.
Experimentally, it was our intention to pursue the behavior
of reacting species in inclusion compounds; hitherto, our
studies have focused on homogeneous crystal lattices. The
advantage of host±guest photochemistry is that one can
obtain a myriad of different crystal environments, while
with simple crystals it is dif®cult to obtain more than one
morphology. The photochemistry of 4-p-cyanophenyl-4-
phenylcyclohexenone (1) as a test case promised to be
especially interesting in view of our past studies on the
compound, both in solution9 and in the crystalline state.10
The solution chemistry is shown in Eq. (1).
The Seebach±Toda host gave two different inclusion
compounds. One was a 1:1 complex 3Ia while the other
was a 2:1 (host:guest) 3Ib. In the crystallizations using the
Seebach±Toda host two observations were made. (a) The
1:1 and 2:1 inclusion complexes (i.e. 3Ia and 3Ib) contained
different enantiomers. (b) Finally, recovered unincluded
enone was found to be essentially enantiomerically pure.
The enantiomers selectively complexed in the different
chiral hosts are outlined in Table 1.
Partially in analogy, Octa±Ph±Tetraol
4 selectively
included the S-enantiomer; and the unincorporated enone,
again, was one enantiomer, the R isomer. This 1:1 complex
4I contained two molecules of ether in addition to the enone.
A cogent point is that in addition to the photochemical
usages of the chiral inclusion compounds, these are useful
for resolution. Thus, as noted, one enantiomer is not incor-
porated into the complex and is available. But, the inclusion
compound is a source of the enantiomer generally used
photochemically.12
ꢀ1†
The remaining hosts were achiral. Both the cyclohexyl-bis-
benzhydrol 5 and benzopinacol 6 afforded 1:1 inclusion
compounds, 5I and 6I, respectively, in quantitative yields;
there was no complication arising from only one of two
enantiomers being incorporated.
Results: Host±Guest Photochemistry
As noted, different hosts and the corresponding different
inclusion compounds can be anticipated to afford different
crystal lattices. For the present study four different host
molecules were employed. One is the Seebach±Toda
With ®ve inclusion compounds in hand, we obtained X-ray
structures from four of theseÐthe Seebach±Toda 1:1 (3Ia),

H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
6823
Scheme 1. Regiochemical results of photolysis of the chiral inclusion complexes. Note: T±S refers to the Seebach±Toda Complex 3I and Octa±Ph refers to
the Octa±Ph±Tetraol Complex 4I.
Seebach±Toda 2:1 (3Ib), the Cy±Hex (5I) 1:1, and the
benzopinacol 1:1 (6I) complexes. Next, we proceeded
with the irradiation of each of these. In contrast to the solu-
tion behavior with preferential cyanophenyl migration, the
reaction regiochemistry in the ®ve inclusion compounds
was a function of the host structure and the composition
of the inclusion compounds.
preferentially was equatorial. This differed from our earlier
examples.4 Note discussion below.
Finally, we note that in these irradiations, the photo-
chemistry focussed on conversions below ten percent
although rather high conversions were possible without
change in the chemistry for several cases. For example, in
the case of the benzopinacol complex conversions up to
seventy percent led to unchanged regiochemistry. With
the inclusion complex 5I with cyclohexyl-bis-benzhydrol
5 the ratio of phenyl migration diminished. Details on
conversions are given in Experimental Section.
In the case of the inclusion compound 5I derived from the
cyanophenyl enone 1 and the cyclohexyl-bis-benzhydrol 5,
phenyl migration was preferred 5:1. This, of course, is the
reverse of the behavior seen in solution photochemistry. In
the case of the Seebach±Toda 2:1 complex 3Ib, a 1:1 ratio
of phenyl to cyanophenyl migration was encountered.
Results: Theoretical and Computational Treatment of
the Reactions
The remaining inclusion compoundsÐthe 1:1 complexes
derived from the Seebach±Toda host 3, the octaphenyl-tetraol
4 host, and from benzopinacolÐled to exclusive p-cyano-
phenyl migration. The photolyses of the chiral inclusion
compounds are outlined in Scheme 1. The achiral cases gave
racemates and their photolyses are depicted in Scheme 2. The
mechanistic source of the different reaction regiochemistries
is considered in the Interpretative Discussion section.
Recently, very convenient combined quantum mechanics-
molecular mechanics programming has become available.8
This permits ab initio computation on part of a very large
molecule while utilizing molecular mechanics on the
remainder. This is the method of Morokuma8a,b and termed
`Oniom'.13 Several force ®elds are incorporated in Oniom.
Dreiding14 and UFF15 are capable of treating cyano
compounds. But, the latter did not have parameterization
for hydrogen bonding which is ignored and neither method
Another item of interest came from the X-ray analyses of
the reactants. In only one case the group which migrated
Scheme 2. Regiochemical results of photolysis of the achiral inclusion complexes. Note: Cy±Hex refers to the cyclohexyl-bis-benzhydrol complex 5I and
Bz±Pin refers to the benzopinacol complex.

6824
H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
Figure 1. Energies as a function of number of inner shell molecules. diamonds for phenyl, circles for cyanophenyl migration and triangles give the energy
difference for the two migrations: A, MM3 values; B, Ab initio energy of the guest.
gave consistent results with the type molecules typical of
our organic photochemistry. Additionally, our total number
of atoms was beyond that handled by ONIOM, circa 4000.
In this computation, three layers were included: (a) the
reactant molecule; (b) an inner shell of molecules surround-
ing the reactant; and (c) an outer shell surrounding the inner
shell.
metry of the included guest molecule. With this approach
ONIOM now needed to deal only with the reactant guest
molecule and the inner shell, since the inner shell already
had been optimized within the rigid outer shell by MM3;
and the 4000 atom limitation was circumvented. A major
effect of the pre-optimization is to relax the inner lattice
shell surrounding the central, reacting guest and thus
`soften' the reaction medium. Any change in the geometry
of reacting guest by the pre-optimization is unimportant
since this is followed by the ONIOM ab initio geometry
optimization. In addition to providing a reasonable lattice,
this approach gave information on the role of distant molec-
ular stress effects (vide infra). Finally, following the pre-
optimization, Oniom optimization was carried out on the
We used MM316 pre-optimization of the reacting guest and
a next (inner) shell of circa 800 surrounding molecules,
keeping the outer shell frozen. The inner shell was taken
such that increasing their number did not change the total
MM3 energy. (e.g. See Fig. 1). Addition of further layers to
the outer shell also did not affect the guest energy or geo-
Table 2. Energies of alternative intermediates for the inclusion compounds
Inclusion
compound
Ph migration
intermediatea
CN±Ph migration
intermediatea
DEb
Type of computationc
Seebach±Toda 1:1
3Ia
2851.4088
2850.6172
2855.4092
1.5988
2851.4272
2850.6317
2855.4249
1.5861
11.5
9.1
9.9
8.0
ONIOM(3-21G)/Dreiding MM
ONIOM(3-21G)/MM3 Substitution
ONIOM(6-31Gp)/MM3 Substitution
Dreiding MM
2.0984
2.0841
8.9
MM3
2857.4235
2851.1435
2850.1239
2854.9183
1.8659
2857.4230
2851.1285
2850.1292
2854.9208
1.8853
20.3
29.4
3.3
ROHF/6-31Gp, isolated biradical
ONIOM(3-21G)/Dreiding MM
ONIOM(3-21G)/MM3 Substitution
ONIOM(6-31Gp)/MM3 Substitution
Dreiding MM
Seebach±Toda 2:1
3Ib
1.6
212.2
1.1
6.6
17.7
0.0
22.1
10.9
22.9
6.6
2.5912
2.5895
MM3
2857.4193
2851.2798
2850.7607
2855.5556
1.7268
2857.4298
851.3080
2850.7607
2855.5523
1.7095
ROHF/6-31Gp, isolated biradical
ONIOM(3-21G)/Dreiding
ONIOM(3-21G)/MM3 Substitution
ONIOM(6-31Gp)/MM3 Substitution
Dreiding MM
Cyclohexyldiol
5I
1.9535
1.9581
MM3
2857.4189
2851.4613
2850.4788
2855.2728
1.5470
2857.4294
2851.4659
2850.4986
2855.2893
1.5530
ROHF/6-31Gp, isolated biradical
ONIOM(3-21G)/Dreiding
ONIOM(3-21G)/MM3 Substitution
ONIOM(6-31Gp)/MM3 Substitution
Dreiding MM
Benzopinacol
6I
2.9
12.5
10.4
23.8
9.0
2.2351
2857.4181
2.2207
2857.4284
MM3
ROHF/6-31Gp, isolated biradical
6.5
a Absolute energies are in Hartrees (627.5 kcal mol21).
b Energy differences are in kcal mol21
.
c Six methods, each for Oniom (Rohf/3-21G)/Dreiding geometry but subsequent computation as indicated, (i) Original computation, (ii) Substitution of
MM3 for Dreiding energies, (iii) Same as (ii) but with 6-31Gp, (iv) Dreiding MM on entirety, (v) MM3 on entirety, (vi) ROHF on the central, triplet guest
alone as optimized within the shell. The bold lines contain the preferred computation.

H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
6825
Scheme 3. Type B enone rearrangement mechanism. Note: 8a AˆH, BˆCN for Cyanophenyl Migration; 8b AˆCN, BˆH for phenyl migration. The three-
dimensional n±pp con®guration is shown with y's as in-plane py electrons, solid dots as p-system electrons, and small circles as sp-hybrid electrons.17
central lattice system (guest1inner shell) with the outer
lattice shell being held ®xed.
tional methods correctly predict experiment for one or more
of the four inclusion complexes, only entry three ONIOM
(6-31Gp)/MM3 ®ts all four cases.
We note that Oniom affords its energies in three components
which comprise the total; note Eq. (2). Thus there is ¯exi-
bility in taking an Oniom computation and replacing one or
more of the three ONIOM components obtained. This, in
effect, is an improvement of the original Oniom results with
the molecular mechanics components being replaced by
their MM3 counterparts (note computation details below).
It is of interest to consider the importance of the MM3 pre-
optimization used in these computations. Table 3 gives the
DE values for Oniom/(6-31Gp)/MM3 without pre-optimiza-
tion. A more complete table is given in Appendix A. We
note that preferential phenyl migration in inclusion
compound 5I is never predicted without this use of MM3.
Etot ˆ Einnerꢀab initio† 1 EouterꢀMM† 2 EinnerꢀMM†
ꢀ2†
Table 2 gives energies for the phenyl and cyanophenyl
migration intermediates (8a and 8b, note also Scheme 3)
and compares the two alternative reaction routes. For each
route, six entries corresponding to different theoretical treat-
ments are given. Thus, the ®rst treatment of the six corre-
sponds to utilization of Oniom computations without
modi®cation. The second entry has the calculated MM3
energy replacing the Dreiding component. The third entry
differs only in a 6-31Gp basis replacing the 3-21G of the
previous line. The fourth and ®fth items compare Dreiding
and MM3 molecular mechanics computations for the entire
lattice without use of quantum mechanics. The last entry
gives the ROHF/6-31Gp energy of the triplet reaction inter-
mediate with geometry optimized within the shell. The DE`s
given correspond to the difference in energies between the
phenyl migration and cyanophenyl migration species in the
medium speci®ed. A negative DE corresponds to preferen-
tial phenyl migration while a positive value would predict
cyanophenyl bridging. As is discussed in the Interpretative
Discussion Section (vide infra), although several computa-
Interpretative Discussion
In the rearrangement of 4-cyanophenyl-4-phenylcyclohex-
enone the preferential cyanophenyl migration has been
noted9 to arise from the greater delocalization resulting
when the cyano substituent in intermediate 8 (see Scheme
3) is on the migrating group. The reasoning agrees with the
experimental observation18 that cyano substitution on the
migrating group has a large positive rate effect while there
is a minimal acceleration with substitution on the non-
migrating; this demonstrates that the rate-limiting portion
on the hypersurface comes early and involves the migration.
This reasoning has now been con®rmed by ab initio compu-
tations at the ROHF/6-31Gp level which indicate that the
lowest energy triplet bridged species has cyanophenyl
migrating. A minor complication is that, for each migrating
group, there are two local minima, differing in the dihedral
angles of the ±CH2±CH2 group. These derive computation-
ally, using least motion, from equatorial and axial groups
migrating. However, for each conformation, cyanophenyl
migration is preferred as noted in Table 4.
Table 3. Energies of alternative migrating species without pre-optimization
DEa,b
Inclusion compound
DEa,b
Table 4. Gas phase energies of the bridged triplet species 8
Inclusion compound
Structure of intermediate
E (ROHF/6-31Gp)
DE ( kcal mol21
Seebach±Toda 1:1 3Ia
Seebach±Toda 2:1 3Ib
21.4
7.0
Cyclohexyldiol 5I
Benzopinacol 6I
14.7
35.8
8a (Equatorial PhCN)
8b (Equatorial Ph)
8a (Axial PhCN)
8b (Axial Ph)
2857.4331
2857.4280
2857.4292
2857.4241
0
a Computations using the ab initio/6-31Gp with MM3 replacement of
Dreiding energies but without pre-optimization.
3.2
2.5
5.7
b kcal mol21
.

6826
H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
Ê
Table 5. RMS motion values (RMS values are in A)
Inclusion
Migrating group
RMS motion
Inclusion complex
Migrating group
RMS motion
Seebach±Toda 3Ia
Seebach±Toda 3Ia
Seebach±Toda 3Ib
Seebach±Toda 3Ib
Phenyl
0.94
1.23
1.39
0.80
Cy±Hex 5I
Cy±Hex 5I
Benzopin 6I
Benzopin 6I
Phenyl
1.52
0.97
1.58
0.82
Cyano±Ph
Phenyl
Cyano±Ph
Cyano±Ph
Phenyl
Cyano±Ph
Some explanation of Table 4 is needed. Thus, there are two
conformers of the enone reactant, one with the cyanophenyl
pseudo equatorial and the other with this group pseudo
axial. In our previous studies we have found that often it
is the pseudo equatorial group which migrates preferen-
tially. Not only is the ipso carbon of an equatorial aryl
group marginally closer to the enone b-carbon (circa by
formations by the shape of surrounding crystal lattices.
Thus, a reaction less favored in solution or the gas phase
by an energy of at least 3.2 kcal mol21 (see Table 4) can be
obtained in the solid state.
A ®nal aspect deals with the concept of least motion. Least
motion needs to be thought about in terms of the required
motion in proceeding from the actual reactant geometry in
the inclusion compound to afford the bridged triplet diradi-
cals 8a and 8b. This principle is violated in three of the four
cases presently studied. The RMS Motion Values obtained
from Macromodel19 are given in Table 5. In agreement with
the qualitative discussion less motion is required. In fact, in
some of our previous studies4 there was a preference for
migration of the equatorial group in accord with least
motion effects. More recently least motion was found to
be unreliable5 and the present study con®rms that unrelia-
bility.
Ê
0.1 A from AM1 geometry), but more importantly the
preferred conformation is found to have the p-system of
the equatorial aryl group facing the enone b-carbon while
the edge of the axial aryl tends to be oriented towards the
b-carbon. This is observed both for isolated molecules and
in most enone crystal lattices we have studied.4
In solution one often has a myriad of potential conforma-
tions of a reacting species to consider. In contrast, for each
inclusion compound there will be a speci®c frozen geo-
metry. Of the four inclusion compounds presently studied
one, the Seebach±Toda 1:1 complex 3Ia, had the cyano-
phenyl group pseudo axial. The remaining three-the
Seebach±Toda 2:1 complex 3Ib, the benzopinacol 6I
complex, and the Cy±Hex 5I-all possessed pseudo equa-
torial cyanophenyl groups. Hence, one unique facet of crys-
tal lattice photochemistry is that one is dealing with
reactivity of isolated frozen conformers, here each with an
axial or an equatorial arrangement. Nevertheless, the axial
versus equatorial relationship did not directly correlate with
the regioselectivity of migration in the present study.
Having discussed least motion as not being controlling, we
now need to discuss what is controlling. One clue comes
from the `Pairs Analysis' which determines which atom
pairs are closest and contribute most to the stabilization or
destabilization of the host±rearranging triplet inclusion
complex. One observation from the pairs analysis is that
the preferred intermediate species invariably has a close
hydrogen bonding interaction between a host hydroxyl
and the triplet oxygen; these are stabilizing interactions.
But additionally, the preferred intermediates imbedded in
the relaxed lattice, but not in the frozen one, lack the close
repulsive atom±atom proximities seen for the unfavored
species.
Thus, the Cy±Hex inclusion complex 5I, which had a
pseudo equatorial cyanophenyl group exhibited mainly
axial phenyl migration. The Seebach±Toda 2:1 complex
3Ib which also had a pseudo equatorial cyanophenyl
group led to 1:1 axial phenyl and equatorial cyanophenyl
migration while the benzopinacol complex with a similar
conformation gave only equatorial cyanophenyl migration.
Finally, the Seebach±Toda 1:1 complex 3Ia which had an
axial cyanophenyl group exhibited axial cyanophenyl
migration.
Another item requiring discussion is the relaxation of the
inner shell of molecules immediately surrounding the react-
ing guest. In the present study, this role was played by the
pre-optimization. Thus, pre-optimization permits the inner-
shell to relax and takes lattice stress into account. Not only
shape but also the mechanical properties of the lattice are
included since the cavity becomes adaptable to the guest.
The most striking result is the preferential phenyl migration
in the Cy±Hex complex 5I. The 1:1 regioselectivity of the
Seebach±Toda 2:1 complex 3Ib is also rather interesting.
We see that the electronic factor favoring cyanophenyl
migration is overridden by the necessity of reacting in an
irregular cavity. Overall, there is an extreme variation of
regioselectivity with a total dependence on the speci®c
host employed.
Closely related and of real interest is the evidence bearing
on long range stress effects. This comes from Fig. 1 which
gives the total energy as a function of number of molecules
considered as inner shell entities. Remembering that the
inner shell molecules are subject to relaxation in the pre-
optimization, we note that with a relatively small number of
inner shell molecules, the energy diminishes as we increase
the number of molecules subject to relaxation. However, a
limit is rapidly reached and addition of further molecules
which may relax does not lower the total energy further.
This provides a good assessment of distance at which stress
effects are relevant. This distance is in the range of 20 to
It was especially interesting to ®nd that in the crystal lattice
photochemistry, not only did the migratory behavior depend
on the host and the crystal lattice, but also that the crystal
lattice forces could override the gas phase electronic effect
with the result of predominant phenyl migration in one case.
This then further illustrates the control of excited state trans-
Ê
30 A and this corresponds to one layer of neighbors
surrounding the reacting species. This result accords with

H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
6827
the suggestions of McBride20 regarding the role of crystal
stress on the reaction course.
Mercury lamp with a water cooling well. The samples
were photolyzed through a 2-mm Pyrex ®lter. The reaction
mixtures were dissolved in CDCl3 after photolysis, dried
over sodium sulfate, analyzed by H NMR and separated
by column chromatography on silica gel.
1
Conclusion
This study has presented new host±guest photochemistry
and has focussed on the use of ONIOM computations to
permit ab initio geometry optimization of open shell
species. It has dealt with the role of crystal lattice relaxation,
stress effects and cavity dominance over electronic effects.
(R,R,R,R)-Tetraethyl-1,4,9,12-tetraoxadispiro[4.2.4.2]-
tetradecane-2,3,10,11-tetracarboxylate. A solution of
2.24 g (20 mmol) of 1,4-cyclohexanedione, 8.6 g
(40.5 mmol) of the diethyl ester of (R,R)-tartaric acid and
0.40 g of p-toluenesulfonic acid in benzene was re¯uxed
with azeotropic removal of water for 30 h. The cooled reac-
tion mixture was washed with saturated sodium bicarbonate,
then water, extracted with methylene chloride and dried
over magnesium sulfate. Removal of solvent in vacuo
afforded 9.4 g (96%) of oily product which was pure by
NMR analysis and could be directly used in the next reac-
tion with phenylmagnesium bromide. An analytical sample
was obtained by extraction of the oil with boiling heptane to
give white thin plates, mp 63±768C, followed by crystal-
lization from ethanol. White needles, mp 75±778C. 1H
NMR (CDCl3, TMS): d 4.79 (s, 4H), 4.30±4.23 (q, 8H,
Jˆ6.9 Hz), 1.95 (s, 8H), 1.31 (t, 12H, Jˆ6.9 Hz,). 13C
NMR: 169.78, 113.25, 77.0, 61.88, 32.76, 14.07. IR
(KBr): 2940, 2978, 1753, 1584, 1221, 1128. [a]2D0ˆ227
(acetone, c2ˆ0.080). HRMS, m/e (M1): 488.4815. Calcd
for C22H32O12, 488.4823. Anal. Calcd for C22H32O12: C,
54.09, H, 6.60. Found: C, 53.85, H, 6.73. The (S,S,S,S)-
isomer was obtained in a similar manner.
Over the last decade our aim in solid-state photochemistry
has been to determine what factors are controlling. A wealth
of superb literature efforts have had the same pursuit.21 We
note that about a decade ago Gavezzotti commented21a that
`The general problem of obtaining information on solid-
state reactivity from a theoretical calculation has not yet
been tackled in a systematic way'.21a Nevertheless, over
this last decade there has been a variety of approaches,21
many elegant.
Experimental
General procedures
All reactions were performed under an atmosphere of dry
nitrogen. Column chromatography was performed on silica
gel (Matheson, Coleman and Bell, grade 62, 60±200 mesh)
mixed with Sylvania 2282 phosphor and slurry packed into
quartz columns to allow monitoring with a hand-held UV
(R,R,R,R)-Diphenyl(3,10,11-tri(hydroxy(diphenyl)methyl)-
1,4,9,12-tetraoxadispiro[4.2.4.2]tetradec-2-yl)methanol (4).
To a solution of 9.0 mmol of phenylmagnesium bromide in
25.0 ml of THF at 0±58C was added a solution of 488 mg
(1.0 mmol) of (R,R,R,R)-tetraethyl-1,4,9,12-tetraoxadi-
1
lamp. Melting points are uncorrected. H and 13C NMR
spectra were recorded at 300 and 75 MHz respectively,
using CDCl3 as solvent and are reported in ppm down®eld
from tetramethylsilane. Solvents were dried following
standard methods.
spiro[4.2.4.2]tetradecane-2,3,10,11-tetracarboxylate
in
8.0 ml of THF. The solution was stirred at 258C for 1 h,
warmed to room temperature and re¯uxed for 9 h. The reac-
tion mixture was poured into 100 ml of saturated ammoni-
um chloride solution and extracted with methylene chloride.
The combined organic fractions were washed with water
and dried over magnesium sulfate. Solvent was removed
in vacuo at 508C to afford a 1:4 complex of the product
and THF as a viscous oil. The oil was dissolved in 80 ml
of carbon tetrachloride, concentrated again in vacuo and
dried at 8 mm Hg/808C for 3 h to afford 0.95 g of an yellow
solid. The solid was dissolved in hot methanol, re¯uxed for
5 min and cooled to 258C. White crystals of complex with
methanol were ®ltered, mp 2738C. The guest-free host was
obtained by removal of the guest solvent in vacuo. The yield
was 0.69 g (74%) of white solid, mp 269±2718C. 1H NMR
(CDCl3, 300 MHz): d 7.3 (m, 40H), 4.6 (s, 4H), 4.0 (s, 4H),
1.4 (m, 8H). 13C NMR: 145.86, 142.43, 128.50, 128.18,
127.56, 127.18, 109.15, 80.40, 78.15, 33.37. IR (KBr):
3389, 3089, 3058, 3033, 2961, 2924, 1635, 1494, 1447,
1376, 1139, 1100. [a]D20ˆ219 (acetone, cˆ0.054).
HRMS, m/e (M1): 929.1041. Calcd for C62H56O8,
929.1032. Anal. Calcd for C62H56O8: C, 80.15, H, 6.08.
Found: C, 79.98, H, 6.19. This structure was con®rmed by
X-ray analysis of complex with CH2Cl2. The (S,S,S,S)-
isomer was obtained in a similar manner.
General procedure for X-ray crystallography analysis
X-Ray diffraction data were collected on a Siemens P4/CCD
diffractometer for single crystals of each compound.
Lorentz and polarization corrections were applied and
each structure was solved under the appropriate space
group symmetry by direct methods using Shelx86 or
shelxtl.22 Hydrogen atom positions were calculated
and re®ned with a rigid model. Full-matrix least-square
re®nement on F2 was carried out employing anisotropic
displacement parameters for all non-hydrogen atoms and
isotropic displacement parameters for all hydrogen atoms.
The coordinates for all compounds studied by X-ray crystal-
lography were deposited with the Cambridge Crystallo-
graphic Data Center. The coordinates can be obtained,
on request, from the Director, Cambridge Crystallo-
graphic Data Center, 12 Union Road, Cambridge, CB2
1EZ, UK.
General procedure for solid-state photolysis
Large crystals were gently cracked and placed between two
Pyrex slides. The slides were placed in a Pyrex beaker ®lled
with water and placed in a larger beaker ®lled with mixture
of ice and water. The slides were 2.5 cm from a 400 W
cis-1,4-Di(diphenylhydroxymethyl)cyclohexane (5). A

6828
H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
solution of 7.0 g (0.035 mol) of dimethyl 1,4-cyclohexane-
dicarboxylate (a 1.7:1 mixture of cis- and trans-isomers) in
45 ml of THF was added to a solution of 5 equivalents of
phenylmagnesium bromide in 50 ml of THF at room
temperature. The reaction mixture was re¯uxed for 13 h,
then poured into 100 ml of saturated NH4Cl solution and
chloroform extracted. The combined organic fractions
were washed with water, dried over magnesium sulfate,
and concentrated in vacuo to afford 12.3 g (78.5%) of
white crystals of a mixture of cis- and trans-isomeric
products. The cis-isomer, 2.6 g (33.6%, based on cis-
ester), was isolated by fractional crystallization from 7:1
remained in solution (48 mg, 88%). [a]2D0ˆ211 (chloro-
form, cˆ0.06).
Analysis of the guest left in mother liquor. A solution of
186 mg (0.20 mmol) of the host (4) and 110 mg
(0.40 mmol) of the enone (1) in 15 ml of 1:1 ether:hexane
mixture was left covered with `Para®lm' with several small
apertures. The crystals of the 1:1:2Et2O complex (240 mg,
98%) were ®ltered. Guest left in the mother liquor (chloro-
form, cˆ0.034) had [a]2D0ˆ12.5. Two crystallizations
raised the rotation to 113 (chloroform, cˆ0.028).
1
ethanol:chloroform; mp 195±1968C. H NMR (CDCl3): d
Crystallization behavior of 4-cyanophenyl-4-phenyl-
cyclohexenone/`Toda±Seebach' (3) system. Using a 1:1
host:guest ratio. A solution of 0.273 g (1.0 mmol) of the
enone and 0.466 g (1.0 mmol) of (R,R)-host (3) in 35 ml
of 1:1 ether:hexane was covered with `Para®lm' with
several small apertures. Three crystalline fractions were
obtained. (a) After 3 days 360 mg (97%, based on one
enantiomer of the guest) of 4±6 mm prismatic crystals of
the 1:1 inclusion compound (3Ia) were formed, mp 123±
124.58C. The mother liquor left in an open ¯ask afforded a
mixture of (b) small microcrystalline aggregates (40 mg,
29% of the remaining enantiomer of 4-cyanophenyl-4-
phenylcyclohexenone (1), mp 758C) and (c) ®ne needles
of the 2:1 complex (3Ib), mp 150±1538C, (170 mg, 28%
based on the remaining enantiomer) after standing over-
night. Mass-balance 77%, the rest of the reaction mixture
was a mixture of small crystals of the enone (1) and the 2:1
complex (3Ib).
7.3 (m, 20H), 2.7 (br.s., 2H), 2.15 (s, 2H), 1.5 (m, 8H).
HRMS m/z (M1) Calcd for C32H32O2: 448.6101, Obsd:
448.6113. 13C NMR (CDCl3): 146.8, 128.1, 126.3, 125.6,
81.4, 40.4, 23.8. The structure was established by X-ray
analysis of the inclusion compound (vide infra).
Crystallization behavior of 4-cyanophenyl-4-phenylcyclo-
hexenone and (R,R,R,R2)-`octaphenyltetraol' (4) system.
Using a 1:1 host:guest ratio. A solution of 186 mg
(0.20 mmol) of (R,R,R,R)-(4) and 55 mg (0.20 mmol) of
the enone (1) in 15 ml of 1:1 ether:hexane was left covered
with `Para®lm' with several small apertures for one day.
The crystals had a 0.55:1 guest:host ratio. The complex
contained two equivalents of ether as a third component.
Using 2:1 a host:guest ratio. (a) A solution of 186 mg
(0.20 mmmol) of the host (4) and 110 mg (0.40 mmol) of
the enone (1) in 15 ml of 1:1 ether:hexane was left covered
with `Para®lm' with several small apertures. The crystals of
the 1:1:2Et2O complex (240 mg, 91%) were formed over-
night. The melting behavior of the complex was interesting.
At 1388C, the compound melted and very intensive evapo-
ration of ether was observed. Then the melt solidi®ed again
and melted in the range of 230±2448C. (b) A solution of
1.696 g (1.82 mmmol) of the host (4) and 998 mg
(3.65 mmol) of the enone (1) in 40 ml of 1:1 ether:hexane
was left covered with `Para®lm' with several small aper-
tures. When 15 ml of the solvent was left crystals of the 1:1
complex were formed as it was described above (2.420 g, 98
%). After ®ltration the mother liquor was left covered with
`Para®lm' with several small apertures. White crystals
(128 mg, 28.5% on one enantiomer) of the enone were
formed overnight, mp 73±768C, [a]2D0ˆ112 (cˆ0.10,
chloroform).
Using
a 2:1 host:guest ratio. Solution of 0.506 g
(1.853 mmol) of the enone and 0.432 g (0.926 mmol) of
(R,R)-host (3) in 40 ml of 1:1 ether:hexane was left covered
with `Para®lm' with several small apertures for 4 days. The
solvent volume decreased to 10 ml, and a mixture of large
prismatic crystals and small microcrystalline aggregates
formed (690 mg). The big prisms were separated by hand
(470 mg, 69%) and found to be the 1:1 inclusion compound,
mp 123±124.58C. The microcrystalline aggregates (120 mg,
24%) were found to be 4-cyanophenyl-4-phenylcyclohexe-
none, mp 758C, [a]2D0ˆ22.6 (chloroform, cˆ0.06). The
separated mother liquor was left in an open ¯ask and ®ne
needles of the 2:1 host (3): enone (1) complex were formed
after standing overnight (50 mg), mp 151.5±1538C. The
mass-balance 68%, the rest of the reaction mixture was a
mixture of small crystals.
Using 4:1 a host:guest ratio. A solution of 186 mg
(0.20 mmol) of the host (4) and 220 mg (0.80 mmol) of
the enone (1) in 15 ml of 1:1 ether:hexane was prepared.
After 15 min, shiny crystals of the inclusion complex
formed. In several hours the precipitation was completed
to give 238 mg (98%) of ®ne transparent crystals of the
1:1 complex.
Using 2:3 host:guest ratio; the 2/1 inclusion compound
(3Ib). A solution of 0.273 g (1.0 mmol) of the enone and
0.699 g (1.5 mmol) of (R,R)-(3) in 35 ml of 1:1 ether:hexane
was kept covered with `Para®lm' with several small aper-
tures at 2208C. Well-shaped needles of the 2:1 complex
(3Ib) were formed ®rst, 393 mg, 65% (based on one
enantiomer of the guest), mp 151.5±153.58C. Prisms of
the 1:1 inclusion compound crystallized subsequently
(3Ia), 355 mg, 96% (based on the one enantiomer of the
guest), mp 119±1228C. The mass-balance 77%, the rest of
the reaction mixture was a mixture of small crystals.
Resolution of racemic 4-cyanophenyl-4-phenylcyclohex-
enone; enantioselectivity of the inclusion process.
Analysis of the guest inside the inclusion complex. Re¯ux-
ing of 270 mg of the complex (4I) in methanol resulted
in a relatively slow decomposition of the starting inclu-
sion compound and formation of a new complex with
methanol. The 4-cyanophenyl-4-phenyl-cyclohexenone
1:1 Complex of cis-1,4-di(diphenylhydroxymethyl)cyclo-
hexane and 4-p-cyanophenyl-4-phenylcyclohexenone
(5I). cis-1,4-Di(diphenylhydroxymethyl)cyclohexane 0.10 g

H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
6829
(0.22 mmol) and 0.061 g (0.22 mmol) of the enone were
dissolved in 30 ml of 4:1 ether:hexane and left covered
with `Para®lm' with several small apertures. After 2 days
crystals of a 1:1 complex were obtained quantitatively; mp
Exploratory solid-state photolysis of 2:1 complex of
`Toda±Seebach' host (3) and 4-cyanophenyl-4-phenyl-
cyclohex-2-en-1-one. The 2:1 complex (3Ib) (120 mg,
0.10 mmol) was photolyzed for 15 min. NMR analysis of
the slightly yellow solid reaction mixture indicated forma-
tion of 2:1 mixture of trans-5-phenyl-6-p-cyanophenyl-
bicyclo[3.1.0]hex-2-one and trans-5-p-cyano-phenyl-6-
phenylbicyclo[3.1.0]hex-2-one (conversion 58%, analyzed
by 1H NMR). The products were isolated by column
chromatography on silica gel using 10% ether in hexane
as eluent. Fraction 1 contained 6.0 mg (22%) of trans-5-
phenyl-6-p-cyanophenylbicyclo[3.1.0]hex-2-one. Fraction
2 contained 3.0 mg (11%) of trans-5-p-cyanophenyl-6-
phenylbicyclo[3.1.0]hex-2-one with all physical and
spectral properties identical to the literature data.9 Fraction
3 contained 9.5 mg (35%) of the starting enone. Fraction 4
contained 84 mg (90%) of the host.
1
102±1058C. H NMR (CDCl3): d 7.65±7.1 (m, 30H), 6.25
(d, Jˆ7 Hz, 1H), 2.85±2.55 (m, 3H), 2.5±2.25 (m, 2H), 2.15
(s, 2H), 1.75±1.5 (m, 4H), 1.5±1.3 (m, 4H). The structure of
the inclusion complex was determined by X-ray analysis.
1:1 Complex of benzopinacol and 4-p-cyanophenyl-4-
phenylcyclohexenone (6I). Benzopinacol, 0.040 g
(0.10 mmol) and 0.030 g (0.10 mmol) of the enone were
dissolved in 30 ml of 1:1 ether:hexane mixture and left
covered with `Para®lm' with several small apertures.
After 2 days crystals of a 1:1 complex were obtained
1
quantitatively; mp 135±1608C. H NMR (CDCl3): d 7.7±
7.15 (m, 30H), 6.25 (d, Jˆ7 Hz, 1H), 3.05 (s, 2H), 2.85±2.6
(m, 2H), 2.5±2.35 (m, 2H). The structure of the inclusion
complex was determined by X-ray analysis.
Irradiation of 70 mg of the complex (3Ib) under similar
conditions for 5 min afforded a 1.2:1 mixture of trans-5-
phenyl-6-p-cyanophenylbicyclo[3.1.0]hex-2-one and trans-
5-p-cyanophenyl-6-phenylbicyclo[3.1.0]hex-2-one (conver-
sion 16%, analyzed by 1H NMR).
Exploratory solid-state photolysis of the 1:1 complex of
`octaphenyltetraol' (4) and 4-cyanophenyl-4-phenyl-
cyclohex-2-en-1-one. Crystals of the 1:1 complex (4I)
(120 mg) were photolyzed for 5 h. NMR analysis of the
slightly yellow solid reaction mixture indicated formation
of trans-5-phenyl-6-p-cyano-phenylbicyclo[3.1.0]hex-2-one
as the only product at a conversion of 26%. The product
was isolated by column chromatography on silica gel using
10% ether in hexane as eluent. Fraction 1 contained 6.1 mg
(22%) trans-5-phenyl-6-p-cyano-phenylbicyclo[3.1.0]hex-
2-one with all physical and spectral properties identical to
the literature data.9 Fraction 2 contained 19 mg (70%) of the
starting material. Fraction 3 contained 88 mg (95%) of the
host.
Irradiation of 70 mg of the complex (3Ib) for 3 min afforded
1.2:1 mixture of trans-5-phenyl-6-p-cyanophenylbicyclo-
[3.1.0]hex-2-one and trans-5-p-cyanophenyl-6-phenylbi-
cyclo-[3.1.0]hex-2-one (conversion 3 or 7% in two separate
runs, analyzed by 1H NMR).
Irradiation of 70 mg of the complex (3Ib) for 25 min
afforded 2:1 mixture of trans-5-phenyl-6-p-cyanophenyl-
bicyclo[3.1.0]hex-2-one and trans-5-p-cyanophenyl-6-
phenylbicyclo-[3.1.0]hex-2-one
(conversion
.77%,
analyzed by 1H NMR). Small amount of unindenti®ed
products, presumably the result of secondary photochemical
products, were also detected by NMR at this conversion.
Irradiation under similar conditions of 70 mg of the
complex (4I) for 8 h afforded only the trans-5-phenyl-6-p-
cyanophenylbicyclo[3.1.0]hex-2-one (conversion 44%,
1
analyzed by H NMR).
Photolysis of the inclusion complex (5I). In typical runs,
0.10 g (0.10 mmol) of the crystalline inclusion compound
(5I) was irradiated between glass slides or in a Pyrex tube
(3 mm in diameter) for different periods of time, ranging
from 5 to 200 min at 15±188C. The ratios of the photo-
products obtained by NMR analysis are given in the Table 6.
Irradiation of 70 mg of the complex (4I) for 14 h afforded a
4:1 mixture of trans-5-phenyl-6-p-cyanophenylbicyclo-
[3.1.0]hex-2-one and trans-5-p-cyano-phenyl-6-phenyl-
1
bicyclo[3.1.0]hex-2-one (conversion 92%, analyzed by H
NMR).
The NMR data of the three photoproducts corresponded to
the known isomeric bicyclic photoketones from photolysis
of 4-p-cyano-4-phenylcyclohexenone in solution9 without
extraneous peaks from by-products.
Exploratory solid-state photolysis of 1:1 complex of
`Toda±Seebach' host (3) and 4-cyanophenyl-4-phenyl-
cyclohex-2-en-1-one. Irradiation of 50 mg of the complex
(3Ia) under similar conditions for 6 h afforded trans-5-
phenyl-6-p-cyanophenylbicyclo[3.1.0]hex-2-one (conver-
Photolysis of the benzopinacol inclusion complex (6I).
Crystals of the inclusion complex (6I) were irradiated
using portions of 0.050±0.070 g. In the NMR spectra the
singlet of the benzopinacol hydroxyl groups overlaps with
1
sion 6%, analyzed by H NMR).
Irradiation of 50 mg of the complex (3Ia) under similar
conditions for 12 h afforded trans-5-phenyl-6-p-cyano-
phenylbicyclo[3.1.0]hex-2-one (conversion 19%, analyzed
1
Table 6.
by H NMR).
Time
(min)
Conversion
(%)
trans±Ph/
trans-p-CN±Ph
trans±Ph/
cis±Ph
Irradiation of 50 mg of the complex (3Ia) under similar
conditions for 19 h afforded 3:1 mixture of trans-5-phenyl-
6-p-cyanophenylbicyclo[3.1.0]hex-2-one and trans-5-p-
cyanophenyl-6-phenylbicyclo[3.1.0]hex-2-one (conversion
75
135
200
,10
16
26
5/1
3/1
3/2
13/2
6/1
3/1
1
47%, analyzed by H NMR).

6830
H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
the doublets of the bridgehead protons in the photoproducts;
thus, the NMR analyses were carried out in the presence of a
0.05 ml of deuterated water for the proton exchange. The
exclusive formation of trans-5-p-cyano-phenyl-6-phenyl-
bicyclo[3.1.0]hexan-2-one has been observed at all conver-
sions.
A sequence of steps necessary to study theoretical
aspects of solid-state photochemistry
1. Solving X-ray structure.
2. Using `Smartpack' or `Icpack' (for inclusion com-
pounds) to generate the minicrystal lattice.
3. Superposition of the `ab initio optimized' intermediates
on central molecule of the minicrystal lattice.
Acknowledgements
4. Choosing optimized (inner) and frozen (outer) shells.
5. Preoptimization of the inner shell (reaction cavity) and
the intermediate by MM3 inside the frozen outer shell.
6. Cutting the preoptimized reaction cavity (the inner shell)
and the intermediate out of the large minicrystal lattice.
7. Optimization of the triplet intermediates (high level,
ROHF) inside the frozen preoptimized reaction cavity
(low level, Dreiding) using two-level ONIOM calcu-
lations as incorporated in Gaussian98.
Support of this research by the National Science Foundation
is gratefully acknowledged with special appreciation for its
support of basic research. Also, we wish to acknowledge
helpful discussions with, and advice from, Dr Douglas
Powell regarding X-ray diffractometry. We are also pleased
to acknowledge Departmental NSF grants CHE-9310428
and CHE-9709005 supporting X-ray facilities and NSF
CHE-9208463, NIH 1 510 RR0 8391-01 supporting NMR.
8. Single point calculations substituting Dreiding energies
to MM3 and ROHF(3-21G) energies to ROHF(6-31Gp).
9. Recalculating the ONIOM energies.
Appendix A. Supplementary Material
Analysis results without pre-optimization of the inner
shell lattice (no preoptimization of the crystal cavity)
References
1. This is Paper 188 of our photochemical series and Paper 254 of
our general series.
See Tables A1 and A2.
2. For Paper 252 see Zimmerman, H. E.; Ignatchenko, A. J. Org.
Chem. 1999, 64, 6635±6645. For Paper 253 see below, Ref. 10.
3. (a) Zimmerman, H. E.; Zuraw, M. J. J. Am. Chem. Soc. 1989,
111, 2358±2361. (b) Zimmerman, H. E.; Zuraw, M. J. J. Am.
Chem. Soc. 1989, 111, 7974±7989.
Table A1. MM3 optimizations of the triplet intermediates inside rigid
crystal lattices
Inclusion compound
DE (Ph±CN±Ph), kcal mol21
Seebach±Toda 1:1 (3Ia)
Seebach±Toda 2:1 (3Ib)
Cyclohexyldiol (5I)
212.9
158.2
7.3
4. (a) Zimmerman, H. E.; Zhu, Z. J. Am. Chem. Soc. 1994, 116,
9757±9758. (b) Zimmerman, H. E.; Zhu, Z. J. Am. Chem. Soc.
1995, 117, 5245±5262.
Benzopinacol (6I)
19.3
5. Zimmerman, H. E.; Sebek, P. J. Am. Chem. Soc. 1997, 119,
3677±3690.
6. Zimmerman, H. E.; Sebek, P.; Zhu, Z. J. Am. Chem. Soc. 1998,
120, 8549±8550.
7. With regard to this approach, one referee's comment was that
Ê
lattice constraints, possibly 50 A distant, may play a role not
included in the inert gas model.
Table A2. ONIOM optimizations of the triplet intermediates inside the
rigid crystal lattice (no preoptimization of the crystal cavity)
Inclusion
compound
DE (Ph±CN±Ph),
Method
kcal mol21
8. (a) Maseras, F.; Morokuma, K. J. Comp. Chem. 1995 16, 1170±
1179. (b) Matsubara, T.; Sieber. S.; Morokuma, K. Int. J. Quantum
Chem., 1996, 60, 1101±1109. (c) Frisch, M. J.; Trucks, G. W.;
Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.;
Zakrzewski, V. G.; Montgomery, Jr., J. A.; Stratmann, R. E.;
Burant, J. C.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain,
M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.;
Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, C.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q. K.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko,
A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox,
D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.6, Gaussian,
Inc., Pittsburgh PA, 1998. (d) The literature has a number of
QM/MM programs described but mainly dealing with solvation
and similar aspects.
Toda±Seebach 1:1
Toda±Seebach 2:1
Cyclohexyldiol
Benzopinacol
25.7
22.9
23.9
1.5
21.4
14.7
8.2
4.1
5.8
7.0
28.6
16.5
12.8
7.7
14.7
22.5
37.5
33.9
7.8
ONIOM/Dreiding
ONIOM/MM3
MM3
ROHF/6-31Gp, biradical
ONIOM(6-31Gp)/MM3
ONIOM/Dreiding
ONIOM/MM3
MM3
ROHF/6-31Gp, biradical
ONIOM(6-31Gp)/MM3
ONIOM/Dreiding
ONIOM/MM3
MM3
ROHF/6-31Gp, biradical
ONIOM(6-31Gp)/MM3
ONIOM/Dreiding
ONIOM/MM3
MM3
ROHF/6-31Gp, biradical
ONIOM(6-31Gp)/MM3
35.8

H. E. Zimmerman et al. / Tetrahedron 56 (2000) 6821±6831
6831
9. Zimmerman, H. E.; Rieke, R. D.; Scheffer, J. R. J. Am. Chem.
Soc. 1967, 89, 2033±2047.
J. N., Eds.; Advances in Photochemistry; Interscience: New York,
1963, Vol. 1, pp 183±208.
10. Zimmerman, H. E.; Alabugin, I. V.; Chen, W.; Zhu, Z. J. Am.
Chem. Soc. 1999, 121, 11930±11931.
18. Zimmerman, H. E.; Lewin, N. J. Am. Chem. Soc. 1969, 91,
879±886.
11. (a) Beck, A. K.; Bastani, B.; Plattner, D. A.; Petter, W.;
Seebach, D.; Braunschweiger, H.; Gysi, P.; Vecchia, L. L. Chimia
1991, 45, 238±244. (b) Seebach, D.; Beck, A. K.; Imwinkelried,
R.; Roggo, S.; Wonnacott, A. Helv. Chim. Acta. 1987, 70, 954±
974.(c) Toda, F.; Tanaka, K. Tetrahedron Lett. 1988, 29, 551±554.
12. Chiral hosts have been used for resolution. See: (a) See Ref.
11b. (b) Toda, F. Acc. Chem. Res. 1995, 28, 1059±1062;
(c) Korkas, P. P.; Weber, E.; Czugler, M.; Naray-Szabo, G.
J. Chem. Soc. Chem. Commun. 1995, 21, 2229±2230.
19. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liscamp, R.;
Lipton, M.; Cau®eld, C.; Chang, G.; Hendrickson, T.; Still, W. C.
MacroModel V6.5. J. Comput. Chem. 1990, 11, 440.
20. (a) McBride, J. M. Acc. Chem. Res. 1983, 16, 304±312;
(b) Kearsley, S. K.; McBride, J. M. Mol. Liq. Cryst. Inc. Nonlin.
Opt. 1988, 156 109±122.
21. (a) Gavezzotti, A.; Simonetta, M. Chem. Rev. 1982, 82, 1±13.
(b) Peachey, N. M.; Eckhardt, C. J. J. Phys. Chem. 1994, 98, 7106±
7115; (c) Leibovitch, M.; Olovsson, G.; Sundarababu, G.;
Ramamurthy, V.; Scheffer, J. R.; Trotter, J. J. Am. Chem. Soc.
1996, 118, 1219±1220; (d) Ariel. S.; Askarai, S.; Scheffer, J. R.;
Trotter, J.; Walsh, L. J. Am. Chem. Soc. 1984, 106 5726±5728.
(e) Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433±
481; (f) Ito, Y. Synthesis, 1998, 1±32. (g) Shin, S. H.; Huang, F. K.;
Garcia-Garibay, M. Tetrahedron Lett. 1999, 261±264. (h) Keating,
A. E.; Shin, S. H.; Houk, K.; Garcia-Garibay, M. J. Am. Chem. Soc.
1997, 119, 1474±1475; (i) Leibovitch, M.; Olovsson, G.; Scheffer,
J. R.; Trotter, J. J. Am. Chem. Soc. 1998, 120, 12550±12551;
(j) Singh, N. B.; Singh, R. J.; Singh, N. P. Tetrahedron 1994, 50,
6441±6493.
13. We thank Dr Douglas J. Fox of Gaussian, Inc. for informal
advice on the use of Oniom.
14. Mayo, S. L.; Olafson, B. D.; Goddard, W. A. J. Phys. Chem.
1990, 94, 8897±8909.
15. Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.;
Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024±10035.
16. Allinger, N. L. MM3, University of Georgia, December 1990.
17. (a) This convenient notation was introduced along with theory
correlating photochemical reactivity to excited state electronic
structure. The notation permits thinking in three dimensions
while writing in two. (b) Zimmerman, H. E. Abstracts of the Seven-
teenth National Organic Symposium of the American Chemical
Society, Bloomington, Indiana, 1961; pp 31±41. (b) Zimmerman,
H. E.; Schuster, D. I. J. Am. Chem. Soc. 1962, 84, 4527±4540.
(c) Zimmerman, H. E.; Noyes, Jr., Hammond, A G. S., Pitts, Jr.,
22. Sheldrick. G. M. Shelxtl Version 5 Reference Manual.
Siemens Analytical X-ray Instruments, 6300 Enterprise Dr.,
Madison, WI 53719-1173, USA, 1994.