Geometric requirements in the ferriin oxidation of benzylic 1,2-diols

Article abstract of DOI:10.1039/a905728e

The rates of ferriin [i.e. tris(1,10-phenanthroline)iron(III)] oxidation of cis- and trans-1,2-diphenylcyclohexane-1,2-diol have been found to be dramatically different; cis-1,2-diphenylcyclohexane-1,2-diol reacts a minimum of 10⁴ times faster than the corresponding trans-isomer; implications for the oxidation of benzylic diols by ferriin are discussed.

Full text of DOI:10.1039/a905728e

Geometric requirements in the ferriin oxidation of benzylic 1,2-diols
John H. Penn,* Robert C. Plants and An Liu
Department of Chemistry, West Virginia University, Morgantown, WV 26506-6045, USA. E-mail: jpenn2@wvu.edu
Received (in Corvallis, OR, USA) 12th July 1999, Accepted 8th October 1999
The rates of ferriin [i.e. tris(1,10-phenanthroline)iron(iii)]
oxidation of cis- and trans-1,2-diphenylcyclohexane-1,2-diol
have been found to be dramatically different; cis-1,2-di-
phenylcyclohexane-1,2-diol reacts a minimum of 104 times
faster than the corresponding trans-isomer; implications for
the oxidation of benzylic diols by ferriin are discussed.
behavior during a kinetics experiment on appropriately substi-
tuted compounds. This has served as an indication that the
reaction mechanism is more complex than originally
thought.6
As part of our reinvestigation of this reaction mechanism, we
have studied the geometric relationship of the hydroxy groups
in this oxidation reaction. Our preliminary investigation of the
reactivity differences exhibited by dl- and meso-1,2-diphenyl-
ethane-1,2-diol indicated to us that steric requirements existed
in this reaction and that the geometry of the reacting groups
might be important. The free rotation of the central C–C bond in
the dl- and meso-isomers of 1,2-diphenylethane-1,2-diol does
not allow for a definitive analysis of the various intramolecular
interactions, which may be responsible for the observed
reactivity. In order to lock the geometry of all components in
place, we chose to study cis- and trans-1,2-diphenylcyclohex-
ane-1,2-diol. These compounds were synthesized based on
Tomboulian’s method.7 In brief, cis-3 was obtained from
1,4-dibenzoylbutane in a McMurry coupling reaction with
TiCl3/LiAlH4 followed by OsO4 oxidation to give the cis-
1,2-diphenylcyclohexane-1,2-diol in 27% yield [mp
86.5–87.5 °C (lit.,7 86–87 °C)]. trans-1,2-Diphenylcyclohex-
ane-1,2-diol was obtained from reaction of cyclohexane-
1,2-dione with excess (3 equiv.) of PhLi at reflux in dry THF for
20 h, followed by appropriate work-up to yield the resultant
white needle-like crystals, mp 122–123 °C (lit.,7
121–122 °C).
The oxidative cleavage of 1,2-diols by a variety of reagents has
been extensively studied.1 The mechanisms are conventionally
classified into several types.1 In the first class of oxidative
cleavage of vic-diols [eqn. (1)], a bidentate complex between
Quantitative analysis of the reaction of ferriin (10 mM) with
cis- and trans-3 (5 mM) in MeCN at room temperature was
accomplished by reverse-phase HPLC on a C-18 column eluted
with MeCN–H2O, using methyl benzoate as an internal
standard. A base (2,6-di-tert-butylpyridine) was added to avoid
the possible carbocation-induced rearrangement product (i.e.
the pinacol rearrangement). cis-3 reacted quantitatively (mass
balance = 98.9% at 34% conversion) to form the anticipated
bond cleavage product (i.e. 1,4-dibenzoylbutane). The error
limits for this determination are ±5%, which means that a small
amount of the trans-1,2-diphenylcyclohexane-1,2-diol could
have been formed in this reaction. The data were carefully
analyzed for the appearance of the trans-1,2-diphenyl-1,2-
cyclohexanediol. Since no trans-diol was observed and the
appearance of 1,4-dibenzoylbutane was near quantitative, we
conclude that the major reaction pathway (by a factor of
> 20+1) is oxidative cleavage of the 1,2-diol C–C bond.
In contrast, the trans-1,2-diphenylcyclohexane-1,2-diol did
not react within a time period of two days. In fact, we were
unable to see a reaction under any conditions employed by us.
Therefore, we were unable to confirm that the expected
oxidation product arose from this reaction. Under controlled
conditions where both cis-3 and trans-3 were reacted in the
same solution, we were able to show that a minimum reactivity
difference of 104 exists between these isomers. Thus, there is a
significant reactivity difference between cis-3 and trans-3 for
ferriin oxidation.
the oxidant and the glycol breaks down by an apparent six-
electron process (two electrons to the oxidant).1 The second
general class of oxidative cleavage mechanisms of vic-diols
[eqn. (2)] invokes a monodentate complex, which yields an
intermediate radical via a one-electron process, followed by
C–C fission and then further oxidation of the resulting radical.1
A third general class of oxidative cleavage reactions for
appropriately substituted vic-diols [eqn. (3)] is a single electron
transfer to form a radical cation, that can subsequently undergo
bond cleavage to form fragments similar to those formed in
[eqn. (2)].2,3
In our laboratories, we hoped to use vic-diols that undergo
cleavage by the general mechanism shown in [eqn. (3)] in order
to determine fundamental reaction rate constants. Our initial
studies demonstrated that ferriin readily oxidized tetraaryl-
ethane-1,2-diols to the corresponding ketones [eqn. (4)].4 By
applying a mechanism in which a slow rate-determining outer-
sphere electron transfer was followed by rapid bond cleavage
{general mechanism of [eqn. (3)]}, we reported endergonic
electron transfer rate constants that were faster than expected,
based upon estimates of the solution phase oxidation potential.5
As we have varied functionality in standard structure–reactivity
probes in an attempt to understand this anomalously fast
electron transfer rate, we have observed non-isosbestic point
This extreme reactivity difference between the two isomers
required an analysis of the possible conformations of cis- and
trans-3. These conformations are shown in Fig. 1. For cis-3,
both available ring-flipped conformations have one phenyl
group in the axial conformation and one phenyl group in the
equatorial conformation. Thus, the energetics of these two
Chem. Commun., 1999, 2359–2360
This journal is © The Royal Society of Chemistry 1999
2359

by two experimental results. First, there is no trace of trans-
1,2-diphenylcyclohexane-1,2-diol in the reactions starting from
the cis-1,2-diphenylcyclohexane-1,2-diol isomer. This argues
against equilibration of cis-3•+ and trans-3•+. Further, prelimi-
nary experimental results in which cis-3•+ and trans-3•+ were
generated by single electron transfer to excited state dicyanoan-
thracene showed similar non-reactivity for both the cis- and
trans-isomers.
The reactivity differences between cis- and trans-3 seem to
preclude reaction through a general mechanism such as that
shown by the general cleavage mechanism of eqn. (2). A
mechanistic interpretation that there is a sufficient energetic
difference for complexation of an equatorial OH group as
compared to an axial OH group is, in principle, possible.
Combined with evidence that the monomethyl ethers of both
meso- and dl-1,2-diphenylethane-1,2-diol are unreactive in
comparison10 allows us to definitively rule out the general
mechanism of eqn. (2), since two OH groups are required for the
reaction.
To the best of our knowledge, such geometric requirements
for the two hydroxy groups are unprecedented in the cleavage of
vic-diols by one-electron oxidizing reagents [general oxidative
cleavage mechanism shown in eqn. (2)]. We are continuing our
efforts to fully understand the mechanism of this reaction in
order to better understand the oxidation reaction mechanisms of
one electron oxidants in reactions such as the Belousov–
Zhabotinski (B–Z) reaction.11,12
The geometric requirements for the ferriin oxidation of
aromatic 1,2-diols involve the orientation of the two OH groups
in a syn-position or an anti-periplanar phenyl–OH arrangement.
The differing reactivity of 3a and 3b with ferriin indicates a very
unexpected and unusual non-outer sphere oxidation mechanism
for this type of compound.
Fig. 1 Conformations of cis- and trans-1,2-diphenylcyclohexane-1,2-diol.
alternative conformations are equivalent. In contrast, trans-3
does not have equivalent conformational energies. Since the
phenyl group is larger than the hydroxy group,8 one would
anticipate that the conformation in which the phenyl groups are
equatorial would be quite a bit more stable than the conforma-
tion in which the hydroxy groups are in the axial position. This
conformational preference tends to ‘lock’ the cyclohexanediol
into the conformation with both phenyl groups in the equatorial
position in the same way that the tert-butyl group is said to
‘lock’ the conformation of substituted cyclohexanes. In this
conformation, the hydroxy groups are in an anti-relationship to
each other. This expectation is consistent with molecular
mechanics calculations in which the energetic difference
between the conformation in which the phenyl groups are axial
and the conformation in which the phenyl groups are equatorial
is 5.4 kcal mol21. From this relative arrangement of groups, we
conclude that either syn-OH groups or an antiperiplanar phenyl–
OH arrangement can lead to the observed bond cleavage
reactions.
The results of this study have important implications for the
ferriin oxidation of vic-diols. A simple outer-sphere electron
transfer reaction, in which an electron is transferred from the
electron donor to the ferriin molecule, cannot explain the
observed reactivity differences. One would anticipate that the
oxidation potentials of cis-3 and trans-3 would be identical,
since cis- and trans-1,2-diphenylcyclohexane have identical
photoelectron spectra (identical ionization potentials).9 Our
estimate of a minimum rate differential for these two com-
pounds of 104 would be inconsistent with the energetic
equivalence of these two reactions.
We gratefully acknowledge the financial support of the West
Virginia University Department of Chemistry.
Notes and references
1 R. A. Sheldon and J. K. Kochi, Metal-Catalyzed Oxidations of Organic
Compounds, Academic Press, New York, 1981.
2 L. W. Reichel, G. W. Griffin and A. J. Muller, Can. J. Chem., 1984, 62,
424.
3 H. F. Davis, P. K. Das, L. W. Reichel and G. W. Griffin, J. Am. Chem.
Soc., 1984, 106, 6968.
4 J. H. Penn, D.-L. Deng and K.-J. Chai, Tetrahedron Lett., 1988, 29,
3635.
5 H. J. Penn, Z. Lin and D.-L. Deng, J. Am. Chem. Soc., 1991, 113,
1001.
6 J. H. Penn, A. Liu, S. Svarovsky and R. H. Simoyi, submitted.
7 P. J. Tomboulian, J. Org. Chem., 1961, 26, 2652.
8 E. L. Eliel and M. Manoharan, J. Org. Chem., 1981, 46, 1959.
9 S. Ruppel, PhD Thesis, Universitaet zu Koeln, 1995.
10 J. H. Penn, A. Liu, S. Svarovsky and R. H. Simoyi, manuscript in
preparation.
11 Y.-C. Chou, H.-P. Lin, S. S. Sun and J. J. Jwo, J. Phys. Chem., 1993, 97,
8450.
12 J. Ungvarai, Z. Nagy-Ungvarai, J. Enderleihn and S. C. Mueller,
J. Chem. Soc., Faraday Trans., 1997, 93, 69.
Alternatively, one could argue that electron transfer occurs in
both species, but that trans-3•+ is a stable species and does not
lead ultimately to the dione, while cis-3•+ is relatively unstable
and reacts rapidly to form the dione. This argument is ruled out
Communication 9/05728E
2360
Chem. Commun., 1999, 2359–2360