Photophysics and photodeprotection reactions of p -methoxyphenacyl phototriggers: An ultrafast and nanosecond time-resolved spectroscopic and density functional theory study

Article abstract of DOI:10.1021/jo100848b

Time-resolved spectroscopic experiments were performed to investigate the kinetics and mechanisms of the photodeprotection reactions for p-methoxyphenacyl (pMP) compounds, p-methoxyphenacyl diethyl phosphate (MPEP) and diphenyl phosphate (MPPP). The experimental results reveal that compared to the previous reports for the counterpart p-hydroxyphenacyl (pHP) phosphates, the ³nπ/ππ* mixed character triplet of pMP acts as a reactive precursor that leads to the subsequent solvent and leaving group dependent chemical reactions and further affects the formation of photoproducts. The MPPP triplet in H₂O/CH₃CN and in fluorinated alcohols shows a rapid heterolytic cleavage (τ ≈ 5.4 ns) that results in deprotection and formation of a solvolytic rearrangement product, whereas the MPPP triplet in CH₃CN and the MPEP triplet in CH₃CN and H₂O/CH₃CN and fluorinated alcohols decay on a much longer time scale (τ ≈ 100 ns) with little observation of the rearrangement product. The density functional theory (DFT) calculations reveal a substantial solvation effect that is connected with the methoxy versus hydroxyl substitution in accounting for the different deprotection reactivity of pMP and pHP compounds. The results reported here provide new insight in elucidating the solvent and leaving group dependent dual reactivity of pMP compounds on the formation of the rearrangement versus reductive photoproduct.

Full text of DOI:10.1021/jo100848b

pubs.acs.org/joc
Photophysics and Photodeprotection Reactions of p-Methoxyphenacyl
Phototriggers: An Ultrafast and Nanosecond Time-Resolved
Spectroscopic and Density Functional Theory Study
Hui-Ying An,† Wai Ming Kwok,‡ Chensheng Ma,† Xiangguo Guan,† Jovi Tze Wai Kan,†
Patrick H. Toy,† and David Lee Phillips*,†
†Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong S. A. R., P. R. China,
and ‡Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University,
Hung Hom, Kowloon, Hong Kong S. A. R., P. R. China
phillips@hkucc.hku.hk
Received April 30, 2010
Time-resolved spectroscopic experiments were performed to investigate the kinetics and mechanisms of the
photodeprotection reactions for p-methoxyphenacyl (pMP) compounds, p-methoxyphenacyl diethyl
phosphate (MPEP) and diphenyl phosphate (MPPP). The experimental results reveal that compared to
3
the previous reports for the counterpart p-hydroxyphenacyl (pHP) phosphates, the nπ*/ππ* mixed
character triplet of pMP acts as a reactive precursor that leads to the subsequent solvent and leaving group
dependent chemical reactions and further affects the formation of photoproducts. The MPPP triplet in
H2O/CH3CN and in fluorinated alcohols shows a rapid heterolytic cleavage (τ ≈ 5.4 ns) that results in
deprotection and formation of a solvolytic rearrangement product, whereas the MPPP triplet in CH3CN
and the MPEP triplet in CH3CN and H2O/CH3CN and fluorinated alcohols decay on a much longer time
scale (τ ≈ 100 ns) with little observation of the rearrangement product. The density functional theory
(DFT) calculations reveal a substantial solvation effect that is connected with the methoxy versus hydroxyl
substitution in accounting for the different deprotection reactivity of pMP and pHP compounds. The
results reported here provide new insight in elucidating the solvent and leaving group dependent dual
reactivity of pMP compounds on the formation of the rearrangement versus reductive photoproduct.
Introduction
due to their desirable ability in providing a very fast photo-
deprotection and the accompanying formation of biologically
benign rearrangement byproducts.1a,2 As a result, pHP com-
pounds have been developed as an efficient “cage” for the
Several kinds of R-keto groups, especially those based on
phenacyl groups, have received a great deal of attention due to
their favorable properties for potential use as photochemically
removable protecting groups (PRPGs) for the rapid release of
various biological effectors.1 Among the wide variety of phena-
cyl based PRPGs that have been examined in the literature,
para-substituted phenacyls bearing the oxygen-containing elec-
tron donors of the p-hydroxyphenacyl (pHP) and p-methoxy-
phenacyl (pMP) groups were found to be particularly attractive
(2) (a) Anderson, J. C.; Reese, C. B. Tetrahedron Lett. 1962, 3, 1–4. (b)
Epstein, W. W.; Garrossian, M. J. Chem. Soc., Chem. Commun. 1987, 532–
533. (c) Baldwin, J. E.; Mcconnaughie, A. W.; Moloney, M. G.; Pratt, A. J.;
Shim, S. B. Tetrahedron 1990, 46, 6879–6884. (d) Park, C. H.; Givens, R. S. J.
Am. Chem. Soc. 1997, 119, 2453–2463. (e) Banerjee, A.; Lee, K.; Falvey, D. E.
Tetrahedron 1999, 55, 12699–12710. (f) Zhang, K.; Corrie, J. E. T.; Muna-
singhe, V. R. N.; Wan, P. J. Am. Chem. Soc. 1999, 121, 5625–5632. (g)
Conrad, P. G.; Givens, R. S.; Hellrung, B.; Rajesh, C. S.; Ramseier, M.; Wirz,
J. J. Am. Chem. Soc. 2000, 122, 9346–9347. (h) Givens, R. S.; Weber, J. F. W.;
Conrad, P. G.; Orosz, G.; Donahue, S. L.; Thayer, S. A. J. Am. Chem. Soc.
2000, 122, 2687–2697. (i) Specht, A.; Loudwig, S.; Peng, L.; Goeldner, M.
Tetrahedron Lett. 2002, 43, 8947–8950.
(1) (a) Givens, R. S.; Park, C. H. Tetrahedron Lett. 1996, 37, 6259–6262.
(b) Givens, R. S.; Jung, A.; Park, C. H.; Weber, J.; Bartlett, W. J. Am. Chem.
Soc. 1997, 119, 8369–8370.
DOI: 10.1021/jo100848b
r
Published on Web 08/04/2010
J. Org. Chem. 2010, 75, 5837–5851 5837
2010 American Chemical Society

An et al.
JOCArticle
liberation of a number of biological stimulants.1,2d,h,3 Previous
investigations on pHP and pMP compounds observed that the
photochemistry of these two PRPGs is highly sensitive to sol-
vent properties. The photodeprotection reaction of pHP caged
compounds (eqs 1 and 2) occurs as the primary process in aque-
ous or largely aqueous solvents but does not take place noti-
ceably in neat organic solvents such as acetonitrile (CH3-
CN).1,2d,f-h,3b,4 Our previous mechanistic studies utilizing
direct time-resolved spectroscopic methods and indirect steady
state quenching methods have presented explicit and consistent
evidence indicating a direct triplet cleavage pathway for pHP
photodeprotection.5 Ultrafast experiments on a series of pHP
caged phosphates and carboxylates (eq 2) in a H2O/CH3CN
(1:1, v/v) solvent have found a stepwise water-assisted triplet
deprotection-rearrangement pathway with the deprotection rate
depending on the lability of the leaving group and ranging from
∼7 ꢀ 109 s-1 for HPPP (X = OPO(OPh)2) to ∼2 ꢀ 108 s-1 for
HPH (X = OCO(CH2)5CH3).5
reduction (2) and rearrangement (3) byproducts. The reduction
accompanied pMP photorelease parallels the photodeprotec-
tion reaction reported for nonsubstituted phenacyl esters and
electron-withdrawing substituted p-phenacyl chlorides.2a
A
laser flash photolysis (LFP) study by Falvey and co-workers
found that such reactions of the nonsubstituted phenacyl
PRPG arises from a very fast hydrogen abstraction (rate of
∼9 ꢀ 106 M-1 s-1) of the phenacyl triplet from the surrounding
solvent.7 The prevalence of the pMP reduction product (2) in
solvents with better hydrogen-donating capacity as well as
the relevant solvent isotope effects implies a primary role of
the analogous hydrogen abstraction and associated reaction
(denoted as “photoreduction” hereafter) in accounting for the
observed pMP photodeprotection in these solvents.
In contrast to the knowledge determined for the pMP photo-
reduction, little is known about the nature of the major photo-
chemical steps and key environmental factors accounting for the
more biologically important pMP rearrangement reaction. It is
relevant to note that a photochemistry study of MPEP shows its
quantum efficiency in methanol is about only half of that of its
pHP counterpart HPEP;6a in addition, pMP acetate was found
to have low reactivity in a H2O/CH3CN (1:1, v/v) solvent, in
sharp contrast to the efficient photorelease reaction of the corres-
ponding HPA observed under the same experimental condi-
tions.1,2d,f-h,3b,4 As the pHP and pMP cages apparently have
similar configurations, the substantially different photochemis-
try and reactivity of the corresponding phototriggers may
suggest a crucial effect associated with replacing the hydroxy
group in pHP by the methoxy group in pMP. Up to now, the
relevant photophysics and photochemistry of pMP have re-
mained little understood, and the underlying cause of this substi-
tution effect has not been investigated in great detail to the best
of our knowledge. Such information is, however, important in
terms of understanding the mechanism of pMP deprotection as
well as helping to clarify some uncertainty in the pHP rearrange-
ment reaction. The pHP versus pMP substitution effect and the
complex solvent and leaving group sensitive pMP reactivity
toward the deprotection and the competitive hydrogen absorp-
tion is an interesting example of the widely known substituent
and solvent dependent photoreactivity of the general family of
aromatic carbonyl compounds. In this sense, pMP compounds
offer a good exemplar to explore this ubiquitous but not fully
understood substituent and solvent dependent photoreactivity
phenomenon in phenacyl type aromatic carbonyl compounds.
As an extension of our previous work on various pHP photo-
triggers,5 we have conducted a combined femtosecond time-
resolved fluorescence (fs-TRF), femtosecond transient absorp-
tion (fs-TA), and picosecond and nanosecond time-resolved
Raman (ps-, ns-TR3) study on two synthesized pMP com-
pounds, MPEP and MPPP, in a range of solvents including
neat CH3CN, H2O/CH3CN (1:1, v/v), CF3CH2OH, and CF3-
CHOHCF3 with the aim of helping to address the issues men-
tioned in the preceding paragraph. DFT calculations were
performed to help map the reaction pathways and locate tran-
sient states and reaction barriers for the pMP deprotection reac-
tion in water-containing solvents. These experimental and com-
putational results taken together suggest a mechanism where
solvation of the leaving group initiates a triplet liberation and
rearrangement pathway for the pMP photodeprotection. The
data provide explicit dynamic information and a mechanistic
Unlike the relatively simple solvent dependence and clean-
ness of the product distribution displayed by the pHP photo-
triggers, photodeprotection of pMP caged compounds exhibits
a more complicated solvent dependence and results in a more
complex product distribution with a relatively low efficiency for
deprotection.2a-d,f,4,6 For example, it was reported that the
photolysis of pMP dihydrogen phosphate (1b), pMP diethyl
phosphate (MPEP, 1c), and pMP diphenyl phosphate (MPPP,
1d) in the hydrogen-donating solvent dioxane leads to depro-
tection along with the reduction product p-methoxyacetophe-
none (2) as the sole ketone byproduct.2b,c Different from this,
photodeprotection of MPEP (1c)6a,b and pMP chloride (1a)2a
in the less hydrogen-donating solvents of methanol, ethanol, or
tert-butyl alcohol produces the corresponding rearranged es-
ters p-methoxyphenylacetates (3) as the major byproduct and a
minor amount of 2. These observations suggest a solvent
dependent dual reactivity of photoexcited pMP in giving two
competing deprotection pathways that lead to, respectively, the
(3) (a) Gee, K. R.; Kueper, L. W.; Barnes, J.; Dudley, G.; Givens, R. S. J.
Org. Chem. 1996, 61, 1228–1233. (b) Conrad, P. G.; Givens, R. S.; Weber,
J. F. W.; Kandler, K. Org. Lett. 2000, 2, 1545–1547.
(4) Brousmiche, D. W.; Wan, P. J. Photochem. Photobiol., A 2000, 130,
113–118.
(5) (a) Ma, C. S.; Kwok, W. M.; Chan, W. S.; Zuo, P.; Kan, J. T. W.; Toy,
P. H.; Phillips, D. L. J. Am. Chem. Soc. 2005, 127, 1463–1472. (b) Ma, C. S.;
Kwok, W. M.; Chan, W. S.; Du, Y.; Kan, J. T. W.; Toy, P. H.; Phillips, D. L.
J. Am. Chem. Soc. 2006, 128, 2558–2570.
(6) (a) Givens, R. S.; Athey, P. S.; Matuszewski, B.; Kueper, L. W.; Xue,
J. Y.; Fister, T. J. Am. Chem. Soc. 1993, 115, 6001–6012. (b) Givens, R. S.;
Kueper, L. W. Chem. Rev. 1993, 93, 55–66. (c) Sheehan, J. C.; Umezawa, K.
J. Org. Chem. 1973, 38, 3771–3774.
(7) Banerjee, A.; Falvey, D. E. J. Am. Chem. Soc. 1998, 120, 2965–2966.
5838 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
FIGURE 2. Normalized fluorescence decay profiles of MPEP at
the labeled fluorescence wavelengths observed with the 267 nm
excitation in (a) CH3CN and (b) H2O/CH3CN (1:1, v/v) are shown.
The solid lines show the instrumental response function (IRF)
convolved two exponential fittings to the experimental data.
FIGURE 1. Fs-TRF spectra of MPEP obtained with 267 nm
excitation in (a) CH3CN and (b) H2O/CH3CN (1:1, v/v). The arrows
indicate the spectral evolution in the two solvents. The inset in panel
a shows a comparison of the normalized fluorescence spectra at 0.05
ps (λmax at 330-360 nm) and 1.2 ps ((λmax at 400-420 nm) in
CH3CN (blue) and H2O/CH3CN (purple).
fluorescence intensity varies significantly and exhibits an increas-
ingly slower decay dynamics as the fluorescence wavelength
increases. Figure 2 shows the time dependences of the fluores-
cence intensity at three selected wavelengths and the associated
exponential fitting results obtained from simulation of the
experimental data containing a convolution of the instrumental
response function (IRF). This fitting found that the wavelength
dependent decay dynamics in the two solvents can both be
described satisfactorily by similar sets of two time constants of
∼0.1 and ∼1.0 ps for those in CH3CN and ∼0.1 and ∼1.3 ps in
H2O/CH3CN. The traces at different wavelengths shown in
Figure 2 were fit to an instrumental response function (IRF)
convolved two exponential fittings to the experimental data. The
trace at 330 nm in Figure 2 has mainly a large contribution from
the shorter lived species responsible for the fluorescence with
description for each of the steps involved in the overall reaction
scheme of the pMP photodeprotection. In addition, since the
hydroxyl proton involved excited-state processes that may occur
in the pHP deprotection are unlikely to be included in the pMP
photodeprotection, a direct comparison of the dynamics and
reaction pathways obtained here for the pMP phosphates with
those reported for the counterpart pHP phosphates will provide
direct evidence for identifying the particular excited state step(s)
that associate(s) with the effect of methoxy versus hydroxyl sub-
stitution. This will in turn help to reveal the molecular origin of
the different deprotection photochemistry displayed by the pMP
and pHP phototriggers.
λ
max at ∼330 nm in CH3CN and ∼360 nm in H2O/CH3CN and
hence has a faster decay, while the trace at 450 nm in Figure 2
mainly has contributions from the longer lived species respon-
sible for the fluorescence band with the λmax at ∼420 and ∼400
nm in the respective solvents and thus has a comparatively
longer decay at this wavelength.
Results and Discussion
Femtosecond Time-Resolved Fluorescence (fs-TRF) Spec-
troscopy. Comparative fs-TRF measurements were done for
the compounds in CH3CN and H2O/CH3CN to characterize the
spectral properties, dynamics, and the effect of water on the
singlet character excited state(s) populated by the 267 nm photo-
excitation. Representative spectra recorded for MPEP in the two
solvents are displayed in Figure 1a and b, respectively. In both of
the solvents, the TRF spectra display a similar very rapid decay
from an initial relatively strong fluorescence (profiles of this
fluorescence have λmax at ∼330 nm in CH3CN and ∼360 nm in
H2O/CH3CN) into a rather weak and red-shifted fluorescence
band with the λmax at ∼420 and ∼400 nm in the respective sol-
vents followed by a quick decay of the latter on a few picose-
conds time scale. As a result of this, the time profile of the
It is known that, for typical substituted aromatic carbonyl
systems including the pMP chromophore (e.g., the spectra at
t = 0 in Figure 1S in the Supporting Information), the lon-
gest wavelength absorption is due to the combined contribu-
tions from a strongly allowed 1ππ* transition (La type)
featured by the absorption band with a λmax at ∼280 nm
and a weakly allowed 1nπ* transition that constitutes to the
tail of the absorption band and extends to ∼340 nm.5,6,8
(8) (a) Ma, C. S.; Du, Y.; Kwok, W. M.; Phillips, D. L. Chem.;Eur. J.
2007, 13, 2290–2305. (b) Ma, C. S.; Kwok, W. M.; An, H. Y.; Guan, X. G.;
Fu, M. Y.; Toy, P. H.; Phillips, D. L. Chem.;Eur. J. 2010, 16, 5102–5118.
J. Org. Chem. Vol. 75, No. 17, 2010 5839

An et al.
JOCArticle
1
The ππ* transition is mainly responsible for the 267 nm
character excited states involved in the photodeprotection of
the R-keto PRPGs.
photoabsorption, and it is straightforward that the observed
initial strong ∼330/360 nm transient fluorescence in the two
solvents (see Figure 1) is from this excited state. The rather
weak nature of the red-shifted (∼400/420 nm) fluorescence
seen at later times and especially the spectral blue shift of this
fluorescence from ∼420 nm in CH3CN to ∼400 nm in H2O/
CH3CN solutions are diagnostic for its origination from a
1nπ* character state. The relative blue shift of the electronic
spectrum in a solvent capable of hydrogen-bonding (HB)
interaction (such as the H2O/CH3CN solution used here)
relative to that in a solvent that has little hydrogen-bonding
ability (such as the neat CH3CN solvent here) is a known
Femtosecond Transient Absorption (fs-TA) Spectroscopy.
Figure 3 shows a comparison of the fs-TA results obtained
for MPEP in neat CH3CN and H2O/CH3CN mixed solvents
recorded at various time intervals ranging from 0 to 5000 ps
after the 267 nm photoexcitation. The TA spectra recorded
before and after a 10 ps time delay between the pump and
probe pulses are presented separately to better display the
spectral evolution at the respective time regimes. Parallel
results acquired for MPPP in the two solvents as well as that
in CF3CHOHCF3 are displayed in Figure 4. The TA spectra
recorded for MPPP in CF3CH2OH are very similar to those
observed in H2O/CH3CN, and these data are given in Figure
2S in the Supporting Information. Comparison of the spec-
tra obtained within 10 ps time delays for the two compounds
in the different solvents reveals a common feature of the
spectral evolution from an initial weak ∼325 nm absorption
(0.1 ps spectrum) into a much stronger but structure-less
absorption with a λmax at ∼410 nm (10 ps spectrum). This
spectral change is accompanied by an isosbestic point at
∼335 nm indicating the occurrence of a dynamic state-to-
state conversion. Examination of the temporal growth of the
∼410 nm absorption as seen in the insets in Figures 3 and 4
reveals that the kinetics of the spectral conversion can be
described by a single exponential function with a ∼2-3 ps
time constant. The spectral profiles of the 10 ps TA closely
resemble the triplet spectra reported for related para-sub-
stituted aromatic carbonyl compounds including the refer-
ence compound methoxyacetophenone (2) (see Figure 3S in
the Supporting Information) and pHP compounds.5b The
spectra can thus be attributed to the pMP localized triplet
(3ππ*) absorption of the two compounds in the various
solvents. Such an assignment is corroborated further by
the TR3 results presented in the next section. From this
and the general correlation between the kinetics of the TA
spectral transformation and the TRF dynamics of the 1nπ*
decay, the observed TA evolution can be ascribed to arise
from the ISC of the excited state population from the 1nπ*
singlet to the 3ππ* triplet of the two systems in the different
solvents. The subtle difference of the time constants in the
TA (τ ≈ 2-3 ps) and TRF (τ ≈ 1-1.3 ps) may be related to
the fact that the two spectral methods detect different
electronic transitions (up to the upper excited state in the
TA spectra and down to the S0 in the TRF spectra), which
might be affected slightly differently by the accompanying
excited state relaxation processes such as intramolecular
vibrational relaxation (IVR) and IC.
effect characteristic of a nπ* transition.9 This effect arises
1
from the presence in the former solvent of specific HB
interactions between solvent molecule(s) and the lone pair
electrons of the n orbital that causes energy destabilization of
the n f π* transition compared to that in the latter solvents
that do not have this hydrogen-bonding effect. On the other
hand, given the general small influence of the HB interaction
on a 1ππ* character transition, the substantial red shift of the
1ππ* fluorescence in the H2O/CH3CN (∼360 nm) solvent
compared to that in the CH3CN (∼330 nm) solvent may
reflect a solvent polarity induced energy stabilization of the
π f π* in the mixed aqueous solvent compared with in the
neat CH3CN solvent (the dielectric constants are ∼78 and
∼37 for H2O and CH3CN). This implies there is a sizable
increase of the molecular dipole moment (i.e., a somewhat
charge transfer (CT) character) in the 1ππ* state in relation
to the ground (S0) state, which is well within expectation with
regard to the electron-donating and -accepting properties of
the methoxy and carbonyl groups, respectively, and the
relative para-position of the two substituents, which is along
the direction of the transition moment of the La type π f π*
electronic transition.
Using the above assignments, the rapid TRF decay and
evolution can be taken as direct spectral manifestation for a
prompt deactivation of the photoexcited 1ππ* state by
1
mainly an internal conversion (IC) to the nπ* state (τ ≈
0.1 ps) and a subsequent decay of the 1nπ* state in the two
solvents (τ ≈ 1 or 1.3 ps, respectively). The analogous decay
dynamics of the two states in the two solvents suggests that,
though this process induces a wavelength shift of the corre-
lated transient fluorescence spectra, the different properties
(polarity and HB capability) of the two solvents have little
influence on the deactivation behavior (e.g., the pathway and
dynamics) of the two states. As confirmed by the TA results
described below, the major deactivation channel responsible
for the fast 1nπ* depopulation is intersystem crossing (ISC)
to a pMP chromophore localized triplet state. The rapid ISC
is consistent with the close to unity triplet yieldreported in the
literature for many aromatic carbonyl compounds.2f,4,5,10
However, It is noted that, probably due to the very short
In contrast to the very similar spectral evolution displayed
by the TA spectra before 10 ps, the triplet spectra of the two
compounds after 10 ps shows clearly different features in
both the spectral decay dynamics and the evolution of the
spectral profiles in the different solvents. For MPEP and
MPPP in H2O/CH3CN and MPPP in the fluorine substituted
alcohols, the temporal changes of the initial triplet TA in the
time interval of ∼10-50 ps exhibit a sharpening and devel-
opment of some structured features in the ∼400-430 nm
wavelength range of the TA profile (see the insets in
Figures 3 and 4). This is not seen in the counterpart spectra
of the two compounds in neat CH3CN. This solvent depen-
dence suggests that certain solvent effects are present only in
the former type of hydrogen-bonding solvents and that these
1
1
lifetimes, the fluorescence data of the ππ* and nπ* states
given here to best of our knowledge represent one of the few
available time-resolved spectral characterizations of singlet
(9) Turro, N. J. Modern Molecular Photochemistry; Benjamin/Cummings:
Menlo Park, CA, 1978.
(10) (a) Ma, C. S.; Chan, W. S.; Kwok, W. M.; Zuo, P.; Phillips, D. L. J.
Phys. Chem. B 2004, 108, 9264–9276. (b) Ma, C. S.; Zuo, P.; Kwok, W. M.;
Chan, W. S.; Kan, J. T. W.; Toy, P. H.; Phillips, D. L. J. Org. Chem. 2004, 69,
6641–6657.
5840 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
FIGURE 3. Fs-TA spectra of MPEP obtained with 267 nm excitation in (a, b) CH3CN and (c, d) H2O/CH3CN. The insets in panels a and c
show early time TA profiles at the indicated wavelengths; the solid lines are from IRF convolved exponential fitting of the experimental data.
The insets in panels b and d display the temporal evolution of the TA spectra at a time range of ∼10-50 ps. The arrows indicate the direction of
the spectral evolution in the two solvents.
solvents can induce noticeable modification of the spectral
character of the triplet absorption. Examination of the corre-
lated TA time profiles at ∼420 nm (see the insets in Figures 3
and 4) found a rise time of ∼12 and ∼20 ps for the growth
dynamics of the structural profiles in H2O/CH3CN and the
fluorinated alcohol solvents, respectively. To help resolve the
nature of the solvent effects and the particular moiety of MPEP
and MPPP that is responsible for the solute-solvent interac-
tion, control experiments were done on the reference com-
pound 2, which has the same pMP chromophore but has no
leaving group. As displayed in Figure 3S in Supporting
Information, the TA spectra acquired for 2 in neat CH3CN
and H2O/CH3CN mixed solvent are very similar and also
closely resemble the spectra recorded for MPEP and MPPP in
neat CH3CN. The lack of the ∼400-430 nm structured feature
for 2 even in H2O/CH3CN is in contrast to the appearance of
this feature observed for MPEP and MPPP in the same solvent.
This observation of the leaving group dependence suggests a
relevance of the leaving group to the solvent effect that affects
the triplet absorption. These results suggest that the temporal
development of the ∼400-430 nm feature showed exclusively
by the triplet of MPEP and MPPP in the water-containing and
fluorinated alcohol solvents may originate from a dynamic
process associated with the partial solvation of the respective
leaving groups of the two compounds by the surrounding
solvent molecules. This is consistent with the known ability
of water and fluorinated alcohols in promoting solvation of an
anionic group.5b,11
Apart from the different spectral character, the later time
triplet absorption of MPEP and MPPP shows distinctly differ-
ent decay dynamics in the different solvents. The triplet kinetics
of MPEP (Figure 3) in neat CH3CN and H2O/CH3CN are
similar and show little decay within the several ns time window
covered by the fs-TA measurements; unlike this, the MPPP
triplet (Figure 4) exhibits increasingly faster decay as the solvent
changes from CH3CN to H2O/CH3CN or CF3CH2OH and to
CF3CHOHCF3. A comparison of the TA decay traces of the
MPPP triplet in these solvents is displayed in Figure 5. Expo-
nential fitting of the data points resulted in time constants of
.∼20 ns, ∼5.4 ns, and ∼3.1 ns for the decay dynamics in CH3-
CN, H2O/CH3CN or CF3CH2OH, and CF3CHOHCF3 sol-
vents, respectively. The much faster triplet decay of MPPP upon
changing the solvent from neat CH3CN to the H2O/CH3CN
mixed solvent (or CF3CH2OH) and to CF3CHOHCF3 indi-
cates there is a significant involvement of an additional triplet
deactivation channel in the latter solvents as compared to CH3-
CN. Considering that the protic solvents are highly capable of
encouraging heterolytic cleavage of a leaving group bearing a
C-O bond and that the diphenyl phosphate in MPPP is a more
(11) Reichardt, C. Solvent and Solvent Effect in Organic Chemistry; VCH
Verlag: Weinheim, 1988.
J. Org. Chem. Vol. 75, No. 17, 2010 5841

An et al.
JOCArticle
FIGURE 4. Fs-TA spectra of MPPP obtained with 267 nm excitation in (a, b) CH3CN, (c, d) H2O/CH3CN, and (e, f) CF3CH(OH)CF3. The
insets in panels a, c, and e show early time TA profiles at the indicated wavelengths; the solid lines are from IRF convolved exponential fitting of
the experimental data. The insets in panels b, d, and f display the temporal evolution of the TA spectra over a time range of ∼10-50 ps. The
arrows indicate the spectral evolution in different solvents.
labile leaving group than the diethyl phosphate in MPEP,5b,11
this observation may imply the occurrence within a several
nanosecond scale of a competitive solvent assisted heterolytic
liberation of the leaving group from the triplet of MPPP in these
solvents. To further explore this and to get some insight into the
structural nature and reactivity of the relatively long-lived triplet
of MPEP in the various solvents and MPPP in CH3CN, TR3
measurements have been designed and preformed for the two
systems in both the picosecond and nanosecond time regimes.
Picosecond and Nanosecond Time-Resolved Resonance Ra-
man Spectroscopy. The ps-TR3 measurements with 267 nm/
400 nm pump/probe wavelengths and ns-TR3 acquired with
266 nm/416 nm pump/probe wavelengths for MPEP and
MPPP were done in the various solvents. With reference to
the TA spectra obtained for the two compounds (Figures 3
and 4), the 400 and 416 nm probe wavelengths were chosen to
detect selectively the resonance Raman spectra of the triplet
state for the two systems in the different solvents. Figure 6
shows the representative ps-TR3 and ns-TR3 spectra of
MPEP recorded in neat CH3CN and mixed H2O/CH3CN
solvent. The counterpart ns-TR3 results obtained for MPPP
in the two solvents are displayed in Figure 7. Ns-TR3 spectra
5842 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
obtained for MPEP and MPPP in CF3CH2OH are very similar
to the corresponding data acquired in the water-mixed solvents
and are given in Figure 4S in the Supporting Information. It can
be seen from Figure 6 that the same vibrational features were
observed for MPEP in the two solvents both in the late ps-TR3
spectra and the corresponding ns-TR3 spectra. Comparison of
the MPPP ns-TR3 spectra with the counterpart MPEP data
reveals observation of analogous Raman features which in-
dicates a very similar structural character for the two com-
pounds in the respective solvents. It is noted that the general
TR3 feature in the resonance enhancement profile and vibra-
tional frequencies observed for the triplet spectra of MPEP and
MPPP in neat CH3CN and the water-mixed solvent resemble
those reported for the triplet spectra of the reference compound
2 and pHP caged compounds in the corresponding solvents.
This corroborates the assignment of the TR3 spectra to the
triplets of MPEP and MPPP and suggests further that the
vibrational features revealed in the spectra originate mainly
from the common pMP chromophore of the two compounds.
Close examination of the TR3 spectra and spectral evolution
obtained for MPEP and MPPP (Figures 6 and 7) reveals several
important observations in terms of the significant solvent and
leaving group dependent triplet decay dynamics and some
distinct solvent dependent vibrational features of the triplet
spectra for MPEP and MPPP in the solvents studied. Regarding
the triplet decay, Figure 8 displays the decay kinetics obtained
by fitting the Raman features with Lorentzian band shapes for
the two compounds in neat CH3CN and a H2O/CH3CN mixed
solvent. The kinetics can be fit reasonable well by single expo-
nential function with time constants estimated to be ∼80 and
∼130 ns for MPEP in CH3CN and H2O/CH3CN solvents,
FIGURE 5. Normalized TA decay profiles of MPPP at the 400 nm
wavelengths observed with the 267 nm excitation in (4) CH3CN,
(]) H2O/CH3CN, (0) CH3CH2OH, and (O) CF3CH(OH)CF3. The
solid lines show the instrumental response function (IRF) con-
volved exponential fittings to the experimental data.
FIGURE 6. Shown are ns-TR3 (a, c) and ps-TR3 (b, d) spectra of MPEP in neat CH3CN (a, b) and H2O/CH3CN (1:1, v/v) (c, d) obtained with a
266 nm pump excitation wavelength and 400 nm (ps-TR3) and 416 nm (ns-TR3) probe wavelengths at various delay times that are indicated next
to the spectra.
J. Org. Chem. Vol. 75, No. 17, 2010 5843

An et al.
JOCArticle
FIGURE 7. Ns-TR3 spectra of MPPP in (a) neat CH3CN and (b) a H2O/CH3CN (1:1, v/v) mixed solvent obtained with a 266 nm pump
excitation wavelength and a 416 nm probe wavelength at various delay times that are indicated next to the spectra are displayed. The # is due to
a stray laser line.
unexpected but is reminiscent of the solvent effect reported
earlier in a TR3 study of the reference compound 2 in various
solvents including the neat CH3CN and mixed H2O/CH3CN
solvents.12 That study observed a much prolonged triplet
lifetime accompanied by some noticeable changes in the
intensity distribution pattern and vibrational frequencies of
some Raman bands in the triplet TR3 spectra of 2 in the H2O/
CH3CN solvent (∼160 ns lifetime) as compared to that in the
neat CH3CN solvent (∼40 ns lifetime).12 Density functional
theory (DFT) calculations that took into consideration the
specific solvent-solute H-bonding interaction of one or two
water molecules with the compound 2 revealed that this
solvent effect is associated with the well-known increase of
the basicity of the carbonyl group upon triplet formation and is
caused mainly by the development in the triplet (relative to the
FIGURE 8. Time dependences of the MPEP (ο) and MPPP (])
triplet ∼1590 cm-1 band in CH3CN and the MPEP (0) and MPPP
(Δ) triplet ∼1604 cm-1 band in H2O/CH3CN (1:1, v/v) mixed
S0 state) of a strengthened H-bonding interaction between the
carbonyl oxygen and the surrounding solvent molecule(s) that
solvent obtained in ns-TR3 spectra of MPEP and MPPP displayed
could be present only in the protic solvent.5,12,13 The H-bonding
in Figures 6 and 7, respectively. The solid lines are results of one
exponential fitting of the experimental data points.
was observed to induce a profound modification of the electro-
nic and geometric structures of the triplet making the H-bonded
triplet to have a more phenyl ring localized ππ* character with
an enhanced ring quinoidal feature and a tendency to have a lon-
respectively and ∼40 and ∼10 ns for MPPP in the same two
solvents, respectively. The kinetics data and time constants
derived for the two compounds in CF3CH2OH were found to
ger lifetime than that of the “free” triplet of the quantum mecha-
be very close to the corresponding results obtained in the water-
mixed solvent, and these data are given in Figure 5S in the
nical mixed nπ*/ππ* character in the aprotic solvent.10,12-14 To
further examine the situation for the triplet of MPEP, a careful
Supporting Information. It is remarkable from Figure 8 that the
comparison has been made between the ns-TR3 (or >50 ps late
decay time of the triplet shows a substantial increase for MPEP
in neat CH3CN as compared to that in the water-mixed solvent
time ps-TR3) spectra of MPEP in CH3CN with those in the
H2O/CH3CN or CF3CH2OH solvents and this comparison is
or the CF3CH2OH solvent. This is in stark contrast to the
shown in Figure 9. Inspection of the spectra in Figure 9 shows
marked decrease in the triplet lifetime observed for MPPP upon
some clear differences between the spectra in the two types of
changing from the former to latter solvents. The very fast triplet
solvents; several features show frequency shifts and there are
decay observed for MPPP in H2O/CH3CN and CF3CH2OH
obvious divergences in the relative intensities of the Raman
coincides with the rapid quenching of the triplet absorption
revealed by the fs-TA measurements for the same compound in
bands especially in the 1350-1500 cm-1 spectral region. These
spectral variations are analogous to that reported for 2 and
the corresponding solvents. Considering the ∼10 ns time reso-
lution of the present ns-TR3 experiments, it is reasonable that,
imply that the carbonyl H-bonding effect may also apply to the
MPEP triplet and contribute to the spectral and lifetime differ-
ences observed in the different solvents here. In particular, the
for the measurements of triplet decay dynamics for MPPP, the
∼10 ns decay time constant estimated on the basis of the ns-TR3
frequency upshift (by ∼15 cm-1) displayed by the ring central
spectra can be considered as an upper limit and the ∼5.4 ns
decay time constant provided by the fs-TA data may serve more
reliably in these solvents.
(13) (a) Zuo, P.; Ma, C. S.; Kwok, W. M.; Chan, W. S.; Phillips, D. L.
J. Org. Chem. 2005, 70, 8661–8675. (b) Rauh, R. D.; Leermakers, P. A. J. Am.
Chem. Soc. 1968, 90, 2246–2249. (c) Nakayama, T.; Otani, A.; Hamanoue,
K. J. Chem. Soc., Faraday Trans. 1995, 91, 855–861. (d) Ramseier, M.; Senn,
P.; Wirz, J. J. Phys. Chem. A 2003, 107, 3305–3315.
The observation that the triplet of MPEP shows a slower
decay in the protic than aprotic solvent appears to be rather
(14) (a) Fang, W. H.; Phillips, D. L. ChemPhysChem 2002, 3, 889–892.
(b) Chen, X. B.; Ma, C. S.; Kwok, W. M.; Guan, X. G.; Du, Y.; Phillips, D. L.
J. Phys. Chem. A 2006, 110, 12406–12413.
(12) Chan, W. S.; Ma, C. S.; Kwok, W. M.; Phillips, D. L. J. Phys. Chem.
A 2005, 109, 3454–3469.
5844 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
in both of the solvents. This observation of a simultaneous
frequency and bandwidth variation of the vibrational bands
on the tens of ps time scale is typical for energy relaxation
processes of the related excited-state species.10,15 The TR3
intensity and spectral evolution observed here can thus be
ascribed to the convolved processes of ISC (τ ≈ 2-3 ps) with
some energy and/or conformation relaxation in the different
solvents. Considering the massive excess energy that the 267
nm photoexcitation could introduce into the nascent triplet
state (a typical triplet state has an energy of 69-71 kcal/mol),
it is quite certain that there is some dynamical relaxation of
the excess vibrational energy in the triplet manifold contri-
buting to the processes observed in the ps spectra. For the
cases taking place in the water-mixed solvent and also in the
fluorinated alcohol solvents, in addition to the excess energy
relaxation, solvation related relaxation such as dynamical
reorientation of solvent molecules in response to the solva-
tion of the leaving group and the development of the
carbonyl located H-bonding could also be involved and
account jointly for the observed early time TR3 temporal
evolution. Taken together with the observed solvent specific
spectral features revealed in the later time TR3 spectra, it is
noticeable that, among the various vibrational bands seen in
the TR3 spectra, the ∼1590-1605 cm-1 band of the ring
central CdC stretching mode appears to be distinctive in
terms of being representative of a particular solvation con-
figuration and the consequently induced subtle modifica-
tion of the pMP triplet structure in the protic versus aprotic
environments.
FIGURE 9. Comparison of the ns-TR3 of MPEP obtained in
CH3CN and H2O/CH3CN (1:1, v/v) mixed solvent with a 266 nm
pump excitation wavelength and a 416 nm probe excitation wave-
length at a time delay of 10 ns.
CdC stretching vibration of the MPEP triplet on going from
CH3CN (1590 cm-1) to the mixed H2O solvent (1604 cm-1) is
consistent with this. This is one of the characteristic indications
for the enhanced quinoidal ring feature of the triplet in the latter
protic solvent compared to the former nonprotic solvent. It is
noted that a similar variation in the spectral appearance is also
observed for MPPP in these two solvents as well (see the 5 ns
TR3 spectra in Figure 7a and b), implying there is also the
involvement of the carbonyl H-bonding irrespective of the
constituent of the leaving group.
It is interesting to note that the spectral differences shown by
MPEP and MPPP in the protic versus aprotic solvents appear to
be not as much as that reported for 2 in the corresponding
solvents. An example of this is that the corresponding frequency
upshift of the ring CdC stretching was observed to be ∼30 cm-1
in the triplet of 2, which is nearly double the amount seen here for
the triplet of MPEP and MPPP. This difference may be asso-
ciated with the additional presence of the leaving group in MPEP
and MPPP in comparison to 2, which could influence the charge
density of the carbonyl oxygen and affect the strength of the
H-bonding interaction. Provided that the strength of the carbo-
nyl H-bonding correlates with the degree of the ππ* electronic
component of the triplet state, the lesser extent of H-bonding
may imply that MPEP and MPPP, like most of the related
aromatic carbonyls,10,14 have triplet states of largely mixed nπ*/
ππ* character with a modestly increased ππ* character relative
to the nπ* contribution in the protic solvent compared to the
aprotic solvent. This intrinsic electronic feature of the triplet has
important implications in understanding the solvent dependent
reductive versus rearrangement deprotection reactivity of the
two compounds reported in the literature.2b,c,6a,b
Combining the fs-TA and ps/ns-TR3 results for the triplet
states of MPEP and MPPP indicates that these triplet states have
nπ*/ππ* mixed character and are present in the solvent asso-
ciated form with the decay dynamics governed by the lability of
the leaving group and the properties of solvent. The tens to
hundred ns decay dynamics observed for the two compounds in
neat CH3CN solvent and MPEP in the other tested solvents
occur on a similar time scale as that reported for the reference
compound 212 and pHP compounds in the inert CH3CN sol-
vent.5,10,13a It can be inferred from this that the major deactiva-
tion event(s) underlying the triplet decay for these systems would
be the diffusion controlled oxygen quenching of the triplet9 and
other possible decay channel(s), such as the deprotection via
either the reductive or rearrangement reactions, if involved,
could occur only to a very limited extent. The much faster triplet
decay (τ ≈ 5.4 ns) of MPPP in the H2O/CH3CN and fluorinated
alcohols compared to that in neat CH3CN (Figures 5 and 8) is
analogous to that reported for the pHP phosphates (HPPP and
HPEP, eq 2) in the corresponding solvents.5,16 The latter is
caused by heterolytic liberation of the leaving group followed by
a stepwise solvolytic rearrangement with the time constant of the
The TR3 spectra recorded at early picoseconds time delay
(Figure 6b and d, <50 ps) exhibit substantial temporal
changes that characterize directly the dynamical evolution
from the initial (<3 ps) to the final (the later time) triplet
spectrum of MPEP in neat CH3CN and mixed H2O/CH3CN
solvents. Comparison of the spectra shows a rapid growth of
the triplet spectrum accompanied by a gradual frequency
upshift and bandwidth narrowing of some Raman features
within ∼50 ps after the photoexcitation. For instance, on
going from 3 to 50 ps, the feature associated with the ring
CdC stretching exhibits a frequency shift from ∼1565 to
∼1585 cm-1 in CH3CN and from 1580 to 1600 cm-1 in H2O/
CH3CN along with a ∼20 cm-1 narrowing of the bandwidth
(15) (a) Liard, D. J.; Busby, M.; Matousek, P.; Towrie, M.; Vlcek, A. J.
Phys. Chem. A 2004, 108, 2363–2369. (b) Liard, D. J.; Busby, M.; Farrell,
I. R.; Matousek, P.; Towrie, M.; Vlcek, A. J. Phys. Chem. A 2004, 108, 556–
567. (c) Hester, R. E.; Matousek, P.; Moore, J. N.; Parker, A. W.; Toner,
W. T.; Towrie, M. Chem. Phys. Lett. 1993, 208, 471–478. (d) Matousek, P.;
Parker, A. W.; Towrie, M.; Toner, W. T. J. Chem. Phys. 1997, 107, 9807–
9817. (e) Ma, C.; Kwok, W. M.; Matousek, P.; Parker, A. W.; Phillips, D.;
Toner, W. T.; Towrie, M. J. Raman Spectrosc. 2001, 32, 115–123. (f) Weaver,
W. L.; Huston, L. A.; Iwata, K.; Gustafson, T. L. J. Phys. Chem. 1992, 96,
8956–8961. (g) Iwata, K.; Hamaguchi, H. J. Phys. Chem. A 1997, 101, 632–
637.
(16) Cao, Q.; Guan, X. G.; George, M. W.; Phillips, D. L.; Ma, C. S.;
Kwok, W. M.; Li, M. D.; Du, Y.; Sun, X. Z.; Xue, J. D. Faraday Discuss.
2010, 145, 171–183.
J. Org. Chem. Vol. 75, No. 17, 2010 5845

An et al.
JOCArticle
FIGURE 11. Time dependence of the ∼1600 cm-1 band in H2O/
CH3CN (1:1, v/v) mixed solvent obtained in the ns-TR3 spectra
displayed in Figure 10. The solid line is from a one exponential
fitting of the experimental data points.
FIGURE 10. ns-TR3 spectra of MPPP in a H2O/CH3CN (1:1, v/v)
mixed solvent obtained with a 266 nm pump excitation wavelength
and a 223 nm probe wavelength at various delay times that are
indicated next to the spectra.
reaction taking place in the CF3CH2OH solvent. Examination
of the time-dependence of the MPAA TR3 spectra (Figure 11,
Figure 7S in the Supporting Information) shows a time resolu-
tion limited time constant of ∼10 ns for the growth of the pro-
duct in the two different solvents. This product formation dyna-
mics appears to correlate well with the triplet decay dynamics (τ
≈ 5.4 ns) observed in the fs-TA (Figures 4 and 5) and ns-TR3
(Figures 7 and 8) in the corresponding solvents. This is con-
sistent with a concerted triplet deprotection and solvolytic re-
arrangement mechanism for MPPP in the two solvents, though
given the uncertainty associated with the ∼10 ns time resolution
of the ns-TR3 measurements, it might also be possible that there
is a stepwise pathway of the triplet deprotection followed by a
rapid, nearly concurrent rearrangement reaction.
triplet cleavage taking place on the hundreds of ps time scale in
the water/CH3CN mixed solvents. To explore the particular
reaction that is responsible for the rapid triplet quenching of
MPPP in these solvents and also to make a connection with and
to determine the rate of the pMP rearrangement deprotection
reaction, photochemistry experiments and ns-TR3 measure-
ments were done to directly detect the formation dynamics of
the rearrangement product for MPPP in the water/CH3CN
mixed solvent and the CF3CH2OH solvent.
Photochemistry and Nanosecond Time-Resolved Resonance
Raman Spectroscopy Experiments To Probe the Rearrangement
Product Formation for MPPP in a Water/CH3CN Mixed Sol-
vent and a CF3CH2OH Solvent. Consistent with the major
photoproduct reported for MPPP in the literature,1a,2 the
UV-vis spectra and spectral transformation (Figure 1S in the
Supporting Information) observed in the photochemistry ex-
periments for the photolysis of MPPP in the H2O/CH3CN (1:1,
v/v) mixed solvent or in the CF3CH2OH solvent indicate the
photodeprotection and rearrangement reactions leading to
release of the phosphate leaving group and formation of the
solvolytic rearrangement products takes place under these
conditions. Referring to the clearly different absorption profiles
exhibited by the MPPP and the MPAA (3) product (spectrum
at 0 and 60 s in Figure 1S in the Supporting Information), a 223
nm probe wavelength was chosen in the ns-TR3 measurements
to selectively follow the dynamics of the formation of MPAA
after the 267 nm photoexcitation of MPPP in the two different
solvents. The spectra obtained in the H2O/CH3CN mixed sol-
vent are displayed in Figure 10. A very similar result is observed
in a CF3CH2OH solvent, and this data is given in Figure 6S in
the Supporting Information.
On the basis of all of the experimental results presented, it
can be concluded that the following photochemical reactions
are induced by the 267 nm excited MPEP and MPPP in the
various solvents (Scheme 1)
This deprotection reactivity of the pMP triplet differs clearly
from that established for the pHP triplet.5b The heterolytic libe-
ration of the leaving group was observed to take place very
rapidly (τ ≈ 150 ps) for the triplet of HPPP in the mixed H2O/
CH3CN (1:1, v/v) solvent. This is much faster (and therefore
more efficient) than the deprotection of MPPP in the same sol-
vent. In contrast to the minor reactivity of MPEP, the deprotec-
tion of HPEP is very efficient and takes place with τ ≈ 420 ps in
the water-containing solvent. The obviously different reactivity
of the pMP to pHP triplets suggests that a crucial effect is
associated with replacing the hydroxy group by the methoxy
group. To provide molecular insight into this, reaction pathway
calculations using the DFT ((U)B3LYP/6-31G*) method that
take into consideration of the specific solvent effect of the water
molecules have been done to investigate the structures, transi-
tion states, and driving force for the pMP triplet deprotection in
a water-containing solvent.
Density Function Theory (DFT) Calculation for the MPPP
Triplet Deprotection. With reference to our previous calcula-
tions for the triplet deprotection of HPEP and HPA
(eq 2),14,16,17 a simple model with the involvement of four
water molecules was adopted for the comparative DFT
calculations performed for MPPP and HPPP. To mimic
the relevant features of the solvent associated configuration
The ns-TR3 spectra (Figure 10) reveal that there are two main
Raman bands at ∼1178 and ∼1600 cm-1 that can promptly be
observed after excitation and appear to be stable up to the tens of
microseconds time scale after the photoexcitation. These spec-
tral features and the overall TR3 spectral profile are nearly iden-
tical to the resonance Raman spectrum reported for an authentic
sample of HPAA obtained with the same probe wavelength and
under the same solvent conditions.5b This result provides explicit
experimental evidence for the formation of a similar rearrange-
ment product MPAA from the photolysis of MPPP in a water-
mixed solvent. Obviously, a similar perspective applies to the
(17) Chen, X. B.; Ma, C. S.; Kwok, W. M.; Guan, X. G.; Du, Y.; Phillips,
D. L. J. Phys. Chem. B 2007, 111, 11832–11842.
5846 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
FIGURE 12. Optimized geometries obtained from (U)B3LYP/6-31G* computations for the reaction complexes (RC), transition states (TS),
and product complexes (PC) of the triplet deprotection reactions of HPPP and MPPP with four water molecules explicitly solvating the system.
SCHEME 1
revealed in the fs-TA and ns-TR3 results, the water molecules
were arranged in a spatial range that allows interaction with the
diphenyl phosphate leaving group on one side and the hydroxyl
(HPPP) or methoxy (MPPP) group on the other side (see Table
2S in the Supporting Information for the detailed Cartesian
coordinates and total energies of the complexes involved in the
reactions). This simple model for the solvation of the molecule is
not meant to imply there is a well-defined four water solute
system but rather allows the water solvent molecules to simul-
taneously interact with the leaving group on one side and the
hydroxyl or methoxy group on the other side while allowing a
partial hydrogen bonding network between the water molecules
interacting with the two sides of the solute molecule. This is
crudely similar to having the solute molecule enmeshed in a
hydrogen-bonded water network that can interact strongly
simultaneously with the leaving group on one side and the
hydroxyl or methoxy group on the other side while having a
hydrogen bond water network surrounding the solute. This
crude four water-solute model was computationally tractable
for our limited computer resources and allowed an exploration
of the one-side (MPPP) versus two-side (HPPP) solvation
effects so as to better understand the different activity for the
reactions of interest. More sophisticated calculations using
many more water molecules along with a QMMM approach
would better mimic the actual solvent-solute interactions and
may provide even deeper insight in the future into the differing
reactivity observed for the systems examined here with a sim-
pler approach. Figure 12 shows the optimized structures found
for the triplet reaction complexes (RC), the transition states
(TS), and product complex (PC) for the MPPP and HPPP
reactions. The barrier crossing in both of the systems leads to a
heterolytic cleavage of the C10-O11 bond to yield a contact ion
J. Org. Chem. Vol. 75, No. 17, 2010 5847

An et al.
JOCArticle
TABLE 1. Key Geometry Information of HPPP and MPPP Reactant Complexes (RC) and Transition States (TS)a
HPPP-4w complex
MPPP-4w complex
TS
RC
TS
diff
RC
diff
˚
bond length of C10-O11 (A)
1.511
-0.149
-0.519
1.689
-0.198
-0.543
1.14
0.178
-0.049
-0.024
1.500
-0.142
-0.520
1.723
-0.206
-0.549
3.04
0.223
-0.064
-0.029
Mulliken charge of C10
Mulliken charge of O11
ΔG‡(kcal/mol)
aObtained at the (U)B3LYP/6-31G* level of theory for the C and O atoms where the cleavage occurs in the deprotection reactions.
pair character PC which is composed of the water-solvated
phosphate anion and the counterpart cation of the pMP or
solvated pHP moiety. It is clear from Figure 12 that, in both the
RC and TS, the four water molecules involved in the reac-
tion form a water chain via H-bonding, by interacting through
the H-bonding of the two ends of the molecule (e.g., with the
leaving group and the phenolic hydrogen, respectively). This
chain of water molecules forms and works as a water bridge in
the case of HPPP (denoted hereafter as two-sided solvation);
such a bridge is however not observed for MPPP because the
change of the hydroxy moiety by the methoxy moiety makes it
unlikely to have an analogous interaction at the methoxy site
and thus leaves available only the solvation at the leaving group
side of the molecule (denoted hereafter as one-sided solvation).
A direct consequence of the different solvation configurations in
HPPP versus MPPP is the larger extent of the electronic con-
jugation between the ring and the electron-donating oxygen in
the former case than in the latter system. This is manifested by
the shorter bond lengths observed in the RC for the related
configuration (Figure 12), it is reasonable that the less favorable
one-side solvation configuration in MPPP relative to the two-
side solvation in HPPP may constitute an important factor to
account for the reduced reactivity of the MPPP triplet com-
pared to the HPPP triplet.
The above results underscore the necessity of the specific
solvation interaction in driving and assisting the triplet
cleavage reactions. Indeed, our previous calculations on
the HPEP triplet deprotection reaction found that, com-
pared to the case without the explicit involvement of the
hydrogen-bonded water molecules, the incorporation of the
four water molecules leads the free energy barrier to drop
substantially.16 Analogous calculations for MPEP and
MPPP observed a similar trend (see Table 1S in the Support-
ing Information). The strong increase in the activation
barrier height brought about by the absence of the water
molecules explains the experimentally observed negligible
triplet reactivity of these compounds in neat CH3CN.
Additionally, in line with the experimental results, our
preliminary comparative calculations have also observed a
trend of a much lower barrier for the MPPP than MPEP
triplet reactions. This provides theoretical evidence to in-
dicate that a more labile leaving group favors the triplet
cleavage, or in other words, solvation of the leaving group
plays an important role in promoting the triplet liberation of
the leaving group. The leaving group solvation can help to
stabilize the ion-pair character product complex; such an
effect has been observed widely for many reactions that
involve heterolytic cleavage of chemical bond(s).5b,11,18 In
fact, for pMP, due to the one-sided solvation configuration,
the solvation of the leaving group supplies the sole driving
force to promote the triplet cleavage. This provides an
interpretation for the experimentally observed faster triplet
decay of MPPP in CF3CH(OH)CF3 than in H2O/CH3CN or
CF3CH2OH solvents (Figure 5), since the former solvent has
a stronger ability in solvating an anion than the latter
solvents. More importantly, this predicts a critical depen-
dence of the pMP triplet cleavage upon the lability of the
leaving group, which is consistent with the experimental
observation that, with a better leaving group, the MPPP
triplet undergoes fairy efficient cleavage, in sharp contrast to
the minor reactivity displayed by the MPEP triplet under the
same experimental conditions (Figures 8).
˚
bonds of the C5-O9 and ring center CdC in HPPP (∼1.365 A
˚
for C4-C3/C1-C6) than found in MPPP (∼1.375 A for C4-
C3/C1-C6). This is in agreement with the vibrational frequency
results observed for the corresponding modes in the TR3
spectra, for example, the CdC stretching is at higher frequency
in HPPP (∼1620 cm
)
-1 5b than MPPP (∼1605 cm-1, Figure 7b)
reflecting more tightening of the bond in HPPP than in MPPP.
The calculated free energy barriers are ∼1.14 kcal/mol for
HPPP and ∼3.04 kcal/mol for MPPP. These low activation
energies are consistent with the generally rapid triplet reactions
observed experimentally for HPPP (∼150 ps) and MPPP (∼5.4
ns) in the water/CH3CN mixed solutions. The substantial
higher deprotection barrier in MPPP than in HPPP is in agre-
ement with the experimentally observed slower triplet decay of
MPPP compared to HPPP. Comparison of the calculated re-
sults on the relevant structures and charge distributions of the
RCs and TSs for MPPP and HPPP contains further useful
information about the origin of the different barrier heights
exhibited by the two systems. Table 1 lists the representative
data of the calculated bond lengths and associated charge den-
sities for the related C-O bond cleavage deprotection reac-
tions. It can be seen from the table that the two systems gene-
rally display modest changes in the structure and charge
distribution upon going from their respective RC to TS. This
suggests there is an early transition state feature for the barrier
leading to deprotection in MPPP and HPPP. A closer compar-
ison observes a higher degree of similarity between the relevant
RC to TS in HPPP than in MPPP. This means there is an easier
access to the TS in HPPP than in MPPP. This corroborates the
lower barrier found for HPPP than MPPP and implies further
that the structure of the RC plays a key role in determining the
accessibility of the activation barrier and the dynamics of the
deprotection reaction. Within the context of this and the fact
that the RC structure is affected sensitively by the solvation
Solvent Dependent Deprotection Mechanism of MPPP and
MPEP. Using all of the experimental and calculated results
reported here in conjunction with relevant results reported in
(18) (a) Newcomb, M.; Horner, J. H.; Whitted, P. O.; Crich, D.; Huang,
X. H.; Yao, Q. W.; Zipse, H. J. Am. Chem. Soc. 1999, 121, 10685–10694. (b)
Whitted, P. O.; Horner, J. H.; Newcomb, M.; Huang, X. H.; Crich, D. Org.
Lett. 1999, 1, 153–156. (c) Horner, J. H.; Taxil, E.; Newcomb, M. J. Am.
Chem. Soc. 2002, 124, 5402–5410. (d) Cameron, J. F.; Willson, C. G.;
Frechet, J. M. J. J. Am. Chem. Soc. 1996, 118, 12925–12937. (e) Lipson,
M.; Deniz, A. A.; Peters, K. S. J. Am. Chem. Soc. 1996, 118, 2992–2997.
5848 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
SCHEME 2
the literature for some closely related systems, the following
picture can be developed for the photophysical and photo-
chemical reactions for 267 nm excited MPEP and MPPP.
After photoexcitation into the strongly absorbing 1ππ* state,
the system evolves promptly (100 fs) by IC to the energetically
nearby 1nπ* state that is followed by rapid ISC (τ≈2-3ps) into
a 3nπ*/ππ* mixed character triplet (Scheme 2) that is present in
a solvent associated form and acts as the reactive precursor to
the subsequent solvent and leaving group dependent chemical
reactions. The triplet state is subject to several dynamically
competing decay channels. Besides the illustrated heterolytic
cleavage (path i) that gives the deprotection and solvolytic
rearrangement product (3 in eq 3), the triplet also has an
intrinsic reactivity to perform a H-abstraction owing to the
involvement of the 3nπ* electronic character of the triplet
configuration.7,9,19 The H-abstraction (with a typical rate of
∼106-109 s-1) may initiate a series of radical processes that can
lead to a second deprotection channel (path ii) which is accom-
panied by formation of the reductive byproduct 2 (eq 3). Of
course, other generally involved triplet deactivation channels
such as an oxygen quenching (path iii, with a typical rate of
non-H-donating solvent (such as MPEP in CH3CN and H2O/
CH3CN and fluorinated alcohols), neither pathway i nor ii
could occur appreciably and consequently the oxygen quen-
ching will turn out to be the prevailing channel to deactivate
the triplet. This perception of the triplet reaction channel
(Scheme 2) provides a unified picture for understanding the
overall mechanism and the factors that may affect the pathway
and reactivity of the pMP triplet.
The reaction mechanisms in Scheme 2 are based on our
experiments and previous reports.2b,c,6a,b Our TA and TR3
experimental results provide the first determination of the
time constants for the elementary steps involved in the
overall deprotection reaction. It is noted that Givens offered
a similar photorearrangement reaction mechanism of pMP
compounds in methanol and tert-butyl alcohol that involved
a spiroketone species as an intermediate;6a,b however, our
experimental results show apparently correlated time con-
stants between the decay of the MPPP triplet (∼5.4 ns) and
the growth of the photorearrangement product (<10 ns)
with little observation of the involvement of any intermedi-
ate. This seems to suggest a concerted triplet deprotection
and solvolytic rearrangement reactions of MPPP in aqueous
solution and fluorinated alcohols.
9
∼108-109 M-1 s-1) that leads to recovery of S0 will also
contribute to the experimentally observed decay of the triplet.
The actual fate and outcome of the triplet decay is governed
by the relative efficiency of these processes, which is in turn
affected strongly by the properties of solvent and lability of the
leaving group as demonstrated in this work. Like the case of
MPPP in the H2O/CH3CN mixed solvent and in the fluori-
nated alcohols, for the direct heterolytic deprotection to be the
predominant triplet decay channel, the system needs to meet a
combined requirement of having a good leaving group and use
a solvent with strong solvating capability. To include the
indirect reductive H-abstraction initiated deprotection, the
system needs to use a solvent that is able to donate a H atom.
The reductive deprotection reported for MPEP in solvents such
as dioxane2b,c or methanol6a,b provide examples for this. In
cases where there is an inert solvent (with little ability in solvat-
ing the ion or donating a H atom) used (MPPP in CH3CN
for instance) or with an unfavorable leaving group and in a
Comparison of the Triplet Deprotection Mechanisms of
pMP with pHP. Comparison of the deprotection mechan-
isms proposed for pMP and that reported for pHP2,5,16,17,20
reveals some similarities and differences. The major simila-
rities are about the spectral features and the dynamics of the
involved photophysical processes that occur before the
triplet reaction. This includes the common feature of a very
rapid IC, ISC, and the accompanying early time relaxation
processes including the excess energy relaxation and solva-
tion related solvent reorganization. The close resemblance of
the corresponding TRF, TA, and TR3 spectra recorded for
pMP and pHP suggests that the corresponding states (1ππ*,
1nπ*, and 3nπ*/ππ*) of the two classes of compounds have
generally similar electronic and structural properties. This is
well within expectation since the hydroxyl and methoxy
groups have similar electronic effects on the intrinsic proper-
ties of the electronic states. The differences in the deprotec-
tion mechanisms are manifested in two aspects. The first is
that unlike the single rearrangement associated deprotection
(19) (a) Du, Y.; Ma, C. S.; Kwok, W. M.; Xue, J. D.; Phillips, D. L. J. Org.
Chem. 2007, 72, 7148–7156. (b) Wagner, P. J.; Kemppain., Ae; Schott, H. N.
J. Am. Chem. Soc. 1973, 95, 5604–5614. (c) Wagner, P. J.; Truman, R. J.;
Scaiano, J. C. J. Am. Chem. Soc. 1985, 107, 7093–7097. (d) Wagner, P. J.;
Nakahira, T. J. Am. Chem. Soc. 1973, 95, 8474–8475. (e) Das, P. K.; Encinas,
M. V.; Scaiano, J. C. J. Am. Chem. Soc. 1981, 103, 4154–4162. (f) Scaiano,
J. C. J. Am. Chem. Soc. 1980, 102, 7747–7753.
(20) Givens, R. S.; Heger, D.; Hellrung, B.; Kamdzhilov, Y.; Mac, M.;
Conrad, P. G.; Cope, E.; Lee, J. I.; Mata-Segreda, J. F.; Schowen, R. L.;
Wirz, J. J. Am. Chem. Soc. 2008, 130, 3307–3309.
J. Org. Chem. Vol. 75, No. 17, 2010 5849

An et al.
JOCArticle
generally displayed by the pHP triplet, the triplet of pMP
shows dual reactivity that can give both the rearrangement
and the reductive associated deprotection pathways. Rele-
vant to this is the fact that the occurrence of the pMP
rearrangement related deprotection requires a much stricter
condition and when it happens, it takes place at a much
slower rate than the corresponding system of pHP. The
second aspect is that previous studies on the pHP deprotec-
tion have established a stepwise deprotection-rearrange-
ment mechanism; the results here on pMP appear to suggest
that the corresponding reaction of pMP is an essentially
concerted process without an apparent clear separation.
The cause of the first aspect is related to the methoxy
versus hydroxyl substitution and can be connected with the
different solvation effects observed for the two kinds of
systems. The two-sided solvation adopted in the pHP triplet
allows for a strong electronic coupling between the triplet
and the surrounding water molecules (or other protic solvent
molecules). This coupling enables the water molecules to
exert a cooperative interaction of both the electronic “push”
through solvation of the hydroxyl hydrogen and the electro-
nic “pull” through solvation of the leaving group on the
triplet. This concerted operation of these two forces favors
strongly the heterolytic cleavage pathway and accounts for
the exclusive high efficiency of this pathway relative to the
other decay channels. However, in the case of the triplet of
pMP, the one-sided solvation allows only the “pull” interac-
tion to be built up significantly so the lack of the electronic
“push” may be responsible for the relatively higher barrier
and the reduced reaction rate of the triplet cleavage in the
pMP systems. The slow dynamics suppresses the efficiency of
this pathway and leads to other channels (H-abstraction,
oxygen quenching) to become involved to compete in the
deactivation of the triplet. Owing to the general low effi-
ciency, the overall deprotection yield of the pMP triplet is
much lower than that of the pHP triplet.
As to the second aspect, early work on the pHP deprotection
have found a sequential deprotection and rearrangement path-
way T f M f HPAA that involves a transient species M to
mediate the phenacyl to phenylacetate solvolytic rearrange-
ment.1,2d,f-h,3b,4,5b,20 Several species, such as the triplet biradi-
cal character p-methide quinine (QM)2f,4,20 and the solvent
associated ion character triplet,2g,5b have been proposed to help
making an assignment to the intermediate M. It is notable that,
although further explicit evidence is still needed to differentiate
and justify the validity of these species, the species were sug-
gested on the basis of a common consideration, which is the
distinct proton-donating capacity of the hydroxyl group in
regard to its marked enhanced acidity or deprotonation capacity
in the triplet compared to the S0 state.2f,5b,13d,21 On this basis, the
hydroxyl group has been taken as being crucially important in
determining the reactivity and the stepwise pathway of the pHP
deprotection and solvolytic rearrangement processes. The cur-
rently observed apparently concerted deprotection-rearrange-
ment of the MPPP triplet implies that while the hydroxyl group
is not totally necessary for the deprotection to take place, the
hydroxyl group may act to contribute a favorable factor to
facilitate lowering the free energy barrier of the deprotection
reaction so as to make it faster. To the best of our knowledge,
our observation of the close to concerted deprotection and
rearrangement for MPPP presents the first direct monitoring
of reactions of this type. In summary, the different reaction
pathways exhibited by the pMP and pHP triplets adds a new
example for the versatile reactivity of the triplets of aromatic
carbonyl compounds. The results here demonstrate that the
substituent group and solvent properties play an important part
in modifying the reaction pathways and determining the photo-
product distribution.
Conclusion
Femtosecond time-resolved fluorescence (fs-TRF), femtose-
cond transient absorption (fs-TA), and picosecond and nano-
second time-resolved Raman (ps-, ns-TR3) experiments were
conducted to study the kinetics and mechanisms of the photo-
deprotection and rearrangement reactions for the pMP com-
pounds p-methoxyphenacyl diethyl phosphate (MPEP) and
diphenyl phosphate (MPPP) in acetonitrile, in water/acetoni-
trile mixed solvents, and in fluorinated alcohol solvents. The
results reported here show that rapid IC (100 fs) and ISC (∼2-3
3
ps) processes lead to a nπ*/ππ* mixed character triplet after
267 nm photoexcitation. The triplet is present in a solvent asso-
ciated form and acts as the reactive precursor to the subsequent
solvent and leaving group dependent chemical reactions seen in
the experimental data. The DFT calculation results provide a
theoretical explanation for the effect of the methoxy versus
hydroxyl substitution on the triplet reactivity and indicate that
the solvent and lability of the leaving group play a crucial role in
determining the rate and pathways of the triplet deprotection.
The one-sided leaving group solvation configuration observed
for pMP triplets in the water-containing solvents is responsible
for the modest triplet heterolytic deprotection that also gives
some rearrangement product. This combined with the nπ*/ππ*
character of the pMP triplets helps to interpret the inclusion of
the alternative reduction associated deprotection pathway ex-
hibited by pMP in some hydrogen-donating solvents.
Experimental and Computational Methods
MPEP and MPPP were prepared following the methods
previously described in the literature.2b,e The identity and purity
of these prepared compounds were confirmed by analysis of the
MS, NMR, and UV-vis absorption spectra. Details of the
preparation and characterization of these compounds are given
in the Supporting Information. The solution samples used for
the time-resolved and steady-state measurements described in
this work were prepared using commercial organic solvents
(HPLC-grade CH3CN and GC grade (g99%) for others) and
deionized water.
The apparatus and methods used for the fs-TA, fs-TRF, and
ps-TR3 experiments were described previously.5 Briefly, the
third harmonic (267 nm) and fundamental laser pulses (800
nm) from a commercial Ti:Sapphire regenerative amplifier laser
system were used as pump and probe laser pulses in the TA
experiment. The time resolution of the TA measurement was
about 200 fs. The spot sizes of the pump and probe beams at the
sample were about 200 and 100 μm, respectively.5b The same Ti:
Sapphire regenerative amplifier laser system was employed in
the TRF experiments with a 150 fs pulse duration and 1 kHz
repetition rate as the excitation and gating pulses, respectively.
The excitation pulses (∼0.5 μJ) were focused (∼100 μm) into a
0.5 mm thick jet stream of sample placed at one focus of an
ellipsoidal mirror.5a For the ps-TR3 measurements, the third
(21) (a) Huck, L. A.; Wan, P. Org. Lett. 2004, 6, 1797–1799. (b) Tolbert,
L. M.; Haubrich, J. E. J. Am. Chem. Soc. 1994, 116, 10593–10600. (c) Webb,
S. P.; Yeh, S. W.; Philips, L. A.; Tolbert, M. A.; Clark, J. H. J. Am. Chem.
Soc. 1984, 106, 7286–7288.
5850 J. Org. Chem. Vol. 75, No. 17, 2010

An et al.
JOCArticle
harmonic (267 nm) and the second harmonic (400 nm) from the
Ti:Sapphire regenerative amplifier laser system with ∼1.5 ps
pulse duration and 1 kHz output were utilized as pump and
probe laser pulses. The pump and probe beams were set at the
magic angle and focused to ∼200 μm spot size on the sample
with pulse energies of 2-3 μJ.5 The apparatus and methods
employed for the ns-TR3 experiments have been described
previously,12,13 and only a short description will be provided
here. The harmonics and/or their hydrogen Raman shifted laser
lines from two nanosecond Nd:YAG laser systems supplied the
pump (266 nm) and probe (223 or 416 nm) wavelengths used in
the ns-TR3 experiments, and a near colinear geometry was
employed to focus them onto a flowing liquid stream of sample.
The time resolution of the experiments was approximately 5-10
ns. The signals for all the experiments were collected and
detected by a nitrogen-cooled CCD detector, and then the data
were sent to an interfaced PC computer. Sample solutions for
these experiments were typically at 1-1.5 mM concentration.
The photochemistry of the reactions for MPPP in H2O/
CH3CN and in CF3CH2OH were monitored by measuring the
UV-vis absorption spectrum of the sample solution (∼0.1 mM
concentration) before and after exposure to an unfocused
266 nm laser line (∼3.5 mJ) from the fourth harmonic of a
nanosecond Nd:YAG Q-switch laser. The sample solution was
put in a 1 cm UV grade cell and excited by the 267 nm laser beam.
As the reaction progressed, the UV-vis absorption spectrum of
the photolyzed sample was measured by using a Perkin-Elmer
Lambda 19 UV-vis spectrometer.
saddle points. Intrinsic reaction coordination (IRC) computa-
tions23 were done to confirm that the transition states connected
the corresponding reactants and products. The Cartesian coordi-
nates, the total energies, and selected output from the calculations
for the precursor species, intermediates, transition states, and final
byproduct are provided in the Supporting Information. Additional
calculations using the 6-311þG(d, p) basis set were done for the for
HPEP and MPEP reactions, and these results were compared to the
results from the calculations already done for the 6-31G* basis set
(see Table 3S in the Supporting Information). The results for HPEP
and MPEP found that the free energy differences from using the
6-311þG(d,p) basis set are both similar to the corresponding ones
obtained from the 6-31G* set and suggest that the data from
6-31G* may be useful to gain some insight into the reactions
examined here.
Acknowledgment. The research was supported by grants
from the Research Grants Council of Hong Kong (HKU 1/
01C and HKU 7039/07P) to D.L.P. and (PolyU/7029/07p
and PolyU/5007/08p) to W.M.K. D.L.P. thanks the Crou-
cher Foundation for the Award of a Senior Research Fellow-
ship and the University of Hong Kong for the Outstanding
Researcher Award.
Supporting Information Available: Recorded spectra of the
MPPPphotochemistry experimentsinthe H2O/CH3CN(1:1, v/v)
mixed solvent and in the CF3CH2OH solvent; fs-TA spectra of
MPPP in CF3CH2OH and the fs-TA spectra acquired for p-
methoxyacetophenone in neat MeCN and MeCN/H2O; ns-TR3
spectra obtained for MPEP and MPPP in CF3CH2OH obtained
with 266 nm pump and 416 nm (ns-TR3) probe wavelengths at
various delay times; time dependences of the MPEP and MPPP
triplets in CF3CH2OH; ns-TR3 spectra of MPPP in CF3CH2OH
obtained with 266 nm pump and 223 nm probe wavelengths at
various delay times; time dependence of MPPP in CF3CH2OH;
key geometry information of the HPEP and MPEP reactant
complexes (RC) and transition states (TS) obtained at the
(U)B3LYP/6-31G* level of theory for the C and O atoms where
the cleavage occurs in the deprotection reaction; Cartesian
coordinates, total energies, and vibrational zero-point energies
determined for the optimized geometries of the HPEP, HPPP,
MPEP and MPPP complexes with four water molecules. This
material is available free of charge via the Internet at http://
pubs.acs.org.
The reactant, product, and transition state geometries for the
MPPP triplet deprotection with the involvement of explicit solvent
water molecules were optimized at the (U)B3LYP/6-31G* level of
theory using the Gaussian 03 program suite.22 Vibrational fre-
quency calculations were performed for all of the stationary points
in order to obtain corrections for the zero point energies (ZPEs) and
to ensure that the computed reactant and product are the corre-
sponding local minimum and the transition states were first-order
(22) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A. J.; Strat-
mann, R. E.; Burant, J. C.; Dapprich, S.; 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, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Baboul, A. G.; 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.7;
Gaussian 03s, revision B.05; Gaussian Inc.: Pittsburgh, PA, 1998, 2003.
(23) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154–2161.
(b) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523–5527.
J. Org. Chem. Vol. 75, No. 17, 2010 5851