Ultrafast time-resolved transient absorption and resonance Raman spectroscopy study of the photodeprotection and rearrangement reactions of p-hydroxyphenacyl caged phosphates

Article abstract of DOI:10.1021/ja0532032

The kinetics and mechanism of the photodeprotection and rearrangement reactions for the pHP phototrigger compounds p-hydroxyphenacyl diethyl phosphate (HPDP) and diphenyl phosphate (HPPP) were studied using transient absorption (TA) and picosecond time-resolved resonance Raman (ps-TR³) spectroscopy. TA spectroscopy was employed to detect the dynamics of the triplet precursor decay as well as to investigate the influence of the solvent and leaving group on the triplet quenching process. Ps-TR³ spectroscopy was used to directly monitor the formation dynamics for the photosolvolytic rearrangement product and its solvent and leaving group dependence. The TA and TR³ spectroscopy experiments were also used to characterize the structural and electronic properties of the triplet precursor to the HPDP and HPPP deprotection reactions. The solvent effect observed in conjunction with the leaving group dependence of the triplet decay dynamics are consistent with a concerted solvent assisted triplet cleavage through a heterolytic mechanism for the HPDP and HPPP photodeprotection process. Correlation of the dynamics between the deprotection and rearrangement processes reveals there is a consecutive mechanism and the involvement of an intermediate between the two reaction steps. The reaction rate of the deprotection and rearrangement steps and the influence of the solvent and leaving group were determined and evaluated based on kinetic modeling of the dynamical data obtained experimentally for HPDP and HPPP in H₂O/MeCN mixed solvents with varying water concentration in the solvent system. A solvation complex with a contact ion pair character was proposed to be the intermediate involved in the deprotection and rearrangement pathway. The results here combined with our previous study on the photophysical events occurring on the early picosecond time scale (Ma; et al. J. Am. Chem. Soc. 2005, 127, 1463-1472) provide a real time overall mechanistic description for the photodeprotection and rearrangement reactions of pHP caged phosphate phototrigger compounds.

Full text of DOI:10.1021/ja0532032

Published on Web 02/04/2006
Ultrafast Time-Resolved Transient Absorption and Resonance
Raman Spectroscopy Study of the Photodeprotection and
Rearrangement Reactions of p-Hydroxyphenacyl
Caged Phosphates
Chensheng Ma, Wai Ming Kwok, Wing Sum Chan, Yong Du, Jovi Tze Wai Kan,
Patrick H. Toy, and David Lee Phillips*
Contribution from the Department of Chemistry, The UniVersity of Hong Kong,
Pokfulam Road, Hong Kong S.A.R., P. R. China
Received May 17, 2005; E-mail: phillips@hkucc.hku.hk
Abstract: The kinetics and mechanism of the photodeprotection and rearrangement reactions for the pHP
phototrigger compounds p-hydroxyphenacyl diethyl phosphate (HPDP) and diphenyl phosphate (HPPP)
were studied using transient absorption (TA) and picosecond time-resolved resonance Raman (ps-TR3)
spectroscopy. TA spectroscopy was employed to detect the dynamics of the triplet precursor decay as
well as to investigate the influence of the solvent and leaving group on the triplet quenching process. Ps-
TR3 spectroscopy was used to directly monitor the formation dynamics for the photosolvolytic rearrangement
product and its solvent and leaving group dependence. The TA and TR3 spectroscopy experiments were
also used to characterize the structural and electronic properties of the triplet precursor to the HPDP and
HPPP deprotection reactions. The solvent effect observed in conjunction with the leaving group dependence
of the triplet decay dynamics are consistent with a concerted solvent assisted triplet cleavage through a
heterolytic mechanism for the HPDP and HPPP photodeprotection process. Correlation of the dynamics
between the deprotection and rearrangement processes reveals there is a consecutive mechanism and
the involvement of an intermediate between the two reaction steps. The reaction rate of the deprotection
and rearrangement steps and the influence of the solvent and leaving group were determined and evaluated
based on kinetic modeling of the dynamical data obtained experimentally for HPDP and HPPP in H2O/
MeCN mixed solvents with varying water concentration in the solvent system. A solvation complex with a
contact ion pair character was proposed to be the intermediate involved in the deprotection and
rearrangement pathway. The results here combined with our previous study on the photophysical events
occurring on the early picosecond time scale (Ma; et al. J. Am. Chem. Soc. 2005, 127, 1463-1472) provide
a real time overall mechanistic description for the photodeprotection and rearrangement reactions of pHP
caged phosphate phototrigger compounds.
Introduction
group has received much attention in this regard due to its
practical potential as a rapid and efficient “cage” for the
There is increasing interest in developing efficient photo-
triggers for real-time monitoring of physiological responses in
biological systems.1-7 The p-hydroxyphenacyl (pHP) protecting
liberation of various biological stimulants.8,9 Previous studies
found that the photodeprotection reaction of pHP caged
compounds is highly solvent-dependent and occurs only in
aqueous or aqueous containing media while the photodepro-
tection reaction does not occur in neat organic solvents such as
acetonitrile (MeCN) where the photoexcitation leads solely to
photophysical processes resulting in the recovery of the ground-
(1) Givens, R. S.; Kueper, L. W. Chem. ReV. 1993, 93, 55-66 and references
therein.
(2) (a) Givens, R. S.; Athey, P. S.; Matuszewski, B.; Kueper, L. W., III; Xue,
J. Y.; Fister, T. J. Am. Chem. Soc. 1993, 115, 6001-6012 and references
therein. (b) Givens, R. S.; Athey, P. S.; Kueper, L. W., III; Matuszewski,
B.; Xue, J.-Y. J. Am. Chem. Soc. 1992, 114, 8708-8710. (c) Gee, K. R.;
Kueper, L. W., III; Barnes, J.; Dudley, G.; Givens, R. S. J. Org. Chem.
1996, 61, 1228-1233.
(3) (a) Il’ichev, Y. V.; Schworer, M. A.; Wirz, J. J. Am. Chem. Soc. 2004,
126, 4581-4595. (b) Rajesh, C. S.; Givens, R. S.; Wirz, J. J. Am. Chem.
Soc. 2000, 122, 611-618. (c) Hangarter, M.-A.; Ho¨rmann, A.; Kamdzhilov,
Y.; Wirz, J. Photochem. Photobiol. Sci. 2003, 2, 524-535.
(4) Rock, R. S.; Chan, S. I. J. Am. Chem. Soc. 1998, 120, 10766-10767.
(5) Namiki, S.; Arai, T.; Rujimori, K. J. Am. Chem. Soc. 1997, 119, 3840-
3841.
(6) (a) Lee, K.; Falvey, D. E. J. Am. Chem. Soc. 2000, 122, 9361-9366. (b)
Banerjee, A.; Lee, K.; Yu, Q.; Fan, A. G.; Falvey, D. E. Tetrahedron lett.
1998, 39, 4635-4638. (c) Banerjee, A.; Falvey, D. E. J. Org. Chem. 1997,
62, 6245-6251.
(7) Zou, K.; Miller, W. T.; Givens, R. S.; Bayley, H. Angew. Chem., Int. Ed.
2001, 40, 3049-3051
(8) (a) Givens, R. S.; Weber, J. F. W.; Conrad, P. G., II; Orosz, G.; Donahue,
S. L.; Thayer, S. A. J. Am. Chem. Soc. 2000, 122, 2687-2697 and
references therein. (b) Conrad, P. G., II; Givens, R. S.; Weber, J. F. W.;
Kandler, K. Org. Lett. 2000, 2, 1545-1547. (c) Gives, R. S.; Park, C.-H.
Tetrahedron Lett. 1996, 37, 6259-6262. (b) Park, C.-H.; Givens, R. S. J.
Am. Chem. Soc. 1997, 119, 2453-2463. (d) Givens, R. S.; Jung, A.; Park,
C.-H.; Weber, J.; Bartlett, W. J. Am. Chem. Soc. 1997, 119, 8369-8370.
(9) Specht, A.; Loudwig, S.; Peng, L.; Goeldner, M. Tetrahedron Lett. 2002,
6, 8947-8950.
9
2558
J. AM. CHEM. SOC. 2006, 128, 2558-2570
10.1021/ja0532032 CCC: $33.50 © 2006 American Chemical Society

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
state substrate molecules.8,10,11 In aqueous or aqueous containing
solvents, along with the liberation of the leaving group, the
photolysis also causes photosolvolytic rearrangement of the pHP
cage into a p-hydroxyphenylacetic acid (HPAA) final product
(eq 1):
phosphate is a much better leaving group than diethyl phos-
phate.14 Since the dynamics of a cleavage process occurring by
a heterolytic mechanism has a distinctly different dependence
on the leaving group from that of a homolytic cleavage,15 it is
expected that comparison of the real time reaction dynamics
displayed by the HPDP and HPPP phototrigger compounds can
provide important evidence to help elucidate the nature of the
photorelease and rearrangement mechanism of the pHP caged
phototriggers.
TA spectroscopy was used to examine the solvent influence
and leaving group dependence of the triplet reactivity in terms
of the photodeprotection reaction. Ps-TR3 experiments were
performed to directly probe and determine the formation
dynamics of the HPAA rearrangement product from photoex-
cited HPDP and HPPP in H2O/MeCN mixed solvents. To our
knowledge, this is the first characterization of the formation
time of the final rearrangement product for pHP phototrigger
compounds. The solvent effect on the deprotection and rear-
rangement reactions was investigated by performing the TA and
ps-TR3 measurements for the two compounds in the mixed
solvents with varying water concentration as well as in other
typical solvents such as dimethyl sulfoxide (DMSO) and 2,2,2-
trifluoroethanol (CF3CH2OH). The TA and TR3 measurements
were also done to characterize the properties of the triplet state
for the new HPPP compound, and these results were compared
to those for HPDP. Correlation of the triplet decay dynamics
with the temporal evolution of the rearrangement product leads
to a quantitative description of the dynamics of the deprotection
and rearrangement reactions that was based on their respective
reaction rates. The solvent and leaving group dependence was
also derived and evaluated for the deprotection and rearrange-
ment reactions. The results presented here shed more light on
the overall mechanistic details of the pHP photochemistry and
enables us to clarify the pertinent controversy of the relevant
reactive intermediate(s) involved in the deprotection and rear-
rangement pathways of pHP caged phototrigger compounds
given in previous studies.8,10,11
Although the products and conditions for the pHP deprotec-
tion have been well established in previous work,10-12 the
reaction mechanism is not well understood and there remains
some uncertainty about the events and reactive intermediates
involved in the photochemical pathway. By using transient
absorption spectroscopy, previous mechanistic work done by
Givens, Wirz and co-workers,10 and Wan and co-workers11
report the observation of several short-lived species after UV
light (∼300 nm) excitation of pHP caged acetate (HPA) and
diethyl phosphate (HPDP) in aqueous containing solvent
systems. However, these studies present some conflicting
interpretations in terms of identities of the relevant intermediates
as well as the processes leading to their formation. Consequently,
two distinct reaction pathways have been proposed for the pHP
deprotection and rearrangement reaction. To clarify this ambigu-
ity in the reaction mechanism, additional work is required. In
particular, the following three issues need to be resolved: (i)
multiplicity for the initial reactive precursor in order to
differentiate between the triplet vs singlet mechanism suggested
by Givens8,10 and Wan,11 respectively; (ii) dynamics and
mechanism of the deprotection (in terms of homolytic vs
heterolytic cleavage8,10) and rearrangement reactions; (iii) role
played by the solvent water molecules in the deprotection
reaction so as to address whether the water molecules work as
a necessary mediator to assist the intramolecular proton transfer
(ESIPT) that consequently facilitates the deprotection and
rearrangement11 or the water molecules simply provide a
favorable solvent environment to encourage a direct cleavage
of the C-O bond connecting the pHP cage and the leaving
group.8,10
Experimental Methods
HPDP was synthesized following the method given in ref 13. The
synthesis of HPPP is similar to that of HPDP, and the details are given
in the Supporting Information. The identity and purity of these
compounds were confirmed by analysis of MS, NMR, and UV
absorption spectroscopy. HPAA is commercially available and was used
after recrystallization. Model compounds p-hydroxyacetophenone (HA)
and p-methoxyacetophenone (MAP) were purchased and used after
recrystallization 3 times in MeCN.
Our recent study employing a combination of femtosecond
Kerr gated time-resolved fluorescence (KTRF) and ps-TR3
spectroscopy13 on HPA and HPDP provided compelling evi-
dence that the triplet is the reactive precursor for the pHP
photorelease reaction. This furnishes an answer to the first issue,
but the next two subjects have not been addressed explicitly to
the best of our knowledge. In this regard, we report here a
subpicosecond TA and ps-TR3 investigation on HPDP and a
newly synthesized pHP caged diphenyl phosphate (HPPP). As
scaled by the pKa of a leaving group anion, the diphenyl
(10) Conrad, P. G., II; Givens, R. S.; Hellrung, B.; Rajesh, C. S.; Ramseier,
M.; Wirz, J. J. Am. Chem. Soc. 2000, 122, 9346-9347.
(11) Zhang, K.; Corrie, J. E. T.; Munasinghe, V. R. N.; Wan, P. J. Am. Chem.
Soc. 1999, 121, 5625-5632.
(12) Brousmiche, D. W.; Wan, P. J. Photochem. Photobiol., A 2000, 130, 113-
118.
(13) Ma, C.; Kwok, W. M.; Chan, W. S.; Zuo, P.; Kan J. T. W.; Toy, P. H.;
Phillips, D. L. J. Am. Chem. Soc. 2005, 127, 1463-1427.
Except for the HPDP and HPPP molecules, the molecular structures
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2559

A R T I C L E S
Ma et al.
for the compounds used in this work (HA and MAP) and those (HPA
and MPDP) important for discussion of our present results are shown
below. Spectroscopic grade MeCN, CF3CH2OH, and DMSO as well
as deionized water were used as solvents for the experiments presented
in this work.
A. Photochemistry Experiments. The photochemistry of the
photoinduced deprotection and solvolytic rearrangement reactions for
the pHP caged compounds was monitored by measuring the UV-vis
absorption spectrum of the sample solution (∼0.1 mM concentration)
before and after exposure to an unfocused 267 nm laser line (∼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. The absolute quantum
yield for the product formation (eq 1) was estimated based on the
decrease of the optical density for the characteristic sample absorption
band (see below), the period of irradiation time, and the laser energy
used for the photolysis measurement.16
Figure 1. UV/vis absorption spectra for the conversion from HPDP to
HPAA upon photolysis of HPDP in a H2O/MeCN (1:1) mixed solvent. The
time periods the sample solution is exposed to the excitation laser pulse
are labeled for each of the spectral traces.
B. Femtosecond Transient Absorption (TA) Experiments. Both
the TA and ps-TR3 experiments were performed based on a commercial
Ti:Sapphire regenerative amplifier laser system equipped with a
homemade OPA system that provides a tunable femtosecond/picosecond
light source.16-18 The TA experiments were done utilizing a newly
developed femtosecond TA setup with a 267 nm pump (the third
harmonic of the 800 nm regenerative amplifier fundamental) and probed
by a white light continuum pulse produced from a rotating CaF2 plate
pumped by the 800 nm fundamental laser pulses.19 After generation,
the white light continuum was collected and focused by an elliptical
mirror and separated into two beams by a beam splitter. One of the
beams (the sample beam) was used as the probe pulse to overlap the
pump pulse (267 nm) in the sample. The other beam (the reference
beam) passes through an unpumped spot in the sample and was used
as a reference to monitor the intensity and spectral variation of the
white light continuum pulse. After these beams passed through the
sample, the two beams were focused into a monochromator and detected
separately in different strips by a liquid nitrogen cooled CCD camera.
The signals from the CCD were then downloaded to a PC for spectral
analyzed. The transient absorbance difference spectrum (the TA
spectrum) was obtained by comparison of the sample-beam spectrum
recorded with the pump beam blocked and unblocked, respectively.
The reference-beam spectrum permits correction of the intensity and
spectral fluctuations of the white light continuum during the TA
experiments. The temporal difference between the pump (267 nm) and
probe (the white light continuum) pulses was controlled using an optical
delay line. The time resolution of the TA measurement was about 200
fs. Sample solutions of ∼1 mM concentration were recycled and
contained in a quartz cell with a 1 mm thickness. The spot sizes of the
pump and probe beams at the sample were about 200 and 100 µm,
respectively. TA experiments were done for HPDP and HPPP and for
the model compounds HP and MAP in neat MeCN and H2O/MeCN
(1:1 by volume) mixed solvents. To examine specific solvent effects,
typical solvents DMSO and CF3CH2OH and the mixture of these two
were also used in the TA experiments for HPDP and HPPP.
C. Picosecond Time-Resolved Resonance Raman (ps-TR3) Ex-
periments. For the ps-TR3 measurements, the samples were pumped
with the 267 nm laser pulse and probed by a 200 nm laser pulse made
by mixing the 800 nm fundamental and the 267 nm output. The pump
and probe pulse durations were ∼1.5 ps. The pulse energies at the
sample were 2-3 µJ. A recirculated solution with a sample concentra-
tion of ∼1-1.5 mM was used with the pump and probe beams loosely
focused onto the thin stream of the sample solution. The Raman light
was collected in a backscattering configuration and detected by a
nitrogen cooled CCD detector. Each ps-TR3 spectrum shown here was
obtained by accumulating over 2 min with an appropriately scaled pump
only and probe only spectra being subtracted from a pump-probe
spectrum. Acetonitrile Raman bands were used to calibrate the TR3
spectrum with an estimated accuracy of (5 cm-1 in absolute frequency.
Ns-TR3 measurements with 267 nm pump and 416 nm probe laser
pulses were also performed for HPPP in MeCN solvent. Ps-TR3
experiments with 267 nm excitation and 400 probe wavelengths were
done for HPDP and HPPP in neat MeCN and H2O/MeCN (1:1) solvent
to examine the influence of carbonyl H-bonding on the structure of
the triplet precursor20 to the pHP photochemistry. Details of the
experimental apparatus and methods employed for these ns- and ps-
TR3 experiments are the same as those described in refs 17, 18, 20,
and 21.
Results
A. Photochemistry of the pHP Caged Phosphates in H2O/
MeCN (1:1) Solvent. Photochemistry experiments were per-
formed for HPDP and HPPP in H2O/MeCN (1:1) solvent to
estimate the quantum yields of the photodeprotection and
rearrangement reactions. Figure 1 shows the UV/vis absorption
spectra obtained after 267 nm photolysis of HPDP in the mixed
solvent. It can be seen that, with an increase of the photolysis
time, the two characteristic absorption bands (maxima at ∼280
and ∼220 nm) of the substrate HPDP decrease in optical density
and shift slightly in position. The absorption spectrum finally
converts into an absorption profile (the absorption trace of the
1 h spectra in Figure 1) that is identical to the absorption
(14) (a) Newcomb, M.; Horner, J. H.; Whitted, P. O.; Crich, D.; Huang, X.;
Yao, Q.; Zipse, H. J. Am. Chem. Soc. 1999, 121, 10685-10694. (b) Whitted,
P. O.; Horner, J. H.; Newcomb, M.; Huang, X.; Crich, D. Org. Lett. 1999,
1, 153-156. (c) Horner, J. H.; Taxil, E.; Newcomb, M. J. Am. Chem. Soc.
2002, 124, 5402-5410.
(15) Reichardt, C. SolVent and SolVent Effect in Organic Chemistry; VCH
Verlagsgesellschaft mbH, D-6940, Weinheim, 1988.
(16) Kwok, W. M.; Zhao, C.; Li, Y.-L.; Guan, X.; Wang, D.; Phillips, D. L. J.
Am. Chem. Soc. 2004, 126, 3119-3131.
(17) Ma, C.; 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.
(18) (a) Ma, C.; Chan, W. S.; Kwok, W. M.; Zuo, P.; Phillips, D. L. J. Phys.
Chem. B 2004, 108, 9264-9276. (b) Kwok, W. M.; Chan, P. Y.; Phillips,
D. L. J. Phys. Chem. B 2004, 108, 19068-19075. (c) Kwok, W. M.; Chan,
P. Y.; Phillips, D. L. J. Phys. Chem. A 2005, 109, 2394-2400.
(19) (a) Ma, C.; Kwok, W. M.; Matousek, P.; Parker, A. W.; Phillips, D.; Toner,
W. T.; Towrie, M. J. Raman Spectrosc. 2001, 32, 115-123. (b) Ma, C.;
Kwok, W. M.; Matousek, P.; Parker, A. W.; Phillips, D.; Toner, W. T.;
Towrie, M. J. Phys. Chem. A 2002, 106, 3294-3305. (c) Ma, C.; Kwok,
W. M.; Matousek, P.; Parker, A. W.; Phillips, D.; Toner, W. T.; Towrie,
M. J. Phys. Chem. A 2001, 105, 4648-4652.
(20) Chan, W. S.; Ma, C.; Kwok, W. M.; Phillips, D. L. J. Phys. Chem. A 2005,
109, 3454-3469.
9
2560 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
Figure 2. Transient absorption spectra of HPDP at early (a) (from 0.25 to
12 ps) and late (b) (from 12 to 5000 ps) picosecond times recorded with
267 nm excitation in MeCN solvent.
Figure 3. Transient absorption spectra of HPDP at early (a) (from 0.1 to
12 ps) and late (b) (from 12 to 2000 ps) picosecond times recorded with
267 nm excitation in H2O/MeCN (1:1) mixed solvent. The right-hand part
in (a) shows profile change of the transient absorption spectra recorded
over 4-12 ps.
spectrum of an authentic sample of the HPAA product. A similar
spectral transformation was also observed for HPPP in the same
solvent. This spectral change reflects the photoinduced genera-
tion of the HPAA rearrangement product and is consistent with
the pHP photoinduced reaction displayed in eq 1. Since
essentially only the photodeprotection and rearrangement reac-
tions occur upon photolysis of pHP caged phototriggers in the
water/MeCN mixed solvent, the quantum yield for the HPAA
formation should be identical to the disappearance quantum yield
of the HPDP reactant. The quantum yield for HPAA formation
can be estimated based on the large decrease of the lowest HPDP
absorption band (∼280 nm) displayed in Figure 1 along with
the irradiation time and laser energy used for the photolysis
experiments.16 This was done by selectively monitoring the
variation of the HPDP absorbance in the 290-310 nm region
where the HPAA product shows little absorption. The quantum
yields obtained are 0.40 ( 0.09 and 0.41 ( 0.01 for HPDP and
HPPP, respectively. The quantum yield obtained here based on
the disappearance of the starting material is very close to the
corresponding value reported previously based on the appearance
of the HPAA product for pHP phototriggers bearing phosphate
esters as the leaving group.8 This implies there is a high
efficiency for the photochemical steps leading to the rearrange-
ment reaction of the pHP caged phototriggers.
changes in the early and late time regimes in the two solvents
are given in Figure 4.
From Figures 2a and 3a, it can be seen that the temporal
evolution of the early time spectra are very similar in both
solvent systems: two absorption bands (a strong one around
400 nm and a broad and weak one at 470-620 nm) grow in
rapidly at the expense of the middle strong band at ∼320 nm.
An isobestic point at ∼330 nm indicates a dynamical conversion
between two distinct states. From Figure 4a, the kinetics of the
∼330 nm decay and the 400 nm and 470-620 nm growth can
be fitted simultaneously by an one exponential function with a
2.5 ps time constant. This value matches well with the value of
the intersystem crossing (ISC) time-constant obtained from our
recent KTRF study13 and indicates that the spectral transforma-
tion observed here is due to the ISC conversion from the lowest
singlet (S1) to triplet (T1). This suggests the very early time
spectra (such as the 0.25 and 0.1 ps spectra in Figure 2a and
3a, respectively) should be assigned to the S1 f Sn absorption
from the S1 (nπ*) singlet state and the fully developed spectra
(12 ps spectra in Figures 2a and 3a) should be assigned to the
T1 f Tn absorption from the T1 triplet (ππ*) state. The triplet
assignment is also corroborated by the late time decay behavior
in the two solvents and the spectral resemblance to the
nanosecond LFP absorption spectra recorded for the pHP caged
compounds in MeCN.10,11 The 267 nm excitation leads to
instantaneous population of the S3 (ππ*) state. The time constant
for internal conversion (IC) from S3 to S1 has been found to be
∼80 fs in both solvent systems according to our previous KTRF
work.13
B. Dynamics of the HPDP Triplet Studied by TA Spec-
troscopy. Figure 2 shows the TA spectra for HPDP obtained
in neat MeCN (267 nm excitation) with time delays from 0 to
5000 ps after excitation. To clearly indicate the spectral changes
at different time scales, the spectra of the early (before 12 ps)
and late (12-5000 ps) times are given separately in the figure.
Corresponding TA spectra in the H2O/MeCN (1:1) mixed
solvent are displayed in Figure 3. The kinetics of the spectral
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2561

A R T I C L E S
Ma et al.
Figure 4. Temporal dependence of transient absorption spectra for HPDP recorded at early picosecond times in MeCN (a) at ∼330 nm (4) and ∼400 nm
(O); at late picosecond times (b) in MeCN (hollow circles) and H2O/MeCN (1:1) (b) at ∼400 nm. The solid lines indicate the kinetics fittings using a one
exponential function to the experimental data points.
The TA spectrum in MeCN solvent shows very little change
at late picosecond times up to 5000 ps (Figure 2b). This is
consistent with the HPDP triplet lifetime that has been observed
to be ∼150 ns in ns-TR3 spectroscopy experiments.17 However,
the TA spectrum in the water mixed solvent shows almost
complete decay at 2000 ps. As displayed in Figure 4b, the triplet
decay kinetics in the water mixed solvent can be fitted by a
single exponential with a time constant of 350 ( 20 ps. This
value parallels the triplet decay rate reported by Wirz and co-
workers employing the TA method (∼400 ps)10 and the value
reported by us using ps-KTR3 spectroscopy (∼420 ( 20 ps).13
This solvent-dependent TA decay dynamics confirms the triplet
Figure 5. Nanosecond time-resolved Resonance Raman (ns-TR3) spectrum
assignment on one hand and validates the triplet as a reactive
of HPPP (a) and HPDP (b) obtained with 267 nm excitation and a 416 nm
probe wavelength in MeCN solvent at 20 ns delay time. The sharp feature
labeled by “#” is due to a stray laser line.
precursor state to the photochemistry events observed previously
in water containing solvent on the other hand.
A careful examination of the TA spectral evolution in the
water mixed solvent (Figure 3) indicates that, besides the overall
decrease in the optical density at late picosecond times, the TA
spectrum displays a small but obvious change in its profile over
the ∼4 to 35 ps time period. The ∼400 nm absorption band is
broad and structureless before 4 ps but appears to gain some
structure afterward. This temporal variation in the profile is not
observed in neat MeCN (Figure 2). According to our previous
ps-TR3 work on pHP related compounds17,18a and other relevant
studies,13,20,22 this phenomenon could have two possible sources:
(i) relaxation of excess energy of the initially formed energetic
triplet state produced from the rapid ISC conversion;13,17,18a (ii)
specific solute-solvent interaction, most possibly the intermo-
lecular H-bond or solvation of the HPDP triplet by the solvent
water molecule(s).20,21 Although source (i) is certainly operative
on the triplet manifold as substantiated by the HPDP ps-TR3
study,13,17 it is unlikely to be the particular reason for the profile
change since this process is not that sensitive to the kind of
solvent used in the system and is expected to result in similar
spectral variations, such as a narrowing or slight spectral shift
of the TA band, in both of the solvent systems. In this regard,
the specific solvent effect would probably be the predominant
factor to account for the profile change observed in Figure 3.
TA measurements on the HA and MAP model compounds that
bear no leaving group found that there is hardly any change in
the spectral shape in both the water/MeCN mixed solvents and
neat MeCN solvent. The TA spectra of HA and MAP in these
two solvents are presented as Figure 1S and 2S in the Supporting
Information. Comparing these results with the HPDP TA result
implies that the profile changes in HPDP might be mainly
associated with solvation of the phosphate leaving group
(possibly through H-bonding interaction with the solvent water
molecules).
C. Triplet Properties and Decay Dynamics of HPPP
Characterized by TA and TR3 Spectroscopy. The TA
experiments analogous to those done for HPDP were also
performed for the newly synthesized HPPP compound in neat
MeCN and mixed MeCN/H2O solvents. The TA results for
HPPP are generally similar to those for HPDP and the recorded
spectra are presented in Figures 3S and 4S in the Supporting
Information. It is obvious that the TA spectra and temporal
evolution at early picosecond times (before 50 ps) closely
resemble those observed for HPDP in the respective solvents.
This suggests that similar photophysical processes like those
observed for HPDP, i.e., rapid ISC conversion (with ∼2 ps time
constant) from the S1 (nπ*) to T1 (ππ*) state, are also being
detected for HPPP in the two solvent systems. The resemblance
in the spectral character for the T1 f Tn absorption observed
for both HPPP and HPDP implies further that these two
compounds have triplet states with similar properties. This is
firmly confirmed by comparison of the ns-TR3 spectra (see
Figure 5) from the two compounds obtained by directly probing
the triplet absorption using a 416 nm probe wavelength and
employing a 267 nm pump wavelength in neat MeCN solvent.
The ns-TR3 spectrum of HPDP displayed in Figure 5 is from
the HPDP triplet as indicated by our previous TR3 spectra
combined with a density functional theoretical (DFT) study.17
(21) Zuo, P.; Ma, C.; Kwok, W. M.; Chan, W. S.; Phillips, D. L. J. Org. Chem.
2005, 70, 8661-8675.
(22) Matousek, P.; Parker, A. W.; Towire, M.; Toner, W. T. J. Chem. Phys,
1997, 107, 9807-9817.
9
2562 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
Figure 6. Temporal dependence of the transient absorption intensity for
HPPP in MeCN (]) and H2O/MeCN (1:1) ([) at ∼400 nm. Data labeled
by the filled circles are from the intensity of the HPDP TA spectra at ∼400
nm obtained in the H2O/MeCN (1:1) solvent. Solid lines indicate kinetics
fitting using a one exponential function to fit the experimental data points.
Figure 7. Triplet decay kinetics observed for HPDP by transient absorption
measurements in water/MeCN mixed solvent with water concentration of
0% (b), 10% ([), 15% (]), 25% (4), and 50% (O). Data labeled by the
filled triangles are obtained in a DMSO/CF3CH2OH (1:1 by volume) mixed
solvent. Solid lines indicate dynamic fittings using a one exponential decay
function to the experimental data points.
Table 1. Time Constants of the Triplet Decay Dynamics and the
Rearrangement Reaction of HPDP and HPPP in H2O/MeCN Mixed
Solvent
corresponding group as a nucleofuge (leaving group that carries
away the bond pair),15 this correlation provides strong evidence
in favor of a direct triplet heterolytic cleavage mechanism for
the pHP deprotection process.
time constant of the triplet decay (τ1
)
a (ps)
15%
H2O%
(vol.)
d
75%
50%
40%
30%
25%
10%
τ2c (ps)
pKa
HPPP
150
830
600
470
0.4
1.39
4.76
HPDP 290
350 530 800 1280 2400 9000
2130
D. Solvent Effect on the Triplet Decay Dynamics. Influ-
ence of Water Concentration on the Triplet Decay Dynam-
ics: To gain further mechanistic insight into the pHP deprotec-
tion reaction, the solvent effect on the triplet decay dynamics
was examined for HPDP and HPPP by using TA spectroscopy.
The TA measurements were performed for the compounds in
H2O/MeCN mixed solvents with various water concentrations.
These experiments resulted in similar TA spectra and analogous
triplet decay kinetics as those observed in the H2O/MeCN (1:
1) solvent (shown in Figure 3). Single exponential dynamics
were observed for the triplet decay in all solvent combinations
with the triplet decay time becoming substantially longer as the
concentration of the water component decreased. Figure 7
displays representative results for the water concentration
dependence of the HPDP triplet decay dynamics. A similar water
concentration dependence was also observed from TA experi-
ments on HPPP. The time constants for the triplet decay
dynamics are summarized in Table 1 for HPDP and HPPP in
the mixed solvents with varying water content. It is interesting
to note that, for solvents with a water content larger than 50%
(by volume ratio), the increase of the triplet decay rate becomes
much more modest upon further addition of water to the solvent
(Table 1). The dependence of the triplet decay dynamics on
water concentration reflects the effect of the local solvent
environment on the triplet quenching process. This may be
associated with the preferential solvation of this solute molecule
in the water/MeCN mixed solvent.27-30 The absence of signifi-
cant improvement of the triplet quenching rate at high water
HPAb
a Uncertainty of the τ1 time is (20 ps. bData for HPA (from ref 13) are
listed for comparison. cCorresponds to time constants of the solvolytic
rearrangement with an uncertainty of (40 ps. dValues of pKa are from ref
14.
The nearly identical features of the HPPP spectrum to that of
the HPDP spectrum indicates unambiguously they have almost
the same triplet structure for the two compounds in terms of
the chromophore (the pHP cage) responsible for the T1 f Tn
absorption probed by the ns-TR3 spectra. This reinforces our
earlier suggestion that photoinduced generation of the triplet
state for pHP caged compounds leads to a localized excitation
on the pHP moiety and is not strongly influenced by the leaving
group.13,17,18a In this sense, examination of the leaving group
and solvent dependence of the triplet decay dynamics can
provide important information for the elucidation of the pho-
todeprotection reaction.
The triplet decay dynamics of HPPP in the two solvents is
displayed in Figure 6. The decay dynamics for HPDP in the
H2O/MeCN (1:1) mixed solvent is also shown for comparison
purposes. Like HPDP, the triplet of HPPP in MeCN shows little
decay within 5000 ps that is consistent with the ∼27 ns triplet
lifetime estimated from a ns-TR3 measurement performed for
HPPP under open air conditions. These spectra and the derived
triplet lifetime decay curve are presented in Figure 5S in the
Supporting Information. In H2O/MeCN (1:1) mixed solvent, the
HPPP triplet decays quickly with an ∼150 ps time constant.
This is faster than the ∼350 ps decay time of the HPDP triplet
in the same solvent. These time constants together with ∼2200
ps triplet decay time found for HPA in this solvent combination
reveals a trend for the correlation of the triplet decay dynamics
with the pKa value of the leaving group anion (Table 1). The
pKa’s of the diphenyl phosphate anion, the diethyl phosphate
anion, and the acetate anion are 0.40, 1.39, and 4.76, respec-
tively.14 Since comparison of the pKa values manifests the
relative stability of the respective anion in aqueous containing
environments as well as reflecting the leaving group ability in
terms of a heterolytic dissociation of the bond bearing the
(23) (a) Shin, D. N.; Wijnen, J. W.; Engberts, J. B. F. N.; Wakisaka, A. J. Phys.
Chem. B 2002, 106, 6014-6020. (b) Shin, D. N.; Wijnen, J. W.; Engberts,
J. B. F. N.; Wakisaka, A. J. Phys. Chem. B 2001, 105, 6759-6762.
(24) Jamroz, D.; Stangret J.; Lindgren J. J. Am. Chem. Soc. 1993, 115, 6165-
6168.
(25) Bago, A.; Scorrano, G.; Stiz, S. J. Am. Chem. Soc. 1997, 119, 2299-
2300.
(26) Takamuku, T.; Tabata, M.; Yamaguchi, A.; Nishimoto, J. J. Phys. Chem.
B 1998, 102, 8880-8888.
(27) Cameron, J. F.; Willson, C. G.; Frechet, J. M. J. J. Am. Chem. Soc. 1996,
118, 12925-12937.
(28) Lipson, M.; Deniz. A. A.; Peters, K. S. J. Am. Chem. Soc. 1996, 118, 2992-
2997.
(29) Pincock, J. A. Acc. Chem. Res. 1997, 30, 43-49.
(30) Richard, J. P.; Amyes, T. L.; Bei, L.; Stubblefield, V. J. Am. Chem. Soc.
1990, 112, 9513-9519.
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2563

A R T I C L E S
Ma et al.
concentration (>50% H2O) implies the predominance of the
solvent water molecules in the HPDP solvation shell at these
large water concentrations.
yield in the H2O/MeCN (1:1) solvent (∼350 ps triplet decay
time), the lower yield in the DMSO/CF3CH2OH solvent mixture
is consistent with the much slower triplet decay rate (∼25 ns
triplet decay time). The lower yield and slower triplet quenching
in the DMSO/CF3CH2OH solvent mixture compared to the H2O/
MeCN (1:1) solvent can be associated with the stronger HBA
and HBD ability and high polarity of water compared to DMSO
and CF3CH2OH.34
Stern-Volmer analysis based on the above triplet decay
measurements leads to a curved and roughly quadratic depen-
dence on water concentration (see Figure 6S in the Supporting
Information). Providing that the addition of water generally
affects a heterolytic dissociation in a nonlinear way, the curved
water concentration dependence observed here corroborates
further the triplet heterolytic cleavage pathway for release of
the phosphate leaving groups in photoexcited HPDP and HPPP.
This water concentration effect implies that the solvent water
molecules not only participate in the pHP photochemistry at
the solvolytic rearrangement step (eq 1) but also play an
important role at a very early stage, and this is crucial to the
deprotection process.
It is reasonable to suggest that the double-bonded oxygen of
the phosphate leaving group and the hydrogen atom of the
hydroxy moiety are the relevant basic and acidic sites, respec-
tively, involved in the concerted triplet solvation and associated
deprotection reaction. On one hand, the importance of the
leaving group solvation is consistent and corroborated with
results that show the triplet decay dynamics for the pHP
compounds investigated (HPPP, HPDP,13 and HPA13) is highly
leaving group dependent, with the decay rate correlating with
the stability of leaving group anion (Table 1). The observation
that the ability of a good leaving group to stabilize the
corresponding solvent solvated anion leads to an encouragement
of the photoinduced dissociation has long been established to
be characteristic of an excited-state heterolytic cleavage
reaction.14,15,27-31 Furthermore, the phosphate anions in HPDP
and HPPP are good nucleofuges,1,2a,32 and a photoheterolytic
cleavage has been proposed for several systems bearing the
phosphates as leaving groups.14 The observed solvent and
leaving group dependent nature of the triplet decay dynamics
can thus be taken as convincing evidence for a triplet quenching
process leading to direct heterolytic cleavage assisted by water
solvation of the leaving group anion. The necessity of the leaving
group solvation for the occurrence of the pHP photodeprotection
also helps to explain the complete lack of reactivity for HPDP
in MeCN10,11 and DMSO solvents. The two solvents are highly
dipolar, but they cannot act as H-bond donors so they are poor
at solvating the departing group.15
On the other hand, solvation of the phenolic proton by a
solvent with HBA capability is associated with the much-
increased acidity of the hydroxy group in the triplet state
compared to the ground state (pKa values of the HA model
compound are 7.9 and 3.6, respectively, in the ground state and
triplet state10).33-38 Water solvation of the phenolic proton is
also required to interpret the previous observation that photolysis
of MPDP (the p-methoxy counterpart of HPDP) in a H2O/MeCN
(1:1) solvent results in hardly any deprotection products.8,11
Since the hydroxy and methoxy group have similar electronic
effects on the intrinsic property of the triplet states,17,20,21 the
absence of reactivity for MPDP argues against a primary step
involving simply the C-O bond heterolysis in the HPDP triplet
and suggests that the proton donating function of the hydroxy
moiety in the pHP triplet10,11,21,35,36 could be crucial to under-
Concerted Solvation of the Triplet State: To be consistent
with the fact that the triplet quenching does not happen in MeCN
solvent, the observed significant triplet quenching especially by
water could be associated with the special properties of solvent
water as well as the substantial change of acid-base properties
(closely related to the electronic structure) for the HPDP
triplet17,20,21 in relation to the ground state. It is well-known
that water may act simultaneously as an HBD (hydrogen bond
donor) and HBA (hydrogen bond acceptor) solvent. This
together with the high relative permittivity (ꢀ ) 78.36) makes
water an extraordinary solvent with both a good ionizing and
dissociating capability that is greatly important for many
reactions involving the breaking of a polarized covalent bond.15
To further elucidate the solvent effect on the triplet quenching
effect in terms of the photodeprotection reaction, we have
performed TA measurements for 267 nm photolysis of HPDP
in solvents of DMSO and CF3CH2OH and a mixed solvent of
DMSO/CF3CH2OH (1:1 by volume). DMSO is a typical HBA
solvent with nearly zero ability to be an HBD solvent while
CF3CH2OH is a good HBD solvent that is not able to act as a
good HBA solvent.15 This particular solvent property enables
selective H-bonding with the HPDP triplet at the acidic site
(phenolic proton) in the DMSO solvent and at the basic site(s)
(the carbonyl oxygen and phosphate anion as leaving group) in
the CF3CH2OH solvent. Solvation by H-bonding at both the
acidic and basic sites by the respective components is possible
in the DMSO/CF3CH2OH (1:1) mixed solvent. The TA experi-
ment in DMSO reveals no triplet quenching, and the recorded
triplet decay kinetics is the same as that found in MeCN. A
small extent of triplet quenching was observed in CF3CH2OH
solvent, but a much more obvious quenching effect was found
in the DMSO/CF3CH2OH (1:1) mixed solvent. The triplet decay
dynamics obtained in this solvent mixture is given in Figure 7
and is displayed together with the quenching dynamics caused
by water in the MeCN mixed solvent. These results show that
the triplet can be quenched rather efficiently in the DMSO/
CF3CH2OH mixed solvent, but not so much in the respective
neat solvent may mean that the presence of solvent with both
the hydrogen bond donating and accepting capacities is essential
for the triplet quenching step. This implies further that concerted
solvation at both the acidic and basic sites is needed for the
deprotection reaction. UV/vis photochemistry measurements
found that the disappearance quantum yield of HDPD is ∼0.17
in the DMSO/CF3CH2OH solvent. Compared with the ∼0.4
(31) Dektar, J. L.; Hacker, N. P. J. Am. Chem. Soc. 1990, 112, 6004-6015.
(32) (a) Givens, R. S.; Matuszewski, B.; Athey, P. S.; Stoner, M. R. J. Am.
Chem. Soc. 1990, 112, 6016-6021. (b) Givens, R. S.; Matuszewski, B. J.
Am. Chem. Soc. 1984, 106, 6860-6861.
(33) Ramseier, M.; Senn, P.; Wirz, J. J. Phys. Chem. A 2003, 107, 3305-3315.
(34) Huck, L. A.; Wan, P. Org. Lett. 2004, 6, 1797-1799.
(35) (a) Tolbert, L. M.; Haubrich, J. E. J. Am. Chem. Soc. 1994, 116, 10593-
10600. (b) Webb, S. P.; Yeh, S. W.; Phillips, L. A.; Tolbert, M. A.; Clark,
J. H. J. Am. Chem. Soc. 1984, 106, 7286-7288.
(36) (a) Porter, G.; Suppan, P. Trans. Faraday Soc. 1965, 61, 1664-1673. (b)
Beckett, A.; Porter, G. Trans. Faraday Soc. 1963, 59, 2051-2057.
(37) Sonnta, J. v.; Mvuia, E.; Hildenbrand, K.; Sonnta, C. V. Chem.sEur. J.
2004, 10, 440-451.
(38) Zepp, R. G.; Gumz, M. M.; Miller, W. L.; Gao, H. J. Phys. Chem. A 1998,
102, 5716-5723.
9
2564 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
Figure 8. Ps-TR3 spectrum of HPDP in H2O/MeCN (1:1) (a) and neat
MeCN (b) obtained with 267 nm excitation and a 400 nm probe wavelength
at a 50 ps delay time.
Figure 9. UV-vis absorption spectra of the HPAA product (solid line)
stand the different photochemical reactivities of MPDP from
HPDP in the same water/MeCN mixed solvent. Additionally,
the concerted solvation perspective has important implications
for the identification of the key intermediate involved in the
deprotection and rearrangement reactions (see below).
and the HPDP substrate (dash line) in a H2O/MeCN (1:1) mixed solvent.
structural parameters for the optimized structures are also listed
in Table 1S in the Supporting Information.
We note that the triplet decay leading to the deprotection
reaction occurs on the hundreds of picoseconds or even longer
time scale (depending on the water concentration). However,
the solvent reorganization required for changing the carbonyl
H-bonding configuration on the triplet manifold takes place in
the early picosecond time regime (within ∼10 ps).20,40 These
preceding results suggest that it is the carbonyl H-bonded triplet
that actually acts as the reactive precursor to the HPDP and
HPPP deprotection reactions. The intrinsic ring localized ππ*
nature of such a precursor triplet state determines that the
deprotection chemistry of pHP caged compounds is fundamen-
tally different from the radical cleavage mechanism proposed
for the deprotection of certain phenacyl caged phototriggers that
occurs from an nπ* triplet and initiated by a hydrogen
abstraction reaction.41
E. Dynamics of the Rearrangement Reaction Studied by
ps-TR3 Spectroscopy. Ps-TR3 measurements were performed
to determine the formation dynamics of the rearrangement
product HPAA for photoexcited HPDP and HPPP in a H2O/
MeCN (1:1) mixed solvent. The UV absorption spectrum of
HPAA (displayed in Figure 9) shows that its lowest two
absorption bands are weak but it absorbs strongly around 190-
200 nm. A 200 nm probe wavelength was therefore chosen for
the ps-TR3 measurements. Figure 10 displays representative ps-
TR3 spectra of HPDP at various delay times obtained with a
267 nm pump and 200 nm probe wavelengths in a H2O/MeCN
(1:1) solvent. An overview of all the TR3 spectra recorded under
these conditions is given as Figure 8S in the Supporting
Information. The resonance Raman spectrum from an authentic
sample of HPAA obtained under the same conditions and using
a 200 nm excitation wavelength is also shown in Figure 10 to
compare with and to help identify the transient observed in the
TR3 spectra. The most important observation in the TR3 spectra
is the simultaneous growth of a series of a new feature at later
picosecond times (∼200 ps and afterward) that reflects the
generation of a new species in this time regime. It is obvious
that the newly developed transient spectra are essentially
identical to the 200 nm resonance Raman spectrum of an
authentic sample of HPAA. This indicates explicitly that the
Carbonyl H-Bonding Effect on the Properties of the
Triplet: The carbonyl oxygen of the pHP cage becomes more
basic in the triplet state compared to ground state.11,34,37-39 This
leads to strengthened intersolute-solvent carbonyl H-bonding
interaction in the triplet state relative to the ground state. Our
previous combined TR3 and DFT study on the MAP20 and HA21
model compounds reveals that the carbonyl H-bonding can cause
a profound modification of the triplet property and is responsible
for the prolonged triplet lifetime of MAP observed in water
containing MeCN solvents compared to neat MeCN solvent.20
The free triplet state observed in neat MeCN has a delocalized
ππ* character (π* electron delocalizes over the ring and
carbonyl subgroups)17,20,21 while the H-bonded triplet recorded
in a H2O/MeCN solvent has a phenyl ring localized biradical
ππ* character with an enhanced quinoidal ring configuration.20,21
The small influence of the leaving group on the properties of
the triplet13,17 implies that the H-bonding effect and associated
property variation may apply generally to the pHP caged
compounds. This is confirmed by comparison of the ps-TR3
spectra for HPDP and HPPP triplet states obtained in neat MeCN
and H2O/MeCN (1:1) solvents. The corresponding spectra for
HPDP are displayed in Figure 8 (similar results were also
obtained for HPPP). It can be seen that there are distinct
differences, though modest, between the two spectra. Especially,
the frequency upshift of the ring center C-C stretching vibration
observed in the mixed solvent (∼1620 cm-1) from the neat
MeCN (∼1600 cm-1) is a characteristic indication for the
increased quinoidal feature in the former than latter solvent
environment.20,21 The broad character of the recorded Raman
bands limits a detailed discussion on the differences of the two
spectra. Nevertheless DFT calculations that consider the car-
bonyl H-bonding interaction on the HPDP triplet confirms it
has a nearly planar biradical ππ* character and indicates that
the change of the HPDP triplet property upon carbonyl
H-bonding closely resembles those observed for the HP and
MAP model compounds. The DFT optimized structure of the
carbonyl H-bonded triplet and comparison of the corresponding
calculated spectrum with the experimental spectrum (Figure 8a)
are given in Figure 7S in the Supporting Information. The
(40) Chudoba, C.; Nibbering, E. T. J.; Elsaesser, T. J. Phys. Chem. A 1999,
103, 5625-5628.
(41) Banerjee, A.; Falvey, D. E. J. Am. Chem. Soc. 1998, 120, 2965-2966.
(39) Wolfbeis, O. S.; Fu¨rlinger, E. J. Am. Chem. Soc. 1982, 104, 4069-4072.
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2565

A R T I C L E S
Ma et al.
Figure 11. Time dependence of the HPAA 850 cm-1 band areas observed
in the ps-TR3 spectra obtained with 267 nm excitation and 200 nm probe
wavelengths for HPDP in H2O/MeCN mixed solvent with H2O concentration
of 50% (O), 25% (4), 15% (]), and 10% (3) and for HPPP in a H2O/
MeCN (1:1) solvent (0). The solid lines represent fitting results of the data
points according to eq 3 (see text for details).
spectra (Figure 10) is displayed in Figure 11 (the data series
indicated by circles). A rough fit of these data by a one
exponential growth function results in a time constant of ∼1100
ps. Comparison of this time constant with the ∼350 ps time
constant triplet delay time observed for HPDP in the same
solvent indicates clearly that the occurrence of the rearrangement
reaction is actually delayed relative to the triplet quenching
process. The delayed formation of the rearrangement product
was observed similarly for HPPP in a H2O/MeCN (1:1) mixed
solvent (Figure 11, data points in squares). Since the triplet decay
time can be considered as the time constant for the release of
the phosphate leaving groups, the delayed rearrangement
dynamics implies a consecutive mechanism and the existence
of an additional intermediate. This additional intermediate is
the direct product of the triplet deprotection and acts simulta-
neously as the immediate precursor to the rearrangement process.
This new intermediate is denoted temporally as “M” in the
following kinetic analysis. The above correlation of the dynam-
ics leads to the following equation (eq 2) to describe the
deprotection and rearrangement reactions of HPDP and HPPP
in a H2O/MeCN mixed solvent. The “T” in the equation
represents the triplet precursor of the deprotection reaction.
Figure 10. Picosecond time-resolved resonance Raman (ps-TR3) spectra
of HPDP obtained with a 267 nm pump and 200 nm probe wavelengths in
a H2O/MeCN (1:1) mixed solvent. Resonance Raman spectrum of an
authentic sample of HPAA recorded with 200 nm excitation is displayed
at the top.
new species can be attributed unambiguously to the HPAA
rearrangement product. The spectral evolution displayed in
Figure 10 thus represents a real time monitoring of the dynamic
transformation from the photoexcited HPDP to the HPAA final
product. To the best of our knowledge, the TR3 results presented
here provide the first direct time-resolved detection of the
solvolytic rearrangement reaction for pHP phototrigger com-
pounds.
Considering the ∼2.5 ps ISC time,13,17 the weak Raman
feature (mainly the band at ∼1550 cm-1) at very early times
(before 5 ps, Figure 10) can be attributed to the S1 singlet state.
The weakness of the S1 feature implies the absorption of S1 f
Sn is weak at the 200 nm probe wavelength. Since the HPDP
triplet has an ∼350 ps lifetime in H2O/MeCN (1:1) solvent,
the observation that TR3 spectra in Figure 10 display no
recognizable features over the 10-200 ps time period implies
that the triplet T1f Tn absorption is either absent or very weak
around the 200 nm probe wavelength. To further confirm the
dynamic observation for the rearrangement reaction in the water
mixed solvent, parallel ps-TR3 measurements (267 nm pump
and a 200 nm probe) were also done for HPDP in neat MeCN
solvent. The TR3 spectra obtained in MeCN (presented as Figure
9S in the Supporting Information) is essentially featureless at
the corresponding late delay times. This is in agreement with
the absence of the HPDP deprotection and rearrangement
reaction in MeCN solvent.8,10,11
k1
k2
T
8 M
8
deprotection
rearrangement
HPAA + phosphate anion (solvated) (2)
In this reaction scheme, the temporal variation of the HPAA
concentration ([HPAA]) is expressed by eq 342
[T]0
-k2t
1
[HPAA] )
[k2(1 - e-k t) - k1(1 - e )] (3)
k2 - k1
where [T]0 stands for the initial concentration of the triplet
precursor; k1 and k2 represent the rate constants of the depro-
tection and rearrangement reactions, respectively.
Providing that the decay of the HPDP triplet due to the
occurrence of the deprotection is much faster (∼500-20 times
faster depending on the water concentration in the range of 75%
-10 vol % water) than the natural triplet decay time (∼150
ns), it is reasonable to assume that the contribution of the triplet
The temporal change of the transient HPAA band area
obtained in the TR3 spectra (Figure 10) reflects the kinetics of
the rearrangement reaction. The time-dependence of the Lorent-
zian fitted ∼850 cm-1 band area of the transient HPAA TR3
(42) Laidler, K. J. Chemical Kinetics; Happer & Row, Publishers: 1987.
9
2566 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
decay by natural photophysical processes can be ignored and
the k1 in eq 3 can be taken as the triplet decay rate determined
in the preceding TA measurements (k1 ) 1/τ1, Table 1 for τ1).
We note that similar results are also observed for HPPP. To
estimate the rearrangement rate of k2 and its influence by the
solvent, the ps-TR3 measurements with a 267 nm pump and a
200 nm probe wavelengths were done for HPDP in water mixed
solvent with the water content varying from 10% to 75% (by
volume). The HPAA formation dynamics obtained represented
by the integrated HPAA ∼850 cm-1 band area are also displayed
in Figure 11. Dynamic fitting of the data series was done using
eq 3, and it was found that the k2 is not that sensitive to the
water concentration. The data obtained in all of the examined
H2O/MeCN combinations can actually be fitted fairly well by
a k2 of ∼2.13 × 10-9 s-1 rearrangement rate (with k2 ) 1/τ2,
this corresponds to a τ2 of ∼470 ps for the time constant). The
lines in Figure 11 display these fitting results. The insensitivity
of k2 on the water concentration may imply that, for solvents
with a water content ranging from 10% to 75% (by volume),
the local solvent environments around the HPDP and HPPP
excited precursor are quite similar in the hundreds to thousand
picosecond time regime (when the rearrangement occurs). Thus,
the difference in the water concentration shows no significant
water concentration dependence on the solvolytic rearrangement
reaction.
An analogous fitting procedure was also adopted to obtain
the k2 for the HPPP rearrangement reaction, and this results in
a k2 of ∼1.67 × 10-9 s-1 rate (corresponding to a τ2 time
constant of 600 ps). The fitting result is shown in Figure 11) in
a H2O/MeCN (1:1) solvent. The time constants of the depro-
tection (τ1) and the rearrangement (τ2) for HPDP and HPPP
are summarized in Table 1. The slower rearrangement time in
HPPP (∼600 ps) than in HPDP (∼470 ps) appears to be
unexpected since our TA result shows that, in a H2O/MeCN
(1:1) solvent, the deprotection time is faster for HPPP (∼150
ps) than for HPDP (∼350 ps). However, this leaving group
dependence of the rearrangement rate has important implications
for understanding the rearrangement process and may help in
the identification of the “M” intermediate. Considering the more
bulky nature of the diphenyl phosphate moiety than the diethyl
phosphate leaving group, the slower k2 in HPPP may simply
reflect the steric influence of the leaving group on the rear-
rangement reaction. It could also imply there is a close
involvement of the leaving group in the “M” precursor to the
solvolytic rearrangement step.
appreciable absorption over the 300-700 nm detection range.
Point (ii) comes from the ps-TR3 data (Figure 10) displaying
only vibrational features from the HPAA product (late times)
and the S1 state (very early times) and no interference from
other unidentified species. Point (iii) is based on the consecutive
reaction scheme of the deprotection and rearrangement reaction
(eq 2). According to the reaction scheme (eq 2), the temporal
dependence of the “M” concentration ([M]) follows42
k1
1
2
[M] ) [T]
(e-k t - e-k t
)
(4)
0k2 - k1
It is obvious from this equation that a significant accumulation
of [M] can only be observed when k1 . k2. In the case where
k1 , k2 a steady-state treatment becomes appropriate and M
cannot be easily observed experimentally due to a rapid
conversion of the nascent M into the HPAA product. This
provides a quantitative correlation to the dynamics data we
obtained for the triplet decay (Figure 7) and HPAA growth
(Figure 11) from HPDP in a solvent with varying water
concentrations. That is, the development time of HPAA shows
a noticeable delay relative to the triplet decay dynamics in high
water content solvents (such as 50% and 75% water content),
while the two times become close in low water content solvent
(such as those with a 15% and 10% water content). Based on
this, detection of the M intermediate is only possible in largely
water containing solvent (>50 vol %) where k1 > k2.
As to attribution of the M species, it is instructive to consider
the importance of a concerted leaving group and hydroxy
solvation to the deprotection reaction. The solvation of phos-
phate leaving groups in HPDP and HPPP can operate by
H-bonding interaction between the double-bonded oxygen (as
a H-bonding acceptor) and the solvent water molecule (as a
H-bonding donor) so as to exert an electrophilic pull on the
departing leaving group anion. This consequently facilitates
dissociation of the corresponding C-O bond. Simultaneously,
solvation of the phenolic proton (as a H-bonding donor) by the
solvent water (as a HBA) results in a certain loss of the proton
from the hydroxy group to the associated water molecule. This
induces an increased electron density on the phenyl ring due to
an improved electronic conjugation from the hydroxy oxygen
to the ring system.21 The electron-enriched ring may in turn
become capable of performing a good nucleophilic attack to
the carbon atom connecting the leaving group so as to exert an
electronic push and facilitate departure of the water solvated
leaving group. This cooperative push-pull scenario is in
excellent agreement with the well-established neighboring group
participation mechanism proposed for the intermolecular rear-
rangement and solvolytic reaction of â-anisylethyl, tosylates,
brosylates, bromides, etc.43-45 The push-pull interaction leads
directly to generation of a transient species analogous to the
spiroketone intermediate (shown below) which has been pro-
posed to be essential for the occurrence of the solvolytic
rearrangement reaction.8,11,46 This in conjunction with the steric
effect of the leaving group on the solvolytic rearrangement step
as well as the effect of the carbonyl H-bonding leads us to
Discussion
A. Spectral Character and Probable Identification of the
Intermediate “M”. Because of its important role, it is necessary
to discuss the spectral features and probable attribution of the
“M” intermediate. Although our present data contain no direct
indication for the identity of this intermediate, it is quite certain
that the species (i) is transparent in the region from 300 to 700
nm and (ii) shows very weak or no appreciable absorption
around 200 nm and (iii) the temporal occurrence of this species
is highly water concentration dependent. Point (i) is based on
the spectral and dynamics of the relevant TA spectra. In the
H2O/MeCN (1:1) solvent (Figure 3b and 4b), the entire TA
spectrum (T1 f Tn absorption) follows a one exponential decay
function and decays back to the prepulse baseline at later times.
This indicates that there is no subsequent species exhibiting an
(43) (a) Cram, D. J. J. Am. Chem. Soc. 1949, 71, 3863-3870. (b) Cram, D. J.
J. Am. Chem. Soc. 1952, 74, 2129-2137. (c) Cram, D. J. J. Am. Chem.
Soc. 1949, 71, 3875-3883.
(44) Swain, C. G.; Eddy, R. W. J. Am. Chem. Soc. 1948, 70, 2989-2994.
(45) Creary, X.; Geiger, C. C. J. Am. Chem. Soc. 1982, 104, 4151-4162.
(46) Park, C-H.; Givens, R. S. J. Am. Chem. Soc. 1997, 119, 2453-2463.
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2567

A R T I C L E S
Scheme 1
Ma et al.
propose a likely candidate for this intermediate species. This
species is suggested to be a solvation complex with contact ion
pair character (shown below).
may be due to the following: (i) The concentration of M is too
low to allow easy detection by TR3. This is possible since the
requirement of k1 . k2 is not fulfilled even under the most
favorable conditions of the present experiment (Table 1). (ii)
The species does not have a strong enough absorption at the
273 nm probe wavelength.
B. Mechanism for the Photodeprotection and Rearrange-
ment Reaction of pHP Caged Phosphate Compounds. The
data presented here provide strong evidence suggesting a direct
triplet heterolytic cleavage mechanism for the deprotection
reaction. It also provides evidence that concerted solvation of
the leaving group and phenolic hydrogen are important and may
work as a driving force to promote the cleavage reaction.
Correlation of the dynamics of the deprotection and rearrange-
ment reactions indicates a consecutive reaction scheme and
involvement of an intermediate that has been tentatively
attributed to a solvation complex with contact ion pair character.
Reaction rates of the deprotection and rearrangement reaction
have been derived based on a kinetic analysis of the dynamics
data recorded experimentally. It was found that the deprotection
rate depends strongly on both the leaving group and water
concentration while the rearrangement rate is only leaving group
dependent but insensitive to the water content. The present work
represents one of the first in terms of real time monitoring and
determination of the reaction pathway and reaction rates for
the pHP photodeprotection and rearrangement reactions and
addressing the important effect of water solvation on the overall
reaction mechanism. This in conjunction with our earlier
femtosecond KTRF13 and relevant TR3 studies17,18a,20,21 on the
photophysics of the pHP compounds leads to the following
mechanistic scheme (Scheme 1). The proposed mechanism
provides an overall account in terms of the early time photo-
physics and deprotection and rearrangement photochemistry for
267 nm photolysis of pHP caged phosphates in H2O/MeCN
mixed solvents (1:1 by volume for example).
Recombination of the contact ion pair may lead to the recovery
of the HPDP and HPPP substrate; separation of the ion pair
accompanied by water solvolysis at the carbonyl carbon leads
to the HPAA product. The solvent water with its high dissociat-
ing ability and nucleophilic capability is ideal for the successful
and efficient ion pair separation and solvolytic rearrangement
reaction to occur.43-46
Considering the nearly planar structure of the precursor triplet
(Table 7S and Table 1S in the Supporting Information), it is
reasonable to suggest that the major chromophore (the phenyl
ring and the joined cyclopropane and solvated hydroxy group)
of the M intermediate may also be mainly planar. It is interesting
that this chromophore may resemble the spiroketone species8,10,11
and they may have parallel electronic properties. DFT computa-
tions at the random-phase approximation (RPA) employing the
Gaussian 98 program suite47 suggest that the spiroketone species
has an electronic transition carrying a reasonably large oscillator
strength at ∼274 nm (the RPA calculated results are given as
Table 2S in the Supporting Information). The ∼274 nm
electronic transition appears to be consistent with the spectral
character found for the M intermediate (points (i) and (ii) above).
We have thus performed a Kerr-gated ps-TR3 experiment with
a 273 nm pump and 273 nm probe wavelength aiming to detect
the M intermediate for both HPDP and HPPP in largely water
containing solvent. However, these spectra reveal no clear
features that can be readily assigned to the M intermediate. This
In the proposed mechanism, the solvent assisted triplet
heterolytic cleavage is the key initiating step to the photochem-
istry of pHP caged phosphates. Special structural and dynamical
effects of the solvent water to solvate the HPDP triplet state
thus play a crucial role in such a process. In this aspect, the
pHP triplet deprotection appears to be analogous to the cleavage
reaction of anion radicals in a series of R-substituted acetophe-
none compounds studied by Save´ant and co-workers48 and
Wagner and co-workers.49 In both cases, the mode of the
(48) Andersen, M. L.; Long, W.; Wayner, D. D. M. J. Am. Chem. Soc. 1997,
119, 6590-6595.
(49) Andrieux, C. P.; Save´ant, J.-M.; Tallec, A.; Tardivel, R.; Tardy, C. J. Am.
Chem. Soc. 1996, 118, 9788-9789.
(47) Frisch, M. J.; et al. Gaussian 98, revision A.7; Gaussian, Inc.: Pittsburgh,
PA, 1998.
9
2568 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006

Photochemistry for p-Hydroxphenacyl Diethyl Phosphate
A R T I C L E S
cleavage can be viewed as an intramolecular dissociative charge
transfer, where the electron is transferred from a π* orbital on
the p-hydroxyphenacyl moiety (for the triplet of the pHP caged
phosphates) or the acetophenone moiety (for the radical anion
cleavage) to a σ* orbital of the breaking bond between the
phenacyl CH2 and the leaving group. A significant degree of
solvent reorganization is required to accommodate the flow of
charge from the π* to the dissociating σ* orbital, and solvent
reorganization has been found to be the main factor in dictating
the height of the intrinsic reaction barrier.48,49 This is consistent
with the solvent effect and relevant interpretation we have
presented here for the HPDP and HPPP triplet quenching
process. For example, the concerted solvation perspective
suggested in this work requires the dynamic rebuilding of the
triplet solvation shell through rearrangement of the surrounding
water molecules that may promote a favorable electronic shift
and structural modification so as to deactivate the triplet
efficiently and consequently account for the deprotection
reaction. It is thus reasonable to suggest that the barrier height
of the triplet photodeprotection for the pHP phosphates might
also be controlled largely by the dynamical reorganization of
the nearby solvent molecules. We note in this regard that a recent
combined experimental and theoretical work by Phillips and
co-workers demonstrates nicely that water-catalyzed solvation
of a neutral leaving group into ions can act as a driving force
to account for photodecomposition of isopolyhalomethanes in
water solvated environments.16 Further theoretical work could
be helpful to provide a detailed molecule-level description of
the reaction pathway for the solvent assisted pHP triplet
deprotection and its subsequent rearrangement reaction.
absorption band at 300-360 nm has been reported in a H2O/
MeCN (1:1) mixed solvent (this band was not seen in neat
MeCN).11 Based on the two exponential decay kinetics of this
absorption band and the observation that the relative contribu-
tions of the two exponential components were sensitive to the
solvent pH value, the band has been associated with two
intermediates, the p-quinone methide (pQM) and its conjugated
base. The pQM intermediate was suggested to be formed by a
water mediated formal proton transfer of the phenolic proton
to the ketone carbonyl and proposed to be the key reactive
precursor for the deprotection reaction (ESIPT model).11 It is
obvious that this band does not appear for photoexcited HPDP
and HPPP in a H2O/MeCN (1:1) solvent (Figure 3, 4, 3S, and
4S) especially in the time period (0-10 ns) when the photo-
release and rearrangement reaction occurs. This means that the
so-called pQM intermediate must not be relevant to deprotection
of these pHP caged phosphates. This suggests that the pQM
may not necessarily be the reactive intermediate to the pHP
photodeprotection reaction, even for the case of HPA. This
reasoning is consistent with the calculation results reported by
Wirz and co-workers that show that the energy of the pQM
triplet is too low to be capable of expelling a leaving group.10
The mechanism shown in the Scheme 1 represents a general
description of the dynamics and reaction pathway followed by
the photophysical and deprotection related photochemical
processes for the pHP caged phosphates. It is believed that this
deprotection and rearrangement mechanism is applicable gener-
ally to various pHP caged phototriggers. It is however worth-
while to point out that although the early picosecond photo-
physical events (the IC and ISC) are relatively leaving group
independent and thus the corresponding pathway is expected
to apply generally to all pHP caged compounds, there might be
additional channel(s) to deactivate the triplet state and the
relative importance of these channel(s) to the deprotection
pathway may depend highly on the ability of the leaving group.
This is based on the triplet heterolytic cleavage mechanism and
the consideration that, in the case of a slow triplet deprotection,
the p-hydroxyphenacyl prototropic related process(es) (such as
the deprotonation and tautomerization of the pHP cage occur
on the nanosecond time regime10,11,21) could become an effective
competing channel to depopulate the triplet. This prospect
enables one to interpret the observation of additional transient
species (assigned to be responsible for the 300-360 nm TA
band observed in the nanosecond to microsecond time regime)
observed for photolysis of HPA11 compared to the absence of
this band for photolysis of HPDP and HPPP in the H2O/MeCN
(1:1) solvent. In contrast to the high sensitivity of the depro-
tection dynamics on the stability of the leaving group anion,
the prototropic related processes localize on the pHP moiety
and its dynamics are expected to be relatively insensitive to
that kind of leaving group.13,21 On one hand, for a good leaving
group such as the diphenyl and diethyl phosphates, respectively,
in HPDP and HPPP, the cleavage process (∼150 and 350 ps
for HPPP and HPDP, respectively) is rapid so that it becomes
the predominant pathway to deactivate the triplet state. Con-
sequently no transient species related to the prototropic process
could be observed after photolysis of HPDP and HPPP in water
containing solvent. On the other hand, for the acetate leaving
group in HPA, due to the slower deprotection rate (2200 ps
triplet decay time), the prototropic related reaction could turn
C. Relevance of Previously Suggested Intermediates to the
Deprotection Reaction. The spectral and dynamical properties
of the intermediate M obtained in the present work enable us
to comment on one of the relevant issues raised from previous
work8,10,11 about the pHP deprotection mechanism, that is, the
possible involvement of the pHP deprotonation related inter-
mediates in the photodeprotection reaction. A TA study on the
photochemistry of the model compound HA in water or water
mixed MeCN solvent found that the HA triplet deprotonation
reaction leads to formation of the ionized HA triplet anion
(3HA-) and subsequently the ground-state anion (HA-) by ISC
conversion.10,21 The 3HA- and HA- transient species was
observed to have an absorption band with λmax at ∼400 and
350 nm, respectively.10 Our conclusion here that the reactive
intermediate is transparent at the 300-700 nm spectral range
precludes involvement of any of the two species in the
deprotection and rearrangement pathway for the pHP caged
phosphates. This is further corroborated by comparison of the
reaction rate between the phenolic deprotonation and release
of the phosphate leaving groups. Our recent ps- and ns-TR3
work found that time constants for the generation of the 3HA-
and HA- are ∼10 ns and ∼90 ns, respectively, in a H2O/MeCN
(1:1) solvent.21 In so far as the leaving group has little influence
on the deprotonation kinetics, the deprotonation is too slow to
compete with the release of phosphate leaving groups that have
been found to occur with time constants of ∼350 and 150 ps,
respectively, for HPDP and HPPP in the same H2O/MeCN
solvent.
In this aspect, we note also that, in the nano- to micro-second
LFP study of the photorelease mechanism for HPA, a transient
9
J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006 2569

A R T I C L E S
Ma et al.
into an efficient competing process and they both make
contributions to the triplet depopulation dynamics. This may
provide an explanation for the observation of the prototropic
related transient species for the photolysis of HPA in the water
mixed solvent.11 Further work is in progress to better understand
the interesting variation of the triplet deactivation pathway with
the nature of the solvent medium (such as the pH condition)
and on the nature of the leaving group.
physical and photochemical events occurring after photolysis
of pHP caged phosphates in various solvent environments.
Acknowledgment. This research was done in the Ultrafast
Laser Facility at the University of Hong Kong and supported
by grants from the Research Grants Council of Hong Kong
(HKU/7108/02P and HKU 1/01C) to D.L.P. W.M.K. thanks the
University of Hong Kong for the award of a Research Assistant
Professorship.
Conclusion
Supporting Information Available: Details for the synthesis
and characterization of HPPP. Figure 1S. Transient absorption
spectra of the HA model compounds obtained in MeCN (a) and
H2O/MeCN (1:1) solvent (b) with 267 nm photoexcitation;
Figure 2S. Transient absorption spectra of the MAP model
compounds obtained in MeCN (a) and H2O/MeCN (1:1) solvent
(b) with 267 nm photoexcitation; Figure 3S. Transient absorption
spectra of HPPP in MeCN at early (a) and late (b) picosecond
times. Figure 4S. Transient absorption spectra of HPPP in H2O/
MeCN (1:1) at early (a) and late (b) picosecond times. Figure
5S. (a) Nanosecond time-resolved resonance Raman (ns-TR3)
spectra of HPPP obtained in MeCN with 267 nm excitation and
416 nm probe wavelength at various delay times. (b) Lifetime
decay of HPPP triplet fitted by one exponential function of the
∼1600 cm-1 Raman band. Figure 6S. Stern-Volmer plot for
triplet quenching dynamics of HPDP by water (in MeCN).
Figure 7S. DFT (UB3LYP/6-311G**) optimized structure for
carbonyl H-bonded triplet of HPDP (a) and comparison of the
calculated spectrum with experimental result (b). Figure 8S.
Picosecond time-resolved resonance Raman spectra of various
delay times for HPDP obtained with 267 nm pump and 200
nm probe wavelength in H2O/MeCN (1:1) solvent; Figure 9S.
Picosecond time-resolved resonance Raman spectra of various
delay times for HPDP obtained with 267 nm pump and 200
nm probe wavelength in MeCN solvent; Table 1S. DFT
(UB3LYP/6-311G**) optimized structural parameters for car-
bonyl H-bonded triplet state of HPDP. Table 2S. Time depend-
ent DFT (B3LYP/6-311G**) RPA calculated result obtained
for excited-state energies and oscillator strengths from the
ground state of the spiroketone species; Complete ref 47. This
material is available free of charge via the Internet at
http://pubs.acs.org.
We report here a TA and ps-TR3 spectroscopy study of the
photodeprotection and rearrangement reactions for HPDP and
HPPP in H2O/MeCN mixed solvents. TA spectroscopy was used
to detect the solvent dependent decay dynamics of the triplet
precursor, and TR3 was used to monitor the formation dynamics
of the HPAA rearrangement product in the mixed solvent
systems. The solvent effect observed in the TA experiments is
consistent with a solvent assisted triplet heterolytic cleavage
pathway for the release of the phosphate leaving groups. These
results provide an indication for the strong coupling of water
that may manifest itself as site-specific concerted solvation of
the hydroxy proton and phosphate leaving group anion through
their respective intermolecular H-bondings with solvent water
molecules to promote the photodeprotection reaction. Correlation
of the dynamics of the deprotection and rearrangement reactions
reveals there is a consecutive mechanism with the rearrangement
occurring subsequently to the deprotection and there is an
involvement of an intermediate between the two reactions. The
reaction rates for the two processes are derived based on a
kinetic modeling of the dynamics data obtained for HPDP and
HPPP in H2O/MeCN mixed solvents with varying water
concentrations. The deprotection rate was found to be sensitive
to the solvent water concentration and is faster for HPPP than
HPDP in the same solvent environment. The rearrangement rate
is generally insensitive to the water content (for water concen-
tration ranges from 10% to 75 vol %) but was estimated to be
slower for HPPP (∼600 ps) than for HPDP (∼470 ps). The
observed dynamics dependence on the solvent and the leaving
group leads to the suggestion of a solvation complex with
contact ion pair character as an intermediate involved in the
pathway of the deprotection and rearrangement reactions. The
results here provide important kinetics and structural data that
allows an overall mechanistic characterization for the photo-
JA0532032
9
2570 J. AM. CHEM. SOC. VOL. 128, NO. 8, 2006