C-O bond fragmentation of 4-picolyl- and N-methyl-4-picolinium esters triggered by photochemical electron transfer

Article abstract of DOI:10.1021/jo049501j

Photochemical reduction of several 4-picolyl- and N-methyl-4-picolinium esters was examined using product analysis, laser flash photolysis, and fluorescence quenching. It is demonstrated that the radical (anions) formed in these reactions readily fragment to yield a carboxylic acid and a 4-pyridylmethyl radical intermediate. The high chemical and quantum yields observed for these photoreactions suggests that these esters can be used as photolabile protecting groups.

Full text of DOI:10.1021/jo049501j

C-O Bon d F r a gm en ta tion of 4-P icolyl- a n d N-Meth yl-4-p icolin iu m
Ester s Tr igger ed by P h otoch em ica l Electr on Tr a n sfer
Chitra Sundararajan and Daniel E. Falvey*
Department of Chemistry and Biochemistry, University of Maryland, College Park, Maryland 20742-2021
falvey@umd.edu
Received March 26, 2004
Photochemical reduction of several 4-picolyl- and N-methyl-4-picolinium esters was examined using
product analysis, laser flash photolysis, and fluorescence quenching. It is demonstrated that the
radical (anions) formed in these reactions readily fragment to yield a carboxylic acid and a
4-pyridylmethyl radical intermediate. The high chemical and quantum yields observed for these
photoreactions suggests that these esters can be used as photolabile protecting groups.
In tr od u ction
protecting groups are important in applications such as
photolithographic fabrication of DNA chips,12-14 photo-
regulation of proteins and cellular signaling pathways,15-17
time-resolved X-ray crystallography,18 and solid-phase
peptide and nucleotide synthesis.14,19 Photolabile groups
are also being used in multistep organic syntheses20,21
and in biological systems.22,23
A specific interest in photoremovable protecting groups
(PRPGs) has lead us to explore PET-triggered C-O bond
scission in phenacyl esters (1, Scheme 1). It was dem-
onstrated that the phenacyl group could be released with
high efficiencies and with excellent chemical yields of the
released substrate.24,25 A particularly attractive feature
of this PET-based PRPG system is that the light absorp-
tion step is decoupled from the bond-breaking step.
Through choice of an appropriate PET sensitizer, it is,
in principle, possible to make the system responsive to
any desired photolysis wavelength. Because the same
protecting group anion is generated with each sensitizer,
the kinetics and mechanism of the release step remain
constant. This contrasts with more conventional PRPG
systems, such as those based on nitrobenzyl, or benzoin
Photoinduced electron transfer (PET) processes have
attracted the interest of a generation of chemists. Previ-
ous work has elucidated many of the factors that control
the rates of these reactions.1,2 Therefore, many current
efforts are aimed at identifying and demonstrating useful
applications of these processes. Applications include solar
energy conversion,3-5 photolithography,6 molecular elec-
tronics,7,8 and the design of novel synthetic transforma-
tions. With regard to the latter, it has long been recog-
nized that PET is capable of generating high-energy
radical and radical ion pairs. Such radical pairs are
interesting because they are often capable of undergoing
bond-forming or bond-fragmentation reactions that are
difficult to carry out using conventional techniques. For
example, Mariano and others have explored PET trig-
gered C-Si fragmentations.9,10 These reactions form
nucleophilic R-amino radicals, which were demonstrated
to be useful in the formation of new carbon-carbon
bonds. Other examples include the reactions of phthal-
imide derivatives triggered by PET from an alkyl-
tethered donor group to the phthalimide chromophore.11
We have had a longstanding interest in applying
photoinduced electron-transfer reactions to the controlled
photorelease of functional molecules. Photoreleasable
(12) Fodor, S. P. A.; Read, J . L.; Pirrung, M. C.; Stryer, L.; Lu, A.
T.; Solas, D. Science (Washington, DC) 1991, 251, 767-773.
(13) Pirrung, M. C. Chem. Rev. 1997, 97, 473-488.
(14) Pirrung, M. C.; Fallon, L.; McGall, G. J . Org. Chem. 1998, 63,
241-246.
(1) Gould, I. R.; Farid, S. Acc. Chem. Res. 1996, 29, 522-528.
(2) Karki, S. B.; Dinnocenzo, J . P.; Farid, S.; Goodman, J . L.; Gould,
I. R.; Zona, T. A. J . Am. Chem. Soc. 1997, 119, 431-432.
(3) Gust, D.; Moore, T. A.; Moore, A. L. Acc. Chem. Res. 2001, 34,
40-48.
(15) Curley, K.; Lawrence, D. S. J . Am. Chem. Soc. 1998, 120, 8573-
8574.
(16) Wood, J . S.; Koszelak, M.; Liu, J .; Lawrence, D. S. J . Am. Chem.
Soc. 1998, 120, 7145-7146.
(17) Pollit, S. K.; Schultz, P. G. Angew. Chem., Int. Ed. 1998, 37,
2104-2107.
(4) Sutin, N. J . Photochemistry 1979, 10, 19.
(5) Maverick, A. W.; Gray, H. B. Pure Appl. Chem. 1980, 52, 2339.
(6) Lockhart, D. J .; Dong, H.; Byrne, M. C.; Follitie, M. T.; Gallo,
M. V.; Chee, M. S.; Mittmann, M.; Wang, C.; Kobayashi, M.; Horton,
H.; Brown, E. L. Nat. Biotechnol. 1996, 14, 1675-1680.
(7) Lukas, A. S.; Bushard, P. J .; Wasielewski, M. R. J . Am. Chem.
Soc. 2001, 123, 2440-2441.
(8) Hayes, R. T.; Wasielewski, M. R.; Gosztola, D. J . Am. Chem. Soc.
2000, 122, 5563-5567.
(9) Yoon, U. C.; J in, Y. X.; Oh, S. W.; Park, C. H.; Park, J . H.;
Campana, C. F.; Cai, X.; Duesler, E. N.; Mariano, P. S. J . Am. Chem.
Soc. 2003, 125, 10664-10671.
(18) Schlichting, I.; Almo, S. C.; Rapp, G.; Wilson, K.; Petratos, K.;
Lentfer, A.; Wittinghofer, A.; Kabsch, W.; Petsko, G.; Goody, R. S.
Nature 1990, 345, 309-315.
(19) Gallop, M. A.; Barrett, R. W.; Dower, W. J .; Fodor, S. P. A.;
Gordon, E. M. J . Med. Chem. 1994, 37, 1233-1251.
(20) Pillai, V. N. R. Synthesis 1980, 1-26.
(21) Pillai, V. N. R. Organic Photochemistry; Marcel Dekker, Inc.:
New York, 1987; Vol. 9.
(22) Nerbonne, J . M.; Richard, S.; Nargeot, J .; Lester, H. A. Nature
1984, 310, 74-76.
(23) McCray, J . A.; Herbette, L.; Kihara, T.; Trentham, D. R. Proc.
Natl. Acad. Sci. U.S.A 1980, 77, 7273-7241.
(24) Banerjee, A.; Lee, K.; Yu, Q.; Fang, A. G.; Falvey, D. E.
Tetrahedron Lett. 1998, 39, 4635-4638.
(10) Yoon, U. C.; Oh, S. W.; Lee, J . H.; Park, J . H.; Kang, K. T.;
Mariano, P. S. J . Org. Chem. 2001, 66, 939-943.
(11) Griesbeck, A. G.; Heinrich, T.; Oelgemoller, M.; Lex, J .; Molis,
A. J . Am. Chem. Soc. 2002, 124, 10972-10973.
(25) Banerjee, A.; Falvey, D. E. J . Org. Chem. 1997, 62, 6245-6251.
10.1021/jo049501j CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/17/2004
J . Org. Chem. 2004, 69, 5547-5554
5547

Sundararajan and Falvey
SCHEME 1
the corresponding carboxylic acids with 4-pyridylcarbinol
and dicyclohexylcarbodiimide (DCC) in methylene chlo-
ride. The photosensitizers employed in this study are
28
listed in Table 1, along with some relevant photophysical
parameters including their excited-state oxidation po-
tentials (Eox* in V). The latter are derived from published
values for the singlet-state energies (Eoo in kcal/mol) and
their (ground-state) oxidation potentials (Eox in V vs SCE)
using eq 1.33
SCHEME 2
Eox* ) Eox - Eoo/23.06
(1)
As shown in Table 1, TMB is the strongest excited-
state reductant, having an Eox* ) -3.17 V (vs SCE). The
least powerful excited-state reductant is triphenylamine
(TPA), which has an Eox* value of -2.06 V. Pyrene and
9-methylcarbazole have intermediate values of -2.17 and
-2.46 V, respectively.
SCHEME 3
Preparative photolyses of the 4-picolyl esters in the
presence of various electron-donating sensitizers were
carried out to determine if C-O bond scission would
occur, releasing the corresponding carboxylic acids. The
4-picolyl esters listed in Table 2 were subjected to filtered
(>320 nm) irradiation with several photosensitizers,
using a 200 W Hg lamp. At these wavelengths the
sensitizers, but not the esters, absorb the light. The
photolysis solution consisted of 4.0 mM of ester and 4.0-
4.5 mM of sensitizer. In each case, an identical solution
was held in the dark as a control sample. The photolysis
and dark control solutions were then analyzed by 1H
NMR and the yields of the carboxylic acids, as well as
the amount of ester consumed, were determined by
integration of the acid peaks relative to an internal
standard.
The 4-picolyl esters release carboxylic acids only when
TMB is used as the sensitizer. When other sensitizers
such as 9-MC and pyrene were employed, no detectable
conversion of the ester was observed, even upon longer
exposure times (upto 8 h). We attribute this to the fact
that the picolyl group is difficult to reduce. It has been
reported that pyridine has Ered of -2.62 V vs SCE.33 The
4-picolyl esters are likely to have a similar potential.
Thus, TMB is able to reduce these esters with an
exergonic PET step, whereas PET is predicted to be
endergonic for pyrene and 9-MC.
The 4-picolyl esters 2 were converted to N-methylpi-
colinium esters in an attempt to increase the reduction
potential of the group, thereby allowing for PET with a
wider array of sensitizers. The reduction potentials of
N-alkylpyridinium salts are reported to be about -1.7
V,33 suggesting that the N-alkylated derivatives would
be better electron acceptors. Picolyl esters 2 were N-
methylated with methyl iodide to give the corresponding
iodide salts 5. These iodide salts display a charge-transfer
band around 350-400 nm, depending on the solvent (see
below). To avoid any ambiguities resulting from compet-
ing absorption by the sensitizers and the charge-transfer
band the iodide counterion was exchanged for perchlor-
ate, yielding 6 as illustrated in Scheme 4. These perchlo-
chromophores, which have the light absorption and bond-
breaking steps occurring at the chromophore. In these
cases, attempts to tune the light absorption properties
of the protecting group often adversely affect the rate of
bond scission. However, one significant limitation of the
phenacyl-based protecting groups is the fact that they
are reduced only at very negative potentials (-2.2 V vs
SCE). This led us to investigate alternative radical ion
bond scission reactions.
The 4-picolyl group (i.e., 4-pyridylmethyl) has been
used as a carboxyl protecting group in peptide synthe-
sis.26,27 Previous reports indicate that they can be re-
leased using cold alkali, catalytic hydrogenation, reducing
metals, or electrolytic reduction28 (Scheme 2). The latter
two reactions were of specific interest to us. Although the
mechanisms had not been examined in detail, we hy-
pothesized that these bond cleavages might occur via the
anion radical of the ester (Scheme 3). Some related
radical heterolytic cleavage reactions have been charac-
terized for â-(phosphatoxy)alkyl and â-(acyloxy)alkyl
radicals. However, simple examples of the latter also
undergo a 1,2-shift.29-31
If the mechanism in Scheme 3 is correct, then it should
be possible to effect a similar reaction via photochemical
reduction. Such a reaction would be interesting not only
from a mechanistic point of view, but it might lead to
the development of improved PRPGs. What follows
describes experiments on the PET reduction of picolyl and
N-methylpicolinium esters. Through photoproduct analy-
sis, fluorescence quenching, laser flash photolysis, and
related experiments it is demonstrated that these esters
do in fact suffer a C-O fragmentation following one-
electron reduction.
Resu lts a n d Discu ssion
Various 4-picolyl esters (2) can be readily prepared by
the reaction of 4-pyridylcarbinol with acid chlorides and
triethylamine in benzene32 or alternately by combining
(26) Camble, R.; Garner, R.; Young, G. T. Nature 1968, 217, 247.
(27) Garner, R.; Young, G. T. J . Chem. Soc. C 1971, 50.
(28) Camble, R.; Garner, R.; Young, G. T. J . Chem. Soc. C 1969,
1911-1916.
(29) Crich, D.; Yao, Q. J . Am. Chem. Soc. 1994, 116, 2631-2632.
(30) Koch, A.; Lamberth, C.; Wettrich, F.; Giese, B. J . Org. Chem.
1993, 58, 1083-1089.
(31) Beckwith, A., L. J .; Duggan, P. J . J . Am. Chem. Soc. 1996, 118,
12838-12839.
(32) Abele, E.; Abele, R.; Gaukhman, A.; Lukevics, E. Synthesis of
Alkyl Heteryl Ethers from Acetates under Interphase Catalysis Condi-
tions in a Liquid/ Solid System; Wiley-Interscience: New York, 1998;
Vol. 34.
(33) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry; 2nd ed.; Marcel Dekker, Inc.: New York, 1993.
5548 J . Org. Chem., Vol. 69, No. 17, 2004

C-O Bond Fragmentation of 4-Picolyl- and N-Methyl-4-picolinium Esters
TABLE 1. P r op er ties of P h otosen sitizer s Used in P h otoclea va ge Exp er im en ts33
sensitizer
Eox*(V vs SCE)
Eoo(kcal/mol)
τ33 (ns)
λ
max (nm)
N,N,N′,N′-tetramethylbenzidine (TMB)
9-methylcarbazole (9-MC)
pyrene
-3.17
-2.46
-2.17
-2.06
83.0
82.1
76.8
70.1
9.4
16.0
190.0
17.0
350
345
370
305
triphenylamine (TPA)
TABLE 2. Sen sitized P h otofr a gm en ta tion of 4-P icolyl Ester s 2
compd
R
sensitizer
conditions
% acid formeda,b
% ester consumeda,b
2d
2e
2f
C(CH3)3
CH(C6H5)2
CH2C6H4CH3
TMB
TMB
TMB
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
14
43
29
31
53
30
a
b
Photolysis solution was 2.5-3.0 mL of a 4 mM solution of ester and 4.0-4.5 mM sensitizer. Yields were determined by 1H NMR
integration of the carboxylic acid peak relative to an internal standard. c Estimated error ( 5%.
TABLE 3. Sen sitized P h otofr a gm en ta tion of N-Meth yl-4-p icolin iu m Ester s 6
compd
R
sensitizer
conditionsa
% acid formedb,c
% ester consumedb,c
6a
6b
6c
6c
6c
6c
6d
6e
6f
CH3
C6H5
CH2C6H5
CH2C6H5
CH2C6H5
CH2C6H5
C(CH3) 3
CH(C6H5) 2
CH2C6H4CH3
CH2C6H4Br
CHdCHC6H5
9-MC
TMB
9-MC
TMB
Pyrene
TPA
9-MC
9-MC
9-MC
9-MC
9-MC
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
3 h, rt, MeOH
62
30
45
92
84
77
43
74
21
49
22
65
26
41
87
92
79
43
78
25
58
64
6g
6h
a
b
Photolysis solution was 2.5-3.0 mL of a 4 mM solution of ester and 4.0-4.5 mM sensitizer. Yields were determined by 1H NMR
integration of the carboxylic acid peak relative to an internal standard. c Estimated error ( 5%.
SCHEME 4
corresponding phenacyl esters demonstrated the difficul-
ties of reducing the protecting group preferentially over
the aryl bromide part of the ester (aryl bromides readily
expel the halide ion following one-electron reduction).20
Phenacyl 4-bromophenylacetate yields a mixture of prod-
ucts upon sensitized photolysis. The latter results from
competing electron transfer to the aryl bromide. Along
with 4-bromophenylacetic acid, the debrominated phena-
cyl phenylacetate and phenylacetic acid were also ob-
tained. In contrast, sensitized photolysis of the corre-
sponding picolinium ester 6g gives only the desired
product, 4-bromophenylacetic acid, in good yields. This
indicates that the picolinium group is far easier to reduce
and can be used for protection of reductively labile
carboxylic acids.
rate salts absorb below 320 nm, with some esters
displaying a weak absorption tail above 320 nm.
Eight examples of 6, listed in Table 3, were prepared.
Sensitized photolyses of these salts were carried out using
a 200 W Hg lamp, with a 320 nm cutoff filter and 9-MC
as the sensitizer. Control experiments were also carried
out on several esters without the sensitizers, under
identical photolysis conditions. Analysis of the photo-
products was carried out in the same manner as de-
scribed for 2. The carboxylic acids were chosen with a
view to demonstrate photorelease in various aliphatic as
well as aromatic esters. Also, as the photoproduct analy-
The cinnamate ester 6h was studied to observe the
effect of the alkene functionality on efficiency of photo-
release. Alkenes, especially highly conjugated ones such
as the cinnamate ester, are known to undergo several
photochemical reactions and are difficult to photodepro-
tect. Upon photolysis, we observed release of cinnamic
acid, although in low yields and with several byproducts.
At higher conversions of the ester, the yields of the free
acids were found to diminish. This may be due to the
photoreactivity of the cinnamate ester or due to photo-
reactions of the product cinnamic acid itself, which could
be undergoing secondary photochemical processes.
1
sis was carried out by H NMR peak integration, it was
desirable to use carboxylic acids with peaks distinct from
those of the sensitizers.
In the absence of the sensitizer, irradiation of the
picolinium salts afforded no photoproducts. Under sen-
sitized irradiation, the esters released the free carboxylic
acids efficiently. For the simple aryl and alkyl esters
6a -f the conversions were clean and the amount of
carboxylic acid formed was, within experimental error,
the same as the amount of ester consumed. It was
observed that the amount of ester consumed was not the
same for the different esters, although the chemical yields
of the carboxylic acids were consistently quantitative.
It is possible to initiate the fragmentation of these
esters using a variety of sensitizers. As shown in Table
3, four different sensitizers, N,N,N′,N′-tetramethylben-
zidine (TMB), 9-methylcarbazole (9-MC), pyrene, and
triphenylamine (TPA), were employed with ester 6c and
An interesting case was N-methylpicolinium 4-bro-
mophenylacetate 6g. Previous PET studies with the
J . Org. Chem, Vol. 69, No. 17, 2004 5549

Sundararajan and Falvey
TABLE 5. Qu a n tu m Yield s for Sen sitized P h otolysis of
6
compd
R
sensitizer
solvent
quantum yield (%)
6c
6c
6c
CH2C6H5
CH2C6H5
CH2C6H5
9-MC
9-MC
TMB
MeOH
MeCN
MeCN
0.23
0.14
0.05
SCHEME 6. P r op osed Mech a n ism for th e
P h otofr a gm en ta tion Rea ction
F IGURE 1. Steady-state UV-vis absorption spectrum of ester
5c in methanol (A), dichloromethane (B), and acetonitrile (C).
To compare the relative efficiencies of sensitized ir-
radiation and direct irradiation of the iodide salts, several
quantum yields were determined. First, the quantum
yield for the sensitized photorelease was determined for
N-methylpicolinium phenylacetate. (6c). A solution of the
ester and an equal amount of sensitizer was irradiated
by a xenon lamp for different periods of time, and the
SCHEME 5
1
TABLE 4. P h otolysis of th e P icolin iu m Iod id es 5
rate of disappearance of the ester was determined by H
NMR. The quantum yields were found to be around 0.14
for the photorelease process (Table 5). The quantum yield
for photorelease from the picolinium iodide salts was
found using N-methylpicolinium benzoate to be 0.15 in
methanol, comparable to the better quantum yields
obtained in the sensitized photolysis.
Having established that picolinium esters release the
free acids upon irradiation, we now turned our attention
to the mechanistic pathway of this photofragmentation
reaction. Scheme 6 describes the anticipated reaction
mechanism. Upon irradiation, the photosensitizer mol-
ecule (Sens) absorbs light preferentially over the sub-
strate and is promoted to its excited singlet state (Sens*1).
The latter transfers an electron to the picolinium ester.
The pyridyl radical 7 subsequently releases the carboxy-
late anion, along with a pyridylmethyl radical fragment
8.
As a rule, PET reactions compete with rapid photo-
physical deactivation of the sensitizer excited state. As
a consequence such reactions are expected to proceed
efficiently only when the initial electron-transfer step is
exergonic. Values for ∆GET can be estimated for the
mechanism in Scheme 6, using the reduction potential
of the N-methylpicolinium group (Ered) and the excited-
state oxidation potential of the sensitizers (Eox*). The
latter are given in Table 1. In polar solvents, such as
MeCN and MeOH, eq 2 provides a reasonable estimate
% acid
% ester
compd
R
conditionsa
formedb,c consumedb,c
5a
5b
5c
5c
5f
CH3
C6H5
3 h, MeOHd
3 h, MeOHd
3 h, MeOHd
67
32
64
30
73
63
74
48
92
45
90
71
CH2C6H5
CH2C6H5
e
3 h, MeCl2
CH2C6H4CH3 3 h, MeOHd
5h
CHdCHC6H5 3 h, MeOHd
a
Photolysis solution was 2.5-3.0 mL of a 4 mM solution of ester.
Yields were determined by 1H NMR integration of the carboxylic
b
acid peak relative to an internal standard. c Estimated error (
5%. Irradiation by Hg lamp, >320 nm cutoff filter. e Irradiation
d
by Xe lamp, >390 nm cutoff filter.
in each case efficient release of the free acid was observed
upon photolysis.
The successful bond cleavage with electron-donating
sensitizers led us to explore bond cleavage through direct
photolysis. Although the perchlorate salts absorb only at
low wavelengths (<320 nm), it was found that the iodide
salts absorb between 350 and 450 nm (Figure 1). This is
attributed to a charge-transfer absorption of the pi-
colinium/iodide ion pair. Irradiation of such charge-
transfer bands creates a species that is a caged radical
pair consisting of radical 7 and iodine atom (Scheme 5).
Therefore, it was anticipated that photolysis of these salts
would also lead to fragmentation of the C-O bond.
Photolysis of the N-methyl-4-picolinium iodides in
methanol (>320 nm, 200 W Hg lamp) releases the free
carboxylic acids in good yields (Table 4). Using the less
polar solvents, acetonitrile (MeCN) or dichloromethane
(MeCl2) allows for irradiation at higher wavelengths.
Entry 4 in Table 4 was obtained by irradiation of ester
5c in MeCl2 above 390 nm using a Xe lamp. However, in
practice, the photolyses proceed more cleanly with MeOH
as the solvent. We attribute this to the ability of MeOH
to scavenge the radical intermediates resulting from the
initial C-O bond scission.
for ∆GET
.
∆GET ) 23.06(Eox* - Ered) + 0.06 eV
(2)
Reduction potentials, Ered, of the various N-methylpi-
colinium esters were determined by cyclic voltammetry
(CV) experiments. In each case, an irreversible reduction
wave is observed with Ered between -1050 and -1200
mV (Table 6). The Ered values were determined at a scan
rate 100 mV/s.
5550 J . Org. Chem., Vol. 69, No. 17, 2004

C-O Bond Fragmentation of 4-Picolyl- and N-Methyl-4-picolinium Esters
TABLE 6. Red u ction P oten tia ls for
N-Meth yl-4-p icolin iu m Ester s
a
compd
R
Ered (V) vs SCE
6a
6b
6c
6d
6e
6f
CH3
C6H5
CH2C6H5
C(CH3)3
CH(C6H5)2
CH2C6H4CH3
CH2C6H4Br
CHdCHC6H5
-1.09
-1.01
-1.08
-1.11
-1.17
-1.15
-1.09
-1.05
6g
6h
a
Potential at half peak current, obtained from a irreversible
wave at a scan rate of 100 mV/sec.
Thus, for the sensitizers employed in this study, ∆GET
is predicted to range between -22.1 kcal/mol (in the case
of Ph3N) and -47.7 kcal/mol (in the case of TMB), using
-1.1 V as the Ered for esters 6. Although the reorganiza-
tion energies (λ) have not been characterized for these
particular donor-acceptor pairs, these driving forces are
sufficiently large that one would expect diffusion-limited
kET values, given typical values of λ. Thus, we conclude
that the thermodynamics of this system are consistent
with the mechanism in Scheme 6.
F IGURE 2. Transient absorption spectra from pulsed laser
photolysis (355 nm, 50-100 mJ , 6 ns) of 9-MC with N-
methylpicolinium diphenylacetate perchlorate 6e in N2-purged
CH3CN. Times are in microseconds after the laser pulse.
The possibility of ground-state complexation between
picolinium salts 6 and the sensitizers led us to examine
the UV-vis spectra of mixtures of the perchlorate salts
and sensitizers. The UV spectra of the sensitizers were
found to be unchanged upon addition of increasing
amounts of esters 6. No charge-transfer bands were
observed, and the UV spectra of sensitizers showed little
change with addition of esters up to 20 mM (1-2 mM
solutions of sensitizer in 3.5 mL methanol were used).
Fluorescence quenching experiments also support the
proposed mechanism. Scheme 6 implies that electron
transfer occurs from the singlet-excited state of the
sensitizer to the picolinium ester. If this occurs, then the
fluorescence of the sensitizer should be quenched upon
addition of the picolinium substrate. To test this, the
fluorescence of the sensitizers was measured with addi-
tion of increasing concentrations of picolinium esters 6
and was observed to diminish as a function of the ester
concentration. The rate constants for fluorescence quench-
ing were found using a Stern-Volmer analysis and
published lifetimes of the sensitizers,33 to be at or near
the diffusion limit (∼1.9 × 1010 in MeCN). The rate
constant for fluorescence quenching of 9-MC by ester 6e
is 2.4 × 1010 M-1 s-1. This result indicates that the
sensitizers rapidly react with the esters upon photolysis.
The Stern-Volmer plots in a few cases show small
deviations from linearity. This observation, along with
the higher than expected rate constants for quenching,
could indicate some contribution from precomplexation
of the sensitizer with the esters, although we could not
detect such complexes by UV-vis absorption.
F IGURE 3. Transient absorption spectra from pulsed laser
photolysis (355 nm, 50-100 mJ , 6 ns) of 9-MC with 9 in N2-
purged CH3CN. Times are given in microseconds after the
laser pulse.
UV-vis spectra were obtained following pulsed laser
photolysis (355 nm, 50-100 mJ /pulse, 4-8 ns) of MeCN
solution containing both the sensitizer (9-MC) and the
picolinium ester 6e. These spectra show strong absorp-
tion bands at 410, 690, and 770 nm (Figure 2). The latter
two peaks correspond to previously characterized absorp-
tions for the 9-MC cation radicals.34 The peak at 410 nm
is assigned to the reduced N-methylpicolinium ion,
radical 7. The other esters (6a -h ) were examined using
the same LFP experiment and essentially identical
spectra are observed.
The assignment of the 410 nm peak is supported by
similar laser flash photolysis experiments carried out on
N-methylpicolinium perchlorate 9 (Figure 3). This latter
ion is expected to be reduced by excited-state sensitizers,
forming a radical similar to 7. As expected, 9 efficiently
quenched the fluorescence of 9-MC, with a rate constant
of 2.4 × 1010 M-1 s-1. However, preparative photolyses
carried out with N-methyl picolinium perchlorate 9 and
The unalkylated esters 2 likewise quench the fluores-
cence of the most reactive donor, TMB (kq ) 1.6 × 1010
M-1 s-1). The less reactive sensitizer 9-MC is also
quenched, but far less efficiently (kq ) 8.6 × 105 M-1 s-1).
The least reactive sensitizers, pyrene and Ph3N, show
no fluorescence quenching when 2 is added. This obser-
vation is consistent with our observation of no photof-
ragmentation using the latter sensitizers.
Laser flash photolysis (LFP) experiments provide
further support for the proposed mechanism. Transient
(34) Shida, T. Electronic Absorption Spectra of Radical Ions; Elsevi-
er: Amsterdam, 1988.
J . Org. Chem, Vol. 69, No. 17, 2004 5551

Sundararajan and Falvey
TABLE 7. Yield s of Acid s a n d N-Meth ylp icolin iu m Ion in th e P r esen ce of CHD
% N-methylpicolinium
compd
R
sensitizer
conditionsa
2 h, MeOH
2 h, MeOH, CHD
2 h, MeCN
2 h, MeCN, CHD
2 h, MeOH
2 h, MeOH, CHD
2 h, MeOH
% acid formedb,c
% ester consumedb,c
ion formedb,c
6c
6c
6c
6c
6f
CH2C6H5
CH2C6H5
CH2C6H5
CH2C6H5
CH2C6H4CH3
CH2C6H4CH3
CHdCHC6H5
CHdCHC6H5
CH2C6H5
9-M.C
9-M.C
9-M.C
9-M.C
9-M.C
9-M.C
9-M.C
9-M.C
pyrene
pyrene
37
86
26
47
17
44
26
42
52
94
40
94
28
46
22
49
38
57
51
99
76
43
38
43
6f
6h
6h
6c
6c
2 h, MeOH, CHD
2 h, MeOH
2 h, MeOH, CHD
CH2C6H5
80
a
b
Photolysis solution was 2.5-3.0 mL of a 4 mM solution of ester and 4.0-4.5 mM sensitizer and 20% CHD. Yields were determined
by 1H NMR integration of the carboxylic acid peak relative to an internal standard. c Estimated error ( 5%.
SCHEME 7. P ET fr om Sen sitizer to P icolin iu m
Moiety
SCHEME 8. Tr a p p in g of 4-N-Meth yl-4-p icolin iu m
Meth yl Ra d ica l 8 by CHD
TABLE 8. Qu a n tu m Yield s w ith a n d w ith ou t CHD
quantum
compd
ester
sensitizer
conditions
MeOH
MeOH, CHD
MeOH/H2O, CHD
yield (%)
9-MC showed that there is an insignificant change in the
solution composition over several hours of photolysis.
Thus, we conclude that 9 undergoes reversible electron
transfer from the sensitizer to form the N-methyl-4-
picolyl radical 8 (Scheme 7). The transient absorption
spectra obtained by LFP on 9 with 9-MC as sensitizer
show the 9-MC cation radical peaks at 690 and 770 nm
and a peak at 410 nm as with the esters 6. Absorption
maxima of similar radicals in the literature also support
this assignment.35
LFP experiments were carried out on esters 6 using
the other sensitizers, TMB, pyrene, and TPA as well. In
all cases, the observed transient signals agree with the
known spectra for the cation radicals of the sensitizers.34
With 9-MC, pyrene, and TPA, we were able to detect the
expected 410 nm peak for the reduced ester. In the case
of TMB, the strong absorbance of the TMB cation radical
at 470 nm obscures the expected 410 nm peak for the
reduced ester.
The unalkylated esters 2 were also examined using
LFP. However, we were limited to using TMB as the
sensitizer since the other three sensitizers are unable to
reduce the ester. No transient peak corresponding to
anion radical 3 could be detected. The anion radical of
pyridine is known to have a weak absorption around 330
nm,34 and if radical 3 has a similar absorption, then
under our LFP conditions we would be unable to observe
it.
The formation of the carboxylic acids in the sensitized
photolysis experiments leaves little doubt that the pre-
dicted C-O bond scission reaction occurs following PET.
However, we were unable to detect the N-methyl-4-
picolinium methyl radical 8. To our knowledge, the
absorption spectrum of this radical is unknown. Attempts
to independently generate and detect radical 8 through
LFP (266 nm, 5-15 mJ , 4-6 ns) experiments on N-meth-
yl-4-chloropicolinium perchlorate were unsuccessful. The
6c
6c
6c
CH2C6H5
CH2C6H5
CH2C6H5
9-MC
9-MC
9-MC
0.23
0.39
0.36
•
benzyl radical (PhCH2 ) is known to absorb near 320
nm.36 Thus, it is possible that radical 8 absorbs in a
similar region and its signal is obscured by absorptions
of the sensitizer and/or the N-methyl-4-picolinium group.
To verify the formation of 8, we carried out trapping
experiments using 1,4-cyclohexadiene (CHD). This latter
additive is a good H atom donor, having a C-H bond
dissociation energy of 76.0 kcal/mol.37 Thus, it was
anticipated that CHD would efficiently trap radical 8 as
the N-methylpicolinium ion 9 (Scheme 8). This proved
to be the case. As shown in Table 7, good yields of the
radical product 9 are observed when CHD is included in
the photolysis solutions. In the absence of CHD, 9 was
1
not detected in the H NMR of the photolysis solutions.
The fate of radical fragment 8 under these conditions is
not known with certainty. The absence of a single
significant byproduct suggests that this radical decays
through a variety of pathways.
Another interesting effect of CHD is the apparent
increase in the efficiencies of the sensitized photolysis
reported in Table 7. Under comparable, although not
rigorously identical, photolysis conditions, a ca. 2-fold
increase in the absolute yield of carboxylic acid is
observed when CHD is included. This is accompanied by
a corresponding increase in the amount of ester that is
converted. In the case of ester derivative 6c this effect
was verified by more accurate quantum yield measure-
ments. As shown in Table 8, the quantum yield for ester
conversion increases from 0.23 in the absence of CHD to
0.39 when the trap is added.
(36) Turro, N. J .; Chow, M.-F.; Chung, C.-J .; Tanimoto, Y.; Weed,
G. C. J . Am. Chem. Soc. 1981, 103, 4574-4576.
(37) Luo, Y.-R. Handbook of Bond Dissociation Energies in Organic
Compounds; CRC Press: Boca Raton, 2003.
(35) Hermolin, J .; Levin, M.; Kosower, E. M. J . Am. Chem. Soc. 1981,
103, 4808-4813.
5552 J . Org. Chem., Vol. 69, No. 17, 2004

C-O Bond Fragmentation of 4-Picolyl- and N-Methyl-4-picolinium Esters
[PF6] as the supporting electrolyte. The electrodes used were
SCHEME 9. Secon d a r y Red u ction by CHD
Ra d ica l
a carbon working, a platinum auxiliary, and an Ag/AgCl
reference. Ferrocene was used as the internal standard. The
CV data were taken in spectroscopic grade dry MeCN after
N2 purging the samples for 10 min. The ferrocene/ferrocenium
couple was found around 590 mV and the typical scan speed
was 100 mV/s.
F lu or escen ce Qu en ch in g Exp er im en ts. Fluorescence
quenching experiments were performed on an luminescence
spectrometer. Samples in MeCN solvent were placed in 1 cm
quartz cuvettes, sealed with rubber septums and purged with
N2 for 10-15 min. Sample concentrations were prepared such
that the optical densities at the excitation wavelength were
about 0.1-0.4.
SCHEME 10. Qu en ch in g of 9-MC Ca tion Ra d ica l
by CHD
F lu or escen ce Lifetim e Mea su r em en t. Fluorescence life-
times were determined by measuring the fluorescence decay
on a nanoflash fluorescence apparatus. Samples in MeCN
solvent were placed in 1 cm quartz cuvettes, sealed with rubber
septums, and purged with N2 for 10-15 min. Sample concen-
trations were prepared such that the optical densities at the
excitation wavelength were about 0.1-0.4. The fluorescence
lifetime of the sensitizer TPA was determined by this proce-
dure.
Qu a n tu m Yield Deter m in a tion . The solutions were pho-
tolyzed by the output of a 1000 W Hg-Xe lamp after passing
through a spectral energy monochromator set to 355 nm ( 10
nm. Solutions of ester and sensitizer (3.3 mM, OD at 355 about
0.4-0.6) were placed in a 1 cm quartz cuvette and purged with
N2 for 10-15 min. The samples were irradiated for different
times, 1.5, 2, 3, and 4 h and the amount of ester converted
was determined by 1H NMR integration. Light intensities were
measured by a radiometer.
Dep r otection P h otolysis. A solution of 4-10 mg of the
picolinium ester and an equal amount or slight excess of
sensitizer was prepared in about 5 mL of CH3OH (4.0 mM of
ester and 4.0-4.5 mM of sensitizer). A 1 mL aliquot from this
solution was used as dark control. The solution was purged
with nitrogen and irradiated using a 200 W Hg lamp with
continuous stirring. A Pyrex filter was used to cut off light
below 320 nm to ensure light absorption by the sensitizer
alone. A 1 mL aliquot was drawn out of the photolyzed
solution. The solvents in the aliquots were evaporated, and
the residues were redissolved in CD3OD and hexamethyldisi-
loxane was added as the internal standard. The percent yields
were determined by 1H NMR integration of the carboxylic acid
peaks relative to the internal standard.
One explanation for this is a secondary, nonphoto-
chemical reduction step carried out by the 1, 4-cyclo-
hexadienyl radical byproduct of the H atom transfer
reaction. It is possible that this radical reduces an
unreacted N-methylpicolinium ester 6, creating the same
intermediate radical 7 generated in the PET step (Scheme
9).
The presence of CHD has another interesting effect.
The sensitizer (9-MC) cation radical is quenched by this
trap (Scheme 10). Repeating this experiment with vary-
ing concentrations of CHD and observing the pseudo-
first-order decay of the 9-MC cation radical provides a
rate constant of 5 × 105 M-1 s-1 for this process. Although
this process has not been studied in detail, it likely
involves H atom transfer to carbazole, followed by depro-
tonation of the highly acidic carbazole conjugate acid.
Such a quenching process would prevent 9-MC cation
radicals from engaging in back electron transfer with 7.
This suppression of back electron transfer would also
increase the quantum yield for the PET process. The
relative contribution of this effect and the CHD radical
pathway shown in Scheme 9 has yet to be determined.
Con clu sion s
In some cases, 4-10 mg each of the picolinium ester and
the sensitizer was dissolved in 5 mL of CH3OH. Half of the
solution was used as dark control, and the remaining half was
N2 purged and photolyzed. The solutions were then evaporated
and redissolved in CD3OD with hexamethyldisiloxane as the
internal standard and 1H NMR spectra were obtained. The
yields were determined as before.
The experiments described herein demonstrate that
the 4-picolyl- and 4-picolinium esters can function as
electron acceptors in PET reactions. It is further dem-
onstrated that the resulting (anion) radicals fragment to
give a carboxylate anion and pyridylmethyl radical. The
high chemical and quantum yields from these fragmen-
tation reactions suggest that such reactions will find use
in photorelease applications. These reactions have been
carried out by sensitization and as well as by direct
photolysis in the case of the iodide salts. The mechanism
of the photorelease has been investigated, and the data
support the proposed PET based mechanism.
La ser F la sh P h otolysis. Laser flash photolysis experi-
ments were performed using an Nd:YAG laser as the pump
beam source. The laser used was capable of 266 or 355 nm
pulses between 4 and 6 ns duration. A 350 MHz digital
oscilloscope was used to observe the waveforms. The samples
were prepared such that the absorbances were between 1.5
and 2.0 at the excitation wavelength, 355 nm. The samples
were placed in 1 cm quartz cuvettes, N2 (or O2) purged for 15
min, and stirred continuously during the photolysis.
Gen er a l P r oced u r es for th e Syn th esis of Ester s 2a -
h . Meth od 1. Esters 2a -d and 2h were prepared using this
procedure.32 4-Pyridylcarbinol (3.45 g, 31.6 mmol) was dis-
solved in 40 mL of benzene, and triethylamine (7.97 mL, 56.8
mmol) was added. A solution of the corresponding acyl chloride
(50.5 mmol) in 15 mL of benzene was added dropwise. The
reaction mixture was stirred at room temperature for 16 h.
H2O (30 mL) was added, and the organic layer was separated,
dried over MgSO4, and concentrated in vacuo.
Exp er im en ta l Section
1
Gen er a l P r oced u r es. All H and 13C NMR were obtained
using a 400 MHz spectrometer. Chemical shifts (δ) are
reported in parts per million (ppm) and coupling constants (J )
in hertz (Hz). Deuterated CD3CN or CD3OD were used for all
NMR experiments. Acetonitrile was distilled from calcium
hydride under nitrogen atmosphere. UV-vis spectra were
recorded on a spectrophotometer.
Cyclic Volta m m etr y Exp er im en ts. All electrochemical
experiments were done on a voltammetry analyzer with [Bu4N]-
Meth od 2. Esters 2e, 2f, and 2g were prepared using this
J . Org. Chem, Vol. 69, No. 17, 2004 5553

Sundararajan and Falvey
procedure.28 N,N′-Dicyclohexylcarbodiimide (DCC) (3.40 g, 16.4
mmol) was added to a solution of the carboxylic acid (16.4
mmol) and 4-pyridylcarbinol (1.62 g, 14.8 mmol) in 50 mL of
dichloromethane. The reaction mixture was stirred at room
temperature for about 24 h and checked by TLC for disap-
pearance of starting material. The urea precipitate was filtered
off, and the filtrate was washed with NaHCO3 (2 × 25 mL)
and H2O (2 × 10 mL). The organic layers were collected and
dried over MgSO4, and the solvent was evaporated.
Gen er a l P r oced u r e for th e Syn th esis of Meth yla ted
Ester s 5a -h . A 0.02 mol portion of the corresponding picolyl
ester was dissolved in 15 mL of CH3OH, and methyl iodide
(4.26 g, 0.03 mol) was added slowly. The reaction mixture was
refluxed at 75 °C for 8 h. The solvent was removed under
reduced pressure, and the residue was recrystallized from hot
methanol or ethanol.
Gen er a l P r oced u r e for Cou n ter -Ion Exch a n ge. The
picolyl methiodide (5 mmol) was dissolved in a minimum
amount of acetonitrile. A solution of silver perchlorate (1.04
g, 5 mmol) in acetonitrile was added, and the reaction mixture
was stirred overnight at room temperature. The yellow AgI
precipitate was filtered off, and the filtrate was concentrated
in vacuo and recrystallized from hot ethanol.
Ack n ow led gm en t. This work was supported by the
National Science Foundation.
1
Su p p or tin g In for m a tion Ava ila ble: Copies of H NMR
spectra of 2a -h , 5a -h , 6a -h and 9, fluorescence data, LFP
spectra, and UV spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
J O049501J
5554 J . Org. Chem., Vol. 69, No. 17, 2004