Photoprocesses of molecules with 2-nitrobenzyl protecting groups and caged organic acids

Article abstract of DOI:10.1111/j.1751-1097.2007.00215.x

The 308 nm photoinduced formation of the nitroso product and the intermediacy of the aci-nitro form(s) were studied for a series of 2-nitrobenzyl alkyl and aryl esters (1a-4e) and bis-(nitrophenyl)methyl acetates (5a-6b) by time-resolved UV-vis spectroscopy. A triplet state appears as major transient, when 2-nitrobenzyl derivatives 1 are substituted by 4,5-dimethoxy (2) and 4,5-methylenedioxy (3/4) groups. This triplet of charge transfer character is, however, not part of the route via the aci-nitro into the 2-nitroso form. The activation energy and preexponential factor of the longest lifetime component (τ_aci), i.e. the major part of the aci-nitro decay, were determined. The carboxylic acids as leaving groups have rather small effects on τ_aci. An additional nitrated phenyl ring in α-position (5) leads generally to shorter τ_aci value. Otherwise, the photogeneration of nitroso products is similar. The quantum yield (Φ_d) varies only moderately with structure, the yield of the aci-nitro form and Φ_d are correlated and little affected by solvent properties.

Full text of DOI:10.1111/j.1751-1097.2007.00215.x

Photochemistry and Photobiology, 2008, 84: 162–171
Photoprocesses of Molecules with 2-Nitrobenzyl Protecting Groups and
Caged Organic Acids
1
Filiz Bley , Klaus Schaper and Helmut Gorner*
1
2
¨
1
¨
¨
Institut fu¨r Organische Chemie und Makromolekulare Chemie, Universitatsstr. 1, Dusseldorf, Germany
2Max-Planck-Institut fu¨r Bioanorganische Chemie, Mu¨lheim an der Ruhr, Germany
Received 5 April 2007, accepted 24 August 2007, DOI: 10.1111 ⁄ j.1751-1097.2007.00215.x
N-(2-nitrobenzyl)ureas (R: H, Me, carboxy) urea release occurs
ABSTRACT
with quantum yields of 0.4–0.8; kaci = 440 nm and saci = 10 ls
for R: carboxy (15). Radicals, observed upon photolysis of
different 2-nitrobenzyl-caged-ATP derivatives in the 1 s range,
were proposed to be generated by electron transfer from the
aci-nitro anion (A)) to the nonexcited substrate (16).
The 308 nm photoinduced formation of the nitroso product and
the intermediacy of the aci-nitro form(s) were studied for a series
of 2-nitrobenzyl alkyl and aryl esters (1a–4e) and bis-(nitrophe-
nyl)methyl acetates (5a–6b) by time-resolved UV-vis spectros-
copy.
A triplet state appears as major transient, when
The photochemical properties of 2-nitrobenzyl acetate (1a)
and several bis-nitrophenyl acetates were studied (12). Photol-
ysis of 2-nitrobenzyl benzoate (1b) yields the aci-nitro form as
intermediate and 2-nitrosobenzaldehyde and benzoic acid as
products with F = 0.08 (6). Excited 4,5-dimethoxy-2-nitrob-
enzyl acetate (2a) has been used as electron donor to photore-
duce cytochrome C (7). Recently, for 2-nitrobenzyl alcohol and
1-(2-nitrophenyl)ethanol a novel pathway via the nitroso
hydrate without cyclization has been established (21). This
mechanism should not contribute to the photoreaction ob-
served, because here the transfer of an acylium cation instead of
a proton would be necessary. The knowledge concerning
2-nitrobenzyl ethers and carboxylates is still rather limited. It
seemed therefore desirable to obtain a deeper insight into the
photoinduced reaction of the 2-nitrobenzyl moiety (Chart 1).
This article aims at the photoreactions of 2-nitrobenzyl-
based derivatives into 2-nitroso compounds and the leaving
group HX, which is acetic acid for the acetates (a: 1, 2, 3 and
4), benzoic acid for the benzoates (b) and 4-toluic acid for the
toluates (c). The effects of 4,5-dimethoxy (2) and 4,5-methy-
lenedioxy (3,4) substituents on the photoinduced generation of
the aci-nitro and the nitroso forms were examined, whereby
2-(4-aminobutyryloxy)-2-(4,5-methylenedioxy-1-nitrophen-2-yl)
acetic acid (4d) and 2-(4-aminoglutaryloxy)-2-(4,5-methylen-
edioxy-1-nitrophen-2-yl)acetic acid (4e) contain amino acid
building blocks, which are known to be important neurotrans-
mitters. Moreover, bis-nitrophenyl acetates (5, 6) were studied.
For biological application the chromophore of the protecting
group should absorb at wavelengths above 300 nm, such that
the bioactive material is not photochemically damaged.
Therefore compounds of type 2–4 were designed to have
absorption maxima at 340–360 nm.
2-nitrobenzyl derivatives 1 are substituted by 4,5-dimethoxy (2)
and 4,5-methylenedioxy (3 ⁄ 4) groups. This triplet of charge
transfer character is, however, not part of the route via the aci-
nitro into the 2-nitroso form. The activation energy and
preexponential factor of the longest lifetime component (saci),
i.e. the major part of the aci-nitro decay, were determined. The
carboxylic acids as leaving groups have rather small effects on
saci. An additional nitrated phenyl ring in a-position (5) leads
generally to shorter saci value. Otherwise, the photogeneration of
nitroso products is similar. The quantum yield (Fd) varies only
moderately with structure, the yield of the aci-nitro form and Fd
are correlated and little affected by solvent properties.
INTRODUCTION
2-Nitrobenzyl-caged compounds have become important as
the irreversible photoreaction of the protecting group can be
used to trigger the release of an appropriate biochemically
active substrate (1–7). A large number of photolabile protect-
ing groups with the 2-nitrobenzyl function have been devel-
oped that can release bioactive compounds rapidly upon
photolysis with UV ⁄ visible light (8). The photoreaction is
triggered by intramolecular benzylic H-atom transfer to the
excited nitro group. Efficient formation of the aci-nitro form is
known for several cases (5,9–15).
The photoreaction of 2-nitrobenzyl (HXCRArNO2 in
Scheme 1) into 2-nitroso compounds has been extensively
studied (16–28). 2-Nitrosobenzenes (OCRArNO) are the prod-
ucts and the aci-nitro form (denoted as HA ⁄ A)) is the major
intermediate in this photoreaction, whereas the cyclic benzis-
oxazoline (HB ⁄ B)) and the hemiacylal (HC ⁄ C)) has not been
observed in most cases. The aci-nitro form is characterized by
an absorption maximum (kaci) at 400–440 nm and a lifetime
(inverse rate constant: saci) in the ls–ms range. Eventually,
the leaving group (HX) is released. For different a-substituted
MATERIALS AND METHODS
The synthesis of a 2-nitrobenzyl acetate containing a rigid amino group
(7) and the other 18 compounds is described in detail elsewhere (27).
The sample purity was ascertained by HPLC analyses, for which a
125 · 4.6 mm Inertsil C18 5 lm column was used and MeOH-water
(2:1, vol) as eluent, flow rate: 0.8 mL min)1. The solvents, e.g.
*Corresponding author email: goerner@mpi-muelheim.mpg.de (Helmut
Gorner)
ꢀ 2007 The Authors. JournalCompilation. The AmericanSocietyofPhotobiology 0031-8655/08
162

Photochemistry and Photobiology, 2008, 84 163
Scheme 1.
Chart 1. 1: 2-nitrobenzyl compounds, 2: 4,5-dimethoxynitrobenzyl compounds, 3: 4,5-methylenedioxynitrobenzyl compounds, 4: a-carboxy-4,5-
methylenedioxynitrobenzyl compounds; (a) acetic acid esters, (b) benzoic acid esters and (c) toluic acid esters.
)1
dimethylsulfoxide: DMSO (Fisher), ethanol and acetonitrile (Merck,
Uvasol), were used as received; water was purified by a milliQ system.
The molar absorption coefficients of 1a, 1b and 1c in acetonitrile at
e335 = 0.6 · 104
M
cm)1
monitored on two spectrophotometers,
.
The UV–Vis absorption spectra were
diode array (Hewlett
a
Packard, 8453) and one from Perkin Elmer (Lambda 19). For
photoconversion the 254 nm line of an Hg lamp was utilized,
absorbances of 0.8–2.5 were used for kirr = 254 nm. The quantum
)1
254 nm are e254 = 0.8 · 104, 1.0 · 104 and 1.4 · 104
M
cm)1
,
respectively. The long-wavelength band of 2a–3c has typically

164 Filiz Bley et al.
yield of decomposition Fd was determined using optically matched
solutions (A254 = 1.5–2) and the uridine ⁄ water ⁄ oxygen actinometer
(29). For a few cases, e.g. in DMSO, the 313 nm line of a 1000 W Hg
lamp, a monochromator and the aberchrome 540 actinometer were
used (30). The retention times under our conditions were 3.0, 8.8, 14.0,
2.8, 8.4, 12.6, 3.2, 10.6 and 16.2 min for 1a–3c, respectively, 10.0 min
for 4b and 10.4 min for 5a. An excimer laser (Lambda Physik,
Gottingen, Germany; EMG 200, pulse width of 20 ns and energy
<100 mJ) was used for excitation at 308 nm. The absorption signals
were measured with two digitizers (Tektronix 7912AD and 390AD).
ethanol at the end of the 308 nm pulse has a maximum at
kaci = 410 nm. The decay follows essentially first-order kinet-
ics (rate constant: 1 ⁄ saci) with two components or one,
depending on the solvent. The spectra of the acetic acid 1a
and toluic acid 1c esters in acetonitrile (Fig. 1a) are similar.
For 1a–1c in dichloromethane and water-free polar solvents
the decay kinetics could not be fitted mono-exponentially in
most cases, but the major decay is the longer lived component
with saci in the 1–10 ms range. A triplet nature is excluded on
the basis that triplets of unsubstituted 2-nitrobenzyl com-
pounds have sub-ns lifetimes (1) and oxygen does not affect the
decay. The transient is therefore assigned to the aci-nitro
structure. The two components of the aci-nitro forms of 1c in
acetonitrile have saci values of 5 ms (70%) and 0.05 ms and
deprotonation of the aci-nitro form into A) has been excluded
for 1b in acetonitrile by conductivity studies (6). A biphasic
decay with saci in the 0.01–0.1 s range was also found for
a-carboxy-4,5-methylenedioxynitrobenzyl compounds 4a–e in
DMSO ⁄ water (Table 1).
Absorbances of 0.3–1.5 at kexc were used, corresponding to [1a] of 0.1–
max
0.5 mM. The DAT, values were measured with respect to DAT
M
= 1
cm)1) for the triplet state of benzophenone
)1
(e520 · Fisc = 6 · 103
in acetonitrile. The respective experimental errors of Ea and Fd are
8
and 12% and those of sT and saci are typically 20%. Phospho-
rescence of singlet molecular oxygen at 1269 nm was detected as
described previously (31). Oxygen concentrations of 2.2, 9.1 and
1.3 mM were used for saturation values in DMSO, acetonitrile and
water, respectively. The quantum yield (FD) of formation of O2(1Dg)
was obtained using optically matched solutions (A308 = 0.8) and
ref
phenazine as reference FD = 0.9 (32).
RESULTS AND DISCUSSION
On the other hand, when water was added to DMSO,
acetonitrile or ethanol, a faster decay at kaci = 410 nm and a
grow-in due to conversion into a secondary transient with
maximum at 500 nm was detected. The conversion into the
500 nm transient is shown in Fig. 2a for 1b in wet DMSO.
A major aim of this study was to better understand the
function–structure relationship concerning the formation and
decay of aci-nitro forms (HA ⁄ A)) and the release of the
leaving group. HX is acetic acid (a: 1–4), benzoic acid (b: 1–4)
and methyl-benzoic acid (c: 1–4). The effects of 4,5-dimethoxy
(2) and 4,5-methylenedioxy (3,4) groups and an additional
nitrophenyl ring (5,6) were examined. A classification may be
the following: Compounds of group 1 have no 4,5-substitu-
ents, of groups 2 ⁄ 3 have 4,5-dimethoxy and 4,5-methylenedi-
oxy substituents and of group 4 are similar, but contain a
carboxy group and for 4d and 4e additionally an amino acid
building block. Two nitro groups without and with a carboxy
group constitute compounds of groups 5 and 6, respectively.
All 2-nitrobenzyl-caged compounds follow a general pattern
except for a 2-nitrobenzyl acetate 7 which was found to be
photochemically inert. This is almost to be expected for
substitution by an amino group.
Table 1. Lifetime saci (ms) of the longer lived aci-nitro transient.*
Comp. CH2Cl2
MeCN
DMSO DMSO-H2O
EtOH
1a
1b
1c
3b
4a
4b
4c
4d
4e
5a
5b
6a†
6b
3
2
3
<1 ⁄ 7
6
50
60
90
100
>90
>50
1
4
4
14
14
8
0.03 ⁄ 6
5
0.01 ⁄ 4
0.03 ⁄ 3
30
30
40
5 ⁄ 50
5 ⁄ 60
5 ⁄ 80
5 ⁄ >50
3 ⁄ >50
0.01
140
100
160
12
0.015
0.01
0.02
0.02
0.001 ⁄ 50
0.03
0.03
0.001 ⁄ 200
0.2 ⁄ 3
0.012
0.016
1.2
0.01
3
5
Transients
*At 25ꢁC using kexc = 308 nm, two transients as indicated. †See
Figs. 7 and 8 and text.
The transient absorption spectrum of 2-nitrobenzyl benzoate
1b in dichloromethane (Fig. 1b), DMSO, acetonitrile or
Figure 2. Transient absorption spectra of 1b in (a) DMSO-H2O (10:1)
Figure 1. Transient absorption spectra of (a) 1a in acetonitrile and
)
and (b) DMSO-H2O (1:1) at 20 ns (s), 10 ls (h), 0.1 ms (d), 1 ms ( )
and 10 ms ( ) after the 308 nm pulse; insets: kinetics at 400 nm (left)
and 480 nm (right).
(b) 1b in dichloromethane at 20 ns (s), 1 ls (D), 0.1 ms (d), 1 ms (
and 10 ms ( ) after the 308 nm pulse; insets: kinetics at 400 nm.

Photochemistry and Photobiology, 2008, 84 165
)1
Addition of 50% water does not significantly change the yield,
but redshifts kaci after a few ls and decreases the lifetime
(Fig. 2b). A redshift of kaci within a few microseconds, due to
deprotonation of the aci-nitro form into A), has been reported
for 2-nitrophenylethyl acetate-caged compounds in a 1:1
acetonitrile–water mixture (22). The presence of water in polar
solvents should favor the deprotonation. The two components
of the aci-nitro forms of 1a in DMSO-phosphate buffer (1:1)
each contribute 50% and have saci values of 0.05 and 0.36 ms,
which is in good agreement with earlier results of 0.07 and
0.65 ms (68%) in water (25). Formation of the aci-nitro form
with similar properties occurs gradually also for 1a and for 1c
(Table 1). In this work, the curves for several compound–
solvent combinations were fitted by one or two first-order
components. For 1b in acetonitrile the first-order fits were
appropriate for analysis of the data presented, but at higher
laser intensity, e.g. larger than 1 MWcm)2 or 25 mJ per pulse
(which was avoided), a mixed first- and second-order decay
was found. Inspired by the results for 2-nitrobenzyl alcohol in
hexane, where a deprotonation of the aci-nitro form was
attributed to a molecule of the starting material (21), we take
this idea a step further and propose that a second-order
process is due to an encounter of two aci-nitro forms,
reminiscent to the case of T-T annihilation. Such a reaction
with a proton transfer catalyst can, in principle, take place in
any solvent.
kox = (0.9–1.5) · 109
controlled limit.
M
s)1, i.e. slightly below the diffusion-
To illustrate characteristic examples, transient absorption
spectra and typical kinetics are shown in Figs. 4–7. For
Figure 3. Transient absorption spectra in acetonitrile (under air) of (a)
2b and (b) 3a at 20 ns (s), 1 ls (D) and 10 ls (h) after the 308 nm
pulse; insets: kinetics at 500 nm (argon-, air- and oxygen-saturated:
upper to lower, respectively).
The transient absorption spectra of 2a–3c in acetonitrile are
much broader and the decay is rather short-lived (Fig. 3a,b and
Table 2). The major difference is the effect of quenching by
oxygen. Similar results were found in dichloromethane, chlo-
roform, DMSO or ethanol. Therefore we propose to assign this
transient to a triplet state. This assignment is supported by the
observed formation of singlet oxygen. The lifetime (inverse
first-order rate constant: sT) in these solvents and DMSO-water
is in the 0.4–2 ls range, i.e. much shorter than typical saci
values. As accompanying weak signal the longer lived aci-nitro
form was also obtained. The ratio of absorbances of the triplet
state at the pulse end (DAT) and the aci-nitro form at a few ls
(DAaci) is not changed by the presence or absence of oxygen, i.e.
the triplet state is not the precursor of the aci-nitro form. The
alternative that triplet quenching by oxygen also populates
the aci-nitro form is rather unlikely. The rate constant for
triplet quenching of 2a–3c in acetonitrile by oxygen is
Figure 4. Transient absorption spectra of 4b in argon-saturated (a)
acetonitrile and (b) DMSO-H2O (1:1, pH 4) at 20 ns (s), 10 ls (h),
10 ms ( ) and 100 ms (¤) after the 308 nm pulse; insets: kinetics at
400 nm (upper) and 500 nm (lower).
Table 2. Lifetime (ls), relative efficiency of the triplet transient and quantum yields of O2(1Dg) formation.*
Comp.
sT CH2Cl2
sT DMSO
sT MeCN
sT DMSO-H2O†
DAT‡
FD§ CH2Cl2
2a
2b
2c
3a
3b
3c
4a
4b
4c
4d
4e
0.7 (0.3)||
1.0 (0.2)
1.1 (0.3)
0.4 (0.2)
0.4 (0.2)
0.4 (0.2)
0.7 (0.3)
0.8 (0.3)
0.7 (0.3)
1 (0.4)
1 (0.4)
1 (0.4)
0.6 (0.3)
0.8 (0.3)
0.6 (0.3)
1
2 (0.5)
1 (0.3)
2 (0.3)
2 (0.3)
3 (0.3)
2 (0.3)
3 (0.3)
3 (0.3)
2
0.3
0.4
0.7
0.4
0.5
0.4
0.8
0.95
0.9
0.9
0.95
1
0.3
0.3
0.4
0.3
0.3
0.5 [0.2]
0.4 [0.2]
0.4 [0.2]
0.3 [0.1]
0.3 [0.1]
0.5 [0.2]
<0.1
>2
>2
1
1
1
1
*Using kexc = 308 nm, sT refers to argon-saturated solution. †1:1 mixture. ‡Relative value in DMSO at 440 nm and the pulse end; similar values
for 2a–4c in dichloromethane, acetonitrile and DMSO-H2O (1:1). §Values outside and within brackets refer to oxygen- and air-saturated solution,
respectively. ||Values in parentheses refer to oxygen-saturated solution.

166 Filiz Bley et al.
The aci-nitro form was also detected as major long-lived
transient. For 4d in aqueous solution at pH 3–4 (absence of a
buffer) or at pH 7 (phosphate buffer) the changes in DAaci are
small (data not shown). The aci-nitro form of 5b (Fig. 6a) can
be considered as a clean example with only one component,
while in DMSO-water a conversion into another aci-nitro form
is indicated (Fig. 6b). 6a is more complex in that the aci-nitro
form is converted into another in which kaci is redshifted
(Fig. 7). Moreover, the presence of water has a large effect, i.e.
a much shorter saci value. This effect is new and points to a
water-mediated behavior in contrast to that of most other
cases, e.g. for 1a–c or 5a ⁄ b. Plots of DA420 vs log time are
shown in Fig. 8 for 6a in acetonitrile, DMSO and 1:1 mixtures
with H2O. The DAaci values as a measure of e410 · Faci of 1–3
or 4a–e in acetonitrile are 0.05–0.1 and 0.3 (Table 3), respec-
tively.
The results so far refer to room temperature and almost
nothing is known concerning the effect of temperature on the
kinetics of the 2-nitrobenzyl function. The saci values of 1a–6b
become shorter at elevated temperatures and semilogarithmic
plots of 1 ⁄ saci vs 1 ⁄ T are linear (Fig. 9). In those cases, where a
second shorter lived saci component was detected at room
temperature, its contribution became smaller or not observable
Figure 5. Transient absorption spectra of 4d in (a) DMSO and (b)
DMSO-H2O (1:1, pH 4) at 20 ns (s), 10 ls (h), 10 ms ( ) and 100 ms
(¤) after the 308 nm pulse; insets: kinetics at (a) 400 ⁄ 500 nm and (b)
420 nm.
Figure 6. Transient absorption spectra of 5b in (a) DMSO and (b)
DMSO-H2O (1:1, pH 7) at 20 ns (s), 1 ls (D), 10 ls (h) and 0.1 ms
(d) after the 308 nm pulse; insets: kinetics at (a) 450 nm and (b) 340
and 420 nm.
Figure 7. Transient absorption spectra of 6a in (a) DMSO and (b)
DMSO-H2O (1:1, pH 4) at 20 ns (s), 1 ls (D), 10 ls (h), 1 ms ( ) and
100 ms (¤) after the 308 nm pulse.
a-carboxy-4,5-methylenedioxynitrobenzyl compounds 4a–c
(Fig. 4) as well as for 4d (Fig. 5) and 4e the triplet state
appears as a major short-lived transient. For the acetic acid
ester 4a in DMSO with sT = 0.6 ls the shoulder in T-T
absorption is replaced by a maximum at 530 nm. Observation
of the phosphorescence of singlet molecular oxygen, O2(1Dg),
at 1269 nm for 2a–3c and the absence for 1b in oxygen- and
air-saturated dichloromethane (and for 2a–3c under argon) is
further evidence for the triplet nature in these cases. The
quantum yield FD, which is a minimum for that of intersystem
crossing (Fisc), is substantial, FD = 0.3–0.5 (Table 2). The
enhancement of FD on going from air- to oxygen-saturated
dichloromethane is due to the short sT of 0.4–1.2 ls which,
under air, leads to more deactivation without quenching. It is
worth mentioning that for the photoconversion of 1-nitro-2-
naphthaldehyde into the nitrosonaphthoic acid it has been
reported that UV–Vis spectroscopy does not reveal any
observable aci-nitro form but a triplet state with sT in the ns
range (33).
Figure 8. Plots of DA460 vs log time for 6a in acetonitrile (a: circles),
DMSO (triangles), dry (b: open) and 1:1 mixtures with H2O (c: full).

Photochemistry and Photobiology, 2008, 84 167
)1
Table 3. Molar absorption coefficient (103 · M cm)1), relative effi-
ciency and quantum yield of conversion.*
Table 4. Activation energy (in kJ mol)1) and preexponential factor
(in s)1).*
Comp.
e254 ⁄ e308
Fd†
0.1
0.12 (0.15)§
0.12
0.08
0.1
0.1
0.1
0.1
DAaci
‡
Comp. MeCN: Ea log A0 DMSO: Ea log A0 EtOH: Ea log A0
1a
1b
1c
2a
2b
2c
3a
3b
3c
4a
4b
4c
4d
4e
5a
5b
6a
6b
7
7.8 ⁄ 2.0
10.3 ⁄ 2.5
15.2 ⁄ 3.1
10.7 ⁄ 6.3
11.0 ⁄ 5.7
18.1 ⁄ 6.7
11.0 ⁄ 4.5
13.5 ⁄ 4.7
19.3 ⁄ 5.1
14.4 ⁄ 6.6
16.2 ⁄ 7.1
20.7 ⁄ 8.7
15.0 ⁄ 5.7
15.3 ⁄ 5.1
15.8 ⁄ 5.4
16.1 ⁄ 5.9
21.1 ⁄ 5.2
33.3 ⁄ 6.6
13.4 ⁄ 4.2
0.1
0.1
0.1
<0.08
<0.08
<0.08
<0.08
<0.08
<0.08
0.3
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.11
1a
1b
1c
3b
4a
4b
4c
4d
5a ⁄ b
6a
6b
48
52
50
11.1
10.8
10.9
50
51
49
51
50
50
50
10.9
10.6
10.6
10.7
9.6
50
50
50
10.3
10.3
10.4
50
50
48
10
10
9.8
51
52
51
52
50
50
51
9.2
9.4
9.2
10.3
13.2
11.3
10.7
9.4
9.4
0.08
0.23
50
13
50
48 (44)†
52 (46)
12.9
9.0
8.5
0.2
0.2
0.2
0.2
0.2
*From plots of log 1 ⁄ saci vs 1 ⁄ T, cf. Fig. 9. †Values in parentheses
refer to DMSO-water (1:1).
increased dipole compared to the aci-nitro-intermediate. It is
noteworthy that values in a similar range, A0 = 1011–1013 s)1
and Ea = 48–54 kJ mol)1
, were found for 2-nitrobenzyl
<0.02
<0.02
alcohol and two derivatives in the same solvents (26). For 1-
(2-nitrophenyl)ethyl ether at pH 8.5 a much smaller value of
Ea = 25 kJ mol)1 has been reported (17).
*In air-saturated acetonitrile, except where indicated. †Using
kirr = 254 nm, see Fig. 10. ‡Relative value at 410 nm and 10 ls after
the pulse. §In DMSO-water.
Continuous irradiation
UV irradiation in acetonitrile (air- or argon-saturated at room
temperature) leads to substrate decomposition, as registered by
HPLC for the cases of 1a–4e. The spectra of nitrobenzyl
compounds 1a–c are characterized by two characteristic
wavelengths (quasi-isosbestic points) at 240 and 290 nm,
absorption decreases between and increases in the range above
300 nm. For 4,5-dimethoxynitrobenzyl 2 and 4,5-methylenedi-
oxynitrobenzyl compounds 3 the quasi-isosbestic points are
shifted to 320 and 370 nm (27). The spectral changes in 2a–4e
are, in principle, similar and redshifted with respect to
methoxy nonsubstituted 1.
The quantum yield of decomposition (Fd) was obtained
either from the increase in absorbance at an appropriate
wavelength, e.g. 260 nm (Fig. 10a), or the decrease in the
Figure 9. Plots of log 1 ⁄ saci vs 1 ⁄ T for 1b (circles), 4b (triangles), 5b
(squares) and 6a (diamonds) in DMSO (open) and ethanol (full).
(in yield) at elevated temperatures. The activation energy (Ea)
and the A-factor (A0) of the Arrhenius dependences for the
reaction from HA to HB (Scheme 1) are listed in Table 4. The
Ea value for 2-nitrobenzyl acetate 1a in solution can be
compared with a calculation of Ea = 62 kJ mol)1 in the gas
phase (13). Generally, Ea and A0 do not vary much with the
nature of the solvent. The obvious effects are a markedly larger
A0 value for 5 and smaller Ea when water is mixed with
DMSO. Therefore, one can assume a transition state with an
Figure 10. Change in (a) Aobs at 400 nm and (b) substrate concentra-
tion vs time due to 254 nm photoconversion of 1a (D), 1b (h), 1c (e),
4a (s, d), 4b ( ) and 4c ( ) in acetonitrile.

168 Filiz Bley et al.
concentration using HPLC data (Fig. 10b). The literature
values are Fd = 0.08 for 1b in acetonitrile and Fd = 0.15 for
1a in buffered DMSO-water (6,12). For comparison,
Fd = 0.1 for 2b in acetonitrile using kirr = 313 nm. Fd is
typically 0.2 for 4 ⁄ 5 type acids (Table 3) and zero only for 7,
which contains the rigid amino group and shows no
transients at all. The photodecomposition of 1a yields
2-nitrosobenzaldehyde and acetic acid, that of 1b or 1c
benzoic acid or 4-methylbenzoic acid, respectively. These
three organic acids are present in neutral aqueous solution as
anion due to pKa ca 3–5. They are also formed upon
irradiation of 2 ⁄ 3, together with the 4,5-dimethoxy and 4,5-
methylenedioxy derivatives of 2-nitrosobenzaldehyde. It has
been reported that the quantum yield of photocleavage is
markedly lower than the unsubstituted 2-nitrobenzyl caged
compound (23).
Nitroso compounds, which are formed upon irradiation
of nitrobenzyl compounds may yield the corresponding
dimer, e.g. at low temperatures (1,34), but the equilibrium
between the monomer and the corresponding dimer is
difficult to predict. Detection of a dimer within ca 10 s
has been reported for a 2-nitrophenylethoxyethyl phosphate
(17), but the substrate concentrations under these IR
measurements in water are typically a thousand times larger
than for UV–Vis spectroscopy. It has been argued that the
absorption at 320 nm of 2-nitrosobenzaldehyde is due to the
dimer and at 750 nm due to the monomer (4,13,17), which
however also absorbs around 300 nm (6). Arguments against
thermal dimer formation under the spectroscopic conditions
of nitrosobenzenes have been given (4,19–21). Here, no
dimers could be detected by UV–Vis spectroscopy and
HPLC for any photoproduct in acetonitrile under the
applied conditions. The photoconversion into 2-nitrosobenz-
aldehyde and benzoic acid is the main process, e.g. for
1a–4e; other peaks were only registered upon prolonged
irradiation, e.g. for 5a–6b.
distinguished, whereby it was proposed that the longer lived
component is stabilized by the interaction of the a-hydrogen
with an oxygen of the HA form (6). Earlier, we have
considered sterical reasons as an alternative explanation (12).
The further rearrangement leads via the cyclic benzisoxazo-
line (HB in Scheme 1) and the hemiacylal (HC) into the
2-nitroso product OCRArNO and HX (MeCO2H, PhCO2H
and p-TolCO2H for acetic, benzoic and toluic acid esters,
respectively). For the specific case of 2-nitrobenzyl methyl
ether in aqueous solution these species have recently been
detected (20). In aqueous solution the equilibria (HB ⁄ B) and
HC ⁄ C)) could be important and eventually the leaving group
HX is released. The most frequently concluded mechanism is
that the aci-nitro acid form (HA) is the precursor of both
OCRArNO and HX. Walbert et al. suggested that the anion
of the 2-(2-nitrophenyl)ethyl acetate-caged compounds under-
goes b-elimination and the acid nitrosobenzene formation
(22).
It is a widely accepted assumption among researchers using
caged compounds, that the decay of the aci-nitro intermediate
HA (Scheme 1) is the rate-limiting step in the photorelease of
the active species (35,36). However, detailed studies on model
systems (20) have revealed that this assumption is not
necessarily correct. If the leaving group is methoxy, the release
is rather slow (about saci = 1–100 s) and strongly dependent
on the reaction condition. The rate changes more than 2 orders
of magnitude as a function of pH. Earlier studies on a caged
glutamate (37) reveal that the release of glutamate from
a-carboxynitrobenzyl glutamate in phosphate buffer at pH 7 is
as fast as the decay of the aci-nitro intermediate and saci is ca
20 ls. From these data one can only get an upper limit for the
lifetime of the hemiacylal HC, which directly releases the free
acid. The difference in behavior between compounds with
methoxy or carboxy as leaving groups is in good agreement
with the fact that carboxylates are much better leaving groups
than alcoholates, or that carboxylic acids are much better
leaving groups than alcohols. We observed lifetimes of 0.001–
160 ms for the aci-nitro intermediates of different compounds.
The assumption that the decay of the aci-nitro intermediate is
the rate-limiting step seems to be likely in many cases but is not
true in all.
General mechanism
The photoreaction of compounds with 2-nitrobenzyl entity is
initiated by intramolecular benzylic H-atom transfer to the
nitro group, reactions 1–3 in Scheme 1. Nitroarenes generally
show a substantial Fisc and intramolecular H-atom transfer
takes place in the triplet state. In specific cases the H-atom
can also be transferred in the excited singlet state (1,21).
Hydrogen shift gives the aci-nitro form as main observable
intermediate. For 2-nitrobenzyl aryl esters (6) and 2-nitrob-
enzyl-methyl-ether (20) in dry acetonitrile it has been argued
that no deprotonation takes place. More than one aci-nitro
form has been proposed in many cases (9–15), but in some
cases the decay at kaci is mono-exponential (13). The known
reason for an increase in the 430–500 nm absorbance (grow-
in component) in the ls range is the equilibrium between HA
and A) in aqueous solution or mixtures of organic solvents
with water; formation of A) is delayed and its maximum can
be redshifted (6,20,22). The aci-nitro form can be present in
any of the four isomers, two concerning the orientation of the
aci-nitro group (tautomeric forms) and two concerning the
orientation of the benzylic ⁄ quinoid group (diastereomers).
Only one or two decay components are usually observed
(12,13). The two stereoisomers (HAE and HAZ) can be
Properties of 2-nitrobenzyl acetate, benzoate and toluate
For 2-nitrobenzyl benzoate 1b two isomers can be distin-
guished kinetically (6). Two components of the aci-nitro
forms have also been proposed in other cases (13,14). It has
been argued that the longer lived component of 1b is
stabilized by the interaction of the a-hydrogen with an
oxygen of the AH form (6). For 2-nitrophenylethyl acetates
the redshift of the absorption maximum and the long-
wavelength side of aci-nitro has been attributed to the
conversion of AH into A) (22). We suggest that both
reasons, depending on the structure and the medium, account
for the two components: hydrogen bonding, as in dichlo-
romethane (Fig. 1b), dry DMSO or dry acetonitrile, and
deprotonation. This latter reaction (2 in Scheme 1) is
proposed to account for the two aci-nitro species of 1a–c in
mixtures of DMSO (Fig. 2a,b), acetonitrile or ethanol with
water (Table 1). Apparently, the redshift is larger for 10:1
mixture than in neat water.

Photochemistry and Photobiology, 2008, 84 169
group is unimportant. These compounds show the largest
spectral changes among the 2-nitrobenzyl esters (27). We
therefore tried to detect the formation of OCRArNO (R:
CO2H) of 4b. In fact, after complete transient decay around
400 nm, a conversion into a permanent absorption was
observed (Fig. 4a,b). T-T absorption and aci-nitro form
absorption are also overlapping for 4d (Fig. 5a,b) and 4e,
containing amino acid building blocks. A dimethyl compound
similar to 4d was used in the photoreduction of cytochrome C
(7). Some effects of substitution were reported (13).
Bis-nitrophenyl acetates 5b (Fig. 6a,b) and 5a reveal mono-
exponential decay at kaci which is slightly redshifted with
respect to a classical aci-nitro spectrum. A rather short lifetime
(Table 1) can be attributed to the sterical stress due to the
bulky groups. Introduction of a carboxy group gives rise to a
complex behavior for 6a, where a grow-in within ca 1 ls and a
redshift of kaci was detected (Fig. 8a,b). This conversion of an
initial aci-nitro form into a second with a rather long saci is
probably due to the two carboxy groups as otherwise the
structure of 6a is identical to 5a. The reason cannot be the
carboxy group as such, as this conversion does not occur for
4a–e. One possibility could be proton transfer between the aci-
nitro and meta-carboxy groups. The reason is not fully
understood.
In principle, the triplet state could be the precursor of the
aci-nitro form or not. If the triplet state is part of the pathway
into the nitroso product, then quenching by oxygen should
lower Fd. The triplet state seems not to be in the pathway via
the aci-nitro form as Fd is not affected by the presence or
absence of oxygen. This is different to the case of 1-nitro-2-
naphthaldehyde, where involvement of the triplet state, in
addition to another (singlet) pathway to the aci-nitro form,
had been concluded, but only 30% of the excited 1-nitro-2-
naphthaldehyde molecules are deactivated via the observed
triplet state (33).
Properties of 4,5-dimethoxynitrobenzyl and
4,5-methylenedioxynitrobenzyl compounds
The 4,5-methylenedioxy or methoxy groups lead to several
surprising effects of the transients. One is the broad spectrum
(Fig. 3a,b), one the much shorter lifetime and another the
effect of quenching by oxygen and conversion of triplet into
O2(1Dg). FD of 2 ⁄ 3 in dichloromethane (Table 2) account for
30–50% of the absorbed photons. These results are not in
agreement with the efficiency of aci-nitro formation. The
suggested explanation is that the transient of 2 ⁄ 3 is assigned to
a triplet state, as illustrated in Scheme 2. The photoconversion
of 2-nitrobenzaldehyde into 2-nitrosobenzoic acid occurs via
the aci-nitro form but does not reveal any triplet state in the ns
time scale (1). For 1-nitro-2-naphthaldehyde, however, Dvor-
nikov et al. showed that the photoconversion into the nitroso
acid does not reveal any aci-nitro form but the precursor; this
triplet state has a lifetime of 50–600 ns at room temperature
and 65 ms in ether ⁄ pentane ⁄ ethanol at )196ꢁC (33). The
relatively high DAT values of the triplet in 2 ⁄ 3 and the lower
DAaci values of the longer lived transient(s) are compatible with
comparable F values and much higher e values for T-T
absorption than for absorption of the aci-nitro form. The
charge transfer, induced by the electron donating groups in
positions 4 and 5 shifts not only the absorption maximum to ca
350 nm but also lowers the triplet level. For 2 ⁄ 3 we propose
two reaction sequences with comparable yield: one is as for 1
and the other is charge-transfer induced, but chemically inert.
It is worth mentioning that also for 4,5-dimethoxy- and
4,5-methylenedioxy-2-nitrobenzyl alcohol, in contrast to
parent 2-nitrobenzyl alcohol, a triplet state with similar
features (albeit low FD) has been detected (26). For 4-amino-
4¢-nitroarenes, a related donor-substituted aromatic nitrocom-
pound, a charge transfer triplet state with a lifetime of a few ns
and a molar absorption coefficient in the range of eT = (2–4) ·
M
)1
104
cm)1 at 600–900 nm are known (38).
CONCLUSION
Properties of nitro and dinitro compounds: 4-6
The photoreactions of 2-nitrobenzyl-type into nitroso com-
pounds and release of the leaving groups acetic acid (a: 1, 2,
3 and 4), benzoic acid (b) and methyl-benzoic acid (c) were
studied. The role of the organic acid HX within a group is
minor, Fd is relatively small and no principal changes were
found for saci under the same excitation conditions. Surpris-
ingly, the methylenedioxy and methoxy groups lead to
The results for a-carboxy-4,5-methylenedioxynitrobenzyl com-
pounds 4 are also in accord with Scheme 2. The short-lived
transient of 4b (Fig. 4a,b) is assigned to the triplet state due to
the effect of oxygen. The decay at kaci is mono-exponential,
introduction of a carboxy group in 4 gives rise to rather long
saci values which are similar for a–c, i.e. a change in the leaving
Scheme 2.

170 Filiz Bley et al.
for chemical kinetic investigations of the glycine receptor. Bio-
chemistry 39, 2063–2070.
population of a triplet state. In most of the other cases both
triplet state and aci-nitro form are observable in the ns ⁄ ls
and ls ⁄ ms range, respectively. Group 1 and 5 compounds
follow the classical pathway (reactions 1–6 in Scheme 1),
where only aci-nitro isomers were detected. For group 2 ⁄ 3
the triplet state is a major transient. The aci-nitro form, but
not the triplet state, is postulated to be an intermediate on
the pathway in the nitroso product. For a-carboxy-4,5-
methylenedioxynitrobenzyl compounds 4 both the triplet
state and the aci-nitro form were observed, but only the
latter is intermediate on the pathway into the nitroso
product. The effects of temperature on the decay kinetics of
the aci-nitro form reveal Arrhenius dependences and common
trends for a given molecule in solution. The changes in saci at
room temperature are due to both, Ea and A0. Ea is around
50 kJ mol)1 for several compounds in DMSO, acetonitrile or
ethanol, but A0 varies with medium and structure; it is
smallest for 6 and largest for 5. The solvent plays a decisive
role in the properties of the various steps, especially the
presence of water. When deprotonation can be excluded in
dichloromethane, intrinsic reasons have to be assumed. On
the other hand, in neutral aqueous solution the aci-nitro form
is present as anion.
12. Grewer, C., S. A. M. Mobarekeh, N. Wartzke, T. Rauen and K.
Schaper (2001) Substrate translocation kinetics of excitatory
amino acid carrier 1 probed with laser-pulse photolysis of a new
photolabile precursor of D-aspartic acid. Biochemistry 40, 232–
240.
13. Schaper, K., D. Dommaschke, S. Globisch and S. A. Madani-
Mobarekeh (2000) AM1 calculations on the mechanism of the
o-nitrobenzyl photochemistry. J. Inf. Rec. 25, 339–354; note that
for 1a Ea = 80 kJ mol)1 rather than an erroneous: 62 kJ mol)1
.
14. Schaper, K., S. A. M. Mobarekeh and C. Grewer (2002) Synthesis
and photophysical characterization of a new, highly hydrophilic
caging group. Eur. J. Org. Chem. 1037–1046.
15. Gee, K. R., L. Niu, K. Schaper, V. Jayaraman and G. P. Hess
(1999) Synthesis and photochemistry of a photolabile precursor
of N-methyl-D-aspartate (NMDA) that is photolyzed in the
microsecond time region and is suitable for chemical kinetic
investigations of the NMDA receptor. Biochemistry 38, 3140–
3147.
16. Corrie, J. E. T., B. C. Gilbert, V. R. N. Munasinghe and A. C.
Whitwood (2000) EPR studies of the structure of transient radicals
formed in photolytic reactions of some 2-nitrobenzyl compounds.
Characterisation of aryl alkoxy aminoxyls and nitroaromatic
radical-anions in the photolysis of caged ATP and related com-
pounds. J. Chem. Soc. Perkin Trans. 2, 2483–2491.
17. Corrie, J. E. T., A. Barth, V. R. N. Munasinghe, D. R. Trentham
and M. C. Hutter (2003) Photolytic cleavage of 1-(2-nitrophe-
nyl)ethyl ethers involves two parallel pathways and product
release is rate-limited by decomposition of a common hemiacetal
intermediate. J. Am. Chem. Soc. 125, 8546–8554.
Acknowledgements—We thank Professor Wolfgang Lubitz for his
support and Mr. Leslie J. Currell, Christian Lemsch and Horst Selbach
for technical assistance.
18. Dunkin, I. R., J. Gebicki, M. Kiszka and D. Sanin-Leira (2001)
Phototautomerism of o-nitrobenzyl compounds: o-Quinonoid aci-
nitro species studied by matrix isolation and DFT calculations.
J. Chem. Soc. Perkin Trans. 2, 1414–1425.
19. Schworer, M. and J. Wirz (2001) Photochemical reaction mecha-
nisms of 2-nitrobenzyl compounds in solution I. 2-nitrotoluene:
Thermodynamic and kinetic parameters of the aci-nitro tautomer.
Helv. Chim. Acta 84, 1441–1458.
20. Il’ichev, Y. V., M. A. Schworer and J. Wirz (2004) Photochemical
reaction mechanisms of 2-nitrobenzyl compounds: Methyl ethers
and caged ATP. J. Am. Chem. Soc. 126, 4581–4595.
21. Gaplovsky, M., Y. V. Il’ichev, Y. Kamdzhilov, S. V. Kombarova,
M. Mac, M. A. Schworer and J. Wirz (2005) Photochemical
reaction mechanisms of 2-nitrobenzyl compounds: 2-Nitrobenzyl
alcohols form 2-nitroso hydrates by dual proton transfer. Photo-
chem. Photobiol. Sci. 4, 33–42.
22. Walbert, S., W. Pfleiderer and U.E. Steiner (2001) Photolabile
protecting groups for nucleosides: Mechanistic studies of the 2-(2-
nitrophenyl)ethyl group. Helv. Chim. Acta 84, 1601–1611.
23. Woll, D., S. Walbert, K.-P. Stengele, T. J. Albert, T. Richmond, J.
Norton, M. Singer, R. D. Green, W. Pfleiderer and U. E. Steiner
(2004) Triplet-sensitized photodeprotection of oligonucleotides in
solution and on microarray chips. Helv. Chim. Acta 87, 28–45.
24. Smirnova, J., D. Woll, W. Pfleiderer and U. E. Steiner (2005)
Synthesis of caged nucleosides with photoremovable protecting
groups linked to intramolecular antennae. Helv. Chim. Acta 84,
891–904.
REFERENCES
1. Morrison, H. A. (1969) The photochemistry of the nitro and
nitroso groups. In The Chemistry of the Nitro and Nitroso Groups
(Edited by H. Feuer), pp. 165–213. Wiley, New York.
2. Dopp, D. (1995) CRC Handbook of Photochemistry and Photobi-
ology (Edited by W. M. Horspool and P.-S. Song), pp. 1019–1062.
CRC Press, Boca Raton, FL.
3. McCray, J. A. and D. R. Trentham (1989) Properties and uses of
photoreactive caged compounds. Annu. Rev. Biophys. Biophys.
Chem. 18, 239–270.
4. Pelliccioli, A. P. and J. Wirz (2002) Photoremovable protecting
groups: Reaction mechanisms and applications. Photochem. Pho-
tobiol. Sci. 1, 441–458.
5. Schupp, H., W. K. Wong and W. Schnabel (1987) Mechanistic
studies of the photorearrangement of o-nitrobenzyl esters. J. Pho-
tochem. 36, 85–97.
6. Zhu, Q. Q., W. Schnabel and H. Schupp (1987) Formation and
decay of nitronic acid in the photorearrangement of o-nitrobenzyl
esters. J. Photochem. 39, 317–332.
7. DiMagno, T. J., M. H. B. Stowell and S. I. Chan (1995) Nitro-
benzene caged compounds as irreversible photoreductants: A ra-
tional approach to studying photoinduced intermolecular
electron-transfer reactions in proteins. J. Phys. Chem. 99, 13038–
13047.
25. Globisch, S. (2002) Ph.D. thesis, Heinrich Heine Universitat,
Dusseldorf.
8. Buhler, S., I. Lagoja, H. Giegrich, K.-P. Stengele and W. Pflei-
derer (2004) New types of very efficient photolabile protecting
groups based upon the [2-(2-nitrophenyl)propoxy]carbonyl
(NPPOC) moiety. Helv. Chim. Acta 87, 620–659.
26. Gorner, H. (2005) Effects of 4,5-dimethoxy groups on the time-
resolved photoconversion of 2-nitrobenzyl alcohols and 2-nitro-
benzaldehyde into nitroso derivatives. Photochem. Photobiol. Sci.
4, 822–828.
9. Gee, K. R., L. Niu, K. Schaper and G. P. Hess (1995) Caged
bioactive carboxylates. Synthesis, photolysis studies, and biologi-
cal characterization of a new caged N-methyl-D-aspartic acid.
J. Org. Chem. 60, 4260–4263.
10. Niu, L., K. R. Gee, K. Schaper and G. P. Hess (1996) Synthesis
and photochemical properties of a kainate precursor and activa-
tion of kainate and AMPA receptor channels on a microsecond
time scale. Biochemistry 35, 2030–2036.
27. Bley, F. (2005) Ph.D. thesis, Heinrich Heine Universitat, Dussel-
dorf.
28. Schaper, K., F. Bley and H. Gorner (2006) Photochemical prop-
erties of o-nitrobenzyl compounds with a bathochromic absorp-
tion. Abstr. Am. Chem. Soc., 231, 676.
29. Fisher, G. J. and H. E. Johns (1976) Pyridine photohydrates. In
Photochemistry and Photobiology of Nucleic Acids (Edited by S. Y.
Wang), pp. 169–224. Academic Press, New York.
11. Grewer, C., J. Jager, B. K. Carpenter and G. P. Hess (2000) A new
photolabile precursor of glycine with improved properties: A tool
30. Heller, H. G. and J. R. Langan (1981) Photochromic heterocyclic
fulgides. Part 3. The use of (E)-a-(2,5-dimethyl-3-furylethylidene)

Photochemistry and Photobiology, 2008, 84 171
(isopropylidene)succinic anhydride as a simple convenient chemi-
cal actinometer. J. Chem. Soc. Perkin Trans. 2, 341–343.
31. Gorner, H. and D. Dopp (2002) Photoinduced demethylation
of 4-nitro-N,N-dimethylaniline. Photochem. Photobiol. Sci. 1, 270–
277.
32. Wilkinson, F., W. P. Helman and A. B. Ross (1993) Quantum
yields for the photosensitized formation of the lowest electroni-
cally excited singlet state of molecular oxygen in solution. J. Phys.
Chem. Ref. Data 22, 113–275.
33. Dvornikov, A. S., C. M. Taylor, Y. C. Liang and P. M. Rentzepis
(1998) Photorearrangement mechanism of 1-nitro-naphthaldehyde
and application to three-dimensional optical storage devices.
J. Photochem. Photobiol. A, Chem. 112, 39–46.
34. Gowenlock, B. G., M. J. Maidment, K. G. Orrell, I. Prokes and J.
A. Roberts (2001) Nitrosoanisoles. Sensitive indicators of dimer-
isation criteria for C-nitrosoarenes. J. Chem. Soc. Perkin Trans. 2,
1904–1911.
35. Corrie, J. E. T. and D. R. Trentham (1993) Caged nucleotides
and neurotransmitters. In Biological Applications of Photochem-
ical Switches (Edited by H. Morrison), pp. 243–306. Wiley, New
York.
36. Marriott, G., Ed. (1998) Caged compounds. In Methods of
Enzymology, Vol. 291, pp. 1–495. Academic Press, New York.
37. Cheng, Q., M. G. Steinmetz and V. Jayaraman (2002) Photolysis
of c-(a-carboxy-2- nitrobenzyl)-L-glutamic acid investigated in the
microsecond time scale by time-resolved FTIR. J. Am. Chem. Soc.
124, 7676–7677.
38. Piotrowiak, P., R. Kobetic, T. Schatz and G. Strati (1995) Tran-
sient charge transfer absorption bands as probes of ion-pairing
dynamics and energetics. J. Phys. Chem. 99, 2250–2253.