Photolabile protecting groups for nucleosides: Mechanistic studies of the 2-(2-nitrophenyl)ethyl group

Article abstract of DOI:10.1002/1522-2675(20010613)84:6<1601::AID-HLCA1601>3.0.CO;2-S

The photochemistry of several 2-(2-nitrophenyl)ethyl-caged compounds including caged thymidine nucleosides was studied by nanosecond laser flash photolysis and stationary illumination experiments with quantitative HPLC analysis for quantum yields and prod

Full text of DOI:10.1002/1522-2675(20010613)84:6<1601::AID-HLCA1601>3.0.CO;2-S

Helvetica Chimica Acta ± Vol. 84 (2001)
1601
Photolabile Protecting Groups for Nucleosides: Mechanistic Studies
of the 2-(2-Nitrophenyl)ethyl Group
by Stefan Walbert, Wolfgang Pfleiderer, and Ulrich E. Steiner*
Fachbereich Chemie der Universität Konstanz, D-78457 Konstanz
Dedicated with best wishes to Prof. Edgar Heilbronner on the occasion of his 80th birthday
The photochemistry of several 2-(2-nitrophenyl)ethyl-caged compounds including caged thymidine
nucleosides was studied by nanosecond laser flash photolysis and stationary illumination experiments with
quantitative HPLC analysis for quantum yields and product distribution. Effects of solvent basicity and acidity
were investigated by varying the H2O content and HCl concentration, respectively, in MeCN/H2O mixtures. For
all compounds 1 ± 7 investigated, intramolecular H abstraction by the nitro group from the exocyclic a-position
with respect to the aryl moiety was found to be the primary process. The protolytic dissociation equilibrium of
the resulting aci-nitro compound was kinetically characterized in the 0.1 ± 10 ms time region. In general, two
reaction channels compete for the aci-nitro compound and its anion: b-elimination of the caged compound
occurs from the anion, while from the undissociated aci-nitro compound, a nitrosobenzene derivative is formed
with no release of the caged compound. The yield ratio of these two reaction channels can be controlled through
shifts in the protolytic dissociation equilibrium of the aci-nitro compound. In solutions with either low basicity
(H2O-free MeCN) or high acidity (higher concentration of HCl in H2O/MeCN), two as yet unidentified
products are formed, each one specifically for one of the mentioned conditions.
Introduction. ± Photolabile protecting groups play an important role in synthetic
organic chemistry [1], for caging of biologically active molecules [2], and for light-
directed, combinatorial solid-phase syntheses of biopolymers [3]. A particularly
attractive application is the generation of so-called high-density DNA chips [4][5]
which are needed in various types to perform the sequencing by hybridisation (SBH)
method for oligonucleotide and gene sequencing [6]. The o-nitrobenzyl (oNB) type of
protecting groups has been commonly used since 1901 when the photoreaction was
discovered [7], and its reaction mechanism has been investigated in detail [8]. So far,
the [(a-methyl-2-nitropiperonyl)oxy]carbonyl ([1-(6-nitro-1,3-benzodioxol-5-yl)-
ethoxy]carbonyl; MeNPOC) group has been the preferred choice in DNA-chip
production [4].
The photochemical cleavage mechanism of the (o-nitrobenzyl)oxy function is
triggered by the abstraction of a benzylic H-atom by the excited nitro group [9].
Recently, a new type of photolabile protecting group was developed on the basis of the
[2-(2-nitrophenyl)ethoxy]carbonyl functionality [10][11] which is cleaved by a light-
induced b-elimination process with formation of o-nitrostyrene, CO2, and the
corresponding alcohol. It was suggested [10] that, in analogy to the o-nitrobenzyl
groups, the cleavage of the [2-(2-nitrophenyl)ethoxy]carbonyl moiety, too, should start
with the formation of an aci-nitro intermediate, through H abstraction by the nitro
group at the exocyclic a-position with respect to the aryl moiety. Based on results from
laser flash photolysis and quantitative stationary irradiation experiments, the present

1602
Helvetica Chimica Acta ± Vol. 84 (2001)
paper presents evidence of the suggested reaction mechanism and provides further
insight into the cleavage mechanism of the new type of photolabile protecting groups.
Results. ± Laser Flash Photolysis. Laser flash photolysis studies on time scales from
nanoseconds to seconds were carried out with the ring- and chain-substituted
derivatives 1 ± 7 of (2-nitrophenyl)ethane (oNPE) (Fig. 1) in MeOH/H2O, MeCN,
and MeCN/H2O. In all cases, the typical spectra of the aci-nitro compounds with their
characteristic maxima around 400 nm [12] were observed. As examples, in Fig. 2, we
present transient spectra for 2-(2-chloro-6-nitrophenyl)ethyl acetate (1) and 2-(2-
nitrophenyl)propyl acetate (2) in MeCN/H2O 1:1 (v/v).
It is of interest to note that, while the differences in the extinction coefficients of the
two compounds 1 and 2 at the laser wavelength of 355 nm (e ˆ 340 and 400 m 1 cm 1 for
1 and 2, resp.) are only minor and the same laser intensity was applied, the initial signal
amplitudes after the laser flash differ by a factor of ca. 4 between the two compounds.
This indicates that the primary reaction for the a-methylated compound (a denotes the
exocyclic position with respect to the aryl moiety) is much more efficient. During the
first microseconds after the laser flash, the initial transient absorption band undergoes a
slight bathochromic shift that is indicative of protolytic dissociation of the aci-nitro
compound, i.e., appearance of the corresponding anion. The extent and rate of this
dissociation process can be modified by the nature of the solvent and the acidity of the
solution. This is demonstrated in Fig. 3 where kinetic traces are shown for a variation of
Fig. 1. 2-(2-Nitrophenyl)ethyl derivatives 1 ± 7 investigated in this work
Fig. 2. Transient absorption spectra observed at various delay times after laser flash photolysis of 0.5 mm
solutions of a) 1 and b) 2 in MeCN/H2O 1 :1 (v/v)

Helvetica Chimica Acta ± Vol. 84 (2001)
1603
the MeCN/H2O ratio (Fig. 3,a) and for varying concentrations of HCl (Fig. 3,b),
respectively. The sample compound used in this study was 2-(5-bromo-2-nitrophenyl)-
propyl methyl carbonate (3), which was synthesized in several steps from N-(3-
ethylphenyl)acetamide (see Exper. Part).
Fig. 3. Kinetic traces of transient absorption observed at 450 nm upon flash photolysis of compound 3: a) in
MeCN/H2O mixtures with variable mixing ratios and b) in MeCN/H2O 1:1 (v/v) with various concentrations of
HCl
At 450 nm the absorbance of the anion generated from 3 greatly exceeds the
absorbance of the aci-nitro form. Thus the rise of the signal intensity indicates the
formation of the aci-nitro anion.
As can be seen in Fig. 3,a for 3, the appearing rate and the amplitude of the anion
absorption increase with increasing content of H2O. On the other hand, in a solvent rich
in H2O, the rate and degree of dissociation can be reduced by lowering the pH (cf.
Fig. 3,b). The transient spectra of the aci-nitro compound and its anion completely
disappear after some milliseconds. In case of compound 1, no further transient is
detectable. In case of compound 2, however, a new persistent absorption band appears
during fractions of seconds. The new band is very similar to what is known from
investigations with the oNB type of protecting groups where it is assigned to the
nitrosobenzaldehyde derivative. For the oNPE protecting group, such a product has not
been seen before. Therefore, a more detailed investigation of the photolysis products
obtained under stationary illumination conditions was undertaken.
Continuous-Illumination Experiments. The quantitation of the photoreactions and
the photoproducts under stationary illumination conditions is conveniently based on
HPLC analysis. An example of our results with compounds 4 and 5 is shown in Fig. 4.
Each peak in the chromatogram is characterized and, with the pertinent reference
compound at hand, identified by its retention time and its UV spectrum.
From the concentrations of the unreacted caged compounds and of photochemi-
cally released thymidine after given illumination intervals, photochemical quantum
yields of total photodecomposition (ftot) and release of the caged thymidine (frel) were
determined (cf. Table 1). The caged thymidine compounds 4 and 5 differ from their
analogues 1 and 2 investigated by laser flash photolysis only by the replacement of the
acetate moiety by the thymidin-5'-yl carbonate function. The overall decomposition
quantum yields for compound 5 carrying an a-methyl substituent is ca. 5 times the value
of compound 4 that is unsubstituted at this a-position (exocyclic position with respect
to the aryl moiety). This observation is in good agreement with the ratio of absorption-

1604
Helvetica Chimica Acta ± Vol. 84 (2001)
Fig. 4. HPLC Traces and UV spectra of the corresponding peaks for a) 4 and b) 5. Solutions were photolysed to
ca. 20% loss of starting material in MeCN/H2O 1:1 (v/v). The peaks at tR 1.8 and 7.0 min are artefacts due to
injection of air and passage of the gradient front across the detector.
band intensities of the primary photoproducts as seen in the laser flash photolysis of
compounds 1 and 2. The similarity suggests that i) the yield of the photochemical
primary product determines the final photochemical quantum yield, and that ii) the
quantum yields of photoreactions of the photolabile protecting groups considered here
are only little affected by the structure of the moiety to which they are attached. The
latter assertion is further supported by the quantum yields observed for compounds 2,
6, and 7 having the anions MeCOO , MeOC(O)O , and OH as formal leaving groups.
Their quantum yields are found in a fairly narrow range of 0.39 to 0.48 which also
includes the quantum yield of compound 5.
Table 1. Quantum Yields in MeCN/H2O 1 : 1 (v/v) under Stationary Illumination at 366/355 nm
4
5
6
2
7
ftot
frel
0.075
0.073
0.40
0.30
0.42
±
0.48
±
0.39
±
a
)
a) Appearance of free thymidine.
With thymidine as the caged moiety in compounds 4 and 5, we also assessed the
quantum yield of the appearance of free thymidine. For 4, this quantum yield is
practically identical to the overall quantum yield ftot of photodecomposition of the
starting material, and no other peaks than for thymidine and the nitrostyrene appear in
the chromatogram. This means that there are no side reactions. On the other hand, for
5, the quantum yield of free thymidine is clearly lower than for the overall
photoreaction. Furthermore, the chromatogram exhibits a new peak at tR 11.5 min
which, from its UV spectrum, can be assigned to a nitrosobenzene derivative. This
observation meets with our finding of a nitroso-product band in the laser flash
photolysis of compound 2 carrying the same protecting group as 5. Corresponding by-
products are formed for all caged substances carrying the 2-(2-nitrophenyl)propyl
moiety such as 2, 6, and 7 included in Table 1. In case of compound 7, with OH as the

Helvetica Chimica Acta ± Vol. 84 (2001)
1605
formal leaving group, nitroso-product formation is the only observed reaction path.
Here no styrene, the indicator of the b-elimination pathway, is found at all.
To support this UV-spectroscopic evidence for nitroso-product formation, a larger
scale photolysis of 7 was carried out, and the photoproduct 8 was isolated by column
1
chromatography and identified by H-NMR spectroscopy (see Exper. Part). Further-
1
more, photolysing 7 in an NMR sample tube and measuring H-NMR spectra after
certain illumination periods directly showed the conversion of 7 into the nitroso
product 8.
pH and Solvent Effects on Product Distribution. The consequences of pH and
solvent variation as borne out in pronounced changes of kinetics and equilibrium
position of the aci-nitro compoundsꢀ protolytic dissociation in the laser flash photolysis
experiments were also explored on the level of final-product formation. Pertinent
investigations were carried out with the same compound 3 that was used in the laser
flash photolysis studies (cf. Fig. 3). In Table 2, the quantum yields and product-yield
profiles observed after 40 ± 50% photochemical conversion under continuous illumi-
nation at 366 nm are listed.
Table 2. Overall photochemical quantum yields and relative product yields for compound 3 under variation of
H2O content in MeCN/H2O mixtures and HCl concentration in H2O/MeCN 1 : 1
ftot
aa)
Yieldsb)
styrene
nitroso
X
Y
compound
0% H2O
20% H2O
40% H2O
0.51
0.49
0.51
0.37
0.37
0.41
0
35
60
0
20
80
3.5
23.9
26.6
0
6.7
24.9
34.2
9.1
7.3
2.7
19.0
12.7
14.5
0.9
0
0
0
0
0
0
18.1
8.6
0
1
10 m HCl
10 3m HCl
10 5m HCl
0
a) Degree of protonation estimated from the laser flash photolysis experiments. b) The values given represent
peak integrals obtained after continuous photolysis of 0.5 mm solutions of 3 at 366 nm for 300 s. For each
product, the UV absorbance was measured at the wavelength of its optimum absorbance. Due to different e
values, the data for different compounds can be compared on a qualitative basis only. For each product,
however, the given data give correct ratios of the yields under different conditions. X, Yˆ unknown compounds.
The following observations are of significance. i) In each of the two series (variation
of H2O content or of HCl concentration), the total quantum yield of photodecompo-
sition is rather invariable. ii) With a variation of the percentage of H2O in a MeCN/H2O
solvent mixture, the yield of styrene (b-elimination pathway) and nitroso product are
complementary. b-Elimination is favoured under conditions where the deprotonation
of the aci-nitro compound is efficient. The reverse is true for the reaction path yielding

1606
Helvetica Chimica Acta ± Vol. 84 (2001)
the nitroso product. At low H2O content, the drop in b-elimination efficiency and the
rise in nitroso product formation efficiency is accompanied by the occurrence of a new,
as yet unidentified, reaction product X. iii) With an increase of acidity (concentration
of HCl) in MeCN/H2O 1:1 (v/v), the b-elimination pathway is gradually suppressed
and even reaches zero level at 0.1m HCl concentration. Nitroso-product formation is
first increasing with acidity but at high concentration of HCl, it is significantly reduced
while, concomitantly, another as yet unidentified product Y is formed.
Discussion. ± The essential findings of our investigations of oNPE compounds may
be summarized as follows: i) The primary product in the photochemistry is the aci-nitro
compound generated by H-transfer from the exocyclic a-position (with respect to the
aryl moiety) to the nitro group. ii) The overall quantum yield of photodecomposition is
proportional to the yield of primary intramolecular H-abstraction. iii) The kinetics and
equilibrium states of the protolytic dissociation of the aci-nitro forms can be observed.
iv) The reaction branching between b-elimination and nitroso-product formation can
be assessed. v) The product ratios can be controlled by the protolytic dissociation of the
aci-nitro form. Based on these conclusions, an overall reaction mechanism as shown in
the Scheme is suggested.
Upon irradiation into the first absorption band, the S1(np*) of the oNPE
chromophore is populated. It decays by intersystem crossing to the triplet state
T1(np*) and by internal conversion to the singlet ground state. To some extent the H-
transfer to the nitro group may already take place in the excited singlet state, but this
process is probably a main decay pathway for the triplet state. The primary product
produced this way must be a triplet biradical that will rapidly relax to the o-quinonoid
singlet ground state of the aci-nitro form, as shown in the Scheme.
In the literature, evidence has been reported both for H-transfer in the singlet and
the triplet state of np*-excited oNB compounds [13]. For the oNPE compounds we
cannot make a definite statement about the roles of S1 and T1 as precursors of the
primary H-transfer product in the direct photolysis process. We have evidence, however,
that the same kind of primary process can be induced by triplet sensitization [14].
Since the aci-nitro compound represents an unstable tautomer of the starting
compound, its reversion to the stable tautomer must be included into the reaction
scheme. In general, this reaction should compete with any other reaction path of the
aci-nitro compound. However, the observation ii) that the photodecomposition
quantum yield is proportional to the primary yield of the aci-nitro product for
compounds with or without a Me substituent at the exocyclic a-position, suggests that
this backward tautomerism cannot kinetically compete with the further reaction
pathways leading to irreversible product formation. Similar conclusions have been
reached for the photodecomposition of oNB compounds [15].
The kinetics of the protolytic dissociation of the aci-nitro compound and the
establishment of a medium-dependent equilibrium is readily observable in the 0.1 ±
10 ms time region. The acidity of the aci-nitro form is not high enough to allow
dissociation in pure MeCN. However, already weak bases such as H2O can act as
effective proton acceptors. Thus the rate and equilibrium position of the protolytic
dissociation is quite sensitive to the H2O content in MeCN/H2O mixtures. On the other
hand, if H2O-containing solutions are acidified, suppression of the dissociation is

Helvetica Chimica Acta ± Vol. 84 (2001)
1607
Scheme
observed, which is unequivocal evidence of the reversibility of the protolytic process.
For compound 3, e.g. (cf. Fig. 3), in MeCN/H2O 1:1, we estimate the following rate
1
1
1
parameters: 2 ´ 106 s for the rate constant of protolytic dissociation and 2 ´ 1010m s ,
‡
i.e. close to diffusion-controlled, for the rate constant of protonation by H3O . These
rate constants correspond to a pKa of 4 which is similar to what has been published for
other o-quinonoid aci-nitro compounds [16].
From the correlation of the final product distribution with the position of the
protolytic dissociation equilibrium, we conclude that the aci-nitro anion is the starting
point for the b-elimination pathway, whereas the undissociated aci-nitro compound
gives rise to formation of the nitroso product. The connection between b-elimination
and the aci-nitro anion is in fact what would have been reasonably anticipated. It is well-

1608
Helvetica Chimica Acta ± Vol. 84 (2001)
known that in the presence of strong bases, b-elimination from oNPE compounds can
be achieved in thermal reactions [17]. The photochemical way represented in the
Scheme just reduces the required basicity of the proton acceptor. Under photochemical
conditions, even the basicity of H2O is sufficient. It must be borne in mind, however,
that, even if b-elimination can be supported by shifting the protolytic equilibrium from
the aci-nitro compound to the anion, since the latter process is reversible, the yield into
the b-elimination channel is ultimately determined by the kinetics of the elimination
step itself. The rate of this process is expected to depend on the properties of the leaving
group. For example, the rate should decrease in the series ROC(O)O , RCOO , HO .
So far, we did not attempt to quantify the difference between the former two of them,
but HO is clearly so slow that in this case, b-elimination cannot compete with the
nitroso channel.
The reaction path to the nitroso product is usually assumed to involve a cyclic
intermediate [8] that mediates the transfer of an O-atom from the nitro group to the
exocyclic C(a). We assume that the same happens in case of the oNPE compounds.
However, in contrast to the oNB case, where the protected moiety is released from the
exocyclic C(a) center after the O-transfer, in the oNPE case, the reaction stops at this
point, because the protected moiety is attached to the b-position and, therefore, not
affected by the O-transfer. For oNPE compounds to be used as photolabile protecting
groups, it is, therefore, important to avoid the nitroso reaction path. Nevertheless, our
mechanistic findings regarding the nitroso pathway with the oNPE compounds can be
applied to the oNB case too, and they contribute valuable details for a more general
insight into the reaction mechanism of the nitroso-product formation. In particular, our
results provide firm evidence that, contrary to the general notion, the ring closure
reaction on the way to the nitroso product involves the undissociated aci-nitro form
rather than its anion. This finding confirms a prediction recently published by Schaper
et al. [18] on the basis of semi-empirical (AM1) quantum-chemical calculations. So far,
no direct spectroscopic evidence of the cyclic intermediate has been reported. We, too,
did not see a spectroscopic transient in the UV/VIS that could be assigned to it, but we
were able to detect the appearance of the nitroso compound, which occurs in about a
second (cf. Fig. 2,b for compound 2). Therefore, it seems reasonable to assign this value
to the lifetime of the cyclic dihydroisoxazol intermediate.
Finally, we address the two unidentified products X and Y. Since their formation is
found under conditions where the undissociated aci-nitro compound prevails, the latter
species may function as their precursor. However, further investigations are needed to
clarify their structure and mechanisms of formation.
Experimental Part
General. TLC: precoated silica-gel thin layer sheets 60 F 254 from Merck. Prep. column chromatography
1
(CC): silica gel 0.04 mm from Baker. UV/VIS: Perkin-Elmer Lambda-15; lmax in nm (lg e). H-NMR (unless
otherwise noted): Bruker AC-250; d in ppm rel. to SiMe4.
1. 2-(2-Chloro-6-nitrophenyl)ethyl Acetate (1). A mixture of 2-(2-chloro-6-nitrophenyl)ethanol [19] (0.53 g,
2.5 mmol) and Ac2O (3 ml) in pyridine (3 ml) was stirred at r.t. for 4 h. After evaporation, the residue was co-
evaporated with MeOH (3 Â 5 ml) and toluene (2 Â 5 ml), dissolved in CHCl3 (30 ml), and treated with sat.
NaHCO3 soln. (30 ml). The org. phase was dried (Na2SO4) and evaporated and the residue purified by CC
(silica gel, 10 Â 2.5 cm, CH2Cl2): 0.62 g (97%) of 1. Colorless oil. 1H-NMR (CDCl3): 7.73 ± 7.26 (m, 3 arom. H);
4.33 (t, PhCH2CH2O); 3.39 (m, PhCH2CH2O); 2.02 (s, Ac).

Helvetica Chimica Acta ± Vol. 84 (2001)
1609
2. 2-(2-Nitrophenyl)propyl Acetate (2). As described for 1, with 2-(2-nitrophenyl)propan-1-ol [20] (7; 1.0 g,
5.5 mmol), Ac2O (10 ml), and pyridine (10 ml). Workup with MeOH (3 Â 10 ml), toluene (2 Â 10 ml), CHCl3
(50 ml), sat. NaHCO3 soln. (50 ml), and CC as described gave 1.2 g (97%) of 2. Colorless oil. 1H-NMR
(CDCl3): 7.75 (d, 1 H o to NO2); 7.50 (m, 3 arom. H); 4.21 (m, CHCH2O); 3.72 (m, MeCHCH2O); 1.99 (s, Ac);
1.36 (d, MeCHCH2O).
3. 3-Ethyl-4-nitrobenzenamine [21]. N-(3-Ethylphenyl)acetamide [21] (33 g, 0.2 mol) was slowly added
under intense mechanical stirring to 100 ml of conc. sulfuric acid. To the resulting soln., cooled to 158, fumic
nitric acid (9.5 g, 6.2 ml, 0.15 mol) was added dropwise, while keeping the temp. below 108. After 75 min at
20 to 108, the soln. was poured on ice, neutralized with solid Na2CO3, and extracted with Et2O (3 Â 250 ml).
The org. phase was washed with H2O (600 ml), dried (Na2SO4), and evaporated. The crude dark yellow oil
(39 g) was heated in the presence of conc. HCl soln. (125 ml) under reflux for 2.5 h. After cooling, the
precipitate was washed with H2O, suspended in 1m NaOH (300 ml), and extracted with Et2O (2 Â 150 ml). The
org. phase was dried (Na2SO4) and evaporated to give again a brown solid (14.2 g). The combined filtrates and
the NaOH phase were neutralized and again extracted with Et2O. The org. phase yielded, after drying (Na2SO4)
and evaporation, 16 g of a dark oil. The crude product was purified by CC (each oily fraction separately, with
silica gel (235 g), 7 Â 14 cm, petroleum ether/AcOEt 20 :1 ! 2 :1): 18.09 g (54%) of amine. Yellow solid. M.p.
84 ± 858 ([21]: 80 ± 818); UV (MeOH): 203 (4.21), 231 (3.82), 245 (sh, 3.69), 372 (sh, 4.09). 1H-NMR (CDCl3):
7.96 (m, H C(5)); 6.47 (m, H C(2), H C(6)); 4.23 (br. s, NH2); 2.93 (q, MeCH2); 1.24 (t, MeCH2). 13C-NMR
(600 MHz, CDCl3): 151.30 (C(1)); 143.25 (C(3)); 139.42 (C(4)); 128.35 (C(5)); 115.35 (C(2)); 111.72 (C(2));
27.48 (MeCH2); 14.67 (MeCH2).
4. 5-Bromo-1-ethyl-2-nitrobenzene. Finely powdered 3-ethyl-4-nitrobenzenamine (16.6 g, 0.1 mol) was
added to a warm (608) mixture of 48% HBr soln. (95 ml) and H2O (155 ml). The resulting soln. was cooled
rapidly to 08, and the resulting suspension was diazotized at 0 ± 38 with NaNO2 (7.0 g, 0.101 mol) in H2O (20 ml).
After stirring for 5 min at 08, the soln. was treated with urea (1 g) and the mixture stirred for another 10 min and
filtered off. The filtrate was added at r.t. to a stirred suspension of copper sulfate pentahydrate (15 g, 60 mmol)
and copper (6 g, 94 mmol) in HBr (63 ml) and H2O (37 ml). The mixture was heated to 758 for 1 h and, after
cooling, extracted with AcOEt. The org. phase was washed with 1m NaOH (150 ml) and H2O (200 ml), dried
(Na2SO4), and evaporated: 18.3 g (80%) of the bromo derivative. Yellow oil. B.p. 75 ± 1038/0.005 mbar. UV
(MeOH): 203 (4.19), 217 (sh, 3.98), 267 (3.91), 320 (sh, 3.21). 1H-NMR (250 MHz, CDCl3): 7.76 (d, H C(3));
7.50 (d, H C(6)); 7.45 (dd, H C(4)); 2.89 (q, MeCH2); 1.27 (t, MeCH2). Anal. calc. for C8H8BrNO2 (230.07):
C 41.77, H 3.50, N 6.09; found: C 41.71, H 3.59, N 6.08.
5. 2-(5-Bromo-2-nitrophenyl)propan-1-ol. A mixture of 5-bromo-1-ethyl-2-nitrobenzene (1.39 g, 6 mmol)
and paraformaldehyde (210 mg, 7 mmol) in dry DMSO (10 ml) was treated with KOtBu (100 mg, 0.89 mmol) in
tBuOH (4 ml). After stirring for 15 min at r.t. and 2 h at 808, the mixture was neutralized, diluted with sat. NaCl
soln. (25 ml), and extracted with AcOEt. The org. phase was dried (Na2SO4) and evaporated and the crude
product purified by CC (silica gel, (45 g), 4 Â 12 cm, toluene ! toluene/AcOEt 8 :1): 1.32 g (85%) of a yellow
1
oil. UV (MeOH): 202 (4.16), 218 (sh, 3.95), 263 (3.76), 330 (sh, 2.96). H-NMR (CDCl3): 7.63 (m, H C(3),
H
C(6)); 7.47 (dd, H C(4)); 4.74 (m, MeCHCH2OH); 3.52 (sext., MeCHCH2OH); 1.68 (br. s, OH); 1.30
(d, MeCHCH2OH). Anal. calc. for C9H10BrNO3 (260.09): C 41.56, H 3.88, N 5.39; found: C 41.74, H 3.95,
N 5.27.
6. 2-(5-Bromo-2-nitrophenyl)propyl Carbonochloridate. To
a cold soln. (08) of diphosgene
(ˆtrichloromethyl carbonochloridate; 740 mg, 455 ml, 3.8 mmol) in dry THF (4 ml), 2-(5-bromo-2-nitro-
phenyl)propan-1-ol (650 mg, 2.5 mmol) and Et3N (255mg, 350 ml, 2.5 mmol) in dry THF (4 ml) were added.
After stirring for 1 h at 08, the solvent was evaporated: 805 mg (99%) of the carbonochloridate. Yellow oil.
1H-NMR (CDCl3): 7.72 (d, H C(3)); 7.55 (m, 2 H, H C(6), H C(4)); 4.45 (m, MeCHCH2O); 3.80 (sext.
MeCHCH2O); 1.39 (d, MeCHCH2O).
7. 2-(5-Bromo-2-nitrophenyl)propyl Methyl Carbonate (3). To a cold soln. of MeOH (20 ml) and Et3N
(2 ml), 2-(5-bromo-2-nitrophenyl)propyl carbonochloridate (1.5 g, 4.7 mmol) was added dropwise at 0 ± 58.
After stirring for 45 min at r.t., the solvent was evaporated and the residue purified by CC (silica gel (60 g),
3.5 Â 16 cm, CH2Cl2): 1.39 g (93%) of 3. Yellow oil. 1H-NMR (CDCl3): 7.66 (d, H C(3)); 7.58 (d, H C(6));
7.49 (dd, H C(4)); 4.27 (m, MeCHCH2O); 3.73 (m, MeCHCH2O, MeO); 1.34 (d, MeCHCH2O).
8. Methyl 2-(2-Nitrophenyl)propyl Carbonate (6). To a cold soln. of MeOH (20 ml) and Et3N (2 ml), 2-(2-
nitrophenyl)propyl carbonochloridate [10] (1.5 g, 6.1 mmol) in CH2Cl2 (5 ml) was added at 0 ± 48. After stirring
for 1 h, the mixture was evaporated and the residue purified by CC (silica gel, 10 Â 2.5 cm, CHCl3): 1.35 g (92%)
of 6. Colorless oil. 1H-NMR ((D6)DMSO): 7.82 (d, 1 H o to NO2); 7.68 ± 7.47 (m, 3 arom. H); 4.26
(m, MeCHCH2O); 3.63 (s, MeO); 3.47 (m, MeCHCH2O); 1.26 (d, MeCHCH2O).

1610
Helvetica Chimica Acta ± Vol. 84 (2001)
9. Photochemical Synthesis of 2-(2-Nitrosophenyl)propan-1,2-diol (8). 2-(2-Nitrophenyl)propan-1-ol (7;
548 mg, 3 mmol) was dissolved in MeCN/H2O 1:1 (300 ml) and irradiated for 70 min at r.t. in a photochemical
reaction vessel equipped with a 150-W high-pressure Hg lamp. Throughout the reaction, the lamp was cooled,
and the soln. was purged with N2 (TLC and UV/VIS monitoring). The solvent was evaporated and the residue
purified by CC (silica gel (18 g), 2 Â 15 cm, petroleum ether/AcOEt 8 :1): 8. Greenish oil. 1H-NMR (400 MHz,
CD3CN): 8.07 (d, H C(3)); 7.76 (t, H C(5)); 7.27 (t, H C(4)); 6.08 (d, H C(6)); 4.03 (m, MeCCH2); 1.85
(s, MeCCH2).
10. Direct Photolysis of 7 in an NMR Tube. In an NMR tube, 7 (50 mg, 0.3 mmol) was dissolved in CD3CN
(0.6 ml) and D2O (0.6 ml). The tube was irradiated by a cooled 150-W high-pressure Hg lamp. 1H-NMR Spectra,
taken before irradiation and after 23, 50, and 75 min, showed the formation of the photoproduct 8 as
characterized above. After 75 min of irradiation, the 1H-NMR showed further secondary photoproducts or by-
products which were not characterized.
11. Time-Resolved Experiments. Nanosecond laser-flash-photolysis experiments were carried out on the
laser system described in [22], however, with a Nd-YAG laser (Spectra Physics, Quanta Ray GCR, 150-5)
operated at its third harmonics (355 nm) replacing the excimer laser. The laser pulse width was 5 ± 7 ns. The
pulse energy was adjusted to ca. 20 mJ over a cross section of about 25 mm2. The laser was run at a repetition rate
of 1 Hz and the transient signals, recorded in the kinetic mode at fixed wavelength, were averaged over typically
4 pulses. The sample solution was contained in a cylindrical flow cell (5-mm cross section, 10-mm path length)
and the flow rate between laser shots adjusted such that the soln. was replaced between any two laser shots.
Details of the laser flash apparatus can be found in [23].
12. Continuous Illumination. For the determination of quantum yields by continuous illumination, a 200-W
HBO lamp combined with a heat filter (5-cm optical pathlength filled with 0.3m CuSO4 in H2O) was used. The
366-nm Hg line was selected by a Schott interference filter. The light intensity was measured by chemical
actinometry by means of the azobenzene method [24]. Standard 10-mm-pathlength stoppered quartz cuvettes
were used. Some quantum yields were determined by irradiation with the pulsed Nd-YAG laser (355 nm). In
these cases, the laser was run at low energy (5 mJ), and the absorbed dose of radiation was measured by an
energy meter (Laser precision RK 3230). HPLC: Merck-Hitachi-7000 equipment with interface, autosampler,
pump, and diode array detector; column Lichrosorb RP-18, linear gradient of MeCN in H2O.
Quantum yields were calculated by Eqn. 1, where the index i represents the starting material in case of ftot
or the released caged compound in case of frel. The concentration change Dci during the time interval Dt was
determined by HPLC. The quantities A, e1, and d denote the total absorbance at the irradiation wavelength in
the middle of the considered time interval of irradiation, the molar absorption coefficient of the starting
compound at the irradiation wavelength, and the optical path length of the cell, respectively; c1 is the
concentration of the starting compound in the middle of the considered time interval of irradiation.
j Dci j
Fi ˆ
e1c1d
A
DtIꢀ1 10
†
A
The quantity c1 was obtained from the HPLC experiment by comparing the peak area of the caged substrate
of an irradiated sample at the wavelength of its maximum absorption with an unirradiated one. In experiments
where thymidine recovery was determined, an aliquot of a thymidine soln. of the same concentration as the
caged substrate was used as a reference. In a typical experiment, 6 ± 8 aliquots of 0.5 mm solns. were irradiated
for different periods to a maximum extent of 50% loss of starting material. Aliquots of 10 or 20 ml were analysed
as described above, and the resulting quantum yields were averaged.
We thank Prof. Dr. Evgeny Kvasyuk and Dr. Sigrid Bühler for the supply of various 2-(2-nitro-
phenyl)ethanol derivatives and Karsten Obert for the synthesis and characterization of 2-(2-nitrosophenyl)-
propane-1,2-diol.
REFERENCES
[1] V. N. R. Pillai, Synthesis 1980, 1; V. N. R. Pillai, in ꢁOrganic Photochemistryꢀ, Ed. A. Padwa, Marcel
Dekker, New York and Basel, 1987, Vol. 9, p. 225.
[2] J. E. T. Corrie, D. R. Trentham, in ꢁBioorganic Photochemistry Seriesꢀ, Ed. H. Morrison, Wiley-
Interscience, 1993, Vol. 2, p. 243.

Helvetica Chimica Acta ± Vol. 84 (2001)
1611
[3] S. P. A. Fodor, J. L. Read, M. C. Pirrung, L. Stryer, A. T. Lu, D. Solas, Science (Washington D.C.) 1991, 251,
767.
[4] G. H. McGall, A. D. Barone, M. Diggelmann, S. P. A. Fodor, E. Gentalen, N. Ngo, J. Am. Chem. Soc. 1997,
119, 5081.
[5] R. J. Lipshutz, S. P. A. Fodor, T. R. Gingeras, D. J. Lockhart, Nat. Gen. Suppl. 1999, 21, 20.
[6] W. G. Bains, C. Smith, J. Theor. Biol. 1988, 135, 303; A. C. Pease, D. Solas, E. J. Sullivan, M. T. Cronin, C. P.
Holmes, S. P. A. Fodor, Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 5022.
[7] G. Ciamician, P. Silber, Chem. Ber. 1901, 34, 2040.
[8] J. A. McCray, D. R. Trentham, Annu. Rev. Biophys. Chem. 1989, 18, 239.
[9] H. Morrison, B. H. Migdalof, J. Org. Chem. 1965, 30, 3996.
[10] A. Hasan, K.-P. Stengele, H. Giegrich, P. Cornwell, K. R. Isham, R. A. Sachleben, W. Pfleiderer, R. S.
Foote, Tetrahedron 1997, 53, 4247.
[11] H. Giegrich, S. Eisele-Bühler, Chr. Hermann, E. Kvasyuk, R. Charubala, W. Pfleiderer, Nucleosides/
Nucleotides 1998, 17, 1987.
[12] G. Wettermark, J. Phys. Chem. 1962, 66, 2560.
[13] R. W. Yip, D. K. Sharma, R. Giasson, D. Gravel, J. Phys. Chem. 1985, 89, 5328; R. W. Yip, D. K. Sharma,
Res. Chem. Intermed. 1989, 11, 109.
[14] S. Walbert, U. E. Steiner, to be published.
[15] Q. Q. Zhu, W. Schnabel, H. Schupp, J. Photochem. 1987, 39, 317.
[16] G. Wettermark, E. Black, L. Dogliotti, Photochem. Photobiol. 1965, 4, 229.
[17] E. Uhlmann, W. Pfleiderer, Helv. Chim. Acta 1981, 64, 1688; S. Alunni, R. Ruzziconi, O. Teofrasti, Res.
Chem. Intermed. 1999, 25, 483.
[18] K. Schaper, D. Dommaschke, S. Globisch, S. A. Madani-Mobarekeh, J. Inf. Recording 2000, 25, 339.
[19] Y. Tsuji, S. Kotachi, K.-T. Huh, J. Org. Chem. 1990, 55, 580.
[20] S. Torii, Y. Murakami, H. Tanaka, K. Okamoto, J. Org. Chem. 1986, 31, 3143.
[21] H. Wieland, L. Horner, Liebigs Ann. 1938, 536, 89.
[22] T. Ulrich, U. E. Steiner, W. Schlenker, Tetrahedron 1986, 42, 6131.
[23] M. Linsenmann, Doctoral Dissertation, University of Konstanz, 1997.
[24] G. Gauglitz, S. Hubig, J. Photochem. 1981, 15, 255; G. Gauglitz, ibid. 1976, 5, 41.
Received April 14, 2001