Photochemistry of 6-amino-2-azido, 2-amino-6-azido and 2,6-diazido analogues of purine ribonucleosides in aqueous solutions

Article abstract of DOI:10.1039/c3pp50385b

The photochemistry of 6-amino-2-azidopurine, 2-amino-6-azidopurine and 2,6-diazidopurine ribonucleosides has been investigated in aqueous solutions under aerobic and anaerobic conditions. Near UV irradiation of 6-amino-2-azido-9-(2′,3′,5′-tri-O-acetyl-β-d- ribofuranosyl)purine and 2-amino-6-azido-9-(2′,3′,5′-tri-O- acetyl-β-d-ribofuranosyl)purine in the presence of oxygen leads to efficient formation of 6-amino-2-nitro-9-(2′,3′,5′-tri-O- acetyl-β-d-ribofuranosyl)purine and 2-amino-6-nitro-9-(2′,3′, 5′-tri-O-acetyl-β-d-ribofuranosyl)purine. Under anaerobic conditions, both azidopurine ribonucleosides preferentially undergo photoreduction to 2,6-diamino-9-(2′,3′,5′-tri-O-acetyl-β-d-ribofuranosyl) purine. The structures of the photoproducts formed were confirmed by UV, NMR and HR ESI-TOF MS spectral data. The photoproducts observed in this study for the aminoazidopurines are distinctly different from those observed previously for 6-azidopurine. When no amino group is present, the photochemistry of 6-azidopurine leads to the formation of a 1,3,5-triazepinone nucleoside. The energetics of the 6-nitreno moiety along both oxidation and ring expansion pathways was calculated using the nudged elastic band (NEB) method based on density functional theory (DFT) using DMol3. The role of the 2-amino group in regulating the competition between these pathways was elucidated in order to explain how the striking difference in reactivity under irradiation arises from the greater spin density on the 6-nitreno-9-methyl-9H-purin-2-amine, which essentially eliminates the barrier to oxidation observed in 6-nitreno-9-methyl- 9H-purine. Finally, the importance of tetrazolyl intermediates for the photochemical activation of azide bond cleavage to release N₂ and form the 6-nitreno group was also corroborated using the DFT methods. The Royal Society of Chemistry and Owner Societies.

Full text of DOI:10.1039/c3pp50385b

Photochemical &
Photobiological Sciences
View Article Online
View Journal | View Issue
PAPER
Photochemistry of 6-amino-2-azido, 2-amino-
6-azido and 2,6-diazido analogues of purine
ribonucleosides in aqueous solutions†
Cite this: Photochem. Photobiol. Sci.,
2014, 13, 563
Krzysztof Komodziński,a Jolanta Lepczyńska,a Zofia Gdaniec,b Libero Bartolotti,c
Bernard Delley,d Stefan Franzene and Bohdan Skalski*a
The photochemistry of 6-amino-2-azidopurine, 2-amino-6-azidopurine and 2,6-diazidopurine ribo-
nucleosides has been investigated in aqueous solutions under aerobic and anaerobic conditions. Near UV
irradiation of 6-amino-2-azido-9-(2’,3’,5’-tri-O-acetyl-β-D-ribofuranosyl)purine and 2-amino-6-azido-
9-(2’,3’,5’-tri-O-acetyl-β-D-ribofuranosyl)purine in the presence of oxygen leads to efficient formation of
6-amino-2-nitro-9-(2’,3’,5’-tri-O-acetyl-β-D-ribofuranosyl)purine and 2-amino-6-nitro-9-(2’,3’,5’-tri-
O-acetyl-β-D-ribofuranosyl)purine. Under anaerobic conditions, both azidopurine ribonucleosides prefer-
entially undergo photoreduction to 2,6-diamino-9-(2’,3’,5’-tri-O-acetyl-β-D-ribofuranosyl)purine. The
structures of the photoproducts formed were confirmed by UV, NMR and HR ESI-TOF MS spectral data.
The photoproducts observed in this study for the aminoazidopurines are distinctly different from those
observed previously for 6-azidopurine. When no amino group is present, the photochemistry of 6-azido-
purine leads to the formation of a 1,3,5-triazepinone nucleoside. The energetics of the 6-nitreno moiety
along both oxidation and ring expansion pathways was calculated using the nudged elastic band (NEB)
method based on density functional theory (DFT) using DMol3. The role of the 2-amino group in regulat-
ing the competition between these pathways was elucidated in order to explain how the striking
difference in reactivity under irradiation arises from the greater spin density on the 6-nitreno-9-methyl-
9H-purin-2-amine, which essentially eliminates the barrier to oxidation observed in 6-nitreno-9-methyl-
9H-purine. Finally, the importance of tetrazolyl intermediates for the photochemical activation of
azide bond cleavage to release N2 and form the 6-nitreno group was also corroborated using the
DFT methods.
Received 5th November 2013,
Accepted 27th December 2013
DOI: 10.1039/c3pp50385b
www.rsc.org/pps
with a variety of biomolecular constituents in their immediate
environment upon near UV activation, their straightforward
Introduction
Nucleosides bearing azido group(s) attached directly to a synthesis and the availability of both enzymatic and chemical
heterocyclic ring belong to a wide range of aryl azides that are procedures for their incorporation into DNA and RNA poly-
commonly used as photoaffinity labeling probes for the study nucleotides, the close structural similarity to their parent
of the structure and function of biological macromolecules.1 nucleobases and their good chemical stability in the absence
Several features of these azidonucleosides, such as their high of light, make these compounds unique in their use as photo-
photochemical reactivity and ability to form covalent bonds crosslinking reagents.2
Understanding the photochemical behavior of azidonucleo-
sides alone, by determining their structures and mechanisms
of photoproduct formation, is important in relation to under-
standing the mechanisms of the reactions and the prediction
aFaculty of Chemistry, Adam Mickiewicz University, Umultowska 89b,
61-614 Poznań, Poland. E-mail: bskalski@amu.edu.pl; Fax: +48-61-829-6761;
Tel: +48-61-829-1585
of structures of the adducts formed under crosslinking con-
ditions. Surprisingly, in spite of the extensive application of
azidonucleosides as photoaffinity probes, their photochemistry
remains rather poorly understood.2 Several photochemical
studies of various azido substituted pyrimidine compounds
have been carried out.3 However, among azidopurine ribonu-
cleosides, only the photochemical reactions of 2-azidoadenosine4
in alcohols and 8-azido- and 6-azidopurine ribonucleosides in
bInstitute of Bioorganic Chemistry, Polish Academy of Science,
Z. Noskowskiego 12/14, 61-704 Poznań, Poland
cDepartment of Chemistry, East Carolina University, Greenville, NC 27858, USA
dPaul-Scherer Institute, CH-5232 Villigen, Switzerland
eDepartment of Chemistry, North Carolina State University, Raleigh, NC 27695, USA
†Electronic supplementary information (ESI) available: The HR ESI-TOF MS,
1H NMR, 13C NMR and 1H–13C HMBC spectral data of photoproducts 5 and 7.
The HPLC analyses of photoirradiated solutions of compound 3. See DOI:
10.1039/c3pp50385b
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 563

View Article Online
Paper
Photochemical & Photobiological Sciences
calculations. Application of computational methods to these
problems has synergy with experimental approaches leading to
a deeper understanding of the mechanistic aspects, which
govern the reactivity of azido-nucleosides.
Scheme 1 The structures of compounds 1–3.
Results and discussion
Photochemistry of azidopurine nucleosides 1–3
aqueous solutions have been investigated.5,6 It has been
shown that the photochemistry of these aryl azides is quite
complex and highly dependent on the position of the azide
group in a heterocyclic ring, the presence of various substitu-
ents, the solvents used and the tendency to exist in the solvent
and pH dependent azide-tetrazole tautomeric equilibria.3 In
the case of 2-azidoadenosine, the formation of 8-substituted
products in addition to the reduction of the nitrene at the
2-position was observed.4 Efficient reduction of the nitrene
occurs also in the case of 8-azidoadenosine, which undergoes
transformation into 8-aminoadenosine upon UV irradiation in
aqueous solutions,5 whereas this is only a minor process in
the case of 6-azidopurine ribonucleoside.6 As we have shown,
the photochemical behavior of the latter resembles closely that
of 4- and 6-azidouracils,3e,h namely it undergoes efficient
photochemical purine ring expansion leading to the novel imi-
dazole-fused 1,3,5-triazepinone nucleoside.6 Furthermore, our
recent study showed that selective excitation of 6-azidopurine
incorporated into single stranded DNA oligonucleotide leads
to the analogous purine ring expanded product.7 These large
differences in reactivity have not been explained.
Aqueous solutions of compounds 1–3 were irradiated at λ >
300 nm under aerobic and anaerobic conditions at room temp-
erature and the progress of the photoreactions was monitored
by both UV-Vis spectroscopy and HPLC. As shown in Fig. 1,
irradiation of an aqueous solution of 1, under both aerobic
and anaerobic conditions, results in almost complete conver-
sion after 12 min, indicating that the presence of oxygen has
virtually no effect on the relative quantum yield of disappear-
ance of 1. The HPLC analysis of the solution of 1 irradiated
under aerobic conditions showed the presence of two peaks, 4
and 5, corresponding to the minor and the major photopro-
duct, respectively, whereas in the case of irradiation of 1 under
anaerobic conditions, the formation of 4 as the only major
photoproduct was observed. The photoproducts 4 and 5 were
isolated and purified by preparative reversed-phase HPLC. The
In this contribution, we report on a study of the photo-
chemistry of 2-azido-6-aminopurine-, 6-azido-2-aminopurine-
and 2,6-diazidopurine ribonucleosides 1–3 (Scheme 1) in
aqueous solution under aerobic and anaerobic conditions.
As will be shown, none of these azidopurines undergoes
ring expansion. Instead, depending on the conditions of
irradiation, photoreactions of the two amino-substituted
derivatives are dominated by photooxidation and/or photo-
reduction of the nitrene leading to respective nitro and/or amino
derivatives (Scheme 2). We have rationalized the differences
in photochemical reactivity of the 6-azidopurine and 6-azido-
2-aminopurine systems, based on quantum chemical
Fig. 1 Changes in the absorption spectrum of an aqueous solution of 1
measured with 1 min increments during irradiation at λ > 300 nm under
aerobic (A) and anaerobic (B) conditions together with HPLC analyses of
the solutions (inset) before (dotted red lines) and after 12 min irradiation
(solid black lines). (C) The normalized absorption spectra of the photo-
products 4 and 5.
Scheme 2 Azide-tetrazole tautomeric equilibrium of
1 in aqueous
solution10 and structures of photoproducts of its irradiation under
aerobic conditions.
564 | Photochem. Photobiol. Sci., 2014, 13, 563–573
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Photochemical & Photobiological Sciences
Paper
photoproduct 4 was identified as 2,6-diamino-9-(2′,3′,5′-tri-O-
acetyl-β-D-ribofuranosyl)purine based on HR ESI-TOF MS and
NMR spectral data and comparison of its UV spectrum with
the published one.8
The molecular formula of 5 (C16H18N6O9), inferred from the
high-resolution ESI-TOF MS spectrum, indicated that the for-
mation of this photoproduct involves the loss of a N2 molecule
and addition of an O2 molecule. As shown in Fig. 1C the UV
spectrum of photoproduct 5 displays three maxima at λ
236 nm, 262 nm, and 332 nm characteristic of the 2-nitroade-
nosine chromophore.9 The identity of 5 with the latter was
further supported by the 1H and 13C NMR spectra.
The 1H NMR spectrum of 5 showed the signal patterns
typical of the intact O-acetylated ribose and two signals at 8.19
and 6.18 ppm corresponding to C8–H and C6–NH2 of the
purine aglycone, respectively. The 13C NMR spectrum dis-
played five carbon signals in its aromatic region, one methine
carbon (C8) at 141.8 ppm and four quaternary carbons at
155.9, 155.4, 149.8 and 122.1 ppm. The 1H–13C HSQC spec-
trum showed a correlation of H8 (8.19 ppm) with C8
(141.8 ppm) and the 1H–13C HMBC spectrum showed corre-
lations of H8 (8.19 ppm) with C4 (149.8 ppm) and C5
(122.1 ppm); H1′ (6.26 ppm) with C4 (149.8 ppm) and C8
(141.8 ppm).
The photochemical behavior of the 2-amino-6-azidopurine
nucleoside 2 is similar to that of its 6-amino-2-azido isomer 1.
As shown in Fig. 2, irradiation under both aerobic and anaero-
bic conditions leads to complete conversion of 2 after 20 min
with the relative quantum yield of disappearance similar to
that of 1. As in the case of 1, the presence of oxygen has an
effect only on the type and distribution of the photoproducts
formed.
Fig. 2 Changes in the absorption spectrum of an aqueous solution
of 2, measured with 1 min increments, during irradiation at λ > 300 nm
under aerobic (A) and anaerobic (B) conditions together with HPLC
analyses of the solutions (inset) before (dotted red lines) and after
20 min irradiation (solid black lines). (C) The normalized absorption
spectra of photoproducts 4, 6 and 7.
In the case of anaerobic irradiation of 2 the formation of
2,6-diaminopurine nucleoside 4 as the only major product
(>90%) is observed, whereas under aerobic conditions, photo-
oxidation of the azido group leading to the 6-nitro-2-amino-
purine nucleoside 7 (50%) dominates, and photoreduction to 4
occurs to a much lesser extent (40%) (Scheme 3).
The proposed structure of 7, which is a structural isomer of
5, is in full agreement with the molecular formula C16H18N6O9
determined by high-resolution ESI-TOF MS. As in the case of
5, the 1H NMR and 13C NMR spectra of 7 show the peaks
characteristic of the sugar and purine units of the molecule. In
the 1H NMR spectrum, two signals corresponding to C8–H and
N–H2 protons appear at 8.17 ppm and 5.62 ppm, respectively,
in addition to resonances due to the ribose unit. In the aro-
matic region of the 13C NMR spectrum of 7, five signals
appear, one from the methine carbon C8 (144.8 ppm) and four
from quaternary carbons at 158.6, 158.3, 152.4 and 119.5 ppm.
Heteromolecular carbon–proton one bond correlation was
detected between C8 (144.8 ppm) and H8 (8.17 ppm). The
1H–13C HMBC spectrum showed correlations between H8
(8.17 ppm) and both C5 (119.5 ppm) and C4 (152.4 ppm);
anomeric H1′ (6.11 ppm) with C4 (152.4 ppm) and C8
(144.8 ppm). Similarly, like in the case of 5, the UV absorption
spectrum of 7 displays the presence of an intense absorption
Scheme 3 Azide-tetrazole tautomeric equilibrium of 211 and structures
of the photoproducts of its irradiation under aerobic conditions.
band at 230 nm (ε = 14 900 M−1 cm−1) and a low intensity,
broad absorption band centered at 386 nm (ε = 2900 M−1
cm−1) (Fig. 2C). To the best of our knowledge, the 2-amino-6-
nitropurine chromophore has not been described previously in
the literature.12
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 565

View Article Online
Paper
Photochemical & Photobiological Sciences
The minor photoproduct 6 of the aerobic irradiation of 2 Comparison of the formation of the 6-nitreno-9-methyl-9H-
(Fig. 2a) was identified as 2′,3′,5′-tri-O-acetylguanosine by com- purine (6NP) and 6-nitreno-9-methyl-9H-purin-2-amine
parison with authentic samples (UV spectrum and HPLC). (2A6NP)
Upon prolonged irradiation of the reaction mixture, an
increase in the amount of 6 formed with a concomitant,
gradual disappearance of 7 was observed indicating that the
The photochemical reactions of 6-azido-9-methyl-9H-purine
(6AzP) and 6-azido-9-methyl-9H-purin-2-amine (2A6AzP) shown
in Scheme 5 compare the formation of a tetrazole structure in
latter undergoes further phototransformation into 6. This was
the two species, called TP and 2ATP, respectively, and the pho-
further supported by direct irradiation of an isolated, pure
tolysis process that leads to the formation of 6NP in each case.
First, the cyclizations of 6AzP and 2A6AzP to form a tetrazole
sample of 7 under identical conditions (Fig. S9, ESI†).
In contrast to monoazide substituted purine nucleosides 1
ring in the cyclized form TP and 2ATP were considered. Follow-
and 2, irradiation of the 2,6-diazidopurine derivative 3 leads to
ing irradiation, the stable tetrazole intermediates, TP and
2ATP, lose N2 to form reactive nitrene intermediates, 6-nitreno-
its very rapid disappearance with the formation of a complex
mixture of more than 20 photoproducts (Fig. S10, ESI†) indi-
9-methyl-9H-purine (6NP) and 6-nitreno-9-methyl-9H-purin-2-
cating its extensive degradation.
amine (2A6NP). Thus, this stage can be compared to the direct
loss of the N2 molecule from photoexcited 6AzP.
The formation of nitrene has two alternative pathways.
Direct cleavage of the N–N bond can lead to loss of N2 and for-
mation of 6-nitreno intermediate as shown in Fig. 3. The
thermal energy barrier is 1.82 eV for this process.
Computational study of the photochemical pathways of
6-azidopurine and 2-amino-6-azidopurine
The above described photochemical pathways of amino-substi-
The alternative pathway for the formation of 6NP and
tuted azidopurines 1,2 differ significantly from that of the
2A6NP involves cyclization to form a tetrazole ring as shown in
unsubstituted 6-azidopurine nucleoside (8) despite the fact
that the primary photochemical event in each of these com-
pounds involves elimination of N2 leading to a highly reactive
intermediate nitrene in the 6-position. It is indeed remarkable
that 2-amino-6-nitrenopurine proceeds with high yield to form
the 6-nitro group (7), while 6-nitrenopurine under analogous
photoirradiation conditions reacts nearly quantitatively to
form the 1,3,5-triazepin-2-one (9). These distinct paths are
shown in Scheme 4.
Scheme 4 and the NEB calculation for this pathway is shown
in Fig. 4.
The loss of N2 from TP and 2ATP is shown in Fig. 5. This
process parallels the loss of N2 from 6AzP shown in Fig. 3. The
difference is seen in the first few frames of the NEB calcu-
lation. The small increase in the energy represents the
In order to determine the origin of the difference in reactiv-
ity between 6-azidopurine (8) and 2-amino-6-azidopurine (2)
nucleosides, the photochemical reactions of 9-methyl analogs
of both nucleosides (6AzP, 2A6AzP) were studied using density
functional theory (DFT) as implemented in the DMol3 code13
and nudged elastic band (NEB) calculation of the reaction
path. There are two distinct phases for comparison. First, the
Scheme 5 Azide-tetrazole tautomeric equilibrium of 6AzP and 2A6AzP
formation of a tetrazole structure and the photolysis process
with TP and 2ATP, respectively, in aqueous solution.
that leads to formation of the 6-nitrenoadenine were studied
for both the purine and 2-aminopurine. Then the analysis of
the reaction path for oxidation was conducted by comparison
with the alternative pathway for the formation of an imidazole-
fused 1,3,5-triazepinone nucleoside.
Fig. 3 Thermal dissociation of 6AzP calculated using the NEB method.
The structures corresponding to the reactant, transition state and
Scheme 4
product for the dissociation are shown immediately above the plot.
566 | Photochem. Photobiol. Sci., 2014, 13, 563–573
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Photochemical & Photobiological Sciences
Paper
transition corresponds to a dissociative process with respect to
loss of N2 for both molecules. The energies of the states
involved in the photolysis of the 2-aminopurine tetrazole struc-
ture, 2ATP, are shown in Fig. S13 (ESI†). It is evident that the
features of TP are similar to those in 2ATP. There are excited
state curve crossings associated with the breaking of the first
bond of the tetrazole ring (Fig. S13†). These facilitate passage
of the system into the photolytic state. While this analysis is
not rigorous it provides sufficient justification for our pur-
poses, which is to differentiate the bent and cyclized (tetrazo-
lyl) forms of the purines and provide a reasonable hypothesis
for the photolytic state. We conclude that the photolytic state
is most likely the tetrazolyl form (TP or 2ATP). Secondly, our
goal was to determine whether there is a substantial difference
due to the presence of a 2-amino group on the purine photo-
chemistry. Based on the similarity of the NEB trajectories and
the excited state curve crossings observed at this level of
theory, we conclude that the mechanisms of photolysis for
both TP and 2ATP are essentially the same. Clearly the excited
state behavior will need to be studied using multiconfiguration
methods to confirm the details of this conclusion and such
studies will be forthcoming in our research. Since the for-
mation of the tetrazole ring and photolysis appear quite
similar in TP and 2ATP these calculations support the hypoth-
esis that the 6-nitreno group is prepared in a similar way irre-
spective of the presence of the 2-amino group. We conclude
that the differences in reactivity of 6AzP and 2A6AzP are due to
events that occur after photolysis, but there are no substantive
differences in the photolytic step.
Fig. 4 Comparison of the cyclization of the tetrazole ring in 6AzP and
2A6AzP calculated using the NEB method. The 6AzP structures corres-
ponding to the reactant, transition state and product for the cyclization
are shown immediately above the plot. The analogous 2A6AzP struc-
tures are omitted for clarity.
Oxidation of nitrene intermediates
The oxidation path was chosen for analysis since O2 is the
obvious oxidant and there is a clear starting point that involves
approach by O2 to the 6-nitreno group. Thus, oxidation can be
readily compared to the ring expansion pathway, while
reduction is more difficult to model since the reductant is not
known.14 In order to understand the two competing reaction
pathways, parallel calculations were set up as three distinct
NEB calculations along the two competing reaction paths for
each molecule. Path 1, which involves oxidation to the 6-nitro
group, was separated into two stages since both 6NP and
2A6NP have a stable but high-energy 6-dioxaziridine inter-
Fig. 5 Comparison of the thermal dissociation of N2 from TP to form
6NP and 2ATP to form 2A6NP calculated using the NEB method. The TP
structures corresponding to the reactant, transition state and product
for the dissociation are shown immediately above the plot. The ana-
logous 2ATP structures are omitted for clarity.
breaking of the first bond between the γ or terminal nitrogen,
Nγ, of the azide and the N(5) nitrogen on the purine ring. Once
this has taken place, the trajectory is quite similar to the pho- mediate, designated 6IP, along the oxidation pathway
tolysis of N2 from azide in 6AzP except that the thermal bar-
riers are 1.69 eV and 1.65 eV for TP and 2ATP, respectively. In
all cases the release of N2 is highly endothermic.
(Scheme 6). Thus, it is reasonable to divide the NEB calcu-
lation for the oxidation part into two separate stages with the
6AzP, TP and 2ATP can each undergo photolysis of N2 to
form the corresponding 6NP intermediate. There are two
reasons why TP may be favored over the 6AzP form as the
species that leads to photolysis. First, the TP isomer may be
favored in solution. Second, the transition to the photolytic
state may be favored in TP. The photoexcitations of 6AzP and
TP were compared using TD-DFT calculations in order to
determine which of the transition moments is dominant. The
nominal transitions are mainly 45→46 (i.e. HOMO→LUMO)
Scheme 6 Steps involved in path 1, which is the oxidation of the
for both 6AzP and TP. As shown in Fig. S12 (ESI†), this 6-nitreno group to the 6-nitro group.
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 567

View Article Online
Paper
Photochemical & Photobiological Sciences
intermediate as a stable point in the path. First is the associ- structure (6NO2P) is rotated with respect to the plane of the
ation of O2 to form the stable high-energy 6-dioxaziridine purine so that the N–C–N–O dihedral angle is 122.9°. This
intermediate, which we call 6IP from a transition state struc- angle is essentially the same for the 2A6NO2P structure.
ture called 6SP. Scheme 6 shows the parallel processes for the
On the path to the formation of either 6NO2P or 2A6NO2P
2-aminopurine, which involves the transformation from 2A6SP from the starting complexes 6NP + O2 or 2A6NP + O2, respect-
to 2A6IP. The second step on path 1 is the formation of the ively, there is a stable intermediate, designated 6IP or 2A6IP,
6-nitro group from the stable 6-dioxaziridine intermediate. shown in Fig. 6. Based on this observation, the oxidation
Path 2 is the competing ring expansion, which occurs in a pathway was divided into two NEB calculations, which are
single step.
shown together in Fig. 7. In the first calculation the starting
The high energy intermediate (6IP) shown in Fig. 6 has the point has O2 approximately 4 Å from the nitreno group, 6NP +
appearance of a triangular molecule bent away from the plane O2 or 2A6NP + O2, and an endpoint at the stable intermediate,
of the purine ring. It is designated 6IP to distinguish it from 6IP or 2A6IP. The second NEB calculation begins from 6IP or
the product 6-nitropurine, 6NO2P. The intermediate 6IP can be 2A6IP and ends at 6NO2P or 2A6NO2P. Thus, the molecule
geometry optimized and is clearly at a local minimum in the must pass through the stable 6-dioxaziridine intermediate
potential energy surface. It has defined vibrational frequencies structure on the way from the addition of O2 to the formation
with no aberrant eigenvalues meaning that all of its normal of 6NO2P or 2A6NO2P. It is clear from inspection that the final
modes are well defined (see ESI†). However, the energy of 6IP stage of the process involving the rearrangement from 6IP/
is higher than that of the 6NO2P structure. Table 1 shows that 2A6IP → 6NO2P/2A6NO2P has the same energy change in both
the 2-amino group does not appear to have a large effect on the 6NP and 2A6NP. Fig. 7 strongly suggests that the 6NO2P/
the bond lengths or angles of either the intermediate 6IP or 2A6NO2P product will be formed if it reaches 6IP/2A6IP. Since
the equilibrium 6-nitropurine structure (6NO2P).
this product is only formed in 2-aminopurine derivative, and
The effective change along the reaction path from 6IP to the alternative ring-expansion takes place in the purine deriva-
6NO2P is a rotation of one of the O atoms (labeled O(1)). The tive, we conclude that the differences in reactivity occur in the
remaining atoms are in nearly the same positions in the two preceding steps that lead to the formation of the 6IP/2A6IP
structures. The O atoms in the intermediate 6IP form are tilted intermediates.
back away from the plane of the purine. The final equilibrium
In order to investigate the significance of the different
peaks and valleys in the energy of the trajectory leading to the
formation of the intermediate, three points in the NEB trajec-
tory that were near maxima or minima were chosen for further
refinement using the program DIMER, which minimizes the
gradient of the structures along the reaction path and thereby
locates a saddle point along the trajectory.15 Once the saddle
points were located, the Hessian was calculated for the saddle
point geometry. In theory, there should be a single negative
eigenvalue corresponding to the point where the minimum
energy path from reactants to products reaches the transition
Fig. 6 Stick bond figure showing (A) the stable 6-dioxaziridino-
9-methyl-9H-purine intermediate (6IP), (B) 6-nitro-9-methyl-9H-purine
(6NO2P), (C) stable 6-dioxaziridino-9-methyl-9H-purin-2-amine inter-
mediate (2A6IP), and (D) 6-nitro-9-methyl-9H-purin-2-amine (2A6NO2P)
structures.
Table 1 Bond lengths and angles of the saddle point structures (6SP,
2A6SP), intermediates 6IP/2A6IP and equilibrium 6-NO2 structure in
both 6NO2P and 2A6NO2P
Bond/Å
Angle/°
N–O(1)
N–O(2)
O(1)–O(2)
C–N–O(1)
C–N–O(2)
6SP
2A6SP
6IP
2A6IP
6NO2P
2A6NO2P
1.38
1.44
1.44
1.45
1.23
1.23
1.39
1.44
1.45
1.44
1.25
1.23
1.74
1.48
1.49
1.49
2.21
2.21
117.9
110.1
109.8
109.9
115.9
115.9
116.6
110.0
109.5
109.8
116.9
117.1
Fig. 7 Energy of the structures in the NEB calculation of the oxidation
pathway for 6-nitreno-9-methyl-9H-purine (6NP) and 6-nitreno-9-
methyl-9H-purin-2-amine (2A6NP). The saddle points calculated using
DIMER were added to the NEB calculation to obtain a more accurate
estimate of the barriers along the reaction path.
568 | Photochem. Photobiol. Sci., 2014, 13, 563–573
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Photochemical & Photobiological Sciences
Paper
state. In practice, there can be additional negative eigenvalues bound O2 rearranges to intermediate 6IP. The structure in
and it is advisable to examine the nature of the saddle point Fig. 8D is a saddle point formed after 6IP on the path to
using intrinsic reaction coordinate (IRC) analysis. IRC analysis 6NO2P. There is a barrier that must be surmounted on the
consists of projections at the transition state along specific path from 6IP to 6NO2P, but this further aspect of the reactivity
normal modes whose Hessian is negative until the gradient will be considered in a future publication that focuses on a
along that mode is zero. Such a point should correspond to detailed look at the intrinsic reaction coordinate.
the reactant or product for reverse and forward projections of
The trajectory for O2 addition in the case of 6-nitreno-9-
the mode, respectively. IRC analysis validates the transition methyl-9H-purin-2-amine consists of a series of 3 saddle
state and identifies the mode that leads from the reactants to points shown in Fig. 9. As was done above for the NEB trajec-
the transition state (and on to the products).
tory of 6NO2P, the structures in Fig. 9 were obtained by choos-
In the calculations presented here, the maximum number ing selected pairs of structures in the NEB calculation shown
of negative eigenvalues was three. When more than one nega- in Fig. 7 and subjecting them to further refinement using the
tive eigenvalue was found, there was a mode that involved program DIMER.13 There is a complex set of rearrangements
motions that appear to be along a reaction coordinate, but one as the O2 atom moves into position to form the meta-stable
of the additional negative eigenvalues was found to be a 2A6NO2P form. The structures along this pathway for the
methyl rotation. In the two cases where three negative eigen- 2A6NP + O2 calculation given in Fig. 6 show that the O2 atom
values were found, the third eigenvalue corresponded to a low begins the process of addition to nitrene by an attack perpen-
frequency torsional mode. Using IRC analysis one can validate dicular to the purine ring. Subsequently, the O2 atom rotates
the saddle points and determine whether the character of the to become coplanar with the purine ring. Finally, the O2 moves
saddle points (transition states) in each molecule is similar or to an orientation that is parallel, but shifted out of the purine
different. Such an analysis provides insights into the nature of ring plane. While the structures in Fig. 9B and 9C resemble
saddle points and helps to define which points along the reac- the structures in Fig. 9A and 9B, respectively, they differ in
tion coordinate to true transition states.
order and orientation of the O2 molecule.
Three distinct saddle points were found for both the 6-
The second part of the trajectory involves the rearrange-
nitrenopurine and 6-nitreno-9-methyl-9H-purin-2-amine systems ment of the structure 6IP/2A6IP, an intermediate O2 bonded to
when O2 is allowed to react with these molecules. However, the nitrene N atom in a triangular fashion. While there are
despite their superficial similarities, the intrinsic reaction comparable barriers prior to the formation of this intermedi-
coordinates and barrier heights for these saddle points were ate, the greatest energy barrier is found for the structure in
significantly different. The calculated structure for the local Fig. 8D, which is formed after 6IP in the trajectory. The
energy minimum at the frame of the trajectory for 6NP + O2 6-nitreno-9-methyl-9H-purine has a barrier that is 0.42 eV higher
led to a saddle point structure shown in Fig. 8A. This structure than the barrier for 6-nitreno-9-methyl-9H-purin-2-amine. No
consists of an addition of O2 to the 6-nitreno group from the unique saddle point structure was found for a similar calcu-
side with an attack perpendicular to the plane of the purine. lation following 2A6IP. Instead the saddle point structure con-
This structure rearranges to the one shown in Fig. 8B, which verged on the stable intermediate. This difference in barrier
consists of a linear O2 that oscillates from side to side main- following the formation of the stable intermediate is signifi-
taining an orientation parallel to the C–N(6) bond. Finally the cant since this barrier must slow the formation of the 6-NO2
group. Consequently, the competing formation of the imida-
zole-fused 1,3,5-triazepinone nucleoside can take place in
6-nitreno-9-methyl-9H-purine, whereas the 6-NO2 group is
formed with high yield in 6-nitreno-9-methyl-9H-purin-2-amine.
In fact, there are three significant differences between the trajec-
tories in the two cases. There is an energy barrier between
Fig. 8 Saddle point structures along the reaction pathway for the
6NP + O2 reaction. The binding energies of these saddle point structures
are (A) −88.10 eV, (B) −88.60 eV, (C) −88.6 eV and (D) −87.75 eV,
respectively, compared to −91.52 eV for 6-nitro-9-methyl-9H-purine
(6NO2P). The energies of these structures are (A) 3.42 eV, (B) 2.92 eV,
(C) 2.92 eV and (D) 3.77 eV greater than the energy of 6NO2P.
Fig. 9 Saddle point structures along the reaction pathway for the
2A6NP + O2 reaction. The binding energies of these saddle point struc-
tures are (A) −97.15 eV, (B) −96.86 eV and (C) −96.25 eV, respectively,
compared to −99.80 eV for 6-nitro-9-methyl-9H-purin-2-amine
(2A6NO2P). The energies of these structures are (A) 2.65 eV, (B) 2.94 eV,
and (C) 3.55 eV greater than the energy of 2A6NO2P.
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 569

View Article Online
Paper
Photochemical & Photobiological Sciences
intermediates A and B in 6NP, but 2A6NP. This barrier appears in the case of purine and 2-aminopurine, respectively. In both
in the NEB calculation as well and is evident in Fig. 7. The cases, the transition state structures consist of a bent 6-nitrene
second difference is at intermediate C. In 6NP, intermediate C that is poised to insert into the ring with a similar geometry.
consists of a double well potential. Although it is quite shallow, The Mulliken charge and spin density for reactant structures,
it is still significantly deeper than the potential in 8C.
transition state and ring-expanded product structures are given
in Tables S3 and S4 of the ESI.† The major difference between
the two species is the spin density on the nitrene, which is sig-
nificantly greater for 2-aminopurine than for purine.
The reaction trajectory along the ring expansion pathway
The alternative pathway involves insertion of the nitrene into
the purine ring. The starting and ending structures for this
pathway are shown in Fig. 10, along with the transition state
structure. Fig. 10 shows the purine structures for comparison.
It is evident that for the ring expansion to occur in the case of
6-nitreno-9-methyl-9H-purin-2-amine there must be an inver-
sion of the 2-amino group. However, this does not necessarily
result in an increased energy barrier for the ring opening
process. The energies of these structures and other intermedi-
ate structures along this pathway are shown in Fig. 11. The
starting energies are 1.08 and 1.33 eV, respectively, for the
purine and 2-aminopurine structures. Relative to these initial
energies, the barriers to ring expansion are 0.197 and 0.094 eV
The energy barrier for ring expansion of 2A6NP shown in
Fig. 11 is smaller than that for 6NP, which may appear to favor
the formation of a ring-expanded form. However, the differ-
ence in spin density shown in Tables S3 and S4† creates a spin
barrier insertion of the nitrene into an expanded ring for
2A6NP relative to 6NP. The required spin change contributes
to a smaller electronic factor for 6-nitreno-9-methyl-9H-purin-
2-amine than for 6-nitreno-9-methyl-9H-purine, which slows
the kinetics of ring expansion. Both the oxidation and ring
expansion pathways have lower barriers for 6-nitreno-9-methyl-
9H-purin-2-amine. However, it is the competition between the
pathways in each molecule that determines the product
distribution.
Conclusions
In conclusion, the above results showed that the near UV light
(λ
> 300 nm) irradiation of 2-amino-6-azidopurine and
6-amino-2-azidopurine ribonucleosides in aqueous solutions in
the presence of oxygen led to the efficient formation of their
2- and 6-nitro derivatives. As in the case of aryl azides, the most
likely mechanism accounting for the formation of nitropurine
ribonucleosides involves a triplet nitrene intermediate that
reacts with the O2 molecule.16 DFT calculations that compare
the reactivity of the 2A6NP with 6NP show that the nitrene
spin density is significantly higher in the former species.
Although the 2-amino group lowers the barrier for oxidation
and the barrier for expansion, oxidation is kinetically favored
for two reasons. First, the spin density on 6NP is much larger
than in 2A6NP. Second, there is essentially no barrier for oxi-
dation, while there is still a small barrier for ring expansion in
2A6NP. The computational results provide a clear interpret-
ation of the experimental results and explain the remarkable
difference in reactivity of these two species, which react quanti-
tatively along different pathways.
Fig. 10 Three structures along the ring-opening pathway are shown.
The reactant, transition state and product for both 6-nitreno-9-methyl-
9H-purin-2-amine (A, B and C respectively) and 6-nitreno-9-methyl-
9H-purine (D, E and F, respectively).
For the first time, these types of photoproducts were
observed via UV irradiation of azidonucleosides. The above
observations make a significant contribution to understanding
the photochemical behavior of the azidopurine ribonucleo-
sides incorporated into oligonucleotides used as photoaffinity
labeling probes in the photocross-linking technique.
Experimental section
Synthetic and analytical methods
Fig. 11 Energy of the structures in the NEB calculation of the purine
ring expansion pathway for 6AzP and 2A6AzP. Three key structures for
this pathway, reactant, transition state and product for the 6AzP are
shown. The analogous 2A6AzP structures are omitted for clarity.
HPLC analysis was performed with an Agilent 1260 Infinity
system equipped with diode-array UV-vis detectors. The
570 | Photochem. Photobiol. Sci., 2014, 13, 563–573
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Photochemical & Photobiological Sciences
Paper
column was a ZORBAX SB-C18, 5.0 μm, 9.1 × 150 mm, eluted
2,6-Diazido-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)purine
with H2O–CH3OH using a linear gradient of 16–50% of (3). Starting nucleoside: 6-chloro-2-iodo-9-(2′,3′,5′-tri-O-acetyl-
CH3OH over 25 min at a flow rate of 1.8 mL min−1. Absorption β-D-ribofuranosyl)purine. Compound 3 was obtained as yellow
spectra were measured using a Cary 300 Bio Varian spectro- foam (0.61 mmol, 61%). UV (CH3CN) λmax/nm (ε/M−1 cm−1):
photometer in CH3CN. The HRMS analyses of starting com- 244.0 (21 000), 266.0 (8700) and 296.0 (11 200). 1H NMR:
pounds 1–3 and the formed photoproducts were performed (CDCl3) δ 8.06 (s, 1H, H8), 6.15 (d, J = 5.1 Hz, 1H, H1′), 5.83 (t,
using an ESI-TOF system (Bruker). The NMR spectra were J = 5.5 Hz, 1H, H2′), 5.60 (t, J = 5.3 Hz, 1H, H3′), 4.47 (m, 1H,
measured in CDCl3 on a Bruker Avance III 700 spectrometer H4′), 4.42 (m, 2H, H5′, H5″), 2.17–2.08 (s, 9H, Ac). 13C NMR:
1
operating at 700 MHz for H and 175 MHz for 13C. The QCI-P (CDCl3) δ 170.2–169.2 (CvO, Ac), 156.4 (C2), 154.1 (C6), 153.3
CryoProbe™, 5 mm, was used.
(C4), 141.5 (C8), 121.7 (C5), 86.2 (C1′), 80.2 (C4′), 73.0 (C3′),
70.3 (C2′), 62.8 (C5′), 20.7–20.3 (CH3–Ac). HRMS (ESI-TOF):
calcd for C16H17N10O7 [M + H]+ 461.1276, found 461.1280.
General procedure for synthesis and spectral data for the
substrates
General procedure for UV irradiation
Azidopurine nucleosides 1–3 were synthesized via their
respective 2- and 6-halogenopurine analogs17 according to a Solutions of compounds 1–3 in an H2O–CH3CN mixture (9/1,
published procedure.18 The respective, starting tri-O-acetylated v/v) were photoirradiated at λ > 300 nm under aerobic and
halogenopurine nucleoside (1.0 mmol) was dissolved in anhy- anaerobic (Ar atmosphere) conditions in a 0.2 cm UV cell
drous DMF (20 mL) and treated with NaN3 (4.0 mmol). The using a 200 W, high-pressure Hg lamp (HBO 200) equipped
reaction mixture was heated at 50 °C for 2 h, cooled, filtrated with a 313 nm interference filter, at room temperature. The
and concentrated to dryness. The solid residue was dissolved progress of the photoreaction was monitored by the disappear-
in CH2Cl2 and washed with water. The organic phase was ance of the absorption band in the UV spectra of compounds
dried over MgSO4 and evaporated to dryness. The crude 1–3 and by HPLC. To isolate and fully characterize the photo-
product was purified by silica gel column flash chromato- products by spectral methods, solutions of compounds 1–3
graphy using a gradient of CH3OH in CH2Cl2.
(5.0 × 10−4 mol dm−3) were irradiated in a 200 mL photoreac-
6-Amino-2-azido-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)- tor with a 200 W, high-pressure immersion Hg lamp (HBO
purine (1). Starting nucleoside: 6-amino-2-iodo-9-(2′,3′,5′-tri-O- 200) equipped with a cut-off Pyrex filter (λ > 300 nm). The
acetyl-β-D-ribofuranosyl)purine. Compound 1 was obtained as photoproducts formed were isolated and purified by prepara-
white foam (0.75 mmol, 75%). UV (CH3CN) λmax/nm (ε/M−1 tive reversed-phase HPLC.
cm−1): 232.0 (25 600), 268.0 (9000), 310.0 (3500) and 322.0
(2700). H NMR: (CDCl3) δ 7.88 (s, 1H, H8), 6.13 (d, J = 4.9 Hz,
1
Spectral data for the photoproducts
1H, H1′), 5.87 (t, J = 5.2 Hz, 1H, H2′), 5.74 (s, 2H, NH2), 5.67
2,6-Diamino-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)purine
(m, 1H, H3′), 4.47 (m, 1H, H5′), 4.45 (m, 1H, H4′), 4.40 (m, 1H, (4). UV (CH3CN) λmax/nm (ε/M−1 cm−1): 216.0 (16 300), 258.0
H5″), 2.18–2.11 (s, 9H, CH3–Ac). 13C NMR: (CDCl3)
δ
(8600) and 280.0 (8800). H NMR: (CDCl3) δ 7.64 (s, 1H, H8),
1
170.3–169.4 (CvO–Ac), 157.2 (C6), 156.0 (C2), 151.0 (C4), 138.4 5.99 (m, 1H, H1′), 5.97 (t, J = 4.7 Hz, 1H, H2′), 5.81 (t, J = 5.2
(C8), 117.5 (C5), 86.3 (C1′), 80.0 (C4′), 73.2 (C2′), 70.4 (C3′), Hz, 1H, H3′), 5.69 (s, 2H, NH2-2), 4.92 (s, 2H, NH2-6), 4.46 (m,
63.1 (C5′), 20.8–20.4 (CH3–Ac). HRMS (ESI-TOF): calcd for 1H, H5′), 4.42 (m, 1H, H4′), 4.37 (m, 1H, H5″), 2.14–2.09 (s,
C16H19N8O7 [M + H]+ 435.1377, found 435.1377.
9H, CH3–Ac). 13C NMR: (CDCl3) δ 170.6–169.5 (CvO–Ac), 159.9
2-Amino-6-azido-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)- (C6), 155.9 (C2), 151.6 (C4), 136.4 (C8), 114.6 (C5), 86.2 (C1′),
purine (2). Starting nucleoside: 2-amino-6-chloro-9-(2′,3′,5′-tri- 79.8 (C4′), 72.9 (C2′), 70.6 (C3′), 63.1 (C5′), 20.8–20.5 (CH3–Ac).
O-acetyl-β-D-ribofuranosyl)purine. Compound 2 was obtained HRMS (ESI-TOF): calcd for C16H21N6O7 [M + H]+ 409.1472,
as white foam (0.7 mmol, 70%). UV (CH3CN) λmax/nm (ε/M−1 found 409.1471.
cm−1): 238.0 (6200), 271.0 (7800) and 299.0 (8500). 1H NMR:
6-Amino-2-nitro-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)-
(CDCl3) δ tetrazole form: 8.02 (s, 1H, H8), 6.62 (s, 2H, NH2), purine (5). UV (CH3CN) λmax/nm (ε/M−1 cm−1): 236.0 (12 400),
1
6.01 (m, 1H, H1′), 5.97 (m, 1H, H2′), 5.78 (m, 1H, H3′), 4.45 262.0 (8000) and 332.0 (2700). H NMR: (CDCl3) δ 8.19 (s, 1H,
(m, 1H, H5′), 4.43 (m, 1H, H4′), 4.38 (m, 1H, H5″), 2.17–2.10 (s, H8), 6.26 (d, J = 5.5 Hz, 1H, H1′), 6.18 (s, 2H, NH2), 5.79 (t, J =
9H, CH3–Ac); azido form: 7.99 (s, 1H, H8), 6.13 (m, 1H, H1′), 5.5 Hz, 1H, H2′), 5.67 (m, 1H, H3′), 4.54 (m, 1H, H4′), 4.50 (m,
6.00 (m, 1H, H2′), 5.74 (m, 1H, H3′), 5.13 (s, 2H, NH2), 4.53 (m, 1H, H5′), 4.48 (m, 1H, H5″), 2.21–2.12 (s, 9H, CH3–Ac). 13C
1H, H5′), 4.49 (m, 1H, H4′), 4.47 (m, 1H, H5″), 2.17–2.10 (s, NMR: (CDCl3) δ 170.3–169.6 (CvO–Ac), 155.9 (C6), 155.4 (C2),
9H, CH3–Ac). 13C NMR: (CDCl3) δ tetrazole form: 170.8–170.6 149.8 (C4), 141.8 (C8), 122.1 (C5), 86.6 (C1′), 80.9 (C4′), 73.7
(CvO–Ac), 159.3 (C6), 153.7 (C2), 143.2 (C4), 139.6 (C8), 118.8 (C2′), 70.8 (C3′), 63.2 (C5′), 20.8–20.4 (CH3–Ac). HRMS
(C5), 86.5 (C1′), 79.8 (C4′), 72.7 (C2′), 70.5 (C3′), 62.9 (C5′), (ESI-TOF): calcd for C16H19N6O9 [M + H]+ 439.1214, found
20.8–20.4 (CH3–Ac); azido form: 169.7–169.4 (CvO–Ac), 162.6 439.1215.
(C6), 146.1 (C2), 142.6 (C4), 138.9 (C8), 114.9 (C5), 87.2 (C1′),
2-Amino-6-nitro-9-(2′,3′,5′-tri-O-acetyl-β-D-ribofuranosyl)-
80.1 (C4′), 72.9 (C2′), 70.6 (C3′), 63.1 (C5′), 20.8–20.4 (CH3–Ac). purine (7). UV (CH3CN) λmax/nm (ε/M−1 cm−1): 230.0 (14 900)
HRMS (ESI-TOF): calcd for C16H19N8O7 [M + H]+ 435.1377, and 386.0 (2900). 1H NMR: (CDCl3) δ 8.17 (s, 1H, H8), 6.11
found 435.1374.
(d, J = 4.9 Hz, 1H, H1′), 5.97 (t, J = 5.2 Hz, 1H, H2′), 5.74
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 571
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Paper
Photochemical & Photobiological Sciences
(t, J = 4.9 Hz, 1H, H3′), 5.62 (s, 2H, NH2), 4.49 (m, 1H, H4′),
4.47 (m, 1H, H5′), 4.42 (m, 1H, H5″), 2.16–2.10 (s, 9H, CH3–Ac).
13C NMR: (CDCl3) δ 170.5–169.3 (CvO–Ac), 158.6 (C6), 158.3
(C2), 152.4 (C4), 144.8 (C8), 119.5 (C5), 86.8 (C1′), 80.2 (C4′), 72.7
(C2′), 70.4 (C3′), 62.8 (C5′), 20.7–20.4 (CH3–Ac). HRMS (ESI-TOF):
calcd for C16H19N6O9 [M + H]+ 439.1214, found 439.1239.
TP
2ATP
6NP
2A6NP
6IP
2A6IP
6SP
7-Methyl-7H-tetrazolo[5,1-i]purine
7-Methyl-7H-tetrazolo[5,1-i]purin-5-amine
6-Nitreno-9-methyl-9H-purine
6-Nitreno-9-methyl-9H-purin-2-amine
6-Dioxaziridino-9-methyl-9H-purine
6-Dioxaziridino-9-methyl-9H-purin-2-amine
6-Dioxaziridino-9-methyl-9H-purine transition state
6-Dioxaziridino-9-methyl-9H-purin-2-amine
sition state
6-Nitro-9-methyl-9-H-purine
Computational methods
2A6SP
tran-
Solutions DFT calculations were carried out using the DMol3
code, which implements a numerical basis set. The geometry
optimizations, the NEB method and thermodynamic free ener-
gies were calculated using DMol313 with the PBE functional19
and the DNP basis set.
6NO2P
2A6NO2P 6-Nitro-9-methyl-9H-purin-2-amine
The NEB method20 was employed using a linear inter-
polation of structures as the starting point. A linear inter-
polation was achieved using the Reaction Preview tool in
Materials Studio, which ensures a one-to-one correspondence
of the atoms in the initial and final structures. The initial and
final structures were geometry optimized using the same basis
set and functional as those used for the NEB calculation with
a convergence criterion of <10−6 change in the energy (in Har-
trees) between two optimization steps. The NEB chains used in
the calculation consisted of 16–30 structures. The end point
structures were held fixed and the intervening structures were
iteratively subjected to the NEB method until the convergence
criterion of a less than 5 × 10−4 Hartrees Bohr−1 change in the
gradient was reached. Refinement of saddle points was calcu-
lated using a version of the program DIMER.15a The FIRE opti-
mizer was employed in the iterative calculations.21
Three paths were defined using the Reaction Preview tool.
The starting structures consisted of geometry optimized
purines with an O2 molecule poised approximately 4 Å from
the 6-nitrene of the purine ring system. The end point for the
first stage consists of a geometry optimized intermediate (6IP
or 2A6IP). A second path begins at 6IP/2A6IP and ends at
6NO2P/2A6NO2P. These paths were compared to the ring
expansion path that involves insertion of the 6-nitrene into the
ring. Structures along these paths that appear to be local
maxima were selected in pairs and submitted to the program
DIMER for further optimization of the structures to locate
saddle points. In order to test whether saddle points have
been located, the structures refined in DIMER were submitted
for a vibrational frequency calculation. Ideally the saddle point
has one negative eigenvalue. In cases where more than one
negative eigenvalue was observed, the potential surfaces and
projections along relevant eigenvectors were studied. Finally,
intrinsic reaction coordinate (IRC) analysis was conducted
to ascertain that the transition state was correctly identified
and to determine which eigenvector follows the minimum
energy path.
Notes and references
1 (a) M. M. Hanna, Methods Enzymol., 1989, 180, 383;
(b) L. A. Sylvers and J. Wower, Bioconjugate Chem., 1993, 4,
411; (c) J. Wower, K. V. Rosen, S. S. Hixon,
R. A. Zimmermann, L. A. Sylvers and Y. Xing, Bioconjugate
Chem., 1994, 5, 158; (d) M. M. Hanna, L. Bensten,
M. Lucido and A. Sapre, Methods Mol. Biol., 1999, 118, 21.
2 K. M. Meisenheimer and T. H. Koch, Crit. Rev. Mol. Biol.,
1997, 32, 101.
3 (a) S. Senda, K. Hirota, M. Suzuki, T. Asao and
K. Maruhashi, J. Chem. Soc., Chem. Commun., 1976, 731;
(b) S. Senda, K. Hirota, T. Asao and K. Maruhashi, J. Am.
Chem. Soc., 1977, 99, 7358; (c) S. Senda, K. Hirota, T. Asao
and K. Maruhashi, J. Chem. Soc., Chem. Commun., 1978,
367; (d) S. Senda, K. Hirota, T. Asao and K. Maruhashi,
J. Am. Chem. Soc., 1978, 100, 7661; (e) S. Senda, K. Hirota,
T. Asao and K. Maruhashi, J. Chem. Soc., Perkin Trans. 1,
1984, 1719; (f) H. Tanaka, H. Hayakawa, K. Haraguchi
and T. Miyasaka, Nucleosides Nucleotides, 1985, 4, 607;
(g) T. Miyasaka, H. Tanaka, K. Satoh, M. Imahashi,
K. Yamaguchi and Y. Iitaka, J. Heterocycl. Chem., 1987, 24,
873; (h) F. Peyrane, M. Cesario and P. Clivio, J. Org. Chem.,
2006, 71, 1742; (i) S. Goudain, A. Martinez, C. Petermann,
D. Harakat and P. Clivio, J. Org. Chem., 2009, 74, 6885;
( j) S. Goudain, C. Petermann, A. Martinez, D. Harakat and
P. Clivio, J. Org. Chem., 2011, 76, 1906.
4 Y. Xing, S. S. Hixson and R. A. Zimmermann, Tetrahedron
Lett., 1990, 31, 5849.
5 S. N. Bose, Indian J. Chem., 1984, 23B, 677.
6 K. Komodziński, J. Nowak, J. Lepczyńska, J. Milecki and
B. Skalski, Tetrahedron Lett., 2012, 53, 2316.
7 K. Komodziński, J. Lepczyńska, P. Ruszkowski, J. Milecki
and B. Skalski, Tetrahedron Lett., 2013, 54, 3781.
8 J. A. Montgomery and L. B. Holum, J. Am. Chem. Soc., 1958,
80, 404.
9 T. Kaiya, H. Tanaka and K. Kohda, Nucleosides, Nucleotides
Nucleic Acids, 2002, 21, 427.
10 (a) C. Temple, M. C. Thorpe, W. C. Colburn and
J. A. Montgomery, J. Org. Chem., 1966, 31, 935; (b) C. Mathe,
T. Lioux and G. Gosselin, Nucleosides, Nucleotides Nucleic
Acids, 2003, 22, 605; (c) T. Lioux, G. Gosselin and C. Mathe,
Eur. J. Org. Chem., 2003, 3997.
Abbreviations
6AzP
6-Azido-9-methyl-9H-purine
2A6AzP
6-Azido-9-methyl-9H-purin-2-amine
572 | Photochem. Photobiol. Sci., 2014, 13, 563–573
This journal is © The Royal Society of Chemistry and Owner Societies 2014

View Article Online
Photochemical & Photobiological Sciences
Paper
11 (a) C. Temple, L. C. Kussner and J. A. Montgomery, J. Org. 16 (a) B. Singh and J. S. Brinen, J. Am. Chem. Soc., 1971, 93,
Chem., 1966, 31, 2210; (b) A. Masternak, B. Skalski and
J. Milecki, J. Labelled Compd. Radiopharm., 2007, 50, 43;
(c) M. K. Lakshman, M. K. Singh, D. Parrish,
R. Balachandran and B. W. Day, J. Org. Chem., 2010, 75,
2461.
540; (b) J. S. Brinen and B. Singh, J. Am. Chem. Soc., 1971,
93, 6623; (c) R. A. Abramovitch and S. R. Challand, J. Chem.
Soc., Chem. Commun., 1972, 964; (d) T. Hander, P. Wessig,
J. Bendig and R. Stosser, J. Am. Chem. Soc., 1999, 121, 6580;
(e) J. Liu, C. M. Hadad and M. S. Platz, Org. Lett., 2005, 7, 549.
17 (a) M. J. Moris and B. Uznański, Can. J. Chem., 1981, 59,
2601; (b) V. Nair and S. G. Richardson, Synthesis, 1982, 670;
(c) A. Matsuda, M. Shizozaki, M. Suzuki, K. Watanabe and
T. Miyasaka, Synthesis, 1986, 385.
12 T. Fujii and T. Itaya, Heterocycles, 1999, 51, 1971.
13 (a) B. Delley, J. Chem. Phys., 1990, 92, 508; (b) B. Delley,
J. Chem. Phys., 2000, 113, 7756.
14 (a) R. A. McClelland, P. A. Davidse and G. Hadzialic,
J. Am. Chem. Soc., 1995, 117, 4173; (b) B. Cheng 18 L. P. Kotra, K. K. Manouilov, E. Cretton-Scott,
and R. A. McClelland, Can. J. Chem., 2001, 79, 1881;
(c) V. Voskresenska, R. M. Wilson, M. Panov, A. N. Tarnovsky,
J.-P. Sommadossi, F. D. Boudinot, R. F. Schinazi and
C. K. Chu, J. Med. Chem., 1996, 39, 5202.
J. A. Krause, S. Vyas, A. H. Winter and C. M. Hadad, J. Am. 19 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
Chem. Soc., 2009, 131, 11535. 1996, 77, 3865.
15 (a) G. Henkenlman and H. Jonsson, J. Chem. Phys., 1999, 20 D. Sheppard, R. Terrell and G. Henkelman, J. Chem. Phys.,
111, 7010; (b) R. A. Olsen, G. J. Kroes, G. Henkelman, 2008, 128, 134106.
A. Arnaldsson and H. Jonsson, J. Chem. Phys., 2004, 21 E. Bitzek, P. Koskinen, F. Gahler, M. Moseler and
121, 9776.
P. Gumbsch, Phys. Rev. Lett., 2006, 97, 170201.
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Photochem. Photobiol. Sci., 2014, 13, 563–573 | 573