Distribution of photo-cross-linked products from 3-aryl-3-trifluoromethyldiazirines and alcohols

Article abstract of DOI:10.1016/j.tet.2008.04.031

The photolysis of N-[4-(3-trifluoromethyl-3H-diazirin-3-yl)benzoyl]-2,2′-ethylenedioxybis(ethylamine) (1) in EtOH, PrOH, i-PrOH, BuOH, and i-BuOH was carried out in both solution (at room temperature) and solid phase (at -196 °C), and the resultant C-H and O-H insertion products were analyzed semi-quantitatively by LC/ESI-MS/MS. The carbene insertion reactions at low temperature produced all possible C-H and O-H insertion products in a relatively uniform distribution, which could not be accomplished by the solution-phase reaction.

Full text of DOI:10.1016/j.tet.2008.04.031

Tetrahedron 64 (2008) 5692–5698
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Distribution of photo-cross-linked products from 3-aryl-3-trifluoro-
methyldiazirines and alcohols
,
b
, ,
, ,
Naoki Kanoh a
y, Takemichi Nakamura c y, Kaori Honda a, Hiroyuki Yamakoshi b,
*
*
Yoshiharu Iwabuchi b, Hiroyuki Osada a
a Antibiotics Laboratory, Advanced Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan
b Graduate School of Pharmaceutical Sciences, Tohoku University, Aobayama, Sendai 980-8578, Japan
c Molecular Characterization Team, Advanced Science Institute, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan
a r t i c l e i n f o
a b s t r a c t
The photolysis of N-[4-(3-trifluoromethyl-3H-diazirin-3-yl)benzoyl]-2,20-ethylenedioxybis(ethylamine)
(1) in EtOH, PrOH, i-PrOH, BuOH, and i-BuOH was carried out in both solution (at room temperature) and
solid phase (at ꢀ196 ꢁC), and the resultant C–H and O–H insertion products were analyzed semi-
quantitatively by LC/ESI-MS/MS. The carbene insertion reactions at low temperature produced all
possible C–H and O–H insertion products in a relatively uniform distribution, which could not be
accomplished by the solution-phase reaction.
Article history:
Received 5 February 2008
Received in revised form 8 April 2008
Accepted 8 April 2008
Available online 11 April 2008
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
dioxybis(ethylamine) (1) not only reacted and was introduced
into a variety of small molecules5,6 but also generated multiple
3-Aryl-3-trifluoromethyldiazirines have been used primarily in
photoaffinity labeling, a biochemical approach to identify a specific
receptor for a small-molecule ligand and a binding site within the
receptor molecule.1 Upon UV irradiation, the photoreactive group
introduced into the ligand is promoted to a carbene, which in turn
binds irreversibly to the receptor at the interaction site. Photo-
generated carbenes are very reactive and are thought to have a low
functional-group selectivity, which makes them essentially useless
for conventional organic synthesis, but recently it has been sug-
gested that they may have many applications in chemical biology.
For example, Richards et al.2 and Delfino et al.3,4 have used meth-
ylene carbene as a probing agent for protein surfaces, structures,
and folding. We and others have utilized carbenes generated from
immobilized aryldiazirines to photo-cross-link small molecules
onto solid surfaces as microarrays5–8 and affinity beads (Fig. 1a).9,10
These platforms are useful for screening protein–small molecule
interactions.
conjugates from a molecule.7 The latter is very important to detect
possible binding proteins for small molecules because there are
examples in which one small molecule interacts with multiple
proteins by using different areas of its surface.11,12 To maximize
cross-linking efficiency and randomness toward functional groups,
we carry out the photo-cross-linking reaction in a semi-solid and
highly concentrated state (Fig. 1a):13 for example, when we prepare
microarrays, each small molecule solution spotted as an array was
dried and concentrated on the solid surface prior to photolysis.
Upon photolysis, the reactive carbene and a proximal small mole-
cule (or a functional group), which usually diffuse away in a fluid
solution, are expected not to diffuse apart because of limited dif-
fusibility. Therefore, even a small molecule (or a functional group)
having low reactivity would react efficiently with a proximal car-
bene to give a conjugate.8 Unless the small molecule and carbene
are aligned regularly, the cross-linking reaction occurs with low
functional group selectivity. To the best of our knowledge, there
is no precedent for producing a variety of surface-immobilized
conjugates from a small molecule.14
However, to explore the scope and limitations of the photo-
cross-linked platforms, it is important to know what functional
groups on a small molecule are used to produce conjugates with
photo-generated carbenes and to what extent these reactions
proceed.
In 1970s and 1980s, groups led by Tomioka, Plats, Moss, and
Gaspar studied the reactivity of photo-generated aryl carbenes
extensively. The carbenes were mostly derived from parent diazo
compounds in solidified organic molecules.15,16 Relatively few
The photo-cross-linking strategy for immobilizing small mole-
cules onto a solid surface is unique in terms of attachment site
selectivity. We have shown that a carbene photo-generated from
N-[4-(3-trifluoromethyl-3H-diazirin-3-yl)benzoyl]-2,20-ethylene-
* Corresponding authors. Tel./fax: þ81 22 795 6847 (N.K.); tel.: þ81 48 467 9364;
fax: þ81 48 467 4629 (T.N.).
E-mail addresses: kanoh@mail.pharm.tohoku.ac.jp (N. Kanoh), takemi@
postman.riken.go.jp (T. Nakamura).
y
Contributed equally to this work.
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.04.031

N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
5693
(a)
(b)
N N
F3C
N
N
F3C
H
N
H
N
O
O
NH2
O
N
H
O
O
O
1
X-Y (alcohol)
365 nm (4 J/cm2)
small
molecule
concentrated
in vacuo
small
molecule
r.t. or - 196°C
365 nm (4 J/cm2)
X
Y
F3C
F3C
H
N
H
N
O
O
O
N
O
2
H
O
O
C-H and O-H insertion products
Figure 1. Production of photoadducts.
studies have provided detailed and quantitative data regarding the
selectivity of diazirine-derived carbenes, particularly those, which
are linked to the other structures such as fatty acyl chains and
peptides.17–21 Specifically, information on the reactivity of photo-
generated carbenes from 1 would be very important for us in the
development of photo-cross-linked platforms. However, to analyze
the structures of the multiple conjugates generated from the photo-
cross-linkingreaction, sensitiveandsystematicmethodsareneeded.
Toward these ends, we have initiated studies to analyze and
quantify the functional-group selectivity of photo-generated car-
benes from 1. As preliminary results from these efforts, we describe
the semi-quantitative and systematic analysis of photolysis prod-
ucts of 1 in alcohols at ambient and low temperatures by LC/ESI-
MS/MS and MS/MS/MS (Fig. 1b).
reaction of the carbene and dissolved oxygen.25 Compounds 4
(MHþ: m/z 335) and 9 (MHþ: m/z 351) were identified as a formally
reduced product and a formal H2O adduct, respectively.23
The same reaction performed at ꢀ196 ꢁC, i.e., in the solid state,
gave almost the same products but with significantly different
product distributions. As clearly shown in Figure 2c, the carbene
insertion reaction with ethanol predominated over the other type
of reaction in the solid state: the ratio of the EtOH adducts (shaded
peaks) increased dramatically. In addition, MS/MS experiments
revealed that a large peak at w33 min contained two products:
primary alcohol 523 and the secondary alcohol 6.
The product distribution of the EtOH adducts was estimated on
the basis of the total ion chromatogram generated from the MS data
set for m/z 379. The relative ratios of peak intensities derived from
each compound were highly reproducible. Estimation using a UV
chromatogram was another possibility, but it was difficult to de-
duce the distribution from the UV trace because of signal overlap
and a lack of sensitivity. In the case of the overlapping of cross-
linked products in the chromatogram (for example, compounds 5
and 6), we estimated that diastereomers having the same atom
connectivity should be generated as a 1:1 mixture. This estimation
is thought to be acceptable because most pairs of diastereomers
throughout in this study were found to be produced approximately
in a 1:1 ratio when good peak separation was observed (vide infra).
As a result, the product distributions of EtOH adducts (3/5/6/8)
were estimated to be 87:0:6:726 and 21:13:33:33 in the liquid and
solid states, respectively.
2. Results and discussions
2.1. Photolysis of 1 in EtOH: characterization of all reaction
products and observation of dramatic differences in product
distribution depending on the reaction state (i.e., reaction
temperature)
The photolysis of aryldiazirines in organic solvents has been
reported to produce a variety of products.18–20,22 Indeed, as shown
in Figure 2a and b, we observed eight new peaks (2–10) after ir-
radiation (365 nm, 4 J/cm2) of an EtOH solution of 1 (w2 mM) at
room temperature. Among those eight products, three compounds
(3, 6, and 8; shaded peaks in Fig. 2b) were each shown by LC/ESI-MS
to have a molecular weight of 378 (observed as MHþ: m/z 379),
which was expected when the photo-generated carbene was
inserted into an EtOH molecule. The structures of the EtOH adducts
3, 6, and 8 were identified as ethyl ether and two diastereomers of
secondary alcohols as shown in Figure 2d, respectively.23
Structures of the other products were also elucidated. Com-
pound 2 was found to be identical to 1 in molecular weight but to
have a different retention time in LC. The UV spectrum of 2 revealed
that the compound has an absorption maximum at 293 nm, sug-
gesting that it was a diazo isomer as shown in Figure 2d.24 Com-
pounds 7 and 10 were found to have the same molecular weight of
366 (MHþ: m/z 367), corresponding to formal adducts between the
photo-generated carbene and H2O2. Because peak 7 disappeared
when triphenylphosphine was added to the reaction mixture after
irradiation, we assigned compound 7 as a hydroperoxide. Com-
pound 10 was identified as a hydrate of aryl trifluoromethyl ketone
11.23 Compounds 7 and 11 would have been generated by the
2.2. MS fragmentation pattern analysis of the EtOH adducts
The collision-induced dissociation (CID) MS/MS spectra of
compounds 3, 5, 6, and 8 are shown in Figure 3a–d. The fragmen-
tation patterns of the structural isomers were clearly different,
whereas those of the diastereomers (6 and 8) were almost identical.
The major product ions in the CID spectrum of each structural
isomer were assigned as shown in Figure 3e–g on the basis of MS/
MS and MS/MS/MS experiments, accurate mass measurements, and
deuterium-labeling experiments.
Two primary fragmentation pathways that give m/z 274 and 231
were dominant in the CID mass spectrum of compound 3 (Fig. 3a).
The product ion at m/z 176 may be generated by a secondary
fragmentation from m/z 274, i.e., losses of CF3 and C2H5 from the
product ion at m/z 274 (Fig. 3e). In contrast to this, different sec-
ondary fragmentation pathways that give m/z 211 and 213, i.e.,
a loss of HF or H2O from m/z 231, emerged in the CID spectrum of 5,

5694
N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
C2H4O, which was originally in an ethanol molecule, but not in-
(a)
cluding the hydrogen of the OH group (see Supplementary data,
Figs. S2–S4). This observation implies that the hydrogen atom of the
hydroxyl group specifically migrated to the ionic fragment as the
neutral C2H4O was lost. Therefore, the characteristic product ions at
m/z 230, 210, 187, and 167 can be regarded as a common feature of
the cross-linked compounds having both 1-hydroxyalkyl and tri-
fluoromethyl groups at the benzylic position. In other words, these
ions may be useful for the LC/MS/MS-based characterization of the
homologous products generated by the photo-cross-linking of 1
with other alcohols.
(b)
irradiated
@ r.t.
(solution)
(c)
2.3. Reaction with other alcohols: LC/MS/MS-based
characterization of the reaction products
irradiated
@ -198 °C
(solid state)
We then performed photo-cross-linking experiments of 1 with
PrOH, i-PrOH, BuOH, and i-BuOH in both liquid and solid states, and
the resultant distributions of the photoadducts were analyzed. The
total ion chromatograms and the MS/MS spectrum for each adduct
(generated from the MS data sets for MHþ at m/z 393 for propanols;
MHþ at m/z 407 for butanols) were obtained (Fig. 4), and their MS/
MS fragmentation patterns were compared with those of the EtOH
adducts (3, 5, 6, and 8).
For example, the photo-cross-linking of 1 with i-PrOH in the
solid state gave four isomeric products as expected, and all of them
were separated by LC as shown in the chromatogram (Fig. 4a). The
most hydrophobic isomer (12) gave a secondary fragmentation
product with m/z 176 in its CID spectrum (Fig. 5a), and the presence
of isopropyl ether was suggested. The ions at m/z 246 and 203 can
be attributed to a loss of 42 u from m/z 288 and 245, respectively.
These 42 u (C3H6: propene equivalent) losses are consistent with
the presence of an isopropyl group. The second and third peaks in
the chromatogram (14, 13) showed secondary losses of HF and H2O
from m/z 245 in their CID spectra (Fig. 5b and c), and the presence of
a 2-hydroxypropyl group at the benzylic position was suggested.
The CID spectrum of the first peak (15) showed product ions at m/z
230, 210, 187, and 167, which are characteristic for a 1-hydroxyalkyl
unit (i.e., 1-hydroxy-1,1-dimethyl group) at the benzyl position
(Fig. 5d), as described previously.
(d)
N
N
O
N
N
F3C
F3C
F3C
R
R
R
R
R
1
2
3
OH
OH
F3C
F3C
F3C
R
4
5
6 and 8
OOH
OH
F3C
F3C
R
R
R
R
The photo-cross-linking experiments of 1 with i-PrOH in the
liquid state also gave four products, although the ratios between
the products were quite different, i.e., the relative abundances of 13
and 14 were very low (Supplementary data, Fig. S5).
7
9
O
OH
HO
F3C
F3C
Similarly, the photo-cross-linking experiment of 1 with i-BuOH
in the solid state gave the six expected isomeric products as shown
in the chromatogram (Fig. 4b). The MS/MS spectrum of compound
16 showed characteristics of O–H insertion products, whereas the
spectra of 17 and 21 clearly exhibited secondary fragment ions at
m/z 230, 210, 187, and 167, revealing the presence of a pair of di-
astereomeric 1-hydroxyalkyl derivatives (Supplementary data,
Fig. S6). Compounds 19 and 20 were suggested to be another dia-
stereomeric pair containing a hydroxyalkyl side-chain since their
MS/MS spectral patterns, including the secondary loss of H2O, were
almost identical. The remaining compound 18 may be assigned to
the remaining isomer. The structures of 16–21 are summarized in
Figure 6. Compounds 16–21 were also observed in the liquid state
experiment but the relative abundances of the isomers were dif-
ferent from the solid state experiment. Again, the relative abun-
dance of 16 was very high while the abundances of 19 and 20 were
very low (Supplementary data, Fig. S5).
11
10
H
N
O
R =
O
NH2
O
Figure 2. LC/ESI-MS chromatograms of 1 and the photolysis products in EtOH. The
total ion chromatograms before photolysis (a) and after the photolysis of 1 performed
at (b) room temperature and (c) ꢀ196 ꢁC are shown; (d) structure of the photolysis
products.
which does not have an ethoxy but rather a hydroxyethyl group
(Fig. 3b and f). The presence of the hydroxyl group is responsible for
the facile loss of HF in 5, at least in part, since roughly 40% of the
deuterium was lost in the MS/MS/MS spectrum of the deuteroxy
analog of 5 (see Supplementary data, Fig. S1). Although the dia-
stereomers 6 and 8 also have a hydroxyl group, the CID spectra of
these compounds (Fig. 3c and d) were clearly different from the CID
spectrum of 5 (Fig. 3b). In the CID spectra of 6 and 8, secondary
fragmentation channels, loss of 44 u from both m/z 274 and 231,
were dominant. The loss of 44 u was shown to be attributable to
In the case of the photo-cross-linking experiment of 1 with
PrOH in the solid state, six products were expected but only five
peaks were observed in the chromatogram (Fig. 5c). We deduced all
of the six expected isomers including two diastereomeric pairs
were actually present but only five peaks were observed due to
peak overlap. We assumed that each diastereomeric pair consisted

N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
5695
MS/MS of 3
(a) 100
274
(e)
O
+H+
88
O
231
F3C
[M+H]+
(m/z 379)
H
N
50
0
NH2
O
O
3
88
100
176
231
274
50
150
200
250
300
350
400
, -C
- CF
2H5
3
+
C10H10NO2
274
(m/z 176)
MS/MS of 5
100
274
(b)
(c)
(d)
(f)
+H+
OH
231
88
50
0
[M+H]+
(m/z 379)
F3C
H
N
211
200
213
O
NH2
O
88
100
O
5
50
150
250
300
300
300
350
350
350
400
400
400
231
- HF
- H2O
C10H9F2O2+
(m/z 211)
C11H8F3O+
(m/z 213)
MS/MS of 6
100
274
187
230
231
50
0
210
(g)
167
OH
+H+
88
88
100
F C
[M+H]+
(m/z 379)
3
H
N
50
150
200
250
O
NH2
O
O
231
6 and 8
274
MS/MS of 8
100
274
-
-
C2H
C2H4O
4O
C9H6F3O+
(m/z 187)
C
11 H11F NO+
(m/z 230)
3
210
187
230
231
50
0
88
167
-HF
-HF
C9H5F O+
C
11H10F NO+
50
100
150
200
250
2
2
m/z
(m/z 167)
(m/z
Figure 3. CID MS/MS spectra, chemical structure, and proposed fragmentation pathways of 3 (a and e), 5 (b and f), 6 (c and g), and 8 (d and g). These spectra were obtained by LC/
MS/MS of the synthetic materials. The precursor ions at m/z 379 (protonated molecule of each compound) were converted to the product ions completely under the CID conditions
used.
of roughly equal amounts of diastereomers, and that the two di-
astereomers in a pair were reasonably well separated by the LC
system as in the previous cases. Based on these assumptions, the
data could be interpreted as follows. The first and third peaks were
assigned to the diastereomeric pair having a 1-hydroxyalkyl group
at the benzylic position (27 and 24) as they showed almost identical
MS/MS spectra containing m/z 230, 210, 187, and 167 ions (Sup-
plementary data, Fig. S7). The last peak 22 was attributed to the O–
H insertion product with m/z 176 in its MS/MS spectrum. The MS/
MS spectra of the second and the fourth peak were similar, and
both showed the characteristics of hydroxyalkyl derivatives. The
synthetic 3-hydroxy-1-propyl derivative (25; see Supplementary
data) eluted at the second peak position, and the MS/MS spectrum
of the synthetic sample (Supplementary data, Fig. S8) was similar to
that of the second peak. Therefore, the majority of the intensity of
the second peak was considered to be due to 25. The remaining
fourth peak can be assigned to one of the 3-hydroxy-2-propyl de-
rivatives (23); however, it is also very likely that a similar amount of
diastereomer was present somewhere, as in the other cases. The
second peak decreased to the size nearly equal to that of the fourth
peak in the liquid state experiment (Supplementary data, Fig. S5).
The MS/MS spectrum of the second peak in the liquid state prod-
ucts was similar to that of 23. These observations suggested that
most of the intensity of the second peak in the liquid state products
was attributable to one of the 3-hydroxy-2-propyl derivatives (26).
Hence, we concluded that the second peak in the solid state
products was a mixture of 25 (major component) and 26 (minor
component).

5696
N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
w/ i-PrOH
(a)
14
288
(a)
100
50
15
O
13
245
F3C
246
12
203
200
12
(b)
w/ i-BuOH
88
100
176
19
18
20
0
16
50
150
250
300
350
400
21
17
(b)
288
OH
100
25 (major) and 26 (minor)
(c)
w/ PrOH
F3C
24
23
27
227
225
22
50
0
245
13
270
250
88
100
31 (major) and 32 (minor)
(d)
w/ BuOH
33
34
50
50
50
150
200
300
288
350
400
400
400
30
28
35
29
(c)
100
OH
20
30
40
50
227
F3C
time, min
50
0
245
225
Figure 4. Total ion chromatograms for the photo-cross-linking products of 1 with
i-PrOH (a), i-BuOH (b), PrOH (c), and BuOH (d) performed at ꢀ196 ꢁC. As the chro-
matograms were obtained by LC/MS/MS analysis on the protonated molecules of each
compound (m/z 393 for the chromatograms a and c; m/z 407 for b and d), only the
photoadducts, which have those masses (i.e., isomeric products of carbene insertion
into C–H or O–H) are present in each trace.
14
88
100
270
250
230
150
200
300
350
(d)
OH
100
Similarly, only seven peaks were observed in the chromatogram
of the photo-cross-linking of 1 with BuOH in the solid state
(Fig. 4d), although eight isomeric products were expected. The first
and fifth peaks were assigned to the diastereomeric 1-hydroxyalkyl
derivatives (35 and 30) as they showed almost identical MS/MS
spectra that included m/z 230, 210, 187, and 167 ions (Supple-
mentary data, Fig. S9). The last peak was attributable to the O–H
insertion product 28 because of m/z 176 in its MS/MS spectrum. The
second, third, fourth, and sixth peaks gave MS/MS spectra having
the characteristics of hydroxyalkyl derivatives. The third and sixth
peaks showed similar peak areas, and their MS/MS spectra were
similar to each other’s as well. This suggested that the third and
sixth peaks represent a pair of diastereomers. On the other hand,
the synthetic 4-hydroxy-1-butyl derivative 31 eluted at the fourth
peak position. The MS/MS spectra of the synthetic 31 (Supple-
mentary data, Fig. S8) and the second and fourth peaks were similar
to one another. We therefore concluded that another dia-
stereomeric pair eluted in the second and fourth peaks, and the 4-
hydroxy-1-butyl derivative 31 co-eluted with the fourth peak,
which is significantly bigger than the second peak, since the second
peak and the fourth peak showed similar peak areas in the chro-
matogram of the liquid state products. We assumed the amount, if
any, of 31 in the liquid state products was very small. Unfortunately,
we could not find any evidence to indicate that the diastereomeric
pairs of 29 and 33 or of 32 and 34 correspond to the second and
fourth peaks, or to the third and sixth peaks, respectively. There-
fore, structural assignments for these diastereomeric pairs in Fig-
ure 6 may be reversed.
210
F3C
187
167
50
0
15
288
88
100
150
200
250
300
350
m/z
Figure 5. CID MS/MS spectra (LC/MS/MS) of the four isomeric photo-cross-linking
products of 1 with i-PrOH. The precursor ions at m/z 393 (protonated molecule of each
compound) were converted to the product ions completely under the CID conditions
used.
C–H bonds highlighted in green; and (4) compounds from the
carbene insertion at the internal methylene groups highlighted in
purple. As described, we assigned two compounds that gave the
same fragmentation pattern as a pair of diastereomers having the
same atom connectivity in those cases where the reaction can
produce a pair of diastereomers (such as 6 and 8). In case one of the
diastereomers overlapped with other products, the ratios of dia-
stereomers having the same atom connectivity were estimated as
1:1 mixtures.
As can be clearly seen in Figure 7, all photolysis experiments
performed in solution gave O–H insertion products in more than
70% of the total insertion reactions. It is generally accepted that
alcohols in solution react with carbenes in the singlet state to give
O–H insertion products as major products.27 Interestingly, when
a C–H bond into which the carbene inserts is on a methyl or
methylene group, the level of C–H insertion products in solution
tends to decrease as the C–H bond moves away from the hydroxyl
group. On the other hand, insertions into the methine C–H bond in
solution tend to dominate over those of other C–H bonds even
when the methine C–H bond is away from the hydroxyl group. This
Based on the structural assignment discussed it was possible to
categorize the cross-linked products into four classes (Fig. 7): (1)
ether-type compounds that were generated by the insertion of the
carbene into the O–H bonds highlighted in red; (2) compounds
from the carbene insertion at C1–H bonds highlighted in light blue;
(3) compounds from the carbene insertion at the terminal methyl

N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
5697
HO
O
HO
F3C
H
N
R =
O
O
NH2
F3C
F3C
O
R
R
R
12
15
13 and 14
HO
HO
O
O
O
HO
F3C
F3C
F3C
F3C
R
R
R
R
R
R
R
R
R
R
16
18
17 and 21
19 and 20
HO
HO
HO
HO
F3C
F3C
F3C
F3C
R
22
25
24 and 27
23 and 26
HO
HO
F3C
HO
F3C
F3C
F3C
F3C
R
R
28
31
30 and 35
29 and 33
(or 32 and 34)
32 and 34
(or 29 and 33)
Figure 6. Summary of the structural assignments for the products of photo-cross-linking of 1 with i-PrOH, i-BuOH, PrOH, and BuOH.
tendency can be explained when the triplet carbene is thought of as
an intermediate: the carbene’s rate-limiting abstraction of a hy-
drogen atom produces a radical pair, both of which in turn bind to
the formal insertion product.16
accomplished by conventional solution-phase reactions. Further
evaluation of the nonselectiveness of the reaction of the carbene
generated from 1 with other molecules will be reported in due
course.
Solid-phase reactions are different from those occurring in so-
lution in many respects. Striking differences were found in the
percentage of terminal methyl C–H insertion products and O–H
insertion products. In every case performed in the solid state, the
percentage of total C–H insertions increased dramatically at the
expense of O–H insertion reactions. In other words, the insertion
reactions in the solid state became more nonselective. Further-
more, it should be noted that all possible isomers were produced
under these conditions. Among the C–H insertion products, the
percentage of C–H insertions at the terminal methyl groups was
significantly increased.
4. Experimental
4.1. Materials and methods
All chemicals and solvents were purchased from commercial
vendors and used without further purification. N-[4-(3-Tri-
fluoromethyl-3H-diazirin-3-yl)benzoyl]-2,20-ethylenedioxybis-
(ethylamine) (1) was prepared according to the reported
procedure.5
4.2. Photolysis of 1 in alcoholic solution (general procedure)
3. Conclusion
An alcoholic solution of compound 1 (trifluoroacetic acid salt,
1 mg/mL, w1 mL) in a 10 mL glass vial was exposed to UV irradi-
ation of 4 J/cm2 at 365 nm using a CL-1000L ultraviolet cross-linker
(UVP Inc., CA, USA). The photolyzed solution was immediately
stored at ꢀ20 ꢁC until analyzed by LC/ESI-MS.
In summary, we analyzed the photolysis products from 1 and
five alcohols using LC/ESI-MS/MS and MS/MS/MS and found that
the reactions in the solution phase were different from those in the
solid phase. Especially in the solid phase, all C–H and O–H insertion
products were formed in close to equivalent amounts. This obser-
vation is basically in accord with that of Tomioka et al., although
they used other carbene species and alcohols that produced much
simpler combinations of photo-cross-linked products.22
4.3. Photolysis of 1 in solidified alcohol at low temperature
(general procedure)
Although temperature effects should be considered for strict
discussion, the results of this study support at least in part that the
photo-cross-linking process for immobilizing small molecules on
solid supports, in which solutions of small molecules were con-
centrated on the solid support in a solid or highly viscous state after
which the small molecules were photo-cross-linked, takes place in
a highly functional group-independent manner, which cannot be
An alcoholic solution of compound 1 (trifluoroacetic acid salt,
1 mg/mL, w1 mL) in a 10 mL glass vial was cooled in a Dewar flask
containing liquid N2. The Dewar flask containing the solidified
sample and liquid N2 was set in a CL-1000L ultraviolet cross-linker
and exposed to UV irradiation of 1 J/cm2 at 365 nm. The surface of
the sample became yellow. The sample was allowed to thaw, during
which time the yellow color faded. The freeze–irradiation–thaw

5698
N. Kanoh et al. / Tetrahedron 64 (2008) 5692–5698
In the typical LC/MS runs, MS/MS of selected precursor ions
were also monitored. This facilitated matching of peaks between
separate runs, even if there were some variations in retention time,
which may be caused by a limited pumping reproducibility at the
low-flow rate gradient.
Acknowledgements
The authors thank Ms. A. Asami for her technical support. This
work was supported in part by the Daiichi Pharmaceutical Award in
Synthetic Organic Chemistry; the Chemical Biology Project of
RIKEN; grants from the MEXT, Japan; and a Grant-in-Aid for Sci-
entific Research (C) (No. 17510187) from JSPS, Japan.
Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2008.04.031.
References and notes
1. Brunner, J. Annu. Rev. Biochem. 1993, 62, 483–514.
2. Richards, F. M.; Lamed, R.; Wynn, R.; Patel, D.; Olack, G. Protein Sci. 2000, 9,
2506–2517.
3. Craig, P. O.; Ureta, D. B.; Delfino, J. M. Protein Sci. 2002, 11, 1353–1366.
4. Gomez, G. E.; Cauerhff, A.; Craig, P. O.; Goldbaum, F. A.; Delfino, J. M. Protein Sci.
2006, 15, 744–752.
5. Kanoh, N.; Kumashiro, S.; Simizu, S.; Kondoh, Y.; Hatakeyama, S.; Tashiro, H.;
Osada, H. Angew. Chem., Int. Ed. 2003, 42, 5584–5587.
6. Kanoh, N.; Kyo, M.; Inamori, K.; Ando, A.; Asami, A.; Osada, H. Anal. Chem. 2006,
78, 2463–2468.
7. Dilly, S. J.; Bell, M. J.; Clark, A. J.; Marsh, A.; Napier, R. M.; Sergeant, M. J.;
Thompson, A. J.; Taylor, P. C. Chem. Commun. 2007, 2808–2810.
8. Kanoh, N.; Asami, A.; Kawatani, M.; Honda, K.; Kumashiro, S.; Takayama, H.;
Simizu, S.; Amemiya, T.; Kondoh, Y.; Hatakeyama, S.; Tsuganezawa, K.; Utata, R.;
Tanaka, A.; Yokoyama, S.; Tashiro, H.; Osada, H. Chem. Asian J. 2006, 1, 789–797.
9. Kanoh, N.; Honda, K.; Simizu, S.; Muroi, M.; Osada, H. Angew. Chem., Int. Ed.
2005, 44, 3559–3562.
10. Kuramochi, K.; Haruyama, T.; Takeuchi, R.; Sunoki, T.; Watanabe, M.; Oshige,
M.; Kobayashi, S.; Sakaguchi, K.; Sugawara, F. Bioconjugate Chem. 2005, 16,
97–104.
Figure 7. Distribution of C–H and O–H insertion products from the photolysis of 1 in
alcohols. The reactions were performed at room temperature (liquid phase: designated
as L) and at ꢀ196 ꢁC (solid phase: S). Graphs and numerical values within or near each
bar refer to the percentages of the products that were generated by the insertion of the
carbenes into the chemical bonds shown in the same color. *Total percentage of C–H
insertion product from three equivalent C–H bonds.
11. Choi, J. W.; Chen, J.; Schreiber, S. L.; Clardy, J. Science 1996, 273, 239–242.
12. Ki, S. W.; Ishigami, K.; Kitahara, T.; Kasahara, K.; Yoshida, M.; Horinouchi, S.
J. Biol. Chem. 2000, 275, 39231–39236.
13. Importance of high reactivity and low functional group selectivity of the photo-
generated carbenes as small molecule capturing agents have been discussed
earlier. See Refs. 5 and 8.
14. Quite recently, Schreiber and Koehler et al. reported a new capture strategy that
allows immobilization of common reactive functional groups present in both
synthetic and natural compounds. See: (a) Bradner, J. E.; McPherson, O. M.;
Mazitschek, R.; Barnes-Seeman, D.; Shen, J. P.; Dhaliwal, J.; Stevenson, K. E.;
Duffner, J. L.; Park, S. B.; Neuberg, D. S.; Nghiem, P.; Schreiber, S. L.; Koehler, A.
N. Chem. Biol. 2006, 13, 493–504; (b) Bradner, J. E.; McPherson, O. M.; Koehler,
A. N. Nat. Protocol 2006, 1, 2344–2352; (c) Schmitz, K.; Haggarty, S. J.;
McPherson, O. M.; Clardy, J.; Koehler, A. N. J. Am. Chem. Soc. 2007, 129, 11346–
11347.
cycle was repeated another three more times. The photolyzed
sample was stored at ꢀ20 ꢁC until analyzed by LC/ESI-MS.
4.4. LC/MS and LC/MS/MS analyses of the photolysis products
The photolyzed solution was typically diluted 1/100 with water
or 0.05% aq formic acid and analyzed with a 4000Q TRAP mass
spectrometer (MSD Sciex, Foster City, CA, USA) equipped with an
Agilent 1100 LC system using the following conditions.
Flow rate: 100 ml/min.
15. Platz, M. S. Kinetics and Spectroscopy of Carbenes and Biradicals; Platz, M. S., Ed.;
Plenum: New York, NY, 1990; pp 143–212.
16. Tomioka, H. Res. Chem. Intermed. 1994, 20, 605–634.
17. Gupta, C. M.; Costello, C. E.; Khorana, H. G. Proc. Natl. Acad. Sci. U.S.A. 1979, 76,
3139–3143.
18. Brunner, J.; Senn, H.; Richards, F. M. J. Biol. Chem. 1980, 255, 3313–3318.
19. Nassal, M. J. Am. Chem. Soc. 1984, 106, 7540–7545.
Solvent A: 0.05% formic acid in water (HPLC grade).
Solvent B: 0.05% formic acid in acetonitrile (HPLC grade).
Gradient: 5%B/0–5 min; 5–45%B/5–65 min; 45–80%B/65–70 min.
Column: Inertsil ODS-3 1.5 mm i.d.ꢂ200 mm (GL Science, Tokyo,
Japan).
20. Platz, M.; Admasu, A. S.; Kwiatkowski, S.; Crocker, P. J.; Imai, N.; Watt, D. S.
Bioconjugate Chem. 1991, 2, 337–341.
Column oven temperature: 25 ꢁC.
21. Delfino, J. M.; Schreiber, S. L.; Richards, F. M. J. Am. Chem. Soc. 1993, 115, 3458–
3474.
22. Tomioka, H.; Suzuki, S.; Izawa, Y. J. Am. Chem. Soc. 1982, 104, 3156–3162.
23. The structures of these compounds were determined by comparison with
authentic synthetic samples. See Supplementary data.
Typical injection volume: 1
Ionization mode: ESIþ.
LC/MS scan mode: Q3.
LC/MS/MS scan mode: Triple Quadrupole (Q3 scan with Q1 pre-
cursor selection).
mL.
24. Diderich, G. Helv. Chim. Acta 1972, 55, 2103–2111.
25. Bethell, D.; Stevens, G.; Tickle, P. J. Chem. Soc. D 1970, 792–794.
26. Production of 5 was below the detection limit in this case.
27. Closs, G. L.; Rabinow, B. E. J. Am. Chem. Soc. 1976, 98, 8190–8198.
MS/MS (low-energy CID) condition: 45 V collision energy, N2 col-
lision gas.