Pd(0)/Cu(I)-mediated direct arylation of 2′-deoxyadenosines: Mechanistic role of Cu(I) and reactivity comparisons with related purine nucleosides

Article abstract of DOI:10.1021/jo9012282

(Chemical Equation Presented) Pd/Cu-mediated direct arylation of 2′-deoxyadenosine with various aryl iodides provides 8-arylated 2′-deoxyadenosine derivatives in good yields. Following significant reaction optimization, it has been determined that a subst

Full text of DOI:10.1021/jo9012282

pubs.acs.org/joc
Pd(0)/Cu(I)-Mediated Direct Arylation of 20-Deoxyadenosines: Mechanistic
Role of Cu(I) and Reactivity Comparisons with Related Purine Nucleosides
Thomas E. Storr,†,‡ Christoph G. Baumann,‡ Robert J. Thatcher,† Sara De Ornellas,†,‡
Adrian C. Whitwood,† and Ian J. S. Fairlamb*,†
†Department of Chemistry, University of York, Heslington, York YO10 5DD, U.K., and
‡Department of Biology, Area 10, University of York, Heslington, York YO10 5YW, U.K.
ijsf1@york.ac.uk
Received June 10, 2009
Pd/Cu-mediated direct arylation of 20-deoxyadenosine with various aryl iodides provides 8-arylated
20-deoxyadenosine derivatives in good yields. Following significant reaction optimization, it has been
determined that a substoichiometric quantity of piperidine (secondary amine) in combination with cesium
carbonate is necessary for effective direct arylation. The general synthetic protocol allows lower tempera-
ture direct arylations, which minimizes deglycosylation. The origin of the piperidine effect primarily derives
from the in situ generation of Pd(OAc)2[(CH2)5NH]2. Various copper(I) salts have been evaluated; only
CuI provides good yields of the 8-arylated-20-deoxyadenosines. Copper(I) appears to have a high binding
affinity for 20-deoxyadenosine, which explains the mandatory requirement for stoichiometric amounts of
this key component. The conditions are compared with more general direct arylation protocols, e.g.,
catalytic Pd, ligand, acid additives, which do not employ copper(I). In each case, no detectable arylation of
20-deoxyadenosine was noted. The conformational preferences of the 8-aryl-20-deoxyadenosine products
have been determined by detailed spectroscopic (NMR) and single crystal X-ray diffraction studies. Almost
exclusively, the preferred solution-state conformation was determined to be syn-C20-endo (ca. 80%). The
presence of a 2-pyridyl group at the 8-position further biases the solution-state equilibrium toward this
conformer (ca. 88%), due to an additional H-bond between H10 and the pyridyl nitrogen atom. The Pd/Cu
catalyst system has been found to be unique for adenosine type substrates, the reactivity of which has been
placed into context with the reported direct arylations of related 1H-imidazoles. The reactivity of other
purine nucleosides has been assessed, which has revealed that both 20-deoxyguanosine and guanosine are
incompatible with the Pd/Cu-direct arylation conditions. Both substrates appear to hinder catalysis, akin to
the established inhibitory effects in Suzuki cross-couplings with arylboronic acids.
Introduction
agents,3 novel ligands for metals,4 and supramolecular build-
ing blocks.5 These analogues allow biomolecular structure
and biochemical mechanisms to be probed,6 which in turn
provides the impetus to develop more efficient synthetic
routes to structurally distinct nucleosides, particularly
purine-based compounds.7 Indeed, Sonogashira, Suzuki,
and Stille cross-coupling processes, catalyzed by Pd(0),8
C-modified non-natural nucleosides are widely employed
as fluorescent markers,1 sensors to detect light,2 therapeutic
*To whom correspondence should be addressed. Tel: +44 (0)1904
434091. Fax: (0)1904 432516.
5810 J. Org. Chem. 2009, 74, 5810–5821
Published on Web 07/27/2009
DOI: 10.1021/jo9012282
r
2009 American Chemical Society

Storr et al.
JOCFeatured Article
provide C-functionalized purine nucleosides of varying sub-
stitution patterns.9 There is a particular interest in the
synthesis of 8-modified purine nucleosides and the following
bioapplications: conformational studies to assess DNA/
RNA base pairing,10 cytostatic properties against certain
tumor cell lines,11 antagonist effects at the A3 adenosine
receptor,12 thrombin inhibitory activity,13 fluorescence app-
lications,14 components in supramolecular assembly,15
among other possibilities.16 In addition to these bioapplica-
tions, other structurally diverse 8-aryl-20-deoxyadenosines
act as pH-sensing fluorescent probes,17 as biomarkers for
exposure to chemical carcinogens,18 and as luminescent and
electroactive labels19 (selected examples given in Figure 1).
Protection of both the sugar hydroxyl groups and the reac-
tive heteroaromatic substituents was considered mandatory
until recently developed conditions showed that unprotected
FIGURE 1. Selected 8-aryl-20-deoxyadenosines.
halogenated nucleosides can be effectively cross-coupled with
various nucleophilic components.20 Quite remarkably, haloge-
nated nucleoside triphosphates are also viable substrates for
Suzuki-Miyaura cross-couplings.21,22
(1) (a) Burgess, K.; Cook, D. Chem. Rev. 2000, 100, 2047. (b) Asseline, U.
Curr. Org. Chem. 2006, 10, 491. (c) Rist, M. J.; Marino, J. P. Curr. Org. Chem.
2002, 6, 775. For a fascinating special issue dedicated to this area, see: (d)
Tor, Y. Tetrahedron 2007, 63, 3425.
Significant progress has been made in the synthetic chemistry
of prefunctionalized nucleosides. A new challenge has emerged
in nucleoside chemical synthesis, which involves the use of
nonfunctionalized derivatives, e.g., 20-deoxyadenosine rather
than 8-bromo-20-deoxyadenosine. This approach would allow
one to greatly diversify the portfolio of C-functionalized purine
nucleosides with the ultimate goal of selectively synthesizing
C-functionalized nucleotides and oligonucleotides. The emer-
gence of direct arylation (functionalization) strategies for
aromatic and heteroaromatic compounds, which is a popular
replacement23 for classical cross-coupling methodologies,
allows us to assess the potential for selective C-H functiona-
lization in nucleosides, particularly purines.
(2) Gao, J. M.; Watanabe, S.; Kool, E. T. J. Am. Chem. Soc. 2004, 126,
12748.
(3) Alvarez-Salas, L. M. Curr. Top. Med. Chem. 2008, 8, 1379.
(4) Ropartz, L.; Meeuwenoord, N. J.; van der Marel, G. A.; van Leeuwen,
P. W. N. M.; Slawin, A. M. Z.; Kamer, P. C. J. Chem. Commun. 2007, 1556.
(5) (a) Cowley, M. J.; Lynam, J. M.; Whitwood, A. C. Dalton Trans. 2007,
4427. (b) Lynam, J. M. Dalton Trans. 2008, 4067.
(6) (a) De Clercq, E. Biochem. Pharmacol. 1984, 33, 2159. (b) Minakawa,
N.; Kojima, N.; Sasaki, T.; Matsuda, A. Nucleosides Nucleotides 1996, 15,
251. (c) De Clercq, E.; Descamps, J.; De Somer, P.; Barr, P. J.; Jones, A. S.;
Walker, R. T. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 2947. (d) Manfredini,
S.; Baraldi, P. G.; Bazzanini, R.; Marangoni, M.; Simoni, D.; Balzarini, J.;
De Clercq, E. J. Med. Chem. 1995, 38, 199. (e) Hocek, M.; Holy, A.; Votruba,
ꢀ
ꢀ
I.; Dvorakova, H. J. Med. Chem. 2000, 43, 1817.
(7) For a review covering the recent advances in the synthesis of purine
derivatives, see: Legraverend, M. Tetrahedron 2008, 64, 8585.
(8) Fairlamb, I. J. S. Tetrahedron 2005, 61, 9661.
(9) (a) Agrofoglio, L. A.; Gillaizeau, I.; Saito, Y. Chem. Rev. 2003, 103,
1875. (b) Hocek, M. Eur. J. Org. Chem. 2003, 245. (c) Lakshman, M. K. Curr.
Org. Synth. 2005, 2, 83.
Some methods have been described for the Pd-catalyzed
intermolecular direct arylation of similar compounds, e.g.,
(benz)imidazoles derivatives.24 However, rather curiously there
is a requirement for stoichiometric CuI additives, although
not exclusively.25 Furthermore, some success has been had
in arylating suitably protected purine derivatives, e.g., N9-
benzyl derivatives.26 In parallel with the Hocek group,27
(10) (a) Bhardwaj, P. K.; Vasella, A. Helv. Chim. Acta 2002, 85, 699. (b)
ꢀ
Eppacher, S.; Solladie, N.; Bernet, B.; Vasella, A. Helv. Chim. Acta 2000, 83,
1311. (c) Saenger, W. Principles of Nucleic Acid Structure; Springer Verlag:
Berlin, 1984; pp 69-78. (d) Catalanotti, B.; Galeone, A.; Paloma, L. G.; Mayol,
L.; Pepe, A. Bioorg. Med. Chem. Lett. 2000, 10, 2005.
(11) Manfredini, S.; Baraldi, P. G.; Bazzanini, R.; Marangoni, M.;
Simoni, D.; Balzarini, J.; De Clercq, E. J. Med. Chem. 1995, 38, 199.
(12) Volpini, R.; Costanzi, S.; Lambertucci, C.; Vittori, S.; Klotz, K.-N.;
Lorenzenc, A.; Cristalli, G. Bioorg. Med. Chem. Lett. 2001, 11, 1931.
(13) He, G. -X.; Krawczyk, S. H.; Swaminathan, S.; Shea, R G.;
(21) Capek, P.; Pohl, R.; Hocek, M. Org. Biomol. Chem. 2006, 4, 2278.
(22) (a) Collier, A.; Wagner, G. Org. Biomol. Chem. 2006, 4, 4526. (b)
Collier, A.; Wagner, G. Chem. Commun. 2008, 178.
ꢀ
Dougherty, J. P.; Terhorst, T.; Law, V. S.; Griffin, L. C.; Coutre, S.;
Bischofberger, N. J. Med. Chem. 1998, 41, 2234.
(14) (a) Srivatsan, S. G.; Greco, N. J.; Tor, Y. Angew. Chem., Int. Ed.
2008, 47, 6661. (b) Srivatsan, S. G.; Weizman, H.; Tor, Y. Org. Biomol.
Chem. 2008, 6, 1334. (c) Firth, A. G.; Fairlamb, I. J. S.; Darley, K.; Baumann,
C. G. Tetrahedron Lett. 2006, 47, 3529.
(15) (a) Sessler, J. L.; Sathiosatham, M.; Doerr, K.; Lynch, V.; Abboud,
K. A. Angew. Chem., Int. Ed. 2000, 39, 1300. (b) Davis, J. T. Angew. Chem.,
Int. Ed. 2004, 43, 668.
(23) For specific reviews concerning the direct arylation of heteroaro-
matics, see: (a) Seregin, I. V.; Gevorgyan, V. Chem. Soc. Rev. 2007, 36, 1173.
(b) Satoh, T.; Miura, M. Chem. Lett. 2007, 36, 200. Regioselective aspects in
the direct arylation of heteroaromatics are covered in: (c) Fairlamb, I. J. S.
Chem. Soc. Rev. 2007, 36, 1036. For general reviews, see: (d) Godula, K.;
Sames, D. Science 2006, 312, 67. (e) Campeau, L.-C.; Fagnou, K. Chem.
Commun. 2006, 1253. (f) Dick, A. R.; Sanford, M. S. Tetrahedron 2006, 62,
2439. For an excellent special issue dedicated to catalytic aromatic C-H
activation, see: Bedford, R. B. Tetrahedron 2008, 64, 5973.
(16) (a) Crisp, G. T.; Gore, J. Tetrahedron 1997, 53, 1523. (b) Gillet, L. C.
J.; Scha1rer, O. D. Org. Lett. 2002, 4, 4205.
(17) (a) Sun, K. M.; McLaughlin, C. K.; Lantero, D. R.; Manderville, R.
A. J. Am. Chem. Soc. 2007, 129, 1894. For a related application for 8-phenol-
20-deoxyguanosines, see: (b) Weishar, J. L.; McLaughlin, C. K.; Baker, M.;
Gabryelski, W.; Manderville, R. A. Org. Lett. 2008, 10, 1839.
(18) (a) Dai, Q.; Xu, D.; Lim, K.; Harvey, R. G. J. Org. Chem. 2007, 72,
(24) (a) Bellina, F.; Cauteruccio, S.; Mannina, L.; Rossi, R.; Viel, S. J.
Org. Chem. 2005, 70, 3997. (b) Bellina, F.; Cauteruccio, S.; Mannina, L.;
Rossi, R.; Viel, S. Eur. J. Org. Chem. 2006, 693. (c) Bellina, F.; Cauteruccio,
S.; Rossi, R. Eur. J. Org. Chem. 2006, 1379. (d) Bellina, F.; Cauteruccio, S.;
Rossi, R. J. Org. Chem. 2007, 72, 8543. (e) Bellina, F.; Calandri, C.;
Cauteruccio, S.; Rossi, R. Tetrahedron 2007, 63, 1970. (f) Bellina, F.;
Cauteruccio, S.; Di Fiore, A.; Rossi, R. Eur. J. Org. Chem. 2008, 543,
5445. (g) Bellina, F.; Cauteruccio, S.; Di Fiore, A.; Marchetti, C.; Rossi, R.
Tetrahedron 2008, 64, 6060.
ꢀ
4856. (b) Vrabel, M.; Hocek, M.; Havran, L.; Fojta, M.; Votruba, I.;
ꢀꢁ ꢁ ꢀ
´
sek, L.; Zendlova, L.; Hobza, P.; Shih,
ꢀ
Klepetarova, B.; Pohl, R.; Rulı
I.-h.; Mabery, E.; Mackman, R. Eur. J. Inorg. Chem. 2007, 12, 1752.
ꢀ
(19) (a) Vrabel, M.; Horakova, P.; Pivokova, H.; Kalachova, L.;
Cernocka, H.; Cahova, H.; Pohl, R.; Sebest, P.; Havran, L.; Hocek, M.;
ꢀ
ꢀ
ꢀ
ꢁ
ꢁ
ꢀ
ꢀ
(25) In ref 24a, it was shown that C2-arylation can occur as a minor
reaction pathway using catalytic Pd(OAc)2 (5 mol %), AsPh3 (10 mol %),
and CsF (2 equiv) in DMF at 140 °C. C5-arylation is usually the preferred
and major reaction pathway.
ꢀ
Fojta, M. Chem.;Eur. J. 2009, 15, 1144. (b) Vrabel, M.; Pohl, R.;
ꢀꢁ
ꢀ
Klepetarova, B.; Votruba, I.; Hocek, M. Org. Biomol. Chem. 2007, 5, 2849.
(20) (a) Collier, A.; Wagner, G. Synth. Commun. 2006, 36, 3713. (b)
ꢁ
Shaughnessy, K. H. Eur. J. Org. Chem. 2006, 1827. (c) Capek, P.; Hocek, M.
ꢁ
ꢁ
(26) Cerna, I.; Pohl, R.; Klepetarova, B.; Hocek, M. Org. Lett. 2006, 8,
ꢀꢁ
ꢀ
Synlett 2005, 3005. (d) Western, E. C.; Daft, J. R.; Johnson, E. M. II;
Gannett, P. M.; Shaughnessy, K. H. J. Org. Chem. 2003, 68, 6767.
5389.
ꢁ
(27) Cerna, I.; Pohl, R.; Hocek, M. Chem. Commun. 2007, 4729.
ꢁ
J. Org. Chem. Vol. 74, No. 16, 2009 5811

Storr et al.
JOCFeatured Article
we have investigated28 the reactivity of adenosine toward
selective C8-H arylation mediated by a Pd-Cu catalyst system
in the presence of aryl iodides using an appropriate base (in
our case, Cs2CO3; Hocek employed piperidine only) in DMF
at ∼120 °C. In a single example, the Hocek group reported
the Pd-Cu-mediated arylation of 20-deoxyadenosine using
4-iodotoluene heating to 125 °C for 5 h, which gave the
C8-arylated product in 31% yield (an 8% yield of a N6/C8-
diarylation product was also reported). Indeed, the N6-
arylation pathway29 is always a problem for this chemis-
try, particularly when run at higher temperatures.
SCHEME 1. Arylation of 20-Deoxyadenosine 1
Building on these findings, we reported a single result
concerning the C8-arylation of unprotected 20-deoxyadeno-
sine (1) under slightly modified conditions to that required
for adenosine (a change in temperature from 120 to 80 °C, to
avoid extensive deglycosylation in the case of 1). Subsequent
studies in our laboratories have revealed that the efficacy of
this specific reaction is dependent on how the DMF solvent is
purified. Therefore, we have chosen to comprehensively
explore the reactivity of 1 as a substrate for this type of
reaction toward iodobenzene initially under a variety of
conditions. We have determined that secondary amines
are necessary additives for reactions employing Pd(OAc)2
when run at 80 °C: either dimethylamine (generated by
degradation of DMF f NHMe2 + CO) or intentionally
added piperidine (as a dimethylamine mimic). The optimized
conditions have been applied against a range of aryl iodides
to deliver a series of novel 8-aryl-20-deoxyadenosine (2a-j)
compounds. A structural analysis of these compounds has
been carried out (by single-crystal X-ray diffraction studies
and solution-based NMR spectroscopic studies). The reac-
tivity of 1 has been compared with other related purine
nucleosides, e.g., guanosine, inosine, and fluorinated adeno-
sines. Defined trends emerge concerning the efficient C8-
arylations of these purine nucleosides.
(see Scheme 1), substantial deglycosylation for 1 was ob-
served, giving mainly 8-phenyladenine (2aa), which confirms
that direct arylation has taken place but that the resulting
8-phenyl-20-deoxyadenosine product (2a) is more susceptible
to deglycosylation than 1. At 80 °C, a higher yield of 2a was
achieved (84%).28 Therefore, milder conditions facilitate the
C8-arylation of the less stable 20-deoxyribose compound 1,
providing increased yields of the arylation product 2a.
Curiously, we noted that reaction efficacy was severely
affected by the use of a thoroughly dried and degassed batch
of DMF. The typical protocol for DMF purification is
distillation from MgSO4 under a N2 atmosphere, which
usually provides a sufficiently dry, oxygen-free solvent ready
for use. This procedure does not fully remove trace water,
but degassing by the freeze/pump/thaw method and storing
the distillate over 4 A molecular sieves essentially provides
“water-free” DMF. When using DMF purified in this
manner, we noticed that the yields for 2a were severely
affected (ca. 10% ( 5% isolated yields). It was also found
that if DMF was used directly from the distillation (without
degassing), the reaction yields were much higher (ca. 65%
yields). This allowed us to identify that low concentrations of
dimethylamine, a well-known degradation product of DMF
on prolonged heating, were necessary for efficient direct
arylation of 1. Interestingly, the use of thoroughly dried
and degassed DMF did not affect the efficacy of the arylation
of adenosine at 120 °C. This is most likely due to the higher
temperatures either producing the necessary amine in situ
(more likely) or as a result of a change in mechanism. Given
the low boiling point of dimethylamine (bp 7 °C)33 and the
practicalities associated with its handling, we chose to assess
other amine additives for optimization of this reaction using
thoroughly dry and degassed DMF (Table 1).
Results and Discussion
20-Deoxyadenosine 1 can be arylated at the C8-position
under the modified reaction conditions determined for ade-
nosine 3 vide infra. The instability of the glycosyl bond at
higher temperatures, e.g., >100 °C, is a significant problem
for this type of chemistry (note:3 is over 2 orders of magnitude
more stable toward deglycosylation than 1).30 Indeed, the
use of high-temperature microwave conditions [Pd(OH)2/C
(5 mol %), CuI (2 equiv), Cs2CO3 (2 equiv), NMP at 160 °C
with MWI] developed by Alami and co-workers31 for non-
ribose-based purine C8-arylation is ultimately precluded due
to this key detail.32 At 120 °C, using our standard conditions
(28) Storr, T. E.; Firth, A. G.; Wilson, K.; Darley, K.; Baumann, C. G.;
Fairlamb, I. J. S. Tetrahedron 2008, 64, 6125.
(29) One can selectively N-arylate protected 20-deoxyadenosines and 20-
deoxyguanosines with aryl bromides using a Pd-Xantphos catalytic system;
see: Ngassa, F. N.; DeKorver, K. A.; Melistas, T. S.; Yeh, E. A.-H.;
Lakshman, M. K. Org. Lett. 2006, 8, 4613.
(30) The t1/2 value for deglycosylation is expected to fall dramatically in
the presence of a base.
(31) Sahnoun, S.; Messaoudi, S.; Peyrat, J. -F.; Brion, J. -D.; Alami, M.
Tetrahedron Lett. 2008, 49, 7279.
(32) Using the conditions described by Alami and co-workers (ref 31), the
arylation of 1 (1 equiv) with iodoanisole (1.5 equiv) resulted in the formation
of a trace amount of 8-(p-methoxyphenyl)-20-deoxyadenosine and adenine
(the deglycosylation product derived from 1) in our hands [other conditions:
Pd(OH)2/C (5 mol %), CuI (1 equiv), Cs2CO3 (2 equiv), NMP, μW, 160 °C,
15 min].
(33) Aston, J. G.; Eidinoff, M. L.; Forster, W. S. J. Am. Chem. Soc. 1939,
6, 1539.
5812 J. Org. Chem. Vol. 74, No. 16, 2009

Storr et al.
JOCFeatured Article
TABLE 1. Effect of Amine Additives on the Reaction of 1 with Iodobenzene To Give 2aa
entry
amine additive
equivb
yield (%)c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
12
84,e 65,f 64f
50
14
7
47
57
65
41
41
28
5g
45
26
<5
dimethylamine (Me2NH)
traced
0.4
0.4
0.4
0.1
0.2
0.4
0.8
1.2
1.6
2.5
0.4
0.4
0.2
diethylamine (Et2NH)
diisopropylamine (i-Pr2NH)
diphenylamine (Ph2NH)
piperidine
piperidine
piperidine
piperidine
piperidine
piperidine
piperidine (without Cs2CO3)
pyrrolidine
triethylamine (Et3N)
N,N,N0,N0-tetramethylethylenediamine (TMEDA)
aReaction conditions as for Scheme 1 at 80 °C, using thoroughly dry and degassed DMF (entry 2 employed freshly distilled DMF). bWith respect to 1
(0.37 mmol scale). cIsolated yield, following chromatography on silica gel. dDue to solvent distillation (not quantified). ePreviously reported yield28 obtained
using freshly distilled solvent. fRepeat runs of this reaction. gProduct 2a was contaminated by a significant amount of piperidine HX (ca. 10 equiv).
3
SCHEME 2. Synthesis of Cu[N(CH2)5]
The reaction of 1 with iodobenzene under the standard
conditions using thoroughly dry and degassed DMF gave
compound 2a in 12% yield (entry 1, Table 1). In the presence
of trace dimethylamine (from freshly distilled, nondegassed
DMF), the yields obtained were substantially higher (yields
for three runs are shown in entry 2). In the presence of the
secondary amines diethylamine, diisopropylamine, and di-
phenylamine, yields of 50, 14, and 7% were recorded for 2a,
respectively (entries 3-5). Piperidine is a superior secondary
amine base for these reactions (entries 6-11). By varying the
concentration of this base additive, the most efficient reac-
tion was determined to be that employing a substoichio-
metric quantity of piperidine (0.4 equiv; entry 8). Higher
concentrations of piperidine lead to diminishing yields
(entries 9-11). When 2.5 equiv of piperidine in the absence
of Cs2CO3 was used, 2a was formed in ca. 5% yield, which
on heating to give Pd0 species in situ. We similarly determined
that piperidine (2 equiv) reacts with Pd(OAc)2 (1 equiv) in
DMF to afford trans-Pd(OAc)2[(CH2)5NH]2 at 20 °C in a
few seconds (Scheme 3). Pd0/Pd colloids are slowly formed,
following prolonged heating (15 h) of this complex in DMF
at 80 °C (with formation of 2,3,4,5-tetrahydropyridine),
indicating that this complex is likely a precatalyst. For the
actual direct arylation, other components in the reaction
mixture may serve to reduce PdII to Pd0. 36
Direct arylation of 1 proceeds to a certain extent using
Cu[N(CH2)5] complex (either 1 or 3 equiv; entries 1 and 2,
respectively) as a substitute for CuI (Table 2). However, the
yields of 2a do not mirror the identical reaction conducted in
the presence of CuI and piperidine.
The direct arylation of 1 with PhI mediated by trans-Pd-
(OAc)2[(CH2)5NH]2 (5 mol %), in the absence of added
piperidine, gave 2a in 64% yield (entry 3). Running an other-
wise identical reaction with 0.3 equiv of piperidine gave 2a
in 68% yield (entry 4). A similar reaction using 2.4 equiv
of piperidine gave 2a in 54% yield, which was contaminated
was contaminated with piperidine HX (entry 12). It was also
3
noted that pyrrolidine is an effective secondary base for this
reaction (entry 13); however, piperidine is superior at equiva-
lent loadings (compare entries 8 and 13). Tertiary amines
(triethylamine and TMEDA) are generally poor base addi-
tives for these reactions (entries 14 and 15, respectively). In
the case of TMEDA, it is plausible that the chelating
bidentate nature of this ligand could interfere with catalysis
(at PdII or CuI; or facilitate oxidation of CuI to CuII).
The use of piperidine as the sole base for the direct arylation
reaction of 1 has been employed by Hocek and co-workers
(also tested at 80 °C; entry 12, Table 1).27 However, our studies
allow us to conclude that the combination of Cs2CO3
(stoichiometric) and piperidine (substoichiometric) is more
effective in reactions run at lower temperatures, e.g., 80 °C.
At this stage in our study, the origin of the secondary amine
effect remained unclear. There were several possibilities; how-
ever, generation of a CuI-amide type complex or slower
reduction of “Pd(OAc)2L2” to give the active “Pd(0)” catalyst
seemed like the most plausible explanation.
by piperidine HX following chromatography on silica gel
3
(entry 5). In the absence of both CuI and piperidine additive
no product formation was observed (entry 6). To summarize,
the enhancing effect of piperidine (secondary amine) in the direct
arylations of 1 derives from N-ligation to PdII, most likely
providing a more effective precatalyst in situ.
To further determine the role of CuI, other CuI sources
have been evaluated, i.e., CuI-thiophene carboxylate and
CuOTf. These electrophilic CuI sources were unable to
promote the direct arylation of 1, highlighting the impor-
tance of the iodide anion. Of further interest is the finding
that tetra-n-butylammonium iodide is an ineffective substi-
tute for CuI.
The Cu[N(CH2)5] complex34 was prepared from mestyl-
copper(I) and piperidine in 49% yield (Scheme 2) vide infra.
It is established that secondary amines react with Pd(OAc)2
to give Pd(OAc)2(R2NH)2 complexes,35 which can also reduce
(34) Tsuda, T.; Watanabe, K.; Miyata, K.; Yamamoto, H.; Saegusa, T.
Inorg. Chem. 1981, 20, 2728.
(35) Tao, B.; Boykin, D. W. J. Org. Chem. 2004, 69, 4330.
(36) Prolonged incubation of Pd(OAc)2[(CH2)5NH]2 in DMF (at 20 °C)
revealed palladium adducts containing 2,3,4,5-tetrahydropyridine (by ESI-
MS/MS).
J. Org. Chem. Vol. 74, No. 16, 2009 5813

Storr et al.
JOCFeatured Article
SCHEME 3. Synthesis of trans-Pd(OAc)2[(CH2)5NH]2
a
aX-ray: the thermal ellipsoids shown at 50% probability; this and all subsequent X-ray crystal structure images were prepared using X-Seed and POVRay
(v. 3.6).
TABLE 2. Effect of Pd and Cu Sources on the Reaction of 1 with Iodobenzene To Give 2aa
aReaction conditions were otherwise identical to footnote a (Table 1). bWith respect to 1 (0.37 mmol scale). cIsolated yield following chromatography
on silica gel. dProduct 2a was contaminated by piperidine HX (ca. 1 equiv). eUnreacted 1 remaining (by TLC analysis).
3
As late transition metals are well-known to form stable
metal complexes with purines, a CuI complex contain-
ing 1 was synthesized. Reaction of CuI (1 equiv) with
1 (1 or 2 equiv) in DMF at 80 °C for 1 h gave in both cases
new CuI complexes (Scheme 4). ESI-MS/MS analysis
of both reaction mixtures revealed the [Cu(N6,N7-1)2]+
cation (m/z 564.9), which upon MS-MS fragmentation
gave [Cu(N6,N7-1)]+ (m/z 313.9) (Figure 2). Where the
CuI:1 ratio was 1:1 the product has been formulated as
[Cu(N6,N7-1)2]CuI2, whereas at a 1:2 ratio it was [Cu(N6,
N7-1)2]I.
The 1H NMR spectrum of [Cu(N6,N7-1)2]I in DMSO-d6
at 298 K (c=0.015 M) exhibits differences with 1 (Figure 3).
The purine protons (C2-H and C8-H) both exhibit line
broadening and become deshielded on coordination to CuI
[C8-H=δ 8.33, s, Δδ=C8-Hcomplex - C8-Hligand(1)=0.11;
Δv1/2 =25 Hz; C2-H=δ 8.25, s, Δδ=C2-Hcomplex - C2-
Hligand(1) = 0.12; Δv1/2 = 40 Hz], which confirms that 1 is
5814 J. Org. Chem. Vol. 74, No. 16, 2009

Storr et al.
JOCFeatured Article
SCHEME 4. Synthesis and Evaluation of CuI-20-Deoxyade-
nosyl Complexes
FIGURE 2. ESI-MS spectrum of [Cu(N6,N7-1)2]I (MS/MS on
m/z = 564.9 gives m/z 313.9 as the major fragment ion).
binding in a bidentate coordination mode.37 A similar
chemical shift difference and line broadening is exhibited
for the coordinating purine amino group [NH2=δ 7.43, s,
Δδ=NH2complex- NH2ligand(1)=0.14; Δv1/2 =30 Hz]. It is
interesting to note that the remaining chemical shifts and line
widths of the 20-deoxyribose motif in the complex are nearly
identical to those observed for 1. 13C NMR spectroscopic
data revealed the presence of sugar carbons in this complex;
however, the apparent exchange process that is occurring
results in none of the carbons of the purine ring being
observed (see the Supporting Information). At higher con-
centrations of [Cu(N6,N7-1)2]I in DMSO-d6 we noted
further line broadening and iodine liberation (c=0.15 M).
[Cu(N6,N7-1)2]CuI2 and [Cu(N6,N7-1)2]I were indepen-
dently subjected to the general direct arylation conditions
(Scheme 4). A 41% yield for 2a was recorded with the former
complex, whereas for the latter complex 2a only a trace
amount of product was detected by TLC analysis. This key
experiment confirms that a CuI:1 ratio of at least 1:1 is
necessary for effective direct arylation.
FIGURE 3. 1H NMR spectra: A, [Cu(N6,N7-1)2]I (c = 0.015 M);
B, compound 1. Both samples were dissolved in DMSO-d6 and run at
500 MHz (note: the singlet species at ca. δ 7.85 ppm is trace DMF).
direct arylation of caffeine [1 (1 equiv), PhI (2 equiv), Pd(OAc)2
(5 mol %), PivOH (10 mol %), Cs2CO3 (2.5 equiv), DMF,
80 °C, 15 h] “protocol A”; Larrosa and co-workers39, a method
for room-temperature direct arylation of indoles [1 (1 equiv),
PhI (2 equiv), Pd(OAc)2 (5 mol %), PivOH (1.5 equiv), Ag2O
(0.75 equiv), DMF, 80 °C, 15 h] “protocol B”; and Fagnou
and co-workers, a general method for the direct arylation of
heteroaromatic compounds [1 (1 equiv), PhI (1 equiv), Pd-
(OAc)2 (1mol%), PCy3 (2mol%), PivOH(30mol%), K2CO3
(1.5 equiv), DMF, 80 °C, 15 h] “protocol C”.
We have also evaluated three synthetic protocols which
utilize a Pd(OAc)2 precatalyst in the presence of pivalic acid
(PivOH), specifically similar conditions to those recently
described as follows: You and co-workers, a method for
(37) In DMSO-d6, adenosine acts as
a monodentate ligand (N7-
coordinated) in PtII complexes. Importantly, only a chemical shift difference
is observed at C8-H (Δδ=0.21) on coordination to PtII (the C2-H chemical
shift was unaffected). Similar (small) chemical shift differences are observed
for PdII-adenosine complexes; see: (a) Dehand, J.; Jordanov, J. J. Chem.
(39) Lebrasseur, N.; Larrosa, I. J. Am. Chem. Soc. 2008, 130, 2926. The
reported reactions in this paper were generally run at lower temperature than
our experiment.
ꢀ
(40) Liegault, B.; Lapointe, D.; Caron, L.; Vlassova, A.; Fagnou, K. J. Org.
Chem. 2009, 74, 1826. Note: this paper generally employed a reaction tempera-
ꢀ
Soc., Chem. Commun. 1976, 598. (b) Quirtk, M.; Salas, J. M.; Sanchez, M. P.;
Beauchamp, A. L.; Solans, X. Inorg. Chim. Acta 1993, 204, 213.
(38) Zhao, D.; Wang, W.; Lian, S.; Yang, F.; Lan, J.; You, J. Chem.;
Eur. J. 2009, 15, 1337.
ture of 100 °C, PCy3 HBF4 and DMA as the solvent. In our experiment, PCy3
3
(stored in a drybox, <1 ppm O2) and DMF were employed.
J. Org. Chem. Vol. 74, No. 16, 2009 5815

Storr et al.
JOCFeatured Article
SCHEME 5. Deuteration of the C8-Position in 1
TABLE 3. Substrate Scope in the Direct Arylation of 1 Using Various
Aryl Iodides
For comparison purposes, and because of the risk of
deglycosylation, a temperature of 80 °C was used in this
study. We anticipated that protocol A would be most
suited to the adenine ring system. The high activity of
protocol B encouraged us to test it on our very different
heteroaromatic ring system (and at higher temperature).
Finally, we anticipated that protocol C would be ineffec-
tive because similar substrates, e.g., imidazoles and related
ring systems do not effectively couple at the C2-H position
(C8-H in purines) under these conditions.41 In brief, for
each set of conditions, 1 was not arylated (note: predomi-
nantly recovered starting materials after 15 h); these sur-
prising results are placed into context vide infra.
In keeping with the previously reported28 base deprotona-
tion of adenosine 3, compound 1 is deprotonated in the
absence of metal under “arylation conditions”. Acetone-d6
provides a suitable source of deuterium in the presence of
Cs2CO3 in DMF at 80 °C for 1 h (Scheme 5). A nearly
complete exchange occurred at C8, and there was no evi-
dence for deuterium incorporation at C2. Running this
reaction in the presence of Cs2CO3 (2.5 equiv) and piperidine
(0.4 equiv) resulted in the same outcome (96% deuteration at
C8). Under identical conditions, 20-deoxyguanosine was not
deuterated at the C8 position.
Substrate Scope. Having established the optimized reac-
tion conditions for the direct arylation of 1, further reac-
tions of other aryl iodides have been evaluated (Table 3). In
addition to direct arylation working well for iodobenzene,
bromobenzene was also found to be compatible with the
reaction conditions (entry 1). The direct arylation of 1 using
aryl iodides which possessed p-OMe and m-OMe substitu-
ents proceeded well (entries 2 and 3) and compared favorably
against iodobenzene (entry 1). The p-Me- and p-F-substi-
tuted aryl iodides gave good yields of the arylated products
2d and 2e, respectively (entries 4 and 5). p-Nitroiodobenzene
reacted with 1 to give compound 2f in modest yield (46%),
which was accompanied by a trace quantity of the N6-
arylation product (2f0). It is perhaps then surprising that
the m-nitroiodobenzene arylated compound 1 effectively
(entry 7).
aYield after chromatography on silica gel, reactions performed on a
1.88 mmol scale (with respect to 1) unless otherwise stated. bUsing PhBr.
cN-Arylated product (2f’) formed (ca. 1%). dPerformed on a 1.71 mmol
scale (with respect to 1).
Wishing to expand the scope of products formed under
the direct arylation conditions, two arbitrarily selected
brominated heteroaromatics were evaluated as coup-
ling partners (Scheme 6). 6-Methoxy-2-bromopyridine
was found to couple with 1 to give 2k in 53% yield. 3-Bro-
mothiophene coupled with 1 to give 2l in a modest 32% yield.
Conformational and Structural Analysis of the Arylation
Products. The NMR spectroscopic data of the 8-aryl-20-deox-
yadenosines provided informative details concerning (i) the
orientation (conformation) of the purine ring with respect to
the glycosyl bond in the 20-deoxyribose moiety and (ii) the ring
conformation of the 20-deoxyribose moiety (Figure 4).
Both p-trifluoromethyliodobenzene and m-trifluoro-
methyliodobenzene provided arylated products 2h and 2i
in 95 and 99% yields, respectively (entries 8 and 9).
2-Iodonapthalene essentially mirrored the reactivity of iodo-
benzene, affording the arylated product 2j in 58% yield
(entry 10).
(41) One can get around the 1H-imidazole problem by using N-oxide
derivatives; see: (a) Campeau, L. -C.; Stuart, D. R.; Leclerc, J. -P.; Bertrand-
Laperle, M.; Villemure, E.; Sun, H. -Y.; Lasserre, S.; Guimond, N.;
Lecavallier, M.; Fagnou, K. J. Am. Chem. Soc. 2009, 131, 3291. Note: it
would be difficult to selectively prepare a similar N-oxide of compound 1.
5816 J. Org. Chem. Vol. 74, No. 16, 2009

Storr et al.
JOCFeatured Article
20-deoxyadenosine 1 (75% C20-endo and 25% C30-endo; for
adenosine 3, 67% C20-endo and 33% C30-endo). We presume
that the addition of the 8-aryl group affects the 20-deoxyr-
ibose conformation, possibly as a result of these compounds
preferring a syn-orientation (as opposed to the anti orienta-
tion, which is observed for 1).46
Atomic-level information on the conformational prefer-
ences of the 8-aryl-20-deoxyadenosines has been gained by
single-crystal X-ray diffraction studies. Compound 2a was
crystallized from a MeOH solution using the slow evapora-
tion method. The X-ray data47 revealed two different mole-
cules in the asymmetric unit cell (Figure 5).
The syn-conformation in molecule 1 is stabilized by an
intramolecular H-bond48 [O(3A)-H(3A) N(2)=2.00(3) A;
3 3 3
173(3)°], whereas molecule 2 does not possess such a H-bond
despite still being in a syn-conformation. The torsion angle,
FIGURE 4. Conformational preferences in 8-aryl-20-deoxyadenosines.
4
0
e.g., O -C10-N9-C4 (Figure 4), is an effective measure for
the orientation of the purine relative to the 20-deoxyribose
(∼<90°=syn;>90°=anti). For molecule 1, markedly different
torsion angles are observed [O(1A)-C(6)-N(3)-C(2) =
63.8(3)°, syn-C20-endo; O(1B)-C(6)-N(3)-C(2) =101.1(3)°,
syn-C30-exo]. The latter can be considered as moving toward an
anti-orientation. In molecule 2, the torsion angle O(4)-C(22)-
N(8)-C(18) is 85.83(19)°, syn-C30-endo/C40-exo.
SCHEME 6. Evaluation of Brominated Heteroaromatic Cou-
pling Partners
For comparison, 8-phenyladenosine28 4 was crystallized
from a MeOH/pyridine (1:1, v/v) solution using the slow
evaporation method (see the Supporting Information). The
X-ray data reveals that the molecule adopts a syn-conforma-
tion with respect to the purine ring and ribose motif [torsion
angle O(1)-C(6)-N(3)-C(2) = 49.76(17)°, syn-C20-endo],
which is stabilized49 by an intramolecular H-bond [O(4)-
H(4) N(2)=1.92(3) A; 174(2)°].
3 3 3
Compound 2j was crystallized from MeOH using the slow
evaporation method (Figure 6). The X-ray data50 revealed
a fascinating structure exhibiting strong intermolecular π-π
stacking interactions between the two naphthyl groups
(the centroid51 distance was found to be ca. 3.865 A).
According to established rules,42 the 8-aryl-20-deoxyade-
nosines (2a-j) can be found to preferentially adopt the
expected syn-orientation (by comparison of the chemical
shifts of C10, C20, C30, C40, and H20; see the Supporting
Information for further details). It is well established that the
20-deoxyribose moiety can adopt various ring conforma-
tions.43 The two low energy conformers are referred to as
C20-endo (south type) and C30-endo (north type), for which
there is a small barrier to interconversion (ca. 2 kcal mol-1).44
These conformations exhibit unique spectroscopic proper-
ties, and while the different conformations cannot be resolved
(46) (a) Hocquet, A.; Leulliot, N.; Ghomi, M. J. Phys. Chem. B 2000, 18,
4560. (b) Stolarski, R.; Dudycz, L.; Shugar, D. Eur. J. Biochem. 1980, 108,
111. (c) Uesugi, S.; Ikehara, M. J. Am. Chem. Soc. 1977, 10, 3250. (d)
Gannett, P. M.; Sura, T. P. Chem. Res. Toxicol. 1993, 5, 690.
(47) X-ray data for compound 2a (CCDC 721575): C32H38N10O8.50; Mw=
698.69; T=110(2) K; λ=0.71073 A; triclinic; P1 space group; a=7.3242(7) A,
b=10.7071(11) A, c=11.3268(12) A; R=96.261(2)°, R=99.177(2)°, γ=
108.178(2)°; V=821.10(14) A3; Z=1, D=1.413 Mg/m3; crystal size=0.45 ꢀ
0.26 ꢀ 0.24 mm3; R1=0.0467, wR2=0.0961 (all data)
1
(48) For solid-state XRD studies illustrating intramolecular H-bonding
and a syn-conformation in related purine nucleosides, see: (a) Lesyng, B.;
Marck, C. H.; Saenger, W. Z. Naturforsch. 1984, 39c, 720. (b) Koyama, G.;
Nakamura, H.; Umezawa, H.; Iitaka, Y. Acta Crystallogr. 1976, B32, 813. (c)
Koyama, G.; Nakamura, H.; Umezawa, H. Acta Crystallogr. 1976, B32, 969.
For related solution NMR spectroscopic studies, see: (d) Koole, L. H.; de
Boer, H.; de haan, J. W.; Haasnoot, C. A. G.; van dael, P.; Buck, H. M. J.
Chem. Soc., Chem. Commun. 1986, 362. (e) Hamm, M. L.; Rajguru, S.;
Downs, A. M.; Cholera, R. J. Am. Chem. Soc. 2005, 127, 12220. (f) Ghosh, A.
K.; Lagisetty, P.; Zajc, B. J. Org. Chem. 2007, 72, 8222.
on the NMR time scale, the H NMR spin-spin coupling
constants can be used to calculate the relative populations of
each conformer in solution at ambient temperature.45 The
relative population of the C20-endo conformer was deter-
mined by the eq C20-endo= 100[J1 2 /(J1 2 + J3 4 )] (see the
Supporting Information for further details). The C20-endo
conformation is adopted preferentially by the new 8-aryl-
20-deoxyadenosines, representing ∼80% of the molecular
population in all cases, which is slightly higher than
0
0
0
0
0 0
(49) It is established that the rotation about the C(10)-N(9) and C(40)-
C(50) bonds is influenced by the intramolecular H-bond from HO-C(50) to
N(3) in purines; see: (a) Fuji, S.; Fujiwara, T.; Tomita, K. Nucleic Acids Res.
1976, 3, 1985. (b) Gunji, H.; Vasella, A. Helv. Chim. Acta 2000, 83, 1.
(50) X-ray data for compound 2j (CCDC 721576): C41H44N10O8; Mw=
804.86; T=120(2) K; λ=1.5418 A; triclinic; P1 space group; a=7.8036(3) A,
b = 10.1562(3) A, c = 12.3581(3) A; R = 93.768(2)°, β= 94.058(3)°, γ =
101.822(3)°; V=953.03(5) A3; Z=1; D=1.402 Mg/m3; crystal size=0.75 ꢀ
0.17 ꢀ 0.13 mm3; R1=0.0335, wR2=0.0929 (all data).
(42) (a) Stolarski, R.; Dudycz, L.; Shugar, D. Eur. J. Biochem. 1980, 108,
111. (b) Uesugi, S.; Ikehara, M. J. Am. Chem. Soc. 1977, 99, 3250. (c)
Gannett, P. M.; Sura, T. P. Chem. Res. Toxicol. 1993, 6, 690.
(43) Van Roey, P.; Salerno, J. M.; Chu, C. K.; Schinazi, R. F. Proc. Natl.
Acad. Sci. U.S.A. 1989, 86, 3929.
(44) Brameld, K. A.; Goddard, W. A. III. J. Am. Chem. Soc. 1999, 121, 985.
(45) (a) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1973, 95, 2333.
(b) Esho, N.; Desaulniers, J. P.; Davies, B.; Chui, H. M. P.; Rao, M. S.;
Chow, C. S.; Szafert, S.; Dembinski, R. Bioorg. Med. Chem. 2005, 13, 1231.
(51) The following centroids were found for 2j (molecule 1: centroid=
C11, C12, C17, C18, C19, and C20; molecule 2: centroid=C31, C32, C37,
C38, C39, C40) using Mercury 2.2 (Build RC5), copyright CCDC 2001-
2008, http://www.ccdc.cam.ac.uk/mercury/.
J. Org. Chem. Vol. 74, No. 16, 2009 5817

Storr et al.
JOCFeatured Article
FIGURE 5. X-ray structure of compound 2a (note: arbitrary number-
ing used in 3D structure). Thermal ellipsoids are shown at 50% prob-
ability. An alterative conformer for the structure on the right is shown in
black bonds with the atoms shown as spheres (C-H atoms and solvent
molecules removed for clarity). The syn-conformer in molecule 1 (right)
is stabilized by an intramolecular H-bond, while molecule 2 (left) adopts
this conformation in the absence of this interaction.
FIGURE 6. X-ray structure of compound 2j (note: arbitrary num-
bering used in 3D structure). Thermal ellipsoids are shown at 50%
probability.
The syn-orientation is adopted in both molecules 1 and 2
[molecule 1: torsion angle O(1)-C(6)-N(3)-C(2) =
67.5(2)°, syn-C20-endo; molecule 2: torsion angle O(1)-
C(6)-N(3)-C(2)=53.7(2)°, syn-C20-endo]. Intramolecular
H-bonding is apparent [molecule 1: O(3)-H(3) N(2)=
3 3 3
1.94(4) A; 165(3)°; molecule 2: O(6)-H(6A) N(7) =
3 3 3
1.84(4) A; 168(3)°], while intermolecular H-bonding inter-
actions with methanol and water molecules for molecule 1
[O(8)-H(8A) N(4) = 1.96(3) A; 169(3)°] and 2 [O(5)-
3 3 3
H(5) O(7) = 1.86(4) A; 168(3)°], respectively, are in-
3 3 3
volved in the crystal packing. X-ray structure determina-
tions of compounds 2b and 2f0 have also been made (see the
Supporting Information).
The 1H and 13C NMR spectroscopic data for 2k indicate
that the molecule shows the highest preference for the syn-
C20-endo conformation (88%). In addition, there is a marked
difference in the chemical shift of C10-H in 2k (δ 7.34) relative
to the other compounds in this series (2a-j: C10-H δ ∼6.11-
6.17; avg = δ 6.15). This difference is attributed to an
intramolecular H-bond between the 8-pyridyl nitrogen and
H10. The proposal is supported by the increased C50O-H
chemical shift (δ 6.11) which also appears downfield rela-
tive to the other compounds in this series (2a-j: C50O-H
δ ∼5.39-5.61; avg=δ 5.52). These intramolecular H-bonds
serve to stabilize the syn-C20-endo conformation (Figure 7).
Reactivity Comparisons with Other Purines. We compared
the reactivity of various purines relative to 20-deoxyadeno-
sine 1 and adenosine 3 in order to probe the possible
“templating” role of the exocyclic amine in adenosine. We
envisaged that 20-deoxyguanosine 5 and guanosine 7 could
be problematic substrates because CuI-coordination can
occur at sites distal to the C8 position. Using our optimized
conditions, poor yields of the direct arylation products 6 and
FIGURE 7. Interaction of the 8-pyridyl group with H10 in 2k.
8, respectively, were obtained (entry 1, Table 4). We also note
that C8-H deuteration experiments performed on com-
pounds 5 and 7 (using conditions in Scheme 5) did not show
any appreciable incorporation of deuterium at C8.
The ionizable proton at N2 in guanine-type substrates is
an issue in these reactions. Therefore, we chose to also
evaluate 20-deoxyinosine 9 and inosine 11. While 9 gave the
direct arylation product 10 in low yield (19%), the latter gave
product 12 in 60% yield (entry 2). An adenosine type
substrate, i.e., 20-fluoro-20-deoxyadenosine 13, was arylated
most effectively, affording 14 in 94% yield (entry 3). As
noted in previous reports52 the 20-fluoro-substituent induces
a conformational flip to syn-C30-endo (ca. 61%). Arylation
of 2-fluoro-20-deoxyadenosine 15 also occurred, although
nucleophilic displacement of fluorine by piperidine is a
(52) Uesugi, S.; Miki, H.; Ikehara, M.; Iwahashi, H.; Kyogoku, Y.
Tetrahedron Lett. 1979, 20, 4073.
5818 J. Org. Chem. Vol. 74, No. 16, 2009

Storr et al.
JOCFeatured Article
TABLE 4. Reactivity Comparison for Related Purine Nucleosides:
Direct Arylation with Iodobenzenea,b
SCHEME 7. Effects of 20-Deoxyguanosine 5 on the Arylation
of 1
exhibit similar, if not identical reactivity, and would prefer
to undergo functionalization at C5 rather than C2.54 We are
able to comment on the mechanistic details of the Pd-Cu
catalyst system by comparing our findings herein with pre-
vious work showing that imidazole derivatives are generally
resistant to C2-H functionalization (in the absence of Cu).41
Despite the recent mechanistic proposal by You and co-
workers38 for the Pd-catalyzed C8-H functionalization of
caffeine, we propose that a concerted metalation-deproto-
nation (CMD) process55 is unlikely for the Pd-Cu catalyst
system described here and elsewhere24 for related “imida-
zole-like” substrates. We have demonstrated that a combi-
nation of Cs2CO3 and piperidine (secondary amine) is
necessary for obtaining good yields of the 8-arylated 20-de-
oxyadenosines at 80 °C. Crucially, substoichiometric quan-
tities of piperidine (0.4 equiv) were required for optimal
yields. The mechanism whereby piperidine (the secondary
amine) acts as a promoter for the reaction primarily derives
from the generation of a Pd(OAc)2[(CH)5NH]2 in situ. The
independent preparation of this complex and evaluation in
the direct arylation of 1 with PhI gave 2a in yields compar-
able to those reactions mediated by Pd(OAc)2/piperidine.
We have shown that Cu[N(CH2)5] resulted in lower yields as
compared to otherwise identical direct arylation reactions
run in the presence of CuI. Other more reactive and “naked”
CuI sources, e.g., CuI-thiophene carboxylate or CuOTf,
resulted in negligible reaction. In addition, tetra-n-butylam-
monium iodide was a poor substitute for CuI. To probe the
CuI role further, the independent preparation of a CuI
complex containing two 20-deoxyadenosine ligands (1) was
undertaken. Altering the CuI:N6,N7-(1) stoichiometry to either
1:1 or 1:2 gave a cationic CuI species, e.g., [CuI(N6,N7-1)2]+,
which was unambiguously characterized by ESI-MS/MS
aReagents and conditions. bReactions of 2’-deoxypurines run at 80 °C
using the optimized conditions (otherwise at 120 °C, as previously
reported). cYield following chromatography on silica gel. dYield is based
on piperidine.
competing reaction giving 16 in 61% yield (entry 4; yield
based on piperidine, the limiting reagent).
The poor reactivity of 20-deoxyguanosine 5 in these reactions
led us to test whether the guanine moiety was inhibiting the
catalytic direct arylation process, akin to the insightful observa-
tions made by Western and Shaughnessy in aqueous Pd-catalyzed
Suzuki cross-coupling of 8-bromo-20-deoxyguanosine (and other
unprotected halonucleosides) with arylboronic acids.53 The yield
of 2a was diminishedslightly in the presence of a substoichiometric
quantity of 5 (0.1 equiv), while a stoichiometric quantity of 5
hindered the reaction, diminishing the yield of 2a (no arylation
products of 5 were detected; 2a was isolated following chromato-
graphy on silica gel) (Scheme 7).
1
and H NMR spectroscopy. The direct arylation only pro-
ceeded satisfactorily in the presence of an additional equiva-
lent of CuI, e.g., in [CuI(N6,N7-1)2]CuI2, to give 2a in 41%
yield. This outcome confirms that CuI has a high binding
affinity for 1 under the direct arylation conditions. Crucially,
this “templating” indicates that CuI assists the C-H func-
tionalization process, and excess CuI is required to disrupt
an otherwise stable complex, e.g., I f II (Scheme 8). It is
interesting to note that a similar Pd-Cu catalyst system
was found to be broadly ineffective for the direct arylation of
9-benzyl-6-phenyl-7-deazapurine at 160 °C,56 suggesting
Several important observations have been made in this
study. “General” direct arylation reaction conditions proved
ineffective for the reaction of 1 with iodobenzene to give 2a.
Related 1-aryl-1H-imidazoles or 1-benzyl-1H-imidazoles
(55) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am. Chem. Soc.
2005, 127, 16754. (b) Garcia-Cuadrado, D.; Braga, A. A. C.; Maseras, F.;
Echavarren, A. M. J. Am. Chem. Soc. 2006, 128, 1066. (c) Garcia-Cuadrado,
D.; de Mendoza, P.; Braga, A. A. C.; Maseras, F.; Echavarren, A. M. J. Am.
Chem. Soc. 2007, 129, 6880. (d) Pascual, S.; De Mendoza, P.; Braga, A. A. C.;
Maseras, F.; Echavarren, A. M. Tetrahedron 2008, 64, 6021. (e) Ackermann,
L.; Vicente, R.; Althammer, A. Org. Lett. 2008, 10, 2299.
(53) (a) Western, E. C.; Shaughnessy, K. H. J. Org. Chem. 2005, 70, 6378.
For further coordination chemistry studies on guanosine and related purines,
see: (b) Dehand, J.; Jordanov, J. J. Chem. Soc., Dalton Trans. 1977, 1588.
(54) The reaction of either 4-bromoanisole or 2-bromonaphthalene with
N-methylimidazole gave exclusively the C5-arylated products in 34% and
40% yields, respectively (in both cases, following lengthy reaction times, e.g.,
25 and 27 h, respectively); see ref 40.
ꢁ
ꢀꢁ ꢀ
(56) Klecka, M.; Pohl, R.; Klepetarva, B.; Hocek, M. Org. Biomol. Chem.
2009, 7, 866.
J. Org. Chem. Vol. 74, No. 16, 2009 5819

Storr et al.
JOCFeatured Article
SCHEME 8. Proposed Mechanism for the Pd-Cu-Mediated
Direct Arylation of 1
III could be in equilibrium with a nucleophilic C8-carbene
III0 (which can also be described as an imidazolium ylide).60
We anticipate III0 would be a strong σ-donating ligand for
PdII. The established reductive elimination61 of certain or-
gano groups from “R-PdII-NHC” complexes (R=Me, aryl
or acyl; NHC = N-heterocyclic carbene) to give Pd0 and
R-NHC products highlights this as alternative mechanistic
path for the reaction.
The optimized direct arylation conditions for 1 are essen-
tially incompatible with 20-deoxyguanosine 5 (or guanosine
7). Shaughnessy and Western53 were able to show that the
ionization of the imide group in guanine is a problem for
Suzuki cross-couplings, leading to Pd catalyst inhibition.
The pH of our direct arylation is comparable to these Suzuki
cross-couplings (pH ∼10), and we anticipate similar catalyst
inhibition. Our competition experiments confirm that
5 slows down the direct arylation of 1; separate reactions
of 5 or 7 with iodobenzene provide only low yields of the
8-phenylated products (6 and 8, respectively).
At a reaction temperature of 80 °C, the direct arylation
conditions described herein have been completely optimized
for the functionalization of adenine-type ring systems. Small
changes in the structure of the purine ring affect the direct
arylation yields significantly (see Table 4), and this has to be
balanced against the inevitable deglycosylation problem for
reactions run at higher temperatures, e.g., >100 °C, which is
an issue for 20-deoxyribose purine derivatives.
On a general note, the yields of 8-aryl-20-deoxyriboses
obtained from the Pd/Cu-mediated direct arylation of 1
described herein are similar to those reported53 for Suzuki
cross-coupling of 8-bromo-20-deoxyribose with aryl or
heteroaryl boronic acids. However, one needs to consider
that C8-bromination of 1 by either (1) NBS, DMF, rt62 or (2)
Br2, NaOAc, H2O, rt63 gives 8-bromo-20-deoxyadenosine in
40% yield (in both cases). Suzuki cross-couplings are effi-
cient in their own right, but they do require bromination
(prefunctionalization) of 1.
In summary, we have comprehensively studied the reactivity
of 20-deoxyadenosine 1 in Pd-Cu-mediated direct arylations.
Following reaction optimization, we have found that second-
ary amines promote direct arylations, and are necessary
additives at 80 °C. The binding affinity of CuI for 20-deoxy-
adenosine dictates that stoichiometric amounts of CuI are
necessary to deliver an activated CuI-nucleoside complex
(e.g., II in Scheme 8). Reactivity comparisons with structurally
related purines showed that the guanine moiety is disfavored in
the C8 direct arylation reaction, possibly due to CuI “templat-
ing” at nonadjacent sites and a different rate of base-assisted
that the “imidazole-like” purine (N7) is necessary to aid
coordination to either Pd and/or Cu. This coordination
hypothesis is supported by our studies and by the compre-
hensive work of Bellina and Rossi on 1H-imidazoles and
related compounds.24 The base-assisted deprotonation step
leads to the formation of 8-cuprio-20-deoxyadenosine III,57
which can then enter a standard Pd(0) catalytic cycle. The
process can be related to alkynylcuprate formation in the
Sonogashira reaction,58 and although this proposal suggests
that CuI ought to be catalytic, the high binding affinity of 1,
and presumably product 2a, for CuI leads to the requirement
for excess CuI in the direct arylation reaction. The CsCO3H
formed during step II f III can in principle be metered out
by piperidine (assuming that the irreversible generation of
CsOH and CO2 does not occur) to give a carbamic acid/
carbamate salt derivative, representing a possible secondary
role for piperidine. This argument is supported by the
established solution behavior59 of similar DMF soluble
carbamic acid salt derivatives. Finally, it is plausible that
(60) Professor Fabio Bellina (University of Pisa, Italy) commented on this
possibility for the related Pd/Cu-mediated direct arylation of imidazoles at
the Catalytic C-H Activation of Organic Molecules Conference held in
Lyon, France (5-8 April 2009). See also: Chodowska-Palicka, J.; Milsson,
M. Synthesis 1974, 128.
(61) Reductive elimination from Me-PdII-NHC: (a) McGuinness, D. S.;
Saendig, N.; Yates, B. F.; Cavell, K. J. J. Am. Chem. Soc. 2001, 123, 4029 and
references cited therein. Ar-PdII-NHC: (b) McGuinness, D. S.; Cavell, K. J.;
Skelton, B. W.; White, A. H. Organometallics 1999, 18, 1596. (c) Caddick, S.;
Cloke, F. G. N.; Hitchcock, P. B.; Leonard, J.; de, K.; Lewis, A. K.; McKerrecher,
D.; Titcomb, L. R. Organometallics 2002, 21, 4318. Acyl-PdII-NHC: (d)
McGuinness, D. S.; Cavell, K. J. Organometallics 2000, 19, 4918.
(57) Similar “cuprated” purines have been recently utilized in oxidative
amination processes; see: Boudet, N.; Dubbaka, S. R.; Knochel, P. Org. Lett.
2008, 10, 1715.
(58) Cai, C.; Vasella, A. Helv. Chim. Acta 1995, 78, 2053. Note: we do not
rule out the possible involvement of bimetallic “Pd-Cu” species in the
reaction mechanism, as suggested in this paper concerning Sonogashira
cross-coupling of alkynes or haloalkyne derivatives with aryl halides.
(59) (a) Salvatore, R. N.; Chu, F.; Nagle, A. S.; Kapxhiu, E. A.; Cross,
R. M.; Jung, K. W. Tetrahedron 2002, 58, 3329. (b) Masuda, K.; Ito, Y.;
Horiguchi, M.; Fujita, H. Tetrahedron 2005, 61, 213.
(62) Nandanan, E.; Camaioni, E.; Jang, S. -Y.; Kim, Y. -C.; Cristalli, G.;
Herdewijn, P.; Secrist, J. A. III; Tiwari, K. N.; Mohanram, A.; Harden, T.
K.; Boyer, J. L.; Jacobson, K. A. J. Med. Chem. 1999, 42, 1625.
(63) Chen, Y.; Shen, L.; Zhang, F.; Lau, S. S.; van Breemen, R. B.;
Nikolic, D.; Bolton, J. L. Chem. Res. Toxicol. 1998, 11, 1105.
5820 J. Org. Chem. Vol. 74, No. 16, 2009

Storr et al.
JOCFeatured Article
deprotonation at the C8-position. Our results provide greater
insight into the direct arylation reactivity of not only purines,
e.g., the adenine moiety, but also 1H-imidazoles in general.
We are currently identifying milder reaction conditions for
metal-mediated direct functionalization of purines.
which was dried under high vacuum (ca. 0.8 mmHg). The crude
mixture was redissolved/suspended in MeOH/CH2Cl2 (1:1,
20 mL) and adsorbed onto silica gel (approximately 0.5 g) with
reduction in vacuo. A short silica gel column (approximately
10 g) was eluted using MeOH/CH2Cl2 (2:98 v/v, moving in
stepwise increments to 10:90 by gradient elution). The fractions
containing the product were combined, the solvents were re-
moved in vacuo, and CH2Cl2 (10 mL) was added and then
removed. The isolated product was then dried under high
vacuum and the purity ascertained by HPLC.
Experimental Section
General details can be found in the Supporting Information.
All X-ray crystal structures were processed in X-seed (author:
Prof. L. J. Barbour, version 2.0; with March 2008 update),64
then traced using Pov-Ray for Windows (author: C. Cason,
version 3.6) and labeled using POV-label (part of the X-seed
suite of software).
Acknowledgment. We thank Professor A. Vasella (ETH,
Zurich) for suggesting the possible role of carbamic acid/
salts under our direct arylation conditions in the presence of
piperidine. We also thank Drs. I. Larrosa (QMUL, London),
J. Slattery (York), and J. Lynam (York) for useful discus-
sions. EPSRC and Replizyme Ltd. (T.E.S. studentship),
BBSRC (grant award to C.G.B. and S.D.O. studentship),
and the Royal Society (University Research Fellowship to
I.J.S.F.) are acknowledged for funding this work. Astra-
Zeneca (Dr. David Hollinshead) is thanked for an unrest-
ricted research award (to I.J.S.F.). The EPSRC Service
at Durham University (UK) is acknowledged for running
13C solid-state NMR experiments (Dr. D. C. Apperley).
Dr. R. G. Sturmey and Professor H. J. Leese are acknow-
ledged for the use of an Agilent 1100 HPLC system.
General Procedure for the Direct Arylation of 1. To a vacuum-
dried Schlenk tube were added 20-deoxyadenosine (473 mg,
1.88 mmol, 1.0 equiv), Cs2CO3 (1.53 g, 4.68 mmol, 2.5 equiv),
CuI (1.07 g, 5.61 mmol, 3.0 equiv), Pd(OAc)2 (21 mg, 94 μmol,
5 mol %), and the aryl iodide (3.74 mmol, 2.0 equiv). The
reaction vessel was evacuated under high vacuum at 25 °C with
stirring and then flushed with N2 (three cycles). Dry distilled and
degassed DMF (10 mL) or “extra dry” DMF (Acros) was then
added along with degassed piperidine (stored over 3 A molecular
sieves, 74 μL, 0.75 mmol, 0.4 equiv). The vessel was sealed and
heated in an oil bath at 80 °C and stirred continuously for 15 h.
The mixture was then allowed to cool to 25 °C and 1 M HCl
solution (10 mL) added. The pH was then adjusted to 6.5 with
1 M NaOH and the aqueous solution extracted with a iPrOH/
EtOAc (1:9, v/v, 5 ꢀ 50 mL) mixture by decanting from the
reaction mixture. The organic extracts were combined, dried
(MgSO4), filtered, and reduced in vacuo to yield a thick gum,
Supporting Information Available: Experimental details
(including characterization data for all compounds) and
X-ray data (for 1, 2a,b,f0,j, 4, and trans-Pd(OAc)2[(CH2)5NH]2,
including CIF files). This material is available free of charge
via the Internet at http://pubs.acs.org.
(64) Barbour, L. J. J. Supramol. Chem. 2001, 1, 189-191. Also see http://
x-seed.net/.
J. Org. Chem. Vol. 74, No. 16, 2009 5821