Synthesis of nucleoside analogues in a ball mill: Fast, chemoselective and high yielding acylation without undesirable solvents

Article abstract of DOI:10.1039/c1gc15131b

The chemoselective acylation of primary aliphatic amines has been achieved in under ten minutes (and for aromatic amines under 120 min) using vibration ball-milling, avoiding undesirable solvents which are typically employed for such reactions (e.g. DMF).

Full text of DOI:10.1039/c1gc15131b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 1778
www.rsc.org/greenchem
PAPER
Synthesis of nucleoside analogues in a ball mill: fast, chemoselective and
high yielding acylation without undesirable solvents†
Francesco Ravalico, Stuart L. James and Joseph S. Vyle*
Received 2nd February 2011, Accepted 17th March 2011
DOI: 10.1039/c1gc15131b
The chemoselective acylation of primary aliphatic amines has been achieved in under ten minutes
(and for aromatic amines under 120 min) using vibration ball-milling, avoiding undesirable
solvents which are typically employed for such reactions (e.g. DMF). Under optimised conditions,
the synthesis of amides in the presence of both primary and secondary alcohol functions was
achieved in high to excellent yields (65–94%). Overall, the methods described have significant
practical advantages over conventional approaches based upon bulk solvents including greater
yields, higher chemoselectivity and easier product separation.
milling conditions in order to better understand and exploit the
Introduction
technique.
Mechanochemistry is a well-established technique in materials
The current work relates to azobenzene-(AB-)derivativised
science1 but has only recently begun to be developed for
nucleoside and nucleotide mimetics. These have enabled the
molecular inorganic2 and organic synthesis.3 The E-factor4 of
organic reactions can be considerably decreased when using
vibrational or planetary ball-milling as minimal or even no
solvent is required. In addition, improvements in the yield and
selectivity can be observed,5 the reaction times and temperatures
irradiation-dependent manipulation of several oligonucleotide
properties through reversible E → Z photoisomerisation.14
The introduction of AB chromophores into oligonu-
cleotides has most commonly been performed with the
para-phenylazobenzoyl (p-AB-CO-) moiety appended to D-
can be reduced6,7 and especially in the context of reactions
threoninol.15 However, in our hands, the synthesis of the
using inorganic catalysts, expeditious and atom-efficient work-
up procedures can be employed.8
Ball-milling remains particularly under-exploited for the
preparation or derivatisation of biological molecules and their
corresponding ortho- and meta-AB derivatives of threoninol
under literature conditions in DMF met with limited success.
In particular, difficulties arose during their isolation due to their
susceptibility to photoisomerisation to the Z-isomer, their high
analogues. In this regard, recent reports have described the
water solubility and the similar retention characteristics of the
application of this technique for the preparation or derivati-
multiple side-products and isomers on silica gel in the presence
sation of amino acid analogues,9 peptide linkages,10 sec-
of residual solvent.
ondary metabolites11,12 and functionalised nucleosides.13 The
Herein, we report the synthesis of o-, m-, and p-AB-derivatised
amphiphilic nature of nucleosides and their analogues, cou-
aminonucleoside analogues and propargylamine carried out
pled with typically high levels of hydration, means that their
in a vibration ball mill. Chemoselective acylation of these
derivatisation through conventional methods usually requires
polyfunctional substrates was rapid, completely avoided the
bulk quantities of purified, high boiling and toxic or carcinogenic
use of DMF, required minimal work-up and thereby avoided
solvents such as DMF, DMSO or pyridine. There is considerable
undesirable photoisomerisation.
scope therefore for reducing the time, energy and toxicity of
the materials used in nucleoside chemistry by applying ball-
milling. In addition, more generally, it is important to conduct
the systematic optimisation of reaction parameters under ball-
Results and Discussion
N-Hydroxysuccinimidyl (NHS) esters are well established for
amine-selective bioconjugation due to the relative insensitivity
of such activated esters towards hydrolysis even at alkaline
pH.16 In order to optimise the ball-milling conditions, a model
reaction between propargylamine (1) and p-AB-CO-NHS (2a,
Fig. 1a) was therefore investigated. The progress of this reaction
was monitored using C18 RP-HPLC at 360 nm (Fig. 1b).
At this wavelength only those compounds containing the
School of Chemistry and Chemical Engineering, Queen’s University
Belfast, David Keir Building, Stranmillis Road, BT9 5AG, Belfast,
Northern Ireland, UK. E-mail: j.vyle@qub.ac.uk; Fax: (+)44 (0) 28
9097 6524; Tel: (+)44 (0) 28 9097 5485
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c1gc15131b
1778 | Green Chem., 2011, 13, 1778–1783
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 1 a) Model reaction of propargylamine (1) with N-hydroxysuccinimidyl para-phenylazobenzoate (p-AB-CO-NHS, 2a) used for optimising the
ball-milling conditions; b) typical HPLC chromatograms of the ball-milling reaction between 1 (2 eq.) and 2a in the presence of DMAP (0.2 eq.) at
28 Hz, monitored at 360 nm at different time points (0.5–30 min).
AB chromophore were detected and isocratic elution readily
separated the more hydrophobic NHS-ester (2a) from the major
product p-AB-propargylamide (3a) and minor quantities of the
hydrolysis product (p-AB-CO2H, 4).
Reaction conditions were thereby optimised in a 25 mL
stainless steel jar containing a 15.0 mm stainless steel ball on
a 50 mg scale (of 2a) by varying the following factors:
1. frequency of milling
2. equivalents of propargylamine and N,N-dimethyl-4-
aminopyridine (DMAP)
3. nature of the base
Influence of milling frequency
We have previously described solvent-free ball-milling for
reactions of nucleoside amine and hydroxyl functions at a
high vibration frequency (30 Hz).13a At lower frequencies (e.g.
25 Hz), incomplete reactions were observed even with prolonged
reaction times. In the present study, we monitored the reaction
between excess propargylamine and the NHS-ester (2a) in the
presence of DMAP over 30 min over a range of frequencies
(20 to 30 Hz).
Fig. 2 Rate of reaction to p-AB-propargylamide (3a) (
10 mins) overall level of hydrolysis to p-AB-acid (4) ( ; after 10 mins
using different vibration frequencies (2 eq. 1, 0.2 eq. DMAP, 20–30 Hz).
; after
Consistent with our previous work, increasing the vibration
frequency to 28 Hz led to a more rapid reaction (Fig. 2). At
28 Hz, a maximum yield of 95% of the propargylamide 3a was
achieved within 10 min and no further reaction was observed
upon further milling. Surprisingly, at 30 Hz, a slower reaction
rate was observed such that after ten minutes there was only
77% consumption of 2a accompanied by enhanced levels of the
corresponding acid (4). Although milling at 28 Hz gave the most
rapid reaction and highest yield, it is notable that all of the other
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1778–1783 | 1779
©

View Article Online
Table 1 Influence of the equivalents of propargylamine (1) and DMAP
on the yield of p-AB-propargylamide (3a) at 28 Hz
Table 2 Influence of the nature of the base on the yield of p-AB-
propargylamide (3a) at 28 Hz, using 2 equivalents of propargylamine,
and 1.5 equivalents of base
time (min)
time (min)
1
2
5
10
20
30
1
2
5
10
20
30
entry
1 (eq)a
DMAP (eq)a
yield (%) of 3ab
a
entry
base
pKa
yield (%) of 3ab
1
2
3
4
5
6
7
8
1.0
1.0
1.5
1.5
1.5
1.5
2.0
2.0
2.0
2.0
2.0
3.0
3.0
0.2
0.5
0.2
0.5
1.0
1.5
0
0.2
0.5
1.0
1.5
0
23
30
45
45
64
47
30
58
70
82
89
69
78
25
41
47
54
79
71
46
61
79
55
91
72
88
27
41
53
57
77
84
55
77
80
81
92
78
91
27
42
58
70
82
84
56
84
88
90
95
85
93
27
43
59
70
82
89
60
88
88
91
95
88
93
27
46
69
70
82
89
61
91
88
91
95
91
93
1
2
3
4
DMAP
DABCO
DBU
9.20
8.93
11.50
10.25
89
56
68
60
91
83
75
76
92
87
72
87
95
90
83
88
95
90
83
89
95
90
83
89
Na2CO3
a Aqueous pKa is quoted; b by HPLC, determined at 360 nm.
9
of other dinitrogen bases as well as sodium carbonate to
support the reaction. 1,4-Diazabicyclo[2.2.2]octane (DABCO)
has proved to be an effective catalyst under solvent-free
conditions6,18 but in our hands, the DABCO-catalysed reaction
was neither as rapid nor as high yielding as that using DMAP
(Table 2, entry 2). In contrast, 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU) is more basic than DMAP but has recently
been reported to act as a nucleophilic catalyst for acylation
reactions.19 In these reports, the DBU-catalysed reactions were
considerably faster and higher yielding than those employing
DMAP. However, we observed slower acylation accompanied
by higher levels of hydrolysis (up to 15%) leading to lower
yields of 3a (Table 2, entry 3) which may be accounted for
by the hygroscopicity of DBU. Overall, in the ball mill, the
presence of DMAP provided a significantly faster reaction
than DBU and a slight rate enhancement compared with
DABCO.
10
11
12
13
0.2
a equivalents quoted relative to 2a; b by HPLC, determined at 360 nm
vibration frequencies investigated gave 90% or greater yield after
30 min (although with higher levels of hydrolysis).
In a seminal work on the variation in energy requirements
for a solvent-free Suzuki–Miyaura reaction, Ondruschka and
coworkers compared the input for a microwave reactor, a
planetary ball-mill and a vibration ball-mill.17 Related to our
observation, in their study a lower yield in the vibration mill at
15 Hz compared with that at 13.3 Hz was observed.
Influence of the amounts of propargylamine and DMAP
Mayr and coworkers have reported that the nucleophilicity of
DABCO towards benzhydrilium ions in acetonitrile is 103 times
that of DMAP and that the nucleofugalities of the resulting
cationic alkylated-DABCO intermediates are ca. 106 times
greater than those of the corresponding DMAP intermediates.20
No such rate enhancement in the ball mill reaction was ob-
served. Furthermore, the same group have described the carbon
basicities of these cationic intermediates which follow the order
DBU > DMAP > DABCO.21 In the ball mill, under conditions
in which the concentration of intermediate acyl-ammonium
ions were determined by the rate of acylation, we should
anticipate the rate order to follow this order of carbon basicities.
We therefore assume that under solvent-free conditions, other
factors also operate. In particular, the low melting points
of DMAP (108–110 ◦C) and DABCO (156 ◦C) may mean
that the faster rate of reaction with DMAP is a function
of the speed with which liquifaction of the reaction mixture
occurs—both reactions formed pastes in the early stages of ball-
milling but subsequently became powders when the reaction
finished.
N-hydroxysuccinimide is acidic and therefore will protonate
both propargylamine and DMAP following the acylation re-
action. The lowering of both the reaction rate and the extent of
the reaction by such protonation was clearly demonstrated by
using only one equivalent of propargylamine in the presence of
substoichiometric DMAP (Table 1 entries 1 and 2).
Optimisation of the total amount of base present with
different ratios of DMAP and propargylamine was therefore
performed. Erratic yields were observed for some reactions
during the first two minutes of ball-milling which can be
accounted for by inhomogenous mixing. However, clear trends
were apparent after five minutes, indicating the influence of
DMAP as a catalyst. Thus, using either 1.5 or 2 equivalents
of propargylamine, significant enhancements in the rates of
reaction were observed, as determined by the yield of 3a after
five minutes, with increasing amounts of added DMAP (Table
1 entries 3–6 and 7–11). Using 1.5 equivalents of DMAP, the
reaction was complete within 10 min (entry 11), accompanied
by minimal hydrolysis (ca. 3%) to 4.
Finally, we note that using a larger excess of propargylamine
with either 0 or 0.2 equivalents of DMAP also gave excellent
yields, although the rate of the uncatalysed reaction was slower
(Table 1, entries 12 and 13).
Sodium carbonate buffers are commonly employed in the
acylation of amines by NHS-esters requiring a mixed aqueous-
organic solvent system such as those utilised in the post-synthetic
labelling of amino-functionalised DNA.22
Typically, long reaction times are required and the isolations
are therefore often associated with high levels of NHS-ester hy-
drolysis. Although the addition of anhydrous sodium carbonate
gave a reaction which was also slower than that observed in the
Influence of DMAP and other bases
Based upon the use of excess propargylamine and 1.5 equivalents
of DMAP (Table 2, entry 1), we next investigated the capacity
1780 | Green Chem., 2011, 13, 1778–1783
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 3 Preparative-scale reactions under optimised conditions for the synthesis of AB-amides
entry
1
R1
R2
product
yield (%)a
94%
2
3
92%
90%
4b
79%
5b
77%
83%
6b
7b
90%
79%
8
9
78%
65%
10
a Isolated yields; b 50 ml Ethyl acetate added; U = N-1-uracilyl.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1778–1783 | 1781
©

View Article Online
DMAP-catalysed reaction, nearly 90% conversion was observed
(Table 2, entry 4) compared with 61% in the absence of the base
(Table 1, entry 7).
Conclusions
We have demonstrated that, under optimised conditions, the
acylation of amines using NHS esters can be performed in equal
or greater yields than by conventional DMF-based approaches;
indeed the motivation for this work was that efficient isolation
from DMF of ortho- or meta-AB D-threoninol adducts was not
feasible in our hands.
Preparative-scale reactions and chemoselective
acylation
A survey of the ball-milling reaction conditions in terms of
frequency, ratio of reactants and the nature of the additional
base on a small-scale enabled preparative-scale reaction condi-
tions to be developed without a significant difference between
the isolated yield and the conversion observed using HPLC. We
note that a distinct advantage of using ball-milling for reactions
involving photosensitive groups is that exposure to incident UV-
Vis radiation is minimised.
The optimisation of solvent and temperature conditions solely
for the purpose of rendering substrates and reagents soluble
is often required in nucleoside and nucleotide chemistry using
molecular solvents. Solvent-free methods have considerable
scope to facilitate this chemistry in general, reducing both the
E-factor and the amount of energy required, as well as providing
higher yields, faster processes and easier product isolation.
To test the effectiveness of this solvent-free process further, the
optimised DMAP-catalysed conditions determined above were
applied to preparative-scale reactions using para- (2a), meta- (2b)
and ortho- (2c) AB-esters (Table 3, entries 1–3). In contrast to the
solution-phase reactions in DMF, it was gratifying to note that
the less reactive o-AB- and m-AB-esters also completely reacted
in under 10 min to the corresponding propargylamides (3a–c)
upon ball-milling. These products could be isolated readily in
excellent yields following extraction.
Applying the same reaction conditions to the preparation of
D-threoninol AB-derivatives (6a–c), was found to give incom-
plete reactions even after 30 min ball-milling at 28 Hz.
Solvent-assisted grinding (often called liquid-assisted grind-
ing, or LAG) has been used to improve ball-milling conversions
in both co-crystallisations23 and metal complexations.24 In the
present case, the addition of stoichiometric quantities of ethyl
acetate significantly enhanced the rates and yields to give 6a–
c with only minor amounts of the side-products observed
by TLC analysis. We ascribe these less polar, AB-containing
side-products to the reaction of the activated esters at both
the amine and hydroxyl functions, since they gave negative
ninhydrin tests. Ethyl acetate was then added to the reaction
vessel and the resulting solution applied to a short silica
gel column. Pure products were isolated following isocratic
elution with ethyl acetate in high yields. Although this work-
up and isolation procedure required larger volumes of solvent
compared with those used for 3a–c, ethyl acetate is categorised
as a preferred solvent according to the Pfizer solvent selection
guide.25
Acknowledgements
This work has been funded by the School of Chemistry and
Chemical Engineering, QUB.
Notes and references
1 (a) T. Frisˇcˇic´, J. Mater. Chem., 2010, 20, 7599–7605; (b) A. Pichon and
S. L. James, CrystEngComm, 2008, 10, 1839–1847; (c) A. L. Garay,
A. Pichon and S. L. James, Chem. Soc. Rev., 2007, 36, 846–855; (d) D.
Braga, S. L. Giaffreda, F. Grepioni, A. Pettersen, L. Maini, M. Curzi
and M. Polito, Dalton Trans., 2006, 1249–1263.
2 (a) P. Bala´zˇ, in Mechanochemistry in Nanoscience and Minerals
Engineering, Springer Berlin Heidelberg, 2008, 1–102; (b) L. Takacs,
J. Miner. Met. Mater. Soc., 2000, 52, 12.
3 (a) G. Kaupp, in Organic Solid State Reactions, Springer-Verlag
Berlin, Berlin, 2005, vol. 254, pp. 95–183; (b) B. Rodr´ıguez, A.
Bruckmann, T. Rantanen and C. Bolm, Adv. Synth. Catal., 2007,
349, 2213–2233; (c) A. Bruckmann, A. Krebs and C. Bolm, Green
Chem., 2008, 10, 1131–1141; (d) F. M. Kerton, in Alternative Solvents
for Green Chemistry, The Royal Society of Chemistry, Cambridge,
2009, 23–43.
4 R. A. Sheldon, Pure Appl. Chem., 2000, 72, 1233–1246.
5 G. W. Wang, Y. W. Dong, P. Wu, T. T. Yuan and Y. B. Shen, J. Org.
Chem., 2008, 73, 7088–7095.
6 J. Mack and M. Shumba, Green Chem., 2007, 9, 328–330.
7 T. H. Zhang, G. W. Wang, P. Lu, Y. J. Li, R. F. Peng, Y. C. Liu, Y.
Murata and K. Komatsu, Org. Biomol. Chem., 2004, 2, 1698–1702.
8 F. Bernhardt, R. Trotzki, T. Szuppa, A. Stolle and B. Ondruschka,
Beilstein J. Org. Chem., 2010, 6(No 7).
Likewise, using solvent-assisted grinding, the protected 2¢-
amino-2¢-deoxynucleoside analogue (7) completely reacted with
minimal side-products. The corresponding p-AB adduct (8) was
isolated with excellent yields (Table 3, entry 7).
The Fmoc protecting group is widely applied for the solid-
phase synthesis of peptides and often requires mixed solvent
systems and a problematic work-up. Even though it is thermally
sensitive, the NHS ester (9) reacted selectively with D-threoninol
to give 10 in high yields (Table 3, entry 8).
Finally, the acylation of aromatic amines using solvent-free
ball-milling was investigated (Table 3, entries 9 and 10). The
relative reactivities of p-anisidine (11) and naphthylamine (13)
compared with the aliphatic amines were found to be consistent
with Cline and Hanna’s report of NHS-ester aminolysis in 1,4-
dioxan. In this study, p-anisidine was found to be approximately
an order of magnitude more reactive than aniline and half as
reactive as benzylamine.26 It is perhaps noteworthy that the
time for the reaction to complete between p-nitrobenzoyl-NHS
9 A. Baron, J. Martinez and F. Lamaty, Tetrahedron Lett., 2010, 51,
6246–6249.
10 V. Declerck, P. Nun, J. Martinez and F. Lamaty, Angew. Chem., Int.
Ed., 2009, 48, 9318–9321.
11 (a) T. Szuppa, A. Stolle, B. Ondruschka and W. Hopfe, Chem-
SusChem, 2010, 3, 1181–1191; (b) T. Szuppa, A. Stolle, B.
Ondruschkaa and W. Hopfe, Green Chem., 2010, 12, 1288–
1294.
◦
12 X. Y. Zhu, Z. H. Li, Q. F. Shu, C. F. Zhou and W. K. Su, Synth.
and aniline at 24.6 C was reported to be 60 days. Using ball-
Commun., 2009, 39, 4199–4211.
milling, reaction times of 20 or 120 min were required to effect
the complete reaction to the corresponding anisidide (12) or
napthylamide (14), respectively.
13 (a) N. Giri, C. Bowen, J. S. Vyle and S. L. James, Green Chem., 2008,
10, 627–628; (b) S. A. Sikchi and P. G. Hultin, J. Org. Chem., 2006,
71, 5888–5891.
1782 | Green Chem., 2011, 13, 1778–1783
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
14 (a) S. Keiper and J. S. Vyle, Angew. Chem., Int. Ed., 2006, 45, 3306–
3309; (b) H. Asanuma, D. Tamaru, A. Yamazawa, M. Z. Liu and M.
Komiyama, ChemBioChem, 2002, 3, 786–789; (c) D. Matsunaga, H.
Asanuma and M. Komiyama, J. Am. Chem. Soc., 2004, 126, 11452–
11453; (d) H. Asanuma, T. Takarada, T. Yoshida, D. Tamaru, X. G.
Liang and M. Komiyama, Angew. Chem., Int. Ed., 2001, 40, 2671–
2673.
15 H. Asanuma, X. Liang, H. Nishioka, D. Matsunaga, M. Liu and M.
Komiyama, Nat. Protoc., 2007, 2, 203–212.
16 G. W. Anderson, J. E. Zimmerman and F. M. Callahan, J. Am. Chem.
Soc., 1964, 86, 1839–1842.
17 F. Schneider, T. Szuppa, A. Stolle, B. Ondruschka and H. Hopf,
Green Chem., 2009, 11, 1894–1899.
18 E. M. C. Ge´rard, H. Sahin, A. Encinas and S. Bra¨se, Synlett, 2008,
2702–2704.
19 (a) K. E. Price, C. Larrive´e-Aboussafy, B. M. Lillie, R. W.
McLaughlin, J. Mustakis, K. W. Hettenbach, J. M. Hawkins and
R. Vaidyanathan, Org. Lett., 2009, 11, 2003–2006; (b) C. Larrive´e-
Aboussafy, B. P. Jones, K. E. Price, M. A. Hardink, R. W.
McLaughlin, B. M. Lillie, J. M. Hawkins and R. Vaidyanathan, Org.
Lett., 2010, 12, 324–327.
20 (a) M. Baidya and H. Mayr, Chem. Commun., 2008, 1792–1794;
(b) F. Brotzel, Y. C. Chu and H. Mayr, J. Org. Chem., 2007, 72, 3679–
3688.
21 M. Baidya, S. Kobayashi, F. Brotzel, U. Schmidhammer, E.
Riedle and H. Mayr, Angew. Chem., Int. Ed., 2007, 46, 6176–
6179.
22 S. H. Weisbrod and A. Marx, Chem. Comm., 2008, 5675–5685 and
references cited therein.
23 (a) T. Frisˇcˇic´ and W. Jones, J. Pharm. Pharmacol., 2010, 62, 1547–
1559; (b) S. Jones, J. Chem. Soc. Perkin Trans. 1, 2002, 1–21.
24 (a) W. Yuan, J. O’Connor and S. L. James, CrystEngComm, 2010,
12, 3515–3517; (b) W. Yuan, T. Frisˇcˇic´, D. Apperley and S. L. James,
Angew. Chem. Int. Ed., 2010, 49, 3916–3919; (c) T. Frisˇcˇic´ and L.
Fa´bia´n, CrystEngComm, 2009, 11, 743–745; (d) G. A. Bowmaker, J.
V. Hanna, R. D. Hart, B. W. Skelton and A. H. White, Dalton Trans.,
2008, 5290–5292.
25 K. Alfonsi, J. Colberg, P. J. Dunn, T. Fevig, S. Jennings, T. A. Johnson,
H. P. Kleine, C. Knight, M. A. Nagy, D. A. Perry and M. Stefaniak,
Green Chem., 2008, 10, 31–36.
26 G. W. Cline and S. B. Hanna, J. Am. Chem. Soc., 1987, 109, 3087–
3091.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1778–1783 | 1783
©