A New Microfluidic Phase-Transfer Reaction Using HPLC Guard Columns as the Reactor for the N3-Protection of Uridine Derivatives

Article abstract of DOI:10.1055/s-0035-1560264

N³-Acylation of uridine derivatives with acyl chlorides, mediated by a phase-transfer reaction, was studied using a new microfluidic device containing an HPLC guard as an effective reactor. The acylated products were obtained in more than 80% y

Full text of DOI:10.1055/s-0035-1560264

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2015, 26, 2578–2582
letter
2578
N. Tago et al.
Letter
Synlett
A New Microfluidic Phase-Transfer Reaction Using HPLC Guard
Columns as the Reactor for the N3-Protection of Uridine Deriva-
tives
Nobuhiro Tagoa
organic solution
O
Yoshiaki Masakia
Hiroshi Nagasawab
Takashi Kanamoria
Akihiro Ohkuboa
Kohji Seio*a
O
O
NH
O
N
N
O
O
Si
O
Cl
N
O
O
Si
O
O
O
OH
Si
HPLC
guard
column
O
Mitsuo Sekine*a
O
OH
Si
a Department of Life Science, Tokyo Institute of Technology,
4259 Nagatsuta, Midoriku, Yokohama, Kanagawa, 226-8501,
Japan
selective acylation
reaction time <2 s
yield >80%
aqueous solution
Na2CO3
TBAB
kseio@bio.titech.ac.jp
msekine@bio.titech.ac.jp
b Kankyo Resilience Co., Ltd, 79-7 Tokiwadai, Hodogaya-ku,
Yokohama, 240-0067, Japan
resilience_info@yahoo.co.jp
Received: 13.07.2015
Accepted after revision: 12.08.2015
Published online: 22.09.2015
N3- (3a) and 4-O- (2a) benzoylated derivatives, compound
2a is converted into 3a by heating at 60 °C in 1,2-dichlo-
roethane. Although this reaction is very fast and effective in
small-scale preparations, the scaling up of the PT reaction
in a batch reactor decreases the reaction rate and lowers the
product yields. It is generally accepted that the PT reaction
rate is highly dependent on the interfacial area, and also
corresponds to the transfer rate of reactants from the aque-
ous phase to the organic phase promoted by the PT cata-
lyst.8
DOI: 10.1055/s-0035-1560264; Art ID: st-2015-u0542-l
Abstract N3-Acylation of uridine derivatives with acyl chlorides, medi-
ated by a phase-transfer reaction, was studied using a new microfluidic
device containing an HPLC guard as an effective reactor. The acylated
products were obtained in more than 80% yields in very short reaction
times of several seconds.
Key words microfluidic system, phase-transfer reaction, nucleoside
chemistry, flow chemistry, HPLC guard column
To accelerate PT reactions, several research groups have
applied microfluidic devices to achieve the large interfacial
areas that are necessary for efficient PT reactions.8–13 For
example, Kitamori et al. reported a notable improvement in
the diazo coupling reaction in a microfluidic system. The
reaction proceeded with quantitative conversion in the mi-
crofluidic system, whereas the yield was 80% with a batch
reactor.9 This result was rationalized in terms of the in-
creased interfacial area in the microfluidic system, which
was estimated to be twice that in the batch reactor.
In addition to the expanded interfacial area, Colin et al.
and Yuan et al. have reported that the segmented flow pat-
tern formed when the organic and aqueous solutions were
mixed at a Y-branch10 also contributes to the acceleration of
the PT reaction, because the efficient mixing inside each
solution refreshes the reactants near the interfacial area.11
Herein, we report that the use of a microfluidic reaction
system improves the reaction rate of N3-acylation of uridine
derivatives under PT conditions. We also show that the in-
sertion of porous materials such as HPLC guard columns
into the reaction tubing further increases the reaction rate.
In recent decades, modified nucleic acids have shown
potential as gene therapeutic drugs1 and gene-detection
probes.2 2′-O-Modification of RNAs has been used to im-
prove the nuclease resistance and binding affinity for target
nucleic acids.3 For the modification of the 2′-hydroxy group
of uridine, 3′,5′-O-(1,1,3,3-tetraisopropyldisiloxane-1,3-di-
yl)uridine has been widely used as the intermediate.4 How-
ever, because the N3- or 4-O-positions are reactive moi-
eties, in some cases, protection of these positions is neces-
sary.5 Various acyl groups have been used as the protecting
group for the uracil base,6 and we have previously reported
that acyl groups can be introduced to the N3-position very
effectively by using a phase-transfer (PT) reaction.7
The PT reaction of uridine derivatives is performed as
shown in Scheme 1 (R1 = R2 = H). In these reactions, com-
pound 1 is benzoylated with benzoyl chloride using CH2Cl2
as the organic phase and an aqueous solution of sodium
carbonate as the aqueous phase. The carbonate anion is
transferred into the organic phase by the PT catalyst tetra-
butylammonium bromide (TBAB). After the formation of
© Georg Thieme Verlag Stuttgart · New York — Synlett 2015, 26, 2578–2582

2579
N. Tago et al.
Letter
Synlett
O
O
a: R1 = H, R2
=
=
O
N
R2
O
N
O
O
R1
R1
R1
NH
O
N
N
R2
b: R1 = H, R2
O
Cl
R2
TBAB
CH2Cl2–Na2CO3 aq
O
N
O
O
O
Si
O
Si
Si
O
O
and
O
OMe
O
O
O
OH
Si
O
OH
O
OH
Si
Si
c: R1 = H, R2
=
d: R1 = H, R2 =
e: R1 = Me, R2
1a–e
2a–e
3a–e
heat
=
Scheme 1 Phase-transfer N3-acylation of uridine derivatives
The reactions were carried out by using the microfluidic
system shown in Figure 1. We prepared the organic solution
by dissolving compound 1 (0.05 M) and benzoyl chloride
(1.3 equiv) in dichloromethane, and the aqueous solution
by dissolving TBAB (0.01 M) in 0.2 M sodium carbonate
solution. The flow system consisted of two plunger pumps
and branch unit, linked to polyether ether ketone (PEEK)
tubing. Initially, the length of the PEEK tubing was set to
260 cm, and the total flow rate, which is the sum of the
flow rate of the organic solution and that of the aqueous
solution, was set to 2.0 mL min−1. Under these conditions,
the reaction time (i.e., the residence time) was estimated to
be 1.9 seconds. Under these conditions, we studied the re-
action yields by changing the ratio of the flow rate of the
aqueous solution and that of the organic solution (AO ratio)
from 0.6 to 5.4. The reaction yield of conversion of 2a into
3a was determined by using 1H NMR spectroscopic analysis
to compare the residual amount of starting material 1 and
product 3a by reference to well-separated 5-H signals. We
studied the effect of the AO ratio, because it has been re-
ported that this ratio is an important factor affecting the re-
action yields in microfluidic reactions.14
The results are summarized in Figure 2 (A). The highest
reaction yield of 84% was obtained when the AO ratio was
5.4. As the AO ratio was decreased from 5.4 to 3.0 and 0.6,
the reaction yield also decreased from 84 to 65, and 29%, re-
spectively. Thus, the higher the AO ratio, the higher the re-
action yield. These results are consistent with those report-
ed previously by Schouten et al.14 They reported that an in-
crease in the AO ratio represents an increase in the surface-
to-volume ratio of the organic solution when the flow pat-
tern is slug flow. In our experiments, the flow pattern is ex-
pected to be slug flow in the PEEK tubing section of the re-
actor.
We also studied the effect of the length of the PEEK tub-
ing by using tubes of different lengths (65, 130, 260, and
520 cm), with the AO ratio set to 3.0. The results are shown
in Figure 2 (B). We expected that the use of longer PEEK
tubing would lead to higher yields, because the reaction
time is equal to the residence time during which the reac-
tion mixture passes through the PEEK tubing. The reaction
yields were found to be 59 ± 5.3%, 57 ± 6.7%, 65 ± 7.4%, and
71 ± 11.6% when the PEEK tube lengths were 65, 130, 260,
and 520 cm, respectively [Figure 2 (B)]. Statistically, there
was no significant difference between these yields. Thus,
these results show that the yields are not affected by the
length of PEEK tubing in this reaction, and that it is more
important to optimize the AO ratio than to extend the reac-
tion time (i.e., the PEEK tubing length) to obtain better
yields.
To improve the yields, we inserted a porous material
into the 50-cm-long PEEK tubing. Porous materials have
been widely applied in microfluidic systems as catalysts15
or reagents.16 In this study, we applied a porous material to
expand the interfacial area. The diameter of the PEEK tub-
ing used in the above experiments was 0.25 mm. Therefore,
if microporous or nanoporous materials are inserted within
Figure 1 Microfluidic system used in this study
© Georg Thieme Verlag Stuttgart · New York — Synlett 2015, 26, 2578–2582

2580
N. Tago et al.
Letter
Synlett
Table 1 Details of HPLC Columns and Reaction Yields
Entry
Surface modification
Pore size
Yield (%)
1
2
3
4
5
6
7
–
–
65 (±7.4)
76a (±4.5)
83b (±4.5)
82b (±3.0)
83b (±4.7)
82b (±2.8)
74a (±1.9)
octadodecyl
carbamoyl
octadodecyl
silanol
10 nm
10 nm
11 nm and 1.4 μm
11 nm and 1.4 μm
2 μm
silanol
silanol
5 μm
a p value ≤ 0.05
b p value ≤ 0.01.
al) was tested. The yield in this case was 82%, which is com-
parable to that of the hydrophilic carbamoyl column (entry
3). Interestingly, this result is better than that obtained
with the 10 nm pore-sized octadodecyl column (entry 2),
even though the surface modification in both cases is hy-
drophobic. Thus, the nano-microporous column may be
more suitable for the microfluidic PT reaction. In addition,
the surface modification by silanol shown in Table 1, entry
5, does not change the effect of the nano-microporous col-
umn. Notably, when the pore size of the materials was
changed to 2 μm, the yield was comparable to that obtained
with the nano-microporous column (entry 6). However,
when the pore size was increased to 5 μm, the yield de-
creased to 74%. Although details of the mechanisms
through which the surface modifications and the various
sized pores change the yield are unclear, it is apparent that
the insertion of an HPLC guard column improves the yields.
We then examined the applicability of this microfluidic
system to other reagents such as isobutyryl chloride (i-PrCOCl),
toluoyl chloride (TolCl), and p-anisoyl chloride (AnCl). In
these reactions, the same microfluidic system detailed in
Table 1, entry 6, was used. The reactions with BzCl and Tol-
Cl gave the same yield of 80% (Table 2, entries 1 and 2). Con-
versely, AnCl (entry 3) afforded N3-anisoylated product 3c
in 69% yield, which was lower than that obtained when BzCl
and TolCl were used (entries 1 and 2). This is probably be-
cause of the relatively low reactivity of AnCl having a
strongly electron-donating methoxy group in comparison
to that of unsubstituted BzCl and TolCl, which have a less
electron-donating methyl group.18 The highest yield of 89%
was obtained when iPrCOCl was used as acylating agent
(entry 4). In addition, N3-isobutyrylated compound 3d was
generated without forming intermediate 4-O-acylated
compound 2d. It should be noted that the benzoylation of
5-methyluridine derivative 1e resulted in a low isolated
yield of 3e (25%; entry 5). This result is probably due to the
steric effect of the 5-methyl group, which prevents benzoy-
lation on the 4-O-position.
Figure 2 (A) Correlation between conversion yield and the AO ratio.
(B) Correlation between the conversion yield and length of PEEK tubing.
Error bars indicate standard deviations.
the PEEK tubing, the solutions are forced to pass through
narrower pores, resulting in a significantly increased inter-
facial area. In addition, connecting additional materials
changes the flow pattern of the reaction mixture, which
could lead to more efficient mixing.
Six types of HPLC guard column with different surface
modifications and pore sizes were chosen as the porous
materials for these experiments. Details of the columns are
shown in Table 1 and Table S1 (Supporting Information). As
shown in Table 1, entry 1, the reaction yield was 65% when
no guard column was inserted. As expected, the yield im-
proved to 76% when an octadodecyl guard column with 10
nm pores was inserted (Table 1, entry 2). When the surface
modification in the column was changed to hydrophilic car-
bamoyl groups, the yield increased further to 83% (Table 1,
entry 3). This result could be because of the nature of the
surface modification. Kitamori et al. reported that a hydro-
phobic surface allowed the organic droplets to merge with
the aqueous droplets at the interface,17 which resulted in a
decrease in the AO ratio and a lower yield.
As shown in Table 1, entry 4, the inclusion of an octado-
decyl guard column containing a mixture of nanoporous
and microporous material (i.e., a nano-microporous materi-
© Georg Thieme Verlag Stuttgart · New York — Synlett 2015, 26, 2578–2582

2581
N. Tago et al.
Letter
Synlett
Table 2 Conversion and Isolated Yields of Reagents Using the Micro-
fluidic System
Cummins, L. L.; Freier, S.; Greig, M. J.; Guinosso, C. J.; Lesnik, E.;
Manalili, S. M.; Mohan, V.; Owens, S.; Ross, B. R.; Sasmor, H.;
Wancewicz, E.; Weiler, K.; Wheeler, P. D.; Cook, P. D. J. Med.
Chem. 1996, 39, 5100. (c) Prakash, T. P.; Kawasaki, A. M.; Lesnik,
E. A.; Owens, S. R.; Manoharan, M. Org. Lett. 2003, 5, 403.
(d) Oeda, Y.; Iijima, Y.; Taguchi, H.; Ohkubo, A.; Seio, K.; Sekine,
M. Org. Lett. 2009, 11, 5582.
Entry
Reagent
Product
Isolated yield (%)
1
2
3
4
5
benzoyl chloride
toluoyl chloride
p-anisoyl chloride
isobutyryl chloride
benzoyl chloride
3a
3b
3c
3d
3e
80
80
69
89
25
(4) Markiewicz, W. T. J. Chem. Res., Miniprint 1979, 173.
(5) (a) Inoue, H.; Hayase, Y.; Imura, A.; Iwai, S.; Miura, K.; Ohtsuka,
E. Nucleic Acids Res. 1987, 15, 6131. (b) Yamada, T.; Okaniwa, N.;
Saneyoshi, H.; Ohkubo, A.; Seio, K.; Nagata, T.; Aoki, Y.; Takeda,
S.; Sekine, M. J. Org. Chem. 2011, 76, 3042. (c) Saneyoshi, H.;
Seio, K.; Sekine, M. J. Org. Chem. 2005, 70, 10453.
(6) (a) Beijer, B.; Grøtli, M.; Douglas, M. E.; Sproat, B. S. Nucleosides
Nucleotides 1994, 13, 1905. (b) Zhou, X.-X.; Chattopadhyaya, J.
Tetrahedron 1986, 42, 5149.
In conclusion, we have developed a new microfluidic
method for the N3-protection of uridine derivatives by the
PT reaction, which was accelerated by the insertion of HPLC
guard columns into the flow system.19 We found that ben-
zoylation succeeded with 83% conversion within a resi-
dence time (i.e., reaction time) of 1.9 seconds with the
guard column, whereas the yield was 65% without the col-
umn. Our data clearly show the usefulness of attaching a
microporous guard column to a microflow reaction system
for the synthesis of N3-protected uridine derivatives. This
microfluidic method may be applied to other liquid–liquid
PT reactions such as alkylation7 and β-glycosilation.20 As a
potential application, a one-pot sequential synthesis21 such
as a 2′-O-modification following base protection using this
microfluidic method is considered to be viable.
(7) Sekine, M. J. Org. Chem. 1989, 54, 2321.
(8) Yadav, G. D.; Jadhav, Y. B. J. Mol. Catal. 2003, 192, 41.
(9) Hisamoto, H.; Saito, T.; Tokeshi, M.; Hibara, A.; Kitamori, T.
Chem. Commun. 2001, 2662.
(10) (a) Guillot, P.; Colin, A. Phys. Rev. E: Stat., Nonlinear, Soft Matter
Phys. 2005, 72, 066301. (b) Zhao, Y.; Chen, G.; Yuan, Q. AIChE J.
2006, 52, 4052.
(11) (a) Taha, T.; Cui, Z. F. Chem. Eng. Sci. 2004, 59, 1181. (b) Burns, J.
R.; Ramshaw, C. Lab Chip 2001, 1, 10.
(12) Ueno, M.; Hisamoto, H.; Kitamori, T.; Kobayashi, S. Chem.
Commun. 2003, 936.
(13) Reichart, B.; Kappe, C. O.; Glasnov, T. N. Synlett 2013, 24, 2393.
(14) Jovanović, J.; Rebrov, E. V.; Nijhuis, T. A.; Hessel, V.; Schouten, J.
C. Ind. Eng. Chem. Res. 2010, 49, 2681.
(15) (a) Yamada, Y. M. A.; Watanabe, T.; Torii, K.; Uozumi, Y. Chem.
Commun. 2009, 5594. (b) Kobayashi, J.; Mori, Y.; Okamoto, K.;
Akiyama, R.; Ueno, M.; Kitamori, T.; Kobayashi, S. Science 2004,
304, 1305. (c) Fukuyama, T.; Rahman, T.; Sato, M.; Ryu, I. Synlett
2008, 151. (d) Noël, T.; Musacchio, A. J. Org. Lett. 2011, 13, 5180.
(16) Opalka, S. M.; Park, J. K.; Longstreet, A. R.; McQuade, D. T. Org.
Lett. 2013, 15, 996.
Acknowledgment
This work was supported by the Japan Society for the Promotion of
Science grant No. 21106001 and grant No. 15H01062 and grant No.
26810086.
(17) Aota, A.; Hibara, A.; Kitamori, T. Anal. Chem. 2007, 79, 3919.
(18) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96.
(19) General procedure for the preparation of organic solution:
3′,5′-O-(1,1,3,3-Tetraisopropyldisiloxane-1,3-diyl)uridine (1a;
730 mg, 1.46 mmol) and an appropriate acid chloride (BzCl,
AnCl, TolCl, iPrCOCl; 1.90 mmol) were dissolved in anhydrous
dichloromethane (30 mL).
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0035-1560264.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
General procedure for the preparation of aqueous solution:
Sodium carbonate (4.24 g, 40 mmol) and tetrabutylammonium
bromide (645 mg, 2 mmol) were dissolved in degassed H2O (200
mL).
References and Notes
(1) (a) Deleavey, G. F.; Damha, M. J. Chem. Biol. 2012, 19, 937.
(b) Sharma, V. K.; Sharma, R. K.; Singh, S. K. MedChemComm
2014, 5, 1454. (c) Ma, H.; Liu, J.; Ali, M. M.; Mahmood, M. A. I.;
Labanieh, L.; Lu, M.; Iqbal, S. M.; Zhang, Q.; Zhao, W.; Wan, Y.
Chem. Soc. Rev. 2015, 44, 1240.
(2) (a) Li, J.; Tan, S.; Kooger, R.; Zhang, C.; Zhang, Y. Chem. Soc. Rev.
2014, 43, 506. (b) Dong, H.; Lei, J.; Ding, L.; Wen, Y.; Ju, H.;
Zhang, X. Chem. Rev. 2013, 113, 6207. (c) Navani, N. K.; Li, Y.
Curr. Opin. Chem. Biol. 2006, 10, 272. (d) Yan, L.; Zhou, J.; Zheng,
Y.; Gamson, A. S.; Roembke, B. T.; Nakayama, S.; Sintim, H. O.
Mol. BioSyst. 2014, 10, 970.
Procedure for the synthesis of 3c: The flow system was con-
structed as shown in Figure 1 using a guard column of monolith
2 μm (Table 1 and Table S1, entry 6). The organic solution and
the aqueous solution were flowed for 15 min with the flow rate
of 0.43 mL min–1 and 1.5 mL min–1, respectively. The organic
solution was washed three times with saturated aqueous
sodium bicarbonate and then dried over sodium sulfate. The
resulting solution was evaporated under reduced pressure. The
residue was dissolved in 1,2-dichloroethane (3 mL) and the
solution was heated at 60 °C for 15 min. The resulting solution
was evaporated under reduced pressure and then the residue
was purified by column chromatography on silica gel (EtOAc–
hexane, 20 to 30%) to give the product (149 mg, 69%) as a white
powder. The analytical data of 3c: 1H NMR (500 MHz, CDCl3): δ
(3) (a) Geary, R. S.; Watanabe, T. A.; Truong, L.; Freier, S.; Lesnik, E.
A.; Sioufi, N. B.; Sasmor, H.; Manoharan, M.; Levin, A. A. J. Phar-
macol. Exp. Ther. 2001, 296, 890. (b) Griffey, R. H.; Monia, B. P.;
© Georg Thieme Verlag Stuttgart · New York — Synlett 2015, 26, 2578–2582

2582
N. Tago et al.
Letter
Synlett
= 7.88 (d, J = 8.5 Hz, 2 H), 7.78 (s, 1 H), 6.95 (d, J = 8.7 Hz, 2 H),
5.79 (d, J = 8.2 Hz, 1 H), 5.75 (s, 1 H), 4.36 (s, 1 H), 4.20 (d,
J = 5.0 Hz, 2 H), 4.11 (d, J = 8.8 Hz, 1 H), 4.01 (dd, J = 13.3, 2.8 Hz,
1 H), 3.87 (s, 3 H), 2.87 (s, 1 H), 1.21–0.93 (m, 27 H); 13C NMR
(126 MHz, CDCl3): δ = 167.6, 165.5, 162.4, 149.2, 139.5, 133.3,
124.3, 114.8, 102.1, 90.9, 82.3, 77.5, 77.3, 77.0, 75.4, 69.1, 60.4,
55.9, 17.7, 17.6, 17.5, 17.5, 17.3, 17.2, 17.2, 17.1, 13.6, 13.2, 13.1,
12.7; HRMS (ESI): m/z [M+Na]+ calcd for C29H44N2O9Si2Na:
643.2478; found: 643.2476. Other procedures and the analyti-
cal data are described in the Supporting Information.
(20) (a) Seela, F.; Winkeler, D. Angew. Chem. Int. Ed. 1979, 18, 536.
(b) Seela, F.; Westermann, B.; Bindig, U. J. Chem. Soc., Perkin
Trans. 1 1988, 697.
(21) Suga, S.; Yamada, D.; Yoshida, J. Chem. Lett. 2010, 39, 404.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2015, 26, 2578–2582