Parallel microflow photochemistry: Process optimization, scale-up, and library synthesis

Article abstract of DOI:10.1021/ol301773r

A novel, multimicrocapillary flow reactor (MμCFR) was constructed and applied to a series of sensitized photoadditions involving 2(5H)-furanones. The reactor allowed for rapid and energy-, time-, and space-efficient sensitizer screening, process optimizat

Full text of DOI:10.1021/ol301773r

ORGANIC
LETTERS
2012
Vol. 14, No. 17
4342–4345
Parallel Microflow Photochemistry:
Process Optimization, Scale-up,
and Library Synthesis
Alexander Yavorskyy,† Oksana Shvydkiv,† Norbert Hoffmann,‡ Kieran Nolan,† and
,§
€
Michael Oelgemoller*
Dublin City University, School of Chemical Sciences, Dublin 9, Ireland, UMR 7312
ꢀ
ꢀ
CNRS et Universite de Reims Champagne-Ardenne, Institut de Chimie Moleculaire de
Reims, UFR Sciences, B.P. 1039, 51687, Reims, Cedex 02, France, and James Cook
University, School of Pharmacy and Molecular Sciences, CBMDT, Townsville,
Queensland 4811, Australia
michael.oelgemoeller@jcu.edu.au
Received June 28, 2012
ABSTRACT
A novel, multimicrocapillary flow reactor (MμCFR) was constructed and applied to a series of sensitized photoadditions involving 2(5H)-
furanones. The reactor allowed for rapid and energy-, time-, and space-efficient sensitizer screening, process optimization, validation, scale-up,
and library synthesis.
Microflow reactors have recently emerged as a new
technology in chemical synthesis and have seen a growing
number of applications.1 The small inner dimensions of
these devices, in combination with their continuous flow
operation, make them especially attractive for photochem-
ical studies.2 In particular, the narrow reaction channels
enable extensive penetration by light, even at high chromo-
phore concentrations (as dictated by the BeerꢀLambert law).
Also, the small dimensions allow for precise temperature
control and thus superior regio- and stereoselectivity.3
One of the major drawbacks of current microflow photo-
reactors is the need to perform individual reactions
separately in-series. Automated reactor systems have
been constructed but do not allow for parallel operation.4
Likewise, small libraries were generated using segment
flow, but this in-series synthesis does not reduce operation
times significantly.5 Clustering (‘numbering-up’) of reac-
tors, as done successfully by Heraeus for the synthesis of
† Dublin City University.
(3) (a) Terao, K.; Nishiyama, Y.; Aida, S.; Tanimoto, H.; Morimoto,
T.; Kakiuchi, K. J. Photochem. Photobiol. A: Chem. 2012, 242, 13.
(b) Neumann, M.; Zeitler, K. Org. Lett. 2012, 14, 2658. (c) Tsutsumi, K.;
Terao, K.; Yamaguchi, H.; Yoshimura, S.; Morimoto, T.; Kakiuchi, K.;
Fukuyama, T.; Ryu, I. Chem. Lett. 2010, 39, 828. (d) Mukae, H.; Maeda,
H.; Nashihara, S.; Mizuno, K. Bull. Chem. Soc. Jpn. 2007, 80, 1157.
(e) Sakeda, K.; Wakabayashi, K.; Matsushita, Y.; Ichimura, T.; Suzuki, T.;
Wada, T.; Inoue, Y. J. Photochem. Photobiol. A: Chem. 2007, 192, 166.
(4) (a) Vasudevan, A.; Villamil, C.; Trumbull, J.; Olson, J.; Sutherland,
D.; Pan, J.; Djuric, S. Tetrahedron Lett. 2010, 51, 4007. (b) Sugimoto, A.
Fukuyama, T.; Sumino, Y.; Takagi, M.; Ryu, I. Tetrahedron 2009, 65,
1593.
‡
ꢀ
CNRS et Universite de Reims Champagne-Ardenne.
§ James Cook University.
(1) For some recent reviews, see: (a) Malet-Sanz, L; Susanne, F.
J. Med. Chem. 2012, 55, 4062. (b) Wegner, J.; Ceylan, S.; Kirschning, A.
Adv. Synth. Catal. 2012, 354, 17. (c) Wiles, C.; Watts, P. Green Chem.
2012, 14, 38. (d) Houlton, S. Manuf. Chem. 2012, 83, 41. (e) Baraldi,
P. T.; Hessel, V. Green Process Synth. 2012, 1, 149.
€
(2) (a) Oelgemoller, M. Chem. Eng. Technol. 2012, 35, 1144.
€
(b) Oelgemoller, M.; Shvydkiv, O. Molecules 2011, 16, 7522. (c) Coyle,
€
E. E.; Oelgemoller, M. Photochem. Photobiol. Sci. 2008, 7, 1313.
(d) Matsushita, Y.; Ichimura, T.; Ohba, N.; Kumada, S.; Sakeda, K.;
Suzuki, T.; Tanibata, H.; Murata, T. Pure Appl. Chem. 2007, 79, 1959.
(e) Shvydkiv, O.; Oelgemo.ller, M. In CRC Handbook of Organic Photo-
chemistry and Photobiology, 3rd ed.; Griesbeck, A. G.; Oelgemo.ller, M.;
Ghetti, F., Eds.; CRC Press: Boca Raton, FL, 2012; Ch. 3, Vol. 1, pp. 49ꢀ72.
(5) (a) Thompson, C. M.; Poole, J. L.; Cross, J. L.; Akritopoulou-
Zanze, I.; Djuric, S. W. Molecules 2011, 16, 9161. (b) Goodell, J. R.;
McMullen, J. P.; Zaborenko, N.; Maloney, J. R.; Ho, C.-X.; Jensen,
K. F.; Porco, J. A., Jr.; Beeler, A. B. J. Org. Chem. 2009, 74, 6169.
r
10.1021/ol301773r
Published on Web 08/20/2012
2012 American Chemical Society

anticancer drug precursors, causes significant installation
costs.6 Flexible polymer-based microcapillaries have been
increasingly used for the construction of continuous
flow reactors.7 Their reaction capillaries are chemically
inert, UV-transparent, inexpensive and can be almost any
length. Using this general reactor concept, we have pre-
viously described a dual-microcapillary system for photo-
chemical transformations in duplicates.8 To further
improve the utility for typical R&D processes,9 we
have constructed a novel multimicrocapillary flow reactor
(MμCFR, Figure 1).10
surface-to-volume ratio per capillary was calculated to be
2514 m2/m3. The nonexposed ends (ca. 75 cm each) were
connected to the syringe pump (influent) and an array of
round-bottom flasks (effluent), which were protected from
light during irradiations. UVA fluorescent tubes (λmax
=
365 nm; 2 ꢁ 18 W; height: 60 cm) were placed in the center of
the glass columns, and small cooling fans were mounted in
their bases. The entire reactor system was kept behind a light-
tight curtain during operation. To avoid cross-irradiation, a
black cardboard screen was placed in between the two parallel
microcapillary towers. The MμCFR system was subsequently
utilized to investigate sensitized additions of alcohols to
furanones.11 This transformation is well understood and has
been used previously as a model reaction for microreactor
evaluations.12 Three typical R&D scenarios were investigated:
• Process optimization using 10 different reaction
conditions;
• Process validation and scale-up using 10 identical
conditions and
• Library synthesis using 10 different reagent mixtures.
Using the addition of isopropanol 2a to the parent
2(5H)-furanone 1a as a representative example, the micro-
capillary reactor was first applied to sensitizer screening
(Scheme 1). The original reaction protocol utilized acetone
as the sensitizer,11a which could not be used due to the poor
overlap of its absorption spectrum with the emission
maximum of the UVA fluorescent tube.13 Hence, a range
of aromatic ketones that are typically employed as sensitizers
were screened (Table 1). Previously degassed solutions of 1a
and the sensitizer (except for the blank experiment) in
isopropanol were pumped through the microcapillaries at a
constant flow rate of 1 mL/min, thus giving an irradiation
time of 5 min. The conversion rates were subsequently
determined by 1H NMR spectroscopic analysis of the crude
reaction mixture. The highest conversion of 72% was ob-
served when 4,40-dimethoxybenzophenone (DMBP) was
used as the sensitizer. In contrast, xanthone, 4-benzoylben-
zoic acid, and benzophenone gave moderate conversion rates
of 30ꢀ47%. Consumption of 1a remained low with 4% for
acetophenone, while all other sensitizers failed to induce any
photoreactivity under the chosen conditions. Partial photo-
reduction was observed for benzophenone, DMBP, and
xanthone.14 No reaction was observed in the absence of
sensitizer, and 1a was recovered quantitatively.15,16
Figure 1. Multimicrocapillary flow reactor (MμCFR): (a) col-
lection flasks; (b) FEP microcapillaries; (c) 10-syringe pump.
To realize a practical number of experiments in parallel,
a 10-syringe pump was selected as the delivery system. Two
bundles of five fluorinated ethylene propylene copolymer
(FEP; outer/inner diameter: 1.6/0.8 mm) capillaries were
wrapped tightly around two Pyrex glass columns (λ g
300 nm; height: 60 cm; outer diameter: 6 cm; thickness:
2.2 mm). Each microtube had a total length of 11.5 m. Of
these, 10 m covered the glass body therefore creating an
irradiated volume of 5 mL inside each capillary. Assuming
that only half of each capillary is irradiated, the effective
(6) Werner, S.; Seliger, R.; Rauter, H.; Wissmann, F. EP-2065387A2;
Chem. Abstr. 2009, 150, 376721.
ꢀ
(7) (a) Levesque, F.; Seeberger, P. H. Angew. Chem., Int. Ed. 2012, 51,
Scheme 1. Sensitized Addition of Isopropanol to Furanone 1a
ꢀ
1706. (b) Levesque, F.; Seeberger, P. H. Org. Lett. 2011, 13, 5008.
(c) Gutierrez, A. C.; Jamison, T. F. Org. Lett. 2011, 13, 6414. (d) Hook,
B. D. A.; Dohle, W.; Hirst, P. R.; Pickworth, M.; Berry, M. B.; Booker-
Milburn, K. J. Org. Chem. 2005, 70, 7558.
(8) (a) Shvydkiv, O.; Yavorskyy, A.; Tan, S. B.; Nolan, K.;
€
Hoffmann, N.; Youssef, A.; Oelgemoller, M. Photochem. Photobiol.
Sci. 2011, 10, 1399. (b) Yavorskyy, A.; Shvydkiv, O.; Nolan, K.;
€
Hoffmann, N.; Oelgemoller, M. Tetrahedron Lett. 2010, 52, 278.
(9) (a) Chin, P.; Barney, W. S.; Pindzola, B. A. Curr. Opin. Drug
Discovery Dev. 2009, 12, 848. (b) Rubin, A. E.; Tummala, S.; Both,
D. A.; Wang, C. C.; Delaney, E. J. Chem. Rev. 2006, 106, 2794.
(10) For thermal reactions, see: (a) Hornung, C. H.; Hallmark, B.;
Baumann, M.; Baxendale, I. R.; Ley, S. V.; Hester, P.; Clayton, P.;
Mackley, M. R. Ind. Eng. Chem. Res. 2010, 49, 4576. (b) Comer, E.;
Organ, M. G. Chem.;Eur. J. 2005, 11, 7223.
With DMBP as the best sensitizer, its concentration was
optimized next. The furanone/isopropanol (1a/2a) pair
was again chosen as a model system (Scheme 2). The
conversion to 3a increased steadily with increasing
Org. Lett., Vol. 14, No. 17, 2012
4343

The light penetration profiles at 365 nm were subse-
quently calculated from the adsorption spectra and the
experimental conditions and were compared to the inner
diameter of the FEP microcapillary of 0.8 mm (Figure 2).18
For all concentrations studied, complete transmission
through the microcapillary was achieved. In contrast, batch
systems typically have much larger path lengths (g1 cm)
and thus show significantly lower transmission efficiencies.8
Table 1. Sensitizer Screeninga
conv of
conv of
sensitizer
(none)
1a (%)b
sensitizer
xanthone
1a (%)b
0c
47
4
0c
benzophenone
4,40-dimethoxy-
benzophenone
4-benzoylbenzoic
acid
30
72
acetophenone
4-tert-butyl-
acetophenone
4-methoxy-
37
0c
0c
0c
acetophenone
4-(dimethylamino)-
benzaldehyde
4,40-bis(dimethyl-
amino)benzophenone
a Conditions: [1a] = 33.3 mM; [sens] = 6.7 mM; Vol = 15 mL; flow
rate: 1 mL/min; residence time = 5 min. b Conversion determined by 1H
NMR spectroscopy ((3%). c No reaction (recovery of 1a: >85%).
amounts of sensitizer until a maximum of 90% was
achieved at a DMBP concentration of 10 mM (Table 2).
Above this critical value, DMBP started to precipitate
within the reactor. As a result, the conversion dropped to
73% due to scattering from the solid particles and the
decrease of sensitizer dissolved in solution.17
Figure 2. Light-penetration profiles of DMBP solutions at 365 nm.
The vertical broken line represents the inner diameter of the micro-
capillary (0.8 mm).
Scheme 2. DMBP-Sensitized Addition to Furanone 1a
The influence of the furanone (1a) concentration on
the conversion rate was likewise investigated. Applying
a fixed standard concentration of DMBP (10 mM)
in isopropanol, [1a] was varied stepwise from 33.3 to
200 mM. A standard irradiation time of 5 min was set
for this experimental run. As would be expected, the
consumption of 1a dropped significantly from 80% to
2% with increasing concentrations (Table 3). An acceptable
Table 2. DMBP Concentration Studya
[DMBP]
(mM)
conv of
[DMBP]
(mM)
conv of
1a (%)b
1a (%)b
Table 3. Furanone (1a) Concentration Studya
0.0
1.0
2.0
4.0
6.0
0c
7.0
76
81
86
90
73
26
40
55
70
8.0
[1a]
conv of
1a (%)b
[1a]
conv of
1a (%)b
9.0
(mM)
(mM)
10.0
11.0d
33.3
50.0
66.7
83.3
100.0
80
51
34
23
14
116.7
133.3
150.0
166.7
200.0
9
7
5
3
2
a Conditions: [1a] = 33.3 mM; Vol = 15 mL; flow rate: 1 mL/min;
residence time = 5 min. b Conversion determined by 1H NMR spectro-
scopy ((3%). c No reaction (recovery of 1a: >85%). d Precipitation of
DMBP within microreactor creating light scattering.
a Conditions: [DMBP] = 10 mM; Vol = 15 mL; flow rate: 1 mL/min;
residence time = 5 min. b Conversion determined by 1H NMR spectro-
scopy ((3%).
(11) (a) Hoffmann, N. Tetrahedron: Asymmetry 1994, 5, 879.
(b) Ohga, K.; Matsuo, T. J. Org. Chem. 1974, 39, 106. (c) Mann, J.;
Weymouth-Wilson, A. Synlett 1992, 67. (d) For a review, see: Hashem,
A. I.; Senning, A.; Hamad, A.-S. S. Org. Prep. Proc. Int. 1998, 30, 401.
(12) (a) Shvydkiv, O.; Yavorskyy, A.; Nolan, K.; Youssef, A.;
(15) Dimerization of 1a was also not observed: Ohga, K.; Matsuo, T.
Bull. Chem. Soc. Jpn. 1970, 43, 3505.
€
Riguet, E.; Hoffmann, N.; Oelgemoller, M. Photochem. Photobiol. Sci.
(16) The results parallel those for analogue additions of tertiary
amines to furanones; see: (a) Bertrand, S.; Hoffmann, N.; Humbel, S.;
Pete, J. P. J. Org. Chem. 2000, 65, 8690. (b) Bertrand, S.; Hoffmann, N.;
Pete, J. P. Eur. J. Org. Chem. 2000, 2227.
2010, 9, 1601. (b) van den Broek, S. A. M. W.; Becker, R.; Koch, K.;
Nieuwland, P. J. Micromachines 2012, 3, 244.
(13) Montali, M.; Credi, A.; Prodi, L.; Gandolfi, M. T. Handbook of
Photochemistry, 3rd ed.; CRC Press: Boca Raton, FL, 2006.
(17) A DMBP concentration of 6.7 mM was considered ‘safe’, and
sensitizer precipitation was never observed.
(18) Braun, A. M.; Maurette, M.; Oliveros, E. Photochemical Tech-
nology; Wiley: Chichester, 1991.
(14) (a) Li, J.-T.; Yang, J.-H.; Han, J.-F.; Li, T.-S. Green Chem. 2003,
5, 433. (b) Pitts, N. J., Jr.; Letsinger, R. L.; Taylor, R. P.; Patterson,
J. M.; Recktenwald, G.; Martin, R. B. J. Am. Chem. Soc. 1959, 81, 1068.
4344
Org. Lett., Vol. 14, No. 17, 2012

conversion of >20% was nevertheless maintained up to
[1a] = 83.3 mM.
Table 4. Experimental Details of Library Synthesisa
The photoaddition of isopropanol to 1a was again
selected for a validation and scale-up study. A solution
of 1a (33.3 mM) and DMBP (6.7 mM) in isopropanol was
distributed over the 10 syringes and irradiated using a
residence time of 5 min. The crude products were analyzed
R
R0
time [min]
yield (%)b
H
CH3
10
10
10
20
20
20
20
20
20
94 (3a)
60 (3b)
90 (3c)
80 (4a)
57 (4b)
61 (4c)
71 (5a)
73 (5b)
89 (5c)
OEtc
OMentd
H
CH3
CH3
1
by H NMR spectroscopy, and conversions of 66ꢀ75%
C2H5
OEtc
OMentd
H
OEtc
OMentd
C2H5
were achieved. The reaction showed good reproducibility
with an average conversion and a relative standard devia-
tion (RSD) of 70.8 ( 4.2%. In a separate run, the residence
time was increased to 10 min to achieve complete conver-
sion. The product mixtures from all runs were combined,
and 3a was isolated in a reasonable quantity (ca. 0.5 g) and
an excellent yield of 94%.
C2H5
-(CH2)4-
-(CH2)4-
-(CH2)4-
a Conditions: [1] = 33.3 mM; [DMBP] = 6.7 mM; Vol = 15 mL;
flow rate: 0.5 and 0.25 mL/min; residence time = 10 and 20 min.
b Isolated yields after column chromatography. c As racemate. d (ꢀ)-
OMent.
The multicapillary reactor was likewise applied to
the synthesis of a small 3 ꢁ 3 product library (Scheme 3;
Table 4). Furanone (1a) and its two 5-substituted derivatives
1b (rac-OEt) and 1c ((ꢀ)-OMent) were selected as model
compounds. Isopropanol (2a), 3-pentanol (2b), and cyclo-
pentanol (2c) were chosen as representative alcohols. The
residence times were increased to 10 min (2a) and 20 min
(2b and 2c) to achieve high conversions and thus isolated
yields. The photoaddition products 3/4/5aꢀc were ob-
tained in good to excellent yields of 57ꢀ94% after column
chromatography. Due to the complete consumption of
1aꢀc in most runs, small amounts of photoreduction and
photopinacolizationproducts ofthesensitizerDMBPwere
In conclusion, we have constructed a simple multimi-
crocapillary flow reactor that allowed for space-, time-,
and resource-efficient process optimization and library-
synthesis. As would be expected, the reactor offered
significant operation time savings compared to in-series
operations with a single-capillary reactor.8,12a The energy
consumption for the synthesis of 1 kg of 3a in the MμCFR
setup was furthermore compared to a conventional cham-
ber reactor (equipped with 16 ꢁ 8 W fluorescent tubes).19
The microreactor consumed ∼30% less energy than the
batch reactor and did not require any cooling water. These
features, together with the small reaction scales and the
possibility of using higher concentrations, make microflow
photochemistry a resource-efficient and green technology.20
A current disadvantage is the usage of a single multisyringe
pump with identical flow rates for all capillaries. Automated
pump systems with individual flow rate settings would easily
overcome this drawback. It is thus hoped that this advanced
microflow technology presented will be rapidly implemented
into chemical R&D processes as a parallel photochemical
synthesis tool.1a,9
1
detected in the crude products by H NMR analysis.14 In
contrast, the corresponding batch reactions performed in a
Pyrex test tube (inner diameter: 0.9 cm) and using a Rayonet
chamber reactor (16 ꢁ 8 W) required prolonged irradiation
times of up to 1 h to reach complete conversions.8a
Scheme 3. Library Synthesis
Acknowledgment. This work was financially supported
by the Environmental Protection Agency (EPA, 2008-ET-
MS-2-S2) and the Department of Environment, Heritage
and Local Government (DEHLG, 2008-S-ET-2). The
authors thank Miss Adeline Fon for technical assistance.
Compared to the original protocol involving acetone
as the sensitizer,11a DMBP generally gave higher con-
versions and yields. Likewise, the addition of DMBP
was never observed, probably due to the stability of its
ketyl radical. When acetone is employed, its undesired
addition to furanones (to yield 3aꢀc) is commonly
found.11a
Supporting Information Available. Experimental pro-
cedures and NMR spectra of all photoproducts. This
material is available free of charge via the Internet at
http://pubs.acs.org.
€
(20) (a) Hoffmann, N. ChemSusChem 2012, 5, 352. (b) Oelgemoller,
(19) (a) Yavorskyy, A.; Shvydkiv, O.; Limburg, C.; Nolan, K.;
ꢀ
M.; Jung, C.; Mattay, J. Pure Appl. Chem. 2007, 79, 1939. (c) Albini, A.;
Fagnoni, M.; Mella, M. Pure Pure Appl. Chem. 2000, 72, 1321.
€
Delaure, Y. M. C.; Oelgemoller, M. Green Chem. 2012, 14, 888.
(b) Protti, S.; Ravelli, D.; Fagnoni, M.; Albini, A. Chem. Commun.
€
2009, 7351. (c) Haggiage, E.; Coyle, E. E.; Joyce, K.; Oelgemoller, M.
Green Chem. 2009, 11, 318.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 17, 2012
4345