Evaluation of a flow-photochemistry platform for the synthesis of compact modules

Article abstract of DOI:10.1016/j.tetlet.2012.01.010

A custom-made meso-scale continuous flow-photochemistry platform was evaluated with the aim to enable the synthesis of synthetically relevant amounts (>5 g, with the option of 10-100 g) of novel compact modules, whose synthesis is based upon a photochemical transformation. The flow-photochemistry concept relies on the irradiation of thin-layers (20-90 μm) of reactant dissolved in a suitable solvent that is pumped through the single-pass photoreactor. The commercially available Ehrfeld Photoreactor XL system was equipped with standard UV lamps (emission maximum at 254 nm) which are cheap, durable and low in power consumption (8 W). The flow photochemistry system was evaluated using a known intramolecular [2+2] cycloaddition reaction investigating influence of flow rate (irradiation time), layer-thickness and reactant concentration. After a short initial optimisation phase, the system delivered in a first run 6 g of tricycle 4 which was further successfully up-scaled to 48 g, demonstrating the robustness and reliability of this flow photoreactor-platform.

Full text of DOI:10.1016/j.tetlet.2012.01.010

Tetrahedron Letters 53 (2012) 1363–1366
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Evaluation of a flow-photochemistry platform for the synthesis
of compact modules
⇑
Matthias Nettekoven , Bernd Püllmann, Rainer E. Martin, David Wechsler
F. Hoffmann-La Roche Ltd, Pharma Research & Early Development, Medicinal Chemistry, Grenzacher Str. 124, 4070 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
A custom-made meso-scale continuous flow-photochemistry platform was evaluated with the aim to
enable the synthesis of synthetically relevant amounts (>5 g, with the option of 10–100 g) of novel com-
pact modules, whose synthesis is based upon a photochemical transformation. The flow-photochemistry
Received 2 December 2011
Revised 21 December 2011
Accepted 4 January 2012
Available online 12 January 2012
concept relies on the irradiation of thin-layers (20–90 lm) of reactant dissolved in a suitable solvent that
is pumped through the single-pass photoreactor. The commercially available Ehrfeld Photoreactor XL sys-
tem was equipped with standard UV lamps (emission maximum at 254 nm) which are cheap, durable
and low in power consumption (8 W). The flow photochemistry system was evaluated using a known
intramolecular [2+2] cycloaddition reaction investigating influence of flow rate (irradiation time),
layer-thickness and reactant concentration. After a short initial optimisation phase, the system delivered
in a first run 6 g of tricycle 4 which was further successfully up-scaled to 48 g, demonstrating the robust-
ness and reliability of this flow photoreactor-platform.
Keywords:
Flow-chemistry
Photochemistry
Modules
Bicyclo[3.2.0]heptane
Ó 2012 Elsevier Ltd. All rights reserved.
Compact modules have recently received much attention with-
in the pharmaceutical community as they can serve as surrogates
for well-established medicinal chemistry scaffolds or building
blocks.1 They usually are low-molecular-weight mono-, bi-, or
fused heterocyclic scaffolds which are intended to render the prop-
erties of the new molecules in a beneficial way with respect to
physico-chemical, metabolic and safety related properties.2 In
addition, they impart versatility in view of shape diversity as well
as, due to their novelty, establishing an exclusive IP position.3
In the ongoing search for such novel topologies both modern
and traditional techniques have been used to provide a quick ac-
cess to such building blocks. Photochemically induced organic
reactions provide a powerful methodology to access complex prod-
ucts often from simple starting materials which are otherwise dif-
ficult to obtain.4 This methodology appears to be somewhat
underutilized due to some disadvantages of the traditionally used
equipment. Powerful light sources such as high-pressure mercury
lamps (or high performance LEDs) are routinely employed to irra-
diate the reaction mixtures. However, due to the rapid absorption
of light in liquids which affect photochemical reactions (irradiation
typically at: 200–400 nm) the dimension of the reaction vessels is
rather limited. This directly affects the quantity of material which
can be efficiently produced in one batch. Using strong light sources
not only exposes the substrate to the reaction conditions required
for the photochemical reaction, but also the product which itself
can undergo undesired side reactions, thereby reducing the yield
of the desired product. Further, such high-power light sources pro-
duce considerable heat, making it necessary to efficiently cool the
reaction vessel. With the advent of flow chemistry5 the combina-
tion of this technology with photochemistry promised to solve
some of those issues due to continuous removal of the product
from the irradiation zone. By accurate control of the flow rate,
the starting material solution is only irradiated as long as neces-
sary to effect the photochemical reaction, a fundamental concept
of any flow microreactor system.6
An attractive synthesis of a desired new compact module based
on a photochemical transformation7 needs to be reproducible on a
multi-gram scale (10–100 g) in order to ensure access to sufficient
quantities of the building block for further modification. In the
search for a practical and simple protocol to operate a flow-photo-
chemistry platform for large scale laboratory synthesis, a recent re-
port8 caught our attention. It comments on a single-pass large scale
synthesis (ꢀ500 g) of a heterocycle with a traditional water cooled
reactor having wrapped around UV transparent tubing through
which the reaction solution is pumped at a certain flow rate. The
light emitted from a high power UV lamp is used to penetrate
the tubing and affect the reaction.9
To avoid the use of high-power UV lamps and still embrace the
principle of a single pass reactor, the Ehrfeld Photoreactor XL10
(Fig. 1) was considered as a viable alternative. This reactor is
designed as a single pass photoreactor where starting material
dissolved in any suitable solvent is pumped through a chamber
⇑
Corresponding author. Tel.: +41 61 6886227; fax: +41 61 6886459.
E-mail address: matthias.nettekoven@roche.com (M. Nettekoven).
as a thin film (20–90 lM) at varying flow rates. The chamber is
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2012.01.010

1364
M. Nettekoven et al. / Tetrahedron Letters 53 (2012) 1363–1366
Figure 1. Schematic representation of the Ehrfeld Photoreactor XL using thin-layer technology.
Figure 2. In-house modified Ehrfeld Photoreactor XL equipped with 4 Â 254 nm 8 W 30 cm low-pressure mercury arcs.
built from a stainless steel plate and a quartz-glass which are sep-
arated from each other by a thin distance holder (20–90 M; vol-
ume of the chamber: 0.41–1.94 mL). According to Lambert–Beer
law the intensity of the UV-light to penetrate these thin films
should be considerably reduced.
Therefore, the Ehrfeld Photoreactor XL was adapted in-house
with 4 conventional 254 nm 8 W 30 cm low-pressure mercury
arcs, commonly used as TLC light bulbs11 (flexibly exchangeable
with 302 or 356 nm) placing them at a distance of 4 cm in front
of the quartz glass plate (Fig. 2), optionally covered with an alu-
minium lid. These light bulbs are in-expensive, work reliably with
the same intensity of light for multiple 1000 h time periods and are
easy to exchange for any reason (damage, use of different wave-
lengths or combination of different wave-lengths to affect orthog-
onal organic photochemistry in the same reactor).
l
To establish the versatility and reliability of this photo-flow
platform for the synthesis of new chemical modules a prototypical
intramolecular [2+2] cycloaddtion12 to form a cyclobutane deriva-
tive was chosen. The formation of bicyclo[3.2.0]heptane 4 from
cyclopentenone 3 in its’ achiral version was studied. The chiral ver-
sion of this particular reaction is described in a batch mode process
at 254 nm with 97 mg (60%) isolated yield. Our process started
with the isomerisation of furfuryl alcohol 1 to the respective cyclo-
pentenone derivative 2 under basic conditions in varying yields.13

M. Nettekoven et al. / Tetrahedron Letters 53 (2012) 1363–1366
1365
batch
flow
flow
O
O
H
H
O
Cl
O
pH 4 with HOAc
hν (254 nm)
acetone
O
240 °C
DIPEA, TBAI
DCM
OH
O
O
O
tR, nominal = 1 min
O
OH
66%
73%
1
2
3
4
Scheme 1. Synthesis of annelated cyclopentanone 4.
However, since scaling-up of these dilute reaction conditions
proved inconvenient, a slightly adapted protocol of a recently
established continuous flow procedure14 was employed, which
delivered the starting material 2 on a 100 g scale in a 66% yield.15
Reaction of 2 with 3-chloromethoxy-propene yielded the starting
material 3 for the photocycloaddition in a 73% yield (Scheme 1).16
Subsequently, the photocycloaddition at 254 nm of 3 to access
annelated cyclopentanone 4 was studied. Initially, at a layer-thick-
ness of 20 lm the influence of the concentration of the dissolved
cyclopentenone 3 in acetone in respect to varying flow rates on
the amount of product 4 formed was studied.17 A concentration
of c = 0.125% (G/V, i.e., 125 mg dissolved in 100 mL) of 3 yielded
generally the highest yield of product 4, up to 70% at a flow rate
of 4 mL/min (i.e., 6.6 s exposure of the reactant to the UV light).
Increasing the concentration up to c = 2% decreased the amount
of product formed. However, at a concentration of c = 0.5% and a
flow rate of 2–3 mL/min ꢀ60% product 4 was formed, seemingly
an optimum of amount of product 4 formed per time unit (Scheme
2). These conditions would enable the synthesis of ꢀ400 mg/h of
product 4.
Scheme 2. Amount of product 4 formed dependent on concentration of 3 in
acetone and the flow rate at 254 nm.
Subsequently, the influence of the layer-thickness on the
amount of product 4 formed was studied at different flow rates.
The optimal flow rate for a layer thickness of 20
at 2–3 mL/min producing ꢀ60% of 4. With a layer-thickness of
45
m yields of ꢀ60% were realized at a flow rate of 2–3 mL/min
as well, however, at higher flow rates yields dropped by 10–20%.
Layer-thicknesses of 70 and 90
lm was confirmed
l
lm produced ꢀ60% of product 4
with flow rates optimal around 3 mL/min as well. The product for-
mation declined steadily from 4 mL/min flow rate onwards
(Scheme 3). With these optimal conditions (3 mL/min flow rate;
0.5% acetone, 60% isolated yield) theoretically 540 mg/h of com-
pounds 4 can be produced.
With these conditions (flow rate: 3 mL/min, c = 0.5% of 3 in ace-
tone, layer-thickness: 90 lm, 254 nm) our setup to produce larger
amounts of compound 4 was studied. On three consecutive days
the reaction from 3 (13.5 g in 3075 mL acetone) to 4 was run for
6–7 h per day resulting in 17 h total reaction time with interrup-
tions at the end of day 1 and day 2. A constant production of
ꢀ60% of 4 was achieved with this set-up during day all three days
irrespective of interrupting the reaction. The reaction course was
monitored with GC, thus being un-precise regarding any undetect-
able side-products and as a consequence the yield of the reaction.
Therefore, the collected acetone solutions from these 3 days were
evaporated and the residue was purified by silica gel chromatogra-
phy yielding 6.4 g (38 mmol, 48%) of the tricycle 4.18
Scheme 3. Amount of product 4 formed dependent on various layer-thicknesses
and flow-rates of 3 dissolved in acetone at a concentration of c = 0.5% and 254 nm.
are envisaged to access new chemical modules stretching out into
new pharmacological space. They will serve as new replacements
for known chemical entities with rendered properties for medici-
nal chemists to design and synthesize optimized new compounds.
To finally demonstrate the capacity for meso-scale synthesis
and long-term use of this flow-photochemistry platform 100 g of
starting material 3 in acetone (c: 0.5%) were processed during
Acknowledgments
111 h at a flow rate of 3 mL/min and a layer-thickness of 90
l
m
For the adaptation of the photoreactor we are grateful to Thomas
Zumstein. We thank Adam Schweizer and Falk Morawitz for the syn-
thesis of gram quantities of starting material 2. For discussions, and
proof-reading we thank Drs. Henner Knust, Mark Rogers-Evans, Tan-
ja Schulz-Gasch, and Thomas Woltering. Thanks to Drs. Carmine Raf-
fa and Frank Schael from Ehrfeld for initial support.
to yield after evaporation of the acetone solution and silica-gel
chromatography 48 g (288 mmol, 48%) of the desired product
4.19 This result clearly shows the robustness and reliability of the
set-up chosen for this flow-photochemistry platform. To further
validate this system a variety of photochemical transformations

1366
M. Nettekoven et al. / Tetrahedron Letters 53 (2012) 1363–1366
12. Le Liepvre, M.; Ollivier, J.; Aitken, D. J. Eur. J. Org. Chem. 2009, 5953–5962.
References and notes
13. Minai, M. Jpn. Kokai Tokkyo Koho, 57062236, 1982.
14. Ulbrich, K.; Kreitmeier, P.; Reiser, O. Synlett 2010, 2037–2040.
1. (a) Wuitschik, G.; Rogers-Evans, M.; Müller, K.; Fischer, H.; Wagner, B.; Schuler,
F.; Polonchuk, L.; Carreira, E. M. Angew. Chem., Int. Ed. 2006, 45, 7736–7739; (b)
Burkhard, J. A.; Wuitschik, G.; Rogers-Evans, M.; Müller, K.; Carreira, E. M.
Angew. Chem., Int. Ed. 2010, 49, 9052–9067.
15. Synthesis of 2: FlowSyn (Uniqsis) connected to
a HP 6890 series gas
chromatography (GC) oven equipped with a 5 mL home-made reactor coil
using Supelco 304 stainless steel tubing (length  ID = 1.45 m  2.1 mm) and a
750 psi back-pressure regulator (usage of a reactor coil with larger inner
diameter proofed to be more favorable with respect to potential blockages than
commercial available standard reactors with ID = 1 mm). Flow stream A (25 g
2. (a) Wuitschik, G.; Rogers-Evans, M.; Buckl, A.; Bernasconi, M.; Märki, M.; Godel,
T.; Fischer, H.; Wagner, B.; Parrilla, I.; Schneider, J.; Alker, A.; Schweizer, W. B.;
Müller, K.; Carreira, E. M. Angew. Chem., Int. Ed. 2008, 47, 4512–4515; (b)
Burckhard, J. A.; Guérot, C.; Knust, H.; Rogers-Evans, M.; Carreira, E. M. Org. Lett.
2010, 12, 1944–1947; (c) Burkhard, J. A.; Wagner, B.; Fischer, H.; Schuler, F.;
Müller, K.; Carreira, E. M. Angew. Chem., Int. Ed. 2010, 49, 3524–3527; (d)
Wuitschik, G.; Carreira, E. M.; Wagner, B.; Fischer, H.; Parrilla, I.; Schuler, F.;
Rogers-Evans, M.; Müller, K. J. Med. Chem. 2010, 53, 3227–3246.
3. Nisius, B.; Bajorath, J. Expert Opin. Drug Discov. 2011, 6, 1–7.
4. (a) Ralph, M. L. Curr. Org. Chem. 2011, 15, 2658–2672; (b) Bach, T.; Hehn, J. P.
Angew. Chem., Int. Ed. 2011, 50, 1000–1045.
5. (a) Wegner, J.; Ceylan, S.; Kirschning, A. Chem. Commun. 2011, 4583–4592; (b)
Baumann, M.; Baxendale, I. R.; Ley, S. V. Mol. Divers. 2011, 613–630.
6. (a) Sugimoto, A.; Fukuyama, T.; Sumino, Y.; Takagi, M.; Ryu, I. Tetrahedron
2009, 1593–1598; (b) Tsutsumi, K.; Terao, K.; Yamaguchi, H.; Yoshimura, S.;
Morimoto, T.; Kakiuchi, K.; Ryu, I. Chem. Lett. 2010, 828–829; (c) Vasudevan,
A.; Villamil, C.; Trumbull, J.; Olson, J.; Sutherland, D.; Pan, J.; Djuric, S.
Tetrahedron Lett. 2010, 51, 4007–4009; (d) Levesque, F.; Seeberger, P. H. Org.
Lett. 2011, 13, 5008–5011; (e) Bou-Hamdan, F. R.; Levesque, F.; O’Brien, A.
G.; Seeberger, P. H. Beilstein J. Org. Chem. 2011, 7, 1124–1129; (f) Fukuyama,
T.; Kajihara, Y.; Hino, Y.; Ryu, I. J. Flow Chem. 2011, 1, 40–45; (g) Pimparkar,
K.; Yen, B.; Goodell, J. R.; Martin, V. I.; Lee, W.-H.; Porco, J. A., Jr.; Beeler, A.
B.; Jensen, K. F. J. Flow Chem. 2011. doi:10.1556/jfchem.2011.00006.
7. (a) [2+2] cycloaddition: Mann, G.; Muchall, H. M. In Methoden Org. Chem.
(Houben-Weyl), 1997; pp 667–684.; (b) Hehn, J. P.; Müller, C.; Bach, T.
Handbook of Synthetic Photochemistry In Albini, A., Fagnoni, M., Eds.; Wiley-
VCH: Weinheim, 2010; pp 171–215.
furfuryl alcohol 1/1000 mL water at pH 4): 4.5 mL/min; flow stream
B
(toluene): 0.5 mL/min. Nominal residence time (tR) = 1 min. Isolated yield
after extraction and distillation of 2: 66%. Please note: due to thermal solvent
expansion at these subcritical conditions, the effective residence time is
smaller than the nominally calculated value of 1 min (see: Martin, R. E.;
Morawitz, F.; Kuratli, C.; Alker, A. M.; Alanine, A. I. J. Eur. Org. Chem. 2012, 1,
47–52.).
16. Synthesis of 3: In a 1 L round bottomed flask, 16 g (163 mmol) alcohol 2 was
dissolved in methylene chloride (160 mL) under argon. DIPEA (126 g,
979 mmol), tetrabutylammonium iodide (6.02 g, 16.3 mmol) and 3-
(chloromethoxy)prop-1-ene (50.5 g, 474 mmol) was added at 0 °C. The
mixture was stirred for 60 h, washed with 2 Â 150 mL HCl, 150 mL of
saturated NaHCO3 and 150 mL brine. The organic phase was dried over
MgSO4, filtered and concentrated. The residue was purified by column
chromatography on silica eluting with a gradient formed from heptane and
ethyl acetate to yield after evaporation of the product containing fractions
19.9 g (73%) of the cyclopentenone 3.
17. All analyses to the amount of product formed (with all starting material
consumed) was measured by GC in duplicates and the means are reported. The
yields are given in % AUC in correlation to total % AUC.
18. Synthesis of 4: A solution of 4-(allyloxymethoxy)cyclopent-2-enone (13.5 g,
80 mmol) in 3.075 L acetone was pumped at a constant flow rate of 3 mL/min
through the Ehrfeld XL Photoreactor. The thin film of 90 lm was irradiated
with 4 Â 8 W light bulbs at 254 nm. The collected acetone solution was
evaporated and the residue was purified by column chromatography on silica
eluting with a gradient formed from hexane and ethyl acetate. Evaporation of
the product containing fractions yielded 6.38 g of the tricycle 4. NMR data are
in accordance with published data.
8. Hook, B. D. A.; Dohle, W.; Hirst, P. R.; Pickworth, M.; Berry, M. B.; Booker-
Milburn, K. I. J. Org. Chem. 2005, 70, 7558–7564.
9. A similar apparatus was built in-house and evaluated, however not found to be
suitable.
10. http://www.ehrfeld.com; Ehrfeld company suggests the potential combination
of the Photoreactor XL with high-power LED light sources. Figure 1: courtesy of
Ehrfeld Company.
11. http://www.uvp.com; 34-0006-01 8-Watt replacement tube, 365 nm; 34-
0042-01 8-Watt replacement tube, 302 nm; 34-0007-01 8-Watt replacement
tube, 254 nm.
19. It proved advantageous for maintaining the yield of the reaction to pump
methanol (15 min, 3 mL/min) through the reaction chamber about every 24 h.
This cleaned the chamber from any polymeric side-product which might have
precipitated and affected the transformation.