A Microflow Electrolysis Cell for Laboratory Synthesis on the Multigram Scale

Article abstract of DOI:10.1021/acs.oprd.5b00260

A large microflow electrolysis cell for laboratory synthesis on a multigram scale is described. It is based on two circular electrodes with a diameter of 149 mm and a spiral electrolyte flow channel 2000 mm long, 5 mm wide, and 0.5 mm interelectrode gap. Using the methoxylation of N-formylpyrrolidine as a model reaction, it is demonstrated that the cell approaches 100% conversion in a single pass, and it is possible to achieve a reaction selectivity >95% and a product formation rate of >20 g h^-1.

Full text of DOI:10.1021/acs.oprd.5b00260

Communication
pubs.acs.org/OPRD
A Microflow Electrolysis Cell for Laboratory Synthesis on the
Multigram Scale
Robert A. Green,* Richard C. D. Brown, and Derek Pletcher
Department of Chemistry, University of Southampton, Southampton SO17 1BJ, U.K.
Bashir Harji
Cambridge Reactor Design Ltd., Unit D2, Brookfield Business Centre, Cottenham, Cambridgeshire CB24 8PS, U.K.
S
* Supporting Information
ABSTRACT: A large microflow electrolysis cell for
laboratory synthesis on a multigram scale is described. It
is based on two circular electrodes with a diameter of 149
mm and a spiral electrolyte flow channel 2000 mm long, 5
mm wide, and 0.5 mm interelectrode gap. Using the
methoxylation of N-formylpyrrolidine as a model reaction,
it is demonstrated that the cell approaches 100%
conversion in a single pass, and it is possible to achieve
a reaction selectivity >95% and a product formation rate of
>20 g h−1.
1. INTRODUCTION
Usually, laboratory electrolysis cells for synthesis achieve
chemical conversion at a very slow rate (for example, a full
conversion may take many hours in a beaker cell), while
industrial flow cells generally operate with a low conversion per
pass of reactant through the cell and hence require extensive
recycling of reactant solutions.1,2 During the past 15 years,
there has been a growing interest in the application of
microflow chemistry to organic synthesis in the laboratory.3,4
The introduction of electrolysis into flow synthesis required the
design of systems capable of a high fractional conversion in a
single pass. This has led to significant innovation in electrolysis
cell design that has been reviewed recently.5,6 Commonly,
however, the high fractional conversion of reactant to product
in a single pass has been achieved by restricting the electrolyte
flow to very slow rates;7 this allows novel and interesting
syntheses but limits severely the amount of product that can be
formed. Recently, we have demonstrated another approach to
the design of microflow electrolysis cells where the electrolyte
channel has an extended length.6,8,9 When operated under
appropriate conditions, these cells can combine a high
fractional conversion and reaction selectivity with a product
formation rate of multiple grams per hour.
Figure 1. (a) Components of the microflow electrolysis cell. 1. Central
bolt. 2. Washer. 3. Insulating tube. 4. Peripheral bolt. 5. Perspex top
plate. 6. Cu backing plate. 7. Carbon/polymer anode plate. 8.
Perfluoroelastomer gasket. 9. Insulating tube around central bolt. 10.
Stainless steel cathode plate with spiral groove. 11. Al base plate. (b)
Photograph of reactor with perspex top. (c) Photograph of reactor
with gasket fitted into cathode plate to create the spiral channel.
formylpyrrolidine (1) to give 2 as the test reaction in MeOH at
a carbon/polymer anode (Scheme 1).10,11 The cathode reaction
was the reduction of MeOH to hydrogen and methoxide, so
that the overall chemical change is formally a dehydrogenative
coupling.
2. RESULTS
The electrolysis cell was based on two circular electrodes,
diameter 149 mm, with the spiral electrolyte channel between;
see Figure 1. The spiral channel design allows an extended
channel length within a small device and maximizes the active
area of the electrode plate. In addition it avoids corners that
disrupt and modify the electrolyte flow regime. In fact, the
channel was 2000 mm in length and 5 mm in width giving an
active electrode area of 100 cm2. The machining of the spacer
groove and the thickness of the spacer control the
interelectrode gap, which was 0.5 mm for most of this work.
Here we describe a cell design intended to allow the
synthesis of larger amounts of product without loss of reaction
selectivity or fractional conversion in a single pass of the
solution through the cell. As shown in Figure 1, the cell has a
parallel plate configuration with circular electrodes but the
interelectrode gap is divided up into a spiral microchannel by a
polymer spacer. As in previous papers,6,8,9 the performance of
the cell was demonstrated using the methoxylation of N-
Received: August 12, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.5b00260
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
1 passing through the cell/hour). The productivities with the
0.1 M reactant solution were several g h−1, similar to those
reported with the earlier cell, but these rather high
productivities were achieved in a different way.6,9 In the
current work they resulted from the high flow rate of reactant
solution, whereas in the earlier work higher productivities were
achieved by increasing the N-formylpyrrolidine concentration
to 0.75 M. In the larger cell, increased reactant concentration
can be used to increase the rate of product formation further,
and it was not expected to influence significantly the fractional
conversion. This was confirmed by electrolysis of a 0.2 M
solution of 1. Using a flow rate of 16 mL min−1 and a cell
current of 12 A, the conversion was 88%, and the productivity
was increased to 20.7 g h−1 (equivalent to 0.5 kg or 5 mol in a
day, entry 6, Table 1).
Scheme 1. Anodic and Cathodic Processes in the Formal
Dehydrogenative Methoxylation of N-Formylpyrrolidine (1)
to Give 2
Table 2 reports data from longer time scale experiments,
again with 0.1 M solutions of N-formylpyrrolidine. It can be
a
The narrow interelectrode gap allows electrolysis with a low
electrolyte concentration. The conditions for the anodic
methoxylation of N-formylpyrrolidine were largely established
in our previous papers.6,9 In the new, larger cell the experiments
were limited to demonstrating the performance at higher flow
rates and with larger amounts of reactant.
The percentage conversion and yield in a single pass and the
product formation rate for a series of electrolyses carried out
with 0.1 M solutions of 1 in MeOH with 0.05 M Et4NBF4 were
established (Table 1). The currents employed were generally
slightly higher than the theoretical values to meet the charge
input demanded by Faraday’s law for a full consumption of the
reactant in a single pass. This leads to a charge efficiency below
100% (in fact, varying between 50 and 90% in the experiments
of Table 1), but in a laboratory cell this is not important, and
high conversion and selectivity are the goals. A cell current of 1
A corresponds to an average current density of 10 mA cm−2,
but current density is a parameter of limited value in a cell of
this type. It is inevitable that the local current density drops
strongly along the channel (in theory, the decay is
exponential)9 as the reactant is consumed.
A full consumption was obtained at the slowest flow rate
(entry 1, Table 1), and it decreased marginally at higher flow
rates. The percentage consumption was, however, always
significantly higher than predicted by a simple model,9 which
is believed to originate from the high rate of hydrogen
formation at the counter electrode leading to an increased mass
transfer coefficient in the microchannel (the gas volume
increases the linear flow rate, and the bubbles enhance
turbulence in the flow). On the other hand, the higher flow
rate permitted a higher rate of product formation; indeed,
productivity is proportional to the flow rate (i.e., the weight of
Table 2. Electrolyses of Larger Volumes of Solution
volume of
reactant solution
(mL)
cell
mass of
product
flow rate
current reactant 1
b
b
entry
(mL min−1
)
(A)
consumed
(g)
1
2
3
4
125
480
8
8
5
5
6
5
92%
92%
74%
87%
1.4
c
5.4
480
16
8
4.1
24
2500
a
0.1 M N-formylpyrrolidine in MeOH with 0.05 M Et4NBF4.
b
c
Interelectrode gap 0.5 mm. Determined by calibrated GC. Isolated
product.
seen that the consumption of reactant varied little with the
volume of reactant solution (hence, electrolysis time) and the
weight of product formed was determined only by the weight of
reactant used. The cell performance was stable with time, and
with this reactant concentration and electrolyte flow rate allows
the formation of >100 g of 2-methoxy-N-formylpyrrolidine (2)
in a day (entries 2−3, Table 2).
A few electrolyses were carried out in a cell where the steel
cathode plate had a deeper (0.75 mm) spacer groove in order
to create a narrower interelectrode gap (0.25 mm). This results
in a higher linear flow rate (at constant volumetric flow rate)
and consequent increased mass transfer coefficient and
conversion. Methoxylation of 1 (0.1 M) in the narrower
interelectrode gap at a flow rate of 16 mL min−1 resulted in
94% consumption of the starting material.
3. DISCUSSION
A microflow electrolysis cell for multigram synthesis is
described. It is easily set up and operates with inexpensive
a
Table 1. Electrolyses of 0.1 M N-Formylpyrrolidine (1) in MeOH with 0.05 M Et4NBF4
flow rate/
volume of
cell
current/A
charge
passed/C
reactant 1
b
b
c
entry
cm3 min−1
solution/cm3
consumed
yield of 2 product 2 rate of formation (g h−1
)
1
2
3
4
5
6
2.0
3.0
10
10
1
1
300
200
100%
84%
84%
83%
77%
88%
73%
84%
84%
69%
77%
84%
1.1
2.0
5.0
10
2
240
3.3
8.0
10
3
225
4.3
16.0
128
240
6
2880
10800
8.7
d
16.0
12
20.7
a
b
Interelectrode gap of 0.5 mm. The theoretical charge for full conversion of 10 cm3 of 0.1 M of the reactant solution is 200 C. Determined by
c
d
calibrated GC. Calculated from the yield, which was determined by calibrated GC. Using a 0.2 M solution of reactant.
B
DOI: 10.1021/acs.oprd.5b00260
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
auxiliary equipment. It is particularly convenient to use for
extended electrolysis where the objective is the synthesis of
larger quantity of product. When operated under appropriate
conditions,9 it allows high conversions in a single pass and the
synthesis of tens of grams of product. The cell is also
straightforward and rapid to dismantle, clean, and reassemble.
This design of microflow electrolysis cell also has the
advantage of a low residence time (150−19 s for flow rates of
2−16 mL min−1) in the reactor for high conversion. This
minimizes competing reactions in homogeneous solution
(often a problem during lengthy electrolyses in beaker cells)
and aids high selectivity. The high conversion also greatly
simplifies pure product isolation.
compressed between an aluminum base plate and a perspex top
each of diameter 180 mm via a central bolt (tightened to 20 N
m) and 8 bolts around the perimeter (each tightened to 4 N
m). The reaction solution entered and exited the cell via steel
tubing, 3/16th inch diameter to which connection could be
made with standard fittings. There were separate reservoirs for
reactant and product solutions, and the solution was pumped
with an Ismatec Reglo digital peristaltic pump, with flow rates
generally in the range 1−20 mL min−1. Electrolyses were
carried out with constant currents controlled by either a TTi 35
V/10A power supply (type TSX3510P) or a Farnell AP60-150
regulated power supply. The cell is straightforward to
dismantle, clean, and reassemble.
With high flow rates, a high conversion in a single pass
demands a high current (the charge/unit volume demanded by
Faraday’s law must be passed during the residence time of the
reactant within the cell). In fact, the performance is limited by
the current density that can be used with the present anode
material. When operated with a cell current above 10 A (an
average current density of 0.1 A cm−2), high conversions were
still obtained although the carbon/PVDF composite was found
to undergo some pitting with black powder appearing in the
product reservoir. In the case of the cell possessing a narrower
interelectrode gap (0.25 mm), electrical shorting was observed
at the higher current densities, although this was not observed
when the interelectrode gap was 0.5 mm. Despite the observed
pitting, the anode material was reused multiple times (after
polishing between reactions) without observed detriment to
conversion or productivity. Clearly, alternative more stable
anode materials would be attractive and are being investigated.
Nonetheless, the larger cell reported here with the carbon/
polymer composite anode is an effective tool for multigram
laboratory electrosynthesis. It is particularly advantageous
compared to our earlier cell design for the electrolysis of
substrates with limited solubility or when the electrolysis target
is a large weight of product.
The cell was always operated in the single pass mode. The
volume of the electrolyte solution channel in the microflow cell
was 5 mL. Hence, with a flow rate of 10 mL min−1 the
residence time of reactant in the cell was only 30 s (excluding
the increased flow rate as a consequence of bubble formation)
necessitating a cell current of 3.3 A for full conversion with 0.1
M reactant undergoing a 2e− oxidation.
Chemicals and Analysis. Methanol (Fisher Scientific,
HPLC grade), tetraethylammonium tetrafluoroborate (Alfa
Aesar, 99%), and N-formylpyrrolidine (Sigma-Aldrich 98%+)
were used without purification.
In general, conversions were determined by gas chromatog-
raphy of the cell effluent. A Shimadzu GC-2014 equipped with
an autosampler, FID detector, and an Agilent technologies HP5
column (length 30 m, I.D 0.32 mm, film thickness 0.25 μm)
was used. The results were processed using GC Solution Lite
software. Separations were carried out using He as a carrier gas
with a flow rate of 2.48 mL min−1 through the column. A split
injection was conducted using a split ratio of 100:1. The
injection and detector temperatures were maintained at 280
and 295 °C, respectively. The oven temperature was initially
held at 60 °C and then programmed to increase at 10 °C min−1
to 180 °C, where it was held for 1 min. 1 and 2 were observed
at 5.2 and 6.2 min, respectively. The GC was calibrated using
serial dilutions of a known concentration of both the starting
material and the product.
Electrochemical Synthesis of 2-Methoxy-N-formylpyrroli-
dine. A solution containing 0.10 M of 1 (4.95 g, 0.05 mol, 1.00
equiv) in MeOH (500 mL) with 0.05 M of Et4NBF4 (5.42 g,
0.025 mol, 0.5 equiv) present as electrolyte was sonicated prior
to the electrolysis to ensure complete dissolution. Before
assembly of the reactor, the working electrode (carbon filled
PVDF) was polished with cotton wool. The cell was filled with
MeOH at a flow rate of 8 mL min−1 and the power supply set
to a constant current of 5 A (a constant voltage limit was also
set at 12 V). The cell feed was then switched to the reactant
solution (note: the cell current does not actually reach the set
value until the MeOH in the channel is displaced by electrolyte
solution). The reaction was continued for 1 h, by which point
480 mL of reaction solution has passed, whereupon the cell
feed was switched back to the MeOH reservoir. The product
reservoir solution was analyzed by GC to determine the
conversion (92%), the yield (89%) and the charge efficiency
(49%). The MeOH was removed under reduced pressure, and
the resultant oil was treated with EtOAc causing the Et4NBF4
to precipitate. The solid was removed by filtration (and could
be reused after recrystallization from a minimum amount of hot
MeOH, and drying overnight in a vacuum oven at 90 °C, ∼10
mbar). EtOAc was removed to give a yellow oil, which was
purified by vacuum distillation (100 °C at 15 mbar to remove
4. EXPERIMENTAL SECTION
Microflow Electrolysis Cell. The cell was manufactured by
Cambridge reactor Design Ltd. It was designed to allow the
conversion of larger amounts of product with high selectivity
and conversion in a single pass. The cell design is based on two
circular plate electrodes, diameter 149 mm and a spiral solution
channel, width 5 mm where the electrolyte flows from the
center of the discs to their perimeter (Figure 1). The spiral
solution channel was created by machining a spiral groove (2
mm in width and 0.5 mm in depth) into one of the electrodes
so that there was a 5 mm spacing between neighboring sections
of the groove. A polymer gasket/spacer, thickness 1 mm, was
lazer cut so that it fitted into the groove. When compressed
against a flat plate electrode, this creates a channel 2 m long, 5
mm wide with an interelectrode gap of 0.5 mm. The
interelectrode gap may be adjusted via the depth of the groove
and/or the thickness of the gasket/spacer.
In the particular cell used in this paper, the groove was
machined into the stainless steel (grade 316L, Castle Metals
UK Ltd.) cathode plate, and the anode plate was carbon filled
polyvinylidene fluoride (C/PVDF, type BMA5, Wilhelm
Eisenhuth GmbH, Germany) sheet, thickness 3 mm or 5
mm. The gasket/spacer was cut from a sheet of KALREZ
perfluoroelastomer (James Walker Ltd., 1 mm thick). The
carbon/polymer composite electrode had a copper backing
plate to improve the potential distribution. The cell was
C
DOI: 10.1021/acs.oprd.5b00260
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Communication
starting materials, then 170 °C at 15 mbar), to give 2 as a
colorless oil (5.4 g, 0.042 mol, 88%).
1H NMR data are consistent with reported values.12 FT-IR
1
(cm−1) neat; 3499, 2892, 1671, 1588. H NMR (400 MHz,
CDCl3) δ ppm; Spectra presented as a mixture of rotamers
(∼5:1). 8.40maj and 8.29min (1H, s), 5.37min and 4.92maj (1H, d, J
= 4.8 Hz), 3.58−3.40 (2H, m), 3.37min and 3.26maj (3H, s),
2.13−1.79 (4H, m). 13C NMR (400 MHz, CDCl3) δ ppm;
Spectra presented as a mixture of rotamers. 162.5min and
161.2maj, 89.5maj and 85.4min, 56.5min and 54.2maj, 45.0min and
42.5maj, 31.8min and 31.7maj, 22.0min and 21.2maj. LRMS (ESI) m/
z 130.1 [M + H]+.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.5b00260.
1H and 13C NMR spectra of compound 2 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: rag206@soton.ac.uk.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors acknowledge financial support from the EPSRC
for Factory in a Fumehood: Reagentless Flow Reactors as
Enabling Techniques for Manufacture (EP/L003325/1) and
Core Capability for Chemistry Research in Southampton (EP/
K039466/1) and the European Regional Development Fund
(ERDF) for cofinancing the AI-Chem project through the
INTERREG IV A France (Channel) - England cross-border
cooperation Programme. The authors would also like to thank
the members of the Factory in a Fumehood consortium, for
their helpful discussions during the project.
REFERENCES
■
(1) Lund, H.; Practical Problems in Electrolysis. In Organic
Electrochemistry, 4th ed.; Lund, H., Hammerich, O., Eds.; Marcel
Dekker: New York, 2001; pp 223−292.
(2) Pletcher, D.; Walsh, F. C. Industrial Electrochemistry, 2nd ed.;
Chapman & Hall: London, UK, 1990.
(3) Watts, K.; Haswell, S. J. Chem. Soc. Rev. 2005, 34, 235.
(4) Yoshida, J.-i.; Nagaki, A. Electrochemical Reactions in Micro-
reactors. In Microreactors in Preparative Chemistry; Reschetilowski, W.,
Ed.; Wiley-VCH: Weinheim, Germany, 2013; pp 231−242.
(5) Watts, K.; Baker, A.; Wirth, T. J. Flow Chem. 2015, 4, 2.
(6) Kuleshova, J.; Hill-Cousins, J. T.; Birkin, P. R.; Brown, R. C. D.;
Pletcher, D.; Underwood, T. J. Electrochim. Acta 2012, 69, 197.
(7) Yoshida, J. Chem. Commun. 2005, 4509.
(8) Kuleshova, J.; Hill-Cousins, J. T.; Birkin, P. R.; Brown, R. C. D.;
Pletcher, D.; Underwood, T. J. Electrochim. Acta 2011, 56, 4322.
(9) Green, R. A.; Brown, R. C. D.; Pletcher, D. J. Flow Chem. 2015, 5,
31.
(10) Nyberg, K.; Servin, R. Acta Chem. Scand. 1976, 30b, 640.
(11) Eberson, L.; Hlavaty, J.; Jonsson, L.; Nyberg, K.; Servin, R.;
̈
Sternerup, H.; Wistrand, L. H. Acta Chem. Scand. 1979, 33b, 113.
(12) Mitzlaff, M.; Warning, K.; Jensen, H. Liebigs Ann. Chem. 1978,
1978, 1713.
D
DOI: 10.1021/acs.oprd.5b00260
Org. Process Res. Dev. XXXX, XXX, XXX−XXX