Continuous flow processing of bismuth-photocatalyzed atom transfer radical addition reactions using an oscillatory flow reactor

Article abstract of DOI:10.1039/d0gc03070h

Metal oxides represent an abundant and non-toxic class of photocatalysts for organic transformations. However, their use in larger scale processes is complicated by incompatibilities with continuous flow processing - a proven scale-up route for photochemi

Full text of DOI:10.1039/d0gc03070h

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: P. Bianchi, J.
Williams and C. O. Kappe, Green Chem., 2020, DOI: 10.1039/D0GC03070H.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Please do not adjust margins
Page 1 of 9
Green Chemistry
View Article Online
DOI: 10.1039/D0GC03070H
ARTICLE
Continuous flow processing of bismuth-photocatalyzed atom
transfer radical addition reactions using an oscillatory flow
reactor
Received 00th January 20xx,
Accepted 00th January 20xx
Pauline Bianchi,a Jason Williams*a,b and C. Oliver Kappe*a,b
DOI: 10.1039/x0xx00000x
Metal oxides represent an abundant and non-toxic class of photocatalysts for organic transformations. However, their use
in larger scale processes is complicated by incompatibilities with continuous flow processing – a proven scale-up route for
photochemistry. We detail the development of an efficient atom transfer radical addition protocol using a sustainable
solvent system (acetone:PEG 400) and a low loading (2 mol%) of Bi2O3, which can be handled in an oscillatory flow reactor.
Optimization of the reaction and oscillatory parameters led to high throughput (36 g in 4 h, 89% yield, 599 g L-1 h-1), with a
process mass intensity (PMI) of just 8.5. The process also facilitates high recyclability (3 cycles with no loss of yield), and was
demonstrated to be applicable to a range of other substrates on multigram scale, in moderate to excellent yields.
profile than HBr or Br2. Since bromination is also an important
reaction in organic synthesis,6 these transformations are
attractive from a sustainability viewpoint. Accordingly, less
toxic, more abundant, and cheaper catalysts than iridium or
ruthenium complexes have been proposed for such reactions,
Bi2O3 being one of these candidates.3a
Introduction
Bismuth, the 83rd element of the periodic table, is a metal which
is increasingly considered for catalysis because of its low cost,
non-toxic nature, large reserves and good availability.1 Among
bismuth species, the semiconductor bismuth oxide (Bi2O3) has
gained some interest as a photocatalyst in recent years. This is,
in part, owing to its lower bandgap compared to the other
commonly used metal oxide semiconductor photocatalysts,
such as titanium oxide (Bi2O3 = 2.6 eV vs 3.0 eV or 3.2 eV for TiO2
in its rutile or anatase forms, respectively). This advantage
enables the use of visible light irradiation in place of high energy
UV wavelengths.2 A small number of synthetically useful
transformations using Bi2O3 as a photocatalyst have already
been demonstrated, in particular by Pericàs et al.3
Despite the attractive profile of visible light-driven
reactions, the requirement for efficient light penetration to
homogeneously irradiate the reaction vessel renders them
difficult to manage, especially at synthetically useful
concentrations. As described by the Beer-Lambert law,
increasing the size of a batch vessel leads to drastic light
attenuation, potentially leading to degradation and side-
product formation. To tackle this issue, continuous flow has
been proven as a powerful alternative processing method.7 The
small dimensions of the reactor channels ensure a short
irradiation path length (which allows quasi homogeneous
irradiation), thus eliminating potential dark zones and
increasing the efficiency and scalability of these reactions.
Surprisingly, despite the well-explored benefits of flow
photochemistry, relatively few ATRA reactions in flow have
been reported.8
Performing visible-light-driven reactions is appealing since it
can enable mild conditions, selectivity, sustainability, and
catalytic amounts of the light-absorbing species. Therefore,
various photocatalysts have been developed for a broad range
of organic reactions.4 Atom transfer radical addition (ATRA)
reactions have received substantial attention as
a
photocatalyzed synthetic transformation, whereby a substrate
R-X inserts across an unsaturated compound, resulting in
complete atom incorporation. The addition of haloalkyl,
aminoperfluoroalkyl, hydroperfluoroalkyl, or haloperfluoroalkyl
moieties, to name a few, have been successfully performed
upon light irradiation.5
Bi2O3 (2 mol%)
2,6-lutidine
Br
Among these reactions, the use of organobromide
substrates has interested researchers because of their safer
R2
+
R2 Br
2
R1
R1
1
3
multigram per
hour throughput
acetone:PEG 400 (0.5 M)
405 nm LED
a. Institute of Chemistry, University of Graz, NAWI Graz, Heinrichstrasse 28, 8010
Graz, Austria.
5 bar, 75 °C, 15 min
b. Center for Continuous Flow Synthesis and Processing (CC FLOW), Research Center
Pharmaceutical Engineering (RCPE), Inffeldgasse 13, 8010 Graz, Austria.
E-mail: jason.williams@rcpe.at; oliver.kappe@uni-graz.at
Fig 1. General scheme of the bismuth-photocatalyzed ATRA across unsaturated
compounds in flow.
†Electronic
Supplementary
Information
(ESI)
available:
See
DOI: 10.1039/x0xx00000x
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 9
ARTICLE
Green Chemistry
The use of continuous flow reactors, however, can raise radical process promoted by light. This transformation strongly
View Article Online
DOI: 10.1039/D0GC03070H
numerous issues, particularly when handling solids. Depending depends on the solvent properties (see Fig. S5b). However, this
on the flow rate, the properties of the reactor and the reaction reduction is hypothesized to be slower than the ATRA reaction,
medium, settling and even clogging will occur unless since it was only observed after complete reaction conversion
preventative measures are taken.9 An appealing development (i.e. by overirradiation of the reaction medium), further
for efficiently handling solids in flow is that of oscillatory flow emphasizing the benefits of continuous flow to prevent such
reactors (OFRs). The use of a superimposed pulsation (meaning phenomena.
the addition of a symmetric pulse on top of the net flow rate)
Finally, Bi2O3 has a bandgap that depends on the size and
enhances mixing, hence ensuring a stable suspension of the morphology of the particles.19 Between the particles of 90-210
solids along the reactor. Other advantages of this technology nm diameter and the ones of 10 µm diameter that were
include: mixing independent of the net flow rate, ease of scale- commercially available, the larger ones were chosen, owing to
up, and improved heat transfer.10 Interest in this promising their significantly higher reactivity (see Fig. S5c and Table S1).
strategy has grown significantly over the last 20 years, Incidentally, these were also more readily available (EUR55 /
facilitating the development of various oscillatory flow 100 g).20
processes for multiphasic and/or slow reactions, which are
difficult to handle using conventional flow reactors.11
Specifically, an OFR using static mixing elements (the HANU
reactor commercialized by Creaflow)12 has recently been
demonstrated as an effective flow reactor for photochemical
reactions containing solids,13 or two immiscible liquid phases.14
This reactor has previously been characterized actinometrically,
demonstrating a high volumetric photon flux of 1.97 × 10-
3 mol L-1 s-1.13b
1a
Bi2O3 (size, loading)
additive (equiv.)
Br
CO2Et
CO2Et
+
EtO2C
CO2Et
wavelength, time,
temperature, solvent
3a
Br
2a
Fig. 2 Bismuth-photocatalyzed ATRA reaction model between 1-hexene (1a) and diethyl
bromomalonate (2a).
In this work, the inherent assets of an OFR were combined
with the potential of Bi2O3 as a sustainable and non-toxic visible
light photocatalyst, to provide an efficient protocol for ATRA
reactions of organobromides 2 with terminal olefins 1 (Fig. 1).
The OFR used in this study has previously been exploited for
handling two different solids in flow, a carbon nitride derivative
(density = 2.34 g cm-3)13a and Na2CO3 (density = 0.86 g cm-3).13b
The innovative process, presented here, pushes the limit of this
reactor setup in terms of handling solid particles and enables
processing of a homogeneous suspension of the much denser
solid, Bi2O3 (density = 8.90 g cm-3).15
It has been decided to use visible light (λ >380 nm) instead
of UV since it is less harmful and less energetic, hence fitting
within the context of greenness and improved substrate
compatibility. LEDs of 405 nm and 460 nm were available for the
flow setup (or 400 nm and 455 nm for batch experiments). The
400 nm and 405 nm wavelengths were selected, since the
absorption profile of Bi2O3 decreases drastically between 400
and 500 nm.19b This choice was reinforced with comparative
experiments in batch, which showed that the reaction was
much faster at 400 nm than at 455 nm (5 h instead of 20 h; see
Table S1 for initial reaction scoping experiments in batch).
The context of sustainability requires a conscientious choice
of solvent, considering its impact on health, safety, and
environment.21 Furthermore, its inert profile towards the
photocatalyst, as well as performance in the ATRA reaction
must be considered. As an additional parameter for this
application, a high solvent viscosity may significantly decrease
the potential settling of the catalyst particles. A screen of
solvents was, therefore, performed and the corresponding
results are presented in Table 1, alongside the solvent’s
properties.
The only way to fulfill all these requirements was to mix two
solvents. Dimethylsulfoxide, dimethylformamide, acetonitrile,
and dimethylacetamide (entries 1, 2, 4 and 12) enabled good
conversion, but are considered as hazardous or problematic
solvents. The use of ester (ethyl acetate and isopropyl acetate;
entries 3 and 9) or alcohol (ethanol, isopropanol, ethylene
glycol, and polyethylene (PEG); entries 7, 8, 10 and 11) solvents
provided a slow reaction, aside from benzyl alcohol, which
strongly reduced Bi2O3 particles (entry 13). No conversion was
obtained with water (entry 6). Acetone (entry 5) was a
promising solvent in terms of its reaction efficiency and
Results and discussion
Batch optimization
For the purpose of reaction optimization, the reaction between
1-hexene 1a and diethyl bromomalonate 2a was chosen as a
model (Fig. 2).
A
few observations regarding the
physicochemical properties of Bi2O3 were first considered.
Firstly, Bi2O3 can react with HBr to form BiOBr or BiBr3,
depending on the content of HBr in the medium.16 Starting
material 2a contains ~5% HBr due to debromination into diethyl
malonate. Furthermore, HBr can be formed during elimination
as a side reaction following the ATRA,17 leading to acidification
of the reaction mixture and reaction of the bismuth species. The
resulting complexes are far more soluble and possess different
optical properties than Bi2O3.18 It is, therefore, recommended
to add an organic base to the medium, such as 2,6-lutidine to
keep Bi2O3 intact (see Fig. S5a). As proposed in other studies,
this base may also interact with the alkyl halide substrate,
activating the C-Br bond to cleavage.8b
Another important point to consider is the observation of
black solids, attributed to Bi2O3 reduction to Bi0 through a
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Page 3 of 9
Green Chemistry
ARTICLE
Green Chemistry
View Article Online
DOI: 10.1039/D0GC03070H
Table 1 Solvent screening for the bismuth-photocatalyzed ATRA reaction between 1a and 2a a
Entry
1
Solvent
Dimethylsulfoxide
Dimethylformamide
Ethyl acetate
Greennessb
Problematic
Price for 1 Lc (EUR) Viscosity (cP) Bi2O3 reduction Conv.d (%) Yieldd (%)
42.3
1.99
0.92
0.46
0.37
0.32
1.00
1.10
2.86
0.52
16.10
109.42
0.95
5.47
3.48
22.14
Slow
Slow
Slow
Medium
Slow
Fast
>99
97
11
>99
>99
<1
97
95
5
2
Hazardous
22.5
3
Recommended
Problematic
17.7
4
Acetonitrile
22.3
98
98
<1
4
5
Acetone
Recommended
Recommended
Recommended
Recommended
Recommended
Recommended
Recommended
Hazardous
13.5
6
Water
/
7
Ethanol
18.5 (denaturated)
Slow
Medium
Slow
Fast
7
8
Isopropanol
21
19
2
16
1
9
Isopropyl acetate
Ethylene glycol
Polyethylene (PEG) 400
Dimethylacetamide
Benzyl alcohol
36.1
42.4
26.7
37.6
20.2
19.28
16.14
10
11
12
13
14
15
3
1
Slow
Slow
Fast
30
>99
>99
24
>99
25
94
95
20
95
Problematic
Acetone:ethylene glycol (8:2)
Acetone:PEG 400 (8:2)
Recommended
Recommended
Fast
Slow
a Conditions: [1a] = 0.675 M, [2a] = 0.5 M and [2,6-lutidine] = 0.5 M, reaction volume = 2 mL, Bi2O3 loading = 1 mol%, 400 nm LED irradiation, room temperature, 5 h; b according to
;
Prat et al.21 c from VWR (for technical grade when available, accessed on 3rd September 2020); d consumption of starting material 2a and yield of desired product 3a, based on GC-
FID area with decane as the internal standard.
greenness (according to the CHEM21 selection guide,21 but no Bi2O3 or 2,6-lutidine, the reaction proceeded (albeit very
exemption from “volatile organic compound” classification,22 slowly) in the mixture of acetone:PEG 400 (entry 7), but not in
and general non-toxic and biodegradable character) but was pure acetone (entry 8). Nevertheless, the selectivity towards 3a
not viscous enough to suspend the solid particles. It was was far lower than during the photocatalyzed reaction (81% for
therefore decided to add a small portion of PEG to increase the entry 7, compared with 95% for entry 5). By the addition of 2,6-
viscosity of the mixture.
lutidine, the reaction slowed down drastically (entry 9), perhaps
The choice of the PEG chain length and its content in the due to the basic rather than acidic conditions.
mixture was determined through laser diffraction experiments.
In these experiments, the mixture was stirred, then stirring
Table 2 Batch optimization of the bismuth-photocatalyzed ATRA reaction between 1a
and 2a a
ceased and the resulting optical density was recorded over 20
seconds, as a measure of how quickly the solids settle (see
section B.1.3 in the ESI). Such experiments showed that a high
amount of PEG was advantageous and that short chain lengths
were more effective in improving particle suspendability.
However, the higher content of Bi0 obtained upon irradiation of
a reaction mixture with PEG 200 than PEG 400 restricted the
choice to the latter (see Fig. S5b). Therefore, the mixture of
acetone and PEG 400 (8:2 in volume) was selected as the best
candidate for the next experiments (entry 15). The selectivity
towards 3a was slightly reduced, when compared with pure
acetone (entry 5), but the yield remained high (95%).
Having considered all of these aspects, some further
optimization experiments and control reactions were carried
out (Table 2). Increasing the Bi2O3 loading showed no
improvement in reaction performance, likely owing to a higher
amount of Bi0 particles produced, or to a more turbid
suspension with poor light transmission (entry 1, see also Table
S1 entries 5 vs 6). An excess of olefin 1a was also required to
ensure complete conversion (entries 2, 3, and 4). Usually the
limiting reagent during ATRA reaction is the olefin but, as in the
method reported by Pericàs, the opposite trend can occur.3a
High conversion and yield were obtained with Bi2O3 and
lutidine upon visible light irradiation (entry 5), whilst no
reaction occurred without irradiation (entry 6). With irradiation,
Bi2O3
Additive [1a]
(equiv.) (M)
[2a]
(M)
λirr. Conv. Yield
(nm) (%)b (%)b
Entry Solvent loading
(mol%)
lutidine
0.675
(1)
1
2
3
4
5
6
acetone
acetone
acetone
acetone
5
1
1
1
1
1
0.5
0.5
0.5
0.5
0.5
0.5
400 97
93
lutidine
0.675
(1)
400 >99 98
lutidine
0.5
400 83
400 49
82
49
(1)
lutidine
0.44
(1)
acetone:PEG
400 (8:2)
lutidine
0.675
(1)
400 >99 95
acetone:PEG
400 (8:2)
lutidine
0.675
(1)
none <1
<1
acetone:PEG
400 (8:2)
7
8
9
none
none
none
none
none
0.675
0.675
0.675
0.5
0.5
0.5
400 36
400 <1
29
<1
2
acetone
acetone:PEG
400 (8:2)
lutidine
(1)
400
2
a Reaction volume = 2 mL, room temperature, 5 h; b consumption of starting material 2a
and yield of desired product 3a, based on GC-FID area with decane as the internal
standard.
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 9
ARTICLE
Green Chemistry
As an additional control experiment to determine whether suspension, even at low flow rates. In addition toVtiheweAprtuicllesaOtnolinre,
DOI: 10.1039/D0GC03070H
this catalysis is truly heterogeneous, an experiment was carried the use of thin tubes and strategically placed vibrator motors
out whereby the reaction was started, then filtered after 2 h, ensured better particle suspension in the auxiliary tubing
and initiated again. This serves to determine whether a (outside the photoreactor itself). It was also important to use a
homogeneous catalytic species is formed as the reaction back-pressure regulator (BPR, Swagelok) and pump (Vapourtec
progresses. At 2 h reaction time, 5% conversion was observed, SF-10) capable of solid handling (see section B.2.1. in the ESI).
then after filtration and a further 3 h reaction, the level of
After compiling the oscillatory flow setup, and optimizing
conversion increased to 76%. This result suggests that the the parameters that enable a homogeneous solid suspension, a
reaction’s initiation period is likely caused by the formation of brief optimization study in flow was performed, to achieve
homogeneous species over time, by either dissolution, or similar results to batch experiments (Table 3). Despite the more
derivatization of Bi2O3.
efficient irradiation in flow, keeping room temperature with a
short residence time was not enough to overcome the
reaction’s induction period (entry 1) and the observation was
the same at 50 °C (entry 2).
Transition to continuous flow
The main challenge during the transposition to continuous flow
conditions was to avoid the settling and deposition of the dense
Bi2O3 particles along the reactor. To overcome such an issue and
benefit from the sustainable profile of continuous flow
technology, a few considerations had to be taken. It was
hypothesized that the particle suspendability would be
enhanced with a viscous solvent mixture, a high particle loading,
and a small particle diameter. Therefore, the viscous mixture of
acetone and PEG 400 (8:2 in volume) was maintained and an
increase of Bi2O3 loading from 1 to 2 mol% was applied.
However, the low reactivity of the smaller particles discouraged
their use in the flow process, so the large particles (10 µm
diameter) were kept.
By increasing the temperature to 70 °C and the Bi2O3 loading
to 5 mol%, a conversion of 46% was obtained (entry 3).
However, to minimize Bi0 formation, the Bi2O3 loading was
reduced to 2 mol% and both the temperature and residence
time were increased to 75 °C and 15 min, respectively, to
provide excellent conversion and selectivity (entry 4). From that
point, it was considered necessary to perform additional control
reactions, since the light irradiation and the temperature have
changed. No conversion was observed without irradiation at
75 °C (entry 5) or without Bi2O3 (entry 6).
Table 3 Flow optimization of the bismuth-photocatalyzed ATRA reaction between 1a and
2a a
Bi2O3
Irradiation Residence Temperature Conv. Yield
Entry loading
(nm)
time (min)
(°C)
(%)d
(%)d
(mol%)
1b
2b
3b
4b
5c
6c
2
405
405
405
405
none
405
7.5
7.5
7.5
15
25
50
70
75
75
75
<1
<1
46
99
<1
<1
<1
<1
42
94
<1
<1
2
5
2
2
15
none
15
a [1a] = 0.675 M, [2a] = 0.5 M and [2,6-lutidine] = 0.5 M, solvent = acetone:PEG 400 (8:2),
b
processed reaction volume = 25 mL, pressure = 5 bar; oscillation parameters: 2.4 Hz
c
frequency, 0.265 mL amplitude; oscillation parameters: 1.5 Hz frequency, 0.12 mL
amplitude; d consumption of starting material 2a and yield of desired product 3a, based
on GC-FID area with decane as the internal standard.
Preliminary experiments were in accordance with studies in
the literature, which demonstrated a stronger effect of the
oscillation amplitude than the frequency on the RTD (see
section B.2.2. in the ESI).13,23 A low oscillation amplitude was
required to maintain a similar profile for both reaction phases.
By setting the frequency to 1.5 Hz (50%) and the amplitude to
0.12 mL (20%), it was possible to observe a similar response
profile (Fig. 4). An increase in conversion at the end of the
process was also observed, owing to a higher catalyst loading
present at that time point (different residence time profile for
the solid phase compared to the liquid phase).24
Fig. 3 Schematic of the oscillatory flow setup for the bismuth-photocatalyzed ATRA
reaction.
The design of the oscillatory flow setup was also paramount
to avoid particle settling (Fig. 3). The OFR used for this study,
the HANU reactor,12 includes a series of cubic static mixing
elements. When combined with a pulsator pump, with tuneable
oscillation character, the split-and-recombine mixing obtained
by the cubic elements can be generated during each pulsation.
Furthermore, the resulting narrow channels (2 mm x 2 mm)
ensure homogeneous irradiation and reinforce stable particle
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 9
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
represents <10% of the initial Bi2O3 loading. AfterVrieewcyArctilcinlegOntlhinee
solid 3 times, no loss of activity wDaOs I:o10b.s1e03rv9/eDd0G(CF0ig3.0760H),
representing another promising feature for the sustainability
profile of this methodology.
Fig. 4 Distribution of conversion and reaction concentration over time, when working
under continuous flow conditions. Conditions: [1a] = 0.675 M, [2a] = 0.5 M and [2,6-
lutidine] = 0.5 M, solvent = acetone:PEG 400 (8:2), processed volume = 50 mL, Bi2O3
loading = 2 mol%, 405 nm LED irradiation, pressure = 5 bar, temperature = 75 °C,
residence time = 15 min, flow rate = 1 mL min-1, frequency = 1.5 Hz (50%), amplitude =
0.12 mL (20%). Conversion was determined by GC-FID using decane as the internal
standard.
Scale-out, Bi2O3 recycling and substrate scope
The likelihood of clogging and settling increases with the time
of processing, so a long run experiment was tested to show the
stability of this process. The reaction was run for 4 h (250 mL
reaction mixture processed) without any issues. Following a
simple workup, 35.9 g of the model substrate 3a was obtained,
corresponding to an 89% yield (accounting for 92.5% purity by
GC-FID analysis). This corresponds to a process mass intensity
(PMI: total mass involved in the reaction, divided by the product
mass) of 8.5, without taking into account any solvent
recyclability, which is low for a photochemical process such as
this.
During the scale-out experiment, 97% conversion was
observed, with a fluctuation of only ±2% conversion,
demonstrating the stability of the protocol (Fig. 5). The space-
time yield of this experiment (productivity divided by 15 mL
reactor volume) was calculated to be 599 g L-1 h-1. This high
productivity, achieved by high concentration, high conversion
and short residence time, demonstrates new possibilities for
industrial scale processing of ATRA reactions.
Since the use of solid catalysts is advantageous in terms of
cost and time needed for downstream purification, the
reuseability of Bi2O3 particles was assessed. After centrifugation
and washing (acetone, water, then acetone, with centrifugation
between each solvent wash; 5000 rpm for 15 min) and drying
overnight (105 °C in an oven), high amounts of Bi2O3 were
recovered (89%, 94% and 90% of the input mass, for three
cycles). This high recovery demonstrates minimal solid settling
or dissolution along with the flow setup, and only minor losses
in the particle recovery process.
Fig. 5 Long run experiment of the ATRA procedure under continuous flow conditions.
Conditions: [1a] = 0.675 M, [2a] = 0.5 M and [2,6-lutidine] = 0.5 M, solvent = acetone:PEG
400 (8:2), processed volume = 250 mL, Bi2O3 loading = 2 mol%, 405 nm LED irradiation,
pressure = 5 bar, temperature = 75 °C, residence time = 15 min, flow rate = 1 mL min-1,
oscillation frequency = 1.5 Hz, amplitude = 0.12 mL. (a) Reaction scheme. (b) Stability of
the long run experiment over time. Conversion was determined by GC-FID using decane
as the internal standard.
The protocol was extended to other substrates (Fig. 7, see
ESI for procedures and characterization data). In this reaction
scope, a selection of olefins were reacted with 2a as the
brominated donor, under the optimized conditions (75 °C and
20 min residence time). Increasing the residence time (30 min)
and temperature (80 °C) was, however, required to facilitate
complete reaction with allyl benzene 1b (93% conv.), allyl
alcohol 1c (97% conv.), and 1-hexyne 1d (79% conv.). The
selectivity towards the brominated product 3 depends on the
olefin 1 and was lower for 3c and 3d than for 3a and 3b. In these
cases, the major side product was proposed to be from
elimination of the desired product. Moderate to excellent
isolated yields were obtained for 3a (94%), 3b (90%), 3c (66%),
and 3d (55%), all of which were performed on multigram scale.
Since the Bi2O3 loading was kept at 2 mol% to minimize
settling issues, some fresh Bi2O3 (~20 mg per cycle) was added
to replace the mass lost during the previous cycle. As previously
demonstrated, a homogeneous bismuth species is likely to be
involved in the reaction, but the lost mass from each recycle
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 9
ARTICLE
Green Chemistry
through a metal oxide photocatalyzed ATRA rVeieawctAiroticnle.OTnlhinee
added species, Bi2O3, fulfills the expectatDioOnI:s1o0.f10a3g9r/eDe0nGCc0a3ta07ly0sHt
since it is cheap, widely available, easy to procure, non-toxic and
in the solid state (allowing straightforward reuse). Furthermore,
the activation of the radical reaction has been promoted by
visible light, less energetic and harmful than UV irradiation. The
choice of the solvent, a mixture of acetone and PEG 400, has a
low cost, minimal safety concerns, negligible extent of bismuth
reduction, and no reaction quenching. The use of oscillatory
flow technology enables experiments at larger scale, with
better control of the photochemical process and efficient
handling of the dense Bi2O3 species (d = 8.9 g cm-3).
R1
Fig. 6 Recycling experiments of the ATRA procedure under the optimized flow conditions.
Conditions: [1a] = 0.675 M, [2a] = 0.5 M and [2,6-lutidine] = 0.5 M, solvent = acetone:PEG
400 (8:2), processed volume = 50 mL, Bi2O3 loading = 2 mol%, 405 nm LED irradiation,
pressure = 5 bar, temperature = 75 °C, residence time = 20 min, flow rate = 0.75 mL min-1,
frequency = 2.1 Hz, amplitude = 0.12 mL. Conversions were determined by GC-FID using
decane as the internal standard.
1a-d
(1.35 equiv.)
+
X
R2
2a-d
(1 equiv.)
X
R2
R1
3a-h
acetone:PEG 400 (8:2)
405 nm, 75 °C, 5 bar
20 min, 0.12 mL, 2.1 Hz
52-97% yield
+
Bi2O3 (10 µm, 2 mol%)
The applicability of different ATRA donors was also
examined. Bromoacetonitrile 2b and carbon tetrabromide 2d
were assessed for new bromofunctionalization reactions and
+
2,6-lutidine (1 equiv.)
perfluorohexyl
iodide
2c
was
tested
for
an
iodoperfluoroalkylation.8b,25 An excellent yield was obtained for
the iodoperfluoroalkylation under the less forcing set of
conditions, owing to the weaker C-I bond. In this case, the
desired compound 3f was the only product detected (97% yield,
with no chromatography required). Iodoperfluoroalkylation is
another interesting ATRA transformation since perfluorinated
compounds play a key role in different fields such as medicinal
chemistry.26
Using the more forcing conditions (80 °C and 30 min
residence time) facilitated an excellent yield of 3e (95%). The
selectivity towards 3g and 3h was, however, much lower, due
to the high propensity for elimination from the desired product.
This decreased further at higher temperature and residence
time. Moderate yields were obtained for 3g and 3h under the
optimized conditions at 75°C and 20 min residence time (52%
and 57% respectively). These results were similar to those
reported in the literature,3a but it must also be stated that
photocatalyst-free ATRA reactions of CBr4 have been
reported.26
HO
1a
1d
CO2Et
1c
1b
EtO2C
Br
CN
CBr4
2d
I
C6F13
2c
2a
2b
Br
CO2Et
CO2Et
Br CO2Et
3a
94% (>99%)
CO2Et
Br CO2Et
3b
90% (93%)a
CO2Et
Br CO2Et
HO
Br CO2Et
3c
66% (97%)a
3d
55% (79%)a
CN
C6F13
Br
I
3e
3f
95% (>99%)a
97% (>99%)
CBr3
CBr3
As a comparison of reaction productivity, the protocol of
Pericàs et al. performed the synthesis of 3h in 54% yield after
48 h irradiation, leading to a space-time yield of 2.5 g L-1 h-1.3a
Although not directly comparable (different light intensity, for
example), the space-time yield of 3h using this flow protocol
was determined as 385 g L-1 h-1, showing far higher productivity
than previously reported. Such an increase of this throughput
reinforces the green profile and larger scale applicability of this
continuous flow procedure.
Br
3g
52% (86%)
Br
3h
57% (71%)
Fig. 7 Substrate scope of the ATRA procedure under continuous flow conditions.
Processed volume = 25 mL except for 3a (50 mL). Isolated yield of the desired product
on multigram scale is displayed. Value in parentheses shows consumption of starting
material 2, based on GC-FID area with decane as the internal standard. reaction
performed at 80 °C, with 30 min residence time.
a
Careful optimization of the flow setup, oscillation
parameters, and chemical conditions (in batch and flow) avoids
clogging or settling of the particles, while maintaining a narrow
RTD. In this way, the model product was synthesized in high
Conclusions
An oscillatory flow process has been designed for the
bromofunctionalization of terminal unsaturated compounds
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 9
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
yield after 4 h of processing without any reactor clogging 1 mL), and submitted for GC analysis (optimized conditions
View Article Online
(35.6 g, 89% yield, 599 g L-1 h-1). To our knowledge, this is the provided conv. >99%, yield 95%).
DOI: 10.1039/D0GC03070H
highest throughput obtained for a photochemical ATRA Representative flow procedure: A sealed Duran bottle
bromination. No loss of activity was observed after recycling the containing 2 mol% of Bi2O3 (233 mg, 0.5 mmol), 25 mmol of 2a
solid Bi2O3 3 times, with high recovery. Finally, a substrate scope (1 equiv.), 25 mmol of 2,6-lutidine (1 equiv.), 5 mmol of decane
was performed, with moderate to excellent yields on multigram (20 mol%) and 50 mL of a degassed mixture of acetone:PEG 400
scales. It is envisaged that the success demonstrated here will (8:2 in volume) was purged with argon for 20 min. A sample of
serve to encourage the use of metal oxide semiconductors, and 33.7 mmol (1.35 equiv.) of 1a was added through the septum
other sustainable inorganic photocatalysts, in continuous flow. via a syringe. The feed was then pumped at a flow rate of 0.75
mL min-1 through the flow reactor pressurized at 5 bar, heated
to 75 °C and irradiated at 405 nm (residence time = 20 min).
During the reaction, the pulsator was set to 20% (0.12 mL)
Experimental section
pulsation amplitude and 70% (2.1 Hz) pulsation frequency. The
collected yellow turbid suspension was then centrifuged for 15
min at 5000 rpm. PEG 400 was removed by extraction with
50 mL of water and 50 mL of diethyl ether. The aqueous phase
was washed with 3 × 50 mL of diethyl ether. The combined
organic fractions were washed with 50 mL of water and then
dried over Na2SO4 before being concentrated on a rotary
evaporator. Although product purity was generally high, to
obtain analytically pure samples, the obtained oil was purified
by flash chromatography on silica gel (40-60 petroleum
ether:ethyl acetate, 95:5) to afford the corresponding product
(>99% conv., 94% yield, 7.6 g). In the meantime, the solid
particles were washed with acetone, then water and finally with
acetone, with centrifugation steps in between (15 min at 5000
rpm), before being dried overnight in an oven at 105 °C.
For the scale-out experiment, this procedure was repeated, but
with a total solution volume of 250 mL.
General information
Bismuth(III) oxide, 2,6-lutidine, decane (internal standard), 1-
hexene 1a, diethyl bromomalonate 2a, allylbenzene 1b, allyl
alcohol 1c, 1-hexyne 1d, bromoacetonitrile 2b, perfluorohexyl
iodide 2c, carbon tetrabromide 2d, polyethylene glycol 400 and
acetone (technical grade) were received from commercial
sources and used without further purification. Conversion and
optimization yields were determined by gas chromatography
1
coupled to a flame ionization detector (GC-FID) or by H NMR
spectroscopy, with decane or 1,3,5-trimethoxybenzene
respectively as the internal standards. Structural
characterization was performed by 1H and 13C NMR
spectroscopy (300 MHz Bruker spectrometer) in CDCl3 (see ESI).
Product masses were confirmed by high-resolution mass
spectrometry (HRMS). Particle size and suspendability
measurements were carried out using laser diffraction (LD)
analysis (see ESI).
Experimental oscillatory flow setup
Conflicts of interest
Flow experiments were carried out in a commercially available
plug flow HANU reactor from Creaflow12 (Hastelloy C276, 15 mL
internal volume, containing of 2 x 2 x 2 mm cubic static mixers
and irradiated at the top by a Peschl Ultraviolet module
composed of 36 x 405 nm LEDs). The feed mixture was pumped
using a Vapourtec SF-10 peristaltic pump, while the oscillations
were delivered by a Creaflow customized ProMinent Beta/4
pump (PTFE/carbon pump head). Pressure was applied by a
Swagelok back pressure regulator (KCB Series, up to 25.8 bar).
Thermal regulation was managed with a Huber CC304
thermostat. All connections between the parts of the setup
were achieved with 1/8’’ (1.6 mm i.d.) or 1/16’’ (0.8 mm i.d.)
o.d. PTFE/PFA tubes with PEEK or stainless steel (Swagelok)
fittings.
Representative batch procedure: A sealed vial containing
1 mol% of Bi2O3 (5 mg, 0.01 mmol), 1 mmol of 2a (1 equiv.),
1 mmol of 2,6-lutidine (1 equiv.), 0.5 mmol of decane (50 mol%)
and 2 mL of a degassed mixture of acetone:PEG 400 (8:2 in
volume) was purged with argon for 10 min. 1.35 mmol of 1a was
added through the septum via a syringe. The vial was then
irradiated for 5 h at room temperature at 10 cm from the
400 nm LEDs. After reaction completion, 100 µL of the yellow
turbid solution was taken and filtered (0.45 μm syringe filter).
The organic phase was extracted from 2 mL of brine by 2 mL of
ethyl acetate, dried over Na2SO4, diluted with MeCN (100 µL in
There are no conflicts to declare.
Acknowledgements
The CC FLOW Project (Austrian Research Promotion Agency FFG
No. 862766) is funded through the Austrian COMET Program by
the Austrian Federal Ministry of Transport, Innovation and
Technology (BMVIT), the Austrian Federal Ministry of Science,
Research and Economy (BMWFW), and by the State of Styria
(Styrian Funding Agency SFG). P.B. acknowledges support by the
University of Liege through the Erasmus+ funding program. The
authors gratefully acknowledge Creaflow for the generous loan
of the HANU Reactor used in this study. The authors would also
like to thank Michael Piller for assistance with laser diffraction
measurements.
Notes and references
1
2
3
R. Mohan, Nat. Chem., 2010, 2, 336.
P. Riente and T. Noël, Cat. Sci. Technol., 2019, 9, 5186–5232.
For selected studies using Bi2O3 as a photocatalyst, see: a) P.
Riente and M. A. Pericàs, ChemSusChem, 2015, 5, 1841–1844;
b) P. Riente, A. Matas Adams, J. Albero, E. Palomares and M.
A. Pericàs, Angew. Chem., Int. Ed., 2014, 53, 9613–9616; c) L.
Buglioni, P. Riente, E. Palomares and M. A. Pericàs, Eur. J. Org.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 9
ARTICLE
Green Chemistry
Chem, 2017, 2017, 6986–6990; d) O. O. Fadeyi, J. J. Mousseau, 20 EUR55 / 100 g for 10 µm particles, vs EUR185 /V1ie0w0ArgticfleoOrn9li0ne-
Y. Feng, C. Allais, P. Nuhant, M. Z. Chen, B. Pierce and R.
210 nm particles. Prices taken from the Sigma-Aldrich
DOI: 10.1039/D0GC03070H
Robinson, Org. Lett., 2015, 17, 5756–5759; e) K. Hakobyan, T.
catalogue, on 3rd September 2020.
Gegenhuber, C. S. P. McErlean and M. Müllner, Angew. Chem., 21 D. Prat, A. Wells, J. Hayler, H. Sneddon, C. R. McElroy, S. Abou-
Int. Ed., 2019, 58, 1828–1832.
Shehada and P. J. Dunn, Green Chem., 2015, 18, 288–296.
4
For selected reviews of modern photocatalysis and catalyst 22 United States Federal Register, 1995, 60, 31633
development, see: a) C. K. Prier, D. A. Rankic and D. W. C. 23 R. Kacker, S. I. Regensburg and H. J. M. Kramer, Chem. Eng. J.,
MacMillan, Chem. Rev., 2013, 113, 5322–5363; b) X. Deng, Z.
2017, 317, 413–423.
Li and H. García, Chem. - Eur. J., 2017, 23, 11189–11209; c) L. 24 L. N. Ejim, S. Yerdelen, T. McGlone, I. Onyemelukwe, B.
Marzo, S. K. Pagire, O. Reiser and B. König, Angew. Chem., Int.
Ed., 2018, 57, 10034–10072; d) Q. Q. Zhou, Y. Q. Zou, L. Q. Lu
Johnston, A. J. Florence and N. M. Reis, Chem. Eng. J., 2017,
308, 669–682.
and W. J. Xiao, Angew. Chem., Int. Ed., 2019, 58, 1586–1604; 25 For selected examples of perfluoroalkylation reactions, see: a)
e) S. Sharma and A. Sharma, Org. Biomol. Chem., 2019, 17,
4384–440513; f) B. M. Hockin, C. Li, N. Robertson and E.
Zysman-Colman, Catal. Sci. Technol., 2019, 9, 889–915.
For a review of photocatalyzed ATRA reactions, see: T.
Courant and G. Masson, J. Org. Chem., 2016, 81, 6945–6952.
a) I. Saikia, A. J. Borah and P. Phukan, Chem. Rev., 2016, 116,
6837–7042; b) D. Cantillo and C. O. Kappe, React. Chem. Eng.,
2017, 2, 7–19.
For selected reviews of flow photochemistry, see: a) D.
Cambié, C. Bottecchia, N. J. W. Straathof, V. Hessel and T.
Noël, Chem. Rev., 2016, 116, 10276–10341; b) T. H. Rehm, 26 K. Matsuo, E. Yamaguchi and A. Itoh, J. Org. Chem., 2020, 85,
ChemPhotoChem, 2019, 3, 1–21; c) J. D. Williams and C. O.
Kappe, Curr. Opin. Green Sustain. Chem., 2020, 25, 100351; d)
T. Noël, J. Flow Chem., 2017, 7, 87–93; e) J. P. Knowles, L. D.
Elliott and K. I. Booker-Milburn, Beilstein J. Org. Chem., 2012,
8, 2025–2052.
C. J. Wallentin, J. D. Nguyen, P. Finkbeiner and C. R. J.
Stephenson, J. Am. Chem. Soc., 2012, 134, 8875–8884; b) L. L.
Mao and H. Cong, ChemSusChem, 2017, 10, 4461–4464; c) G.
Magagnano, A. Gualandi, M. Marchini, L. Mengozzi, P. Ceroni
and P. G. Cozzi, Chem. Commun., 2017, 53, 1591–1594; d) G.
Wu, J. Börger and A. Jacobi von Wangelin, Angew. Chem., Int.
Ed., 2019, 58, 17241–17245; e) M. Inoue, Y. Sumii and N.
Shibata, ACS Omega, 2020, 5, 10633–10640; b) Y. Zhou, J.
Wang, Z. Gu, S. Wang, W. Zhu, J. L. Acenã, V. A. Soloshonok,
K. Izawa and H. Liu, Chem. Rev., 2016, 116, 422–518.
5
6
7
10574–10583.
8
9
a) J. W. Tucker, Y. Zhang, T. F. Jamison and C. R. J. Stephenson,
Angew. Chem., Int. Ed., 2012, 51, 4144–4147; b) C. Rosso, J. D.
Williams, G. Filippini, M. Prato and C. O. Kappe, Org. Lett.,
2019, 21, 5341–5345.
M. Schoenitz, L. Grundemann, W. Augustin and S. Scholl,
Chem. Commun., 2015, 51, 8213–8228.
10 T. McGlone, N. E. B. Briggs, C. A. Clark, C. J. Brown, J. Sefcik
and A. J. Florence, Org. Process Res. Dev., 2015, 19, 1186–
1202.
11 For a recent review of oscillatory flow reactors, including a
comparison versus other solid-handling methods in flow, see:
P. Bianchi, J. D. Williams and C. O. Kappe, J. Flow Chem., 2020,
10, 475–490.
12 For more information on the HANU reactor, see: http://www.
creaflow.be/
13 For other examples of handling solids using this reactor, see:
a) C. Rosso, S. Gisbertz, J. D. Williams, H. P. L. Gemoets, W.
Debrouwer, B. Pieber and C. O. Kappe, React. Chem. Eng.,
2020, 5, 595–604; b) W. Debrouwer, W. Kimpe, R. Dangreau,
K. Huvaere, H. P. L. Gemoets, M. Mottaghi, S. Kuhn and K. van
Aken, Org. Process Res. Dev., 2020, 24, DOI:
10.1021/acs.oprd.0c00150.
14 Z. Wen, A. Maheshwari, C. Sambiagio, Y. Deng, G. Laudadio, K.
Van Aken, Y. Sun, H. P. L. Gemoets and T. Noël, Org. Process
Res. Dev.,2020, 24, DOI: 10.1021/acs.oprd.0c00235.
15 P. Patnaik, Handbook of Inorganic Chemicals, McGraw-Hill,
2003, New York.
16 a) S. Wu, Y. Xie, X. Zhang, Z. Huang, Y. Liu, M. Fang, X. Wu and
X. Min, J. Mater. Res., 2019, 1–12; b) A. Han, J. Sun, G. K.
Chuah and S. Jaenicke, RSC Adv., 2017, 7, 145–152; c) M. N.
Novokreshchenova, T. V. Daminova and Y. M. Yukhin, Russ. J.
Appl. Chem., 2007, 80, 2015–2019.
17 E. Arceo, E. Montroni and P. Melchiorre, Angew. Chem., Int.
Ed., 2014, 53, 12064–12068.
18 X. Meng and Z. Zhang, J. Mol. Catal. A: Chem., 2016, 423, 533–
549.
19 a) L. Leontie, M. Caraman, M. Alexe and C. Harnagea, Surf. Sci.,
2002, 507–510, 480–485; b) S. Anandan and J. J. Wu, Mater.
Lett., 2009, 63, 2387–2389.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 9
Green Chemistry
View Article Online
DOI: 10.1039/D0GC03070H
An oscillatory flow photoreactor facilitates the use of Bi2O3 as a green and recyclable photocatalyst
for ATRA reactions.