Finding the Perfect Match: A Combined Computational and Experimental Study toward Efficient and Scalable Photosensitized [2 + 2] Cycloadditions in Flow

Article abstract of DOI:10.1021/acs.oprd.8b00375

With ever-evolving light-emitting diode (LED) technology, classical photochemical transformations are becoming accessible with more efficient and industrially viable light sources. In combination with a triplet sensitizer, we report the detailed exploration of [2 + 2] cycloadditions, in flow, of various maleic anhydride derivatives with gaseous ethylene. By the use of a flow reactor capable of gas handling and LED wavelength/power screening, an in-depth optimization of these reactions was carried out. In particular, we highlight the importance of matching the substrate and sensitizer triplet energies alongside the light source emission wavelength and power. Initial triplet-sensitized reactions of maleic anhydride were hampered by benzophenone's poor absorbance at 375 nm. However, density functional theory (DFT) calculations predicted that derivatives such as citraconic anhydride have low enough triplet energies to undergo triplet transfer from thioxanthone, whose absorbance matches the LED emission at 375 nm. This observation held true experimentally, allowing optimization and further exemplification in a larger-scale reactor, whereby >100 g of material was processed in 10 h. These straightforward DFT calculations were also applied to a number of other substrates and showed a good correlation with experimental data, implying that their use can be a powerful strategy in targeted reaction optimization for future substrates.

Full text of DOI:10.1021/acs.oprd.8b00375

Subscriber access provided by University of South Dakota
Full Paper
Finding the Perfect Match: A Combined Computational
and Experimental Study Towards Efficient and
Scalable Photosensitized [2+2] Cycloadditions in Flow
Jason Williams, Momoe Nakano, Romaric Gérardy, Juan A. Rincon, Oscar
de Frutos, Carlos Mateos, Jean-Christophe M Monbaliu, and C. Oliver Kappe
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00375 • Publication Date (Web): 27 Dec 2018
Downloaded from http://pubs.acs.org on December 27, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Finding the Perfect Match: A Combined Computational and
Experimental Study Towards Efficient and Scalable Photosensitized
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[2+2] Cycloadditions in Flow
†
,‡
‡⊥
§
∥
Jason D. Williams, Momoe Nakano, Romaric Gérardy, Juan A. Rincón, Óscar
∥
∥
§
*,†,‡
de Frutos, Carlos Mateos, Jean-Christophe M. Monbaliu, C. Oliver Kappe
†
Center for Continuous Flow Synthesis and Processing (CC FLOW), Research Center
Pharmaceutical Engineering GmbH (RCPE), Inffeldgasse 13, 8010 Graz, Austria
‡
Institute of Chemistry, University of Graz, NAWI Graz, Heinrichstrasse 28, 8010 Graz, Austria
§
Center for Integrated Technology and Organic Synthesis, Research Unit MolSys, University of
Liège, B-4000 Liège (Sart Tilman), Belgium
∥
Centro de Investigación Lilly S.A., Avda. de la Industria 30, 28108 Alcobendas-Madrid, Spain
*
Corresponding author. E-mail: oliver.kappe@uni-graz.at
⊥
Present address: Division of Materials Science, Graduate School of Science and Technology,
Nara Institute of Science and Technology (NAIST), 8916-5 Takayamacho, Ikoma, Nara, 630-
101, Japan
0
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
TABLE OF CONTENTS ENTRY
Triplet
sensitizer
Substrate
O
O
O
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
S
Well matched triplet
energies
Efficient and
scalable [2+2]
cycloaddition
Optimal light source
narrow band LEDs)
(
(DFT calculations)
ACS Paragon Plus Environment
2

Page 3 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ABSTRACT
With ever-evolving LED technology, classical photochemical transformations are becoming
accessible with more efficient and industrially viable light sources. In combination with a triplet
sensitizer, we report the detailed exploration of [2+2] cycloadditions, in flow, of various maleic
anhydride derivatives with gaseous ethylene. Using a flow reactor capable of gas handling and
LED wavelength/power screening, an in-depth optimization of these reactions was carried out. In
particular, we highlight the importance of matching substrate and sensitizer triplet energies,
alongside light source emission wavelength and power. Initial triplet sensitized reactions of maleic
anhydride were hampered by benzophenone’s poor absorbance at 375 nm. However, DFT
calculations predicted that derivatives such as citraconic anhydride have low enough triplet
energies to undergo triplet transfer from thioxanthone, whose absorbance matches LED emission
at 375 nm. This observation held true experimentally, allowing optimization and further
exemplification in a larger scale reactor, whereby >100 g of material was processed in 10 hours.
These straightforward DFT calculations were also applied to a number of other substrates, showing
a good correlation with experimental data, implying a powerful strategy in targeted reaction
optimization for future substrates.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KEYWORDS flow photochemistry; DFT calculations; cycloadditions; photosensitizers; gas-
liquid reactions
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
A recent surge in efforts towards new photochemical methodologies for organic synthesis1
has also brought about renewed interest in “classical” photochemical transformations, such as the
2
synthesis of cyclobutanes by [2+2] cycloaddition. Although photochemical [2+2] cycloadditions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
a
have been continually studied since their discovery in 1877, their applications have remained
almost exclusively academic. In recent years, though, the modern “Escape from Flatland” era of
3
medicinal chemistry has driven a paradigm shift, favoring sp -rich compounds with an increased
3
number of chiral centers and more three-dimensional complexity. Within this framework,
building blocks with distinct and varied expansion vectors are highly desirable, which has driven
4
5
synthetic research in synthesis of compounds such as cubanes and bicyclo[1.1.1]pentanes. The
cyclobutane moiety, however, also fulfils these prerequisites, as a saturated linking motif with
multiple distinct expansion vectors. In contrast to more complex building blocks, cyclobutanes
may be accessed by means of [2+2] cycloaddition; a route which instils modularity, or adds
complexity, depending upon the degree of starting material functionalization.
Maleic anhydride (1) is a cheap and widely available bulk chemical (EUR13.40 per kg),6
which has been demonstrated to undergo photochemical [2+2] cycloaddition with ethylene,
7
,8
furnishing a cyclobutane-based cyclic anhydride. This 5-membered anhydride can be opened (in
a symmetric or asymmetric fashion), revealing two carboxylic acid moieties, with many possible
modes of further functionalization such as reductions, esterification or amidation. Furthermore,
many recent cross-coupling methodologies have begun to invoke carboxylic acid-derived
9
electrophiles, making maleic anhydride derivatives versatile cyclobutane synthons, towards a vast
2b
range of diversely functionalized molecules (Scheme 1). Synthesis of these building blocks on
small scale may be readily achieved in medicinal chemistry labs using modern synthetic
photochemistry equipment, yet performing the same chemistry on larger scales is expected to
become problematic due to issues with light penetration and gas-liquid mixing.
ACS Paragon Plus Environment
4

Page 5 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
CONHR1
r
e
st
/
e
CO R2
2
amide
formation
O
CO H
R3
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1) h, cycloaddition
C-C couplings
O
2) ring opening/
CO H
R4
2
desymmetrization
O
1
reductio
HO
functiona
cheap starting
material
Versatile building blocks
n/alcohol
lization
High degree of saturation
Novel chemical space
5
R O
Scheme 1. Depiction of possible uses for maleic anhydride-derived cyclobutanes in API synthesis.
Implementation of flow technology has been identified as the most practical solution to
10
performing photochemical reactions in a scalable manner and, accordingly, there are many
examples of photochemical [2+2] reactions performed in flow.8,11,12 These include very few
1
2
examples of large scale processing, and almost all employ UV light sources such as pressurized
mercury lamps. These light sources suffer from numerous drawbacks that make them unsuitable
for large scale applications. High energy consumption coupled with low efficiency results in an
immense heat output, requiring substantial cooling. Furthermore, the lifetime of these lamps is
unpredictable, and relatively short, requiring regular replacement and hence, material “sent to
waste” in a manufacturing setting. Finally, these lamps consist of multiple discrete energy bands
across the entire UV-visible spectrum, exacerbating energy wastage by unused light and
introducing issues with reaction selectivity, particularly in molecules containing delicate
functionalities, such as APIs (active pharmaceutical ingredients).10a,b LED light sources offer an
excellent alternative; incorporating improved efficiency, predictable performance over long
1
3
lifetimes and a narrow pseudo-monochromatic light output. In addition, LED technology is
developing at a fast rate, implying that their use will continue to become more widespread within
1
4
all areas of synthetic photochemistry. Indeed, the use of LEDs in place of higher energy (or other
15
broad-emission light sources) has already been reported in numerous cases and this trend is ever
increasing.13
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The main drawback of LEDs in synthetic photochemistry is the fact that those with
synthetically useful light output are currently limited to wavelengths >340 nm. Therefore, in order
to carry out most cycloadditions, an added photosensitizer is required, due to the short absorption
wavelength of most organic substrates. However, this targeted excitation of a photocatalytic
species can also allow for more efficient reactions under milder conditions. Since triplet
photosensitizers typically have high attenuation coefficients (efficient light absorption) and an
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
6
intersystem crossing (ISC) yield close to unity, even a small quantity of sensitizer can enhance
the rate at which the reaction substrate reaches its required excited state, when compared to direct
irradiation. Triplet-sensitized organic transformations have seen significant interest in recent years,
as alternative or complimentary methods to the more common single electron transfer (SET)
1
7
photoredox catalysis. It is also envisioned that computational methods may be used to assist in
identifying suitable triplet sensitizers to be paired with different reaction substrates through simple
1
8
comparison of triplet energies. Therein, it may be possible for a powerful predictive method to
be developed, allowing expeditious reaction optimization.
Within our interests to develop and showcase scalable flow photochemistry towards API
synthesis, the formation of cyclobutanes is of significant interest. To our knowledge, scale-up of
8
photosensitized cycloadditions has very little literature precedence thus far, and so a detailed
study into this type of synthetic methodology would prove to be of interest to process development
chemists within the pharmaceutical and fine chemical industries. Ethylene is an excellent choice
as an alkene cycloaddition partner to further explore this reaction, since it acts as a reactive and
cheap route into simple cyclobutanes, but also further exemplifies the benefits of performing gas-
liquid reactions under flow conditions. Inspired by recent work examining the scalability of the
8
cycloaddition of maleic anhydride (1) with ethylene, using ketone photosensitizers, we opted to
undertake a broader study of this class of reactions. The aforementioned study initially employed
a medium pressure Hg lamp, but later moved on to LED irradiation to achieve a more energy-
efficient and industrially-applicable transformation, assumedly for the reasons mentioned above.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment
6

Page 7 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Optimization of Maleic Anhydride Cycloaddition
Our studies using the same substrate began using a commercially available plate based
1
9
photoreactor (Corning Advanced-Flow Lab Photo Reactor). This reactor implements two LED
panels to irradiate the reaction plate and uses integrated pumps, mass flow controller and back
pressure regulator, along with temperature regulators for both the reaction plate and LED panels
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Scheme 2). Initial reactions were performed using 1-hexene as a model liquid olefin, in order to
explore reaction parameters, prior to involving a gaseous reagent feed. These studies showed that
temperature had no significant effect on reaction rate and that almost complete reaction could be
achieved in a 10 min residence time, using an excess of 1-hexene (4 equiv) and an elevated loading
of benzophenone (see Supporting Information, Table S5-S7 for details of reaction optimization
with 1-hexene). Our attention then shifted to introducing ethylene as a gaseous alkene. In this case
it was expected to be less straightforward to apply such an excess of the alkene partner, since the
concentration in solution is solubility-limited. Nevertheless, it was found that by applying a
moderate pressure of 12 bar to these reactions using the in-built back pressure regulator (Zaiput
BPR-10) the ethylene stream was entirely dissolved in the reaction medium, resulting in simpler
monophasic fluid behavior. Surprisingly, a single phase was observed in all reactions, even those
at elevated temperatures and when using 2.5 equiv of ethylene. Although multiphasic flow regimes
2
0
have been shown to improve productivity in some cases, having all of the input gas dissolved in
the reaction medium allows a longer irradiation time (based on the same flow rates), hence
improving overall productivity (see Supporting Information, Table S8 for details of reaction
optimization with ethylene). Based on the optimized conditions (Scheme 2) a greatly increased
–1
–
space-time-yield was calculated in comparison to the previously reported procedure (301 g L h
1
–1 –1 8
vs 90 g L h ). This can be attributed to the increased light intensity provided by the LED
arrays and smaller reactor size (2.77 mL vs previously reported 20 mL reactor).
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
.5 equiv
0
.625 mL/min
MFC
O
O
O
375 nm
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
+
O
O
O
25 °C
12 bar
0.25 mL/min
3
Ph
Ph
2.77 mL
10.4 min)
2
(
89%
4
0 mol%
01 g L h1

1
3
(vs. 90 g L h , see ref. 8)
1
1
0.5 M in EtOAc
Scheme 2. Flow schematic showing the optimal reaction conditions for the cyclization of maleic
anhydride (1) with ethylene. GC yield is shown in the scheme and an 84% isolated yield was
achieved.
Unfortunately, the limit of productivity for this reaction quickly became apparent and the
optimal conditions (Scheme 2) could not be improved upon further. Our optimal conditions closely
8
reflected those used in the aforementioned publication, with the main deviation being the use of
ambient temperature, which avoids energy costs associated with cooling the reactor. Using a panel
of 375 nm LEDs, very high loadings (40-60%) of the sensitizer benzophenone (2) were required
to achieve a fast rate of reaction, leading to poor atom-economy. Recovery and reuse of the
sensitizer is possible (see Supporting Information Section 1.5.1), yet this purification process
would add multiple unit operations to the process and is undesirable when considering large scale
processing, particularly in recovering a commercially available material. Of particular interest in
this case, was the comparative performance of differing LED wavelengths under these conditions.
The absorption spectrum of benzophenone peaks at ~340 nm (λmaz = 343 nm) and the efficacy of
pseudo-monochromatic LEDs at multiple wavelengths was examined (Table 1). LEDs with a
maximum emission at 343 nm were found to be ineffective in this transformation (Table 1, entry
1
), likely due to their poor photon flux, which is substantially lower than that of other wavelengths
examined. As noted previously, the development of UV LEDs is a rapidly advancing area of
14
research and as a result it is expected that this power will improve in the coming years. LEDs of
365 nm and 375 nm showed roughly equal yield, despite benzophenone’s significantly higher
ACS Paragon Plus Environment
8

Page 9 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
absorbance at 365 nm. This higher absorbance at 365 nm may bring the reaction into a regime
where all light is absorbed before penetrating the entire reaction medium, hence showing only
limited improvement. As expected, irradiation at 385 nm shows very little conversion and 405 nm
is completely outside of benzophenone’s absorbance range, so no reaction was observed.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Results of Benzophenone-sensitized Reaction Protocol at varying LED Wavelengths,
Including a Graphical Representation of Benzophenone Absorption Spectrum with the
Examined LED Wavelengths Appended.
Entry
Wavelength [nm]a
Radiant flux
Photon flux
[mmol h ]
Yield 3
[%]
b
–1
c
[W]
1
2
3
4
5
343
365
375
385
405
2.9
30
14
90
89
48
43.6
46.4
53.6
56.8
477
522
619
690
0
a
The UV/vis spectrum was measured in ethyl acetate at a concentration of 12.5 mM. Wavelengths
listed are the maximum output wavelength, as taken from the LED specification sheet; the actual
b
value may vary by ±5 nm. Radiant flux values were obtained from the LED specification sheets
for a single LED, then multiplied by 40 (the total number of LEDs irradiating the reactor plate).
Note: Two separate panels were used for various experiments in this study (panel 1 = 340, 375,
3
95, 422, 450, 540 nm; panel 2 = 365, 385, 405, 485, 610 nm and “4000K” white light; see
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
c
Supporting Information Figure S2 and Figure S3 for emission spectra). Yield was calculated by
the ratio of starting material to product by GC, since only one product was observed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Since the reaction performance could not be improved by selecting a different LED
wavelength, six alternative photosensitizers were examined, in attempt to achieve better overlap
with the available LED output (Table 2). None gave a comparable level of conversion, with only
a small extent of reaction observed with 4,4’-dimethoxybenzophenone (entry 4) and thioxanthone
(entry 6). This can easily be explained by the poor spectral overlap (when a sensitizer’s triplet
energy is high), or by insufficient triplet energy (when good spectral overlap is achieved). In order
to enable a predictive method of matching substrates and sensitizers, the triplet energies of these
sensitizers were obtained using DFT calculations and compared to literature values. A screen of
methods and basis sets revealed M06-2X at a TZVP level of theory to be most suitable, whilst still
allowing these calculations to be completed quickly on a standard desktop computer.21
2
2
Experimental determination of triplet energies requires specialized equipment, so these data are
not readily available. Accordingly, this method of comparison will be especially beneficial,
particularly for reaction substrates and more exotic sensitizers. The trend between sensitizer triplet
energy and absorption spectrum can be seen clearly and, as would be expected, higher triplet
energies require absorption of more energetic (shorter) wavelengths.
ACS Paragon Plus Environment
10

Page 11 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Table 2. A Selection of Triplet Sensitizers, Displaying their Triplet Energies as Found Using
DFT Calculations Compared with Experimental Literature Values, Alongside their
Wavelength of Maximum Absorption (λmax), and Extent of Reaction of Maleic Anhydride (1)
with Ethylene.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Entry
Sensitizer
Calculated ET
Experimental
λmax
[nm]
Experimental
yield 3 [%]
1
–1 23
–1
a
[kcal mol ]
ET [kcal mol ]
1
1
2
3
4
5
6
7
Xanthone (4)
Acetophenone (5)
Benzaldehyde (6)
74.6
72.4
70.7
72.7
69.2
65.5
54.8
74.222a
73.622a
71.922a
69.415c
68.522a
65.522a
53.322a
338
318
325
335
343
377
397
0
0
0
4,4’-Dimethoxy
benzophenone (7)
4
Benzophenone (2)
Thioxanthone (8)
Fluorenone (9)
12
4
0
Reaction conditions: 0.05 equiv sensitizer loading, 0.5 M concentration in EtOAc, 0.5 mL/min
a
substrate flowrate, 1.25 mL/min ethylene flowrate. Yield was calculated by the ratio of starting
material to product by GC, since only one product was observed.
This observation appears to limit the synthetic application of light sources emitting at
wavelengths >340 nm, since longer wavelengths only allow lower energy triplet states to be
achieved. From the obtained UV-vis absorption spectra (Table 2), thioxanthone (8) (entry 6)24
appears to provide a good overlap with the output of several available LEDs (365, 375, 385 nm),
with an appreciably high triplet energy. Furthermore, multiple recent publications have found
thioxanthone and its derivatives to be effective triplet sensitizers, at low loadings, for a range of
2
5
26
27
transformations including intra- and intermolecular [2+2] cycloadditions, halogenations and
28
radical polymerizations. Accordingly, further DFT calculations were undertaken, in order to
guide the choice of substrate towards one which might undergo effective triplet transfer with
thioxanthone (8). The addition of a single methyl group to the backbone of maleic anhydride (1)
was calculated to reduce the triplet energy by 3.2 kcal/mol, which prompted experimental
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
examination of its performance in this [2+2] cycloaddition reaction. Pleasingly, the observation
held true experimentally and complete conversion to the cyclobutane product was observed within
a short residence time, using a significantly lower sensitizer loading (Scheme 3Error! Reference
source not found.). The lower sensitizer loading can be explained by comparing the extinction
coefficients of both sensitizers, since poorer light absorption will ultimately result in a lower triplet
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
1
–1
state population. Benzophenone (2) has a very low coefficient of 34 M cm , whilst that of
–
1
–1
29
thioxanthone (8) is 6265 M cm (roughly 200 times higher), meaning that a far lower sensitizer
concentration is required in the reaction mixture to absorb all of the incident light.
O
Unproductive using
O
thioxanthone
Extinction coefficient at 375 nm:
O
benzophenone 2 = 34 M-1 cm-1
thioxanthone 8 = 6265 M-1 cm-1
1
E = 65.8 kcal/mol
T
O
O
O
O
O
S
8
(2.5 mol%)
O
375 nm LEDs
5 °C, 12 bar
2.5 equiv
2
O
EtOAc
0.5 M)
1
0
11
(
E = 62.6 kcal/mol
T
0.5 mL/min
1.25 mL/min
>
99% conversion
space-time-yield = 759 g L-1 h-1
estimated residence time = 5 min 12 sec
Scheme 3. Modification of starting material from maleic anhydride (1) to citraconic anhydride
10), and accompanying improvement in reaction performance, when using a low loading of
(
thioxanthone (8) as a photosensitizer.
A detailed examination of the reaction parameters was carried out (Table 3), following
complete conversion observed at 0.5 mL/min (entry 1). The residence time could not be reduced
any further, as shown by the incomplete conversion observed at higher flow rates (entry 2 and 3),
ACS Paragon Plus Environment
12

Page 13 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
and use of acetone as an alternative solvent resulted in no change (entry 4). Increasing the
concentration was attempted to improve productivity, but was observed to reduce conversion
(entry 5), thought to be due to the shorter residence time caused by the increased ethylene flow
rate. Accordingly, the ethylene flow rate was reduced to provide only 1.5 equiv, yet this further
reduced the reaction yield (entry 6). A lower photosensitizer loading gave a slightly reduced yield
(entry 7 vs entry 2), yet a higher loading also gave a decrease in yield (entry 8 vs entry 2), implying
that complete light absorbance within this channel depth (0.4 mm) is achieved between 2.5-5 mol%
sensitizer loading (12.5 to 25 mM concentration). Furthermore, the same quantity of thioxanthone
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
was observed by GC before and after the reaction, so no decomposition or photobleaching
occurred under the reaction conditions. Direct translation to the more widely available 385 nm
LEDs showed a slight decrease in yield (entry 9), so a lower flow rate of 0.45 mL/min was
necessitated to achieve full conversion (entry 10). To test the influence of incident oxygen upon
the reaction, it was performed without degassing the reaction mixture (entry 11). The resulting
lack of any reaction (all starting material remained unchanged) clearly demonstrates the
importance of excluding oxygen from these reactions; assumedly due to competitive deactivation
of the excited thioxanthone species. These studies led to the optimal conditions for this
–
1
–1
transformation, providing an exceptional space-time-yield of 759 g L h (productivity of 2.1 g
–
1
h ) with complete starting material conversion.
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. Optimization of the [2+2] Photocycloaddition of Citraconic Anhydride (10) with
Ethylene.
Entry Reaction Sensitizer Solution Ethylene Residence Irradiation Yield Productivity Space-
–1
conc.
loading
[mol%] [mL/min] [mL/min]
equiv
flow rate flow rate
time
wavelength
[nm]
11
[g h ]
time-
yield
[g h–1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[M]
[min]a
[%]b
/
–
1
L ]
1
2
3
4c
5
6
7
8
9
0.5
0.5
0.5
0.5
0.6
0.6
0.5
0.5
0.5
0.5
0.5
2.5
2.5
2.5
2.5
2.5
2.5
1.25
5
0.5
1
1.25 / 2.5
2.5 / 2.5
1.5 / 2.5
2.5 / 2.5
1.5 / 2.5
0.9 / 1.5
2.5 / 2.5
1.5 / 1.5
1.25 / 2.5
1.13 / 2.5
1.13 / 2.5
5.2
2.6
4.3
2.6
5.2
5.2
2.6
2.6
5.2
5.7
5.7
375
375
375
375
375
375
375
375
385
385
385
>99
69
97
69
96
87
59
64
95
>99
0
2.10
2.90
2.45
2.90
2.42
2.19
2.48
2.69
2.02
1.89
-
759
1 047
883
1 047
874
792
896
972
729
683
-
0.6
1
0.5
0.5
1
1
2.5
2.5
2.5
0.5
0.45
0.45
1
1
0
1d
aResidence times are estimated because they cannot be accurately defined, since the change in
b
volume upon dissolution and consumption of ethylene is complex. Yield was calculated by the
c
ratio of starting material to product by GC, since only one product was observed. Acetone was
d
used as reaction solvent Reaction was run without prior degassing.
ACS Paragon Plus Environment
14

Page 15 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Application to Other Substrates
Having previously observed an excellent correlation between calculated and experimental
triplet energies of triplet sensitizers (Table 2), it was decided that a useful expansion of this
computational system would be to include reaction substrates for comparison and choice of
sensitizer. In this manner a sound estimate can be made, of which triplet sensitizer is suitable to
undergo triplet transfer towards a chosen substrate. Following this strategy could have a significant
impact upon the time and resource spent optimizing a [2+2] cycloaddition (or other triplet
sensitized reaction). Furthermore, an estimate of the most applicable triplet sensitizers will help
guide the choice of light source, potentially limiting wasted investment in poorly-matched
equipment. A range of substrates known to undergo [2+2] cycloadditions were selected, and their
triplet energies calculated by DFT, by the same method and basis set used previously (M06-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
X/TZVP). Subsequently, their reactivity was examined using two sets of conditions, similar to
those described in the benzophenone-sensitized reaction of maleic anhydride (1) and the
thioxanthone-sensitized reaction of citraconic anhydride (10) (Table 4). Due to the low extinction
coefficient of benzophenone (2) at 375 nm (see discussion above), a far higher loading (40 mol%
vs 2.5 mol%) was used, as decided upon in previous optimization studies (see Supporting
Information Table S8).
Of the substrates examined, maleimides 12 and 13 (entry 1 and 2) had the lowest triplet
energies, and were both observed to undergo complete conversion under both sets of reaction
conditions. The triplet energies of N-phenylmaleimide was also calculated, but was found to be
insoluble under the standard conditions, so could not be examined experimentally. Citraconic
anhydride (10) (entry 3) was found to have a very similar triplet energy to maleimide (13) and was
already been shown to react smoothly under sensitization by thioxanthone (8), but also undergoes
effective cycloaddition under benzophenone-sensitized conditions. Isophorene (14) (entry 4) was
calculated to have a significantly higher triplet energy of 72.1 kcal/mol, yet undergoes a reasonable
extent of reaction (52% conversion) with thioxanthone sensitization. This result implies that the
calculated triplet energy in this case could be significantly erroneous, particularly when
considering the lower triplet energy of cyclohexanone (15) (entry 5). Based on previous trends and
the observed experimental results, the inclusion of a methyl group at the enone carbon should
lower the triplet energy. Some reaction of cyclohexanone (15) was still observed using
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
thioxanthone, but far lower than the corresponding benzophenone-sensitized reaction. Finally, the
triplet energy of maleic anhydride (1) was too higher for effective reaction with thioxanthone, but,
as discussed earlier, underwent productive reaction with benzophenone.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
16

Page 17 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Table 4. Calculated Triplet energies of Selected Substrates Compared with their Reaction
Performance, Using Thioxanthone (2.5 mol%) or Benzophenone (40 mol%) as the
Photosensitizer
Entry
Substrate
Calculated E
Conversion with
Conversion with
T1
1 a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
[kcal mol ]
thioxanthone 8
benzophenone 2
b
c
(ET = 65.5) [%]
(E = 68.5) [%]
T
1
O
N
60.0
>99
>99
>99
>99
O
O
Bn
1
2
2
3
62.5
62.6
>99
>99
O
N
H
1
3
O
O
O
O
10
4
72.1d
52
84
14
5
6
65.2
65.8
12
4
72
89
O
15
O
O
O
1
aTriplet energies calculated by DFT: M06-2X/TZVP. Conditions: LED irradiation at 375 nm,
bthioxanthone conditions: 2.5 mol% sensitizer, 0.5 mL/min solution flow rate, 1.25 mL/min (2.5
c
equiv) gas flow rate, residence time ≈ 5.2 min. benzophenone conditions: 40 mol% sensitizer,
0
.25 mL/min solution flow rate, 0.625 mL/min (2.5 equiv) gas flow rate, residence time ≈ 10.4
d
min. Calculated triplet energy is likely erroneous, particularly when considered with the
experimental conversion observed and versus entry 5.
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
From the examined substrates, it was observed that maleimides generally have a lower
triplet energy than their anhydride analogues. Accordingly, further optimization using N-
benzylmaleimide (12) was carried out to determine the maximum productivity which could be
achieved with this substrate. Analogous to the example described earlier, multiple LED
wavelengths were examined for their efficacy in this thioxanthone-sensitized reaction (Table 5),
using a faster flowrate to improve the impact of different wavelengths on these incomplete
reactions. 365 nm LEDs displayed a reasonable rate of reaction, yet this was surpassed by both
375 nm and 385 nm. A wavelength of 385 nm provides a poor overlap with catalyst absorbance,
yet the LED power is considerably higher. On the other hand, 375 nm lies close to the catalyst’s
λmax and these LEDs still provide a high radiant flux. Following additional optimization
experiments (see Supporting Information, Table S13) almost complete conversion (98% yield)
was still observed at a minimum residence time of ~2.23 min (1.15 mL/min substrate flow rate,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
1
–1
2
.88 mL/min ethylene flow rate), yielding an impressive space-time-yield of 2285 g L h .
ACS Paragon Plus Environment
18

Page 19 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Table 5. Results of Thioxanthone-sensitized Reaction Protocol at Varying LED Wavelengths
at 1.5 mL/min Flowrate, Including a Graphical Representation of Thioxanthone Absorption
Spectrum with the Examined LED Wavelengths Appended.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Entry
Wavelength [nm]a
Radiant flux
[W]b
Photon flux
[mmol h–1]
Yield 12
[%]c
1
2
3
365
375
385
43.6
46.4
53.6
477
522
619
59
76
66
The UV/vis spectrum was measured in ethyl acetate at a concentration of 0.225 mM. Conditions:
.5 mol% thioxanthone, 1.5 mL/min solution flow rate, 3.75 mL/min (2.5 equiv) gas flow rate,
2
a
residence time ≈ 1.7 min. Wavelengths listed are the target maximum output wavelength, as
taken from the LED specification sheet; the actual value may vary by ±5 nm. Radiant flux
b
values were obtained from the LED specification sheets for a single LED, then multiplied by 40
(
the total number of LEDs irradiating the reactor plate). Note: Two separate panels were used for
various experiments in this study (panel 1 = 340, 375, 395, 422, 450, 540 nm; panel 2 = 365,
85, 405, 485, 610 nm and “4000K” white light; see Supporting Information Figure S2 and
3
c
Figure S3 for emission spectra). Yield was calculated by the ratio of starting material to product
by GC, since only one product was observed.
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scale-up
With the optimal conditions for the cycloaddition of citraconic anhydride (10) in hand
(Table 3), our interest turned to the scalability of this system, due to its improved commercial
availability versus maleimides. Flow chemistry is known to be an enabling technology for a wide
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
0
variety of chemical transformations, as exemplified in numerous cases. However, the use of
photochemistry introduces further complexity, and these processes have only scarcely been
demonstrated on industrially relevant scales. The Corning Advanced-FlowTM Lab Photo Reactor,
utilized in these small scale trials, has an irradiated volume of only 2.77 mL. In order to produce
material to satisfy demands, for use in API synthesis for example, a system with a larger irradiated
volume must be introduced to implement faster flow rates. To tackle this issue, the Corning G1
Photo Reactor suggests to provide seamless scale up from the “Lab” model, by allowing the
19
placement of multiple plates in series. For initial scale-up endeavors, five “low flow” plates (2.77
mL volume each, as used in the optimization experiments) were connected in series, essentially
elongating the reactor to give a total irradiated volume of 13.85 mL (Scheme 4, Figure 1).
Figure 1. Reaction setup for larger scale demonstration, using a Corning G1 Photo Reactor, with
5
× 2.77 mL reaction plates installed.
The reaction conditions from the lab scale optimization were maintained by simply scaling
the flow rates by a factor of 5 for the G1 experiments (2.5 mL/min), providing a consistent
irradiated residence time of ~5.5 min (an estimate, due to changes in volume caused by dissolution
ACS Paragon Plus Environment
20

Page 21 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
and consumption of ethylene gas along the reaction pathway). Gratifyingly, the reaction performed
to an identical standard when compared with the lab scale results. Complete conversion of the
starting material was observed in all of the eleven analytical samples, taken at increments across a
1
0 h runtime (see Supporting Information, Figure S5). In this time, 1.5 L of reaction mixture was
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
processed, resulting in 101.3 g (96%) of the desired product, after evaporation of solvent. Aside
from the added triplet sensitizer (2.5 mol%), this material was found to be pure by NMR analysis.
6.25 mL/min
O
MFC
O
O
LED irradiation (375 nm)
0
.5 M in EtOAc
O
O
O
+
O
2.5 mL/min
12 bar
2.77 mL
25 °C 4 x 2.77 mL
96%, 101.3 g
processed in 10 h
S
2.5 mol%
Scheme 4. Flow schematic showing the conditions used for an extended run in the Corning G1
Photo Reactor.
3. Conclusion
We have demonstrated an in-depth study of photochemical [2+2] cycloadditions, using
simple in silico calculations to reinforce experimental observations and guide reaction
optimization. Initial optimization work using maleic anhydride as a reaction substrate had only
limited success, which was caused by a poor overlap of LED output with sensitizer
(benzophenone) absorbance. The relatively high triplet energy of maleic anhydride means that
other sensitizers were ineffective in triplet transfer, resulting in no reaction – an effect corroborated
by DFT-calculated triplet energies. Modification of the reaction substrate was shown to have a
dramatic effect on the calculated triplet energy. It was found that addition of a methyl group to the
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
substrate (citraconic anhydride) reduced its triplet energy by 3.2 kcal/mol. Subsequently, this
brings the substrate into the range of triplet transfer from thioxanthone, whose absorbance (λmax
77 nm) overlaps well with available LEDs. This effective overlap allows the sensitizer loading to
be lowered to 2.5 mol%, giving complete conversion at higher flowrates up to 0.5 mL/min.
=
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The triplet energies of additional substrates were also examined using DFT calculations,
and their reactivities compared with the two distinct sensitizer systems. In general, these calculated
energies correlated well with the experimental conversion observed, implying that using a
straightforward DFT method (with a very low computational cost) could be very beneficial in the
design of this type of chemical system, particularly in the choice of triplet sensitizer. Implementing
such computationally-guided sensitizer selection has the potential to speed up development of
chemical processes and save a great deal of resource – both material and FTE time. Finally, scale-
up of the citraconic anhydride cycloaddition was achieved by simply running the reaction mixture
through 5 × 2.77 mL reactor plates in series, using a Corning G1 Photo Reactor. This strategy
enabled production of >100 g of material over 10 h, demonstrating a stable system over a long
period of time.
EXPERIMENTAL SECTION
General Procedure for Small Scale [2+2] Cycloadditions Lab Photo Reactor.
Using a Zaiput back pressure regulator, a pressure of 12 bar was set. The array of 40 × 375 nm
LEDs was switched on and its thermal fluid was set to 15 °C. The reaction plate thermostat was
set to 25 °C. A solution of citraconic anhydride (10) (280 mg, 2.5 mmol) and thioxanthone (8)
(13.3 mg, 0.0625 mmol) was made up in degassed ethyl acetate (5 mL total volume in a volumetric
flask, solvent was degassed prior to use by sparging with argon for 15 min) and loaded into a 5
mL injection loop (with an Upchurch 6-way valve, see Supporting Information for further details).
Ethyl acetate was pumped through the 2.77 mL reaction plate at a flowrate of 0.5 mL/min.
Simultaneously, ethylene was introduced from a pressurized cylinder, via a Bronkhorst mass flow
controller, at a flowrate of 15 mL/min (2.5 equiv, roughly 1.25 mL/min at 12 bar). After allowing
the reactor to equilibrate for roughly 15 min, the reaction mixture was injected from the sample
loop. The central 1 mL portion of this sample was collected for analysis by gas chromatography.
ACS Paragon Plus Environment
22

Page 23 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Long Run Synthesis of 3-Oxabicyclo[3.2.0]heptane-2,4-dione (3) in Lab Photo Reactor,
Scheme 2:
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Using a Zaiput backpressure regulator, a pressure of 12 bar was set. The 40 LED array was
switched on at 375 nm and its thermal fluid was set to 15 °C. The reaction plate thermostat was
set to 25 °C. A 0.5 M solution of maleic anhydride (1) (2.45 g, 25 mmol) and benzophenone (2)
(2.73 g, 15 mmol) was made up in degassed ethyl acetate (~50 mL) (solvent was degassed prior to
use by sparging with argon for 15 min). Reaction solution was pumped through the 2.77 mL
reaction plate at 0.2 mL/min. Simultaneously, ethylene was introduced from a pressurized
cylinder, via a Bronkhorst mass flow controller, at 0.5 mL/min. After allowing the reactor to
equilibrate for roughly 15 min, the processed reaction mixture was collected for 175 min,
amounting to a total collected volume of 35 mL. From the resulting solution, solvent was removed
in vacuo, yielding a white solid. To this solid was added cyclohexane (5 mL), and the solid
collected again by filtration, then washed with further cyclohexane (5 × 5 mL). The liquor was
evaporated in vacuo, yielding recovered benzophenone (2) (1.75g, 92%). The filtered product was
8
identified as 3-oxabicyclo[3.2.0]heptane-2,4-dione (3) as an off-white solid (1.86 g, 84% yield,
1
adjusted for 5.5% benzophenone content by NMR). H NMR (300 MHz, CDCl ) δ 3.53–3.43 (m,
3
1
3
2
H), 2.79–2.64 (m, 2H), 2.44–2.28 (m, 2H) ppm. C NMR (75 MHz, CDCl ) δ 173.6, 38.9, 23.3
3
ppm.
Synthesis of 3-Benzyl-3-azabicyclo[3.2.0]heptane-2,4-dione (16) in Lab Photo Reactor, Table
4
, entry 1:
Using a Zaiput back pressure regulator, a pressure of 12 bar was set. The array of 40 × 375 nm
LEDs was switched on and its thermal fluid was set to 15 °C. The reaction plate thermostat was
set to 25 °C. A solution of N-benzylmaleimide (12) (468 mg, 2.5 mmol) and thioxanthone (8) (13.3
mg, 0.0625 mmol) was made up in degassed ethyl acetate (5 mL total volume in a volumetric flask,
solvent was degassed prior to use by sparging with argon for 15 min) and loaded into a 5 mL
injection loop (with an Upchurch 6-way valve, see Supporting Information for further details).
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ethyl acetate was pumped through the 2.77 mL reaction plate at a flowrate of 0.5 mL/min.
Simultaneously, ethylene was introduced from a pressurized cylinder, via a Bronkhorst mass flow
controller, at a flowrate of 15 mL/min (2.5 equiv, roughly 1.25 mL/min at 12 bar). After allowing
the reactor to equilibrate for roughly 15 min, the reaction mixture was injected from the sample
loop. 10 mL of reaction mixture was collected, to take into account for forward and backward
diffusion within the reactor. Solvent was removed in vacuo to afford a yellow oil, which was
purified by column chromatography; elution gradient 0-50% ethyl acetate in cyclohexane. The
appropriate fractions were combined and solvent removed in vacuo to afford 3-benzyl-3-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
d
1
azabicyclo[3.2.0]heptane-2,4-dione (16) as a pale yellow solid (512 mg, 95% yield). H NMR
(
300 MHz, CDCl ) δ = 7.41–7.37 (m, 2H), 7.35–7.26 (m, 3H), 4.70 (s, 2H), 3.30–3.25 (m, 2H),
3
1
3
2.66–2.63 (m, 2H), 2.13–2.10 (m, 2H) ppm. C NMR (75 MHz, CDCl ) δ = 179.5, 136.0,
3
128.7, 128.6, 127.9, 42.5, 38.4, 22.9 ppm.
Scale-up Synthesis of 1-Methyl-3-oxabicyclo[3.2.0]heptane-2,4-dione (11) in G1 reactor,
Scheme 4:
Using a Zaiput back pressure regulator, a pressure of 12 bar was set. The arrays of 375 nm LEDs
were switched on and their thermal fluid was set to 10 °C. The reaction plate thermostat was set
to 25 °C. A solution of citraconic anhydride (10) (89.66 g, 800 mmol) and thioxanthone (8) (4.29
g, 20 mmol) was made up in degassed ethyl acetate (1.6 L) (sparged with an argon for 30 min).
This reaction solution was pumped through the 14 mL series of reaction plates (5 × 2.8 mL) at a
flowrate of 2.5 mL/min. Simultaneously, ethylene was introduced from a cylinder, via a BROOKS
mass flow controller, at a flowrate of 74 mL/min (2.5 equiv, roughly 6.2 mL/min at 12 bar). An
equilibration period of 15 min was allowed, then the output solution was collected for 10 hours,
with samples taken for GC analysis roughly every hour. 1.5 L of product solution was collected
and solvent was removed in vacuo, yielding 1-methyl-3-oxabicyclo[3.2.0]heptane-2,4-
31
dione (11) as a pale yellow solid (101.3 g, 96% yield, adjusted for 2.5% thioxanthone
1
content by NMR). H NMR (400 MHz, CDCl ) δ 3.19–3.16 (m, 1H), 2.74–2.69 (m, 1H), 2.59–
3
ACS Paragon Plus Environment
24

Page 25 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
3
2
.56 (m, 1H), 2.34–2.27 (m, 2H), 1.52 (s, 3H) ppm. C NMR (100.6 MHz, CDCl ) δ 176.3, 173.0,
3
45.6, 43.3, 30.3, 20.1, 19.3 ppm.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACKNOWLEDGMENT
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). The authors gratefully
acknowledge Corning SAS for the generous loan of the Corning Advanced-Flow Lab Photo
Reactor used in this study in Graz.
Supporting Information
Further details of optimization experiments, general experimental procedures, equipment details
and setup, NMR and GC data and full details of DFT calculations. This material is available free
of charge via the Internet at http://pubs.acs.org.
REFERENCES
(1)
For reviews on modern photochemical methods: (a) Shaw, M. H.; Twilton, J.; MacMillan,
D. W. C. Photoredox Catalysis in Organic Chemistry. J. Org. Chem. 2016, 81, 6898–6926.
b) Romero, N. A.; Nicewicz, D. A. Organic Photoredox Catalysis. Chem. Rev. 2016, 116,
0075–10166. (c) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Visible Light
(
1
Photoredox Catalysis with Transition Metal Complexes: Applications in Organic Synthesis.
Chem. Rev. 2013, 113, 5322–5363.
(2)
(a) Liebermann, C. Ueber Polythymochinon. Ber. Dtsch. Chem. Ges. 1877, 10, 2177−2179.
(
b) Poplata, S.; Tröster, A.; Zou, Y. Q.; Bach, T. Recent Advances in the Synthesis of
Cyclobutanes by Olefin [2 +2] Photocycloaddition Reactions. Chem. Rev. 2016, 116, 9748–
815. (c) Chen, T.; Barton, L. M.; Lin, Y.; Tsien, J.; Kossler, D.; Bastida, I.; Asai, S.; Bi,
C.; Chen, J. S.; Shan, M.; Fang, H.; Fang, F. G.; Choi, H.; Hawkins, L.; Qin, T.; Baran, P.
S. Building C(sp )-Rich Complexity by Combining Cycloaddition and C–C Cross-Coupling
9
3
Reactions. Nature 2018, 560, 350–354. (d) Poplata, S.; Bach, T. Enantioselective
Intermolecular [2+2] Photocycloaddition Reaction of Cyclic Enones and Its Application in
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a Synthesis of (–)-Grandisol. J. Am. Chem. Soc. 2018, 140, 3228–3231. (e) Skalenko, Y.
A.; Druzhenko, T. V.; Denisenko, A. V.; Samoilenko, M. V.; Dacenko, O. P.; Trofymchuk,
S. A.; Grygorenko, O. O.; Tolmachev, A. A.; Mykhailiuk, P. K. [2+2]-Photocycloaddition
of N-Benzylmaleimide to Alkenes As an Approach to Functional 3-
Azabicyclo[3.2.0]heptanes. J. Org. Chem. 2018, 83, 6275–6289.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(3)
(4)
(a) Lovering, F.; Bikker, J.; Humblet, C. Escape from Flatland: Increasing Saturation as an
Approach to Improving Clinical Success. J. Med. Chem. 2009, 52, 6752–6756. (b)
Lovering, F. Escape from Flatland 2: Complexity and Promiscuity. Med. Chem. Commun.
2
013, 4, 515–519. (c) Barker, A.; Kettle, J. G.; Nowak, T.; Pease, J. E. Expanding Medicinal
Chemistry Space. Drug Discov. Today 2013, 18, 298–304.
(a) Reekie, T. A.; Williams, C. M.; Rendina, L. M.; Kassiou, M. Cubanes in Medicinal
Chemistry. J. Med. Chem. 2018, 10.1021/acs.jmedchem.8b00888. (b) Wlochal, J.; Davies,
R. D. M.; Burton, J. Cubanes in Medicinal Chemistry: Synthesis of Functionalized Building
Blocks. Org. Lett. 2014, 16, 4094–4097.
(5)
Stepan, A. F.; Subramanyam, C.; Efremov, I. V.; Dutra, J. K.; O’Sullivan, T. J.; Dirico, K.
J.; McDonald, W. S.; Won, A.; Dorff, P. H.; Nolan, C. E.; Becker, S. L.; Pusilnik, L. R.;
Riddell, D. R.; Kauffman, G. W.; Kormos, B. L.; Zhang, L.; Lu, Y.; Capetta, S. H.; Green,
M. E.; Karki, K.; Sibley, E.; Atchison, K. P.; Hallgren, A. J.; Oborski, C. E.; Robshaw, A.
E.; Sneed, B.; O’Donnell, C. J. Application of the Bicyclo[1.1.1]Pentane Motif as a
Nonclassical Phenyl Ring Bioisostere in the Design of a Potent and Orally Active γ-
Secretase Inhibitor. J. Med. Chem. 2012, 55, 3414–3424.
(
6)
7)
https://www.sigmaaldrich.com/catalog/product/aldrich/m188 (accessed 27th October
2
018).
(
Selected examples of the photochemical [2+2] cycloaddition of maleic anhydride with
ethylene: (a) Bloomfield, J.J.; Owsley, D. G. Photochemical [2 + 2] Cycloaddition
Reactions at Low Temperatures. Synthesis of Bridgehead Substituted
Bicyclo[n.2.0]dicarboxylates from Maleic Acid Derivatives and Ethylene J. Org. Chem.
1
971, 36, 3768–3773. (b) Faure, S.; Jensen, A. A.; Maurat, V.; Gu, X.; Sagot, E.;
Aitken, D. J.; Bolte, J.; Gefflaut, T.; Bunch, L. Stereoselective Chemoenzymatic
Synthesis of the Four Stereoisomers of L-2-(2-Carboxycyclobutyl)glycine and
Pharmacological Characterization at Human Excitatory Amino Acid Transporter
Subtypes 1, 2, and 3 J. Med. Chem. 2006, 49, 6532–6538.
(8)
Precedence of the photosensitized [2+2] cycloaddition of maleic anhydride with
ethylene in flow: Corcoran, E. B.; Lévesque, F.; McMullen, J. P.; Naber, J. R. Studies
Toward the Scaling of Gas-Liquid Photocycloadditions ChemPhotoChem 2018, 2,
931–937.
ACS Paragon Plus Environment
26

Page 27 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
(
9)
Selected reviews of cross-coupling methodologies using carboxylic acids, or their
derivatives, as starting materials: (a) Jin, Y.; Fu, H. Visible-Light Photoredox
Decarboxylative Couplings. Asian J. Org. Chem. 2017, 6, 368–385. (b) Weaver, J. D.;
Recio, A, III.; Grenning, A. J.; Tunge, J. A. Transition Metal-Catalyzed Decarboxylative
Allylation and Benzylation Reactions. Chem. Rev. 2011, 111, 1846–1913. (c) Murarka, S.
N-(Acyloxy)phthalimides as Redox-Active Esters in Cross-Coupling Reactions. Adv. Synth.
Catal. 2018, 360, 1735–1753. (d) Dzik, W. I.; Lange, P. P.; Gooßen, L. J. Carboxylates as
Sources of Carbon Nucleophiles and Electrophiles: Comparison of Decarboxylative and
Decarbonylative Pathways. Chem. Sci. 2012, 3, 2671–2678. (e) Patra, T.; Maiti, D.
Decarboxylation as the Key Step in C−C Bond-Forming Reactions. Chem. Eur. J. 2017, 23,
7382–7401. (f) Schwarz, J.; König, B. Decarboxylative reactions with and without light – a
comparison. Green Chem. 2018, 20, 323–361. (g) Li, Y.; Ge, L.; Muhammad, M. T.; Bao,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
H. Recent Progress on Radical Decarboxylative Alkylation for Csp -C Bond Formation.
Synthesis 2017, 49, 5263–5284. (h) Gooßen, L. J.; Rodríguez, N.; Gooßen, K. Carboxylic
Acids as Substrates in Homogeneous Catalysis. Angew. Chem. Int. Ed. 2008, 47, 3100–
3
120. (i) Rodríguez, N.; Goossen, L. J. Decarboxylative coupling reactions: a modern
strategy for C–C-Bond formation. Chem. Soc. Rev. 2011, 40, 5030–5048. (j) Cornella, J.;
Larrosa, I. Decarboxylative Carbon-Carbon Bond-Forming Transformations of
(Hetero)Aromatic Carboxylic Acids. Synthesis 2012, 44, 653–676.
(
10) Reviews on photochemistry in flow: (a) Su, Y.; Straathof, N. J. W.; Hessel, V.; Noël, T.
Photochemical Transformations Accelerated in Continuous-Flow Reactors: Basic Concepts
and Applications. Chem. Eur. J. 2014, 20, 10562–10589. (b) Cambié, D.; Bottecchia, C.;
Straathof, N. J. W.; Hessel, V.; Noël, T. Applications of Continuous-Flow Photochemistry
in Organic Synthesis, Material Science, and Water Treatment. Chem. Rev. 2016, 116,
10276–10341. (c) Knowles, J. P.; Elliott, L. D.; Booker-Milburn, K. I. Flow
Photochemistry: Old Light through New Windows. Beilstein J. Org. Chem. 2012, 8, 2025–
2
052. (d) Elliott, L. D.; Knowles, J. P.; Koovits, P. J.; Maskill, K. G.; Ralph, M. J.; Lejeune,
G.; Edwards, L. J.; Robinson, R. I.; Clemens, I. R.; Cox, B.; Pascoe, D. D.; Koch, G.; Eberle,
M.; Berry, M. B.; Booker-Milburn, K. I. Batch versus Flow Photochemistry: A Revealing
Comparison of Yield and Productivity. Chem. Eur. J. 2014, 20, 15226–15232. (e) Politano,
F.; Oksdath-Mansilla, G. Light on the Horizon: Current Research and Future Perspectives
in Flow Photochemistry. Org. Process Res. Dev. 2018, 22, 1045–1062. (f) Mizuno, K.;
Nishiyama, Y.; Ogaki, T.; Terao, K.; Ikeda, H.; Kakiuchi, K. Utilization of Microflow
Reactors to Carry out Synthetically Useful Organic Photochemical Reactions. J.
Photochem. Photobiol. C Photochem. Rev. 2016, 29, 107–147. (g) Plutschack, M. B.;
Pieber, B.; Gilmore, K.; Seeberger, P. H. The Hitchhiker’s Guide to Flow Chemistry. Chem.
Rev. 2017, 117, 11796–11893. (h) Noël, T. A Personal Perspective on the Future of Flow
Photochemistry. J. Flow Chem. 2017, 7, 87–93.
(11) Examples of [2+2] cycloadditions in flow: (a) Terao, K.; Nishiyama, Y.; Tanimoto, H.;
Morimoto, T.; Oelgemöller, M.; Morimoto, T. Diastereoselective [2+2] Photocycloaddition
of a Chiral Cyclohexenone with Ethylene in a Continuous Flow Microcapillary Reactor. J.
Flow Chem. 2012, 2, 73–76. (b) Fukuyama, T.; Kajihara, Y.; Hino, Y.; Ryu, I. Continuous
Microflow [2 + 2] Photocycloaddition Reactions Using Energy-Saving Compact Light
Sources. J. Flow Chem. 2012, 1, 40–45. (c) Conradi, M.; Junkers, T. Efficient [2 + 2]
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Photocycloadditions under Equimolar Conditions by Employing a Continuous UV-Flow
Reactor. J. Photochem. Photobiol. A Chem. 2013, 259, 41–46.
(
12) (a) Elliott, L. D.; Berry, M.; Harji, B.; Klauber, D.; Leonard, J.; Booker-Milburn, K. I. A
Small-Footprint, High-Capacity Flow Reactor for UV Photochemical Synthesis on the
Kilogram Scale. Org. Process Res. Dev. 2016, 20, 1806–1811. (b) Elliott, L. D.; Knowles,
J. P.; Stacey, C. S.; Klauber, D. J.; Booker-Milburn, K. I. Using batch reactor results to
calculate optimal flow rates for the scale-up of UV photochemical reactions React. Chem.
Eng., 2018, 3, 86–93.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
13) Sender, M.; Ziegenbalg, D. Light Sources for Photochemical Processes – Estimation of
Technological Potentials. Chem. Ing. Tech. 2017, 89, 1159–1173.
(
14) (a) Muramoto, Y.; Kimura, M.; Nouda, S. Development and Future of Ultraviolet Light-
Emitting Diodes: UV-LED Will Replace the UV Lamp. Semicond. Sci. Technol. 2014, 29,
0
84004. (b) Hirayama, H.; Maeda, N.; Fujikawa, S.; Toyoda, S.; Kamata, N. Recent
Progress and Future Prospects of AlGaN- Diodes. Jpn. J. Appl. Phys. 2014, 53, 100209.
(
15) Selected examples of recent photochemical reactions using LEDs in place of traditional
higher energy light sources: (a) Loponov, K. N.; Lopes, J.; Barlog, M.; Astrova, E. V.;
Malkov, A. V.; Lapkin, A. A. Optimization of a Scalable Photochemical Reactor for
Reactions with Singlet Oxygen. Org. Process Res. Dev. 2014, 18, 1443–1454. (b) Bonfield,
H. E.; Williams, J. D.; Ooi, X.; Leach, S. G.; Kerr, W. J.; Edwards, L. J. A Detailed Study
of Irradiation Requirements Towards an Efficient Photochemical Wohl-Ziegler Procedure
in Flow. ChemPhotoChem 2018, 2, 938–944. (c) Gorges, R.; Meyer, S.; Kreisel, G.
Photocatalysis in Microreactors. J. Photochem. Photobiol. A Chem. 2004, 167, 95–99. (d)
Meyer, S.; Tietze, D.; Rau, S.; Schäfer, B.; Kreisel, G. Photosensitized Oxidation of
Citronellol in Microreactors. J. Photochem. Photobiol. A Chem. 2007, 186, 248–253. (e)
Ziegenbalg, D.; Kreisel, G.; Weiß, D.; Kralisch, D. OLEDs as Prospective Light Sources
for Microstructured Photoreactors. Photochem. Photobiol. Sci. 2014, 13, 1005–1015. (f)
Manabe, Y.; Kitawaki, Y.; Nagasaki, M.; Fukase, K.; Matsubara, H.; Hino, Y.; Fukuyama,
T.; Ryu, I. Revisiting the Bromination of C-H Bonds with Molecular Bromine by Using a
Photo-Microflow System. Chem. Eur. J. 2014, 20, 12750–12753.
(16) (a) Zhao, J.; Wu, W.; Sun, J.; Guo, S. Triplet Photosensitizers: From Molecular Design to
Applications. Chem. Soc. Rev. 2013, 42, 5323. (b) Aloïse, S.; Ruckebusch, C.; Blanchet, L.;
Réhault, J.; Buntinx, G.; Huvenne, J. P. The Benzophenone S (n,π*) →T (n,π*) States
1
1
Intersystem Crossing Reinvestigated by Ultrafast Absorption Spectroscopy and
Multivariate Curve Resolution. J. Phys. Chem. A 2008, 112, 224–231.
(
17) Selected examples of photoredox methodologies utilizing triplet photosensitizers: (a)
Hörmann, F. M.; Chung, T. S.; Rodriguez, E.; Jakob, M.; Bach, T. Evidence for Triplet
Sensitization in the Visible-Light-Induced [2+2] Photocycloaddition of Eniminium Ions.
Angew. Chem. Int. Ed. 2018, 57, 827–831. (b) Zhao, J.; Brosmer, J. L.; Tang, Q.; Yang, Z.;
Houk, K. N.; Diaconescu, P. L.; Kwon, O. Intramolecular Crossed [2+2]
Photocycloaddition through Visible Light-Induced Energy Transfer. J. Am. Chem. Soc.
3
2
017, 139, 9807–9810. (c) Shen, Y.; Gu, Y.; Martin, R. sp C−H Arylation and
ACS Paragon Plus Environment
28

Page 29 of 31
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Alkylation Enabled by the Synergy of Triplet Excited Ketones and Nickel Catalysts
J. Am. Chem. Soc. 2018, 140, 12200–12209. (d) Strieth-Kalthoff, F.; James, M. J.;
Teders, M.; Pitzer, L.; Glorius, F. Energy transfer catalysis mediated by visible light:
principles, applications, directions. Chem. Soc. Rev. 2018. 47, 7190-7202.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(18) Selected publications in which triplet energies are estimated using DFT calculations: (a)
James, M. J.; Schwarz, J. L.; Strieth-Kalthoff, F.; Wibbeling, B.; Glorius, F.
Dearomative Cascade Photocatalysis: Divergent Synthesis through Catalyst
Selective Energy Transfer J. Am. Chem. Soc. 2018, 140, 8624–8628. (b)
Benniston, A. C.; Harriman, A.; Li, P.; Patel, P. V; Sams, C. A. Enantioselective
photochemistry through Lewis acid−catalyzed triplet energy transfer Science 2016,
354, 1391–1395. (c) Zhao, J.; Brosmer, J. L.; Tang, Q.; Yang, Z.; Houk, K. N.;
Diaconescu, P. L.; Kwon, O. Intramolecular Crossed [2+2] Photocycloaddition
through Visible Light-Induced Energy Transfer J. Am. Chem. Soc. 2017, 139, 9807–
9810.
(
19) Selected uses of Corning Photo Reactors: (a) Gérardy, R.; Winter, M.; Horn, C. R.; Vizza,
A.; Van Hecke, K.; Monbaliu, J.-C. M. Continuous-Flow Preparation of γ-Butyrolactone
Scaffolds from Renewable Fumaric and Itaconic Acids under Photosensitized Conditions.
Org. Process Res. Dev. 2017, 21, 2012–2017. (b) Emmanuel, N.; Mendoza, C.; Winter, M.;
Horn, C. R.; Vizza, A.; Dreesen, L.; Heinrichs, B.; Monbaliu, J.-C. M. Scalable
Photocatalytic Oxidation of Methionine under Continuous-Flow Conditions. Org. Process
Res. Dev. 2017, 21, 1435–1438.
(
20) Examples of accelerated reaction by using biphasic slug-flow regimes: Nakano, M.;
Nishiyama, Y.; Tanimoto, H.; Morimoto, T.; Kakiuchi, K. Remarkable Improvement of
Organic Photoreaction Efficiency in the Flow Microreactor by the Slug Flow Condition
Using Water. Org. Process Res. Dev. 2016, 20, 1626–1632. (b) Telmesani, R.; White, J. A.
H.; Beeler, A. B. Liquid-Liquid Slug-Flow-Accelerated [2+2] Photocycloaddition of
Cinnamates. ChemPhotoChem 2018, 2, 865–869.
(
21) See Supporting Information Table S14 for a comparison of DFT methods, leading to the
conclusion that M06-2X/TZVP is suitable.
(
22) Examples of experimental triplet energy measurements under cryogenic conditions, using
specialist equipment: (a) Herkstroeter, W. G.; Lamola, A. A.; Hammond, G. S. Mechanisms
of Photochemical Reactions in Solution. XXVIII. Values of Triplet Excitation Energies of
Selected Sensitizers. J. Am. Chem. Soc. 1964, 86, 4537–4540. (b) McClure, D. S. Triplet-
Singlet Transitions in Organic Molecules. Lifetime Measurements of the Triplet State. J.
Chem. Phys. 1949, 17, 905–913. (c) Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T.
Handbook of Photochemistry, 3rd ed.; CRC Press: Boca Raton, US, 2006.
ACS Paragon Plus Environment
29