Photochemical Flow Oximation of Alkanes

Article abstract of DOI:10.1055/s-0040-1707281

The nitrosation of several alkanes using tert-butyl nitrite has been performed in flow showing a remarkable reduction in the reaction time compared with batch processing. Due to the necessity for large excesses of the alkane component a continuous recycling process was devised for the preparation of larger quantities of material.

Full text of DOI:10.1055/s-0040-1707281

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart
2020, 31, A–F
cluster
A
Integrated Synthesis Using Continuous-
O. M. Griffiths et al.
Cluster
Synlett
Photochemical Flow Oximation of Alkanes
Oliver M. Griffithsa,b
Michele Ruggeria
Ian R. Baxendale*a
a Department of Chemistry, University of Durham, South Road,
Durham, Durham, DH1 3LE, UK
i.r.baxendale@durham.ac.uk
b Department of Chemistry, Cambridge University, South Road,
Cambridge, Cambridgeshire, CB2 1EW, UK
Published as part of the Cluster
Integrated Synthesis Using Continuous-Flow Technologies
Corresponding Author
IDeaMenpaRail.rtBi.mra.xbeeannxtdeonafldCeaahlem@disutrhy,aUmn.aivce.urksity of Durham, South Road, Durham, Durham, DH1 3LE, UK
Received: 24.07.2020
Accepted after revision: 19.08.2020
Published online: 21.09.2020
In 2019 Lebl et al.8 reported a continuous version of the
Toray process claiming 57% overall yield for the photochem-
ical step and showing how the transformation could benefit
from flow processing in terms of scaling-up, sustainability,
reaction time, and safety. Despite the power of this strategy
it has several drawbacks, primarily amongst these is the
need to generate the unstable nitrosyl chloride (2) and the
highly corrosive character of its reaction byproduct hydro-
chloric acid. Consequently, several other nitrosylating
agents have been studied, with alkyl nitrites9 showing good
synthetic utility and being readily available especially tert-
butyl nitrite. The corresponding oxamination reaction
(Scheme 2), like the Toray process, also involves the photo-
promoted dissociation of the nitrosylating agent 4, the
alkoxy radical formed then abstracts a hydrogen atom from
the hydrocarbon generating a second radical which can re-
act with the remaining nitroso radical leading to the de-
sired product 3 after tautomerization. However, it has been
shown that the main product of the reaction is actually the
trans-configured dimer 510 whilst the desired oxime 3 is
obtained as only a minor product; however, the reaction
shows good yields (>80% varying ratios 3/5).11 As a result
several batch methods have been devised to convert 5 into
the synthetically more valuable oxime 3.12
DOI: 10.1055/s-0040-1707281; Art ID: st-2020-u0416-c
Abstract The nitrosation of several alkanes using tert-butyl nitrite has
been performed in flow showing a remarkable reduction in the reaction
time compared with batch processing. Due to the necessity for large ex-
cesses of the alkane component a continuous recycling process was de-
vised for the preparation of larger quantities of material.
Key words nitrosation, oximes, flow chemistry, Toray process, photo-
chemistry
Oximes are an important class of molecules that have
found numerous applications in many different fields such
as coordination chemistry,1 material science,2 and medici-
nal chemistry.3 The further importance of these molecules
is also highlighted by the large number of important biolog-
ically active compounds possessing this chemical moiety.4
However, the greatest application of oximes is often as in-
termediates in cascades, such as the named Beckmann rear-
rangement or the related fragmentation, and in their reac-
tions leading to nitriles.5
Cyclohexanone oxime is a compound that exemplifies
the aforementioned rearrangement chemistry being a pre-
cursor of caprolactam, itself a starting material for nylon-6
synthesis. Because of the immense industrial demand for
nylon there has been significant research into the prepara-
tion of all the intermediates along the chemical pipeline.
Since the discovery of the Toray photonitrosation of cyclo-
hexane (PNC) process (Scheme 1),6 which enables oxime
formation directly from cyclohexane in a single step with-
out going through the related ketone, interest in photo-oxi-
mation7 has grown considerably.
HO
O
N
N
N
hν
O
O
+
N
O
1
5
4
3
Scheme 2 Nitrosation of cyclohexane with tert-butyl nitrile
Interested in the synthetic potential of being able to ac-
tivate normally inert alkanes and the inherent benefits of-
fered by conducting photochemical reactions in flow13 we
embarked upon a more in-depth study of this chemistry in
flow.
HO
high-pressure
N
mercury lamp
NOCl
+
+
HCl
HCl, 15 °C
86%
2
1
For our study we utilized a commercial Vapourtec E-se-
ries flow reactor in combination with a UV 150 photochem-
3
Scheme 1 The Toray process towards nylon-6 synthesis
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F
Georg Thieme Verlag KG, Rüdigerstraße 14, 70469 Stuttgart, Germany

B
O. M. Griffiths et al.
Cluster
Synlett
Table 1 Selected Data for the Evaluation of the Molar Ratio 1/t-BuOH/4a
ical add-on allowing constant monitoring of temperature
and application of external cooling if required.14 The photo-
reactor was equipped with a set of 365 nm LEDs (9 W) pro-
viding irradiation to a 10 mL FEP coiled tube flow reactor (1
mm ID). The choice of the light source was influenced by
the absorption spectra of the species involved in the trans-
formation. The absorption band responsible for the homo-
lytic breaking of t-BuONO lies between 320–430 nm, but is
partially overlapped, with the absorption band of the oxime
product 3 and the dimer 5 which have a maximum at 300
nm but extends to around 360 nm. Wysocki et al.15 has
shown that efficient activation of t-BuONO can be achieved
using an emission between 365 and 405 nm and therefore a
set of 365 nm LEDs were used. However, this does introduce
some limitation regarding photon flux and can therefore
lead to the requirement for longer reaction times to ensure
high conversions.
To establish the flow process, we used cyclohexane and
screened several molar ratios of reagents, temperature, and
flow rates in order to find optimum conditions for the
transformation. In their study, Wysocki et al.15 had deter-
mined that the addition of t-BuOH as an additive was highly
beneficial to the reaction progress in batch. We therefore
additionally wished to validate this finding with respect to
the flow process (Table 1).
Entry
Molar ratio
1/tBuOH/4
Temp (°C)
tR (min)
Total NMR Ratio of
conv. (%)
5/3
1
2
100:0:1
100:0:1
100:0:1
100:0:1
100:15:1
100:15:1
100:15:1
100:15:1
200:30:1
30:0:1
50
28
18
18
28
28
28
50
50
18
18
28
50
50
50
50
50
50
10
10
10
20
10
2.5
1.25
5
66
55
42
41
58
53
37
60
62
32
46
57
68
68
67
69
55
55
5.4:1
6.4:1
9.5:1
4.3:1
6.3:1
12.3:1
6.4:1
6.2:1
10.5:1
8.9:1
7.9:1
5.9:1
4.6:1
8.8:1
10.1:1
6.2:1
4.5:1
4.5:1
3
4
5
6
7
8
9
5
10
11
12
13
14
15
16
17
18
10
10
10
10
10
10
10
10
10
30:15:1
30:15:1
30:15:1
45:15:1
60:15:1
45:0:1
15:30:1
30:60:1
a 10 mL flow coil; conversion vs an internal standard.
Several general observations can be made. Firstly, the
reaction produces mainly the dimer 5 in accordance with
previous-literature-reported batch results.16 Furthermore
and in validation of Wysocki’s study15 the addition of t-
BuOH does have an impact on conversion (Table 1; cf. en-
tries 3 and 5, 10 and 11) and also changes the ratio between
5 and 3 in favor of the desired monomer 3; again in accor-
dance with their report (albeit not as significantly). Howev-
er, the influence of the t-BuOH seems limited (cf. entries 13
and 14, 17 and 18) and seemingly dilution (cf. entries 3 and
10) and more significantly temperature (cf. entries 1–3, 11–
13) has a more pronounced effect on the reaction outcome
with regards to overall conversion and product composi-
tion. The fact that equitable results were obtained at higher
reactor temperature (higher temperatures failed to give
better results) without the addition of the t-BuOH suggests
that under these conditions it does not play an important
role behaving only as a diluent (entries 1, 8, 9, 14, and 16).
Finally, residence time indicates an optimal reaction win-
dow of between 5–10 min based upon this initial scoping.
Several of these observations can be rationalized. A high
dilution of the tert-butyl nitrite reduces competing termi-
nation and side reactions which are competitive when us-
ing non-activated reactants such as cyclohexane.6–8 For ex-
ample, it was noted that nitrocyclohexane was formed in
small amounts (2–9%) as a byproduct increasing propor-
tionally with higher concentration of tert-butyl nitrite
tent with studies of Mackor et al. when high nitric oxide
concentrations were employed.12 The impact of tempera-
ture is less clear but to discount a purely thermal fragmen-
tation process the reaction was also conducted (repeat of
Table 1, entry 9) without irradiation which resulted in re-
covery of only unreacted starting materials. As higher tem-
peratures yield higher compositions of the monomer 3 it
may be the monomer is more stable than the dimer 5 to de-
composition and hence this accounts for the improved con-
versions.
We have previously found value in screening a range of
alkyl nitrites in other projects,17 therefore additional alkyl
nitrite sources were evaluated (e.g. isoamyl nitrite, butyl ni-
trite) but all gave inferior results (conversion/purity) com-
pared to tert-butyl nitrite. This is consistent with other lit-
erature studies12,15,18 explaining that non-tertiary alkyl ni-
trites undergo undesirable side reactions, such as the
Barton reaction, when irradiated.
With good initial results from the direct reaction of tert-
butyl nitrite (4) and cyclohexane (1, Table 1, entry 16, 48%
isolated yield of 5 by recrystallization from cyclohexane)
we decided to target these conditions for further optimiza-
tion focusing on concentration and residence time (Table 2,
Figure 1). Reviewing the data indicated we had already ser-
endipitously identified the best conditions (Table 1 entry
16, Table 2 entry 5).
1
(identified by GC–MS and H NMR spectroscopy); consis-
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F

C
O. M. Griffiths et al.
Cluster
Synlett
Table 2 Molar Ratio Studya
leaving the products 5/3; which in a second step can be
heated to convert them into 3, we wished to establish the
principles of a more continuous operation. To create a via-
ble set-up, we also had to consider some further reaction
aspects: During our work we had found that prolong heat-
ing of oxime 3 neat (>45 min) led to its slow but progressive
decomposition. This was nevertheless completely sup-
pressed if a solvent such as cyclohexane with quantities of
t-BuOH was present. However, the low solvent reflux tem-
perature of cyclohexane resulted in the need for longer
heating periods to effect full conversion of 5 into 3. There-
fore, and to avoid any addition issues with needing to regu-
late the rate of a continuous distillation (not drying out the
product) but allowing constant heating (conversion of 5
into 3), we adopted a mixed flow and continuous distilla-
tion set-up (Figure 2).
Entry
Molar ratio 1/4
Product ratio 5/3
Total NMR
conv. (%)
1
2
10:1
20:1
30:1
40:1
45:1
50:1
60:1
70:1
80:1
90:1
100:1
8.1:1
7.2:1
6:5:1
6.4:1
6.2:1
6.1:1
7.0:1
7.0:1
6.9:1
7.0:1
5.4:1
35
57
59
65
69
64
64
64
63
64
66
3
4
5
6
7
8
9
10
11
a Flow rate of 1 mL/min; conversion vs an internal standard.
Figure 2 Continuous-flow recycling process for cyclohexane
An assembled Soxhlet extraction apparatus (a) func-
tioned to separate the product solution (b) from unreacted
cyclohexane (a) whilst converting the dimer 5 into oxime 3
(b) and simultaneously acting to regulate an essentially
constant liquid level (a/b). Unreacted cyclohexane was con-
tinuously distilled and could be pumped from the top of the
Soxhlet collection chamber (a, containing pure cyclohex-
ane) to a stirred stock flask (d) where it was continuously
refreshed with additional tert-butyl nitrite at a flow rate
commensurate to maintain the established 45:1 reagent ra-
tio. The stock solution (d) was pumped through the photo-
reactor with an optimized residence time of 10 min and a
theoretical throughput of 1.24 g/h.
Figure 1 Residence time optimization, 10 mL FEP coil reactor using
45:1 of 1/4; conversion vs an internal standard
Our next consideration was the transformation of dimer
5 to the corresponding monomer 3. Donaruma had report-
ed18 that isomerization could be performed by dissolution
and heating in MeOH, indeed 20 min reflux resulted in
complete conversion (2 M, 50 mmol). As a note, the solution
changed from colourless to blue, indicating the formation of
the monomer species. By comparison, testing t-BuOH gave
only 15–18% conversion after 90 min (65 °C) indicating a
much slower process. Interestingly, whilst performing melt-
ing point purity analysis on 5, it was observed that when
heated neat it readily interconverted into the corresponding
oxime 3. Thus when a purified sample of 3 was heated neat
at 100 °C for 30 min a quantitative conversion into the
monomer 3 was achieved.20 However, when a sample of a
1
Interval sampling and H NMR analysis of the photore-
actor output indicated that the sum (69% ± 1.8 total conver-
sion) and ratio 5/3 (1:4.2 ±0.08) remained essentially con-
stant (4 h run). However, the resulting isolated yield of 3
was 45% (via recrystallisation) rather disappointing when
compared to previous reported batch processes (53–
82%).12c,15 However, it should be noted that these previous
batch experiments typical involved prolonged irradiation
times of >16 h. Wysocki15 had noted that in batch after 180
min irradiation (high power 365 nm NVSU233A LED diodes
from Nichia Corp) conversion reached a maximum consid-
ering the sum of 5/3 (with 3 only being detected at >45
min), whereas it took an additional 13 h of irradiation to
completely convert the dimeric species into oxime 3. The
1
crude reaction product (69% H NMR conversion; 4.2:1 of
5/3) was heated similarly a less clean transformation was
observed and only 44% isolated yield of 3 was obtained fol-
lowing chromatographic purification. We later found that
dimer 5 can be easily isolated and recrystallized from cyclo-
hexane which offers a simple processing sequence.
So, although it is trivial to batch collect the reactor out-
put and remove the excess low boiling point cyclohexane
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F

D
O. M. Griffiths et al.
Cluster
Synlett
conversion of dimer 5 into the corresponding monomer 3 is
therefore also photochemically induced. Indeed, we also
demonstrated that reducing the flow rate (1 → 0.5 mL/min)
led to a higher proportion of the monomer (3.1:1 vs 4.3:1
5/3) in the photoreactor output but this was associated
with a reduced overall conversion (69 → 60%; Figure 1).
Fortunately, our continuous distillation system should due
to the longer processing times overcome this need for addi-
tional irradiation, due to the thermally conducted splitting
of the dimer.
The Soxhlet system (Figure 2) was therefore run in con-
tinuous mode for an extended 20 h and generated after re-
crystallisation from cyclohexane 14.72 g of pure product 3
(59%) demonstrating the scalability achievable with this
simple system. We also believe it should be possible to cou-
ple multiple photoreactors to a single Soxhlet system to
drastically increase throughput, currently the photoreactor
is the limiting processing component in this set-up.
Having determined the feasibility of the process we next
explored the scope of the transformation in terms of other
viable substrates. First, we considered the value of introduc-
ing unsaturation to the reactant, namely testing cyclohex-
ene, as the product would offer value as an aniline precur-
sor via the Semmler–Wolff reaction.21 In addition, the
alkene should make cyclohexene a better substrate by as-
sisting the proton-abstraction step due to the lower bond
dissociation energy of the allylic hydrogen compared with
the purely alkyl bond (85 vs 96 kcal mol–1).22
was advantageous (see the Supporting Information). A sub-
strate-to-nitrite ratio of 100:1 gave the best results again at
50 °C. For this substrate, the rate of the reaction was much
faster, with higher conversions being achieved at increased
flow rates, ultimately, 2 mL/min was deemed optimal (83%
conv. 61:22 7+8/9). Also, it was noted that under shorter
residence times higher proportions of compound 9 was ob-
served which tautomerized slowly upon standing but was
shown to be rapidly interconverted upon further irradiation
presumably by a photochemical [1,3]-sigmatropic hydride
shift.
As performed previous, we investigated the reaction of
cyclohexene (6) in the integrated Soxhlet flow reactor
which allowed easy scaling of the reaction and isolation of
21.98 g of a mixture of 7/8 (3.8:1) in 77% isolated yield from
a 22 h run.
Given these encouraging results we decided to pursue a
small expansion of the reaction scope by investigating sev-
eral other alkanes (Figure 3). The new substrates selected
were by comparison all non-volatile therefore removal of
unreacted starting material was no longer achievable by
simple evaporation requiring all crude products to be sepa-
rated by chromatographic purification. To enact the flow
process a standard stock solution comprising substrate/t-
BuOH/t-BuONO 20:20:1 (10 mmol t-BuONO scale) was pre-
pared and pumped through the photoreactor at 1 mL/min
and 50 °C (no optimization was performed). The reactor
output was collected, and the solvent evaporated before the
products were isolated via chromatography.
HO
OH
HO
OH
N
N
N
N
N
10 (68%)
11 (71%)
13 (47%)
12 (40%)
Scheme 3 Photopromoted oximation of cyclohexene
R
OH
R
HO
HO
N
N
N
When the reaction was run in flow it produced predom-
inantly compound 7 (Scheme 3) which exists in equilibrium
with 8 (3.8:1). This was proven by the partial conversion of
a pure solution of 7 back into a 3.8:1 mixture of 7/8 (24 h in
CDCl3). The preference for the anti-isomer 7 can be ratio-
nalized by a 1,4-interaction between the hydroxyl group
and the vinyl proton in 8 (Scheme 4).
Ph
+
N
R
HO
Ph
14 (47%)
18 R = H (64%)
19 R = Me (38%)
16 R = H (47%)
17 R = OMe (49%)
15 (11%)
Figure 3 Expansion of substrate scope
Compounds 10 and 11 were obtained starting from te-
traline and indane as single regio- and diastereoisomers
(confirmed by single-crystal X-ray determination; see the
Supporting Information) in 68% and 71% yield, respectively.
The regiospecificity for the benzylic proton abstraction
arises from the lower bond dissociation energy (83 vs 96
kcal abstraction per mol).23 The associated diastereoselec-
tivity is accounted for by conformational preference arising
from the varying interactions between the hydroxy group
and the -geminal protons vs the -peri-interaction with
the aromatic proton (Figure 4, 20 and 21).
HO
O
O
OH
N
N
N
N
H
H
H
H
7
9a
9b
8
Scheme 4 Interconversion of 2-cyclohexenone oxime isomers
As the reagent ratio had been an important parameter
in the reaction of the cyclohexane (1) a rapid screening was
performed finding again that a large excess of cyclohexene
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F

E
O. M. Griffiths et al.
Cluster
Synlett
(3) (a) Logan, R. T.; Redpath, J.; Roy, R. G. EP 0199393, 1986.
(b) Huang, C.-T.; Pelosi, S. S. Jr.; Bayless, A. V. US 4882354, 1989.
(c) Shahid, M.; Martorana, M. G.; Cottney, J. E.; Marshall, R. J.
J. Pharmacol. 1990, 100, 735.
H
H
H
HO
H
H
OH
N
N
H
H
H
H
H
H
H
(4) (a) Soga, S.; Neckers, L. M.; Schulte, T. W.; Shiotsu, Y.; Akasaka,
K.; Narumi, H.; Agatsuma, T.; Ikuina, Y.; Murakata, C.; Tamaoki,
T.; Akinaga, S. Cancer Res. 1999, 59, 2931. (b) Nikitjuka, A.;
Jirgensons, A. Chem. Heterocycl. Compd. 2014, 49, 1544.
(c) Hasaneen, M. N. Herbicides Properties, Synthesis and Control
of Weeds (2012), (accessed Apr 15, 2020); IntechOpen: London,
2012, https://www.intechopen.com/books/herbicides-proper-
ties-synthesis-and-control-of-weeds. (d) Fylaktakidou, K. C.;
Hadjipavlou-Litina, D. J.; Litinas, K. E.; Varella, E. A.; Nicolaides,
D. Curr. Pharm. Des. 2008, 14, 1001. (e) Marrs, T. C. Pharmacol.
Ther. 1993, 58, 51. (f) Dawson, R. M. J. Appl. Toxicol. 1994, 14,
317. (g) Taylor, P. Anticholinesterase Agents, In The Pharmacolog-
ical Basis of Therapeutics, 9th ed; Hardman, J. G.; Limbird, L. E.,
Ed.; McGraw Hill: New York, 1996, 161–176.
(5) Wang, Z. Beckmann Rearrangement and Beckmann Fragmenta-
tion, In Comprehensive Organic Name Reactions and Reagents;
John Wiley & Sons: Hoboken, NJ, 2010, 288–295.
(6) (a) Ito, Y. Bull. Chem. Soc. Jpn. 1956, 29, 227. (b) Ito, Y.; Matsuda,
S. Ann. N.Y. Acad. Sci. 1969, 147, 618. (c) Fischer, M. Angew.
Chem., Int. Ed. Engl. 1978, 17, 16.
H
22
20
21
Figure 4 Disfavored interactions conformational interactions
Compound 12 was obtained from the (1R,6S)-3-carene.
Again, excellent regio- and stereoselectivity were observed
arising from the steric hindrance between the allylic posi-
tions (orange favored abstraction site) on the cyclohexene
ring and the methyl group (Figure 4, 22). Interestingly no
oximation was observed on the allylic methyl arguably be-
cause the primary radical obtained is less stable compared
with the alternative secondary positions. In order to test
the reaction on a heterocyclic core, tetrahydroquinoline
was processed. Of the two possible kinetic regioisomeric
products only compound 13 forms as a single stereoisomer
probably due to the presence of stabilizing hydrogen bond-
ing between the hydroxyl proton and the pyridine nitrogen.
Starting from 1-phenylcyclohexene a 4.3:1 regioisomeric
mixture of compounds 14 and 15 were obtained with pref-
erence for abstraction at the less hindered allylic proton. Fi-
nally, oximes 16 and 17 were generated from their symmet-
rical precursor alkanes and similarly the systems 18 and 19
could be prepared from toluene and ethyl benzene, respec-
tively.
(7) (a) Lynn, E. V. J. Am. Chem. Soc. 1919, 41, 368. (b) Lynn, E. V.;
Hilton, O. J. Am. Chem. Soc. 1922, 44, 645.
(8) Lebl, R.; Cantillo, D.; Kappe, C. O. React. Chem. Eng. 2019, 4, 738.
(9) (a) Weiß, R.; Wagner, K.; Hertel, M. Chem. Ber. 1984, 117, 1965.
(b) Haub, E. K.; Lizano, A. C.; Noble, M. E. Inorg. Chem. 1995, 34,
1440. (c) Grossi, L.; Strazzari, S. J. Org. Chem. 1999, 64, 8076.
(d) Monbaliu, J.-C.; Jorda, J.; Chevalier, B.; Morvan, B. Chim. Oggi
2011, 29, 50.
(10) Smith, D. B. Photochemistry, Vol 2; RSC: London, 1997.
(11) Hong, W. P. Iosub A. V. Stahl S. S. J. Am. Chem. Soc. 2013, 137,
13664.
(12) (a) Pape, M. Fortschr. Chem. Forsch. 1967, 7, 559. (b) Mackor, A.;
Veenland, J. U.; de Boer, T. J. Recl. Trav. Chim. Pays-Bas 1969, 88,
1249. (c) Mackor, A.; de Boer, T. J. Recl. Trav. Chim. Pays-Bas
1969, 89, 151. (d) Mackor, A.; de Boer, T. J. Recl. Trav. Chim. Pays-
Bas 1969, 89, 159. (e) Mackor, A.; de Boer, T. J. Recl. Trav. Chim.
Pays-Bas 1970, 89, 164.
In conclusion, we have reported the photoflow oxima-
tion of several alkanes using tert-butyl nitrite. The flow
process allows for a considerable reduction in the reaction
time and enables easy scale-up of the transformation. An
integrated continuous distillation process was also investi-
gated in order to gauge the possibility of recycling the vola-
tile unreacted starting materials.
(13) (a) 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. Chem. Eur. J. 2014, 20, 15226. (b) Baumann, M.;
Baxendale, I. R. Beilstein J. Org. Chem. 2015, 11, 1194.
(c) Movsisyan, M.; Delbeke, E. I. P.; Berton, J. K. E. T.;
Battilocchio, C.; Ley, S. V.; Stevens, C. V. Chem. Soc. Rev. 2016, 45,
4892. (d) Fanelli, F.; Parisi, G.; Degennaro, L.; Luisi, R. Beilstein J.
Org. Chem. 2017, 13, 520. (e) Fuse, S.; Otake, Y.; Nakamura, H.
Eur. J. Org. Chem. 2017, 44, 6466. (f) Plutschack, M. B.; Pieber, B.;
Gilmore, K.; Seeberger, P. H. Chem. Rev. 2017, 117, 11796.
(g) Shen, G.; Osako, T.; Nagaosa, M.; Uozumi, Y. J. Org. Chem.
2018, 83, 7380. (h) Akwi, F. M.; Watts, P. Chem. Commun. 2018,
54, 13894. (i) Sambiagio, C.; Noël, T. Trends Chem. 2020, 2, 92.
(j) Filippo, M. D.; Bracken, C.; Baumann, M. Molecules 2020, 25,
356.
Acknowledgment
We would like to thank Dr Andrej Batsanov and Dr Dmitry S. Yufit of
Durham University for solving the X-ray structures.
Supporting Information
Supporting information for this article is available online at
https://doi.org/10.1055/s-0040-1707281. Su
p
p
ortingInformationS
u
p
p
ortingI
n
formatio
n
References and Notes
(1) Bolotin, D. S.; Bokach, N. A.; Kukushkin, V. Y. Coord. Chem. Rev.
2016, 313, 62.
(2) (a) Tamada, M.; Seko, N.; Yoshii, F. Radiat. Phys. Chem. 2004, 71,
221. (b) Seko, N.; Tamada, M.; Yoshii, F. Nucl. Instrum. Methods
Phys. Res., Sect. B 2005, 236, 21.
(14) https://www.vapourtec.com/products/e-series-flow-chemis-
try-system/the-easy-photochem-features/(accessedSeptember
12, 2020).
(15) Wysocki, D.; Teles, J. H.; Dehn, R.; Trapp, O.; Schäfer, B.; Schaub,
T. ChemPhotoChem 2018, 2, 22.
(16) Smith, D. B. Photochemistry, Vol. 1, 2nd ed; RSC: London, 1970.
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F

F
O. M. Griffiths et al.
Cluster
Synlett
(17) (a) Browne, D. L.; Baxendale, I. R.; Ley, S. V. Tetrahedron 2011,
67, 10296. (b) Hu, T.; Baxendale, I. R.; Baumann, M. Molecules
2016, 21, 918. (c) Röder, L.; Nicholls, A. J.; Baxendale, I. R. Mole-
cules 2019, 24, 1996.
(18) Donaruma, L. G.; Carmody, D. J. J. Org. Chem. 1957, 22, 635.
(19) Donaruma, L. G. J. Org. Chem. 1958, 23, 1338.
(20) Burrell, E. J. J. Phys. Chem. 1962, 66, 401.
(21) (a) Semmler, W. Chem. Ber. 1892, 25, 3352. (b) Wolff, L. Liebigs
Ann. 1902, 322, 351.
(22) (a) Alfassi, Z. B.; Feldman, L. Int. J. Chem. Kinet. 1981, 13, 771.
(b) CRC Handbook of Chemistry and Physics, 85th ed; Lide, D. R.,
Ed.; CRC Press: Boca Raton, FL, 2004.
(23) Laarhoven, L. J. J.; Mulder, P. J. Phys. Chem. 1997, 101, 73.
© 2020. Thieme. All rights reserved. Synlett 2020, 31, A–F