Controlling the photochemical reactions of alkenes by light-path length effects of a microchannel reactor

Article abstract of DOI:10.1007/s11164-012-0636-4

Photoirradiation of Me₂CO-H₂O solution of pent-4-en-1-ol (1a) with a high-pressure mercury lamp in a test tube gave 8-hydroxyoctan-2-one (3a) in 66 % yield along with oxetane (4a) and the isomer (4a') in 10 % yield. Irradiation of the running Me₂CO-H₂O solution of 1a in the flow system of a microchannel reactor (MCR) gave mainly 4a. The photoreaction of 1,1-diphenylethene (2a) with triethylamine gave a Markovnikov-type adduct (5a) and an anti-Markovnikov-type adduct (6a). The use of the MCR enhanced the production of 5a. These phenomena were explained by the light-path length effects of the MCR.

Full text of DOI:10.1007/s11164-012-0636-4

Res Chem Intermed (2013) 39:111–126
DOI 10.1007/s11164-012-0636-4
Controlling the photochemical reactions of alkenes
by light-path length effects of a microchannel reactor
•
Toshiaki Yamashita Shuhei Matsushita
•
•
Takuya Nagatomo Ryosuke Yamauchi
Masahide Yasuda
•
Received: 31 August 2011 / Accepted: 9 November 2011 / Published online: 23 June 2012
ꢀ Springer Science+Business Media B.V. 2012
Abstract Photoirradiation of Me2CO–H2O solution of pent-4-en-1-ol (1a) with a
high-pressure mercury lamp in a test tube gave 8-hydroxyoctan-2-one (3a) in 66 %
yield along with oxetane (4a) and the isomer (4a0) in 10 % yield. Irradiation of the
running Me2CO–H2O solution of 1a in the flow system of a microchannel reactor
(MCR) gave mainly 4a. The photoreaction of 1,1-diphenylethene (2a) with trieth-
ylamine gave a Markovnikov-type adduct (5a) and an anti-Markovnikov-type
adduct (6a). The use of the MCR enhanced the production of 5a. These phenomena
were explained by the light-path length effects of the MCR.
Keywords Microchannel reactor ꢀ Photoaddition ꢀ Alkene ꢀ Aliphatic amine ꢀ
Acetone
Introduction
Microchannel reactors (MCRs), which consist of a narrow channel, are microre-
action devices which have characteristic features such as short molecular diffusion
distances, excellent heat transfer, large surface-to-volume ratio, and laminar flow.
It is expected that use of MCRs in photochemical reactions will lead to the
following effects:
T. Yamashita (&) ꢀ S. Matsushita ꢀ T. Nagatomo ꢀ R. Yamauchi
Department of Chemical Science and Engineering, Miyakonojo National College of Technology,
Miyakonojo, Miyazaki 885-8567, Japan
e-mail: photoya@cc.miyakonojo-nct.ac.jp
M. Yasuda
Department of Applied Chemistry, Faculty of Engineering, University of Miyazaki,
Gakuen-Kibanadai, Miyazaki 889-2192, Japan
e-mail: yasuda@cc.miyazaki-u.ac.jp
123

112
T. Yamashita et al.
(i) Temperature-controlled photoreaction by the MCR: We have reported
temperature dependence of the redox sensitization of the pair of phenanthrene
and p-dicyanobenzene towards the nucleophilic addition reaction of amines
and alcohols to simple alkenes using a MCR where the reaction temperature
was precisely controlled [1]. It was found that the redox-sensitized nucleo-
philic addition proceeded efficiently at lower temperature.
(ii) Suppression of further reaction by the flow system of the MCR: For example,
Yoshida and coworkers [2] reported that use of a micromixing apparatus
depressed dialkylation in Friedel–Crafts monoalkylation of aromatic com-
pounds with N-acyliminium ions. Also, Matsushita et al. [3] have investigated
that the process of N,N-dialkylation in the photocatalytic N-alkylation of
benzylamine was inhibited by the use of a microreactor.
(iii) Effects of light-path length of the MCR: The MCR has a narrow channel
where incident light passes through a few microns of path length, being very
short compared with the few centimeters of path length for a usual batch
system. When irradiated in a batch, it is thought that absorption of the incident
light takes place by the part of the substrate located near the vessel walls.
Therefore, molecules in the ground state remain in the vessel. If the light-path
length of the vessel is made very thin, the whole of the substrate present in the
vessel can be excited. Therefore, it can be expected that a microscale reactor
such as an MCR will change the reaction pathways. Mizuno and coworkers
[4] reported that the efficiency and regioselectivity of intramolecular
[2 ? 2] and [2 ? 3] photocycloaddition of alkenylnaphthalenes were
enhanced by the effects of the light-path length of the MCR.
Here we investigate the light-path length effects of the MCR on the photoreaction
of acetone with pent-4-en-1-ol (1a) (Scheme 1, reaction A) and 1,1-diphenylethene
(2a) with triethylamine (TEA) (reaction B). The product distribution of 8-hydrox-
yoctan-2-one (3a) and oxetane (4a) in reaction A and 2-(diethylamino)-3,
Reaction A
Me
OH
OH
Me
OH
OH
hν
O
Me
+
+
+
+
+
Me2CO
O
Me
O
Me
1a
3a
4a
4a'
Reaction B
Ph
Ph NEt2
Me
Ph Me
hν
Me
Et3N
Ph
Ph
Ph
NEt2
2a
5a
6a
Scheme 1 Photoreactions of 1a with acetone (reaction A) and 2a with TEA (reaction B)
123

Controlling the photochemical reactions of alkenes
113
3-diphenylbutane (5a) and 3-(diethylamino)-1,1-diphenylbutane (6a) in reaction B
were compared between batch and MCR photoreactions.
Experimental
General
1
Melting points were measured using open capillary tubes and are uncorrected. H
and 13C nuclear magnetic resonance (NMR) spectra were recorded in CDCl3 using
tetramethylsilane as internal standard on a Bruker AV 400 M at 400 and 100 MHz,
respectively. Matrix-assisted laser desorption/ionization mass spectra (MALDI–
MS) were measured on a Bruker Daltonics Autoflex III TOF/TOF in positive ion
mode. Gas–liquid chromatography (GLC) analysis was performed using a 25-m
fused-silica capillary column. Gas chromatography (GC)–mass spectrometry (MS)
analysis was performed on a Shimadzu GCMS-QP5050 using a 25-m fused-silica
capillary column.
Spectral-grade acetone, MeCN, tetrahydrofuran (THF), benzene, hexane, and
cyclohexane were used without further purification. A special grade of alkenols
(1a–f), triethylamine (TEA), and tert-butylamine (t-BuNH2) were employed without
further distillation. Commercially available 1,1-diphenylethene (2a) was distilled
from sodium under reduced pressure before use.
Batch photoreaction of acetone with pent-4-en-1-ol (1a)
The typical reaction procedure for the photoreaction of acetone with pent-4-en-1-ol
(1a) was as follows: An aqueous solution (13 mL) containing 1a (0.15 mmol) and
acetone (3 mL) was introduced into a Pyrex test tube of 22 mm diameter and was
then bubbled with argon for 10 min. The solution was irradiated for 40 min with
a 300-W high-pressure mercury lamp (PIH 300; Eikosha, Japan). After the
photoreaction, the reaction mixture was extracted with chloroform, and then the
extract was subjected to column chromatography on silica gel using hexane and
hexane–ethyl acetate (10:1 and 4:1, v/v) as eluents. The structure of 8-hydroxyoc-
tan-2-one (3a) was determined by comparisons of the spectral data with those
published in literature [5]. The oxetane (4a) and the isomer (4a0) were decomposed
at room temperature over a few days. The minor product, 6-methylhept-3-ene-1,
6-diol (7), was obtained as a mixture with 4a. The structure of 7 was determined by
1H and 13C NMR spectra. The yields of the products were determined by GLC using
heptadecane as internal standard.
8-Hydroxyoctan-2-one (3a)
Colorless oil; 1H NMR d/ppm = 1.25–1.42 (m, 4H), 1.50–1.64 (m, 4H), 2.14
(s, 3H), 2.44 (t, J = 7.3 Hz, 2H), 3.16 (brs, 1H), 3.59 (t, J = 6.5 Hz, 2H); 13C NMR
d/ppm = 23.8, 25.6, 28.92, 30.3, 32.5, 43.6, 62.4, 209.8; HRMS (EI) m/z calcd. for
C8H16O2 144.1150, found 144.1070.
123

114
T. Yamashita et al.
2,2-Dimethyl-3-(3-hydroxypropyl)oxetane (4a)
1
Colorless oil; H NMR: d/ppm = 1.33 (s, 3H), 1.41 (s, 3H), 1.37–1.52 (m, 2H),
1.58–1.68 (m, 2H), 2.65 (m, 1H), 3.61 (t, J = 6.4 Hz, 2H), 3.62 (brs, 1H), 4.13 (dd,
J = 7.2, 6.1 Hz, 1H), 4.53 (dd, J = 8.2, 6.0 Hz, 1H); 13C NMR: d/ppm = 23.2,
25.7, 30.5, 30.8, 44.2, 62.5, 71.0, 86.8; HRMS (CI) m/z calcd. for C8H16O2
(M ? H)? 145.1228, found 145.1239.
2,2-Dimethyl-4-(3-hydroxypropyl)oxetane (4a0)
1
Colorless oil; H NMR: d/ppm = 1.40 (s, 3H), 1.46 (s, 3H), 1.67–1.71 (m, 2H),
1.74–1.85 (m, 2H), 2.08 (dd, J = 10.8, 7.1 Hz, 1H), 2.44 (dd, J = 10.8, 7.6 Hz,
1H), 3.65 (brs, 1H), 3.65 (brt, J = 5.8 Hz, 2H), 4.51–4.65 (m, 1H); 13C NMR:
d/ppm = 28.7, 29.1, 31.0, 35.9, 39.5, 62.7, 73.9, 80.4; HRMS (CI) m/z calcd. for
C8H16O2 (M ? H)? 145.1228, found 145.1237.
6-Methylhept-3-ene-1,6-diol (7)
1H NMR: d/ppm = 1.42 (s, 6H), 2.26 (d, J = 7.0 Hz, 2H), 2.28–2.39 (m, 2H), 2.82
(brs, 2H), 3.64 (t, J = 6.2 Hz, 2H), 5.53–5.72 (m, 2H); 13C NMR: d/ppm = 29.2,
29.2, 30.6, 40.9, 61.7, 70.9, 127.7, 129.3.
6-Methylheptane-1,6-diol (8)
Colorless oil; 1H NMR: d/ppm = 1.21 (s, 6H), 1.34–1.42 (m, 4H), 1.43–1.49
(m, 2H), 1.53–1.65 (m, 2H), 2.35 (brs, 2H), 3.63 (t, J = 6.6 Hz, 2H); 13C NMR:
d/ppm = 24.1, 26.3, 29.2, 29.2, 32.7, 43.8, 62.7, 71.0; HRMS (CI) m/z calcd. for
C8H18O2 (M ? H)? 147.1385, found 147.1395.
Photoreactions in the presence of t-BuNH2 were performed as follows: An
aqueous solution (30 mL) containing 1a–f (0.15 mmol) and acetone (9 mL) was
introduced into a Pyrex test tube of 22 mm diameter. The solution was bubbled with
argon for 10 min, then t-BuNH2 (3 mL) was added. The solution was irradiated for
20–60 min with a 300-W high-pressure mercury lamp at 20 ꢁC. The products were
isolated by column chromatography on silica gel using hexane–ethyl acetate (10:1,
v/v) as eluent. The structures of 7-hydroxyheptan-2-one (3b) [6, 7], 7-hydroxyoc-
tan-2-one (3d) [6], and nonane-2,8-dione (3f) [8, 9] were determined by
comparisons with the spectral data published in literature. The structure determi-
1
nation of 9-hydroxynonan-2-one (3c) was performed by H and 13C NMR spectra.
The structure of 7-oxooctanoic acid (3e) was determined by comparison with the
spectroscopic properties of an authentic sample.
7-Hydroxyheptan-2-one (3b)
1
Colorless oil; H NMR d/ppm = 1.32–1.42 (m, 2H), 1.52–1.66 (m, 4H), 2.15 (s,
3H), 2.46 (t, J = 7.2 Hz, 2H), 2.98 (brs, 1H), 3.62 (t, J = 6.5 Hz, 2H); 13C NMR d/
123

Controlling the photochemical reactions of alkenes
115
ppm = 23.5, 25.3, 29.9, 32.4, 43.6, 62.3, 209.5; HRMS (EI) m/z calcd. for C7H14O2
130.0994, found 130.1044.
9-Hydroxynonan-2-one (3c)
Colorless oil; 1H NMR d/ppm = 1.26–1.37 (m, 6H), 1.50–1.63 (m, 4H), 2.14
(s, 3H), 2.20 (brs, 1H), 2.43 (t, J = 7.3 Hz, 2H), 3.62 (t, J = 6.5 Hz, 2H); 13C NMR
d/ppm = 23.8, 25.6, 29.1, 29.2, 29.8, 32.7, 43.7, 62.8, 209.5; HRMS (EI) m/z calcd.
for C9H18O2 158.1307, found 158.1327.
7-Hydroxyoctan-2-one (3d)
Colorless oil; 1H NMR d/ppm = 1.19 (d, J = 6.3 Hz, 3H), 1.35–1.68 (m, 7H), 2.14
(s, 3H), 2.45 (t, J = 7.3 Hz, 2H), 3.81 (m, 1H); 13C NMR d/ppm = 23.6, 23.7, 25.3,
29.9, 39.0, 43.6, 67.9, 209.1; HRMS (EI) m/z calcd. for C8H16O2 144.1150, found
144.1141.
Nonane-2,8-dione (3f)
White solid, mp 46.0–47.0 ꢁC (from benzene–hexane); 1H NMR d/ppm =
1.22–1.35 (m, 2H), 1.57 (quint., J = 7.5 Hz, 4 H), 2.12 (s, 6H), 2.45 (t,
J = 7.3 Hz, 4H); 13C NMR d/ppm = 23.5, 28.6, 29.8, 43.3, 208.6; HRMS (EI)
m/z calcd. for C9H16O2 156.1150, found 156.1128.
Batch photoreaction of 2a with TEA and tert-butylamine (t-BuNH2)
An MeCN (9 mL) solution of 2a (0.1 mmol) and the amine (1 mL) was poured into
a rectangular quartz glass cell with 20 mm path length, 10 mm width, and 44 mm
height. The solution was bubbled with argon for 3 min and then sealed with a Teflon
septum. Irradiation was performed with a 100-W high-pressure mercury lamp
(UVL-100HA; Riko, Japan) through a glass filter (k [ 250 nm) for 20–110 min at
20 ꢁC. The products were isolated by column chromatography on silica gel using
benzene and benzene–ethyl acetate (10:1, v/v) as eluents. The structures of
3-(diethylamino)-1,1-diphenylbutane (6a) [10], 1-(tert-butylamino)-2,2-diphenyl-
ethane (6b) [11], 1,1-diphenylethane (9) [11], and 2,2,3,3-tetraphenylbutane (10)
[12] were determined by comparisons of the spectral data with those published in
literature. The yields of the products were determined by GLC using heptadecane as
internal standard.
2-(Diethylamino)-3,3-diphenylbutane (5a)
1
Colorless oil; H NMR d/ppm = 0.89 (t, J = 7.0 Hz, 6H), 1.05 (d, J = 7.2 Hz,
3H), 1.70 (s, 3H), 1.82–1.91 (m, 2H), 2.17–2.25 (m, 2H), 3.56 (q, J = 7.2 Hz, 1H),
7.10–7.35 (m, 2H); 13C NMR d/ppm = 10.2, 13.8, 30.0, 44.8, 52.3, 61.6, 125.1,
125.8, 126. 9, 127.6, 128.1, 129.8, 145.3, 150.7.
123

116
T. Yamashita et al.
1-(tert-Butylamino)-1,1-diphenylethane (5b)
1
White solid, mp 48.5–49.5 ꢁC (from benzene–hexane); H NMR d/ppm = 1.02 (s,
9H), 1.70 (s, 1H), 1.90 (s, 3H), 7.12–7.39 (m, 10H); 13C NMR d/ppm = 29.7, 32.8,
51.8, 62.4, 126.0, 127.2, 127.6, 151.8; HRMS (EI) m/z calcd. for C18H23N 253.1830,
found 253.1833.
MCR photoreaction
A schematic drawing of the microchannel reactor (MCR; ICC-SY500) purchased
from the Institute of Microchemical Technology (Kanagawa, Japan) is shown in
Scheme 2. The channel shape was as follows: channel width, 92.3 lm; channel
depth, 38.1 lm; cross-section area, 2.76 9 10-5 cm2; length, 511 mm; channel
volume, 1.41 lL. The MCR was made from Pyrex glass which could transmit light
with k [ 250 nm. The MCR was set in a water bath which was kept at a given
temperature by a temperature-controlling circulator.
An aqueous solution (10 mL) containing 1a (0.15 mmol) and acetone (3 mL)
was prepared and then bubbled with argon for 10 min. The solution was injected
into the MCR by a syringe pump at a rate of 0.71–5.65 lL min-1 under irradiation.
In the case of the photoreaction of 2a, an MeCN (9 mL) solution containing 2a
(0.1 mmol) and TEA (1 mL) was injected into the MCR by a syringe pump at
a rate of 0.14–24.5 lL min-1 under irradiation. The yields of the products were
determined by GLC using heptadecane as internal standard or by GC–MS by
absolute calibration method.
Results
Photoreaction of acetone with alkenols (1a–f) in aqueous solution (reaction A)
Batch photoreaction
Acetone is well known as a highly reactive source which can generate enolate anion,
•
ketyl radical, acetonyl radical (MeCOCH2 ) [10, 13–19], and oxetanes in thermal
Scheme 2 a Top view and b cross section of the MCR: width = 92.3 lm, depth = 38.1 lm,
length = 511 mm, and volume = 1.41 lL
123

Controlling the photochemical reactions of alkenes
117
or photochemical reactions. Therefore, controlling reactivity is a requisite for use of
Me2CO as a C3 source in organic synthesis. Acetonyl radical can be generated
through a hydrogen abstraction from acetone by the triplet acetone (3Me2CO*,
•
Scheme 3, Eq. 2). By using MeCOCH2 , the acetonylation towards water-soluble
pent-4-en-1-ol (1a) was performed in aqueous solution. Mixed solution of Me2CO–
H2O (3:10, 13 mL) containing 1a (0.15 mmol) was irradiated with a high-pressure
mercury lamp in a batch system. After irradiation for 40 min, 8-hydroxyoctan-2-one
(3a) was formed in 66 % yield along with oxetane (4a) and the isomer (4a0) in 10 %
yield (Table 1, entry 4). When the photoreactions of Me2CO (2.33 mmol) with 1a
(29.5 mmol) were performed for 23 h without water as a co-solvent [20], 3a was not
formed, but complicated products involving 4a, 4a0, 6-methylhept-3-ene-1,6-diol
(7), 6-methylheptane-1,6-diol (8), and many other products were formed. The
irradiation of Me2CO–benzene (3:10, 13 mL) solution of 1a (0.15 mmol) for
40 min resulted in the recovery of 1a (entry 1). Though the photoreaction in hexane
proceeded smoothly, 3a and 4a were not formed (entry 2). Photoreaction in MeCN
resulted in the consumption of 1a only (entry 3). These results indicate that water is
requisite for efficient formation of 3a. The present method was applied to direct
Me
OH
OH
OH
O
Me
∗
∗
3Me2CO
+
+
(1)
+
+
Me
O
1a
4a Me
4a'
O
3Me2CO
Me C CH2
(2)
(3)
Me2CO
Me2C OH
13
14
OH
Me
OH
Me
13
1a
+
O
O
14
3a
Me2CO
Me Me
OH
OH
Me Me
OH
7
dispro.
OH
(4)
(5)
14
13
+
+
1a
13
Me Me
15
OH
OH
8
O
Me
Me
O
16
Scheme 3 Reaction mechanism for reaction A
123

118
T. Yamashita et al.
photoacetonylation of water-soluble alkenols (1b–f) in aqueous solution
(Scheme 4). The products (3b–f) were obtained in fairly good yields.
MCR photoreaction
The MCR photoreaction of acetone with 1a was performed under continuous flow.
Irradiation of the running Me2CO–H2O (3:10, 13 mL) solution containing 1a
(0.15 mmol) at 0.71–5.65 lL min-1 in the MCR gave mainly 4a (entries 9–11 in
Table 1). The formation of 3a was diminished in the MCR photoreaction, suggesting
•
that MCR photoirradiation reduced the generation of MeCOCH2 . Moreover, the
yields of 4a increased with increases in the flow rate of the MCR, as shown in Fig. 1.
Photoreaction of 1,1-diphenylethene (2a) with triethylamine (TEA) (reaction B)
Batch photoreaction
Photoreaction of 1,1-diphenylethene (2a) with triethylamine (TEA) was performed
in a batch reaction (Scheme 5). The product yields based on the consumed 2a are
Table 1 Photoreactions of acetone with 1a in MCR and in batch (reaction A)
Entry
Reactor
Solvent
Time (min)
Conv. (%)
Product yield (%)a
4 (4a:4a0)
3a
1
Batchb
Batchb
Batchb
Batchb
Batchb
Batchc
Batchc
Batchc,d
MCRe
MCRe
MCRe
Benzene
Hexane
MeCN
H2O
40
12
63
0
0
0
0
2
40
3
40
53
0
0
4
40
100
17
10 (5:1)
14 (7:1)
26 (6:1)
17 (5:1)
17 (6:1)
24 (9:1)
42 (9:1)
67 (10:1)
66
60
21
16
33
5
5
H2O
7
6
H2O
12
100
100
100
92
7
H2O
40
8
H2O
24
9
H2O
2.0f
1.0g
0.25h
10
11
H2O
36
6
H2O
10
8
Irradiation of Me2CO–solvent (3:10, 13 mL) solution containing 1a (0.15 mmol)
a
Yields based on consumed 1a given by quantitative analyses by GLC (entries 1–8) and GC–MS
(entries 9-11)
b
In a Pyrex test tube (diameter 22 mm)
c
In a cell (thickness 1.5 mm)
d
Irradiation was performed through a filter containing Me2CO–H2O (3:10, 13 mL) solution containing
1a (thickness 5 mm)
e
In a MCR (depth 0.038 mm, width 92.3 mm, length 511 mm)
Flow rate = 0.71 lL min-1
f
Flow rate = 1.41 lL min-1
g
Flow rate = 5.65 lL min-1
h
123

Controlling the photochemical reactions of alkenes
119
Me
Scheme 4 The
photoacetonylations of
1a–f with acetone in the
presence of t-BuNH2
1a-f
hν
Me
Me
+
R
R
O
O
3a-f
a: R = CH2CH2OH (72%), b: R = CH2OH (56%),
c: R = CH2CH2CH2OH (76%), d: R = CH(Me)OH (47%),
e: R = CH2CO2H (77%), f: R = CH2COMe (49%)
Fig. 1 Effect of flow rate on
yields of 3a (open circles) and 4
(filled circles) in the MCR
photoreactions of 1a (open
squares) in Me2CO–H2O.
Product yields based on
consumed 1a
presented in Table 2. Markovnikov-type adduct (5a) and anti-Markovnikov-type
adduct (6a), 1,1-diphenylethane (9), and 2,2,3,3-tetraphenylbutane (10) were
formed in 14, 19, 44, and 11 % yields, respectively (entry 1 in Table 2). The
5a/6a ratio was 0.72. The addition of 0.7 mL H2O to the reaction system enhanced
the formation of 6a, but decreased the formation of 5a (entry 2). The photoreaction
in THF required longer irradiation to consume 2a, compared with the case of MeCN
(entry 3). The addition of n-Bu4NBF4 shortened the irradiation time (entry 5). In the
case of photoreaction in hexane, the 5a/6a ratio was enhanced to 2.10 (entry 6).
Figure 2 shows the time-conversion plots of the yields of 5a, 6a, and 9 in the batch
photoreaction in hexane. The formations of 5a and 9 were dependent upon the
irradiation time, whereas the formation of 6a stopped at a low yield. At low
conversion of 2a in hexane, the 5a/6a ratio reached 5.47 (entry 7).
MCR photoreaction
The photoreaction of 2a with TEA was performed using the MCR. A reactant was
introduced to the MCR by a microsyringe pump at a constant flow rate
(0.56 lL min-1) under irradiation. The photoreaction of 2a with TEA in MeCN
123

120
T. Yamashita et al.
Ph
Ph Me
Ph NEt2
Me
Ph
Ph Me
h
ν
Me
Ph
Ph
+
+
+
+
Et3N
Ph
Ph
Ph
Ph
Ph
Me
Me
NEt2
2a
5a
6a
9
10
Scheme 5 The photoreaction of 2a with TEA
Table 2 Photoreactions of 2a with TEA (reaction B)
Entry
Reactor
Solvent
Time (min)
Conv. (%)
Product yield (%)a
Ratio
5a
6a
9
10
5a/6a
1
Batch
Batch
Batch
Batch
Batch
Batch
Batch
MCR
MCR
MCR
MCR
MeCN
MeCN–H2Ob
60
93
90
51
97
96
89
23
96
39
34
21
14
7
19
34
25
24
19
11
9
44
32
28
29
28
49
27
45
25
27
35
11
15
11
16
13
4
0.72
0.19
0.38
0.39
0.49
2.10
5.47
1.65
10.7
11.5
16.9
2
60
3
THF
THF–H2Oc
THF
300
45
10
10
9
4
5d
45
6
Hexane
110
20
23
51
24
33
46
61
7
Hexane
0
8e
9e
10e
11e,f
MeCN
2.5
2.5
2.5
2.5
16
3
9
Hexane
2
Cyclohexane
Cyclohexane
4
0
4
0
Irradiation (k [ 250 nm) of a solution (9 mL) containing 2a (0.1 mmol) and TEA (1 mL) at 20 ꢁC
a
Yields based on consumed 2a
b
MeCN–H2O = 8.3:0.7
c
THF–H2O = 8:1
d
In presence of Et4NBF4 (0.1 mmol)
Flow rate = 0.56 lL min-1
e
f
Irradiation of cyclohexane solution (9 mL) of 2a (0.05 mmol) and TEA (1 mL) at 10 ꢁC
gave 5a more than 6a. The product distribution was different from the batch
photoreaction (entry 8). Moreover, the 5a/6a ratio was higher in hexane solution,
although the conversion of 2a was low (entry 9). Figure 3 shows the time-conversion
plots in the MCR photoreaction in hexane. With increases in conversion, the yield of
6a increased linearly. With increases in irradiation time, the yields of 5a increased
until the yield reached 20 %. After that, the yield kept constant. Therefore, it is
expected that 5a is decomposed during the photoreaction. The photolysis of 5a was
carried out by the MCR irradiation of 5a (0.1 mmol) in the presence of TEA (1 mL)
in hexane (9 mL). Apparently, the 5a was converted to 9 accompanying the
formation of a small amount of 2a and 6a, as shown in Fig. 4. With cyclohexane
instead of hexane, the 5a/6a ratio was increased to 11.5 (entry 10).
Figure 5 shows the temperature dependence of the product yields in the MCR
photoreactions in cyclohexane at 0.56 lL min-1. The yields of 5a and 6a were
dependent upon reaction temperature, but the yield of 9 was constant. Irradiation at
10 ꢁC made the yield of 5a and the 5a/6a ratio higher (entry 11).
123

Controlling the photochemical reactions of alkenes
121
Fig. 2 Time-conversion plots
for yields of 5a (filled circles),
6a (open circles), and 9 (filled
triangles) in the batch
photoreactions of 2a (open
squares) in hexane
Batch and MCR photoreactions of 2a with tert-butylamine (t-BuNH2)
The photoreaction of 2a with TEA caused the proton transfer from a-C–H of TEA to
2a. In the case of tert-butylamine (t-BuNH2) not having a-C–H, the proton transfer
•
from N–H is expected to generate the tert-butylaminyl radical (t-BuNH ). Batch
photoreaction of 2a (0.1 mmol) with t-BuNH2 (5 mL) without a solvent gave many
complicated products involving 5b, 9, and 10 in 31, 2, and 15 % yields, respectively
Fig. 3 Time-conversion plots
of the yields of 5a (filled
circles), 6a (open circles), and 9
(filled triangles) in the MCR
photoreactions of 2a (open
squares) in hexane
123

122
T. Yamashita et al.
Fig. 4 Time-conversion plots
for the yields of 2a (open
squares), 6a (open circles),
9 (filled triangles) in the MCR
photoreaction of 5a (filled
circles) in hexane
Fig. 5 Temperature effects on
the yields of 5a (filled circles),
6a (open circles), and 9 (filled
triangles) in the MCR
photoreactions of 2a (open
squares) in cyclohexane; the
flow rate was 0.56 lL min-1
Product yields based on
consumed 2a
.
(Scheme 6, entry 3 in Table 3). The use of H2O as a co-solvent increased the yield
of the anti-Markovnikov-type adduct (6b) (entry 5).
MCR photoreaction enhanced the selectivity of the Markovnikov-type adduct
(5b) (entry 4). Figure 6 shows the time-conversion plots of the product yields for
the MCR photoreaction of 2a with t-BuNH2. The formation of 5b decreased with
increasing irradiation time. The MCR irradiation of 5b (0.04 mmol) in t-BuNH2
123

Controlling the photochemical reactions of alkenes
123
Ph Me
Ph
Ph
h
ν
2a
+
t-BuNH2
+
NHBut
+
9
+
10
+
CH2CN
Ph
NHBut
Ph
Ph
5b
6b
11
Scheme 6 The photoreaction of 2a with t-BuNH2
Table 3 Photoreactions of 2a in the presence of t-BuNH2
Entry
Reactor
Additive
Time (min)
Conv. (%)
Product yield (%)a
5b
6b
9
10
15
11
1b
2b,c
Batch
MCR
Batch
MCR
Batch
MCR
MeCN
MeCN
None
None
H2O
90
68
65
29
29
31
32
15
44
31
51
20
45
0
0
3
3
2
5
6
4
18
0
0.25
30
21
15
19
5
3
4d
5e
6c,e
0
–
0.13
60
0
–
16
0
–
H2O
0.25
10
–
Irradiation (k [ 250 nm) of t-BuNH2 (5 mL) containing 2a (0.1 mmol) at 5 ꢁC
a
Yields based on consumed 2a
b
In presence of MeCN (5 mL)
Flow rate = 6.13 lL min-1
c
Flow rate = 12.3 lL min-1
d
e
In presence of H2O (1 mL)
Fig. 6 Time-conversion plots
of the yields of 5b (filled
circles), 6b (open circles),
9 (filled triangles), and 10 (open
diamonds) in the MCR
photoreactions of 2a (open
squares) in t-BuNH2
(5 mL) for 15 s induced the decomposition of 52 % of 5b and the formation of
10 % yield of 2a and 20 % yield of 10. However, the batch irradiation of 5b for
90 min resulted in low decomposition (11 %).
123

124
T. Yamashita et al.
Discussion
Reaction mechanisms of reactions A and B
Under irradiation by k [ 300 nm light, Me2CO was selectively excited to form the
excited triplet state of Me2CO (3Me2CO*). The resulting 3Me2CO* reacted directly
with 1a to give oxetane (4a) (Eq. 1 in Scheme 3). Simultaneously, 3Me2CO*
abstracted an H atom from Me2CO in the ground state to generate the acetonyl
•
•
radical (MeCOCH2 , 13) and 2-hydroxy-2-propyl radical (Me2C(OH) , 14) (Eq. 2).
The addition reaction of 13 to 1a followed by an H-atom transfer from 14 gave 3a
(Eq. 3). The formations of 7 and 8 took place through the disproportionation of 15
(Eq. 4). The formation of 13 was confirmed by the formation of hexane-2,5-dione
(16) by irradiation of the aqueous solution of Me2CO (Eq. 5).
In the case of 2a, almost all incident light (k [ 250 nm) was absorbed by 2a
under the reaction conditions. Therefore, the initiation step should be a hydrogen
transfer from amines to the excited singlet states of 2a, resulting in the formation of
17 and 18 (Scheme 7). The free energy changes (DG) for the electron transfer from
TEA and t-BuNH2 to 2a* were calculated to be negative (-1.20 eV) and slightly
negative (-0.099 eV) by the Rehm–Weller equation, respectively [21–23]. In the
case of TEA, the proton transfer might proceed through exciplex formation, as has
been previously reported in the photoadditions of TEA to stilbenes and styrenes
[24–26]. The coupling reaction between 17 and 18 in a solvent cage gave a
Markovnikov-type adduct, 5a (Scheme 7). The formation of 6a would proceed
through the addition of 18 escaped from the solvent cage to the ground state of 2a to
generate 19 followed by a hydrogen abstraction from 17 [10, 18]. The escape of 18
from the solvent cage was accelerated in MeCN due to the well-known fact that
cage escape is favored in a polar solvent. Also, the addition of water increased the
formation of 6a. Indeed, the deuterio-incorporated 9 (Ph2CHCH2D) was obtained in
39 % yield in MeCN–D2O. This fact indicates that the protonation of 2a•2 with
∗
Ph
Ph
Ph Me
Me
+
Et2NCHMe
Et3N
Ph
Ph
Ph
Me
in cage
∗
NEt2
5a
2a
17
18
2a
Ph NEt2
Ph NEt2
17
2a
+
+
Ph
Me
Ph
Me
19
6a
∗
Ph
Ph Me
Ph Me
δ -
hν
Me
Me
5a
+
H2C=CHNEt2
δ+
Ph
Ph
Ph
Me
9
NEt2
NEt2
Scheme 7 Reaction mechanism for reaction B
123

Controlling the photochemical reactions of alkenes
125
H2O accelerates the escape of TEA•? before the proton transfer between initially
formed ion pairs (TEA•?/2a•-) [24]. Outside of the cage, 18 was formed by the
deprotonation of TEA•? by TEA. We have previously reported that the addition of
such salts as Et4NBF4 and Et4NOAc was effective to prevent back electron transfer
from the radical anion of electron acceptor to the radical cation of the substrates
[27]. Therefore, the addition of Et4NBF4 was effective to prevent the competitive
back electron transfer from TEA•?/2a•-. Compound 9 was obtained by the
disproportionation of 17 or hydrogen abstraction by 17 from amine. Moreover,
photoexcitation of 5a would make the C–C bond electronically polarized to give 9,
as previously reported (Scheme 7; Fig. 4) [28].
The effect of the MCR
When 3.14 mol dm-3 of Me2CO (e313 = 7 mol-1 dm3 cm-1 at 313 nm) was
irradiated in a test tube of 22 mm diameter, it was thought that Me2CO entirely
absorbed the incident light near the vessel wall since the absorbance (A) was
3
calculated to be up to 48. Therefore, Me2CO* preferentially reacted with Me2CO
in the ground state to generate 13 and 14 since the concentration of Me2CO was
3
higher than those of Me2CO* as well as 1a.
In the MCR where the light-path length was up to 0.0381 mm, on the other hand,
almost all Me2CO molecules were excited in a narrow channel since the A was
calculated to be up to 0.08. Since the concentration of Me2CO in the ground state
under MCR irradiation conditions would be much lower than the concentration of
3
1a, Me2CO* preferentially reacted with 1a rather than Me2CO in the ground state
to form oxetane 4a in the MCR. Similar effects would also occur in reaction B. The
MCR photoreaction enhanced the concentration of the excited state of 2a to lower
the possibility of the attack of the ground state of 2a to a 17/18 radical pair. As a
result, 5a was formed predominantly over 6a in the MCR photoreaction.
Conclusions
Oxetanes have been extensively used for preparation of a wide range of polymeric
`
materials as well as epoxides [29]. The Paterno–Bu¨chi reaction is a convenient
method for preparation of oxetanes [20]. This reaction is usually performed in
nonpolar solvents or in neat. The use of the MCR was effective for oxetane
formation rather than acetonylation in an aqueous solution. The MCR reaction
slowed down the reaction of the excited acetone with the ground state of acetone.
Also, the MCR photoreaction of 2a with TEA gave 5a more predominantly, which
was produced by a coupling reaction of a 17/18 radical pair in a solvent cage. The
MCR photoreaction prevented the attack of the ground state of 2a to the 17/18
radical pair.
Thus, the MCR photoreaction prevented the reaction of the excited state of the
substrate with the ground state of the same substrate. The operation of the light-path
length effects of a microchannel reactor was confirmed in reactions A and B.
123

126
T. Yamashita et al.
References
1. J. Matsumoto, Y. Yoshinaga, A. Hamasaki, T. Kawasaki, T. Yamashita, T. Shiragami, M. Yasuda,
Bull. Chem. Soc. Jpn. 84, 1130 (2011)
2. S. Suga, A. Nagai, J. Yoshida, Chem. Commun. 354 (2003)
3. Y. Matsushita, N. Ohba, S. Kumada, T. Suzuki, T. Ichimura, Catal. Commun. 8, 2194 (2007)
4. H. Mukae, H. Maeda, S. Nashihara, K. Mizuno, Bull. Chem. Soc. Jpn. 80, 1157 (2007)
5. J. Tsuji, M. Kaito, T. Takahashi, Bull. Chem. Soc. Jpn. 51, 547 (1978)
6. G. Bianchinia, M. Crucianellia, C. Canevalib, C. Crestinic, F. Morazzonib, R. Saladinod, Tetrahedron
62, 12326 (2006)
7. K. Sato, M. Aoki, J. Takagi, K. Zimmermann, R. Noyori, Bull. Chem. Soc. Jpn. 72, 2287 (1999)
8. C. Peppe, E.S. Lang, G.N. Ledesma, L.B. Castro, O.S.R. Barros, P.A. Mello, Synlett. 3091 (2005)
9. I.F. Belskii, Z.K. Minashkina, Russ. Chem. Bull. 19, 2195 (1970)
10. T. Yamashita, M. Yasuda, M. Watanabe, R. Kojima, K. Tanabe, K. Shima, J. Org. Chem. 61, 6438
(1996)
11. T. Yamashita, K. Shiomori, M. Yasuda, K. Shima, Bull. Chem. Soc. Jpn. 64, 366 (1991)
12. W.D. Hounshell, D.A. Dougherty, J.P. Hummel, K. Mislow, J. Am. Chem. Soc. 99, 1916 (1977)
13. K.G. Stirk, H.I. Kenttamaa, J. Am. Chem. Soc. 113, 5880 (1991)
14. W. Reusch, J. Org. Chem. 27, 1882 (1962)
15. S.H. Schroeter, Liebigs Ann. Chem. 1890 (1974)
16. P.G. Gassman, J.L. Smith, J. Am. Chem. Soc. 48, 4439 (1983)
17. K.M. Majerski, D. Pavlovic, M.S. Kulyk, J. Org. Chem. 58, 252 (1993)
18. T. Yamashita, M. Watanabe, R. Kojima, T. Shiragami, K. Shima, M. Yasuda, J. Photochem.
Photobiol. A Chem. 118, 165 (1998)
19. Y. Shiraishi, D. Tsukamoto, T. Hirai, Org. Lett. 10, 3117 (2008)
20. G. Bu¨chi, C.G. Inman, E.S. Lipinsky, J. Am. Chem. Soc. 76, 4327 (1954)
21. G.J. Kavarnos, N.J. Turro, Chem. Rev. 86, 401 (1986)
22. R.S. Ruoff, K.M. Kadish, P. Boulas, E.C.M. Chen, J. Phys. Chem. 99, 8843 (1995)
23. D. Rehm, A. Weller, Isr. J. Chem. 8, 259 (1970)
24. S. Toki, H. Shintaro, S. Takamuku, H. Sakurai, Nippon Kagaku Kaishi. 152 (1984)
25. F.D. Lewis, T.-I. Ho, J. Am. Chem. Soc. 99, 7991 (1977)
26. F.D. Lewis, D.M. Bassani, J. Photochem. Photobiol. A Chem. 81, 13 (1994)
27. T. Yamashita, T. Tsurusako, N. Nakamura, M. Yasuda, K. Shima, Bull. Chem. Soc. Jpn. 66, 857
(1993)
28. D.D.M. Wayner, L. Gravelle, Tetrahedron Lett. 29, 431 (1988)
29. H. Hailesa, J. Behrendtb, in Comprehensive Heterocyclic Chemistry III, vol. 2, ed. by A. Katritzky,
C. Ramsden, E. Scriven, R. Taylor (Elsevier, Amsterdam, 2008), pp. 321–364
123