Continuous Flow Liquid-Liquid Separation Using a Computer-Vision Control System: The Bromination of Enaminones with N -Bromosuccinimide

Article abstract of DOI:10.1055/s-0035-1560975

Incorporating open-source software components (Python, OpenCV), a computer-vision system was used to control the interface level in a gravity-based inline liquid-liquid separation device. This was used in the continuous flow bromination of a series of enaminone substrates. The main byproduct of the reaction, succinimide, was efficiently extracted into the aqueous stream, providing clean products without the need for further purification.

Full text of DOI:10.1055/s-0035-1560975

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2016, 27, 164–168
letter
164
M. O’Brien, D. Cooper
Letter
Synlett
Continuous Flow Liquid–Liquid Separation Using a Computer-
Vision Control System: The Bromination of Enaminones with
N-Bromosuccinimide
inline
liquid-liquid
extraction
NHR3
O
Matthew O’Brien*
Dennis Cooper
R1
R2
NHR3
R2
O
Lennard-Jones Building, School of Physical and Geographical Sciences,
Keele University, Keele, Borough of Newcastle-under-Lyme, Staffordshire,
ST5 5BG, UK
•
R1
O
O
Br
m.obrien@keele.ac.uk
N
Br
12 examples
76–97% yield
O
Dedicated to Professor Steven V. Ley, mentor and friend, on the occasion of
his 70th birthday
N
H
O
Received: 02.10.2015
Accepted after revision: 04.11.2015
Published online: 07.12.2015
mised once.4 Whilst reaction processes themselves are
clearly important to any chemical transformation, the over-
all synthesis operation can involve many other post-reaction
processes (including workup, isolation, and purification).
The development of flow chemistry ‘equivalents’ to these
batch protocols is clearly an important area of research and
currently enjoys significant activity. This is particularly im-
portant in the context of truly continuous flow multistep
processes, where the byproducts from one process must be
removed before the flow stream enters a downstream reac-
tion zone, where they may be incompatible with the in-
tended chemical transformation. In terms of workup and
isolation, the incorporation into flow-chemistry systems of
cartridges of solid-supported reagents for in-line ‘scaveng-
ing’, ‘catch-and-release’, or ‘phase switching’ has been a
hugely successful tactic.5 This technique is particularly suit-
able for small-scale reactions. On larger scales, however, the
need to use greater quantities of solid-supported reagent
leads to issues of cost and also to problems of material dis-
persion.6
DOI: 10.1055/s-0035-1560975; Art ID: st-2015-d0789-l
Abstract Incorporating open-source software components (Python,
OpenCV), a computer-vision system was used to control the interface
level in a gravity-based inline liquid–liquid separation device. This was
used in the continuous flow bromination of a series of enaminone sub-
strates. The main byproduct of the reaction, succinimide, was efficient-
ly extracted into the aqueous stream, providing clean products without
the need for further purification.
Key words continous flow, liquid–liquid separation, bromination,
enaminone, open source, computer-vision, OpenCV, Python
In recent years, flow chemistry has emerged as an at-
tractive alternative to traditional batch protocols for chemi-
cal synthesis.1 Although batch chemistry is perfectly suit-
able for many synthesis processes, flow chemistry often of-
fers significant advantages. As reactants and reagents are
continuously pumped through a relatively small reaction
zone, only a small quantity of material is being processed at
any one time. Reactions are then scaled over time (or
through parallelisation) rather than by increasing dimen-
sion. This provides an enhanced safety profile and is partic-
ularly important for reactions that involve the build-up of
hazardous, unstable, or explosive intermediates, or that use
hazardous conditions (e.g., high temperatures or pres-
sures).2 Another related issue is the scale-variant nature of
several important reaction parameters (e.g., surface-area to
volume ratio). These can play a crucial role in the outcome
of a reaction as they determine the rate of heat, light, and
material transfer across interfacial boundaries.3 As a result,
for many batch processes, scaling up can often involve sig-
nificant re-optimisation. As the reaction/contact zones of a
flow-chemistry system are essentially fixed, this leads to ef-
ficient, scale-invariant processes that need only be opti-
aqueous
out
control
system
....
....
aqueous
in
....
interface
detection
mixer
organic
in
(CH2Cl2)
separator
organic
out (CH2Cl2)
Figure 1 General schematic for a gravity-based inline liquid–liquid
separator
In batch chemistry, gravity-based liquid–liquid ex-
traction is one of the most commonly used workup/purifi-
cation protocols, often providing significant purification
using very inexpensive reagents and apparatus. The incor-
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 164–168

165
M. O’Brien, D. Cooper
Letter
Synlett
poration of inline liquid–liquid extraction into continuous
flow chemical systems can provide a cost-effective purifica-
tion stage which does not suffer from depletion (as it is con-
tinually replenished) and potentially has scale-invariant
dispersion properties. There are two general strategies for
achieving this. The selective wetting of materials (such as
expanded, porous PTFE) leads to membranes which are es-
sentially impermeable to aqueous solutions but highly per-
meable to immiscible organic solvents and solutions. This
effect has been used successfully in a number of devices
and systems.7 Alternatively, we have been interested in the
development of continuous flow gravity-based separations
(see Figure 1 for a general schematic). In general, the organ-
ic flow stream is mixed in with an aqueous ‘quench/ex-
traction’ stream. After mixing, the combined stream enters
a separation chamber where the dense phase leaves
through the bottom and the light phase leaves through the
top. As variation in material transfer between the phases
will lead to variation in their relative rate of accumulation
in the separation chamber, dynamic positional control of
the liquid–liquid interface (which must remain within the
separation chamber) is vital. To achieve this, we developed
a computer-vision8 video system which monitors the posi-
tion of a coloured ‘float’ that sits at the liquid–liquid inter-
face (having a density between that of the two phases).9
This positional information is then fed back to control the
speed of the pump which extracts the aqueous phase. A
similar system has recently been used by the Ley group as
part of a multistep continuous flow synthetic sequence.10
An alternative approach to interface detection using imped-
ance has also been used in continuous flow systems.11
Here we report the use of this system in the continuous
flow bromination of a series of enaminones using N-bromo-
succinimide (NBS) as the halogenating agent.12 The major
byproduct from the reaction, succinimide, can be extracted
using aqueous alkali solution. The enaminone substrates
3a–l were synthesised using a straightforward condensa-
tion reaction between the symmetrical diketones dimedo-
ne (1a), cyclohexane-dione (1b), or pentane-2,4-dione (1c),
and a series of amines 2a–g (Scheme 1,Figure 2). Following
reflux in a toluene–ethanol mixture,13 all but one of the
enaminone products solidified after removal of solvent and
were easily purified by recrystallisation from toluene. Com-
pound 3l required the excess amine to be removed by aque-
ous extraction (saturated ammonium chloride) prior to re-
crystallisation. The yields were moderate to very high, and
the products were obtained in high purity (by NMR analy-
sis). The flow-chemical apparatus setup used for the bromi-
nation reactions is shown in Figure 3. The starting material
and NBS were introduced into separate flow streams via in-
jection loops. Dichloromethane was used as the organic sol-
vent for all reactions. The volume of the NBS loop used
was longer than that of the starting material. This was to
allow for a slight ‘overlap’ in the ‘pulses’ which met at the
t-junction, ensuring that the starting material would al-
ways be accompanied by NBS during each run. After the t-
junction, a residence loop controlled (along with the flow
rate) the amount of time that the mixture had to react. All
reactions were carried out at room temperature.
O
O
R3
R2
R1
R2
i) toluene-EtOH
(20:1), reflux
O
HN
1a–c
R1
+
ii) recrystallisation
from toluene
R3
3a–l
H2N
2a–g
Scheme 1 Condensation reaction between diketones 1a–c and
amines 2a–g to form enaminones 3a–l
Following reaction, a second t-junction allowed the in-
troduction of an aqueous reagent stream. An active mixer
(several small PTFE-coated stirrer bars in a glass column
over a magnetic stirrer) then facilitated rapid mixing of the
two phases, before settling and entry into the separation
vessel (made from the barrel of a 5 mL disposable syringe).
A webcam pointed horizontally at the separation chamber
O
O
Me
Me
Me
Me
N
N
N
H
H
3b
3a
89%
86%
O
O
Me
Me
Me
Me
Me
Me
N
N
H
H
3c
3d
93%
97%
O
O
Br
Me
Me
Me
Me
N
N
H
H
3f
87%
3e
88%
Cl
O
O
Me
Me
O
N
N
H
H
3g
3h
91%
82%
O
O
N
N
H
H
3j
82%
3i
Cl
71%
O
HN
O
HN
Me
Me
Cl
Me
3l
Me
3k
50%
89%
Figure 2 Enaminones formed using the condensation reaction shown
in Scheme 1
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 164–168

166
M. O’Brien, D. Cooper
Letter
Synlett
was used to feed a video stream to the control computer.
This used a Python script, incorporating the OpenCV li-
brary,14 which permitted ‘observation’ of the vertical posi-
tion of the interfacial float (see Supporting Information file
for details of all scripts used).15 For this work, we found that
a float made by fusing a mixture of 210 mg cut from the
green plunger of a Norm-Ject 1 mL PP/PE disposable syringe
(which is less dense than water) and 67 mg cut from a
green Keck clip (which is denser than dichloromethane)
was able to maintain its position at the liquid–liquid inter-
face under all conditions used. The two pieces of plastic
were fused by holding them together with a pair of twee-
zers and heating with a heat gun until they melted. The co-
lour that the computer ‘looks for’ could be selected by the
user at the start of each run (by selecting an object on the
screen). The script essentially filters the images it receives
based on the hue value (HSV image encoding) of the select-
ed colour (within certain tolerances). Some simple process-
ing (involving dilation and erosion) removes ‘noise’, and the
position of the centroid of the largest object with matching
hue is determined. The control computer was interfaced to
the ‘aqueous out’ pump. As the pump has a maximum
speed (in this case 9.9 mL/min) and cannot pump back-
wards, a true ‘proportional’ response to positional error is
not appropriate. The algorithm used generated a pumping
speed of (2c)3.32 mL/min, where c is the fractional distance
along the line between the upper and lower bounds. This
way, the speed would be zero when the float was at (or
above) the upper bound, 1.0 mL/min when the float was
halfway between the upper and lower bounds, and at max-
imum speed (9.9 mL/min) when the float was at (or below)
the lower bound. Although it would be possible to scale the
response curve for different rates of the ‘aqueous in’ pump,
we found that this simple setup provided a robust automat-
ed system which adequately maintained the position of the
float within the desired bounds, using a wide range of
‘aqueous in’ flow rates.
The control script allows the user to manually select the
upper and lower bounds (from a single video frame) at the
start of operation. For all the reactions described below, ap-
proximately the same upper and lower bounds were chosen
for each run and were around 1 mL apart, with the lower
bound around 2 mL above the bottom of the vessel. Follow-
ing a brief survey of conditions, using 3a as our initial test
substrate, we found that concentrations of 0.10 M and
0.106 M, for the starting material and NBS, respectively,
with flow rates of 1.0 mL/min each and a residence loop of
4.7 mL (2 min 21 s reaction time) led to complete conver-
sion of starting material into product. The NBS reagent loop
(4.1 mL) was ‘injected’ into the flow stream 20 seconds be-
fore the starting material loop (3.0 mL), providing an over-
lap of 20 seconds at the start and 46 seconds at the end of
the run. For the aqueous extraction, the use of a 0.1 M solu-
tion (in each) of sodium thiosulfate and potassium carbon-
O
NHR3
R2
3a–l
Na2S2O3
K2CO3
(0.1 M)
(0.10 M CH2Cl2)
3.0 mL
R1
aqueous
in
1.0
mL/min
4.0 mL/min
CH2Cl2
O
(0.106 M CH2Cl2)
N
Br
NBS
4.7 mL
(t = 2.35 min)
O
1.0
mL/min
4.1 mL
CH2Cl2
O
aqueous out
mixer
N
H
O
webcam
separation
vessel
Python
O
NHR3
R2
OpenCV
R1
....
....
....
Br
4a–l
Figure 3 Apparatus schematic for the continuous flow bromination of
enaminones 3a–l
ate, injected at a flow rate of 4.0 mL/min cleanly extracted
all of the succinimide byproduct. With these conditions, the
yield of 4a was almost quantitative. Pleasingly, the yields for
the continuous flow bromination of a series of substrates,
using the same conditions, were in general very high (Fig-
ure 4). Products were obtained in high purity after removal
of solvent on a rotary evaporator. Shown in Figure 5 are rel-
evant sections of the NMR spectrum for compound 4f pro-
duced with (red spectrum) and without (black spectrum)
the aqueous extraction step.
As can be seen, no succinimide can be observed in the
extracted sample. Clearly, some dispersion will take place in
the heavy phase of the separation vessel. To avoid any loss
of material, or cross contamination between runs, we col-
lected the output from each reaction for 20 minutes. As
several of the product solutions had a significant yellow co-
louration, this provided a useful visual indication/confirma-
tion of the time distribution of product concentration in the
separation vessel. Shown in Figure 6 are images from
screengrabs at various times during the continuous flow
synthesis of 4d. The processed images (ii) show (in white)
what the computer identifies as having the ‘correct’ hue, as
selected by the user. The centroid of the identified float is
shown as a small white square in (i), and the upper and
lower bounds are shown as small blue circles. As can be
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 164–168

167
M. O’Brien, D. Cooper
Letter
Synlett
O
O
Br
Br
Me
Me
Me
N
N
H
N
Me
H
4a 97%
4b 94%
O
O
Br
Br
Me
Me
Me
Me
Me
N
H
N
Me
H
4c 82%
4d 76%
O
O
Br
Br
Br
Me
Me
Me
N
N
H
Me
H
Cl
4e 85%
4f 85%
O
O
Br
Br
Me
Me
O
N
N
H
H
4g 84%
4h 78%
O
O
Br
Br
N
H
N
H
Cl
4i 93%
4j 90%
O
HN
O
HN
Figure 5 1H NMR spectrum of the reaction product for the continuous
flow bromination of 3f with (red) and without (black) aqueous ex-
traction
Me
Me
4k 91%
Cl
Me
Me
4l 83%
Br
Br
Figure 4 Products 4a–l and isolated yields for the continuous flow bro-
mination of enaminones 3a–l; all products were analytically pure (by
NMR spectroscopy)
ration of the volatile dichloromethane on standing). At the
end of each day, the ‘aqueous-out’ pump had distilled water
pumped through it to avoid any possible settling/build-up
of particulates and to prevent possible corrosion.
seen, despite significant colouration that builds up during
the run, as well as some cloudiness in the upper (aqueous)
phase, the computer-vision system is able to distinguish
and keep track of the green float during the entire run (the
size of the float is magnified by the cylindrical vessel, mak-
ing it appear larger in these images than it actually is).
As the float still appears to be the only ‘green’ object in
the image, the system is able to operate properly with sig-
nificant colouration using suitable tolerance parameters.
Clearly, if the colouration became strong, or if the solutions
are themselves the same colour as the float, then the sys-
tem might not be able to properly locate the float. It is note-
worthy that no observable chemical cross contamination
was observed (by NMR spectroscopy) during these reac-
tions, even though the system was never dismantled
throughout the entire study. Several consecutive runs were
carried out in each laboratory session. At the start of each
set of runs, the system was ‘primed’ for a few minutes by
running dichloromethane and aqueous solution through to
remove any ‘air’ from the system/pumps (caused by evapo-
Figure 6 Camera images (i) and corresponding processed images (ii)
showing a close up of the separator vessel at: a) 0 min b) 5 min 28 s c) 7
min 43 s d) 10 min 17 s.
In conclusion, we have successfully used an inline liq-
uid–liquid extraction system, incorporating a computer-vi-
sion control system, for the continuous flow bromination of
a series of enaminones.16 The reactions all proceeded with
high chemical yield and, importantly, the extraction system
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 164–168

168
M. O’Brien, D. Cooper
Letter
Synlett
completely removed the succinimide byproduct to afford
analytically pure products. We are currently incorporating
this system into several multistep flow syntheses. Addition-
ally, we are quantitatively investigating the dynamic perfor-
mance of the system under a range of chemical and physical
conditions and will report our findings in due course.
V. Synthesis 2011, 1157. (d) Polyzos, A.; O’Brien, M.; Petersen, T.
P.; Baxendale, I. R.; Ley, S. V. Angew. Chem. Int. Ed. 2011, 50,
1190.
(6) Lange, H.; Carter, C. F.; Hopkin, M. D.; Burke, A.; Goode, J. G.;
Baxendale, I. R.; Ley, S. V. Chem. Sci. 2011, 2, 765.
(7) (a) Kralj, J. G.; Sahoo, H. R.; Jensen, K. F. Lab Chip 2007, 7, 256.
(b) Atallah, R. H.; Ruzicka, J.; Christian, G. D. Anal. Chem. 1987,
59, 2909. (c) Castell, O. K.; Allender, C. J.; Barrow, D. A. Lab Chip
2009, 9, 388. (d) Kolehmainen, E.; Turunen, I. Chem. Eng. Process.
2007, 46, 834. (e) Hornung, C. H.; Mackley, M. R.; Baxendale, I.
R.; Ley, S. V. Org. Process Res. Dev. 2007, 11, 399.
(8) For a recent review on the use of cameras in organic synthesis,
see: Ley, S. V.; Ingham, R. J.; O’Brien, M.; Browne, D. L. Beilstein J.
Org. Chem. 2013, 9, 1051.
(9) O’Brien, M.; Koos, P.; Browne, D. L.; Ley, S. V. Org. Biomol. Chem.
2012, 10, 7031.
(10) Ingham, R. J.; Battilocchio, C.; Fitzpatrick, D. E.; Sliwinski, E.;
Hawkins, J. M.; Ley, S. V. Angew. Chem. Int. Ed. 2015, 54, 144.
(11) Sprecher, H.; Payán, M.; Weber, M.; Yilmaz, G.; Wille, G. J. Flow
Chem. 2012, 2, 20.
Acknowledgment
We would like to thank the Keele University Acorn Fund for funding.
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0035-1560975.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
References and Notes
(12) Jirkovsky, I. Can. J. Chem. 1974, 52, 55.
(13) Gerasyuto, A. I.; Hsung, R. P.; Sydorenko, N.; Slafer, B. J. Org.
Chem. 2005, 70, 4248.
(14) Bradski, G. Dr. Dobbs J. 2000, 25, 120.
(15) For an alternative approach that relates the size of a coloured
disk to the vertical distance from a camera, see: Wang, T. H.; Lu,
M. C.; Hsu, C. C.; Chen, C. C.; Tan, J. D. Measurement 2009, 42,
604.
(16) Representative Procedure for the Formation of the Enami-
nones (3i)
(1) (a) Pastre, J. C.; Browne, D. L.; Ley, S. V. Chem. Soc. Rev. 2013, 42,
8849. (b) Wegner, J.; Ceylan, S.; Kirschning, A. Adv. Synth. Catal.
2012, 354, 17. (c) Brzozowski, M.; O’Brien, M.; Ley, S. V.;
Polyzos, A. Acc. Chem. Res. 2015, 48, 349. (d) Gutmann, B.;
Cantillo, D.; Kappe, C. O. Angew. Chem. Int. Ed. 2015, 54, 6688.
(e) Pathak, S.; Kundu, A.; Pramanik, A. RSC Adv. 2014, 4, 10180.
(f) Ley, S. V. Chem. Rec. 2012, 12, 378. (g) Webb, D.; Jamison, T.
F. Chem. Sci. 2010, 1, 675. (h) McQuade, D. T.; Seeberger, P. H.
J. Org. Chem. 2013, 78, 6384. (i) Wiles, C.; Watts, P. Chem.
Commun. 2011, 47, 6512. (j) Newman, S. G.; Jensen, K. F. Green
Chem. 2013, 15, 1456. (k) Yoshida, J. I. Chem. Rec. 2010, 10, 332.
(l) Knowles, J. P.; Elliott, L. D.; Booker-Milburn, K. I. Beilstein J.
Org. Chem. 2012, 8, 2025. (m) Ley, S. V.; Fitzpatrick, D. E.;
Ingham, R. J.; Myers, R. M. Angew. Chem. Int. Ed. 2015, 54, 3449.
(2) (a) Brandt, J. C.; Wirth, T. Beilstein J. Org. Chem. 2009, 5, 30.
(b) Muller, S. T. R.; Wirth, T. ChemSusChem 2015, 8, 245.
(c) O’Brien, M.; Baxendale, I. R.; Ley, S. V. Org. Lett. 2010, 12,
1596. (d) Mastronardi, F.; Gutmann, B.; Kappe, C. O. Org. Lett.
2013, 15, 5590. (e) Maurya, R. A.; Park, C. P.; Lee, J. H.; Kim, D. P.
Angew. Chem. Int. Ed. 2011, 50, 5952. (f) Malet-Sanz, L.;
Madrzak, J.; Ley, S. V.; Baxendale, I. R. Org. Biomol. Chem. 2010,
8, 5324. (g) O’Brien, M.; Taylor, N.; Polyzos, A.; Baxendale, I. R.;
Ley, S. V. Chem. Sci. 2011, 2, 1250. (h) Poh, J.-S.; Tran, D. N.;
Battilocchio, C.; Hawkins, J. M.; Ley, S. V. Angew. Chem. Int. Ed.
2015, 54, 7920. (i) Razzaq, T.; Kappe, C. O. Chem. Asian. J. 2010,
5, 1274.
(3) (a) Adolf, W. G. Mass Transfer Effects on Liquid-Phase Chemical
Reaction Rates, In Homogeneous Catalysis; Luberoff, B. A., Ed.;
American Chemical Society: Washington DC, 1974, ; Advances
in Chemistry Series; Vol. 70: Chap. 3,: 35. (b) Hobbs, C. C.;
Onore, M. J.; Van’t Hof, H. A.; Mesich, F. G.; Drew, E. H. Ind. Eng.
Chem. Prod. Res. Dev. 1972, 11, 220. (c) Markos, J.; Pisu, M.;
Morbidelli, M. Comput. Chem. Eng. 1998, 22, 627.
(4) (a) Jensen, K. F. Chem. Eng. Sci. 2001, 56, 293. (b) Renken, A.;
Kiwi-Minsker, L. Adv. Catal. 2010, 53, 47. (c) Mansur, E. A.; Ye,
M.; Wang, Y.; Dai, Y. Chin. J. Chem. Eng. 2008, 16, 503.
(5) (a) Ley, S. V.; Baxendale, I. R.; Bream, R. N.; Jackson, P. S.; Leach,
A. G.; Longbottom, D. A.; Nesi, M.; Scott, J. S.; Storer, R. I.; Taylor,
S. J. J. Chem. Soc., Perkin Trans. 1 2000, 3815. (b) Alza, E.;
Rodriguez-Escrich, C.; Sayalero, S.; Bastero, A.; Pericas, M. A.
Chem. Eur. J. 2009, 15, 10167. (c) O'Brien, M.; Denton, R.; Ley, S.
Cyclohexanedione (1.55 g, 13.8 mmol, 1 equiv) and 4-chloro-
benzylamine (1.8 mL, 14.8 mmol, 1.1 equiv) were added to a
flask under nitrogen. To this was added toluene (50 mL) and
EtOH (2.5 mL), and the mixture was stirred at reflux for 3 h.
Upon completion the solvent was removed under reduced pres-
sure, forming a yellow-brown solid, which was recrystallised
from toluene. The product was made up of yellow-brown crys-
tals; yield 71% (2.31 g); mp 166.8–168.4 °C (lit.: 170–172 °C).
1H NMR (300 MHz, CDCl3): δ = 7.32–7.26 (2 H, m), 7.18 (J = 8.54
Hz, 2 H, m), 5.34 (1 H, br s), 5.07 (1 H, s), 4.18 (J = 5.36 Hz, 2 H,
d), 2.38 (J = 6.15 Hz, 2 H, t), 2.27 (J = 6.53 Hz, 2 H, t), 1.99–1.91
(2 H, m, H-2). 13C NMR (100 MHz, CDCl3): δ = 197.4, 165.2,
135.3, 133.6, 128.9, 97.5, 46.3, 36.4, 29.5, 21.9. IR: ν = 3247
(NH), 3049, 1538 (C=O), 683 cm–1
.
Representative Procedure for the Continuous Flow Bromina-
tion (4i)
Using the apparatus shown in Figure 3, the system was primed
with CH2Cl2 and the aqueous extraction solvent for several
minutes until there were no air gaps in the flow path. 3-[(4-
chlorobenzyl)amino]cyclohex-2-enone (3i, 0.100 M in CH2Cl2)
was loaded in to the 3 mL injection loop. NBS (0.106 M in
CH2Cl2) was loaded into the 4.1 mL injection loop. The NBS loop
was injected into the system 20 s prior to the substrate loop.
Organic solution exiting the system was collected for 20 min.
The solvent was removed under reduced pressure to afford a
yellow-brown solid; yield 93% (87.5 mg): mp 121.5–122.3 °C.
1H NMR (400 MHz, CDCl3): δ = 7.32 (J = 7.43 Hz, 2 H, d), 7.18
(J = 7.65 Hz, 2 H, d), 6.12 (1 H, br s), 4.49 (J = 5.17 Hz, 2 H, d),
2.51 (J = 6.34 Hz, 2 H, t), 2.45 (J = 6.03 Hz, 2 H, t), 1.98–1.84 (2 H,
m). 13C NMR (100 MHz, CDCl3): δ = 187.8, 161.0, 136.0, 133.7,
129.2, 128.0, 96.6, 46.5, 36.6, 26.7, 20.7. IR: ν = 3260 (NH), 2988,
2955, 2939, 2901, 2884, 1584 (C=O), 800, 715 cm–1
.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, 164–168