Synthesis of 4,4′-Disubstituted and Spiro-tetrahydroquinolines via Photochemical Cyclization of Acrylanilides and the First Synthesis of (±)- trans -Vabicaserin

Article abstract of DOI:10.1055/s-0035-1562621

The synthesis of vabicaserin analogues bearing a quaternary center or spiro substitution at the 4-position has been studied via a [6π]-acrylanilide cyclization employing flow photochemistry in a mesoscale and microfluidic flow photoreactor. The method is also used to synthesize 4,4′-disubstituted tetrahydroquinolines and, furthermore, enables the first synthesis of (±)-trans-vabicaserin.

Full text of DOI:10.1055/s-0035-1562621

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, A–F
letter
A
H. F. Koolman et al.
Letter
Synlett
Synthesis of 4,4′-Disubstituted and Spiro-tetrahydroquinolines
via Photochemical Cyclization of Acrylanilides and the First Syn-
thesis of (±)-trans-Vabicaserin
Hannes F. Koolman*a
Hg lamp
Wilfried M. Brajeb
Andreas Hauptb
Boc
Boc
NH
N
O
NH
N
O
a AbbVie, Discovery Chemistry and Technology,
1 North Waukegan Road, North Chicago, IL 60064,
USA
R4
R4
R4
N
N
N
meso-scale or microfluidic
photoflow reactor
N
R1
R2
4
4
R1
R2
R1
R2
hannes.koolman@abbvie.com
acetone/Pyrex®
b AbbVie Deutschland GmbH & Co. KG, Neuroscience
Discovery Research, Knollstrasse, 67061 Lud-
wigshafen, Germany
R3
R3
R3
(
)-trans-Vabicaserin
possible quaternary
carbon center
up to multi-gram scale
R1, R2, R3 = D, CF3, alkyl,
cycloalkyl
R4 = F, Cl
13 examples, 40–92%
Received: 17.06.2016
Accepted after revision: 28.07.2016
Published online: 16.08.2016
O
F
O
N
H2N
H
N
NH
N
DOI: 10.1055/s-0035-1562621; Art ID: st-2016-r0394-l
S
Cl
N
HN
N
O
O
Abstract The synthesis of vabicaserin analogues bearing a quaternary
center or spiro substitution at the 4-position has been studied via a
[6π]-acrylanilide cyclization employing flow photochemistry in a meso-
scale and microfluidic flow photoreactor. The method is also used to
synthesize 4,4′-disubstituted tetrahydroquinolines and, furthermore,
enables the first synthesis of (±)-trans-vabicaserin.
N
N
H
Cl
MeO
2, GSK2163632A
1
O
HO
Key words flow photochemistry, photocyclization, electrocyclization,
OEt
electrocyclic reactions, tetrahydroquinoline, vabicaserin
O
NH
N
O
N
N
1,2,3,4-Tetrahydroquinolines are important substruc-
tures in many biologically active molecules, and a vast
number of synthetic methodologies have been established
over the past decades.1 The general motive can be found in
marketed drugs like the anthelmintic oxamniquine2 or the
antiarrhythmic nicainoprol,3 as well as a large number of
natural products.1 Despite the obvious relevance of the gen-
eral substructure, a comparably limited number of reports
in the literature show a precedence for 4,4′-dialkyl and
even a smaller precedence for 4-spirocyclic tetrahydroquin-
olines.4 Recent reports about pharmacologically useful
tetrahydroquinolines bearing such a substitution pattern
include, for example, retinoic acid receptor related orphan
receptor C (RORc) inverse agonist 1,5 G protein coupled
receptor kinase inhibitor 2 (GSK2163632A),6 PPARα/γ ago-
nists 37 and 11β-hydroxysteroid-dehydrogenase type 1 in-
hibitor 48 (Figure 1).
N
N
OH
3
4
Figure 1 Examples of pharmaceutically relevant 1,2,3,4-tetrahydro-
quinolines bearing disubstitution in the 4-position
quinoline substructure, and besides its potential pharmaco-
logical activity, its synthesis has recently gained more at-
tention.10
We envisioned that a 4,4′-dialkyl or spirocylic-substi-
tuted tetrahydroquinoline as part of a diazepino-annelated
heterocycle 6 or by itself (7) would not only be a challeng-
ing target with its quaternary center adjacent to the aro-
matic ring but could also offer an unexplored position for
substituents to probe different pharmacological activity
(Figure 2).
The 1,2,3,4-tetrahydroquinoline core is also often incor-
porated into a polycyclic system. For example, vabicaserin
(5), a clinically investigated 5-HT2C agonist for the treat-
ment of schizophrenia,9 features a diazepino-hexahydro-
For the construction of the quaternary center in the
4-position of tetrahydroquinolines, several synthetic routes
are commonly employed.11 They include predominantly
Lewis acid12 but also Brønsted acid4a mediated cyclizations
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F

B
H. F. Koolman et al.
Letter
Synlett
Boc-deprotected diazepine was recovered quantitatively. By
subjecting the acrylanilide 9 to a quartz-glass-filtered UV
light from a mercury middle-pressure lamp in a mesoscale
tube-based flow reactor (see Supporting Information) in
acetone, the desired dihydroquinolinone 10 was gratifying-
ly isolable, albeit in low yields and with small amounts of
recovered starting material (Scheme 1 and Table 1, entry 1).
Reduction with borane and subsequent Boc deprotection
with trifluoroacetic acid yielded the desired diazepino-
annelated heterocycle 11 in good yields.21
NH
NH
R
N
N
NH
4
5
6
7
Vabicaserin
Figure 2 Vabicaserin (5) and potential substituted tetrahydroquinoline
targets
of acrylanilides followed by amide reduction. A direct syn-
thesis of 4,4′-dimethyl tetrahydroquinoline has also been
accomplished by intramolecular Friedel–Crafts alkylation.13
Another approach that is employed to yield spirocyclic sys-
tems is an intramolecular Heck reaction of homo-acrylani-
lides followed by a reduction of the olefin and amide.14 An
intramolecular radical addition to quinolinones has been
used to generate a spirocyclic dihydroquinolinone.15 Anoth-
er well-established approach that yields suitable dihydro-
quinolinone precursors is the ruthenium-catalyzed 1,4-ad-
dition of 2-aminoboronic acid to α,β-unsaturated esters,16
which was recently reported in water at room tempera-
ture.17 The Povarov reaction, an inverse electron-demand
aza Diels–Alder reaction of N-aryl imines with electron-
rich alkenes, is another elegant route to furnish the quater-
nary center.18 A more exotic reaction to obtain a quaternary
center at the 4-position was also achieved by deprotonation
of ruthenium complexes of 1,4-dimethyl-1,2,3,4-tetra-
hydroquinolines and subsequent reaction with electro-
philes.19 Furthermore, the well-defined photochemical
[6π]-acrylanilide cyclization can be used with suitably
substituted precursors to generate the 3,4-dihydroquinolin-
2-one20 after thermally allowed suprafacial hydrogen
1,5-shift or intermolecular proton transfer, respectively,
depending on the reaction conditions.
Initially, we were interested in synthesizing 4,4′-disub-
stituted tetrahydroquinolines as part of a diazepino-annel-
ated heterocycle 6 (Figure 2). In order to establish a robust
route, we based our first attempts on a Lewis acid mediated
cyclization with a dimethyl substitution as the initial model
substrate. Commercially available N-Boc-protected benzo-
diazepine 8 was therefore acylated with dimethylacrylic
acid chloride and the resulting anilide 9 treated with AlCl3
in dichloromethane at 0 °C. Unfortunately, no cyclized dihy-
droquinolinone 10 was observed even with prolonged reac-
tion times or higher reaction temperatures, and only the
Table 1 Optimization of Reaction Conditions for the Photochemical
[6π]-Scrylanilide Cyclization in a Photochemical Flow Reactora
Boc
Boc
Boc
N
O
Boc
N
O
N
N
conditions
N
N
N
H
N
H
O
9
10
12
8
Entry
Solvent/filter
Conv.
(%)b
Yield of
Recovered
10/12/8 (%)c
yield of 9 (%)c
1
acetone/quartz
MeCN/quartz
toluene/quartz
Et2O/quartz
39
52
62
48
40
95
49
18/ND/NDd
17/16/10
30/trace/4
13/21/5
32/–/–
18
9
2
3
24
12
60
4
4
5
acetone/Pyrex®
acetone/Pyrex®
acetone/Pyrex®
6e
7f
61/–/–
30/–/–
51
a General reaction conditions: 9 (0.25 mmol, 0.05 M), irradiated once at 0.2
mL/min (5 mL tube reactor, 25 min residence time) with a 150 W Hg lamp,
if not otherwise noted. See Supporting Information.
b Based on relative HPLC integrals at 254 nm.
c Isolated yield.
d ND = not determined.
e Reaction mixture was re-irradiated for 15 h in continuous flow.
f Reaction mixture was irradiated for 15 h in an immersion well batch reactor
employing the same lamp (3.3 mmol scale, 0.02 M).
We then turned to the optimization of the photochemi-
cal step to enable the generation of larger quantities and to
eventually explore a broader substrate scope. Summarized
in Table 1, several triplet- and nontriplet-sensitizing sol-
vents and borosilicate (Pyrex®) as an optional optical filter
were screened in a flow format. Ultimately, those results
were compared to a reaction in batch (Table 1).
O
Boc
Boc
Boc
Cl
1.) BH3⋅THF
r.t.
N
N
NH
N
Et3N
hν
CH2Cl2
r.t.
96%
acetone/quartz
18%
2.) TFA, CH2Cl2
r.t.
72% (2 steps)
N
N
H
N
N
O
O
8
9
10
11
Scheme 1 Initial synthetic route to diazepino-tetrahydroquinolines with quaternary center adjacent to the aromatic ring
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F

C
H. F. Koolman et al.
Letter
Synlett
Compared to the reaction in acetone (Table 1, entry 1),
when employing acetonitrile, a higher conversion of start-
ing material was possible, albeit significant quantities of
not only the photo-Fries rearrangement product 12 but also
the deacetylated product 8 were isolated (Table 1, entry 2).
The triplet-sensitizing solvent toluene resulted in the for-
mation of a larger amount of unidentifiable low-molecular
weight side products and coloring of the employed FEP-
tubing to a larger extent compared to acetonitrile (Table 1,
entry 3). When utilizing diethyl ether, a comparable result
to acetonitrile was obtained, confirming that more photo-
Fries rearrangement product 12 is formed in nontriplet-
sensitizing solvents (Table 1, entry 4). Switching to the sen-
sitizing solvent acetone in combination with filtering off
lower wavelengths using a Pyrex® filter, a clean reaction
profile with 32% isolable yield and good starting material
recovery was obtained after irradiation for just 25 minutes
(Table 1, entry 5). Re-irradiation of the reaction mixture in a
continuous fashion for 15 hours resulted in almost full con-
version of the starting material yielding 61% of the desired
product 10 (Table 1, entry 6).22 When comparing this result
to a batch reaction employing a regular immersion well re-
actor (Table 1, entry 7), the mesoscale flow setup shows a
higher space-time yield confirming the advantage of a flow
reactor.23
With the optimized reaction conditions in hand, a series
of acrylanilides were cyclized to their respective dihydro-
quinolinones and also converted into the ultimately desired
diazepino-tetrahydroquinolines in a similar fashion as de-
scribed in Scheme 1.
Table 2 Flow Photocyclization of Diazepino-acrylanilides to the Corresponding Dihydroquinolinonesa
R1
O
Boc
R2
Cl
Boc
1.) BH3⋅THF
r.t.–60 °C
NH
N
O
N
O
7
R4
8
R3
hν
acetone/Pyrex®
R4
8
R4
Et3N
CH2Cl2
r.t.
2.) TFA, CH2Cl2
r.t.
N
N
N
R1
R1
R2
R1
R2
R2
R3
R3
14a–m
R3
13a–m
6a–m
Entry
R1
Me
R2
R3
R4
Time (h)
Conv. (%)b
14
Yield (%)c
45
6
Yield (%)d
1
2
Me
Me
Me
Me
CF3
CF3
H
7-Cl
16.5
17.25
17
95
65
95
95
91
71
43
98f
99i
98i
0
14a
14b
14c
14d
14e
14f
6a
6b
6c
6d
6e
6f
27
Me
H
H
H
H
H
H
8-Cl
7-F
8-F
H
33 (52 brsm)
35
3
Me
50
64
4e
5
Me
12
47
63
Me
20
33 (36 brsm)
27 (49 brsm)
20 (67 brsm)
49/17
43
6e
7
Me
7-F
H
32
46
Me
CH3CH2
-(CH2)3-
-(CH2)3-
-(CH2)3-
Ph
21
14g
14h
14i
6g
6h
6i
78
8
Me
H
5
67/53g
70/75j
12/4k
–
9
H
H
11 minh
5
82
10
11
12
13
D
H
14j
69/23
6j
Me
H
H
H
H
6
14k
14l
–
6k
6l
-(CH2)3-
-(CH2)3-
H
11 minl
11 minl
90
85
40
76
F
14m
48 (69 brsm)
6m
95
a General reaction conditions: 13 (1.0 mmol, 0.05 M) was continuously irradiated for the time specified at 0.2 mL/min (5 mL tube reactor, 25 min residence time)
with a 150 W Hg lamp. See Supporting Information.
b Based on relative HPLC integrals at 254 nm.
c Isolated yield; yields of separated cis and trans isomer if applicable after photocylization; 14i not separable.
d Unoptimized isolated yield over two steps; yields of cis/trans if isomers from previous step were reacted separately.
e 0.7 mmol scale.
f A cis/trans ratio of 2.2:1 was determined by relative HPLC integrals at 254 nm, later confirmed by NMR analysis.
g Unoptimized isolated yields over two steps of cis and trans isomer, taken forward from 14h separately.
h Reaction performed with a 1 mL microfluidic reactor with a 300 W Hg lamp at 2.19 mmol scale (0.05 M) and a single irradiation residence time of 11 min (total
runtime of 6.7 h). See Supporting Information and ref. 24 for details.
i A cis/trans ratio of 3:1 was determined by NMR analysis of the crude reaction and taken forward as mixture.
j Unoptimized isolated yields over two steps of cis and trans isomer; isomers separated after amide reduction.
k Combined yields of all respective enantiomers separated by chiral SFC (cis/trans, taken forward from 14j separately).
l Reaction performed with a 1 mL microfluidic reactor with a 300 W Hg lamp at 9.4 mmol scale (0.05 M) and a single irradiation residence time of 11 min (total
runtime of 35 h). See Supporting Information and ref. 24 for details.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F

D
H. F. Koolman et al.
Letter
Synlett
Employing the optimized combination of using acetone
as solvent and Pyrex® as a filter, several diazepino-acrylani-
lides were cyclized effectively to the corresponding dihy-
droquinolinones (Table 2). Substrates with electron-with-
drawing substituents on the aromatic ring in either posi-
tion showed good conversions based on consumed starting
material, but isolated yields were considerably lower com-
pared to the unsubstituted analogue 9 even without the
formation of distinct side products (Table 2, entries 1–4).
Acrylanilides 13e–f bearing a trifluoromethyl substituent
on the olefin also showed lower yields, which were even
further impacted by an additional fluoro substituent on the
aromatic ring (Table 2, entries 5 and 6). The ethyl-substitut-
ed derivative 13g showed sluggish conversion compared to
the dimethyl substrate and might have benefited from an
increased runtime based on the amount of recovered start-
ing material (Table 2, entry 7). The cyclic anilides, however,
showed remarkably higher reactivity in this electrocyclic
reaction. Full conversion could be obtained after four and
five hours, respectively (Table 2, entries 8 and 10), or with a
residence time of just 11 minutes in a microfluidic flow
photoreactor (LOPHTOR) that has previously been report-
ed.24 Surprisingly, even if conducted in an aprotic solvent
without the addition of a Brønsted acid the cis isomer was
preferentially formed as identified by NMR analysis of the
crude mixture or isolation and subsequent analysis of the
products.20c Nevertheless, further reduction and Boc depro-
tection to the final derivatives 6h–j yielded (±)-trans-vabi-
caserin (6i, Table 2, entry 9) for the first time. Further sub-
stituted vabicaserin analogues with unprecedented substit-
uents adjacent to the ring could be obtained as well (Table
2, entries 8 and 10, 6h,j).21 When attempting a methyl-cin-
namoyl-substituted benzodiazepine, no product could be
obtained after UV-light irradiation (Table 2, entry 11, see
Supporting Information). To attempt the synthesis of a spi-
rocyclic substituent on the acrylic moiety via [6π]-acrylani-
lide cyclization, we challenged the cyclobutyl precursors
13l,m (Table 2, entries 12 and 13). These reactions were
also performed in the above-mentioned microfluidic flow
photoreactor. Since this system employs a higher light in-
tensity to flow-cell ratio, a single pass with a residence time
of just 11 minutes was also sufficient in this case to yield a
conversion comparable to that of the mesoflow reactor and
readily allowed for the generation of multigram amounts of
cyclized products 14l,m. Overall, to the best of our knowl-
edge, this is the first report of an acryl-anilide cyclization
being utilized to yield a spirocyclic dihydroquinoline moi-
ety.
Encouraged by the fact that a more general synthesis of
the spiro-substituted tetrahydroquinoline (i.e., 7, Figure 1)
would enable a more flexible and diversity-oriented con-
struction of the diazepine ring,25 we tried to transfer the
findings from the aforementioned photocyclization to more
simple aniline precursors (Table 3).26
Table 3 Photocylization of Simple Cyclobutyl-acrylanilide Precursorsa
O
R2
R2
R2
Cl
table
conditions
Et3N
R1
NH
CH2Cl2
r.t.
R1
NH2
R1
NH
O
O
16a R1, R2 = H, 97%
16b R1 = F; R2 = H, 77%
16c R1, R2 = F, 71%
17a R1, R2 = H
17b R1 = F; R2 = H
17c R1, R2 = F
15a R = H
15b R = F
Entry
R1
R2
mmol
Solvent/filter
Conv. (%)b
17
17a
Yield (%)c
1
2
3
4
5
H
H
H
H
H
H
0.05
0.05
0.08
0.92
1.95
acetone/Pyrex®
acetone/quartz
toluene/quartz
toluene/Vycor®
toluene/Vycor®
5
20
85
60
50
NDd
ND
50
H
H
H
F
17a
17a
17a
17b
42
16
(46 brsm)
6
7
F
F
H
F
1.17
1.83
acetone/Vycor®
acetone/Vycor®
50
50
17b
17c
23
(43 brsm)
17
a General reaction conditions: 16 (0.1 M) was irradiated in a 1 mL microfluidic reactor with a 300 W Hg lamp at the indicated scale by a single irradiation resi-
dence time of 11 min. See ref. 24 for details.
b Based on relative HPLC integrals at 254 nm. The conversion is >90% at an early time point but then declines over time indicating a change in the reactor
conditions.
c Isolated yield; combined yield of regioisomers where applicable.
d ND = not determined.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F

E
H. F. Koolman et al.
Letter
Synlett
When applying the previously optimized reaction con-
ditions, it became apparent that the system would require
some direct irradiation as the use of a Pyrex® filter in com-
bination with the absorbing solvent acetone did not yield
any significant conversion (Table 3, entry 1).27 Limiting the
optical filter to very short wavelengths only (Table 3, entry
2, quartz) increased the conversion to 20%. Using toluene as
solvent resulted in higher conversion after a short irradia-
tion (Table 3, entry 3). However, the formation of decompo-
sition products and subsequent staining of the reactor
membrane28 led to an isolated yield of only 50%. To some
extent this could be minimized by applying a Vycor® filter
(Table 3, entry 4) where an isolated yield of 42% of the de-
sired spirocycle 17a could be obtained on a 0.9 mmol scale.
Using the latter conditions for the cyclization of the fluoro-
substituted derivatives 16b and 16c on even larger scale ei-
ther employing toluene or acetone as solvent enabled the
synthesis of these valuable building blocks, however, isolat-
ed yields were modest (16–23%, Table 3, entries 5–7).
In summary, we have demonstrated that 4,4′-disubsti-
tuted tetrahydroquinolines as part of diazepino-annelated
heterocycles can be synthesized efficiently by applying a
photochemical [6π]-acrylanilide cyclization in a meso- and
microfluidic photoreactor, even on multigram scale.
The highlighted synthetic route was suitable to synthesize
(±)-trans-vabicaserin for the first time. It tolerates substitu-
ents on the aromatic, as well as the acrylic moiety. In this
context, the 4-spirosubstituted dihydroquinoline moiety is
synthesized for the first time via a photocyclization. At-
tempts to transfer the method to the cyclization of a sim-
pler spiro-substituted acrylanilide precursor were essen-
tially successful, but performing those reactions on a larger
scale would require further optimization.
References and Notes
(1) (a) Sridharan, V.; Suryavanshi, P. A.; Menendez, J. C. Chem. Rev.
2011, 111, 7157. (b) Katritzky, A. R.; Rachwal, S.; Rachwal, B.
Tetrahedron 1996, 52, 15031. (c) Nammalwar, B.; Bunce, R. A.
Molecules 2014, 19, 204.
(2) Ferrari, M. L. A.; Coelho, P. M. Z.; Antunes, C. M. F. Tavares C. A.
P.; da Cunha, A. S. Bull. W.H.O. 2003, 81, 190.
(3) Imanishi, S.; Kimura, T.; Arita, M. Cardiovasc. Drug Rev. 1991, 9,
223.
(4) For recent relevant examples, see: (a) Chu, D. T. W.; Chen, J.;
Zhang, W.; Li, X.; Song, J.; Wang, B.; Cong, Q.; James, D. R. US
20060128790, 2006; Chem. Abstr. 2006, 145, 62795
(b) Frackenpohl, J.; Bojack, G.; Helmke, H.; Lehr, S.; Mueller, T.;
Willms, L.; Dietrich, H.; Schmutzler, D.; Baltz, R.; Bickers, U. WO
2015155154, 2015; Chem. Abstr. 2015, 163, 628222 (c) Feng, L.;
Huang, M.; Liu, Y.; Wu, G.; Yan, S.; Yun, H.; Zhou, M. US
20120190677, 2012; Chem. Abstr. 2012, 157, 294945
(5) Fauber, B. P.; Gobbi, A.; Savy, P.; Burton, B.; Deng, Y.; Everett, C.;
La, H.; Johnson, A. R.; Lockey, P.; Norman, M.; Wong, H. Bioorg.
Med. Chem. Lett. 2015, 25, 4109.
(6) Homan, K. T.; Larimore, K. M.; Elkins, J. M.; Szklarz, M.; Knapp,
S.; Tesmer, J. J. G. ACS Chem. Biol. 2015, 10, 310.
(7) Parmenon, C.; Guillard, J.; Caignard, D.-H.; Hennuyer, N.; Staels,
B.; Audinot-Bouchez, V.; Boutin, J.-A.; Ktorza, A.; Dacquet, C.;
Viaud-Massuard, M.-C. Bioorg. Med. Chem. Lett. 2008, 18, 1617.
(8) Venier, O.; Pascal, C.; Braun, A.; Namane, C.; Mougenot, P.;
Crespin, O.; Pacquet, F.; Mougenot, C.; Monseau, C.; Onofri, B.;
Dadji-Faïhun, R.; Leger, C.; Ben-Hassine, M.; Van-Pham, T.;
Ragot, J. L.; Philippo, C.; Güssregen, S.; Engel, C.; Farjot, G.; Noah,
L.; Maniani, K.; Nicolaï, E. Bioorg. Med. Chem. Lett. 2011, 21,
2244.
(9) Dunlop, J.; Watts, S. W.; Barrett, J. E.; Coupet, J.; Harrison, B.;
Mazandarani, H.; Nawoschik, S.; Pangalos, M. N.; Ramamoorthy,
S.; Schechter, L.; Smith, D.; Stack, G.; Zhang, J.; Zhang, G.;
Rosenzweig-Lipson, S. J. Pharm. Exp. Tech. 2011, 337, 673.
(10) (a) Megati, S.; Bhansali, S.; Dehnhardt, C.; Deshmukh, S.; Fung,
P.; Macewan, M.; Tinder, R. J. WO 2009039362, 2009; Chem.
Abstr. 2009, 150, 329853 (b) Dragan, V.; McWilliams, J. C.;
Miller, R.; Sutherland, K.; Dillon, J. L.; O'Brien, M. K. Org. Lett.
2013, 15, 2942. (c) Dziechciejewski, W. J.; Weber, R.; Sowada,
O.; Boysen, M. M. K. Org. Lett. 2015, 17, 4132.
Acknowledgment
The authors thank Tanja Lindner for supporting chemical synthesis.
Rickie S. Yarbrough, David N. Whittern, and Claudia Krack are ac-
knowledged for analytical support. Frank L. Wagenaar and Anita K.
McGreal are acknowledged for supporting purifications and Dave
Blanchard for engineering troubleshooting on the LOPHTOR system.
James Stambuli and Amanda W. Dombrowski are acknowledged for
useful comments regarding this manuscript. All are employees of
AbbVie. This study was sponsored by AbbVie. AbbVie contributed to
the study design, research, and interpretation of data, writing, re-
viewing, and approving the manuscript.
(11) It is noteworthy, that the commonly employed approach, a Pd-
catalyzed amidation of aryl halides, should allow the synthesis
of dihydroquinolinones bearing a quarternary center in the 4-
position if a suitable precursor is employed. Until now, this has
not been shown. For the underlying chemistry, see: (a) Zhang,
L.; Qureshi, Z.; Sonaglia, L.; Lautens, M. Angew. Chem. Int. Ed.
2014, 53, 13850. For the synthesis of such precursors, the pur-
posive Rh-based conjugate addition of ortho-halo-arylboronic
acids onto β-substituted acrylamides, however, fails even if only
a tertiary center is attempted, as stated in: (b) Zhang, L.;
Sonaglia, L.; Stacey, J.; Lautens, M. Org. Lett. 2013, 15, 2128; and
references cited therein.
(12) Parmenon, C.; Guillard, J.; Caignard, D.-H.; Hennuyer, N.; Staels,
B.; Audinot-Bouchez, V.; Boutin, J.-A.; Dacquet, C.; Ktorza, A.;
Viaud-Massuard, M.-C. Bioorg. Med. Chem. Lett. 2008, 18, 1617.
(13) Bunce, R. A.; Cain, N. R.; Cooper, J. G. Org. Prep. Proced. Int. 2013,
45, 28.
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0035-1562621.
S
u
p
p
ortioInfgrmoaitn
S
u
p
p
ortiInfogrmoaitn
(14) (a) Pereira, J.; Barlier, M.; Guillou, C. Org. Lett. 2007, 9, 3101.
(b) Koolman, H. F.; Braje, W. M.; Mack, H.; Haupt, A.; Relo, A. L.;
Drescher, K.; Bakker, M. H. M.; Lakics, V.; Hoft, C.; Xu, R.; Zhao,
X. WO 2014041131, 2014; Chem. Abstr. 2014, 160, 457941
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F

F
H. F. Koolman et al.
Letter
Synlett
(15) Dressel, M.; Bach, T. Org. Lett. 2006, 8, 3145.
in heptane) to yield 202 mg (61%) of 10 as a beige solid. ESI-MS:
m/z (%) = 275 (80) [M – C4H9 + 2H]+, 331 (20) [M + H]+. 1H NMR
(500 MHz, CDCl3): δ = 7.23–7.15 (m, 1 H), 7.13–6.99 (m, 2 H),
4.68 (s, 1 H), 4.54 (s, 1 H), 4.18–4.09 (m, 2 H), 3.82–3.66 (m, 2
H), 2.43 (d, J = 8.8 Hz, 2 H), 1.41 (s, 6 H), 1.27 (s, 9 H).
(16) Horn, J.; Li, H. Y.; Marsden, S. P.; Nelson, A.; Shearer, R. J.;
Campbell, A. J.; House, D.; Weingarten, G. G. Tetrahedron 2009,
65, 9002.
(17) Linsenmeier, A. M.; Braje, W. M. Tetrahedron 2015, 71, 6913.
(18) (a) Povarov, L. S. Russ. Chem. Rev. 1967, 36, 656. (b) Xie, M.; Liu,
X.; Zhu, Y.; Zhao, X.; Xia, Y.; Lin, L.; Feng, X. Chem. Eur. J. 2011,
17, 13800.
(19) Pigge, F. C.; Fang, S.; Rath, N. P. Tetrahedron Lett. 1999, 40, 2251.
(20) (a) For a general review, see: Laarhoven, W. H. Org. Photochem.
1989, 10, 163. (b) Barbera, C.; Garcia, H.; Miranda, M. A.; Primo,
J. An. Quim. 1995, 91, 95. (c) Bach, T.; Grosch, B.; Strassner, T.;
Herdtweck, E. J. Org. Chem. 2003, 68, 1107.
(21) The pharmacologic activities and the relevance of the final
compounds presented herein will be published elsewhere.
(22) Typical Procedure for the Preparation of tert-Butyl 8,8-
Dimethyl-6-oxo-3,4,7,8-tetrahydro-1H-[1,4]diazepino[6,7,1-
ij]quinoline-2(6H)-carboxylate (10)
(23) For a review on effective space time yields (STYeff) to compare
efficiencies of reactors with different geometries see:
(a) Shvydkiv, O.; Yavorskyy, A.; Tan, S. B.; Nolan, K.; Hoffmann,
N.; Youssef, A.; Oelgemöller, M. Photochem. Photobiol. Sci. 2011,
10, 1399. (b) The flow reaction (Table 1, entry 6) has a STYeff
=
0.21 mmol L–1 min–1 (based on a 5 mL volume), whereas the
batch reaction (Table 1, entry 7) has a STYeff = 7.1·10–3 mmol L–1
min–1 (based on a 250 mL volume) and is thus considered less
efficient.
(24) Vasudevan, A.; Villamil, C.; Trumbull, J.; Olson, J.; Sutherland,
D.; Pan, J.; Djuric, S. Tetrahedron Lett. 2010, 51, 4007.
(25) For example via a known Pictet–Spengler-type cyclization, see:
Neelamegam, R.; Hellenbrand, T.; Schroeder, F. A.; Wang, C.;
Hooker, J. M. J. Med. Chem. 2014, 57, 1488.
(26) The [6π]-acrylanilide cyclization of dimethylphenyl-acrylamide
is known and described in ref. 20b but the yield reported
therein is elusive.
A degassed solution of 0.33 g (1.0 mmol) of tert-butyl 1-(3-
methylbut-2-enoyl)-2,3-dihydro-1H-benzo[e][1,4]diazepine-
4(5H)-carboxylate (9) in acetone (20 mL), 0.05 M) was continu-
ously irradiated with a 150 W Hg lamp in a mesoscale photo-
chemical flow reactor (5 mL reactor volume, FEP-tubing)
equipped with a Pyrex® filter at 10–15 °C until completion of
the reaction monitored by liquid chromatography. The solution
was concentrated in vacuo, and the residue was directly puri-
fied by column chromatography on silica (eluent: 10–30% EtOAc
(27) Acetone has a UV cutoff of ca. 330 nm (although this effect is
diminished in
a microfluidic setup with very short path
lengths), see: Reichardt, C. Solvents and Solvent Effects in
Organic Chemistry; Wiley-VCH: Weinheim, 2003, 3rd ed., 482.
(28) See ref. 24 for the setup of the flow cell.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2016, 27, A–F