An acid-catalysed hydroamination approach to isoindolines

Article abstract of DOI:10.1055/s-0031-1291155

Acid-catalysed intramolecular hydroaminations of 2-alkenylarylethylamine derivatives lead smoothly to isoindolines via benzylic carbenium ion generation. Georg Thieme Verlag Stuttgart · New York.

Full text of DOI:10.1055/s-0031-1291155

LETTER
▌1667
lAettner Acid-Catalysed Hydroamination Approach to Isoindolines
Acid-Catalysed Hydroamination Approach to Isoindolines
Laura Henderson,a David W. Knight,*a Andrew C. Williamsb
a
School of Chemistry, Cardiff University, Main College, Park Place, Cardiff, CF10 3AT, UK
Fax +44(29)20874030; E-mail: knightdw@cf.ac.uk
b
Eli Lilly and Co. Ltd., The Lilly Research Centre, Erl Wood Manor, Windlesham, Surrey, GU20 6PH, UK
Received: 16.03.2012; Accepted: 12.04.2012
R1
Abstract: Acid-catalysed intramolecular hydroaminations of 2-
alkenylarylethylamine derivatives lead smoothly to isoindolines via
benzylic carbenium ion generation.
TfOH
CO2Me
R1
CO2Me
TsHN
N
CH2Cl2
Ts
3
4
Key words: cyclisation, sulfonamides, heterocycles, hydroamina-
tion, isoindolines
Scheme 2 Initial model reactions
In our recently reported synthesis of the indole alkaloid
α-cyclopiazonic acid,1 a central key step was an acid-cat-
alysed cascade cyclisation of the indole-4-methanol 1
which resulted in isolation of an excellent yield of the
hoped-for tetracyclic derivative 2 (Scheme 1).
Herein, we wish to report our initial findings in this pro-
jected series of ideas, which has led to a new and relative-
ly simple approach to the synthesis of substituted
isoindolines.
Reducing the cascade sequence shown in Scheme 1 to a
simpler form results in the idea of making isoindolines 64
by cyclisation of 2-alkenyl-benzylamine derivatives 5.
Specifically, this would feature sequential alkene proton-
ation and trapping of the resulting benzylic carbenium ion
by the pendant amine nitrogen (Scheme 3). Our initial re-
search had revealed that if the amine group in related cy-
NHNs
NNs
H
TBDPSO
CO2Me
CO2Me
H
TfOH
CH2Cl2, 20 °C
NTs
NTs
clisation precursors was derivatised as either
a
1
2
sulfonamide or an acid-stable carbamate, then such cycli-
sations under acidic conditions were indeed viable and did
not simply result in complete protonation of the amine
group, as might be expected. Such cyclisations leading to
pyrrolidines (Scheme 2) were indeed very efficient, espe-
cially when using the super acid trifluoromethanesulfonic
(triflic) acid to protonate the alkene group.2
Scheme 1 A cascade cyclisation in the synthesis of α-cyclopiazonic
acid
We assumed that the mechanism of the cascade was the
obvious one: protonation of the pendant oxygen, loss of
the silyloxy group as the silanol, resulting in formation of
a benzylic carbenium ion, and cascade cyclisation with
generation of an intermediate tertiary aliphatic carbenium
ion which is finally trapped by the sulfonamide function.
In the light of this, we speculated whether similar but
somewhat simpler cyclisations onto benzylic carbenium
ions would be equally facile. Suitable benzylic carbenium
ions could be generated by alkene protonation, as was
used in our initial model studies, which demonstrated the
viability of this methodology in the synthesis of pyrro-
lidines [e.g., 4] from alkenyl sulfonamides 3 by exposure
to triflic acid (Scheme 2).2
NHR1
H+
?
NR1
R2
R2
5
6
Scheme 3 The proposed new approach to isoindolines
However, an immediate concern was that partial proton-
ation of the modified amine function NHR1, which we an-
ticipated would either be a sulfonamide or an acid-stable
carbamate, rather than the alkene group in the precursors
5 could lead to the ammonium salts 7 and thence to loss of
the former and nonproductive generation of an alternative
benzylic carbenium ion 8 (Scheme 4).
Such methodology has the potential to be applied to the
elaboration of a variety of other saturated and semisaturat-
ed nitrogen-based heteroaromatic systems in what, over-
all, is an acid-catalysed alkene hydroamination reaction.3
In addition, nonregioselective cyclisation of the modified
amine group onto either end of the pendant alkene could
lead to mixtures of products. In the event, all of these con-
cerns proved to be unfounded. The results obtained thus
far are summarized in Table 1.
SYNLETT 2012, 23, 1667–1669
Advanced online publication: 11.06.2012
DOI: 10.1055/s-0031-1291155; Art ID: ST-2012-D0246-L
© Georg Thieme Verlag Stuttgart · New York
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6

1668
L. Henderson et al.
LETTER
Table 1 Acid-Catalysed Cyclisations of 2-Alkenylphenyl Sulfon-
amides 9 to the Isoindolines 10 (continued)
NH2R1
?
H+
?
5
NHTs
H+
NTs
R2
R2
R
7
8
R
9
10
Scheme 4 A possible alternative pathway
Entry Product 10
Acid
Time TempYield
(h) (°C) (%)
Table 1 Acid-Catalysed Cyclisations of 2-Alkenylphenyl Sulfon-
amides 9 to the Isoindolines 10
NTs
NHTs
H+
NTs
TfOH
concd H2SO4
3
3
40 100
40 98
8
R
R
9
10
10h
Entry Product 10
Acid
Time TempYield
(h) (°C) (%)
All of the precursor sulfonamides 9a–h were prepared
using optimized Suzuki–Miyaura couplings5 between N-
tosyl-2-bromobenzylamine6 and the corresponding alke-
nylboronic acid. This employed a premix of the palladium
catalyst [Pd(OAc)2], dtbpf [1,1′-bis(di-tert-butylphosphi-
no)ferrocene] as ligand, buffered by excess potassium
phosphate in 1:1 aqueous ethanol under microwave heat-
ing.
NTs
TfOH
24
0
98
92
1
concd H2SO4 48 20
10a
10b
10c
10d
10e
10f
NTs
NTs
In the first series of trials, we used the original method for
the related acid-catalysed pyrrolidine syntheses which
consisted of exposing the precursors 9 to 0.5 equivalents
of triflic acid as a ca. 3 mM solution in dry dichlorometh-
ane for various times and at various temperatures. The
progress of the cyclisations was monitored both by TLC
and 1H NMR spectroscopy. Once completed, the cyclisa-
tions were worked up by quenching the acid using excess
2 M aqueous potassium carbonate, separating the dichlo-
romethane layer, which was then dried, filtered through
silica gel, and evaporated. The yields quoted in Table 1 re-
TfOH
24 40
81
70
2
3
4
5
6
7
Cl
concd H2SO4 48 40
TfOH
24 40
84
73
F
concd H2SO4 24 40
1
fer to products which were pure according to both H
NTs
NTs
NTs
NTs
NMR and 13C NMR spectroscopy.
TfOH
24 40
71
67
CF3
concd H2SO4 24 40
We were delighted that the first substrate 9a, an otherwise
unsubstituted stilbene derivative, underwent slow but
very clean cyclisation at 0 °C to give an essentially quan-
titative yield of the hoped-for product 10a. In all probabil-
ity, mild heating would deliver a similarly excellent yield
in a (much) reduced time. Heating was required for the
conversion of the much more electron-deficient substrates
9b–d, but in each case, carrying out the reaction in gently
refluxing dichloromethane delivered high yields of the
isoindolines 10b–d.7 Presumably, this type of substituent
significantly retards the rate of alkene protonation by re-
ducing its electron density. Quite why the 4-methyl deriv-
ative 9e required similar conditions remains an oddity.
None of these or the following products 10 contained any
of the alternative six-membered tetrahydroisoquinoline
products, which might be formed from the isomeric carbe-
nium ion resulting from protonation at the alternate ben-
zylic centre, followed by addition of the sulfonamide
group to the distal end of the alkene.
TfOH
24 40
86
67
concd H2SO4 24 40
TfOH
concd H2SO4
2
4
40
40 100
93
TfOH
concd H2SO4 24
6
0
0
70
78
10g
Synlett 2012, 23, 1667–1669
© Georg Thieme Verlag Stuttgart · New York

LETTER
Acid-Catalysed Hydroamination Approach to Isoindolines
1669
The conditions required to achieve cyclisations onto ali-
phatic alkenes were significantly milder, presumably be-
cause in such substrates, an extended stilbene
chromophore is not ruptured during protonation. Thus, the
cyclohexylethenyl derivative 9f required a much briefer
period of reflux to deliver a respectable yield of the ex-
pected product 10f, while the cyclohexenyl analogue 9g
underwent smooth cyclisation at 0 °C to give the spiro de-
rivative 10g presumably because, in this case, the interme-
diate carbenium ion is tertiary. This is a particularly useful
result, as such highly substituted derivatives are not read-
ily accessible using alternative approaches. Finally, a
purely aliphatic derivative 10h was also obtained relative-
ly rapidly as would be expected in the light of the forego-
ing results.
References
(1) Griffiths-Jones, C. M.; Knight, D. W. Tetrahedron 2011, 67,
8515.
(2) (a) Griffiths-Jones, C. M.; Knight, D. W. Tetrahedron 2010,
66, 4150. (b) Schlummer, B.; Hartwig, J. F. Org. Lett. 2002,
4, 1471.
(3) For recent reviews of alkene hydroamination, see:
(a) Müller, T. E.; Beller, M. Chem. Rev. 1998, 98, 675.
(b) Nobis, M.; Driessen-Hölscher, B. Angew. Chem. Int. Ed.
2001, 40, 3983. (c) Pohlki, F.; Doye, S. Chem. Soc. Rev.
2003, 32, 104. (d) Müller, T. E.; Hultzsch, K. C.; Yus, M.;
Foubelo, F.; Tada, M. Chem Rev. 2008, 108, 3795.
(e) Majumdar, K. C.; Debnath, P.; De , N.; Roy, B. Curr.
Org. Chem. 2011, 15, 1760.
(4) Bergman, J.; Janosik, T. Comp. Heterocycl. Chem.; Vol. 3;
Katritzky, A. R.; Ramsden, C. A.; Scriven, E. F. V.; Taylor,
R. J. K., Eds.; Elsevier: Oxford, 2008, 269–351; and
references cited therein.
An irritating drawback with this methodology is that it re-
quired triflic acid. Obviously a highly corrosive and also
relatively expensive reagent, this has the additional com-
plication of being rather difficult to store and to keep com-
pletely anhydrous, meaning that its quantification and
indeed its reactivity (anhydrous vs. hydrate) is always a
matter of some uncertainty. For these reasons, we chose to
determine whether concentrated sulfuric acid2b could be
used with similar efficacy, as it is so much cheaper and
also more easily preserved in an anhydrous state. A disad-
vantage appeared to be that it is not especially soluble in
dichloromethane. This turned out not to be relevant, as
shown by the results presented in Table 1. Each of the pre-
cursors 9 also underwent smooth cyclisations when ex-
posed to similar amounts of concd H2SO4 suspended in
dichloromethane. In general, such cyclisations were
slightly slower and delivered slightly inferior yields.
However, it should be kept in mind that the conditions de-
tailed in Table 1 are not completely optimized and that al-
ternative combinations of both time and temperature
could be more efficacious in some of the examples.
(5) Henderson, L.; Knight, D. W.; Williams, A. C. Tetrahedron
Lett. 2012, submitted.
(6) Zhu, M.; Fujita, K.; Yamaguchi, R. Org. Lett. 2010, 12,
1336.
(7) Preparation of Isoindolines 10b–d; Typical Procedures:
1-(4-Chlorobenzyl)-2-tosylisoindoline (10b)
i) (E)-N-[2-(4-Chlorostyryl)benzyl]-4-methylbenzene-
sulfonamide 9b (95 mg, 0.238 mmol, 1.0 equiv) was
dissolved in dry dichloromethane (10 ml), to which was
added TfOH (0.20 ml, 0.119 mmol, 0.5 equiv). The resulting
solution was heated to 40 °C for 24 h, then cooled to room
temperature and quenched using saturated aqueous K2CO3.
The quenched solution was then separated and the aqueous
layer was extracted with dichloromethane (3 x 10 ml). The
combined organic layers were dried over K2CO3, filtered
through a plug of silica gel and the filtrates and washings
evaporated to give the isoindole 10b as a yellow oil (77 mg,
yield: 81%).
ii) (E)-N-[2-(4-Chlorostyryl)benzyl]-4-methylbenzene-
sulfonamide 9b (106 mg, 0.267 mmol, 1.0 equiv) was
dissolved in dichloromethane (10 ml), to which was added
concentrated sulfuric acid (2 drops). The resulting
suspension was stirred at 40 °C for 48 h, then cooled to room
temperature and quenched with saturated aqueous K2CO3.
The quenched solution was then separated and the aqueous
layer was extracted with dichloromethane (3 x 10 ml). The
combined organic layers were dried over K2CO3, filtered
through a plug of silica gel and the filtrates and washings
evaporated to give the isoindole 10b as a yellow oil (74 mg,
yield: 70%).
The two samples showed identical spectroscopic and
analytical data. IR (film): 3277, 3062, 3028, 2923, 2855,
1598, 1448, 1274, 815, 750, 658 cm–1. 1H NMR (400 MHz,
CDCl3) δ = 7.75–7.52 (m, 4 H, 2 × Ts CH and 2 × Ar CH),
7.35–6.88 (m, 8 H, 2 × Ts CH and 6 × Ar CH), 5.31 (d, 1 H,
J = 15.4 Hz, CHaHbN), 4.59 (d, 1 H, J = 15.4 Hz, CHaHbN),
4.11 (t, 1 H, J = 6.2 Hz, CHN), 2.98 (app. br s, 2 H, CH2),
2.28 (s, 3 H, Ts CH3). 13C NMR (400 MHz, CDCl3) δ =
143.2 (C), 139.8 (C), 137.0 (C), 132.6 (C), 132.4 (C), 132.2
(C), 129.6 (ArCH), 129.5 (2 × TsCH), 128.6 (ArCH), 128.5
(ArCH), 128.4 (ArCH), 127.4 (ArCH), 127.2 (ArCH), 127.1
(2 × TsCH), 126.5 (ArCH), 125.9 (ArCH), 54.7 (CHN), 44.0
(CH2N), 32.3 (CH2), 21.6 (TsCH3). HRMS (APCI): m/z
calcd. for C22H21ClNO2S [M + H]+ = 398.0982 (Cl35); found
398.0983 (Cl35).
The present method should find many applications in this
area as it is simple and requires only two steps – a robust
palladium-catalysed coupling and a cyclisation. Many al-
ternative approaches to isoindolines are known4 and some
are equally efficient and relatively brief but, at the least,
the present method does not produce substantial amounts
of byproducts and noxious spent reagents and therefore is
relatively ‘green’. Obviously, a significant concern with a
method which employs such strong acids is its compatibil-
ity with potentially sensitive functional groups. In our pre-
vious work, we have shown that, at least, carboxylic acid
methyl and ethyl esters, sulfones, and isolated alkenes sur-
vive unmolested. More recently,8 we have found that
acetyloxy groups are especially good at masking alcohol
groups which could otherwise interfere with such cyclisa-
tions; the highly acidic conditions do not interfere with
this type of ester, even when heated. We therefore do not
view this as an especially serious drawback.
(8) AlHadi, A.; Knight, D. W. unpublished observations.
Acknowledgment
We are grateful to The Lilly Research Centre Ltd. and the EPSRC
for financial support.
© Georg Thieme Verlag Stuttgart · New York
Synlett 2012, 23, 1667–1669

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.