Synthesis of novel isochromene derivatives by tandem Ugi reaction/ nucleophilic substitution

Article abstract of DOI:10.1055/s-0029-1218553

Unprecedented 3-amino-4-amido-lH-isochromenes have been prepared in three steps from protected ortho-(hydroxymethyl)benzaldehyde, with the introduction of three diversity inputs, through a Ugi reaction followed by intramolecular nucleophilic substitution. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-0029-1218553

LETTER
85
Synthesis of Novel Isochromene Derivatives by Tandem Ugi Reaction/
Nucleophilic Substitution
I
sochromenes fr
u
om Tandem
c
U
gi-SN a Banfi,*a Andrea Basso,a Francesco Casuscelli,b Giuseppe Guanti,a Farah Naz,a Renata Riva,a Paola Zitoa
a
Department of Chemistry and Industrial Chemistry, University of Genova, Via Dodecaneso 31, 16146 Genova, Italy
Fax +39(010)3536118; E-mail: banfi@chimica.unige.it
b
Nerviano Medical Sciences Srl, Viale Pasteur 10, 20014 Nerviano (MI), Italy
Received 23 October 2009
(Scheme 1) was anticipated. Although reports of intramo-
lecular Mitsunobu reactions of unactivated secondary
amides are rare in the literature,9 our previous experience
suggested that this process could be feasible. In particular,
we have noticed that ketopiperazines6 or b-lactams7 could
be generated by this type of cyclization under Mitsunobu
or Mitsunobu-like conditions.
Abstract: Unprecedented
have been prepared in three steps from protected ortho-(hydroxy-
methyl)benzaldehyde, with the introduction of three diversity in-
3-amino-4-amido-1H-isochromenes
puts, through
a Ugi reaction followed by intramolecular
nucleophilic substitution.
Key words: multicomponent reactions, heterocycles, iso-
chromenes, enzymatic acylation, nucleophilic substitutions
From the point of view of atom economy, the use of un-
protected hydroxy aldehydes 1 (R5 = H), which obviously
exists as the corresponding lactols, would be ideal. How-
ever, when we prepared known lactol 410 by reduction of
phthalide 3, and submitted it to a model Ugi reaction, the
desired adduct 10a was isolated in poor yields, and its pu-
rification was very troublesome (Scheme 2).
The isocyanide-based multicomponent reactions (IM-
CRs), named after Passerini1 and Ugi,2 are very useful for
the generation of drug-like libraries, as demonstrated by
their extensive applications in the pharmaceutical realm.3
The classical versions of the Passerini and Ugi reactions
are perfectly suited for exploring substituent diversity, but
not for generating different (especially heterocyclic) scaf-
folds. Many attempts have been made to overcome this
drawback, but probably the most fruitful approach in-
volves coupling the Ugi and Passerini reactions with post-
condensation transformations.4 Over the last years, we
have been particularly interested in tandem Ugi-SN2 or
Ugi-SN2¢ processes, where the leaving group is an alco-
holic hydroxy group that is activated in situ, while the nu-
cleophile may be either an additional function suitably
introduced into one of the initial components,5–7 or the
isocyanide-derived secondary amide.8
b
c
OH
OH
OH
O
OAc
85%
91%
5
6
O
3
4
d,e
a
92%
70%
OSiMe2t-Bu
OH
f
OSiMe2t-Bu
O
O
100%
7
OH
8
g
(see Table)
g
< 20%
OH
OSiMe2t-Bu
R1
R1
OH
h
OR5
CHO
HN
HN
R1
(see Table)
HN
?
O
O
R4
N
R3
N
R3
1
O
R3
R4
R2
10a–h
R2
R1 NC
N
R2
R2 NH2
R3 CO2H
9a–h
O
O
2
O
Scheme 2 Reagents and conditions: (a) DIBAL-H, CH2Cl2, –78 °C;
(b) LiAlH4, Et2O–THF, 0 °C (30 min)→r.t., 3 h; (c) Supported PPL13
(50 mg/mmol of 6), vinyl acetate–i-Pr2O, 1:2, r.t., 40 h; (d) TB-
DMSCl, imidazole, DMF, 0 °C, 75 min; (e) KOH (0.4 M in MeOH),
r.t., 30 min; (f) (COCl)2, DMSO, Et3N, CH2Cl2, –78 °C → –50 °C; (g)
R2NH2, R1NC, R3CO2H, MeOH, 3 Å MS, r.t., 48 h; (h) MeCN–aq
40% HF (20:1), 0 °C, 3–5 h.
Scheme 1
In a continuation of these studies, we planned to use (hy-
droxymethyl)benzaldehydes of general formula 1 (either
protected or unprotected) as the hydroxy group-contain-
ing component, in a sequence of Ugi MCR’s followed by
nucleophilic substitution. In this design, replacement of
the hydroxy group with the secondary amide nitrogen of 2
Thus, instead of the lactol, we preferred to use protected
aldehyde 8, which we planned to obtain from 5 through
monoprotection to mono-silyl ether 7. However, after
phthalide 3 was reduced to diol 5,11 monoprotection to
mono-silyl ether 7 was found to proceed with almost no
selectivity and a maximum yield of 42%.12 Thus, a longer,
SYNLETT 2010, No. 1, pp 0085–0088
Advanced online publication: 09.12.2009
DOI: 10.1055/s-0029-1218553; Art ID: G35309ST
x
x.
x
x.
2
0
1
0
© Georg Thieme Verlag Stuttgart · New York

86
Banfi et al.
LETTER
but much more efficient, route was devised that took ad-
vantage of a selective enzymatic acetylation (Scheme 2).
After screening various enzymes, the best results were ob-
tained with supported pig pancreatic lipase (PPL),13 which
gave monoacetate 6 in excellent yield, despite the fact that
the acylating agent (vinyl acetate) was used in large ex-
cess. After protecting-group exchange14 and Swern oxida-
tion, aldehyde 8 was produced in 71% overall yield from
phthalide 3. This method is simpler, cheaper, and more ef-
ficient than previously reported synthetic approaches to
monoprotected diol 5.15
OH
HN
OMs
HN
Cl
R1
R1
R1
b
a
HN
O
R3
O
R3
O
R3
N
N
N
R2
R2
R2
O
O
O
10a–h
11a–h
12a–h
c
O
R1
O
O
O
N
H
This time the Ugi condensations with a variety of isocya-
nides, amines and carboxylic acids proceeded in satisfac-
tory yield (see Table 1); alcohols 10a–h were then
obtained by desilylation. During this last reaction, care
must be taken in the case of aromatic amide 10c because
of the formation of variable amounts of phthalide deriva-
tives by intramolecular SNAc.
d
R1
N
H
N
N
R3
R3
R2
R2
O
O
14a–h
13a–h
Scheme 3 Reagents and conditions: (a) MsCl, Et3N, CH2Cl2,
–15 °C; (b) warming to r.t.; (c) DMF, TBAI, Et3N/Et3NH+ buffer,
100 °C; (d) exposure to air at r.t. in the dry state.
With hydroxy amides 10a–h in hand, we studied their cy-
clization under a variety of conditions; we focused espe-
cially on compounds 10a–c. However, neither typical
Mitsunobu conditions (DEAD, DIAD or TBAD with
Ph3P), nor the modified Hanessian procedure (sulfuryl di-
imidazole, NaH, DMF)6,16 gave any cyclization product.
Apart from MS data (which cannot clearly discriminate
between different isomers), the structures of 13a and 13b
were unambiguously proved by NMR (1H, 13C, COSY,
HSQC, HMBC)19 and IR studies. The reaction conditions
were optimized in order to allow a ‘one-pot’ procedure
(Method A), and the methodology was extended to gener-
ate the new isochromenes 13d–h listed in Table 1, by
varying the isocyanide, the amine, and the carboxylic ac-
id. The methodology proved to be fairly general, although
chloride 12c, which is derived from an aromatic isocya-
nide, failed to afford 13c under the same conditions. The
overall yields were sometimes only moderate, especially
when bulky carboxylic acids were employed as Ugi in-
puts. However, it is worth noting that isochromenes 13
Conversion of alcohols 10a and 10b into the correspond-
ing methanesulfonates 11a and 11b, by treatment with
MsCl and Et3N at –15 °C in CH2Cl2, followed by warming
to room temperature, gave the corresponding chlorides
12a and 12b (Scheme 3).17 Heating of these chlorides at
100 °C in DMF in the presence of a buffer eventually led
to cyclization18 but, to our surprise, the resulting products
were the unprecedented isochromenes 13a and 13b, stem-
ming from O-cyclization instead of N-cyclization.
Table 1 Results of Combinatorial Synthesis of Isochromenes 13
Compd R1
R2
R3
Et
Yield of 9 (%)
Yield of 10 (%) Cyclization methodaYield of 13 (%)
a
n-Pent
i-Bu
61
87
A
B
B
–
61
68
35
–
b
c
c-Hex
n-Bu
n-Bu
n-Bu
Bn
3-BrC6H4
3-BrC6H4
3-BrC6H4
Et
75
53
60
70
93
45
67
75
4-MeOC6H4
n-Pent
d
e
A
A
B
A
B
A
A
25
54
68
35
48
59
75
t-Bu
f
c-Hex
i-Bu
Et
73
68
g
c-Hex
c-Hex
2-furyl-CH2
Bn
MeOCH2
Et
80
84
82
91
h
a Method A: MsCl, Et3N, CH2Cl2, –15 °C → r.t.; evaporation, then DMF, Et3N (2 equiv), TBAI (0.5 equiv), 100 °C, 2–3 h (overall yields from
10). Method B: MsCl, Et3N, CH2Cl2, –15 °C → r.t.; chromatography of chloride 12, then DMF, Et3N (2 equiv), pTSA (0.5 equiv), TBAI (0.5
equiv), 100 °C, 2–3 h.
Synlett 2010, No. 1, 85–88 © Thieme Stuttgart · New York

LETTER
Isochromenes from Tandem Ugi-SN
87
(7) Banfi, L.; Basso, A.; Guanti, G.; Lecinska, P.; Riva, R. Mol.
Divers. 2008, 12, 187.
(8) Banfi, L.; Basso, A.; Cerulli, V.; Guanti, G.; Riva, R. J. Org.
Chem. 2008, 73, 1608.
(9) Swamy, K. C. K.; Kumar, N. N. B.; Balaraman, E.; Kumar,
K. V. P. P. Chem. Rev. 2009, 109, 2551.
(10) Mikami, K.; Ohmura, H. Org. Lett. 2002, 4, 3355.
(11) Brown, H. C.; Narasimhan, S.; Choi, Y. M. J. Org. Chem.
1982, 47, 4702.
(12) Schomaker, J. M.; Travis, B. R.; Borhan, B. Org. Lett. 2003,
5, 3089.
(13) Banfi, L.; Guanti, G.; Riva, R. Tetrahedron: Asymmetry
1995, 6, 1345.
(14) The aldehyde derived from oxidation of monoacetate 6
could not be used for the preparation of alcohols 10, because
acetyl removal after the Ugi reaction brought about
cyclization to the phthalide derivatives with elimination of
R1NH2
(15) Baciocchi, E.; Bernini, R.; Lanzalunga, O. Chem. Commun.
1993, 1691.
can be obtained in just three steps from aldehyde 8, with
the introduction of three diversity inputs.20
The double bond of isochromenes 13 is expected to be
electron-rich and, therefore, able to donate electrons in
ionic or radical processes. We accidentally discovered
that these compounds are indeed reactive towards molec-
ular oxygen. Whereas they appear to be stable if stored in
the dry state at –20 °C, they slowly decompose when left
dry in the air at room temperature. For example, when
compound 13a was left on the bench for five days in an
open flask, complete decomposition took place. The main
decomposition product was unambiguously identified by
13
NMR (1H, C, COSY, HSQC, HMBC), MS and IR21 as
the urethane 14a, which is derived from incorporation of
one molecule of oxygen. In contrast, when 13a was left in
CDCl3 solution at r.t. for five days, only 10% conversion
into 14a was observed. In the other cases the behavior was
similar, with decomposition being faster for 13a, 13b,
13d, and 13e (oils) than for 13f–h (solids). This reaction,
which most likely proceeds through single-electron trans-
fer from 13 to O2, unveils the uncommon tendency of
these isochromenes to donate single electrons, as happens,
for example, with FADH2. In fact, 3-aminoisochromenes
have been shown to be potential photocaged com-
pounds.22
(16) Hanessian, S.; Couture, C.; Wiss, H. Can. J. Chem. 1985, 63,
3613.
(17) 12a and 12b were stable enough to be purified by
chromatography and characterized by NMR. However,
compound 12a, upon GC-MS analysis, gave a peak
corresponding to the cyclized products 13a.
(18) Interestingly, treatment of either 11a,b or 12a,b with strong
bases (NaH, t-BuOK, LiHMDS) failed to afford any
cyclization product. Buffered thermal conditions were
needed for this cyclization to occur. In the one-pot procedure
(Method A), Et3NH+Cl– was already present in the crude
chloride, whereas, in the two-step process (Method B), a
triethylammonium salt had to be added.
(19) In compounds 13, the protons of all CH2 groups, which
should be enantiotopic since there are no stereogenic centres,
behave as if they were diastereotopic instead. A partial
coalescence of these non-isochronous protons was observed
only at 130 °C. This phenomenon is likely due to restricted
rotation around the C4–N bond, which generates axial
chirality.
(20) Typical experimental procedure: cyclization of 10h to give
13h. A solution of alcohol 10h (0.50 mmol) in anhydrous
CH2Cl2 (3.5 mL) was cooled to –15 °C, and treated with
Et3N (153 mL, 1.1 mmol) and methanesulfonyl chloride (46
mL, 0.60 mmol). After 1 h, the cooling bath was removed and
Et3N (139 mL, 1.0 mmol) was added. After stirring for 1 h at
r.t., the solvent was removed under reduced pressure. The
residue was taken up in anhydrous DMF (6.5 mL), treated
with Et3N (139 mL, 1.0 mmol) and TBAI (92 mg, 0.25
mmol), and stirred at 100 °C for 2 h. After cooling, the
mixture was treated with sat. aq NH4Cl (25 mL) and H2O (25
mL) and extracted with EtOAc. The organic phases were
washed with sat. aq NaCl, evaporated, and purified by
chromatography (PE–EtOAc, 85:15) to give pure 13h. Rf =
0.69 (PE–acetone, 70:30). 1H NMR (CDCl3):19 d = 7.45–
7.35 (m, 2 H, ArH), 7.33–7.25 (m, 3 H, ArH), 7.21 (dt, Jd =
1.2 , Jt = 7.5 Hz, 1 H, H-7), 7.02 (d, J = 7.2 Hz, 1 H, H-9),
6.96 (dt, Jd = 0.9, Jt = 7.3 Hz, 1 H, H-8), 6.65 (d, J = 7.8 Hz,
1 H, H-6), 5.62 (d, J = 13.2 Hz, 1 H, CHHPh), 5.03 and 4.90
(AB system, J = 12.6 Hz, 2 H, H-1), 3.71 (d, J = 13.5 Hz, 1
H, CHHPh), 3.39 (d, J = 8.4 Hz, 1 H, HN), 3.03 (dtt, Jt = 4.0,
11.5, Jd = 8.0 Hz, 1 H, CHNH), 2.24 (q, J = 7.2 Hz, 2 H,
CH2CH3), 1.75–1.40 (m, 4 H, c-Hexeq CHH), 1.30 (m, 1 H,
eq. CHH of c-Hex), 1.13 (tq, Jt = 3.1 Hz, Jq = 12.9 Hz, 1 H,
axial CH2CHHCH2), 1.13–1.02 (m, 1 H, axial CH2CHHCH2),
1.08 (t, J = 7.2 Hz, 3 H, CH3CH2), 0.94 (tq, Jt = 3.0 Hz, Jq =
12.3 Hz, 1 H, axial CH2CHHCH2), 0.71 (dq, Jd = 3.3 Hz,
This reactivity, together with the likely ability of the elec-
tron-rich double bond to undergo electrophilic additions,
cycloadditions and so on, could make compounds 13 use-
ful intermediates for various organic transformations.
Acknowledgment
This research was financially supported by Nerviano Medical Sci-
ences Srl. We thank Dr. Francesca Musumeci and Dr. Romina Vi-
tale for their experimental contribution to this work, and
Fondazione San Paolo for a contribution for the purchase of the
NMR instrument.
References and Notes
(1) (a) Banfi, L.; Riva, R. The Passerini Reaction, In Organic
Reactions, Vol. 65; Overman, L. E., Ed.; Wiley: Hoboken,
2005, 1–140. (b) Passerini, M. Gazz. Chim. Ital. 1921, 51,
126.
(2) Ugi, I.; Steinbrueckner, C. Angew. Chem. 1960, 72, 267.
(3) (a) Akritopoulou-Zanze, I. Curr. Opin. Chem. Biol. 2008,
12, 324. (b) Doemling, A. Chem. Rev. 2006, 106, 17.
(c) Hulme, C.; Gore, V. Curr. Med. Chem. 2003, 10, 51.
(4) (a) Akritopoulou-Zanze, I.; Djuric, S. W. Heterocycles
2007, 73, 125. (b) Hulme, C.; Nixey, T.; Bienaymé, H.;
Chenera, B.; Jones, W.; Tempest, P.; Smith, A. L. Methods
Enzymol. 2003, 369, 469.
(5) (a) Banfi, L.; Basso, A.; Cerulli, V.; Guanti, G.; Monfardini,
I.; Riva, R. Mol. Divers. 2009, in press; DOI: 10.1007/
s11030-009-9210-4. (b) Banfi, L.; Basso, A.; Guanti, G.;
Lecinska, P.; Riva, R. Org. Biomol. Chem. 2006, 4, 4236.
(c) Banfi, L.; Basso, A.; Guanti, G.; Lecinska, P.; Riva, R.;
Rocca, V. Heterocycles 2007, 73, 699.
(6) Banfi, L.; Basso, A.; Guanti, G.; Kielland, N.; Repetto, C.;
Riva, R. J. Org. Chem. 2007, 72, 2151.
Synlett 2010, No. 1, 85–88 © Thieme Stuttgart · New York

88
Banfi et al.
LETTER
Jq = 11.7 Hz, 1 H, axial CH-CHH-CH2), 0.22 (dq, Jd = 3.3
Hz, Jq = 11.8 Hz, 1 H, axial CH-CHH-CH2). 13C NMR
(CDCl3): d (attribution of signals was made with the aid of
HSQC and HMBC experiments) = 177.4 (C=O), 155.1 (C-
3), 139.3 (quarternary C of benzyl), 133.0 (C-5), 129.7 (×2),
128.8 (×2), 127.7 (CH of benzyl), 128.8 (C-7), 124.3 (C-10),
123.8 (C-9), 122.5 (C-8), 116.0 (C-6), 93.4 (C-4), 69.2 (C-
1), 50.7 (CH2Ph), 50.7 (CHN), 34.3, 33.1, 25.6, 25.1, 24.9
(c-Hex CH2), 25.9 (CH2CH3), 9.6 (CH3). HRMS (EI): m/z
[M]+ calcd for C25H30N2O2: 390.2307; found: 390.2235. GC-
MS: m/z (%) = 390 (41.1) [M]+, 334 (5.9), 333 (8.2), 299
(71.5), 243 (63.1), 207 (10.4), 206 (8.5), 161 (35.9), 144
(27.6), 143 (4.8), 118 (18.6), 117 (14.0), 116 (19.3), 106
(5.1), 91 (100.0), 89 (8.7), 65 (9.6), 57 (20.7), 55 (19.5), 41
(13.6). IR (CHCl3): 3399, 3012, 2924, 2850, 1625, 1594,
1569, 1447, 1397, 1362, 1347, 1226, 1097, 1073, 917 cm–1.
Hz, 2 H, CH2NH), 2.44 (q, J = 7.4 Hz, 2 H, O=CCH2CH3),
2.00 [nonuplet, J = 6.8 Hz, 1 H, CH(CH3)2], 1.48 (quint., J =
7.0 Hz, 2 H, NCH2CH2), 1.38–1.24 (m, 4 H, CH2CH2CH3),
1.05 (t, J = 7.4 Hz, 3 H, CH3CH2), 0.89 [t (covered by the d
at 0.88), J = 7.2 Hz, 3 H, CH3CH2CH2], 0.88 [d, J = 6.8 Hz,
6 H, (CH3)2CH]. 13C NMR (CDCL3): d (attribution was
made also by HSQC and HMBC) = 177.9 (O=CCH2), 173.3
(CCON), 155.9 (OCON), 135.7, 135.2 (C-1, C-2), 131.0,
129.7, 128.1, 127.6 (C-3, C-4, C-5, C-6), 63.9 (CH2O), 52.7
(CH2 of isobutyl), 41.1 (NHCH2), 32.0 (O=CCH2CH3), 29.6
(NCH2CH2), 28.8 (NCH2CH2CH2), 28.2 [CH(CH3)2], 22.3
(CCH2CH3), 20.2 [(CH3)2CH], 14.0 (CH3CH2C), 9.7
(CH3CH2CO). HMRS (EI): m/z [M]+ C21H32N2O4:
376.2362; found: 376.2375. GC-MS: m/z (%) = 376 (0.3)
[M]+, 347 (5.6), 232 (11.4), 189 (4.0), 146 (19.5), 135
(100.0), 134 (60.1), 133 (9.9), 130 (16.3), 118 (24.4), 105
(6.8), 91 (6.7), 90 (8.3), 86 (12.4), 77 (5.9), 74 (10.5), 57
(25.6), 56 (13.0), 43 (9.9), 41 (9.4). IR (CHCl3): 3445, 2952,
2867, 1709 (s), 1654 (s), 1602, 1461, 1359, 1326, 1183,
1117, 1047 cm–1.
(21) A pure dry sample of 13a (87.5 mg) was left for 5 days in an
open flask on the bench at r.t. After this time, TLC showed
that it had completely decomposed, giving a new main spot
on the TLC plate. This was isolated by chromatography
(PE–acetone, 85:15→70:30). Rf = 0.46 (PE–acetone, 75:25).
Yield: 45.5 mg (48%). 1H NMR (CDCl3): d = 7.54–7.43 (m,
2 H), 7.41–7.30 (m, 2 H), 5.22 (s, 2 H, CH2O), 4.72 (br t, 1
H, NH), 3.58 (d, J = 7.5 Hz, 2 H, CH2-iPr), 3.16 (q, J = 6.7
(22) Cory, M. G.; Richards, N. G. J.; Zerner, M. C. Modeling the
Hydrogen Bond, ACS Symposium Series #569; American
Chemical Society: Washington DC, 1994, 222–234.
Synlett 2010, No. 1, 85–88 © Thieme Stuttgart · New York

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.