Long-range diastereoselectivity in an Ugi reaction: Stereocontrolled and diversity-oriented synthesis of tetrahydrobenzoxazepines

Article abstract of DOI:10.1002/ejoc.201300541

Salicylaldehydes and protected 1,2-amino alcohols have been convergently converted into a series of 2,3-dihydrobenzo[f][1,4]oxazepines, which undergo an Ugi-Joullie multicomponent reaction with unusual long-range diastereoselectivity. This protocol allows

Full text of DOI:10.1002/ejoc.201300541

FULL PAPER
DOI: 10.1002/ejoc.201300541
Long-Range Diastereoselectivity in an Ugi Reaction: Stereocontrolled and
Diversity-Oriented Synthesis of Tetrahydrobenzoxazepines
Luca Banfi,*[a] Alessandro Bagno,[b] Andrea Basso,[a] Carlo De Santis,[a] Renata Riva,[a] and
Federico Rastrelli[b]
Dedicated to the memory of Stefano Marcaccini[‡]
Keywords: Multicomponent reactions / Medicinal chemistry / Computational chemistry / Nitrogen heterocycles / Oxygen
heterocycles
Salicylaldehydes and protected 1,2-amino alcohols have
been convergently converted into a series of 2,3-dihydro-
benzo[f][1,4]oxazepines, which undergo an Ugi–Joullié
ration, in only two steps, of a series of drug-like tetrahydro-
benzo[f][1,4]oxazepines, with the combinatorial introduction
of four diversity points. The possibility of obtaining these
multicomponent reaction with unusual long-range diastereo- compounds in enantiomerically pure form was also demon-
selectivity. This protocol allows the diastereoselective prepa- strated.
severe stereochemical problems: a) the use of chiral isocy-
Introduction
anides, carbonyl compounds, or carboxylic acids as reaction
The Ugi multicomponent reaction is very useful in diver-
components has so far failed to result in any respectable
sity-oriented synthesis, since it allows the combination of
four diversity inputs in a single step, starting from sub-
strates that are easily available with a high degree of struc-
level of diastereoselectivity;[3] b) most α-chiral aldehydes[4]
and some α-chiral isocyanides[5] are prone to undergo race-
mization under the conditions of the Ugi reaction. In con-
tural variation. Although the classical products of an Ugi
trast, the steric biases imposed by a cyclic transition state
reaction are acyclic peptide-like compounds, the reaction
can facilitate the achievement of good to excellent dia-
has also been shown to be very useful for the combinatorial
stereoselectivities in intramolecular Ugi reactions.[6]
assembly of drug-like, conformationally biased nitrogen
In particular, several examples of diastereoselective Ugi
heterocycles. This has been mainly accomplished by two al-
reactions of cyclic imines (Ugi–Joullié reactions)[7] have
ternative strategies: a) coupling the Ugi reaction with a
been reported. These preformed imines typically also give
post-condensation cyclization;[1] b) performing an intramo-
higher reaction rates in relatively non-polar solvents such
lecular variant of the Ugi reaction.[2] For example, oxoacids,
as dichloromethane, and the neutral or slightly acidic con-
amino acids, and cyclic imines (corresponding to amino al-
ditions make both the imines[8] and the chiral isocyanides
dehydes or amino ketones) have been widely used as bifunc-
used[5c] less sensitive to racemization/epimerization pro-
tional components in this context.
cesses when they have a stereogenic centre at the α position.
Notwithstanding the reduction of the number of compo-
Moreover, chiral cyclic imines often give good to excellent
nents from four to three, intramolecular Ugi reactions are
diastereoselectivities in Ugi reactions, in contrast to the be-
attractive not only because they give nitrogen heterocycles
haviour of chiral acyclic imines derived from chiral carb-
directly in a single step, but also from a stereochemical
onyl compounds.
point of view. The classical Ugi reaction suffers from two
Most reported examples of the Ugi–Joullié reaction con-
cern five-membered rings (chiral pyrrolines or thiazolines).
[a] Department of Chemistry and Industrial Chemistry, University
Good diastereoselectivities have only been obtained when
the stereogenic centre is at the α position of the imine car-
bon,[8–9] although this attribute does not guarantee a high
stereoselectivity in all cases.[9c,10] When the stereogenic cen-
tre is α to nitrogen, only moderate diastereoselectivities have
been achieved.[11] Other examples involve six-membered
rings (tetrahydropyridines, oxazines, dehydromorpholines).
of Genova,
via Dodecaneso 31, 16146 Genova, Italy
Fax: +39-010-3536118
E-mail: banfi@chimica.unige.it
[b] Department of Chemical Sciences, University of Padova,
via Marzolo 1, 35131 Padova, Italy
[‡] Italian pioneer of multicomponent reactions
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201300541.
5064
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
Scheme 1. The two approaches for the synthesis of dihydrobenzoxazepines 1. a) TBAD (di-tert-butyl azodicarboxylate), PPh3, THF,
–15 °CǞroom temp.; b) LiAlH4, Et2O/THF, 0 °C; c) HCl, CH2Cl2/H2O, room temp.; d) (HCHO)n, MgCl2, Et3N; e) HC(OMe)3, MeOH,
Amberlyst 15; f) see Table 1.
In these cases, good diastereoselectivities have been ob-
Results and Discussion
tained when the stereogenic centre is α to the imine car-
Initially,[16] we synthesized dihydrobenzoxazepines 5a (R1
bon,[9d,12] but also when it is α[13] or even β[14] to the nitro-
= R2 = H; R3 = Me) and 5b (R1 = R2 = H; R3 = iBu)
gen. The latter report is the only example of long-range
diastereoselectivity in an Ugi–Joullié reaction.
At the outset of this work, no examples of diastereoselec-
tive Ugi reactions on chiral seven-membered cyclic imines
had been reported. Seven-membered rings are expected to
experience more complex conformational equilibria than
six-membered rings, and therefore it is anticipated to be
more difficult to obtain good diastereoselectivity, especially
according to Route A, starting from salicylic acid, via Wein-
reb hydroxamate 1 (Scheme 1). The key step was an inter-
molecular Mitsunobu reaction with protected alcohols 2a
(R3 = Me) or 2b (R3 = iBu) to give ethers 3a and 3b. Then,
the hydroxamates were reduced to give the corresponding
aldehydes (i.e., 4a and 4b), which were finally smoothly de-
protected to give cyclic imines 5. These were not isolated
but used directly in the Ugi–Joullié reaction.
when the chirality centre is far away from the electrophilic
Boc-protected 2-amino alcohol 2a (Boc = tert-butoxy-
imine carbon. Nevertheless, seven-membered nitrogen het-
carbonyl; R3 = Me) can be easily synthesized by protection
erocycles are very attractive from the point of view of me-
of commercially available 1-amino-2-propanol.[17] On the
other hand, protected 2-amino alcohol 2b was prepared
dicinal chemistry, where they represent “privileged scaf-
folds”. In a continuation of our previous research into the
starting from the Weinreb hydroxamate of Boc-glycine
diversity-oriented synthesis of seven-membered nitrogen
10,[15a,18] as shown in Schme 2, by addition of a Grignard
heterocycles using the Ugi reaction,[11a,15] we searched for
reagent, followed by ketone reduction.[15a,19]
new classes of chiral seven-membered cyclic imines that:
a) can be easily prepared, also in a combinatorial manner;
b) can be prepared in both enantiomerically pure forms;
c) are able to give good levels of diastereoselectivity in their
Ugi reactions. After some experimentation (unreported
work) we found that dihydrobenzoxazepines 1 (Scheme 1)
fulfil all these requirements, in particular giving an unex-
pectedly high (from 5:1 to 9:1) long-range diastereoselectiv-
ity in their Ugi reactions. We have already reported this
finding in a preliminary communication.[16] In this full pa-
per, we now extend these results by reporting: a) a new and
more efficient way to obtain this class of cyclic imines;
Scheme 2. Synthesis of alcohols 2. a) R3MgX, THF, –15 °C;
b) the introduction of diversity inputs in all possible posi-
b) NaBH4, MeOH, –40 °C; c) NaN3, acetone; d) H2, Pd/C, MeOH.
HCl; e) Boc2O, Et3N, CH2Cl2.
tions; c) a thorough study of the relationship between stere-
oselectivity and reaction conditions; d) an assessment of the
relative configuration of the major diastereomer through a
To expand the diversity of our dihydrobenzoxazepine li-
combination of NMR spectroscopic data and computa- brary, and to explore the effect of steric bulk on the stereo-
tional data; e) a thorough conformational study of the selectivity, we also prepared protected amino alcohols 2c
products; f) the enantioselective synthesis of these cyclic (R3 = iPr) and 2d (R3 = tBu). Despite the claims of a pre-
imines using a biocatalytic kinetic resolution.
vious report that the reaction of 10 with iPrMgCl failed to
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5065

L. Banfi, A. Bagno, A. Basso, C. De Santis, R. Riva, F. Rastrelli
FULL PAPER
give any of the desired ketone,[19b] in our hands, simply by Table 1. Results of intermolecular Mitsunobu reactions to give
ethers 9.
performing the Grignard addition at a lower temperature
(–15 °C), the procedure used for the synthesis of 2b also
worked for 2c, albeit in a lower yield. However, with a tert-
butyl Grignard reagent, none of the ketone product was
obtained by this method. Therefore, we chose a completely
different route for the synthesis of 2d, starting from 1-
bromopinacolone (11).[20] Substitution with azide followed
by hydrogenation and Boc protection gave ketone 12, which
was reduced to give alcohol 2d.
Before starting the preparation of a series of differently
substituted imines 5, we reconsidered our synthetic strategy.
Route A was satisfactory, but it involved four steps from
salicylic acid. While the last two steps took place in very
high yield, the Mitsunobu reaction gave products 3 in only
40–60% yield. This step was further complicated by the
troublesome separation of coupling product 3 from unre-
acted starting hydroxamate 1. The conversion of salicylic
acid into hydroxamate 1 proceeded in only moderate yield.
Finally, substituted salicylic acids are not so easily synthe-
sized in the laboratory (Kolbe carboxylation requires high
pressures and an autoclave reactor), whereas the corre-
sponding salicylaldehydes (i.e., 7) can be more easily pre-
pared in a huge variety from phenols 6 by Skattebøl for-
mylation.[21] This was an important consideration, since we
wanted to fully explore the structural diversity arising from
introducing substituents R1 and R2 onto the ring of the
salicylic derivative.
Entry
7
R1
7b 3-Cl
7a
R2
2
R3
iBu 9c
iPr 9d
iBu 9e
iBu 9f
tBu 9g
9
Cond.[a] Yield [%][b]
1
2
3
4
5
H
H
H
2b
2c
2b
A
82
60
43
40
0
H
B
7c 5-Br
7d 3-Cl
7a
A
6-Me 2b
2d
A
H
H
A,B
[a] Conditions: A: 1) (MeO)3CH, MeOH, Amberlyst 15, room
temp., overnight; 2) 2 (1.15 equiv.), THF, PPh3/TBAD (di-tert-butyl
azodicarboxylate) (added in three portions), 0 °CǞr.t., overnight.
B: 1) (MeO)3CH, MeOH, Amberlyst 15, room temp., overnight;
2) 2 (1.5 equiv.), THF, PPh3/DIAD (diisopropyl azodicarboxylate),
0 °CǞr.t., overnight. [b] Isolated yields from 7 after chromatog-
raphy.
group in precursor 8b is more sterically encumbered than
that in 8c. Maybe the higher bulk can suppress an un-
wanted side-reaction leading to an unreactive phenoxytri-
phenylphosphonium derivative. The reaction in Table 1, en-
try 1 went to completion, but a substantial amount of unre-
acted phenol 7 was observed in the crude mixture in the
other reactions. Bulky tert-butyl-substituted amino alcohol
2d failed to give the desired product (i.e., 9g).
Aryl ethers 9c–9f were smoothly converted into the cor-
responding dihydrobenzoxazepines (i.e., 5c–5f) by acidic hy-
drolysis of both the Boc group and the acetal. Also here,
the crude cyclic imines were used directly in the Ugi reac-
Therefore, to prepare dihydrobenzoxazepines 5c–5f, we
decided to use route B, starting from a series of salicylalde-
hydes 7, which were either commercially available or pre-
pared by Skattebøl formylation of phenols 6. This new syn-
thetic strategy was in part inspired by a synthesis of dihy-
drobenzoxazepines reported by Del Buttero et al., where the
aldehyde functionality was protected as an acetal.[22] This
protection can be introduced and cleaved in quantitative
yields in an operationally simple manner. It is important to
note that protection of the aldehyde function is necessary,
since attempted Mitsunobu reactions starting from salicyl-
aldehydes failed to give the desired phenyl ethers.
Table 2. Effect of experimental variables on the diastereoselectivity
of the Ugi reaction of 5c (R1 = 3-Cl, R2 = H, R3 = iBu) to give 13j–
14j (R1 = 9-Cl, R2 = H, R3 = iBu, R4 = tBu, R5 = 5-Cl-thienyl).[a]
Salicylaldehydes 7 were first protected as dimethyl acetals
8a and 8d by reaction with trimethyl orthoformate. These
acetals were found to be unstable during silica gel
chromatography, and thus were used directly after solvent
evaporation, in the reaction with protected amino alcohols
2 under Mitsunobu conditions to give phenyl ethers 9c–
9f, which were obtained in pure form by chromatography.
Interestingly, the acetal moieties in 9c–9f were completely
stable, which suggests that the phenol group may have a
role in the easy hydrolysis of the salicylaldehyde acetals. The
lower polarity of both 7 and 8 than 9 meant that separation
of the products from any unreacted phenols was quite easy.
The results of the Mitsunobu reactions are reported in
Table 1.
Entry Solvent
Conc. [m] Temp. [°C] Yield [%][b] 13:14 ratio[c]
1
2
3
4
MeOH
MeOH
MeOH
EtOH
0.17
0.017
1
25
25
25
25
25
25
25
25
25
25
25
25
–10
–40
76
66
40
78
58
42
40
43
54
48
4
87:13
85:15
85:15
83:17
80:20
72:28
85:15
83:17
78:22
83:17
79:21
72:28
88:12
87:13
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
5
iPrOH
6
7
8
9
CF3CH2OH
CH3CN
CH2Cl2
THF
10
11
12
13
14
toluene
H2O
CH2Cl2/ZnI2
MeOH
MeOH
9
66
34
The yields ranged from excellent (for 9c) to moderate (in
all the other cases). The higher yield obtained for 9c com-
pared to the other cases is surprising, because the phenol
[a] R1 = 9-Cl, R2 = H, R3 = Bu, R4 = tBu, R5 = 5-Cl-thienyl.
[b] Isolated yields. All reactions were carried out for 48 h with the
exception of the reaction of entry 9 (7 d). [c] Determined by NMR
spectroscopy.
5066
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
tions. It is worth noting that an intermediate purification CH2Cl2 and toluene, the reaction still proceeded at a rea-
was needed only at the stage of the ethers (i.e., 6). Thus, sonable rate, but in THF, the reaction was quite slow and
complex tetrahydrobenzoxazepines 13–14, decorated with required several days. The effect of concentration was
five diversity points, could be accessed in just two steps nearly insignificant (Table 2, entries 1–3), and in any case it
from salicylaldehydes 7. Despite the moderate yield ob- was not linear. The most striking result was the effect of
tained in the Mitsunobu step, this protocol is therefore temperature (Table 2, entries 1, 13, and 14). Apart from a
quite efficient from the point of view of step economy.
reduction in yield due to the lower reaction rate, the stereo-
Having synthesized a series of dihydrobenzoxazepines selectivity seemed completely unaffected by the tempera-
5a–5f decorated with diversity points in both the benzene ture! Exactly the same outcome was observed in the Ugi
and heterocyclic rings, we submitted them to a series of Ugi reaction of dihydrobenzoxazepine 5d (R3 = iPr) with the
reactions with isocyanides and carboxylic acids to give ad- same two components (tert-butyl isocyanide and 5-chloro-
ducts 13–14 (Table 3). As we reported in our preliminary thiophenecarboxylic acid; Table 3, entry 11). This suggests
communication,[16] these Ugi reactions were characterized an entropically driven diastereoselectivity, with ΔΔH‡
≈
by remarkable long-range diastereoselectivities. To optimize 0.[23] A similar result was recently observed by Orru and co-
the diastereoselectivity, and to get some insight into this workers in another Ugi–Joullié reaction on a chiral pyrro-
surprising behaviour, we performed a series of reactions on line.[8]
benzoxazepine 5c using tert-butyl isocyanide and 5-chloro-
thiophenecarboxylic acid as the other two components. The tion was established preparing a series of tetrahydrobenz-
results are shown in Table 2. oxazepines 13–14. The results are shown in Table 3. As can
Having found the best conditions, the scope of the reac-
The best solvent was found to be methanol. When the be seen, for all of the combinations tested, the diastereo-
polarity of the solvent was increased, the diastereoselec- selectivity remained good, ranging from 84:16 to 90:10 with
tivity dropped (compare Table 2, entries 1 and 6). However, one notable exception, represented by Table 3, entry 14. The
less polar alcoholic solvents, such as ethanol or 2-propanol presence of a peri substituent at C-7 in the dihydrobenz-
(Table 2, entries 4 and 5), also led to a slight decrease in the oxazepinone led to a remarkable decrease in the stereoselec-
dr. In aprotic solvents, the reaction rate decreased. In tivity. Apart from this case we can conclude that the nature
Table 3. Results of Ugi–Joullié reactions on dihydrobenzoxazepines 5.
Entry
5
13–14
R1
R2
R3
R4
R5
Yield [%][a]
dr[b]
1
2
3
4
5
6
7
8
a[c]
a[c]
a[c]
b[c]
b[c]
b[c]
b[c]
b[c]
b[c]
c[d]
d[d]
d[d]
e[d]
f[d]
a
b
c
d
e
f
g
h
i
j
k
l
m
n
H
H
H
H
H
H
H
H
H
9-Cl
H
H
7-Br
9-Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
6-Me
Me
Me
Me
iBu
iBu
iBu
iBu
iBu
iBu
iBu
iPr
cyHex
tBu
Bn
cyHex
tBu
tBu
Et
70
77
71
59
64
78
57
70
56
76
56
53
56
55
85:15
87:13
84:16
86:14
88:12
89:11
88:12
88:12
90:10
87:13
89:11
90:10
84:16
63:37
MeOCH2
BocNHCH2
Et
PhCH2
5-Cl-thienyl
3-BrC6H4
ZNHCH2CH2
nBu
nBu
9
4-BnOC6H4CH2CH2 MeOCH2
10
11
12
13
14
tBu
tBu
nBu
tBu
tBu
5-Cl-thienyl
5-Cl-thienyl
Ph
5-Cl-thienyl
5-Cl-thienyl
iPr
iBu
iBu
[a] Isolated yields from aldehydes 4 or acetals 9 after chromatography. Reactions were carried out at 25 °C in MeOH at a concentration
of 0.17 m. [b] The 13/14 ratio was determined by NMR spectroscopy or HPLC. [c] Obtained from Route A. [d] Obtained from Route B.
Z = benzyloxycarbonyl.
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5067

L. Banfi, A. Bagno, A. Basso, C. De Santis, R. Riva, F. Rastrelli
FULL PAPER
of the isocyanide and of the carboxylic acid, the presence fered from those calculated for 14j. In the first case, we saw
of a substituent at C-10, and even the nature of the R3 a typical half-chair conformation (Figure 1), whereas in the
substituent have very little effect on the diastereoselectivity, trans isomer, a twist-like conformation was preferred. In
making this long-range asymmetric induction quite general. both cases, the substituent at C-5 (i.e., the aminocarbonyl
It should be stressed that dr values ranging from 5:1 to 9:1 group derived from the isocyanide) has a strong tendency
should be considered excellent, bearing in mind that this is to stay in an axial or pseudo-axial position to avoid peri
a 1,4 long-range asymmetric induction, that good dr values interactions with 6-H. This preferred arrangement is also
are obtained even with R3 substituents of low steric bulk demonstrated by the strong n.O.e. measured between 5-H
(e.g., Me), and that isocyanide-based multicomponent reac- and 6-H in both diastereoisomers for some of compounds
tions are known to be typically less stereoselective than 13 and 14. When the substituent at C-5 is placed in an axial
other nucleophilic additions,[6d] most probably because of position, in 13j, adoption of a half-chair allows the isobutyl
the linear structure of the isonitriles.
group to be positioned in a favourable equatorial position.
In most instances (with the exception of compounds 13k However, in 14j, in the same half-chair conformation, the
and 14k), we were unable to separate the two diastereoiso- isobutyl group would necessarily go into an axial position,
mers by chromatography. However, the major diastereoiso- and this may drive the change of the conformation to a
mer 13 could be obtained in pure form by crystallization; twisted one.
this was demonstrated for 13j and 13m.
1H and 13C chemical shifts of the lowest-energy conform-
The relative configuration of the major adduct 13 was ers were then calculated,[24] along with the proton–proton
proved by comparison of a series of experimental NMR coupling constants of the spin systems corresponding to the
spectra with the same spectra as predicted by DFT meth- seven-membered ring and the isobutyl group. The values
ods. The calculations were performed using adducts 13j and obtained were then averaged for each of the cis and trans
14j (hereafter referred to as cis and trans, respectively).
isomers according to their statistical weights, and then the
A conformational analysis provided 19 conformers for subspectra of the spin systems were simulated. Figure 2
the cis isomer and 15 conformers for the trans isomer. All shows the experimental and theoretical subspectra of cis
of the output geometries were then reoptimized (see com- isomer 13j. In particular, the subspectra show (from left to
putational details). For each isomer, the new geometries right): the signal of 5-H (singlet), the signal of equatorial 3-
were sorted by increasing energy, and the respective par- H, the signal of 2-H (a complex multiplet coupling with 3-
tition functions were built. In a Maxwell–Boltzmann analy- H and with the isobutyl CH2), and finally the signal of axial
sis at 298 K, the two lowest-energy conformers were found 3-H (in CDCl3, the NH singlet at δ = 5.15 ppm is also vis-
to account for 95% of the overall population in the cis iso- ible). Note that sharp singlets due to residual CH2Cl2 (at δ
mer. Similarly, the four lowest-energy conformers (two of = 5.27 and 5.68 ppm) as well as the water signal (only in
which were isoenergetic within the accuracy of the calcula- [D6]DMSO) are also present in the experimental spectra.
tions) accounted for 86% of the overall population in the
trans isomer (Figure 1).
The calculated spectrum of trans isomer 14j is given in
the Supporting Information. It appears to be quite different
from the experimental spectrum of cis isomer 13j, and sim-
ilar to the experimental spectrum of trans isomer 14j. Note
that while a sample of diastereomerically pure cis isomer
13j could be obtained by crystallization, the experimental
spectrum of trans isomer 14j could only be deduced from
the spectrum of an enriched mixture, since separation by
chromatography was not possible.
The calculated NMR data were generally found to be in
agreement with the experimental data. Particularly for cis
isomer 13j, the predicted and experimental J2,3 values were
in good agreement, along with the hyperfine patterns of the
relevant signals. Moreover, in 13j, the axial 3-H proton,
which experiences a higher J2,3, appears upfield in both the
experimental and predicted spectra, whereas the opposite is
true for trans compound 14j. Finally, in trans isomer 14j,
the chemical shift of 2-H was shifted downfield by 0.85–
1.3 ppm in both the experimental and predicted spectra.
On the other hand, the differences between the 13C chem-
ical shifts were so small that the spectra of 13j and 14j are
too similar for a meaningful comparison. These results are
reported in the Supporting Information only.
Figure 1. Top: overlap of the two lowest-energy conformers in 13j
(cis; left, ΔE = 0.23 kcalmol–1) and in 14j (trans; right, ΔE =
0.74 kcalmol–1). Bottom: the two lowest-energy conformers for 13j
(cis) and 14j (trans) viewed from another perspective: note the half-
chair and twist conformations of the seven-membered ring. Hydro-
gen atoms are omitted for clarity. Note that the conformation of
the seven-membered ring is preserved in each of the two isomers.
The relative configurations of the other adducts 13 was
Interestingly, the calculations showed that the lowest en- then established based on clear patterns in the NMR spec-
ergy conformations of the benzoxazepine moiety in 13j dif- tra. For example: a) in all cases, the chemical shift of 2-H
5068
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
Figure 2. Experimental (top, in CDCl3; middle, in [D6]DMSO) and simulated (bottom) 1H subspectra of cis isomer 13j. See text for
details.
was much higher in the trans isomer than in the cis isomer;
Recently, a computational study[25] has suggested the
b) in all cases, in cis compounds, 3-H trans to R3 (equato- possibility that the nitrilium ion may undergo a fast reac-
rial) appeared further upfield than 3-H cis to R3 (axial), tion with the carboxylate to give the α-addition product di-
whereas the opposite was true for the trans compounds; rectly. In this scenario, the isocyanide addition would not
c) J2,3cis was always lower (ca. 0) in the cis compounds than be reversible, and the α-addition would take place in a sin-
in the trans compounds (ca. 3.0 Hz).
gle step, although the involvement of a reversibly formed
The conformational analysis of the final products (i.e., iminium ion may depend on the solvent.[25] An alternative
13–14), and the fact that the diastereoselectivity was not rationalization based on this mechanistic hypothesis is that
influenced by the temperature (entropically driven stereose- trans isomer 14 is formed by a mechanism involving a con-
lectivity), led us to propose the scenario shown in Scheme 3. formational rearrangement to 19 at the cyclic imine stage.
Starting cyclic imine 5 is expected to prefer a half-chair con- In this way, lower-face attack would give a twist conforma-
formation with the R2 substituent in an equatorial position. tion directly, with the C-5 substituent in an axial position.
Reversible attack of the isocyanide can then take place on Also in this case, the route leading to 14 should be entropi-
either face, since its linear structure makes the isocyanide cally disfavoured.
insensitive to steric bulk, and also because the R2 substitu-
The compounds listed in Table 3 are racemic. However,
ent is far away. Attack from the upper face leads to axial in principle, it should be possible to obtain the same com-
iminium ion 15, whereas attack from the lower face leads to pounds in enantiomerically pure form, provided that the
equatorial iminium ion 17. However, the α-addition greatly starting Boc amino alcohols are available. Compound 2a
increases the steric bulk of the substituent at C-5, which (R3 = Me) can easily be obtained from commercially avail-
prefers an axial arrangement in order to avoid peri interac- able (R)- or (S)-1-amino-2-propanols. The asymmetric syn-
tions. In 15, this is not a problem, because the nitrilium ion thesis of both enantiomers of compound 2c (R3 = iPr) has
is already in an axial position, and thus 16 is readily (and been reported previously.[26] Therefore, we concentrated on
irreversibly) formed. Finally, OǞN acyl migration gives the an efficient method for the preparation of alcohol 2b (R3 =
final cis adduct (i.e., 13). In 17, in contrast, the α-addition iBu). In particular, we decided to explore an enzymatic ki-
requires a prior conformational rearrangement to take netic resolution (Scheme 4). However, alcohol 2b turned out
place to give twist conformation 18, in order to place the to be rather reluctant to undergo lipase-catalysed acetyl-
nitrilium ion in an axial position. From an enthalpic point ation. We tested several commercially available lipases, but
of view, both processes are feasible, but the need for a con- the only one that gave a good conversion in a reasonable
formational rearrangement makes the route that leads to 14 time was Novozym (Candida antarctica lipase). The selectiv-
less entropically favourable.
ity turned out to be excellent (E = 215). Therefore, with a
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5069

L. Banfi, A. Bagno, A. Basso, C. De Santis, R. Riva, F. Rastrelli
FULL PAPER
95%. Since (S)-2b can be obtained in quantitative yield by
saponification of (S)-20, both enantiomers are accessible
with high ee. The absolute configuration was determined by
Mosher’s method[27] (see Exp. Sect.). The outcome is in per-
fect agreement with Kazlauskas’ model for the prediction
of enantioselectivity in lipase-catalysed acylation reactions
of secondary alcohols.[28]
Alcohol (R)-2b was then converted in two steps into
tetrahydrobenzoxazepine (S)-13m following the same
method used for the racemic compound. The Mitsunobu
reaction was shown to be completely stereospecific by
HPLC analysis of compound (S)-9e on a chiral stationary
phase (using racemic 9e as standard), which showed an ee
of 99%. The absolute configuration of (S)-9e was presumed
on the reasonable assumption that the Mitsunobu reaction
proceeds with complete inversion of configuration.
Conclusions
In conclusion, we have reported a step-economical, di-
versity-oriented, and stereoselective method for the synthe-
sis of tetrahydrobenzo[f][1,4]oxazepines 13. The value of
these compounds for medicinal chemistry is demonstrated
by several reports concerning the pharmacological activity
of this class of heterocycle.[29] The method involves just two
steps from readily available salicylaldehydes, and allows the
introduction of four diversity inputs. The diastereoselectiv-
ity is good, and pure diastereomers may be obtained by
a single crystallization. Moreover, the overall procedure is
stereospecific, and therefore also enantiomerically pure fi-
nal products can be obtained starting from enantiomer-
ically pure Boc-protected amino alcohols. One such product
has been efficiently prepared using an enzyme-catalysed ki-
netic resolution.
Scheme 3. Rationalization of the observed diastereoselectivity.
reaction time of 44 h (41.2% conversion), we could isolate
(S) acetate 20 with 98.1% ee, whereas with a reaction time
of 192 h (52.6% conversion), unreacted alcohol (R)-2b was
obtained with 99.5% ee. In both cases, the overall yield was
Experimental Section
General Remarks: NMR spectra were recorded at room temp. in
CDCl3 or in [D6]DMSO at 300 MHz (1H) and 75 MHz (13C),
using, as internal standard, TMS (tetramethylsilane; 1H NMR in
CDCl3; 0.000 ppm) or the central peak of DMSO (1H NMR in [D6]-
DMSO; 2.506 ppm) or the central peak of CDCl3 (13C in CDCl3;
77.02 ppm), or the central peak of DMSO (13C in [D6]DMSO;
39.43 ppm). Chemical shifts are reported in ppm on the δ scale.
Peak assignments were made with the aid of gCOSY and gHSQC
experiments. In ABX systems, proton A is considered to be upfield
and proton B to be downfield. GC–MS was carried out using an
HP-1 column (12 m long, 0.2 mm wide), electron impact at 70 eV,
and a mass temperature of about 170 °C. Only ions with m/z Ͼ 33
were detected. All analyses were performed (unless otherwise
stated) with a constant He flow of 0.9 mLmin–1 with initial temp.
of 100 °C, initiation time 2 min, rate 20 °Cmin–1, final temp.
280 °C, injection temp. 250 °C, detection temp. 280 °C. Accurate
masses were recorded on a Synapt G2 qTOF instrument using Leu-
cine Enkephalin (2 ngmL–1) as reference mass. Samples were dis-
solved at 5 mm concentration in 50% acetonitrile. TLC analyses
were carried out on silica gel plates and visualized with UV light
(254 nm) or with ninhydrin [ninhydrin (900 mg) in nBuOH
Scheme 4. Enantioselective synthesis of tetrahydrobenzoxazepine
13m. a) Candida antarctica Lipase, vinyl acetate, room temp., 44–
192 h. b) KOH, MeOH, room temp. c) see Table 1. d) see Table 3.
5070
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
(300 mL) and AcOH (9 mL), followed by warming]. Rf values were
measured after an elution of 7–9 cm. Column chromatography was
carried out using the “flash” method with 220–400 mesh silica. IR
spectra were recorded as CHCl3 solutions. For optical rotations,
the units are: degcm3 g–1 dm–1 for [α] and 100ϫgcm–3 for concen-
tration (c). Petroleum ether (40–60 °C) is abbreviated as PE. In ex-
tractive work-up, aqueous solutions were always reextracted three
times with the appropriate organic solvent. Organic extracts were
always dried with Na2SO4 and filtered, before evaporation of the
solvent under reduced pressure. All reactions using dry solvents
were carried out under a nitrogen atmosphere.
THF (30 mL) and cooled to –15 °C. Isopropylmagnesium chloride
(2 m in THF; 18.0 mL, 36.0 mmol) was then slowly added over
15 min using a dropping funnel. The resulting solution was stirred
for 2 h at room temp. NH4H2PO4 (5% aq.; 15 mL) was then slowly
added using a dropping funnel. The temperature was allowed to
rise to room temp. The mixture was poured into NH4H2PO4 (5%
aq.; 35 mL), NH4Cl (saturated aq.; 50 mL), and HCl (37% aq.;
1 mL), and the resulting mixture was extracted with Et2O. Evapora-
tion gave a crude product that was purified by chromatography
(PE/EtOAc, 70:30) to give a colourless oil (2.020 g). This oil was
dissolved in dry MeOH (15 mL), cooled to –40 °C, and NaBH4
(857 mg, 22.65 mmol) was added in portions. After 90 min at
–40 °C, the reaction mixture was treated with NH4Cl (saturated
aq.; 10 mL), and most of the methanol was evaporated without
warming. The residue was diluted with NH4H2PO4 (5% aq.), and
this mixture was treated with HCl (37% aq.), and extracted with
Et2O. Evaporation and chromatography (PE/EtOAc, 80:20 + 0.5%
MeOH) gave pure 2c (1.67 g, 53% from 10) as an oil. Rf = 0.28
(PE/EtOAc, 80:20). HRMS (ESI+): calcd. for C10H21NNaO3 [M +
Na]+ 226.1419; found 226.1418. The other spectroscopic data (1H
and 13C NMR, IR) were identical to those previously reported for
the enantiomerically pure (R) compound.[26a,26b]
N-Methyl-N-methoxy-2-hydroxybenzamide (1): A solution of sal-
icylic acid (1.688 g, 12.22 mmol) in dry tetrahydrofuran (THF;
25 mL) was cooled to 0 °C and treated with carbonyl diimidazole
(2.38 g, 14.67 mmol) in four portions. The mixture was stirred at
0 °C for 2 h. Then N,O-dimethylhydroxylamine hydrochloride
(1.43 g, 14.67 mmol) was added, followed by triethylamine
(2.21 mL, 15.89 mmol). After stirring for 6 h at room temp., the
reaction mixture was poured into (NH4)H2PO4 (5% aq.; 70 mL)
containing HCl (1 m; 23 mL). The resulting pH was 6. Extraction
with Et2O (1ϫ) and EtOAc (3ϫ), and evaporation of the organic
extracts gave an oil, which was purified by chromatography (PE/
EtOAc, 70:30) to give pure 1 (1.400 g, 63%) as an oil. The spectro-
scopic and analytical data were identical to those already re-
ported.[30]
(؎)-tert-Butyl (2-Hydroxy-3,3-dimethylbutyl)carbamate (2d): So-
dium azide (1.085 g, 16.7 mmol) was suspended in dry acetone
(30 mL), and this mixture was treated with 1-bromo-3,3-dimeth-
ylbutanone (1-bromopinacolone; 1.73 mL, 12.85 mmol). The sus-
pension was stirred for 4 h at room temp. and then filtered. The
filtrate was evaporated to dryness, and the residue was dissolved in
MeOH (35 mL). HCl (37% aq.; 1.07 mL) and Pd/C (10%; 300 mg)
were added, and the mixture was hydrogenated at room temp. un-
der the slight overpressure of an inflated balloon for 18 h. The cata-
lyst was removed by filtration, and the filtrate was evaporated to
dryness. The residue was dissolved in Et2O and iPrOH and evapo-
rated again to azeotropically remove residual water. The resulting
solid was triturated with Et2O to give pure 1-amino-3,3-dimeth-
ylbutanone hydrochloride (1.79 g, 93%), m.p. 205–206 °C (ref.[20]
205 °C). The spectroscopic data were identical to those reported.
This hydrochloride was suspended in dry CH2Cl2 (30 mL) and
treated with Et3N (4.5 mL, 32.11 mmol) and di-tert-butyl dicarbon-
ate (3.36 g, 15.4 mmol). The suspension was stirred at room temp.
for 18 h, then the mixture was diluted with Et2O and washed first
with NaHCO3 (saturated aq.) and then with NH4H2PO4 (5% aq.).
The organic phase was evaporated to dryness and the residue was
purified by chromatography (PE/EtOAc, 90:10Ǟ75:25) to give
ketone 12 (1.90 g, 74%; 69% from 1-bromopinacolone) as an oil.
Ketone 12 was dissolved in dry MeOH (20 mL). The solution was
cooled to –40 °C, and NaBH4 (1.12 g, 29.5 mmol) was added in
portions (caution: gas evolution was observed). After 90 min at
–40 °C, the reaction mixture was treated with NH4Cl (saturated
aq.; 10 mL), and most of the methanol was evaporated without
warming. The residue was diluted with NH4H2PO4 (5% aq.), and
the resulting mixture was treated with HCl (37% aq.), and ex-
tracted with Et2O. Evaporation and chromatography (PE/EtOAc,
80:20 + 0.5% MeOH) gave pure 2d (1.63 g, 85%) as a white solid,
m.p. 65–66 °C. Rf = 0.35 (PE/EtOAc, 70:30). The spectroscopic
data (1H and 13C NMR, IR) were identical to those previously
reported for the enantiomerically pure (R) compound.[26a,26b]
(R)-tert-Butyl (2-Hydroxy-4-methylpentyl)carbamate (R)-(2b): A
solution of (Ϯ)-2b[15a] (820 mg, 3.77 mmol) in diisopropyl ether
(49 mL), was treated with vinyl acetate (25 mL) and Novozym 435
(lipase from Candida antarctica; 2.136 g). The suspension was
shaken in an orbital shaker at 25 °C for 192 h. Then it was filtered
and evaporated to dryness, and the residue was purified by
chromatography to give pure acetate (S)-20 (486.8 mg, 49.8%) and
pure unreacted (R)-2b (367.3 mg, 44.8%). The enantiomeric excess
of 2b was determined by its conversion into the Mosher ester with
the (R) acyl chloride (see Supporting Information), and integration
1
of the OCH3 signal in the H NMR spectrum [which resonates at
3.57 ppm for the ester derived from (S)-2b and at 3.53 ppm for the
other diastereomer]. The ee was shown to be Ͼ99.5%. Acetate (S)-
20 was converted into (S)-2b by saponification (see below). Alcohol
(S)-2b was also converted into its Mosher ester using the (R) acyl
1
chloride. H NMR analysis indicated an ee of 88.6%. From these
values we could calculate the E factor to be Ͼ97. Data for (R)-2b:
[α]D = –4.5 (c = 2, CHCl3). The other spectroscopic data were iden-
tical with those of the racemic compound.[15a]
(S)-tert-Butyl (2-Hydroxy-4-methylpentyl)carbamate (S)-(2b): The
enzymatic kinetic resolution was carried out on racemic 2b
(500 mg) exactly as described above, but with a reaction time of
44 h. Work-up and chromatography gave pure acetate (S)-20
(243 mg, 40.7%) and pure unreacted (R)-2b (257 mg, 51.4%). Acet-
ate (S)-20 was converted into (S)-2b as follows: it was dissolved in
MeOH (10 mL) and treated with NaOH (2 m aq.; 0.90 mL) at room
temp for 2 h. Then the solution was poured into NH4Cl (saturated
aq.) and extracted with Et2O to give, after evaporation and
chromatography, pure (S)-2b in quantitative yield. Both the (S)-2b
obtained by saponification of (S)-20 and the recovered (R)-2b were
1
converted into their Mosher esters using the (R) acyl chloride. H
NMR analysis indicated an ee of 98.1% for the (R)-2b derived from
acetate (S)-20, and an ee of 68.7% for the recovered (S)-2b. From
these data, the E factor was calculated to be 215. Data for (S)-2b:
[α]D = +4.5 (c = 2, CHCl3).
(؎)-N-Methyl-N-methoxy-2-{[1-(tert-butoxycarbonylamino)prop-2-
yl]oxy}benzamide (3a): A solution of phenol 1 (507.0 mg,
2.797 mmol) and (Ϯ)-1-(tert-butoxycarbonylamino)propan-2-ol
(2a)[17] (637.5 mg, 3.6375 mmol, 1.3 equiv.) in dry THF (10 mL)
was cooled to –15 °C, and treated with triphenylphosphane
(1.100 g, 4.195 mmol, 1.5 equiv.) and di-tert-butyl azodicarboxylate
(؎)-tert-Butyl (2-Hydroxy-3-methylbutyl)carbamate (2c): Weinreb
hydroxamate 10[15a,18] (3.40 g, 15.6 mmol) was dissolved in dry
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5071

L. Banfi, A. Bagno, A. Basso, C. De Santis, R. Riva, F. Rastrelli
FULL PAPER
(TBAD; 965 mg, 4.195 mmol, 1.5 equiv.). The temperature was al-
lowed to rise slowly to 20 °C over 4 h. The mixture was stirred
overnight at room temp., then the solvents were evaporated, and
the residue was purified by chromatography (PE/EtOAc, 50:50) to
give pure 3a (467.7 mg, 49%) as an oil. Rf = 0.32 (PE/EtOAc, 1:1).
1H NMR (300 MHz, CDCl3, 55 °C): δ = 1.30 [d, 3J(H,H) = 6.3 Hz,
3 H, H3C-CH], 1.40 [s, 9 H, (CH3)3C], 3.15 [ddd, 3J(H,H) = 5.0,
(؎)-2-{[1-(tert-Butoxycarbonylamino)prop-2-yl]oxy}benzaldehyde
(4a): A solution of hydroxamate 3a (1.354 g, 4.00 mmol) in dry
THF (4 mL) and dry Et2O (16 mL) was cooled to 0 °C and treated
with LiAlH4 (218.6 mg, 5.76 mmol). After 2 h at 0 °C, the re-
duction was complete. The mixture was diluted with THF (15 mL)
and slowly treated (caution!) with a solution of KHSO4 (1.725 g,
12.67 mmol) in H2O (8 mL). The biphasic system was stirred for
15 min, then it was treated with Na,K tartrate (25%; 22 mL) and
2
7.2, J(H,H) = 14.2 Hz, 1 H, CHHNH], 3.30 (s, 3 H, NCH3), 3.45
3
2
[ddd, J(H,H) = 3.0, 7.2, J(H,H) = 14.2 Hz, 1 H, CHHNH], 3.55 stirred for a further 1.5 h. Extraction with THF/Et2O (1:1) and
3
(s, 3 H, OCH3), 4.50 [dquint, J(H,H) = 3.0 (d), 6.4 (quint) Hz, 1
concentration of the organic phase gave a crude product that was
H, CH-O], 5.51 (br. s, 1 H NH), 6.93 [d, J(H,H) = 8.1 Hz, 1 H, purified either by chromatography (PE/EtOAc, 80:20) (90% yield)
6-H], 6.96 [t, 2J(H,H) = 7.5 Hz, 1 H, 4-H], 7.26 [dd, 3J(H,H) = 7.5,
or trituration (EtOAc/PE) (888.2 mg, 83%). Rf = 0.60 (PE/EtOAc,
3
4J(H,H) = 1.5 Hz, 1 H, 3-H], 7.32 [td, J(H,H) = 7.8 (t), J(H,H)
70:30). H NMR (300 MHz, CDCl3, 55 °C): δ = 1.35 [d, J(H,H)
3
4
1
3
= 1.5 (d) Hz, 1 H, 5-H] ppm. 13C NMR (75 MHz, CDCl3, 55 °C): = 6.0 Hz, 3 H, CH3CH], 1.44 [s, 9 H, (CH3)3C], 3.36 [ddd, 3J(H,H)
2
3
δ = 17.5 (CH3CH), 28.4 [(CH3)3C], 33.3 (br., CH3N), 45.8 = 5.7, 6.9, J(H,H) = 14.4 Hz, 1 H, CHHNH], 3.48 [ddd, J(H,H)
(CH2NH), 61.1 (CH3O), 75.0 (CH-O), 79.0 [C(CH3)3], 114.0, 120.8, = 4.2, 6.9, 2J(H,H) = 14.4 Hz, 1 H, CHHNH], 4.65 [sextuplet,
127.9, 130.6 (aromatic CH), 126.4 (C-2), 154.6, 156.2 (C=O of ure-
thane, C-1) ppm. Note: the signal of the hydroxamate C=O was
very broad and was undetectable. GC–MS: tR = 8.57; m/z (%) =
294 [M – 44]+ (0.3), 278 [M – 60]+ (1.7), 265 [M – 73]+ (7.1), 222
(31.7), 204 (39.0), 178 (80.1), 149 (7.6), 139 (9.8), 133 (10.1), 121
(81.4), 120 (7.2), 105 (6.5), 102 (6.4), 93 (8.7), 92 (91.7), 88 (12.0),
3J(H,H) = 5.8 Hz, 1 H, CH-O], 4.89 (br. s, 1 H, NH), 7.02 [t,
3J(H,H) = 7.5 Hz, 1 H, 5-H], 7.04 [d, 3J(H,H) = 8.7 Hz, 1 H, 3-
H], 7.51 [td, 3J(H,H) = 7.9 (t), 4J(H,H) = 2.0 (d) Hz, 1 H, 4-H],
3
4
7.82 [dd, J(H,H) = 7.9, J(H,H) = 1.7 Hz, 1 H, 6-H], 10.45 (s, 1
H, CH=O) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ = 17.15
(CH3CH), 28.4 [C(CH3)3], 45.6 (CH2NHBoc), 74.1 (CH-O), 79.7
84 (12.2), 65 (10.6), 61 (5.9), 57 (100.0), 56 (16.5), 43 (8.5), 41 [C(CH3)3], 114.1 (C-3), 121.0 (C-5), 125.7 (C-1), 129.0 (C-6), 135.9
(26.5), 39 (8.4). C17H26N2O5 (338.40): calcd. C 60.34, H 7.74, N (C-4), 156.0, 160.2 (C-2, tBuOC=O), 189.8 (CHO) ppm.
8.28; found C 60.5, H 7.8, N 8.1.
C15H21NO4 (279.34): calcd. C 64.50, H 7.58, N 5.01; found C 64.7,
H 7.7, N 4.95.
(؎)-N-Methyl-N-methoxy-2-{[1-(tert-butoxycarbonylamino)-4-meth-
ylpent-2-yl]oxy}benzamide (3b): A solution of phenol 1 (659 mg,
3.64 mmol) and (Ϯ)-1-(tert-butoxycarbonylamino)-4-methyl-
pentan-2-ol (2b)[15a] (988.5 mg, 4.55 mmol) in dry THF (10 mL)
was cooled to 0 °C, and triphenylphosphane (602 mg, 2.30 mmol)
and TBAD (529 mg, 2.30 mmol) were added. The mixture was
stirred at 0 °C and, after 100 min, it was treated again with tri-
phenylphosphane (602 mg, 2.30 mmol) and TBAD (529 mg,
2.30 mmol). After a further 100 min, a third addition of tri-
phenylphosphane (602 mg, 2.30 mmol) and TBAD (529 mg,
2.30 mmol) was made. After stirring for a further 3.5 h, the solvents
were evaporated, and the residue was purified by chromatography
(PE/EtOAc, 75:25Ǟ60:40) to give pure 3b (857 mg, 62%) as an
oil. Rf = 0.36 (PE/EtOAc, 65:35). 1H NMR (300 MHz, CDCl3,
55 °C): δ = 0.94 [d, 3J(H,H) = 6.3 Hz, 3 H, CH3CH], 0.96 [d,
3J(H,H) = 6.6 Hz, 3 H, CH3CH], 1.37 [s, 9 H, (CH3)3CO], 1.43
3-Chloro-2-hydroxy-6-methylbenzaldehyde (7d): Compound 7d was
prepared from commercially available 2-chloro-5-methylphenol
using the Skattebøl method.[21a,31] It was obtained in pure form by
chromatography as a yellow solid (38% yield). This compound was
already reported.[32] M.p. 50–51 °C. Rf = 0.50 (PE/acetone, 8:2). 1H
NMR (300 MHz, CDCl3, 20 °C): δ = 2.61 (s, 3 H, CH3), 6.70 [d,
3J(H,H) = 8.1 Hz, 1 H, 5-H], 7.47 [d, 3J(H,H) = 8.1 Hz, 1 H, 4-
H], 10.30 (s, 1 H, OH), 12.42 (s, 1 H, CH=O) ppm. 13C NMR
(75 MHz, CDCl3, 20 °C): δ = 17.9 (CH3), 119.1, 120.1 (C-1, C-3),
140.8 (C-6), 158.6 (C-2), 122.1 (C-5), 137.1 (C-4), 195.2 (CHO)
ppm. IR: ν = 3017, 1640, 1402, 1312, 1282, 1133 cm–1. GC–MS:
˜
tR = 5.45; m/z (%) = 170 [M]+ (91.5), 169 (100.0), 152 (12.5), 141
(6.2), 124 (18.6), 107 (6.3), 105 (5.6), 89 (10.5), 77 (37.8), 53 (9.3),
51 (18.6), 50 (6.1), 39 (7.8). HRMS (ESI–): calcd. for C8H6ClO2
[M – H]– 169.0056; found 169.0052.
3
2
[ddd, J(H,H) = 6.0, 6.9, J(H,H) = 14.0 Hz, 1 H, CHH-iPr], 1.62
(؎)-tert-Butyl 2-[2-Chloro-6-(dimethoxymethyl)phenoxy]-4-methyl-
pentylcarbamate (9c): A solution of 3-chlorosalicylaldehyde[21b]
3
3
[dt, J(H,H) = 6.9, J(H,H) = 14.0 Hz, 1 H, CHH-iPr], 1.79 [sep-
tuplet, 3J(H,H) = 6.7 Hz, 1 H, CH(CH3)2], 3.14 [dt, 3J(H,H) = 6.0, (1.50 g, 7.40 mmol) in dry MeOH (8 mL) was treated with tri-
2J(H,H) = 14.1 Hz, 1 H, CHHNHBoc], 3.29 (s, 3 H, CH3N), 3.44
methyl orthoformate (3.5 mL) and Amberlyst 15 resin (100 mg),
and the resulting mixture was stirred at room temp. for 20 h. The
mixture was diluted with CH2Cl2 and filtered. The filtrate was
treated with solid NaHCO3 (25 mg) and filtered again. The filtrate
3
2
[ddd, J(H,H) = 3.3, 6.6, J(H,H) = 14.1 Hz, 1 H, CHHNHBoc],
3
3.55 (s, 3 H, CH3O), 4.49 [qd, J(H,H) = 3.3 (d), 6.6 (q) Hz, 1 H,
CH-O], 5.58 (s, 1 H, NH), 6.95 [t, 3J(H,H) = 7.5 Hz, 1 H, 3-H],
6.96 [d, 3J(H,H) = 7.8 Hz, 1 H, 5-H], 7.25 [dd, 3J(H,H) = 7.8, was evaporated, the residue was dissolved in CH2Cl2/toluene, and
4J(H,H) = 1.5 Hz, 1 H, 2-H], 7.31 [ddd, 3J(H,H) = 7.5, 9.3, 4J(H,H)
the solution was evaporated again to azeotropically remove all of
= 1.8 Hz, 1 H, 4-H] ppm. 13C NMR (75 MHz, CDCl3, 55 °C): δ = the methanol. The resulting crude oil was dissolved in dry THF
22.9 (CH3CHCH3), 24.8 [C(CH3)2], 28.4 [C(CH3)3], 33.5 (CH3N), (15 mL). This solution was cooled to 0 °C, and racemic alcohol 2b
41.1 (CH2iPr), 44.4 (CH2NHBoc), 61.1 (CH3O), 77.2 (CH-O), 78.9 (1.83 g, 8.51 mmol), PPh3 (970 mg, 3.70 mmol), and TBAD
[C(CH3)3], 114.0 (C-5), 120.7 (C-3), 126.5 (C-1), 127.8 (C-2), 130.6 (872 mg, 3.70 mmol) were added in that order. The mixture was
(C-4), 155.1, 156.3 (C-6, OC=O) ppm. Note: the signal of hydrox-
stirred at 0 °C, and after 60 min, further PPh3 (970 mg, 3.70 mmol)
and TBAD (872 mg, 3.70 mmol) were added. After a further
60 min, a third addition of PPh3 (970 mg, 3.70 mmol), and TBAD
amate C=O was very broad and was undetectable. IR: ν = 3668,
˜
3595, 3379, 3016, 2960, 2712, 1702, 1637, 1598, 1441, 1366, 1166,
1036, 917, 828 cm–1. GC–MS: tR = 9.30; m/z (%) = 380 [M]+ (0.1), (872 mg, 3.70 mmol) was made. The solution was allowed to reach
320 (0.7), 307 (3.7), 264 (18.9), 246 (10.9), 220 (100.0), 191 (6.5),
182 (5.5), 139 (19.6), 134 (9.5), 130 (11.4), 126 (10.0), 121 (94.7),
room temperature and then it was stirred overnight. It was diluted
with Et2O and washed with NaOH (0.1 m aq.; 40 mL) to remove
100 (5.9), 93 (6.8), 92 (6.1), 84 (6.3), 83 (12.3), 74 (6.2), 65 (7.5), the unreacted phenols. After washing with NH4Cl (saturated aq.),
57 (91.7), 56 (8.2), 43 (7.2), 41 (20.6). HRMS (ESI+): calcd. for the solvents were evaporated, and the residue was purified by
C20H33N2O5 [M + H]+ 381.2389; found 381.2390.
5072
chromatography (PE/EtOAc, 90:10Ǟ80:20) to give pure 9c (2.44 g,
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
82%) as an oil. Rf = 0.60 (PE/EtOAc, 80:20). 1H NMR (300 MHz,
55 (8.8), 44 (14.2), 43 (6.4), 41 (24.3), 39 (10.3). HRMS (ESI+):
3
CDCl3, 20 °C): δ = 0.88 [d, J(H,H) = 6.6 Hz, 3 H, CH3CH], 0.92 calcd. for C19H31NO5Na [M + Na]+ 376.2100; found 376.2103.
3
[d, J(H,H) = 6.3 Hz, 3 H, CH3CH], 1.47 [s, 9 H, (CH3)3C], 1.42–
(؎)-tert-Butyl 2-[4-Bromo-2-(dimethoxymethyl)phenoxy]-3-methyl-
1.57 (m, 2 H, CH2iPr), 1.73 [nonuplet, 3J(H,H) = 6.8 Hz, 1 H,
butylcarbamate (9e): Compound 9e was prepared from commer-
cially available 5-bromosalicylaldehyde (671 mg) following the same
procedure used for 9c. Chromatography (PE/EtOAc, 90:10Ǟ80:20)
CH(CH3)2], 3.28 [dt, 3J(H,H) = 6.2, 2J(H,H) = 14.2 Hz, 1 H,
CHHNHBoc], 3.35 (s, 3 H, OCH3), 3.39 (s, 3 H, OCH3), 3.44 [ddd,
3J(H,H) = 2.8, 5.7, 3J(H,H) = 14.2 Hz, 1 H, CHHNHBoc], 4.79
gave pure 9e (673 mg, 43%) as an oil. Rf = 0.60 (PE/EtOAc, 70:30).
[qd, 3J(H,H) = 6.5 (q), 2.8 (d) Hz, 1 H, CH-O], 5.47 (br. t, 1 H,
1H NMR (300 MHz, CDCl3, 20 °C): δ = 0.95 [d, 3J(H,H) = 6.3 Hz,
3
NH), 5.60 [s, 1 H, CH(OCH3)2], 7.05 [t, J(H,H) = 7.8 Hz, 1 H, 5-
3
3 H, CH3CH], 0.98 [d, J(H,H) = 6.3 Hz, 3 H, CH3CH], 1.40 [s, 9
H], 7.34 [dd, 3J(H,H) = 8.0, 4J(H,H) = 1.6 Hz, 1 H, 4-H], 7.45 [dd,
H, (CH3)3C], 1.38–1.50 (m, 1 H, CHH-iPr), 1.67 [dt, 3J(H,H) = 6.9
3J(H,H) = 7.8, 4J(H,H) = 1.6 Hz, 1 H, 6-H] ppm. 13C NMR
(t), 2J(H,H) = 13.5 (d) Hz, 1 H, CHH-iPr], 1.78 [nonuplet, 3J(H,H)
(75 MHz, CDCl3, 20 °C):
δ = 22.1, 23.6 [(CH3)2C], 24.7
= 6.6 Hz, 1 H, CH(CH3)2], 3.15 [dt, 3J(H,H) = 6.1, 2J(H,H) =
13.8 Hz, 1 H, CHHNHBoc], 3.23 (s, 3 H, OCH3), 3.45 (s, 3 H,
OCH3), 3.50 [ddd, 3J(H,H) = 3.2, 6.7, 2J(H,H) = 13.8 Hz, 1 H,
CHHNHBoc], 4.42 [qd, 3J(H,H) = 6.5 (q), 3.2 (d) Hz, 1 H, CH-O],
5.59 (br. t, 1 H, NH), 5.66 [s, 1 H, CH(OCH3)2], 6.80 [d, 3J(H,H) =
[CH(CH3)2], 28.4 [C(CH3)3], 40.2 (CH2iPr), 43.6 (CH2NH), 53.0,
54.4 (OCH3), 79.0 [C(CH3)3], 80.3 (CH-O), 99.3 [CH(OMe)2],
124.0 (C-5), 126.4 (C-6), 126.9 (C-3), 131.2 (C-4), 133.1 (C-2), 151.1
(C-1), 156.2 (C=O) ppm. IR: ν = 3447, 3016, 2922, 2829, 1705,
˜
1589, 1435, 1359, 1255, 1140, 1087, 986, 905 cm–1. GC–MS: tR
=
4
8.7 Hz, 1 H, 6-H], 7.37 [dd, 3J(H,H) = 8.7, J(H,H) = 2.7 Hz, 1 H,
9.11; m/z (%) = 386 [M – 15]+ (0.1), 369 [M – 32]+ (2.4), 296 (10.9),
282 (5.9), 281 (5.3), 238 (5.8), 172 (39.5), 171 (61.1), 170 (100.0),
155 (18.0), 144 (97.9), 141 (16.8), 100 (25.5), 83 (18.2), 74 (11.3),
59 (8.0), 57 (58.0), 56 (9.3), 55 (10.8), 43 (7.4), 41 (20.5). HRMS
(ESI+): calcd. for C20H32ClNO5Na [M + Na]+ 424.1867; found
424.1869.
5-H], 7.63 [d, 3J(H,H) = 2.7 Hz, 1 H, 3-H] ppm. 13C NMR
(75 MHz, CDCl3, 20 °C):
δ = 22.9, 23.0 [(CH3)2C], 24.7
[CH(CH3)2], 28.4 [C(CH3)3], 41.1 (CH2iPr), 43.8 (CH2NH), 50.7,
55.0 (OCH3), 77.1 (CH-O), 79.1 [C(CH3)3], 98.1 [CH(OMe)2],
122.9 (C-4), 115.0 (C-6), 128.7 (C-2), 130.7 (C-3), 132.5 (C-5),
155.1, 156.2 (C-1, C=O) ppm. GC–MS: tR = 9.75; m/z (%) = 447
[M]+ (0.3), 445 [M]+ (0.3), 390 (0.5), 388 (0.5), 342 (4.2), 340 (3.0),
328 (4.3), 326 (4.7), 284 (4.1), 282 (4.6), 216 (38.1), 214 (36.6), 201
(12.3), 199 (13.0), 187 (5.6), 185 (6.0), 144 (100.0), 100 (16.9), 83
(20.2), 74 (15.3), 69 (8.3), 57 (75.7), 56 (13.0), 55 (14.9), 45 (6.3),
44 (9.1), 43 (10.7), 41 (30.8), 39 (7.8). HRMS (ESI+): calcd. for
C20H32BrNO5Na [M + Na]+ 468.1362; found 468.1351.
(؎)-tert-Butyl 2-[2-(Dimethoxymethyl)phenoxy]-3-methylbutylcarb-
amate (9d): A solution of salicylaldehyde (1.00 g, 8.19 mmol) in dry
MeOH (4 mL) was treated with trimethyl orthoformate (1.25 mL)
and Amberlyst 15 resin (125 mg), and the resulting mixture was
stirred at room temp. for 20 h. The mixture was diluted with
CH2Cl2 and filtered. The filtrate was treated with solid NaHCO3
(25 mg) and filtered again. The filtrate was evaporated, the residue
was dissolved in CH2Cl2/toluene, and the solution was evaporated
again to azeotropically remove all of the methanol. The resulting
crude oil was dissolved in dry THF (10 mL), and treated with race-
mic alcohol 2c (1.91 g, 9.42 mmol). A premixed solution of PPh3
(2.58 g, 9.83 mmol), and diisopropyl azodicarboxylate (DIAD;
1.93 mL, 9.42 mmol) in dry THF (10 mL) was the added dropwise
over 5 min. The resulting solution was stirred at room temp. over-
night. Then it was diluted with Et2O and washed with NaOH
(0.1 m aq.; 40 mL) to remove the unreacted phenols. After washing
with NH4Cl (saturated), the organic phase was evaporated, and the
residue was purified by chromatography (PE/EtOAc,
90:10Ǟ85:15) to give pure 9d (1.74 g, 60%) as an oil. Rf = 0.52
(PE/EtOAc, 80:20). 1H NMR (300 MHz, CDCl3, 0 °C): δ = 1.02
(S)-tert-Butyl 2-[4-Bromo-2-(dimethoxymethyl)phenoxy]-3-methyl-
butylcarbamate (9e): Compound (S)-9e was prepared from 5-bro-
mosalicylaldehyde and (R)-2b in 43% yield, following the same pro-
cedure used for 9c. [α]D = +33.6 (c = 2.75, CHCl3). The enantio-
meric excess was measured using HPLC on a chiral stationary
phase. Column: Chiralpak AD 250ϫ4.6 mm. Flow: 0.8 mLmin–1;
temp.: 35 °C, isocratic elution with n-hexane/iPrOH, 97:3. tR (S) =
6.67 min. tR (R) = 7.29 min. ee = 99%.
General Procedures for the Synthesis of Tetrahydrobenzoxazepines
13 and 14
Method A (from aldehydes 4): A solution of 4a or 4b (3.2 mmol) in
CH2Cl2 (20 mL) was treated with conc. HCl (37%; 2.80 mL). The
biphasic system was stirred for 1 h at room temp. (the aqueous
phase became first fluorescent yellow and then wine red). Then the
mixture was diluted with CH2Cl2 and treated with a solution of
Na2CO3 (2.026 g, 19.11 mmol, 0.5 equiv. relative to the HCl used)
in H2O (60 mL). After checking that the pH was Ͼ9, the phases
were separated, and the aqueous phase was re-extracted twice with
CH2Cl2. The combined organic extracts were washed with brine,
and the solvents were evaporated to dryness. This crude product
was used at once in the Ugi reaction. The residue was dissolved in
dry MeOH (5 mLmmol–1) and treated, at room temp., with carbox-
ylic acid (1.2 equiv.) and isocyanide (1.2 equiv.). After 48 h, the
mixture was evaporated to dryness, the residue was dissolved in
EtOAc, and the resulting solution was washed with NaHCO3 (satu-
rated aq.) to remove excess carboxylic acid. After evaporation, the
crude product was purified by chromatography (PE/EtOAc). The
two diastereomers could not be separated.
3
3
[d, J(H,H) = 6.9 Hz, 3 H, CH3CH], 1.06 [d, J(H,H) = 6.9 Hz, 3
H, CH3CH], 1.36 [s, 9 H, (CH3)3C], 2.08 [octuplet, 3J(H,H) =
6.7 Hz, 1 H, CH(CH3)2], 3.19 [ddd, J(H,H) = 4.8, 7.8, J(H,H) =
13.8 Hz, 1 H, CHHNHBoc], 3.24 (s, 3 H, OCH3), 3.47 (s, 3 H,
OCH3), 3.54 [ddd, 3J(H,H) = 2.7, 7.2, 2J(H,H) = 13.8 Hz, 1 H,
CHHNHBoc], 4.20 [ddd, J(H,H) = 2.7, 4.8, 6.8 Hz, 1 H, CH-O],
5.62 (br. t, 1 H, NH), 5.76 [s, 1 H, CH(OCH3)2], 6.89 [d, J(H,H)
3
2
3
3
3
4
= 8.4 Hz, 1 H, 6-H], 6.94 [td, J(H,H) = 7.6 (t), J(H,H) = 0.6 (d)
3
4
Hz, 1 H, 4-H], 7.26 [td, J(H,H) = 8.0 (t), J(H,H) = 1.8 (d) Hz, 1
H, 5-H], 7.51 [dd, 3J(H,H) = 7.6, 4J(H,H) = 1.6 Hz, 1 H, 3-H] ppm.
13C NMR (75 MHz, CDCl3, 20 °C): δ = 18.0, 18.6 [(CH3)2C], 28.4
[C(CH3)3], 30.2 [CH(CH3)2], 41.2 (CH2NH), 50.6, 55.0 (OCH3),
78.9 [C(CH3)3], 82.6 (CH-O), 98.8 [CH(OMe)2], 113.3 (C-6), 120.1
(C-4), 126.3 (C-2), 127.6 (C-3), 129.7 (C-5), 156.2, 156.5 (C-1,
C=O) ppm. IR: ν = 3443, 3342, 3040, 2952, 2927, 2829, 1701, 1599,
˜
1524, 1478, 1452, 1365, 1275, 1219, 1158, 1118, 1044, 1036, 974,
956, 910 cm–1. GC–MS: tR = 8.34; m/z (%) = 353 [M]+ (0.2), 322
(1.6), 296 (4.1), 250 (4.8), 248 (11.0), 234 (24.4), 190 (17.2), 161
(9.2), 145 (5.7), 137 (100.0), 136 (51.3), 130 (42.5), 121 (18.2), 119
Method B (from acetals 9): A solution of acetal 9 (2.64 mmol) in
CH2Cl2 (10 mL) was treated with conc. HCl (37%; 2.17 mL), and
the mixture was stirred for 6 h at room temp. The solution became
fluorescent yellow. The mixture was diluted with CH2Cl2 and
(8.2), 107 (27.9), 91 (7.4), 87 (10.7), 74 (10.4), 57 (42.0), 56 (9.8), treated with a solution of Na2CO3·H2O (1.78 g, 14.4 mmol) in H2O
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5073

L. Banfi, A. Bagno, A. Basso, C. De Santis, R. Riva, F. Rastrelli
FULL PAPER
(40 mL). After checking that the pH was Ͼ9, the phases were sepa-
rated, and the aqueous phase was re-extracted twice with CH2Cl2.
The combined organic extracts were washed with brine, and the
solvents were evaporated to dryness. This crude product was used
at once in the Ugi reaction. The residue was dissolved in dry
MeOH (5 mLmmol–1) and treated, at room temp., with carboxylic
acid (1.2 equiv.) and isocyanide (1.2 equiv.). After 48 h, the mixture
was evaporated to dryness, the residue was dissolved in EtOAc, and
the resulting solution was washed with NaHCO3 (saturated aq.) to
remove excess carboxylic acid. After evaporation, the crude prod-
uct was purified by chromatography (PE/EtOAc). The two dia-
stereomers could not be separated, with the exception of com-
pounds 13k–14k.
[CH(CH3)2], 28.0 [C(CH3)3], 41.3 (CH2iPr), 50.5 (br., C-3), 50.7
[C(CH3)3], 62.1 (C-5), 77.2 (C-2), 167.0, 162.5 (C=O).
Computational Details: The conformational analysis was performed
with Spartan ’02[33] using the Conformational Search module. Rel-
evant options were set as follows: SearchMethod = MonteCarlo,
MaxConfs = 400, Window = 10 kcalmol–1. All of the output geo-
metries were then reoptimized with Gaussian 09[34] at the B3LYP/
6-31G(d,p) level of theory. 1H and 13C shieldings (σ) of the lowest-
energy conformers, as well as J-couplings, were calculated at the
B3LYP/cc-pVTZ level.[24] Chemical shifts were calculated as δ =
1
σ – σref, where σref is the reference shielding of TMS (31.6 for H
and 183.6 for 13C).[24] Full data are given in the Supporting Infor-
mation.
N-(tert-Butyl)-9-chloro-4-(5-chlorothiophene-2-carbonyl)-2-isobutyl-
2,3,4,5-tetrahydrobenzo[f][1,4]oxazepine-5-carboxamides 13j and
14j: White solid, yield 76%. cis:trans ratio = 87:13 (1H NMR). Rf
= 0.60 (PE/EtOAc, 70:30). A sample of pure 13f was obtained by
trituration with Et2O/pentane followed by crystallization from
methanol, m.p. 128.5–129.5. 1H NMR (300 MHz, [D6]DMSO,
70 °C; Note. at 30 or 70 °C, only one rotamer is visible. At 70 °C
Supporting Information (see footnote on the first page of this arti-
cle): Details of computational conformational analysis and theoret-
ical NMR predictions. Details of the Mosher ester analysis of (R)-
2b. Spectroscopic data for compounds 4b, 9f, 14a–i, and 14k–n.
1
Copies of H and 13C NMR spectra of all new compounds.
3
Acknowledgments
the peaks are sharper): δ = 0.91 [d, J(H,H) = 6.6 Hz, 3 H, CH3],
0.92 [d, 3J(H,H) = 6.6 Hz, 3 H, CH3], 1.27 [s, 9 H, (CH3)3C], 1.25–
3
2
The authors thank Dr. Benedetta Pollarolo, Miss Alice Brambilla,
and Dr. Valentina Cerulli for their contributions to this work, Dr.
Andrea Armirotti and Dr. Sine Mandrup Bertozzi for performing
HRMS analyses, and Dr. Valeria Rocca for HPLC analyses. The
authors thank Ministero dell’Università e della Ricerca (MIUR)
(PRIN 2009 project, Synthetic Methodologies for Generation of
Biologically Relevant Molecular Diversity) for financial assistance,
and Novo Nordisk for the gift of Novozym enzyme.
1.37 (m, 1 H, CHH-iPr), 1.71 [ddd, J(H,H) = 5.4, 8.4, J(H,H) =
14.1 Hz, 1 H, CHH-iPr], 2.07 (mc) [m, 1 H, CH(CH3)2], 3.62 [dd,
3J(H,H) = 9.3, J(H,H) = 15.3 Hz, 1 H, 3-H], 3.87 [tdd, J(H,H) =
8.7 (t), 1.5 (d), 4.8 (d) Hz, 1 H, 2-H], 4.23 [d, J(H,H) = 15.3 Hz,
1 H, 3-H], 5.92 (s, 1 H, 5-H), 6.84 (s, 1 H, NH), 7.10 [t, J(H,H) =
2
3
2
3
3
7.8 Hz, 1 H, 7-H], 7.14 [d, J(H,H) = 3.9 Hz, 1 H, H thienyl meta
to C=O], 7.24 [dd, 3J(H,H) = 7.5, 4J(H,H) = 1.5 Hz, 1 H, 6-H],
7.28 [d, 3J(H,H) = 3.9 Hz, 1 H, H thienyl ortho to C=O], 7.45
[dd, J(H,H) = 8.0, J(H,H) = 1.7 Hz, 1 H, 8-H] ppm. 1H NMR
(300 MHz, CDCl3, 55 °C): δ = 0.94 [d, 3J(H,H) = 6.3 Hz, 3 H,
CH3CH], 0.96 [d, 3J(H,H) = 6.6 Hz, 3 H, CH3CH], 1.26 [s, 9 H,
(CH3)3C], 1.29–1.38 (m, 1 H, CHH-iPr), 1.82 [ddd, 3J(H,H) = 4.5,
3
4
[1] a) L. Banfi, A. Basso, R. Riva, Top. Heterocycl. Chem. 2010,
23, 1–39; b) L. Banfi, R. Riva, A. Basso, Synlett 2010, 23–41.
[2] C. Hulme, J. Dietrich, Mol. Diversity 2009, 13, 195–207.
8.4, 2J(H,H) = 14.4 Hz, 1 H, CHH-iPr], 2.19 (mc) [m, 1 H, [3] L. Banfi, A. Basso, G. Guanti, R. Riva, in: Multicomponent
CH(CH3)2], 3.65 [dd, J(H,H) = 10.8, J(H,H) = 14.7 Hz, 1 H, 3-
H], 3.84 [td, 3J(H,H) = 8.7 (t), 4.2 (d) Hz, 1 H, 2-H], 4.35 [d,
2J(H,H) = 14.7 Hz, 1 H, 3-H], 5.19 (s, 1 H, NH), 6.00 (s, 1 H, 5-
3
2
Reactions (Eds.: J. Zhu, H. Bienaymé), Wiley-VCH, Weinheim,
Germany, 2005, p. 1–32.
[4] C. L. Kelly, K. W. M. Lawrie, P. Morgan, C. L. Willis, Tetrahe-
dron Lett. 2000, 41, 8001–8005. We have recently observed
complete epimerization during Ugi reactions of cyclic chiral
aldehydes with two stereogenic centres (unpublished work).
[5] a) A. G. Zhdanko, V. G. Nenajdenko, Eur. J. Org. Chem. 2009,
74, 884–887; b) L. Banfi, A. Basso, R. Riva, in: Isocyanide
Chemistry (Ed.: V. Nenajdenko), Wiley-VCH, Weinheim, Ger-
many, 2012, p. 1–33; c) D. W. Carney, J. V. Truong, J. K. Sello,
J. Org. Chem. 2011, 76, 10279–10285.
3
3
H), 6.86 [d, J(H,H) = 3.9 Hz, 1 H, thienyl CH], 7.04 [t, J(H,H)
3
= 7.8 Hz, 1 H, 7-H], 7.12 [br. d, J(H,H) not measurable, 1 H, 6-
H], 7.15 [d, J(H,H) = 3.9 Hz, 1 H, thienyl CH], 7.42 [dd, 3J(H,H)
3
= 7.8, 4J(H,H) = 1.8 Hz, 1 H, 8-H] ppm. 13C NMR (75 MHz, [D6]-
DMSO, 70 °C): δ = 22.6, 21.4 (CH3), 23.3 [CH(CH3)2], 28.0
[C(CH3)3], 41.7 (CH2iPr), 50.5 (br., C-3), 50.6 [C(CH3)3], 63.0 (C-
5), 80.8 (C-2), 124.1 (C-7), 126.2 (C-9), 126.6 (C thienyl meta to
C=O), 128.9 (C thienyl ortho to C=O), 129.8, 129. 7 (C-6 and C-
8), 131.8 (C-C=O thienyl), 132.1 (C-5a), 135.5 (C-Cl thienyl), 152.2
(C-9a), 162.7 (thienyl C=O), 165.4 (NH-C=O) ppm. HRMS
(ESI+): calcd. for C23H29Cl2N2O3S [M + H]+ 483.1276; found
483.1276.
[6] a) A. Basso, L. Banfi, R. Riva, Eur. J. Org. Chem. 2010, 1831–
1841; b) V. G. Nenajdenko, A. L. Reznichenko, E. S. Balen-
kova, Tetrahedron 2007, 63, 3031–3041; c) K. Sung, F.-L. Chen,
M.-J. Chung, Mol. Diversity 2003, 6, 213–221; d) S. S.
van Berkel, B. G. M. Bögels, M. A. Wijdeven, B. Westermann,
F. P. J. T. Rutjes, Eur. J. Org. Chem. 2012, 3543–3559; e) T. M.
Vishwanatha, N. Narendra, V. V. Sureshbabu, Tetrahedron
Lett. 2011, 52, 5620–5624; f) I. Kanizsai, S. Gyonfalvi, Z.
Szakonyi, R. Sillanpaa, F. Fulop, Green Chem. 2007, 9, 357–
360; g) V. P. Mehta, S. G. Modha, E. Ruijter, K. Van Hecke, L.
Van Meervelt, C. Pannecouque, J. Balzarini, R. V. A. Orru, E.
Van der Eycken, J. Org. Chem. 2011, 76, 2828–2839.
Selected spectroscopic data for 14j: 1H NMR (300 MHz, [D6]-
3
DMSO, 70 °C): δ = 0.84 [d, J(H,H) = 6.6 Hz, 3 H, CH3], 0.92 [d,
3J(H,H) = 6.6 Hz, 3 H, CH3], 1.30 [s, 9 H, (CH3)3C], 1.37–1.25 (m,
1 H, CHH-iPr), 1.70–1.57 (m, 1 H, CHH-iPr), 1.95 (mc) [m, 1 H,
CH(CH3)2], 3.81 [dd, 3J(H,H) = 3.2, 2J(H,H) = 13.8 Hz, 1 H, 3-
H], 3.99 [dd, 3J(H,H) = 9.9, 3J(H,H) = 13.8 Hz, 1 H, 3-H], 4.64
[7] R. F. Nutt, M. M. Joullie, J. Am. Chem. Soc. 1982, 104, 5852–
5853.
3
[ddt, J(H,H) = 9.0 (d), 9.9 (d), 3.6 (t) Hz, 1 H, 2-H], 5.95 (s, 1 H,
[8] A. Znabet, E. Ruijter, F. J. J. de Kanter, V. Kohler, M. Helliwell,
N. J. Turner, R. V. A. Orru, Angew. Chem. 2010, 122, 5417–
5420; Angew. Chem. Int. Ed. 2010, 49, 5289–5292.
[9] a) K. M. Bonger, T. Wennekes, D. V. Filippov, G. Lodder, G. A.
van der Marel, H. S. Overkleeft, Eur. J. Org. Chem. 2008, 3678–
3688; b) V. Cerulli, L. Banfi, A. Basso, V. Rocca, R. Riva, Org.
3
5-H), 6.84 (s, 1 H, NH), 7.03 [t, J(H,H) = 7.8 Hz, 1 H, 7-H], 7.14
[d, 3J(H,H) = 3.9 Hz, 1 H, H thienyl], 7.24 [dd, 3J(H,H) = 7.5,
4J(H,H) = 1.5 Hz, 1 H, 8-H], 7.28 [d, 3J(H,H) = 3.9 Hz, 1 H, H
thienyl], 7.38 [dd, 3J(H,H) = 8.0, 4J(H,H) = 1.7 Hz, 1 H, 6-H] ppm.
13C NMR (75 MHz, [D6]DMSO, 70 °C): δ = 22.5, 21.1 (CH3), 23.6
5074
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 5064–5075

Long-Range Diastereoselectivity in an Ugi Reaction
Biomol. Chem. 2012, 10, 1255–1274; c) T. M. Chapman, I. G. [22]
Davies, B. Gu, T. M. Block, D. I. C. Scopes, P. A. Hay, S. M.
P. Del Buttero, G. Molteni, A. Papagni, L. Miozzo, Tetrahe-
dron: Asymmetry 2004, 15, 2555–2559.
Courtney, L. A. McNeill, C. J. Schofield, B. G. Davis, J. Am.
Chem. Soc. 2005, 127, 506–507; d) M. S. M. Timmer, M. D. P.
Risseeuw, M. Verdoes, D. V. Filippov, J. R. Plaisier, G. A.
van der Marel, H. S. Overkleeft, J. H. van Boom, Tetrahedron:
Asymmetry 2005, 16, 177–185; e) A. Znabet, J. Zonneveld, E.
Janssen, F. J. J. De Kanter, M. Helliwell, N. J. Turner, E. Ru-
ijter, R. V. A. Orru, Chem. Commun. 2010, 46, 7706–7708.
[10] a) D. M. Flanagan, M. M. Joullié, Synth. Commun. 1989, 19,
1–12; b) I. Schlemminger, H. H. Janknecht, W. Maison, W.
Saak, J. Martens, Tetrahedron Lett. 2000, 41, 7289–7292.
[11] a) L. Banfi, A. Basso, G. Guanti, S. Merlo, C. Repetto, R.
Riva, Tetrahedron 2008, 64, 1114–1134; b) T. Iizuka, S. Takigu-
chi, Y. Kumakura, N. Tsukioka, K. Higuchi, T. Kawasaki, Tet-
rahedron Lett. 2010, 51, 6003–6005.
[12] S. Murrison, S. K. Maurya, C. Einzinger, B. McKeever-Abbas,
S. Warriner, A. Nelson, Eur. J. Org. Chem. 2011, 2354–2359.
[13] a) H. Gröger, M. Hatam, J. Martens, Tetrahedron 1995, 51,
7173–7180; b) W. Maison, A. Lützen, M. Kosten, I. Schlem-
minger, O. Westerhoff, J. Martens, J. Chem. Soc. Perkin Trans.
1 1999, 3515–3525; c) W. Maison, A. Lützen, M. Kosten, I.
Schlemminger, O. Westerhoff, W. Saak, J. Martens, J. Chem.
Soc. Perkin Trans. 1 2000, 1867–1871; d) D. G. Zhu, R. J. Chen,
H. B. Liang, S. Li, L. Pan, X. C. Chen, Synlett 2010, 897–900.
[14] C. A. Sperger, P. Mayer, K. T. Wanner, Tetrahedron 2009, 65,
10463–10469.
[23]
[24]
[25]
[26]
G. Cainelli, P. Galletti, D. Giacomini, Chem. Rev. 2009, 38,
990–1001.
A. Bagno, F. Rastrelli, G. Saielli, Chem. Eur. J. 2006, 12, 5514–
5525.
N. Cheron, R. Ramozzi, L. El Kaim, L. Grimaud, P. Fleurat-
Lessard, J. Org. Chem. 2012, 77, 1361–1366.
a) N. J. Adderley, D. J. Buchanan, D. J. Dixon, D. I. Lainé, An-
gew. Chem. 2003, 115, 4373–4376; Angew. Chem. Int. Ed. 2003,
42, 4241–4244; b) D. Enders, A. Haertwig, G. Raabe, J. Run-
sink, Eur. J. Org. Chem. 1998, 1771–1792; c) D. M. Hodgson,
N. S. Kaka, Angew. Chem. 2008, 120, 10106–10108; Angew.
Chem. Int. Ed. 2008, 47, 9958–9960.
[27]
[28]
J. A. Dale, H. S. Mosher, J. Am. Chem. Soc. 1973, 95, 512–519.
R. J. Kazlauskas, A. N. E. Weissfloch, A. T. Rappaport, L. A.
Cuccia, J. Org. Chem. 1991, 56, 2656–2665.
[29] a) N. K. Anand, C. M. Blazey, O. J. Bowles, C. A. Buhr, J. Bus-
senius, J. K. Curtis, S. C. Defina, L. Dubenko, J. R. Harris, E.
Jackson-Ugueto, A. Joshi, A. I. Kim, A. L. Tsuhako, S. Ma, J.-
C. L. Manalo, S. Ng, C. J. Peto, K. D. Rice, T. H. Tsang, C. A.
Zaharia, WO2010135524, 2010; b) M. D. Brewer, M. N. Bur-
gess, R. J. J. Dorgan, R. L. Elliott, P. Mamalis, B. R. Manger,
R. A. B. Webster, J. Med. Chem. 1989, 32, 2058–2062; c) M.
Díaz-Gavilán, J. A. Gómez-Vidal, A. Entrena, M. A. Gallo, A.
Espinosa, J. M. Campos, J. Org. Chem. 2006, 71, 1043–1054;
d) T. Fujimoto, J. Kunitomo, Y. Tomata, K. Nishiyama, M.
Nakashima, M. Hirozane, S.-i. Yoshikubo, K. Hirai, S. Marui,
Bioorg. Med. Chem. Lett. 2011, 21, 6414–6416.
[15] a) L. Banfi, A. Basso, V. Cerulli, G. Guanti, P. Lecinska, I.
Monfardini, R. Riva, Mol. Diversity 2010, 14, 425–442; b) L.
Banfi, A. Basso, G. Guanti, N. Kielland, C. Repetto, R. Riva,
J. Org. Chem. 2007, 72, 2151–2160; c) L. Banfi, A. Basso, G.
Guanti, P. Lecinska, R. Riva, Org. Biomol. Chem. 2006, 4,
4236–4240.
[30] F. Silva, M. Reiter, R. Mills-Webb, M. Sawicki, D. Klaer, N.
Bensel, A. Wagner, V. Gouverneur, J. Org. Chem. 2006, 71,
8390–8394.
[16] L. Banfi, A. Basso, V. Cerulli, V. Rocca, R. Riva, Beilstein J.
Org. Chem. 2011, 7, 976–979.
[31] K. Tamaki, T. Yamaguchi, K. Oda, N. Terasaka, D. Makai,
M. Nakadai, WO2006046593, 2006.
[17] a) Y. Basel, A. Hassner, J. Org. Chem. 2000, 65, 6368–6380; b)
Z. Guo, Y.-F. Zhu, T. D. Gross, F. C. Tucci, Y. Gao, M. Moor-
jani, P. J. Connors, M. W. Rowbottom, Y. Chen, R. S. Stru-
thers, Q. Xie, J. Saunders, G. Reinhart, T. K. Chen, A. L. K.
Bonneville, Chen, J. Med. Chem. 2004, 47, 1259–1271; c) F. C.
Tucci, Y.-F. Zhu, Z. Guo, T. D. Gross, P. J. Connors Jr, R. S.
Struthers, G. J. Reinhart, J. Saunders, C. Chen, Bioorg. Med.
Chem. Lett. 2003, 13, 3317–3322.
[18] C. Incarvito, M. Lam, B. Rhatigan, A. L. Rheingold, C. J. Qin,
A. L. Gavrilova, B. Bosnich, J. Chem. Soc., Dalton Trans. 2001,
3478–3488.
[19] a) B. W. Dymock, P. S. Jones, J. H. Merrett, M. J. Parratt,
US2002/32221A1, 2002; b) J. Liu, N. Ikemoto, D. Petrillo, J. D.
Armstrong, Tetrahedron Lett. 2002, 43, 8223–8226.
[20] R. N. Misra, H. Y. Xiao, K. S. Kim, S. F. Lu, W. C. Han, S. A.
Barbosa, J. T. Hunt, D. B. Rawlins, W. F. Shan, S. Z. Ahmed,
L. G. Qian, B. C. Chen, R. L. Zhao, M. S. Bednarz, K. A. Kel-
lar, J. G. Mulheron, R. Batorsky, U. Roongta, A. Kamath, P.
Marathe, S. A. Ranadive, J. S. Sack, J. S. Tokarski, N. P. Pavlet-
ich, F. Y. F. Lee, K. R. Webster, S. D. Kimball, J. Med. Chem.
2004, 47, 1719–1728.
[21] a) T. V. Hansen, L. Skattebol, Org. Synth. 2005, 82, 64–67; b)
N. U. Hofslokken, L. Skattebol, Acta Chem. Scand. 1999, 53,
258–262.
[32] S. Bertini, A. De Cupertinis, C. Granchi, B. Bargagli, T. Tucci-
nardi, A. Martinelli, M. Macchia, J. R. Gunther, K. E. Carl-
son, J. A. Katzenellenbogen, F. Minutolo, Eur. J. Org. Chem.
2011, 46, 2453–2462.
[33] Spartan ’02, Wavefunction, Inc. Irvine, CA.
[34] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, Montgomery Jr, J. A. J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09, rev.
A.1, Gaussian, Inc., Wallingford CT, 2009.
Received: April 14, 2013
Published Online: July 2, 2013
Eur. J. Org. Chem. 2013, 5064–5075
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5075