Synthesis of 3-aryl/benzyl-4,5,6,6a-tetrahydro-3a H-pyrrolo[3,4-d]isoxazole derivatives: A comparison between conventional, microwave-assisted and flow-based methodologies

Article abstract of DOI:10.1021/jo1014323

Two modern synthetic technologies to perform 1,3-dipolar cycloaddition reactions were compared. This study puts in evidence the power of microwave-assisted and flow-based methodologies compared to the conventional one in terms of reaction time and yield,

Full text of DOI:10.1021/jo1014323

pubs.acs.org/joc
a solution of the stable precursor hydroximoyl halide and the
Synthesis of 3-Aryl/benzyl-4,5,6,6a-tetrahydro-3aH-
pyrrolo[3,4-d ]isoxazole Derivatives: A Comparison
between Conventional, Microwave-Assisted and
Flow-Based Methodologies
dipolarophile.3 Alternatively, an efficient strategy involves
the use of a heterogeneous mixture of an organic solvent,
e.g., ethyl acetate, and an inorganic base, e.g., NaHCO3 or
KHCO3.4 Both methods allow the maintenance of a low
concentration of the dipole, thus preventing dimerization
and promoting its reaction with the dipolarophile. The draw-
back of the above-described strategy is the slowness of the
reaction, which can take up to several days or weeks.
Microwave-assisted methodology is a well established way
to improve a reaction outcome and speed up the process. The
advantage of such a methodology applied to 1,3-dipolar
cycloaddition has been highlighted.5 Besides, flow chemistry
is an emerging technology to implement and expedite classi-
cal organic reactions.6 Very recently, this methodology has
been successfully applied to cycloaddition reactions.7
In this paper we analyzed and compared the usefulness of
these two modern methodologies applied to the synthesis of
bicyclic-Δ2-isoxazolines of general structure 3, derived from
1,3-dipolar cycloaddition of differently substituted nitrile
oxides to N-Boc-Δ3-pyrroline 2 (Scheme 1).
Sabrina Castellano,† Lucia Tamborini,‡ Monica Viviano,†
Andrea Pinto,‡ Gianluca Sbardella,† and Paola Conti*,‡
†Dipartimento di Scienze Farmaceutiche, Universitaꢀ degli
Studi di Salerno, Via Ponte Don Melillo, 84084 Fisciano,
Italy, and ‡Dipartimento di Scienze Farmaceutiche
ꢀ
“Pietro Pratesi”, Universita degli Studi di Milano,
Via Mangiagalli 25, 20133 Milano, Italy
paola.conti@unimi.it
Received July 21, 2010
SCHEME 1. General Reaction Scheme
Two modern synthetic technologies to perform 1,3-dipo-
lar cycloaddition reactions were compared. This study
puts in evidence the power of microwave-assisted and
flow-based methodologies compared to the conventional
one in terms of reaction time and yield, and demonstrates
the potential of flow chemistry in terms of time, automa-
tion, and scaling up opportunities.
3-Aryl-(or benzyl-)4,5,6,6a-tetrahydro-3aH-pyrrolo[3,4-d]-
isoxazole derivatives are biologically interesting molecules,
since they can be considered as frozen analogues of arylalk-
ylamines, thus being useful tools for structure-activity
studies in different medicinal chemistry areas. As a matter
of fact, such a scaffold has been used in several biologically
active compounds, including nicotinic receptor ligands,8
antibacterial agents,9 and neuroleptics.10
The target molecules can be obtained according to the reac-
tion depicted in Scheme 1. As previously reported with other
types of hydroxamoyl halides, alkene 2 has a poor reactivity,
and unless the generated dipole is highly reactive (e.g.,
bromonitrileoxide), the cycloaddition reaction gives low
The 1,3-dipolar cycloaddition offers a convenient one-step
route for the construction of a variety of complex five-
membered heterocycles. 1,3-Dipolar cycloadditions of in situ
generated nitrile oxides with alkenes are well-documented
and provide access to Δ2-isoxazolines.1 Aldoximes are estab-
lished precursors of nitrile oxides, and different classes of
reagents have been used in the literature for the conversion of
aldoximes to nitrile oxides.2 The outcome of the reaction is
strongly dependent on the nature of the dipolarophile. With
sluggish dipolarophiles, the 1,3-dipole must be generated
slowly so as to disfavor dimerization of the nitrile oxide to
give furoxan (1,2,5-oxadiazol-2-oxide) as an unwanted side
product. Slow generation of the nitrile oxide can be achieved
by addition of an organic base by means of a syringe pump to
(3) Christl, M.; Huisgen, R. Chem. Ber. 1973, 106, 3345–3367.
(4) Vyas, D. M.; Chiang, Y.; Doyle, T. W. Tetrahedron Lett. 1984, 25,
487–490.
(5) (a) Spiteri, C.; Mason, C.; Zhang, F.; Ritson, D. J.; Sharma, P.;
Keeling, S.; Moses, J. E. Org. Biomol. Chem. 2010, 8, 2537–2542. (b) Pineiro,
M.; Pinho e Melo, T. M. V. D. Eur. J. Org. Chem. 2009, 5287–5307.
(6) New Avenues to Efficient Chemical Synthesis. Emerging Technologies;
Seeberger, P. H., Blume, T., Eds.; Sprinter-Verlag: Berlin, Heidelberg, 2007.
(7) (a) Smith, C. D.; Baxendale, I. R.; Lanners, S.; Hayward, J. J.; Smith,
S. C.; Ley, S. V. Org. Biomol. Chem. 2007, 5, 1559–1561. (b) Baumann, M.;
Baxendale, I. R.; Ley, S. V. Synlett 2010, 5, 749–752.
(1) (a) Caramella, P.; Grunanger, P. Nitrile oxides and imines. In 1,3-
Dipolar Cycloaddition Chemistry; Padwa, A., Ed.; John Wiley & Sons Inc.:
New York, 1984. (b) Nitrile Oxides, Nitrones and Nitronates in Organic
Synthesis: Novel Strategies in Synthesis, 2nd ed.; Fewer, E., Ed.; John Wiley
& Sons Inc.: Hoboken, NJ, 2008.
(8) Dallanoce, C.; Bazza, P.; Grazioso, G.; De Amici, M.; Gotti, C.;
Riganti, L.; Clementi, F.; De Micheli, C. Eur. J. Org. Chem. 2006, 16, 3746–
3754.
(9) Macielag, M. J.; Weidner-Wells, M. A.; Lin, S.-C. US Patent Applica-
tion, US 20090029980.
(2) (a) Dubrovskiy, A. V.; Larock, R. C. Org. Lett. 2010, 12, 1180–1183.
(b) Lee, G. A. Synthesis 1982, 508–509.
(10) Steiner, G.; Munschauer, R.; Unger, L.; Teschendorf, H.-J.; Hoeger,
T. PCT Int. Appl. WO95/15327.
DOI: 10.1021/jo1014323
r
Published on Web 09/30/2010
J. Org. Chem. 2010, 75, 7439–7442 7439
2010 American Chemical Society

Castellano et al.
JOCNote
TABLE 1. Reaction of Chloroximes 1 with N-Boc-Δ3-Pyrroline 2 under
Conventional Reaction Conditions at Room Temperature
TABLE 3. Microwave-Assisted Cycloaddition of Chloroximes 1a-f to
N-Boc-Δ3-Pyrroline 2
time
equiv of 1a solvent (days) yieldb (%)
entrya
R
time (h)
yieldb (%)
entry
R
a
b
c
d
e
f
Ph
p-NO2-C6H4
p-CH3O-C6H4
PhCH2
p-NO2-C6H4-CH2
p-CH3O-C6H4-CH2
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
66
60
67
50
47
57
a
b
c
d
e
f
Ph
p-NO2-C6H4
p-CH3O-C6H4
PhCH2
p-NO2-C6H4-CH2
p-CH3O-C6H4-CH2
4
4
4
4
4
4
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
EtOAc
7
7
7
7
7
7
40
28
45
32
<2
35
aResults obtained using the optimized conditions reported in Table 2,
aAdded portionwise to the reaction mixture. bIsolated yield.
entry 7. bIsolated yield.
ranged from 5 to 20 W, and the registered pressure ranged from
30 to 50 psi. Cooling was kept off while irradiating. Heating
cycles of 30 min each were applied. The first equivalent of
chloroxime was added at the beginning of the experiment, and
further 0.5 equiv portions were added before every additional
heating cycle, since monitoring by TLC showed the presence of
unreacted alkene 2, while chloroxime 1a was completely con-
sumed. The highest yield was obtained after the addition of 2
equiv of chloroxime and 3 heating cycles (Table 2, entry 4). As
observed under conventional heating conditions, the addition of
a small percentage of water (1% v/v) to the reaction medium had
a positive effect on the reaction outcome (Table 2, entry 7 vs 4).
Increasing the temperature over 80 °C was counterproductive
(Table 2, entries 9 and 10 vs 7).
TABLE 2. Optimization of Reaction Conditions for the Cycloaddition
of Chloroxime 1a (R = Ph) to N-Boc-Δ3-Pyrroline 2, under Conventional
Heating or Microwave Irradiation
equiv
of 1aa
time
(h)
yieldd
(%)
entry
solvent
EtOAc
T (°C)
1
2
3
4
5
6
7
8
9
10
2.0
2.0
1.5
2.0
2.5
3.0
2.0
2.0
2.0
2.0
72
72
refluxb
refluxb
80c
42
51
35
48
41
34
66
47
56
44
EtOAc þ H2O (1%)
EtOAc
EtOAc
2 ꢀ 0.5
3 ꢀ 0.5
4 ꢀ 0.5
5 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
3 ꢀ 0.5
80c
EtOAc
EtOAc
EtOAc þ H2O (1%)
EtOAc þ H2O (2%)
EtOAc þ H2O (1%)
EtOAc þ H2O (1%)
80c
80c
80c
80c
90c
100c
aAdded portionwise to the reaction mixture. bConventional heating.
cMicrowave irradiation. dIsolated yield.
Thus, we selected as optimal conditions those reported in
entry 7. While operating under these conditions, we extended
our procedure to a series of differently substituted chlorox-
imes 1a-f. The data reported in Table 3 demonstrate that this
methodology is versatile and can be fruitfully applied to the
synthesis of a series of 3-substituted isoxazolines 3. In fact, all
derivatives were obtained in acceptable yields (47-67%).
Notably, the reaction time was significantly reduced from
7 days (at room temperature) or 3 days (refluxing under
conventional heating) to 1.5 h, and very importantly, the out-
come of the reaction was not anymore dependent on the nature
of the chloroxime. Indeed, we were able to obtain derivative 3e
in 47% yield (compare entry e in Table 1 and in Table 3).
The main limitation of the described methodology is
related to the possibility of scaling up the reaction. In fact,
microwave devices commonly available in a research labora-
tory are designed to fit 0.2-35 mL sealed tubes, allowing
performance of the reaction up to a maximum total volume
of 25 mL. This entails the need to repeat the reaction several
times in order to produce the desired cycloadduct in multi-
gram scale. Moreover, as pointed out above, the procedure,
albeit optimized, requires multiple operator interventions to
add additional aliquots of chloroxime over time.
yields.11 Thus, a slow generation of the dipole is required. We
initially carried out the reaction under conventional reaction
conditions: the nitrile oxide was generated in situ by treating
chloroxime 1 with excess solid NaHCO3 in EtOAc, in the
presence of alkene 2, and the reaction was carried out at room
temperature.9,11 As a result of the partial dimerization of the
dipole, the addition of further aliquots of chloroxime over time
was necessary. The series of differently substituted chloroximes
1a-f was prepared from the corresponding aldehydes according
to a procedure reported in the literature.2a The cycloaddition
reaction proceeded slowly, yielding less than 50% of the desired
product after 7 days (Table 1). The yield was strongly dependent
on the nature of the chloroxime ranging from 45% (entry c, R=
p-CH3O-C6H4) toaslowas<2%(entrye,R=p-NO2-C6H4-CH2).
The detailed results are reported in Table 1.
We then optimized the reaction conditions using chloroxime
1a as a model substrate. When the reaction was carried out
under conventional heating by refluxing in EtOAc, the yield
was comparable to that obtained at room temperature (Table 2,
entry 1 vs Table 1, entry a), but reaction time was reduced from
7 to 3 days and only 2 equiv of 1a were consumed. The yield can
be further improved up to 51% by adding a small percentage of
water (1% v/v) to the reaction medium (Table 2, entry 2). The
potential usefulness of the microwave-assisted methodology
was then investigated. We performed a series of experiments by
varying the reaction conditions, i.e., the number of equivalents
of 1a, time, temperature, and percentage of added water, with
the aim of fine-tuning the best reaction conditions (Table 2).
The use of solid NaHCO3 (4 equiv) as a base was kept as a
constant in all the experiments.
Thus we decided to investigate the feasibility of this type of
cycloaddition reaction under continuous-flow conditions.12
For this purpose, we used the R2þ/R4 flow system commer-
cially available from Vapourtec.
First, we chose the appropriate base among a selection of
polymer-supported (PS) bases (i.e., PS-sodium carbonate,
Amberlyst A-21, PS-DBU,13 PS-BEMP14), as well as solid
(12) (a) Bedore, M. W.; Zaborenko, N.; Jensen, K. F.; Jamison, T. F. Org.
Process Res. Dev. 2010, 14, 432–440. (b) Damm, M.; Glasnov, T. N.; Kappe,
C. O. Org. Process Res. Dev. 2010, 14, 215–224. (c) Glasnov, T. N.; Findenig,
S.; Kappe, C. O. Chem.;Eur. J. 2009, 15, 1001–1010. (d) Mennecke, K.;
Solodenko, W.; Kirschning, A. Synthesis 2008, 10, 1589–1599.
(13) 1,8-Diazabicyclo[5.4.0]undec-7-ene, polymer bound (PS-DBU).
(14) 2-tert-Butylimino-2-diethylamino-1,3-dimethyl-perhydro-1,3,2-dia-
zaphosphorine, polymer bound (PS-BEMP).
Microwave irradiation was performed by setting the tempera-
ture to the desired value. The average microwave output power
(11) Conti, P.; De Amici, M.; Pinto, A.; Tamborini, L.; Grazioso, G.;
Frølund, B.; Nielsen, B.; Thomsen., C.; Ebert, B.; De Micheli, C. Eur. J. Org.
Chem. 2006, 24, 5533–5542.
7440 J. Org. Chem. Vol. 75, No. 21, 2010

Castellano et al.
JOCNote
TABLE 4. Cycloaddition of Chloroxime 1a (R = Ph) to N-Boc-Δ3-
TABLE 5. Cycloaddition of Chloroxime 1a-f to N-Boc-Δ3-Pyrroline 2
Pyrroline 2 under Continuous-Flow Conditionsa
under Continuous-Flow Conditionsa
entry
equiv of 1ab
residence time (min)
T (°C)
yieldc
entrya
R
time (min)
yieldb (%)
1
2
3
4
5
6
7
1.0
1.5
2.0
1.5
1.5
4.5
1.5
30
30
30
10
10
10
10
80
80
80
80
50
60
63
45
69
87
45
a
b
c
d
e
f
Ph
p-NO2-C6H4
p-CH3O-C6H4
PhCH2
p-NO2-C6H4-CH2
p-CH3O-C6H4-CH2
10
10
10
10
10
10
69
60
73
71
67
64
90
90
100d
aResults obtained using the optimized conditions reported in Table 4,
aReactions conducted on a 0.25 mmol scale of N-Boc-Δ3-pyrroline 2 in
EtOAc (0.25 M solution), using solid K2CO3 as a base (4.0 equiv). bAll
solutions were prepared in 1 mL of EtOAc. cIsolated yield. dA further increase
in the temperature up to 130 °C causes a dramatic reduction of the yield (8%).
entry 5. bIsolated yield.
Since unreacted starting material was recovered in all of the
performed experiments, we resolved to test the reaction using a
larger excess of chloroxime 1a (entry 6). In this case, complete
conversion of the alkene was observed, and the yield was
improved to 87%, but the isolation of the product was hampered
by the presence of a large amount of furoxan and other
byproducts and required intensive and time-consuming purifica-
tions by column chromatography. Therefore, we selected the
conditions reported in Table 4, entry 5 as the optimal ones and
validated our synthetic protocol by applying such optimized con-
ditions to the series of chloroximes 1a-f. In all cases (Table 5),
the cycloadducts were obtained in good yield (60-73%), very
short time (10 min), and using as low as 1.5 equiv of chloroxime.
In conclusion, we have compared three different methodolo-
gies to prepare 3-substituted-4,5,6,6a-tetrahydro-3aH-pyrrolo-
[3,4-d ]isoxazole derivatives by means of 1,3-dipolar cyclo-
addition. Under conventional reaction conditions the reaction
proceeded very slowly (7 days) with low yield (e45%), and it
was sometimes impossible to obtain the desired product (i.e.,
compound 3e). Running the reaction under microwave irradia-
tion resulted in both a yield increase, especially in the case of
poorly reactive dipoles, and a considerable acceleration of the
process from 7 days to 1.5 h. Yet, multiple operator interventions
and excess of chloroxime were needed. Notably, if conventional
heating was applied, the reaction was speeded up to a much
lesser extent (3 days) and no increase in the yield was observed.
Finally, through a flow-chemistry approach, we further
reduced reaction times from 1.5 h to 10 min and additionally
increased the yields. This process could be easily scaled up by
simply letting the instrument run for longer time and using a
larger base-containing column, without any intervention
from the operator.16
SCHEME 2. Schematic Representation of the Cycloaddition
Reaction in Flow
NaHCO3 and K2CO3.15 Initially, the reaction was per-
formed using a 0.25 M solution of 2 (0.25 mmol) in EtOAc
(1 mL) and an equimolar solution of chloroxime 1a (Table 4,
entry 1). The solutions were introduced into the flow stream
(EtOAc) through two injection loops, mixing was achieved
with a simple T-piece, and the combined output was then directed
through a glass column containing the base (4 equiv), heated at
80 °C, in which the reaction took place (Scheme 2). The total flow
rate was set in a way that the residence time of the reagents in the
column was 30 min. The final flow line was then collected and
evaporated to give the crude material, which was purified by
column chromatography. A positive system pressure was main-
tained by using an in-line 100 psi back-pressure regulator.
All of the PS-bases turned out to be inefficient in the
cycloaddition reaction, since the product 3 was isolated in
unacceptable yield (<10%). Only solid K2CO3 gave a good
result (50% yield), and so it was selected for the subsequent
optimization phase. A rapid screening of reaction param-
eters was performed, which included reaction temperature,
residence time (flow rate), and stoichiometry.
As shown in Table 4, under continuous-flow conditions it was
possible to obtain cycloadduct 3a in good yield (69%, entry 5) in
only 10 min using 1.5 equiv of chloroxime 1a (0.37 M in EtOAc).
This represents a significant improvement over the above-
described microwave-assisted methodology, since the reaction
time was considerably reduced from 1.5 h to 10 min with a
comparable satisfactory yield (69% vs 66%). Increasing the
temperature to 100 °C was detrimental for the reaction yield
(Table 4, entry 7 vs 5).
Our analysis clearly emphasizes the power of the flow-
based technique and the unparalleled advantages of its
application to the cycloaddition chemistry, in terms of time,
yield, automation, and scaling-up opportunities. In perspec-
tive, the combination of microwave-assisted and flow-based
methodologies could enable the achievement of further
progresses in the field of cycloaddition chemistry.
Experimental Section
General Procedure for the Cycloaddition Reaction. To a solu-
tion of N-Boc-Δ3-pyrroline 2 (169 mg, 1.0 mmol) in EtOAc (5 mL)
were added solid NaHCO3 (420 mg, 5.0 mmol) and the proper chlo-
roxime 1 (1a-f, 4 equiv added portionwise: 1 equiv the first day,
0.5 equiv/day for the next six days). The mixture was vigorously
(15) (a) Hopkin, M. D.; Baxendale, I. R.; Ley, S. V. Chem. Commun. 2010, 46,
2450–2452. (b) Kirschning, A.; Solodenko, W.; Mennecke, K. Chem.;Eur. J.
2006, 12, 5972–5990. (c) Wiles, C.; Watts, P.; Haswell, S. J. Tetrahedron 2004, 60,
8421–8427. (d) Baxendale, I. R.; Deeley, J.; Griffiths-Jones, C. M.; Ley, S. V.;
Saaby, S.; Tranmer, G. K. Chem. Commun. 2006, 2566–2568.
(16) (a) Baxendale, I. R.; Ley, S. V.; Smith, C. D.; Tamborini, L.; Voica,
A. F. J. Comb. Chem. 2008, 10, 851–857. (b) Baumann, M.; Baxendale, I. R.;
Ley, S. V.; Smith, C. D.; Tranmer, G. K. Org. Lett. 2006, 8, 5231–5234. (c)
Tamborini, L.; Conti, P.; Pinto, A.; De Micheli, C. Tetrahedron: Asymmetry
2010, 21, 222–225.
J. Org. Chem. Vol. 75, No. 21, 2010 7441

Castellano et al.
JOCNote
stirred at room temperature for 7 days. Water was added to the
reaction mixture, and the organic layer was separated. The aqueous
layer was further extracted with EtOAc, and the combined organic
phase was dried over anhydrous sodium sulfate. The solvent was
evaporated, and the crude material was purified by silica gel
chromatography (cyclohexane/EtOAc 7:3) to give the correspond-
ing cycloadduct 3a-f.
General Procedure for the Microwave-Assisted Cycloaddition
Reaction. The proper chloroxime (1a-f, 1.0 mmol) was dis-
solved in EtOAc (5 mL) at room temperature in a 10 mL CEM
pressure vessel equipped with a stirrer bar. N-Boc-Δ3-pyrroline 2
(169 mg, 1.0 mmol), solid NaHCO3 (336 mg, 4.0 mmol), and H2O
(50 μL) were added, and the vial was sealed and heated in a CEM
Discover microwave synthesizer to 80 °C (measured by the
vertically focused IR temperature sensor) for 30 min. The reaction
cycle was repeated another two times, each time adding additional
0.5 equiv of chloroxime 1 (2 equiv total). The solid was filtered off,
and the solution was dried over anhydrous sodium sulfate. The
solvent was evaporated, and the crude material was purified by
silica gel column chromatography (cyclohexane/EtOAc 7:3) to
give the corresponding cycloadduct 3a-f.
prepared. The two reactant streams were mixed using a simple
T-piece and delivered to a glass column (6.6 mm i.d. by 100 mm
length) filled with K2CO3 (4.0 mmol, 540 mg) heated at 90 °C at
a total flow rate of 0.1 mL min-1, equating to a residence time of
about 10 min. A 100 psi backpressure regulator was applied to
the system. The solvent was evaporated, and the product was
purified by silica gel column chromatography (cyclohexane/
EtOAc 7:3) to give the corresponding cycloadduct 3a-f.
tert-Butyl 3-Phenyl-6,6a-dihydro-3aH-pyrrolo[3,4-d ]isoxazole-
1
5(4H)-carboxylate (3a). White solid; mp 146-147 °C (dec). H
NMR (300 MHz, CDCl3): δ 1.40 (s, 9H); 3.51-3.70 (m, 3H); 3.90
(d, J = 12.6 Hz, 1H); 4.15-4.25 (m, 1H); 5.28 (ddd, J = 1.1; 5.5;
9.3 Hz, 1H); 7.35-7.45 (m, 3H); 7.55-7.65 (m, 2H). 13C NMR
(75.5 MHz, CDCl3): δ 28.5; 49.8; 51.6; 53.7; 80.4; 85.7; 127.1; 128.5;
129.2; 130.5; 154.3; 158.1.
Acknowledgment. We are very grateful to Prof. Carlo De
Micheli for helpful discussions. This work was partially sup-
ꢀ
ported by grants from Universita di Salerno, Italy (G.S.) and
Universita degli Studi di Milano, Italy (P.C.).
ꢀ
General Procedure for the Cycloaddition Reaction Using the
R2þ/R4 Flow Reactor. A 0.25 M solution of N-Boc-Δ3-pyrro-
line 2 (1.0 mmol) in EtOAc (4 mL) and a 0.37 M solution of the
proper chloroxime 1a-f (1.5 mmol) in EtOAc (4 mL) were
Supporting Information Available: General experimental
procedures, characterization data, and copies of 1H and 13C
NMR spectra of synthesized compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
7442 J. Org. Chem. Vol. 75, No. 21, 2010