Diazo transfer reaction to 1,3-dicarbonyl compounds with sulfonyl azides catalyzed by molecular sieves

Article abstract of DOI:10.1016/j.molcata.2014.02.011

A simple and effective heterogeneous catalyst based on zeolite-type materials has been developed for the diazo transfer reaction involving 1,3-dicarbonyl compounds and tosyl azide. α-Diazo carbonyl compounds were obtained under mild conditions in good to high yields using commercial molecular sieve 4A or analogues as the catalyst. The best catalyst was found to be 4A-1000, a synthetic potassium-free nepheline obtained by heating molecular sieve 4A at 1000 C. Characterization of the resulting aluminosilicate by XRD, FTIR and SEM-EDS analysis confirmed the change of the crystal structure. Besides being nontoxic and inexpensive, the heterogeneous catalyst was readily removed by filtration and could be reused at least for four runs without any special treatment.

Full text of DOI:10.1016/j.molcata.2014.02.011

Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Diazo transfer reaction to 1,3-dicarbonyl compounds with sulfonyl
azides catalyzed by molecular sieves
a
a
b
a,∗
Luiz G. Dutra , Cristine Saibert , Denice S. Vicentini , Marcus M. Sá
a
Departamento de Química, Universidade Federal de Santa Catarina, Florianópolis 88040-900, SC, Brazil
Poliinova Pesquisa e Desenvolvimento Tecnológico Ltda, Rua Patrício Antônio Teixeira, 317, sala 408A, Jardim Carandai, Bigua c¸ u 88160-000, SC, Brazil
b
a r t i c l e i n f o
a b s t r a c t
Article history:
A simple and effective heterogeneous catalyst based on zeolite-type materials has been developed for
the diazo transfer reaction involving 1,3-dicarbonyl compounds and tosyl azide. ␣-Diazo carbonyl com-
pounds were obtained under mild conditions in good to high yields using commercial molecular sieve
Received 24 December 2013
Accepted 8 February 2014
Available online 18 February 2014
4
A or analogues as the catalyst. The best catalyst was found to be 4A-1000, a synthetic potassium-free
◦
nepheline obtained by heating molecular sieve 4A at 1000 C. Characterization of the resulting alumi-
nosilicate by XRD, FTIR and SEM-EDS analysis confirmed the change of the crystal structure. Besides
being nontoxic and inexpensive, the heterogeneous catalyst was readily removed by filtration and could
be reused at least for four runs without any special treatment.
Keywords:
Nepheline
Molecular sieve
Heterogeneous catalysis
␣
-Diazo carbonyl
© 2014 Elsevier B.V. All rights reserved.
Diazo transfer
1
. Introduction
-Diazo carbonyl compounds are useful pharmaceutical build-
cation issues, the methods for their preparation can be complex
and/or the cost of the reagent is high when compared to tosyl azide.
␣
The basic catalyst also requires particular consideration, since
it is very often a tertiary amine (triethylamine, DABCO, DBU) or
an alkaline hydroxide/carbonate (NaOH, K CO , Cs CO ) used in
ing blocks since they can undergo a wide variety of chemical
transformations such as cyclopropanation, Wolff rearrangement,
2
3
2
3
1
,3-dipolar cycloaddition, and X–H insertion (X = C, N, O, S), often
excess [5–7,13,14]. The need for an aqueous work-up to separate
out the water-soluble bases from the reaction mixture, besides this
being a tedious process, precludes any practical recovery of the
catalyst and thus represents a main source of residue generation.
The increase in environmental consciousness worldwide has led
to a considerable interest in developing more efficient and cleaner
methods for chemical synthesis [15–18]. Among them, the use of
heterogeneous catalysis (in which the active sites are supported
on an insoluble solid) offers many advantages over the use of con-
ventional homogeneous catalysis, even in industrial scale [19–22].
Besides the generally higher reaction ratio and selectivity observed
for the heterogeneous process, the insoluble catalysts are easily
recovered by filtration and thus can be reused instead of being
disposed of as waste. In spite of these benefits, the use of basic
inorganic solids as heterogeneous catalysts for the diazo transfer
reaction is so far limited to one report, wherein natural clays pre-
treated with NaOH furnished the ␣-diazo carbonyl compounds in
good yields, although the reaction was slow (1–3 days) for all but
one case [23]. Therefore, the development of more effective hetero-
geneous catalysts for basic catalysis is of considered interest. In this
regard, molecular sieves are particularly useful since they are inex-
pensive and can be modified to attain the desired basic properties
[24–29].
under mild conditions [1–3]. The transfer of the diazo group to
an active methylene compound is one of the most prevalent syn-
thetic approaches to the ␣-diazo carbonyl scaffold [4,5]. The diazo
transfer reaction to 1,3-dicarbonyl compounds usually involves a
combination of a sulfonyl azide and a base in a given organic sol-
vent to furnish 2-diazo-1,3-dicarbonyl derivatives in good yields
(Scheme 1). para-Toluenesulfonyl azide (tosyl azide, TsN ) is com-
3
monly used as the transfer reagent due to its easy accessibility
TsN3 is commercially available or can be readily prepared from
(
tosyl chloride) and relatively high reactivity among sulfonyl azides.
In spite of these advantages, the reaction using tosyl azide is fre-
quently associated with problems related to the isolation of the
desired diazo product as an equivalent mixture with tosyl amide
(
TsNH ), which is the byproduct of the diazo transfer using TsN . To
2
3
overcome the drawbacks associated with the use of tosyl azide and
its byproduct tosyl amide, several modified diazo transfer reagents
have been reported [6–12]. However, some present similar purifi-
∗ Corresponding author. Tel.: +55 48 37216844; fax: +55 48 37216850.
E-mail address: marcus.sa@ufsc.br (M.M. Sá).
http://dx.doi.org/10.1016/j.molcata.2014.02.011
1
381-1169/© 2014 Elsevier B.V. All rights reserved.

3
6
L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
O
O
O
O
Base
R1
R2
+ TsN
3
R1
R2
+ TsNH
2
Solvent
N2
Scheme 1. Diazo transfer reaction of tosyl azide and 1,3-dicarbonyl compounds.
(
a)
b)
As a continuation of our research interest in synthetic trans-
formations involving ␣-diazo carbonyl compounds [30–32] and
heterogeneous catalysis with molecular sieves [33–35], we report
herein a general and more benign approach to the preparation of
(
␣
-diazo carbonyl compounds 2 through the diazo transfer reaction
catalyzed by recyclable molecular sieves.
2
. Experimental
1
0
20
30
40
50
60
70
80
Position (2θ)
2.1. General
Fig. 1. XRD patterns for MS 4A (a) and 4A-1000 (b).
The FTIR spectra for the samples were recorded on a Bruker
−
1
spectrophotometer in the 4000–600 cm
range with 32 scans.
1
3
C NMR (100 MHz, CDCl ): ı 185.4 (C), 160.1 (C), 149.3 (C), 142.4
3
The samples were prepared in KBr pellets. The X-ray diffrac-
tograms were measured in the angular range of 2ꢀ = 5–80 , using a
(C), 129.1 (2 × CH), 122.8 (2 × CH), 61.8 (CH ), 14.0 (CH ).
2
3
◦
Philips X’Pert diffractometer equipped with a copper tube (Cu K␣,
ꢁ
3. Results and discussion
= 1.54056 A˚ ). Scanning electron microscopy (SEM) images were
acquired with a Jeol JSM-6701F microscope. The elemental anal-
ysis was based on the scan area of the SEM and the identification
was carried our by energy dispersive X-ray spectroscopy (EDS) at an
accelerating voltage of 15 kV and magnification of 13000×. Samples
were prepared as suspensions in ultrapure water (1 g/L) applying
sonication at 500 W for 20 min. The suspensions were added drop-
In the search for a catalyst of basic character that is able to medi-
ate the diazo transfer reaction of model substrates (see Section
.2) we selected ten molecular sieves and zeolite-type derivatives
for this study, comprising four commercially available (3A, 4A, 5A,
3X), two exchanged with potassium ion (4A/KCl, 13X/KCl) [34,35]
and four thermally-treated (4A-1000, 4A/KCl-1000 and 13X-1000
3
1
◦
wise to aluminum stubs, dried at 70 C and sputter-coated with
gold.
◦
◦
heated at 1000 C, and 4A-600 heated at 600 C) materials. The
treatment of MS with a KCl solution causes the partial replacement
of Na ions with K ions, which is usually associated with the enhance-
ment of the basic character due to an increase in the negative charge
over the peripheral oxygen atoms [24]. Another strategy to mod-
ify zeolite-type materials involves conventional heating at high
The molecular sieves (MS) 3A, 4A, 5A and 13X were com-
mercially available and obtained from Sigma-Aldrich. Potassium-
exchanged MS 4A/KCl and 13X/KCl were prepared by the
diffusion method using aqueous KCl as previously reported [34,35].
Thermally-modified 4A-600, 4A-1000, 4A/KCl-1000 and 13X-1000
samples were prepared by heating the corresponding MS in an elec-
◦
temperatures (600 or 1000 C), which induces various structural
changes in the crystalline materials, including phase transitions
and amorphization which can lead to enhanced properties or even
novel features.
◦
tric furnace at 1000 C (for 4A-1000, 4A/KCl-1000 and 13X-1000) or
◦
at 600 C (for 4A-600) for 1 h. After the samples had been cooled, the
final solids were transferred to a desiccator and stored for months
without any significant loss of activity.
3.1. Characterization of catalysts
2.2. General procedure for the diazo transfer reaction
The analysis of the X-ray diffraction (XRD) data for MS 4A and
4
A-1000 (Fig. 1) revealed that changes occurred in the crystalline
◦
To a solution of the 1,3-dicarbonyl compound 1 (1.00 mmol) and
system of 4A heated at 1000 C. The crystalline arrangement of
cubic zeolite 4A was modified to a hexagonal mode, which was
identified as the nepheline mineral (JCPDS card N. 88-1231) [40].
p-toluenesulfonyl azide (0.197 g, 1.00 mmol) in THF (2.0 mL) was
◦
added the MS (0.300 g) and the mixture was stirred at 25 C for the
time stated in Table 2. After the reaction was completed (monitored
by TLC [8:2 hexane:EtOAc] by following the consumption of the
starting 1,3-dicarbonyl compound 1), the catalyst was separated by
Nepheline [(Na,K)AlSiO ] is a natural mineral comprised of alu-
4
minosilicate, sodium and potassium which is closely related to
feldspar [41,42]. Nepheline has been previously synthesized by
heating 4A-type zeolites in an electric furnace or under microwave
filtration followed by sequential washing with 10 mL CH Cl2 and
2
◦
1
0 mL EtOAc. The filtrate was concentrated under reduced pres-
irradiation at around 1000 C [43–45].
sure and the residue was taken up in ethyl ether. The mixture was
concentrated again and the final residue was triturated with hex-
Energy-dispersive X-ray spectroscopy (EDS) analysis furnished
the chemical composition for representative 4A-type materials. The
mineral obtained by treating MS 4A at 1000 C (4A-1000) showed
◦
ane (3 × 20 mL). The solid material formed (TsNH ) was filtered and
2
the solvent was removed under reduced pressure to give the diazo
carbonyl compound 2 in >95% purity. Diazo compounds 2a–d, f–k
are known and their spectroscopic characterizations were in agree-
ment with published data [7,12,32,36–39]. Spectroscopic data for
the novel ethyl 2-diazo-3-(4-nitrophenyl)-3-oxopropanoate (2e):
colorless oil; IR (neat): ꢂmax = 3110, 2985, 2147, 1718, 1629, 1523,
essentially the same composition of Si and Al found for the com-
mercial 4A used as the precursor (Fig. 2a and b). A small decrease
(<1.5%) in the oxygen content was observed for the 4A-1000 com-
position. The Si/Al ratio of 1 was maintained, which means that
the basic character of the material may have been retained due to
−
the high number of available Al–O sites. In contrast to the natu-
−
1 1
1
314, 745 cm ; H NMR (400 MHz, CDCl ): ı 8.27 (d, J = 8.6 Hz, 2H),
ral nepheline, 4A-1000 is potassium-free and also absent of other
common metallic impurities such as iron. Therefore, 4A-1000 can
3
7
.74 (d, J = 8.6 Hz, 2H), 4.24 (q, J = 7.0 Hz, 2H), 1.26 (t, J = 7.0 Hz, 3H);

L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
37
Table 1
Optimizing the reaction conditions for the diazo transfer reactiona.
O
O
TsN3
O
O
Catalyst
Solvent, 25 o
OEt
OEt + TsNH
N2 2a
2
C
1a
No.
Molecular sieve
Solvent
Time (h)
Conversionb (%)
1
2
3
4
5
6
7
8
9
3A
4A
5A
4A
13X
4A
4A
4A
4A
4A
4A
4A
4A
4A
4A
4A/KCl
13X/KCl
4A-1000
4A-1000
4A-600
4A/KCl-1000
13X-1000
EtOAc
EtOAc
Acetone
Acetone
THF
THF
CH3CN
DMF
55
47
50
48
24
48
48
50
50
48
45
48
50
52
52
24
45
6
53
72
14
85
30
94
91
93
10
70
75
36
93
96
91
37
35
95
90
24
34
33
H2O
Isopropanol
Hexane
1
1
1
1
1
1
1
1
1
1
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
CH2Cl2
c
c
c
THF
THF
THF
CH3CN
THF
10
46
6
THF
THF
6
a
Conditions: ethyl acetoacetate (1a; 1.00 mmol), TsN3 (1.00 mmol), MS (300 mg),
solvent (2.0 mL).
b
Conversion (%) was determined by 1H NMR integration (400 MHz, CDCl3) of the
crude mixture.
Fig. 2. EDS data for 4A (a), 4A-1000 (b) and 4A/KCl-1000 (c).
c
Solvent-free conditions; mass catalyst (mg) per substrate 1a (mmol): 300 (entry
13), 200 (entry 14) and 100 (entry 15).
be considered as a synthetic nepheline, but with improved purity.
The EDS analysis was also useful for confirming the introduction of
potassium by Na/K-exchange in the 4A/KCl-1000 sample, at levels
comparable to those found in natural nepheline (Fig. 2c).
As expected for the 4A-1000 sample, the bands at around 3440
−
1
and 1668 cm , which are related to the O–H bond, disappeared or
dramatically diminished its intensity. This observation supports the
possibility that at high temperatures the Brønsted acid sites from
the crystal lattice undergo dehydroxylation (loss of O–H groups)
with the release of water and the consequent formation of Lewis
acid sites [48]. In fact, the most prominent band for the 4A-1000
The FTIR spectrum of MS 4A showed the characteristic bands of
−
1
the hydrated zeolite (Fig. 3a). The large band at around 3440 cm
is attributed to the combination of symmetric and asymmetric OH
−
1
oscillation of the water molecule. The band at 1668 cm
refers
to the interaction of the OH group with the oxygen in the zeo-
lite structure [46,47]. Also, the bands centered around 1000 and
−
1
sample, at around 1000 cm , was related to the asymmetrical
stretching vibration of Si–O–Si and Si–O–Al modes of the inor-
ganic framework [47,49]. Finally, the appearance of a shoulder in
−
1
6
68 cm are attributed, respectively, to the asymmetric and sym-
metric vibration bands of the TO4 mode (T = Al or Si), which are
characteristic of this zeolite [45,47].
−
1
−1
1104 cm and a relatively strong band at around 714 cm can be
related to the displacement of the original bands of asymmetric and
symmetric vibration stretching (TO mode) to longer wavelengths.
4
The SEM images show that the morphology of MS 4A (Fig. 4a)
consists of cubic particles with sizes ranging from 2 to 5 ␮m. How-
ever, as supported by the XRD analysis (see Fig. 1) the particle
morphology was significantly altered from a cubic shape for 4A to
a nearly spherical shape for 4A-1000 (Fig. 4b). These spherical-type
particles were also observed for the Na/K-exchanged 4A/KCl-1000
(a)
(
1
(
Fig. 4c), as well as for 4A-1000 after four consecutive uses (4A-
000 × 4, Fig. 4d) verifying the recycling potential of this material
see Section 3.2). This morphological alteration may be primarily
(b)
responsible for differences in the catalytic performance of the cat-
alyst due to the increase in the surface area of the particles on
changing from cubic to spherical shapes.
3.2. Catalyst profiles
Ethyl acetoacetate (1a) and TsN3 were selected as the model
3500
3000
2500
2000
1500
1000
substrates for screening a variety of molecular sieves as potential
catalysts for the diazo transfer reaction under different reaction
conditions (Table 1). The formation of the expected ␣-diazo-␤-
keto ester 2a was dependent on both the catalyst and the solvent
-1
Wavenumber (cm )
Fig. 3. FTIR spectra for 4A (a) and 4A-1000 (b).

3
8
L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
Fig. 4. SEM micrographs for MS 4A (a), 4A-1000 (b), 4A/KCl-1000 (c) and 4A-1000 × 4 (d).
used. Clearly, MS 4A was the most efficient catalyst with all solvents
tested (compare entries 1 and 2; 3 and 4; 5 and 6). In combination
with 4A, THF and acetonitrile were equally useful solvents for the
reaction (entries 6 and 7), although DMF (entry 8) also led to >90%
conversion. However, the latter was not selected for further stud-
ies due to environmental concerns and toxicity issues [50]. With
regard to the other classes of solvents tested, neither the hydrox-
ylic (entries 9 and 10) nor the less polar (entries 11 and 12) solvents
gave satisfactory results. On the other hand, solvent-free trans-
formations were successfully achieved (entries 13–15), even with
reduced quantities of the catalyst, although for small-scale reac-
tions it was difficult to control the stirring due to the heterogeneous
nature of the medium. Therefore, THF (in minimum quantities) was
chosen as the preferred solvent for further developments.
catalytic profile. The results in Table 2 were generated from reac-
tions run on the mmol scale, although scaling up to gram quantities
provided comparable results.
In all cases, the use of 4A-1000 furnished better isolated yields
and shorter reaction times compared to 4A. In general, ␤-keto esters
are reactive substrates for the diazo transfer reaction using 4A or
4A-1000, with a clear preference for the latter. Remarkably, in one
case (the preparation of the 4-nitrobenzoyl-substituted diazo ester
1e) the use of 4A-1000 was successful while the commercial 4A
was found to be completely inactive. The enhanced reactivity of
methyl over ethyl ester was evidenced by comparing ␤-keto ester
1b with 1a and, much more dramatically, the malonate 1f with
1g. While this pronounced difference in reactivity may be initially
attributed to steric effects (the ethyl group being larger than the
methyl group), the preparation of dibenzyl diazomalonate (1h) in
reasonable yields clearly contradicts this hypothesis. Instead, the
difference in the pKa of the active ␣-methylene group and the rela-
tive ability to coordinate to Lewis acid sites from the catalyst should
be considered (see discussion below).
Although the use of KCl-exchanged molecular sieves, such as
4
A/KCl and 13X/KCl, has been associated with enhanced basic
properties [24,35], in this study both catalysts led to low conver-
sions (<40%) even after prolonged periods (Table 1, entries 16 and
1
7). On the other hand, the reaction efficiency was considerably
improved on using 4A-1000 as the catalyst, which is a potassium-
free nepheline obtained by heating MS 4A at 1000 C. In these cases,
the conversion was almost completed after only 6 h in THF (entry
Cyclic substrates also worked well and gave the corresponding
diazo product from the diester (1i) and the diketone (1j). Similarly,
1,3-pentanodione (1k) was very reactive and could be entirely con-
verted to the expected diazo diketone 2k, although the isolated
yield was only moderate due to mass loss during the work-up.
This may be ascribed, in part, to the ability of some 1,3-dicarbonyl
compounds to become associated with the MS, possibly through
coordination with the Al sites as represented in Scheme 2. Control
experiments in the absence of TsN3 showed that the loss of the
1,3-dicarbonyl substrate during the catalyst removal is dependent
on both the type of MS and the solvent used in the filtration step.
Therefore, it is advisable to wash the catalyst thoroughly with two
different solvents of medium polarity, such as dichloromethane and
ethyl acetate (see Section 2).
The presence of a diversity of functional groups is well tolerated,
including chloro, azido and nitro groups (Table 2, entries 1c, 1d and
1e, respectively). The preparation of ␥-halo-␣-diazo-␤-keto esters
is of particular interest as they are useful building blocks for the syn-
thesis of a great variety of multifunctionalized compounds [30–32].
However, the prior synthesis of a ␥-bromo-␣-diazo-␤-keto ester
derivative involved potentially toxic mercury-based reagents and
the highly sensitive bromoacetyl bromide [32], which restricts
◦
1
8) or 10 h in acetonitrile (entry 19). Furthermore, the separation
of the catalyst by filtration was much more convenient with the
use of 4A-1000 than with commercial 4A due to the tendency of
4
A-1000 to be in the form of large aggregates that are more easily
retained by regular paper filter than the powder-like 4A. The higher
activity observed for 4A-1000 was unexpected because nepheline-
type materials are generally of no interest as catalysts due to their
lack of porosity [47]. The application of temperature during the
preparation of the catalysts was found to greatly affect their acti-
◦
vation, since heating 4A at 600 C produced a much poorer reaction
promoter (4A-600, entry 20). As anticipated, prior K/Na-exchange
◦
followed by heating at 1000 C also failed to furnish a catalyst that
was able to promote high conversions to the expected product 2a
(
4A/KCl-1000, entry 21), and the same was true for another type of
◦
MS heated at 1000 C (13X-1000, entry 22).
Motivated by this preliminary study, the procedure using 4A-
1
000 as the catalyst for the diazo transfer reaction was extended to
a representative group of 1,3-dicarbonyl compounds. Commercial
MS 4A was also included in this study in order to compare their

L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
39
Table 2
Diazo transfer reactiona promoted by 4A and 4A-1000.
O
O
TsN3
O
O
Catalyst
THF, 25 o
R1
R2
R1
R2 + TsNH
2
2
C
1
N2
No.
Product
2
MS 4A
MS 4A-1000
Time (h)
Time (h)
Yield (%)
80
Yield (%)
O
O
O
1
1
1
1
a
b
c
OEt
25
6
6
95
81
80
76
N2
O
OMe
O
70
75
64
5
N
2
O
O
Cl
OEt
24
46
18
26
N
2
O
N3
d
OEt
N
2
O
O
OEt
0b
1
e
40
26
83
N2
2
O N
O
O
1
1
1
f
MeO
EtO
OMe
27
80
25
80
46
94
12
73
N
2
O
O
g
h
OEt
50 [25]c
10 [15]c
N2
O
O
O
O
80
48d
N
2
O
O
O
1
i
24
51d
22
68
O
N
2
1
1
j
20
5
75
65
12
4
82
69
O
O
O
N
2
O
k
N
2
a
b
c
◦
Conditions: 1 (1.00 mmol), TsN3 (0.197 g, 1.00 mmol), catalyst (0.300 g), THF (2.0 mL), 25 C.
Recovery of >90% starting material 1.
◦
Numbers in brackets refer to the results obtained under conventional heating at 65 C.
d
1
Under these conditions the compound was obtained in ca. 50% conversion (by H NMR).
its broader application. Accordingly, a one-step synthesis of the
related ␥-chloro-␣-diazo-␤-keto ester 2c from the commercially
available keto ester 1c in gram scale is now feasible via an MS-
catalyzed diazo transfer reaction (Table 3). As expected, the use of
extensive formation of water-soluble ammonium salt 3 through
chlorine displacement from 2c (Table 3). The nucleophilic character
of the base could be severely diminished by replacing triethylamine
with DIPEA (N,N-diisopropylethylamine). Although better results
were achieved with the bulkier base, in this case the use of an
organic catalyst required an aqueous work-up, which was not only
tedious but led to acyl cleavage of the diazo product 2c to give ethyl
4
A and 4A-1000 led to 2c in high yields under mild conditions. It is
worth noting that, in our study, the conventional method using tri-
ethylamine as the catalyst led to unsatisfactory results due to the

4
0
L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
Table 4
O
O
O
O
Recovery of the 4A-1000 catalyst for the preparation of diazo ester 2a.
R1
R2
R1
R2
2
Entry
Recycle number
Time (h)
Yield (%)
1
2
3
4
5
Fresh
6
6
7
8
8
95
92
84
75
67
1
N2
Recycle 1
Recycle 2
Recycle 3
Recycle 4
Na+ -O
Al
δ-
Al
δ-
Na+
δ-
Na+ -
O
HO
Al
the case of the diacid 1m, the presence of acidic protons might lead
to extensive inactivation of the basic sites in the catalyst. However,
no reasonable explanation was so far been found for the complete
lack of reactivity of the trifluoromethyl-substituted 1,3-diketone
1n, since other diketones (Table 2, 1j and 1k) and aryl-containing
substrates (1e and 1h) were readily transformed to the correspond-
ing diazo products 2.
δ-
δ-
δ-
O
O
O
O
R1
R2
R1
R2
δ-
δ-
N2+
TsN3 TsNH2
H
Scheme 2. Plausible mechanism for the diazo transfer reaction catalyzed by 4A-
000.
Catalyst reuse is a highly desirable feature in industrial pro-
cesses and thus we turned our attention to the possibility of reusing
1
4
A-1000 as the catalyst for the diazo transfer reaction with ethyl
Table 3
acetoacetate (1a) and TsN3 under the conditions given in Table 1.
Thus, after the first reaction cycle the catalyst was separated by fil-
tration, washed with CH Cl2 and EtOAc, dried in an oven (100 C)
and used in the next run at least twice without appreciable loss of
catalytic activity (Table 4). However, it was necessary to progres-
sively extend the reaction time in order to sustain the high product
yields obtained in the first two runs. Attempts to reestablish the
high catalyst activity by reheating a few recycled samples of the
Preparation of ␥-chloro-␣-diazo-␤-keto ester 2c from the diazo transfer reaction
with different catalysts (TEA = triethylamine; DIPEA = N,N-diisopropylethylamine).
◦
2
O
O
TsN3
Catalyst
THF, 25 o
O
O
Cl
Cl
OEt
OEt
C
1c
N2 2c
O
O
O
N+
Cl-
H
◦
4A-1000 catalyst at 1000 C did not give any improvement in the
efficiency.
OEt
OEt
N2
3
N2
4
The partial inactivation of the catalyst is possibly related, in part,
to the persistence of small quantities of substrate through coordi-
nation of the 1,3-dicarbonyl group with the Al sites which act as
Lewis acids (Scheme 2).
In combination, the observations described herein support the
dual role of the catalyst as a Lewis acid and also a Brønsted base
promoter. Of particular importance is the finding that in the case
Catalyst
Time (h)
Yield 2c (%)
TEA
DIPEA
20
21
24
18
<30
65
75
4
4
A
A-1000
80
◦
of the 4A-1000 catalyst the heating at 1000 C enhanced its perfor-
mance both as a Lewis acid and as a Brønsted base. Complexation
of the bidentate 1,3-dicarbonyl substrate at the acidic Al site is
facilitated by the concurrent deprotonation in the presence of a
Brønsted base such as the electron-rich oxygen atom in close prox-
imity (Scheme 2). Reaction of the substrate-catalyst complex with
the diazo transfer reagent followed by product release ultimately
led to the ␣-diazo carbonyl compound 2 with the regeneration of
the heterogeneous catalyst.
diazoacetate (4) as a side product, thus having a negative impact
on the yield.
The use of 4A-1000 allowed the facile separation of the het-
erogeneous catalyst from the reaction mixture, which is not only
economically advantageous and environmentally benign but also
bypass the aqueous work-up, which may cause purification issues.
After recovering the catalyst through the filtration and evapora-
tion of the solvents, it was easy to separate out the TsNH2 formed
as the byproduct, which is usually associated with difficulties
when the conventional work-up is employed [9,12–14]. Our sim-
ple and reproducible protocol involved the trituration of the crude
product mixture with ether followed by hexane (see Section 2),
which causes the precipitation of TsNH2 while the diazo product
4. Conclusions
Heterogeneous catalysts based on zeolite-type materials were
developed for the diazo transfer reaction from a variety of sub-
strates under mild conditions in good to high yields. Due to the
simplicity of the process, the work-up and purification steps were
also straightforward and provided high-quality diazo products in
the gram scale. Subtle changes in the substrate structure may lead
to considerable differences in the reactivity; however, the reactions
were generally carried out within 4–24 h when 4A-1000 (a syn-
thetic K-free nepheline) was used as the catalyst. In fact, heating
2
remains soluble and can be isolated in high purity.
Other substrates containing active methylene groups (Fig. 5)
were also tested in the diazo transfer reaction but were not reac-
tive under the stated conditions. The lack of reactivity for the cyano
ester 1l is probably related to the absence of a 1,3-dicarbonyl
moiety which excluded the postulated pathway involving the for-
mation of a catalyst-substrate coordination complex (Scheme 2). In
◦
the commercially available MS 4A at 1000 C did generate a more
efficient catalyst, not only enhancing its activity but allowing it to
be more easily separated from the reaction medium by filtration. In
addition, the reuse of 4A-1000 was found to be potentially viable
at least for a few runs without any special treatment. Besides being
useful due to its catalytic activity and overall efficiency, 4A-1000
is an aluminosilicate which is nontoxic and inexpensive. These are
O
O
O
O
O
F
NC
OEt
HO
OH
F
1
l
1m
F
1n
Fig. 5. Substrates that failed to react with TsN3 and 4A or 4A-1000 as the catalyst.

L.G. Dutra et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 35–41
41
important features to be considered in relation to the development
of more sustainable processes involving acid-base catalysis.
[19] S. Chassaing, A. Alix, T. Boningari, K.S.S. Sido, M. Keller, P. Kuhn, B. Louis, J.
Sommer, P. Pale, Synthesis (2010) 1557–1567.
[
[
20] H.-J. Federsel, Acc. Chem. Res. 42 (2009) 671–680.
21] A. Corma, H. Garcia, Chem. Rev. 103 (2003) 4307–4365.
Acknowledgments
[22] Y. Ono, T. Baba, Catal. Today 38 (1997) 321–337.
[
[
[
23] R.S. Rianelli, M.C.B.V. Souza, V.F. Ferreira, Synth. Commun. 34 (2004) 951–959.
24] M. Wallau, U. Schuchardt, J. Braz. Chem. Soc. 6 (1995) 393–403.
25] S.E. Sen, S.M. Smith, K.A. Sullivan, Tetrahedron 55 (1999) 12657–12698.
The authors wish to thank the Central de Análises (Departa-
mento de Química, UFSC, Florianópolis) for spectroscopic analysis
and the support of Laborat o´ rio Central de Microscopia Eletr oˆ nica
[26] G. Sartori, R. Ballini, F. Bigi, G. Bosica, R. Maggi, P. Righi, Chem. Rev. 104 (2004)
99–250.
1
[
[
27] Z. Hell, A. Cwik, Z. Finta, Z. Horváth, J. Mol. Catal. A: Chem. 184 (2002) 191–195.
28] M. Adinolfi, G. Barone, A. Iadonisi, M. Schiattarella, Tetrahedron Lett. 44 (2003)
4661–4663.
(
LCME-UFSC) for the SEM-EDS analysis. L.G.D. and C.S. are grate-
ful to CAPES (Coordena c¸ ão de Aperfei c¸ oamento de Pessoal de
Nível Superior) and CNPq (Conselho Nacional de Desenvolvi-
mento Científico e Tecnológico) for academic grants. M.M.S. is
grateful to CNPq for research fellowship. Financial support from
INCT-Catalysis/CNPq (Instituto Nacional de Catálise de Sistemas
Nanoestruturados, Brazil) is also gratefully acknowledged.
[29] K.P.R. Kartha, B. Mukhopadhyay, R.A. Field, Carbohydr. Res. 339 (2004)
29–732.
7
[
30] M.M. Sá, G.P. Silveira, A.J. Bortoluzzi, A. Padwa, Tetrahedron 59 (2003)
5441–5447.
[31] A. Padwa, M.M. Sá, J. Braz. Chem. Soc. 10 (1999) 231–236.
[
[
[
32] A. Padwa, M.M. Sá, M.D. Weingarten, Tetrahedron 53 (1997) 2371–2386.
33] T.B. Bisol, A.J. Bortoluzzi, M.M. Sá, J. Org. Chem. 76 (2011) 948–962.
34] M.M. Sá, L. Meier, L. Fernandes, S.B.C. Pergher, Catal. Commun. 8 (2007)
1625–1629.
References
[
[
35] M.M. Sá, L. Meier, Synlett (2006) 3474–3478.
36] M.C. Bagley, R.T. Buck, S.L. Hind, C.J. Moody, J. Chem. Soc., Perkin Trans. 1 (1998)
[
[
[
[
[
[
[
[
1] G. Maas, Angew. Chem. Int. Ed. 48 (2009) 8186–8195.
2] Y. Zhang, J. Wang, Chem. Commun. (2009) 5350–5361.
3] Z. Zhang, J. Wang, Tetrahedron 64 (2008) 6577–6605.
4] M. Regitz, Angew. Chem. Int. Ed. 6 (1967) 733–749.
5] J.B. Hendrickson, W.A. Wolf, J. Org. Chem. 33 (1968) 3610–3618.
6] D.F. Taber, R.E. Ruckle Jr., M.J. Hennessy, J. Org. Chem. 51 (1986) 4077–4078.
7] G.M. Green, N.P. Peet, W.A. Metz, J. Org. Chem. 66 (2001) 2509–2511.
8] V.V. Popic, S.M. Korneev, V.A. Nikolaev, I.K. Korobitsyna, Synthesis (1991)
5
91–600.
37] P. Wyatt, A. Hudson, J. Charmant, A.G. Orpen, H. Phetmung, Org. Biomol. Chem.
(2006) 2218–2232.
[
4
[
[
38] F. Nanteuil, J. Loup, J. Waser, Org. Lett. 15 (2013) 3738–3741.
39] A.M. Harned, W.M. Sherrill, D.L. Flynn, P.R. Hanson, Tetrahedron 61 (2005)
12093–12099.
[
[
[
[
40] V. Kahlenberg, H. Böhm, Am. Mineral. 83 (1998) 631–637.
41] F. Burat, O. Kangal, G. Onal, Miner. Eng. 19 (2006) 370–371.
42] S. Hansen, J. Fälth, Zeolites 2 (1982) 162–166.
43] Z. Pilter, S. Szabó, M. Hasznos-Nezdei, E. Pallai-Varsányi, Microporous Meso-
porous Mater. 40 (2000) 257–262.
1
95–198.
[9] M.K. Muthyala, S. Choudhary, A. Kumar, J. Org. Chem. 77 (2012) 8787–8791.
[
[
[
10] R.P. Wurz, W. Lin, A.B. Charette, Tetrahedron Lett. 44 (2003) 8845–8848.
11] J.L. Chiara, J.R. Suárez, Adv. Synth. Catal. 353 (2011) 575–579.
12] M. Kitamura, N. Tashiro, S. Miyagawa, T. Okauchi, Synthesis (2011)
[
[
[
[
44] T. Ohgushi, K. Ishimaru, S. Komarneni, J. Am. Ceram. Soc. 84 (2001) 321–327.
45] S. Markovic, V. Dondur, R. Dimitrijevic, J. Mol. Struct. 654 (2003) 223–234.
46] A. Adamczyk, E. Dlugon, Spectrochim. Acta, Part A 89 (2012) 11–17.
47] A. Caovilla, R.T. Rigo, F.G. Penha, S.B.C. Pergher, Quim. Nova 32 (2009)
1
037–1044.
[
[
[
[
13] J.C. Lee, J.Y. Yuk, Synth. Commun. 25 (1995) 1511–1515.
14] M. Presset, D. Mailhol, Y. Coquerel, J. Rodriguez, Synthesis (2011) 2549–2552.
15] J.L. Tucker, Org. Process Res. Dev. 14 (2010) 328–331.
1818–1821.
[
[
[
48] J. Weitkamp, Solid State Ionics 131 (2000) 175–188.
49] P. Innocenzi, J. Non-Cryst. Solids 316 (2003) 309–319.
50] K. Alfonsi, J. Colberg, P.J. Dunn, T. Fevig, S. Jennings, T.A. Johnson, H.P. Kleine, C.
Knight, M.A. Nagy, D.A. Perry, M. Stefaniak, Green Chem. 10 (2008) 31–36.
16] N.R. Candeias, L.C. Branco, P.M.P. Gois, C.A.M. Afonso, A.F. Trindade, Chem. Rev.
1
09 (2009) 2703–2802.
[
[
17] I.T. Horváth, P.T. Anastas, Chem. Rev. 107 (2007) 2169–2173.
18] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.