Preparation and use of polystyryl-DABCOF2: An efficient recoverable and reusable catalyst for β-azidation of α,β- unsaturated ketones in water

Article abstract of DOI:10.1002/adsc.201100705

Novel solid fluorides were prepared to optimize the β-azidation of α,β-unsaturated ketones. The higher loading of these catalysts compared to that of commercially available fluorides has allowed the use of a smaller mass of catalyst helping the mixing of

Full text of DOI:10.1002/adsc.201100705

FULL PAPERS
DOI: 10.1002/adsc.201100705
Preparation and Use of Polystyryl-DABCOF2: An Efficient
Recoverable and Reusable Catalyst for b-Azidation of
a,b-Unsaturated Ketones in Water
Tommaso Angelini,a Daniela Lanari,a Raimondo Maggi,b Ferdinando Pizzo,a
Giovanni Sartori,b and Luigi Vaccaroa,*
a
Laboratory of Green Synthetic Organic Chemistry, CEMIN – Dipartimento di Chimica, Universitꢀ di Perugia Via Elce di
Sotto, 8; I-06123 Perugia, Italy
Fax : (+39)-075-5855-560; e-mail: luigi@unipg.it
“Clean Synthetic Methodologies Group”, Dipartimento di Chimica Organica e Industriale dell’Universitꢀ, and Consorzio
b
Interuniversitario “La Chimica per l’Ambiente” (INCA), UdR PR2, Parco Area delle Scienze 17A, I-43124 Parma, Italy
Received: September 7, 2011; Revised: October 21, 2011; Published online: February 28, 2012
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201100705.
Abstract: Novel solid fluorides were prepared to op- prove the efficiency of our approach we have devel-
timize the b-azidation of a,b-unsaturated ketones. oped a protocol operating in a continuous-flow
The higher loading of these catalysts compared to manner that has allowed us to achieve an E-factor of
that of commercially available fluorides has allowed 1.7–1.9, with a reduction of ca. 80% of the corre-
the use of a smaller mass of catalyst helping the sponding batch conditions. The continuous-flow pro-
mixing of the reaction mixture. Porous polymeric tocol has allowed us to minimize the use of trime-
supports have proved to be more efficient in the thylsilyl azide making the recovery and reuse of
presence of water as reaction medium. Water has water and catalyst 5f very efficient and simple. Final-
played a crucial role showing a beneficial effect on ly, a novel reduction system using palladium on alu-
the reactivity by improving dispersion of the reaction mina (5 mol%) and equimolar amount of formic acid
mixture and also by avoiding organic fouling caused has been used in the presence of 1 equivalent of di-
by the retention of the reaction mixture within the tert-butyl pyrocarbonate to set a multistep protocol
polymeric matrix. This has facilitated the recovery of operating in continuous-flow conditions for the prep-
the products from the catalyst. The protocol reported aration of two representative N-Boc-b-amino ke-
has allowed a significant reduction in the organic sol- tones starting from the corresponding enones with
vent required for the complete recovery of the pure E-factors of 3.2 and 2.7, respectively.
product whilst leaving the catalyst clean and reusa-
ble. E-factors are in the range of 5.9–10.5 and there-
fore ca. 3 times smaller than previous procedures op- Keywords: azides; enones; green chemistry; polystyr-
erating under solvent-free conditions. To further im- ene-supported catalysts; solvent-free conditions
Introduction
useful approach for the selective introduction of a pri-
mary amino group.[3]
ThÀe conjugate addition of an N-nucleophile (e.g.,
Despite the large number of practical applications,
N3 ) to a,b-unsaturated ketones represents a well-es- efficient procedures for the preparation of b-azido
tablished method for the preparation of b-amino car- carbonyl compounds are currently limited and gener-
bonyl compounds[1] which are versatile intermediates ally they use a combination of sodium azide and a
for the synthesis of a variety of target molecules in- strong acid as hydrazoic acid source.[4] Miller et al. re-
cluding b-amino acids, pharmaceuticals, and natural ported an excellent protocol using a combination of
products.[2]
trimethylsilyl azide (TMSN3) and acetic acid as azido
Among all the different nucleophiles, addition of ion source and a tertiary amine as catalyst, developed
azido ion and subsequent reduction is certainly a in order to avoid the use of the highly toxic and ex-
plosive hydrazoic acid.[5] Further similar examples
908
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 908 – 916

Preparation and Use of Polystyryl-DABCOF2: An Efficient Recoverable and Reusable Catalyst
have been also reported[6] and this approach has been
Our intention is to exploit water as medium to in-
exploited by Miller et al. for the realization of an crease the reactivity of the system but also to improve
asymmetric procedure.[7]
dispersion of the reaction mixture. The use of a spe-
We have been interested in the use of organic am- cific porous solid support in water should avoid the
[8]
organic fouling[14] caused by the sticking of the reac-
À
À
monium fluorides for activating Si N and Si O
bonds and within this research we have reported an tion mixture within the polymeric matrix, allowing a
alternative protocol for the b-azidation of a,b-unsatu- better catalytic efficiency and facilitating the recovery
rated ketones based on the use of recoverable Am- of the products from the catalyst.
berlite IRA900F as catalyst under solvent-free condi-
tions (SolFC).[8c]
Our research program is committed to the optimi- Results and Discussion
zation of synthetic procedures by employing eco-
friendly reaction protocols, including the use of In this paper, we report our results on the design and
water,[9] SolFC,[10] and supported organocatalysts.[11]
preparation of a novel ammonium fluoride supported
We have investigated the use of a polymer-support- on polystyrene, and its application for the develop-
ed organocatalyst under SolFC proving that this is a ment of a chemically and environmentally efficient
very attractive approach to realize efficient synthetic procedure for the b-azidation of a,b-unsaturated ke-
procedures with minimal waste production.[8,11] In tones executable on a large scale and featuring a low
fact, while the adoption of SolFC is necessary to in- E-factor.
crease the efficiency of the polymer-supported recov-
erable catalyst, this approach imposes the use of
We have focused our attention on 1,4-diazabicyclo-
AHCTUNGTREGUN[NN 2.2.2]octane (DABCO, 1), an interesting diamine
strictly catalytic amounts of catalyst to avoid crucial moiety that, if supported on a polymer resin, may
issues relative to the mixing of equimolar amounts of carry a high-loading of fluoride as counter ion to two
reactants without employing a reaction medium. The ammonium functionalities (Table 1).
resulting protocol is intrinsically characterized by a
very high environmental efficiency.
Starting from different types of polystyrene (PS)-Cl
2, the DABCO moiety was introduced to form the
We have also been interested in useful applications corresponding mono-ammonium chlorides 3 that were
of continuous-flow chemistry.[12] Accordingly, to fur- subsequently transformed into the bis-ammonium io-
ther reduce the E-factor[13] we have also reported the dides 4 that, after treatment with potassium fluoride,
use of cyclic continuous-flow reactors operating in gave the desired bis-fluoride PS-DABCOF2 catalysts
SolFC or highly concentrated conditions running on a 5.
multigram scale.[8a,11a] Highlights of these protocols
Besides the experimental details available as Sup-
are the recovery of the product with a minimal porting Information, it should be pointed that the
amount of organic solvent, as well as recovery and conversion of the mono-ammonium salt 3 to the cor-
reuse of the solid catalyst without decreasing the effi- responding bis-ammonium iodide 4 with MeI in tolu-
ciency.
ene, has been performed by using a continuous-flow
In our previous report on the b-azidation of a,b-un- procedure, because under normal stirring (although
saturated ketones catalyzed by Amberlite IRA900F, very gentle) the solid porous catalyst was immediately
we have highlighted that the adoption of SolFC was pulverized making very difficult its manipulation.
necessary to achieve the maximum efficiency of the
All the polymer-supported ammonium salts 3, 4,
process. Although satisfactory E-factor values were and 5 have been fully characterized. The fluoride
achieved (ca. 22), we also noticed that in this process loadings for the catalysts 5a–f prepared are reported
the reaction mixture showed an evident tendency to in Table 1.
stick with the polymer resin. Therefore compared to
These catalysts have been tested in the representa-
other processes studied by our group, larger amounts tive reaction of (E)-hept-3-en-2-one (6a) and TMSN3
of organic solvent were needed to isolate the prod- (1.5 equivalents) by using different reaction media at
A
608C. The results are reported in Table 2.
We have decided to design a novel fluoride source
We were pleased to find that water has a beneficial
supported on polystyrene featuring a higher loading influence on the process allowing us to achieve better
compared to that of commercially available fluorides results than those obtained in the presence of an or-
in order to use a smaller mass of catalyst and help ganic medium (Table 2, entries 8 and 9).
mixing of the reaction mixture.
All the catalysts 5a–f efficiently promoted the con-
In addition to this and considering our experience version of 6a to 7a using water as reaction medium
in the use of water as reaction medium,[9] we have de- (Table 1, entries 1–6). Better results were obtained in
cided to prepare solid fluoride catalysts suitable for the case of catalysts 5d–f. (Table 1, entries 4–6) that
being used in the presence of water.
were prepared by using a porous support, confirming
Adv. Synth. Catal. 2012, 354, 908 – 916
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
909

Tommaso Angelini et al.
FULL PAPERS
Table 1. Preparation of PS-DABCOF2 catalysts 5.
Entry
Resin[a]
Catalyst (F mmolgÀ1)[b]
1
2
3
4
5
6
gel -type (1% DVB cl) 100–200 mesh
gel -type (2% DVB cl) 200–400 mesh
gel -type (2% DVB cl) 200–400 mesh
porous (5.5 DVB cl) 16–50 mesh
porous (5.5 DVB cl) 16–50 mesh
porous (5.5 DVB cl) 16-50 mesh
5a (2.42)[c]
5b (4.69)[c]
5c (6.29)[d]
5d (4.28)[c]
5e (6.20)[d]
5f (7.39)[d]
[a]
DVB cl =divinyl benzene cross-linked.
[b]
5 was prepared by washing 4 with aqueous KF in continuous-flow until after treating the eluted solution with AgNO3 no
precipitation of AgCl was observed. 4 was prepared by using 5 equiv. of MeI in toluene for 1 day at room temperature in
continuous-flow (see Supporting Information).
Conditions for the preparation of 3: 2 equiv. of DABCO (1), 3 days, 808C.
Conditions for the preparation of 3: 10 equiv. of DABCO (1), 5 days, reflux.
[c]
[d]
Table 2. b-Azidation of 6a using catalysts 5a–f.
Under SolFC the results were satisfactorily but the
conversion of 6a to 7a was slower than in the pres-
ence of water (Table 2, entry 7).
According to our previous results,[8a–c] the recovered
catalysts resulted from the complete substitution of
both fluoride ions by azido ions with formation of vol-
atile trimethylsilyl fluoride, b.p. 168C. Therefore in
the case of catalyst 5f, it was recovered as the corre-
sponding PS-DABCO(N3)2 that was still equally effi-
cient (see below).
Also iodide 4f was an effective catalyst but less effi-
cient than fluoride, and complete conversion was not
reached even after longer reaction time. This is proba-
bly due to the formation of trimethylsilyl iodide (by
azido/iodide exchange) that probably decomposes in
water to form HI which causes the hydrolysis of
TMSN3.
Entry
Catalyst
Medium[b]
Conversion [%][c]
1
2
3
4
5
6
7
8
5a
5b
5c
5d
5e
5f
5f
5f
5f
4f
5f
H2O
H2O
H2O
H2O
H2O
H2O
–
CH3CN
CH2Cl2
H2O
H2O[f]
89
90
92
>99[d]
>99[d]
>99[d]
91[e]
67
9
10
11
74
70
60
The pH of the reaction medium is ca. 3.0 at the be-
ginning of the process and becomes ca. 4.0 when the
conversion of 6a to 7a is complete. It should also be
noticed that the reaction does not proceed satisfacto-
rily when NaN3 was used as alternative source of
azido ion (Table 2, entry 11).
Therefore, in agreement to our previously reported
results,[8c] it can be concluded that the reaction pro-
ceeds according to the plausible mechanism depicted
in Scheme 1.
[a]
Referred to the amount of fluoride used.
0.5 mL for 1 mmol of 6a.
[b]
[c]
1
Conversion of 6a to 7a was measured by H NMR analy-
ses, the remaining material was the unreacted 6a.
91% of isolated yield of the pure product 7a (see Experi-
mental Section for further details).
Complete conversion to 7a was achieved after 12 h.
By using NaN3 and keeping the pH value between 3.0
and 4.0 by adding H2SO4 (see Supporting Information
for details).
[d]
[e]
[f]
Although these preliminary reactions were conduct-
ed on a small scale (0.5–1.0 mmol), we have observed
that, for our process, these resins are adequate to be that, as we have envisaged, water has a positive influ-
used in the presence of water ence on the dispersion of the reaction mixture facili-
910
asc.wiley-vch.de
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 908 – 916

Preparation and Use of Polystyryl-DABCOF2: An Efficient Recoverable and Reusable Catalyst
Scheme 1. Plausible mechanism for the process involving PS-DABCOF2 catalysts.
amount of organic solvent is necessary (ca.
4.0 mLmmolÀ1) for the recovery of the product and
reuse of catalyst with an obvious increase of the E-
factor.
In addition, PS-DABCOF2 catalysts 5a–f (in the re-
actions using water or organic solvent or SolFC)
during the magnetic stirring are crunched into a fine
powder that is rather difficult to be manipulated
during filtration. The use of water as reaction medium
offers a solution to this problem.
It is interesting to notice that by increasing the fluo-
ride loading, the possible reduction of the efficiency
of the catalyst due to crowding of the catalytic sites,
was not observed. Therefore, highly loaded 5f was
chosen for the continuation of this study. The main
advantage of the use of 5f is the minimal amount of
solid resin needed for running the reactions.
Scheme 2. Schematic representation of the small-scale pro-
tocol for the b-azidation of enone 6a in water with recovery
and reuse of catalyst 5f.
In Table 3 are summarized the results obtained in
tating the recovery of the catalyst and of the product the reactions of a,b-unsaturated ketones 6a–j cata-
7a.
lyzed by 20 mol% of 5f in water.
In fact, when the reaction is performed in water,
Reactions were carried out at 608C except in the
the catalyst 5f remains dispersed in water while upon case of enones 6c, 6g and 6h, where the reactions
adding EtOAc (1.0 mLmmolÀ1) to the reaction mix- were performed at 308C (Table 3, entries 3, 7 and 8)
ture the product is dissolved in the organic phase and since these substrates decompose at higher tempera-
therefore can be separated from aqueous layer by ture. Sterically hindered ketones 6e, 6f and 6i react
simple suction. The product can be isolated after slowly and require a higher excess of TMSN3 to com-
evaporation. Both water and catalyst are therefore plete the reaction (Table 3, entries 5, 6 and 9).
easily recovered and have been reused in further 10
runs with no change in the results. Under these condi- scale is low and ca. 3 times smaller than that of our
tions E-factor for 7a is 7.4 (Scheme 2). previously reported procedure performed under
In all cases the E-factor obtained on a 1.0 mmol
On the contrary, when the process is performed SolFC and based on the use of Amberlite IRA900F
under SolFC or in an organic solvent, a filtration is as catalyst.[8c]
needed to recover the catalyst and in the case of
To prove the efficiency of our approach we have
SolFC, the reaction mixture is strongly stuck on the also decided to perform the azidation of a,b-unsatu-
solid catalyst. Therefore in these cases a larger rated ketones 6a, 6f, and 6h on a large-scale setting in
Adv. Synth. Catal. 2012, 354, 908 – 916
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
911

Tommaso Angelini et al.
FULL PAPERS
Table 3. PS-DABCOF2-catalyzed b-azidation of enones 6 at 608C in water.
Entry
1
Enone
Time [h]
8
Adduct
Yield [%][a]
91
E[b]
8.6
6a
6b
7a
7b
2
3
20
16
95
6.7
6c
7c
93[c]
10.2
4
6d
18
7d
97[d]
5.9
5
6
6e
6f
150
24
7e
7f
95[d,e]
6.5
9.7
89[e]
7
6g
24
7g
80[c]
10.5
8
9
6h
6i
18
72
48
7h
7i
95[c]
93[e]
97
7.8
7.8
7.0
10
6j
7j
[a]
Isolated yield of the pure products 7 (see Experimental Section for further details).
E-factor.[13]
Reaction performed at 308C.
60/40 mixture of diastereoisomers.
3.0 equivalents of TMSN3 were used.
[b]
[c]
[d]
[e]
a continuous-flow reactor. According to our previous er and reuse the aqueous layer. Water can be a very
reports in this field, this reactor should be designed to expensive waste and the realisation of a procedure for
optimize the recovery and reuse of the catalyst, to its recovery and reuse is highly desirable especially in
minimize waste and in particular the amount of or- a large-scale protocol.
ganic solvent needed to isolate the final products.
The schematic representation of the reactor is illus-
In addition, in the context of this contribution, it is trated in Scheme 3 (thermostatted box is not shown
crucial to set the reactor in order to be able to recov- for clarity).
912
asc.wiley-vch.de
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 908 – 916

Preparation and Use of Polystyryl-DABCOF2: An Efficient Recoverable and Reusable Catalyst
evaporation of the organic solvent under reduced
pressure). Additional use of EtOAc (3ꢂ2 mL) al-
lowed the isolated yields of the pure products 7a, 7f
and 7h to be increased to 93, 91 and 96%, with a very
low E-factor of 1.7, 1.9 and 1.8, respectively.
It is important to notice that the adoption of a
cyclic continuous-flow procedure has allowed us to
minimize the use of water and TMSN3. In fact, to
allow the complete fluoride-azido ions exchange oc-
curring during the process, in the first run a 20 mol%
excess of TMSN3 has been used. Starting from the
second run the E-factor slightly decreased because no
more TMSN3 was necessary for the ion exchange (see
mechanism in Scheme 1) and strictly 1.0 equiv. of this
reactant can be used. The catalytic system has been
reused in three further runs with unchanged efficien-
cy. In batch conditions the use of a stoichiometric
amount of TMSN3 was not possible and an incom-
plete conversion was observed due to a partial de-
composition and loss of TMSN3.
In order to define a protocol for a representative
further manipulation of the intermediate b-azido ke-
tones 7, we have searched for a procedure to directly
reduce the azido group and prepare the correspond-
ing b-amino ketones 8 (Scheme 4). In order to set-up
a multistep procedure starting directly from a,b-unsa-
turated ketones 6, we focused our attention on a re-
Scheme 3. Preparation of b-azido ketones 7a, 7f and 7h
using a cyclic continuous-flow reactor.
The mixture of water (20 mL) and reactants ductive method that could operate in the presence of
(100 mmol of 6a, 6f or 6h, 1.2 equiv. of TMSN3) were water.
charged into a glass column functioning as reservoir
For this purpose, we directed our attention to the
and equipped with a mechanical stirring system ap- Pd-catalyzed reductive system that in our case, should
propriately designed and built to fit the column. PS- be used in the presence of water with a hydrogen
DABCOF2 (5f) (20 mol% measured on the ketone 6) source alternative to H2. Formic acid or hydrazine are
was charged into a glass column and the reaction mix- valuable options although in the literature their use is
ture was continuously pumped through it. In the cases reported with a large excess (10–20 equiv.).[15]
of 6a and 6h the complete conversion to the corre-
Pd-catalyzed reduction of azido group is a widely
sponding 7a and 7h was achieved under the same con- reported method in literature,[4a,16] but to the best of
ditions reported in Table 3. In the case of 6f a longer our knowledge for this reaction, there are no reports
reaction time (72 h) was expectedly needed for the on the use of Pd on alumina/formic acid or hydrazine
complete conversion to 7f due to the significantly re- systems.
duced amount of TMSN3 used (1.0 instead of
3.0 equivalents).
Preliminary experiments confirmed that the direct
reduction of b-azido ketones 7 to b-amino ketone 8 is
At this point, still operating under mechanical stir- not possible because these compounds are very unsta-
ring, the pump was left to run in order to recover the ble. In fact, after trying several sets of reaction condi-
aqueous reaction mixture into the reservoir. The me- tions, running the reactions in water or in organic sol-
chanical stirring was stopped and after some minutes vent, always only decomposition products were isolat-
water/organic phase separation occurred. Considering ed. Actually, to the best of our knowledge compounds
the amount remained adsorbed into the catalyst (ca.
40%), water was completely recovered into a separate
reservoir. It should be noticed that the reaction prog-
ress includes the consumption of 1 equiv. of H2O for
the formation of the desired b-azido ketone 7 (see
Scheme 1).
Then EtOAc was added (3ꢂ2 mL) to wash the cat-
alyst and to isolate the pure product 7a, 7f or 7h,
after passing it through the dry column (10 g of
Na2SO4) in 86, 85 and 92% yield, respectively (after Scheme 4. Preparation of N-Boc-b-amino ketones 9.
Adv. Synth. Catal. 2012, 354, 908 – 916
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
913

Tommaso Angelini et al.
FULL PAPERS
such as 8 are not known in the literature except as N-
protected derivatives 9.
We have then created a multistep continuous-flow
reactor (Scheme 5) by adding a third column contain-
For this reason we have directed our efforts to set- ing 5 mol% (calculated on 100 mmol of 6a) of Pd on
up a procedure for the reduction of the azido group Al2O3 to the reactor previously described in
of 7 in the presence of the protective agent Boc2O, as Scheme 3.
an amino protecting group N-Boc is a generally ap-
preciable choice due to its easy manipulation and continuous-flow according to the procedure described
chemical stability. before, the air/solvent valve was opened and still
At the end of the azidation process performed in
In the representative case of 7a we have devised an under operating mechanical stirring, the pump was set
efficient procedure for its reduction to the corre- to transfer the reaction mixture into the reservoir and
sponding 9a. In water, by using 5 mol% of Pd/Al2O3 empty the PS-DABCOF2 (5f) column. EtOAc was
in the presence of just 1.0 equivalent of HCOOH and added (6ꢂ2 mL) to wash the catalyst.
1.0 equivalent of Boc2O, the desired N-Boc-b-amino
ketone 9a was obtained in 80% yield.
By setting appropriately the valves, Boc2O
(0.95 equiv.) was charged into the reservoir, HCOOH
The protocol defined is very efficient mainly be- (0.95 equiv.) was added as a concentrated solution in
cause of the use of equimolar amounts of HCOOH, H2O (4 mL) into the column containing the Pd/Al2O3
Boc2O and 7a. It should also be highlighted that the and then the mixture was cyclically flowed through
use of this ratio is also crucial for the success of the catalyst Pd/Al2O3 for 18 h at 308C until the reduction/
procedure, otherwise decomposition processes of both protection process was complete.
the b-amino ketone 8 and the protecting agent occur
giving very discouraging results.
Similarly to the previously described reactor
(Scheme 3), at this point and still under operating me-
In addition, the reductive system showed no prob- chanical stirring, the pump was left to run in order to
lem in the reduction of 7a in clean water, but when in recover the aqueous reaction mixture into the reser-
batch conditions we tried the one-pot protocol start- voir. The mechanical stirring was stopped and after
ing from 6a, very poor results were obtained. This some minutes when water/organic phase separation
was ascribable to the excess of TMSN3 used in the b- occurred, water was recovered into a separate reser-
azidation step in batch conditions that is responsible voir. The pure product was then flowed through a dry
for the poisoning of the Pd catalyst.
column (10 g of Na2SO4) to be recovered after organ-
The adoption of the continuous-flow protocol fur- ic solvent evaporation under reduced pressure. Addi-
nished the solution to this problem. In fact by operat- tional use of EtOAC (3ꢂ2 mL) allowed the isolated
ing in continuous-flow conditions a stoichiometric yields of the pure products 9a and 9b in 74 and 77%,
amount of TMSN3 could be used and no poisoning of respectively with a very low E-factor of 3.2 and 2.7,
Pd catalyst could occur.
respectively.
Conclusions
In conclusion, we have reported the preparation of
novel solid fluorides 5a–f to optimize the b-azidation
of a,b-unsaturated ketones 6a–j.
The catalysts have been designed to obtain a higher
loading compared to that of commercially available
fluorides allowing the use of a smaller mass of catalyst
and help mixing of the reaction mixture. Gel-type and
porous polymeric supports have been compared and
more reticulated resins have proved to be more effi-
cient in the presence of water as reaction mixture.
According to our expectations, water has proved to
play a crucial role by increasing the reactivity and by
improving the dispersion of the reaction mixture. The
use of a porous solid support suitable for aqueous
conditions has allowed us to avoid organic fouling
caused by the sticking of the reaction mixture within
the polymeric matrix, allowing a better catalytic effi-
ciency and facilitating the recovery of the products
from the catalyst.
Scheme 5. Cyclic continuous-flow reactor for the prepara-
tion of N-Boc-b-amino ketones 9a and 9b.
914
asc.wiley-vch.de
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 908 – 916

Preparation and Use of Polystyryl-DABCOF2: An Efficient Recoverable and Reusable Catalyst
and trimethylsilyl azide (0.100 mL, 0.75 mmol) were consec-
As an evident consequence, this approach has al-
lowed to us minimize the use of the organic solvent
needed for the complete recovery of product and
leave the catalyst clean and reusable, while in the pro-
tocol performed under SolFC[8c] the reaction mixture
is strongly stuck on the solid catalyst requiring a
larger amount of organic solvent for the isolation of
the product. E-factors are in the range of 5.9–10.5 and
therefore ca. 3 times smaller than our previous proce-
dure operating under SolFC.[8c]
utively added and the resulting mixture was left under stir-
ring at 608C for 8 h. Then, ethyl acetate (0.5 mL) was
added, the mixture was stirred for 5 min, the organic phase
was separated with a Pasteur pipette and the solvent evapo-
rated under vacuum to give pure 7a; yield: 0.071 g (91%).
The aqueous phase and the catalyst remaining in the vial
can be reused. E-factor: 7.4 [0.056 g (ketone)+0.086 g
(TMSN3)+0.450 g
(product).
(AcOEt)À0.070 g
(product)]/0.070 g
To further improve the efficiency of our approach
we have reported a protocol operating in a continu-
ous-flow manner that has allowed to further reduce
the E-factor to 1.7–1.9 (ca. 80% compared to our
batch conditions and 93[8c]–98%[8c] compared to litera-
ture procedures) for the representative substrates 6a,
6f, and 6h. The continuous-flow protocol has allowed
us to minimize the use of TMSN3 making the recov-
ery and reuse of water and catalyst 5f very efficient
and simple.
Acknowledgements
We gratefully acknowledge the Ministero dell’Istruzione,
dell’Universitꢀ e della Ricerca (MIUR) within the projects
PRIN 2008 and “Firb-Futuro in Ricerca” and the Universitꢀ
degli Studi di Perugia and Parma for financial support.
Finally a novel reduction system using Pd on Al2O3
(5 mol%) and equimolar amount of HCOOH has
been used in the presence of 1 equiv. of Boc2O to set-
up a multistep protocol operating in continuous-flow
conditions for the preparation of representative N-
Boc-b-amino ketones 9a and 9b starting from the cor-
responding enones 6a and 6b with an E-factor of 3.2
and 2.7, respectively.
References
[1] a) D. Enders, C. Wang, J. X. Liebich, Chem. Eur. J.
2009, 15, 11058–11076; b) G. Bartoli, M. Bartolacci, A.
Giuliani, E. Marcantoni, M. Massaccesi, E. Torregiani,
J. Org. Chem. 2005, 70, 169–174; c) S. Kobayashi, K.
Kakumoto, M. Sugiura, Org. Lett. 2002, 4, 1319–1322;
d) L. W. Xu, C. G. Xia, Tetrahedron Lett. 2004, 45,
4507–4510.
[2] a) M. Arend, B. Westermann, N. Risch, Angew. Chem.
1998, 110, 1096–1122; Angew. Chem. Int. Ed. 1998, 37,
1044–1070; b) M. P. Sibi, J. J. Shay, M. Liu, C. P. Jas-
perse, J. Am. Chem. Soc. 1998, 120, 6615–6616; c) Y.
Bandala, E. Juaristi, in: Amino Acids, Peptides and
Proteins in Organic Chemistry, (Ed.:A. B Hughes),
Wiley-VCH, Weinheim, 2009; d) D. C. Cole, Tetrahe-
dron 1994, 50, 9517–9582.
[3] a) S. Brꢃse, C. Gil, K. Knepper, V. Zimmermann,
Angew. Chem. 2005, 117, 5320–5374; Angew. Chem.
Int. Ed. 2005, 44, 5188–5240; b) M. E. C. Biffin, S. J.
Miller, D. B. Paul, in: The Chemistry of the Azido
Group, (Ed.: S. Patai), Wiley-Interscience, London,
1971, pp 57–180; c) E. F. V. Scriven, K. Turbull, Chem.
Rev. 1988, 88, 297–368.
[4] a) M. S. Taylor, D. N. Zalatan, A. M. Lerchner, E. N.
Jacobsen, J. Am. Chem. Soc. 2005, 127, 1313–1317;
b) L.-W. Xu, C.-G. Xia, J.-W. Li, S.-L. Zhou, Tetrahe-
dron Lett. 2004, 45, 1219–1221; c) L.-W. Xu, C.-G. Xia,
J.-W. Li, S.-L. Zhou, Synlett 2003, 2246–2248.
[5] D. J. Guerin, T. E. Horstmann, S. J. Miller, Org. Lett.
1999, 1, 1107–1109.
Experimental Section
General Remarks
All chemicals were purchased and used without any further
purification. All 1H NMR and 13C NMR spectra were re-
corded at 400 MHz and 100.6 MHz, respectively, using a
convenient deuterated solvent (reported in the characteriza-
tion charts) and the residual peak as internal standard, or
TMS in the case of CDCl3. IR spectra were recorded with
an FT-IR Bruker IFS 112 v spectrometer using CHCl3 as
solvent. Elemental analyses were realized by using a Fisons
instrument EA 1108 CHN. SEM pictures of gold coated
polymers (instrument EMITECH K55OX sputter coater)
were taken on a Philips SEM XL30.
All polystyrene-supported ammonium salts 3a–f, 4a–f,
5a–f are new compounds. All b-azido ketones 7[8c] and N-
Boc-b-amino ketone 9a[4a] are known compounds while 9b is
a new compound.
All the detailed experimental procedures, characterization
data of compounds 3a–f, 4a–f, 5a–f, 7a–j and 9a, b, along
1
[6] For other protocols using TMSN3, see: a) I. Adamo, F.
Benedetti, F. Berti, P. Campaner, Org. Lett. 2006, 8. 51–
54; b) A. Dondoni, A. Marra, A. C. Boscarato, Eur. J.
Org. Chem. 1999, 5, 3562–3572; c) A. Dondoni, A. Bo-
scarato, A. Marra, Tetrahedron: Asymmetry 1994, 5,
2209–2212; d) R. Thiergardt, G. Rihs, P. Hug, H. H.
Peter, Tetrahedron 1995, 51, 733–742.
with copies of the SEM images of 5f as well as of the H
and 13C NMR spectra for compounds 7a–j and 9a, b are
given in the Supporting Information.
Representative Batch Experimental Procedure
In a screw-capped vial equipped with a magnetic stirrer, PS-
DABCOF2 (5f) (0.014 g, 0.1 mmol, 7.4 mmolgÀ1), water
(0.250 mL), (E)-3-hepten-2-one (6a) (0.066 mL, 0.5 mmol)
[7] a) D. J. Guerin, S. J. Miller, J. Am. Chem. Soc. 2002,
124, 2134–2136; b) T. E. Horstmann, D. J. Guerin, S. J.
Adv. Synth. Catal. 2012, 354, 908 – 916
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
915

Tommaso Angelini et al.
FULL PAPERS
Miller, Angew. Chem. 2000, 112, 3781–3784; Angew.
Chem. Int. Ed. 2000, 39, 3635–3637.
[8] a) F. Fringuelli, D. Lanari, F. Pizzo, L. Vaccaro, Green
Chem. 2010, 12, 1301–1305; b) F. Fringuelli, D. Lanari,
F. Pizzo, L. Vaccaro, Eur. J. Org. Chem. 2008, 3928–
3932; c) L. Castrica, F. Fringuelli, L. Gregoli, F. Pizzo,
L. Vaccaro, J. Org. Chem. 2006, 71, 9536–9539.
[9] a) S. Bonollo, D. Lanari, F. Pizzo, L. Vaccaro, Org. Lett.
2011, 13, 2150–2152; b) S. Bonollo, D. Lanari, L. Vac-
caro, Eur. J. Org. Chem. 2011, 2587–2598.
[10] For leading examples, see: a) D. Lanari, R. Ballini, A.
Palmieri, F. Pizzo, L. Vaccaro, Eur. J. Org. Chem. 2011,
2874–2884; b) F. Fringuelli, F. Pizzo, S. Tortoioli, L.
Vaccaro, J. Org. Chem. 2004, 69, 8780–8785.
[11] For leading examples, see: a) A. Zvagulis, S. Bonollo,
D. Lanari, F. Pizzo, L. Vaccaro, Adv. Synth. Catal. 2010,
352, 2489–2496; b) L. Soldi, W. Ferstl, S. Loebbecke, R.
Maggi, C. Malmassari, G. Sartori, S. Yada, J. Catal.
2008, 258, 289–295.
[12] a) M. Baumann, I. R. Baxendale, S. V. Ley, Mol. Diver-
sity 2011, 15, 613–630; b) C. J. Smith, C. D. Smith, N.
Nikbin, S. V. Ley, I. R. Baxendale, Org. Biomol. Chem.
2011, 9, 1927–1937; c) M. OꢄBrien, N. Taylor, A. Poly-
zos, I. R. Baxendale, S. V. Ley, Chem. Sci. 2011, 2,
1250–1257; d) C. Aranda, A. Cornejo, J. M. Fraile, E.
Garcꢅa-Verdugo, M. J. Gil, S. V. Luis, J. A. Mayoral, V.
Martinez-Merino, Z. Ochoa Green Chem. 2011, 13,
983–990; e) V. Sans, N. Karbass, . M. I. Burguete, V.
CompaÇ, E. Garcꢅa-Verdugo, S. V. Luis, M. Pawlak
Chem. Eur. J. 2011, 17, 1894–1906; f) M. Baumann,
I. R. Baxendale, M. Brasholz, J. J. Hayward, S. V. Ley,
N. Nikbin, Synlett 2011, 1375–1380; g) L. J. Martin,
A. L. Marzinzik, S. V. Ley, I. R. Baxendale, Org. Lett.
2011, 13, 320–323.
[13] a) R. A. Sheldon, Chem. Ind. (London, U.K.) 1997, 12–
15; b) R. A. Sheldon, Green Chem. 2007, 9, 1273–1283;
c) J. Augꢆ, Green Chem. 2008, 10, 225–231; d) R. A.
Sheldon, Chem. Commun. 2008, 3352–3365.
[14] http://www.dow.com/assets/attachments/business/ier/
ier_for_industrial_water_treatment/amberlite_ira900_cl/
tds/amberlite_ira900_cl.pdf.
[15] a) A. Kulkarni, R. Gianattasio, B. Torok, Synthesis
2011, 1227–1232; b) K. Nadrah, M. S. Dolenc, Synlett
2007, 1257–1258; c) Z. Paryzek, H. Koeng, B. Tabaczka,
Synthesis 2003, 2023–2026; d) N. A. Cortes, R. F. Heck,
J. Org. Chem. 1977, 42, 3491–3494; e) A. Furst, R. C.
Berlo, S. Hooton, Chem. Rev. 1965, 65, 51–68.
[16] a) S. Chandrasekhar, S. J. Prakash, C. L. Rao, J. Org.
Chem. 2006, 71, 2196–2199; b) H. Sajiki, Tetrahedron
Lett. 1995, 36, 3465–3468; c) Y. Gaony, J. Org. Chem.
1994, 59, 6853–6855; d) E. J. Corey, J. O. Link, J. Am.
Chem. Soc. 1992, 114, 1906–1908.
916
asc.wiley-vch.de
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 908 – 916