Unmodified nano-powder magnetite catalyzes a four-component aza-Sakurai reaction

Article abstract of DOI:10.1002/adsc.200800089

A new catalyst for an old material: magnetite is an excellent Lewis acid catalyst for the four-component aza-Sakurai reaction. The process could be repeated up to 15-times without losing effectiveness, with the catalyst recycling being as easy as the use of a simple magnet. The catalyst is selective and could discriminate between aldehyde and ketone functionalities, catalyzing first the reaction with the higher electrophilic aldehyde.

Full text of DOI:10.1002/adsc.200800089

COMMUNICATIONS
DOI: 10.1002/adsc.200800089
Unmodified Nano-Powder Magnetite Catalyzes a Four-
Component Aza-Sakurai Reaction
Ricardo Martínez,a Diego J. Ramón,a,* and Miguel Yusa,*
a
Instituto de Síntesis Orgµnica (ISO), and Departamento de Química Orgµnica, Facultad de Ciencias, Universidad de
Alicante, Apdo. 99, 03080 Alicante, Spain
Fax : (+35)-965-9035; e-mail: djramon@ua.es or yus@ua.es
Received: February 11, 2008; Published online: May 9, 2008
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: A new catalyst for an old material: mag-
netite is an excellent Lewis acid catalyst for the
four-component aza-Sakurai reaction. The process
could be repeated up to 15-times without losing ef-
fectiveness, with the catalyst recycling being as easy
as the use of a simple magnet. The catalyst is selec-
tive and could discriminate between aldehyde and
ketone functionalities, catalyzing first the reaction
with the higher electrophilic aldehyde.
ing reagents and catalysts the more difficult is the suc-
cess of the reaction. Very recently, a new four-compo-
nent aza-Sakurai reaction catalyzed by substoichio-
metric amounts of FeSO4 and using aldehyde deriva-
tives has been published, the results being in general
moderate.[11]
With this study we would like to show the first ex-
ample in which the unmodified commercial nano-
powder magnetite can be used (and re-used) as a typi-
cal heterogeneous Lewis acid catalyst in usual and
mild organic synthetic protocols.
Our starting point is the known surface of the mag-
netite Fe3O4 (111), which is terminated by a hexago-
nal oxygen layer covered by one quarter monolayer
of iron cations.[12] These Lewis acid centres on the sur-
face could catalyze the reaction. So, the multicompo-
nent reaction outlined in Table 1 was used as standard
Keywords: heterogeneous catalysis; iron; Lewis
acids; magnetic properties; multicomponent reac-
tions; synthetic methods
Magnetic nano- and microparticles are of great inter- for the reaction condition optimization. The reaction
est for a wide range of disciplines, and their develop- using 40 mol% of commercial nanopowder magnetite
ment is highly demanded nowadays,[1] with the suc- gave the expected carbamate 5a with a good result,
cessful applications of such particles depending whereas the reaction under similar conditions in the
strongly on their stability. So, for many applications, absence of catalyst did not take place, even after a
the protection of the naked magnetic particle, includ- longer reaction time (compare entries 1 and 2 in
ing grafting or coating with new organic or inorganic Table 1). The reaction using a lower amount of cata-
layers, is crucial,[2] and permits their further function- lyst provoked an appreciable and unusual increase on
alization.[3]
the result (entry 3 in Table 1), whereas further de-
Among the different magnetic nanoparticles, Fe3O4 creases resulted in a continuous decline of the results
is the most important one, and although naked mag- (Table 1, entries 5 and 6). The reaction could be also
netite seems to play a capital role in pre-biotic performed without the stirring bar and without a sig-
chemistry,[4] its use as a catalyst has been quite limited nificant effect (Table 1, entry 4). The reaction using
to only a few examples, including carbon-carbon Fe2O3, which presents a similar atom surface distribu-
double bond isomerization,[5] modified Fischer– tion to magnetite only changing Fe(II) by Fe
ACHTREUNG(III),
Tropsch reaction,[6] water-gas shift reaction,[7] dehy- failed (entry 7). However, the reaction using FeO in
drogenation[8] and epoxidation[9] processes, many of which, excluding imperfections, the last surface layer
them performed under very extreme reaction condi- is terminated purely by oxygen atoms, gave a signifi-
tions.
On the other hand, multicomponent reactions allow the reaction seems to be catalyzed by Fe(II) and not
one to create complicated molecules using only one by Fe(III) centres. The reaction performed at room
cant result (entry 8). These results pointed out that
AHCTREUNG
process in a very fast, efficient and time-saving temperature took a longer times in order to obtain
manner,[10] albeit with the higher the number of start- similar results (Table 1, entries 9 and 10). Other reac-
Adv. Synth. Catal. 2008, 350, 1235 – 1240
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1235

Ricardo Martínez et al.
COMMUNICATIONS
Table 1. Reaction-conditions optimization.
cyclohexanone in only two hours (entries 23 and 24).
Finally, it is worthy of note that the catalyst is selec-
tive and could discriminate between an aromatic alde-
hyde and the related ketone. Thus, the reaction with
4-acetylbenzaldehyde gave only the product resulting
from the reaction with the aldehyde groupeven when
using two equivalents of compounds 2, 3 and 4, the
related one resulting from the reaction with the
ketone groupnot being detected (entry 25).
Once the catalytic activity of the unmodified mag-
netite was demonstrated, we faced the problem of re-
use, finding that the chemical yields were constant in
a range between 83 and 98%, after 15 cycles of reac-
tion, for the preparation of compound 5a (Figure 1),
with the magnetite being maintained inside the flask
by the helpof a magnet.
In order to explore the possible degradation of the
naked magnetite under the reaction conditions, the
BET surface area was determined at the beginning of
the process (52.69 m2 gÀ1) and after 15 cycles
(50.43 m2 gÀ1), with this concordant result showing
that there is no significant sinterization process under
the assayed reaction conditions. A similar result could
be extracted from the observation of TEM microscop-
ic images, which did not show any difference between
the unmodified magnetite and the 15-fold reused one
(Figure 2).
Entry Solvent T [8C] Catalyst [mol%] t [h] Yield [%][a]
1
2
3
4
5
6
7
8
PhMe 110
PhMe 110
PhMe 110
PhMe 110
PhMe 110
PhMe 110
PhMe 110
PhMe 110
PhMe 25
CH2Cl2 25
-
8
2
2
2
2
2
2
2
24
72
2
2
2
0
Fe3O4 [40]
Fe3O4 [20]
Fe3O4 [20]
Fe3O4 [10]
Fe3O4 [1]
Fe2O3 [20]
FeO [20]
Fe3O4 [20]
Fe3O4 [20]
75
85
80[b]
79
<5
0
40
65
75
30
55
11
0
9
10
11
12
13
14
15
CH2Cl2 110[c] Fe3O4 [20]
THF
110[c] Fe3O4 [20]
DMF
110
Fe3O4 [20]
Fe3O4 [20]
Fe3O4 [20]
DMSO 110
MeCN 110
2
2
38
[a]
Isolated yields after column chromatography (silica gel:
hexane/ethyl acetate).
Reaction performed without stirring bar.
Reaction performed in a sealed tube.
[b]
[c]
To finish with the study on the stability of the mag-
tion conditions or solvents did not improve the prece- netite nanopowder, we studied the obtained liquid
dent result (entries 11–15). phase after one cycle of the reaction for the prepara-
After completing the optimization study, we exam- tion of compound 5a. The magnetite was isolated by a
ined the reaction’s scope, starting by changing the magnet and washed with toluene; the resulting liquid
nature of chloroformate 2. The reaction using differ- mixture without filtration was used as medium for the
ent substituted chloroformates gave practically the preparation of compound 5m in the absence of mag-
same results (Table 2, entries 1–6). These results are netite, affording the expected allyl carbamate in only
similar using benzoyl chloride or even using pivaloyl 21% yield. This result indicated that the isolation pro-
chloride, which is clearly a more hindered compound cesses permitted the presence of iron species in the
(entries 7 and 8). Unfortunately, the reaction using medium. In a parallel experiment, after the standard
N,N-dimethylcarbamoyl chloride failed (entry 9). preparation of compound 5a and isolation of magnet-
Then, we followed with a study using different sour- ite as above, the resulting mixture without filtration
ces of the silylated nucleophile 3. Whereas the reac- was concentrated and re-dissolved in methanol. The
tion with trimethylsilyl cyanide (3b) gave the expect- flame atomic absorption spectroscopy (FAAS)
ed amido nitrile 5j with a very low yield, the reaction showed the presence of around 6 mg of iron in the
using the corresponding allyltrimethylsilane (3c) gave methanolic solution (less than 2% of initial magnetite
the expected carbamic compound 5k in high yield added). However, if prior to the concentration, a
(compare entries 2, 10 and 11). Similar allylation pro- simple filtration is done, the FAAS did not show the
cess using a chiral chloroformate did not show any presence of iron atoms, excluding the presence of ho-
diastereoselectivity (entry 14).
mogeneous iron species. Therefore, we believe that
The nature of the carbonyl compound 1 has a cer- the possible degradation of magnetite is not true, and
tain impact on the results. So, whereas aromatic or the initial catalytic and FAAS results only showed the
a,b-unsaturated aldehydes gave excellent results, ali- capacity of the magnet used to trap the nanoparticles.
phatic aldehydes gave only good chemical yields
The above protocol could be expanded to the relat-
(compare entries 2, 11–13, 15–19 in Table 2). In the ed ketals 6, as depicted in Scheme 1. The reaction
case of using acetophenone or 2-hexanone (en- using the masked benzaldehyde derivative 6a gave
tries 20–22) the reaction time had to be increased in the corresponding carbamate 5b in 90% yield, which
order to improve on the previous results. However, is a similar result to that obtained using benzaldehyde
the expected carbamate 5u and v were obtained using (compare to entry 2 in Table 2). In order to study the
1236
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1235 – 1240

Unmodified Nano-Powder Magnetite Catalyzes a Four-Component Aza-Sakurai Reaction
COMMUNICATIONS
Table 2. Magnetite catalyzes the four-component aza-Sakurai reaction.
Entry
R1
R2
R3
Nu
Product
Yield [%][a]
1
2
3
4
5
6
7
8
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
H
OCH2Ph
OCH2CH=CH2
H[b]
5a
5b
5c
5d
5e
5f
5g
5h
5i
85
98
93
70
70
80
81
56
H[b]
OCH2C CH
OPh
H[b]
ꢀ
H[b]
OEt
H[b]
(À)-menthyloxy
H[b]
Ph
H[b]
t-Bu
NMe2
H[b]
9
Ph
H
H[b]
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Ph
Ph
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
(À)-menthyloxy
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
OCH2CH=CH2
CN
5j
5k
5l
5m
5n
5o
5p
5q
5r
5s
5t
5u
5v
5w
5x
5y
<5
96
94
CH2CH=CH2
CH2CH=CH2
CH2CH=CH2
CH2CH=CH2
H[b]
4-ClC6H4
4-MeOC6H4
Ph
PhCH=CH
PhCH=CH
i-Pr
c-C6H11
c-C6H11
Ph
n-Bu
n-Bu
84
82[c]
71
CH2CH=CH2
85
61
55
66
H[b]
H[b]
CH2CH=CH2
H[b]
71[d]
89[d]
45[d]
53
H[b]
ꢀ
OCH2C CH
ꢀ
OCH2C CH
CH2CH=CH2
-(CH2)5-
-(CH2)5-
4-MeCOC6H4
OCH2CꢀCH
H[b]
ꢀ
OCH2C CH
CH2CH=CH2
CH2CH=CH2
49
93
H
OCH2CH=CH2
[a]
Isolated yields after column chromatography (silica gel: hexane/ethyl acetate).
Reaction performed using Et3SiH.
1:1 Mixture of diastereoisomers.
[b]
[c]
[d]
Yield after 16 h of reaction time.
Figure 1.
Adv. Synth. Catal. 2008, 350, 1235 – 1240
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1237

Ricardo Martínez et al.
COMMUNICATIONS
Figure 2. TEM microscopic images of commmercial nanopowder magnetite (left) and the same magnetite after being used in
15 cycles (right).
Scheme 1. Multicomponent reaction using ketals 6.
possible selectivity between the free aldehyde group submitted to reaction under different catalyst condi-
and the masked one, the reaction was performed tions. Compound 10 was alternatively prepared in a
[13]
using compound 6b, giving the biscarbamate deriva- classical two-stepreaction
tive 7 with a poor yield when only one equivalent of
reagents 2b, 3a and 4 were used and a higher yield
when two equivalents were added.
using benzaldehyde, lith-
Finally, we tried to identify some aspects of the pos-
sible catalytic pathway and therefore we performed
several parallel reactions with different reagents.
Firstly, we performed the reaction of compounds 4
and 2b in toluene at 1108C finding that after only one
hour the carbamate 8 was formed, independently of
the presence or absence of magnetite [Eq. (1) in
Scheme 2], the chemical yield being increased after
two hours. To the above mixture of the carbamate 8
and the silyl derivative 9 was added benzaldehyde
(1a) and, as in the previous case, the formation of
imine derivative 10 [Eq. (2) in Scheme 2] was inde-
pendent on the presence of magnetite. This step was
previously described as the only Fe(II)-catalyzed step
under homogeneous conditions.[11] However, under
higher temperature conditions the catalyst seems to
be unnecessary. Finally, the isolated carbamate 10 was Scheme 2. Possible reaction pathway.
1238
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1235 – 1240

Unmodified Nano-Powder Magnetite Catalyzes a Four-Component Aza-Sakurai Reaction
COMMUNICATIONS
ium hexamethyldisilylamide to form the correspond- Acknowledgements
ing imine and then reaction with the corresponding
allyl chloroformate.
The reaction of the imine compound 10 with tri-
ACHTREUNGethylsilane (3a) at 1108C in toluene [Eq. (3) in
We are grateful to the Spanish Ministerio de Educación y
Ciencia (Consolider Ingenio 2010 CSD2007–00006) as well
as to the Generalitat Valenciana (GV; grant no. GRUPOS03/
135) for the financial support. R.M. thanks GV for a predoc-
toral grant.
Scheme 2] failed after two hours. The parallel reaction
using one equivalent of hydrochloric acid-free trime-
thylsilyl chloride also failed under the same condi-
tions. The reaction using only 20 mol% of magnetite
gave the expected product 5b, but with a low chemical
yield (<10%). However, the reaction using one
equivalent of hydrochloric acid-free trimethylsilyl
chloride and 20 mol% of magnetite gave the expected
product 5b in the usual chemical yields (90%). Finally,
and in order to avoid the hypothesis that trimethyl-
silyl chloride reacted with magnetite to form soluble
FeCl2 and this was the true catalyst, the final reaction
of compounds 10 and 3a was repeated using 0.3
mol% of FeCl2 (which is a higher amount of iron than
was found in an FAAS experiment without filtration;
see above) giving the compound 5b in a miserable
yield (<5%). After these experiments, we concluded
that the reaction pathway seems to follow the equa-
tions outlined in Scheme 2, with the magnetite cata-
lyst being only important in the reaction of silyl deriv-
ative 3 with the imine intermediate 10, the nature of
the true catalyst is still elusive.
References
[1] A.-H. Lu, E. L. Salabas, F. Schüth, Angew. Chem. 2007,
119, 1242–1266; Angew. Chem. Int. Ed. 2007, 46, 1222–
1244.
[2] See, for instance: a) Y. Deng, W. Yang, C. Wang, S. Fu,
Adv. Mater. 2003, 15, 1729–1732; b) D. Horµk, F. Ledn-
´
´
ˇ
icky, E. Petrovsky, A. Kapicka, Macromol. Mater. Eng.
2004, 289, 341–348; c) Y. Ding, Y. Hu, X. Jiang, L.
Zhang, C. Yang, Angew. Chem. 2004, 116, 6529–6532;
Angew. Chem. Int. Ed. 2004, 43, 6369–6372; d) D.-J.
Kim, Y.-K. Lyu, H. N. Choi, I.-H. Min, W.-Y. Lee
Chem. Commun. 2005, 2966–2968; e) H. Wakamatsu,
K. Yamamoto, A. Nakao, T. Aoyagi, J. Magn. Magn.
Mater. 2006, 302, 327–333; f) L. Guo, G.-L. Pei, T.-J.
Wang, Z.-W. Wang, Y. Jin, Coll. Surf. A: Physicochem.
Eng. Asp. 2007, 293, 58–62.
[3] See, for instance: a) A. Hu, G. T. Yee, W. Lin, J. Am.
Chem. Soc. 2005, 127, 12486–12487; b) S. Ko, J. Jang,
Angew. Chem. 2006, 118, 7726–7729; Angew. Chem.
Int. Ed. 2006, 45, 7564–7567; c) M. Kawamura, K.
Sato, Chem. Commun. 2006, 4718–4719; d) C. ÓDµ-
laigh, S. A. Corr, Y. Gun’ko, S. J. Connon, Angew.
Chem. 2007, 119, 4407–4410; Angew. Chem. Int. Ed.
2007, 46, 4329–4332.
[4] Q. W. Chen, C. L. Chen, Curr. Org. Chem. 2005, 9,
989–998.
[5] W. I. Gilbert, J. Turkevich, E. S. Wallis, J. Org. Chem.
1938, 3, 611–617.
In conclusion, unmodified commercial magnetite
nanopowder has been shown to be an active, stable
and highly selective catalyst for the uncommon four-
component aza-Sakurai reaction. This is the first time
that unmodified magnetite is used as catalyst in
normal organic reactions. The simple recyclability
makes this catalyst suitable for continuous industrial
processes.
[6] L. S. Glebov, G. A. Kliger, T. P. Popova, V. E. Shiryae-
va, V. P. Ryzhikov, E. V. Marchevskaya, O. A. Lesik,
S. M. Loktev, V. G. Beryezkin, J. Mol. Catal. 1986, 35,
335–348.
ExperimentalSection
[7] Q. Liu, W. Ma, R. He, Z. Mu, Catal. Today 2005, 106,
52–56.
[8] A. Schüle, U. Nieken, O. Shekhah, W. Ranke, R.
Schlçgl, G. Kolios, Phys. Chem. Chem. Phys. 2007, 9,
3619–3634.
[9] a) J. Liang, Q. Zhang, H. Wu, G. Meng, Q. Tang, Y.
Wang, Catal. Commun. 2004, 5, 665–669; b) D. Guin,
B. Baruwati, S. V. Manorama, J. Mol. Catal. A: Chem.
2005, 242, 26–31.
[10] For recent reviews, see: a) Multicomponent Reactions,
(Eds.: J. Zhu, H. Bienaymꢁ), Wiley-VCH, Weinheim,
2005; b) L. A. Wessjohann, E. Ruijter, Mol. Diversity
2005, 9, 159–169; c) D. J. Ramón, M. Yus, Angew.
Chem. 2005, 117, 1628–1661; Angew. Chem. Int. Ed.
2005, 44, 1602–1634; d) A. Dçmling, Chem. Rev. 2006,
106, 17–89; e) A. Dondoni, A. Massi, Acc. Chem. Res.
2006, 39, 451–463; f) F. Lieby-Muller, C. Simon, T.
Constantieux, J. Rodriguez, QSAR Comb. Sci. 2006, 25,
432–438; g) G. Guillena, D. J. Ramón, M. Yus, Tetrahe-
dron: Asymmetry 2007, 18, 693–700.
GeneralProcedure for the Four-Component Aza-
Sakurai Reaction Catalyzed by Magnetite
To a solution of the carbonyl compound (1, 4 mmol) in dry
toluene (5 mL) were added chloroformate derivative (2,
4.8 mmol), silyl derivative (3, 4.8 mmol), hexamethyldisila-
zane (4, 4.8 mmol, 1.01 mL) and Fe3O4 (0.86 mmol, 0.2 g)
under an argon atmosphere. The reaction mixture was
heated at 1108C and stirred for 2 h. Then, the catalyst and
stirrer bar were removed with a magnet. The solvent was re-
moved under reduced pressure and the resulting residue was
purified by flash chromatography on silica gel (hexane/ethyl
acetate) to give the corresponding product 5.
Supporting Information
Experimental procedure and characterization data for all
compounds are given in the Supporting Information.
Adv. Synth. Catal. 2008, 350, 1235 – 1240
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1239

Ricardo Martínez et al.
COMMUNICATIONS
[11] a) Q.-Y. Song, B.-L. Yang, S.-K. Tian, J. Org. Chem.
2007, 72, 5407–5410; b) B.-L. Yang, S.-K. Tian Eur. J.
Org. Chem. 2007, 4646–4650.
klewski, W. Ranke; W. Weiss, C. Wçll, R. Schlçgl,
Phys. Chem. Chem. Phys. 2000, 2, 5314–5319.
[13] a) T. R. Wu, J. M. Chong, Org. Lett. 2006, 8, 5–18;
b) A. Ting, S. Lou, S. E. Schaus, Org. Lett. 2006, 8,
2003–2006.
[12] a) Y. Joseph, W. Ranke, W. Weiss, J. Phys. Chem. B
2000, 104, 3224–3236; b) Y. Joseph, M. Wühn, A. Ni-
1240
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1235 – 1240