Halide substituted Schiff-bases: Different activities in methyltrioxorhenium(VII) catalyzed epoxidation via different substitution patterns

Article abstract of DOI:10.1016/j.jorganchem.2011.12.010

This report shows the influence of halide substituted Schiff-bases as ligands of methyltrioxorhenium (MTO) in epoxidation catalysis. Therefore, selected Schiff-bases were prepared by the reaction of hydroxy-benzaldehydes and aniline derivates. These differently substituted Schiff-bases were tested as MTO-ligands in cyclooctene-and 1-octene-epoxidation. Although no great disparities among the substitution patterns have been found, some conclusions can be drawn. Flourines are inferior to chlorines or bromines as substituents. Halides in ortho-position lead to higher activities than in para-or meta-position. The balance between electron donating and withdrawing influences at the Schiff-base plays a prominent role in their utility as ligand to MTO in epoxidation catalysis.

Full text of DOI:10.1016/j.jorganchem.2011.12.010

Journal of Organometallic Chemistry 701 (2012) 51e55
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Halide substituted Schiff-bases: Different activities in methyltrioxorhenium(VII)
catalyzed epoxidation via different substitution patterns
*
Philipp Altmann, Mirza Cokoja, Fritz E. Kühn
Chair of Inorganic Chemistry and Molecular Catalysis, Catalysis Research Center, Technische Universität München, Ernst-Otto-Fischer-Str. 1, 85747 Garching bei München, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
This report shows the influence of halide substituted Schiff-bases as ligands of methyltrioxorhenium
(MTO) in epoxidation catalysis. Therefore, selected Schiff-bases were prepared by the reaction of
hydroxy-benzaldehydes and aniline derivates. These differently substituted Schiff-bases were tested as
MTO-ligands in cyclooctene- and 1-octene-epoxidation. Although no great disparities among the
substitution patterns have been found, some conclusions can be drawn. Flourines are inferior to chlorines
or bromines as substituents. Halides in ortho-position lead to higher activities than in para- or meta-
position. The balance between electron donating and withdrawing influences at the Schiff-base plays
a prominent role in their utility as ligand to MTO in epoxidation catalysis.
Received 11 November 2011
Received in revised form
5 December 2011
Accepted 7 December 2011
Keywords:
Epoxidation
Homogeneous catalysis
Methyltrioxorhenium (MTO)
Schiff-base
Ó 2011 Published by Elsevier B.V.
Hydrogen peroxide
1. Introduction
studies. Many scientific reports are available based on this substrate
to compare the obtained results with respect to the general activity
Methyltrioxorhenium (MTO) has proven to be a highly versatile
catalyst for a plethora of organic reactions [1e5]. Best examined is
olefin oxidation [2,3,6e33]. The particular advantage of MTO is its
compatibility with hydrogen peroxide as oxidizing agent, leading to
water as the only byproduct. However, diol formation in the pres-
ence of Lewis acidic MTO and catalyst decomposition in aqueous
solutions are major problems [34,35]. When trying to overcome
both problems it was found that the presence of Lewis-bases in
significant excess stabilizes MTO, accelerates the oxidation reaction
and leads to significant diol reduction [16,22,25,27,36e39]. It was
further observed that Schiff-bases as additives need not to be
applied in excess to reach similar effects [21,26,30,32,33,40e42].
However, it has also been shown that the scope of the Schiff-base
supported epoxidation catalyzed by MTO appears to have a more
limited substrate scope than the reaction supported by aromatic
Lewis-bases [20,43,44]. Furthermore, the so far published results
are not all matching well.
of the catalytic system. 1-octene is chosen since its oxidation is
much slower than that of cyclooctene and comparison of the two
substrates should reveal the influence of the different substituents
quite well.
2. Experimental
2.1. Catalysis
Method A: 441 mg cyclooctene (4 mmol, 0.52 mL), 0.5 mL
mesitylene (internal standard), 1 mol% catalyst (0.04 mmol
MTO þ 0.04 mmol Schiff-base), 2 mL nitromethane (MeNO2).
Method B: 449 mg 1-octene (4 mmol, 0.63 mL), 0.2 mL mesity-
lene (internal standard), 0.2 mL toluene (internal standard), 1 mol%
catalyst (0.04 mmol MTO
nitromethane.
þ
0.04 mmol Schiff-base), 2 mL
The compounds from methods A and B were combined in
tempered glas vessles and vigorously stirred. The reaction was
started by addition of 0.91 mL aqueous H2O2 solution (27%,
8 mmol). Samples of 0.2 mL were taken after 5 min, 30 min and 1 h.
Samples were treated with MnO2 to destroy unreacted hydrogen
peroxide and MgSO4 to remove water. They were subsequently
filtered and afterwards diluted with dichloromethane (DCM) for GC
analysis. Analysis was carried out by calibrated GC methods in all
cases.
In this work the influence of different electron withdrawing
groups, especially halides at different positions is examined. Two
different substrates, cyclooctene and 1-octene have been chosen.
Cyclooctene was selected because of its widespread use in former
* Corresponding author. Fax: þ49 (0) 89 289 13473.
E-mail address: fritz.kuehn@ch.tum.de (F.E. Kühn).
0022-328X/$ e see front matter Ó 2011 Published by Elsevier B.V.
doi:10.1016/j.jorganchem.2011.12.010

52
P. Altmann et al. / Journal of Organometallic Chemistry 701 (2012) 51e55
Scheme 1. Reaction of hydroxy-benzaldehydes and aniline derivates to Schiff-bases.
2.2. General procedure for the synthesis of ligands
All obtained Schiff-bases react with MTO to form stable
complexes (Scheme 2). In accordance with earlier examinations the
1H NMR data show a high field shift of the methyl group of MTO
adducts of 1e14 compared to free MTO (Table 1). This observations
A solution of salicylaldehyde derivative (1.0 equiv.) in ethanol
(5.0 mL mmolꢀ1) was added into a solution of aniline derivative (1.0
equiv.) in ethanol (5.0 mL mmolꢀ1). After stirring at room
temperature for 1 h, the mixture was refluxed until complete
consumption of the starting materials. Subsequently, ethanol was
removed under reduced pressure and the obtained imines were
purified by crystallization. The obtained Schiff-bases were charac-
terized by NMR and elemental analysis. All analytical data is in
accordance with literature [21,26,30,32,33,43].
Scheme 2. Formation of MTO-Schiff-base complex.
3. Results and discussion
Table 1
Selected 1H NMR spectroscopic data for MTO-Schiff-base complexes in CDCl3.
A set of substituted Schiff-bases was synthesized (Scheme 1). All
Schiff-bases were obtained by reacting different hydroxy-
benzaldehydes with aniline derivates in ethanol according to
published procedures [21,26,30,32,33,43]. Based on former
research results, indicating that electron donating groups decrease
the catalyst activity and reduce the lifetime of the catalyst, electron
withdrawing groups, especially halides were chosen for this study.
Compound
d
(1H) MTO-CH3
Compound
d
(1H) MTO-CH3
MTO
þ2
2.68
2.60
2.62
2.62
2.62
þ7
þ11
þ12
þ13
2.62
2.62
2.62
2.62
þ4
þ5
þ6

P. Altmann et al. / Journal of Organometallic Chemistry 701 (2012) 51e55
53
Table 2
Substituting both aromatics does also not increase the catalyst’s
activity (Table 2, entries 12e14).
Cyclooctene-epoxidation yields and TOFsa catalyzed by MTO.
Lower catalyst concentrations of 0.1 mol% give 40% yield after
30 min with a TOF of 2600 hꢀ1 for 8 as ligand of MTO. When
comparing to an earlier study of Sharpless et al. who also worked
with nitromethane as solvent, activities are quite similar. Using
MTO and pyridine in a ratio of 1:1 leads to 40% yield after 5 min in
the experiments performed by Sharpless et al. A 70% yield is ob-
tained after 5 min with pure MTO and doubled catalyst concen-
tration as applied in ref. 16. Experiments with 4-tert-butylpyridine
(1:5) yielded 95% product after 5 min whereas 90% yield are re-
ported for the MTO-pyridine ratio of 1:12 [27].
Compared to other experiments with Schiff-bases, carried out
earlier in our group, the results in nitromethane are much better
than those in dichloromethane or without solvent (Table 4)
[21,32,33]. A yield of 78% is obtained with Schiff-base 1 after 30 min
compared to 4 h in earlier studies without solvent, which equals to
an acceleration by factor eight [33]. When the Schiff-base is
replaced by a 4-tert-butylpyridine ligand, the expected increase in
activity can be observed. After 5 min a yield of 95% and a TOF of
1100 hꢀ1 are reached (Table 2, entry 15). Changing the solvent from
nitromethane to dichloromethane results in a slightly lower
activity (Table 2, entries 19, 20 vs. entries 3, 4). Comparing the
nitromethane-experiments to earlier studies nitromethane seems
to be a better or at least an equal solvent if the different experi-
mental setups are taken into account (Table 4).
Entry
Ligand
MTO
[mol%]
Solvent
Yield
30 min [%]
TOF
[hꢀ1
]
1
2
3
4
5
6
7
8
e
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
100
78
99
100
99
99
99
88
100
91
99
840
650
790
840
750
720
760
660
700
670
740
770
770
710
1100
1
2
3
4
5
6
7
8
9
10
12
13
9
10
11
12
13
14
15
100
96
92
14
tbutylpyridine
100
(1:5)b
16
17
18
19
20
8
9
10
2
3
0.1
0.1
0.1
1
MeNO2
MeNO2
MeNO2
DCM
39
33
33
90
83
2600
2100
2100
670
1
DCM
490
a
TOF calculated after 5 min: [(mol epoxide)/(mol catalyst *h)].
ratio MTO:ligand.
b
proof the tethering of the Schiff-base to MTO [32,33]. Among the
Schiff-bases-adducts no significant difference between halide or
nitro (7) substitution can be observed.
To rule out solvent molecules as ligands to MTO instead of
Schiff-bases we chose nitromethane (MeNO2) as a very weak
coordinating solvent for the catalysis tests [43]. Furthermore, blank
experiments without catalyst show no formation of epoxide
neither with cyclooctene nor with 1-octene.
Using a terminal alkene like 1-octene as substrate leads to
similar results as obtained with cyclooctene with respect to the
differences between pure MTO and MTO plus Schiff-base. The yield
after 3 h ranges between 61% and 74% and the TOFs vary from
70 hꢀ1e80 hꢀ1 (Table 3). All examined systems are almost equal
except nitro-substituted compound 7 (56% yield, 50 hꢀ1) (Table 3,
entry 8). The reasons are most likely the same as described for
cyclooctene. Again, using nitromethane as solvent shows very good
results compared to other systems working with 1-octene (Table 4).
Slight differences occur between chlorine-substituted ligands
leading to an activity order o > p > m (Table 3, entries 9e11). The
highest activity can be achieved with the trichloro-substituted
compound 13 (74% yield). Interestingly, with 4-tert-butylpyridine
as additive the catalyst system decomposes rapidly. In the begin-
ning it is more active indicated by the higher TOF of 100 hꢀ1 but
after 3 h the yield remains at 31%, whereas the stability of the
Schiff-base supported catalyst lasts longer and higher conversions
For the first part of the catalysis examination, cyclooctene was
applied as substrate, as it is a very active starting material, leading
to a comparatively stable epoxide. Based on earlier experiments,
MTO and the Schiff-base-ligands have been used in a ratio of 1:1 for
this study [21,32,33]. The catalyst:substrate:oxidant ratio was
chosen to be 1:100:200 and the substrate concentration in the
reaction solution was 1 M. Hydrogen peroxide (aqueous solution
27%) was used as oxidant. All of the ligands examined with MTO
show similar results (Table 2). The turn-over-frequencies (TOFs) are
ranging between 700 hꢀ1 and 800 hꢀ1 and the yield is quantitative
after 30 min reaction time in almost every case. Only the non-
substituted Schiff-base 1 as well as the nitro-substituted Schiff-
base 7 are worse, reaching TOFs of around 650 hꢀ1 and yields of 78%
and 88%, respectively (Table 2, entries 2 and 8). As indicated in
earlier examinations, a reason for that might be the relatively good
donor ability of unsubstituted Schiff-base 1 compared to the halide
substituted Schiff-bases [33]. On the other hand, the nitro-group of
Schiff-base 7, as very strong electron acceptor, seems to withdraw
too much electron density from the Schiff-base-system to operate
the catalysis satisfactorily. The catalytic performance of MTO-
Schiff-base adducts relies most likely on an equilibrium of elec-
tron donating and electron accepting effects. If the equilibrium is
shifted too much towards one side, the activity of the catalyst
decreases.
Table 3
1-Octene epoxidation yields and TOFsa catalyzed by MTO.
Entry
Ligand
MTO
[mol%]
Solvent
Yield
3 h [%]
TOF
[hꢀ1
]
1
2
3
4
5
6
7
8
e
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
MeNO2
69
61
72
72
69
62
69
56
72
67
69
66
73
74
67
31
70
70
70
70
70
80
70
50
70
70
70
80
80
80
70
100
1
2
3
4
5
6
7
8
9
10
11
12
13
14
9
10
11
12
13
14
15
16
Chlorine and bromine substituents lead to similar results
(Table 2, entries 3e5, 9e11). The performance of the fluorine-
substituted compounds 5 and 6 is slightly lower (Table 2, entries
6 and 7). The position of the halide has no visible influence on the
fast cyclooctene-epoxidation (Table 2, entries 9e11). It does also
not matter if the halide is located at the hydroxy-benzaldehyde
moiety or at the aniline moiety (Table 2, entries 3e5, 9e11).
tbutylpyridine
(1:5)b
a
TOF calculated after 5 min: [(mol epoxide)/(mol catalyst *h)].
ratio MTO:ligand.
b

54
P. Altmann et al. / Journal of Organometallic Chemistry 701 (2012) 51e55
Table 4
Collected data concerning MTO catalyzed epoxidation of cyclooctene and 1-octene.
Ref
Ligand
Substrate c [mol/l]a
MTO [mol%]
0.5
Solvent
MeNO2
Reaction time [h]
0.5
Yield [%]
95
[27]
pyridine
cyclooctene
0.8
[27]
[16]
[33]
[33]
[38]
[37]
[45]
[17]
[46]
[21]
[20]
[23]
[23]
[45]
[17]
[21]
[46]
pyridine
cyclooctene
2.0
cyclooctene
2.3
cyclooctene
2.0
cyclootene
2.0
cyclooctene
0.9
cyclootene
1.3
cyclootene
1.1
cyclootene
3.5
cyclooctene
1.2
cyclooctene
0.2
cyclooctene
0.7
cyclooctene
0.8
cyclooctene
0.8
1-octene
1.1
1-octene
3.5
1-octene
0.2
0.5
1
DCM
2
4
99
75
3-cyanopyridine
1
e
1
e
4
70
10
1
e
4
79
4,40-dimethyl-2,20-bipyridine
pyridine
1
DCM
4
78
0.5
0.1
0.1
0.1
1
CHCl3
0.5
2
79
3-methylpyrazole
3-methylpyrazole
pyrazole
DCM
> 99
> 99
100
70
e
3
trifluoroethanol
1
Schiff-base
DCM
4
4-tert-butylpyridine
pyrazole
1
tert-butanol
0.5
2
78
0.5
0.5
0.5
0.5
1
DCM
96
pyridine
DCM
2
96
3-methylpyrazole
3-methylpyrazole
Schiff-base
DCM
8
91
e
8
62
DCM
24
21
100
100
pyrazole
1-octene
1.2
0.1
trifluoroethanol
a
If not given directly, calculated from experimental sections.
are reached. For an equal stabilization a higher amount of pyridine
is needed for the MTO/pyridine catalyst system.
Acknowledgment
PA thanks TUM-GS for financial support.
References
4. Conclusion
Several halide substituted Schiff-bases, have been examined in
epoxidation for cyclooctene- and 1-octene-epoxidation in the
presence of MTO and with H2O2 as oxidizing agent. The addition of
Schiff-bases has not a pronounced influence on the catalytic
activity and selectivity. In some cases a small increase of yield and
TOF is observed if Schiff-bases are added to the reaction solution
(Tables 2, 3). Fluorine substitution (5, 11) leads to slightly lower
yields than substitution with other halides. No significant differ-
ences occur when chlorine and bromine moieties are applied. The
yields of such systems range between 67% and 74% after 5 min
reaction time. It also has no significant influence whether the
hydroxy-benzaldehyde or the aniline moiety or both of them are
substituted. The ortho-, meta-, para-substitution at the aniline
moiety shows a reaction order of o > p > m. The exchange of
Schiff-bases as ligands by 4-tert-butylpyridine leads to a higher
conversion and turn-over-frequency if cyclooctene is used as
substrate. The epoxidation of 1-octene displays a higher activity
only early in the reaction. The overall yield after 3 h is lower than
that of MTO plus Schiff-bases. The catalytic optimum of MTO-
Schiff-base-systems is based on fine tuning the electron
donating and the electron accepting character of the Schiff-base.
The utilization of nitromethane as solvent for the catalysis
turned out to be much more rewarding than catalysis with other
or without solvents.
[1] W.A. Herrmann, F.E. Kühn, Acc. Chem. Res. 30 (1997) 169.
[2] F.E. Kühn, A.M. Santos, W.A. Herrmann, Dalton Trans. (2005) 2483.
[3] F.E. Kühn, A. Scherbaum, W.A. Herrmann, J. Organomet. Chem. 689 (2004)
4149e4164.
[4] C.C. Romão, F.E. Kühn, W.A. Herrmann, Chem. Rev. 97 (1997) 3197e3246.
[5] G.S. Owens, J. Arias, M.M. Abu-Omar, Catal. Today 55 (2000) 317.
[6] W. Adam, C.M. Mitchell, Angew. Chem. Int. Ed. 35 (1996) 533.
[7] A.M. Al-Ajlouni, J.H. Espenson, J. Organomet. Chem. 61 (1996) 3969e3976.
[8] J.H. Espenson, Chem. Commun. (1999) 479.
[9] W.A. Herrmann, R.W. Fischer, W. Scherer, M.U. Rauch, Angew. Chem. Int. Ed.
32 (1993) 1157e1160.
[10] H. Adolfsson, C. Copéret, J.P. Chiang, A.K. Yudin, J. Organomet. Chem. 65 (2000)
8651e8658.
[11] A.M. Al-Ajlouni, J.H. Espenson, J. Am. Chem. Soc. 117 (1995) 9243e9250.
[12] M. Carril, P. Altmann, W. Bonrath, T. Netscher, J. Schütz, F.E. Kühn, Catal. Sci.
Technol. (2012).
[13] J.J. Haider, R.M. Kratzer, W.A. Herrmann, J. Zhao, F.E. Kühn, J. Organomet.
Chem. 689 (2004) 3735e3740.
[14] W.A. Herrmann, R.W. Fischer, D.W. Marz, Angew. Chem. Int. Ed. 30 (1991)
1638e1641.
[15] W.A. Herrmann, R.W. Fischer, M.U. Rauch, W. Scherer, J. Mol. Catal. 86 (1994)
243e266.
[16] F.E. Kühn, A.M. Santos, P.W. Roesky, E. Herdtweck, W. Scherer, P. Gisdakis,
I.V. Yudanov, C.D. Valentin, Chem. Eur. J. 5 (1999) 3603e3615.
[17] S. Yamazaki, Tetrahedron 64 (2008) 9253e9257.
[18] W. Adam, W.A. Herrmann, J. Lin, C.R. Saha-Möller, J. Org. Chem. 59 (1994)
8281e8283.
[19] W. Adam, W.A. Herrmann, J. Lin, C.R. Saha-Möller, R.W. Fischer, J.D.G. Correia,
Angew. Chem. Int. Ed. 33 (1994) 2476e2477.
[20] P. Altmann, F.E. Kühn, J. Organomet. Chem. 694 (2009) 4032e4035.

P. Altmann et al. / Journal of Organometallic Chemistry 701 (2012) 51e55
55
[21] A. Capapé, M.-D. Zhou, S.-L. Zang, F.E. Kühn, J. Organomet. Chem. 693 (2008)
3240e3244.
[22] C. Copéret, H. Adolfsson, K.B. Sharpless, Chem. Commun. (1997) 1565e1566.
[23] E. daPalmaCarreiro, A.J. Burke, M.J.M. Curto, A.J.R. Teixeira, J. Mol. Catal. A:
Chem. 217 (2004) 69e72.
[34] M.M. Abu-Omar, P.J. Hansen, J.H. Espenson, J. Am. Chem. Soc. 118 (1996) 4966.
[35] F.E. Kühn, A.M. Santos, I.S. Gonçalves, C.C. Romão, A.D. Lopes, Appl. Organo-
met. Chem. 15 (2001) 43.
[36] W.A. Herrmann, J.D.G. Correia, M.U. Rauch, G.R.J. Artus, F.E. Kühn, J. Mol. Catal.
A: Chem. 118 (1997) 33.
[24] E. daPalmaCarreiro, G. Yong-En, A.J. Burke, J. Mol. Catal. A: Chem. 235 (2005)
285e292.
[37] W.A. Herrmann, H. Ding, R.M. Kratzer, F.E. Kühn, J.J. Haider, R.W. Fischer,
J. Organomet. Chem. 549 (1997) 319.
[25] W.A. Herrmann, R.M. Kratzer, H. Ding, W.R. Thiel, H. Glas, J. Organomet. Chem.
555 (1998) 293e295.
[38] P. Ferreira, W.-M. Xue, É Bencze, E. Herdtweck, F.E. Kühn, Inorg. Chem. 40
(2001) 5834e5841.
[26] C.-J. Qiu, Y.-C. Zhang, Y. Gao, J.-Q. Zhao, J. Organomet. Chem. 694 (2009)
3418e3424.
[39] C.D. Nunes, M. Pillinger, A.A. Valente, I.S. Gonçalves, J. Rocha, P. Ferreira,
F.E. Kühn, Eur. J. Inorg. Chem. 5 (2002) 1100.
[27] J. Rudolph, K.L. Reddy, J.P. Chiang, K.B. Sharpless, J. Am. Chem. Soc. 119 (1997)
6189e6190.
[40] M.D. Zhou, S.L. Zang, E. Herdtweck, F.E. Kühn, J. Organomet. Chem. 693 (2008)
2473.
[28] S. Stankovic, J.H. Espenson, Chem. Commun. (1998) 1579e1580.
[29] W.-D. Wang, J.H. Espenson, J. Am. Chem. Soc. 120 (1998) 11335e11341.
[30] Z. Xu, M.-D. Zhou, M. Drees, H. Chaffey-Millar, E. Herdtweck, W.A. Herrmann,
F.E. Kühn, Inorg. Chem. 48 (2009) 6812e6822.
[31] M.-D. Zhou, K.R. Jain, A. Günyar, P.N.W. Baxter, E. Herdtweck, F.E. Kühn, Eur. J.
Inorg. Chem. (2009) 2907e2914.
[32] M.-D. Zhou, Y. Yu, A. Capapé, K.R. Jain, E. Herdtweck, X.-R. Li, J. Li, S.-L. Zang,
F.E. Kühn, Chem. Asian J. 4 (2009) 411e418.
[33] M.-D. Zhou, J. Zhao, J. Li, S. Yue, C.-N. Bao, J. Mink, S.-L. Zang, F.E. Kühn, Chem.
Eur. J. 13 (2007) 158e166.
[41] Y. Gao, Y.M. Zhang, G.J. Qui, J.Q. Zhao, Appl. Organomet. Chem. 25 (2011) 54.
[42] S. Yue, J. Li, Z.H. Yu, Q. Kang, X.P. Gu, S.L. Zang, Russ. J. Coord. Chem. 36 (2010)
847.
[43] M. Carril, P. Altmann, M. Drees, W. Bonrath, T. Netscher, J. Schütz, F.E. Kühn,
J. Catal. 283 (2011) 55e67.
[44] T. Michel, D. Betz, M. Cokoja, V. Sieber, F.E. Kühn, J. Mol. Catal. A-Chem. 340
(2011) 9.
[45] S. Yamazaki, Org. Biomol. Chem. 5 (2007) 2109e2113.
[46] M.C.A. vanVielt, I.W.C.E. Arends, R.A. Sheldon, Chem. Commun. (1999)
821e822.