Strong Lewis acids of air-stable metallocene bis(perfluorooctanesulfonate)s as high-efficiency catalysts for carbonyl-group transformation reactions

Article abstract of DOI:10.1002/chem.201103874

Strong Lewis acids of air-stable metallocene bis(perfluorooctanesulfonate)s [M(Cp)₂][OSO₂C₈F₁₇] ₂·nH₂O·THF (M=Zr (2 a·3 H ₂O·THF), M=Ti (2 b·2 H₂O·THF)) were synthesized by the reaction of [M(Cp)₂]Cl₂ (M=Zr (1 a), M=Ti (1 b)) with nBuLi and C₈F₁₇SO₃H (2 equiv) or with C₈F₁₇SO₃Ag (2 equiv). The hydrate numbers (n) of these complexes were variable, changing from 0 to 4 depending on conditions. In contrast to well-known metallocene triflates, these complexes suffered no change in open air for a year. thermogravimetry-differential scanning calorimetry (TG-DSC) analysis showed that 2 a and 2 b were thermally stable at 300 and 180 °C, respectively. These complexes exhibited unusually high solubility in polar organic solvents. Conductivity measurement showed that the complexes (2 a and 2 b) were ionic dissociation in CH₃CN solution. X-ray analysis result confirmed 2 a·3 H₂O·THF was a cationic organometallic Lewis acid. UV/Vis spectra showed a significant red shift due to the strong complex formation between 10-methylacridone and 2 a. Fluorescence spectra showed that the Lewis acidity of 2 a fell between those of Sc³⁺ (λ_em=474 nm) and Fe³⁺ (λ_em=478 nm). ESR spectra showed the Lewis acidity of 2 a (0.91 eV) was at the same level as that of Sc³⁺ (1.00 eV) and Y ³⁺ (0.85 eV), while the Lewis acidity of 2 b (1.06 eV) was larger than that of Sc³⁺ (1.00 eV) and Y³⁺ (0.85 eV). They showed high catalytic ability in carbonyl-compound transformation reactions, such as the Mannich reaction, the Mukaiyama aldol reaction, allylation of aldehydes, the Friedel-Crafts acylation of alkyl aromatic ethers, and cyclotrimerization of ketones. Moreover, the complexes possessed good reusability. On account of their excellent catalytic efficiency, stability, and reusability, the complexes will find broad catalytic applications in organic synthesis. Copyright

Full text of DOI:10.1002/chem.201103874

DOI: 10.1002/chem.201103874
Strong Lewis Acids of Air-Stable Metallocene
Bis(perfluorooctanesulfonate)s as High-Efficiency Catalysts for Carbonyl-
Group Transformation Reactions
Renhua Qiu,[a] Xinhua Xu,*[a, b] Lifeng Peng,[a] Yalei Zhao,[a] Ningbo Li,[a] and
Shuangfeng Yin*[a]
6172
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182

FULL PAPER
Abstract: Strong Lewis acids of air-
stable metallocene bis(perfluorooctane-
unusually high solubility in polar or-
ganic solvents. Conductivity measure-
ment showed that the complexes (2a
and 2b) were ionic dissociation in
CH3CN solution. X-ray analysis result
confirmed 2a·3H2O·THF was a cationic
organometallic Lewis acid. UV/Vis
spectra showed a significant red shift
due to the strong complex formation
between 10-methylacridone and 2a.
Fluorescence spectra showed that the
Lewis acidity of 2a fell between those
478 nm). ESR spectra showed the
Lewis acidity of 2a (0.91 eV) was at
the same level as that of Sc3+ (1.00 eV)
and Y3+ (0.85 eV), while the Lewis
acidity of 2b (1.06 eV) was larger than
that of Sc3+ (1.00 eV) and Y3+
(0.85 eV). They showed high catalytic
ability in carbonyl-compound transfor-
mation reactions, such as the Mannich
reaction, the Mukaiyama aldol re-
sulfonate)s
[M(Cp)2][OSO2C8F17]2·
nH2O·THF (M=Zr (2a·3H2O·THF),
M=Ti (2b·2H2O·THF)) were synthe-
sized by the reaction of [M(Cp)2]Cl2
(M=Zr (1a), M=Ti (1b)) with nBuLi
and C8F17SO3H (2 equiv) or with
C8F17SO3Ag (2 equiv). The hydrate
numbers (n) of these complexes were
variable, changing from 0 to 4 depend-
ing on conditions. In contrast to well-
known metallocene triflates, these com-
plexes suffered no change in open air
for a year. thermogravimetry–differen-
tial scanning calorimetry (TG-DSC)
analysis showed that 2a and 2b were
thermally stable at 300 and 1808C, re-
spectively. These complexes exhibited
AHCTUNGTRENNGaUN ction, allylation of aldehydes, the Frie-
del–Crafts acylation of alkyl aromatic
ethers, and cyclotrimerization of ke-
tones. Moreover, the complexes pos-
sessed good reusability. On account of
their excellent catalytic efficiency, sta-
bility, and reusability, the complexes
will find broad catalytic applications in
organic synthesis.
of Sc3+ (lem =474 nm) and Fe3+ (lem
=
Keywords: carbonyl-group transfor-
mations · homogeneous catalysis ·
Lewis acids · metallocenes · sustain-
able chemistry
Introduction
other reactions due to requirement of different Lewis acidi-
ty. With the emphasis on the “green” and “sustainable”
chemistry,[7] the design and synthesis of highly catalytic effi-
cient, cheap, stable, versatile, and green Lewis acids for car-
bonyl-group transformation is a key issue in Lewis acid
chemistry.
Cationic Group 4 metallocene compounds as Lewis acid
catalysts have attracted increasing attention in modern or-
ganic synthesis recently.[8] Such Lewis acids should be as
strongly acidic as possible to acquire higher activity; howev-
er, they become more susceptible to hydrolysis upon in-
creasing the acidity. For example, metallocene triflates
Reactions of carbonyl groups that are activated by Lewis
acids are involved in several important organic transforma-
tions, such as in the synthesis of b-amino carbonyl com-
pounds, b-hydroxyl ketones, b-hydroxyl esters, allyl alcohols,
aryl ketones, multiaromatic compounds, and so forth.[1] The
Mannich reaction,[2] allylation,[3] the Mukaiyama aldol re-
ACHTUNGTRENNUNG
action,[4] the Friedel–Crafts acylation,[5] and cylcotrimeriza-
tion of ketones[6] are typical carbonyl-group transformation
reactions for the synthesis of the above compounds. Coordi-
nation of the carbonyl oxygen to a Lewis acid site induces
polarization in the molecule, effectively enhancing its re-
[M(Cp)2]ACHTNUTRGNEUNG[OSO2CF3]2 (M=Zr and Ti) were successfully em-
A
ployed as catalysts for Diels–Alder reaction,[9] Mukaiyama
aldol reaction,[10] glycosylation,[11] and so forth. Unfortunate-
ly, these complexes are not stable in air, suffering from
facile hydrolysis.[12] Thus, they must be handled under strict-
ly anhydrous conditions. Another notable species are puta-
ACHTUNGTRENNUNG
more disadvantages, such as: 1) air or moisture sensitive,
2) poor selectivity or functional intolerance, and 3) unrecov-
erable. In addition, although some catalysts show high cata-
lytic efficiency in one of these reactions, they are poor in
tive metallocene perchlorates [M(Cp)2]ACHTUNTRGENUNG[ClO4]2 (M=Hf and
Zr), which have found extremely versatile application in gly-
cosylation technology.[13] These species cannot be isolated
due to their hydrolytic instability, and thus they are usually
generated in situ prior to reaction by treating [M(Cp)2]Cl2
with potentially explosive AgClO4. Therefore, improvement
of the hygroscopic character of the cationic metallocene de-
rivatives is highly demanding from the standpoint of practi-
cal utilization as catalysts.
Recently, the use of longer perfluoroalkylsulfonate groups
as effective counter anions was found to provide air-stable
and water-tolerant organometallic species, which are in
sharp contrast to the corresponding hygroscopic organome-
tallic triflates.[6b,14] This finding has led us to postulate that
longer perfluoroalkanesulfonate groups serve to overcome
[a] Dr. R. Qiu, Prof. Dr. X. Xu, Dr. L. Peng, Dr. Y. Zhao, Dr. N. Li,
Prof. Dr. S. Yin
State Key Laboratory of Chemo/Biosensing and Chemometrics
College of Chemistry and Chemical Engineering
Hunan University, Changsha, 410082 (P.R. China)
Fax : (+86)731-88821310
E-mail: xhx1581@hnu.edu.cn
sf_yin@hnu.edu.cn
[b] Prof. Dr. X. Xu
State Key Laboratory of Elemento-Organic Chemistry
Nankai University, Tianjing, 300071 (P.R. China)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103874.
Chem. Eur. J. 2012, 18, 6172 – 6182
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6173

X. Xu, S. Yin et al.
the hydrolytic instability of the cationic organometallic spe-
cies in a general sense.[14a] With this in mind, we herein suc-
cessfully isolated two air-stable mononuclear metallocene
bis(perfluorooctanesulfonate)
complexes
[M(Cp)2]
[OSO2C8F17]2 (M=Zr, Ti),[15] which enabled facile assess-
ment of Lewis acidity and catalytic activity in carbonyl-
group transformation reactions; for example, the Mannich
reaction, allylation of aldehydes, the Muakaiyama aldol re-
action, the Friedel–Crafts acylation, and cyclotrimerization
of ketones.
Results and Discussion
There are two paths for synthesis of zirconocene and titano-
cene bis(perfluoroalkanesulfonate) [M(Cp)2][OSO2C8F17]2
(M=Zr (2a), Ti (2b)) (Scheme 1). Path A is treatment of
[M(Cp)2]Cl2 (M=Zr (1a), Ti (1b)) with AgOSO2C8F17
Scheme 1. Synthesis of zirconocene and titanocene bis(perfluoroalkane-
sulfonate)s.
(2 equiv) in THF,[15] while path B is treatment of [M(Cp)2]-
ACHTUNGTRENNUNG[nBu]2 {prepared from [M(Cp)2]Cl2 (M=Zr (1a), Ti (1b))
with nBuLi (2 equiv) in THF} with C8F17SO3H (2 equiv) in
THF; both of the two paths allowed us to isolate
2a·3H2O·THF and 2b·2H2O·THF (after recrystallization
from THF/hexane) as stable hydrated species. The hydration
Figure 1. The TG-DSC curves of complex 2a·3H2O·THF (top) and
2b·2H2O·THF (bottom).
1
number (n) of 2 was variable upon standing. H NMR spec-
troscopy (in anhydrous CH3CN) and elemental analysis
showed that freshly prepared samples after recrystallization
contained 2 or 3 water molecules along with solvating THF
for 2a and 2b. Pumping these complexes in vacuum for
a week at room temperature caused complete dehydration
giving n=0. Standing the freshly prepared compounds 2a
and 2b in open air for two days induced desolvation of the
THF and the water content increased to n=4. Such durabili-
ty, though being hydrated, presents a sharp contrast to the
loss. The endothermic step below 2208C can be assigned to
the removal of water molecules. The material is stable up to
about 3008C. Over 3008C two weight losses of exothermic
nature appear, plausibly due to the oxidation of organic en-
tities. We observed the removal of perfluorooctanesulfuryl
ligands at 4008C, leaving zirconium fluoride behind. As
shown in Figure 1 (bottom), the TG curve of 2b·2H2O·THF
shows three stages of weight loss. The endothermic step
below 1008C can be assigned to the removal of water mole-
cules. The material is stable up to about 2008C, and then
two weight losses of exothermic nature appear, plausibly
due to the oxidation of organic entities. We observed the re-
moval of perfluorooctanesulfuryl ligands at 2508C, leaving
titanium fluoride behind. We hence deduce that 2a and 2b
are thermally stable at 300 and 1808C, respectively. We also
employed 1H and 19F NMR techniques to analyze the
2b·2H2O·THF sample that had undergone thermal treat-
ment at 1808C for two days, and observed no change.
corresponding triflate [Zr(Cp)2]
dergoes facile hydrolysis to binuclear [Zr2(Cp)
OH)2]
[OSO2CF3]4·4THF.[16] Notably, 2 exhibited no sign of
ACHTUNGTRENNUNG
C
CHTUNGTRENNUNG
ACHTUNGTRENNUNG
structural change after being kept in open air for a year.
The solid samples remained as dry crystals or powder and
suffered no color change. Therefore, the metallocene per-
fluorooctanesulfonates are storable in open air, showing
a great advantage over the metallocene triflates or perchlo-
rates from an operational point of view.
As depicted in Figure 1 (top), the thermogravimetric
(TG) curve of 2a·3H2O·THF shows three stages of weight
Other notable features of complexes 2a·3H2O·THF and
2b·2H2O·THF were their unusual high solubility in MeOH,
6174
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182

Lewis Acid Chemistry
FULL PAPER
Table 1. The solubility of the mononuclear metallocene perfluorooctane-
could be attributed to the long fluoroalkyl chain in the sul-
fonate ligand. The unusual solubility reflects the amphiphilic
nature of the long fluoroalkyl chain.[14a] Upon dissolving the
metallocene complex in polar solvents, the solvent mole-
cules can approach the coordination sphere of the metal
atom to replace the hydrated water on account of their com-
patibility, and the resulting solvated species are highly solu-
ble in the same polar solvents. By contrast, CH2Cl2 as well
as hydrocarbons cannot reach the metal atoms because of
the interference of water with these hydrophobic solvents. It
can be concluded that the metallocene perfluorooctanesulfo-
nates were neither hydrophilic nor lipophilic in their hydrat-
ed form, and accordingly they were not soluble in both
strongly hydrophobic organic solvents and water, while the
water was readily replaced by the polar organic solvents to
generate the highly soluble solvated species.
The high solubility of the complexes 2a and 2b in polar
organic solution encouraged us to investigate their structure
in a solid state. We take 2a as an example. The structure of
2a·3H2O·THF in the solid state was confirmed by X-ray
analysis.[8c] The crystals suitable for the X-ray diffraction
were obtained by diffusion of hexane into saturated solution
of 2a·3H2O·THF in THF. The crystal structure and packing
together with selected bond lengths and angles are shown in
Figure 2. The zirconium atom in the cationic ion is coordi-
sulfonates in organic solvents at 258C.[a]
Solvent
Solubility [gLꢀ1
]
2a·3H2O·THF
2b·2H2O·THF
acetone
THF
EtOAc
MeOH
CH3CN
Et2O
CH2Cl2
toluene
hexane
957
244
729
186
143
21
0
0
0
275
121
19
128
530
11
0
0
0
[a] All samples were freshly prepared and recrystallized.
acetone, THF, EtOAc, and CH3CN (Table 1). In acetone,
the solubility of 2a·3H2O·THF was extraordinarily high and
amount up to 957 gLꢀ1; it is better to say that they were
miscible with each other to form a slightly viscous liquid.
Both complexes exhibited normal solubility in less polar
Et2O, but were poorly soluble in CH2Cl2. This is quite sur-
prising behavior because CH2Cl2 was usually the best sol-
vent for similar metallocene bisACTHNUTRGENUGN(triflate) derivatives. Consis-
tently, they were not soluble in much less polar toluene and
apolar n-hexane. Furthermore, as they were apparently in-
soluble in water, these complexes were hydrophobic, which
Figure 2. An ORTEP view showing 50% probability ellipsoids and packing of [Zr(Cp)2ACTHUNGTRNEUG(N H2O)3][OSO2C8F17]2·THF (2a·3H2O·THF).
Chem. Eur. J. 2012, 18, 6172 – 6182
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6175

X. Xu, S. Yin et al.
nated by three water molecules and not by THF. Hence, the
geometry of the [Zr(Cp)2A
(H2O)3]2+ ion is similar to that of
(H2O)3]2+
[Zr(Cp)2A [OSO2CF3]2 The three H2O mole-
CTHUNGTRENNUNG
2ꢀ [17]
C
R
.
cules lie on the plane that bisects the angle between the Cp
ꢀ
ring planes. The OSO2C8F17 ions and the THF molecule are
packed around the zirconium cation in such a way that their
oxygen atoms point towards the H2O ligands. The C8F17
chains of the anion, on the other hand, are clustered togeth-
er so as to produce hydrophobic domains in the crystal
structure. Unfortunately, in spite of many attempts we were
unable to obtain suitable crystals of 2b for X-ray analysis.
However, due to the similar chemical structure, we deduced
that the complex 2b could be also cationic organometallic
species which was confirmed by conductivity measurements.
We applied conductivity measurements to investigate
their ionic dissociation behavior in solution. The molar con-
ductivities (L) of 2a·3H2O·THF (L=136.1 mScmꢀ1 in
CH3CN (1.0 mmolLꢀ1) at 158C) and 2b·3H2O·THF (L=
114.5 mScmꢀ1 in CH3CN (1.0 mmolLꢀ1) at 158C) were
nearly in the same level as that of complex [Zr(Cp)2]-
ACHTUNGTRENNUNG
[(OSO2CF3)2]·3H2O·THF (L=158.1 mScmꢀ1 in CH3CN
(1.0 mmolLꢀ1) at 158C). The large molar conductivity
values are consistent with complete ionization into a 1:2
electrolyte in CH3CN, implying that these complexes are
cationic in the solid state and in solution.
Since the metallocene Lewis acid is desired to be as
strongly acidic as possible to acquire higher activity, we de-
cided to estimate the Lewis acidity of 2a and 2b. We first
applied UV/Vis spectra (Figure 3, top) to qualitatively deter-
mine the Lewis acidity of 2a, and observed a significant red
shift due to the strong complex formation between 10-meth-
ylacridone and 2a, illustrating a large Lewis acidity ability
of 2a. Previously, Fukuzumi et al. reported that the Lewis
acidity of the metal complexes can be quantitatively estimat-
ed by the red shift (lem) of Lewis acid metal ions with 10-
methylacridone on the basis of fluorescence spectra.[18] Thus,
we measured the fluorescence spectra of 10-methylacridone
and 2a to determine the lem values, which are shown in
Figure 3 (middle). The significant red shift of fluorescence
spectrum maximum due to the complex formation between
10-methylacridone and 2a (lem =476 nm) indicates that the
Figure 3. Top: UV/Vis spectral changes of complex formation between
10-methylacridone and 2a. Middle: Fluorescence spectral changes of
complex formation between 10-methylacridone and 2a. Bottom: ESR
Cꢀ
Cꢀ
spectra of a) O2 /Zr4+ (2a) and b) O2 /Ti4+ (2b).
Lewis acidity of 2a falls between those of Sc3+ (lem
=
to be capable for inducing carbon–carbon bond-forming re-
actions,[14a] it is reasonable to expect that the Lewis acidity
of 2a and 2b is high enough to trigger synthetically useful
reactions. We also employed Hammett indicator method to
determine the acidity of the complex 2a and 2b, and found
both of them have relatively strong acidity with acid
strength of 0.8<Ho ꢁ3.3 (Ho being the Hammett acidity
function). The characteristics of 2a and 2b stimulated us to
evaluate its performance as a Lewis acid catalyst for carbon-
yl-group transformation reactions, such as the solvent-free
Mannich reaction, the Mukaiyama aldol reaction, allylation
of aldehydes, the Friedel–Crafts acylation of alkylaromatic
ethers, and cylcotrimerization of ketones.
474 nm) and Fe3+ (lem =478 nm). The Lewis acidity of the
metal complexes can be also quantitatively estimated by the
binding energies (DE values) of Lewis acid metal ions with
O2 on the basis of ESR spectra.[19] To estimate the Lewis
·ꢀ
acidity of 2a and 2b exactly, we measured the ESR spectra
Cꢀ
Cꢀ
of the O2 -2a and O2 -2b complexes to determine their DE
values. As shown in Figure 3 (bottom), the DE value of the
titanium complex (O2 -2b) is significantly larger (Ti4+: gzz =
Cꢀ
2.0289, DE=1.06 eV) than that of ScACHTNURGTNE(UNG OTf)3 (gzz =2.0304,
DE=1.00 eV) and is the largest among the DE values ever
reported by Fukuzumi group.[19b] The DE of the zirconium
Cꢀ
complex (O2 -2a) exhibits a relatively large value (Zr4+
:
gzz =2.0331, DE=0.91 eV), which falls between those of Sc-
(OTf)3 and (OTf)3 (gzz =2.0349, DE=0.85 eV). Since
Lewis acids with a DE value larger than 0.88 were presumed
The Mannich reaction is one of the most useful carbonyl
transformation reactions for the synthesis of nitrogen-con-
taining compounds as well as an effective method for the
N
YACHTUNGTRENNUNG
6176
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182

Lewis Acid Chemistry
FULL PAPER
[20]
ꢀ
ꢀ
construction of C C, C N bonds. A proper control of the
three components (unmodified ketones donor, amine, and
aldehyde acceptor) of the Mannich reaction is challenging,
because side reactions such as aldol reaction could lower the
product yield. Impressive achievements have been made in
Mannich reaction over various Lewis acid or Lewis base cat-
alysts.[21] However, aldol addition and condensation products
were also observed, longer reaction times were needed, and
the use of imine is “environmentally unfriendly” and not
“atom economic”. Furthermore, the methods have disadvan-
tage(s) such as low yield, poor selectivity, and use of organic
solvent and/or need of strictly anhydrous reaction condi-
tions.
the para-position exhibits higher activity, owing to the large
electron-withdrawing ability of -NO2, which enhanced the
nucleophilic ability of the amine in Mannich reaction,
whereas that with a methyl group in the para-position shows
lower activity, since the electron-donating ability of p-CH3
reduced the nucleophilic ability of the amine in Mannich re-
action. Without the catalyst, the reaction basically does not
take place; only 8% yield of the desired product is ob-
tained.
The allylation of aldehydes and the Mukaiyama aldol re-
action mediated by a Lewis acid are among the most con-
venient processes for carbonyl-compound transformation in
organic synthesis. The addition reactions are intrinsically ef-
ficient in the production of useful building blocks such as
homoallylic alcohols[23] and b-hydroxyl ketone or ester de-
rivatives,[24] because new stereogenic centers and new
carbon–carbon bonds can be formed in a single operation.
Several efficient Lewis acid catalysts that are based on
boron, titanium, zirconium, copper, and so forth, have been
reported. In most cases, temperatures of ꢀ20 to ꢀ788C and
strictly anhydrous conditions are required.[25] In the present
investigation, we address these problems by fabricating the
extremely air-stable metallocene perfluorooctanesulfonate
catalysts 2a and 2b, and report the highly efficient catalytic
activity over the catalyst in allylation and Mukaiyama aldol
reactions of aldehydes.
It is highly desirable to develop catalytic systems that can
affect organic reactions in solvent-free conditions with ex-
cellent efficiency and stereoselectivity. Despite the recent
emphasis on solvent-free synthesis, reports on the Mannich
reaction conducted in solvent-free conditions are few.[22]
Due to the fact that the Mannich reaction only needs
a weak Lewis acidic catalyst, and with the unique characters
of 2a in mind, we performed the direct Mannich reaction in
solvent-free conditions. As expected, high efficiency and
chemoselectivity were obtained (Table 2).
Table 2. Product yields in Mannich reactions of aldehyde (4a–e) with
amine (5a–c) and acetophenone (6a) catalyzed by 2a.[a]
The reactions of aldehydes (4a–d, f–h) with nucleophiles,
such as tetrallytin (8), ketene silyl acetals (9) were examined
over 2a and 2b (Table 3). As expected, the reactions result-
ed in good yields of homoallyl alcohols (10a–h) and b-hy-
droxy ester (11a–h) derivatives in CH3CN or Et2O. A simi-
lar catalytic activity of complexes 2a and 2b was observed
when aliphatic aldehydes and aromatic aldehydes with elec-
tron-donating or electron-withdrawing groups were investi-
Entry
1[b]
2
3
4
5
6
7
8
R1
R2
Product (7)
Yield [%]
Ph (4a)
Ph (4a)
Ph (5a)
Ph (5a)
Ph (5a)
Ph (5a)
Ph (5a)
Ph (5a)
p-MeC6H4 (5b)
p-O2NC6H4 (5c)
7a
7a
7b
7c
7d
7e
7 f
7g
8
97
95
99
99
99
99
96
p-MeC6H4 (4b)
p-ClC6H4 (4c)
p-CF3C6H4 (4d)
p-O2NC6H4 (4e)
Ph (4a)
Table 3. Product yields in reactions of aldehyde (4) with nucleophiles (7,
8) catalyzed by 2a and 2b.[a]
Ph (4a)
[a] Catalyst 2a (0.05 mmol); R1CHO (4; 1.0 mmol); R2NH2 (5;
1.0 mmol); acetophenone (6a; 1.0 mmol); solvent-free conditions; RT;
5 min, isolated yield. [b] No Catalyst, 120 min.
In order to demonstrate the excellent catalytic efficiency
of 2a, we examined direct Mannich reaction of different al-
dehydes with one equivalent of aniline and acetophenone in
the presence of 5.0 mol% catalyst 2a. Aromatic aldehydes
with electron-donating and electron-withdrawing groups
were employed (Table 2, entries 2–6). Despite variation of
reaction rate, the yields are high. The aldehydes with elec-
tron-withdrawing groups (such as chloride and trifluoro-
methyl) in the phenyl plane exhibit higher reaction activity
than the aldehydes with electron-donating groups (for exam-
ple, methyl) (Table 2, entries 2–6). Two other amines were
also examined with benzaldehyde and acetophenone in the
presence of 5.0 mol% 2a under solvent-free conditions
(Table 2, entries 7 and 8). As expected, an -NO2 group in
Entry
R (4)
Yield [%]
2b
10[a]
2a
11[b]
2a
2b
1
2
3
4
5
6
7
8
Ph (4a)
10a
10b
10c
10d
10e
10 f
10g
10h
94
91
90
93
94
90
93
87
82
79
80
82
83
82
82
75
11a
11b
11c
11d
11e
11 f
11g
11h
92
92
93
94
96
90
87
83
80
81
80
84
85
80
76
70
p-MeC6H4 (4b)
p-MeOC6H4 (4 f)
p-ClC6H4 (4c)
p-CF3C6H4 (4d)
PhCH=CH (4 f)
PhCH2CH2 (4g)
C7H15 (4h)
[a] RCHO (4; 1.0 mmol); tetraallyltin (8; 0.3 mmol); 2a or 2b
(0.05 mmol); CH3CN (3.0 mL); 1 h, isolated yield. [b] RCHO (4;
1.0 mmol); ketene silyl acetals (9; 1.2 mmol); 2a or 2b (0.05 mmol);
Et2O (3.0 mL); RT, 6 h, isolated yield.
Chem. Eur. J. 2012, 18, 6172 – 6182
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6177

X. Xu, S. Yin et al.
gated. It is worth noting that 2a with a more appropriate
Lewis acidity is better than 2b for these two carbon–carbon
forming reactions, since the nucleophiles 8 and 9 are sensi-
tive to high Lewis acidic catalysts and easily decompose
rather than react with the aldehyde (4a–d, f–i).
can be considered to be an excellent catalyst in terms of
good yields and para-regioselectivity in Friedel–Crafts acyla-
tion.
The cyclotrimerization reaction of ketones has provided
a powerful strategy for the construction of polyaromatic
compounds, such as 1,3,5-trisubstitued benzene moieties or
hexasubstituted (trisannulated) benzenes.[28] The common
approach for preparation of polyaromatic compounds is the
cyclotrimerization of ketones. The reported methods to
induce cyclotrimerization of ketones involve using strong
proton acids and metallic halides as Lewis acids as promot-
ers. Other approaches include cyclotrimerization of alkynes
in the presence of organometallic reagents, for example, the
Pd-, Ir- and Ru-catalyzed multifold sequential Suzuki–
Miyaura reaction between oligohalorenes with organometal-
lic partners.[29] However, most of these procedures are often
complicated by inefficient, harsh reaction conditions, lack of
regioselectivity, formation of undesirable 1,2,4-trisubstituted
by-products, and requirement of expensive or specialized or-
ganometallic catalysts. Moreover, most of these Lewis acids
are highly sensitive to air, and cannot be reused. Thus, there
is a need to develop an air-stable, highly efficient and recy-
clable catalyst for the cyclotrimerization of ketones.
Because of the high tolerance of 2a and 2b towards hy-
drolysis, the solvents adopted in these reactions were used
as received and were not subject to any kind of drying pro-
cedure. Usually a highly reactive reagent would cause poor
chemoselectivity. For example, Grignard, lithium, and titani-
um reagents often fail to make discrimination between alde-
hydes and ketones. With 2a and 2b, however, we observed
good chemoselectivity. Ketones were less reactive: the addi-
tion reaction of acetophenone (6a) with nucleophiles 8 and
9 did not occur at room temperature, but the corresponding
homoallyl alcohol (12) was obtained in 87 or 70% yields at
658C for 24 h in the presence of 2a or 2b, respectively
The Friedel–Crafts acylation of aromatic compounds is
one of the most important reactions in organic chemistry.[5a]
Traditionally, AlCl3 is used as a Lewis acid promoter, and
stoichiometric amounts of AlCl3 are needed due to the for-
mation of stable adduct compound resulting from the inter-
action between the catalyst and the carbonyl oxygen of the
ketone product. The compound Sc
reaction and a good result was obtained; however, the high
price of Sc (OTf)4/
(OTf)3 limits its further utilization.[26] Hf
LiClO4 and NbCl5/AgClO4 were also reported as Friedel–
Crafts acylation catalysts.[27] However, larger equivalents of
potentially explosive perchloride salts have to be used as co-
catalyst. Furthermore, most of these Lewis acids are air or
moisture sensitive.
G
We assessed complex 2a (5.0 mol%) as a catalyst for the
cyclotrimerization of structurally diverse ketones in reflux-
ing toluene, and good-to-excellent yields were obtained
(Table 5). All the results showed that the reaction proceed-
ed well for a variety of acetophenones (Table 5, entries 1–6).
E
Table 5. Yields of 1,3,5-trisubstitutebenzenes synthesized from acetophe-
nones catalyzed by 2a in refluxing toluene.[a]
We assessed the 2b (5.0 mol%) catalyst for the Friedel–
Crafts acylation of structurally diverse alkyl aromatic ether
with 2.0 equivalents of acetic chloride at room temperature
in CH3CN, and good-to-excellent product yields were ob-
tained (Table 4). The electron-donating groups attached to
the aromatic ring enhanced the reaction activity. As expect-
ed, this reaction showed high regioselectivity (para-isomer
>99%). In addition, the activity of AlCl3 was markedly low
under the present mild reaction conditions. No product was
obtained in the absence of the catalyst. Thus, complex 2b
Entry RCOR’
t [h] Product (15) Yield [%][b]
1
2
3
4
5
6
7
PhC(O)CH3 (6a)
3
6
6
15a
15b
15c
98
89
85
95
93
78
85
p-Cl-C6H4C(O)CH3 (6b)
p-Br-C6H4C(O)CH3 (6c)
p-Me-C6H4C(O)CH3 (6d)
p-MeO-C6H4C(O)CH3 (6e)
p-NO2-C6H4C(O)CH3 (6 f)
cyclohexanone (6g)
3.5 15d
3
24
3
15e
15 f
15g
Table 4. Product yields in reactions of Friedel–Crafts acylation of alkyl
[a] Ketone (1.0 mmol); 2a (0.05 mmol); toluene (3.0 mL); reflux. [b] Iso-
lated yield.
aromatic ether (13) catalyzed by 2b.[a]
The acetophenones with electron-donating groups in the
para-position of the phenyl plane (e.g., methyl and methox-
yl) exhibited higher reactivity in the cyclotrimerization re-
Entry
Substrate (13)
PhOMe(13a)
PhOBu (13b)
PhOBz (13c)
o-MeC6H4OMe (13d)
o-MeC6H4OBu (13e)
Product (14)
Yield [%]
1
2
G
14a
14b
14c
14d
14e
70
65
63
62
60
AHCTUNGTREGaNNUN ction than that with electron-withdrawing groups (e.g., Cl,
Br, and NO2). It can be deduced that the electron-donating
groups enhanced the density of super-conjugated p electrons
in the phenyl plane and carbonyl group, leading to a stronger
interaction of carbonyl group of the acetophenones with the
zirconium atom of complex of 2a. We then examined the cy-
clotrimerization reaction of cyclohexanone (Table 5,
3[b]
4
5
[a] Substrate (13; 1.0 mmol); acetyl chloride (2.0 mmol); 2b (0.05 mmol);
CH3CN (5 mL); RT, 4 h; isolated yield of para-substituted product.
[b] 808C, 10 h.
6178
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182

Lewis Acid Chemistry
FULL PAPER
Table 7. Yields of Mannich reaction (4a+5a+6a!7a), allylation (4a+
8!10a) and Mukaiyama-aldol reactions (4a+9!11a) catalyzed by re-
covered catalyst 2a.[a]
entry 7). In this case, it gave the desired triannulated ben-
zene (15g) in yields of 85% for 3 h in refluxing toluene.
This fact illustrates that this method can be also applied in
cyclotrimerization of aliphatic ketones.
Cycle
Yield
[%][b]
Cat.
Yield
[%][b]
Cat.
Yield
[%][b]
Cat.
[%][c]
[%][c]
[%][c]
To show the advantage of metallocene perfluoro-
4a+5a+6a!7a
4a+8!10a
4a+9!11a
ACHTUNGTRENNUNG
1
2
3
4
5
97
99
97
99
98
96
94
94
95
94
93
99
98
97
98
97
92
90
91
91
90
99
98
96
97
96
ACHTUNGTRENNUNG
96
97
98
96
and a control cationic species were assessed for the Mannich
reaction (4a+5a+6a!7a), allylation (4a+8!10a), the
Mukaiyama aldol reactions (3a+8!10a), and for the Frie-
del–Crafts acylation (13a+AcCl!14a). The yields of the
respective reactions are compiled in Table 6. High yields
were constantly attained over 2a/2b, while the other cata-
[a] 4a (10 mmol); 5a (10 mmol) and 6a (10 mmol) or 7 (0.3 mmol) or 8
(1.2 mmol); 2a (0.5 mmol); RT. [b] Isolated yield of desired product.
[c] Isolated yield of recovered catalyst.
According to the findings reported so far,[15] the mecha-
nism of the nucleophilic reaction of carbonyl compounds,
such as acetic chloride, aldehydes, or acetophenone in the
presence of solid 2a/2b is postulated (Scheme 2). When
solid 2a/2b is added to the reaction solution, the zirconium
or titanium cationic groups dissociate occur immediately. At
this stage, there is frequent intra- and intermolecular ex-
change of substrate and organic solvent/other substrate. The
Table 6. Yields in reactions of carbonyl compounds (En) with nucleo-
philes (Nu) catalyzed by 2a/2b, 1a, and 3.[a]
Entry
En
Nu
Product
Yield [%]
2a/2b
1a
3[a]
1[a]
2[b]
3[c]
4[d]
5[f]
4a+5a
4a
4a
6a
8
9
7a
97
94
92
70
98
15
14
30
12
–
95
89
93
51[e]
–
10a
11a
14a
15a
AcCl
13a
6a
[a] 4a (1.0 mmol); 5a (1.0 mmol); 6a (1.0 mmol); 2a/2b/1a/3
(0.05 mmol); RT, 5 min; isolated yield. [b] 4a (1.0 mmol); 8 (0.3 mmol);
2a/2b/1a/3 (0.05 mmol); CH3CN (3.0 mL); RT, 1 h; isolated yield. [c] 4a
(1.0 mmol); 9 (1.2 mmol); 2a/2b/1a/3 (0.05 mmol); Et2O (3.0 mL); 08C
to RT, 6 h; isolated yield. [d] 4a (1.0 mmol); AcCl (2.0 mmol); 2a/2b/1a/
3 (0.05 mmol); CH3CN (5.0 mL); RT, 4 h; isolated yield of para-isomer.
[e] Mixture of o-and p-isomers. [f] 6a (1.0 mmol); 2a/2b (0.05 mmol); tol-
uene (3.0 mL); reflux, isolated yield.
lysts showed much lower yields, plausibly due to their lower
Lewis acidity and/or more air- and moisture-sensitive fea-
tures. However, CF3SO3H, which is considered as a powerful
catalyst by Hollis and Bosnich, can be formed through facile
hydrolysis of [Zr(Cp)2]ACTHNUTRGNEUNG
[OSO2CF3]2 (3),[10] and there is en-
hancement of catalytic activity as a result. It is worth point-
ing out that only a 35% yield of homoallyl alcohol (10a)
and no desired aldolate products (11a) were obtained in the
absence of catalyst under similar conditions, an explicit indi-
cation of the effectiveness of the catalysts. With regards to
metal sulfonate catalysis, it must be taken into account that
a sulfonic acid arising in situ during the reaction may possi-
bly work as an active species.[30] Thus, the reaction of 4a
with 8 or 9 was conducted in the presence of C8F17SO3H
(5.0 mol%). The reaction was rather complex because many
spots were detected in TLC analysis, and only 17–19% yield
of desired products were reported.
To test the reusability of the catalyst and reproducibility
of catalytic performance, 2a was subjected to several cycles
of the Mannich reaction (4a+5a+6a!7a), allylation
(4a+8!10a) and the Mukaiyama aldol reactions (4a+9!
11a) (Table 7). The detected change in product yield was
minimal in a trial of five cycles, indicating that the catalyst
is stable and suitable for reuse. Therefore, complex 2a has
the advantages of being high in activity, selectivity, stability,
and reusability.
Scheme 2. A proposed mechanism of reaction of carbonyl compounds
transformation catalyzed by 2a or 2b.
Chem. Eur. J. 2012, 18, 6172 – 6182
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6179

X. Xu, S. Yin et al.
fraction analysis was performed on a SMART-APEX diffractometer in
Shanghai Institute Organic Chemistry, China Academy of Science. The
Lewis acidity estimation was performed by means of ESR, UV/Vis and
fluorescence spectroscopy in Okayama University of Science and Osaka
University (Japan). The acidity was measured by Hammett indicator
method as described previously.[15a] Acid strength was expressed in terms
of Hammett acidity function (Ho) as scaled by pKa value of the indica-
tors.
carbonyl compound coordinates with the metal cations and
is activated. Path A: The Mannich reaction of the amine (5)
component with the activated non-enolizable carbonyl com-
pound (4) to give a hemiaminal, which after losing a mole-
cule of water to give the electrophilic intermediate
(ArCH=NAr, confirmed by NMR spectroscopy), then
reacts with the enolized carbonyl compound (6, nucleophile)
at its a-carbon atom in an aldol-type reaction to give rise to
the Mannich base (7). Due to the solvent-free condition, the
reaction is much faster than that of the other three reac-
tions. Path B: The nucleophilices of alkyl aromatic ether
(13) attack the activated carbonyl compounds (acetic chlo-
ride) to produce the desired para-alkoxylaromatic ketone
(14), while in Path C nucleophilices such as tetrallytin (8)
and ketene silyl acetals (9) attack the activated carbonyl
compounds [aldehydes (4) or ketones (6)] to produce the
corresponding homoallyl alcohols (10) and b-hydroxyl ester
(11) derivatives. Path D: This pathway involves the in situ
formation of the central benzene moiety (15) by a triple al-
dolization and dehydration of three equivalents of adoliza-
ble ketones (6). The catalysts 2a/2b are renewed with the
intake of solvating ligand or substrate. At the present stage
of investigation, we are still not sure of the transient inter-
mediates that are possibly formed during the reaction. Cur-
rent work is still being conducted for the understanding of
this aspect.
Preparation of [M(Cp)2][OSO2C8F17]2 (M=Zr, 2a; M=Ti, 2 b)
[14a]
Path A:
A
solution of AgOSO2C8F17
(1.21 g, 2.0 mmol) in THF
(10 mL) was added to a solution of [M(Cp)2]Cl2 (1.0 mmol) in THF
(20 mL). After the mixture was stirred at room temperature for an hour,
it was filtered. Dry n-hexane (40 mL) was laid over the colorless filtrate,
and after keeping in the refrigerator for 24 h, the colorless crystal was
collected. (M=Zr, 2a·3H2O·THF, 0.794 g, 65%; M=Ti, 2b·2H2O·THF,
693 mg, 54%).
Path B: A solution of C8F17SO3H (1.0 g, 2.0 mmol) in THF (10 mL) was
added to a solution of [M(Cp)2][Bu]2 in THF (20 mL) prepared from
[M(Cp)2]Cl2 (1 mmol) with nBuLi (1 mL, 1m) at ꢀ788C for 1 h. After
the mixture was stirred at ꢀ788C to room temperature for an hour, it
was evaporated. The residue was dissolved with Et2O, filtrated, evaporat-
ed, then the resulted residue was dissolved with THF (20 mL). Dry n-
hexane (40 mL) was laid over the colorless filtrate, and after keeping in
the refrigerator for 24 h, the colorless crystal was collected (M=Zr,
2a·3H2O·THF, 1.18 g, 97%; M=Ti, 2b·2H2O·THF, 1.11 g, 87%).
Data for 2a·3H2O·THF: M.p. 133–1368C; 1H NMR (400 MHz, TMS,
258C, CD3CN): d=6.65 (s, 10H; Cp), 3.78 (s, 4H; THF), 2.37 (s, nH;
H2O), 1.93 ppm (t, J=6.4 Hz, 4H; THF); 19F NMR (376 MHz, CD3CN):
d=ꢀ79.96 to ꢀ79.00 (t, J=9.0 Hz, 3F; CF3-), ꢀ112.79 (s, 2F; -CF2-),
ꢀ118.66 (s, 2F; -CF2-), ꢀ119.62 to ꢀ119.81 (d, J=71.4 Hz, 6F; -(CF2)3-),
ꢀ120.65 (s, 2F; -CF2-), ꢀ124.03 to ꢀ124.13 ppm (m, 2F; -CF2-); elemen-
tal analysis calcd (%) for 2a (C26H10F34O6S2Zr) after pumping for
a week: C 25.60, H 0.83; found: C 25.67, H 0.82; elemental analysis calcd
(%) for 2a·4H2O (C26H18F34O10S2Zr) after standing in open air for 2
days: C 24.18, H 1.40; found: C 24.34, H 1.33; X-ray data: formula:
Conclusion
C30H24F34O10S2Zr; prismatic, colorless, Mr =1345.83, 1calcd =1.905 Mgmꢀ3
,
¯
In summary, we have successfully synthesized and character-
ized the cationic mononuclear zirconocene and titanocene
perfluorooctanesulfonate complexes. They are strongly
acidic and air-stable, and show highly catalytic activity in the
Mannich reaction under solver-free conditions, allylation of
aldehydes, Mukaiyama aldol reaction, and Friedel—Crafts
acylation of alkyl aryl ethers, and could be reused. On ac-
count of their excellent catalytic efficiency as well as reusa-
bility, the complexes will find broad catalytic applications in
organic synthesis.
triclinic, P1, a=9.9295(13), b=11.9551(16), c=20.269(3) ꢁ, a=
82.833(3), b=79.491(3), g=87.397(3)8, V=2346.7(5) ꢁ3, Z=2, q=1.7–
25.58, 0.50ꢂ0.24ꢂ0.21 mm, T=293(2) K, measured reflections/independ-
ent reflections 12436/8583, Rint =0.104, h=ꢀ12–11, k=ꢀ14–11, l=ꢀ24–
22, R1 =0.095, wR2 =0.277, S=0.95.
Data for 2b·2H2O·THF: M.p. 206–2108C; 1H NMR (400 MHz, CD3CN,
TMS): d=6.95 (s, 10H; Cp), 3.62–3.67 (m, 4H; THF), 3.56 (s, nH; H2O),
1.78–1.83 ppm (m, 4H; THF). 19F NMR (376 MHz, 258C, CD3CN): d=
ꢀ79.66 to ꢀ79.75 (m, 3F; CF3-), ꢀ113.28 to ꢀ113.38 (t, J=16.5 Hz, 2F;
-CF2-), ꢀ119.36 to ꢀ119.43 (t, J=3.8 Hz, 2F; -CF2-), ꢀ120.34 to ꢀ120.54
(m, 6F; -(CF2)-), ꢀ121.36 to ꢀ121.41 (t, J=8.6 Hz, 2F; -CF2-), ꢀ124.72
to ꢀ124.87 (m, 2F; -CF2-); elemental analysis calcd (%) for 2b
(C26H10F34O6S2Ti) after pumping for a week: C 26.55, H 0.86; found: C
26.76,
H
0.86; elemental analysis calcd (%) for 2b·4H2O
(C26H18F34O10S2Ti) after standing in open air for 2 days: C 25.02, H 1.45;
found: C 25.28, H 1.42.
Experimental Section
Determination of hydration number: Molecular sieves (4 ꢁ, 11.0 g, dried
at 3558C in a muffle furnace for 5 h) were added to CD3CN (25.0 g), and
the mixture was kept under nitrogen atmosphere overnight. In this
General: The chemicals were purchased from Aldrich. Co. as well as
from other chemical companies, such as TCI (Shanghai), Alfa Aesar
(Tianjing), and Sinopharm Chemical Reagent, and were used as received
unless otherwise specified. The preparation of catalyst was carried out
under nitrogen atmosphere with freshly distilled solvents unless other-
wise noted. THF was distilled from sodium/benzophenone, while CH3CN
was distilled from CaH2. Acetone was refluxed with KMnO4 for 4 h and
distilled, and then dried with K2CO3; after that, it was distilled and kept
inside a dry box. The catalytic reactions were carried out in air, and the
solvents were used as received. NMR spectra were recorded at 258C on
an INOVA-400M spectrometer (USA) calibrated with tetramethylsilane
(TMS) as an internal reference. Elemental analyses were performed on
a VARIO EL III instrument (Germany). Conductivity was measured on
a REX conductivity meter DDS-307 (China). TG-DSC analysis was per-
formed on a NETZSCH-STA-449C instrument. X-ray single crystal dif-
1
CD3CN, water was not detected by H NMR spectroscopy. The dehydrat-
ed CD3CN was added to a freshly prepared (2a·3H2O·THF) (10.0 mg, re-
crystallized from THF/hexane (3:4) followed by drying under reduced
pressure for a week), and the solution was analyzed by 1H NMR spec-
troscopy. Then the sample was subjected to elemental analysis.
UV/Vis spectra detection of the 10-methylacridone/2a complex: The for-
mation of metal ion complexes with 10-methylacridone was examined
from the change in the UV/Vis spectra in the presence of various concen-
trations of metal ions (Mn+) by using a Hewlett–Packard 8452 A diode
array spectrophotometer. The formation constants were determined from
linear plots of (AꢀA0)ꢀ1 versus [Mn+
]
ꢀ1, in which A and A0 are the ab-
sorbance at lmax in the presence of the metal ion and the absorbance at
the same wavelength in the absence of the metal ion, respectively.
6180
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182

Lewis Acid Chemistry
FULL PAPER
Fluorescence spectra detection of the 10-methylacridone/2a complex:
The fluorescence measurement of 10-methylacridone/2a complex was
performed on a spectrofluorophotometer. The excitation wavelength of
10-methylacridone/2a or 2b complexes was 413 nm in MeCN. The MeCN
solutions were deaerated by argon purging for 7 min prior to the meas-
urements.
Typical procedure for Mukaiyama-aldol reaction of benzaldehyde (4a)
with ketene silyl acetals (9) catalyzed by 2a: Complex 2a (61 mg,
0.05 mmol) and (1-methoxy-2-methylprop-1-enyloxy)trimethylsilane (9;
208 mg, 1.2 mmol) were added to
a solution of benzaldehyde (4a;
1.0 mmol) in diethyl ether (3.0 mL) at 08C. Then the temperature was
raised to room temperature slowly. After the mixture was stirred at room
temperature for 6 h and monitored by TLC, it was subject to evaporation
in vacuum at room temperature, the residue was dissolved in n-hexane
(10 mLꢂ3) and the catalyst was collected by means of filtration for the
next cycle of reaction. To the combined hexane solution, MeOH and HCl
(aq) were added and the mixture was stirred for 15 min. NaHCO3 (aq)
was added for neutralization. The mixture was subject to evaporation,
and the solids thus obtained were dissolved in AcOEt and water. After
extraction with AcOEt (three times), the organic layer was washed with
NaCl (aq) and dried over MgSO4. After evaporation, the GLC yield was
measured. Alternatively, the residue was subject to silica gel column
chromatography (petroleum ether/ethyl acetate=10:1), colorless crystals
of 11a were obtained (191 mg, isolated yield 92%). Aldehydes such as
4a–d, and 4 f–h and nucleophiles silica enol ether 9 are commercially
available.
Cꢀ
ESR detection of O2 /2a or 2b complex: A quartz ESR tube (4.5 mm
i.d.) containing an oxygen-saturated solution of dimeric 1-benzyl-1,4 di-
hydronicotinamide (BNA)2 (1.0ꢂ10ꢀ2 m) and 2a·3H2O·THF or
2b·2H2O·THF (1.0ꢂ10ꢀ3 m) in MeCN was irradiated in the cavity of the
ESR spectrometer with the focused light of a 1000 W high-pressure Hg
lamp through an aqueous filter. Dimeric (BNA)2, which was used as an
electron donor to reduce oxygen, was prepared according to the litera-
Cꢀ
ture. The ESR spectra of O2 /2a or 2b complex in frozen MeCN were
measured at 143 K with a JEOL X-band apparatus under nonsaturating
microwave power conditions. The g values were calibrated precisely with
a Mn2+ marker, which was used as a reference.
Crystal data refinements details: Refinement of F2 was against all reflec-
tions. The weighted R factor (wR) and goodness of fit (S) were based on
F2, conventional R factors (R) were based on F, with F set to zero for
negative F2. The threshold expression of F2 >2s(F2) was used only for
calculating R factors(gt) and so forth and was not relevant to the choice
of reflections for refinement. R factors based on F2 were statistically
about twice as large as those based on F, and R factors based on all data
were even larger. Data collection: Bruker SMART; cell refinement:
Bruker SMART; data reduction: Bruker SHELXTL; program used to
solve structure: SHELXS97;[31] program used to refine structure:
SHELXL97;[31] molecular graphics: Bruker SHELXTL;[31] software used
to prepare material for publication: Bruker SHELXTL.[31] CCDC-631296
(2a·3H2O·THF) contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
Typical procedure for Friedel–Crafts acylation of anisole (13a) with
acetic chloride catalyzed by 2a: Anisole (13a) (108 mg, 1.0 mmol) was
placed in a 50 mL round-bottomed flask and a solution of catalyst
(61 mg, 0.05 mmol) in CH3CN (5.0 mL) and acetic chloride (157 mg,
2.0 mmol) was added through a syringe. Then the mixture was stirred at
room temperature for 4 h. After that, the mixture was subject to silica
gel column chromatograph; the Friedel–Crafts acylation product (14a)
was obtained: 105 mg, isolated yield 70%. Alkyl aryl ethers 13a–e and
nucleophile acetic chloride are commercially available.
All products have been reported, and the experimental details and
1H NMR spectra data of the products can be found in the Supporting In-
formation.
Typical procedure for Mannich reaction of benzaldehyde (4a) with ani-
line (5a) and acetophenone (6a) catalyzed by 2a: Catalyst 2a (61 mg,
0.05 mmol), PhCHO (106 mg, 1.0 mmol), PhNH2 (93 mg, 1.0 mmol), and
acetophenone (120 mg, 1.0 mmol) were placed in a 50 mL round-bot-
tomed flask. Then the mixture was stirred for 1 h under the TLC analysis
until the PhCHO and PhNH2, as well as the intermediate N-benzylide-
neaniline obtained from PhCHO and PhNH2, were consumed completely.
Then the solvents of the resulting mixture were removed by evaporation
in vacuum, and the residue was dissolved in CH2Cl2 and filtered. The fil-
trate was subject to volatilization at room temperature and the colorless
Acknowledgements
This work was partially supported by the National Natural Science Foun-
dation of China (21172061, 21003040, and 20973056). We thank the help-
ful discussion of Profs. J. Otera, A. Orita (Okayama University of Sci-
ence), C.-T. Au, and W.-Y. Wong (Hong Kong Baptist University). We
also thank Prof. S. Fukuzumi for the Lewis acidity estimation of 2a and
2b.
1
crystals were obtained that were suitable for H NMR measurements. An
alternative workup procedure is as follows: the mixture was extracted
with CH2Cl2 (10 mLꢂ3), and the resulting residue was subject to column
chromatograph (petroleum ether/ethyl acetate=5:1). Yield, 291 mg,
97%. Aldehydes and amines such as 4a–g, 5a–c and nucleophiles aceto-
phenone 6a are commercially available.
[1] a) A. Corma, H. Garcia, Chem. Rev. 2002, 102, 3837–3892; b) A.
Corma, H. Garcia, Chem. Rev. 2003, 103, 4307–4365.
[2] a) K. Juhl, N. Gathergood, K. A. Jorgensen, Angew. Chem. 2001,
113, 3083–3085; Angew. Chem. Int. Ed. 2001, 40, 2995–2997; b) S.
Kobayashi, R. Matsubara, Y. Nakamura, H. Kitagawa, M. Sugiura,
J. Am. Chem. Soc. 2003, 125, 2507–2515.
[3] a) K. Shibatomi, Synthesis 2010, 2679–2702; b) C. Pan, Z. Wang,
Coord. Chem. Rev. 2008, 252, 736–750; c) R. B. Kargbo, G. R.
Cook, Curr. Org. Chem. 2007, 11, 1287–1309.
[4] T. D. Machajewski, C. H. Wong, Angew. Chem. 2000, 112, 1406–
1430; Angew. Chem. Int. Ed. 2000, 39, 1352–1374.
[5] a) K. Binnemans, Chem. Rev. 2007, 107, 2592–2614; b) A. Corma,
H. Garcia, Adv. Synth. Catal. 2006, 348, 1391–1412; c) H. Gaspard-
Iloughmane, C. Le Roux, Eur. J. Org. Chem. 2004, 2517–2532.
[6] a) Y. Zhao, J. Li, C. Li, K. Yin, D. Ye, X. Jia, Green Chem. 2010, 12,
1370–1372; b) W.-B. Yi, C. Cai, X. Wang, J. Fluorine Chem. 2007,
128, 919–924; c) X. B. Jing, F. Xu, Q. H. Zhu, X. F. Ren, C. G. Yan,
L. Wang, J. R. Wang, Synth. Commun. 2005, 35, 3167–3171; d) I.
Amer, H. Schumann, V. Ravindar, W. Baidossi, N. Goren, J. Blum,
J. Mol. Catal. 1993, 85, 163–171.
Typical procedure for allylation of benzaldehyde (4a) with tetrallyltin (8)
catalyzed by 2a: Complex 2a (61 mg, 0.05 mmol) was added to a solution
of benzaldehyde (4a; 1.0 mmol) in CH3CN (3.0 mL). Tetrallyltin (8;
0.3 mmol) was then added to the mixture at room temperature. After the
mixture was stirred at room temperature for an hour and monitored by
TLC, it was evaporated in vacuum at room temperature. n-Hexane
(10 mLꢂ3) was added to the residue; the catalyst precipitated and was
recovered by filtration for the next reaction cycle. The combined n-
hexane solution was concentrated, and then MeOH and HCl (aq) was
added and stirred for 15 min. NaHCO3 (aq) was added for neutralization.
After the mixture was subject to evaporation, the as-obtained solids were
dissolved in AcOEt and water. After extraction with AcOEt (three
times), the organic layer was washed with NaCl (aq) and dried over
MgSO4. After evaporation, GLC yield was measured. Alternatively, the
residue was subject to silica gel column chromatography (petroleum
ether/ethyl acetate=8:1) and 10a was obtained as a colorless oil (139 mg,
isolated yield 94%). Aldehydes such as 4a–d and 4 f–h and nucleophiles
tetraallylatin (8) are commercially available.
[7] a) M. North, F. Pizzato, P. Villuendas, ChemSusChem 2009, 2, 862–
865; b) A. A. Kiss, A. C. Dimian, G. Rothenberg, Adv. Synth. Catal.
2006, 348, 75–81; c) R. A. Sheldon, Green Chem. 2005, 7, 267–278.
Chem. Eur. J. 2012, 18, 6172 – 6182
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6181

X. Xu, S. Yin et al.
[8] a) Y. X. Chen, C. L. Stern, S. T. Yang, T. J. Marks, J. Am. Chem. Soc.
1996, 118, 12451–12452; b) L. Jia, X. M. Yang, C. Stern, T. J. Marks,
Organometallics 1994, 13, 3755–3757; c) A. Decken, E. G. Ilyin,
H. D. B. Jenkins, G. B. Nikiforov, J. Passmore, Dalton Trans. 2005,
3039–3050; d) R. Qiu, G. Zhang, X. Xu, K. Zou, L. Shao, D. Fang,
Y. Li, A. Orita, R. Saijo, H. Mineyama, T. Suenobu, S. Fukuzumi, D.
An, J. Otera, J. Organomet. Chem. 2009, 694, 1524–1528.
[9] a) W. Odenkirk, B. Bosnich, J. Chem. Soc. Chem. Commun. 1995,
1181–1182; b) T. K. Hollis, N. P. Robinson, B. Bosnich, Organome-
tallics 1992, 11, 2745–2748; c) W. Odenkirk, A. L. Rheingold, B.
Bosnich, J. Am. Chem. Soc. 1992, 114, 6392–6398; d) T. K. Hollis,
N. P. Robinson, B. Bosnich, J. Am. Chem. Soc. 1992, 114, 5464–
5466.
[10] T. K. Hollis, B. Bosnich, J. Am. Chem. Soc. 1995, 117, 4570–4581.
[11] a) K. Ohmori, K. Hatakeyama, H. Ohrui, K. Suzuki, Tetrahedron
2004, 60, 1365–1373; b) M. I. Barrena, R. Echarri, S. Castillon, Syn-
lett 1996, 675–676; c) M. I. Matheu, R. Echarri, S. Castillon, Tetrahe-
dron Lett. 1993, 34, 2361–2364; d) M. I. Matheu, R. Echarri, S. Cas-
tillon, Tetrahedron Lett. 1992, 33, 1093–1096.
[12] T. K. Hollis, N. P. Robinson, B. Bosnich, Tetrahedron Lett. 1992, 33,
6423–6426.
[13] a) H. Kuyama, T. Nukada, Y. Ito, Y. Nakahara, T. Ogawa, Carbo-
hydr. Res. 1995, 268, C1; b) T. Matsumoto, T. Hosoya, K. Suzuki,
Synlett 1991, 709; c) P. Wipf, J. T. Reeves, J. Org. Chem. 2001, 66,
7910–7914.
[14] a) D. L. An, Z. Peng, A. Orita, A. Kurita, S. Man-e, K. Ohkubo, X.
Li, S. Fukuzumi, J. Otera, Chem. Eur. J. 2006, 12, 1642–1647; b) C.
Cai, W.-B. Yi, W. Zhang, M.-G. Shen, M. Hong, L.-Y. Zeng, Mol.
Diversity 2009, 13, 209–239; c) T. Hanamoto, Y. Sugimoto, T. Z. Jin,
J. Inanaga, Bull. Chem. Soc. Jpn. 1997, 70, 1421–1426; d) W.-B. Yi,
C. Cai, J. Fluorine Chem. 2008, 129, 524–528.
[15] a) R. Qiu, G. Zhang, Y. Zhu, X. Xu, L. Shao, Y. Li, D. An, S. Yin,
Chem. Eur. J. 2009, 15, 6488–6494; b) R. Qiu, X. Xu, Y. Li, G.
Zhang, L. Shao, D. An, S. Yin, Chem. Commun. 2009, 1679–1681;
c) R. Qiu, Y. Zhu, X. Xu, Y. Li, L. Shao, X. Ren, X. Cai, D. An, S.
Yin, Catal. Commun. 2009, 10, 1889–1892; d) R. Qiu, Y. Qiu, S.
Yin, X. Xu, S. Luo, C.-T. Au, W.-Y. Wong, S. Shimada, Adv. Synth.
Catal. 2010, 352, 153–162; e) R. Qiu, S. Yin, X. Zhang, J. Xia, X. Xu
and S. Luo, Chem. Commun. 2009, 4759–4761; f) R. Qiu, Y. Qiu, S.
Yin, X. Song, Z. Meng, X. Xu, X. Zhang, S. Luo, C.-T. Au, W.-Y.
Wong, Green Chem. 2010, 12, 1767–1771; g) R. Qiu, G. Zhang, X.
Ren, X. Xu, R. Yang, S. Luo, S. Yin, J. Organomet. Chem. 2010, 695,
1182–1188; h) J. Xia, R. Qiu, S. Yin, X. Zhang, S. Luo, C.-T. Au, K.
Xia, W.-Y. Wong, J. Organomet. Chem. 2010, 695, 1487–1492; i) R.
Qiu, S. Yin, X. Song, Z. Meng, Y. Qiu, N. Tan, X. Xu, S. Luo, F.-R.
Dai, C.-T. Au, W.-Y. Wong, Dalton Trans. 2011, 40, 9482–9489; j) R.
Qiu, Z. Meng, S. Yin, X. Song, N. Tan, Y. Zhou, K. Yu, X. Xu, S.
Luo, C.-T. Au, W.-Y. Wong, ChemPlusChem, 2012, DOI: 10.1002/
cplu.201200030.
[19] a) S. Fukuzumi, K. Ohkubo, Chem. Eur. J. 2000, 6, 4532–4535; b) K.
Ohkubo, S. C. Menon, A. Orita, J. Otera, S. Fukuzumi, J. Org.
Chem. 2003, 68, 4720–4726.
[20] a) K. Mikami, S. Fustero, M. Sanchez-Rosello, J. Luis Acena, V. So-
loshonok, A. Sorochinsky, Synthesis 2011, 3045–3079; b) G. Della
Sala, A. Russo, A. Lattanzi, Curr. Org. Chem. 2011, 15, 2147–2183;
c) A. Bartoli, F. Rodier, L. Commeiras, J.-L. Parrain, G. Chouraqui,
Nat. Prod. Rep. 2011, 28, 763–782; d) B. Weiner, W. Szymanski,
D. B. Janssen, A. J. Minnaard, B. L. Feringa, Chem. Soc. Rev. 2010,
39, 1656–1691.
[21] a) B. M. Trost, J. Jaratjaroonphong, V. Reutrakul, J. Am. Chem. Soc.
2006, 128, 2778–2779; b) B. M. Trost, L. R. Terrell, J. Am. Chem.
Soc. 2003, 125, 338–339; c) J. W. Yang, C. Chandler, M. Stadler, D.
Kampen, B. List, Nature 2008, 452, 453–455; d) J. W. Yang, M. Sta-
dler, B. List, Nat. Protoc. 2007, 2, 1937–1942; f) B. List, Acc. Chem.
Res. 2004, 37, 548–557; e) T. B. Poulsen, C. Alemparte, S. Saaby, M.
Bella, K. A. Jørgensen, Angew. Chem. 2005, 117, 2956–2959;
Angew. Chem. Int. Ed. 2005, 44, 2896–2899.
[22] M. A. P. Martins, C. P. Frizzo, D. N. Moreira, L. Buriol, P. Machado,
Chem. Rev. 2009, 109, 4140–4182.
[23] a) Y. Nagano, A. Orita, J. Otera, Adv. Synth. Catal. 2003, 345, 643–
646; b) P. Sreekanth, S.-W. Kim, T. Hyeon, B. M. Kim, Adv. Synth.
Catal. 2003, 345, 936–938; c) A. Yanagisawa, H. Inoue, M. Moro-
dome, H. Yamamoto, J. Am. Chem. Soc. 1993, 115, 10356–10357;
d) X. Zhang, D. H. Chen, X. H. Liu, X. M. Feng, J. Org. Chem.
2007, 72, 5227–5233.
[24] a) H. Ishitani, Y. Yamashita, H. Shimizu, S. Kobayashi, J. Am.
Chem. Soc. 2000, 122, 5403–5404; b) S. Kobayashi, M. Ueno, S.
Saito, Y. Mizuki, H. Ishitani, Y. Yamashita, Proc. Natl. Acad. Sci.
USA 2004, 101, 5476–5481; c) J. J. Song, Z. Tan, J. T. Reeves, N. K.
Yee, C. H. Senanayake, Org. Lett. 2007, 9, 1013–1016; d) M. Takeu-
chi, R. Akiyama, S. Kobayashi, J. Am. Chem. Soc. 2005, 127, 13096–
13097.
[25] R. Mahrwald, Chem. Rev. 1999, 99, 1095–1120.
[26] A. Kawada, Bull. Chem. Soc. Jpn. 2000, 73, 2325–2333.
[27] a) S. Arai, Y. Sudo, A. Nishida, Tetrahedron 2005, 61, 4639–4642;
b) I. Hachiya, M. Moriwaki, S. Kobayashi, Tetrahedron Lett. 1995,
36, 409–412.
[28] A. W. Amick, L. T. Scott, J. Org. Chem. 2007, 72, 3412–3418.
[29] a) A. W. Fatland, B. E. Eaton, Org. Lett. 2000, 2, 3131–3133; b) F. C.
Pigge, F. Ghasedi, Tetrahedron Lett. 2000, 41, 6545–6549.
[30] R. Dumeunier, I. E. Marko, Tetrahedron Lett. 2004, 45, 825–829.
[31] a) G. M. Sheldrick, SADABS, An Empirical Absorption Correction
Program for Area Detector Data, University of Gçttingen, Gçttin-
gen, 1996; b) G. M. Sheldrick, SHELXS-97, University of Gçttingen,
Gçttingen, 1997; c) SMART, Version 5.628; Bruker AXS Inc.;
d) SAINTB, Version 6.22a; Bruker AXS Inc.; e) SHELXTL NT/
2000, Version 6.1.[1]; f) G. M. Sheldrick, Acta Cryst., Sect. A, 2008,
64, 112–122..
[16] W. Lasser, U. Thewalt, J. Organomet. Chem. 1986, 311, 69–75.
[17] U. Thewalt, L. Wiltrand, J. Organomet. Chem. 1984, 276, 341–348.
[18] S. Fukuzumi, K. Ohkubo, J. Am. Chem. Soc. 2002, 124, 10270–
10271.
Received: December 10, 2011
Published online: April 13, 2012
6182
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 6172 – 6182