Ferrocene-stabilized silicon cations as catalysts for diels-alder reactions: Attempted experimental quantification of lewis acidity and reactIR kinetic analysis

Article abstract of DOI:10.1021/om401040y

The ²⁹Si NMR chemical shifts of ferrocene-stabilized silicon cations span a wide range depending on the substituents at the silicon atom. These pronounced differences in deshielding of the silicon atom do not translate into significant differences in their catalytic activity in Diels-Alder reactions. It was shown by Lewis pair formation with Lewis base probes (Et ₃PO and pyridine-d₅) that there is hardly any difference between these silicon cations after coordination to a Lewis base. This finding not only thwarts experimental quantification of the Lewis acidity of the free Lewis acids but also demonstrates that the reactivity differences are largely due to steric effects for a given counteranion. These observations are further verified by a ReactIR kinetic analysis. The Lewis acidity of silicon cations and their performance as catalysts cannot be correlated with ²⁹Si NMR chemical shifts as well as resonances of adducts with Lewis base probes, not even for a subset of silicon Lewis acids.

Full text of DOI:10.1021/om401040y

Article
pubs.acs.org/Organometallics
Ferrocene-Stabilized Silicon Cations as Catalysts for Diels−Alder
Reactions: Attempted Experimental Quantification of Lewis Acidity
and ReactIR Kinetic Analysis
Alexander R. Nodling,†,‡ Kristine Muther,† Volker H. G. Rohde,† Gerhard Hilt,*,‡
̈
̈
and Martin Oestreich*,†
†Institut fur Chemie, Technische Universitat Berlin, Strasse des 17. Juni 115, 10623 Berlin, Germany
̈
̈
‡Fachbereich Chemie, Philipps-Universitat Marburg, Hans-Meerwein-Strasse, 35043 Marburg, Germany
̈
S
* Supporting Information
ABSTRACT: The 29Si NMR chemical shifts of ferrocene-
stabilized silicon cations span a wide range depending on the
substituents at the silicon atom. These pronounced differences
in deshielding of the silicon atom do not translate into
significant differences in their catalytic activity in Diels−Alder
reactions. It was shown by Lewis pair formation with Lewis
base probes (Et3PO and pyridine-d5) that there is hardly any
difference between these silicon cations after coordination to a
Lewis base. This finding not only thwarts experimental
quantification of the Lewis acidity of the free Lewis acids
but also demonstrates that the reactivity differences are largely due to steric effects for a given counteranion. These observations
are further verified by a ReactIR kinetic analysis. The Lewis acidity of silicon cations and their performance as catalysts cannot be
correlated with 29Si NMR chemical shifts as well as resonances of adducts with Lewis base probes, not even for a subset of silicon
Lewis acids.
Scheme 1. Diels−Alder Reaction Catalyzed by Silicon-Based
Lewis Acids
INTRODUCTION
■
The performance of tetracoordinate silicon-based Lewis acids
of the general formula R3SiX as catalysts largely depends on the
X group (typically Cl, OTf, and NTf2) and, to lesser extent, on
the R groups (usually alkyl or aryl groups).1,2 A particularly
striking example is the reactivity difference of Me3SiOTf (4)
and Me3SiNTf2 (5) in low-temperature Diels−Alder reactions.
It was Ghosez and co-workers who demonstrated that 5
catalyzes Diels−Alder reactions at 0 °C where conventional 4 is
not sufficiently potent (e.g., 1 + 2 → 3, Scheme 1).3 Sawamura
and co-workers later showed that [Et3Si(toluene)]+[B-
(C6F5)4]− (6) is an even better catalyst for the same
transformation (1 + 2 → 3, Scheme 1).4 The coordinating
ability order of the involved counteranions X− is TfO− > Tf2N−
> [B(C6F5)4]−, and that greatly enhances the Lewis acidity at
the silicon atom in reverse order.2 Also, the silicon atom in 6 is
cationic, and the neutral arene solvent lends stabilization to an
otherwise tricoordinate, tetravalent silicon cation. Such solvent-
stabilized silylium ions are nevertheless exceptionally strong
Lewis acids.2,5 Although it is not allowed to directly correlate
the 29Si NMR chemical shift with the Lewis acidity of the
silicon atom, the degree of deshielding is still believed to be a
good qualitative measure of its Lewis acidity (43.5 ppm for 4 <
55.9 ppm for 5 < 81.8 ppm for 6).6
→ 3, Scheme 1).8 We were recently able to prepare and
characterize 10 additional new members of this family (7b−7k,
Figure 1).9 These span a relatively wide chemical shift range of
77.4−120.9 ppm. We assumed that a maximum Δδ value of
43.5 ppm would also be reflected in different Lewis acidities
Our laboratory had introduced the ferrocene-stabilized
silicon cation 7a,7 which emerged as an excellent catalyst for
the above and other challenging Diels−Alder reactions (1 + 2
Received: October 25, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
coordination of a strong Lewis acid to the oxygen atom of 8,
but absolute values are almost identical (δ 89.5 1.5 ppm).
The adduct of [Et3Si(1,2-Cl2C6H4)]+[B(C6F5)4]− 6′·8 (29Si
NMR: δ 101.3 ppm) with the same counteranion fits also into
this scale (Table 1, entry 7). Even the chemical shifts found for
5·8 and 4·8 generated from nonionic 5 and 4 are of the same
order of magnitude (Table 1, entries 8 and 9). This is in stark
contrast to the large range of 29Si NMR chemical shifts of
silicon cations 7 (δ 77.4−120.9 ppm, cf. Figure 1). The 31P{1H}
NMR chemical shifts show no correlation with other
spectroscopic or structural data. Hence, determination of the
Lewis acidity by the Gutmann−Beckett method is not
applicable within our family of silylium ions. Differences in
the 29Si NMR chemical shift as well as in the Lewis acidity are
largely controlled by the degree of interaction between the
cationic silicon atom and the ferrocene backbone. Coordination
of phosphine oxide 8 (or any other Lewis base) cancels that
interaction, and the thus-formed cations with a tetracoordi-
nated silicon atom do not show pronounced differences in the
31P{1H} NMR chemical shifts, as the substituent effects alone
are relatively minor.
Figure 1. The family of ferrocene-stabilized silicon cations: 29Si NMR
chemical shifts.
and, hence, catalytic activity. We, therefore, embarked on an
experimental verification of the Lewis acidities of ferrocene-
stabilized silicon cations 7 in comparison with 4−6 by adduct
formation with Lewis base probes and by kinetic measurements
of a selected Diels−Alder reaction.10
RESULTS AND DISCUSSION
■
As a consequence of the poor validity of the Gutmann−
Beckett method, we moved to Lewis base probes 10-d1 and 11-
d5 for the quantification of Lewis acidity, which were
introduced by Hilt (Figure 2).15,16 Coordination of either the
Quantification of Lewis Acidity. To assess the Lewis
acidity of 7,12 we chose the established Gutmann−Beckett
method where Lewis pair formation with triethylphosphine
oxide 8 as Lewis base is used to determine relative Lewis
acidities.13 The degree of deshielding in the 31P NMR spectrum
related to free 8 (δ 47.6 ppm in 1,2-Cl2C6D4) is a measure of
the Lewis acidity. The 29Si and 31P{1H} NMR chemical shifts of
selected adducts 914 are summarized in Table 1 (entries 1−6;
Table 1. Gutmann−Beckett Analysis: Adducts with
ab
,
Triethylphosphine Oxide (8) as Lewis Base Probe
15
16
Figure 2. Lewis base probes 10-d1 and 11-d5 introduced by Hilt.
tertiary amine or the pyridine nitrogen atom to the Lewis acid
had been shown to result in deshielding of 2H resonance signals
depending on the strength of the Lewis acid.
29Si NMR
(ppm)
31P{1H} NMR
(ppm)
Amine 10-d1 was too sterically hindered to form stable
adducts with ferrocene-stabilized silicon cations 7. Decom-
position was usually observed, and that was in agreement with
previous findings from our laboratory.8b The use of non-
hindered 11-d5 was more promising, as Manners and co-
workers had already reported the preparation of stable pyridine
adducts of 7 (29Si NMR: δ 36.2 ppm in CD2Cl2 for 12e and δ
entry compound
R1
R2
c
1
2
3
4
5
6
7
8
9
9a
tBu
tBu
iPr
Ph
Fc
Me
tBu
iPr
tBu
Fc
27.2
88.7
87.9
88.4
90.5
88.0
89.0
88.8
91.0
92.5
c
9d
9e
25.0
c
24.3
c
9f
14.3
d
9h
9i
13.5
25.2 ppm in CD2Cl2 for 12h).11 Moreover, Hilt and Nodling
c
Fc
Me
17.8
̈
d
had employed 11-d5 to quantify the Lewis acidities of several
triorganosilyl triflates and had found a qualitative correlation
6′·8
5·8
4·8
35.8
c
32.5
2
c
32.3
between the deshielding of the Hpara nucleus and the rate
a
constants of a Diels−Alder reaction catalyzed by these silicon
Lewis acids.16 Not surprisingly, the targeted adducts 12 cleanly
formed with six selected silicon cations 714 (Table 2, entries 1−
6). 29Si NMR chemical shifts were again diagnostic of Lewis
pair formation. However, the Δδ values, particularly those of
the Hpara nucleus, obtained from the H NMR measurements
did not show the same trends as seen for the aforementioned
screening of triorganosilyl triflates.16 Δδ(2Hpara) values are
within less than 0.1 ppm, even for 6′·11-d5 and 5·11-d5 (Table
2, entries 7 and 8). Also, Δδ(2Hortho) values were not positive
throughout; negative values correspond to a shielding of the
2Hortho nucleus. The Δδ(2Hmeta) values were distinguishable but
Lewis pairs generated according to General Procedures 1 (for 9 and
b
6′·8) and 2 (for 5·8 and 4·8). 31P{1H} NMR chemical shift of 8 in
c
d
1,2-Cl2C6D4: δ 47.6 ppm. 29Si DEPT NMR. 1H,29Si HMQC NMR.
2
2
for the NMR spectroscopic characterization of 9b−9c, 9g, and
9j−9k, see the Supporting Information). The 29Si NMR
chemical shifts δ 13.5−27.2 ppm of adducts 9 are clear evidence
for the formation of a tetracoordinated silicon atom, and the
2JSi,P coupling constants ranging from 12.5 to 21.1 Hz obtained
from 29Si DEPT NMR spectra are further proof of Lewis pair
formation. The 31P{1H} NMR chemical shifts of 9 indicate
B
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
of trans-14 was observed (Figure 3). It is apparent that full
conversion is usually reached within half an hour. Neither the
Table 2. Adducts with Pyridine-d5 (11-d5) as Lewis Base
Probe
a
bc
,
2
Δδ in H NMR
29Si NMR
(ppm)
entry compound
R1
R2
ortho
meta
para
d
1
2
3
4
5
6
7
8
9
12a-d5
12d-d5
12e-d5
12f-d5
tBu
tBu
iPr
Ph
Fc
Me
−0.41 0.38
0.60
0.67
0.57
0.60
0.61
0.58
0.61
0.64
0.28
36.8
d
tBu 0.57
iPr −0.36 0.44
tBu −0.37 0.38
0.65
35.0
d
34.5
d
23.4
d
12h-d5
12i-d5
Fc
0.62
0.60
22.8
d
Fc
Me
0.03
0.47
26.8
d
6′·11-d5
5·11-d5
4·11-d5
−0.40 0.54
−0.04 0.67
−0.07 0.24
42.6
Figure 3. Comparison of the kinetic profiles of the model Diels−Alder
reaction catalyzed by various silicon Lewis acids.
e
41.9
e
42.7
a
Lewis pairs generated according to General Procedures 3 (for 12-d5
technical equipment nor the solvent (1,2-Cl2C6H4 solidifies at
−18 °C, and 7 are not stable above −40 °C in CH2Cl2) allowed
running these reactions at lower temperature. The setup was
extremely sensitive toward minor variations of the temperature.
The cycloaddition even occurs instantaneously with Me3SiNTf2
(5) and [Et3Si(1,2-Cl2C6H4)]+[B(C6F5)4]− (6′), and 5 and 6′
are better catalysts than 7 at 12.7 °C.17 The kinetic profiles of
the catalyses with ferrocene-stabilized silicon cations 7 are all
different and might be grouped into those of silicon cations
with and without additional ferrocenyl groups (7h and 7i versus
7a, 7d, 7e, and 7f). Within this family, the steric demand of R1
and R2 might account for the different kinetic profiles: 7a (R1 =
tBu and R2 = Me) is by far the best catalyst. As discussed above,
the initial deshielding of the silicon atom is canceled by
coordination of a Lewis base, here dienophile E-13. 7f (R1 =
tBu and R2 = Ph, δ 98.6 ppm) is the better catalyst than more
hindered 7d (R1 = tBu and R2 = tBu, δ 120.9 ppm) but
significantly less deshielded. Currently, we cannot explain the
course of the curve for catalysts 7h and 7i with more than one
ferrocenyl substituent but speculate that the excellent
stabilization and secondary effects9 exerted by the electron-
rich ferrocenyl groups makes them poorer Lewis acids.
No simple overall reaction order could be deduced from
these kinetic profiles.18 We think that this is a reflection of the
unclear mechanism of Diels−Alder reactions catalyzed by
highly Lewis acidic silicon cations. Our group proved that the
Diels−Alder reaction is indeed stepwise and diastereoconver-
gent; both E-13 and Z-13 yield trans-14 with d.r. = 99:1.19
Moreover, Prakash, Olah, and co-workers had shown for the
silylcarboxonium/silyloxycarbenium ion of cyclohex-2-enone
that the positive charge accumulates at the β-position.20 We
interpret these findings in support of a polar stepwise rather
than a concerted mechanism.3 We exclude here the
intermediacy of pentacoordinate silicon cations, as there was
no 29Si NMR spectroscopic evidence when treating 7a with
excess Lewis base (benzophenone or acetonitrile).
b
and 6′·11-d5) and 4 (for 5·11-d5 and 4·11-d5). Δδ values relative to
c
2
free 11-d5 (for H NMR chemical shifts, see Figure 2). Line widths
were determined for the resonance signal of the deuteron in the meta
position (ortho for 4·11-d5): 17.7 Hz (average) for 12-d5 and 4.8 Hz
d
e
(average) for 4·11-d5−6′·11-d5. 1H,29Si HMQC NMR. 29Si DEPT
NMR.
did not allow for any correlation with kinetic data (vide infra).
As with the Gutmann−Beckett analysis, these results are
inconclusive.
Kinetic Analysis. The marked chemical shift differences in
the 29Si NMR spectra of ferrocene-stabilized silicon cations 7
did not translate into meaningful trends in the Lewis pair
formation with 31P{1H} NMR (Table 1) or 2H NMR (Table 2)
probes.12 We had observed though that Lewis acids 7 catalyze
Diels−Alder reactions at slightly different rates, and it therefore
appeared reasonable to evaluate the reactivity of selected 7 in a
kinetic analysis of a representative Diels−Alder reaction. The
CO group in the various dienophiles tested in the past8
suggested in situ IR spectroscopy with a ReactIR as a suitable
tool for monitoring these air- and moisture-sensitive reactions.
We chose the Diels−Alder reaction of cyclohexa-1,3-diene (1)
and chalcone (E-13) as a model reaction (1 + E-13 → trans-14,
Scheme 2). The carbonyl stretching bands of E-13 and trans-14
Scheme 2. Model Diels−Alder Reaction for the ReactIR
Analysis
were sufficiently separated, and this rather difficult Diels−Alder
reaction would not be catalyzed by protons8b at the reaction
temperature (approximately 15 °C) required for our ReactIR
measurements.
The reaction progress was constantly monitored until no
further increase in the absorption intensity of the carbonyl band
As quantification of the reactivity of 7 from rate constants
was not possible, we determined the initial rates (Figure 4 and
Table 3) and turnover frequencies (TOFs) at 50% and 95%
conversion (Table 3) to establish a reactivity order.
Comparison of the initial rates results in a relative order 7a
C
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Lewis acid (Figure 1). Steric and to a lesser extent electronic
effects govern the reactivity, as verified by ReactIR kinetic
measurements. The key finding of this work is that the Lewis
acidity of silicon cations and their performance as catalysts
cannot be correlated with 29Si NMR chemical shifts nor with
those of adducts with various Lewis base NMR probes, not
even for a small subset of silicon Lewis acids.
EXPERIMENTAL SECTION
■
General Remarks. All reactions were performed in flame-dried
glassware using a glovebox (O2 < 0.5 ppm, H2O < 0.5 ppm) or
conventional Schlenk techniques under a static pressure of argon or
nitrogen. Solvents and solutions were transferred with syringes.
Benzene was purified and dried using a solvent system. Toluene was
dried over CaH2 and stored over molecular sieves. 1,2-Cl2C6H4 and
1,2-Cl2C6D4 were dried over CaH2 prior to use and stored over
molecular sieves in a glovebox. Cyclohexa-1,3-diene (1) was distilled
from NaBH4. E-Chalcone (E-13) was recrystallized from ethanol and
dried by azeotropic distillation with benzene or toluene. Ferrocenyl-
substituted silanes, i.e., precursors of silylium ions 7a−7k, were
prepared according to previously reported procedures and dried by
azeotropic distillation with benzene.9 Triethylsilane and allyltrime-
thylsilane were distilled from CaH2 or LiAlH4 and stored over
molecular sieves. [Ph3C]+[B(C6F5)4]− was prepared according to a
reported procedure, recrystallized from CH2Cl2/n-pentane, and stored
in a glovebox.21 Triethylphosphine oxide (8) was used as received and
stored in a glovebox. Quinolizidine-d1 (10-d1) was prepared according
to a reported procedure,15 and pyridine-d5 (11-d5) was dried over
CaH2; both were stored over molecular sieves. Me3SiOTf (4) was
purified by fractional distillation under an inert atmosphere in the
presence of Me4Si. Me3SiNTf2 (5) was obtained from commerical
sources or prepared according to a known procedure from
allyltrimethylsilane3b and used immediately thereafter. 1H, 2H, 11B,
19F, 29Si, and 31P{1H} NMR spectra were recorded in 1,2-Cl2C6H4 and
Figure 4. Determination of the initial rates (cf. Scheme 2).
≈ 7i > 7e ≈ 7f > 7h ≈ 7d (column 3). Similar trends are seen
with the TOFs at different conversion: 7a > 7i ≈ 7e ≈ 7f > 7d
> 7h (column 4) and 7a > 7f ≈ 7e ≈ 7d > 7i (column 5). We
interpret these data to mean that the different reactivities of 7
are largely due to steric effects. The performance of catalysts 7h
and 7i with multiple ferrocenyl substituents at higher
conversion is not understood. It is worthy of note that 7a
(R1 = tBu and R2 = Me), which we selected almost arbitrarily
when we had started this chemistry five years ago,8a indeed is
the best catalyst within the family of ferrocene-stabilized silicon
cations.
CONCLUSION
■
The initial purpose of this study was an experimental
quantification of the Lewis acidity of our ferrocene-stabilized
silicon cations 712 by Lewis pair formation utilizing the
methods established by Gutmann and Beckett13 (with Et3PO as
1,2-Cl2C6D4 at the Westfalische Wilhelms-Universitat Munster, the
̈
̈
̈
Technische Universitat Berlin, and the Philipps-Universitat Marburg.
̈
̈
Chemical shifts are reported in parts per million (ppm) and are
referenced to the residual solvent resonance as the internal standard
16
2
a
31P NMR probe) and Hilt (with pyridine-d5 as a H NMR
1
(1,2-Cl2C6D3H: δ 6.94 and 7.20 ppm for H NMR; 1,2-Cl2C6DH3: δ
probe). We soon learned that the wide range of 29Si NMR
6.94 and 7.20 ppm for 2H NMR). Data are reported as follows:
chemical shift, multiplicity (br s = broad singlet, s = singlet, d =
doublet, t = triplet, q = quartet, sept = septet, br m = broad multiplet,
m = multiplet, mc = centrosymmetric multiplet), coupling constants
chemical shifts of 7 does not translate into significant
2
differences in the 31P NMR and H NMR chemical shifts,
respectively, in the Lewis pairs 9 (Table 1) and 12 (Table 2).
The adduct formation cancels the interaction between the
cationic silicon atom and the ferrocene backbone, thereby
reducing pronounced to minor differences that are simply due
to the steric environment around the silicon atom. When
applying 7 as catalysts in Diels−Alder reactions, it is the
dienophile that slips into the role of the Lewis base. For that
reason, it now comes as no surprise that Lewis acids 7 display
similar catalytic activity in Diels−Alder reactions despite
markedly different deshielding of the silicon atom in the free
1
(Hz), and integration. H,29Si HMQC NMR spectra are measured
3
with a coupling constant of 7.0 Hz for the JH,Si coupling. The peak
intensities in the 1H,29Si HMQC NMR spectra cannot be correlated to
the amount of compound. Gas liquid chromatography (GLC) was
performed on a capillary column (30 m × 0.32 mm, 0.25 μm film
thickness) using the following programs: N2 carrier gas, injection
temperature 240 or 250 °C, detector temperature 300 °C, flow rate
1.74 mL/min or 1.70 mL/min; temperature program: start temper-
ature 40 °C, heating rate 10 °C/min, end temperature 280 °C for 10
min. In situ FT-IR spectroscopy was performed using a ReactIR with
Table 3. Average Initial Rates and Turnover Frequencies (cf. Scheme 2 and Figure 4)18
entry
Lewis acid
initial rate (mol·L−1·s−1
)
TOF at 50% conv (s−1
)
TOF at 95% conv (s−1
)
1
2
3
4
5
6
7
8
7a
7d
7e
7f
7h
7i
(1.23 0.10) × 10−3
(0.43 0.10) × 10−3
(0.86 0.13) × 10−3
(0.84 0.22) × 10−3
(0.53 0.03) × 10−3
(1.17 0.16) × 10−3
n.d.
(4.08 0.29) × 10−2
(1.45 0.29) × 10−2
(2.32 0.25) × 10−2
(2.18 0.27) × 10−2
(0.69 0.17) × 10−2
(2.82 0.37) × 10−2
37.2 × 10−2
(2.86 0.20) × 10−2
(1.07 0.26) × 10−2
(1.22 0.19) × 10−2
(1.46 0.28) × 10−2
a
(0.63 0.27) × 10−2
32.0 × 10−2
39.3 × 10−2
6′
5
n.d.
37.6 × 10−2
a
Full conversion required 16 h at room temperature. n.d. = not determined.
D
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
DS AgX FiberConduit technology applying DiComp diamond-based
ATR sensors. The probe head was cleaned before every use by
sonication at 40 °C in distilled toluene followed by rinsing with
distilled CH2Cl2 afterward.
triphenylmethane formed in silicon cation generation was used as
internal standard.
tert-Butylferrocenylmethylsilylium Tetrakis(penta-
fluorophenyl)borate Triethylphosphine Oxide Adduct (9a).
This was prepared from tert-butylferrocenylmethylsilane (5.00 mg,
17.5 μmol, 1.00 equiv), [Ph3C]+[B(C6F5)4]− (16.1 mg, 17.5 μmol,
1.00 equiv), and triethylphosphine oxide (8, 2.30 mg, 17.1 μmol, 0.980
General Procedure 1: Preparation of Triethylphosphine
Oxide Adducts 6′·8 and 9 of Silylium Ions 6′ and 7. In a
glovebox, a solution of the requisite silane (1.00 equiv) in 1,2-Cl2C6D4
(0.25 mL) was added to a suspension of [Ph3C]+[B(C6F5)4]− (1.00
equiv) in 1,2-Cl2C6D4 (0.15 mL) in an 8 mL vial equipped with a
magnetic stir bar. The resulting red-brown solution was stirred for 1
min, and a solution of triethylphosphine oxide (8, 0.870−0.980 equiv)
in 1,2-Cl2C6D4 (0.25 mL) was added. The sample was transferred to
an NMR tube and directly subjected to NMR spectroscopic analysis.
General Procedure 2: Preparation of Triethylphosphine
Oxide Adducts 4·8 and 5·8 of Trimethylsilyl Precursors 4 and
5. In a glovebox, triethylphosphine oxide (8, 0.806−0.895 equiv) was
weighed into a vial equipped with a magnetic stir bar and sealed with a
septum. The vial was transferred out of the glovebox and connected to
a Schlenk line, and 1,2-Cl2C6D4 (0.60 mL) was added. 4 or 5 (1.00
equiv) was added, and the solution was stirred for 1 min. The sample
was transferred to an NMR tube, the vial was washed with 1,2-Cl2C6H4
(0.20 mL), and the washings were transferred to the NMR tube. The
sample was directly subjected to NMR spectroscopic analysis.
General Procedure 3: Preparation of Pyridine-d5 Adducts 6′·
11-d5 and 12-d5 of Silylium Ions 6′ and 7. In a glovebox, a
solution of the requisite silane (1.00 equiv) in 1,2-Cl2C6H4 (0.40 mL)
was added to a suspension of [Ph3C]+[B(C6F5)4]− (1.00 equiv) in 1,2-
Cl2C6H4 (0.30 mL) in an 8 mL vial equipped with a magnetic stir bar.
The resulting red-brown solution was stirred for 1 min and transferred
to a vial containing pyridine-d5 (11-d5, 0.639−0.661 equiv). The
sample was transferred to an NMR tube, the vials were washed with
1,2-Cl2C6H4 (0.30 mL), and the washings were transferred to the
NMR tube. The sample was directly subjected to NMR spectroscopic
analysis.
1
equiv) according to General Procedure 1. H NMR (300 MHz, 1,2-
Cl2C6D4): δ 0.38 (s, 3H), 0.80 (dt, JH,P = 19.5 Hz, J = 7.7 Hz, 9H),
0.99 (s, 9H), 1.57 (dq, JH,P = 11.5 Hz, J = 7.7 Hz, 6H), 3.90 (mc, 1H),
4.01 (s, 5H), 4.04 (mc, 1H), 4.41 ppm (mc, 2H). 11B NMR (96 MHz,
1,2-Cl2C6D4): δ −16.5 ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ
−166.3, −162.2, −132.0 ppm. 29Si DEPT NMR (60 MHz, 1,2-
Cl2C6D4): δ 27.2 ppm (d, JSi,P = 17.5 Hz). 31P{1H} NMR (162 MHz,
1,2-Cl2C6D4): δ 88.7 ppm.
Di-tert-butylferrocenylsilylium Tetrakis(pentafluorophenyl)-
borate Triethylphosphine Oxide Adduct (9d). This was prepared
from di-tert-butylferrocenylsilane (5.73 mg, 17.5 μmol, 1.00 equiv),
[Ph3C]+[B(C6F5)4]− (16.1 mg, 17.5 μmol, 1.00 equiv), and
triethylphosphine oxide (8, 2.30 mg, 17.1 μmol, 0.980 equiv)
according to General Procedure 1. 1H NMR (300 MHz, 1,2-
Cl2C6D4): δ 0.81 (dt, JH,P = 19.5 Hz, J = 7.7 Hz, 9H), 1.06 (s,
18H), 1.60 (dq, JH,P = 12.1 Hz, J = 7.7 Hz, 6H), 4.05 (mc, 2H), 4.06 (s,
5H), 4.44 ppm (mc, 2H). 11B NMR (96 MHz, 1,2-Cl2C6D4): δ −15.9
ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ −165.8, −161.9, −131.5
ppm. 29Si DEPT NMR (60 MHz, 1,2-Cl2C6D4): δ 25.0 ppm (d, JSi,P
21.1 Hz). 31P{1H} NMR (162 MHz, 1,2-Cl2C6D4): δ 87.9 ppm.
=
Ferrocenyldiisopropylsilylium Tetrakis(pentafluorophenyl)-
borate Triethylphosphine Oxide Adduct (9e). This was prepared
from ferrocenyldiisopropylsilane (5.23 mg, 17.5 μmol, 1.00 equiv),
[Ph3C]+[B(C6F5)4]− (16.1 mg, 17.5 μmol, 1.00 equiv), and
triethylphosphine oxide (8, 2.30 mg, 17.1 μmol, 0.980 equiv)
according to General Procedure 1. 1H NMR (300 MHz, 1,2-
Cl2C6D4): δ 0.82 (dt, JH,P = 19.4 Hz, J = 7.7 Hz, 9H), 1.06−1.20
(m, 14H), 1.53 (dq, JH,P = 11.6 Hz, J = 7.6 Hz, 6H), 3.95 (mc, 2H),
4.02 (s, 5H), 4.41 ppm (mc, 2H). 11B NMR (96 MHz, 1,2-Cl2C6D4): δ
−15.9 ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ −165.8, −161.9,
−131.5 ppm. 29Si DEPT NMR (60 MHz, 1,2-Cl2C6D4): δ 24.3 ppm
(d, JSi,P = 17.6 Hz). 31P{1H} NMR (162 MHz, 1,2-Cl2C6D4): δ 88.4
ppm.
General Procedure 4: Preparation of Pyridine-d5 Adducts 4·
11-d5 and 5·11-d5 of Trimethylsilyl Precursors 4 and 5. A
septum-sealed screw-cap NMR tube was charged with pyridine-d5 (11-
d5, 0.602−0.624 equiv), and a solution of 4 or 5 (1.00 equiv) in 1,2-
Cl2C6H4 (1.00 mL) was added. The sample was directly subjected to
NMR spectroscopic analysis.
tert-Butylferrocenylphenylsilylium Tetrakis(penta-
fluorophenyl)borate Triethylphosphine Oxide Adduct (9f).
This was prepared from tert-butylferrocenyphenylsilane (6.08 mg,
17.5 μmol, 1.00 equiv), [Ph3C]+[B(C6F5)4]− (16.1 mg, 17.5 μmol,
1.00 equiv), and triethylphosphine oxide (8, 2.30 mg, 17.1 μmol, 0.980
General Procedure 5: ReactIR Analysis of the Model Diels−
Alder Reaction. Stock solutions in 1,2-Cl2C6H4 of all reactants were
used: silanes (0.1 M), [Ph3C]+[B(C6F5)4]− (62.5 mM), and E-13
(1.67 M). The stock solutions of [Ph3C]+[B(C6F5)4]− and E-13 were
prepared for a maximum of 10 measurements and used for at least two
different catalysts to eliminate errors. A 10 mL two-neck flask with a
magnetic stir bar was equipped with the ReactIR probe head and
connected to a Schlenk line through a rubber septum. The joint for the
probe head was air tightened using a PTFE sleeve and the ReactIR
PTFE adapter. The flask was heated under inert atmosphere,
evacuated, and flushed with inert gas after cooling to room
temperature at least five times. Appropriate amounts of the stock
solutions were transferred into vials sealed with rubber septa inside the
glovebox before being connected to a Schlenk line outside the
glovebox. The flask was cooled to 12.7 °C (para-xylene/CO2 cooling
bath), an aliquot of the stock solution of [Ph3C]+[B(C6F5)4]− (0.40
mL, 0.025 mmol, 5.0 mol %) was added, and the measurement was
initiated. An aliquot of the silane stock solution (0.30 mL, 0.030 mmol,
6.0 mol %) was added. After 1 min, an aliquot of the stock solution of
E-13 (0.30 mL, 0.50 mmol, 1.0 equiv) was added, and the solution was
stirred for approximately 4 min. Precooled 1 (0.10 mL, 1.0 mmol, 2.0
equiv) was added. The reaction progress was monitored until no
further increase of the carbonyl band absorption of trans-14 at 1687
cm−1 was observed (8 h was required for 7h as full conversion was
reached after more than 16 h at room temperature). To confirm full
conversion, a sample was hydrolyzed with saturated aqueous NaHCO3
solution (2 mL) and extracted with tert-butyl methyl ether (2 mL). An
aliquot of the organic phase (100 μL) was eluted over a small pad of
silica gel with tert-butyl methyl ether and subjected to GLC analysis;
1
equiv) according to General Procedure 1. H NMR (300 MHz, 1,2-
Cl2C6D4): δ 0.78 (dt, JH,P = 19.4 Hz, J = 7.7 Hz, 9H), 1.03 (s, 9H),
1.51 (dq, JH,P = 11.6 Hz, J = 7.7 Hz, 6H), 3.89 (ddd, J = 2.5 Hz, J = 1.2
Hz, J = 1.2 Hz, 1H), 4.17 (s, 5H), 4.22 (ddd, J = 2.5 Hz, J = 1.2 Hz, J =
1.2 Hz, 1H), 4.45 (ddd, J = 2.4 Hz, J = 2.4 Hz, J = 1.0 Hz, 1H), 4.50
(ddd, J = 2.4 Hz, J = 2.4 Hz, J = 1.1 Hz, 1H), 7.44−7.50 (m, 3H),
7.75−7.79 ppm (m, 2H). 11B NMR (96 MHz, 1,2-Cl2C6D4): δ −15.9
ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ −165.8, −161.9, −131.5
ppm. 29Si DEPT NMR (60 MHz, 1,2-Cl2C6D4): δ 14.3 ppm (d, JSi,P
16.9 Hz). 31P{1H} NMR (121 MHz, 1,2-Cl2C6D4): δ 90.5 ppm.
=
Triferrocenylsilylium Tetrakis(pentafluorophenyl)borate
Triethylphosphine Oxide Adduct (9h). This was prepared from
triferrocenylsilane (10.2 mg, 17.5 μmol, 1.00 equiv), [Ph3C]+[B-
(C6F5)4]− (16.1 mg, 17.5 μmol, 1.00 equiv), and triethylphosphine
oxide (8, 2.30 mg, 17.1 μmol, 0.980 equiv) according to General
1
Procedure 1. H NMR (300 MHz, 1,2-Cl2C6D4): δ 0.84 (dt, JH,P
=
19.4 Hz, J = 7.7 Hz, 9H), 1.65 (dq, JH,P = 11.6 Hz, J = 7.7 Hz, 6H),
4.08 (s, 15H), 4.44 (mc, 6H), 4.56 ppm (mc, 6H). 11B NMR (96 MHz,
1,2-Cl2C6D4): δ −16.0 ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ
1
−165.8, −162.0, −131.6 ppm. H,29Si HMQC NMR (99 MHz, 1,2-
Cl2C6D4): δ 13.5 ppm. 31P{1H} NMR (202 MHz, 1,2-Cl2C6D4): δ
88.0 ppm.
Diferrocenylmethylsilylium Tetrakis(pentafluorophenyl)-
borate Triethylphosphine Oxide Adduct (9i). This was prepared
from diferrocenylmethylsilane (7.23 mg, 17.5 μmol, 1.00 equiv),
E
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
2
[Ph3C]+[B(C6F5)4]− (16.1 mg, 17.5 μmol, 1.00 equiv), and
triethylphosphine oxide (8, 2.30 mg, 17.1 μmol, 0.980 equiv)
according to General Procedure 1. 1H NMR (300 MHz, 1,2-
Cl2C6D4): δ 0.75 (s, 3H), 0.89 (dt, JH,P = 19.3 Hz, J = 7.7 Hz, 9H),
1.64 (dq, JH,P = 11.6 Hz, J = 7.7 Hz, 6H), 4.05 (s, 10H), 4.06 (mc, 4H),
4.42 (ddd, J = 2.4 Hz, J = 2.4 Hz, J = 1.1 Hz, 2H), 4.45 ppm (ddd, J =
2.4 Hz, J = 2.4 Hz, J = 1.2 Hz, 2H). 11B NMR (96 MHz, 1,2-Cl2C6D4):
δ −16.0 ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ −165.9, −162.0,
−131.6 ppm. 29Si DEPT NMR (60 MHz, 1,2-Cl2C6D4): δ 17.8 ppm
(d, JSi,P = 12.5 Hz). 31P{1H} NMR (162 MHz, 1,2-Cl2C6D4): δ 89.0
ppm.
General Procedure 3. H NMR (77 MHz, 1,2-Cl2C6H4): δ 7.43, 8.00,
8.18 ppm. 1H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6H4): δ 23.4 ppm.
Triferrocenylsilylium Tetrakis(pentafluorophenyl)borate
Pyridine-d5 Adduct (12h-d5). This was prepared from triferroce-
nylsilane (46.7 mg, 80.0 μmol, 1.01 equiv), [Ph3C]+[B(C6F5)4]− (73.0
mg, 79.1 μmol, 1.00 equiv), and pyridine-d5 (11-d5, 4.40 mg, 52.3
2
μmol, 0.661 equiv) according to General Procedure 3. H NMR (77
MHz, 1,2-Cl2C6H4): δ 7.65, 8.01, 9.17 ppm. 1H,29Si HMQC NMR (99
MHz, 1,2-Cl2C6H4): δ 22.8 ppm.
Diferrocenylmethylsilylium Tetrakis(pentafluorophenyl)-
borate Pyridine-d5 Adduct (12i-d5). This was prepared from
diferrocenylmethylsilane (33.1 mg, 80.0 μmol, 1.01 equiv), [Ph3C]+[B-
(C6F5)4]− (73.4 mg, 79.5 μmol, 1.00 equiv), and pyridine-d5 (11-d5,
4.40 mg, 52.3 μmol, 0.658 equiv) according to General Procedure 3
Triethylsilylium Tetrakis(pentafluorophenyl)borate Triethyl-
phosphine Oxide Adduct (6′·8). This was prepared from
triethylsilane (11.6 mg, 100 μmol, 1.00 equiv), [Ph3C]+[B(C6F5)4]−
(92.2 mg, 100 μmol, 1.00 equiv), and triethylphosphine oxide (8, 12.7
mg, 95.0 μmol, 0.950 equiv) according to General Procedure 1 except
for stirring the silylium ion solution for 24 h before the addition of 8.
The formation of 6′ was not entirely complete; thus the adduct of
2
along with adduct 12h-d5. H NMR (77 MHz, 1,2-Cl2C6H4): δ 7.52,
7.98, 8.58 ppm. 1H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6H4): δ 26.8
ppm.
Triethylsilylium Tetrakis(pentafluorophenyl)borate Pyri-
dine-d5 Adduct (6′·11-d5). This was prepared from triethylsilane
(9.30 mg, 80.0 μmol, 1.00 equiv), [Ph3C]+[B(C6F5)4]− (73.8 mg, 80.0
μmol, 1.00 equiv), and pyridine-d5 (11-d5, 4.40 mg, 52.3 μmol, 0.654
1
[Ph3C]+[B(C6F5)4]− and 8 formed additionally. H NMR (300 MHz,
1,2-Cl2C6D4): δ 0.51−0.56 (m, 6H), 0.75−0.81 (m, 9H), 0.91−0.99
(m, 9H), 1.69−1.81 ppm (m, 6H). 11B NMR (96 MHz, 1,2-Cl2C6D4):
δ −15.8 ppm. 19F NMR (282 MHz, 1,2-Cl2C6D4): δ −165.8, −161.9,
2
equiv) according to General Procedure 3. H NMR (77 MHz, 1,2-
Cl2C6H4): δ 7.59, 8.01, 8.15 ppm. H,29Si HMQC NMR (99 MHz,
1
1
−131.4 ppm. H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6D4): δ 35.8
ppm. 31P{1H} NMR (202 MHz, 1,2-Cl2C6D4): δ 88.8 ppm.
Trimethylsilylium Bis(trifluoromethanesulfonyl)imide Trie-
thylphosphine Oxide Adduct (5·8). This was prepared from N-
trimethylsilyl bis(trifluoromethanesulfonyl)imide (5, 49.4 mg, 32.0 μL,
139.7 μmol, 1.00 equiv) and triethylphosphine oxide (8, 16.8 mg,
125.1 μmol, 0.895 equiv) according to General Procedure 2. 1H NMR
(400 MHz, 1,2-Cl2C6D4): δ 0.23 (s, 9H), 1.03 (dt, JH,P = 19.3 Hz, J =
7.7 Hz, 9H), 2.03 ppm (dq, JH,P = 11.7 Hz, J = 7.7 Hz, 6H). 19F NMR
(376 MHz, 1,2-Cl2C6D4): δ −78.6 ppm. 29Si DEPT NMR (79 MHz,
1,2-Cl2C6D4): δ 32.5 ppm (d, JSi,P = 14.5 Hz). 31P{1H} NMR (162
MHz, 1,2-Cl2C6D4): δ 91.0 ppm.
1,2-Cl2C6H4): δ 42.6 ppm.
Trimethylsilylium Bis(trifluoromethanesulfonyl)imide Pyri-
dine-d5 Adduct (5·11-d5). This was prepared from N-trimethylsilyl
bis(trifluoromethanesulfonyl)imide (5, 28.3 mg, 80.0 μmol, 1.00
equiv) and pyridine-d5 (11-d5, 4.20 mg, 49.9 μmol, 0.624 equiv)
according to General Procedure 4. 2H NMR (61 MHz, 1,2-Cl2C6H4):
δ 7.72, 8.04, 8.51 ppm. 29Si DEPT NMR (79 MHz, 1,2-Cl2C6H4): δ
41.9 ppm.
Trimethylsilylium Trifluoromethanesulfonate Pyridine-d5
Adduct (4·11-d5). This was prepared from trimethylsilyl trifluor-
omethanesulfonate (4, 18.4 mg, 82.9 μmol, 1.00 equiv) and pyridine-
d5 (11-d5, 4.20 mg, 49.9 μmol, 0.602 equiv) according to General
Trimethylsilylium Trifluoromethanesulfonate Triethylphos-
phine Oxide Adduct (4·8). This was prepared from trimethylsilyl
trifluoromethanesulfonate (4, 29.5 mg, 24.0 μL, 132.7 μmol, 1.00
equiv) and triethylphosphine oxide (8, 14.4 mg, 107.3 μmol, 0.806
2
Procedure 4. H NMR (61 MHz, 1,2-Cl2C6H4): δ 7.29, 7.68, 8.46
ppm. 29Si DEPT NMR (79 MHz, 1,2-Cl2C6H4): δ 42.7 ppm.
1
equiv) according to General Procedure 2. H NMR (400 MHz, 1,2-
ASSOCIATED CONTENT
■
Cl2C6D4): δ 0.25 (s, 9H), 1.05 (dt, JH,P = 19.2 Hz, J = 7.6 Hz, 9H),
2.23 ppm (dq, JH,P = 12.0 Hz, J = 7.6 Hz, 6H). 19F NMR (376 MHz,
1,2-Cl2C6D4): δ −77.4 ppm. 29Si DEPT NMR (79 MHz, 1,2-
Cl2C6D4): δ 32.3 ppm (d, JSi,P = 14.4 Hz). 31P{1H} NMR (162 MHz,
1,2-Cl2C6D4): δ 92.5 ppm.
S
* Supporting Information
Experimental details, characterization data, as well as H, 11B,
1
19F, 29Si, and 31P NMR spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
tert-Butylferrocenylmethylsilylium Tetrakis(penta-
fluorophenyl)borate Pyridine-d5 Adduct (12a-d5). This was
prepared from tert-butylferrocenylmethylsilane (22.9 mg, 80.0 μmol,
1.00 equiv), [Ph3C]+[B(C6F5)4]− (73.8 mg, 80.0 μmol, 1.00 equiv),
and pyridine-d5 (11-d5, 4.40 mg, 52.3 μmol, 0.654 equiv) according to
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: hilt@chemie.uni-marburg.de.
*E-mail: martin.oestreich@tu-berlin.de.
2
General Procedure 3. H NMR (77 MHz, 1,2-Cl2C6H4): δ 7.43, 8.00,
8.14 ppm. 1H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6H4): δ 36.8 ppm.
Di-tert-butylferrocenylsilylium Tetrakis(pentafluorophenyl)-
borate Pyridine-d5 Adduct (12d-d5). This was prepared from di-
tert-butylferrocenylsilane (26.3 mg, 80.0 μmol, 1.00 equiv),
[Ph3C]+[B(C6F5)4]− (73.8 mg, 80.0 μmol, 1.00 equiv), and
pyridine-d5 (11-d5, 4.40 mg, 52.3 μmol, 0.654 equiv) according to
Author Contributions
A.R.N. and K.M. contributed equally.
Notes
2
The authors declare no competing financial interest.
General Procedure 3. H NMR (77 MHz, 1,2-Cl2C6H4): δ 7.70, 8.07,
9.12 ppm. 1H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6H4): δ 35.0 ppm.
Ferrocenyldiisopropylsilylium Tetrakis(pentafluorophenyl)-
borate Pyridine-d5 Adduct (12e-d5). This was prepared from
ferrocenyldiisopropylsilane (24.0 mg, 80.0 μmol, 1.00 equiv),
[Ph3C]+[B(C6F5)4]− (73.8 mg, 80.0 μmol, 1.00 equiv), and
pyridine-d5 (11-d5, 4.40 mg, 52.3 μmol, 0.654 equiv) according to
ACKNOWLEDGMENTS
■
This research was supported by the Deutsche Forschungsge-
meinschaft (Oe 249/9-1) and the Fonds der Chemischen
Industrie (predoctoral fellowship to V.H.G.R., 2012−2014).
M.O. is indebted to the Einstein Foundation (Berlin) for an
endowed professorship. We thank the Cluster of Excellence
“Unifying Concepts in Catalysis” of the Deutsche Forschungs-
gemeinschaft (EXC 314) for providing a Mettler Toledo
ReactIR 15. We are grateful to Dr. Sebastian Kemper (TU
Berlin) for expert advice with the NMR measurements.
2
General Procedure 3. H NMR (77 MHz, 1,2-Cl2C6H4): δ 7.49, 7.97,
8.19 ppm. 1H,29Si HMQC NMR (99 MHz, 1,2-Cl2C6H4): δ 34.5 ppm.
tert-Butylferrocenylphenylsilylium Tetrakis(penta-
fluorophenyl)borate Pyridine-d5 Adduct (12f-d5). This was
prepared from tert-butylferrocenylphenylsilane (27.8 mg, 80.0 μmol,
1.00 equiv), [Ph3C]+[B(C6F5)4]− (73.8 mg, 80.0 μmol, 1.00 equiv),
and pyridine-d5 (11-d5, 4.30 mg, 51.1 μmol, 0.639 equiv) according to
F
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(21) (a) Wang, C.; Erker, G.; Kehr, G.; Wedeking, K.; Frohlich, R.
Organometallics 2005, 24, 4760−4773. (b) Lambert, J. B.; Lin, L.;
Keinan, S. Org. Biomol. Chem. 2003, 1, 2559−2565.
̈
REFERENCES
■
(1) (a) Hosomi, A.; Miura, K. In Acid Catalysis in Modern Organic
Synthesis; Yamamoto, H., Ishihara, K., Eds.; Wiley-VCH: Weinheim,
Germany, 2008; pp 469−516. (b) Dilmann, A. D.; Ioffe, S. L. Chem.
Rev. 2003, 103, 733−772. (c) Oishi, M. In Lewis Acids in Organic
Synthesis; Yamamoto, H., Ed.; Wiley-VCH: Weinheim, Germany,
2000; pp 355−393.
(2) Klare, H. F. T.; Oestreich, M. Dalton Trans. 2010, 39, 9176−
9184.
(3) (a) Mathieu, B.; Ghosez, L. Tetrahedron Lett. 1997, 38, 5497−
5500. (b) Mathieu, B.; Ghosez, L. Tetrahedron 2002, 58, 8219−8226.
(4) Hara, K.; Akiyama, R.; Sawamura, M. Org. Lett. 2005, 7, 5621−
5623.
(5) (a) Schulz, A.; Villinger, A. Angew. Chem., Int. Ed. 2012, 51,
4526−4528. (b) Muller, T. Adv. Organomet. Chem. 2005, 53, 155−215.
̈
(c) Lambert, J. B.; Zhao, Y.; Zhang, S. M. J. Phys. Org. Chem. 2001, 14,
370−379. (d) Reed, C. A. Acc. Chem. Res. 1998, 31, 325−332.
(6) Olah, G. A.; Field, L. D. Organometallics 1982, 1, 1485−1487.
(7) Muther, K.; Frohlich, R.; Muck-Lichtenfeld, C.; Grimme, S.;
̈
̈
̈
Oestreich, M. J. Am. Chem. Soc. 2011, 133, 12442−12444.
(8) (a) Klare, H. F. T.; Bergander, K.; Oestreich, M. Angew. Chem.,
Int. Ed. 2009, 48, 9077−9079. (b) Schmidt, R. K.; Muther, K.; Muck-
̈
̈
Lichtenfeld, C.; Grimme, S.; Oestreich, M. J. Am. Chem. Soc. 2012,
134, 4421−4428.
(9) Muther, K.; Hrobar
M. Chem.−Eur. J. 2013, 19, 16579−16594.
́
ik, P.; Hrobar
́ ́
ikova, V.; Kaupp, M.; Oestreich,
̈
(10) The obvious choice of ferrocenyl-substituted silicon Lewis acids
7 with TfO− and Tf2N− as counteranions was thwarted by our inability
to access these in pure form. We are aware that Manners and co-
workers had prepared such triflates as well as their pyridine adducts.11
While the data of the latter were a useful reference point for our
purposes, we decided not to use impure Lewis acids in the kinetic
analysis.
(11) Bourke, S. C.; MacLachlan, M. J.; Lough, A. J.; Manners, I.
Chem.−Eur. J. 2005, 11, 1989−2000.
(12) We had previously calculated fluoride ion affinities as a measure
of the Lewis acidity of 7, but these correlated poorly with the 29Si
NMR chemical shifts.9 Likewise, there was no correlation of those
numbers with the observed catalytic activities of 7 later obtained in the
ReactIR kinetic analysis.
(13) (a) Mayer, U.; Gutmann, V.; Gerger, W. Monatsh. Chem. 1975,
106, 1235−1257. (b) Gutmann, V. Coord. Chem. Rev. 1976, 18, 225−
255. (c) Beckett, M. A.; Strickland, G. C.; Holland, J. R.; Varma, K. S.
Polym. Commun. 1996, 37, 4629−4631. (d) Beckett, M. A.;
Brassington, D. S.; Light, M. E.; Hursthouse, M. B. J. Chem. Soc.,
Dalton Trans. 2001, 1768−1772.
(14) Ferrocene-stabilized silicon cations 7 are able to exchange
substituents with unreacted silane precursors.9 The ligand scrambling
is dependent on the steric situation around the silicon atom and the
silane concentration. We selected those 7 that would not show this
undesired side-reaction.
(15) Hilt, G.; Punner, F.; Mobus, J.; Naseri, V.; Bohn, M. A. Eur. J.
̈
̈
Org. Chem. 2011, 5962−5966.
(16) Hilt, G.; Nodling, A. Eur. J. Org. Chem. 2011, 7071−7075.
̈
(17) Interestingly, the Diels−Alder reaction outlined in Scheme 2
cannot be performed at temperatures below −25 °C when using
[Et3Si(toluene)]+[B(C6F5)4]− (6) as the catalyst because the solution
began to solidify (cf. ref 4).
(18) We cannot exclude temporary catalyst inhibition by the Diels−
Alder adduct formed during the course of the reaction. Also, the Lewis
acid might undergo undesired side-reactions with the diene
component. We, therefore, report TOF values in addition to the
initial rates as a qualitative measure to compare and distinguish
between the reactivity of the silicon cations at higher conversion
(Table 3).
(19) Schmidt, R. K. Ph.D. Thesis, Westfalische Wilhelms-Universitat
̈
̈
Munster, Germany, 2012.
̈
(20) Prakash, G. K. S.; Bae, C.; Rasul, G.; Olah, G. A. J. Org. Chem.
2002, 67, 1297−1301.
G
dx.doi.org/10.1021/om401040y | Organometallics XXXX, XXX, XXX−XXX