The use of Karstedt's catalyst to study the reduction of Me2NCHO (DMF) by the popular "dual SiH"-containing tetramethyldisiloxane, HMe2SiOSiMe2H (1), has revealed that the first step in the process involves an initial single hydrosilylation to form HSiMe2OSiMe2OCH2NMe2 (3). This intermediate is readily isolated and purified via distillation. In the continued presence of the catalyst, 3 transforms to the transient tetrasiloxane HMe2SiOSiMe2OSiMe2OSiMe2OCH2NMe2 (4), along with the formation of Me3N. The tetrasiloxane 4 itself transforms to Me3N and (Me2SiO) n (n = 4-6). Despite the demonstrated reactivity of 3, it can also be used to perform the expected metal-catalyzed hydrosilylation chemistry of the SiH group as well as reactions of the SiOCH2NMe2 functionality involving siloxane chain extension and is thus an important new reagent for siloxane chemistry.
The use of Karstedt's catalyst to study the reduction of Me2NCHO (DMF) by the popular "dual SiH"-containing tetramethyldisiloxane, HMe2SiOSiMe2H (1), has revealed that the first step in the process involves an initial single hydrosilylation to form HSiMe2OSiMe2OCH2NMe2 (3). This intermediate is readily isolated and purified via distillation. In the continued presence of the catalyst, 3 transforms to the transient tetrasiloxane HMe2SiOSiMe2OSiMe2OSiMe2OCH2NMe2 (4), along with the formation of Me3N. The tetrasiloxane 4 itself transforms to Me3N and (Me2SiO) n (n = 4-6). Despite the demonstrated reactivity of 3, it can also be used to perform the expected metal-catalyzed hydrosilylation chemistry of the SiH group as well as reactions of the SiOCH2NMe2 functionality involving siloxane chain extension and is thus an important new reagent for siloxane chemistry.
Since the initial,
and generally uncited, report by the Voronkov group that silanes,
R3SiH, could effectively reduce DMF to Me3N
with concomitant formation of disiloxanes,[1] their use as reducing agents to transform amides to amines has been
well-studied and widely used due to the generally mild reaction conditions
employed.[2] Using monosilanes, R3SiH, to reduce DMF as a model amide, we recently demonstrated that
such reductions proceed via an initial hydrosilylation reaction to
produce siloxymethylaminesR3SiOCH2NMe2 (O-silylated hemi-aminals).[3] This class of
compounds can further react with silanes in the presence of catalysts,
and also in excess DMF, to form the amine and appropriate disiloxane
(eq 1).A particularly versatile silane for
such reductions is 1,1,3,3-tetramethyldisiloxane, HSiMe2OSiMe2H (1), and using a range of catalysts,
including Karstedt’s catalyst (bis[1,3-bis(η2-ethenyl)-1,1,3,3-tetramethyldisiloxane]platinum), this reagent possesses
some unique reduction characteristics.[4−6] For example, it has been
demonstrated that 1 can effectively reduce amides under
conditions where other silanes are ineffective and this special property
has been associated with a “dual SiH effect”.[5a] Furthermore, the same disiloxane has been noted
as a poor reagent for other reactions.[5c] We have previously reported that using 1 in the presence
of (Me3N)Mo(CO)5 as catalyst, for the reduction
of DMF, a double hydrosilylation occurs to form Me2NCH2OSiMe2OSiMe2OCH2NMe2, (2), which can be isolated and characterized.[7] As a continuation of our studies in this arena
we have now used Karstedt’s catalyst to study the reaction
of 1 with DMF, monitoring the reaction with 29Si, 13C, and 1H NMR spectroscopy. Typical monitoring
sequences of such a reaction in C6D6 at room
temperature (∼295 K) are presented in Figures 1 (29Si) and 2 (13C). Along with the disappearance of the 29Si resonance
at −4.6 ppm due to 1, there is a growth of two
resonances at −6.4 and −12.2 ppm, and under the reaction
conditions used, no further significant chemistry takes place.
Figure 1
29Si NMR monitoring of the reaction between HSiMe2OSiMe2H (−4.6 ppm) and DMF (1:5 molar ratio) catalyzed by
1 mol % of Karstedt’s catalyst, illustrating the formation
of 3 (−6.4 and −12.2 ppm).
Figure 2
13C NMR monitoring of the reaction between HSiMe2OSiMe2H (0.1 ppm) and DMF (1:5 molar ratio) catalyzed
by 1 mol % of Karstedt’s catalyst, illustrating the formation
of 3 (81.7 (CH2), 41.0 (NMe2),
0.84 and −0.89 ppm (SiMe)). Me resonances for DMF appear at
30.6 and 35.2 ppm.
29Si NMR monitoring of the reaction between HSiMe2OSiMe2H (−4.6 ppm) and DMF (1:5 molar ratio) catalyzed by
1 mol % of Karstedt’s catalyst, illustrating the formation
of 3 (−6.4 and −12.2 ppm).The related 13C spectral sequence (Figure 2) exhibits the appearance
of new resonances at 81.7, 41.0, 0.84, and −0.89 ppm typical
of the SiOCH2NMe2 group and two new Me2Si units. The new material is the single hydrosilylation product,
Me2NCH2OSiMe2OSiMe2H (3; eq 2).13C NMR monitoring of the reaction between HSiMe2OSiMe2H (0.1 ppm) and DMF (1:5 molar ratio) catalyzed
by 1 mol % of Karstedt’s catalyst, illustrating the formation
of 3 (81.7 (CH2), 41.0 (NMe2),
0.84 and −0.89 ppm (SiMe)). Me resonances for DMF appear at
30.6 and 35.2 ppm.The same chemistry takes
place on a larger preparative scale using either hexane or benzene
as solvent, and we have been able to isolate this material by distillation
at 49 °C/15 mmHg in excellent yield. Compound 3 is
relatively stable at room temperature; however, when it is subjected
to the presence of Karstedt’s catalyst the 29Si
resonances at −6.4 and −12.2 ppm transform to four new
signals at −6.7, −13.9, −19.8, and −21.5
ppm and via 13C NMR we observe the concurrent formation
of Me3N. The new silicon-containing material is the tetrasiloxane
HSiMe2OSiMe2OSiMe2OSiMe2OCH2NMe2 (4). While we have been
unable to obtain 4 as an analytically pure material,
we have been able to prepare and isolate it in ∼95% purity
and study its further chemistry, as described in the Supporting Information. Continued exposure of 4 to the catalytic conditions results in formation of more Me3N and a mixture of cyclic dimethylpolysiloxanes, (Me2SiO) (n = 4, D4; n = 5, D5; n = 6, D6), as noted by comparison of their 29Si NMR and GC/MS spectra with those of known materials (Figures S-5–S-8,
respectively (Supporting Information)).
The 29Si and 13C monitoring of this transformation
is presented in Figures 3 and 4, respectively.
Figure 3
Transformation of 4 (−6.7,
−13.9, −19.8, and −21.5 ppm) to predominantly D4 (−19.1 ppm) and D6 (−21.6 ppm)
monitored by 29Si NMR.
Figure 4
Transformation of 4 (81.8 ppm (CH2); 41.1 ppm (Me2N) to Me3N (47.5 ppm) and D4–6 monitored by 13C NMR.
The minor resonances at ∼20 ppm are associated with the xylene
solvent of the catalyst solution.
Transformation of 4 (−6.7,
−13.9, −19.8, and −21.5 ppm) to predominantly D4 (−19.1 ppm) and D6 (−21.6 ppm)
monitored by 29Si NMR.Transformation of 4 (81.8 ppm (CH2); 41.1 ppm (Me2N) to Me3N (47.5 ppm) and D4–6 monitored by 13C NMR.
The minor resonances at ∼20 ppm are associated with the xylene
solvent of the catalyst solution.To prove the structure and formulation, we have reacted 4 with Me3SiCl. This procedure results in the formation
and high-yield isolation of the expected pentasiloxane HSiMe2OSiMe2OSiMe2OSiMe2OSiMe3 (5), a known compound.[8]Overall during the reduction of DMF by 1, the initially
formed hydrosilylation product 3 reacts with itself to
liberate Me3N and 4, which further eliminates
Me3N along with formation of D, both reactions demonstrating the reactivity of the siloxymethylamines
with SiH species involving formation of Me3N and siloxanes
(Scheme 1).[3] This
is a further illustration of the capacity of hydrosilanes to reduce
siloxymethylamines (O-silylated hemi-aminals), as illustrated in eq 1, and the more general capacity to reduce aminals.[9]
Scheme 1
Sequence of Reactions Leading to Production
of Me3N and D
The transformation of 4 to Me3N poses the question as to whether this
is an intramolecular elimination reaction or involves a bimolecular
reaction with a second SiH functionality. Since the silicon-containing
compound formed is predominantly D4 and we see no significant
amounts of long-chain polysiloxanes, we favor the intramolecular amine
elimination. We attempted to answer that question by D-labeling experiments
using deuteriotetramethyldisiloxane, DSiMe2OSiMe2D[10] (1D). Use of 1D to reduce DMF resulted, as expected, in the formation of Me2NCHD2. However, using an equimolar mixture of 1 and 1D resulted predominantly in the formation
of Me2NCH2D and Me3N and smaller
amounts of Me2NCHD2: i.e., a significant H/D
scrambling had occurred (Figure S-4 (Supporting
Information)). Separate experiments mixing 1D and
Et3SiH resulted in a very rapidly established equilibrium
illustrating the metal-catalyzed H/D exchange in hydrosilanes,[11] thus ruling out the mechanistic clarification
we sought.As we previously reported, the use of (Me3N)Mo(CO)5 as catalyst for the reduction of DMF by 1 proceeds via a different route, involving the intermediacy
of the double-hydrosilylation intermediate (Me2NCH2OSiMe2)2O.[7] We have treated the new intermediate 3 with DMF in
the presence of the molybdenum catalyst and observed the rapid formation
of (Me2NCH2OSiMe2)2O (2) with no initial Me3N formation, illustrated
in Figure 5. The two catalysts clearly have
distinctive properties in this chemistry, and studies to tease out
these distinctions, along with those of other catalysts, are in progress.
Figure 5
29Si NMR monitoring of the reaction between DMF and HSiMe2OSiMe2OCH2NMe2 (3)
catalyzed by 1 mol % of (Me3N)Mo(CO)5 leading
to Me2NCH2OSiMe2OSiMe2OCH2NMe2, (2) at −13.9
ppm.
29Si NMR monitoring of the reaction between DMF and HSiMe2OSiMe2OCH2NMe2 (3)
catalyzed by 1 mol % of (Me3N)Mo(CO)5 leading
to Me2NCH2OSiMe2OSiMe2OCH2NMe2, (2) at −13.9
ppm.Treatment of 3 under
varying conditions of the functional group reactivity was performed
to illustrate that the two terminal silyl groups (SiH and SiOCH2NMe2) retain their established chemistry in the
presence of each other, even in the presence of metal catalysts. Thus,
the reaction of 3 with Me3ECl (E = Si, Ge)
led to the high-yield formation of the siloxane chain extension products,
resulting in either trisiloxane 6a or disiloxygermoxane 6b (eq 5).The trisiloxane 6a is a useful, commercially
available reagent,[12] whereas the germanium
analogue is unreported. All analytical and spectroscopic data are
in accord with either the published data or those expected. For example,
the 29Si NMR data for 6a exhibits three resonances
at 7.4, −6.9, and −19.4 ppm for the Me3SiO,
HMe2SiO, and Me2SiO silicon atoms, respectively.
The Ge analogue 6b exhibits 29Si resonances
at −8.2 and −17.6 ppm, in accord with expectation.The chemistry of the SiH functionality of 3 was initially
expected to be complicated, since the hydrosilylation reaction normally
needs a catalytic species similar to that required for the amine elimination/amide
reduction product (eq 2). Hence, a competition
between hydrosilylation and amine elimination was anticipated. However,
in a test hydrosilylation reaction we treated 3 with
Me3SiCH=CH2 in the presence of Karstedt’s
catalyst and obtained a high yield of the “expected”
hydrosilylation product 7 (eq 6). No amine elimination chemistry was observed and the hydrosilylation
appears to be regiospecific.Compound 7 can be readily reacted with chlorosilanes
for further siloxane chain extension (eq 7).Thus, the newly observed and isolated material 3 has three distinct and very useful modes of chemistry: elimination
of the reduced amide as the amine, hydrosilylation using the SiH group,
and siloxane chain extension using the SiOCH2NMe2 functionality (Scheme 2).