Mechanistic Investigation of Bis(imino)pyridine Manganese Catalyzed Carbonyl and Carboxylate Hydrosilylation

Article abstract of DOI:10.1021/jacs.7b00879

We recently reported a bis(imino)pyridine (or pyridine diimine, PDI) manganese precatalyst, (^Ph2PPrPDI)Mn (1), that is active for the hydrosilylation of ketones and dihydrosilylation of esters. In this contribution, we reveal an expanded scope for 1-mediated hydrosilylation and propose two different mechanisms through which catalysis is achieved. Aldehyde hydrosilylation turnover frequencies (TOFs) of up to 4900 min^-1 have been realized, the highest reported for first row metal-catalyzed carbonyl hydrosilylation. Additionally, 1 has been shown to mediate formate dihydrosilylation with leading TOFs of up to 330 min^-1. Under stoichiometric and catalytic conditions, addition of PhSiH₃ to (^Ph2PPrPDI)Mn was found to result in partial conversion to a new diamagnetic hydride compound. Independent preparation of (^Ph2PPrPDI)MnH (2) was achieved upon adding NaEt₃BH to (^Ph2PPrPDI)MnCl₂ and single-crystal X-ray diffraction analysis revealed this complex to possess a capped trigonal bipyramidal solid-state geometry. When 2,2,2-trifluoroacetophenone was added to 1, radical transfer yielded (^Ph2PPrPDI·)Mn(OC·(Ph)(CF₃)) (3), which undergoes intermolecular C-C bond formation to produce the respective Mn(II) dimer, [(μ-O,N_py-4-OC(CF₃)(Ph)-4-H-^Ph2PPrPDI)Mn]₂ (4). Upon finding 3 to be inefficient and 4 to be inactive, kinetic trials were conducted to elucidate the mechanisms of 1- and 2-mediated hydrosilylation. Varying the concentration of 1, substrate, and PhSiH₃ revealed a first order dependence on each reagent. Furthermore, a kinetic isotope effect (KIE) of 2.2 ± 0.1 was observed for 1-catalyzed hydrosilylation of diisopropyl ketone, while a KIE of 4.2 ± 0.6 was determined using 2, suggesting 1 and 2 operate through different mechanisms. Although kinetic trials reveal 1 to be the more active precatalyst for carbonyl hydrosilylation, a concurrent 2-mediated pathway is more efficient for carboxylate hydrosilylation. Considering these observations, 1-catalyzed hydrosilylation is believed to proceed through a modified Ojima mechanism, while 2-mediated hydrosilylation occurs via insertion.

Full text of DOI:10.1021/jacs.7b00879

Subscriber access provided by University of Newcastle, Australia
Article
Mechanistic Investigation of Bis(imino)pyridine Manganese
Catalyzed Carbonyl and Carboxylate Hydrosilylation
Tufan K. Mukhopadhyay, Christopher L. Rock, Mannkyu Hong, Daniel
Charles Ashley, Thomas L. Groy, Mu-Hyun Baik, and Ryan J. Trovitch
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b00879 • Publication Date (Web): 10 Mar 2017
Downloaded from http://pubs.acs.org on March 10, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Mechanistic Investigation of Bis(imino)pyridine Manganese
Catalyzed Carbonyl and Carboxylate Hydrosilylation
9
Tufan K. Mukhopadhyay,† Christopher L. Rock,† Mannkyu Hong,‡,§ Daniel C. Ashley, Thomas
∥
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
L. Groy,† Mu-Hyun Baik,‡,§,* Ryan J. Trovitch†,*
†School of Molecular Sciences, Arizona State University, Tempe, Arizona 85287, United States
‡Department of Chemistry, Korea Advanced Institute of Science and Technology (KAIST),
Daejeon 34141, South Korea
§Center for Catalytic Hydrocarbon Functionalizations, Institute for Basic Science (IBS),
Daejeon 34141, South Korea
∥Department of Chemistry, Indiana University, Bloomington, Indiana 47405, United States
ABSTRACT
We recently reported a bis(imino)pyridine (or pyridine diimine, PDI) manganese precatalyst,
2
(
Ph PPrPDI)Mn (1), that is active for the hydrosilylation of ketones and dihydrosilylation of esters.
In this contribution, we reveal an expanded scope for 1-mediated hydrosilylation and propose
two different mechanisms through which catalysis is achieved. Aldehyde hydrosilylation
turnover frequencies (TOFs) of up to 4,900 min-1 have been realized, the highest reported for
first row metal-catalyzed carbonyl hydrosilylation. Additionally, 1 has been shown to mediate
formate dihydrosilylation with leading TOFs of up to 330 min-1. Under stoichiometric and
catalytic conditions, addition of PhSiH3 to (Ph PPrPDI)Mn was found to result in partial
2
conversion to a new diamagnetic hydride compound. Independent preparation of (Ph PPrPDI)MnH
2
(2) was achieved upon adding NaEt3BH to (Ph PPrPDI)MnCl2 and single-crystal X-ray diffraction
2
analysis revealed this complex to possess a capped trigonal bipyramidal solid-state geometry.
When 2,2,2-trifluoroacetophenone was added to 1, radical transfer yielded
2
(
Ph PPrPDI·)Mn(OC·(Ph)(CF3)) (3), which undergoes intramolecular C-C bond formation to
produce the respective Mn(II) dimer, [(μ-O,Npy-4-OC(CF3)(Ph)-4-H-Ph PPrPDI)Mn]2 (4). Upon
finding 3 to be inefficient and 4 to be inactive, kinetic trials were conducted to elucidate the
mechanisms of 1- and 2-mediated hydrosilylation. Varying the concentration of 1, substrate, and
2
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 36
1
2
3
4
5
6
7
8
9
PhSiH3 revealed a first order dependence on each reagent. Furthermore, a kinetic isotope effect
(KIE) of 2.2 ± 0.1 was observed for 1-catalyzed hydrosilylation of diisopropyl ketone, while a
KIE of 4.2 ± 0.6 was determined using 2, suggesting 1 and 2 operate through different
mechanisms. Although kinetic trials reveal 1 to be the more active precatalyst for carbonyl
hydrosilylation, a concurrent 2-mediated pathway is more efficient for carboxylate
hydrosilylation. Considering these observations, 1-catalyzed hydrosilylation is believed to
proceed through a modified Ojima mechanism, while 2-mediated hydrosilylation occurs via
insertion.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Hydrosilylation catalysts have long been used to mediate Si-C and Si-O bond formation in an
atom-efficient fashion.1 Since Speier’s initial observation of chloroplatinic acid-catalyzed olefin
hydrosilylation,2 precious metal catalysts have been widely used for the production of silicone
adhesives, rubbers, and release coatings.3 Olefin hydrosilylation is known to proceed through the
Chalk-Harrod mechanism, which features Si-H oxidative addition, alkene insertion into M-H,
and reductive elimination of the respective alkylsilane.4 A modified Chalk-Harrod mechanism
featuring alkene insertion into M-SiR3 has also been documented to account for the formation of
unsaturated biproducts.5 Conversely, Si-O formation by way of carbonyl hydrosilylation has
emerged as a mild route to alcohols that can be conducted in a stereoselective manner.6 In the
presence of monohydrosilanes, precious metal catalysts including Wilkinson’s catalyst7 are
believed to mediate carbonyl hydrosilylation via the Ojima Mechanism (Fig. 1), whereby Si-H
oxidative addition, carbonyl insertion into M-SiR3, and reductive elimination yields the
respective silyl ether.8 Recent investigations have suggested that electrophilic silylene
intermediates are responsible for Rh-catalyzed carbonyl reduction in the presence of dihydro-
and trihydrosilanes.9 Similarly, σ-silane complexes of Ru and Ir are believed to mediate carbonyl
hydrosilylation by promoting substrate nucleophilic attack at Si.10
2
ACS Paragon Plus Environment

Page 3 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Ojima mechanism for carbonyl hydrosilylation.
Although precious metal catalyzed hydrosilylation pathways have been extensively
investigated, efforts to develop low-cost first row metal hydrosilylation catalysts have only
recently started to intensify.11-16 Base metal catalysts are advantageous from a cost and toxicity
perspective;17 however, they can be difficult to study since they tend to engage in one electron
reaction pathways and often adopt high-spin electronic configurations.18 This has led to a relative
dearth of mechanistic information regarding first row metal carbonyl hydrosilylation, and the
catalysts which have been thoroughly studied operate through widely varied mechanisms.19 For
example, although a fair number of Mn-based hydrosilylation catalysts have been reported,20 the
mechanisms through which they operate remain poorly understood. Moreover, it is unclear
whether low-valent Mn0,21 MnI,22 and MnII catalysts23 proceed via similar or different
mechanisms than each other or the lone high-valent Mn catalyst that is known, (3,5-tBu2-
salen)MnN.24
In 2014, we reported that formally zerovalent (Ph PPrPDI)Mn (Fig. 2, 1) achieves ketone
2
hydrosilylation TOFs of up to 76,800 h-1 (more accurately expressed as 1,280 min-1) and ester
dihydrosilylation TOFs of up to 18 h-1 under ambient conditions (Fig. 2).25 Herein, we describe
an expanded scope with higher activities for 1-mediated aldehyde and formate hydrosilylation,
the identification of an in situ generated Mn-H co-catalyst, and a comprehensive kinetic analysis
that suggests concurrent modified Ojima and Mn-H insertion mechanisms are responsible for
catalysis. With an updated understanding of catalyst electronic structure and reactivity, the role
of PDI chelate denticity and non-innocence is also discussed.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Ketone hydrosilylation (A) and ester dihydrosilylation (B) using 1.25
RESULTS AND DISCUSSION
Expansion of Substrate Scope. Upon evaluating the ketone and ester hydrosilylation activity
of 1, we hypothesized that higher TOFs might be achieved when screening less sterically
demanding aldehyde- and formate-containing substrates. Knowing that 1 mediates efficient and
exothermic ketone hydrosilylation under neat conditions using PhSiH3 (other silanes afford
slower conversion),25 an equimolar solution of benzaldehyde and PhSiH3 was added to 0.1 mol%
1 at 25 °C. After 2 min, the reaction was exposed to air to deactivate the catalyst and NMR
spectroscopic analysis revealed complete benzaldehyde hydrosilylation (TOF = 490 min-1) to a
98:2 ratio of PhSi(OCH2Ph)3:PhSiH(OCH2Ph)2. Treatment with 10% aq. NaOH allowed for the
isolation of benzyl alcohol (94%, Table 1, a) following extraction and solvent evaporation (Fig.
S3 and S4 of the Supporting Information). For comparison, 13 additional aldehydes were
evaluated for hydrosilylation under these conditions (b-n). Notably, p-fluoro, p-chloro, and p-
bromobenzaldehyde were efficiently hydrosilylated after 2 min (b-d) while p-iodobenzaldehyde
reached only 32% conversion (e). Although electron-withdrawing and -donating groups were
found to slow the rate of 1-catalyzed ketone hydrosilylation in benzene-d6 solution,25
benzaldehyde substitution does not adversely influence the hydrosilylation rate under neat
conditions (f-i). Substrate conversion was also unaffected by the N- and O-heterocyclic moieties
of pyridine-3-carboxaldehye and furfural (k-l) and the alkene functionalities of 3-cyclohexene-1-
4
ACS Paragon Plus Environment

Page 5 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
carboxaldehyde and citral (m-n). Importantly, 1 was found to exhibit chemoselectivity for the
hydrosilylation of aldehydes over nitro (f), nitrile (g), and alkene functionalities (m-n). Catalyst
homogeneity was confirmed by conducting 1-mediated benzaldehyde hydrosilylation in the
presence of excess Hg0, which did not affect conversion. Benzaldehyde hydrosilylation was also
observed using 1 in absence of light, disfavoring the possibility of light-induced catalysis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Hydrosilylation of aldehydes using 0.1 mol% 1 and PhSiH3 at 25 °C.a,b
aPercent conversion as determined by 1H NMR spectroscopy (TOF = 490 min-1 for all substrates
except e). bIsolated yields of the corresponding alcohol in parentheses (spectroscopic analysis provided in
SI).
To determine the highest attainable TOF for 1-catalyzed aldehyde hydrosilylation, efforts
were made to further reduce the catalyst loading. Adding 0.01 mol% 1 to an equimolar solution
of benzaldehyde and PhSiH3 (Table 2, a) resulted in vigorous bubbling due to the exothermic
nature of this reaction. After 2 min, the catalyst was quenched upon opening to air and NMR
spectroscopic analysis revealed greater than 99% conversion to a mixture of silyl ethers, equating
to a TOF of 4,900 min-1. The same conversion was observed upon adding a stock solution of 1 in
PhSiH3 to a mixture of PhSiH3 and benzaldehyde. Comparable hydrosilylation activity was noted
for 4-fluorobenzaldehyde, 2-naphthaldehyde, and furfural (b-d) and hydrolysis of the silyl ether
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 36
1
2
3
4
5
6
7
8
9
products yielded the corresponding alcohols in good yield (Table 2). To determine the maximum
TON for 1-catalyzed aldehyde hydrosilylation, a solution containing 10,000 equiv. of dry PhSiH3
and benzaldehyde was added to 1. After allowing the reaction to cool to ambient temperature for
15 min, a second solution of 10,000 equiv. PhSiH3/benzaldehyde was added. This process was
repeated three additional times until a total of 50,000 equiv. of substrate and silane were added.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Analysis of the product mixture by H NMR spectroscopy revealed 63% overall conversion,
equating to a TON of 31,000. The observed hydrosilylation TOFs are approximately 4-times
higher than previously reported for 1-mediated ketone hydrosilylation25 and are significantly
higher than those reported for leading Fe (393 min-1),12f Co (49.5 min-1),13f and Ni (38.4 min-1)14c
carbonyl hydrosilylation catalysts.23 The maximum TON of 31,000 is also more than double the
benzaldehyde hydrosilylation TON achieved using (κ5-N,N,N,N,N-PyEtPDEA)Mn.23
Table 2. Aldehyde hydrosilylation at low catalyst loading.a,b
aPercent conversion as determined by 1H NMR spectroscopy (TOF = 4,900 min-1 for all substrates).
bIsolated yields of the corresponding alcohol in parentheses.
Knowing that 1 catalyzes the dihydrosilylation of esters with TOFs of up to 18 h-1,25 it was
anticipated that a comparable activity enhancement might be observed upon investigating the
dihydrosilylation of formates. When a neat equimolar mixture of either methyl or ethyl formate
and PhSiH3 was added to 0.02 mol% of 1, an exothermic reaction ensued, resulting in complete
conversion to a mixture of methoxy and ethoxy (in the case of ethyl formate) silyl ethers.
Attempts to hydrolyze each product mixture and separate the product alcohols by fractional
6
ACS Paragon Plus Environment

Page 7 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
distillation were unsuccessful; therefore, higher molecular weight formates were chosen to allow
formate dihydrosilylation optimization. Adding an equimolar quantity of benzyl formate and
PhSiH3 to 0.02 mol% of 1 resulted in complete substrate reduction within 15 min, and hydrolysis
using 10% aq. NaOH followed by extraction and evaporation afforded pure benzyl alcohol with
good yield (Table 3, a). Complete conversion was also observed upon adding a stock solution of
1 in PhSiH3 to benzyl formate. Similarly, p-anisyl formate, phenylethyl formate, hexyl formate,
and isoamyl formate were fully hydrosilylated using 0.02 mol% 1 over the course of 15 min and
the corresponding alcohols were obtained in good-to-modest yield (b-e). For each of these
transformations, 1 achieved formate dihydrosilylation with a TOF of 330 min-1, orders of
magnitude faster than the best known carboxylate dihydrosilylation catalysts.26 Although
citronellyl formate and geranyl formate were distilled and scrupulously dried prior to screening,
it was not possible to reproducibly transform these substrates with 0.02 mol% loadings of 1.
These substrates were cleaved in 15 min to yield the corresponding alcohols when using 0.1
mol% 1 (f-g, TOF = 66 min-1), indicating that substrate purity is critically important when
maximizing the efficacy of Mn-catalyzed hydrosilylation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. Hydrosilylation of formates using 0.02-0.1 mol% 1 or 2 and PhSiH3 at 25 °C.a,b
aPercent conversion as determined by 1H NMR spectroscopy (TOF = 330 min-1 for a-e; 66 min-1 for f-
g). bIsolated yields of the corresponding alcohol in parentheses (spectroscopic analysis provided in SI).
cTrial conducted using 0.1 mol% 1 or 2.
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 36
1
2
3
4
5
6
7
8
9
Identification of catalytically relevant species. When analyzing unquenched post-catalysis
solutions by multinuclear NMR spectroscopy, partial conversion of 1 to a new Mn complex
31
featuring a 1H NMR triplet at -2.98 ppm (J = 112.4 Hz) and P NMR singlet at 69.19 ppm was
observed. In an attempt to characterize this compound, one equiv. of PhSiH3 was added to a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
solution of 1 in benzene-d6. Approximately 20% of the desired product was observed after 24 h
at 25 °C; however, longer reaction times did not result in further conversion. Fortunately,
25
treating (Ph PPrPDI)MnCl2 with 2 equiv. of NaEt3BH allowed for the independent preparation
2
and isolation of the C2-symmetric diamagnetic hydride compound, (Ph PPrPDI)MnH (eqn 1, 2).
2
Upon formulating 2 as a six-coordinate monohydride, efforts were made to determine if this
compound possesses a traditional octahedral solid-state geometry featuring a low-spin Mn(I)
center. Crystallization of 2 from a concentrated toluene solution layered with diethyl ether
afforded single crystals suitable for X-ray diffraction analysis, which revealed the geometry
about Mn to be a hydride-capped trigonal bipyramid (Fig. 3) with N(1)-Mn(1)-N(3) and P(1)-
Mn(1)-P(2) angles of 158.81(9) and 123.85(3), respectively (metrical parameters provided in
Table 4 and Table S2). The Mn(1)-N(1), Mn(1)-N(2), and Mn(1)-N(3) distances were found to
be 1.945(2), 1.860(2), and 1.947(2) Å, respectively, suggestive of a low spin Mn center. The
imine bond lengths of 1.351(3) and 1.342(3) Å are significantly elongated relative to unreduced
PDI chelates (1.28 Å),27 while the C(2)-C(3) and C(7)-C(8) bonds are contracted to 1.424(4) and
1.418(3) Å, respectively. Comparing these parameters to those of 1 (Table 4)25 and literature
values consistent with PDI reduction,27 our crystallographic data suggests that 2 possesses a
Mn(III) center that is supported by a doubly-reduced PDI chelate.
8
ACS Paragon Plus Environment

Page 9 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. The molecular structure of 2 shown at 30% probability ellipsoids. The asymmetric unit
was found to possess a second molecule of 2, which has been omitted for clarity. Hydrogen
atoms other than H1M are also omitted. The inset highlights the capped trigonal bipyramidal
geometry about Mn (considering axial imine N atoms) with aryl groups removed for clarity.
Table 4. Bond lengths (Å) and angles () determined for 1 (previously reported),25 2, and 3.
2a
1
3
Mn(1)-N(1)
Mn(1)-N(2)
Mn(1)-N(3)
Mn(1)-P(1)
Mn(1)-P(2)
Mn(1)-H(1M)
Mn(1)-O(1)
Mn(1)-N(2A)
N(1)-C(2)
1.944(2)
1.887(2)
1.949(2)
2.2697(8)
2.2634(8)
-
1.945(2) [1.951(2)]
1.860(2) [1.862(2)]
1.947(2) [1.948(2)]
2.2067(8) [2.2117(8)]
2.2158(7) [2.2064(7)]
1.57(3) [1.51(3)]
-
2.303(2)
2.159(2)
2.270(3)
2.8055(9)
-
-
-
2.025(2)
2.429(2)
1.288(4)
1.291(4)
1.486(4)
1.485(4)
-
1.354(3)
1.355(3)
1.416(4)
1.414(3)
1.351(3) [1.343(3)]
1.342(3) [1.346(3)]
1.424(4) [1.418(4)]
1.418(3) [1.419(4)]
N(3)-C(8)
C(2)-C(3)
C(7)-C(8)
N(1)-Mn(1)-N(3)
P(1)-Mn(1)-P(2)
P(1)-Mn(1)-N(2A)
N(2)-Mn(1)-H(1M)
N(2)-Mn(1)-O(1)
157.54(9)
105.93(3)
158.81(9) [158.64(9)]
123.85(3) [128.30(3)]
139.32(9)
-
-
-
-
-
167.55(6)
-
174.27(9)
175.0(10) [172.9(10)]
-
aMetrical parameters for second molecule of 2 given in brackets. For additional information, see Fig. S1
and Table S2 of the SI.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 36
1
2
3
4
To elucidate the electronic structures of 1 and 2, density functional theory (DFT) calculations
5
6
7
8
were performed according to methods previously used for our study of
[(Ph PPrPDI)Mn(CO)][Br].28 We sampled various possible spin states for 1 and found a doublet
2
state that is energetically preferred in pseudo-trigonal bipyramidal coordination geometry. This
formally low-spin doublet (SMn = ½) configuration also gave the best agreement with the
experimental crystal structure, as enumerated in Table S4. The quartet state, where three Mn-
based electrons remain unpaired, is 10.5 kcal/mol higher in energy and is therefore not relevant
for the ground state electronic structure of 1. This finding is in good agreement with our spin
state assignment based on experimental observations only.25 Interestingly, a more detailed
analysis of the doublet ground state reveals non-classical electronic features that were previously
not recognized. The calculated Mulliken spin densities of ~1.5 for Mn and the shapes of the
Kohn-Sham orbitals suggest that the electronic structure of complex 1 is best described as a
superposition between the two resonance structures shown in Fig. 4. The qualitative MO diagram
at the left of Fig. 4a describes a classical low-spin system that standard ligand field theory may
suggest and that we had originally envisioned.25 There are two doubly occupied MOs that are
metal-centered and one that is singly occupied. The PDI ligand electrons reside in one of two
nearly degenerate π*-orbitals (Fig. 4b), specifically, the one possessing two nodal planes within
the pyridyl moiety labeled as 1π*(1b1*).
Alternatively, the Mn(II)-d5 complex may use an open-shell configuration where the Mn-
center formally adopts an intermediate-spin configuration exposing three unpaired metal-based
electrons that are antiferromagnetically coupled to unpaired electrons in the 1π* and 2π* PDI-
ligand orbitals as illustrated in Fig. 4b, explaining the aforementioned α-electron spin density of
~1.5 on Mn. Based on our calculations, we estimate the J-coupling to be -401 cm-1. This non-
classical electronic structure gives rise to biradical character on the PDI moiety, although the
redox state of the ligand has not changed. One plausible way of visualizing how the classical
low-spin Mn(II) complex shown in Fig. 4a may transform to the non-classical open-shell
configuration is a double electron transfer whereby one α-electron is transferred from the PDI-π*
orbital to the Mn-dxy orbital, while a β-electron from the Mn-dxz orbital is transferred back to the
PDI-2π* orbital, as shown in blue and red in Fig. 4, respectively (refer to Fig. S61).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10
ACS Paragon Plus Environment

Page 11 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Resonance configurations describing the electronic structure of 1. (a) Original low spin
electronic structure featuring PDI electron pairing. (b) Intermediate spin electronic structure with
antiferromagnetic coupling. (c) The frontier MOs of 1π*(b1*) and 2π*(a2*).
Formally adding a hydrogen atom oxidatively to 1 affords the Mn(III)-d4 hydride complex 2.
Five different spin states, namely closed-shell singlet, open-shell singlet, triplet, quintet and
septet, were considered. Interestingly, the open-shell singlet state is electronically unstable in the
sense that the antiferromagnetically coupled electrons on the PDI ligand and the metal center are
unable to remain in different orbitals. The presence of the hydride ligand in the trans-position to
the PDI-pyridyl moiety impacts the two π* orbitals in a notably different way: because there is
significant N-p orbital component in 1π*, whereas there is no N-character in 2π*, the former is
pushed to a much higher energy than the latter. As a consequence, the two excess electrons on
the PDI(-II) ligand move into 2π* and are paired, as illustrated qualitatively in Fig. 5. As a result,
the radical character on the PDI ligand that was predicted for 1 is lost; in 2 the two ligand-based
electrons are spin-paired.
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Energy-level diagram for 2.
The hydride ligand has a similarly profound impact on the metal-centered MOs, as
highlighted in Fig. 5. The dx2-y2-based MO, which was the lowest energy frontier orbital in 1,
engages in a strong σ-bonding interaction with the hydride ligand, thus pushing the
complementary antibonding combination to be the highest energy frontier orbital. Among all
spin-states considered, we found that this closed-shell singlet state featuring a diamagnetic PDI
dianion29 is energetically most favorable with the next lowest energy state being the triplet at a
relative solution phase free energy of 10.0 kcal/mol. It also gave the best agreement with the
experimental structure. The calculated P(1)–Mn(1)–P(2) angle was markedly different between
the states; the singlet state calculation gave an angle of 128.2°, which most closely reproduces
the experimentally determined values of 123.85 and 128.30°, while the higher spin states provide
angles of 148.4° or greater (Table S5). The collection of the metal-centered frontier orbitals can
be visualized as that of a classical low-spin system, as illustrated in Fig. 5. We suggest that the
preference for the low spin state arises from the interaction between the relatively hard Lewis
acidic Mn(III) center and the strong field hydride ligand. Moreover, comparing the doublet
complex 1 with singlet complex 2 reveals that the latter possesses shorter computed Mn-PDI
bond lengths, with Mn(1)–N(1) and Mn(1)-P(1) distances of 1.950 and 2.202 Å, respectively.
Our updated electronic structure descriptions for both compounds are provided in Fig. 6; the
12
ACS Paragon Plus Environment

Page 13 of 36
Journal of the American Chemical Society
1
2
3
4
experimental and computational investigations on 2 reveal a superposition of both available
5
6
resonance structures as opposed to the localized structure shown.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Lewis structure of 1 and 2 most consistent with the computed electronic structures.
During our investigation of ketone hydrosilylation,25 adding 1 mol% 1 to a benzene-d6
solution containing PhSiH3 and 2,2,2-trifluoroacetophenone resulted in an uncharacteristic
bluish-green solution. Complete reduction of this substrate to the quaternary silane product,
PhSi(OCH(CF3)(Ph))3, was observed after 12 h at 25 °C; however, delayed turnover onset was
also noted (no conversion after 2 h). Given that all other ketone, aldehyde, ester, and formate
solutions remain light brown throughout catalysis, we set out to further investigate the origin of
disparate reactivity between 2,2,2-trifluoroacetophenone and the other substrates. When a
stoichiometric quantity of 2,2,2-trifluoroacetophenone was added to a THF solution of 1, an
instantaneous color change from brownish-orange to deep blue was observed. The resulting
compound was isolated and found to be 1H NMR silent, although evidence for an uncoordinated
phosphine arm was observed by 31P NMR spectroscopy (Fig. S47). The solution phase magnetic
moment of this product was determined to be 4.4 B (benzene-d6), which implies an S = 3/2 Mn
center. Importantly, its electronic spectrum was found to exhibit two absorption maxima at 360
nm ( = 3,410 M-1cm-1) and 612 nm ( = 3,610 M-1cm-1) (Fig. S48). Although the extinction
coefficients are slightly higher, the wavelengths of these Mn-to-PDI chelate charge transfer
bands are nearly identical to those observed for the 5-coordinate monobromide complex, (κ4-
P,N,N,N-Ph PPrPDI)MnBr (361 nm,  = 3,040 M-1cm-1; 614 nm,  = 2,650 M-1cm-1).28 The
2
magnetic moment and electronic spectrum provide strong evidence that this blue product is (κ4-
P,N,N,N-Ph PPrPDI·)Mn(OC·(Ph)(CF3)) (eqn 2, 3), which features a high-spin Mn(II) center that
2
is antiferromagnetically coupled to both PDI and 2,2,2-trifluoroacetophenone radical anions.
Observation of radical transfer from one redox-active ligand (PDI) to another (2,2,2-
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 36
1
2
3
4
5
6
7
8
trifluoroacetophenone) is unique and the electron-withdrawing trifluoromethyl substituent
renders this substrate more easily reducible than the others that have been screened (which do
not react with 1 in the absence of silane).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To further confirm the electronic structure of 3, attempts were made to obtain single crystals
of this compound for X-ray diffraction analysis. Although unsuccessful, storing a concentrated
THF solution of 3 at -35 C for 24 h afforded light amber crystals that were suitable for X-ray
diffraction. The solid-state structure of the light amber complex revealed [(μ-O,Npy-4-
OC(CF3)(Ph)-4-H-Ph PPrPDI)Mn]2 (4) (Fig. 7), which arises from coupling between the PDI
2
radical anion of 3 and the 2,2,2-trifluoroacetophenone radical anion of a neighboring 3 molecule.
Although the PDI radical anion of 3 is delocalized, C-C bond formation takes place at the para-
position of the central pyridine ring to yield 4. Complex 4 features a distorted octahedral
geometry about Mn(II) consisting of an alkylated κ4-P,N,N,N-Ph PPrPDI anion, an alkoxide moiety
2
with a short Mn(1)-O(1) contact of 2.025(2) Å, and a weak σ-bonding interaction to the amido
functionality of a neighboring chelate [Mn(1)-N(2A) of 2.429(2) Å]. Notably, dimerization of 3
by way of radical coupling destroys the extended π-system of Ph PPrPDI. This is clearly seen in
2
the central pyridine ring, where N(2) and C(5) become sp3-hybridized. The C(3)-C(4) and C(6)-
C(7) distances of 1.350(4) and 1.353(4) Å represent true double bonds, while the C(4)-C(5) and
C(5)-C(6) contacts of 1.504(5) and 1.503(4) Å indicate a C-C single bond. Likewise, the
Ph PPrPDI imine substituents in 4 [1.288(4) and 1.291(4) Å] retain double bond character while
2
the Cimine-Cpyridine distances are representative of single bonds [1.486(4) and 1.485(4) Å]. The
Mn(1)-NPDI and Mn(1)-P(1) distances are consistent with a high spin Mn(II) center. Once
formed, crystals of 4 are completely insoluble in common solvents.
14
ACS Paragon Plus Environment

Page 15 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. The solid-state structure of 4 shown at 30% probability ellipsoids. Atom labels
ending with “A” were generated by symmetry. Hydrogen atoms and three co-crystallized THF
molecules per asymmetric unit have been omitted for clarity.
Kinetic Experiments. Upon isolating compounds relevant to catalysis, the mechanism of 1-
mediated hydrosilylation was investigated. Knowing that 1 possesses an unpaired Mn-based
electron (S = 1/2), the possibility that this precatalyst promotes a radical chain mechanism had to
be considered. To gain insight, AIBN, a radical chain initiator with t1/2 = 1 h at 85 C, was added
to an equimolar mixture of PhSiH3 and diisopropyl ketone under neat conditions. No evidence
for carbonyl hydrosilylation was observed after 2 h at 90 C, suggesting that 1 does not achieve
hydrosilylation through a radical initiated pathway. This assessment is further supported by the
absence of coupled silane or substrate following catalysis (unused PhSiH3 remains),25 which are
likely termination products of a radical chain mechanism. Adding a stoichiometric quantity of
PhSiH3 and PhSiD3 (50 equiv. each) to 1 in the absence of substrate did not result in label
scrambling after 12 h at ambient temperature, as judged by 1H NMR spectroscopy.
Considering that 3 is observed upon adding 2,2,2-trifluoroacetophenone to 1, the possibility
that catalysis is achieved following electron transfer to the substrate also had to be evaluated.30
Radical transfer from a Co/Zr heterobimetallic complex to benzophenone was recently reported
by Thomas and co-workers and the resulting ketyl radical is known to undergo concerted
homolytic cleavage in presence of silane to generate silyl ethers.31 To determine if 1 mediates
hydrosilylation in this fashion, kinetic trials were performed to compare the relative rates of
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 36
1
2
3
4
5
6
7
8
9
carbonyl hydrosilylation using 1, 2, and 3. Because 1-mediated aldehyde hydrosilylation is
exceptionally fast, equimolar solutions of PhSiH3 and diisopropyl ketone in 0.7 mL of benzene-
d6 were added to 1.0 mol% 1-3, and the disappearance of substrate was monitored over time by
1H NMR spectroscopy (Fig. 8). As anticipated, 1 was found to efficiently convert diisopropyl
ketone to a mixture of silyl ethers at a rate of 4.48 x 10-4 M/s. Surprisingly, 2 was slightly less
active for this transformation (3.87 x 10-4 M/s), while radical-transferred complex 3 was
considerably less effective (0.36 x 10-4 M/s). These rates suggest that 3 must first react with
PhSiH3 to enter the catalytic cycle (in benzene-d6, it takes 48 h for 3 to react with stoichiometric
PhSiH3), which is consistent with our prior observation of delayed onset during 2,2,2-
trifluoroacetophenone hydrosilylation.25 Moreover, attempts to achieve diisopropyl ketone
hydrosilylation using isolated 4 as the catalyst were unsuccessful, indicating that radical transfer
to the substrate is best considered a catalyst deactivation pathway.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0
100
200
300
400
500
600
Time (s)
Figure 8. Diisopropyl ketone hydrosilylation at 1.0 mol% catalyst loading in 0.7 mL benzene-d6.
Rates of 1 (), 2 (), and 3 () catalyzed reactions are 4.48 x 10-4, 3.87 x 10-4, and 0.36 x 10-4
M/s, respectively.
With radical chain and radical transfer pathways excluded, 1-catalyzed diisopropyl ketone
hydrosilylation was further examined to establish a rate law. Using 1.0 mol% 1 and a constant
concentration of PhSiH3 in benzene-d6, the rate of disappearance for six different concentrations
16
ACS Paragon Plus Environment

Page 17 of 36
Journal of the American Chemical Society
1
2
3
4
1
of diisopropyl ketone was determined by H NMR spectroscopy with respect to anisole as an
5
6
7
8
internal standard (each data point was collected in triplicate, see Fig. S49-S51 for additional
information). A first order dependence on diisopropyl ketone concentration was found according
to Fig. 9 (A). Likewise, variation of silane and catalyst concentration also revealed a linear
dependence on the observed rate (Fig. 9, B and C). Using the same procedure, the rate law of 1-
catalyzed isopropyl formate dihydrosilylation was determined. When isopropyl formate
disappearance was followed under varying concentrations of substrate, PhSiH3, and 1 in
benzene-d6, a first order dependence on each reagent was observed (Fig. 9, D, E, and F).
Therefore, the rate expressions for 1-catalyzed ketone and formate hydrosilylation can be
formulated as:
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Rate = kobs[1][ketone][PhSiH3] (3)
Rate = kobs[1][formate][PhSiH3] (4)
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 36
1
2
3
4
5
6
7
8
9
12
10
8
8
A
B
7
6
5
4
3
2
1
0
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
2
0
0
0.2
0.4
0.6
0.8
0
0.2
0.4
0.6
0.8
0.015
0.6
[Diisopropylketone] (M)
[PhSiH3] (M)
12
10
8
3.5
3
C
D
2.5
2
6
1.5
1
4
2
0.5
0
0
0
0.005
[(Ph2PPrPDI)Mn] (M)
0.01
0
0.1
0.2
0.3
0.4
0.5
0.6
[Isopropylformate] (M)
3.5
3
6
5
4
3
2
1
0
E
F
2.5
2
1.5
1
0.5
0
0
0.1
0.2
0.3
0.4
0.5
0
0.1
0.2
0.3
0.4
0.5
0.6
[(Ph2PPrPDI)Mn] (M)
[PhSiH3] (M)
Figure 9. Dependence of [ketone] (A, R2 = 0.97), [silane] (B, R2 = 0.98), and [1] (C, R2 = 0.94)
on the rate of 1-catalyzed diisopropyl ketone hydrosilylation. Dependence of formate (D, R2 =
0.98), silane (E, R2 = 0.98), and 1 (F, R2 = 0.96) on the rate of 1-catalyzed isopropyl formate
hydrosilylation.
Under identical conditions, a kinetic isotope effect (KIE) for 1-catalyzed diisopropyl ketone
hydrosilylation was established by comparing the rates of substrate disappearance in the presence
of stoichiometric PhSiH3 and PhSiD3. Diisopropyl ketone was hydrosilylated using PhSiH3 at a
rate of 5.65 x 10-4 M/s, and at a modestly slower rate of 2.53 x 10-4 M/s using PhSiD3, equating
to a KIE of 2.2 ± 0.1 (Fig. 10, A). Primary KIEs of this magnitude are consistent with Si-H bond
18
ACS Paragon Plus Environment

Page 19 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
oxidative addition at (or contributing to) the rate determining step of the transformation. Since
independently prepared 2 mediates carbonyl hydrosilylation at appreciable rates, the same
transformations were conducted using this catalyst. A KIE of 4.1 ± 0.6 was observed for 2-
mediated diisopropyl ketone hydrosilylation (Fig. 10, B), indicating that Si-H bond breaking
using 2 requires higher reorganization energy, and that 1 and 2 achieve carbonyl hydrosilylation
through different mechanisms.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Upon finding that 1 and 2 operate independently, the carboxylate dihydrosilylation efficiency
of each catalyst in the presence of stoichiometric PhSiH3 was assayed by 1H NMR spectroscopy.
At 25 °C in benzene-d6, it was found that 1.0 mol% 1 mediates the reductive cleavage of
isopropyl formate at a rate of 2.26 x 10-4 M/s. Surprisingly, 2 was found to be considerably more
efficient for this transformation, with an observed rate of 7.37 x 10-4 M/s (Fig. 10, C). Likewise,
2 was found to catalyze ethyl acetate dihydrosilylation at a faster rate than 1 (Fig. 10, D),
suggesting the mechanism for 2-mediated carbonyl hydrosilylation is better suited for
carboxylate reduction than the carbonyl hydrosilylation mechanism for 1. The synthetic utility of
2 for formate dihydrosilylation has been further demonstrated by repeating each trial in Table 3
using 0.02-0.1 mol% 2. Each transformation was exothermic and found to reach greater than
99% conversion within 15 min.
Finally, the role of the chelate phosphine donors during catalysis was probed according to
different, yet complementary approaches. Since 1 and 2 feature a pentadentate Ph PPrPDI chelate,
2
it is logical to hypothesize that phosphine arm dissociation must preclude Si-H activation (in the
case of 1) or substrate binding (at 2). To determine if exogenous phosphine inhibits the rate of 1-
mediated diisopropyl ketone hydrosilylation, kinetic trials were conducted upon adding 20, 40,
and 80 equiv. of triphenylphosphine. Phosphine addition had no observable effect on the initial
rate of diisopropyl ketone consumption (6.07 ± 0.035 x 10-4 Ms-1, Fig. S55), suggesting that
phosphine coordination does not inhibit catalysis. Moreover, adding 20 equiv. of PPh3 to 1 did
not result in observable phosphine displacement by 31P NMR spectroscopy. To determine
whether these groups are essential for enabling (PDI)Mn-based carbonyl hydrosilylation, an aryl-
substituted PDI precatalyst, (2,6-iPr PhPDI)Mn(THF)2,32 was screened for benzaldehyde
2
hydrosilylation activity. At 1.0 mol% loading in the presence of stoichiometric PhSiH3, complete
conversion of the substrate to a mixture of silyl ethers was observed after 10 min, indicating that
phosphine co-donors are not required for (PDI)Mn hydrosilylation activity. Rather, we have
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 36
1
2
3
4
found that the phosphine co-donors of 1 and 2 serve to stabilize the precatalyst, rendering it less
5
6
susceptible to decomposition and slightly more tolerant of substrate impurity.
7
8
9
0.5
0.45
0.4
0.5
0.45
0.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.35
0.3
0.35
0.3
0.25
0.2
0.25
0.2
0.15
0.1
0.15
0.1
0.05
0
0.05
0
A
B
0
100
200
300
400
500
600
0
100
200
300
400
500
600
Time (s)
Time (s)
0.5
0.45
0.4
0.5
0.45
0.4
0.35
0.3
0.35
0.3
0.25
0.2
0.25
0.2
0.15
0.1
0.15
0.1
0.05
0
0.05
0
C
D
0
100
200
300
Time (s)
400
500
0
1000
2000
3000
4000
Time (s)
5000
6000
7000
Figure 10. The disappearance of diisopropyl ketone in the presence of PhSiH3 () and PhSiD3
() catalyzed by 1.0 mol% 1 (A, KIE = 2.2 ± 0.1). Rates are 5.654 x 10-4 M/s (PhSiH3) and
2.534 x 10-4 M/s (PhSiD3). The disappearance of diisopropyl ketone in the presence of PhSiH3
() and PhSiD3 () catalyzed by 1.0 mol% 2 (B, KIE = 4.1 ± 0.6). Rates are 4.96 x 10-4 M/s
(PhSiH3) and 1.19 x 10-4 M/s (PhSiD3). Isopropyl formate dihydrosilylation using PhSiH3 and
1.0 mol% of 1 (, 2.26 x 10-4 M/s) and 2 (, 7.37 x 10-4 M/s) at 25 °C in benzene-d6 (C). Ethyl
acetate dihydrosilylation using PhSiH3 and 1.0 mol% of 1 (, 0.16 x 10-4 M/s) and 2 (, 0.41 x
10-4 M/s) at 25 °C in benzene-d6 (D).
Hydrosilylation Mechanisms. Given the first order dependence of substrate, silane, and 1 on
the rate of carbonyl hydrosilylation and carboxylate dihydrosilylation (eqn 3-4), a mechanism
that incorporates both transformations can be proposed. Considering the spectator role of the
Ph PPrPDI phosphine co-donors and the fact that 1 can hydrosilylate sterically demanding ketones
such as diisopropyl ketone and dicyclohexyl ketone,25 this catalyst is believed to operate though
2
an Ojima-like mechanism (Fig. 11) that is modified from the general example in Fig. 1. The
20
ACS Paragon Plus Environment

Page 21 of 36
Journal of the American Chemical Society
1
2
initial step of the transformation involves phosphine dissociation33 and reversible Si-H oxidative
addition to generate the 5-coordinate silyl hydride intermediate A. Although formation of A is
consistent with a primary Si-H(D) KIE of 2.2 ± 0.1, it is not believed to be the rate-determining
step. Rather, A is believed to reversibly bind substrate to give intermediate B, which undergoes
insertion into the Mn-H bond to generate C. The insertion step is believed to be rate-determining,
consistent with having a first-order dependence on both substrate and silane. Moreover, insertion
of substrate into the Mn-Si bond (as in Fig. 1), is disfavored since it would require formation of a
high-energy intermediate that features a bulky tertiary alkyl ligand. In the case of ketone
hydrosilylation, C undergoes fast reductive elimination to furnish the hydrosilylated product
while regenerating catalyst 1. If Si-H bonds remain, the resulting silyl ether is more reactive than
PhSiH3 and further participates in hydrosilylation to form doubly and triply condensed products,
explaining the observation of unreacted PhSiH3 after each catalytic reaction.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
During 1-catalyzed formate or ester dihydrosilylation, fast -alkoxide elimination is believed
to occur following insertion to give alkoxide intermediate D. Like C, D reductively eliminates
the corresponding silyl ether, but also releases aldehyde upon reforming 1. Efforts to observe
aldehyde produced in situ by 1H NMR spectroscopy have been unsuccessful since aldehydes are
quickly reduced in the presence of excess PhSiH3. The carboxylate dihydrosilylation cycle in
Fig. 11 features -alkoxide elimination after the rate-determining step, consistent with the rate
law established for 1-catalyzed isopropyl formate dihydrosilylation (eqn 4). Throughout the
carbonyl hydrosilylation and carboxylate dihydrosilylation cycles in Fig. 11, Ph PPrPDI acts as a
2
redox non-innocent ligand, allowing the Mn center to remain divalent during catalysis.34 As
indicated, both phosphine donors of Ph PPrPDI are free to dissociate and do not inhibit the rate of
2
hydrosilylation.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. Modified Ojima mechanism for 1-catalyzed aldehyde (R’ = H, R’’ = alkyl or aryl),
ketone (R’, R’’ = alkyl or aryl), formate (R’ = H, R’’ = OR’’’; where R’’’ = alkyl), and ester (R’
= alkyl, R’’ = OR’’’; where R’’’ = alkyl or aryl) hydrosilylation.
Knowing PhSiH3 reacts with 1 to generate small quantities of a Mn-H compound under
stoichiometric and catalytic conditions, and that independently prepared 2 is active for carbonyl
hydrosilylation and carboxylate dihydrosilylation, we believe a second pathway contributes to
the overall mechanism. It is proposed that intermediate A occasionally undergoes silyl ligand
loss or transfer to chelate, as observed for isoelectronic Mn dialkyl complexes,35 to generate 2 or
a yet to be identified silylated-PDI variant of 2, allowing catalysis to proceed through the Mn-H
insertion pathway in Fig. 12. Starting from 2, chelate phosphine donor displacement by an
incoming carbonyl moiety33 leads to E, which undergoes migratory insertion to yield Mn(II)
alkoxide intermediates of type F. Since a KIE of 4.1 ± 0.6 was observed for 2-mediated
diisopropyl ketone hydrosilylation, it is believed that the rate determining step involves σ-bond
metathesis between the Mn-O bond of F and the Si-H bond of an incoming silane, a pathway that
requires more reorganization energy than Si-H oxidative addition. This assessment is further
supported by kinetic data which shows a dependence on substrate, silane, and 2 during
diisopropyl ketone hydrosilylation (Fig. S56-S60 of the SI).
22
ACS Paragon Plus Environment

Page 23 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 12. Insertion mechanism for 2-catalyzed aldehyde (R’ = H, R’’ = alkyl or aryl), ketone
(R’, R’’ = alkyl or aryl), formate (R’ = H, R’’ = OR’’’; where R’’’ = alkyl), and ester (R’ = alkyl,
R’’ = OR’’’; where R’’’ = alkyl or aryl) hydrosilylation.
In the case of formate or ester dihydrosilylation, fast -alkoxide elimination at F occurs to
form intermediate G. Upon losing aldehyde, -bond metathesis between the Mn-O bond of H
and silane can occur to yield the silyl ether product and starting catalyst. As in Fig. 11, the
aldehyde released in the carboxylate dihydrosilylation cycle in Fig. 12 is quickly hydrosilylated
under the reaction conditions to yield the corresponding silyl ether. Once phosphine substituent
dissociation occurs, the carbonyl and carboxylate hydrosilylation pathways accessed by 2 are
believed to proceed at Mn(II); however, the PDI chelate likely remains a radical monoanion
throughout catalysis. Considering the mechanistic pathways for 1 and 2, our kinetic data affirms
that the carbonyl hydrosilylation cycle mediated by 1 (Fig. 11, right) is faster than the carbonyl
hydrosilylation cycle mediated by 2 (Fig. 12, right). Conversely, comparing the rates of 1- and 2-
catalyzed isopropyl formate (Fig. 10, C) and ethyl acetate reduction (Fig. 10, D) demonstrates
that 2 is the more efficient catalyst for carboxylate dihydrosilylation.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 36
1
2
3
4
CONCLUSION
5
6
7
8
By optimizing the conditions of 1-mediated aldehyde hydrosilylation, TOFs of up to 4,900
min-1 have been demonstrated and a benzaldehyde hydrosilylation TON of 31,000 has been
achieved. Extending this investigation to the reduction of formates revealed dihydrosilylation
TOFs of up to 330 min-1, considerably faster than what has been reported for comparable first
row metal catalyzed reactions. Importantly, the characterization and study of a Mn-H compound
relevant to catalysis, 2, revealed that this precursor also exhibits exceptional formate
dihydrosilylation activity. In contrast, radical transfer from the chelate of 1 to incoming 2,2,2-
trifluoroacetophenone has been found to result in catalyst deactivation through intermolecular
radical coupling to yield 4. Kinetic analysis of 1-mediated carbonyl and formate hydrosilylation
revealed a first order dependence on substrate, silane, and catalyst, a primary KIE, and that
catalysis is unaffected by exogenous phosphine, indicating that both transformations are
achieved through a modified Ojima mechanism. Kinetic analysis of 2-mediated hydrosilylation
revealed a large KIE that has been attributed to rate-determining σ-bond metathesis after
substrate insertion. Comparing the activity of 1- and 2-mediated hydrosilylation has revealed that
1 is more active for carbonyl hydrosilylation, while 2 is more efficient for carboxylate
dihydrosilylation. The mechanistic complexity underlying (PDI)Mn hydrosilylation is surprising
and the efforts described herein may prove influential for the study of other leading first row
metal catalysts.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General Considerations. Unless otherwise stated, all synthetic manipulations were
performed in an MBraun glovebox under an atmosphere of purified nitrogen. Aldrich or Acros
anhydrous solvents were purified using a Pure Process Technology solvent system and stored in
the glovebox over activated 4Å molecular sieves and sodium before use. Benzene-d6 was
purchased from Cambridge Isotope Laboratories and dried over 4Å molecular sieves and
potassium before use. (THF)2MnCl2 and 4-fluorobenzaldehyde were purchased from Acros.
1,3,5,7-Cyclooctatetraene was used as received from Strem Chemicals. Mercury, sodium
triethylborohydride, benzaldehyde, p-anisaldehyde, p-tolualdehyde, o-nitrobenzaldehyde,
furfural, citral, diisopropyl ketone, 2,2,2-trifluoroacetophenone, p-anisyl formate, and citronellyl
formate were obtained from Sigma Aldrich. Phenylsilane, pyridine-3-carboxaldehyde, 4-
cyanobenzaldehyde, 4-iodobenzaldehyde, and 4-bromobenzaldehyde were purchased from
24
ACS Paragon Plus Environment

Page 25 of 36
Journal of the American Chemical Society
1
2
3
4
Oakwood
Chemicals.
4-Chlorobenzaldehyde,
2-naphthaldehyde,
3-cyclohexene-1-
5
6
7
8
carboxaldehyde, and anisole were acquired from TCI America. Methyl formate, ethyl formate,
benzyl formate, 2-phenylethyl formate, hexyl formate, and isoamyl formate were purchased from
TCI America and dried over sieves before use. Geranyl formate was obtained from Santa Cruz
Chemicals and dried over sieves before use. 3-(Diphenylphosphino)-1-propylamine,36
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
25
38
2,6-
Ph PPrPDI,37
(
Ph PPrPDI)MnCl2,
(
Ph PPrPDI)Mn
(1),25
phenylsilane-d3,
and
(
2
2
2
iPr PhPDI)Mn(THF)232 were prepared according to literature procedure.
2
1
Solution H nuclear magnetic resonance (NMR) spectra were recorded at room temperature
on a Bruker 400, Varian 400, or Varian 500 MHz NMR spectrometer. All H and 13C NMR
chemical shifts (ppm) are reported relative to Si(CH3)4 using H (residual) and 13C chemical
shifts of the solvent as secondary standards. 31P NMR data is reported relative to H3PO4.
Elemental analyses were performed at Robertson Microlit Laboratories Inc. (Ledgewood, NJ)
and the Arizona State University CLAS Goldwater Environmental Laboratory (Tempe, AZ).
Solution state magnetic susceptibility was determined from Evans method using the Varian 400
MHz NMR spectrometer.
1
1
X-ray Crystallography. Single crystals suitable for X-ray diffraction were coated with
polyisobutylene oil in the glovebox and transferred to glass fiber with Apiezon N grease, which
was then mounted on the goniometer head of a Bruker APEX Diffractometer equipped with Mo
Kα radiation (Arizona State University). A hemisphere routine was used for data collection and
determination of the lattice constants. The space group was identified and the data was processed
using the Bruker SAINT+ program and corrected for absorption using SADABS. The structures
were solved using direct methods (SHELXS) completed by subsequent Fourier synthesis and
refined by full-matrix, least-squares procedures on [F2] (SHELXL). The solid-state structure of 2
features two unique molecules in the asymmetric unit. The hydride ligand was located in the
difference map for each molecule. The solid-state structure of 4 features three co-crystallized
THF molecules per asymmetric unit, one of which was modeled with two partial occupants. The
molecular structure of 4 is a dimer that is symmetry generated by an inversion center midway
between the two metal atoms. Crystallographic parameters for 2 and 4 can be found in Table S1
of the SI.
General method of aldehyde hydrosilylation using 0.1 mol% 1: In the glove box, a 20 mL
scintillation vial was charged with 0.0022 g of 1 (0.0033 mmol). An equimolar mixture of
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 36
1
2
3
4
5
6
7
8
PhSiH3 (3.3 mmol) and aldehyde (3.3 mmol) was then added. CAUTION: This reaction is
exothermic and vigorous bubbling occurs. After 2 min, the solution was exposed to oxygen to
1
deactivate the catalyst, filtered through Celite, and analyzed by H NMR spectroscopy to
determine conversion. The silyl ether products were then hydrolyzed by stirring with 2 mL of
10% aqueous NaOH solution for 2 h. The organic layer was extracted with diethyl ether (5 x 2
mL), dried over anhydrous Na2SO4, and the solvent was removed by rotary evaporation (40 °C)
to isolate the corresponding alcohol.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General method of aldehyde hydrosilylation using 0.01 mol% 1: In the glove box, a 100
mL round bottom flask was charged with 0.0022 g of 1 (0.0033 mmol). An equimolar mixture of
PhSiH3 (33 mmol) and aldehyde (33 mmol) was then added. CAUTION: This reaction is
exothermic and vigorous bubbling occurs. After 2 min, the solution was exposed to oxygen to
1
deactivate the catalyst, filtered through Celite, and analyzed by H NMR spectroscopy to
determine conversion. The silyl ether products were then hydrolyzed by stirring with 2 mL of
10% aqueous NaOH solution for 2 h. The organic layer was extracted with diethyl ether (5 x 2
mL), dried over anhydrous Na2SO4, and the solvent was removed by rotary evaporation (40 °C)
to isolate the corresponding alcohol.
General method of formate dihydrosilylation using 0.02 mol% 1 or 2: In the glove box a
100 mL round bottom flask was charged with 0.0020 g of catalyst 1 or 2 (0.0033 mmol). An
equimolar mixture of PhSiH3 (16 mmol) and aldehyde (16 mmol) was then added. CAUTION:
This reaction is exothermic and vigorous bubbling occurs. After 15 min, the solution was
1
exposed to oxygen to deactivate the catalyst, filtered through Celite, and analyzed by H NMR
spectroscopy to determine conversion. The silyl ether products were then hydrolyzed by stirring
with 2 mL of 10% aqueous NaOH solution for 2 h. The organic layer was extracted with diethyl
ether (5 x 2 mL) and dried over anhydrous Na2SO4. The solvent and co-produced methanol were
removed by rotary evaporation to isolate the higher molecular weight alcohol.
General method of formate hydrosilylation using 0.1 mol% 1 or 2: In the glove box a 20
mL scintillation vial was charged with 0.0013 g of catalyst 1 or 2 (0.0019 mmol). An equimolar
mixture of PhSiH3 (1.9 mmol) and aldehyde (1.9 mmol) was then added. CAUTION: This
reaction is exothermic and vigorous bubbling occurs. After 15 min, the solution was exposed to
oxygen to deactivate the catalyst, filtered through Celite, and analyzed by 1H NMR spectroscopy
26
ACS Paragon Plus Environment

Page 27 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
to determine conversion. The silyl ether products were then hydrolyzed by stirring with 2 mL of
10% aqueous NaOH solution for 2 h. The organic layer was extracted with diethyl ether (5 x
2 mL) and dried over anhydrous Na2SO4. The solvent and co-produced methanol were removed
by rotary evaporation to isolate the higher molecular weight alcohol.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of (Ph PPrPDI)MnH (2): Under N2 atmosphere, a 250 mL round bottom flask
2
25
was charged with 0.344 g (0.466 mmol) of (Ph PPrPDI)MnCl2 in approximately 80 mL of
2
toluene. The slurry was cooled to liquid N2 temperature in a cold well for 30 min. A 20 mL
scintillation vial was charged with 0.114 g (0.932 mmol) of NaEt3BH (1 M in toluene) in
approximately 10 mL of toluene and also cooled for 30 min. Then the NaEt3BH solution was
added to the slurry of (Ph PPrPDI)MnCl2 slowly while stirring. The flask was allowed to warm to
2
25 °C and a deep brownish-green color was observed after 20 min, which continued to darken
over time. After 6 h the brownish-green solution was filtered through Celite and the toluene was
evacuated to obtain a dark solid. The solid was washed with pentane (4 x 5 mL) and dried again.
It was washed quickly with 10 mL of ether (2 x 5 mL) to remove any free ligand generated in the
reaction. It was then filtered through Celite using 15 mL of toluene. The toluene was removed in
vacuo and the residue was dissolved in a minimum amount of diethyl ether and placed at -35 °C
overnight. Dark green crystals identified as 2 (0.125 g, 0.187 mmol, 40%) were collected after
drying. Analysis for C39H42N3P2Mn: Calcd. C, 69.94%; H, 6.32%; N, 6.27%. Found: C, 68.26%;
1
H, 6.44%; N, 5.88%. H NMR (benzene-d6): δ (ppm) = 8.17 (d, JH-H = 8.1 Hz, 2H, pyridine),
7.76 (t, JH-H = 8.1 Hz, 1H, pyridine), 7.47 (t, JH-H = 8.1 Hz, 4H, phenyl), 7.09 (m, 4H, phenyl),
6.61 (m, 6H, phenyl), 5.66 (t, JH-H = 8.1 Hz, 4H, phenyl), 4.50 (d, JH-H = 11.4 Hz, 2H, -CH2),
2.83 (t, JH-H = 11.2 Hz, 2H, -CH2), 2.48 (broad m, 4H, -CH2), 2.34 (s, 6H, CH3), 1.78 (broad m,
2H, -CH2), 1.36 (broad m, 2H, -CH2), 1.26 (broad m, 2H, -CH2), -2.98 (t, JP-H = 112.4 Hz, 1H,
Mn-H). 13C NMR (benzene-d6): δ (ppm) = 154.57 (C=N), 151.46 (Ar), 144.10 (dd, J = 22.3, 18.6
Hz, Ar), 137.64 (Ar), 133.48 (dd, J = 18.3, 13.9 Hz, Ar), 133.01 (t, J = 6.4 Hz, Ar), 131.35 (t, J =
3.7 Hz, Ar), 129.42 (Ar), 129.21 (t, J = 5.1 Hz, Ar), 128.98 (Ar), 115.29 (Ar), 110.37 (Ar), 57.47
31
(CH2), 33.18 (d, J = 19.2, 16.9 Hz, PCH2), 29.70 (CH2), 14.40 (CH3). P{1H} NMR (benzene-
d6): δ (ppm) = 69.19 (s).
Preparation of (Ph PPrPDI·)Mn(OC·(Ph)(CF3)) (3): Under an inert atmosphere, a 20 mL
2
scintillation vial was charged with 0.108 g (0.162 mmol) of 1 in approximately 10 mL benzene.
To the dark brown solution, 0.023 mL (0.162 mmol) of 2,2,2-trifluoroacetophenone was added.
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 36
1
2
3
4
5
6
7
8
9
The solution turned deep blue in color immediately. It was stirred for 1 h and then filtered
through Celite. The filtrate was dried under vacuum to isolate a blue residue. After repeated
washing with diethyl ether (5 x 3 mL) followed by drying in vacuo, 0.078 g (0.092 mmol, 56%)
of a dark blue solid identified as 3 was isolated. Analysis for C47H46N3P2MnOF3: Calcd. C,
66.77%; H, 5.48%; N, 4.96%. Found: C, 66.69%; H, 5.35%; N, 4.78%. 1H NMR (benzene-d6): 
(ppm) No resonances observed. Magnetic moment (Evans method, benzene-d6): eff = 4.4 B.
UV-vis (from 7 independent concentrations in toluene, Fig. S48): λmax = 360 nm (ε = 3,410 M-1
cm-1), 612 nm (ε = 3,610 M-1 cm-1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of [(μ-O,Npy-4-OC(CF3)(Ph)-4-H-Ph PPrPDI)Mn]2 (4): A concentrated THF
2
solution of 3 (0.078 g, 0.092 mmol) was placed at -35 °C for 24 h. The blue solution was
removed and the remaining solid was washed with toluene (3 x 2 mL) and Et2O (2 x 2 mL).
Drying in vacuo yielded 0.027 g (0.016 mmol, 17%) of amber crystals identified as 4. Analysis
for C94H92N6P4Mn2O2F6: Calcd. C, 66.77%; H, 5.48%; N, 4.96%. Found: C, 66.02%; H, 5.66%;
N, 4.75%. 1H NMR (benzene-d6):  (ppm) No resonances observed.
General method for kinetic data collection: In the glove box, a J. Young tube was charged
with a benzene-d6 (0.7 mL) solution of 0.002 g of catalyst. The tube was placed into a liquid N2
jacketed cold well to freeze the catalyst solution. While frozen, PhSiH3 was added, followed by
the substrate (ketone or ester). Finally, 10 L of anisole was added as an internal standard. The
tube was sealed under N2 atmosphere and promptly transferred into a dewar filled with liquid N2.
1
After warming the reaction mixture to room temperature, a H NMR experiment was quickly
started and spectral data was collected at 1 min (or 30 s) intervals. The substrate consumption
rate was calculated by integrating the starting material and product with respect to anisole. Each
experiment was duplicated or triplicated to determine an average rate and standard deviation.
Computational Data. All calculations were performed using density functional theory39
(DFT) implemented in the Jaguar 9.140 suite of ab initio quantum chemistry programs. The
geometry optimization calculations were carried out with the PBE41 functional. 6-31G** basis
set42 was used for all atoms, except for I and Br, which were represented by the Los Alamos
LACVP basis set that consist of relativistic effective core potentials.43 The energies of the
optimized structures were reevaluated by single-point calculations using PBE-D344 functional
28
ACS Paragon Plus Environment

Page 29 of 36
Journal of the American Chemical Society
1
2
along with Dunning’s correlation consistent triple-ζ basis set cc-pVTZ(-f)45 that includes a
double set of polarization functions.
3
4
5
6
7
8
9
The optimized stationary points were confirmed through analytic computation of harmonic
force constants. The analytical vibrational frequencies computed with the PBE/6-31G** level of
theory were used to derive zero point energy and vibrational entropy corrections from the
unscaled frequencies which has only one imaginary frequency. Solvation energies were
evaluated by a self-consistent reaction field (SCRF) approach46 using a dielectric constant of ε =
2.284 along with the optimized gas phase structures.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The energy components were computed with the protocol below (eqn 5-8). G(gas) being the
Gibbs free energy in gas phase; G(solv) being the free energy of solvation; H(gas) being the
enthalpy in gas phase; T being the temperature; S(gas) the entropy in gas phase; E(SCF) being
the self-consistent field energy; ZPE being the zero point energy. The spin-spin coupling
constant J for the antiferromagnetic coupling in complex 1 has been evaluated using
Noodleman’s spin-projection technique (eqn 9).47
∆G(Sol) = ∆G(gas) + ∆G(solv) (5)
∆G(gas) = ∆H(gas) - T∆S(gas) (6)
∆H(gas) = ∆E(SCF) + ∆ZPE + ∆CvT + ∆PV (7)
∆H(gas) ≈ ∆E(SCF) + ∆ZPE (8)
E
–E
BS HS
퐽 =
(9)
2S
S
+S
HS BS BS
ASSOCIATED CONTENT
Supporting Information. Detailed hydrosilylation procedures, crystallographic information for
2 (CCDC - 1518909) and 4 (CCDC - 1518910), NMR spectroscopic data, Cartesian coordinates
of the DFT-optimized structures, and computational details. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
Email: mbaik2805@kaist.ac.kr; ryan.trovitch@asu.edu
Phone: (480) 727-8930
29
ACS Paragon Plus Environment