Mechanistic studies of the dehydrocoupling and dehydropolymerization of amine-boranes using a [Rh(Xantphos)]+ catalyst

Article abstract of DOI:10.1021/ja503335g

A detailed catalytic, stoichiometric, and mechanistic study on the dehydrocoupling of H₃B·NMe₂H and dehydropolymerization of H₃B·NMeH₂ using the [Rh(Xantphos)]⁺ fragment is reported. At 0.2 mol % catalyst loadings, dehydrocoupling produces dimeric [H₂B-NMe₂]₂ and poly(methylaminoborane) (M_n = 22 700 g mol^-1, PDI = 2.1), respectively. The stoichiometric and catalytic kinetic data obtained suggest that similar mechanisms operate for both substrates, in which a key feature is an induction period that generates the active catalyst, proposed to be a Rh-amido-borane, that reversibly binds additional amine-borane so that saturation kinetics (Michaelis-Menten type steady-state approximation) operate during catalysis. B-N bond formation (with H₃B·NMeH ₂) or elimination of amino-borane (with H₃B· NMe₂H) follows, in which N-H activation is proposed to be turnover limiting (KIE = 2.1 ± 0.2), with suggested mechanisms that only differ in that B-N bond formation (and the resulting propagation of a polymer chain) is favored for H₃B·NMeH₂ but not H ₃B·NMe₂H. Importantly, for the dehydropolymerization of H₃B·NMeH₂, polymer formation follows a chain growth process from the metal (relatively high degrees of polymerization at low conversions, increased catalyst loadings lead to lower-molecular-weight polymer), which is not living, and control of polymer molecular weight can be also achieved by using H₂ (M_n = 2 800 g mol^-1, PDI = 1.8) or THF solvent (M_n = 52 200 g mol^-1, PDI = 1.4). Hydrogen is suggested to act as a chain transfer agent in a similar way to the polymerization of ethene, leading to low-molecular-weight polymer, while THF acts to attenuate chain transfer and accordingly longer polymer chains are formed. In situ studies on the likely active species present data that support a Rh-amido-borane intermediate as the active catalyst. An alternative Rh(III) hydrido-boryl complex, which has been independently synthesized and structurally characterized, is discounted as an intermediate by kinetic studies. A mechanism for dehydropolymerization is suggested in which the putative amido-borane species dehydrogenates an additional H₃B·NMeH₂ to form the real monomer amino-borane H₂B=NMeH that undergoes insertion into the Rh-amido bond to propagate the growing polymer chain from the metal. Such a process is directly analogous to the chain growth mechanism for single-site olefin polymerization.

Full text of DOI:10.1021/ja503335g

Subscriber access provided by the Henry Madden Library | California State University, Fresno
Article
Mechanistic Studies of the Dehydrocoupling and Dehydropolymerization
of Amine-Boranes using a [Rh(Xantphos)]+ Catalyst
Heather C. Johnson, Erin M. Leitao, George R. Whittell,
Ian Manners, Guy C. Lloyd-Jones, and Andrew S. Weller
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja503335g • Publication Date (Web): 20 May 2014
Downloaded from http://pubs.acs.org on May 26, 2014
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 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Mechanistic Studies of the Dehydrocoupling and Dehydropolymerization of
Amine-Boranes using a [Rh(Xantphos)]+ Catalyst
9
Heather C. Johnson,a Erin M. Leitao,b George R. Whittell,b Ian Manners,b*
Guy C. Lloyd-Jones,c* and Andrew S. Wellera*
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
a
Department of Chemistry, Chemistry Research Laboratories, University of Oxford,
Mansfield Road, Oxford. OX1 3TA. U.K. Email: andrew.weller@chem.ox.ac.uk
b
School of Chemistry, University of Bristol, Cantock’s Close, Bristol. BS8 1TS. U.K.
Email: ian.manners@bristol.ac.uk
c School of Chemistry, University of Edinburgh, West Mains Road, Edinburgh. EH9 3JJ.
U. K. Email: guy.lloyd-jones@ed.ac.uk
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 53
1
2
3
4
Abstract
5
6
7
8
A detailed catalytic, stoichiometric and mechanistic study on the dehydrocoupling of
H3B·NMe2H and dehydropolymerization of H3B·NMeH2 using the [Rh(Xantphos)]+
fragment is reported. At 0.2 mol% catalyst loadings dehydrocoupling produces dimeric
[H2B=NMe2]2 and poly(methylaminoborane) (Mn = 22 700 g mol–1, PDI = 2.1),
respectively. The stoichiometric and catalytic kinetic data obtained suggest that similar
mechanisms operate for both substrates, in which a key feature is an induction period
that generates the active catalyst, proposed to be an amido–borane, that reversibly
binds additional amine–borane so that saturation kinetics (Michaelis-Menten type
steady-state approximation) operate during catalysis. B–N bond formation (with
H3B·NMeH2) or elimination of amino–borane (with H3B·NMe2H) follows, in which N–H
activation is proposed to be turn–over limiting (KIE = 2.1 ± 0.2), with suggested
mechanisms that only differ in that B–N bond formation (and the resulting propagation of
a polymer chain) is favoured for H3B·NMeH2 but not H3B·NMe2H. Importantly, for the
dehydropolymerization of H3B·NMeH2 polymer formation follows a chain growth process
from the metal (relatively high degrees of polymerization at low conversions, increased
catalyst loadings lead to lower molecular weight polymer), that is not living, and control
of polymer molecular weight can be also achieved by using H2 (Mn = 2 800 g mol–1, PDI
= 1.8) or THF solvent (Mn = 52 200 g mol–1, PDI = 1.4). Hydrogen is suggested to act as
a chain transfer agent in a similar way to the polymerization of ethene, leading to low
molecular weight polymer, while THF acts to attenuate chain transfer and accordingly
longer polymer chains are formed. In situ studies on the likely active species present
data that support an amido–borane intermediate as the active catalyst. An alternative
hydrido–boryl complex, which has been independently synthesised, and structurally
characterized, is discounted as an intermediate by kinetic studies. A mechanism for
dehydropolymerization is suggested in which the putative amido–borane species
dehydrogenates an additional H3B·NMeH2 to form the “real monomer” amino–borane
H2B=NMeH that undergoes insertion into the Rh–amido bond to propagate the growing
polymer chain off the metal. Such a process is directly analogous to the chain growth
mechanism for single–site olefin polymerization.
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
2
ACS Paragon Plus Environment

Page 3 of 53
Journal of the American Chemical Society
1
2
3
4
1. Introduction
5
6
7
8
Catalytic routes for the formation of main–group/main–group bonds are important for the
targeted construction of new molecules and materials. However, enabling catalytic
methodologies for such bond forming events lag behind those developed for the
construction of C–C and C–X bonds.1 The development of reliable, robust and
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
controllable
processes
is
thus
an
important
challenge.2-5
Catalytic
dehydropolymerization6 of amine–boranes to give polyaminoboranes presents one such
opportunity, as this produces new BN polymeric materials that are isoelectronic with
technologically pervasive polyolefins. Such new materials have potential applications as
high performance polymers and as precursors to BN-based ceramics and single layer
hexagonal BN thin films (white graphene).7 Although ill-defined branched, oligomeric
materials that have been termed “polyaminoborane” have historically been prepared by
non–catalytic methods,8-11 it is only recently that high molecular weight, essentially linear
polyaminoboranes have been produced by catalytic methods from amine–boranes such
as H3B·NH3 and H3B·NMeH2 (Scheme 1), initially using Brookhart’s catalyst
Ir(tBuPOCOPtBu)H2 [tBuPOCOPtBu = κ3-PCP-1,3-(OPtBu2)2C6H3].12
tBu2
O
P
when R = Me
Mn = 55 200
PDI = 2.9
H
Ir
R = H, Me
H
O
P
tBu2
≤ 0.5 mol%
H3B·NRH2
H2B·NRH
n
– H2
Scheme 1. Dehydropolymerization of amine–boranes using the Ir(tBuPOCOPtBu)H2 catalyst.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 53
1
2
3
4
5
6
7
8
9
In 2006 Goldberg, Heinekey and co-workers demonstrated that H3B·NH3 could be
dehydrooligomerized using this Ir catalyst to afford an insoluble material tentatively
reported as [H2BNH2]5,13,14 but later assigned as linear polyaminoborane [H2BNH2]n (n =
ca. 20) on the basis of solid–state 11B NMR spectroscopy by Manners and coworkers.15
In 2008 the former group16 also described that the dehydrooligomerization of
H3B·NMeH2 at low relative concentrations of amine–borane, or mixtures of the latter with
H3B·NH3, gave low molecular weight but soluble oligomers (Mn less than ca. 2,500 g
mol-1). Independently in 2008, Manners and co-workers17 reported the production of high
molecular weight [H2BNMeH]n (Mn = 55,200 g mol-1, PDI = 2.9) and related materials at
low catalyst loadings (0.3 mol%) using both high and low concentrations of
substrates.15,17 More recently photoactivated catalysts based upon [CpFe(CO)2]2 have
been reported to dehydropolymerize H3B·NMeH2 to [H2BNMeH]n (Mn = 64,500 g mol-1,
PDI = 1.83),18 as have Mn(η5–C5H5)(CO)3, Cr(η6–C6H6)(CO)3 and Cr(CO)6 for the cases
of H3B·NRH2 (R = Me or Et) under similar conditions.19,20 Catalysts based upon
[Rh(Ph2P(CH2)4PPh2)]+ also show good activities (0.2 mol%) in producing high
molecular weight poly(methylaminoborane), [H2BNMeH]n, from H3B·NMeH2 (Mn =
144,000 g mol-1, PDI = 1.25).21 Fe(PhNCH2CH2NPh)(Cy2PCH2CH2PCy2)22 and
complexes based upon “bifunctional” Ru(PNP)(H)(PMe3) [PNP = HN(CH2CH2PiPr2)2]23
also catalyze polyaminoborane formation, the latter at very low (less than 0.1 mol%)
loadings. Ionic liquids have also been shown to support the formation of
polyaminoboranes from H3B·NH3 when used in conjunction with metal–based
catalysts.24 It is also noteworthy that anionic oligomerization approaches to both linear
and branched short chain aminoboranes have recently been described.25,26
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
ACS Paragon Plus Environment

Page 5 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Mechanistic studies focussing on the dehydropolymerization of H3B·NMeH2 or H3B·NH3
substrates are few in number. Nevertheless important observations and overarching
rationales have been suggested from these studies. This relative dearth can be
compared to studies with H3B·NMe2H, which are considerably more numerous, and
often demonstrate subtle differences in likely mechanistic pathways depending on
identity of the metal–ligand fragment.2,18,27-33 The polymer growth kinetics (molecular
weight versus conversion) using the Ir(tBuPOCOPtBu)H2 / H3B·NMeH2 system suggest
the operation of a modified chain–growth mechanism that involves both a slow metal–
based dehydrogenation of amine–borane and faster insertion/polymerization of the
resulting amino–borane.15 Using the same system, sigma–bound amine–borane
intermediates for catalytic redistribution of oligomeric diborazanes have recently been
proposed on the basis of kinetic modelling.34 Using catalyst systems based upon
Fe(PhNCH2CH2NPh)(Cy2PCH2CH2PCy2) / H3B·NH3 an initiation mechanism that
invokes an Fe–amido–borane has been suggested, which then undergoes
dehydrogenative insertion of additional H3B·NH3 to form polyaminoborane.22 For
Ru(PNP)(H)(PMe3) / H3B·NH3 a mechanism is proposed, based upon experimental and
DFT studies, in which amino–borane is formed in a low, but steady state, concentration
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
that undergoes catalysed polymerization by an enchainment reaction that relies upon
16
metal–ligand cooperatively.23 Kinetic studies using the Ir(tBuPOCOPtBu)H2
and
Ru(PNP)(H)(PMe3)23 systems demonstrate a first order dependence on both amine–
borane and catalyst concentrations, although for the latter catalyst when H3B·ND3 was
used there was a zero–order dependence on this substrate suggesting a change in the
turnover limiting step. A number of apparently homogeneous35 catalyst systems show
kinetic profiles that might suggest induction periods prior to rapid dehydropolymerisation
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 53
1
2
3
4
5
6
7
8
9
of H3B·NH3 or H3B·NMe2H,14,21-23 although the underlying reasons for this have only
been addressed in detail for a dehydocoupling catalyst based upon Shvo’s catalyst that
produces borazine rather than polyaminoborane.36
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
On–Metal Dehydropolymerization
Cy2B NRH
RHN
[M]
BH2
NHR
n H3B·NRH2
H
H2RN H2B·NRH
BH3
H3B·NRH2
H3B·NRH2
n
[M]
– H2
– n H2
H3B
[M]
– H2
H
B
R
H
R
H
N
B
N
B
1
/
H2B NRH
3
– 3 H2
N
R
Cy2B NRH
Off–Metal Dehydrocoupling
Scheme 2. Suggested pathways for dehydropolymerization, dehydrogenation and hydroboration.
Adapted from reference 37.
The role of free, transient,38 amino–borane in dehydropolymerization, such as H2B=NH2
or H2B=NMeH, which arises from initial dehydrogenation of amine–borane has attracted
particular attention as these (or very closely related metal–bound species) are likely
boron–containing intermediates. Baker, Dixon and co-workers have suggested that
selectivity in the dehydropolymerization of H3B·NH3 depends on whether the
intermediate amino–borane remains associated with the metal.37 Release from the metal
ultimately results in the formation of borazine by trimerization, whereas if strong
coordination/rapid insertion of amino–borane into the growing polymer chain occurs then
polymerization is favoured (Scheme 2). The generation of transient amino–boranes,
such as H2B=NH2 or H2B=NMeH, during catalysis can also be probed by addition of
6
ACS Paragon Plus Environment

Page 7 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
exogenous cyclohexene, which undergoes hydroboration to form Cy2B=NHR (R = Me,
H).37 Catalyst systems in which amino–borane is suggested to not be released from the
metal do not form the hydroborated product during dehydropolymerization, while for
those that form borazine from trimerization of free amino–borane, or when amino–
borane is produced thermally in the absence of a metal-ligand fragment,34 the
hydroborated product is observed in significant quantities. However, recent experimental
and computational studies using Ir(tBuPOCOPtBu)H2 or Ru(PNP)(H)(PMe3) suggest that
if hydroboration or borazine formation are not kinetically competitive with metal–
promoted B–N coupling then Cy2B=NH2 will not be observed, even if free amino–borane
is formed transiently.23,34 Adding to this complexity, hydrogen redistribution reactions
can also occur, in which amino–boranes take part in hydrogen–transfer with amine–
boranes,34,39 while a nucleophilic solvent (e.g. THF) can also potentially catalyse
polyaminoborane formation from amino–boranes.40
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
Mechanistic insight that comes from the direct observation of intermediates in
dehydropolymerization is also very rare, although off–cycle products have been
reported.13,29,41 The product of the first insertion event of H3B·NMeH2 using the
[Ir(PCy3)2(H)2]+
fragment
has
been
shown
to
be
[Ir(PCy3)2(H)2(η2–
H3B·NMeHBH2·NMeH2)][BArF ] [ArF = 3,5–(CF3)2C6H3],42 in which the resulting
4
diborazane forms a sigma43 complex with the Ir–centre (Scheme 3a). Studies on
closely–related phosphine–borane dehydrocoupling44 using the [Rh(Ph2P(CH2)3PPh2)]+
fragment (which is also an excellent catalyst for amine–borane dehydropolymerization21)
provide complementary insight, and intermediates that sit each side of the
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 53
1
2
3
4
5
6
7
8
9
dehydrocoupling step have been characterised, allowing for activation parameters for
the P–B bond forming event to be determined (Scheme 3b).45-47 These intermediates
show that P–H activation has occurred to give a Rh(III) phosphino hydride with
supporting intra and intermolecular sigma (B–H····Rh) interactions. Using the
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
[Rh(Xantphos)]+
fragment
[Xantphos
=
4,5-bis(diphenylphosphino)-9,9-
dimethylxanthene],48,49 that is valence isoelectronic to [Ir(tBuPOCOPtBu)],15,34 B–B
homocoupling of H3B·NMe3 gives a diborane(4) complex (Scheme 3c). Computation and
experiment point to a pathway in which a low energy reversible B–H activation of
amine–borane is followed by a second, higher energy, B–H activation and B–B
coupling,50 these steps being related to those generally invoked in B–N bond formation
in dehydropolymerization.
(a)
PCy3
PCy3
MeH
N
H3B·NMeH2
NMeH2
NMeH2
BH2
H
H
H
H
H
H
H
H
Ir
B
Ir
B
– H2
H
H
PCy3
PCy3
H
H2
B
H
(b)
F
Ph2
Ph
P
2 P
Rh
P
B
PPh2
Ph2
2 H3B·PPh2H
– H2
H
P
P
H
H
H
H
Rh
Ph2
Ph2
Rh
P
B
P
P
BH2
H
H
Ph2
PPh2H
Ph2
Ph2
(c)
Me
Me
Me
Me
NMe3
BH2
B–H activation
B–B coupling
H shift
H
O
NMe3
O
2 H3B·NMe3
Ph2
P
[Rh]
B
B
H
H
H
H
Ph2
P
H
Rh
P
– H2
Ph2
H
Rh
P
NMe3
not observed
B
Ph2
NMe3
H
Scheme 3. Isolated intermediates in amine–borane, and related, dehydrocoupling.
[BArF ]– anions are not shown. (a) H3B·NMeH2 oligomerization;42 (b) H3B·PPh2H
4
oligomerization; 45,47 (c) B–B homocoupling.50
8
ACS Paragon Plus Environment

Page 9 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Encouraged by the [Rh(Xantphos)]+ fragment’s ability to B–B homocouple amine–
boranes we now report its use in a detailed stoichiometric, catalytic and
mechanistic/kinetic investigation into the dehydropolymerisation of H3B·NMeH2 to form
polyaminoborane. Additional mechanistic and structural data on the processes occurring
comes from the reactions of this fragment with H3B·NMe3, H2B=NiPr2 and H3B·NMe2H.
These studies lead to an overall mechanistic framework for dehydropolymerization using
transition metal fragments that supports, and puts detail upon, the
dehydrogenation/coordination/insertion mechanism proposed by others.15,22,23,28,37 This
insight leads to the gross control of the degree of dehydropolymerization, allowing for
both low and higher molecular weight polyaminoborane to be obtained.
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
2.
Results
2.1
Stoichiometric
Reactivity
of
Precatalyst
[Rh(κ2-POP-Xantphos)(η2-
t
H2B(NMe3)CH2CH2 Bu][BArF ]
4
H3B·NMe3
The stoichiometric reactivity of the [Rh(Xantphos)]+ fragment with amine–boranes is
described first, as this provides base–line reactivity with which to contextualize
subsequent catalytic studies. Many of our previous investigations into the coordination,
reaction and catalytic chemistry of amine and phosphine–boranes with cationic Rh(I)
fragments have used [Rh(L)2(η-arene)][BArF ] (L = phosphine; arene = C6H5F or C6H4F2)
4
precursors as a convenient latent source of the {Rh(L)2}+ fragment, these being formed
from hydrogenation of the corresponding NBD (norbornadiene) adduct in fluorobenzene,
or 1,2–difluorobenzene, solvent.21,45,51,52 Surprisingly, in these solvents, we have not
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 53
1
2
3
4
5
6
been able to make the corresponding Rh(I)–Xantphos fluoroarene precatalyst, as
decomposition to as yet unidentified product(s) occurs. Thus we turned to the previously
7
8
9
reported and structurally characterized50 Rh(I) species [Rh(κ2-PP-Xantphos)(η2-
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
t
H2B(NMe3)(CH2CH2 Bu)][BArF ], 1, and the Rh(III) complex [Rh(κ2-POP-Xantphos)(H)2(η2-
4
H3B·NMe3)][BArF ], 2, as reliable and relatively stable [Rh(Xantphos)]+ precatalysts
4
t
(Scheme 4). Complex 1 has the hydroborated alkene, H2B(NMe3)3CH2CH2 Bu, I, ligated
to the metal centre through two Rh···H–B sigma interactions, while 2 has a H3B·NMe3
bound through a single Rh···H–B interaction. These complexes also demonstrate the
variability in the Xantphos coordination mode, mer–κ3-POP and cis–κ2-PP,53,54 and are
also related to recently reported cationic53,55 and neutral56,57 rhodium dihydride
complexes with Xantphos–type ligands.
tBu
Me
Me
+ H2
H
Ph
P
tBu
2
H3B·NMe3
[Rh(Xanthpos)(NBD)]+
Me
Me
O
Rh
H
O
–
[BArF
]
Ph2
P
4
H
H
H
P
Ph2
Rh
H
B
H
P
B
Ph2
NMe3
NMe3
2
1
Scheme 4. Formation of Rh(I) and Rh(III) starting materials.50 [BArF ]– anions are not
shown. 1,2–F2C6H4 solvent.
4
In solution under an Ar atmosphere complex 2 forms as yet unidentified products
(Scheme 5, 50% in 24 hours), while under an H2 atmosphere it is stable showing no
change after 24 hrs. These observations suggest that irreversible H2 loss from 2 on the
timescale of catalysis (~90 minutes, vide infra) is slow. Addition of the dimeric amino–
borane [H2B=NMe2]2, II, to 2, which has previously been shown to promote H2 loss from
other Rh(III) dihydride species,27,58 resulted in no significant H2 loss over the course of a
10
ACS Paragon Plus Environment

Page 11 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
few hours, although over 24 hours a new species becomes dominant that results from
the reaction of H2B=NMe2, II, with 2 (see Section 2.2). Addition of excess NCMe to 2
forms the previously reported NCMe adduct, 3,55 and free H3B·NMe3, while addition of
excess THF forms a (45:55) mixture of 2 and a complex spectroscopically characterized
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
as the THF adduct: [Rh(κ3-POP-Xantphos)(H)2(THF)][BArF ] 4 (Scheme 5).59 Complex 4
4
also shows very similar NMR data to the analogous acetone adduct: [Rh(κ3-POP
-
Xantphos)(H)2(acetone)][BArF ].55 THF and H3B·NMe3 binding are thus competitive.
4
Although irreversible H2 loss is proposed to be slow, H/D exchange at Rh–H and B–D is
shown to be rapid (on time of mixing) by 1H and 2H NMR spectroscopy when [Rh(κ3-POP
-
Xantphos)(H)2(η1-D3B·NMe3)][BArF ], d3-2, is generated in situ by addition of H2 to a 1:1
4
mixture of [Rh(κ2-PP-Xantphos)(NBD)][BArF ] and D3B·NMe3. Presumably this occurs via
4
B–H activation at the Rh(III) dihydride fragment, via a sigma–CAM mechanism (σ–
complex–assisted metathesis),60 to give a base-stabilised dihydrogen–boryl species61-64
that can then reform to give an alternative isotopomer. However any equilibira operating
must sit far to the side of 2 as there is no evidence by NMR spectroscopy for the
formation of a new species when 2 is placed under H2 (4 atm). Addition of H3B·NMe3 to
1 results in the slow formation of the corresponding diborane(4) complex (Scheme 3c)
that comes from sequential B–H activation in two amine–boranes.50
– H2
SLOW
H
H
Ph
P
2
Ph
P
2
Me
Me
Me
Me
L
Rh
H
O
Rh
H
O
+ H3B·NMe3
H
P
P
Ph2
H
H
B
L
Ph2
NO REACTION
OVER 2 HOURS
NMe3
3 L = NCMe
4 L = THF
2
Me2N BH2
H2B NMe2
Scheme 5. Reactivity of 2. [BArF ]– anions are not shown. 1,2–F2C6H4 solvent.
4
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 53
1
2
3
4
H3B·NMe2H
5
6
Addition of 2 equivalents of H3B·NMe2H to 1 results the immediate (time of mixing)
7
8
9
generation of the analogous complex to 2, [Rh(κ3-POP-Xantphos)(H)2(η1-
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
t
H3B·NMe2H)][BArF ], 5, alongside free H2B(NMe3)CH2CH2 Bu, I (Scheme 6). Complex 5
4
has been characterized by NMR spectroscopy by analogy with 2 (Supporting Materials),
and other sigma borane complexes.65 In particular, in the H NMR spectrum, relative
1
integral 1 H signals are observed at δ –14.11 (br) and δ –19.05 (doublet of triplet of
doublets) for the inequivalent Rh–hydrides, and a broad integral 3 H signal at δ –1.31 is
assigned to the sigma–bound H3B·NMe2H Rh···H–B groups that are interconverting
between bridging and terminal positions.43,61 The 31P{1H} NMR spectrum shows a single
11
environment at δ 44.5 [J(RhP) = 115 Hz], while the B NMR spectrum shows a broad
signal at δ –12, barely shifted from free H3B·NMe2H (δ –12.8), consistent with an η1–
coordination of the amine–borane.51 The amino–borane H2B=NMe2, and its consequent
66
dimer [H2B=NMe2]2, II,
are also formed, that arise from dehydrogenation of
H3B·NMe2H with concomitant transfer of H2 to Rh. Complex 5 is not stable, and is slowly
consumed so that after 5 hours the Rh(III)-dihydride [Rh(κ3-POP-Xantphos)(H)2(η1-
H2B(NMe3)CH2CH2 Bu)][BArF ] 6 is formed, alongside [H2B=NMe2]2 (Scheme 6).
t
4
Complex 6 has been spectroscopically characterized (see Supporting Materials), and
shows very similar data to 2 and 5, but now has the borane I bound to the metal centre.
6 presumably forms after dehydrogenation of 5 (and release of H2) in the absence of
excess H3B·NMe2H. Interestingly 1 and 6 are shown to be in equilibrium with one
another, as addition of H2 (4 atm) to 1 results in a 3:1 mixture of 6 to 1, which is biased
back in favour of 1 on removal of H2. However we discount a significant role for the
12
ACS Paragon Plus Environment

Page 13 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
equilibrium between 6 and 1 during catalysis, based on the following observations: (i) 1
rapidly reacts with H3B·NMe2H to form 5, (ii) 6 only forms slowly at low [H3B·NMe2H]
from 5, (iii) the temporal evolution of catalysis is the same whether starting from Rh(I) or
Rh(III) precursors, and (iv) excess I does not does change the observed temporal profile
of catalysis (vide infra). This is contrast to the auto–catalytic role that the final product
[H2B=NMe2]2 has been shown to take in dehydrocoupling of H3B·NMe2H as catalyzed by
the [Rh(PCy3)2(H)2]+ fragment.27 Addition of D2 to 5/H3B·NMe2H results in H/D exchange
at the B–H and Rh–H positions as well as in the free amine–borane (as measured by 2H
NMR spectroscopy) indicating that reversible B–H activation is a relatively low energy
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
2
process. No H/D exchange was observed at nitrogen (by H NMR spectroscopy),
suggesting that reversible N–H activation is considerably higher in energy, as has been
noted before in related systems.66,67 Slow dehydrogenation of H3B·NMe2H is also
observed.
– H2
+ H2
H
H
Ph
P
2
Ph
Me
Me
slow
P
2
2 H3B·NMe2H
Me
Me
Me2N BH2
H2B NMe2
Rh
H
O
+
1
Rh
H
NMe3
O
H
fast
– H2
P
Ph2
H
P
Ph2
H
H
B
B
– H2B=NMe2
– H2B=NMe2
H
NMe2H
– I
+ I
5
6
II
tBu
Scheme 6. Dehydrocoupling of H3B·NMe2H. [BArF ]– anions are not shown. C6H5F
4
solvent.
H3B·NMeH2
Addition of 2 equivalents of H3B·NMeH2 to 1 resulted in the immediate formation of the
Rh(III) dihydride complex [Rh(κ3-POP-Xantphos)(H)2(η1-H3B·NMeH2)][BArF ] 7 (Scheme
4
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 53
1
2
3
4
5
6
7
8
9
7). Complex 7 was characterised by NMR spectroscopy, and these data are very similar
to those for 2, 5 and 6. The amino–borane that would arise from initial dehydrogenation,
H2B=NMeH, was not observed,38 however, the ultimate thermodynamic product of
dehydrocoupling, N–trimethylborazine III, was formed [δ(11B) 33.3, doublet; lit. δ 33.268].
There was no evidence for the formation of polymeric BN materials or the potential
cyclic triborazane intermediate, [H2BNMeH]3.69 We have recently39 shown that when the
amino–borane H2B=NHtBu is released from a metal center it undergoes trimerisation to
form [HBNtBu]3 by an (unresolved) mechanism in which hydrogen redistribution
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
processes are occurring,34 and it is possible that such processes are also operating here.
As found for 5, complex 7 undergoes a second, slower, dehydrogenation. This process
is a little faster than for 5, taking 2 hours to fully consume 7 to afford III and an
equilibrium mixture of 6/1. Addition of NCMe (excess) to 7 affords the corresponding
MeCN adduct, 3, and free H3B·NMeH2.
H
H
B
Ph
P
2
Me
Me
Me
Me
– H2
2 H3B·NMeH2
N
B
N
B
Rh
H
O
1
6/1 +
H
slow
+ I
fast
H
N
H
P
Ph2
H
H
B
– I
Me
NMeH2
III
7
Scheme 7. Borazine formation at low [H3B·NMeH2]. [BArF ]– anions are not shown.
4
C6H5F solvent.
General Comments on the Stoichiometric Reactivity.
These observations show that under non–catalytic conditions, dehydrogenation of
H3B·NMe2H or H3B·NMeH2 at a Rh(I) centre (i.e. 1) is rapid, while at a Rh(III) dihydride
centre (i.e. 5 or 7) it is slower, even though B–H activation (as measured by H/D
exchange experiments for H3B·NMe3) is fast at the RhH2 center. These observations are
14
ACS Paragon Plus Environment

Page 15 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
similar to those previously reported for the [Rh(PR3)2]+ and [Rh(PR3)2(H)2]+ fragments
respectively.27,51 As will be demonstrated, this slower rate of dehydrogenation of 5 and 7
is in contrast to the fast consumption of H3B·NMe2H or H3B·NMeH2 under catalytic
conditions (e.g. 0.2 mol% 1, H3B·NMe2H 0.072 M). In addition, under catalytic conditions
H3B·NMeH2 is dehydropolymerized to give [H2BNMeH]n rather than forming
trimethylborazine III, and there is an induction period observed before catalysis. These
observations suggest additional mechanistic considerations need to be adopted under
the conditions of high ratios of amine–borane to metal–ligand fragment, and these are
discussed next.
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
2.2 Catalysis.
Initial Experiments using H3B·NMe2H and H3B·NMeH2
Under catalytic conditions (0.2 mol% 1, 0.072 M H3B·NMe2H, 1,2–F2C6H5 solvent, open
system to a slow flow of Ar) complex 1 catalyzes the dehydrogenation of H3B·NMe2H to
ultimately form dimeric II (Scheme 8a). Following this reaction by 11B NMR
spectroscopy using periodic sampling of the reaction mixture shows that there was an
induction period of approximately 400–500 seconds, and H2B=NMe2 was also observed
as an intermediate during the productive phase of catalysis. Turnover is relatively fast
once the induction period is over, with an overall ToF ~1200 hr–1 (ToN = 500); a rate that
is comparable to [Rh(Ph2PCH2CH2CH2PPh2)(η6–C6H5F)][BArF ],21 which also shows an
4
induction period and is suggested to be homogeneous in character. Very similar
temporal profiles are observed starting from the Rh(III) complex, 2 (Supporting
Materials), suggesting that the induction period is not due to the formation of the simple
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 53
1
2
3
4
5
6
7
8
9
Rh(III) analog (i.e. 5), consistent with the rapid formation of 5 from 1 (Scheme 6). This
also argues against the involvement of I during the induction period or catalysis, as 2 is
generated without I being present. At ~30% conversion (~900 s) addition of Hg to the
catalyst solution, or filtration of the catalyst mixture though a 0.2 µm filter and addition of
a further 500 equivalents of H3B·NMe2H, did not result in the termination of catalysis
(see Supporting Materials): observations that suggest a homogeneous system.54 The
catalyst can also be recycled, in that addition of a further 500 equivalents of H3B·NMe2H
to the catalytic mixture directly at the end of catalysis resulted in essentially the same
rate and overall turnover number. There is no induction period observed in this
recharging experiment, or in the filtration experiment, suggesting that the catalyst
remains in its active form in both. No significant amount of the linear diborazane
H3B·NMe2BH2·NMe2H68 was observed, similar to [Rh(Ph2PCH2CH2CH2PPh2)(η6–
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
C6H5F)][BArF ],21 but different from [Rh(PR3)2H2]+ systems where it is observed in
4
significant amounts.27,51,58
16
ACS Paragon Plus Environment

Page 17 of 53
Journal of the American Chemical Society
1
2
3
4
(a)
Open System
5
6
7
8
9
1 0.2 mol%
Me2N BH2
H2B NMe2
H3B·NMe2H
0.072 M
H2
–
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
(b) Closed System
Scheme 8. 11B Time/Concentration plot of the dehydrocoupling of H3B·NMe2H; ▼
H3B·NMe2H, ■ H2B=NMe2, ▲ [H2B=NMe2]2, ● BH(NMe2)2. 0.2 mol% 1, [1] = 1.44 x 10
–4, 0.072 M H3B·NMe2H, 1,2–F2C6H4 solvent, (a) Open system; (b) closed system. Inset
shows the induction period.
In a closed system (New Era© high pressure NMR tube) catalysis is significantly slower
(Scheme 8b), ToF ~130 hr–1 (ToN = 500). A very similar induction period to the open
system is observed, and H2B=NMe2 is also an intermediate. We27 and others23 have
commented previously on the rate inhibition by H2 in amine–borane dehydrocoupling.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 53
1
2
3
4
5
6
7
8
9
For example, with the [Rh(PCy3)2]+ catalyst, H2 build-up forces the system to sit in a
Rh(III)/Rh(III) cycle that turns over considerably slower than the Rh(I)/Rh(III) cycle. The
more active Rh(I) oxidation state is generated by addition of the product II to
[Rh(PCy3)2(H)2]+ that promotes H2 reductive elimination, i.e. autocatalysis. In our system
addition of 200 equivalents of II prior to catalysis (0.2 mol% 1, 0.072 M amine–borane,
open system) resulted in no significant change in the reaction profile, consistent with the
lack of reaction between 1 and II under stoichiometric conditions on the timescale of
catalysis (Scheme 5). Addition of 55 equivalents of I also did not change the catalytic
temporal profile (Supporting Materials) demonstrating that it does not act to modify
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
18
ACS Paragon Plus Environment

Page 19 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Entry
Conditions
1, 0.2 mol%
2, 0.2 mol%
Mn (g mol–1)
22 700
PDI
2.1
1.9
1
2
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
24 800
3
4
1, 0.4 mol%, 0.22 M, H3B·NMeHBH2·NMeH2,
1, 0.2 mol%, further 500 equivs.
15 400
17 900
1.8
1.8
5
6
7
9
1, 1 mol%
7 500
1.5
1, 0.2 mol%, closed
2 800
1.8a
1.4b
1.8
1, 0.2 mol%, THF solvent
1, 0.2 mol%, excess cyclohexene
52 200
38 600
Table 1. Dehydropolymerization data, Mn by GPC. 100% conversion after first measured
point (2 hrs) as determined by 11B NMR spectroscopy. 0.44 M [H3B·NMeH2], open
system, C6H5F unless otherwise stated. greater than 95% conversion, 24 hrs. 85%
conversion, 19 hrs.
a
b
Catalyst 1 also dehydropolymerizes H3B·NMeH2 (0.2 mol% 1, 0.44 M amine–borane,
open system, 2 hrs, C6H5F as solvent) to afford polyaminoborane [H2BNMeH]n (Mn = 22
700 g mol-1, PDI = 1.8 using polystyrene standards for GPC column calibration). This is
lower molecular weight than typically formed using the [Ir(tBuPOCOPtBu)H2] catalyst (Mn
= 55,200 g mol-1, PDI = 2.9) in THF as solvent.15 The Rh(III) catalyst 2 also produced
very similar polymer to that for 1 (Mn = 24 800 g mol-1, PDI = 1.9). These polymers
11
formed show B NMR spectra very similar to that reported for high molecular weight
[H2BNMeH]n produced from [Ir(tBuPOCOPtBu)H2] 17 and [Rh(Ph2PCH2CH2CH2PPh2)(η6–
C6H5F)][BArF ] 21 catalysts, with a broad, symmetrical, peak observed at δ –5.4 (fwhm =
4
720 Hz, Figure 1a).15 No significant signals were observed around δ 0 which might
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 53
1
2
3
4
indicate chain branching,23 although such a feature if small could be lost in the peak
5
6
11
width of the main signal. To the detection limit of B NMR spectroscopy (ca. 5 %) no
7
8
9
signals were observed between δ 30–40 that could be assigned to free MeHN=BH(R).
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. (a) 11B{1H} NMR spectrum of the material that is isolated after
dehydropolymerization of H3B·NMeH2 using 1 (0.2 mol% 0.44 M H3B·NMeH2, open
system, 2 hrs). The signal at δ –17 is assigned to entrained H3B·NMeH2 which reduces
significantly in relative intensity on addition of more 1 (0.2 mol%, Supporting Materials).
(b) Under sealed conditions (H2 build up). The signals at ~ δ 5 and ~ δ –17 split into a
triplet and quartet respectively (Supporting Materials), reminiscent of the signals
observed for H3B·NMeHBH2·NMeH2,42 suggesting the presence of short–chain
oligomers.
20
ACS Paragon Plus Environment

Page 21 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
H
1 0.2 mol%
H
H3B·NMeH2
B
0.44 M
– H2
N
n
Me
H
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
Scheme 9. Polymer conversion plot (triangles), and H2 evolution (squares, gas burette,
calculated at 26 ºC), for the dehydrocoupling of H3B·NMeH2. For polymer conversion
each point is a separate experiment in C6H5F, with the product precipitated with hexane.
The conversion of H3B·NMeH2 (δ –17.8, q) relative to [H2BNMeH]n (δ –5.4, br) measured
by 11B{1H} NMR spectroscopy (THF solvent).
A time/conversion plot for H3B·NMeH2 dehydrocoupling to form polyaminoborane using
catalyst 1 in an open system is shown in Scheme 9 alongside a hydrogen evolution plot,
as measured by gas–burette. As for H3B·NMe2H there is a significant induction period
(10 minutes) before the rapid dehydrocoupling occurs. Polymer formation and hydrogen
evolution track one another, and by the end of catalysis (7200 seconds, 98% conversion,
ToF ~250 hr–1) just over 1 equivalent of H2 has been produced, consistent with the
formation of polyaminoborane of empirical formula approximating to [H2BNMeH]n. This
reaction is considerably slower than for H3B·NMe2H, but this might reflect the poorer
solubility of H3B·NMeH2 in C6H5F. Neither trimethylborazine, III, nor signals assignable
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 53
1
2
3
4
5
to free H2B=NMeH, were observed during the reaction using 11B NMR spectroscopy
when interrogated by regular sampling of the catalysis mixture.
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
H
H
H
MeH
N
Me
H
B
Me
H
cat. = 1
N
B
N
B
NMeH2
B
III
H3B
B
N
n
– H2
H2
N
Me
H
Me
both observed at 0.4 mol% (~ 10:1)
no polyaminoborane observed at 20 mol%
Scheme 10. Redistribution reactions. Closed system: [1]
[H3B·NMeHBH2·NMeH2] 0.22 M. Open system; [1]
=
=
0.4 mol%,
20 mol%,
=
[H3B·NMeHBH2·NMeH2] = 0.11 M.
68
Addition of the linear diborazane H3B·NMeHBH2·NMeH2
to 1 (20 mol%) in a sealed
NMR tube resulted in the formation of N–trimethylborazine III, alongside unidentified
metal products. No significant amounts of polyaminoborane or cyclic triborazane
69
[MeHNBH2]3 were observed under these near-stoichiometric conditions. However, at
0.4 mol% of 1 significant amounts of polyaminoborane were observed (Mn = 15 400 g
mol-1, Mw = 27 800 g mol-1, PDI = 1.8), so that this is now the major species formed
(~90% by 11B NMR spectroscopy, Scheme 10). This process presumably occurs via
metal–promoted B–N bond cleavage, possibly via a Rh sigma amine–borane
intermediate,27,51 to give H2B=NMeH and H3B·NMeH2 which both proceed under the
appropriate conditions of substrate concentration to give polyaminoborane and / or III.
The formation of only III at low substrate concentration is consistent with the
stoichiometric experiments using H3B·NMeH2 (i.e. Scheme 7). A very similar
redistribution of H3B·NMeHBH2·NMeH2 to afford poly(methylaminoborane) has been
reported using the [Ir(tBuPOCOPtBu)H2] catalyst,34 that is also suggested to operate via
B–N bond cleavage and an amino–borane intermediate, although this catalyst produces
22
ACS Paragon Plus Environment

Page 23 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
polyaminoborane of higher Mw (67, 400 g mol-1, PDI = 1.44) under the conditions used.
Ru(PNP)(H)(PMe3)–based systems have also been shown, by cyclohexene trapping
experiments, to promote redistribution of polyaminoborane. 23 Addition of the secondary
linear diborazane H3B·NMe2BH2·NMe2H to 1 (20 mol%) in a sealed NMR tube ultimately
formed [H2B=NMe2]2 after 24 hours. After 100 minutes of reaction 55% of the linear
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
diborazane has been consumed, with H2B=NMe2, [H2B=NMe2]2, boranediamine
70
HB(NMe2)2
and the amidodiborane (H2B)2(µ–H)(NMe2)34 all being observed in
significant amounts. These last two species suggest B–N bond cleavage is occurring to
form free NMe2H, as has been explored computationally and kinetically in thermal
rearrangements of linear diborazanes.34 That both primary and secondary linear
diborazanes react with complex 1 to ultimately form the final products of
dehydrocoupling shows that although they are not observed during catalysis, their
formation, either transiently metal–bound or free, cannot be discounted.
Effect of Solvent on Polymerization
Changing the solvent to THF produced polyaminoborane (catalyst = 1, 0.2 mol%) with
higher molecular weight (Mn = 52 200 versus 22 700 g mol-1) than for C6H5F solvent, but
now taking a significantly longer time to reach near completion (19 hr versus 2 hr, Table
1). This suggests THF slows the rate of dehydropolymerization, possibly by the
reversible formation of an adduct (cf 4), and this may also have a role to play in
attenuating any chain termination events if competitive with H2 binding71 (see below).
Alternatively, more of the catalyst could sit as the simple adduct species 4 leading to
fewer active metal sites, and thus longer polymer chains growing from the metal. THF
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 53
1
2
3
4
5
6
7
8
9
may also solvate the growing polymer better leading to longer chain growing from the
metal. Only a very small quantity of trimethylborazine, III, was observed (1–2%). THF
solvent might also result in a change in mechanism to one which involves hydride
donation to the metal to form a THF–stabilized borenium, i.e. [(NMeH2)(THF)BH2]+.32
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
Polymer growth kinetics and control over molecular weight using hydrogen.
A plot of number-averaged degree of polymerization, DPn [DPn = Mn / Mw(H2B=NMeH)]
versus conversion for the dehydrocoupling of H3B·NMeH2 using 1 (0.2 mol%, open
system) shows a relationship that is suggestive of a predominately chain growth
mechanism for the growing polymer (Scheme 11). Such a process has been proposed
previously for the [Ir(tBuPOCOPtBu)H2] catalyst system for which a modified chain
growth mechanism is invoked, in which slow dehydrogenation to form amino–borane is
followed by faster metal–mediated polymerization of this unsaturated fragment.15 This
suggestion is on the basis of the polymer conversion kinetics that show that high
molecular weight polymers are present at low (less than 40%) conversion; coupled with
the observation that higher catalyst loadings lead to higher molecular weight polymer. A
similar mechanism has been proposed for the dehydropolymerization of ammonia–
borane using bifunctional Ru–catalysts.23 Our polymer conversion kinetics suggest a
similar mechanism is operating, in that there is a high degree of polymerization at low
conversion (Mn = 30 800 g mol-1, PDI = 1.4 at 20% conversion; Mn = 25 300 g mol-1, PDI
= 1.6 at 100% conversion).72 However, in contrast to the [Ir(tBuPOCOPtBu)H2] systems,
when the catalyst loading is increased (i.e. x 5 the loading, 1 mol%) the polymer that
results is now of significantly lower molecular weight, but similar polydispersity, (Mn = 7
500 g mol-1, PDI = 1.5). This strongly suggests a metal-centered process, as initially
24
ACS Paragon Plus Environment

Page 25 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
proposed by Baker and Dixon for the catalytic dehydrogenation of ammonia–borane.37
11B{1H} NMR data for each conversion point show broadly similar peak profiles centred
around δ –5. In particular those at low conversions and high conversions are
qualitatively the same, suggesting the nature of the polymer in each is similar.
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
Addition of a further 500 equivalents of H3B·NMeH2 to a reaction post polymerization
resulted in further dehydropolymerization, to yield polymer with similar molecular weight
and polydispersity to before (Mn = 17 900 g mol-1, PDI = 1.8), over a similar timescale.
This shows that the catalyst remains active directly after catalysis has finished, but it is
not a living system and there must be some chain transfer/termination process occurring.
In a closed system (Youngs flask, ~ 30 cm3 volume, stirred) dehydropolymerization also
proceeds essentially to completion (Scheme 11, Table 1), but over a much longer
timescale than in an open system (24 hrs versus 2 hrs) The resulting isolated solid is
waxy in appearance, suggesting a lower Mn polymer compared with the free flowing
11
solid produced in an open system. A B{1H} NMR spectrum of this material shows a
broad, poorly resolved peak centred around δ –5 that also shows evidence for shorter
chain oligomeric species, cf. H3B·NMeHBH2·NMeH2,39 by an overlaid sharper signal that
11
becomes a broad triplet in the B NMR spectrum (Figure 1b). There is also a smaller
intensity signal ca. δ –18 in the region associated with BH3 groups,29 which is also
coincident with residual H3B·NMeH2. Analysis of this material by GPC showed that the
polymer produced under these conditions of exogenous hydrogen was considerably
shorter than that produced in an open system, Mn = 2 800 g mol-1, PDI = 1.8. This
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 53
1
2
3
4
5
6
7
8
9
demonstrates that hydrogen potentially acts as a modifier in catalysts, and we suggest it
acts as a chain transfer reagent, as in Ziegler Natta ethene polymerization where
hydrogen can used control polymer molecular weight.1,73
(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
(b)
H
H
H
1 0.2 mol%
H
H3B·NMeH2
H3B·NMeH2
B
B
N
n
0.44 M
– H2
N
n
0.44 M
– H2
Me
H
Me
H
Mn = 17 900
PDI = 1.8
Mn = 22 700
PDI = 1.9
open system
(c)
H
H
H
1 0.2 mol%
1 0.2 mol%
H
B
H3B·NMeH2
B
N
n
0.44 M
0.44 M
– H2
N
n
Me
H
Me
– H2
H
Mn = 2 800
PDI = 1.8
open system,
THF solvent
Mn = 52 200
PDI = 1.4
closed system
Scheme 11. (a) Degree of polymerization versus conversion: 0.2 mol% 1, 0.44 M
[H3B·NMeH2], open system. Each point is a separate experiment in C6H5F with varying
time, with the product precipitated with hexane. Degree of polymerization determined by
GPC. Polymer conversion measured by B{1H} NMR spectroscopy. Data points come
11
from three repeat analyses on the same sample, with the mean value and standard error
shown. (b) Addition of a further 500 equivalents of H3B·NMeH2 to 1 after catalysis. (c)
Control over molecular weight using H2 (C6H5F solvent) or THF solvent.
Probing free H2B=NMeH as an intermediate
26
ACS Paragon Plus Environment

Page 27 of 53
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
As discussed in the Introduction, the hydroboration of exogenous cyclohexene has
previously been shown act as a marker for the presence of free amino–borane
H2B=NMeH in dehydropolymerization reactions.22,34,37 In the presence of cyclohexene
using 50 mol% of 1 with H3B·NMeH2, the hydroborated product Cy2B=NMeH is
observed as the major boron–containing product, alongside III as the minor product
(Scheme 12). This suggests that at low substrate concentration free amino–borane is
generated, that has sufficient lifetime for reaction with cyclohexene. By contrast, at high
substrate concentrations (0.2 mol% 1) no hydroborated product is observed. Instead
polymer is produced, interestingly with a significantly higher molecular weight than
formed in the absence of cyclohexene (Mn = 38 600 g mol-1, PDI = 1.8). A small amount
of cyclohexane is also formed (~5% conversion). This suggests that under this
concentration regime free amino–borane is not produced in concentrations that allow for
hydroboration of cyclohexene. As 2 has been reported to reduce cyclohexene to
cyclohexane while becoming a Rh(I) species,50 the longer polymer chain length could be
a result of a lower concentration of the Rh(III) precatalyst (e.g. 7), that would
concomitantly result in fewer active sites for polymerization. Alternatively, cyclohexene
could simply attenuate chain transfer by being competitive with H2 for binding to the
active catalyst (vide infra).
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
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 53
1
2
3
4
5
6
H
Me
B
Me
1 50 mol%
N
N
B
H3B·NMeH2
Cy2B NMeH
major
B
0.044 M
– H2
N
Me
H
H
2.5 equivs.
7
8
minor
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
H
B
1 0.2 mol%
H
H3B·NMeH2
0.44 M
– H2
N n
H
2.5 equivs.
Me
~ 5 % conversion
Mn = 38 600
PDI = 1.8
Scheme 12. Cyclohexene trapping experiments. Solvent = C6H5F
Kinetic Studies on H3B·NMe2H: Open system
The low solubility of H3B·NMeH2, and resulting polyaminoborane, preclude detailed
solution–based kinetic investigations. We have thus conducted more detailed studies on
the catalytic process occurring using soluble H3B·NMe2H, which ultimately
dehydrogenates to give II. That both primary and secondary amine–borane systems
show very similar reaction profiles [induction period, same binding mode and reactivity
with the {Rh(Xantphos)H2}+ fragment] suggests that this approximation is a reasonable
one.
28
ACS Paragon Plus Environment

Page 29 of 53
Journal of the American Chemical Society
1
2
3
4
5
cat. = 1
Me2N BH2
H2B NMe2
H3B·NMe2H
– H2
6
7
(a)
8
9
0.018 M
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
(b)
0.288 M
Scheme 13. Time concentration plots for different [H3B·NMe2H] using 1 as a catalyst
(open system, 1,2–F2C6H4, [1] = 1.44 x 10–4 M). (a) [H3B·NMe2H] = 0.018 M; (b)
[H3B·NMe2H] = 0.288 M. Refer for Scheme 8a for [H3B·NMe2H] = 0.072 M. H3B·NMe2H;
▼ H3B·NMe2H, ■ H2B=NMe2, ▲ [H2B=NMe2]2, ● BH(NMe2)2.
Following the temporal evolution of the dehydrocoupling of H3B·NMe2H in an open
system (i.e. under a slow flow of Ar) under different substrate concentration regimes
[0.018 M to 0.288M74] while keeping [1] constant (1.44 x 10–4 M, i.e. 0.2 mol% at
[H3B·NMe2H] = 0.072 M) led to the concentration/time plots as exemplified in Scheme
13 (also Supporting Materials and Scheme 8a). All of these plots show very similar
induction periods (~ 400 s) and the formation of H2B=NMe2 as an intermediate. At higher
29
ACS Paragon Plus Environment