"spontaneous" ambient temperature dehydrocoupling of aromatic amine-boranes

Article abstract of DOI:10.1002/chem.201103241

The dehydrocoupling/dehydrogenation behavior of primary arylamine-borane adducts ArNH₂·BH₃ (3-a-c; Ar=a: Ph, b: p-MeOC ₆H₄, c: p-CF₃C₆H₄) has been studied in detail both in solut

Full text of DOI:10.1002/chem.201103241

FULL PAPER
DOI: 10.1002/chem.201103241
“Spontaneous” Ambient Temperature Dehydrocoupling of Aromatic
Amine–Boranes
Holger Helten,[a, b] Alasdair P. M. Robertson,[a] Anne Staubitz,[a, c] James R. Vance,[a]
Mairi F. Haddow,[a] and Ian Manners*[a]
Abstract: The dehydrocoupling/dehy-
drogenation behavior of primary aryl-
ACHTUNGTRENNUNGamine–borane adducts ArNH2·BH3
observed in the reaction of 3a with
BH3·THF and was characterized in situ.
The reaction of 3a with PhNH2 (2a)
was found to provide a new, convenient
method for the preparation of dianili-
noborane (PhNH)2BH (5a), which has
potential generality. This observation,
together with further studies of reac-
tions of 4a, 5a, and 7a,b, provided in-
sight into the mechanism of the cata-
lyst-free ambient temperature dehydro-
coupling of 3a–c in solution. For exam-
ple, the reaction of 4a with 5a yields
6a and 7a. It was found that borazines
(ArN-BH)3 (6a–c) are not simply
formed via dehydrogenation of cyclo-
triborazanes 7a–c in solution. The
transformation of 7a to 6a is slowly in-
duced by 5a and proceeds via regener-
ation of 3a. The adducts 3a–c also un-
derwent rapid dehydrocoupling in the
solid state at elevated temperatures
and even very slowly at ambient tem-
perature. From aniline–borane deriva-
tive 3c, the linear iminoborane oligo-
(3a–c; Ar=a: Ph, b: p-MeOC6H4, c: p-
CF3C6H4) has been studied in detail
both in solution at ambient tempera-
ture as well as in the solid state at am-
bient or elevated temperatures. The
presence of a metal catalyst was found
to be unnecessary for the release of H2.
From reactions of 3a,b in concentrated
solutions in THF at 228C over 24 h cy-
clotriborazanes (ArNH-BH2)3 (7a,b)
were isolated as THF adducts,
7a,b·THF, or solvent-free 7a, which
could not be obtained via heating of
3a–c in the melt. The m-(anilino)dibor-
mer
(p-CF3C6H4)NACTHUNGTERNNU[G BH-NH(p-
Keywords: amine-borane dehydro-
genation · amines · boranes · bor-
on–nitrogen compounds · dehydro-
coupling
CF3C6H4)]2 (11) was obtained. The
single-crystal X-ray structures of 3a–c,
5a, 7a, 7b·THF, and 11 are discussed.
ane [H2BACHTUNGTRENNUNG(m-PhNH)ACHTUNGTRENNUNG(m-H)BH2] (4a) was
Introduction
ers,[9] over the last decade various transition-metal cata-
lysts[1,2,4,6–8,10] and also main group systems[11] have been de-
veloped that enable reactions to proceed at moderate tem-
peratures.
The dehydrogenation and/or dehydrocoupling of amine–
borane adducts,[1] R(R’)NH·BH3 (R, R’=H, alkyl or aryl),
has attracted considerable interest from the perspectives of
chemical hydrogen storage,[2,3] transfer hydrogenation of or-
ganic substrates,[4,5] and the preparation of new inorganic
oligomeric[6] or polymeric materials.[7,8] Since thermal dehy-
drogenation reactions generally exhibit high kinetic barri-
Upon dehydrogenation with heterogeneous[10a,b] or homo-
geneous[10l] metal catalysts, primary amine–boranes (I) react
via cyclotriborazanes (II) as intermediates to ultimately give
borazines (III) with release of two equivalents of dihydro-
gen overall (Scheme 1). The use of Brookhartꢀs dehydrogen-
ation catalyst [Ir(tBu POCOP)(H)2] (tBu4POCOP=k3-1,3-
4
(OPtBu2)2C6H3) (1)[12] with ammonia–borane leads to the
elimination of 1 equiv of H2 to give insoluble polyaminobor-
ane, [NH2-BH2]n (nꢀ20).[7b,c] Application of this catalyst[7b,c]
and various other[7c–e] to primary alkylamine–borane adducts
provides access to high molecular weight linear polyamino-
boranes, [RNH-BH2]n (IV). To date, homopolymers with
R=H, Me, and nBu as well as random copolymers with
R=H or Me and R=Me or nBu have been described.[7b,c]
In striking contrast to their aliphatic analogues, knowl-
edge of the chemistry of primary arylamine–boranes, and
the parent compound, aniline–borane, is very limited.[13] It
has been noted that adducts having aryl groups at nitrogen
are significantly less stable than aliphatic amine–bora-
nes.[13b–d,14] This has been attributed to i) the reduced donor
ability of the nitrogen lone pair of aryl amines, and thus, the
[a] Dr. H. Helten, A. P. M. Robertson, Prof. Dr. A. Staubitz, J. R. Vance,
Dr. M. F. Haddow, Prof. Dr. I. Manners
School of Chemistry, University of Bristol
Cantockꢀs Close, Bristol, BS8 1TS (UK)
Fax : (+44)117-929-0509
E-mail: ian.manners@bristol.ac.uk
[b] Dr. H. Helten
Current address:
Institute of Inorganic Chemistry
RWTH Aachen University
Landoltweg 1, 52056 Aachen (Germany)
[c] Prof. Dr. A. Staubitz
Current address:
Otto-Diels-Institute, University of Kiel
Otto-Hahn-Platz 3, 24118 Kiel (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103241.
ꢁ
relative weakness of the B N bond in their borane ad-
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4665

aniline derivative employed. This could be avoided by ap-
plying modified reaction conditions: for the synthesis of 3a
and 3c, the liquid amine was added to a suspension of
frozen BH3·SMe2 in hexanes at ꢁ788C. The reactions started
while warming the mixtures to ambient temperature when
the formation of a fine precipitate of 3a or 3c was observed.
Scheme 1. Dehydrocoupling of primary amine–borane adducts (I) (R: ali-
phatic substituents).
ducts,[13,14] and ii) the enhanced N–H acidity.[13c] N,N-Diiso-
propylaniline–borane loses borane above its melting point at
36–388C, so that it can serve as a convenient source of di-
borane.[14d] Diphenylamine–borane[14a] and tertiary aryla-
mine–boranes[14b–d] have been used as reducing and hydrobo-
ration reagents. It has been suggested that dissociation of
the adducts is operative in such reactions.[14b] In 2003, we re-
ported that aniline–borane underwent dehydrocoupling with
Scheme 2. Synthesis of aromatic amine–borane adducts 3a–c (reaction
conditions: 2a,c, hexanes, ꢁ788C; 2b, Et2O, ꢁ308C).
The advantage of this protocol was that the products were
virtually insoluble in hexanes, whereby the formation of fur-
ther products (see below) could be largely avoided, and 3a,c
could be isolated by filtration. In the case of 3b, this
method was not successful as the reaction of BH3·SMe2 with
the solid amine 2b (which was insoluble in hexanes) re-
mained incomplete. In this case, BH3·SMe2 was added to a
solution of 2b in diethyl ether at ꢁ308C. Upon warming to
ambient temperature 3b precipitated from the reaction mix-
ture and was separated by filtration. The products, 3a–c,
were isolated in good to moderate yields, and their purity
was examined by elemental analysis, which gave satisfactory
results. Their identification was based on EI mass spectro-
[RhACHTUNGTRENNUNG(m-Cl)ACHTUNGTRENNUNG(1,5-cod)]2 as a precatalyst to give N,N’,N’’-triphe-
nylborazine under heterogeneous catalytic conditions.[10b]
The intermediacy of N,N’,N’’-triphenylcyclotriborazane was
suggested on the basis of observations made by 11B NMR
spectroscopy. Very recently, Michalak, Kang, and co-workers
showed that a series PdII complexes were able to catalyze
the dehydrocoupling of ammonia–borane and a range of pri-
mary amine–borane adducts, including aniline–borane,
under homogeneous conditions with release of 2 equiv of
H2.[10n] The characterization of the B–N-containing products
was hampered by their poor solubility in common organic
AHCTUNGTRENNUNG
metry and 1H and 11B NMR spectra, the latter of which
showed characteristic quartet resonances for the BH3 groups
1
ꢁ
solvents. However, the products contained no B H bonds,
at about ꢁ16 ppm. The H NMR chemical shift of the N-H
and their constitution was suggested to be [RHNB]x, with
protons (in CDCl3) followed the order 3b (d=5.29) <3a
(d=5.33) <3c (d=5.55), which reflects the trend of increas-
ing acidity in moving from the more electron donating p-
C6H4OMe group via phenyl to the strongly electron-with-
drawing p-C6H4CF3 group. Furthermore, the chemical shift
of the N-H protons showed a strong influence of the solvent
employed, as exemplified by derivative 3a: in CDCl3 and
CD2Cl2 the resonance was detected at d=5.33, in [D8]THF
it was shifted to d=6.75. The molecular structures of 3a–c
were also determined by single-crystal X-ray diffractometry
(Figures 1, 2, and S4 in the Supporting Information).
ꢁ
ꢁ
two B B bonds and one B N bond formed for every two
H2 molecules lost. It should be noted that the bisborane
adduct of diaminobenzene[13c] could be a promising candi-
date for hydrogen storage as it contains about 6 wt% of ac-
cessible hydrogen.
Herein, we report on our studies of the behavior of ani-
line–borane and related primary arylamine–borane adducts
both in solution and in the bulk at ambient or elevated tem-
perature, with the initial aim to preparing [ArNH-BH2]n, in-
organic B–N analogues of (para-substituted) polystyrenes.
The formation and characterization of the dehydrocoupling/
dehydrogenation products is described.
ꢁ
The B N bond lengths of the three adducts in the solid
state are virtually identical (1.619(2) (3a), 1.618(2) (3b),
and 1.614(4) ꢂ (3c)), but significantly longer than that of
MeNH2·BH3 (1.587(3)[16a] or 1.5936(13) ꢂ[16b]). Note that the
ꢁ
Results
B N bonds of 3a–c are nearly perpendicular to the aryl
rings, with interplanar angles between the aryl rings and the
planes given by C1-N1-B1 of 76.7 (3a), 78.6 (3b), and 84.88
(3c), respectively. This conformation allows conjugation of
Synthesis and characterization of aromatic amine–borane
adducts 3a–c: Adducts 3a–c were synthesized by the reac-
tion of 2a–c with borane dimethyl sulfide complex
(Scheme 2). The preparative procedures described in the lit-
erature for 3a,[10b,13b–d,15] via reaction of the respective ani-
line derivative 2a–c with BH3·THF in THF or with
BH3·SMe2 in THF or CH2Cl2, yielded in our hands signifi-
cant amounts of byproducts; their ratio depended on the
ꢁ
the dative B N bond into the aromatic p-system, which
causes weakening of the B N bond. In the solid state the
molecules of 3a–c assemble via B-H···H-N interactions to
form a three-dimensional network, in which intermolecular
H···H distances are significantly shorter than the sum of the
van der Waals radii for two hydrogen atoms, 2.4 ꢂ (see Sup-
ꢁ
4666
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
107.6 and 124.38 (3a), 97.1, 95.6, 95.8, and 94.18 (3b), and
104.4, 92.4, 97.1, and 94.08 (3c), whereas the H···H-N angles
were 179.6 and 148.18 (3a), 152.6, 147.3, 163.1, and 146.48
(3b), and 152.3, 135.8, 165.9, and 141.58 (3c). This type of
linear N-H···H and bent B-H···H geometry is consistent with
other examples of dihydrogen bonds for crystallographically
characterized amine–boranes.[17]
Dehydrocoupling/dehydrogenation of 3a–c in solution: With
the initial aim to preparing [PhNH-BH2]n, an inorganic B–N
analogue of polystyrene, aniline–borane (3a) was treated
with a THF solution of Brookhartꢀs IrIII pincer complex 1
(0.6 mol%) at ꢁ208C (initial concentration of 3a: 5 molLꢁ1
;
Scheme 3).
Figure 1. a) Molecular structure and b) association of the molecules of 3a
via H…H interactions in the solid state with thermal ellipsoids at the
50% probability level. All hydrogen atoms bonded to carbon are omitted
for clarity.
Scheme 3. Dehydrocoupling of 3a–c in solution.
Upon warming to ambient temperature 3a dissolved with
vigorous bubbling. An 11B{1H} NMR spectrum recorded
after 100 min revealed that 88% (by signal integration[18]) of
3a had been consumed and a variety of products were
formed. Two major resonances were detected at d=ꢁ21.9
and 25.6, assigned to m-(anilino)diborane (4a) and dianilino-
borane (5a), respectively.[19] The spectrum showed further
resonances for BH3·THF (d=ꢁ1.5), N,N’,N’’-triphenylbora-
zine (6a) (d=32.2),[10b] a broad resonance centered at d=
ꢁ3.6 assigned to N,N’,N’’-triphenylcyclotriborazane (7a),[19]
and a resonance at d=ꢁ11.6 corresponding to an unidenti-
fied product. Additionally, a small signal at d ꢀ 36 (<1%)
was detected, which was tentatively assigned to monomeric
anilinoborane (8a) (cf., e.g., PhMeN=BH2: d =40;[19] iPr2N=
BH2: d=34.2[21] or 34.7[10b]); no evidence was obtained for
the formation of polyanilinoborane.
Surprisingly, when THF was added to 3a at ꢁ208C and
the mixture was then warmed to ambient temperature, the
solution thus formed (5m starting concentration of 3a)
showed vigorous gas evolution even in the absence of a cat-
alyst such as 1. After 100 min an 11B{1H} NMR spectrum
(Figure 3) showed, within the experimental error, an identi-
cal distribution of products to that observed when 1 was
present.
Figure 2. a) Molecular structure and b) association of the molecules of 3c
via H…H interactions in the solid state with thermal ellipsoids at the
50% probability level. All hydrogen atoms bonded to carbon are omitted
for clarity. The CF3 group is disordered; only the prevailing conformation
(90%) is shown.
porting Information, Table S1). Each molecule associates
via both its B-H and N-H hydrogen atoms with another
molecule and with two other molecules via either its B-H or
N-H hydrogen atoms. Unlike 3a, in 3b and 3c the N-H hy-
drogens interact with two B-H hydrogen atoms of the same
borane fragment. The H···H-B angles were found to be
Analysis of the mixture of the reaction in the presence of
1 after 100 min by 31P{1H} NMR spectroscopy revealed that
the dihydride complex 1 (d=204.9)[12] had completely been
transformed into complex
9 a
(Figure 4; d=171.3),[6a]
s-bound BH3 adduct of 1.[22] The constitution of 9 was
confirmed by the addition of an authentic sample, independ-
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4667

I. Manners et al.
starting adducts followed the order Ar=p-C6H4OMe
<C6H5 <p-C6H4CF3. 2) The rate increased significantly
with the initial concentration of adduct. For example, at the
highest dilution in THF (0.25m), no significant conversion of
3a was detected within 1 d, and less than 11% had reacted
after 2 d. However, in a 1 molar solution in THF after 1 d
79% of 3a was consumed, while 98% was consumed in a 5
molar solution over the same reaction time. 3) In different
solvents the rate of consumption of 3a followed the order
THF ! Et2O < CH2Cl2:[25,26] negligible conversion was ob-
served within 1 d in a 0.25 molar solution in THF, in 0.25
molar solutions in Et2O about 73%, and in CH2Cl2 about
98% of 3a was consumed after 1 d. 4) The signals of 6a–c
and 7a,b were only observed in the later stages of the reac-
tions and grew at the expense of those of 3a–c, 4a–c, and
5a–c; cyclotriborazane derivative 7c was not observed. 5)
Borazines 6a–c seemed to be the final reaction products,
though certain amounts of 5a–c also remained present to-
wards the later stages of the reactions.[27] For instance, in re-
actions of 3a (1m in THF, Et2O, or CH2Cl2) at longer reac-
tion times (>7 d) the 11B NMR signal of 7a decreased
again, and after three weeks significant amounts of 6a and
5a were identified as the major products, and 7a was no
longer detected (Supporting Information, Figures S13–S15).
From reactions of 3a,b in THF (10m) cyclotriborazanes
7a,b were isolated as THF adducts, 7a,b·THF, in 14 and
20% yield, respectively, by preciptitation of the reaction
mixtures into hexanes. In the case of the phenyl derivative,
from the remaining hexanes solution, borazine 6a[10b] was
obtained in 47% yield by recrystallization. Reaction of 3a
in CH2Cl2 afforded, after the same separation procedure,
solvent-free 7a in 17% isolated yield.
Figure 3. 11B{1H} NMR spectrum of the reaction of 3a (5m) in THF after
100 min.
ently prepared from
1 and
BH3·THF,[6a] suggesting the cat-
alyst had been deactivated
through the transfer of BH3
from 3a. Goldberg, Heinekey,
and co-workers have reported
that 9 was a poor catalyst for
the dehydrogenation of ammo-
Figure 4. s-Borane complex 9.
nia–borane.[6a] Hence, with aniline–borane the active catalyst
is poisoned by the amine–borane substrate itself.[23]
Therefore, we decided to study the reaction in the pres-
ence of catalyst 1 with derivative 3b, wherein the electron
donating OMe group in the para position at the phenyl ring
ꢁ
was expected to stabilize the B N bond. Though the reac-
tion was significantly slower, with 73% conversion after
100 min, analogous dehydrocoupling products 4b–8b[24] and
BH3·THF were formed (Table 1); complete formation of
complex 9 was observed as well. When 3c, as a representa-
tive example with a very different electronic structure, was
reacted in the same manner, about 97% of the starting
adduct was consumed within the same reaction time. In this
case, the products 4c–6c and BH3·THF were formed, but
neither the cyclotriborazane 7c nor the monomeric amino-
borane species 8c was observed.
Upon dissolving 7a,b·THF in CD2Cl2, 1H and 13C NMR
signals of 1 equiv of THF were observed, which were very
close in chemical shift to those of free THF. The 1H and
13C{1H} NMR spectra of 7a and 7a,b·THF displayed only
one set of resonances in the aromatic region, which points
to a single environment for the three aryl groups. The
proton NMR spectra showed in each case a single broad N-
H resonance centered at d=4.16 (7a), 4.28 (7a·THF) and
4.16 (7b·THF), respectively. The slight shift of this signal for
7a vs 7a·THF evidences that in the latter compound the
THF molecule interacts with, but is not tightly bound to, the
N-H proton in solution. The 1H NMR spectra further
showed two B-H resonances at d=2.82 and 2.36 (7a), 2.86
and 2.40 (7a·THF), and 2.65 and 2.28 (7b·THF), which
could not be detected without 11B broadband decoupling; in
the 11B NMR spectra the coupling with the B-bonded pro-
tons was not resolved. These data are consistent with the
formation of highly symmetric isomers of 7a,b.[28] Solvent-
free single crystals of 7a suitable for X-ray crystallography
(Figure 5) were attained by slow diffusion of hexanes into a
concentrated solution of 7a·THF in CH2Cl2, which demon-
strates that THF dissociates from 7a·THF in CH2Cl2 solu-
tion. Single-crystals of 7b·THF were obtained by diffusion
of hexanes into a concentrated THF solution of 7b·THF
(Figure 6).
Table 1. 11B NMR chemical shifts (d, in THF) of the species observed
during reactions shown in Scheme 3.
3
4
5
6
7
8
Ar=C6H5
Ar=p-C6H4OMe
Ar=p-C6H4CF3
ꢁ16.9
ꢁ17.6
ꢁ16.4
ꢁ21.9
ꢁ22.3
ꢁ21.6
25.6
25.7
25.9
32.2
31.4
32.3
ꢁ3.6
36.2
ꢁ3.9
35.5
[a]
[a]
[a] Not observed.
Reactions of 3a–c in THF (1m) and reactions of 3a with
three different starting concentrations (5, 1, and 0.25m) and
in three different solvents, THF, diethyl ether, and dichloro-
methane, were monitored by 11B{1H} NMR spectroscopy.[18]
The main results were: 1) The rate of conversion of the
4668
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
In the solid state 7a and 7b·THF adopt a chair conforma-
tion with all aryl substituents in equatorial positions (eee
isomers). The structures can be compared to that of the eee
isomer of N,N’,N’’-trimethylcyclotriborazane, which also
shows this type of conformation.[29] In 7a the B N bond
ꢁ
lengths are 1.591(5) (B1–N1) and 1.597(5) ꢂ (B1–N1i), and
ꢁ
the average B N bond length of 7b·THF is 1.594(9) ꢂ, thus
ꢁ
slightly longer than the average B N bond lengths of eee-
(MeNH-BH2)3 (1.565 ꢂ)[29] and (NH2-BH2)3 (1.576 ꢂ),[30]
but in the same range as that of (Me2N-BH2)3 (1.59 ꢂ).[31]
The boron and nitrogen atoms are in distorted tetrahedral
environments (7a: N-B-N 106.7(3)8, B-N-B 114.8(4)8;
7b·THF: N-B-Nav 106.5(5)8, B-N-Bav 114.4(5)8).
Interestingly, the molecules of crystallographically C3 sym-
metric 7a assemble via H···H interactions involving all axial
B-H and N-H hydrogen atoms to form 1D chains along the
crystallographic c axis.[32] The H···H distance is 2.081 ꢂ. If
ꢁ
ꢁ
normalized values for B H (1.21 ꢂ) and N H bonds
(1.03 ꢂ) as established by neutron diffraction are used to
ꢁ
minimize the errors associated with X H bond lengths ob-
tained from X-ray diffraction data the H···H distance is de-
termined as 1.83 ꢂ. Interestingly, both the H···H-N and
H···H-B moieties show almost linear arrangements with
angles of 165.68 for H···H-B and 168.18 for H···H-N. Com-
pound 7b·THF does not show this type of intermolecular in-
teractions. Here, each molecule of 7b is bound to one THF
molecule via O···H hydrogen bonds involving all three axial
N-H hydrogens (N–Oav 3.130(3) ꢂ). It is noteworthy that
polyaminoboranes also appear to bind THF. Samples of
poly(N-methylaminoborane) and random copolymers [NH2-
BH2]n-r-[MeNH-BH2]m, which were purified in the same
manner as 7a,b (i.e., by precipitation from THF into hex-
anes), contained about 10–15% of residual THF.[7b,c] Conse-
quently, 7a,b may serve as appropriate molecular model sys-
tems for these types of polymers.
Figure 5. a) Molecular structure of 7a in the solid state with thermal el-
lipsoids at the 50% probability level. All hydrogen atoms bonded to
carbon are omitted for clarity. b) Assembly of 7a in chains along c via
H…H interactions. The phenyl groups are truncated (only Cipso shown).
i
ꢁ
ꢁ
Selected bond lengths [ꢂ] and angles [8]: N1 B1 1.591(5), N1 B1
1.597(5), B1-N1-B1ii 114.8(4), N1-B1-N1i 106.7(3).
Reaction of aniline–borane (3a) with BH3·THF: We at-
tempted to confirm the constitution of 4a–c and 5a–c, the
products initially formed from 3a–c in solution, and investi-
gated pathways for their formation. As the three derivatives
3a–c showed qualitatively the same behavior at early stages
of their reaction in solution, we chose for our studies the
parent aniline–borane (3a) as a representative example.
The observation of BH3·THF in reactions in THF provid-
ꢁ
ed direct evidence for B N bond cleavage of 3a–c without
the action of a catalyst. This suggests that 4a–c and 5a–c
ꢁ
might be formed via reactions of 3a–c with its B N bond
cleavage products, BH3 (as BH3·THF in THF) and ArNH2,
respectively. Similarly to the preparation of m-(NH2)B2H5,
reported recently by Chen, Zhao, and Shore,[33] 3a was treat-
ed with a solution of BH3·THF in THF (Scheme 4). Within
2 d at ambient temperature the formation of 4a was almost
complete (95% by 11B{1H} signal integration; see Supporting
Information, Figure S21).
Figure 6. Molecular structure of 7b·THF in the solid state with thermal
ellipsoids at the 50% probability level. All hydrogen atoms bonded to
carbon are omitted for clarity. The asymmetric unit contains two inde-
pendent 7b·THF units; only one unit is shown. Selected bond lengths [ꢂ]
ꢁ
ꢁ
ꢁ
ꢁ
and angles [8]: N1 B2 1.591(10), N1 B3 1.601(9), N2 B1 1.604(9), N2
ꢁ
ꢁ
B2 1.589(8), N3 B1 1.594(10), N3 B3 1.584(8), B2-N1-B3 114.3(4), B1-
N2-B2 115.8(5), B1-N3-B3 113.1(5), N2-B1-N3 108.8(5), N1-B2-N2
105.7(6), N1-B3-N3 104.9(5), O7…N1 3.147(4), O7…N2 3.094(1), O7…N3
3.137(4).
By 11B NMR, the monomeric anilinoborane (8a) was
identified in the reaction mixture, which is a plausible inter-
mediate in the formation of 4a, as the latter could be
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4669

I. Manners et al.
2ꢁ
(DIPP)]2
(DIPP=2,6-iPr2C6H3),[11b] and lanthanide
ACHTUNGTRENNUNG(III)
complexes thereof have been presented.[38] The complexes
were prepared via the neutral diaminoborane species
HB[NHACTHUNTRGEN(UNG DIPP)]2, which was obtained by the reaction of
Scheme 4. Reaction of 3a with BH3·THF (in THF) or BH3·SMe2 (in
CD2Cl2).
(DIPP)NH2 with BH3·SMe2 (ratio 2:1) at higher tempera-
ture (608C) and in the presence of nBu2Mg as a catalyst.
Our method for the synthesis of HBACTHNUTRGNEN(UG NHPh)2 (5a) repre-
sents a new, convenient, and potentially general route to
such species that proceeds at ambient temperature without
the requirement for a catalyst.
formed via hydroboration of 8a with BH3. Additionally,
small amounts of 5a, 6a, and 7a were observed in the mix-
ture. Upon removal of the volatiles in vacuo their ratio in-
creased dramatically, and all attempts to separate and purify
4a failed.[34] Therefore, the reaction was carried out in
CD2Cl2 using BH3·SMe2 as the borane source (which gave a
virtually identical result), and the mixture was then analyzed
The constitution of 5a was unambiguously confirmed by
multinuclear NMR spectroscopy, mass spectrometry, and
single-crystal X-ray diffraction (Figure 7). Following
HB[NHACHTNUGRTNEUNG
(DIPP)]2,[11b] 5a is the second example of this class
1
by H NMR spectroscopy. This revealed a broad high field
resonance at d=0.38, which is characteristic of the bridging
hydrogen[30] of 4a. The signal assigned to the terminal B-H
1
protons was detected at d=1.83 (4H) with a JB,H coupling
constant of 133 Hz in magnitude. The 11B NMR spectrum of
4a appeared as a doublet (1JB,H =26 Hz) of triplets (1JB,H
=
Figure 7. Molecular structure of 5a in the solid state with thermal ellip-
soids at the 50% probability level. The asymmetric unit contains two in-
dependent molecules, of which one is disordered; only one molecule is
shown. All hydrogen atoms bonded to carbon are omitted for clarity. Se-
132 Hz), due to couplings with the bridging and terminal
protons, respectively.[35] By in situ EI-MS the molecular ion
of 4a was detected (m/z 119), though in very low abundance
(2%). A higher intensity was found for the signal for the
anilinoborane fragment 8a·+ (m/z 105, 44%).
ꢁ
ꢁ
lected bond lengths [ꢂ] and angles [8]: N1 B1 1.413(1), B1 N2 1.412(1);
N1-B1-N2 120.3(1).
Reaction of aniline–borane (3a) with PhNH2
(2a): The dia-
of compounds to be structurally characterized. It contains
two independent molecules in the asymmetric unit with sim-
ilar structural data. The geometric parameters of the N-B-N
nilinoborane species 5a–c are formally derived from the sto-
ichiometric combination of anilines 2a–c with their respec-
tive anilinoborane derivative 8a–c with release of H2; 8a–c
may be formed via dehydrogenation of 3a–c. The observa-
tion of 8a upon dissolution of aniline–borane (3a) prompted
us to attempt the synthesis of 5a[36] by treatment of 3a with
a concentrated solution of aniline (2a; 1 equiv) at ꢁ408C
(Scheme 5). During subsequent warming to ambient temper-
ature vigorous gas evolution was observed. This resulted in
a clean reaction with quantitative formation of 5a within
3 d. Upon evaporation of the solvent and washing the crude
product with hexanes pure 5a was isolated in 88% yield.
frameworks of 5a and HB[NH
the error margins. The average B N bond length for 5a is
1.414(2) ꢂ (cf. HB[NH(DIPP)]2: 1.415(2) and 1.414(2) ꢂ)
and the mean N-B-N angle is 120.7(1)8 (HB[NH(DIPP)]2:
ACHTUNGERTN(NUNG DIPP)]2 are identical within
ꢁ
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
121.9(1)8). Both the B and N centers are in trigonal-planar
environments (the average values for the sum of the angles
at B and N are 359.9 and 359.78, respectively), hence, they
can be regarded as sp2-hybridized. This allows for delocali-
zation of the p-lone pairs on both N atoms into the vacant
p-orbital on boron, which imparts some degree of double
ꢁ
bond character to the B N bonds. Compound 5a displays
E,E configuration in the solid state. This is in contrast to the
crystallographically characterized diastereomer of HB[NH-
AHCTUNGTERGNUN(N DIPP)]2, which showed E,Z configuration. For HB[NH-
AHCTUNGTERGNUN(N DIPP)]2 in solution at room temperature, two sets of
1H NMR resonances have been found, whereas for 5a we
detected only one set of N-H and phenyl resonances, which
points to either the presence of only one diastereomer or to
Scheme 5. Reaction of 3a with aniline (2a).
ꢁ
Boraamidinate (bam) ligands, [RB
N
fast rotation about the B N bonds in solution. Furthermore,
deprotonated form of diaminoboranes, have recently attract-
ed considerable attention with respect to their coordination
chemistry. Currently, their potential for the synthesis of
novel paramagnetic coordination polymers and the activity
of complexes containing this ligand system as catalysts in
alkene polymerization are under active investigation.[37]
Harder and co-workers have recently established the first
bam ligand with an H substituent in the backbone, HB[N-
unlike in the case of HB[NH
A
phenyl substituents of 5a adopt a largely coplanar arrange-
ment with the respective N-B-N unit, having an average
twist angle of only 8.48, thus pointing to extended p-conju-
gation.
Reactions of m-(PhNH)B2H5
ACHTUNGTERN(UNGN 4a) and HBACHTUNREGTG(NNUN PhNH)2 ACHTUNGTRENNUNG(5a): In
order to obtain more information about the processes that
4670
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
occur in the later stages of the ambient temperature cata-
lyst-free dehydrocoupling of 3a–c, we studied the reaction
between 4a and 5a. Furthermore, reactions of 4a and 5a
clude that 3a may have been formed from the reaction of
2a with an excess of BH3·THF left in solution from the
preparation of 4a, the reaction was repeated via generation
of 4a from 3a and BH3·SMe2 in CH2Cl2. First, all volatiles
were removed in vacuo, then, the residue was re-dissolved
in CH2Cl2, with analysis by 11B NMR spectroscopy revealing
that the mixture was free of BH3·SMe2. It contained about
48% of 4a besides the subsequent products 5a–7a and less
than 7% of unreacted 3a. Upon addition of PhNH2 at
ꢁ408C and subsequent warming to ambient temperature,
the formation of about 41% of 3a was observed. The consti-
tution of 3a was confirmed by addition of an authentic
sample.
ꢁ
with the products of B N bond cleavage of 3a in THF (i.e.,
PhNH2 and BH3·THF) were investigated.
Upon generation of m-(anilino)diborane (4a) from 3a and
BH3·THF as described above, pure 5a was added to the mix-
ture (Scheme 6). 11B{1H} NMR spectroscopic monitoring re-
vealed that the resonances of 6a and 7a increased at the ex-
Calculations on DFT niveau (B3LYP/def2-TZVPP,
COSMO (e=7.4257, THF)//RI-BP86/def2-SVP, COSMO
(THF)) revealed that the structure 10 is a local energy mini-
mum, and a viable pathway for its formation was found to
be the nucleophilic attack of nitrogen of 2a at a boron
Scheme 6. Reaction of 4a with 5a.
pense of those of 4a and 5a (see Supporting Information,
Figure S22). Over longer reaction times the signal of 6a fur-
ther increased while that of 7a decreased again. This sug-
gests that cyclotriborazane 7a is the primary product of the
combination of 4a and 5a; for its formation no further hy-
drogen has to be eliminated. Borazine 6a could then be
formed via dehydrogenation of 7a. However, as both 6a
and 7a were observed at an early stage of the reaction
(after 3 h), the presence of another pathway to 6a that does
not involve the formation of 7a as an intermedate cannot be
excluded.
Chen, Zhao, and Shore have recently demonstrated that
m-(NH2)B2H5 undergoes ring opening upon reaction with
NH3, with formation of the linear species NH3-BH2-NH2-
BH3.[33] Accordingly, it seemed plausible that the reaction of
4a with PhNH2 (2a) could yield the linear diborazane
PhNH2-BH2-NHPh-BH3 (10). Aniline (2a) was added to a
solution of 4a, freshly prepared from 3a and BH3·THF, at
08C (Scheme 7). After the mixture was stirred for 1 h at am-
bient temperature, an 11B NMR spectrum showed that about
49% (by signal integration) of 3a was formed. The mixture
further contained ca. 27% of 5a, 21% of 4a, 2% of 7a, and
1% of 8a; no evidence for the formation of 10 was obtained
(see Supporting Information, Figure S23). At longer reac-
tion times the signal of 7a increased, and the signal of bora-
zine derivative 6a was additionally observed. In order to ex-
ꢁ
center of 4a from the backside of the B H bond involving
¼
the bridging hydrogen (DG 298 = +92.4 kJmolꢁ1; see Sup-
porting Information, Table S2 and Scheme S1). However,
the most stable conformation (gauche) is 33.6 kJmolꢁ1
higher in energy than the reactants, 4a and 2a.[39] Conse-
quently, if 10 is formed it will either react further via 1,4-hy-
¼
drogen elimination to give two equivalents of 8a (DG
=
298
+129.7 kJmolꢁ1; DRG298 =ꢁ126.1 kJmolꢁ1), or it will react
back to 4a and 2a. Subsequently, the latter can attack 4a
ꢁ
from the backside of its B N bond, which results in transfer
of a BH3 fragment from 4a to 2a and formation of
¼
PhNH2·BH3 (3a) and PhNH=BH2 (8a) (DG
=
298
+129.4 kJmolꢁ1; DRG298 =ꢁ29.4 kJmolꢁ1). This explains the
observation of 3a in the reaction mixture. Computationally,
we found an alternative, stepwise pathway for the formation
of 3a and 8a from 4a and 2a in THF solution. A BH3 frag-
ment can first be abstracted from 4a by a THF solvent mol-
ecule, which is subsequently transferred to PhNH2. The acti-
¼
vation barrier for the first step of this reaction, DG
=
298
+95.8 kJmolꢁ1, is significantly lower than that for the direct
transfer of BH3 from 4a to 2a.
Cyclotriborazane 7a could be a product derived from 8a.
It should further be noted that the observation of significant
amounts of 4a following the reaction of 4a with 2a does not
necessarily mean that it had remained partially unreacted.
Compound 4a may also have been re-formed from 3a via
disproportionation of the latter into 4a and 5a (see above).
Furthermore, dianilinoborane (5a) could alternatively be
formed by the reaction of 8a with an excess of PhNH2.
Reaction of 5a with BH3·THF gave a mixture of 3a
(ca. 9% by 11B{1H} signal integration), 4a (ca. 44%), 5a, 6a,
8a (in total ca. 45%), and 7a (ca. 2%), (Scheme 8; Support-
ing Information, Figure S24).
Dehydrogenation of (PhNH-BH2)3 ACHTUNTRGNEUNG(7a): In order to exam-
ine if the formation of borazines 6a–c at a later stage in the
course of reactions of 3a–c in solution occurs via dehydro-
genation of cyclotriborazanes 7a–c, phenyl derivative 7a, as
a selected example, was stirred in THF at ambient tempera-
Scheme 7. Reaction of 4a with aniline (2a).
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4671

I. Manners et al.
after the melting process (Supporting Information, Figures
S44 and S45). The DSC trace of derivative 3c displayed an
exotherm at about 408C, that is, below the endotherm for
the melting process (Figure S46). In an attempt to detect in-
termediates during this transformation, we carried out reac-
tions of 3a–c slightly above their melting temperatures
(3a,c: 908C; 3b: 1008C) and, in the case of 3a and 3c, addi-
tionally at 608C (Scheme 10). After heating for 1.5 h in the
Scheme 8. Reaction of 5a with BH3·THF.
ture and monitored by 11B and 11B{1H} NMR spectroscopy
(Scheme 9).
Scheme 10. Dehydrocoupling of 3a–c at elevated temperatures (3a,c:
608C or 908C, 3b: 1008C).
melt 3a and 3b were completely converted into borazines
6a,b together with minor amounts of 5a,b (see Supporting
Information, Figures S31 and S35). However, when the reac-
tions were stopped after 1 min, upon dissolving the mixtures
in THF, 11B NMR spectroscopy showed besides 5a,b, 6a,b,
and small amounts of unreacted 3a,b certain amounts of
4a,b and 7a,b (4a: 20%; 7a: 2%; 4b: 11%; 7b: 11%; Fig-
ures S29 and S33). Already after 15 min in the melt 3a,b,
4a,b, and 7a,b were largely consumed (4a: 1%; 7a: <1%;
4b and 7b: <1%; Figures S30 and S34). The reaction of 3c
initially showed an analogous course: after 1 min at 908C
about 15% of 4c and in total about 80% of 5c and 6c was
formed (Figure S36). Additionally, the 11B{1H} NMR spec-
trum displayed a small signal at d=ꢁ3.7 (<1%), which is
tentatively assigned to cyclotriborazane derivative 7c. It is
noteworthy that significant amounts of BH3·THF were also
detected (Figures S36 and S37). At longer reaction times (>
30 min) the content of the reaction vessel became increas-
ingly dark, and some amount of a solid material was
formed, which was insoluble in common organic solvents.
An 11B NMR spectrum of the THF-soluble fraction of the
product mixture revealed that the ratios of 4c–7c and
BH3·THF were decreased again, and a complex mixture of
unidentified products was formed, giving rise to 11B NMR
resonances at d=0.9, 0.6, 0.3, ꢁ0.6, and ꢁ1.4 ppm that dis-
Scheme 9. Dehydrogenation of 7a.
Surprisingly, no reaction was observed after 40 days at
228C (see Supporting Information, Figure S25). This further
supports the hypothesis that there might be another path-
way from 3a–c (or 4a–c and 5a–c) to 6a–c, which does not
proceed via 7a–c as intermediates. The observation of the
subsequent disappearance of 7a,b initially formed during
catalyst-free reactions of 3a,b suggests that the dehydrogen-
ation of 7a,b may be caused by a species generated from
3a,b in solution. As dianilinoborane 5a is derived from
2 equiv of 3a via twofold H2 elimination, we suspected that
this species might be able to dehydrogenate 7a. Therefore,
cyclotriborazane 7a was stirred in THF in the presence of
5a (Scheme 9).[40] Interestingly, 11B{1H} NMR reaction moni-
toring revealed that within 16 days about 3% of aniline–
borane 3a was formed (Figure S26). At longer reaction
times the formation of 4a, 6a, and BH3·THF was observed
(Figure S27).[41] Pure 7a could be transformed into 6a by
thermolysis in the melt at 1608C, where quantitative conver-
sion was observed after 30 min (Scheme 9; Supporting Infor-
mation, Figure S28). At 908C in the solid state no conver-
sion of 7a occurred over 1.5 h.
1
played no H couplings (Figure S38). Analysis of a sample
that was heated for 1.5 h at 908C by 19F{1H} NMR spectros-
copy revealed that reactions had occurred at the CF3 group.
Compound 3c underwent fast dehydrocoupling even at
608C in the solid phase. Within 30 min about 95% of the
adduct was consumed to give predominantly 5c and 6c
(Figure S39). Solid aniline–borane 3a also showed some
degree of dehydrocoupling when heated at 608C; after
30 min about 16% of 3a had reacted (Figure S32).
Dehydrocoupling of 3a–c in the solid state: For comparative
purposes, reactions of 3a–c in the bulk were investigated. It
has been previously shown that aniline–borane 3a is quanti-
tatively transformed into borazine 6a in the melt at 1208C
within 1.5 h.[13d] Differential scanning calorimetry (DSC)
performed on 3a and 3b showed an exotherm immediately
All three arylamine–borane adducts studied, 3a–c, were
found to be perfectly stable when kept at ꢁ408C under an
4672
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
argon atmosphere. However, dehydrocoupling occurred in
the solid state at ambient temperature (Scheme 11). Within
four months under inert atmosphere 3a was converted into
Figure 8. Molecular structure of 11 in the solid state with thermal ellip-
soids at the 50% probability level. All hydrogen atoms bonded to carbon
ꢁ
are omitted for clarity. Selected bond lengths [ꢂ] and angles [8]: N2 B1
ꢁ
ꢁ
ꢁ
1.407(3), B1 N1 1.441(3), N1 B2 1.437(3), B2 N3 1.413(3); N2-B1-N1
122.3(2), B1-N1-B2 119.9(2), N1-B2-N3 121.2(2).
1.437(3) ꢂ) are somewhat longer than those of 5a. The ter-
minal aryl substituents occupy trans positions and have
small twist angles with the N2-B1-N1-B2-N3 plane (for Ar
at N2: 24.38, for Ar at N3: 4.18). The central aryl ring (at
N1) is out of the plane of the B-N chain (twist angle: 70.98).
Compound 11 is the first five-membered iminoborane oligo-
mer to be crystallographically characterized.
Scheme 11. Dehydrocoupling of 3a–c at ambient temperature in the solid
state.
Discussion
The observed trend in rate of dehydrocoupling/dehydrogen-
ation of aromatic amine–borane adducts 3a–c in solution in
the order Ar=p-C6H4OMe < C6H5 < p-C6H4CF3 follows
a mixture containing unreacted 3a (40%, by 11B{1H} signal
integration), 4a (4%), 5a, 6a (5a and 6a in total ca. 27%),
and 7a (29%; Supporting Information, Figure S40). After
one year 3a was completely consumed; the mixture con-
tained 5a, 6a (in total ca. 78%), and 7a (22%; Figure S41).
Derivative 3b dehydrocoupled significantly slower under
these conditions. After four months only about 3% of 7b
was formed (Figure S42). On the other hand, 3c was largely
consumed within one month in the solid state at ambient
temperature. An 11B{1H} NMR spectrum showed, besides
the resonances of 4c, 7c, and unreacted 3c (ca. 2%), over-
lapping signals centered at d=31.0 and 26.4 ppm (Figure
S43). While the former is assigned to borazine derivative 6c
the latter is the result of a superposition of the resonances
of 5c and linear-chain iminoborane oligomer (p-CF3C6H4)N-
ꢁ
the anticipated trends of decreasing stability of their B N
ꢁ
bonds and increasing acidity of their N H bonds. The obser-
vation that the rate of conversion of 3a–c increased signifi-
cantly with the initial concentration of adduct as well as the
observation that the reactions were considerably faster in
Et2O and CH2Cl2 than in THF suggest that the dehydrogen-
ation is an intermolecular[10b] process. Presumably the aggre-
gation of molecules via B-H···H-N interactions, as found in
the solid state for 3a–c, is, to some extent, retained in Et2O
and CH2Cl2 solution, but is largely diminished in the strong-
ly donating solvent THF. This would lead to a decrease in
rate of dehydrogenation in the latter solvent. Evidence sup-
porting this hypothesis was obtained by the observation of a
strong downfield shift of the N-H proton resonance of 3a
when changing the solvent from CDCl3 or CD2Cl2 to
[D8]THF. The observed formation of BH3·THF in THF solu-
ACHTUNGTRENNUNG
lution of the product mixture derived from 3c yielded a
small crop of crystals of 11 suitable for X-ray crystallogra-
phy (Figure 8). It should be noted that one derivative of this
ꢁ
tions of 3a–c provides direct evidence for B N bond cleav-
age. This can be ascribed to the weakness of this bond in
3a–c, as suggested by their molecular structures in the solid
state, which showed relatively long B–N distances. If the de-
hydrogenation catalyst [Ir(tBu4POCOP)(H)2] (1) is present
in solution, it is poisoned by BH3·THF—and accordingly, in-
directly by the adducts themselves—through formation of
complex 9, the s-bound BH3 adduct of 1.
class of compounds, PhNACTHNUTRGNE(UNG BH-NHPh)2, has been prepared
before via reaction of aniline with triethylamine–borane at
90–1008C.[36]
The all-trans configured B–N backbone of 11 is essentially
planar. The mean deviation from the least-squares plane
given by N2-B1-N1-B2-N3 is 0.105 ꢂ. As for 5a, compound
ꢁ
11 exhibits rather short B N bonds. These features point to
extended p-conjugation over the B-N chain. It is noteworthy
The m-(amino)diborane and diaminoborane species, 4a–c
and 5a–c, respectively, which are formed at an early stage of
the reactions of 3a–c in solution, are the products of com-
bined disproportionation and dehydrogenation of the ad-
ꢁ
that the terminal B N bonds (1.407(3) and 1.413(3) ꢂ) are
slightly shorter while the central B N bonds (1.441(3) and
ꢁ
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4673

I. Manners et al.
ducts. Independent preparations of 4a from 3a and
BH3·THF and of 5a from 3a and PhNH2 (2a) demonstrated
that 4a–c and 5a–c are formed in solutions of 3a–c from re-
actions of the adducts with the primary products of their
after complete consumption of 3a–c, though compounds
4a–c were no longer detected.
It is noteworthy that the release of more than one equiv
of dihydrogen from aromatic amine–borane adducts 3a–c
proceeds rapidly upon melting (or below in the case of 3c),
which was also evidenced by their DSC traces. This is in
marked contrast to the behavior of ammonia–borane[43] and
many simple alkylamine–borane adducts.[13d,16a,44] For exam-
ple, from MeNH2·BH3 the first equivalent of H2 is eliminat-
ed at 1008C with formation of cyclotriborazane (MeNH-
BH2)3,[13d,16a,44] which can be further dehydrogenated at
much higher temperature (ca. 1908C) to give the borazine
derivative (MeN-BH)3.[13d,16a] In the cases of 3a–c, cyclotri-
borazanes 7a–c were initially formed at 90–1008C or 608C
(3c), but they were quickly consumed without the tempera-
ture being increased. The observation that pure 7a showed
no evidence of dehydrogenation within 1.5 h at the tempera-
ture at which 3a was completely transformed into 6a and 5a
may be due to the fact that 7a has a higher melting point,
and the formation of a mobile phase may be a prerequisite
for dehydrogenation of 7a to proceed in a reasonable time
scale.
In the solid state 3a–c underwent slow dehydrocoupling
even at ambient temperature. Here, the trend in rate in the
order Ar=p-C6H4OMe < C6H5 < p-C6H4CF3 was particular-
ly pronounced. While only 3% of 7b was formed from 3b
within four months, 3a was converted into a mixture of 3a–
7a over the same period of time. After one year 3a was
completely consumed, and only 5a–7a were found in the
mixture. The persistence of 5a, combined with the absence
of 4a, reveals that some amount of B2H6 must have been
evolved over time. Derivative 3c almost completely convert-
ed within one month into a mixture of 4c–7c and the linear
oligoiminoborane 11, which also points to elimination of
borane in this case.
ꢁ
B N bond cleavage, BH3 and ArNH2. It seems likely that
monomeric anilinoboranes 8a–c are involved in these pro-
cesses, as 8a,b were observed during reactions of 3a,b in so-
lution.
The formation of borazines 6a–c and cyclotriborazanes
7a,b in the later stages of the catalyst-free dehydrocoupling
of 3a–c suggests that they are products derived from 4a–c
and 5a–c. This was supported by the reaction of pure 5a
with freshly generated 4a. Towards the end of this reaction
the ratio of 6a increased at the expense of that of 7a. This
suggests that 6a may be formed via dehydrogenation of 7a
and that the latter is the primary product of the combination
of 4a with 5a. However, significant amounts of 6a were
found even at very early stages of the reaction, hence, there
might be another pathway to 6a that does not proceed via
7a as an intermediate. Through thermolysis of 7a in the
melt it was shown in principle that the dehydrogenation of
7a can yield 6a. However, pure 7a did not undergo dehy-
drogenation in THF solution at 228C. During the reaction
of 7a in the presence of 5a the generation of 3a was initially
observed. In the further course of the reaction, the dehydro-
coupling products of the latter were formed, namely 4a, 5a,
and 6a, which suggests that the formation of borazines 6a–c
does not simply proceed via dehydrogenation of cyclotribor-
azanes 7a–c in solution. It therefore appears that the forma-
tion of 6a–c and 7a,b in the ambient temperature dehydro-
coupling of 3a–c occurs via alternative pathways, whereas
the one that leads to 7a,b is reversible.
In the very late stages of the reactions of 3a–c in solution
besides the final products, 6a–c, minor amounts of 5a–c re-
mained unreacted. On the other hand, compounds 4a–c
were completely consumed. As 4a–c and 5a–c are formed
together in 1:1 ratio by combined disproportionation and
dehydrogenation of 3a–c, and both may combine stoichio-
metrically to give 6a–c and 7a–c, the disappearance of 4a–c,
which have a higher boron content than 5a–c, suggests that
some amount of borane may have been released from the
mixtures at longer reaction times.
In the bulk, slightly above their melting points 3a,b, and
initially also 3c, showed a similar behavior as in solution (3c
even in the solid phase at 608C), though the reactions were
much faster. Within 1.5 h 3a and 3b were completely trans-
formed into borazines 6a,b together with minor amounts of
5a,b; prolonged thermolysis of derivative 3c resulted in the
formation of a complex mixture due to reactions at the CF3
group. After short reaction times besides 6a–c and 5a–c
compounds 4a–c and 7a–c were also observed. Further-
more, the 11B NMR spectra, measured upon dissolving the
reaction mixtures in THF, showed significant amounts of
BH3·THF, which is most clearly seen in mixtures derived
from 3c.[42] Presumably, some amount of borane (i.e., in the
form of B2H6) was released during the reactions in the bulk
as well. This explains why the samples still contained 5a–c
Conclusion
Primary arylamine–borane adducts, 3a–c, have been pre-
pared and structurally characterized. The molecules form a
complex network via intermolecular dihydrogen bonding in
the solid state. They underwent dehydrocoupling in solution
at ambient temperature without the action of a catalyst. The
rate of conversion followed the order Ar=p-C6H4OMe <
C6H5 < p-C6H4CF3, which corresponds to the anticipated
ꢁ
trends of decreasing stability of their B N bonds and in-
ꢁ
creasing acidity of their N H bonds. The observed solvent
and concentration dependence points to intermolecular de-
ꢁ
hydrogenation processes. The occurrence of B N bond
cleavage was evidenced by the observation of BH3·THF in
THF solutions of 3a–c. The primary dehydrocoupling prod-
ucts, m-(anilino)diboranes 4a–c and dianilinoboranes 5a–c,
may be formed via dehydrogenation of 3a–c (presumably
via the monomers 8a–c) and subsequent reaction with the
ꢁ
products of B N bond cleavage of the adducts, BH3·THF
and PhNH2, respectively. This was reinforced by independ-
4674
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
4-aminobenzotrifluoride were dried and degassed by inert-gas distillation
from CaH2, p-anisidine was recrystallized from diethyl ether, and borane
tetrahydrofuran complex solution was distilled under static vacuum at
ambient temperature prior to use; borane dimethyl sulfide complex was
used as received. Complexes 1[12] and 9[6a] were prepared according to
methods described in the literature. NMR spectra were collected on a
Jeol Lambda 300, a Jeol ECP 300, or a Jeol ECP 400 spectrometer.
Chemical shifts were referenced to residual protic impurities in the sol-
vent (1H) or the deuterio solvent itself (13C) and reported relative to ex-
ternal SiMe4 (1H, 13C), BF3·OEt2 (11B), CFCl3 (19F), or H3PO4 (31P) stand-
ards. Mass spectra were obtained with the use of a VG Analytical Auto-
Spec mass spectrometer employing electron ionisation (EI) using a 70 eV
electron impact ionisation source. Melting points (uncorrected) were ob-
tained using a Stuart SMP3 melting point apparatus in 0.5 mm (o.d.)
glass capillaries, which were sealed under argon. Elemental analysis was
performed with a Eurovector EA 3000 Elemental Analyser at the Uni-
versity of Bristol Microanalysis Laboratory. GPC chromatograms were
recorded on a Viscotek VE2001, using a flow rate of 1 mLminꢁ1 in THF
with 0.1 w/w% [nBu4N]Br, calibrated for polystyrene standards, with an
approximate injection concentration of 2 mgmLꢁ1. The columns used
were of grade GP5000HHR followed by GP2500HHR (Viscotek) at a con-
stant temperature of 308C. Differential scanning calorimetry (DSC) was
performed on a TA Instruments DSC Q100 apparatus coupled to a refri-
gerated cooling system (RCS90). Both heating and cooling scans were
performed at 108Cminꢁ1 under an N2 atmosphere with the samples con-
tained in hermetic aluminium pans.
ent preparations of 4a and 5a; the latter was structurally
characterized. In the further course of the catalyst-free am-
bient temperature dehydrocoupling of 3a–c, borazines 6a–c
and cyclotriborazanes 7a,b were formed through reactions
between 4a–c and 5a–c; derivative 7c was not observed,
here. Cyclotriborazanes 7a,b further transformed into bora-
zines 6a,b as the final products. However, 6a–c are not
simply the products of the dehydrogenation of 7a–c in solu-
tion. The slow transformation of 7a into 6a was found to be
induced by 5a and proceeded via intermediate re-generation
of 3a, which was finally converted into 6a. Consequently,
the formation of 7a,b in the ambient temperature dehydro-
coupling of 3a,b is the result of a reversible alternative path-
way. This unexpected result reveals that fundamental ques-
tions such as the pathway of the transformation of cyclic
borazane oligomers into borazines, in general, are still unre-
solved. Further investigations will be necessary to clarify
these issues.
Reactions of 3a,b in concentrated solutions provided con-
venient access to N,N’,N’’-triarylcyclotriborazanes 7a,b. This
methodology is preferred over heating of 3a,b in the melt, a
method that can generally be applied for the synthesis of
N,N’,N’’-trialkylcyclotriborazanes from the corresponding
primary alkylamine–borane adducts.[13d,44] In reactions of
3a,b slightly above their melting points (at 90–1008C) 7a,b
were formed initially, but rapidly transformed into borazines
6a,b as the major products. Derivative 3c underwent rapid
dehydrocoupling even at 608C in the solid state. Cyclotri-
borazanes 7a,b are appropriate model systems for polyami-
noboranes and tend to bind THF in the solid state; the crys-
tal structure of 7b·THF showed that the THF solvent mole-
cules are bound via N-H···O interactions.
X-ray structural characterization: X-ray diffraction experiments on 3a,
7a, 7b·THF and 5a were carried out at 100 K on a Bruker APEX II dif-
fractometer using MoKa radiation (l=0.71073 ꢂ). X-ray diffraction ex-
periments on 3b, 3c, and 11 were carried out at 100 K on a Bruker Mi-
crostar rotating anode diffractometer using CuKa radiation (l=
1.54178 ꢂ). Data collections were performed using a CCD area detector
from a single crystal mounted on a glass fiber. Intensities were integrat-
AHCTUNGTRENNUNG
ed[46a] from several series of exposures measuring 0.58 in w or f. Absorp-
tion corrections were based on equivalent reflections using SADABS or
TWINABS.[46b] The structures were solved using SHELXS and refined
2
against all Fo data with hydrogen atoms attached to carbon atoms riding
in calculated positions using SHELXL.[46c] The position and thermal pa-
rameters on hydrogen atoms on boron and nitrogen were allowed to
refine freely, except for 7b·THF, where a riding model with fixed lengths
was used. Crystal structure and refinement data are given in Table 2.
Crystals of 7b·THF was a non-merohedral twin containing two domains,
respectively. These were indexed independently and integrated simulta-
neously using the APEX II software. Refinement proceeded smoothly to
give the structures shown to give ratios of 0.62:0.38, respectively for the
different domains. For 7a and 7b·THF, Friedel pairs were merged, so the
absolute handedness of the structures is unknown and has been assigned
arbitrarily. In 5a, one of the molecules is disordered over an inversion
centre, but resolved well into two positions with only a few restraints. In
7b·THF, two of the para-methoxy phenyl groups (on independent mole-
cules) are disordered, in a ratio of 0.78:0.22 (the disorder is related).
Some restraints and constraints were used to ensure a smooth refine-
ment.
Aromatic amine–borane adducts 3a–c underwent dehy-
drocoupling in the solid state even at ambient temperature.
Here, the trend in rate in the order Ar=p-C6H4OMe <
C6H5 < p-C6H4CF3 was particularly pronounced. While 3b
showed only slow formation of 7b, and 3a took one year to
be consumed, derivative 3c dehydrocoupled almost com-
pletely within 1 month to afford besides 4c–7c the linear
iminoborane oligomer 11, which is to date the longest-chain
derivative of this class of compounds to be crystallographi-
cally characterized.
CCDC-850031 (3a), CCDC-850032 (3b), CCDC-850033 (3c), CCDC-
850035 (7a), CCDC-850036 (7b·THF), CCDC-850034 (5a), and CCDC-
850030 (11) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.pccdc.cam.ac.uk/data_request/cif.
Experimental Section
General procedures and materials: All manipulations were performed
under an atmosphere of dry nitrogen using standard Schlenk techniques
or in a MBraun glovebox filled with dry argon. Solvents (tetrahydrofur-
an, diethyl ether, dichloromethane, and hexanes) were dried and de-
gassed by means of a Grubbs-type solvent purification system immediate-
ly prior to use.[45] Diethyl ether was further dried and degassed at reflux
over Na and freshly distilled prior to use. Deuterated CDCl3 and CD2Cl2
for NMR spectroscopy were purchased from the Cambridge Isotope Lab-
oratories and dried and degassed by inert-gas distillation from CaH2;
[D8]THF was purchased from Aldrich and used as received. Aniline, p-
anisidine, borane dimethyl sulfide complex, and borane tetrahydrofuran
complex (1m solution in THF) were purchased from Aldrich. 4-Amino-
benzotrifluoride was purchased from Apollo Scientific Ltd. Aniline and
General procedure for the synthesis of PhNH2·BH3
ACHTUNGTRENUN(NG 3a) and (p-
CF3C6H4)NH2·BH3 (3c): To a suspension of frozen BH3·SMe2 (5.0 mL,
AHCTUNGTRENNUNG
52.72 mmol) in hexanes (50 mL) was added the appropriate aniline deriv-
ative (2a: 4.5 mL, 49.38 mmol; 2c: 6.3 mL, 50.17 mmol) via syringe at
ꢁ788C. After warming the reaction mixture to ambient temperature the
white solid product was separated by filtration using a filter cannula and
subsequently washed with hexanes (5ꢃ40 mL) and dried in vacuo.
For PhNH2·BH3 (3a): Yield: 4.800 g (44.88 mmol, 91%); m.p. 708C
(decomp.); 1H NMR (300.53 MHz, CDCl3, 294.25 K): d=7.42 (m, 2H;
1
CHPh), 7.30 (m, 3H; CHPh), 5.33 (br, 2H; NH2), 1.89 ppm (brq, JBH
=
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4675

I. Manners et al.
Table 2. Crystal structure and refinement data for 3a–c, 7a, 7b·THF, 5a, and 11.
3a 3b 3c
colorless block colorless lath colorless needle colorless rod
7a
7b THF
5a
11
color, habit
size [mm]
empirical formula
M
colorless plate
colorless block colorless block
0.25ꢃ0.08ꢃ0.08 0.25ꢃ0.05ꢃ0.01 0.22ꢃ0.01ꢃ0.01 0.44ꢃ0.08ꢃ0.08 0.25ꢃ0.18ꢃ0.03 0.50ꢃ0.49ꢃ0.45 0.22ꢃ0.21ꢃ0.17
C6H10BN
106.96
C7H12BNO
136.99
C7H9BF3N
174.96
C18H24B3N3
314.83
C25H38B3N3O4
477.01
C12H13BN2
196.05
C21H16B2F9N3
502.99
crystal system
space group
a [ꢂ]
b [ꢂ]
c [ꢂ]
monoclinic
P21/c
8.372 (2)
10.807 (3)
7.712 (2)
90.00
orthorhombic
Pbca
10.068 (4)
7.871 (3)
20.411 (8)
90.00
monoclinic
P21/c
11.764 (4)
8.983 (3)
8.088 (3)
90.00
trigonal
R3c
18.8586(3)
18.8586(3)
8.9992 (4)
90.00
monoclinic
Pc
17.8347(8)
12.9685(6)
11.6784(5)
90.00
monoclinic
P21/c
5.9829 (2)
17.7604(5)
15.0569(4)
90.00
monoclinic
P21/c
8.5949 (16)
12.557 (2)
20.293 (4)
90.00
a [8]
b [8]
113.102(13)
90.00
641.8 (3)
4
0.063
100
2.65, 27.48
0.995 to q
=27.488
9870/1464
90.00
90.00
1617.4 (11)
8
0.572
95.01 (2)
90.00
851.4 (5)
4
1.101
100
6.21, 66.61
0.965 to q
=66.618
4890/1452
90.00
120.00
2771.74(14)
6
0.065
97.117 (3)
90.00
2680.3 (2)
4
0.078
100
1.95, 27.49
0.981 to q
=27.498
7435/7435
93.7260(10)
90.00
1596.54(8)
6
0.072
100
2.29, 27.50
1.000 to q
=27.508
26394/3671
96.543 (5)
90.00
2175.8 (7)
4
1.284
100
4.15, 66.55
0.964 to q
=66.558
29646/3698
g [8]
V [ꢂꢁ3
]
Z
m [mmꢁ1
T [K]
qmin,max
]
100
100
7.47, 66.26
0.970 to q
=66.268
11488/1373
2.16, 27.48
1.000 to q
=27.488
28582/713
completeness
reflections: total/inde-
pendent
Rint
0.1057
0.0653
0.0716
0.0455
0.0000[a]
0.0240
0.0564
final R1 and wR2
0.0546, 0.1566
0.186, ꢁ0.261
1.107
0.0459, 0.1433
0.186, ꢁ0.206
1.125
0.0551, 0.1497
0.250, ꢁ0.241
1.365
0.0574, 0.1432
0.186, ꢁ0.238
1.132
0.0732, 0.2199
0.405, ꢁ0.236
1.182
0.0348, 0.0905
0.189, ꢁ0.205
1.223
0.0519, 0.1453
0.389, ꢁ0.562
1.535
largest peak, hole [eꢂꢁ3
]
1calcd [gcmꢁ3
]
[a] Non-merohedral twin.
110 Hz, 3H; BH3); 1H NMR (300.53 MHz, CD2Cl2, 293.75 K): d=7.42
tions of 3a,b and the formation of 4a,b–6a,b, BH3·THF, and 9 in the case
of the reaction of 3c. An authentic sample of complex 9 was added to
each mixture.
(m, 2H; CHPh), 7.30 (m, 3H; CHPh), 5.33 (br, 2H; NH2), 1.78 ppm (brq,
1
1JBH =107 Hz, 3H; BH3); H NMR (300.53 MHz, [D8]THF, 294.65 K): d=
7.42 (m, 2H; CHPh), 7.30 (m, 3H; CHPh), 6.75 (br, 2H; NH2), 1.72 ppm
(brq, 3H; BH3); 11B NMR (96.42 MHz, CDCl3, 294.65 K): d=ꢁ16.9 (q,
1JBH =99 Hz); MS (EI, 70 eV): m/z (%): 105 (35) [M+ꢁ2H], 93 (100)
[M+ꢁBH3], 78 (42) [PhH+]; elemental analysis calcd (%) for C6H10BN:
C 67.37, H 9.42, N 13.10; found: C 67.16, H 9.45, N 12.89.
Catalyst-free dehydrocoupling/dehydrogenation of 3a–c in solution: To
2.50 mmol of the respective solid aniline–borane derivative (3a: 0.267 g;
3b: 0.342 g; 3c: 0.437 g) was added the appropriate amount (0.5, 2.5, or
10 mL) of solvent (THF, Et2O, or CH2Cl2) at ambient temperature while
vigorous bubbling was observed. The reaction was monitored by 11B and
11B{1H} NMR spectroscopy, which revealed the formation of 4a,b–8a,b
and BH3·THF (if THF was the solvent) in the cases of reactions of 3a,b
and the formation of 4a,b–6a,b and BH3·THF (if THF was the solvent)
in the case of the reaction of 3c.
For (p-CF3C6H4)NH2·BH3 (3c): Yield: 6.852 g (39.16 mmol, 78%); m.p.
728C (decomp.); 1H NMR (300.53 MHz, CDCl3, 294.15 K): d=7.69 (d,
3
3JHH =8.4 Hz, 2H; CHAr), 7.41 (d, JHH =8.4 Hz, 2H; CHAr), 5.55 (br, 2H;
NH2), 1.91 ppm (brq, 1JBH =114 Hz, 3H; BH3); 11B NMR (96.42 MHz,
CDCl3, 294.35 K): d=ꢁ16.0 ppm (q, 1JBH =94 Hz); 19F{1H} NMR
(282.78 MHz, CDCl3, 294.55 K): d=ꢁ62.5; MS (EI, 70 eV): m/z (%): 173
(35) [M+ꢁ2H], 78 (100) [M+ꢁBH3], 146 (75) [CF3C6H5+]; elemental
analysis calcd (%) for C7H9BF3N: C 48.05, H 5.18, N 8.01; found: C
47.64, H 5.15, N 8.03.
Synthesis of (PhN-BH)3 ACTHNUGTRNEG(UN 6a) and (PhNH-BH2)3·THF (7a·THF): To solid
aniline–borane (3a) (2.674 g, 25.00 mmol) was added THF (2.5 mL) at
ambient temperature. After 10 min a clear solution was formed which
was vigorously bubbling. The mixture was stirred for 24 h at ambient
temperature while the formation of a white precipitate was observed.
Then THF (2 mL) was added and the mixture was precipitated into hex-
anes (180 mL). The overlying solution was transferred into another flask
using a filter cannula, and the remaining white solid was dried in vacuo
and identified as 7a·THF. Yield: 0.443 g (1.14 mmol, 14%); m.p. 1528C
Synthesis of (p-MeOC6H4)NH2·BH3 ACTHNUTRGNEUGN(3b): To a solution of 2b (1.149 g,
9.33 mmol) in diethyl ether (38 mL) was added BH3·SMe2 (0.95 mL,
10.02 mmol) at ꢁ308C. Upon warming to ambient temperature a white
precipitate was formed, which was separated by filtration using a filter
cannula, washed with hexanes (4ꢃ20 mL), and dried in vacuo. Yield:
0.878 g (6.41 mmol, 69%); m.p. 918C (decomp.); 1H NMR (300.53 MHz,
CDCl3, 294.65 K): d=7.24 (m, 2H; CHAr), 6.91 (m, 2H; CHAr), 5.29 (br,
ACTHNUTRGNEUNG
(decomp.); 1H{11B} NMR (300.53 MHz, CD2Cl2, 295.35 K): d=7.31–7.38
(m, 12H; ortho+meta-CHPh), 7.22 (m, 4H; para-CHPh), 4.28 (br, 3H;
NH), 3.73 (m, 4H, a-CH2-THF), 2.86 (br, 3H; BH), 2.40 (br, 3H; BH),
1.85 ppm (m, 4H, b-CH2-THF); 11B NMR (96.42 MHz, CD2Cl2,
297.55 K): d=ꢁ4.4 ppm (br); 13C{1H} NMR (100.63 MHz, CD2Cl2,
295.95 K): d=144.6 (s, ipso-CPh), 129.5 (s, ortho/meta-CPh), 126.5 (s, para-
2H; NH2), 3.81 (s, 3H; OCH3), 1.87 (brq, 1JBH =112 Hz, 3H; BH3);
1
11B NMR (96.42 MHz, CDCl3, 294.95 K): d=ꢁ17.2 ppm (q, JBH
=
99 Hz); MS (EI, 70 eV): m/z (%): 135 (60) [M+ꢁ2H], 123 (87) [M+
ꢁBH3], 108 (100) [MeOC6H5·+]; elemental analysis calcd (%) for
C7H12BNO: C 61.37, H 8.83, N 10.22; found: C 61.27, H 8.88, N 10.30.
C
Ph), 123.6 (brs, ortho/meta-CPh), 68.4 (s, a-C-THF), 26.2 ppm (s, b-C-
THF); MS (EI, 70 eV): m/z (%): 309 (9) [M+ꢁ6H], 196 (37)
[(PhNH)2BH+], 105 (79) [PhNHBH2+], 93 (95) [PhNH2+], 78 (100)
[PhH+]; elemental analysis calcd (%) for C22H32B3N3O: C 68.29, H 8.34,
N 10.86; found: C 67.44, H 8.27, N 10.76. From the remaining hexanes
solution within 2 months at ꢁ658C a white precipitate was formed. The
overlying solution was removed by a filter cannula, and the remaining
white solid was dried in vacuo and identified as 6a.[10b] Yield: 1.220 g
(3.95 mmol, 47%).
Dehydrocoupling/dehydrogenation of 3a–c in the presence of catalyst 1:
To 2.50 mmol of the respective solid aniline–borane derivative (3a:
0.267 g; 3b: 0.342 g; 3c: 0.437 g) was added a solution of 1 (9 mg,
0.015 mmol, 0.6 mol%) in THF (0.5 mL) at ꢁ208C. While warming to
ambient temperature vigorous bubbling was observed. The reaction was
monitored by 11B, 11B{1H}, and 31P{1H} NMR spectroscopy, which re-
vealed the formation of 4a,b–8a,b, BH3·THF, and 9 in the cases of reac-
4676
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
Synthesis of [(p-MeOC6H4)NH-BH2]3·THF (7b·THF): To solid 4-me-
thoxyaniline–borane (3b) (1.370 g, 10.00 mmol) was added THF (1 mL)
at ambient temperature. After 10 min a clear solution was formed which
was vigorously bubbling. The mixture was stirred for 24 h at ambient
temperature while the formation of a white precipitate was observed.
Then THF (17 mL) was added and the mixture was precipitated into hex-
anes (180 mL). The overlying solution was removed using a filter cannu-
la, and the remaining white solid was dried in vacuo and identified as
Reaction of m-(PhNH)B2H5
ACHTUNGTNER(UGNN 4a) with HBACHTUNRTGENN(GUN PhNH)2 ACHTNUGTREN(NNGU 5a): To solid 3a
(0.267 g, 2.50 mmol) solution of BH3·THF (2.50 mmol) in THF
a
(2.5 mL) was added at ambient temperature, while vigorous bubbling was
observed. After the mixture was stirred for 3 d at ambient temperature,
11B{1H} NMR spectroscopic monitoring revealed the formation of 4a (ca.
96% by signal integration). Then, solid dianilinoborane (5a) (0.489 g,
2.49 mmol) was added at ambient temperature. The reaction was moni-
tored by 11B and 11B{1H} NMR spectroscopy, which revealed the forma-
tion of 6a and 7a.
ACTHNUTRGNEUNG
7b·THF. Yield: 0.318 g (0.67 mmol, 20%); m.p. 1308C (decomp.); 1H{11B}
NMR (300.53 MHz, CD2Cl2, 299.25 K): d=7.21 (m, 6H; CHAr), 6.85 (m,
6H; CHAr), 4.16 (br, 3H; NH), 3.74 (m, 4H, a-CH2-THF), 2.65 (br, 3H;
BH), 2.28 (br, 3H; BH), 1.86 ppm (m, 4H, b-CH2-THF); 11B NMR
(96.42 MHz, CD2Cl2, 295.85 K): d=ꢁ4.5 (br); 13C{1H} NMR (75.57 MHz,
CD2Cl2, 297.65 K): d=158.0 (s, C4 Ar), 137.4 (s, C1 Ar), 124.4 (s, CHAr),
114.6 (brs, CHAr), 68.5 (s, a-C-THF), 26.2 ppm (s, b-C-THF); MS (EI,
70 eV): m/z (%): 399 (53) [M+ꢁ6H], 256 (34) [(p-MeOC6H4NH)2BH+],
135 (100) [p-MeOC6H4NHBH2+], 123 (78) [p-MeOC6H4NH2+], 120 (79)
Reaction of m-(PhNH)B2H5 ACTHNUTRGENN(GU 4a) with PhNH2 AHCTUNGTRNE(NUGN 2a)
Via the generation of 4a from 3a and BH3·THF: To solid 3a (0.267 g,
2.50 mmol) a solution of BH3·THF (2.50 mmol) in THF (2.5 mL) was
added at ambient temperature, while vigorous bubbling was observed.
After the mixture was stirred for 2d at ambient temperature, 11B{1H}
NMR spectroscopic monitoring revealed the formation of 4a (ca. 95%
by signal integration). Then, aniline 3a (230 mL, 2.52 mmol) was added at
08C. After 15 min the mixture was warmed to ambient temperature and
monitored by 11B and 11B{1H} NMR spectroscopy, which revealed the for-
mation of 3a, 5a, 7a and 8a.
+
[p-MeOC6H4NHBH2 ꢁCH3], 108 (93) [C6H5OMe+], 93 (48) [PhNH2+];
elemental analysis calcd (%) for C18H24B3N3: C 62.95, H 8.03, N 8.81;
found: C 63.10, H 8.05, N 8.79.
Via the generation of 4a from 3a and BH3·SMe2: To solid 3a (0.214 g,
Synthesis of (PhNH-BH2)3 ACHTNUTRGNE(NUG 7a): To solid aniline–borane 3a (2.674 g,
2.00 mmol)
a solution of BH3·SMe2 (190 mL, 2.00 mmol) in CH2Cl2
25.00 mmol) was added CH2Cl2 (2.5 mL) at ambient temperature to give
a suspension that was vigorously bubbling. The mixture was stirred for
24 h at ambient temperature. Then CH2Cl2 (15 mL) was added and the
mixture was precipitated into hexanes (200 mL). The overlying solution
was removed using a filter cannula. This procedure was repeated two fur-
ther times. The remaining white solid was dried in vacuo and identified
(3 mL) was added at ꢁ408C. While warming to ambient temperature vig-
orous bubbling was observed. After the mixture was stirred for 3 h at am-
bient temperature, all volatiles were removed in vacuo. Upon dissolving
the residue in CH2Cl2, 11B NMR spectroscopy revealed the formation of
4a (ca. 48% by signal integration). Then, aniline 3a (182 mL, 2.00 mmol)
was added at 08C. After 15 min the mixture was warmed to ambient tem-
perature and monitored by 11B and 11B{1H} NMR spectroscopy, which re-
vealed the formation of 3a, 5a, 7a and 8a. An authentic sample of 3a
was added to the mixture.
ACTHNUTRGNEUNG
as 7a. Yield: 0.436 g (1.38 mmol, 17%); m.p. 1378C (decomp.); 1H{11B}
NMR (300.53 MHz, CD2Cl2, 294.95 K): d=7.31–7.38 (m, 12H; ortho+
meta-CHPh), 7.21 (m, 4H; para-CHPh), 4.16 (br, 3H; NH), 2.82 (br, 3H;
BH), 2.36 ppm (br, 3H; BH); 11B NMR (96.42 MHz, CD2Cl2, 297.15 K):
d=ꢁ4.4 ppm (br); 13C{1H} NMR (75.57 MHz, CD2Cl2, 296.95 K): d=
144.5 (s, ipso-CPh), 129.5 (s, ortho/meta-CPh), 126.5 (s, para-CPh),
123.6 ppm (brs, ortho/meta-CPh); elemental analysis calcd (%) for
C18H24B3N3: C 68.67, H 7.68, N 13.35; found: C 68.49, H 7.70, N 13.21.
Reaction of HBACHTNUGRTENNUG(PhNH)2 ACHTUTGNREN(NGUN 5a) with BH3·THF: To solid 5a (0.588 g,
3.00 mmol) a solution of BH3·THF (3.00 mmol) in THF (3.0 mL) was
added at ambient temperature. The mixture was stirred at ambient tem-
perature and monitored by 11B and 11B{1H} NMR spectroscopy, which re-
vealed the formation of 3a, 4a, 6a, 7a and 8a.
Attempted synthesis of m-(PhNH)B2H5 ACTHNUTRGNEUNG(4a): To solid 3a (0.465 g,
Attempted dehydrogenation of (PhNH-BH2)3 ACTHNUTRGENUG(N 7a) in THF: Cyclotribor-
4.35 mmol) a solution of BH3·THF (5.00 mmol) in THF (5 mL) was
added at ambient temperature, while vigorous bubbling was observed.
After the mixture was stirred for 2 d at ambient temperature, 11B{1H}
NMR reaction monitoring revealed that it contained ca. 95% of 4a.
Then all volatiles were removed in vacuo. Upon re-dissolving the crude
product mixture in THF (2 mL), an 11B{1H} NMR showed the formation
of significant amounts of 5a, 6a, and 7a and a decrease of the ratio of 4a
to ca. 34%.
AHCTUNGTREGaNNUN zane 7a·THF (0.05 g, 0.13 mmol) was suspended in THF (0.5 mL) at
ambient temperature. The mixture was frequently subjected to 11B and
11B{1H} NMR spectroscopy, which revealed no reaction within one
month.
Reaction of (PhNH-BH2)3 ACTHNUTRGNEUGN(7a) in the presence of 5a: Cyclotriborazane
7a (0.063 g, 0.20 mmol) and 5a (0.006 g, 0.03 mmol) was suspended in
THF (1 mL) at ambient temperature. The mixture was frequently sub-
jected to 11B and 11B{1H} NMR spectroscopy, which revealed the forma-
tion of 3a, 4a, 6a, and BH3·THF.
Characterization of m-(PhNH)B2H5
ACHTUNGERTN(NUNG 4a): To solid 3a (0.107 g,
1.00 mmol) solution of BH3·SMe2 (95 mL, 1.00 mmol) in CD2Cl2
a
Dehydrogenation of (PhNH-BH2)3 ACTHNUTRGNE(UNG 7a) in the melt: Cyclotriborazane 7a
(1.5 mL) was added at ꢁ408C. While warming to ambient temperature
(0.02 g, 0.06 mmol) was heated in a Schlenk flask at 1608C under N2 flow
for 30 min. During the melting process vigorous bubbling was observed.
Upon cooling to ambient temperature the product mixture was dissolved
in THF and subjected to 11B and 11B{1H} NMR spectroscopy, which re-
vealed the formation of 6a.
vigorous bubbling was observed. After the mixture was stirred for 3 h at
1
ambient temperature, it was analyzed by H, 11B, and 11B{1H} NMR spec-
troscopy and EI-MS spectrometry. 1H NMR (300.53 MHz, CD2Cl2,
294.75 K): d=1.83 (brq, 1JBH =133 Hz, 4H; 2ꢃBH2), 0.38 ppm (br, 1H;
B-H-B); 11B NMR (96.42 MHz, CD2Cl2, 294.75 K): d=ꢁ21.6 ppm (td,
1JBH =132 and 26 Hz); MS (EI, 70 eV): m/z (%): 119 (2) [M+], 105 (44)
[M+ꢁBH3], 93 (100) [PhNH2+].
Attempted dehydrogenation of (PhNH-BH2)3 ACTHNUTRGENUG(N 7a) at 908C: Cyclotribor-
azane 7a (0.02 g, 0.06 mmol) was heated in a Schlenk flask at 908C under
N2 flow for 1.5 h. Upon cooling to ambient temperature the content of
the Schlenk flask was dissolved in THF and subjected to 11B and 11B{1H}
NMR spectroscopy, which revealed no reaction.
Synthesis of HBACHTUNGTRENNUNG(PhNH)2 ACHTUNGTERN(NUGN 5a): To solid 3a (1.100 g, 10.28 mmol) a solu-
tion of aniline (1 mL, 10.97 mmol) in THF (1 mL) was added at ꢁ408C.
While warming to ambient temperature vigorous bubbling was observed.
The mixture was stirred for 3 d at room temperature. Then all volatiles
were removed in vacuo. The remaining white solid was washed with hex-
anes (3ꢃ5 mL) and dried in vacuo. Yield: 1.770 g (9.03 mmol, 88%);
Dehydrocoupling of 3a–c in the melt: Aniline–borane derivatives 3a–c
(0.30 mmol, 0.032 g (3a), 0.041 g (3b), 0.052 g (3c)) were heated in a
Schlenk flask above their melting points (3a,c at 908C; 3b at 1008C)
under N2 flow for 1 min, 15 min, or 1.5 h. During the melting process vig-
orous bubbling was observed. In the case of 3c, upon prolonged heating
(> 30 min) the mixture turned yellow to brown and a brown solid materi-
al was formed, which was found to be insoluble in THF, CH2Cl2, or ace-
tone. Upon cooling to ambient temperature the product mixtures were
dissolved in THF (0.8 mL) and subjected to 11B and 11B{1H} NMR spec-
troscopy, which revealed the formation of 4a–c, 5a–c, 6a–c, 7a–c, and
BH3·THF after 1 or 15 min of heating, in the case of 3a,b, the formation
m.p. 1088C (decomp.); 1H
d=7.25 (m, 4H; meta-HPh), 7.04 (m, 4H; ortho-HPh), 6.94 (m, 2H; para-
Ph), 5.34 (br, 2H; NH), 5.02 ppm (t, 3JHH =8.0 Hz, 1H; BH); 11B NMR
{11B} NMR (300.53 MHz, CDCl3, 296.35 K):
ACHTUNGTRENNUNG
H
(96.42 MHz, CDCl3, 298.45 K): d=25.7 (d, 1JBH =101 Hz); 13C{1H} NMR
(75.57 MHz, CDCl3, 297.45 K): d=144.2 (s, ipso-CPh), 129.3 (s, meta-CPh),
121.1 (s, para-CPh), 117.8 (brs, ortho-CPh); MS (EI, 70 eV): m/z (%): 196
(16) [M+], 93 (100) [PhNH2+]; elemental analysis calcd (%) for
C12H13BN2: C 73.51, H 6.68, N 14.29; found: C 73.39, H 6.80, N 14.30.
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4677

I. Manners et al.
of 5a,b and 6a,b after 1.5 h, and in the case of 3c, the formation of 5c,
6c, and unidentified products with 11B resonances at d =0.9, 0.6, 0.3,
ꢁ0.6, and ꢁ1.4; 19F{1H} resonances were detected at d =ꢁ61.0, ꢁ61.1,
ꢁ62.3, ꢁ62.5, ꢁ62.7, ꢁ62.8, ꢁ62.9, ꢁ149.7, ꢁ150.3 to ꢁ150.9, ꢁ156.1, and
ꢁ156.2 ppm.
Dehydrocoupling of 3a,c at 608C: Aniline–borane derivatives 3a,c
(0.30 mmol, 0.032 g (3a), 0.052 g (3c)) were heated in a Schlenk flask at
608C under N2 flow for 30 min. Upon cooling to ambient temperature
the product mixtures were dissolved in THF (0.8 mL) and subjected to
11B and 11B{1H} NMR spectroscopy, which revealed the formation of 4a–
7a and BH3·THF in the mixture derived from 3a and the formation of
4c–6c and BH3·THF in the mixture derived from 3c.
2009, 28, 5493–5504; d) M. E. Sloan, A. Staubitz, K. Lee, I. Man-
ners, Eur. J. Org. Chem. 2011, 672–675.
[5] X. Yang, T. Fox, H. Berke, Chem. Commun. 2011, 47, 2053–2055.
[6] a) M. C. Denney, V. Pons, T. J. Hebden, D. M. Heinekey, K. I. Gold-
berg, J. Am. Chem. Soc. 2006, 128, 12048–12049; b) B. L. Dietrich,
K. I. Goldberg, D. M. Heinekey, T. Autrey, J. C. Linehan, Inorg.
Chem. 2008, 47, 8583–8585; c) H. C. Johnson, A. P. M. Robertson,
A. B. Chaplin, L. J. Sewell, A. L. Thompson, M. F. Haddow, I. Man-
ners, A. S. Weller, J. Am. Chem. Soc. 2011, 133, 11076–11079.
[7] a) T. J. Clark, K. Lee, I. Manners, Chem. Eur. J. 2006, 12, 8634–
8648; b) A. Staubitz, A. P. Soto, I. Manners, Angew. Chem. 2008,
120, 6308–6311; Angew. Chem. Int. Ed. 2008, 47, 6212–6215; c) A.
Staubitz, M. E. Sloan, A. P. M. Robertson, A. Friedrich, S. Schneid-
er, P. J. Gates, J. Schmedt auf der Gꢄnne, I. Manners, J. Am. Chem.
Soc. 2010, 132, 13332–13345; d) R. Dallanegra, A. P. M. Robertson,
A. B. Chaplin, I. Manners, A. S. Weller, Chem. Commun. 2011, 47,
3763–3765; e) J. R. Vance, A. P. M. Robertson, K. Lee, I. Manners,
Chem. Eur. J. 2011, 17, 4099–4103; f) T. Kakizawa, Y. Kawano, K.
Naganeyama, M. Shimoi, Chem. Lett. 2011, 40, 171–173; g) A. P. M.
Robertson, E. M. Leitao, I. Manners, J. Am. Chem. Soc. 2011, 133,
19322–19325.
Dehydrocoupling of 3a–c in the solid state at ambient temperature: Ani-
line–borane derivatives 3a–c (1.00 mmol, 0.107 g (3a), 0.137 g (3b),
0.175 g (3c)) were stored in a vial under inert atmosphere in a glovebox.
Frequently, a small amount (ca. 15 mg) of the product mixture was dis-
solved in THF (0.5 mL) and subjected to 11B and 11B{1H} NMR spectros-
copy, which revealed the formation of 4a–7a and 7b from 3a and 3b, re-
spectively, after 4 months, the formation of 5a, 6a and 7a from 3a after 1
year, and the formation of 4c–7c and 11 from 3c after one month. Upon
slow diffusion of hexanes into a solution of the product mixture derived
from 3c, a small crop of colorless crystals of compound 11 was obtained.
[8] V. Pons, R. T. Baker, Angew. Chem. 2008, 120, 9742–9744; Angew.
Chem. Int. Ed. 2008, 47, 9600–9602.
Computational methods: DFT calculations were carried out with the
TURBOMOLE V5.9.1 program package.[47a] Optimizations[47b] were done
with the local Slater-Dirac exchange[48] and the correlation energy density
functional (No.V) by Vosko, Wilk, and Nusair (VWN (V))[49] together
with Beckeꢀs gradient corrected exchange functional B88[50] in combina-
tion with the gradient corrected correlation functional by Perdew
(P86)[51] within the RI approximation[52] and the valence-double-z basis
set def2-SVP.[53] During the optimizations the influence of the polar sol-
vent was taken into account by employing the COSMO approach[54] with
e=7.4257. For cavity construction the atomic radii of Bondi,[55] obtained
from cystallographic data, were used. Transition states were located by
[9] K. Niedenzu, J. W. Dawson, Boron-Nitrogen Compounds, Academic
Press Inc., New York, 1965.
[10] a) C. A. Jaska, K. Temple, A. J. Lough, I. Manners, Chem. Commun.
2001, 962–963; b) C. A. Jaska, K. Temple, A. J. Lough, I. Manners,
J. Am. Chem. Soc. 2003, 125, 9424–9434; c) R. J. Keaton, J. M. Blac-
quiere, R. T. Baker, J. Am. Chem. Soc. 2007, 129, 1844–1845;
d) J. L. Fulton, J. C. Linehan, T. Autrey, M. Balasubramanian, Y.
Chen, N. K. Szymczak, J. Am. Chem. Soc. 2007, 129, 11936–11949;
e) D. Pun, E. Lobkovsky, P. J. Chirik, Chem. Commun. 2007, 3297–
3299; f) T. M. Douglas, A. B. Chaplin, A. S. Weller, J. Am. Chem.
Soc. 2008, 130, 14432–14433; g) N. Blaquiere, S. Diallo-Garcia, S. I.
Gorelsky, D. A. Black, K. Fagnou, J. Am. Chem. Soc. 2008, 130,
14034–14035; h) M. E. Sloan, T. J. Clark, I. Manners, Inorg. Chem.
2009, 48, 2429–2435; i) T. M. Douglas, A. B. Chaplin, A. S. Weller,
X. Yang, M. B. Hall, J. Am. Chem. Soc. 2009, 131, 15440–15456;
j) M. Kꢅß, A. Friedrich, M. Drees, S. Schneider, Angew. Chem.
2009, 121, 922–924; Angew. Chem. Int. Ed. 2009, 48, 905–907; k) A.
Friedrich, M. Drees, S. Schneider, Chem. Eur. J. 2009, 15, 10339–
10342; l) Y. Kawano, M. Uruichi, M. Shimoi, S. Taki, T. Kawaguchi,
T. Kakizawa, H. Ogino, J. Am. Chem. Soc. 2009, 131, 14946–14957;
m) M. E. Sloan, A. Staubitz, T. J. Clark, C. A. Russell, G. C. Lloyd-
Jones, I. Manners, J. Am. Chem. Soc. 2010, 132, 3831–3841; n) S.-K.
Kim, W.-S. Han, T.-J. Kim, T.-Y. Kim, S. W. Nam, M. Mitoraj, Ł.
using
a
TRIM algorithm.[56] Excellent initial guesses were obtained
through relaxed surface scans along the major reaction coordinates. All
stationary points were characterized by numerical vibrational frequencies
calculations.[57] Single point calculations were carried out using the Three
Parameter Hybrid Functional Becke3[58] (B3) in combination with the
gradient corrected correlation functional by Lee, Yang and Parr[59] (LYP)
using the valence-triple-z basis set def2-TZVP.[60] The COSMO approach
was employed with the same parameters as used for optimizations. Zero
point corrections and thermal corrections to free energies were adopted
from frequencies calculations on the optimization level (RI-BP86/def2-
SV(P), COSMO). The practicability of this approach has been demon-
strated before.[61]
´
Piekos, A. Michalak, S.-J. Hwang, S. O. Kang, J. Am. Chem. Soc.
2010, 132, 9954–9955; o) G. Alcaraz, L. Vendier, E. Clot, S. Sabo-
Etienne, Angew. Chem. 2010, 122, 930–932; Angew. Chem. Int. Ed.
2010, 49, 7170; p) M. S. Hill, G. Kociok-Kçhn, T. P. Robinson,
Chem. Commun. 2010, 46, 7587–7589; q) C. Y. Tang, A. L. Thomp-
son, S. Aldridge, J. Am. Chem. Soc. 2010, 132, 10578–10591; r) C. J.
Stevens, R. Dallanegra, A. B. Chaplin, A. S. Weller, S. A. Macgre-
gor, B. Ward, D. McKay, G. Alcaraz, S. Sabo-Etienne, Chem. Eur. J.
2011, 17, 3011–3020; s) M. Vogt, B. de Bruin, H. Berke, M. Trinca-
do, H. Grꢄtzmacher, Chem. Sci. 2011, 2, 723–727; t) A. M. Chap-
man, M. F. Haddow, D. F. Wass, J. Am. Chem. Soc. 2011, 133, 8826–
8829.
Acknowledgements
H.H. thanks the Deutsche Forschungsgemeinschaft (DFG) for a research
fellowship. H.H., A.P.M.R.,J.R.V., and I. M. thank the EPSRC for sup-
port. We thank Dr. Sanjib K. Patra for DSC measurements.
[1] A. Staubitz, A. P. M. Robertson, M. E. Sloan, I. Manners, Chem.
Rev. 2010, 110, 4023–4078.
[11] a) J. Spielmann, G. Jansen, H. Bandmann, S. Harder, Angew. Chem.
2008, 120, 6386–6391; Angew. Chem. Int. Ed. 2008, 47, 6290–6295;
b) J. Spielmann, M. Bolte, S. Harder, Chem. Commun. 2009, 6934–
6936; c) D. J. Liptrot, M. S. Hill, M. F. Mahon, D. J. MacDougall,
Chem. Eur. J. 2010, 16, 8508–8515; d) A. J. M. Miller, J. E. Bercaw,
Chem. Commun. 2010, 46, 1709–1711; e) G. R. Whittell, E. I. Bal-
mond, A. P. M. Robertson, S. K. Patra, M. F. Haddow, I. Manners,
Eur. J. Inorg. Chem. 2010, 3967–3975; f) H. J. Cowley, M. S. Holt,
R. L. Melen, J. M. Rawson, D. S. Wright, Chem. Commun. 2011, 47,
2682–2684; g) M. M. Hansmann, R. L. Melen, D. S. Wright, Chem.
Sci. 2011, 2, 1554–1559; h) P. Bellham, M. S. Hill, D. J. Liptrot, D. J.
[2] a) T. B. Marder, Angew. Chem. 2007, 119, 8262–8264; Angew.
Chem. Int. Ed. 2007, 46, 8116–8118; b) F. H. Stephens, V. Pons,
R. T. Baker, Dalton Trans. 2007, 2613–2626; c) A. Staubitz, A. P. M.
Robertson, I. Manners, Chem. Rev. 2010, 110, 4079–4124.
[3] a) A. D. Sutton, A. K. Burrell, D. A. Dixon, E. B. Garner III, J. C.
Gordon, T. Nakagawa, K. C. Ott, J. P. Robinson, M. Vasiliu, Science
2011, 331, 1426–1429; b) W. Luo, L. N. Zakharov, S.-Y. Liu, J. Am.
Chem. Soc. 2011, 133, 13006–13009.
[4] a) C. A. Jaska, I. Manners, J. Am. Chem. Soc. 2004, 126, 2698–2699;
b) Y. Jiang, H. Berke, Chem. Commun. 2007, 3571–3573; c) Y.
Jiang, O. Blacque, T. Fox, C. M. Frech, H. Berke, Organometallics
4678
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680

Dehydrocoupling of Aromatic Amine–Boranes
FULL PAPER
MacDougall, M. F. Mahon, Chem. Commun. 2011, 47, 9060–9062;
i) M. S. Hill, M. Hodgson, D. J. Liptrot, M. F. Mahon, Dalton Trans.
2011, 40, 7783–7790.
cannot be excluded that longer-chain iminoborane oligomers were
formed, in addition, which might resonate in the same region. Note
that the reaction of 3c in the solid state yielded the oligomer p-
[12] I. Gçttker-Schnetmann, P. S. White, M. Brookhart, Organometallics
2004, 23, 1766–1776.
CF3C6H4NHACHTUNGETRN[NUG BH-NH(p-CF3C6H4)]2 (11) in the mixture with 5c and
minor amounts of other products (see Section on Dehydrocoupling
of 3a–c in the solid state). The resonances of the two compounds
overlapped and were not separately resolved.
[13] a) H. C. Kelly, F. R. Marchelli, M. B. Giusto, Inorg. Chem. 1964, 3,
431–437; b) C. J. Foret, M. A. Chiusano, J. D. OꢀBrien, D. R.
Martin, J. Inorg. Nucl. Chem. 1980, 42, 165–169; c) C. Camacho,
M. A. Paz-Sandoval, R. Contreras, Polyhedron 1986, 5, 1723–1732;
d) E. Framery, M. Vaultier, Heteroat. Chem. 2000, 11, 218–225.
[14] a) C. Camacho, G. Uribe, R. Contreras, Synthesis 1982, 1027–1030;
b) H. C. Brown, L. T. Murray, Inorg. Chem. 1984, 23, 2746–2753;
c) A. M. Salunkhe, E. R. Burkhardt, Tetrahedron Lett. 1997, 38,
1519–1522; d) H. C. Brown, M. Zaidlewicz, P. V. Dalvi, Organome-
tallics 1998, 17, 4202–4205.
[28] It cannot completely be excluded that also the eea diastereomers of
7a,b were formed. However, by precipitation only the eee isomers
were isolated.
[29] C. K. Narula, J. Fr. Janik, E. N. Duesler, R. T. Paine, R. Schaeffer,
Inorg. Chem. 1986, 25, 3346–3349.
[30] P. W. R. Corfield, S. G. Shore, J. Am. chem. Soc. 1973, 95, 1480–
1487.
[31] L. M. Trefonas, F. S. Mathews, W. N. Lipscomb, Acta Crystallogr.
1961, 14, 273–278.
[32] Along the linear N-H···H-B bound polymer chain of 7a in the solid
state the molecules have alternating propeller chirality. This results
in perpendicular orientation of the phenyl rings between the mole-
cules in which there are short C-H···Cp interactions.
[33] X. Chen, J.-C. Zhao, S. G. Shore, J. Am. Chem. Soc. 2010, 132,
10658–10659.
[15] Compound 3b (and also 3a) has been prepared via a different meth-
od.[13a]
[16] a) M. E. Bowden, I. W. M. Brown, G. J. Gainsford, H. Wong, Inorg.
Chim. Acta 2008, 361, 2147–2153; b) S. Aldridge, A. J. Downs, C. Y.
Tang, S. Parsons, M. C. Clarke, R. D. L. Johnstone, H. E. Robertson,
D. W. H. Rankin, D. A. Wann, J. Am. Chem. Soc. 2009, 131, 2231–
2243.
[17] W. T. Klooster, T. F. Koetzle, P. E. M. Siegbahn, T. B. Richardson,
R. H. Crabtree, J. Am. Chem. Soc. 1999, 121, 6337–6343.
[34] The reaction of 3a with BH3·SMe2 in CD2Cl2 yielded ca. 93% (by
11B{1H} NMR signal integration) of 4a after 3 h. Upon evaporation
of the solvent, the ratio of further products formed was somewhat
lower than in the reaction of 3a with BH3·THF, but still about 50%,
which prevented the isolation of 4a. The preparative method recent-
ly described by Wrackmeyer et al. for the synthesis of m-(amino)di-
boranes starting from diaminoboranes and BH3·THF[20] could not be
applied to the synthesis of 4a; no clean reaction occurred between
5a and three equiv of BH3·THF.
[18] The given ratios based on 11B{1H} NMR signal integration corre-
spond to total ratios of boron atoms. It should be noted that the
values are only rough estimates due to severe overlap of the signals.
The spectra are provided in the Supporting Information.
[19] For the characterization of 4a, 5a, and 7a see below.
[20] B. Wrackmeyer, E. Molla, P. Thoma, E. V. Klimkina, O. L. Tok, T.
Bauer, R. Kempe, Z. Anorg. Allg. Chem. 2011, 637, 401–405.
[21] R. K. Kanjolia, L. K. Krannich, C. L. Watkins, J. Chem. Soc. Dalton
Trans. 1986, 2345–2348.
[22] T. J. Hebden, M. C. Denney, V. Pons, P. M. B. Piccoli, T. F. Koetzle,
A. J. Schultz, W. Kaminsky, K. I. Goldberg, D. M. Heinekey, J. Am.
Chem. Soc. 2008, 130, 10812–10820.
[35] In the 11B NMR spectra the coupling pattern was only resolved
when the reaction was carried out in CH2Cl2 (or CD2Cl2); in THF or
Et2O a broad resonance was detected for 4a. Similar observations
have been made for m-(amino)diborane species before and have
been assigned to dynamic effects: a) W. D. Phillips, H. C. Miller,
E. L. Muetterties, J. Am. Chem. Soc. 1959, 81, 4496–4500; b) D. F.
Gaines, R. Schaeffer, J. Am. Chem. Soc. 1964, 86, 1505–1507.
[36] Dianilinoborane (5a) has been prepared before by the reaction of
aniline with Et3N·BH3 at 50–608C: R. Kçster, H. Bellut, S. Sattori,
Liebigs Ann. Chem. 1968, 720, 1–22. Under these conditions 5a was
[23] We also attempted to prepare a random copolymer via the reaction
of a mixture of 3a and MeNH2·BH3 (ratio 1:9) in THF (10m) in the
presence of 1 (0.1 mol%). An 11B{1H} NMR spectrum recorded
after stirring the mixture for 17.5 h at ambient temperature showed,
besides the dehydrocoupling products of 3a, negligible conversion
of MeNH2·BH3. It should be noted that in the absence of 3a under
these conditions complete formation of [MeNH-BH2]n was observ-
ed,[7b,c] hence, the homopolymerization of MeNH2·BH3 was com-
pletely quenched due to the presence of 3a; the 31P{1H} NMR of the
reaction mixture showed only the resonance of 9. When the catalytic
activity of 9 (0.5 mol%) was tested for the dehydrocoupling/dehy-
drogenation of MeNH2·BH3 (2m in THF), within 17 h less than 3%
of MeNH2·BH3 had reacted. However, after 5 d almost complete
conversion was observed, to give a mixture of the cyclotriborazane
(MeNH-BH2)3 and high molecular weight polymer [MeNH-BH2]n
(Mw =77000 by GPC, PDI 1.2, in THF with 0.1 w/w% [nBu4N]Br,
calibrated with polystyrene standards). BH3·THF (1 mol%) showed
no catalytic activity towards MeNH2·BH3 (2m in THF) within 4 d.
[24] Borazine 6b has been prepared via a different route, see: L. F.
Hohnstedt, D. T. Haworth, J. Am. Chem. Soc. 1960, 82, 89–92.
[25] It is noteworthy that 3a did not dissolve immediately upon solvent
addition. When THF (2.5 mL) was added to 2.5 mmol of 3a a clear
solution was formed within a few minutes. When using Et2O dissolv-
ing was observed within 1 h, and with CH2Cl2 a solution was formed
after a period of 2 h.
formed in mixtures with BACHTUNRTGENNUG(NHPh)3 and PhNHACHTUGNTERN(NUGN BH-NHPh)2. It was
suggested that anilinoborane (8a) was an intermediate of the reac-
tion. Although 5a has been isolated in that study, the spectroscopic
information provided was not sufficient to allow for an unambiguous
assignment of the data that we observed in reactions of 3a in solu-
tion to the structure of 5a.
[37] C. Fedorchuk, M. Copsey, T. Chivers, Coord. Chem. Rev. 2007, 251,
897–924.
[38] S. Harder, Dalton Trans. 2010, 39, 6677–6681.
[39] Recently, various ring opening pathways have been calculated for
the parent m-(amino)diborane, m-(NH2)B2H5, with NH3: W. R. Nutt,
M. L. McKee, Inorg. Chem. 2007, 46, 7633–7645. In their gas phase
calculations the most stable conformation of the linear species NH3-
BH2-NH2-BH3 (gauche) was positive in Gibbs free energy 1.6 kcal
molꢁ1 with respect to the reactants, which has been criticized in
ref. [33]. We have re-investigated this reaction computationally. In
the gas phase we obtained
a similar value with our method,
+9.2 kJmolꢁ1. However, when the polar environment was modeled
using the COSMO approach
a negative value was obtained:
ꢁ9.3 kJmolꢁ1
.
[26] In order to exclude that the enhanced rate of dehydrocoupling in
Et2O compared to THF may be due to impurities in the former sol-
vent, diethyl ether was redistilled from sodium wire prior to use.
[27] The 11B NMR signals observed were assigned to 5a–c through com-
parison with the resonance of isolated 5a (see Section on Reaction
of aniline–borane (3a) with PhNH2 (2a)). With prolonged time
during reactions of 3a–c in solution the respective signal assigned to
5a–c showed no significant change in chemical shift. However, it
[40] It should be mentioned that cyclotriborazane 7a (and also 7b) is
very poorly soluble in THF; in Et2O and CH2Cl2 its solubility is
even lower. Therefore, although substoichiometric amounts (15
mol%) of 5a were employed in this experiment, the ratio of 5a to
7a in solution was approximately 1:1 (estimated by 11B{1H} signal in-
tegration).
[41] Conversion of 7a could also be achieved to some extent at ambient
temperature by using catalyst 1. In the presence of 3 mol% of 1 in
Chem. Eur. J. 2012, 18, 4665 – 4680
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4679

I. Manners et al.
THF after 1 d ca. 9% (by 11B{1H} signal integration) of 6a was
formed, together with 5a (ca. 7%) and 3a (<1%). However, the
31P{1H} NMR spectrum of the reaction mixture revealed that 1 was
completely transformed to 9, and no significant further change was
observed by NMR spectroscopic reaction monitoring within 4 d.
[42] It should be noted that the mixtures derived from 3c at elevated
temperatures might contain compound 11, in addition.
[43] a) F. Baitalow, J. Baumann, G. Wolf, K. Jaenicke-Rçßler, G. Leitner,
Thermochim. Acta 2002, 391, 159–168; b) M. Bowden, T. Autrey, I.
Brown, M. Ryan, Curr. Appl. Phys. 2008, 8, 498–500.
[51] J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824.
[52] a) K. Eichkorn, O. Treutler, H. ꢆhm, M. Hꢅser, R. Ahlrichs, Chem.
Phys. Lett. 1995, 240, 283–290; b) K. Eichkorn, O. Treutler, H.
ꢆhm, M. Hꢅser, R. Ahlrichs, Chem. Phys. Lett. 1995, 242, 652–660;
c) K. Eichkorn, F. Weigend, O. Treutler, R. Ahlrichs, Theo. Chem.
Acc. 1997, 97, 119–124.
[53] A. Schꢅfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–
2577.
[54] A. Klamt, G. Schꢄꢄrmann, J. Chem. Soc. Perkin Trans. 2 1993, 799–
805.
[44] a) T. C. Bissot, R. W. Parry, J. Am. Chem. Soc. 1955, 77, 3481–3482;
b) M. P. Brown, R. W. Heseltine, L. H. Sutcliffe, J. Chem. Soc. A
1968, 612–616.
[45] A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen, F. J.
Timmers, Organometallics 1996, 15, 1518–1520.
[46] a) Bruker-AXS SAINT, Madison, Wisconsin; b) G. M. Sheldrick,
SADABS V2008/1 or TWINABS V2008/4, University of Gçttingen,
Germany; c) G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64,
112–122.
[47] a) R. Ahlrichs, M. Bꢅr, M. Hꢅser, H. Horn, C. Kçlmel, Chem. Phys.
Lett. 1989, 162, 165–169; b) M. von Arnim, R. Ahlrichs, J. Chem.
Phys. 1999, 111, 9183–9190.
[55] A. Bondi, J. Phys. Chem. 1964, 68, 441–451.
[56] T. Helgaker, Chem. Phys. Lett. 1991, 182, 503–510.
[57] a) P. Deglmann, F. Furche, R. Ahlrichs, Chem. Phys. Lett. 2002, 362,
511–518; b) P. Deglmann, F. Furche, J. Chem. Phys. 2002, 117,
9535–9538; c) P. Deglmann, K. May, F. Furche, R. Ahlrichs, Chem.
Phys. Lett. 2004, 384, 103–107.
[58] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[59] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[60] a) F. Weigend, M. Hꢅser, H. Patzelt, R. Ahlrichs, Chem. Phys. Lett.
1998, 294, 143–152; b) F. Weigend, R. Ahlrichs, Phys. Chem. Chem.
Phys. 2005, 7, 3297–3305.
[61] H. Helten, G. Schnakenburg, J. Daniels, A. J. Arduengo III, R.
Streubel, Organometallics 2011, 30, 84–91.
[48] a) P. A. M. Dirac, Proc. R. Soc. London Ser. A 1929, 123, 714–733;
b) J. C. Slater, Phys. Rev. 1951, 81, 385–390.
[49] S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200–1211.
[50] A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100.
Received: October 13, 2011
Published online: March 5, 2012
4680
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4665 – 4680