Donor-Appended N,C-Chelate Organoboron Compounds: Influence of Donor Strength on Photochromic Behaviour

Article abstract of DOI:10.1002/chem.201602410

Recently, four-coordinated N,C-chelate organoboron compounds have been found to show many interesting photochemical transformations depending on the nature of their chelating framework. As such, the effect of substitution on the chelate ligand has been well-established and understood, but the impact of the aryl groups attached to the boron atom remains less clear. To investigate the effect of enhanced charge-transfer character, a series of new N,C-chelate organoboron compounds with donor-functionalized aryl groups have been synthesized and characterized using NMR, UV/Vis, and electrochemical methods. These compounds were found to possess bright and tunable charge-transfer luminescence which is dependent on the donor strength of the amino substituent. In addition, some of these compounds undergo photochromic switching, producing dark isomers of various colors. This work establishes that donor-functionalization of the aryl groups in N,C-chelate boron compounds is an effective strategy for tuning both the photophysical and photochemical properties of such systems. The new findings also help elucidate the influence of electronic structure on the photoreactivity of N,C-chelate organoboron compounds which appears to be as important as steric crowding around the boron atom.

Full text of DOI:10.1002/chem.201602410

DOI: 10.1002/chem.201602410
Full Paper
&
Inorganic Chemistry
Donor-Appended N,C-Chelate Organoboron Compounds:
Influence of Donor Strength on Photochromic Behaviour
Soren K. Mellerup,[a] Kang Yuan,[a] Carmen Nguyen,[b] Zheng-Hong Lu,[b] and Suning Wang*[a]
Abstract: Recently, four-coordinated N,C-chelate organobor-
on compounds have been found to show many interesting
photochemical transformations depending on the nature of
their chelating framework. As such, the effect of substitution
on the chelate ligand has been well-established and under-
stood, but the impact of the aryl groups attached to the
boron atom remains less clear. To investigate the effect of
enhanced charge-transfer character, a series of new N,C-che-
late organoboron compounds with donor-functionalized aryl
groups have been synthesized and characterized using NMR,
UV/Vis, and electrochemical methods. These compounds
were found to possess bright and tunable charge-transfer lu-
minescence which is dependent on the donor strength of
the amino substituent. In addition, some of these com-
pounds undergo photochromic switching, producing dark
isomers of various colors. This work establishes that donor-
functionalization of the aryl groups in N,C-chelate boron
compounds is an effective strategy for tuning both the pho-
tophysical and photochemical properties of such systems.
The new findings also help elucidate the influence of elec-
tronic structure on the photoreactivity of N,C-chelate orga-
noboron compounds which appears to be as important as
steric crowding around the boron atom.
Introduction
Over the past several decades, organoboron chemistry has
gathered significant attention owing to the diverse range of
properties available to this class of chemical species. Whereas
applications of three-coordinate organoboron compounds fre-
quently make use of the empty pp orbital on the boron atom
(e.g. electron transport/p-conjugated materials[1] or compo-
nents of catalytic systems[2]), those of four-coordinate organo-
boron tend to focus on their rich optoelectronic properties;
that is, organic light-emitting diodes (OLEDs),[3] nonlinear
optics (NLO),[4] and much more recently, memory devices.[5]
With respect to the latter, we have had a long-standing inter-
est in the photochemical behaviour of N,C-organoboron che-
lates (A; Scheme 1) which readily undergo thermally reversible
intramolecular CꢀC bond forming/breaking to quantitatively
generate the dark isomer B.[6] Following this initial discovery,
numerous modifications to the chelating ligands have estab-
lished this reactivity as a general transformation[7] of N,C-orga-
noboron chelates, as well as revealing several alternative iso-
Scheme 1. Previously studied systems A and C, as well as proposed 1–3 and
1’–4’.
merization pathways depending on the ring size and donation
strength of the neutral donor in the N,C-chelate unit.[8] Despite
the success of different chelating ligands, less consideration
has been given to the aryl groups attached to boron. This is
due, in part, to the synthetic challenges associated with the
preparation and isolation of XBR2 reagents, as well as their lim-
ited availability from commercial sources. To date, the boron
center has been substituted with less bulky groups such as
phenyl or pentafluorophenyl,[7] that yield photochemically inert
molecules. Although this clearly establishes the importance of
the flanking methyl groups on the aryl ring in the photoisome-
rization process, the electronic effect of the boron substituents
remains less clear. Given that the charge-transfer (CT) transition
from the aryl p!chelate p* (e.g. Mes-p!py-p*, ppy=2-phe-
nylpyridyl) is believed to initiate the photochemical transfor-
[a] S. K. Mellerup, K. Yuan, Prof. Dr. S. Wang
Department of Chemistry, Queen’s University
90 Bader Lane, Kingston, Ontario, K7L 3N6 (Canada)
E-mail: suning.wang@chem.queensu.ca
[b] C. Nguyen, Prof. Dr. Z.-H. Lu
Department of Materials Science and Engineering
University of Toronto, 27 King’s College Circle, Toronto, Ontario,
M5S 1A1 (Canada)
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under http://dx.doi.org/10.1002/
chem.201602410.
Chem. Eur. J. 2016, 22, 1 – 10
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
mation of these species, we postulate that substituting the Ar
groups with strong electron donors may alter or enhance the
photochromic properties of such systems. Furthermore, the
use of amine donors could impart interesting electrochemical[9]
or photophysical properties such as thermally activated de-
layed fluorescence (TADF),[10] which has been observed for re-
lated four-coordinated organoboron analogues with N,N-pyrid-
yl-pyrrolo chelates.[10a] We therefore set out to design and pre-
pare various donor-appended organoboron derivatives of the
general form (ppy)BAr2, in which the Ar substituents bear an
electron-donating moiety bound to m-xylene or phenyl
(Scheme 1, right).
X-ray diffraction analyses were performed for 2 and 2’ (see the
Supporting Information). The 11B chemical shifts of all seven
compounds fall between 3–5 ppm, which are typical for this
class of compounds.[7,8] The X-ray crystal structures of 2 and 2’
(Figure 1) reveal a significant difference between their BꢀN
Results and Discussion
Synthesis and structures
The donor-functionalized organoboron compounds 1–3 and
1’–4’ were prepared using one of the two methods shown in
Scheme 2: (a) reacting FBAr2 reagents, generated in a one-pot
Figure 1. X-ray crystal structure of 2 (top) and 2’ (bottom). Key bond lengths
[ꢁ] and angles [8] for 2: B(1)ꢀN(1) 1.638(2), B(1)ꢀC(1) 1.647(2), B(1)ꢀC(12)
1.638(2), B(1)ꢀC(20) 1.646(2), N(1)-B(1)-C(1) 95.03(1), C(12)-B(1)-C(20)
113.65(1); for 2’: B(1)ꢀN(1) 1.614(4), B(1)ꢀC(1) 1.606(4), B(1)ꢀC(12) 1.620(4),
B(1)ꢀC(18) 1.611(4), N(1)-B(1)-C(1) 96.5(1), C(12)-B(1)-C(18) 113.4(1).
Scheme 2. Synthesis of 1–3 and 1’–4’.
procedure, with 2-(2-lithiophenyl)pyridine, or (b) nucleophilic
[11]
substitution of (ppy)BBr2 with appropriate Grignard reagents.
and BꢀC bond lengths, much like their donor-absent counter-
parts A and C.[7a] The steric bulk provided by the o-methyl
groups results in longer bonds around the boron atom (e.g.
B1ꢀN1=1.638(2) and 1.614(4) ꢁ; B1ꢀC1=1.647(2) and
1.606(4) ꢁ; B1ꢀC12/C20=1.638(1) and 1.646(2), and B1ꢀC12/
C18=1.620(4) and 1.611(4) ꢁ for 2 and 2’ respectively), which
is known to be a major contributing factor to the photochro-
mic behavior observed in A and its derivatives. Compared to A
and C, the BꢀCAr bond lengths of 2 and 2’ are shorter by
around 0.01 and 0.02 ꢁ, respectively,[7] whereas the various
angles around the boron atom remain similar. Due to the lack
of suitable crystals, the remaining compounds were compared
to their DFT-optimized structures, which matched well with
the structural data obtained for 2 and 2’. In all seven of the
calculated structures, the nitrogen atom of the amine donor
adopts a trigonal-planar geometry as a result of lone-pair de-
localization through the adjacent p-system(s) (see the Support-
ing Information).
The use of the two separate routes was necessary to obtain
good yields. The sterically congested molecules 1–3 were ob-
tained in good yields using method (a) whereas compounds
1’–4’ were prepared in moderate yields using method (b) (45–
80%). Method (a) is an improved procedure compared to that
reported for (ppy)BMes2 (A) and analogues,[7,8] in which FBAr2
reagents from commercial sources were used. As such, our
one-pot FBAr2 protocol can be used as an inexpensive and effi-
cient technique for incorporating various BAr2 or BR2 units into
N,C-organoboron chelates. It is worth noting that other trans-
metallation procedures[12] may also be effective in obtaining
the new donor-functionalized N,C-chelate boron compounds,
such as the borenium-cation-mediated boro-destannylation or
desilylation reported by Ingleson and coworkers.[12c] Four rep-
resentative nitrogen-based donors, namely carbazolyl (Cbz, 1,
and 1’), diphenylamino (NPh2, 2, and 2’), dimethylamino (NMe2,
3, and 3’), and phenoxazine (pxz, 4’) were chosen to study
a range of donor strengths, and to vary the HOMO energy
levels, within these systems.
Compounds 1–3 and 1’–4’ were fully characterized by 1H,
13C, and 11B NMR, HRMS, and elemental analysis. Single-crystal
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Photophysical and electrochemical properties
backbone. The CT band of 3 is significantly red-shifted, com-
pared to 3’, which can be attributed to the two additional
methyl groups on the aryl ring in 3 that raise the HOMO
energy and decrease the HOMO–LUMO gap. Similar differences
are also observed for the pairs 1/1’ and 2/2’, although they are
less obvious.
Shown in Figure 2 are the UV/Vis and fluorescence spectra of
all seven compounds with their pertinent photophysical and
electrochemical data summarized in Table 1. With the excep-
tion of 3, which possesses a weak absorption band at 410 nm,
Compounds 1–3 and 1’–4’ are all moderately fluorescent in
solution with emission color ranging from deep-blue to orange
depending on the amine donor (see the Supporting Informa-
tion). Interestingly, the emission energy of the congested part-
ner in each pair (i.e. 1/1’, 2/2’ and 3/3’) is consistently red-shift-
ed by 25–30 nm relative to the noncongested one, which
agrees with the trend of their CT absorption bands. Further-
more, the emission energy decreases in the order of 1>2>3
and 1’>2’>4’>3’, indicating that the donor strength of the
amino groups likely follows the order of Cbz<NPh2 <Pxz<
NMe2. This is reasonable since the greater delocalization of the
lone pair in the extended p-system of carbazole would reduce
its donor strength compared to that of NMe2. When doped in
PMMA films (5 wt%), the emission energy of all seven com-
pounds is hypsochromically shifted by 25–30 nm and exhibits
much higher and impressive quantum yields (0.56–0.79) with
the exception of 1 and 2, which have a low emission quantum
efficiency caused by the effective competing photoisomeriza-
tion pathway (see next section).
Figure 2. UV/Vis and fluorescence spectra of 1–3 (solid lines) and 1’– 4’
(dashed lines) in toluene at 10ꢀ5 m. Inset: Main orbital contribution to the
calculated S1 transition of 1.
Table 1. Photophysical and electrochemical properties of 1–3 and 1’–4’.
The cyclic voltammetry (CV) diagrams of all seven com-
pounds display a single reduction peak at around ꢀ2.30 to
ꢀ2.40 V (Figure 3), which is attributed to the reduction of the
red
labs [nm]
lem [nm]
Ff
(solution[a]/DF[b]
E1/2ox/E1/2
(e [Mꢀ1 cmꢀ1])
(solution[a]/DF[b]
)
)
[V][c]
1
2
3
1’
2’
3’
4’
347 (20503)
313 (29682)
410 (1951)
345 (19104)
308 (27077)
378 (10730)
329 (35016)
480/457
552/516
598/562
452/438
518/491
575/537
532/506
0.12/0.05
0.17/0.15
0.29/0.56
0.20/0.79
0.22/0.76
0.31/0.77
0.12/0.61
0.94/ꢀ2.40
0.85/ꢀ2.33
0.59/ꢀ2.31
1.03/ꢀ2.43
0.67/ꢀ2.34
0.48/ꢀ2.39
0.60/ꢀ2.39
[a] 10ꢀ5 m in toluene. [b] Doped films; 5 wt% in PMMA. [c] From CV meas-
urements recorded in mixed solvents of CH3CN :C6H6 (1:2) relative to FcH/
FcH+.
each species absorbs strongly in the UV region with their ab-
sorption maxima (labs) between 305–380 nm. Unlike the previ-
ously reported photochromic (ppy)BMes2 system[7a] that has
a typical low-energy shoulder peak corresponding to the Mes-
p!ppy-p* charge-transfer (CT) transition, compounds 1–3 and
1’–4’ have well-resolved low-energy peaks in their absorption
spectra. To understand this difference, TD-DFT calculations
were performed at the M06-2X/6-31g(d) level of theory as this
combination was able to most accurately reproduce the UV/Vis
spectra of the boron compounds. In all cases, the primary con-
tribution to the calculated S1 vertical excitation was HOMO!
LUMO with the HOMO being located predominantly on the
amine donor p-system and the LUMO situated on the ppy
backbone (p*) as shown in Figure 2. These well-separated mo-
lecular orbitals are reminiscent of the CT character in similar
pyridyl–pyrrolo N,N-chelate organoboron compounds, which
have been shown to be applicable as TADF emitters.[10a] There-
fore, the low-energy absorption band in compounds 1–3 and
1’–4’ is assigned as CT from the amine donors to the ppy
Figure 3. CV diagrams showing the reduction and oxidation of 1–3 and 1’–
4’, recorded in mixtures of CH3CN :C6H6 (1:2) with scan rates of 200 mVsꢀ1
.
p* orbital of the N,C-chelating ligand and matches well with
those of (ppy)BMes2 and related compounds.[7a] The oxidation
peaks of 1–3 and 1’-4’ are all fairly similar to those of their re-
spective amine donors,[9a,d–f] with pseudo-reversible oxidation
of the aryl amines at around 1.30 to 0.50 V. For some of the
compounds, two well-resolved oxidation peaks corresponding
to the two amino groups were observed. The oxidation poten-
tials of 1–3 and 1’–4’ decrease in the same order as that ob-
served for their emission energies, again supporting the in-
creasing donor strength of Cbz<NPh2 <Pxz<NMe2. Utilizing
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
the CV and absorption data, the HOMO–LUMO energies of 1–3
and 1’–4’ were determined experimentally using the oxida-
tion/reduction peak potentials and compared to TD-DFT calcu-
lated data (Figure 4). In general, the calculated values follow
the same trend as the experimental ones, in which changing
the donor from Cbz!NPh2!Pxz!NMe2 results in a gradual
increase in HOMO energy and thus a narrowing of the optical
band gap. Furthermore, the flanking methyl groups in 1–3
slightly destabilize the HOMO energies compared to 1’–4’ as
a result of the increased electron richness of the phenyl ring.
photochromic behaviour similar to that of A and its analogues
under 350 nm light. Upon monitoring the transformation by
UV/Vis and NMR spectra (Figure 5 and the Supporting Informa-
Figure 4. DFT calculated at the M06-2X/6-31g(d) level of theory (grey) versus
experimental HOMO–LUMO energies of 1–3 and 1’–4’.
Photochromic properties
Upon exposure to UV irradiation, 1–3 and 1–4’ each give dis-
tinct responses as shown in Scheme 3. Both 1 and 2 exhibit
Figure 5. UV/Vis time-lapsed spectra of 1 (top) and 2 (bottom) in toluene at
10ꢀ4 m exposed to 365 nm irradiation at 5 s intervals. Inset: photographs
showing the color of 1a and 2a.
tion), it is clear that the desired dark isomer species 1a and 2a
formed. For example, the UV/Vis spectrum of 1 (ca. 10ꢀ4 m in
toluene) shows a steady decrease of the main absorption band
at 345 nm whereas a new characteristic[7,8] absorption band,
assigned to 1a, grows in at around 558 nm. TD-DFT calcula-
tions agree with previous findings[7,8] which suggest that the
origin of this new band is a CT transition in 1a and 2a from
HOMO (p) located on the boracycle and cyclohexadiene ring
to LUMO (ppy-p*) (see the Supporting Information). This is ac-
companied by a change in solution appearance from colorless
to purple for 1a and yellow to deep-blue for 2a (see the Sup-
porting Information), demonstrating that varying the amino
group on the aryl unit of the boron compounds is one possi-
1
ble approach to tune the color of the dark isomers. The H and
11B NMR show that these transformations are clean, with two
distinct cyclohexadiene singlet peaks, well-resolved o-methyl
peaks, and a sharp 11B chemical shift at around ꢀ10 ppm (see
the Supporting Information). Using (ppy)BMes2 (A) as an inter-
nal standard (which has a photoisomerization quantum effi-
ciency of 0.85), competitive NMR photolysis experiments were
carried out at 365 nm for 1 and 2. Their relative rates of photo-
isomerization (kPI) were determined to be 1.1 and 6.5 for kPI(A)/
Scheme 3. Photoreactivity of 1–3 and 1’–4’.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
kPI(1) and kPI(A)/kPI(2), respectively. Using these relative rate
constants and absorption data, the relative photoisomerization
efficiency[13] (FPI) of 1 and 2 versus A were approximated to be
0.64 and 0.10, respectively. The much lower photoisomeriza-
tion efficiency of 2 may be caused by an increase in the nonra-
diative rate constant of 2 due to additional rotational degrees
of freedom, which competes with the photochemical transfor-
mation. Therefore, enhancing the charge-transfer character of
such systems does not increase the rate or efficiency of photoi-
somerization. The dark isomers 1a and 2a are both air-sensi-
tive, decomposing into their anticipated[7] deborylated/CꢀC-
coupled products as confirmed by HRMS (see the Supporting
Information). Just as with the dark isomer B, 1a and 2a can be
readily converted back to 1 and 2 through the use of heat
with comparable activation energies (see the Supporting Infor-
mation).[6]
butions from the BꢀCAr and BꢀCppy bonds, which are very simi-
lar to the HOMO and HOMOꢀ4 orbitals of A. Thus, upon exci-
tation of A, 1, and 2 near their absorption maxima, it is possi-
ble that the BꢀCAr and BꢀCppy bonds are significantly weakened
which would ultimately facilitate the CArꢀCppy and BꢀCAr cou-
pling as well as BꢀCPPy bond breaking. Even though these orbi-
tals are present in 3, they only contribute to higher energy ver-
tical excitations (e.g. S5), suggesting that the ꢀN(CH3)2 donor
allows the fluorescence CT transition to dominate the relaxa-
tion from the excited state, as supported by its much greater
fluorescent quantum efficiency compared to A, 1, and 2 (see
Table 1). This increase in Ffl can be attributed to the less-bulky
ꢀN(CH3)2 groups of 3 compared to the Cbz or NPh2 groups of
1 and 2 that allow the donor moiety to be coplanar with the
phenyl ring (see calculated structures in the Supporting Infor-
mation), which reduces twisted intramolecular charge-transfer
(TICT) quenching.[14] In addition, the narrow optical gap of 3
should increase its non-radiative decay rate because of the
energy-gap law.[15] The combination of these two effects are
what likely suppress the photoreactivity of 3 as kPI becomes
negligible compared to kr +knr as shown in Figure 6. Further-
more, the balance between competing rate constants could
explain the FPI trend of 1–3, as 1 has the band gap that
matches most closely with A. Overall, these findings indicate
that the electronic structure of the Ar groups on boron are ex-
tremely important to the photoisomerization event and could
provide an alternative strategy[16] for preparing photochemical-
ly inert N,C-organoboron chelates.
In sharp contrast to all other reported N,C-chelate organo-
boron compounds bearing o-methyl groups, compound 3 was
found to be stable towards 300 and 350 nm irradiation (see
the Supporting Information). This finding is quite surprising, as
the steric strain and HOMO destabilization induced by the o-
methyl groups are known to be some of the most significant
contributors to the photochromic behavior of these com-
pounds. Given that the experimentally determined HOMO level
and steric congestion of 3 are comparable to those of A, 1,
and 2, the lack of photochromic switching in 3 must be
caused by the impact of ꢀN(CH3)2 on the electronic structure
of the molecule. TD-DFT data of A and 1–3 reveal that the
photostable system 3 has different electronic contributions to
its vertical excitations. In the case of 3, the S1 vertical excitation
only has one contribution, HOMO (Ar-p)!LUMO (ppy-p*),
which was assigned to the radiative decay pathway of these
compounds (Figure 2). A, 1, and 2 all possess the same transi-
tion as the primary contribution to S1, but also a secondary
transition to S1 from either HOMOꢀ4 or HOMOꢀ3 to LUMO
(ppy-p*) (see Figure 6 and the Supporting Information). The
HOMOꢀx (x=3 or 4) orbitals in 1 and 2 both have large contri-
To further probe the electronic influence of the amine
donors, analogues lacking the o-methyl functionality were in-
vestigated. Initially, colorless, solutions of both 1’ and 2’
changed to orange and red respectively upon exposure to
1
300 nm light. H NMR reveals numerous new and broad peaks
in the aromatic region over time, with the peaks of 1’ slowly
disappearing over the course of 24 h (see the Supporting Infor-
mation). Heating of the colored samples at 908C for 18 h re-
sulted in the reappearance of the chemical shifts belonging to
1’ along with numerous unassignable peaks which were pres-
ent after irradiation. This implies that both 1’ and 2’ are capa-
ble of undergoing some form of photochemical change along
with irreversible decomposition. Since the orange and red
colors of the solutions persist through the thermal reversal
back to 1’ and 2’, they likely belong to the decomposition
products. Irreversible photochemical oxidation of the amino
groups may be responsible for the poor photostability of 1’
and 2’ that is not observed for analogous systems such as C,
which lacks the electron-donating amino group.[7a] The lack of
photoisomerization in 1’ and 2’ can be attributed to the ab-
sence of the o-methyl groups, which, in addition to weakening
the BꢀCAr bonds, may also stabilize the various three-coordi-
nated boron intermediates formed along the photoisomeriza-
tion pathways as illustrated previously for A and its deriva-
tives.[8b,c,e] Much like for 3, compounds 3’ and 4’ were found to
be photochemically inert.
Figure 6. Energy diagram explaining the differing photochemical behavior
of A and 1–3 in terms of competing rate constants, in which kr =rate of radi-
ative decay, knr =rate of nonradiative decay, and kPI =rate of photoisomeriza-
tion. It is important to note that the isomerization process has been shown
to proceed through a photoactive triplet state[17] and therefore kPI includes
contributions from kISC (rate of intersystem crossing).
Based on these findings and inspired by the work of Chou
et al. on related N,N-chelate boron compounds,[10a] we attempt-
ed to investigate compound 4’ as a thermally activated de-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
layed fluorescence emitter (TADF). Despite meeting most of
the criteria, such as well-separated HOMO and LUMO, a small
DESꢀT and a high solid-state quantum efficiency, 4’ was found
to degrade under standard OLED operating conditions (see the
Supporting Information) and therefore displayed very poor
performance. Given the propensity for such systems to under-
go photo-oxidation or exciton-driven transformations within
OLED devices,[5a, 18] it is entirely possible that the instability of
4’ is due to a similar phenomenon. As a result, N,C-chelate or-
ganoboron compounds bearing an amino group are not suita-
ble for use as emitters in OLEDs.
ing on the donor strength of the amino group, the lumines-
cence of these chelate compounds can be tuned from deep-
blue to orange as a result of varying HOMO energy levels. Of
the compounds bearing o-methyl groups, only 1 and 2 were
found to be photochromic with 3 being completely inert to
UV irradiation. This is the first example of a photochemically
inert N,C-chelate organoboron molecule with flanking methyl
groups and thus revealed that the electronic nature of the
charge-transfer transitions is equally as important as steric con-
gestions in terms of governing the photochromic switching in
this class of compounds. This is further corroborated by the
photoinstability of the nonsterically congested molecules 1’
and 2’, which are not yet fully understood. Despite the photo-
stable nature of 3, 3’, and 4’, their electrochemical instability
under standard device operating conditions prevents the use
of this class of molecules as TADF emitters unlike their pyridyl
pyrrolo-N,N-chelate analogues.
Finally, the electrochemical properties of the dark isomers
were investigated using CV measurements as shown in
Figure 7. Comparable to the dark isomer B and its analogue-
Experimental Section
General Procedures
All reactions were carried out under an inert atmosphere of dry ni-
trogen while employing standard Schlenk techniques. All starting
materials were purchased from Sigma–Aldrich or Matrix Chemical
Company (4-bromophenyl-N,N-diphenylamine and 4-bromo-N,N-di-
methylaniline) and used without further purification. 9-(4-bromo-
phenyl)-9H-carbazole,[19] 10-(4-bromophenyl)-10H-phenoxazine,[20]
4-bromo-3,5-dimethyl-N,N-diphenylaniline,[21] 9-(4-bromo-3,5-dime-
thylphenyl)-9H-carbazole,[21] and 4-bromo-N,N-3,5-tetramethylani-
line[22] were prepared according to literature procedures. Solvents
were dried over Na, degassed, and stored over 4 ꢁ molecular
Figure 7. CV diagrams showing the oxidation peaks of 1a (bottom) and 2a
(top), recorded in CH3CN with scan rates of 200 mVsꢀ1. Inset: Assignment of
the oxidation peaks to the HOMO and HOMOꢀx orbitals of 1a and 2a.
1
sieves prior to use. H, 13C, and 11B NMR spectra were recorded on
s,[9a,d] the dark isomers 1a and 2a display one new, nonreversi-
ble oxidation peak at ꢀ0.32 and ꢀ0.48 V, respectively, in addi-
tion to oxidation peaks at 1.24 and 0.85 V (Figure 7). The peaks
at negative potential are assigned to the HOMO energy level
of 1a (ꢀ4.33 eV) and 2a (ꢀ4.17 eV), which are dominated by
the cyclohexadiene p-system and “bent” bonds of the three-
membered boracyclic rings. The peaks appearing at positive
potential are assigned to lower-lying molecular orbitals (e.g.
HOMOꢀ1 and HOMOꢀ2) that are centered on amine donor p-
systems just as in 1 and 2, as supported by DFT data.
Although the relatively high HOMO of 1a and 2a explains
their sensitivity towards oxygen, the oxidation potential of 1a
in fact is lower than those of previously reported dark isomer-
s.[7a] Attempts to reverse the photoisomerization and convert
1a back to 1, and 2a back to 2 by bulk electrolysis proved un-
successful, as prolonged oxidation near the oxidation potential
of the respective triarylamino groups resulted in no observable
spectral or color change.
a Bruker Avance 400 MHZ spectrometer using deuterated solvents
that were purchased from Cambridge Isotopes. High resolution
mass spectra (HRMS) were obtained using a Micromass GCT TOF-EI
Mass Spectrometer. Elemental analyses were performed by the Ele-
mental Analysis Service at the University of Montreal. Excitation
and emission spectra were recorded using a Photon Technologies
International QuantaMaster Model 2 spectrometer. UV/Vis spectra
were recorded a Varian Cary 50 spectrometer. Photoluminescent
quantum yields were measured using the optically dilute method
(Absorbanceꢁ0.1) relative to quinine sulfate in 0.5m H2SO4 (Fsol
=
0.546)[23] and an integration sphere. Cyclic voltammetry was per-
formed using a BAS CV-50W analyzer at a scan rate of 200 mVsꢀ1
with sample concentrations of 5 mg/3 mL in CH3CN :C6H6 (1:2),
0.1m NBu4PF6 as supporting electrolyte, and glassy carbon refer-
ence electrode. The ferrocenium/ferrocene couple was used as an
internal standard (Eo =0.342 V).[24] HOMO–LUMO energies were de-
termined from CV data using the E1/2 for pseudo-reversible peaks
and oxidation/reduction peak maxima for nonreversible peaks.
Synthesis of FBR2 reagents and boron chelates 1–3
Conclusions
A 125 mL oven-dried Schlenk flask under N2 was charged with
magnesium turnings (0.034 g, 2.1 mmol), 4-bromo-3,5-dimethyl-
phenyl-N-heterocycle (0.90 g, 2.1 mmol), one iodine granule, THF
(30 mL), and a magnetic stirrer bar. The magnesium was activated
using a heat gun and the resulting mixture refluxed for 2 h or until
all of the magnesium had dissolved. The solution was cooled to
In conclusion, seven new donor-appended N,C-chelate organo-
boron compounds have been prepared by two different meth-
ods which give good yields and provide a facile approach for
introducing various Ar groups onto the boron atom. Depend-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ꢀ788C using
a
dry-ice bath (30 min) and BF3·Et2O (0.30 g,
and concentrated in vacuo. Flash column chromatography employ-
ing gradient elution (4:1!1:1 hexane :CH2Cl2) afforded the desired
products in moderate yields (40–50%). The compounds were fur-
ther purified by recrystallization from the slow evaporation of
a hexane :CH2Cl2 (~1:1) solution.
2.1 mmol) was added dropwise, after which the dry-ice bath re-
moved. After stirring at room temperature for 18 h, the freshly pre-
pared FBR2 solution was cooled to ꢀ788C. In a separate 125 mL
oven-dried Schlenk flask under N2 was added 2-(2-bromophenyl)-
pyridine[6] (0.49 g, 2.1 mmol) and Et2O (15 mL). After cooling the
mixture to ꢀ788C (30 min), nBuLi (0.92 mL, 2.3 mmol, 2.5m) was
added dropwise to give a bright-yellow solution that was kept at
ꢀ788C for 1.5 h. The FBR2 solution was then transferred dropwise
using a cannula to the lithiated PhPy solution, and the completed
reaction was slowly warmed to room temperature over 18 h. Ex-
traction with CH2Cl2 (30 mL), concentration in vacuo, and flash
column chromatography employing gradient elution (10:1!1:1;
hexane :CH2Cl2) afforded the desired products in good yields (60–
80%). The compounds were further purified by recrystallization
from a hexane :CH2Cl2 (~1:1) solution.
1
(PhPy)B(p-Ph-Cbz)2 (1’): white solid, 50% yield. H NMR (400 MHz,
CD2Cl2): d=8.75 (d, J=5.7 Hz, 1H, Py-H), 8.26–8.22 (m, 2H, Py-H
and Ph-H), 8.16 (d, J=7.7 Hz, 4H, Cbz-H), 8.08 (d, J=7.7 Hz, 1H,
Ph-H), 7.93 (d, J=8.0 Hz, 1H, Py-H), 7.64–7.55 (m, 6H, B-Ph-N and
Py-H), 7.53–7.45 (m, 13H, Cbz-H and B-Ph-N), 7.29 ppm (t, J=
7.3 Hz, 4H, Cbz-H); 13C NMR (100 MHz, CD2Cl2): d=150.8, 140.6,
140.3, 134.6, 133.6, 130.6, 129.9, 125.7, 125.1, 125.0, 122.3, 121.8,
121.4, 120.6, 119.3, 118.8, 118.0, 115.4, 110.0, 109.2 ppm; 11B NMR
(128 MHz, CD2Cl2): d=3.1 ppm; HRMS (EI): m/z calcd for C47H32BN3
[M]+: 649.2697; found 649.2693; elemental analysis calcd (%) for
C47H32BN3: C 86.90; H 4.97; N 6.47; found: C 84.11; H 5.03; N 6.31.
Low carbon analysis is common for boron-containing compounds
due to the formation of refractory boron carbide during analysis.
(PhPy)B(2,6-diMe-Ph-4-Cbz)2 (1): white solid, 60% yield. 1H NMR
(400 MHz, CD2Cl2): d=8.85 (d, J=5.5 Hz, 1H, Py-H), 8.24–8.10 (m,
6H, Py-H and Cbz-H), 8.06 (dd, J=7.6 Hz, 1H, Ph-H), 8.01 (d, J=
7.3 Hz, 1H, Ph-H), 7.57–7.35 (m, 11H, Ph-H, Py-H, and Cbz-H), 7.27
(t, J=7.7 Hz, 4H, Cbz-H), 7.10 (bs, 4H, B-Ph-N), 2.04 ppm (bs, 12H,
Ph-CH3); 13C NMR (100 MHz, CD2Cl2): d=159.5, 146.1, 141.3, 141.0,
140.7, 138.4, 135.2, 134.3, 131.4, 131.1, 126.9, 125.9, 125.6, 123.3,
123.0, 122.1, 120.2, 120.0, 119.4, 118.3, 110.1, 110.0, 22.5 ppm;
11B NMR (128 MHz, CD2Cl2): d=4.5 ppm; HRMS (EI): m/z calcd for
C51H40BN3 [M]+: 705.3324; found 705.3320; elemental analysis calcd
(%) for C51H40BN3: C 86.80; H 5.71; N 5.95; found: C 86.96; H 5.76;
N 5.09.
(PhPy)B(p-Ph-NPh2)2 (2’): pale-yellow needles, 47% yield. 1H NMR
(400 MHz, CD2Cl2): d=8.59 (dd, J=5.7, 1.2 Hz, 1H, Py-H), 8.14–8.10
(m, 2H, Py-H), 7.98 (d, J=7.5 Hz, 1H, Ph-H), 7.78 (d, J=7.2 Hz, 1H,
Ph-H), 7.50 (q, J=7.2 Hz, 1H, Ph-H), 7.48 (qd, J=5.8, 2.1 Hz, 1H, Py-
H), 7.41 (td, J=7.6, 1.1 Hz, 1H, Ph-H), 7.23 (t, J=7.8 Hz, 8H, NPh2-
H), 7.16 (d, J=8.4 Hz, 4H, B-Ph-N), 7.06 (d, J=8.0 Hz, 8H, NPh2-H),
7.02–6.94 ppm (m, 8H, NPh2-H and B-Ph-N); 13C NMR (100 MHz,
CD2Cl2): d=158.5, 148.6, 145.8, 144.5, 141.3, 136.5, 134.2, 131.4,
130.9, 129.6, 129.4, 126.4, 124.5, 124.3, 124.0, 123.1, 122.6, 122.4,
122.3, 118.8 ppm; 11B NMR (128 MHz, CD2Cl2): d=3.3 ppm; HRMS
(EI): m/z calcd for C47H36BN3 [M]+: 653.3010; found 653.3005; ele-
mental analysis calcd (%) for C47H36BN3: C 86.37; H 5.55; N 6.43;
found: C 86.69; H 5.78; N 6.24.
(PhPy)B(2,6-diMe-Ph-4-NPh2)2 (2): bright-yellow needles, 80%
1
yield. H NMR (400 MHz, CD2Cl2): d=8.72 (d, J=5.8 Hz, 1H, Py-H),
8.09 (d, J=4.0 Hz, 2H, Py-H), 7.97 (d, J=7.9 Hz, 1H, Ph-H), 7.88 (d,
J=7.29 Hz, 1H, Ph-H), 7.43 (t, J=7.1 Hz, 1H, Ph-H), 7.40–7.34 (m,
2H, Py-H and Ph-H), 7.22 (t, J=7.7 Hz, 8H, NPh2-H), 7.04 (d, J=
7.7 Hz, 8H, NPh2-H), 6.95 (t, J=7.7 Hz, 4H, NPh2-H), 6.62 (bs, 4H, B-
Ph-N), 1.79 ppm (bs, 12H, Ph-CH3); 13C NMR (100 MHz, CD2Cl2): d=
158.5, 149.7, 147.4, 145.3, 143.4, 140.7, 140.1, 134.2, 130.3, 130.2,
128.3, 128.1, 124.9, 124.7, 123.1, 122.5, 121.1, 121.0, 120.9, 117.3,
30.8 ppm; 11B NMR (128 MHz, CD2Cl2): d=4.3 ppm; HRMS (EI): m/z
calcd for C51H44BN3 [M]+: 709.3637; found 709.3633; elemental
analysis calcd (%) for C51H44BN3: C 86.31; H 5.86; N 5.92; found: C
85.82; H 5.37; N 5.81.
(PhPy)B(p-Ph-NMe2)2 (3’): yellow solid, 42% yield. 1H NMR
(400 MHz, CD2Cl2): d=8.52 (dt, J=5.7, 1.2 Hz, 1H, Py-H), 8.10–8.06
(m, 2H, Py-H and Ph-H), 7.95 (dt, J=7.6, 0.9 Hz, 1H, Ph-H), 7.66 (d,
J=7.1 Hz, 1H, Ph-H), 7.48–7.33 (m, 3H, Ph-H and Py-H), 7.06 (d, J=
8.6 Hz, 4H, B-Ph-N), 6.65 (d, J=8.6 Hz, 4H, B-Ph-N), 2.88 ppm (s,
12H, N-CH3); 13C NMR (100 MHz, CD2Cl2): d=158.3, 149.7, 144.4,
140.8, 136.4, 134.1, 131.1, 130.8, 126.0, 122.4, 122.1, 119.9, 119.3,
118.5, 113.0, 41.0 ppm; 11B NMR (128 MHz, CD2Cl2): d=3.6 ppm;
HRMS (EI): m/z calcd for C27H28BN3 [M]+: 405.2381; found:
405.2385; elemental analysis calcd (%) for C27H28BN3: C 80.00; H
6.96; N 10.37; found: C 80.12; H 7.10; N 10.24.
(PhPy)B(2,6-diMe-Ph-4-NMe2)2 (3): yellow-orange solid, 41% yield.
1H NMR (400 MHz, CD2Cl2): d=8.63 (d, J=5.9 Hz, 1H, Py-H), 8.03 (d,
J=4.5 Hz, 2H, Py-H), 7.91 (d, J=7.9 Hz, 1H, Ph-H), 7.77 (d, J=
7.5 Hz, 1H, Ph-H), 7.40–7.22 (m, 3H, Ph-H and Py-H), 6.30 (s, 4H, B-
Ph-N), 2.84 (s, 12H, B-Ph-N), 1.83 ppm (bs, 12H, Ph-CH3); 13C NMR
(100 MHz, CD2Cl2): d=159.4, 148.5, 146.5, 141.2, 140.8, 138.8, 135.1,
131.1, 125.3, 122.1, 121.8, 118.9, 118.1, 114.6, 111.1, 40.9, 25.6 ppm;
11B NMR (128 MHz, CD2Cl2): d=4.6 ppm; HRMS (EI): m/z calcd for
C31H36BN3 [M]+: 461.3008; found: 461.3003; elemental analysis
calcd (%) for C31H36BN3: C 80.69; H 7.86; N 9.11; found: C 80.76; H
7.91; N 9.32.
(PhPy)B(p-Ph-Pxz)2 (4’): pale-yellow solid, 41% yield. 1H NMR
(400 MHz, CD2Cl2): d=8.42 (dt, J=5.8, 1.0 Hz, 1H, Py-H), 8.24–8.15
(m, 2H, Py-H and Ph-H), 7.87 (d, J=7.6 Hz, 1H, Ph-H), 7.64 (d, J=
7.3 Hz, 1H, Ph-H), 7.45–7.36 (m, 4H, Py-H and Pxz-H), 7.35–7.28 (m,
4H, B-Ph-N), 7.24 (d, J=8.2 Hz, 2H, Pxz-H), 7.10–6.99 (m, 6H, B-Ph-
N and Pxz-H), 6.62–6.41 (m, 9H, Pxz-H), 5.83 ppm (d, J=7.8, 1.7 Hz,
2H, Pxz-H); 13C NMR (100 MHz, CD2Cl2): d=158.8, 144.3, 141.7,
136.6, 135.6, 135.4, 135.2, 131.7, 130.8, 129.6, 126.8, 123.6, 122.9,
122.4, 121.3, 120.5, 119.1, 115.4, 113.8 ppm; 11B NMR (128 MHz,
CD2Cl2): d=3.4 ppm; HRMS (EI): m/z calcd for C47H32BN3O2 [M]+:
681.2613; found: 681.2619; elemental analysis calcd (%) for
C47H32BN3O2: C 82.82; H 4.73; N 6.17; found: C 82.74; H 4.91; N
6.09.
Synthesis of Boron Chelates 1’–4’
Grignard reagents were prepared as described above (2.1 mmol)
and subsequently cooled to room temperature. Next, (PhPy)BBr2
[11]
(0.33 g, 1.0 mmol) was added under a stream of N2 and the solu-
tion was heated to reflux overnight. The completed reaction was
diluted with CH2Cl2 (30 mL), passed through a short plug of Celite,
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[1] a) T. Noda, Y. Shirota, J. Am. Chem. Soc. 1998, 120, 9714–9715; b) T.
Noda, H. Ogawa, Y. Shirota, Adv. Mater. 1999, 11, 283–285; c) Y. Shirota,
J. Mater. Chem. 2000, 10, 1–25; d) Y. Shirota, H. Kageyama, Chem. Rev.
2007, 107, 953–1010; e) X. Yin, J. Chen, R. A. Lalancette, T. B. Marder, F.
Jꢃkle, Angew. Chem. Int. Ed. 2014, 53, 9761–9765; Angew. Chem. 2014,
126, 9919–9923; f) T. Kushida, A. Shuto, M. Yoshio, T. Kato, S. Yamagu-
chi, Angew. Chem. Int. Ed. 2015, 54, 6922–6925; Angew. Chem. 2015,
127, 7026–7029; g) Z. Zhang, R. M. Edkins, J. Nitsch, K. Fucke, A. Steffan,
L. E. Longobardi, D. W. Stephan, C. Lambert, T. B. Marder, Chem. Sci.
2015, 6, 308–321; h) Z. M. Hudson, S. Wang, Acc. Chem. Res. 2009, 42,
1584–1596; i) C. R. Wade, A. E. J. Broomsgrove, S. Aldridge, F. P. Gabba¨ı,
Chem. Rev. 2010, 110, 3958–3984; j) F. Jꢃkle, Chem. Rev. 2010, 110,
3985–4022; k) A. Wakamiya, S. Yamaguchi, Bull. Chem. Soc. Jpn. 2015,
88, 1357–1377.
Computational studies
DFT and TD-DFT calculations were performed using the Gaussi-
an 09 suite of programs[25] on the High Performance Computing
Virtual Laboratory (HPCVL) at Queen’s University. Input coordinates
of 2 and 2’ were taken from their X-ray crystal structures and all re-
maining compounds were built from the optimized geometries of
2 and 2’. Geometry optimizations and vertical excitations of all
compounds were obtained at the M06-2X[26]/6-31g(d)[27] level of
theory and the resulting structures were confirmed to be station-
ary points through vibrational frequency analysis.
Single-crystal X-ray diffraction
[2] a) D. McArthur, C. P. Butts, D. M. Lindsay, Chem. Commun. 2011, 47,
6650–6652; b) V. Bagutski, A. D. Grosso, J. A. Carrillo, I. A. Cade, M. D.
Helm, J. R. Lawson, P. J. Singleton, S. A. Solomon, T. Marcelli, M. J. Ingle-
son, J. Am. Chem. Soc. 2013, 135, 474–487; c) T. Stahl, K. Muther, Y.
Ohki, K. Tatsumi, M. Oestreich, J. Am. Chem. Soc. 2013, 135, 10978–
10981; d) P. Eisenberger, B. P. Bestvater, E. C. Keske, C. M. Crudden,
Angew. Chem. Int. Ed. 2015, 54, 2467–2471; Angew. Chem. 2015, 127,
2497–2501; e) D. W. Stephan, G. Erker, Angew. Chem. Int. Ed. 2015, 54,
6400–6441; Angew. Chem. 2015, 127, 6498–6541 and references there-
in.
[3] a) Y. Li, Y. Liu, W. Bu, J. Guo, Y. Wang, Chem. Commun. 2000, 1551–
1552; b) S. Wang, Coord. Chem. Rev. 2001, 215, 79–98; c) Y. Cui, Q. D.
Liu, D. R. Bai, W. L. Jia, Y. Tao, S. Wang, Inorg. Chem. 2005, 44, 601–609;
d) H. Y. Chen, Y. Chi, C. S. Liu, J. K. Yu, Y. M. Cheng, K. S. Chen, P. T. Chou,
S. M. Peng, G. H. Lee, A. J. Carty, S. J. Yeh, C. T. Chen, Adv. Funct. Mater.
2005, 15, 567–574; e) A. Loudet, K. Burgess, Chem. Rev. 2007, 107,
4891–4932; f) M. Santra, H. Moon, M. H. Park, T. W. Lee, Y. K. Kim, K. H.
Ahn, Chem. Eur. J. 2012, 18, 9886–9893; g) D. Li, H. Zhang, Y. Wang,
Chem. Soc. Rev. 2013, 42, 8416–8433.
Pale-yellow needles suitable for single-crystal X-ray diffraction were
grown by slow evaporation of a concentrated dichloromethane so-
lution of 2 or 2’. The crystals (0.01ꢂ0.01ꢂ0.10 mm3) were mounted
on a glass fiber and diffraction data were collected on a Bruker
Apex II single-crystal X-ray diffractometer with graphite-monochro-
mated MoKa radiation operating at 50 kV and 30 mA (T=180 K).
Data were processed using the Bruker SHELXTL software package
(version 6.10)[28] and corrected for absorption effects. The crystals
of 2 and 2’ belong to the triclinic space group P and all nonhy-
ꢀ1
drogen atoms were refined anisotropically. CCDC 1479414 and
1479413 for 2 and 2’, respectively, contain the supplementary
crystallographic data for this paper. These data are provided free
of charge by The Cambridge Crystallographic Data Centre.
OLED device fabrication
[4] a) C. D. Entwistle, T. B. Marder, Angew. Chem. Int. Ed. 2002, 41, 2927–
2931; Angew. Chem. 2002, 114, 3051–3056 and references therein;
b) M. Zhu, L. Jiang, M. Yuan, X. Liu, C. Ouyang, H. Zheng, X. Yin, Z. Zuo,
H. Liu, Y. Li, J. Polym. Sci. Part A 2008, 46, 7401–7410; c) J. M. Hales, S.
Barlow, H. Kim, S. Mukhopadhyay, J. L. Bredas, J. W. Perry, S. R. Marder,
Chem. Mater. 2014, 26, 549–560.
[5] a) Y. L. Chang, Y. L. Rao, S. Gong, G. L. Ingram, S. Wang, Z. H. Lu, Adv.
Mater. 2014, 26, 6729–6733; b) C. T. Poon, D. Wu, W. H. Lam, V. W. W.
Yam, Angew. Chem. Int. Ed. 2015, 54, 10569–10573; Angew. Chem.
2015, 127, 10715–10719; c) C. T. Poon, D. Wu, V. W. W. Yam, Angew.
Chem. Int. Ed. 2016, 55, 3647–3651; Angew. Chem. 2016, 128, 3711–
3715.
[6] Y. L. Rao, H. Amarne, S. B. Zhao, T. M. McCormick, S. Martic, Y. Sun, R. Y.
Wang, S. Wang, J. Am. Chem. Soc. 2008, 130, 12898–12900.
[7] a) H. Amarne, C. Baik, S. K. Murphy, S. Wang, Chem. Eur. J. 2010, 16,
4750–4761; b) H. Amarne, C. Baik, R. Y. Wang, S. Wang, Organometallics
2011, 30, 665–668; c) Y. L. Rao, S. Wang, Inorg. Chem. 2011, 50, 12263–
12274.
[8] a) Y. L. Rao, H. Amarne, S. Wang, Coord. Chem. Rev. 2012, 256, 759–770;
b) Y. L. Rao, L. D. Chen, N. J. Mosey, S. Wang, J. Am. Chem. Soc. 2012,
134, 11026–11034; c) Y. L. Rao, H. Amarne, L. D. Chen, M. L. Brown, N. J.
Mosey, S. Wang, J. Am. Chem. Soc. 2013, 135, 3407–3410; d) Y. L. Rao, T.
Kusamoto, R. Sakamoto, H. Nishihara, S. Wang, Organometallics 2014,
33, 1787–1793; e) Y. L. Rao, C. Hçrl, H. Braunschweig, S. Wang, Angew.
Chem. Int. Ed. 2014, 53, 9086–9089; Angew. Chem. 2014, 126, 9232–
9236.
OLED devices were fabricated in a Kurt J. Lesker LUMINOS cluster
tool with a base pressure of 10ꢀ8 Torr without breaking vacuum.
The indium tin oxide (ITO) anode was commercially patterned and
coated on glass substrates 50ꢂ50 mm2 with a sheet resistance less
than 15 Wsqꢀ1. Substrates were ultrasonically cleaned with a stan-
dard regimen of Alconox, acetone, and methanol followed by UV–
ozone treatment for 15 min. The active area for all devices was
2 mm2. The film thicknesses were monitored by a calibrated quartz
crystal microbalance and were further verified for single-carrier de-
vices using capacitance–voltage measurements (Agilent 4294A).
Current–voltage (I–V) characteristics were measured using
a HP4140B picoammeter in ambient air. Luminance measurements
and EL spectra were taken using a Minolta LS-110 luminance meter
and an Ocean Optics USB200 spectrometer with bare fiber, respec-
tively. The external quantum efficiency of EL devices was calculated
following the standard procedure.[29] After deposition, single carrier
devices were transferred to a homebuilt variable temperature cryo-
stat for measurement at 298 K.
Acknowledgements
We thank the Natural Sciences and Engineering Research
Counsil of Canada for financial support, the High Performance
Computing Virtual Laboratory at Queen’s for computational fa-
cilities, and X. Wang for aid with X-ray crystal data collection. S.
K. Mellerup thanks the Government of Canada for the Vanier
CGS award.
[9] a) E. T. Seo, R. F. Nelson, J. M. Fritsch, L. S. Marcoux, D. W. Leedy, R. N.
Adams, J. Am. Chem. Soc. 1966, 88, 3498–3503; b) R. Reynolds, L. L.
Line, R. F. Nelson, J. Am. Chem. Soc. 1974, 96, 1087–1092; c) K. Y. Chiu,
T. X. Su, J. H. Li, T. H. Lin, G. S. Liou, S. H. Cheng, J. Electroanal. Chem.
2005, 575, 95–101; d) J. F. Ambrose, R. F. Nelson, J. Electrochem. Soc.
1968, 115, 1159–1164; e) F. Cao, J. Kim, A. J. Bard, J. Am. Chem. Soc.
2014, 136, 18163–18169; f) J. Kulys, R. Vidziunaite, R. Janciene, A. Palai-
ma, Electroanalysis 2006, 18, 1771–1777.
[10] Examples of boron-based TADF emitters: a) Y. J. Shiu, Y. C. Cheng, W. L.
Tsai, C. C. Wu, C. T. Chao, C. W. Lu, Y. Chi, Y. T. Chen, S. H. Liu, P. T. Chou,
Angew. Chem. Int. Ed. 2016, 55, 3017–3021; Angew. Chem. 2016, 128,
3069–3073; b) K. Suzuki, S. Kubo, K. Shizu, T. Fukushima, A. Wakamiya,
Keywords: boron
photoluminescence · TADF
·
OLED
·
photochromism
·
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Y. Murata, C. Adachi, H. Kaji, Angew. Chem. Int. Ed. 2015, 54, 15231–
15235; Angew. Chem. 2015, 127, 15446–15450; c) T. Hatakeyama, K.
Shiren, K. Nakajima, S. Nomura, S. Nakatsuka, K. Kinoshita, J. Ni, Y. Ono,
T. Ikuta, Adv. Mater. 2016, 28, 2777–2781; d) M. Numata, T. Yasuda, C.
Adachi, Chem. Commun. 2015, 51, 9443–9446; e) Y. Kitamoto, T. Nami-
kawa, D. Ikemizu, Y. Miyata, T. Suzuki, H. Kita, T. Sato, S. Oi, J. Mater.
Chem. C 2015, 3, 9122–9130; f) I. S. Park, M. Numata, C. Adachi, T.
Yasuda, Bull. Chem. Soc. Jpn. 2016, 89, 375–377.
[18] S. Wang, D. T. Yang, J. Lu, H. Shimogawa, S. Gong, X. Wang, S. K. Meller-
up, A. Wakamiya, Y. L. Chang, C. Yang, Z. H. Lu, Angew. Chem. Int. Ed.
2015, 54, 15074–15078; Angew. Chem. 2015, 127, 15289–15293.
[19] C. Gu, N. Huang, J. Gao, F. Xu, Y. Xu, D. Jiang, Angew. Chem. Int. Ed.
2014, 53, 4850–4855; Angew. Chem. 2014, 126, 4950–4955.
[20] H. Tan, C. Pan, G. Wang, Y. Wu, Y. Zhang, G. Yu, M. Zhang, Dyes Pigm.
2014, 101, 67–73.
[21] R. Stahl, C. Lambert, C. Kaiser, R. Wortmann, R. Jakober, Chem. Eur. J.
2006, 12, 2358–2370.
[22] S. Ladouceur, D. Fortin, E. Zysman-Colman, Inorg. Chem. 2010, 49,
5625–5641.
[23] A. M. Brouwer, Pure Appl. Chem. 2011, 83, 2213–2228.
[24] V. V. Pavlishchuk, A. W. Addison, Inorg. Chim. Acta 2000, 298, 97–102.
[25] Gaussian 09 Revision C.01, M. J. Frisch et al., 2010 (see Supporting Infor-
mation for full reference).
[26] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215–241.
[27] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724–
728; b) A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980, 72, 5639–
5648.
[11] N. Ishida, T. Moriya, T. Goya, M. Murakami, J. Org. Chem. 2010, 75, 8709–
8712.
[12] a) D. L. Crossley, A. Cade, E. R. Clark, A. Escande, M. J. Humphries, S. M.
King, I. Vitorica-Yrezabal, M. J. Ingleson, M. L. Turner, Chem. Sci. 2015, 6,
5144–5151; b) A. C. Shaikh, D. S. Ranade, S. Thorat, A. Maity, P. P. Kulkar-
ni, R. G. Gonnade, P. Munshi, N. T. Patil, Chem. Commun. 2015, 51,
16115–16118; c) D. L. Crossley, J. Cid, L. D. Curless, M. L. Turner, M. J. In-
gleson, Organometallics 2015, 34, 5767–5774; d) M. Yusuf, K. Liu, F.
Guo, R. A. Lalancette, F. Jꢃkle, Dalton Trans. 2016, 45, 4580–4587.
[13] D. Abdallah, J. Whelan, J. M. Dust, S. Hoz, E. Buncel, J. Phys. Chem. A
2009, 113, 6640–6647.
[14] Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 2003, 103, 3899–
4031.
[15] a) M. Bixon, J. Jortner, J. Chem. Phys. 1968, 48, 715–726; b) Q. Zhang, H.
Kuwabara, W. J. Potscavage, Jr., S. Huang, Y. Hatae, T. Shibata, C. Adachi,
J. Am. Chem. Soc. 2014, 136, 18070–18081.
[28] SHELXTL, version 6.14, Bruker AXS, Madison, WI, 2000–2003.
[29] S. R. Forrest, D. D. C. Bradley, M. E. Thompson, Adv. Mater. 2003, 15,
1043–1048.
[16] C. Baik, Z. M. Hudson, H. Amarne, S. Wang, J. Am. Chem. Soc. 2009, 131,
14549–14559.
Received: May 19, 2016
[17] Z. M. Hudson, S. B. Ko, S. Yamaguchi, S. Wang, Org. Lett. 2012, 14,
5610–5613.
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Inorganic Chemistry
S. K. Mellerup, K. Yuan, C. Nguyen,
Z.-H. Lu, S. Wang*
&& – &&
Donor-Appended N,C-Chelate
Organoboron Compounds: Influence
of Donor Strength on Photochromic
Behaviour
Let there be chelates! A series of
nescence and distinct photoreactivity,
depending on the donor strength of
the amino group and the steric effect of
the aryl unit (see figure).
amino-functionalized N,C-chelate orga-
noboron compounds have been found
to display bright charge-transfer lumi-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!