Tunable fluorophores based on 2-(N-Arylimino)pyrrolyl chelates of diphenylboron: Synthesis, structure, photophysical characterization, and application in OLEDs

Article abstract of DOI:10.1002/chem.201303607

Reactions of 2-(N-arylimino)pyrroles (HNC₄H₃C(H)=N- Ar) with triphenylboron (BPh₃) in boiling toluene afford the respective highly emissive N,N-boron chelate complexes, [BPh₂{κ ²N,N-NC₄H₃C(H)=N-Ar}] (Ar=C₆H ₅ (12), 2,6-Me₂-C₆H₃ (13), 2,6-iPr₂-C₆H₃ (14), 4-OMe-C₆H ₄ (15), 3,4-Me₂-C₆H₃ (16), 4-F-C₆H₄ (17), 4-NO₂-C₆H₄ (18), 4-CN-C₆H₄ (19), 3,4,5-F₃-C ₆H₂ (20), and C₆F₅ (21)) in moderate to high yields. The photophysical properties of these new boron complexes largely depend on the substituents present on the aryl rings of their N-arylimino moieties. The complexes bearing electron-withdrawing aniline substituents 17-20 show more intense (e.g., φ_f=0.71 for Ar=4-CN-C₆H₄ (19) in THF), higher-energy (blue) fluorescent emission compared to those bearing electron-donating substituents, for which the emission is redshifted at the expense of lower quantum yields (φ_f=0.13 and 0.14 for Ar=4-OMe-C₆H₄ (15) and 3,4-Me₂-C₆H₃ (16), respectively, in THF). The presence of substituents bulkier than a hydrogen atom at the 2,6-positions of the aryl groups strongly restricts rotation of this moiety towards coplanarity with the iminopyrrolyl ligand framework, inducing a shift in the emission to the violet region (λ_max=410-465-nm) and a significant decrease in quantum yield (φ_f=0.005, 0.023, and 0.20 for Ar=2,6-Me₂-C₆H₃ (13), 2,6-iPr ₂-C₆H₃ (14), and C₆F₅ (21), respectively, in THF), even when electron-withdrawing groups are also present. Density functional theory (DFT) and time-dependent DFT (TD-DFT) calculations have indicated that the excited singlet state has a planar aryliminopyrrolyl ligand, except when prevented by steric hindrance (ortho substituents). Calculated absorption maxima reproduce the experimental values, but the error is higher for the emission wavelengths. Organic light-emitting diodes (OLEDs) have been fabricated with the new boron complexes, with luminances of the order of 3000-cdm^-2 being achieved for a green-emitting device. Fluorescent N,N-boron chelate complexes: Mononuclear boron complexes of 2-(N-arylimino)pyrrolyl emit violet to bluish-green colors in solution (see figure, ITO=indium tin oxide, PEDOT:PSS=poly(3,4- ethylenedioxythiophene):poly(styrene sulfonic acid)), depending on the substituents on the N-aryl group. Organic light-emitting diodes have been successfully fabricated with the new boron complexes, achieving luminances of the order of 3000-cdm^-2.

Full text of DOI:10.1002/chem.201303607

DOI: 10.1002/chem.201303607
Full Paper
&
Fluorescence
Tunable Fluorophores Based on 2-(N-Arylimino)pyrrolyl Chelates
of Diphenylboron: Synthesis, Structure, Photophysical
Characterization, and Application in OLEDs
D. Suresh,[a] Patrꢀcia S. Lopes,[a] Bruno Ferreira,[a] Clꢁudia A. Figueira,[a] Clara S. B. Gomes,[a]
Pedro T. Gomes,*[a] Roberto E. Di Paolo,[a] Antꢂnio L. MaÅanita,[a] M. Teresa Duarte,[a]
Ana Charas,[b] Jorge Morgado,[b, c] and Maria Josꢃ Calhorda[d]
Abstract: Reactions of 2-(N-arylimino)pyrroles (HNC4H3C(H)= a hydrogen atom at the 2,6-positions of the aryl groups
N-Ar) with triphenylboron (BPh3) in boiling toluene afford
the respective highly emissive N,N’-boron chelate complexes,
[BPh2{k2N,N’-NC4H3C(H)=N-Ar}] (Ar=C6H5 (12), 2,6-Me2-C6H3
(13), 2,6-iPr2-C6H3 (14), 4-OMe-C6H4 (15), 3,4-Me2-C6H3 (16),
4-F-C6H4 (17), 4-NO2-C6H4 (18), 4-CN-C6H4 (19), 3,4,5-F3-C6H2
(20), and C6F5 (21)) in moderate to high yields. The photo-
physical properties of these new boron complexes largely
depend on the substituents present on the aryl rings of
their N-arylimino moieties. The complexes bearing electron-
withdrawing aniline substituents 17–20 show more intense
strongly restricts rotation of this moiety towards coplanarity
with the iminopyrrolyl ligand framework, inducing a shift in
the emission to the violet region (lmax =410–465 nm) and
a significant decrease in quantum yield (f =0.005, 0.023,
f
and 0.20 for Ar=2,6-Me2-C6H3 (13), 2,6-iPr2-C6H3 (14), and
C6F5 (21), respectively, in THF), even when electron-with-
drawing groups are also present. Density functional theory
(DFT) and time-dependent DFT (TD-DFT) calculations have
indicated that the excited singlet state has a planar arylimi-
nopyrrolyl ligand, except when prevented by steric hin-
drance (ortho substituents). Calculated absorption maxima
reproduce the experimental values, but the error is higher
for the emission wavelengths. Organic light-emitting diodes
(OLEDs) have been fabricated with the new boron com-
plexes, with luminances of the order of 3000 cdmꢀ2 being
achieved for a green-emitting device.
(e.g., f =0.71 for Ar=4-CN-C6H4 (19) in THF), higher-energy
f
(blue) fluorescent emission compared to those bearing elec-
tron-donating substituents, for which the emission is red-
shifted at the expense of lower quantum yields (f =0.13
f
and 0.14 for Ar=4-OMe-C6H4 (15) and 3,4-Me2-C6H3 (16), re-
spectively, in THF). The presence of substituents bulkier than
Introduction
in order to prepare organometallic complexes that might be
applied as pre-catalysts for various organic reactions.[1] Biden-
tate 2-iminopyrrolyl chelating ligands, containing a pyrrolyl
anionic ring and a neutral imine as donor moieties, are one
such type of ligands,[2] which can be considered to be structur-
ally similar to salicylaldiminate, anilidoimine, or 2-(2-pyridyl)in-
dolyl ligands.[3] Although the first syntheses of homoleptic
metal complexes of CoII, NiII, PdII, CuII, and ZnII containing li-
gands of this kind were reported back in the 1960s,[4] there has
been a recent resurgence of interest in such systems.[5] The
ease of steric and electronic tuning through the introduction
of aryl substituents on the imine group makes these chelates
highly flexible in terms of ligand design, which renders them
very interesting for various applications. For instance, bulky
aryl substituents on the imine group provide steric tuning and
shielding of the metal centers and hence several mono- or bis-
[2-(N-arylimino)pyrrolyl] early and late transition-metal com-
plexes have been reported in the literature as efficient a-olefin
oligo-/polymerization catalysts.[2,6]
There has been considerable interest in the synthesis of biden-
tate chelating ligands containing an imine and a donor moiety,
[a] Dr. D. Suresh, P. S. Lopes, B. Ferreira, C. A. Figueira, Dr. C. S. B. Gomes,
Dr. P. T. Gomes, Dr. R. E. Di Paolo, Prof. A. L. MaÅanita, Dr. M. T. Duarte
Centro de Quꢀmica Estrutural
Departamento de Engenharia Quꢀmica
Instituto Superior Tꢁcnico, Universidade de Lisboa
Av. Rovisco Pais, 1049-001 Lisboa (Portugal)
E-mail: pedro.t.gomes@tecnico.ulisboa.pt
[b] Dr. A. Charas, Dr. J. Morgado
Instituto de TelecomunicaÅꢂes
Av. Rovisco Pais, 1049-001 Lisboa (Portugal)
[c] Dr. J. Morgado
Department of Bioengineering, Instituto Superior Tꢁcnico
Universidade de Lisboa
Av. Rovisco Pais, 1049-001 Lisboa (Portugal)
[d] Prof. M. J. Calhorda
Departamento de Quꢀmica e Bioquꢀmica, CQB
Faculdade de CiÞncias, Universidade de Lisboa
Campo Grande, Ed. C8, 1749-016 Lisboa (Portugal)
Besides catalytic investigation, another important focus of
interest in recent years has been the synthesis of coordination
and organometallic compounds with photoluminescent (PL)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303607.
Chem. Eur. J. 2014, 20, 1 – 16
1
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
3
properties, due to their potential applications in electrolumi-
nescent (EL) displays.[7] Luminescent chelate complexes have
proved to be particularly useful in electroluminescent displays
because of their relatively high stability and volatility.[8] Much
effort has been devoted to the design and synthesis of new
phosphorescent materials with heavy metals such as iridium-
(III)[9] and platinum(II).[10] Recently, the synthesis of fluorescent
tricoordinate[11] and tetracoordinate[12] boron complexes has at-
tracted a great deal of attention because of their potential ap-
plications in functional materials and in sensors.[13] Earlier re-
ports on boron-based complexes suggested that variations in
the chromophore part of the molecule would influence the
HOMO–LUMO levels and thereby the color of emission.[12a,d] In
addition, some of these compounds exhibited good charge-
transport properties and, consequently, could be employed in
the fabrication of organic light-emitting diodes (OLEDs) and
sensors.[14]
states of which could be assigned to mixed intraligand (p–p*)
and metal-to-ligand charge-transfer (MLCT) transitions. Their
phosphorescence quantum yields were, however, very low
(fp <3%). Experimental and theoretical results indicated that
the photophysical properties of these complexes largely de-
pended on the N,N or N,O anionic ancillary ligands.
Very recently, in a preliminary communication, we reported
the synthesis and characterization of new fluorescent four-co-
ordinate mono- and binuclear organoboron complexes con-
taining 2-(N-arylformimino)pyrrolyl bidentate ligands (D and E,
respectively) and their successful application in single-layer,
non-doped OLEDs.[17] In particular, the mononuclear boron
complex [BPh2{k2N,N’-NC4H3C(H)=N-Ph}] D (denoted as 12 in
the present work; see Scheme 2 below) exhibited blue fluores-
cence (lmemax =479 nm) characterized by a quantum yield of
34%, this emission being assigned to a virtually pure ligand-
1
based (p–p*) transition.[18]
Recently, we have developed new homoleptic 2-(N-arylformi-
mino)pyrrolyl zinc(II) complexes of the type [Zn{k2N,N’-2-(N-
arylformimino)pyrrolyl}2] (A1 and A2) or [Zn{k2N,N’-2-(N-
arylformimino)phenanthro[9,10-c]pyrrolyl}2] (A3 and A4), the
latter containing a phenanthrene fragment fused on the
These results prompted us to synthesize new boron com-
plexes containing 2-(N-arylformimino)pyrrolyl anionic ligands
and, in order to assess their emissivity and color tunability, to
study in detail their photoluminescent properties under the in-
fluence of different structural parameters of the N-aryl ring. In
the present paper, we report a series of highly emissive mono-
nuclear 2-(N-arylformimino)pyrrolyl chelates of diphenylboron,
[BPh2{k2N,N’-NC4H3C(H)=N-Ar}], and their characterization by
multinuclear NMR, X-ray diffraction analysis, and cyclic voltam-
metry. Photophysical characterization of the newly synthesized
complexes was accomplished by steady-state and time-re-
solved luminescence techniques in solution. Density functional
theory (DFT) and time-dependent DFT (TD-DFT) calculations
(ADF program) were also carried out on these new boron com-
plexes, to assign their electronic transitions and to determine
their singlet excited-state properties (geometry, lifetime, emis-
sion). Furthermore, these compounds have been utilized as
emissive layers in OLED devices.
Results and Discussion
Synthesis and characterization of iminopyrrole ligand
precursors
Most of the ligand precursors were synthesized and character-
ized according to literature methods (Scheme 1).[5d,19–24] The
new ligand precursors 6 and 10 were synthesized, and their
characterization data are presented in the Experimental Sec-
tion. In general, the 2-(N-arylformimino)pyrrole ligand precur-
C3ꢀC4 bond of the pyrrolyl ring of the ligand, and studied
their photoluminescent properties.[15] The two phenanthrene-
fused iminopyrrolyl zinc complexes showed mainly a ligand-
1
based (p–p*) emission in the blue/green spectral region, but
their fluorescence efficiencies were low (f =3.9 and 8.8%, re-
f
spectively). Eisenberg and co-workers reported a series of neu-
tral phosphorescent copper(I) complexes of the types [Cu(N,N)-
(P,P)] and [Cu(N,O)(P,P)] containing phosphine or diphosphine
ligands and coordinated bidentate N,N ancillary ligands such
as 2-(N-phenylformimino)pyrrolyl or 2-(N-phenylformimino)in-
dolyl (B), or the corresponding parent bidentate N,O ancillary
ligands 2-formylpyrrolyl or 2-formylindolyl, respectively (C).[16]
These complexes exhibited long-lived emissions, the excited
Scheme 1. Synthesis of 2-(N-arylformimino)pyrrole ligand precursors 2–11.
p-TSA=para-toluenesulfonic acid.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
2
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1
sors (HNC4H3C(H)=N-Ar), with Ar=C6H5 (2), 2,6-Me2-C6H3 (3),
2,6-iPr2-C6H3 (4), 4-OMe-C6H4 (5), 3,4-Me2-C6H3 (6), 4-F-C6H4 (7),
4-NO2-C6H4 (8), 4-CN-C6H4 (9), 3,4,5-F3-C6H2 (10), and C6F5 (11),
were synthesized by condensation reactions of 2-formylpyrrole
(1) with the requisite substituted aryl amines, namely aniline,
2,6-dimethylaniline, 2,6-diisopropylaniline, 4-methoxyaniline,
3,4-dimethylaniline, 4-fluoroaniline, 4-nitroaniline, 4-aminoben-
zonitrile, 3,4,5-trifluoroaniline, and 2,3,4,5,6-pentafluoroaniline,
employing standard reaction conditions (Scheme 1). The syn-
theses and purifications of these ligand precursors were
straightforward, and those with sterically bulky and electron-
donating substituents were typically obtained in good yields.
However, the ligand precursors bearing electron-withdrawing
groups were always obtained with unreacted starting amines
and thus had to be purified by vacuum sublimation, leading to
low to moderate yields. All compounds were characterized by
1H, 13C, and, in the case of fluorinated derivatives, 19F NMR
spectroscopies, their spectra being consistent with those re-
ported in the literature.
and C6F5 (21), were characterized by H, 13C, 11B, and, in the
case of fluorinated compounds, 19F NMR spectroscopies, as
well as by elemental analysis. Where possible, single-crystal
X-ray diffraction analysis was also performed. Although the or-
ganoboron compound [BPh2{k2N,N’-NC4H3C(H)=N-C6H5}] (12)
has already been reported in our earlier communication,[17] it is
also included in this full paper. The absence of NH proton sig-
nals in the 1H NMR spectra of the complexes confirmed the for-
mation of the neutral tetracoordinate boron mononuclear
complexes. The 11B NMR chemical shifts of these compounds
are in the range d=4.88–5.81 ppm, consistent with those of
other four-coordinate boron derivatives[12] (resonances for
three-coordinate boron are typically downfield shifted, appear-
ing at dꢁ +25 ppm and above[11]). In general, these iminopyr-
rolyl organoboron compounds could be handled in air for peri-
ods of at least an hour, but, upon prolonged exposure, under-
went a relatively slow decomposition process leading to regen-
eration of the corresponding free ligand precursor. However,
compound 21 proved to be extremely sensitive to air and
moisture (in the solid state or in solution) as well as very unsta-
ble in solution even under nitrogen, undergoing very fast de-
composition with concomitant regeneration of the ligand pre-
cursor. This instability, most likely due to the high electronega-
tivity of the C6F5 group, could be monitored by UV/Vis absorp-
tion spectroscopy and affected the electrochemical measure-
ments. Compounds 13–21 were found to be soluble in most
of the common organic solvents, such as dichloromethane,
tetrahydrofuran, and toluene, and moderately soluble in
n-hexane.
Crystals suitable for single-crystal X-ray diffraction analysis
were obtained for the ligand precursors 3–6 and 8. The struc-
tural features of these compounds are similar to those of other
previously reported iminopyrrole molecules.[5d,e,h,15,19] In fact,
the crystal structure of compound 4 has been reported pre-
viously.[5h] However, a new polymorphic form was obtained in
the present work when the compound was crystallized from
toluene double-layered with n-hexane. Perspective views of
the molecular structures of the ligand precursors 3–6 and 8
are given in Figures S1–S5 in the Supporting Information. Se-
lected bond lengths and angles are given in the captions to
these figures. Details of the crystal structure determinations
are given in Table S1 in the Supporting Information.
Crystals suitable for X-ray diffraction analysis were obtained
for compounds 13–17, 19, and 21, enabling the determination
of their crystal structures. Perspective views of the molecular
structures of these iminopyrrolyl boron complexes are shown
in Figure 1 (and Figure S6 in the Supporting Information) and
selected bond distances and angles are given in Table 1.
In general, the boron center in each of the complexes dis-
plays a typical distorted tetrahedral geometry with C1 symme-
try. The iminopyrrolyl ligand is chelated to the boron atom
through the two nitrogen atoms to form a five-membered
ring. The respective bite angles (N1-B1-N2) range from
93.81(17)8 to 95.30(8)8. Throughout the series [BPh2{k2N,N’-
NC4H3C(H)=N-Ar}] (13–17, 19, and 21), the BꢀC and BꢀN bond
distances are rather invariant, with average B1ꢀCipso and
B1ꢀN1(pyrrolyl) distances of 1.610 and 1.561 ꢅ, respectively,
the latter being shorter by 0.073 ꢅ compared to the average
B1ꢀN2(imine) distance (1.634 ꢅ), as previously observed for
compound 12.[17]
Synthesis and characterization of iminopyrrolyl boron
complexes
The organoboron compounds 13–21 (Scheme 2) were ob-
tained in good to moderate yields by reactions of the corre-
sponding iminopyrrole ligand precursors with triphenylboron.
Furthermore, the iminopyrrolyl fragments coordinated to
boron are virtually planar in all of the compounds, but the re-
spective aryl groups bonded to the iminic nitrogen are not co-
planar. The dihedral angle (defined as C6-N2-C7-C12) depends
on the bulkiness of the substituents at 2,6-positions of the N-
aryl groups. For instance, within the group of complexes bear-
ing 2,6-substituted aryl groups (13, 14, and 21, with Me, iPr
and F substituents, respectively), the N-aryl plane is nearly per-
pendicular to that of the iminopyrrolyl fragment in 13 and 14,
as reflected in very high dihedral angles (81.6(3)8 for 13,
Scheme 2. Synthesis of iminopyrrolyl-BPh2 complexes 12–21.
The mixtures were heated overnight in refluxing toluene under
nitrogen atmosphere. The iminopyrrolyl organoboron com-
pounds [BPh2{k2N,N’-NC4H3C(H)=N-Ar}], with Ar=2,6-Me2-C6H3
(13), 2,6-iPr2-C6H3 (14), 4-OMe-C6H4 (15), 3,4-Me2-C6H3 (16), 4-F-
C6H4 (17), 4-NO2-C6H4 (18), 4-CN-C6H4 (19), 3,4,5-F3-C6H2 (20),
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
3
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 1. Perspective views of the molecular structures of iminopyrrolyl boron complexes 13–17, 19, and 21. Ellipsoids are drawn at the 30% probability
level. All calculated hydrogen atoms have been omitted for clarity.
Photoluminescence properties
75.3(6)8 for 14). A dihedral angle of 70.3(3)8 is observed for 21,
owing to the lower steric volume of the F atom compared to
the Me and iPr groups. In the remaining complexes 15–17 and
19, in which H atoms occupy the 2,6-positions of the N-aryl
ring, the dihedral angles are considerably smaller (41.82(19)8
for 15, 39.84(16)8 for 16, 45.7(2)8 for 17, and 27.4(2)8 for 19).
The other structural features of these boron complexes resem-
ble those of similar analogues[12a,d,17,19] and will be further dis-
cussed below, when comparing them with those of excited
states.
The UV/Vis absorption and fluorescence spectra of complexes
12–21 in THF are shown in Figure 2 (and Figure S8 in the Sup-
porting Information) and relevant photophysical data, listed in
order of increasing absorption wavelength maxima, are sum-
marized in Table 3. In contrast to their non-emissive ligand pre-
cursors 2–11, most of the boron complexes are highly emissive
in solution.
The absorption spectra show wavelength maxima (corre-
!
sponding to the S1,1 S0,0 vibronic transition) within the range
353–421 nm in THF, which depend on the steric and electronic
properties of the N-aryl substituents. The absorption spectra of
the 2,6-substituted compounds (13, 14, and 21) are considera-
bly blue-shifted with respect to that of the unsubstituted
parent compound (12). This feature is due to the large torsion
angle between the N-aryl ring and the iminopyrrolyl fragment
(>708, see the last row of Table 1), which prevents p-conjuga-
tion (resonance) between the two fragments. On the other
hand, substitution at the 3-, 4-, or 5-positions of the N-aryl ring
leads to red-shifts of the absorption maxima, the magnitudes
of which also depend on the electron-donating or electron-
withdrawing properties of these substituents. The correlation
of the absorption energy maxima with IPꢀEA (Table 2 and Fig-
ure S9a in the Supporting Information), which assumes propor-
tionality between the S0-S1 and HOMO–LUMO energy gaps, is
surprisingly good (except for compound 21), given the ab-
sence of CIS (configuration interaction of states),[27] which is re-
quired for conversion of orbital energies into state energies.
The fluorescence emission maxima (412–505 nm) correspond
to the S1,0!S0,1 vibronic transition, except in the case of the
Electrochemical properties
Cyclic voltammetry measurements on compounds 12–21 were
performed in electrolyte solutions in dichloromethane. The re-
spective onset oxidation and reduction potentials were used,
after conversion to the absolute scale taking the Fc/Fc+ (ferro-
cene/ferrocenium ion redox couple) as a reference, to estimate
their ionization potentials (IP) and electron affinities (EA).[25]
The values obtained are summarized in Table 2.
With the exception of compounds 13, 14, and 21, bearing
particularly twisted aryl groups owing to restricted rotation
about the N-aryl bond, the ionization potentials are clearly cor-
related with the electron-donating or electron-withdrawing
properties of the substituents, as represented by the sum of
their classical Hammett s constants[26] (Table 2 and Figure S7a
in the Supporting Information). The electron affinities also
show a correlation, although that of the strongly conjugated
nitro substituent (compound 18) falls outside the relevant
trend line (Figure S7b in the Supporting Information).
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
4
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
is, a rather planar ground-state
geometry, in agreement with the
lowest computationally calculat-
ed dihedral angle between the
iminopyrrolyl and aryl moieties
in this complex (25.078, see
Figure 3). Accordingly, com-
pound 19, which has the second
lowest dihedral angle (25.88),
shows the second best vibration-
ally resolved emission (Figure 2).
As observed for their absorp-
tions, the fluorescence of the
Table 1. Selected bond distances [ꢅ] and angles [8] for complexes 13–17, 19, and 21.
13[a]
14[a]
15
16
17
19
21[a]
Distances
N1ꢀC2
N1ꢀC5
N2ꢀC6
N2ꢀC7
N1ꢀB1
1.376(3)
1.347(3)
1.297(3)
1.439(3)
1.561(4)
1.645(4)
1.594(4)
1.615(4)
1.364(6)
1.356(6)
1.299(6)
1.452(6)
1.541(7)
1.625(7)
1.605(7)
1.591(8)
1.375(2)
1.347(2)
1.304(2)
1.424(2)
1.566(2)
1.634(2)
1.618(2)
1.604(2)
1.371(2)
1.407(2)
1.3749(15)
1.3465(15)
1.3083(14)
1.4301(14)
1.5660(15)
1.6306(16)
1.6206(18)
1.6189(16)
1.5061(17)
1.3794(19)
1.3467(19)
1.3172(18)
1.4285(18)
1.565(2)
1.633(2)
1.621(2)
1.608(2)
1.3593(17)
1.381(2)
1.337(2)
1.319(2)
1.426(2)
1.569(3)
1.630(2)
1.603(3)
1.613(3)
1.439(3)
1.145(2)
1.381(3)
1.343(3)
1.313(3)
1.418(3)
1.561(3)
1.644(3)
1.603(4)
1.611(4)
1.343(3)
N2ꢀB1
Caryl1ꢀB1
Caryl2ꢀB1
[b]
XꢀCN-aryl-p
[b,c]
YꢀXꢀCN-aryl-p
[d]
XꢀCN-aryl-m
1.5047(17)
1.342(3)
1.339(3)
1.332(3)
1.347(3)
violet-emitting
2,6-substituted
[e]
XꢀCN-aryl-o
1.501(4)
1.496(4)
1.508(7)
1.508(7)
N-aryl sterically hindered com-
pounds 13, 14, and 21 is sub-
stantially shifted to shorter
wavelengths with respect to that
of 12, while that of the remain-
ing blue to blue-greenish-
emitting compounds is shifted
to longer wavelengths. Again,
some correlation with the
LUMO-HOMO energy gap is ob-
served (Table 2 and Figure S9b
in the Supporting Information),
except in the cases of com-
pounds 18 and 21.
Angles
C2-N1-C5
C2-N1-B1
C5-N1-B1
C6-N2-C7
C6-N2-B1
C7-N2-B1
N1-B1-N2
107.2(2)
112.4(2)
139.2(2)
120.6(2)
110.7(2)
128.6(2)
94.69(19)
116.0(2)
113.8(2)
109.0(2)
111.3(2)
110.0(2)
81.6(3)
107.1(5)
112.7(4)
139.1(5)
119.7(4)
110.9(4)
128.4(4)
94.9(4)
114.7(5)
108.5(4)
115.8(5)
111.6(4)
109.7(4)
75.3(6)
107.88(13)
112.14(12)
139.68(14)
121.64(13)
111.03(12)
126.86(12)
94.90(11)
117.82(13)
112.91(12)
110.28(12)
107.25(12)
111.26(11)
41.82(19)
107.46(9)
112.00(9)
140.45(10)
121.03(9)
110.52(9)
128.28(9)
95.30(8)
112.91(9)
111.61(9)
110.72(9)
110.07(9)
114.99(9)
39.84(16)
107.78(12)
112.51(12)
139.50(13)
121.43(12)
110.87(12)
127.51(11)
94.90(11)
115.76(12)
112.07(12)
109.96(12)
109.59(12)
112.70(12)
45.7(2)
107.70(15)
112.20(14)
140.05(16)
122.30(15)
110.76(14)
126.89(14)
95.35(13)
118.88(16)
110.45(15)
108.82(15)
111.02(14)
110.02(14)
27.4(2)
107.7(2)
113.00(18)
139.3(2)
122.52(19)
111.30(18)
124.78(18)
93.81(17)
119.2(2)
111.77(19)
112.2(2)
110.81(17)
105.93(18)
70.3(3)
Caryl1-B1-Caryl2
N1-B1-Caryl1
N1-B1-Caryl2
N2-B1-Caryl1
N2-B1-Caryl2
C6-N2-C7-C12[f]
[a] Due to the similarity of the parameters for the molecules in the asymmetric unit, only bond distances and
angles for molecule A are presented. [b] X=O (15), C (16, 19), F (17, 21). [c] Y=C (15), F (17). [d] X=C (16), F
(17). [e] X=C (13, 14), F (17). [f] For molecule B, the dihedral angle C6-N2-C7-C12 is 80.9(3)8 in 13, 79.4(6)8 in
14, and 75.1(3)8 in 21.
The fluorescence quantum
yields and lifetimes are dictated
by the non-radiative rate con-
stants (knr) (the variation of the
radiative rate constant kf among
all of the compounds is relatively
small). The 2,6-substituted compounds 13 and 14 show the
Table 2. Ionization potentials (IP) and electron affinities (EA) of com-
plexes 12–21, estimated from cyclic voltammetry measurements, derived
largest values of knr (ca. 1010 ꢀ1) and, consequently, very low
s
IPꢀEA values, energies of the absorption maximum (Eambasx) and of the first
fluorescence quantum yields (0.005 and 0.023). Conversely, the
highest quantum yields are observed for 18, 19, and 20, which
show the lowest knr values (ca. 108 sꢀ1). In order to clarify the
origin of these differences in the knr values, triplet formation
quantum yields fT were measured for 13, 12, and 19, as com-
pounds representative of high, medium, and low values of knr,
respectively (Table 4).
0ꢀ0
vibronic transition (E ), and the sum of Hammett s constants of the cor-
em
responding N-aryl substituents.
max
abs
0ꢀ0
[a]
Cmpd N-aryl ring
IP
EA
IPꢀEA
E
E
so +sm +sp
em
substituents [eV] [eV] [eV]
[eV] [eV]
12
13
14
15
16
17
18
19
20
21[b]
none
5.64 2.82 2.82
5.82 2.56 3.26
5.83 2.56 3.27
5.45 2.74 2.71
5.52 2.74 2.78
5.66 2.86 2.80
5.83 3.43 2.40
5.78 3.06 2.72
5.79 2.99 2.80
5.58 3.01 2.57
3.24 2.75
0
2,6-Me2
2,6-iPr2
4-OMe
3,4-Me2
4-F
4-NO2
4-CN
3,4,5-F3
2,3,4,5,6-F5
3.51 3.20 ꢀ0.34
3.51 3.21 ꢀ0.46
3.21 2.65 ꢀ0.29
3.23 2.72 ꢀ0.02
For the 2,6-dimethyl-substituted compound 13, no triplet
formation was found, indicating that exclusively internal con-
version is responsible for the non-radiative decay of the singlet
excited state S1 of this compound. The exceptionally high
value of kic is also responsible for the abnormally low fluores-
cence quantum yield and lifetime (36 ps) of 13 within this
series of compounds. For the parent (unsubstituted) com-
pound 12, fT is also low (ca. 1%) and internal conversion is
similarly the dominant mechanism of non-radiative decay.
Finally, the cyano-substituted compound 19 has a tenfold
lower kic relative to 12, most likely due to the extended p-con-
jugation of the cyanophenyl and iminopyrrolyl moieties, which
may increase the rigidity of the complex in the S1 excited
state.
3.25 2.77
2.95 2.67
3.07 2.74
3.18 2.79
3.38 2.87
0.05
0.78
0.66
0.73
1.23
[a] Sum of the Hammett constants (ref. [26]). [b] Compound unstable in
solution.
nitro-substituted compound 18, for which the maximum corre-
sponds to the S1,0!S0,0 transition. The higher intensity of the
S1,0!S0,0 transition of 18 indicates a smaller change in the
equilibrium geometries of its ground and excited states, that
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
5
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
of the boron compounds 12–21. The geometries of the
ground state and the first excited singlet state (obtained by
promoting one electron from the HOMO to the LUMO) were
fully optimized. The BP86 functional and a triple-z basis set, as
used for geometry optimization of complex 12 (only hydrogen
atoms on the arene ring) and others in a previous study,[18]
were also employed here in order to compare the present re-
sults with previously obtained data relating to other boron
compounds. Time-dependent DFT (TD-DFT)[30] was used to
obtain the absorption spectra and the lifetimes of the first ex-
cited singlet states (same functional and basis set). Some calcu-
lations were performed in the gas phase and others also in
THF, as specified in the text below.
Molecular geometries
The optimized geometries of the ground states and the first
excited singlet states are shown in Figure 3 (and Figure S12 in
the Supporting Information), emphasizing the dihedral angle
between the iminopyrrolyl and aryl moieties. The agreement
between the calculated and experimental dihedral angles of
the ground-state compounds is generally very good, consider-
ing that rotation about the NꢀC bond is essentially unhindered
(Figure 3). The best agreements are observed for 19 (CN), with
values of 27.48 and 25.88, and 14, with values of 75.38 and
74.28. In 15, 16, and 17, the deviations range from about 10 to
158. No experimental structures could be obtained for 18 and
20. The largest deviations are observed in 13 (Me2) and 21 (F5),
and can probably be attributed to dimeric association of the
molecules through weak CꢀH···p hydrogen bonds in the corre-
sponding crystal structures. The aryl ring is distorted in order
to maximize hydrogen bonding. This interaction energy was
not calculated, since it is outside the scope of this work. For
the compound with the bulkiest ortho substituents, 14 (iPr2),
a perfect agreement between the calculated and experimental
dihedral angles was observed, despite also existing as a dimer.
Based on the good agreement between the experimental and
calculated structural parameters, we trust this methodology to
provide reliable estimates of the geometrical features of com-
pounds 18 and 20, for which no X-ray structures are available.
Some trends are evident in these structures, namely that the
dihedral angle is always higher than 258 in the ground-state
structures and drops to close to zero (planar ligand) in the sin-
glet excited states of all of the compounds, except those with
two ortho substituents (13 (Me2), 14 (iPr2), 21 (F5)). As men-
tioned above, the reason for this is the steric hindrance of the
substituents, which prevents the attainment of planarity in this
state, and is reflected in the higher angles of 41.678 in 14
(iPr2), 35.788 in 13 (Me2), and 22.048 in 21 (F5).
Figure 2. Normalized a) absorption and b) emission spectra of complexes
12–21 in THF.
It is interesting to note that, except for the compounds
bearing aryl groups with substituents at the 2,6-positions, 13,
14, and 21, general trends are observed for the emissions of
complexes 12–21: the use of increasingly electron-withdrawing
N-aryl substituents shifts the emission wavelength maxima to
higher energies and increases the fluorescence quantum yields
(see Figures S10a and S10b in the Supporting Information, re-
spectively). These electronic effects, along with the steric effect
of 2,6-substitution, enable color tuning of this molecular
system in the range from violet to blue to bluish-green (see
Figure S11 in the Supporting Information).
Interestingly, these geometrical features seem to correlate
well with the low experimental fluorescence quantum yields f
f
(Table 3). In particular, the nonplanarity of the ligand in the sin-
glet excited state of the hindered 2,6-substituted N-aryl deriva-
tives may be associated with very low f values of 0.005 for 13
(Me2) and 0.023 for 14 (iPr2). Compounds 15–20 (and also
12[18]), with virtually planar ligands in their singlet excited
f
Density functional calculations
DFT calculations[28] (ADF[29] program) were performed in order
to understand the influence of substituents on the properties
states, display much higher values of f. The ligand in the sin-
f
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
6
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
glet state of compound 21 is far from planar (22.048),
although less so than in 13 and 14 (35.88 and 41.78,
respectively), but its fluorescence quantum yield is
higher than those of other compounds, such as 15
and 16. In this compound, the steric effect is accom-
panied by the electronic effect of the five electroneg-
ative fluorine atoms, and it cannot be analyzed sepa-
rately.
Table 3. Wavelength maximum (lambasx) and molar extinction coefficient (emax) of the
first absorption band, wavelength maximum (lmemax) and wavelength of the first vibron-
ic transition (l0emꢀ0) of the emission band, fluorescence quantum yield (f), fluorescence
f
lifetime (tf), radiative rate constant (kf), and sum of non-radiative rate constants (knr) of
boron complexes 12–21 in THF at 293 K.
max
abs
max
em
0ꢀ0[b]
[e]
Cmpd N-aryl ring Ssi[a]
l
emax
l
l
f
f
tf[c]
kf[d]
knr
em
substituents
[nm] [104 Lmolꢀ1 [nm] [nm]
[ns] [nsꢀ1] [nsꢀ1
]
cmꢀ1
]
The energies and three-dimensional representa-
tions (Molekel[31]) of the HOMOs and LUMOs of all of
the compounds (determined in THF) are shown in
Figure 4 (and Figure S13 in the Supporting Informa-
tion). Both the HOMO and the LUMO of all of the
compounds are essentially located on the iminopyr-
rolyl ligand, the contribution of the aryl moiety being
negligible in 14 (iPr2) and 13 (Me2), as well as in the
LUMO of 21 (F5).
14
2,6-iPr2
2,6-Me2
2,3,4,5,6-F5
4-F
ꢀ0.46 353 1.7
412 386 0.023 0.13 0.18 7.52
413 388 0.005 0.036 0.13 27.6
462 432 0.20 1.04 0.19 0.77
478 448 0.25 1.53 0.16 0.49
479 451 0.34 1.90 0.18 0.35
492 456 0.14 1.22 0.11 0.70
505 468 0.13 1.07 0.12 0.81
474 445 0.47 2.72 0.17 0.19
482 453 0.71 3.28 0.22 0.09
466 464 0.50 1.83 0.27 0.27
13
ꢀ0.34 353 1.7
1.22 367 2.1
0.06 381 1.9
21[f]
17
12[g] none
0
383 1.7
16
15
20
19
18
3,4-Me2
ꢀ0.24 384 2.0
ꢀ0.27 386 2.1
0.74 390 2.0
0.67 404 2.2
0.78 421 2.0
4-OMe
3,4,5-F3
4-CN
4-NO2
[a] Ssi =so +sm +sp (ref. [26]). [b] From spectral decomposition with sums of four
Gaussians. [c] From single exponential decays. [d] kf =f/tf. [e] knr =(1ꢀf)/tf. [f] Com-
f
f
pound unstable in solution. [g] Ref. [17].
Electronic state transitions
The absorption maxima were calculated using the
TD-DFT approach described above, both in the gas
Figure 3. Optimized geometries of the iminopyrrolyl boron complexes 13–21 in the ground state and in the first excited singlet state, showing the dihedral
angles (8).
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
7
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
duce very well the experimental values for most of the com-
pounds, with deviations between 0 and 0.11 eV, exceptions
being 14 (0.25 eV), 15 (0.25 eV), and 18 (0.35 eV). The SO
values show better agreement for compound 15 (0.04 eV), but
greatly overestimate the energy of the absorption maxima by
a mean deviation of 0.29 eV for all the other compounds. The
introduction of the solvent (THF) effect in the calculation of ab-
sorption energies also led to large deviations from the experi-
mental values, depending on the nature of the substituents.
With electronegative substituents, the calculated energies
were lower than the experimental values, the largest devia-
tions being observed for compounds 18 (NO2) and 15 (OMe)
(0.76 and 0.44 eV, respectively), followed by the fluorinated de-
rivatives (17, 20, 21) with smaller deviations (0.23, 0.16,
0.20 eV). For the other species, the shifts were in the opposite
direction and smaller (0.1–0.2 eV). In summary, the gas-phase
values were, in general, closer to the experimental ones.
The low-energy transition was 90% HOMO (H) to LUMO (L)
for compound 12 (H), both orbitals being delocalized over the
aryliminopyrrolyl moiety.[18] Two other compounds, 13 and 19,
were studied in more detail.
Table 4. Fluorescence quantum yield (f) and lifetime (tf), triplet forma-
f
tion quantum yield (fT) and lifetime (tT), and radiative (kf), internal con-
version (kic), and intersystem crossing (kisc) rate constants of boron com-
plexes 13, 12, and 19 in THF at 293 K.
[a]
[b]
Cmpd N-aryl ring
f
f
tf
fT tT
kf
knr
kisc
kic
substituents
[ns]
[ms] [nsꢀ1] [nsꢀ1] [nsꢀ1] [nsꢀ1
]
13
12
19
2,6-Me2
none
4-CN
0.005 0.036 0
0.34 1.9 0.01 1.87 0.18
0.71 3.28 0.2 89.5 0.22
–
0.13 27.6 0.00 27.60
0.35 0.01
0.09 0.06
0.34
0.03
[a] kisc =fT/tf. [b] kic =knrꢀkisc
.
In 13 (2,6-Me2), the low-energy band has contributions from
three transitions: H-5!L (98%); H-7!L (49%)+H!L (40%);
H-6!L (71%)+H!L (19%). The HOMO and LUMO are local-
ized on the iminopyrrolyl part of the ligand, while H-5 and H-7
are essentially on the aryl ring and H-6 has some contribution
from boron (see Figure 5, as well as Figure S14 in the Support-
ing Information). Thus, the band is characterized by significant
intraligand charge transfer from the aryl to the iminopyrrolyl
moiety, which may contribute to the measured low fluores-
cence yield (Table 3).
Figure 4. The energies (eV) and three-dimensional representations of the
HOMOs and LUMOs (in THF) of the iminopyrrolyl boron complexes 13–21.
phase (GP) and in THF, since the solvent often influences the
absorption features. This methodology, however, does not
allow the calculation of the radiative lifetime of the singlet ex-
cited state (1/kf), which requires the introduction of spin-orbit
coupling. It can be obtained from a scalar relativistic TD-DFT
approach, which uses a perturbative method including the in-
fluence of spin-orbit coupling, SOPERT (SO),[32] and has been
shown to yield better results when using a hybrid PBE0 func-
tional in conjunction with an all-electron basis set (TZ2P was
still used). Having performed this calculation, we used other
available information on absorption spectra for comparison
with our standard approach.
The calculated absorption energy maxima are given along
with the experimental values in Table 5. The GP values repro-
Table 5. Calculated (GP, THF, and SO) and experimental energy maxima
[eV] of the lowest energy band.
Figure 5. The nature of the transitions leading to the lower-energy absorp-
tion bands for 13 (2,6-Me2) (left) and 19 (CN) (right).
max
abs
max
abs
max
abs
Cmpd
N-aryl ring
Eambasx(exp)
E
(GP)
E
(THF)
E
(SO)
substituents
[eV]
[eV]
[eV]
[eV]
In 19 (CN), the low-energy band has contributions from two
transitions: H!L (84%)+H-5!L (5%); H-5!L (88%)+H!L
(6%). The HOMO and LUMO are localized on the cyanoarylimi-
nopyrrolyl ligand, while H-5 is almost completely on the pyrrol-
yl moiety, with a small contribution from boron (Figure 5). In
this compound, the intraligand contribution of the pyrrolyl
group is dominant, even compared to the unsubstituted com-
pound 12, which may be responsible for the increased fluores-
cence yield (Tables 3 and 4).
14
13
21
17
12
16
15
20
19
18
2,6-iPr2
2,6-Me2
2,3,4,5,6-F5
4-F
3.51
3.51
3.38
3.25
3.24
3.23
3.21
3.18
3.07
2.95
3.76
3.62
3.38
3.25
3.29
3.16
2.96
3.23
3.14
2.60
3.79
3.75
3.18
3.02
3.34
3.17
2.77
3.02
3.16
2.19
3.92
3.86
3.70
3.47
3.48
3.41
3.25
3.47
3.40
3.24
none
3,4-Me2
4-OMe
3,4,5-F3
4-CN
4-NO2
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
8
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The emission wavelength was calculated from the energy
difference between the excited singlet state and the ground
state with the same geometry, both in the gas phase (GP) and
in THF. The geometry was optimized using the standard ap-
proach in ADF (BP86 functional) as described above. Repro-
ducing emission energies is difficult, as has been reported by
other authors. Since new and (in principle) better methodolo-
gies have become available for optimizing the excited states
with TD-DFT, combining several excitations and affording the
emission energy directly, we thought that it would be interest-
ing to compare them with the established approach involving
only the HOMO to LUMO excitation. The radiative lifetimes
(1/kf) of the singlet excited states were obtained from SO cal-
culations, as mentioned above, allowing calculation of the kf
values. All of the results are collected in Table 6 and compared
with the experimental values (emission maxima).
agreement is perfect for compounds 12 and 13. For the com-
pounds with p-donating substituents, such as F (17, 20) or
OMe (15), the calculated fluorescence rate constants are very
close to the experimental values, whereas for those bearing p-
accepting groups such as cyanide (19) or nitro (18), the agree-
ment is poor. Compound 21, with five electronegative F sub-
stituents, two of them in ortho positions, represents the largest
challenge for the calculation method, and the calculated excit-
ed singlet-state lifetime was the furthest away from the experi-
mental value.
Electroluminescence properties
Light-emitting diodes, LEDs, with the simple ITO/PEDOT:PSS/
complex (12–21)/Ca/Al structure, were firstly prepared, with
films of the emissive complexes being deposited by spin-coat-
ing inside a glove box. The relative positions of the
frontier energy levels of the components used in the
max
Table 6. Calculated (GP, THF, and TD-DFT) and experimental energy maxima (E ) of
the emission bands and calculated (SO) and experimental fluorescence rate constants
(kf) of boron complexes 12–21.
various LED structures are shown in Figure S17 in the
Supporting Information. The emissions obtained from
LEDs based on the complexes with the lowest solu-
tion PL quantum yields, 13 and 14 (containing 2,6-
substituted N-aryl rings), were very dim. We could
not record their electroluminescence (EL) spectra. In
addition, LEDs based on the complexes with strongly
electron-accepting groups (18, 20, and 21) also
showed very low emission intensities, in spite of their
em
max
em
max
em
max
em
max
em
Cmpd N-aryl ring
E
(exp)
E
(GP)
E
(THF)
E
(TD-DFT) kf (exp) kf (SO)
substituents [eV]
[eV]
[eV]
[eV]
[nsꢀ1
]
[nsꢀ1
]
14
13
21
17
12
16
15
20
19
18
2,6-iPr2
2,6-Me2
2,3,4,5,6-F5
4-F
3.01
3.00
2.68
2.59
2.59
2.52
2.46
2.62
2.57
2.66
2.12
1.97
2.05
1.86
1.89
1.83
1.76
1.88
1.87
2.02
2.17
2.02
2.09
1.92
1.96
1.88
1.79
1.95
1.95
1.93
nc[a]
3.02
nc[a]
3.00
2.94
3.01
3.25
3.23
2.73
nc[a]
0.18
0.13
0.19
0.16
0.18
0.11
0.12
0.17
0.22
0.27
0.01
0.13
0.004
0.22
0.18
0.25
0.19
0.14
0.08
0.11
none
moderate solution fluorescence quantum yields (f).
3,4-Me2
4-OMe
3,4,5-F3
4-CN
f
Among these, it is worth mentioning the results ob-
tained for 18, which showed a solution quantum
yield of 50%, yet the emission intensity of its LED
was the lowest among the series. It is also interesting
to note that for these complexes there is no elec-
tron-injection barrier with respect to the calcium
cathode, pointing to an “excess” electron injection
4-NO2
[a] Not converged.
The GP energy values are systematically much higher for all
of the compounds, which is not surprising since they rely on
a single HOMO–LUMO excitation. The solvent correction (THF
values) shifts all calculated values closer to the experimental
ones, but only by a small amount. For instance, the calculated
emission of compound 13 (Me2) is increased from 1.97 to
2.02 eV, but is still too far from the experimental value of
3.00 eV. On the other hand, the TD-DFT values reproduce the
observed energies more closely, almost perfectly for 13, but
with deviations of 0.16 eV for 19 (CN) and 0.35–0.79 eV for the
other compounds. Although these calculations require much
longer times, the inclusion of several excitations leads to
a good agreement between calculated and experimental emis-
sions.
with respect to the hole injection from PEDOT:PSS. It is also
possible that there may be significant PL quenching in the
solid state compared to solution, contributing to the very poor
device performance. However, LEDs based on complexes 15,
16, 17, and 19 showed significant emission intensities, with 16
exhibiting a maximum luminance of 614 cdmꢀ2 with a peak EL
efficiency of 0.2 cdAꢀ1 (see Table 7). Although 16 showed only
a modest solution photoluminescence (PL) quantum yield (f)
f
of 14% (see Table 3), it was the complex that performed best
in this simple LED structure. From Figure S17 (Supporting Infor-
mation), we conclude that, among these four complexes, 15
and 16 have similar electron- and hole-injection barriers. The
difference in LED emission efficiencies may possibly result from
differences in charge transport, assuming that these complexes
The calculation of emission energies is still a challenge and
therefore several computational approaches must be tested.
Owing to their limitations, it is not yet possible to give
a simple picture of emission that has a reliable predictive capa-
bility.
have similar f values in the solid state and in solution, with 16
exhibiting more equilibrated hole and electron mobilities.
f
Complexes 15–20, those in this new series that showed the
highest solution f values, were further tested in LEDs, for
f
which they were deposited by sublimation on top of either PE-
DOT:PSS or N,N’-bis(3-methylphenyl)-N,N’-diphenylbenzidine
(TPD)-coated PEDOT:PSS. Barium was used as cathode material,
protected with an overlayer of Al. Among the ITO/PEDOT:PSS/
The excited singlet-state lifetimes are given as fluorescence
rate constants (kf). The values are generally close to the experi-
mental ones, with small deviations in both directions. The
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
9
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
tron-hole recombination. In fact,
we observed a significant im-
provement in performance of
the LEDs based on these six
complexes (15–20), although the
devices based on 18 and 20 still
showed very poor performances.
Among the other LEDs, those
based on 16 showed the best
performance, as observed for
the other device types, with
a maximum luminance of nearly
3000 cdmꢀ2 and a maximum EL
efficiency of 1.4 cdAꢀ1. In fact,
the devices based on complexes
15 and 16, with electron-donat-
ing groups on the N-aryl moiety,
displayed the highest luminan-
ces, in spite of the modest solu-
Table 7. Maximum luminance (Lmax) in [cdmꢀ2], luminous EL efficiency (fEL,max) in [cdAꢀ1], and external quan-
tum efficiency (EQEmax) in [%] of LEDs based on complexes 12–21. Films were prepared by spin-coating (in the
case of the Ca-based LEDs) or by sublimation.
Complex N-aryl ring
ITO/PEDOT:PSS/complex/ ITO/PEDOT:PSS/com-
ITO/PEDOT:PSS/TPD/com-
plex/Ba (sublimed)
substituents
Ca (spin-coating)
plex/Ba (sublimed)
12
13
14
15
16
17
18
19
20
21
none
Lmax
0.35
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
fEL,max 3.8ꢆ10ꢀ4
EQEmax 1.5ꢆ10ꢀ4
2,6-Me2
2,6-iPr2
4-OMe
3,4-Me2
4-F
Lmax
0.096
fEL,max 5.6ꢆ10ꢀ6
EQEmax 6.6ꢆ10ꢀ6
Lmax
0.085
fEL,max 3.56ꢆ10ꢀ6
EQEmax 1.92ꢆ10ꢀ6
–
22
–
500
Lmax
fEL,max 0.036
EQEmax 0.013
Lmax
fEL,max 0.2
EQEmax 0.079
Lmax
fEL,max 0.005
31
0.006
0.002
755
0.31
0.12
53
0.03
0.01
0.28
4.8ꢆ10ꢀ4
1.6ꢆ10ꢀ4
80
0.072
0.027
0.26
2.09ꢆ10ꢀ5
8.2ꢆ10ꢀ6
–
0.16
0.058
2965
1.4
0.48
384
0.22
0.087
2.1
2.85ꢆ10ꢀ3
9.9ꢆ10ꢀ4
246
0.089
0.036
1.4
8.5ꢆ10ꢀ5
3.4ꢆ10ꢀ5
–
614
43
tion f values (13% and 14%, re-
f
EQEmax 0.002
spectively) of these complexes.
4-NO2
4-CN
Lmax
0.066
fEL,max 2.36ꢆ10ꢀ5
LEDs based on 19, the complex
EQEmax 8.2ꢆ10ꢀ6
showing the highest solution f
f
Lmax
–
–
–
(71%), were not the best per-
forming. Figure 6 (and Fig-
ure S15 in the Supporting Infor-
mation) compares the perform-
ances of LEDs based on 16 and
19, having barium cathodes, and
with or without an HT/EBL made
of TPD. In both cases, the inser-
tion of TPD led to an increase in
fEL,max
EQEmax
Lmax
fEL,max 0.002
EQEmax 0.001
3,4,5-F3
0.012
2,3,4,5,6-F5 Lmax
0.28
fEL,max 6.4ꢆ10ꢀ5
–
–
–
–
EQEmax 3.3ꢆ10ꢀ5
sublimed complex/Ba devices, those based on 16 showed the
best performance, as found in the Ca-based LEDs, with a maxi-
mum luminance of 755 cdmꢀ2 and a maximum EL efficiency of
0.31 cdAꢀ1. As observed for the first series of LEDs, those
based on complexes 18 and 20 showed the poorest perform-
ances. The use of Ba in place of Ca should improve the elec-
tron-injection efficiency in the case of complexes 15, 16, and
possibly also 17 (albeit to a lesser extent), maintaining the
ohmic electron injection for the remaining complexes of the
studied series. We attribute the observed differences in per-
formance for the devices based on complexes 18–20 to the
different film preparation processes (sublimation vs solution
spin-coating). We consider that the variation in the performan-
ces of the devices based on complexes 15–17 is inconsistent
with reduction of the electron-injection barrier upon the use of
Ba (which should intrinsically improve the EL efficiency, but
was not realized for 15). Again, we believe that the change in
the film preparation process is the main contributing factor.
The insertion of a TPD hole-transporting/electron-blocking
layer (HT/EBL), with a HOMO lying slightly lower in energy
than the PEDOT:PSS workfunction, should improve hole injec-
tion for all of the complexes by creating an intermediate level
between the PEDOT:PSS workfunction and their HOMO levels.
In addition, its high-lying LUMO prevents electrons from escap-
ing to the PEDOT:PSS electrode, thereby improving the elec-
the maximum luminance and also enhanced the EL efficiency
(in the case of 16, it led to an increase in the maximum EL effi-
ciency from 0.31 to 1.4 cdAꢀ1, whereas in the case of 19 it led
to a much more modest increase from 0.07 to 0.09 cdAꢀ1). It is
also interesting to note that while the solution PL efficiency of
Figure 6. Current (I, filled symbols) and luminance (L, open symbols) for
LEDs based on 16 (film thickness 80 nm) and 19 (film thickness 60 nm). Re-
sults are shown for ITO/PEDOT:PSS/complex/Ba/Al and ITO/PEDOT:PSS/TPD/
complex/Ba/Al device structures.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
10
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
19 was higher than that of 16 (71% versus 14%), the reverse
was true for the EL efficiencies. This behavior could be due to
differences in PL efficiency in the solid state (with detrimental
stronger dipole interactions in the case of 19), higher hole-in-
jection barrier, due to a higher ionization potential of 19 with
respect to that of PEDOT:PSS (5.2 eV), and possibly also differ-
ences in hole and/or electron mobilities.
LED characteristics of which are shown in Figure 6) and 18,
which bears a strongly electron-withdrawing group (NO2) and,
in spite of its reasonable solution PL efficiency of 50%, yielded
poorly performing LEDs. For complex 16, the solution and
thin-film PL spectra are similar (with a red-shift of the emission
maximum from 490 to 505 nm on going from solution to
a thin film), with the EL spectrum having a maximum at an in-
termediate wavelength (Figure 7A). The solid-state PL spec-
trum of complex 19 shows a significantly red-shifted maximum
(516 nm) with respect to the solution PL (maximum at
482 nm), and a stronger tail at longer wavelengths (Figure 7B).
Surprisingly, the EL spectrum is not as shifted toward lower en-
ergies as the PL spectrum. This is somewhat unexpected, as
we would suppose the accessible excited states to be the
same under both photo and electrical excitation. We do not
yet have a definitive explanation for this observation as, in
view of the spectral shapes, interference effects are unlikely to
be the cause. For complex 18, the red-shift of the thin-film PL
spectrum with respect to that obtained from the solution is
even more pronounced, with the maximum increasing from
about 466 nm to 524 nm (Figure 7C). This effect, combined
with the loss of vibronic resolution, indicates that a different
lowest-energy state appears in the solid state, which is most
likely of intermolecular origin. The emission of the 18-based
LEDs was too weak to allow us to record an EL spectrum,
which indicates that, in spite of the high solution PL efficiency
of 50%, the lowest-energy state formed in the solid state has
insufficient emission efficiency.
From the results shown in Table 7, it can be concluded that
the presence of electron-withdrawing substituents on the N-
aryl fragment is generally detrimental to device performance.
Complexes 17 and 19, with fluoro and cyano substituents in
the para position, respectively, show intermediate performan-
ces. With an increase in the number of groups and/or their
electron-withdrawing strengths, the performance decreases
even further.
Figure 7 (and also Figure S16 in the Supporting Information)
compares the PL spectra of three complexes, 16 and 19 (the
Conclusion
Several tetracoordinate mononuclear organoboron complexes
containing 2-(N-arylformimino)pyrrolyl ligands with N-aryl ring
substituents of different electronic and steric natures have
been synthesized and characterized by multiple methods. X-
ray crystallography studies have confirmed chelation of these
ligands to the Lewis acidic diphenylboron moiety. The resulting
complexes have a pseudo-tetrahedral geometry about the
boron atoms, and the dihedral angle between the N-aryl plane
and the iminopyrrolyl fragment depends upon the bulkiness of
the substituents in the 2,6-positions of the aryl groups. It was
found that the electronic properties of these fluorescent com-
pounds could be varied in a predictable manner by changing
the electron-donating or electron-withdrawing character of the
aniline substituents. For instance, the color of emission could
be tuned from blue to bluish-green by increasing the electron-
donating strength of the substituents. This change was at the
expense of the fluorescence quantum yield. DFT calculations
have shown the geometry of the iminopyrrolyl ligand in the
excited singlet state to be almost planar, except when ortho
substituents are present, in contrast to the nonplanar ground-
state structure. The low-energy band leading to emission is es-
sentially a HOMO to LUMO transition. The more delocalized
frontier orbitals seem to be associated with higher fluores-
cence quantum yield, whereas the introduction of some
charge transfer between the ligands has the opposite effect.
DFT methods lead to very good estimates of the absorption
Figure 7. Comparison of emission spectra (PL recorded in solution and from
thin films, and EL of the thin films) for complexes (A) 16, (B) 19, and (C) 18.
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
11
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
whereupon it turned yellow-orange. It was then allowed to cool,
CH2Cl2 was added, and the suspension was filtered through Celite
and washed through with more CH2Cl2. After removal of all vola-
tiles from the filtrate in vacuo, the product was dissolved in reflux-
ing n-hexane. The dark-brown oil was separated from the superna-
tant orange solution, and the latter was stored at ꢀ208C to yield
orange-yellow crystals of the iminopyrrole ligand precursor. For fur-
ther details of the experimental procedure and product characteri-
zation, see the Supporting Information.
maxima, but emission is more difficult to predict, even by
using the more recent methodologies. Single-layer devices,
with a PEDOT:PSS hole-injection layer, have been fabricated
with the prepared complexes as emissive layers, films of which
were prepared either by spin-coating from solution or by subli-
mation. As expected, the complexes that showed the lowest
solution photoluminescence (PL) efficiencies (12, 13, 14, and
21) yielded very poorly emissive devices, and LEDs based on
18, which has a solution PL efficiency of 50%, also showed
very poor performance. The best performing devices were ob-
tained with 16, even though it has only a modest solution PL
efficiency of 14%. Upon insertion of a hole-transporting/elec-
tron-blocking layer of TPD, the performance was greatly im-
proved, with the devices based on 16 showing a maximum lu-
minance of 2965 cdmꢀ2 with a maximum EL efficiency of
General procedure for the syntheses of diphenylboron
derivatives, [BPh2{k2N,N’-NC4H3C(H)=N-Ar}] (13–21)
In a typical experiment, an equimolar mixture of triphenylboron
and the requisite iminopyrrolyl ligand precursor in toluene (15–
25 mL) was stirred and heated at reflux, under nitrogen atmos-
phere, overnight. The reaction mixture was allowed to cool to
room temperature, then concentrated under vacuum to a volume
of 5 mL and double-layered with hexane. The resulting solution
was kept at ꢀ208C to afford the corresponding crystalline boron
complex. For further details of the experimental procedure and
product characterization, see the Supporting Information.
1.4 cdAꢀ1
.
Experimental Section
General procedures
X-ray data collection
All experiments involving air- and/or moisture-sensitive materials
were carried out under inert atmosphere using a dual vacuum/ni-
trogen line and standard Schlenk techniques. Nitrogen gas was
supplied in cylinders (Air Liquide) and purified by passage through
4 ꢅ molecular sieves. Unless otherwise stated, all reagents were
purchased from commercial suppliers (e.g., Acros, Aldrich, Fluka)
and used without further purification. All solvents to be used
under inert atmosphere were thoroughly deoxygenated and dehy-
drated before use. They were dried and purified by refluxing over
a suitable drying agent followed by distillation under nitrogen. The
following drying agents were used: sodium (for toluene, diethyl
ether, and tetrahydrofuran (THF)) and calcium hydride (for
n-hexane and dichloromethane). Solvents and solutions were trans-
ferred using a positive pressure of nitrogen through stainless steel
cannulae and mixtures were filtered in a similar way using modi-
fied cannulae that could be fitted with glass fiber filter disks.
Crystallographic and experimental details of the crystal structure
determinations are listed in Tables S1 and S2 in the Supporting In-
formation. The crystals were selected under an inert atmosphere,
covered with polyfluoroether oil, and mounted on a nylon loop.
Crystallographic data for compounds 3–6, 8, 13–17, 19, and 21
were collected at 150 K using graphite-monochromated MoKa radi-
ation (l=0.71073 ꢅ) on a Bruker AXS-Kappa APEX II diffractometer
equipped with an Oxford Cryosystems open-flow nitrogen cryostat.
Cell parameters were retrieved using Bruker SMART[33] software
and refined using Bruker SAINT[34] on all observed reflections. Ab-
sorption corrections were applied using SADABS.[35] Structure solu-
tion and refinement were performed by direct methods with the
programs SIR2004[36] and SHELXL[37] included in the package of pro-
grams WINGX-Version 1.80.05.[38] Except for the NH hydrogen
atoms in the compounds, all hydrogen atoms were inserted in ide-
alized positions and allowed to refine riding on the parent carbon
atom. Graphic representations were prepared with ORTEP-III.[39]
Data have been deposited with the CCDC under deposition num-
bers CCDC-957737 for 3, 957738 for 4, 957739 for 5, 957740 for 6,
957741 for 8, 957742 for 13, 957743 for 14, 957744 for 15, 957745
for 16, 957746 for 17, 957747 for 19, and 957748 for 21. CCDC-
957737–957748 contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Nuclear magnetic resonance (NMR) spectra were recorded on
a Bruker Avance III 300 (1H, 13C, 11B, and 19F) spectrometer. Deuter-
ated solvents were dried by storage over 4 ꢅ molecular sieves and
degassed by the freeze-pump-thaw method. Spectra were refer-
enced internally to residual protio-solvent (1H) or solvent (13C) reso-
nances and are reported relative to tetramethylsilane (d=0). 11B
and 19F NMR spectra were referenced to Et2O·BF3 and CFCl3, respec-
tively. All chemical shifts are quoted in d (ppm) and coupling con-
stants are given in Hz. Multiplicities are abbreviated as follows:
broad (br), singlet (s), doublet (d), triplet (t), quartet (q), heptet (h),
and multiplet (m). For air- and/or moisture-sensitive materials, sam-
ples were prepared in J. Young NMR tubes in a glovebox. Elemen-
tal analyses were obtained from the IST elemental analysis service.
Cyclic voltammetry studies
Cyclic voltammetry measurements were carried out with a Solartron
potentiostat using a standard three-electrode cell, with a saturated
calomel reference electrode, a platinum wire as counter electrode,
and a platinum disk as working electrode. The compounds were
dissolved in freshly distilled dichloromethane containing 0.2m elec-
trolyte salt (tetrabutylammonium tetrafluoroborate or tetrabutyl-
ammonium perchlorate).
General method for the synthesis of iminopyrrole ligand
precursors
This synthetic procedure was based on other similar methods de-
scribed in the literature.[5d,19–24] In a typical experiment, an equimo-
lar ratio of 2-formylpyrrole and a substituted aniline, a catalytic
amount of p-toluenesulfonic acid, and MgSO4 (to remove any
water from the reaction mixture) were suspended in absolute etha-
nol (20 mL) in a round-bottomed flask fitted with a condenser and
a CaCl2 guard tube. The mixture was heated to reflux overnight,
Ionization potential (IP) and electron affinity (EA) values were esti-
mated from the onset of oxidation and reduction potentials, re-
spectively. To convert the values on the electrochemical scale to an
absolute scale, referred to the vacuum, we used ferrocene as a ref-
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
12
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
benzophenone
TT
sample
DOD
max
erence and considered the energy level of ferrocene/ferrocenium
(Fc/Fc+) to be 4.80 eV below the vacuum level, as detailed in
ref. [17].
e
ꢀsTample
¼
benzophenone ꢀTbenzophenone
ð2Þ
sample
DODmax
e
TT
benzophenone
TT
with e
¼ 7200 mꢀ1 cmꢀ1 and ꢀTbenzophenone ¼ 1:
Spectroscopic measurements
Computational studies
Absorption and fluorescence spectra were measured on a Beckman
DU-70 spectrophotometer and a SPEX Fluorolog 212I, respectively.
The fluorescence spectra were collected with right-angle geometry,
in the S/R mode, and corrected for instrumental wavelength de-
pendence.
Density functional theory calculations[28] were performed using the
Amsterdam density functional program package (ADF).[29] Gradient-
corrected geometry optimizations,[42] without symmetry con-
straints, were performed using the local density approximation of
the correlation energy (Vosko–Wilk–Nusair),[43] and the generalized
gradient approximation (Becke’s[44] exchange and Perdew’s[45] corre-
lation functionals). Relativistic effects were treated with the ZORA
approximation.[46] Unrestricted calculations were performed for ex-
cited singlet states. The core orbitals were frozen for B, C, and N
(1s). Triple-z Slater-type orbitals (STO) were used to describe the
valence shells of B, C, and N (2s and 2p). A set of two polarization
functions was added to B, C, and N (single-z, 3d, 4f). Triple-z
Slater-type orbitals (STO) were used to describe the valence shells
of H (1s), augmented with two polarization functions (single-z 2s,
2p). Time-dependent DFT calculations in the ADF implementation
were performed to determine the excitation energies.[30] The sol-
vent effect was included with the COSMO approach in ADF in
single-point calculations on the optimized geometries. The geome-
try of the excited state was calculated by promoting one electron
from the HOMO to the LUMO with S=0. The perturbative method
in the time-dependent density functional theory (TD-DFT) formal-
ism, with the influence of spin-orbit coupling effect (SOPERT),[32]
was used to calculate the excited-state lifetimes. In these calcula-
tions, complete basis sets were used for all elements (as above,
but without any frozen cores) with the hybrid PBE0 functional.[47]
We checked that the absorption spectra calculated by this ap-
proach were the same as those obtained under the same condi-
tions but without including spin-orbit coupling since all the atoms
are light. TD-DFT optimizations of the first singlet excited state
were also performed, using the Gaussian 09 software,[48] for techni-
cal reasons, with the PBE0 functional[47] and a 6–31G** basis set for
all atoms.[49]
Fluorescence decays were measured using the time-correlated
single-photon counting (TCSPC) technique as previously de-
scribed.[40] The pulsed (82 MHz) excitation source was a Ti:sapphire
Tsunami laser pumped with a Millennia Xs solid-state laser (Spectra
Physics). Decays longer than 2 ns were remeasured at reduced rep-
etition rate (4 MHz) using a pulse picker (Spectra Physics, Model
3980). The Tsunami output (720–900 nm) was frequency-doubled
and vertically polarized. The sample emission was passed through
a polarizer set at the magic angle and a Jobin–Yvon H10 mono-
chromator, and finally detected with a microchannel plate photo-
multiplier (Hamamatsu, R3809u-50 MCP-PT). A fraction of the Tsu-
nami output was detected with a PHD-400-N photodiode (Becker
and Hickl, GmbH) for generation of the start signal. Start and stop
signals were processed with an SPC-630 acquisition board (Becker
and Hickl, GmbH). The instrumental response function (IRF) was
measured using a LUDOX scattering solution in water with trans-
mittance at the excitation wavelength matched to that of the
sample (FWHM=19 ps). The IRF and sample signals were collected
until 5ꢆ103 counts at the maximum were reached. Fluorescence
decays were deconvoluted from the IRF using the modulation
functions method (Sand program).[41]
Flash photolysis
Triplet-state absorption spectra and quantum yields were mea-
sured with a laser flash photolysis apparatus (Applied Photophy-
sics) pumped by an Nd:YAG laser (Spectra Physics). Transient spec-
tra were obtained by monitoring the optical density (OD) change
at intervals of 10 nm over the range 300–700 nm and averaging at
least ten decays at each wavelength. Second-order kinetics was ob-
served for the decay of the lowest triplet state. Excitation was at
355 nm with an unfocused beam. Special care was taken in deter-
mining triplet yields to have optically matched dilute solutions
(absꢁ0.2–0.3 in a 10 mm square cell) and low laser energy (2 mJ)
to avoid multiphoton and triplet-triplet (T-T) annihilation effects.
The triplet molar absorption coefficients were determined by the
singlet depletion technique, according to the well-known relation-
ship:
The structures were modeled on those of compounds 13–17, 19,
and 21 described above. Three-dimensional representations of the
orbitals were obtained with Molekel[31] and electronic spectra with
Chemcraft.[50]
Light-emitting diode studies
Light-emitting diodes were fabricated on glass/ITO substrates
(ITO=indium tin oxide), which were cleaned with detergent, dis-
tilled water, acetone, and isopropanol. They were treated with
oxygen plasma, prior to the deposition of PEDOT:PSS (poly(3,4-eth-
ylenedioxythiophene) doped with polystyrene sulfonic acid,
CLEVIOS P VP.AI 4083 from Heraeus Clevios GmbH) by spin-coat-
ing. The PEDOT:PSS films (40 nm thick, as measured with a DEKTAK
profilometer) were annealed in air for 2 min at 1208C, and then
transferred to a nitrogen-filled glove box. Some devices were pre-
pared with an approximately 20 nm thick, hole-transporting/elec-
tron-blocking layer of TPD (N,N’-bis(3-methylphenyl)-N,N’-diphenyl-
benzidine, from Aldrich), which was thermally deposited on top of
the PEDOT:PSS.
ðeSÞðDODTÞ
ð1Þ
eT ¼
ðDODSÞ
where both DODS and DODT were obtained from the triplet transi-
ent absorption spectra. Triplet formation quantum yields were de-
rived from these and actinometry with benzophenone. The inter-
system-crossing yields (fT) for the compounds were obtained by
comparing the DOD at 530 nm of solutions of benzophenone in
benzene (standard) optically matched (at the laser excitation wave-
length) with that of the compound using the following equation:
Films of the complexes 12–21 were deposited on top of the PE-
DOT:PSS by spin-coating from solutions in THF, inside the glove
box. The thicknesses of the films of the complexes were in the
range 60–100 nm. The substrates were then placed inside an evap-
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
13
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Y. Wang, Chem. Commun. 2000, 1551–1552; d) S. Lamansky, P. Djuro-
vich, D. Murphy, F. Abdel-Razzaq, H. E. Lee, C. Adachi, P. E. Burrows, S. R.
Forrest, M. E. Thompson, J. Am. Chem. Soc. 2001, 123, 4304–4312.
[9] a) J. M. Fernꢁndez-Hernꢁndez, C.-H. Yang, J. I. Beltran, V. Lemaur, F. Polo,
R. Frohlich, J. Cornil, L. De Cola, J. Am. Chem. Soc. 2011, 133, 10543–
10548; b) C.-H. Yang, Y.-M. Cheng, Y. Chi, C.-J. Hsu, F.-C. Fang, K.-T.
Wong, P.-T. Chou, C.-H. Chang, M.-H. Tsai, C.-C. Wu, Angew. Chem. 2007,
119, 2470–2473; Angew. Chem. Int. Ed. 2007, 46, 2418–2421; c) L. He, L.
Duan, J. Qiao, D. Zhang, L. Wang, Y. Qiu, Org. Electron. 2010, 11, 1185–
1191.
oration chamber, and the top metal cathode (calcium), about
40 nm thick, was deposited at a base pressure of 2ꢆ10ꢀ6 mbar
through a shadow mask, defining pixel areas of 4 mm2. A protect-
ing, approximately 60 nm thick, aluminum layer was deposited on
top. Alternatively, films of complexes 15–20 (with thicknesses rang-
ing from 50 to 115 nm) were prepared by thermal deposition on
top of either PEDOT:PSS or TPD-coated PEDOT:PSS. Device struc-
tures were completed by the deposition of barium (ca. 40 nm
thick) protected by an overlayer of aluminum (ca. 60 nm thick).
[10] a) J. Kavitha, S. Y. Chang, Y. Chi, J. K. Yu, Y. H. Hu, P. T. Chou, S. M. Peng,
A. J. Carty, Adv. Funct. Mater. 2005, 15, 223–229; b) X. Yang, Z. Wang, S.
Madakuni, J. Li, G. E. Jabbour, Adv. Mater. 2008, 20, 2405–2409; c) M. A.
Baldo, D. F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thompson, S. R.
Forrest, Nature 1998, 395, 151–154.
Devices were tested under vacuum using a K2400 Source Meter
and a calibrated silicon photodiode, as described previously.[51]
Electroluminescence (EL) spectra were obtained with a CCD spec-
trograph (from Ocean Optics or from ScanSci). External quantum
efficiencies were estimated as detailed in ref. [51].
[11] a) Z. M. Hudson, S. Wang, Acc. Chem. Res. 2009, 42, 1584–1596; b) P.
Chen, R. A. Lalancette, F. Jꢇkle, Angew. Chem. 2012, 124, 8118–8122;
Angew. Chem. Int. Ed. 2012, 51, 7994–7998; c) M. Halik, W. Wenseleers,
C. Grasso, F. Stellacci, E. Zojer, S. Barlow, J. L. Bredas, J. W. Perry, S. R.
Marder, Chem. Commun. 2003, 1490–1491; d) C. D. Entwistle, T. B.
Marder, Chem. Mater. 2004, 16, 4574–4585; e) C. D. Entwistle, T. B.
Marder, Angew. Chem. 2002, 114, 3051–3056; Angew. Chem. Int. Ed.
2002, 41, 2927–2931; f) R. Bernard, D. Cornu, J. P. Scharff, R. Chiriac, P.
Miele, Inorg. Chem. 2006, 45, 8743–8748; g) R. Bernard, D. Cornu, P. L.
Baldeck, J. Caslavsky, J. M. Letoffe, J. P. Scharff, P. Miele, Dalton Trans.
2005, 3065–3071; h) F. Jꢇkle, Chem. Rev. 2010, 110, 3985–4022.
[12] a) Q.-D. Liu, M. S. Mudadu, R. Thummel, Y. Tao, S. Wang, Adv. Funct.
Mater. 2005, 15, 143–154; b) H. Amarne, C. Baik, S. K. Murphy, S. Wang,
Chem. Eur. J. 2010, 16, 4750–4761; c) S. Kappaun, S. Rentenberger, A.
Pogantsch, E. Zojer, K. Mereiter, G. Trimmel, R. Saf, K. C. Mçller, F. Stelzer,
C. Slugovc, Chem. Mater. 2006, 18, 3539–3547; d) B. J. Liddle, R. M.
Silva, T. J. Morin, F. P. Macedo, R. Shukla, S. V. Lindeman, J. R. Gardinier, J.
Org. Chem. 2007, 72, 5637–5646; e) Y. Liu, J. Guo, H. Zhang, Y. Wang,
Angew. Chem. 2002, 114, 190–192; Angew. Chem. Int. Ed. 2002, 41,
182–184; f) Z. Zhang, H. Bi, Y. Zhang, D. Yao, H. Gao, Y. Fan, H. Zhang, Y.
Wang, Z. Chen, D. Ma, Inorg. Chem. 2009, 48, 7230–7236; g) D. Li, Z.
Zhang, S. Zhao, Y. Wang, H. Zhang, Dalton Trans. 2011, 40, 1279–1285;
h) D. Li, Y. Yuan, H. Bi, D. Yao, X. Zhao, W. Tian, Y. Wang, H. Zhang, Inorg.
Chem. 2011, 50, 4825–4831; i) C.-C. Cheng, W.-S. Yu, P.-T. Chou, S.-M.
Peng, G.-H. Lee, P.-C. Wu, Y.-H. Song, Y. Chi, Chem. Commun. 2003,
2628–2629; j) J. Ugolotti, S. Hellstrom, G. J. P. Britovsek, T. S. Jones, P.
Hunt, A. J. P. White, Dalton Trans. 2007, 1425–1432; k) J. Yoshino, A.
Furuta, T. Kambe, H. Itoi, N. Kano, T. Kawashima, Y. Ito, M. Asashima,
Chem. Eur. J. 2010, 16, 5026–5035; l) N. Kano, A. Furuta, T. Kambe, J.
Yoshino, Y. Shibata, T. Kawashima, N. Mizorogi, S. Nagase, Eur. J. Inorg.
Chem. 2012, 1584–1587; m) J. Yoshino, N. Kano, T. Kawashima, Chem.
Commun. 2007, 559–561; n) P. T. Gomes, D. Suresh, C. S. B. Gomes, P. S.
Lopes, C. A. Figueira (IST), World Patent WO2013039413, 2013; o) D.
Suresh, P. T. Gomes in Advances in Organometallic Chemistry and Cataly-
sis: The Silver/Gold Jubilee International Conference on Organometallic
Chemistry Celebratory Book, 1st ed. (Ed.: A. J. L. Pombeiro), John Wiley &
Sons, Hoboken, 2014, Chap. 36, pp. 485-492; p) D. Li, H. Zhang, Y.
Wang, Chem. Soc. Rev. 2013, 42, 8416–8433.
Acknowledgements
We thank the Fundażo para a CiÞncia e Tecnologia, Portugal,
for financial support (Projects PTDC/QUI/65474/2006, PEst-OE/
QUI/UI0100/2013, PEst-OE/QUI/UI0612/2013, and PEst-OE/EEI/
LA0008/2013) and for fellowships to D.S., P.S.L., B.F., C.A.F., and
C.S.B.G. (SFRH/BPD/47853/2008, SFRH/BD/88639/2012, SFRH/
BD/80122/2011, SFRH/BD/47730/2008, and SFRH/BPD/64423/
2009, respectively).
Keywords: boranes · fluorescence · N-iminopyrrolyl ligands ·
OLEDs · density functional calculations
[1] a) G. Parshall, S. Ittel, Homogeneous Catalysis: The Applications of
Chemistry of Catalysis by Soluble Transition Metal Complexes, 2nd ed.,
Wiley, New York, 1992; b) B. Cornils, W. A. Herrmann, Applied Homoge-
neous Catalysis with Organometallic Compounds, 2nd ed., Wiley-VCH,
Weinheim, 2002.
[2] K. Mashima, H. Tsuguri, J. Organomet. Chem. 2005, 690, 4414–4423, and
references therein.
[3] T. Li, J. Jenter, P. W. Roesky, Z. Anorg. Allg. Chem. 2010, 636, 2148–2155.
[4] R. H. Holm, A. Chakravorty, L. J. Theriot, Inorg. Chem. 1966, 5, 625–635.
[5] See, for instance: a) T. K. Panda, K. Yamamoto, K. Yamamoto, H. Kaneko,
Y. Yang, H. Tsurugi, K. Mashima, Organometallics 2012, 31, 2268–2274;
b) J.-S. Mu, Y.-X. Wang, B.-X. Li, Y.-S. Li, Dalton Trans. 2011, 40, 3490–
3497; c) H. Tsurugi, Y. Matsuo, K. Mashima, J. Mol. Catal. A 2006, 254,
131–137; d) R. M. Bellabarba, P. T. Gomes, S. I. Pascu, Dalton Trans.
2003, 4431–4436; e) S. A. Carabineiro, L. C. Silva, P. T. Gomes, L. C. J.
Pereira, L. F. Veiros, S. I. Pascu, M. T. Duarte, S. Namorado, R. T. Henri-
ques, Inorg. Chem. 2007, 46, 6880–6890; f) S. A. Carabineiro, P. T.
Gomes, L. F. Veiros, C. Freire, L. C. J. Pereira, R. T. Henriques, J. E. Warren,
S. I. Pascu, Dalton Trans. 2007, 5460–5470; g) S. A. Carabineiro, R. M.
Bellabarba, P. T. Gomes, S. I. Pascu, L. F. Veiros, C. Freire, L. C. J. Pereira,
R. T. Henriques, M. C. Oliveira, J. E. Warren, Inorg. Chem. 2008, 47, 8896–
8911; h) C. S. B. Gomes, D. Suresh, P. T. Gomes, L. F. Veiros, M. T. Duarte,
T. G. Nunes, M. C. Oliveira, Dalton Trans. 2010, 39, 736–748; i) C. S. B.
Gomes, S. A. Carabineiro, P. T. Gomes, M. T. Duarte, Inorg. Chim. Acta
2011, 367, 151–157.
[13] a) S. Sole, F. P. Gabbai, Chem. Commun. 2004, 1284–1285; b) W. Yang, H.
He, D. G. Drueckhammer, Angew. Chem. 2001, 113, 1764–1768; Angew.
Chem. Int. Ed. 2001, 40, 1714–1718; c) J. Killoran, L. Allen, J. F. Gallagher,
W. M. Gallagher, D. F. O’Shea, Chem. Commun. 2002, 1862–1863;
d) Z. Q. Liu, Q. Fang, D. Wang, G. Xue, W. T. Yu, Z. S. Shao, H. M. Jiang,
Chem. Commun. 2002, 2900–2901; e) S. Yamaguchi, T. Shirasaka, S.
Akiyama, K. Tamao, J. Am. Chem. Soc. 2002, 124, 8816–8817.
[14] a) S. L. Hellstrom, J. Ugolotti, G. J. P. Britovsek, T. S. Jones, A. J. P. White,
New J. Chem. 2008, 32, 1379–1387; b) T.-R. Chen, R.-H. Chien, M.-S. Jan,
A. Yeh, J.-D. Chen, J. Organomet. Chem. 2006, 691, 799–804; c) K.
Rurack, M. Kollmannsberger, J. Daub, Angew. Chem. 2001, 113, 396–
399; Angew. Chem. Int. Ed. 2001, 40, 385–387.
[15] C. S. B. Gomes, P. T. Gomes, R. E. Di Paolo, A. L. MaÅanita, M. T. Duarte,
M. J. Calhorda, Inorg. Chem. 2009, 48, 11176–11186.
[16] M. G. Crestani, G. F. Manbeck, W. W. Brennessel, T. M. McCormick, R. Ei-
senberg, Inorg. Chem. 2011, 50, 7172–7188.
[6] a) G. J. P. Britovsek, V. C. Gibson, D. F. Wass, Angew. Chem. 1999, 111,
448–468; Angew. Chem. Int. Ed. 1999, 38, 428–447 and references
therein; b) V. C. Gibson, S. K. Spitzmesser, Chem. Rev. 2003, 103, 283–
313 and references therein.
[7] a) T. K. S. Wong, Handbook of Organic Electronics and Photonics, Vol. 2
(Ed.: H. S. Nalwa), American Scientific Publishers, Stevenson Ranch, CA,
2008, pp. 413–472; b) C. H. Chen, J. Shi, Coord. Chem. Rev. 1998, 171,
161–174; c) S. Wang, Coord. Chem. Rev. 2001, 215, 79–98.
[8] a) Z. Shen, P. E. Burrows, V. Bulovic, S. R. Borrest, M. E. Thompson, Sci-
ence 1997, 276, 2009–2011; b) H. Aziz, Z. D. Popovic, N.-X. Hu, A.-M.
Hor, G. Xu, Science 1999, 283, 1900–1902; c) Y. Li, Y. Liu, W. Bu, J. Guo,
[17] a) D. Suresh, C. S. B. Gomes, P. T. Gomes, R. E. Di Paolo, A. L. MaÅanita,
M. J. Calhorda, A. Charas, J. Morgado, M. T. Duarte, Dalton Trans. 2012,
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
14
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
41, 8502–8505; and the following two errata: b) D. Suresh, C. S. B.
Gomes, P. T. Gomes, R. E. Di Paolo, A. L. MaÅanita, M. J. Calhorda, A.
Charas, J. Morgado, M. T. Duarte, Dalton Trans. 2012, 41, 14713; c) D.
Suresh, C. S. B. Gomes, P. T. Gomes, R. E. Di Paolo, A. L. MaÅanita, M. J.
Calhorda, A. Charas, J. Morgado, M. T. Duarte, Dalton Trans. 2013, 42,
16969.
[36] M. C. Burla, R. Caliandro, M. Camalli, B. Carrozzini, G. L. Cascarano, L.
De Caro, C. Giacovazzo, G. Polidori, R. Spagna, J. Appl. Crystallogr. 2005,
38, 381–388.
[37] G. M. Sheldrick, SHELXL-97, A Computer Program for Refinement of
Crystal Structure, University of Gçttingen, Gçttingen, 1997.
[38] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[39] M. N. Burnett, C. K. Johnson, ORTEP-III: Oak Ridge Thermal Ellipsoid Plot
Program for Crystal Structure Illustration, Oak Ridge National Laboratory
Report ORNL-6895, 1996.
[18] M. J. Calhorda, D. Suresh, P. T. Gomes, R. E. Di Paolo, A. L. MaÅanita,
Dalton Trans. 2012, 41, 13210–13217.
[19] C. S. B. Gomes, C. A. Figueira, P. S. Lopes, D. Suresh, P. T. Gomes, M. T.
Duarte, Acta Crystallogr. Sect. C 2011, 67, o315–o318.
[20] T. Yasumoto, T. Yamagata, K. Mashima, Organometallics 2005, 24, 3375–
3377.
[40] R. F. Rodrigues, P. F. da Silva, K. Shimizu, A. A. Freitas, S. A. Kovalenko,
N. P. Ernsting, F. H. Quina, A. MaÅanita, Chem. Eur. J. 2009, 15, 1397–
1402.
[21] P. J. Figiel, A. Sibaouih, J. U. Ahmad, M. Nieger, M. T. Raeisaenen, M. Le-
skelae, T. Repo, Adv. Synth. Catal. 2009, 351, 2625–2632.
[22] A. Ali-Bey, H. F. Loukil, G. Gosselin, A. Mathieu, A. Mesli, J. Heterocycl.
Chem. 1982, 19, 617–624.
[41] G. Striker, V. Subramaniam, C. A. M. Seidel, A. Volkmer, J. Phys. Chem. B
1999, 103, 8612–8617.
[42] a) L. Versluis, T. Ziegler, J. Chem. Phys. 1988, 88, 322–328; b) L. Fan, T.
Ziegler, J. Chem. Phys. 1991, 95, 7401–7408.
[23] M. Asadi, M. H. Ghatee, S. Torabi, K. Mohammadi, F. Moosavi, J. Chem.
Sci. 2010, 122, 539–548.
[43] S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200–1211.
[44] A. D. Becke, J. Chem. Phys. 1988, 88, 1053–1062.
[24] B.-C. Xu, T. Hu, J.-Q. Wu, N.-H. Hu, Y.-S. Li, Dalton Trans. 2009, 8854–
8863.
[45] a) J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824; b) J. P. Perdew, Phys.
Rev. B 1986, 34, 7406–7406.
[25] J. Pommerehne, H. Vestweber, W. Guss, R. F. Mahrt, H. Bꢇssler, M. Porsh,
J. Daub, Adv. Mater. 1995, 7, 551–554.
[46] E. van Lenthe, A. Ehlers, E. J. Baerends, J. Chem. Phys. 1999, 110, 8943–
8953.
[26] a) C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195; b) M.
Segala, Y. Takahata, D. P. Chong, J. Mol. Struct. THEOCHEM 2006, 758,
61–69.
[47] a) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868; b) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1997, 78,
1396–1396.
[27] See, for example: a) B. Foresman, M. Head-Gordon, J. A. Pople, M. J.
Frisch, J. Phys. Chem. 1992, 96, 135–149; b) M. Head-Gordon, R. J. Rico,
M. Oumi, T. J. Lee, Chem. Phys. Lett. 1994, 219, 21–29; c) M. Head-
Gordon, D. Maurice, M. Oumi, Chem. Phys. Lett. 1995, 246, 114–121.
[28] R. G. Parr, W. Yang, Density Functional Theory of Atoms and Molecules,
Oxford University Press, New York, 1989.
[29] a) SCM, ADF2009, Theoretical Chemistry, Vrije Universiteit, Amsterdam,
The Netherlands, http://www.scm.com; b) G. te Velde, F. M. Bickelhaupt,
E. J. Baerends, C. F. Guerra, S. J. A. van Gisbergen, J. G. Snijders, T. Zieg-
ler, J. Comput. Chem. 2001, 22, 931–967; c) C. F. Guerra, J. G. Snijders,
G. te Velde, E. J. Baerends, Theor. Chem. Acc. 1998, 99, 391–403.
[30] a) S. J. A. van Gisbergen, J. A. Groeneveld, A. Rosa, J. G. Snijders, E. J.
Baerends, J. Phys. Chem. A 1999, 103, 6835–6844; b) A. E. Rosa, J. Baer-
ends, S. J. A. van Gisbergen, E. van Lenthe, J. A. Groeneveld, J. G. Snijd-
ers, J. Am. Chem. Soc. 1999, 121, 10356–10365; c) S. J. A. van Gisbergen,
A. Rosa, G. Ricciardi, E. J. Baerends, J. Chem. Phys. 1999, 111, 2499–
2507; d) S. J. A. van Gisbergen, J. G. Snijders, E. J. Baerends, Comput.
Phys. Commun. 1999, 118, 119–138.
[48] Gaussian 09, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Pe-
tersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J.
Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bear-
park, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V.
Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Moro-
kuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, K. J. Dannenberg, S. Dap-
prich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski,
D. J. Fox, Gaussian, Inc., Wallingford, CT, 2009.
[49] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724–
728; b) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56,
2257–2261; c) P. C. Hariharan, J. A. Pople, Mol. Phys. 1974, 27, 209–214;
d) M. S. Gordon, Chem. Phys. Lett. 1980, 76, 163–168; e) P. C. Hariharan,
J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222.
[50] http://www.chemcraftprog.com/index.html.
[51] J. Morgado, A. Charas, J. A. Fernandes, I. S. GonÅalves, L. D. Carlos, L.
Alcꢁcer, J. Phys. D: Appl. Phys. 2006, 39, 3582–3587.
[31] S. Portmann, H. P. Lꢈthi, Chimia 2000, 54, 766–770.
[32] F. Wang, T. Ziegler, J. Chem. Phys. 2005, 123, 154102–154113.
[33] SMART software for the CCD detector system, Version 5.625, Bruker AXS
Inc., Madison, WI, USA, 2001.
[34] SAINT software for the CCD detector system, Version 7.03, Bruker AXS
Inc., Madison, WI, USA, 2004.
[35] G. M. Sheldrick, SADABS, Program for Empirical Absorption Correction,
University of Gçttingen, Gçttingen, 1996.
Received: September 11, 2013
Published online on && &&, 0000
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
15
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
Fluorescent N,N’-boron chelate com-
plexes: Mononuclear boron complexes
of 2-(N-arylimino)pyrrolyl emit violet to
bluish-green colors in solution (see
figure, ITO=indium tin oxide, PE-
DOT:PSS=poly(3,4-ethylenedioxythio-
phene):poly(styrene sulfonic acid)), de-
pending on the substituents on the N-
aryl group. Organic light-emitting
diodes have been successfully fabricat-
ed with the new boron complexes, ach-
ieving luminances of the order of 3000
cdmꢀ2.
&
Fluorescence
D. Suresh, P. S. Lopes, B. Ferreira,
C. A. Figueira, C. S. B. Gomes,
P. T. Gomes,* R. E. Di Paolo,
A. L. MaÅanita, M. T. Duarte, A. Charas,
J. Morgado, M. J. Calhorda
&& – &&
Tunable Fluorophores Based on 2-(N-
Arylimino)pyrrolyl Chelates of
Diphenylboron: Synthesis, Structure,
Photophysical Characterization, and
Application in OLEDs
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
16
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!