A series of blue-green-yellow-red emitting Cu(I) complexes: Molecular structure and photophysical performance

Article abstract of DOI:10.1016/j.saa.2019.117280

In this work, we designed a series of [Cu(N–N)(PPh₃)₂]BF₄ complexes with different optical edge values and emission colors from blue to red, where N–N and PPh₃ denoted a diamine ligand and triphenylphosphine, respectively. Six N–N ligands with various conjugation chains (short π chain, modest π chain and long π chain) were selected. A systematical comparison between these Cu(I) complexes was performed, so that the correlation between N–N structure and [Cu(N–N)(PPh₃)₂] photophysical performance was tentatively discussed. Their single crystal structure was found consistent with literature ones, forming a typical tetrahedral coordination geometry. Density functional theory calculation indicated that their onset electronic transition showed a mixed character of metal-to-ligand-charge-transfer and ligand-to-ligand-charge-transfer. Detailed analysis on photophysical parameters suggested that the absorption edge of [Cu(N–N)(PPh₃)₂]BF₄ complex was controlled by conjugation length in diamine ligand. A wide absorption edge needed a short conjugation chain in diamine ligand. Similar tendency was found for their emission spectra. In addition, a long conjugation chain in diamine ligand widened emission spectra obviously. Emission dynamics showed slim correlation with diamine ligand conjugation length since the excited state was controlled mainly by dynamic procedure and steric factor of diamine ligands.

Full text of DOI:10.1016/j.saa.2019.117280

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular
Spectroscopy
journal homepage: www.elsevier.com/locate/saa
A series of blue-green-yellow-red emitting Cu(I) complexes: Molecular
structure and photophysical performance
b,
Liming Zhang a, , Qinghui Zuo
⁎
⁎
a
State Key Laboratory of Luminescence and Applications, Changchun Institute of Optics Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, PR China
School of Materials Science and Engineering, Changchun University of Science and Technology, No. 7989, Weixing Road, Changchun 130022, PR China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 April 2019
Received in revised form 14 June 2019
Accepted 15 June 2019
Available online 17 June 2019
In this work, we designed a series of [Cu(N\\N)(PPh3)2]BF4 complexes with different optical edge values and
emission colors from blue to red, where N\\N and PPh3 denoted a diamine ligand and triphenylphosphine, re-
spectively. Six N\\N ligands with various conjugation chains (short π chain, modest π chain and long π chain)
were selected. A systematical comparison between these Cu(I) complexes was performed, so that the correlation
between N\\N structure and [Cu(N\\N)(PPh3)2] photophysical performance was tentatively discussed. Their sin-
gle crystal structure was found consistent with literature ones, forming a typical tetrahedral coordination geom-
etry. Density functional theory calculation indicated that their onset electronic transition showed a mixed
character of metal-to-ligand-charge-transfer and ligand-to-ligand-charge-transfer. Detailed analysis on
photophysical parameters suggested that the absorption edge of [Cu(N\\N)(PPh3)2]BF4 complex was controlled
by conjugation length in diamine ligand. A wide absorption edge needed a short conjugation chain in diamine
ligand. Similar tendency was found for their emission spectra. In addition, a long conjugation chain in diamine
ligand widened emission spectra obviously. Emission dynamics showed slim correlation with diamine ligand
conjugation length since the excited state was controlled mainly by dynamic procedure and steric factor of di-
amine ligands.
Keywords:
Phosphorescent Cu(I)
Molecular structure
Conjugation plane
Frontier molecular orbitals
© 2019 Elsevier B.V. All rights reserved.
1. Introduction
their occupied frontier molecular orbitals (FMOs) are composed of Cu
(I) d orbital and P\\P ligand, which endows these FMOs with dominant
Luminescent metal complexes have attracted more and more re-
search attention owing to their wide application in optoelectronic de-
vices such as light-emitting panels and solar cells [1–3]. Although
heavy metal complexes generally have better performance in aspect of
emission variety, emission lifetime and efficiency, more and more con-
cern has been focused on their toxicity and environmental impact [4]. In
this case, Cu(I) complexes have been highly recommended owing to
their environmental friendliness and economic cost [5–7]. Among
them, phosphorescent ones with heteroleptic ligands have been inten-
sively studied. Their ligands include diamine aromatic ligands (N\\N)
such as 1,10-phenanthroline and phosphorous ligands (P\\P) such as
PPh3, forming a representative tetrahedral geometry with molecular
formula of [Cu(N\\N)(P\\P)]+. Such structures have been firstly re-
ported by McMillin in 2002 [5]. Analogous structures have been synthe-
sized and reported since they have shown efficient emission with long-
lived emissive center. Feng and his group have performed theoretical
calculation on [Cu(N\\N)(P\\P)] structures [8]. It is reported that
metal character. While the unoccupied FMOs are basically the π* of
N\\N ligands. As a result, the emissive state of [Cu(N\\N)(P\\P)] struc-
tures is generally described as a character of metal-to-ligand-charge-
transfer (MLCT).
It is then found that regardless of their intense absorption in UV re-
gion, these [Cu(N\\N)(P\\P)] structures have large Stokes shift and
emit low energy emission in yellow region. Detailed analysis on MLCT
excited state suggests that there exists an energy-exhausting procedure
of structural transformation from tetrahedral geometry (in ground
state) to tetragonally flattened geometry (in excited state), leading to
decreased energy content and low photoluminescence quantum yield
(PLQY) [8]. To control such structural relaxation, Zhang and coworkers
incorporated bulky substituents in 2,9-positions of 1,10-phenanthroline
[9]. As expected, Stokes shift has been decreased dramatically, and PLQY
has been improved greatly. Moreover, Zhang has reported an interest-
ing phenomenon of inner- and inter-molecular π-π stacking which effi-
ciently limits such structural relaxation as well, showing improved PL
performance [10]. In the past a few years, continuous efforts have
been devoted to reveal the correlation between ligand structure and
PL performance. For example, Cabrera and coworkers report a series of
[Cu(N\\N)(P\\P)] structures with MLCT absorption between 400 nm
⁎
Corresponding authors.
E-mail addresses: zhangliming@ciomp.ac.cn (L. Zhang), zuoqinghui@cust.edu.cn
(Q. Zuo).
https://doi.org/10.1016/j.saa.2019.117280
1386-1425/© 2019 Elsevier B.V. All rights reserved.

2
L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
and 450 nm and discuss their HOMO-LUMO energy gaps with DFT
methods [11]. Their emission is reported as a mixture of triplet MLCT
and triplet LC (ligand-centered) state. Ren and coworkers then present
their theoretical work about isomerization and methyl modification to
decrease non-radiative decay rate and thus improve emissive perfor-
mance of triplet MLCT state [12]. Sessolo, Orti and Housecroft have re-
ported ligands of 1,10-phenanthroline and bipyridine substituted at
only 2-position, showing red emission with quantum yields of ~38%
and emission lifetime of ~10 μs from their [Cu(N\\N)(P\\P)] complexes
[13–15]. Mallick and Sinha have reported phosphino-bridged com-
plexes with MLCT absorption at ~500 nm [16]. Tschierlei and Karnahl
have found long-lived excited states in their [Cu(N\\N)(P\\P)] com-
plexes owing to the extended π-system in anthracene-based diamine li-
gand [17].
Regardless of these numerous reports, although the key factors
affecting MLCT excited state of [Cu(N\\N)(P\\P)] have been re-
vealed as above stated, high-energy emission, such as deep blue,
from [Cu(N\\N)(P\\P)] structures is still rare [18,19]. This is because
their optical edge (ΔE = ELUMO-EHOMO) is usually a narrow one
(b450 nm), which technically denies their high energy emission,
given the presence of Stokes shift. There is even rare systematical re-
search on how to adjust the ΔE of [Cu(N\\N)(P\\P)] structures.
Given this circumstance, we intend to devote our effort to the
correlation between [Cu(N\\N)(P\\P)] molecular structure and
its ΔE. PPh3 is selected in this work instead of bis(2-
diphenylphosphiophenyl)ether since PPh3 flexible structure avoids
π-π stacking and its effect on PL performance. Six N\\N ligands
with various conjugation chains (short π chain, modest π chain and
long π chain) are selected. A systematical comparison between
these Cu(I) complexes is performed, so that the correlation between
N\\N structure and [Cu(N\\N)(P\\P)] photophysical performance
can be tentatively discussed (Scheme 1).
(1H-benzo[d]imidazol-2-yl)thiazole, ethane-1,2-diamine, benzene-
1,2-diamine, 4-methylbenzenesulfonic acid and PPh3. NMR data
were collected on
a
Varian INOVA 300 spectrometer.
Photoluminescence and absorption spectra were recorded using a
Hitachi F-7000 spectrometer and a Shimadzu UV-3101 PC spectro-
photometer, respectively. Photodynamic data were collected by a
FL920 spectrometer equipped with
a
hydrogen lamp.
Photoluminescence quantum yield was determined using the combi-
nation of the Hitachi F-7000 spectrometer and an integrating sphere
with solid samples. Single crystal data were collected by the combi-
nation of a Siemens P4 single-crystal X-ray diffractometer and a
Smart CCD-1000 detector (Mo Kα radiation at 298 K). Density func-
tional theory (DFT) calculation was performed on [Cu(N\\N)(PPh3)
2] complexes using their single crystal structure as initial geometry.
This DFT calculation was performed with Firefly in vacuum at
RB3LYP/SBKJC level. Graphical plotting for frontier molecular or-
bitals was done by wxMacmolplt software package (contour value
= 0.03).
2.2. Synthesis of N\\N ligands
L1A. 1H,1′H-2,2′-biimidazole (L1A) was obtained following below
procedure [20]. Aqueous glyoxal (20%) was placed in water bath at 45
°C, then ammonia was bubbled into this solution for 2 h. The mixture
was further stirred overnight. Solid product was collected, washed
with water and then dissolved in ethylene glucol. After being decolor-
ized by C powder, the solution was cooled. White solid product was ob-
tained as L1A. 1H NMR (DMSO d6): δ 7.08 (s).
L1B. 2-(1H-tetrazol-5-yl)pyridine (L1B) was obtained as follows. A
mixture of 2-cyanopyridine (10 mmol), NaN3 (15 mmol), ZnBr
(10 mmol) and water (30 mL) was prepared and heated at 100 °C for
8 h. After cooling, solid NaOH (25 mmol) was added. The solution was
filtered and then mixed with HCl until pH = 1. Crude product was col-
lected and purified on a silica gel column to give L1B as white solid. 1H
NMR (CDCl3): δ 8.21 (1H, t), 8.25 (1H, t), 8.33 (1H, d, J = 6.0), 8.85
(1H, d, J = 3.6).
2. Experimental section
2.1. Equipment and reagents
L2A. 4-(1H-benzo[d]imidazol-2-yl)thiazole (L2A) was purchased
and used as received as above mentioned.
L2B. 2-phenyl-5-(pyridin-2-yl)-1,3,4-oxadiazole (L2B) was obtained
as follows. A mixture of L1B (10 mmol), benzoyl chloride (15 mmol)
All chemicals were AR grade ones and commercially obtained
with no further purifications, including aqueous glyoxal (20%), am-
monia, 2-cyanopyridine, sodium azide, 1,10-phenanthroline, 4-
Scheme 1. Molecular structure of [Cu(N\\N)(PPh3)2]BF4 complexes used in this work.

L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
3
and 35 mL of pyridine was prepared and stirred under N2 atmosphere at
3. Result and discussion
120 °C for 2 days. After cooling, crushed ice was added. Crude product
was collected and purified on a silica gel column to give L2B as white
powder. 1HNMR (CDCl3): δ 7.45 (1H, m), 7.57 (3H, m), 7.91 (1H, m),
8.22 (1H, t), 8.28 (1H, t), 8.35 (1H, d, J = 6.0), 8.84 (1H, d, J = 3.6).
L3A. 1,10-phenanthroline-5,6-dione (Phen-O) was firstly prepared
following a literature procedure [21]. Then a mixture of Phen-O
(10 mmol), ethane-1,2-diamine (3 mL), ethanol (25 mL) and 4-
methylbenzenesulfonic acid (1 mmol) was prepared and stirred at 80
°C overnight. Crude product was collected, washed with ethanol and
purified in hot ethanol to give L3A as white powder. 1H NMR (CDCl3):
δ7.81(m, 2H), 9.02 (s, 2H), 9.33(d, 2H, J = 8.0 Hz), 9.53(d, 2H, J =
8.0 Hz).
L3B. Dipyrido[3,2-a:2′,3′-c]phenazine (L3B) was synthesized follow-
ing a similar procedure to L3A, except that ethane-1,2-diamine was re-
placed with benzene-1,2-diamine in this run. 1H NMR (CDCl3): δ, 7.81
(m, 2H), 7.94 (d, 2H, J = 6.4 Hz), 8.32 (d, 2H, J = 6.4 Hz), 9.24 (d, 2H,
J = 8.0 Hz), 9.61 (d, 2H, J = 8.0 Hz).
3.1. Molecular structure
For a clear understanding on the design strategy of this work, a de-
tailed explanation is given as follows. As above mentioned, although
bis(2-diphenylphosphinophenyl)ether (POP) is a widely used auxiliary
phosphorous ligand for [Cu(N\\N)(P\\P)] complexes, its phenyl rings
are restricted by its O atom and close to N\\N ligand, resulting in
inner-molecular π-π stacking which affects PL performance of [Cu
(N\\N)(POP)] complexes. To eliminate such effect, PPh3 is selected as
the auxiliary ligand in this work since the free rotation of its phenyl
rings compromises such highly ordered π-π stacking. As for the six di-
amine ligands, they are all coplanar ones with similar coordination per-
formance. There are electron-donors or electron-acceptors in these
coplanar π chains. They are divided into three groups according to the
number of their conjugation π electrons, so that the correlation between
the optical edge of [Cu(N\\N)(PPh3)2]BF4 and N\\N ligand structure
can be discussed.
2.3. Synthesis of [Cu(N\\N)(PPh3)2] complexes
Owing to the rigid coplanar diamine ligands, [Cu(N\\N)(PPh3)2]BF4
complexes tend to crystalize in solid state. Fig. 1 shows their ORTEP
plotting, corresponding key structural parameters are listed in Table 1.
It is clear that these [Cu(N\\N)(PPh3)2]BF4 complexes all adopt a similar
coordination geometry to literature cases [8]. Two N atoms from a N\\N
ligand and two P atoms from two PPh3 ligands coordinate with a Cu
(I) center, forming a typical tetrahedral coordination geometry. This ge-
ometry is slightly distorted owing to its heterogeneous ligands, as sug-
gested by their structural parameters. All Cu\\P bonds are quite
similar to each other owing to their identical coordination P atom. The
Cu\\N bonds, however, are slightly different from each other but still
comparable to literature ones [8]. It is found that the Cu\\N bond be-
tween Cu(I) and the N atom from a five-membered ring (imidazole
etc., N2.1 Å) is longer than that between Cu(I) and the N atom from a
six-membered ring (pyridine etc., b2.1 Å). This tendency is independent
of electron-donors or acceptors in diamine ligands. We thus attribute its
causation to the strong coordination tension in a five-membered ring
[20]. Coordination bite angles of N-Cu-N (b80°) are typically smaller
than literature values (N80°). In addition, P-Cu-P bite angles vary in a
wide region ranging from 118° to 126°. This fact actually means a
crowded coordination environment, and both diamine ligand and
PPh3 ligand try to decrease coordination hindrance by leaving the Cu
(I) center.
All [Cu(N\\N)(PPh3)2] complexes were synthesized following a
classical procedure [19]. Their identity was confirmed by NMR, ele-
mental analysis and single crystal analysis. These single crystals
were used for later characterization, including photophysical
analysis.
[Cu(L1A)(PPh3)2]BF4. Starting compound [Cu(CH3CN)4]BF4 was pre-
pared as follows. A mixture of [Cu(CH3CN)4]BF4 (5 mmol), PPh3
(10 mmol) and CHCl3 (10 mL) was prepared and stirred for 30 min
under ambient condition. Then L1A (5 mmol) was added. The resulting
solution was stirred for another 60 min. 5 mL of ethanol was added,
then this solution was filtered. The natural evaporation of solvent gave
bulk solid. 1H NMR(CDCl3): δ 8.05 (t, 2H), 7.97 (t, 2H), 7.35 (t, 11H),
7.26 (m, 4H), 7.23 (m, 4H), 7.05 (m, 8H), 6.60 (m, 3H). Elemental anal-
ysis for C42H36B1Cu1F4N2P2: C, 62.35, H, 4.48, N, 6.92. Found: C, 62.27, H,
4.56, N, 6.84.
[Cu(L1B)(PPh3)2]BF4. Its synthetic procedure was similar to [Cu
(L1A)(PPh3)2]BF4, except that L1A was replaced by L1B in this run. 1H
NMR(CDCl3): δ 8.18 (d, 2H), 8.03 (t, 2H), 7.36 (t, 11H), 7.28 (m, 4H),
7.25 (m, 4H), 7.07 (m, 8H), 6.62 (m, 3H). Elemental analysis for
C42H35B1Cu1F4N5P2: C, 61.36, H, 4.29, N, 8.52. Found: C, 61.24, H, 4.44,
N, 8.47.
[Cu(L2A)(PPh3)2]BF4. Its synthetic procedure was similar to [Cu
Regardless of the close distance between aromatic rings in these [Cu
(N\\N)(PPh3)2]BF4 complexes, no obvious inner-molecular and inter-
molecular π-π stacking is observed, as above expected. This is because
the highly ordered π-π stacking involves the re-arrangement and orien-
tation adjustment of aromatic rings. These [Cu(N\\N)(PPh3)2]BF4 com-
plexes cannot satisfy this requirement owing to the free rotation of
phenyl rings in PPh3 ligand. The absence of such π-π stacking ensures
that MLCT excited state performance is controlled by diamine ligand
electronic structure solo, with neglectable geometric factor to be
considered.
(L1A)(PPh3)2]BF4, except that L1A was replaced by L2A in this run. 1H
NMR(CDCl3): δ 7.98–7.94 (m, 4H), 7.62–7.58 (m, 2H), 7.37 (t, 11H),
7.29 (m, 4H), 7.24 (m, 4H), 7.02 (m, 8H), 6.59 (m, 3H). Elemental anal-
ysis for C46H37B1Cu1F4N3P2S1: C, 63.06, H, 4.26, N, 4.80. Found: C, 62.95,
H, 4.35, N, 4.71.
[Cu(L2B)(PPh3)2]BF4. Its synthetic procedure was similar to [Cu
(L1A)(PPh3)2]BF4, except that L1A was replaced by L2B in this run. 1H
NMR(CDCl3): δ 8.02–7.98 (m, 4H), 7.36–7.28 (m, 13H), 7.25 (m, 4H),
7.21 (m, 4H), 7.01 (m, 8H), 6.60 (m, 3H). Elemental analysis for
C
49H39B1Cu1F4N3P2O1: C, 65.53, H, 4.38, N, 4.68. Found: C, 65.57, H,
4.42, N, 4.73.
3.2. FMO analysis by DFT
[Cu(L3A)(PPh3)2]BF4. Its synthetic procedure was similar to [Cu
(L1A)(PPh3)2]BF4, except that L1A was replaced by L3A in this run. 1H
NMR(CDCl3): δ 9.13 (d, 2H), 9.01 (d, 2H), 8.81 (m, 2H), 7.75 (m, 2H),
7.33 (t, 11H), 7.23 (m, 4H), 7.16 (m, 4H), 7.04 (m, 8H), 6.63 (m, 3H). El-
emental analysis for C50H38B1Cu1F4N2P2: C, 66.20, H, 4.22, N, 6.18.
Found: C, 66.27, H, 4.31, N, 6.15.
[Cu(L3B)(PPh3)2]BF4. Its synthetic procedure was similar to [Cu
(L1A)(PPh3)2]BF4, except that L1A was replaced by L3B in this run. 1H
NMR(CDCl3): δ9.24 (d, 2H), 9.11 (d, 2H), 8.04 (d, 2H), 7.74 (d, 2H),
7.60 (m, 2H), 7.34 (t, 11H), 7.25 (m, 4H), 7.21 (m, 4H), 7.03 (m, 8H),
6.60 (m, 3H). Elemental analysis for C54H40B1Cu1F4N2P2: C, 67.76, H,
4.21, N, 5.85. Found: C, 67.83, H, 4.33, N, 5.79.
Feng's work suggests that, unlike Zn(II) complexes, for most [Cu
(N\\N)(P\\P)] complexes, Cu(I) d orbitals contribute to FMO obviously
and participate in their MLCT photophysical procedures [8]. For a confir-
mation on the MLCT nature in our [Cu(N\\N)(PPh3)2]BF4 complexes, a
DFT (density functional theory) calculation is performed on their crystal
structures. Fig. 2 shows graphic presentation of their HOMO (the
highest occupied molecular orbital) and LUMO (the lowest unoccupied
molecular orbital). Their energy levels and transition energy values are
listed in Table 2. It is clear that the occupied FMOs have mixed character
with contribution from Cu(I) center and the P atom of PPh3 ligand. As
for the unoccupied FMOs, they are completely composed of diamine

4
L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
b
a
c
d
e
Fig. 1. ORTEP plotting of [Cu(N\\N)(PPh3)2]BF4 complexes, a, [Cu(L1B)(PPh3)2]BF4, b, [Cu(L2A)(PPh3)2]BF4, c, [Cu(L2B)(PPh3)2]BF4, d, [Cu(L3A)(PPh3)2]BF4, e, [Cu(L3B)(PPh3)2]BF4. H
atoms were not plotted for clarity.

L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
5
Table 1
Table 2
Selected geometric parameters of [Cu(N\\N)(PPh3)2]BF4 complexes.
Energy levels and transition energy values of [Cu(N\\N)(PPh3)2]BF4 complexes.
N-N=
N-N=
L1B
L2A
L2B
L3A
L3B
Energy (eV)
L1B
L2A
L2B
L3A
L3B
Bond length (Å)
Cu(1)-N(1)
Cu(1)-N(2)
Cu(1)-P(1)
Cu(1)-P(2)
Bond angle (°)
N(1)-Cu-N(2)
N(1)-Cu-P(1)
N(2)-Cu-P(1)
N(1)-Cu-P(2)
N(2)-Cu-P(2)
P(1)-Cu-P(2)
LUMO+1
LUMO
HOMO
−4.41
−5.22
−8.22
−8.47
2.42
−3.65
−4.33
−7.63
−8.14
2.62
−3.97
−4.92
−8.04
−8.48
2.55
−4.59
−4.77
−7.96
−8.46
2.65
−4.52
−4.91
−8.00
−8.31
2.63
2.186
2.103
2.258
2.254
2.157
2.042
2.229
2.281
2.154
2.095
2.278
2.237
2.068
2.083
2.237
2.249
2.105
2.069
2.272
2.275
HOMO-1
S
S
0 → S1
0 → S2
2.49
3.08
2.77
2.84
2.82
78.15
114.15
103.40
110.71
112.65
126.75
79.04
125.70
119.17
107.99
106.38
118.19
78.47
100.37
109.33
121.63
117.69
120.96
80.22
119.87
108.93
108.93
117.00
120.63
79.36
111.23
108.62
111.15
113.59
124.09
Based on this analysis, it is found that there are two dominant factors
for a diamine ligand in affecting the energy levels of these [Cu(N\\N)
(PPh3)2]BF4 complexes: (1) the length of conjugation chain and (2) elec-
tron donors or acceptors in conjugation chain. Their effect is briefed as
follows. A long conjugation chain decreases HOMO and LUMO energy
levels along with onset electronic transition energy, while a short one
does otherwise. Electron donors in conjugation chain increases HOMO
and LUMO energy levels, along with onset electronic transition energy,
while electron acceptors do otherwise. This hypothesis is firstly con-
firmed by the DFT result shown in Table 2 and will be further confirmed
by below discussion.
ligands π*. The onset electronic transitions between occupied FMOs and
unoccupied ones are thus endowed with a mixed character of MLCT
(metal-to-ligand-charge-transfer) and LLCT (ligand-to-ligand-charge-
transfer). This conclusion is consistent with literature reports on similar
Cu(I) complexes [22]. Such kind of excited state (MLCT&LLCT), espe-
cially MLCT one, generally suffers from a structural relaxation from tet-
rahedral geometry to tetragonally flattened geometry, accompanied by
non-radiative decay. As a consequence, excited state energy is de-
creased and quenched, leading to low emission efficiency and low-
energy emission (large Stokes shift) [9,10]. This statement will be fur-
ther discussed below.
It is still noticed that the energy levels of occupied FMOs are usually
localized in a narrow region (−8.47 eV to −8.31 eV for HOMO-1,
−8.22 eV to −7.96 eV for HOMO). On the other hand, those of unoccu-
pied FMOs vary in a wide region (−5.22 eV to −4.33 eV for LUMO,
−4.52 eV to −3.65 eV for LUMO+1). This is well explained by their na-
ture. Since occupied FMOs have dominant Cu(I) contribution, their en-
ergy levels are thus limited in a fixed region, despite of different
diamine ligands and their slim contribution. As for the unoccupied
FMOs, they are diamine ligand π* orbitals in nature, which endows
them with a great variety. It is still observed that electron-donors
with lone pairs in diamine ligand, such as the S atom in L2A ligand,
lead to obvious increase of both HOMO and LUMO energy levels. Cor-
respondingly, onset electronic transition energy is increased. While,
a long conjugation chain, such as L3A and L3B ligands, tends to de-
crease HOMO and LUMO energy levels. In addition, electron-
acceptors tend to decrease both HOMO and LUMO energy levels as
well, such as L1B ligand, but its short conjugation chain compromises
such decrease effect.
3.3. Absorption spectra
The onset electronic transition corresponds to an absorption edge. In
this case, absorption spectra of these [Cu(N\\N)(PPh3)2]BF4 complexes
are recorded and shown in Fig. 3. Similar to literature case, each absorp-
tion spectrum of these [Cu(N\\N)(PPh3)2]BF4 complexes consists of
two major sections, a strong absorption in high energy UV region from
230 nm to 330 nm and a weak absorption in low energy region from
350 nm to ~450 nm. The strong absorption peaks have been revealed
by DFT calculation result as π → π* transition of diamine and phospho-
rous ligands [8,18]. As for the weak absorption band, no such absorption
band is observed from diamine or phosphorous ligands. It is thus named
as the electronic transition of MLCT&LLCT, as above mentioned [8]. Cor-
responding absorption edge (λabs) data are summarized in Table 3 for
comparison convenience. It is obvious that the optical edge values of
the first three [Cu(N\\N)(PPh3)2]BF4 complexes (N405 nm) are much
higher than those of literature reports (b450 nm) [8,9,23]. These high
absorption edge values guarantee their high energy emission, which
will be confirmed below. It is clear that λabs follows an obvious order
of [Cu(L1A)(PPh3)2]BF4 (366 nm) b [Cu(L1B)(PPh3)2]BF4 (401 nm) b
[Cu(L2A)(PPh3)2]BF4 (405 nm) b [Cu(L2B)(PPh3)2]BF4 (448 nm) b [Cu
(L3A)(PPh3)2]BF4 (460 nm) b [Cu(L3B)(PPh3)2]BF4 (466 nm). This
Fig. 2. Graphic presentation for HOMO and LUMO, and energy levels of FMOs.

6
L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
Fig. 3. Absorption spectra of [Cu(N\\N)(PPh3)2]BF4 complexes in CH2Cl2 solution with
concentration of 1 μM.
Fig. 4. Emission spectra of [Cu(N\\N)(PPh3)2]BF4 complexes in solid state.
order perfectly matches the order of π electron number of their diamine
ligands, L1A (8) b L1B (10) b L2A (12) b L2B (16) b L3A (18) b L3B (22),
which can be explained as follows. As revealed by DFT calculation result
of this work and literature reports, HOMO of [Cu(N\\N)(PPh3)2]BF4
complexes is mainly composed of Cu(I) orbitals and the P atom from
PPh3 ligand [8]. There is rather slim contribution from N\\N ligand.
The HOMO energy level is thus restricted in a narrow region. As for
LUMO, it is diamine ligand π* orbital in nature. In other words, LUMO
energy level is variable and controlled just by diamine structure (π*).
In this case, the absorption edge between HOMO and LUMO is actually
controlled by diamine ligand (π*). A short conjugation chain in diamine
ligand leads to a high level π* of this diamine ligand, and consequently a
wide absorption edge of its [Cu(N\\N)(PPh3)2]BF4 complex. Given a di-
amine ligand with a longer conjugation chain, more π electrons are in-
troduced, which decreases π* energy level of this diamine ligand and
consequently the absorption edge of its [Cu(N\\N)(PPh3)2]BF4 com-
plex. It is thus concluded that the absorption edge of [Cu(N\\N)(PPh3)
2]BF4 complex is controlled by diamine conjugation length. A wide ab-
sorption edge needs a short conjugation chain in diamine ligand.
is reasonable since its emissive state is originated from MLCT&LLCT ex-
cited state [8]. Interestingly, the emission peak of these [Cu(N\\N)
(PPh3)2]BF4 complexes follows the same order of their absorption
edge, [Cu(L1A)(PPh3)2]BF4 (466 nm) b [Cu(L1B)(PPh3)2]BF4 (499 nm)
b [Cu(L2A)(PPh3)2]BF4 (522 nm) b [Cu(L2B)(PPh3)2]BF4 (553 nm) b
[Cu(L3A)(PPh3)2]BF4 (607 nm) b [Cu(L3B)(PPh3)2]BF4 (620 nm). This
fact tentatively suggests that the emission energy of [Cu(N\\N)(PPh3)
2]BF4 complexes is mainly controlled by their intrinsic energy gap be-
tween HOMO and LUMO. Steric factors, such as incorporating bulky sub-
stituents into 2,9-positions of 1,10-phenanthroline, only minimize the
energy loss of MLCT excited state, but fail to increase or even control
the energy content of MLCT excited state.
For comparison convenience, Stokes shift between absorption edge
and emission peak of each [Cu(N\\N)(PPh3)2]BF4 complex is listed in
Table 3. FWHM (full-width-at-half-maximum) values are given in
Table 3 as well. It is clear that all these [Cu(N\\N)(PPh3)2]BF4 com-
plexes have large Stokes shift values. This is because their MLCT&LLCT
excited state tends to experience a structural relaxation from tetrahe-
dral geometry to tetragonally flattened geometry before reaching the
lowest emissive S1 state [8]. In generally, a short conjugation chain has
a minimal steric hindrance to stop such structural relaxation, which is
consistent with the highest Stokes shift value of Cu(L1A)(PPh3)2]BF4
(0.727 eV). Upon a longer conjugation chain, its steric hindrance is in-
creased, leading to decreased Stokes shift values of the other five [Cu
(N\\N)(PPh3)2]BF4 complexes. As for FWHM values, a similar order is
found, [Cu(L1A)(PPh3)2]BF4 (86 nm) b [Cu(L2A)(PPh3)2]BF4 (90 nm)
b [Cu(L1B)(PPh3)2]BF4 (93 nm) b [Cu(L2B)(PPh3)2]BF4 (96 nm) b [Cu
(L3A)(PPh3)2]BF4 (156 nm) b [Cu(L3B)(PPh3)2]BF4 (177 nm). It seems
that a short conjugation chain in diamine ligand leads to a narrow emis-
sion band, while a long one in diamine ligand widens emission band
greatly. This tendency is even more obvious in [Cu(L3A)(PPh3)2]BF4
and [Cu(L3B)(PPh3)2]BF4, considering their large FWHM values. This is
because excited electrons are mainly localized on π* orbitals of diamine
ligands. A short conjugation chain in diamine ligand means a restricted
distribution of excited state electrons. On the other hand, a long one
leads to a broad distribution of these excited state electrons, resulting
in multiple sublevels, which consequently gives a broad emission band.
3.4. Emission spectra
Upon photoexcitation on their MLCT&LLCT state, bright emission is
observed for all these [Cu(N\\N)(PPh3)2]BF4 complexes. As shown in
Fig. 4, each complex shows a broad emission band. Various emission
colors from blue to red is observed from these [Cu(N\\N)(PPh3)2]BF4
complexes. There are no vibronic details or shoulder peaks, indicating
that the excited state has a charge transfer character. This conclusion
Table 3
Photophysical parameters of [Cu(N\\N)(PPh3)2]BF4 complexes in CH2Cl2 solution or in
solid state.
N-N=
L1A
L1B
10
L2A
12
L2B
16
L3A
18
L3B
22
Number of π electrons
aλabs (nm)
8
366
466
0.727
86
401
499
0.607
93
405
522
0.686
90
448
553
0.526
96
460
607
0.653
156
466
620
0.661
177
bλem (nm)
Stokes shift (eV)
FWHM (nm)
3.5. Lifetime and PLQY
τ
τ
1 (μs)
2 (μs)
1.638
8.366
6.873
0.10
1.615
6.210
5.212
0.22
13.211 12.179
71.089 121.530
66.711 107.448
5.907
5.908
5.907
0.23
38.935
130.346 10.386
5.868
88.751
70.285
0.27
For a better understanding on the excited state of these [Cu(N\\N)
(PPh3)2]BF4 complexes, their emission decay dynamics are recorded
and shown in Fig. 5. Detailed fitting parameters are listed in Table 3. It
is observed that all six [Cu(N\\N)(PPh3)2]BF4 complexes follow
biexponential decay mode with two long-lived decay centers (τ1 and
τ2). Their lifetimes are obviously different from each other, regardless
τ (μs)
Φ
K
K
0.17
0.59
5.491
3.816
r (×103 S−1
)
14.455
42.210
2.548
3.841
nr (×103 S−1
)
130.941 149.655 12.442
a
Absorption edge.
Emission peak.
b

L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
7
short conjugation chain offers a high Knr value, while a large diamine li-
gand with a long conjugation chain tends to present a low Knr value.
This is because the non-radiative decay procedure of MLCT&LLCT ex-
cited state is mainly dominated by its structural relaxation, as above
mentioned. Here, steric factor plays an important role. Be more specific,
a small ligand offers limited steric hindrance to stop the structural relax-
ation, so that an intense geometric relaxation is occurred, leading to a
high Knr value. On the other hand, a large ligand with obvious steric hin-
drance and rigid structure may restrict such geometric relaxation, de-
creasing Knr value. It is thus concluded that the PL performance of [Cu
(N\\N)(PPh3)2]BF4 complexes is mainly controlled by dynamic proce-
dure and steric factor of diamine ligands. No direct correlation is
found with diamine ligand conjugation length.
4. Conclusion
To sum up, this paper presented six [Cu(N\\N)(PPh3)2] complexes
with various absorption edge values and emission colors from blue to
red. Their six diamine ligands were divided into three groups (short
conjugation chain, modest conjugation chain and long conjugation
chain) by the number of their π electrons. Their single crystals were
firstly obtained and discussed. It was found that these complexes all
adopted a typical tetrahedral coordination geometry. DFT calculation
on them suggested that their onset electronic transition showed a
mixed character of MLCT and LLCT, which was consistent with literature
case. In this case, the correlation between diamine ligand structure and
[Cu(N\\N)(PPh3)2] photophysical performance was carefully checked.
Detailed analysis on photophysical parameters suggested that the ab-
sorption edge of [Cu(N\\N)(PPh3)2]BF4 complex was controlled by di-
amine conjugation length. A wide absorption edge needed a short
conjugation chain in diamine ligand. Similar tendency was found for
their emission spectra. In addition, a long conjugation chain in diamine
ligand widened emission spectra obviously. Emission dynamics showed
slim correlation with diamine ligand conjugation length since the ex-
cited state was controlled mainly by dynamic procedure and steric fac-
tor of diamine ligands. This conclusion is expected to help further design
for [Cu(N\\N)(PPh3)2] complexes with controllable optical edge. For
later efforts, diamine ligands should be modified with bulky substitu-
ents so that the non-radiative decay procedure can be minimized, lead-
ing to improved PL performance.
Fig. 5. Emission decay dynamics of [Cu(N\\N)(PPh3)2]BF4 complexes in solid state.
of their microsecond scale. This observation is consistent with the mix-
ture of MLCT&LLCT in excited state. Their mean lifetime (τ) values listed
in Table 3 suggest that their emission has a triplet nature. As a conse-
quence, the emissive center of these [Cu(N\\N)(PPh3)2]BF4 complexes
can be named as triplet MLCT&LLCT, which is consistent with literature
case [8]. It has been above mentioned that there is strong π → π* absorp-
tion in each [Cu(N\\N)(PPh3)2]BF4 complex. Generally, in an emissive
center with biexponential decay behavior, there exists a potential sur-
face crossing from π → π* sate (localized at higher energy region) to
MLCT state (localized at lower energy region) [23]. In this case, the
short-lived emissive center τ1 is named as the radiative decay of π →
π* excited sate, while the long-lived emissive center τ2 is named the ra-
diative decay of MLCT excited state.
Photoluminescence quantum yields (PLQYs, Φ) of these [Cu(N\\N)
(PPh3)2]BF4 complexes are measured and listed in Table 3. No surprise,
similar to the case of Stokes shift, [Cu(L1A)(PPh3)2]BF4 has the lowest
PLQY owing to its intense structural relaxation. As for the other [Cu
(N\\N)(PPh3)2]BF4 complexes, their diamine ligands offer additional
steric hindrance, leading to decreased structural relaxation in excited
state and consequently increased PLQYs. For example, it is observed
that the PLQY of [Cu(L2B)(PPh3)2]BF4 is higher than those of other [Cu
(N\\N)(PPh3)2]BF4 complexes obviously. This observation may be ex-
plained by the large rigid coplanar structure in L2B ligand. There may
be even inner- or inter-molecular stacking in [Cu(L2B)(PPh3)2]BF4,
which effectively limits its structural relaxation and results in a high
PLQY value [10].
Acknowledgments
The authors gratefully thank the financial supports of the NSFC
(Grant No. 51572256) and the Science and Technology Innovation Fund
Projects of Changchun University of Science and Technology (No. XJJLG-
2016-04).
For a better understanding on the radiative and non-radiative decay
procedures in [Cu(N\\N)(PPh3)2]BF4 complexes, their decay probability
values (Kr and Knr) are calculated with Formula (1) and Formula (2).
References
kr
Φ ¼
ð1Þ
ð2Þ
[1] M.W. Mara, K.A. Fransted, L.X. Chen, Coord. Chem. Rev. 282–283 (2015) 2–18.
[2] F. Dumur, Org. Electron. 21 (2015) 27–39.
[3] C.E. Housecroft, E.C. Constable, Chem. Soc. Rev. 44 (2015) 8386–8398.
[4] Y. Zhang, M. Schulz, M.W. chltler, M. Karnahl, B. Dietzek, Coord. Chem. Rev. 356
(2018) 127–146.
kr þ knr
1
τ
¼ kr þ knr
[5] D.G. Cuttell, S.-M. Kuang, P.E. Fanwick, D.R. McMillin, R.A. Walton, J. Am. Chem. Soc.
124 (2002) 6–7.
With their PLQY and τ values on hand, corresponding decay proba-
bility values are calculated and listed in Table 3. It is observed that Kr
varies in a wide region from 2.548 × 103 S−1 to 42.210 × 103 S−1
[6] S.M. Kuang, D.G. Cuttell, D.R. McMillin, P.E. Fanwick, R.A. Walton, Inorg. Chem. 41
(2002) 3313–3322.
[7] E. Leoni, J. Mohanraj, M. Holler, M. Mohankumar, I. Nierengarten, F. Monti, A.
Sournia-Saquet, B. Delavaux-Nicot, J.-F. Nierengarten, N. Armaroli, Inorg. Chem. 57
(2018) 15537–15549.
.
There is, however, no obvious correlation between Kr and conjugation
chain length. The existence of electron donor or acceptor exerts no di-
rect influence on Kr as well. This is because the emissive decay of triplet
MLCT&LLCT is actually controlled by a dynamic procedure of the lowest
emissive state and the potential surface crossing from π → π* sate to
MLCT state. This procedure has little effect with the energy level of di-
amine ligand. As for Knr, its value depends on the geometric structure
of diamine ligand greatly. Generally, a small diamine ligand with a
[8] L. Yang, J.K. Feng, A.M. Ren, M. Zhang, Y.G. Ma, X.D. Liu, Chem. Eur. J. 10 (2005)
1867–1879.
[9] Q. Zhang, Q. Zhou, Y. Cheng, L. Wang, D. Ma, X. Jing, F. Wang, Adv. Mater. 16 (2004)
432–436.
[10] L. Zhang, B. Li, Z. Su, Langmuir 25 (2009) 2068–2074.
[11] A.R. Cabrera, I.A. Gonzalez, D. Cores-Arriagada, M. Natali, H. Berke, C.G. Daniliuc, M.B.
Camarada, A. Toro-Labbe, R.S. Rojas, C.O. Salas, RSC Adv. 6 (2016) 5141–5153.
[12] L. Shen, X.L. Ding, T.F. He, X.L. Hao, J.F. Guo, L.Y. Zou, A.M. Ren, New J. Chem. 42
(2018) 3660–3670.

8
L. Zhang, Q. Zuo / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 223 (2019) 117280
[13] M. Alkan-Zambada, S. Keller, L. Martinez-Sarti, A. Prescimone, J.M. Junquera-
Hernandez, E.C. Constable, H.J. Bolink, M. Sessolo, E. Orti, C.E. Housecroft, J. Mater.
Chem. C 6 (2018) 8460–8471.
[14] S. Keller, A. Prescimone, H. Bolink, M. Sessolo, G. Longo, L. Martinez-Sarti, J.M.
Junquera-Hernandez, E.C. Constable, E. Orti, C.E. Housecroft, Dalton Trans. 47
(2018) 14263–14276.
[15] F. Brunner, A. Babaei, A. Pertegas, J.M. Junquera-Hernandez, A. Prescimone, E.C.
Constable, H.J. Bolink, M. Sessolo, E. Orti, C.E. Housecroft, Dalton Trans. 48 (2019)
446–460.
[17] R. Giereth, I. Reim, W. Frey, H. Junge, S. Tschierlei, M. Karnahl, Sustainable Energy
Fules 3 (2019) 692–700.
[18] L. Zhang, B. Li, Z. Su, J. Phys. Chem. C 113 (2009) 13968–13973.
[19] L. Zhang, B. Li, L. Zhang, Z. Su, ACS Appl. Mater. Interfaces 1 (2009) 1852–1855.
[20] M.A. Baldo, C. Adachi, S.R. Forrest, Phys. Rev. B 62 (2000) 10967–10977.
[21] P. Chen, L. Zhao, Y. Duan, Y. Zhao, W. Xie, G. Xie, S. Liu, L. Zhang, B. Li, J. Lumin. 2001,
131, 2144–2147.
[22] L. Zhang, S. Yue, B. Li, D. Fan, Inorg. Chim. Acta 384 (2012) 225–232.
[23] L. Chen, S. Sun, J. Li, Z. Fang, Inorg. Chim. Acta 446 (2016) 24–31.
[16] B. Chodhury, D. Mallick, K.K. Sarkar, K. Naskar, C. Sinha, Inorg. Chim. Acta 488 (2019)
125–130.