Systematic investigation of the metal-structure-photophysics relationship of emissive d10-complexes of group 11 elements: The prospect of application in organic light emitting devices

ARTICLE

pubs.acs.org/JACS

Systematic Investigation of the Metal-StructureꢀPhotophysics

Relationship of Emissive d¹⁰-Complexes of Group 11 Elements:

The Prospect of Application in Organic Light Emitting Devices

Chien-Wei Hsu,^†Chao-Chen Lin,^‡Min-Wen Chung,^‡Yun Chi,*^,†Gene-Hsiang Lee,^‡Pi-Tai Chou,*^,‡

Chih-Hao Chang,*^,§and Pin-Yang Chen^§

^†Department of Chemistry, National Tsing Hua University, Hsinchu 30013, Taiwan

^‡Department of Chemistry, National Taiwan University, Taipei 10617, Taiwan

^§Department of Photonics Engineering, Yuan Ze University, Chungli 32003, Taiwan

S Supporting Information

b

ABSTRACT: A series of new emissive group 11 transition metal

d¹⁰-complexes 1ꢀ8 bearing functionalized 2-pyridyl pyrrolide

together with phosphine ancillary such as bis[2-(diphenylphosp-

hino)phenyl] ether (POP) or PPh₃are reported. The titled

complexes are categorized into three classes, i.e. Cu(I) complexes

(1ꢀ3), Ag(I) complexes (4 and 5), and Au(I) metal complexes

(6ꢀ8). Via combination of experimental and theoretical ap-

proaches, the group 11 d¹⁰-metal ions versus their structural

variation, stability, and corresponding photophysical properties

have been investigated in a systematic and comprehensive manner. The results conclude that, along the same family, how much a

metal d-orbital is involved in the electronic transition plays a more important role than how heavy the metal atom is, i.e. the atomic

number, in enhancing the spinꢀorbit coupling. The metal ions with and without involvement of a d orbital in the lowest lying

electronic transition are thus classiﬁed into internal and external heavy atoms, respectively. Cu(I) complexes 1ꢀ3 show an

appreciable metal d contribution (i.e., MLCT) in the lowest lying transition, so that Cu(I) acts as an internal heavy atom. Despite its

smallest atomic number among group 11 elements, Cu(I) complexes 1ꢀ3 exhibit a substantially larger rate of intersystem crossing

(ISC) and phosphorescence radiative decay rate constant (k_r^p) than those of Ag(I) (4 and 5) and Au(I) (6ꢀ8) complexes

possessing pure π f π* character in the lowest transition. Since Ag(I) and Au(I) act only as external heavy atoms in the titled

complexes, the spinꢀorbit coupling is mainly governed by the atomic number, such that complexes associated with the heavier

Au(I) (6ꢀ8) show faster ISC and larger k_r^pthan the Ag(I) complexes (4 and 5). This trend of correlation should be universal and

has been ﬁrmly supported by experimental data in combination with empirical derivation. Along this line, Cu(I) complex 1 exhibits

intensive phosphorescence (Φ_p= 0.35 in solid state) and has been successfully utilized for fabrication of OLEDs, attaining peak EL

eﬃciencies of 6.6%, 20.0 cd/A, and 14.9 lm/W for the forward directions.

higher ligand ﬁeld strength, and the lower oxidation states of the

central metal ion all play pivotal roles in facilitating the heavy-metal

induced spinꢀorbit interaction as well as in the promotion of

MLCT contributions in the emissive states.⁴

The above interest then stimulates parallel studies aimed at

the investigation of emissive metal complexes that possess the

d¹⁰-electronic conﬁguration.⁵However, in contrast to the lumi-

nescent Cu(I) cluster complexes,⁶the majority of monometallic

Cu(I) complexesare cationic in nature, and their assembly requires

the incorporation of two diimine chelates for full occupation of the

tetrahedral coordination environment.⁷Recently, attention was

paid to the alternative architecture possessing one diimine plus

an ancillary bis-phosphine chelate,⁸due to the greatly enhanced

emission performance imposed by the highly rigid and stronger

1. INTRODUCTION

There has been considerable interest in the fundamental proper-

ties of luminescent transition metal complexes, for which the

characteristics of the lowest lying triplet excited states are domi-

nated by both ligand-centered ππ* and metal-to-ligand charge

transfer (MLCT) transitions.¹In moving toward realization of all

phosphorescent organic-light emitting devices (PhOLEDs), eﬀorts

have been made to design metal-based phosphors with emission

wavelengths spanning the whole visible region.²These research

works are mainly focused on the late transition metal containing

materials, for which the most recognizable paradigms involve the

third-row transition metal complexes with either d⁶- or d⁸-electro-

nic conﬁguration. Guided by this initiative, our group has initiated

researchonOs(II), Ir(III), and Pt(II) metal complexes withpyridyl

azolate and respective functionalized chelates, which are known to

constitute an integral component for harnessing eﬀective radiative

transitions.³It is believed that the rigid metalꢀchelate bonding,

Received: March 24, 2011

Published: June 28, 2011

r

2011 American Chemical Society

12085

dx.doi.org/10.1021/ja2026568 J. Am. Chem. Soc. 2011, 133, 12085–12099

|

Journal of the American Chemical Society

ARTICLE

metal-phosphine bonding. With respect to the photophysics,

their room temperature photoluminescence is attributed to the

MLCT transition, which is enhanced by both the low oxidation

potential of the Cu(I) core and the absence of a higher lying

metal centered dd-transition, the latter of which would result in

the weakening of metalꢀligand bonding and in turn promote the

radiationless deactivation in d⁶- and d⁸-metal complexes. In other

words, for these Cu(I) complexes, the highest occupied molec-

ular orbital (HOMO) has a predominant metal d character,

possibly mixed with a small contribution from the bis-phosphine,

while the lowest unoccupied molecular orbital (LUMO) is

essentially residing on the π* orbital localized on the diimine.

In the absence of the aforementioned metal dd state induced

quenching, they are thus capable of generating eﬃcient lumines-

cence and are prospective in the fabrication of OLEDs⁹as well as

the light-emitting electrochemical cell (LEC)¹⁰with relatively

low cost compared to the case of those constituted of Os(II),

Ir(III), and Pt(II) complexes. Another important issue of applic-

ability lies in that the Cu(I) complexes, in comparison to the late

transition metal complexes, are more environmentally benign.

Also, it is worthy of note that mobile ionic species are used in the

electroluminescent layers of LEC to yield eﬃcient light output;

thus, the employment of cationic Cu(I) complexes is especially

suitable.

Herein, we report the synthesis and characterization of a new

class of Cu(I) metal complexes with a series of tailor-made

pyrrolide anions, demonstrating various degrees of π-conjuga-

tion and spatial bulkiness. All of these compounds are lumines-

cent, and they are unique in their high quantum eﬃciency (Φ >

0.3 for complex 1) and tunable emission peak wavelengths

(550ꢀ605 nm) in CH₂Cl₂solution at room temperature, for

which their luminescent behaviors are comparable to those of the

three-coordinate Cu(I) complexes recently documented in the

literature.¹¹Experiments were then conducted to demonstrate

their usefulness in fabrication of OLEDs. Moreover, for further

analysis of the eﬀects exerted by the central Cu(I) metal, we have

also synthesized the Ag(I) as well as Au(I) counterparts of the

group 11 triads, cf. complexes 4ꢀ8, and then systematically

investigated their structural versus corresponding spectroscopic

and photophysical properties. In theory, the second and third

row analogues of group 11 d¹⁰-metal elements, on the one hand,

would reveal a steady increase of atomic number, molecular mass,

and ligandꢀmetal coordination bond strength. On the other

hand, they may concomitantly display a decrease of coordination

numbers and an increase of oxidation potential that play adverse

eﬀects toward achieving eﬀective luminescence. On this basis, we

then provide an in-depth analysis to elucidate a puzzling excep-

tion regarding why the Cu(I) complexes are deviated from the

“heavy-atom eﬀect”, a theory which states that increased inter-

system crossing (ISC) rates typically accompany complexes with

atoms of higher atomic number (Z). With the possession of

whole group 11 metal complexes, the interplay of countering

eﬀects on the luminescent characteristics can thus be judiciously

identiﬁed via the relevant photophysical studies. Such an inves-

tigation is complementary to the recent study on the relativistic

eﬀects in homogeneous Au(I) catalysis versus that of its lighter

congeners.¹²

mode or fast atom bombardment (FAB) mode. ¹H and ¹³C NMR spectra

were recorded on Varian Mercury-400 or INOVA-500 instruments;

chemical shifts are quoted with respect to the internal standard tetra-

methylsilane for ¹H and ¹³C NMR data. Elemental analyses were carried

out at the NSC Regional Instrumentation Center at National Chao Tung

University, Hsinchu, Taiwan. The chelating pyrroles, namely, 3,5-bis-

(trifluoromethyl)-2-(2⁰-pyridyl) pyrrole (fpyroH), 3,5-bis(trifluorome-

thyl)-2- (6⁰-phenyl-2⁰-pyridyl) pyrrole (fphroH), and 3,5-bis(trifluo-

romethyl)-2-(2⁰-isoquinolinyl) pyrrole (fiqroH), were prepared accord-

ing to the literature procedures,¹³while the copper complex [Cu(CH₃-

CN)₄]BF₄was synthesized by dropwise addition of aqueous HBF₄(50%)

into a suspension of Cu₂O in acetonitrile.¹⁴All reactions were conducted

under N₂atmosphere using anhydrous solvents or solvents treated with

an appropriate drying reagent.

Preparation of Cu(POP)(fpyro) (1). A 25 mL round bottle flask

was charged with [Cu(CH₃CN)₄]BF₄(50 mg, 0.16 mmol) in 10 mL of

acetonitrile and bis[2-(diphenylphosphino)phenyl] ether (POP, 85.6

mg, 0.16 mmol) in 10 mL of THF. After stirring at RT for 1 h, a solution

of 3,5-bis(trifluoromethyl)-2-(2⁰-pyridyl) pyrrole (fpyroH, 45 mg, 0.16

mmol) in 10 mL of THF was added. This mixture was stirred for another

5 h and solvent was then evaporated. Light green crystals were obtained

by diffusion of hexanes into a CH₂Cl₂solution of product at RT. Yield:

87 mg, 62%.

1

Spectral data for 1: MS (FAB, ⁶³Cu): m/z 881 [M⁺]. H NMR

(400 MHz, d₆-acetone, 298 K): δ 7.87ꢀ7.84 (m, 2H), 7.68 (t, J = 7.2 Hz,

1H), 7.40ꢀ7.38 (m, 3H), 7.30ꢀ7.22 (m, 10H), 7.16ꢀ7.08 (m, 7H),

6.98 (t, J = 7.6 Hz, 3H), 6.86ꢀ6.74 (m, 3H), 6.68ꢀ6.67 (m, 3H), 6.60

(t, J = 6.4 Hz, 1H). ³¹Pꢀ{¹H} NMR (202 MHz, d₆-acetone, 298 K):

δ ꢀ12.57 (br, 2P). ¹⁹Fꢀ{¹H} NMR (470 MHz, d₆-acetone, 298 K):

δ ꢀ54.19 (s, 3F), ꢀ59.22 (s, 3F). Anal. Calcd for C₄₇H₃₃CuF₆N₂OP₂:

C, 64.06; H, 3.77; N, 3.18. Found: C, 64.08; H, 3.89; N, 3.18.

Selected crystal data of 1: C₅₁H₄₁CuF₆N₂O₂P₂, M = 953.34, mono-

clinic, space group C2/c, T = 150(2) K, a = 37.5399(17), b = 12.9489(6),

c = 22.7366(10) Å, β = 127.247(1)ꢀ, V = 8798.0(7) Å³, Z = 8, F_calcd

=

,

1.439 mg/m³, F(000) = 3920, λ(Mo KR) = 0.7107 Å, μ = 0.639 mm^ꢀ1

crystal size = 0.50 ꢁ 0.40 ꢁ 0.15 mm³, 10105 independent reﬂections

collected with (R_int= 0.0324), GOF = 1.107, ﬁnal R₁[I > 2σ(I)] = 0.0584,

wR₂(all data) = 0.1742, and D-map, max./min. = 1.644/ꢀ1.121 e/Å³.

Preparation of Cu(POP)(fphro) (2). Similar procedures are

employed for the reaction of [Cu(CH₃CN)₄]BF₄with POP and

fphroH, giving light green crystals with a yield of 67%.

Spectral data for 2: MS (FAB, ⁶³Cu): m/z 957 [M⁺]. ¹H NMR (400

MHz, d₆-acetone, 298 K): δ 7.68 (d, J = 8.0 Hz, 1H), 7.59 (t, J = 7.6 Hz,

1H), 7.56 (d, J = 7.2 Hz, 1H), 7.40ꢀ7.20 (m, 20H), 6.92ꢀ6.84 (m, 5H),

6.80ꢀ6.72 (m, 5H), 6.60ꢀ6.58 (m, 4H). ³¹Pꢀ{¹H} NMR (202 MHz,

d₆-acetone, 298 K): δ ꢀ16.03 (br, 2P). ¹⁹Fꢀ{¹H} NMR (470 MHz,

d₆-acetone, 298 K): δ ꢀ53.57 (s, 3F), ꢀ60.74 (s, 3F). Anal. Calcd for

C₅₃H₃₇CuF₆N₂OP₂: C, 66.49; H, 3.90; N, 2.93. Found: C, 66.33; H,

4.25; N, 3.27.

Selected crystal data of 2: C₅₅H₄₁Cl₄CuF₆N₂OP₂, M = 1127.18,

monoclinic, space group P2₁/n, T = 150(2) K, a = 13.9226(9),

b = 20.2681(12), c = 18.1130(10) Å, β = 90.146(1)ꢀ, V = 5111.2(5)

Å³, Z = 4, F_calcd= 1.465 mg/m³, F(000) = 2296, λ(Mo KR) = 0.7107 Å,

μ = 0.763 mm^ꢀ1, crystal size = 0.42 ꢁ 0.40 ꢁ 0.35 mm³, 11736

independent reﬂections collected with (R_int= 0.0303), GOF = 1.029,

ﬁnal R₁[I > 2σ(I)] = 0.0443, wR₂(all data) = 0.1092, and D-map,

max./min. = 1.285/ꢀ0.910 e/Å³.

Preparation of Cu(POP)(fiqro) (3). Similar procedures are

employed for the reaction of [Cu(CH₃CN)₄]BF₄with POP and fiqroH,

giving yellow crystals with yield of 76%.

1

Spectral data for 3: MS (FAB, ⁶³Cu): m/z 931 [M⁺]. H NMR

2. EXPERIMENTAL SECTION

(500 MHz, d₆-acetone, 298 K): δ 8.31 (d, J = 9.5 Hz, 1H), 7.96 (d, J = 9.0

Hz, 1H), 7.85 (d, J = 8.5 Hz, 1H), 7.72 (d, J = 7.5 Hz, 1H), 7.50ꢀ7.36

(m, 4H), 7.36ꢀ7.26 (m, 4H), 7.24ꢀ7.08 (m, 8H), 7.08ꢀ6.96 (m, 6H),

General Information and Materials. Mass spectra were ob-

tained on a JEOL SX-102A instrument operating in electron impact (EI)

12086

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

1

6.92ꢀ6.80 (m, 5H), 6.79ꢀ6.69 (m, 3H), 6.63 (t, J = 7.7 Hz, 1H).

³¹Pꢀ{¹H} NMR (202 MHz, d₆-acetone, 298 K): δ ꢀ12.24 (br, 2P).

¹⁹Fꢀ{¹H} NMR (470 MHz, d₆-acetone, 298 K): δꢀ53.62 (s, 3F), ꢀ58.81

(s, 3F). Anal. Calcd for C₅₁H₃₅CuF₆N₂OP₂: C, 65.77; H, 3.79; N, 3.01.

Found: C, 65.88; H, 3.98; N, 3.37.

Spectral data for 7: MS (FAB, ¹⁹⁶Au): m/z 814 [M⁺]. H NMR

(500 MHz, CD₂Cl₂, 298 K): δ 7.93 (d, J = 7.0 Hz, 2H), 7.80 (t, J = 7.5

Hz, 1H), 7.68 (d, J = 8.0 Hz, 2H), 7.48 (t, J = 6.0 Hz, 3H), 7.35 (td, J = 7.5

Hz, 2 Hz, 6H), 7.24ꢀ7.16 (m, 6H), 7.09 (s, 1H), 6.95ꢀ6.92 (m, 3H).

³¹Pꢀ{¹H} NMR (202 MHz, CD₂Cl₂, 298 K): δ 32.3 (s, 1P). ¹⁹Fꢀ{¹H}

NMR (470 MHz, CD₂Cl₂, 298 K): δ ꢀ53.8 (s, 3F), ꢀ59.3 (s, 3F). Anal.

Calcd for C₃₅H₂₄AuF₆N₂P: C, 51.16; H, 2.97; N, 3.44. Found: C, 51.21;

H, 3.03; N, 3.26.

Preparation of Ag(POP)(fpyro) (4). A 50 mL round bottle flask

was charged with Ag₂O (22.5 mg, 0.10 mmol), POP (104.8 mg,

0.20 mmol), and fpyroH (54.5 mg, 0.20 mmol) in 30 mL of CH₂Cl₂.

After the mixture was stirred for 6 h at RT, the solvent was evaporated.

A colorless crystalline sample was obtained by diffusion of hexanes into a

CH₂Cl₂solution at RT. Yield: 150.8 mg, 81%.

Selected crystal data of 7: C₃₅H₂₄AuF₆N₂P, M = 814.50, triclinic, space

group P1, T= 150(2) K, a= 10.3486(5), b = 11.5288(5), c =13.3286(6) Å,

R = 83.087(1), β = 85.517 (1ꢀ, γ = 76.607(1)ꢀ, V = 1533.61(12) Å³, Z = 2,

F_calcd= 1.764 mg/m³, F(000) = 792, λ(Mo KR) = 0.7107 Å, μ =

4.913 mm^ꢀ1, crystal size = 0.40 ꢁ 0.32 ꢁ 0.25 mm³, 7022 independent

reﬂections collected with (R_int= 0.0282), GOF = 1.025, ﬁnal R₁[I > 2σ-

(I)] = 0.0220, wR₂(all data) = 0.0494, and D-map, max./min. =

1.202/ꢀ0.680 e/Å³.

Spectral data for 4: MS (FAB, ¹⁰⁷Ag): m/z 1032 [M⁺

+

¹⁰⁷Ag],

1

646 [M⁺ꢀ fpyro]. H NMR (500 MHz, d₆-acetone, 298 K): δ 7.89

(d, J = 4.5 Hz, 1H), 7.82 (d, J = 7.5 Hz, 1H), 7.71 (t, J = 7.5 Hz, 1H), 7.36

(t, J = 7.0 Hz, 4H), 7.32ꢀ7.20 (m, 18H), 7.30ꢀ7.22 (m, 10H), 7.01

(t, J = 7.5 Hz, 2H), 6.92 (d, J = 8.5 Hz, 3H), 6.83ꢀ6.74 (m, 4H).

³¹Pꢀ{¹H} NMR (202 MHz, d₆-acetone, 298 K): δ ꢀ8.62 (d, J_AgP= 400

and 347 Hz, 2P). ¹⁹Fꢀ{¹H} NMR (470 MHz, d₆-acetone, 298 K):

δ ꢀ53.7 (s, 3F), ꢀ60.3 (s, 3F). Anal. Calcd for C₄₇H₃₃AgF₆N₂OP₂:

C, 60.99; H, 3.59; N, 3.03. Found: C, 61.21; H, 3.87; N, 2.84.

Selected crystal data of 4: C₅₁H₄₁AgF₆N₂O₂P₂, M = 925.56, orthor-

hombic, space group Pbca, T = 150(2) K, a = 15.8784(7), b = 21.1674(9),

c = 23.6769(10) Å, V = 7957.9(6) Å³, Z = 8, F_calcd= 1.545 mg/m³,

F(000) = 3744, λ(Mo KR) = 0.7107 Å, μ = 0.656 mm^ꢀ1, crystal size =

0.30 ꢁ 0.15 ꢁ 0.07 mm³, 9149 independent reﬂections collected with

(R_int= 0.0700), GOF = 1.216, ﬁnal R₁[I>2σ(I)] =0.0690, wR₂(alldata) =

0.1385, and D-map, max./min. = 0.810/ꢀ0.862 e/Å³.

Spectral data for 8: MS (FAB, ¹⁹⁶Au): m/z 789 [M⁺+ 1]. ¹H NMR

(500 MHz, CD₂Cl₂, 298 K): δ 8.18 (d, J_HH= 8.0 Hz, 1H), 7.92 (d, J_HH

=

8.0 Hz, 1H), 7.78 (d, J_HH= 8.0 Hz, 1H), 7.54ꢀ7.47 (m, 10H), 7.43ꢀ7.34

(m, 6H), 7.32 (d, J_HH= 8.0 Hz, 1H), 7.12 (t, J_HH= 8.0 Hz, 1H), 6.96 (s,

1H). ³¹P NMR (202 MHz, CD₂Cl₂, 298 K): δ 32.4 (s, 1P). ¹⁹F NMR (470

MHz, CD₂Cl₂, 298 K): δ ꢀ53.7 (s, 3F), ꢀ59.5 (s, 3F). Anal. Calcd for

C₃₃H₂₂AuF₆N₂P: C, 50.27; H, 2.81; N, 3.55. Found: C, 50.09; H, 3.20;

N, 3.78.

Selected crystal data of 8: C₃₃H₂₂AuF₆N₂P, M = 788.46, monoclinic,

space group C2/c, T = 150(2) K, a = 13.9852(4), b = 10.6777(4),

c = 38.8186(14) Å, β = 97.0350(10)ꢀ, V = 5753.1(3) Å³, Z = 8,

F

_calcd= 1.821 mg/m³, F(000) = 3056, λ(Mo KR) = 0.7107 Å, μ = 5.236

Preparation of Ag(POP)(fiqro) (5). Similar procedures are

employed for the reaction of Ag₂O with POP and fiqroH in CH₂Cl₂,

giving yellow crystals with yield: 75%.

mm^ꢀ1, crystal size = 0.40 ꢁ 0.36 ꢁ 0.22 mm³, 6565 independent

reﬂections collected with (R_int= 0.0284), GOF = 1.120, ﬁnal

R₁[I > 2σ(I)] = 0.0253, wR₂(all data) = 0.0517, and D-map, max./

min. = 0.999/ꢀ1.311 e/Å³.

Spectral data for 5: MS (FAB, ¹⁰⁷Ag): m/z 1082 [M⁺

+

¹⁰⁷Ag],

646 [M⁺ꢀ ﬁqro]. H NMR (500 MHz, d₆-acetone, 298 K): δ 8.33

(d, J = 8.5 Hz, 1H), 8.04 (d, J = 9.0 Hz, 1H), 7.74 (d, J = 8.0 Hz, 1H), 7.60

(d, J = 8.5 Hz, 4H), 7.34ꢀ7.08 (m, 23H), 7.03 (t, J = 7.5 Hz, 2H), 6.98

(d, J = 8.5 Hz, 2H), 6.81ꢀ6.76 (m, 3H), 6.74 (t, J = 7.5 Hz, 1H).

³¹Pꢀ{¹H} NMR (202 MHz, d₆-acetone, 298 K): δ ꢀ7.40 (d, J_AgP= 393

and 341 Hz, 2P). ¹⁹Fꢀ{¹H} NMR (470 MHz, d₆-acetone, 298 K):

δ ꢀ52.9 (s, 3F), ꢀ59.8 (s, 3F). Anal. Calcd for C₅₁H₃₅AgF₆N₂OP₂: C,

62.78; H, 3.62; N, 2.87. Found: C, 62.56; H, 3.86; N, 3.26.

1

Photophysical Measurements. Steady-state absorption and

emission spectra in both solution and solid were recorded with a Hitachi

(U-3310) spectrophotometer and an Edinburgh (FS920) fluorimeter,

respectively. Phosphorescence lifetime measurements were performed

with an Edinburgh FL 900 photon-counting system. Further details,

including the procedures for calculating emission quantum yields, were

elaborated in a previous report.¹⁵As for quantum yield measurements in

the solid state, an integrating sphere (Labsphere) was applied, inwhich the

solid sample film was prepared via a vapor deposition method and was

excited by a series of Ar⁺laser lines, depending on the sample absorption

region. The resulting luminescence was acquired with an intensified

charge-coupled device for subsequent quantum yield analyses.¹⁶

The <300 ps lifetime of the ﬂuorescence bands in dual-emitting

complexes was measured with an Edinburgh OB 900-L time-correlated

single photon counting system.¹⁷The excitation pulses were derived

from the third or fourth harmonics of the signal output from a traveling-

wave optical parametric ampliﬁer of the white-light continuum

(TOPAS-C, Spectra Physics) system, seeded by an 1 kHz Ti:sapphire

ampliﬁer (Spitﬁre Pro, Spectra Physics).¹⁸After convolution, a system

response time of ∼50 ps can be conﬁdently achieved. For measuring the

relaxation dynamics of <50 ps, we then switched to the femtosecond

transient absorption system. Details of UV/vis/NIR transient absorp-

tion measurements have been elaborated in our previous report.¹⁹

Brieﬂy, a similar femtosecond Ti:sapphire ampliﬁer was used as the

light source (800 nm), which is split (50%) into two beams; one beam is

then converted to the designated excitation wavelength, i.e. 400 nm (the

pump beam), by coupling it into a second-harmonic generator. The

other 800 nm probe beam is then focused on a 1 mm thick sapphire plate

to generate a white light continuum (450ꢀ1000 nm), which after

passing through the sample cell, is coupled into a 100 μm optical ﬁber

connected to a diode array. The eﬀective time resolution of the ultrafast

spectrometer is determined to be about 250 fs.

Preparation of Au(PPh₃)(fpyro) (6). A 25 mL round bottle flask

was charged with NaH (7 mg, 0.29 mmol) in 5 mL of THF, and a

solution of 3,5-bis(trifluoromethyl)-2-(2⁰-pyridyl) pyrrole (fpyroH, 51.3

mg, 0.18 mmol) in 10 mL of THF. After stirring at 0 ꢀC for 15 min, the

resulting solution was transferred to the solution of ClAuPPh₃(90.7 mg,

0.18 mmol) in 20 mL of THF. This mixture was stirred for 6 h, and the

solvent was then evaporated. Colorless crystals were obtained by

diffusion of pentanes into THF at RT. Yield: 115 mg, 85%.

Spectral data for 6: MS (FAB, ¹⁹⁶Au): m/z 738 [M⁺]. ¹H NMR (500

MHz, CD₂Cl₂, 298 K): δ 8.08 (d, J = 5.0 Hz, 1H), 7.80 (d, J = 8.0 Hz, 1H),

7.70 (t, J = 8.0 Hz, 1H), 7.64ꢀ7.52 (m, 10H), 7.52ꢀ7.46 (m, 5H), 7.12

(t, J = 6.5 Hz, 1H), 6.92 (s, 1H). ³¹Pꢀ{¹H} NMR (202 MHz, CD₂Cl₂,

298 K): δ 32.9 (s, 1P). ¹⁹Fꢀ{¹H} NMR (470 MHz, CD₂Cl₂, 298 K):

δ ꢀ54.4 (s, 3F), ꢀ59.5 (s, 3F). Anal. Calcd for C₂₉H₂₀AuF₆N₂P: C,

47.17; H, 2.73; N, 3.79. Found: C, 47.06; H, 3.11; N, 3.92.

Selected crystal data of 6: C₂₉H₂₀AuF₆N₂P, M = 738.41, triclinic, space

group P1, T = 150(2) K, a = 8.7933(4), b = 10.1944(4), c = 16.6221(7) Å,

R = 73.1230(10), β = 75.2700(10), γ = 64.7300(10)ꢀ, V = 1274.71(9) Å³,

Z = 2, F_calcd= 1.924 mg/m³, F(000) = 712, λ(Mo KR) = 0.7107 Å,

μ = 5.900 mm^ꢀ1, crystal size = 0.16 ꢁ 0.12 ꢁ 0.12 mm³, 5819

independent reﬂections collected with (R_int= 0.0363), GOF = 1.030,

ﬁnal R₁[I > 2σ(I)] = 0.0292, wR₂(all data) = 0.0634, and D-map, max./

min. = 1.497/ꢀ0.689 e/Å³.

Both Au(PPh₃)(fphro) (7) and Au(PPh₃)(ﬁqro) (8) were prepared

in an analogous manner with yields of 87% and 85%, respectively.

12087

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Scheme 1. Structures of Group 11 Transition Metal Complexes: Cu(I) Complexes 1ꢀ3, Ag(I) Complexes 4 and 5, and Au(I)

Complexes 6ꢀ8

Computational Methodology. Calculations on electronic sing-

let states of all titled complexes were carried out using the density

functional theory (DFT) with B3LYP hybrid functional.²⁰A “double-ζ”

quality basis set consisting of Hay and Wadt’s effective core potentials

(LANL2DZ)²¹was employed for the Cu(I), Ag(I), and Au(I) metal

ions, and a 6-31G* basis set²²for the rest of the atoms. The relativistic

effective core potential (ECP) replaced the inner core electrons of all

metal atoms, leaving only the outer core valence electrons (ns²np⁶nd¹⁰,

n = 3, 4, and 5 for Cu(I), Ag(I), and Au(I), respectively) with which to be

concerned. Time-dependent DFT (TDDFT) calculations using the

B3LYP functional were then performed on the basis of the optimized

structures at ground states.²³Typically, 10 lower triplet and singlet roots

of the nonhermitian eigenvalue equations were obtained to determine

the vertical excitation energies. Oscillator strengths were then deduced

from the dipole transition matrix elements (for singlet states only). All

calculations were carried out using Gaussian 03.²⁴

the anticipated MLCT excited state.²⁵On the other hand, the

selection of POP chelate was also based on its greater rigidity and

ability to prevent the solvent attack at the central Cu(I) metal ion

by exerting excessive steric protection relative to two monodentate

phosphines.²⁶These measures are expected to enhance the

chemical stability under normal laboratory conditions as well

as the emission efficiency (vide infra). Without the ancillary POP

chelate, ready aggregation of the copperꢀpyrrolide fragment

occurred, affording a racemic mixture of zwitterionic trimer, such

as complexes with formula [Cu(fpyro)]₃.²⁷

The molecular drawings of Cu(I) metal complexes 1ꢀ3 are

depicted in Scheme 1. Their formulas were ﬁrst established by

mass analysis. Then their spectroscopic characteristics were further

conﬁrmed by ¹H, ¹⁹F, and ³¹P NMR, and the purity was veriﬁed by

subsequent microanalysis. In all cases, the ³¹Pꢀ{¹H} NMR

spectra exhibit a broad peak at around δ = ꢀ12, which is

independent of the measurement temperature. Apparently, the

broadening of the ³¹P NMR spectrum is caused by the quadruple

nuclei ⁶³Cu and ⁶⁵Cu; both possess a magnetic spin quantum

Electrochemical Measurement. Cyclic voltammetry (CV) mea-

surements were performed using a BAS 100 B/W electrochemical

analyzer. The potentials were recorded using Pt reference electrodes, in

anhydrous CH₂Cl₂with 0.1 M TBAPF₆as the supporting electrolyte, and

at the typical scan rate of 100 mV s^ꢀ1. All data were reported versus NHE

and calibrated with ferrocene as the internal standard (0.53 V).

3

number I = /₂and have natural abundances of 69% and 31%,

respectively.

Crystals of 1 and 2 suitable for X-ray analysis were obtained by

slow recrystallization from a CH₂Cl₂/hexane mixture at room

temperature. Their ORTEP diagrams are shown in Figure 1,

respectively, together with the selected bond lengths and angles.

In these two compounds, each of the copper atoms is located in

an N₂P₂distorted tetrahedral environment, in which both the

POP and the fphro (or fpyro) moieties are adopting a normal

coordination mode. In complex 1, the CuꢀP distances are nearly

equal, i.e. CuꢀP(1) = 2.2563(9) and CuꢀP(2) = 2.2344(9) Å,

and it possesses a much widened PꢀCuꢀP bite angle of ∼111ꢀ,

which is expected for the tetrahedral geometry. Moreover, the

molecule possesses two distinctive CuꢀN distances, with the

CuꢀN(pyrrolide) distance of 2.019(3) Å being comparatively

shorter than the CuꢀN(pyridine) distance of 2.102(3) Å, the

3. RESULTS AND DISCUSSION

3.1. Synthesis and Characterization. Cu(I) metal complexes

1ꢀ3 were obtained in good yield by mixing an equimolar amount

of bis[2-(diphenylphosphino)phenyl] ether (POP) and [Cu-

(CH₃CN)₄]BF₄, followed by addition of 1 equiv of appropriate

pyrrole precursors N^∧NH dissolved in THF solution, where

N^∧NH = fpyroH, fphroH, and fiqroH, for 1, 2, and 3, respec-

tively. On the one hand, these functionalized pyrrole chelates

designed according to the supposition that the 2-trifluoromethyl

group could increase the steric hindrance around the Cu(I) core

and hence inhibit the formation of the flattened conformation at

12088

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Figure 2. ORTEP diagram of Ag(I) complex 4 with ellipsoids shown at

the 30% probability level. Selected bond distances: AgꢀP(1) = 2.431(1),

AgꢀP(2) = 2.473(1), AgꢀN(1) = 2.406(4), and AgꢀN(2) = 2.265(4) Å.

Bond angles: N(1)ꢀAgꢀN(2) = 70.79(14) and P(1)ꢀAgꢀP(2) =

108.03(4)ꢀ.

doublets of nearly equal intensity, a consequence of the direct

coupling to both ¹⁰⁷Ag and ¹⁰⁹Ag nuclei with I = ¹/₂, owing to

their natural abundance of 51% and 49%, respectively. This

spectral feature also provides the initial spectral conﬁrmation to

the formation of Ag(I) metal complexes.

The structure of 4 was then established by X-ray diﬀraction

analysis. As revealed in Figure 2, complex 4 is essentially isostruc-

tural to the Cu(I) metal complex 1, except that all silverꢀligand

bond distances increase according to the increase of the ionic

radius of Ag(I) ion: AgꢀP(1) = 2.431(1), AgꢀP(2) = 2.473(1),

AgꢀN(1) = 2.406(4), and AgꢀN(2) = 2.265(4) Å. Moreover,

the Ag(I) analogue of Cu(POP)(fphro) (2) could not be

isolated, despite numerous attempts to optimize the conditions.

This suggests that the instability very likely originates from the

internal steric interaction between the ortho-phenyl substituent

and its adjacent ancillary POP, which then weakens the coordi-

native bonding of the anticipated Ag(I) metal complex.

Figure 1. ORTEP diagram of Cu(I) complexes with ellipsoids shown at

the 30% probability level. (a) ORTEP diagram of 1 and metric data:

CuꢀP(1) = 2.2563(9), CuꢀP(2) = 2.2344(9), CuꢀN(1) = 2.102(3),

CuꢀN(2) = 2.019(3) Å; N(2)ꢀCuꢀN(1) = 80.04(11) and P-

(2)ꢀCuꢀP(1) = 111.95(3)ꢀ. (b) ORTEP diagram of 2 and metric data:

CuꢀP(1) = 2.2500(6), CuꢀP(2) = 2.3325(6), CuꢀN(1) = 2.0278(17),

and CuꢀN(2) = 2.2653(17) Å; N(1)ꢀCuꢀN(2) = 77.44(7) and

P(1)ꢀCuꢀP(2) = 111.07(2)ꢀ.

The eﬀectiveness of our endeavors in obtaining all the required

d¹⁰-metal complexes is further exempliﬁed by the successful

preparation of third-row Au(I) metal complexes involving all three

pyrrolide chelates, forming Au(PPh₃)(fpyro) (6), Au(PPh₃)-

(fphro) (7), and Au(PPh₃)(ﬁqro) (8) depicted in Scheme 1. They

were best prepared by ﬁrst deprotonation of respective pyrrole

chelates with NaH in THF, followed by ﬁltration and transferring

this solution to a suspension of gold reagent ClAuPPh₃in THF and

stirring at RT for an extended period of time (see the Experimental

Section for details).

The crystal structures of Au(I) complexes 6ꢀ8 are then

determined, for which the ORTEP diagram and crucial metric

parameters are showed in Figure 3 with the metric data enclosed to

aid the discussion on the variation of metalꢀligand distances and

coordination modes. It is notable that all complexes show

result of which reﬂects the anionic nature of the pyrrolide versus

the neutral pyridyl group. In sharp contrast, due to the presence

of an ortho-substituted phenyl group on the pyridyl fragment of

complex 2, the unfavorable steric repulsion, as demonstrated by

the twisted arrangement of the phenyl substituent (see Figure 1b),

pushes the adjacent PPh₂fragment away from the coordinated

pyridyl group. The weakening of bonding is clearly shown by the

further elongation of the CuꢀP(2) and CuꢀN(2) distances (c.f.

complex 1) to 2.3325(6) Å and 2.2653(17) Å, respectively,

whereas the other CuꢀP(1) and CuꢀN(pyrrolide) distances are

still within the normal range, i.e. 2.2500(6) Å and 2.0278(17) Å,

respectively. Notably, the gross conformation of the POP moieties

remains unchanged and the respective metric parameters are

compatible to those reported in the literature.²⁶

In yet another approach, the synthesis of the isoelectronic,

second-row Ag(I) analogues was then executed using Ag₂O

instead of [Cu(CH₃CN)₄]BF₄in the presence of both POP

and respective pyrrolide chelates at RT, giving isolation of two

air-stable complexes, Ag(POP)(fpyro) (4) and Ag(POP)(ﬁqro)

(5), as depicted in Scheme 1. The key spectral features of 4 and 5,

again, involve the ³¹P NMR splitting pattern of two well resolved

monomeric structures with no obvious intermolecular Au Au

3 3 3

contacts. In addition, the PꢀAuꢀN(pyrrolide) bond angle is

found to be essentially linear (163.4ꢀ174.1ꢀ), which reﬂects the

retention of the typical 2-coordinate metalꢀligand arrangement

of Au(I) complexes.²⁸Meanwhile, the AuꢀN(pyridyl) distances

(2.631ꢀ3.004 Å) are elongated versus AuꢀN(pyrrolide) dis-

tances (2.061ꢀ2.086 Å). This variation of AuꢀN distances again

12089

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Figure 4. Absorption and normalized emission spectra of Cu(I) com-

plexes 1, 2, and 3 recorded in CH₂Cl₂solution at RT and the emission of

1 in a CH₂Cl₂matrix at 77 K.

POP ligands. As for complex 3, the absorption band at 350ꢀ

400 nm is assigned to the ππ* transition of the fiqro chelate,

which is red-shifted with respect to 1 and 2, due to the elongation

of π-conjugation provided by the quinolinyl fragment. On the

other hand, the addition of a phenyl group at the pyridyl site in 2

does not alter the spectral position with respect to that of 1,

indicating that the phenyl ring is not in conjugation with the

fpyro moiety, as can be seen from Figure 1. According to the

computational results (vide infra), the lowest lying transition for

1ꢀ3 should contain substantial MLCT (metal-to-ligand charge-

transfer) character (Table 2 and Figure 5 for selected frontier

orbitals of complex 3). However, the MLCT band is obscure in

the absorption spectra, possibly owing to its small transition

probability, i.e. the rather low absorptivity. For instance, in

complex 3, the higher lying transition, e.g. S₂, is essentially

localized on the fiqro ligand (ππ*), with oscillator strength f =

0.1628, while the involvement of the Cu(I) metal d-orbital is

clearly visualized in S₁, with relatively much lower transition

probability (f = 0.0363) and a small difference in energy versus

that of the higher lying S₂state.

In degassed CH₂Cl₂solution at RT, Cu(I) complexes 1ꢀ3 all

show a single emission band, maximized at 559, 551, and 603 nm

(Figure 4) with good quantum yields of 0.34, 0.28, and 0.12,

respectively. The emission intensity is drastically quenched by the

presence of oxygen in aerated solution. This, together with the

calculated radiative decay rate constant of ∼10⁴s^ꢀ1(see

Table 1), points out unambiguously that the emission is phos-

phorescence in origin.

To gain more insight into the emission properties, lumines-

cence spectra of 1ꢀ3 were also acquired in the 77 K CH₂Cl₂solid

matrix. The resulting emission for complex 1 is shown in Figure 4,

while those for 2 and 3 are depicted in Figure S1 of the Supporting

Information. For all 1ꢀ3, the emission is substantially blue-

shifted in the 77 K matrix relative to that recorded at 298 K. Note

that the phosphorescence origin of these emission bands remains

ascertained due to their >80 μs (instrumental range limited)

lifetime. Bearing in mind that the MLCT transition of d¹⁰Cu(I)

complexes 1ꢀ3 essentially removes a d electron from Cu(I), thus

forming a d⁹Cu(II) metal center, the results can be rationalized

by the fact that the original geometry of the tetrahedron-like d¹⁰

Cu(I) complexes is prone to distortion toward the square-planar-

like coordination required by the oxidized d⁹Cu(II) metal center

Figure 3. ORTEP diagram of Au(I) complexes with ellipsoids shown at

the 30% probability level: (a) ORTEP diagram of 6 and metric data:

AuꢀP(1) = 2.2160(9), AuꢀN(1) = 2.631(3), and AuꢀN(2) = 2.082(3)

Å; N(1)ꢀAuꢀN(2) = 69.28(11), P(1)ꢀAuꢀN(1) = 123.61(7),

P(1)ꢀAuꢀN(2) = 166.91(9)ꢀ. (b) ORTEP diagram of 7 and metric

data: AuꢀP(1) = 2.2345(7), AuꢀN(1) = 2.061(2), Au N(2) =

3 3 3

3.004(2) Å; P(1)ꢀAuꢀN(1) = 174.05(6)ꢀ. (c) ORTEP diagram of 8

and metric data: AuꢀP(1) = 2.2207(7), AuꢀN(1) = 2.681(2), AuꢀN(2)

= 2.086(2) Å; N(1)ꢀAuꢀN(2) = 69.25(8), P(1)ꢀAuꢀN(1) =

127.36(6), P(1)ꢀAuꢀN(2) = 163.38(7)ꢀ.

is consistent with the stronger nucleophilicity of the pyrrolide

fragment. Furthermore, complex 7 possesses the longest Auꢀ

N(pyridyl) distance of 3.004 Å, which is in good agreement with

the enhanced intramolecular steric constraint exerted by the

ortho-phenyl substituent.

3.2. MetalꢀLigandꢀPhotophysics Relationship. The room

temperature absorption and emission spectra of Cu(I) comple-

xes 1ꢀ3 are shown in Figure 4. Pertinent photophysical data are

listed in Table 1. In CH₂Cl₂, complexes 1 and 2 exhibit multiple

absorption peaks in the 200ꢀ370 nm region (ε > 10⁴M^ꢀ1cm^ꢀ1

)

that can be attributed to the ππ* absorption of both N^∧N and

12090

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Table 1. Photophysical and Electrochemical Data of Group 11 Complexes 1ꢀ8 Recorded in CH₂Cl₂at Room Temperature

d

λ_abs(nm) (ε ꢁ 10^ꢀ3M^ꢀ1cm^ꢀ1

)

λ_em(nm)^a

em Q.Y.^a,^b

τ_obs(μs)^a

k_r^p(s^ꢀ1

)

E^ox(ΔE)^c

E_HOMO

1

2

3

4

228 (47), 289 (23), 325 (20)

228 (57), 296 (24), 357 (10)

228 (89), 289 (48), 375 (26)

228 (39), 282 (24), 326 (12)

559; (481)

0.34; (0.35)

23.9

1.4 ꢁ 10⁴

1.1 ꢁ 10⁴

3.2 ꢁ 10³

6.88 ꢁ 10¹

0.43 (81)

0.54 (67)

0.53 (76)

0.74 (irr)

0.53

0.56

0.61

0.65

551; (490)

0.28; (0.05)

24.8

603; (553)

0.12; (0.51)

0.0075 [ﬂ], 0.0095; (0.14)^b

37

403 [ﬂ], 460, 477, 507;

(394 [ﬂ], 446, 476)

449 [ﬂ], 510, 546, 588;

(429 [ﬂ], 495, 532, 573)

491; (484)

176 ps [ﬂ], 138

5

228 (68), 284 (38), 369 (18)

0.014 [ﬂ], 0.021; (0.09)^b

75 ps [ﬂ], 157

1.34 ꢁ 10²

0.76 (irr)

0.62

6

7

8

228 (49), 308 (14)

0.037; (0.40)

0.01; (0.01)

0.12; (0.04)

124

17.4

79.7

2.98 ꢁ 10²

5.75 ꢁ 10²

1.46 ꢁ 10³

1.04 (irr)

1.07 (irr)

0.79

0.84

0.82

227 (42), 241 (41), 307 (11)

463, 490; (449, 475)

228 (59), 275 (26), 349 (14)

513, 542, 585; (503, 536)

^aData recorded in solid state are given in parentheses. Information regarding the ﬂuorescence band is marked with [ﬂ]. ^bThe emission quantum yield in

solid state is the sum of both ﬂuorescence and phosphorescence bands. ^cE^oxrepresents oxidation potential in V measured in CH₂Cl₂, using 0.1 M

(nBu)₄NPF₆as electrolyte and calibrated with ferrocene, and ΔE = |E_paꢀ E_pc| in mV or irreversible peak while marked with “(irr)”. ^dThe computed

energy level of the HOMO in terms of the normal hydrogen electrode (NHE).

in the emissive state, after MLCT excitation.²⁹As a result,

structural relaxation should take place in the low viscosity solution

at 298 K, rendering a bathochromic shift to the emission. On the

other hand, such structural distortion is hindered in the 77 K rigid

medium so that the phosphorescence emitting states of 1ꢀ3 retain

their ground-state geometries. Similar retardation of structural

relaxation is also observed in solid state at 298 K, as evidenced by

the >50 nm blue-shift of the phosphorescence peak wavelength

relative to that measured in CH₂Cl₂solution at 298 K (see

Table 1 and Figure S2). Thus, the rigidochromism, which has

been documented for other transition metal complexes,^29,30

seems to be operative for this class of Cu(I) complexes 1ꢀ3.

Supplementary support of the above viewpoint is given by the

computational approach. Based on the optimized ground-state

tetrahedral geometry, the energy of the lowest lying triplet state is

substantially deviated from the phosphorescence peak energy

acquired in CH₂Cl₂at 298 K. For example, the T₁energy of

complex 1 is calculated to be 22,676 cm^ꢀ1(441 nm in terms of

wavelength; see Table 2), which is much higher than the phos-

phorescence peak energy of 17889 cm^ꢀ1(559 nm). Similar

deviation is resolved for 2 and 3. However, as shown in Tables 1

and 2, and Figures 4 and S1, the calculated T₁energies of 441,

435, and 517 nm for 1, 2, and 3 are nearly identical with the ﬁrst

vibronic peaks of 446, 453, and 510 nm resolved in the 77 K

CH₂Cl₂matrix, further supporting the proposed rigidochromism

for the titled Cu(I) complexes.

because of directinvolvement ofthe metal d_πorbital, should invoke

signiﬁcant spinꢀorbit coupling matrix.^1d,32Moreover, mixing of

MLCT and ππ* in both singlet (S) and triplet (T) states leads to

an S f T intersystem crossing (ISC) process incorporating the

Æ¹d_ππ*|H_so|³ππ*æ or Æ³d_ππ*|H_so|¹ππ*æ term. The combination

of these two factors induces changes of orbital angular momen-

tum, i.e., d_πf π or vice versa, which may eﬀectively couple with

the ﬂip of electron spin. As a result, the transition has a

signiﬁcantly large ﬁrst-order spinꢀorbit coupling term, which

enhances the corresponding singletꢀtriplet mixing as well as the

rate of intersystem crossing.³³Conversely, for both S and T states

possessingpureππ* (eitherintraligand(ILCT) or ligand-to-ligand

(LLCT) character), the core metal can be empirically treated as an

element that virtually executes a small “external” heavy atom eﬀect

as opposed to the “internal” heavy atom eﬀect where a metal d_π

orbital is directly involved in MLCT. Together with the lack of

changes in orbital angular moment for S(ππ*)ꢀT(ππ*) con-

version, the S(ππ*)ꢀT(ππ*) mixing is small, and hence, a large

energy gap between S(ππ*) and T(ππ*) states (cf. S(MLCT/

ππ*) and T(MLCT/ππ*)) is expected.

Stemming from the above empirical approaches, for Cu(I)

complexes 1ꢀ3, due to the substantial MLCT contribution

(>20%) in S₁and, hence, a large spinꢀorbit coupling matrix, the

energy gap between S₁and its corresponding triplet state is

expected to be much reduced. Conversely, an S_m(ππ*) state

with negligible MLCT percentage is anticipated to have a larger

energy gap with its corresponding triplet state. Thus, despite that

the S₁(MLCT/ππ*) state is ascribed to the lowest excited singlet

state for all Cu(I) complexes, its correlated triplet state may not

necessarily be the lowest lying triplet state, i.e. T₁. In a more

general term, it could be ascribed as the T_nstate (n g 2) shown in

Scheme 2. For 1 and 3, as shown by the calculated data, the T₁state

possesses an essentially pure ππ* conﬁguration, while the T₂state

is comprised of appreciable MLCT character (see Table 2).

We thus conclude that the lowest lying T₁and second lowest

(T₂) triplet states correlate with S_m(ππ*) and S₁(MLCT/ππ*),

respectively. This viewpoint is reaﬃrmed by a good correlation

of the frontier orbitals involved for S₁and T₂, being mainly

contributed by HOMO-1 f LUMO with substantial MLCT%, as

well as for S_m(e.g., m = 2 for 3; see Table 2) and T₁, being mainly

associated with the HOMOf LUMO ππ* transition. As for 2, this

higher lying T_nis assigned to be T₃according to the correlation

shown in Table 2. This assignment is reasonable, since certain

More intriguingly, despite the fact that the S₁state has a

substantial MLCT (>20%) contribution, revealed by the frontier

orbital analyses (see Table 2 and Figure 5), the T₁state for 1ꢀ3

has ππ* conﬁguration in nature. Spectroscopically, the domination

of ππ* character of the T₁state is aﬃrmed by the well-resolved

vibronic progressions in the 77 K rigid medium. Furthermore,

upon careful inspection of the lower-lying electronic excited

states for 1ꢀ3, it is found that there exists a speciﬁc excited state

S_m(m g 2) mainly possessing intraligand ππ* charge transfer

(ILCT) character, which is similar to T₁(see Table 2).

The above results are of greater value in fundamental photo-

physics and can be rationalized on the following basis. First of all,

due to the good <π|π*> overlap integral and thus greater electron

exchange energy,³¹the diﬀerence in electron repulsion and, hence,

1

3

the energy splitting between ππ* and ππ* should be greater

than that between ¹MLCT and ³MLCT. Second, it has been well

established that states incorporating signiﬁcant MLCT character,

12091

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Table 2. Calculated Energy Levels, Oscillator Strengths (f), and Orbital Transition Analyses for Selected Lower-Lying Transitions

of Cu(I) Complexes 1ꢀ3

states

λ_cal

f

assignments

MLCT (%)

1

2

3

S₁

S₂

S₃

S₄

T₁

T₂

S₁

S₂

T₁

T₂

T₃

S₁

S₂

T₁

T₂

362.6

359.8

347.5

345

0.0021

HOMO-1 f LUMO (60%)

HOMO f LUMO (31%)

HOMO f LUMO (59%)

HOMO-1 f LUMO (29%)

HOMO-1 f LUMO+1 (44%)

HOMO f LUMO+1 (32%)

HOMO f LUMO (5%)

22.73

10.88

14.92

0

0.0021

0.0042

0.0091

0.0812

HOMO f LUMO+1 (85%)

HOMO f LUMO (16%)

HOMO f LUMO+1 (95%)

HOMO f LUMO (10%)

HOMO-1 f LUMO (26%)

HOMO f LUMO+6 (24%)

HOMO-1 f LUMO+1 (13%)

HOMO-1 f LUMO (74%)

HOMO f LUMO (11%)

HOMO-1 f LUMO+1 (7%)

HOMO f LUMO+1 (75%)

HOMO f LUMO (9%)

345

0.0812

0

441.4 (446)^a

364.5

0

14.33

31.12

0

364.5

0

364.5

0

356.1

0.006

356.1

0.006

356.1

0.006

353.2

0.0774

353.2

0.0774

0

434.5 (453)^a,^b

391.7

0

HOMO f LUMO+1 (54%)

HOMO f LUMO+3 (35%)

HOMO f LUMO+4 (12%)

HOMO f LUMO+1 (41%)

HOMO f LUMO+4 (21%)

HOMO f LUMO+3 (16%)

HOMO-1 f LUMO (76%)

HOMO-1 f LUMO+3 (6%)

HOMO-1 f LUMO+4 (5%)

HOMO-1 f LUMO (91%)

HOMO f LUMO (91%)

HOMO-1 f LUMO (5%)

HOMO f LUMO (86%)

HOMO-6 f LUMO (5%)

HOMO-1 f LUMO (67%)

HOMO-6 f LUMO (17%)

0

391.7

0

391.7

0

365.6

0

33.41

36.41

1.41

0

365.6

0

365.6

0

399.9

0.0363

388.7

0.1628

388.7

0.1628

510.3 (517)^a

430.9

0

26.96

430.9

^aValues in parentheses denote the ﬁrst vibronic peaks of the emission spectra measured under 77 K. ^bValue obtained from the halfway point of the slope

at the shorter wavelength side.

Scheme 2. Inverted Order of the Lower Lying Singlet-to-

T₁(ππ*) for 1. It is thus evident that for Cu(I) complexes 1ꢀ3,

relative to the singlet manifold, the sequence of triplet states has

been reversed in the order of energetics shown in Scheme 2. More

speciﬁcally, according to the data listed in Table 2, the lowest lying

S₁(MLCT/ππ*) versus its correlated T_n(MLCT/ππ*) state is

delineated to be S₁ꢀT₂, S₁ꢀT₃, and S₁ꢀT₂for Cu(I) complexes

1, 2, and 3, and their energy gap is calculated to be 0.42, 1.98, and

4.9 kcal/mol, respectively. These values are substantially smaller

than the energy gap between S_m(ππ*) and T₁(ππ*) for the same

series of complexes 1, 2, and 3, which is calculated to be 18.1

(S₄ꢀT₁), 15.3 (S₂ꢀT₁), and17.4 kcal/mol (S₂ꢀT₁), respectively.

Although the current theoretical level and basis sets applied in this

study are still subject to uncertainty and, thus, should be treated

as a semiquantitative approach, the resulting trend and relation-

ship ﬁrmly support the mechanism of large SꢀT (MLCT/ππ*)

versus small SꢀT (ππ*) mixing proposed above.

Triplet Levels for Cu(I) Complexes 1ꢀ3^a

^am and n are integer numbers (g 2).

higher lying S_m(ππ*) states may correlate with T₁(ππ*) due to

the presumably large energy gap. Likewise, there is no reason to

have T₁(ππ*) exactly correlated with S₂. For example, as shown

in Table 2, good correlation is found between S₄(ππ*) and

12092

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Figure 5. Selected frontier orbitals involved in the lower-lying electronic transitions of group 11 metal complexes 3, 5, and 8.

As the central atom goes down the d¹⁰group 11 family from

Cu(I) to Ag(I), due to the increase in nuclei charge and, hence,

stronger net attracting force, the d orbital should be further lowered

in energy, which, in turn, results in an increase of the metal

oxidation potential (see section 3.3). As a result, for the lowest

lying excited state in both singlet and triplet manifolds, null MLCT

contribution is found for Ag(I) complexes 4 and 5 (see Table 3 and

Figure 5 for the selected frontier orbitals for 5), which are iso-

structural to the Cu(I) complexes 1 and 3, respectively. For 5, the

ﬁrst singlet excited state carrying considerable MLCT is S₃(16%,

not shown here), and it is as high as S₆for complex 4 (14%, not

shown). Thus, in brief, for the titled Ag(I) complexes, both S₁and

T₁are dominated by the ligand-centered ππ* transition.

where H_sois the Hamiltonian for spinꢀorbit coupling and

ΔE_{S ꢀT}is the energy diﬀerence between S₁and its correlated

T_ns¹tates. Though pending quantitative assessment, due to the

direct involvement of the metal d_πorbital, the spinꢀorbit

coupling matrix of |ÆT_n(MLCT/ππ*)|H_so|S₁(MLCT/ππ*)æ|

for Cu(I) complexes 1ꢀ3 may be greater than that of

|ÆT₁(ππ*)|H_so|S₁(ππ*)æ| for Ag(I) complexes 4 and 5. In other

words, since the metal d_πorbitals are not involved in the lowest

lying transitions in both singlet and triplet manifolds, the Ag(I)

metal center in complexes 4 and 5 simply acts as an external

heavy atom, which does not bolster the spinꢀorbit coupling as

eﬀectively as the Cu(I) metal core, deﬁned in this study as an

internal heavy atom due to the extensive incorporation of MLCT.

Experimentally, the above viewpoint is fully justiﬁed by the

steady-state luminescent spectra and relevant photophysical prop-

erties. As shown in Figure 6 and Table 1, the Ag(I) complexes 4

and 5 clearly exhibit dual emission, consisting of a ﬂuorescence

[λ_max: 403 nm (4) and 449 nm (5)] and a phosphorescence band

[λ_max: 477 nm (4) and 546 nm (5)], supporting the slow rate of

n

As mentioned above, the rate of ISC is strongly inﬂuenced by

the types of transition involved. More speciﬁcally, the corre-

sponding ISC rate constant, k_isc, is expressed as

ꢀ

ꢁ

2

ψ jH_sojψ

T_n

S₁

k_isc

ꢁ

ð1Þ

2

ðΔE

Þ

S₁ꢀ T_n

12093

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Table 3. Calculated Energy Levels, Oscillator Strengths (f), and Orbital Transition Analyses for Selected Lower-Lying Transitions

of Ag(I) and Au(I) Complexes 4ꢀ8

a

states

λ_cal

f

assignment

MLCT (%)

4

5

6

S₁

S₂

364.6

340.1

0.0001

HOMO f LUMO (+100%)

HOMO f LUMO+1 (70%)

HOMO f LUMO+2 (27%)

HOMO f LUMO+4 (+91%)

HOMO f LUMO+9 (+10%)

HOMO f LUMO (90%)

HOMO f LUMO+9 (7%)

HOMO f LUMO (96%)

HOMO f LUMO+1 (99%)

HOMO f LUMO (83%)

HOMO-3 f LUMO (6%)

HOMO-3 f LUMO (51%)

HOMO-2 f LUMO (14%)

HOMO f LUMO (11%)

HOMO f LUMO (97%)

HOMO f LUMO+1 (97%)

HOMO f LUMO+3 (80%)

HOMO f LUMO (9%)

0.00

2.56

0.00

0.0008

S₂

340.1

0.0008

T₁

T₂

S₁

430.0

0

430.0

0

365.3

0

365.3

0

385.0

0.1725

S₂

353.0

0.0026

T₁

T₂

S₁

499.9 (492)

414.2

0

414.2

0

414.2

0

345.5

0.0093

S₂

343.6

0.0057

T₁

T₂

S₁

426.7

0

426.7

0

426.7

0

HOMO f LUMO+7 (8%)

HOMO f LUMO+1 (7%)

HOMO f LUMO+1 (49%)

HOMO f LUMO (35%)

HOMO f LUMO+7 (13%)

HOMO f LUMO (100%)

HOMO f LUMO+5 (61%)

HOMO f LUMO+6 (19%)

HOMO f LUMO+3 (12%)

HOMO f LUMO+1 (5%)

HOMO f LUMO (87%)

HOMO f LUMO (80%)

HOMO-2 f LUMO (11%)

HOMO-1 f LUMO (7%)

426.7

0

346.2

0

346.2

0

346.2

0

7

8

350.0

0.005

T₁

S₁

409.8

0

409.8

0

409.8

0

409.8

0

360.4

0.1917

T₁

482.7 (493)

0

^aValues in parentheses denote the ﬁrst vibronic peaks of the phosphorescence bands measured under 77 K.

S₁ꢀT₁intersystem crossing. Dynamically, as listed in Table 1,

the observed ﬂuorescence decay time (τ_obs) was resolved to be

176 and 75 ps for complexes 4 and 5, respectively. By knowing

the ﬂuorescence yield (Φ_f; see Table 1), the S₁f S₀ﬂuorescence

radiative decay time τ_r(τ_r= τ_obs/Φ_f) is deduced to be 23.4 and

5.35 ns for 4 and 5, respectively. Assuming that radiationless

deactivation pathways other than ISC are minor, the (176 ps)^ꢀ1

(4) and (75 ps)^ꢀ1(5) for the ﬂuorescence decay could thus be

considered as the rate constant of the S₁ꢀT₁intersystem crossing

excitation wavelength prohibits transient absorption measure-

ments for Cu(I) complexes 1 and 2. Alternatively, the transient

absorption experiment was carried out using 3. By monitoring the

tripletꢀtriplet absorption at 590 nm, as shown in Figure 7, a

single exponential rise component of 1.1 ps is resolved. Taking

the fast growth of tripletꢀtriplet absorption to be mainly

associated with S₁ꢀT_nintersystem crossing, the k_iscof (1.1 ps)^ꢀ1

for 3 is larger than its Ag(I) counterpart 5 by more than 50-fold.

It is also noteworthy that the fast S₁f T₁intersystem crossing

process of ∼10 ps has recently been reported for other Cu(I)

complexes.³⁵Thus, despite the fact that the Cu(I) core (Z = 29)

is less of a heavy atom than the Ag(I) counterpart (Z = 47), the

signiﬁcantly larger k_iscfor Cu(I) complexes 1ꢀ3 (c.f. Ag(I)

analogues 4 and 5) can be rationalized by fast S₁(MLCT/ππ*)

f T_n(MLCT/ππ*) ISC for 1ꢀ3 versus slow S₁(ππ*) f

T₁(ππ*) for 4 and 5.

process, which is apparently much smaller than that (g10¹²s^ꢀ1

)

reported for, e.g., typical Ir(III) complexes.^19,34

As for Cu(I) complexes 1ꢀ3, supported by the sole phos-

phorescence observed in steady-state (see Figure 4), k_iscis

expected to be faster than that of Ag(I) complexes 4 and 5.

Unfortunately, the intersystem crossing rate cannot be resolved

from the ﬂuorescence decay (or rise of the phosphorescence)

due to the response limit of ∼50 ps (see Experimental Section)

for the current time-correlated single photon counting setup.

Hence, further resolution of k_iscis provided by the femto-

second transient absorption measurement. The lack of suitable

Going further down the periodic table while replacing the

central metal atom with the heaviest core, Au(I) complexes 6ꢀ8

exhibit absorption spectra resembling those of their Ag(I)

counterparts 4 and 5 (see Figure 8 and Table 1, and also Figure 5

12094

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Figure 6. Absorption and normalized emission spectra of Ag(I) com-

plexes 4 and 5 recorded in CH₂Cl₂solution at RT and the emission of 5

at 77 K.

Figure 8. Absorption and normalized emission spectra of Au(I) com-

plexes 6, 7, and 8 recorded in CH₂Cl₂solution at RT and the emission of

8 at 77 K.

Scheme 3. Correlation between the Lower Lying Singlet and

Triplet Levels for Ag(I) and Au(I) Complexes 4ꢀ8^a

^am and n are integer numbers (g2).

On the basis of the above discussion, the correlation between

the lower lying single and triplet levels for Ag(I) and Au(I)

complexes 4ꢀ8 can be depicted in Scheme 3. Upon excitation,

the S₁state of the Ag(I) and Au(I) complexes, which possesses

essentially only the ligand-centered ππ* character, in theory,

should not alter the optimized geometry for the d¹⁰conﬁgura-

tion. Therefore, unlike the cases of Cu(I) complexes 1ꢀ3,

structural deformation is not expected to take place for 4ꢀ8 in

the excited state. Experimentally, this viewpoint is ﬁrmly sup-

ported by the emission spectra acquired in the 77 K rigid

medium. As shown in Figures 6 and 8, the emission spectra of

Ag(I) complex 5 and Au(I) complex 8, for example, show similar

peak wavelengths and spectral proﬁles for their phosphorescence

bands between 298 and 77 K in CH₂Cl₂, as opposed to the

substantial spectral diﬀerence observed for the Cu(I) complexes

1ꢀ3 (vide supra).

Figure 7. Rise dynamics of tripletꢀtriplet absorption of 3 monitored at

590 nm in CH₂Cl₂. A single exponential rise time is ﬁtted to be 1.1 ps.

Inset: The selected transient absorption spectra at various delay times,

with excitation wavelength λ_ex= 400 nm.

for the selected frontier orbitals of 8). It has been reported that

the metal d orbital of the Au(I) core is even lower in energy

compared to those of the Cu(I) and Ag(I) complexes,³⁶so that

the lowest lying state should be dominated by the ligand-

centered ππ* transition in both singlet and triplet manifolds.

This viewpoint is also ﬁrmly supported by the computational

results listed in Table 3, concluding null MLCT contribution for

both S₁and T₁states. In sharp contrast to the dual emission

(ﬂuorescence/phosphorescence) resolved in 4 and 5, however,

the emission spectrum of Au(I) complexes 6ꢀ8 consists solely of

phosphorescence (see Figure 8). In fact, for 6ꢀ8, the large

energy separation between S₁(ππ*) and T₁(ππ*) states can be

perceived via the large Stokes-shift for the phosphorescence

emission, deﬁned as the diﬀerence in peak wavelength between

(the lowest lying) absorption and emission. The large S₁ꢀT₁

(ππ*) energy gap is also evidenced by the calculated values 15.5,

12.3, and 20 kcal/mol for complexes 6, 7, and 8, respectively,

according to Table 3.

Evidently, without MLCT involvement in the lower lying

excited states, the Au(I) core also acts as an external heavy atom.

Approximating the metal ion as hydrogen-like, a simpliﬁed ap-

proach predicts that the rate constant of ISC is proportional

to Z⁸, where Z stands for the atomic number.³⁷With the highest

atomic number among group 11 elements, the Au(I) core

(Z = 79) is expected to induce a stronger external heavy atom

eﬀect than that of the Ag(I) core (Z = 47), facilitating the

intersystem crossing for complexes 6ꢀ8 (c.f. Ag(I) complexes 4

and 5). An attempt to resolve k_iscfor 6ꢀ8 via monitoring the

lifetime for the ﬂuorescence band (which is not observable

in steady state) or the rise time of the phosphorescence by a

12095

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

time-correlated single photon counting experiment was not suc-

cessful due to the much larger k_isc, i.e. >(50 ps)^ꢀ1. We also made

eﬀorts in an attempt to resolve k_iscvia femtosecond transient

absorption for complex 8 in CH₂Cl₂(λ_ex: 266 nm). Unfortu-

nately, complex 8 was subject to photodecomposition under

such high energy excitation. Nevertheless, the instrumental

response-limited time constant (>(50 ps)^ꢀ1) of k_iscfor 6ꢀ8

attests to accelerated intersystem crossing relative to that of

Ag(I) complexes 4 and 5.

Moreover, the oxidation potentials of both Ag(I) and Au(I)

complexes were also moved to the more positive region and

became irreversible. Such irreversible oxidation versus the re-

versible nature for Cu(I) complexes (1ꢀ3) indirectly reaﬃrms

the dominant metal participation for Cu(I) complexes as op-

posed to the pure ligand-centered nature for the Ag(I) and Au(I)

counterparts (4ꢀ8), regarding the highest occupied molecular

orbitals. This viewpoint is supported by the 0.74 and 0.76 V

measured for 4 and 5 in the Ag(I) series and the 1.04, 1.07, and

1.07 V for 6, 7, and 8, respectively, in the Au(I) series. Overall, the

oxidation potentials follow the order 6ꢀ8 > 4ꢀ5 > 1ꢀ3. We also

calculated both HOMO and LUMO energies of the titled

complexes using the DFT method. As shown in Table 1, the

trend of the HOMO energy is also consistent with the above

electrochemical measurement (also see Table S1 for the LUMO).

Since the metal d orbital should be even lower in energy than the

ligand π orbitals for the titled Ag(I) and Au(I) series, the results

are consistent with the periodic properties of group 11 metal

triads, for which the oxidation potential is in the order Au(I) >

Ag(I) > Cu(I). On the other hand, the observed discrepancy

between Ag(I) and Au(I) complexes may be alternatively

attributed to the diﬀerence in coordination environment (four-

versus three-coordination).

As to harvest the T₁f S₀phosphorescence via its radiative

transition, the rate constant k^p_rcan be expressed as follows:

ꢀ

ꢁ

ψ jH_sojψ ²μ²_S

ꢀ

ꢁ

k^pꢁ ψ jH_erjψ ²≈ γ

,

S₁

T₁

1

S₀

T₁

r

2

ðΔE_S

Þ

ꢀ T₁

1

16π³10⁶n³E³_em

γ ¼

ð2Þ

3hε₀

where H_erdenotes the electric dipole operator created by the

electric magnetic ﬁeld, μ_Sis the S₀f S₁transition dipole

1

moment, E_emrepresents the T₁ꢀS₀energy gap in cm^ꢀ1, and n, h,

and ε₀are the refractive index, Planck’s constant, and the

permittivity in vacuum, respectively.^30a,38

3.4. OLED Device Fabrication. Following the aforementioned

photophysical results and discussion, once the nonradiative decay

rate is of a similar magnitude, harvesting phosphorescence de-

pends strongly on the radiative decay rate constant. This seems

to be the case for the titled complexes. In degassed CH₂Cl₂

solution, Cu(I) complexes exhibit decent to intense (Φ_p= 0.34)

emission, while Au(I) and Ag(I) complexes render weaker and

the weakest emission yield, respectively, for which the result

correlates well with respect to the order of the radiative decay

rate constants (vide supra, see Table 1). Apparently, parts of the

nonradiative deactivation pathways are caused by the large

amplitude vibration motions that occur in solution, which are

expected to reduce or even diminish in solid state, as evidenced

by the intense emission for, e.g., Ag(I) complex 4 (Φ_p= 0.14)

and Au(I) complex 6 (Φ_p= 0.4) recorded as solid powder. Note

that similar dual emissions (fluorescence and phosphorescence)

are resolved for the Ag(I) complexes 4 and 5 in solid state, as

shown in Figure S3. Nonetheless, for practical concerns such as

cost and natural abundance, application to OLEDs is only

focused on Cu(I) based phosphors, among which complex 1 is

paid special attention due to its high emission quantum yield in

both solution and solid state (see Table 1).

In this approach, 1,3-bis(carbazol-9-yl)benzene (mCP) host

material and a series of electron transport materials were utilized

for fabrication ofOLEDs as well asto investigatethe carrierbalance

during device operation. All studied devices consisted of a simpli-

ﬁed three-layer architecture of ITO/TAPC (40 nm)/mCP doped

with 8 wt % of 1 (30 nm)/electron-transport layer (ETL)

(40 nm)/LiF (0.8 nm)/Al (150 nm), for which the graphical

conﬁgurations and structural drawings of organic materials are

depicted in Figure 9. Host material mCP was selected in these

studies because it is well-known to allow eﬃcient host-to-guest

energy transfer and eﬀective exciton conﬁnement in the emitting

layer.⁴⁰After optimization, satisfactory performance was obtained

using 1,3,5-tri[(3-pyridyl)phen-3-yl]benzene (TmPyPB) and tris-

(2,4,6-trimethyl-3-(pyridin-3-yl)phenyl)borane (3TPyMB) as the

electron transport/hole blocking layer for devices A and B, res-

pectively.⁴¹Device A showed the largest peak brightness of 4988

cd/m²at 12.8 V, while peak eﬃciencies of 2.2%, 6.3 cd/A, 3.7 lm/

As evidenced by the similar absorption extinction coeﬃcients

(see Figures 4, 6, and 8 and Tables 1), for simplicity, the S₀f S₁

transition dipole moment μ_Sis assumed to be the same for all

1

complexes 1ꢀ8. Together with the approximation that E_emis the

same for all of them, the simpliﬁcation makes it feasible to

2

correlate Æψ_S|H_so|ψ_Tæ²/(ΔE_{S ꢀT}

)

with the phosphores-

cence radiativ¹e decay ¹rate constant (k_r). As derived above,

1

Æψ_S|H_so|ψ_Tæ²/(ΔE_{S ꢀT}

)

increases in the order of Cu(I)

2

1

complexes (1 ꢀ 3) > Au(I) complexes (6ꢀ8) > Ag(I) complexes

(4 and 5), which correlates well with the trend of k^p_r: 10⁴s^ꢀ1

(1ꢀ3) > 3 ꢁ 10²∼10³s^ꢀ1(6ꢀ8) > ∼10²s^ꢀ1(4 and 5) (see

Tables 1). For both the Ag(I) and Au(I) analogues, negligible

MLCT contribution in the emitting T₁state and, hence, reduc-

tion of the T₁ꢀS₁coupling matrix is manifested by the rather

small T₁f S₀radiative (phosphorescence) decay rate constants

(k^p_r), on the order of <10³s^ꢀ1. While apparently attributed to the

involvement of MLCT, the k^p_rvalue of 10⁴s^ꢀ1for the Cu(I)

complexes is still less than that of typical Ir(III) complexes

(10⁵ꢀ10⁶s^ꢀ1) by more than 1 order of magnitude.^19,39This is

mainly dueto the much lighterCu(I) core (Z = 29) compared with

Z = 77 for the Ir(III) core, resulting in a signiﬁcant decrease of the

H_soterm (see eq 2).

3.3. Electrochemistry. The electrochemical behavior of these

group 11 metal complexes was investigated by cyclic voltamme-

try using ferrocene as the internal standard. The respective redox

data are also listed in Table 1. During the anodic scan in CH₂Cl₂,

the Cu(I) metal complexes 1, 2, and 3 exhibited reversible peak

potential at 0.43, 0.54, and 0.53 V, respectively. It is noted that

complex 1, with the smallest degree of π-conjugation at the

pyrrolide chelate, shows the lowest potential. This contradictory

observation indicates that the oxidation is a localized metal-

centered process. In addition, the pyrrolide chelate of 2 and 3

may exert greater steric hindrance on the formation of the square-

planar geometry required by the oxidized d⁹Cu(II) metal center.

Consequently, their oxidation potentials are expected to be more

positive than that of the parent Cu(I) complex 1; in other words,

1 should give the highest propensity and least amount of barrier

in forming the flattened geometry.

12096

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Figure 9. Schematic device structures, and the structural drawing and

energy levels of the materials employed for OLED fabrication.

Figure 10. (a) EL spectra, (b) current densityꢀvoltageꢀluminance

(IꢀVꢀL) characteristics, and (c) external quantum eﬃciency/luminance

eﬃciency vs luminance for the OLED devices A, B, and C fabricated using

Cu(I) metal complex 1 as dopant.

W, and 5.3%, 15.4 cd/A, 11.7 lm/W were realized with devices A

and B, respectively. The electroluminescence (EL) characteristics

and the numeric data are shown in Figure 10 and Table 4. Other

tested devices using lower-gap ETLs showed much inferior

eﬃciencies, which are not shown here to save space.

conﬁrming the aforementioned quenching eﬀect at higher cur-

rent densities. In yet another approach, it seems reasonable that

one can further reduce the exciton quenching while preserving

the carrier balance simultaneously by inserting a thin 3TPyMB

layer between the emitting layer and the TmPyPB layer. This

thinking led us to fabricate the third OLED, namely device C,

which incorporates a double ETL architecture: ITO/TAPC

(40 nm)/mCP doped with 8 wt % of 1 (30 nm)/3TPyMB

(3 nm)/TmPyPB (37 nm)/LiF (0.8 nm)/Al (150 nm). The total

thickness of the ETL was then ﬁxed at 40 nm to make a fair

comparison, while 3 nm of 3TPyMB gave the optimal experi-

mental outcome.

As expected, the IꢀVꢀL characteristics of device C were

similar to those of device B (Figure 10 and Table 4), and the EL

eﬃciencies were higher than those of device B, indicating that the

addition of a thin 3TPyMB layer had indeed retarded the electron

transport. Although the improvement in peak luminance was only

moderate in device C, other device eﬃciency data have shown

It is noteworthy that both TmPyPB and 3TPyMB possess

much larger triplet-energy gap and ionization potential, such that

excessive holes are expected to accumulate at the emitting layer

(EML)/ETL interface.⁴¹The accumulation of carriers at the

heterojunction causes the build-up of a localized electrical ﬁeld,

which not only alters the carrier transportation but also results in

severe exciton quenching.⁴²This phenomenon is also related to

the diﬀerence in the electron transport ability. The electron

transport of TmPyPB (∼10^ꢀ3cm²/(V s)) is faster than that

(∼10^ꢀ5cm²/(V s)) of 3TPyMB by 2 orders of magnitude.⁴¹The

better electron transport capability in TmPyPB might reduce the

carrier accumulation at the interface and diminish the localized

electrical ﬁeld. As a consequence, device A exhibited a substan-

tially higher peak luminance of ∼5000 cd/m². In stark contrast,

the peak luminance of device B was merely ∼560 cd/m², thus

12097

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Table 4. EL Characteristics of OLEDs with Diﬀerent ETLs

max. B (cd/m²) (V)

CIE coordinates^b

η_ext(%)

η_l(cd/A)

η_p(lm/W)

V_on(V)

a

b

a

b

a

b

c

device A (TmPyPB)

2.2

5.3

6.6

7.8

2.2

1.9

3.8

0.5

6.3

15.4

20.0

23.2

6.1

5.5

3.7

11.7

14.9

13.4

3.3

2.6

5.7

0.3

4.5

4.1

4.2

5.2

4988 (12.8 V)

558 (15.0 V)

598 (12.6 V)

415 (20.0 V)

(0.339, 0.500)

(0.312, 0.515)

(0.308, 0.522)

(0.322, 0.518)

device B (3TPyMB)

device C (3TPyMB/TmPyPB)

11.3

1.4

device D (nondoped)

^aMaximum eﬃciencies. ^bEﬃciency data recorded at 10²cd/m². ^cTurn-on voltage measured at 1 cd/m².

signiﬁcant improvement. Overall, device C exhibited peak EL

eﬃciencies of 6.6%, 20.0 cd/A, and 14.9 lm/W for the forward

directions. At the practical brightness of 100 cd/m², the forward

eﬃciency remained high at around 3.8%, 11.3 cd/A, and 5.7

lm/W. With these encouraging results, these newly developed Cu(I)

metal complexes show promising potential in the ﬁeld of OLEDs.

Nevertheless, according to the energy levels shown in Figure 9c,

the HOMO of mCP and the LUMO of 3TPyMB appear

well suited to generate exciplex-type emission near 2.6 eV

(∼ 480 nm).⁴³Electroluminescence from such an interfacial

exciplex would overlap with the emission of Cu(I) complex 1

and contribute to the higher eﬃciency observed. Accordingly, the

nondoped device was fabricated to further identify the origin of

emission in device C. Thus, a new architecture D was fabricated,

which consisted of ITO/TAPC (40 nm)/1 (30 nm)/3TPyMB

(3 nm)/TmPyPB (37 nm)/LiF (0.8 nm)/Al (150 nm). The

relevant graphical device diagram and associated eﬃciency

data are shown in Figure S5 of the Supporting Information. As

shown in Table 4, the listed data of D were essentially identical

to that of all other devices, eliminating the possible formation

of exciplex. Moreover, although the peak eﬃciencies of device

D were slightly higher than that of C, the eﬃciencies drop-

ped notably at higher brightness and the turn-on voltage of

D increased to 5.2 V. Such a feature unambiguously con-

ﬁrmed the necessity of the hostꢀguest system used in the

present study.

group 11 metal complexes increase upon increasing the atomic

number down the group. While pending quantitative proof of

concept, a similar mechanism is believed to be operative on other

transition metal families. Finally, from the application point of

view, the low-cost Cu(I) metal element is highly desired for use

as lighting materials. As such, complex 1 has been successfully

applied as a dopant to fabricate high eﬃciency OLEDs. In the

preliminary approach, the as-prepared device is capable of exhibit-

ing peak EL eﬃciencies of 6.6%, 20.0 cd/A, and 14.9 lm/W for

the forward directions. The combination of fundamental synthetic

strategy and encouraging results in potential OLEDs applications

should make these group 11 d¹⁰metal complexes highly attrac-

tive to a broad spectrum of research ﬁelds.

’ ASSOCIATED CONTENT

S

Supporting Information. X-ray crystallographic data ﬁle

b

(CIF) and solid state and lower temperature emission spectra of

the studied complexes. This material is available free of charge via

the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION

Corresponding Author

ychi@mx.nthu.edu.tw; chop@ntu.edu.tw; chc@saturn.yzu.edu.tw

’ ACKNOWLEDGMENT

This work was supported by the NSC and MEA of Taiwan.

We are also grateful to the National Center for High-Performance

Computing for computer time and facilities.

4. CONCLUSION

In conclusion, we have designed and synthesized a series of

emissive, group 11 d¹⁰transition metal complexes bearing various

tailor-made 2-pyridyl pyrrolide chromophores. The compilation of

Cu(I), Ag(I), and Au(I) analogues renders systematic discussion

on the core-metal dependent variation of the structures feasible.

Also, for the ﬁrst time, the group 11 d¹⁰metal core versus its

associated photophysical properties is investigated in a compre-

hensive manner. We have proved unambiguously that, for the

titled metal complexes, direct involvement of the d orbital in the

lowest lying electronic transition plays a major role in inducing

spinꢀorbit coupling compared to the cases of complexes without

metal d orbital involvement in the same transition. The former

and latter are dubbed respectively as the internal and external

heavy atom eﬀect in this study. As a result, despite the larger

atomic numbers for the Ag(I) and Au(I) metal cores, the rate

constants of S₁ꢀT₁intersystem crossing and T₁ꢀS₀radiative

decay of the associated complexes, due to the dominant ππ*

character, are signiﬁcantly smaller than those of the Cu(I)

complexes having appreciable MLCT contribution. Once acting

as external heavy atom, our experimental data indicate that the

intersystem crossing (ISC) and radiative decay rate constants of

’ REFERENCES

(1) (a) Baba, A. I.; Shaw, J. R.; Simon, J. A.; Thummel, R. P.;

Schmehl, R. H. Coord. Chem. Rev. 1998, 171, 43. (b) Armaroli, N. Chem.

Soc. Rev. 2001, 30, 113. (c) Hay, P. J. J. Phys. Chem. A 2002, 106, 1634.

(d) Chou, P.-T.; Chi, Y. Chem.—Eur. J. 2007, 13, 380. (e) Vlcek, A., Jr.

Top. Organomet. Chem. 2010, 29, 73.

(2) (a) Xiao, L.; Chen, Z.; Qu, B.; Luo, J.; Kong, S.; Gong, Q.; Kido,

J. Adv. Mater. 2011, 23, 926. (b) Sasabe, H.; Kido, J. Chem. Mater. 2011,

23, 621. (c) Chen, Z.-Q.; Bian, Z.-Q.; Huang, C.-H. Adv. Mater. 2010,

22, 1534. (d) Zhao, Q.; Li, F.; Huang, C. Chem. Soc. Rev. 2010, 39, 3007.

(e) Wong, W.-Y.; Ho, C.-L. Coord. Chem. Rev. 2009, 253, 1709. (f)

Wong, W.-Y.; Ho, C.-L. J. Mater. Chem. 2009, 19, 4457. (g) Ulbricht, C.;

Beyer, B.; Friebe, C.; Winter, A.; Schubert, U. S. Adv. Mater. 2009,

21, 4418. (h) You, Y.; Park, S. Y. Dalton Trans. 2009, 1267.

(3) (a) Chou, P.-T.; Chi, Y. Eur. J. Inorg. Chem. 2006, 3319. (b) Chi,

Y.; Chou, P.-T. Chem. Soc. Rev. 2007, 36, 1421. (c) Chi, Y.; Chou, P.-T.

Chem. Soc. Rev. 2010, 39, 638.

(4) (a) Tung, Y.-L.; Lee, S.-W.; Chi, Y.; Chen, L.-S.; Shu, C.-F.; Wu,

F.-I.; Carty, A. J.; Chou, P.-T.; Peng, S.-M.; Lee, G.-H. Adv. Mater. 2005,

17, 1059. (b) Tung, Y.-L.; Lee, S.-W.; Chi, Y.; Tao, Y.-T.; Chien, C.-H.;

12098

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Journal of the American Chemical Society

ARTICLE

Cheng, Y.-M.; Chou, P.-T.; Peng, S.-M.; Liu, C.-S. J. Mater. Chem. 2005,

15, 460. (c) Fang, C.-H.; Chen, Y.-L.; Yang, C.-H.; Chi, Y.; Yeh, Y.-S.; Li,

E. Y.; Cheng, Y.-M.; Hsu, C.-J.; Chou, P.-T.; Chen, C.-T. Chem.—Eur.

J. 2007, 13, 2686.

Casida, M. E. J. Chem. Phys. 1998, 108, 4439. (e) Stratmann, R. E.;

Scuseria, G. E.; Frisch, J. M. J. Chem. Phys. 1998, 109, 8218.

(24) Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford, CT,

2004.

(5) (a) Vogler, A.; Kunkely, H. Top. Curr. Chem. 2001, 213, 143. (b)

Che, C.-M.; Lai, S.-W. Coord. Chem. Rev. 2005, 249, 1296. (c) Barbieri,

A.; Accorsi, G.; Armaroli, N. Chem. Commun. 2008, 2185. (d) Deaton,

J. C.; Switalski, S. C.; Kondakov, D. Y.; Young, R. H.; Pawlik, T. D.;

Giesen, D. J.; Harkins, S. B.; Miller, A. J. M.; Mickenberg, S. F.; Peters,

J. C. J. Am. Chem. Soc. 2010, 132, 9499.

(25) Tsuboyama, A.; Kuge, K.; Furugori, M.; Okada, S.; Hoshino,

M.; Ueno, K. Inorg. Chem. 2007, 46, 1992.

(26) (a) Cuttell, D. G.; Kuang, S. M.; Fanwick, P. E.; McMillin, D. R.;

Walton, R. A. J. Am. Chem. Soc. 2002, 124, 6. (b) Kuang, S. M.; Cuttell,

D. G.; McMillin, D. R.; Fanwick, P. E.; Walton, R. A. Inorg. Chem. 2002,

41, 3313.

(6) (a) Ford, P. C.; Cariati, E.; Bourassa, J. Chem. Rev. 1999,

99, 3625. (b) Yam, V. W.-W.; Lo, K. K.-W.; Fung, W. K.-M.; Wang,

C.-R. Coord. Chem. Rev. 1998, 171, 17.

(7) (a) Williams, R. M.; De Cola, L.; Hartl, F.; Lagref, J.-J.; Planeix,

J.-M.; De Cian, A.; Hosseini, M. W. Coord. Chem. Rev. 2002, 230, 253.

(b) Armaroli, N.; Accorsi, G.; Cardinali, F.; Listorti, A. Top. Curr. Chem.

2007, 280, 69. (c) Lavie-Cambot, A.; Cantuel, M.; Leydet, Y.; Jonusauskas,

G.; Bassani, D. M.; McClenaghan, N. D. Coord. Chem. Rev. 2008, 252, 2572.

(d) Accorsi, G.; Listorti, A.; Yoosaf, K.; Armaroli, N. Chem. Soc. Rev. 2009,

38, 1690.

(8) (a) Song, D.; Jia, W. L.; Wu, G.; Wang, S. Dalton Trans.

2005, 433. (b) Chou, C.-H.; Yeh, W.-Y.; Lee, G.-H.; Peng, S.-M. Inorg.

Chim. Acta 2006, 359, 4139. (c) Saito, K.; Arai, T.; Takahashi, N.;

Tsukuda, T.; Tsubomura, T. Dalton Trans. 2006, 4444. (d) McCormick,

T.; Jia, W.-L.; Wang, S. Inorg. Chem. 2006, 45, 147. (e) Wong, E.; Li, J.;

Seward, C.; Wang, S. Dalton Trans. 2009, 1776. (f) Wei, F.; Fang, L.;

Huang, Y. Inorg. Chim. Acta 2010, 363, 2600.

(27) Andino, J. G.; Flores, J. A.; Karty, J. A.; Massa, J. P.; Park, H.;

Tsvetkov, N. P.; Wolfe, R. J.; Caulton, K. G. Inorg. Chem. 2010, 49, 7626.

(28) (a) Yam, V. W.-W.; Chan, C.-L.; Li, C.-K.; Wong, K. M.-C.

Coord. Chem. Rev. 2001, 216ꢀ217, 173. (b) Schwerdtfeger, P.; Hermann,

H. L.; Schmidbaur, H. Inorg. Chem. 2003, 42, 1334. (c) De Fremont, P.;

Scott, N. M.; Stevens, E. D.; Nolan, S. P. Organometallics 2005, 24, 2411. (d)

Ovejero, P.; Mayoral, M. J.; Cano, M.; Lagunas, M. C. J. Organomet. Chem.

2007, 692, 1690.

(29) (a) Chen, L. X.; Shaw, G. B.; Novozhilova, I.; Liu, T.; Jennings,

G.; Attenkofer, K.; Meyer, G. J.; Coppens, P. J. Am. Chem. Soc. 2003,

125, 7022. (b) Chen, L. X.; Jenings, G.; Liu, T.; Gosztola, D. J.; Hessler,

J. P.; Scaltrito, D. V.; Meyer, G. J. J. Am. Chem. Soc. 2002, 124, 10861.

(30) (a) Sacksteder, L.; Zipp, A. P.; Brown, E. A.; Streich, J.; Demas,

J. N.; DeGraﬀ, B. A. Inorg. Chem. 1990, 29, 4335. (b) Colombo, M. G.;

Hauser, A.; Gudel, H. U. Top. Curr. Chem. 1994, 171, 143. (c) Weinstein,

J. A.; van Slageren, J.; Stufkens, D. J.; Zalis, S.; George, M. W. J. Chem.

Soc., Dalton Trans. 2001, 2587.

(9) (a) Ma, Y. G.; Che, C.-M.; Chao, H.-Y.; Zhou, X. M.; Chan,

W.-H.; Shen, J. C. Adv. Mater. 1999, 11, 852. (b) Zhang, Q.; Zhou, Q.;

Cheng, Y.; Wang, L.; Ma, D.; Jing, X.; Wang, F. Adv. Mater. 2004,

16, 432. (c) Zhang, L.; Li, B.; Su, Z. J. Phys. Chem. C 2009, 113, 13968.

(d) Min, J.; Zhang, Q.; Sun, W.; Cheng, Y.; Wang, L. Dalton Trans. 2011,

40, 686.

(10) (a) Armaroli, N.; Accorsi, G.; Holler, M.; Moudam, O.;

Nierengarten, J.-F.; Zhou, Z.; Wegh, R. T.; Welter, R. Adv. Mater. 2006,

18, 1313. (b) Moudam, O.; Kaeser, A.; Delavaux-Nicot, B.; Duhayon, C.;

Holler, M.; Accorsi, G.; Armaroli, N.; Seguy, I.; Navarro, J.; Destruel, P.;

Nierengarten, J.-F. Chem. Commun. 2007, 3077.

(11) (a) Lotito, K. J.; Peters, J. C. Chem. Commun. 2010, 46, 3690.

(b) Krylova, V. A.; Djurovich, P. I.; Whited, M. T.; Thompson, M. E.

Chem. Commun. 2010, 46, 6696.

(12) Gorin, D. J.; Toste, F. D. Nature 2007, 446, 395.

(13) (a) Klappa, J. J.; Rich, A. E.; McNeill, K. Org. Lett. 2002, 4, 435.

(b) Ohkura, H.; Berbasov, D. O.; Soloshonok, V. A. Tetrahedron Lett.

2003, 44, 2417. (c) Chen, J.-L.; Lin, C.-H.; Chen, J.-H.; Chi, Y.; Chiu,

Y.-C.; Chou, P.-T.; Lai, C.-H.; Lee, G.-H.; Carty, A. J. Inorg. Chem. 2008,

47, 5154.

(31) (a) Turro, N. J. Modern Molecular Photochemistry; University

Science Books: Mill Valley, CA, 1991. (b) Griﬃths, D. J. Introduction to

Quantum Mechanics, 2nd ed.; Prentice Hall: Upper Saddle River, NJ,

2005.

(32) Cheng, Y.-M.; Li, E. Y.; Lee, G.-H.; Chou, P.-T.; Lin, S.-Y.; Shu,

C.-F.; Hwang, K.-C.; Chen, Y.-L.; Song, Y.-H.; Chi, Y. Inorg. Chem. 2007,

46, 10276.

(33) Chou, P.-T.; Chi, Y.; Chung, M.-W.; Lin, C.-C. Coord. Chem.

Rev. 2011, DOI: 10.1016/J.CCR.2010.12.013.

(34) (a) Hedley, G. J.; Ruseckas, A.; Samuel, I. D. W. J. Phys. Chem. A

2009, 113, 2. (b) DeArmond, M. K. Acc. Chem. Res. 1974, 7, 309.

(35) (a) Iwamura, M.; Takeuchi, S.; Tahara., T. J. Am. Chem. Soc.

2007, 129, 5248. (b) Leydet, Y.; Bassani, D. M.; Jonusauskas, G.;

McClenaghan, N. D. J. Am. Chem. Soc. 2007, 129, 8688. (c) Iwamura,

M.; Watanabe, H.; Ishii, K.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc.

2011, 133, 7728.

(36) Nakanishi, W.; Yamanaka, M.; Nakamura, E. J. Am. Chem. Soc.

2005, 127, 1446.

(37) McGlynn, S. P.; Azumi, T.; Kinoshita, M. Molecular Spectros-

copy of the Triplet State; Prentice Hall Inc.: Englewood Cliﬀs, NJ, 1969;

Chapter 5, pp 189.

(14) Gubas, G. J. Inorg. Synth. 1990, 28, 68.

(15) Lin, C.-C.; Chen, C.-L.; Chung, M.-W.; Chen, Y.-J.; Chou, P. T.

J. Phys. Chem. A 2010, 114, 10412.

(16) De Mello, J. C.; Wittmann, H. F.; Friend, R. H. Adv. Mater.

1997, 9, 230.

(17) Hsieh, C.-C.; Cheng, Y.-M.; Hsu, C.-J.; Chen, K.-Y.; Chou,

P.-T. J. Phys. Chem. A 2008, 112, 8323.

(18) Hsieh, C.-C.; Chou, P.-T.; Shih, C.-W.; Chuang, W.-T.; Chung,

M.-W.; Lee, J.; Joo, T. J. Am. Chem. Soc. 2011, 133, 2932.

(19) Duan, H. S.; Chou, P. T.; Hsu, C. C.; Hung, J. Y.; Chi, Y. Inorg.

Chem. 2009, 48, 6501.

(38) Haneder, S.; Da Como, E.; Feldmann, J.; Lupton, J. M.;

Lennartz, C.; Erk, P.; Fuchs, E.; Molt, O.; M€unster, I.; Schildknecht,

C.; Wagenblast, G. Adv. Mater. 2008, 20, 3325.

(39) Tsuboyama, A.; Iwawaki, H.; Furugori, M.; Mukaide, T.;

Kamatani, J.; Igawa, S.; Moriyama, T.; Miura, S.; Takiguchi, T.; Okada,

S.; Hoshino, M.; Ueno, K. J. Am. Chem. Soc. 2003, 125, 12971.

(40) Holmes, R. J.; Forrest, S. R.; Tung, Y.-J.; Kwong, R. C.; Brown,

J. J.; Garon, S.; Thompson, M. E. Appl. Phys. Lett. 2003, 82, 2422.

(41) (a) Su, S.-J.; Chiba, T.; Takeda, T.; Kido, J. Adv. Mater. 2008,

20, 2125. (b) Tanaka, D.; Agata, Y.; Takeda, T.; Watanabe, S.; Kido, J.

Jpn. J. Appl. Phys. 2007, 46, L117.

(20) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (b)

Becke, A. D. J. Chem. Phys. 1993, 98, 5648.

(42) (a) Kalinowski, J.; Stampor, W.; Cocchi, M.; Virgili, D.; Fattori,

V. J. Appl. Phys. 2006, 100, 034318. (b) Reineke, S.; Walzer, K.; Leo, K.

Phys. Rev. B 2007, 75, 125328.

(43) (a) Kolosov, D.; Adamovich, V.; Djurovich, P.; Thompson,

M. E.; Adachi, C. J. Am. Chem. Soc. 2002, 124, 9945. (b) Gebler, D. D.;

Wang, Y. Z.; Blatchford, J. W.; Jessen, S. W.; Fu, D.-K.; Swager, T. M.;

MacDiarmid, A. G.; Epstein, A. J. Appl. Phys. Lett. 1997, 70, 1644.

(21) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. (b)

Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (c) Hay, P. J.; Wadt,

W. R. J. Chem. Phys. 1985, 82, 299.

(22) Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209.

(23) (a) Jamorski, C.; Casida, M. E.; Salahub, D. R. J. Chem. Phys.

1996, 104, 5134. (b) Petersilka, M.; Grossmann, U. J.; Gross, E. K. U.

Phys. Rev. Lett. 1996, 76, 1212. (c) Bauernschmitt, R.; Ahlrichs, R.;

Hennrich, F. H.; Kappes, M. M. J. Am. Chem. Soc. 1998, 120, 5052. (d)

12099

dx.doi.org/10.1021/ja2026568 |J. Am. Chem. Soc. 2011, 133, 12085–12099

Article Doi

Systematic investigation of the metal-structure-photophysics relationship of emissive d10-complexes of group 11 elements: The prospect of application in organic light emitting devices

DOI: 10.1021/ja2026568

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja2026568

Full text of DOI:10.1021/ja2026568

Products guided by the article

R&D Labs maybe for 1318967-90-6

Relevant to this article

Hot Product