Open Access Article
Andreas M.
Bünzli
a,
Antonio
Pertegás
b,
Cristina
Momblona
b,
José M.
Junquera-Hernández
b,
Edwin C.
Constable
a,
Henk J.
Bolink
b,
Enrique
Ortí
*b and
Catherine E.
Housecroft
*a
aDepartment of Chemistry, University of Basel, Spitalstrasse 51, CH-4056 Basel, Switzerland. E-mail: catherine.housecroft@unibas.ch
bInstituto de Ciencia Molecular, Universidad de Valencia, Catedrático José Beltrán 2, Paterna, E-46980, Spain. E-mail: enrique.orti@uv.es
First published on 23rd September 2016
The synthesis of four cyclometallated [Ir(C^N)2(N^N)][PF6] compounds in which N^N is a substituted 2,2′-bipyridine (bpy) ligand and the naphthyl-centred ligand 2,7-bis(2-(2-(4-(pyridin-2-yl)phenoxy)ethoxy)ethoxy)naphthalene provides the two cyclometallating C^N units is reported. The iridium(III) complexes have been characterized by 1H and 13C NMR spectroscopies, mass spectrometry and elemental analysis, and their electrochemical and photophysical properties are described. Comparisons are made with a model [Ir(ppy)2(N^N)][PF6] compound (Hppy = 2-phenylpyridine). The complexes containing the naphthyl-unit exhibit similar absorption spectra and excitation at 280 nm leads to an orange emission. The incorporation of the naphthalene unit does not lead to a desirable blue contribution to the emission. Density functional theory calculations were performed to investigate the geometries of the complexes in their ground and first triplet excited states, as well as the energies and compositions of the highest-occupied and lowest unoccupied molecular orbital (HOMO and LUMO) manifolds. Trends in the HOMO–LUMO gaps agree with those observed electrochemically. The energy difference between the LUMO and the lowest unoccupied MO located on the naphthyl unit (LUMO+7) is large enough to explain why there is no contribution from the naphthyl-centred triplet excited state to the phosphorescence emission. Singlet excited states were also investigated. Light-emitting electrochemical cells (LECs) using the [Ir(C^N)2(N^N)][PF6] and [Ir(ppy)2(N^N)][PF6] complexes in the emissive layer were made and evaluated. The presence of the naphthyl-bridge between the cyclometallating units does not significantly alter the device response.
One approach to white-emitting LECs is to combine blue-green with red-emitting complexes in the active layer of the device or to mix three iTMCs of different colours (blue, red and orange).10–13 Other strategies (not uniquely based on iTMCs) include multifluorophoric conjugated polymers,14,15 a combined polymer-composite blue-light emitting layer with an orange ionic iridium complex,16,17 or employ a colour conversion layer.18,19 However, all these LEC devices only operate at low luminance values. The reason for the small number and poor performances of white-emitting LECs reported is largely ascribed to the lack of highly efficient and stable deep-blue-emitting complexes required for colour-mixing.
We now explore a possible approach to dual-emitting iTMCs for white-light emission which follows the principle of combining complementary colours by combining a blue-emitting naphthalene domain and orange-emitting [Ir(C^N)2(N^N)]+ complexes. The naphthalene component is accommodated within a linker between the two cyclometallating units in ligand H21 (Scheme 1).
![]() | ||
| Scheme 1 Synthesis of cyclometallating ligand H21 with atom labelling for NMR assignments. Conditions: (i) NaH, DMF, 120 °C, 24 h. | ||
2,7-Bis[2-(2-hydroxyethoxy)ethoxy]naphthalene,20 ligands 2–521,22 and [Ir(ppy)2(4)][PF6]23 were prepared by standard procedures. Silica gel 60 was bought from Fluka.
000), 266 (33
000), 285 (38
000), 325 (6000), 350 sh (2000). Emission (CH2Cl2, 9.50 × 10−6 mol L−1, λexc = 237 nm) λem = 331, 344, 410 nm. ESI-MS m/z 643.6 [M + H]+ (base peak, calc. 643.3). Found C 74.44, H 5.91, N 4.58; C40H38N2O6 requires C 74.75, H 5.96, N 4.36%.
:
MeOH 100
:
1) and [Ir2(1)2(μ-Cl)2] was isolated as a yellow solid (90.0 mg, 0.052 mmol, 33.4%). 1H NMR (500 MHz, CD2Cl2) δ/ppm 9.11 (dd, J = 5.8, 0.9 Hz, 2H, HB6), 7.71–7.63 (overlapping m, 4H, HB4+N4), 7.56 (d, J = 7.9 Hz, 2H, HB3), 7.20 (d, J = 8.7 Hz, 2H, HA3), 6.97 (dd, J = 8.8, 2.5 Hz, 2H, HN3), 6.90 (d, J = 2.4 Hz, 2H, HN1), 6.71 (ddd, J = 7.3, 5.8, 1.4 Hz, 2H, HB5), 6.33 (dd, J = 8.5, 2.5 Hz, 2H, HA4), 5.32 (d, J = 2.6 Hz, 2H, HA6), 4.04–3.94 (overlapping m, 4H, Ha+a′), 3.87 (m, 2H, Hd/d′), 3.81–3.73 (overlapping m, 4H, Hb/b′+d/d′), 3.65 (m, 2H, Hb/b′), 3.61–3.57 (overlapping m, 4H, Hc+c′). 13C NMR (126 MHz, CD2Cl2) δ/ppm 168.2 (CB2), 159.5 (CA5), 157.7 (CN2), 151.7 (CB6), 147.3 (CA1), 137.6 (CA2), 137.0 (CB4), 136.4 (CN8a), 129.3 (CN4), 125.5 (CA3), 124.9 (CN4a), 121.8 (CB5), 118.4 (CB3), 117.2 (CN3), 116.6 (CA6), 108.6 (CA4), 107.5 (CN1), 70.3 (Cc), 70.1 (Cb), 67.6 (Ca+d). ESI-MS m/z 833.5 [Ir(1)]+ (base peak, calc. 833.2). The complex was used in the next step without further purification.
:
MeOH 100
:
0.25, then to CH2Cl2
:
MeOH 100
:
1). [Ir(1)(2)][PF6] was isolated as an orange solid (96.0 mg, 0.075 mmol, 55.3%). 1H NMR (500 MHz, CD2Cl2) δ/ppm 8.37 (dd, J = 8.1, 1.2 Hz, 2H, HE3), 8.10 (t, J = 7.9 Hz, 2H, HE4), 8.06 (dd, J = 5.9, 0.8 Hz, 2H, HB6), 7.70 (ddd, J = 8.4, 7.5, 1.5 Hz, 2H, HB4), 7.66 (d, J = 8.8 Hz, 2H, HN4), 7.30 (overlapping m, 4H, HB3+E5), 7.01 (ddd, J = 7.3, 5.9, 1.4 Hz, 2H, HB5), 6.98–6.90 (overlapping m, 6H, HN3+N1+G4), 6.75–6.69 (overlapping m, 6H, HB4+G3), 6.55 (br, 4H, HG2), 5.99 (dd, J = 8.6, 2.5 Hz, 2H, HA4), 4.62 (d, J = 2.5 Hz, 2H, HA6), 4.06–3.97 (overlapping m, 4H, Ha+a′), 3.77 (m, 2H, Hb/b′), 3.71 (m, 2H, Hd/d′), 3.66 (m, 2H, Hb/b′), 3.55 (m, 2H, Hd/d′), 3.53–3.45 (overlapping m, 4H, Hc+c′). 13C NMR (126 MHz, CD2Cl2) δ/ppm 168.1 (CB2), 165.7 (CE6), 159.4 (CE2), 159.2 (CA5), 157.9 (CN2), 150.3 (CB6), 149.3 (CA1), 139.8 (CE4), 138.3 (CG1), 138.3 (CB4), 136.4(CN8a), 135.4 (CA2), 130.1 (CE5), 129.3 (CN4), 128.8 (CG4), 127.9 (CG2), 127.9 (CG3), 125.9 (CA3), 125.0 (CN4a), 124.7 (CE3), 121.2 (CB5), 119.0 (CB3), 117.3 (CN3), 116.2 (CA6), 109.4 (CA4), 107.8 (CN1), 70.6 (Cb+b′), 69.8 (Cc+c′), 67.9 (Ca+a′), 67.3 (Cd+d′). UV-Vis λ/nm (ε/L mol−1 cm−1) (CH2Cl2, 1.00 × 10−5 mol L−1) 237 (102
000), 280 (44
000), 312 (36
000), 325 (33
000), 350 sh (17
000), 400 sh (6000). ESI-MS m/z 1141.3 [M − PF6]+ (base peak, calc. 1141.4). Found C 57.75, H 4.61, N 4.56; C62H52F6IrN4O6P requires C 57.89, H 4.07, N 4.36%.
:
MeOH 100
:
0.25 then 100
:
1. [Ir(1)(3)][PF6] was isolated as an orange solid (60.0 mg, 0.050 mmol, 45.5%). 1H NMR (500 MHz, CD2Cl2) δ/ppm 8.53–8.47 (overlapping m, 2H, HF3+E3), 8.19 (t, J = 7.9 Hz, 1H, HF4), 8.10 (td, J = 8.0, 1.6 Hz, 1H, HE4), 7.90 (dd, J = 5.5, 1.0 Hz, 1H, HE6), 7.74 (ddd, J = 8.4, 7.5, 1.5 Hz, 1H, HB4), 7.70–7.65 (overlapping m, 2H, HN4+N5), 7.64 (ddd, J = 8.2, 7.5, 1.5 Hz, 1H, HD4), 7.54 (d, J = 8.1 Hz, 1H, HB3), 7.50–7.43 (overlapping m, 3H, HB6+D6+F5), 7.40 (d, J = 8.2 Hz, 1H, HD3), 7.37 (ddd, J = 7.6, 5.6, 1.1 Hz, 1H, HE5), 7.10 (d, J = 8.7 Hz, 1H, HC3), 7.00–6.89 (overlapping m, 8H, HN3+N6+D5+G4+A3+N1+N8+B5), 6.76 (broadened t, J ∼ 7.0 Hz, 2H, HG3), 6.57 (br, 2H, HG2), 6.41 (dd, J = 8.6, 2.5 Hz, 1H, HC4), 6.11 (dd, J = 8.6, 2.5 Hz, 1H, HA4), 5.38 (d, J = 2.5 Hz, 1H, HC6), 4.91 (d, J = 2.5 Hz, 1H, HA6), 4.10–3.93 (overlapping m, 5H, Ha+a′+e+e′+h/h′), 3.88–3.77 (overlapping m, 4H, Hh/h′+d/d′+b/b′+f/f′), 3.74–3.61 overlapping (m, 7H, Hd/d′+b/b′+f/f′+c+c′+g+g′). 13C NMR (126 MHz, CD2Cl2) δ/ppm 169.0 (CB2), 167.4 (CD2), 166.3 (CF6), 160.6 (CC5), 159.8 (CA5), 157.9 (CN2/N7), 157.9 (CN2/N7), 157.4 (CE1), 157.2 (CF2), 153.8 (CA1), 151.0 (CE6), 149.3 (CD6), 149.1 (CC1), 148.9 (CB6), 140.1 (CF4), 139.8 (CE4), 138.5 (CD4), 138.4 (CG1), 138.4 (CB4), 136.5 (CN8a), 136.4 (CA2), 136.3 (CC2), 130.6 (CF5), 129.4 (CG4), 129.4 (CN5), 129.3 (CN4), 128.2 (CE5), 128.2 (CG3), 128.0 (CG2), 126.6 (CC3), 126.3 (CA3), 125.5 (CE3), 125.0 (CN4a), 124.1 (CF3), 122.5 (CD5), 121.6 (CB5), 119.5 (CB3), 119.4 (CD3), 117.6 (CA6), 117.3 (CN3+N6), 116.5 (CC6), 109.9 (CC4), 108.7 (CA4), 107.9 (CN8), 107.7 (CN1), 70.6 (CCH2), 70.4 (CCH2), 70.3 (CCH2), 70.0 (CCH2), 68.0 (CCH2), 67.9 (CCH2), 67.7 (CCH2), 67.5 (CCH2). UV-Vis λ/nm (ε/L mol−1 cm−1) (CH2Cl2, 1.00 × 10−5 mol L−1) 237 (99
000), 280 (49
000), 312 (35
000), 326 (28
000), 350 sh (16
000), 400 sh (7000). ESI-MS m/z 1065.3 [M − PF6]+ (base peak, calc. 1065.3). Found C 55.58, H 4.60, N 4.89; C56H48F6IrN4O6P requires C 55.58, H 4.00, N 4.63%.
:
MeOH 100
:
0.5 and finally CH2Cl2
:
MeOH 100
:
1. [Ir(1)(2)][PF6] was isolated as an orange solid (101 mg, 0.072 mmol, 75.0%). 1H NMR (500 MHz, CD2Cl2) δ/ppm 8.15–8.07 (overlapping m, 4H, HB6+E3), 7.69 (ddd, J = 8.4, 7.5, 1.5 Hz, 2H, HB4), 7.66 (d, J = 8.6 Hz, 2H, HN4), 7.32 (d, J = 8.1 Hz, 2H, HB3), 7.25 (d, J = 2.0 Hz, 2H, HE5), 7.00 (ddd, J = 7.3, 5.9, 1.3 Hz, 2H, HB5), 6.98–6.91 (overlapping m, 6H, HG4+N1+N3), 6.77–6.69 (overlapping m, 6H, HA3+G3), 6.58 (br, 4H, HG2), 6.00 (dd, J = 8.6, 2.5 Hz, 2H, HA4), 4.65 (d, J = 2.5 Hz, 2H, HA6), 4.08–3.96 (m, 4H, Ha), 3.80–3.74 (m, 2H, Hb/b′), 3.74–3.69 (m, 2H, Hd/d′), 3.69–3.63 (m, 2H, Hb/b′), 3.59–3.53 (m, 2H, Hd/d′), 3.53–3.45 (m, 4H, Hc), 1.40 (s, 18H, HtBu). 13C NMR (126 MHz, CD2Cl2) δ/ppm 168.1 (CB2), 165.3 (CE6), 164.7 (CE4), 160.0 (CE2), 159.2 (CA5), 157.9 (CN2), 150.2 (CB6), 149.5 (CA1), 138.5 (CG1), 138.2 (CB4), 136.4 (CN8a), 135.4 (CA2), 129.3 (CN4), 128.8 (CG4), 128.0 (CG2), 127.8 (CG3), 126.8 (CE5), 126.0 (CA3), 124.9 (CN4a), 121.6 (CE3), 121.0 (CB5), 118.8 (CB3), 117.3 (CN3), 116.2 (CA6), 109.2 (CA4), 107.8 (CN1), 70.6 (Cb), 69.8 (Cc), 67.9 (Ca), 67.3 (Cd), 36.0 (Cquat-tBu-E), 30.5 (CtBu-E). UV-Vis λ/nm (ε/L mol−1 cm−1) (CH2Cl2, 1.00 × 10−5 mol L−1) 236 (110
000), 280 (45
000), 311 (38
000), 325 (35
000), 350 sh (17
000), 395 sh (9000). ESI-MS m/z 1254.0 [M − PF6]+ (base peak, calc. 1253.5). Found C 60.49, H 5.37, N 4.16; C70H68F6IrN4O6P requires C 60.12, H 4.90, N 4.01%.
:
MeOH 100
:
0.25 then 100
:
1). [Ir(1)(5)][PF6] was isolated as an orange solid (33.0 mg, 0.025 mmol, 44.1%). 1H NMR (500 MHz, CD2Cl2) δ/ppm 8.30 (d, J = 1.9 Hz, 1H, HF3), 8.28 (d, J = 1.6 Hz, 1H, HE3), 7.79 (d, J = 5.9 Hz, 1H, H6), 7.75 (ddd, J = 8.4, 7.3, 1.3 Hz, 1H, HB4), 7.71–7.65 (overlapping m, 2H, HN4+N5), 7.64 (ddd, J = 8.6, 7.5, 1.3 Hz, 1H, HD4), 7.54 (d, J = 8.1 Hz, 1H, HB3), 7.47 (d, J = 5.3 Hz, 1H, HB6), 7.45 (d, J = 5.6 Hz, 1H, HD6), 7.41 (d, J = 1.9 Hz, 1H, HF5), 7.39 (d, J = 8.2 Hz, 1H, HD3), 7.36 (dd, J = 5.9, 1.8 Hz, 1H, HE5), 7.09 (d, J = 8.6 Hz, 1H, HC3), 7.01–6.88 (overlapping m, 8H, HN3+N6+G4+A3+D5+N1+N8+B5), 6.76 (broadened t, J ∼ 7.3 Hz, 2H, HG3), 6.58 (br, 2H, HG2), 6.41 (dd, J = 8.6, 2.5 Hz, 1H, HC4), 6.12 (dd, J = 8.6, 2.5 Hz, 1H, HA4), 5.38 (d, J = 2.5 Hz, 1H, HC6), 4.89 (d, J = 2.5 Hz, 1H, HA6), 4.09–3.94 (overlapping m, 5H, Ha+a′+e+e′+h/h′), 3.88–3.77 (overlapping m, 4H, Hh/h′+d/d′+b/b′+f/f′), 3.74–3.59 (overlapping m, 7H, Hd/d′+b/b′+f/f′+c+c′+g+g′), 1.46 (s, 9H, HtBu-ring F), 1.42 (s, 9H, HtBu-ring E). 13C NMR (126 MHz, CD2Cl2) δ/ppm 1v69.0 (CB2), 167.5 (CD2), 166.0 (CF6), 164.7 (CF4), 164.3 (CE4), 160.6 (CC5), 159.8 (CA5), 157.9 (CN7), 157.9 (CN2), 157.3 (CF2), 157.2 (CE2), 154.2 (CA1), 150.5 (CE6), 149.5 (CC1), 149.3 (CD6), 148.8 (CB6), 138.7 (CG1), 138.3 (CD4), 138.3 (CB4), 136.5 (CN8a), 136.5 (CA2), 136.4 (CC2), 129.3 (CN4/N5), 129.3 (CN4/N5), 129.3 (CG4), 128.1 (CG2), 128.1 (CG3), 127.6 (CF5), 126.6 (CC3), 126.3 (CA3), 125.7 (CE5), 125.0 (CN4a), 122.3 (CD5), 121.8 (CE3), 121.4 (CB5), 120.7 (CF3), 119.4 (CB3), 119.4 (CD3), 117.5 (CA6), 117.3 (CN3+N6), 116.5 (CC6), 109.8 (CC4), 108.6 (CA4), 107.9 (CN8), 107.7 (CN1), 70.6 (CCH2), 70.4 (CCH2), 70.3 (CCH2), 70.0 (CCH2), 67.9 (CCH2), 67.9 (CCH2), 67.7 (CCH2), 67.5 (CCH2), 36.1 (Cquat-tBu-ring E), 36.1 (Cquat-tBu-ring F), 30.6 (CtBu-ring F), 30.6 (CtBu-ring E). UV-Vis λ/nm (ε/L mol−1 cm−1) (CH2Cl2, 1.00 × 10−5 mol L−1) 236 (107
000), 280 (52
000), 311 (37
000), 326 (28
000), 350 sh (16
000), 400 sh (7000). ESI-MS m/z 1177.4 [M − PF6]+ (base peak, calc. 1177.5). Found C 58.52, H 5.25, N 4.58; C64H64F6IrN4O6P requires C 58.13, H 4.88, N 4.24%.
:
1 molar ratio (iTMC
:
IL). The devices were then transferred to an inert atmosphere glovebox (<0.1 ppm O2 and H2O, MBraun), where a layer (70 nm) of aluminium (the top electrode) was thermally evaporated onto the devices using an Edwards Auto500 evaporator integrated in the inert atmosphere glovebox. The area of the device was 6.5 mm2. The devices were not encapsulated and were characterized inside the glovebox at room temperature.
The conventional methodology39 for preparing complexes of the type [Ir(C^N)2(N^N)][PF6] is treatment of a [Ir(C^N)2(μ-Cl)2] dimer with two equivalents of an N^N ligand followed by anion exchange. However, the unfavourable effect that residual chloride ion has on the performance of LEC devices40 has led us to develop an alternative strategy in which the chlorido dimer is converted to an intermediate solvento complex41 by treatment of the dimer with AgPF6 in methanol (Scheme 2).42 The dimer [Ir2(1)2(μ-Cl)2] was prepared from H21 and IrCl3·xH2O under dilute conditions to ensure that the two C^N coordination sites of H21 underwent cyclometallation at the same iridium(III) centre. The dimer was then treated with AgPF6 in MeOH to give the solvento complex [Ir(1)(MeOH)2][PF6] which was used for the final reaction step without further purification. Both [Ir2(1)2(μ-Cl)2] and [Ir(1)(MeOH)2][PF6] were characterized by ESI MS and NMR spectroscopy. The base peak envelope in the mass spectrum of each complex corresponded to the [Ir(1)]+ cation. Evidence for the coordinated MeOH in [Ir(1)(MeOH)2][PF6] came from δ(1H) and δ(13C) NMR signals at 3.34 and 49.9 ppm, respectively. These values compare with δ(1H) = 3.35 ppm and δ(13C) = 49.9 ppm in [Ir(Phppy)2(MeOH)2][PF6] and [Ir(Ph2ppy)2(MeOH)2][PF6] (HPhppy = 2-(3-phenyl)phenylpyridine and HPh2ppy = 2-(3,5-diphenyl)phenylpyridine).42 Going from H21 to the coordinated ligand in [Ir2(1)2(μ-Cl)2] and [Ir(1)(MeOH)2][PF6] results in a loss in rotational freedom within the polyethyleneoxy-chain, with the consequence that each pair of CH2 protons appears as a diastereotopic pair (see Experimental section and discussion of the [Ir(1)(N^N)]+ complexes).
![]() | ||
| Fig. 1 500 MHz 1H NMR spectrum of a CDCl3 solution of H21; scale in δ/ppm. * = residual CHCl3. See Scheme 1 for ring labels. | ||
![]() | ||
| Fig. 2 The aromatic and alkyl regions of the 500 MHz 1H NMR spectrum of [Ir(1)(2)][PF6] in CD2Cl2; scale δ/ppm. See Scheme 2 for atom labels. | ||
On going from [Ir(1)(2)]+ and [Ir(1)(4)]+ to [Ir(1)(3)]+ and [Ir(1)(5)]+, the complex is desymmetrized. This is apparent from the 1H NMR spectrum for [Ir(1)(3)][PF6] (Fig. 3) compared to that of [Ir(1)(2)][PF6] (Fig. 2). Signals for both HG2 (δ = 6.57 ppm) and HG3 (δ = 6.76 ppm) are broadened in the 1H NMR spectrum of [Ir(1)(3)][PF6] (Fig. 3), and similarly broadened signals are observed for HG2 (δ = 6.58 ppm) and HG3 (δ = 6.76 ppm) in [Ir(1)(5)][PF6]. As in [Ir(1)(2)][PF6] and [Ir(1)(4)][PF6], this indicates hindered rotation of the 6-phenyl substituent.
![]() | ||
| Fig. 3 The aromatic and alkyl regions of the 500 MHz 1H NMR spectrum of [Ir(1)(3)][PF6] in CD2Cl2; scale δ/ppm. See Scheme 2 for atom labels. * = overlapping signals for residual CDHCl2 and CH2Cl2. | ||
![]() | ||
| Fig. 4 Cyclic voltammogram for [Ir(1)(5)][PF6] in CH2Cl2 (see Table 1 for conditions) referenced to Fc/Fc+; → = direction of scan. | ||
| Compound | E ox1/2/V | E red1/2/V | ΔE1/2/V |
|---|---|---|---|
| [Ir(1)(2)][PF6] | +0.82irr | −1.81qr | 2.63 |
| [Ir(1)(3)][PF6] | +0.80irr | −1.84qr | 2.64 |
| [Ir(1)(4)][PF6] | +0.76irr | −1.96qr | 2.72 |
| [Ir(1)(5)][PF6] | +0.77irr | −1.92qr | 2.69 |
| [Ir(ppy)2(4)][PF6] | +0.87irr | −1.94irr | 2.81 |
The reduction processes centred on the N^N ligand are similar for [Ir(1)(4)][PF6] and [Ir(1)(5)][PF6] (−1.96 V and −1.92 V, respectively) compared to −1.94 V for [Ir(ppy)2(4)][PF6]. On going from [Ir(1)(2)][PF6] to [Ir(1)(4)][PF6] and from [Ir(1)(3)][PF6] to [Ir(1)(5)][PF6], electron-releasing tert-butyl groups are introduced into the N^N domain and there is a significant shift of Ered1/2 to more negative potentials. This is consistent with the LUMO of the complex being localized on the N^N ligand as discussed below and in agreement with, for example, [Ir(ppy)2(2)][PF6].43
![]() | ||
| Fig. 5 Solution absorption spectra of [Ir(1)(N^N)][PF6] complexes and reference compound [Ir(ppy)2(4)][PF6] in CH2Cl2 (1.00 × 10−5 mol L−1). | ||
Upon excitation at 280 nm, the complexes all exhibit orange emission. The solution emission spectra are shown in Fig. 6 and emission maxima, quantum yields and lifetimes are given in Tables 2 and 3. Introducing the electron-releasing tert-butyl substituents results in the expected blue-shift in λmaxem (23 nm on going from [Ir(1)(3)][PF6] to [Ir(1)(5)][PF6], and 26 nm from [Ir(1)(2)][PF6] to [Ir(1)(4)][PF6]). Introducing the naphthyl-containing linker between the two ppy moieties has little effect on λmaxem (558 nm in model compound [Ir(ppy)2(4)][PF6] and 564 nm in [Ir(1)(4)][PF6]) consistent with the emission originating from a mixture of MLCT and LLCT transitions6 involving the Ir(III) metal centre and the ppy units of the C^N ligands in the HOMO and the N^N ligand in the LUMO as discussed below. This is consistent with the electrochemical data (Table 1). In degassed solutions, the [Ir(1)(N^N)][PF6] orange emissions exhibit photoluminescence quantum yields (PLQYs) in the range 6–18% (Table 2) with lifetimes of between 135 and 370 ns (Table 3). Fig. 7 depicts the emission spectra of solid-state samples of the complexes. For [Ir(1)(2)][PF6] and [Ir(1)(3)][PF6], going from solution to solid leads to a blue-shift in λmaxem, whereas for the complexes containing the tert-butyl substituents, λmaxem is little affected. Both the PLQY and emission lifetimes are enhanced from solution to solid state (Tables 2 and 3). We note that the solution PLQY recorded for the reference complex [Ir(ppy)2(4)][PF6] of 1% (Table 2) is consistent with the 0.07% previously reported.23
![]() | ||
| Fig. 6 Normalized solution emission spectra of [Ir(1)(N^N)][PF6] complexes and reference compound [Ir(ppy)2(4)][PF6] in CH2Cl2 (1.00 × 10−5 mol L−1) excited at 280 nm. | ||
![]() | ||
| Fig. 7 Normalized solid state emission spectra of [Ir(1)(N^N)][PF6] complexes and reference compound [Ir(ppy)2(4)][PF6] excited at 280 nm. | ||
| Complex cation | CH2Cl2 solution | Powder | ||
|---|---|---|---|---|
| λ maxem [nm] | PLQYb [%] | λ maxem [nm] | PLQY [%] | |
| a λ exc = 280 nm. b Argon degassed. c A value of 555 nm has been reported for a CH2Cl2 solution of [Ir(ppy)2(4)][PF6].23 | ||||
| [Ir(1)(2)]+ | 434, 590 | 16 | 578 | 35 |
| [Ir(1)(3)]+ | 361, 601 | 6 | 581 | 19 |
| [Ir(1)(4)]+ | 425, 564 | 9 | 567 | 20 |
| [Ir(1)(5)]+ | 435, 578 | 18 | 575 | 21 |
| [Ir(ppy)2(4)]+ | 420, 558c | 1 | 540 | 12 |
| CH2Cl2 solution | Powder | ||||||
|---|---|---|---|---|---|---|---|
| Complex cation | λ maxem [nm] |
τ
av b [ns] |
τ 1 [ns] (A1) | τ 2 [ns] (A2) |
τ
av b [ns] |
τ 1 [ns] (A1) | τ 2 [ns] (A2) |
| a λ exc = 280 nm; lifetimes were measured using a Hamamatsu Quantaurus-Tau (see Experimental section) and some values of λmaxem differ slightly from those in Table 2 where spectra were recorded on a Shimadzu RF-5301PC spectrofluorometer. b Biexponential fit using the equation τav = ∑Aiτi/∑Ai where Ai is the pre-exponential factor of the lifetime. | |||||||
| [Ir(1)(2)]+ | 587 | 381 | 370 (17 296) |
576 (624) | 545 | 436 (9095) | 703 (3884) |
| [Ir(1)(3)]+ | 601 | 159 | 145 (7394) | 649 (49) | 407 | 374 (11 665) |
686 (740) |
| [Ir(1)(4)]+ | 564 | 144 | 135 (72 891) |
245 (3871) | 678 | 569 (28 144) |
949 (6768) |
| [Ir(1)(4)]+ | 420 | 4 | 3 (44 137) |
56 (27) | |||
| [Ir(1)(5)]+ | 579 | 365 | 344 (16 410) |
543 (1241) | 462 | 337 (8836) | 579 (5543) |
The emission spectra of the [Ir(1)(N^N)][PF6] complexes as amorphous thin films with the composition used in LEC devices were also recorded (Fig. S1†). The spectra are similar and follow the same trends discussed above for the powder spectra with the emission maxima slightly red-shifted ([Ir(1)(2)][PF6]: 583 nm, [Ir(1)(3)][PF6]: 589 nm, [Ir(1)(4)][PF6]: 574 nm, [Ir(1)(5)][PF6]: 579 nm). The PLQY values (see Table 4) obtained in thin film are comparable to those determined in powder (21.9% for [Ir(1)(5)][PF6], 17.5% for [Ir(1)(4)][PF6] and 15.5% for [Ir(1)(3)][PF6]) except for [Ir(1)(2)][PF6] which shows a significantly lower PLQY of 22.6% compared to 35% in the powder (Table 2).
:
[Bmim][PF6] 4
:
1 molar ratio
| LEC | Ir-iTMC |
t
max a/h |
L max/cd m−2 |
t
1/2 b/h |
Efficacyc/cd A−1 | PCEd/lm W−1 | EQEe/% | PLQYf/% |
|---|---|---|---|---|---|---|---|---|
| a Time to reach the maximum luminance Lmax. b Time to reach one-half of the maximum luminance. c Maximum efficacy. d Maximum power conversion efficiency. e Maximum external quantum efficiency. f Photoluminescence quantum yield in thin film using the same composition than for the device active layer (λexc = 320 nm). | ||||||||
| A | [Ir(1)(2)][PF6] | 48.6 | 91 | 785 | 4.2 | 2.1 | 1.8 | 22.6 |
| B | [Ir(1)(3)][PF6] | 52.1 | 67 | 581 | 3.1 | 1.7 | 1.3 | 17.5 |
| C | [Ir(1)(4)][PF6] | 61.8 | 54 | 626 | 2.5 | 1.3 | 0.8 | 15.5 |
| D | [Ir(1)(5)][PF6] | 0.6 | 133 | 31 | 5.4 | 2.2 | 2.1 | 21.9 |
| E | [Ir(ppy)(4)][PF6] | 3.3 | 146 | 26 | 6.4 | 3.7 | 2.6 | — |
A comparison of Fig. 6 and 7 reveals that each complex in solution, most noticeably [Ir(1)(4)][PF6], exhibits a second emission at higher energies than the dominant band, and that this emission band is absent in the powder samples. Since the aim of introducing the naphthyl domain into the [Ir(C^N)2(N^N)]+ complex was to generate a dual emitter, the appearance of the second band, albeit weak, demanded further investigation. The absence of the band in the solid state samples suggested that the high-energy emission might arise from dissociated ligand. Fig. 8 shows an overlay of the normalized emission spectra of [Ir(1)(4)][PF6], [Ir(ppy)2(4)][PF6], H21 and 4. Both [Ir(1)(4)][PF6] and [Ir(ppy)2(4)][PF6] exhibit a band with λmaxem ∼ 420 nm (Table 2), suggesting that the origin of the emission is not the naphthyl domain. Note that the previously reported emission spectrum of [Ir(ppy)2(4)][PF6] did not extend to below 450 nm.23 Model compounds HOnaphth (HOnaphth = 2,7-bis[2-(2-hydroxyethoxy)ethoxy]naphthalene) and H21 both show a structured emission at 331 and 344 nm typical of naphthalene;46 in addition H21 shows an emission at 410 nm. Unexpectedly, excitation of a CH2Cl2 solution of ligand 4 gives rise to an extremely similar PL spectrum as that of H21 and HOnaphth (Fig. 8). Repeated measurements for 4 and comparisons with the emission spectrum of 5 and with the reported emission spectrum of 4,4′-di-tert-butyl-2,2′-bipyridine (λmaxem = 357 nm)47 validated the result.
![]() | ||
| Fig. 8 Normalized solution emission spectra of [Ir(1)(4)][PF6], [Ir(ppy)2(4)][PF6], H21, HOnaphth and 4 in CH2Cl2 (∼1.00 × 10−5 mol L−1). Compounds were excited at 280 nm. | ||
Excitation spectra of [Ir(1)(4)][PF6] and [Ir(ppy)2(4)][PF6] (Fig. S2†) reveal the origins of the 564 and 425 nm emissions in [Ir(1)(4)][PF6], and the 558 and 420 nm bands in [Ir(ppy)2(4)][PF6]. The low-energy MLCT emissions for both complexes arise from absorptions over the whole region from 230 to 500 nm, and the profiles of the excitation spectra are similar to the absorption spectra shown in Fig. 5, but, as expected, lack the naphthyl band at 237 nm. The 425 or 420 nm emission results from two absorption bands at ∼258 and ∼358 nm (Fig. 9, black and cyan spectra). The excitation spectra for [Ir(1)(4)][PF6] and [Ir(ppy)2(4)][PF6] resemble that of 4 (Fig. 9). In contrast, the excitation spectrum of H21 for the emission at 410 nm (Fig. 9, green spectrum) is broader. The similarity between the excitation spectra of [Ir(1)(4)][PF6], [Ir(ppy)2(4)][PF6] and free ligand 4 indicate that the high-energy emission around 420 nm arises from a fluorescent excited state of coordinated ligand 4. Although an extremely low concentration impurity could be responsible for such observations, the method of measuring and the reproducibility of recorded solutions of independently synthesized compounds makes this scenario unlikely.
![]() | ||
| Fig. 9 Normalized excitation spectra for CH2Cl2 solutions (1.00 × 10−5 M) of [Ir(1)(4)][PF6], [Ir(ppy)2(4)][PF6], H21 and 4 each at a fixed emission wavelength of 420 nm. | ||
The PLQY increases from 1% to 9% on going from [Ir(ppy)2(4)][PF6] to [Ir(1)(4)][PF6], but this is not necessarily a consequence of partial energy transfer from the naphthyl unit (PLQY = 15% for H21) to the Ir(III) coordination sphere. Excited-state lifetime measurements of [Ir(1)(4)][PF6] (τav = 144 ns for the 564 nm and τav = 4 ns for the 420 nm emission band, Table 3) corroborate the fact that the higher-energy emission arises from a fluorescent excited state being in the same range as for H21 (τav = 6 ns for the 420 nm emission band).
The geometry of the complexes in their ground electronic state (S0) was optimized without imposing any symmetry restriction. The calculated geometries reproduce the trends observed typically on this type of complexes, showing a distorted octahedral coordination of the iridium atom.6,48Fig. 10 displays the minimum-energy optimized structures of complexes [Ir(1)(2)]+ and [Ir(1)(3)]+ as representative examples. Table S1† summarizes the values of the geometrical parameters defining the iridium coordination sphere and of selected distances and dihedral angles. For all the complexes, the pendant phenyl rings introduced as R and R1 substituents in the N^N ligand (Scheme 2) present intracation face-to-face π-stacking interactions with the phenyl rings of the closest ppy in the cyclometallating ligand. The calculated centroid–centroid distances between rings A and G in Scheme 2 range from 3.40 to 3.56 Å, in good agreement with the X-ray values reported for complexes [Ir(ppy)2(2)]+ and [Ir(ppy)2(3)]+.48 The steric hindrance produced by the phenyl substituents induces a twisting between the rings of the bpy ligand. The twisting angle is higher for [Ir(1)(2)]+ (32.8°), [Ir(1)(4)]+ (38.3°) and [Ir(ppy)2(4)]+ (37.1°) than for [Ir(1)(3)]+ (21.3°) and [Ir(1)(5)]+ (24.9°) because in the former two pendant phenyl rings are introduced in R and R1 positions. For all the [Ir(1)2(N^N)]+ complexes, the cyclometallating ligand 1 remains in an expanded disposition (Fig. 10), featuring a distance from the centroid of the naphthalene group to the iridium atom of 10.6–10.8 Å.
The geometry of the complexes in their first triplet excited state (T1) was also optimized using the spin-unrestricted UB3LYP-D3/(6-31G**+LANL2DZ) approach. It is worth to note that the face-to-face intracation interactions observed in the ground state are preserved in T1, and that the naphthalene-iridium distance remains almost constant. The most important changes are found for the bpy ligand that becomes more planar in T1 (twisting angles in the 13–18° range, Table S1, ESI†). This points to a higher electron density on the bpy ligand in T1, which would stabilize the planar form as it favors the electronic delocalization.
Fig. 11 displays the isovalue contours calculated for the highest-occupied (HOMO) and lowest-unoccupied (LUMO) molecular orbitals of complex [Ir(1)(2)]+ as a representative example. Orbitals HOMO−1 and LUMO+7, which respectively correspond to the highest-occupied and lowest-unoccupied molecular orbitals centred on the naphthalene group of ligand 1, are also displayed. The topology of the corresponding MOs of all the [Ir(1)(N^N)]+ complexes fully reproduces that of the selected example. The table inserted in Fig. 11 summarizes the MO energies calculated for the [Ir(1)(N^N)]+ complexes and compares them with those obtained for the HOMO and LUMO of [Ir(ppy)2(4)]+. As it is usually found for ppy-based cyclometallated Ir-iTMCs,6,44,45 the HOMO results from a mixture of dπ orbitals of Ir(III) and phenyl π orbitals, with some contribution from the pyridine rings, of the cyclometallating ligand, whereas the LUMO is located over the bpy of the ancillary ligand. The similar energy values estimated for the HOMO of [Ir(1)(N^N)]+ are in good agreement with the experimental Eox1/2 values (Table 1), which do not change greatly along the series. The introduction of electron-releasing tert-butyl groups in the N^N ligand slightly destabilizes the HOMO of [Ir(1)(4)]+ and [Ir(1)(5)]+, in good agreement with the slightly lower Eox1/2 recorded for these complexes. The effect of the substituents is higher in the LUMO, centred on the ancillary ligand. It becomes stabilized along the series [Ir(1)(4)]+ (−2.30 eV) > [Ir(ppy)2(4)]+ (−2.34 eV) > [Ir(1)(5)]+ (−2.41 eV) > [Ir(1)(2)]+ (−2.47 eV) > [Ir(1)(3)]+ (−2.56 eV), in quite good correlation with the less negative Ered1/2 values recorded along this series (Table 1).
The HOMO–LUMO gaps calculated for [Ir(1)(2)]+ (3.20 eV) and [Ir(1)(3)]+ (3.13 eV) are lower than those computed for [Ir(1)(4)]+ (3.33 eV) and [Ir(1)(5)]+ (3.24 eV) in good accord with the trends observed for the electrochemical gap ΔE1/2. If light emission in these complexes originates from a state described by the HOMO → LUMO excitation, MO calculations predict that [Ir(1)(4)]+ would emit more in the blue that the rest of the complexes and close to the emission of [Ir(ppy)2(4)]+. This is in fact the trend observed in solution for the λmaxem values (Table 2). The significant energy difference between the LUMO and the LUMO+7 centred on the naphthalene moiety leads us to expect that states associated to this moiety will appear at high energies and will not contribute to the emission.
The nature of the low-lying triplet states was first studied by performing TD-DFT calculations at the optimized geometry of the ground state (S0). The vertical excitation energies and electronic descriptions computed for the lowest-lying triplet states of the [Ir(1)(N^N)]+ complexes and those of [Ir(ppy)2(4)]+ are given in Table S2.† All the complexes present a first triplet excited state (T1) mainly defined by the HOMO → LUMO excitation, which implies an electron transfer from the Ir-ppy environment, where the HOMO is localized, to the bpy ligand, where the LUMO resides (see Fig. 11). The T1 state therefore shows a mixed 3MLCT/3LLCT character, and the calculated excitation energies ([Ir(ppy)2(4)]+: 2.56 eV > [Ir(1)(4)]+: 2.55 eV > [Ir(1)(5)]+: 2.53 eV > [Ir(1)(2)]+: 2.48 eV > [Ir(1)(3)]+: 2.43 eV) follow the trend expected on the basis of the HOMO–LUMO gap. The T1 state was further examined by optimizing its structure at the spin-unrestricted UB3LYP-D3 level as described above. The unpaired-electron spin density distribution calculated for this state (Fig. 12a) matches the topology of the HOMO → LUMO excitation and therefore corroborates the electron transfer from the Ir-ppy environment to the N^N ligand, thus confirming the 3MLCT/3LLCT character of T1. The electronic nature predicted for T1 is in good agreement with the broad an unstructured shape of the emission bands observed experimentally (Fig. 6 and 7).
Back to the TD-DFT calculation of triplet states, the first excited state involving the naphthyl group of ligand 1 is found 0.3–0.4 eV above T1 (Table S2†). As was to be expected, it is mainly described by the HOMO−1 → LUMO+7 excitation, and its vertical excitation energy from S0 remains almost constant (2.85–2.86 eV) for all the [Ir(1)(N^N)]+ complexes, as the structural differences between these complexes concern the N^N ligand from which the naphthyl is far apart. The spin-density distribution calculated for this state is shown in Fig. 12b and clearly reflects the localization of the electronic transition on the naphthyl unit. After full-geometry relaxation at the UB3LYP-D3 level, the naphthyl-centred triplet state is calculated 0.38, 0.41, 0.28 and 0.30 eV above the HOMO → LUMO T1 triplet for [Ir(1)(2)]+, [Ir(1)(3)]+, [Ir(1)(4)]+ and [Ir(1)(5)]+, respectively. This energy difference is large enough to discard the contribution of the naphthyl-centred triplet excited state to the phosphorescence emission, which is predicted to occur from the lowest-energy, HOMO → LUMO, 3MLCT/3LLCT T1 state.
TD-DFT calculations were also performed for singlet excited states to investigate the nature of the bands appearing in the experimental absorption spectra (Fig. 5). Singlet excited states with mainly MLCT character and small oscillator strengths are found in the 400 nm region, followed by ligand-centred (LC) π → π* states involving both the ppy and bpy moieties with some MLCT/LLCT character between 400 and 300 nm. No excited state implying the naphthyl moiety of ligand 1 is found below 300 nm. According to the theoretical results, it is unlikely that emission, fluorescent or phosphorescent, takes place from naphthyl-centred excited states.
:
IL ratio of 4
:
1.50–52
Fig. 13 displays the luminance versus time curves for LECs A–E, and Table 4 collects the parameters that summarize the device performances. The turn-on time (tmax) is defined as the time to reach the maximum luminance (Lmax) and the device lifetime (t1/2) is the time to reach one-half of Lmax after this value is achieved. LEC devices can be divided in two groups with LECs D and E, which incorporate complexes [Ir(1)(5)][PF6] and [Ir(ppy)(4)][PF6], showing shorter tmax and faster luminance decays than LECs A–C (Table 4). LECs D and E show tmax, and t1/2 below 3.5 and 35 hours, respectively, whereas devices A–C have longer turn-on times above 45 hours and a much more stable behaviour with t1/2 values ranging from 581 to 785 hours. In contrast, LECs D and E achieve higher luminance values above 100 cd m−2 (133 and 146 cd m−2, respectively) compared with LECs A–C, which show maximum luminances of 91, 67 and 54 cd m−2, respectively. The maximum values obtained for the efficacy, the power conversion efficiency (PCE) and the external quantum efficiency (EQE) of LEC D (5.4 cd A−1, 2.2 lm W−1 and 2.1%, respectively) are slightly lower than those found for LEC E (6.4 cd A−1, 3.7 lm W−1 and 2.6%), and are similar to those reported for other orange-emitting Ir-iTMC-LECs under pulsed current operation.49,53 The efficiency, PCE and EQE values found for LECs A–C are smaller than those obtained for D, and decrease in passing from A to B and to C (Table 4). This trend is in agreement with the trend of the PLQY for the amorphous thin film (device environment) of the complexes [Ir(1)(2)][PF6], [Ir(1)(3)][PF6] and [Ir(1)(4)][PF6].
![]() | ||
| Fig. 13 Luminance vs. time for the LECs driven using a pulsed current of 25 A m−2 (average current density) at a frequency of 1000 Hz and duty cycles of 50%. | ||
The main differentiator between the complexes [Ir(1)(5)][PF6] and [Ir(ppy)(4)][PF6] used to prepare LECs D and E is the naphthyl-bridged functionalization of the ppy ligands in the case of LEC D. Yet both the device responses tmax and t1/2 are comparable for these two devices, which indicates that the ppy functionalization does not influence the ionic movement in the films. Therefore, the longer tmax and t1/2 found for LECs A–C should not be ascribed to the large size of ligand 1. Moreover, the [Ir(1)(2)][PF6] complex, used in the fabrication of LEC A, exhibits the highest PLQY in thin film (22.6%), whereas the device efficiency is superior for D incorporating a complex with a slightly lower PLQY (21.9%). Considering these PLQY values and a typical outcoupling of 20%, the theoretical maximum EQEs predicted for LECs A and D when all injected electrons and holes combine have very similar values of 4.5% and 4.4%, respectively. As the EQE achieved for D (2.1%) is slightly closer to the theoretical value than that obtained for A (1.8%), we can hypothesize that the lower exciton-quenching for D could be the result of a better-balanced carrier injection due to the faster response. Hence, for the devices with slower response (devices A–C), lower efficiencies were achieved. Comparable characteristics have been observed for similar orange complexes reported in LECs.43
The electroluminescence (EL) spectrum of all the LECs was registered during the operation of the devices (Fig. S3†). All the LECs show orange electroluminescence with maxima in the 580–590 nm range but for LEC C ([Ir(1)(4)][PF6]) which emits at 575 nm. The EL spectra are similar to the PL spectra recorded in powder (Fig. 7) and in thin film (Fig. S1†).
Footnote |
| † Electronic supplementary information (ESI) available: Tables S1 and S2: Computed geometrical parameters for Ir-iTMCs and lowest triplet excited states. Fig. S1–S3: Photoluminescence, excitation and electroluminescence spectra. See DOI: 10.1039/c6dt03082c |
| This journal is © The Royal Society of Chemistry 2016 |