Highly Efficient Photo- and Electroluminescence from Two-Coordinate

Feb 15, 2019 - A series of six luminescent two-coordinate Cu(I) complexes were investigated bearing nonconventional N-heterocyclic carbene ligands, ...
0 downloads 0 Views 2MB Size
Article Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

pubs.acs.org/JACS

Highly Efficient Photo- and Electroluminescence from TwoCoordinate Cu(I) Complexes Featuring Nonconventional N‑Heterocyclic Carbenes Shuyang Shi,† Moon Chul Jung,‡ Caleb Coburn,§ Abegail Tadle,† Daniel Sylvinson M. R.,† Peter I. Djurovich,† Stephen R. Forrest,§,∥,⊥ and Mark E. Thompson*,†,‡

Downloaded via WASHINGTON UNIV on February 15, 2019 at 19:25:58 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



Department of Chemistry, and ‡Mork Family Department of Chemical Engineering and Materials Science, University of Southern California, Los Angeles, California 90089, United States § Department of Physics, ∥Department of Electrical and Computer Engineering, and ⊥Department of Materials Science and Engineering, University of Michigan, Ann Arbor, Michigan 48109, United States S Supporting Information *

ABSTRACT: A series of six luminescent two-coordinate Cu(I) complexes were investigated bearing nonconventional N-heterocyclic carbene ligands, monoamido-aminocarbene (MAC*) and diamidocarbene (DAC*), along with carbazolyl (Cz) as well as mono- and dicyano-substituted Cz derivatives. The emission color can be systematically varied over 270 nm, from violet to red, through proper choice of the acceptor (carbene) and donor (carbazolyl) groups. The compounds exhibit photoluminescent quantum efficiencies up to 100% in fluid solution and polystyrene films with short decay lifetimes (τ ≈ 1 μs). The radiative rate constants for the Cu(I) complexes (kr = 105−106 s−1) are comparable to state of the art phosphorescent emitters with noble metals such as Ir and Pt. All complexes show strong solvatochromism due to the large dipole moment of the ground states and the transition dipole moment that is in the opposite direction. Temperature-dependent studies of (MAC*)Cu(Cz) reveal a small energy separation between the lowest singlet and triplet states (ΔES1−T1 = 500 cm−1) and an exceptionally large zero-field splitting (ZFS = 85 cm−1). Organic light-emitting diodes (OLEDs) fabricated with (MAC*)Cu(Cz) as a green emissive dopant have high external quantum efficiencies (EQE = 19.4%) and brightness of 54 000 cd/m2 with modest roll-off at high currents. The complex can also serve as a neat emissive layer to make highly efficient OLEDs (EQE = 16.3%).



INTRODUCTION Organic light-emitting diode (OLED) technology is widely used in commercial applications such as displays and lighting. For luminescent materials to be applied in OLEDs, it is essential that both singlet (S1) and triplet (T1) excitons be harvested in the emission layer to achieve high efficiency. Organometallic complexes with noble metals such as Ir(III), Pt(II), Re(I), and Ru(II) can induce efficient spin−orbit coupling (SOC) between T1 and S1, and can achieve high radiative rate constants from T1 to S0.1−8 Thermally activated delayed fluorescence (TADF) is an alternate mechanism to harvest both singlet and triplet excitons that does not require heavy metals for efficient SOC. Cu complexes have been reported to undergo efficient phosphorescence from charge transfer (CT) states via TADF.9−19 Most of the reported luminescent mononuclear Cu(I) complexes are four-13,15,20−23 or three-coordinate,24−30 and typically undergo structural distortion in the excited state. The geometric change leads to large Franck−Condon factors and an increase in nonradiative decay rates due to strong vibronic coupling between the ground and excited states.31 Therefore, although photoluminescence quantum yields (ΦPL) of copper complexes © XXXX American Chemical Society

can be high in the crystalline solid, the efficiency markedly decreases in fluid solution. Interestingly, we recently reported a two-coordinate cationic Cu carbene complex [(DAC)2Cu][BF4] (DAC = 1,3-bis(mesityl)-5,5-dimethyl-4,6-diketopyrimidinyl-2-ylidene) that is brightly emissive in fluid solution with ΦPL = 65%,32 which is among the highest values reported for mononuclear Cu(I) complexes.21,29,33 The high ΦPL of this complex demonstrates that a two-coordinated Cu center can achieve high efficiency in fluid media if the ligands provide sufficient steric hindrance to prevent structural distortion in the excited state. Although the nonradiative rate constant (knr = 103−104 s−1) of [(DAC)2]Cu(BF4) in solution is low as compared to most of the other luminescent Cu complexes,20,25 the radiative rate constant (kr = ∼104 s−1) is at least an order of magnitude smaller than those of Ir- and Pt-based emitters (kr = 105−106 s−1).2 Recent studies have also appeared describing highly efficient emission from two-coordinated Cu(I) complexes bearing cyclic alkyl-amino carbene (CAAC) liReceived: November 26, 2018

A

DOI: 10.1021/jacs.8b12397 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society gands.27,34−37 These complexes have fast radiative rates (kr > 105 s−1) and highlight two-coordinate Cu(I) structures, particularly (CAAC)Cu(amide) complexes,35,37 as promising luminescent complexes for application to OLEDs and other optoelectronic applications, provided that the appropriate carbene is chosen as a ligand. In this work, we characterize a series of two-coordinate neutral Cu complexes (Scheme 1) bearing nonconventional cyclic

C(CH3)2), 2.95 (sept, J = 6.5 Hz, 4H, CH(CH3)2), 3.77 (s, 2H, CCH2Cl), 7.08 (m, 3H, Ar−H), 7.27 (m, 2H, Ar−H), 7.39 (m, 1H, Ar−H), 8.59 (s, 1H, NCHN). 13C NMR δC (CDCl3, 101 MHz, 298 K): 23.17 (s, CH(CH3)2), 24,11 (s, CH(CH3)2), 24.22 (s, C(CH3)2), 24.83 (s, CH(CH3)2), 27.35 (s, CH(CH3)2), 28.94 (s, CH(CH3)2), 46.25 (s, C(CH3)2), 53.25 (s, CCH2Cl), 123.05 (s, m-ArH), 124.01 (s, m-ArH), 124.20 (s, p-ArH), 129.18 (s, p-ArH), 132.95 (s, o-NAr), 138.91 (s, o-NAr), 145.21 (ipso-NAr), 145.51 (ipso-NAr), 148.51 (s, NCN), 174.74 (s, CO). MALDI-TOF: m/z calculated, 447.34 [M − Cl−]+; found, 447.64 [M]+.

Scheme 1. Molecular Structures of 1−6

1,3-Bis(2,6-diisopropylphenyl)-5,5-dimethyl-4-keto-tetrahydropyrimidin-1-ium Chloride (1b). 1c (650 mg, 1.34 mmol) was dissolved in toluene (20 mL), and the solution was refluxed for 16 h at 110 °C during which a white precipitate formed. The reaction mixture was cooled to rt, and the white precipitate was collected by vacuum filtration and washed with cold toluene. Yield: 450 mg (69%). 1H NMR δH (CDCl3, 400 MHz, 298 K): 1.20 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.30 (dd, J = 9.9, 6.8 Hz, 12H, CH(CH3)2), 1.40 (d, J = 6.6 Hz, 6H, CH(CH3)2), 1.76 (s, 6H, C(CH3)2), 3.04 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.25 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 4.66 (s, 2H, CCH2N), 7.20−7.30 (m, 4H, Ar−H), 7.45 (td, J = 7.9, 4.0 Hz, 2H, Ar− H), 9.82 (s, 1H, NCH−N). 13C NMR δC (CDCl3, 101 MHz, 298 K): 24.00 (s, CH(CH3)2), 24.41 (s, CH(CH3)2), 24.59 (s, C(CH3)2), 24.77 (s, CH(CH3)2), 24.87 (s, CH(CH3)2), 28.98 (s, CH(CH3)2), 29.23 (s, CH(CH3)2), 39.17 (s, C(CH3)2), 61.76 (s, CCH2N), 124.86 (s, m-ArH), 125.59 (s, m-ArH), 131.66 (s, p-ArH), 131.97 (s, p-ArH), 129.56 (s, o-NAr), 134.48 (s, o-NAr), 144.35 (ipso-NAr), 145.98 (ipsoNAr), 159.29 (s, NCN), 169.67 (s, CO). Anal. Calcd for C30H43ClN2O: C, 74.58; N, 5.80; H, 8.97. Found: C, 74.31; N, 6.17; H, 8.96. (N,N′-Bis(diisopropylphenyl)-5,5-dimethyl-4-keto-tetrahydropyrimidin-2-ylidene)-Cu(I) Chloride (MAC*CuCl) (1a). KHMDS (136 mg, 0.68 mmol) was added to a THF solution (20 mL) of 1b (300 mg, 0.62 mmol) at rt, and the solution was stirred for 1 h before CuCl (67 mg, 0.68 mmol) was added. The reaction mixture was stirred at rt for 16 h, filtered through Celite, and the solvent was concentrated to 3 mL under reduced pressure. Hexane (20 mL) was added to the solution, and a white precipitate formed. Yield: 300 mg (88%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.17 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.35 (m, 18H, CH(CH3)2), 1.59 (s, 6H, C(CH3)2), 3.13 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.38 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 4.14 (s, 2H, CCH2N), 7.30 (d, J = 7.6 Hz, 2H, m-ArH), 7.37 (d, J = 7.7 Hz, 2H, m-ArH), 7.42 (t, J = 7.7 Hz, 1H, p-ArH), 7.47 (t, J = 7.7 Hz, 1H, p-ArH). 13 C NMR δC (acetone-d6, 101 MHz, 298 K): 23.25 (s, CH(CH3)2), 23,67 (s, CH(CH3)2), 23.69 (s, CH(CH3)2), 23.75 (s, C(CH3)2), 24.25 (s, CH(CH3)2), 28.27 (s, CH(CH3)2), 28.58 (s, CH(CH3)2), 37.87 (s, C(CH3)2), 60.75 (s, CCH2N), 124.19 (s, m-ArH), 125.11 (s, m-ArH), 129.65 (s, p-ArH), 130.05 (s, p-ArH), 136.19 (s, o-NAr), 140.14 (s, o-NAr), 144.55 (ipso-NAr), 145.62 (ipso-NAr), 171.16 (s, CO), 208.89 (s, NCN). Anal. Calcd for C30H42ClCuN2O: C, 66.03; N, 5.13; H, 7.76. Found: C, 66.02; N, 5.43; H, 7.73. Synthesis of [(DAC*)Cu]2Cl2 (2a). KHMDS (450 mg, 2.28 mmol) was added to a THF solution (20 mL) of 2b (1.39 g, 2.28 mmol) at rt, and the solution was stirred for 1 h before CuCl (230 mg, 2.28 mmol) was added. The reaction mixture was stirred at rt for 16 h. The solvent was evaporated under reduced pressure, and the obtained red solid was redissolved in toluene (20 mL) and filtered through Celite. The filtrate

amidocarbenes38−41 including (MAC*)Cu(CzCN2) (1), (MAC*)Cu(CzCN) (2), (MAC*)Cu(Cz) (3), (DAC*)Cu(CzCN2) (4), (DAC*)Cu(CzCN) (5), and (DAC*)Cu(Cz) (6) (MAC = cyclic monoamido-aminocarbene, DAC = cyclic diamidocarbene, “*” indicates that the aryl group bound to N is 2,6-diisopropylphenyl, Cz = carbazole). The carbonyl groups incorporated into MACs and DACs increase the π-accepting properties of the carbene, and concomitantly lower the energy of the LUMO. The diisopropylphenyl group on the carbene ligands provides sufficient steric hindrance to suppress the formation of an exciplex between the complexes and the solvent. All of the complexes show efficient TADF with short decay lifetime in polystyrene films and solution. The radiative rates (kr = 105−106 s−1) are comparable to those of (CAAC)Cu(amide) compounds,37and, more surprising, to state-of-the-art emitters bearing noble metal such as Ir and Pt.2 OLEDs using complex 3 as dopants and neat emitters were grown using vapor deposition, and show an external quantum efficiency up to 19.4%, with only modest efficiency roll-off at high brightness.



EXPERIMENTAL SECTION

Synthesis. All reactions were carried out using standard Schlenk and glovebox techniques using dried and degassed solvents. The DAC*Cl42 precursor, 3-cyanocarbazole,43 and 3,6-dicyanocarbazole44 were synthesized by following the literature procedure. CuCl and carbazole were purchased from Sigma-Aldrich. 3-Chloropivaloyl chloride was purchased from Arctom Chemicals. NMR spectra were recorded on a Varian 400 NMR spectrometer. Synthesis of (MAC*)CuCl (1a). 3-Chloro-N-(2,6-diisopropylphenyl)-N′-((2,6-diisopropylphenylimino)methyl)-2,2-dimethylpropanamide (1c). N,N′-Bis(2,6-diisopropylphenyl) formamidine (500 mg, 1.37 mmol) and triethylamine (287 mL, 2.06 mmol) were dissolved in dichloromethane (20 mL) and stirred at 0 °C for 10 min, after which 3chloropivaloyl chloride (0.195 mL, 1.51 mmol) was added dropwise. The solution mixture was stirred for 3 h at 0 °C. The solvent was removed under reduced pressure to afford a white powder, which was extracted with toluene and filtered through Celite. Removal of the residual solvent afforded the product as a white solid. Yield: 650 mg (98%). 1H NMR δH (CDCl3, 400 MHz, 298 K): 1.14 (m, 18H, CH(CH3)2), 1.30 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.54 (s, 6H, B

DOI: 10.1021/jacs.8b12397 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

CH(CH3)2), 23.60 (s, CH(CH3)2), 23.82 (s, C(CH3)2), 24.38 (s, CH(CH3)2), 28.49 (s, CH(CH3)2), 28.71 (s, CH(CH3)2), 38.01 (s, C(CH3)2), 61.12 (s, CCH2N), 96.68 (s, CN−Cz), 115.07 (s, CH7(Cz)), 115.16 (s, CH1(Cz)), 117.01 (s, CH5(Cz)), 119.21 (s, CH4(Cz)), 121.42 (s, ipso-CN(Cz)), 123.21 (s, ipso-C(Cz)), 123.71 (s, CH3(Cz)), 124.15 (s, ipso-C(Cz)), 124.48 (s, CH6(Cz)), 124.90 (s, m-ArH), 125.59 (s, CH2(Cz)), 125.81 (s, m-ArH), 130.13 (s, p-ArH), 130.50 (s, p-ArH), 136.48 (s, o-Ar), 140.47 (s, o-Ar), 145.45 (ipso-N− Ar), 146.56 (ipso-N−Ar), 150.50 (s, ipso-N(Cz)), 151.59 (s, ipsoN(Cz)), 171.23 (s, CO), 209.25 (s, NCN). Anal. Calcd for C43H49CuN4O: C, 73.63; N, 7.99; H, 7.04. Found: C, 73.45; N, 8.16; H, 6.88. (MAC*)Cu(Cz) (3). The complex was made from (MAC*)CuCl (2.0 g, 3.67 mmol), Cz (613 mg, 3.67 mmol), and NaOtBu (353 mg, 3.67 mmol) as a yellow solid. Yield: 2.1 g (85%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.14−1.36 (m, 18H, CH(CH3)2), 1.43 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.66 (s, 6H, C(CH3)2), 3.30 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.55 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 4.26 (s, 2H, CCH2N), 5.56 (d, J = 8.1 Hz, 2H, CH1(Cz)), 6.71 (t, J = 7.3 Hz, 2H, CH3(Cz)), 6.81 (t, J = 6.9 Hz, 2H, CH2(Cz)), 7.53 (d, J = 7.8 Hz, 2H, m-ArH), 7.58 (d, J = 7.8 Hz, 2H, m-ArH), 7.71 (d, J = 6.9 Hz, 2H, CH4(Cz)), 7.73−7.79 (m, 2H, ArH). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.46 (s, CH(CH3)2), 23,58 (s, CH(CH3)2), 23.61 (s, CH(CH3)2), 23.82 (s, C(CH3)2), 24.37 (s, CH(CH3)2), 28.50 (s, CH(CH3)2), 28.72 (s, CH(CH3)2), 37.97 (s, C(CH3)2), 61.16 (s, CCH2N), 114.60 (s, CH1(Cz)), 114.95 (s, CH3(Cz)), 118.39 (s, CH4(Cz)), 122.73 (s, CH2(Cz)), 123.91 (s, ipso-C(Cz)), 124.78 (s, mArH), 125.70 (s, m-ArH), 129.93 (s, p-ArH), 130.30 (s, p-ArH), 136.47 (s, o-Ar), 140.51 (s, o-Ar), 145.37 (ipso-N−Ar), 146.46 (ipso-N−Ar), 149.85 (s, ipso-N(Cz)), 171.29 (s, CO), 209.83 (s, NCN). Anal. Calcd for C42H50CuN3O: C, 74.58; N, 6.21; H, 7.45. Found: C, 74.21; N, 6.01; H, 7.47. (DAC*)Cu(CzCN2) (4). The complex was made from (DAC*)CuCl (200 mg, 0.36 mmol), [Cz(CN)2] (78 mg, 0.36 mmol), and NaOtBu (35 mg, 0.36 mmol) as a yellow solid. Yield: 180 mg (68%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.25 (dd, J = 11.3 Hz, 6.8 Hz, 24H, CH(CH3)2), 1.95 (s, 6H, C(CH3)2), 3.23 (sept, J = 6.8 Hz, 4H, CH(CH3)2), 5.59 (d, J = 8.0 Hz, 2H, CH1(Cz)), 7.28 (d, J = 8.5 Hz, 2H, CH2(Cz)), 7.68 (d, J = 7.8 Hz, 4H, m-ArH), 7.97 (t, J = 7.8 Hz, 2H, pArH), 8.36 (s, 2H, CH3(Cz)). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.34 (s, CH(CH3)2), 23.72 (s, CH(CH3)2), 24.36 (s, C(CH3)2), 28.80 (s, CH(CH3)2), 52.37 (s, C(CH3)2), 99.53 (s, CN−Cz), 115.72 (s, CH1(Cz)), 120.49 (s, ipso-CN(Cz)), 123.56 (s, ipso-C(Cz)), 124.81 (s, CH3(Cz)), 125.62 (s, m-ArH), 127.39 (s, CH2(Cz), 131.41 (s, p-ArH), 135.20 (s, o-Ar), 146.59 (ipso-N−Ar), 152.14 (s, ipsoN(Cz)), 172.11 (s, CO), 213.90 (s, NCN). Anal. Calcd for C44H46CuN5O2 + 0.5H2O: C, 70.52; N, 9.34; H, 6.32. Found: C, 70.81; N, 9.26; H, 6.29. (DAC*)Cu(CzCN) (5). The complex was made from (DAC*)CuCl (200 mg, 0.36 mmol), CzCN (69 mg, 0.36 mmol), and NaOtBu (35 mg, 0.36 mmol) as an orange solid. Yield: 200 mg (76%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.25 (dd, J = 8.3 Hz, 6.8 Hz, 24H, CH(CH3)2), 1.94 (s, 6H, C(CH3)2), 3.22 (sept, J = 6.8 Hz, 4H, CH(CH3)2), 5.51 (d, J = 8.5 Hz, 1H, CH7(Cz)), 5.61 (d, J = 8.1 Hz, 1H, CH1(Cz)), 6.91 (t, J = 7.8 Hz, 1H, CH5(Cz)), 7.00 (t, J = 7.6 Hz, 1H, CH6(Cz)), 7.13 (d, J = 8.5 Hz, 1H, CH2(Cz)), 7.65 (d, J = 7.8 Hz, 4H, m-ArH), 7.89 (d, J = 7.8 Hz, 1H, CH4(Cz)), 7.91 (t, J = 7.8 Hz, 2H, pArH), 8.17 (s, 1H, CH3(Cz)). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.33 (s, CH(CH3)2), 23.69 (s, CH(CH3)2), 24.37 (s, C(CH3)2), 28.80 (s, CH(CH3)2), 52.20 (s, C(CH3)2), 97.27 (s, CN−Cz), 115.06 (s, CH7(Cz)), 115.09 (s, CH1(Cz)), 117.40 (s, CH5(Cz)), 119.31 (s, CH4(Cz)), 121.24 (s, ipso-CN(Cz)), 123.33 (s, ipso-C(Cz)), 123.77 (s, CH3(Cz)), 124.34 (s, ipso-C(Cz)), 124.67 (s, CH6(Cz)), 125.51 (s, m-ArH), 125.84 (s, CH2(Cz)), 131.22 (s, p-ArH), 135.15 (s, o-Ar), 146.49 (ipso-N−Ar), 150.33 (s, ipso-N(Cz)), 151.49 (s, ipso-N(Cz)), 172.17 (s, CO), 214.12 (s, NCN). Anal. Calcd for C43H47CuN4O2: C, 72.19; N, 7.83; H, 6.62. Found: C, 72.04; N, 7.88; H, 6.62. (DAC*)Cu(Cz) (6). The complex was made from (DAC*)CuCl (100 mg, 0.18 mmol), Cz (30 mg, 0.18 mmol), and NaOtBu (18 mg, 0.19 mmol) as a purple solid. After the precipitation, the compound was

was concentrated to 3 mL under reduced pressure. Hexane (20 mL) was added to the solution, and a red precipitate formed. Yield: 400 mg (31%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.18 (d, J = 6.8 Hz, 24H, CH(CH3)2), 1.33 (d, J = 6.8 Hz, 24H, CH(CH3)2), 1.86 (s, 12H, C(CH3)2), 3.03 (sept, J = 6.8 Hz, 8H, CH(CH3)2)), 7.38 (d, J = 7.8 Hz, 8H, m-ArH), 7.53 (d, J = 7.8 Hz, 4H, p-ArH). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.35 (s, CH(CH3)2), 23.58 (s, CH(CH3)2), 24.37 (s, C(CH3)2), 28.70 (s, CH(CH3)2), 51.92 (s, C(CH3)2), 124.78 (s, mArH), 130.69 (s, p-ArH), 135.02 (s, o-Ar), 145.42 (ipso-N−Ar), 172.23 (s, CO), 213.64 (s, NCN). Anal. Calcd for C60H80Cl2Cu2N4O4: C, 64.38; N, 5.01; H, 7.19. Found: C, 64.16; N, 5.29; H, 7.19.

Synthesis of Complexes 1−6. General Procedure. The carbazole ligand and NaOtBu were dissolved in THF and stirred for 3 h at rt. (carbene)CuCl was added to the reaction mixture and stirred for 16 h. The resulting mixture was filtered through Celite, and the solvent was removed under reduced pressure to afford a solid. The solid was redissolved in dichloromethane, and hexane was added to precipitate the desired product.

(MAC*)Cu(CzCN2) (1). The complex was made from (MAC*)CuCl (160 mg, 0.29 mmol), [Cz(CN)2] (64 mg, 0.29 mmol), and NaOtBu (29 mg, 0.30 mmol) as a white solid. Yield: 168 mg (85%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.20−1.28 (m, 18H, CH(CH3)2), 1.43 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.68 (s, 6H, C(CH3)2), 3.30 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 3.55 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 4.34 (s, 2H, CCH2N), 5.59 (d, J = 8.5 Hz, 2H, CH1(Cz)), 7.22 (d, J = 10.8 Hz, 2H, CH2(Cz)), 7.58 (d, J = 7.8 Hz, 2H, m-ArH), 7.64 (d, J = 7.8 Hz, 2H, m-ArH), 7.84 (m, 2H, p-ArH), 8.34 (s, 2H, CH3(Cz)). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.43 (s, CH(CH3)2), 23.58 (s, CH(CH3)2), 23.63 (s, CH(CH3)2), 23.81 (s, C(CH3)2), 24.40 (s, CH(CH3)2), 28.48 (s, CH(CH3)2), 28.70 (s, CH(CH3)2), 38.06 (s, C(CH3)2), 61.08 (s, CCH2N), 99.09 (s, CN-Cz), 115.76 (s, CH1(Cz)), 120.61 (s, ipso-CN(Cz)), 123.43 (s, ipso-C(Cz)), 124.74 (s, CH3(Cz)), 125.02 (s, m-ArH), 125.91 (s, m-ArH), 127.16 (s, CH2(Cz), 130.31 (s, p-ArH), 130.69 (s, p-ArH), 136.51 (s, o-Ar), 140.44 (s, o-Ar), 145.53 (ipso-N−Ar), 146.66 (ipso-N−Ar), 152.27 (s, ipso-N(Cz)), 171.18 (s, CO), 208.74 (s, NCN). Anal. Calcd for C44H48CuN5O: C, 72.75; N, 9.64; H, 6.66. Found: C, 72.70; N, 9.36; H, 6.83. (MAC*)Cu(CzCN) (2). The complex was made from (MAC*)CuCl (200 mg, 0.37 mmol), CzCN (71 mg, 0.37 mmol), and NaOtBu (36 mg, 0.37 mmol) as a white solid. Yield: 200 mg (78%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.21−1.31 (m, 18H, CH(CH3)2), 1.43 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.67 (s, 6H, C(CH3)2), 3.30 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.55 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 4.31 (s, 2H, CCH2N), 5.51 (d, J = 8.5 Hz, 1H, CH7(Cz)), 5.63 (d, J = 7.9 Hz, 1H, CH1(Cz)), 6.87 (t, J = 7.4 Hz, 1H, CH5(Cz)), 6.95 (t, J = 7.6 Hz, 1H, CH6(Cz)), 7.09 (d, J = 8.5 Hz, 1H, CH2(Cz)), 7.55 (d, J = 7.8 Hz, 2H, m-ArH), 7.61 (d, J = 7.8 Hz, 2H, m-ArH), 7.79−7.84 (m, 2H, pArH), 7.87 (d, 1H, CH4(Cz)), 8.15 (s, 1H, CH3(Cz)). 13C NMR δC (acetone-d6, 101 MHz, 298 K): 23.44 (s, CH(CH3)2), 23.59 (s, C

DOI: 10.1021/jacs.8b12397 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Table 1. Redox Data for Complexes 1−6 and Carbene−CuCl and Dipole Moments Calculated for 1−6 in the Ground State (S0), Lowest Ligand-Localized Triplet State (3Cz), and Charge Transfer Singlet State (1ICT) dipole moment μ (μZ)c

potentials complex 1 2 3 4 5 6 (MAC*)CuCl (DAC*)CuCl

Eoxa

(V)

0.60 0.38 0.17 0.60 0.38 0.17 0.91 1.03

a

Ered (V)

HOMO (eV)

LUMO (eV)

μgs S0

μes 3Cz

μes 1ICT

−2.45 −2.45 −2.50 −1.60 −1.56 −1.60 −2.50 −1.60

−5.48 −5.23 −4.98 −5.48 −5.23 −4.98 −5.84 −5.97

−1.94 −1.94 −1.88 −2.94 −2.99 −2.94 −1.88 −2.94

19.1 (18.9) 14.9 (14.8) 10.6 (10.2) 17.2 (17.2) 13.7 (13.0) 8.3 (8.3)

19.1 (18.8) 12.5 (12.4) 9.9 (9.4) 16.7 (16.7) 11.4 (10.3) 0.60 (0.5)

5.5 (−4.3) 8.6 (−8.6) 13.7 (−13.2) 8.9 (−8.9) 13.5 (−12.8) 17.0 (−17.0)

b

b

Redox potentials obtained in acetonitrile with 0.1 M TBAPF6 versus internal Fc+/Fc. bHOMO = 1.15(Eox) + 4.79; LUMO = 1.18(Ered) − 4.83.38 Obtained from TD-DFT calculations (CAM-B3LYP/LACVP**) using geometry optimized structures. μZ is the projection of μ along the Cu−N bond axis. Negative values indicate dipole moments are opposite in direction from the dipole moments in the ground state. All dipole moment values are reported in Debye. a c

purified by sublimation. Yield: 86 mg (70%). 1H NMR δH (acetone-d6, 400 MHz, 298 K): 1.25 (dd, J = 6.8, 5.8 Hz, 24H, CH(CH3)2), 1.93 (s, 6H, C(CH3)2), 3.20 (sept, J = 6.8 Hz, 4H, CH(CH3)2)), 5.55 (d, J = 8.1 Hz, 2H, CH1(Cz)), 6.75 (t, J = 7.7 Hz, 2H, CH3(Cz)), 6.86 (t, J = 7.5 Hz, 2H, CH2(Cz)), 7.62 (d, J = 7.9 Hz, 4H, m-ArH), 7.74 (d, J = 7.6 Hz, 2H, CH4(Cz)), 7.86 (d, J = 7.8 Hz, 2H, p-ArH). 13C NMR δC (acetoned6, 101 MHz, 298 K): 23.34 (s, CH(CH3)2), 23.65 (s, CH(CH3)2), 24.38 (s, C(CH3)2), 28.84 (s, CH(CH3)2), 51.98 (s, C(CH3)2), 114.59 (s, CH1(Cz)), 115.50 (s, CH3(Cz)), 118.40 (s, CH4(Cz)), 122.97 (s, CH2(Cz)), 124.13 (s, ipso-C(Cz)), 125.37 (s, m-ArH), 131.01 (s, pArH), 135.12 (s, o-Ar), 146.39 (ipso-N−Ar), 149.71 (s, ipsoN(Cz)),172.24 (s, CO), 214.46 (s, NCN). MALDI-TOF: m/z calcd, 689.30 [M]+; found, 690.54 [M]+. X-ray Crystallography. The X-ray intensity data were measured on a Bruker APEX DUO system equipped with a TRIUMPH curvedcrystal monochromator and a Mo Kα fine-focus tube (λ = 0.71073 Å). The frames were integrated with the Bruker SAINT software package using a SAINT V8.37A (Bruker AXS, 2013) algorithm. Data were corrected for absorption effects using the multiscan method (SADABS). All non-hydrogen atoms were refined anisotropically. CCDC 1873677, 1873680, 1873678, and 1873679 contain the supplementary crystallographic data for 1−4, respectively. These data can be obtained free of charge at http://www.ccdc.cam.ac.uk/conts/ retrieving.html, or from the Cambridge Crystallographic Data Centre, CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax, 44-1223-336033; or e-mail, [email protected]). Electrochemical Measurements. Cyclic voltammetry and differential pulsed voltammetry were performed using a VersaSTAT 3 potentiostat. Anhydrous acetonitrile (DriSolv) was used as the solvent under inert atmosphere, and 0.1 M tetra(n-butyl)ammonium hexafluorophosphate (TBAF) was used as the supporting electrolyte. A glassy carbon rod was used as the working electrode, a platinum wire was used as the counter electrode, and a silver wire was used as a pseudoreference electrode. The redox potentials are based on values measured from differential pulsed voltammetry and are reported relative to a ferrocene/ferrocenium (Cp2Fe/Cp2Fe+) redox couple used as an internal reference, while electrochemical reversibility was determined using cyclic voltammetry. Photophysical Characterization. The UV−visible spectra were recorded on a Hewlett-Packard 8453 diode array spectrometer. Photoluminescent emission measurements were performed using a Photon Technology International QuantaMaster model C-60 fluorimeter. Emission lifetimes at room temperature were measured by time-correlated single-photon counting using an IBH Fluorocube instrument equipped with an LED (λ = 405 nm) excitation source. Quantum yield measurements were carried out using a Hamamatsu C9920 system equipped with a xenon lamp, calibrated integrating sphere, and model C10027 photonic multichannel analyzer (PMA). Temperature-dependent measurements in the range of 5−320 K were performed using a JANIS ST-100 Standard Optical Cryostat instrument

equipped with an intelligent temperature controller. Emission lifetimes were measured using the Photon Technology International QuantaMaster model C-60 fluorimeter in the range of 5−150 K and the IBH Fluorocube instrument in the range of 150−320 K. All samples in fluid solution were deaerated by extensive sparging with N2. Density Functional Theory (DFT) Calculations. Ground-state geometries of all complexes reported were optimized at the B3LYP/ LACVP** level. Time-dependent DFT (TD-DFT) calculations were performed on the optimized structures at the CAM-B3LYP/LACVP** level. These DFT and TD-DFT calculations were performed using QChem 5.1.45 TD-DFT calculations with solvent effects were performed using the IEF-PCM solvation model in the nonequilibrium limit using the ptSS (perturbative state-specific) approach as implemented in QChem 5.1. OLED Fabrication and Characterization. Glass substrates with 1 mm wide indium tin oxide (ITO) strips were cleaned by sequential sonication in tergitol, deionized water, acetone, and isopropanol, followed by 15 min UV ozone exposure. Organic materials and metals were deposited in a vacuum thermal evaporator with a base pressure of 10−7 Torr at rates of 0.2−1 Å/s. A 2 mm2 device area was defined by deposition through shadow masks to define 1 mm wide cathodes consisting of 100 nm Al that were oriented perpendicular to the ITO strips. The device structure was: glass substrate/70 nm ITO/5 nm hexaazatriphenylene hexacarbonitrile (HATCN)/40 nm 4,4′-cyclohexylidenebis [N,N-bis(4-methylphenyl)benzenamine] (TAPC)/10 nm N,N′-dicarbazolyl-3,5-benzene (mCP)/EML/45 nm 2,2′,2″(l,3,5-benzenetriyl)-tris(L-phenyl-l-H-benzimidazole) (TPBi)/1.5 nm 8-hydroxyquinolinato lithium (LiQ)/100 nm Al. Here, the EML is 25 nm of compound 3, either neat or doped into 3,3′-di(9H-carbazol-9yl)-1,1′-biphenyl (mCBP) at 10 or 40 vol %. A HP4156A semiconductor parameter analyzer was used to source voltage while measuring the photocurrent using a calibrated large area Thorlabs FDS1010-CAL photodiode that collected all light exiting the bottom of the flat glass substrate. A fiber-coupled OceanOptics USB4000-VIS-NIR spectrometer was used to measure the output spectra.



RESULTS AND DISCUSSION Synthesis and Characterization. The pyrimidinium salt precursor to MAC* was synthesized by condensation of N,N′bis(2,6-diisopropylphenyl) formamidine and 3-chloropivaloyl chloride in the presence of excess triethylamine at 0 °C, followed by intramolecular cyclization in refluxing toluene for 16 h. The overall yield of MAC*Cl was ca. 70%. The DAC*Cl salt was synthesized following the literature procedure.42 The (carbene)CuCl precursors were synthesized from the free MAC* and DAC* carbenes generated in situ, followed by addition of CuCl. MAC*CuCl was isolated as a white solid, whereas the DAC* D

DOI: 10.1021/jacs.8b12397 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society analogue is a deep red solid. The red color for the DAC* complex is likely due to the formation of a dimeric species [(DAC*)Cu(μCl)2Cu(DAC*)], as reported for the mesitylsubstituted DAC analogue.46 The (carbene)CuCz complexes were synthesized from reaction of the (carbene)CuCl with sodium carbazolide and isolated in 68−85% yield. Complexes 1 and 2 are white solids, whereas complexes 3−6 are yellow, orange-yellow, orange, and purple, respectively. Complexes 1−5 are air- and moisture-stable and can be sublimed under vacuum, whereas complex 6 reacts with ambient moisture to generate free CzH and a yellow solid, presumably DAC*CuOH. Single crystals of 1−4 suitable for X-ray diffraction studies were grown by slow evaporation of CH2Cl2/pentane solutions. The complexes show a near linear geometry at the copper center (CNHC−Cu−NCz = 173−180°) (Figure S1). Bond lengths to the ligands (Cu−CNHC = 1.868(2)−1.902(2) Å, Cu−NCz = 1.852(1)−1.855(2) Å) are comparable to the values reported for mononuclear Cu−CNHC and Cu−Cz complexes.25−27,32,47 Dihedral angles between ligand planes are in the range between 1−16° and lead to a near parallel orientation of 2p orbitals on CNHC and NCz. This geometric arrangement, in light of the close distance between these 2p orbitals on the ligated atoms (CNHC··· NCz = 3.719(2)−3.762(2) Å), suggests that a long-range π interaction could be present across the metal center.48,49 The electrochemical properties of the complexes and their precursors (carbene−CuCl) were examined by cyclic voltammetry and differential pulse voltammetry in acetonitrile solution (Table 1). The redox potentials are referenced to an internal ferrocene (Fc+/Fc) couple, and converted to HOMO and LUMO energies.50 All of the complexes display quasi-reversible reduction and irreversible oxidation waves. The reduction potentials are similar in complexes with the same carbene ligand and are largely unchanged from the precursors MAC*CuCl and DAC*CuCl. Reduction potentials for the DAC* analogues are anodically shifted relative to the MAC* analogues by ca. 0.90 V due to the second carbonyl group in DAC*. The oxidation potentials increase from complexes 3 to 1 (and 6 to 4) consistent with the donor strength of the Cz ligands decreasing with addition of the nitrile groups. Oxidation potentials of chloride precursors (Eox = 0.91 and 1.03 V for MAC*CuCl and DAC*CuCl) are markedly larger than the Cu−Cz complexes. The redox behavior suggests that the reduction potential is determined by the identity of the carbene ligand, whereas the oxidation potential is controlled by carbazolyl ligand. Theoretical calculations support the assignment of the electrochemical results. The contours for the LUMO and HOMO obtained from the Density Functional Theory (DFT) calculations of the (carbene)Cu(Cz) complexes (for example, see Figure 1) show the LUMO is primarily localized on the carbene ligand, whereas the HOMO is predominantly on the Cz ligand. Both LUMO and HOMO have little, but not insignificant, metal character. DFT and TD-DFT calculations were carried out at the CAM-B3LYP/LACVP** level to predict the lowest energy excited states for these complexes.51 The CAM-B3LYP functional was found to provide an accurate description of both the charge transfer (CT) and the ligandlocalized (Cz) states (vide infra). One state corresponds to an intramolecular charge transfer (ICT) involving the carbazolyl (donor) and the carbene (acceptor). In compounds 1 and 2, the ICT state lies higher in energy than the 3Cz state, whereas in 3− 6 the ICT state is lowest in energy. The singlet and triplet (1ICT and 3ICT) levels for these configurations are predicted to be

Figure 1. Frontier orbitals. LUMO (left) and HOMO (right) of complexes 3 (top) and 6 (bottom).

close in energy (