Enhancement of the Luminescent Efficiency in Carbene-Au(I)-Aryl Complexes by the Restriction of Renner-Teller Distortion and Bond Rotation

Article abstract of DOI:10.1021/jacs.9b13755

A series of (carbene)Au^(I)(aryl) complexes are reported. The nature of the lowest excited state in these complexes changes character from metal-to-ligand charge transfer (MLCT) to interligand charge transfer (ICT) with increasing electron-donating strength of the aryl ligand. Complexes that have the MLCT lowest excited state undergo a Renner-Teller bending distortion upon excitation. Such a distortion leads to a large rate of nonradiative decay, on the order of 10⁸ s^-1. Renner-Teller-based nonradiative decay does not occur in chromophores with an ICT emissive state. Introducing a julolidine moiety and ortho-methyl substituents to the aryl group makes the molecule rigid and hinders the rotation along the Au-C_aryl-coordinate bond. Consequently, the nonradiative decay rates of these ICT emitters are decreased and become lower than the radiative decay rate constants (k_r = 10⁵ s^-1). Thus, high-luminescent efficiencies (φ_PL = 0.61 and 0.77) along with short lifetimes (τ < 2 μs) are obtained for yellow and green emitters, respectively. Thermally assisted delayed fluorescence behavior is observed, owing to the small exchange energy (?"E_ST < 1600 cm^-1) in these emitters.

Full text of DOI:10.1021/jacs.9b13755

Subscriber access provided by WESTERN SYDNEY U
Article
Enhancement of the Luminescent Efficiency in Carbene-Au(I)-Aryl
Complexes by the Restriction of Renner–Teller Distortion and Bond Rotation
Tian-yi Li, Daniel Sylvinson Muthiah Ravinson, Ralf Haiges, Peter I. Djurovich, and Mark E. Thompson
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b13755 • Publication Date (Web): 02 Mar 2020
Downloaded from pubs.acs.org on March 2, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
Enhancement of the Luminescent Efficiency in Carbene-Au(I)-Aryl Complexes
by the Restriction of Renner–Teller Distortion and Bond Rotation
7
8
9
Tian-yi Li, Daniel Sylvinson Muthiah Ravinson, Ralf Haiges, Peter I. Djurovich, Mark E. Thompson*
Department of Chemistry, University of Southern California, Los Angeles, CA 90089, United States
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Abstract
A series of (carbene)Au(I)(aryl) complexes are reported. The nature of the lowest excited state in these
complexes changes character from metal-to-ligand charge transfer (MLCT) to interligand charge transfer
(ICT) with increasing electron donating strength of the aryl ligand. Complexes that have an MLCT lowest
excited state undergo a Renner–Teller bending distortion upon excitation. Such a distortion leads to a large
rate for nonradiative decay, on the order of 108 s-1. Renner–Teller based nonradiative decay does not occur
in chromophores with an ICT emissive state. Introducing a julolidine moiety and ortho-methyl substituents
to the aryl group makes the molecule rigid and hinders the rotation along Au−Caryl coordinate bond.
Consequently, nonradiative decay rates of these ICT emitters are decreased and become lower than the
radiative decay rate constants (kr = 105 s-1). Thus, high luminescence efficiencies (PL = 0.61 and 0.77)
along with short lifetimes ( < 2 μs) are obtained for yellow and green emitters, respectively. Thermally
assisted delayed fluorescence behavior is observed owing to the small exchange energy (EST < 1600 cm-1)
in these emitters.
Introduction
Linear, two coordinate, neutral, d10 coinage metal (Cu, Ag and Au) complexes with a general formula
(carbene)M(I)(amide) have emerged as promising luminescent materials.1 These complexes have high
photoluminescence quantum yields (PL) and short luminescent decay lifetimes (). By judicious choice of
the carbene ligand, metal atom and amide, highly efficient luminescence (PL ~ 1.0) can be obtained with
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 38
1
2
3
4
5
6
7
8
9
fast radiative decay ( < 1 μs), values comparable to properties of phosphorescent complexes using noble
metals (Ir and Pt).1a-c The origin of their distinguished photophysical properties is the emissive
amide-to-carbene interligand charge transfer (ICT) state, which emits via thermally assisted delayed
fluorescence (TADF). Since the highest occupied molecular orbital (HOMO) localized on the π electron
donating amide and the lowest unoccupied molecular orbital (LUMO) on carbene acceptor are well
separated (Ccarbene···Namid distance is ~4 Å), the energy gap between the S1 and T1 states (EST) is small.
Therefore, in conjunction with rapid intersystem crossing (ISC) induced by the heavy metal atom, fast rates
of T1S1 ISC up to 108-109 s-1 are possible.1a TADF emitters without metal ions show bimodal emission
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
decay, with both prompt ( < 100 ns) and delayed emission ( > 10 s) signatures, where the relative
intensity of the former is largely determined by the slow ISC rate and the latter by size of EST.2 In contrast,
the (carbene)M(amide) complexes present mono-exponential emissive decays on the 0.5–2 s timescale
(neglecting the 20–200 ps component associated with ISC).3
Many recent investigations of luminescent Au complexes are focused on square-planar d8 Au(III) with
bidentate4 or pincer tridentate ligands.5 Several reports have also recently appeared that describe Au(III)
complexes which display luminescent properties associated with TADF emission.6 However, due to the
unfilled d orbital in Au(III) complexes, these complexes can undergo nonradiative decay through low-lying
d-d states. Such a nonradiative decay route can be avoided in two-coordinate Au(I) complexes with a filled
d10 configuration. Several reports of emissive two-coordinate Au(aryl) complexes have also appeared.7
Previous studies on these derivatives have described materials that luminesce from metal-metal states
induced by aurophilic interactions or from intraligand π-π* states;8 however, there do not appear to be any
reports of mononuclear, two-coordinate Au complexes where emission occurs from metal-to-ligand charge
transfer (MLCT) states.9 Here we investigate changing the donor ligands in the previous reported
(carbene)Au(amide) complexes to aryl ligands, establishing a new series of luminescent two-coordinate
Au(I) chromophores with a general structure of (carbene)Au(aryl).7b, 10
2
ACS Paragon Plus Environment

Page 3 of 38
Journal of the American Chemical Society
1
2
3
4
5
Most previously reported d10 coinage metal
Jahn-Teller distortion
flattening
emitters have three- or four-coordinate
geometry.11 Upon photoexcitation to the MLCT
state, the metal center is transiently oxidized to a
d9 configuration. Consequently, these compounds
suffer from Jahn−Teller distortion as shown in
Figure 1. This process involves breaking the
degeneracy of the orbitals due to asymmetric
occupancy of the d-orbitals in the d9 configuration.
To minimize the overall molecular energy, four
coordinate species with tetrahedral ground state
6
7
8
9
L
L
L
h
L
L
M(II)
d9
M(I)
d10
oxidation of M
L
L
L
L
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
T-shape
L
L
L
L
h
M(II)
d9
M(I)
d10
L
oxidation of M
Renner-Teller distortion
bending
L
h
L M(II)
d9
M(I)
d10
L
L
oxidation of M
Figure 1. Schematically illustrations of molecular
distortion mechanisms for d10 coinage metal
complexes upon excitation to a MLCT excited
state.
undergo
a
flattening
distortion
after
photoexcitation,12 whereas the trigonal planar, three coordinate complexes undergo T-shaped structural
reorganization in the excited state.13 In such cases, the nonradiative decay rate is sufficiently fast to severely
limit luminescence in fluid environments, owing to the enhanced Franck−Condon factors between vibronic
states of the excited and ground electronic state.14 Similarly, when the MLCT state is the lowest excited
state, linear d10 coinage complexes can also nonradiatively decay through a geometric distortion, but in this
case, the deformation involves bending,9d, 15 referred to as the Renner–Teller effect (Figure 1c).16 The
driving force for such distortion comes from stabilization of the excited state caused by orbital mixing in
the new geometry as a result of high-order vibrational and electronic coupling in the open-shell system.17
To preclude significant nonradiative energy loss induced by Renner–Teller distortion in two coordinate
complexes, the lowest excited state (the emissive state) must avoid a change in valence of the metal ion,
i.e. not be an MLCT state.15 This energetic condition was met in our previous (carbene)M(I)(amide)
compounds by virtue of using electrophilic diamidocarbene (DAC) and monoamido-aminocarbene (MAC)
ligands as acceptors and amides as a strong electron donors.1a, 1c Similarly, by employing the same DAC
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 38
1
2
3
4
5
6
7
8
9
and MAC ligands, and by adding electron donating substituents to the energetically low-lying π orbital of
the phenyl ligand to destabilize the energy level of the HOMO, a (carbene)Au(aryl) complex should have
a lowest emissive state with ICT character. To validate this hypothesis, a series of (DAC)Au(aryl) and
(MAC)Au(aryl) complexes were synthesized where substituents with increasing electron donor strength on
the aryl ligand were used to destabilize the energy level of the aryl π orbital, thus, shifting the HOMO
energy away from the Au center and onto the aryl ligand. On the basis of theoretical calculations and
photophysical studies, we conclude that Renner–Teller distortions occur in complexes with MLCT excited
states, but not in compounds that have an ICT lowest excited state. Such a distortion leads to a significant
increase in nonradiative decay rate, which can enhance nonradiative decay and make the chromophore
poorly or non-emissive. We also investigate the role of aryl- and amide-group rotation in nonradiative decay
of these complexes. We find that restricting rotation in addition to eliminating Renner−Teller distortion
leads to a marked increase in the luminescence efficiency (PL increasing from < 10-3 to 0.77), owing to a
decrease in nonradiative decay (from >108 to ca. 5×104 s-1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Synthesis and characterization. All the boronic acid except 4-dimethylamino-2,6-dimethyl boronic acid
are
purchased
from
commercial
sources
and
used
without
further
purification.
4-dimethylamino-2,6-dimethyl boronic acid and 3,5-dimethyl-4-julolidine boronic acid pinacol ester were
synthesized using modified methods according to the previous reports, synthetic details are provided in the
electronic supporting information (ESI).18 The carbene ligands were synthesized according to the reported
methods.1c All the reactions are carried out using standard Schleck techniques, and the purification of the
products are carried out under air. All the solvents were used as received from commercial sources except
1
where individually mentioned. H NMR and 13C NMR spectra were recorded on a Varian Mercury 400
instrument. Elemental analyses were performed using a Thermo Scientific FlashSmart CHNS elemental
analyzer.
4
ACS Paragon Plus Environment

Page 5 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
General synthesis of (carbene)Au(aryl) from boronic acid (Method A): In an oven-dried glass dram vial,
the (carbene)AuCl precursor (50 mg) and KOH (1.5 equiv) were stirred in a mixed solvent of benzene and
toluene (2 ml, 1:1 v/v) for 2 h at room temperature while covered with aluminum foil. Then, the
corresponding boronic acid (1.5 equiv) was added in one portion and the mixture was kept stirring at room
temperature. Upon complete conversion of the precursor (monitored by 1H NMR), excess CH2Cl2 (20 ml)
was added and the suspension was filtered through a pad of Celite, providing a clear filtrate. After removal
of solvents under vacuum, a small amount of CH2Cl2 was added to dissolve the raw solid and pentane was
added to precipitate the desired product as a microcrystalline powder.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General synthesis of (carbene)Au(aryl) from boronic acid pinacol ester (Method B): In an oven-dried glass
dram vial, the (carbene)AuCl precursor (50 mg), the corresponding boronic acid ester (1.2 equiv) and KOH
(1.5 equiv) were suspended in a mixed solvent of benzene and toluene (2 ml, 1:1 v/v). The mixture was
stirred at room temperature for 12 h. Upon complete conversion of the precursor (monitored by 1H NMR),
excess CH2Cl2 (20 ml) was added and the suspension was filtered through a pad of Celite, providing a clear
filtrate. After removal of solvents under vacuum, a small amount of CH2Cl2 was added to dissolve the raw
solid and pentane was added to precipitate the desired products as a microcrystalline powder.
1
(DAC)Au(I)(phenyl) (1a): Method A, dark yellow microcrystalline powder, yield 44 mg (83%). H NMR
(400 MHz, acetone) δ 7.54 (t, J = 7.8 Hz, 2H, p-ArH(dipp)), 7.39 (d, J = 7.8 Hz, 4H, m-ArH(dipp)), 6.82
– 6.73 (m, 2H, m-ArH), 6.65 (ddt, J = 8.0, 6.8, 1.5 Hz, 1H, p-ArH), 6.53 – 6.48 (m, 2H, o-ArH), 3.04 (sept,
J = 6.7 Hz, 4H, CH(CH3)2), 1.87 (s, 6H, C(CH3)2), 1.39 (d, J = 6.8 Hz, 12H, CH(CH3)2), 1.19 (d, J = 6.8 Hz,
12H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 227.57 (NCN), 172.65 (C=O), 166.18 (ipso-Au-Ar),
145.54 (ipso-N-Ar(dipp)), 140.25 (o-Ar), 134.87 (o-Ar(dipp)), 130.09 (p-Ar(dipp)), 125.96 (p-Ar), 124.44
(m-Ar), 124.30 (m-Ar(dipp)), 51.43 (C(CH3)2), 28.78 (CH(CH3)2), 24.38 (C(CH3)2), 23.63 (CH(CH3)2),
23.13 (CH(CH3)2). Elemental analysis calculated for C36H45AuN2O2: C, 58.85; H, 6.17; N, 3.81. found: C,
58.46; H, 6.57; N, 4.09.
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 38
1
2
3
4
5
6
7
8
9
(DAC)Au(I)(4-carbazolylphenyl) (1b): Method A, bright yellow microcrystalline powder, yield 51 mg
(78%). 1H NMR (400 MHz, acetone) δ 8.14 (dt, J = 7.6, 1.0 Hz, 2H, CH4(Cz)), 7.58 (t, J = 7.8 Hz, 2H, p-
ArH(dipp)), 7.45 (d, J = 7.7 Hz, 4H, m-ArH(dipp)), 7.34 (ddd, J = 8.3, 7.1, 1.2 Hz, 2H, CH2(Cz)), 7.24 (dt,
J = 8.3, 0.8 Hz, 2H, CH1(Cz)), 7.19 (ddd, J = 8.0, 7.1, 1.0 Hz, 2H, CH3(Cz)), 7.07 – 7.03 (m, 2H, m-ArH),
6.86 – 6.79 (m, 2H, o-ArH), 3.10 (sept, J = 6.7 Hz, 4H, CH(CH3)2), 1.91 (s, 6H, C(CH3)2), 1.47 (d, J =
6.8 Hz, 12H, CH(CH3)2), 1.22 (d, J = 6.8 Hz, 12H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 226.97
(NCN), 172.65 (C=O), 166.26 (ipso-Au-Ar), 145.63 (ipso-N-Ar(dipp)), 141.03 (ipso-N-Cz), 141.00 (o-Ar),
134.91 (o-Ar(dipp)), 133.99 (m-Ar), 130.24 (p-Ar(dipp)), 125.65 (p-Ar), 124.42 (m-Ar(dipp)), 124.18
(ipso-C-Cz), 122.90 (CH2(Cz)), 119.95 (CH4(Cz)), 119.35 (CH3(Cz)), 109.76 (CH1(Cz)), 51.52 (C(CH3)2),
28.84 (CH(CH3)2), 24.42 (C(CH3)2), 23.66 (CH(CH3)2), 23.25 (CH(CH3)2). Elemental analysis calculated
for C48H52AuN3O2: C, 64.06; H, 5.82; N, 4.67. found: C, 64.32; H, 6.04; N, 4.42.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(DAC)Au(I)(4-diphenylaminophenyl) (1c): Method A, bright orange microcrystalline powder, yield 55 mg
(85%). 1H NMR (400 MHz, acetone) δ 7.53 (t, J = 7.8 Hz, 2H, p-ArH(dipp)), 7.40 (d, J = 7.7 Hz, 4H, m-
ArH(dipp)), 7.21 – 7.11 (m, 4H, N(ArH)2), 6.93 – 6.82 (m, 6H, N(ArH)2), 6.62 – 6.55 (m, 2H, m-ArH),
6.53 – 6.45 (m, 2H, o-ArH), 3.05 (sept, J = 6.7 Hz, 4H, CH(CH3)2), 1.88 (s, 6H, C(CH3)2), 1.41 (d, J =
6.8 Hz, 12H, CH(CH3)2), 1.19 (d, J = 6.8 Hz, 12H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 227.28
(NCN), 172.65 (C=O), 161.91 (ipso-Au-Ar), 148.33 (p-Ar), 145.54 (ipso-N-Ar(dipp)), 144.37 (ipso-N-Ar),
140.73 (o-Ar), 134.89 (o-Ar(dipp)), 130.12 (p-Ar(dipp)), 128.88 (p-N(Ar)2), 124.32 (m-Ar(dipp)), 123.25
(m-Ar), 123.01 (o-N(Ar)2), 121.63 (m-N(Ar)2), 51.39 (C(CH3)2), 28.80 (CH(CH3)2), 24.39 (C(CH3)2), 23.63
(CH(CH3)2), 23.20 (CH(CH3)2). Elemental analysis calculated for C48H54AuN3O2: C, 63.92; H, 6.03; N,
4.66. found: C, 63.52; H, 6.16; N, 4.66.
1
(MAC)Au(I)(phenyl) (2a): Method A, white microcrystalline powder, yield 43 mg (81%). H NMR
(400 MHz, acetone) δ 7.50 – 7.38 (m, 2H, p-ArH(dipp)), 7.38 – 7.33 (m, 2H, m-ArH(dipp)), 7.32 – 7.26
(m, 2H, m-ArH(dipp)), 6.77 – 6.70 (m, 2H, m-ArH), 6.63 – 6.60 (m, 1H, p-ArH), 6.59 – 6.55 (m, 2H, o-
ArH), 4.08 (s, 2H, CCH2N), 3.41 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.16 (sept, J = 6.8 Hz, 2H, CH(CH3)2),
6
ACS Paragon Plus Environment

Page 7 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1.60 (s, 6H, C(CH3)2), 1.44 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.41 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.36 (d, J
= 6.9 Hz, 6H, CH(CH3)2), 1.17 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 221.42
(NCN), 172.23 (C=O), 167.65 (ipso-Au-Ar), 145.81 (ipso-N-Ar), 144.57 (ipso-N-Ar), 140.31 (o-Ar),
140.17 (o-Ar(dipp)), 136.06 (o-Ar(dipp)), 129.52 (p-Ar(dipp)), 129.11 (p-Ar(dipp)), 125.82 (p-Ar), 124.75
(m-Ar(dipp)), 123.69 (m-Ar(dipp)), 123.67 (m-Ar), 61.19 (CCH2N), 37.91 (C(CH3)2), 28.69 (CH(CH3)2),
23.97 (CH(CH3)2), 23.92 (CH(CH3)2), 23.75 (C(CH3)2), 23.70 (CH(CH3)2), 23.13 (CH(CH3)2). Elemental
analysis calculated for C36H47AuN2O: C, 59.99; H, 6.57; N, 3.89. found: C, 59.83; H, 6.69; N, 3.79.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(MAC)Au(I)(4-carbazolylphenyl) (2b): Method A, white microcrystalline powder, yield 54 mg (83%).
1H NMR (400 MHz, acetone) δ 8.13 (dt, J = 7.8, 1.9 Hz 2H, CH4(Cz)), 7.54 – 7.43 (m, 2H, p-ArH(dipp)),
7.43 – 7.38 (m, 2H, CH2(Cz)), 7.37 – 7.29 (m, 4H, m-ArH(dipp)), 7.22 (dt, J = 8.3, 0.9 Hz, 2H, CH1(Cz)),
7.18 (ddd, J = 8.0, 7.1, 1.0 Hz, 2H, CH3(Cz)), 7.03 – 6.97 (m, 2H, m-ArH), 6.92 – 6.85 (m, 2H, o-ArH), 4.13
(s, 2H, CCH2N), 3.47 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.22 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 1.63 (s,
6H, C(CH3)2), 1.52 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.48 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.39 (d, J = 6.9 Hz,
6H, CH(CH3)2), 1.21 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 220.69 (NCN),
172.25 (C=O), 167.91 (ipso-Au-Ar), 145.90 (ipso-N-Ar), 144.64 (ipso-N-Ar), 141.04 (ipso-N-Cz), 141.05
(o-Ar), 140.18 (o-Ar(dipp)), 136.09 (o-Ar(dipp)), 133.28 (m-Ar), 129.68 (p-Ar(dipp)), 129.27 (p-Ar(dipp)),
125.61 (p-Ar), 124.86 (m-Ar(dipp)), 124.08 (ipso-C-Cz), 123.82 (m-Ar(dipp)), 122.84 (CH2(Cz)), 119.92
(CH4(Cz)), 119.26 (CH3(Cz)), 109.79 (CH1(Cz)), 61.24 (CCH2N), 37.96 (C(CH3)2), 28.42 (CH(CH3)2),
24.10 (CH(CH3)2), 23.95 (C(CH3)2), 23.77 (CH(CH3)2), 23.74 (CH(CH3)2), 23.26 (CH(CH3)2). Elemental
analysis calculated for C48H54AuN3O: C, 65.07; H, 6.14; N, 4.74. found: C, 64.67; H, 6.30; N, 4.68.
(MAC)Au(I)(4-diphenylaminophenyl) (2c): Method A, beige microcrystalline powder, yield 55 mg (85%).
1H NMR (400 MHz, acetone) δ 7.49 – 7.38 (m, 2H, p-ArH(dipp)), 7.36 (d, J = 7.4 Hz, 2H, m-ArH(dipp)),
7.29 (d, J = 7.6 Hz, 2H, m-ArH(dipp)), 7.18 – 7.11 (m, 4H, N(ArH)2), 6.90 – 6.82 (m, 6H, N(ArH)2), 6.59
– 6.54 (br, 4H, ArH), 4.09 (s, 2H, CCH2N), 3.42 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 3.17 (sept, J = 7.0 Hz,
2H, CH(CH3)2), 1.61 (s, 6H, C(CH3)2), 1.46 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.42 (d, J = 6.9 Hz, 6H,
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 38
1
2
3
4
5
6
7
8
9
CH(CH3)2), 1.37 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.18 (d, J = 6.8 Hz, 6H, CH(CH3)2). 13C NMR (101 MHz,
acetone) δ 221.16 (NCN), 172.25 (C=O), 163.66 (ipso-Au-Ar), 148.41 (p-Ar), 145.82 (ipso-N-Ar), 144.58
(ipso-N-Ar(dipp)), 143.57 (ipso-N-Ar(dipp)), 140.77 (o-Ar), 140.18 (o-Ar(dipp)), 136.07 (o-Ar(dipp)),
129.56 (p-Ar(dipp)), 129.15 (p-Ar(dipp)), 128.81 (p- N(Ar)2), 124.79 (m-Ar(dipp)), 123.73 (m-Ar(dipp)),
123.61(m-Ar), 122.76 (o-N(Ar)2), 121.36 (m-N(Ar)2), 61.21 (CCH2N), 37.91 (C(CH3)2), 28.70 (CH(CH3)2),
28.37 (CH(CH3)2), 24.05 (CH(CH3)2), 23.91 (C(CH3)2), 23.75 (CH(CH3)2), 23.70 (CH(CH3)2), 23.21
(CH(CH3)2). Elemental analysis calculated for C48H56AuN3O: C, 64.93; H, 6.36; N, 4.73. found: C, 64.91;
H, 6.40; N, 4.84.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(MAC)Au(I)(4-dimethylaminophenyl) (2d): Method A, beige microcrystalline powder, yield 45 mg (80%).
1H NMR (400 MHz, acetone) δ 7.49 – 7.38 (m, 2H, p-ArH(dipp)), 7.35 (d, J = 7.6 Hz, 2H, m-ArH(dipp)),
7.28 (d, J = 7.6 Hz, 2H, m-ArH(dipp)), 6.47 – 6.41 (m, 2H, m-ArH), 6.32 – 6.26 (m, 2H, o-ArH), 4.04 (s,
2H, CCH2N), 3.40 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 3.16 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 2.66 (s, 6H,
N(CH3)2), 1.60 (s, 6H, C(CH3)2), 1.44 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.40 (d, J = 6.9 Hz, 6H, CH(CH3)2),
1.36 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.17 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ
222.06 (NCN), 172.29 (C=O), 155.16 (ipso-Au-Ar), 148.31 (p-Ar), 145.79 (ipso-N-Ar(dipp)), 144.58
(ipso-N-Ar(dipp)), 140.42 (o-Ar), 140.26 (o-Ar(dipp)), 136.12 (o-Ar(dipp)), 129.42 (p-Ar(dipp)), 129.01
(p-Ar(dipp)), 124.70 (m-Ar(dipp)), 123.63 (m-Ar(dipp)), 112.20 (m-Ar), 61.16 (CCH2N), 40.28 (N(CH3)2),
37.90 (C(CH3)2), 28.68 (CH(CH3)2), 23.97 (CH(CH3)2), 23.92 (C(CH3)2), 23.75 (CH(CH3)2), 23.70
(CH(CH3)2), 23.13 (CH(CH3)2). Elemental analysis calculated for C38H52AuN3O: C, 59.75; H, 6.86; N, 5.50.
found: C, 59.47; H, 6.92; N, 5.53.
(MAC)Au(I)(julolidine) (2e): Method A, yellow microcrystalline powder, yield 51 mg (55%). 1H NMR (400
MHz, acetone) δ 7.47 – 7.37 (m, 2H, p-ArH(dipp)), 7.35 – 7.31 (m, 2H, m-ArH(dipp)), 7.27 (d, J = 7.7 Hz,
2H, m-ArH(dipp)), 5.97 (s, 2H, o-ArH), 4.01 (s, 2H, CCH2N), 3.39 (sept, J = 6.9 Hz, 2H, CH(CH3)2), 3.15
(sept, J = 6.9 Hz, 2H, CH(CH3)2), 2.89 – 2.83 (m, 4H, ArCH2CH2), 2.34 (t, J = 6.6 Hz, 4H, NCH2CH2),
1.79 – 1.71 (m, 4H, CH2CH2CH2), 1.58 (s, 6H, C(CH3)2), 1.44 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.40 (d, J =
8
ACS Paragon Plus Environment

Page 9 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
6.9 Hz, 6H, CH(CH3)2), 1.35 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.17 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR
(101 MHz, acetone) δ 222.36 (NCN), 172.29 (C=O), 154.39 (ipso-Au-Ar), 145.73 (ipso-N-Ar(dipp)),
144.52 (ipso-N-Ar(dipp)), 140.29 (p-Ar) , 140.24 (o-Ar(dipp)), 138.93 (o-Ar), 136.13 (o-Ar(dipp)), 129.36
(p-Ar(dipp)), 128.94 (p-Ar(dipp)), 124.68 (m-Ar(dipp)), 123.61 (m-Ar(dipp)), 119.07 (m-Ar), 61.22
(CCH2N), 50.18 (NCH2CH2), 37.86 (C(CH3)2), 28.67(CH(CH3)2), 27.46 (CH2CH2CH2), 24.02 (CH(CH3)2),
23.90 (C(CH3)2), 23.76 (CH(CH3)2), 23.68 (CH(CH3)2), 23.16 (CH(CH3)2), 22.72 (CH2CH2CH2). Elemental
analysis calculated for C42H56AuN3O: C, 61.83; H, 6.92; N, 5.15. found: C, 61.48; H, 6.89; N, 5.15.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(MAC)Au(I)(4-dimethylamino-2,5-dimethylphenyl) (2d-Me): Method A, light yellow microcrystalline
powder, yield 38 mg (65%). 1H NMR (400 MHz, acetone) δ 7.47 – 7.38 (m, 2H, p-ArH(dipp)), 7.38 – 7.34
(m, 2H, m-ArH(dipp)) 7.31 – 7.27 (m, 2H, m-ArH(dipp)), 6.08 (s, 2H, m-ArH), 4.02 (s, 2H, CCH2N), 3.42
(sept, J = 6.8 Hz, 2H, CH(CH3)2), 3.17 (sept, J = 6.8, 2H, CH(CH3)2), 2.64 (s, 6H, N(CH3)2), 1.59 (s, 6H,
C(CH3)2), 1.42 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.39 – 1.33 (m, 18H, CH(CH3)2 and o-ArCH3), 1.17 (d, J =
6.9 Hz, 6H, CH(CH3)2). 13C NMR (101 MHz, acetone) δ 223.71 (NCN), 172.47 (C=O), 156.91 (ipso-Au-
Ar), 148.80 (p-Ar), 147.28 (o-Ar), 145.78 (ipso-N-Ar(dipp)), 144.62 (ipso-N-Ar(dipp)), 140.83 (o-
Ar(dipp)), 136.53 (o-Ar(dipp)), 129.44 (p-Ar(dipp)), 129.13 (p-Ar(dipp)), 124.85 (m-Ar(dipp)), 123.81 (m-
Ar(dipp)), 110.17 (m-Ar), 61.54 (CCH2N), 40.28 (N(CH3)2), 37.77 (C(CH3)2), 28.63 (CH(CH3)2), 28.40
(CH(CH3)2), 25.73 (ArCH3), 23.89 (CH(CH3)2), 23.85 (C(CH3)2), 23.80 (CH(CH3)2), 23.54 (CH(CH3)2),
23.17 (CH(CH3)2). Elemental analysis calculated for C40H56AuN3O: C, 60.67; H, 7.13; N, 5.31. found: C,
60.67; H, 7.43; N, 5.17.
(MAC)Au(I)(2,5-dimethyljulolidine) (2e-Me): Method B, yellow microcrystalline powder, yield 35 mg
1
(56%). H NMR (400 MHz, acetone) δ 7.47 – 7.37 (m, 2H, p-ArH(dipp)), 7.36 – 7.33 (m, 2H, m-
ArH(dipp)), 7.30 – 7.26 (m, 2H, m-ArH(dipp)), 4.00 (s, 2H, CCH2N), 3.42 (sept, J = 6.8 Hz, 2H,
CH(CH3)2), 3.17 (sept, J = 6.8 Hz, 2H, CH(CH3)2), 2.80 – 2.74 (m, 4H, ArCH2CH2), 2.24 (t, J = 6.8 Hz,
4H, NCH2CH2), 1.79 (m, 4H, CH2CH2CH2), 1.59 (s, 6H, C(CH3)2), 1.41 (d, J = 6.8 Hz, 6H, CH(CH3)2),
1.36 (d, J = 6.9 Hz, 12H, CH(CH3)2), 1.24 (s, 6H, o-ArCH3), 1.18 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 38
1
2
3
4
5
6
7
8
9
(101 MHz, acetone) δ 223.76 (NCN), 172.50 (C=O), 158.73 (ipso-Au-Ar), 145.75 (ipso-N-Ar(dipp)),
144.58 (ipso-N-Ar(dipp)), 142.49 (p-Ar), 141.38 (o-Ar), 140.96 (o-Ar(dipp)), 136.63 (o-Ar(dipp)), 129.44
(p-Ar(dipp)), 129.13 (p-Ar(dipp)), 124.90 (m-Ar(dipp)), 123.86 (m-Ar(dipp)), 116.23 (m-Ar), 61.51
(CCH2N), 50.25 (NCH2CH2), 37.77 (C(CH3)2), 29.67 (CH2CH2CH2), 28.62 (CH(CH3)2), 28.40 (CH(CH3)2),
26.01 (ArCH3), 23.89 (CH(CH3)2), 23.79 (C(CH3)2), 23.56 (CH(CH3)2), 23.21 (CH(CH3)2), 23.15
(CH(CH3)2), 22.99 (CH2CH2CH2). Elemental analysis calculated for C44H60AuN3O: C, 62.62; H, 7.17; N,
4.98. found: C, 62.64; H, 7.47; N, 4.83.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
X-ray Crystallography. The X-ray intensity data were collected on a Bruker APEX DUO 3-circle platform
diffractometer with the χ-axis fixed at 50.74°, and using Mo K radiation (λ = 0.71073 Å) from a fine-focus
tube monochromatized by a TRIUMPH curved-crystal monochromator. The diffractometer was equipped
with an APEX II CCD detector and an Oxford Cryosystems Cryostream 700 apparatus for low-temperature
data collection adjusted to 100(2) K. The crystal was mounted in a Cryo-Loop using Paratone oil. A
complete hemisphere of data was scanned on omega (0.5°) at a detector distance of 50 mm and a resolution
of 512 x 512 pixels. The frames were integrated using the SAINT algorithm to give the hkl files corrected
for Lp/decay. Data were corrected for absorption effects using the multi-scan method (SADABS). The
structures were solved by intrinsic phasing and refined with the Bruker SHELXTL Software Package.
Supplementary crystallographic data can be downloaded from Cambridge Crystallographic Data Centre
(CCDC) with the registration number: 1948368 (2a), 1948367 (2b), 1948366 (1a), 1948365 (1b), 1948370
(1c), 1948369 (2d), 1948371 (2d-Me) and 1948372 (2e-Me).
Theoretical Calculations. All Density Functional Theory (DFT) calculations reported in this work were
performed using the Q-Chem 5.1 program.19 Ground state geometry optimization of the complexes were
performed at the B3LYP/LACVP* level. Time-dependent density functional theory (TD-DFT) calculations
were performed on the ground state optimized geometries at the CAM-B3LYP/LACVP* level which has
been found to offer good agreement with experimental results for the class of materials under study striking
a balanced description of both charge-transfer and locally excited (LE) states.1a-c Relaxed potential energy
10
ACS Paragon Plus Environment

Page 11 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
surface (PES) scans were performed at the B3LYP/LACVP** level. The PES scans were performed for the
complexes in 15o increments of the dihedral angle between the C-N bond of the carbene and the C-C bond
of the Aryl ligand that includes the coordinating C atom and the adjacent C atom. Single-point calculations
were performed on the geometries optimized for each dihedral angle constraint using the B3LYP functional
with Grimme’s D3 dispersion correction20 and the LACVP** ECP/basis set to account for dispersion effects
which are vital to reliably predict the potential energy landscape.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electrochemical Measurements. Cyclic voltammetry (CV) and differential pulsed voltammetry (DPV)
were conducted using a VersaSTAT 3 Potentiostat Galvanostat with a standard three-electrode system. A
glassy carbon rod was used as the working electrode, a platinum wire as the counter electrode and a silver
wire as a pseudo-reference electrode. Anhydrous acetonitrile (CH3CN, DriSolv from VWR) was used as
solvent and 0.1 M tetra-n-butyl ammonium hexafluorophosphate (TBAF from Acros) was used as the
supporting electrolyte. Decamethylferrocene was used as internal reference. The redox potentials were
obtained based on values measured from DPV and are reported relative to a ferrocene/ferrocenium (Fc+/Fc)
redox couple.
Photophysical Characterization. UV-visible absorption spectra were recorded using a Hewlett-Packard
8453 diode array spectrometer. Steady-state photoluminescent emission spectra were measured on a Photon
Technology International QuantaMaster model C-60 fluorimeter. Transient photoluminescent lifetimes
were measured on an IBH Fluorocube instrument using time-correlated single-photon counting method
(TCSPC) for measurements at ambient conditions and multichannel scaling method (MSC) for
measurements at low temperature. Photoluminescent quantum yields were determined using a Hamamatsu
C9920 system equipped with a xenon lamp, calibrated integrating sphere and model C10027 photonic
multichannel analyzer (PMA). Temperature-dependent lifetime measurements from 77 to 310 K were
measured using IBH Fluorocube instrument in an OptistatDN Oxford cryostat. All fluid samples for
photoluminescent measurements were deaerated by bubbling N2 in a glass cuvette fitted with a Teflon
stopcock. Doped polymer films (1 wt%) were prepared in toluene solution of polystyrene (PS). Solutions
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 38
1
2
3
4
5
6
were dropcast onto a quartz substrate and the films were air-dried for 3 h before drying under vacuum
overnight. The polymer films were measured under a stream of N2 during the photophysical measurements.
7
8
9
Results and Discussion
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis and Characterization
A series of two-coordinate (carbene)Au(aryl) complexes with DAC (1a−c) and MAC (2a−e, 2d-Me and
2e-Me) as an electron acceptor ligands were synthesized as shown in Figure 2. The aryl moieties are phenyl
(1a and 2a), 4-carbazolyl phenyl (1b and 2b), 4-diphenylamino-phenyl (1c and 2c),
4-dimethylamino-phenyl (2d), julolidine (2e), 2,6-dimethyl-4 dimethylamino-phenyl (2d-Me) and 3,5-
dimethyl-julolidine (2e-Me). All of the complexes were isolated in good yields (56−85%) via a modified
one-pot reaction by reaction of the (carbene)AuCl precursor with aryl boronic acid or pinacol ester in the
presence of KOH, as shown in Figure 2.21 The conversion of the Au precursor complex was monitored by
1H NMR and the final desired products were purified by recrystallization. When boronic acid is a reactant,
100% conversion of the precursor can be realized within 3 h. However, the reaction time needs to be
extended to 12 h if pinacol-borane is used instead, due to the relatively lower reactivity of the boronic acid
ester. Less boronic acid ester (1.2 equiv) is required than boronic acid (1.5 equiv.) for 100% conversion of
OH
OH
O
O
1.5 equiv. KOH
or
(carbene)Au(aryl)
aryl
B
aryl B
(carbene)AuCl
+
toluene/benzene
room temperature
O
iPr
X
R
R
R
R
iPr
N
N
aryl =
iPr
iPr
Au
N
N
N
N
aryl
X = CO (DAC)
X = CH2 (MAC)
1a
2a
1b
2b
1c
2c
2d
2e
(R=H)
(R=H)
2d-Me
(R=Me)
2e-Me
(R=Me)
Figure 2. Preparation of two-coordinate (carbene)Au(aryl) complexes and the molecular structures of
complexes 1a–c, 2a–e, 2d-Me and 2e-Me.
12
ACS Paragon Plus Environment

Page 13 of 38
Journal of the American Chemical Society
1
2
3
4
the (carbene)AuCl precursor. The complexes
5
6
7
8
were isolated as beige to orange solids and are
stable to air in solution and solid state.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Selected metrical parameters from single
crystal structures of 1a−c, 2a−b and 2d are given
in Figure 3. All of the molecules have a linear
coordination geometry at the Au center
(Ccarbene−Au−Caryl = 175−178°). Bond lengths
between Au and Ccarbene (2.01−2.06 Å) are similar
to values between Au and Caryl (2.02−2.05 Å).
The dihedral angle between the carbene ligand
and the coordinated phenylene is smaller in 1a
(17.6°), 2a (13.3°), 1b (2.7°) and 2b (4.1°) than
in 1c (52.0°) and 2d (62.5°). Space-filling models
of 1a–c in Figure 3 illustrate that the coordinated
phenyl moieties are not sterically encumbered by
the bulky dipp (diisopropyl phenyl) moieties, and
thus a small energy barrier is expected for the
rotation around the Au−Caryl bond in these
derivatives. Steric interactions among aryl
groups lead to larger dihedral angles between the
coordinated aryl and the amines in 1b, 2b and 1c
(48.8°, 52.6° and 34.2°, respectively) than the
dimethylamine in 2d (10.3°).
Figure 3. (top) Simplified geometries of 1a−1c, 2a,
2b and 2d with significant crystallographic distances
(in Å) and angles, C: grey; N: blue, Au: gold;
(bottom) space-filling diagrams of 1a, the isopropyl
moieties are highlighted in purple.
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 38
1
2
3
4
5
6
7
8
9
Electrochemical properties of complexes 1a–c and 2a–d were investigated by cyclic voltammetry and
differential pulse voltammetry (DPV) in acetonitrile. Redox potentials determined using
decamethylferrocene as internal reference are reported here along with those for the (DAC)AuCl and
(MAC)AuCl precursors relative to ferrocene (Table 1). The complexes display irreversible oxidation and
quasi-reversible reduction peaks (see ESI). The oxidation potentials of 1a, 1b and 1c decrease from 0.93 to
0.72 to 0.28 V, respectively, and are similar to values found for the corresponding MAC derivatives (2a–
c). The complexes with phenyl ligands (1a and 2a) have oxidation potentials within 0.04 V of the
(carbene)AuCl analogs, indicating that the oxidation is metal centered. The small cathodic shift in oxidation
potential of the phenyl-carbazole derivatives (1b and 2b, Eox = 0.2 V) suggests that the HOMO in these
complexes is comprised of contributions from both the metal and aryl ligand. The oxidation potentials for
the remaining complexes (1c, 2c and 2d) are shifted cathodically from the (carbene)AuCl analogs by
0.6−0.9 V, indicating that the HOMO in these complexes is primarily based on the arylamine ligand and
have little metal character. The reduction potentials for 1a–c fall between -1.53 V and -1.71 V, close to that
of (DAC)AuCl, and similarly, potentials for 2a and 2c are close to that of (MAC)AuCl (Ered = -2.42 V).
The redox properties indicate that the oxidation is gradually tuned from metal centered to aryl-amine ligand
based, depending on the amine, whereas the reduction potential is dictated mainly by the carbene ligand.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14
ACS Paragon Plus Environment

Page 15 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table 1. Electrochemical data of (carbene)AuCl precursors and (carbene)Au(aryl) complexes.a
Complex
Eox (V)a
0.97
Ered (V)a
−1.56
−1.53
−1.71
−1.55
−2.44
Eox-Ered (V)
2.53
(DAC)AuCl
1a
0.93
2.46
1b
0.72
2.43
1c
0.28
1.83
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(MAC)AuCl
0.95
3.39
2a
2b
2c
2d
2d-Me
2e
2e-Me
0.92
0.73
0.26
−0.03
−0.10
−0.20
−0.28
−2.42
3.34
b
b
−2.41
−2.45
b
2.67
2.42
b
b
b
b
b
a Values obtained from DPV measurements carried out using decamethylferrocene as an internal reference,
with redox potentials adjusted to 0 V versus ferrocene.22 The redox potential measured for
decamethylferrocene relative to ferrocene can be found in the ESI. b No reduction peak is observed.
Turning the Renner–Teller Effect On and Off
The Renner–Teller effect will bring about a bending distortion in a linear two coordinate complex that have
a lowest excited state with MLCT character (Au dz2-to-ligand) such as a (carbene)Au(aryl) complex. In
order to probe this distortion in a basic system we considered a representative analog of the compounds
investigated here, containing a simplified MAC ligand with the dipp moieties replaced by methyl groups.
Two triplet optimized structures were identified using unrestricted DFT (UDFT) calculations at the
B3LYP/LACVP* level, one remains linear with elongated Au–C bonds, whereas the other is bent (Ccarbene
–
Au–Caryl = 145º). Here we find major differences between the SOMO and SOMO-1 orbitals of the two
structures (SOMO = highest occupied singly occupied molecular orbital) as shown in Figure 4(a). The
geometric change from linear to bent is the result of - mixing in the excited state. The SOMO of the
linear triplet is largely localized on a -orbital that has carbene-Au antibonding character, whereas the
SOMO-1 is a combination of Au dz2 and Au–ligand -orbitals. Bending the Ccarbene–Au–Caryl angle stabilizes
the SOMO sufficiently to overcome a smaller destabilization of the SOMO-1 in the T1 state and leads to a
structure that is 0.35 eV lower in total energy than the linear one. The excited state of the bent molecule
can then undergo effective nonradiative deactivation via vibronic coupling with the ground state. In
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 38
1
2
3
4
5
6
7
8
9
contrast, no equivalent stabilization in the excited state is generated by bending in complexes that have an
ICT excited state. The carbene and arene  orbitals involved in the ICT transition are spatially well-
separated and there is only a minor contribution of metal dxz orbitals in the HOMO. Such an excited
molecule with ICT character can thus decay radiatively back to the ground state with a high efficiency as
deactivation by a vibronic coupling pathway is suppressed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To probe the Renner–Teller effect in these
complexes, we consider 1a–c and 2a–d.
According to the electrochemical measurements
in this series of complexes, electron donating
substituents on the aryl ring can be used to alter
the composition of the HOMO from one that is
Au d orbital in nature to one dominated by the
aryl  system. Thus, the position of the lowest
excited state can be inverted from MLCT to ICT
state. It is noteworthy that the higher lying
intraligand π-π* states never become the lowest
excited state in these complexes.
Time-
dependent density functional theory (TD-DFT)
calculations were carried out using the optimized
ground state geometries to clarify the nature of
the singlet and triplet excited states (Table 2).
Representative natural transition orbitals (NTOs)
for the S1 state of 1a and 1c are shown in Figure
4(b). These isosurfaces represent the electronic
structure of the molecules immediately after
Figure 4. (a) Schematic diagram of the frontier
molecular orbitals of simplified
a
MAC:Au(C6H5) model complex illustrating the
Renner–Teller effect. (b) NTOs for the lowest
singlet excited states of 1a and 1c (green: hole,
yellow: electron), with both molecules in the ground
state geometry (note that 1a will ultimately distort
to the bent structure in the excited state) (c)
Qualitative energy level diagram for ICT and
MLCT based T1 potential energy surfaces (GS =
ground state geometry), based on TD-DFT
calculations.
16
ACS Paragon Plus Environment

Page 17 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
excitation, before either ISC or a Renner-Teller distortion can take place. The NTOs of 1a show substantial
MLCT character whereas those of 1c are characteristic of an ICT state. In addition to investigating the S1
and T1 states in the optimized ground state geometry, we have optimized the geometry of the triplet state
for each molecule (Table 2). The T1 states predicted for complexes 2b and 2c in the ground state geometry
are localized on the aryl ligands, but become MLCT and ICT based T1 states, respectively, in the optimized
triplet geometry. The optimized T1 geometries in 1a-b and 2a-b present a significant bending distortion at
the Au center (the Ccarbene–Au–Caryl angle is as small as 137° in 2a), which is attributed to the Renner–Teller
effect. In contrast, this angle close to 180º in the complexes that have ICT based T1 states, i.e. 1c, 2c, 2d-2e.
Stabilization of the geometry optimized triplet is much greater for the (carbene)Au(aryl) complexes with
MLCT T1 states (1.6-2.1 eV) than for the ICT based triplet states (0.5-0.8 eV). The increased stabilization
of the bent structure for the MLCT triplets leads to enhanced vibronic coupling between the excited and
ground states, resulting in the markedly higher nonradiative decay rates observed for MLCT based triplet
states. A qualitative energy level diagram with the ground state geometry S1/T1, as well as ICT and MLCT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. The lowest singlet and triplet excited state properties determined from TD-DFT calculations.
Based on optimized S0 geometry
S1 T1
Based on optimized T1 geometry
T1
Complex
Energy
(eV / ƒ)
Energy
(eV)
3.14
3.04
3.06
2.66
3.43
3.40
3.06
3.06
3.03
2.96
2.89
2.83
C-Au-X
(º) a
Energy
(eV)
2.62
1.11
1.36
2.12
2.80
1.33
1.43
2.55
2.26
2.19
2.34
2.12

(eV)d
0.52
1.93
1.70
0.54
0.63
2.07
1.63
0.51
0.77
0.77
0.55
0.71
State
State
State
(DAC)AuCl 3.46 / 0.003 MLCT
MLCT
MLCT
MLCT
ICT
MLCT
MLCT
LC(Cz)b
LC(TPA)c
ICT
168
138
141
179
162
137
140
179
167
177
179
179
MLCT
MLCT
MLCT
ICT
MLCT
MLCT
MLCT
ICT
ICT
ICT
ICT
ICT
1a
1b
1c
3.15 / 0.003 MLCT
3.28 / 0.004 MLCT
2.93 / 0.248
ICT
(MAC)AuCl 4.12 / 0.005 MLCT
2a
2b
2c
2d
2d-Me
2e
3.92 / 0.006 MLCT
4.00 / 0.007 MLCT
3.72 / 0.346
3.47 / 0.316
3.36 / 0.299
3.36 / 0.306
3.25 / 0.283
ICT
ICT
ICT
ICT
ICT
ICT
ICT
ICT
2e-Me
a The bond angle is for Ccarbene–Au–Cl or Ccarbene-Au-Caryl b Cz = carbazolyl. c This state is predominantly localized
.
on the triphenyl amine (TPA) with a small amount of ICT character. d The difference in vertical transition energy
between triplet states in the optimized S0 and T1 structures.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 38
1
2
3
4
T1 surfaces, is shown Figure 4(c). The figure illustrates how bending in the MLCT state brought about by
5
6
the Renner-Teller effect promotes the nonradiative T1S0 transition.
7
8
9
Absorption spectra of 1a–c and 2a–d in THF solution are given in Figure 5a. The intense absorption
bands in high energy region (below 280 nm) correspond to the ligand-based π-π* transitions for all the
chromophores. The broad featureless band (ε ~3000 L/(mol×cm)) peaking at 350 nm in 1a and 290 nm in
2a is assigned to 1ICT absorption from the phenyl to the carbene ligand. Weak bands (ε ~500 L/(mol×cm))
at lower energy (390–440 nm in 1a and 330–360 nm in 2a) are assigned to the 1MLCT absorption, which
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
agree well with the MLCT bands of the corresponding (carbene)AuCl precursors (see Figure S9 in the
ESI). Structured bands between 295–350 nm in 1b and 2b are assigned to the π-π* transitions localized on
the carbazolyl moiety. The lowest-lying 1MLCT bands are overlapped by the close-lying and more intense
1ICT band peaking at 390 nm in 1b and 330 nm in 2b. For 1c and 2c, intense bands at 300 nm are assigned
to π-π* transitions on the TPA moiety and the lowest lying 1ICT bands peak at 460 nm in 1c and at 360 nm
in 2c. The ICT band of 2d is slightly redshifted in comparison to that of 2c, peaking at 365 nm. This shift
is attributed to the stronger electron-donating strength of the dimethyl amino group and the resultant
1
destabilized HOMO energy level. The distinct ICT bands in these complexes present characteristic
bathochromic shifts in solvents with decreasing polarity (from CH3CN to THF to cyclohexane) as shown
in Figure 5b. In contrast, the ligand-based absorption bands at 340 nm in 1b and 300 nm in 1c do not shift
in these solvents. The pronounced negative solvatochromism of the 1ICT bands is rationalized to the effects
of opposing molecular dipole moments in the ground and excited states, which result in an increase in
excitation energy in polar solvents.1a, 1c
Luminescence at room temperature from 1a–b and 2a–b is broad and featureless (see ESI). The
emission is weak (PL < 0.001) with short luminescence decay lifetimes ( < 10 ns) in both fluid solution
and polystyrene (PS) film. Emission from 2b at room temperature (max = 389 nm in PS) is significantly
higher in energy than that from 1a–b and 2a (max > 450 nm in PS). The emission energy in PS at 77 K
remains relatively constant for 1a–b and 2a and the luminescence lifetimes are several tens of microseconds,
18
ACS Paragon Plus Environment

Page 19 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
(a) 10000
(b)
1.2
1a
2a
1a
Cyclohexane
THF
0.8
5000
7
0.4
MeCN
8
9
0
0.0
1.6
10000
1b
1c
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1b
1.2
2b
5000
0.8
0.4
0.0
0
10000
6
1c
2c
2d
x4
4
5000
0
2
0
300
350
400
450
500
550
600
250
300
350
400
450
500
550
Wavelength (nm)
Wavelength (nm)
Figure 5. (a) UV-visible absorption spectra of complexes 1a–c and 2a–d in THF at RT; (b) bathochromic
shift of the ICT absorption bands with decreasing solvent polarity in complexes 1a–c.
indicative of phosphorescence. In contrast, emission from 2b at 77 K is red shifted (max = 410 nm) and
highly structured, with a decay lifetime of seconds. The luminescence data for 1a–b and 2a in PS are
consistent with assignment to MLCT transitions, whereas emission from 2b is ascribed to a transition
localized on the PhCz ligand. We propose that the poor luminescence efficiency of 1a–b and 2a is a result
of rapid nonradiative decay caused by a bending distortion of the Ccarbene−Au−Caryl bond in the excited state
(Renner-Teller effect). This deactivation pathway is so effective that the process is not suppressed by the
relatively rigid polymer matrix. It is inferred that the large MAC/DAC ligands provide sufficient steric
bulk to allow the bending distortion of the smaller aryl ligands within the free volume of the polymer matrix,
resulting in rapid decay of the excited state (knr ~108 s-1) in both polymer film and fluid solution.
In contrast, complexes 1c, 2c and 2d present bright orange, blue and green emission, respectively,
featuring broad unstructured bands in both fluid solution and PS film at room temperature. Emission spectra
in 2-methyltetrahydrofuran (MeTHF) and methylcyclohexane (MeCy) are shown in Figure 6 and the
corresponding photophysical properties are summarized in Table 2. At room temperature, the broad ICT
emission bands undergo hypsochromic shifts with decreasing solvent polarity, as well as in frozen solvents
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 38
1
2
3
4
5
6
7
8
9
at 77 K.
The solvatochromism and
1.0
0.5
MeTHF
rigidochromism is caused by the dipolar solvent
organizing around the large dipole of the S0
ground state, destabilizing the opposing dipole
of the emissive ICT state (see ESI). Thus,
hypsochromic shifts at 77 K are greater in a
polar solvent than a non-polar solvent. Complex
2c displays a structured emission band at 77 K,
presumably owing to mixing between the ICT
state and ligand-centered (3LC) state of the TPA
moiety, whereas spectra from 1c and 2d at 77 K
remain featureless. All three complexes present
broad emission spectra at both room temperature
and low temperature when doped into a PS film
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.0
1.0
MeCy
0.5
0.0
1.0
PS film
1c
2c
2d
0.5
Solid: RT
Dash: 77K
0.0
400 450 500 550 600 650 700 750 800 850
Wavelength (nm)
Figure 6. Emission spectra of complex 1c, 2c and 2d
at room temperature and 77 K in dilute fluid solution
and polystyrene (PS) film.
(Figure 6). Large increases in luminescence lifetimes (from s to tens of s) upon cooling to 77 K are
indicative of emission from a thermally populated 1ICT state at room temperature and 3ICT state at 77 K.
Since the excited state has ICT character, a Renner−Teller distortion is avoided upon photoexcitation as the
metal ion remains effectively in a d10 electronic configuration in derivatives with electron-rich aryl ligands.
Consequently, the magnitude of knr is limited to values of 106−107 s-1 in fluid solution and decreases further
to 105 s-1 in PS film. This change suggests that the bending distortion is greatly diminished in these ICT
emitters.
20
ACS Paragon Plus Environment

Page 21 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
Table 3. Luminescent properties of complexes 1c and 2c–g.
Room temperature
77 K
Complex
λem (nm)
kr (105 s-1) knr (105 s-1) λem (nm)
PL  (s)
 (μs)
MeTHF
5.0
7
8
9
1c
2c
2d
2d-Me
2e
2e-Me
680
522
583
594
604
620
0.002 0.004
0.12 0.210
0.035 0.083
0.055 0.310
0.013 0.015
0.021 0.128
2500
42
120
31
680
76
542
432, 456
474
83
66
46
63
60
87
5.8
4.2
1.8
8.9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
496
496
522
1.6
MeCy
5.4
1c
2c
2d
2d-Me
2e
2e-Me
586
460
492
520
532
558
0.15 0.270
0.004 0.016
0.009 0.052
32
640
190
1.7
3.7
3.0
557
436, 456
470
86
50
39
75
53
94
2.6
1.9
1.3
2.6
0.42
0.42
0.35
3.3
1.6
2.2
508
506
544
1.6
1 wt% doped PS film
1c
2c
2d
2d-Me
2e
2e-Me
620
462
504
520
544
554
0.38 0.650
5.8
0.84
2.2
1.9
2.4
2.0
9.6
3.5
5.5
0.55
1.6
1.3
612
460
502
526
540
560
54
41
45
56
65
72
0.19
0.28
0.77
0.61
0.61
2.3
1.3
4.1
2.5
3.0
The Role of Ligand Rotation in Nonradiative Decay
Even though Renner−Teller distortion is suppressed in the ICT emitters (1c and 2c−d), the luminescence
efficiencies remain low in fluid solution and modest even in PS film (PL < 0.40). Other effective
nonradiative decay routes must still be present, with a likely candidate involving rotation around the
coordinating Au−Caryl and C−N single bonds of the arylamine moiety.23 The single crystal structures of
these complexes show that there are no steric restrictions to rotation of the coordinated phenylene, which
may act to increase losses from nonradiative decay. Consistent with this proposal, values for knr in 1c and
2c−d increase as much as two orders of magnitude when measured in fluid solution (MeCy) as opposed to
the rigid media of PS (Table 3). Thus, to test this rotational deactivation hypothesis, we prepared complexes
2d, 2d-Me, 2e and 2e-Me. These complexes have a common MAC carbene acceptor ligand but three
different aryl donor ligands. In complex 2d-Me methyl substituents at the ortho-positions of the coordinated
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 38
1
2
3
4
phenyl ring serve to hinder rotation around the
5
6
7
8
Au−Caryl bond, whereas a julolidine moiety is used to
prevent rotation around the C−N bond in 2e. Finally,
ortho-methyl substituents are also introduced to a
julolidine moiety in complex 2e-Me, inhibiting
rotation around both the Au−Caryl and C−N bond.
(Figure 7)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The effect of the ortho methyl groups on the
orientation of the coordinated aryl ligand is evident
in the single crystal structures of 2d, 2d-Me and 2e-
Figure 7. Molecular structure and design
strategy for restricted rotation.
Me shown in Figure 8, top. The dihedral angle between the carbene and the aryl ligand in complex 2d is
~62.5°, comparable to that found in 1c, which suggests a low energy barrier for rotation around the Au−Caryl
bond. In contrast, this dihedral angle is much smaller in complexes 2d-Me (5.4°) as well as in 2e-Me (5.8°).
The distance between the ortho-methyl carbon to the center of the adjacent phenyl rings in the dipp moieties
is between 3.57−3.65 Å in 2d-Me and 3.57–3.82 Å in 2e-Me. This separation places these substituents well
within the Van der Waals radii of the methyl group (2.0 Å).24 As revealed in space-filling diagrams
(Figure 8, bottom), confinement of the two ortho-methyl groups inside the steric shield of the isopropyl
groups of the dipp substituents enforces a coplanar geometry between the ligands. This orientation is also
1
evidenced by the chemical shifts of the ortho-methyl signals in the H NMR spectra. Resonances for the
ortho-methyl groups in the free ligands N,N-3,5-tetramethyl aniline ( = 2.21 ppm) and 3,5-dimethyl
julolidine ( = 2.03 ppm) shift upfield in 2e-Me ( = 1.35 ppm) and 2e-Me ( = 1.24 ppm). These shifts
suggest that the methyl substituents experience a shielding influence from being pointed toward the center
of the phenyl rings of dipp moieties, thus indicating a molecular configuration in fluid solution close to that
found in single crystals. To further evaluate whether the rotation around the Au−Caryl bond is hindered by
the ortho-methyl substituents, DFT calculations were carried out to scan the molecular potential energy at
22
ACS Paragon Plus Environment

Page 23 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. (top) Single crystal thermal ellipsoid diagrams of complex 2d, 2d-Me and 2e-Me, all the non-
H atoms are plotted at 50% probability ellipsoid (C: black; N: blue; O: red; Au: gold). H atoms are
omitted for clarity; (bottom) space-filling diagrams, the isopropyl moieties are highlighted in purple and
the ortho-methyl groups in orange.
different ligand conformations (Figure 9). With
the arrangement of the carbene and aryl ligands
varied from coplanar (dihedral angle = 0°) to
perpendicular (dihedral angle = 90°), energetic
barriers for unsubstituted 2d and 2e peak at
around 2 kcal/mol, whereas the barriers are
increased by 4.0 kcal/mol in 2d-Me and 2e-Me.
Thus, it is inferred that rotation around the
Au−Caryl bond can occur at rates competitive
6
5
4
3
2
1
0
2d
2d-Me
2e
2e-Me
0
15
30
45
60
75
90
Dihedral angle (deg)
Figure 9. Theoretically calculated molecular energy
of 2d/2d-Me and 2e/2e-Me with dihedral angle
between carbene and aryl moieties varied from 0 to
90°.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 38
1
2
3
4
with radiative decay in 2d and 2e, whereas the ligands in 2d-Me and 2e-Me maintain a coplanar molecular
5
6
geometry within the microsecond lifetimes of their respective excited states.
7
8
9
The oxidation potentials for 2e and 2e-Me are cathodically shifted compared with 2d and 2d-Me owing
to the stronger electron donating strength of the julolidine moiety and ortho-methyl groups (Table 1). The
electrochemical reduction of complexes 2d-Me and 2e/2e-Me are shifted outside of the observable
electrochemical window. These derivatives are expected to be ICT emitters due to strong electron donation
from the amine-modified aryl ligands. Results
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
25.0k
from TD-DFT calculations carried out for
complexes 2d/2d-Me and 2e/2e-Me are
summarized in Table 2. Their optimized
geometries in the lowest T1 state remain linear,
indicating that Renner−Teller distortion is
avoided (see ESI). Absorption spectra of
2d/2d-Me and 2e/2e-Me recorded in dilute
cyclohexane solution are shown in Figure 10a.
Intense bands at high energy ( < 325 nm) are
assigned to π-π* transitions on the ligands,
whereas broad bands at longer wavelengths are
attributed to the spin-allowed ICT absorptions.
A plot of the oxidation potential versus the
onset energy of these low energy bands follows
a linear relationship (see inset in Figure 10a),
which supports the ICT character of these
transitions where the acceptor is the carbene
ligand. Notably, the ICT bands of 2d-Me and
2d
2d-Me
2e
0.0
-0.1
-0.2
-0.3
2d-Me
2e
20.0k
15.0k
10.0k
5.0k
2d
2e-Me
2e-Me
2.8
2.9
3.0
3.1
3.2
ECT (eV)
0.0
250
300
350
400
450
500
550
Wavelength (nm)
(b)
at 77K
1.0
0.5
0.0
1.0
at RT
2d
2d-Me
2e
0.5
0.0
2e-Me
450
500
550
600
650
700
750
Wavelength (nm)
Figure 10. (a) Absorption spectra of complexes
2d/2d-Me and 2e/2e-Me in cyclohexane solution at
room temperature. Inset: correlation of the low-energy
band energy (ECT, determined by the peak maxima)
and the oxidation potential (Vox obtained from DPV
measurement); (b) Emission spectra of complexes
2d/2d-Me and 2e/2e-Me at RT and at 77 K in 1 wt%
doped PS film.
24
ACS Paragon Plus Environment

Page 25 of 38
Journal of the American Chemical Society
1
2
3
4
2e-Me have extinction coefficients ( >5500 L/(mol×cm)) almost double those of 2d and 2e (
5
6
~3000 L/(mol×cm)). A coplanar configuration enhances overlap between the ligand  orbitals in 2d-Me
7
8
9
and 2e-Me and is reflected by the extinction coefficients of their ICT transition being higher than 2d and
2e.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The luminescent properties of complexes 2d-Me and 2e/2e-Me were recorded in both solution and PS
film (Figure 10b and ESI) and summarized in Table 3. The complexes display broad, featureless emission
bands in both MeCy solution and the polymer matrix consistent with an ICT character for the emissive
excited state. Consequently, luminescence from 2d-Me and 2e/2e-Me displays solvatochromic and
rigidochromic behavior similar to that observed for 2c−d in solution with different polarities and
temperature (see SI). An important difference in 2d-Me and 2e/2e-Me is a dramatic increase of nearly two
orders of magnitude in luminescent efficiency in fluid solution (PL  0.40) compared to their unconstrained
derivatives (2c−d). This increase does not come about from a change in the radiative rate constants as
values for 2d-Me and 2e/2e-Me in MeCy solution (kr = 1.3−2.6×105 s-1) are similar to those of 2c−d,
consistent with a Strickler−Berg analysis of their respective absorption spectra (see the ESI). Instead, the
high efficiency of 2d-Me and 2e/2e-Me in MeCy is due to a corresponding decrease in the rate of
nonradiative decay. Since nonradiative decay attributed to Renner−Teller distortion is absent in 2d-Me and
2e/2e-Me as ICT emitters, the decrease in knr is ascribed to restricted rotation of the amine in 2e and the
aryl ligand in 2d-Me/2e-Me around the C−N and Au−Caryl single bonds. A further increase in quantum
efficiency, up to 0.77 for 2d-Me, is achieved upon dispersing in PS film and is caused by an additional
decrease in knr in the rigid media.
Informed with the photophysical data above, a semilog plot of knr versus emission energy in PS for the
complexes with ICT emissive states is shown in Figure 11. Compounds that share common promoting
vibrational modes which couple the excited and ground states will have values for knr that increase
exponentially with decreasing emission energy (Energy Gap Law).25 Deviations from this behavior are
indicative of a change in the vibrational/rotational mechanism responsible for nonradiative decay. Taking
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 38
1
2
3
4
5
6
7
8
9
as a reference the known analog having a carbazole donor [(MAC)Au(Cz)], the fastest knr is found for the
complex with lowest emission energy (1c). The high knr for this complex is likely due to effects from
vibronic coupling in addition to bond rotation. Notably, complexes 2c and 2d, despite having higher
emission energies, also have higher knr values than (MAC)Au(Cz), which emphasizes the effects of excited
state deactivation via bond rotation of the aryl ligand. A significant decrease in knr is found in complex
2d-Me with hindered rotation along Au−Caryl bond. Finally, complexes 2e and 2e-Me have knr values
comparable to (MAC)Au(Cz) even though these aryl derivatives have lower emission energies. Such results
indicate that molecular design to restrict rotation of the aryl ligand can effectively decrease energy loss
from nonradiative decay in these two coordinate (carbene)Au(aryl) ICT emitters.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1c
106
2d
2c
2e
105
2e-Me
2d-Me
104
2.0
2.2
2.4
2.6
2.8
E0-0 (eV)
Figure 11. Semilog plot of knr versus the emission energy for complexes 1c, 2c−e, 2d-Me and 2e-Me (using
data in PS film). The value reported for (MAC)Au(carbazole) is marked with a black star.1a
Temperature dependence of TADF
Luminescent decay lifetimes of all of the (carbene)Au(aryl) ICT emitters in PS film vary within
0.65−4.1 s at room temperature yet increase to 42−72 s at 77 K. This large change with temperature is
consistent thermal activation from triplet to singlet states that is the hallmark of TADF behavior. Thus,
transient emission decays were recorded between 77−310 K to further characterize these photophysical
processes. A representative example of the emission lifetime plotted as a function of temperature is shown
for 2e-Me in Figure 12; similar plots for 1c, 2c, 2d/2d-Me and 2e are shown in the ESI. Emission from all
of these complexes at low temperature is governed by a long-lived phosphorescence from 3ICT (T1S0, >
26
ACS Paragon Plus Environment

Page 27 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
99%). With increasing temperature, endothermic ISC from T1 to S1 is promoted and the radiative decay
from S1 becomes dominant (> 90%), resulting in a drastically decreased lifetime (see ESI for the ratio of
phosphorescence vs. TADF). Since luminescence from these emitters decays monoexponentially in the
measured temperature range, it is inferred that thermal equilibrium is established well before either
phosphorescence or TADF because of the fast rate for ISC in these types of complexes.3c, 26 Therefore, two
Boltzmann distribution models were employed to fit the temperature-dependent lifetime curves as
illustrated in Figure 12.1a, 1c The two-level model assumes only S1 and T1 states having an energy gap of
EST, whereas the three-level model takes into consideration zero-field-splitting (ZFS) of the triplet
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
sublevels. Both models provide the same trends for the exchange energy EST, but the three-level model
gives a better fit in the low temperature regime. Kinetic and thermodynamic parameters derived from the
photophysical measurements using the three-level model are given in Table 4. Data from these fits show
the smallest EST in 1c (686 cm-1, 85.0 meV) and largest in 2c (1640 cm-1, 203 meV), whereas 2d/2d-Me
and 2e/2e-Me have comparable values ranging between 1120 cm-1 (139 meV) and 1280 cm-1 (159 meV).
The ZFS values derived from the three-level model exceed 130 cm-1 (except for 1c, ZFS = 11 cm-1) and are
comparable to values previously reported two coordinate (carbene)M(carbazole) complexes.1a, 1c Such
impressive ZFS effects explain the poor fits of the two-level model at low temperature. However, it should
be noted that the values given for the ZFS in Table 4 are somewhat unreliable as large fitting errors can be
generated due to the paucity of data collected at temperatures where the effects of ZFS are most pronounced
(<120 K).
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 38
1
2
3
4
5
6
7
8
9
70
60
50
40
30
20
2e-Me
O
iPr
iPr
N
N
iPr
iPr
Au
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Two-level fitting
Three-level fitting
10
0
100
150
200
250
300
Temperature (K)
Figure 12. Two-level and three-level models for the description of emission in (carbene)Au(aryl)
complexes (the fitting equations are given in the previous report1a) and the temperature-dependent lifetime
of 2e-Me in PS film with fitting curves.
Table 4. Photophysical data derived from three-level Boltzmann fits to temperature-dependent lifetimes
EST (cm-1)
τ (μs)
Ⅰ,Ⅱ
τ (μs)
Ⅲ
Complex
ZFS (cm-1)
τfl (ns)
1c
2c
2d
2d-Me
2e
2e-Me
690
11
14
0.53
2.9
7.0
5.9
4.7
68
44
45
57
71
73
34
5.7
0.30
20
22
3.1
1600
1300
1200
1100
1200
290
600
230
140
420
Conclusion
The work presented here shows (carbene)Au(aryl) complexes are promising chromophores for
luminescent applications. These two coordinate (carbene)Au(aryl) complexes can be prepared using readily
accessible aryl boronic acids or pinacol esters. Their emission color is tuned from sky blue to red by
changing carbene ligand and electron donation strength of the substituent on the aryl ligand. A bending
distortion is expected upon excitation in species with an MLCT excited state, owing to transient oxidation
of the Au center from d10 to d9, referred to as the Renner-Teller effect. However, by increasing the electron
donating strength of the substituent on the phenyl ligand, the energy of the aryl  electrons can be raised
above that of the Au d electrons. Consequently, the excited state alignment can be tuned, and the lowest
excited state character changed from MLCT to ICT. Single crystal X-ray analysis shows that rotation
28
ACS Paragon Plus Environment

Page 29 of 38
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
around the Au−Caryl and amine C−N bond is unencumbered and expected to be facile. The conformational
flexibility of aryl ligands can be restricted by chemical modifications, such as introducing ortho-methyl
groups or the julolidine moiety. Increasing the molecular rigidity by restricting single bond rotation
improves the luminescence efficiency by further decreasing the rate of nonradiative decay.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
These two-coordinate complexes provide new models for investigation of dynamic change of molecular
geometry in the excited state and are complementary to three- and four-coordinated d10 chromophores that
distort in the excited state via the Jahn-Teller effect. It is noteworthy that the luminescent properties of these
two-coordinate complexes with ICT emission are comparable to the recently reported four-coordinate d8
Au TADF derivatives where the metal center is in a formal Au(III) oxidation state.6c The similarity in
photophysical properties supports the assumption that oxidation of the formally Au(I) metal center does not
occur upon excitation in (carbene)Au(aryl) complexes with ICT emissive states. This work highlights the
importance of avoiding a situation where the lowest excited state with MLCT character when aiming for
highly luminescent candidates in metals with d10 electronic configurations since the geometric distortion
opens up efficient nonradiative decay pathways. Enhanced quantum yields are therefore observed in
complexes that have lowest energy excited states with ICT character, where the Au center is free from
transient oxidation, and which keeps the excited molecules in geometries similar to the ground state. Our
(carbene)Au(aryl) complexes thus enrich the library of the luminescent coinage metal complexes, providing
additional opportunities to discover new photodynamic and optoelectronic properties.
Acknowledgment
This work was supported by the Universal Display Corporation.
Supporting Information
Detailed synthetic procedures and characterization; NMR spectra; single crystal X-ray diffraction data; CV
and DPV curves; photophysical data including temperature-dependent lifetime measurements and data
from theoretical calculations.
29
ACS Paragon Plus Environment