Photophysical characteristics and reactivity of bis(2.9-di-tert-butyl-1,10- phenanthroline) copper(l)

Article abstract of DOI:10.1021/ic802361q

The recently synthesized sterically constrained copper(l) complex [Cu(dtbp)₂]⁺ (1), where dtbp is 2,9-di-tert-butyl-1,10- phenanthroline, exhibits unique photophysical and reactivity properties. Complex 1 (λ_abs, 425 nm; ε,

Full text of DOI:10.1021/ic802361q

5704 Inorg. Chem. 2009, 48, 5704–5714
DOI: 10.1021/ic802361q
Photophysical Characteristics and Reactivity of Bis(2,9-di-tert-butyl-1,10-phenanthroline)
copper(I)
Omar Green, Bhavesh A. Gandhi, and Judith N. Burstyn*
Department of Chemistry, University of Wisconsin, Madison, Wisconsin 53706
Received December 10, 2008
The recently synthesized sterically constrained copper(I) complex [Cu(dtbp)2]+ (1), where dtbp is 2,9-di-tert-butyl-1,10-
phenanthroline, exhibits unique photophysical and reactivity properties. Complex 1 (λabs, 425 nm; ε, 3100 L M-1 cm-1;
λemission, 599 nm) has the longest metal-to-ligand charge-transfer (MLCT) emission lifetime (τ, 3260 ns) and largest quantum
yield (φ, 5.6%) of all [Cu(R2phen)2]+ complexes. Complex 1 also exhibits a large positive reduction potential for the
[Cu2+(dtbp)2]|[Cu+(dtbp)2] couple (E1/2 = 0.70 V vs Fc+/0) and a large negative excited-state reduction potential for the
[Cu2+(dtbp)(dtbp-•)]|[Cu2+(dtbp)2] couple (E1/2 = -1.66 V vs Fc+/0), indicating that this complex is a potent photoreductant in
the excited state. The steric constraint imposed by the t-butyl substituents in 1 enables unusual ligand replacement reactivity.
Either CH3CN or CO replaces one of the dtbp ligands, a reaction that is readily followed by loss of the unique emission signature
of 1. Monodentate CH3CN binds to the copper(I) center with an affinity 2 orders of magnitude greater than that of the displaced
dtbp, despite the fact that the displaced ligand is bidentate. CO-induced displacement of dtbp from 1is reversible, but only in the
presence of 1 equiv of unbound dtbp. The exceptionally strong donor ligand CH3NC displaces both dtbp ligands from 1. In
contrast to the facile ligand displacement reactivity with good donor ligands, 1 does not react readily with O2, by either a ligand
displacement or an oxidative pathway. Rather, O2 induces partial quenching of emission via an outer-sphere interaction with 1.
Introduction
geometric and electronic control.5,6,8,9 In this paper we
present the characterization of a homoleptic phenanthroline
complex of copper(I) with photophysical properties compar-
Luminescent transition metal complexes are useful in
sensors, in solid-state lighting devices, and in dye-sensitized
solar energy conversion devices.1-3 In these contexts, dii-
mine-ligated ruthenium(II) complexes are the most com-
monly used compounds. There are major disadvantages to
these metals in practical applications: they are rare, expen-
sive, and toxic. Because of these drawbacks there is interest in
developing luminescent complexes of the abundant, cheap,
and non-toxic metal, copper.4-6 Copper-functionalized dye-
sensitized solar cells, employing diimine-copper(I) complexes
as dyes, are remarkably efficient.7 Phenanthroline-ligated
copper(I) complexes are potentially useful dye candidates,
as their photophysical properties may be tailored through
able to those of tris(bipyridyl)ruthenium(II), [Ru(bpy)3]2+
a complex commonly applied in device design.
,
In phenanthroline-copper(I) complexes, metal-to-ligand
charge transfer excitation formally oxidizes the copper(I)
ion to copper(II), and relaxation may occur through non-
emissive geometric reorganization or by radiative decay.5,6,8,9
For copper complexes to be useful in photochemical devices,
the tendency of copper to undergo geometric reorganization
toward square planar geometry on excitation must be inhibited
to maximize the radiative decay of the excited state. Copper(I)
complexes that meet this requirement include the homoleptic
[Cu(R2Phen)2]+ complexes, where R2Phen is a 2,9-disubsti-
tuted phenanthroline.10 Substitution at the 2 and 9 positions,
those closest to the copper(I) ion, inhibit the geometric
reorganization and thus maximize radiative emission.5,6,9
The first example of room temperature emission from
this class of complexes was reported by Blaskie and
*To whom correspondence should be addressed. E-mail: burstyn@
chem.wisc.edu. Phone: 608-262-0328. Fax: 608-262-6143.
(1) Rogers, C. W.; Wolf, M. O. Coord. Chem. Rev. 2002
233–234, 341-350.
(2) O’Regan, B.; Graetzel, M. Nature (London) 1991, 353, 737–40.
(3) Evans, R. C.; Douglas, P.; Winscom, C. J. Coord. Chem. Rev. 2006,
250, 2093–2126.
(8) Armaroli, N. Chem. Soc. Rev. 2001, 30, 113–124.
(9) Scaltrito, D. V.; Thompson, D. W.; O’Callaghan, J. A.; Meyer, G. J.
Coord. Chem. Rev. 2000, 208, 243–266.
(10) Abbreviations used: dtbp, 2,9-di-tert-butyl-1,10-phenanthroline; dmp,
2,9-dimethyl-1,10-phenanthroline; dpp, 2,9-diphenyl-1,10-phenanthroline; xop,
2-(2-methylphenyl)-9-(2,6-dimethylphenyl)-1,10-phenanthroline; dnpp, 2,9-dineo-
pentyl-1,10-phenanthroline; dipp, 2,9-di-isopropyl-1,10-phenanthroline; dbp,
2,9-butyl-1,10-phenanthroline; dsbp, 2,9-sec-butyl-1,10-phenanthroline; phen,
1,10-phenanthroline.
(4) Barbieri, A.; Accorsi, G.; Armaroli, N. Chem. Commun. 2008
2185–2193.
(5) Lavie-Cambot, A.; Cantuel, M.; Leydet, Y.; Jonusauskas, G.; Bassani,
D. M.; McClenaghan, N. D. Coord. Chem. Rev. 2008, 252, 2572–2584.
(6) Armaroli, N.; Accorsi, G.; Cardinali, F.; Listorti, A. Top. Curr. Chem.
2007, 280, 69–115.
(7) Bessho, T.; Constable, E. C.; Graetzel, M.; Hernandez Redondo, A.;
Housecroft, C. E.; Kylberg, W.; Nazeeruddin, M. K.; Neuberger, M.;
Schaffner, S. Chem. Commun. 2008, 3717–3719.
r
2009 American Chemical Society
pubs.acs.org/IC
Published on Web 06/04/2009

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5705
Table 1. Photophysical Characteristics of [Cu(R2Phen)2]+ Complexes in CH2Cl2 Solution and Comparison with Those of [Ru(bpy)3]+a,b
ελmax, L mol-1
ΕAbs - ΕEm,
eV
complex
λAbs, nm
cm-1
λEm, nm
τ, ns
φem ꢀ 103
kr103 s-1
knr 107 s-1
references
[Ru(bpy)3]2+
[Cu(dtbp)2]+
[Cu(dtbp)(dmp)]+
[Cu(dsbp)2]+
[Cu(dipp)2]+
[Cu(dnpp)2]+
[Cu(dbp)2]+
[Cu(dmp)2]+
[Cu(dpp)2]+
[Cu(bfp)2]+
446-454
425
440
455
445
449
457
457
440
∼14000
3100
7000
6600
7500
5700
7000
7800
3800
10900
3000
580-630
599
646
690
650
715
725
730
690
0.617-0.799
0.847
0.899
0.928
0.879
1.027
1.003
1.015
1.021
0.819
530-1250
3260
730
400
365
260
150
90
243
29-270
56
10
4.5
4.0
1.6
0.9
0.4
59
this work
17.18
13.70
11.25
10.90
6.15
6.00
4.44
3.58
20.00
6.71
0.03
0.14
0.25
0.27
0.38
0.66
1.11
0.41
0.60
0.67
20
12
60
12
12
12
19
42
19
0.87
3.3
1.0
462
452
665
673
165
149
[Cu(xop)2]+
0.901
a The data for [Ru(bpy)3]+ are presented as a range because they were collected in a variety of solvents. b Abbreviations: dtbp, 2,9-di-tert-butyl-
1,10-phenanthroline; dmp, 2,9-dimethyl-1,10-phenanthroline; dsbp, 2,9-sec-butyl-1,10-phenanthroline; dipp, 2,9-di-isopropyl-1,10-phenanthroline;
dnpp, 2,9-dineopentyl-1,10-phenanthroline; dbp, 2,9-butyl-1,10-phenanthroline; xop, 2-(2-methylphenyl)-9-(2,6-dimethylphenyl)-1,10-phenanthroline;
dpp, 2,9-diphenyl-1,10-phenanthroline;bfp, 2,9-bis(trifluoromethyl)-1,10-phenanthroline.
McMillin,11 who contrasted the luminescence of [Cu(dmp)2]+
with that of non-emissive [Cu(phen)2]+. The methyl groups at
the 2 and 9 positions provided sufficient steric bulk that [Cu
(dmp)2]+ emitted at room temperature in CH2Cl2 solution.
Emission from [Cu(dmp)2]+ is characterized by a short
excited-state lifetime (90 ns), a modest quantum yield
(4.0 ꢀ 10-4), and facile quenching by exogenous ligands.12
When [Cu(dmp)2]+ distorts toward square planar geometry
in the excited state, a new coordination site becomes acc-
essible, and small molecule ligands may bind and quench the
emission.13,14 [Cu(dmp)2]+ does not emit in coordinating sol-
vents, and addition of Lewis bases in non-coordinating solvents
quenches the emission.11 Exciplex formation through ligand
Figure 1. Ball and stick and space filling diagrams of [Cu(dtbp)2]+
binding to the flattened copper(II)-like excited state is a primary
derived from the crystallographic coordinates.26
reaction path for [Cu(dmp)2]+.13,15,16 This pathway shuts down
when sufficient steric bulk is added on either the phenanthroline
of bulkier substituents, both the lifetime and the quantum
ligand, as in the case of 2,9-diphenyl-1,10-phenanthroline, or the
incoming Lewis base, as in the case of 2,6-dimethylpyridine.17
yield increase. To date, the most sterically demanding com-
bination of substituents was in a complex with heteroleptic
Subsequent studies have shown that the photophysical
ligation, [Cu(dmp)(dtbp)](PF6).20 This complex exhibited
properties of [Cu(R2Phen)2]+ complexes vary with the steric
demand at the 2 and 9 positions.11,12,17-25 As the geometric
the longest lifetime (730 ns) and the largest quantum yield
(1 ꢀ 10-2) among complexes with 2,9-dialkyl substituents.
reorganization of the excited state is inhibited by the addition
As illustrated in Table 1, the lifetime and quantum yield
increase with increasing steric demand of the substituents
(11) Blaskie, M. W.; McMillin, D. R. Inorg. Chem. 1980, 19, 3519–3522.
(12) Eggleston, M. K.; McMillin, D. R.; Koenig, K. S.; Pallenberg, A. J.
Inorg. Chem. 1997, 36, 172–176.
because of inhibition of non-radiative vibrational relaxation.
Increased ligand bulk also shuts down Lewis base-induced
quenching via exciplex formation; homoleptic complexes
with bulky phenanthroline ligands emit even in acetonitrile,
tetrahydrofuran, or methanol solutions.12,19
(13) McMillin, D. R.; Kirchhoff, J. R.; Goodwin, K. V. Coord. Chem. Rev.
1985, 64, 83–92.
(14) Chen, L. X.; Shaw, G. B.; Novozhilova, I.; Liu, T.; Jennings, G.;
Attenkofer, K.; Meyer, G. J.; Coppens, P. J. Am. Chem. Soc. 2003, 125,
7022–7034.
(15) Everly, R. M.; McMillin, D. R. J. Phys. Chem. 1991, 95, 9071–9075.
(16) Stacy, E. M.; McMillin, D. R. Inorg. Chem. 1990, 29, 393–396.
(17) Dietrich-Buchecker, C. O.; Marnot, P. A.; Sauvage, J. P.; Kirchoff
J. R.; McMillin, D. R. Chem. Commun. 1983, 513–515.
(18) Dietrich-Buchecker, C. O.; Nierengarten, J. F.; Sauvage, J. P.;
Armaroli, N.; Balzani, V.; De Cola, L. J. Am. Chem. Soc. 1993, 115,
11237–11244.
(19) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1999, 38,
3414–3422.
(20) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. J. Am. Chem. Soc.
1999, 121, 4292–4293.
(21) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1998, 37,
Recently, we designed a strategy that enabled synthesis of
[Cu(dtbp)2][B(C6F5)4], (1), the most sterically encumbered
homoleptic [Cu(R2Phen)2]+ complex prepared to date.26 As
illustrated by the space filling diagram in Figure 1, the tert-
butyl substituents impart limits on the potential for geometric
reorganization in [Cu(dtbp)2]+. Consistent with substantial
steric constraint, this complex exhibits an average copper(I)-
˚
N bond length of 2.11 A, which is significantly longer than
˚
the 2.04 A average copper(I)-N bond length for all other
complexes in this class. We report herein photophysical
characterization of 1 and exploration of the unique reactivity
of 1 that is afforded by the sterically weakened copper(I)-
N bonds. The reactivity of 1, particularly the role of dtbp in
this reactivity, is investigated by comparison with a complex
2285–2290.
(22) Cunningham, C. T.; Cunningham, K. L. H.; Michalec, J. F.;
McMillin, D. R. Inorg. Chem. 1999, 38, 4388–4392.
(23) Eggleston, M. K.; Fanwick, P. E.; Pallenberg, A. J.; McMillin, D. R.
Inorg. Chem. 1997, 36, 4007–4010.
(24) Felder, D.; Nierengarten, J.-F.; Barigelletti, F.; Ventura, B.; Armar-
oli, N. J. Am. Chem. Soc. 2001, 123, 6291–6299.
(25) Kalsani, V.; Schmittel, M.; Listorti, A.; Accorsi, G.; Armaroli, N.
Inorg. Chem. 2006, 45, 2061–2067.
(26) Gandhi, B. A.; Green, O.; Burstyn, J. N. Inorg. Chem. 2007, 46
3816–3825.

5706 Inorganic Chemistry, Vol. 48, No. 13, 2009
Green et al.
that has only a single 2,9-di-tert-butyl-1,10-phenanthroline
ligand, [Cu(dtbp)(acetone)](SbF6).26
electrode, a Ag|AgCl reference electrode and a Pt auxiliary
electrode. The electrochemical analyses were performed under
an Ar atmosphere in a homemade cell designed for low volume
applications.31 The sample, a 0.1 M solution of 1 in CH2Cl2 with
0.1 M tetrabutylammonium hexafluorophosphate as the sup-
porting electrolyte, was contained within a Teflon sleeve on the
working electrode. Data were collected at a scan rate of 50 mV/s.
The E1/2 of the ferrocene0/+ couple was 500 mV versus Ag|AgCl
in CH2Cl2 under identical experimental conditions to those used
for studying 1.
Experimental Section
All reagents were purchased from Aldrich Chemical Co.
and purified according to literature procedures unless
otherwise noted. CH3CN, CH2Cl2 and n-hexane were
purchased as spectroscopic grade (Burdick & Jackson)
and distilled from CaH2 under N2. All solvents were further
degassed by three freeze-pump-thaw cycles or by spar-
ging with Ar. The complexes [Cu(dtbp)2][B(C6F5)4] (1),
[Cu(dtbp)(NCCH3)](PF6), [Cu(dtbp)(acetone)](SbF6), and
[Cu(dtbp)(dmp)](BF4) were synthesized using a recently
developed method, which is described elsewhere.26
[Ru(bpy)3][(PF6)2] was synthesized and purified accord-
ing to literature procedure.27 CH3NC was synthesized as
described28 and was used immediately. O2 (99.5%) and
CO (99.5%) were purchased from BOC and AGA Gases,
respectively, and were used as received. All spectra were
recorded on samples that were maintained under an inert
atmosphere (N2) unless otherwise noted. Igor Pro 4.08
was used to process and fit spectrophotometric data to
obtain peak positions and integrated intensities.
Measurement of the Binding Constant for the Second
dtbp Ligand in 1. The binding constant for the second dtbp
ligand in complex 1 was determined. In this experiment, [Cu
(dtbp)(acetone)](SbF6) (3.6 ꢀ 10-5 M in CH2Cl2) was titrated
with aliquots (3 μL, 0.2 equiv) of a solution of dtbp (4.8 ꢀ 10-3
M in CH2Cl2), and binding was followed by the growth of the
425 nm metal-to-ligand charge-transfer (MLCT) peak. The
experiment was performed in triplicate. The binding constant
at room temperature was calculated employing a method for
complexes with high binding efficiency.32
Spectrophotometric Studies of Ligand Reactivity. The
reactivity of 1 with exogenous ligands CH3CN and CH3NC was
studied using absorption and emission spectrophotometric ti-
trations. A solution of 1 (2 mL of a 3.6 ꢀ 10-5 M solution in
CH2Cl2) was placed in a photoluminescence cuvette and titrated
with ligand (3.6 ꢀ 10-3 M). An aliquot of the titrant (0.20 or
0.25 equiv, 4 or 5 μL, respectively) was added to the sample and
reference cells, and the solutions were stirred vigorously for
5 min. After each addition, the absorption spectrum was re-
corded followed immediately by the photoluminescence spec-
trum. The binding affinities of CH3CN and CH3NC were
calculated, from emission and absorption data independently
for their respective titrations, according to the method for high
binding constants.32
Reaction of 1 with the gaseous ligands, CO and O2, was also
followed by absorption and emission spectroscopy. In these
experiments the gaseous ligands (50 μL, 31 equiv O2, 46 equiv
CO) were bubbled into a solution of 1 in CH2Cl2 (3.6 ꢀ 10-5 M)
via a gastight syringe, after which the absorption and emission
spectra were recorded. To test the reversibility of these reactions,
the solution was sparged with Ar for 5 min, the lost solvent was
replaced, and again absorption and emission spectra were
recorded.
1H NMR Studies of Ligand Reactivity. 1H NMR was used
to follow the reaction of 1 with CH3CN. 1H NMR spectra were
obtained on a Bruker AC+ 300a spectrometer at room tem-
perature. Chemical shifts were referenced to residual deuterated
solvent protons and are reported in parts per million (ppm)
versus (CH3)4Si. Complex 1 (0.01 M in CD2Cl2) was titrated
with CH3CN (0.2 M in CD2Cl2) in 0.25 equiv increments. After
addition of each aliquot of CH3CN, the tube was agitated and
allowed to equilibrate for 5 min before the spectrum was
recorded.
Spectrophotometric Measurements. Electronic absorp-
tion spectra were obtained with a Varian Cary 4 Bio spectro-
photometer, and photoluminescence data were collected with an
ISS PC-1 spectrofluorometer outfitted with a 300 W high
pressure Xe arc lamp source. All emission spectra were corrected
by applying correction factors provided with the instrumental
software (Vinci; v. 0.0797, ISS Inc.), which are specific to
the instrument. The quantum yield of 1 was estimated29 using
[Ru(bpy)3][(PF6)2] in water as a standard (Φem= 0.042).30
The emission spectrum of 1 in degassed CH2Cl2 was recorded
using 425 nm excitation. Igor Pro 4.08 was used to fit the
corrected emission spectrum to two Gaussians and to measure
the area under the curve. Integrated emission intensities were
obtained for 1 at three different concentrations, with 425 nm
absorbance values between 0.08 and 0.04 to avoid inner fil-
ter effects. An approximate quantum yield was calculated
according to eq 1, where Q is the quantum yield, I is the
integrated emission intensity, OD is the optical density, and
n is the refractive index. The subscript R denotes the refer-
ence values for the standard [Ru(bpy)3][(PF6)2]. The integrated
emission intensity of 1 was similarly compared to that of
[Cu(dtbp)(dmp)](BF4).20
Q=QR ¼ ðI=IRÞðODR=ODÞðn2=nR
Þ
ð1Þ
2
Emission lifetime data for 1 in CH2Cl2 were collected with an
ISS Chronos lifetime spectrometer equipped with a 370 nm LED
source. Frequency and phase-modulated data were collected
with a 550 kV long pass filter. The reference standard was
Ru(bpy)2Cl2. The experimental data for 1 were best fit with a
single exponential decay time of 3260 ns. Radiative and non-
radiative relaxation rates were calculated according to literature
procedure.9
FT Raman Studies of Ligand Reactivity. The reactions
of 1 with CH3CN and CO were followed by FT Raman spec-
troscopy. FT Raman data were collected with a RFS 100
spectrometer. A vial charged with 100 μL of 0.1 M solution of
1 in CH2Cl2 was titrated with 0.5 M CH3CN in CH2Cl2 in 5 μL
(0.25 equiv) aliquots. After each addition the solution was
agitated for 5 min, and an FT Raman spectrum was recorded.
To follow the reaction of 1 with CO, 100 μL aliquots of pure
gaseous CO were introduced through a septum into a vial
containing 100 μL of a 0.1 M solution of 1 in CH2Cl2. The vial
Electrochemical Measurements. Cyclic voltammetry to
determine the redox potential of 1 was performed using a BAS
Epsilon potentiostat outfitted with a glassy carbon working
(27) Lee, K. W.; Slinker, J. D.; Gorodetsky, A. A.; Flores-Torres, S.;
Abruna, H. D.; Houston, P. L.; Malliaras, G. G. Phys. Chem. Chem. Phys.
2003, 5, 2706–2709.
(28) Cappon, J. J.; Witters, K. D.; Baart, J.; Verdegem, P. J. E.; Hoek, A.
C.; Luiten, R. J. H.; Raap, J.; Lugtenburg, J. Recl. Trav. Chim. Pays-Bas
1994, 113, 318–328.
(29) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; 3rd ed.;
Springer: New York, NY, 2006.
(30) Caspar, J. V.; Meyer, T. J. J. Am. Chem. Soc. 1983, 105, 5583–5590.
(31) Feinberg, B. A.; Liu, X.; Ryan, M. D.; Schejter, A.; Zhang, C.;
Margoliash, E. Biochemistry 1998, 37, 13091–13101.
(32) Bourson, J.; Pouget, J.; Valeur, B. J. Phys. Chem. 1993, 97
4552–4557.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5707
was agitated vigorously for 5 min and an FT Raman spectrum
was recorded.
FT-IR Studies of Ligand Reactivity. The reactions of CO
with the complexes 1and [Cu(dtbp)(acetone)](SbF6)weremonitored
by FT-IR spectroscopy. Spectra were obtained on a Bruker Vertex
70 spectrometer using a liquid IR cell (NaCl windows, 0.1 mm path
length), into which the sample aliquot was introduced via a gastight
syringe. Solutions of 1and [Cu(dtbp)(acetone)](SbF6)inCH2Cl2 (0.1
M) were prepared in septum-sealed vials under N2, and the IR
spectrum of an aliquot of each sample was recorded. The solutions
were sparged with CO gas for approximately 5 min until the color
change was complete: from yellow-orange to pale yellow in the case
of 1, and from bright to pale yellow in the case of [Cu(dtbp)(acetone)]
(SbF6). Fresh CH2Cl2 was added as needed to restore the sample
volume, and the spectrum of an aliquot was recorded. The solvent
was then removed under vacuum, the solid was redissolved in fresh
CH2Cl2, and the spectrum of an aliquot was recorded. To the
solution that originally contained [Cu(dtbp)(acetone)](SbF6) was
added excess dtbp. This solution was then taken to dryness under
vacuum, the solid was redissolved in fresh CH2Cl2, and the spectrum
of an aliquot was recorded.
Figure 2. Absorption, excitation, and emission profiles of 1 in CH2Cl2.
(a) The absorption spectrum exhibits intense πfπ* transition centered at
275 nm and weaker MLCT transition centered at 425 nm. (b) The
excitation profile of 1 is overlaid on the absorption spectrum. Maximal
emission intensity (599 nm) correlates with excitation into the MLCT
absorption band. (c) The emission spectrum obtained with 425 nm
excitation shows a broad emission peak centered at 599 nm.
including intense πfπ* ligand absorption (275 nm) and
less intense dπfπ* MLCT absorption (425 nm).9 The
magnitude of the extinction coefficient of the MLCT
transition (Table 1) is on the order expected for a
dπfπ* transition.35 Excitation into this absorption
band produces a broad emission band with a maximum at
599 nm, typical of emission from an MLCT excited state.9
The 425 nm feature is absent from the spectrum of the
dtbp ligand, supporting the assignment of this band to the
MLCT absorption in the complex.
Synthesis of [Cu(dtbp)(CO)](SbF6). CO was allowed to
flow through a vial charged with a bright yellow CH2Cl2
solution of [Cu(dtbp)(acetone)](SbF6) until the solvent comple-
tely evaporated. The resulting pale yellow solid was dissolved in
fresh CH2Cl2, and CO was bubbled again until nothing re-
mained but a white solid. Crystallization of this solid from
CH2Cl2/hexanes yielded long colorless plates of [Cu(dtbp)
1
(CO)](SbF6). H NMR (300 MHz, CD2Cl2): δ 1.805 (s, 18H,
Complex 1 exhibits the longest lifetime and highest quan-
tum yield for this class of compounds (Table 1), substan-
tially larger than those observed for [Cu(dtbp)(dmp)]+.20
MLCT-derived emission from 1 is characterized by a life-
time (τ) of 3260 ns and an estimated quantum yield
(φ) of 5.6 ((0.4)% (5.6 ꢀ 10-2). The improvement in
the emission characteristics of 1, relative to those of
[Cu(dtbp)(dmp)]+,20 appears to be due to a reduction in
the non-radiative relaxation rate by almost a full order
of magnitude. The calculated radiative relaxation rate
(kr = φ/τ) for 1 is 1.7 ((0.2) ꢀ 104 s-1, while the non-
radiative relaxation rate (knr = (1 - φ)/τ)) is 3.0 ((0.2) ꢀ
105 s-1.9 The quantum yield of 1in CH2Cl2 is approximately
50% larger than that of [Ru(bpy)3][(PF6)2]inwater.30 Direct
comparison of the MLCT-derived emission intensity of
complex 1 with that of [Cu(dtbp)(dmp)](BF4), under the
same solution conditions, indicates that the quantum yield
of 1 is approximately five times larger (Supporting Informa-
tion, Figure S1).
Electrochemical Analysis. The reduction potential of
complex 1 (0.1 M) in CH2Cl2 solution was measured by
cyclic voltammetry (Supporting Information, Figure S2).
The voltammetric behavior of complex 1 was not ideal: in
a single cycle the forward and reverse currents were not
the same (ipa/ipc= 0.75, where ipa and ipc are the intensities
of the anodic and cathodic peak, respectively), and the
current flow in both sweep directions decreased upon a
second cycle. These observations suggest that the complex
degraded during the course of the experiment. Reasonable
behavior was obtained at a sweep rate of 50 mV/sec. The
separation of 180 mV between the anodic and cathodic
peaks was within the 100-200 mV range that has been
observed for others of this class of complexes.9 The apparent
CH3), δ 8.001 (s, 2H, CH), δ 8.159 (d, J = 8.4 Hz, 2H, CH), δ
8.621 (d, J = 9 Hz, 2H, CH) ppm. 13C NMR (125 MHz,
CD2Cl2) δ 34.01, 41.29, 125.66, 129.32, 130.72, 143.86, 145.95,
172.72, 173.24 ppm. IR (cm-1): SbF6-, 654; CO, 2130.
X-ray Structure Determination for[Cu(dtbp)(CO)](SbF6).
Under the microscope, several seemingly single crystals of
[Cu(dtbp)(CO)](SbF6), of which none showed any optical evi-
dence of twinning, were selected under oil in air at room
temperature. All examined crystals were twinned, and the
structure was solved based on the data set collected on a non-
merohedrally twinned crystal using CellNow.33,34 The twin
components were related by a 179.6° rotation about the c axis, with
the minor component contribution of 47.85%. Data frames were
integrated with the Bruker SAINT-Plus software package, and
unit-cell parameters were obtained from CellNow. Data were
correct for absorption effects with TWINABS [μ(Mo KR) =
2.216 mm-1; max/min transmission, 0.660/0.808]. The crystal
structure of [Cu(dtbp)(CO)](SbF6) was solved by direct methods.
All non-hydrogen atoms were identified on the initial electron
density map, and refined anisotropically by full-matrix least-
squares methods; all hydrogen atoms were calculated at idealized
positions and then refined as riding atoms with individ-
ual isotropic coefficients. Details of the data collection and
refinement are listed in Supporting Information, Table S1, and
structural parameters are provided in Supporting Information,
Tables S2-S6.
Results
Photophysical and Electrochemical Attributes of
[Cu(dtbp)2][B(C6F5)4] (1). Photophysical Studies. The
absorption spectrum, emission spectrum and excitation
profile of 1 are shown in Figure 2. The absorption features
are typical of bis(phenanthroline)copper(I) complexes,
(33) SADABS V.2.05, T., SAINT V.6.22, SHELXTL V.6.10, SMART
V.5.622 Software Reference Manuals; Bruker-AXS: Madison, WI,
2000-2004
(35) Harris, D. C.; Bertolucci, M. D. Symmetry and Spectroscopy: An
Introduction to Vibrational and Electronic Spectroscopy; Dover Publications:
New York, 1978.
::
::
(34) Sheldrick, G. M.; University of Gottingen: Gottingen, Germany,
2005.

5708 Inorganic Chemistry, Vol. 48, No. 13, 2009
Green et al.
Table 2. Cu2+/+ Potentials of Selected [Cu(R2Phen)2]+ Complexes and Their
Derived Excited-State Reduction Potentialsa,b
complex
E(Cu2+/+),cV ΔGes, eV E(Cu2+/+*),d V reference
[Cu(dtbp)2]+
[Cu(dsbp)2]+
[Cu(dbp)2]+
[Cu(dmp)2]+
[Cu(bfp)2]+
0.70
0.68
0.61
0.50
1.1
2.36
2.21
2.14
2.04
2.14
-1.66
-1.53
-1.53
-1.54
-1.04
this work
12
12
12
42
a The first four complexes are listed in order of decreasing bulk
of the alkyl ligand. b Abbreviations: bfp, 2,9-bis(trifluoromethyl)-1,10-
phenanthroline; dtbp, 2,9-di-tert-butyl-1,10-phenanthroline; dsbp, 2,9-
sec-butyl-1,10-phenanthroline; dbp, 2,9-dibutyl-1,10-phenanthroline;
dmp, 2,9-dimethyl-1,10-phenanthroline. c All values listed were
Figure 3. Reaction of 1 with CH3CN followed by absorption spectro-
scopy. Electronic absorption spectra of 1 (36 μM in CH2Cl2) titrated with
CH3CN (0.2 equiv aliquots up to 1 equiv). The spectra obtained after
addition of 0 (blue solid line), 1 (green solid line) and 3 (red dashed line) of
CH3CN are highlighted. Inset: the MLCT absorption of 1.
measured versus ferrocene (Fc+/0
) ) =
in CH2Cl2. d E(Cu2+/+*
E(Cu2+/+) - ΔGes; calculated versus ferrocene (Fc+/0
)
E1/2 of 1200 mV vs Ag|AgCl, 700 mV vs Fc+/0, was assigned
to the Cu2+/+ couple (Table 2).
An estimate of the reduction potential of the excited
state may be calculated from the emission spectral data
and the ground-state reduction potential. The maximum
free energy of an emissive state (ΔGes, eV) may be
determined by extending a tangent from the high-energy
side of the emission spectrum to the energy axis.36
The value for ΔGes of 1 obtained by use of this method
and those reported for related complexes in the litera-
ture are listed in Table 2. The excited-state potential
[Cu2+(R2Phen)2]|[Cu2+(R2Phen)(R2Phen-•)] may be es-
timated from ΔGes, which is the maximum energy differ-
ence between photoexcited [Cu2+(R2Phen)(R2Phen-•)]
and ground [Cu+(R2Phen)2], and E1/2, which mea-
sures the energy difference between [Cu2+(R2Phen)2]
and [Cu+(R2Phen)2]. The difference between E1/2
and ΔGes is then the potential for the [Cu2+(R2Phen)2]|
[Cu2+(R2Phen)(R2Phen-•)] reduction process.9 The ex-
cited-state reduction potential of 1 was estimated from
this method to be -1.66 V versus ferrocene+/0. A com-
parison of estimated potentials for the [Cu2+(R2Phen)2]|
[Cu2+(R2Phen)(R2Phen-•)] process for 1 and related
complexes is given in Table 2.
Reactivity of 1 and Related Complexes with Exogenous
Ligands. Binding Affinity of the Second dtbp Ligand.
Complex 1 undergoes facile ligand replacement reactions,
where one dtbp ligand is lost. When 1 was dissolved in
coordinating solvents, its characteristic bright orange
color disappeared and was replaced by a yellow color.
This color change occurred in methanol, acetone, and
CH3CN, but not in CH2Cl2. Since the bright orange color
is characteristic of bis(phenanthroline) coordination to
copper(I), it appeared that one of the ligands was dis-
placed upon dissolution, even in a weakly coordinating
solvent like acetone (Reaction 2, Y is solvent).
Figure 4. Reaction of 1 with CH3CN followed by emission spectro-
scopy. Photoluminescence spectra of 1 (36 μM in CH2Cl2) titrated with
CH3CN (0.25 equiv aliquots up to 1 equiv). The spectra obtained after
addition of 0 (solid line), 1 (dashed line) and 3 (dashed-dotted line) equiv
of CH3CN are highlighted. The excitation wavelength was 425 nm.
appropriate for complexes with high stability con-
stants;32 the resulting binding constant for dtbp was 9.9
((0.3) ꢀ 105. This binding constant defines the relative
affinities of acetone and dtbp for the Cu center, as it was not
possible to prepare a complex bearing one dtbp and no
other ligand.
½CuðdtbpÞðacetoneÞꢁðSbF6Þ þ dtbp T
½CuðdtbpÞ2ꢁðSbF6Þ þ acetone ð3Þ
Reactivity of 1 with CH3CN. The reaction of 1 with
CH3CN was studied in more detail. Changes attributable
to the displacement of dtbp from 1 were observed in the
absorption and emission spectra of 1 upon titration with
CH3CN (Reaction 4, Figures 3 & 4). Most noteworthy
was the loss of the MLCT absorption at 425 nm upon
addition of 1 equiv of CH3CN, and the appearance of new
features between 300 and 325 nm (Figure 3). Changes in
the πfπ* transition region were also observed. As
CH3CN was added, the 275 nm band split into two new
bands: one at 271 nm, the position of the πfπ* absorp-
tion of the free dtbp ligand, and one at 279 nm. Indepen-
dent synthesis and crystallographic26 and spectroscopic
characterization (Supporting Information, Figure S3) of
[Cu(dtbp)(NCCH3)](PF6) confirmed that the absorptions
at 279, 310, and 325 nm were from the complex [Cu(dtbp)
(NCCH3)]+. Addition of CH3CN to 1 also resulted in loss
of the MLCT-derived emission at 599 nm (Figure 4). Note
that in both absorption and emission experiments spec-
tral changes are observed up to 1 equiv of CH3CN;
½CuðdtbpÞ2ꢁþ þ Y T ½CuðdtbpÞðYÞꢁþ þ dtbp ð2Þ
The binding affinity of the second dtbp ligand in 1 was de-
termined by measuring a binding constant for Reaction 3.
In CH2Cl2, [Cu(dtbp)(acetone)](SbF6) was titrated with
dtbp, and the growth of the MLCT absorption at 425 nm
was monitored. The data were fit to an expression
(36) Arnold, D. R.; Baird, N. C.; Bolton, J. R.; Brand, J. C. D.; Jacobs, P.
W. M.; DeMayo, P.; Ware, W. R. Photochemistry: An Introduction;
Academic Press: New York, NY, 1974.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5709
Figure 5. Reaction of 1 with CH3CN followed by 1H NMR. Shown are
the changes in the aliphatic region upon titration of 1 with CH3CN.
Spectra proceed from bottom to top: 1 in CD2Cl2 was titrated with
0-1.5 equiv. of CH3CN (in CD2Cl2) in 0.25 equivalent aliquots. The
chemical shifts of the various tert-butyl protons of dtbp are indicated:
1 (*; δ 1.21 ppm), [Cu(dtbp)(NCCH3)]+ ((; δ 1.55 ppm), and dtbp (9;
δ 1. 73 ppm). The resonance that is initially observed at δ 2.43 ppm (b) is
assigned to the methyl group of CH3CN in [Cu(dtbp)(NCCH3)]+
.
minimal further changes are observed upon addition of
up to 3 equiv CH3CN. The equilibrium constant for
reaction 4, describing the relative binding affinities of
CH3CN and dtbp, was calculated to be 4 ((2) ꢀ 107.
Figure 6. Titration of 1 with CH3CN followed by FT-Raman spectro-
scopy. Spectra show the addition of (a) 0, (b) 0.5, (c) 1.0, (d) 1.5, (e) 2.0,
and (f) 3.5 equiv. of CH3CN to 1 in CH2Cl2 solution. Highlighted
vibrations include: νCN of [Cu(dtbp)(NCCH3)]+ (*; 2283 cm-1); νCN
of free CH3CN in CH2Cl2 solution (b; 2253 cm-1); an a1 vibration
of free dtbp ((; 1404 cm-1), and analogous modes of dtbp in
[Cu(dtbp)(NCCH3)]+ (9; 1424 cm-1) and dtbp in 1 (); 1391 cm-1).
½CuðdtbpÞ2ꢁþ þ CH3CN T
½CuðdtbpÞðCH3CNÞꢁþ þ dtbp ð4Þ
an additional peak corresponding to free CH3CN was
observed at 2253 cm-1. Notably, the νC-N of bound
CH3CN in [Cu(dtbp)(NCCH3)]+ is 30 cm-1 above the
value of νC-N observed for free CH3CN, indicating a
non-classical interaction of copper(I) with CH3CN in this
complex.37,38
The displacement of dtbp by CH3CN was explicitly
demonstrated by 1H NMR (Figure 5), following changes
in the distinct resonances of the tert-butyl substituents of
dtbp. The tert-butyl groups of 1 appear at 1.21 ppm, the
tert-butyl groups of the free dtbp ligand appear at
1.55 ppm, and those of the complex [Cu(dtbp)(NCCH3)]+
appear at 1.75 ppm (Supporting Information, Figure S4). As
CH3CN is added to 1, the intensity of the resonance at
1.21 ppm decreases, and new resonances grow in simul-
taneously at 1.55 ppm and 1.73 ppm. This observation
clearly reveals that CH3CN displaces one dtbp ligand
from the coordination sphere of copper(I). The reaction
is essentially complete upon addition of 1 equiv CH3CN.
At 1 equiv, the 1.21 ppm resonance of 1 is barely observed.
Minimal further change in the intensity of the resona-
nces of free dtbp (1.55 ppm) and [Cu(dtbp)(NCCH3)]+
(1.73 ppm) occurs upon addition of up to 3 equiv CH3CN.
The methyl protons of CH3CN appear at 2.43 ppm in
[Cu(dtbp)(NCCH3)]+ as it is formed from 1 upon addition
of 1 equiv CH3CN. When excess CH3CN is added, this
resonance shifts toward that of free CH3CN (2.04 ppm)
suggesting that there is rapid exchange between the free
and bound CH3CN.
Reactivity of 1 with CO. CO displaces one dtbp ligand
from 1 in a reaction analogous to that observed for
CH3CN (Reaction 5). The [Cu(dtbp)(CO)]+ complex
was independently synthesized from [Cu(dtbp)(acetone)]+
and the characterization of this new complex will be
discussed below as it pertains to the reactivity of 1.
Absorption, emission, FT Raman, and IR spectroscopies
were used to monitor the reaction of 1 with excess CO; the
spectroscopic features characteristic of 1 are replaced by
those characteristic of [Cu(dtbp)(CO)]+ and free dtbpupon
reaction. A color change from bright orange to an almost-
colorless pale yellow is observed as gaseous CO is added.
This color change is complete within seconds. As in the
reaction with CH3CN, the MLCT absorption at 425 nm
characteristic of 1 is lost, and the πfπ* transition at 275
nm is replaced by two new absorption features at 284 and
271 nm, due to [Cu(dtbp)(CO)]+ and free dtbp, respectively
(Supporting Information, Figure S5). Loss of the MLCT
absorption band of 1 upon reaction with CO results in near
complete loss of the MLCT-derived emission at 599 nm
(Table 3). Also notable are changes in the ligand based ring
vibrations in the FT Raman spectrum (Supporting Infor-
mation, Figure S6), analogous to those observed in reac-
tion with CH3CN (Figure 6), with the appearance of a new
FT Raman analysis corroborates the displacement of a
single dtbp ligand by CH3CN (Figure 6). In the absence of
CH3CN, the Raman spectrum of 1 revealed a ligand-
based ring vibration at 1391 cm-1, with a shoulder at
higher energy (Figure 6a). The analogous vibration
appears at 1404 cm-1 in free dtbp and at 1420 cm-1 in
[Cu(dtbp)(NCCH3)]+. As CH3CN is added to 1, the
peak at 1391 cm-1 is replaced by two peaks, at the
positions characteristic of the free ligand (1404 cm-1) and
[Cu(dtbp)(NCCH3)]+ (1421 cm-1). The reaction can also
be followed by the appearance of the C-N mode of
C-O stretch at 2130 cm-1
.
½CuðdtbpÞ2ꢁþ þ CO T ½CuðdtbpÞðCOÞꢁþ þ dtbp ð5Þ
The reaction of 1 with CO is reversible. When the
solution obtained after reaction of 1 with CO, which
bound CH3CN in [Cu(dtbp)(NCCH3)]+ at 2283 cm-1
:
this mode is completely absent in 1 and grows in only as
CH3CN is added. When 3 equiv of CH3CN were added,
(37) Thomas, J. C.; Peters, J. C. Polyhedron 2004, 23, 2901–2913.
(38) Strauss, S. H. Dalton Trans. 2000, 1–6.

5710 Inorganic Chemistry, Vol. 48, No. 13, 2009
Table 3. Excess CO and O2 Reversibly Quench the Emission of 1a
percent initial emission (I/I0 ꢀ 100)
Green et al.
CO
O2
1 in CH2Cl2
1st exposure
1st Ar sparge
2nd exposure
2nd Ar sparge
100%
3.7%
87%
2.9%
70%
100%
44%
89%
42%
83%
a A 36 μM solution of 1 in CH2Cl2 was exposed to excess CO or O2.
The gas was removed by sparging with Ar, and the solvent volume was
restored. This procedure was repeated for a total of two times. Absorp-
tion (Supporting Information, Figures S5 and S9) and emission data
were collected after each manipulation.
Figure 7. Reactions of 1 and [Cu(dtbp)(acetone)](SbF6) with CO fol-
lowed by solution FT-IR spectroscopy. Left column: FT-IR spectra of a
solution of 1 in CH2Cl2: (a) in the absence of CO; (b) upon exposure to
CO; (c) after removal of solvent in vacuo and addition of fresh solvent.
Right column: FT-IR spectra of a solution of [Cu(dtbp)(acetone)](SbF6)
in CH2Cl2: (d) in the absence of CO; (e) upon exposure to CO; (f) after
removal of solvent in vacuo and addition of fresh solvent; (g) the solution
from (f) to which excess dtbp was added, the solvent was removed in
vacuo, and fresh solvent was added.
contained [Cu(dtbp)(CO)]+ and free dtbp, was sparged
with Ar, the characteristic absorption and emission fea-
tures of 1 returned. When CO was reintroduced, the
products [Cu(dtbp)(CO)]+ and free dtbp were once again
observed (Supporting Information, Figure S5). Exposing
solutions of 1 to cycles of CO and Ar, with restoration of
evaporated solvent, revealed that after two complete
cycles (CO exposure, degassing with Ar, addition of lost
solvent) 70% of the initial photoluminescence intensity
was recovered (Table 3). Evidence of incomplete rever-
sion is present in the absorption spectra (Supporting
Information, Figure S5); the intensity of the MLCT
absorption of 1 was not fully restored after addition
and removal of CO. The reversibility of Reaction 5 was
also studied by FT-IR. A series of spectra were recorded
(1) before exposure of 1 to CO, (2) after bubbling with
CO, and (3) after the reaction solution had been taken to
dryness in vacuo and the resulting solid redissolved in
fresh CH2Cl2. Figure 7 shows the νC-O stretching region
of these spectra: no νC-O band was seen before exposure
of 1 to CO (Figure 7a) but upon exposure of 1 to CO, a
C-O stretch appeared at 2130 cm1 (Figure 7b). When the
products were dried in vacuo and redissolved in CH2Cl2,
the C-O stretch disappeared from the spectrum
(Figure 7c).
Reversible CO binding is only observed when a
second dtbp ligand is present. FT-IR studies, parallel
to those described for 1 above, were performed using
the complex [Cu(dtbp)(acetone)](SbF6). Excess CO
reacted readily with [Cu(dtbp)(acetone)](SbF6) to form
[Cu(dtbp)(CO)]+, as illustrated in Figure 7, spectra d and
e. When the product [Cu(dtbp)(CO)]+ was treated in the
same manner as the product of the reaction of 1 with CO,
that is, taken to dryness under vacuum and redissolved in
fresh solvent, the CO remained bound to the metal center
(Figure 7f). Excess dtbp ligand was then added to the
same solution, the solvent was removed in vacuo, and the
resulting solid was redissolved. The solution product in
the presence of excess dtbp ligand bore no CO (Figure 7g).
These experiments reveal that reversible CO binding is
enabled by the presence of uncoordinated dtbp, which
replaces CO as the ligand.
ligation to the copper(I) center. As shown in Supporting
Information, Figure S7, the IR spectrum of this complex
in CH2Cl2 solution reveals a νC-O mode at 2130 cm-1
,
illustrating classical binding of CO to copper(I).38 The
band at 2130 cm-1 shifts to 2080 cm-1 when the solution
of [Cu(dtbp)(CO)]+ is equilibrated with 13CO, thus con-
clusively identifying this as the νC-O mode. The 1H NMR
spectrum of [Cu(dtbp)(CO)](SbF6) showed no evidence
of acetone methyl groups, and the tert-butyl methyl
proton resonance is shifted downfield relative to that of
[Cu(dtbp)(acetone)](SbF6). The quaternary carbon of the
CO ligand appeared at 173.24 ppm in the 13C NMR
spectrum.
½CuðdtbpÞðacetoneÞꢁðSbF6Þ þ CO T
½CuðdtbpÞðCOÞꢁðSbF6Þ þ acetone ð6Þ
[Cu(dtbp)(CO)](SbF6) crystallized in a centrosym-
metric triclinic unit cell of P1 symmetry containing two
formula units (Z = 2), such that one [Cu(dtbp)(CO)]+
cation and one SbF-6 anion comprise the crystallographi-
cally independent part. The cation possesses pseudo-C2v
symmetry, with a trigonal planar three-coordinate Cu(I)
(Figure 8). This complex is the first three-coordinate
[Cu(dtbp)X]+/0 complex in which the third ligand lies in
the phenanthroline plane, in contrast to the distorted
asymmetry of previously characterized [Cu(dtbp)X]+/0
complexes.26,39 The methyl groups of the tert-butyl
substituents are positioned symmetrically so as to limit
crowding about the CO ligand. The coordination sphere
of the metal ion is composed of the two nitrogen atoms of
the phenanthroline and the carbon atom of the CO
˚
ligand. The mean of 2.04 A for the two pseudoequi-
˚
˚
valent Cu-N distances (2.045(2) A, 2.034(3) A) in the
[Cu(dtbp)(CO)]+ cation is similar to those of sterically
unconstrained (2,9-phenanthroline)Cu(I) complexes, but
The complex [Cu(dtbp)(CO)]+ was independently
synthesized to corroborate the spectral assignments made
in the reaction of 1 with CO. Reaction 6 was carried out
in the non-coordinating solvent CH2Cl2 to prepare
[Cu(dtbp)(CO)](SbF6). Evaporation of the solvent under
a CO flow was necessary to completely remove the
acetone ligand, thus eliminating any competition for
˚
˚
0.07 A shorter than the mean Cu-N distance of 2.11 A in
sterically encumbered 1.26 A salient structural feature of
[Cu(dtbp)(CO)]+ is the linkage between the terminal CO
and the d10 Cu(I) ion, with Cu-C and C-O bond
(39) Pallenberg, A. J.; Koenig, K. S.; Barnhart, D. M. Inorg. Chem. 1995,
34, 2833–2840.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5711
Figure 8. Molecular drawings of the cation in [Cu(dtbp)(CO)](SbF6) shown with 50% probability thermal ellipsoids. The Cu atom and the CO ligand are
perfectlyinthe planeofthe phenanthrolineligand. H atoms have beenomitted forclarity. Three views are shown:(a)sideview, (b)end view, and (c) top view.
˚
distances of 1.814(3) A and 1.132(3) A, respectively.
˚
Notably, the C-O bond is only slightly longer than the
1.128 A of free CO, which in conjunction with the
˚
high C-O stretching frequency reveal the minimal back
bonding in this complex.
Reactivity of 1 with O2. Dioxygen does not displace
dtbp from 1 (Reaction 7), but O2 does partially quench the
luminescence of 1. Upon exposure to excess O2 the
absorption spectrum of 1 changes slightly; a minor
diminution of the intensity of the πfπ* ligand absorp-
tion (275 nm) is observed (Supporting Information,
Figure S8). There is no evidence of an absorption band
at 271 nm from the free ligand or of any other new bands.
Similarly, there is a modest change in the intensity of
the MLCT absorption of 1 upon O2 addition. Since the
MLCT-derived emission intensity is very sensitive to the
presence of the second dtbp ligand, we conclude that
O2 does not displace dtbp from 1. Excess O2 does quench
the luminescence of 1, albeit to a substantially lesser
extent than excess CO (Table 3). The luminescence
quenching observed on exposure of 1 to O2 is reversible
upon sparging the solution with Ar. The quenching
induced by O2 is more reversible than that induced by
CO; after two exposures of 1 to O2, each followed by
removal with an Ar sparge, 83% of the initial emission
intensity is restored.
Figure 9. Reaction of 1 with CH3NC followed by absorption spectro-
scopy. Electronic absorption spectra of 1 (36 μM in CH2Cl2) titrated with
CH3NC. The spectra obtained after addition of 0 (red solid line),
1 (green solid line) and 6 (blue solid line) equiv of CH3NC are shown.
Inset: the MLCT absorption of 1.
The binding constant for the additional CH3NC ligands
(Reaction 8b) was calculated to be 4 ꢀ 103 using inten-
sities at 269 nm from 1 to 6 equiv CH3NC. The identity of
the final copper(I) product of reaction 8b was not verified
but is believed to be a homoleptic isonitrile complex,
analogous to the common copper(I) starting material
[Cu(CH3CN)4]+. Thus, the sum of Reactions 8a and 8b,
which is Reaction 8c, exhibits an equilibrium constant of
1.2 ꢀ 1013 M-2
.
½CuðdtbpÞ2ꢁþ þ CH3NC T
½CuðdtbpÞðCH3NCÞꢁþ þ dtbp ð8aÞ
½CuðdtbpÞ2ꢁþ þ O2 T No Reaction
ð7Þ
½CuðdtbpÞðCH3NCÞꢁþ þ 3CH3NC þ dtbp T
Reactivity of 1 with CH3NC. CH3NC displaces both
dtbp ligands from 1 (Reactions 8a,b). As was observed
for reaction of 1 with CH3CN and CO, addition of
1 equiv CH3NC is accompanied by changes in the absorp-
tion spectrum consistent with displacement of one dtbp
ligand from the metal (Figure 9). However, unlike
CH3CN or CO, addition of CH3NC beyond 1 equiv
results in a continued increase in the absorption at
271 nm corresponding to free dtbp, indicating that
CH3NC is capable of displacing the second dtbp. Equili-
brium constants for reactions 8a and 8b were calculated
from the absorption data. The binding constant for
the first CH3NC (Reaction 8a) was found to be 3 ꢀ 109
using the intensity at 425 nm from 0 to 1 equiv CH3NC.
½CuðCH3NCÞ ꢁþ þ 2dtbp ð8bÞ
4
½CuðdtbpÞ2ꢁþ þ 4CH3NC T
½CuðCH3NCÞ ꢁþ þ 2dtbp ð8cÞ
4
Discussion
The properties of 1 extend established trends in absorption
and emission characteristics of bis(phenanthroline)copper(I)
complexes.5,6,8,9 Table 1 lists photophysical characteristics of
selected complexes with varied substituents at the 2 and 9

5712 Inorganic Chemistry, Vol. 48, No. 13, 2009
Green et al.
positions of the ligand and organized in order of decreasing
emission lifetime and decreasing quantum yield. The lifetime
and quantum yield correlate with the apparent steric bulk of
the ligand, with complex 1 showing the longest observed
lifetime and highest quantum yield in this class of complexes.
The lifetime increases with the energy of emission in keeping
with the energy gap law.29 The presence of the bulky tert-
butyl groups in complex 1 modestly increases the radiative
relaxation rate, kr, and dramatically decreases the non-
radiative relaxation rate, knr. This substantial decrease in
knr is consistent with inhibition of the geometric reorganiza-
tion in the excited state, as complexes in which the tetrahe-
dral-to-square planar structural change is inhibited are
consistently more emissive.5,6,8,9 Further evidence for mini-
mal reorganization in the excited state is provided by the
difference between the energies of absorption and emission,
whose value is smaller for 1 than for other [Cu(R2Phen)2]+
complexes. Luminesence in complexes with MLCT-derived
excited states may occur from states of singlet or triplet
multiplicity, and the magnitude of EAbs - EEm is related to
the energy losses that occur with internal conversion, inter-
system crossing, and geometric reorganization.40,41 The
0.847 eV difference observed in 1 implies that there are
smaller energy losses, presumably because the bulky tert-
butyl groups minimize structural changes in the excited
states. The oscillator strength of the complex 1 MLCT
absorption, with an ε of 3100 L M-1 cm-1, is among the
smallest of this series because of the poorer overlap between
the metal and ligand orbitals that is a consequence of the
elongated Cu-N bonds.26
Differential effects of sterics and electronics on the proper-
ties of [Cu(R2Phen)2]+ complexes may be deduced from a
comparison between 1 and [Cu(bfp)2]+.42,43 The last three
entries of Table 1 feature ligands with electron withdrawing
groups at the 2 and 9 positions, for which there is no obvious
correlation between the photophysical properties and
electronic effects. We note that [Cu(bfp)2]+, with the most
electron withdrawing ligands, exhibits a value of EAbs - EEm
that is smaller than that of complex 1. One might conclude
that [Cu(bfp)2]+ exhibits less excited-state distortion than 1;
however, the ground-state crystallographic data show that
[Cu(bfp)2]+ is in fact more easily distorted.43 The data thus
suggest that the behavior of [Cu(bfp)2]+ is largely dictated by
electronic, not steric, effects. This conclusion is supported by
a comparison of the electrochemical behavior of [Cu(bfp)2]+
and 1. Both complexes exhibit high positive ground-state
reduction potentials. The electrochemical behavior of
[Cu(bfp)2]+ is perfectly reversible, and the potential is the
mostpositiveamongall[Cu(R2Phen)2]+ complexes(Table2),
as the highly electron withdrawing groups stabilize the
copper(I) state.42 The apparent E1/2 for complex 1 is the most
positive among complexes with 2,9-dialkyl-1,10-phenanthro-
line ligands, illustrated by the first four entries in Table 2. The
reduction potential increases with steric bulk, which stabilizes
copper(I) relative to copper(II) by inhibition of the geometric
reorganization required in the latter.9 In contrast to
[Cu(bfp)2]+, the electrochemical behavior of 1 is not rever-
sible. With modestly sized CF3 substituents, [Cu(bfp)2]+ is
able to accommodate the tetragonal geometry preferred by
copper(II),43 while the exceptional bulk of the tert-butyl
groups renders the oxidized form of complex 1 unstable.
Also listed in Table 2 are the excited-state reduction poten-
tials for the [Cu2+(R2Phen)(R2Phen-•)]|[Cu2+(R2Phen)2]
couple. We note that the high ΔGes of complex 1 contributes
a large negative value in the calculation of the excited-
state reduction potential, and the calculated potential indi-
cates that the excited state of 1, [Cu2+(dtbp)(dtbp-•)], is
a potent photoreductant. In contrast, the highly positive
reduction potential of [Cu(bfp)2]+ coupled with a much
lower ΔGes value renders [Cu2+(bfp)(bfp-•)] a weak photo-
reductant.42
Copper(I) complexes are generally labile; however, the
exceptionally bulky tert-butyl substituents at the 2 and 9
positions of the phenanthroline enable unique ligand sub-
stitution reactivity in complex 1. In coordinating solvents,
such as acetone or acetonitrile, one of the bidentate dtbp
ligands is replaced by solvent. For most complexes in
this family, the only observable species in solution is the
bis-phenanthroline species, [Cu(R2Phen)2]+, because of the
chelating nature of the ligands.39 However, the dtbp ligand is
so bulky that accommodating four tert-butyl groups about
the copper(I) center is more challenging, and only the novel
reaction conditions under which complex 1 is synthesized
allow for the binding of two dtbp ligands.26,44 Complex 1 is
prepared inacetone by reaction of 2 equivofdtbp with 1 equiv
of Ag[B(C6F5)4] and excess copper metal. With an excess of
acetone present, the immediate product is the acetone solvato
complex, [Cu(dtbp)(acetone)]+. When the acetone solvent is
removed upon workup, the second equivalent of dtbp binds
to form complex 1. Spectrophotometric titration of the
acetone solvato complex with the second dtbp to form 1
revealed that the binding constant of dtbp relative to the
bound acetone was 9.9 ((0.3) ꢀ 105. In stark contrast,
titration of 1 equiv of CH3CN into complex 1 revealed the
binding constant of CH3CN relative to the second dtbpligand
to be 4 ((2) ꢀ 107. These binding affinities dramatically
demonstrate that
a ligand displacement reaction on
[Cu(CH3CN)4]+, in which a minimum 4 equiv of CH3CN
are present regardless of solvent, is a poor choice for the
synthesis of complex 1.20,26
An important consequence of low affinity for the second
dtbp ligand is its facile displacement in solution. The solid-
state and solution-state FT-Raman spectra of complex 1
revealed differences in the intense feature, which we assign to
an a1 symmetry phenanthroline ring vibration.45 The solid-
state spectrum of 1 reveals a peak at 1400 cm-1, while the
solution-state spectrum in CH2Cl2 reveals two frequencies
associated with this feature: a peak at 1391 cm-1, with a
higher energy shoulder at 1402 cm-1 (Supporting Informa-
tion, Figures S9 and S10). The single peak in the solid-state
spectrum indicates an equality in the binding of the two dtbp
ligands about the copper(I) center, confirming what is ob-
served crystallographically where the Cu-N distances are
similar.26 In contrast, the two peaks of the solution-state
(40) Iwamura, M.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2007, 129,
5248–5256.
(41) Bock, C. R.; Connor, J. A.; Gutierrez, A. R.; Meyer, T. J.; Whitten,
D. G.; Sullivan, B. P.; Nagle, J. K. J. Am. Chem. Soc. 1979, 101, 4815–4824.
(42) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Angew. Chem., Int. Ed.
1998, 37, 1556–1558.
(44) Morgan, H. H. J. Chem. Soc. 1923, 2901–2907.
(45) Reiher, M.; Brehm, G.; Schneider, S. J. Phys. Chem. A 2004, 108,
734–742.
(43) Kovalevsky, A. Y.; Gembicky, M.; Coppens, P. Inorg. Chem. 2004,
43, 8282–8289.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 5713
spectrum suggest an inequality in the binding of the two
phenanthroline ligands about the metal center. The position
of the shoulder, 1402 cm-1, is within 2 cm-1 of the
a1 vibration observed in a CH2Cl2 solution of free dtbp,
1404 cm-1 (Supporting Information, Figure S6e). The higher
energy shoulder suggests the presence of a population of
weakly associated dtbp in solution that arises as a result of the
increased steric congestion about the copper(I) center. These
data also suggest that while the two dtbp ligands of complex 1
are bound in similar fashion in the solid state (1400 cm-1), the
solid-state binding is more similar to that of the weakly
bound dtbp of 1 in solution (1402 cm-1) than to that of the
strongly bound dtbp (1391 cm-1).
The intrinsic instability of the homoleptic complex 1 may
be of value in the design of molecular-based devices, as labile
copper(I) complexes have been exploited to great effect in the
synthesis of controlled molecular assemblies.46-51 In parti-
cular, the coordination geometry and coordination number
preferences of copper(I), and the distinctly different prefer-
ences of copper(II), have been used in the design of photo- or
electrochemically addressable molecular machines.52 Bulky
copper(I) phenanthroline complexes have proved particu-
larly useful in the synthesis of predictable assemblies of
heteroleptic complexes in which it is essential to control the
thermodynamics.53-55 This HETPHEN approach is char-
acterized by judicious choice of bulky phenanthroline
ligands, such that the two homoleptic [Cu(R2Phen)2]+ spe-
cies are less stable than the heteroleptic combination. This
method has exploited bulky aryl-substituted phenanthro-
line ligands; however, the instability of complex 1 suggests
that tert-butyl-phenanthroline may also be useful for this
purpose. The synthetic method used to prepare 1,26 which
avoids [Cu(CH3CN)4]+ as a starting material, was superior
for the preparation of the heteroleptic complex [Cu(dmp)
(dtbp)]+,20 and this method may be similarly advantageous
in the clean preparation of more complex heteroleptic
assemblies.
luminescence quenching is also observed with O2; however,
the absence of any change in the absorption spectrum
was inconsistent with a ligand exchange reaction. Rather,
the data suggest an outer-sphere interaction between
between 1 and O2, similar to that observed previously
for [Cu(dmp)(dtbp)]+.56 Interestingly, complex 1 shows
virtually no tendency to react with O2,57,58 which is un-
usual for copper(I) complexes but consistent with the
large positive reduction potential that inhibits oxidative
reactivity.
Using the qualitative and quantitative data collected in this
study, we may construct a ligand affinity series for the
[Cu(dtbp)]+ fragment. The binding affinity for an additional
ligand increases in the following order: acetone < dtbp < CO
≈ CH3CN < CH3NC. Strongly donating monodentate
ligands, CO, CH3CN and CH3NC, react with 1 to displace
one bidentate dtbp ligand. This reactivity is unique in that it
defies the chelate effect, in which bidentate ligands typically
enhance the stability of metal complexes relative to mono-
dentate ligands. Furthermore, in the case of CO and CH3CN,
the greater affinity of the monodentate ligands over the
bidentate dtbp is observed in the absence of significant back-
bonding between the copper(I) and π acid ligand.38 CH3NC
very efficiently displaces the first dtbp ligand from complex 1,
and also displaces the remaining bound dtbp ligand, although
considerably less efficiently.
Conclusion
We have characterized the photophysical properties of
the most sterically constrained copper(I)-phenanthroline
complex, [Cu(dtbp)2]+. The photophysical measurements
revealed a quantum yield on par with and an excited-state
lifetime longer than those of [Ru(bpy)3]2+. The exceptional
steric constraints in complex 1 weaken the metal-ligand
bonding, which in turn afforded a unique type of reactivity.
One of the chelating dtbp ligands is readily replaced by
strongly donating, monodentate ligands such as acetonitrile
and CO. The unique combination of excellent photophysical
properties and ligand displacement reactivity renders com-
plex 1 attractive for use in sensors, molecular machines, or
photoelectronic devices.
The excellent photophysical properties of complex 1 and
its instability toward ligand substitution set the stage for
the exploitation of this complex in sensing applications.
For example, CO has a greater binding affinity for the
copper ion than does the second bidentate dtbp ligand, and
the binding of CO results in near complete quenching
of luminescence. The excess dtbp remaining in solution aids
in driving off the bound CO, reversing the reaction and
restoring the luminescence. This facile reversibility is only
possible in the presence of the extra equivalent of dtbp to
fill the coordination sphere of the copper ion. Reversible
Acknowledgment. The authors are grateful to
Dr. Ewald Terpetschnig of ISS Inc. for the emission
lifetime measurements. We would also like to thank
Dr. Umesh Agarwal and Mr. Richard Reiner of the
Forest Products Laboratory (Madison, WI) for use of
the FT-Raman spectrometer. Acknowledgment is made
to the Donors of the American Chemical Society
Petroleum Research Fund for support of this research
(ACS-PRF Grant 42041-AC3 to J.N.B.).
(46) Dietrich-Buchecker, C. O.; Sauvage, J. P. Chem. Rev. 1987, 87
795–810.
(47) Sauvage, J. P. Acc. Chem. Res. 1990, 23, 319–327.
(48) Collin, J.-P.; Dietrich-Buchecker, C.; Gavina, P.; Jimenez-Molero,
M. C.; Sauvage, J.-P. Acc. Chem. Res. 2001, 34, 477–487.
(49) Dietrich-Buchecker, C.; Jimenez-Molero, M. C.; Sartor, V.; Sauvage,
J.-P. Pure Appl. Chem. 2003, 75, 1383–1393.
Supporting Information Available: Electronic absorption and
NMR spectra of [Cu(dtbp)NCCH3]PF6. Cyclic voltammogram
of 1, and comparison of corrected emission spectrum to that of
(50) Nitschke, J. R. Acc. Chem. Res. 2007, 40, 103–112.
(51) Schultz, D.; Nitschke, J. R. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
11191–11195.
(52) Livoreil, A.; Sauvage, J.-P.; Armaroli, N.; Balzani, V.; Flamigni, L.;
Ventura, B. J. Am. Chem. Soc. 1997, 119, 12114–12124.
(53) Schmittel, M.; Ganz, A. Chem. Commun. 1997, 999–1000.
(54) Schmittel, M.; Michel, C.; Liu, S.-X.; Schildbach, D.; Fenske, D. Eur.
J. Inorg. Chem. 2001, 1155–1166.
(56) Miller, M. T.; Karpishin, T. B. Sens. Actuators, B 1999, B61, 222–224.
(57) Lewis, E. A.; Tolman, W. B. Chem. Rev. 2004, 104, 1047–1076.
(58) Mirica, L. M.; Ottenwaelder, X.; Stack, T. D. P. Chem. Rev. 2004,
104, 1013–1045.
(59) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.
Von Zelewsky, A. Coord. Chem. Rev. 1988, 84, 85–277.
(60) Cunningham, K. L.; McMillin, D. R. Inorg. Chem. 1998, 37
4114–4119.
(55) Kalsani, V.; Bodenstedt, H.; Fenske, D.; Schmittel, M. Eur. J. Inorg.
Chem. 2005, 1841–1849.

5714 Inorganic Chemistry, Vol. 48, No. 13, 2009
Green et al.
[Cu(dtbp)(dmp)]BF4. The reaction of 1 with CO followed by
electronic absorption, FT-Raman and FT-IR spectroscopy, and
FT Raman band assignments. This material is available free
of charge via the Internet at http://pubs.acs.org. Crystallo-
graphic data collection parameters, atomic coordinates, bond
parameters, and spectral parameters for [Cu(dtbp)CO]SbF6.
CCDC 724725 contains the supplementary crystallographic
data; these data may be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.