Luminescent ruthenium tripod complexes: Properties in solution and on conductive surfaces

Article abstract of DOI:10.1021/ic1002868

Two luminescent ruthenium complexes containing tripod-type end groups linked through a rigid spacer to a phenanthroline derivative, able to confer an axial geometry to the complexes, are described. One of the compounds is functionalized with thioacetate groups in order to link the metal complex to metallic surfaces. The photophysical and electrochemical behavior of the complexes are studied in solution and on conductive substrates and, furthermore, self-assembled monolayers are investigated in a junction using gold and an indium gallium eutectic, as electrodes, and by time-resolved confocal microscopy. The results show that the complexes form very stable and well-ordered monolayers because of the tripod system, which can anchor the complex almost perpendicular to the surfaces.

Full text of DOI:10.1021/ic1002868

Inorg. Chem. 2011, 50, 1581–1591 1581
DOI: 10.1021/ic1002868
Luminescent Ruthenium Tripod Complexes: Properties in Solution
and on Conductive Surfaces
Srinidhi Ramachandra,† Klaus C. Schuermann,‡ Fabio Edafe,§ Peter Belser,§ Christian A. Nijhuis,^
William F. Reus,^ George M. Whitesides,^ and Luisa De Cola*,†,‡
†Laboratory of Supramolecular Chemistry and Technology, University of Twente, P.O. Box 217, 7500 AE
‡
€
€
€
Enschede, The Netherlands, Physikalisches Institut, Westfalische Wilhelms-Universitat Munster,
§
ꢀ
Mendelstrasse 7, 48149 Mu€nster, Germany, Department of Chemistry, University of Fribourg, Chemin du Musee 9,
1700 Fribourg, Switzerland, and ^Department of Chemistry and Chemical Biology, Harvard University, 12
Oxford Street, Cambridge, Massachusetts 02138, United States
Received February 10, 2010
Two luminescent ruthenium complexes containing tripod-type end groups linked through a rigid spacer to a
phenanthroline derivative, able to confer an axial geometry to the complexes, are described. One of the compounds
is functionalized with thioacetate groups in order to link the metal complex to metallic surfaces. The photophysical and
electrochemical behavior of the complexes are studied in solution and on conductive substrates and, furthermore, self-
assembled monolayers are investigated in a junction using gold and an indium gallium eutectic, as electrodes, and by
time-resolved confocal microscopy. The results show that the complexes form very stable and well-ordered
monolayers because of the tripod system, which can anchor the complex almost perpendicular to the surfaces.
Introduction
osmium ions coordinated to polypyridyl derivatives and their
assemblies on conductive surfaces.3 The advantages of using
these molecules rely on their rich, often reversible, electro-
chemical properties, which are often localized on the metal ion,
for oxidation processes, and on the ligands, for the reduction
processes, on their photophysical properties, and, in partic-
ular, on the nature of their emitting excited state (triplet) and, as
a consequence, long-lived excited-state lifetimes. Most of the
complexes also exhibit good emission quantum yield, for
In the last decades, a large number of papers dealing with
various types of molecular assemblies on metallic and semi-
conductor surfaces have been published because the anchor-
ing of molecules on surfaces is the first step toward the creation
of molecular devices.1 In particular, stable electro- and photo-
active species possessing a thiol derivative have been inves-
tigated to understand the role played by the chemical structures
in important phenomena such as charge injection and con-
ductivity in the general context of molecular electronics.2 Even
though most of the investigated systems are organic molecules,
more recently some effort has been concentrated in the
use of metal complexes containing ruthenium, iridium, and
(3) (a) Figgemeier, E.; Constable, E. C.; Housecroft, C. E.; Zimmermann,
Y. C. Langmuir 2004, 20, 9242–9248. (b) Silva, M. J. J. P.; Bertoncello, P.;
Daskalakis, N. N.; Spencer, N.; Kariuki, B. M.; Unwin, P. R.; Pikramenou, Z.
Supramol. Chem. 2007, 19, 115–127. (c) Bertoncello, P.; Kefalas, E. T.;
Pikramenou, Z.; Unwin, P. R.; Forster, R. J. J. Phys. Chem. B 2006, 110,
10063–10069. (d) Obeng, Y. S.; Bard, A. J. Langmuir 1991, 7, 195–201.
(e) Zhang, X.; Bard, A. J. J. Phys. Chem. 1988, 92, 5566–5569. (f) Meyer,
T. J.; Meyer, G. J.; Pfennig, B. W.; Schoonover, J. R.; Timpson, C. J.; Wall, J. F.;
Kobusch, C.; Chen, X.; Peek, B. M. Inorg. Chem. 1994, 33, 3952–3964.
(g) Tuccitto, N.; Torrisi, V.; Cavazzini, M.; Morotti, T.; Puntoriero, F.; Quici,
S.; Campagna, S.; Licciardello, A. ChemPhysChem 2007, 8, 227–230. (h) Cattabriga,
M.; Ferri, V.; Tran, E.; Galloni, P.; Rampi, M. A. Inorg. Chim. Acta 2007, 360, 1095–
1101. (i) Weiss, E. A.; Kriebel, J. K.; Rampi, M. A.; Whitesides, G. M. Phil. Trans. R.
Soc., A 2007, 365, 1509–1537. (j) Albrecht, T.; Guckian, A.; Ulstrup, J.; Vos, J. G.
Nano Lett. 2005, 5, 1451–1455. (k) Forster, R. J.; Figgemeier, E.; Lees, A. C.; Hjelm,
J.; Vos, J. G. Langmuir 2000, 16, 7867–7870. (l) Forster, R. J.; Figgemeier, E.;
Loughman, P.; Lees, A.; Hjelm, J.; Vos, J. G. Langmuir 2000, 16, 7871–7875.
(m) Otsuki, J.; Tokimoto, T.; Yano, Y. N. T.; Hasegawa, T.; Chen, X.; Okamoto, Y.
Chem.;Eur. J. 2007, 13, 2311–2319. (n) Terada, K.; Kobayashi, K.; Haga, M.-A.
Dalton Trans. 2008, 4846–4854. (o) Haga, M.-A.; Kobayashi, K.; Terada, K. Coord.
Chem. Rev. 2007, 251, 2688–2701. (p) Haga, M.-A.; Wang, K.; Kato, N.; Monjushiro,
H. Mol. Cryst. Liq. Cryst. 1999, 337, 89–92.
*To whom correspondence should be addressed. E-mail: decola@
uni-muenster.de.
(1) (a) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides,
G. M. Chem. Rev. 2005, 105, 1103–1170. (b) Dong, T. Y.; Huang, C. L.; Chen,
C. P.; Lin, M. C. J. Organomet. Chem. 2007, 692, 5147–5155. (c) Ulman, A.
Chem. Rev. 1996, 96, 1533–1554.
(2) (a) Mativetsky, J. M.; Pace, G.; Elbing, M.; Rampi, M. A.; Mayor, M.;
Samori, P. J. Am. Chem. Soc. 2008, 130, 9192–9193. (b) Reichert, J.; Ochs, R.;
Beckmann, D.; Weber, H. B.; Mayor, M.; Lohneysen, H. v. Phys. Rev. Lett. 2002,
88, 176804–176807. (c) Aviram, A.; Ratner, M. A. Chem. Phys. Lett. 1974, 29,
277–283. (d) Mirkin, C. A.; Ratner, M. A. Annu. Rev. Phys. Chem. 1992, 43,
719–754. (e) Dichtel, W. R.; Heath, J. R.; Fraser Stoddart, J. Philos. Trans. R.
Soc., A 2007, 365, 1607–1625. (f) Joachim, C.; Gimzewski, J. K.; Aviram, A.
Nature 2000, 408, 541. (g) Huang, Z. F.; Chen, F.; Bennett, P. A.; Tao, N. J.
J. Am. Chem. Soc. 2007, 129, 13225–13231. (h) Chen, F.; Hihath, J.; Huang, Z.;
Li, X.; Tao, N. J. Annu. Rev. Phys. Chem. 2007, 58, 535–564.
€
r
2010 American Chemical Society
Published on Web 12/31/2010
pubs.acs.org/IC

1582 Inorganic Chemistry, Vol. 50, No. 5, 2011
Ramachandra et al.
iridium even close to unity, and excellent stability.4 In addition,
most of these systems have lately been fully investigated as
materials for electroluminescent devices because their lumi-
nescent excited state can be electrically populated.5 Their
properties as organized monolayers have, however, received
much less attention than many organic systems. The lack of
data is related to the difficulties to anchor the luminescent
metal complexes in a well-organized monolayer on conduc-
tive surfaces, mainly because of their, in most cases, octahe-
dral geometry and therefore arrangement of their chelating
ligands, and to the synthetic difficulties to introduce thiol
groups.
Solvents were distilled from the appropriate drying agents. The
palladium catalysts were purchased from Strem. All other reagents
were obtained exclusively from Fluka, Aldrich, and Acros.
Chromatography. Thin-layer chromatography was performed
using aluminum sheets precoated with silica gel 60 F254 purchased
from Merck. Preparative plates were made by using glass sheets
precoated with silica gel 60 F254 with a layer thickness of 2 mm
purchased from Merck. Column chromatography was carried
out using silica gel 60, 230-400 mesh, from Chemie Brunschwig
AG and a neutral aluminum oxide gel from Fluka.
NMR. 1H and 13C NMR spectra were obtained with a Bruker
1
Avance DRX-360 (360.13 MHz for H) or a Bruker Avance
DRX-400 (400.13 MHz for 1H and 100.62 MHz for 13C) spec-
trometer. Chemical shifts (δ) are given in parts per million and
coupling constants (J) in Hertz, using the solvent itself as an
internal standard. Assignment of the 1H and 13C NMR signals
was performed by COSY and DEPT techniques.
The requirements to have good coverage and control of the
distance of the luminescent center from the substrate are not
easy to fulfill because the large headgroup containing the
metal complex must somehow match the anchoring area in
order to align the molecules as rod-type systems and the need
of rigid ligands, which could hold the metal complexes perpen-
dicular to the substrates and dictate the distance, also demands
a rather long synthesis. The most commonly employed ligands
to bind molecules to conductive substrates, through both cova-
lent and noncovalent interactions, include amino, thiol, and
carboxylic groups as anchoring units.6 Multisite binding is
also desirable from the application point of view because the
stability of these assemblies is of critical importance. Recently,
few groups have reported the use of tripodal molecules in order
to strongly bind chromophores to the surface and to confer a
better arrangement for the molecule to stand on the substrate.7
Following the same rationale, we have employed a rigid
tripodal ligand, functionalized with thioacetyl groups at each of
the three legs, coordinated to ruthenium(II) to form a lumi-
nescent complex. In this paper, we report the synthesis and
photophysical characterization of tripodal complexes, with
and without anchoring groups, both in solution and on the
surface and, for the thioacetyl derivative, self-assembled
monolayer (SAM) formation on conductive metallic sub-
strates. We have also examined the electrochemical behavior
of these assemblies in solution and as SAMs. In addition, we
also report the preliminary results on the conductance measure-
ments of the ruthenium monolayer between two electrodes
using a setup containing an eutectic alloy of indium and gallium
(EGaIn) as one of the electrodes and a gold surface as the other
one, to demonstrate the possible application of these mole-
cules in molecular electronics.
MS. Mass spectra were recorded either on a Vacuum Gen-
erators Micromass VG 70/70E (fast atom bombardment ioniza-
tion; nitrobenzyl alcohol or dithranol AgOTf matrix of the
sample) or on a HP 5988A Quadrupol (electron impact ioniza-
tion, 70 eV) mass spectrometer. Electrospray ionization (ESI)
and high-resolution (HR) mass spectrometry (MS) spectra
were recorded on a Bruker FTMS 4.7T BioAPEXII spec-
trometer.
Synthesis of Ru-SAc. A new polypyridyl ligand, 2-{4-[2-(tri-
methylsilyl)ethynyl]phenyl}-1H-imidazo[4,5-f][1,10]phenanthroline
(TMS-EPIP), its ruthenium(II) complex, [Ru(bpy)2TMS-EPIP]2þ
(bpy = 2,20-bipyridine), and the target system Ru-SAc have been
synthesized and characterized. The corresponding compound
Ru-tert-Bu was prepared in a way similar to that of the caltrop with
thioacetate functions.
1,10-Phenanthroline-5,6-dione,8 cis-[Ru(bpy)2Cl2] 2H2O,9
3
and a trithiolacetate tripod base10 were prepared according to
literature procedures. Other reagents were purchased commer-
cially from Fluka, Aldrich, and Acros and used without further
purification unless otherwise noted.
TMS-EPIP. A mixture of 1,10-phenanthroline-5,6-dione
(0.53 g, 2.50 mmol), 4-[(trimethylsilyl)-ethynyl]benzaldehyde
(0.71 g, 3.50 mmol), ammonium acetate (3.88 g, 50.00 mmol),
and glacial acetic acid (15 mL) was refluxed for 4 h and then
cooled to room temperature (298 K). It was diluted with water,
and the dropwise addition of concentrated aqueous ammonia gave
a yellow precipitate, which was collected, washed with water, and
dried. The crude product obtained was purified by recrystalliza-
tion from CHCl3/MeOH (4:1, v/v) and dried. Yield: 0.71 g
(72%). 1H NMR [(CD3)2SO, 360 MHz]: δ 13.88 (br, 1H, NH),
9.04 (dd, 2H, Ha), 8.92 (dd, 2H, Hc), 8.30 (d, 2H, He), 7.84
(q, 2H, Hb), 7.71 (d, 2H, Hd), 0.27 (s, 9H, Si(CH3)3). 13C NMR
[(CD3)2SO, 100 MHz]: δ 149.62, 147.94, 143.70, 132.25, 130.14,
129.58, 126.22, 123.31, 122.90, 104.82, 96.10, -0.17. ESI-MS. Calcd
for C24H20N4Si: m/z 392.15. Found: m/z 393.15 [M þ H]þ.
Experimental Section
General Procedures. Reactants. All reactions were carried
out under an argon atmosphere and in oven-dried glassware.
Ru(bpy)2TMS-EPIP. A mixture of cis-[ Ru(bpy)2Cl2] 2H2O
3
(0.5 mmol, 0.260 g), TMS-EPIP (0.5 mmol, 0.196 g), EtOH
(10 mL), and water (5 mL) was refluxed under argon for 2 h to
give a clear red solution. After most of the EtOH solvent was
removed under reduced pressure, a red precipitate was obtained
by the dropwise addition of a saturated aqueous NH4PF6 solution.
The product was purified by column chromatography on alumina
using acetonitrile/toluene (1:1, v/v) as the eluent and then dried
in vacuo. Yield: 0.345 g (63%). 1H NMR [(CD3)2SO, 400 MHz]:
δ 14.41 (br, 1H, NH), 9.09 (dd, 2H, Hc), 8.88 (d, 2H, H3), 8.84
(4) Aldachi, C. B., M. A.; Forrest, M. R.; Thompson, M. E. Appl. Phys.
Lett. 2000, 77, 904–906.
(5) Yersin, H. Highly Efficient OLEDs with Phosphorescent Materials;
Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2008; p 438.
(6) (a) Fox, M. A.; Whitesell, J. K.; McKerrow, A. J. Langmuir 1998, 14,
816–820. (b) Kitagawa, T.; Idomoto, Y.; Matsubara, H.; Hobara, D.; Kakiuchi, T.;
Okazaki, T.; Komatsu, K. J. Org. Chem. 2006, 71, 1362–1369. (c) Park, J. S.; Vo,
A. N.; Barriet, D.; Shon, Y. S.; Lee, T. R. Langmuir 2005, 21, 2902–2911.
(7) (a) Myahkostupov, M.; Piotrowiak, P.; Wang, D.; Galoppini, E.
J. Phys. Chem. C 2007, 111, 2827–2829. (b) Galoppini, E.; Guo, W.; Zhang,
W.; Hoertz, P. G.; Qu, P.; Meyer, G. J. J. Am. Chem. Soc. 2002, 124, 7801–7811.
(c) Hoertz, P. G.; Carlisle, R. A.; Meyer, G. J.; Wang, D.; Piotrowiak, P.;
Galoppini, E. Nano Lett. 2003, 3, 325–330. (d) Nikitin, K.; Lestini, E.; Lazzari,
M.; Altobello, S.; Fitzmaurice, D. Langmuir 2007, 23, 12147–12153. (e) Ichimura,
A. S.; Lew, W.; Allara, D. L. Langmuir 2008, 24, 2487–2493. (f) Weidner, T.; Kramer,
A.; Bruhn, C.; Zharnikov, M.; Shaporenko, A.; Siemeling, U.; Trager, F. Dalton Trans.
2006, 2767–2777.
0
0
(d, 2H, H3 ), 8.33 (d, 2H, He), 8.22 (t, 2H, H4), 8.11 (t, 2H, H4 ),
(8) Amouyal, E.; Homsi, A.; Chambron, J.-C.; Sauvage, J.-P. Dalton
Trans. 1990, 1841–1845.
(9) Lay, P. A.; Sargeson, A. M.; Taube, H.; Chou, M. H.; Creutz, C. In
Inorganic Syntheses; Jean'ne, M. S., Ed.; 2007; Vol. 28, pp 291-299.
(10) Jian, H.; Tour, J. M. J. Org. Chem. 2003, 68, 5091–5103.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1583
8.07 (dd, 2H, Ha), 7.94 (q, 2H, Hb), 7.85 (d, 2H, H6), 7.76
7.38 (d, 6H, J = 8.4), 3.89 (q, 2H, J = 7.0), 1.33 (s, 27H), 1.26
(t, 3H, J = 7.1). 13C NMR (CD2Cl2, 100 MHz): δ 151.9, 135.3,
134.0, 131.6, 131.0, 125.5, 120.2, 91.0, 60.1, 35.0, 31.3, 18.6.
MALDI-MS (matrix = DCTB). Calcd for C56H56IOSi:
m/z 772.41. Found: m/z 772.42.
0
0
(d, 2H, Hd), 7.6 (m, 4H, H6 and H5), 7.35 (t, 2H, H5 ), 0.28
(s, 9H, Si(CH3)3). 13C NMR [(CD3)2SO, 100 MHz]: δ 156.76,
156.55, 151.43, 151.35, 149.87, 145.11, 137.93, 137.78, 132.44,
130.42, 128.86, 128.16, 127.85, 127.71, 126.62, 126.33, 125.27,
124.43, 124.34, 123.63, 104.62, 96.57, -0.17. ESI-MS. Calcd
for C44H36F12N8P2RuSi: m/z 1096.12. Found: m/z 805.18
[M - 2PF6 þ e-]þ, 951.15 [M - PF6]þ.
Tris[4-(4-tert-butylphenylethynyl)phenyl]-40-iodophenylsilane
(Iodophenyl-tert-Bu). A solution of Li[C6H4I] was prepared first
by the addition of 1.6 M LiBun in hexane (8.9 mL, 1.11 equiv) to
p-diiodobenzene (0.495 g, 1.5 mmol) in diethyl ether (10 mL) and
stirred for 1 h at room temperature under argon. The resulting
solution was transferred dropwise to a solution of ethoxy-tert-Bu
(1.00 g, 1.28 mmol) in 20 mL of dry pentane. The resulting
mixture was stirred at room temperature overnight and then poured
into H2O, and the mixture was extracted with CH2Cl2. The com-
bined organic layer was dried over MgSO4 and filtered, and the
solvent was removed in vacuo. The residue was purified by column
chromatography to afforda slightly yellow clear oil (0.655 g, 55%).
1H NMR (CD2Cl2, 400 MHz): δ 7.79 (d, 2H, J = 8.1), 7.57-
7.7.52 (m, 12H), 7.48 (d, 6H, J = 8.2), 7.41 (d, 6H, J = 8.4), 7.30
(d, 2H, J = 8.2), 1.33 (s, 27H). 13C NMR (CD2Cl2, 100 MHz):
δ 152.60, 138.43, 137.79, 136.75, 133.74, 133.37, 131.87, 131.47,
126.05, 125.68, 120.42, 97.92, 91.45, 88.94, 35.28, 31.44. MAL-
DI-MS (matrix = DCTB). Calcd for C60H55OSi: m/z 930.31.
Found: m/z 930.32.
Ru(bpy)2EPIP. To a stirring solution of [Ru(bpy)2TMS-
EPIP](PF6)2 (0.164 mmol, 0.180 g) in tetrahydrofuran (THF;
10 mL) was added at room temperature a solution of K2CO3
(0.5-1 equiv, 12 mg) in MeOH (10 mL). After the reaction mixture
was stirred at room temperature for 4 h, it was filtered and the
solvent was removed in vacuo. A red precipitate was obtained by
dropwise addition of a saturated aqueous NH4PF6 solution.
The product was purified by column chromatography on alu-
mina using acetonitrile/toluene (1:1, v/v) as the eluent and then
dried in vacuo. Yield: 0.143 g (85%). 1H NMR [(CD3)2SO,
400 MHz]: δ 14.45 (br, 1H, NH), 9.09 (d, 2H, Hc), 8.88 (d, 2H, H3),
0
8.85 (d, 2H, H3 ), 8.33 (d, 2H, He), 8.22 (t, 2H, H4), 8.10 (m, 4H,
0
H4 þ Ha), 7.93 (q, 2H, Hb), 7.85 (d, 2H, H6), 7.77 (d, 2H, Hd),
0
0
7.60 (m, 4H, H6 and H5), 7.35 (t, 2H, H5 ), 4.43 (s, 1H,
acetylenic). 13C NMR [(CD3)2SO, 100 MHz]: δ 156.78, 156.57,
151.47, 151.39, 149.91, 145.14, 137.96, 137.81, 132.57, 130.46,
129.82, 127.89, 127.75, 126.69, 126.36, 124.46, 124.38, 123.31,
83.07, 82.94, 30.69. ESI-MS. Calcd for C41H28F12N8P2Ru:
m/z 1024.08. Found: m/z 733.14 [M - 2PF6 þ e-]þ, 879.11
[M - PF6]þ.
[Ru(bpy)2EPIP](PF6)2-Si-Tripod-tert-Bu] (Ru-tert-Bu). An
oven-dried screw-cap tube was charged with tripod-tert-Bu
(0.150 g, 0.146 mmol), palladium catalyst bis(dibenzylidene-
acetone)palladium(0) (5.6 mg, 0.01 mmol), CuI (1.9 mg,
0.01 mmol), and PPh3 (10.3 mg, 0.039 mmol). The tube was capped
with a septum, evacuated, and backfilledwithargonthree times. N,
N-Diisopropylethylamine (5 mL) was added via a syringe. A
solution of [Ru(bpy)2EPIP](PF6)2 (200 mg, 0.195 mmol) in THF
(5 mL) was transferred via a cannula to the tube. The tube was
then capped with its screw cap, and the solution was stirred at
room temperature for 2 days. The solvent was removed under
reduced pressure to give a solid residue, which was dissolved in
water (10 mL), NH4PF6 (0.5 g) was added, and the resulting red
precipitate was isolated by suction filtration. The red precip-
itate was first purified by column chromatography on alumina
(1:1 acetonitrile/toluene) to remove the unreacted tripod-OMe
and then eluted with a gradient of acetonitrile/water/saturated
aqueous KNO3 from 100:1:1 to 100:18:2. The collected orange
fractions were combined and evaporated. An aqueous solution
of saturated ammonium hexafluorophosphate was added. Fil-
tration under vacuum with Celite, washing with H2O, and
dissolution in acetone afforded the desired Ru-tripod as a red
powder. Yield: 0.148 g (45%). 1H NMR (DMSO-d6, 400 MHz):
δ 14.55 (br, 1H), 9.12 (d, 2H, J = 8.1), 8.85 (d, 2H, J = 8.3), 8.81
(d, 2H, J = 8.1), 8.39 (d, 2H, J = 8.3), 8.21 (t, 2H, J = 7.8),
8.12-8.06 (m, 2H), 7.88 (d, 2H, J = 8.6), 7.84 (d, 2H, J = 5.0),
7.71 (d, 2H, J = 8.1), 7.67-7.45 (m, 34H), 7.34 (d, 2H, J = 6),
1.29 (s, 27H).
Ru-SAc. An oven-dried screw-cap tube was charged with a
trithiolacetate tripod base (0.146 mmol, 0.150 g), palladium
catalyst Pd(dba)2 [dba = bis(dibenzylideneacetone)palladium;
3-5 mol %, 4.9 mg], CuI (3-5 mol %, 1.2 mg), and PPh3 (12-
20 mol %, 7.7 mg). The tube was capped with a septum,
evacuated, and backfilled with argon three times. Triethylamine
(5 mL) was added via a syringe. A solution of [Ru(bpy)2EPIP]-
(PF6)2 (0.146 mmol, 0.150 g) in THF (5 mL) was transferred via
a cannula to the tube. The tube was then capped with its screw
cap, and the solution was stirred at room temperature for 3 days.
After most of the solvent was removed under reduced pressure,
an orange precipitate was obtained by the dropwise addition of a
saturated aqueous NH4PF6 solution. The product was purified
by column chromatography on silica, eluting with a gradient of
acetonitrile/water/saturated aqueous KNO3 from 100:1:1 to
100:18:2. The collected orange fractions were combined and
dissolved in acetone. An aqueous solution of saturated ammonium
hexafluorophosphate was added, and acetone was evaporated.
Filtration under vacuum and washing with H2O of the resulting
precipitate afforded the desired caltrop (0.18 g, 64%) as an
orange solid. 1H NMR [CD2Cl2, 400 MHz]: δ 9.16 (d, 2H),
8.47 (d, 2H,), 8.43 (d, 2H), 8.33 (d, 2H), 8.11 (t, 2H), 7.80 (t, 2H),
7.93 (d, 2H), 7.80 (m, 6H), 7.70 (d, 2H), 7.60 (m, 16H), 7.51
(m, 5H), 7.43 (m, 3H), 7.30 (m, 8H), 4.11 (s, 6H), 2.35 (s, 9H).
ESI-MS. Calcd for C98H70F12N8O3P2RuS3Si: m/z 1922.28.
Found: m/z 816.18 ([M - 2PF6]2þ)/2.
Synthesis of Ru-tert-Bu. Ethoxytris[4-(4-tert-butylphenyl-
ethynyl)phenyl]silane (Ethoxy-tert-Bu). To an oven-dried glass
vessel containing 4-tert-butylphenylacetylene (2.4 mL, 13.2 mmol),
ethoxytris(p-iodophenyl)silane (1.53 g, 2.2 mmol), Pd(dba)2
(0.19 g, 0.33 mmol), copper iodide (0.06 g, 0.3 mmol), and
triphenylphosphine (0.29 g, 1.1 mmol) were added 150 mL of
THF and 50 mL of triethylamine, which was filled with argon. The
mixture was stirred at room temperature for 2 days. The reaction
mixture was then poured into water, and the aqueous layer was
extracted with ethyl acetate three times. The combined organic
solution was washed with water and dried over magnesium sulfate.
The solvent was removed in vacuo, and the residues were purified by
column chromatography on silica gel (hexanes/CH2Cl2, 5:1) to give
ethoxy-tert-Bu as a yellow sticky oil (1.65 g, 95%). 1H NMR
(CD2Cl2, 400 MHz): δ 7.58-7.47 (m, 12H), 7.48 (d, 6H, J = 8.4),
HR ESI-MS. Calcd for C101H82N8RuSi: m/z 768.27454.
Found: m/z 768.27429 ([M - 2PF6]2þ/2).
13C NMR (CD3CN, 500 MHz): δ138.77, 137.43, 137.35,
135.38, 134.45, 133.44, 133.12, 133.09, 132.80, 132.72, 132.35,
132.13, 131.97, 131.66, 130.40, 129.79, 129.70, 128.62, 128.48,
127.86, 127.14, 126.77, 126.08, 125.81, 125.41, 125.31, 125.23,
120.76, 92.20, 91.86, 91.03, 89.25, 31.40.
Photophysics. Absorption spectra were measured on a Varian
Cary 5000 double-beam UV-vis-near-IR (NIR) spectrometer
and baseline-corrected. Steady-state emission spectra were re-
corded on a Horiba Jobin-Yvon IBH FL-322 Fluorolog 3 spec-
trometer equipped with a 450 W xenon arc lamp, double-grating
excitation and emission monochromators (2.1 nm/mm disper-
sion; 1200 grooves mm-1), and a Hamamatsu R928 photomulti-
plier tube or a TBX-4-X single-photon-counting detector. Emission
spectra were corrected for the source intensity (lamp and grating)
and emission spectral response (detector and grating) by standard

1584 Inorganic Chemistry, Vol. 50, No. 5, 2011
Ramachandra et al.
correction curves. For monolayers, the sample was mountedona
commercially available solid-state sample holder provided by
Horiba Jobin-Yvon, and the emission was collected with the front-
face geometry. Time-resolved measurements were performed using the
time-correlated single-photon-counting option on the Fluorolog 3.
NanoLEDs [402 nm; full width at half-maximum (fwhm) <
750 ps] with repetition rates between 10 kHz and 1 MHz were
used to excite the sample. The excitation sources were mounted
directly on the sample chamber at 90ꢀ to a double-grating emission
monochromator (2.1 nm/mm dispersion; 1200 grooves mm-1) and
collected by a TBX-4-X single-photon-counting detector. The
photons collected at the detector are correlated by a time-
to-amplitude converter (TAC) to the excitation pulse. Signals
were collected using an IBH DataStation Hub photon-counting
module, and data analysis was performed using the commercially
available DAS6 software (Horiba Jobin Yvon IBH). The good-
ness of fit was assessed by minimization of the reduced χ2
function and visual inspection of the weighted residuals.
Quantum Yield. Luminescence quantum yields (Φem) were
measured in optically dilute solutions (O.D. < 0.1 at excitation
wavelength) and compared to reference emitters by the following
equation:
Surface Analysis. Atomic Force Microscopy (AFM) Imaging.
AFM images of the monolayers on flat gold substrates were
acquired in air at room temperature with a commercial instrument
(Digital Instruments, Nanoscope IIIa, Dimension 3000, Santa
Barbara, CA) operating in tapping mode. AFM images are flat-
tened and shown without further modification. Analysis was
performed using WSxM 4.0 Develop.12
Fluorescence Lifetime Microscopy (FLIM). FLIM images and
the fluorescence decays on surfaces were recorded using a Micro-
time 200 (PicoQuant) attached to an Olympus IX 71 microscope
with a 100ꢀ oil-immersion objective and a scanning speed of 6 μs
per point at excitation with a 440 nm laser (fwhm 80 ps). Fluores-
cent lifetimes were calculated from the whole area by the software
SymphoTime (PicoQuant).
Formation of the Junctions. Ultraflat gold surfaces were formed
by a template-stripping (TS) procedure published previously.13 All
of the details can be found in ref 14, but a brief description is given
here. On silicon wafers with their native SiO2 layer present, a layer
of 500 nm of gold was thermally deposited by electron beam
(e-beam) at (2-3) ꢀ 10-6 Torr at a rate of 8-10 A s-1. Glass
˚
slides, which were cleaned by washing with EtOH and by a plasma
of air (500 Torr, 5 min), of typically 1 cm2 were glued at the silver
surface using an optical adhesive (Norland, No. 61). The optical
adhesive was cured by exposure to UV light for 2 h. The glass sub-
strates were cleaved off the silicon wafer by using a razor blade, after
which the TS gold substrates were immersed in a solution of 2 mM
Ru-SAc in EtOH/acetonitrile for 24 h at room temperature. After
SAM formation, the samples were rinsed with EtOH.
"
#
ꢀ
ꢁꢀ
ꢁ
ꢀ
ꢁ
2
2
ArðλrÞ IrðλrÞ nx
AxðλxÞ IxðλxÞ nr
Dx
Dr
Φx ¼ Φr
where A is the absorbance at the excitation wavelength (λ), I is
the intensity of the excitation light at the excitation wavelength
(λ), n is the refractive index of the solvent, D is the integrated
intensity of the luminescence, and Φ is the quantum yield. The
subscripts r and x refer to the reference and sample, respectively.
All quantum yields were performed at identical excitation wave-
lengths for the sample and reference, canceling the I(λr)/I(λx) term
in the equation. The quantum yield of the reference compound,
Ru(bpy)3, was obtained from the literature (Φ = 0.016).11
Deaerated samples were prepared by the freeze-pump-thaw
technique.
Conical-shaped EGaIn alloy (75.5% gallium and 24.5% indium
by weight, 15.7 ꢀC melting point, with a surface layer of Ga2O3)
was used as the top electrode. A detailed description of the
formation and contact of the SAMs by Ga2O3/EGaIn top elec-
trodes has been reported previously.15,16 Ga2O3/EGaIn is a non-
Newtonian fluid. On the micrometer scale, Ga2O3/EGaIn behaves
as a solid, but when sheer pressure is applied, Ga2O3/EGaIn
behaves as a liquid. Ga2O3/EGaInwill flowuntilthe sheer pressure
is relieved. This behavior, unlike mercury, allows one to shape
Ga2O3/EGaIn into nonspherical shapes. A drop of Ga2O3/EGaIn
hanging at a 26S-guage needle was brought into contact with a
surface that is wettable by Ga2O3/EGaIn, such as poly(dimethyl-
siloxane), glass, or silver surfaces. Ga2O3/EGaIn adheres both
to the surface and to the needle. Slowly retracting the needle
from the EGaIn drop, by using a micromanipulator, deforms the
EGaIn drop in such a way that two conical-shaped Ga2O3/EGaIn
structures connected head-to-head arise. Further retraction of the
needle results in separation of the conical-shaped Ga2O3/EGaIn
structures, one at the needle and one at the surface. Subsequently,
the substrate was discarded and replaced by a TS silver surface
with the SAM of interest, and the conical-shaped Ga2O3/EGaIn
at the needle was brought into contact with the SAM.
Cyclic Voltammetry (CV). CV was performed in a gastight
single-compartment three-electrode cell using a Voltalab 40 system
from Radiometer Analytical that consists of a PGZ301 potentio-
stat and Voltamaster 4 software. The working electrode was a
1 mm platinum disk, the counter electrode was a platinum wire, and
silver wire was used as a pseudoreference electrode. All glass-
ware was dried prior to use. The compounds (electrolyte, analyte,
and reference) were placed in a Schlenk flask, which was then
evacuated and heated with a heat gun to eliminate any moisture
and oxygen that had entered during the addition. The flask was
then evacuated and filled three times with dry N2(g). The solvent
was added via syringe directly to the sealed Schlenk flask and
then degassed for 10 min with a gentle stream of dry N2. After
degassing, the solution was added, via a syringe, to the electro-
chemical cell under a positive N2 pressure and the electrodes
were then added. The solution was kept under a positive N2 pres-
sure during the measurements, but no flow was allowed through
the cell. For electrochemistry of the surfaces, silver and platinum
wires were used as reference and counter electrodes, respectively.
The reason for employing the platinum electrode was the fact that
the electrochemical window of gold is not large enough for the
ruthenium complex. The surface coverage on platinum is pre-
sented in the Supporting Information. Subsequent experiments
were carried out using a neat acetonitrile solution with tetra-
butylammonium hexafluorophosphate (TBAPF6; Sigma Aldrich)
as the electrolyte (0.1 M). Measurements with both gold and
indium/tin oxide (ITO) functionalized with Ru-SAc were per-
formed in the same experimental conditions as mentioned
above.
Results and Discussion
All of the ligands and ruthenium(II) complexes as well as
their abbreviations are shown in Scheme 1.
Synthesis of the Trithiolacetate Tripod Base. This molec-
ular base with thiol anchoring groups is prepared in the
same way as that described by Jian and Tour10 with a
(12) Horcas, I.; Fernandez, R.; Gomez-Rodriguez, J. M.; Colchero, J.;
Gomez-Herrero, J.; Baro, A. M. Rev. Sci. Instrum. 2007, 78, 013705–8.
(13) Weiss, E. A.; Kaufman, G. K.; Kriebel, J. K.; Li, Z.; Schalek, R.;
Whitesides, G. M. Langmuir 2007, 23, 9686–9694.
(14) Weiss, E. A.; Chiechi, R. C.; Kaufman, G. K.; Kriebel, J. K.; Li, Z.;
Duati, M.; Rampi, M. A.; Whitesides, G. M. J. Am. Chem. Soc. 2007, 129,
4336–4349.
(15) Chiechi, R. C.; Weiss, E. A.; Dickey, M. D.; Whitesides, G. M.
Angew. Chem., Int. Ed. 2008, 47, 142–144.
(16) Nijhuis, C. A.; Reus, W. F.; Whitesides, G. M. J. Am. Chem. Soc.
2009, 131, 17814–17827.
(11) Issberner, J.; Vogtle, F.; De Cola, L.; Balzani, V. Chem.;Eur. J. 1997, 3,
706.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1585
Scheme 1. Schematic Formulas of All of the Complexes Investigated and Their Precursorsa
a The molecules will be named with the abbreviation shown under each formula.
modification in the preparation of the tris[4-[3-(tert-
butyldimethylsilanyloxymethyl)phenylethynyl]phenyl]-
40-iodophenylsilane in which dry pentane was used as the
solvent instead of THF.

1586 Inorganic Chemistry, Vol. 50, No. 5, 2011
Ramachandra et al.
Synthesis of the tert-Butyl Tripod Base. The construc-
tion of the molecular base without anchoring groups starts
with the synthesis of ethoxytris(p-iodophenyl)silane17 fol-
lowed by a Sonogashira coupling reaction of this molecule
with 3 equiv of 4-tert-butylphenylacetylene. The desired
molecule is afforded by lithiation of 1,4-diiodobenzene
and the addition of 4-iodophenyllithium to ethoxytris-
[4-(4-tert-butylphenylethynyl)phenyl]silane.
A previous description of the synthesis of a similar
tripod base in the literature10 mentioned some problems
during the addition of 4-iodophenyllithium (prepared in
situ in dry ether) to ethoxytris[4-[3-(TBDMSO)phenyl-
ethynyl)phenyl]silane (dissolved in dry THF) or also
by the reverse addition (TBDMSO = tert-butyldimethyl-
silyloxy). The preparation of tris[4-[(3-(TBDMSO)methyl)-
phenylethynyl]phenyl]-40-iodophenylsilane as described10
gave a very poor yield, which is far from the predicted 88%
yield. Finally, we solved that problem by using dry pentane as
the solvent in the reaction between 4-iodophenyllithium and
ethoxytris[4-(4-tert-buylphenylethynyl)phenyl]silane.
Under such conditions, we got an acceptable yield of
roughly 50%.
Synthesis of Ruthenium(II) Complexes. Ruthenium(II)
complexes containing the TMS-EPIP ligand were pre-
pared by the direct reaction of yellow TMS-EPIP with the
appropriate mole ratios of the purple Ru(bpy)2Cl2 in
MeOH/water for 2 h. The desired orange ruthenium(II)
complex was isolated as its hexafluorophosphate and was
purified by column chromatography. [Ru(bpy)2EPIP]-
(PF6)2 was obtained by removal of the trimethylsilyl
protecting group from [Ru(bpy)2TMS-EPIP](PF6)2 with
K2CO3 in MeOH. The Ru-SAc and Ru-tert-Bu complexes
have been synthesized by a Sonogashira cross-coupling
reaction between [Ru(bpy)2EPIP](PF6)2 and respectively
the trithiolacetate tripod base10 or tert-butyl tripod base in
the presence of Pd(dba)2 palladium catalysts [Pd(dba)2 =
bis(dibenzylideneacetone)palladium(0)]. See the Experimen-
tal Section for the preparation of all of the precursors and the
full characterization of the molecules.
Figure 1. Absorption spectra of Ru-SAc (red line), Ru-tert-Bu (blue
line), Ru-(bpy)2EPIP (green line), and Ru-(bpy)2TMS-EPIP (red line) in
acetonitrile solutions.
orbitals of the ruthenium and the π* orbitals of the bipyridine
and phenanthroline ligands.
Because of the similar energy levels involving the bipyr-
idine and phenanthroline ligands and the broadness of the
spectra, it is difficult to attribute the lowest excited state
from the absorption characteristics. We can therefore
expect that for complexes containing the tripod with thio-
acetate and tert-butyl groups the lowest MLCT involves the
bipyridine ligands or the substituted phenanthroline. On
the other hand, for Ru(bpy)2EPIP and Ru(bpy)2TMS-EPIP,
because of higher conjugation and the lack of electron-
donating groups, we expect the lowest MLCT absorption
on the phenanthroline moiety. Such a hypothesis is sup-
ported by the emission spectra and electrochemistry data of
the complexes.
The room temperature emission spectra of the reference
compound, Ru-tert-Bu, and the Ru-SAc complex were
recorded in acetonitrile solutions upon excitation at
452 nm (Figure 2a). The emission spectra have broad
structureless bands centered at 609 and 614 nm for
Ru-SAc and Ru-tert-Bu, respectively. These bands are
attributed to the radiative decay of the triplet metal-
to-ligand charge-transfer (3MLCT) state of the ruthe-
nium complexes to the ground state. The red shift of the
emission band for the Ru-tert-Bu complex versus the
Ru-SAc complex can be due to the slightly electron-donating
nature of the tert-butyl groups, which can then increase the σ
donation of phenanthroline, favoring the involvement of
Photophysical Characterization in Solution. The absorp-
tion spectrum of Ru-SAc in an acetonitrile solution is
shown in Figure 1. The high-energy band at 291 nm (ε =
1.08 ꢀ 105 M-1 cm-1) can be assigned to the bipyridine
and phenanthroline singlet intraligand (1IL) π-π* transi-
tions, while the bands at 312 nm (ε = 8.1 ꢀ 104 M-1 cm-1
)
and 341 nm (ε = 3.7 ꢀ 104 M-1 cm-1) are due to the
π-π* absorptions of the highly conjugated moieties
containing the phenyleneethylene groups. In particular,
the lowest-energy band is due to 2-[4-(2-ethynyl)phenyl]-
1H-imidazo[4,5-f][1,10]phenanthroline,18 while the 312 nm
feature is attributed to the tripod species. This assignment
is corroborated by the observation that, for the two
reference non-tripodal complexes Ru(bpy)2EPIP and
Ru(bpy)2TMS-EPIP, this 312 nm band is missing (see
Figure 1). The lowest-energy bands around 458 nm (ε=
1.5 ꢀ 104 M-1 cm-1) are assigned to singlet metal-to-ligand
charge-transfer (1MLCT) transitions, which are typical
for ruthenium polypyridyl complexes, involving the d
3
the bpy ligands in the MLCT lowest excited state. The
excited-state lifetimes for both complexes are very similar
in acetonitrile solutions (see Table 1), and in deaerated
solutions, both compounds show much longer excited-
state lifetimes, as expected because of the triplet character
of the lowest excited states. The emission quantum yields
2þ
of the complexes were measured using the Ru(bpy)3
complex as the reference (Φem = 0.016)11 in aerated
acetonitrile, and the values are reported in Table 1. In the
deaerated solutions, as expected, the emission quantum
yields of the complexes are higher than those for analogous
aerated complexes. The precursor complexes also emit at
room temperature, and their emission spectra are quite
similar to those of the other compounds. In the same
solvent, they show emission maxima that are slightly red-
shifted compared with those of the tripod systems, λ = 617
(17) Yao, Y.; Tour, J. M. J. Org. Chem. 1999, 64, 1968–1971.
(18) Huang, W. Y.; Gao, W.; Kwei, T. K.; Okamoto, Y. Macromolecules
2001, 34, 1570–1578.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1587
Figure 2. (a) Emission spectra of Ru-SAc (red line), Ru-tert-Bu (blue line), Ru(bpy)2EPIP (light-green line), and Ru(bpy)2TMS-EPIP (dark-green line) in
acetonitrile solutions. (b) Emission spectra measured at 77 K in butyronitrile glass. λex = 452 nm.
Table 1. Photophysical Data in an Acetonitrile Solution, Unless Otherwise Specified, for All of the Complexes Investigated
emission
room temperature
77 K
complex
Ru-SAc
Ru-tert-Bu
Ru(bpy)2EPIP
Ru(bpy)2TMS-EPIP
Ru-SH on gold
Ru-SAc on glass
λ
max (nm)
φa
φb
τ ( μs)a
τ (ns)b
λ
max (nm)c
τ ( μs)c
609
614
617
621
619
615
0.12
0.14
0.08
0.10
0.014
0.018
0.015
0.020
1.04
1.14
1.08
1.05
2.0
155
196
178
203
8.0
579, 626
5.2
5.6
5.3
5.2
580, 628
586, 635
585, 637
0.09
812
a In a degassed solution. b In an air-equilibrated solution. c In butyronitrile glass. For lifetime measurements, a 402 nm laser diode was used as the
excitation source. For lifetimes on gold and glass, a 440 nm laser excitation was used.
and 621 nm for Ru(bpy)2EPIP and Ru(bpy)2TMS-EPIP,
respectively. The 77 K emission spectra, recorded in butyr-
onitrile glass (Figure 2b), show a blue shift relative to the
room temperature solution measurements, as expected for
an MLCT emission spectrum similar to what is known for
ruthenium polypyridyl complexes.19 The excited-state life-
times at 77 K (Table 1) are longer than those at room
temperature, and such behavior is attributed to the lack
of thermal population of the triplet metal-centered (3MC)
states, which would otherwise quench the luminescent
excited state.20
Preparation of the SAMs. In order to perform local
measurements on the SAMs, the prerequisite is a surface
with a very high degree of flatness. Hence, we prepared
the ultraflat gold substrates on a microscope coverslip
using the TS method as described by Weiss and co-
workers.14
The samples were immediately immersed in an EtOH
solution of the tripod molecules and left for 24 h for mono-
layer formation. A total of 10 μL of a hydrazine solution
was added to deprotect the thioacetyl groups in situ to
facilitate Ru-SH monolayer formation. The samples were
subsequently rinsed with EtOH to remove any unbound
molecules. As the AFM picture shows (Figure 3), the
monolayer is not homogeneous but forms islands, which is
Figure 3. AFM picture of the SAM of Ru-SH on a gold substrate and a
(19) Campagna, S.; Puntoriero, F.; Nastasi, F.; Bergamini, G.; Balzani,
height profile, revealing the formation of islands with heights correspond-
ing to the dimensions of the molecule.
V. Photochemistry and Photophysics of Coordination Compounds: Ruthenium.
Photochemistry and Photophysics of Coordination Compounds I; 2007;
pp 117-214.
well-known for other thiolate complexes on surfaces.1b,3n
The same procedure was applied to the reference com-
(20) Juris, A. B. V.; Barigelletti, F.; Campagna, S.; Belser, P.;
von Zelewsky, A. Coord. Chem. Rev. 1988, 84, 85–277.

1588 Inorganic Chemistry, Vol. 50, No. 5, 2011
Ramachandra et al.
pound Ru-tert-Bu, but no monolayer was found because
the lack of anchoring groups does not allow any strong
bond to the gold surface, and therefore after the first wash,
the complex was rinsed away. The AFM picture indicates,
however, that the islands of Ru-SAc are sufficiently large
to perform local measurements and the quality and height
of the layers confirm monolayer formation, their stabil-
ity, and good packing, as is also indicated by the high
yields in the Ga2O3/EGaIn measurements (vide infra).
Furthermore, when a control experiment was performed, by
immersion of a bare gold surface in the EtOH solution under
identical conditions, it revealed no such island forma-
tion, supporting our claim that the islands are indeed due
to monolayer formation. This experiment also suggested
that the depth profile of about 3 nm, which is indicated by
AFM, is not due to the surface roughness of the gold.
Photophysical Measurements of the Ru-SH Monolayer.
The emission measurements of the Ru-SH monolayer on
an ultraflat gold substrate were successfully carried out in
the spectrofluorimeter, indicating that despite the quench-
ing induced by the gold surface the signal strength from
the monolayer was detectable. The emission of the mono-
layer (Figure 4) resembled the emission profile of Ru-SH
in solution. However, we see a red shift of 4 nm in the
emission with respect to the solution measurements. This
could suggest that the lowest excited state involves the phen-
anthroline ligand, which, because of high conjugation,
feels the electronic interaction with the gold surface.
In other words, upon anchoring of the complexes on the
substrate, the 3MLCT state would be localized on the chelating
2-[4-(2-ethynyl)phenyl]-1H-imidazo[4,5-f ][1,10]phenanthroline
moiety and the electron-withdrawing effect of the gold
substrate lowers the lowest unoccupied molecular orbital
(LUMO) of the ligand, shifting the emission at lower
energy. In order to clarify the effect of the binding on
the gold surface, we have also drop-casted the solution
on glass (Figure 4). We were also able to measure the
excitation spectrum of this monolayer because the signal
strength was high enough, indicating a high packing
density (Figure S1 in the Supporting Information).
Figure 4. Emission spectra of the Ru-SH monolayer on anultraflat gold
substrate (9) and Ru-SH drop cast on glass (O).
Figure 5. Decay profiles of (A) Ru-SH on glass, (B) Ru-SH on gold, (C)
and bare gold. λex = 440 nm.
almost monoexponential behavior, because of the fact that
the short component has a weight of less than 10% on the
entire signal. The long component has an excited-state
lifetime of 812 ns and the shorter minor component of about
90 ns. This is in good agreement with the value reported for
other ruthenium complexes on glass (700 ( 50 ns).21 Because
the glass cannot play any role in the quenching of the emis-
sion, we attribute the short component to a possible high
local concentration of ruthenium complexes, which could
lead to self-quenching. The long excited-state lifetime can
be, therefore, taken as a good reference for the unquenched
ruthenium complex on an inert substrate.
The excited-state lifetimes of the Ru-SH molecules
after their immobilization, through binding of the thiol
groups on the gold, were measured using a time-resolved
confocal microscope. The lifetimes were fit to a biexponen-
tial, with a shorter component of 2 ns and a longer compo-
nent of 8 ns (Figure 5). This dual-exponential decay could be
due to the different orientations of the molecules on the
surfaces, for instance, coordination of two out of the three
thiols, leading to a different tilt angle of the ruthenium
complex versus the surface upon coordination of all of the
three anchoring groups. The X-ray photoelectron spec-
troscopy (XPS) measurements indicate that at least two
out of the three groups surely are attached to the gold
surface (Figure S2 in the Supporting Information), and
this is consistent with what was reported for similar sys-
tems.10 Another possible explanation for the presence of
two components is triplet-triplet annihilation due to a
strong packing of the complexes with a consequent quench-
ing of the emission. In order to evaluate the extent of the
quenching, we compared the film of Ru-SH by the drop
cast method on a microscopy glass plate, measuring the
lifetime of the ruthenium emission. The lifetime was fit to
a biexponential decay even though we can also assume
In the case of the SAM, we can, therefore, conclude that
quenching of the emission is due to a photoinduced energy
and/or electron transfer from the metal complex to the gold
surface as reported earlier.22 In particular, for similar
ruthenium complexes anchored onto the gold nanocrys-
talline surface, it has been reported that the origin of
quenching comes mainly from an electron transfer from
the excited-state ruthenium complex to the gold surface.23
The quenching rate can be calculated using the following
equation and the 8 ns lifetime for Ru-SH on gold and 812 ns
for Ru-SH on glass.
1
1
keT
¼
-
τ
τ0
The rate constant was determined to be 1.23 ꢀ 108 s-1
.
This estimated rate constant is about 3 orders of magnitude
(21) Wei, S.; Gafney, H. D.; Clark, J. B.; Perettie, D. J. Chem. Phys. Lett.
1983, 99, 253–257.
(22) D’Aleo, A.; Williams, R. M.; Chriqui, Y.; Iyer, V. M.; Belser, P.;
Vergeer, F.; Ruiz, V.; Unwin, P. R.; De Cola, L. Open Inorg. Chem. J. 2007,
1, 26–36.
(23) Pramod, P.; Sudeep, P. K.; Thomas, K. G.; Kamat, P. V. J. Phys.
Chem. B 2006, 110, 20737–20741.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1589
higher than that reported by Unwin and co-workers for
a thiolated ruthenium complex on a gold surface
(1 ꢀ 105 s-1 3c
)
and about 2 orders of magnitude larger
than that reported by Kamat and co-workers for elec-
tron transfer from a ruthenium complex to gold nano-
particles (1.1 ꢀ 106 s-1).23 However, in both of the above
cases, they have alkyl linkers to the gold surface, which
surely slow down the electron-transfer processes com-
pared to our fully conjugated rigid tripod system.
Electrochemical Measurements. Electrochemistry of
the reference compound, Ru-tert-Bu, as well as the thiolated
Ru-SH was performed in solutions. The Ru-SH electro-
chemistry was compared with the results obtained on
SAMs on a platinum-disk electrode. In an acetonitrile
solution with TBAPF6 as the supporting electrolyte, both
of the complexes show reversible reduction and oxidation
waves. In particular, the first oxidation occurs at þ1.31 V
(vs saturated calomel electrode, SCE). The peak-to-peak
separation is 90 mV. The observed peak-to-peak separa-
tion is larger than the expected peak-to-peak separation
for an ideal Nernstian behavior (ca. 59 mV). However,
because the reference redox species ferrocene also showed
this, we can attribute the observed effect to the ohmic drop in
such systems; similar behavior is also reported by Bard and
co-workers.24
Figure 6. CV of Ru-SH adsorbed onto the platinum electrode recorded
in an acetonitrile solution (0.1 M TBAPF6) at different scan rates:
(A) 100 mV s-1; (B) 200 mV s-1; (C) 500 mV s-1; (D) 1 V s-1
.
ruthenium and the platinum surface.28 The relationship
between the peak height and scan rate was not linear, as
was expected for a redox-active species covalently attached
on the electrode surface (diffusionless-controlled process),
but rather showed a quadratic dependence.
Conductivity Measurements Using a Ga2O3/EGaIn Setup.
In order to gain a better picture on how the electrons and
holes can flow through the molecules, we formed tunnel-
ing junctions with SAMs of Ru-SAc. We formed SAMs of
Ru-SAc on template-striped gold (AuTS) bottom electrodes
and used cone-shaped top electrodes of a liquid metal alloy of
gallium and indium (Ga2O3/EGaIn)15,16,29 to contact and to
complete the tunneling junctions.14 A surface layer of Ga2O3
on EGaIn readily forms under ambient conditions.30 A sche-
matic picture of the tunneling junctions is shown in Figure 7.
In all experiments, we biased the Ga2O3/EGaIn top electrode
and grounded the AuTS electrode.
We used top electrodes Ga2O3/EGaIn for three rea-
sons: (i) Junctions with top electrodes of Ga2O3/EGaIn
are easy to assemble.15 (ii) Junctions with top electrodes
of Ga2O3/EGaIn have J(V) characteristics that are domi-
nated by the chemical structure of the SAM.16 Thus,
Ga2O3/EGaIn does not destroy the molecules in the SAM,
unlike, for instance, direct metal deposition methods,31 and
the junctions are not dominated by artifacts, such as the
formation and dissolution of filaments.32 (iii) Junctions with
top electrodes of Ga2O3/EGaIn are relatively mechan-
ically stable.16 This stability makes it possible to measure
hundreds of J(V) curves of single junctions and to collect
statistically large numbers of data. Statistically large
numbers of data have to be recorded and analyzed to
discriminate the artifact from real data and to determine
the reproducibility and yield of working devices.16,33
Early work has shown that a liquid electrode, such as a
The reversible oxidation can be attributed to oxidation
of the ruthenium center RuIII/RuII and occurs at poten-
tials similar to that of [Ru(bpy)2(phen)]2þ 25
The first
.
reduction at -1.33 V can be attributed to reduction of the
bipyridine ligand.
For electrochemistry on monolayers, a 1 mm platinum-
disk electrode, precleaned thoroughly by repeated soni-
cation with deionized water and acetonitrile (spectro-
scopic grade), was used to form Ru-SH monolayers by
immersion of this precleaned platinum electrode in a con-
centrated solution of Ru-SAc in acetonitrile (∼10-3 M). The
thioacetate groups were deprotected in situ by adding ∼
10 μL of a hydrazine solution to obtain the Ru-SH complex.
The electrode was later (after 48 h of immersion) rinsed
with acetonitrile before measurements to eliminate any
adsorbed complex. The cyclic voltammogramms were mea-
sured with different scan rates and referenced against SCE
with a silver wire as a quasi-reference electrode. Oxidation of
the ruthenium ion occurs at þ1.38 V, very similar to the
value obtained in solution. It is interesting to note that the
peak separation ΔEp is greater than 0 and increases with
increasing scan rates (210 mV at 1000 mV s-1 compared
to 52 mV at 100 mV s-1; see Figure 6). This is consistent
with similar trends observed for osmium bipyridine com-
plexes on gold surfaces.26 In addition to this, the fwhm is
greater than 90.6 mV, which is the expected value for an
ideal one-electron redox process. Both of these observa-
tions can be attributed to repulsive interactions between
neighboring redox sites, which becomes significant at a
reasonably higher packing density of the monolayer,27
and to rather slow electron-transfer reactions between the
(28) Laviron, E. J. Electroanal. Chem. 1979, 101, 19–28.
(29) Dickey, M. D.; Chiechi, R. C.; Larsen, R. J.; Weiss, E. A.; Weitz,
D. A.; Whitesides, G. M. Adv. Funct. Mater. 2008, 18, 1097–1104.
(30) The thickness, composition, and surface roughness of the layer of
Ga2O3 have been characterized. The electrical properties also have been
characterized. The layer of Ga2O3 is about 2 orders of magnitude more
resistive than bulk EGaIn, is about 1-2 nm thick, and is rough (the actual
contact area is 20-40% of the measured contact area). We will report these
findings in a separate paper.
(31) Walker, A. V.; Tighe, T. B.; Cabarcos, O. M.; Reinard, M. D.;
Haynie, B. C.; Uppili, S.; Winograd, N.; Allara, D. L. J. Am. Chem. Soc.
2004, 126, 3954–3963.
(32) Beebe, J. M.; Kushmerick, J. G. Appl. Phys. Lett. 2007, 90, 083117–3.
(33) Kim, T. W.; Wang, G.; Lee, H.; Lee, T. Nanotechnology 2007, 18,
315204–8.
(24) Khalid, O. M.; Ku, S.-Y.; Wong, K.-T.; Bard, A. J. Angew. Chem.,
Int. Ed. 2009, 48, 9300–9303.
(25) Kalyanasundaram, K. Coord. Chem. Rev. 1982, 46, 159–244.
(26) Ricci, A.; Rolli, C.; Rothacher, S.; Baraldo, L.; Bonazzola, C.; Calvo, E.;
Tognalli, N.; Fainstein, A. J. Solid State Electrochem. 2007, 11, 1511–1520.
(27) (a) Laviron, E.; Roullier, L. J. Electroanal. Chem. 1980, 115, 65–74.
(b) Kondo, T.; Kanai, T.; Uosaki, K. Langmuir 2001, 17, 6317–6324.

1590 Inorganic Chemistry, Vol. 50, No. 5, 2011
Ramachandra et al.
Figure 7. Schematic ofthe AuTS-Ru-tripod//Ga2O3/EGaIn tunneljunc-
tions. The EGaIn top electrode is biased, andthe AuTS bottomelectrode is
grounded.
mercury drop, could be the answer to measuring rather
fragile monolayers.34 However, the mechanical stability
of the mercury-based junction is not so high, and indeed
the use of this electrode is rather impractical for further
developments. Junctions based on mercury-drop top electro-
des only were stable, in best cases, up to 15 scans before the
mercury top electrode amalgamates with the bottom
electrode.14 Ga2O3/EGaIn as a top electrode overcomes
some of these problems and proved to be a possible answer in
terms of stability and reproducibility.15
These junctions also have three uncertainties that are
all related to the layer of Ga2O3. (i) The thickness of the
layer of Ga2O3 is uncertain. We estimated the thickness of
the layer of Ga2O3 to be 1-2 nm by time-of-flight secondary
ion mass spectrometry and angle-resolved XPS. (ii) The
electrical properties of the layer of Ga2O3 are uncertain.
We estimated the resistivity of the layer of Ga2O3 to be
about 2 orders of magnitude more resistive than that of
bulk EGaIn but to be about 4 orders of magnitude less
resistive than that of a SAM of SC10CH3.16 Thus, we believe
that the resistivity of the layer of Ga2O3 is negligible. (iii) The
topography of the contact of Ga2O3 with the SAM is
uncertain. We recorded optical micrographs of cone-shaped
tips of Ga2O3 in contact with ITO. We estimated that the
actual contact area is ∼25% of the measured contact area.
The Ru-SAc monolayers were formed at the AuTS
electrodes for 24 h. The junctions were completed by
Figure 8. (top, a) Average |J|(V) semilogplot of the TSAu-Ru//EGaIn
junctions. One trace = 0 V f þ2.0 V f -2.0 V f 0 V, and the arrows
indicate the scan direction. (bottom, b). Energy levels of the Ru-SH
molecule and the electrodes in a EGaIn setup at an open circuit with
respect to a vacuum.36
contacting the monolayer with a conically shaped tip of
Ga2O3/EGaIn. Typical junction sizes were 500-1000 μm2.
A total of 17 junctions were assembled at three different
AuTS surfaces. Of these 17 junctions, 14 were working and
3 were shorting; thus, the yield of the working devices was
82%.35 The junctions were then scanned between -2andþ2V
as an example of a J(V) curve. We performed a statistical
analysis of the data by a procedure reported earlier.16 The
semilog plot of the average |J|(V) curve is shown in Figure 8.
The junctions were stable against these large potentials,
and they could be measured for more than 1 h, obtaining
20-30 scans before they shorted or the experiments were
stopped. We measured 60 traces per junction before we
terminated the experiment. The relatively large stability
and high yield in the working devices suggest that the tripod
molecules are extremely stable and, most likely, constantly
standing up and that the SAMs are densely packed.
The Ru-tripod has accessible HOMO and LUMO
levels, which may come in resonance with the Fermi levels
of the electrodes at large bias. The tripod molecule studied
in this paper is a highly conjugated molecule; thus, the
HOMO and LUMO orbitals will, to a certain degree, be
delocalized. Given the length of the molecule, it is fair to
assume that both the HOMO and LUMO will be located
unsymmetrically inside the junction. For this reason, they
are, in principle, unsymmetrically coupled to the electro-
des and should be coupled more strongly to the Ga2O3/
EGaIn top electrode. We believe that the HOMO and
(34) Tran, E.; Rampi, M. A.; Whitesides, G. M. Angew. Chem., Int. Ed.
2004, 43, 3835–3839.
(35) Because the sizes of the AFM tip and EGaln droplet are not
comparable and are roughly 3 orders of magnitude apart, an uneven surface
on the nanoscale does not necessarily mean a macroscopic effect because the
area measured under EGaIn is several micrometers while the defect measured
with AFM is on the order of a tenth of a nanometer. The measurements are
highly reproducible, and the scan can be repeated several times without any
degradation or destruction of the layer, indicating a very stable surface.

Article
Inorganic Chemistry, Vol. 50, No. 5, 2011 1591
LUMO levels are unsymmetrically coupled to the bottom
and top electrodes, but the presence of mobile PF6- anions in
the junction will, at least partially, compensate for that.
The presence of mobile PF6- anions in the junctions may
cause large potential drops at the tripod//Au interface,
facilitating the hole-injection process.
geometry of the system. The structure of the complexes has
been, in fact, designed in a way that the ruthenium center is
coordinated to two bipyridine ligands and a chelating phe-
nanthroline bearing a five-membered ring for its axial func-
tionalization. This approach allows the complexes to be lumines-
cent and, for the thiol derivative, to stand almost vertically
when assembled to a surface. We have demonstrated that the
gold surface quenches the emission of the ruthenium com-
plexes self-assembled on the metal surface. The SAMs of
ruthenium complexes have been employed to construct a
redox-active junction to investigate the conductivity of the
assembled molecules on the gold surface by using an EGaIn
eutectic as a second electrode. The results showed that the
monolayers are stable and a rectification behavior is ob-
served. In view of the fact that the complexes are also
electroluminescent, an understanding of the charge injection
and transport could lead to important consequences for the
design of LEEC and other electroluminescent devices.
Figure 8 shows the approximate energy diagram. The
HOMO and LUMO levels were estimated by electro-
chemistry data36 and found to be -5.8 and -3.1 eV,
respectively. Ga2O3/EGaIn has a work function of ∼-4.3 eV
and gold of -5.1 eV. We do not know the details of the
mechanism of charge transport, but we hypothesize that
charge can be injected in the HOMO and LUMO levels of
the molecules. Currently, we are investigating tripod mol-
ecules with different metal centers and ligands, hence differ-
ent HOMO and LUMO levels, to gain more insight into
the tunneling processes across these junctions. Furthermore,
as a possible extension of this work, the great stability of
the TS Au/Ru-tripod//EGaIn tunnel junctions could lead
to investigations of the charge-transfer processes of electro-
luminescent molecules in tunnel junctions (single-layer
electroluminescence).
Acknowledgment. Funding from NanoNed (Project
AMM7010), which is an initiative from the Dutch ministry
of economic affairs, is acknowledged. The authors acknowl-
€
edge Dr. Federico Polo (Westfalische Wilhelms-Universitat
Munster) for helpful discussions.
€
Conclusions
€
We have prepared and investigated a series of ruthenium
complexes, two of them containing a tripod system. The
tripod can be further chemically functionalized with thiol
groups for its attachment to metallic surfaces through more
thiols to stabilize the binding and allow a perfect perpendicular
Supporting Information Available: Excitation spectrum of a
Ru-SAc monolayer on a gold substrate (Figure S1), XPS spectrum
of Ru-SH bound to a gold substrate (Figure S2), statistical analysis
of the data from tunneling junctions (Figure S3), and estimation
of the surface coverage. This material is available free of charge
via the Internet at http://pubs.acs.org.
(36) Bard, A. J.; Faulkner, L. R. Electrochemical Methods;Fundamentals
and Applications; John Wiley & Sons: New York, 2001.