Redox behavior of rhodium 9,10-phenanthrenediimine complexes

Article abstract of DOI:10.1021/ic1007632

New square-planar rhodium complexes of the redox-active ligand 9,10-phenenthrenediimine (phdi) have been prepared and studied. The complexes [dpp-nacnac^CH3]Rh(phdi) (2a; [dpp-nacnac^CH3]^- = CH[C(Me)(N-ⁱPr₂

Full text of DOI:10.1021/ic1007632

Inorg. Chem. 2011, 50, 13–21 13
DOI: 10.1021/ic1007632
Redox Behavior of Rhodium 9,10-Phenanthrenediimine Complexes
David W. Shaffer, Scott A. Ryken, Ryan A. Zarkesh, and Alan F. Heyduk*
Department of Chemistry, University of California, Irvine (UCI), California 92697-2025
Received April 20, 2010
New square-planar rhodium complexes of the redox-active ligand 9,10-phenenthrenediimine (phdi) have been prepared
and studied. The complexes [dpp-nacnacCH3]Rh(phdi) (2a; [dpp-nacnacCH3]- =CH[C(Me)(N-iPr2C6H3)]2-) and [dpp-
nacnacCF3]Rh(phdi) (2b; [dpp-nacnacCF3]- =CH[C(CF3)(N-iPr2C6H3)]2-) have been prepared from the corresponding
[nacnac]Rh(CO)2 synthons by treatment with Me3NO in the presence of the phdi ligand. Complexes 2a and 2b are
diamagnetic, and their absorption spectra are dominated by intense charge-transfer transitions throughout the visible region.
Electrochemical studies indicate that both the phdi ligand and the rhodium metal center are redox-active, with the [nacnac]-
ligands serving to modulate the one-electron-oxidation and -reduction redox potentials. In the case of 2a, chemical oxidation
and reduction reactions provided access to the one-electron-oxidized cation, [2a]þ, and one-electron-reduced anion, [2a]-,
the latter of which has been characterized in the solid state by single-crystal X-ray diffraction. Solution electron paramagnetic
resonance spectra of [2a]þ and [2a]- are consistent with S =1/2 spin systems, but surprisingly the low-temperature spectrum
of [2a]- shows a high degree of rhombicity, suggestive of rhodium(II) character in the reduced anion.
Introduction
been an increase in the study of the synthesis and reactivity of
these complexes.1,3,4 Surprisingly, despite the importance of
rhodium complexes to coordination, bioinorganic, and orga-
nometallic chemistry, little effort has been directed toward
understanding the electronic properties of rhodium complexes
with such redox-active ligands.1d,5,6
Square-planar rhodium(I) complexes occupy a special
position in inorganic and organometallic chemistry for their
dynamic stoichiometric and catalytic bond activation
reactivity.7 In addition to the catalysis of a variety of H-X
additions to unsaturated organic substrates,8 rhodium(I)
complexes have been exploited in hydrogen production from
protons,9 CO2 reduction,10 and C-F activation11 and have
even been shown to bind N2.12 One of the key features of
square-planar rhodium(I) complexes that leads to this varied
Late-transition-metal ions interact with redox-active li-
gands to afford complexes with complicated electronic prop-
erties owing to the energetic proximity of the metal d orbitals
and the frontier ligand orbitals.1 Ambiguity in the assignment
of metal and ligand oxidation states often arises in these types
of complexes, prompting them to be described as noninno-
cent.2 Protocols have been established for utilizing structural,
spectroscopic, and theoretical techniques to probe the electro-
nic structures of these complexes,1e,3 and as such, there has
*To whom correspondence should be addressed. E-mail: aheyduk@
uci.edu.
(1) (a) Ghosh, M.; Weyhermuller, T.; Wieghardt, K. Dalton. Trans. 2010, 39,
€
1996–2007. (b) Chirik, P. J.; Wieghardt, K. Science 2010, 327, 794–795. (c) Boyer,
J. L.; Rochford, J.; Tsai, M.-K.; Muckerman, J. T.; Fujita, E. Coord. Chem. Rev. 2010,
254, 309–330. (d) van der Vlugt, J. I.; Reek, J. N. H. Angew. Chem., Int. Ed. 2009, 48,
8832–8846. (e)Kaim, W.;Lahiri, G. K. Angew. Chem., Int. Ed. 2007, 46, 1778–1796.
(f) de Bruin, B.; Hetterscheid, D. G. H. Eur. J. Inorg. Chem. 2007, 211–230.
(g) Butin, K. P.; Beloglazkina, E. K.; Zyk, N. V. Russ. Chem. Rev. 2005, 74, 531–553.
(h) Ward, M. D.; McCleverty, J. A. J. Chem. Soc., Dalton Trans. 2002, 275–288.
(2) (a) Kaim, W.; Schwederski, B. Pure Appl. Chem. 2004, 76, 351–364.
(b) Jørgensen, C. K. Coord. Chem. Rev. 1966, 1, 164–178. (c) Jørgensen, C. K.
Oxidation Numbers and Oxidation States; Springer: Heidelberg, Germany, 1969.
(3) (a) Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Trans.
2007, 1552–1566. (b) Remenyi, C.; Kaupp, M. J. Am. Chem. Soc. 2005, 127,
11399–11413. (c) Lever, A. B. P.; Gorelsky, S. I. Struct. Bonding (Berlin) 2004,
107, 77–114. (d) Chaudhuri, P.; Verani, C. N.; Bill, E.; Bothe, E.; Weyherm€uller,
T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213–2223.
(4) (a) Lippert, C. A.; Arnstein, S. A.; Sherrill, C. D.; Soper, J. D. J. Am.
Chem. Soc. 2010, 132, 3879–3892. (b) Kaim, W.; Sieger, M.; Greulich, S.;
Sarkar, B.; Fiedler, J.; Zalis, S. J. Organomet. Chem. 2010, 695, 1052–1058. (c)
Roquette, P.; Maronna, A.; Peters, A.; Kaifer, E.; Himmel, H.-J.; Hauf, C.; Herz,
V.; Schiedt, E.-W.; Scherer, W. Chem.;Eur. J. 2010, 16, 1336–1350. (d)
McKinnon, S. D. J.; Patrick, B. O.; Lever, A. B. P.; Hicks, R. G. Chem. Commun.
2010, 46, 773–775.
(5) Allgeier, A. M.; Mirkin, C. A. Angew. Chem., Int. Ed. 1998, 37, 895–
908.
(6) Palmer, J. H.; Mahammed, A.; Lancaster, K. M.; Gross, Z.; Gray,
H. B. Inorg. Chem. 2009, 48, 9308–9315.
(7) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem. Rev. 2010, 110,
624–655.
(8) (a) Lazzaroni, R.; Settambolo, R.; Alagone, G.; Ghio, C. Coord.
Chem. Rev. 2010, 254, 696–706. (b) Hayashi, T.; Yamasaki, K. Chem. Rev. 2003,
103, 2829–2844. (c) Rhodium-Catalysed Hydroformylation; van Leeuwen, P.
W. N. M., Claver, C., Eds.; Kluwer CMC: Dordrecht, The Netherlands, 2000. (d)
Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. Rev. 2002, 102, 1731–1769.
ꢀ
(9) Grass, V.; Lexa, D.; Saveant, J.-M. J. Am. Chem. Soc. 1997, 119, 7526–
7532.
(10) Huang, K.-W.; Han, J. H.; Musgrave, C. B.; Fujita, E. Organome-
tallics 2007, 26, 508–513.
(11) Torrens, H. Coord. Chem. Rev. 2005, 249, 1957–1985.
(12) (a) Masuda, J. D.; Stephan, D. W. Can. J. Chem. 2005, 83, 324–327.
(b) Budzelaar, P. H. M.; de Gelder, R.; Gal, A. W. Organometallics 1998, 17,
4121–4123. (c) Vigalok, A.; Ben-David, Y.; Milstein, D. Organometallics 1996,
15, 1839–1844.
ꢀ ꢁ
r
2010 American Chemical Society
Published on Web 07/14/2010
pubs.acs.org/IC

14 Inorganic Chemistry, Vol. 50, No. 1, 2011
Shaffer et al.
Chart 1
(trimethylsilyl)-9,10-phenanthrenediimine,16 [dpp-nacnacCF3]H,17
and [dpp-nacnacCH3]Li Et2O18 were prepared according to
3
established procedures. Graphite (Alfa Aesar) was dried under
a dynamic vacuum at <0.1 mtorr for 1 day prior to use.
Potassium metal (Alfa Aesar) and [RhCl(CO)2]2 (Heraeus) were
used without further purification.
Physical Methods. NMR spectra were collected on Bruker
Avance 400 and 600 MHz spectrometers in dry, degassed CDCl3
1
or C6D6. H NMR spectra were referenced to tetramethylsilane
reactivity is the nucleophilicity of the metal center, which
provides a mechanism for the binding of substrates in the
metal coordination sphere.
(TMS) using the residual proteo impurities of the solvent; 13C
NMR spectra were referenced to TMS using the natural abun-
dance 13C impurities of the solvent. 19F NMR spectra were refe-
renced to CFCl3 using C6F6 as an internal standard at -164.9
ppm. All chemical shifts are reported using the standard δ nota-
tion in parts per million; positive chemical shifts are to a higher
frequency from the given reference. IR spectra were recorded as
KBr pellets with a Perkin-Elmer Spectrum One FTIR spectro-
photometer. Electronic absorption spectra were recorded with a
Perkin-Elmer Lambda 800 UV-vis spectrophotometer. APCI-
MS data were collected on a Waters LCT Premier mass spectro-
meter. Perpendicular-mode X-band EPR spectra were collected
using a Bruker EMX spectrometer equipped with an ER041XG
microwave bridge.
The introduction of a redox-active ligand into the coordina-
tion sphere of a rhodium(I) center offers a unique opportunity
to compliment the nucleophilicity of rhodium(I) with a ligand
capable of further two-electron reactivity. 9,10-Phenanthrene-
diimine (phdi) is a redox-active ligand derived from phenan-
threnequinone in which the quinone oxygen atoms have been
replaced with imines to afford an R-diimine for coordination to
metal ions. While phdi has been utilized extensively as a ligand
for rhodium(III) and other metal ions for DNA intercalation
studies,13 its use as a platform for redox activity with lower-
valent late transition metals is underexplored.14,15 While phdi is
a closed-shell neutral ligand, one-electron reduction affords an
open-shell diiminosemiquinonate ([phdisq•]-) and two-elec-
tron reduction affords the closed-shell phenanthrenediamide
form ([phda]2-), as shown in Chart 1. Thus, a RhIphdi complex
could be bifunctional in that the rhodium(I) metal center would
be nucleophilic because of its d8 electron configuration and yet
the complex could still be reduced by up to two electrons be-
cause of the redox-active nature of the phdi ligand. This paper
reports the synthesis of a new type of rhodium(I) complex that
includes a redox-active phdi ligand and a chelating [nacnac]-
ligand. While the neutral complex displays properties consis-
tent with a “normal” square-planar, d8 rhodium(I) coordinated
to a phdi ligand, one-electron oxidation and reduction afford
complexes that display noninnocent behavior. Notably, the
electron paramagnetic resonance (EPR) spectra for the one-
electron-reduction product suggest oxidation of the metal
center up to rhodium(II) with concomitant reduction of the
phdi ligand to the [phda]2- oxidation state.
Electrochemical Methods. Electrochemical experiments were
performed on a Gamry Series G 300 potentiostat/galvanostat/
ZRA (Gamry Instruments, Warminster, PA) using a 3.0 mm
glassy carbon working electrode, a platinum wire auxiliary elec-
trode, and a silver wire reference electrode. Electrochemical ex-
periments were performed at ambient temperature (20-24 °C),
either in a nitrogen-filled glovebox or under an atmosphere of
argon. Sample concentrations were 1.0 mM in THF with 0.10 M
NBu4PF6 as the supporting electrolyte. All potentials are referenced
to [Cp2Fe] ,
þ/0 19 using decamethylferrocene as an internal standard
(-0.49 V vs [Cp2Fe]þ/0). The typical solvent system window
with our configuration was þ1.0 V for the oxidation limit and
-3.4 V for the reduction limit (vs [Cp2Fe]þ/0). Decamethyl-
ferrocene (Acros) was purified by sublimation under reduced
pressure,20 and tetrabutylammonium hexafluorophosphate
(Acros) was recrystallized from ethanol three times and dried
under vacuum.21
Synthesis of 9,10-Phenanthrenediimine (phdi) and 9,10-Phe-
nanthrenediimine-d2 (phdi-d2). The synthesis was through mod-
ification of a published procedure.16 Dry methanol (50 mL) was
added to a Schlenk flask containing 487 mg of N,N0-bis-
(trimethylsilyl)-9,10-phenanthrenediimine (1.39 mmol) and stir-
red for 40 min. The solvent and methoxytrimethylsilane bypro-
duct were then removed under reduced pressure. Pure phdi was
isolated in quantitative yield as a light-green solid. The deuterated
version, phdi-d2, was prepared by an analogous procedure using
methanol-d. Anal. Calcd for C14H10N2: C, 81.53; H, 4.89; N,
13.58. Found: C, 81.06; H, 4.69; N, 13.37. ESI-MS (methanol):
m/z 207.0 ([M þ H]þ). 1H NMR (C6D6, 600 MHz): δ 12.39 [s, 1H,
N-H, (E,Z)-phdi], 10.47 [s, 1H, N-H, (E,Z)-phdi], 10.18 [s, 2H,
Experimental Section
General Considerations. Many of the complexes described
below are air- and moisture-sensitive, necessitating that manipula-
tions be carried out under an inert atmosphere of argon or nitrogen
using standard Schlenk, vacuum-line, and glovebox techniques. All
reactions were carried out at ambient temperature (20-24 °C)
unless otherwise noted. Hydrocarbon solvents were sparged with
argon, then deoxygenated, and dried by successive passage through
Q5 and activated alumina columns, respectively. Ethereal and
halogenated solvents were sparged with argon and then dried by
passage through two activated alumina columns. To test for effective
oxygen and water removal, nonchlorinated solvents were treated
with a few drops of a purple solution of sodium benzophenone
ketyl in tetrahydrofuran (THF). The ligand precursors N,N0-bis-
3
N-H, (Z,Z)-phdi], 8.89 [d, JHH = 7 Hz, 1H, Ar-H, (E,Z)-
3
phdi], 8.60 [d, JHH = 7 Hz, 2H, Ar-H, (Z,Z)-phdi], 7.48 (m,
Ar-H ), 7.10 (t, Ar-H ), 7.07 (t, Ar-H ), 7.00 (t, Ar-H ), 6.85 (d,
Ar-H ), 6.78 (t, Ar-H ). IR (KBr): νCO 3252 (N-H), 3208
(N-H), 3065 (N-H), 1624 (CdN), 1602 (CdN) cm-1. UV-vis
[CH2Cl2; λmax/nm (ε/M-1 cm-1)]: 296 (3200), 373 (1900).
(13) (a) Fu, P. K.-L.; Bradley, P. M.; Turro, C. Inorg. Chem. 2003, 42, 878–
884. (b) Hall, D. B.; Holmlin, R. E.; Barton, J. K. Nature 1996, 382, 731–735. (c)
Krotz, A. H.; Kuo, L. Y.; Barton, J. K. Inorg. Chem. 1993, 32, 5963–5974.
(14) Belser, P.; Von Zelewsky, A.; Zehnder, M. Inorg. Chem. 1981, 20,
3098–3103.
(17) van Belzan, R.; Klein, R. A.; Smeets, W. J. J.; Spek, A. L.; Benedix,
R.; Elsevier, C. J. Recl. Trav. Chim. Pays-Bas 1996, 115, 275–285.
(18) Stender, M.; Wright, R. J.; Eichler, B. E.; Prust, J.; Olmstead, M. M.;
Roesky, H. W.; Power, P. P. J. Chem. Soc., Dalton Trans. 2001, 3465–3469.
(19) (a) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 92, 877–910. (b)
Gritzner, G.; Kuta, J. Pure Appl. Chem. 1984, 56, 461–466.
(15) (a) Chern, S.-S.; Liaw, M.-C.; Peng, S.-M. J. Chem. Soc., Chem.
Commun. 1993, 359–361. (b) Peng, S. M.; Chen, C.-T.; Liaw, D.-S.; Chen, C.-I.;
Wang, Y. Inorg. Chim. Acta 1985, 101, L31–L33.
(20) (a) Zahl, A.; van Eldik, R.; Matsumoto, M.; Swaddle, T. W. Inorg.
Chem. 2003, 42, 3718–3722. (b) Coddington, J.; Wherland, S. Inorg. Chim. Acta
1996, 242, 159–164.
€
(16) Tuchtenhagen, G.; Ruhlmann, K. Liebigs Ann. Chem. 1968, 711,
174–183.
(21) Bedard, R. L.; Dahl, L. F. J. Am. Chem. Soc. 1986, 108, 5933–5942.

Article
Inorganic Chemistry, Vol. 50, No. 1, 2011 15
Synthesis of [dpp-nacnacCF3]Li Et2O. A solution of [dpp-
([M]þ).1HNMR(CDCl3, 600 MHz): δ7.98-7.90 (m, 4H, Ar-H),
7.82 (s, 2H, N-H), 7.75(d, 2JHH =8.1Hz, 2H, Ar-H), 7.71-7.67
(m, 6H, Ar-H), 7.22 (ddd, 4JHH = 1.9 Hz, 3JHH = 6.2 and 8.1 Hz,
2H, Ar-H), 4.53 (s, 1H, -CH-), 3.32 [sept, 3JHH = 6.9 Hz, 4H,
-CH(CH3)(CH3)0], 1.99 (s, 6H, -CH3), 1.34 [d, 3JHH = 7.0 Hz,
3
nacnacCF3]H (151 mg, 287 mmol, 1 equiv) in 9 mL of pentane
and 1 mL of diethyl ether was cooled to -35 °C. n-BuLi (2.60 M,
121 μL, 1.1 equiv) in hexane was diluted to 5 mL in pentane and
added dropwise to the [dpp-nacnacCF3]H solution. After stirring at
room temperature for 12 h, the solvent was reduced to 2 mL, when
the product began to precipitate. The suspension was warmed to
redissolve all solids and then cooled to -35 °C to precipitate the
product as yellow crystals (yield 119 mg, 69%). 1H NMR (CDCl3,
600 MHz): δ 7.06 (d, 3JHH=7.3 Hz, 4H, Ar-H), 7.06 (t, 3JHH=7.6
3
12H, -CH(CH3)(CH3)0], 0.93 [d, JHH = 6.8 Hz, 12H, -CH-
(CH3)(CH3)0]. 13C NMR (CDCl3, 125.8 MHz): δ 154.6 (C-N),
153.3 (C-N), 151.6 (aryl-C-N), 145.8 (aryl-C), 130.8 (aryl-C),
129.1 (aryl-C), 129.0 (aryl-C), 126.6 (aryl-C), 125.6 (aryl-C),
124.8 (aryl-C), 122.8 (aryl-C), 118.4 (aryl-C), 97.0 (-CH-),
28.7 [-CH(CH3)(CH3)0], 23.8 [-CH(CH3)(CH3)0], 23.7 [-CH-
Hz, 2H, Ar-H), 5.52 (s, 1H, -CH), 2.97 [sept, 3JHH=6.9 Hz, 4H,
-CH(CH3)(CH3)0], 2.69 (br, 4H, -OCH2CH3), 1.22 [d, JHH
=
(CH3)(CH3)0], 23.6 (-CH3). IR (KBr) ν 3313 (N-H) cm-1
.
3
UV-vis [CH2Cl2; λmax/nm (ε/M-1 cm-1)]: 362 (23 000), 593
(22 700), 796 (550).
3
6.8 Hz, 12H, -CH(CH3)(CH3)0], 1.03 [d, JHH = 6.9 Hz, 12H,
-CH(CH3)(CH3)0], 0.67 (br, 6H, -OCH2CH3). 13C NMR (CD-
Cl3, 125.8 MHz): δ 152.5 (q, 2JCF=68 Hz, CN), 145.9 (aryl-CN),
139.9 (o-aryl-C), 124.3 (aryl-C), 123.5 (aryl-C), 83.4 (-CH-),
28.3 [-CH(CH3)(CH3)0], 25.0 [-CH(CH3)(CH3)0], 23.1 [-CH-
(CH3)(CH3)0]. 19F NMR (CDCl3, 376.5 MHz); δ -64.0 (s, 6F,
-CF3).
Synthesis of [dpp-nacnacCF3]Rh(phdi) (2b). N,N0-Bis(tri-
methylsilyl)-9,10-phenanthrenediimine (80.5 mg, 0.230 mmol,
1.2 equiv) and Me3NO (28.9 mg, 0.385 mmol, 2.0 equiv) were
stirred in 10 mL of dry MeOH for 30 min. The solvent was then
removed under reduced pressure, and a 3 mL benzene solution of 1b
(131.6 mg, 0.192 mmol, 1 equiv) was added with a cannula. A dark-
green suspension formed and was stirred for 22 h to afford a dark-
blue solution. The solution was layered with 5 mL of dry methanol
to precipitate a purple solid, which was filtered in air, washed with 1
mL methanol, and dried under vacuum (yield 141.0 mg, 88%).
Anal. Calcd for C43H45N4F6Rh: C, 61.87; H, 5.43; N, 6.71. Found:
C, 61.53; H, 5.46; N, 6.60. 1H NMR (CDCl3, 600 MHz): δ
7.96-7.90 (m, 4H, Ar-H), 7.67 (t, 3JHH = 7.5 Hz, 2H, Ar-H),
7.59 (d, 3JHH = 7.2 Hz, 4H, Ar-H), 7.54 (d, 3JHH = 7.8 Hz, 2H,
Ar-H), 7.24 (d, 3JHH = 9.6 Hz, 2H, Ar-H), 7.19 (s, 2H, NH),
5.56 (s, 1H, -CH-), 3.13 [sept, 3JHH = 6.7 Hz, 4H, -CH(CH3)-
(CH3)0], 1.35 [d, 3JHH = 6.8 Hz, 12H, -CH(CH3)(CH3)0], 0.88 [d,
3JHH = 6.7 Hz, 12H, -CH(CH3)(CH3)0]. 13CNMR(CDCl3, 125.8
MHz): δ 155.7 (C-N), 149.2 (aryl-C-N), 145.4 (aryl-C), 142.4
(q, 2JCF = 26 Hz, C-N), 130.3 (aryl-C), 129.6 (aryl-C), 129.3
(aryl-C), 128.7 (aryl-C), 126.5 (aryl-C), 124.7 (aryl-C), 122.4
(aryl-C), 119.2 (aryl-C), 90.3 (-CH-), 29.1 [-CH(CH3)(CH3)0],
24.2 [-CH(CH3)(CH3)0], 23.1 [-CH(CH3)(CH3)0]. 19F NMR
(CDCl3, 376.5 MHz): δ -64.2 (s, 6F, -CF3). IR (KBr): ν 3309
(N-H) cm-1. UV-vis [CH2Cl2; λmax/nm (ε/M-1 cm-1)]: 312
(20 000), 357 (11 100), 450 (10 700), 587 (21 100).
Synthesis of [dpp-nacnacCH3]Rh(CO)2 (1a). A 5 mL ether
solution containing 387 mg of [dpp-nacnacCH3]Li Et2O (776
3
μmol, 2.0 equiv) was added to 151 mg of [Rh( μ-Cl)(CO)2]2 (389
μmol, 1 equiv) dissolved in 5 mL of ether. The yellow solution was
stirred for 8 h. After filtration to remove LiCl, the solvent was
removed under reduced pressure. Yellow crystals (yield 347 mg,
78%) of the product were obtained after recrystallization of the
crude product from THF at -35 °C. Anal. Calcd for C31H41-
N2O2Rh: C, 64.58; H, 7.17; N, 4.86. Found: C, 64.45; H, 7.25; N,
4.80. 1H NMR (CDCl3, 500 MHz): δ 7.20-7.13 (m, 6H, Ar-H),
5.12 (s, 1H, -CH-), 3.23 [sept, 3JHH = 6.9 Hz, 4H, -CH(CH3)-
(CH3)0], 1.79 (s, 6H, -CH3), 1.37 [d, 3JHH = 6.8 Hz, 12H, -CH-
(CH3)(CH3)0], 1.22 [d, 3JHH = 7.0 Hz, 12H, -CH(CH3)(CH3)0].
13C NMR (CDCl3, 125.8 MHz): δ 184.2 (d, 1JRhC = 66 Hz, CO),
160.1 (CN), 155.1 (aryl-CN), 140.3 (o-aryl-C), 125.9 (aryl-C),
123.4 (aryl-C), 96.7 (-CH-), 27.7 [-CH(CH3)(CH3)0], 24.0
[-CH(CH3)(CH3)0], 23.6 [-CH(CH3)(CH3)0], 22.7 (-CH3). IR
(KBr): νCO 2053, 1988 cm-1. UV-vis [CH2Cl2; λmax/nm (ε/M-1
cm-1)]: 352 (19600).
Synthesis of [dpp-nacnacCF3]Rh(CO)2 (1b). An 8 mL ether
solution containing [dpp-nacnacCF3]Li Et2O (162.0 mg, 267 μmol,
3
Synthesis of [(dpp-nacnac)Rh(phdi)][BF4] ([2a][BF4]). A 20 mL
scintillation vial was charged with 2a (104.5 mg, 144 μmol,
1 equiv) and tri-p-tolylaminium tetrafluoroborate (53.8 mg, 144
μmol, 1 equiv). Diethyl ether (10 mL) was added to the solids,
and the blue-purple suspension was stirred for 24 h. The purple
product was collected by filtration, washed with 3 ꢀ 1.5 mL of
diethyl ether, and dried under vacuum (yield 96.2 mg, 82%). IR
(KBr) ν 3296 (N-H) cm-1. UV-vis [CH2Cl2; λmax/nm (ε/M-1
cm-1)]: 323 (16 334), 571 (19 700), 855 (880). EPR (CH2Cl2, 298
K): g = 2.004. EPR (CH2Cl2, 77 K): g = 2.003.
2.1 equiv) was added to a 2 mL ether suspension of [Rh( μ-Cl)-
(CO)2]2 (49.1 mg, 126 μmol, 1.0 equiv). The yellow solution was
stirred for 16 h. After filtration to remove LiCl, the solvent was
removed under reduced pressure. Yellow-orange crystals of the
product were obtained after recrystallization of the crude product in
pentane at -35 °C (yield 158.6 mg, 92%). Anal. Calcd for C31H35-
N2O2F6Rh: C, 54.39; H, 5.15; N, 4.09. Found: C, 54.24; H, 5.21; N,
4.02. 1H NMR (CDCl3, 600 MHz): δ 7.22-7.11 (m, 6H, Ar-H),
6.08 (s, 1H, -CH-), 3.00 [sept, 3JHH = 6.8 Hz, 4H, -CH(CH3)-
(CH3)0], 1.43 [d, 3JHH = 6.8 Hz, 12H, -CH(CH3)(CH3)0], 1.26 [d,
3JHH = 6.8 Hz, 12H, -CH(CH3)(CH3)0]. 13C NMR (CDCl3, 125.8
MHz): δ 180.0 (d, 1JRhC = 68 Hz, CO), 154.1 (aryl-CN), 149.2 (d,
2JCF = 27 Hz, CN), 139.9 (aryl-C), 127.2 (aryl-C), 123.3 (aryl-
C), 87.9 (-CH-), 28.4 [-CH(CH3)(CH3)0], 24.3 [-CH(CH3)-
(CH3)0], 23.3 [-CH(CH3)(CH3)0]. 19F NMR (CDCl3, 376.5 MHz):
δ -63.5 (s, 6F, -CF3). IR (KBr): νCO 2078, 2020 cm-1. UV-vis
[CH2Cl2; λmax/nm (ε/M-1 cm-1)]: 308 (6300), 391 (12 600).
Synthesis of [K(18-crown-6)][(dpp-nacnac)Rh(phdi)] ([K(18-
crown-6)][2a]). In a 20 mL scintillation vial, 2a was dissolved
in THF (105.5 mg, 145 μmol, 1 equiv) and cooled to ca. -35 °C. To
this deep-blue solution was added a freshly prepared KC8 solid
(20.1 mg, 149 μmol, 1.02 equiv). The resulting suspension was
stirred for 45 min, over which time the color changed from blue to
green. The solid products were then filtered off and washed with
1 mL of THF. To the filtrate was added solid 18-crown-6 (60 mg,
227 μmol, 1.6 equiv). The volume of the resulting brown solution
was reduced to 4 mL in vacuo, and pentane (6 mL) was added to
precipitate the product, which was filtered off, washed with pentane
(3 ꢀ 1.5 mL), and dried under vacuum (yield 143.8 mg, 96%). IR
(KBr): ν 3356 (N-H) cm-1. EPR (THF, 298 K): g = 2.09. EPR
(THF, 77 K): g = 2.395, 1.975, and 1.941.
Synthesis of [dpp-nacnacCH3]Rh(phdi) (2a). A 50 mL metha-
nolic solution of N,N0-bis(trimethylsilyl)-9,10-phenanthrenediimine
(486 mg, 1.39 mmol, 1.2 equiv) and Me3NO (174 mg, 2.32 mmol,
2.0 equiv) was stirred for 30 min. The solvent was removed under
reduced pressure, and a 10 mL benzene solution of 1a (666 mg,
1.16 mmol, 1 equiv) was added with a cannula. The yellow suspen-
sion was stirred for 24 h to afford a dark-blue suspension of the
product. Dry methanol (20 mL) was added to aid product pre-
cipitation, and the product was isolated as a purple solid by filtration
in air, washing with methanol, and drying under vacuum (yield 720
mg, 86%). Anal. Calcd for C43H51N4Rh: C, 71.06; H, 7.07; N, 7.71.
Found: C, 70.94; H, 7.23; N, 7.62. APCI-MS (toluene): m/z 726.1
Crystallographic Methods. X-ray diffraction data were
collected on crystals mounted on glass fibers using a Bruker
CCD platform diffractometer equipped with a CCD detector.
Measurements were carried out at 163 K using Mo KR (λ =
˚
0.710 73 A) radiation, which was wavelength-selected with a

16 Inorganic Chemistry, Vol. 50, No. 1, 2011
Table 1. X-ray Diffraction Data Collection and Refinement Parameters
[K(18-crown-6)]
Shaffer et al.
Scheme 1. Synthesis of 2a and 2b
[dpp-nacnacCH3]-
[(dpp-nacnac)Rh(phdi)]
([K(18-crown-6)][2a])
Rh(CO)2 (1a)
empirical formula
fw
cryst syst
C31H41N2O2Rh
576.57
monoclinic
C2/c
35.218(6)
9.3465(16)
20.861(4)
90
120.613(2)
90
5909.8(17)
8
31 653
7286
0.0209
C55H75KN4O6Rh•(C4H8O)3
1246.51
monoclinic
P21/c
12.5891(7)
24.1291(13)
22.6328(12)
90
105.2230(10)
90
6633.8(6)
4
76 342
space group
˚
a/A
˚
b/A
˚
c/A
R/deg
β/deg
γ/deg
3
˚
V/A
Z
reflns collected
indep reflns
R1 [I > 2σ(I )]a
wR2 (all data)a
GOFa
15 768
0.0376
0.0874
0.997
0.0554
1.052
˚
Table 2. Selected Bond Lengths (A) for 1a and [K(18-crown-6)][2a]
P
P
P
P
1a
[K(18-crown-6)] [2a]
a R1 = ||Fo| - |Fc||/ |Fo|; wR2= [ [w(Fo2 - Fc2)2]/ [w(Fo2)2]]1/2
;
P
GOF = [ w(|Fo| - |Fc|)2/(n - m)]1/2
.
Rh(1)-N(1)
Rh(1)-N(2)
Rh(1)-N(3)
Rh(1)-N(4)
N(1)-C(1)
N(2)-C(3)
C(1)-C(2)
2.0491(11)
2.0514(11)
2.0097(16)
2.0111(16)
1.9881(17)
1.9832(17)
1.332(3)
1.331(2)
1.394(3)
1.401(3)
1.350(2)
single-crystal graphite monochromator. The SMART program
package was used to determine unit-cell parameters and to
collect data.22 The raw frame data were processed using
SAINT23 and SADABS24 to yield the reflection data files.
Subsequent calculations were carried out using the SHELXTL25
program suite. Structures were solved by direct methods and
refined on F2 by full-matrix least-squares techniques. Analytical
scattering factors for neutral atoms were used throughout the
analyses.26 Hydrogen atoms were included using a riding model.
ORTEP diagrams were generated using ORTEP-3 for
Windows.27 Diffraction data are shown in Table 1.
1.3308(16)
1.3316(17)
1.3994(16)
1.3965(17)
C(2)-C(3)
N(3)-C(30)
N(4)-C(43)
C(30)-C(43)
1.349(3)
1.409(3)
equal-intensity 1H NMR resonances. The N-H protons
of the E,E isomer resonate at 10.2 ppm, but hydrogen
bonding in the E,Z isomer splits the N-H proton reso-
nances, which appear at 10.5 and 12.4 ppm, respectively.
Results
Synthesis and Characterization of RhIphdi Complexes.
phdi was synthesized in quantitative yield by stirring N,
N0-bis(trimethylsilyl)-9,10-phenanthrenediimine in dry
methanol.16 Removal of the methanol in vacuo resulted
in a quantitative yield of phdi, which is pure by 1H NMR
spectroscopy. Incorporation of deuterium into the imine
group was readily achieved using CH3OD for depro-
tection. The free ligand exists as a mixture of E,Z and
E,E isomers in a 5.5:1 ratio (eq 1), as determined by
1H NMR spectroscopy at 298 K in deuterated benzene
(ΔG° = 1.0 kcal mol-1).17 Deuterium substitution to
make phdi-d2 favors the E,Z isomer further (6.6:1), con-
sistent with stronger hydrogen bonding in the deuterated
compound.28 The symmetric E,E isomer shows five
equal-intensity 1H NMR resonances in the aromatic
region, whereas the asymmetric E,Z isomer shows 10
Metalation of phdi with rhodium(I) was achieved by
oxidative decarbonylation of 1a and 1b (Scheme 1).
Complex 1a was prepared from [Rh( μ-Cl)(CO)2]2 and
[dpp-nacnac]Li Et2O in 78% yield according to proce-
3
dures developed for other Rh-nacnac derivatives.12a Si-
milarly, the derivative with the fluorinated nacnac ligand,
1b, was synthesized in 92% yield by stirring [Rh( μ-Cl)-
(CO)2]2 with [dpp-nacnacCF3]Li Et2O overnight at room
3
temperature. The geometry of the Rh-nacnac unit in 1a,
determined by single-crystal X-ray diffraction (see the
Supporting Information) is roughly square-planar and
analogous to other previously reported RhInacnac deri-
vatives.12a,b,29 Selected metrical data for 1a are shown
in Table 2. IR spectra of 1a and 1b revealed two CtO
stretching frequencies, observed at 2053 and 1988 cm-1
for 1a and at 2078 and 2020 cm-1 for 1b (Δν = 25 and
32 cm-1, respectively). The observed blue shift in the
(22) SMART Software Users Guide, version 5.1; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 1999.
(23) SAINT Software Users Guide, version 6.0; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 1999.
(24) Sheldrick, G. M. SADABS, version 2.10; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 2002.
(25) Sheldrick, G. M. SHELXTL, version 6.12; Bruker Analytical X-ray
Systems, Inc.: Madison, WI, 2001.
(26) International Tables for X-Ray Crystallography; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1992; Vol. C.
(29) (a) Geier, S. J.; Stephan, D. W. Chem. Commun. 2008, 99–101.
(c) Willems, S. T. H.; Budzelaar, P. H. M.; Moonen, N. N. P.; de Gelder, R.; Smits,
J. M. M.; Gal, A. W. Chem.;Eur. J. 2002, 8, 1310–1320. (d) Budzelaar,
P. H. M.; Moonen, N. N. P.; de Gelder, R.; Smits, J. M. M.; Gal, A. W. Eur.
J. Inorg. Chem. 2000, 753–769.
(27) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(28) (a) Pal, H.; Nagasawa, Y.; Tominaga, K.; Kumazaki, S.; Yoshihara,
K. J. Chem. Phys. 1995, 102, 7758–7760. (b) Rao, C. N. R. J. Chem. Soc.,
ꢀ
Faraday Trans. 1 1975, 71, 980–983. (c) Nemethy, G.; Scheraga, H. A. J. Chem.
Phys. 1964, 41, 680–689.

Article
Inorganic Chemistry, Vol. 50, No. 1, 2011 17
CtO stretching frequencies for 1a and 1b indicates that
the electronic properties of the nacnac ligands have a
pronounced impact on the electron density at the metal
center, and the installation of CF3 groups into the nacnac
backbone affords a more electron-deficient rhodium
center that cannot π-backbond as strongly to the carbon-
yl ligands. For comparison, the two CtO stretching
frequencies for the analogous iridium complexes, [dpp-
nacnacCH3]Ir(CO)2 and [dpp-nacnacCF3]Ir(CO)2, only
blue shift by 20 and 16 cm-1, respectively,30 indicating
a much smaller interaction of the nacnac ligand with the
metal valence orbitals.
Compounds 1a and 1b are robust in the presence of
phdi; however, Me3NO can be added to promote reaction
through the oxidative removal of carbonyl ligands
(Scheme 1). The addition of 2 equiv of Me3NO to a
benzene solution of 1a and phdi resulted in the release
of CO2 and formation of 2a as a microcrystalline purple
solid in 73% isolated yield. Complex 2a can also be
prepared directly using the crude phdi product obtained
upon removal of methanol from deprotection of N,N0-
bis(trimethylsilyl)-9,10-phenanthrenediimine. The yield
of 2a prepared in this manner is 86% based on 1a. The
reaction of 1b with phdi and 2 equiv of Me3NO proceeds
similarly to afford 2b as a purple solid in 88% yield.
Complexes 2a and 2b are proposed to be square-planar
rhodium complexes. The formulation of both complexes was
confirmed by elemental analysis and atmospheric-pressure
chemical ionization mass spectrometry (APCI-MS), which
showed [M]þ peaks with the appropriate isotopic distribu-
tion at m/z 726.1 for 2a and m/z 834.2 for 2b. The 1H NMR
spectra of 2a and 2b are consistent with C2v-symmetric,
square-planar complexes and show chemical shifts similar
to those of the dicarbonyl complexes 1a and 1b. Spectra of 2a
and 2b show four equivalent isopropyl groups for the dpp-
nacnac ligands with single methine resonances at 3.32 and
3.13 ppm, respectively. The diastereotopic methyl resonances
are observed at 0.93 and 1.34 ppm for 2a and at 0.88 and 1.35
ppm for 2b. In both cases, the phdi ligand is symmetric with
four aromatic resonances. The most significant differences in
the 1H NMR spectra of the two complexes are in the chemi-
cal shifts of the phdi N-H resonances and the vinylic nacnac
C-H resonance. Whereas the vinylic C-H resonance of 2b
(5.56 ppm) shifts downfield relative to the same resonance in
2a (4.53 ppm), the phdi N-H resonance of 2b (7.19 ppm) is
shifted upfield compared to 2a (7.82 ppm). While it would be
desirable to track the synthesis and reactivity of 2a and 2b by
monitoring the CdN stretches of the phdi ligand, these
stretches are less prominent than the CO ligands of 1a and
1b and are red-shifted into the fingerprint region of the com-
plexes, making definitive assignments difficult. The N-H
stretches are readily visible in the IR spectrum at 3313 cm-1
for 2a and at 3309 cm-1 for 2b.
Figure 1. UV-vis spectra of 2a (solid line) and 2b (dashed line) at 298 K
in CH2Cl2.
Figure 2. Cyclic voltammograms of (a) 2a and (b) 2b measured at 200
mV s-1. All measurements were made in a THF solution with 1.0 mM
analyte and 0.10 M TBAPF6 under an inert atmosphere. Potentials were
referenced to [Cp2Fe]0/þ
.
showing transitions at 312 nm (20 000 M-1 cm-1) and 587
nm (21 100 M-1 cm-1); furthermore, two additional strong
transitions are observed at 357 nm (11 100 M-1 cm-1) and
450 nm (10 700 M-1 cm-1). That the high-energy band shifts
from 362 nm in 2a to 312 nm in 2b suggests that this transition
originates on the metal or the nacnac ligand, with the CF3
groups of [dpp-nacnacCF3] causing the 50 nm blue shift in 2b.
The small blue shift of the low-energy band on going from 2a
to 2b suggests a metal-to-ligand charge-transfer assignment
involving only the rhodium center and the phdi ligand.
Electrochemical Studies of 2a and 2b. Cyclic voltamme-
try studies were conducted on solution samples of 2a and 2b
to probe their outer-sphere redox properties. The juxtaposi-
tion of a low-valent rhodium metal center and a reducible
phdi ligand in complexes 2a and 2b suggests that both
oxidative and reductive processes might be accessible. The
cyclic voltammogram of 2a in THF is shown in Figure 2a. A
reversible one-electron reduction was observed at -2.03 V
along with a reversible one-electron oxidation at 0.06 V (all
potentials vs [Cp2Fe]þ/0).31 Furthermore, an irreversible
oxidative process and a partially reversible reductive process
can be observed near the edges of the solvent window at
þ0.79 and -3.08 V, respectively. The cyclic voltammogram
of 2b, shown in Figure 2b, is similar, showing a reversible
The UV-vis spectra of 2a and 2b, shown in Figure 1, dis-
play several intense transitions throughout the visible region
of the spectrum. In the case of 2a, two strong charge-transfer
transitions were observed at 362 nm (23 000 M-1 cm-1) and
593 nm (22 700 M-1 cm-1) along with two or more shoulders
in the 400-450 nm range and a weak transition at 796 nm
(546 M-1 cm-1). The UV-vis spectrum of 2b is similar,
(30) Bernskoetter, W. H.; Lobkovsky, E.; Chirik, P. J. Organometallics
2005, 24, 6250–6259.
(31) All potentials referenced to the [Cp2Fe]þ/0 redox couple using
Cp*2Fe (E01/2 = -0.49 V) as an internal standard.

18 Inorganic Chemistry, Vol. 50, No. 1, 2011
Shaffer et al.
Figure 4. X-band EPR spectrum of [2a]- in THF at 77 and 298 K
(inset). The best fit for the 77 K spectrum (dashed line) was obtained with
gx = 2.39, gy = 1.97, and gz = 1.94.
suggestive of a doubly reduced phdi ligand in [2a]-. Similar
C-N and C-C bond distances in redox-active diimine
ligands have been taken as evidence for the reduction of
the ligand to the semiquinonate or even the catecholate
oxidation state.15,33,34 For comparison, rhodium complexes
Figure 3. ORTEP plot of the rhodium anion [2a]-. Ellipsoids are shown
at 50% probability. The [K(18-crown-6)] cation, a THF molecule, and all
carbon-bound hydrogen atoms have been omitted for clarity.
one-electron reduction at -1.79 V. The first one-electron
oxidation, at þ0.37 V, is only partially reversible, as is the
second reduction at -2.69 V. The second oxidative process
(þ0.75 V) is irreversible, as with 2a. Finally, the cyclic
voltammogram of 2b shows daughter reductions at -0.27
and -0.6 V spawned from the second irreversible oxidation
at þ0.75 V.32 The reductive processes observed for 2a and 2b
are consistent with the cyclic voltammetry results obtained
for the free phdi ligand, which showed a partially rever-
sible reduction at -1.94 V and an irreversible reduction at
-2.66 V. As expected, the incorporation of CF3-substituted
nacnac ligand into the rhodium coordination sphere makes
the complex both easier to reduce and harder to oxidize,
consistent with a more electron-deficient metal center.
Outer-Sphere Oxidation and Reduction of 2a. The
reversibility of the first oxidation and reduction processes
in the cyclic voltammogram of 2a suggested that the pro-
ducts of these process species might be isolable. The chemical
reduction of 2a requires a powerful reductant and was
achieved by reaction with 1 equiv of KC8. The addition of
18-crown-6 to a THF solution of the reduction product
afforded green crystals of [K(18-crown-6)][2a] in 96% yield.
Single crystals of [K(18-crown-6)][2a] were grown from a
saturated THF solution and analyzed by X-ray diffraction.
The structure of the rhodium anion is depicted as an ORTEP
plot in Figure 3. Table 2 lists selected bond distances and
angles for the rhodium complex anion. The Rh-N bond
with phdi ligands typically show a CdN bond distance of
13c,35
˚
˚
1.29 A and a C-C bond distance of 1.49 A.
Complex [2a]- was characterized by solution EPR and
solid-state IR spectroscopy. The IR spectrum showed the
N-H stretch at 3356 cm-1; however, no clearly assignable
CdN stretch was observed. Figure 4 shows the solution
X-band EPR spectrum at 77 and 298 K (inset), obtained for
crystals of [K(18-crown-6)][2a] dissolved in THF. At 298 K,
the EPR spectrum of [2a]- shows a broad signal at g = 2.09.
Upon cooling to 77 K, the signal becomes rhombic with
features at g = 2.39, 1.97, and 1.94. For comparison, related
square-planar rhodium(I) semiquinonate complexes, [sq•]-
RhI(CO)2 ([sq•] = 3,5-di-tert-butyl-1,2-semiquinonate and
substituted derivatives), have been reported to show iso-
tropic EPR spectra (g = 2.002) with well-defined hyperfine
coupling to the semiquinonate ligand substituents, indicative
of the unpaired electron residing on the redox-active
ligand.36 Similarly, rhodium(0) complexes typically show
isotropic or axial EPR spectra with g values close to 2.0 at
both room and low temperature.37-40 In contrast, rhodium-
(II) complexes typically show rhombic EPR spectra at low
€
€
(33) (a) Justel, T.; Bendix, J.; Metzler-Nolte, N.; Weyhermuller, T.;
Nuber, B.; Wieghardt, K. Inorg. Chem. 1998, 37, 35–43. (b) Chopek, K.;
Bothe, E.; Neese, F.; Weyherm€uller, T.; Wieghardt, K. Inorg. Chem. 2006, 45,
6298–6307.
(34) Bhattacharya, S.; Gupta, P.; Basuli, F.; Pierpont, C. G. Inorg. Chem.
2002, 41, 5810–5816.
distances to the nacnacCH3 ligand in {[dpp-nacnacCH3]Rh-
(35) (a) Kisko, J. L.; Barton, J. K. Inorg. Chem. 2000, 39, 4942–4949. (b)
Krotz, A. H.; Barton, J. K. Inorg. Chem. 1994, 33, 1940–1947.
(36) (a) Bubnov, M. P.; Nevodchikov, V. I.; Fukin, G. K.; Cherkasov,
V. K.; Abakumov, G. A. Inorg. Chem. Commun. 2007, 10, 989–992. (b)
Abakumov, G. A.; Nevodchikov, V. I. Dokl. Acad. Nauk 1982, 266, 1407–1410.
-
(phdi)}- ([2a] ) are significantly shorter (2.01 A) than the
˚
analogous Rh-N bond distances to the nacnacCH3 ligand in
the neutral dicarbonyl complex 1a (2.05 A). Other metrical
˚
parameters for the nacnacCH3 ligand of [2a]- match up well
ꢂ
ꢂ
(37) Caldararu, H.; Dearmond, M. K.; Hanck, K. W.; Sahini, V. E. J. Am.
Chem. Soc. 1976, 98, 4455–4457.
with those in 1a. Within the phdi ligand, the C-N distances
are long (1.35 A), consistent with the reduction of the diimine
(38) (a) Kunin, A. J.; Nanni, E. J.; Eisenberg, R. Inorg. Chem. 1985, 24,
1852–1856. (b) Mueller, K. T.; Kunin, A. J.; Greiner, S.; Henderson, T.; Kreilick,
R. W.; Eisenberg, R. J. Am. Chem. Soc. 1987, 109, 6313–6318.
(39) Longato, B.; Coppo, R.; Pilloni, G.; Corvaja, C.; Toffoletti, A.;
Bandoli, G. J. Organomet. Chem. 2001, 637, 710–718.
˚
functionality of the ligand. Furthermore, the C(30)-C(43)
˚
distance of 1.41 A is consistent with an aromatic C-C bond
€
(40) (a) Deblon, S.; Liesum, L.; Harmer, J.; Schonberg, H.; Schweiger, A.;
€
(32) Daughter peaks are also observed on the second cyclic scan of 2a at
-0.45, -0.85, and -1.28 V.
Grutzmacher, H. Chem.;Eur. J. 2002, 8, 601–611. (b) de Bruin, B.; Russcher,
J. C.; Gr€utzmacher, H. J. Organomet. Chem. 2007, 692, 3167–3173.

Article
Inorganic Chemistry, Vol. 50, No. 1, 2011 19
nucleophilic metal center coordinated to an electrophilic or
reducible ligand. Typically, square-planar d8 transition-metal
complexes are nucleophilic because of the nonbonding elec-
tron pair held within the metal dz2 orbital. On the other hand,
the coordinated phdi ligand, an R-diimine analogous to
phenanthrenequinone, has a low-energy π* orbital capable
of accepting two electrons. The juxtaposition of these two
motifs may offer the ability to design new catalysts that take
advantage of both the nucleophilic properties and the ability
to act as two-electron oxidants. In order to develop our
understanding of how metal and ligand fragments can work
together, we examined the structural, spectroscopic, and
electronic properties of 2a and 2b.
Comparisons between 2a and 2b were used to determine to
what extent thenacnacligand impacts the electronic structure
of the Rh-phdi unit. According to the CO stretching frequen-
cies in 1a and 1b, substitution of CF3 groups into the nacnac
backbone has a strong impact on the ability of the rhodium
center to participate in π-backbonding with the CO ligands.
The nacnac ligands should have an analogous impact in 2a
and 2b and modulate the electronic interaction between the
rhodium centers and the redox-active phdi ligand. While the
CdN stretching frequencies of 2a and 2b cannot be used to
probe the difference between 2a and 2b, the N-H stretching
frequency decreases upon substitution of the fluorinated
ligand, suggesting that νN-H is a strong indicator of Rh f
phdi electron donation. Cyclic voltammetry data support the
contention that 2b is more electron-deficient than 2a, with
one-electron-oxidation and -reduction potentials shifting by
þ310 and þ240 mV, respectively, for 2b relative to 2a. These
data suggest that subtle changes to an auxiliary ligand
coordinated to the Rh-phdi fragment can have dramatic
effects on the electronic properties of the overall complex.
The key aspect of these studies was to probe the nature of the
Rh-phdi interaction. The “formal” oxidation state assignment
for 2a and 2b would be that of a rhodium(I) coordinated to a
neutral phdi ligand (A in Chart 2);43 however, this formal
oxidation state assignment does not necessarily correspond to
the “experimental” oxidation state of the complex.44 An alter-
native, electronic isomer45-47 for 2a and 2b would be an
antiferromagnetically coupled, biradical state with a rhodium-
(II) coordinated to a [phdisq•]- ligand (B in Chart 2). In the
absence of high-resolution structural data for 2a or 2b, distin-
guishing between A and B is difficult,48 and our spectroscopic
Figure 5. X-band EPR spectrum of [2a]þ in CH2Cl2 at 298 and 77 K
(inset).
temperatures.41 The EPR parameters for [2a]- are similar to
those of the recently reported planar rhodium dianion [Rh-
{bdt}2]2- ({bdt}2-=3,6-bis(trimethylsilyl)benzene-1,2-dithio-
late), which was assigned a rhodium(II) oxidation state based
on a rhombic EPR spectrum at 25 K with g = 2.50, 2.00, and
1.97.42
One-electron oxidation of 2a was possible using tri-p-
tolylaminium as an oxidant. While the electrochemical data
shown in Figure 2 suggest that silver salts would be compe-
tent oxidants for 2a, reactions using these reagents did not
lead to clean oxidation of the rhodium complex. On the other
hand, [N(p-tolyl)3][BF4] reacted cleanly with 2a to provide
the oxidized product as a purple paramagnetic solid, {[dpp-
nacnacCH3]Rh(phdi)}{BF4} ([2a][BF4]) in 82% yield. The
new species was characterized by solid-state IR spectroscopy,
UV-vis electronic spectroscopy, and solution-phase EPR
spectroscopy. The N-H stretch is shifted to 3296 cm-1 in the
IR spectrum. The UV-vis spectrum of {[dpp-nacnacCH3]-
Rh(phdi)}þ ([2a]þ) is similar to that of 2a, showing two
intense transitions at 327 nm (16 000 M-1 cm-1) and 571 nm
(19 700 M-1 cm-1). The X-band EPR spectrum of [2a]þ at
298 K, shown in Figure 5, displayed an isotropic 15-line
signal centered at g= 2.004. Upon cooling to 77 K, this com-
plex signal simplified to a single, sharp, isotropic signal cen-
tered at g = 2.003. The EPR data for [2a]þ is again indicative
of an S = 1/2 complex; however, the g value and isotropic
nature of the signal are not consistent with localization of the
unpaired electron on the metal center to afford a rhodium(II)
complex. Instead, the EPR signal, with sharp hyperfine
couplings that disappear at low temperature, is consistent
with delocalization of the unpaired electron onto the organic
ligand framework.
(43) (a) Hegedus, L. S. In Transition Metals in the Synthesis of Complex
Organic Molecules; University Science Books: Mill Valley, CA, 1994; p 3. (b)
Shriver, D.; Atkins, P. In Inorganic Chemistry, 3rd ed.; W. H. Freeman: New
York, 1999; p 69.
€
(44) (a) Jorgensen, C. K. In Oxidation Numbers and Oxidation States;
Springer: Heidelberg, Germany, 1969. (b) Chaudhuri, P.; Verani, C. N.; Bill, E.;
Bothe, E.; Weyherm€uller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213–
2223.
(45) We use the term “electronic isomer” to indicate a specific distribution
of electrons for the complex as represented by a given Lewis structure. The
terms “valence tautomer” (see ref 46) and “electromer” (see ref 47) also have
been used in the literature to indicate species with the same arrangement of
nuclei but different distributions of electrons.
Discussion
Complexes 2a and 2b present an interesting opportunity
to evaluate the electronic structure and reactivity of a
(41) (a) Wayland, B. B.; Sherry, A. E.; Bunn, A. G. J. Am. Chem. Soc. 1993,
115, 7675–7684. (b) Dixon, F. M.; Farrell, J. R.; Doan, P. E.; Williamson, A.;
Weinberger, D. A.; Mirkin, C. A.; Stern, C.; Incarvito, C. D.; Liable-Sands, L. M.;
Zakharov, L. N.; Rheingold, A. L. Organometallics 2002, 21, 3091–3093. (c) Gerisch,
M.; Krumper, J. R.; Bergman, R. G.; Tilley, T. D. Organometallics 2003, 22, 47–58.
(d) Hetterscheid, D. G. H.; Klop, M.; Kicken, R. J. N. A. M.; Smits, J. M. M.; Reijerse,
E. J.; de Bruin, B. Chem.;Eur. J. 2007, 13, 3386–3405.
(46) Pierpont, C. G. Coord. Chem. Rev. 2001, 216-217, 99–125.
(47) Bally, T. Nat. Chem. 2010, 2, 165–166.
(48) Attempts to resolve the issue of the two electronic isomers through
high-level density functional theory calculations have led to ambiguous
results. Both closed-shell RhIphdi and biradical RhIIphdisq• structures have
been located; however, the energy differences are within the error of the
calculation. A further complication arises in that the electronic structures of
each electronic isomer show a strong dependence on the functional employed
for the calculation.
€
(42) Benedito, F. L.; Petrenko, T.; Bill, E.; Weyhermuller, T.; Wieghardt,
K. Inorg. Chem. 2009, 48, 10913–10925.

20 Inorganic Chemistry, Vol. 50, No. 1, 2011
Shaffer et al.
Chart 2. Possible Electronic Isomers for 2a and [2a]-
isomer Bof 2a. The ligand-based reduction of Bwould generate
isomer E, comprising a rhodium(II) metal center and a fully
reduced [phda]2- ligand.
Both EPR spectroscopic and X-ray structural data support E
as the best description for [2a]-. Most proposed rhodium(0)
complexes in the literature are highly delocalized, with the
unpaired electron spread out over the metal and ligands.40 Both
electronic isomers C and D would be expected to give rise to an
isotropic or axial EPR spectrum at high and low temperatures,
with g value(s) very close to 2.0, as has been observed for other
delocalized rhodium(0) complexes and square-planar metal
complexes with semiquinonate ligands.37-40 Conversely, anion
[2a]- displays an isotropic EPR spectrum at room temperature
that becomes rhombic upon cooling to 77 K. The observed g
values and degree of rhombicity are consistent with EPR
spectra of d7 rhodium(II) complexes reported in the
literature.41,42 It is surprising that the reduction of 2a affords
an anion with clear rhodium(II) character, but two factors
could work to stabilize electronic isomer E relative to C/D.
First, the [phda]2- ligand is a closed-shell phenanthrene ring
system, which would benefit from full aromatic stabilization
energy. Second, E places two electrons in a π-type ligand orbital
rather than the dz2 orbital that is localized on the rhodium
center. Placing two electrons in a π-type orbital should result in
less electron-pair repulsion because the electrons are delocalized
over a larger volume.
data provide scant evidence for favoring biradical B over the
formal electronic isomer A. Neither complex 2a nor 2b shows
evidence for paramagnetism that would arise from breaking of
the antiferromagnetic interaction in biradical isomer B. Simi-
larly, the NMR spectra of 2a and 2b show no evidence for
paramagnetically shifted resonances that might occur for a
[nacnac]- ligand coordinated to a rhodium(II) center; further-
more, the [nacnac]- features in the NMR spectra of 2a and 2b
are very similar to those in the spectra of 1a and 1b, where a
biradical representation like B is not plausible. The resonances
for the redox-active ligand in 2a and 2b also are consistent with
those for the free phdi ligand. While the lowest-energy transi-
tion in the UV-vis spectra of 2a and 2b shows barely any
change between 2a and 2b, opening the possibility of an
intraligand charge-transfer transition within an open-shell
[phdisq•]- ligand, the energy of this transition is in line with
other rhodium(I) R-diimine complexes,49 suggesting a better
assignment as metal-to-diimine-ligand charge transfer.
Conclusion
Definitive assignments of experimental oxidation states for
2a and 2b are problematic. Given the results of related work
on square-planar d8 complexes of palladium and platinum,50,51
it is tempting to assign a closed-shell, RhIphdi ground-state
electron configuration for 2a and 2b. On the other hand,
square-planar rhodium complexes with redox-active dithiolene
ligands seem to show a proclivity for the d7 rhodium(II)
oxidation state.42 While there is no definitive structural or
spectroscopic evidence for assigning the RhII[phdisq•]- electro-
nic isomer (B of Chart 2) as the ground state for 2a and 2b, the
electrochemically reversible reduction of 2a and characteriza-
tion of the product, [2a]-, as a RhII[phda] complex suggest that
2a has at least some biradical character in the ground state.
Hence, the individual RhIphdi (A) and RhII[phdisq•] (B)
electronic isomers of 2a are too simple and a limitation of
Lewis structures in representing the electronic structure of
complex molecules. Instead, the ground state of 2a likely has
character of both electronic isomers A and B and should be
represented as a multiconfiguration ground state.51 Within the
framework of Lewis structures, this multiconfigurational char-
acter can be represented with resonance, as is often done to
overcome the limitations of Lewis structures in explaining the
reactivity of organic molecules.
Characterization of the anion [2a]- is straightforward be-
cause of the characteristic S=1/2 EPR signature of the reduced
complex and proves the accessibility of a rhodium(II) oxidation
state. The extra electron in the anion [2a]- should be contained
within a molecular orbital localized on the phdi ligand or on the
rhodium metal center. The Lewis structures for three potential
electronic isomers of [2a]- are shown in Chart 2. If 2a is viewed
as RhIphdi isomer A, reduction at the metal center would leave
the phdi ligand in the diimine oxidation state while reducing the
metal center to a d9 rhodium(0) isomer, C. The other possibility
for the reduction of A is to place the electron into the π system
of the phdi ligand, reducing it from a diimine to a diiminose-
miquinonate ligand and leaving the metal as a d8 rhodium(I)
isomer, D. The same electronic isomer of [2a]-, D, would
be generated by the metal-based reduction of RhII[phdisq•],
Undoubtedly, coordination of the phdi ligand to a square-
planar rhodium(I) generates a molecular platform with
interesting electronic properties. As expected, the proximity
of metal- and phdi-based valence orbitals leads to electronic
structures that include a wealth of charge-transfer character
and that are capable of reversible outer-sphere oxidation and
(50) (a) Ghosh, P.; Begum, A.; Herebian, D.; Bothe, E.; Hildenbrand, K.;
€
Weyhermuller, T.; Wieghardt, K. Angew. Chem., Int. Ed. 2003, 42, 563–567.
(b) Herebian, D.; Bothe, E.; Neese, F.; Weyherm€uller, T.; Wieghardt, K. J. Am.
Chem. Soc. 2003, 125, 9116–9128.
(51) Ray, K.; Weyhermuller, T.; Neese, F.; Wieghardt, K. Inorg. Chem.
2005, 44, 5345–5360.
(49) (a) Shinozaki, K.; Takahashi, N. Inorg. Chem. 1996, 35, 3917–3924.
(b) Fordyce, W. A.; Crosby, G. A. Inorg. Chem. 1982, 21, 1023–1026. (c) Kaim,
W.; Reinhardt, R.; Sieger, M. Inorg. Chem. 1994, 33, 4453–4459. (d) Ladwig, M.;
Kaim, W. J. Organomet. Chem. 1991, 419, 233–243.

Article
Inorganic Chemistry, Vol. 50, No. 1, 2011 21
reduction processes. Regardless of the electronic configura-
tion assignment that is made for the members of the [2a]þ, 2a,
and [2a]- redox series, the stability of the [nacnac]Rh(phdi)
core through multiple one-electron processes motivates
further studies into the reactivity of the Rh-phdi platform.
is an Alfred P. Sloan Foundation Fellow and a Camille
Dreyfus Teacher-Scholar. The authors thank Heraeus for
donation of the rhodium starting material.
Supporting Information Available: X-ray crystal structure,
listing of bond lengths, and cyclic voltammetry data for 1a. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Acknowledgment. This work was supported by the UCI
School of Physical Sciences Center for Solar Energy. A.F.H.