Solvent and temperature induced switching between structural isomers of RhI phosphinoalkyl thioether (PS) complexes

Article abstract of DOI:10.1021/ic101021t

To develop functional systems based on the weak-link approach (WLA), it is important to understand how solvent and ligand binding strength alter the coordination geometry of complexes formed from this method. A series of phosphinoalkyl thioether (PS) hemi

Full text of DOI:10.1021/ic101021t

7188 Inorg. Chem. 2010, 49, 7188–7196
DOI: 10.1021/ic101021t
Solvent and Temperature Induced Switching Between Structural Isomers of RhI
Phosphinoalkyl Thioether (PS) Complexes
Michael J. Wiester, Adam B. Braunschweig, Hyojong Yoo, and Chad A. Mirkin*
Department of Chemistry and the International Institute for Nanotechnology, Northwestern University,
Evanston, Illinois 60208-3113
Received May 20, 2010
To develop functional systems based on the weak-link approach (WLA), it is important to understand how solvent and
ligand binding strength alter the coordination geometry of complexes formed from this method. A series of
phosphinoalkyl thioether (PS) hemilabile ligands with varying electron donating abilities were synthesized and
incorporated into homoligated RhI(PS)2Cl complexes to help understand the effects of solvent and ligand binding
strength on the preferred coordination modes. The switching between closed and semiopen structural isomers of these
RhI(PS)2Cl complexes was studied by variable temperature 31P NMR spectroscopy in different solvent mixtures of
CH2Cl2 and tetrahydrofuran (THF) to obtain thermodynamic parameters (ΔG°, ΔH°, TΔS°, and Keq). The isomers
differ in the position of the chloride counterion. In the closed isomer, the Cl- anion occupies the outer coordination
sphere, while in the semiopen isomer, the Cl- has moved inner sphere and displaced one of the Rh-S bonds. The
closed isomer is favored in CH2Cl2 and the semiopen isomer is favored in THF. The preference for either isomer at
equilibrium depends on the solvent polarity, based upon the ENT solvent polarity scale, as was determined from 15
different solvents, with more polar solvents favoring the closed isomer. The isomer preference also depends on the
electron donating ability of the group attached to the sulfur of the PS ligand, with electron donating groups favoring the
closed isomers and electron withdrawing groups favoring the semiopen isomers. The formation of the semiopen
isomer from the closed isomer is entropically favored but enthalpically disfavored under all conditions studied.
Elucidation of the principles and environments that determine the equilibrium between the two isomers will aid in the
design of functional complexes prepared by the WLA.
Introduction
used in ELISA- and PCR-like signal amplification and detec-
tion schemes for small molecules and elemental ions.3,4 In the
WLA, flexible hemilabile ligands containing chelating metal
binding sites consisting of a diphenylphosphine moiety and a
second heteroatom metal binder (PX, where X = O, S, or N)
are combined with tetra-coordinate metal ions, such as RhI,
PtII, PdII, or CuI, to form supramolecular coordination
complexes such as macrocycles and tweezers.1 Homoligated
complexes, in which the two ligands bound to the metal center
are identical, have previously been interconverted between
two coordination geometries, open (1) and closed (2), through
small molecule reactions that occur at metal centers
(Scheme 1A).1 In this process, the weaker metal-heteroatom
bonds are cleaved by incoming small molecule ligands (such
The weak-link approach (WLA) is a method for preparing
macrocyclic, tweezer, and triple-decker coordination comp-
lexes,1,2 which behave as allosteric enzyme mimics that can be
*To whom correspondence should be addressed. E-mail: chadnano@
northwestern.edu.
(1) (a) Holliday, B. J.; Mirkin, C. A. Angew. Chem., Int. Ed. 2001, 40,
2022–2043. (b) Gianneschi, N. C.; Masar, M. S., III; Mirkin, C. A. Acc. Chem.
Res. 2005, 38, 825–837. (c) Oliveri, C. G.; Ulmann, P. A.; Wiester, M. J.; Mirkin,
C. A. Acc. Chem. Res. 2008, 41, 1618–1629. (d) Wiester, M. J.; Mirkin, C. A.
Inorg. Chem. 2009, 48, 8054–8056. (e) Wiester, M. J.; Ulmann, P. U.; Mirkin,
C. A., Angew. Chem., Int. Ed. 2010, DOI: 10.1002/anie.201000380.
(2) Reviews on other coordination chemistry approaches for preparing
supramolecular assemblies: (a) Caulder, D. L.; Raymond, K. N. Acc. Chem.
Res. 1999, 32, 975–982. (b) Fujita, M.; Tominaga, M.; Hori, A.; Therrien, B. Acc.
Chem. Res. 2005, 38, 369–378. (c) Lee, S. J.; Lin, W. Acc. Chem. Res. 2008, 41,
521–537.
(3) (a) Gianneschi, N. C.; Cho, S.-H.; Nguyen, S. T.; Mirkin, C. A.
Angew. Chem., Int. Ed. 2004, 43, 5503–5507. (b) Oliveri, C. G.; Gianneschi,
N. C.; Nguyen, S. T.; Mirkin, C. A.; Stern, C. L.; Wawrzak, Z.; Pink, M. J. Am.
Chem. Soc. 2006, 128, 16286–16296. (c) Yoon, H. J.; Heo, J.; Mirkin, C. A.
J. Am. Chem. Soc. 2007, 129, 14182–14183. (d) Yoon, H. J.; Mirkin, C. A.
J. Am. Chem. Soc. 2008, 130, 11590–11591. (e) Ulmann, P. A.; Braunschweig,
A. B.; Lee, O.-S.; Wiester, M. J.; Schatz, G. C.; Mirkin, C. A. Chem. Commun.
2009, 5121–5123.
(4) Other examples of synthetic allosteric systems and enzyme mimics:
(a) Thordarson, P.; Bijsterveld, E. J. A.; Elemans, J. A. A. W.; Kasak, P.;
Nolte, R. J. M.; Rowan, A. E. J. Am. Chem. Soc. 2003, 125, 1186–1187.
(b) Pall, T.; Ruud, G. E. C.; Johannes, A. A. W. E.; Paul, J. T.; Johan, V.; Alan,
E. R.; Roeland, J. M. N. Angew. Chem., Int. Ed. 2004, 43, 4755–4759.
(c) Yoshizawa, M.; Tamura, M.; Fujita, M. Science 2006, 312, 251–254.
(d) Pluth, M. D.; Bergman, R. G.; Raymond, K. N. Science 2007, 316, 85–88.
(e) Veling, N.; Thomassen, P. J.; Thordarson, P.; Elemans, J. A. A. W.; Nolte,
R. J. M.; Rowan, A. E. Tetrahedron 2008, 64, 8535–8542.
r
pubs.acs.org/IC
Published on Web 07/09/2010
2010 American Chemical Society

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 7189
Scheme 1. (A) Homoligated Tweezer Complexes Formed via the WLA and (B) Heteroligated Tweezer Complexes Formed via the WLA and HILR
Reaction, Where A and B Are Two Different Functional Groups
as Cl-, CO, MeCN, or CN-), while the stronger phosphine-
metal bonds remain intact, resulting in more flexible open
structures. Heteroligated complexes, in which two different
ligands (either PS/PO or PS/PS0; PS, phosphinoalkyl thio-
ether) are complexed to the metal, also can be prepared by the
WLA and a halide-induced ligand rearrangement (HILR)
reaction,5 and interconverted between multiple geometries
(3-6) through small molecule reactions that occur at the
metal hinge sites (Scheme 1B). If only one of the metal-
heteroatom bonds is cleaved, the molecule adopts a semiopen
conformation (4 and 5, Scheme 1B), termed the semiopen
isomer.1c WLA complexes have only been interconverted
between different geometries by the addition or removal of
small molecules,1 whereas other dynamic molecular systems6a-e
have been actuated by changes in solvent,7a temperature,7b or
irradiation,7c as well as by the addition of small molecules.7d,e
While Cl- is often utilized with complexes prepared by the
WLA as a structural regulatory ligand that binds to the metal
center to create more flexible open structures, in some cases
Cl- is present as a counterion.1,3 This indicated to us that
under certain conditions Cl- may move from outer to inner
sphere coordination or vice versa, thereby interconverting the
two closed and semiopen structural isomers (eq 1). Herein we
report the observation that by altering solvent and tempera-
ture we can induce the opening and closing of various
homoligated RhI(PS)2Cl complexes as a result of the Cl-
anions moving from the outer to inner sphere of the coordi-
nation complex (eq 1). When designing functional architec-
tures based on the WLA, it is important to understand how
different environmental conditions and ligand properties
affect the geometry and ultimately the performance of those
systems. Because the coordination geometry is central to all
of the functional complexes prepared via the WLA, we have
systematically studied the thermodynamic basis for this
switching by variable temperature (VT) 31P NMR spectros-
copy and related the preference of the Cl- anion for inner and
outer sphere coordination to solvent polarity and electron
donating ability of the thioether ligands. Taken together,
these complexes are excellent model systems for understand-
ing how to design the much more sophisticated and compli-
cated WLA catalysts and allosteric enzyme mimics.
(5) (a) Ovchinnikov, M. V.; Brown, A. M.; Liu, X.; Mirkin, C. A.;
Zakharov, L. N.; Rheingold, A. L. Inorg. Chem. 2004, 43, 8233–8235.
(b) Brown, A. M.; Ovchinnikov, M. V.; Stern, C. L.; Mirkin, C. A. J. Am. Chem.
Soc. 2004, 126, 14316–14317. (c) Brown, A. M.; Ovchinnikov, M. V.; Mirkin, C. A.
Angew. Chem., Int. Ed. 2005, 44, 4207–4209. (d) Jeon, Y. M.; Heo, J.; Brown, A. M.;
Mirkin, C. A. Organometallics 2006, 25, 2729–2732. (e) Ulmann, P. A.; Brown,
A. M.; Ovchinnikov, M. V.; Mirkin, C. A.; Di Pasquale, A. G.; Rheingold, A. L.
Chem.—Eur. J. 2007, 13, 4529–4534. (f) Oliveri, C. G.; Heo, J.; Nguyen, S. T.;
Mirkin, C. A.; Wawrzak, Z. Inorg. Chem. 2007, 46, 7716–7718. (g) Oliveri, C. G.;
Nguyen, S. T.; Mirkin, C. A. Inorg. Chem. 2008, 47, 2755–2763.
(6) (a) Collin, J.-P.; Dietrich-Buchecker, C.; Gavina, P.; Jimenez-Molero,
M. C.; Sauvage, J.-P. Acc. Chem. Res. 2001, 34, 477–487. (b) Feringa, B. L.
Acc. Chem. Res. 2001, 34, 504–513. (c) Braunschweig, A. B.; Northrop, B. H.;
Stoddart, J. F. J. Mater. Chem. 2006, 16, 32–44. (d) Kay, E. R.; Leigh, D. A.;
Zerbetto, F. Angew. Chem., Int. Ed. 2007, 46, 72–191. (e) Olson, M. A.;
Braunschweig, A. B.; Ikeda, T.; Fang, L.; Trabolsi, A.; Slawin, A. M. Z.; Khan,
S. I.; Stoddart, J. F. Org. Biomol. Chem. 2009, 7, 4391–4405.
ꢀ
ꢀ
(7) (a) Leigh, D. A.; Morales, M. A. F.; Perez, E. M.; Wong, J. K. Y.; Saiz,
C. G.; Slawin, A. M. Z.; Carmichael, A. J.; Haddleton, D. M.; Brouwer, A. M.;
ꢀ
Buma, W. J.; Wurpel, G. W. H.; Leon, S.; Zerbetto, F. Angew. Chem., Int. Ed.
2005, 44, 3062–3067. (b) Jeppesen, J. O.; Perkins, J.; Becher, J.; Stoddart, J. F.
Angew. Chem., Int. Ed. 2001, 40, 1216–1221. (c) Irie, M. Bull. Chem. Soc. Jpn.
2008, 81, 917–926. (d) Marlin, D. S.; Cabrera, D. G.; Leigh, D. A.; Slawin, A. M. Z.
Angew. Chem., Int. Ed. 2006, 45, 1385–1390. (e) Marlin, D. S.; Cabrera, D. G.;
Leigh, D. A.; Slawin, A. M. Z. Angew. Chem., Int. Ed. 2006, 45, 77–83.

7190 Inorganic Chemistry, Vol. 49, No. 15, 2010
Wiester et al.
(s, CH3, 3H), 2.29 (s, CH3, 3H). 13C{1H} NMR (CD2Cl2), δ =
138.4 (s, ArC), 138.0 (d, JC-P = 13.6 Hz, ArC), 136.2 (s, ArC),
132.6 (d, JC-P = 18.8 Hz, ArC), 131.2 (s, ArC), 131.0 (s, ArC),
129.7 (s, ArC), 128.7 (s, ArC), 128.5 (d, JC-P = 6.6 Hz, ArC),
Experimental Section
General Methods. All reactions were carried out under an
inert atmosphere of nitrogen using standard Schlenk techniques
or an inert-atmosphere glovebox unless otherwise noted. Tetra-
hydrofuran (THF), dichloromethane (CH2Cl2), diethyl ether,
and hexanes were purified according to published methods.8
Chlorocyclohexane, 2,4,6-trimethylpyridine, 1,2-dimethoxyethane,
diethylene glycol dimethyl ether, methyl acetate, triethylene glycol
dimethyl ether, 1,1,2,2-tetrachloroethane, dibromomethane,
1-decanol, 1-nonanol, and 2-chlorophenol were purchased from
127.1 (s, ArC), 30.0 (d, JC-P = 22.0 Hz, CH2), 28.9 (d, JC-P
=
15.0 Hz, CH2), 20.6 (s, CH3), 20.1 (s, CH3). 31P{1H} NMR
(CD2Cl2) δ=-17.3 (s). APPIMS (m/z): Calcd: 351.1331 [MþH]þ.
Found: 351.1318.
(2-(p-Tolylthio)ethyl)diphenylphosphine (d). 1HNMR (CD2Cl2),
δ = 7.33-7.39 (br m, Ph2P, 10 H), 7.16 (d, JH-H = 8.4 Hz, ArH,
2 H), 7.09 (d, JH-H = 8.0 Hz, ArH, 2 H), 2.91 (m, PCH2CH2S,
2 H), 2.33 (m, PCH2CH2S, 2 H), 2.30 (s, CH3, 3H). 13C{1H}
NMR (CD2Cl2), δ = 137.9 (d, JC-P = 13.8 Hz, ArC), 136.4 (s,
ArC), 132.7 (d, JC-P = 18.8 Hz, ArC), 132.1 (s, ArC), 130.0 (s,
ArC), 129.6 (s, ArC), 128.7 (s, ArC), 128.5 (d, JC-P = 6.6 Hz,
ArC), 30.8 (d, JC-P = 22.3 Hz, CH2), 28.1 (d, JC-P = 15.1 Hz,
CH2), 20.7 (s, CH3). 31P{1H} NMR (CD2Cl2) δ = -16.4 (s).
APPIMS (m/z): Calcd: 337.1174 [MþH]þ. Found: 337.1166.
(2-(Phenylthio)ethyl)diphenylphosphine (e). 1HNMR(CD2Cl2),
δ = 7.33-7.41 (br m, Ph2P, 10 H), 7.17-7.27 (br m, ArH, 5 H),
2.96 (m, PCH2CH2S, 2 H), 2.36 (m, PCH2CH2S, 2 H). 13C{1H}
NMR (CD2Cl2), δ = 137.9 (d, JC-P = 13.8 Hz, ArC), 136.0 (s,
ArC), 132.7 (d, JC-P = 18.8 Hz, ArC), 129.1 (s, ArC), 128.9 (s,
ArC), 128.8 (s, ArC), 128.5 (d, JC-P = 6.6 Hz, ArC), 126.0 (s,
ArC), 30.1 (d, JC-P = 22.7 Hz, CH2), 28.1 (d, JC-P = 15.1 Hz,
CH2). 31P{1H} NMR (CD2Cl2) δ = -17.2 (s). APPIMS (m/z):
Calcd: 323.1023 [MþH]þ. Found: 323.1009.
˚
Alfa Aesar, dried over 4 A molecular sieves and deoxygenated
prior to use. Deuterated solvents were purchased from Cam-
bridge Isotope Laboratories Inc. and used as received. All other
chemicals were used as received from Aldrich Chemical Co. or
TCI America unless otherwise stated.
All NMR spectra were recorded on a Bruker Avance III 400
MHz FTNMR spectrometer, and all chemical shifts are re-
ported in parts per million. H NMR spectra were referenced
relative to residual solvent proton resonances in deuterated
solvents. 31P{1H} and 31P (161 MHz) spectra were referenced
relative to an external 85% H3PO4 standard. 13C{1H} (101 MHz)
spectra were referenced relative to residual solvent resonances.
19F{1H} (376 MHz) spectra were referenced relative to an external
CFCl3 in CDCl3 standard. High resolution atmospheric pres-
sure photoionization mass spectra (APPIMS) and high resolution
electrospray ionization mass spectra (ESIMS) were recorded on
an Agilent 6210 LC-TOF with an Agilent 1200 HPLC introduc-
tion mass spectrometer system.
1
(2-(4-Fluorophenylthio)ethyl)diphenylphosphine (f). 1H NMR
(CD2Cl2), δ = 7.38-7.42 (br m, Ph2P, 10 H), 7.31 (m, ArH,
2 H), 7.05 (d, JH-H = 7.0 Hz, ArH, 2 H), 2.97 (m, PCH2CH2S,
2 H), 2.39 (m, PCH2CH2S, 2 H). 13C{1H} NMR (CD2Cl2), δ =
161.8 (d, JC-F = 244.4 Hz, ArC), 137.8 (d, JC-P = 13.8 Hz,
ArC), 132.6 (d, JC-P = 18.8 Hz, ArC), 132.4 (d, JC-F = 8.0 Hz,
ArC), 130.9 (s, ArC), 128.7 (d, JC-F = 9.5 Hz, ArC), 128.5 (d,
Preparation of PS Ligands. Ligands a-g were prepared
according to literature methods.5c A general procedure for the
preparation of these compounds is given below.
2-Chloroethyldiphenylphosphine (0.34 g, 1.4 mmol, 1.0 equiv),
cesium carbonate (0.88 g, 2.8 mmol, 2.0 equiv), and the appropriate
thiol precursor (1.4 mmol, 1.0 equiv) were added to 25 mL of
acetonitrile in a 50 mL Schlenk flask. The solution was degassed with
nitrogen for 10 min and then heated under reflux for 16 h. The
solution was cooled to room temperature, and all subsequent purifi-
cation was performed in air. The solution was filtered through a fine
fritted funnel, and the solvent was removed under vacuum. The
residue was dissolved in CH2Cl2 (50 mL) and washed with water
(3 ꢀ 50 mL). The organic layers were collected, dried over MgSO4,
and the solvent was removed under vacuum to yield a colorless oil.
The compound was further purified on a silica gel column (CH2Cl2)
and recrystallized from pentane to yield a white powder.
(2-(Methylthio)ethyl)diphenylphosphine (a). 1HNMR(CD2Cl2),
δ = 7.39-7.47 (m, Ph2P, 10 H), 2.59 (m, PCH2CH2S, 2 H), 2.39
(m, PCH2CH2S, 2 H), 2.13 (s, CH3, 3H). 13C{1H} NMR (CD2Cl2),
δ = 138.2 (d, JC-P = 13.8 Hz, ArC), 132.6 (d, JC-P = 18.8 Hz,
ArC), 128.6 (s, ArC), 128.5 (d, JC-P = 6.6 Hz, ArC), 30.6 (d,
J
J
C-P = 6.6 Hz, ArC), 115.9 (d, JC-P = 21.9 Hz, ArC), 31.4 (d,
C-P = 22.4 Hz, CH2), 28.1 (d, JC-P = 15.2 Hz, CH2). 31P{1H}
NMR (CD2Cl2) δ = -16.6 (s). 19F{1H} NMR (CD2Cl2) δ =
-116.3 (s, ArF). APPIMS (m/z): Calcd: 341.0924 [MþH]þ.
Found: 341.0912.
Diphenyl(2-(2,3,5,6-tetrafluorophenylthio)ethyl)phosphine (g).
1H NMR (CD2Cl2), δ = 7.34-7.40 (br m, Ph2P, 10 H), 7.08 (m,
ArH, 1 H), 2.98 (m, PCH2CH2S, 2 H), 2.32 (m, PCH2CH2S,
2 H). 13C{1H} NMR (CD2Cl2), δ = 147.7 (m, ArC), 145.3 (m,
ArC), 137.4 (d, JC-P = 13.7 Hz, ArC), 132.6 (d, JC-P = 19.0 Hz,
ArC), 128.9 (s, ArC), 128.6 (d, JC-P = 6.7 Hz, ArC), 114.8 (t,
J
J
C-F = 20.7 Hz, ArC), 106.0 (t, JC-F = 22.9 Hz, ArC), 31.2 (d,
C-P = 21.6 Hz, CH2), 28.9 (d, JC-P = 16.2 Hz, CH2). 31P{1H}
NMR (CD2Cl2) δ = -16.4 (s). 19F{1H} NMR -134.37 (m,
ArF), -139.00 (m, ArF). APPIMS (m/z): Calcd: 395.0641
[MþH]þ. Found: 395.0630.
J
C-P = 21.5 Hz, CH2), 28.2 (d, JC-P = 14.6 Hz, CH2), 15.2 (s,
Preparation of RhI(PS)2Cl Tweezer Complexes. Closed iso-
mers, 7a-f, and semiopen isomers, 8b-g, were all prepared by
the same method; the general procedure for their preparation is
given below.
CH3). 31P{1H} NMR (CD2Cl2) δ = -16.4 (s). APPIMS (m/z):
Calcd: 261.0861 [MþH]þ. Found: 261.0865.
(2-(Mesitylthio)ethyl)diphenylphosphine (b). 1H NMR (CD2Cl2),
δ = 7.31 (m, Ph2P, 10 H), 6.92 (s, ArH, 2 H), 2.65 (m,
PCH2CH2S, 2 H), 2.42 (s, CH3, 6 H), 2.25 (s, CH3, 3 H), 2.23
(m, PCH2CH2S, 2 H). 13C{1H} NMR (CD2Cl2), δ = 142.9 (s,
ArC), 138.3 (s, ArC), 138.0 (d, JC-P = 16.7 Hz, ArC), 132.5 (d,
A solution of the appropriate PS ligand (0.39 mmol, 1.0 equiv)
inTHF(5mL) wasaddedtoasolutionof[Rh(COE)2Cl]2 (0.072g,
0.099 mmol, 0.25 equiv) in THF (5 mL) via pipet over 10 min.
The solution was stirred at room temperature for 2 h. The
orange solution was concentrated under vacuum, and the orange
solid was precipitated by the addition of hexanes (10 mL). Filter-
ing through Celite with subsequent hexane washes (3 ꢀ 5 mL)
resulted in pure semiopen complex. The semiopen isomer could
be quantitatively converted to the closed isomer by removal of
THF under vacuum followed by addition of CH2Cl2.
[(Ph2PCH2CH2SCH3)2Rh][Cl] (7a). 1H NMR (THF-d8), δ =
7.27-7.40 (br m, Ph2P, 12 H), 7.17-7.25 (br m, Ph2P, 8 H), 2.73
(br s, CH3, 6 H), 2.62 (br m, PCH2CH2S, 8 H). 31P{1H} NMR
(THF-d8) δ = 64.3 (d, JRh-P = 160.4 Hz). ESIMS (m/z): Calcd:
623.0627 [M-Cl]þ. Found: 623.0639.
JC-P = 18.7 Hz, ArC), 129.5 (s, ArC), 128.9 (s, ArC), 128.6 (s,
ArC), 128.4 (d, JC-P = 6.5 Hz, ArC), 31.4 (d, JC-P = 20.7 Hz,
CH2), 28.0 (d, JC-P = 15.2 Hz, CH2), 21.7 (s, CH3), 20.7 (s,
CH3). 31P{1H} NMR (CD2Cl2) δ = -16.4 (s). APPIMS (m/z):
Calcd: 365.1487 [MþH]þ. Found: 365.1474.
(2-(2,4-Dimethylphenylthio)ethyl)diphenylphosphine (c). 1H
NMR (CD2Cl2), δ = 7.10-7.43 (br m, Ph2P, 10 H), 7.08 (d,
JH-H = 8.1 Hz, ArH, 1 H), 7.03 (s, ArH, 1 H), 6.95 (br d, ArH,
1 H), 2.89 (m, PCH2CH2S, 2 H), 2.35 (m, PCH2CH2S, 2 H), 2.32
(8) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518–1520.

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 7191
cis-[K2-(Ph2PCH2CH2S(2,4,6-(CH3)3C6H2))-K1-(Ph2PCH2-
following equations: for methanol, T = -23.832Δ2 - 29.46Δ þ
403.0 (where Δ is the shift difference (ppm) between the CH3 and
OH peak) and for ethylene glycol, T = (4.637 - Δ)/0.009967
(where Δ is the shift difference (ppm) between the CH2 and OH
peak). Keq was determined from the integration of the appro-
priate resonances in the 31P NMR spectra corresponding to the
closed and semiopen isomers.
CH2S(2,4,6-(CH3)3C6H2)) ClRh] (8b). 1H NMR (THF-d8), δ =
7.57 (br m, Ph2P, ArH, 4 H), 7.08-7.30 (br m, Ph2P, ArH, 12 H),
6.84-6.93 (br m, Ph2P, ArH, 8 H), 3.03 (br m, PCH2CH2S, 2 H),
2.76 (br m, PCH2CH2S, CH3, 6 H), 2.39 (br m, PCH2CH2S,
CH3, 10 H), 2.23 (br m, PCH2CH2S, CH3, 8 H). 31P{1H} NMR
(THF-d8) δ = 71.4 (dd, JRh-P = 138.4 Hz, JP-P = 31.5 Hz),
34.3 (dd, JRh-P = 128.7 Hz, JP-P = 32.8 Hz). ESIMS (m/z):
Calcd: 831.1879 [M-Cl]þ. Found: 831.1874.
Results and Discussion
cis-[K2-(Ph2PCH2CH2S(2,4-(CH3)2C6H3))-K1-(Ph2PCH2CH2-
HomoligatedRhI(PS)2Clcomplexes (7a-f and 8b-g, eq1)
were prepared to evaluate the influence of the electron with-
drawing/donating ability of the ligand, solvent effects, and
temperature on the equilibrium between the two possible
isomers, closed (7) and semiopen (8), where the Cl- anion is
outer or inner sphere, respectively. PS ligands a-g (eq 1),
which contain substituents on the thioether that vary in their
electron donating/withdrawing ability, were synthesized in
one step from commercially available thiols and (2-chloro-
ethyl)diphenylphosphine by refluxing the reaction mixture
in acetonitrile in the presence of cesium carbonate. The sub-
stituents on the ligands (R) span a range of electron donating
ability, from the strongly electron withdrawing tetrafluoro-
phenyl (g) to the electron donating methyl (a). The RhI-
(PS)2Cl complexes were prepared in THF by adding 4 equiv
of ligand to 1 equiv of [Rh(COE)2Cl]2 (COE = cyclooctene)
followed by precipitation with hexanes. They were character-
ized by 1H, 31P{1H}, 19F{1H} NMR spectroscopy, high reso-
lution mass spectrometry, and in some cases in the solid state
by single crystal X-ray diffraction studies, and all data are
consistent with their proposed structural formulations.
All crystal structures exhibit a distorted tetracoordinate
square-planar geometry with respect to the metal ion
(Figures 1-5). The semiopen isomers exhibit cis-P-Rh-P
coordination environments, where one PS ligand is present
as a five-membered κ2-PS chelate and the second PS ligand is
present in a κ1-PS unchelated form, with the chloride ion
occupying the fourth coordination site of the RhI center
(Figures 1, 2, 4, and 5). The solid state structures of the
semiopen isomers 8c,d,f,g contain two different Rh-P bonds
1
S(2,4-(CH3)2C6H3)) ClRh] (8c). H NMR (THF-d8), δ = 7.20
(br m, Ph2P, ArH, 8 H), 6.95 (br m, Ph2P, ArH, 6 H), 6.82 (br m,
Ph2P, ArH, 8 H), 6.72 (br m, Ph2P, ArH, 2 H), 6.63 (br m, Ph2P,
ArH, 2 H), 2.54 (br m, PCH2CH2S, 2 H), 2.13 (br m,
PCH2CH2S, CH3, 6 H), 1.99 (br m, PCH2CH2S, CH3, 4 H),
1.86 (br m, PCH2CH2S, CH3, 2 H), 1.50 (br m, PCH2CH2S,
CH3, 2 H), 1.23 (br m, PCH2CH2S, CH3, 4 H). 31P{1H} NMR
(THF-d8) δ = 70.9 (dd, JRh-P = 135.9 Hz, JP-P = 30.3 Hz),
31.3 (dd, JRh-P = 127.5 Hz, JP-P = 31.5 Hz). ESIMS (m/z):
Calcd: 803.1566 [M-Cl]þ. Found: 803.1564.
cis-[K2-(Ph2PCH2CH2S(4-(CH3)C6H4))-K1-(Ph2PCH2CH2-
S(4-(CH3)C6H4))ClRh] (8d). 1H NMR (THF-d8), δ = 7.80 (br
m, Ph2P, ArH, 4 H), 7.63 (br m, Ph2P, ArH, 8 H), 7.40 (br m,
Ph2P, ArH, 6 H), 7.27 (br m, Ph2P, ArH, 10 H), 3.08 (br m,
PCH2CH2S, 2 H), 2.56 (br m, PCH2CH2S, 2 H), 2.48 (br, CH3, 3
H), 2.31 (br m, PCH2CH2S, 2 H), 1.91 (br m, PCH2CH2S, 2 H),
1.68 (br, CH3, 3 H). 31P{1H} NMR (THF-d8) δ = 71.5 (dd,
J
Rh-P = 135.2 Hz, JP-P = 29.1 Hz), 30.8 (dd, JRh-P = 125.0
Hz, JP-P = 25.5 Hz). ESIMS (m/z): Calcd: 775.1258 [M-Cl]þ.
Found: 775.1272.
cis-[K2-(Ph2PCH2CH2SPh)-K1-(Ph2PCH2CH2SPh)ClRh] (8e).
1H NMR (THF-d8), δ = 7.77 (br m, Ph2P, ArH, 4 H), 7.48 (br
m, Ph2P, ArH, 8 H), 7.29-7.24 (br m, Ph2P, ArH, 10 H), 7.13 (br
m, Ph2P, ArH, 8 H), 3.01 (br m, PCH2CH2S, 2 H), 2.45 (br m,
PCH2CH2S, 2 H), 2.17 (br m, PCH2CH2S, 2 H), 1.77 (br m,
PCH2CH2S, 2 H). 31P{1H} NMR (THF-d8) δ=73.2(dd, JRh-P
135.9 Hz, JP-P = 27.9 Hz), 32.8 (dd, JRh-P = 125.0 Hz, JP-P
=
=
27.9 Hz). ESIMS (m/z): Calcd: 747.0940 [M-Cl]þ. Found:
747.0937.
cis-[K2-(Ph2PCH2CH2S(4-F1-C6H4))-K1-(Ph2PCH2CH2S(4-
F1-C6H4))ClRh] (8f). 1H NMR (THF-d8), δ = 7.75 (br m, Ph2P,
ArH, 4 H), 7.44 (br m, Ph2P, ArH, 8 H), 7.23 (br m, Ph2P, ArH,
4 H), 7.10 (br m, Ph2P, ArH, 6 H), 7.00 (br m, Ph2P, ArH, 6 H),
2.91 (br m, PCH2CH2S, 4 H), 2.40 (br m, PCH2CH2S, 4 H).
19F{1H} NMR (THF-d8) δ = -118.3 (br m, ArF). 31P{1H}
NMR (THF-d8) δ = 74.6 (dd, JRh-P = 137.1 Hz, JP-P = 30.3
Hz), 35.5 (dd, JRh-P = 127.5 Hz, JP-P = 30.3 Hz). ESIMS
(m/z): Calcd: 783.0757 [M-Cl]þ. Found: 783.0738.
˚
lengths, with the Rh1-P1 bond being around 0.05 A shorter
than the Rh1-P2bond inall structures (Table 1). The shorter
length of the Rh1-P1 bond is most likely a result of the P1
phosphorus atom being a part of the five-membered PS
chelate. The stronger trans effect of the thioether, compared
to chloride, may also contribute to the shortening of the
Rh1-P1 bond. The solid state structures of semiopen iso-
mers 8c,d,f,g only exhibit two significant differences in bond
lengths. The largest difference among the structures is be-
tween the Rh1-S1 bonds of 8f and 8g, with the former being
cis-[K2-(Ph2PCH2CH2S(2,3,5,6-F4-C6H1))-K1-(Ph2PCH2CH2-
1
S(2,3,5,6-F4-C6H1)) ClRh] (8g). H NMR (THF-d8), δ = 7.80
(br m, Ph2P, ArH, 5 H), 7.31-7.46 (br m, Ph2P, ArH, 13 H), 7.16
(br m, Ph2P, ArH, 4 H), 3.65 (br m, PCH2CH2S, 2 H), 2.89 (br m,
PCH2CH2S, 2 H), 2.64 (br m, PCH2CH2S, 4 H). 19F{1H} NMR
(THF-d8) δ = -133.4 (m, ArF), -136.5 (m, ArF), -141.6 (m,
ArF), -141.9 (m, ArF). 31P{1H} NMR (THF-d8) δ = 71.3 (dd,
˚
0.043 A longer than the latter (Table 1). In addition, the
˚
Rh1-P1 bond of 8g is 0.010 A longer than the analogous
J
Rh-P = 133.5 Hz, JP-P = 30.3 Hz), 33.1 (dd, JRh-P = 132.3
Hz, JP-P = 31.5 Hz). ESIMS (m/z): Calcd: 891.0186 [M-Cl]þ.
Found: 891.0200.
˚
bond in 8d. All other bonds are within 0.009 A of each other
(Table 1). In the closed isomer 7e (Figure 3), both PS ligands
are chelated to the RhI in a κ2-PS fashion with a cis-P-Rh-P
coordination environment, and the chloride ion is present as
General Procedure for VT NMR Experiments. All samples
were prepared in an inert atmosphere glovebox and sealed under
nitrogen in NMR tubes with J. Young valves. The NMR
samples were prepared by adding 0.5 mL of a stock solution
containing between 0 and 100% THF-d8 in CD2Cl2 to 15-20
mg of the RhI(PS)2Cl complex. The samples were allowed to
equilibrate for 10 min in the NMR after the temperature reading
from the thermocouple stabilized to within 0.1 K. The tempera-
ture of the probe was determined by using either neat methanol
(for temperatures between 170 and 300 K) or neat ethylene
glycol (for temperatures between 300 and 312 K) using the
˚
an outer-sphere counterion with a Rh-Cl distance of 7.57 A.
The smaller P2-Rh1-S1 angle in the closed isomer, 161°, as
compared to the semiopen isomers, around 174°, is more
distorted from an ideal square-planar geometry, which is
possibly a result of the increased steric congestion around the
RhI center when both PS ligands are chelated to the metal as
is 7e (further crystallographic data are presented in the Sup-
porting Information). Taken together, these data show

7192 Inorganic Chemistry, Vol. 49, No. 15, 2010
Wiester et al.
Figure 5. X-ray crystal structure of semiopen isomer 8g, with thermal
ellipsoids drawn at 50% probability. The crystal of 8g was grown by slow
diffusion ofdiethyl ether intoa THF solutionof 8g. Hydrogenatomshave
been omitted for clarity. Pink = Rh, Yellow = S, Orange = P, Green =
Cl, Light Blue = F.
Figure 1. X-ray crystal structure of semiopen isomer 8c, with thermal
ellipsoids drawn at 50% probability. The crystal of 8c was grown by slow
diffusion of diethyl ether into a THF solution of 8c. Hydrogenatoms have
been omitted for clarity. Pink = Rh, Yellow = S, Orange = P, Green = Cl.
˚
Table 1. Selected Bond Lengths (A) and Angles (deg)
8c
8d
7e
8f
8g
Rh1-P1
Rh1-P2
Rh1-S1
Rh1-S2
Rh1-Cl1
2.1949(6) 2.1895(3)
2.2463(7) 2.2486(3)
2.3537(7) 2.3551(3)
2.2325(9) 2.1903(6) 2.1996(4)
2.2325(9) 2.2477(6) 2.2539(4)
2.3100(9) 2.3592(6) 2.3160(4)
na
na
2.3100(9) na
na
2.4006(6) 2.3918(3)
na
2.3976(6) 2.3924(3)
P1-Rh1-P2
P1-Rh1-S1
P1-Rh1-S2
P1-Rh1-Cl1 172.81(2) 174.220(11) na
P2-Rh1-S1
P2-Rh1-S2
98.88(2)
86.83(2)
na
97.657(12)
87.397(12)
na
98.18(5)
85.59(3)
97.33(2)
86.95(2)
99.689(13)
86.993(13)
161.72(3) na
na
Figure 2. X-ray crystal structure of semiopen isomer 8d, with thermal
ellipsoids drawn at 50% probability. The crystal of 8d was grown by slow
diffusionof diethyl ether into a THF solution of8d. Hydrogen atoms have
been omitted for clarity. Pink = Rh, Yellow = S, Orange = P, Green = Cl.
174.82(2) 170.728(13)
174.27(2) 174.062(11) 161.72(3) 175.24(2) 173.006(13)
na
na
85.59(3)
na
na
na
P2-Rh1-Cl1 87.93(2)
S1-Rh1-S2 na
87.775(11)
na
87.72(2)
na
89.315(13)
na
96.45(4)
the semiopen isomer (8b) is observed, indicating that the
coordination of the Cl- is strongly influenced by solvent. To
study this effect further, the equilibria distribution between
semiopen and closed isomers were measured in a series of
different solvents. Complexes 7b and 8b were used in these
studies because the two mesityl ligands provide enhanced
solubility and therefore access to more solvents, and the
equilibrium distribution between the two isomers was ob-
served to be strongly influenced by solvent polarity, as deter-
mined by the ENT solvent polarity scale.9 This scale is based
upon a solvatochromatic shift in the absorption spectra of
2,6-dipenyl-4-(2,4,6-triphenyl-1-pyridinio)phenolate
Figure 3. X-ray crystal structure of closed isomer 7e, with thermal
ellipsoids drawn at 50% probability. The crystal of 7e was grown by slow
diffusion of pentane into a CH2Cl2 solution of 7e. Hydrogen atoms and
non-coordinating counterions have been omitted for clarity. Pink = Rh,
Yellow = S, Orange = P.
(diphenyl betaine), a dye molecule that is zwitterionic in the
ground state and less polar in the excited state because of charge
redistribution, and the scale ranges from 0.0 for tetramethyl-
silane to 1.0 for water.9 Solvents were chosen such that they
spanned a range of polarities from 0.009 to 0.741.9 The 31
P
NMR spectra of complexes 7b and 8b were obtained in 15
solvents at room temperature, and the percentage of semi-
open 8b was plotted as a function of solvent polarity (Figure 6).
The two different coordination environments of the two
isomers can be distinguished by their resonances in the 31P
NMR spectra, with the closed isomer 7b exhibiting a doublet
around δ 61-66 and the semiopen isomer 8b exhibiting two
doublets of doublets around δ 71-75 and δ 31-35, corre-
sponding to the phosphines of the chelated and non-chelated
ligands, respectively (Figure 8).10 Although, the solvents
chosen vary in physical properties from coordinating to
Figure 4. X-ray crystal structure of semiopen isomer 8f, with thermal
ellipsoids drawn at 50% probability. The crystal of 8f was grown by slow
diffusion of diethyl ether into a THF solution of 8f. Hydrogen atoms have
been omitted for clarity. Pink = Rh, Yellow = S, Orange = P, Green =
Cl, Light Blue = F.
that the structures are remarkably similar in the solid state,
lending themselves to reasonable comparisons (vide infra).
Room temperature 31P NMR spectra of the RhI(PS)2Cl
complex made from ligand b, show that theclosedisomer(7b)
is the only isomer present in CD2Cl2, while in THF-d8 only
(9) (a) Dimroth, K.; Reichardt, C. Fresenius’ Z. Anal. Chem. 1966, 215,
344–350. (b) Reichardt, C. Chem. Rev. 1994, 94, 2319–2358.
(10) (a) Sanger, A. R.; Lobe, C. G.; Weiner-Fedorak, J. E. Inorg. Chim.
Acta 1981, 53, L123–L124. (b) Sanger, A. R. Can. J. Chem. 1983, 61, 2214–
2219.

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 7193
Figure 6. Effect of solvent polarity on the ratio of semiopen 8b to closed
7b. The dashed line is at 0.253.
Figure 8. 31P NMR spectra of 7b and 8b in different solvent mixtures of
CD2Cl2 and THF-d8 at 298 K. Spectra are shifted by 2.5 ppm.
Figure 7. Effect of hydrogen-bond accepting basicity (β) of the solvent
on the preference for either isomer. No reported values for 1-nonanol,
diglyme, or triglyme could be found.
non-coordinating, hydrogen bond donating/accepting, protic
and aprotic, solvent polarity seems to be the most significant
factor affecting the preference for one isomer over the other.
In all solvents less polar than diethylene glycol dimethyl ether
(diglyme), which has an ENT value of 0.244, only the semiopen
isomer, 8b, was observed in the 31P NMR spectra (Figure 6).
In contrast, only the closed isomer, 7b, was observed in all
solvents more polar than 1,1,2,2-tetrachloroethane, which
has an ENT value of 0.269. Only in the case of the two solvents
with polarities between these two values, methyl acetate and
triethylene glycol dimethyl ether (triglyme) (which both have
ENT values of 0.253), were both isomers observed in solution at
room temperature. The difference in the ratios of 7b to 8b in
methyl acetate and triglyme may result from the RhI(PS)2Cl
complex being more sensitive to small changes in polarity in
this region than the dye molecule used to calculate ENT values,
diphenyl betaine, which gives the same value for methyl
acetate and triglyme. Interestingly, the switch in preference
for one isomer takes place over a remarkably narrow polarity
range (0.244-0.269), and does not seem to show a significant
dependence on any other physical properties of the solvents.
The same trend is seen with other solvent polarity scales (Z
and π*), although fewer values for different solvents have
been reported for such scales (see Supporting Information).11
Figure 9. Polarity of different solvent ratios of THF and CH2Cl2.
It should also be noted that no correlation was observed
between isomer preference and hydrogen-bond accepting
basicity (β) (Figure 7), hydrogen-bond donating acidity (R),
or dielectric constant of the solvents, although not all hydro-
gen-bond donating acidities R values for the solvents used in
this study are known (see Supporting Information).11 Inter-
estingly, the two solvents of choice for working with WLA
complexes historically have been CH2Cl2 and THF,1,3,5
which happen to fall on opposite sides of the value of solvent
polarity defining preference for inner or outer sphere Cl-
coordination (Figure 6).
To further understand the origin of the preference for the
closedisomer inmore polar solvents and thesemiopen isomer
in less polar solvents, the switching between the closed and
semiopen isomers, 7b and 8b, was studied by 31P NMR
spectroscopy in different solvent mixtures of CD2Cl2 and
THF-d8 (Figure 8). The absorption spectra of diphenyl
betaine in different solvent mixtures of CH2Cl2 and THF
were taken and used to calculate ENT values for the solvent
mixtures (Figure 9 and Supporting Information).12 The
(11) (a) Kamlet, M. J.; Abboud, J.-L. M.; Abraham, M. H.; Taft, R. W.
J. Org. Chem. 1983, 48, 2877–2887. (b) Marcus, Y. Chem. Soc. Rev. 1993, 22,
409–416. (c) Zou, H.; Zhang, Y.; Lu, P. Anal. Chim. Acta 1995, 310, 461–471.
(d) El-Sayed, M.; Spange, S. J. Phys. Chem. B 2007, 111, 7224–7233.
(12) (a) Balakrishnan, S.; Easteal, A. J. Aust. J. Chem. 1981, 34, 943–947.
(b) Balakrishnan, S.; Easteal, A. J. Aust. J. Chem. 1981, 34, 933–941.
(c) Langhals, H. Angew. Chem., Int. Ed. 1982, 21, 724–733.

7194 Inorganic Chemistry, Vol. 49, No. 15, 2010
Wiester et al.
solvent polarity values measured were 0.213 for THF and
0.308 for CH2Cl2, which match well with literature values
(0.207 and 0.309, respectively).9 The polarity of the solution
does not change as the amount of THF relative to CH2Cl2
increases from 0 to 20%, and only the closed isomer 7b is
observed in the 31P NMR spectra. In this range, the value of
ENT remains constant at 0.308 (Figure 9). There is a slight
decrease in the polarity of the solution as the amount of THF
increases from 20 to 60%, ETN values decrease from 0.308 to
0.301, and a small amount of 8b begins to appear in the 31P
NMR spectra at 40% THF-d8 (Figure 8). A dramatic
decrease in the polarity of the solvent mixture occurs as the
amount of THF increases from 60 to 100%, as evidenced by
decreasing ENT values from 0.301 to 0.213, and 8b begins to
dominate the equilibrium. In solvent mixtures between 50
and 80% THF-d8, the range at which the ENT value decreases
significantly, both isomers can be observed in solution. For
solvent mixtures above 80% THF-d8, only the semiopen
isomer is observed by 31P NMR spectroscopy at room
temperature, because the ETN value defining the preference
for outer sphere Cl- coordination has been exceeded. The
range of solvent polarities where both isomers (7b and 8b) are
observed in mixtures of CD2Cl2 and THF-d8 (0.275-0.302)
is slightly shifted from the solvent polarity range for the pure
solvents (0.244-0.269) where both isomers are observed
(Figure 6). The size of this window is very similar for both
cases, 0.027 for CD2Cl2 and THF-d8 solvent mixtures and
0.025 for the pure solvents, indicating that the preference for
one isomer switches over a very narrow range of solvent
polarity. The 0.03 shift in the rangeof solvent polarities where
both isomers are observed in solution between the pure
solvents as compared to the binary CH2Cl2: THF solvent
mixtures may result from the preferential solvation of the RhI
complex or the diphenyl betaine in the solvent mixtures.13
These results have clear implications for functional systems
involving the WLA. If one wants to take advantage of
through space interactions between the two ligands in a
closed tweezer-like complex, one should use solvents more
polar than dibromomethane, while using a solvent with a
slightly lower polarity, such as diglyme, will result in a
semiopen complex, with little communication between the
ligands.
Figure 10. 31P NMR spectra of 7b and 8b at various temperatures in
THF-d8. Spectra are shifted by 2.5 ppm.
tion of the closed isomer. Van’t Hoff plots were used to
estimate ΔH° and TΔS°.14
Indeed, interconversion between the two isomers can be
observed by measuring the intensities of the peaks corre-
sponding to 7b and 8b in THF-d8 in the their 31P NMR
spectra as a function of temperature. At 299 K, only the
resonances corresponding to the semiopen isomer, 8b, are
observed (Figure 10), however, upon cooling the solution, the
resonance corresponding to the closed isomer, 7b, is first
observed at 285 K, and then increases in relative intensity
compared to resonances for the semiopen isomer as the
temperature is progressively decreased. In THF-d8, Keq
increases as the temperature increases, indicating that the
movement of Cl- to the inner coordination sphere is endo-
thermic. It should be noted that the resonances for both
isomers do not shift or broaden significantly over the tem-
perature range where switching is observed, indicating that
the two complexes are exchanging slowly on the NMR time
scale. ΔH° and TΔS° were determined from a van’t Hoff plot
of the different Keq values obtained as a function of temperature,
neglecting the changes in solvent heat capacity (Figure 11),14
and were found to be 2.7 ( 0.2 and 8.8 ( 0.3 kcal/mol,
respectively, at 298 K, signifying that the formation of the
semiopen isomer from the closed isomer is entropically
favored, but enthalpically disfavored. Because only the
semiopen complex is observed by NMR spectroscopy at
298 K in THF-d8, the value of Keq had to be calculated from
ΔH° and TΔS° and was found to be 2.9 ꢀ 104 ( 9 ꢀ 103 M-1
at 298 K. Using these values, ΔG° was found to be -6.1 ( 0.3
kcal/mol in favor of the semiopen isomer in THF-d8.
The electron donating/withdrawing ability of the PS
ligands also likely influences the equilibrium distribution
between the two isomers. To quantitatively compare the
effects of the electronic contribution of the ligand on the
equilibrium between the two possible isomers, the thermo-
dynamic parameters (Keq, ΔG°, ΔH°, and TΔS°) for the
interconversion between the closed and semiopen isomers
were determined by VT 31P NMR spectroscopy in mixtures
of CD2Cl2 and THF-d8. The 31P NMR spectra were taken
over a temperature range of 170 to 312 K, and the relative
integration of the resonances corresponding to the complexes
with different coordination geometries was used to calculate
Keq at each temperature (Figure 10). When calculating Keq,
the concentration of chloride counterions in the outer co-
ordination sphere was assumed to be equal to the concentra-
The equilibria between the closed (7b) and semiopen (8b)
isomers were also investigated by VT 31P NMR spectroscopy
as a function of CD2Cl2 and THF-d8 solvent ratio to under-
stand how ΔG°, ΔH°, TΔS°, and Keq change with solvent
polarity (Figure 8). These experiments were run in triplicate
to determine the uncertainty associated with each value. As
discussed above, in solvent mixtures containing less than
40% THF-d8, only the closed isomer (7b) could be observed
(13) (a) Reichardt, C. Solvents and Solvent Effects in Organic Chemistry,
3rd ed.; Wiley-VCH: Weinheim, Germany, 2003. (b) Marcus, Y. Solvent
Mixtures: Properties and Selective Solvation; Marcel Dekker: New York, 2002.
(14) (a) Clarke, E. C. W.; Glew, D. N. Trans. Faraday Soc. 1966, 62, 539–
547. (b) Stauffer, D. A.; Barrans, R. E.; Dougherty, D. A. J. Org. Chem. 1990, 55,
2762–2767.

Article
Inorganic Chemistry, Vol. 49, No. 15, 2010 7195
Table 2. Thermodynamic Parameters for the Switching between the Closed and Semi-Open Isomers of Complexes 7b and 8b, Respectively, at 298 Ka,b
complex
solvent ratio (CD2Cl2:THF-d8)
Keq (M-1
)
ΔG° (kcal/mol)
ΔH° (kcal/mol)
TΔS° (kcal/mol)
7b,8b
60:40
50:50
40:60
30:70
20:80
10:90
0:100
3.5 ( 0.8
-0.7 ( 2.6
-1.3 ( 1.6
-2.0 ( 2.3
-3.0 ( 0.2
-4.7 ( 0.3
-5.7 ( 0.4
-6.1 ( 0.3
13.1 ( 1.9
11.4 ( 1.1
13.3 ( 1.7
9.4 ( 0.2
8.1 ( 0.2
7.2 ( 0.2
2.7 ( 0.2
13.8 ( 1.8
12.7 ( 1.1
15.3 ( 1.6
12.4 ( 0.2
12.8 ( 0.3
13.0 ( 0.3
8.8 ( 0.3
9 ( 3
31 ( 17
150 ( 18
3 ꢀ 103 ( 1 ꢀ 103
1.6 ꢀ 104 ( 6 ꢀ 103
2.9 ꢀ 104 ( 9 ꢀ 103
a Enthalpy and entropy values were obtained by neglecting the solvent heat capacity -ΔCp°.14 b Thermodynamic parameters were extrapolated from
van’t Hoff plots (Figure 11). Uncertainties were determined by a least-squares analysis from three different samples.
sponds tothe point whenpolarity begins to change. Similarly,
no change in the equilibrium position is observed until the
solvent ratio of 60:40, CD2Cl2/THF-d8 is surpassed, strongly
suggesting this change is the result of changes in polarity.
From 70 to 100% THF-d8, there is a large decrease in solvent
polarity and also a decrease observed in the values of ΔH°
and TΔS°. Interestingly, formation of the semiopen isomer
from the closed isomer is entropically favored in all solvent
mixtures of THF-d8 and CD2Cl2. The favorable entropic
changes could arise because the charged closed complex and
counterion have an ordered solvation sphere, while the
resulting neutral semiopen complex produces considerable
less ordering of the solvent.15 The positive enthalpy in all
solvent mixtures is most likely a result of the difference in
energies between the Rh-S bond, which is broken, and
Rh-Cl bond, which is formed, with the former being stronger.
It is also likely that there is some contribution to the positive
enthalpy from the energy required for solvent reorganization.
The switching between inner and outer sphere Cl- coordi-
nation of RhI(PS)2Cl complexes 7c-f and 8c-f in different
solvent mixtures of THF-d8 and CD2Cl2 was also studied by
VT 31P NMR spectroscopy (Table 3). No error values are
reported because of the excessive experimental time to run
these in triplicate, although the uncertainties can be approxi-
mated from those calculated for complexes 7b,8b. It should
be noted that no switching was observed for the complexes
formed with the most electron donating (a) and most electron
withdrawing (g) ligands. For all solvent ratios and tempera-
tures investigated, only the closed isomer 7a and only the
semiopen isomer 8g were observed by 31P NMR spectros-
copy, indicating that strongly electron donating ligands favor
the closed isomer, and strongly electron withdrawing ligands
favor the semiopen isomer. For some of the complexes 7c-f
and 8c-f, switching was only observed at temperatures
between 170-180 K in 100% THF-d8, close to the freezing
point of THF, limiting the range over which switching could
be studied as well as the magnitude of the observed changes in
the 31P NMR spectra, and thus increasing the experimental
error beyond the acceptable limit. To compare the thermo-
dynamic parameters of these complexes, we chose to inves-
tigate switching in 80 and 90% THF-d8 because in this region
there is a significant change in polarity and as a consequence,
the equilibrium position. For complexes 7c-f and 8c-f, ΔH°
decreases as the amount of THF-d8 increases, similar to
complexes 7b and 8b, signifying that there is a lower enthalpy
of reaction in less polar solvent mixtures, consistent with the
hypothesis that less polar solvent mixtures cannot shield the
charges between the RhI complex and the Cl- counterion.
Figure 11. van’t Hoff plots of 7b and 8b in different solvent mixture of
CD2Cl2 and THF-d8.
in the 31P NMR spectra at all temperatures investigated. In
solutions of 40-100% THF-d8 in CD2Cl2 both isomers,
closed (7b) and semiopen (8b), are present in solution, and
the equilibrium thermodynamic parameters could be deter-
mined for each solvent mixture. As the amount of THF-d8
increases relative to CD2Cl2 from 40 to 60%, the values of
ΔG°, ΔH°, and TΔS° remain almost constant, within the
uncertainty of the experiment (Table 2). This is consistent
with the observation that the ENT value does not vary
significantly over this range, and therefore no change in the
equilibrium position is expected. In contrast, as the amount
of THF-d8 increases above 60%, ΔH° decreases with increas-
ing THF-d8 concentration, from 13.3 ( 1.7 kcal/mol in 60%
THF-d8 to 2.7 ( 0.2 kcal/mol in 100% THF-d8, indicating
that the enthalpic cost in going from the closed to semiopen
isomer decreases as the solvent polarity decreases. This is
most likely the result of the neutral semiopen isomer being
favored because the less polar solvent mixture is less able to
solvate the ions. The values of TΔS° remain consistent within
the uncertainty of the experiment in all mixtures of THF-d8
and CD2Cl2, with the values fluctuating around 13 kcal/mol.
The value of TΔS° in 100% THF-d8 is significantly less than
the corresponding values calculated for the solvent mixtures,
8.8 ( 0.3 kcal/mol, possibly a result of the less polar THF-d8
forming a less ordered solvent shell around the ions or fewer
solvent molecules being able to solvate the ions, because of
the slightly larger size of THF-d8 as compared to CD2Cl2,
which produces a lower entropic contribution when liberated
(Table 2). The changes in Keq, ΔG°, ΔH°, and TΔS° can be
explained by considering the plot of solvent ratio to polarity:
no significant changes in the ENT values are seen between
ratios of 100:0 and 60:40, CD2Cl2/THF-d8, which corre-
(15) Basolo, F.; Johnson, R. C. Coordination Chemistry: The Chemistry of
Metal Complexes; W. A. Benjamin, Inc.: New York, 1964.

7196 Inorganic Chemistry, Vol. 49, No. 15, 2010
Wiester et al.
Table 3. Thermodynamic Parameters for the Switching between the Closed and Semi-Open Isomers of Complexes 7b-f and 8b-f, Respectively, at 298 Ka,b
complexes
solvent ratio (CD2Cl2:THF-d8)
Keq (M-1
3 ꢀ 103
)
ΔG° (kcal/mol)
ΔH° (kcal/mol)
TΔS° (kcal/mol)
7b,8b
20:80
10:90
20:80
10:90
20:80
10:90
20:80
10:90
20:80
10:90
-4.7
-5.7
-4.1
-5.6
-4.5
-4.7
-8.8
-6.4
-8.8
-5.8
8.1
7.2
8.0
6.7
7.5
6.2
7.2
5.9
6.4
5.9
13
13
12
12
12
11
16
12
15
12
1.6 ꢀ 104
1 ꢀ 103
7c,8c
7d,8d
7e,8e
7f,8f
1.2 ꢀ 104
2 ꢀ 103
3 ꢀ 103
2.7 ꢀ 106
4.7 ꢀ 104
2.8 ꢀ 106
1.7 ꢀ 104
a Enthalpy and entropy values were obtained by neglecting the solvent heat capacity -ΔCp°.14 b Thermodynamic parameters were extrapolated from
van’t Hoff plots (see Supporting Information).
Although for reactions involving 7c,d and 8c,d, respectively,
the values of TΔS° do not vary significantly as the amount
of THF-d8 increases, for reactions involving complexes 7e,f
and 8e,f, respectively, the values of TΔS° are 3-4 kcal/mol
larger in 80% THF-d8 than in 90% THF-d8. These data
show that the reaction is always entropically favored, but
there is no obvious reason for the abrupt change in magni-
tude of TΔS° for 7e,f and 8e,f. ΔH° also can be compared
between complexes to correlate the electron donating/with-
drawing abilities of the ligands to the preference of one
isomer over the other and the strength of the Rh-S bond.
As mentioned above, for the complexes with the most
electron donating and withdrawing groups on the ligands,
only the closed isomer 7a and semiopen isomer 8g were
observed for all solvent ratios and temperatures investi-
gated. In both 80 and 90% THF-d8 similar trends are
observed for ΔH° as the electron donating ability of the
substituents on the sulfur are changed. ΔH° can be com-
pared among the different complexes at each solvent ratio to
quantify the differences in the enthalpic cost of breaking the
Rh-S bond as a result of changing the electron donating
ability of the ligand. As the energy required to form the Rh-Cl
bond is nearly identical in all cases and the energy associated
with solvent reorganization should be similar in all cases, the
change in this parameter should only reflect differences in
strength of the Rh-S bond. There is a decrease in ΔH° as the
electron withdrawing ability of the ligands decreases, ΔH°b >
ΔH°c > ΔH°d > ΔH°e > ΔH°f, signifying that it requires less
energy to break the Rh-S bond when more electron with-
drawing ligands are present (Table 3).
by VT 31P NMR spectroscopy. This switching is reversible
and depends upon the solvent polarity, the solvent ratio, the
temperature, and the group attached to the sulfur of the PS
ligand. These results suggest that when designing functional
systems using the WLA, it is necessary to understand
the effects of solvent, temperature, and the balance between
counterion coordination and the electron donating ability
of the ligand to predict the coordination geometry under
experimental conditions. As a consequence of these stud-
ies one can anticipate the solvent polarity, ligand, or
temperature needed to achieve a desired coordination
geometry. If a closed geometry is desired, a more polar
solvent and electron donating ligands should be used,
whereas less polar solvents, such as THF, and electron
withdrawing ligands favor the semiopen geometry. This
study focused on RhI homoligated tweezer-like com-
plexes; however, the results are likely extendable to the
more sophisticated and complicated WLA catalysts and
allosteric enzyme mimics.
Acknowledgment. We acknowledge the NSF, AFOSR,
ARO, and DDRE (through the MURI program) for
financial support of this research and IMSERC (North-
western U.) for analytical services. A.B.B. is grateful for a
NIH Postdoctoral fellowship (5F32CA136148-02).
Note Added after ASAP Publication. This paper was
published on the Web on July 9, 2010. An additional co-
author, Hyojong Yoo, was added to the paper and the
corrected version was reposted on July 26, 2010.
Supporting Information Available: Tables of crystallographic
data, UV/vis spectra of diphenyl betaine, 31P VT NMR spectra,
and van’t Hoff plots. This material is available free of charge via
the Internet at http://pubs.acs.org.
Conclusion
In conclusion, the switching between closed and semiopen
structural isomers of RhI(PS)2Cl complexes has been studied