Probing the mechanistic and energetic basis for the Weak-Link approach to supramolecular coordination complexes

Article abstract of DOI:10.1021/om020739c

Through the Weak-Link Synthetic Approach, the unsymmetric hemilabile ligand, 1,4-bis-(2-diphenylphosphinoethoxy)naphthalene (7), has been used to prepare binuclear Rh(I) macrocycles in a two-step fashion. The intermediate structure, [(μ²,η¹:η⁴:η¹-(1,4-bi s(2-(diphenylphosphino)ethoxy)naphthalene))₂Rh₂] [BF₄]₂ (8), and several macrocyclic complexes, [(μ²-1,4-bis(2-(diphenylphosphino)ethoxy)naphthalene)₂R h₂(CH₃CN)₄] [BF₄]₂ (9), [(μ²-1,4-bis-(2-(diphenylphosphino)ethoxy)naphthalene)₂ Rh₂(CO)₆][BF₄]₂ (10), and [(μ²-1,4-bis(2-(diphenylphosphino)ethoxy)naphthalene)₂( CO)₂Rh₂(CH₃CN)₂][BF₄] ₂ (11), are reported herein. Additionally, the mononuclear Rh(I) complexes, [(η¹:η⁴:?η¹-1,4-bis(2-(diphenylphos phino)ethoxy)-naphthalene)Rh] [BF₄] (12) and [(η¹:η¹-1,4-bis(2-(diphenylphosphino)ethoxy)naphth alene)-(CO)Rh(CH₃CN)] [BF₄] (13), were isolated and fully characterized. Single-crystal X-ray diffraction structures of 8, 9 (cis-, anti-), 11, and 12 were determined. 2-D solution NMR studies and the solid-state crystal structures of 8 and 12 suggest that the arenes bind to Rh(I) in predominantly an η⁴-fashion. The conversion of dimer 8 into monomer 12 was monitored under a variety of conditions. The kinetic investigation of this process as a function of temperature with a series of phosphine-based ligands present has afforded the thermo-chemical parameters of this conversion, which are consistent with an associative mechanism. Significantly, this work represents the first study aimed at gaining mechanistic insight into the conversion of the binuclear intermediates into the mononuclear products formed in the Weak-Link Approach. The high activation barrier and associative nature of this step is what allows one to efficiently synthesize targeted macrocyclic and higher-ordered structures via the approach.

Full text of DOI:10.1021/om020739c

Organometallics 2002, 21, 5713-5725
5713
P r obin g th e Mech a n istic a n d En er getic Ba sis for th e
Wea k -Lin k Ap p r oa ch to Su p r a m olecu la r Coor d in a tion
Com p lexes
Bradley J . Holliday, You-Moon J eon, Chad A. Mirkin,* and Charlotte L. Stern
Department of Chemistry and Center for Nanofabrication and Molecular Self-Assembly,
2145 Sheridan Road, Northwestern University, Evanston, Illinois 60208-3113
Christopher D. Incarvito, Lev N. Zakharov, Roger D. Sommer, and
Arnold L. Rheingold
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
Received September 6, 2002
Through the Weak-Link Synthetic Approach, the unsymmetric hemilabile ligand, 1,4-bis-
(2-diphenylphosphinoethoxy)naphthalene (7), has been used to prepare binuclear Rh(I)
macrocycles in a two-step fashion. The intermediate structure, [(µ2,η1:η4:η1-(1,4-bis(2-
(diphenylphosphino)ethoxy)naphthalene))2Rh2][BF4]2 (8), and several macrocyclic complexes,
[(µ2-1,4-bis(2-(diphenylphosphino)ethoxy)naphthalene)2Rh2(CH3CN)4][BF4]2 (9), [(µ2-1,4-bis-
(2-(diphenylphosphino)ethoxy)naphthalene)2Rh2(CO)6][BF4]2 (10), and [(µ2-1,4-bis(2-(diphen-
ylphosphino)ethoxy)naphthalene)2(CO)2Rh2(CH3CN)2][BF4]2 (11), are reported herein. Ad-
ditionally, the mononuclear Rh(I) complexes, [(η1:η4:η1-1,4-bis(2-(diphenylphosphino)ethoxy)-
naphthalene)Rh][BF4] (12) and [(η1:η1-1,4-bis(2-(diphenylphosphino)ethoxy)naphthalene)-
(CO)Rh(CH3CN)][BF4] (13), were isolated and fully characterized. Single-crystal X-ray
diffraction structures of 8, 9 (cis-, anti-), 11, and 12 were determined. 2-D solution NMR
studies and the solid-state crystal structures of 8 and 12 suggest that the arenes bind to
Rh(I) in predominantly an η4-fashion. The conversion of dimer 8 into monomer 12 was
monitored under a variety of conditions. The kinetic investigation of this process as a function
of temperature with a series of phosphine-based ligands present has afforded the thermo-
chemical parameters of this conversion, which are consistent with an associative mechanism.
Significantly, this work represents the first study aimed at gaining mechanistic insight into
the conversion of the binuclear intermediates into the mononuclear products formed in the
Weak-Link Approach. The high activation barrier and associative nature of this step is what
allows one to efficiently synthesize targeted macrocyclic and higher-ordered structures via
the approach.
In tr od u ction
coordination chemistry-based synthetic approaches have
been developed for preparing such structures: the
Directional-Bonding, Symmetry-Interaction, and Weak-
Link Approaches.2 With these approaches one can
construct a wide variety of architectures that vary in
size, chemical complexity, and structural flexibility.
Coordination chemistry is used in these approaches
because of the large number of available and tailorable
metal-ligand coordination modes and accessible metal-
ligand bond strengths when compared with the covalent
chemistry of carbon.
From a mechanistic standpoint, these three strategies
can be broken into two general categories: those under
thermodynamic control and those under kinetic control.
All of them rely on the concept of thermally annealing
the reactants to avoid potential oligomeric and poly-
meric byproducts, which are often the initial complexes
formed. However, in the case of the Directional-Bonding
and Symmetry-Interaction Approaches, the overall ther-
The design and synthesis of inorganic and organo-
metallic macrocycles with tailorable host-guest chem-
istry and catalytic properties have been the subject of
intense research over the past decade.1 Three general
* To whom correspondence should be addressed. Fax: (847) 467-
5123. E-mail: camirkin@chem.nwu.edu.
(1) (a) Lehn, J .-M. Supramolecular Chemistry; VCH: New York,
1995. (b) Vo¨gtle, F. Supramolecular Chemistry; Wiley: Chichester,
1991. (c) Caulder, D. L.; Raymond, K. N. J . Chem. Soc., Dalton Trans.
1999, 1185-1200. (d) MacGillivray, L. R.; Atwood, J . L. Angew. Chem.,
Int. Ed. 1999, 38, 1018-1033. (e) Lawrence, D. S.; J iang, T.; Levett,
M. Chem. Rev. 1995, 95, 2229-2260. (f) Hunter, C. A. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1079-1081. (g) Kuehl, C. J .; Kryschenko, Y.
K.; Radhakrishnan, U.; Seidel, S. R.; Huang, S. D.; Stang, P. J . Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 4932-4936. (h) Leininger, S.;
Olenyuk, B.; Stang, P. J . Chem. Rev. 2000, 100, 853-908. (i) Leininger,
S.; Fan, J .; Schmitz, M.; Stang, P. J . Proc. Natl. Acad. Sci. U.S.A. 2000,
97, 1380-1384. (j) Cotton, F. A.; Lin, C.; Murillo, C. A. Acc. Chem.
Res. 2001, 34, 759-771. (k) Caulder, D. L.; Raymond, K. N. Acc. Chem.
Res. 1999, 32, 975-982. (l) Fujita, M. Struct. Bonding 2000, 96, 177-
201. (m) Fujita, M.; Umemoto, K.; Yoshizawa, M.; Fujita, N.; Kusuka-
wa, T.; Biradha, K. Chem. Commun. 2001, 509-518. (n) Balzani, V.;
Credi, A.; Raymo, F. M.; Stoddart, J . F. Angew. Chem., Int. Ed. 2000,
39, 3348-3391.
(2) Holliday, B. J .; Mirkin, C. A. Angew. Chem., Int. Ed. 2001, 40,
2022-2043.
10.1021/om020739c CCC: $22.00 © 2002 American Chemical Society
Publication on Web 11/20/2002

5714 Organometallics, Vol. 21, No. 26, 2002
Holliday et al.
polynuclear three-dimensional cage architectures.4-10
This approach relies on hemilabile ligands,11 1a ,b, to
organize metal centers in the form of rigid condensed
intermediates (2a and 3b) with intentionally designed
strong and weak metal links (Scheme 1). These struc-
tures can be subsequently opened into flexible macro-
cycles, 4a,b, through high-yielding, small-molecule ligand
substitution reactions which break the weak links but
not the strong ones. For the systems studied thus far,
two distinct structural isomers of the condensed inter-
mediate, descriptively termed the “bow-tie” and “slipped”
intermediates (2a and 3b, respectively), have been
characterized; they differ in the ligand-binding mode to
the metal centers. Moreover, both isomers, under the
appropriate conditions, often will convert to the ther-
modynamic mononuclear two-legged piano-stool com-
plexes, 5a ,b (Scheme 1).7
Sch em e 1
In this paper we report the design and synthesis of a
new naphthalene-based hemilabile ligand that allows
us to systematically examine some of the important
factors that control the formation of both condensed
intermediates and the thermodynamic mononuclear
products (Scheme 1). The motivation for synthesizing
the naphthalene-based ligand 7 (Scheme 2) comes from
the observed reaction pathways of the related ligand
systems based on phenyl, biphenyl, anthracenyl, and
duranyl moieties. In the case of Rh(I), when phenyl- or
biphenyl-based phosphinoalkyl ether ligands were used,
the slipped intermediates (binuclear Rh(I) structures
that have two-legged piano-stool geometries with η6-
arenes and bis(phosphine)-Rh(I) bonds) are the pre-
dominant products.7 However, when the duranyl- or
anthracyl-based ligands were used, the bow-tie inter-
mediates are the predominant products under virtually
identical conditions.4 As the intermediate structure
between benzene and anthracene, we have studied the
naphthalene-based ligand and its tendency to form
either the bow-tie or slipped condensed intermediate.
Importantly, since naphthalene groups undergo ring
slippage more readily then single ring aromatic groups,
binuclear intermediates formed from 7 undergo more
modynamic product, or global minimum on the reaction
coordinate diagram, is the desired product of the reac-
tion. Hence, reactions that rely on these approaches are
referred to as being under thermodynamic control. In
the case of the Weak-Link Approach, the desired
product is a metastable intermediate structure, such as
2a or 3b, that is a local minimum on the reaction
coordinate diagram with the mononuclear product, 5a ,b,
being the ultimate thermodynamic sink of the reaction
(Scheme 1). Through the careful selection of reaction
conditions, the intermediates can be formed to the
exclusion of both oligomeric and monomeric byproducts.
Therefore, this approach is said to be under kinetic
control. The binuclear kinetic intermediates in this
reaction scheme can be reacted with small molecules,
before they are converted to mononuclear complexes, to
form the large ring structures that are typically the
synthetic targets. While higher ordered and structurally
dynamic architectures have been observed in all three
approaches,3 little mechanistic information is available
regarding the possible pathways and the relative ener-
getics of these conversions. The purpose of this paper
is to examine a novel system relevant to the Weak-Link
Approach in an attempt to understand the factors that
control Pathways 1 and 2 (Scheme 1). Elucidating the
reaction pathways is essential for fully understanding
and utilizing this synthetic strategy.
(4) Farrell, J . R.; Mirkin, C. A.; Guzei, I. A.; Liable-Sands, L. M.;
Rheingold, A. L. Angew. Chem., Int. Ed. 1998, 37, 465-467.
(5) Farrell, J . R.; Mirkin, C. A.; Liable-Sands, L. M.; Rheingold, A.
L. J . Am. Chem. Soc. 1998, 120, 11834-11835.
(6) Holliday, B. J .; Farrell, J . R.; Mirkin, C. A.; Lam, K.-C.;
Rheingold, A. L. J . Am. Chem. Soc. 1999, 121, 6316-6317.
(7) (a) Farrell, J . R.; Eisenberg, A. H.; Mirkin, C. A.; Guzei, I. A.;
Liable-Sands, L. M.; Incarvito, C. D.; Rheingold, A. L.; Stern, C. L.
Organometallics 1999, 18, 4856-4868. (b) Indeed, the direct reaction
of the Rh(I) transition metal starting material with ligand (7) in the
presence of a coordinating small molecule (i.e. CH3CN) results in the
formation of insoluble oligomers. When this same reaction mixture is
heated to 80 °C for 3.5 days we see quantitative formation of monomer
12 as confirmed by 31P{1H} NMR spectroscopy (see Supporting
Information). This demonstrates the necessity of the weak-links to
direct the formation of discrete closed rings in high yield via chelation.
(8) Dixon, F. M.; Eisenberg, A. H.; Farrell, J . R.; Mirkin, C. A.;
Liable-Sands, L. M.; Rheingold, A. L. Inorg. Chem. 2000, 39, 3432-
3433.
(9) Eisenberg, A. H.; Dixon, F. M.; Mirkin, C. A.; Stern, C. L.;
Incarvito, C. D.; Rheingold, A. L. Organometallics 2001, 20, 2052-
2058.
(10) (a) Liu, X.; Eisenberg, A. H.; Stern, C. L.; Mirkin, C. A. Inorg.
Chem. 2001, 40, 2940-2941. (b) Liu, X.; Stern, C. L.; Mirkin, C. A.
Organometallics 2002, 21, 1017-1019. (c) Ovchinnikov, M. V.; Holli-
day, B. J .; Mirkin, C. A.; Zakharov, L. N.; Rheingold, A. L. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 4927-4931.
Our group has extensively utilized the Weak-Link
Approach to construct flexible binuclear macrocyclic and
(3) (a) Schweiger, M.; Seidel, S. R.; Arif, A. M.; Stang, P. J . Inorg.
Chem. 2002, 41, 2556-2559. (b) Campos-Ferna´ndez, C. S.; Cle´rac, R.;
Koomen, J . M.; Russell, D. H.; Dunbar, K. R. J . Am. Chem. Soc. 2001,
123, 773-774. (c) Albrecht, M.; Napp, M.; Schneider, M.; Weis, P.;
Fro¨hlich, R. Chem. Commun. 2001, 409-410. (d) Beissel, T.; Powers,
R. E.; Parac, T. N.; Raymond, K. N. J . Am. Chem. Soc. 1999, 121,
4200-4206. (e) Chi, X.; Guerin, A. J .; Haycock, R. A.; Hunter, C. A.;
Sarson, L. D. Chem. Commun. 1995, 2563-2565. (f) Fuss, M.; Siehl,
H.-U.; Olenyuk, B.; Stang, P. J . Organometallics 1999, 18, 758-769.
(g) Hiraoka, S.; Fujita, M. J . Am. Chem. Soc. 1999, 121, 10239-10240.
(h) Marquis-Rigault, A.; Dupont-Gervais, A.; Baxter, P. N. W.; Van
Dorsselaer, A.; Lehn, J .-M. Inorg. Chem. 1996, 35, 2307-2310. (i)
Navarro, J . A. R.; J anik, M. B. L.; Freisinger, E.; Lippert, B. Inorg.
Chem. 1999, 38, 426-432. (j) Yamanoi, Y.; Sakamoto, Y.; Kusukawa,
T.; Fujita, M.; Sakamoto, S.; Yamaguchi, K. J . Am. Chem. Soc. 2001,
123, 980-981.
(11) (a) Slone, C. S.; Weinberger, D. A.; Mirkin, C. A. In Progress
In Inorganic Chemistry; Karlin, K. D., Ed.; J ohn Wiley & Sons: New
York, 1999; Vol. 48, pp 233-350. (b) Bader, A.; Lindner, E. Coord.
Chem. Rev. 1991, 108, 27-110.

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5715
Sch em e 2
ethoxyaryl chelates yield eight distinct resonances in
the 1-D proton spectrum. Additionally, the aromatic
region displays multiple resonances associated with the
break in symmetry of the naphthalene ring in this
coordination environment. The COSY spectrum of 8 in
CD2Cl2 was taken to elucidate the complicated shifts
and couplings shown in the standard proton experiment
(see Supporting Information). On the basis of the
correlation peaks in the COSY spectrum, the methylene
resonances of the carbons adjacent to the phosphorus
(1.7-2.7 ppm) and oxygen (3.7-5.4 ppm) atoms have
been unambiguously assigned (see Experimental Sec-
tion). All of the methylene resonances display coordina-
tion shifts (∆coord) between (0.1 and 1.3 ppm.
To confirm the methylene assignments and aid in the
analysis of the aromatic region of the 1H NMR spectrum,
the 13C{1H} NMR and HMQC (see Supporting Informa-
tion) spectra were recorded. The 13C{1H} NMR spectrum
of 8 displays distinct resonances for each carbon in the
ligand skeleton. The carbons adjacent to the phosphorus
atom display the expected C-P coupling with coupling
constants on the order of 30 Hz. Interestingly, of the
other carbons within three bonds of the phosphorus
atom, only the carbons on the chelated portion of the
ligand display further C-P coupling (2J P-C-C and
2J P-Rh-C), consistent with a Karplus-type effect.12,13 The
13C{1H} NMR spectrum of 8 shows large ∆coord values
for the carbon resonances of the 1/4 and 2/3 position
carbon atoms of the naphthalene ring relative to free
ligand, 30 and 20 ppm, respectively. The 9/10 carbon
resonances show only a minimal shift of 3 ppm. The
large shifts in the observed resonances of carbons 1-4
are consistent with a preference for an η4-bonding
interaction for the naphthalene rings,14 as observed in
the solid-state structure (vide infra).
Sch em e 3
facile conversion to the thermodynamic mononuclear
product. Therefore, one can study the kinetics of conver-
sion of condensed intermediates formed from 7 into the
respective mononuclear thermodynamic product to gain
insight into some of the important mechanistic factors
underlying this process.
31P{1H} NMR spectroscopy of 8 shows two resonances
at 25.7 (dd, J P-P ) 40 Hz, J Rh-P ) 205 Hz) and 34.9
ppm (dd, J P-P ) 40 Hz, J Rh-P ) 216 Hz) in dichlo-
romethane-d2. These chemical shifts, splitting patterns,
and coupling constants are characteristic of slipped
intermediate structures, which consist of Rh(I) metal
centers with two-legged piano-stool geometries.7 Even
though the arene-Rh(I) interaction in bisphosphine Rh-
(I) complexes with two-legged piano-stool geometries is
known to be relatively weak,15 ring slippage or other
fluxional behavior associated with the naphthalene
rings of 8, if they occur, are too fast to be observed on
the 31P{1H} NMR time scale in the temperature range
between -60 and +70 °C.
Resu lts
The new unsymmetric ligand 7, 1,4-bis(2-diphenyl-
phosphinoethoxy)naphthalene, was synthesized in two
steps. First, 1,4-dihydroxynaphthalene was reacted with
excess 1,2-dichloroethane in the presence of anhydrous
potassium carbonate and a catalytic amount of 18-
crown-6 in refluxing acetone to give 1,4-di(2-chloroeth-
oxy)naphthalene 6. Second, the alkoxy-chloro derivative
6 was converted to the corresponding diphenylphos-
phine 7 by reacting it with KPPh2 in THF (Scheme 3).
All spectroscopic, high-resolution mass spectrometry
and combustion analysis data for 6 and 7 are consistent
with the proposed structures (vide infra).
The binuclear Rh(I) slipped intermediate 8 was
synthesized in near-quantitative yield by reacting ligand
7 in THF with a Rh(I) precursor generated by the
reaction of [RhCl(COE)2]2 and AgBF4 in dichloromethane
at room temperature (Scheme 2). This method is analo-
gous to those used for preparing compounds 2a and 3b
from Rh(I) and their respective ligands (Scheme 1).4-8
Notably, this reaction yields the slipped isomeric prod-
uct 8 with no observation of the bow-tie product
containing arene rings stacked directly on top of one
another (Scheme 1).
(12) (a) Samitov, Y.; Safiullin, R. K. Zh. Obshch. Khim. 1988, 58,
983-989. (b) Lazhko, EÄ . I.; Luzikova, E. V.; Mikhailov, G. Y.;
Kazankova, M. A.; Ustynyuk, Y. A. Zh. Obshch. Khim. 1988, 58, 1247-
1258. (c) Ernst, L. Org. Magn. Reson. 1977, 9, 35-43. (d) Kainosho,
M. J . Phys. Chem. 1970, 74, 2853-2855.
(13) (a) Lambert, J . B.; Shurvell, H. F.; Verbit, L.; Cooks, R. G.;
Stout, G. H. Organic Structural Analysis; Macmillan Publishing: New
York, 1976. (b) Karplus, M. J . Chem. Phys. 1959, 30, 11-15. (c)
Karplus, M. J . Am. Chem. Soc. 1963, 85, 2870-2871. (d) Gutowsky,
H. S.; Karplus, M.; Grant, D. M. J . Chem. Phys. 1959, 31, 1278-1289.
(14) (a) Mu¨ller, J .; Gaede, P. E.; Hirsch, C.; Qiao, K. J . Organomet.
Chem. 1994, 472, 329-335. (b) Scha¨ufele, H.; Hu, D.; Pritzkow, H.;
Zenneck, U. Organometallics 1989, 8, 396-401. (c) Bennett, M. A.; Lu,
Z.; Wang, X.; Bown, M.; Hockless, D. C. R. J . Am. Chem. Soc. 1998,
120, 10409-10415. (d) Crabtree, R. H.; Parnell, C. P. Organometallics
1984, 3, 1727-1731. (e) Rieke, R. D.; Henry, W. P.; Arney, J . S. Inorg.
Chem. 1987, 26, 420-427. (f) Chin, R. M.; Dong, L.; Duckett, S. B.;
J ones, W. D. Organometallics 1992, 11, 871-876.
Solu tion Ch a r a cter iza tion of 8. The 1H NMR
spectrum of the condensed intermediate 8 reflects the
unsymmetric nature of the proposed solution and de-
termined solid-state structures (vide infra). The protons
of the methylene arms between the phosphorus and
(15) Singewald, E. T.; Shi, X.; Mirkin, C. A.; Schofer, S. J .; Stern,
C. L. Organometallics 1996, 15, 3062-3069.

5716 Organometallics, Vol. 21, No. 26, 2002
Holliday et al.
Å and Rh-P(bridging)ave ) 2.267 Å) are slightly shorter
than that of the benzene analogue, [µ2,η1:η6:η1- (1,4-(Ph2-
PCH2CH2O)2C6H4)2Rh2][BF4]2 (Rh-P(chelating)ave
)
2.274 Å) and (Rh-P(bridging)ave ) 2.296 Å); therefore,
the average P-Rh-P bond angle (95.38°) is slightly
larger than that of the analogue (95.06°).7
Rea ctivity of 8. Upon addition of acetonitrile to a
dichloromethane-d2 solution of 8, the weak Rh-arene
linkages are selectively cleaved to form a series of open
macrocycles (9a -d ), which are structural and coordina-
tion isomers of each other (Scheme 4). 31P{1H} NMR
spectroscopy of mixture 9 shows four doublets: two at
39.1 (d, J Rh-P ) 173 Hz) and 37.3 ppm (d, J Rh-P ) 173
Hz) for the cis-acetonitrile complexes 9a ,b, and two at
24.4 (d, J Rh-P ) 131 Hz) and 24.1 ppm (d, J Rh-P ) 131
Hz) for the trans-CH3CN-Rh(I)-NCCH3 complexes 9c,d
in a 13:4:2:1 ratio, respectively. These resonances are
assigned as the four structures that can result from the
combination of syn- and anti-conformations of the
naphthalene rings and the cis- and trans-coordination
modes of the phosphine and acetonitrile ligands (Scheme
4). The syn- and anti-confirmations of each pair of
coordination isomers could not be unambiguously dif-
ferentiated in solution. The formation of cis- and trans-
isomers with this class of compounds has been observed
with analogous macrocycles, and the spectroscopy for
those compounds compares well with that for the cis-
and trans-compounds in mixture 9.15,18-20 Interestingly,
mixture 9 consists of the coordination and conforma-
tional isomers associated with the different orientations
of the arene rings and the ligand environment around
the metal center. In contrast, the two macrocycles with
strictly trans-phosphine coordination, 10 and 11 (vide
infra), exist as single well-characterized products. This
can be explained by considering the arene-arene dis-
tance in each system. In X-ray diffraction studies of a
bis-acetonitrile (cis-) adduct analogous to 99 and 9 (cis-,
anti-) (vide infra), the average arene-arene distance is
4.05 Å in comparison to an average of 6.37 Å for 11 and
three other macrocycles previously reported.7-9 The
increased arene-arene distance allows for free rotation
about the 1,4-axis of the naphthalene rings in complexes
10 and 11 and therefore scrambling of the isomers on
the NMR time scale. When the solvent was evaporated
in vacuo and the resulting residue redissolved in dichlo-
romethane-d2, slipped intermediate 8 can be recovered
cleanly as evidenced by 31P{1H} NMR spectroscopy.
When a dichloromethane-d2 solution of slipped inter-
mediate 8 is charged with CO (1 atm) the color changes
from deep red to pale yellow with the concomitant
formation of a white precipitate (Scheme 4). The 31P-
{1H} NMR spectrum of the supernatant shows only one
broad doublet at 26.8 ppm with J Rh-P ) 90 Hz. This
compares well with the 31P{1H} NMR resonances for the
analogous trigonal-bipyramidal Rh(I) complexes which
have three equatorial CO ligands and two apical phos-
phine ligands: [(η1-PhO(CH2)2PPh2)2(CO)3 Rh][BF4],
31.5 ppm (d, J Rh-P ) 71.7 Hz); [(µ2-(1,4-(Ph2PCH2-
CH2O)2-2,3,5,6-((CH3)4C6)))2(CO)6 Rh2][BF4]2, 20 ppm (d,
F igu r e 1. ORTEP diagram of complex 8. Thermal el-
lipsoids are drawn at the 50% probability level. Hydrogen
atoms, solvent molecules, and anions are omitted for
clarity. See Table 1 for selected bond distances and angles.
X-r a y Cr ysta llogr a p h y of 8. The slipped arene
configuration of 8 was confirmed by X-ray crystal-
lography (Figure 1). Pale red single crystals suitable for
X-ray analysis were grown by slow diffusion of ether
into a dichloromethane solution of 8. One dichlo-
romethane and a diethyl ether molecule were cocrys-
tallized with 8 but exhibited severe disorder. Therefore,
they were refined isotropically, while the remaining non-
hydrogen atoms were refined anisotropically. The se-
lected bond lengths and angles are given in Table 1. In
the solid state, the two naphthalene rings of 8 exhibit
a syn conformation, and the observed Rh-C distances
are in the range for Rh(I) compounds with bis(phos-
phine), η6-arene piano-stool geometries (Rh-C ) 2.217-
2.516 Å).16,17 However, the naphthalene ring interaction
with the Rh(I) metal center might best be described as
η4 rather than η6 since there are three short, one
medium, and two long interatomic Rh-C distances: Rh-
(1)-C(1) ) 2.261(6) Å, Rh(1)-C(2) ) 2.211(6) Å, Rh-
(1)-C(3) ) 2.298(6) Å, Rh(1)-C(4) ) 2.419(6) Å, Rh(1)-
C(5) ) 2.569(6) Å, Rh(1)-C(6) ) 2.507(6) Å, Rh(2)-
C(39) ) 2.482(6) Å, Rh(2)-C(40) ) 2.360(6) Å, Rh(2)-
C(41) ) 2.253(6) Å, Rh(2)-C(42) ) 2.265(6) Å, Rh(2)-
C(43) ) 2.514(6) Å, Rh(2)-C(44) ) 2.622(6) Å. While
the η4 hapticity in structures 8 and 12 (vide infra) is
less pronounced than in other naphthalene-containing
systems as judged by the relatively small hinge angles
(6° and 8°, respectively) and 10% difference in bound
and unbound Rh-C bond lengths,14 the Rh(I) metal
center displays an off-center distortion in its coordina-
tion to the naphthalene ring that is statistically mean-
ingful and also displayed in the solution NMR studies.
The two rhodium metal centers are separated by 7.45
Å. The Rh-P bond lengths (Rh-P(chelating)ave ) 2.236
(16) (a) Slone, C. S.; Mirkin, C. A.; Yap, G. P. A.; Guzei, I. A.;
Rheingold, A. L. J . Am. Chem. Soc. 1997, 119, 10743-10753. (b)
Alvarez, M.; Lugan, N.; Donnadieu, B.; Mathieu, R. Organometallics
1995, 14, 365-370. (c) Westcott, S. A.; Taylor, N. J .; Marder, T. B.;
Baker, R. T.; J ones, N. J .; Calabrese, J . C. Chem. Commun. 1991, 304-
305. (d) Townsend, J . M.; Blount, J . F. Inorg. Chem. 1981, 20, 269-
271. (e) Albano, P.; Aresta, M.; Manassero, M. Inorg. Chem. 1980, 19,
1069-1072.
(18) Singewald, E. T.; Mirkin, C. A.; Levy, A. D.; Stern, C. L. Angew.
Chem., Int. Ed. Engl. 1994, 33, 2473-2475.
(19) Sassano, C. A.; Mirkin, C. A. J . Am. Chem. Soc. 1995, 117,
11379-11380.
(20) Appleton, T. G.; Clark, H. C.; Manzer, L. E. Coord. Chem. Rev.
1973, 10, 335-422.
(17) Singewald, E. T.; Slone, C. S.; Stern, C. L.; Mirkin, C. A.; Yap,
G. P. A.; Liable-Sands, L. M.; Rheingold, A. L. J . Am. Chem. Soc. 1997,
119, 3048-3056.

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5717
Ta ble 1. Selected Bon d Dista n ces (Å) a n d Bon d An gles (d eg) for 8, 9 (cis-, a n ti-), 11, a n d 12
9 (cis-, anti-) 11 12
8
Rh1-P1
2.2286(15)
2.2675(16)
2.2438(16)
2.2669(15)
2.261(6)
2.211(6)
2.298(6)
2.419(6)
2.569(6)
2.507(6)
2.482(6)
2.360(6)
2.253(6)
2.265(6)
2.514(6)
2.622(6)
95.82(6)
94.94(6)
Rh1-P1
2.450(15)
2.2412(15)
2.050(5)
2.039(5)
1.137(9)
1.146(10)
1.460(11)
1.465(11)
1.842(6)
1.836(6)
1.831(6)
Rh1-P1
2.3468(9)
2.3281(8)
1.824(4)
2.044(3)
1.141(4)
1.130(5)
1.466(6)
Rh1-P1
Rh1-P2
Rh1-C1
Rh1-C2
Rh1-C3
Rh1-C4
Rh1-C5
Rh1-C10
O1-C1
2.262(7)
2.248(7)
2.261(3)
2.265(2)
2.253(2)
2.252(2)
2.486(3)
2.473(3)
1.355(4)
1.373(3)
1.406(4)
1.399(4)
1.401(4)
1.444(3)
1.426(3)
1.460(4)
98.05(2)
Rh1-P3
Rh1-P2
Rh1-P2
Rh2-P2
Rh1-N1
Rh1-C39
Rh2-P4
Rh1-N2
Rh1-N1
Rh1-C1
N1-C39
C39-O3
Rh1-C2
N2-C41
C40-N1
Rh1-C3
C39-C40
C41-C42
P1-C21
C40-C41
Rh1-C4
P1-Rh1-P2
P1-Rh1-N1
P1-Rh1-C39
P2-Rh1-N1
P2-Rh1-C39
C39-Rh1-N1
Rh1-C39-O3
Rh1-N1-C40
N1-C40-C41
174.31(4)
Rh1-C5
89.95(8)
92.24(11)
90.43(8)
87.45(11)
177.71(13)
177.1(3)
176.3(3)
178.8(5)
Rh1-C6
P1-C15
O2-C4
Rh2-C39
Rh2-C40
Rh2-C41
Rh2-C42
Rh2-C43
Rh2-C44
P1-Rh1-P3
P2-Rh2-P4
P2-C27
C1-C2
P2-C33
1.825(6)
96.21(5)
C2-C3
P1-Rh1-P2
P1-Rh1-N1
N2-Rh1-N1
N2-Rh1-P2
N1-Rh1-P2
N2-Rh1-P1
C3-C4
175.69(18)
85.9(2)
169.44(14)
86.76(16)
91.52(14)
C4-C5
C5-C10
C1-C10
P1-Rh1-P2
related trans-CO-Rh(I)-NCCH3 complexes.4,7,15,21 Com-
plex 11 was also synthesized in quantitative yield by
addition of CO gas (1 atm) to a dichloromethane-d2
solution of 9 (Scheme 4).
Sch em e 4
The 1H NMR spectra of compounds 9 and 10 are
broad, indicating the presence of additional isomers,
bonding conformations, or fluctuations that are not
resolved on the NMR time scale. However, due to the
conformational stability of complex 11, the 1H NMR
spectroscopy signals are sharp and well-resolved, which
allows more detailed analysis of the resonances and
their corresponding shifts relative to free ligand 7. The
1H NMR spectroscopy resonance associated with the
methylene protons in the position R to the phosphine
in 11 was shifted downfield by 0.5 ppm relative to 7 via
an inductive effect of Rh(I)-P coordination, and the
proton signals of the naphthalene ring were shifted
upfield presumably due to the shielding effect of the
naphthalene rings on each other.
The two naphthalene rings in complexes 9, 10, and
11 can have three different conformations analogous to
related organic cyclophanes: syn-, anti-, or edge-face.22
1
The H NMR spectrum of 11 displays three resonances
J Rh-P ) 88 Hz); [(µ2-(1,4-(Ph2PCH2CH2O)2C6H4))2(CO)6
Rh2][BF4]2, 28.0 ppm (d, J Rh-P ) 78 Hz); [(µ2-(4,4′-(Ph2-
PCH2CH2O)2-C6H4-C6H4))2(CO)6 Rh2][BF4]2, 30.2 ppm
(d, J Rh-P ) 79 Hz).7,15 The CO complex was not stable
under ESI (electrospray ionization) mass spectrometry
conditions, thus precluding the gas-phase characteriza-
tion of the parent ion. However, a fragment of the
for the protons on the naphthalene ring. These signals
are all shifted upfield relative to the free ligand,
indicating that there is significant overlap between the
two aromatic rings in solution. More specifically, a
comparison between the signals of free ligand 7 and
macrocycle 11 yields a cyclization shift (∆cyc) for each
set of protons. The ∆cyc for the protons on C2 and C3 is
0.14, that for C5 and C8 is 0.24, and that for C6 and C7
is 0.22 (see numbering in Scheme 3). The perturbation
of all three resonances and the ∆cyc values are consistent
macrocycle with the loss of the 6 CO ligands, [M - 6CO
- +
- BF4
]
) m/z 1461.6, was observed. The FTIR
spectrum of 10 in dichloromethane-d2 exhibits two νCO
bands at 2013 (s) and 2093 (s) cm-1, consistent with the
proposed structure. The white precipitate (mentioned
above) is compound 10 and can be redissolved in
dichloromethane saturated with CO.
with analogous organic syn-naphthalenophanes (∆cyc
)
(21) Siedle, A. R.; Gleason, W. B.; Newmark, R. A.; Skarjune, R. P.;
Lyon, P. A.; Markell, C. G.; Hodgson, K. O.; Roe, A. L. Inorg. Chem.
1990, 29, 1667-1673.
Complex 10 was collected by filtration over Celite and
redissolved in acetonitrile (Scheme 4). As acetonitrile
was added, evolution of CO gas was observed. 31P{1H}
NMR spectroscopy shows the disappearance of the
doublet assigned to 10 and the formation of a new
doublet at 27.5 ppm (J Rh-P ) 118 Hz), which has been
assigned to the acetonitrile/CO adduct complex 11. The
31P{1H} NMR chemical shift and Rh-P coupling con-
stant correspond well with the analogous data for
(22) (a) Adams, S. P.; Whitlock, H. W. J . Org. Chem. 1981, 46, 3474-
3478. (b) J arvi, E. T.; Whitlock, H. W. J . Am. Chem. Soc. 1980, 102,
657-662. (c) Blank, N. E.; Haenel, M. W. Tetrahedron Lett. 1978,
1425-1428. (d) Chandross, E. A.; Ferguson, J .; McRae, E. G. J . Chem.
Phys. 1966, 45, 3546-3553. (e) Katul, J . A.; Zahlan, A. B. J . Chem.
Phys. 1967, 47, 1012-1014. (f) Haigh, C. W.; Mallion, R. B. Org. Magn.
Reson. 1972, 4, 203-228. (g) Chandra, A. K. Chem. Phys. Lett. 1970,
5, 229-231. (h) Murakami, Y.; Aoyama, Y.; Kida, M.; Nakano, A.;
Dobashi, K.; Tran, C. D.; Matsuda, Y. J . Chem. Soc., Perkin Trans. 1
1979, 1560-1567. (i) Shinmyozu, T.; Inazu, T.; Yoshino, T. Chem. Lett.
1978, 405-408. (j) Kawabata, T.; Shinmyozu, T.; Inazu, T.; Yoshino,
T. Chem. Lett. 1979, 315-318.

5718 Organometallics, Vol. 21, No. 26, 2002
Holliday et al.
F igu r e 4. Plot of kobsd vs concentration of 7 for the
conversion of 8 to 12 at T ) 40 °C.
F igu r e 2. ORTEP diagram of complex 9 (cis-, anti-).
Thermal ellipsoids are drawn at the 50% probability level.
Hydrogen atoms, solvent molecules, and anions are omitted
for clarity. See Table 1 for selected bond distances and
angles.
square-planar coordination geometries (L1-Rh-L2
angles of 85.9-96.2°, Table 1). The planes defined by
each metal center and its ligands are parallel planar
with the acetonitrile ligands pointing to opposite sides
of the macrocyclic ring (Figure 2).
X-r a y Cr ysta llogr a p h y of 11. Yellow single crystals
suitable for X-ray diffraction analysis were obtained by
layering diethyl ether over an acetonitrile solution of
11 (Figure 3 and Table 1). The X-ray diffraction data
show that complex 11 is located on a crystallographic
2-fold axis. Although two acetonitrile molecules cocrys-
tallized with 11, there is no evidence of any interaction
between these solvent molecules and the Rh(I) metal
centers. The two naphthalene rings are anti to one
another and parallel planar but do not overlap (dis-
placed by over 3 Å), and therefore do not exhibit
evidence of intramolecular π-π interactions in the solid
state. The coordination spheres of the rhodium metal
centers are best described as distorted square planar
with trans-bisphosphines (P(1)-Rh(1)-P(2) ) 174.31-
(4)°) and trans-CO, CH3CN (C(39)-Rh(1)-N(1) ) 177.71-
(13)°) as predicted by 31P{1H} NMR spectroscopy in
solution. The two rhodium metal centers are separated
by 11.56 Å comparable to that of a related duranyl
complex ([(µ2-(1,4-(Ph2PCH2CH2O)2-2,3,5,6-((CH3)4C6)))2-
(CO)2(CH3CN)2Rh2][BF4]2, 11.61 Å).7
F igu r e 3. ORTEP diagram of complex 11. Thermal
ellipsoids are drawn at the 50% probability level. Hydrogen
atoms, solvent molecules, and anions are omitted for
clarity. See Table 1 for selected bond distances and angles.
0-0.2 ppm), indicating that macrocycle 11 displays a
syn-conformation of the naphthalene rings relative to
one another.22 Interestingly, even though the spectro-
scopic data suggest that the naphthalene rings of 11
are in the syn-conformation in solution, the solid-state
structures of 9 (cis-, anti-) and 11 display an anti,
displaced, and nearly parallel planar conformation of
the naphthalene rings (Figures 2 and 3). This confor-
mational change is presumably due to crystal packing
effects maximizing intermolecular π-π interactions.
X-r a y Cr ysta llogr a p h y of 9 (cis-, a n ti-). Slow
diffusion of pentane into a concentrated 5:1 dichlo-
romethane:acetonitrile solution containing complexes
9a -d yielded pale yellow crystals suitable for X-ray
diffraction analysis. Interestingly, only one of the four
conformers/isomers present in the solution mixture 9
is observed in the solid-state structure (Figure 2). The
cis-phosphine, cis-acetonitrile complex of the anti-
naphthalene conformation 9 (cis-, anti-) cocrystalized
with two molecules of methylene chloride and two
molecules of water. The solvent molecules were severely
disordered and therefore treated in the data analysis
by the SQUEEZE correction method. The remaining
non-hydrogen atoms were refined anisotropically, and
selected bond distances and angles can be found in Table
1. The naphthalene rings are situated anti to one
another in a parallel-planar orientation but do not
overlap, demonstrating a lack of intramolecular π-π
interactions in the solid state. The two rhodium(I) metal
centers are 11.19 Å apart and display slightly distorted
Con ver sion of Bin u clea r Com p lex 8 to Mon o-
n u clea r Com p lex 12. The conversion of the slipped
intermediate 8 into the analogous mononuclear piano-
stool complex 12 was studied to gain insight into the
relative stabilities of Rh(I) complexes and mechanistic
information about this process (Scheme 2). The analo-
gous conversion process has been observed for related
ligand systems but never studied in detail.2,7 The
1
conversion of 8 to 12 was monitored by H and 31P{1H}
NMR spectroscopy with no sign of decomposition or side
reactions. Kinetic studies based on 1H NMR spectros-
copy were undertaken in the presence of phosphorus-
based ligands. The latter catalyze the associative dinu-
clear to mononuclear conversion process. The kinetics
reveal first-order dependence on macrocycle (8) concen-
tration, and the plot of kobsd vs [7] shows a linear
relationship with a zero intercept (Figure 4). The rate
of the conversion was studied over a 30-50 °C temper-
ature range in the presence of ligand 7 (Figure 5A).
From these data, the activation parameters have been
determined for the conversion (Figure 5A, inset). Ad-
ditionally, the kinetics of the conversion of 8 into 12
were performed in the presence of a series of phosphines
of varying steric and electronic parameters (Table 2).

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5719
F igu r e 6. COSY spectrum (500 MHz) of monomer 12 in
CD2Cl2 shown as a contour plot. The 500-MHz one-
dimensional 1H NMR spectrum is shown at the top and
left-hand edge. The signals that can be assigned are
marked accordingly. The dashed lines highlight peak
correlations.
F igu r e 5. (A) Temperature profile of the rate constants
for the conversion of 8 to 12 in the presence of excess ligand
7 plotted according to the Arrhenius model. The inset
shows the calculated activation parameters for this reac-
tion. (B) The reaction coordinate diagram profiling the
formation and catalyzed conversion of binuclear intermedi-
ate structure 8 to the mononuclear thermodynamic prod-
uct, 12.
adjacent aromatic protons, d and e (7.6-8.0 ppm),
display the expected correlations as well. The remaining
nondiphenylphosphine proton, c, displays a singlet
resonance at 6.96 ppm with a self-correlated peak in
the COSY spectrum. The resonances associated with
protons a/a′ and c display significant coordination shifts
(∆coord) of -0.7 and +0.4 ppm ∆coord (respectively), which
are consistent with the proposed solution and solid-state
structures.
The rigid nature of 12 is confirmed by the observation
of only single correlations between the methylene
protons on adjacent carbons (a to b only), a demonstra-
tion of the Karplus effect in this system.13 For example,
based on the X-ray structure of 12 (vide infra), the
dihedral angles between the proton labeled a and the
protons on the adjacent carbon, b and b′, are 169° and
-76° (respectively). According to the Karplus relation-
ship, the coupling of the vicinal (3J H-C-C-H) spins a to
b should be ∼9.5 Hz, and a to b′ should not exhibit spin-
spin coupling. This correlation pattern is observed in
the COSY spectrum of 12 indicating a high level of
agreement between the solution and solid-state struc-
tures. Identical analyses can be performed for the
remaining seven methylene protons, and all dihedral
angle/spin-spin coupling relationships agree with the
observed data.
Ta ble 2. Effect of P h osp h in e Liga n d on kobsd of 8
to 12 Con ver sion a
entry
ligand
PBu3
PCy3
PEtPh2
PPh3
P(C6F5)3
P(OMe)Ph2
BDPOb
7
kobsd (10-5 s-1
)
cone angle (deg)
a
b
c
d
e
f
7.52
1.08
4.42
3.98
0.558
6.97
5.93
7.47
132
170
140
145
184
132
140c
g
h
a
The conversion of 8 to 12 was measured in CD2Cl2 at T ) 40
°C in the presence of 0.2 equiv of monophosphine (a-f) or 0.1 equiv
of diphosphine (g and h) by 1H NMR spectroscopy against a decane
b
internal standard. Bis(1,8-diphenylphosphino)octane. c Estimat-
ed based on PEtPh2.
Solu t ion Ch a r a ct er iza t ion of 12. To elucidate
the solution structure of 12 and make a complete
assignment of the one-dimensional 1H and 13C{1H}
NMR spectra, the H,H-COSY and two-dimensional
H,C-HMQC (see Supporting Information) spectra were
recorded in CD2Cl2. The splitting of each methylene
peak into two distinct resonances in the 1H NMR
spectrum of 12 is diagnostic of the lack of symmetry
about the O-Rh-O plane. All of the methylene reso-
nances (a/a′ and b/b′) are broad with unresolved cou-
pling; however, the COSY spectrum of 12 displays the
expected correlations of a to a′ and b to b′ for methylene
protons on the same carbon atom (Figure 6). The
By comparing the 13C{1H} NMR spectrum of 12 with
the 13C{1H} NMR spectrum of free ligand and analyzing
the 1H NMR (vide infra) and HMQC spectrum of 12 (see
Supporting Information), the complete assignment of
the 13C{1H} NMR spectrum of 12 has been made. The
HMQC spectrum displays the expected correlations
1
between the assigned H NMR spectrum and the 13C-
{1H} NMR resonances. All carbon resonances show a

5720 Organometallics, Vol. 21, No. 26, 2002
Holliday et al.
C6)))Rh][BF4] (Rh-Pave ) 2.2388(12) Å),7 presumably
due to the trans influence of the naphthalene ring via
a strong arene-Rh(I) interaction. This bond distance
falls within the range of previously reported mono-
nuclear Rh(I) piano-stool structures (average Rh-P
bond lengths ) 2.219-2.251 Å).7,16,17,21 The crystal
packing diagram of 12 shows that the naphthalene rings
from two independent molecules of 12 do not π-π stack
on one another, even though a similar η6-anthraceno-
diphosphine-silver(I) complex displays π-π stacking
interactions between the two anthracyl rings in the unit
cell (arene-arene distance ) 3.416 Å).24
R ea ct ivit y of Mon on u clea r Com p ou n d 12. The
mononuclear Rh(I) piano-stool complex 12 does not show
any reactivity with CO or acetonitrile alone in solution
(at room temperature) even though the arene-Rh(I)
bond is known to be rather weak.15,18,19 However, when
complex 12 was charged with CO in the presence of
acetonitrile the color changed from deep red to pale
yellow due to the formation of mononuclear trans-CO-
Rh(I)-NCCH3 complex 13, which results from the cleav-
age of this weak arene-Rh(I) interaction (Scheme 2).
The 31P{1H} NMR spectrum of this solution exhibits a
doublet at 23.6 ppm with coupling constant J Rh-P ) 117
Hz, which compares well with that for binuclear trans-
CO-Rh(I)-NCCH3 complex 11 (J Rh-P ) 118 Hz) and
previously reported related complexes.4,7,15,21 The 1H
NMR spectrum and mass spectroscopy data for 13 are
also consistent with the proposed structure. Finally,
complex 12 undergoes reversible one-electron oxidation
(E1/2 ) 660 mV vs Fc/Fc+) forming a stable Rh(II)
complex (Scheme 2). This is consistent with what has
been observed with other Rh(I) two-legged piano-stool
analogues of 12.17,18,25
F igu r e 7. ORTEP diagram of complex 12. Thermal
ellipsoids are drawn at the 50% probability level. Hydrogen
atoms, solvent molecules, and anions are omitted for
clarity. See Table 1 for selected bond distances and angles.
relatively small ∆coord ((3-6 ppm) except for the two
resonances of the carbons directly bound to the metal
center, which display a ∆coord of -17 (C1/4) and -32 (C2/3
)
ppm. These coordination shifts and assignments are
consistent with assignments for previously studied η4-
coordination modes for untethered naphthalene ligands
and confirm the η4-coordination of the arene ring to the
Rh(I) in 12 in solution.14 In addition to the peak
positions and shifts, the observed coupling patterns are
also consistent with the proposed solution structure. The
resonance associated with the carbon atoms adjacent
to the phosphorus atoms displays a second-order cou-
pling pattern that can be modeled as a AXY system
indicating the expected coupling to the adjacent phos-
phorus atom as well as the 103Rh metal center, 2J Rh-P-C
.
The resonances assigned to the carbons that are directly
interacting with the Rh(I) metal center in an η4-fashion
both display the expected 103Rh-13C spin-spin coupling
(J Rh-C) of 3-4 Hz, which is in agreement with previ-
ously reported carbon atoms π-bound to Rh(I).23 The
lack of a large ∆coord and Rh-C coupling for carbons 9
and 10 further supports an η4, as opposed to η6,
coordination mode for the Rh(I) metal center in solution.
All of these data are consistent with the proposed
structure for complex 12 and demonstrate excellent
agreement between the solution and solid-state struc-
tures.
X-r a y Cr ysta llogr a p h y of 12. Slow diffusion of
pentane into a saturated dichloromethane solution of
12 gave prismatic orange crystals suitable for X-ray
diffraction analysis (Figure 7 and Table 1). The observed
Rh-C distances are in the expected range for mono-
meric Rh(I) compounds with a bis(phosphine), η6-arene
piano-stool geometry (Rh-C ) 2.217-2.516 Å).16,17
However, the naphthalene ring is actually bound to the
Rh(I) metal center via an η4-bonding mode rather than
η6-bonding mode as evidenced by two long (Rh(1)-C(5)
) 2.486(3) Å, Rh(1)-C(10) ) 2.473(3) Å) and four short
(Rh(1)-C(1) ) 2.261(3) Å, Rh(1)-C(2) ) 2.265(2) Å, Rh-
(1)-C(3) ) 2.253(2) Å, Rh(1)-C(4) ) 2.252(2) Å) Rh-C
bonds. The Rh-P bond lengths (Rh(1)-Pave ) 2.2550-
(7) Å) are slightly longer than that of the duranyl
analogue, [(η1:η6:η1-(1,4-(Ph2PCH2CH2O)2-2,3,5,6-((CH3)-
Discu ssion
In addition to the reactivity outlined above, ligand 7
and its binuclear Rh(I) complex, 8, have afforded us the
opportunity to study the conversion of binuclear inter-
mediates to mononuclear complexes and to map out
some of the possible reaction pathways and energetics
associated with the Weak-Link Approach.
As reported previously, for macrocycles prepared via
the Weak-Link Approach, the two-legged piano-stool
mononuclear Rh(I) complex is often the thermodynamic
product with respect to the slipped and bow-tie inter-
mediates (Scheme 1).7 Since the naphthalene complex,
8, has the potential to undergo ring-slippage and
numerous ligand-metal binding modes available to
facilitate ligand substitution, this system is especially
well-suited for studying the conversion of dimer 8 to
monomer complex 12 (Scheme 2). Notably these trans-
formations are much faster compared to the analogous
transformation involving the previously reported durene
complex [(η1:η6:η1-(1,4-(Ph2PCH2CH2O)2-2,3,5,6-((CH3)-
C6)))Rh][BF4], which takes 10 d at 75 °C in ClCD2CD2-
Cl to undergo dimer-to-monomer conversion. Addition-
ally, the latter transformation is accompanied by
significant decomposition (10-20%).7
(24) Xu, F.-B.; Weng, L.-H.; Sun, L.-J .; Zhang, Z.-Z.; Zhou, Z.-F.
Organometallics 2000, 19, 2658-2660.
(23) (a) Bodner, G. M.; Storhoff, B. N.; Doddrell, D.; Todd, L. J .
Chem. Commun. 1970, 1530-1531. (b) Arthurs, M.; Gross, P.; Kubal,
G.; Paniwnyk, L.; Curzon, E. J . Organomet. Chem. 1989, 366, 223-
243.
(25) 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.

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5721
Sch em e 5
Upon extended heating (48 h) of analytically pure 8
under conditions similar to those previously reported
(vide supra), no conversion to monomer was observed.
However, the addition of a small amount of free ligand
7 to this solution brought about complete conversion
within 20 min under identical conditions. In the previ-
ously reported system,7 the small amount of decomposi-
tion that was observed provided a small but increasing
amount of free phosphine ligand, which we now believe
catalyzed the conversion to monomeric product. Given
the dependence of the conversion on the presence of
excess ligand and the absence of phosphine bound to
12 in the final product mixture, the formation of
monomer from binuclear species is thought to be an
associative process. With this in mind, we performed
kinetic experiments aimed at studying the conversion
of dimer 8 to monomer 12 in the presence of 0.1 equiv
of free ligand (7) between 30 and 50 °C in CD2Cl2. The
process was monitored by 1H NMR spectroscopy against
a decane internal standard. These experiments all
displayed clean first-order kinetics based on macrocycle
concentration and a linear temperature dependence,26
which allowed us to calculate the activation parameters
for dimer to monomer conversion (Figure 5). Addition-
ally, the rate of reaction is first order in phosphorus-
based nucleophile. The plot of kobsd vs [7] is linear with
a zero intercept, consistent with an associative process
(Figure 4).
The reaction proceeds with a forward activation
energy of 54.8 kJ /mol and an activation enthalpy of 52.2
kJ /mol (Figure 5B). This relatively small enthalpy of
activation is indicative of a transition state where bond
making is closely accompanied by bond breaking, which
lends support to the hypothesis that this reaction is
associative in nature and that the first steps of ligand
binding are rate limiting. The large negative activation
entropy (-158.1 J /(mol K)) indicates the formation of a
highly ordered transition state, which is consistent with
an associative process that brings 8 and 1 equiv of 7
(or 2 equiv of monophosphine) together prior to dis-
sociation to a mononuclear complex. On the basis of
these observations and earlier studies of associative
substitution reactions at Rh(I) facilitated by ring-
slippage of coordinated arene rings,27-30 we propose a
mechanism that involves the sequential and rate de-
termining coordination of incoming ligands to the metal
centers of dimer 8, with concomitant ring slippage of
the coordinated arenes from η4 to η2 (accommodating
the additional ligand and maintaining the 16-electron
count on the rhodium metal center) (14 in Scheme 5).
This is followed by dissociation of the bridging phos-
phine and ring slippage back to an η4-coordination mode
to give 2 equiv of monomeric intermediate, 15. The
observed product 12 can then be formed by associative
attack of the free chelating phosphine, again with an
η4 to η2 change in arene coordination mode, followed by
loss of the catalytic phosphine ligand and reestablish-
ment of the η4-coordination mode (Scheme 5). Consistent
with this hypothesis, intermediates cannot be observed
even by the addition of excess ethyldiphenylphosphine
at -78 °C and gradual warming to room temperature
while monitoring by 31P{1H} NMR spectroscopy.
By using different phosphines with systematically
varied steric and electronic properties as catalytic
reagents, one can gain further insight into this process;
specifically, information about the transition state and
rate determining step of the reaction can be inferred
from effects on the rate of reaction. Six commercially
available monodentate phosphines of varying steric and
electronic parameters were investigated in addition to
the bidentate ligands 7 and bis(1,8-diphenylphosphino)-
octane (BDPO) (Table 2).31 For comparison purposes,
0.2 equiv (consistent equivalent of phosphine) of the
monodentate phosphines were added to CD2Cl2 solu-
tions of 8, and the conversion was monitored by 1H NMR
spectroscopy at 40 °C against a decane internal stan-
dard. BDPO was used in these experiments in order to
test the role of both the chelation effect and the
hemilabile nature of ligand 7 to facilitate the conversion
of 8 to 12. BDPO is of comparable P-P length to 7 but
lacks the aromatic core and the labile oxygen moieties.
The rate of conversion of condensed intermediate 8
into mononuclear complex 12 varies with the ligand
introduced to catalyze the process, but all processes
follow first order (in macrocycle) kinetics. The data
listed in Table 2 were compared to four parameters for
the phosphine ligands, including cone angle,32,33 elec-
tronic parameter (Tolman),32 QALE sigma donation
ability,34 and basicity.35 The ligand size, as measured
by the cone angle of the phosphines, is the dominate
(31) 1,8-Bis(diphenylphosphino)octane (BDPO) was prepared in 80%
yield by the reaction of KPPh2 with 1,8-dibromooctane in THF at room
temperature followed by filtration through a plug of alumina and
removal of solvent. The crude colorless solid was purified by passing
through a short plug of silica gel in CH2Cl2 and recrystallizing with
pentane at -30 °C. The composition and purity was confirmed by 1H
and 31P{1H} NMR spectroscopy as compared to the following: Hill,
W. E.; Minahan, D. M. A.; Taylor, J . G.; McAuliffe, C. A. J . Chem.
Soc., Perkin Trans. 2 1982, 327-330.
(26) (a) Espenson, J . H. Chemical Kinetics and Reaction Mecha-
nisms, 2nd ed.; McGraw-Hill: New York, 1995. (b) Pilling, M. J .;
Seakins, P. W. Reaction Kinetics; Oxford University Press: New York,
1995.
(27) (a) Kataoka, Y.; Saito, Y.; Shibahara, A.; Tani, K. Chem. Lett.
1997, 621-622. (b) Calhorda, M. J .; Roma˜o, C. C.; Veiros, L. F. Chem.
Eur. J . 2002, 8, 868-875.
(28) (a) Basolo, F. New J . Chem. 1994, 18, 19-24. (b) Basolo, F.
Inorg. Chim. Acta 1985, 100, 33-39. (c) Rerek, M. E.; Basolo, F. J .
Am. Chem. Soc. 1984, 106, 5908-5912. (d) Rerek, M. E.; J i, L.-N.;
Basolo, F. Chem. Commun. 1983, 1208-1209.
(29) Rerek, M. E.; Basolo, F. Organometallics 1983, 2, 372-376.
(30) Zhang, S.; Shen, J . K.; Basolo, F.; J u, T. D.; Lang, R. F.; Kiss,
G.; Hoff, C. D. Organometallics 1994, 13, 3692-3702.
(32) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987.
(33) Tolman, C. A. Chem. Rev. 1977, 77, 313-348.
(34) (a) Fernandez, A. L.; Wilson, M. R.; Prock, A.; Giering, W. P.
Organometallics 2001, 20, 3429-3435. (b) Fernandez, A. L.; Reyes,
C.; Prock, A.; Giering, W. P. J . Chem. Soc., Perkin Trans. 2 2000, 1033-
1041.

5722 Organometallics, Vol. 21, No. 26, 2002
Holliday et al.
would again expect to see an observable amount of
intermediate 15. The data also agree well with the
relatively small value for the enthalpy of activation and
indicate the importance of the bond formation in the
transition state of the reaction, which is directly affected
by the steric bulk of the incoming ligand and its ability
to access the transition metal centers of complex 8.
Additionally, the bidentate phosphines (BDPO and 7)
can be compared to evaluate the role of chelation when
7 is used to catalyze the conversion. Specifically, when
one considers the structural similarities and differences
of BDPO and 7, they can be compared to gain insight
into the importance of the hemilabile nature of ligand
7 in the conversion of 8 into 12. Because the rate of
reaction catalyzed by 7 is only slightly faster than that
of the BDPO-catalyzed reaction (1.25 times larger rate
constant, Table 2), it is concluded that the hemilabile
nature of 7 and the presence of an arene ring capable
of binding to the metal play a minor role, if any, in the
phosphine-catalyzed conversion of 8 into 12.
Con clu sion s
In summary, we report binuclear and mononuclear
rhodium macrocycles based on the unsymmetric 1,4-bis-
(2-diphenylphosphinoethoxy)naphthalene ligand that
have been fully characterized by various means includ-
ing multidimensional NMR and X-ray crystallography.
The use of a naphthalene aromatic core in the title
ligand and its ability to facilitate different binding
modes to the rhodium(I) metal center via ring slippage
have provided the first mechanistic insight into the
process by which this class of kinetically stable binuclear
complexes converts to the corresponding mononuclear
complexes. Specifically, based on kinetic studies of this
process, we propose an associative mechanism for this
conversion catalyzed by the addition of phosphine
ligands. These studies demonstrate the importance of
the hemilabile nature of the ligands used in the Weak-
Link Approach and the significance of the careful control
of reaction conditions, including stiochiometry, for the
synthesis and isolation of the desired metastable bi-
nuclear intermediate structures.
F igu r e 8. (A) Plot of log kobsd vs cone angle of the
phosphorus nucleophile for the conversion of 8 to 12 at T
) 40 °C. The letters correspond to those in Table 2. (B)
Plot of log kobsd vs the electronic parameter33 of the
phosphorus nucleophile for the conversion of 8 to 12 at T
) 40 °C. The letters correspond to those in Table 2.
factor controlling the rate of the reaction, which has
been previously observed with associative substitution
reactions.29 A plot of rate of conversion versus cone angle
is shown in Figure 8A and indicates a linear correlation
between reaction rate and steric bulk of the nucleophile.
Other ligand parameters showed poor correlation to the
reaction rate data (Figure 8B), demonstrating that
basicity and donor ability of the incoming nucleophile
must be of secondary importance in determining the
rate of conversion.30,36 This behavior is well-known and
completely consistent with previous studies of the
associative substitution reactions of late transition
metal centers.28,29,37 If the rate determining step of this
process was formation of monomer 12 from intermediate
15, one would expect the rate of conversion to be
inversely dependent on the strength of the nucleophile-
Rh bond being broken in this step and, therefore, the
basicity of the phosphine used to initiate the reaction.
Additionally, the observation of a build-up of the
concentration of 15 (or its chelated analogue) in solution
would be expected. If the slow step in this reaction
scheme was the attack of the pendant phosphine arm
of 15 on the sterically crowded Rh(I) metal center one
Exp er im en ta l Section
Gen er a l Meth od s. Unless otherwise noted, all reactions
and manipulations were carried out under a nitrogen atmos-
phere using standard Schlenk techniques. Diethyl ether and
tetrahydrofuran (THF) were dried and distilled over sodium
and benzophenone. Acetonitrile, methylene chloride, and pen-
tane were dried and distilled over calcium hydride. [Rh(Cl)-
(COE)2]2 (COE ) cyclooctene) was prepared according to a
literature method using RhCl3 and COE in 2-propanol.38
Commercial purity carbon monoxide was purchased from
Matheson Gas and passed through a 15-cm column of Drierite
before use. All other chemicals were purchased from common
commercial sources (Aldrich, TCI, Acros, and Cambridge
Isotopes) and used as received. The purity of the phosphine
reagents utilized in the kinetic studies was verified by 31P-
{1H} NMR spectroscopy prior to their use. One-dimensional
NMR spectra were recorded on a Varian 300-MHz spectrom-
eter. Two-dimensional and 13C{1H} NMR (125 MHz) spectra
(35) Hudson, H. R. In The Chemistry of Organophosphorus Com-
pounds; Hartley, F. R., Ed.; Wiley & Sons: New York, 1990; Vol. 1, pp
473-488.
(36) (a) Morris, D. E.; Basolo, F. J . Am. Chem. Soc. 1968, 90, 2531-
2535. (b) Schuster-Woldan, H. G.; Basolo, F. J . Am. Chem. Soc. 1966,
88, 1657-1663. (c) Thorsteinson, E. M.; Basolo, F. J . Am. Chem. Soc.
1966, 88, 3929-3936.
(37) (a) Howell, J . A. S.; Dixon, D. T.; Kola, J . C.; Ashford, N. F. J .
Organomet. Chem. 1985, 294, C1-C4. (b) Butler, I. S.; Ismail, A. A.
Inorg. Chem. 1986, 25, 3910-3914.
(38) Reagents for Transition Metal Complex and Organometallic
Synthesis; Angelici, R. J ., Ed.; J ohn Wiley & Sons: New York, 1990;
Vol. 28.

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5723
of 8 and 12 were recorded on a Varian 500-MHz spectrometer.
1H NMR signals are reported relative to residual proton
resonances in deuterated solvents. 31P{1H} NMR spectra were
recorded at 121 MHz and referenced versus an 85% H3PO4
external standard. 13C{1H} NMR spectra were recorded at 75
MHz. All chemical shifts are reported in ppm and coupling
constants in Hz. Elemental analyses were performed by QTI,
Whitehouse, NJ (www.qtionline.com). Combustion analysis
was not used for compounds 9, 10, 11, and 13 due to the
relative lability of the acetonitrile and carbon monoxide ligands
under combustion conditions. Electrochemical measurements
were carried out with a PINE AFRDE5 bipotentiostat/gla-
vanostat using a Au electrode with a Pt mesh counter electrode
and a silver wire quasireference electrode in a 0.1 M n-Bu4-
NPF6/CH2Cl2 solution. Electrochemical data were referenced
using the Fc/Fc+ (Fc ) (η5-C5H5)Fe (η5-C5H5)) redox couple as
an internal standard.
Syn th esis of 1,4-Di(2-ch lor oeth oxy)n a p h th a len e (6). A
mixture of 1,4-dihydroxynaphthalene (2.0 g, 12.49 mmol), 1,2-
dichloroethane (25 mL, 0.32 mol), potassium carbonate (8.0
g, 57.89 mmol), and 18-crown-6 (0.5 g, 1.89 mmol) in acetone
(25 mL) was heated at reflux with vigorous stirring for 24 h.
The mixture was allowed to cool to room temperature and
diluted with dichloromethane (∼300 mL). The reaction mixture
was vacuum filtered through a Celite and silica gel pad and
evaporated to dryness. The crude product was titrated with
ethanol, which gave analytically pure product (2.9 g, 82%). Mp
121-123 °C. 1H NMR (CDCl3): δ 3.95 (t, J H-H ) 6.0, 4H, CH2-
Cl), 4.37 (t, J H-H ) 6.0, 4H, CH2O), 6.71 (s, 2H, C10H6), 7.55
(m, 2H, C10H6), 8.27 (m, 2H, C10H6). 13C{1H} NMR (CD2Cl2):
δ 42.39 (s, CH2Cl), 68.98 (s, OCH2), 105.11 (s, C2/3), 122.05 (s,
solvent was pumped off in vacuo and the resulting solid was
dried overnight at 55 °C under vacuum (80 mg, 92%). Dec pt
184 °C. H NMR (CD2Cl2): δ 1.88 (m, 2H, CH2PPh2), 2.11 (m,
2H, CH2PPh2), 2.43 (m, 2H, CH2PPh2), 2.58 (m, 2H, CH2PPh2),
3.86 (m, 2H, CH2O), 4.00 (m, 2H, CH2O), 4.62 (m, 2H, CH2O),
5.35 (m, 2H, CH2O), 6.22 (m, 4H, P(C6H5)2), 6.62 (m, 4H,
P(C6H5)2 + C10H6), 6.87 (m, 14H, P(C6H5)2), 7.07 (m, 2H,
1
P(C6H5)2), 7.19 (m, 4H, P(C6H5)2), 7.39 (m, 8H, P(C6H5)2
+
C
10H6), 7.67 (m, 2H, P(C6H5)2), 7.78 (m, 6H, P(C6H5)2 + C10H6),
8.15 (m, 4H, P(C6H5)2), 8.34 (d, J H-H ) 8.4, 2H, C10H6), 8.42
(d, J H-H ) 6.6, 2H, C10H6). 13C{1H} NMR (125.641 MHz, CD2-
Cl2): δ 26.31 (d, J C-P ) 31.4, CH2PPh2), 30.46 (d, J C-P ) 26.8,
CH2PPh2), 66.41 (s, CH2O), 69.40 (d, J C-P ) 16.5, CH2O), 78.49
(s, C2/3), 89.15 (s, C2/3), 115.75 (s, C1/4), 117.09 (dd, J C-P ) 7.5,
J C-Rh ) 2.5, C1/4), 119.88 (s, C5/8), 120.35 (s, C5/8), 121.24 (s,
C
9/10), 127.90 (s, C9/10), 128.13 (m, P(C6H5)2), 128.30 (m,
P(C6H5)2), 128.56 (m, P(C6H5)2), 128.89 (s, C6/7), 129.29 (s, C6/7),
129.63 (m, P(C6H5)2), 130.44 (m, P(C6H5)2), 130.59 (m, P(C6H5)2),
130.91 (m, P(C6H5)2), 131.05 (m, P(C6H5)2), 131.36 (m, P(C6H5)2),
131.60 (m, P(C6H5)2), 131.72 (m, P(C6H5)2), 132.50 (m, P(C6H5)2),
132.52 (m, P(C6H5)2), 133.20 (m, P(C6H5)2), 134.55 (m, P(C6H5)2),
137.15 (m, P(C6H5)2). 31P{1H} NMR (CD2Cl2): δ 25.66 (dd, J P-P
) 40, J Rh-P ) 205), 34.91 (dd, J P-P ) 40, J Rh-P ) 216). MS
-
+
(ESI, m/z): [M - BF4
]
+ 1461.7 (calcd for C76H68BF4O4P4Rh2
1461.9), [M - 2BF4
]
- 2+ 687.5 (calcd for C76H68O4P4Rh22+ 687.5).
Anal. Calcd for C76H68B2F8O4P4Rh2: C, 58.91; H, 4.43; P, 8.00.
Found: C, 58.24; H, 4.48; P, 7.84.
Syn th esis of [(µ2-1,4-Bis(2-(diph en ylph osph in o)eth oxy)-
n a p h th a len e)2Rh 2(CD3CN)4][BF 4]2 (9). Three drops of ace-
tonitrile-d3 was added to a CD2Cl2 solution of 8 (10 mg in 0.5
mL). The solution color changed from deep-red to yellow
indicating the formation of 9. 1H NMR (CD2Cl2): δ 2.97 (m,
8H, CH2PPh2), 4.58 (m, 8H, CH2O), 6.66 (m, 4H, C10H6), 6.86-
8.14 (m, 48H, P(C6H5)2 + C10H6). 31P{1H} NMR (CD2Cl2): δ
39.1 (d, J Rh-P ) 172.8), 37.3 (d, J Rh-P ) 172.8), 24.4 (d, J Rh-P
C
9/10), 126.48 (s, C5/8), 126.77 (s, C6/7), 148.82 (s, C1/4). HRMS
(EI, m/z): 284.03715 (calcd for C14H14Cl2O2 284.03708). Anal.
Calcd for C14H14Cl2O2: C, 58.97; H, 4.95. Found: C, 58.72; H,
4.97.
) 130.6), 24.1 (d, J Rh-P ) 130.6). MS (ESI, m/z): [M - 4CD3-
Syn th esis of 1,4-Bis(2-(d ip h en ylp h osp h in o)eth oxy)-
n a p h th a len e (7). KPPh2 (0.5 M, 4.4 mL, 2.20 mmol) in 50
mL of THF was added dropwise to a THF (100 mL) solution
of 6 (0.35 g, 1.07 mmol) under ice-bath cooling for 1 h and the
mixture was stirred overnight at room temperature. After the
evaporation of solvent, the organic materials were extracted
with dichloromethane from water. Combined organic phases
were dried using anhydrous MgSO4 and then in vacuo. The
white solid was suspended in EtOH (30 mL) and isolated via
cannula filtration. The resulting solid was dried under vacuum
overnight (0.55 g, 88%). Mp 145-146 °C. 1H NMR (CD2Cl2):
δ 2.71 (t, J H-H ) 7.2, 4H, CH2PPh2), 4.25 (dt (pseudoquartet),
J H-H ) 7.2, J P-H ) 9.6, 4H, CH2O), 6.57 (s, 2H, C10H6), 7.39
(m, 12H, P(C6H5)2 + C10H6), 7.51 (m, 10H, P(C6H5)2), 7.89 (m,
2H, C10H6). 13C{1H} NMR (CD2Cl2): δ 28.94 (d, J C-P ) 13.1,
CH2PPh2), 66.58 (d, J C-P ) 23.9, CH2O), 104.77 (s, C2/3), 122.28
(s, C5/8), 126.16 (s, C6/7), 126.80 (s, C9/10), 129.07 (d, J C-P ) 6.5,
Cm-P), 129.26 (s, Cp-P), 133.26 (d, J C-P ) 19.1, Co-P), 138.96
(d, J C-P ) 12.8, Ci-P), 148.95 (s, C1/4). 31P{1H} NMR(CD2Cl2):
δ -21.29 (s). HRMS (EI, m/z): 584.20344 (calcd for C38H34O2P2
584.20340). Anal. Calcd for C38H34O2P2: C, 78.06; H, 5.87; P,
10.60. Found: C, 77.85; H, 5.79; P, 10.37.
Syn th esis of [(µ2,η1:η4:η1-(1,4-Bis(2-(diph en ylph osph in o)-
eth oxy)n a p h th a len e))2Rh 2] [BF 4]2 (8). In a glovebox [Rh-
(Cl)(COE)2]2 (40 mg, 1.13 × 10-4 mol) and AgBF4 (22 mg, 1.13
× 10-4 mol) were reacted in 3 mL of CH2Cl2 for 30 min. The
resulting reaction mixture was filtered through Celite, and the
filtrate was diluted with 125 mL of THF. To this, a solution of
7 (66 mg, 1.129 × 10-4 mol) in 125 mL of THF was added
dropwise at -78 °C over 2 h and then warmed to room
temperature over 1 h. The solvent was removed under vacuum
to yield an orange-red powder. The solid was dissolved in a
small amount of CH2Cl2 and precipitated with pentane. The
resulting orange-red solid was collected over Celite and
redissolved in CH2Cl2. A few drops of CH3CN were added to
this solution and stirred for 20 min at room temperature. The
CN - BF4
]
+ 1461.2 (calcd for C76H68BF4O4P4Rh2+ 1461.9), [M
-
-
- 4CD3CN - 2BF4
]
2+ 687.4 (calcd for C76H68O4P4Rh22+ 687.5).
Syn th esis of [(µ2-1,4-Bis(2-(diph en ylph osph in o)eth oxy)-
n a p h th a len e)2Rh 2(CO)6][BF 4]2 (10). A CD2Cl2 solution of 8
(10 mg in 0.5 mL) was charged with CO gas for 3 h at room
temperature. The resulting solution was agitated overnight
with a mechanical shaker. The solution changed from deep-
red to a pale-yellow color with a white precipitate. 31P{1H}
NMR spectroscopic data of the supernatant solution and the
redissolved precipitate show the formation of tricarbonylrhod-
1
ium(I) complex 10 in a quantitative yield. H NMR (CD2Cl2):
δ 3.22 (m(b), 8H, CH2PPh2), 4.50 (m(b), 8H, CH2O), 6.54 (s(b),
4H, C10H6), 7.26 (m, 4H, C10H6), 7.51-7.71 (m, 44H, P(C6H5)2
+ C10H6). 31P{1H} NMR (CD2Cl2): δ 26.81 (d, J Rh-P ) 90). MS
-
+
(ESI, m/z): [M - 6CO - BF4
]
1461.6 (calcd for C76H68-
+
- 2+
BF4O4P4Rh2 1461.9), [M - 6CO - 2BF4
]
687.4 (calcd for
C
76H68O4P4Rh22+ 687.5). FTIR (CD2Cl2): νCO 2013 (s), 2093 (s)
cm-1
.
Syn th esis of [(µ2-1,4-Bis(2-(diph en ylph osph in o)eth oxy)-
n a p h th a len e)2(CO)2Rh 2 (CD3CN)2][BF 4]2 (11). The precipi-
tate of 10 was collected by filtration of the CD2Cl2 suspension
(vide supra) through Celite. The solid was then redissolved
with CD3CN and evolution of CO gas was observed. NMR
spectroscopy shows the quantitative formation of the trans-
1
CO-Rh(I)-NCCD3 complex, 11. H NMR (CD3CN): δ 3.21 (m,
8H, CH2PPh2), 4.44 (m, 8H, CH2O), 6.42 (s, 4H, C10H6), 7.33
(m, 4H, C10H6), 7.50 (m, 24H, P(C6H5)2), 7.64 (m, 4H, C10H6),
7.71 (m, 16H, P(C6H5)2). 31P{1H} NMR (CD3CN): δ 27.52 (d,
-
+
J Rh-P ) 118). MS (ESI, m/z): [M - BF4
] 1605.7 (calcd for
+
- 2+
C
82H68D6BF4N2O6P4Rh2 1606.0), [M - 2BF4
]
759.6 (calcd
2+
for C82H68D6N2O6P4Rh2 759.5), [M - 2CO - 2CD3CN -
2BF4
]
2+ 687.8 (calcd C80H68D6N2O4P4Rh22+ 687.5). FTIR (CD3-
-
CN): νCO 1943 (s) cm-1
.
Syn th esis of [(η1:η4:η1-1,4-Bis(2-(d ip h en ylp h osp h in o)-
eth oxy)n a p h th a len e)Rh ][BF 4] (12). Complex 8 (10 mg, 6.46

5724 Organometallics, Vol. 21, No. 26, 2002
Ta ble 3. Cr ysta llogr a p h ic Da ta for 8, 9 (cis-, a n ti-), 11, a n d 12
Holliday et al.
entry
8
9 (cis-, anti-)
11
12
formula
fw
C
77.25H68B2F8O4.25
P4Rh2Cl1.5
-
C
84H80B2F8N4O4P4Rh2‚
2(CH2Cl2)‚2(H2O)
C
82H74B2F8N2O6 P4Rh2‚
C
38H34BF4O2P2Rh
2(CH3CN)
1620.81
1918.72
1768.86
774.31
color, habit
red, block
yellow, block
0.10 × 0.10 × 0.10
P1h
yellow, block
0.20 × 0.18 × 0.12
P21/c
monoclinic
11.3786(6)
19.4572(10)
18.9976(10)
90
99.8450(10)
90
4144.0(4)
2
red, block
0.30 × 0.30 × 0.20
P21
monoclinic
10.3886(5)
13.9509(6)
11.8759(5)
90
103.9520(10)
90
1670.40(13)
2
cryst dimens (mm3)
0.45 × 0.15 × 0.07
space group
cryst syst
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Cc
monoclinic
22.986(3)
30.121(4)
16.062(2)
90
129.574(2)
90
8572(2)
4
triclinic
10.693(2)
12.056(2)
18.979(3)
84.261(3)
89.892(3)
83.049(3)
2416.3(8)
1
Z
D
calcd, g cm-3
1.256
1.319
1.418
1.539
radiation (λ, D)
F(000)
Mo KR (0.71073)
3292
Mo KR (0.71073)
980
Mo KR (0.71073)
1808
Mo KR (0.71073)
788
2θ range, deg
µ, cm-1
range of trans factors
T, K
2.66-56.54
5.65 (Mo KR)
0.9058-1.00
153
2.16-57.6
5.80 (Mo KR)
0.63-1.00
150(2)
3.02-56.58
5.48 (Mo KR)
0.8983-0.9371
183(2)
4.68-56.48
6.64 (Mo KR)
0.8257-0.8787
173(2)
scan method
abs cor
ω
ω
ω
ω
empirical from SADABS
no. of measd reflns
no. of indep reflns
refinement method
GOF on F2
39692
25467
14897
10565
21022
8984
7128
5505
full-matrix least-squares on F2
1.09
1.150
1.021
1.031
final R indicesa (I > 2σ(I))
R1 ) 0.0483,
wR2 ) 0.1310
R1 ) 0.0558,
wR2 ) 0.1386
R1 ) 0.0786,
wR2 ) 0.1911
R1 ) 0.0947,
wR2 ) 0.2014
R1 ) 0.0454,
wR2 ) 0.1106
R1 ) 0.0735,
wR2 ) 0.1248
R1 ) 0.0201,
wR2 ) 0.0534
R1 ) 0.0208,
wR2 ) 0.0536
R indices (all data)
R1 ) ∑||Fo| - |Fc||/∑|Fo|, wR2 ) [∑(w(Fo - Fc2)2)/∑w(Fo2)2]1/2, w ) 1/σ2(F2).
a
2
× 10-6 mol) was dissolved in 0.5 mL of CD2Cl2 to which ligand
7 was added (2 mg, 3.23 × 10-6 mol). The sealed NMR tube
was warmed to 50 °C for 48 h, which quantitatively gave
monomeric Rh(I) complex and free ligand without any decom-
position as determined by NMR spectroscopy. Repeated pre-
cipitation (3 times) of complex 12 from CH2Cl2 with diethyl
ether afforded analytically pure monomeric complex (9 mg,
90%). Dec pt 270 °C.1H NMR (CD2Cl2): δ 1.98 (m, 2H, CH2-
PPh2), 2.08 (m, 2H, CH2PPh2), 4.30 (m, 2H, CH2O), 4.47 (m,
2H, CH2O), 6.64 (m, 4H, P(C6H5)2), 6.96 (s, 2H, C10H6), 7.00
(m, 4H, P(C6H5)2), 7.15 (m, 10H, P(C6H5)2), 7.25 (m, 2H,
P(C6H5)2), 7.73 (m, 2H, C10H6), 7.93 (m, 2H, C10H6). 13C{1H}
NMR (125.641 MHz, CD2Cl2): δ 24.88 (m, CH2PPh2), 69.03
(s, CH2O), 87.77 (d, J C-Rh ) 3.4, C2/3), 116.62 (d, J C-Rh ) 3.3,
Kin etic NMR Exp er im en ts. Monophosphine (0.2 equiv,
43 µL, 2 × 10-2 M in CD2Cl2) or 0.1 equiv of diphosphine (43
µL, 1 × 10-2 M in CD2Cl2) was added to a solution of 8 (7 mg,
4.5 × 10-6 mol) in 0.5 mL of 1 × 10-3 M decane in CD2Cl2 in
an air-free NMR tube. The mixture was placed in an NMR
spectrometer, and the temperature was maintained at 40 °C
over the duration of the experiment. The 1H NMR spectra were
recorded periodically (every 5-20 min; depending on the
temperature of experiment). Comparison of the relative inte-
grations of decane to 8 provided conversion data of 8 to 12.
The data were fit to rate laws for zero-, first-, and second-
order kinetics. The first-order plots (ln[8] vs time) showed a
linear relationship, and the kobsd were determined from the
slope of this line by the linear least-squares method. The
correlation coefficient of the least-squares line (R2 > 0.997)
was very good. Furthermore, to ensure statistical accuracy,
the variation of macrocycle concentration vs time was fit
directly to the nonlinear rate law using nonlinear least-squares
analysis. The variable-temperature kinetics were performed
in an identical manner at 30, 35, 45, and 50 °C. All reactions
proceed to completion to give 12 and free nucleophile as
C
1/4), 120.84 (s, C5/8), 122.09 (s, C9/10), 128.72 (m, Cm-P), 128.98
(m, Cm-P), 130.14 (s, C6/7), 130.35 (m, Ci-P), 130.74 (m, Ci-
P), 131.06 (s, Cp-P), 131.33 (s, Cp-P), 133.38 (m, Co-P), 134.03
(m, Co-P). 31P{1H} NMR (CD2Cl2): δ 27.29 (d, J Rh-P ) 202).
-
+
MS (ESI, m/z): [M - BF4
]
687.3 (calcd for C38H34O2P2Rh+
687.5). Anal. Calcd for C38H34BF4O2P2Rh: C, 58.91; H, 4.43;
P, 8.00. Found: C, 58.85; H, 4.29; P, 7.75. E1/2 (CH2Cl2, 0.1 M
[(n-Bu)4N][PF6]): 660 mV (vs Fc/Fc+).
1
confirmed by H and 31P{1H} NMR spectroscopies.
Syn t h esis of [(η1:η1-1,4-Bis(2-(d ip h en ylp h osp h in o)-
eth oxy)n a p h th a len e)(CO)Rh (CH3CN)][BF 4] (13). Three
drops of CH3CN was added to a CD2Cl2 solution of 12 (10 mg
in 0.5 mL), which was then charged with CO (1 atm) for 2 h
at room temperature. The resulting NMR solution was shaken
overnight, which gave a monomeric trans-CO-Rh(I)-NCCD3
complex quantitatively as determined by NMR spectroscopy.
1H NMR (CD2Cl2): δ 2.90 (m, 2H, CH2PPh2), 3.14 (m, 2H, CH2-
PPh2), 4.42 (m, 4H, CH2O), 6.40 (s, 2H, C10H6), 7.18-7.79 (m,
24H, P(C6H5)2 + C10H6). 31P{1H} NMR(CD2Cl2): δ 23.56 (d,
X-r a y Cr ysta llogr a p h y. A suitable X-ray quality crystal
of 8 coated with Paratone was mounted on a Bruker SMART-
1000 diffractometer equipped with a graphite-monochromated
Mo KR (λ ) 0.71073) radiation source and a CCD detector.
Data collection was performed with a detector distance of 5
cm. The raw data collected were processed to produce conven-
tional intensity data by the program SAINT. The intensity
data were corrected for Lorentz and polarization effects.
Absorption correction was also applied using the SADABS
empirical method. The structure was solved by direct methods
provided by the program package SIRS92. Unless otherwise
noted, all the non-hydrogen atoms were refined anisotropically.
Hydrogen atom positions were calculated and included in the
final cycle of refinement. Suitable crystals of 9 (cis-, anti-), 11,
-
+
J Rh-P ) 117). MS (ESI, m/z): [M - BF4
] 756.3 (calcd for
-
+
C
41H37NO3P2Rh 756.6), [M - CH3CN - BF4
]
715.3 (calcd
for C39H34O3P2Rh+ 715.5), [M - CO - CH3CN - BF4
]
+ 687.4
-
(calcd for C38H34O2P2Rh+ 687.5).

Weak-Link Approach to Coordination Complexes
Organometallics, Vol. 21, No. 26, 2002 5725
and 12 were selected and glued to a thin glass fiber using
epoxy. The structures were solved using direct methods for
11 and 12 and the Patterson function for 9 (cis-, anti-). Non-
hydrogen atoms were located by Fourier synthesis and were
refined anisotropically. Hydrogen atoms were added at calcu-
lated positions and treated as isotropic contributions with
thermal parameters defined as 1.2 or 1.5 times that of the
parent atom. The systematic absences in diffraction data of
11 and 12 were consistent with the reported space groups. The
E-statistics suggested a centrosymmetric space group for 9
(cis-, anti-). The solvent molecules in the structure of 9 (cis-,
anti-) are disordered. The program SQUEEZE39 was used to
treat the solvent molecules. Corrections of the X-ray data for
9 (cis-, anti-) by SQUEEZE (113 electron/cell) are close to the
required value (104 electron/cell) for the formula above. The
correct absolute structure was unambiguously determined;
Flack parameter ) 0.14(3) and 0.007(16) respectively for 8 and
12. All software and sources of scattering factors are contained
in the SHELXTL program library (version 5.10, G. Sheldrick,
Bruker-AXS, Madison, WI.). Crystallographic data for 8, 9
(cis-, anti-), 11, and 12 are summarized in Table 3.
Ack n ow led gm en t. C.A.M. acknowledges the NSF
(CHE-0071885/001) and AFOSR (F49620-00-1-0230) for
support of this research. B.J .H. acknowledges Sigma Xi,
the Link Foundation, and NSF/NSEC (EEC-0118025)
for fellowship support. Y.-M.J . acknowledges the Korea
Research Foundation (KRF-99-D031) for postdoctoral
fellowship support. Professors J . B. Lambert and Fred
Basolo are gratefully acknowledged for helpful discus-
sions.
Su p p or tin g In for m a tion Ava ila ble: The COSY and
HMQC spectra of complex 8 and the HMQC spectrum of 12;
images of the reaction described in footnote 7b and the 31P-
{1H} NMR spectrum of the crude reaction mixture. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(39) Van der Sluis, P.; Spek, A. L. Acta Crystallogr. 1990, A46, 194-
201.
OM020739C