Thermodynamics of phosphine coordination to the [PNP]Rh(I) fragment: An example of the importance of reorganization energies in the assessment of metal-ligand 'bond strengths'

Article abstract of DOI:10.1021/ja974200v

Reaction enthalpies of the complexes [RPNP]Rh(COE) ([RPNP] = N(SiMe₂CH₂PPh₂)₂, N(SiMe₂-CH₂P(i)Pr₂)₂; COE = cyclooctene) with a series of phosphine ligands and CO have been measured by solution calorimetry. The measured enthalpies span a range of ca. 40 kcal/mol. These systems favor coordination of strong π-acceptor/weak σ-donor ligands as shown by the trend in ΔH(rxn): CO >> Ppyrl'₃ > Ppyrl₃ > PPhpyrl₂ > PPh₂pyrl > PPh₃. This trend is exactly the opposite of that observed in another square planar rhodium(I) system, trans-RhCl(CO)(PZ₃)₂. With the exception of CO, the ligands investigated are isosteric, and so the observed trends are electronic in nature. Single-crystal X-ray diffraction studies on several of theses complexes ([RPNP]RhL where R, L = Ph, PPh₃; Ph, Ppyrl₃; Ph, CO; (i)Pr, PPh₃; (i)Pr, Ppyrl₃; (i)Pr, CO; (i)Pr, COE) have been performed. Although the structural trends are readily understood in terms of the electronic (donor/acceptor) nature of each ligand array, it is not obvious that the structural data predict the trends or, in particular, the trend reversal in ΔH(rxn) in the two Rh(I) systems. Rather, these results illustrate the importance of reorganization energies in thermodynamic analyses of metal-ligand bonding, especially in the presence of synergistic bonding involving σ-donor, π-donor, and π-acceptor ligands, interacting through shared metal orbitals (electron push-pull). In such cases the interpretation of a metal-ligand bond dissociation enthalpy (D) as an intrinsic, universal, and transferable property of that bond (e.g., a 'bond strength') is an invalid proposition.

Full text of DOI:10.1021/ja974200v

7806
J. Am. Chem. Soc. 1998, 120, 7806-7815
Thermodynamics of Phosphine Coordination to the [PNP]RhI
Fragment: An Example of the Importance of Reorganization Energies
in the Assessment of Metal-Ligand “Bond Strengths”
Jinkun Huang,† Christopher M. Haar,† Steven P. Nolan,*,†,§ William J. Marshall,‡ and
Kenneth G. Moloy*,‡,|
Contribution from the Department of Chemistry, UniVersity of New Orleans,
New Orleans, Louisiana 70148, and Central Research and DeVelopment,
E. I. du Pont de Nemours & Co., Inc., Experimental Station, P.O. Box 80328,
Wilmington, Delaware 19880-0328
ReceiVed December 11, 1997
Abstract: Reaction enthalpies of the complexes [RPNP]Rh(COE) ([RPNP] ) N(SiMe2CH2PPh2)2, N(SiMe2-
CH2PiPr2)2; COE ) cyclooctene) with a series of phosphine ligands and CO have been measured by solution
calorimetry. The measured enthalpies span a range of ca. 40 kcal/mol. These systems favor coordination of
strong π-acceptor/weak σ-donor ligands as shown by the trend in ∆Hrxn: CO . Ppyrl′3 > Ppyrl3 > PPhpyrl2
> PPh2pyrl > PPh3. This trend is exactly the opposite of that observed in another square planar rhodium(I)
system, trans-RhCl(CO)(PZ3)2. With the exception of CO, the ligands investigated are isosteric, and so the
observed trends are electronic in nature. Single-crystal X-ray diffraction studies on several of theses complexes
([RPNP]RhL where R, L ) Ph, PPh3; Ph, Ppyrl3; Ph, CO; iPr, PPh3; iPr, Ppyrl3; iPr, CO; iPr, COE) have been
performed. Although the structural trends are readily understood in terms of the electronic (donor/acceptor)
nature of each ligand array, it is not obvious that the structural data predict the trends or, in particular, the
trend reversal in ∆Hrxn in the two Rh(I) systems. Rather, these results illustrate the importance of reorganization
energies in thermodynamic analyses of metal-ligand bonding, especially in the presence of synergistic bonding
involving σ-donor, π-donor, and π-acceptor ligands, interacting through shared metal orbitals (electron push-
pull). In such cases the interpretation of a metal-ligand bond dissociation enthalpy (D) as an intrinsic, universal,
and transferable property of that bond (e.g., a “bond strength”) is an invalid proposition.
Introduction
fundamental understanding of the mechanisms by which phos-
phines modulate metal reactivity is important for the develop-
ment of new catalysts and reagents. To this end, we have been
active in delineating, measuring, and calibrating the steric and
electronic contributions of phosphine ligands to organometallic
thermochemistry by means of solution calorimetry. We have
studied a number of systems which probed the steric (CpRu-
(PZ3)2Cl, Cp*Ru(PZ3)2Cl) and electronic (Fe(CO)3(PZ3)2, RhCl-
(CO)(PZ3)2, Rh(acac)(CO)(PZ3)) contributions of phosphine
ligands, as shown in eqs 1-4.4-6
The quantitative assessment of metal-ligand thermochemistry
has been of interest for some time.1 This interest is driven by
the need for a better understanding of metal-ligand interactions
and thus reactivity. Significant progress has been made toward
the accumulation of a great deal of information and insight on
this subject. The ultimate goal is to apply the knowledge and
understanding thus gained in a predictive manner to stoichio-
metric and catalytic chemistry.
Phosphines, PZ3,2 are a class of ligands that are prevalent in
stoichiometric and catalytic organometallic chemistry.3 They
generally play a role as ancillary ligands used to fine-tune the
(3) See the following and references therein: (a) McAuliffe, C. A.;
Mackie, A. G. P-Donor Ligands. In Encyclopedia of Inorganic Chemistry;
King, R. B., Ed.; J. Wiley and Sons: New York, 1994; Vol. 6, p 2989. (b)
Wilson, M. R.; Woska, D. C.; Prock, A.; Giering, W. P. Organometallics
1993, 12, 1742. (c) Caffery, M. L.; Brown, T. L. Inorg. Chem. 1991, 30,
3907. (d) Dunne, B. J.; Morris, R. B.; Orpen, A. G. J. Chem. Soc., Dalton
Trans. 1991, 653. (e) Corbridge, D. E. C. Phosphorus. An Outline of its
Chemistry, Biochemistry and Technology, 5th ed.; Elsevier: New York,
1995; Chapter 10. (f) Levason, W. In The Chemistry of Organophosphorus
Compounds; Hartley, F. R., Ed.; Wiley: New York, 1990; Vol. 1, Chapter
15. (g) McCauliffe, C. A. ComprehensiVe Coordination Chemistry; Wilkin-
son, G., Gillard, R. D., McCleverty, J. A., Eds.; Pergamon: Oxford, U.K.,
1987; Vol. 2, p 989. (h) 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; p 66. (i)
Pignolet, L. H., Ed. Homogeneous Catalysis with Metal Phosphine
Complexes; Plenum: New York, 1983. (k) Alyea, E. C.; Meek, D. W. AdV.
Chem. Ser. 1982, 196.
reactivity of the metal center and appended ligands.
A
† University of New Orleans.
‡ E. I. du Pont de Nemours & Co., Inc.
§ E-mail: spncm@uno.edu.
| E-mail: moloykg@esvax.email.dupont.com.
Contribution No. 7647.
(1) (a) Nolan, S. P. Bonding Energetics of Organometallic Compounds.
In Encyclopedia of Inorganic Chemistry; King, R. B., Ed.; J. Wiley and
Sons: New York, 1994; Vol. 1, p 307. (b) Hoff, C. D. Prog. Inorg. Chem.
1992, 40, 503. (c) Marks, T. J., Ed. Bonding Energetics in Organometallic
Compounds. ACS Symp. Ser. 1990, 428. (d) Martinho Simo˜es, J. A.
Beauchamp, J. L. Chem. ReV. 1990, 90, 629.
(2) We use the term “phosphine” here to refer in a very generic sense to
P(III)-donor ligands of the type PZ3, where Z may be carbon or a heteroatom,
especially oxygen and nitrogen (e.g., phosphites and phosphoramides).
S0002-7863(97)04200-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/23/1998

Phosphine Coordination to the [PNP]RhI Fragment
J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7807
(PPh3)2.8 Although steric factors certainly contribute to this
observation, it is likely that the reaction is also favored by the
∆Hrxn
Cp′Ru(COD)Cl + PZ3
8 Cp′Ru(PZ3)2Cl + COD (1)
electronic properties of CO Vis-a`-Vis PPh3. A more conclusive
Cp′ ) C5H5, C5(CH3)5
-
example is provided by the example of the complexes ML4
,
where M ) Rh(1-) or Ir(1-) and L ) PZ3 or CO. These
complexes are stabilized by weak donors/good π-acceptors, even
when sterics are not at issue.9 For example, equilibrium 7 lies
too far to the right to be measured spectroscopically.7e These
observations suggested to us that the electronic ordering
described above should be reversed in cases where π-back-
donation is an important component of the molecular bonding,
such as in the case of electron-rich metal-centered systems.
∆Hrxn
Fe(BDA)(CO)3 + 2PZ3
8 Fe(CO)3(PZ3)2 + BDA (2)
BDA ) PhCHCHCOCH3
∆Hrxn
1/2[RhCl(CO)2]2 + 2PZ3
8
trans-RhCl(CO)(PZ3)2 + CO (3)
3
[Rh(CO)3(PPh3)]- + Ppyrl3 y > 10z
[Rh(CO)3(Ppyrl3)]- + PPh3 (7)
∆Hrxn
Keq
Rh(CO)2(acac) + PZ3
8 Rh(CO)(PZ3)(acac) + CO (4)
In the systems examined to date, and where electronic
contributions dominate, phosphine substitution becomes increas-
ingly favored as the basicity of the phosphine increases. Thus,
∆Heq for the equilibrium in eq 5, where LnM ) RhCl(CO) or
Fe(CO)3, follows the order P(p-CH3OC6H4)3 > PPh3 > P(p-
CF3C6H4)3 > Ppyrl3 > Ppyrl′3.7
In an effort to test our hypothesis we sought to examine the
thermodynamics of phosphine substitution at electron-rich metal
centers or coordination sites. As described in this contribution,
solution reaction calorimetry results show that the reaction
enthalpy scale can indeed be reversed. By combining the
thermochemical results with crystallographic studies, we further
show that these results cannot be understood simply in terms
of differences in intrinsic bond strengths. Rather, they provide
a good example of the necessary consideration that must be
given to reorganization energies when evaluating metal-ligand
bond strengths and the role they play in the relative stabilities
of metal complexes.
LnM(PZ3)2 + 2PZ′3 h LnM(PZ′3)2 + 2PZ3
(5)
This series of ligands has been shown7 to be isosteric, and
thus, the observed ordering is due solely to changes in ligand
electronic (donor/acceptor) properties. The range in stabilities
can be quite significant; our calorimetry data show that the
enthalpy for the equilibrium shown in eq 6 favors products by
25.9 kcal/mol when PZ′3 ) P(p-CH3OC6H4)3 and PZ3
)
Results
Ppyrl′3.6a
Calorimetry. During our search for an appropriate electron-
rich metal center, we examined Wilkinson’s catalyst, RhCl-
(PPh3)3. As noted above, it is known that this complex reacts
rapidly and cleanly with CO to yield trans-RhCl(CO)(PPh3)2.8
Thus, replacement of the PPh3 trans to chlorine with a good
π-acceptor ligand is favorable in this example, exactly the
situation sought. Indeed, PPh3 is displaced by the ligand Ppyrl3
according to eq 8. The reaction proceeds rapidly on mixing at
trans-RhCl(CO)(PZ3)2 + 2PZ′3 h
trans-RhCl(CO)(PZ′3)2 + 2PZ3 (6)
While the enthalpic ordering in the systems studied favors
better donors, clearly this will not be the case with all metal
centers. For example, CO, a very good π-acceptor, readily
displaces PPh3 from RhCl(PPh3)3 to give trans-RhCl(CO)-
(4) For organoruthenium systems, see: (a) Serron, S. A.; Luo, L.; Li,
C.; Cucullu, M. E.; Nolan, S. P. Organometallics 1995, 14, 5290. (b) Serron,
S. A.; Nolan, S. P. Organometallics 1995, 14, 4611. (c) Luo, L.; Li, C.;
Cucullu, M. E.; Nolan, S. P. Organometallics 1995, 14, 1333. (d) Luo, L.;
Li, C.; Cucullu, M. E.; Nolan, S. P.; Fagan, P. J.; Jones, N. L.; Calabrese,
J. C. Organometallics 1995, 14, 289. (e) Luo, L.; Nolan, S. P. Organome-
tallics 1994, 13, 4781. (f) Li, C.; Cucullu, M. E.; McIntyre, R. A.; Stevens,
E. D.; Nolan, S. P. Organometallics 1994, 13, 3621. (g) Luo, L.; Zhu, N.;
Zhu, N.-J.; Stevens, E. D.; Nolan, S. P.; Fagan, P. J. Organometallics 1994,
13, 669. (h) Luo, L.; Fagan, P. J.; Nolan, S. P. Organometallics 1993, 12,
4305. (i) Nolan, S. P.; Martin, K. L.; Stevens, E. D.; Fagan, P. J.
Organometallics 1992, 11, 3947.
(5) For organoiron systems, see: (a) Li, C.; Stevens, E. D.; Nolan, S. P.
Organometallics 1995, 14, 3791. (b) Li, C.; Nolan, S. P. Organometallics
1995, 14, 1327. (c) Luo, L.; Nolan, S. P. Inorg. Chem. 1993, 32, 2410. (d)
Luo, L.; Nolan, S. P. Organometallics 1992, 11, 3483.
(6) For organorhodium systems, see: (a) Serron, S.; Nolan, S. P.; Moloy,
K. G. Organometallics 1996, 15, 4301. (b) Serron, S.; Huang, J.; Nolan, S.
P. Organometallics 1998, 17, 534.
(7) We recently reported the results of several studies on N-pyrrolylphos-
phine ligands (pyrl and pyrl′ are defined below). The cumulative data
demonstrate that these ligands are potent π-acceptors and are also isosteric
with P(p-XC6H4)3; see the following references and also ref 6. (a) Huang,
A.; Marcone, J. E.; Mason, K. L.; Marshall, W. J.; Moloy, K. G.; Serron,
S.; Nolan, S. P. Organometallics 1997, 16, 3377. (b) Serron, S.; Nolan, S.
P. Inorg. Chim. Acta 1996, 252, 107. (c) Li, C.; Serron, S.; Nolan, S. P.;
Petersen, J. L. Organometallics 1996, 15, 4020. (d) Moloy, K. G.; Petersen,
J. L. J. Am. Chem. Soc. 1995, 117, 7696.
room temperature and in quantitative yield as determined by
NMR. The product is easily characterized by its 31P NMR
spectrum, which shows two distinct sets of phosphorus reso-
nances: a doublet of doublets due to the pair of mutually trans
PPh3 ligands (δ 31P ) 34.3, JRh-P ) 133 Hz, JP-P ) 42 Hz)
and a doublet of triplets assignable to the unique P(pyrl)3 ligand
(δ 31P ) 101.1, JRh-P ) 278 Hz, JP-P ) 42 Hz).10
A
calorimetric study of this reaction shows that ∆Hrxn ) -3.2(3)
kcal/mol. We previously showed that Ppyrl3 is isosteric with
(8) (a) Hughes, R. P. ComprehensiVe Organometallic Chemistry; Wilkin-
son, G., Stone, F. G. A., Abel, A., Eds.; Pergamon Press: Oxford, U.K.,
1982; Chapter 35. (b) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic
Chemistry, 5th ed.; Wiley-Interscience: New York, 1988; p 900 ff.
(9) (a) Chan, A. S. C. Inorg. Chim. Acta 1993, 210, 5. (b) Chan, A. S.
C. Shieh, H.-S.; Hill, J. R. J. Organomet. Chem. 1985, 279, 171. (c) Chan,
A. S. C. Carroll, W. E.; Willis, D. E. J. Mol. Catal. 1983, 19, 377. (d)
Bogdanovic, B.; Leitner, W.; Six, C.; Wilczok, U.; Wittmann, K. Angew.
Chem., Int. Ed. Engl. 1997, 36, 502.
(10) A single-crystal X-ray determination of the structure of trans-RhCl-
(PPh3)2(Ppyrl3) confirms the atom connectivity. However, a lack of high
angle data resulted in large thermal parameters and unreasonable bond
lengths, precluding any meaningful conclusions regarding the bonding in
this complex.

7808 J. Am. Chem. Soc., Vol. 120, No. 31, 1998
Huang et al.
PPh3 but is a much better π-acceptor and a weaker donor.7 Thus,
eq 8 demonstrates that, in the absence of steric effects,
replacement of ligands trans to chlorine is favored by weak
donors/strong acceptors, exactly the opposite of the systems
previously studied.
Table 1. Enthalpies of Substitution for Eq 9
a
-∆Hrxn
(kcal/mol)
complex
R
L
øb
1
2
3
4
5
6
7
8
9
Ph
Ph
Ph
Ph
Ph
Ph
iPr
iPr
iPr
iPr
iPr
iPr
PPh3
7.2(3)
11.1(2)
13.0(2)
13.8(1)
14.9(2)
48.9(3)
10.9(3)
13.0(4)
15.4(3)
18.9(2)
20.2(3)
53.3(4)
13.25
21
29
37
48
PPh2pyrl
PPhpyrl2
Ppyrl3
Ppyrl′3
CO
However, extension of this chemistry to other isosteric PZ3
ligands (PPhxpyrl3-x, P(p-XC6H4)3) resulted in complex mixtures
of products, presumably due to competing ligand substitution
and scrambling of all sites on rhodium, i.e., positions cis as
well as trans to chlorine.8a This outcome is not surprising
considering the small reaction enthalpy obtained for eq 8 and
that PPh3 and Ppyrl3 lie at opposite extremes of donor/acceptor
character. As a result we were forced to search for a more
suitable, well-behaved system. The result in eq 8 indicated that
the direction investigated showed promise and thus the pos-
sibility of blocking the troublesome scrambling reactions by use
of a multidentate spectator ligand was considered. An ideal
candidate is provided by Fryzuk’s tridentate bis(phosphino)-
amido ligands RPNP, shown below.11 These ligands form a
number of square-planar complexes of the type [PNP]RhL,
where L is a neutral donor such as PPh3, CO, or olefin. In
these complexes L is forced to coordinate trans to the amido
and cis to the bis(phosphine) groups and thus occupies the
position equivalent to that of Ppyrl3 in eq 8. The amido group
is a better donor (σ and π) than chloro,12 and thus rhodium is
also expected to be, overall, more electron rich in [PNP]RhL
than in RhCl(PPh3)2L. The effect on the thermodynamics
should be further magnified because L is forced into a position
trans to the σ/π -donor (trans influence). These assumptions
appear to be confirmed by a comparison of the IR spectra of
trans-RhCl(CO)(PPh2Me)2 (νCO ) 1974 cm-1 7c,13) and [PhPNP]-
Rh(CO) (νCO ) 1950 cm-1 11a).
PPh3
13.25
21
29
37
48
PPh2pyrl
PPhpyrl2
Ppyrl3
Ppyrl′3
CO
10
11
12
a Enthalpy values are provided with 95% confidence limits (paran-
thesis). b ø is Tolman’s electronic parameter (ref 14).
Reaction calorimetry shows conclusively and convincingly
that better acceptor ligands give the more stable complexes in
the [RPNP]RhL system. As shown in Table 1, progression from
PPh3, a relatiVely good σ-donor/weak π-acceptor (ø ) 13.2514),
through PPhxpy3-x and ultimately to Ppyrl′3, a potent acceptor/
weak donor (ø ) 48),7a produces a substantial increase in the
reaction enthalpy for both sets of [RPNP] complexes. The
enthalpies of reaction involving the [iPrPNP] series are larger
than the enthalpy values for the related [PhPNP] series. This
trend is consistent with the expected increase in electron density
on rhodium upon substitution of the bulky alkyl for phenyl, as
demonstrated by the 18 cm-1 drop in νCO for [iPrPNP]Rh(CO)
(νCO ) 1932 cm-1) versus [PhPNP]Rh(CO) (νCO ) 1950 cm-1).
This situation is exaggerated for L ) CO, where binding of
i
CO is favored by some 42 kcal/mol over PPh3 in the PrPNP
system. In the case of CO, however, reduced steric effects
cannot be excluded from influencing the magnitude of ∆Hrxn
.
Sterics are unlikely to play a role in the enthalpy trend obtained
with the series of phosphorus donors examined, as discussed
in further detail below.
Structural Studies of [RPNP]RhL. Crystallographic studies
of complexes 1, 4, 6, 7, 10, 12, and [iPrPNP]Rh(COE) (13)
were performed in an effort to better understand the calorimetry
results. Structural parameters relevant to this discussion are
provided in Tables 2 and 3. More detailed crystallographic
information may be found in the Supporting Information.
The description of these complexes is straightforward. All
are pseudo-square-planar complexes, quite similar in their gross
features to the isoelectronic complexes [PhPNP]MCl, M ) Ni,
Pd.11b Distortions from ideal square-planar geometry are
observed and best described by comparison of the trans-ligand-
Rh-ligand angles. These angles most closely approach 180°
in the case of the sterically unencumbered CO ligand, where
for complexes 6 and 12, both pairs of angles are >172°. In
the remaining complexes these angles decrease from the ideal
180° by up to 23°, presumably in response to the presence of
the bulkier ligands L. This angle compression is larger for
trans-PRhP than for trans-NRhP in all cases. The chelate
rings in all cases are puckered to varying degrees, as judged by
the Si-N-Si angles and intra-ring P-Rh-N-Si torsion angles,
Substitution of cyclooctene in [RPNP]Rh(COE) was shown
previously to provide an efficient route to complexes of the type
[RPNP]RhL, where L ) PPh3, PMe3, and CO.11a We thus chose
this substitution reaction for calorimetric investigation. NMR
monitoring of the reactions in eq 9 shows that they proceed
quantitatively and at 60 °C are sufficiently rapid (ca. 2 h) for
reaction calorimetry.
(11) (a) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J. Organometallics
1986, 5, 2469. (b) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J.; Secco, A.
S.; Trotter, J. Organometallics 1982, 1, 918.
(12) Poulton, J. T.; Sigalas, M. P.; Folting, K.; Streib, W. E.; Eisenstein,
O.; Caulton, K. G. Inorg. Chem. 1994, 33, 1476.
(13) See also ref 8a, Table 13.
(14) ø is Tolman’s electronic parameter;14a the commonly used value
for PPh3 (ø ) 13.25) is taken from ref 14b. (a) Tolman, C. A. Chem. ReV.
1977, 77, 313. (b) Bartik, T.; Himmler, T.; Schulte, H.-G.; Seevogel, K. J.
Organomet. Chem. 1984, 272, 29.

Phosphine Coordination to the [PNP]RhI Fragment
J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7809
Table 2. Bond Lengths and Angles in the Inner Coordination Sphere of the Complexes [RPNP]RhLa,b
1
4
6c
7
10
12
13
Rh-N, Å
Rh-P, Å
2.151(2)
2.130(2)
2.099(2);
2.117(2)
2.2986(7)
2.3270(6);
2.3243(7);
2.2938(6)
1.797(3);
1.811(3)
84.77(6)
88.40(6);
88.77(6);
87.78(6)
92.77(8)
94.13(8);
93.69(8);
89.78(8)
177.1(1);
177.1(1)
172.60(5);
176.47(7)
127.2(1);
123.5(1)
14.0, 0.9;
26.7, 27.8
2.158(2)
2.128(3)
2.097(2)
2.125(2)
2.2773 (6)
2.2918(7)
2.3261(8)
2.2923(8)
2.3329(16)
2.3387(16)
2.3221(10)
2.3523(10)
2.3074(9)
2.3085(8)
2.3061(6)
2.3590(6)
Rh-L, Å
2.2121(6)
2.1404(7)
2.2226(5)
2.1333(9)
1.802(3)
2.140(2)
2.174(2)
84.68(6)
84.64(6)
PRhN, deg
79.96(6)
85.69(6)
82.71(6)
83.86(6)
82.74(5)
84.58(5)
80.30(9)
85.29(9)
87.17(6)
87.20(6)
PRhL, deg
95.70(2)
101.49(2)
95.00(3)
98.72(3)
98.71(2)
97.52(2)
98.22(3)
98.37(3)
93.1(1)
92.5(1)
NRhL, deg
PRhP, deg
SiNSi, deg
PRhNSi, degd
168.41(6)
157.34(3)
128.5(1)
170.92(9)
166.26(4)
123.2(1)
165.21(5)
159.92(2)
122.7(1)
167.00(8)
161.61(3)
124.2(2)
179.2(3)
174.17(6)
124.5(1)
31.3, 25.2
168.76(4)
119.3(1)
27.4, 49.7
26.7, 40.5
17.47, 42.70
26.7, 47.5
20.4, 31.8
a Complete structural details are provided as Supporting Information. b Parameters referring to P involve phosphorus of the PNP chelate only;
those referring to L refer to the ligand trans to nitrogen (PPh3, Ppyrl3, or CO). c Two molecules per asymmetric unit. d Intra-ring torsion angle.
Progression from [PhPNP]RhL to the analogous [iPrPNP]-
RhL complex does not result in significant changes in the Rh-L
distance, although Rh-PPh3 increases slightly, Rh-Ppyrl3
decreases, and Rh-CO is unchanged. The observations indicate
that sterics are not significantly different between the two
chelates. Within each pair of PPh3 and Ppyrl3 complexes,
progression from [PhPNP] to [iPrPNP] causes the P-Z bond
length to increase. This is accompanied by a slight folding of
the Z groups away from rhodium (the sum of ZPZ angles
decreases, and the average RhPZ angle increases). The bending
in the PZ3 fragment is consistent with a greater degree of
π-back-bonding in the [iPrPNP]RhL complexes due to increased
electron donation on replacement of Ph with iPr.1g Further note
that, for each pair of complexes [RPNP]RhPPh3 and [RPNP]-
RhPpyrl3 (i.e., 1 vs 4, and 7 vs 10), the value ∑ ZPZ (Table
3) is always smaller for Ppyrl3. This trend has been observed
in related systems7a,d and is further evidence for a greater degree
of π-back-bonding in the Ppyrl3 complexes.
Table 3. Important Structural Parameters in the Rh-PZ3
Fragment of the Complexes [RPNP]Rh-PZ3
1
4
7
10
av P-Z, Å
ZPZ, deg
av RhPZ, deg
1.845
305.1
116.4
1.724
296.9
118.6
115
1.849
300.7
117.6
110
1.732
294.1
119.3
115
∑
2/3(∑ PRhH),b deg 117
Rh-P, Å
2.2121(6) 2.1404(7) 2.2226(5) 2.1333(9)
a Complete structural details are provided as Supporting Information.
b See the text for a description of this parameter
and there are no clear trends with respect to the ligand L. A
possible exception is for L ) PZ3, where in each case, one
“normal” torsion angle (15-30°) is accompanied by one large
torsion angle (>40°), indicating that one ring is twisted more
severely than the other. Other than the unsurprisingly longer
Rh-P(chelate) bonds in the [iPrPNP] series (ca. 2.33 vs 2.30
Å for [PhPNP]), there are no gross differences between the
phenyl- and isopropyl-substituted chelates.
Consistent with previous work, both by us7 and by others,15
the steric sizes of the PPh3 and Ppyrl3 ligands are found to be
indistinguishable. This conclusion is drawn by a comparison
Trends are apparent in the metal-ligand bond lengths. In
all cases Rh-Ppyrl3 is shorter than Rh-PPh3 by ca. 0.08 Å.
This is indicative of greater π-acceptor character in the case of
Ppyrl3, as we previously demonstrated.7 This shortening may
alternatively be attributed to differences in phosphorus hybrid-
ization resulting from the greater electronegativity of the
N-pyrrolyl substituents relative to phenyl. However, note that
the decreased Rh-P bond length is also accompanied by a
shortening of Rh-N by 0.02-0.03 Å. A reasonable explanation
for these observations is that the degree of π-donation from
nitrogen to rhodium increases in the presence of a superior trans
π-acceptor (electron push-pull). In further support of this
explanation, note that Rh-N shortens by another 0.02-0.03 Å
on replacement of Ppyrl3 with CO, a more potent π-acceptor.
Variations in the Rh-P(chelate) distances from complex to
complex are slightly less than the internal variations within each
complex (these distances are crystallographically distinct in all
cases). Thus, no meaningful conclusions may be drawn
regarding this parameter, other than it appears to be rather
insensitive to the variable ligand L.
2
of /3(∑ PRhH), which measures the sum of angles centered
at rhodium and formed by rhodium, phosphorus, and the
hydrogen of each substituent Z forming the closest rhodium
contact. The parameter thus defined is similar to the Tolman
cone angle14 but is measured directly and more easily from the
X-ray data.16 In all cases the closest Rh-H contact occurs with
a hydrogen â to phosphorus. Values for this parameter are
provided in Table 3 and show that PPh3 and Ppyrl3 possess
equivalent cone angles. By inference we conclude that the series
of PPhxpyrl3-x ligands described in this work form an isosteric
series.
Discussion
As shown in Table 1, ∆Hrxn for the reaction in eq 9 increases
with increasing π-acceptor character of the incoming ligand L.
(15) Trzeciak, A. M.; Glowiak, T.; Grzybek, R.; Ziolkowski, J. J. J.
Chem. Soc., Dalton Trans. 1997, 1831.

7810 J. Am. Chem. Soc., Vol. 120, No. 31, 1998
Huang et al.
The correlation of ∆Hrxn with ø is most consistent with, and
suggestive of, an electronic explanation for the trend. The
evidence indicates that substitution is increasingly favored with
increasing π-acceptor character of the incoming ligand. The
π-acceptor character of PZ3 ligands in turn increases with
increasing ø (and the σ-donor character likewise generally
decreases). The structural data support this idea; for each pair
of complexes 1-4 and 7-10, Rh-Ppyrl3 is shorter than Rh-
PPh3 by ca. 0.08 Å (see Figures 2 and 3). Increased π-back-
bonding in the case of Ppyrl3 is also reflected by structural
differences in the PZ3 fragment (P-Z distance, ∑ ZPZ) as
discussed above. Similar trends have been previously observed7
and, along with spectroscopic data (IR), are attributed to
increased metal-to-ligand back-bonding. The 31P NMR data are
also informative in that it is seen that JRh-PZ steadily increases
3
with increasing ligand ø. In fact, Figure 4 shows that JRh-P
correlates well with ∆Hrxn. Metal-phosphorus coupling con-
stants depend on a number of factors, some related, including
the degree of σ-bonding (and possibly π-bonding), the elec-
tronegativity of the substituents at phosphorus, the metal-P
distance, the s character of the bond, and the polarizability of
phosphorus.18 While the correlation in Figure 4 cannot be
assigned to any single contributing factor, we interpret this
correlation as evidence for an electronic role in the enthalpy
ordering.
The enthalpy trend presented here is best explained in terms
of an electronic preference for placement of a good π-acceptor
trans to nitrogen. Nitrogen is a good σ/π-donor and is expected
to render rhodium relatively electron rich for a 16 e- metal
center.12 As the electron density on rhodium is increased it is
not unexpected that coordination of weak donors and/or good
π-acceptors, particularly those trans to nitrogen, will be favored.
This is corroborated by noting the shortening of the Rh-N bond
length as one progresses from PPh3 to Ppyrl3 to CO, indicative
of the synergistic bonding between nitrogen and the ligand trans
to nitrogen.
Having thus rationalized the enthalpy trends for phosphine
substitution on the [RPNP]Rh(I) fragment, we now turn attention
to the dichotomy presented by the examples of [RPNP]Rh(PZ3)
and trans-RhCl(CO)(PZ3)2. Both complexes are square-planar
Rh(I) and possess similar ligand arrays. However, the thermo-
dynamic trends for phosphine coordination in these systems are
exactly the opposite (eqs 10 and 11). This contrast is clearly
depicted in Figure 5, where Tolman’s ø is plotted versus the
measured substitution enthalpies for eqs 3 and 9.19 The
substitution enthalpy data allow us to write the equilibrium
preferences for phosphine coordination in these systems as
shown in Scheme 1.
The major difference in the two examples presented in
Scheme 1 is the site at which the substitution occurs. In the
case of [RPNP]Rh(PZ3), substitution occurs trans to a good σ/π-
donor (N), as discussed above. For RhCl(CO)(PZ3)2 the
substitution occurs cis to a good σ-donor and modest π-donor
(Cl). In addition, RhCl(CO)(PZ3)2 already contains a potent
π-acceptor ligand, CO. Clearly, the resulting electronic changes
and requirements in each system are significantly different. For
the example of RhCl(CO)(PZ3)2 our explanation of the enthalpic
ordering follows that described previously.6a,7e Rhodium in
these complexes is formally 16 e- and thus electron deficient.
Figure 1. Reaction enthalpy vs Tolman’s electronic parameter (ø) for
eq 9. Squares ) [PhPNP]RhL, diamonds ) [iPrPNP]RhL.
These data are plotted versus Tolman’s electronic parameter ø
in Figure 1, which shows a fairly steady increase in ∆Hrxn with
increasing ø for both the [PhPNP] and [iPrPNP] systems
studied.17
We ascribe the trends depicted in Figure 1 to an electronic
preference for π-acceptor ligands and not to steric effects. It
would be expected that sterics would contribute significantly
to the observed ∆Hrxn because the coordination site at which
substitution occurs is rather congested, due to the presence of
the pair of bulky cis-R2P groups. This likely explains in part
the large ∆Hrxn values obtained with CO. However, we, and
others, have shown that the ligands PPhxpyrl3-x are isosteric,7,15
differing substantially only in their electronic properties. Ppyrl3
thus possesses the same cone angle as PPh3 but is a much more
potent π-acceptor. This argument is supported by a comparison
of the crystallographically determined structures of complexes
1, 4, 7, and 10, where the steric sizes of Ppyrl3 and PPh3 are
indistinguishable (Table 3). Furthermore, the calorimetric data
for Ppyrl′3 also correlate with Tolman’s ø. Evidence has been
presented7a which indicates that Ppyrl′3 is essentially isosteric
with PPhxpyrl3-x. This is a result of both the remoteness of
the -CO2Et groups from the metal center and the way in which
they are directed away from the metal by the five-membered
ring. Most importantly, it is unlikely that Ppyrl′3 could be
smaller than PPhxpyrl3-x, which would be required if steric
changes (toward smaller cone angles) were responsible for the
enthalpy trend found in the [RPNP] systems presented herein.
(16) There are two major differences between our method and that used
to measure the Tolman cone angle θ. First, θ is the angle defined by
phosphorus, the metal, and the line defined by the metal and the tangent to
van der Waals radius of the hydrogen atom closest to the metal.14 Our
measure is from the center of the hydrogen atom and thus does not
incorporate the van der Waals radius. Second, θ is the maximum cone angle
obtained upon rotating the Z groups about the P-Z bond, whereas our
measure is derived from the structure as observed and without such
manipulation. The estimation of phosphine sterics directly from X-ray data,
but incorporating the hydrogen van der Waals radius correction, has been
previously reported.16a-c We suggest only that our method is useful for
comparing relative steric sizes and that Ppyrl3 possesses the same, well-
established Tolman cone angle as PPh3 (θ ) 145°). (a) DeSanto, J. T.;
Mosbo, J. A.; Storhoff, B. N.; Bock, P. L.; Bloss, R. E. Inorg. Chem. 1980,
19, 3086. (b) Alyea, E. C.; Dias, S. A.; Ferguson, G.; Restino, R. J. Inorg.
Chem. 1977, 16, 2329. (c) Immirzi, A.; Musco, A. Inorg. Chim. Acta 1977,
25, L41.
(17) The curvature in the data plotted in Figure 1 suggests that more
than one electronic parameter may be required to fully explain the observed
trends, particularly in the [PhPNP] series. However, we consider the
correlation with ø to be sufficiently good for the purposes of this discussion.
For a recent and thorough discussion of phosphine ligand parametrization,
see the following and references therein: Fernandez, A.; Reyes, C.; Wilson,
M. R.; Woska, D. C.; Prock, A.; Giering, W. P. Organometallics 1997, 16,
342.
(18) Alyea, E. C.; Song, S. Inorg. Chem. 1995, 34, 3864.
(19) The enthalpy data reported herein are given on the basis of 1 mol
of the product trans-RhCl(CO)(PZ3)2 and therefore correspond to eq 3 as
written. They are one-half the values given in our initial report on this
system,6a which was based on 1 mol of the dimeric starting material, [RhCl-
(CO)2]2.

Phosphine Coordination to the [PNP]RhI Fragment
J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7811
Figure 3. ORTEP drawings of the complexes (a) [iPrPNP]RhPPh3 (7),
(b) [iPrPNP]RhPpyrl3 (10), and (c) [iPrPNP]Rh(CO) (12); thermal
ellipsoids are drawn at the 50% probability level.
Figure 2. ORTEP drawings of the complexes (a) [PhPNP]RhPPh3 (1),
(b) [PhPNP]RhPpyrl3 (4), and (c) [PhPNP]Rh(CO) (6); thermal
ellipsoids are drawn at the 50% probability level.
This is at least partially compensated for by σ/π-donation from
chlorine as determined by the shortening of Rh-Cl as PZ3
becomes less donating (ø increases). Coordination of a weak
donor/strong π-acceptor ligand (CO) trans to chlorine is

7812 J. Am. Chem. Soc., Vol. 120, No. 31, 1998
Huang et al.
Scheme 2
Figure 4. Reaction enthalpy vs JRh-P for eq 9. Circles ) [iPrPNP]-
RhL, diamonds ) [PhPNP]RhL; individual points are labeled according
to eq 9.
of CO and formation of stronger σ-donor bonds (µ-Cl) via
dimerization to [RhCl(CO)2]2.20
Comparing the two sets of complexes it is concluded that
the balance of σ-donor, π-donor, and π-acceptor ligands
ultimately determines the position of a ligand exchange equi-
librium such as eq 5. This is a direct consequence of the
synergistic bonding between these donor/acceptor ligands and
shared metal orbitals.21
The dichotomy presented by these two examples highlights
a fundamental requirement when analyzing thermochemical
information, whether derived from equilibrium measurements,
calorimetry or other physicochemical measurements, or kinetics
(rates of forward and reverse reactions). Very often this
information is used to derive conclusions, qualitative and/or
quantitative, regarding the relative and absolute strengths of the
bonds being broken and formed. However, before meaningful
conclusions regarding bond strengths can be drawn from such
thermochemical data, consideration must be given to reorganiza-
tion energies. The contributions of reorganization energies to
the overall thermodynamic picture for the present system is
shown in Scheme 2. Following the convention described by
Martinho Simo˜es and Beauchamp,1d it is seen that the bond
dissociation enthalpy, D(Rh-PZ3), is composed of contributions
from both an “intrinsic” bond enthalpy term E(Rh-PZ3) (i.e.,
the “bond strength”) and the reorganization energies ER(Rh) and
ER(PZ3). The position of an equilibrium such as that shown in
eq 12, and thus the difference in bond dissociation enthalpies
(D), is determined by the combination of six factors: two bond
enthalpy terms (E) and four reorganization energy terms (ER)
(entropic contributions aside). As the reorganization energy
terms become large the interpretation of D as an intrinsic
property of a particular bond becomes increasingly meaningless.
From the structural data accumulated on the complexes
[RPNP]Rh(PZ3) and trans-RhCl(CO)(PZ3)2, it is easy to see why
the reorganization energies are substantial. In both cases
progression from phosphines with low ø values to those with
high ø values is accompanied by extensive changes in bond
lengths and angles in the rhodium and PZ3 fragments. More-
over, the similar structural trends for the complexes [RPNP]-
Figure 5. Reaction enthalpy vs Tolman’s electronic parameter (ø) for
eq 9 ([RPNP] ) [iPrPNP]) and eq 3. Circles ) [iPrPNP]RhL (eq 9);
diamonds ) trans-RhCl(CO)L2 (eq 3). Letters correspond to the ligands
L ) PPh3 (a), PPh2pyrl (b), PPhpyrl2 (c), Ppyrl3 (d), and Ppyrl′3 (e).
Scheme 1
stabilizing due to efficient, synergistic π-donation/π-back-
bonding, which is the same argument used to explain the
enthalpy trends in the [RPNP]Rh systems. Addition of further
weak σ-donors/strong π-acceptors PZ3 cis to CO and Cl is
unfavorable due to competition between PZ3 and CO as π-acids.
This conclusion is based on the observation that the Rh-CO
bond length increases in trans-RhCl(CO)(PZ3)2 as ø(PZ3)
increases (e.g., PZ3 becomes increasingly more “CO-like”). This
argument also explains why RhCl(CO)3 is only observed under
high CO pressure, being more stable with respect to elimination
(20) Morris, D. E.; Tinker, H. B. J. Organomet. Chem. 1973, 49, C53.
(21) For a related argument relating to exchange of anionic σ/π-donor
ligands, see: Poulton, J. T.; Hauger, B. E.; Kuhlman, R. L.; Caulton, K. G.
Inorg. Chem. 1994, 33, 3325.

Phosphine Coordination to the [PNP]RhI Fragment
J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7813
calorimetric conditions. These conditions are necessary for accurate
and meaningful calorimetric results and were satisfied for all reactions
investigated.
Solution Calorimetry. In a representative experimental trial, the
mixing vessels of the Setaram C-80 were cleaned, dried in an oven
maintained at 120 °C, and then taken into the glovebox. A sample of
[iPrPNP]Rh(COE) (21.1 mg, 34.8 µmol) was weighed into the lower
Rh(PZ3) and trans-RhCl(CO)(PZ3)2 offer no clue to the reversed
enthalpic ordering between the two families. In each case, as
ø increases, the Rh-PZ3 and Rh-X bond lengths decrease, the
PZ3 group adopts more of an inverted umbrella shape, and the
P-Z bond lengths increase. Bond lengths to other ligands
generally increase, e.g., Rh-CO in trans-RhCl(CO)(PZ3)2.
Breaking the Rh-PZ3 bond will thus be expected to result in
significant changes in the bonding between rhodium and the
remaining ligands in the fragment LnRh (ER(Rh)) and also in
the liberated ligand PZ3 (ER(PZ3). The structural parameters,
and especially the Rh-PZ3 bond distance, are of no use in
predicting the reversed enthalpic ordering between these two
families, indicating that the use of structural data to infer
conclusions regarding thermodynamics can be risky and mis-
leading.
An extensive amount of effort is required, and a number of
assumptions and simplifications are often made, to account for
the ER terms. In many cases they are assumed to be insignificant
and/or are ignored. These contributions are significant when
the bonds being broken and formed are involved in synergistic
bonding. Our results, showing that trends in thermodynamic
data can be completely reversed in very related systems
emphasizes the importance of reorganization energies and the
consideration they must be given. These results also underscore
the caution that must be employed when attempting to interpret
experimental thermochemical results and bond dissociation
enthalpies as intrinsic, transferable properties of individual
bonds. In cases involving synergistic bonding between σ-do-
nors, π-donors, and π-acceptors (electron push-pull), even the
concept of an individual bond strength becomes increasingly
ambiguous.22
vessel, which was closed and sealed with 1.5 mL of mercury.
A
solution of PPh3 (11.1 mg, 42.3 µmol) in toluene (4 mL) was added,
and the remainder of the cell was assembled, removed from the
glovebox, and inserted into the calorimeter. The reference vessel was
loaded in an identical fashion with the exception that no organorhodium
complex was added to the lower vessel. After the calorimeter had
reached thermal equilibrium at 60.0 °C (ca. 2 h), it was inverted, thereby
allowing the reactants to mix. The reaction was considered complete
after the calorimeter had once again reached thermal equilibrium (ca.
2 h). Control reactions with Hg and phosphine show no reaction. The
enthalpy of ligand substitution (-10.9 ( 0.3 kcal/mol) listed in Table
1 represents the average of at least three individual calorimetric
determinations with all species in solution. The enthalpy of solution
of [iPrPNP]Rh(COE) (+9.6 ( 0.2 kcal/mol) in neat toluene was
determined using identical methodology. Other examples were per-
formed in an identical fashion with the exception of [PhPNP]Rh(COE)
+ PPh3, which was conducted at 80.0 °C.
[PhPNP]Rh(PPh2pyrl) (2). In the glovebox, a 25-mL flask fitted
with a frit and a magnetic stir bar was charged with [PhPNP]Rh(COE)
(100.0 mg, 0.135 mmol) and PPh2pyrl (33.9 mg, 0.135 mmol). Toluene
(3-5 mL) was added; the vessel was sealed, removed from the
glovebox, and heated with stirring in a 60 °C oil bath for 1-2 h. The
reaction vessel was then interfaced to a high-vacuum line, and the
toluene was removed in vacuo. The oily residue was triturated several
times with pentane (3-5 mL) to ensure complete removal of cy-
clooctene. The resulting orange powder was then taken up in pentane
(ca. 5 mL) and filtered and the solvent reduced to ca. 1 mL. Slow
cooling of this solution afforded 2 as orange crystals, which were
collected on a frit and dried under flowing argon. Yield: 83 mg (70%).
1H NMR (toluene-d8): δ 0.03(s, 12 H, Si(CH3)2), δ 1.66 (m, 4 H, PCH2-
Si), δ 6.15 (m, 2 H, pyrrolyl), δ 7.52(m, 2 H, pyrrolyl), δ 6.45 (m, 4H,
phenyl), δ 6.48 (m, 2 H, phenyl), δ 7.01 (m, 4 H, phenyl), δ 6.80(m,
12 H, PNP-phenyl), δ 7.42(m, 8 H, PNP-phenyl). 31P{1H} NMR
(toluene-d8): δ 32.5 (dd, JPP ) 43 Hz, JRh ) 140 Hz), δ 89.6 (dt, JRh
) 183 Hz). Anal. Calcd for C46H50N2P3RhSi2: C, 62.58; H, 5.71; N,
3.17. Found: C, 62.71; H, 6.03; N, 2.94.
[PhPNP]Rh(PPhpyrl2) (3). This complex was prepared in a manner
similar to that for 2, using [PhPNP]Rh(COE) (80.0 mg, 0.108 mmol)
and PPhpyrl2 (25.9 mg, 0.108 mmol). Yield: 54 mg (57%) as yellow
crystals. 1H NMR (toluene-d8): δ 0.02 (s, 12 H, Si(CH3)2), δ 1.72
(m, 4 H, PCH2Si), δ 6.05 (m, 4 H, pyrrolyl), δ 7.34 (m, 4 H, pyrrolyl),
δ 6.21 (m, 2 H, phenyl), δ 6.42 (m, 1 H, phenyl), δ 7.04 (m, 2 H,
phenyl), δ 6.86 (m, 12 H, PNP-phenyl), δ 7.48 (m, 8 H, PNP-phenyl).
31P{1H} NMR (toluene-d8): δ 32.9 (dd, JPP ) 44 Hz, JRh ) 137 Hz),
δ 105.8 (dt, JRh ) 194 Hz). Anal. Calcd for C44H49N3P3RhSi2: C,
60.61; H, 5.66; N, 4.82. Found: C, 60.23; H, 5.68; N, 4.62.
[PhPNP]Rh(Ppyrl3) (4). This complex was prepared in a manner
similar to that for 2, using [PhPNP]Rh(COE) (80.0 mg, 0.108 mmol)
and Ppyrl3 (24.7 mg, 0.108 mmol). Yield: 78 mg (84%) as yellow
crystals. 1H NMR (toluene-d8): δ 0.08 (s, 12 H, Si(CH3)2), δ 1.78
(m, 4 H, PCH2Si), δ 5.85 (m, 6 H, pyrrolyl), δ 6.72 (m, 6 H, pyrrolyl),
δ 6.94 (m, 12 H, phenyl), δ 7.61 (m, 8 H, phenyl). 31P{1H} NMR
(toluene-d8): δ 33.0 (dd, JPP ) 47 Hz, JRh ) 134 Hz), δ 103.5 (dt, JRh
) 243 Hz). Anal. Calcd for C42H48N4P3RhSi2: C, 58.60; H, 5.62; N,
6.51. Found: C, 58.49; H, 5.62; N, 6.41.
Experimental Section
General Considerations. All manipulations were performed under
inert atmospheres of argon or nitrogen using standard high-vacuum or
Schlenk line techniques or in a glovebox containing less than 1 ppm
oxygen and water. Solvents, including deuterated solvents for NMR
analyses, were dried by standard methods23 and distilled under nitrogen
or vacuum transferred before use. NMR spectra were recorded using
Varian Gemini 300-MHz or Varian Unity 400-MHz spectrometers.
Elemental analyses were performed by Desert Analytics (Tucson, AZ).
Only materials of high purity as indicated by NMR spectroscopy were
used in the calorimetric experiments. Calorimetric measurements were
performed using a Calvet calorimeter (Setaram C-80) which was
periodically calibrated using the TRIS reaction24 or the enthalpy of
solution of KCl in water.25 This calorimeter has been previously
described,26 and typical procedures are described below. Experimental
enthalpy data are reported with 95% confidence limits. The complexes
[iPrPNP]Rh(COE), [PhPNP]Rh(COE), [PhPNP]RhPPh3, and [PhPNP]-
Rh(CO) were synthesized according to literature procedures,11a as were
the ligands PPh2pyrl, PPhpyrl2, Ppyrl3,7d and Ppyrl′3.7a Triphenylphos-
phine (Aldrich) was recrystallized from ethanol prior to use, and carbon
monoxide (Matheson, UHP grade) was used as received.
NMR Titrations. Prior to every set of calorimetric experiments
involving a new ligand, an accurately weighed amount ((0.1 mg) of
the [RPNP]Rh(COE) complex was placed in an NMR tube along with
toluene-d8 and >1.2 equiv of ligand. The sample was heated at 60 °C
in an oil bath for 1 h, after which both 1H and 31P NMR spectra indicated
that the reactions were clean and quantitative under experimental
[PhPNP]Rh(Ppyrl′3) (5). This complex was prepared in a manner
similar to that for 2 using [PhPNP]Rh(COE) (40.0 mg, 0.0539 mmol)
and Ppyrl′3 (35.7 mg, 0.0540 mmol). After trituration, the product was
washed with pentane to obtain pure 5 as a yellow power, yield: 57
mg (82%). 1H NMR (toluene-d8): δ 0.02 (s, 12 H, Si(CH3)2), δ 0.95
(t, 12 H, CO2CH2CH3), δ 1.10 (t, 6 H, CO2CH2CH3), δ 1.82 (m, 4 H,
PCH2Si), δ 3.94 (q, 8 H, CO2CH2CH3), δ 4.15 (q, 4 H, CO2CH2CH3),
δ 7.03(m, 12 H, phenyl), δ 7.58 (m, 8 H, phenyl), δ 7.34 (s, 6 H,
(22) Purcell, K. F.; Kotz, J. C. Inorganic Chemistry; W. B. Saunders:
Philadelphia, PA, 1977; pp 119-123.
(23) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press: New York, 1988.
(24) Ojelund, G.; Wadso¨, I. Acta Chem. Scand. 1968, 22, 1691-1699.
(25) Kilday, M. V. J. Res. Natl. Bur. Stand. (U.S.) 1980, 85, 467-481.
(26) (a) Nolan, S. P.; Lopez de la Vega, R.; Hoff, C. D. Inorg. Chem.
1986, 25, 4446. (b) Nolan, S. P.; Hoff, C. D. J. Organomet. Chem. 1985,
282, 357.
pyrrolyl). 31P{1H} NMR (toluene-d8): δ 31.8 (dd, JPP ) 46 Hz, JRh
)

7814 J. Am. Chem. Soc., Vol. 120, No. 31, 1998
Huang et al.
Table 4. Crystallographic Data for the Complexes [RPNP]RhL
1
4
6
7
10
12
13
formula
fw
color
space group P21/n (no. 14)
a, Å
b, Å
c, Å
C48H51NP3RhSi2 C42H48N4P3RhSi2 C31H36NOP2RhSi2 C36H59NP3RhSi2 C30H56N4P3RhSi2 C19H44NOP2RhSi2 C26H58NP2RhSi2
893.95
red-orange
860.88
yellow
P21/n (no. 14)
13.218(1)
12.697(1)
25.079(1)
659.67
757.88
724.81
yellow
523.60
yellow
Pca21 (no. 29)
14.653(1)
14.850(1)
12.365(1)
605.79
orange
yellow-orange
P1h (no. 2)
18.140(1)
18.570(1)
10.014(1)
99.04(1)
93.27(1)
107.10(1)
4
red-orange
P21/n (no. 14)
11.084(1)
16.615(1)
21.243(1)
P1h (no. 2)
11.876(1)
15.462(1)
11.493(1)
101.11(1)
117.52(1)
89.87(1)
2
P1h (no. 2)
11.026(1)
15.045(1)
10.874(1)
109.76(1)
110.16(1)
89.45(1)
2
11.716(1)
21.023(1)
18.958(1)
R, deg
â, deg
γ, deg
formula units
per cell
Ra
104.86(1)
4
99.26(1)
4
91.75(1)
4
4
0.031
0.036
1.98
0.027
0.029
1.41
0.032
0.038
2.12
0.028
0.033
1.87
0.040
0.051
2.96
0.020
0.026
1.58
0.032
0.042
2.55
a
Rw
GOF
2
a R ) ∑(||Fo| - |Fc||)/∑|Fo|; Rw ) ∑w(|Fo| - |Fc|)2/∑w|Fo| .
125 Hz), δ 97.2 (dt, JRh ) 259 Hz). Anal. Calcd for C60H72N4P3-
RhSi2: C, 55.72; H, 5.61; N, 4.33. Found: C, 55.53; H, 5.47; N, 4.52.
[iPrPNP]RhPPh3 (7). In the glovebox, a storage tube fitted with a
Teflon valve and a magnetic stir bar was charged with [iPrPNP]Rh-
(COE) (13) (100.1 mg, 0.1652 mmol) and PPh3 (43.3 mg, 0.165 mmol).
Toluene (3-5 mL) was added; the vessel was sealed, removed from
the glovebox, and heated with stirring in a 60 °C oil bath for 1-2 h.
The reaction vessel was then interfaced to a high-vacuum line, and the
toluene was removed in vacuo. The residue was triturated several times
with pentane (3-5 mL) to ensure complete removal of cyclooctene.
The resulting yellow-orange powder was then taken up in pentane (ca.
5 mL) and filtered via cannula into another vessel, and the solvent
was reduced to ca. 1 mL. Slow cooling of this solution afforded 3a as
large, transparent, ruby red blocks, which were isolated by cannula
decantation of the mother liquor and drying under flowing argon.
Yield: 63.8 mg (51%). An additional crop of smaller crystals (20.1
mg, 16%) was recovered from the mother liquor. Thorough drying
under high vacuum caused the crystals to become opaque and return
to the yellow-orange color observed in the crude product. 1H NMR
(C6D6): δ 0.52 (s, 12 H, Si(CH3)2), δ 0.90 (m, 4 H, PCH2Si; m 12 H,
CH(CH3)2), δ 1.08 (m 12 H, CH(CH3)2), δ 1.17 (m, 4 H, CH(CH3)2),
δ 7.04 (m, 9 H, phenyl), δ 8.05 (m, 6 H, phenyl). 31P{1H} NMR
(C6D6): δ 40.1 (dd, JPP ) 40 Hz, JRh ) 130 Hz), δ 44.8 (dt, JRh ) 173
Hz). Anal. Calcd: C, 57.05; H, 7.85; N, 1.85 Found: C, 56.77; H,
8.03; N, 1.54.
[iPrPNP]Rh(Ppyrl′3) (11). This complex was prepared in a manner
similar to that for 7 using [iPrPNP]Rh(COE) (85.9 mg, 0.142 mmol)
and Ppyrl′3 (102.1 mg, 0.154 mmol). Yield: 58 mg (35%) as pale
yellow microcrystals. 1H NMR (C6D6): δ 0.36 (s, 12 H, Si(CH3)2), δ
0.8-1.1 (m, 4 H, PCH2Si; m 12 H, CH(CH3)2; m, 12 H, CH(CH3)2; t,
18 H, CO2CH2CH3), δ 1.44 (m, 4 H, CH(CH3)2), δ 4.07 (q, 12 H,
CO2CH2CH3), δ 7.91 (s, 6 H, pyrrolyl). 31P{1H} NMR (C6D6): δ 46.6
(dd, JPP ) 45 Hz, JRh ) 115 Hz), δ 90.6 (dt, JRh ) 274 Hz). Anal.
Calcd: C, 49.82; H, 6.97; N, 4.84. Found: C, 50.05; H, 7.08; N, 4.63.
[iPrPNP]Rh(CO) (12). This complex was prepared and isolated
in a manner similar to that for 7. After removing the storage tube
from the glovebox, the solution of [iPrPNP]Rh(COE) (104.4 mg, 0.172
mmol) was exposed to ca. 1 atm of CO for 30 min with efficient stirring.
The bright yellow-orange of [iPrPNP]Rh(COE) faded to pale yellow
almost immediately upon exposure to the reactant gas. Yield: 70 mg
(78%) as yellow crystals. IR (toluene): νCO 1932 cm-1 (vs). 1H NMR
(toluene-d8): δ 0.28 (s, 12 H, Si(CH3)2), δ 0.66 (m, 4 H, PCH2Si), δ
1.00 (m, 12 H, CH(CH3)2), δ 1.18 (m, 12 H, CH(CH3)2), δ 1.81 (m, 4
H, CH(CH3)2). 31P{1H} NMR (toluene-d8): δ 53.3 (d, JRh ) 122 Hz).
Anal. Calcd for C19H44NOP2RhSi2: C, 43.59; H, 8.47; N, 2.68.
Found: C, 43.91; H, 8.93; N, 2.68.
X-ray Structural Analyses. The structures were determined from
intensity data collected on a Rigaku RU300 R-AXIS image plate area
detector equipped with Mo KR radiation and a low-temperature
apparatus. Data sets were corrected for Lorentz and polarization effects
but not for absorption. All data sets were collected at -100 °C. The
structures were solved by direct methods and refined by full-matrix
least-squares techniques. The refinement and analysis of the structure
was carried out using a package of local programs.27 The atomic
scattering factors were taken from the tabulations of Cromer and Waber;
anomalous dispersion corrections were by Cromer.28 In the least-
squares refinement, the function minimized was ∑w(|Fo| - |Fc|)2, with
the weights, w, assigned as [σ2(I) + 0.0009I2]-1/2. The crystallographic
highlights for each complex are given in Table 4. Selected distances
and angles are given in Tables 2 and 3, as described in the text. Because
in each complex the refinement produced some unreasonable C-H bond
lengths or hydrogen thermal parameters, the hydrogen atoms were
placed in idealized positions close to their previously refined positions.
All of the non-hydrogen atoms were refined with anisotropic thermal
parameters. Complete structural details including data collection and
refinement, atomic coordinates, anisotropic thermal parameters, and
hydrogen atom positions, are available as Supporting Information.
[iPrPNP]Rh(PPh2pyrl) (8). This complex was prepared in a manner
similar to 7 using [iPrPNP]Rh(COE) (79.5 mg, 0.131 mmol) and PPh2-
pyrl (33.0 mg, 0.131 mmol). Yield: 66 mg (67%) as yellow crystals.
1H NMR (C6D6): δ 0.54 (s, 12 H, Si(CH3)2), δ 0.93 (m, 4 H, PCH2Si;
m 12 H, CH(CH3)2), δ 1.11 (m 12 H, CH(CH3)2), δ 1.29 (m, 4 H,
CH(CH3)2), δ 6.42 (s, 2 H, pyrrolyl), δ 6.95 (br s, 6 H, phenyl), δ 7.72
(s, 2 H, pyrrolyl), δ 7.78 (m, 4 H, phenyl). 31P{1H} NMR (C6D6): δ
42.0 (dd, JPP ) 42 Hz, JRh ) 127 Hz), δ 84.7 (dt, JRh ) 193 Hz).
Anal. Calcd: C, 57.05; H, 7.85; N, 1.85 Found: C, 56.77; H, 8.03;
N, 1.54.
[iPrPNP]Rh(PPhpyrl2) (9). This complex was prepared in a manner
similar to that for 7 using [iPrPNP]Rh(COE) (67.2 mg, 0.111 mmol)
and PPhpyrl2 (26.7 mg, 0.111 mmol). Yield: 80 mg (98%) as a foamy
1
yellow solid. Thorough drying afforded a product pure by NMR. H
NMR (C6D6): δ 0.49 (s, 12 H, Si(CH3)2), δ 0.94 (m, 4 H, PCH2Si; m
12 H, CH(CH3)2), δ 1.08 (m 12 H, CH(CH3)2), δ 1.36 (m, 4 H,
CH(CH3)2), δ 6.32 (s, 4 H, pyrrolyl), δ 6.86 (m, 3 H, phenyl), δ 7.17
(m, 2 H, phenyl), δ 7.58 (br s, 4 H, pyrrolyl). 31P{1H} NMR (C6D6):
δ 43.3 (dd, JPP ) 43 Hz, JRh ) 125 Hz), δ 101.1 (dt, JRh ) 217 Hz).
[iPrPNP]Rh(Ppyrl3) (10). This complex was prepared in a manner
similar to that for 7 using [iPrPNP]Rh(COE) (100.2 mg, 0.165 mmol)
and Ppyrl3 (37.9 mg, 0.165 mmol). Yield: 74 mg (62%) as light yellow
crystals. 1H NMR (C6D6): δ 0.43 (s, 12 H, Si(CH3)2), δ 0.91 (m, 4 H,
PCH2Si; m 12 H, CH(CH3)2), δ 1.09 (m 12 H, CH(CH3)2), δ 1.39 (m,
4 H, CH(CH3)2), δ 6.18 (s, 6 H, pyrrolyl), δ 7.18 (br s, 6 H, pyrrolyl).
31P{1H} NMR (C6D6): δ 45.1 (dd, JPP ) 46 Hz, JRh ) 123 Hz), δ
96.7 (dt, JRh ) 256 Hz).
Acknowledgment. The National Science Foundation (CHE-
9631611) and Du Pont (Educational Aid Grant) are gratefully
acknowledged for support of this research. We also thank
Alistair J. McGhie for expert technical assistance.
(27) Calabrese, J. C. Central Research and Development, E. I. Du Pont
de Nemours and Co., Inc., P.O. Box 80228, Wilmington, DE 19880-0228,
1991.
(28) International Tables for X-ray Crystallography; Kynoch Press:
Birmingham, England, 1974; Vol. IV: (a) Table 2.2B; (b) Table 2.3.1.

Phosphine Coordination to the [PNP]RhI Fragment
J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7815
Supporting Information Available: Full crystallographic
details including listings of atomic coordinates, B values,
selected distances and angles, anisotropic thermal parameters,
and ORTEP drawings for complexes 1, 4, 6, 7, 10, 12, and 13
(62 pages, print/PDF). See any current masthead page for
ordering information and Web access instructions.
JA974200V