Phosphinorhodium-catalyzed dehalogenation of chlorinated and fluorinated ethylenes: Distinct mechanisms with triethylsilane and dihydrogen

Article abstract of DOI:10.1021/om900718s

Catalytic dehalogenation of chlorinated and fluorinated ethylenes by (PR₃)₃RhCl complexes is described. The C-Cl and C-F bonds are activated by the catalyst in the presence of triethylsilane (Et ₃SiH) or dihydrogen (H

Full text of DOI:10.1021/om900718s

5982 Organometallics 2009, 28, 5982–5991
DOI: 10.1021/om900718s
Phosphinorhodium-Catalyzed Dehalogenation of Chlorinated and
Fluorinated Ethylenes: Distinct Mechanisms with Triethylsilane and
Dihydrogen
Alicia A. Peterson, Kristen A. Thoreson, and Kristopher McNeill*,†
Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455. †Present address:
Department of Environmental Science, ETH Zurich, Switzerland.
Received August 13, 2009
Catalytic dehalogenation of chlorinated and fluorinated ethylenes by (PR3)3RhCl complexes is
described. The C-Cl and C-F bonds are activated by the catalyst in the presence of triethylsilane
(Et3SiH) or dihydrogen (H2). Spectroscopic studies, in addition to substrate preference, indicate that
rhodium hydride species are important intermediates. Kinetic parameters and product distribution
for dehalogenation reactions were determined using NMR spectroscopy. Evidence for sequential
chlorine removal was obtained, and the rates of dehalogenation were found to increase with
decreasing halogen content. It was also shown that this catalytic system has a preference for sp2-
over sp3-hybridized carbon-halogen bonds. Dechlorination using (PPh3)3RhCl and either H2 or
Et3SiH supports an insertion/β-chloride elimination mechanism; however, the two systems display
distinct differences. On the basis of these differences, the dominant pathway for Et3SiH is proposed to
involve rhodium(I), while the H2 system is proposed to primarily involve rhodium(III). This is
supported with isotopic labeling studies using D2, Et3SiD, and (PPh3)3RhD, which yield different
stereochemistry of dechlorinated products. With D2, only products consistent with syn-β-chloride
elimination were observed, while with Et3SiD and (PPh3)3RhD both syn- and anti-β-chloride
elimination products were observed. In addition, NMR spectroscopic evidence of different hydride
intermediates in the H2 and Et3SiH systems was obtained. Different pathways for dehalogenation
with Et3SiH and H2 is further supported by the observation of 1,2-addition (hydrogenation) products
using H2 and the lack of 1,2-addition (hydrosilylation) products using Et3SiH.
Introduction
Remediation of these pollutants by dechlorination has
involved the use of zerovalent metals,11-19 electrochemical
processes,20-25 and microbial degradation.5,7,26-28 Several
heterogeneous noble metals have also been shown to degrade
Thousands of tons of persistent chlorocarbons are produced
annually, many of which are toxic and bioaccumulative.1-3
Chlorinated ethylenes, specifically perchloroethylene (PCE)
and trichloroethylene (TCE), are potential carcinogens that
are found in both soil and water systems.1-8 The toxic effects
associated with these complexes are related to the chlorine
substituents. Accordingly, removal of the chlorine atoms leads
to detoxification.9 For instance, vinyl chloride, a known carci-
nogen, yields nontoxic ethylene upon dechlorination.10
(11) Song, H.; Carraway, E. R. Appl. Catal., B 2008, 78, 53–60.
(12) Dries, J.; Bastiaens, L.; Springael, D.; Agathos, S. N.; Diels, L.
Environ. Sci. Technol. 2004, 38, 2879–2884.
(13) Dayan, H.; Abrajano, T.; Sturchio, N. C.; Winsor, L. Org.
Geochem. 1999, 30, 755–763.
(14) Kim, Y. H.; Carraway, E. R. Environ. Technol. 2003, 24, 809–819.
(15) Arnold, W. A.; Roberts, A. L. Environ. Sci. Technol. 1998, 32,
3017–3025.
(16) Arnold, W. A.; Roberts, A. L. Environ. Sci. Technol. 2000, 34,
1794–1805.
(17) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.;
Campbell, T. J. Environ. Sci. Technol. 1996, 30, 2654–2659.
(18) Gillham, R. W.; O’Hannesin, S. F. Ground Water 1994, 32, 958–
967.
(19) O’Hannesin, S. F.; Gillham, R. W. Ground Water 1998, 36, 164–
170.
(20) Shimomura, T.; Sanford, R. A. J. Environ. Qual. 2005, 34, 1435–
1438.
(21) Farrell, J.; Melitas, N.; Kason, M.; Li, T. Environ. Sci. Technol.
*Corresponding author. E-mail: kris.mcneill@env.ethz.ch.
(1) Moran, M. J.; Zogorski, J. S.; Squillace, P. J. Environ. Sci.
Technol. 2007, 41, 74–81.
(2) Doherty, R. E. Environ. Forensics 2000, 1, 83–93.
(3) Doherty, R. E. Environ. Forensics 2000, 1, 69–81.
(4) Glod, G.; Angst, W.; Holliger, C.; Schwarzenbach, R. P. Environ.
Sci. Technol. 1997, 31, 253–260.
(5) Glod, G.; Brodmann, U.; Angst, W.; Holliger, C.; Schwarzenbach,
R. P. Environ. Sci. Technol. 1997, 31, 3154–3160.
(6) Squillace, P. J.; Moran, M. J.; Lapham, W. W.; Price, C. V.;
Clawges, R. M.; Zogorski, J. S. Environ. Sci. Technol. 1999, 33, 4176–4187.
(7) Gantzer, C. J.; Wackett, L. P. Environ. Sci. Technol. 1991, 25, 715–722.
(8) Leisinger, T. B., W. Poorly Degradable Substances; VCH Berlags-
gesellschaft: Weinheim, 1986; Vol. 8, pp 475-513.
2000, 34, 2549–2556.
(22) Lowry, G. V.; Reinhard, M. Environ. Sci. Technol. 2000, 34,
3217–3223.
(23) Lowry, G. V.; Reinhard, M. Environ. Sci. Technol. 2001, 35, 696–
702.
(24) Liu, Z.; Betterton, E. A.; Arnold, R. G. Environ. Sci. Technol.
2000, 34, 804–811.
(25) Lin, X. Q.; Boisselier-Cocolios, B.; Kadish, K. M. Inorg. Chem.
1986, 25, 3242–3248.
(9) Alonso, F.; Beletskaya, I. P.; Yus, M. Chem. Rev. 2002, 102, 4009–
4091.
(10) Clewell, H. J.; Gentry, P. R.; Gearhart, J. M.; Allen, B. C.;
Andersen, M. E. Sci. Total Environ. 2001, 274, 37–66.
r
pubs.acs.org/Organometallics
Published on Web 09/11/2009
2009 American Chemical Society

Article
Organometallics, Vol. 28, No. 20, 2009 5983
chlorinated ethylenes and other chlorocarbons using H2 as
the reductant.22,23,29-35 In these systems the noble metal is
the catalyst and H2 the reducing agent. We propose that
metal hydrides play an active role in these systems, as H2 is
needed for degradation to occur. In order to probe this
mechanistic hypothesis, we have explored the use of homo-
geneous rhodium hydrides as catalysts for the degradation of
chlorinated ethylenes.
Phosphinorhodium complexes also catalytically activate
sp2-hybridized C-F bonds, including fluorinated arenes46-51
and alkenes.52-56 In these systems, H2 and various silanes have
been employed as reducing agents, and alkyl and aryl phos-
phines have been used as ancillary ligands.46-56
Previously, we reported that dehalogenation of vinyl
fluoride, vinyl chloride, and chlorofluoroethylenes occurs
in the presence of (PPh3)3RhCl and Et3SiH.52 This reaction
was found to have an intermolecular preference for C-F
bond activation versus C-Cl bond activation, with vinyl
fluoride degrading the fastest of the substrates tested. The
data also support an intramolecular preference for C-Cl
bond activation versus C-F bond activation. Specifically,
dehalogenation of chlorofluorethylenes resulted in the ob-
servation of only vinyl fluoride and not vinyl chloride. The
rate of dehalogenation increased with sequential halogen
removal, such that dihaloalkenes were degraded slower than
monohaloalkenes. The substitution pattern also affected the
rates; 1,1-dihaloalkenes were degraded faster than the 1,2-
dihaloalkenes. No isomerization of the 1,2-dichloroethy-
lenes was observed during the dehalogenation. This system
showed marked preference for sp2-hybridized C-X bonds
versus sp3-hybridized bonds. Halogenated ethylenes were
readily degraded under the conditions, but 1-fluorooctane,
1,1,2-trichloroethane, dichloromethane, and 1,2-dichloro-
ethane were not. While this earlier study established sub-
strate scope and preferences, a firm mechanistic understand-
ing was not attained.
Heterogeneous rhodium-based catalysts have been used to
dechlorinate chlorinated alkenes and alkanes. For example,
Rh/alumina, Rh/SiO2, or RhCl3 catalysts in the presence of
H2 fully dechlorinate dichloropropane,36 dichlorobutenes,37
perchloroethylene,38 trichloroethylene,29 and 1,2-dichloro-
ethane.29,39,40 However due to difficulties associated with
studying reaction mechanisms in heterogeneous systems,
little is known about the mechanism of these reactions.
There are also a few examples of homogeneous catalytic
systems that use a phosphinorhodium catalyst and hydrogen
to degrade chlorofluoromethanes41 and chloroalkanes.40,42
Furthermore, Esteruelas et al. employed chlorotris(triphenyl-
phosphino)rhodium, (PPh3)3RhCl, as a homogeneous cata-
lyst in the presence of triethylsilane (Et3SiH) to dehalogenate
chlorinated alicyclic43 and aryl44,45 compounds (i.e., γ-hexa-
chlorocyclohexane, chlorobenzene, and polychlorinated
benzenes). The proposed first step of this reaction is the
production of Et3SiCl and (PPh3)3RhH as the active catalytic
intermediate, and for hexachlorocyclohexanes, an oxidative
addition/β-chloride elimination mechanism was proposed.43
Herein we report further studies of the catalytic dehalo-
genation of halogenated ethylenes by (PR3)3RhCl. We ex-
pand on our preliminary communication of the dehalogena-
tion reactions catalyzed by (PPh3)3RhCl in the presence of
Et3SiH and investigate the dehalogenation using H2 as the
reducing agent. These two systems were found to exhibit
significant differences. We propose mechanisms for the
reduction with Et3SiH and H2, which favor Rh(I) and Rh-
(III) pathways, respectively.
(26) Middeldorp, P. J. M.; Luijten, M. L. G. C.; Van De Pas, B. A.;
Van Eekert, M. H. A.; Kengen, S. W. M.; Schraa, G.; Stams, A. J. M.
Biorem. J. 1999, 3, 151–169.
(27) Mohn, W. W.; Tiedje, J. M. Microbiol. Rev. 1992, 56, 482–507.
(28) Smidt, H.; de Vos, W. M. Annu. Rev. Microbiol. 2004, 58, 43–73.
(29) Bozzelli, J. W.; Chen, Y.-M. Chem. Eng. Commun. 1992, 115,
1–11.
(30) Kovenklioglu, S.; Cao, Z.; Shah, D.; Farrauto, R. J.; Balko,
E. N. AIChE J. 1992, 38, 1003–1012.
(31) Lowry, G. V.; Reinhard, M. Environ. Sci. Technol. 1999, 33,
1905–1910.
(32) McNab, W. W. Jr; Ruiz, R. Chemosphere 1998, 37, 925–936.
(33) McNab, W. W. Jr; Ruiz, R.; Reinhard, M. Environ. Sci. Technol.
2000, 34, 149–153.
Experimental Procedures
General Considerations. All manipulations were performed
under a nitrogen atmosphere using standard Schlenk techniques
or in a nitrogen-filled glovebox (MBraun Unilab). All solvents
and reagents were purchased from commercial suppliers unless
otherwise stated; solvents were dried prior to use. Benzene-d6 and
p-xylene were distilled from a sodium/benzophenone ketyl
mixture under nitrogen and stored in a sealed flask. cis-1,2-
Dichloroethylene, trans-1,2-dichloroethylene (TCI America), tri-
chloroethylene, perchloroethylene, 1,1-dichloroethylene, dichlor-
omethane (Sigma-Aldrich), 1,2-dichloroethane, and trichloro-
ethane (Acros) were dried over CaCl2, vacuum transferred into
(34) Nishijima, W.; Ochi, Y.; Tsai, T.-Y.; Nakano, Y.; Okada, M.
Appl. Catal., B 2004, 51, 135–140.
(35) Schreier, C. G.; Reinhard, M. Chemosphere 1995, 31, 3475–3487.
(36) Scharfe, G.; Wilhelms, R. E. Catalytic dechlorination of dichloro-
propane. 71-2164074, 19711223, 1973.
(37) Vojtko, J.; Hrusovsky, M. Chem. Zvesti. 1972, 26, 460–465.
(38) Ordonez, S.; Sastre, H.; Diez, F. V. Thermochim. Acta 2001, 379,
25–34.
(39) Chang, C.-C.; Reo, C. M.; Lund, C. R. F. Appl. Catal., B 1999,
20, 309–317.
(40) Trabuco, E.; Ford, P. C. J. Mol. Catal. A: Chem. 1999, 148, 1–7.
(41) Cho, O.-J.; Lee, I.-M.; Park, K.-Y.; Kim, H.-S. J. Fluorine Chem.
1995, 71, 107–109.
˚
a sealed flask, and stored in the dark over 4 A molecular sieves.
˚
Tetrahydrofuran-d8 was dried over 4 A molecular sieves. Vinyl
(42) Kim, H.-S.; Cho, O.-J.; Lee, I.-M.; Hong, S.-P.; Kwag, C.-Y.;
Ahn, B. S. J. Mol. Catal. A: Chem. 1996, 111, 49–65.
(43) Esteruelas, M. A.; Herrero, J.; Olivan, M. Organometallics 2004,
23, 3891–3897.
chloride and d1-triethylsilane (Et3SiD) (Sigma-Aldrich), all fluo-
rine-containing gases (SynQuest Laboratories, Inc.), H2 (Airgas),
(44) Esteruelas, M. A.; Herrero, J.; Lopez, F. M.; Martin, M.; Oro,
L. A. Organometallics 1999, 18, 1110–1112.
(45) Diaz, J.; Esteruelas, M. A.; Herrero, J.; Moralejo, L.; Olivan, M.
J. Catal. 2000, 195, 187–192.
(46) Aizenberg, M.; Milstein, D. Science 1994, 265, 359–361.
(47) Aizenberg, M.; Milstein, D. J. Am. Chem. Soc. 1995, 117, 8674–
8675.
(51) Fuchibe, K.; Ohshima, Y.; Mitomi, K.; Akiyama, T. J. Fluorine
Chem. 2007, 128, 1158–1167.
(52) Peterson, A. A.; McNeill, K. Organometallics 2006, 25, 4938–
4940.
(53) Noveski, D.; Braun, T.; Schulte, M.; Neumann, B.; Stammler,
H.-G. Dalton Trans. 2003, 4075–4083.
(48) Young, R. J. Jr; Grushin, V. V. Organometallics 1999, 18, 294–
296.
(49) Edelbach, B. L.; Jones, W. D. J. Am. Chem. Soc. 1997, 119, 7734–
7742.
(54) Braun, T.; Noveski, D.; Neumann, B.; Stammler, H.-G. Angew.
Chem., Int. Ed. 2002, 41, 2745–2748.
(55) Braun, T.; Noveski, D.; Ahijado, M.; Wehmeier, F. Dalton
Trans. 2007, 3820–3825.
(50) Ishii, Y.; Chatani, N.; Yorimitsu, S.; Murai, S. Chem. Lett. 1998,
157–158.
(56) Braun, T.; Wehmeier, F.; Altenhoener, K. Angew. Chem., Int.
Ed. 2007, 46, 5321–5324.

5984 Organometallics, Vol. 28, No. 20, 2009
Peterson et al.
and deuterium (Matheson Gas Products, Inc.) were used as
received. (PPh3)3RhCl,57 (PPh3)4RhH,58 (PPh3)3RhH,43 (PPh3)3-
RhF,59 (PMe3)3RhCl,60 and (PEt3)3RhCl60 were prepared ac-
(3.3 mM; used as an internal standard) was prepared in C6D6.
An aliquot (700 μL) of this stock solution was added by syringe
into a sealable NMR tube. The substrate was then either added
via syringe in the glovebox or transferred as a gas on a Schlenk
line using a known volume bulb. The headspace was evacuated
from the tube by freezing the solution and placing it under
vacuum. The mixture was warmed to room temperature, and the
headspace was filled with approximately 1 atm of H2 (or D2).
The tube was inverted three times to mix H2 into the C6D6 and
then immediately frozen in liquid nitrogen. Products of the
dehalogenation reactions were determined by comparing the
NMR resonances, chemical shifts, and coupling constants to
authentic standards.
1
cording to literature procedures and characterized by H and
31P{1H} NMR spectroscopy. Dechlorination reactions were
performed in sealed NMR tubes to prevent volatilization of
reactants or products. NMR spectra were acquired on Varian
UNITY Inova NMR spectrometers (300, 500, and600 MHz) and
referenced to the residual protiated solvent (C6D5H, 7.16 ppm).
19FNMRand 31P{1H} NMR spectrawereacquiredat282.12 and
121.4 MHz, respectively, with 31P referenced to 85% phosphoric
acid and 19F referenced to trichlorofluoromethane, both as
external standards.
[P(p-FC6H4)3]3RhCl and [P(p-MeOC6H4)3]3RhCl. These
syntheses were modified from the procedure described by
Montelatici et al.61 1,5-Cyclooctadienerhodium(I) chloride di-
mer ([(COD)RhCl]2) was dissolved in toluene, and 3 molar equiv
of the desired phosphine was added. The solution was left to stir
for 14 h at ambient temperature. Complexes were purified by
crystallization from toluene. Crystals were isolated by filtration
and dried in vacuo. The complexes were characterized by 1H and
31P{1H} NMR and compared to the literature.61
Dechlorination Using (PPh3)4RhH. A stock solution contain-
ing (PPh3)4RhH (2.4 mM) and p-xylene (3.3 mM; used as an
internal standard) was prepared in C6D6. An aliquot (700 μL) of
this stock solution was added by syringe into a sealable NMR
tube. Substrate (20-30 mM) was added to this solution by
syringe. The tube was mixed by inversion. Products and reaction
progress was monitored using NMR spectroscopy at room
temperature.
Dechlorination Using (PPh3)3RhH (or (PPh3)3RhD). A pro-
cedure similar to the literature was adapted.43 A solution
containing (PPh3)3RhCl (10 mM) and Et3SiH(D) (30 mM)
was prepared in C6D6 in a sealable NMR tube. The solvent
and volatile materials were removed under vacuum on a Schlenk
line. The solid was then redissolved in 700 μL of C6D6 in the
sealable NMR tube, and an NMR spectrum was acquired to
confirm only (PPh3)3RhH(D) was present. Substrate (20 mM)
was then added to this solution by syringe. The tube was mixed
by inversion. Products and reaction progress were monitored
using NMR spectroscopy at room temperature.
(PPh3)nRhCl for Use in Phosphine Ratio Experiments. The
catalyst was prepared in situ as described above using [(COD)-
RhCl]2 and PPh3. Appropriate equivalents of triphenylpho-
sphine were added to a 2.4 mM Rh solution with 1.7 mM
p-xylene in C6D6. The solutions were left overnight to react.
The Et3SiH (125 mM) and cis-DCE (30 mM) were added by
syringe. The sealable NMR tubes were frozen immediately and
stored in liquid nitrogen until inserted in the NMR probe.
Kinetic data collection was performed as described above.
Reaction Kinetics with Et3SiH. Kinetic experiments were per-
formed in sealed NMR tubes monitoring the reaction by 1H NMR
spectroscopy. The reaction mixture was kept frozen until imme-
diately prior to placement in the NMR probe; NMR probe
temperature was set to 35 or 55 °C. The kinetic data were collected
in an arrayed experiment set to scan every 2 to 5 min. The rates of
dehalogenation were determined by quantifying the disappearance
of the substrate resonances, integrating versus internal standard
p-xylene. These data were plotted as concentration of substrate
versus time and fit to a first-order exponential decay function.
Relative Reaction Rates with H2. Substrate disappearance
after 60 min was determined using sealed NMR tubes and
monitoring the reaction by 1H NMR spectroscopy. The reaction
mixture was kept frozen until immediately prior to placement in
the NMR probe; NMR probe temperature was set to 22 °C. A
quantitative NMR spectrum was taken to determine the initial
concentration of the substrate. The NMR tube was then im-
mediately frozen in liquid nitrogen until ready to mix by
inversion. The tube was thawed to room temperature and mixed
by repeated inversion for 60 min, at a rate of 15 inversions per
minute, and then frozen in liquid nitrogen. The final NMR
spectrum was taken with the same parameters as the first NMR
spectrum. The percentages of substrate dehalogenation were
determined by quantifying the disappearance of the substrate
resonances, integrating versus internal standard p-xylene, and
determining the amount reacted by comparing the initial NMR
and final NMR integration.
Dechlorination Using Et3SiH. A stock solution containing
(PPh3)3RhCl (2.4 mM), Et3SiH (125 mM), and p-xylene (3.3
mM; used as an internal standard) was prepared in C6D6. An
aliquot (700 μL) of this stock solution was added by syringe to a
sealable NMR tube. The substrate was then either added via
syringe in the glovebox or transferred as a gas on a Schlenk line
using a known volume bulb. Kinetic data collection was per-
formed as described above. Products of the dehalogenation
reactions were determined by comparing the NMR resonances,
chemical shifts, and coupling constants to authentic standards.
Dechlorination Using H2 (or D2) Gas. A stock solution con-
taining (PPh3)3RhCl (2.4 mM), NEt3 (150 mM), and p-xylene
Results
Dechlorination Using Et3SiH. Treatment of cis-1,2-di-
chloroethylene (cis-DCE, 30 mM in C6D6) with Et3SiH
(125 mM) and
a catalytic amount of (PPh3)3RhCl
(2.4 mM) at 55 °C leads to complete dechlorination in 80
min (t1/2 of 23 min) (eq 1). The reaction was monitored by 1H
NMR spectroscopy. The initial products observed were vinyl
chloride and Et3SiCl (eq 1). Vinyl chloride was quickly
dechlorinated to ethylene. Over the course of the reaction,
1-butene and, eventually, cis- and trans-2-butene were also
produced. After cis-DCE was fully degraded, a small amount
of ethane was also observed. These results suggest that
dechlorination to ethylene is followed by dimerization to
1-butene and ultimately isomerization to 2-butenes. These
dimerization62,63 and isomerization64-66 reactions are
known. The yield of observed products was essentially
(57) Osborn, J. A.; Wilkinson, G. Inorg. Synth. 1990, 28, 77–79.
(58) Ahmad, N.; Levison, J. J.; Robinson, S. D.; Uttley, M. F. Inorg.
Synth. 1990, 28, 81–83.
(59) Grushin, V. V.; Marshall, W. J. J. Am. Chem. Soc. 2004, 126,
3068–3069.
(60) Price, R. T.; Andersen, R. A.; Muetterties, E. L. J. Organomet.
(62) Cramer, R. J. Am. Chem. Soc. 1965, 87, 4717–4727.
(63) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G.
J. Chem. Soc. A 1966, 1711–1732.
(64) Tuner, M.; Von Jouanne, J.; Brauer, H. D.; Kelm, H. J. Mol.
Catal. 1979, 5, 433–445.
(65) Tuner, M.; Von Jouanne, J.; Brauer, H. D.; Kelm, H. J. Mol.
Catal. 1979, 5, 425–431.
Chem. 1989, 376, 407–417.
(61) Montelatici, S.; van der Ent, A.; Osborn, J. A.; Wilkinson, G.
J. Chem. Soc. A 1968, 1054 –1058.
(66) Tuner, M.; Von Jouanne, J.; Kelm, H. J. Mol. Catal. 1979, 5,
447–457.

Article
Organometallics, Vol. 28, No. 20, 2009 5985
Table 1. Observed Rate Constant of Substrate Degradation (kobs
with Et3SiH
)
c
substrate
vinyl fluoride
1,1-chlorofluoroethylene
vinyl chloride
1,1-DCE
cis-DCE
trans-DCE
cis-1,2-chlorofluoroethylene
trans-1,2-chlorofluoroethylene
TCE
kobs (10-5 s-1 a
)
krel
50 ( 5b
24 ( 8b
71.4
34.3
11.4
5.7
3.9
3.6
3.1
2.4
1.0
<0.1
8 ( 1b
4 ( 2b
2.7 ( 0.8b
2.5 ( 0.9
2.2 ( 0.3b
1.7 ( 0.4b
0.7 ( 0.3
<0.1
PCE
a Error is the standard deviation of the replicate experiments. Values
were determined from loss of substrate versus time using H NMR at
1
35 °C. Each sealable NMR tube contained substrate (20-30 mM),
catalyst ((PPh3)3RhCl, 2.4 mM), Et3SiH (125 mM), and internal stan-
dard (p-xylene, 3.3 mM) in C6D6. b Ref 52. c Relative rate constants (krel
are relative to the rate of dehalogenation of TCE (0.7 ꢀ 10-5 s-1).
)
observed rate constants for those previously reported, as well
as for trans-1,2-dichloroethylene (trans-DCE), TCE, and
PCE, are compiled in Table 1. Reinforcing the previous
conclusions about the reaction, the rate of dehalogenation
was found to increase with sequential chlorine removal, such
that PCE is degraded slowest, followed by TCE and the
dihaloalkenes. The most facilely degraded substrates were
the monohaloalkenes. When comparing among the dichlor-
oethylenes, the relative rates of degradation are very similar,
with 1,1-dichloroethylene (1,1-DCE) reacting the fastest of
the three.
Et3SiCl was the major silane product observed during the
dehalogenation reactions; however, it was not the only silane
product observed, as hexaethyldisilane (Et3Si-SiEt3) and
vinyltriethylsilane (Et3SiCHCH2) were also observed. Dur-
ing the dechlorination of 1,1-DCE, TCE, and vinyl chloride,
a small amount of Et3SiCHCH2 was observed by 1H NMR
spectroscopy. With all substrates, Et3SiCHCH2 was ob-
served after the halogenated substrates were completely
consumed in the catalytic reaction. This suggests that the
Et3SiCHCH2 is produced from ethylene and some substrates
inhibit its formation. This was tested by reacting ethylene
under the dehalogenation reaction conditions. When ethy-
lene was reacted with (PPh3)3RhCl and Et3SiH under iden-
tical reaction conditions, Et3SiCHCH2, ethane, and Et3SiCl
(small amount from loss of Cl from (PPh3)3RhCl) were
observed.
Experiments were performed to determine the reaction
order in catalyst, (PPh3)3RhCl, and Et3SiH. Dechlorination
was found to be first-order in catalyst, (PPh3)3RhCl, at
concentrations ranging from 0.1 to 2.4 mM (Figure 2). This
was determined by measuring the initial rates of cis-DCE
degradation under pseudo-first-order reaction conditions at
55 °C in the presence of a constant amount of Et3SiH
(125 mM) and varying amounts of catalyst. The reaction
was found to be zero-order in Et3SiH by measuring the
pseudo-first-order kobs for cis-DCE degradation at 55 °C
and monitoring the reaction by 1H NMR spectroscopy. The
studies were conducted with Et3SiH (20-300 mM) and
(PPh3)3RhCl (2.4 mM) (Figure 3). The rate of rhodium-
mediated dechlorination does not depend on Et3SiH con-
centrations when concentrations above 20 mM are used.
Dehalogenation Using Dihydrogen. Dihydrogen, H2, was
also used as the reducing agent for catalytic dehalogenation.
Treatment of cis-DCE (30 mM in C6D6) with (PPh3)3RhCl
Figure 1. (A) cis-DCE degradation (b) by (PPh3)3RhCl
(2.4 mM) and Et3SiH (125 mM) at 55 °C and the subsequent
appearance of reaction products (9 = sum of reaction products
C2 mass), as well as C2 mass balance (O). (B) Growth of specific
reaction products during the course of cis-DCE degradation:
vinyl chloride (2), ethylene (1), butenes ([). Data are fitted with
first-order decay (cis-DCE), first-order exponential growth and
decay (vinyl chloride and ethylene), and first-order exponential
growth (butenes).
quantitative, as vinyl chloride, ethylene, and butenes com-
prise over 95% of the C2 mass. The product yields were
determined by 1H NMR integration versus an internal
standard, p-xylene.
To further characterize this C-Cl bond activation process,
a study of the reaction kinetics was undertaken. Kinetic
experiments were conducted in a temperature-controlled
NMR probe, and concentrations of the reactants and pro-
1
ducts were determined by integration of the H NMR reso-
nances relative to those of the internal standard (p-xylene).
Pseudo-first-order reaction kinetics were observed for the loss
of cis-DCE under these reaction conditions. A sample kinetic
plot depicting the exponential degradation of cis-DCE with
concomitant growth of products is shown in Figure 1. The C2
mass balance is also depicted.
All products were observed after 10 min of reaction time at
55 °C. This is consistent with fast degradation of the vinyl
chloride intermediate.52 The dimerization and isomerization
reactions occurred while chlorinated substrates were still
present in the reaction mixture.
Recently we reported that dehalogenation of vinyl fluoride,
vinyl chloride, and chlorofluoroethylenes occurs in the
presence of (PPh3)3RhCl and Et3SiH.52 Pseudo-first-order

5986 Organometallics, Vol. 28, No. 20, 2009
Peterson et al.
Table 2. Substrate Degradation Rates with H2 as the Reductanta
b
krel
substrate
percent reacted (%)
vinyl chloride
cis-DCE
trans-DCE
1,1-DCE
TCE
75 ( 1
75 ( 3
44 ( 4
31 ( 6
13 ( 8
10.2
10.1
4.2
2.7
1.0
a Percent reacted values were determined from loss of substrate after
60 min of constant inversion using quantitative 1H NMR spectroscopy
at room temperature, 22 °C. Each sealable NMR tube contained
substrate (20-30 mM), catalyst ((PPh3)3RhCl, 2.4 mM), H2 (1 atm),
NEt3 (150 mM), and internal standard (p-xylene, 3.3 mM) in C6D6. The
estimated error is the standard deviation of three trials. b Relative rate
constants were determined assuming a first-order degradation model
and were normalized to the rate of TCE degradation.
Figure 2. Observed degradation rate of cis-DCE in the presence
of 125 mM Et3SiH and varying concentrations of catalyst
(0.1 to 2.4 mM) in C6D6. Values were determined using initial
1
agent. Kinetic studies in the NMR probe were not con-
ducted due to the low solubility of H2 in C6D6 and slow
mass transfer kinetics, which made it difficult to maintain
pseudo-first-order conditions. The substrates were com-
pared by quantifying substrate lost after 60 min of reaction
at room temperature. Each reaction mixture consisted of
substrate (20-30 mM), (PPh3)3RhCl (2.4 mM), and NEt3
(150 mM); the headspace in the NMR tube was filled with
H2 at 1 atm. The percent reacted after 60 min was calculated
for the chlorinated ethylenes and is compiled in Table 2.
The general reactivity trends were the same as when Et3SiH
was employed. The rate of dechlorination increases with
loss of chlorine substituents, as vinyl chloride degraded the
fastest and TCE the slowest. However, unlike the Et3SiH
system, cis-DCE is degraded at a similar rate to vinyl
chloride, while 1,1-DCE and trans-DCE degraded at a
slower rate. All of the dichloroethylenes produce vinyl
chloride as a reactive intermediate, which was then de-
chlorinated to ethylene and further hydrogenated to
ethane. When ethylene was reacted as the substrate in this
system (with H2, NEt3, and (PPh3)3RhCl), hydrogenation
occurs, and ethane is observed.
The reaction was found to be first-order in catalyst, (PPh3)3-
RhCl, when H2 was used as the hydride source. The fraction of
cis-DCE degraded in 60 min versus catalyst concentration (0 to
2.4 mM) is depicted in Figure 4. The data are plotted on a
logarithmic scale, which should linearize the data set if the
reaction is pseudo-first-order and first-order in catalyst. Indeed,
a linear relationship is observed (Figure 4). There is no effect on
the percent cis-DCE degraded in 60 min when varying the
concentration of NEt3 from 50 to 200 mM when the amount of
(PPh3)3RhCl was kept constant (2.4 mM) (Figure 5).
rates for loss of substrate as monitored by H NMR at 55 °C.
The fit is linear to the initial rate data, R2 = 0.9721. Error bars
were determined from the standard error in the linear fit of the
initial rate plots.
Figure 3. Observed cis-DCE degradation (kobs) in the presence
of 2.4 mM (PPh3)3RhCl and varying concentrations of Et3SiH
(from 22 to 300 mM) in C6D6. Error bars were determined from
the standard error in the exponential fit to the substrate decay
kinetic plots, except when the Et3SiH concentration was 125
mM, in which case the error is the standard deviation of 11
experiments.
(2.4 mM) and H2 (1 atm) at room temperature resulted in
complete dechlorination after 80 min (eq 2). In this case,
NEt3 (150 mM) was added to scavenge the resulting HCl. In
the absence of NEt3, catalyst degradation and formation of
[HPPh3][Cl] were observed. The reaction was mixed by
inverting continuously until analyzed. The suite of products
observed from dehalogenation using H2 was similar to that
observed using Et3SiH. Initially, vinyl chloride and ethylene
were produced. However, unlike the Et3SiH, no butenes were
observed. Hydrogenation was observed during the dechlor-
ination reaction, resulting in ethane and a small amount of
1,2-dichloroethane and 1-chloroethane. In all cases, 93% of
the C2 mass was accounted for. The 1,2-dichloroethane and
1-chloroethane products account for 2% of the C2 mass
balance.
When vinyl fluoride (25 mM) is reacted with (PPh3)3RhCl
(2.4 mM) and H2 (1 atm) in the presence of NEt3 (150 mM),
the major reaction pathway is hydrogenation, as the major
product is fluoroethane and the minor product ethylene.
Because defluorination was not the major degradation path-
way, the percent reacted in 60 min is not included in Table 2.
Fluoroethane and ethylene were observed from the reaction
of vinyl fluoride with independently prepared (PPh3)3RhF59
and H2. This observation is consistent with (PPh3)3RhF being
a potential intermediate in the catalytic cycle.
Chlorofluoroethylenes were reacted with (PPh3)3RhCl,
H2, and NEt3. The first product observed was vinyl fluoride,
and no vinyl chloride was detected during the dehalogena-
tion, similar to the Et3SiH system. Some hydrogenation
product, chlorofluoroethane, was also observed. The major
final product was fluoroethane, the hydrogenation product
of the vinyl fluoride.
The relative reactivity of (PPh3)3RhCl with halogenated
substrates was further examined using H2 as the reducing

Article
Organometallics, Vol. 28, No. 20, 2009 5987
Figure 4. Natural log of the fraction of cis-DCE degraded in
60 min versus catalyst concentration. Reaction conditions: 22 (
1 °C, 1 atm H2, 150 mM NEt3, and varying concentration of
catalyst (0 to 2.4 mM) in C6D6. The fit is linear to the data, R2 =
0.9799.
Figure 5. Percent cis-DCE degraded in 60 min at room tem-
perature in the presence of H2 (1 atm), (PPh3)3RhCl (2.4 mM),
and varying concentrations of NEt3 (50 to 200 mM) in C6D6.
Values were determined by calculating loss of substrate using 1H
NMR at 22 °C. Error bars are set to 3%, which is based on the
standard deviation of triplicate cis-DCE degradation.
Sp3-Hybridized Carbon-Halogen Bonds. Similar to Et3-
SiH, when H2 is used as the reducing agent under identical
reaction conditions, there is no reaction with 1,2-chloroethane
or dichloromethane after two days at room temperature.
NMR Observation of Intermediates. When Et3SiH and
NMR spectroscopy. The dechlorination products produced
were ethylene and vinyl chloride, and 90% of the C2 mass
was accounted for by these products and unreacted cis-DCE.
When H2 (1 atm) was reacted with (PPh3)3RhCl, only
(PPh3)3RhH2Cl43,63,69 was produced. This was indicated by
1H NMR resonances observed at -9.4 and -16.7 ppm. The
31P{1H} NMR spectrum contained a doublet at 39.4 ppm
(116 Hz) and a broad peak at 20 ppm. During the dehalogena-
tion reactions with H2 and (PPh3)3RhCl, the major species
observed by 31P{1H} NMR spectroscopy was (PPh3)3RhH2Cl.
A small amount of (PPh3)3RhCl and (PPh3)2RhCl(alkene)
(36.6 ppm, 129 Hz)70 were also observed in the 31P{1H}
NMR spectrum. No resonances attributable to (PPh3)3RhH
were observed in either 1H or 31P{1H} NMR spectra.
1
(PPh3)3RhCl were combined in a 1:1 or 2:1 ratio, a new H
NMR resonance consistent with a Rh-H species was obser-
ved at-7.9 ppm. Esteruelaset al. observed the same result and
assigned the resonance to (PPh3)3RhH.43,67 In our system,
when excess Et3SiH (>10 equiv) is allowed to react with
(PPh3)3RhCl, the only rhodium hydride complex observed
initially is (PPh3)2RhH(Cl)SiEt3, identified by the upfield
hydride resonance (δ -14.7, JRh-H = 22 Hz) in the 1H{31P}
NMR spectrum and its31P{1H} resonance (δ 39.7, d, JRh-P
=
128 Hz).68 With time, the (PPh3)2RhH(Cl)SiEt3 1H resonance
at -14.7 ppm disappears, and the (PPh3)3RhH resonance,
-7.9 ppm, appears.67 The 31P{1H} NMR spectrum shows the
same behavior, with the (PPh3)2RhH(Cl)SiEt3 doublet at
39.7 ppm giving way to the (PPh3)3RhH doublet at 42 ppm
(158 Hz). We propose that the solution of Et3SiH and
(PPh3)3RhCl is in equilibrium, such that LnRh(I)-H and
LnRh(III)-H complexes are present during the catalytic cycle.
This is supported by the 31P{1H} NMR resonances observed
during dechlorination reactions; both (PPh3)2RhH(Cl)-
SiEt3 (δ 39.2, d, JRh-P = 130 Hz) and (PPh3)3RhH (δ 42, d,
JRh-P = 156 Hz) are observed. This differs from observations
made by Esteruelas et al. during the dechlorination of hexa-
chlorocyclohexane. They reported a doublet at 39.4 ppm
(115 Hz) in the 31P{1H} NMR spectrum and two hydride
peaks in the 1H NMR spectra (-9.38 and -16.73 ppm) that
they assigned to (PPh3)3RhH2Cl.43,63,69
(PPh3)3RhH can be isolated by reacting (PPh3)3RhCl with
excess Et3SiH and then removing Et3SiCl and unreacted
Et3SiH under vacuum. (PPh3)3RhH isolated in this way was
found to react immediately with cis-DCE to yield vinyl
chloride and ethylene, as well as (PPh3)3RhCl. Further-
more, independently synthesized and isolated (PPh3)4RhH
(2.4 mM) reacted readily with cis-DCE (20 mM) in C6D6
upon mixing at room temperature. (PPh3)3RhCl and free
PPh3 were formed in these reactions, as indicated by 31P{1H}
Effect of Phosphine:Rhodium Ratio on Rate. The phosphine
dependence of the dechlorination of cis-DCE was explored in
order to gain insight into the active catalytic species and to
determine the optimum rhodium to phosphine ratio for in situ
catalyst generation. Reactions were performed with cis-DCE
(30 mM) and Et3SiH (125 mM). The catalyst was formed in situ
by mixing PPh3 with[(COD)RhCl]2 at various ratios from 1:0 to
1:5 Rh:PPh3. ThetotalRhandEt3SiH concentrations were kept
constant in all experiments, and the amount of PPh3 was varied
accordingly. The effect of phosphine concentration on the rate
constant was determined by measuring the pseudo-first-order
1
kobs for cis-DCE degradation at 55 °C using H NMR spec-
troscopy.
The optimal rate constant was achieved with 3 equiv of
phosphine. The reaction rate constant decreased with addi-
tional phosphine (Figure 6), and no dehalogenation was
observed in the absence of phosphine. This result differs
from previous reports on hydrogenation and dechlorination
reactions where an initial ligand dissociation step is thought
to produce (PPh3)2RhCl, with a 2:1 PPh3:Rh ratio, yielding
the optimal rate of reaction (eq 3).43,61,63,71
The inhibitory effect of greater than 3 equiv of phosphine
was further verified in an independent experiment where
1 equiv of PPh3 was added to a reaction mixture containing
(67) Strauss, S. H.; Diamond, S. E.; Mares, F.; Shriver, D. F. Inorg.
Chem. 1978, 17, 3064–3068.
(68) Duckett, S. B.; Galvez-Lopez, M.-D.; Perutz Robin, N.; Schott,
D. Dalton Trans. 2004, 2746–2749.
(69) Duckett, S. B.; Newell, C. L.; Eisenberg, R. J. Am. Chem. Soc.
1994, 116, 10548–10556.
(70) Naaktgeboren, A. J.; Nolte, R. J. M.; Drenth, W. J. Am. Chem.
Soc. 1980, 102, 3350–3354.
(71) Ohtani, Y.; Fujimoto, M.; Yamagishi, A. Bull. Chem. Soc. Jpn.
1977, 50, 1453–1459.

5988 Organometallics, Vol. 28, No. 20, 2009
Peterson et al.
Table 3. Rate Constants Obtained for Various (PR3)3RhCl
Complexes with cis-DCE and Et3SiH and the Corresponding
Phosphine Properties72
isolated
catalyst
kobs (10-4 s-1) kobs (10-4 s-1
)
pKa
(PR3Hþ)
cone angle
θ
35 °C
55 °C
n.d.a
[P(p-FC6H4)3]3-
RhCl
1.9 ( 0.7
1.97
145
(PPh3)3RhCl
[P(p-MeO-
0.27 ( 0.08
5.0 ( 0.5
0.40 ( 0.02
2.73
4.59
145
145
n.d.a
C6H4)3]3RhCl
(PEt3)3RhCl
(PMe3)3RhCl
<0.01
<0.01
<0.01
<0.01
8.69
8.65
132
118
a n.d. = not determined.
Figure 6. Observed rate constants (kobs) for cis-DCE degrada-
tion are compared to the phosphine:rhodium ratio for the
reaction of cis-DCE with Et3SiH and the catalyst at 55 °C in
C6D6. In these studies, the catalyst was made from the in situ
reaction of [(COD)RhCl]2 with PPh3. Error bars are the error in
the exponential fit of the 1H NMR data.
and cis-DCE. The syn-β-chloride elimination product, E-1-
chloro-2-deuteroethylene (JH-H = 15 Hz), was observed in
the H NMR spectrum (eq 5) as hypothesized. No other
1
chlorodeuteroethylenes were observed. These results, along
with the reaction preferences previously discussed, support
an insertion/syn-β-chloride elimination mechanism when
hydrogen is used as the hydride source. This is also the only
product observed when either Et3SiCl or Et3N was added to
the reaction mixture of cis-DCE, D2, and (PPh3)3RhCl,
indicating that these species play no role in the β-Cl elimina-
tion product formed.
Et3SiH, cis-DCE, and (PPh3)3RhCl. Under these conditions,
a 3-fold decrease in the observed rate constant was observed
compared to the same experiment without added phosphine.
A similar result was also observed by Esteruelas et al. for
chlorinated benzene dechlorination.43,44
The cis-DCE degradation rate constant, kobs, for in situ
generated (PPh3)3RhCl (PPh3:Rh ratio of 3:1) was an order
of magnitude lower than when the reaction was performed
with isolated and purified (PPh3)3RhCl. The source of this
discrepancy was found to be 1,5-cyclooctadiene, which acted
as an inhibitor for cis-DCE dechlorination by (PPh3)3RhCl
and Et3SiH (data not shown).
In the H2 system, the percentage of cis-DCE dechlorinated
in 60 min in the presence of (PPh3)3RhCl (2.4 mM), NEt3
(150 mM), 1 equiv of PPh3 (2.4 mM), and H2 (1 atm) was
found to be 42 ( 6%. This was a relative decrease of 2.5 times
compared to the reaction without the extra equivalent of
PPh3, signifying that the addition of PPh3 inhibited dechlori-
nation of cis-DCE when H2 was used.
Effect of Phosphine Structure on Rate. We also examined
the effect of the electronic nature of the phosphine used in
(PR3)3RhCl on the reaction kinetics with cis-DCE and
Et3SiH by preparing (PPh3)3RhCl analogues with trimethyl-
phosphine, triethylphosphine, tris(p-methoxyphenyl)pho-
sphine, and tris(p-fluorophenyl)phosphine. The observed
rate of reaction with cis-DCE decreased monotonically with
the basicity of the phosphine, with [P(p-FC6H4)3]3RhCl
being the most active (Table 3), while (PMe3)3RhCl and
(PEt3)3RhCl complexes were found to be inactive. This trend
was also observed by Montelatici et al. for the activation of
hydrogen by (PPh3)2RhX(C6H6).61 It was proposed that this
was due to both steric and electronic reasons.61
A similar result was obtained in the reaction of trans-DCE
with (PPh3)3RhCl and D2 in C6D6. Only resonances for the
syn-β-chloride elimination product were observed, Z-1-
chloro-2-deuteroethylene (JH-H = 7 Hz).
However, in a similar experiment using Et3SiD with
(PPh3)3RhCl and cis-DCE, a 1:1 mixture of E- and Z-1-
chloro-2-deuteroethylene was observed in the 1H NMR
spectrum, as previously reported (eq 6).52 When isolated
(PPh3)3RhD was reacted with cis-DCE, the same 1:1 mixture
of products was observed (eq 6). Switching solvents to
tetrahydrofuran-d8, for the reaction of Et3SiD with
(PPh3)3RhCl and cis-DCE, changed the product ratio to a
1:7 mixture of E- and Z-1-chloro-2-deuteroethylene.
Fully protonated (undeuterated) vinyl chloride is also
observed with time in the Et3SiD reaction and when the
(PPh3)3RhD is reacted with cis-DCE. The independent
reaction of vinyl chloride with Et3SiD and (PPh3)3RhCl
resulted in no deuterium incorporation into the residual
vinyl chloride. This indicates that vinyl chloride is not
isomerized under these conditions.
Isotope Labeling. Our leading hypothesis for the dehalo-
genation process is an insertion/β-chloride elimination mechan-
ism. The first step of this mechanism involves insertion of the
alkene into the Rh-H bond, yielding a β-chloroalkyl complex,
which then undergoes syn-β-chloride elimination. This process
gives rise to stereospecific incorporation of the hydride-derived
hydrogen. This mechanism is illustrated in eq 4.
When the (PPh3)3RhCl/Et3SiD system was treated with
trans-DCE, a 1:1.5 ratio of E- to Z-1-chloro-2-deuteroethy-
lene was observed. The same product ratio (1:1.5, E:Z) was
also observed when (PPh3)3RhD was reacted withtrans-DCE.
Discussion
Proposed Mechanism. The data for dehalogenation of
halogenated ethylenes using (PPh3)3RhCl with Et3SiH and
H2 support similar, yet distinct, mechanisms for the two
To test this mechanistic hypothesis, an experiment was
performed using D2 as the hydride source with (PPh3)3RhCl

Article
Organometallics, Vol. 28, No. 20, 2009 5989
Figure 7. Proposed mechanism for the catalytic dehalogenation
of chlorinated and fluorinated ethylenes using (PPh3)3RhCl and
Et3SiH. The product elimination step is labeled A. It is proposed
that reduction with Et3SiH occurs primarily through a Rh(I)
pathway, and the Rh(III) complex is highlighted in boldface.
Figure 8. Proposed mechanism for the catalytic dehalogenation
of chlorinated and fluorinated ethylenes using (PPh3)3RhCl
and H2. It is proposed that reduction with H2 occurs primarily
through a Rh(III) pathway with two possible product-forming
reactions: syn-β-X-elimination (B) and reductive elimination
(C). All Rh(III) complexes are highlighted in boldface.
reducing agents. The Et3SiH and H2 systems are character-
ized by four major differences:
1. Hydride complexes observed during reaction
2. Stereochemistry of elimination
3. Occurrence of hydrogenation/hydrosilylation products
4. Vinyl fluoride products
With fluorinated alkenes, the same mechanism is proposed.
However, following β-F-elimination (PPh3)3RhF is gener-
ated. This species has been shown to be a competent catalyst.52
When H2 is used, the dominant pathway is proposed to be
the Rh(III)-H pathway. Here, the initial step in the reaction
is oxidative addition of H2 to (PPh3)3RhCl (either before or
after phosphine dissociation). Alkene association at the open
coordination site of (PPh3)2RhH2Cl followed by insertion
into one of the Rh(III)-H bonds leads to a Rh-alkyl complex
(Figure 8). From here, reductive elimination (step C,
Figure 8) yields hydrogenation products, and syn-β-X elim-
ination (step B, Figure 8) yields dehalogenation products. As
shown, the chloro ligand remains bound to the Rh through-
out. Not shown explicitly is the hydrodehalogenation pro-
cess that leaves the haloalkene halogen bound to the
rhodium. In the case of fluoro substituents, this produces
(PPh3)2RhF. We have verified that (PPh3)3RhF is also a
competent catalyst using H2 as the reducing agent.
We propose two mechanisms that account for these
differences, with the key distinction between them being that
reduction with Et3SiH occurs through a Rh(I)-type pathway
(Figure 7), while the H2 reduction occurs through a Rh(III)-
type pathway (Figure 8). The mechanisms are presented first,
followed by a discussion of how they account for each of the
observations in the Et3SiH and H2 systems.
In the proposed mechanism, illustrated using vinyl chloride,
when Et3SiH is employed, the first step is oxidative addition of
Et3SiH to (PPh3)3RhCl to form (PPh3)3RhH(Cl)SiEt3, which
undergoes reductive elimination of Et3SiCl, to give (PPh3)3-
RhH. Alkene association with (PPh3)3RhH followed by inser-
tion into the Rh-H bond leads to a Rh-alkyl complex. The
alkyl complex may then undergo β-X elimination (step A),
yielding the dehalogenated alkene as well as the return of the
catalyst (PPh3)3RhX (Figure 7). This Rh-alkyl complex can
undergo either syn- or anti-β-X elimination, consistent with the
D-labeling data, as described in the stereochemistry of elim-
ination section below. We have depicted the Rh-alkyl complex
as a tris(triphenylphosphine) complex; however, there is no
direct evidence for this. Elimination could occur from, for
example, bis- or tris-phosphine complexes. The observation of
Et3SiCHCH2 suggests the possibility that a similar minor
pathway involving (PPh3)3Rh-SiEt3 may exist (eq 7).
C-X Oxidative Addition Mechanism. Before considering
in detail how the proposed mechanism accounts for the
experimental observations, it is worth pointing out how the
data do not support an alternative mechanistic hypothesis
involving C-X oxidative addition (eq 8).
In an oxidative addition/reductive elimination mecha-
nism, isotopic labeling studies would be expected to show
retention of stereochemistry (eq 8). In fact, the isotopic
labeling studies show inversion (with D2) and scrambling
(with Et3SiD). Oxidative addition is also expected to show a
preference for weaker sp3-hybridized C-X bonds over sp2-
hybridized C-X bonds, as well as similar reactivity toward
vinyl C-X and aryl C-X bonds. Neither was observed.
Vinyl fluoride was degraded faster than vinyl chloride, which
is also contrary to the expectations of an oxidative addition
The observation of Et3Si-SiEt3 lends credence to the
presence of Rh-silyl species, as these are known intermedi-
ates in dehydrocoupling of silanes.73-77
(72) Rahman, M. M.; Liu, H. Y.; Prock, A.; Giering, W. P. Organo-
metallics 1987, 6, 650–658.
(73) Goikhman, R.; Milstein, D. Chem.-Eur. J. 2005, 11, 2983–2988.
(74) Brown-Wensley, K. A. Organometallics 1987, 6, 1590–1591.
(75) Rosenberg, L.; Davis, C. W.; Yao, J. J. Am. Chem. Soc. 2001,
123, 5120–5121.
(77) Turculet, L.; Feldman, J. D.; Tilley, T. D. Organometallics 2004,
23, 2488–2502.
(76) Mitchell, G. P.; Tilley, T. D. Organometallics 1998, 17, 2912–2916.

5990 Organometallics, Vol. 28, No. 20, 2009
Peterson et al.
mechanism, as insertion into C-F bonds is more difficult
than C-Cl insertion. Finally, the observation of hydrogena-
tion products when H2 is used as the reducing agent also
supports insertion of the alkene into a Rh-H bond.
leaving group by rhodium would lead directly to an alkene
complex (eq 10).
Hydride Complexes Observed during Reaction. Numerous
lines of evidence support the intermediacy of rhodium
hydride species in both the Et3SiH and H2 dehalogenation
processes. First, conditions that produce Rh-H species were
needed for dehalogenation. In the Et3SiH system, no reac-
tion was observed when cis-DCE was treated with either
Et3SiH or (PPh3)3RhCl alone. Similarly in the H2 catalytic
system, no reaction occurs when cis-DCE was reacted with
either (PPh3)3RhCl or H2 alone. Both (PPh3)3RhCl and
either Et3SiH or H2 need to be present for reaction to occur.
Second, the isolated Rh-H complex, (PPh3)4RhH, reacts
directly with the halogenated ethylenes to give dehalogena-
tion products. Finally, Rh-H complexes were observed dur-
ing the dehalogenation reactions by 1H NMR spectroscopy.
However, the spectroscopic data suggest that H2 and
Et3SiH favor different pathways. During the course of the
dehalogenation reactions with Et3SiH, the Rh(I) hydride,
(PPh3)3RhH, is the primary hydride species observed. Under
H2, both 1H and 31P{1H} NMR spectra show that the Rh(III)
complex, (PPh3)3RhH2Cl, is present. Presumably, the differ-
ence in the two systems is the strong driving force for reductive
elimination of Et3SiCl (or Et3SiF in defluorination reactions)
which pushes the silane system toward (PPh3)3RhH.
Stereochemistry of Elimination. When D2 is reacted with
(PPh3)3RhCl and cis-DCE or trans-DCE, the stereochemistry of
the 1-chloro-2-deuteroethylene products (E-1-chloro-2-deuter-
oethylene or Z-1-chloro-2-deuteroethylene, respectively) is con-
sistent with exclusive syn-β-Cl-elimination (step B, Figure 8).
When Et3SiD is reacted with (PPh3)3RhCl and cis-DCE in
C6D6, a 1:1 mixture of E- to Z-1-chloro-2-deuteroethylene is
observed, and when trans-DCE is reacted, a slightly different
product ratio of 1:1.5 (E:Z) is observed. The same product
ratios are observed when (PPh3)3RhD is used as the Rh-D
source, supporting the conclusion that Et3SiD/H follows a
Rh(I)-H pathway (step A, Figure 7).
A noteworthy difference between the proposed syn- and
anti-elimination pathways is that significant charge buildup
is expected in the anti-elimination pathway en route to the
product ion pair. No such charge separation is expected in
the syn-β-Cl-elimination pathway. Support for this model
was obtained by switching to a more polar solvent. Reacting
cis-DCE with Et3SiD and (PPh3)3RhCl in tetrahydrofuran-
d8 resulted in a product ratio of 1:7 E- to Z-1-chloro-2-
deuteroethylene, contrasting the 1:1 ratio observed in C6D6.
We propose that the shift to increased amounts of the anti-
elimination product (Z-1-chloro-2-deuteroethylene) is due
to better stabilization of the anti-elimination transition state
with the more polar solvent.
The difference in the E:Z ratio of the 1-chloro-2-deutero-
ethylene products starting from either the cis-DCE or trans-
DCE is puzzling and informative. From a Newman projection
analysis, the same three rotomeric conformers are available
starting from insertion of cis- or trans-DCE into a Rh-D
bond (see Supporting Information). If bond rotation inter-
converting theserotomers is fast comparedto elimination, this
analysis suggests that cis- and trans-DCE should give the same
ratio of products. We conclude that the converse must be true,
that elimination is competitive with C-C bond rotation. The
kinetic product mixture is observed, as the rotomers do not
have time to fully equilibrate.78
Occurrence of Hydrogenation/Hydrosilylation Products.
Hydrogenation (1,2-addition) products are observed only
in the H2 catalytic system, providing further support for the
involvement of a Rh(III)-H pathway with H2. Hydrogena-
tion products (dichloroethane, chloroethane, fluoroethane,
chlorofluoroethane, and ethane) were observed during the
degradation of various halogenated ethylenes with H2 and
(PPh3)3RhCl, supporting the involvement of a Rh(III)-alkyl
complex in this catalytic system, as reductive elimination is
not favorable from a Rh(I) center. Therefore, observation of
hydrogenation products supports the assertion that hydro-
genation occurs from a Rh(III) complex.
No 1,2-addition (hydrosilylation) products were observed
during the degradation of halogenated ethylene when Et3SiH
was used as the reducing agent. Hydrosilylation products
would arise from insertion of the alkene into the Rh-SiEt3
bond followed by reductive elimination to form Et4Si. By
contrast, Braun et al. recently reported hydrosilylation pro-
ducts in the degradation of hexafluoropropene catalyzed by
a triethylphosphine rhodium complex.56
The mixture of syn- and anti-elimination products with
Et3SiH rules out the possibility that it reacts by the same
pathway as H2. Furthermore, spectroscopic evidence shows
a mixture of Rh(III)-H and Rh(I)-H species with Et3SiH and
only Rh(III)-H with H2. Taken together, the data suggest
that β-haloalkyl complexes of Rh(III) give exclusively syn-
elimination products, while β-haloalkyl complexes of Rh(I)
can give both syn- and anti-elimination products (eq 9). This
is based on experiments with (PPh3)3RhD and cis-DCE that
show a 1:1 mixture of E-1-chloro-2-deuteroethylene and Z-
1-chloro-2-deuteroethylene. In this system, it seems unlikely
that Rh(III) species are involved and that both products
come from the same dichloroethylrhodium(I) intermediate.
Vinyl Fluoride Products. Differences in the products ob-
served when (PPh3)3RhCl reacts with vinyl fluoride in the
presence of Et3SiH and H2 provide further support for
different degradation pathways with the two reducing
agents. When Et3SiH, (PPh3)3RhCl, and vinyl fluoride were
reacted, the major degradation product is ethylene. How-
ever, when vinyl fluoride is reacted with H2 and (PPh3)3-
RhCl, the major degradation product was fluoroethane, and
We speculate that the nucleophilicity of the metal center
enhances the favorability of the anti-elimination pathway
with Rh(I) over Rh(III). Pseudobackside displacement of the
(78) Another possibility is that conformer interconversion is fast but
that the position of the conformer equilibrium is perturbed by the site of
deuteration.

Article
Organometallics, Vol. 28, No. 20, 2009 5991
the minor degradation product was ethylene. With Et3SiH,
the active Rh(I)-H would favor β-F-elimination, giving
ethylene as the product (step A, Figure 7). With H2, the
active Rh(III)-H allows for either β-F-elimination (produc-
ing ethylene, step B, Figure 8) or reductive elimination
(producing fluoroethane, step C, Figure 8) after the insertion
of the vinyl fluoride into the Rh-H bond.
Other Mechanistic Observations. In the Et3SiH system, all
of the dichloroethylenes are degraded at a similar rate
(Table 1), while in the H2 system, cis-DCE is degraded more
readily than trans-DCE and 1,1-DCE (Table 2). A possible
explanation for this difference in reactivity is that in the H2
system the active catalyst, Rh(III)-H, is more sterically
crowded and more selective in terms of alkene association.
with an insertion/β-chloride elimination mechanism, but
due to distinct differences in the two systems, it is proposed
that a Rh(I) pathway is favored for Et3SiH and a Rh(III)
pathway is favored for H2. From the Rh(I) pathway both
syn- and anti-β-halo-elimination can occur, and from the
Rh(III) pathway, only syn-β-halo-elimination occurs. The
difference in reactivity between Et3SiH and H2 leads to
different stereochemistry of elimination products and dif-
ferent propensities for giving 1,2-addition products. There-
fore, direct comparisons between H2 and Et3SiH in this
system need to be considered carefully.
Acknowledgment. This work was supported by the
National Science Foundation (CHE-0239461, CHE-
0809575). NMR instrumentation was provided with
funds from the National Science Foundation (BIR-
961477), the University of Minnesota Medical School,
and the Minnesota Medical Foundation.
Conclusions
Catalytic dechlorination of chlorinated ethylenes by
(PR3)3RhCl occurs with Et3SiH and H2 as reducing agents.
These are therefore effective catalytic systems for the
degradation of an important class of environmental con-
taminants. Rhodium hydride complexes are the active
intermediates for the degradation of the halogenated ethy-
lenes studied. The use of both Et3SiH and H2 is consistent
Supporting Information Available: Newman projections of
dichloroalkylrhodium complexes from cis- and trans-DCE,
representative halogenated ethylene degradation plots. This
material is available free of charge via the Internet at http://
pubs.acs.org.