Reactions of the schiff base rhodium hydride complex RhH(Bu4salophen) with olefins. Observations and mechanistic studies

Article abstract of DOI:10.1021/om950595k

The tetradentate dianionic Schiff base ligand Bu₄salophenH₂ reacts with [RhCl(C₂H₄)₂]₂ and NR₄OH (R = n-Bu, Et) to produce the complexes RhR(Bu₄salophen) (1, R = n-Bu; 2, R = Et), which undergo photolysis (λ > 475 nm) under a hydrogen atmosphere to generate RhH(Bu₄salophen) (3) and the corresponding alkanes. In benzene at room temperature, olefins insert into the metal-hydride bond of 3 to produce Rh(CH2_CH₂R)(Bu₄salophen) (2, R = Et; 4, R = Ph; 5, R = CN). In a competition experiment, insertion of styrene is only slightly favored over that of ethylene. For olefinic substrates capable of radical rearrangement (CH₂=CHR where R = CH₂OCH₂CH=CH₂ and c-C₃H₅), insertion products are formed that do not correspond to radical rearrangements. Specifically, Rh(CH₂CH₂OCH₂CH=CH₂)(Bu ₄salophen) (6) is observed as the only product of the reaction of 3 with bis(allyl) ether while Rh(CH₂CH₂-C-C₃H₅)(Bu ₄Salophen) (7) is seen as the major product in the reaction with vinylcyclopropane. For the vinylcyclopropane reaction, additional products are also identified by 2D NMR methods as Rh(CH₂CH₂CH=CHCH₃)(Bu₄salophen) (8) and (Bu₄salophen)Rh-(λ-CH₂CH₂CH=CHCH ₂)Rh(Bu₄Salophen) (9). Both the poor selectivity for styrene versus ethylene insertion and the absence of radical rearranged product with bis(allyl) ether and vinylcyclopropane indicate that, in the reaction of 3 with olefins, an intermediate containing a pure carbon-based radical is not involved.

Full text of DOI:10.1021/om950595k

Organometallics 1996, 15, 1697-1706
1697
Rea ction s of th e Sch iff Ba se Rh od iu m Hyd r id e Com p lex
Rh H(Bu 4sa lop h en ) w ith Olefin s. Obser va tion s a n d
Mech a n istic Stu d ies
D. J oe Anderson and Richard Eisenberg*
Department of Chemistry, University of Rochester, Rochester, New York 14627
Received J uly 31, 1995X
The tetradentate dianionic Schiff base ligand Bu4salophenH2 reacts with [RhCl(C2H4)2]2
and NR4OH (R ) n-Bu, Et) to produce the complexes RhR(Bu4salophen) (1, R ) n-Bu; 2, R
) Et), which undergo photolysis (λ > 475 nm) under a hydrogen atmosphere to generate
RhH(Bu4salophen) (3) and the corresponding alkanes. In benzene at room temperature,
olefins insert into the metal-hydride bond of 3 to produce Rh(CH2CH2R)(Bu4salophen) (2,
R ) Et; 4, R ) Ph; 5, R ) CN). In a competition experiment, insertion of styrene is only
slightly favored over that of ethylene. For olefinic substrates capable of radical rearrange-
ment (CH2dCHR where R ) CH2OCH2CHdCH2 and c-C3H5), insertion products are formed
that do not correspond to radical rearrangements. Specifically, Rh(CH2CH2OCH2CHdCH2)(Bu4-
salophen) (6) is observed as the only product of the reaction of 3 with bis(allyl) ether while
Rh(CH2CH2-c-C3H5)(Bu4salophen) (7) is seen as the major product in the reaction with
vinylcyclopropane. For the vinylcyclopropane reaction, additional products are also identified
by 2D NMR methods as Rh(CH2CH2CHdCHCH3)(Bu4salophen) (8) and (Bu4salophen)Rh-
(µ-CH2CH2CHdCHCH2)Rh(Bu4salophen) (9). Both the poor selectivity for styrene versus
ethylene insertion and the absence of radical rearranged product with bis(allyl) ether and
vinylcyclopropane indicate that, in the reaction of 3 with olefins, an intermediate containing
a pure carbon-based radical is not involved.
In tr od u ction
Similar radical and radical chain mechanisms have
been proposed by Wayland to explain the versatile
reactivity of related rhodium(II) porphyrin systems. In
1981, Wayland and co-workers reported the synthesis
and isolation of an unusually stable formyl complex Rh-
(CHO)(OEP) produced by the carbonylation of [Rh-
(OEP)]2 in the presence of H2.5-11 Halpern and co-
workers performed a kinetic study of this CO insertion,
as well as of the styrene insertion, and a radical chain
mechanism was implicated for both.4 In subsequent
studies of the metalloradical character of rhodium(II)
porphyrin complexes, Wayland reported the extraordi-
nary activation of the CH bond in methane by Rh(TMP)
(TMP ) tetramesitylporphyrin) under extremely mild
conditions in hydrocarbon solvent.12 The same Rh-
(porphyrin) complexes were found to attack benzylic CH
bonds in alkylarenes as well.13-15 The observed chem-
istry was explained by cooperative attack of two Rh(II)
Olefin insertion into a metal-hydride bond plays a
fundamental role in many organometallic reactions.
This insertion is thought to play a key role in metal-
catalyzed olefin hydrogenation, hydroformylation, and
possibly hydrosilylation and hydrocyanation. The prin-
cipal mechanism proposed for olefin insertion involves
olefin binding to a coordinatively unsaturated metal
hydride to generate an intermediate in which the olefin
and hydride ligands occupy mutually cis positions about
the metal center; the insertion reaction then proceeds
through a four-center transition state.1 Evidence exists,
however, for olefin insertion by metal hydride complexes
in which no such open site is available or in which an
open site appears to be of lesser importancesi.e., trans
to the hydride ligand.1 Specifically, the bis(dimethyl-
glyoximato)cobalt and -rhodium systems first investi-
gated by Schrauzer and co-workers were found to
exhibit reversible olefin insertion,2,3 while styrene inser-
tion into the rhodium hydride bond of RhH(OEP) (OEP
) octaethylporphyrin) was observed by Halpern and co-
workers.4 In these systems, the coordination sites cis
to the hydride ligand were occupied by the chelating
dimethylglyoxime or macrocyclic OEP ligands. The
Halpern study invoked a radical mechanism to achieve
the observed reaction chemistry.
(5) Wayland, B. B.; Woods, B. A. J . Chem. Soc., Chem. Commun.
1981, 700-701.
(6) Wayland, B. B.; Duttaahmed, A.; Woods, B. A. J . Chem. Soc.,
Chem. Commun. 1983, 142-143.
(7) Wayland, B. B.; Sherry, A. E.; Coffin, V. L. In Homogeneous
Transition Metal Catalyzed Reactions; W. R. Moser, W. R., Slocum, D.
W., Eds.; American Chemical Society: Washington, DC, 1992; Vol. 230;
pp 249-259.
(8) Wayland, B. B.; Coffin, V. L.; Sherry, A. E.; Brennen, W. R. ACS
Symp. Ser. 1990, No. 428, 148-158.
(9) Wayland, B. B.; Van Voorhees, S. L.; Wilker, C. Inorg. Chem.
1986, 25, 4039-4042.
(10) Farnos, M. D.; Woods, B. A.; Wayland, B. B. J . Am. Chem. Soc.
1986, 108, 3659-3663.
X Abstract published in Advance ACS Abstracts, February 1, 1996.
(1) Crabtree, R. H. The Organometallic Chemistry of the Transition
Metals, 2nd ed.; J ohn Wiley & Sons, Inc.: New York, 1994.
(2) Schrauzer, G. N.; Holland, R. J . J . Am. Chem. Soc. 1971, 93,
1505-1506.
(11) Wayland, B. B.; Woods, B. A.; Pierce, R. J . Am. Chem. Soc. 1982,
104, 302-303.
(12) Sherry, A. E.; Wayland, B. B. J . Am. Chem. Soc. 1990, 112,
1259-1261.
(3) Weber, J . H.; Schrauzer, G. N. J . Am. Chem. Soc. 1970, 92, 726-
727.
(13) Wayland, B. B.; Del Rossi, K. J . J . Organomet. Chem. 1984,
276, C27C30.
(4) Paonessa, R. S.; Thomas, N. C.; Halpern, J . J . Am. Chem. Soc.
1985, 107, 4333.
(14) Del Rossi, K. J .; Wayland, B. B. J . Am. Chem. Soc. 1985, 107,
7941-7944.
0276-7333/96/2315-1697$12.00/0 © 1996 American Chemical Society

1698 Organometallics, Vol. 15, No. 6, 1996
Anderson and Eisenberg
centers on the alkyl C-H bond to split it homolytically
under ambient temperature conditions.15 Further stud-
ies of this reaction system utilized two rhodium por-
phyrins linked by an alkoxy chain in order to modify
the interaction of the rhodium centers with methane
from a termolecular process to a bimolecular one, and
in keeping with the proposed mechanism, this decrease
in molecularity boosted the rate of homolytic methane
CH activation.16 Similar radical and radical chain
mechanisms have been used to explain other reactions
of rhodium porphyrins with hydrogen, olefins, carbon
monoxide, alkyl isocyanides, trialkylsilanes, and tri-
alkylstannanes.11,17,18
corresponding rhodium hydride species, and a study of
olefin addition to this hydride species.
Resu lts a n d Discu ssion
Syn th esis of Rh R(Bu 4sa lop h en ). Reaction of the
chloro-bridged Rh(I) dimer [Rh(µ-Cl)(C2H4)2]2 with the
free Bu4salophenH2 Schiff base ligand and excess tet-
rabutylammonium hydroxide affords a dark purple
solution that yields a like colored precipitate upon
solvent removal (eq 1). The resultant compound (1) is
We have recently sought to extend the diverse reac-
tion chemistry observed for these porphyrin systems to
rhodium complexes containing derivatives of the dian-
ionic tetradentate Schiff base ligand salophen. Schiff
base ligands lend themselves readily to structural
modifications that offer considerable control over com-
plex solubility and reactivity. Such control, for example,
has been demonstrated over the last few years with
great success in the use of manganese Schiff base
complexes to catalyze olefin epoxidationsa reaction
previously observed for the analogous manganese por-
phyrin complex.19,20 Prior to our initial communication
on the RhR(Bu4salophen) complex I and a related report
air stable and light sensitive in the solid state, and
forms bright red-purple solutions in acetone, ether,
THF, methylene chloride, ethanol, methanol, and hy-
drocarbon solvents. It is insoluble in, but stable to,
deoxygenated water.
The coordinated Schiff base ligand of 1 exhibits a
1
number of characteristic signals by H and 13C NMR
spectroscopy, the most prominent of which are two
singlets in the H NMR spectrum for each of two sets
by Wayland,21,22 most tetradentate Schiff base com-
plexes of rhodium were reported as soluble only in donor
solvents.23-31 In this article, we describe in full the
synthesis and characterization of I for R ) n-Bu and
Et, photolysis of I under hydrogen to produce the
1
of tert-butyl groups at δ 1.99 (18 H) and 1.45 (18 H) ppm.
All other ligand resonances fall much further downfield.
The resonance furthest downfield, at δ 8.40 ppm (2 H),
is assignable to the imine protons and shows a small
splitting (2 Hz) not seen in the free ligand due to 3J Rh-H
coupling. Two signals at δ 7.86 (2 H) and 7.33 (2 H)
show a through-ring H-H coupling of 2 Hz and are
assigned to the aromatic protons adjacent to the tert-
butyl groups. The remaining two 1H NMR signals
belonging to the Schiff base ligand appear as doublets
of doublets, one at δ 7.13 (2 H) and the other at 6.86
(15) Wayland, B. B.; Ba, S.; Sherry, A. E. J . Am. Chem. Soc. 1991,
113, 5305-5311.
(16) Zhang, X.-X.; Wayland, B. B. J . Am. Chem. Soc. 1994, 116,
7897-7898.
(17) Poszmik, G.; Carroll, P. J .; Wayland, B. B. Organometallics
1993, 12, 3410-3417.
(18) Mizutani, T.; Uesaka, T.; Ogoshi, H. Organometallics 1995, 14,
341-346.
(19) J acobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J . R.; Deng, L.
J . Am. Chem. Soc. 1991, 113, 7063-7064.
4
ppm (2 H; 3J HH ) 6 Hz, J HH ) 3 Hz), corresponding to
(20) Zhang, W.; Loebach, J . L.; Wilson, S. R.; J acobsen, E. N. J . Am.
Chem. Soc. 1990, 112, 2801-2803.
the four aromatic protons of the phenylenediimine
1
moiety. Three sets of H NMR resonances that belong
(21) Bunn, A. G.; Wei, M.; Wayland, B. B. Organometallics 1994,
13, 3390-3392.
to the complex but are not assignable to the Schiff base
ligand appear at δ 2.38 (complex multiplet, 2 H), 0.84-
0.95 (broad multiplet, 4 H), and 0.44 ppm (triplet, 3 H).
These resonances correspond to the butyl ligand and
exhibit a pattern similar to that of the butyl group in
n-Bu4N+. The R-methylene protons of the coordinated
butyl group are shown in Figure 1a. Their unique
pattern serves as an important guide to the identifica-
tion and characterization of other Rh(III) alkyl com-
plexes generated in subsequent reactions between RhH-
(Bu4salophen) and olefins.
(22) Anderson, D. J .; Eisenberg, R. Inorg. Chem. 1994, 33, 5378-
5379.
(23) Levitin, I.; Sigan, A. L.; Vol’Pin, M. E. J . Chem. Soc., Chem.
Commun. 1975, 469-470.
(24) Nikitaev, A. T.; Sigan, A. L.; Levitin, Y.; Belyi, A. A.; Vol’Pin,
M. E. Inorg. Chim. Acta 1982, 64, L29-L31.
(25) Aleinikova, V. N.; Vasil’ev, V. V.; Shagisultanova, G. A. Russ.
J . Phys. Chem. 1986, 60, 935-937.
(26) Aleinokova, V. N.; Vasil’ev, V. V.; Timonon, A. M.; Shagisul-
tanova, G. A. Russian J . Phys. Chem. 1987, 61, 580-582.
(27) Calmotti, S.; Pasini, A. Inorg. Chim. 1984, 85, L55-L56.
(28) Cozens, R. J .; Murray, K. S.; West, B. O. J . Chem. Soc., Chem.
Commun. 1970, 1262-1263.
(29) Cozens, R. J .; Murray, K. S.; West, B. O. J . Organomet. Chem.
1972, 38, 391-402.
All of the 13C NMR spectra reported herein were
recorded using the J -modulated spin-echo pulse se-
quence, also known as the attached proton test (APT),
in which signals for carbons with an even number of
(30) Rogers, C. A.; West, B. O. J . Organomet. Chem. 1974, 70, 445-
453.
(31) Van Den Bergen, A. M.; Elliott, R. L.; Lyons, C. J .; MacKinnon,
K. P.; West, B. O. J . Organomet. Chem. 1985, 297, 361-370.

Reactions of RhH(Bu4salophen) with Olefins
Organometallics, Vol. 15, No. 6, 1996 1699
F igu r e 2. Electronic spectra of Rh(n-Bu)(Bu4salophen)
(1): (a) absorption spectrum at 298 K in benzene; (b)
emission spectrum in toluene glass at 77 K, 510 nm
excitation wavelength; (c) excitation spectrum in toluene
glass at 77 K, emission monitored at 670 nm.
previously by Aleinikova et al.,25 and an analysis of the
emitting state in 1 will be presented elsewhere in
conjunction with a discussion about related luminescent
compounds.32
Use of tetraethylammonium hydroxide in place of
tetrabutylammonium hydroxide in the above reaction
gives a similar dark purple product which is distinct
by 1H NMR spectroscopy from 1 only in the appearance
of NMR signals at δ 2.37 (qd, J RhH ) 3.2 Hz, J HH ) 7.4
Hz) and 0.23 (td, J RhH ) 1.1 Hz, J HH ) 7.4 Hz) and the
absence of the n-butyl group signals. The two reso-
nances unique to this product are assigned to a rhodium-
bound ethyl group in RhEt(Bu4saloph) (2) (eq 1, R )
Et). Formation of the ethyl complex in this reaction
indicates that the tetralkylammonium ion serves as the
source of the alkyl groups in compounds 1 and 2.18,33
Rea ctivity of 1 a n d 2. Benzene solutions of 1 are
stable at room temperature for several days in the dark,
even in the presence of hydrogen or carbon monoxide.
However, in situ photolysis of a benzene-d6 solution of
1 under hydrogen with long-wavelength visible light (λ
> 475 nm) generates a red solution which exhibits a
F igu r e 1. 1H NMR (400 MHz) resonances for rhodium
bound methylene units: (a) [Rh]nBu (1); (b) [Rh]CH2-
CH2C6H5 (4); (c) product from allyl ether addition to 3,
[Rh]CH2CH2CH2OCH2CHdCH2 (6); (d) products from vi-
nylcyclopropane addition to 3, [Rh]CH2CH2(c-C3H5) (7) (left
multiplet) and [Rh]CH2CH2CHdCHCH3 (8) (right multi-
plet); (e) product from trans-1,3-pentadiene addition, [Rh]-
CH2CH2CHdCHCH3 (8). Peak “x” is from the â-protons of
4, while “y” is from unreacted starting material 1.
attached protons have a phase opposite to signals for
carbons with an odd number of attached protons. The
13C NMR spectrum of the complex shows resonances
assignable to all of the carbon atoms of the Schiff base
ligand as well as to the four distinct butyl carbons at δ
34.36, 25.14 (1J RhC ) 31 Hz), 22.14 (2J RhC ) 2 Hz) and
13.65 ppm. The phase of the first three carbon signals
indicates that each carbon has an even number of
attached protons, whereas the last has an odd number
of attached protons. It is thus possible from the NMR
spectroscopic data to identify unequivocally the product
of eq 1 as Rh(n-Bu)(Bu4salophen) (1). A crystal struc-
ture analysis, obtained using crystals grown from a
benzene/methylene chloride solution and described in
detail elsewhere,22 confirms our assignment of 1 and
shows that it possesses a square pyramidal geometry,
with the n-butyl group occupying the apical position and
the four Schiff base donor atoms comprising the base
of the pyramid about Rh(III).
Complex 1 exhibits a weak luminescence in fluid
solution that intensifies upon freezing the solution to
77 K. The low-temperature emission maximum at 671
nm is accompanied by a shoulder at 745 nm. Features
in the low-temperature excitation spectrum of this
compound closely parallel features in the room-temper-
ature absorbance spectrum. Figure 2 shows the low-
temperature emission and excitation spectra for 1 in
toluene glass (77 K) as well as the room-temperature
absorption spectrum in benzene. Luminescence from
Rh(III) alkyl Schiff base complexes has been reported
1
broad featureless H NMR signal at δ -25.8 ppm (1 H)
in addition to resonances assignable to the coordinated
Bu4salophen ligand and to free n-butane in solution
(multiplets at δ 1.23 ppm and 0.86 ppm). Production
of n-butane is confirmed by GC/MS analysis of the
volatiles from this solution. Photolysis of 2 under a
hydrogen atmosphere also produces the characteristic
1H NMR signal at δ -25.8 ppm (1 H) and a sharp singlet
at 0.80 ppm assignable to ethane. In the presence of
trace amounts of donor solvents in benzene-d6, the
1
upfield H NMR resonance resolves into a broad doublet
with 51 Hz coupling, while photolysis of 1 in THF-d8
leads to a sharp doublet at δ -22.52 ppm with 37 Hz
1
coupling. The presence in these H NMR spectra of a
signal in the metal-hydride region along with produc-
tion of the alkane derived from 1 and 2 is consistent
with homolytic metal-alkyl bond hydrogenolysis and
formation of the hydride product RhH(Bu4salophen) (3),
eq 2. When a solution of 3 is placed under an atmo-
sphere of D2, immediate formation of HD (δ 2.42 ppm,
J HD ) 43 Hz) accompanies the disappearance of the
(32) Anderson, D. J .; Eisenberg, R. To be submitted for publication.
(33) Collman, J . P.; Brauman, J . I.; Madonik, A. M. Organometallics
1986, 5, 215-218.

1700 Organometallics, Vol. 15, No. 6, 1996
Anderson and Eisenberg
For the reaction of 3 with ethylene, the dark purple
product is identified as 2 by the characteristic splitting
pattern (vide supra) of the ethyl group seen in the
1
product’s H NMR spectrum along with the Schiff base
resonances. However, for olefins other than ethylene,
unambiguous identification of the reaction product
necessitated further spectroscopic characterization. In
each case, a 13C spectrum was acquired which revealed
all Bu4salophen resonances (i.e., nine aromatic carbons
at δ ∼168-120 ppm, the imine carbon at δ ∼115 ppm,
two tert-butyl quaternary carbons at δ ∼37 and ∼34
ppm, and two tert-butyl methyl carbons at δ ∼32 and
∼30 ppm) as well as those of the alkyl ligand generated
upon olefin insertion.
(2)
The reaction of 3 with styrene leads to the observation
1
of two complex H NMR multiplets in addition to those
upfield resonance at δ -25.8 ppm within minutes. The
production of HD and the disappearance of the hydride
resonance is consistent with the formation of RhD(Bu4-
salophen) (3-d1) through rapid, thermally promoted
exchange between D2 and RhH(Bu4salophen), eq 3.
of the Bu4salophen ligand. The coupling pattern of the
multiplet at δ 2.47 ppm (2 H, Figure 1b) resembles that
of the rhodium-bound methylene unit in 1 and is
therefore assigned to the protons of an analogous
rhodium-bound methylene carbon. The remaining com-
plex multiplet appears as a distorted triplet at δ 2.29
ppm (2 H). Homonuclear decoupling of either of these
resonances collapses the other into a broad singlet. By
13C NMR spectroscopy, two signals are found at δ 25.22
(J RhC ) 25.2 Hz) and δ 38.55 (6 Hz) ppm that correspond
to the carbon atoms of a â-phenylethyl group in the Rh-
(CH2CH2Ph)(Bu4salophen) insertion product (4). Con-
sistent with this formulation, the 13C NMR spectrum
shows six peaks in the aromatic region assignable to
the phenyl group. A series of overlapping 1H NMR
multiplets from δ 6.67 to 6.79 ppm (5 H) were assigned
to the aromatic protons of this phenyl group. These
resonances, along with the multiplet at δ 2.47 ppm, are
1
not seen in the H NMR spectrum of the product which
results from the addition of styrene-d8 to 3, although a
broad singlet could still be seen at δ 2.29 ppm (1 H)
corresponding to the former hydride ligand after inser-
tion.
Although solutions of the hydride complex 3 are air-
sensitive and labile in the absence of a dihydrogen
atmosphere, the alkyl complexes are not. The thermal
insertion of styrene was therefore used to trap pho-
tolytically-generated hydride 3 and to monitor the
progress of preparative-scale photolysis. During the
course of the photolysis, the photolysis vessel was
temporarily removed from the photon flux of the lamp
while samples were withdrawn under H2 into an NMR
tube containing excess styrene. NMR spectra of these
samples showed unreacted n-Bu complex 1 and the
â-phenylethyl complex 4 from the thermal reaction of
hydride 3 and styrene. The relative amounts of these
complexes in the NMR sample were taken as indication
of the progress of the photolysis in the bulk solution. If
the NMR spectra indicated that starting material
remained, photolysis of the bulk solution was continued.
When NMR analysis of samples taken from the bulk
solution showed that all of the starting n-Bu complex 1
was gone, photolysis was halted and excess styrene was
added to the bulk solution to produce Rh(CH2CH2Ph)-
(Bu4salophen) (4) in ∼100 mg quantities.
Rea ction of Rh H(Bu 4sa lop h en ) (3) w ith Olefin s.
Benzene solutions of 3 generated in situ react readily
at room temperature with a number of olefins to form
the corresponding alkyl product as shown in eq 4 for R
1
) H, C6H5, and CN. Inspection of the H NMR signals
In the reaction of hydride 3 with acrylonitrile, chemi-
for the Schiff base ligand in each of these reactions
indicates that in every case a single diamagnetic product
with approximate Cs or mirror symmetry predominates.
1
cal shifts within the H NMR spectrum of the product
were highly dependent on the constitution of the solu-

Reactions of RhH(Bu4salophen) with Olefins
Organometallics, Vol. 15, No. 6, 1996 1701
Sch em e 1
For example, in tin hydride additions to the selectively
deuterated olefins styrene, 1-hexene, and cis- and trans-
2-butene, Kuivila and Sommer found that initial attack
of Bu3Sn• on hexene and butene was reversible whereas
attack of the tin radical on styrene was irreversible.34
For radically-promoted hydrothiolation of olefins, Sivertz
and co-workers observed that, for RS• addition, ka-
(styrene) is much greater than ka(1-pentene) but, for the
hydrogen transfer step, ktr(1-pentene) exceeds ktr(sty-
rene).35 In each of these cases, benzylic stabilization of
either the transition state for X• addition to the olefin
or the carbon-based radical intermediate accounted for
the differences in olefin reactivity. For example, ben-
zylic stabilization of Bu3SnCH2CHPh hindered dissocia-
tion to styrene and Bu3Sn• leading to the observed
irreversibility, while, for thiol addition, benzylic stabi-
lization of the RS• addition product led to a slower
H-atom transfer step.
Sch em e 2
In this context, we sought evidence for participation
of metal alkyl radical intermediates in the addition of
olefins to RhH(Bu4salophen) (3). To that end, a com-
petition reaction between styrene and ethylene with 3
was conducted. When a 4:1 ratio of ethylene to styrene
was introduced to a benzene solution of hydride 3, the
styrene addition product 4 was favored over the ethyl-
ene product 2 by only a 2:1 ratio, yielding an overall
preference of 8:1 for styrene over ethylene.
tion. Complex multiplets assignable to the methylene
units of a â-cyanoethyl ligand in Rh(CH2CH2CN)(Bu4-
salophen) (5) were identified at δ 1.94 and 2.12 ppm in
the presence of excess acrylonitrile. However, removal
of all solvent and acrylonitrile from this solution under
vacuum and introduction of fresh solvent shifted these
resonances to positions where they were obscured by
large signals from the tert-butyl groups of the Schiff base
ligand. In neither case was accurate integration pos-
sible. Even so, a cross-peak in the 1H-1H COSY
spectrum of this product at (δ 1.81 ppm, 1.66 ppm)
confirmed the adjacency of the two groups giving rise
to these signals, while two resonances at δ 12.73 (J RhC
) 32.6 Hz) and 17.61 ppm in the 13C NMR spectrum of
5 had phases and chemical shifts consistent with two
methylene groups, one of which was directly bound to
rhodium.
For the mechanism of Rh-H addition to olefins via
Scheme 2, two limiting cases can be considered. In the
first, the Rh• addition step is fast with subsequent
H-atom transfer slow, whereas, in the second, the
relative rates of the two steps are reversed. For the first
of these limiting cases, ethylene addition would be
expected to be more favorable whereas for the second
styrene addition should predominate. The relatively
meager 8:1 preference that is observed for styrene
addition to 3 (a ∆∆G‡ of <0.9 kcal) suggests that if
Scheme 2 is operative, the competition is not determined
solely by one or the other of the mechanism’s steps but
rather that both steps influence the observed preference.
Alternatively, and in keeping with the observations
described below, we think that the 8:1 styrene:ethylene
preference indicates that the species resulting from
metalloradical addition to the olefin is not adequately
described as the C-based radical B and instead that
stabilization from the metal center affects the reactivity
of the Rh• addition species for both of the olefins.
Rea ction of 3 w ith Bis(a llyl) Eth er . Further
investigation of the possible radical nature of olefin
addition to hydride 3 utilized the well-studied rear-
rangements of certain unsaturated organic radicals.
These rearrangements have been used to “clock” com-
petitive reaction pathways for both organic and orga-
nometallic systems.36-39 For olefin addition to 3 by the
radical chain pathway of Scheme 1, the species subject
to rearrangement is produced by the addition of the
metalloradical A to the olefin. Subsequent H atom
transfer to B then competes with the rearrangement so
Mech a n istic P r obes of Olefin In ser tion : Com -
p etitive Eth ylen e a n d Styr en e Ad d ition . The re-
gioselectivity observed in the addition of styrene and
acrylonitrile to the hydride complex 3 is consistent with
the radical chain mechanism proposed by Halpern for
the addition of styrene to RhH(OEP).4 In the propaga-
tion steps of this mechanism, shown with the overall
reaction in Scheme 1, olefin addition is catalyzed by a
Rh(II) radical species A that binds olefin to yield a metal
alkyl radical B having significant unpaired spin on the
carbon atom â to the rhodium center. The observed
kinetics of the addition are consistent with a facile
equilibrium yielding radical B that in turn abstracts a
H atom irreversibly from the starting Rh(III) hydride
to form the olefin insertion product while regenerating
the chain-propagating Rh(II) species A.
Addition of styrene to the Rh-H bond of RhH(OEP)
as formulated by Halpern and co-workers belongs to the
more general class of XY addition to olefins by the
radical chain pathway shown in Scheme 2. The litera-
ture provides numerous examples of such additions in
which X ) halide, N, P, RS, R3Si, R3Ge, and R3Sn and
Y ) H.33 Typically for these reactions, addition of XY
to styrene contrasts markedly with that to olefins
lacking aromatic substituents.
(34) Kuivila, H. G.; Sommer, R. J . Am. Chem. Soc. 1967, 89, 5616-
5619.
(35) Sivertz, C. J . Phys. Chem. 1959, 63, 34-38.
(36) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317-323.
(37) Newcomb, M. Tetrahedron 1993, 49, 1151-1176.
(38) Bullock, R. M.; Samsel, E. G. J . Am. Chem. Soc. 1990, 112,
6886-6898.
(39) Hembre, R. T.; Ramage, D. L.; Scott, C. P.; Norton, J . R.
Organometallics 1994, 13, 2995-3001.

1702 Organometallics, Vol. 15, No. 6, 1996
Anderson and Eisenberg
Sch em e 3
that the ratio of rearranged to unrearranged product
indicates the relative rates of hydrogen atom transfer
and radical rearrangement as shown in Scheme 3 for
bis(allyl) ether. For bis(allyl) ether, the rearrangement
corresponds to ring closure that would yield a (4-methyl-
3-tetrahydrofuranyl)methylrhodium derivative. Inde-
pendent studies have shown that the parent organic
radical for C rearranges with a unimolecular rate
constant of k ) 4 × 106 s-1 at 25 °C.37
When excess bis(allyl) ether was added to a benzene-
d6 solution of 3, however, only a single major product
(56%) was identified that corresponded to anti-Mark-
ovnikov addition of the rhodium-hydride bond across
the substrate double bond without rearrangement. The
identity of this product, Rh(CH2CH2CH2OCH2CHd
CH2)(Bu4salophen) (6), was established by 1H NMR
spectroscopy, and, in particular, by Schiff base ligand
resonances and a characteristic rhodium-bound meth-
ylene signal at δ 2.38 ppm (m, 2 H, Figure 1c). The
absence of any rearranged product suggested that on
the basis of Scheme 3, the hydride atom abstraction
reaction proceeded much faster than the rearrangement.
On the basis of the first-order rate constant for the
parent radical cyclization of 4 × 106 s-1, the H atom
abstraction reaction would have to proceed near the
diffusion controlled limit to avoid forming observable
amounts of rearranged product. With a concentration
of ca. 1 × 10-3 M for hydride 3, the second-order rate
constant would have to be ca. 1 × 1010 M-1 s-1 to explain
the observed result using Scheme 3.
F igu r e 3. 1H-1H COSY (500 MHz) spectrum of products
from vinylcyclopropane addition to rhodium hydride 3.
Correlations are for (a) Rh(CH2CH2(c-C3H5))(Bu4salophen)
(7), (b) Rh(CH2CH2CHdCHCH3)(Bu4salophen) (8), and (c)
(Bu4saloph)Rh(CH2CHdCHCH2CH2)Rh(Bu4saloph) (9).
products. A 1H-1H COSY spectrum of this reaction
mixture, acquired using a double-quantum filter to
reduce near-diagonal effects of the large tert-butyl
singlets, led to assignment of the two major products.
As shown in Figure 3a, a COSY walk originating at the
rhodium-bound methylene signal at δ 2.42 ppm proceeds
through several cross peaks located at (2.42, 0.95 ppm),
(0.95, 0.23 ppm), (0.23, -0.058 ppm), (0.23, -0.35 ppm),
and (-0.058, -0.35 ppm). The two multiplets at δ
-0.058 and -0.35 ppm have coupling patterns which
are very similar to those seen for the cyclopropyl
methylene groups in vinylcyclopropane, although shifted
significantly upfield. On the basis of this similarity,
Rea ction of 3 w ith Vin ylcyclop r op a n e. In order
to test the proposed radical insertion mechanism of
Scheme 1 more definitively, attention next focused on
the cyclopropylcarbinyl radical and the use of vinylcy-
clopropane as a substrate in reaction with 3. Since the
rate of ring opening for the cyclopropylcarbinyl radical,
eq 5, is one of the fastest of the standard radical “clocks”
1
together with the rhodium-bound methylene H NMR
with k ) 4 × 107 s -1, reaction of 3 with vinylcyclopro-
pane should yield significant ring opened productsi.e.,
Rh(CH2CHdCHCH2CH3)(Bu4salophen)sif the radical
mechanism of Scheme 1 was followed. Surprisingly,
this turned out not to be the case.
signal and the connectivity information provided by the
COSY spectrum, these resonances are assigned to the
product Rh(2-cylopropylethyl)(Bu4salophen) (7). Inte-
When vinylcyclopropane was reacted with hydride 3,
two major sets of product resonances were observed in
1
the H NMR spectrum, but neither corresponded to the
2-pentenylrhodium complex expected from radical ring
opening. Associated with two sets of Schiff base ligand
resonances were two complex multiplets at δ 2.42 and
2.31 ppm (Figure 1d), which, through comparison with
similar signals in the spectra of compounds 1, 4, and 6
(Figure 1a-c), were assigned to rhodium-bound meth-
ylene units of two separate square pyramidal addition
gration of the two high-field resonances at δ -0.058 and

Reactions of RhH(Bu4salophen) with Olefins
Organometallics, Vol. 15, No. 6, 1996 1703
-0.35 ppm and of the methylene resonance at δ 2.42
salophen) was seen in either 1D or 2D spectra acquired
with this reaction mixture.
1
ppm versus the imine region in the H NMR spectrum
indicated that 7 was the major observable product
formed and accounted for 32% of the total Schiff base
species in the product mixture.
A COSY walk (Figure 3b) originating at the other
rhodium-bound methylene peak, at δ 2.32 ppm, passes
through cross peaks at (2.32, 1.65 ppm), (1.65, 4.90
ppm), and (4.90, 1.16 ppm). These resonances (4.90,
2.32, 1.65, and 1.16 ppm) are assigned to the 3-pentenyl
group of the ring-opened addition product Rh(CH2CH2-
CHdCHCH3)(Bu4salophen) (8), in which the two olefinic
However, an additional product was sometimes seen
in this reaction that did possess an allyl group but was
not Rh(2-pentenyl)(Bu4salophen). In reactions where
this product was present, the yield of the unrearranged
insertion product 7 was reduced by a proportionate
amount. The identity of this compound was established
by careful NMR analysis as the binuclear alkenyl
bridged species (Bu4salophen)Rh(CH2CHdCHCH2CH2)-
Rh(Bu4salophen) (9). This compound exhibited two
proton signals overlap at δ 4.90 ppm. Integration of the
δ 2.32 ppm signal indicated that 8 accounts for 17-24%
of the total Schiff base containing product. Compound
8 is the product expected for addition of hydride 3 across
the terminal double bond of 1,3-pentadiene. Support
for the assignment of 8 was obtained by reacting excess
trans-1,3-pentadiene with hydride 3 in a benzene-d6
solution and confirming that the resonances of 8 are the
same as those of the terminal olefin insertion product.
In the original reactions of vinylcyclopropane with 3,
no 1,3-pentadiene was observed in the initial reaction
system, although afterwards small quantities (trace to
no more than 15% of total olefin) of 1,3-pentadiene could
be seen. Its formation may result from rhodium-
catalyzed isomerization of vinylcyclopropane to penta-
diene,40 or via â-elimination from 8, but in neither case
does it arise from radical-promoted cyclopropylcarbinyl
ring opening.
distinct sets of Schiff base resonances in equal intensity
and shifted slightly upfield from those of other Bu4-
salophen products. The alkenyl bridge of 9 was deter-
mined by a rhodium-bound methylene signal at δ 2.84
1
(2 H, dd 3.5, 8.2 Hz) that according to H-1H COSY
data (Figure 3c) was coupled to two complex olefinic
multiplets at δ 4.56 (1 H) and 4.39 (1 H), which in turn
were coupled to each other and to a signal at δ 0.4 ppm
that was further coupled to another resonance at δ 1.9
ppm. Formation of 9 is a matter of speculation that is
best done on the basis of the mechanistic conclusions
drawn from the reactions of RhH(Bu4salophen) with
various olefins. We address this matter below.
Mech a n istic Con clu sion s Rega r d in g th e Olefin
In ser tion Rea ction s of Rh H(Bu 4sa lop h en ). Reac-
tions of hydride 3 with substrates capable of radical
rearrangement produce insertion products which exhibit
no radical rearrangement. The lack of such rearrange-
ment strongly suggests that Scheme 1 must be modified
as an accurate description of the insertion mechanism.
Specifically, it appears that while Rh(II) species un-
doubtedly play a role in the reaction chemistry, the
generation of a simple carbon-based radical by olefin
addition to Rh(II) seems problematic. If such a radical
were generated, radical-promoted rearrangement should
have occurred, but experimental evidence indicates
otherwise. This is particularly compelling in the reac-
tion of vinylcyclopropane with 3 that gave major prod-
ucts devoid of radical-based rearrangements.
In addition to the two major products, a minor product
was seen which could not be identified reliably due to
1
the proximity of its H NMR signals to residual ridges
of F1 noise in the COSY spectrum. Integration of the
R-methylene 1H resonance of this product at δ 2.24 ppm
showed that it constituted 1-8% of the observed prod-
ucts containing the Bu4salophen ligand.
An essential point of the reaction of vinylcyclopropane
with hydride 3 is that despite the mixture of products,
no major component corresponded to the 2-pentenyl-
rhodium complex expected from the radical promoted
ring-opening of the substrate. While the alkyl (∼0-2
ppm) and aryl (∼6-8 ppm) regions of the 1H NMR
spectrum were quite crowded and dominated by Schiff
base resonances, the spectrum between these regions
(∼2-6 ppm) was by comparison quite clean. Protons
of the rhodium-bound -CH2- group (i.e., RhCH2-
CHdCHCH2CH3) predicted for the ring-opened product
were not observed in this region downfield of the
rhodium-bound methylene protons described above, and
no allylic group assignable to Rh(2-pentenyl)(Bu4-
Support for the view that olefin addition to a Rh(II)
metalloradical does not generate the simple carbon
radical shown as B in Scheme 1 is eloquently provided
by Wayland through both experiment and analysis. In
the low-temperature EPR spectrum of the 13C2H4 adduct
of Rh(TTiPP) (TTiPP ) tetrakis(2,4,6-triisopropylphen-
yl)porphyrin), Wayland observed a doublet of triplets
assignable to hyperfine coupling to rhodium and to both
(40) Khusnutdinov, R. I.; Dzhemilev, U. M. J . Organomet. Chem.
1994, 47, 1-18.

1704 Organometallics, Vol. 15, No. 6, 1996
Anderson and Eisenberg
olefinic carbons, thus providing clear evidence that
unpaired spin density in this adduct is distributed
across both the metal and the olefinic carbon atoms.41,42
Furthermore, in reactions of Rh(II) porphyrins with
olefinic substrates capable of radical polymerization,
such as acrylic acid and methyl acrylate (eq 6 for RdH,
Me), no such polymerization was found although cou-
pling products were generated that were consistent with
initial olefin addition to a metalloradical species.43 An
analysis of radical stability indicated that the addition
of Rh(II) porphyrin to a simple olefin to generate a
C-based radical would be thermodynamically un-
use. Diethyl ether and benzene were distilled under nitrogen
from dark blue or purple solutions of sodium benzophenone
ketyl. Benzene-d6 was vacuum distilled from a potassium
mirror or from sodium benzophenone ketyl. Methanol and
ethanol were refluxed over magnesium and distilled under
nitrogen. All NMR spectra were recorded in benzene-d6
solution, unless otherwise noted. 1H NMR spectra were
recorded on a Bruker WP200, a GE QE-300, a Bruker AMX-
400, or a Varian VXR 500 at 200, 300, 400.13, and 500 MHz,
respectively. 13C NMR spectra were recorded on a Bruker
AMX-400 at 100.62 MHz, using a J -modulated spin echo pulse
sequence, in which quaternary carbons are detected and
appear with the same phase as methylene carbons. Chemical
shifts for 1H NMR spectra are reported relative to TMS but
were referenced from the residual proton impurity peak in
benzene-d6 at δ 7.15 ppm. Chemical shifts for 13C NMR
spectra are also reported relative to TMS and were referenced
favorablesformation of a Rh-C bond (∼50 kcal mol-1
)
does not provide the necessary driving force for breaking
an alkene π bond (∼64 kcal mol-1) which is necessary
to form a C-based radical.43,44 The driving force for
olefin activation by these compounds derives from the
formation of additional bonds. In eq 6, additional Rh-C
and C-C σ bonds contribute to that driving force. It
thus appears that the species generated upon Rh•
addition to olefin in the present study must be a
significantly stabilized radical that is not best described
by that shown in Scheme 1.
to the solvent triplet for benzene-d6 at δ 128.0 ppm (J CD
)
With this background, we can now address the one
reaction that shows evidence of radical rearrangement,
namely the formation of 9 from the reaction of hydride
3 with vinylcyclopropane. We view the formation of 9
as a coupling reaction that leads to ring opening only
when two Rh(II) centers are involved in the reaction
system. Similar coupling reactions were described by
Wayland leading to C2- and C4-bridged products in the
reactions of olefins with Rh(porphyrins), but again no
polymerization products resulting from pure C-based
radicals were seen.41-43,45 In the formation of 9, the
initial vinylcyclopropane addition does not result in ring
opening, but in the presence of a second Rh(II), ring
opening occurs concomitant with formation of the second
Rh-C bond, eq 7. The intermediate shown in eq 7 is a
metal-stabilized radical having a delocalized structure
that is not expected to undergo the traditional radical
rearrangements prior to bimolecular H-atom transfer
or radical coupling reactions.
24.2 Hz). Infrared spectra were recorded on a Mattson Galaxy
6020 spectrometer. Electronic absorption data were recorded
in benzene on a Hitachi U-2000 spectrophotometer in benzene
solution. Luminescence spectra were recorded in toluene at
room temperature and at 77 K in 5 mm glass NMR tubes on
a SPEX Fluorolog-2 fluorimeter equipped with a 450 W Xenon
lamp and a Hamamatsu R929 phtomultiplier tube. Elemental
analyses were performed by Desert Analytics of Tucson, AZ.
Hydrogen (UHP grade) and ethylene (CP grade) were used
as purchased from Air Products. Vinylcyclopropane,46 3,5 di-
48
tert-butylsalicylaldehyde,47 Bu4salophenH2 and [Rh(µ-Cl)-
49
(C2H4)2]2 were prepared according to literature methods.
Other olefins were purchased from Aldrich. Styrene was dried
over calcium hydride and vacuum distilled prior to use. All
other olefins were vacuum transferred prior to use.
Rh (n -Bu )(Bu 4sa lop h en ) (1). A three neck round bottom
flask containing Bu4salophenH2 (1.8675 g, 3.453 mmol) was
fitted with a reflux condenser and a side arm containing 0.6450
g (1.658 mmol) of [Rh(µ-Cl)(C2H4)2]2. This system was flushed
with nitrogen, after which ca. 30 mL of diethyl ether was
added under nitrogen to give a transparent yellow solution of
Bu4salophenH2. This solution turned transparent red-orange
upon addition of 1 M TBA(OH) (Aldrich) in methanol (16.8
mL, 16.8 mmol). Solid [Rh(µ-Cl)(C2H4)2]2 was added slowly
over the course of several minutes to give an opaquely dark,
yet clear, deep green solution. Upon gentle reflux over the
course of 4 h, this solution turned brown, and then deep red-
purple. This purple solution was cooled, reduced in volume
under a stream of nitrogen, filtered, and washed with metha-
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All manipulations were per-
formed under a nitrogen atmosphere, either in a Vacuum
Atmospheres glovebox or using Schlenck techniques, unless
otherwise noted. Solvents were dried and distilled prior to
(41) Bunn, A. G.; Wayland, B. B. J . Am. Chem. Soc. 1992, 114,
6917-6919.
(42) Wayland, B. B.; Sherry, A. E.; Bunn, A. G. J . Am. Chem. Soc.
1993, 115, 7675-7684.
(46) Tsunoda, T.; Hudlicky, T. Syn. Lett. 1990, 322.
(47) Casiraghi, G.; Casnati, G.; Puplia, G.; Sartori, G.; Terenghi, G.
J . Chem. Soc., Perkin 1 1980, 1862-1865.
(43) Wayland, B. B.; Poszmik, G.; Fryd, M. Organometallics 1992,
11, 3534-3542.
(48) Kawato, T.; Kanatomi, H.; Koyama, H.; Igarashi, T. J . Photo-
chem. 1986, 33, 199-208.
(44) Wayland, B. B. Polyhedron 1988, 7, 1545-1555.
(45) Wayland, B. B.; Feng, Y.; Ba, S. Organometallics 1989, 8, 1438-
1441.
(49) Parshall, G. W. In Inorganic Syntheses; Parshall, G. W., Ed.;
McGraw-Hill Book Co.: New York, 1974; Vol. XV, p 14.

Reactions of RhH(Bu4salophen) with Olefins
Organometallics, Vol. 15, No. 6, 1996 1705
nol to give 1.6715 g (72% yield) of dark purple powder upon
vacuum drying. This powder was generally used without
further purification, though analytically pure, diffraction
quality crystals were grown by slow evaporation of solvent
from a benzene:methylene chloride solution at room temper-
ature. Anal. Calcd for C40H55N2O2Rh: C, 68.75; H, 7.93; N,
4.01. Found: C, 68.28; H, 7.80; N 3.80. Electronic absorp-
tion: λmax ) 583 nm, log ꢀ 3.90; 548 nm, log ꢀ 3.85; 497 nm,
log ꢀ 3.97; 394 nm, log ꢀ 4.53. Emission (λexcitation ) 510 nm):
λmax ) 671 nm, 745 nm (sh). Exitation (λobserved ) 670 nm):
475 nm (sh), 503 nm, 551 nm, 592 nm. 1H NMR: δ 8.40 (d, 2
H, J Rh-H ) 1.9 Hz, -NdCH-); 7.86, 7.33 (d, 4 H, J H-H ) 2.4
were needed for complete loss of the Rh(n-Bu)(Bu4salophen),
at which time the solution was used for subsequent reactions.
Rea ction of D2 w ith 3. A solution of 3 (method A) in a J .
Young vacuum-line-adapted NMR tube was attached to a
vacuum line and frozen in liquid N2. The headspace of the
tube was evacuated then filled with ca. 700 Torr of D2. The
tube was thawed and removed from the line, and a 1H NMR
spectrum was immediately acquired. HD was observed in this
first spectrum as a 1:1:1 triplet at δ 4.43 ppm with J HD ) 43
Hz. Within 10 min, the hydride resonance at δ -25.77 ppm
had completely vanished while the HD triplet continued to
grow in.
Ad d ition of Olefin s to Rh H(Bu 4sa lop h en ). Olefin ad-
dition products were prepared from solutions of 3 generated
by either method A or B as detailed above and, in some cases,
both. For olefins liquid at room temperature, a drop of the
olefin was added under the nitrogen atmosphere of a glovebox
to an NMR tube containing 3 prepared via method A. The
tube was then shaken and removed immediately from the
glovebox, and progress of the reaction was monitored by 1H
NMR. Once the reaction was complete, solvent and excess
olefin were removed under vacuum, fresh solvent was vacuum
transferred into the tube, and the NMR data listed below were
acquired. Alternatively, solutions of liquid olefins in benzene
were prepared in a glovebox and subsequently transferred
under hydrogen into a Schlenk tube containing 4 prepared by
method B. The reaction solution was allowed to stir for ∼8 h,
at which point solvent was removed under vacuum and the
residue collected.
3
Hz, -C6H2(C(CH3)3)2 ); 7.13, 6.86 (dd, 4 H, J H-H ) 6.2 Hz,
4J H-H ) 3.2 Hz, -NC6H4N-); 2.36 (m, 2 H, RhCH2CH2CH2-
CH3); 1.99, 1.45 (s, 36 H, -C(CH3)3); 0.90 (m, 2 H, RhCH2-
CH2CH2CH3); 0.44 (t, 3 H, J H-H ) 7.0 Hz, RhCH2CH2CH2CH3).
13C NMR: δ 167.52, 151.95, 144.38, 143.26, 136.03, 130.31,
129.1, 125.47, 120.36 (2 Hz), 114.85, 36.7, 34.25, 31.62, 30.38,
34.36, 25.14 (30.5 Hz), 22.53 (2.4 Hz), 13.65.
Syn th esis of Rh Et(Bu 4sa lop h en ) (2). The same proce-
dure as for 1 was used. A 0.0775 g (0.143 mmol) amount of
Bu4salophenH2, 0.034 g (0.087 mmol) of [Rh(µ-Cl)(C2H4)2]2, and
0.86 mL (0.69 mmol) of 0.8 M TEA(OH) in methanol were used
to give 30 mg (31%) of product. 1H NMR: δ 8.35 (d, 2 H, J Rh-H
) 1.9, -NdCH-); 7.87, 7.32 (d, 4 H, J H-H ) 2.6, -C6H2-
3
4
(C(CH3)3)2-); 7.10, 6.85 (dd, 4 H, J H-H ) 6.3 Hz, J H-H ) 3.3
2
Hz, -NC6H4N-); 2.37 (qd, 2 H, RhCH2CH3, J Rh-H ) 3.2
Hz,3J H-H ) 7.4 Hz); 1.98, 1.46 (s, 36 H, -C(CH3)3); 0.23 (td, 3
3
3
H, RhCH2CH3, J Rh-H ) 1.1 Hz, J Rh-H ) 7.4 Hz).
Ad d ition of C2H4. Ethylene was reacted only with 3
prepared by method A. An NMR sample of 4 was attached to
a high-vacuum line and frozen in liquid nitrogen to the top of
the solution. The headspace of the tube was then evacuated
and filled with 700 Torr of ethylene. Progress of the reaction
In Situ p r ep a r a tion of Rh H(Bu 4sa lop h en ). (A) NMR
Sca le. In a glovebox, 5 mg (7 µmol) of 1 in 0.5 mL (14 mM) of
benzene-d6 was loaded into a 5 mm diameter NMR tube fitted
with a concentric J . Young vacuum-line-adapted Teflon valve.
The tube was closed, removed from the glovebox, attached to
a high-vacuum manifold, immersed to the valve in liquid
nitrogen, and opened to vacuum to evacuate the headspace,
which was subsequently filled with ∼700 Torr of hydrogen.
The valve was closed and the tube brought to room tempera-
ture. (Safety note: A simple ideal gas law calculation shows
that the headspace of a tube so prepared holds in excess of 3
atm. Any sample so prepared should be handled with care.)
The sample was irradiated using light from an Oriel lamp
fitted with a 475 nm cutoff filter. Progress of the photolysis
was monitored spectroscopically via 1H NMR. Typically,
photolysis was complete within 1 h. No reaction was observed
in a sample prepared as detailed above when it was kept
rigorously in the dark, though solutions so prepared would
photolyze slowly over the course of several days if kept in
ambient room light. Yields (NMR) typically range 50-75%.
1H NMR: δ 8.26 (bs, 2 H, -NdCH-); 7.85, 7.20 (bs, 4 H,
-C6H2(C(C H3)3)2-); 7.02, 6.81 (dd, 4H, -NC6H4N-); 1.98, 1.48
1
was monitored by H NMR spectroscopy.
Ad d ition of Styr en e. Rh (CH2CH2P h )(Bu 4sa lop h en ) (4).
Styrene was added to samples of 3 prepared by both methods.
For the preparative-scale reaction (B) 155 mg of 2 was used.
The product residue was collected on sintered glass, washed
with several <2 mL portions of methanol, and dried under
vacuum to afford 94 mg of product (57% yield). 1H NMR: δ
8.35 (d, 2 H, J Rh-H ) 1.8 Hz, -NdCH-); 7.88, 7.3 (d, 4H, J H-H
3
) 2.6 Hz, -C6H2(C(CH3)3)2 ); 7.1, 6.86 (dd, 4 H, J H-H ) 6.3
4
Hz, J H-H ) 3.4 Hz, -NC6H4N-); 6.67-6.79 (m, 5 H, -CH2-
CH2C6H5); 2.48 (m, 2 H, RhCH2CH2C6H5); 2.29 (“t”, 2 H,
RhCH2CH2C6H5); 2.00, 1.46 (s, 36 H, -C6H2(C(CH3)3)2). 13C
NMR: δ 167.48, 152.15, 144.27, 143.24, 136.1, 130.44, 129.21,
128.73, 128.32, 128.46, 126.55, 125.55, 125.41, 120.32, 114.91,
1
36.71, 34.26, 31.63, 30.64, 38.55 (d, 0.8 Hz), 25.22 (d, J RhC
)
31.4 Hz). Anal. Calcd for C44H55N2O2Rh: C, 70.76; H, 7.42;
N, 3.75. Found: C, 70.73; H, 7.60; N, 3.74.
Ad d ition of Styr en e-d 8. A drop of styrene-d8 was added
1
(s, 36 H, -C(CH3)3); -25.77 (bd, 1 H, RhH, J Rh-H ) 51 Hz).
to a solution of 4 (method A). 1H NMR: δ 8.34 (d, 2 H, J RhH
)
13C NMR: δ 168.32, 152.72, 144.64, 142.97, 136.19, 130.3,
2.1 Hz, -NdCH-); 7.88, 7.30 (d, 4 H, J HH ) 2.6 Hz, -C6H2-
129.07, 125.3, 120.61, 114.56, 36.73, 34.25, 31.65, 30.24.
3
4
(C(CH3)3)2); 7.10, 6.86 (dd, 4 H, J HH ) 6.3 Hz, J HH ) 3.3 Hz,
-NC6H4N-); 2.27 (“t”, 1 H, RhCD2CHD2C6D5); 2.00, 1.46 (s,
36 H, -C6H2(C(CH3)3)2).
(B) P r ep a r a tive Sca le. A cylindrical Schlenk tube of
approximate dimension 15 cm long and 2.5 cm in diameter
was loaded with a magnetic stir bar and with 155 mg of Rh-
(n-Bu)(Bu4salophen) (1) in approximately 50 mL of solvent (0.4
mM). The tube was closed with a rubber septum and removed
from the glovebox. On a standard Schlenk line, the headspace
of the vessel was purged under a steady stream of hydrogen
for several minutes. It was then again closed and with stirring
was photolyzed as above. Progress of the photolysis was
monitored by periodic removal, under hydrogen pressure via
cannula, of small (<1 mL) reaction mixture samples. These
samples were cannulated through the rubber septum of a
septum-closed NMR tube containing benzene-d6 (>0.25 mL)
and a drop of styrene. Proton NMR spectra of these samples
were inspected for loss of Rh(n-Bu)(Bu4salophen) (1) butyl-
group signals and for growth of the methylene signals of Rh-
(CH2CH2Ph)(Bu4salophen) (4). Photolysis times of about 2 h
Com p etitive Ad d ition : Eth ylen e vs Styr en e. A solution
of 4 (method A) was taken into a glovebox and frozen. The
tube was opened, and a drop of styrene was placed on the top
wall of the tube. The tube was closed, removed from the box,
and placed on the high vacuum line. It was immediately
immersed in liquid nitrogen before it thawed. Ethylene was
then placed over the sample as above. The sample was thawed
and immediately placed in the NMR probe, in which the
reaction was followed by 1H NMR spectroscopy. Large ex-
cesses of both styrene and ethylene were seen in solution by
integration of the respective 1H NMR signals for the olefins
and the metal complexes.
Ad d ition of Acr ylon itr ile. Rh (CH2CH2CN)(Bu 4sa lo-
p h en ) (5). 5 was prepared from 3 (method A) by addition of
excess (988 equiv) acrylonitrile. Yield (1H NMR): 73%. 1H

1706 Organometallics, Vol. 15, No. 6, 1996
Anderson and Eisenberg
NMR (with excess acrylonitrile): δ 8.46 (d, 2 H, J RhH ) 1.7
Hz, -NdCH-); 7.69, 7.24 (d, 4 H, J HH ) 2.4 Hz, -C6H2-
2 H, RhCH2CH2(c-C3H5)), 0.214, -0.054, -0.35 (cm, 5 H,
RhCH2CH2(c-C3H5)).
3
4
(C(CH3)3)2); 7.38, 6.95 (dd, 4 H, J HH ) 6.2 Hz, J HH ) 3.4 Hz,
-NC6H4N-); 2.12 (m); 1.94 (m); 1.75, 1.39 (s, 36 H, -C6H2-
(C(CH3)3)2). 13C NMR: δ 167.17, 152.64, 143.83, 142.8, 135.79,
130.52, 129.47, 125.77, 119.7, 115.19, 36.38, 34.16, 31.59,
Rh (CH2CH2CHdCHCH3)(Bu 4sa lop h en ) (8) (17-24%).
1H NMR: δ 8.36 (d, J Rh-H ) 1.7 Hz, 2 H, -NdCH-); 7.87,
7.32 (d, J H-H ) 2.2 Hz, 4 H, -C6H2(C(CH3)3)2 ); 7.10 (dd, 2 H,
4
3J H-H ) 6.1 Hz, J H-H ) 3.4 Hz, -NC6H4N-); 6.85 (obscured
1
30.14, 17.61, 12.73 (d, J RhC ) 32.6 Hz).
dd, -NC6H4N-); 4.94 (cm, RhCH2CH2CHdCHCH3); 2.32 (cm,
RhCH2CH2CHdCHCH3); 1.98, 1.46 (s, -C6H2(C(CH3)3)2), 1.65
(RhCH2CH2CHdCHCH3), 1.16 (bm, RhCH2CH2CHdCHCH3).
Un id en tified Rh CH2-Con ta in in g P r od u ct. (1-8%). 1H
NMR: δ 2.24 (cm).
Ad d ition of Bis(a llyl) Eth er . Rh CH2CH2CH2OCH2CHd
CH2(Bu 4sa lop h en ) (6). Approximately 285 equiv of bis(allyl)
ether was vacuum transferred into a previously prepared
solution of 3 (method A). The amount of excess allyl ether
and yields were determined spectroscopically, and then the
solvent and excess allyl ether were removed under vacuum.
NMR data listed are for product in the absence of excess allyl
ether. Yield (1H NMR): 56%. 1H NMR: δ 8.32 (d, 2 H, J RhH
) 2.2 Hz, -NdCH-); 7.86, 7.30 (d, 4 H, J HH ) 2.5 Hz, -C6H2-
(Bu 4sa lop h )R h CH 2CH dCHCH 2CH 2R h (Bu 4sa lop h ) (9)
(0-6%). 1H NMR: δ 8.28 (d, 2 H, J RhH ) 2.1 Hz, -NdCH-);
8.18 (d, 2 H, J RhH ) 1.3 Hz, -NdCH-); 7.85 (d, 2 H, J HH
)
2.7 Hz, -C6H2(C(CH3)3)2 ); 7.76 (d, 2 H, J HH ) 2.4 Hz, -C6H2-
(C(CH3)3)2 ); 7.29 (d, 4 H, J HH ) 2.4 Hz, -C6H2(C(CH3)3)2 );
7.19 (d, partially obscured by solvent, -C6H2(C(CH3)3)2 ), 7.05
(dd, 2 H, 3J HH ) 6.1 Hz, 4J HH ) 3.4 Hz, -NC6H4N-), 7.01 (dd,
3
4
(C(CH3)3)2); 7.07, 6.83 (dd, 4 H, J HH ) 6.1 Hz, J HH ) 3.2 Hz,
-NC6H4N-); 5.47 (m, 1 H, -CH2CH2CH2OCH2CHdCH2), 4.86
3
(dm, J HH ) 17.2, -CH2CH2CH2OCH2CHdCH2 (trans)), 4.74
3
4
4 H, J HH ) 6.1 Hz, J HH ) 3.7 Hz, -NC6H4N-); 6.83 (dd,
(dm, 3J HH ) 10.4, -CH2CH2CH2OCH2CHdCH2 (cis)), 2.38 (m,
2 H, -CH2CH2CH2OCH2CHdCH2); 1.98, 1.45 (s, 36 H, -C6H2-
(C(CH3)3)2), 1.22 (m, 2 H, -CH2CH2CH2OCH2CHdCH2), 2.82
3
partially obscured by other products), 6.71 (dd, 4 H, J HH
)
6.1 Hz, 4J HH ) 3.4 Hz, -NC6H4N-); 4.56 (m, 1 H, RhCH2CHd
CHCH2CH2Rh), 4.39 (m, 1 H, RhCH2CHdCHCH2CH2Rh), 2.84
(dd, 2 H, J RhH ) 3.5 Hz, J HH ) 8.2 Hz, RhCH2CHdCHCH2-
CH2Rh), 1.9 (RhCH2CHdCHCH2CH2Rh), 0.4 (RhCH2CHdC-
HCH2CH2Rh).
3
(t, 2 H, J HH ) 7 Hz, -CH2CH2CH2OCH2CHdCH2), 3.37 (d, 2
3
H, J HH ) 5.3, -CH2CH2CH2OCH2CHdCH2).
Ad d ition of Vin ylcyclop r op a n e. Addition of a slight
excess (1.8 equiv) of vinylcyclopropane to a solution of 3
(method A) gave a mixture of products. Due to the position of
the cyclopropyl resonances in the 1H NMR spectrum, typical
nonreactive alkane or alkyl silane internal standards were not
used. In the presence of THF as a standard, the overall yield
of Schiff-base-containing product was 94% (1H NMR) as shown
by integration of the spectral regions in which the imine
protons of the Schiff base resonate (8.5-8.1 ppm), though
selectivity for the identified products was lower (34%) in this
reaction than in those carried out in the absence of THF
(typically 57%). Percentage composition ranges for the prod-
ucts listed below were based on integration of resonances
unique for each product versus integration of the imine region.
Rh (CH2CH2(c-C3H5))(Bu 4sa lop h en ) (7). (26-32%). 1H
NMR: δ 8.38 (bs, -NdCH-); 7.86, 7.31 (obscured, -C6H2-
(C(CH3)3)2 ); 7.1, 6.86 (obscured, -NC6H4N-); 2.42 (cm, 2 H,
RhCH2CH2(c-C3H5)), 1.97, 1.44 (s, -C6H2(C(CH3)3)2), 0.941 (cm,
Ad d it ion of tr a n s-1,3-P en t a d ien e. R h CH 2CH 2CH d
CHCH3(Bu 4sa lop h en ) (8). Approximately 2 equiv of 1,3-
pentadiene were dropped into a previously prepared solution
of 3 (method A). The amount of excess 1,3-pentadiene and
yields were determined spectroscopically, and then the solvent
and excess 1,3-pentadiene were removed under vacuum. NMR
data listed are for the product in absence of excess pentadiene.
Yield (1H NMR): 59%.
Ack n ow led gm en t. We wish to thank the Depart-
ment of Energy, Division of Chemical Sciences, for
support of this research. A generous loan of rhodium
salts from Alfa/AESAR J ohnson Matthey Co. is also
gratefully acknowledged.
OM950595K