Reactions of (Trimethyltriazacyclononane)Rh(vinyl)3, - Rh(Z-propenyl)3, and - Rh(vinyl)2Me with protic acids. The relative migratory aptitude of methyl and vinyl groups to an (Ethylidene)Rh alkylidene carbon

Article abstract of DOI:10.1021/om980423y

Preparations of CnRh(CH=CH₂)₃, CnRh(Z-CH=CHMe)₃, and CnRhMe(CH=CH₂)₂ (Cn = 1,4,7-trimethyl-1,4,7-triazacyclononane) are reported. An X-ray crystal structure determination of CnRh(CH=CH₂)₃ revealed C₃ molecular symmetry with average lengths Rh-N, 2.25 ?; Rh-C, 2.00 ?; and C=C, 1.31 ?. Protonations of these compounds were carried out with HOTf (triflic acid, HOSO₂CF₃), [H(Et₂O)₂][BAr₄] (BAr₄ = ^(-)B[3,5-(CF₃)₂C₆H₃] ₄), and HCl in organic solvents as well as with methanol as solvent and acid. CnRh(CH=CH₂)₃ gave [CnRh(CH=CH₂)(η³-E-CH₂CHCHMe)]X in isolated yields up to ca. 80%. Use of CD₃OD as acid/solvent led to formation of [CnRh(CH=CH₂)(η³-CH₂CHCHCH₂D)]X (X = ^(-)OCD₃ assumed) as the only isotopomeric product, suggesting [CnRh(=CHCH₂D)-(CH=CH₂)₂]⁺ as a probable intermediate. No Rh(=CH-CH₃) functional group could be detected by NMR at -60 °C or above. Triflic acid protonation of CnRh(Z-CH=CHMe)₃ generated [CnRh(Z-CH=CHMe)(η³-E,E-MeCHCHCHEt)], establishing that product formation is by alkenyl migration and not via an electrocyclic ring closure within the [Rh(=CHEt)(CH=CHMe)]⁺ moiety. Protonation of CnRhMe(CH=CH₂)₂ generated [CnRhMe(η³-CH₂CHCHMe)]⁺ as the sole product (96% by NMR). Reaction of CnRh(CH=CH₂)₂Me in CD₃OD generated [CnRhMe(η³-CH₂CHCHCH₂D)]⁺ as the sole isotopomer. Thus, the migratory competition established between methyl and vinyl groups for migration to the carbene carbon in proposed intermediate [CnRh(=CHMe)(CH=CH₂)Me]⁺ is dominated by the vinyl group.

Full text of DOI:10.1021/om980423y

Organometallics 1998, 17, 5397-5405
5397
Rea ction s of (Tr im eth yltr ia za cyclon on a n e)Rh (vin yl)3,
-Rh (Z-p r op en yl)3, a n d -Rh (vin yl)2Me w ith P r otic Acid s.
Th e Rela tive Migr a tor y Ap titu d e of Meth yl a n d Vin yl
Gr ou p s to a n (Eth ylid en e)Rh Alk ylid en e Ca r bon
Hongshi Zhen, Chunming Wang, Yonghan Hu, and Thomas C. Flood*
Department of Chemistry, University of Southern California,
Los Angeles, California 90089-0744
Received May 26, 1998
Preparations of CnRh(CHdCH2)3, CnRh(Z-CHdCHMe)3, and CnRhMe(CHdCH2)2 (Cn )
1,4,7-trimethyl-1,4,7-triazacyclononane) are reported. An X-ray crystal structure determi-
nation of CnRh(CHdCH2)3 revealed C3 molecular symmetry with average lengths Rh-N,
2.25 Å; Rh-C, 2.00 Å; and CdC, 1.31 Å. Protonations of these compounds were carried out
with HOTf (triflic acid, HOSO2CF3), [H(Et2O)2][BAr4] (BAr4
)
(-)B[3,5-(CF3)2C6H3]4), and
HCl in organic solvents as well as with methanol as solvent and acid. CnRh(CHdCH2)3
gave [CnRh(CHdCH2)(η3-E-CH2CHCHMe)]X in isolated yields up to ca. 80%. Use of
CD3OD as acid/solvent led to formation of [CnRh(CHdCH2)(η3-CH2CHCHCH2D)]X (X )
(-)OCD3 assumed) as the only isotopomeric product, suggesting [CnRh(dCHCH2D)-
(CHdCH2)2]+ as a probable intermediate. No Rh(dCH-CH3) functional group could be
detected by NMR at -60 °C or above. Triflic acid protonation of CnRh(Z-CHdCHMe)3
generated [CnRh(Z-CHdCHMe)(η3-E,E-MeCHCHCHEt)], establishing that product for-
mation is by alkenyl migration and not via an electrocyclic ring closure within the [Rh-
(dCHEt)(CHdCHMe)]+ moiety. Protonation of CnRhMe(CHdCH2)2 generated [CnRhMe-
(η3-CH2CHCHMe)]+ as the sole product (96% by NMR). Reaction of CnRh(CHdCH2)2Me in
CD3OD generated [CnRhMe(η3-CH2CHCHCH2D)]+ as the sole isotopomer. Thus, the
migratory competition established between methyl and vinyl groups for migration to the
carbene carbon in proposed intermediate [CnRh(dCHMe)(CHdCH2)Me]+ is dominated by
the vinyl group.
In tr od u ction
alkylidene to an alkene substrate to form a cyclopro-
pane. However, potential also exists to use the elec-
trophilic alkylidene fragment for intramolecular C-C
bond formation in the coordination sphere of the metal.
A few metal-to-alkylidene hydrocarbyl migratory inser-
tions are known, although in those cases the alkylidene
fragment has generally been generated by means other
than vinyl protonation. Examples include transiently
generated (inferred) carbenes [(Me3P)3Ir(dCH2)-
MeBr]+,2a [Cp2W(dCH2)R]+,2b [(η6-C5H4CH2CH2)-
(η6-C5H4CH2CHd)W]+,2c and [(η6-C6H6)Ru(dCH2)Me]+.2d
Heating of Cp2Ta(dCHMe)Me at 80 °C led to an
apparent methyl migration to the alkylidene carbon.3a
One of the established methods to generate electro-
philic metal-alkylidene (metal-“carbene”) complexes
is the protonation of vinyl-metal groups (eq 1).1 Most
known examples involve CpFe (“Fp”)-coordinated vinyl
groups, but examples involving (dmpe)2Fe,1b Ru and
Os,1e,f Re,1m and Pt1c are known. In most cases, the
objective has been to transfer the metal-coordinated
Heating (dmpe)W(CH2 Bu)(dCHtBu)(tCtBu) in neat
t
dmpe at 120 °C gave trans-(dmpe)2W(H)(tCtBu) and
E-tBuCHdCHtBu by a postulated migration of the
neopentyl group to the neopentylidene.3b J ones et al.3c
demonstrated a reversible migration of an alkyl group
(1) (a) Davison, A.; Reger, D. L. J . Am. Chem. Soc. 1972, 94, 9237.
(b) Bellerby, J . M.; Mays, M. J . J . Organomet. Chem. 1976, 117, C21.
(c) Bell, R. A.; Chisholm, M. H. Inorg. Chem. 1977, 16, 698. (d) Davison,
A.; Selegue, J . P. J . Am. Chem. Soc. 1978, 100, 7763. (e) Bruce, M. I.;
Wallis, R. C. Aust. J . Chem. 1979, 32, 1471. (f) Bruce, M. I.; Swincer,
A. G. Aust. J . Chem. 1980, 33, 1471. (g) Marten, D. F. J . Org. Chem.
1981, 46, 5422. (h) Bodnar, T.; Cutler, A. R. J . Organomet. Chem. 1981,
213, C31. (i) Selegue, J . P. J . Am. Chem. Soc. 1982, 104, 119. (j) Casey,
C. P.; Miles, W. H.; Tukada, H.; O’Connor, J . M. J . Am. Chem. Soc.
1982, 104, 3761. (k) Kremer, K. A. M.; Kuo, G.-H.; O’Connor, E. J .;
Helquist, P.; Kerber, R. C. J . Am. Chem. Soc. 1982, 104, 6119. (m)
Hatton, W. G.; Gladysz, J . A. J . Am. Chem. Soc. 1983, 105, 6157. (n)
Kuo, G.-H.; Helquist, P.; Kerber, R. C. Organometallics 1984, 3, 806.
(o) Casey, C. P.; Miles, W. H.; Tukada, H. J . Am. Chem. Soc. 1985,
107, 2924.
(2) (a) Thorn, D. L.; Tulip, T. H. J . Am. Chem. Soc. 1981, 103, 5984.
(b) Hays, J . C.; Cooper, N. J . J . Am. Chem. Soc. 1982, 104, 5570. (c)
Chong, K. S.; Green, M. L. H. J . Chem. Soc., Chem. Commun. 1982,
991. (d) Roder, K.; Werner, H. Angew. Chem., Int. Ed. Engl. 1987, 26,
686.
(3) (a) Sharp, P. R.; Schrock, R. R. J . Organomet. Chem. 1979, 171,
43. (b) Holmes, S. J .; Clark, D. N.; Turner, H. W., Schrock, R. R. J .
Am. Chem. Soc. 1982, 104, 6322. (c) Stenstrum, Y.; Koziol, A. E.;
Palenik, G. J .; J ones, W. M. Organometallics 1987, 6, 2079.
10.1021/om980423y CCC: $15.00 © 1998 American Chemical Society
Publication on Web 11/06/1998

5398 Organometallics, Vol. 17, No. 24, 1998
Zhen et al.
from the metal to a stabilized alkylidene center as
shown in eq 2.
Z-LiCHdCHMe and CnRhCl3 react to form CnRh-
(Z-CHdCHMe)3 (2), which was isolated in 9% yield. The
1H and 13C NMR spectra of 2 in acetone-d6 are very
similar to those of 1, with the single NMe proton
resonance at δ 2.53, indicating C3v molecular symmetry.
The geminal (δ 6.94) and trans (δ 5.90) protons with
respect to Rh are Rh coupled with J ) 4.0 and 3.4 Hz,
respectively), but the propenyl Me (δ 1.70) is not. In
the 13C NMR spectrum, CR of the propenyl group (δ
150.88) is Rh coupled with J ) 42 Hz: Câ (δ 128.16)
and Me (Cγ, δ 19.70) are not.
In the context of this approach to C-C bond forma-
tion, it could be useful to establish the relative aptitudes
of various types of hydrocarbyl ligands for migration
from the metal to the alkylidene moiety (eq 3). We report
As shown in eq 5, CnRhMe(OTf)2 (OTf ) triflate or
-OSO2CF3) is also vinylated by vinyllithium. Benzene
here the synthesis and protonation of CnRh(CHdCH2)3
(1), CnRh(Z-CHdCHMe)3 (2), and CnRhMe(CHdCH2)2
(3) (Cn ) 1,4,7-trimethyl-1,4,7-triazacyclononane). Evi-
dence is presented that protonation forms transient
rhodium carbene complexes, and the preference for vinyl
migration over methyl migration to the electrophilic
ethylidene carbon is demonstrated. This system has the
potential to allow competition experiments among a
variety of migrating hydrocarbyl groups to a variety of
alkylidene fragments, contingent on the synthesis of
their precursors.
extraction of the residue after solvent removal affords
ca. 70% yield of CnRhMe(CHdCH2)2, 3. It is important
not to use CH2Cl2 or CHCl3 for workup of 3, since this
generates side products that are difficult to separate.
It is interesting that 3 has less thermal stability than
either 1 or CnRhMe3. On standing at room temperature
as a solid or in solution, Cn is liberated and metal
precipitates. The approximate half-time for decomposi-
tion in benzene is 2 days, and as the solid ca. 1 week.
For this reason, no combustion analysis was obtained
and the crude material was used for the protonation
Resu lts
Syn th eses a n d Sp ectr a . Wieghardt4 has described
the preparation of CnRhCl3, which serves as the starting
material for all of our CnRh complexes. Vinylation of
CnRhCl3 produces CnRh(CHdCH2)3, 1, in 30 min with
ca. 90% yield after workup (eq 4). Trivinyl 1 is a
1
experiments. The H and 13C NMR spectra in C6D6 are
consistent with Cs molecular symmetry, but are other-
wise very similar to the spectra of 1 and 2. The Rh-
methyl group appears at δ 0.57 (J RhH ) 2.5 Hz) in the
proton spectrum and at δ 3.61 (J RhC ) 37 Hz) in the
13C NMR. The spectra showed the crude material to
be at least 95% free of organic impurities. Complex 3
is soluble in DMSO, CH2Cl2, and THF, moderately
soluble in benzene, and insoluble in alkanes.
CnRhMe(OTf)2, starting material for the synthesis of
3, is generated by acid cleavage of CnRhMe3,6 so we
wish to point out that we have found a much improved
method for the preparation of the latter complex. The
previously reported method which used MeLi required
several days, and the yield and purity of the product
were very dependent on the source of the MeLi.6 More
recently, we have found that MgMe2 is a much better
reagent for the methylation of CnRhCl3 (see the Ex-
perimental Section).
P r oton a tion s of Rh od iu m Vin yls. For other syn-
thetic purposes, it was intended to cleave one or two
vinyl groups from 1 to prepare CnRh(CHdCH2)2(OTf)
and CnRh(CHdCH2)(OTf)2. In fact, when a CH2Cl2
solution of 1 at -78 °C was treated with 1 equiv of a
titrated ether solution of triflic acid (CF3SO3H or HOTf)
and allowed to warm to room temperature, the 1H NMR
spectrum was initially complex and uninterpretable.
Over a period of 20-30 h, this mixture converted to a
single material whose 1H and 13C NMR spectra are
consistent only with the structure [CnRh(CHdCH2)(η3-
CH2CHCHMe)]+ (4) (eq 6). The 1H NMR spectrum
crystalline solid that is soluble in THF, CH2Cl2, and
DMSO, but not in ether, benzene or alkanes.
Slow evaporation of solvent from a methylene chloride
solution of 1 afforded X-ray diffraction-quality crystals.
1
The H NMR spectrum in DMSO-d6 is consistent with
effective C3v symmetry with a single N-methyl singlet
and a single set of vinyl resonances, all three of which
are rhodium coupled (J RhH: cis ) 1.9, trans ) 3.8, gem
) 3.0 Hz). The 13C NMR spectrum is similarly simple,
with the vinyl methylene (Câ) at δ 115.12 and the
rhodium-vinyl carbon (CR) doublet at δ 166.51 (J RhC
)
42 Hz). Spectral parameters appear to be typical. For
example, the 1H and 13C parameters for the vinyl
groups in 1 (and 3 below) are very similar to those in
Cp*(CH2dCH)Rh(µ-CH2)2Rh(CHdCH2)Cp*, including
1J RhC of 40-43 Hz for the vinyl R-carbons.5
(4) (a) Wieghardt, K.; Chaudhuri, P.; Nuber, B.; Weiss, J . Inorg.
Chem. 1982, 21, 3086. (b) Chaudhuri, P.; Wieghardt, K. Prog. Inorg.
Chem. 1987, 35, 329.
(5) Martinez, J . M.; Gill, J . B.; Adams, H.; Bailey, N. A.; Saez, I.
M.; Sunley, G. J .; Maitlis, P. M. J . Organomet. Chem. 1990, 394,
583.
(6) Wang, L.; Wang, C.; Bau, R.; Flood, T. C. Organometallics 1996,
15, 491.

Reactions of (Trimethyltriazacyclononane)Rh(vinyl)3
Organometallics, Vol. 17, No. 24, 1998 5399
in the spectra from 210 to 240 K. The vinyl, ligand
1
NCH3, and butenyl ligand regions of the H spectra at
210-240 K were very cluttered and changed continu-
ously with temperature, so structural assignments could
not be made. Resonances attributable to product 4 were
increasingly evident at these temperatures, and after
2 days at room temperature only 4 was observed: no
starting material or intermediates were present.
A -78 °C solution of CnRh(CHdCH2)3 (1) in CH2Cl2
was treated with an ether solution of DOTf, and the
product was worked up as usual after completion of the
reaction. 1H and 13C NMR spectra of the product in
DMSO-d6 were identical to those of 4, except that the
butenyl Me proton resonance was substantially broad-
ened (D coupling and shifting). Part of the butenyl
methyl 13C NMR singlet at δ 16.62 was split into a 1/1/1
triplet (J DC ) 19.7 Hz) and isotopically shifted to δ
16.54. Integration10 showed about 30% 4-d1. Also, part
of the intensity of the resonance for the adjacent allyl
carbon at δ 60.19 (d, J RhC ) 10.9 Hz) was shifted upfield
by 6 Hz (at 90 MHz). In a second experiment, 1 equiv
of DOTf was added as neat acid to a solution of 1. NMR
spectra were identical to those just described except that
the level of monodeuteration was now ca. 50%. In a
final attempt to maximize the degree of monodeutera-
tion in butenyl complex 4, a small amount of 1 was
suspended in CD3OD, and progress of the reaction was
suggests nearly quantitative formation of 4, but it was
isolated in 79% yield. The presence of three distinct
NMe groups in the 1H NMR spectrum indicates that the
metal center is chiral, in contrast to the Cs symmetry
anticipated for CnRh(CHdCH2)2(OTf). In addition, the
proton resonance intensities in 4 integrate as one vinyl
group, not two. Proton resonances of the vinyl group
in 4 are very similar to those in 1 and 3, but the vinyl
1
R-carbon of 4 exhibits a J RhC of 26 Hz, in contrast to
the values of 40-43 Hz for 1-3 and Cp*(CH2dCH)Rh-
(µ-CH2)2Rh(CHdCH2)Cp*.5 η3-Butenyl proton reso-
nances of 4 range from the RhMe doublet at δ 1.14 to
the central allylic multiplet at 4.33 ppm and are very
similar to those of a number of known η3-allylic rhodium
complexes.5,7 The three π-allyl carbons of 4 have
rhodium couplings in the carbon spectrum which are
normal for allylic rhodium groups, but chemical shifts
are slightly different from most CpRh or Cp*Rh allylic
1
monitored by H and 13C NMR spectroscopy. The solid
cationic or neutral complexes:5,7a labeled as C(1)-C(2)
-
dissolved slowly as the reaction progressed, and after
ca. 3 days the sample was homogeneous. Both the
proton and carbon resonances were shifted various
small amounts in methanol-d4, compared to DMSO-d6,
but were otherwise the same. The butenyl CH3 reso-
nance at δ 17.10 was very small, and the deuterium
coupled CH2D 1/1/1 triplet at δ 16.82 was dominant. The
allyl carbon adjacent to CH3 was not detectable in the
baseline noise, while the carbon adjacent to CH2D was
now a clean doublet at δ 62.31 (J RhC ) 11 Hz). Integra-
tion10 of the butenyl methyl resonances at δ 17.10/16.82
indicated that 4 was 92% monodeuterated under these
conditions. No evidence of deuteration in any other
position of 4 could be detected. The counterion could
not be determined, and solvent removal led to decom-
position.
C(3)-Me; C(1) δ 46.64 (typical literature values δ 47-
61), C(2) 109.28 (85-99), C(3) 60.17 (70-82). The geom-
etry of the η3-butenyl group was initially assigned by
examination of the central allylic multiplet at δ 4.33
ppm, which appeared as a doublet of triplets with J anti
-
(triplet) ) 11 Hz and J syn(doublet) ) 7.8 Hz. Hence,
the methyl group must be syn to the central hydrogen.
Attempted X-ray crystal structure determinations of 4
were plagued by disorder (see below), but the geometry
assignment was corroborated by the spectra and X-ray
structure of 6 (below).
Since it is believed that the protonation-induced
rearrangements reported here proceed via rhodium
alkylidene intermediates, an attempt was made to
detect the [RhdCHCH3]+ group by NMR spectroscopy
at low temperature. An NMR sample of CnRh(CHd
CH2)3 (1) and Brookhart’s acid,8 [H(Et2O)2][BAr4] (BAr4
Reaction of CnRh(Z-CHdCHMe)3 (2) with [H(Et2O)2]-
[BAr4] in CH2Cl2 gave ca. 77% yield of [CnRh-
(CHdCHMe)(η3-MeCHCHCHEt)]+ (5), the structure of
)
(-)B[3,5-(CF3)2C6H3]4) in CD2Cl2 was monitored at
210, 220, 240, and 300 K by NMR. 1H spectra were
acquired from δ 0 to 20 ppm and 13C from δ 0 to 350
ppm.9 At least 14 000 transients were acquired for the
13C spectrum at each temperature. No trace of reso-
nances attributable to the [RhdCHCH3]+ group could
be detected at any of these temperatures: that is,
nothing was seen below δ 8 ppm in the 1H spectrum
and nothing below δ 170 ppm in the 13C spectrum.
Decreasing amounts of starting material were visible
1
which, shown in eq 6, was unequivocal from its H and
13C NMR spectra. A clear triplet at δ 4.62 (J ) 11 Hz)
for the central allylic proton indicates that the methyl
and ethyl groups of the η3-MeCHCHCHEt ligand have
the same geometry, and the coupling constant of 11 Hz
is indicative of anti coupling in these molecules.
Protonation of CnRhMe(CHdCH2)2 (3) by [H(Et2O)2]-
[BAr4] gave an 84% isolated yield of [CnRh(η3-
CH2CHCHMe)Me]+ (6) (eq 6), which was recrystallized
from ether/hexane and fully characterized, including an
(7) (a) Powell, P.; Russell, L. J . J . Organomet. Chem. 1977, 129, 415.
(b) Lee, H.-B.; Maitlis, P. M. J . Chem. Soc., Dalton Trans. 1975, 2316.
(8) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920.
(9) Proton chemical shifts for R-protons of alkylidene complexes are
found from δ -2 to 20 ppm; at higher field (δ -2 to 12) for “Schrock”-
type alkylidenes (Schrock, R. R. Acc. Chem. Res. 1979, 12, 98) and lower
field (δ 10 to 20) for electrophilic alkylidene complexes (Brookhart, M.;
Tucker, J . R.; Flood, T. C.; J ensen, J . J . Am. Chem. Soc. 1980, 102,
1203). Carbon chemical shifts for electrophilic alkylidene complexes
appear in the general range of δ 250 to 350 ppm (same references).
1
X-ray crystal structure (below). The H and 13C NMR
spectra were very similar to those of 4 and 5 except for
the absence of the vinyl group and the presence of a
rhodium-methyl doublet at δ -0.14 (J RhH ) 1.8 Hz) in
(10) No corrections for NOE and relaxation effects on the 13C
NMR integrals were made since the exact integral ratios are not
significant.

5400 Organometallics, Vol. 17, No. 24, 1998
Zhen et al.
present case. The average CdC length and average
Rh-CdC angle of 1 are 1.312 Å and 133.4°. Solid-state
structures of several vinyl or alkenyl rhodium complexes
have been determined.5,11 Of the examples cited, all
contain RhIII except one,11c but there are no statistically
significant differences in their parameters on the aver-
age. The Rh-C lengths range from 2.00 to 2.12 Å, the
CdC lengths from 1.25 to 1.36 Å (the shortest of which
is surprisingly short5), and the Rh-CdC angles from
121° to 130°. This puts the Rh-C length of 1 on the
short end of the range and the Rh-CdC angle at the
high end of the range, but there is no obvious depen-
dence of the metrics on the ligand array or oxidation
state among these examples.
F igu r e 1. The molecular structure (50% probability el-
lipsoids) of the heavy atoms of CnRh(CHdCH2)3 (1).
X-ray diffraction data were acquired on single crystals
of 4(BAr4) and 6(BAr4), but both structures were of poor
quality despite low-temperature data collection on two
or three different crystals of each of them. Unfortu-
nately, neither 4(X) nor 6(X) (X ) Cl or OTf) could be
induced to give crystals suitable for X-ray diffraction.
During the solution of both structures 4(BAr4) and
Ta ble 1. Selected Bon d Len gth s (Å) a n d Bon d
An gles (d eg) of Cn Rh (CHdCH2)3 (1)
Rh-N(1)
2.253(4)
2.250(4)
2.254(5)
1.995(5)
2.001(5)
1.990(5)
1.313(8)
1.308(8)
1.316(8)
N(1)-Rh-N(2)
N(2)-Rh-N(3)
N(1)-Rh-N(3)
C(1)-Rh-C(2)
C(2)-Rh-C(3)
C(3)-Rh-C(1)
Rh-C(10)-C(11)
Rh-C(12)-C(13)
Rh-C(14)-C(15)
80.2(2)
80.0(3)
80.0(3)
87.6(3)
87.6(3)
87.2(5)
132.9(4)
133.4(4)
133.9(5)
Rh-N(2)
Rh-N(3)
-
Rh-C(10)
Rh-C(12)
Rh-C(14)
C(10)-C(11)
C(12)-C(13)
C(14)-C(15)
6(BAr4), the large BAr4 anion presented itself almost
intact from the first E-map. Most of the CF3 groups
were disordered, but this was easily treated as described
in the Experimental Section, and the anions of both 4
and 6 refined very well. The cations however, did not.
Both cations showed signs of several types of disorder,
which we suspect arise from the disparity in cation and
anion sizes and from the approximately spherical shape
of the cations. Bond lengths and angles of the ligands
of cation 4 did not converge to reasonable values with
any scheme devised to deal with its disorder. Accord-
ingly, those data are not reported here even though the
final R factor was below 6%. However, the quality of
the solution was good enough to corroborate the struc-
ture assigned to the cation on the basis of spectroscopy.
The methyl group of the butenyl ligand of cation 6 was
disordered at the two ends of the allyl group. Refine-
ment of both methyl sites led to C(5)/C(5A) populations
of 0.56/0.44. With this minimal accommodation for
disorder, the final parameters for the cation of 6 were
ultimately good enough to establish its structure as
shown in Figure 2. Table 2 contains selected bond
lengths and bond angles. The structure is essentially
octahedral at the metal with the butenyl ligand occupy-
ing two cis sites as expected. Rh-C distances of the
butenyl group of 6 are typical in comparison with
literature examples of π-allyl rhodium structures.12
the proton spectrum and δ 3.85 (J RhC ) 25 Hz) in the
carbon spectrum. The resonance at δ 3.68 of the central
proton of the η3-butenyl group appeared as a doublet of
triplets (J anti(triplet) ) 11 Hz and J syn(doublet) ) 7.4
Hz) just as in 4. In combination with the X-ray
structure of 6, these 1H NMR data confirm the geometry
assignments of 4, 5, and 6. Protonation of an NMR
sample was carried out in DMSO-d6 with 1 equiv of a
titrated solution of HCl in ether. Integration of the
methyl proton resonances in starting 3 and product 6
with respect to internal Ph2Si(CH3)2 indicated that 6
had formed in 96% yield. Suspension of CnRhMe(CHd
CH2)2 (3) in CD3OD in an NMR tube at 56 °C led to its
gradual dissolution over 2 days and formation of [CnRh-
(η3-CH2CHCH-CH2D)Me]+ (6-d1). The 1H resonance of
the butenyl -CH2D group at δ 1.28, a doublet for CH3,
was a broad multiplet, and the adjacent allylic CH group
at δ 2.40, normally a doublet of quartets, was a doublet
of triplets. No other changes were visible for any of the
resonances, including that of the RhMe group.
Str u ctu r e Deter m in a tion s of Cn Rh (CHdCH2)3,
1, a n d [Cn Rh (η3-CH2CHCH-Me)Me][BAr 4], 6(BAr 4).
Solution of the X-ray diffraction data proceeded very
well for 1. A drawing of its molecular structure is shown
in Figure 1, and Table 1 contains selected bond lengths
and bond angles. Average Rh-N and Rh-C bond
lengths are 2.252 and 1.995 Å, respectively. Average
angles of N-Rh-N, 80.1°, and C-Rh-C, 87.5°, clearly
indicate that the Cn ligand is moved slightly away from
the metal along the molecular 3-fold axis (not a crystal-
lographic axis), and the Rh-vinyls are forced slightly
toward one another, probably because of the steric
congestion between the Rh-vinyls and the N-methyls.
These angles and Rh-N lengths are all similar to those
of CnRhMe3 and [tris(N-neohexyl)Cn]RhMe3 reported
earlier.6 The Rh-C lengths, though, are ca. 0.05 Å
shorter than those of the previous two, as is reasonable
for the change in hybridization from sp3 to sp2 in the
Discu ssion
It has been reported that [CnRhMe2(η2-CH2dCH2)]+
is observed by NMR at -40 °C and that above that
(11) (a) Werner, H.; Wolf, J .; Schubert, U.; Ackermann, K. J .
Organomet. Chem. 1986, 317, 327. (b) Marder, T. B.; Chan, D. M.-T.;
Fultz, W. C.; Milstein, D. J . Chem. Soc., Chem. Commun. 1988, 996.
(c) Wiedemann, R.; Steinert, P.; Schafer, M.; Werner, H. J . Am. Chem.
Soc. 1993, 115, 9864. (d) Dunaj-J urco, M.; Kettmann, V.; Steinborn,
D.; Ludwig, M. Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 1994,
50, 1427.
(12) (a) Tanaka, I.; J in-No, N.; Kushida, T.; Tsutsui, N.; Ashida, T.;
Suzuki, H.; Sakurai, H.; Moro-Oka, Y.; Kawa, T. I. Bull. Chem. Soc.
J pn. 1983, 56, 657. (b) Murray, B. D.; Power, P. P. Organometallics
1984, 3, 1199. (c) Batsanov, A. S.; Struchkov, Y. T. J . Organomet.
Chem. 1984, 265, 305. (d) King, S. A.; Van Engen, D.; Fischer, H. E.;
Schwartz, J . Organometallics 1991, 10, 1195. (e) Paiaro, G.; Pandolfo,
L.; Ganis, P.; Valle, G.; Fioretto, S. D. J . Organomet. Chem. 1995, 488,
249.

Reactions of (Trimethyltriazacyclononane)Rh(vinyl)3
Organometallics, Vol. 17, No. 24, 1998 5401
Sch em e 2
their mechanisms were not examined.14 To continue,
ethylene could undergo insertion into the Rh-vinyl
bond, and hydride rearrangements would then give
butenyl complex 4. At least one apparent precedent
exists for this rearrangement in iridium chemistry.15
On the other hand, as mentioned in the Introduction
and illustrated in eq 1, protonation of vinyl-metal
complexes is a known method of generating alkylidene
complexes1 and forms the basis of an alternative mech-
anism shown in Scheme 2, which shall be called the
“alkylidene path”. The divinyl(alkylidene)Rh cation 8
would undergo rearrangement with migration of a vinyl
group to the alkylidene carbon. The resulting butenyl
species need only coordinate the other two carbons to
form 4.
The ethylene and alkylidene paths can be easily
distinguished in principle by the use of deuterated acid.
Schemes 1 and 2 include the predicted labeling results
assuming the system would be well behaved, that is,
that it would not exhibit randomization of the label by
irrelevant paths. Dark circles represent positions where
part of the monodeuteration label would reside if 1 were
cleanly monodeuterated. The ethylene path should
generate ethylene-d1 coordinated to rhodium regardless
of whether the vinyl protonolysis would proceed by
initial metal protonation or by direct attack on the
Rh-C bond. Because of the symmetry of 7-d1, in effect
it is labeled in two sites, and the subsequent ethylene
insertion would lead to 4-d1, which would be labeled in
at least two sites as shown. Depending on how the
hydride shifting would occur, the label could be more
widely dispersed, but two would be the minimum
number of sites containing label. Now, considering the
alkylidene path and assuming that the deuteration were
irreversible, ethylidene cation 8-d1 would be labeled in
the methyl group only, and the subsequent vinyl migra-
tion and η3-coordination would leave the label unmoved.
Our attempts to deuterate trivinyl 1 with deuterated
triflic acid were frustrated by the tendency of very small
amounts of such a strong acid to undergo exchange with
adventitious proton sources. The problem would un-
doubtedly be exacerbated by a large kinetic isotope effect
favoring transfer of any H+ over D+. It was fortunate
F igu r e 2. The molecular structure (30% probability el-
lipsoids) of the heavy atoms of [CnRhMe(η3-CH2CHCHMe)]+
(6). The BAr4- anion and the second site of the disordered
methyl group, C(5A), bonded to C(2), have been omitted
for clarity.
Ta ble 2. Selected Bon d Len gth s (Å) a n d Bon d
An gles (d eg) of [Cn Rh Me(η3-CH2CHCHMe)][BAr 4]
(6)
Rh-N(1)
Rh-N(2)
Rh-N(3)
Rh-C(1)
Rh-C(2)
Rh-C(3)
Rh-C(4)
C(2)-C(3)
C(3)-C(4)
C(4)-C(5)
C(2)-C(5A)
2.255(10)
2.177(11)
2.200(10)
2.025(11)
2.172(12)
2.078(15)
2.216(16)
1.379(18)
1.488(21)
1.375(22)
1.344(23)
N(1)-Rh-N(2)
N(2)-Rh-N(3)
N(1)-Rh-N(3)
C(1)-Rh-C(2)
C(1)-Rh-C(3)
C(1)-Rh-C(4)
N(1)-Rh-C(1)
N(2)-Rh-C(2)
N(3)-Rh-C(4)
C(2)-C(3)-C(4)
C(3)-C(4)-C(5)
C(5A)-C(2)-C(3)
81.3(4)
80.1(4)
81.2(4)
92.6(5)
77.0(3)
93.3(5)
171.7(4)
174.7(4)
170.8(5)
127.6(12)
137.7(14)
137.4(15)
Sch em e 1
temperature it undergoes insertion of the ethylene
ligand into the Rh-Me bond, ultimately giving [CnRh-
(H)(η3-CH2CHCH2)]+.13 In view of these facts, on find-
ing that triflic acid cleavage of CnRh(CHdCH2)3 (1) does
not yield CnRh(CHdCH2)2(OTf) but instead gives [CnRh-
(CHdCH2)(η3-CH2CHCHMe)]+ (4), one might propose
the mechanism shown in Scheme 1, which shall be
called the “ethylene path”. In this sequence, the protic
cleavage would generate ethylene, but rather than
leave, the ethylene could remain π-coordinated to the
metal as shown in intermediate 7. Examples of protic
cleavage of metal vinyl groups are known, although
(14) (a) Tripathy, P. B.; Renoe, B. W.; Adzamli, K.; Roundhill, D.
M. J . Am. Chem. Soc. 1971, 93, 4406. (b) Mann, B. E.; Shaw, B. L.;
Tucker, N. I. J . Chem. Soc., A 1971, 2667. (c) Trocha-Grimshaw, J .;
Henbest, H. B. J . Chem. Soc., Chem. Commun. 1968, 757.
(15) (Tpb*)Ir(H)(CHdCH2)(CH2dCH2) (tpb* ) tris(3,5-dimethylpyra-
zolyl)borate anion) rearranges to (tpb*)Ir(H)(η3-CH2CHCHMe) via a
postulated ethylene insertion into the Rh-vinyl bond: Boutry, O.;
Gutierrez, E.; Monge, A.; Nicasio, M. C.; Perez, P. J .; Carmona, E. J .
Am. Chem. Soc. 1992, 114, 7288.
(13) (a) Wang, L.; Flood, T. C. J . Am. Chem. Soc. 1992, 114, 3169.
(b) Wang, L.; Lu, R. S.; Bau, R.; Flood, T. C. J . Am. Chem. Soc. 1993,
115, 6999.

5402 Organometallics, Vol. 17, No. 24, 1998
Zhen et al.
Sch em e 3
that the protonation could be effected by methanol-d4
as solvent, since it gave clean monodeuteration. How-
ever, the counterion could not be determined in solution,
and solvent removal gave only intractable materials.
The obvious candidate for counterion is CD3O-, but it
was surprising to us that such a strong nucleophile
would not exhibit some kind of reactivity with the π-allyl
cation. That the product in methanol clearly was the
(π-butenyl)rhodium cation is in contrast to similar
systems that undergo nucleophilic additions to either
the central or terminal carbon of π-allyl cations of Rh,
Ir, and Ru.16
The presence of only one deuterium in only the methyl
group of the butenyl ligand gives a clear answer that
the formation of 4 proceeds via an alkylidene path.
Since no label was observed in remaining starting
material late in the reaction (in the slow reaction in
CD3OD) or in the CH2 of the vinyl group in 4, it is also
clear that protonation of the vinyl group is irreversible;
that is, migration of the vinyl group to the alkylidene
carbon in 8-d1 is much faster than deprotonation of 8-d1
back to 1.
Sch em e 4
Attempts to observe any alkylidene-containing inter-
mediates in the protonation of 1 at temperatures as low
as 210 K by NMR spectroscopy revealed no signals
between δ 8 and 20 ppm in the 1H spectrum and no
resonances between δ 170 and 350 ppm in the 13C
spectrum.9 In addition, starting material was present
in slowly diminishing quantity between 210 and 240 K.
Thus, vinyl group migration to the alkylidene carbon is
also faster than the initial alkylidene-forming protona-
tion. To conclude that vinyl migration to the ethylidene
carbon in 6 is faster than both protonation of 1 and
deprotonation of 8 unfortunately is not very informative,
since proton transfers to and from carbon when they
involve rehybridization of the carbon center can be
several orders of magnitude slower than the diffusion
limit.17 The same is true for proton transfers to and
from metal centers.18
So far, rearrangement of ethylidene intermediate 8
to product 4 has been rationalized as migration of a
vinyl group to the alkylidene carbon. Another possible
mechanism for this rearrangement is an electrocyclic
ring closure followed by rearrangement of the resulting
metallacyclobutene to the π-allyl product. These two
mechanisms can be differentiated by an appropriate
labeling experiment. Scheme 3 shows the results
predicted for each mechanism in the case where the
label is a methyl group as in CnRh(cis-CHdCHMe)3 (2).
Protonation of 2 presumably generates the propylidene
intermediate [CnRh(dCHEt)(CHdCHMe)2]+, repre-
sented in Scheme 3 by the partial structure 9, and the
vinyl migration path then is predicted to afford hexenyl
product 5a . Electrocyclic ring closure would generate
metallacycle 10, which, because hydrogen is well-known
to migrate faster than methyl, would result in the
2-methylpentenyl product 5b. The experimentally ob-
served formation of only 5a gives an unequivocal answer
regarding this rearrangement. The geometry observed
for the η3-hexenyl ligand can be rationalized to result
from a series of η3-η1-η3 interconversions with C-C
bond rotations in the η1 intermediates.
The nature of the hydrocarbyl migration to the
alkylidene ligand is of particular interest. Specifically,
if a vinyl group could be protonated in the presence of
two other hydrocarbyl ligands which were different from
one another, then in principle a series of relative
migratory aptitudes could be measured. This would
depend, of course, on one’s ability to synthesize the
mixed-ligand starting materials, on selective protona-
tion of the vinyl ligand, and on the ultimate products
of the migrations being unique and characterizable. To
determine the feasibility of such a competition, CnRhMe-
(CHdCH2)2, 3, was prepared and its protonation exam-
ined. The predicted ethylidene intermediate, 11, and
subsequent migratory competition is depicted in Scheme
4. The fact that π-allyl complex 6 was formed in 96%
yield by NMR and was the only product detectable
indicates that the vinyl/methyl migratory rate ratio
must be at least 24.
(16) (a) Periana, R. A.; Bergman, R. G. J . Am. Chem. Soc. 1986,
108, 7346. (b) Tjaden, E. B.; Stryker, J . M. Organometallics 1992, 11,
16. (c) McNeill, K.; Andersen, R. A.; Bergman, R. G. J . Am. Chem.
Soc. 1995, 117, 3625. (d) Curtis, D. M.; Eisenstein, O. Organometallics
1984, 3, 887.
(17) J encks, W. P. Catalysis in Chemistry and Enzymology; McGraw-
Hill: New York, 1969; pp 175-178.
(18) Edidin, R. T.; Sullivan, J . M.; Norton, J . R. J . Am. Chem. Soc.
1987, 109, 3945.
A referee suggested the possibility that the methyl
migration might be faster than vinyl migration, but if
it were reversible and only the vinyl migration product
were stable, then the observed result could be mis-
leading. Subsequent observation that dissolution of 3

Reactions of (Trimethyltriazacyclononane)Rh(vinyl)3
Organometallics, Vol. 17, No. 24, 1998 5403
in CD3OD yields only [CnRh(η3-CH2CHCH-CH2D)Me]+
(6-d1) is inconsistent with this hypothesis. Thus,
reversible formation of the intermediate {CnRh-
(CHdCH2)[CH(CH3)(CH2D]}+ would require the forma-
tion of some [CnRh(η3-CH2CHCH-Me)(CH2D)]+ product.
No deuterium is observed in the RhMe group of 6-d1.
Thus, it is confirmed that 6 is the kinetic product and
vinyl group migration is faster than methyl group
migration.
therefore resorted to the alternative of high-resolution FAB
MS, which were recorded at the Mass Spectrometry Facility
of the University of California at Riverside. 1H and 13C NMR
spectra of 1, 2, and 3 have been submitted with the Supporting
Information as a demonstration of bulk purity. The hydro-
carbyl ligand 1H and 13C NMR spectra given below for 1-6
are all labeled corresponding to the structures shown here,
independent of X-ray structure labeling and independent of
the actual combination of ligands in a given complex.
It is well-known in organic chemistry that sp2 carbon
centers migrate better than sp3.19 This is generally
attributed to the greater steric accessibility of an sp2
carbon at its π face. In addition, π participation during
migration allows delocalization into the π system of any
electron deficiency. Such delocalization can be so
stabilizing that it leads to a discrete intermediate with
σ bonding to both the departed and receiving atoms, as
in the case of the well-known phenonium ion intermedi-
ates which intervene in phenyl migrations of carboca-
tions. In the last mentioned systems, phenyl groups
migrate much faster than either alkyl groups or hy-
drides.19 Thus it is consistent with these precedents
that sp2 vinyl should migrate faster than sp3 methyl to
the electrophilic alkylidene group in intermediate 11.
In summary, alkenyl rhodium groups in CnRhR-
(CHdCHR′)2 are efficiently and irreversibly protonated
at the â position to form electrophilic rhodium alky-
lidene intermediates. These intermediates show a very
high reactivity for intramolecular migration of a second
vinyl group to the ethylidene carbon. In an intramo-
lecular competition for migration of a methyl or vinyl
group to the ethylidene carbon, vinyl migration is at
least an order of magnitude faster.
Im p r oved P r ep a r a tion of Cn Rh Me3. A solution of 5
mmol of MgMe2 (prepared by the standard dioxane method)
in 30 mL of THF was added dropwise to a stirred suspension
of CnRhCl3 (0.9 g, 2.4 mmol)4a in 20 mL of THF, and the
mixture was then heated at reflux for 12 h, during which the
mixture turned to a clear yellow solution. Wet THF (THF/
H2O ca. 3:1 by volume) was added dropwise until gas evolution
ceased. From this point, the mixture can be manipulated in
air. The solvent was removed at reduced pressure, and the
residue was extracted with 70 mL of CH2Cl2 by stirring for
0.5 h at 22 °C. The extract was filtered and the solvent
removed at reduced pressure and then under vacuum to give
1
the light yellow product in 92% yield. The H NMR shows no
CnRhMeCl2 and only a trace of CnRhMe2Cl, so the material
is suitable for most purposes. If purer material is desired, it
can be recrystallized from benzene, in which CnRhMe2Cl is
much less soluble.
Exp er im en ta l Section
Cn Rh (CHdCH2)3 (1). A THF solution of vinyllithium (25
mL of a 0.26 M solution, 6.5 mmol) was added slowly at -78
°C to a stirred suspension of CnRhCl3 (0.42 g, 1.3 mmol)4a in
45 mL of THF under an N2 atmosphere. The mixture was
allowed to warm to 22 °C, and the clear solution was stirred
for an additional 25 min. Then CH2Cl2 was slowly added until
gas evolution ceased. Volatiles were evaporated at reduced
pressure, and the residue was extracted with 30 mL of
CH2Cl2. The extract was filtered and the solvent was evapo-
rated at reduced pressure to give a light brown solid (0.36 g,
93%). Slow evaporation of a CH2Cl2 solution of product gave
off-white rectangular prisms, which were collected and washed
with ether. 1H NMR (DMSO-d6, 360 MHz): δ 2.53 (s, NCH3),
2.47-2.74 (m, NCH2), 4.78 (ddd, J ab ) 3.5 Hz, J ac ) 17 Hz,
J RhH ) 1.9 Hz, Ha), 5.20 (ddd, J bc ) 10 Hz, J RhH ) 3.8 Hz, Hb),
7.33 (ddd, J RhH ) 3.0 Hz, Hc). 13C{1H} NMR (DMSO-d6, 90
MHz): δ 47.73 (NCH3), 56.83 (NCH2), 115.12 (Câ), 166.51 (d,
J RhC ) 42 Hz, CR). Anal. Calcd for C15H30N3Rh: C, 50.70; H,
8.51. Found: C, 48.88; H, 8.48. FAB mass spectrum: calcd
for [C15H30N3Rh]H+ 356.1573; found 356.1566 (2 ppm); also
m/ e ) 328 (20%) [M+ - (C2H3)].
Cn Rh (Z-CHdCHMe)3 (2). A Schlenk flask charged with
CnRhCl3 (300 mg, 0.79 mmol) and cis-CH3CHdCHLi (260 mg,
5.4 mmol) was placed in an acetone/dry ice bath, and 60 mL
of ether was added. This mixture was stirred for 30 min at
-78 °C and 3 h at 0 °C, and the solvent was removed under
vacuum. The residue was extracted with dry benzene (3 ×
25 mL), and solvent was again removed, leaving a dark solid.
The solid was washed with cold methanol and dried to afford
34 mg (9%) of crude product. Colorless crystals were obtained
either by slow evaporation of an acetone solution or by slow
diffusion of pentane into an ether solution. 1H NMR (acetone-
Gen er a l. All reactions involving organometallic com-
pounds, unless otherwise mentioned, were carried out under
an atmosphere of N2 or Ar purified over reduced Cu catalyst
(BASF R3-11) and Aquasorb. Flamed-out glassware and
standard vacuum line, Schlenk, and N2-atmosphere box tech-
niques were employed. Benzene, ether, hexanes, pentanes,
and THF were distilled from purple solutions of sodium/
benzophenone, and CH2Cl2 was distilled twice from CaH2.
Vinyllithium20a and Z-propenyllithium20b were prepared by
standard procedures. Elemental analyses were performed by
Galbraith Laboratories, Knoxville, TN. Hydrogen combustion
analyses of CnRh and (tacn)Rh molecules generally are near
or within convention tolerances, but there have been problems
obtaining acceptable C analyses, particularly for neutral
hydrocarbyls, which are usually 1-4% low. Rhodium com-
pounds are prone to formation of nitrides and carbides that
are stable at the combustion temperatures accessible with
conventional instruments. The problem is highly dependent
on the structure of the compound involved. In the present
case, all three of the salts [CnRh(hydrocarbyl)(allylic)][BAr4]
yielded acceptable results with the submission of a single
sample. Neither CnRh(vinyl)3 (1) nor CnRh(propenyl)3 (2)
yielded any carbon analysis closer than 2% (always low) after
multiple sample submissions. Vigorous conditions and pro-
moters were used, but with no improvement. We have
(19) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper & Row: New York, 1987. Carey,
F. A.; Sundberg, R. J . Advanced Organic Chemistry, Part A: Structure
and Mechanism, 3rd ed.;., Plenum Press: New York, 1984.
(20) (a) Seyferth, D.; Weiner, M. A. J . Am. Chem. Soc. 1961, 83,
3585. (b) Seyferth, D.; Vaughan, G. L. J . Am. Chem. Soc. 1964, 86,
883.

5404 Organometallics, Vol. 17, No. 24, 1998
Zhen et al.
d6, 360 MHz): δ 1.70 (dd, J ab ) 6.5 Hz, J ac ) 1.3 Hz, CH3(a)),
mg, 0.044 mmol) under N2 in a Schlenk flask in a -78 °C bath.
The mixture was stirred at room temperature for 2 days.
Volatiles were removed in vacuo to afford 43 mg (77% yield)
of a light yellow solid. The crude was extracted with 25 mL
of Et2O, the solvent was evaporated, and the residue was
washed with pentane. The solid was dissolved in Et2O, and
diffusion of pentane into the solution gave colorless rectangular
crystals. Cation resonances: 1H NMR (acetone-d6, 360 MHz)
δ 1.18 (t, J ih1 ) J ih2 ) 7.4 Hz, CH3(i)), 1.75 (ddq, J gh1 ) 3.8 Hz,
J h1h2 ) 13 Hz, Hh1), 1.98 (ddq, J gh2 ) 3.8 Hz, Hh2), 1.42 (d, J de
) 6.8 Hz, CH3(e)), 1.70 (dd, J ab ) 6.8 Hz, J ac ) 1.4 Hz, CH3(a)),
2.67, 3.03, 3.17 (3s, NCH3), 2.4 (m, Hd and Hg), 2.65-3.70 (m,
NCH2), 4.62 (t, J df ) J fg ) 11 Hz, Hf), 5.37 (ddq, J bc ) 9.4 Hz,
J RhH ) 3.2 Hz, Hb), 6.34 (ddq, J RhH ) 5.2 Hz, Hc); 13C{1H} NMR
(acetone-d6, 90 MHz) δ 16.08 (C6), 17.38 (C5), 18.64 (Cγ), 26.59
(C1), 53.12, 53.52, 54.77 (3 NCH3), 58.01, 58.14, 58.76, 59.56,
60.85 64.10 (6 NCH2), 66.72 (d, J RhC ) 13 Hz, C2), 61.28 (d,
J RhC ) 7.9 Hz, C4), 108.83 (d, J RhC ) 6.6 Hz, C3), 127.95 (Câ),
145.28 (d, J RhC ) 30 Hz, CR). Anion resonances: 1H NMR
(acetone-d6) δ 7.67 (br s, 4H, Hpara), 7.79 (br s, 8H, Hortho);
2.53 (s, NCH3), 2.82 (s, NCH2), 5.90 (ddq, J bc ) 10 Hz, J RhHb
)
3.4 Hz, Hb), 6.94 (ddq, J RhHc ) 4.0 Hz, Hc). 13C{1H} NMR
(acetone-d6, 90 MHz): δ 19.70 (Cγ), 48.81 (NCH3), 57.87
(NCH2), 128.16 (Câ), 150.88 (d, J RhC ) 42 Hz, CR). The 1H NMR
revealed ca. 10% of resonances corresponding to CnRh(Z-CHd
CHMe)2(E-CHdCHMe). Anal. Calcd for C18H36N3Rh: C,
54.40; H, 9.13. Found: C, 52.69; H, 9.18. FAB mass spec-
trum: calcd for [C18H36N3Rh]H+ 398.2043; found 398.2036 (2
ppm); also m/ e ) 356 (15%) [M+ - (C3H5)].
Cn Rh (CHdCH2)2Me (3). A THF solution of vinyllithium
(20 mL of a 0.26 M solution, 5.2 mmol) was added slowly to a
stirred suspension of CnRhMe(OTf)2 (0.50 g, 0.85 mmol)6 in
25 mL of THF at -78 °C under an N2 atmosphere. The
mixture was stirred for 5 min at -78 °C and then allowed to
warm to 22 °C. The suspension quickly changed to a clear
yellow solution. After stirring an additional 30 min, the
volatiles were removed under vacuum and the resulting solid
was extracted with 50 mL of benzene. Removal of the benzene
gave 0.21 g (72%) of a yellow solid. This molecule slowly
decomposes with liberation of Cn ligand at 22 °C in the solid
state (t1/2 ∼ 1 week) or in solution (t1/2 ∼ 2 days), so further
13C{1H} NMR (acetone-d6) δ 118.42 (Cpara), 125.40 (q, J FC
)
273 Hz, CF3), 130.06 (q, J FC ) 33 Hz, Cmeta), 135.54 (Cortho),
162.60 (1:1:1:1 q, J BC ) 50 Hz, Cipso). Anal. Calcd for
1
purification and combustion analysis were not attempted. H
C
50H49N3F24BRh: C, 47.60; H, 3.91. Found: C, 47.32; H, 4.25.
NMR (C6D6, 250 MHz): δ 0.57 (d, J RhH ) 2.5 Hz, RhCH3),
1.50-1.65 (m, 6H, NCH2), 1.86-2.10 (m, 6H, NCH2), 2.29 (s,
6H, NCH3), 2.42 (s, 3H, NCH3), 5.63 (ddd, 2H, J ab ) 3.3 Hz,
J ac ) 17 Hz, J RhH ) 2.0 Hz, Ha), 6.18 (ddd, 2H, J bc ) 10 Hz,
J RhH ) 3.4 Hz, Hb), 7.80 (ddd, 2H, J RhH ) 3.1 Hz, Hc). 13C{1H}
NMR (C6D6, 90 MHz): δ 3.61 (d, J RhC ) 37 Hz, RhCH3), 48.20
(2 NCH3), 48.46 (1 NCH3), 56.57, 56.83, 57.03 (each 2 NCH2),
116.34 (Câ), 167.66 (d, J RhC ) 42 Hz, CR). Integration of the
1H and 13C NMR spectra indicated that crude 3 was at least
95% pure. FAB mass spectrum: calcd for [C14H30N3Rh]H+
344.1573; found 344.1571 (1 ppm); also m/ e ) 328 (20%) [M+
- (CH3)].
FAB mass spectrum: calcd for [C18H37N3Rh]+ 398.2043; found
398.2053 (3 ppm).
[Cn Rh Me(η3-CH2CHCHMe][BAr 4], 6(BAr 4). A 40 mL
portion of CH2Cl2 at -78 °C was transferred by cannula to a
mixture of CnRh(CHdCH2)2Me (3, 0.17 g, 0.50 mmol) and
HBAr4‚(Et2O)2 (0.47 g, 0.47 mmol) under N2 in a Schlenk flask
in a -78 °C bath. The mixture was stirred for 5 min and then
allowed to warm to 22 °C. The resulting bright yellow solution
was stirred overnight. Volatiles were removed in vacuo, and
the resulting yellow solid was extracted with 25 mL of Et2O
to afford 0.47 g (84% yield) of a pale yellow solid. Diffusion of
hexane vapor into a CH2Cl2 solution of 6 afforded off-white
P r oton a tion
Exp er im en ts.
[Cn Rh (CHdCH2)(η3-
crystals. 1H NMR (DMSO-d6, 250 MHz): δ -0.14 (d, J RhH
)
CH2CHCHMe)][BAr 4], 4(BAr 4). A 15 mL portion of CH2Cl2
was added to a mixture of CnRh(CHdCH2)3 (1, 0.12 g, 0.33
mmol) and HBAr4‚(Et2O)2 (0.32 g, 0.32 mmol)8 under N2 in a
Schlenk flask in a -78 °C bath. The mixture was stirred for
5 min and then allowed to warm to 22 °C. The resulting bright
yellow solution was stirred overnight. Volatiles were removed
in vacuuo to afford 0.30 g (79% yield) of a slightly yellow solid.
Use of HCl or HOTf in ether solution to react with 1 also
yielded 4(Cl) or 4(OTf). Diffusion of hexane vapor into a
CH2Cl2 solution of 4 afforded off-white crystals. Cation
1.8 Hz, RhCH3), 1.14 (d, J gh ) 6.0 Hz, Hh), 1.58 (d, J df ) 11
Hz, Hd), 2.03, 2.59, 2.73 (3s, NCH3), 2.40 (dq, J fg ) 11 Hz, Hg),
2.5-3.4 (m, NCH2), 2.63 (d, J ef ) 7.4 Hz, He), 3.68 (dt, J ef
)
7.4 Hz, J df ) J fg ) 11 Hz, Hf). 13C{1H} NMR (DMSO-d6, 90
MHz): δ 3.85 (d, J RhC ) 25 Hz, RhCH3), 15.62 (C5), 44.84 (d,
J RhC ) 14 Hz, C2), 47.22, 50.81 53.51 (3 NCH3), 56.21, 56.69,
58.03, 58.55 60.76, 61.13 (6 NCH2), 58.50 (d, J RhC ) 11 Hz,
C4), 107.01 (d, J RhC ) 6.5 Hz, C3). BAr4 anion resonances as
above for 4. Anal. Calcd for C46H43N3F24BRh: C, 45.75; H,
3.59. Found: C, 45.64; H, 3.29.
resonances: 1H NMR (DMSO-d6, 250 MHz) δ 1.16 (d, J gh
)
6.0 Hz, butenyl CH3(h)), 1.74 (d, J df ) 11 Hz, Hd), 2.06, 2.77,
2.86 (3s, NCH3), 2.5-3.5 (m, NCH2), 2.39 (dq, J fg ) 11 Hz, Hg),
P r oton a tion s w ith Deu ter a ted Acid s. [Cn Rh (CHd
CH2)(η3-CH2CHCH-CH2D)]+ (4-d ). (a ) Deu ter a ted Tr iflic
Acid . DOTf was purchased from Aldrich and used as received.
A -78 °C solution of 50 mg of CnRh(CHdCH2)3 (1) in 10 mL
of CH2Cl2 was treated with 0.84 mL of a 0.1703 N solution of
DOTf in ether. The solution was allowed to warm to 22 °C
and was stirred for an additional hour. Solvent removal under
vacuum gave an off-white solid, which was examined by 1H
and 13C NMR in DMSO-d6. Over a period of hours, the initial
complex spectrum simplified to that of a single material.
Overall, the spectra were identical to those given above for 4,
except that the butenyl Me group was partially deuterated.
See the Results section for details. In a second experiment, a
-78 °C solution of 50 mg of CnRh(CHdCH2)3 (1) in 8 mL of
CH2Cl2 was treated directly with 12 µL of neat DOTf. The
solution was allowed to warm to 22 °C and was stirred for an
additional hour. Solvent removal under vacuum gave an off-
3.00 (d, J ef ) 7.8 Hz, He), 4.33 (ddd, J df ) J fg ) 11 Hz, J ef
)
7.8 Hz, Hf), 4.74 (d, J ac ) 16 Hz, Ha), 5.13 (d, J bc ) 8.5 Hz,
Hb), 6.51 (ddd, J RhH ) 4.6 Hz, Hc). 13C{1H} NMR (DMSO-d6,
90 MHz) δ 16.51 (C5), 46.64 (d, J RhC ) 14 Hz, C2), 47.49, 50.13,
52.97 (3 NCH3), 55.66, 57.38, 57.73, 58.76, 59.32, 61.18 (6
NCH2), 60.17 (d, J RhC ) 11 Hz, C4), 109.28 (d, J RhC ) 4.9 Hz,
C3), 119.32 (Câ), 156.44 (d, J RhC ) 26 Hz, CR). Anion reso-
nances: 1H NMR (DMSO-d6) δ 7.60 (br s, 8H, Hortho), 7.66 (br
s, 4H, Hpara); 13C{1H} NMR (DMSO-d6) δ 117.64 (Cpara), 123.98
(q, J FC ) 273 Hz, CF3), 128.46 (q, J FC ) 33 Hz, Cmeta), 134.00
(Cortho), 160.92 (1:1:1:1 q, J BC ) 50 Hz, Cipso). Anal. Calcd for
C
47H43N3F24BRh: C, 46.29; H, 3.55. Found: C, 45.95; H, 3.65.
NMR Sea r ch for [Rh dCHCH3]+. CnRh(CHdCH2)3, 1 (10
mg), and solid HBAr4‚(Et2O)2 (28 mg) were combined in an
NMR tube, 0.5 mL of CD2Cl2 was added by vacuum transfer,
and the tube was sealed. The cold sample was transferred to
an NMR probe, which was precooled at 210 K. 1H and 13C
NMR spectra were recorded at 210, 220, 240, and 300 K. See
the Results section for details.
[Cn Rh (CHdCHMe)(η3-MeCHCHCHEt][BAr 4], 5(BAr 4).
A 10 mL portion of CH2Cl2 was added to a mixture of CnRh-
(CHdCHMe)3 (2, 18 mg, 0.045 mmol) and HBAr4‚(Et2O)2 (45
1
white solid for which the H and 13C NMR (DMSO-d6) spectra
were identical to those just described except that the level of
monodeuteration was now ca. 50%.
(b) Rea ction of 1 in Meth a n ol-d 4. About 10 mg of 1 was
suspended in CD3OD in an NMR tube. 1H and 13C NMR
spectra were acquired periodically. The solid dissolved slowly
as the reaction progressed, and after ca. 3 days the solid was

Reactions of (Trimethyltriazacyclononane)Rh(vinyl)3
Organometallics, Vol. 17, No. 24, 1998 5405
Ta ble 3. Selected Cr ysta llogr a p h ic In for m a tion
for Str u ctu r es of Cn Rh (CHdCH2)3 (1) a n d
[Cn Rh Me(η3-CH2CHCHMe)][BAr 4] (6)
gone and NMR showed the reaction to be complete and clean.
13C{1H} NMR, hydrocarbyl ligand resonances (methanol-d4, 90
MHz): δ 16.82 (1:1:1 t, J CD ) 20 Hz, butenyl CH2D (C5)), 17.10
(s, butenyl CH3 (C5), ∆δ CH2D vs CH3 ) 0.28 ppm), 47.82 (d,
J RhC ) 15 Hz, C2), 62.31 (d, J RhC ) 11 Hz, C4), 111.43 (d, J RhC
) 5.1 Hz, C3), 120.61 (Câ), 156.62 (d, J RhC ) 28 Hz, CR). The
counterion could not be determined, and solvent removal led
to decomposition.
(c) Exch a n ge Con tr ol of Bu ten yl Com p lex 6 in Meth -
a n ol-d 4. About 10 mg of the butenyl complex 6(BAr4) was
dissolved in CD3OD in an NMR tube and was let stand for
several days at room temperature. The 1H NMR revealed
slight changes in the chemical shifts from the same material
in DMSO-d6, but no changes in intensity or coupling pattern
were observed, indicating no H/D exchange with the solvent.
1H NMR (methanol-d4, 360 MHz): δ 0.03 (d, J RhH ) 1.8 Hz,
RhCH3), 1.28 (d, J gh ) 6.0 Hz, Hh), 1.72 (d, J df ) 11 Hz, Hd),
2.13, 2.73, 2.85 (3s, NCH3), 2.40 (dq, J fg ) 11 Hz, Hg), 2.7-3.4
(1)
(6)
empirical formula
cryst syst
space group
a (Å)
b (Å)
c (Å)
C
15H30N3Rh
C46H43N3F24BRh
orthorhombic
Pbca (No. 61)
17.755(2)
19.718(3)
28.136(3)
monoclinic
P21/n (No. 14)
7.679(2)
15.828(2)
13.505(2)
92.59(2)
1639.8(5)
4
â (deg)
volume (Å3)
Z
9850(2)
8
fw
355.3
1.439
183
2455
1708
172
R ) 0.0330
Rw ) 0.0371
1207.6
1.629
163
6173
3560
884
R ) 0.0623
Rw ) 0.0880
calcd density (g/cm3)
temp (K) of collection
reflns collcd
obs reflns [F > 4σ(F)]
no. of params refined
final R indices (obs data)
(m, NCH2), He obscured by NCH2 resonances, 3.86 (dt, J ef
7.4 Hz, J df ) J fg ) 11 Hz, Hf).
)
(d ) Rea ction of 3 in Meth a n ol-d 4. About 10 mg of 3 was
suspended in CD3OD in an NMR tube. 1H and 13C NMR
spectra were acquired periodically. The solid dissolved slowly
as the reaction progressed, and after ca. 2 days the solid was
gone and NMR showed the reaction to be essentially complete.
1H NMR (methanol-d4, 360 MHz): only resonances at δ 1.28
(Hh normally a doublet, now a multiplet) and δ 2.40 (Hg
normally a dq, now a dt from Hf and CH2D (Hh)) show changes
from those of control experiment (c) above: in particular, no
changes were detectable in Hd or Hf. Although He was
obscured by NCH2, the observation of no change in the Hd and
Hf resonances rules out label incorporation into the He site.
Observation of no change in the Hf resonance rules out any
deuteration in the Hg site. There was no deuterium incorpora-
tion into the RhCH3 group.
two independent sets of partially populated F3 groups. All
F-F distances (within F3 groups) and all C-F distances in
the structure were constrained to refine to single values, but
the F-F distances between F3 sets on a given CF3 carbon were
not constrained. In this way, the anion refined very well. The
cation, however, did not refine as well. It showed signs of
several types of disorder, especially the methyl group of the
butenyl ligand, which was disordered at the two ends of the
allyl group. Refinement of both methyl sites led to C(5)/C(5A)
populations of 0.56/0.44. There was also some evidence for
additional orientations of the allyl group and certain carbons
of the Cn ligand, but these did not refine properly and so were
finally ignored.
With the minimal accommodation for butenyl methyl dis-
order, the final parameters for the cation of 6 were ultimately
good enough to warrant reporting. A structural drawing of 6
is shown in Figure 2; selected bond lengths and angles of 6
are given in Table 2; Table 3 lists selected crystallographic
information about the structure determination. Complete
crystallographic data for 6 appear in the Supporting Informa-
tion.
X-r a y Sin gle-Cr ysta l An a lysis of 1. X-ray diffraction
data were acquired at -90 °C on a single crystal of 1, and
solution of the data proceeded uneventfully via standard
heavy-atom techniques. The structure was refined anisotro-
pically to final agreement factors of R ) 3.30% and Rw
)
3.71%. A structural drawing of 1 is shown in Figure 1; selected
bond lengths and angles of 1 are given in Table 1; Table 3
lists selected crystallographic information about the structure
determination. Complete crystallographic data for the mol-
ecule appear in the Supporting Information.
Ack n ow led gm en t. This work was supported by the
National Science Foundation. We thank Prof. Robert
Bau for helpful discussions.
X-r a y Sin gle-Cr ysta l An a lysis of 6(BAr 4). X-ray diffrac-
tion data were acquired at -110 °C on single crystals of
6(BAr4). Solution of the data gave poor-quality structures
despite low-temperature data collection on two or three
different crystals. During solution of the structure, the large
BAr4- anion presented itself almost intact from the first E-map
phased by a direct methods calculation. Most of the CF3
groups were disordered. Each disordered CF3 was refined with
Su p p or tin g In for m a tion Ava ila ble: Complete crystal-
lographic data on 1 and 6 and 1H and 13C NMR spectra of 1,
2, and 3 (21 pages). Ordering information is given on any
current masthead page.
OM980423Y