Cross-coupling of cobalt-complexed propargyl radicals: Metal core- and α- and γ-aryl-induced chemo- and diastereoselectivity

Article abstract of DOI:10.1021/om070180m

Chemo- and diastereoselectivities of homo- and cross-coupling reactions of Co₂(CO)₆-complexed propargyl radicals were studied. The alternative radical generation methods - reduction of respective cations with Zn or Cp₂Co and one-step mediation with THF or Tf₂O - were employed. In the case of cobalt complexes with terminal triple bonds, the product distribution is nearly statistical and dependent upon the reducing agent with the concentration of the cross-coupling product 8 falling in the range 38-49%. The diastereoselectivity varied widely (de 40-92%) with the preponderant formation of d,l-diastereoisomers 7-9 in both homo- and cross-coupling reactions. The highest level of stereocontrol was achieved in THF- and Tf ₂O-mediated reactions (de up to 92%), while the reductions with Cp₂Co were inferior in both homo- and cross-coupling processes (de 40-56%). An introduction of a γ-aromatic ring revealed a kinetic differentiation at the radical generation step that, in turn, resulted in a nonstatistical distribution of homo- and cross-dimers. The d,l- diastereoselectivity is found to be systematically lower for homodimer 16, containing a γ-phenyl substituent at the triple bond (de: 16 14-64%; 7 52-84%). The observed chemoselectivities are accounted for on the basis of the computed values of charge distribution in the requisite cations.

Full text of DOI:10.1021/om070180m

Organometallics 2007, 26, 3173-3182
3173
Cross-Coupling of Cobalt-Complexed Propargyl Radicals: Metal
Core- and r- and γ-Aryl-Induced Chemo- and Diastereoselectivity†
Gagik G. Melikyan,* Arthur Floruti, Lucin Devletyan, Pogban Toure, Norman Dean, and
Louis Carlson
Department of Chemistry and Biochemistry, California State UniVersity Northridge,
Northridge, California 91330-8262
ReceiVed February 26, 2007
Chemo- and diastereoselectivities of homo- and cross-coupling reactions of Co2(CO)6-complexed
propargyl radicals were studied. The alternative radical generation methodssreduction of respective cations
with Zn or Cp2Co and one-step mediation with THF or Tf2Oswere employed. In the case of cobalt
complexes with terminal triple bonds, the product distribution is nearly statistical and dependent upon
the reducing agent with the concentration of the cross-coupling product 8 falling in the range 38-49%.
The diastereoselectivity varied widely (de 40-92%) with the preponderant formation of d,l-diastereoi-
somers 7-9 in both homo- and cross-coupling reactions. The highest level of stereocontrol was achieved
in THF- and Tf2O-mediated reactions (de up to 92%), while the reductions with Cp2Co were inferior in
both homo- and cross-coupling processes (de 40-56%). An introduction of a γ-aromatic ring revealed
a kinetic differentiation at the radical generation step that, in turn, resulted in a nonstatistical distribution
of homo- and cross-dimers. The d,l-diastereoselectivity is found to be systematically lower for homo-
dimer 16, containing a γ-phenyl substituent at the triple bond (de: 16 14-64%; 7 52-84%). The observed
chemoselectivities are accounted for on the basis of the computed values of charge distribution in the
requisite cations.
chemistry of organometallic radicals, in particular those with
an unpaired electron localized on the R-carbon atom in a
π-bonded ligand, has become an emerging interdisciplinary
field.2f The synthetic potential uncovered so far is truly
remarkable: it provides novel methods for inter- and intramo-
lecular radical C-C bond formation that readily occurs, in a
selective manner, in a diverse polyfunctional environment.2d-f,3
Intermolecular homo-coupling radical reactions constitute the
bulk of the experimental material reported so far with a
diastereomeric excess varying in the range 0-94%.3,4 The
Introduction
The ability to control the chemo-, regio-, and stereoselectivity
of radical reactions remains one of the main challenges of
modern synthetic chemistry.1 Among innovative approaches is
the coordination of organic moieties with transition metals,
which facilitates the generation of radical species and provides
for their moderation.2 The very topology of metal complexes
allows for altering of the electronic, steric, and conformational
parameters of π- and σ-bonded ligands by varying the nature
of the transition metal, its oxidation state, and attendant with
it, the mode of interaction with an organic moiety.2 The
(3) (Pentadienyl radical)Fe(CO)3: (a) Mahler, J. E.; Gibson, D. H.; Pettit,
R. J. Am. Chem. Soc. 1963, 85, 3959. (b) Sapienza, R. S.; Riley, P. E.;
Davis, R. E.; Pettit, R. J. Organomet. Chem. 1976, 121, C35. (c) Pearson,
A. J.; Chen, Y.-S.; Daroux, M. L.; Tanaka, A. A.; Zettler, M. J. Chem.
Soc., Chem. Commun. 1987, 155. (d) Hwang, W. S.; Liao, R. L.; Horng,
Y. L.; Ong, C. W. Polyhedron 1989, 8, 479. (propargyl radical)Cp2Mo2-
(CO)4: (e) Le Berre-Cosquer, N.; Kergoat, R.; L’Haridon, P. Organome-
tallics 1992, 11, 721. (benzyl radical)Cr(CO)3: (f) Top, S.; Jaouen, G. J.
Organomet. Chem. 1987, 336, 143. Ferrocenyl radical: (g) Cais, M.;
Eisenstadt, A. J. Chem. Soc. 1965, 30, 1148. (h) Cais, M.; Askhenazi, P.;
Dani, S.; Gottlieb, J. J. Organomet. Chem. 1977, 540, 127. CpCo(η5-C7H9)-
radical: (i) Geiger, W. E.; Genneth, Th.; Lane, G. A. Organometallics 1986,
5, 1352. (propargyl radical)Co2(CO)6: (j) Padmanabhan, S.; Nicholas, K.
M. J. Organomet. Chem. 1981, 212, 115.
(4) (a) Melikyan, G. G.; Vostrowsky, O.; Bauer, W.; Bestmann, H. J. J.
Organomet. Chem. 1992, 423, C24. (b) Melikyan, G. G.; Combs, R. C.;
Lamirand, J.; Khan, M.; Nicholas, K. M. Tetrahedron Lett. 1994, 363. (c)
Melikyan, G. G.; Deravakian, A. J. Organomet. Chem. 1997, 544, 143. (d)
Melikyan, G. G.; Bright, S.; Monroe, T.; Hardcastle, K. M.; Ciurash, J.
Angew. Chem., Int. Ed. 1998, 37, 161. (e) Melikyan, G. G.; Deravakian,
A.; Myer, S.; Yadegar, S.; Hardcastle, K. I.; Ciurash, J.; Toure, P. J.
Organomet. Chem. 1999, 578, 68. (f) Melikyan, G. G.; Sepanian, S.; Riahi,
B.; Villena, F.; Jerome, J.; Ahrens, B.; McClain, R.; Matchett, J.; Scanlon,
S.; Abrenica, E.; Paulsen, K.; Hardcastle, K. I. J. Organomet. Chem. 2003,
683, 324. (g) Melikyan, G. G.; Villena, F.; Sepanian, S.; Pulido, M.;
Sarkissian, H.; Florut, A. Org. Lett. 2003, 5, 3395. (h) Melikyan, G. G.;
Villena, F.; Florut, A.; Sepanian, S.; Sarkissian, H.; Rowe, A.; Toure, P.;
Mehta, D.; Christian, N.; Myer, S.; Miller, D.; Scanlon, S.; Porazik, M.
Organometallics 2006, 25, 4680.
* Corresponding author. E-mail: gagik.melikyan@csun.edu.
† This paper is dedicated to the memory of Dr. Alexander Tatosyan
(1952-2004), a talented scientist, a colleague, and a friend.
(1) (a) Curran, D. P. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Ed.; Pergamon Press: London, 1992, Vol. 5, pp 716, 780. (b) Malacria,
M. Chem. ReV. 1996, 96, 289. (c) Giese, B.; Kopping, B.; Gobel, T.;
Dickhaut, J.; Thoma, G.; Kulicke, K. J.; Trach, F. In Organic Reactions;
Paquette L., Ed.; John Wiley: New York 1996; Vol. 48, p 301. (d) Curran,
D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical Reactions;
VCH: Weinhein, 1997. (e) Sibi, M. P.; Porter, N. A. Acc. Chem. Res. 1999,
32, 163. (f) Stereoselective Radical Reactions. Sibi, M. P.; Ternes, T. R. In
Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley-VCH: New York, 2000.
(g) Radicals in Organic Synthesis; Sibi, M., Ed.; Wiley: New York, 2001;
Vol. 1. (h) Zard, S. Z. Radical Reactions in Organic Synthesis; Oxford
University Press, Inc.: New York, 2003; 256 pp. (i) Togo, H. AdVanced
Free Radical Reactions for Organic Synthesis; Elsevier: Amsterdam, 2004;
258 pp.
(2) (a) Organometallic Radical Processes; Trogler, W. C., Ed.; Elsevi-
er: Amsterdam, 1990; Chapters 3, 4, 9, 10. (b) Collman, J. R.; Hegedus,
L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of
Organotransition Metal Chemistry; University Science Books: Mill Valley,
CA, 1987. (c) Astruc, D. Acc. Chem. Res. 1991, 24, 36. (d) Astruc, D.
Electron Transfer and Radical Processes in Transition-Metal Chemistry;
VCH: New York, 1995; Chapters 3, 5, 6. (e) Torraca, K. E.; McElwee-
White, L. Coord. Chem. ReV. 2000, 206-207, 469. (f) Melikyan, G. G. In
Frontiers in Organometallic Chemistry; Nova Science Publishers: New
York, 2006; p 155, and references therein.
10.1021/om070180m CCC: $37.00 © 2007 American Chemical Society
Publication on Web 05/19/2007

3174 Organometallics, Vol. 26, No. 13, 2007
Melikyan et al.
highest stereoselectivityswith preponderant formation of d,l-
diastereomersswas observed in THF-mediated and spontaneous
dimerizations of Co2(CO)6-complexed propargyl radicals (de
90-94%).4b-e To the contrary, intermolecular radical cross-
coupling reactions, with an unpaired electron localized in an
R-position of a π-bonded ligand, remains practically unknown.2,3
Conceptually related are radical dimerizations of Fe(CO)3-
cyclohexadienyl complexes:3c,d the topology of a π-bonded
ligand is such that the isomeric radicals can be generated at
both ends of an unsaturated five-carbon moiety. Zinc-induced
reduction of the monosubstituted cyclohexadienyl cation gave
rise to isomeric 1,1- and 1,5-dimers (85:15; de 0%) with
preponderant formation of sterically hindered head-to-head
product.3c An electrochemical reduction of a 1,4-disubstituted
analogue lacked both regio- and stereoselectivity.3d Our interest
in this area stems from a systematic study on the chemistry of
transition metal-templated propargyl radicals and cations4
that led us to the discovery of the THF-mediated,4c,e,f and
spontaneous,4g generation and coupling of Co2(CO)6-coordinated
propargyl radicals. It is noteworthy that the level of stereocontrol
achieved in these reactions (up to 97% d,l-) remains unprec-
edented for intermolecular organometallic radical dimeriza-
tions.2,3 Herewith we present the first account on the chemo-
and diastereoselectiVity of the intermolecular radical cross-
coupling reactions of cobalt-complexed propargyl radicals. The
current study was undertaken to determine to what extent the
product distribution can be altered by varying the kinetic and
thermodynamic parameters of the requisite cations and radicals.
Treatment of alcohols 1 and 2 with tetrafluoroboric acid5
provided for concurrent generation of the cations 3 and 4
(Scheme 1). Their reduction with zinc (6-fold excess; 20 °C, 3
h) yielded propargyl radicals 5 and 6, the key transient species
previously employed in homo-coupling reactions.4b,e-g The
concurrent homo- and cross-coupling reactions gave rise to
respective dimeric products (7, 8, and 9) each represented by a
mixture d,l- and meso-diastereomers. Three kinetic parameters
need to be taken into consideration while considering the
chemoselectivity of these reactions: (1) the rate of cation
generation; (2) the rate of radical generation; and (3) the rate
of dimerization itself. Since the two-step experimental protocol
includes an isolation of the cations 3 and 4, the alleged disparity
in the rates of cation generation, caused by a para-substituent
at the aromatic nucleus, becomes a noncontributing factor. The
other two parameters, if nearly equal, should provide for a
statistical product distribution.6,7 By the NMR data, the ratio of
dimers 7:8:9 in the crude mixture was equal to 22:40:38 (Table
1), which may be caused by either the faster generation of radical
6 or a higher rate of its self-dimerization, or the combination
thereof. Given the fact that the para-methoxy group is remotely
located from the reaction site, the assumption was made that
the dimerization rate of radicals 5 and 6 cannot differ signifi-
cantly. To examine if the charge distribution in cations 3 and 4
does, in fact, impact the reduction kinetics, semiempirical and
an initio studies8 were carried out for cobalt-complexed cations
3 and 4 and their organic counterparts, 10 and 11 (Figure 1).
For the former, as expected, a positive charge on the R-carbon
atom is higher than that in p-OMe derivative 11, given the
electron-donating nature of the substituent (C1 +0.126233 vs
C1 +0.042917). The complexation, unexpectedly, reverses the
trend and the positive charge in cation 3 becomes significanty
lower when compared to that of cation 4 (C1 +0.004863 vs C1
+0.382988). A careful consideration of the computational data
reveals more differences in the charge distribution caused by
the interplay between an electron flow toward the cationic center
and the intensity of the back-bonding between the transition
metal and a π-bonded ligand. In particular, an electron donation
from the para-substituent increases the negative charge on the
aromatic carbon directly bonded to the cationic center (3 C1′
-0.086368; 4 C1′ -0.266173), making the C1-C1′ bond more
polar (∆δ 3 0.091231; 4 0.649161) and also shorter (C1-C1′ 3
1.464 Å, 4 1.405 Å). Concurrently, the back-bonding decreases
as indicated by an increased electron density over the cobalt
atoms (3 -0.404561 and -0.421442; 4 -0.472625 and
-0.479224). Overall, the metal cluster acts as an electron
reservoir by accommodating an additional electron density in
response to an electronic shift from the para-substituent and
toward the C1′ carbon atom. Most importantly, calculations
revealed that an increased positive charge on the cationic C1
atom in cation 4 is, most probably, responsible for the
nonstatistical product distribution in the zinc-mediated reaction.
The level of diastereoselection is comparable in homo- and
cross-dimeric products (Scheme 1; de 66-76%) with prepon-
derant formation of d,l-stereoisomers (d,l:meso: 7 83:17; 8 88:
12; 9 88:12). Stereochemical assignments are based on the X-ray
crystallography data4b,f and NMR signatures of methyne
protons.4c,f,g Given the complexity of the mixture, the isolation
of the individual stereoisomers was carried out, by preparative
TLC, in three steps, with no significant changes in the
stereoisomeric compositions (d,l:meso: 7 87:13; 8 82:18; 9 88:
12). Depicted in Figures 2 and 3 are staggered orientations of
Results and Discussion
The current level of knowledge on stereoelectronic parameters
of the π-bonded organometallic radicals is quite limited2,3,5 and
does not provide the reliable basis for predicting the chemo-
and stereoselectivity of the reactions. The hybridization of the
R-radical centerssp2 vs sp3sand, attendant with it, the very
shape of the immediate radical environmentsplanar vs
pyramidalsremain unknown. For comparison, the respective
oxidized species, the Co2(CO)6-doubly-stabilized propargyl
cation, is shown to maintain an sp2-configuration with an
R-cationic center being shifted toward one of the cobalt atoms.4d
Given the lack of literature precedence,2,3 the chemoselectivity
of the intermolecular cross-coupling reactions is hard to predict.
First, the “persistency” of carbon-based radicals6,7 in the
organometallic area is an unexplored category, even on a
conceptual basis, and it remains to be understood how the cobalt-
alkyne complex needs to be modified, at the core and at the
periphery, in order to achieve a synthetically meaningful level
of persistency. Second, using propargyl cations as precursors
to the respective radicals adds a new variable, when compared
to the purely organic environment,6,7 which can also affect the
product distribution.
(5) (a) Melikyan, G. G.; Nicholas, K. M. In Modern Acetylene Chemistry;
Stang, P. J., Diederich, F., Eds.; VCH Publishers: Weinheim, 1995; Chapter
4. (b) McGlinchey, M. J.; Girard, L.; Ruffolo, R. Coord. Chem ReV. 1995,
143, 331. (c) Amouri, H. E.; Gruselle, M. Chem. ReV. 1996, 96, 1077. (d)
Went, M. AdV. Organomet. Chem. 1997, 41, 69. (e) Green, J. R. Curr.
Org. Chem. 2001, 5, 809. (f) Muller, T. J. J. Eur. J. Org. Chem. 2001,
2021. (g) Teobald, B. J. Tetrahedron 2002, 58, 4133.
(6) (a) See: ref 1a, Chapter 4.1, p 758. (b) Griller, D.; Ingold, K. Acc.
Chem. Res. 1976, 9, 13. (c) Fischer, H. Chem. ReV. 2001, 101, 3581, and
references therein.
(7) (a) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925. (b) Allen, A. D.;
Fenwick, M. F.; Riyad, H.; Tidwell, T. T. J. Org. Chem. 2001, 66, 5759.
(c) Studer, A. Chem.-Eur. J. 2001, 7, 1159. (d) Alajarin, M.; Vidal, A.;
Ortin, M.; Bautista, D. New J. Chem. 2004, 28, 570. (e) Wetter, C.; Jantos,
K.; Woithe, K.; Studer, A. Org. Lett. 2003, 5, 2899.
(8) Titan; Wavefunction, Inc. All geometries were optimized at the PM3
and 3-21G* levels, 2000.

Cross-Coupling of Cobalt-Complexed Propargyl Radicals
Organometallics, Vol. 26, No. 13, 2007 3175
Figure 1. Charge distribution and structural parameters derived from the PM3 (3, 4, 13) and 3-21G* (10, 11, 17) calculations.
Scheme 1
converging propargyl radicals prior to the formation of d,l- and
meso-stereoisomers, respectively. For each diastereomer, there
are three staggered orientations (A-F) differing in the number
of supposedly destabilizing gauche interactions (two vs three).
For d,l-diastereomers, the most favorable orientation is repre-
sented by the structure A, with two gauche interactions, as
opposed to the alternative orientations B and C, each featuring
three gauche interactions (Figure 2). It is worthy to mention
that the X-ray crystallography for d,l-diastereomer 94f proves
that an orientation A is energetically most feasible with pairs
of cobalt-alkyne units and aromatic rings opposed to each other.
Among the three orientations that could give rise to meso-
diastereomerssD, E, and Fsthe former two are less favorable,
each featuring three gauche interactions (Figure 3). To interpret
the observed preponderant formation of d,l-stereoisomers 7-9,
one has to compare orientations A and F, each having only two
gauche interactions. The assumption needs to be made that the
repulsion between two cobalt-alkyne units and two aromatic

3176 Organometallics, Vol. 26, No. 13, 2007
Melikyan et al.
Figure 2. Formation of d,l-diastereomers 7-9: staggered orientations of converging propargyl radicals.
Figure 3. Formation of meso-diastereomers 7-9: staggered orientations of converging propargyl radicals.
Table 1. Chemo- and Diastereoselectivity of Radical Cross-Coupling Reactions
7
8
9
reduction method
d,l+meso
d,l:meso
d,l+meso
d,l:meso
d,l+meso
d,l:meso
1.
2.
3.
4.
Zn
Cp2Co
22
33
26
19
83:17
78:22
92:8
40
38
49
47
88:12
70:30
94:6
38
29
25
34
88:12
75:25
95:5
THF-mediated, one-step coupling protocol4h
Tf2O-mediated, one-step coupling protocol4g
88:12
94:6
96:4
rings in A is significantly less than that of two pairs of cobalt-
alkyne units and aromatic rings in F. To what extent an alleged
π-stacking between aromatic rings could provide for an ad-
ditional stabilization remains unclear.
1) is indicative of the contribution made to the assembly of the
converging radicals by an oxidized form of cobaltocene (radical-
ion pair).
The THF-mediated, one-step dimerization protocol4h (Table
1, entry 3) includes the treatment of equimolar amounts of
alcohols 1 and 2 with a 2-fold excess of THF and HBF4. The
reaction introduces another variable, cation generation rate, since
a one-step procedure does not involve isolation of the requisite
cations 3 and 4. A nearly statistical distribution was observed
for radical coupling products (7:8:9, 26:49:25), along with a
higher d,l-diastereoselectivity of 92-95% (de 84-90%). The
THF-mediated reaction has a complex multistep mechanism4h
that involves coordination of two molecules of THF with a
cationic center and the consecutive cluster-to-cluster reduction
between electronically unequivalent metal cores and cluster-
to-ligand electron transfer, with the latter actually generating
radicals 5 and 6. The reaction is known to provide for a better
selectivity in intermolecular dimerization reactions: a ratio of
d,l:meso diastereomers of 7 is equal to 75:254b and 95:5d in
Zn- and THF-mediated reactions, respectively. The spontaneous
radical generation reaction mediated by triflic anhydride4g
yielded 47% of cross-dimer 8; distribution of homo-dimers 7
and 9 significantly deviated from the statistical distribution with
the preponderant formation of the latter (7:9, 19:34; Table 1,
entry 4). The mechanism of the reaction includes an instanta-
neous formation of mixed esters when requisite alcohols are
treated with Tf2O, followed by a slow heterolysis of the R-C-O
bond and an in situ generation of the cations 3 and 4.4g The
level of observed stereoselectivity is comparable to that of the
Replacement of zinc, a heterogeneous reducing agent, with
cobaltocene, a 19e- homogeneous reducing agent, substantially
changes the reaction profile. The latter is a relatively bulky
molecule (199.56 Å3) comparable in volume with that of starting
alcohols (1 327.60 Å3; 2 362.18 Å3). The rate of the radical
generation will then be dependent upon the ability of cobaltocene
to approach the cationic center to effectively overlap a donor
orbital with a p-orbital of the cation, which, in turn, is
interacting, through-space, with an electron-rich cobalt cluster.4d
By the NMR data, the ratio of dimers 7:8:9 in the crude mixture
was equal to 33:38:29 (Table 1). While the concentration of
the cross-dimer 8 remains nearly the same (Zn 40%; Cp2Co
38%), the homo-dimers 7 and 9 are formed in almost equal
quantities (33%; 29%). It is indicative of the stronger reducing
power of cobaltocene, as opposed to that of zinc, which does
not allow for the discrimination of the requisite cations 3 and
4 based on the electrophilicity of the cationic center. The
stereoselectivity of the cobaltocene-mediated reaction (Table 1,
entry 2) is systematically lower (de 40-56%) revealing a
nontrivial dependence of the stereochemical outcome of the
radical coupling reaction on the nature of the reducing agent.
From a theoretical standpoint, one could expect the stereose-
lectivity of the coupling of free radicals to be independent of
the radical generation mode. Its observed dependency (Table

Cross-Coupling of Cobalt-Complexed Propargyl Radicals
Organometallics, Vol. 26, No. 13, 2007 3177
Scheme 2
Table 2. Chemo- and Diastereoselectivity of Radical Cross-Coupling Reactionsa,b
7
15
16
reduction
method
molar ratios
1:12
d,l+meso
d,l:meso
d,l+meso
d,l:meso
d,l+meso
d,l:meso
1. Zn
2. Zn
3. Zn
4. Zn
5. Zn
6. Zn
7. Cp2Co
8. THF
1:1
2:1
3:1
7:1
9:1
1:3
1:1
1:1
39(25)
52(45)
63(56)
83(76)
86(81)
17(6)
85:15
78:22
83:17
85:15
86:14
79:21
76:24
92:8
16(50)
24(44)
21(38)
8(22)
74:26
70:30
72:28
71:29
71:29
70:30
70:30
86:14
45(25)
24(11)
17(6)
9(2)
72:28
70:30
67:33
67:33
67:33
72:28
57:43
82:18
8(18)
6(1)
23(38)
47(50)
14(50)
60(56)
6(25)
33(25)
47(25)
53(25)
a Product distribution and diastereomeric ratios are determined by NMR. b Statistical distribution is shown in parentheses.
THF-mediated reaction (d,l:meso, 8 94:6; 9 96:4), with a slight
decrease for homo-dimer 7 (d,l:meso, 88:12 vs 92:8). The
observed disparity in product distribution and a level of
diastereocontrol allows us to conclude that the substrate structure
and radical generation algorithm are the most critical parameters
of the reaction, and the cross-coupling product is best formed,
and with a highest diastereoselectivity, in THF- and Tf2O-
mediated reactions, with Zn and Cp2Co being inferior by both
criteria chosen.
The impact of the γ-aryl group upon chemo- and diastereo-
selectivity was then studied by employing the alcohol 12, along
with parent alcohol 1, as a second component of the cross-
coupling reaction (Scheme 2). The rationale behind this was
that the presence of the bulky aromatic ring, in tandem with a
bent geometry of the triple bond,5 can make radical 14 more
persistent,6,7 thus retarding the rate of the homo-coupling
reaction and, to the contrary, facilitating the formation of the
cross-coupling product 15. Another outcome could have been
not steric, but an electronic impact of the γ-arylmethoxy group
that could affect the electrophilicity of the cation 13, thus
creating a precondition for a kinetic differentiation at the radical
generation step. The treatment of propargyl alcohols 1 and 12,
in a molar ratio 1:1, with an excess of HBF4 yielded respective
cations 3 and 13, which, upon isolation, were treated with zinc
to produce propargyl radicals 5 and 14. The subsequent radical
coupling afforded the mixture of homo- and cross-dimers 7:15:
16 in the ratio 39:16:45 (Table 2, entry 1). These data represent
a significant departure from a nearly statistical distribution of
the products observed in the absence of a substituent in the
γ-position of propargyl alcohol (Table 1; cross-dimer 8 38-
47%). In fact, a dramatic drop in the concentration of the cross-
product 15 points to a substantial difference in the generation
rate of prerequisite radicals 5 and 14. Most probably, because
of the electronic differences and charge distribution in cations
3 and 13, their reduction is well separated in time with one of
them (3) being reduced at a much higher rate. To interpret these
data, semiempirical and ab initio studies8 were carried out for
cobalt-complexed cations 3 and 13 and their organic counterparts
10 and 17 (Figure 1). For the latter, the electropositivity of the
C1 atom is much lower when compared to that in the parent
cation 10 (17 C1 +0.063326; 10 C1 +0.126233) (Figure 1), a
difference that should be attributed to the donating properties
of a 4′-OMeC6H4 substituent. The same trend is noticeable in
cobalt-complexed cations 3 and 13: for the former, the central
carbon atom (C1) is nearly electroneutral (C1 +0.004863), while
in the latter the cationic center acquires a partial negative charge

3178 Organometallics, Vol. 26, No. 13, 2007
Melikyan et al.
Table 3. Summary of the Crystal Structure Data for
Complex d,l-16
formula
fw
C47H26O15Co4
1066.40
temperature, K
cryst color
cryst dimens, mm
cryst syst
a, Å
100(2)
dark red
0.308 × 0.300 × 0.092
triclinic
11.0937(8)
13.5833(10)
17.9283(12)
69.733(6)
71.212(6)
66.574(7)
2271.5(3)
P1h
b, Å
c, Å
R, deg
â, deg
γ, deg
volume, Å3
space group
Z
2
D(calc), Mg/m3
µ(Mo KR), mm-1
no. of indep reflns
absorp corr
1.559
1.504
10 063 [Ri ) 0.0538]
none
Figure 4. Product distribution as a function of a molar ratio of
alcohols 1 and 12.
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R1
10 063/0/575
1.157
0.0495
wR2
0.1232
1.223, -0.576
largest diff peak, hole, e Å-3
To the contrary, the observed amounts of the cross-product 15
are systematically lower than those theoretically anticipated, with
the numerical data converging with an increase in the molar
ratio of reactants (Table 2: 1:12 1:1, 15 16% vs 50%; 9:1, 8%
vs 18%). Using a larger excess of alcohol 12 (Table 2, entries
3 and 6) resulted in a reversal in the concentration of homo-
dimers 7 and 16, while the amount of cross-product 15 remained
nearly the same (21% vs 23%). A rather dramatic change in
chemoselectivity was observed in the case of Cp2Co, acting as
a reducing agent (Table 2). The concentration of the cross-dimer
15 reaches the highest mark (47%), while the formation of the
homo-dimer 16 suffers a significant decline (6%). In the THF-
mediated reaction, the product distribution is strikingly different
(7:15:16, 53:14:33) from a nearly statistical distribution observed
in the case of propargyl alcohols with terminal triple bonds
(Table 1: 7:8:9, 26:49:25). One of the contributing factors in a
one-step, THF-mediated reaction is the rate of cation generation
since the latter is converted to the respective radicals in situ,
without their isolation in an individual form.
Figure 5. ORTEP diagram of the molecular structure of complex
d,l-16 with 30% probability ellipsoids.
(C1 -0.029867)! The calculation data also indicate that the
geometric and electronic parameters in close vicinity to the
cationic center remain nearly identical (Figure 1; C1′ 3 -0.086368,
13 -0.070943; C1-C1′ 3 1.464 Å, 13 1.463 Å; C1-C2 3 1.340
Å, 13 1.340 Å), while a donation from a 4′-OMeC6H4 substituent
increases the electron density over the metal core (3 Co
-0.404561; -0.421442; 13 -0.422122; -0.433775).
On the basis of the calculation data, the observed chemose-
lectivity (7:15:16, 39:16:45; Table 2, entry 1) can be accounted
for in terms of kinetic differentiation at the radical generation
step: cation 3 is reduced at a higher rate than its counterpart
13 because in the latter the central carbon atom (C1) is less
receptive toward an electron transfer from a reducing agent
(Figure 1: C1 3 +0.004863, 13 -0.029867). The low concen-
tration of the cross-dimer 15 (16%) is indicative of the disparity
in the generation of the respective radicals, 5 and 14. To
establish if the product distribution can be modified by using,
as a tool, the molar ratio of alcohols 1 and 12, the latter was
varied from 1:1 to 9:1 and the ratio of products and their
diastereoselectivity were determined by NMR (Table 2, entries
2-5; Figure 4). The percentage of homo-dimer 7 gradually
increased (from 39% to 86%), while the formation of its
counterpart, 16, dropped significantly (from 45% to 6%). The
formation of cross-product 15 was most favorable at the ratio
of 1:12 equal to 2:1 (24%; Table 2, entry 2), undergoing a
gradual decline with further increase in the molar ratio of the
starting alcohols. Also presented in Table 2, in parentheses, are
the statistical ratios of products 7, 15, and 16. Homo-dimers 7
and 16 are formed in higher quantities than statistically expected,
independent from the molar ratio of starting alcohols 1 and 12.
The diastereoselectivity data are summarized in Table 2.
Stereochemical assignments for dimers 7 and 15 are based on
the X-ray crystallography data4b,f and NMR signatures of
methyne protons.4c,f,g In the case of homo-dimer 16, the NMR
data were not sufficient for an unambiguous stereochemical
assignment and the structure of the major stereoisomer, as d,l-
16, was determined by X-ray crystallography (Figure 5).9 The
synthesis was carried out by the treatment of cation 13 with a
6-fold excess of zinc at ambient temperature and isolation of
the homo-dimer as a mixture of d,l:meso, 93:7 (Scheme 3). From
the conformational point of view, d,l-16 appears to substantially
differ from bis-cobalt complexes with terminal triple bonds.4b,e,f
The latter arrange both acetylenic moieties in close proximity
to each other, while phenyl4b,f or methyl4e groups are oriented
gauche, with a different degree of conformational distortion
(44-46°). Consequently, the internal hydrogen atoms (H3A and
H
4A) are disposed anti to each other (171-172°).4e,f In d,l-16,
given the presence of bulky γ-phenyl substituents, the alkyne
(9) (a) Oxford Diffraction. Crysalis CCD and RED, Version 1.70; Oxford
Diffraction Ltd.: Oxford, UK, 2002. (b) Sheldrick, G. M. SHELX-97,
Program for the Solution and Refinement of Crystal Structures, University
of Go¨ttingen: Go¨ttingen, Germany, 1997.

Cross-Coupling of Cobalt-Complexed Propargyl Radicals
Organometallics, Vol. 26, No. 13, 2007 3179
Scheme 3
Table 4. Selected Bond Lengths (Å) and Bond and Torsion Angles (deg) for Complex d,l-16
Bond Lengths
Co(1)-Co(2)
Co(1)-C(1)
Co(1)-C(2)
C(1)-C(2)
Co(2)-C(1)
Co(2)-C(2)
2.464(5)
1.995(3)
1.986(2)
1.344(4)
1.980(2)
1.983(2)
Co(3)-Co(4)
Co(3)-C(5)
Co(3)-C(6)
C(5)-C(6)
Co(4)-C(5)
Co(4)-C(6)
2.469(5)
1.969(3)
1.996(3)
1.341(4)
1.976(2)
1.970(2)
Bond Angles
C(11)-C(1)-C(2)
C(4)-C(5)-C(6)
147.3(2)
148.7(2)
C(1)-C(2)-C(3)
C(5)-C(6)-C(61)
150.1(2)
146.1(2)
Dihedral Angles
C(1)-C(2)-C(3)-C(31)
C(2)-C(3)-C(4)-C(5)
H(3)-C(3)-C(4)-H(4)
C(61)-C(6)-C(5)-C(4)
C(5)-C(4)-C(3)-C(31)
H(4)-C(4)-C(3)-C(2)
Co(1)-Co(2)-C(1)-C(2)
C(2)-C(1)-C(11)-C(12)
99.0
150.5
81.6
0.6
79.6
34.1
73.6
1.2
C(1)-C(2)-C(3)-C(4)
C(41)-C(4)-C(3)-C(31)
C(11)-C(1)-C(2)-C(3)
C(41)-C(4)-C(3)-H(3)
C(41)-C(4)-C(3)-C(2)
H(4)-C(4)-C(3)-C(31)
Co(3)-Co(4)-C(5)-C(6)
C(5)-C(6)-C(61)-C(62)
65.8
48.2
3.6
162.7
81.7
164.0°
74.4°
127.3°
units are pushed apart from each other, representing an anti
conformation around an internal C3-C4 bond (C2-C3-C4-C5
150.5°). Hydrogen atoms attached to the asymmetric centers
(H3, H4) are located gauche to each other (H3-C3-C4-H4
81.6°), as are the unsubstituted phenyl groups (C31-C3-C4-
C41 48.2°). Metal cores Co2C2 represent tetrahedrons with a
skew geometry where the angles between Co-Co and C-C
triple bonds are significantly deviated from the perpendicular
arrangement (73.6°; 74.4°).5a Spatially, the γ-phenyl groups are
not equivalent and are positioned differently with respect to the
triple bonds: one of the phenyl groups is located in the plane
of the triple bond [C(2)-C(1)-C(11)-C(12) 1.2°], while the
other exhibits a significant twist [C(5)-C(6)-C(61)-C(62)
127.3°]. Other noteworthy structural features of d,l-16 include
(a) an essentially undistorted planarity of alkyne moieties (C11-
C1-C2-C3 3.6°, C61-C6-C5-C4 0.6°); (b) a bent geometry5,10
for coordinated alkyne units (C11-C1-C2 147.3°, C1-C2-C3
150.1°, C4-C5-C6 148.7°, C5-C6-C61 146.1°), reflecting
substantial rehybridization of and strong back-bonding to the
alkyne from the cobalt carbonyl moiety; and (c) a lengthened
coordinated C-C triple bond (1.34 Å vs ca. 1.21 Å for free
ligand) attendant with complexation to the transition metal.
The level of diastereoselection for d,l- and meso-7 varies in
a wide range, from 76:24 to 92:8 (Table 2), with preponderant
formation of the d,l-stereoisomer for all reducing agents and
alternative experimental protocols studied so far. The lowest
stereoselectivity is observed in the case of Cp2Co (de 52%),
while the highest level of stereoselection was reported in the
THF-mediated process (de 84%). The stereoselectivity is
systematically lower for the homo-dimer 16, containing a
γ-phenyl substituent at the triple bond: d,l:meso ratio varies
from 57:43 to 82:18. Curiously, both extremes are again
observed in the case of Cp2Co (de 14%) and the THF-mediated
process (de 64%). The same phenomenon was observed for
cross-dimer 15: Cp2Co d,l:meso, 70:30; THF d,l:meso, 86:14.
Careful analysis of d,l:meso ratios (Table 2) revealed another
tendency that cannot be easily interpreted: for nearly every
reported case, the d,l-diastereoselectivity of cross-dimer 15 is
lower than that for homo-dimer 7 and higher than that for homo-
dimer 16.
d,l-3,4-Diaryl-1,5-alkadiynes are not easily accessible by
alternative means.11 The “classical” propargyl-propargyl cou-
pling reaction11a is accompanied by acetylene-allene rearrange-
ment and, attendant with it, a poor regioselectivity. Mixtures
of three isomeric compounds are usually formedshead-to-head,
head-to-tail, and tail-to-tailswith their separation being a
difficult experimental task.11a d,l-3,4-Diaryl-1,5-hexadiynes are
produced in a ruthenium-catalyzed process,11b although the
reaction is inherently limited in scope, and both yields and
diastereoselectivities drastically decrease with either electron-
withdrawing (CF3) or electron-donating substituents (Me; OMe)
introduced to the aromatic nuclei. The reproducibility of the
experimental data might also be problematic since the authors
claim the presence of “...adventitious molecular oxygen...” to
be essential to the mechanism of the reaction.11b Intermolecular
coupling of propargyl alcohols can also be mediated by a Ti-
(OiPr)2Cl2/Mg mixture,11c although the process, from the
synthetic point of view, remains highly deficient: (a) in the
case of propargyl alcohols with a terminal triple bond, target
1,5-alkadiynes are accompanied by comparable quantities of
acetylenic allenes (45-50%), which are difficult to separate;
(b) the reaction lacks diastereoselectivity and suffers from low
conversions (70-72%); and (c) the isolation of the products in
a homogeneous form is not achievable.11c Palladium-catalyzed
(11) (a) Badanyan, Sh. O.; Voskanyan, M. G.; Chobanyan, Zh. A. Russ.
Chem. ReV. 1981, 50, 1074. (b) Onodera G.; Nishibayashi, Y.; Uemura, S.
Organometallics 2006, 25, 35. (c) Yang, F.; Zhao, G.; Ding, Y.; Zhao, Z.;
Zheng, Y. Tetrahedron Lett. 2002, 43, 1289. (d) Ogoshi, S.; Nishiguchi,
S.; Tsutsumi, K.; Kurosawa, H. J. Org. Chem. 1995, 60, 4650.
(10) Dickson, R. S.; Fraser, P. J. AdVances in Organometallic Chemistry
12; Academic Press: New York, 1974; p 323.

3180 Organometallics, Vol. 26, No. 13, 2007
Melikyan et al.
8; fraction 2 d,l-8 and meso-9; fraction 3 d,l-9. Step 2, reseparation
of fraction 1 (PE; 2 plates): d,l-7 and meso-7 (combined) and meso-
8. Step 3, reseparation of fraction 2 (PE:C6H6, 7:1; 1 plate): d,l-8
and meso-9.
reductive homo-coupling reaction of propargyl carbonates yields
unsubstituted 1,5-alkadiynes as minor products (5-29%), along
with isomeric allene-ynes.11d
d,l- and meso-7: 12.5 mg was obtained (25.0%). By NMR, the
d,l:meso ratio was equal to 87:13, de 74%. Spectral and physico-
chemical characteristics are identical with those reported earlier.4h
Conclusion
Chemo- and diastereoselectivities of homo- and cross-
coupling reactions of Co2(CO)6-complexed propargyl radicals
are dependent upon the structural and electronic parameters of
the substrates. Among alternative radical generation methodss
reduction of respective cations with Zn or Cp2Co and one-step
mediation with THF or Tf2Osthe latter two provided for the
highest level of d,l-diastereocontrol (de up to 92%). A donating,
and bulky, γ-aromatic ring did not increase the persistency of
the propargyl radicals, but caused the kinetic differentiation at
the radical generation step and nonstatistical distribution of the
dimeric products. The chemoselectivity (the ratio of homo- and
cross-coupling products) is consistent with the computational
data: the electrophilicity of the R-cationic center is unexpectedly
enhanced by a p-methoxy substituent at the R-phenyl group and,
to the contrary, is decreased due to the presence of the
γ-aromatic ring.
meso-8: 3.9 mg was obtained (3.8%); dark red oil. TLC (PE:E,
2:1): Rf 0.60. 1H NMR (400 MHz, CDCl3): δ 3.89 (3H, s, OCH3),
4.38 (2H, AB-spectrum, CH, J ) 12.0), 4.96 (1H, s, HCt), 5.06
(1H, s, HCt), 6.90 - 7.55 (9H, m, aromatic H). MS-FAB: m/z
748 (M+ - 3CO), 663 (M+ - 6CO - H), 635 (M+ - 7CO - H),
607 (M+ - 8CO - H), 579 (M+ - 9CO - H), 551 (M+ - 10CO
- H), 524 (M+ - 11CO), 378 (M+ - 12CO - 2Co). d,l-8: 18.0
mg was obtained (17.5%). The ratio of d,l-8:meso-8 was equal to
82:18 (by weight), de 64%; dark red crystals; Tdec 75-115 °C
(sealed capillary; coevaporated with benzene, 3 × 1 mL). TLC (PE:
1
E, 2:1): Rf 0.54. H NMR (400 MHz, CDCl3): δ 3.69 (3H, s,
OCH3), 4.33 (2H, s, CH), 6.31 (2H, d, HCt, J ) 1.4), 6.69 (2H,
d, aromatic H, J ) 8.5), 7.03 (2H, d, aromatic H), 7.13 (5H, m,
aromatic H). MS-FAB: m/z 748 (M+ - 3CO), 720 (M+ - 4CO),
691 (M+ - 5CO - H), 664 (M+ - 6CO), 663 (M+ - 6CO - H),
636 (M+ - 7CO), 635 (M+ - 7CO - H), 608 (M+ - 8CO), 607
(M+ - 8CO - H), 580 (M+ - 9CO), 579 (M+ - 9CO - H), 552
(M+ - 10CO), 551 (M+ - 10CO - H), 523 (M+ - 11CO - H),
495 (M+ - 12CO - H), 437 (M+ - 12CO - Co - H), 378 (M+
- 12CO - 2Co), 319 (M+ - 12CO - 3Co - H). Anal. Calcd for
C31H16O13Co4: C, 44.75; H, 1.92. Found: C, 44.92; H, 2.13.
Experimental Section
All manipulations of air-sensitive materials were carried out in
flame-dried Schlenk-type glassware on a dual-manifold Schlenk
line interfaced to a vacuum line. Argon and nitrogen (Airgas,
ultrahigh purity) were dried by passing through a Drierite tube
(Hammond). All solvents were distilled before use under dry
nitrogen over appropriate drying agents (ether, THF, from sodium
benzophenone ketyl; CH2Cl2, from CaH2; benzene, from sodium).
All reagents were purchased from Sigma-Aldrich and used as
received. Co2(CO)8 was purchased from Strem. NMR solvents were
meso-9: 2.9 mg was obtained (5.4%). d,l-9: 21 mg was obtained
(39.0%). The ratio of d,l-9:meso-9 was equal to 88:12 (by weight),
de 76%. Spectral and physicochemical characteristics are identical
with those reported earlier.4f
Reduction with Cobaltocene. Under an atmosphere of nitrogen,
HBF4·Me2O (201 mg, 1.5 mmol) was added dropwise to a solution
of alcohol 1 (52 mg, 0.125 mmol) and alcohol 2 (56 mg, 0.125
mmol) in dry ether (20 mL) at -20 °C. The reaction mixture was
stirred for 1 h, and the ethereal layer was removed. At -20 °C,
cations 3 and 4 were washed with dry ether (2 × 15 mL), the
residual solvent was removed under reduced pressure, and the
precipitate was dissolved in dry CH2Cl2 (5 mL). Cobaltocene (47
mg, 0.25 mmol) was added at 0 °C, the reaction mixture was stirred
at 20 °C for 3 h, then an additional amount of cobaltocene (9 mg,
0.05 mmol) was introduced, and stirring was continued for an
additional hour (TLC control). By NMR of the crude mixture, the
product contribution was equal to 7:8:9, 33:38:29. The ratio of d,l-
and meso-diastereosomers of 7, 8, and 9 was equal to 78:22 (de
56%), 70:30 (de 40%), and 75:25 (de 50%), respectively.
One-Step, THF-Mediated Dimerization. Under an atmosphere
of nitrogen, at -5 °C, HBF4·Me2O (62 mg, 0.5 mmol) was added
dropwise to a solution of alcohol 1 (52 mg, 0.125 mmol), alcohol
2 (56 mg, 0.125 mmol), and THF (36 mg, 40.5 µL, 0.5 mmol) in
dry CH2Cl2 (5 mL). The reaction mixture was stirred at 20 °C for
8 h (TLC control). The reaction mixture was cooled to 0 °C, diluted
with ether (15 mL), and washed with water (4 × 10 mL). The
ethereal layer was dried (MgSO4), filtered through a short bed of
silica gel (2 cm), and evaporated to dryness under reduced pressure.
By NMR of the crude mixture, the product distribution was equal
to 7:8:9, 26:49:25. The ratio of d,l- and meso-diastereosomers of
7, 8, and 9 was equal to 92:8 (de 84%), 94:6 (de 88%), and 95:5
(de 90%), respectively.
supplied by Cambridge Isotope Laboratories. H and 13C NMR
1
spectra were recorded on a Bruker DRX-400 (1H, 400 MHz)
spectrometer. Chemical shifts were referenced to internal solvent
resonances and are reported relative to tetramethylsilane. Spin-
spin coupling constants (J) are given in hertz. Elemental analyses
were performed by Desert Analytics (Tucson, AZ). Melting
temperatures (uncorrected) were measured on a Mel-Temp II
(Laboratory Devices) apparatus. Silica gel S733-1 (200-425 mesh;
Fisher) was used for flash column chromatography. Analytical and
preparative TLC analyses were conducted on silica gel 60 F254 (EM
Science; aluminum sheets) and silica gel 60 PF254 (EM Science;
w/gypsum), respectively. Eluents are ether (E), petroleum ether
(PE), and benzene (B). Mass spectra were run at the Regional Center
on Mass-Spectroscopy, UC Riverside, Riverside, CA (FAB, ZAB-
SE; CI-NH3, 7070EHF; Micromass).
d,l- and meso-(3,4-Diphenyl-1,5-hexadiyne)bis(dicobalthexa-
carbonyl) (7), d,l- and meso-[3-(4′-Methoxyphenyl)-4-phenyl-
1,5-hexadiyne]bis(dicobalthexacarbonyl) (8), and d,l- and
meso-[3,4-Di(4′-methoxyphenyl)-1,5-hexadiyne]bis(dicobalth-
exacarbonyl) (9). Reduction with Zinc. Under an atmosphere of
nitrogen, HBF4·Me2O (201 mg, 1.50 mmol) was added dropwise
to a solution of alcohol 1 (52 mg, 0.125 mmol) and alcohol 2 (56
mg, 0.125 mmol) in dry ether (20 mL) at -20 °C. The reaction
mixture was stirred for 1 h, and the ethereal layer was removed.
At -20 °C, the cations 3 and 4 were washed with dry ether (2 ×
15 mL), the residual solvent was removed under reduced pressure,
and the precipitate was dissolved in dry CH2Cl2 (5 mL). The
reaction mixture was then treated with zinc (98 mg, 1.5 mmol)
and stirred at 20 °C for 3 h (TLC control). Zinc was filtered off,
and the crude mixture was examined by NMR (d,l-7:meso-7, 83:
17; d,l-8:meso-8, 88:12; d,l-9:meso-9, 88:12; 7:8:9, 22:40:38) and
fractionated on preparative TLC plates (silica gel, 20 × 20 cm).
Step 1 (PE:E, 20:1; 3 plates): fraction 1 d,l-7, meso-7, and meso-
One-Step, Tf2O-Mediated Dimerization. Under an atmosphere
of nitrogen, at -5 °C, Tf2O (141 mg, 84 µL, 0.5 mmol) was added
dropwise to a solution of alcohol 1 (52 mg, 0.125 mmol) and alcohol
2 (56 mg, 0.125 mmol) in dry CH2Cl2 (5 mL). The reaction mixture
was stirred at 20 °C for 24 h (TLC control). The reaction mixture
was cooled to 0 °C and diluted with ether (10 mL) and water (5
mL). The organic layer was repeatedly washed with water (5 × 4

Cross-Coupling of Cobalt-Complexed Propargyl Radicals
Organometallics, Vol. 26, No. 13, 2007 3181
mL). The ethereal layer was dried (Na2SO4), then filtered through
a short bed of silica gel (2 cm) and evaporated to dryness under
reduced pressure. By NMR of the crude mixture, the product
distribution was equal to 7:8:9, 19:47:34. The ratio of d,l- and meso-
diastereosomers of 7, 8, and 9 was equal to 88:12 (de 76%), 94:6
(de 88%), and 96:4 (de 92%), respectively.
(10 063 independent, R ) 0.0538) were collected and used for the
unit cell determination and the structure refinement.9 An initial
Patterson map located the four cobalt atoms, while all other non-
hydrogen atoms were located in the difference maps through
subsequent rounds of least-squares refinement. Hydrogen atoms
were placed in calculated positions with the exception of hydrogens
H3 and H4, which were located in the difference map. All non-
hydrogen atoms were refined anisotropically, calculated hydrogen
atoms were refined as riding atoms, while H3 and H4 were refined
isotropically. Structure solution and refinement were performed
using Shelx97 through the Wingx interface.9
[1-Phenyl-3-(4′-methoxyphenyl)-2-propyn-1-ol]dicobalthexac-
arbonyl (12). Under an atmosphere of nitrogen, a solution of BuLi
in hexane (4.8 mmol, 3 mL/1.6 M) was added dropwise to a solution
of 1-ethynyl-4-methoxybenzene (0.598 g, 4.4 mmol) in dry THF
(10 mL) at -20 °C (10 min). Upon addition, the reaction mixture
was stirred at -20 °C for 5 h, and a solution of benzaldehyde (0.53
mL, 5.2 mmol) was added dropwise to the reaction mixture at -40
°C (15 min). The reaction mixture was stirred 19 h at ambient
temperature, then cooled to 0 °C and quenched with H2O (30 mL)
and saturated NH4Cl (30 mL). An aqueous layer was extracted with
ether (3 × 15 mL), and combined ethereal fractions were dried
over Na2SO4. Upon concentration under reduced pressure (1/3 of
the initial volume), under an atmosphere of nitrogen, the crude
alcohol (1.04 g, 4.4 mmol; assuming 100% yield) was added to a
solution of dicobaltoctacarbonyl (1.65 g, 4.8 mmol) in THF (80
mL). The reaction mixture was stirred at room temperature for 15
h, concentrated under reduced pressure, and fractionated on a silica
gel column (153 g, PE:E, 5:1) to afford 12 (1.99 g, 86.5%) as black
crystals (crystallizes in a freezer in 2 weeks). Mp: 80-88 °C (sealed
capillary; coevaporated with benzene, 3 × 1 mL). TLC (benzene/
acetone, 9:1): Rf 0.60. 1H NMR (200 MHz, CDCl3): δ 2.49 (1H,
d, OH, J ) 2.8), 3.85 (3H, s, OMe), 6.15 (1H, d, CH), 6.90 (2H,
d, arom. H, J ) 8.8), 7.29-7.37 (3H, m, arom. H), 7.50 (4H, t,
arom. H, J ) 7.2). MS-FAB+: m/z M+ 524, 507 (M+ - OH),
496 (M+ - CO), 479 (M+ - OH- CO), 468 (M+ - 2CO), 440
(M+ - 3CO), 423 (M+ - OH - 3CO), 412 (M+ - 4CO), 384
(M+ - 5CO), 356 (M+ - 6CO), 221 (M+ - Co2(CO)6 - OH).
Anal. Found: C, 50.26; H 2.80. C23H16O9Co2 requires: C, 50.42;
H 2.69.
d,l- and meso-(3,4-Diphenyl-1,5-hexadiyne)bis(dicobalthexac-
arbonyl) (7), d,l- and meso-[1-(4′-Methoxyphenyl)-3,4-diphenyl-
1,5-hexadiyne]bis(dicobalthexacarbonyl) (15), and d,l- and
meso-[1,6-Di(4′-methoxyphenyl)-3,4-diphenyl-2,5-hexadiyne]bis-
(dicobalthexacarbonyl) (16). Reduction with Zinc. Substrates’
molar ratio 2:1 (a total of 0.25 mmol): Under an atmosphere of
nitrogen, HBF4·Me2O (201 mg, 1.50 mmol) was added dropwise
to
a
solution of propargyl alcohols
1
(70 mg,
0.167 mmol) and 12 (44 mg, 0.083 mmol) in dry ether (20 mL) at
-20 °C. The reaction mixture was stirred for 1.5 h, an ethereal
layer was removed, and the cations 3 and 13 were washed with
dry ether (2 × 15 mL). The residual ether was removed under
reduced pressure, and the precipitate was dissolved in dry CH2Cl2
(5 mL). The reaction mixture was then treated with zinc (98 mg,
1.50 mmol) and stirred at ambient temperature for 1 h (TLC
control). Zinc was filtered off, and the product composition was
determined by NMR to be equal to 7:15:16, 52:24:24. The
stereoisomeric ratio was determined as follows: d,l-7:meso-7, 78:
22 (de 56%); d,l-15:meso-15, 70:30 (de 40%); d,l-16:meso-16, 70:
30 (de 40%). Individual diastereomers were isolated by fractionation
on preparative TLC plates (silica gel, 20 × 20 cm; 2 plates; PE:E,
20:1): fraction 1 d,l-7 and meso-7; fraction 2 d,l-15, meso-15, and
MeOC6H4CtCCH2C6H5 [Co2(CO)6] (18); fraction 3 d,l-16 and
meso-16.
Fraction 1 (d,l-7 + meso-7): 42 mg was obtained (62.2%) as a
red solid. TLC (PE:E, 5:1): Rf 0.78. By NMR, the ratio of d,l-7:
meso-7 was equal to 80:20, de 60%. Spectral and physicochemical
characteristics are identical with those reported earlier.4h
d,l- and meso-[1,6-Di(4′-methoxyphenyl)-3,4-diphenyl-2,5-
hexadiyne]bis(dicobalthexacarbonyl) (16). Under an atmosphere
of nitrogen, HBF4·Me2O (458 mg, 3.42 mmol) was added dropwise
to a solution of propargyl alcohol 12 (300 mg, 0.57 mmol) in dry
ether (20 mL) at -5 °C. The reaction mixture was stirred for 2 h,
an ethereal layer was removed, and the cation 13 was washed with
dry ether (2 × 20 mL). The residual ether was removed under
reduced pressure, and the precipitate was dissolved in CH2Cl2 (11.4
mL). The reaction mixture was then treated with zinc (222 mg,
3.42 mmol) and stirred at ambient temperature for 16 h (TLC
control). Zinc was filtered off, and the crude mixture was fraction-
ated on the silica gel column (27 g; PE:E, 9:1) to afford 16 (157
mg, 54.2%; d,l:meso, 93:7, de 86%), as black crystals (crystallizes
in a freezer in 2 weeks). TLC (PE:E, 9:1): d,l Rf 0.36; meso Rf
0.42; Tdec 125-130 °C (partial melting; sealed capillary; coevapo-
Fraction 2 (d,l-15 + meso-15 + 18): by NMR data, the ratio of
(d,l-15 + meso-15):18 was equal to 95:5 and the stereoisomeric
ratio of d,l-15:meso-15 was equal to 70:30, de 40%. d,l-15 and
meso-15 were separated from trace amounts of 18 by using
preparative TLC (PE/CH2Cl2, 15:1; 3 runs). An authentic sample
of 18 was synthesized by quenching the cation 13 with tributyltin
hydride.
d,l-15 + meso-15: 4.7 mg was obtained (6.2%; d,l:meso, 70:
30; de 40%) as a dark red oil. TLC (PE:E; 5:1): d,l-15 Rf 0.69,
1
meso-15 Rf 0.69. H NMR (400 MHz, CDCl3; contains residual
amount of solvents): δ meso-15 3.83 (3H, s, OMe), d,l-15 3.86
(3H, s, OMe), meso-15 4.58 (H, d, CH, J ) 11.2), meso-15 4.69
(H, s, CtC-H), d,l-15 4.71 (H, d, CH, J ) 5.6), meso-15 4.77
(H, d, CH, J ) 11.2), d,l-15 4.86 (H, d, CH, J ) 5.6), 5.83 (H, s,
CtC-H), meso-15 6.68 (2H, d, arom., J ) 8.8), d,l-15 6.86 (2H,
d, arom., J ) 8), meso-15 6.84-7.02 (4H, m, 2H, arom.), d,l-15 +
meso-15 7.05-7.6 (20H, m, 10H, arom.). MS-FAB+: m/z 824 (M+
- 3CO), 796 (M+ - 4CO), 768 (M+ - 5CO), 740 (M+ - 6CO),
712 (M+ - 7CO), 684 (M+ - 8CO), 656 (M+ - 9CO), 628 (M+
1
rated with benzene, 3 × 1 mL). H NMR (400 MHz, CDCl3): δ
3.83 (6H, s, 2OMe), meso-16 4.82 (2H, s, 2CH), d,l-16 5.18 (2H,
s, 2CH), meso-16 6.63 (4H, d, arom. H, J ) 8.8), d,l-16 6.84 (4H,
d, arom. H, J ) 8.8), 6.96 - 7.05 (8H, m, arom. H), 7.10 (4H, t,
arom. H, J ) 7.6), 7.25 (2H, t, arom. H, J ) 7.2). MS-FAB+: m/z
930 (MH+ - 3CO), 846 (MH+ - 6CO), 818 (MH+ - 4CO), 762
(MH+ - 9CO), 678 (MH+ - 12CO), 647 (MH+ - 12CO - OCH3),
619 (MH+ - 12CO - Co), 588 (MH+ - 12CO - OCH3 - Co),
560 (MH+ - 12CO - 2 Co - CC6H4OMe), 501 (MH+ - 12CO
- 3Co), 221 (MH+ - 3CO - HCC6H5CCC6H4OMe). Anal.
Found: C, 51.84; H 2.85. C44H26O14Co4 requires: C, 52.12; H 2.58.
Single crystals suitable for X-ray structure analysis (Figure 2) were
obtained by methanol vapor diffusion into a solution of d,l-16 in
acetone.
- 10CO), 600 (M+ - 11CO), 572 (M+ - 12CO), 482 (M+
12CO - C7H6 or C6H2O), 454 (M+ - 12CO - 2Co), 395 (M+
-
-
12CO - 3Co), 221 (M+ - 12CO - 4Co - C16H13O). HR-MS/
FAB: calcd for C34H20O10Co4 M+ - 3CO 823.838438, found
823.841500.
MeOC6H4CtCCH2C6H5[Co2(CO)6] (18): dark red solid. TLC
(PE:E, 5:1): Rf 0.69. 1H NMR (400 MHz, CDCl3): δ 3.86 (3H, s,
OMe), 6.15 (1H, d, CH), 4.29 (2H, s, CH2) 6.92 (2H, d, arom. H,
J ) 8.8), 7.29-7.37 (3H, m, arom. H), 7.50 (4H, t, arom. H, J )
X-ray Crystallography of d,l-16. The crystal was mounted on
a glass fiber and placed directly into the cold stream of an Oxford
Diffraction Xcaliber3 diffractometer. A total of 61 450 reflections

3182 Organometallics, Vol. 26, No. 13, 2007
Melikyan et al.
7.2). MS-FAB+: m/z M+ 508, 480 (M+ - CO), 452 (M+ - 2CO),
424 (M+ - 3CO), 396 (M+ - 4CO), 368 (M+ - 5CO), 340 (M+
- 6CO), 281 (M+ - 6CO - Co), 221 (M+ - Co2(CO)6 - H).
HR-MS/FAB: calcd for C21H14O6Co2 (M+ - CO) 479.945434,
found 479.945900.
Fraction 3 (d,l-16 + meso-16): 20 mg was obtained (46.6%).
By NMR, the stereoisomeric ratio of d,l-16:meso-16 was equal to
70:30, de 40%.
ether (15 mL) was added. The organic layer was washed with H2O
(5 × 10 mL) and dried (Na2SO4). The crude mixture was evaporated
to dryness, then diluted with ether (20 mL) and added dropwise to
a solution of Co2(CO)8 (17 mg, 0.05 mmol) in ether (30 mL) at
room temperature. The reaction was stirred for 2 h at room
temperature. The crude was separated from excess Co2(CO)8 by
column chromatography using silica gel as a stationary phase (20
g; PE, E). By NMR of the crude mixture, the product distribution
was equal to 7:15:16, 53:14:33. The stereoisomeric ratio was
determined as follows: d,l-7:meso-7, 92:8 (de 84%); d,l-15:meso-
15, 86:14 (de 72%); d,l-16:meso-16, 82:18 (de 64%).
Substrates’ molar ratio 1:1 (1 0.125 mmol, 12 0.125 mmol; a
total of 0.25 mmol): 7:15:16, 39:16:45 (NMR); d,l-7:meso-7, 85:
15 (de 70%); d,l-15:meso-15, 74:26 (de 48%); d,l-16:meso-16, 72:
28 (de 44%). Substrates’ molar ratio 3:1 (1 0.1875 mmol, 12 0.0625
mmol; a total of 0.25 mmol): 7:15:16, 63:21:17 (NMR); d,l-7:
meso-7, 83:17 (de 66%); d,l-15:meso-15, 72:28 (de 44%); d,l-16:
meso-16, 67:33 (de 34%). Substrates’ molar ratio 7:1 (1 0.2188
mmol, 12 0.0312 mmol; a total of 0.25 mmol): 7:15:16, 83:8:9
(NMR); d,l-7:meso-7, 85:15 (de 70%); d,l-15:meso-15, 71:29 (de
42%); d,l-16:meso-16, 67:33 (de 34%). Substrates’ molar ratio 9:1
(1 0.225 mmol, 12 0.025 mmol; a total of 0.25 mmol): 7:15:16,
86:8:6 (NMR); d,l-7:meso-7, 86:14 (de 72%); d,l-15:meso-15, 71:
29 (de 42%); d,l-16:meso-16, 67:33 (de 34%). Substrates’ molar
ratio 1:3 (1 0.0625 mmol, 12 0.1875 mmol; a total of 0.25 mmol):
7:15:16, 17:23:60 (NMR); d,l-7:meso-7, 79:21 (de 58%); d,l-15:
meso-15, 70:30 (de 40%); d,l-16:meso-16, 72:28 (de 44%).
Hydrogen atom abstraction product 18 was formed in minute
quantities (3-7%).
Reduction with Cobaltocene. Under an atmosphere of nitrogen,
HBF4·Me2O (201 mg, 1.50 mmol) was added dropwise to a solution
of propargyl alcohols 1 (52 mg, 0.125 mmol) and 12 (66 mg, 0.125
mmol) in dry ether (20 mL) at -20 °C. The reaction mixture was
stirred for 1 h, an ethereal layer was removed, and the cations 3
and 13 were washed with dry ether (2 × 15 mL). The residual
amount of ether was removed under reduced pressure, and the
precipitate was dissolved in CH2Cl2 (3 mL). A solution of Cp2Co
(70.9 mg, 0.375 mmol) was prepared by dissolving Cp2Co in CH2-
Cl2 (2 mL) and stirring it at room temperature under anaerobic
conditions (20 min). This solution was then added dropwise to the
reaction flask at 0 °C. The reaction was stirred for 3 h (TLC
controlled). The crude mixture was then filtered through a short
bed of Florisil (2 cm) and eluted with ether. Ether was stripped by
reduced pressure, and the crude solid was analyzed using a
spectroscopic method. By NMR, the product distribution was equal
to 7:15:16, 47:47:6. The stereoisomeric ratio was determined as
follows: d,l-7:meso-7, 76:24 (de 52%); d,l-15:meso-15, 70:30 (de
40%); d,l-16:meso-16, 57:43 (de 14%).
d,l-7 and meso-7: 20 mg was obtained (40.3%). By NMR, the
ratio of d,l-7:meso-7 is equal to 94:6 (de 88%).
d,l-15, meso-15 and -18: 7.9 mg was obtained. By NMR data,
the ratio of d,l-15 + meso-15:18 was equal to 80:20. Individual
compounds were isolated by using preparative TLC (PE/CH2Cl2,
15:1; 3 runs). Obtained were d,l-15 + meso-15 (3 mg, 2.6%) and
18 (1.5 mg, 2.4%). The stereoisomeric ratio of d,l-15:meso-15 is
equal to 91:9 (de 82%).
Decomplexation of d,l- and meso-16 to d,l- and meso-1,6-Di-
(4′-methoxyphenyl)-3,4-diphenyl-2,5-hexadiyne (19). d,l-16 and
meso-16 were decomplexed in order to separate the individual
compounds from a side product of identical chromatographic
mobility. Under an atmosphere of nitrogen, Ce(NH4)2(NO3)6 (89
mg, 0.162 mmol) in acetone (5 mL) was added dropwise to a
solution of impure d,l-16 and meso-16 (21 mg, 0.0202 mmol) in
acetone (5 mL) at -78 °C. The reaction mixture was stirred for 1
h (TLC control), then poured, at 0 °C, onto a saturated aqueous
solution of NaCl (25 mL) and extracted with ether (3 × 15 mL).
The combined ethereal layers were dried (Na2SO4, +4 °C), the
solvent, upon filtration, was stripped away under reduced pressure,
and the residue was fractionated by PTLC (PE/B, 1:1). Obtained
was d,l-19 + meso-19 (4.9 mg, 55.0%) as a yellowish-white solid.
By NMR, the ratio of d,l-19:meso-19 was equal to 61:39 (de 22%).
The diastereomeric mixture can be crystallized, to release a white
solid, by dissolving it in a mixture of CH2Cl2 (5 mL) and pentane
(15 mL), followed by coevaporation of solvents under reduced
pressure. Mp: 129-130 °C (sealed capillary; coevaporated with
1
benzene, 3 × 1 mL). TLC (PE/E, 1:1): Rf 0.50. H NMR (400
MHz, CDCl3): δ meso-19 3.82 (6H, s, 2OMe), d,l-19 3.83 (6H, s,
2OMe), meso-19 4.28 (2H, s, 2CH), d,l-19 4.31 (2H, s, 2CH), 6.85
(4H, d, arom. H, J ) 8.8), 7.27-7.44 (14H, m, arom. H). MS
DEI: m/z 442 (M+), 221. HR-MS/DEI: calcd for C32H26O2
442.193280, found 442.193564.
Acknowledgment. This material is based upon work sup-
ported by the National Science Foundation under CHE-0707865.
We are also greatly indebted to the Office of Graduate Studies,
Research and International Programs and University Corpora-
tion, California State University Northridge, for generous
support.
d,l-7 and meso-7: 11.5 mg was obtained (23%). By NMR, the
ratio of d,l-7:meso-7 is equal to 71:29 (de 42%).
d,l-15, meso-15 and 18: 46.2 mg was obtained (includes minute
quantities of an unidentified side product). By NMR data, the ratio
of d,l-15 + meso-15:18 was equal to 48:52; d,l-15:meso-15, 70:30
(de 40%). Cross-dimer 15 could not be isolated in a homogeneous
form because of the presence of an unidentified component that
exhibited similar chromatographic behavior.
Supporting Information Available: Calculated heats of forma-
tion (kcal/mol) for propargyl cations 10, 11, and 17 (PM3) (Table
1; Figure 1); calculated total electronic energies (hartrees) for cobalt-
complexed propargyl cations 3, 4, and 13 (HF/3-21G*) (Table 1).
Tables of crystallographic details, bond distances and angles, atomic
coordinates, and equivalent isotropic displacement parameters, as
well as torsion angles for d,l-16, are available in CIF format. This
material is available free of charge via the Internet at http://pubs.
acs.org.
d,l- and meso-16: 4.3 mg was obtained (6.78%). By NMR, the
ratio of d,l-16:meso-16 is equal to 66:34, de 32%.
One-Step, THF-Mediated Dimerization. Under an atmosphere
of nitrogen, THF (36 mg, 0.50 mmol) was added to a solution of
propargyl alcohols 1 (52 mg, 0.125 mmol) and 12 (66 mg, 0.125
mmol) in dry CH2Cl2 (5 mL). The reaction mixture was then cooled
to -5 °C, and HBF4·Me2O (67 mg, 0.50 mmol) was added
dropwise. The mixture was then warmed to 20 °C and stirred for
21 h (TLC control). The reaction was then cooled to 0 °C, and
OM070180M