Transmetalation of aryl units from gold(I) to cobalt (III): A clean route to the synthesis of anion cobaltoreceptors

Article abstract of DOI:10.1021/om0490482

New organocobalt complexes were synthesized through the transmetalation reaction of lithium derivatives of several N,C,N pincer ligands, Li(2,6-(NR ²R¹CH₂)₂C₆H ₂R′), with [(ν⁵-Cp)CoI₂]₂. Changing one or both methyl groups on the nitrogen atoms for ethyl, tert-butyl, or a (CH₂)₄ group led to rather unstable aquo or HX derivatives of the form [(ν⁵-Cp)Co{2,6-(NR¹R ²CH₂)₂C₆H₃}HY]PF ₆ (where R¹, R² = Me, Et, t-Bu, (CH ₂)₄ and Y = OH, Cl, I), which were characterized by ¹H NMR spectroscopy and a crystal structure analysis for the compound in which R¹ = R² = Et. Related organocobalt derivatives could, however, be prepared more readily by using a transauration reaction. Thus, the transmetalation reaction between Ph₃PAu(2,6-(Me ₂NCH₂)₂C₆H₃) and (ν⁵-Cp)CoI₂]₂ afforded the known [(ν₅-Cp)Co{2,6-(Me₂NCH₂)₂C ₆H₃}H₂O]PF₆ in good yield. Two organocobalt complexes containing pincer ligands substituted at the para position by either an electron-withdrawing group, -COMe (5a), or an electron-donating group, -OMe (5b), were thus prepared by the transauration with gold(I) derivatives: Ph₃PAu(2,6-(Me₂NCH₂) ₂(4-COMeC₆H₂) and Ph₃PAu(2,6- (Me₂NCH₂)₂-(4-MeOC₆H₂) and [(ν⁵-Cp)CoI₂]₂. The thermodynamic study of these latter compounds indicated that they display high selectivity for acetate versus chloride anion in water. We found for 5b that the acetate versus chloride selectivity reaches a factor of 40. Such results are without precedent in the literature, if one does not take into account the selectivities found for oxyanions in nonprotic solvents such as acetonitrile, dimethyl sulfoxide, and chloroform.

Full text of DOI:10.1021/om0490482

1756
Organometallics 2005, 24, 1756-1761
Transmetalation of Aryl Units from Gold(I) to
Cobalt(III): A Clean Route to the Synthesis of Anion
Cobaltoreceptors
Mike Robitzer,† Imenne Bouama¨ıed,† Claude Sirlin,*,† Preston A. Chase,‡
Gerard van Koten,‡ and Michel Pfeffer*,†
Laboratoire de Synthe`ses Me´tallo-Induites UMR 7513, Universite´ Louis Pasteur,
4 rue Blaise Pascal, 67070 Strasbourg, France, and Department of Metal-Mediated Synthesis,
Debye Institute-Utrecht University, Padulaan 8, 3584 CH Utrecht, The Netherlands
Received December 3, 2004
New organocobalt complexes were synthesized through the transmetalation reaction of
lithium derivatives of several N,C,N pincer ligands, Li(2,6-(NR2R1CH2)2C6H2R′), with [(η5-
Cp)CoI2]2. Changing one or both methyl groups on the nitrogen atoms for ethyl, tert-butyl,
or a (CH2)4 group led to rather unstable aquo or HX derivatives of the form [(η5-Cp)Co{2,6-
(NR1R2CH2)2C6H3}HY]PF6 (where R1, R2 ) Me, Et, t-Bu, (CH2)4 and Y ) OH, Cl, I), which
were characterized by 1H NMR spectroscopy and a crystal structure analysis for the
compound in which R1 ) R2 ) Et. Related organocobalt derivatives could, however, be
prepared more readily by using a transauration reaction. Thus, the transmetalation reaction
between Ph3PAu(2,6-(Me2NCH2)2C6H3) and [(η5-Cp)CoI2]2 afforded the known [(η5-Cp)Co{2,6-
(Me2NCH2)2C6H3}H2O]PF6 in good yield. Two organocobalt complexes containing pincer
ligands substituted at the para position by either an electron-withdrawing group, -COMe
(5a), or an electron-donating group, -OMe (5b), were thus prepared by the transauration
with gold(I) derivatives: Ph3PAu(2,6-(Me2NCH2)2(4-COMeC6H2) and Ph3PAu(2,6-(Me2NCH2)2-
(4-MeOC6H2) and [(η5-Cp)CoI2]2. The thermodynamic study of these latter compounds
indicated that they display high selectivity for acetate versus chloride anion in water. We
found for 5b that the acetate versus chloride selectivity reaches a factor of 40. Such results
are without precedent in the literature, if one does not take into account the selectivities
found for oxyanions in nonprotic solvents such as acetonitrile, dimethyl sulfoxide, and
chloroform.
Scheme 1
Introduction
We have recently described the synthesis of the novel
organocobalt compound 1,1 which resulted from the cou-
pling between a cyclopentadienyl-cobalt(III) complex
and the monoanionic terdentate pincer ligand (NCN).2,3
A structural analysis performed by X-ray diffraction
studies showed, however, that in compound 1 the NCN
ligand was coordinated to the cobalt center in a biden-
tate fashion via the aryl carbon and one nitrogen atom,
the second N atom interacting with a metal-aquo unit
through a hydrogen bond. We have also observed that
Brønsted acids could be coordinated to the NCN-metal
fragment with no hydrolysis of the Co-C bond being
observed. The pKa determinations4 of the aquo deriva-
tive 1 demonstrated that the proton was acting as an
allosteric effector and that the complexes 2X, derived
from the reaction of 1 with Brønsted acids (HX), are
better described as anion supercomplexes (Scheme 1).5,6
In these latter compounds, the anions are coordinated
both to the cobalt center and to an ammonium unit
contained in the same receiving molecule. Modification
of the cobaltoreceptor was attempted to determine if
modulation of the binding properties was feasible by
changing, for instance, the nature of the nitrogen sub-
stituents or the electronic properties of the organocobalt
center. This was envisioned as a route to possibly tuning
the selectivity of the anion receptor, which is a property
of paramount interest in molecular receptors.7-13 In
addition, anchoring facilities introduced at the para
* To whom correspondence should be addressed. E-mail:
pfeffer@chimie.u-strasbg.fr (M.P.); sirlin@chimie.u-strasbg.fr (C.S.).
† Universite´ Louis Pasteur.
‡ Utrecht University.
(1) Meneghetti, M. R.; Grellier, M.; Pfeffer, M.; De Cian, A.; Fischer,
J. Eur. J. Inorg. Chem. 2000, 1539.
(2) van Koten, G. Pure Appl. Chem. 1989, 61, 1681.
(3) Albrecht, M.; van Koten, G. Angew. Chem., Int. Ed. 2001, 40,
3750.
(5) Bianchi, A.; Bowman-James, K.; Garc´ıa-Espan˜a, E. Supramo-
lecular Chemistry of Anions; Wiley-VCH: New York, 1997.
(6) Schmidtchen, F. P.; Berger, M. Chem. Rev. 1997, 97, 1609.
(4) Robitzer, M.; Sirlin, C.; Kyritsakas, N.; Pfeffer, M. Eur. J. Inorg.
Chem. 2002, 2312.
10.1021/om0490482 CCC: $30.25 © 2005 American Chemical Society
Publication on Web 03/01/2005

Transmetalation of Aryl Units from Au(I) to Co(III)
Organometallics, Vol. 24, No. 7, 2005 1757
Scheme 2
position of the aryl ring of the pincer would allow the
possibility of integrating the organocobalt receptor into
larger structures such as polymers14,15 and den-
drimers.16-18
Results and Discussion
Synthesis. The attempted syntheses of pincer com-
plexes of cobalt(III) related to 1 or 2X, in which the
substituents have been changed from methyl to ethyl,
tert-butyl, or (CH2)4, have led to rather disappointing
results, as the complexes obtained were unstable and
difficult to purify. In this respect, the only remotely
acceptable results were obtained in the synthesis of the
tetraethyl derivative 3.
Figure 1. ORTEP representation of 3.
in 3 than in 2Cl, reflecting the increased steric effects
of the ethyl groups as compared to those of the methyl.
In addition to the difficulty in synthesizing the
aforementioned compounds via Li-halogen exchange,
we also found that the transmetalation of lithiated
derivatives of para-substituted pincer ligands such as
4-R-2-bromo-1,3-bis[(dimethylamino)methyl] benzene (R
) Me, Cl, NMe2, COMe, O-SiMe2tBu) with (CpCoI2)2 did
not lead to clean syntheses of compounds related to 1.
In most of these cases not only was the yield of the
product rather low but also the compounds were difficult
to purify, as some paramagnetic byproducts were formed
that could not be easily removed. We were thus looking
for an alternate method that would avoid these incon-
veniences. Organogold(I) complexes were shown to be
good starting materials for transmetalation reactions
toward mercury, gold, and rhenium complexes.20a-c
Moreover, it was shown recently that the gold(I) complex
of the pincer ligand 4 was an interesting starting
material to transmetalate the latter ligand on Ni(II),
Pd(II), Pt(II), Ti(IV), or Fe(III).20d In addition, this
reagent obviated the reduction of the Ti(IV) and Fe(III)
complexes observed with the corresponding lithium
synthon. As such, this method could afford the desired
Co(III) compounds without the previously observed
difficulties. To test reaction conditions, we repeated the
reaction between 4 and (CpCoI2)2, which gave 1 in a
yield (60%) after workup similar to that of the original
synthesis (Scheme 3).
The transmetalation with 2-lithio-1,3-bis[(diethylami-
no)methyl]benzene19 and [CpCoI2]2 was followed by
treatment of the resulting compound with HCl to afford
1
3 as a green solid (Scheme 2). Its H NMR spectrum
was in line with that of the corresponding N-methyl
complexes. It is worth mentioning, however, that the
methylene protons of the diethylamino unit coordinated
to the cobalt center did not appear as the expected AB
spin system but as a singlet. On the other hand, the
protons of the methylene group linking the aryl ring and
the protonated diethylamino group were indeed dia-
stereotopic and coupled with the proton on the nitrogen
atom. Despite the fact that the compound was difficult
to purify (we could not get an acceptable analysis), we
succeeded in obtaining a few crystals, one of which
among them was amenable to an X-ray diffraction
study. An ORTEP diagram of the structure is depicted
in Figure 1.
When the geometry around the cobalt center in the
related complex 2Cl 1 is compared with that of 3 (Table
3), the main differences arose from the position of the
chloride ion, which was located at longer distances from
both the cobalt atom and the nitrogen N1 in 3 than was
the case in 2Cl. The angle C16-Co-Cl is indeed larger
This method proved to be synthetically simpler than
the corresponding method using lithiated pincer deriva-
tives, as the gold complex is air stable and can be
manipulated without need for inert conditions. More-
over, no redox side reactions were observed between the
cobalt complex and the gold(I). Also, the gold(I) byprod-
uct IAuPPh3 could be recovered almost quantitatively
and was subsequently reused.
(7) Lehn, J.-M. La chimie supramole´culaire; concepts et perspectives,
De Boeck & Larcier: Paris, Bruxelles, 1997.
(8) Graf, E.; Lehn, J.-M. J. Am. Chem. Soc. 1976, 98, 6403.
(9) Metz, B.; Rosalky, J. M.; Weiss, R. J. Chem. Soc., Chem.
Commun. 1976, 533.
(10) Lehn, J.-M.; Sonveaux, E.; Willard, A. K. J. Am. Chem. Soc.
1978, 100, 4914.
(11) Dietrich, B.; Guilhem, J.; Lehn, J.-M.; Pascard, C.; Sonveaux,
E. Helv. Chim. Acta 1984, 67, 91.
(12) Reilly, S. D.; Khalsa, G. R. K.; Ford, D. K.; Brainard, J. R.; Hay,
P. J.; Smith, P. H. Inorg. Chem. 1995, 34, 569.
(13) McNamara, C. A.; Dixon, M. J.; Bradley, M. Chem. Rev. 2002,
102, 3275.
(20) (a) Gregory, J.; Ingold, C. K. J. Chem. Soc. B 1969, 276.
Bennett, M. A.; Bhargava, S. K.; Griffiths, K. D.; Robertson, G. B.;
Wickramasinghe, W. A.; Willis, A. C. Angew. Chem., Int. Ed. Engl.
1987, 26, 258. (b) Vicente, J.; Chicote, M. T.; Abrisqueta, M. D.; Jones,
P. G. Organometallics 1997, 16, 5628. Vicente, J.; Chicote, M. T.;
Abrisqueta, M. D. J. Chem. Soc., Dalton Trans. 1995, 497. Vicente, J.;
Chicote, M. T.; Guerrero, R.; Saura-Llamas, I. M.; Jones, P. G.; Ramirez
de Arellano, M. C. Chem. Eur. J. 2001, 7, 638. Vicente, J.; Chicote, M.
T.; Abrisqueta, M. D.; Alvarez-Falcon, M. M. J. Organomet. Chem.
2002, 663, 40. (c) Ferrer, M.; Rodriguez, L.; Rossell, O.; Lima, J. C.;
Pina, F.; Gomez-Sal, P.; Martin, A. Organometallics 2004, 23, 5096-
5099. (d) Contel, M.; Stol, M.; Casado, M. A.; van Klink, G. P. M.; Ellis,
D. D.; Spek, A. L.; van Koten, G. Organometallics 2002, 21, 4556.
(14) Leadbeater, N. E.; Marco, M. Chem. Rev. 2002, 102, 3217.
(15) Dietrich, B.; Lehn, J.-M.; Guilhem, J.; Pascard, C. Tetrahedron
Lett. 1989, 30, 4125.
(16) van Heerbeek, R.; Kamer, P. J. C.; van Leeuwen, P. W. N. M.;
Reek, J. N. H. Chem. Rev. 2002, 102, 3717.
(17) Astruc, D.; Chardac, F. Chem. Rev. 2001, 101, 2991.
(18) (a) Kreiter, R.; Keij, A. W.; Klein Gebbink, R. J. M.; van Koten,
G. Top. Curr. Chem. 2001, 217, 163. (b) Chase, P. A.; Klein Gebbink,
R. J. M.; van Koten, G. J. Organomet. Chem. 2004, 689, 4016.
(19) Schlengermann, R.; Sieler, J.; Jelonek, S.; Hey-Hawkins, E.
Chem. Commun. 1997, 197.

1758 Organometallics, Vol. 24, No. 7, 2005
Robitzer et al.
Scheme 3
Scheme 5
Scheme 4
Table 1. Association and Selectivity Data for 5a
and 5b Obtained in Water
log Kass in water
selectivity
-
complex pKa Cl- AcO- H2PO4
S(AcO-/Cl-) S(H2PO4-/Cl-)
1
5a
5b
6.1 1.8
5.6 1.8
6.2 1.8
3.0
2.5
3.4
3.1
2.4
2.9
16
5
40
20
4
13
sized the corresponding Au-pincer complex (6) and
performed its transmetalation reaction with [CpCoI2]2.
Complex 6 was obtained in 71% yield (Scheme 5) after
purification via a route analogous to that for 4a, 4b,
and the related dimethyl pincer 4.20d Of note is that 6
is obtained as a highly viscous, clear, colorless oil at
room temperature, which solidifies to a white solid upon
cooling to 4 °C. Interrogation of the oil under polarized
light indicated no liquid crystalline properties. Reaction
1
of 6 with
/ equiv of [CpCoI2]2 in methanol under
2
conditions identical with those used for compounds 5a
and 5b resulted in the generation of a light yellow
solution. Workup and further reaction with HCl gave a
1
product in low yield (5%) that contained, by H NMR
spectroscopy, the pincer fragment. There was no indica-
tion of CpCo resonances. In line with previous observa-
tions during the synthesis of 3 via translithiation, the
product is unstable and difficult to isolate. The slightly
elevated temperatures needed to effect transauration
appear to result in decomposition of the product. Likely
due to steric effects, the NEt2 cannot effectively support
the Co center and, thus, the thermodynamic stability
of the tetraethyl pincer cobalt species is much lower
than that of the tetramethyl compound. To synthesize
3, lower temperatures are required; therefore, compara-
tively unreactive reagents, such as 6, are not compatible.
Thermodynamics. We have studied the thermody-
namic properties of the new compounds 5a and 5b. The
pKa of the pendant amine and the association constants
for selected anions (oxyanions and chloride) have been
determined as previously described.4 A general reaction
for both processes is found in Scheme 1. When the pH
of the medium was decreased by adding HPF6 to the
Using this procedure, two new cobaltoreceptors have
been prepared having either an electron-withdrawing
group (COMe, 5a) or an electron-releasing group (OMe,
5b), at the para position, giving the compounds in 50%
and 60% yield, respectively (Scheme 4). The synthesis
of 5b was straightforward, but the synthesis of 5a
required protection of the carbonyl function prior to
generation of the lithium reagent for the synthesis of
4a, which was obtained in 50% yield.
Cleavage of the C-Au bond was observed in attempts
to deprotect the acetal in 4a, but fortunately, in situ
deprotection after the transmetalation was successful
by addition of HPF6 to the reaction mixture (MeOH/
H2O) until pH 1. Extraction of the cobalt compound 5a
with chloroform afforded a red solution. Concentration
and dilution with water followed by addition of KPF6
led to the precipitation of the expected compound as red-
brown crystals in reasonable yield. No further purifica-
tion was required. It is likely, on the basis of the distinct
color of the solutions (see Experimental Section), that
the chloroform- and water-soluble compound was the
aqua complex having the iodide as a counteranion.
On the basis of the success of the transauration
method for the generation of para-substituted com-
plexes, attempts were made to determine if the synthe-
sis of 3 was amenable with this route. We thus synthe-
1
reaction medium, the H NMR signals of the complex
were shifted to high field. Therefore, variation of the
chemical shifts allowed us to measure the pKa of the
ionizable function. We have also measured the associa-
tion constants of the supercomplexes for the anions Cl-,
-
AcO-, and H2PO4 in water solutions buffered at pH
5.5. The results are gathered and compared to those of
the original cobaltoreceptor 1 in Table 1.
Figures 2 and 3 show the titration curves for 5a and
5b, respectively. The consequence of para substitution
of the aryl unit by a COMe or OMe group results in a

Transmetalation of Aryl Units from Au(I) to Co(III)
Organometallics, Vol. 24, No. 7, 2005 1759
Conclusion
Our goal to modify the original structure of an
organometallic anion receptor has been only partly
fulfilled. The modification of the alkyl substituents on
the nitrogens (NMe2 to NEt2) proved to be unfruitful,
as steric effects render the resulting tetraethyl molecule
less stable. This result leaves modification of the Cp
ring, for example, as the remaining potential site for
tuning receptor properties. The transauration reaction
was demonstrated to be the cleanest method for prepar-
ing series of related para-functionalized compounds.
Para substitution of the aryl ring of the pincer group
offers an excellent route to further functionalize such
structures. High selectivity of the p-methoxy-substituted
receptor toward acetate over chloride (S(AcO-/Cl-) )
40) has been highlighted. Further work is in progress
to investigate the use of chiral pincer cobalt receptors
to achieve the enantiomeric recognition of oxyanions.25,26
Figure 2. Evolution of ∆δ versus pH for compound 5a.
Experimental Section
All reactions affording the aryl bromides, the lithium salts,
and the cobalt complexes were carried out in freshly distilled
solvents, under a dry argon atmosphere using Schlenk tech-
niques. Commercial reagents have been used without further
purification. [η5-CpCoI2]2,27 AuClPPh3,28 and 1-Br-2,6(Et2-
29
NCH2)2C6H4 were prepared by published methods, All the
salts used for Kass measurements were dried in vacuo. Buffer
solutions were prepared from 2-(N-morpholino)ethanesulfonic
acid (MES, 2 equiv of the hemisodium salt and 3 equiv of the
acid, [C] ) 0.125 M). Elemental analyses were performed by
the Service Central d’Analyse de l’Institut de Chimie (Stras-
bourg, France). 1H NMR spectra were recorded at 300.13 MHz,
31P{1H} NMR spectra at 121.49 MHz, and 13C{1H} NMR
spectra at 75.47 MHz on a Bruker FT instrument (AV-300).
Chemical shifts (δ) were referenced to the residual solvent
peak as internal standard,30 except for pKa determinations,
for which a small amount of Et2O was added as internal
standard. δ and coupling constant values (J) are expressed in
ppm and Hz, respectively. pH was measured with a PHN78
pH meter (Tacussel electronique) with an ECS electrode.
Absorption spectra were recorded on a Uvikon XL spectrom-
eter. Association constants were determined via the graphical
representation of Scatchard. Reproducibility of multiple ex-
periments afforded the same values for a determined anion
within a 10% variation. All the selected values have R2 > 0.90,
and uncertainty was estimated between 5 and 10%.
Figure 3. Evolution of ∆δ versus pH for compound 5b.
slight modulation of the pKa of the amino unit. On
comparison of the original receptor 1 (pKa ) 6.1) to the
-COMe-substituted 5a (pKa ) 5.6), the pKa of the amino
fragment decreased by 0.5. The pKa of 5b (6.2) is almost
identical with that of 1. This decrease in the basicity of
the amino fragment for 5a reflects a better bonding of
the OH2 molecule to the receptor. The stronger associa-
tion of the water molecule to the receptor might be
directly correlated with the electron-withdrawing effect
of the para substituent, conferring more electrophilic
character to the cobalt center and thus strengthening
the HOH- - -N interaction. This results in a more
difficult displacement of the water molecule by the
incoming substrate. As the association process is mea-
sured at pH 5.5, half of the water molecules are still
bound to the amino unit. As a consequence the associa-
[Co{2,6-(Et2NCH2)2C6H3}(η5-C5H5)(HCl)]PF6 (3). A 1.8
mL portion of a 1.6 M solution of n-BuLi in n-hexane (2.88
mmol) was added to a solution of 1-bromo-2,6-bis[(diethylami-
no)methyl]benzene (0.9 g, 2.75 mmol) in 20 mL of Et2O cooled
to -78 °C. The solution was then heated to 30 °C and stirred
for 2 h. This solution was added dropwise to a stirred
suspension of [η5-CpCoI2]2 (0.98 g, 1.3 mmol) in toluene (100
mL) at -25 °C. The reaction mixture was heated to 20 °C and
tion constants for OAc- and H2PO4 are lowered by a
-
factor of 3 to 5.
On the basis of the pKa values, the behavior of the
-OMe-substituted receptor 5b is expected to be ap-
proximately the same as that of the original receptor 1.
In fact, the result is a stronger complexation of acetate
anion (2.5 times) and a weaker interaction with dihy-
drogenophosphate anion (1.5 times). The remarkable
consequence is that the acetate versus chloride selectiv-
ity reaches a factor of 40. Such results are without
precedent in the literature, if one does not take into
account the selectivities found for oxyanions in nonprotic
solvents such as acetonitrile, dimethyl sulfoxide, and
chloroform.21-24
(23) Rudkevich, D. M.; Verboom, W.; Brzozka, Z.; Palys, M. J.;
Stauthamer, S. P. R. V.; van Hummel, G. J.; Franken, S. M.; Harkema,
S.; Engbersen, J. F. J.; Reinhoudt, D. N. J. Am. Chem. Soc. 1994, 116,
4341.
(24) Fan, E.; van Arman, S. A.; Kincaid, S.; Hamilton, A. D. J. Am.
Chem. Soc. 1993, 115, 369.
(25) Seel, C.; Galan, A.; De Mendoza, J. Top. Curr. Chem. 1995,
175, 101.
(26) Schmidtchen, F. P.; Gleich, A.; Schummer, A. Pure Appl. Chem.
1989, 61, 1535.
(27) Roe, D. M.; Maitlis, P. M. J. Chem. Soc. A 1971, 3173.
(28) Braunstein, P.; Lehner, H.; Matt, D. Inorg. Synth. 1990, 27,
218.
(21) Azenbacher, P., Jr.; Jurs´ıkova´, K.; Sessler, J. L. J. Am. Chem.
Soc. 2000, 122, 9350.
(22) Beer, P. D.; Dent, S. W. Chem. Commun. 1998, 825.
(29) Trofimenko, S. Inorg. Chem. 1973, 12, 1215.
(30) Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. J. Org. Chem. 1997,
62, 7512.

1760 Organometallics, Vol. 24, No. 7, 2005
Robitzer et al.
stirred for 16 h, during which time it turned green. The solvent
was then removed in vacuo, and the product was extracted by
dissolving it in CH2Cl2 (3 × 50 mL). Filtration of this solution
over Celite and addition of n-hexane to it afforded a powder
that was dried in vacuo. The powder was then dissolved in
MeOH (10 mL), and a solution of HCl in water (36%; 0.1 mL,
1.2 mmol) was added to it, affording a purple solution. A large
excess of NaCl and KPF6 (10 equiv each) were added and the
mixture was filtered after 30 min of stirring. The residue was
washed with MeOH (2 × 15 mL). The product was redissolved
in acetone (3 × 30 mL), and the solvent was removed in vacuo
to afford 3 as a purple solid. Crystals were obtained from a
solution in acetone layered with n-hexane.
1H NMR (AcD6): δ 7.84 (s, 1H, Ar), 7.66 (s, 1H, Ar), 5.53 (s,
2
5H, C5H5), 4.71 and 3.81 (AB system, J ) 12.4, 2H, CH2N),
4.49 and 3.80 (AB system, 2J ) 14.0, 2H, CH2N-Co), 3.50 and
1.33 (2s, 6H, Co-NMe2), 2.59 (s, 3H, COCH3), 2.25 (s, 6H,
NMe2). 13C{1H} NMR (AcD6): δ 197.9 (CO), 172.0 (C-Co),
153.1 (Ar), 149.2 (Ar), 136.0 (C-CO), 129.1 (CH, Ar), 121.5
(CH, Ar), 86.3 (5C, Cp), 77.0 (CH2N-Co), 68.0 (CH2NMe2), 58.9
(NCH3-Co), 49.7 (NCH3-Co), 43.2 (Ls, NCH3), 26.6 (CH3). 31P-
1
{1H} NMR (AcD6): δ -143.0 (sept, JP-F ) 707.6, PF6). Anal.
Calcd for C19H28CoF6N2O2P (520.34): C, 43.86; H, 5.42; N, 5.38.
Found: C, 43.19; H, 5.21; N, 5.37. The lower value found for
1
C is due to the presence of ca. 7% of CH2Cl2 (checked by H
NMR), which could not be removed even after treatment of
5a for several hours in vacuo. MS (FAB+): 375.2 (24%) (5a -
PF6, calcd 375.15), 357.1 (100%) (5a - PF6 - H2O, calcd 357.1).
1H NMR (AcD6): δ 9.36 (br s, 1H, NH+), 7.45 (dd, 3J ) 5.1,
3
3
3J ) 3.9, 1H, Ar), 7.18 (d, J ) 5.1, 1H, Ar), 7.17 (d, J ) 3.9,
1H, Ar), 5.39 (s, 5H, Cp), 4.97 and 4.67 (AB system, 2J ) 12.9,
2H, CH2N-Co), 4.2 (s, 2H, CH2NH+), 3.74 (qd, 1H, 2J ) 13.6,
3J ) 7.1, Co-NCH2CH3), 3.54 (qd, 2H, 2J ) 7.3, 3J ) 4.9,
NH+CH2CH3), 3.00 (qd, 1H, 2J ) 13.6, 3J ) 7.1, Co-NCH2-
CH3), 2.98-2.88 (m, 2H, Co-NCH2CH3), 2.27 (qd, 1H, 2J )
14.6, 3J ) 7.3, NH+CH2CH3), 1.86 (t, 3H, 3J ) 7.1, Co-
NCH2CH3), 1.50 (t, 3H, 3J ) 7.3, NH+CH2CH3), 1.10 (t, 3H, 3J
) 7.3, Co-NCH2CH3), 0.84 (qd, 1H, 2J ) 14.6, 3J ) 7.3,
Au{2,6-(Me2NCH2)2-4-MeO-C6H2}(PPh3) (4b). A 5.6 mL
portion of a 2.3 M solution of n-BuLi in n-hexane (12.9 mmol)
was added to a solution of 4-bromo-3,5-bis[(dimethylamino)-
methyl]anisole (3.9 g, 12.9 mmol) in 25 mL of hexane that was
cooled to -80 °C. The solution was then heated to 25 °C and
stirred for 16 h. This solution was then added dropwise to a
slurry of AuClPPh3 (4.95 g, 9.82 mmol) in Et2O (70 mL) at
-30 °C. The medium was heated (35 °C) for 6 h, and the
solution became brownish. A 60 mL portion of water was then
added, and the mixture was stirred for 30 min. After filtration
over Celite, the aqueous phase was extracted with Et2O (2 ×
50 mL). The organic phases were recovered, dried over MgSO4,
and filtered on a G4 glass frit. The crude product was obtained
as a white powder through removal of the solvent in vacuo.
4b was obtained as white crystals via crystallization from CH2-
Cl2/hexane. Yield: 3.84 g (44%).
1H NMR (CDCl3): δ 7.70-7.61 (m, 6H, Hm PPh3), 7.50-7.44
(m, 9H, Ho,p PPh3), 6.91 (s, 2H, Ar), 3.81 (s, 3H, CH3), 3.80 (s,
4H, CH2), 2.26 (s, 12H, NMe2). 13C{1H} NMR (CDCl3): δ 158.2
(CO), 148.6 (Ar), 140.0 (C-Au), 134.4 (s, Co, PPh3), 131.2 (Cp,
PPh3), 129.1 (s, Cm, PPh3), 113.2 (CH, Ar), 70.2 (CH2N), 55.3
(OCH3), 46.0 (NCH3). 31P{1H} NMR (CDCl3): δ 45.85 (s, PPh3).
Anal. Calcd for C31H36AuN2OP (680.22): C, 54.71; H, 5.33; N,
4.12. Found: C, 54.70; H, 5.31; N, 4.14.
[Co{2,6-(Me2NCH2)2-4-MeO-C6H2}(η5-C5H5)(H2O)]PF6
(5b). Compound 4b (2.8 g, 4.1 mmol) and (CpCoI2)2 (1.56 g,
2.05 mmol) were mixed in 25 mL of methanol. The black-green
solution was heated at 35 °C for 6 h and became green-brown.
The crude product was filtered over Celite (recovering of gold)
and extracted with 2 × 30 mL of methanol. The red-brown
solution was concentrated to ca. 15 mL, and then 20 mL of
water was added and the solution was filtered over Celite
(further recovery of gold). The solution was concentrated again
on a rotary evaporator (T ≈ 40 °C), and the methanol removed.
An additional 50 mL of water and 10 equiv of KPF6 were added
with stirring. After 15 min, a precipitate could be isolated by
filtration over a G4 glass frit. It was washed with 2 × 50 mL
of water, 2 × 50 mL of hexane and 2 × 50 mL of Et2O. 5b was
isolated as a violet solid. Yield: 1.30 g (63%).
3
NH+CH2CH3), 0.68 (t, 3H, J ) 7.1, NH+CH2CH3).
Au{2,6-(Me2NCH2)2-4-MeC(OCH2CH2O)-C6H2}(PPh3) (4a).
A 4.35 mL portion of a 2.3 M solution of n-BuLi in n-hexane
(10 mmol) was added to a solution of 2-[4-bromo-3,5-bis-
[(dimethylamino)methyl]phenyl-2-methyl-1,3-dioxolane (3.51
g, 9.82 mmol) in 20 mL of hexane that was cooled to -80 °C.
The solution was then heated to 25 °C and stirred for 16 h.
This solution was then added dropwise to a slurry of AuClPPh3
(4.95 g, 9.82 mmol) in Et2O (70 mL) at -30 °C. The reaction
mixture was heated (35 °C) for 6 h and the solution became
brownish. A 50 mL portion of water was then added, and the
mixture was stirred for 30 min. After filtration over Celite,
the aqueous phase was extracted with 2 × 50 mL of Et2O. The
organic phases were recovered, dried over MgSO4, and filtered
on a G4 glass frit. The crude product was obtained as a white
powder through removal of the solvent in vacuo. 4a was
obtained as white crystals via crystallization from CH2Cl2/
hexane. Yield: 3.60 g (50%).
1H NMR (CDCl3): δ 7.67-7.61 (m, 6H, Hm PPh3), 7.50-7.42
(m, 9H, Ho,p PPh3), 7.37 (s, 2H, Ar), 4.00-3.92 (m, 2H, OCH2),
3.79 (s, 4H, NCH2), 3.79-3.71 (m, 2H, OCH2), 2.24 (s, 12H,
NMe2), 1.66 (s, 3H, CH3). 13C{1H} NMR (CDCl3): δ ) 175.5
(COO), 147.1 (C, Ar), 139.7 (C, Ar), 134.2 (d, 2JC-P ) 14.3, Co,
3
PPh3), 131.7 (C, Ar), 131.0 (Cp, PPh3), 128.9 (d, JC-P ) 11.2,
Cm, PPh3), 124.1 (CH, Ar), 109.3 (C, Ar), 70.0 (CH2N), 64.2
(2C, OCH2), 45.8 (NCH3), 27.6 (CCH3). 31P{1H} NMR (CDCl3):
δ 45.56 (s, PPh3). Anal. Calcd for C34H40AuN2O2P (736.25): C,
55.44; H, 5.47; N, 3.80. Found: C, 55.67; H, 5.51; N, 3.70.
[Co{2,6-(Me2NCH2)2-4-MeCO-C6H2}(η5-C5H5)(H2O)]-
PF6 (5a). Compound 4a (1.10 g, 1.50 mmol) and (CpCoI2)2 (0.57
g, 0.75 mmol) were mixed in 15 mL of methanol. The black-
green solution was heated to 35 °C for 6 h and became
heterogeneous (purple solution and brown precipitate). The
crude product was filtered over Celite (recovering of gold) and
extracted with 2 × 30 mL of methanol. The dark red solution
was concentrated to ca. 25 mL. A 50 mL portion of water was
added, and HPF6 (60% in H2O) was added dropwise until the
pH of the solution was equal to 1. The green solution was
stirred for 10 min; thereafter the product was extracted with
3 × 40 mL of CHCl3. The red organic phase was concentrated
to ca. 10 mL, and 150 mL of water was added (pH 7). CHCl3
was removed, and the product was totally soluble in water.
Filtration over Celite removed all residual solids (further
recovery of gold). KPF6 (10 equiv) was added to the aqueous
solution, and after 15 min of stirring, a precipitate could be
isolated by filtration over a G4 glass frit. It was washed with
2 × 50 mL of water, 2 × 50 mL of hexane, and 2 × 50 mL of
Et2O. 5a was isolated as a brown solid. Yield: 0.39 g (50%).
4
4
1H NMR (AcD6): δ 6.96 (d, J ) 2.3, 1H, Ar), 6.80 (d, J )
2
2.3, 1H, Ar), 5.44 (s, 5H, C5H5), 4.58 and 3.59 (AB system, J
) 12.2, 2H, CH2N), 4.32 and 3.60 (AB system, J ) 13.8, 2H,
2
CH2N-Co), 3.81 (s, 3H, OCH3), 3.46 and 1.32 (2s, 6H, Co-
NMe2), 2.19 (s, 6H, NMe2). 13C{1H} NMR (AcD6): δ 159.6 (Ar),
152.4 (Ar), 148.8 (Ar), 143.0 (C-Co), 117.2 (CH, Ar), 109.6 (CH,
Ar), 85.6 (Cp), 77.1 (CH2N-Co), 68.0 (CH2N), 58.8 (NCH3-
Co), 55.6 (OCH3), 49.7 (NCH3-Co), 43.3 (Ls, NCH3). 31P{1H}
NMR (AcD6): δ -143.0 (sept, 1JP-F ) 707.6, PF6). Anal. Calcd
for C18H28CoF6N2O2P (508.32): C, 42.53; H, 5.55; N, 5.51.
Found: C, 41.74; H, 5.55; N, 5.55. The lower value found for
1
C is due to the presence of ca. 8% of CH2Cl2 (checked by H
NMR), which could not be removed even after treatment of
5a for several hours in vacuo. MS (FAB+): 363.1 (35%) (5b -
PF6, calcd 363.15), 345.1 (100%) (5b - PF6 - H2O, calcd
345.14).

Transmetalation of Aryl Units from Au(I) to Co(III)
Organometallics, Vol. 24, No. 7, 2005 1761
Table 2. X-ray Diffraction Data for 3
reaction mixture was stirred for 16 h. The reaction was
quenched with water (90 mL), and the layers were separated.
The aqueous layer was washed with Et2O (3 × 50 mL), and
the combined organics were dried over MgSO4, filtered, and
dried in vacuo to obtain a sticky, clear, colorless oil. The oil
was dissolved in hot pentane and placed in the freezer (-30
°C). The cold solution was quickly decanted from the white
solid, which liquefies at room temperature (melting point ∼10
°C), and the oil was dried in vacuo. Yield: 0.510 g, 71%.
formula
mol wt
cryst syst
space group
a (Å)
C21H33CoN2Cl‚PF6
552.86
orthorhombic
Pbca
14.2005(3)
20.7594(5)
16.4489(6)
90
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
color
90
90
3
1H NMR (C6D6): δ 7.74 (d, 2H, J ) 7.8, Ar), 7.57 (m, 6H,
4849.0(4)
8
H
m PPh3), 7.35 (t, 1H, 3J ) 7.7, Ar), 6.92 (br m, 9H, Ho,p PPh3),
3
4.18 (s, 4H, ArCH2N), 2.61 (q, 8H, J ) 7.2, NCH2CH3), 0.92
violet
(t, 12H, J ) 7.1, NCH2CH3). 13C{1H} NMR (C6D6): δ 148.8
3
cryst dimens (mm)
0.20 × 0.20 × 0.05
(ArCH2N), 134.7 (d, 2JC-P ) 14, Co PPh3), 132.4 (d, 1JC-P ) 45,
F
F000
calcd (g cm-3
)
1.51
5
4
2288
Cipso PPh3, 131.0 (d, JC-P ) 0.2, Cp PPh3), 129.1 (d, JC-P
)
µ (mm-1
T (K)
λ (Å)
)
0.942
10, Cm PPh3), 127.5 (AuC), 127.4 (CpH, Ar), 126.0 (CmH, Ar),
64.6 (ArCH2N), 47.1 (NCH2CH3), 12.1 (NCH2CH3). 31P{1H}
NMR (C6D6): δ 45.1. Anal. Calcd for C34H42AuN2P (706.65):
C, 57.79; H, 5.99; N, 3.96. Found: C, 57.92; H, 5.96; N, 3.99.
X-ray Crystallographic Study. Single crystals suitable
for X-ray diffraction analysis were obtained as described above.
These crystals were poorly formed, which explains the low
quality of the crystal structure determination. Data were
collected on a Nonius KappaCCD diffractometer, using Mo KR
graphite-monochromated radiation (λ ) 0.7107 Å) and æ and
ω scans. Absorption corrections were partially included in the
data reduction procedure. The structures were solved using
direct methods and refined against |F|. Hydrogen atoms were
introduced in structure factor calculations as fixed contributors
by their computed coordinates (d(C-H) ) 0.95 Å, B(H) )
1.3Beqv (Å2) of the attached carbon atom). For all computations,
the Nonius OpenMoLeN package31 was used. Further experi-
mental details are given in Table 2.
173
0.710 73
Mo KR
KappaCCD
φ scans
0-18, 0- 21, 0-26
2.5-27.49
6153
radiation
diffractometer
scan mode
h, k, l limits
θ limits (deg)
no. of data measd
no. of data with I > 3σ(I)
3008
no. of variables
298
R
0.082
Rw
0.099
GOF
1.071
largest peak in final diff map (e Å-3
)
2.517
Table 3. Comparison of the Structural Data of
Supercomplexes 3 and 2Cl
bond length (Å)
bond angle (deg)
3
2Cl
3
2Cl
Co-Cp#a
1.718(2) 1.71(1) Cl-Co-N1
97.7(2) 96.8(1)
103.3(2) 95.9(1)
116.8(4) 120.9(2)
81.8(3) 82.8(3)
128.5(5) 125.5(8)
121.4(4) 128.7(6)
100.7(6)
b
Co-C(Cp)m 2.10(5) 2.089(4) Cl-Co-C16
Acknowledgment. We are grateful to Dr. Andre´ de
Cian and Nathalie Kyritsakas for their help in the X-ray
structure determination. The Ministe`re de l’Education
Nationale and the CNRS (France), the Natural Sciences
and Engineering Research Council of Canada (NSERC,
P.A.C.), and the University of Utrecht (P.A.C., G.v.K.)
are thanked for their financial support.
Co-Cl
2.429(2) 2.283(2) Cl-Co-Cp#a
2.094(8) 2.074(3) N1-Co-C16
1.968(9) 1.964(3) N1-Co-Cp#a
1.49(1) 1.487(5) C16-Co-Cp#a
1.50(1) 1.490(5) Co-N1-C16
1.41(1) 1.406(5) N1-C10-C11
3.201(5) 3.02(8) C10-C11-C16
C11-C16-Co
Co-N1
Co-C16
N1-C10
C10-C11
C11-C16
N2-Cl
107.6(8) 107.5(3)
114.7(8) 116.4(3)
111.0(2)
C6-N1-Co-Cl
C16-C15-C17-N2 -75.5(4)
27.0(8) 14.6(6)
Supporting Information Available: CIF files giving
crystallographic data. This material is available free of charge
via the Internet at http://pubs.acs.org.
a Cp# ) cyclopentadienyl centro¨ıd. b (Cp)m ) mean value.
Au{2,6-(Et2NCH2)2-C6H4}(PPh3) (6). Solid 2-lithio-1,3-bis-
[(diethylamino)methyl] benzene (0.259 g, 1.02 mmol) was
dissolved in dry Et2O (15 mL) and cannula-transferred onto a
slurry of ClAuPPh3 (0.504 g, 1.02 mmol) in dry Et2O at 0 °C.
The flask was rinsed with dry Et2O (1 × 10 mL), and the
OM0490482
(31) OpenMoleN, Interactive Structure Solution; Nonius BV, Delft,
The Netherlands, 1997.