Synthesis and decarbonylation reactions of diiron cyclopentadienyl complexes with bent-phosphinidene bridges

Article abstract of DOI:10.1021/om101125g

The phosphinidene complexes [Fe₂Cp₂(μ-PR)(μ-CO) (CO)₂] (Cp = η⁵-C₅H₅; R=Cy, Ph, Mes, Mes* Mes = 2,4,6-C₆H₂Me₃, Mes* = 2,4,6-C₆H₂^tBu₃) were readily prepared in high yields through a two-step procedure starting from the corresponding phosphine complexes [Fe₂Cp₂(μ-CO) ₂(CO)(PH₂R)]. When R = Cy, Ph, Mes, the first step required the oxidation of the phosphine complex with 1 equiv of [FeCp ₂]BF_4, this being followed by spontaneous dehydrogenation to yield the cationic phosphide complexes [Fe₂Cp₂(μ- PHR)(μ-CO)(CO)₂]BF₄, while the formation of the related PMes*H-bridged complex required a double oxidation of the corresponding neutral precursor in the presence of a deprotonating agent. In all cases the final step involved the deprotonation of the above cations with strong bases such as M(OH) (M = Na, K). Attempts to decarbonylate the above phosphinidene complexes by either thermolysis or photolysis in solution gave results strongly dependent on the nature of R. Thus, the photolysis of the cyclohexylphosphinidene complex gave a mixture of the clusters [Fe ₄Cp₄(μ₃-PCy)₂(μ₃- CO)(μ-CO)(CO)] (Fe-Fe distances 2.6115(7) and 2.5332(8) A) and [Fe ₃Cp₃(μ₃-PCy)(μ-PCyH)(μ-CO) ₂], while the thermal decarbonylation of the mesitylphosphinidene complex in refluxing toluene gave the trinuclear derivative [Fe ₃Cp₃(μ-PMes)(μ-PMesH)(μ-CO)₂(CO)], having a trigonal phosphinidene ligand bridging two iron atoms (Fe-P = 2.161(8), 2.165(8) A). The formation of all these compounds presumably involves the participation of the corresponding intermediate complex [Fe₂Cp ₂(μ-PR)(μ-CO)₂], a species likely to contain a trigonal (four-electron-donor) phosphinidene ligand, not detected. In contrast, the supermesitylphosphinidene complex undergoes a C(^tBu)-H oxidative addition to the P atom under mild thermal conditions to give the phosphine derivative [Fe₂Cp₂(μ-CO)₂(CO){PH(CH ₂CMe₂)C₆H₂^tBu ₂}], while its photolysis gave a mixture of the above phosphine complex and the phosphide-hydride derivatives cis- and trans-[Fe ₂Cp₂(μ-H){μ-P(CH₂CMe₂)C ₆H₂^tBu₂}(CO)₂], the latter resulting from the oxidative addition of a P-H bond to the dimetal center.

Full text of DOI:10.1021/om101125g

1102 Organometallics 2011, 30, 1102–1115
DOI: 10.1021/om101125g
Synthesis and Decarbonylation Reactions of Diiron Cyclopentadienyl
Complexes with Bent-Phosphinidene Bridges
ꢀ
M. Angeles Alvarez, M. Esther Garcıa, Rocıo Gonzalez, Alberto Ramos, and
Miguel A. Ruiz*
´
ꢀ
Departamento de Quımica Organica e Inorganica/IUQOEM, Universidad de Oviedo, E-33071 Oviedo, Spain
ꢀ
Received November 30, 2010
The phosphinidene complexes [Fe2Cp2(μ-PR)(μ-CO)(CO)2] (Cp = η5-C5H5; R=Cy, Ph, Mes,
Mes*; Mes = 2,4,6-C6H2Me3, Mes* = 2,4,6-C6H2 Bu3) were readily prepared in high yields through
t
a two-step procedure starting from the corresponding phosphine complexes [Fe2Cp2(μ-CO)2(CO)-
(PH2R)]. When R = Cy, Ph, Mes, the first step required the oxidation of the phosphine complex with
1 equiv of [FeCp2]BF4, this being followed by spontaneous dehydrogenation to yield the cationic
phosphide complexes [Fe2Cp2(μ-PHR)(μ-CO)(CO)2]BF4, while the formation of the related
PMes*H-bridged complex required a double oxidation of the corresponding neutral precursor in
the presence of a deprotonating agent. In all cases the final step involved the deprotonation of the
above cations with strong bases such as M(OH) (M = Na, K). Attempts to decarbonylate the above
phosphinidene complexes by either thermolysis or photolysis in solution gave results strongly
dependent on the nature of R. Thus, the photolysis of the cyclohexylphosphinidene complex gave
a mixture of the clusters [Fe4Cp4(μ3-PCy)2(μ3-CO)(μ-CO)(CO)] (Fe-Fe distances 2.6115(7) and
˚
2.5332(8) A) and [Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2], while the thermal decarbonylation of the
mesitylphosphinidene complex in refluxing toluene gave the trinuclear derivative [Fe3Cp3(μ-PMes)-
(μ-PMesH)(μ-CO)2(CO)], having a trigonal phosphinidene ligand bridging two iron atoms (Fe-P =
2.161(8), 2.165(8) A). The formation of all these compounds presumably involves the participation of
˚
the corresponding intermediate complex [Fe2Cp2(μ-PR)(μ-CO)2], a species likely to contain a
trigonal (four-electron-donor) phosphinidene ligand, not detected. In contrast, the supermesitylpho-
sphinidene complex undergoes a C(tBu)-H oxidative addition to the P atom under mild thermal
t
conditions to give the phosphine derivative [Fe2Cp2(μ-CO)2(CO){PH(CH2CMe2)C6H2 Bu2}], while
its photolysis gave a mixture of the above phosphine complex and the phosphide-hydride derivatives
t
cis- and trans-[Fe2Cp2(μ-H){μ-P(CH2CMe2)C6H2 Bu2}(CO)2], the latter resulting from the oxidative
addition of a P-H bond to the dimetal center.
Introduction
to which these ligands are bound.1d,g,h In a simplified way,
the M-P bonding in these complexes can be described as
single and double, respectively (A and B in Chart 1), with a
lone pair of electrons on the phosphorus atom in any case,
which makes these compounds very reactive toward unsatu-
rated organic molecules and useful in the synthesis of
organophosphorus derivatives.1
In contrast, the chemistry of binuclear complexes bearing
phosphinidene bridges (C-E in Chart 1) has remained
comparatively little explored until recently, although the
presence of multiple M-P bonding or lone pairs should
render these molecules quite reactive toward unsaturated
organic compounds or metal fragments. Within the chem-
istry of binuclear phosphinidene complexes, the number of
diiron species described in the literature is comparatively
scarce. Moreover, most of the reported compounds of this
type were either thermally unstable (as the cyclopentadienyl
complex [Fe2Cp2(μ-PPh)(CO)4])2 or were transient species
The study of the synthesis and reactivity of phosphinidene
(PR) complexes has constituted a very active area of research
within the organometallic chemistry in the last three
decades.1 Most of this work has been devoted to the terminal
bent-phosphinidene complexes, which are usually compared
to carbenes and can be also classified as electrophilic or
nucleophilic, depending on the nature of the metal fragments
*To whom correspondence should be addressed. E-mail: mara@
uniovi.es.
(1) Reviews: (a) Aktas, H.; Slootweg, J. C.; Lammerstma, K. Angew.
Chem., Int. Ed. 2010, 49, 2. (b) Waterman, R. Dalton Trans. 2009, 18. (c)
Mathey, F. Dalton Trans. 2007, 1861. (d) Lammertsma, K. Top. Curr.
Chem. 2003, 229, 95. (e) Streubel, R. Top. Curr. Chem. 2003, 223, 91. (f)
Mathey, F. Angew. Chem., Int. Ed. Engl. 2003, 42, 1578. (g) Lammerts-
ma, K.; Vlaar, M. J. M. Eur. J. Org. Chem. 2002, 1127. (h) Mathey, F.;
Tran-Huy, N. H.; Marinetti, A. Helv. Chim. Acta 2001, 84, 2938. (i)
Stephan, D. W. Angew. Chem., Int. Ed. Engl. 2000, 39, 314. (j) Shah, S.;
Protasiewicz, J. D. Coord. Chem. Rev. 2000, 210, 181. (k) Schrock, R. R.
Acc. Chem. Res. 1997, 30, 9. (l) Cowley, A. H. Acc. Chem. Res. 1997, 30,
445. (m) Cowley, A. H.; Barron, A. R. Acc. Chem. Res. 1988, 21, 81. (n)
Huttner, G.; Knoll, K. Angew. Chem., Int. Ed. Engl. 1987, 26, 743. (o)
Huttner, G.; Evertz, K. Acc. Chem. Res. 1986, 19, 406.
(2) (a) Klasen, C.; Effinger, G.; Schmid, S.; Lorenz, I. P. Z. Nat-
urforsch., B 1993, 48, 705. (b) Lorenz, I. P.; Pohl, W.; Noeth, H.;
Schmidt, M. J. Organomet. Chem. 1994, 475, 211.
r
pubs.acs.org/Organometallics
Published on Web 02/10/2011
2011 American Chemical Society

Article
Organometallics, Vol. 30, No. 5, 2011 1103
Chart 1
possibilityof generating diiron complexes of the type [Fe2Cp2-
(μ-PR)(μ-CO)2] with trigonal-phosphinidene bridges. As will
bediscussedbelow, such dicarbonylderivatives actuallymight
be formed but would in any case be very unstable, and only
different triiron or tetrairon derivatives of these putative
intermediates can be isolated from the corresponding reaction
mixtures, with only one of them displaying a trigonal phos-
phinidene bridge.
Results and Discussion
Synthesis and Structural Characterization of Compounds 1.
The phosphine complexes 1a-d were synthesized through a
two-step procedure. First, a photochemically induced dec-
arbonylation of [Fe2Cp2(CO)4] was carried out in MeCN at
263 K to give the tricarbonyl complex [Fe2Cp2(μ-CO)2-
(CNMe)(CO)],13 this having a coordinated acetonitrile mo-
lecule that is then easily displaced upon addition of the
stoichiometric amount of the appropriate primary phos-
phine PRH2 (R = Cy, Ph, Mes, Mes*) to give the corre-
sponding derivative [Fe2Cp2(μ-CO)2(CO)(PRH2)] (1a-d) in
high yield. The IR spectra in solution for these compounds
(Table 1) display two bands due to the C-O stretch of the
bridging CO ligands, with the expected pattern (weak and
strong, in order of decreasing frequencies) for a transoid
relative positioning of these ligands,14 and a third band in the
region of the terminal C-O stretches of these groups. The
frequency of the latter band suggests a transoid arrangement
of the terminal ligands as well, by comparison with analo-
gous diiron complexes described in the literature.15 As
expected, compounds 1a,b display a single resonance in the
corresponding 31P{1H} NMR spectra (δ 17.2 and 3.4 ppm,
respectively). In contrast, compounds 1c,d exhibited two
resonances with different intensities in each case (20:1 for
1c and 5:1 for 1d in CD2Cl2 at room temperature), which we
attribute to the presence in solution of a small amount of the
corresponding cis isomer, in addition to the major (trans)
isomer. These two isomers are in a solvent-dependent equi-
librium, as denoted by the distinct ratios found in C6D6 at
room temperature (25:1 for 1c and 13:1 for 1d). This type of
equilibrium has been studied in detail previously for analo-
gous complexes such as [Fe2Cp2(μ-CO)2(CO){P(OPh)3}]16
or even the tetracarbonyl dimer [Fe2Cp2(CO)4],17 and there-
fore does not need to be discussed in detail.
generated in situ from suitable precursors, such as [Fe2{μ-
P(NiPr2)}2(CO)6],3 [Fe2(μ-PtBu)(CO)6],4 [Fe3(μ2-PPh)(μ3-
PPh)(CO)9]n- (n=1, 2),5 and [Fe2(μ-PPh)2(CO)6]2-.6 All
these circumstances impose significant restrictions upon the
study of the chemical behavior of the aforementioned complexes.
Finally, there are just a couple of previous examples of isolable
diiron phosphinidene complexes, [Fe2{μ-P(OR)}2(CO)6]7 and
[Fe2(μ-PPh)2(L2)2],8 both of them having bulky protecting
i
groups (R = 4,2,6-C6H2MetBu2,7 L2=HC(CMeNC6H3 Pr2)2)8
and a trigonal geometry around phosphorus, but no reactivity
appears to have been developed around these complexes.
Recently, our group reported a high-yield synthetic proce-
dure for new stable diiron complexes with bridging bent (or
pyramidal)-phosphinidene ligands of the formula [Fe2Cp2(μ-
t
PR)(μ-CO)(CO)2] (R=Cy (3a),9 Ph (3b),9 2,4,6-C6H2 Bu3 or
Mes* (3d)10), using a two-step process. First, the phosphine
complexes [Fe2Cp2(μ-CO)2(CO)(PRH2)] are transformed into
the corresponding phosphide-bridged complexes [Fe2Cp2-
(μ-PHR)(μ-CO)(CO)2]BF4 via oxidation with [FeCp2]BF4
coupled to dehydrogenation (R=Cy, Ph) or deprotonation
(R=Mes*) processes. In the second step, deprotonation of the
cationic phosphide complexes with KOH or other strong bases
gave the corresponding phosphinidene derivatives. Preliminary
studies on the reactivity of the PCy-9,11 and PMes*-bridged10
complexes revealed that these highly nucleophilic complexes
have a strong potential for the synthesis of novel organophos-
phorus ligands. This was confirmed by a recent detailed study
on the reactions of these complexes toward diazoalkanes,12
which led to the uncovering of three novel coordination modes
for the phosphadiazadiene ligands.
In this paper we report full details, including complemen-
tary electrochemical experiments, of the synthetic procedure
leading to the phosphinidene complexes 3a,b,d, which we
have further extended to include the new mesitylphosphini-
dene complex [Fe2Cp2(μ-PMes)(μ-CO)(CO)2] (3c; Mes = 2,4,
6-C6H2Me3). We also report our studies on the decarbonyla-
tion reactions of compounds 3a-d, aimed to explore the
Synthesis and Structural Characterization of Compounds 2.
The removal of a hydrogen atom from the coordinated
PRH2 ligand of compounds 1 to give the corresponding
(13) Labinger, J. A.; Madhaven, S. J. Organomet. Chem. 1977, 134,
381.
(14) Braterman, P. S. Metal Carbonyl Spectra; Academic Press:
London, U.K., 1975.
(15) (a) Haines, R. J.; du Preez, A. L. Inorg. Chem. 1969, 8, 1459. (b)
Klasen, C.; Lorenz, I. P.; Schmid, S.; Beuter, G. J. Organomet. Chem.
(3) (a) King, R. B. J. Organomet. Chem. 1998, 557, 29. (b) King, R. B.;
Bitterwolf, T. E. Coord. Chem. Rev. 2000, 206-207, 563.
(4) Lang, H.; Zsolnai, L.; Huttner, G. Chem. Ber 1985, 118, 4426.
(5) (a) Ohst, H. H.; Kochi, J. K. Inorg. Chem. 1986, 25, 2066. (b) Ohst,
H. H.; Kochi, J. K. J. Am. Chem. Soc. 1986, 108, 2897.
ꢀ
(6) Treichel, P. M.; Douglas, W. M.; Dean, W. K. Inorg. Chem. 1972,
11, 1615.
(7) (a) Flynn, K. M.; Bartlett, R. A.; Olmstead, M. M.; Power, P. P.
Organometallics 1986, 5, 813. (b) Barlett, R. A.; Dias, H. V. R.; Flynn,
K. M.; Olmstead, M. M.; Power, P. P. J. Am. Chem. Soc. 1987, 109, 5699.
(8) Bai, G.;Wei, P.;Das, A. K.;Stephan, D. W.Dalton Trans. 2006, 1141.
1992, 428, 363. (c) Alvarez, C. M.; Galan, B.; Garcıa, M. E.; Riera, V.;
Ruiz, M. A.; Vaissermann, J. Organometallics 2003, 22, 5504. (d) Zhang,
S.; Brown, T. L. Organometallics 1992, 11, 4166. (e) Alvarez, C. M.;
Galan, B.; Garcıa, M. E.; Riera, V.; Ruiz, M. A.; Bois, C. Organome-
tallics 2003, 22, 3039. (f) Alvarez, C. M.; Garcıa, M. E.; Ruiz, M. A.;
Connelly, N. G. Organometallics 2004, 23, 4750.
ꢀ
ꢀ
(9) Alvarez, C. M.; Alvarez, M. A.; Garcıa, M. E.; Gonzalez, R.;
Ruiz, M. A.; Hamidov, H.; Jeffery, J. C. Organometallics 2005, 24, 5503.
(16) Cotton, F. A.; Kruczynski, L.; White, A. J. Inorg. Chem. 1974,
13, 1402.
ꢀ
(10) Alvarez, M. A.; Garcıa, M. E.; Gonzalez, R.; Ramos, A.; Ruiz,
(17) (a) Bryan, R. F.; Greene, P. T. J. Chem. Soc. A 1970, 3064. (b)
Bryan, R. F.; Greene, P. T.; Newlands, M. J.; Field, D. S. J. Chem. Soc.
A 1970, 3068. (c) Hunt, G. R. A. J. Chem. Educ. 1976, 53, 53. (d) Zhang,
S.; Brown, T. L. J. Am. Chem. Soc. 1993, 115, 1779. (e) Vitale, M.; Lee,
K. K.; Hemann, C. F.; Hille, R.; Gustafson, T. L.; Bursten, B. E. J. Am.
Chem. Soc. 1995, 117, 2286.
M. A. Organometallics 2010, 29, 1875.
(11) Alvarez, M. A.; Garcıa, M. E.; Gonzalez, R.; Ruiz, M. A.
ꢀ
Organometallics 2008, 27, 1037.
(12) Alvarez, M. A.; Garcıa, M. E.; Gonzalez, R.; Ruiz, M. A.
Organometallics 2010, 29, 5140.
ꢀ

1104 Organometallics, Vol. 30, No. 5, 2011
Table 1. Selected IR and 31P{1H} NMR Data for New Compounds
Alvarez et al.
b
c
compd
ν(CO)a
δ
P (JPP
)
δH (JHP)
[Fe2Cp2(μ-CO)2(CO)(PCyH2)] (1a)
[Fe2Cp2(μ-CO)2(CO)(PPhH2)] (1b)
1935 (s), 1770 (w), 1730 (vs)
1939 (s), 1767 (w), 1732 (vs)
1938 (m), 1772 (w), 1734 (vs)
17.2d
3.28 (327)d
4.47 (342)d
3.87 (345)
3.4d
[Fe2Cp2(μ-CO)2(CO)(PMesH2)] (trans-1c)
[Fe2Cp2(μ-CO)2(CO)(PMesH2)] (cis-1c)
[Fe2Cp2(μ-CO)2(CO)(PMes*H2)] (trans-1d)
[Fe2Cp2(μ-CO)2(CO)(PMes*H2)] (cis-1d)
[Fe2Cp2(μ-PCyH)(μ-CO)(CO)2]BF4 (2a)
[Fe2Cp2(μ-PPhH)(μ-CO)(CO)2]BF4 (cis, anti-2b)
[Fe2Cp2(μ-PPhH)(μ-CO)(CO)2]BF4 (cis, syn-2b)
[Fe2Cp2(μ-PMesH)(μ-CO)(CO)2]BF4 (2c)
[Fe2Cp2(μ-PMes*H)(μ-CO)(CO)2]BF4 (2d)
[Fe2Cp2(μ-PCy)(μ-CO)(CO)2] (3a)
[Fe2Cp2(μ-PPh)(μ-CO)(CO)2] (3b)
[Fe2Cp2(μ-PMes)(μ-CO)(CO)2] (3c)
[Fe2Cp2(μ-PMes*)(μ-CO)(CO)2] (3d)
[Fe4Cp4(μ3-PCy)2(μ3-CO)(μ-CO)(CO)] (4)
[Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2] (anti-5)
[Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2] (syn-5)
[Fe3Cp3(μ-PMes)(μ-PMesH)(μ-CO)2(CO)] (syn-6)
[Fe3Cp3(μ-PMes)(μ-PMesH)(μ-CO)2(CO)] (anti-6)
-24.7
-27.9
-21.3
-24.7
223.4d
189.6
184.5
156.6
181.7e
531.6g
498.2
495.1
593.4
352.2d
1936 (m), 1772 (w), 1732 (vs)
4.44 (343)
4.51 (343)
8.13 (390)d
9.10 (413)
2019 (vs), 1990 (w), 1833 (m)
2021 (vs), 1995 (w), 1834 (m)
2026 (vs), 1998 (w), 1830 (m)
2028 (vs), 2000 (w), 1825 (m)
1977 (vs), 1940 (w), 1773 (m)
1988 (vs), 1952 (w), 1783 (m)
1989 (vs), 1951 (w), 1782 (m)
1991 (vs), 1958 (w), 1769 (m)
1896 (vs), 1740 (m), 1611 (m)
1772 (vs), 1730 (w)h
9.00 (397)
9.88 (385)f
495.8 (305), -16.1 (305)
501.3 (286), -43.8 (286)
761.4 (55), 25.5 (55)
746.2 (51), 22.3 (51)
36.8g
0.49 (253)
2.33 (304)
6.44 (306)
1767(vs), 1725 (w)h
1953 (vs), 1763 (w), 1724 (s)
t
[Fe2Cp2(μ-CO)2(CO){PH(CH2CMe2)C6H2 Bu2}] (trans-7)
[Fe2Cp2(μ-CO)2(CO){PH(CH2CMe2)C6H2 Bu2}] (cis-7)
[Fe2Cp2(μ-H){μ-P(CH2CMe2)C6H2tBu2}(CO)2] (trans-8)
[Fe2Cp2(μ-H){μ-P(CH2CMe2)C6H2 Bu2}(CO)2] (cis-8)
1934 (m), 1772 (w), 1730 (vs)
4.56 (348)g
37.5g
170.5
167.6
t
t
1926 (w, sh), 1908 (vs)
1939 (vs), 1901 (w)
a Recorded in dichloromethane solution, with C-O stretching bands (ν(CO)) in cm-1 b Recorded in CD2Cl2 solutions at 290 K and 121.50 MHz
.
unless otherwise stated; δ in ppm relative to external 85% aqueous H3PO4 and J in Hz. c 1H NMR resonance for the hydrogen atoms bonded to
phosphorus, recorded in CD2Cl2 solutions at 290 K and 300 MHz unless otherwise stated; δ in ppm relative to internal SiMe4 and J in Hz. d In CDCl3
solution. e Recorded at 162.00 MHz. f Recorded at 400.13 MHz. g In C6D6 solution. h In toluene solution.
Scheme 1. Synthesis of the Phosphinidene Complexes 3
Chart 2. Projections of the Possible Isomers of Compounds 2,
Viewed along the Axis Joining the Bridging PRH and CO Ligands
decarbonylation or carbonylation reactions of phosphide-
bridged precursors.2,18
phosphide-bridged cations 2 was found to be strongly de-
pendent on the particular phosphine ligand (Scheme 1).
Thus, when R = Cy, Ph, Mes, the process was accomplished
by oxidation of the phosphine complexes with 1 equiv of
[FeCp2]BF4 in CH2Cl2 at 273 K, this being followed by a
spontaneous and fast dehydrogenation leading directly
to the corresponding complexes [Fe2Cp2(μ-PHR)(μ-CO)-
(CO)2]BF4 (2a-c). In contrast, when R = Mes*, the forma-
tion of the cationic phosphide complex required the addition
of 2 equiv of the oxidizing agent and the presence of a weak
base (NaHCO3). The best and more reproducible yields were
obtained when using THF as solvent and a temperature of
243 K, although the formation of 2d can be also accom-
plished in CH2Cl2. In any case, it should be noted that the
above reactions differ substantially from previous synthetic
procedures to prepare related diiron cations, these relying on
The IR spectra of compounds 2a-d display three C-O
stretching bands with a similar pattern (Table 1). The low
frequency of the band at ca. 1830 cm-1 is indicative of the
presence of a bridging carbonyl, while the relative intensities
of the bands at around 2000 cm-1 denote the presence of two
terminal carbonyls almost parallel to each other.14 Com-
pounds 2a,c,d display a single, quite deshielded resonance in
the corresponding 31P{1H} NMR spectrum, as expected for a
complex having a phosphide ligand bridging two metal atoms
connected by a metal-metal bond.19 In contrast, compound
2b exhibits two resonances, also quite deshielded: a very
intense one at 189.6 ppm and a weak one at 184.5 ppm, this
being indicativeofthe presence of twoisomersinsolution. The
major isomer is most likely to be the unique isomer present for
the other complexes: that is, that with the R group pointing
away from the Cp ligands (isomer cis,anti in Chart 2),
(19) Carty, A. J.; MacLaughlin, S. A.; Nucciarone, D. In Phosphorus-
31 NMR Spectroscopy in Stereochemical Analysis; Verkade, J. G., Quin,
L. D., Eds.; VCH: Deerfield Beach, FL, 1987; Chapter 16.
(18) Sugiura, J.; Kakizawa, T.; Hashimoto, H.; Tobita, H.; Ogino, H.
Organometallics 2005, 24, 1099.

Article
Organometallics, Vol. 30, No. 5, 2011 1105
Scheme 2. Reaction Pathway in the Formation of Compounds
2a-c
Scheme 3. Reaction Pathway in the Formation of Compound 2d
radical intermediate B would then undergo a fast dehydro-
genation process to give the diamagnetic intermediate C with
a terminal phosphide ligand, this finally rearranging into a
bridging position to give compounds 2, specifically with a cis
positioning of the terminal ligands. We note that a related
dehydrogenation process most likely takes place in the
reactions of the dinuclear radical species [Mo2Cp2(μ-PR2)-
(CO)4] with secondary phosphines HPR02 to give the mixed
phosphide derivatives [Mo2Cp2(μ-PR2)(μ-PR02)(CO)2].22
Clearly, the key step in the formation of compounds 2 is
the homolytic fission of a P-H bond in a radical species, as
opposed to more common processes involving the oxidative
addition of a P-H bond to multiple metal-metal bonds23 or
to an unsaturated mononuclear center,24 and thus represents
a novel synthetic entry to binuclear complexes having brid-
ging organophosphide ligands.
According to the experimental observations, the forma-
tion of compound 2d follows a pathway different from that
outlined above (Scheme 3). In the first place we must note
that, if only 1 equiv of the ferrocenium salt is added to a
solution of 1d in either CH2Cl2 or THF at 243 K, then an
NMR-silent purple solution is formed, probably containing
the paramagnetic species [Fe2Cp2(μ-κ1:η2-PMes*H2)(μ-
CO)(CO)2]BF4 (D) instead of the expected diamagnetic
phosphide complex. The intermediate D is a very air-sensi-
tive species, and it has a lifetime of only a few minutes at
temperatures above 243 K, decomposing into a mixture of
several uncharacterized products and significant amounts of
the tetracarbonyl dimer [Fe2Cp2(CO)4]. Although inter-
mediate D does not react with excess [FeCp2]BF4 itself, when
a solution of D is stirred with a base such as Na2CO3 or
NaHCO3, it then undergoes a clean disproportionation to
give an equimolar mixture of compounds 1d and 2d. At this
point, the addition of another 1 equiv of the oxidizing agent
under these conditions leads to the complete consumption of
expectedly more favored on steric grounds. Out of the two
other possible isomers for these compounds (Chart 2), we
proposethe cis,syn structure for theminor isomer of 2b, due to
the similar 31P chemical shifts displayed by these isomers. It
should benoticed that the R group in 2b (Ph) is that having the
smallest steric demands of the whole family of compounds 2,
thus reinforcing the hypothesis that the isomerism in these
species is a matter governed by steric effects. We finally note
that the above isomerism is reminiscent of that previously
found for carbene-bridged diiron and diruthenium complexes
of the type [M2Cp2(μ-CO)(CO)2(μ-CXY)] (X, Y = H, R,
SR).20
Proposed Reaction Pathways in the Formation of Com-
pounds 2. The formation of compounds 2a-c (Scheme 2) is
obviously initiated by the removal of one electron from the
corresponding phosphine complexes 1a-c, and this would
give initially the paramagnetic cation A. In order to account
for the dominant (cis) positioning of the Cp ligands in the
products 2, as opposed to the trans arrangement dominant in
the starting complexes 1, we propose that the intermediate A
would completely rearrange into the corresponding cis iso-
mer (B) possibly via an isomer A0 with only terminal ligands,
as proposed previously for the neutral complexes [Fe2Cp2(μ-
CO)2(CO){P(OPh)3}]16 and [Fe2Cp2(CO)4].17 That the cis
structure B would be more stable than the trans isomer A in
the paramagnetic cation would not be unusual. For instance,
we have shown previously that the oxidation of the phos-
phide-hydride complexes trans-[Fe2Cp2(μ-H)(μ-PPh2)-
(CO)2] and trans-[M2Cp2(μ-H)(μ-PPh2)(CO)4] (M = Mo,
W) yields the paramagnetic cations [M2Cp2(μ-H)(μ-PR2)-
(CO)x]þ (x = 2, 4) specifically with a cis geometry.15f,21 The
(20) (a) Dyke, A. F.; Knox, S. A. R.; Morris, M. J.; Naish, P. J. J.
Chem. Soc., Dalton Trans. 1983, 1417. (b) Colborn, R. E.; Dyke, A. F.;
Knox., S. A. R.; Mead, K. A.; Woodward, P. J. Chem. Soc., Dalton
Trans. 1983, 2099. (c) Gracey, B. P.; Knox, S. A. R.; Macpherson, K. A.;
Orpen, A. G.; Stobart, S. R. J. Chem. Soc., Dalton Trans. 1985, 1935. (d)
Schroeder, N. C.; Funchess, R.; Jacobson, R. A.; Angelici, R. J.
Organometallics 1989, 8, 521.
(22) Garcıa, M. E.; Riera, V.; Rueda, M. T.; Ruiz, M. A.; Lanfranchi,
M; Tiripicchio, A. J. Am. Chem. Soc. 1999, 121, 4060.
€
(23) (a) Ebsworth, E. A. V.; McIntosh, A. P.; Schroder, M. J.
Organomet. Chem. 1986, 312, C41. (b) Garcıa, M. E.; Riera, V.; Ruiz,
ꢀ
M. A.; Rueda, M. T.; Saez, D. Organometallics 2002, 21, 5515.
(24) Powell, J.; Gregg, M. G.; Sawyer, J. F. Inorg. Chem. 1989, 28,
4451.
ꢀ
(21) Alvarez, C. M.; Garcıa, M. E.; Rueda, M. T.; Ruiz, M. A.; Saez,
D.; Connelly, N. G. Organometallics 2005, 24, 650.

1106 Organometallics, Vol. 30, No. 5, 2011
Alvarez et al.
the phosphine complex 1d and almost quantitative forma-
tion of the phosphide derivative 2d. Of course, 2d is therefore
more conveniently prepared in a single step by adding 2 equiv
of [FeCp2]BF4 to a solution of 1d in the presence of excess
NaHCO3.
From the above data it is clear that intermediate D still
retains a coordinated PR*H2 ligand. In addition, an inspec-
tion of the IR spectra of different reaction mixtures in
CH2Cl2 suggests that at least two C-O stretching bands at
2010 (vs) and 1852 (m) cm-1 can be attributed to this
intermediate. Actually, these bands are not very different
from those of the final phosphide-bridged cation 2d. Hence,
we propose for D a similar structure, with a bridging
carbonyl and two terminal CO ligands arranged in a cis
fashion, with the other bridging position being occupied now
by a phosphine ligand bound to the second metal center
Figure 1. CV of compound 1a, recorded at 200 mV s-1 in
CH2Cl2, two cycles, from -2.0 to 0 V, starting at -0.6 V and
sweeping toward negative potentials.
through an agostic-type P-H
Fe interaction. As noted
3 3 3 3
above, intermediate D does not react to a detectable extent
with [FeCp2]BF4 unless a base is present. It seems, therefore,
that the oxidation of D to the corresponding diamagnetic
dipositive cation [Fe2Cp2(μ-κ1:η2-PMes*H2)(μ-CO)(CO)2]2þ
(E) would be an unfavorable event unless it is coupled to a
deprotonation step of the agostic phosphine, this leading to the
phosphide-bridged cation 2d, a matter to be confirmed through
the electrochemical studies described below. However, we
should stress that a similar deprotonation/protonation re-
lationship linking bridging agostic phosphine ligands to the
corresponding bridging phosphide ligands has been recently
reported by us for cations of the type [Mo2Cp2(μ-κ1:η2-
HPR2)(μ-PR02)(CO)2]þ.25
Cyclic Voltammetry Studies on Compounds 1a,d. In order
to gain further insight into the different reaction pathways
leading to compounds 2, cyclic voltammetry (CV) experi-
ments were carried out for compounds 1a (Figure 1) and 1d
(Figures 2 and 3) in CH2Cl2. We note first that the results
obtained for compound 1a are fully consistent with the
mechanism depicted in Scheme 2. Thus, we observe a nearly
reversible wave at ca. -0.2 V (vs the FeCp2þ/0 couple) when
scanning at 200 mV s-1. However, when the scan rate is
slowed to 50 mV s-1, then the reduction wave almost
disappears, which is consistent with an irreversible chemical
transformation (dehydrogenation) taking place at the para-
magnetic cation initially formed. Indeed, the second cycle
displays a new and reversible wave at -1.23 V, not present in
the first cycle (Figure 1), which is attributed to the formation
of 2a. This was confirmed by an independent CV study
carried out on the phosphide complex.
Figure 2. CV of compound 1d, recorded at 200 mV s-1 in
CH2Cl2, from 0 to -1.1 V, starting at -0.6 V and sweeping
toward positive potentials.
In contrast, the CV diagram of compound 1d shows a
wave at a potential (ca. -0.15 V) slightly higher than that of
1a, as expected, but now this wave remains reversible regard-
less of the scan rate (Figure 2), while successive scans showed
no traces of new waves attributable to the phosphide cation
2d (shown by independent CV experiments to give rise to a
reversible wave at -1.14 V). This is in full agreement with our
proposal that dehydrogenation does not take place after the
initial oxidation of the phosphine complex 1d. A second, now
irreversible, oxidation wave at ca. 0.73 V can be observed for
1d when scanning at higher potentials (Figure 3), this clearly
indicating that the ferrocenium salt itself would not be able
to further oxidize the paramagnetic cation initially formed
to give the dipositive cation E, in agreement with our
Figure 3. CV of compound 1d, recorded at 200 mV s-1 in
CH2Cl2, from 1.15 to -1.0 V, starting at -0.6 V and sweeping
toward positive potentials.
experimental data. Only when this process is coupled to a
proton transfer reaction due to the presence of a base, as
noted above, would the unfavorable thermodynamics to
reach compound 2d be overcome. A similar effect has been
previously found in the chemical and electrochemical studies
of the hydride derivatives [M2Cp2(μ-H)(μ-PPh2)(CO)4]
(M = Mo, W),21 where the second oxidation process to give
[M2Cp2(μ-H)(μ-PPh2)(CO)4]2þ occurred irreversibly at a
(25) Alvarez, M. A.; Garcıa, M. E.; Martınez, M. E.; Ramos, A.;
ꢀ
Ruiz, M. A.; Saez, D.; Vaissermann, J. Inorg. Chem. 2006, 45, 6965.

Article
Organometallics, Vol. 30, No. 5, 2011 1107
Figure 4. ORTEP diagrams of compound 3b, with H atoms omitted for clarity (left, taken from ref 9), and of compound 3d, with H
atoms and the p-tBu group omitted for clarity (right, taken from ref 10).
˚
potential higher than that of the ferrocenium cation
(0.1-0.25 V), but only in the presence of a weak base,
yielding the monopositive cations [M2Cp2(μ-PPh2)(CO)4]þ.
Unfortunately, in the case of 1d, the return sweep to negative
potentials after the second oxidation shows a greatly dimin-
ished reduction wave corresponding to the first oxidation
process, this being indicative of hampered diffusion of the
analyte to the surface of the working electrode, most prob-
ably due to adsorption of the decomposition products
formed after the second and irreversible oxidation of 1d,
thus precluding the extraction of any further information
from this experiment.
Synthesis and Structural Characterization of Compounds 3.
The final step in the formation of compounds 3 involves the
reversible deprotonation of the corresponding PHR-bridged
precursors 2, this being one of the common procedures to
generate pyramidal phosphinidene bridges.2b,26 In our case,
this can be conveniently accomplished using solid Na(OH)
or K(OH), a process easily reversed by addition of
Table 2. Selected Bond Lengths (A) and Angles (deg) for
Compounds 3b,da
3b9
3d10
Fe-Fe
2.5879(4)
1.756(2)
1.925(2)
2.262(1)
1.845(2)
97.7(1)
2.5763(5)
1.760(3)
1.913(3)
2.306(1)
1.893(2)
99.5(1)
94.2(1)
86.0(1)
113.4(1)
Fe-COt
Fe-COb
Fe-P
P-C(aryl)
Fe-Fe-COt
P-Fe-COt
COt-Fe-COb
Fe-P-C(aryl)
91.5(1)
86.8(1)
113.9(1)
a Values involving the iron atoms are averaged figures for Fe(1) and
Fe(2); terminal and bridging CO ligands are identified with t and b
superscripts, respectively.
phosphinidene ligand bearing a lone electron pair not involved
in the binding to the metal centers. There are, however, some
significant differences between both structures, most likely
induced by the steric bulk of the supermesityl moiety in 3d:
˚
(a) a slight elongation (by ca. 0.05 A) in both the P-C(aryl)
HBF4 OEt2 to the resulting complexes 3.
3
and the already mentioned Fe-P lengths in 3d, (b) a strong
deviation of the aryl ring of 3d from planarity, adopting an
The solid-state structures of compounds 3b,d (Figure 4 and
Table 2) were reported in our preliminary studies.9,10 In both
cases the overall geometries are very similar, with two FeCp-
(CO) fragments arranged in a cis fashion and symmetrically
bridged by a CO ligand and an arylphosphinidene ligand. The
aryl group is positioned anti to the Cp ligands, a conformation
presumably more favored on steric grounds and also dominant
in the cationic precursors 2, as discussed above. The inter-
t
incipient boat conformation that takes the ortho Bu groups
˚
away from the terminal Cp (H H = ca. 2.2 A) and CO
˚
(H C/O = ca. 2.6 A) ligands, and (c) a slight bending of the
3 3 3
3 3 3
terminal carbonyls in 3d away from the aryl group. We notice
that similar “boat deformations” of supermesityl groups have
been previously described in the literature for other
molecules,28 although that observed in 3d seems particularly
strong according to the tip angles for the ipso and para C atoms
of the ring (ca. 22 and 10°, respectively).
The spectroscopic data in solution for compounds 3a-d
are fully consistent with the solid-state structures just dis-
cussed. We first note that the IR spectra display three C-O
stretching bands indicative of the retention in solution of a
bridging CO and almost parallel terminal CO ligands, the
pattern being similar to that of the parent cationic phosphide
complexes 2, but with the bands shifted to lower frequencies
(by ca. 40 cm-1), as expected from the increased electron
density in the phosphinidene complexes. The PR ligand in
these compounds gives rise to a characteristic strongly
deshielded resonance in the 31P{1H} NMR spectra (δ
495-595 ppm, Table 1). For comparison, these shifts are
higher than that of the recently reported diruthenium
˚
metallic distances in these compounds (ca. 2.58 A) are con-
sistent with the presence of a metal-metal single bond and are
comparable to those measured for related cyclopentadienyl
compounds having P and C atoms at the bridging positions,
such as the phosphide-bridged complexes [Fe2Cp2{μ-P-
(CHdCHPh)Ph}(μ-CO)(CO)2]OTf18 and [Fe2Cp(μ-PMe2)-
(μ-CO)(CO)4]27 (ca. 2.62 A). The most remarkable feature in
˚
3b,d is of course the bridging bent-phosphinidene ligand, with
the phosphorus atom in a pyramidal environment as denoted
by the C(ipso)-P-Fe2(centroid) angles of ca. 120°. The Fe-P
˚
˚
distances, ca. 2.26 A for 3b and ca. 2.31 A for 3d, are slightly
longer than those found for the phosphide derivatives men-
˚
tioned above (ca. 2.19 A) and are consistent with single
Fe-P bonds, as expected for a two-electron-donor, bent-
(26) (a) Hirth, U. A.; Malisch, W. J. Organomet. Chem. 1992, 439,
C16. (b) Ho, J.; Hou, Z.; Drake, R. J.; Stephan, D. W. Organometallics
1993, 12, 3145. (c) Ho, J.; Drake, R. J.; Stephan, D. W. J. Am. Chem. Soc.
1993, 115, 3792. (d) Maslennikov, S. V.; Glueck, D. S.; Yap, G. P. A.;
Rheingold, A. L. Organometallics 1996, 15, 2483. (e) Kourkine, I. V.;
Glueck, D. S. Inorg. Chem. 1997, 36, 5160.
(28) (a) Aktas, H.; Slootweg, J. C.; Schakel, M.; Ehlers, A. W.; Lutz,
M.; Spek, A. L.; Lammertsma, K. J. Am. Chem. Soc. 2009, 131, 6666. (b)
Jutzi, P.; Schmidt, H.; Neumann, B.; Stammler, H. G. Organometallics
1996, 15, 741and references therein.
(27) Vahrenkamp, H. J. Organomet. Chem. 1973, 63, 399.

1108 Organometallics, Vol. 30, No. 5, 2011
Alvarez et al.
Chart 3
compound [Ru2Cp*2(μ-PPh)(μ-CO)(CO)2] (457 ppm),29
a
Figure 5. ORTEP diagram of compound 4, with H atoms and
Cy groups (except the C1 atoms) omitted for clarity.
difference expected when comparing Fe and Ru complexes
bridged by PR2 or PR ligands.19 However, we should stress
that the huge difference between the chemical shift of the
PMes* complex (593.5 ppm) and those of the aryl com-
pounds 3b,c (ca. 500 ppm) is totally unexpected and is not
observed in the compounds of type 1 and 2 (Table 1).
Although obviously this large difference must be somehow
related to the great steric demands of the Mes* substituent in
the phosphinidene ligand, currently we cannot offer a satis-
factory explanation for this spectroscopic observation.
Decarbonylation Reactions of Compounds 3a,b. Decarbo-
nylation reactions were attempted for both complexes either
by heating of their toluene solutions or by irradiation with
visible-UV light, but the results were quite different. In the
case of the phenylphosphinidene complex 3b, full decom-
position takes place in toluene solution at 333 K after 1.5 h,
with the dimer [Fe2Cp2(CO)4] being the only new organo-
metallic species detected in the solution at any time. Analo-
gously, the photolysis of toluene solutions of 3b from 243 to
293 K gave mainly the dimer [Fe2Cp2(CO)4] in 0.5 h, along
with trace amounts of other uncharacterized products.
In contrast, compound 3a was found to be fairly robust
thermally, because only a small amount of [Fe2Cp2(CO)4]
was formed after refluxing a toluene solution of the complex
for 1 h. The full thermal degradation of 3a under these
conditions required about 5 h, but no other organometallic
species (apart from the mentioned dimer) were present in the
solution in significant amounts at any time. Fortunately,
decarbonylation of 3a could be accomplished photochemi-
cally, although it required prolonged reaction times (ca. 6 h)
and gave modest yields of products, of which we have been
able to isolate and characterize the tetranuclear cluster
[Fe4Cp4(μ3-PCy)2(μ3-CO)(μ-CO)(CO)] (4) and the trinuclear
isomers syn- and anti-[Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2]
(5) (Chart 3).
˚
Table 3. Selected Bond Lengths (A) and Angles (deg) for
Compound 4
Fe(1)-Fe(2)
2.6115(7)
2.5332(8)
2.244(1)
2.239(1)
2.254(1)
Fe(1)-C(1)
Fe(2)-C(1)
Fe(1)-C(2)
Fe(4)-C(3)
2.016(4)
1.956(5)
1.903(4)
1.727(5)
ꢀ
Fe(1)-Fe(1)
Fe(1)-P(1)
Fe(2)-P(1)
Fe(4)-P(1)
Fe(1)-P(1)-Fe(2)
Fe(1)-P(1)-Fe(4)
Fe(2)-P(1)-Fe(4)
71.3(1)
125.1(1)
105.2(1)
Fe(1)-C(1)-Fe(2)
Fe(1)-C(1)-Fe(2)
82.2(2)
77.8(2)
metal fragment (Fe(CO)Cp). The corresponding metal atom
(Fe(4)), however, is not directly bound to the metal triangle
but is connected to it through two μ3-PCy ligands placed over
the longest edges of the triangle. The phosphinidene-bridged
Fe(1)-Fe(2) bonds have an intermetallic distance of
˚
2.6115(7) A, a figure comparable to the values measured
for the phosphinidene derivatives 3b,d, while the remaining
ꢀ
˚
Fe(1)-Fe(1) bond is considerably shorter (2.5332(8) A),
obviously because of the presence of two CO ligands brid-
ging these metal atoms, with this figure being comparable to
other intermetallic distances in complexes having a Fe2(μ-
˚
CO)2 core (cf. 2.534(2) and 2.531(2) A for the trans and
cis isomers of [Fe2Cp2(μ-CO)2(CO)2]).17 All the Fe-P dis-
tances are almost identical (ca. 2.24 A) and only marginally
shorter than the single-bond figure measured for the bi-
˚
˚
nuclear phosphinidene complex 3b (ca. 2.26 A), this indicat-
ing a fairly even distribution of the four bonding electrons
of the PR ligands among the three metal centers in each
case. Other structural parameters in this compound are as
expected.
Spectroscopic data in solution for compound 4 are con-
sistent with the solid-state structure discussed above. The
retention of three carbonyls with different coordination
modes is evident from the presence in the IR spectrum of
three C-O stretching bands with very different frequencies,
which can be respectively assigned to terminal (1898 cm-1),
μ2 (1746 cm-1), and μ3 (1621 cm-1) carbonyl ligands. This is
paralleled by the observation of very differently deshielded
carbonyl resonances in the 13C{1H} NMR spectrum, at 303.0
(μ3-CO), 273.5 (μ2-CO), and 224.2 ppm (terminal CO), these
being in the typical ranges for diiron complexes having
Structural Characterization of Compound 4. The structure
of 4 in the crystal form (Figure 5 and Table 3) displays a
central trimetal core built from three CpFe fragments defin-
ing an isosceles triangle, with a slightly asymmetric CO
ligand bridging this triangle and a second CO ligand bridging
the shorter edge. These carbonyl ligands and the Fe(2) atom
are placed in a plane of symmetry containing the fourth
(29) Takemoto, S.; Kimura, Y.; Kamikawa, K.; Matsuzaka, H.
Organometallics 2008, 27, 1780.

Article
Organometallics, Vol. 30, No. 5, 2011 1109
carbonyl ligands in those coordination modes.9,15c,15e,15f,30
The phosphinidene ligands also remain equivalent in solu-
tion and give rise to a single 31P resonance at 352.2 ppm, a
chemical shift that falls in the expected range for phosphini-
dene ligands bridging three iron atoms.19,31 Other spectro-
scopic data of 4 are as expected and deserve no detailed
comment.
Structural Characterization of Compound 5. Spectroscopic
data for the two isomers of compound 5 (Table 1 and
Experimental Section) are similar to each other, thus indi-
cating that these compounds share the same basic structure
(Chart 3), which we propose to differ only in the relative
conformation of the PHCy bridge with respect to the phos-
phinidene ligand, with the major isomer having the Cy
groups away from each other (anti-5), presumably more
favored on steric grounds. The relative arrangement pro-
posed for the four bridging ligands (PCy, PHCy, and 2CO)
present in both isomers of this 50-electron complex (there-
fore having only two metal-metal bonds according to the
EAN formalism) is based on the solid-state structures found
for related trinuclear clusters with two semibridging CO
ligands, such as the carbyne-bridged complexes [MnMo2-
Cp2Cp0(μ3-CH)(μ-PCy2)(CO)4], and [MnMo2Cp2Cp0(μ3-
COMe)(μ-PCy2)(CO)4]32 and the phosphinidene-bridged
cluster [Fe2MnCp(μ3-PPh)(CO)8],33 although all these species
are 48-electron complexes with three intermetallic bonds. In
the above Mo2Mn clusters, the phosphide ligand is placed
close to the intermetallic plane, while the face-bridging ligand
and the semibridging carbonyls areplaced ondifferent sides of
the intermetallic plane, and this is the overall arrangement
proposed for both isomers of 5.
The IR spectra for both isomers of 5 exhibit two C-O
stretching bands at low frequencies that can be assigned to
two CO ligands bridging two metal atoms an placed on the
same side of the intermetallic plane, according to their
relative intensities (very strong and weak in order of decreas-
ing frequencies). Moreover, these ligands are related by
symmetry, as deduced from the presence of a single and very
deshielded carbonyl resonance at 281.2 ppm in the 13C{1H}
NMR spectrum of the major isomer. The 31P NMR spectra
show in each case a pair of mutually coupled resonances
(JPP = ca. 300 Hz): a quite deshielded one at ca. 500 ppm,
corresponding to a μ3-PCy ligand, and a strongly shielded
one at around -30 ppm, corresponding to a PHCy ligand
(JPH = ca. 250-300 Hz). The low chemical shift of the latter
resonance is characteristic of bridging phosphide ligands in
the absence of metal-metal bonds,19,34 while the observed
P-P couplings are comparable to those measured for the
related cluster [Fe3(μ-H)(μ3-PH)(μ-PH2)(CO)9].35 The main
Figure 6. ORTEP diagram of compound 6, with H atoms
(except the P-H atom) and p-Me groups omitted for clarity.
˚
Table 4. Selected Bond Lengths (A) and Angles (deg) for Com-
pound 6
Fe(1)-Fe(2)
Fe(1)-P(1)
Fe(2)-P(2)
Fe(3)-P(1)
Fe(3)-P(2)
2.541(1)
2.213(7)
2.165(8)
2.315(7)
2.161(8)
Fe(1)-C(1)
Fe(1)-C(2)
Fe(2)-C(1)
Fe(2)-C(2)
Fe(3)-C(3)
1.912(5)
1.905(6)
1.906(5)
1.934(6)
1.729(6)
Fe(1)-P(1)-Fe(3)
Fe(2)-P(2)-Fe(3)
Fe(2)-P(2)-C(5)
Fe(3)-P(2)-C(5)
122.4(3)
132.5(4)
111.2(4)
114.9(4)
Fe(1)-Fe(2)-P(2)
Fe(2)-Fe(1)-P(1)
P(1)-Fe(3)-P(2)
93.0(1)
100.4(1)
83.5(3)
spectroscopic differences between the syn and anti isomers of
compound 5 are observed in the P-bound hydrogen reso-
nance: the major isomer displays a quite shielded resonance
exhibiting no coupling to the phosphinidene ligand (δ 0.49
ppm, 1JHP = 253 Hz), while the minor isomer displays a less
shielded resonance that is weakly coupled to the PCy ligand
1
3
(δ 2.33 ppm, JHP = 304 Hz, JHP = 7 Hz). The observed
differences in the 3JPH values are consistent with the assigned
geometries by recalling the Karplus-type dependence with
the dihedral angles usually observed for three-bond P-H
couplings,36 this implying larger couplings for dihedral
H-P-Fe-P angles close to 180° (syn-5).
Decarbonylation Reactions of Compound 3c. Surprisingly,
the mesitylphosphinidene complex did not undergo photo-
chemically induced decarbonylation or any other process at
a significant extent under ordinary conditions. In contrast,
refluxing a toluene solution of 3c for ca. 6 h gave the trimetal
complex [Fe3Cp3(μ-PMes)(μ-PMesH)(μ-CO)2(CO)] (6) in
medium yield (ca. 40%), along with small amounts of the
dimer [Fe2Cp2(CO)4].
The molecule of 6 (Figure 6 and Table 4) is built from three
FeCp fragments, two of them connected through an inter-
metallic bond and two bridging CO ligands, while the third
one is connected to the other metal atoms through either a
bridging phosphide ligand (to Fe(1)) or a trigonal phosphi-
nidene ligand (to Fe(2)) and completes its coordination
sphere with a terminal carbonyl located opposite to the
mesityl group of the phosphide ligand, presumably to mini-
mize steric repulsions. In the crystal, the environments
around the Fe(1)/Fe(2) and P(1)/P(2) atoms are disordered
but could be satisfactorily refined with 0.5 occupancy factors
(thus, the H(1) atom is actually disordered between P(1) and
(30) (a) Sunick, D. L.; White, P. S.; Schauer, C. K. Organometallics
1993, 12, 245. (b) Borg-Breen, C. C.; Bautista, M. T.; Schauer, C. K.;
White, P. S. J. Am. Chem. Soc. 2000, 122, 3952. (c) Parkinson, J. L.; Baik,
M. H.; Trullinger, G. E.; Schauer, C. K.; White, P. S. Inorg. Chim. Acta
1999, 294, 140.
(31) Yeh, W. Y.; Liu, Y. C.; Peng, S. M.; Lee, G. H. J. Organomet.
Chem. 2004, 689, 1014.
(32) (a) Alvarez, M. A.; Garcıa, M. E.; Martınez, M. E.; Ruiz, M. A.
ꢀ
Organometallics 2010, 29, 904. (b) Garcıa, M. E.; Garcıa-Vivo, D.; Ruiz,
M. A. Organometallics 2009, 28, 4385.
(33) (a) Huttner, G.; Frank, A.; Mohr, G. Z. Naturforsch., B 1978, 31,
1161. (b) Huttner, G.; Schneider, J.; Mohr, G.; von Seyerl, J. J.
Organomet. Chem. 1980, 191, 161.
˚
P(2)). The Fe(1)-Fe(2) length of 2.541(1) A is almost
(34) Decken, A.; Neil, M. A.; Dyker, C. A.; Bottomley, F. Can. J.
Chem. 2002, 80, 55.
(35) Bautista, M. T.; White, P. S.; Schauer, C. K. J. Am. Chem. Soc.
1991, 113, 8963.
(36) Bentrude, W. G.; Setzer, W. N. In Phosphorus-31 NMR Spec-
troscopy in Stereochemical Analysis; Verkade, J. G., Quin, L. D., Eds.;
VCH: Deerfield Beach, FL, 1987; Chapter 11.

1110 Organometallics, Vol. 30, No. 5, 2011
Alvarez et al.
Chart 4
phosphinidene complex [Mo2Cp2(μ-PMes*)(μ-CO)2]37 and
even to those of the isoelectronic phosphide complexes
[Fe2Cp2(μ-PR2)(μ-PR02)(μ-CO)] (R = Ph, R0 = Ph, Et).15f
However, no evidence has been obtained for the presence of
this species in the reaction mixtures after decarbonylation of
compounds 3a,c. The formation of 4, however, can be viewed
as resulting from the coupling of two of such units and a
subsequent decarbonylation of the resulting [Fe4Cp4(μ-
PR)2(CO)4] intermediate. On the other hand, the formation
of 5 can be rationalized by assuming that, as a side process,
homolytic fission of the Fe-Fe bond in 3a might take place
at some extent, this yielding the FeCp(CO)2 and FeCp(CO)-
(PCy) radicals, with the former explaining the appearance of
[Fe2Cp2(CO)4] in the corresponding reaction mixtures. The
phosphinidene radical might then be able to couple to a
molecule of [Fe2Cp2(μ-PR)(μ-CO)2] to give the paramag-
netic intermediate [Fe3Cp3(μ-PR)2(CO)4] that eventually
would yield 5 after hydrogen abstraction (possibly from
the solvent) and further loss of two CO molecules. Indeed,
the formation of compound 6 might be rationalized in the
same way, except that now only one CO ligand would be
removed from the trimetallic intermediate after hydrogen
abstraction.
˚
identical with the value of ca. 2.53 A measured for both
isomers of [Fe2Cp2(μ-CO)2(CO)2],17a,b as expected. On the
basis of simple electron counts, the phosphorus ligands in 6
should provide two electrons to each of the Fe(1) and Fe(2)
atoms and a total of three electrons to the Fe(3) atom. In
agreement with this, the phosphide ligand is quite asymme-
˚
trically placed, substantially closer to Fe(1) (2.213(7) A) than
˚
to Fe(2) (2.315(7) A), while the trigonal phosphinidene
ligand (sum of X-P(2)-Y angles ca. 360°) behaves as a
symmetrical four-electron donor and displays two short and
comparable P-Fe lengths (2.165(8) and 2.161(8) A). The
latter distances are similar to the average Fe-P length in the
trigonal phosphinidene complex [Fe2(CO)6(μ-POR)2]
(2.157(1) A) and are significantly shorter (by 0.10-0.15
A) than the values measured for the pyramidal phosphini-
dene complexes 3b,d. We notice that the only other reported
diiron complex with a trigonal phosphinidene ligand dis-
plays a much longer Fe-P separation of ca. 2.45 A.
˚
7
Decarbonylation Reactions of Compound 3d. This complex
has only a limited thermal stability. In fact, it rearranges
slowly (conversion ca. 80% after 4 day) in toluene solution at
room temperature by undergoing intramolecular insertion of
the P atom into a C-H bond of one of the close o-tBu groups
˚
˚
8
˚
˚
(P H = ca. 2.55 A in the crystal), to give the phosphine
3 3 3
The spectroscopic data for 6 are consistent with the
retention in solution of the essential features of the solid-
state structure discussed above, although two isomers are
present in a ratio of ca. 10:1 at room temperature in
dichloromethane. These would follow from the two possible
conformations (syn and anti) of the Cp ligand in the FeCp-
CO fragment with respect to the mesityl group of the
phosphide ligand (Chart 4). We trust that the syn isomer,
that present in the crystal, would also be the major isomer in
solution. In any case, both isomers display two similar and
largely separated 31P resonances, as found for 5. That
corresponding to the PHMes ligand appears quite shielded
(ca. 25 ppm), consistent with the lack of an intermetallic
interaction between the implied iron atoms. In contrast, the
resonance of the phosphinidene ligand now appears at ca.
750 ppm, some 350 ppm above the figure for the pyramidal
phosphinidene complex 3c, this being consistent with the
presence of a trigonal phosphinidene bridge.19 Other spec-
t
derivative [Fe2Cp2(μ-CO)2(CO){PH(CH2CMe2)C6H2 Bu2}]
(7). As found for compounds 1c,d, complex 7 exists in
solution as an equilibrium mixture of trans (major) and cis
(minor) isomers, these differing in the relative orientation of
the Cp or (CO/phosphine) ligands, and deserves no further
discussion. Expectedly, the transformation leading to 7 is
much faster at higher temperatures, and it can be completed
in ca. 1 h at 338 K in toluene solution, but to our surprise this
transformation could not be suppressed under photolytic
conditions, even at low temperatures. In fact, all photolytic
experiments carried out by us for 3d led to mixtures of
compound 7 and the dicarbonyl hydrides [Fe2Cp2(μ-H){μ-
t
P(CH2CMe2)C6H2 Bu2}(CO)2] (8) (cis and trans isomers),
with the relative amount of hydride complexes increasing at
prolonged reaction times, thus suggesting that they are
generated photochemically from 7, as shown by independent
tests (Scheme 4).
The structures of both isomers of compound 8 can be
deduced easily from the corresponding spectroscopic data
and their comparison with the data available for other diiron
hydride complexes of the type [Fe2Cp2(μ-H)(μ-PRR0)-
(CO)2].15c,f,38 The major difference between the trans and
cis isomers, of course, is found in the reversed pattern of the
C-O stretching bands (Table 1), while the asymmetry of the
1
troscopic features in the H and 13C NMR spectra of the
major isomer (see the Experimental Section) are fully con-
sistent with the structure found in the crystal and deserve no
further comments, while most of the resonances for the
minor isomer were obscured by those of the major isomer.
Reaction Pathways in the Formation of Compounds 4-6.
The removal of a CO molecule from compounds 3 would be
expected to give a dicarbonyl derivative of the type [Fe2Cp2-
(μ-PR)(μ-CO)2], most likely with three bridging groups, one
of them being a four-electron-donor, trigonal phosphinidene
ligand. Such a structure would be comparable to that of the
1
trans isomer is readily apparent in the H NMR spectrum,
displaying separate resonances for each of the Cp or Me
groups. In contrast, the isomer cis-8 has a symmetry plane
relating both metal fragments, and accordingly it displays
single Cp (4.29 ppm), CH2 (2.43 ppm), and CMe2 (1.53 ppm)
resonances. Out of the two possible isomers having a cis
arrangement of the terminal ligands in 8, the observed isomer
most likely corresponds to that having its bulkier aryl
ꢀ
(37) (a) Garcıa, M. E.; Riera, V.; Ruiz, M. A.; Saez, D.; Vaissermann,
J.; Jeffery, J. C. J. Am. Chem. Soc. 2002, 124, 14304. (b) Garcıa, M. E.;
ꢀ
Riera, V.; Ruiz, M. A.; Saez, D.; Hamidov, H.; Jeffery, J. C.; Riis-
Johannessen, T. J. Am. Chem. Soc. 2003, 125, 13044. (c) Amor, I.;
ꢀ
Garcıa, M. E.; Ruiz, M. A.; Saez, D.; Hamidov, H.; Jeffery, J. C.
Organometallics 2006, 25, 4857.
(38) (a) Brunner, H.; Peter, H. J. Organomet. Chem. 1990, 393, 411.
€
(b) Brunner, H.; Rotzer, M. J. Organomet. Chem. 1992, 425, 119.

Article
Scheme 4. Photochemical Decarbonylation of Compound 7
Organometallics, Vol. 30, No. 5, 2011 1111
Scheme 5. Reaction Pathways in the Formation of Compound 7
fragment away from the Cp ligands (as observed for the
phosphinidene complexes 3), more favored on steric
grounds, although we have not investigated this matter in
detail. Finally, we note that the hydride ligand in both
isomers gives rise to a quite shielded resonance at ca. -18
ppm (JHP = 39), comparable to those found in the literature
for related complexes.
Reaction Pathways in the Formation of Compounds 7 and 8.
The formation of compound 7 from 3d obviously requires
the insertion of the P atom into the C-H bond of one of the
closer o-tBu groups at some stage. We recently reported that
the dimolybdenum complex [Mo2Cp2(μ-PR*)(CO)4], having
a trigonal (four-electron-donor) PR* bridge, undergoes a
related C-H bond cleavage reaction possibly via a photo-
excited state having a pyramidal (two-electron-donor) PR*
bridge.37c On this basis, it could be assumed that the C-H
cleavage step might take place directly at the phosphinidene-
bridged molecule of 3d (Scheme 5). This would yield the
unstable intermediate F with a bridging phosphine that
would rearrange into a terminal position, thus generating
one of the isomers of compound 7. In any case, regardless of
the isomer formed initially (cis or trans), they would inter-
convert afterward to reach the corresponding equilibrium
ratio, as already discussed in previous sections. However,
this information together, we might also conceive as a
possible event for 3d the presence in solution of a tiny
amount of terminal isomer T, that being perhaps the active
species inducing the cleavage of the C-H bond, leading
initially to the trans isomer of 7 (Scheme 5). At present,
however, we cannot exclude any of the above proposals.
Once the phosphine complex 7 is formed, its photochemi-
cal decarbonylation to give the hydrides 8 seems completely
analogous to the formation of the hydrides [Fe2Cp2(μ-H)(μ-
PPh2)(CO)2] from the phosphine complex trans-[Fe2Cp2(μ-
CO)2(CO)(PPh2H)] reported previously by us15f and does
not need to be discussed in detail again. We just note that,
according to this previous study, the trans isomer is formed
first, in good agreement with the fact that trans-8 is the only
isomer obtained upon photolysis of 3d using short reaction
times, while the cis isomer would be photogenerated from the
trans isomer at a lower rate (Scheme 4).
t
related activations of a Bu group in a compound with a
t
PMes* ligand to generate a PH(CH2CMe2)C6H2 Bu2 ligand
have been previously reported to occur also on a few
transient mononuclear complexes,39 even with liberation of
the free phosphine.39,40 On this basis, then, one might con-
ceive that such an activation reaction would surely be easier
for a terminal bent phosphinidene, because of its higher
coordinative unsaturation. Interestingly, we found recently
that the oxophosphinidene derivative of compound 3d exists
in solution as a mixture of isomers having either terminal or
bridging oxophosphinidene ligands (P(O)R*).10 Taking all
Concluding Remarks
The diiron phosphinidene complexes 3 have been selec-
tively synthesized from the corresponding PRH2 complexes
1 via the PRH-bridged cations 2. The critical step is the
formation of the phosphide complexes, this being oxidatively
induced in two different ways: either though a one-electron
oxidation followed by spontaneous dehydrogenation (R =
Cy, Ph, Mes) or through a thermodynamically disfavored
two-electron oxidation coupled with deprotonation (R =
Mes*). The final step is a conventional deprotonation of the
bridging PRH ligand in the cationic complexes 2 to yield the
corresponding derivatives with a pyramidal phosphinidene
ligand. To our knowledge, this is a novel and highly efficient
synthetic procedure to obtain pyramidal phosphinidene
ligands bridging two metal atoms from the corresponding
phosphine complexes.
(39) (a) Champion, D. H.; Cowley, A. H. Polyhedron 1985, 4, 1791.
(b) Arif, A. M.; Cowley, A. H.; Pakulski, M.; Pearsall, M. A.; Clegg, W.;
Norman, N. C.; Orpen, A. G. J. Chem. Soc., Dalton Trans. 1988, 2713.
(c) Hey-Hawkins, E.; Kurz, S. J. Organomet. Chem. 1994, 479, 125. (d)
Masuda, J. D.; Hoskin, A. J.; Graham, T. W.; Beddie, C.; Fermin, M. C.;
Etkin, N.; Stephan, D. W. Chem. Eur. J. 2006, 12, 8696. (e) Masuda,
J. D.; Jantunen, K. C.; Ozerov, O. V.; Noonan, K. J. T.; Gates, D. P.;
Scott, B. L.; Kiplinger, J. L. J. Am. Chem. Soc. 2008, 130, 2408.
(40) (a) Shah, S.; Simpson, M. C.; Smith, R. C.; Protasiewicz, J. D. J.
Am. Chem. Soc. 2001, 123, 6925. (b) Smith, R. C.; Shah, S.; Protasiewicz,
J. D. J. Organomet. Chem. 2002, 646, 255.

1112 Organometallics, Vol. 30, No. 5, 2011
Alvarez et al.
Decarbonylation of the PR-bridged complexes 3 led to no
stable complexes with trigonal phosphinidene bridges of the
type [Fe2Cp2(μ-PR)(μ-CO)2], and the results were strongly
dependent on the nature of R. While no stable derivatives
were obtained when R = Ph, intramolecular insertion of the
P atom into the C-H bond of an o-tBu group dominated the
reactions of the sterically congested PMes* complex. In
contrast, the decarbonylation reactions of the PCy and PMes
complexes gave tetranuclear (Cy) or trinuclear (Cy, Mes)
derivatives that might be formed from the putative
[Fe2Cp2(μ-PR)(μ-CO)2] intermediate through either dimer-
ization or reaction with mononuclear radicals FeCpPR
arising from thermally and photochemically induced homo-
lytic cleavage of the metal-metal bond in the parent phos-
phinidene complexes.
Calcd for C19H23Fe2O3P: C, 51.62; H, 5.24. Found: C, 51.23; H,
5.05. 1H NMR (CDCl3): δ 4.72 (s, 5H, Cp), 4.50 (d, JHP = 1, 5H,
Cp), 3.28 (dd, JHP = 327, JHH = 6, 2H, P-H), 2.05-0.97 (m,
11H, Cy).
Preparation of [Fe2Cp2(μ-CO)2(CO)(PPhH2)] (1b). The pro-
cedure is completely analogous to that described for 1a but using
PPhH2 instead (315 μL, 2.864 mmol). Compound 1b was thus
obtained as a dark green microcrystalline solid (1.108 g, 90%).
Anal. Calcd for C19H17Fe2O3P: C, 52.34; H, 3.93. Found: C,
51.90; H, 3.65. 1H NMR (CDCl3): δ 7.70-7.30 (m, 5H, Ph), 4.76
(s, 5H, Cp), 4.47 (d, JHP = 342, 2H, P-H), 4.45 (d, JHP = 1, 5H,
Cp).
Preparation of [Fe2Cp2(μ-CO)2(CO)(PMesH2)] (1c). The pro-
cedure is completely analogous to that described for 1a, but
using PMesH2 instead (5 mL of a 0.62 M solution in CH2Cl2,
3.10 mmol). The green solid obtained after removal of solvent
under vacuum from the reaction mixture was dissolved in the
minimum amount of a dichloromethane/petroleum (1/2) mix-
ture and chromatographed on alumina (activity II) at 288 K.
Elution with dichloromethane/petroleum ether (1/4) gave a
yellow fraction containing excess phosphine and related impu-
rities and a red fraction containing some [Fe2Cp2(CO)4]. Elution
with dichloromethane/petroleum ether (1/4) gave a green frac-
tion which gave, after removal of solvents under vacuum,
compound 1c as a green microcrystalline solid (1.120 g, 83%).
This complex exists in solution as an equilibrium mixture of the
corresponding trans and cis isomers. Anal. Calcd for
C22H23Fe2O3P: C, 55.27; H, 4.85. Found: C, 55.23; H, 4.90.
Spectroscopic data for trans-1c are as follows. 1H NMR: δ 6.91
(s, 2H, C6H2), 4.74, 4.19 (2s, 2 ꢀ 5H, Cp), 3.87 (d, JHP = 345,
2H, P-H), 2.45 (s, 6H, o-Me), 2.26 (s, 3H, p-Me). The reso-
nances due to the isomer cis-1c were obscured by those of the
major isomer in the 1H NMR spectrum. Trans/cis ratio: ca. 20/1
(from the 31P{1H} NMR spectrum).
Preparation of [Fe2Cp2(μ-CO)2(CO)(PMes*H2)] (1d). The
procedure is completely analogous to that described for 1a,
but using PMes*H2 instead (0.787 g, 2.827 mmol). The green
residue obtained after removal of the solvent was washed with
petroleum ether (6 ꢀ 10 mL) to give compound 1d as a dark
green microcrystalline solid. The petroleum ether fractions were
collected and the solvent removed under vacuum. The residue
was then dissolved in dichloromethane/petroleum ether (1/18)
and chromatographed on alumina (activity IV) at 288 K.
Elution with dichloromethane/petroleum ether (1/14) gave a
red fraction containing some [Fe2Cp2(CO)4] and then a green
fraction which yielded, after removal of solvents, a second crop
of compound 1d (combined yield: 1.520 g, 89%). This complex
exists in solution as an equilibrium mixture of the corresponding
trans and cis isomers. Anal. Calcd for C31H41Fe2O3P: C, 61.61;
H, 6.84. Found: C, 61.49; H, 6.88. Spectroscopic data for trans-
1d are as follows. 1H NMR (400.13 MHz) δ 7.45 (s, 2H, C6H2),
4.71 (s, 5H, Cp), 4.44 (d, JHP = 343, 2H, P-H), 3.99 (s, 5H, Cp),
1.59 (s, 18H, o-tBu), 1.31 (s, 9H, p-tBu). Spectroscopic data for
cis-1d are as follows. 1H NMR (400.13 MHz): δ 4.71 (s, 5H, Cp),
4.51 (d, JHP = 343, 2H, P-H), 3.94 (s, 5H, Cp), 1.65 (s, 18H,
o-tBu); other resonances of this isomer were obscured by those
of the major isomer. Trans/cis ratio: 5/1.
Experimental Section
General Procedures and Starting Materials. All manipulations
and reactions were carried out under a nitrogen (99.995%)
atmosphere using standard Schlenk techniques. Solvents were
purified according to literature procedures and distilled prior to
use.41 Petroleum ether refers to that fraction distilling in the
range 338-343 K. Compounds [FeCp2]BF4,42 [Fe2Cp2(μ-CO)2-
(CO)(CNMe)],13 PMesH2,39d and PMes*H243 were synthesized
according to literature procedures. All other reagents were
obtained from the usual commercial suppliers and used as
received. Photochemical experiments were performed using
jacketed quartz or Pyrex Schlenk tubes, cooled by tap water
(ca. 288 K) or by a closed 2-propanol circuit kept at the desired
temperature with a cryostat. A 400 W mercury lamp placed ca. 1
cm away from the Schlenk tube was used for all the experiments.
Chromatographic separations were carried out using jacketed
columns cooled by tap water (ca. 288 K) or by a closed
2-propanol circuit, kept at the desired temperature with a
cryostat. Commercial aluminum oxide (activity I, 150 mesh)
was degassed under vacuum prior to use. The latter was mixed
afterward under nitrogen with the appropriate amount of water
to reach the activity desired. IR C-O stretching frequencies
were measured in solution and are referred to as ν(CO). Nuclear
magnetic resonance (NMR) spectra were routinely recorded at
300.13 (1H), 121.50 (31P{1H}), and 75.48 MHz (13C{1H}) at 290
K in CD2Cl2 solutions unless otherwise stated. Chemical shifts
(δ) are given in ppm, relative to internal tetramethylsilane (1H,
13C) or external 85% aqueous H3PO4 (31P). Coupling constants
(J) are given in Hz. Cyclic voltammetry experiments were
performed in an airtight custom-made electrolytic cell using a
Pt working electrode, an Ag-wire reference electrode and a Pt-
wire auxiliary electrode. [NBu4][PF6] was used as the supporting
electrolyte (0.1 M solutions). The analyte concentrations in a
typical experiment were ca. 10-3 M. All the potentials were
referenced versus the ferrocene/ferrocenium couple (Cp2Feþ/0),
used as internal reference in all the experiments.
Preparation of [Fe2Cp2(μ-CO)2(CO)(PCyH2)] (1a). An acet-
onitrile solution (50 mL) of [Fe2Cp2(μ-CO)2(CO)(CNMe)] was
prepared in situ from [Fe2Cp2(CO)4] (1.000 g, 2.825 mmol). The
solvent was then removed under vacuum and the residue
dissolved in dichloromethane (20 mL) and stirred with PCyH2
(375 μL, 2.825 mmol) at room temperature for 5 min to give a
dark green solution, which was filtered. Solvent was then
removed under vacuum from the filtrate, and the residue was
washed with petroleum ether (5 ꢀ 10 mL) to give compound 1a
as a dark green microcrystalline solid (1.125 g, 90%). Anal.
Preparation of [Fe2Cp2(μ-PCyH)(μ-CO)(CO)2]BF4 (2a). So-
lid [FeCp2]BF4 (0.763 g, 2.798 mmol) was slowly added to a
stirred dichloromethane solution (30 mL) of compound 1a
(1.125 g, 2.545 mmol) at 273 K, and the mixture was further
stirred for 10 min to give a red solution. The solvent was then
removed under vacuum, and the residue was washed with
petroleum ether (7 ꢀ 10 mL) to give compound 2a as a red
microcrystalline solid (1.303 g, 97%). Anal. Calcd for
C19H22BF4Fe2O3P: C, 43.23; H, 4.20. Found: C, 42.92; H,
(41) Armarego, W. L. F.; Chai, C. Purification of Laboratory Che-
micals, 5th ed.; Butterworth-Heinemann: Oxford, U.K., 2003.
(42) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877.
(43) Cowley, A. H.; Norman, N. C.; Pakulski, M. Inorg. Synth. 1990,
27, 253.
1
3.89. H NMR (CDCl3): δ 8.13 (dd, JHP = 390, JHH = 10,
1H, P-H), 5.17 (s, 10H, Cp), 2.25-1.15 (m, 11H, Cy).
Preparation of [Fe2Cp2(μ-PPhH)(μ-CO)(CO)2]BF4 (2b). The
procedure is completely analogous to that described for 2a. With

Article
Organometallics, Vol. 30, No. 5, 2011 1113
1.108 g (2.541 mmol) of compound 1b as starting material and
using 0.762 g of [FeCp2]BF4 (2.795 mmol), compound 2b was
obtained as a red microcrystalline solid (1.272 g, 96%). Anal.
Calcd for C19H16BF4Fe2O3P: C, 43.73; H, 3.09. Found: C,
43.45; H, 3.15. Spectroscopic data for cis,anti-2b are as follows.
1H NMR: δ 9.10 (d, JHP = 413, 1H, P-H), 7.80-7.40 (m, 5H,
Ph), 5.25 (s, 10H, Cp). The resonances due to the isomer cis,syn-
2b were obscured by those of the major isomer in the 1H NMR
spectrum. Anti/syn ratio: ca. 10/1 (from the 31P{1H} NMR
spectrum).
Preparation of [Fe2Cp2(μ-PMesH)(μ-CO)(CO)2]BF4 (2c). So-
lid [FeCp2]BF4 (0.703 g, 2.578 mmol) was slowly added to a
stirred dichloromethane solution (30 mL) of compound 1c
(1.120 g, 2.343 mmol) at 273 K, and the mixture was further
stirred for 10 min to give a red solution. Excess KOH (ca. 0.5 g)
was added to this solution, and the mixture was stirred for a
further 10 min to give an orange solution containing compound
3c. The solvent was then evaporated under vacuum and the
residue was dissolved in the minimum amount of dichloro-
methane/petroleum ether (1/1) and chromatographed on alu-
mina (activity IV) at 288 K. Elution with dichloromethane/
petroleum ether (1/2) gave a yellow fraction containing [FeCp2],
and elution with dichloromethane/petroleum ether (1/1) gave an
orange fraction containing compound 3c. The solvents were
removed under vacuum, and the residue was dissolved in
52.13; H, 3.17. 1H NMR: δ 7.60-7.20 (m, 5H, Ph), 4.90 (s, 10H,
Cp).
Preparation of [Fe2Cp2(μ-PMes)(μ-CO)(CO)2] (3c). Method A
(Small Scale). This synthetic procedure is completely analogous to
that described for 3a. With compound 2c (0.050 g, 0.093 mmol) as
starting material and using KOH (ca. 0.1 g, excess), compound 3c
was obtained as an orange, very air-sensitive microcrystalline solid
(0.042 g, 95%).
Method B (Large Scale). This synthetic procedure is analo-
gous to that described for 2c, but omitting the final step
(protonation). Yield: 0.960 g (86%). Satisfactory elemental
analysis could not be obtained for this very air-sensitive materi-
al. 1H NMR: δ 6.85 (s, 2H, C6H2), 4.91 (s, 10H, Cp), 2.53 (s, 6H,
o-Me), 2.25 (s, 3H, p-Me).
Preparation of [Fe2Cp2(μ-PMes*)(μ-CO)(CO)2] (3d). The
procedure is completely analogous to that described for 3a.
With compound 2d (0.050 g, 0.072 mmol) as starting material
and using KOH (ca. 0.1 g, excess), compound 3d was obtained as
a wine red, very air-sensitive microcrystalline solid (0.042 g,
95%). Satisfactory elemental analysis could not be obtained for
this very air-sensitive material. 1H NMR: δ 6.92 (s, 2H, C6H2),
4.86 (s, 10H, Cp), 1.58 (s, 18H, o-tBu), 1.07 (s, 9H, p-tBu).
Photolysis of Compound 3a. A toluene solution (8 mL) of
compound 3a (0.200 g, 0.455 mmol) was irradiated with vis-UV
light for 6 h at 288 K while N2 was gently bubbled through the
solution, to give a brown mixture. The solvent was then removed
under vacuum, the residue was extracted with dichloromethane/
petroleum ether (1/1), and the extracts were chromatographed
on alumina (activity IV) at 288 K. Elution with the same solvent
mixture gave a green fraction yielding, upon removal of solvents
under vacuum, the compound [Fe4Cp4(μ3-PCy)(μ3-CO)(μ-CO)-
(CO)] (4) as a green microcrystalline solid (0.040 g, 11%). The
crystals of 4 used in the diffraction study were grown by the slow
diffusion of a layer of petroleum ether into a CH2Cl2 solution of
the complex at 253 K. Elution with dichloromethane/petroleum
ether (3/1) gave a red-brown fraction yielding analogously the
compound syn-[Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2] (syn-5) as a
brown solid (0.021 g, 7%) and then a brown fraction yielding
compound anti-[Fe3Cp3(μ3-PCy)(μ-PCyH)(μ-CO)2] (anti-5)
also as a brown solid (0.035 g, 12%). Data for compound 4
are as follows. Anal. Calcd for C35H42Fe4O3P2: C, 52.81; H,
5.32. Found: C, 52.53; H, 5.03. 1H NMR (CDCl3): δ 4.71, 4.70
(2s, 2 ꢀ 5H, Cp), 4.62 (s, 10H, Cp), 2.07-1.25 (m, 22H, Cy).
13C{1H} NMR (CDCl3): δ 303.0 (s, μ3-CO), 273.5 (s, μ-CO),
224.2 (t, JCP = 28, FeCO), 90.5 (s, 2Cp), 87.3, 80.6 (2 s, Cp), 64.0
CH2Cl2 (20 mL). Then, HBF4 OEt2 (325 μL of a 54% solution
3
in Et2O, 2.358 mmol) was added to the latter solution, and the
mixture was stirred for 5 min to give a red solution. The solvent
was then removed under vacuum, and the red residue was
washed with Et2O (6 ꢀ 8 mL) and then petroleum ether (4 ꢀ 8
mL) to give compound 2c as a red microcrystalline solid (1.056 g,
80%). Anal. Calcd for C22H22BF4Fe2O3P: C, 46.86; H, 3.93.
Found: C, 46.78; H, 3.80. 1H NMR: δ 9.00 (d, JHP = 397, 1H,
P-H), 7.00 (d, JHP = 4, 2H, C6H2), 5.39 (s, 10H, Cp), 2.73 (s,
6H, o-Me), 2.27 (s, 3H, p-Me).
Preparation of [Fe2Cp2(μ-PMes*H)(μ-CO)(CO)2]BF4 (2d).
Solid [FeCp2]BF4 (0.400 g, 1.467 mmol) was slowly added to a
stirred suspension of NaHCO3 (1.0 g, 11.9 mmol) in a THF
solution (22 mL) of compound 1d (0.400 g, 0.662 mmol) at 243
K, and the mixture was further stirred for 1 h at the same
temperature to give a red solution containing compound 2d as
the major species. The mixture was then allowed to reach room
temperature (this causing a partial conversion of 2d into the
phosphinidene complex 3d) and filtered with a canula. After
HBF4 OEt2 (150 μL of a 54% solution in Et2O, 1.32 mmol) was
3
(false t, |JCP þ JCP
ꢀ
| = 17, C1(Cy)), 34.5, 31.5 (2s, C2(Cy)), 26.6
added to the filtrate, the mixture was stirred for 5 min to give a
red solution. Removal of the solvent under vacuum from the
latter solution and washing of the residue with petroleum ether
(5 ꢀ 20 mL) gave compound 2d as a red microcrystalline solid
(0.360 g, 79%). Anal. Calcd for C31H40BF4Fe2O3P: C, 53.95; H,
5.84. Found: C, 54.00; H, 5.88. 1H NMR (400.13 MHz): δ 9.88
(d, JHP = 385, 1H, P-H), 7.08 (s, 2H, C6H2), 5.32 (s, 10H, Cp),
1.64 (s, 18H, o-tBu), 1.06 (s, 9H, p-tBu).
(s, br, C3(Cy)), 25.5 (s, C4(Cy)). 13C{1H} NMR (CDCl3, 233 K)
δ 27.4, 27.3 (2s, C3(Cy)). Spectroscopic data for syn-5 are as
follows. 1H NMR: δ 4.58 (s, 5H, Cp), 4.19 (s, 10H, Cp), 3.04 (m,
2H, Cy), 2.48 (m, 2H, Cy), 2.33 (dt, JHP = 304, JHH = JHP = 7,
1H, P-H), 2.23 (m, 2H, Cy), 2.06 (m, 1H, Cy), 1.90 (m, 2H, Cy),
1.78-1.50 (m, 8H, Cy), 1.29 (m, 2H, Cy), 1.14 (m, 2H, Cy), 1.01
(m, 1H, Cy). Data for anti-5 are as follows. Anal. Calcd for
C29H38Fe3O2P2: C, 53.74; H, 5.91. Found: C, 53.41; H, 5.52. 1H
NMR: δ 4.61 (s, 5H, Cp), 4.11 (s, 10H, Cp), 3.11 (m, 2H, Cy),
2.25 (m, 2H, Cy), 2.08 (m, 1H, Cy), 1.92 (m, 2H, Cy), 1.81 (m,
2H, Cy), 1.78-1.66 (m, 5H, Cy), 1.56 (m, 1H, Cy), 1.29-1.15
(m, 7H, Cy), 0.49 (dd, JHP = 253, JHH = 4, 1H, P-H). 13C{1H}
NMR (CDCl3): δ 281.2 (d, JCP = 14, μ-CO), 85.7 (s, Cp), 81.2 (s,
2Cp), 54.3 (d, JCP = 25, C1(Cy)], 37.9 (d, JCP = 12, C1(Cy)),
Preparation of [Fe2Cp2(μ-PCy)(μ-CO)(CO)2] (3a). Solid
KOH (ca. 0.1 g, excess) was added to a dichloromethane
solution (6 mL) of compound 2a (0.050 g, 0.095 mmol) at room
temperature, and the mixture was vigorously stirred for 5 min to
give a red-brown solution which was filtered using a cannula.
The solvent was then removed under vacuum from the filtrate,
and the residue was washed with petroleum ether (2 ꢀ 5 mL) to
give compound 3a as a red-brown, air-sensitive microcrystalline
solid (0.040 g, 95%). Anal. Calcd for C19H21Fe2O3P: C, 51.87;
H, 4.81. Found: C, 51.38; H, 4.47. 1H NMR (C6D6): δ 4.24 (s,
10H, Cp), 2.24-1.47 (m, 11H, Cy).
Preparation of [Fe2Cp2(μ-PPh)(μ-CO)(CO)2] (3b). The pro-
cedure is completely analogous to that described for 3a. With
0.050 g (0.096 mmol) of compound 2b as starting material and
using ca. 0.1 g of KOH (excess), compound 3b was obtained as a
red-brown, air-sensitive microcrystalline solid (0.039 g, 93%).
Anal. Calcd for C19H15Fe2O3P: C, 52.58; H, 3.48. Found: C,
35.4, 32.4 (2s, C2(Cy)), 28.5 (d, JCP = 11, C3(Cy)), 27.7 (d, JCP
=
10, C3(Cy)), 27.1, 26.9 (2s, C4(Cy)).
Preparation of [Fe3Cp3(μ-PMes)(μ-PMesH)(μ-CO)2(CO)]
(6). A solution of 3c (0.080 g, 0.168 mmol) in toluene (4 mL)
was refluxed for 6.5 h to give a dark green-brown solution. After
removal of the solvent under vacuum, the residue was extracted
with dichloromethane/petroleum ether (1/2) and the extracts
were chromatographed on alumina (activity IV) at 288 K.
Elution with the same solvent mixture gave a yellow frac-
tion containing [FeCp2], a red fraction containing some

1114 Organometallics, Vol. 30, No. 5, 2011
Alvarez et al.
Table 5. Crystal Data for New Compounds
4 2CH2Cl2
3
6
mol formula
mol wt
cryst syst
C
37H46Cl4Fe4O3P2
965.88
orthorhombic
Pnma
0.710 73
18.6619(7)
19.2134(6)
10.4001(3)
90
90
90
3729.0(2)
4
1.720
1.935
100(2)
2.12-26.38
C36H38Fe3O3P2
748.15
monoclinic
C2/c
0.710 73
23.646(2)
9.1154(9)
33.408(3)
90
104.087(5)
90
6984.2(11)
8
1.423
1.355
100(2)
1.78, 25.35
space group
˚
radiation (λ, A)
˚
a, A
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
3
˚
V, A
Z
calcd density, g cm-3
abs coeff, mm-1
temp, K
θ range, deg
index ranges (h, k, l)
no. of rflns collected
no. of indep rflns (Rint
no. of rflns with I > 2σ(I)
R indexes (data with I > 2σ(I))a
R indexes (all data)a
GOF
0-23; 0-24; 0-12
48 888
-28 to þ27, 0-10, 0-40
47 239
6380 (0.1392)
3763
)
3928 (0.0936)
2758
R1 = 0.0374, wR2 = 0.0613b
R1 = 0.0593, wR2 = 0.1282c
R1 = 0.1176, wR2 = 0.146c
0.973
R1 = 0.0744, wR2 = 0.0715b
1.003
0/225
no. of restraints/params
˚
1/427
0.736, -0.744
-3
ΔF(max, min), e A
0.878, -0.633
P
P
P
P
a R1 = ||Fo| - |Fc||/ |Fo|. wR2 = [ w(|Fo|2 - |Fc|2)2/ w|Fo|2]1/2. w = 1/[σ2(Fo2) þ (aP)2 þ bP] where P = (Fo þ 2Fc2)/3. b a = 0.0206, b =
2
6.7275. c a = 0.0732, b = 0.0000.
[Fe2Cp2(CO)4], and a third green fraction. Removal of the
solvents from the latter fraction under vacuum gave the green
complex 6 (0.050 g, 40%) as a mixture of syn and anti isomers.
The crystals of 6 (syn isomer) used in the diffraction study were
grown by the slow diffusion of a layer of petroleum ether into a
toluene solution of the complex at 253 K. Anal. Calcd for
C36H38Fe3O3P2: C, 57.79; H, 5.12. Found: C, 57.60; H, 5.10.
o-tBu), 1.20 (s, 9H, p-tBu), 1.00 (s, 3H, Me). Spectroscopic data
for cis-7 are as follows. 1H NMR (C6D6): δ 4.51, 4.20 (2s, 2 ꢀ 5H,
Cp). Other resonances of this isomer were obscured by those of
the major isomer. Trans/cis ratio: ca. 5/1.
Photolysis of Compound 3d. A toluene solution (8 mL) of
compound 3d (0.100 g, 0.166 mmol) was irradiated with visi-
ble-UV light in a quartz Schlenk tube for 1 h while N2 was
gently bubbled through the solution, to give a green-brown
mixture. After removal of the solvent under vacuum, the residue
was extracted with dichloromethane/petroleum ether (1/15) and
the extracts were chromatographed on alumina (activity IV) at
288 K. Elution with the same solvent mixture gave two orange
fractions yielding respectively, after removal of solvents
under vacuum, the complexes trans-[Fe2Cp2(μ-H){μ-P(CH2-
1
Spectroscopic data for syn-6 are as follows. H NMR: δ 6.98,
6.92, 6.90, 6.81 (4s, 4 ꢀ 1H, C6H2), 6.44 (d, JHP = 306, 1H,
P-H), 4.46 (d, JHP = 1, 5H, Cp), 4.43 (s, 5H, Cp), 3.72 (d, JHP
= 1, 5H, Cp), 2.64, 2.38, 2.37, 2.26, 2.18, 2.14 (6s, 6 ꢀ 3H, Me).
13C{1H} NMR (CDCl3): δ 281.5 (dd, JCP = 7, 11, μ-CO), 279.5
(d, JCP = 10, μ-CO), 213.6 (dd, JCP = 28, 19, FeCO), 162.1 (dd,
JCP = 28, 5, C1(Mes)), 142.5 (d, JCP = 15, C1(Mes)), 141.2,
138.8, 137.5, 137.0, 133.5, 132.4 (6s, C2,4(Mes)), 130.1 (d, JCP
=
CMe2)C6H2 Bu2}(CO)2] (trans-8; 0.019 g, 20%) and cis-[Fe2Cp2-
t
t
8, C3(Mes)), 129.1 (d, JCP = 4, C3(Mes)), 128.9 (s, C3(Mes)),
128.8 (s, C3(Mes)), 90.0, 88.9, 86.4 (3s, Cp), 24.6, 24.5, 21.9, 21.7,
21.2, 21.0 (6s, Me). Spectroscopic data for anti-6 are as follows.
1H NMR: δ 4.39 (s, 5H, Cp), 4.36, 3.82 (2d, JHP = 1, 2 ꢀ 5H,
Cp). Other resonances for anti-6 in the 1H and 13C{1H} NMR
spectra could not be assigned, as they were masked by those of
the major isomer. Syn/anti ratio: ca. 10/1.
(μ-H){μ-P(CH2CMe2)C6H2 Bu2}(CO)2] (cis-8; 0.012 g, 13%) as
orange microcrystalline powders. Finally, elution with dichlor-
omethane/petroleum ether (1/1) gave a green fraction analo-
gously yielding compound 7 (0.060 g, 60%). Anal. Calcd for
C30H39Fe2O2P (trans-8): C, 62.74; H, 6.84. Found: C, 62.60; H,
6.83. Anal. Calcd for C30H39Fe2O2P (cis-8): C, 62.74; H, 6.84.
Found: C, 62.52; H, 6.61. Spectroscopic data for trans-8 are as
follows. 1H NMR (C6D6): δ 7.55 (dd, JHH = 2, JHP = 4, 1H,
C6H2), 7.35 (d, JHH = 2, 1H, C6H2), 4.36, 4.24 (2s, 2 ꢀ 5H, Cp),
Preparation of [Fe2Cp2(μ-CO)2(CO){PH(CH2CMe2)C6H2
tBu2}] (7). A toluene solution (5 mL) of compound 3d (0.050 g,
0.083 mmol) was placed in a Schlenk tube equipped with a Young
valve, and it was stirred at 233 K for 1 h to give a brown-green
mixture. The solvent was then removed under vacuum, and the
residue was dissolved in a minimum of dichloromethane/petro-
leum ether (1/6) and chromatographed on alumina (activity IV).
Elution with dichloromethane/petroleum ether (1/4) gave a red
fraction containing the dimer [Fe2Cp2(CO)4]. Elution with di-
chloromethane/petroleum ether (1/1) gave a green fraction yield-
ing, after removal of solvents, compound 7 as a dark green
microcrystalline solid (0.036 g, 71%). This complex exists in
solution as an equilibrium mixture of the corresponding trans
and cisisomers. Anal. Calcd for C31H39Fe2O3P: C, 61.82; H, 6.53.
Found: C, 61.67; H, 6.51. Spectroscopic data for trans-7 are as
follows. 1H NMR (C6D6): δ 7.48, 7.06 (2 m, 2 ꢀ 1H, C6H2), 4.56
(dd, JHP = 348, JHH = 5, 1H, P-H), 4.47, 4.09 (2s, 2 ꢀ 5H, Cp),
2.53, 1.99 (2 m, 2 ꢀ 1H, CH2), 1.66 (s, 3H, Me), 1.64 (s, 9H,
2.60 (t, JHH = JHP = 13, 1H, CH2), 2.03 (dd, JHH = 13, JHP
=
5, 1H, CH2), 1.88 (s, 3H, Me), 1.58 (s, 9H, o-tBu), 1.40 (s, 3H,
Me), 1.35 (s, 9H, p-tBu), -18.30 (d, JHP = 39, 1H, μ-H).
Spectroscopic data for cis-8 are as follows. H NMR (C6D6):
1
δ 7.51 (dd, JHH = 2, JHP = 4, 1H, C6H2), 7.27 (d, JHH = 2, 1H,
C6H2), 4.29 (s, 10H, Cp), 2.43 (d, JHP = 5, 2H, CH2), 1.53 (s, 6H,
=
Me), 1.44 (s, 9H, o-tBu), 1.32 (s, 9H, p-tBu), -18.21 (d, JHP
39, 1H, μ-H).
X-ray Structure Determination for Compound 4. The X-ray
intensity data for 4 2CH2Cl2 were collected on a Kappa-Apex-
3
II Bruker diffractometer using graphite-monochromated Mo
KR radiation at 100 K. The software APEX44 was used for
collecting frames with the ω/ψ scan measurement method. The
(44) APEX 2, version 2.0-1; Bruker AXS Inc., Madison, WI, 2005.

Article
Organometallics, Vol. 30, No. 5, 2011 1115
Bruker SAINT45 software was used for the data reduction, and a
multiscan absorption correction was applied with SADABS.46
Using the program suite WinGX,47 the structure was solved by
direct methods with SIR9248 and refined with full-matrix least
squares on F2 with SHELXL97.49 During the solution process,
the compound was found to be placed on the symmetry opera-
tion x, -y þ 3/2, z and to crystallize with two molecules of
dichloromethane. All the positional parameters and the aniso-
tropic temperature factors of all the non-H atoms were refined
anisotropically, except for the carbon atoms C(14), C(21),
C(31), and C(33), which were refined isotropically to prevent
their temperature factors from becoming nonpositive definite.
All hydrogen atoms were geometrically placed and refined using
a riding model. Crystallographic data and structure refinement
could be modeled satisfactorily by introducing occupation
factors of 0.5 in all atoms of both parts. During the final stages
of the refinement, all the positional parameters and the aniso-
tropic temperature factors of all the non-H atoms were refined
anisotropically. All hydrogen atoms were geometrically placed
and refined using a riding model, except for the hydrogen atoms
H(1) and H(2), which were located in the Fourier map and
refined isotropically; nevertheless, a restriction on the P-
˚
(2B)-H(2) distance (1.2 A) had to be added for a satisfactory
refinement. As the solvent molecule present in the asymmetric
unit could not be conveniently modeled, the SQUEEZE50
procedure, as implemented in PLATON,51 was used. Upon
SQUEEZE application and convergence the strongest residual
peak (0.74 e A-3) was located near the Fe(1) atom. Crystal-
˚
details for 4 2CH2Cl2 are collected in Table 5.
lographic data and structure refinement details for 6 are col-
lected in Table 5.
3
X-ray Structure Determination for Compound 6. The acquisi-
tion of data, absorption corrections, and data reduction for 6
were performed as described for 4. Using the program suite
WinGX,47 the structure was solved by Patterson interpretation
and phase expansion and refined with full-matrix least squares
on F2 using SHELXL97.49 During the solution stages the
compound was found to crystallize with a molecule of solvent,
possibly n-hexane, highly disordered and placed on a symmetry
element. The refinement process made evident that both phos-
phorus atoms were disordered in two positions, but this disorder
Acknowledgment. We thank the DGI of Spain
(Projects CTQ2006-01207 and CTQ2009-09444) and the
COST action CM0802 ‘‘PhoSciNet’’ for supporting this
work and the MEC of Spain for a grant (to R.G.). Dr.
Celedonio M. Alvarez is also acknowledged for the initial
experiments concerning the synthesis of compounds 3a,b.
Supporting Information Available: A CIF file giving crystal-
lographic data for the structural analysis of compounds 4 and 6.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(45) SMART & SAINT Software Reference Manuals, Version 5.051
(Windows NT Version); Bruker Analytical X-ray Instruments: Madison,
WI, 1998.
(46) Sheldrick, G. M. SADABS, Program for Empirical Absorption
€
€
Correction; University of Gottingen, Gottingen, Germany, 1996.
(47) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
(48) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Gualardi, A.
J. Appl. Crystallogr. 1993, 26, 343.
(50) Van der Sluis, P.; Spek, A. L. Acta Crystallogr., Sect. A 1990, 46,
194.
(51) Spek, A. L. PLATON, A Multipurpose Crystallographic Tool;
Utrecht University, Utrecht, The Netherlands, 2010.
(49) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112.