Carbamoyl (Carboxamido) Complexes as Precursors for Metallaoxetene, Isonitrile, and Aminomethylidyne Complexes of Iron

Article abstract of DOI:10.1021/om990385y

The reaction of [Fe{C(=O)NⁱPr₂}(CO)₄]Li with (CF₃CO)₂O in diethyl ether followed by triphenylphosphine provides carbamoyl complexes [Fe{η²-OCNⁱPr₂}(CF₃)(CO) ₂(PPh₃)] (1) or [Fe(η²-OCNⁱPr₂)(CO)₂(PPh ₃)₂](O₂CCF₃) (2a·O₂CCF₃) depending on the reaction conditions. The same procedure performed in dichloromethane provides the aminomethylidyne complex [Fe(≡CNⁱPr₂)(CO)₃(PPh ₃)](O₂CCF₃) (5·O₂CCF₃) as well as traces of [Fe(η²-OCNⁱPr₂)(CO)₂-(PPh ₃)₂](O₂CCF₃) (2a·O₂CCF₃). The reactions of 1 with moist AgBF₄ or HBF₄ provide [Fe-{η²-OCNⁱPr₂}(CO)₃(PPh ₃)]BF₄ (2b·BF₄), via a putative difluorocarbene intermediate, [Fe{η²-OCNⁱPr₂}(=CF₂)(CO) ₂(PPh₃)]BF₄. In a similar manner, [Fe(η₂-OCNⁱPr₂)(CF₃)(CO)(dPPe)] (3) (obtained from 1 and dppe) reacts with HBF₄(aq) to provide [Fe{η²-OCNⁱPr₂}(CO) ₂(dppe)]-BF₄ (4·BF₄). A difluorocarbene complex is also implicated in the formation of the ferraoxetene [Fe{κ²-C(NⁱPr₂)OCF ₂}(CO){HB(pz)₃}] (6) (pz = pyrazol-1-yl), which results from treatment of 1 with K[HB(pz)₃] under high dilution. Under more concentrated conditions [Fe{HB(pz)₃}₂] is obtained. The ferraoxetene undergoes an unusual acid hydrolysis with HBF₄(aq)/CO to provide the isonitrile salt [Fe(CNⁱPr)(CO)₂{HB(pz)₃}]PF₆ (7·PF₆). The salt [Fe(≡CNⁱPr₂)-(CO)₃(PPh₃)]I (5·I) is the product of the reaction of 1 with iodine, via an as yet obscure mechanism, which may however also involve ferraoxetene intermediates.

Full text of DOI:10.1021/om990385y

Organometallics 2000, 19, 15-21
15
Ca r ba m oyl (Ca r boxa m id o) Com p lexes a s P r ecu r sor s for
Meta lla oxeten e, Ison itr ile, a n d Am in om eth ylid yn e
Com p lexes of Ir on
Stephen Anderson, Anthony F. Hill,* and Ying Thong Ng
Centre for Chemical Synthesis, Department of Chemistry, Imperial College of Science,
Technology and Medicine, South Kensington, London SW7 2AY, U.K.
Received May 24, 1999
The reaction of [Fe{C(dO)NiPr2}(CO)4]Li with (CF3CO)2O in diethyl ether followed by
triphenylphosphine provides carbamoyl complexes [Fe{η2-OCNiPr2}(CF3)(CO)2(PPh3)] (1) or
[Fe(η2-OCNiPr2)(CO)2(PPh3)2](O2CCF3) (2a ‚O2CCF 3) depending on the reaction conditions.
The same procedure performed in dichloromethane provides the aminomethylidyne complex
[Fe(tCNiPr2)(CO)3(PPh3)](O2CCF3) (5‚O2CCF 3) as well as traces of [Fe(η2-OCNiPr2)(CO)2-
(PPh3)2](O2CCF3) (2a ‚O2CCF 3). The reactions of 1 with moist AgBF4 or HBF4 provide [Fe-
{η2-OCNiPr2}(CO)3(PPh3)]BF4 (2b‚BF 4), via a putative difluorocarbene intermediate, [Fe{η2-
OCNiPr2}(dCF2)(CO)2(PPh3)]BF4. In a similar manner, [Fe(η2-OCNiPr2)(CF3)(CO)(dppe)] (3)
(obtained from 1 and dppe) reacts with HBF4(aq) to provide [Fe{η2-OCNiPr2}(CO)2(dppe)]-
BF4 (4‚BF 4). A difluorocarbene complex is also implicated in the formation of the ferraoxetene
[Fe{κ2-C(NiPr2)OCF2}(CO){HB(pz)3}] (6) (pz ) pyrazol-1-yl), which results from treatment
of 1 with K[HB(pz)3] under high dilution. Under more concentrated conditions [Fe{HB(pz)3}2]
is obtained. The ferraoxetene undergoes an unusual acid hydrolysis with HBF4(aq)/CO to
provide the isonitrile salt [Fe(CNiPr)(CO)2{HB(pz)3}]PF6 (7‚P F 6). The salt [Fe(tCNiPr2)-
(CO)3(PPh3)]I (5‚I) is the product of the reaction of 1 with iodine, via an as yet obscure
mechanism, which may however also involve ferraoxetene intermediates.
Sch em e 1
In tr od u ction
We have previously described, in brief, the synthesis
of a range of bidentate carbamoyl (carboxamide) com-
plexes of iron which are co-ligated by halide, phosphine,
stannyl, and carbonyl ligands.1 These complexes are
readily obtained by the electrophilic halogenation or
stannylation of [Fe{C(dO)NiPr2}(CO)4]Li. The subse-
quent chemistry is dominated by ligand substitution
reactions, which leave the bidentate mode of coordina-
tion for the carbamoyl ligand ultimately unchanged
(Scheme 1). In this report we describe full details of the
preparation of the trifluoromethyl derivative [Fe(η2-
OCNiPr2)(CF3)(CO)2(PPh3)] (1) and show that the chem-
istry of this species is dominated by reactions that
intimately involve the carbamoyl and/or trifluoromethyl
ligands, such that the carbamoyl ligand is converted into
aminomethylidyne, ferraoxetene, and isonitrile groups.
Aspects of the results described herein have formed the
basis of preliminary reports.1,2
detail.3 Of particular relevance are the studies of
Semmelhack which show that HSAB considerations
may be useful in determining the preference of O- vs
Fe-alkylation in reactions with carbon-based electro-
philes.4 The reactivity of the corresponding carbamoy-
lates [Fe{C(dO)NR2}(CO)4]- has received less atten-
tion;5,6 however Fischer has reported the in situ
preparation of the iron carbamoylate salt [Fe{C(O)-
Resu lts a n d Discu ssion
The reactions of iron acylates [Fe{C(dO)R}(CO)4]-
with electrophiles have been studied in considerable
* To whom correspondence should be addressed. E-mail:
a.hill@ic.ac.uk.
(1) (a) Anderson, S.; Clark, G. R.: Hill, A. F. Organometallics 1992,
11, 1988. (b) Anderson, S.; Hill, A. F.; Slawin, A. M. Z.; White, A. J .
P.; Williams, D. J . Inorg. Chem. 1998, 37, 594.
(2) (a) Anderson, S.; Hill, A. F.; Slawin, A. M. Z.; Williams, J . D. J .
Chem. Soc., Chem. Commun. 1993, 266. (b) Anderson, S.; Hill, A. F.
Organometallics 1995, 14, 1562.
(3) Collman, J . P. Acc. Chem. Res. 1975, 8, 342.
(4) Semmelhack, M. F.; Tamura, R. J . Am. Chem. Soc. 1983, 105,
4099.
(5) Angelici, R. J . Acc. Chem. Res. 1972, 18, 335.
10.1021/om990385y CCC: $19.00 © 2000 American Chemical Society
Publication on Web 12/10/1999

16 Organometallics, Vol. 19, No. 1, 2000
Anderson et al.
Sch em e 2. Th e Ma yr Alk ylid yn e Syn th esis8 (R )
Sch em e 3
CH3, C6H5)
NiPr2}(CO)4]Li from the reaction of [Fe(CO)5] and
lithium diisopropylamide (hereafter LDA) and its reac-
tion with [Et3O]BF4 to provide a neutral carbene
complex [Fe{dC(OEt)NiPr2}(CO)4].7
Our initial studies sought to establish whether the
Mayr alkylidyne synthesis (Scheme 2)8 might be ex-
tended to group 8 metals with aminomethylidyne sub-
stituents. In Mayr’s synthetic approach, oxide is for-
mally abstracted from an anionic pentacarbonyl acyl or
aroyl carbonylmetallate of group 6 metals to produce
an alkylidyne complex. We have been able to success-
fully extend this synthesis to include aminomethyli-
dyne,9 ferrocenebis(methylidyne),10 cymatrenylmethyli-
dyne,11 and alkynylmethylidyne12 complexes of chro-
mium, molybdenum, and tungsten. Furthermore the use
of phosphine dihalides in place of (CF3CO)2O provides
convenient access to amino-, aryl-, and thienyl alkyli-
dynes of tungsten and molybdenum.13 It therefore
appeared reasonable and desirable to extend this chem-
istry to other metal centers and in particular later
transition metals. Comparatively fewer alkylidyne com-
plexes are known for these metals, and the synthetic
routes typically lack generality.14
trum consists of a poorly resolved quartet showing
3J (PF) coupling (27 Hz) also evident in the 19F NMR
spectrum, which shows a doublet, consistent with the
1
trans disposition of CF3 and PPh3 ligands. The H and
13C NMR data associated with the carbamoyl methyl
groups indicate that these are all chemically distinct.
The failure of this approach to provide the desired
aminomethylidyne complex may be interpreted with
reference to Scheme 3. Electrophilic attack at a metal
acyl complex may occur at either the acyl oxygen or the
metal center, and the factors that control such reactions
are still emerging. Those relevant to the present group
6/group 8 dichotomy presumably include 6 vs 5 metal
coordination number, d6 vs d8 electron configuration,
HSAB character of the electrophile, and possible steric
bulk of the acyl substituent. The HSAB aspects of iron
acylate alkylation have been thoroughly investigated by
Semmelhack.4 The conclusions of the study, i.e., that
strongly solvating solvents (HMPA, Et2O) and hard
electrophiles (R-O3SF, R-O3SC6H4Me-4) favor O-alkyl-
ation over Fe-alkylation, were shown to be reminiscent
of those for isolobal enolate anions.15 The remaining
factors pertain to the steric accessibility and nucleophi-
licity of the metal centers involved.
Treating a diethyl ether solution of [Fe{C(dO)NiPr2}-
(CO)4]Li (prepared in situ) with trifluoroacetic anhy-
dride followed by triphenylphosphine does not in fact
lead to the anticipated aminomethylidyne complex [Fe-
(tCNiPr2)(O2CCF3)(CO)2(PPh3)] but rather a trifluo-
romethyl(carbamoyl) derivative formulated as [Fe{η2-
OCNiPr2}(CF3)(CO)2(PPh3)] (1) on the basis of spectro-
scopic and analytical data and a single-crystal X-ray
diffraction study.1a The bidentate carbamoyl ligand
gives rise to a characteristic infrared absorption at 1610
cm-1 (CH2Cl2) in addition to two terminal carbonyl
associated absorptions at 2027, 1953 cm-1, which indi-
cate a cis-dicarbonyl arrangement. The 31P NMR spec-
(6) Weller, F.; Petz, W. Z. Anorg. Allg. Chem. 1994, 620, 343. Petz,
W. J . Organomet. Chem. 1993, 456, 85. Boese, R.; Blaser, D.; Petz, W.
Z. Naturforsch., B 1988, 43, 945. Petz, W. Z. Naturforsch., B 1981,
36b, 335.
(7) Fischer, E. O.; Schneider, J .; Ackerman, K. Z. Naturforsch., B
1984, 39, 468.
HSAB considerations usefully rationalize the regio-
selectivity of aryl and aroyl ferrate alkylation allowing
the optimization of yields of alkoxy carbene complexes.
Unfortunately, employing a range of solvent combina-
tions with the carbamoyl ferrate analogue (thf, HMPA,
and Et2O) including the optimum conditions devised by
Semmelhack (ether/HMPA, 1:4) led only to mixtures of
1 and [Fe(CO)3(PPh3)2]. The formation in many of these
(8) Mayr, A.; McDermott, G. A. J . Am. Chem. Soc. 1986, 108, 458.
(9) Anderson, S.; Hill, A. F. J . Organomet. Chem. 1990, 394, C24.
(10) Davies, S. J .; Hill, A. F.; Pilotti, M. U.; Stone, F. G. A.
Polyhedron 1989, 8, 2265.
(11) Anderson, S.; Hill, A. F. J . Chem. Soc., Dalton Trans. 1993,
587.
(12) Hart, I. J .; Hill, A. F.; Stone, F. G. A. J . Chem. Soc., Dalton
Trans. 1989, 2261
(13) Anderson, S.; Cook, D. J .; Hill, A. F. J . Organomet. Chem. 1993,
463, C3.
(14) Gallop, M. A.; Roper, W. R. Adv. Organomet. Chem. 1986, 25,
121. For a review of group 8 alkylidyne complexes see: Hill, A. F. In
Comprehensive Organometallic Chemistry II; Eds. Abel, E. W., Stone,
F. G. A., Wilkinson, G., Eds.; Pergamon Press: Oxford, 1995; Vol. 7.
(15) House, H. O. Modern Synthetic Reactions, 2nd ed.; Benjamin:
New York; pp 522-529.

Carbamoyl (Carboxamido) Complexes
Organometallics, Vol. 19, No. 1, 2000 17
Sch em e 4. Acid Hyd r olysis of Tr iflu or om eth yl
Liga n d s (d p p e )
reactions of large amounts of [Fe(CO)3(PPh3)2] is per-
haps surprising given that reductive elimination reac-
tions involving trifluoroacyl and trifluoromethyl ligands
are generally not facile, especially in the latter case due
to the characteristically strong metal-carbon bond. The
complex 1 is very stable toward addition of excess
phosphine (vide infra), and so the point of mechanistic
divergence leading to [Fe(CO)3(PPh3)2] is presumably
not later than the initial product of electrophilic attack
at the metal by (CF3CO)2O. A further noteworthy aspect
of the sequence leading to 1 is the extremely facile
decarbonylation of the presumed trifluoroacyl interme-
diate (A, Scheme 2). This occurs at subambient tem-
peratures, while the closely related complex [Fe{C(d
O)CF3}2(CO)4] requires considerably more elevated
temperatures (refluxing heptane, 2 h).16 Indeed, it
appears that low-temperature decarbonylation of the
presumed intermediate [Fe{C(dO)CF3}{C(dO)NiPr2}-
(CO)n] (A, n ) 3, 4, ?) is crucial to the progress of the
reaction in that the addition of PPh3 immediately
following trifluoroacetylation (within 10-15 min, -78
°C) leads to almost exclusive formation of [Fe(η2-OCNi-
Pr2)(CO)2(PPh3)2](O2CCF3) (2a ‚O2CCF 3), while the for-
mation of 1 is maximized by a delay of over an hour (at
-78 °C). The presence of trifluoroacetate in the reaction
mixture may be a significant factor, given that acetate
salts have been used as catalysts for the decarbonylation
and ligand substitution reactions of group 8 metal
carbonyls.17
The apparently robust nature of the bidentate car-
bamoyl coordination is perhaps surprising given the
typically soft nature of low-valent iron centers, although
this feature has also emerged in our study of related
halide/carbamoyl complexes of divalent iron.1 Further-
more, related bidentate aroyl complexes of iron have
been described by Girolami,18 and these readily coordi-
nate CO under comparatively mild conditions with the
assumption of monodentate aroyl coordination. This
behavior is in direct contrast to the complexes [Ru(η2-
OCR)X(CO)(PPh3)2] (R ) Me, Ph, C6H4Me; X ) Cl, Br,
I),19 which do not readily carbonylate to form [Ru{σ-
C(dO)R}X(CO)2(PPh3)2]. In an attempt to induce mono-
dentate carbamoyl coordination 1 was treated with
dppe. A reaction slowly ensues in refluxing thf; however
the product is [Fe(η2-OCNiPr2)(CF3)(CO)(dppe)] (3) rather
than the anticipated [Fe{σ-C(dO)NiPr2}(CF3)(CO)2-
(dppe)]. Spectroscopic data for 3 suggest that the ster-
eochemistry adopted is most likely that indicated in
Scheme 4. In particular, the 31P NMR spectrum consists
of two multiplets [74.4, 71.1 ppm], indicating chemical
inequivalence. The 13C NMR resonance attributable to
the carbamoyl carbon appears as a double doublet (205.4
ppm), showing distinct couplings to cis (28 Hz) and trans
(50 Hz) coordinated phosphorus nuclei.
1,2-bis(d ip h en ylp h osp h in o)eth a n e)
complexes has been a key observation in the renaissance
of metal perfluoroalkyl chemistry.20 Thus for example
trifluoromethyl complexes of ruthenium have been used
extensively by Roper as precursors for dihalocarbene
complexes.20,21 Shriver has illustrated that fluoride
abstracting agents can also be employed with iron-CF3
complexes.22 Thus, for example, treating [Fe(CF3)(CO)-
(PPh3)(η-C5H5)] with BCl3 leads to the trichloromethyl
derivative [Fe(CCl3)(CO)(PPh3)(η-C5H5)] via a difluoro-
carbene intermediate, [Fe(dCF2)(CO)(PPh3)(η-C5H5)]+.
The fluoride substituents of 1 are also nucleofugic; for
example, treatment with aqueous HBF4 leads to the
tricarbonyl salt [Fe(η2-OCNiPr2)(CO)3(PPh3)]BF4 (2b‚
BF 4). The yields obtained are too high for the third
carbonyl ligand to have originated from decomposition
and CO scavenging (Scheme 3). The product is also
obtained using AgBF4 in place of HBF4, and it therefore
appears that the carbonyl ligand results from hydrolysis
of a putative difluorocarbene complex [Fe(η2-OCNiPr2)-
(dCF2)(CO)2(PPh3)]BF4. This possibility was also in-
vestigated using the complex 3, which is also hydrolyzed
by aqueous acid [HBF4(aq)] to provide [Fe(η2-OCNiPr2)-
(CO)2(dppe)]BF4 (4‚BF 4) presumably via the carbene
complex [Fe(η2-OCNiPr2)(dCF2)(CO)(dppe)]+. That this
reaction is facile is also indicated by the absence of a
molecular ion in the FAB-mass spectrum (protic nba
matrix) of 3 in favor of an intense peak corresponding
to 4+.
Acyl and carbamoyl ligands are typically nucleophilic
at the carbonyl oxygen, and difluorocarbene ligands are
strongly oxophilic electrophiles. It is therefore somewhat
surprising that the proposed intermediate carbamoyl-
difluorocarbene complexes do not rearrange with cou-
pling of the two mutually compatible sites of reactivity.
The question of electrophilic attack at monodentate
acyls is less well documented; however it appears
plausible that the bidentate coordination of the carbam-
oyl ligand in the present situation leads to a lowering
Rea ction s of th e Tr iflu or om eth yl Liga n d . The
nucleofugicity of fluoride substituents in metal-CF3
(16) Treichel, P. M.; Stone, F. G. A. Adv. Organomet. Chem. 1964,
1, 143. Morrison, J . A. Adv. Inorg. Radiochem. 1983, 27, 293. King, R.
B.; Bisnette, M. B. Acc. Chem. Res. 1970, 3, 417. King, R. B.; Bisnette,
M. B. J . Organomet. Chem. 1964, 2, 15. Hensley, D. W.; Wurster, W.
L.; Stewart, R. P., J r. Inorg. Chem. 1981, 20, 645.
(20) Brothers, P. J .; Roper, W. R. Chem. Rev. 1988, 88, 1293.
(21) Clark, G. R.; Hoskins, S. V.; Roper, W. R. J . Organomet. Chem.
1982 234, C9. Hoskins, S. V.; Pauptit, R. A.; Roper, W. R.; Waters, J .
M. J . Organomet. Chem. 1984, 269, C55.
(22) Richmond, T. G.; Shriver, D. F. Organometallics 1983, 2, 1061.
Richmond, T. G.; Crespi, A. M.; Shriver, D. F. Organometallics 1984,
3, 313.
(17) Lavigne, G. Eur. J . Inorg. Chem. 1999, 917.
(18) Girolami, G. S.; Hermes, A. R. Organometallics 1988, 7, 394.
(19) Roper, W. R.; Wright, L. J . J . Organomet. Chem. 1977, 142,
C1. Rickard, C. E. F.; Roper, W. R.; Taylor, G. E.; Waters, J . M.; Wright,
L. J . J . Organomet. Chem. 1990, 389, 375.

18 Organometallics, Vol. 19, No. 1, 2000
Anderson et al.
Ch a r t 1. F er r a oxeten e Con n ectivities
Am in om eth ylid yn e F or m a tion . The motivation for
the results described so far was the synthesis of alkyl-
idyne complexes of iron. Alkylidyne complexes of the
later transition metals remain rare, and the routes
applied sucessfully to date tend to be without precedent
in early transition metal chemistry.14 It was also noted
above that the “sensible” solvent combinations chosen
to favor O-trifluoroacetylation of the carbamoylate [Fe-
{σ-C(dO)NiPr2}(CO)4]Li failed comprehensively. We
were therefore somewhat surprised and gratified to find
that [Fe{σ-C(dO)NiPr2}(CO)4]Li (preformed in diethyl
ether and isolated) when treated with (CF3CO)2O and
PPh3 in dichloromethane led to two compounds, the
carbamoyl salt 2a ‚O2CCF 3 described above and the
aminomethylidyne salt [Fe(tCNiPr2)(CO)3(PPh3)](O2-
CCF3) (5‚O2CCF 3). The tetrachloroborate salt of this
complex has been described by Fischer24 and remains
the only isolable mononuclear alkylidyne complex of
iron. Fischer’s preparation of 5‚BCl4 follows his now
classical approach of alkoxide abstraction from an
alkoxy carbene complex [Fe{dC(OEt)NiPr2}(CO)3(PPh3)]
using BCl3. The present (“one-pot”) method offers con-
siderable practical advantage and economy, not with-
standing the unexpected and unusual solvent depen-
dence.
An alternative preparation of the complex 5+ was
subsequently discovered: Reaction of 1 with iodine
provides the iodide salt 5‚I via a most unusual and as
yet obscure mechanism. The yield (36%), while syn-
thetically useful, is not sufficient to exclude CO scav-
enging as the source of the third carbonyl ligand.
Nevertheless we are inclined to suspect that the CF3
ligand is intimately involved in the ultimate reduction
of the carbamoyl ligand to an aminomethylidyne, since
no 5‚I is obtained from the treatment of any of the
complexes [Fe(η2-OCNiPr2)X(CO)2(PPh3)] (X ) Br, I,
SnPh3), [Fe(η2-OCNiPr2)(SnPh3)(CO)(dppe)], [Fe(η2-
OCNiPr2)(SnPh3)(CO)(dppm)], 2a +, or 2b+ with iodine.
One possibility (suggested by the formation of a ferra-
oxetene discussed above) involves the formation of a
ferraoxetene of the form [Fe{κ2dC(NiPr2)OCF2}I(CO)2-
(PPh3)] (A, Scheme 6). Hydrolysis of the metallacyclic
difluoroalkyl group could provide a ferraoxetenone [Fe-
{κ2dC(NiPr2)OC(dO)}I(CO)2(PPh3)] (B, Scheme 6), which
extrudes CO2 to provide the complex [Fe(tCNiPr2)(CO)2-
(PPh3)]+. Capture of adventitious CO could then provide
the final product. The reluctance of iodide to occupy the
free coordination site is perhaps surprising; however it
should be noted that ruthenium alkylidyne complexes
[Ru(tCAr)Cl(CO)(PPh3)2] (Ar ) Ph, C6H4Me-4) are
readily carbonylated under 1 atm of CO to provide
cationic complexes [Ru(tCAr)(CO)2(PPh3)2]+, indicating
that the electron-rich metal center is better stabilized
by cation formation and π-acid ligation.14 In further
support of the ferraoxetenone intermediacy, we refer to
our recent isolation of a ruthenaoxetenone from the
addition of carbon dioxide to the alkylidyne complex
[Ru(tCPh)Cl(CO)(PPh3)2].25
Sch em e 5. F er r a oxeten e F or m a tion (Tp )
K3-HB(p yr a zol-1-yl)3; * ) p r op osed in ter m ed ia te)
of the nucleophilicity of the acyl oxygen atom, enabling
the ligand to act as a potent donor to the metal center.
Accordingly, the reactions of 1 with strongly chelating
ligands were investigated in an attempt to open up the
FeCO metallacycle: Treating 1 with NaC5H5‚dme (dme
) 1,2-dimethoxyethane) provided [Fe(CO)3(PPh3)2)] as
the only isolable organometallic product. The use of
potassium hydrotris(pyrazolyl)borate (K[HB(pz)3]; pz )
pyrazol-1-yl) however proved more informative. In re-
fluxing thf a smooth and high-yielding reaction ensued
under comparatively high dilution; however neither of
the anticipated products [Fe{η1-C(dO)NiPr2}(CO)(dCF2)-
{HB(pz)3}] nor K[Fe{η1-C(dO)NiPr2}(CO)(CF3){HB-
(pz)3}] were obtained. Rather, a novel ferraoxetene was
obtained, [Fe{κ2dC(NiPr2)OCF2}(CO){HB(pz)3}] (6). Un-
der more concentrated conditions [Fe{HB(pz)3}2] was
the only iron-containing product isolated. Spectroscopic
data did not unambiguously distinguish the possible
coupling arrangements of the carbamoyl and difluoro-
carbene components of the metallacycle (Chart 1),
although NOE data suggested a close proximity of one
isopropyl substituent to a pyrazol ring favoring con-
nectivity A. The connectivity (A) within the metallacycle
could however be confirmed by X-ray crystallography2a
as that shown in Scheme 4. The formation of the
ferraoxetene complex supports the plausibility of [Fe-
{η1-C(dO)NiPr2}(CO)(dCF2){HB(pz)3}] as an intermedi-
ate which under the thermal conditions of the reaction
undergoes intramolecular attack at the difluorocarbene
ligand by the monodentate (and accordingly nucleo-
philic) carbamoyl oxygen (Scheme 5). 1,3-Metallaoxe-
tenes remain a very rare class of complex, with the only
other examples being [Ru(κ2dCPhOCH2)(CA)2(PPh3)2]+
(A ) O, NC6H4Me-4), which result from the π-acid (CA)
induced coupling of benzoyl and methylene ligands in
the complex [Ru(η2-OCPh)Cl(dCH2)(PPh3)2],23 a reac-
tion with parallels to that proposed to account for the
formation of 6.
The possibility that hydrolysis of a difluoroferraoxe-
tene might be involved in the formation of 5‚I was
investigated by treating 6 with aqueous HPF6. The
product of this reaction is however the isonitrile salt
(23) Bohle, D. S.; Clark, G. R.; Rickard, C. E. F.; Roper, W. R.;
Wright, L. J . J . Organomet. Chem. 1988, 358, 411. Bohle, D. S.; Clark,
G. R.; Rickard, C. E. F.; Roper, W. R.; Shepard, W. E. B.; Wright, L. J .
J . Chem. Soc., Chem. Commun. 1987, 563.
(24) Fischer, E. O.; Schneider J .; Neugebauer, D. Angew. Chem., Int.
Ed. Engl. 1984, 23, 1820.

Carbamoyl (Carboxamido) Complexes
Organometallics, Vol. 19, No. 1, 2000 19
Sch em e 6. P r op osed Mech a n ism for th e F or m a tion of 5‚I by Ca r ba m oyl Iod in a tion
Sch em e 7. Mech a n istic P r op osa ls for th e F er r a oxeten e/Ison itr ile In ter con ver sion (Tp )
K3-HB(p yr a zol-1-yl)3; * ) p r op osed in ter m ed ia tes)
[Fe(CNiPr)(CO)2{HB(pz)3}]PF6 (7‚P F 6). Notably, rea-
Sch em e 8. Ison itr ile Con ver sion s
sonable yields (60%) are only obtained if the reaction is
carried out under an atmosphere of carbon monoxide.
The mechanism of isonitrile formation has not been
established; however those shown in Scheme 7 are at
least plausible. We have recently had occasion25 to
implicate the metallacycle [Ru{κ2dCPhSC(dS)}Cl(CO)-
(PPh3)2] in the formation of [Ru(η2-SCPh)Cl(CS)(PPh3)2]
from [Ru(tCPh)Cl(CO)(PPh3)2] and carbon disulfide.
The metallacycle [Fe{κ2dC(NiPr2)OC(dO)}(CO){HB-
(pz)3}] is therefore a reasonable intermediate. Protona-
tion of the endocyclic oxygen could then provide the
alkylidene complex [Fe{dC(NiPr2)OH}(CO)2{HB(pz)3}]+,
which subsequently eliminates 2-propanol to provide the
isonitrile ligand. While this transformation is unprec-
edented, the dehydration of primary carbamoyl ligands
to form isonitrile complexes has some precedent in late
transition metal chemistry5 and represents the reverse
of isonitrile hydration as a means of carbamoyl synthe-
sis (Scheme 8). An alternative route for the conversion
of “[Fe{C(NiPr2)OC(dO)}(CO){HB(pz)3}]” to 7‚P F6 might
involve loss of CO2 from the metallacycle to provide the
formally zerovalent alkylidyne complex [Fe(tCNiPr2)-
(CO){HB(pz)3}]. This complex would be prone to acid
oxidation to provide [Fe(tCNiPr2)(L)(CO){HB(pz)3}]2+
(L ) OH2, CO), which due to the high positive charge
might be expected to undergo nucleophilic (HO-, F-?)
dealkylation of the aminomethylidyne ligand. This last
step is the reverse of isonitrile alkylation as a route to
aminomethylidyne complexes, a process only favored for
very electron-rich precursors.26 This latter mechanism
is perhaps more attractive given the yield enhancement
resulting from an external source of CO.
Con clu d in g Rem a r k s. The formation of stable car-
bamoyl complexes of iron is surprisingly facile if the
ligand can be stabilized by bidentate coordination. Once
this has occurred, the bidentate mode of coordination
appears to be favored and ultimately preserved in many
reactions. The role of the carbamoyl ligand is however
not exclusively spectatorial, as illustrated by its conver-
sion to metallaoxetene, aminomethylidyne, and isoni-
trile groups. The mechanisms of these transformations
(25) Bedford, R.; Hill, A. F.; White, A. J . P.; Williams, D. J . Angew.
Chem., Int. Ed. Engl. 1996, 35, 95.
(26) Mayr, A.; Hoffmeister. H. Adv. Organomet. Chem. 1991, 32, 227.
Angelici, R. J .; Kim, H. S. Adv. Organomet. Chem. 1987, 27, 51.

20 Organometallics, Vol. 19, No. 1, 2000
Anderson et al.
and cooled (-30 °C) to provide yellow crystals of the product,
which were dried in vacuo. Yield: 7.60 g (58%). IR (CH2Cl2):
2039, 1968 ν(CO) 1605 cm-1 ν(NCO). (Nujol) 2026, 1960 ν(CO),
1612 cm-1 ν(NCO). NMR (CDCl3, 25 °C), 1H: δ 0.08, 1.03 [d ×
2, 12 H, CH3], 3.06, 5.16 [h × 2, 2 H, NCH], 7.33-7.53 ppm
[m, 30 H, C6H5]. 13C{1H}: 214.0 [t, FeCO, J (PC) ) 30.7 Hz],
211.3 [t, FeCO, J (PC) ) 21.5 Hz], 184.3 [t, OCN, J (PC) ) 23.0
Hz], 133.6-129.4 [C6H5], 56.1, 49.6 [NCH], 21.3, 19.2, ppm
[CH3]. 31P{1H}: 53.9 ppm. FAB-MS: m/z ) 764 [M]+, 708[M -
2CO]+, 580 [Fe(PPh3)2]+. These data are essentially identical
to those obtained for the crystallographically characterized salt
2a ‚P F 6 prepared via an alternative route.27
remain speculative and require further study. The
alkylidyne-forming reaction does however further il-
lustrate that in designing synthetic approaches to
alkylidyne complexes of the later transition metals, one
should not necessarily look to precedents from earlier
triads.
Exp er im en ta l Section
Gen er a l P r oced u r es. All manipulations were routinely
carried out under an atmosphere of prepurified dinitrogen
using conventional Schlenk-tube techniques. Solvents were
purified by distillation from an appropriate drying agent
[ethers and paraffins from sodium/potassium alloy with ben-
zophenone as indicator; halocarbons from CaH2]. All reagents
were commercially available (Aldrich) and used as received.
1H, 13C, and 31P{1H} NMR spectra were recorded on a J EOL
GNM EX270 NMR spectrometer and calibrated against inter-
nal Me4Si (1H), CDCl3 (13C), or external H3PO4 (31P). Infrared
spectra were recorded using a Perkin-Elmer 1720-X FT-IR
spectrometer. FAB mass spectrometry was carried out with
an Autospec Q mass spectrometer using nitrobenzyl alcohol
as matrix. For salts, “M” refers to the cationic complex. Light
petroleum refers to that fraction of bp 40-60 °C. Elemental
microanalyses were carried out by Medac Ltd, Middlesex, or
by the Imperial College Microanalytical Laboratory.
Syn th esis of [F e(η2-OCNiP r 2)(CO)3(P P h 3)]BF 4 (2b‚BF 4).
[Fe(η2-OCNiPr2)(CO)2(PPh3)(CF3)] (1; 1.00 g, 1.75 mmol) was
dissolved in diethyl ether (50 mL) and HBF4(aq) (0.12 mL, {wt/
mL ) 1.31 g}, 1.75 mmol) added. On stirring at room
temperature, a bright yellow precipitate eventually formed (2
h) from the ether liquor, which was then decanted off. The
product was then recrystalized from a mixture of dichlo-
romethane and light petroleum. Yield: 0.89 g (82%). IR (CH2-
Cl2): 2102, 2058, 2023 ν(CO) 1650 cm-1 ν(NCO). (Nujol) 2099,
2053, 2011 ν(CO), 1652 cm-1 ν(NCO). NMR (CDCl3, 25 °C),
1H: δ 0.55, 1.24, 1.36, 1.45 [d × 4 (br), 12 H, CH3], 3.64, 4.68
[h × 2, 2 H, NCH], 7.30, 7.53 ppm [m, 15 H, C6H5]. 13C{1H}:
δ 206.6 [d, FeCO, J (PC) ) 31.4 Hz], 202.2 [d, FeCO, J (PC) )
26.5 Hz], 197.1 [d, FeCO, J (PC) ) 51.0 Hz], 185.9 [d, OCN,
J (PC) ) 18.8 Hz], 133.2-127.5 [C6H5], 56.7, 50.3 [NCH], 21.8,
19.7, 19.4, 19.1 ppm [CH3]. 31P{1H}: 22.4 ppm. FAB-MS: m/z
) 530 [M]+, 502 [M′ - CO]+, 474 [M - 2CO]+, 446 [M - 3CO]+,
318 [FePPh3]+. Anal. Found: C, 54.1; H, 4.7; N. 2.2. Calcd for
Syn th esis of [F e(η2-OCNiP r 2)(CF 3)(CO)2(P P h 3)] (1). [Fe-
(CO)5] (3.00 g, 15 mmol) was placed in a Schlenk tube with
diethyl ether (50 mL), and then LDA was added dropwise (1.5
mol dm-3, 10.2 mL, 15 mmol). On completion of the addition,
the reaction mixture was cooled (dry ice/propanone) and
trifluoroacetic anhydride (2.4 mL, 17 mmol) diluted in ether
(20 mL), added dropwise. After stirring at low temperature
for 90 min, triphenylphosphine (6.00 g, 23 mmol) was added
and the reaction allowed to warm slowly to room temperature
overnight. A bright yellow precipitate formed, which was
isolated by decanting off the supernantant. Lithium trifluo-
roacetate was removed by extraction of the residue with a
mixture of dichloromethane and light petroleum (2:1) and
filtration of the combined extracts through a plug of diatoma-
ceous earth. On removing the solvent, the isolated product was
further purified by column chromatography. Yield: 4.92 g
(56%). The preparation may be carried out on a larger scale
(15.0 g [Fe(CO)5]) without compromising the yield. IR (CH2-
Cl2): 2027, 1953 ν(CO); 1610 ν(NCO) cm-1. (Nujol): 2020, 1949
ν(CO), 1606 ν(NCO) cm-1. NMR (CDCl3, 25 °C), 1H: δ 0.53,
1.04, 1.16, 1.29 [d × 4, 12 H, CH3, J (HH) ) 6.6 Hz], 3.28, 4.79
[h × 2, 2 H, NCH], 7.18-7.41 [m, 15 H, C6H5] ppm. 13C{1H}:
218.3 [d, FeCO, J (PC) not resolved], 211.6 [d, FeCO, J (PC)
not resolved], 197.1 [d, OCN, J (PC) not resolved], 133.6-128.5
[C6H5], 54.7, 48.5, [2 × s, NCH], 21.2, 20.5, 20.2, 19.9 [4 × s,
CH3] ppm. 31P{1H}: 49.6 ppm [q, J (PF) ) 27.2 Hz]. 19F: -7.15
ppm [d, J (PF) ) 29.3 Hz]. FAB-MS: m/z ) 543 [M - CO]+,
515 [M - 2CO]+, 465 [M - 2CO - CF3]+, 318 [FePPh3]+. Anal.
Found: C, 58.8; H, 5.2; N, 2.4. Calcd for C28H29F3FeNO3P: C,
58.9; H, 5.1; N, 2.5. The complex was also characterized by an
X-ray diffraction analysis.1
Syn th esis of [Fe(η2-OCNiP r 2)(CO)2(P P h 3)2](O2CCF3) (2a‚
O2CCF 3). A solution of [Fe(CO)5] (3.00 g, 2.0 mL, 15 mmol) in
diethyl ether (50 mL) was treated with a solution of LDA (10.2
mL, 1.50 mol dm-3, 15 mmol) and then cooled in a dry ice/
propanone bath. Trifluoroacetic anhydride (2.40 mL, 17 mmol)
was added dropwise, and the mixture stirred for 15 min.
Triphenylphosphine (6.00 g, 23 mmol) was then added, and
the mixture allowed to warm to room temperature overnight.
The small amount of precipitate that formed was removed by
filtration, and then all volatiles were removed from the filtrate
under reduced pressure. The residue was extracted with a
mixture of propanone and light petroleum (2:1), and the
combined extracts were concentrated under reduced pressure
C
28H29BF4FeNO4P: C, 54.5; H, 4.7; N, 2.3.
Syn th esis of [F e(η2-OCNiP r 2)(CO)(d p p e)(CF 3)] (3). [Fe-
{η2-OCNiPr2)(CO)2(PPh3)(CF3)] (1; 1.00 g, 1.75 mmol) and 1,2-
bis (diphenylphosphino)ethane (dppe, 0.70 g, 1.75 mmol) were
placed in a Schlenk tube under an atmosphere of nitrogen.
Tetrahydrofuran (30 mL) was added and the mixture heated
under reflux for 3 h. The thf was then removed under reduced
pressure, and diethyl ether (30 mL) added to the oily residue
to give a precipitate, which was washed with diethyl ether by
decantation. The residue was then chromatographed (silica gel,
CH2Cl2) and crystallized by addition of light petroleum,
concentration, and cooling (-30 °C). Yield: 0.82 g (68%). IR
(CH2Cl2): 1914 ν(CO) 1590, ν(NCO) cm-1. (Nujol): 1912 ν(CO),
1
1564 ν(NCO) cm-1. NMR (CDCl3, 25 °C), H: 0.51, 1.05, 1.24,
1.31 [d × 4, 12 H, CH3, J (HH) ) 6.6 Hz], 1.91, 2.31 [m × 2, br,
4 H, PCH2], 3.28, 5.04 [h × 2, 2 H, NCH], 6.74-8.07 [m, 20 H,
C6H5] ppm. 13C{1H}: 221.8 [d, FeCO, J (PC) ) 17.7 Hz], 205.4
[dd, OCN, J (PC) ) 50.0, 28.0 Hz], 137.9-127.6 [C6H5], 54.2,
47.1 [NCH], 30.3, 29.2 [m × 2, PCH2], 22.5, 21.0, 20.5, 20.1
[CH3]. 31P{1H}: 74.4, 71.1 [m × 2]. FAB-MS {X ) FeC(O)-
NiPr2(dppe)(CO)2}: m/z ) 638 [X]+, 610 [X - CO]+, 582[X -
2CO]+, 453 [Fedppe]+, 183 [FeOCNiPr2]+. Anal. Found: C,
57.5; H, 5.3; N, 1.8. Calcd for C35H38F3FeNO2P2.CH2Cl2: C,
56.6; H, 5.3; N, 1.8. Dichloromethane of solvation confirmed
1
by H NMR integration.
Syn th esis of [F e(tCNiP r 2)(CO)3(P P h 3)]O2CCF 3 (5‚O2-
CCF 3) a n d [F e(η2-OCNiP r 2)(CO)2(P P h 3)2]O2CCF 3 (2a ‚O2-
CCF 3). [Fe(CO)5] (3.00 g, 2.0 mL, 15 mmol) was diluted in
diethyl ether (50 mL) in a Schlenk tube and LDA (10.2 mL,
1.5 mol dm-3, 15 mmol) added. On completion of the addition,
the ether and any unreacted [Fe(CO)5] was removed under
reduced pressure and the residue redissolved in dichlo-
romethane (50 mL). This was then cooled (dry ice/acetone),
and a solution of trifluoroacetic anhydride (2.4 mL, 17 mmol)
in diethyl ether (20 mL) was added dropwise. After stirring
at low temperature for 15 min, triphenylphosphine (6.00 g,
23 mmol) was added and the reaction allowed to warm slowly
to room temperature, during which time an oily precipitate
(27) Anderson, S.; Hill, A. F.; Ng, Y. T.; White, A. J . P.; Williams,
D. J ., manuscript in preparation.

Carbamoyl (Carboxamido) Complexes
Organometallics, Vol. 19, No. 1, 2000 21
formed. Petrol (40-60) (25 mL) was added and the reaction
mixture filtered through a plug of diatomaceous earth. The
bright yellow precipitate that collected on the plug was then
extracted into tetrahydrofuran, the extracts were filtered, and
the solvent was removed from the filtrate under reduced
pressure to provide 5‚O2CCF 3. Yield: 1.20 g (12.5%). IR (CH2-
Cl2): 2083, 2034, 2012 ν(CO), 1664 cm-1 ν(CN). 2079, 2041,
1992 ν(CO), 1642 cm-1 ν(CN). NMR (CDCl3, 25 °C), 1H: δ 1.17
[d, 12 H, CH3, J (HH) ) 6.3 Hz], 3.96 [h, 2 H, NCH], 7.58-
7.76 ppm [m, 15 H, C6H5]. 13C{1H}: δ 266.5 [d, FetC, J (PC)
) 42.8 Hz], 206.5 [d, FeCO], 134.4-130.3 [C6H5], 58.5 [NCH],
22.0 ppm [CH3]. 31P{1H}: 58.4 ppm. FAB-MS: m/z ) 514 [M]+,
486 [M - CO]+, 458 [M - 2CO]+, 318 [FePPh3]+. These data
correspond to those for the crystallographically characterized
salt 5‚BCl4.24 The CH2Cl2/petrol liquor that initially passed
through the diatomaceous earth plug was cooled (-30 °C), and
eventually crystals formed. On decanting off the solvent and
drying the crystals under nitrogen, the second product 2a ‚
O2CCF 3 was isolated. Yield (not optimized): 0.24 g (2.5%).
Spectroscopic data were identical to those obtained above via
an alternative procedure.
completion, the thf was removed under reduced pressure and
the residue chromatographed on silica gel, eluting with a
mixture of dichloromethane and light petroleum (2:3). Yield:
0.61 g (73%). IR (CH2Cl2): 2485 ν(BH), 1959 ν(CO), 1557 ν-
(NCO) cm-1. (Nujol) 2488 ν(BH), 1950 ν(CO), 1563 ν(NCO)
1
cm-1. NMR (CDCl3, 25 °C), H: δ 0.74, 1.20 1.38, 1.43 [d × 4,
12 H, CH3], 3.42, 3.93 [h × 2, 2 H, NCH], 6.00, 6.03, 6.23 [t ×
3, 3 H, H4(pz)], 7.47, 7.51, 7.59, 7.64, 7.69, 7.92 [d × 6, 6 H,
H3,5(pz)] ppm. 13C{1H}: 225.0 [FeCO], 221.7 [q, OCN, J (FC)
) 4.7 Hz], 138.4 [dd, FeCF2, J (FC) ) 352, 377 Hz], 143.7, 143.5,
142.8 [C3(pz)], 135.0, 134.5 [C5(pz)], 105.3, 104.9 [C4(pz)], 54.7,
48.7 [NCH], 25.5, 21.3, 21.0, 20.3 [CH3] ppm. 19F: -33.5, -53.2
[AB, J (AB) ) 97.6 Hz]. FAB-MS: m/z ) 447 [M - CO]+, 397
[M - CO - CF2]+, 269 [FeHB(pz)3]+. Anal. Found: C, 45.4; H,
5.1; N, 20.9. Calcd for C18H24BF2FeN7O2: C, 45.5; H, 5.1; N,
20.6. The complex was also charaterized by an X-ray diffraction
analysis.2a
P r ep a r a tion of [F e(CNiP r )(CO)2{HB(p z)3}]P F 6 (7‚P F 6).
The complex [Fe(κ2dC(NiPr2)OCF2}(CO){HB(pz)3}] (6; 0.20 g,
0.42 mmol) was dissolved in diethyl ether (20 mL) and carbon
monoxide passed through the solution for 5 min. The mixture
was then treated with concentrated aqueous hexafluorophos-
phoric acid (0.12 g, 0.08 mL, 0.84 mmol). The mixture was
stirred for 1 h, during which time a pale yellow solid precipi-
tated, which was isolated by decantation, washed with diethyl
ether, and dried in vacuo. Yield: 0.10 g (60%). IR (CH2Cl2):
2222 ν(CN), 2104, 2068 cm-1 ν(CO) (Nujol) 2232 ν(CN), 2104,
2064 cm-1 ν(CO). NMR (CDCl3, 25 °C), 1H: δ 1.46 [d, 6 H,
CH3, J (HH) ) 6.3 Hz], 4.50 [h, 1 H, NCH], 6.40, 6.45 [t × 2, 3
H, H4(pz)], 7.80-7.96 [d × 4, 6 H, H3,5(pz)] ppm. FAB-MS: m/z
) 394 [M]+, 338 [M - 2CO]+, 269 [FeHB(pz)3]+. Anal. Found:
C, 33.7; H, 3.1; N, 18.0. Calcd for C15H17BF6FeN7O2P: C, 33.4;
H, 3.2; N, 18.2.
Syn th esis of [F e(tCNiP r 2)(CO)3(P P h 3)]I (5‚I). [Fe(η2-
OCNiPr2)(CO)2(PPh3)(CF3)] (1; 0.39 g, 0.68 mmol) and iodine
(0.17 g, 0.68 mmol) were placed in a Schlenk tube, which was
then evacuated and refilled with nitrogen. Diethyl ether (30
mL) was then added and the mixture stirred at room temper-
ature for 3 h. The residue which separated from the ether
liquor was isolated by decanting off the supernatant and
washed with diethyl ether. The residue was extracted with a
mixture of dichloromethane and light petroleum (2:1), and the
combined extracts were filtered through diatomaceous earth.
On concentrating the filtrate under reduced pressure, the
product crystallized out of the liquor and was isolated by
decantion. Yield: 0.16 g (36%). Spectroscopic data associated
with the cationic complex were identical to those for 5‚O2CCF 3
described above and 5‚BCl4.24
Ackn owledgm en t. We are grateful to the E.P.S.R.C.
(U.K.) for the award of a studentship (to S.A.). A.F.H.
gratefully acknowledges the Leverhulme Trust and the
Royal Society for the award of a Senior Research
Fellowship.
Syn t h esis of [F e{κ2-CF 2OCNiP r 2}(CO){H B(p z)3}] (6).
[Fe(η2-OCNiPr2)CF3(CO)2(PPh3)] (1; 1.00 g, 1.75 mmol) and
potassium hydrotris(pyrazol-1-yl)borate (0.44 g, 1.75 mmol)
were placed in a Schlenk tube, which was then evacuated and
refilled with nitrogen. Tetrahydrofuran (30 mL) was then
added and the mixture heated under reflux for 1 h. On
OM990385Y