Evidence for Iron-Catalyzed α-Phosphinidene Elimination with Phenylphosphine

Article abstract of DOI:10.1002/chem.201704954

The ubiquitous half-sandwich iron complex [CpFe(CO)₂Me] (Cp=η⁵-C₅H₅) appears to be a catalyst for α-phosphinidene elimination from primary phosphines. Dehydrocoupling reactions provided initial insight into this unusual reaction mechanism, and trapping reactions with organic substrates gave products consistent with an α elimination mechanism, including a rare example of a three-component reaction. The substrate scope of this reaction is consistent with generation of a triplet phosphinidene. In all, this study presents catalytic phosphinidene transfer to unsaturated organic substrates.

Full text of DOI:10.1002/chem.201704954

A Journal of
Accepted Article
Title: Evidence for Iron-Catalyzed α-Phosphinidene Elimination with
Phenylphosphine
Authors: Justin K Pagano, Brandon J Ackley, and Rory Waterman
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201704954
Link to VoR: http://dx.doi.org/10.1002/chem.201704954
Supported by

10.1002/chem.201704954
Chemistry - A European Journal
COMMUNICATION
Evidence for Iron-Catalyzed α-Phosphinidene Elimination with
Phenylphosphine
Justin K. Pagano,[a]‡ Brandon J. Ackley,[a] and Rory Waterman[a]*
Abstract: The ubiquitous half-sandwich iron complex CpFe(CO)2Me
(Cp = η5-C5H5) appears to be a catalyst for α-phosphinidene
elimination from primary phosphines. Dehydrocoupling reactions
provided initial insight into this unusual transformation, and trapping
reactions with organic substrates gave products consistent with an α
remain—most notably that the method is not catalytic. Our
objective has been to leverage hereto unknown -phosphinidene
elimination in the catalytic synthesis of organophosphines akin to
stoichiometric PR transfer. Layfield and coworkers provided a
critical advance in realizing this goal. In 2016, they reported the
catalytic dehydrogenation of primary phosphines with trapping as
phosphaalkenes in the presence of N-heterocyclic carbenes.[12]
Simple iron derivatives are attractive for this reaction. Ruiz
and coworkers have shown that Cp2Fe2(CO)3(PR) (R = Mes,
Mes*) and CpFe(CO)2(PHMes*) are susceptible to apparent
phosphinidene elimination under mild conditions; however, these
putative phosphinidene intermediates were not successfully
trapped.[13] With these results in mind, we hypothesized that
elimination mechanism, including
a rare example of a three
component reaction. The substrate scope of this reaction is consistent
with generation of a triplet phosphinidene. In all, this study presents
catalytic phosphinidene transfer to unsaturated organic substrates.
The catalytic formation of main-group bonds is of growing
importance in the targeted synthesis of value-added molecules
and materials.[1] Among the various elements, bond formation to
phosphorus remains an ongoing field of study due to its increasing
scarcity and importance in a variety of chemical disciplines.[1d, 1f, 2]
An acutely underdeveloped reaction in main group bond
formation is  elimination. Reported initially by Neale and Tilley in
2002, they identified that the dehydropolymerization of stannanes
CpFe(CO)2Me (1) would be
a potential catalyst for α-
phosphinidene elimination. Herein, we demonstrate that 1 is a
catalyst for this reaction and subsequent catalytic trapping of
these fragments with organic substrates is indeed possible. This
system, however, does not exhibit a broad substrate scope, which
we tentatively attribute to generation of a triplet phosphinidene.
Compound 1 is dehydrocoupling catalyst for primary and
secondary phosphines (See SI for full details). For example,
treatment of phenylphosphine with 10 mol % of 1 at 65 °C in a 1:9
benzene:Et2O solution gave hydrogen and primarily (PhPH)2 in
approximately 45% conversion as observed by 31P NMR
spectroscopy.[14] Additionally, the two unusual products for
dehydrocoupling catalysis, a triphosphine, (PhPH)2PPh, and
Ph2PH, were observed in 6% and 1% conversion, respectively, by
31P NMR spectroscopy.[1g][15] The reliable formation of
(PhPH)2PPh is unknown in catalytic dehydrocoupling reactions of
PhPH2,[1g] and Ph2PH was also observed in Layfield’s recent
report on catalytic phosphinidene transfer.[12] These two products,
the triphosphine and diphenylphosphine, hint at α-phosphinidene
elimination, which prompted attempts to trap such putative
fragments with unsaturated substrates.
occurred via the formal de-insertion of a stannylene fragment from
[3]
a group 4 stannyl intermediate.
The examples of bona fide 
elimination were extended to antimony[4] and arsenic[5] in
subsequent years, though a reasonable supposition for -silylene
elimination was reported by McIndoe and Rosenberg.[6] In all
instances, the products from these catalytic reactions were
homocoupled, either as polymers in the case of tin or as small
molecules for antimony and arsenic. However, Neale and Tilley
observed the stoichiometric trapping of a stannylene fragment
during the thermal decomposition of CpCp*HfCl(SnMes2H) in the
presence of excess 2,3-dimethyl-1,3-butadiene, which gave 1,1-
dimesityl-3,4-dimethylhydrostannole.[3a]
Trapping of low valent fragments is of deep synthetic
importance. Returning to phosphorus, Mathey and coworkers
isolated metal-stabilized 7-phosphanorbornadienes,[7] which
initiated a decades-long adventure in organophosphine synthesis
via phosphinidene trapping.[8] Recently, Bertrand and coworkers
have shown stable singlet phosphinidenes to behave in an
electrophilic manner.[9] Indeed, it is well established that
phosphinidene fragments are readily trapped by dienes.[10]
Treatment of equimolar 2,3-dimethyl-1,3-butadiene and
phenylphosphine with 5 mol % of 1 in a 1:9 benzene:Et2O solution
at 65 °C for four hours gave secondary phosphine (2),
dihydrophosphole (3), and tetrahydrodiphosphorin (4) products as
a mixture in 18%, 53%, and 29% conversion, respectively
(Scheme 1).
Despite advances from each Lammerstma and Cummins in
11]
easier-to-prepare phosphinidene precursors,[10c,
limitations
[a]
Dr. Justin K. Pagano, Brandon J. Ackley, Prof. Dr. Rory Waterman
Department of Chemistry, University of Vermont, Burlington, VT
05405, USA
E-mail: rory.waterman@uvm.edu
Current Author Address: Chemistry Division, Los Alamos National
Laboratory, Mail Stop J-514, Los Alamos, NM 87545
Scheme 1. All products from the reaction of phenylphosphine and 2,3-dimethyl-
1,3-butadiene in the presence of 5 mol % of 1. Reaction conditions: PhPH2, 2
equiv. of 2,3-dimethyl-1,3-butadiene, and 0.05 equiv. of 1 at 65 °C for 4 h, but
higher temperatures disfavor formation of 2.
‡
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201704954
Chemistry - A European Journal
COMMUNICATION
Products 3 and 4 are suggestive of phosphinidene trapping.
However, the formation of a hydrophosphination product, 2, avails
the possibility that 3 results from ring closure of 2, as 1 is a known
hydrophosphination catalyst.[16] Monitoring the formation of these
products by 31P NMR spectroscopy does not reveal an apparent
dependence of the formation of 3 on the concentration of 2 (Figure
S-1). Additionally, results from two additional experiments
eliminate a ring-closure pathway (i.e., 2 converted to 3): 1)
Isolated samples of 2[2c] are not converted to 3 by 1 under catalytic
conditions with or without added hydrogen, and 2) using
phenylphosphine-d2 as a substrate yields only D2 and 2-d2 as
observed by 2H NMR spectroscopy, whereas a ring closure event
would result in formation of HD and partial deuteration of 3 and 4.
Given that formation of 2 is independent of the other products, two
pathways to form 4 appeared possible (Scheme 2). In pathway A,
4 is formed by the sequential trapping of phosphinidene
equivalents, the first to give 3 followed by 3 reacting with an
additional phosphinidene equivalent to give 4. In pathway B, 1,2-
diphenyldiphosphine, formed by the condensation of
phosphinidene fragments, undergoes a [4+2] cycloaddition with
2,3-dimethyl-1,3-butadiene to give 4 directly. Two observations
support pathway A over pathway B: 1) Monitoring these reactions
by 31P NMR spectroscopy reveals a decrease in the concentration
of 3 as the concentration of 4 increases (Figure S-1), and 2) when
additional PhPH2 is added to a typical catalytic reaction that has
reached completion (i.e., phosphine and diene substrates have
been consumed), the concentration of 3 decreases with an
increase in the concentration of 4.
pentaphenylphosphole
as
measured
by
31P
NMR
spectroscopy.[10b] Formation of phosphole in greater than 10%
were achieved by running the same reaction under 1 atm of
hydrogen. This observation, while ineffective for synthesis,
illustrates that there are tunable factors in developing this
catalysis for synthesis.
In contrast, 2-butyne proved to be a more successful
substrate. Reaction of PhPH2 with five equivalents of 2-butyne
and 5 mol % of 1 gave 2,3,4,5-tetramethyl-1-phenylphosphole in
53% conversion from starting PhPH2, as identified by 31P NMR
spectroscopy (Scheme 3).[18] Competitive hydrophosphination
was not observed under these reaction conditions. These alkyne
examples indicate that there are substantial steric and electronic
effects to be explored in this catalysis, and are rare examples of
three-component reactions involving phosphinidene chemistry.[19]
Scheme 3. Reaction of phenylphosphine and 2-butyne with 5 mol % of 1, which
gives the 2,3,4,5-tetramethyl-1-phenylphosphole as the primary product (53%
conversion by 31P NMR spectroscopy). Reaction conditions: PhPH2, 5 eq. 2-
butyne, 0.05 eq. 1, 65 °C, 2 h.
Diethyl disulfide also proved to be an effective trap under
these reaction conditions. Treatment of a 1:9 benzene:Et2O
solution of PhPH2 with equimolar (SEt)2 and 5 mol % of 1 gave
PhP(SEt)2 in quantitative conversion after one hour as evidenced
by 31P NMR spectroscopy (Scheme 4). The use of diethyl disulfide
as a trap for phosphinidenes is well established by Schmidt and
coworkers.[20]
Scheme 4. Reaction of phenylphosphine and diethyl disulfide with 5 mol % of
1. Reaction conditions: PhPH2, EtSSEt, 0.05 eq. 1, 65 °C, 1 h.
Further trapping reactions were largely unsuccessful. For
example, reactions run in the presence of or utilizing N-
methylimidazole as a solvent yielded no new phosphorus
containing products as observed by 31P NMR spectroscopy. This
result indicates that involvement of a phosphenium (R2P+)
fragment is not likely in this reaction. Furthermore, some
previously reported phosphinidene traps were not successful.
Simple ketones like acetone or benzophenone provide no
trapping.[20] The similarity of the substrate scope to reactions
reported by Schmidt is an initial indication that any phosphinidene
Scheme 2. Potential paths to tetrahydrodiphosphorin 4. Path A is the sequential
trapping of phosphinidene equivalents, and path B is the direct formation of 4
via a [4+2] cycloaddition with diphosphene. Observation of catalytic reactions
and sequential addition of PhPH2 support path A.
Formation of 3 and 4 do not appear to result from reaction
of (PPh)5 with 2,3-dimethyl-1,3-butadiene under these conditions,
which is a known process in the melt or under UV photolysis
conditions.[17] Indeed, reaction of (PPh)5 with five equivalents of
2,3-dimethyl-1,3-butadiene in the presence of 10 mol % of 1 in a
1:9-benzene:Et2O mixture at 65 °C gave only 11% conversion to
3 after 48 hours (i.e., >10x reaction time; see Supporting
Information for details). In control reactions without 1, formation of
the cyclophosphines (PPh)x (x = 4, 6) are observed after 48 hours
with no trapping or hydrophosphination products as determined
by 31P NMR spectroscopy.
With the success of trapping reactions with 2,3-dimethyl-
1,3-butadiene, more challenging substrates were attempted.
Treatment of PhPH2 with five equivalents of diphenylacetylene
and 5 mol % of 1 afforded double hydrophosphination[16] of the
alkyne in 42% conversion[2d] with trace amounts of 1,2,3,4,5-
intermediate is
a
triplet. The observation that singlet
phosphinidenes are not readily trapped by alkynes or 2,3-
dimethyl-1,3-butadiene further supports this supposition.[21] Thus,
these results are consistent with the catalytic formation of triplet
phosphinidenes as opposed to singlet phosphinidenes, and
provide a potential avenue to develop this kind of group transfer
catalysis.
Literature understanding of 1 and the observed reaction
chemistry provide the basis for a working mechanistic hypothesis
for this catalysis (Scheme 5). Activation of 1 would arise from
migratory insertion of the methyl ligand,[22] rather than protonation,
based on the relatively low acidity of these phosphine
substrates[23] and prior study of protonolysis[24] and migratory
insertion reactions at 1.[25] This proposal is buttressed by the
This article is protected by copyright. All rights reserved.

10.1002/chem.201704954
Chemistry - A European Journal
COMMUNICATION
persistence of 1 at the end of catalysis. For example, catalytic
reactions using 1-d3 result in nearly quantitative restoration of the
pre-catalyst by 2H NMR spectroscopy upon complete conversion
of the substrates. Though it is a negative result, CD3H was not
observed in reactions using 1-d3 as a catalyst. Following
migratory insertion, activation of phosphine substrate by P–H
oxidative addition would give an iron(IV) intermediate that would
engage in α-phosphinidene elimination, though the order of steps
(PR transfer vs. H2 evolution is unknown). It is also possible that
a terminal iron phosphinidene derivative is formed. However, the
transfer reactions herein are inconsistent with the PR-transfer
reactivity of known terminal phosphinidene compounds,[26] though
that area is sorely under-investigated. It is unclear that trapping
with organic substrates need or need not involve interaction with
the metal. Thus, we favor  elimination as a working hypothesis.
The observed dehydrocoupling products including those of P–H
insertion, P–P insertion, and condensation are consistent with
other α-elimination reactions.[3-4] Current evidence prevents more
than speculation at this point. Given the reactivity of :PPh,[17, 20]
the lack of apparent reaction with solvent suggests high
association with iron, consistent with theoretical analysis of α-
stannylene elimination.[27] Reductive elimination of H2 would close
the cycle and is also consistent with the observation of H2 in the
Acknowledgements
This work was supported by the U.S. National Science
Foundation (CHE-1265608 and CHE-1565658). We would also
like to thank Dr. Jaqueline Kiplinger (LANL) for providing
computational resources for preparing data and this manuscript.
Keywords: Phosphorus • Phosphinidene • Iron • Catalysis •
Trapping
References
[1]
[2]
[3]
a) A. Staubitz, A. P. M. Robertson, I. Manners, Chem. Rev. 2010, 110,
4079-4124; b) A. Staubitz, A. P. M. Robertson, M. E. Sloan, I. Manners,
Chem. Rev. 2010, 110, 4023-4078; c) J. I. Bates, J. Dugal-Tessier, D. P.
Gates, Dalton Trans. 2010, 39, 3151-3159; d) R. Waterman, Dalton
Trans. 2009, 18-26; e) R. Waterman, Organometallics 2013, 32, 7249-
7263; f) C. A. Bange, R. Waterman, Chem. Eur. J. 2016, 22, 12598-
12605; g) R. Waterman, Cur. Org. Chem. 2012, 16, 1313-1331.
a) A. J. Roering, S. E. Leshinski, S. M. Chan, T. Shalumova, S. N.
MacMillan, J. M. Tanski, R. Waterman, Organometallics 2010, 29, 2557-
2565; b) A. J. Roering, L. T. Elrod, J. K. Pagano, S. L. Guillot, S. M. Chan,
J. M. Tanski, R. Waterman, Dalton Trans. 2013, 42, 1159-1167; c) M. B.
Ghebreab, C. A. Bange, R. Waterman, J. Am. Chem. Soc. 2014, 136,
9240-9243; d) C. A. Bange, R. Waterman, ACS Catalysis 2016, 6, 6413-
6416.
dehydrocoupling and trapping reactions. Such
a cycle is
consistent with iron-based steps in Si–S coupling involving the
retention of 1 and H2 loss as proposed by Nakazawa.[28]
In conclusion, a rare example of what may catalytic α-
phosphinidene elimination mediated by the ubiquitous iron
complex 1 is reported. Phosphine dehydrocoupling reactions
provided initial insight into this interesting reaction type, and
catalytic trapping reactions were successful using 2,3-dimethyl-
1,3-butadiene, internal alkynes, and diethyldisulfide, which are
consistent with the reactivity expected for a putative triplet
phosphinidene. Deuterium labelling studies, product distributions,
and reaction studies further support the claim that this catalysis is
based largely on the α-elimination of a triplet phosphinidene.
Further work to expand the substrate scope of this reaction to a
broader set of reagents, and investigation of the phosphinidene
spin state are underway.
a) N. R. Neale, T. D. Tilley, J. Am. Chem. Soc. 2002, 124, 3802-3803; b)
N. R. Neale, T. D. Tilley, Tetrahedron 2004, 60, 7247-7260; c) N. R.
Neale, T. D. Tilley, J. Am. Chem. Soc. 2005, 127, 14745-14755.
a) R. Waterman, T. D. Tilley, Angew. Chem. Int. Ed 2006, 45, 2926-2929;
b) T. Pugh, N. F. Chilton, R. A. Layfield, Chem. Sci. 2017, 8, 2073-2080.
A. J. Roering, J. J. Davidson, S. N. MacMillan, J. M. Tanski, R. Waterman,
Dalton Trans. 2008, 4488-4498.
[4]
[5]
[6]
S. M. Jackson, D. M. Chisholm, J. S. McIndoe, L. Rosenberg, Eur. J.
Inorg. Chem. 2011, 2011, 327-330.
[7]
[8]
F. Mathey, Angew. Chem. Int. Ed. 1987, 26, 275-286.
a) C. Compain, B. Donnadieu, F. Mathey, Organometallics 2005, 24,
1762-1765; b) F. Mathey, Dalton Trans. 2007, 1861-1868; c) X. Zhao, Z.
Lu, Q. Wang, D. Wei, Y. Lu, Z. Duan, F. Mathey, Organometallics 2016;
d) Z. Benko, R. Streubel, L. Nyulaszi, Dalton Trans. 2006, 4321-4327.
a) L. Liu, David A. Ruiz, D. Munz, G. Bertrand, Chem 2016, 1, 147-153;
b) M. M. Hansmann, R. Jazzar, G. Bertrand, J. Am. Chem. Soc. 2016,
138, 8356-8359.
[9]
[10] a) F. Mathey, Chem. Rev. 1988, 88, 429-453; b) F. Mathey, Sci. Syn.
2002, 9, 553-600; c) M. J. van Eis, H. Zappey, F. J. J. de Kanter, W. H.
de Wolf, K. Lammertsma, F. Bickelhaupt, J. Am. Chem. Soc. 2000, 122,
3386-3390; d) K. Lammertsma, in New Aspects in Phosphorus
Chemistry III, Vol. 229 (Ed.: J.-P. Majoral), Springer Berlin Heidelberg,
2003, pp. 95-119; e) M. L. G. Borst, R. E. Bulo, D. J. Gibney, Y. Alem, F.
J. J. de Kanter, A. W. Ehlers, M. Schakel, M. Lutz, A. L. Spek, K.
Lammertsma, J. Am. Chem. Soc. 2005, 127, 16985-16999.
[11] A. Velian, C. C. Cummins, J. Am. Chem. Soc. 2012, 134, 13978-13981.
[12] K. Pal, O. B. Hemming, B. M. Day, T. Pugh, D. J. Evans, R. A. Layfield,
Angew. Chem., Int. Ed. 2016, 55, 1690-1693.
[13] a) M. A. Alvarez, M. E. García, M. A. Ruiz, J. Suárez, Angew. Chem., Int.
Ed. 2011, 50, 6383-6387; b) M. A. Alvarez, M. E. García, D. García-Vivó,
A. Ramos, M. A. Ruiz, Inorg. Chem. 2012, 51, 3698-3706.
[14] R. Waterman, Organometallics 2007, 26, 2492-2494.
[15] The results of other dehydrocoupling reactions can be found in the
supplementary information.
Scheme 5. Potential catalytic cycle for the liberation of phosphinidene
fragments by 1 based on prior examples of E–H activation by 1 and related
derivatives. The order of phosphinidene elimination versus H2 loss is not known
and may be reversed. Initial H2 loss may promote formation of an iron-
phosphinidene intermediate that available data can neither prove nor refute.
[16] a) M. Kamitani, M. Itazaki, C. Tamiya, H. Nakazawa, J. Am. Chem. Soc.
2012, 134, 11932-11935; b) M. Itazaki, S. Katsube, M. Kamitani, H.
Nakazawa, Chem. Commun. 2016, 52, 3163-3166.
[17] U. Schmidt, I. Boie, Angew. Chem. Int. Ed. 1966, 5, 1038.
[18] P. J. Fagan, W. A. Nugent, Org. Syn. 1992, 70.
This article is protected by copyright. All rights reserved.

10.1002/chem.201704954
Chemistry - A European Journal
COMMUNICATION
[19] F. Knoch, F. Kremer, U. Schmidt, U. Zenneck, P. Le Floch, F. Mathey,
Organometallics 1996, 15, 2713-2719.
[20] U. Schmidt, C. Osterroht, Angew. Chem. 1965, 77, 455-455.
[21] I. Kalinina, F. Mathey, Organometallics 2006, 25, 5031-5034.
[22] K. H. Pannell, H. K. Sharma, Organometallics 2010, 29, 4741-4745.
[23] K. Issleib, R. Kümmel, J. Organomet. Chem. 1965, 3, 84-91.
[24] N. De Luca, A. Wojcicki, J. Organomet. Chem. 1980, 193, 359-378.
[25] R. George, J.-A. M. Andersen, J. R. Moss, J. Organomet. Chem. 1995,
505, 131-133.
[26] H. Aktaş, J. C. Slootweg, K. Lammertsma, Angew. Chem., Int. Ed. 2010,
49, 2102-2113.
[27] J. Guihaumé, C. Raynaud, O. Eisenstein, L. Perrin, L. Maron, T. D. Tilley,
Angew. Chem., Int. Ed. 2010, 49, 1816-1819.
[28] K. Fukumoto, M. Kasa, T. Oya, M. Itazaki, H. Nakazawa,
Organometallics 2011, 30, 3461-3463.
This article is protected by copyright. All rights reserved.

10.1002/chem.201704954
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents
COMMUNICATION
A simple iron compound can
Justin K. Pagano, Brandon J. Ackley,
and Rory Waterman*
catalytically transfer a phosphinidene
equivalent to a 2,3-dimethyl-1,3-
butadiene, alkynes, and diethyl
disulfide with loss of hydrogen. These
phosphinidene transfer reactions and
dehydrocoupling catalysis indicate -
phosphinidene elimination
Page No. – Page No.
Evidence for Iron-Catalyzed α-
Phosphinidene Elimination with
Phenylphosphine
This article is protected by copyright. All rights reserved.