Properties and reactivity patterns of AsP3: An experimental and computational study of group 15 elemental molecules

Article abstract of DOI:10.1021/ja906294m

Facile synthetic access to the isolable, thermally robust AsP₃ molecule has allowed for a thorough study of its physical properties and reaction chemistry with a variety of transition-metal and organic fragments. The electronic properties of As

Full text of DOI:10.1021/ja906294m

Published on Web 10/02/2009
Properties and Reactivity Patterns of AsP3: An Experimental
and Computational Study of Group 15 Elemental Molecules
Brandi M. Cossairt and Christopher C. Cummins*
Massachusetts Institute of Technology, Department of Chemistry, 77 Massachusetts AVenue,
Cambridge, Massachusetts 02139
Received August 5, 2009; E-mail: ccummins@mit.edu
Abstract: Facile synthetic access to the isolable, thermally robust AsP3 molecule has allowed for a thorough
study of its physical properties and reaction chemistry with a variety of transition-metal and organic fragments.
The electronic properties of AsP3 in comparison with P4 are revealed by DFT and atoms in molecules
(AIM) approaches and are discussed in relation to the observed electrochemical profiles and the phosphorus
NMR properties of the two molecules. An investigation of the nucleus independent chemical shifts revealed
that AsP3 retains spherical aromaticity. The thermodynamic properties of AsP3 and P4 are described. The
reaction types explored in this study include the thermal decomposition of the AsP3 tetrahedron to its
elements, the synthesis and structural characterization of [(AsP3)FeCp*(dppe)][BPh4] (dppe ) 1,2-
bis(diphenylphosphino)ethane), 1, selective single As-P bond cleavage reactions, including the synthesis
and structural characterization of AsP3(P(N(iPr)2)N(SiMe3)2)2, 2, and activations of AsP3 by reactive early
transition-metal fragments including Nb(H)(η2-tBu(H)CdNAr)(N[CH2 Bu]Ar)2 and Mo(N[tBu]Ar)3 (Ar ) 3,5-
t
Me2C6H3). In the presence of reducing equivalents, AsP3 was found to allow access to [Na][E3Nb(ODipp)3]
(Dipp ) 2,6-diisopropylphenyl) complexes (E ) As or P) which themselves allow access to mixtures of
AsnP4-n (n ) 1-4).
Introduction
the tetraatomic tetrahedra AsnP4-n (n ) 1-3) is far from
complete. AsP3 has been suggested as an impurity in white
White phosphorus, P4, is of interest as a soluble, molecular
form of this element obtained industrially via the thermal method
of phosphate rock reduction. P4 is the primary precursor to most
P-containing compounds.1 Melting without decomposition at 44
°C, pure P4 is a white, waxy substance that presents no
difficulties in terms of isolation, storage, and handling, as long
as oxygen is excluded.1,2 On the other hand, yellow arsenic,
As4, is thermally and photochemically unstable in the condensed
phase, reverting readily to the metallic form.2 As such, its
chemistry and application in industry is not well developed.
Recently, the interpnictide molecule AsP3 was synthesized in
pure form and was shown to be robust, isolable, and as easy to
handle as P4 itself.3
Though only recently isolated and studied as a pure material,
AsP3 has long been contemplated by inorganic chemists in both
academic and industrial settings. Ozin was the first to observe
AsP3 as one component of a hot, gas-phase mixture of arsenic
and phosphorus vapors using Raman spectroscopy, inspiring
chemists to speculate on the stability of the AsP3 heteroatomic
tetrahedron.4 The physical properties of solid and liquid
arsenic-phosphorus phases have been described in some
detail,5,6 but our understanding of the nature and properties of
phosphorus as prepared industrially.7 The apatite mineral used
in the industrial production of P4 is known to contain a small
percentage of As for P substitution (0.003-0.03%) which is
not removed prior to reduction and likely leads to some amount
of AsP3 impurity in industrially prepared P4.8,9 With our
discovery of a selective, niobium-mediated synthesis of AsP3
and the present work, we are elucidating both the physical and
chemical properties of this small molecule that bridges the gap
between the soluble, molecular forms of the elements phos-
phorus and arsenic.
The present work delivers an overview of the synthesis and
properties of AsP3 as well as a detailed picture of the structural
and electronic properties of AsP3 as deduced by experiment,
density functional theory (DFT),10 and Bader’s atoms in
molecules (AIM)11 analysis. Also described herein are reactivity
studies of AsP3 with transition-metal complexes and with main-
group element systems, for comparison with related data on P4
(6) Ugai, Y. A.; Semenova, G. V.; Berendt, E.; Goncharov, E. G. Russ.
J. Phys. Chem. 1979, 53, 576–577.
(7) Schipper, W. Research and Development, Thermphos International.
Personal communication.
(8) Stow, S. H. Econ. Geol. 1969, 64, 667–671.
(9) Harnisch, H.; Heymer, G.; Klose, W. ; Schrodter, K. In Winnacker-
Kuchler: Chemische Technik: Prozesse und Produkte; Dittmeyer, R.,
Keim, W., Kreysa, G., Oberholz, A., Eds.; Wiley-VCH: Weinheim,
2005; Vol. 3, pp 343-427.
(1) Emsley, J. The 13th element: the sordid tale of murder, fire and
phosphorus; John Wiley & Sons, Inc.: New York, 2000.
(2) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nd ed.;
Butterworth-Heinemann: Oxford, 1997.
(10) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra,
C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput.
Chem. 2001, 22, 931–967.
(3) Cossairt, B. M.; Diawara, M. C.; Cummins, C. C. Science 2009, 323,
602.
(4) Ozin, G. A. J. Chem. Soc. A. 1970, 2307–2310.
(5) Karakaya, I.; Thompson, W. T. J. Phase Equilib. 1991, 12, 343–346.
(11) Bader, R. F. W. Atoms in Molecules: A Quantum Theory; Oxford
University Press: New York, 1994.
9
10.1021/ja906294m CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 15501–15511 15501

A R T I C L E S
Cossairt and Cummins
Scheme 1
Figure 1. Cyclic voltammogram of P4 (black) and AsP3 (red). 0.25 M
[NBu4][PF6] in THF at 20 °C, sweep rate 300 mV s-1, referenced vs Fc/
Fc+, AsP3 purified by sublimation (red), P4 purified by recrystallization
(black).
reactivity, and to show that under the proper conditions, AsP3
serves as a soluble As0 source. The utility of AsP3 for the
synthesis of mixed arsenic-phosphorus ligands that would be
difficult to synthesize by other means will also be highlighted.
fragment. This is a close analogy to the solution synthesis of a
heteroatomic interpnictide using transfer of P33- from an early
transition metal center to an As3+ fragment.
Results and Discussion
Synthesis and Physical Properties of AsP3. The industrial
synthesis of elemental phosphorus has remained essentially
unchanged for over a century and consists of heating phosphate
rock in excess of 1400 °C with coke and gravel.1,9 White
phosphorus can also be prepared from red phosphorus by
thermolysis at elevated temperature, similar to the synthesis of
yellow arsenic from metallic arsenic.2 White phosphorus has
become difficult to obtain for research purposes in the United
States, and both white and red phosphorus are classified by the
DEA as List I federally regulated substances.12
The physical properties of AsP3 have been probed by a variety
of methods. AsP3 readily sublimes under vacuum at 60 °C and
melts without decomposition at 72 °C.3 AsP3 has been shown
to be thermally stable in refluxing toluene solution for more
than one week.3 Raman spectroscopy obtained on solid samples
of AsP3 shows four resonances at 313 (a1), 345 (e), 428 (a1),
and 557 (e) cm-1, consistent with calculated stretching modes
for this C3V symmetric molecule.3 High-resolution, electron-
impact mass spectroscopy on a solid sample of AsP3 has
provided a mass of 167.8426 m/z.3 A solution molecular weight
determination of AsP3 gives a molecular weight of 167 ( 5
m/z (95% confidence level), confirming that the molecule also
exists in the monomeric form when in solution. Finally,
phosphorus NMR spectroscopy shows a single sharp resonance
at -484 ppm in benzene solution.3 This shift is 36 ppm
downfield of that for elemental phosphorus in the form of P4
(discussed below).
Anionic cyclo-P3 complexes of niobium have proven to be
3-
effective as P3 transfer agents, and when combined with an
E3+ source (E ) P, As, Sb), we have shown that access to P4,
AsP3, and SbP3 is possible.3,13-15 Specifically in the case where
E ) As, treatment of [Na][P3Nb(ODipp)3] (Dipp ) 2,6-
diisopropylphenyl) with 1 equiv of AsCl3 affords AsP3 in greater
than 70% isolated yield as shown in Scheme 1, i.3 This reaction
provides the first solution-phase synthesis of an interpnictide
AsnP4-n (n ) 1-3) tetrahedron.
Electronic Structure of AsP3. The electrochemical profiles
of AsP3 and P4 have been compared using cyclic voltammetry
in THF solution (Figure 1). P4 has a single broad reduction event
with onset at -1.3 V vs Fc/Fc+. This is the only observable
reduction feature in the P4 cyclic voltammogram (CV) and
possibly indicates formation of the P4 radical anion followed
by irreversible bond rupture.19 Scanning cathodically, the CV
of AsP3 displays a similarly broad irreversible reduction event
with onset at -1.0 V vs Fc/Fc+, which likely signifies more
facile generation of AsP3 radical anion and subsequent bond
rupture. The earlier onset of reduction for AsP3 suggests that
the LUMO of AsP3 is lower in energy than the LUMO of P4
by approximately 10 kcal mol-1. The calculated molecular
orbitals for P4 and AsP3 show that the LUMO of AsP3 is lower
in energy by 4.9 kcal mol-1 (Figure 2). Scanning anodically
from the rest potential, no observable oxidation events are
initially observable in the solvent window of the experiment.
Upon reduction however, both the AsP3 and P4 solutions show
a single oxidation wave. Repeated scanning causes the oxidation
Transition-metal chalcogenide chemistry provides us with a
family of reactions bearing close relation to our synthesis of
AsP3. Namely, treatment of Y2X2 (Y ) S, Se, Te; X ) Cl or
Br) with S5TiCp2 results in formation of Y2S5 and X2TiCp2.16-18
The first example of such a reaction (Y ) S, X ) Cl) comes
from a 1968 report by Schmidt and co-workers (Scheme 1, ii).16
When Y ) Se or Te, this reaction is an illustrative example of
the solution synthesis of a heteroatomic interchalcogenide using
2+
transfer of S52- from an early transition metal center to an M2
(12) Advisories to the Public. Red phosphorus, white phosphorus, and hypophos-
phorous acid are being used to make methamphetamine; Drug Enforce-
ment Agency, Office of Diversion Control: Washington, DC, http://www.
deadiversion.usdoj.gov/chem_prog/advisories/phosphorous.htm, June 18,
2009.
(13) Piro, N. A.; Cummins, C. C. J. Am. Chem. Soc. 2008, 130, 9524–
9535.
(14) Piro, N. A.; Cummins, C. C. Angew. Chem., Int. Ed. 2009, 48, 934–
938.
(15) Piro, N. A. Niobium-Mediated Generation of P-P Multiply Bonded
Intermediates. Ph.D. Thesis; Massachusetts Institute of Technology:
Cambridge, MA, June 2009.
(16) Schmidt, M.; Block, B.; Block, H. D.; Kopf, H.; Wilhelm, E. Angew.
Chem., Int. Ed. Engl. 1968, 7, 632–633.
(19) The electrochemistry data were collected in 0.25 M [TBA][PF6] at a
scan rate of 300 mV s-1. A 1 mm diameter Pt disk working electrode,
curly Pt wire counter electrode, and Ag wire pseudo-reference electrode
were used for all meaurements. All measurements were corrected to
Cp2Fe0/+ by the addition of Cp2Fe as an internal standard.
(17) Steudel, R.; Papavassiliou, M.; Strauss, E.; Laitinen, R. Angew. Chem.,
Int. Ed. Engl. 1986, 25, 99–101.
(18) Steudel, R.; Jensen, D.; Baumgart, F. Polyhedron 1990, 9, 1199–1208.
9
15502 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Synthesis and Structural and Electronic Properties of AsP3
A R T I C L E S
Figure 2. Molecular orbital diagrams for AsP3 (left) and P4 (right). Both diagrams were calculated using C3V symmetry to allow for direct comparison of
the calculated orbitals.
Table 1. NMR Shielding Tensors for AsP3 and P4
approximately 40 ppm. This decrease in the paramagnetic term
arises from greater coupling of the virtual and occupied orbitals
due to the decrease in the HOMO-LUMO gap (HLG).22 Thus,
it is the HLG, not the electron density at P, that is responsible
for the observed chemical shift difference between P4 and AsP3.
σdia
σpara
σSO
HLG (eV)
AsP3
P4
954.694
953.301
-151.744
-111.418
17.372
16.116
4.58
4.98
An additional electronic property of note for comparison
between P4 and AsP3 is the degree of spherical aromaticity
harbored by these clusters.23 Hirsch and co-workers have
previously applied the concept of spherical aromaticity to a
variety of Td cage molecules including P4 by calculating the
nucleus-independent chemical shift (NICS) at the cage critical
point.23,24 The NICS value for P4 is known by this method to
be large and negative, indicative of spherical aromaticity and
large diamagnetic ring currents in the cluster. Following location
of the cage critical points in P4 and AsP3, we find that the NICS
value calculated for P4, -59.444, is only one unit more negative
than that found for AsP3, -58.230. This indicates that despite
the lowering in molecular symmetry upon going from P4 to
AsP3, a great deal of spherical aromaticity is retained. The
retention of spherical aromaticity is partially due to the fact that
AsP3, like P4, maintains closed shell σ and π subsystems,
resulting in symmetrically distributed angular momenta (Figure
2).24
events to grow in intensity and shift slightly in potential,
suggesting that they are a result of electropolymerization or some
other complicating process. The observed stabilization of the
AsP3 LUMO renders reduction chemistry more facile and is
certainly a contributor to the enhanced reactivity of AsP3 as
compared to P4.
Closer scrutiny of the molecular orbital diagrams of AsP3
and P4 gives additional insight into the differences of these two
related molecules (Figure 2). We find that the HOMO-LUMO
energy gap of AsP3 is smaller in magnitude by 0.40 eV as
compared to that for P4 (Figure 2). The lower-energy HOMO-
LUMO gap of AsP3 is manifest experimentally in the 31P NMR
spectrum. It might be expected, on the basis of the electron
density distribution in AsP3, that the presence of the more
electropositive As atom would increase the electron density at
the phosphorus nuclei, causing the remaining three phosphorus
atoms to be more shielded and resonate at higher field relative
to P4. This, however, is not the case. In benzene solution, P4
resonates at -520 ppm in the 31P NMR spectrum, while the
phosphorus atoms of AsP3 resonate at -484 ppm. A decom-
position of the NMR shielding terms, as calculated with
ADF,20,21 reveals that the paramagnetic shielding term is the
dominant contributor to the chemical shift (Table 1). The
paramagnetic shielding term of AsP3 is more negative by
That the degree of spherical aromaticity in AsP3 is not
diminished relative to that of P4 implies that the presence of
the single As atom does not greatly perturb the charge
distribution. There are two limiting views of the charge
distribution in AsP3: one view is that AsP3 contains an As3+
3-
ion supported by a P3 unit, while an alternative view is that
(22) Widdifield, C. M.; Schurko, R. W. Concept. Magn. Reson. A 2009,
34A, 91–123.
(20) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra,
C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. J. Comput.
Chem. 2001, 22, 931–967.
(23) Schleyer, P. V.; Maerker, C.; Dransfeld, A.; Jiao, H.; Hommes, N.
J. Am. Chem. Soc. 1996, 118, 6317–6318.
(21) Fonseca Guerra, C.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor.
Chem. Acc. 1998, 99, 391–403.
(24) Hirsch, A.; Chen, Z.; Jiao, H. Angew. Chem., Int. Ed. 2001, 40, 2834–
2838.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15503

A R T I C L E S
Cossairt and Cummins
Table 2. Calculated Atomic Charges for AsP3
AIM (e)
Hirshfeld (e)
Voroni (e)
As
P
+0.126
-0.042
+0.0461
-0.015
+0.051
-0.018
Table 3. Heats of Atomization, Bond Energies, and Standard
Heats of Formation for AsnP4-n
heat of atomization
calculateda
bond energy
calculatedb
kcal mol-1
∆fHo
0
calculatedc
kcal mol-1
kcal mol-1
P4
AsP3
284
266
P-P 47
P-P 47
17
32
As-P 41
P-P 47
As2P2
250
45
As-P 41
As-As 36
As-P 41
As-As 36
As-As 36
As3P
As4
234
220
57
68
a Heat of atomization ) |total bonding energy AsnP4-n| - |(n (energy
4S As atoms)) + (4-n (energy 4S P atom)) | - |zero-point energy|.
Figure 3. Plot of ∆fHo vs n for AsnP4-n
.
0
c
b Heat of atomization/6. ∆fHo ) Σ(∆fHo constituent atoms)32 - heat
experimental values.30 Using this energy for the P-P bonds,
we can extend our method to AsP3. Following subtraction of 3
P-P single bond energies from the heat of atomization we
computed for AsP3, we are left with 3 As-P bonds; with an
estimated bond dissociation energy of 41 kcal mol-1 each.
Similarly, the As-As bonds of As4 were found to have bond
dissociation energies of 36 kcal mol-1. Keeping the P-P and
As-As bond energies constant, extraction of P-As bond
energies from the heats of atomization of As2P2 and As3P give
the same value of 41 kcal mol-1 (Table 3). In summary, our
calculations indicate that the As-P bonds of AsP3 are 6 kcal
mol-1 weaker than the P-P bonds of P4 (and 5 kcal mol-1
stronger than the As-As bonds of As4). This is in agreement
with our observations of the enhanced reactivity of these bonds
(Vide infra) relative to P4 as well as reported values for typical
P-P, P-As, and As-As single bond strengths.31
0
0
of atomization.
AsP3 contains neutral P atoms and a neutral As atom. Calcula-
tions using the AIM method have suggested that the phosphorus
atoms of AsP3 harbor a very slight negative charge of -0.04 e
and the arsenic atom makes up the balance, having a slight
positive charge of +0.12 e. Table 2 compiles the charge
descriptions given by the AIM method, as well as the Hirshfeld
and Voroni Deformation Density methods.25,26 All three meth-
ods give generally good agreement of the assigned charges and
support the description that AsP3 contains a neutral As atom
and three neutral P atoms. This charge distribution informs our
description of AsP3 as a soluble, molecular combination of these
two elements. This analysis is in agreement with more simple
interpretations based on Pauling electronegativities (P, 2.19 and
As, 2.18), as well as with more complicated theoretical
descriptions ( Figure 1S, Supporting Information).27,28
From the calculated values for heats of atomization and the
known heats of formation of the As and P atoms at 0 K, we are
able to extract estimated heats of formation at 0 K for AsnP4-n
by summation of the experimental heats of formation of the
constituent atoms followed by subtraction of the calculated heats
of atomization (Table 3). As would be expected, there is a
monotonic increase in the standard heats of formation with
increasing n (Table 3, Figure 3). This trend in heats of formation
begins to shed light on the thermodynamic stability of these
tetrahedral pnictogen molecules and shows that upon increasing
the arsenic content there is a significant price to pay for
formation of AsnP4-n from the elements in their standard states.
Reactivity Studies. A series of reactivity studies has been
carried out with AsP3. The reactions chosen have previously
been successfully carried out with P4 and include thermolysis,
adduct formation, single bond cleavage reactions by transition
metal and organic radicals, and molecule activation with
In order to obtain a more quantitative depiction of the As-P
and P-P bond energies in AsP3, we investigated the heats of
atomization of AsP3, As2P2, As3P, P4, and As4, in order to
calculate average bond energies for these species.27 Heats of
atomization are computed as an energy difference between the
molecule and the single atoms. The individual atoms are
computed as spherically symmetric and spin-restricted.29 In
order to accurately represent the true atomic ground state, we
calculated the fragment energy of a single P atom with three R
spins (4S ground state).29 After correction for the true atomic
ground state and for the zero-point energy of the molecule, we
obtained reliable and accurate heats of atomization as shown
in Table 3.28 These data permit estimation of the bond
dissociation energies; for example, the P4 molecule is composed
of six P-P bonds, so division of the heat of atomization of P4
by six gives an estimate of the P-P bond energy.27 Our
Nb(H)(η2-tBu(H)CdNAr)(N[CH2 Bu]Ar)2 (Ar ) 3,5-Me2C6H3),
t
calculations put the P-P bond energy of P4 at 47 kcal mol-1
,
Mo(N[tBu]Ar)3, and Cl2Nb(ODipp)3 under reducing conditions.
In looking at such a class of reactions we hope to compare the
which compares perfectly with the 47 kcal mol-1 obtained from
(25) Nalewajski, R. F. Phys. Chem. Chem. Phys. 2002, 4, 1710–1721.
(26) Fonseca Guerra, C.; Handgraaf, J. W.; Baerends, E. J.; Bickelhaupt,
F. M. J. Comput. Chem. 2003, 25, 189.
(30) Tsirelson, V. G.; Tarasova, N. P.; Bobrov, M. F.; Smetannikov, Y. V.
Heteroatom. Chem. 2006, 17, 572–578.
(27) Pauling, L. The Nature of the Chemical Bond, 3rd ed.; Cornell
University Press: Ithaca, 1960.
(31) Norman, N. C. Chemistry of Arsenic, Antimony, and Bismuth; Springer:
New York, 1998.
(28) Please see the Supporting Information for further details.
(29) Baerends, E. J.; Branchadell, V.; Sodupe, M. Chem. Phys. Lett. 1997,
265, 481–489.
(32) Chase, M. W. NIST-JANAF Thermochemical Tables; American
Institute of Physics: Melville, NY (J. Phys. Chem. Ref. Data,
Monograph 9), 1988; Vol. 28, Iss. 6.
9
15504 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Synthesis and Structural and Electronic Properties of AsP3
A R T I C L E S
Figure 4. Thermal decomposition of AsP3 to metallic arsenic and red phosphorus. (A) Raman spectra of the product of heating AsP3 (black) and an
authentic sample of red phosphorus (red). (B) Powder XRD data for the product of heating AsP3. (C) Photographs of the reaction vessel before heating
(bottom) and after heating (top). (D) Elemental mapping by EDS analysis using an SEM microscope.
reactivity patterns of P4 and AsP3 in light of their contrasting
physical and electronic properties as elucidated in the theoretical
studies.
Thermal decomposition of AsP3 to the elements is quite
surprising and the mechanism by which this occurs is not known.
One possibility is that under thermal conditions four molecules
of AsP3 disproportionate to give rise to three molecules P4 and
one molecule of As4, which themselves then revert to red
phosphorus and metallic arsenic. From our calculated heats of
formation (Figure 2), we can get an estimate for the heat of
disproportionation of four molecules of AsP3 to give three
molecules of P4 and one molecule of As4. In so doing we find
that the process is downhill by 7.42 kcal mol, suggesting that
the kinetic stability of AsP3 is quite important in its isolability,
as thermodynamically it is only metastable. An alternative
mechanism might involve the formation of “red AsP3”, which
is not itself thermally robust, reverting to red phosphorus and
metallic arsenic. It may be speculated that an alternative to the
thermal pathway may yet be successful in leading to a red
polymeric form of AsP3, and preliminary studies on AsP3
polymerization by radical initiators and UV light are under way.
Coordination Chemistry of AsP3. In our initial report on the
facile synthesis of AsP3 we used the complex (µ-N2)[Mo(CO)3-
(PiPr3)2]2 to complex the AsP3 tetrahedron and to provide a
structural glimpse of the intact tetrahedron (Scheme 2, i).3
Unfortunately, the AsP3 adduct, (AsP3)Mo(CO)3(PiPr3)2 was
unstable above 0 °C and for that reason was difficult to work
with. Additionally, accurate determination of the As-P inter-
atomic distances in (AsP3)Mo(CO)3(PiPr3)2 was hampered by a
70:30 disorder between the As atom and one of the two
noncomplexed P atoms in the tetrahedron. Thus, we sought to
synthesize a more thermally stable adduct of AsP3 and hoped
to obtain accurate metrical parameters from an ordered crystal
structure.
Thermolysis of AsP3. Amorphous red phosphorus was first
obtained in 1848 by heating P4 in the absence of air for several
days at high temperatures, and is now made on a commercial
scale of 7000 tons per year by a similar thermal conversion
process.2 Red phosphorus is much less toxic and much less
reactive than monomeric white phosphorus and it is extensively
used in the production of matches, flame retardants, and
phosphide materials for semiconductor applications.2 Semicon-
ductor applications account for a majority of red phosphorus
consumption, with aluminum phosphide production accounting
for over 24% of the total consumption. Much of this aluminum
phosphide is further alloyed with species such as gallium
arsenide to tune the semiconductor band gap.1,2
Polymeric forms of P/As alloys are exceedingly rare.33,34
Intrigued by the possibility of a “red AsP3” phase wherein one
out of every four sites in red P would be occupied by an As
atom, we subjected AsP3 to the same conditions under which
P4 converts to red P. Heating a sealed tube containing white
AsP3 to 300 °C for 36 h resulted in apparent segregation of the
elements, producing amorphous metallic arsenic and amorphous
red phosphorus, as determined by Raman spectroscopy (diag-
nostic for red P35), powder XRD analysis (diagnostic for
amorphous metallic As36), and EDS elemental analysis of the
bulk material (Figure 4). The EDS elemental map shown in
Figure 4 shows distinct regions where only phosphorus (red)
and only arsenic (green) are sequestered; however, there are
some regions (yellow) where the two species may be found
together, suggestive of incomplete separation. Repetition of this
experiment on a variety of scales gave reproducible results.
The precursor FeCp*(dppe)Cl, which had been reported to
form a stable P4 adduct,37 has allowed us to realize this goal.
Treatment of dark-green FeCp*(dppe)Cl with 1 equiv of AsP3
(33) Jayasekera, B.; Somaskandan, K.; Brock, S. L. Inorg. Chem. 2004,
43, 6902–6904.
(34) Von Schnering, H. G.; Honle, W. Chem. ReV. 1988, 88, 243–273.
(35) Fasol, G.; Cardona, M.; Honle, W.; Von Schnering, H. G. Solid State
Commun. 1984, 52, 307–310.
(37) de los Rios, I.; Hamon, J. R.; Hamon, P.; Lapinte, C.; Toupet, L.;
Romerosa, A.; Peruzzini, M. Angew. Chem., Int. Ed. 2001, 40, 3910–
3912.
(36) Breitling, G. Mater. Res. Bull. 1969, 4, 19–32.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15505

A R T I C L E S
Cossairt and Cummins
Scheme 2
single As-P bond has been selectively cleaved, as assigned by
31P NMR spectroscopy (Scheme 3, i, and Figure 3S, Supporting
Information). Recrystallization of the crude reaction mixture
from n-hexane/Et2O gave crystalline 2 in 88% yield. X-ray
crystallographic analysis on a single crystal of 2 confirmed the
selective As-P bond cleavage as shown in Figure 6 and Table
2S, Supporting Information. The geometrical parameters of the
arsatriphosphabicyclobutane core of 2 are nearly isomorphous
with the all-phosphorus analogue as reported by Lappert and
co-workers.38 A discussion of accurate P-P and As-P bond
lengths is precluded by disorder in the crystal structure of 2.
Notably, there is a positional disorder of As1 and P1, but there
is no As component to the P2 or P3 positions.
The bright green Ti(III) reagent Ti(N[tBu]Ar)3 (Ar ) 3,5-
Me2C6H3) has proven to be a potent metalloradical and one-
electron reductant since its first synthesis by reduction of the
corresponding ClTi(N[tBu]Ar)3 complex in 1995.39-43 Given
the facile radical opening of the AsP3 tetrahedron with
(P(N(iPr)2)N(SiMe3)2)2, we hypothesized that Ti(N[tBu]Ar)3
would be a good candidate for forming a 1,4-bis(metallo)arsa-
triphosphabicyclobutane. In fact, Ti(N[tBu]Ar)3 does react with
AsP3 to generate the desired 1,4-bis(metallo)arsatriphosphabi-
cyclobutane, AsP3(Ti(N[tBu]Ar)3)2, 3-AsP3, in which a single
As-P bond has been cleaved selectively; however, the reaction
does not go to completion. Instead, an equilibrium is established
between AsP3, Ti(N[tBu]Ar)3, and 3-AsP3 (Scheme 3, ii). This
equilibrium was studied by 31P NMR spectroscopy at a variety
of concentrations and temperatures. In all cases the equilibrium
favored the reactants, with 3-AsP3 never accounting for more
than 30% of the reaction mixture (Figures 4S, 14S, 15S,
Supporting Information).44 The analogous reaction with P4
likewise gave an equilibrium mixture, but the 1,4-bis(metal-
lo)tetraphosphabicyclobutane product, P4(Ti(N[tBu]Ar)3)2, 3-P4,
never accounted for more than 5% of the reaction mixture under
identical conditions. Although no intermediates were detected
in these equilibria, the systems did not conform to the simple
equilibrium expression, Keq ) 3/[Ti(N[tBu]Ar)3]2[EP3] (E ) As,
P), suggesting that more complicated processes are at work. It
should be noted that the concentration of Ti(N[tBu]Ar)3 was
obtained by subtraction of 2[3-E4] from [Ti(N[tBu]Ar)3]0. In
direct competition reactions, the tendency of Ti(N[tBu]Ar)3 to
react with EP3 (E ) P, As) to form the radical-opened
bicyclobutane derivatives, 3-AsP3 and 3-P4, revealed a measur-
able preference for edge-opening of AsP3, (Figures 14S and 15S,
Supporting Information), consistent with heightened reactivity
for AsP3 over P4 (Vide supra). One hypothesis for why
Ti(N[tBu]Ar)3 is unsuccessful at driving the radical cleavage
of either P4 or AsP3 fully to 3-EP3 (E ) P, As) is centered upon
the unfavorable loss in entropy inherent in forming a single-
product molecule from three reactant molecules. The interaction
in THF results in a gradual color change to brown over 20 min.
Subsequent treatment of the reaction mixture with 1 equiv of
NaBPh4 results in an immediate color change to bright magenta.
Following removal of the NaCl byproduct, [(AsP3)Fe-
Cp*(dppe)][BPh4], 1, was isolated in 80% yield by recrystal-
lization of the crude solids from 1:1 Et2O/THF (Scheme 2, ii).
X-ray diffraction quality crystals of 1 were grown from a
mixture of THF and CH2Cl2 at -35 °C over the course of
several days. The X-ray structure of 1 (Figure 5) displays a
fully ordered AsP3 tetrahedron in an η1 binding mode at a
phosphorus vertex (Table 1S, Supporting Information). The
solid-state structure of 1 is consistent with the solution-state
configuration of the AsP3 unit as assigned by NMR spectroscopy
(Supporting Information Figure 2S). The P12-As1 and
P12-P13,14 distances in 1 are 2.283(2) Å and 2.183(3) Å,
respectively, and are noticeably shorter than typical P-As and
P-P single bonds. The other As-P and P-P bonds are
noticeably longer with As1-P13 at 2.334(2) Å, As1-P14 at
2.336(2) Å, and P13-P14 at 2.231(3) Å. The shortened bond
lengths to P12 are the effect of rehybridization that occurs upon
AsP3 binding to the Fe center, producing greater 3s character
in the bonding molecular orbitals of this four-coordinate
phosphorus atom. It is noteworthy that the Fe1-P12 interatomic
distance of 2.172(2) Å is significantly shorter than the
Fe-phosphine bond distances at 2.233(2) Å and 2.242(2) Å.
As-P Bond Cleavage Reactions. The estimated 6 kcal mol-1
difference in energy between the As-P bonds and the P-P
bonds in AsP3 suggests that reactions resulting in selective As-P
bond cleavage may be possible. Several systems are known to
promote radical opening of the P4 tetrahedron to produce
substituted tetraphosphabicyclobutane structures. A noteworthy
example of this reaction type was reported by Lappert and co-
workers and uses (P(N(iPr)2)N(SiMe3)2)2 as a source of the
phosphorus radical •P(N(iPr)2)N(SiMe3)2, which was shown to
activate P4 to produce the corresponding 1,4-bis(phosphido)t-
etraphosphabicyclobutane.38 While activation of P4 by (P(N-
(iPr)2)N(SiMe3)2)2 requires somewhat harsh conditions (refluxing
in toluene for 1.5 h), the corresponding reaction with AsP3 is
rapid at room temperature. Upon mixing (P(N(iPr)2)N(SiMe3)2)2
with 1 equiv of AsP3 in toluene the initially colorless reaction
mixture turns bright yellow. NMR spectroscopic analysis of the
crude reaction mixture shows clean and quantitative conversion
to AsP3(P(N(iPr)2)N(SiMe3)2)2 (set of isomers), 2, in which a
(39) Wanandi, P. W.; Davis, W. M.; Cummins, C. C. J. Am. Chem. Soc.
1995, 117, 2110–2111.
(40) Peters, J. C.; Johnson, A. R.; Odom, A. L.; Wanandi, P. W.; Davis,
W. M.; Cummins, C. C. J. Am. Chem. Soc. 1996, 118, 10175–10188.
(41) Agapie, T.; Diaconescu, P. L.; Mindiola, D. J.; Cummins, C. C.
Organometallics 2002, 21, 1329–1340.
(42) Agarwal, P.; Piro, N. A.; Meyer, K.; Muller, P.; Cummins, C. C.
Angew. Chem., Int. Ed. 2007, 46, 3111–3114.
(43) Kim, E.; Odom, A. L.; Cummins, C. C. Inorg. Chim. Acta 1998, 278,
103–107.
(44) Concentrations of AsP3 and 3-AsP3 were determined by integration
against an internal standard (PPh3). The concentration of Ti(N[tBu]Ar)3
was inferred using the equation [Ti(N[tBu]Ar)3]f ) [Ti(N[tBu]Ar)3]i-
2[3-AsP3]f. This assumes no other degradation pathways are operative
during the course of the reaction.
(38) Bezombes, J. P.; Hitchcock, P. B.; Lappert, M. F.; Nycz, J. E. Dalton
Trans. 2004, 499–501.
9
15506 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Synthesis and Structural and Electronic Properties of AsP3
A R T I C L E S
Figure 5. Solid-state structure of complex 1 shown with 50% probability thermal ellipsoids. (A) View of the [(AsP3)FeCp*(dppe)]+ core. (B) View of the
asymmetric unit. Hydrogen atoms and CH2Cl2 solvent residues have been omitted for clarity.
Scheme 3
between the hard titanium(IV) metal center and the soft
phosphorus (or arsenic) center does not provide a large enough
enthalpic contribution to counterbalance the entropic losses
incurred.
Formation of µ2:η2,η2-E2Nb2 Complexes (E ) As, P). The
activation of white phosphorus by early transition-metal com-
plexes has been a prolific area of investigation in recent
years.45-48 Included among these are niobium complexes
bearing three monoanionic ligands. This class of complexes has
given rise to a wide array of Pn ligands (n ) 1-8).3,45,49,50
The metallaziridine(hydride) derivative, Nb(H)(η2-tBu(H)Cd
NAr)(N[CH2 Bu]Ar)2 (Ar ) 3,5-Me2C6H3), was found to react
t
with 0.25 equiv of white phosphorus to quantitatively give rise
to the corresponding (µ2:η2,η2-P2)[Nb(N[CH2tBu]Ar)3]2 complex,
which has a diagnostic phosphorus NMR resonance at 399 ppm
in C6D6. Because of the high efficiency of this reaction it seemed
to be an ideal candidate for reactivity tests with AsP3, with an
eye toward formation of (µ2:η2,η2-P2)[Nb(N[CH2 Bu]Ar)3]2 and
t
(µ2:η2,η2-AsP)[Nb(N[CH2 Bu]Ar)3]2 in a 1:1 ratio. In fact,
t
treatment of bright yellow Nb(H)(η2-tBu(H)CdNAr)(N[CH2-
Figure 6. Solid-state structure of compound 2 with 50% probability thermal ellipsoids. Hydrogen atoms are omitted for clarity.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15507

A R T I C L E S
Cossairt and Cummins
Scheme 4
Scheme 5
tBu]Ar)2 with a slight excess (0.35 equiv) of AsP3 results in
clean and complete conversion to the expected 1:1 green mixture
but the corresponding 13C NMR spectroscopic features are
distinct and reveal that the reaction cleanly forms 5-As and 5-P
in a 1:3 ratio (Figure 6S, Supporting Information). Interestingly,
it was found that in 1:1 competition experiments of AsP3 and
P4 with Mo(N[tBu]Ar)3, there is no selectivity for reaction with
AsP3 over P4 (Figure 16S, Supporting Information). This is a
striking example of a reaction in which AsP3 does not give
enhanced reactivity over P4, and is an interesting case. The
reaction of AsP3 with Mo(N[tBu]Ar)3 gives us clear proof that
AsP3 functions experimentally as a soluble source of As0 and
P0 and also provides motivation for future studies in the
synthesis of precise 3:1 mixtures of metal phosphides and metal
arsenides for materials applications.56-58
t
of products, (µ2:η2,η2-P2)[Nb(N[CH2 Bu]Ar)3]2, 4-P2, and (µ2:
t
η2,η2-AsP)[Nb(N[CH2 Bu]Ar)3]2, 4-AsP, over the course of 90
min (Scheme 4). 31P NMR spectroscopic analysis of the crude
reaction mixture shows exclusively two resonances in a 2:1 ratio
at 399 and 438 ppm, respectively (Figure 5S, Supporting
Information). Assignment of the resonance at 438 ppm to 4-AsP
was corroborated by DFT NMR shielding calculations.28 This
reaction provides access to a rare AsP ligand bridging two
niobium metal centers, and future work with 4-AsP will be
aimed at identifying reactions in which the AsP unit is
released.51-53
Formation of Terminal Et M Complexes (E ) As, P). P4
degradation to P1 units has been observed in the treatment of
the reactive Mo(III) precursor Mo(N[tBu]Ar)3 with 0.25 equiv
of P4 to give the Pt Mo(N[tBu]Ar)3 complex.54 Access to the
arsenide congener of Pt Mo(N[tBu]Ar)3 is also afforded by the
corresponding reaction with As4; however, this experiment is
challenging as As4 is unisolable.55 A soluble, isolable source
of As0 would be preferable to obtain a stoichiometric transfor-
mation to the terminal arsenide complex, and AsP3 proves to
be an effective source of a single As0 equivalent. Treatment of
Mo(N[tBu]Ar)3 with a slight excess of AsP3 (0.29 equiv) results
in formation of a 1:3 mixture of Ast Mo(N[tBu]Ar)3, 5-As, and
Pt Mo(N[tBu]Ar)3, 5-P, over the course of 1 h (Scheme 5). The
1H NMR spectroscopic features for 5-As and 5-P are coincident,
Formation of cyclo-E3 Complexes (E ) As, P), As2P2, and
As3P. Our original synthesis of AsP3 involved treatment of an
anionic niobium cyclo-P3 complex with AsCl3 to obtain the
tetraatomic tetrahedron. Since that discovery, we have been
interested in preparing other anionic cyclo-E3 compounds, in
particular an anionic cyclo-As2P complex. Such a species could,
in principle, be used to synthesize the previously unexplored
tetraatomic interpnictides As2P2 and As3P by treatment of the
cyclo-As2P derivative with the appropriate ECl3 reagent (E )
P, As). Strides in this direction have been made by investigating
the reaction between AsP3 and Cl2Nb(ODipp)3 in the presence
of a reducing agent.3,59 Combining AsP3 with Cl2Nb(ODipp)3
in a 1:1 ratio in THF followed by treatment with Na/Hg
amalgam gives rise to a mixture of [Na][P3Nb(ODipp)3],
6-P3, [Na][AsP2Nb(ODipp)3], 6-AsP2, [Na][As2PNb(ODipp)3],
6-As2P, and presumably, [Na][As3Nb(ODipp)3], 6-As3, over the
course of 30 min. Evidence for formation of 6-P3, 6-AsP2, and
6-As2P in a 10:5:1 molar ratio (76% yield from Cl2Nb(ODipp)3)
(45) Figueroa, J. S.; Cummins, C. C. Dalton Trans. 2006, 2161–2168.
(46) Ehses, M. P.; Romerosa, A.; Peruzzini, M. Top. Curr. Chem. 2002,
220, 107–140.
(47) Peruzzini, M.; Gonsalvi, L.; Romerosa, A. Chem. Soc. ReV. 2005, 34,
1038–1047.
is provided by the clean appearance of three singlets in the 31
P
(48) Scherer, O. J. Acc. Chem. Res. 1999, 32, 751–762.
(49) Cossairt, B. M.; Cummins, C. C. Angew. Chem., Int. Ed. 2008, 47,
169–172.
NMR spectrum of the crude reaction mixture, spaced as
expected for sequential As-doping of the cyclo-P3 complex (Vide
supra) at -206 (6-P3), -167 (6-AsP2), and -132 ppm (6-As2P)
(Figure 7S, Supporting Information). From the considerable
(50) Piro, N. A.; Figueroa, J. S.; McKellar, J. T.; Cummins, C. C. Science
2006, 313, 1276–1279.
(51) Umbarkar, S.; Sekar, P.; Scheer, M. J. Chem. Soc. Dalton Trans. 2000,
1135–1137.
(52) Davies, J. E.; Mays, M. J.; Raithby, P. R.; Shields, G. P.; Tompkin,
P. K.; Woods, A. D. J. Chem. Soc. Dalton Trans. 2000, 1925–1930.
(53) Davies, J. E.; Kerr, L. C.; Mays, M. J.; Raithby, P. R.; Tompkin, P. K.;
Woods, A. D. Angew. Chem., Int. Ed. 1998, 37, 1428–1429.
(54) Laplaza, C. E.; Davis, W. M.; Cummins, C. C. Angew. Chem., Int.
Ed. Engl. 1995, 34, 2042–2044.
(56) Carenco, S.; Resa, I.; LeGoff, X.; LeFloch, P.; Mezailles, N. Chem.
Commun. 2008, 2568–2570.
(57) Yan, P.; Xie, Y.; Wang, W.; Liu, F.; Qian, Y. J. Mater. Chem. 1999,
9, 1831–1833.
(58) Li, B.; Xie, Y.; Jiaxing, H.; Liu, Y.; Qian, Y. Ultrason. Sono. 2001,
8, 331–334.
(55) Curley, J. C.; Piro, N. A.; Cummins, C. C. Inorg. Chem. 10.1021/
ic9016068, 2009.
(59) Clark, J. R.; Pulvirenti, A. L.; Fanwick, P. E.; Sigalas, M.; Eisenstein,
O.; Rothwell, I. P. Inorg. Chem. 1997, 36, 3623–3631.
9
15508 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Synthesis and Structural and Electronic Properties of AsP3
A R T I C L E S
Scheme 6
ratio of As to P atoms in the final product mixture is 1 to 1.6,
which is very near the 1 to 1.3 ratio that we would expect (3/4
AsP3 + AsCl3) over the course of the two reactions. This is an
important observation that implies that there is no selective loss
of As or P atoms during the synthesis. We are currently
investigating alternative syntheses of As2P2 and As3P for
exploration of these molecules as pure substances.
Conclusions and Future Directions
For the first time, the reaction chemistry of the interpnictide
compound AsP3 has been investigated. In the presence of Lewis
acidic metal fragments, adduct formation is selective for binding
at a single phosphorus vertex, and reducing main group and
transition metal species effect radical cleavage reactions where
a single As-P bond is selectively cleaved. Furthermore, AsP3
can be used to prepare otherwise difficult to access mixed As/P
molecules, both as free entities, in the case of the mixed
interpnictides AsnP4-n (n ) 1-3), or as coordinated ligands, in
the case of 4-AsP. The electronic and structural features that
render AsP3 distinct from P4 are consistent with the reactivity
patterns that have emerged. In particular, we have been able to
distinguish between the vertices of the AsP3 molecule in the
formation of compounds 1 and 2. DFT calculations have
provided us insight into the electronic structure of this unique
tetraatomic molecule and have provided us with estimates of
important thermodynamic parameters. We are currently inves-
tigating the use of AsP3 in the synthesis of solid-state, mixed
arsenide-phosphide materials as well as a “red AsP3” phase,
both of which may be of interest for the development of novel
semiconductor materials. Future work will target the directed
synthesis of the new interpnictides As2P2 and As3P reported
herein as well as further unveil the diverse reaction chemistry
of AsP3.
Table 4. NMR and GC-MS Data for Syntheses of AsP3, As2P2,
As3P, and As4
P4 (internal standard) AsP3
As2P2
As3P
As4
N/A
N/A
31P shift (ppm)
-520
-520
7.8 min
124
-484 -452 -432
-482 -451 -424
calculated 31P shift (ppm)
a
8.9 min 9.8 min 10.6 min 11.2 min
GC-MS retention time
GC-MS mass (m/z)
GC-MS area (au)
yield
168
212
256
300
50994
N/A
30475 11348 3176
∼1977
b
c
d
e
f
49%
39%
50%
2%
a GC-MS data were collected using an Agilent 6890N network GC
system with an Agilent 5973 Network mass selective detector and an
Rtx-1 column from Restek. b Yields of AsP3, As2P2, and As3P are
calculated from the corresponding starting concentration of 6-E3; in each
case the theoretical yield is 100%; for As4 the yield is calculated from
the starting concentration of Cl2Nb(ODipp)3 making the theoretical yield
,100%. c GC-MS yield from 6-P3. d GC-MS yield from 6-AsP2.
e GC-MS yield from 6-As2P. f GC-MS yield from Cl2Nb(ODipp)3.
Experimental Details
General Considerations. All manipulations were performed in
a Vacuum Atmospheres model MO-40 M glovebox under an inert
atmosphere of purified N2. All solvents were obtained anhydrous
and oxygen-free by bubble degassing (N2) and purification through
columns of alumina and Q5. Deuterated solvents were purchased
from Cambridge Isotope Laboratories. Benzene-d6 and toluene-d8
were degassed and stored over molecular sieves for at least 2 d
prior to use. Celite 435 (EM Science) were dried by heating above
200 °C under a dynamic vacuum for at least 24 h prior to use.
AsP3,3 Mo(N[tBu]Ar)3,61 Ti(N[tBu]Ar)3,40 Cl2Nb(ODipp)3,59 Nb(H)
38
(η2-tBu(H)CdNAr)(N[CH2 Bu]Ar)2 (PN(iPr)2N(SiMe3)2)2, and
FeCp*(dppe)Cl37 were synthesized according to reported methods.
All glassware was oven-dried at temperatures greater than 170 °C
prior to use. NMR spectra were obtained on Varian Inova 500
instruments equipped with Oxford Instruments superconducting
magnets and referenced to residual C6D6 (1H ) 7.16 ppm, 13C )
t
Figure 7. GC-MS chromatogram following treatment of in situ generated
6-P3, 6-AsP2, 6-As2P, and 6-As3 with AsCl3. P4 added as an internal standard.
128.06 ppm) or (CD3)2CO (1H ) 2.05 ppm, 13C ) 29.84 ppm). 31
P
NMR spectra were referenced externally to 85% H3PO4 (0 ppm).
Elemental analyses were performed by Midwest Microlab LLC,
Indianapolis, IN. UV-vis spectra were obtained on a Hewlett-
Packard 845x series UV-vis system equipped with a tungsten lamp.
GC-MS data were collected using an Agilent 6890N network GC
system with an Agilent 5973 Network mass selective detector and
an Rtx-1 column from Restek. MALDI-TOF MS data were
collected on a Bruker OmniFlex instrument, and data were
processed using the Bruker FlexControl software package. For
concentration of the unexpected 6-As2P (and from subsequent
analysis, Vide infra), we propose that 6-As3 is also present in
low concentration. Treatment of this crude reaction mixture with
AsCl3 with the exclusion of light rapidly generated a reaction
mixture containing Cl2Nb(ODipp)3(THF) and the tetraatomic
interpnictides AsP3, As2P2, As3P, as well as As4 (Scheme 6).60
The presence of the four tetrahedral molecules is confirmed by
31P NMR spectroscopy and by GC-MS (Table 4, Figure 7 and
Figures 8S-13S, Supporting Information). It is of note that the
(61) Laplaza, C. E.; Cummins, C. C. Science 1995, 268, 861–863.
(62) Figueroa, J. S.; Cummins, C. C. J. Am. Chem. Soc. 2003, 125, 4020–
4021.
(60) P4 was used as an internal standard for monitoring purposes. No P4
was present in the reaction mixture prior to its addition.
(63) Sheldrick, G. M. SADABS; Bruker AXS Inc.: Madison, WI, 2005.
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15509

A R T I C L E S
Cossairt and Cummins
Raman studies, an Invictus solid-state laser at 785 nm, manufactured
by Kaiser Optics, was routed through fiber-optic cables to a Hololab
series 5000 Raman Microscope. The Raman scattering was observed
via 180° reflectance through the objective of the Raman microscope.
Each spectrum was corrected for dark current and cosmic ray
interference using the Hololab software. Powder diffraction data
were collected on a PANalytical X’Pert Pro multipurpose diffrac-
tometer equipped with a 1.8 kW sealed tube X-ray source using
Mo KR radiation (λ ) 0.71073 Å) and equipped with high-speed
Bragg-Brentano optics. Scanning electron microscopy was per-
formed on a JEOL JSM-5910 instrument using a JEOL BEI detector
and a Rontec EDX system for elemental analysis and mapping.
Thermolysis of AsP3. AsP3 (100 mg) was loaded into a thick-
walled glass tube. The tube was sealed under vacuum. The tube
was wired to a thermocouple probe, and the probe and tube were
wrapped completely with heating tape. The heating tape was set to
heat at 290-300 °C for 40 h. After this time the apparatus was
cooled, and the heating tape was removed. The white-yellow AsP3
had been converted to a red-black material as well as a metallic
shiny material which had sublimed partially up the tube. Raman
spectroscopy of the bulk material through the tube revealed
complete consumption of the AsP3 tetrahedron. The tube was scored
and broken open. The contents of the tube were removed and placed
on a zero background silicon 510 surface for powder diffraction.
Powder diffraction data: broad peaks centered around 18°, 32°, 58°,
and 90° 2θ. The product mixture was then also analyzed by Raman
spectroscopy and EDS using a JEOL SEM microscope. Raman
spectroscopy results: broad and weak resonance at 280 cm-1, sharp
and intense resonance at 320 cm-1, broad and intense resonances
extending from 340 to 500 cm-1. SEM data: in a 50 µm × 50 µm
region, elemental composition analysis gave phosphorus 74.01%
(error 0.66%) and arsenic 25.98% (error 0.76%).
NMR analysis which showed clean and complete conversion of
the starting materials to AsP3(P(N(iPr)2)N(SiMe3)2)2. The reaction
mixture was taken to dryness under reduced pressure, and the
resulting residue was dissolved in n-hexane/Et2O (2:1) and recrys-
tallized, affording 196 mg (0.261 mmol, 88% yield) of pale-yellow
crystals of AsP3(P(N(iPr)2)N(SiMe3)2)2 (mixture of diastereomers).
1H NMR (20 °C, benzene-d6, 500 MHz): δ ) 0.43 (36H, br, SiMe3),
i
i
i
1.03 (12H, m, Pr-Me), 1.20 (6H, m, Pr-Me), 1.26 (6H, m, Pr-
Me), 3.34 (2H, m, Pr-CH), 3.48 (2H, m, Pr-CH). 31P{1H} NMR
(20 °C, benzene-d6, 202.5 MHz): δ ) -311 (2P, m, P(PP)As),
-148.5 (1P, m, P(PP)As), 118.5 (1P, m, As bound phosphine),
123 (1P, m, P bound phosphine). 13C{1H} NMR (20 °C, benzene-
d6, 125.8 MHz): δ ) 5.1 (br, SiMe3), 6.53 (br, SiMe3), 23.6 (m,
i
i
i
i
iPr-Me), 24.7 (m, Pr-Me), 48.2 (m, Pr-CH). Elem. Anal. Calcd
for C24H64AsN4P5Si4: C 38.39, H 8.59, N 7.46, P 20.62; Found: C
39.42, H 8.78, N 8.26, P 19.27.
Treatment of AsP3 (and P4) with Ti(N[tBu]Ar)3, Synthesis of
3-P4 and 3-AsP3. To begin the measurements, AsP3 (15 mg, 0.09
mmol), P4 (11 mg, 0.09 mmol), PPh3 (12 mg, 0.045 mmol),
Ti(N[tBu]Ar)3 (68 mg, 0.118 mmol), and toluene (872 mg) were
combined and placed in a J-Young-style NMR tube. The temper-
ature was varied between 5 and 35 °C, allowing the tube to fully
equilibrate before taking the final concentration readings. The
simplest equilibrium expression was not obeyed, but the product
concentration for 3-AsP3 was always higher than for 3-P4. For
example, at 20 °C 3-AsP3 and 3-P4 were present in a 7:1 ratio
(Figure 14S, Supporting Information). These results show that in a
1:1 mixture of AsP3 and P4, a greater percentage of AsP3 undergoes
reaction. This experiment was repeated with only AsP3 in the
reaction mixture and alternatively with only P4 present to obtain
clean NMR spectral data for both 3-AsP3 and 3-P4. 31P{1H} NMR
(AsP3[Ti(N[tBu]Ar)3], 20 °C, benzene-d6, 202.5 MHz): δ ) -275
(2P, d, J1 ) 200 Hz, P(PP)As), -9.6 (1P, t, 1J ) 200 Hz, P(PP)As).
31P{1H} NMR (P4[Ti(N[tBu]Ar)3], 20 °C, benzene-d6, 202.5 MHz):
δ ) -284 (2P, t, 1J ) 188 Hz, P(PP)P), 12.9 (2P, t, 1J ) 188 Hz,
P(PP)P).
[(AsP3)FeCp*(dppe)][BPh4], 1. FeCp*(dppe)Cl (144 mg, 0.230
mmol, 1 equiv) was dissolved in 8 mL of THF and was transferred
to a vial containing solid AsP3 (43 mg, 0.256 mmol, 1.1 equiv)
and a stir bar. The reaction mixture was allowed to stir for 30 min
during which time the originally green reaction mixture went
slightly orange (subtle change). At this point NaBPh4 (79 mg, 0.230
mmol, 1 equiv) was added, resulting in immediate formation of a
bright magenta-purple color. The reaction mixture was allowed to
stir an additional 10 min. At this point the reaction mixture was
concentrated, and n-pentane was added to help precipitate the salt.
The reaction mixture was filtered through a plug of Celite, and the
volatile components were concentrated to 5 mL. Et2O (3 mL) was
added, and the purple solution was placed in the -35 °C freezer to
induce precipitation. After 30 min, a copious magenta precipitate
had formed. This precipitate was isolated atop a frit, resulting in
198 mg (0.183 mmol) of material (80% yield). X-ray diffraction-
quality crystals were grown from an Et2O/CH2Cl2 (1:1) solution at
-35 °C. 1H NMR (20 °C, acetone-d6, 500 MHz): δ ) 1.48 (15H,
s, Cp*-Me), 2.61 (2H, m, dppe-CH2), 2.75 (2H, m, dppe-CH2), 6.77
(m, 8H, Ar), 6.91 (m, 12H, Ar), 7.34 (br, 8H, Ar), 7.50 (br, 4H,
Ar), 7.61 (br, 8H, Ar). 31P{1H} NMR (20 °C, acetone-d6, 202.5
Treatment of AsP3 with Nb(H)(η2-tBu(H)CdNAr)(N[CH2tBu]-
Ar)3, Synthesis of 4-AsP and 4-P2. Nb(H)(η2-tBu(H)CdNAr)
t
(N[CH2 Bu]Ar)2 (113 mg, 0.17 mmol) was dissolved in 5 mL of
Et2O and was added to a vial containing AsP3 (10 mg, 0.06 mmol)
and a stir bar. The reaction mixture was vigorously stirred for 90
min during which time the originally yellow solution took on a
deep-green color. The volatile components of the reaction mixture
were removed under reduced pressure. The resulting green powder
was taken up in C6D6 for NMR analysis. Recrystallization from
Et2O/C5 (1:2) afforded 98 mg (82% yield) of a 1:1 mixture of (µ2:
η2,η2-P2)[Nb(N[CH2 Bu]Ar)3]2 and (µ2:η2,η2-AsP)[Nb(N[CH2-
t
tBu]Ar)3]2. 1H NMR (20 °C, benzene-d6, 500 MHz): δ ) 6.98 (24H,
br, o-Ar), 6.59 (12H, s, p-Ar), 4.23 (24H, br, N-CH2), 2.25 (36H,
s, Ar-CH3, µ2-P2), 2.23 (36H, s, Ar-CH3, µ2-AsP), 0.97 (54H, s,
t
tBu, µ2-P2), 0.95 (54H, s, Bu, µ2-AsP). 31P{1H} NMR (20 °C,
benzene-d6, 202.5 MHz): δ ) 399 (2P, br, µ2-P2), 438 (1P, br, µ2-
AsP). 13C{1H} NMR (20 °C, benzene-d6, 125.7 MHz): δ ) 154.2
(Ar), 138.1 (Ar), 126.6 (Ar), 124.4 (Ar), 73 (N-CH2), 37.3 (C(CH3)),
30.3 (C(CH3), µ2-P2), 30.1 (C(CH3), µ2-AsP), 22.1 (Ar-CH3).
Treatment of AsP3 with Mo(N[tBu]Ar)3, Synthesis of 5-As
and 5-P. Mo(N[tBu]Ar)3 (100 mg, 0.16 mmol) was dissolved in 5
mL of Et2O and was added to a vial containing solid AsP3 (8 mg,
0.047 mmol). The mixture was stirred for 60 min during which
time the reaction mixture assumed a dark-orange color. The volatile
components of the reaction mixture were removed under reduced
pressure. The resulting residue was dissolved in C6D6 and was taken
for NMR analysis. Recrystallization from Et2O/n-pentane (1:2)
afforded 75 mg (70% yield) of 3:1 PMo(N[tBu]Ar)3 and AsMo-
1
MHz): δ ) -450 (2P, d, J ) 245 Hz, non-Fe bound P), -261
(1P, tt, 1J ) 245 Hz, 2J ) 37 Hz, Fe-bound P), 89 (2P, d, 2J ) 37
Hz, dppe-P). 13C{1H} NMR (20 °C, acetone-d6, 125.8 MHz): δ )
10.6 (s, Cp*-Me), 29.5 (m, dppe-CH2), 90.7 (s, Cp*-ring), 122.8
(Ar), 126.6 (Ar), 129.6 (Ar), 131.7 (Ar), 133.2 (Ar), 137.6 (Ar),
164.9 (Ar), 165.7 (Ar). UV-vis: 494 nm (ε 590 M-1 cm-1), 543
nm (ε 480 M-1 cm-1). MALDI-TOF MS: 757.0318 m/z
([AsP3FeCp*(dppe)]+), 589.2216 m/z ([FeCp*(dppe)]+), 319.1650
m/z ([BPh4]-). Elem. Anal. Calcd for C60H59AsBFeP5: C 66.94, H
5.52, P 14.39; Found: C 66.87, H 5.61, P 13.98.
AsP3(P(N(iPr)2)N(SiMe3)2)2, 2. (P(N(iPr)2)N(SiMe3)2)2 (190 mg,
0.295 mmol) was dissolved in 10 mL of toluene and was added
slowly to a vial containing a solution of AsP3 (54 mg, 0.321 mmol)
in 5 mL of toluene. Upon mixing the two colorless reagents, the
reaction mixture took on a vibrant yellow color. The mixture was
stirred for an additional 20 min, and an aliquot was withdrawn for
1
(N[tBu]Ar)3. H NMR (20 °C, benzene-d6, 500 MHz): δ ) 6.61
(6H, s, o-Ar), 5.81 (3H, br, p-Ar), 2.04 (18H, s, Ar-Me), 1.66 (27H,
s, tBu). 31P{1H} NMR (20 °C, benzene-d6, 202.5 MHz): δ ) 1216
(1P, br, Pt Mo). 13C{1H} NMR (20 °C, benzene-d6, 125.7 MHz):
δ ) 150.6 (Ar, Pt Mo), 150.0 (Ar, Ast Mo), 136.85 (Ar, Pt Mo),
9
15510 J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009

Synthesis and Structural and Electronic Properties of AsP3
A R T I C L E S
136.80 (Ar, Ast Mo), 130.6 (Ar, Ast Mo), 130.4 (Ar, Pt Mo),
127.7 (Ar), 60.6 (C(CH3), Pt Mo), 59.8 (C(CH3), Ast Mo), 34.1
(C(CH3), Ast Mo), 33.9 (C(CH3), Pt Mo), 21.5 (Ar-CH3).
Reduction of Cl2Nb(ODipp)3 in the Presence of AsP3 and
Na/Hg Followed by Treatment with AsCl3, Generation of
6-P3, 6-AsP2, 6-As2P, 6-As3, AsP3, As2P2, As3P, and As4.
Cl2Nb(ODipp)3 (278 mg, 0.40 mmol) was combined with solid AsP3
(67 mg, 0.40 mmol) and then dissolved in 10 mL of THF. After
all the AsP3 was dissolved a 0.5% Na/Hg amalgam was added.
The reaction mixture was allowed to stir for 20 min during which
time it assumed a dark orange-brown color. The reaction mixture
was taken to dryness. To the residue was added 5 mL of Et2O. The
resulting residue was filtered to remove any insoluble material. The
filtrate was taken to dryness under reduced pressure. The resulting
residue was dissolved in a stock solution of P4 (0.158 mmol total
as an internal standard) in C6D6 for NMR analysis. 31P NMR (20
°C, benzene-d6, 202.5 MHz): δ ) -520 (P4, s, 0.1581 mmol), -206
([Na][P3Nb(ODipp)3], s, 0.1935 mmol, 48% yield), -167 ([Na][P2-
AsNb(ODipp)3], s, 0.090 mmol, 23% yield), -132 ([Na][PAs2-
Nb(ODipp)3], s, 0.0198 mmol, 5% yield). The total yield of 6-P3,
6-AsP2, and 6-As2P is 76% based on Cl2Nb(ODipp)3, which is
consistent with the yields obtained for the synthesis of 6-P3 from
P4.3 The NMR tube was returned to the glovebox, and its contents
were taken to dryness. The resulting residue was dissolved in 5
mL of THF, and the solution was frozen in the cold well. AsCl3
(72 mg) was added to this thawing solution (as a stock solution in
toluene). The resulting mixture was allowed to stir for 20 min after
which time it was filtered through a pad of Celite and concentrated
to 1 mL. This sample was placed in an NMR tube (wrapped in Al
foil) for analysis. 31P{1H} NMR (20 °C, benzene-d6, 202.5 MHz):
δ ) -520 (P4, s), -484 (AsP3, s), -452 (As2P2, s), -432 (AsP3,
s). A portion of the sample was diluted 100-fold with C6H6 and
was analyzed by GC-MS. GC-MS: P4 [retention time 7.8 min,
parent ion 124 m/z with fragments at 93 (P3) and 62 (P2), total area
50994], AsP3 [retention time 8.9 min, parent ion 168 m/z with
fragments at 137 (AsP2), 106 (AsP), 93 (P3), 75 (As), and 62 (P2),
total area 30475, 49% yield from 6-P3], As2P2 [retention time 9.8
min, parent ion 212 m/z with fragments at 181 (As2P), 150 (As2),
137 (AsP2), 106 (AsP), 75 (As), and 62 (P2), total area 11348,
39% yield from 6-AsP2], As3P [retention time 10.6 min, parent ion
256 m/z with fragments at 225 (As3), 181 (As2P), 150 (As2), 106
(AsP), and 75 (As), total area 3176, 50% yield from 6-As2P], As4
[retention time 11.2 min, parent ion 300 m/z with fragments at 150
(As2) and 75 (As); peaks from neighboring HODipp signal also
present, approximate total area 1977, 2% yield from Cl2Nb-
(ODipp)3].
they are linked to (1.5 times for methyl groups). All disorders were
refined within SHELXL with the help of rigid bond restraints as
well as similarity restraints on the anisotropic displacement
parameters for neighboring atoms and on 1,2- and 1,3-distances
throughout the disordered components. The relative occupancies
of disordered components were refined freely within SHELXL.
Further details are available in the Supporting Information (Table
S1 and a CIF file) and from the CCDC under deposition numbers
730555 (2) and 730556 (1).
Computational Studies. All calculations were carried out using
ADF 2008.01 from Scientific Computing and Modeling (http://
www.scm.com)20,21 on a 32-processor Quantum Cube workstation
from Parallel Quantum Solutions (http://www.pqs-chem.com). All
of the calculations were repeated using the OLYP functionals which
are a combination of the OPTX exchange functional of Handy and
Cohen used with Lee, Yang, and Parr’s nonlocal correlation function
(LYP).66-68 In addition, all calculations were carried out using the
zero-order regular approximation (ZORA) for relativistic effects.69-71
The basis sets used were quadruple-ꢁ with four polarization
functions (QZ4P) for As and P in the calculation of E4 tetrahedra
(E ) As, P) and triple-ꢁ with two polarization functions (TZ2P)
for all other calculations, as supplied with ADF. Chemical shielding
tensors were calculated for the 31P nuclei in the optimized structures
by the GIAO method using the ADF package.72-75 The functionals,
basis sets, and relativistic approximations used were the same as
those described above. The isotropic value of the chemical shielding
was converted to a chemical shift downfield of 85% phosphoric
acid using P4, white phosphorus, as a computational reference; the
computed absolute shielding value of P4 was correlated with a
chemical shift equal to its experimental value (-520 ppm) in dilute
benzene solution.76 For zero-point energy corrections, frequencies
calculations were computed using optimized structures as described
above.
Acknowledgment. We thank the NSF (Grant CHE-719157) and
ThermPhos International for generous funding and support for this
project.
Supporting Information Available: Full crystallographic data
on compounds 1 and 2, complete ADF output files, and
phosphorus NMR spectra for all reactions. This material is
available free of charge via the Internet at http://pubs.acs.org.77
JA906294M
(64) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, A46, 467–473.
(65) Sheldrick, G. M. SHELXL-97: Program for crystal structure deter-
mination; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(66) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789.
(67) Handy, N. C.; Cohen, A. J. Mol. Phys. 2001, 99, 403–412.
(68) Baker, J.; Pulay, P. J. Comput. Chem. 2003, 24, 1184–1191.
(69) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597–4610.
X-ray Diffraction Studies. Diffraction-quality crystals of
[(AsP3)FeCp*(dppe)][BPh4] were grown from CH2Cl2/Et2O at -35
°C over the course of several days. Diffraction-quality crystals of
AsP3(P(N(iPr)2)N(SiMe3)2)2 were grown from n-hexane/Et2O by
cooling at -35 °C over the course of two days. The crystals were
mounted in hydrocarbon oil on a nylon loop. Low-temperature (100
K) data were collected on a Siemens Platform three-circle diffrac-
tometer coupled to a Bruker-AXS Smart Apex CCD detector with
graphite-monochromated Mo KR radiation (λ ) 0.71073 Å)
performing ꢀ and ω scans. A semiempirical absorption correction
was applied to the diffraction data using SADABS.63 The structures
were solved by direct methods using SHELXS64 and refined against
F2 on all data by full-matrix least-squares with SHELXL-97.65 All
non-H atoms were refined anisotropically. All H atoms were
included in the models at geometrically calculated positions and
refined using a riding model. The isotopic displacement parameters
of all H atoms were fixed to 1.2 times the U value of the atoms
(70) van Lenthe, E.; Ehlers, A.; Baerends, E. J. Chem. Phys. 1999, 110,
8943–8953.
(71) van Lenthe, E.; Baerends, E.; Snijders, J. J. Chem. Phys. 1994, 101,
9783–9792.
(72) Schreckenbach, G.; Ziegler, T. J. Phys. Chem. 1995, 99, 606–611.
(73) Schreckenbach, G.; Ziegler, T. Int. J. Quantum Chem. 1997, 61, 899–
918.
(74) Wolff, S. K.; Ziegler, T. J. Chem. Phys. 1998, 109, 895–905.
(75) Wolff, S. K.; Ziegler, T.; van Lenthe, E.; Baerends, E. J. J. Chem.
Phys. 1999, 110, 7689–7698.
(76) van Wullen, C. Phys. Chem. Chem. Phys. 2000, 2, 2137–2144.
(77) CIF files are available from the Cambridge Crystallographic Database
under deposition numbers 730555 (2) and 730556 (1).
9
J. AM. CHEM. SOC. VOL. 131, NO. 42, 2009 15511