Formation of cationic [RP5Cl]+-cages via insertion of [RPCl]+-cations into a P-P bond of the P4 tetrahedron

Article abstract of DOI:10.1021/ic2013304

Fluorobenzene solutions of RPCl₂ and a Lewis acid such as ECl₃ (E = Al, Ga) in a 1:1 ratio are used as reactive sources of chlorophosphenium cations [RPCl]⁺, which insert into P-P bonds of dissolved P₄. This general protocol represents a powerful strategy for the synthesis of new cationic chloro-substituted organophosphorus [RP ₅Cl]⁺-cages as illustrated by the isolation of several monocations (21a-g⁺) in good to excellent yields. For singular reaction two possible reaction mechanisms are proposed on the basis of quantum chemical calculations. The intriguing NMR spectra and structures of the obtained cationic [RP₅Cl]⁺-cages are discussed. Furthermore, the reactions of dichlorophosphanes and the Lewis acid GaCl₃ in various stoichiometries are investigated to obtain a deeper understanding of the species involved in these reactions. The formation of intermediates such as RPCl ₂?GaCl₃ (14) adducts, dichlorophosphanylchlorophosphonium cations [RPCl₂-RPCl]⁺ (16⁺) and [RPCl₂-RPCl-GaCl₃]⁺ (17⁺) in reaction mixtures of RPCl₂ and GaCl₃ in fluorobenzene strongly depends on the basicity of the dichlorophosphane RPCl₂ (R = tBu, Cy, iPr, Et, Me, Ph, C₆F₅) and the reaction stoichiometry.

Full text of DOI:10.1021/ic2013304

Article
pubs.acs.org/IC
Formation of Cationic [RP5Cl]+-Cages via Insertion of [RPCl]+-Cations
into a P−P Bond of the P4 Tetrahedron
Michael H. Holthausen,† Kai-Oliver Feldmann,†,‡ Stephen Schulz,† Alexander Hepp,†
and Jan J. Weigand*,†
‡
†Institut fur Anorganische and Analytische Chemie and Graduate School of Chemistry, Westfaelische Wilhelms Universitaet
̈
Muenster, Corrensstrasse 30, 48149 Muenster, Germany
S
* Supporting Information
ABSTRACT: Fluorobenzene solutions of RPCl2 and a Lewis acid such as ECl3 (E = Al, Ga)
in a 1:1 ratio are used as reactive sources of chlorophosphenium cations [RPCl]+, which insert
into P−P bonds of dissolved P4. This general protocol represents a powerful strategy for the
synthesis of new cationic chloro-substituted organophosphorus [RP5Cl]+-cages as illustrated
by the isolation of several monocations (21a-g+) in good to excellent yields. For singular reac-
tion two possible reaction mechanisms are proposed on the basis of quantum chemical
calculations. The intriguing NMR spectra and structures of the obtained cationic [RP5Cl]+-
cages are discussed. Furthermore, the reactions of dichlorophosphanes and the Lewis acid
GaCl3 in various stoichiometries are investigated to obtain a deeper understanding of the
species involved in these reactions. The formation of intermediates such as RPCl2·GaCl3 (14)
adducts, dichlorophosphanylchlorophosphonium cations [RPCl2−RPCl]+ (16+) and [RPCl2−
RPCl−GaCl3]+ (17+) in reaction mixtures of RPCl2 and GaCl3 in fluorobenzene strongly depends on the basicity of the
dichlorophosphane RPCl2 (R = tBu, Cy, iPr, Et, Me, Ph, C6F5) and the reaction stoichiometry.
has been extended to highly functionalized cation 7+ and the
INTRODUCTION
■
zwitterionic P5-cage 8, which are formed by the insertion of the
corresponding phosphenium cations that are derived from four-
membered phosphorus−nitrogen-metal heterocycles (Chart 2).7
Anionic and neutral phosphorus cage and ring compounds have
been thoroughly investigated.1 However, the diversity of
cationic polyphosphorus derivatives is mostly limited to
catenated cyclic or acyclic polyphosphanylphosphonium salts.2
Cationic [P5X2]+-cages 1+ (X = Cl, Br, I) were reported only
recently. Cations 1+ resulted from the reaction of the silver salt
Ag[Al{OC(CF3)3}4] with PX3 (X = Cl, Br, I) in the presence of
P4.3 Applying a stoichiometric melt approach at elevated
temperatures (60 to 100 °C),4 we were able to extend this
chemistry by intercepting the more stable diorganophosphe-
nium cation [Ph2P]+ with P4. This approach resulted in the
formation of the cationic cages [Ph2P5]+ (2+), [Ph4P6]2+ (32+),
and [Ph6P7]3+ (43+) via the consecutive insertion of up to three
[Ph2P]+ into the P−P bonds of the P4 tetrahedron (Chart 1).5
Chart 2. Cationic Phosphorus Cages from Four-Membered
Heterocycles
Chart 1. Cationic Phosphorus Cages
In continuation of our investigations, we became interested
in protocols for the generation of functionalized P5-cages.
Similarly, the reaction of P4 with the cyclo-diphosphadiazane
[DippNPCl]2 (Dipp = 2,6-diisopropylphenyl) and GaCl3 in
fluorobenzene yielded compound 5[GaCl4]. The addition of an
excess GaCl3 and a second equiv of P4 to a solution of 5[GaCl4]
afforded the dicationic species 62+ as [Ga2Cl7]− salt.6 This approach
Received: June 21, 2011
Published: February 28, 2012
© 2012 American Chemical Society
3374
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Herein, we report a convenient, high yielding one-pot
procedure for the preparation of a range of cationic [RP5Cl]+-
cages. The reaction proceeds formally via chlorophosphe-
nium ions [RPCl]+ that are generated in situ from
dichlorophosphanes and a Lewis acid ECl3 (E = Al, Ga) in
fluorobenzene solution. The subsequent reaction with P4
results in the precipitation of analytically pure products. The
intriguing NMR spectra and molecular structures of the
obtained cationic [RP5Cl]+-cages are discussed. Furthermore,
we present an investigation of the reactions of dichlorophos-
phanes and the Lewis acid GaCl3 in various stoichiometries
which provides a deeper understanding of the species that are
involved in these reactions. To gain an in depth understand-
ing about the underlying reaction mechanisms of the formal
“insertion reaction” we carried out quantum chemical
calculations.
Chart 3. Potential Products from the Reaction of
Chlorophosphanes and Lewis Acids ECl3 (E = Al, Ga)
a
a
Anions are not displayed.
RESULTS AND DISCUSSION
■
The reactions of monochlorophosphanes R2PCl and Lewis
acids ECl3 (E = Al, Ga) in organic solvents have been
previously investigated.8 The expected Lewis acid−base
complexes 10 were observed in these reaction mixtures, but
other compounds were also detected.8 These include
homoatomic coordination complexes of phosphenium
cations, that is, phosphanylphosphonium salts 11+ featuring
P−P bonds,9 and unstabilized cations [R2P]+ (9+).10,2 Two-
coordinate phosphenium cations 9+ are main group carbene
analogues that feature six valence electrons on the central
atom.11 Because of their inherent electrophilicity, most of the
phosphenium ions are highly reactive and transient species.
This hampers their direct observation in solution. However,
structural characterization is possible in solution12 and solid
state13 when the phosphenium cation is stablized by delocalization
of the positive charge. The reactions of dichlorophosphanes
RPCl2 (13) and Lewis acids ECl3 (E = Al, Ga) in organic
solvents have been studied to a much lesser extent. The
formation of Lewis acid−base complexes mRPCl2·nECl3 (E =
Al, m = 1, 2; n = 1, 2)14 has been suggested, although the Lewis
basicity of dichlorophosphanes is much lower in comparison to
the corresponding monochlorophosphanes.15 Such adducts
were used for the in situ formation of phosphenium cations
[RPCl][AlCl4] and the subsequent synthesis of phosphorus
heterocycles.16 Mixtures of mono- and dichlorophosphanes in
the presence of Lewis acids result in the formation of
chlorophosphanylchlorophosphonium cations 15+, which were
spectroscopically characterized by 31P NMR for a variety of
different substituents. In most cases, the spectra are sufficiently
resolved at room temperature to detect the 1JPP coupling in the
chlorophosphanylchlorophosphonium cations 15+.17 However,
mixtures of dichlorophosphanes and a Lewis acid in CH2Cl2
show broad peaks in the 31P NMR spectra. This indicates
dynamic dissociation of the coordinated P−P bonds of the
corresponding chlorophosphanyldichlorophosphonium cations
16+ because of the weak donor abilities of dichlorophos-
phanes.17
of different neutral and cationic compounds are formed which
in some cases show dynamic exchange. We start our discussion
with the influence of various stoichiometries on the
composition. The 31P{1H} NMR spectra of reaction mixtures
of the dichlorophosphanes 13a−g and GaCl3 in fluorobenzene
depend strongly on the ratio of the reactants. Table 1
summarizes the chemical shifts and coupling constants for
the observed species in solution18 for the stoichiometries 2:1,
1:1, 1:2, and 1:3 of dichlorophosphanes 13a−g and GaCl3 at
room temperature.19 Depending on the substituent R and the
reaction stoichiometry, the formation of several structurally
distinct species as well as some decomposition products as a
result of solvent activation are observed (Chart 3). Figure 1
displays the 31P{1H} NMR spectra of the reaction mixtures of
the selected dichlorophosphanes 13a,b,f and GaCl3 in the
mentioned stoichiometries. Similar to previous studies,17 the
classical 1:1 complexes RCl2P−GaCl3 (14) are observed in
reaction mixtures involving alkyl substituted dichlorophos-
phanes with reaction stoichiometries of RPCl2 (13a−e) and
GaCl3 in the range from 1:1 to 1:3. A second possibility of
adduct formation is described by the coordination of the
gallium atom to one of the chlorine substituents of RPCl2
forming compound RClP−Cl−GaCl3 (14′). Similar species
have been described as phosphenium ion equivalents within the
system Ag+/PX3 fomally delivering a PX2 cation.20 However,
+
we found that adduct 14a is favored by ΔG298.15 K = 10.9 kcal/mol
compared to adduct 14a′ according to quantum chemical
calculations (MP2/TZVP, see Supporting Information for
details). Thus, complex 14a is essentially the only product in
the 1:1 reaction mixture of 13a and GaCl3. The Lewis
acid−base adducts 14a-e show a consistent upfield shift of their
31P NMR signals by approximately 60 ppm to higher field with
respect to the free phosphanes 13a−e. Similar high-field shifts
were reported for the structurally related monochlorophos-
phane adducts 10.21 We were able to isolate and structurally
characterize compound 14a (Figure 2) from the 1:1 reaction of
tBuPCl2 with 1 equiv of GaCl3 in fluorobenzene. Diffusion of
n-hexane into the reaction mixture at −35 °C yielded small
amounts of adduct 14a as an extremely air and moisture
sensitive, crystalline material that starts to melt above −25 °C.
Compound 14a crystallizes in the monoclinic space group P21/n
with four formula units in the unit cell. The compound
represents the first structurally characterized Lewis acid-base
adduct of GaCl3 and a dichlorophosphane. The P−Ga bond
31P NMR Investigation of RPCl2/GaCl3 Mixtures. We
have now performed a systematic investigation of the reaction
of dichlorophosphanes RPCl2, 13a−g (a: R = tBu, b: Cy, c: iPr,
d: Et, e: Me, f: Ph, g: C6F5) and ECl3 (E = Al, Ga) in
fluorobenzene, which we use as solvent in this study because of
its beneficial properties during the synthesis of our phosphorus
cages (vide infra). Our results show that, depending on the
ratio of the reactants and the organic group on RPCl2, mixtures
3375
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
a
Table 1. 31P NMR Data and Signal Assignments in Mixtures of Dichlorophosphanes (13a−g) and GaCl3 in Fluorobenzene
c
d
e
stoichiometry
RPCl2:GaCl3
RPCl2−GaCl3 [RPCl2−PRCl]+
[RPCl2−PRCl−
GaCl3]+ (17+)
[RPCl2H]+
[RArFPCl−PRCl]+
[RArFPClH]+
RPCl2
R
(14)
(16+)
(18+)
(19+)
(20)
(13a-g)
tBu
2: 1
1: 1
1: 2
1: 3
2: 1
1: 1
1: 2
1: 3
2: 1
1: 1
1: 2
1: 3
165 (14)
141 (20)
139 (22)
139 (20)
163 (97)
130 (35)
129 (35)
128 (35)
170 (55)
165 (77)
135 (25)
134 (19)
199
99 [630]
99 [630]
68 [550]
Cy
102 (140)
198
200
134, 99 [∼490]
134, 99 [498]
134, 99 [498]
105 (205)
89 [650]
89 [645]
89 [645]
57 [550]
57 [550]
iPr
101 (110)
141, 104 [503]
141, 104 [503]
95 [650]
95 [650]
62 [560]
62 [560]
99, 85 [420];
93, 80 [420]
Et
2: 1
1: 1
1: 2
1: 3
2: 1
1: 1
1: 2
138 (45), 104 (280)
133 (131), 100 (85)
196
192
b
132 (172)
131 (71)
136, 100 [∼420]
136, 100 [∼460]
91 [670]
91 [670]
92, 79 [390]
92, 79 [390]
55 [570]
55 [570]
b
Me
157 (67), 98 (106)
130 (40), 97 (46)
130 (420), 97 (420)
127 (157)
127 (130)
85 [695]
85 [695]
83, 73 [370];
83, 75 [368]
47 [570]
47 [570]
b
1: 3
130, 97 [∼400]
83, 73 [370];
83, 75 [369]
Ph
2: 1
1: 1
1: 2
139 (87), 85 (99)
111 (35), 85 (31)
110 (425), 85 (410)
163
69 [695]
69 [695]
70, 58 [393];
71, 58 [395]
b
1: 3
110, 85 [∼430]
70, 57 [395];
70, 58 [397]
C6F5
2: 1
1: 1
1: 2
1: 3
22 [620]
22 [620]
22 [620]
22 [620]
135
a
b
1
31P{1H}/31P NMR-Data are given as chemical shifts are reported in ppm JPP values [in parentheses] and half widths (in brackets) in Hz. Very
c
d
e
broad resonances. Two sets of AX spin systems are observed, resulting from ArF = o,p-C6H4F. Only small amounts observed. Chemical shift of
the respective dichlorophosphane (13a−g) in fluorobenzene.
length [Ga1−P1 2.4252(9) Å] in 14a is typical for phosphane-
gallane complexes [2.35−2.68 Å]22 and comparable to those of
related complexes such as Cl3Ga−Ph2PPPh2−GaCl3 [2.4285(6)
Å].4 Moreover, the phosphorus atom in 14a shows a distorted
tetrahedral environment [Ga1−P1−C1 121.2(1)°] because of
the presence of the sterically demanding tBu group, which adopts
a staggered conformation with the GaCl3 moiety.
It is noteworthy that, while the 1:1 to 1:3 mixtures of RPCl2
(13a-f) and GaCl3 gave distinct species in the 31P NMR spectra, the
2:1 reactions showed broad signals (Figure 1). Thus, the 2:1
reaction mixture of 13a and GaCl3 gave rise to a broadened
resonance at 165 ppm with a line width of ν1/2 = 14 Hz. We
attribute the observed line broadening to a fast ex-
change process in solution (eq 1) between 13a (δ = 199 ppm)
and 14a (δ = 141 ppm).
GaCl3 in 2:1, 1:1, 1:2 and 1:3 ratios. As observed for the 2:1
mixture of 13a and GaCl3, the 2:1 reaction mixture of 13b and
GaCl3 shows a broad resonance (δ = 163 ppm, ν1/2 = 97 Hz) as
the main species. However, an additional broad signal (δ = 102
ppm, ν1/2 = 140 Hz) with low intensity indicates a second
dynamic process. We believe that this observation may be
attributed to the very fast and reversible reaction of 14b and
[RPCl][GaCl4] to form species 16b+ (eq 2). Compounds of
type 14 and 16+ have previously been proposed as structural
alternatives that form equilibria in solution, and their relative
stabilities depend upon the nature of the substituent (R) as well
as reaction stoichiometry.21
Reducing the steric bulk and the basicity of the used
dichlorophosphanes from tBu (13a) to methyl (13e) results in
significant changes in the 31P{1H} NMR spectra of the respective
reaction mixtures (Table 1). Figure 1 (middle) displays the 31P
NMR spectra of solutions which contained CyPCl2 (13b) and
Phosphanylphosphonium salts 16+ may be generated by the
in situ formation of Lewis acidic phosphenium cations (12+; eq 3)
and subsequent coordination of a dichlorophosphane ligand
according to eq 2.2 This process is repressed by the equilibrium
dissociation of [GaCl4]− to yield chloride as well as the excess of
3376
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Figure 1. 31P{1H} NMR spectra (in fluorobenzene solution, C6D6-capillary, 27 °C, 81.01 MHz) of reaction mixtures of RPCl2 and GaCl3 in various
stoichiometries; top: R = tBu; middle: R = Cy; bottom: R = Ph. Small amounts of side products are indicated by an asterisk.
dichlorophosphanes which we propose undergo nucleophilic
substitution reactions that lead to the dissociation of the formed
complexes. The appearance of only one broad resonance for 16b+
again emphasizes this fast exchange process.23 The addition of
excess GaCl3 suppresses nucleophiles, therefore, the 1:1 reaction
mixture of CyPCl2 and GaCl3 shows, as a minor component, the
two doublets (AB spin system, δA = 99, δB = 134 ppm; 1JPP = 490
Hz) that are typical for phosphanlyphosphonium cations. The
donor−acceptor complex 14b remains the main species in the
reactions mixtures of CyPCl2 and GaCl3. However, this situation
changes drastically when the basicity and steric requirements of
the dichlorophosphanes are further reduced. Reaction mixtures of
the less bulky EtPCl2 and MePCl2 with GaCl3 exhibit the
phosphanlyphosphonium complexes 17d,e+ as main prod-
ucts, and the donor−acceptor complexes 14d,e are now the
minor species. Cations 17d,e+ are probably formed from
14d,e and 12+ according to eq 4. They were identified by two
doublets in the 31P{1H} NMR spectra that are significantly
broadened because of the influence of the quadrupole
moment of the attached gallium atom as well as fast exchange
processes.21
Similar to Cy and tBu derivatives 13a and 13b, the 31P NMR
spectra of the 2:1 mixtures of 13d, 13e, and 13f and GaCl3
3377
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Figure 2. ORTEP plot of the molecular structure of 14a. Thermal
ellipsoids at 50% probability (hydrogen atoms are omitted for clarity).
Selected bond lengths (Å) and angles (°): P1−C1 1.828(4), P1−Cl1
1.999(1), P1−Cl2 1.995(1), Ga1−P1 2.4252(9); C1−P1−Cl2
106.7(2), C1−P1−Cl1 106.6(1), Cl2−P1−Cl1 105.21(6), C1−P1−
Ga1 121.2(1), Cl2−P1−Ga1 109.22(5), Cl1−P1−Ga1 106.83(5).
display two broad singlet resonances that result from the
already discussed equilibria 1 and 4 in solution (eq 1 and 4).
However, in the 1:1, 1:2, and 1:3 mixtures of 13f and GaCl3,
the related adduct 14f is not observed in the 31P{1H} NMR
spectra (Figure 1, bottom). In this case equilibrium 4 (eq 4)
dominates the reaction, and the main observed species is 17f+.
The observed trend in the 1JPP couplings in derivatives 16+ and
17+ (Table 1) is explained by the σ-donor strength of the
dichlorophosphanes,24 which is influenced by inductive as well
as steric effects of the substituent (for both criteria: Ph < Me <
Et < iPr < Cy <tBu).25 No reaction was observed in the
31P{1H} NMR spectra between dichlorophosphane 13g (R =
C6F5) possessing a very low basicity and GaCl3. Phosphonium
cations [RPCl2H]+ (18a−g+) are observed in all reactions of
dichlorophosphanes and GaCl3 were the ratio is at least 1:1
(R = Cy) or an excess of GaCl3 is used (R = tBu, iPr, Et, Me, Ph)
(Table 1). The proton-coupled 31P NMR spectra of these
mixtures show an additional splitting of the corresponding
resonance as a consequence of the one-bond coupling (1JPH
ranges from 620−695 Hz) of the hydrogen atom to the
phosphorus atom in derivatives of 18+. They presumably
originate in the reaction mixtures with an excess of GaCl3 by
direct Friedel−Crafts type arylation reactions of the dichloro-
phosphanes RPCl2. This results in the formation of
chlorophosphanes RArFPCl and one equivalent HCl. Because
of the higher Lewis basicity of chlorophosphanes compared to
dichlorophosphanes they are mainly observed in the reaction
mixtures as [RArFPClH]+ (20a−g+, ArF = o,p-C6H4F).26,27
Furthermore the formation of the corresponding phosphanyl-
phosphonium salts [RArFPCl−PRCl]+ (19c−f+) are observed
for less σ-donating organic substituents R. The latter species
result from the reaction of RPCl2 and GaCl3 with RArFPCl and
are mainly observed in the 1:2 and 1:3 reaction mixtures. The
amount of 18+, 19+, and 20+ increases drastically with
prolonged reaction times. In the case of 13g the only side-
product observed for all reactions stoichiometries is cation
[(C6F5)ArFPClH]+ (20g+). The observed amount of this
product significantly increases in reaction mixtures with an
excess of GaCl3 (Table 1).28
Figure 3. Raman spectra below 450 cm−1 of the reaction mixtures of
RPCl2 and GaCl3 in various stoichiometries in fluorobenzene.
suggested equilibrium between adducts 14a,f and the
corresponding dichlorophosphanes (eq 1). Furthermore, all
spectra display a series of modes that are associated with
Cl3Ga−P vibrations [R = tBu: 434, 406, 310, 199, 160 cm−1;
R = Ph: 441, 405, 320, 207 cm−1], and are therefore expected for
adducts 14a,f and phosphanylphosphonium complexes 17f +
(eq 4). In the Raman spectrum of the 2:1 mixture of PhPCl2
and GaCl3 the characteristic band at 346 cm−1, corresponding
to the A1 vibration mode of the tetrahedral [GaCl4]− species, is
observed. This band significantly decreases in the 1:1, 1:2 and
1:3 spectra, while bands [399, 287 cm−1] appear, which can be
attributed to the [Ga2Cl7]− anion. The presence of the
[Ga3Cl10]− anion can be assumed since there are significant
broader bands observed in the 1:2 and 1:3 mixture at
approximately 390 cm−1 and 265 cm−1. In addition, [GaCl4]−
anions are still present for the mentioned 1:2 and 1:3
stoichiometries. On the basis of these results, we suggest the
equilibrium outlined in eq 5.29
In contrast to the case of PhPCl2 (13f), no bands are
observed in the Raman spectra of tBuPCl2 (13a) and GaCl3
which can be attributed to the presence of the gallate anions
[GaCl4]−, [Ga2Cl7]−, or [Ga3Cl10]−. Furthermore, for mixtures
containing either 13a or 13f around 400 cm−1 and 165 cm−1
are observed, which we attribute to Ga2Cl6. This assignment is
supported by the fact that the intensity of these bands increases
with the concentration of GaCl3. All assignments are in agree-
ment with experimental data from the literature and are sum-
marized in Table 2.4,29,30
Raman Spectra of Selected RPCl2/GaCl3 Mixtures. Our
findings are further supported by Raman spectroscopy. The
Raman spectra of various reaction stoichiometries of tBuPCl2
(13a) and PhPCl2 (13f) with GaCl3 in fluorobenzene are
depicted in Figure 3. In both cases, characteristic bands of the
free dichlorophosphanes 13a,f are observed in the Raman
spectra of the 2:1 mixtures (Table 2). This is in accord with the
3378
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Table 2. Comparison of Bands in the Raman Spectra of Reaction Mixtures from RPCl2 Reaction Mixtures of RPCl2 (R = tBu,
Ph) and GaCl3 in Fluorobenzene at RT in Various Stoichiometries
a
a
R = tBu
R = Ph
b
2: 1
1: 1
1: 2
1: 3
2: 1
1: 1
1: 2
1: 3
assignment
455 (vw)
441 (vw)
418 (w)
RPCl2
d
434 (w)
415 (w)
434 (w)
415 (w)
434 (w)
415 (m)
434 (w)
415 (s)
441 (vw)
417 (w)
441 (vw)
414 (m)
399 (m)
c
439 (vw)
413 (s)
398 (m, sh)
c
νasym[GaCl2], δ[P−GaCl3]
Ga2Cl6, ν1[Ag]
Ga2Cl7−, νsym[Ga(Clt)3]
d
406 (vw, sh)
359 (vs)
406 (vw, sh)
359 (vs)
396 (w, sh)
359 (vs)
396 (m, sh)
359 (vs)
405 (vw)
370 (vs)
c
ν[P−Ga]
νsym[P−GaCl3]
d
370 (vs)
c
370 (vs)
c
369 (vs)
c
Ga2Cl7−, νasym[Ga(Clt)3]
GaCl4−, ν1[A1]
Ga2Cl6, 2ν3
346 (vs)
345 (m)
344 (m)
344 (m)
342 (m)
342 (m)
322 (m, sh)
310 (m)
322 (m, sh)
310 (m)
c
c
Ga2Cl6, ν2[Ag]
d
310 (w)
310 (m)
320 (w)
292 (m)
320 (w)
320 (m)
320 (s)
δ[P−Ga]
RPCl2
297 (w, sh)
287 (w, br)
245 (s)
287 (w, br)
245 (s)
287 (w, br)
245 (s)
Ga2Cl7−, δ[Ga−Clb−Ga]
PhF (solvent)
241 (s)
241 (s)
241 (s)
241 (s)
245 (s)
d
199 (w)
185 (vw)
160 (vw)
199 (m)
199 (m)
199 (m)
207 (m, sh)
197 (m, sh)
207 (m, br)
207 (m, br)
201 (m, br)
δ[P−Ga]
RPCl2
d
163 (vw)
164 (m)
166 (s)
162 (m)
163 (s)
165 (s)
Ga2Cl6, ν3[Ag], νsym[Ga−P]
a
Intensity and appearance of bands are indicated by the following abbreviations: s = strong, m = medium, w = weak, v = very, br = broad, sh =
b
shoulder, t = terminal, b = bridging. Assignments of bands for Ga2Cl6, GaCl4 , Ga2Cl7 , and GanCl3n+1− (n = 1−3) are based on data from the Ga/
GaCl3 system from ref 29. Not observed because of overlapping of bands. Assignments of bands have been carried out with the help of calculated
−
−
c
d
Raman spectra for adduct 14a (see Supporting Information for further details).
Weaker bands may be below the detection limit or may not
be observed because of overlapping bands. The situation is
further complicated by the fact that bands due to gallium(III)-
arene modes are expected to occur in the same spectral range.
Nevertheless, we believe that our findings strongly suggest that
the very basic tBuPCl2 (13a) exclusively forms adduct 14a with
GaCl3, whereas 1:1 mixtures of PhPCl2 (13f) and GaCl3 and
mixtures of 13f which contain an excess of GaCl3 give the
phosphanylphosphonium salt 17f+. A mixture of both types of
products is observed for the other combinations of
dichlorophosphane and GaCl3 (13b−e), in which the ratios
of the observed species in solution strongly depend on the
steric bulk and basicity of the phosphane. An in-depth Raman
spectroscopic and 31P NMR spectroscopic investigation of
mixtures of dichlorophosphanes and AlCl3 is hampered by the
low solubility of the latter in fluorobenzene. Binary systems of
RPCl2/AlCl3 in fluorobenzene are nevertheless reactive sources
for chlorophosphenium cation equivalents [RPCl]+ (eq 3) and
can be used for the generation of functionalized [RP5Cl]+-cages
21+ (vide infra).
Scheme 1. Reaction Pathway for the Formation of 21e[GaCl4]
via a One Step Phosphane Assisted Phosphenium Ion Insertion
Scheme 2. Reaction Pathway for the Formation of
21e[GaCl4] via a Two Step Mechanism Including the
Formation of Butterfly Type Intermediate 22
Quantum Chemical Calculations. According to recent
investigations,20,31 the formation of P5 cage compounds via free
phosphenium ions seemed unlikely. However, attempts to
calculate a feasible reaction mechanism from the adducts of
GaCl3 with RPCl2 (R = tBu, Me) as sources of phosphenium
ions and P4 to form P5 cage cations were unsuccessful at the
B3LYP/6-31G(d) and MP2/6-31G(d) levels of theory.
Therefore, the reaction of P4 with phosphanylphosphonium
salt 16e[GaCl4] as a source of a phosphenium ion was in-
vestigated at the B3LYP/6-31G(d) and subsequently refined
at the MP2/6-31G(d) level of theory. A single step inser-
tion of the phosphenium moiety into the P−P bond of the
P4 tetrahedron (Scheme 1) and a two step reaction path via
butterfly type compound 22 (Scheme 2) were found to be
viable.
Figure 4 shows the calculated energy profile of both reaction
paths. With a barrier of 21.1 kcal/mol (ΔH298, TS1) a single
step transfer of the phosphenium moiety in 16e[GaCl4] and
insertion thereof into a P−P bond of the P4 tetrahedron is
energetically viable (black path). In the light of recent
mechanistic studies on the reaction of an isoelectronic silylene
32
with P4 this is best understood as a combined electrophilic
and nucleophilic attack of the phosphenium moiety at P4. On
the one hand the P−P bond of P4 (highest occupied molecular
orbital, HOMO) nucleophilically attacks the p-type orbital of
the phosphenium moiety. On the other hand the lone pair
of electrons of the phosphenium moiety donates electron
density into the lowest unoccupied molecular orbital (LUMO) of
the P4 tetrahedron which corresponds to p-orbitals situated
3379
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Figure 4. Calculated reaction pathways for the reaction of 16e[GaCl4] and P4. Calculated differences of the enthalpies at 298.15 K (ΔH298) are given
for the B3LYP/6-31G(d) and MP2/6-31G(d) (in parentheses) optimized structures. The optimized structures of 16e[GaCl4] + P4 were defined as
0 kcal/mol.
perpendicular to the P4 lone pairs.32 It was found however, that a
lower barrier reaction path is accessible when the phosphenium
cation does not act as a nucleophile. Instead a chlorine
substituent of the tetrachlorogallate anion may nucleophilically
attack the P4 tetrahedron. Along with the electrophilic attack of
the phosphenium moiety on P4 this leads to the slightly
temperature. In general, highly pure compounds 21a−g[ECl4]
show low solubilities in 1,2-difluorobenzene, fluorobenzene,
and dichloromethane. However, addition of small amounts
(<5 mol-%) of Lewis acid ECl3 (E = Al, Ga) increases their
solubilities drastically.33 Compounds 21a−g[ECl4] are air,
moisture, and slightly light-sensitive. They readily decompose
in coordinating solvents such as MeCN.
endothermic formation of a butterfly type structure 22 (ΔH298
=
2.1 kcal/mol) via transition state TS2 (ΔH298 K = 9.9 kcal/mol,
red path). Intermediate 22 can then rearrange via TS3 (ΔH298 K
= 3.3 kcal/mol) to the P5-cage compound 21e[GaCl4] and
MePCl2. Calculations on the MP2/6-31G(d) level yielded
related pathways with higher but still reasonable barriers
(ΔH298 = 27.4 kcal/mol for TS1, 17.7 kcal/mol for TS2 and
16.3 kcal/mol for TS3). It is therefore most likely that the
formation of the cationic P5-cage compounds proceeds via a two-
step reaction with structures of type 22 as intermediates and may
be best described as an adduct formation, addition, and
subsequent rearrangement reaction.
Syntheses and Characterization of the P5 Cage
Compounds [21a−g][ECl4]. Dissolution of recrystallized P4
in 1:1 mixtures of dichlorophosphanes 13a−f and ECl3 (E = Al,
Ga) in fluorobenzene yielded white to yellowish precipitates of
21a−g[ECl4] (Scheme 3). After workup, which included
The salts are stable in substance for at least six month when
stored at low temperatures in the dark. The 31P{1H} NMR
spectra of the dissolved salts in dichloromethane at room
temperature are depicted in Figure 5. Iterative line shape
fitting36 of the observed spin systems yielded chemical shifts
and coupling constants typical for Cs-symmetric P5 cages with
four chemically unequal phosphorus nuclei (Table 3). Because
of the reduced symmetry of the [RP5Cl]+-cages compared to
the C2v-symmetric [R2P5]+-cages,3,5 ABM2X spin systems are
observed for cations 21a−d+ and ABMX2 spins systems are
detected for cations 21e−g+. Due to similar geometry of the P5
core of cations 21a−g+, the JPP and JPP coupling constants
deviate only marginally from each other (Table 3). However,
the chemical shifts are strongly dependent on the substituent
attached to the P5 cages. The values of Taft’s σ* parameter37
for the alkyl and aryl groups (Table 3) show an approximate
linear correlation with the chemical shift of the respective
phosphorus atom (Figure 6). The differences in chemical shifts
are strongest for the tetra-coordinate phosphonium moieties
(marked by a red dot, Figure 5), which show shifts ranging from
99 ppm (21a+, R = tBu) to 26 ppm (21g+, R = C6F5). A
reverse trend is observed for the resonances of the chemically
equivalent bridging phosphorus atoms, which range from 44 to
83 ppm (marked by a blue dot, Figure 5). The tri-coordinate
phosphorus atoms opposite the tetra-coordinate phosphorus
atoms appear in a chemical shift range from −295 to −246 ppm
(marked by yellow or green dots, Figure 5). It is not trivial to
distinguish the two tricoordinate phosphorus atoms opposite
the tetra-coordinate moiety, but the assignment of the
resonances of each phosphorus nucleus in the cages 21a−g+
1
2
Scheme 3. Preparation of Compounds 21a−g[ECl4] from P4,
RPCl2, and ECl3 in Fluorobenzene at RT (E = Al, Ga)
filtration and washing with n-hexane or recrystallization from
dichloromethane, the salts 21a−g[ECl4] are obtained as
analytically pure compounds (>97% purity) in good to
excellent yields (Table 3). Trace impurities of white
phosphorus can be removed by vacuum sublimation at room
3380
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
a
Table 3. 31P{1H} NMR Parameters for [RP5Cl]+ Cations 21a−g+ , Taft’s σ* Parameter for Alkyl and Aryl Groups, and Yields of
21a−g[ECl4] (E = Al, Ga)
21a+
21b+
21c+
21d+
21e+
21f+
21g+
bc
,
g
spin system
ABM2X
−295
−278
44
ABM2X
−283
−279
54
ABM2X
−283
−281
54
ABM2X
−286
−270
66
ABMX2
−287
−260
69
ABMX2
−285
−284
55
ABMX2
−284
−246
26
d
δA
δB
δM
δX
99
85
89
80
81
76
83
de
,
1
1JAM; JAX
−140
−136
−181
−283
2
−135
−141
−180
−280
23
−135
−142
−180
−281
23
−134
−142
−177
−279
24
−134
−143
−175
−277
25
−144
−133
−179
−281
3
−137
−142
−172
−307
26
1JBM; JBX
1
1JAB
1JMX
2JAM; JAX
2JBM; JBX
2
2
24
−2
0
0
3
23
10
R
tBu
Cy
iPr
Et
Me
Ph
C6F5
1.10
62
f
σ*
−0.30
65
−0.26
90
−0.19
81
−0.10
64
−0.00
79
0.60
90
yield [ECl4]− salts [%]
E = Ga, Al
77
94
74
80
85
94
67
a
b
31P{1H} NMR data of the [GaCl4]− and [AlCl4]− salts are similar (in CD2Cl2, 25 °C). Designation of spin system by convention. Furthest
downfield resonance is denoted by the latest letter in the alphabet, and furthest upfield by the earliest letter: Δδ(PiPii)/JPiPii > 10 (resonance
c
considered to be pseudo first order and the assigned letters are separated) < 10 (consecutive letters are assigned). All parameters were derived by
d
e
iterative fitting of experimental at 161.94 MHz. Chemical shifts (δ) are given in [ppm] and coupling constants (J) in [Hz]. The absolute sign of
f
the 1JPP have been tentatively assigned to be negative. Tafts’s σ* values for alkyl and aryl groups (R = tBu, Cy, iPr, Et, Me, Ph) taken from ref 34 and
for R = C6F5 from ref 35. Additional phosphorus-fluorine coupling has to be considered in the full line shape iteration: JMF = 5.8 Hz, JXF
g
3
4
=
4
32.3 Hz, JMF = 4.7 Hz.
was achieved by a detailed analysis of the respective spin
systems and the 31P NMR coupling constants are presented in
Table 3. According to Figure 7, the unsymmetrical substituted
P5 cage can be divided by a plane, spanned by the tetra-
coordinate and bridging tricoordinate phosphorus atoms, into a
HCl- and HR-hemisphere. The HCl-hemisphere contains the
chloride substituent and the HR-hemisphere the alkyl or aryl
respective chemical shifts strongly depend on R and range from
−295 to −246 ppm (vide supra). For those nuclei, the 1JPP and
2JPP couplings to the bridging and tetra-coordinate phosphorus
atoms are close to ∼−142 Hz and ∼2 Hz, respectively.
Comparison of the 2JPP coupling constants with those observed
2
in the symmetrically substituted R2P5-cations (R = Ph, JPP
=
8 Hz; R = Cl, Br, I, JPP = 26 Hz3) strongly supports our
assignment according to Figures 5 and 8. The JPP coupling
5
2
1
2
1
substituents. The different types of JPP and JPP coupling
constants in each cation are sufficiently different to allow for an
unambiguous assignment of the chemical shift of each
phosphorus atom by comparison. Although it is known that
constants between the tetra-coordinate phosphorus atoms and
the bridging phosphorus atoms show a larger value, typically
found for phosphanylphosphonium cations,40 and range from
−277 to −307 Hz (Table 3).
1
the JPP coupling constants can strongly depend on parameters
The 31P{1H} NMR spectrum of the reaction mixture of 13g,
P4 and GaCl3 in fluorobenzene in a 1:1:1 ratio shows the
ABMX2 spin system of cation 21g+ (Figure 8). An interesting
feature of this spectrum (fitted spectrum as insets in Figure 8)
is the additional coupling to the ortho- and meta-fluoro
substituents of the C6F5-group. The X2 part of the spin system
is, therefore, split into a doublett of triplett of triplett arising
such as oxidation state, coordination number, electronic effects
of the substituents, and the presence or absence of localized
1
electron pairs, we have assigned the absolute value of the JPP
coupling to be negative in agreement with previous studies.38
1
2
With the sign of JPP predetermined, the sign of the JPP
coupling constants is obtained from the fit. Two sets of 1JPP and
2JPP coupling constants are observed between the tricoordinate
phosphorus atoms (whose resonances are marked by blue,
yellow, and green dots in Figure 5) and the tetra-coordinate
phosphorus atoms (whose resonances are marked by a red dot
in Figure 5). The nature of the substituents at the tetra-
coordinate atom controls both the coupling constants 1,2JPP and
to a considerable extent the chemical shifts. This is not only
attributed to the electronic effects of the substituents (vide
supra) but is also caused by “cross-ring through space”
interaction of the lone pairs on the phosphorus atoms and
the respective substituent.39 Thus, the phosphorus signals
marked in green are assigned to the HCl-hemisphere since the
respective chemical shifts are only slightly different and range
4
from the additional “through space” JPF enlarged coupling
constant of 32 Hz.41 The small additional splitting of the
resonance for the tetra-coordinate phosphorus atom
3
originates from the JPF coupling constant with a value of 6
Hz. This small value can be explained as due to cancellation
of a positive “through space” contribution form the lone pair
interaction of the bridging phosphorus atoms and the ortho-
fluorine substituent and a negative “through bond”
contribution. This “through space” coupling path is further
supported by the X-ray analysis of the solid-state structure of
21g[GaCl4].42
Compounds 21b−g[GaCl4] have been crystallographically
characterized, confirming their identities as the first examples of
nonsymmetrically substituted cationic P5-cages. Suitable
crystals for X-ray investigations were obtained by diffusion of
n-hexane into CH2Cl2 solutions of the respective salts. Table 4
1
2
from −278 to −287 ppm. The JPP and JPP couplings of these
phosphorus atoms to the bridging and tetra-coordinate P atoms
are close to ∼−135 Hz and ∼22 Hz, respectively. The signals
marked in yellow are assigned to the HR-hemisphere. The
3381
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Figure 6. δ vs σ* plot for the phosphonium nuclei in cations 21a,c−g+.
The correlation coefficient is 0.9830.
Figure 7. Definition of the HCl- and HR-hemisphere in cages 21a−g+.
The tetra-coordinate and bridging tricoordinate phosphorus atoms
span a plane. The tricoordinate phosphorus atom above the plane lies
within the HCl-hemisphere (marked by green dots, Figure 5), the
phosphors atom below in the HR-hemisphere (marked by a yellow
dots, Figure 5).
Figure 5. 31P{1H} NMR spectra of cations 21a−g+ (in CD2Cl2, 300 K,
161.94 MHz). Signals assigned to unknown side-products are labeled
with asterisks.
lists selected structural parameters of the cages 21b−g+, and
Figures 9 and 10 show ORTEP views of each cation.
Crystallographic data and details of the structure refinement
of compounds 21b−g[GaCl4] are presented in Table 5. The P5
cages in cations 21b−g+ display nearly identical bond lengths
and angles. The P−P bond lengths in cations 21b−g+
(2.1511(9) to 2.258(2) Å) are very close to the values found
in the symmetrically substituted [P5Br2]+ (2.150(7) to 2.262(8)
Å)3 and [Ph2P5]+ ion (2.179(1) to 2.249(1) Å).5 Similar to the
latter C2v-symmetric cages, the bonds between the bridging (P2,
P3) and tetra-coordinate phosphorus atoms (P1) and the P4−
P5 bond in cations 21b−g+ are approximately 0.08 Å shorter
(2.1511(9) to 2.1981(6) Å) than the remaining P−P bonds
(2.257(1) to 2.232(1) Å). A short P4−P5 bond length has been
observed for all reported cationic P5-cages.3,5-7 Similarly short
P−P bond lengths were also detected in a related SiP4 cage
compound (2.159(2) Å)43 and in bicyclo[1.1.0]-tetraphosphanes,
R2P4 (R = organyl) which display relatively short P−P bridgehead
Figure 8. 31P{1H} NMR spectrum of cation 21g+ (C6D6-capillary,
C6H5F, 25 °C, 161.94 Hz). Signals assigned to unknown side-products
are labeled with asterisks. Expansion (inset) show the experimental
(up) and fitted36 (down) spectra of cation 21g+.
bonds (2.120 Å).44 There are several examples of rather short P−P
single bond distances involving cationic four-coordinate phospho-
rus atoms.2,9,45,46 This shortening may be caused by the positive
charge at the phosphonium moiety which also results in a
3382
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
a
Table 4. Selected Bond Length [Å] and Angles [°] of [RP5Cl]-Cations 21b−g+
21b+
21c+
21d+
21e+
21f+
21g+
P1−Cl1
P1−C1
P1−P2
P1−P3
P2−P4
P2−P5
P3−P4
P3−P5
P4−P5
2.013(2)
1.809(4)
2.156(2)
2.163(2)
2.258(2)
2.241(2)
2.256(2)
2.243(2)
2.161(2)
2.0093(9)
1.833(3)
2.1597(9)
2.1594(9)
2.239(1)
2.2546(9)
2.2393(9)
2.2541(9)
2.170(1)
2.0167(5)
1.799(2)
1.996(2)
1.823(7)
2.162(1)
2.163(1)
2.245(1)
2.257(1)
2.250(1)
2.253(1)
2.178(1)
2.0040(5)
1.786(2)
2.0078(9)
1.802(2)
2.1621(9)
2.1511(9)
2.246(1)
2.232(1)
2.250(1)
2.252(1)
2.185(1)
2.1630(6)
2.1559(5)
2.2462(6)
2.2406(6)
2.2358(6)
2.2501(6)
2.1851(6)
2.1644(5)
2.1615(5)
2.2496(6)
2.2428(6)
2.2491(6)
2.2533(6)
2.1689(6)
Cl1−P1−C1
P2−P1−P3
P2−P1−C1
P3−P1−Cl1
P1−P2−P5
P1−P2−P4
P5−P2−P4
P4−P5−P2
P2−P5−P3
P5−P4−P3
104.4(1)
91.00
102.60(9)
91.57(3)
115.91(9)
113.30(4)
81.08(3)
85.18(3)
60.75(3)
61.48(3)
86.72(3)
61.46(3)
105.39(5)
91.54(2)
117.17(6)
110.61(2)
82.05(2)
83.86(2)
58.29(2)
60.99(2)
87.12(2)
61.18(2)
104.3(2)
91.41(4)
119.3(4)
114.2(2)
82.17(4)
84.67(4)
57.87(4)
60.87(4)
86.68(4)
61.14(4)
107.58(5)
90.68(2)
115.39(5)
112.94(2)
83.73(2)
84.60(2)
57.74(2)
61.29(2)
86.37(2)
61.30(2)
103.88(8)
92.59(4)
114.96(8)
115.03(4)
83.89(4)
80.34(3)
58.41(3)
61.13(3)
88.14(4)
60.99(4)
118.3(2)
111.69(6)
83.61(5)
84.58(6)
57.42(5)
61.71(5)
86.81(6)
60.98(5)
a
All bond length and angles in the [GaCl4]− anion show typical values.
Figure 10. ORTEP representation of the molecular structures of the
cations 21f,g+ in compounds 21f,g[GaCl4]. Thermal ellipsoids are set at
50% probability. Hydrogen atoms and anions are omitted for clarity.
4
interaction is in agreement with the observed large JPF
coupling constant in the 31P{1H} NMR spectrum (vide
supra) and is evidence in favor of the proposed “through
space” interaction.
CONCLUSION
■
The composition of mixtures of RPCl2 and GaCl3 in
fluorobenzene strongly depends on the basicity of RPCl2 (R =
tBu, Cy, iPr, Et, Me, Ph, C6F5) and the reaction
stoichiometry. 31P{1H} NMR and Raman experiments show
the formation of several structurally distinct species, including
RPCl2·GaCl3 adducts, (14) and dichlorophosphanylchloro-
phosphonium cations [RPCl2−RPCl]+ (16+) and [RPCl2−
RPCl−GaCl3]+ (17+). Despite their different compositions 1:1
mixtures of RPCl2 and a Lewis acid ECl3 (E = Al, Ga) are
potent sources of reactive chlorophosphenium cation [RPCl]+
equivalents, which insert formally into P−P bonds of dissolved P4.
The formal insertion reaction has been investigated with
quantum chemical calculations, and we suggest a two-stage
reaction mechanism via a butterfly shaped intermediate. This
synthetic approach represents a general method and a powerful
strategy for the synthesis of new cationic chloro-substituted
Figure 9. ORTEP representation of the molecular structures of the
cations 21b−e+ in compounds 21b−e[GaCl4]. Thermal ellipsoids are
set at 50% probability. Hydrogen atoms and anions are omitted for
clarity.
significant shortening of the P−C (ranging from 1.786(2) to
1.833(3) Å) and P−Cl bonds (ranging from 1.996(2) to
2.0167(5) Å) compared to chloro- and alkyl- or aryl-
substituted phosphanes (typical bond length for P−C: 1.87 Å
and P−Cl: 2.04 Å).47 A special feature in cation 21g+ is the
observation of an intramolecular contact between the
phosphorus atom P3 and one of the ortho-fluorine atoms of
the C6F5-substituent (2.947(2) Å) which lies well within the
sum of the van der Waals radii (rF + rP = 3.27 Å).48 This
3383
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
organophosphorus [RP5Cl]+ cages. The reactions proceed cleanly
for a wide range of different substituents on the dichlorophos-
phane to new chloro substituted monocations 21a-g+ isolated in
good to excellent yields. The detailed 31P NMR investigation of
these cations enabled a deeper understanding of the electronic and
steric influences of the substituents on the 31P NMR parameters.
Furthermore, the simplicity and viability of this reaction bodes well
for the development of larger cages using P4 as a starting material.
Because of the presence of chloro substituents such cages are
amenable to a host of subsequent further transformations.
room temperature. P4 dissolves completely accompanied by the
formation of a colorless to yellowish precipitate. After addition of
n-hexane (2 mL), the precipitate was collected, washed with n-hexane
(3 × 2 mL), and dried in vacuo. Note, excess P4 can be removed by
sublimation to yield analytical pure product. Additional manipulations
and reaction times are given for each compound seperately (vide infra).
21a[ECl4]. The reaction mixture was stirred for 2 d. The formed
precipitates (contains unreacted P4) were removed by filtration.
The addition of hexane (4 mL) to the filtrate afforded a colorless
precipitate which was separated by filtration, washed with n-hexane
(3 × 2 mL), and dried in vacuo to yield analytically pure 21a[ECl4];
21a[GaCl4]: Yield: 65% (298 mg); mp 145 °C (decomp.); IR (KBr,
25 °C, cm−1): 2967 (s), 2957 (m), 2925 (w), 1470 (vw), 1455 (vs),
1399 (vw), 1369 (s), 1309 (vw), 1239 (w), 1171 (s), 1012 (m), 939
(w), 791 (m), 619 (m), 586 (s), 573 (vw), 542 (w), 457 (w); Raman
(150 mW, 25 °C, cm−1): 2971 (4), 2944 (3) 2901 (10), 1460 (5),
1175 (3), 794 (5), 566 (6), 542 (100), 460 (50), 440 (27), 392 (41),
364 (34), 342 (25), 213 (23), 159 (39), 123 (21), 106 (6); 1H NMR
EXPERIMENTAL SECTION
■
General Considerations. All reactions were carried out either in a
glovebox or using standard Schlenk techniques under an inert Ar
atmosphere. Dry, oxygen-free solvents were employed. P4 was dried
with Me3SiCl and recrystallized from CS2 prior to use. Reagent grade
GaCl3 and AlCl3 were used as received from commercial suppliers.
PhPCl2, iPrPCl2, and MePCl2 were purchased from Sigma-Aldrich and
3
(CD2Cl2, 25 °C, [ppm]): δ = 1.58 (9H, d, Me, JHP = 23.7 Hz);
13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 24.3 (3C, d, CH3, 2JCP
=
freshly distilled prior to use. CyPCl2,49 tBuPCl2,50 and C6F5PCl2
51
3.9 Hz), 50.2 (1C, s, C); 71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]):
δ = 249 (s, ν1/2 = 158 Hz); elemental analysis for C4H9Cl5GaP5
(458.97): calcd. C: 10.5, H 2.0; found: C 10.3, H 2.1; 21a[AlCl4]:
Yield: 77% (297 mg); mp 130 °C (decomp.); IR (KBr, 25 °C, cm−1):
2968 (vs), 2933 (vw), 1457 (s), 1395 (w), 1370 (s), 1309 (vw), 1241
(w), 1171 (s), 1095 (vw), 998 (m), 939 (w), 875 (w), 834 (vw), 791
(s), 766 (w), 677 (vw), 619 (2), 587 (vw), 571 (m), 542 (w), 481
(vs), 437 (vw); Raman (150 mW, 25 °C, cm−1): 2972 (4), 2944 (4),
2901 (13), 1460 (6), 1175 (4), 794 (5), 567 (6), 542 (100), 458 (36),
440 (26), 392 (37), 366 (12), 359 (47), 348 (13), 213 (23), 161 (37),
were prepared according to literature procedures. All NMR measure-
ments were performed at 300 K on a Bruker AVANCE III 400 or a
Bruker AVANCE II 200 (RPCl2/GaCl3 mixtures) spectrometer, and
spectra were referenced either to residual solvent (1H, 13C) or
externally [19F (CCl3F); 31P (H3PO4); 27Al (Al(NO3)3); 71Ga
(Ga(NO3)3)]. CD2Cl2 was purchased form Sigma-Aldrich and dried
over CaH2, vacuum distilled prior to use, and stored over 4 Å
molecular sieves in the glovebox. To obtain the 31P{1H} NMR data for
reaction mixtures, a C6D6 capillary was inserted in the NMR tube. For
compounds which give rise to higher order spin-systems in their
31P{1H} NMR spectra, the resolution-enhanced 31P{1H} spectra were
transferred to the software program gNMR, version 5.0, by Cherwell
Scientific.36 The full line shape iteration procedure of gNMR was
applied to obtain the best match of the calculated to the experimental
spectra along with the assignment of all the peaks revealed in the
1
128 (5), 105 (18); H NMR (CD2Cl2, 25 °C, [ppm]): δ = 1.57 (9H,
3
d, Me, JHP = 23.8 Hz); 13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ =
24.2 (3C, d, CH3, 2JCP = 3.9 Hz), 50.2 (1C, s, C); 27Al NMR (CD2Cl2,
25 °C, [ppm]): δ = 104 (s, ν1/2 = 133 Hz); elemental analysis for
C4H9Cl5AlP5 (416.23): calcd. C: 11.5, H 2.2; found: C 11.3, H 2.4.
21b[GaCl4]. 2 h reaction time; yield: 90% (437 mg); mp 46.3−48.1 °C;
IR (KBr, 25 °C, cm−1): 2941 (vs), 2919 (s), 2875 (m), 2853 (w),
1444 (vs), 1291 (w), 1266 (s), 1194 (vw), 1176 (w) 1110 (vw), 1040
(vw), 995 (m), 916 (w), 888 (m), 844 (w), 815 (s), 743 (m), 620 (w),
572 (vs), 458 (m), 421 (w); Raman (150 mW, 25 °C, cm−1): 2950
(15), 2878 (6), 2853 (4) 1432 (5), 1293 (7), 1268 (6), 1024 (9), 846
(3), 817 (6), 573 (8), 556 (7), 544 (100), 460 (35), 442 (48), 423 (4),
391 (38), 357 (63), 343 (36), 289 (7), 225 (5), 213 (21), 192 (18),
176 (13), 151 (10), 125 (65); 1H NMR (CD2Cl2, 25 °C, [ppm]): δ =
1.34 (1H, m, C4Hax), 1.50 (2H, m, C2Hax), 1.59 (2H, m, C3Hax), 1.83
(2H, m, C4Heq); 2.05 (2H, m, C3Heq); 2.23 (1H, m, C2Heq), 3.08
(1H, m, C1Hax); 13C{1H}-NMR (CD2Cl2, 25 °C, [ppm]): δ = 25.0
1
resolution-enhanced spectra. The signs for the J(31P,31P) coupling
constants were set negative and all other signs obtained accordingly.38
The connectivity of the phosphorus cage compounds 21a−g+ was
additionally proven by the use of 2-dimensional 31P{1H}31P{1H}-
COSY NMR experiments. Melting points were recorded on an
electrothermal melting point apparatus in sealed capillaries under
Argon atmosphere and are uncorrected. Infrared (IR) spectra were
recorded using a Perkin-Elmer Spectrum One FT-IR instrument.
Raman spectra were recorded on a Bruker Ram II spectrometer, with
Nd:YAG-Laser (1064 nm, 10−200 mW) in sealed glass capillaries.
The intensities are reported in percent relative to the most intense
peak and are given in parentheses. Elemental analyses were performed
on a Vario EL III CHNS elemental analyzer at the IAAC, University of
1
2
(1C, d, C1, JCP = 2.0 Hz), 25.4 (2C, d, C2/C6, JCP = 4.9 Hz), 25.7
(2C, d, C3/C5, JCP = 4.5 Hz), 25.8 (1C, s, C4); 71Ga{1H} NMR
3
(CD2Cl2, 25 °C, [ppm]): δ = 248 (s, ν1/2 = 1500 Hz); elemental
analysis for C6H11Cl5GaP5 (485.01): calcd.: C 14.9, H 2.3, found: C
14.9, H 2.1; 21b[AlCl4]: 2 h reaction time; yield: 84% (372 mg); mp
46.9−49.5 °C; IR (KBr, 25 °C, cm−1): 2940 (vs), 2920 (s), 2880 (m),
2854 (vw), 1444 (s), 1292 (w), 1265 (s), 1194 (w), 1176 (w), 1110
(vw), 1023 (vw), 995 (m), 916 (w), 888 (w), 844 (w), 815 (s), 743
(w), 675 (m), 619 (vw), 572 (w), 542 (w), 481 (s) 387 (w); Raman
(150 mW, 25 °C, cm−1): 2947 (21), 2880 (3), 2855 (10), 1447 (6),
1294 (6), 1268 (8), 1024 (9), 847 (4), 817 (7), 573 (8), 557 (5), 544
(100), 460 (34), 442 (47), 390 (41), 364 (68), 358 (7), 347 (8), 289
(7), 225 (5), 213 (21), 192 (23), 176 (14), 130 (35); 1H NMR
(CD2Cl2, 25 °C, [ppm]): δ = 1.35 (1H, m, C4Hax), 1.51 (2H, m,
C2Hax), 1.58 (2H, m, C3Hax), 1.83 (2H, m, C4Heq), 2.05 (2H, m,
C3Heq), 2.24 (1H, m, C2Heq), 3.09 (1H, m, C1Hax); 13C{1H} NMR
Munster, Germany, except for 21g[GaCl ] and 21g[AlCl ] because of
̈
4
4
the high content of fluorine.
31P{1H} NMR Experiments of Dichlorophosphane/GaCl3
Mixtures. A solution of the dichlorophosphane 13a-f (0.5 mmol)
and GaCl3 (0.25/0.5/1.0/1.5 mmol) in C6H5F (2 mL) was stirred for
15 min at room temperature after which 31P{1H} NMR and Raman
spectra were recorded immediately (Tables 1, 2).
Synthesis of 14a. A solution of tBuPCl2 (159 mg, 1.0 mmol) and
GaCl3 (176 mg, 1.0 mmol) in C6H5F (5 mL) was stirred for 2 h at
room temperature. Diffusion of n-hexane into the reaction mixture at
−35 °C yields small amounts of adduct 14a as crystalline material. 14a
is very temperature sensitive and isolation needs to be performed at
low temperatures (−35 °C); 14a: 18% (60 mg, 0,2 mmol); mp > −25 °C;
3
1H NMR (CD2Cl2, 25 °C, [ppm]): δ = 1.52 (d, 9H, CH3, JHP
=
1
22.7 Hz); 13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 24.1 (3C, d,
(CD2Cl2, 25 °C, [ppm]): δ = 25.0 (1C, d, C1, JCP = 2.6 Hz), 25.6
CH3, JCP= 9.8 Hz), 44.0 (1C, d, C(CH3)3, JCP= 12.7 Hz); 71Ga{1H}
NMR (CD2Cl2, 25 °C, [ppm]): δ = 259 (s, ν1/2 = 4500 Hz); 31P{1H}
NMR (CD2Cl2, 25 °C, [ppm]): δ = 142.4 (1P, s, ν1/2 = 37 Hz).
General Synthesis of Compounds 21a−g[ECl4] (E = Ga, Al). A
mixture of the dichlorophosphane (1.0 mmol), ECl3 (1.0 mmol), and
freshly recrystallized P4 (1.0 mmol) in C6H5F (5 mL) was stirred at
(2C, d, C2/C6), 25.7 (2C, s, C3/C5), 25.8 (1C, s, C4); 27Al-NMR
(CD2Cl2, 25 °C, [ppm]): δ = 104 (s, ν1/2 = 95 Hz); elemental analysis for
C6H11Cl5AlP5 (442.27): calcd. C: 16.3, H 2.5; found: C 16.3, H 2.4.
21c[GaCl4]. 2 h reaction time; yield: 81% (360 mg); mp 125.8−
127.3 °C; IR (KBr, 25 °C, cm−1): 2967 (m) 2915 (w) 2862 (vw),
1456 (m), 1385 (m), 1366 (w), 1233 (m), 1030 (w), 863 (vw), 651
2
1
3384
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
Table 5. Crystallographic Data and Details of the Structure Refinement of Compounds 14a and 21b−g[GaCl4]
14a
21b[GaCl4]
21c[GaCl4]
21d[GaCl4]
21e[GaCl4]
21f[GaCl4]
21g[GaCl4]
formula
C4H9Cl5GaP
335.05
C6H11Cl5GaP5
484.96
C3H7Cl5GaP5
444.91
C2H5Cl5GaP5
430.88
CH3Cl5GaP5
416.85
C6H5Cl5GaP5
478.92
C6Cl5F5GaP5
568.88
Mr [g mol−1
]
dimension [mm3]
0.27 × 0.22 ×
0.10
0.12 × 0.03 ×
0.02
0.32 × 0.07 ×
0.2
0.20 × 0.06 ×
0.03
0.20 × 0.14 ×
0.04
0.09 × 0.05 ×
0.02
0.19 × 0.04 ×
0.02
color, habit
crystal system
space group
a [Å]
colorless plate
monoclinic
P21/n
colorless plate
orthorhombic
Cmc21
colorless plate
monoclinic
P21/c
colorless plate
monoclinic
P21/n
colorless plate
orthorhombic
P212121
6.3462(4)
9.2206(6)
22.395(1)
90
colorless plate
monoclinic
P21/c
colorless rod
monoclinic
P21
12.0362(7)
8.6708(5)
13.2467(8)
90
23.2837(5)
13.2160(3)
16.7780(3)
90
14.3092(9)
21.778(1)
9.5931(6)
90
12.6761(6)
8.9880(4)
13.2882(6)
90
8.4378(5)
15.2665(9)
12.3411(7)
90
6.8661(8)
8.802(1)
14.108(2)
90
b [Å]
c [Å]
α [deg]
β [deg]
γ [deg]
V [Å3]
116.435(1)
90
90
96.350(1)
90
112.951(1)
90
90
92.307(1)
90
95.396(2)
90
90
90
1237.9(1)
4
5162.7(2)
12
2971.3(3)
8
1394.1(1)
4
1310.5 (1)
4
1588.4(2)
4
847.6(2)
2
Z
T [K]
ρc [g cm−3
153(1)
1.798
153(1)
1.872
153(1)
1.989
153(1)
2.053
153(1)
2.113
153(1)
2.003
153(2)
2.229
]
F(000)
656
2856
1728
832
800
928
544
λMoKα [Å]
0.71073
3.377
0.71073
2.816
0.71073
3.252
0.71073
3.462
0.71073
3.679
0.71073
3.050
0.71073
2.918
μ [mm−1
]
absorption correction
reflections collected
reflections unique
Rint
SADABS
12275
SADABS
15069
SADABS
29945
SADABS
137999
3327
SADABS
13457
SADABS
17396
SADABS
9477
2961
6703
7091
3129
4266
4525
0.0257
2584
0.0442
5370
0.0435
5786
0.0238
3008
0.0407
2865
0.0252
3803
0.0185
4325
reflection obs.
[F > 3σ(F)]
residual density [e Å−3
]
2.527, −0.645
103
0.781, −0.592
243
0.600, −0.456
257
0.378, −0.311
119
0.519, −0.647
110
0.366, −0.310
154
0.423, −0.276
199
parameters
GOF
1.039
0.995
1.017
1.061
1.047
1.012
1.041
R1 [I > 3σ(I)]
wR2 (all data)
CCDC
0.0388
0.1036
826493
0.0373
0.0717
826494
0.0284
0.0691
826495
0.0196
0.0496
826496
0.0265
0.0529
826497
0.0191
0.0424
826498
0.0253
0.0621
826499
3
3
(s), 561 (vs), 540 (m), 452 (w), 412 (w); Raman (150 mW, 25 °C,
cm−1): 2967 (23), 2917 (28), 2862 (7), 1439 (5), 1388 (5), 1369 (4),
1238 (3), 1080 (3), 1033 (3), 865 (7), 654 (4), 578 (13), 563 (17),
542 (100), 452 (63), 414 (34), 390 (51), 358 (100), 345 (40), 237
(21), 206 (21), 153 (92), 123 (47); 1H NMR (CD2Cl2, 25 °C,
[ppm]): δ = 1.53 (3H, dt, CH3, JHP3 = 27.4 Hz, JHH = 7.5 Hz), 3.10
(2H, dqu, CH2, JHP = 15.0 Hz, JHH = 7.5 Hz); 13C{1H} NMR
(CD2Cl2, 25 °C, [ppm]): δ = 7.0 (1C, d, CH3, JCP = 5.9 Hz), 39.4
2
2
(1C, m, CH2); 71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 240 (s,
ν1/2 = 7700 Hz); elemental analysis for C2H5Cl5GaP5 (430.92): calcd.
C: 5.6, H 1.2; found: C 6.0, H 1.4; 21d[AlCl4]: 2 h reaction time;
yield: 80% (311 mg); mp 143.3−144.2 °C; IR (KBr, 25 °C, cm−1):
2943 (m), 2904 (m), 1441 (w), 1386 (m), 1242 (vs), 1153 (w), 1015
(w), 1001 (s), 743 (vs), 697 (m), 571 (s), 548 (w), 532 (w), 501 (w),
434 (vw); Raman (150 mW, 25 °C, cm−1): 2976 (12), 2932 (10),
2905 (16), 2866 (5), 1388 (7), 570 (15), 548 (60), 530 (51), 438
(100), 395 (46), 363 (14), 356 (60), 242 (24), 210 (82), 179 (10),
2
[ppm]): δ = 1.49 (6H, dd, CH3, JHP = 26.0 Hz, 3JHH = 7.0 Hz), 3.35
3
(1H, sept., CH, JHH = 7.0 Hz); 13C{1H}-NMR (CD2Cl2, 25 °C,
2
[ppm]): δ = 15.3 (2C, d, CH3, JCP = 1.9 Hz), 45.0 (1C, m, CH);
71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 248 (s, ν1/2 = 1900 Hz);
elemental analysis for C3H7Cl5GaP5 (444.94): calcd. C: 8.1, H 1.6;
found: C 8.4, H 1.6; 21c[AlCl4]: 2 h reaction time Yield: 74% (298
mg); mp 156.9−158.1 °C; IR (KBr, 25 °C, cm−1): 2967 (m), 2916
(vw), 2862 (vw), 1456 (m), 1386 (w), 1366 (w), 1233 (m), 1031 (w),
863 (w), 663 (vw), 652 (m), 563 (s), 541 (m), 487 (vw), 470 (vs);
Raman (150 mW, 25 °C, cm−1) 2970 (20), 2917 (22), 2863 (7), 1464
(5), 1388 (4), 1368 (3), 1235 (4), 1079 (3), 1037 (4), 866 (5), 565
(11), 544 (100), 454 (40), 415 (23), 391 (32), 361 (61), 349 (8), 238
1
127 (14); H NMR (CD2Cl2, 25 °C, [ppm]): δ = 1.52 (3H, dt, CH3,
3JHP = 27.5 Hz, JHH = 7.6 Hz), 3.12 (2H, dqu, CH2, JHP = 15.2 Hz,
3JHH = 7.6 Hz); 13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 7.0 (1C,
d, CH3, 2JCP = 5.9 Hz), 39.4 (1C, m, CH); 27Al NMR (CD2Cl2, 25 °C,
[ppm]): δ = 103.7 (s, ν1/2 = 69 Hz); elemental analysis for
C2H5Cl5AlP5 (388.18): calcd. C: 6.2, H 1.3; found: C 6.7, H 1.5.
21e[GaCl4]. 2 h reaction time; Yield: 79% (329 mg); mp 149.2−152.1
°C; IR (KBr, 25 °C, cm−1): 2985 (m), 2897 (vs), 2237 (vw), 2179
(vw), 2031 (vw), 1384 (m), 1240 (vs), 983 (m), 893 (vs), 749 (s),
638 (vw), 562 (vs), 537 (w), 436 (s); Raman (150 mW, 25 °C, cm−1):
2987 (5), 2898 (24), 1384 (8), 752 (9), 563 (22), 539 (85), 529 (17),
442 (100), 398 (25), 357 (52), 346 (20), 285 (7), 250 (9), 201 (39),
155 (23), 122 (15); 1H NMR (CD2Cl2, 25 °C, [ppm]): δ = 3.00 (3H,
3
2
1
(14), 207 (11), 178 (4), 164 (48), 132 (16); H NMR (CD2Cl2, 25
2
3
°C, [ppm]): δ = 1.52 (6H, dd, CH3, JHP = 26.0 Hz, JHH = 7.0 Hz),
3.38 (1H, sept., CH, JHH = 7.1 Hz); 13C{1H} NMR (CD2Cl2, 25 °C,
3
[ppm]): δ = 15.2 (2C, d, CH3, 2JCP = 2.0 Hz), 45.0 (1C, m, CH); 27Al
NMR (CD2Cl2, 25 °C, [ppm]): δ = 103.5 (s, ν1/2 = 10 Hz); elemental
analysis for C3H7Cl5AlP5 (402.20): calcd. C: 9.0, H 1.8; found: C 9.3,
H 1.8.
21d[GaCl4]. 2 h reaction time; yield: 64% (276 mg); mp 100.6−
101.8 °C; IR (KBr, 25 °C, cm−1): 2942 (m), 2903 (m), 2864 (vw),
1441 (w), 1385 (m), 1243 (w), 1132 (vw), 1015 (w), 1001 (m), 742
(vs), 696 (m), 570 (s), 547 (w), 531 (vs), 435 (s); Raman (150 mW,
25 °C, cm−1): 2973 (10), 2934 (27), 2910 (25), 2873 (6), 1398 (5),
544 (100), 473 (25), 448 (52), 394 (53), 376 (53), 361 (78), 348
(85), 224 (25), 153 (80), 116 (20); 1H NMR (CD2Cl2, 25 °C,
d, Me, JPH = 10.8 Hz); 13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ =
2
32.7 (1C, m); 71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 213 (s,
ν1/2 = 4300 Hz); elemental analysis for CH3Cl5GaP5 (416.89): calcd.:
C 2.9, H 0.7; found: C 3.2, H 1.1; 21e[AlCl4]: 2 h reaction time; yield:
85% (318 mg); mp 144.5−145.9 °C; IR (KBr, 25 °C, cm−1): 2988 (m),
2902 (w), 1384 (m), 1283 (w), 1156 (vw), 892 (vs), 750 (m), 561 (m),
3385
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
540 (w), 477 (s), 437 (vw); Raman (150 mW, 25 °C, cm−1): 2993 (4),
2906 (16), 1387 (7), 750 (8), 559 (23), 541 (48), 437 (100), 397
(21), 361 (9), 353 (45), 281 (6), 251 (9), 204 (52), 181 (5), 130
data for all compounds were collected on a Bruker AXS APEX CCD
diffractometer equipped with a rotating anode at 153(1) K using
graphite-monochromated MoKα radiation (λ = 0.71073 Å) with a scan
width of 0.3° and variable exposure times typically between 5 and 15 s.
In all cases generator settings were 50 kV and 180 mA. Diffraction data
were collected over the full sphere, and the frames were integrated
using the Bruker SMART52 software package using the narrow frame
algorithm. Data were corrected for absorption effects using the
SADABS routine (empirical multiscan method). For further crystal
and data collection details see Table 5.
Structure Solution and Refinement. Atomic scattering factors
for non-hydrogen elements were taken from the literature tabulations.
Structure solutions were found by using direct methods as
implemented in the SHELXS-97 package53and were refined with
SHELXL-9754 against F2 using first isotropic and later anisotropic
thermal parameters for all non-hydrogen atoms. Hydrogen atom
positions were calculated and allowed to ride on the carbon atom
to which they are bonded, assuming a C−H bond length of 0.95 Å.
H-atom temperature factors were fixed at 1.20 times the isotropic
temperature factor of the C atom to which they are bonded. The H-atom
contributions were calculated but not refined. The locations of the largest
peaks in the final difference Fourier map as well as the magnitude of the
residual electron densities in each case were of no chemical significance.
For 21b[GaCl4] and 21e[GaCl4] the crystals were twins (twin law 1 0 0,
0 −1 0, 0 0 −1) with not quite equal components.
1
2
(31); H NMR (CD2Cl2, 25 °C, [ppm]): δ = 2.9 (3H, d, Me, JHP
=
10.8 Hz); 13C{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 32.6 (1C, m);
27Al NMR (CD2Cl2, 25 °C, [ppm]): δ = 104 (s, ν1/2 = 371 Hz);
elemental analysis for CH3Cl5AlP5 (374.15): calcd. C: 3.2, H 0.8;
found: C 3.6, H 1.2.
21f[GaCl4]. 2 h reaction time; reaction time; yield: 90% (431 mg);
mp 124.3−126.0 °C; IR (KBr, 25 °C, cm−1): 3100 (m), 3055 (w),
1996 (vw), 1971 (w), 1898 (w), 1810 (w), 1673 (w), 1576 (s), 1477
(m), 1431 (vs), 1385 (w), 1334 (s), 1309 (m), 1160 (w), 1097 (vs),
997 (m), 927 (vw), 746 (vs), 712 (w), 677 (s), 613 (s), 577 (vs), 543
(w), 500 (s), 447 (m); Raman (150 mW, 25 °C, cm−1): 3064 (13),
1578 (17), 1162 (8), 1098 (9), 1025 (24), 997 (27), 577 (18), 559
(12), 545 (100), 502 (24), 439 (25), 392 (52), 378 (19), 355 (63),
342 (21), 307 (12), 231 (6), 217 (32), 201 (49), 152 (16), 126 (73);
1H NMR (CD2Cl2, 25 °C, [ppm]): δ = 7.85 (2H, m, m-Ph), 7.94 (1H,
m, p-Ph), 8.14 (2H, m, o-Ph); 13C{1H} NMR (CD2Cl2, 25 °C,
3
[ppm]): δ = 131.0 (2C, d, m-Ph, JCP = 13.8 Hz), 131.5 (1C, m, i-Ph),
131.8 (2C, d, o-Ph, 2JCP = 12.9 Hz), 137.5 (1C, d, p-Ph, 4JCP = 3.5 Hz);
71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]): δ = 240 (s, ν1/2 = 10000
Hz); elemental analysis for C6H5Cl5GaP5 (478.96): calcd. C 15.1, H
1.1; found C 15.1, H 1.3; 21f[AlCl4]: 2 h reaction time; yield: 94%
(410 mg); mp 119.5−121.2 °C; IR (KBr, 25 °C, cm−1): 3100 (w),
3054 (vw), 1996 (vw), 1971 (vw), 1898 (vw), 1811 (w), 1672 (w),
1576 (m), 1477 (w), 1431 (s), 1385 (vw), 1334 (m), 1309 (w), 1160
(vw), 1098 (w), 1084 (s), 995 (m), 746 (vs), 712 (m), 678 (s), 613
(s), 580 (vs), 544 (s), 467 (w); Raman (150 mW, 25 °C, cm−1) 3065
(12), 1558 (17), 1162 (7), 1099 (10), 1024 (20), 997 (28), 577 (20),
560 (11), 544 (100), 501 (21), 437 (25), 392 (47), 379 (35), 355
(70), 306 (12), 230 (6), 216 (30), 201 (53), 153 (5), 133 (50), 117
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed information on the quantum chemical calculations and
31P{1H} NMR investigations of reaction mixtures of 13c,d,e,g
and GaCl3 in various stoichiometries. This material is available
free of charge via the Internet at http://pubs.acs.org.
1
(17); H NMR (CD2Cl2, 25 °C, [ppm]): δ = 7.85 (2H, m, m-Ph),
7.95 (1H, m, p-Ph), 8.14 (2H, m, o-Ph); 13C{1H} NMR (CD2Cl2, 25
3
AUTHOR INFORMATION
°C, [ppm]): δ = 131.0 (2C, d, m-Ph, JCP = 13.7 Hz), 131.7 (1C, m,
■
2
4
i-Ph), 131.8 (2C, d, o-Ph, JCP = 13.0 Hz), 137.6 (1C, d, p-Ph, JCP
=
Corresponding Author
*E-mail: jweigand@uni-muenster.de.
3.8 H); 27Al-NMR (CD2Cl2, 25 °C, [ppm]): δ = 104 (s, ν1/2 = 130
Hz); elemental analysis for C6H5Cl5AlP5 (436.22): calcd. C 16.5, H
1.2; found: C 16.6, H 1.5.
Notes
21g[GaCl4]. 6 h reaction time; Yield: 62% (353 mg); mp 78.9−80.3
°C; IR (KBr, 25 °C, cm−1): 1644 (vs), 1561 (vw), 1516 (vs), 1485
(vw), 1395 (w) 1303 (m), 1155 (w), 1115 (w), 1095 (s) 1023 (vw),
981 (vs), 762 (vw), 721 (m), 632 (vs), 589 (w), 572 (vs), 534 (w),
526 (s), 461 (s); Raman (300 mW, 25 °C, cm−1): 1645 (8), 1397 (5),
861 (5), 591 (6), 575 (5), 540 (100), 461 (27), 440 (5), 404 (40), 359
(30), 344 (6), 224 (25), 183 (15), 144 (7), 126 (10), 92 (6); 19F
NMR (CD2Cl2, 25 °C, [ppm]): δ = −153.2 (2F, m, m-F), −134.2
(1F, m, p-F), −130.6 (2F, m, o-F); 13C NMR (CD2Cl2,
25 °C, [ppm]): δ = 106.4 (1C, m, i-p-F), 138.5 (2C, d(br), m-Ph,
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support from the Fonds
der Chemischen Industrie (fellowships to M.H.H. and K.-O.F.),
the DFG (WE 4621/2-1), the European COST network
PhosSciNet (CM0802), and the International Research Train-
ing Group 1444. J.J.W. thanks Prof. F. Ekkehardt Hahn (WWU
Muenster) for his support, and Prof. Robert Wolf (University
of Regensburg) for helpful discussions.
1
1JCF = 262 Hz), 146.4 (2C, d(br), o-Ph JCF = 260 Hz), 147.8 (1C,
d(br), p-Ph, 1JCF = 273 Hz); 71Ga{1H} NMR (CD2Cl2, 25 °C, [ppm]):
δ = 233 (s, ν1/2 = 10000 Hz); 21g[AlCl4]: Yield: 67% (353 mg); mp
93.4−95.5 °C; IR (KBr, 25 °C, cm−1): 1643 (vs), 1515 (w), 1486 (vs),
1395 (m), 1303 (m), 1155 (w), 1116 (w), 1095 (vs), 1023 (w), 980
(vs), 722 (w), 633 (s), 589 (vw), 573 (s), 536 (vw), 526 (w), 473 (m);
Raman (300 mW, 25 °C, cm−1): 1646 (10), 1397 (8), 862 (6), 591
(11), 575 (8), 539 (100), 463 (34), 378 (5), 358 (35), 343 (21), 225
(25), 174 (9), 145 (10), 119 (12), 92 (9); 19F-NMR (CD2Cl2, 25 °C,
[ppm]): δ = −153.5 (2F, m, m-F), −134.3 (1F, m, p-F), −130.8 (2F, m,
o-F); 13C{1H} NMR (CD2Cl21, 25 °C, [ppm]): δ = 106.9 (1C, m,
i-Ph), 138.8 (2C, d(br), m-Ph, JCF = 262 Hz), 146.8 (2C, d(br), o-Ph,
REFERENCES
■
(1) (a) Baudler, M. Angew. Chem., Int. Ed. Engl. 1982, 7, 492.
(b) Baudler, M. Angew. Chem., Int. Ed. Engl. 1987, 26, 419.
(2) Dyker, C. A.; Burford, N. Chem.Asian J. 2008, 3, 28.
(3) (a) Krossing, I.; Raabe, I. Angew. Chem., Int. Ed. 2001, 40, 4406.
(b) Bihlmeier, A.; Gonsior, M.; Raabe, I.; Trapp, N.; Krossing, I.
Chem.Eur. J. 2004, 10, 5041. (c) Krossing, I. J. Chem. Soc., Dalton
Trans. 2002, 500. (d) Gonsior, M.; Krossing, I.; Muller, L.; Raabe, I.;
̈
Jansen, M.; van Wullen, L. Chem.Eur. J. 2002, 8, 4475.
̈
1
1JCF = 260 Hz), 147.9 (1C, d(br), p-Ph, JCF = 269 Hz); 27Al NMR
(4) Weigand, J. J.; Burford, N.; Decken, A. Eur. J. Inorg. Chem. 2008,
(CD2Cl2, 25 °C, [ppm]): δ = 104 (s, ν1/2 = 470 Hz).
4343.
X-ray Data Collection and Reduction. Suitable single crystals
for 21b-g[GaCl4] and 14a·fluorobenzene were obtained by crystal-
lization from a minimum amount of CH2Cl2 (21b-g[GaCl4]) or
fluorobenzene (14a) by n-hexane diffusion at −30 °C. The crystals
were coated with Paratone-N oil, mounted using a glass fiber pin, and
frozen in the cold nitrogen stream of the goniometer. X-ray diffraction
(5) Weigand, J. J.; Holthausen, M. H.; Frohlich, R. Angew. Chem., Int.
Ed. 2009, 48, 295.
(6) Holthausen, M. H.; Weigand, J. J. J. Am. Chem. Soc. 2009, 131,
14210.
(7) Holthausen, M. H.; Richter, C.; Hepp, A.; Weigand, J. J. Chem.
Commun. 2010, 46, 6921.
̈
3386
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387

Inorganic Chemistry
Article
(8) (a) Alton, E. R.; Montemayor, R. G.; Parry, R. W. Inorg. Chem.
1974, 13, 2267. (b) Schultz, C. W.; Parry, R. W. Inorg. Chem. 1976, 15,
3046. (c) Kopp, R. W.; Bond, A. C.; Parry, R. W. Inorg. Chem. 1976,
15, 3042. (d) Thomas, M. G.; Schultz, C. W.; Parry, R. W. Inorg. Chem.
1977, 16, 994. (e) Shagvaleev, F. S.; Zykova, T. V.; Tarasova, R. I.;
Sitdikova, T. S.; Moskva, V. V. Zh. Obshch. Khim. 1990, 60, 1775.
(9) (a) Shagvaleev, F. S.; Zykova, T. V.; Tarasova, R. I.; Sitdikova,
T. S.; Moskva, V. V. J. Gen. Chem. USSR (Engl. Trans.) 1990, 60, 1585;
(b) Z. Obshch. Khim. 1990, 60, 1775. (c) Burford, N.; Ragogna, P. J.;
McDonald, R.; Ferguson, M. J. J. Am. Chem. Soc. 2003, 125, 14404.
(d) Burford, N.; Herbert, D. E.; Ragogna, P. J.; McDonald, R.;
Ferguson, M. J. J. Am. Chem. Soc. 2004, 126, 17067. (e) Burford, N.;
Dyker, C. A. M.; Lumsden, M.; Decken, A. Angew. Chem., Int. Ed.
2005, 44, 6196. (f) Dyker, C. A.; Burford, N.; Lumsden, M. D.;
Decken, A. J. Am. Chem. Soc. 2006, 128, 9632. (g) Weigand, J. J.;
Burford, N.; Lumsden, M.; Decken, A. Angew. Chem., Int. Ed. 2006, 45,
6733. (h) Riegel, S. D.; Burford, N.; Lumsden, M. D.; Decken, A.
Chem. Commun. 2007, 4469. (i) Carpenter, Y.; Dyker, C. A.; Burford,
N.; Lumsden, M. D.; Decken, A. J. Am. Chem. Soc. 2008, 130, 15732.
(10) Burford, N.; Ragogna, P. J. J. Chem. Soc., Dalton Trans. 2002,
4307.
(11) Cowley, A. H.; Kemp, R. A. Chem. Rev. 1985, 85, 367.
(12) Michalik, D.; Schulz, A.; Villinger, A.; Weding, N. Angew. Chem.,
Int. Ed. 2008, 47, 6465.
(13) (a) Dumitrescu, A.; Gornitzka, H.; Schoeller, W. W.; Bourissou,
D.; Bertrand, G. Eur. J. Inorg. Chem. 2002, 1953. (b) Reeske, G.;
Hoberg, C. R.; Hill, N. J.; Cowley, A. H. J. Am. Chem. Soc. 2008, 128,
2800. (c) Dube, J. W.; Farrar, G. J.; Norton, E. L.; Szekely, K. L. S.;
Cooper, B. F. T.; McDonald, C. L. B. Organometallics 2009, 28, 4377.
(d) Reed, R. W.; Xie, Z.; Reed, C. A. Organometallics 1995, 14, 5002.
(28) We assume that this also results from GaCl3 induced Friedel−
Crafts type arylation reactions.
(29) Ulvenlund, S.; Whaetley, A.; Bengtsson, L. A. J. Chem. Soc.,
Dalton Trans. 1995, 2, 245.
(30) Ulvenlund, S.; Bengtsson-Kloo, L.; Stahl, K. J. Chem. Soc.,
̊
Faraday Trans. 1995, 91, 4223.
(31) Gudat, D. Eur. J. Inorg. Chem. 1998, 1087.
(32) Schoeller, W. W. Phys. Chem. Chem. Phys. 2009, 11, 5273.
(33) The additionally added Lewis acid ECl3 (E = Ga, Al) lead to a
significant broadening of the corresponding 71Ga and 27Al NMR
resonances, for similar observations see ref 32.
(34) Hansch, C.; Leo, A. Exploring QSAR: Fundamentals and
Applications in Chemistry and Biology; American Chemical Society:
Washington, DC, 1995; Vols. 1 and 2.
(35) Chang, I. S.; Price, J. T.; Tomlinson, A. J.; Willis, C. Can. J.
Chem. 1972, 50, 512.
(36) Budzelaar, P. H. M. gNMR for Windows (5.0.6.0) NMR
Simulation Program; Cherwell Scientific, 2006.
(37) (a) Taft, R. W. J. Am. Chem. Soc. 1952, 74, 2729. (b) Taft, R. W.
J. Am. Chem. Soc. 1953, 75, 4231.
(38) (a) McFarlane, H. C. E.; McFarlane, W.; Nash, J. A. Dalton
Trans. 1980, 240. (b) Aime, S.; Harris, R. K.; McVicker, E. M.; Fild, M.
Dalton Trans. 1976, 2144. (c) Forgeron, M. A. M.; Gee, M.;
Wasylishen, R. E. J. Phys. Chem. 2004, 108, 4895. (d) Del Bene, J. E.;
Elguero, J.; Alkorta, I. J. Phys. Chem. 2004, 108, 3662.
(39) (a) Tattershall, B. W.; Kendall, N. L. Polyhedron 1994, 10, 1517.
(b) Baudler, M.; Adamek, C.; Opiela, S.; Budzikiewicz, H.; Ouzounis,
D. Angew. Chem., Int. Ed. 1988, 27, 1059. (c) Niecke, E.; Ruger, R.;
̈
Krebs, B. Angew. Chem., Int. Ed. Engl. 1982, 21, 544. (d) Jutzi, P.;
Meyer, U. J. Organomet. Chem. 1987, 333, C18.
(40) Weigand, J. J.; Riegel, S. D.; Burford, N.; Decken, A. J. Am.
Chem. Soc. 2007, 129, 7969.
(e) Schmidtpeter, A.; Jochem, G.; Klinger, C.; Robl, C.; Noth, H.
̈
3
J. Organomet. Chem. 1997, 529, 87.
(41) For values of JPF couplings see, e.g., (a) Westenkirchner, A.;
(14) Symmes, C.; Quin, L. D. J. Org. Chem. 1978, 43, 1250.
(15) (a) Henderson, W. A. Jr.; Streuli, C. A. J. Am. Chem. Soc. 1960,
82, 5791. (b) Laurence, C.; Gal, J.-F. Lewis Basicity and Affinity Scales:
Data and Measurement, 1st. ed.; John Wiley & Sons: London, U.K.,
2010; (c) Quin, L. D. A Guide to Organophosphorus Chemistry, 1st. ed.;
John Wiley & Sons: New York, 2000.
(16) (a) Kashman, Y.; Menachem, Y.; Benary, E. Tetrahedron 1973,
29, 4279. (b) Crews, P. J. Org. Chem. 1975, 40, 1170. (c) Kashman, Y.;
Rudi, A. Tetrahedron Lett. 1976, 32, 2819. (d) Rudi, A.; Kashman, Y.
Tetrahedron Lett. 1978, 25, 2209. (e) Cowley, A. H.; Stewart, C. A.;
Whittlesey, B. R.; Wright, T. C. Tetrahedron Lett. 1984, 25, 815.
(17) (a) Carpenter, Y. Investigation into the Reactivity and Structure of
Phosphinophosphonium Cations and Related Species; Ph.D. Thesis,
Dalhousie University, Halifax, NS, Canada, December 2010;
(b) Carpenter, Y.; Burford, N.; Lumsden, M. D.; McDonald, R.
Inorg. Chem. 2011, 50, 3342.
(18) Only species are reported whose concentration in solution is
significantly higher than 3% (determined by 31P{1H} NMR).
(19) Temperature dependent NMR investigations were not performed
because of the narrow liquid phase of fluorobenzene (m.p. −42.2 °C; b.p. 96.1
°C. Ginzburg, B. M.; Tuichiev, S. Russian J. Appl. Chem. 2009, 82, 1178.
(20) Gonsior, M.; Krossing, I.; Matern, E. Chem.Eur. J. 2006, 12, 1703.
(21) Burford, N.; Cameron, T. S.; LeBlanc, D. J.; Losier, P.; Sereda,
S.; Wu, G. Organometallics 1997, 4712.
(22) Norman, N. C.; Pickett, N. L. Coord. Chem. Rev. 1995, 145, 27.
(23) (a) Slattery, J. M.; Fish, C.; Green, M.; Hooper, T. N.; Jeffery,
J. C.; Kilby, R. J.; Lynam, J. M.; McGrady, J. E.; Pantazis, D. A.; Russel,
C. A.; Willians, C. E. Chem.Eur. J. 2007, 13, 6967. (b) Le, T. D.;
Arquier, D.; Miqueu, K.; Sotiropoulos, J.-M.; Coppel, Y.; Bastin, S.;
Igau, A. J. Organomet. Chem. 2009, 694, 229.
Villinger, A.; Karaghiosoff, K.; Wustrack, R.; Michalik, D.; Schulz, A.
Inorg. Chem. 2011, 50, 2691. (b) Hoge, B.; Panne, P. Chem.Eur. J.
2006, 12, 9025.
(42) (a) Miller, G. R.; Yankowsky, A. W.; Grim, S. O. J. Chem. Phys.
1969, 51, 3185. (b) Fluck, E.; Wachtler, D. Liebigs Ann. Chem. 1979, 1125.
(43) Xiong, Y.; Yao, S.; Brym, M.; Driess, M. Angew. Chem., Int. Ed.
2007, 46, 4511.
(44) (a) Niecke, E.; Ruger, R.; Krebs, B. Angew. Chem., Int. Ed. Engl.
̈
1982, 21, 544. (b) Schoeller, W. W.; Staemmler, V.; Rademacher, P.;
Niecke, E. Inorg. Chem. 1986, 25, 4382.
(45) Wolstenholme, D. J.; Weigand, J. J.; Davidson, R. J.; Pearson,
J. K.; Cameron, T. S. J. Phys. Chem. A 2008, 112, 3424.
(46) Weigand, J. J.; Burford, N.; Davidson, R. J.; Cameron, T. S.;
Seelheim, P. J. Am. Chem. Soc. 2009, 131, 17943.
(47) Sum of covalent radii: r(P) = 1.10, r(C) = 0.77, r(Cl) = 0.99 ;
Wiberg, N.; Wiberg, E.; Hollemann, A. F. Lehrbuch der Anorganischen
Chemie, 102nd ed.; Walter de Gruyter: Berlin, Germany, 2007; Anhang IV.
(48) Bondi, A. J. Phys. Chem. 1964, 68, 441.
(49) Henderson, A. W.; Buckler, S. A.; Day, N. E.; Grayson, M.
J. Org. Chem. 1961, 26, 4770.
(50) Fild, M.; Stelzer, O.; Schmutzler, R. Inorg. Synth. 1973, 14, 4.
(51) Magnelli, D.; Tesi, G.; Lowe, J. U.; McQuiston, W. E. Inorg.
Chem. 1966, 5, 457.
(52) SMART, program package; Bruker AXS: Madison, WI, 2000.
(53) SHELXS-97; Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(54) Sheldrick, G. M. SHELXL-97; Universitat Gottingen: Gottingen,
̈
̈
̈
Germany, 1997.
(24) Drago, R. S.; Joerg, S. J. Am. Chem. Soc. 1996, 118, 2654.
(25) Tolman, C. A. Chem. Rev. 1977, 77, 315.
(26) Nagy, G.; Balde, D. French Patent 1950691, Chem. Abstr. 1967,
67, 3142.
(27) Frank, W.; Gelhausen, B.; Reiss, G. J.; Salzer, R. Z. Naturforsch.
B.: Chem. Sci. 1998, 53, 1149.
3387
dx.doi.org/10.1021/ic2013304 | Inorg. Chem. 2012, 51, 3374−3387