Solvent-free mechanochemical synthesis of phosphonium salts

Article abstract of DOI:10.1039/b111515d

Phosphonium salts have been prepared during high-energy ball-milling of triphenylphosphine with solid organic bromides; the reactions occur at ambient conditions without a solvent; in the case of 2-bromo-2-phenylacetophenone the reaction in a solution usually produces a mixture containing both the C-phosphorylated and O-phosphorylated compounds, while the solvent-free mechanically induced transformation results in the thermodynamically favorable C-phosphorylated product; the occurrence of the observed transformations during mechanical processing of solid reactants is confirmed by the solid-state ³¹P NMR spectroscopy and X-ray powder diffraction.

Full text of DOI:10.1039/b111515d

Solvent-free mechanochemical synthesis of phosphonium salts†
Viktor P. Balema,* Jerzy W. Wiench, Marek Pruski and Vitalij K. Pecharsky*
Ames Laboratory, Iowa State University, Ames, IA 50011-3020, USA. E-mail: balema@ameslab.gov
Department of Materials Science and Engineering, Iowa State University, Ames, IA 50011, USA.
E-mail: vitkp@ameslab.gov
Received (in Corvallis, OR, USA) 17th December 2001, Accepted 7th February 2002
First published as an Advance Article on the web 6th March 2002
1
Phosphonium salts have been prepared during high-energy
ball-milling of triphenylphosphine with solid organic bro-
mides; the reactions occur at ambient conditions without a
solvent; in the case of 2-bromo-2-phenylacetophenone the
reaction in a solution usually produces a mixture containing
both the C-phosphorylated and O-phosphorylated com-
pounds, while the solvent-free mechanically induced trans-
formation results in the thermodynamically favorable C-
phosphorylated product; the occurrence of the observed
transformations during mechanical processing of solid
reactants is confirmed by the solid-state 31P NMR spectros-
copy and X-ray powder diffraction.
yield. Its H and 31P NMR spectra in CDCl3 and the melting
point10 were identical with those of the commercial product.
(1)
The success in the mechanochemical solid-state alkylation of
triphenylphosphine with 2-bromoacetophenone encouraged us
to study its mechanochemical reaction with 2-bromo-2-phenyl-
acetophenone. It was reported previously that 2-bromo-2-phe-
nylacetophenone reacts with triphenylphosphine in solution
yielding a mixture of C- and O-phosphorylated reaction
products.9 The mechanochemical reaction was carried out
following the procedure used for the preparation of salt 1 (eqn.
(1)). Again, the solid-state 31P{1H} CP MAS spectrum of the
ball-milled sample contains only one peak centered at d31P =
26 ppm (Fig. 1), consistent with the position of the signal in the
NMR spectrum of pure (a-benzoyl-a-phenylmethylene)triphe-
nylphosphonium bromide 2 (eqn. (1)). The latter was obtained
in 99% yield, recrystallized from a methanol–diethyl ether
solution and characterized by liquid-state 1H, 31P, and 13C NMR
spectroscopy and elemental analysis.10 The formation of the C-
phosphorylated compound 2 in almost quantitative yield shows
that the mechanochemical solvent-free technique clearly dis-
criminates between the C- (Arbuzov) and O-phosphorylation
(Perkow) pathways in favor of the more thermodynamically
stable Arbuzov-type product. Using the same mechanochemical
approach, we also succeeded in the preparation of two other
phosphonium salts 3 and 4, (eqn. (1)) from corresponding solid
organic bromides. While 2-(naphthalenylmethyl)triphenylphos-
phonium bromide 3 easily forms during ball-milling of the
reactants in hardened-steel equipment for one hour, the
preparation of propane-1,3-diylbis(triphenylphosphonium) di-
bromide 4 requires mechanical processing in a tungsten carbide
Mechanical processing (ball-milling) of solids is a scalable
technique, which is routinely used in materials science1,2 and in
the pharmaceutical industry.3 However, mechanical activation
is seldom attempted in synthetic organic chemistry and when it
is, mechanical treatment is carried out in the presence of a
solvent or is followed by additional processing, e.g., heating.4–6
Furthermore, as far as we are aware, mechanochemical
transformations of organo-phosphorus compounds were pre-
viously unknown.
Here we report the solvent-free preparation of phosphonium
salts during mechanical processing of solid reactants at ambient
conditions. Traditionally, phosphonium salts are synthesized in
organic solutions in a typical SN2 quaternization reaction of
ternary phosphines with organic halides.7,8 These transforma-
tions may be complicated by side-reactions, and thus become
unsuitable for the synthesis of desired compounds. Especially
unreliable are reactions of ternary phosphines with a-bromoke-
tones, where the alkylation of phosphines is accompanied by the
formation of O-phosphorylated products and by the dehydro-
bromination of starting ketones.7–9
To explore the reactivity of triphenylphosphine with solid
organic bromides during mechanical processing in a solvent-
free environment, triphenylphosphine and 2-bromoacetophe-
none, taken in a 1+1 molar ratio, were sealed under helium in a
hardened-steel vial with steel balls and milled for one hour in a
Spex-8000 mill. Ball-milling for one hour was found to be
sufficient for the completion of this reaction.‡ The obtained
powder was analyzed by solid-state 31P{1H} cross polarization
magic angle spinning nuclear magnetic resonance (CP MAS
NMR) spectroscopy and X-ray powder diffraction. Commercial
phenacyltriphenylphosphonium bromide 1 (Lancaster Synthe-
sis) was used as a reference. The solid-state 31P{1H} CP MAS
spectra of both the powder obtained after mechanical processing
and commercial phosphonium salt, are shown in Fig. 1. They
contain only one signal at d31P = 23 ppm, indicating a complete
transformation of triphenylphosphine (d31P = 25 ppm) and
2-bromoacetophenone into salt 1 (eqn. (1)). Also, the X-ray
powder diffraction pattern of the mechanochemically prepared
compound 1† is in an excellent agreement with that of the
commercial phosphonium salt. For an additional purification,
the mechanically prepared phosphonium bromide 1 was
dissolved in hot water and isolated after crystallization in 90%
Fig. 1 The solid-state 31P{1H} CP MAS NMR spectra of the mechano-
chemically prepared bromide 1 (a), the commercial salt 1 (Lancaster
Synthesis) (b) and the phosphonium bromide 2 prepared mechanochem-
ically (c).
† Electronic supplementary information (ESI) available: X-ray diffraction
patterns of 1. See http://www.rsc.org/suppdata/cc/b1/b111515d/
724
CHEM. COMMUN., 2002, 724–725
This journal is © The Royal Society of Chemistry 2002

View Article Online
vial with tungsten carbide balls for 12.5 h. In both cases,
phosphonium salts were isolated in high, or moderate yields (3:
95%, 4+51%) and the comparison10 with the commercially
available products confirmed their formation.
Division (VPB, and VKP) and Chemical Sciences Division
(JWW and MP) of the U.S. DOE.
Several models describing mechanically induced reactions
between triphenylphosphine and organic halides can be sug-
gested. One of the probable mechanisms when low-melting
halides react with triphenylphosphine is a local formation of
low-melting eutectics in the triphenylphosphine–organic bro-
mide systems during ball-milling. In this case, the reactions may
occur in the melt, which forms locally and momentarily in the
areas where the rapidly moving balls collide with both the walls
of the reaction vial and with one other.
To verify the possibility of the eutectics formation in the
triphenylphosphine–bromobenzophenone systems, we studied
two mixtures: triphenylphosphine and 4A-bromobenzophenone
(A), and triphenylphosphine, 4’-bromobenzophenone and salt
1(B), both in an equimolar ratio. Mixture A was ball-milled for
one hour, or melted and then cooled to room temperature.
Mixture B was ball-milled for one hour. Since both 2-bromo-
acetophenone and 4’-bromobenzophenone melt at nearly the
same temperature, 48–50 and 50–52 °C, respectively, and the
chemical structures of these bromoketones are almost identical,
it is reasonable to expect that their eutectics with triphenylphos-
phine (mp 80–82 °C) will show similar melting behavior.
Furthermore, while 2-bromoacetophenone immediately reacts
with triphenylphosphine in the melt forming salt 1, 4A-
bromobenzophenone is inert towards the same phosphine
providing a good model for understanding of the temperature-
induced phase transformations in the triphenylphosphine–
2-bromoacetophenone system.
Notes and references
‡ The starting and reference compounds were purchased from Aldrich or
Lancaster Synthesis. Solid-state 31P{1H} CP MAS NMR spectra were
collected on a Chemagnetics Infinity 400 MHz spectrometer equipped with
a Chemagnetics MAS probe using a variable amplitude cross polarization
scheme (contact time of 0.5 ms, relaxation delay of 5 s, and spinning rate of
20 kHz). 1H, 13C and 31P NMR spectra in CDCl3 or C6D6 were obtained
using Varian VXR-300 and Varian VXR-400 spectrometers. Chemical
shifts are reported with respect to the 85% solution of H3PO4 in water (31P)
or TMS (1H and 13C) as the external standards. Ball-milling (for 15 min to
12.5 h) of various quantities of materials, usually about 1.0 g total, was
performed in a Spex-8000 mill using 21 g of steel balls in a hardened-steel
vial, or 70 g of tungsten carbide balls in a tungsten carbide (CW) vial sealed
under helium as protective atmosphere. Processing times under one hour
usually resulted in lower yields. Chemical reactions described in this work
do not continue after ball-milling has been stopped. Forced-air cooling of
the vial was employed to prevent its heating during ball-milling experi-
ments.
1 C. Suryanarayana, Progr. Mater. Sci., 2001, 46, 1.
2 V. P. Balema, K. W. Dennis and V. K. Pecharsky, Chem. Commun.,
2000, 1665.
3 T. P. Shakhtshneider and V. V. Boldyrev, Mol. Solid State, 1999, 3,
271.
4 F. Toda, Acc. Chem. Res., 1995, 28, 480.
5 V. D. Makhaev, A. P. Borisov and L. A. Petrova, J. Organomet. Chem.,
1999, 590, 222.
6 M. Nüchter, B. Ondruschka and R. Trotzki, J. Prakt. Chem., 2000, 342,
720.
7 W. A. Johnson, Ylides and imines of phosphors, John Wiley & Sons,
Inc., New York, 1993.
8 The Chemistry of Organophosphorus Compounds vol. 3, Phosphonium
Salts, Ylides and Phosphoranes., ed. F. R. Hartley, John Wiley & Sons,
Ltd, Chichester, New York, Brisbane, Toronto, Singapore, 1994.
9 J. Borowitz, P. E. Rusek and R. Virkhaus, J. Org. Chem., 1969, 34,
1995.
In the case of A the onset of the first endothermic (melting)
event is observed at 36 °C, and the second melting event at ~ 50
°C. Therefore, a low-melting eutectic indeed exists in the binary
triphenylphosphine–4A-bromobenzophenone system. The onset
of melting in B (38 °C) is close to that of A, indicating that the
presence of the reaction product should have a minor effect (if
any) on the reactions described above.
10 Mechanochemically prepared phosphonium compounds. 1: yield 90%;
solid-state 31P{1H} CP MAS NMR: d31P (ppm): 23.0; liquid-state NMR
(CDCl3): d31P (ppm): 23.0; d1H/JP-H (ppm/Hz): 8.38/7.4 (d, 2H, Ph),
8.01–7.90 (m, 6H, Ph), 7.67–7.74 (m, 4H, Ph), 7.67–7.54 (m, 6H, Ph),
7.52–7.45 (m, 2H, Ph), 6.40/12.2 (d, 2H, CH2); mp 270–273 °C; 2: yield
99%; solid-state 31P{1H} CP MAS NMR: d31P (ppm): 26.0; liquid-state
NMR (CDCl3): d31P (ppm): 27.4; d1H/JP–H (ppm/Hz): 9.30/12.4 (d, 1H,
CH), 8.42 (m, 2H, Ph), 8.02–7.96 (m, 6H, Ph), 7.71–7.67 (m, 3H, Ph),
7.60–7.55 (m, 6H, Ph), 7.46–7.35 (m, 5H, Ph), 7.22–7.19 (m, 1H, Ph),
7.15–7.11 (m, 2H, Ph); d13C{1H}/JC–P (ppm/Hz): 195.36/3.6 (d, CO),
135.57/9.7 (d, Ph), 134.41 (m, Ph), 133.78/5.2 (d, Ph), 131.89/5.2 (d,
Ph), 131.19 (s, Ph), 129.75/13.2 (d, Ph), 129.51/2 (d, Ph), 129.17/2 (d,
Ph), 128.86 (s, Ph), 128.83/5.9 (d, Ph), 118.78/86.6 Hz (d, CP, Ph),
54.30/54.9 (d, CP). Found (%): C 70.74, H 5.19%. Calculated for
C32H26OPBr (%): C 71.50, H 4.84%; mp 246–248 °C (decomposition);
3: yield 95%; solid-state 31P{1H} CP MAS NMR: d31P (ppm): 25.0;
liquid-state NMR (CDCl3): d31P (ppm): 24.4; d1H/JP-H (ppm): 7.98–7.1
(m, 22H, Aryl), 5.6/12.4 (d, 2H, CH2); mp 239–241 °C; 4: yield 51%;
liquid-state NMR (CDCl3): d31P (ppm): 25.4; d1H (ppm): 7.87–7.83 (m,
10H, Ph ), 7.71–7.57 (m, 20H, Ph), 4.64–4.55 (m, 4H, CH2), 2.3–1.83
(m, 2H, CH2) mp 355–360 °C (decomposition).
The existing data on the local temperature rise in different
types of mills during mechanical processing of inorganic solids1
and our results11 on mechanical treatment of LiAlH4 are
consistent with the ‘localized melting’ hypothesis. Based on the
fact that prolonged ball-milling of pure LiAlH4, which irreversi-
bly loses hydrogen at 150 °C and above, does not result in
noticeable change in this material,11 the local temperature in a
Spex-8000 mill cannot exceed 150 °C. The temperature in the
vial, however, can locally reach ~ 35–40 °C, which is sufficient
for the formation of liquid phases in some of the described
processes.
The mechanochemical formation of the phosphonium salt 4
from high melting (mp 228–230 °C) 3-bromopropyl(triphenyl-
phosphonium) bromide and triphenylphosphine is, however,
difficult to explain by the localized melting mechanism12 and,
therefore, alternative scenarios should be also considered. One
such possibility is that this chemical transformation proceeds in
the solid-state during ball-milling. In this case, the mechanical
treatment creates conditions sufficient for close contact and
subsequent chemical interactions between reacting compounds
and provides the mass transport required for the transformation
to occur. Experiments to verify this hypothesis and establish
possible activation periods are under way, and their results will
be published elsewhere.
In summary, the results described above indicate that solvent-
free mechanochemistry has a potential to become an alternative
to conventional organic synthesis. Continuing research to
establish the suitability of the mechanochemical synthesis to
different reaction types may lead to the development of novel
green technologies13 once its mechanism is better understood.
Ames Laboratory is operated by Iowa State University for the
U.S. Department of Energy (DOE) under contract No. W-
7405-ENG-82. Different aspects of this work were supported by
the Office of Basic Energy Sciences, Materials Sciences
11 V. P. Balema, K. W. Dennis and V. K. Pecharsky, J. Alloys Compd.,
2000, 69, 313.
12 The solvent-free preparation of 4 was achieved only after steel balls
were replaced by heavier CW balls thereby increasing the input of
mechanical energy into the system.1 It is feasible that additional energy
is required due to high melting temperatures of triphenylphosphine and
3-bromopropyl(triphenylphosphonium) bromide. However, the local-
ized melting mechanism offers no explanation for the fact that after a
prolonged ball-milling of these reactants in the presence of an excess of
Ph3P (which is the low melting component in this system) the starting
bromide still remains in the reaction mixture.
13 Solvent-free mechanochemistry may appear to be an energy intensive
technique when compared with the conventional solvent-based chem-
ical synthesis. However, considering life cycles of both processes, e.g.
the energy required to produce, deliver, collect and dispose the solvents,
and restore the environment, the advantages of the mechanochemical
approach are noteworthy.
CHEM. COMMUN., 2002, 724–725
725