Electrochemical synthesis of phosphine from the lower phosphorus acids

Article abstract of DOI:10.1134/S0020168509120152

We have studied the reduction of hypophosphorous and phosphorous acids at mercury and lead cathodes and with potassium and sodium amalgams. Both acids are found to be difficult to reduce. With the mercury cathode, the current efficiency for phosphine is w

Full text of DOI:10.1134/S0020168509120152

ISSN 0020-1685, Inorganic Materials, 2009, Vol. 45, No. 12, pp. 1388–1393. © Pleiades Publishing, Ltd., 2009.
Original Russian Text © V.V. Turygin, M.K. Smirnov, A.V. Khudenko, S.V. Nikolashin, V.A. Fedorov, A.P. Tomilov, 2009, published in Neorganicheskie Materialy, 2009, Vol. 45,
No. 12, pp. 1484–1490.
Electrochemical Synthesis of Phosphine
from the Lower Phosphorus Acids
V. V. Turygina, M. K. Smirnova, A. V. Khudenkoa, S. V. Nikolashinb,
V. A. Fedorovb, and A. P. Tomilova
a State Research Institute of Organic Chemistry and Technology, sh. Entuziastov 23, Moscow, 111024 Russia
b Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences,
Leninskii pr. 31, Moscow, 119991 Russia
e-mail: fedorov@igic.ras.ru
Received June 4, 2009
Abstract—We have studied the reduction of hypophosphorous and phosphorous acids at mercury and lead
cathodes and with potassium and sodium amalgams. Both acids are found to be difficult to reduce. With the
mercury cathode, the current efficiency for phosphine is within 2.5%. The phosphine yield with the Pb cathode
can be raised by adding sulfuric acid or ethylene glycol to the catholyte.
DOI: 10.1134/S0020168509120152
INTRODUCTION
standard thermodynamic reduction potentials of hypo-
phosphorous, phosphorous, and phosphoric acids on
the hydrogen scale [18]. Calculation results suggest
that phosphine can in principle be obtained by electro-
chemical reduction of these acids or their salts. The
three acids are rather close in reduction potential. It
would be expected that the energy input in phosphine
production will be minimized by using hypophospho-
rous acid. To reduce it to phosphine, an electric charge
of 4F per mole of the acid is sufficient.
Baudler and Bongardt [19] reported the reduction
potentials of phosphorus acid anions at a mercury drop-
ping electrode in glycerol solutions of the correspond-
ing sodium salts with tetraethylammonium perchlorate
as the polarographic background (Table 2). The hypo-
phosphite ion is reduced at a rather high negative poten-
tial (–0.64 eV). Phosphine is unlikely to form under
such conditions because the reduction requires less
than 1F per mole.
High-purity phosphine is widely used in the fabrica-
tion of heteroepitaxial structures of compound semi-
conductors. On a commercial scale, phosphine is com-
monly produced by reacting magnesium or zinc phos-
phides with acid solutions [1, 2]. This approach has
serious drawbacks because the resulting phosphine
contains large amounts of harmful impurities, which
can only be removed by sophisticated processes involv-
ing low-temperature column distillations. Moreover,
decomposition of phosphides by acids leads to the for-
mation of large amounts of difficult-to-process wastes.
The search for novel strategies that would allow synthe-
sis of sufficiently pure phosphine with a small amount
of waste is an important practical issue. In this context,
it is of interest to take advantage of electrochemical
synthesis, which has been shown to be effective in pro-
ducing arsine from various raw materials [3].
In the great majority of studies concerned with the
electrochemical synthesis of phosphine at relatively
high temperatures, white phosphorus has been used as
the raw material. The reduction of white phosphorus in
acid solutions was studied systematically in the 1960s
at Hooker Chemical Corp. (USA) [4–13] and also in the
United Kingdom [14] and West Germany [15–17].
High current efficiencies for phosphine were achieved
in concentrated phosphoric acid at 80–90°ë. With
decreasing temperature, the reduction efficiency
decreases even though the phosphorus remains liquid.
A serious drawback to this process is that white phos-
phorus is very toxic and can spontaneously ignite in air
at temperatures above 34°ë.
It is known [20] that coatings produced by electro-
chemical nickel deposition in the presence of sodium
hypophosphite contain phosphorus. The likely cathode
reaction is
ç2êO–2 + 2ç+ + –
ê + 2ç2é,
(1)
and the liberated phosphorus forms a phosphorus–
nickel alloy. It may be that the reduction of the hypo-
phosphite ion leads to the formation of phosphine,
which reacts instantaneously with nickel ions at the
cathode to form nickel phosphide in the coating.
Lurke [21] reported phosphine formation by the
reaction of hypophosphorous acid with hydrogen at the
The possibility of electrochemical phosphine synthesis “instant of release.” In addition, phosphine was found
from phosphorus acids is of some interest. Table 1 lists the to form via the reduction of hypophosphorous acid at
1388

ELECTROCHEMICAL SYNTHESIS OF PHOSPHINE
Table 1. Standard reduction potentials of phosphorus acids [18]
1389
Acid
Reaction
Standard potential, V
–0.174
H3PO2 + 4H+ + 4e–
PH3 + 2H2O
H3PO4
H3PO–2 + 5H+ + 4e–
H3PO3 + 6H+ + 6e–
PH3 + 2H2O
PH3 + 3H2O
–0.145
H3PO3
H3PO4
–0.282
H2PO–3 + 7H+ + 6e–
PH3 + 3H2O
–0.265
HPO23– + 8H+ + 6e–
H3PO4 + 8H+ + 8e–
PH3 + 3H2O
PH3 + 4H2O
–0.205
–0.281
–0.265
H2PO–4 + 9H+ + 8e–
HPO24– + 10H+ + 8e–
PO34– + 11H+ + 8e–
PH3 + 3H2O
PH3 + 4H2O
PH3 + 4H2O
–0.212
–0.123
lead, zinc, and mercury cathodes [22]. The reactivity of
the lead and zinc drops rapidly, and the phosphine yield
decreases. According to Baudler and Schellenberg [22],
the phosphine yield at a mercury cathode is rather high,
but they presented no quantitative data. Sodium hypo-
phosphite cannot be reduced under such conditions
[23].
H
H
HO P O
HO P OH
,
H
(I)
(2)
(II)
which contain penta- and trivalent phosphorus. Tau-
tomer I is nonreactive [25].
Under ordinary conditions, the reaction equilibrium
Ayers [24] described phosphine formation as a is displaced to I, and H3PO2 is titrated as a monobasic
acid. Hypophosphorous acid is a strong acid (K = 8.9 ×
10–2) and is almost fully dissociated in aqueous solu-
tions. Like hypophosphorous acid, phosphorous acid
exists in two forms: HP(O)(OH)2 and P(OH)3. Equilib-
rium is displaced to the former, which is most likely
also nonreactive [25].
result of phosphorous acid electrolysis, but he did not
specify the yield of the process. The lack of quantitative
data may indicate that he failed to achieve a good phos-
phine yield.
The objective of this work is to examine the possi-
bility of phosphine synthesis by electrochemical reduc-
tion of lower (hypophosphorous and phosphorous)
acids.
EXPERIMENTAL
Current–voltage measurements were made with an
IPC-2000 programmed potentiostat in a diaphragmless
cell thermostated at 25 1°ë. Before and during the
measurements, the free space of the cell was purged
with argon. The potential was scanned at a rate of
0.2 V/s and was measured with respect to a silver/silver
chloride reference electrode. We used stationary elec-
trodes from various materials. The auxiliary electrode
was of platinum. Prior to measurements, the electrodes
were polished. Their working surface area was 100 to
160 mm2.
The experiments were conducted in a filter-press
electrolyzer with Teflon frames. The cathodic and
anodic compartments were separated by a Nafion cat-
ion-exchange membrane. The membrane–electrode
separation was 2 to 5 mm, depending on the thickness
Hypophosphorous acid, H3PO2, is known to exist in
two tautomers,
Table 2. Polarographic reduction potentials of sodium salts
in glycerol with (C2H5)4NClO4 background electrolyte [19]
Reduction
potential, V
Quantity of elec-
tricity, F/mol
Salt
NaH2PO2 · H2O
NaH2PO3 · 2H2O
Na2HPO3 · 5H2O
Na2HPO4 · 2H2O
Na3PO4 · 6H2O
–0.64
–0.605
–0.42
–0.225
–0.31
0.7
0.4
0.7
0.5
0.3
INORGANIC MATERIALS Vol. 45 No. 12 2009

1390
TURYGIN et al.
For amalgam reduction, we used a cylindrical reac-
Ar
9
tor–electrolyzer system, shown schematically in Fig. 1.
The outer beaker contained an alkali solution, and the
inner beaker contained an acid solution. During elec-
trolysis, the alkali metal deposited on the mercury cath-
ode, and the forming amalgam was driven to the inner
space by stirring, where it decomposed. The volumes of
the outer and inner spaces were 30 and 50 cm3, respec-
tively.
PH + H
8
3
3
2
O
2
9
5
3
10
11
In preliminary experiments, we used hypophospho-
rous acid prepared from a sodium hypophosphite solu-
tion via ion exchange with the H-form of a cation-
exchange resin. In further studies of the reduction of
hypophosphorous acid, a 50% solution of the acid
(Aldrich) was used.
Phosphine in gas mixtures was determined by iodo-
metric titration of mercury(II) phosphochloride [26].
The forming gas was bubbled through a HgCl2 solution.
The phosphine was collected as the insoluble com-
plex(HgCl)3P, which was dissolved in an excess of a
titrated iodine solution after the experiment. The unre-
acted iodine was titrated with a thiosulfate solution.
1
2
6
7
4
45
mm
Fig. 1. Reactor–electrolyzer system for amalgam reduction:
(1) glass beaker, (2) inner glass tube, (3) rubber stoppers, (4)
mercury, (5) current lead, (6) platinum anode, (7) stirrer, (8)
inert gas inlet, (9) gas outlets, (10) solution of the substance
to be reduced, (11) electrolyte (alkali solution).
RESULTS AND DISCUSSION
Since data on the electrochemical behavior of the
lower phosphorus acids at solid electrodes have not
been reported in the literature, we performed current–
voltage measurements in aqueous 1.0 M H2SO4 with
lead, zinc, cadmium, copper, amalgamated copper,
mercury, and glassy carbon cathodes.
At the Pb cathode (Fig. 2), as the potential is
scanned from 0 to –2.0 V hydrogen release begins at
about –0.9 V and reaches 710 mA (current density,
0.42 A/cm2) at –2.0 V. In the presence of 0.1 M H3PO2,
the current increases considerably, up to 1000 mA.
of the frame. The electrodes were 1.0 × 2.0 cm in dimen-
sions. When a liquid cathode was used, the electrolyzer
was mounted horizontally, and a Teflon plate with holes
for mercury was used instead of the cathode frame. The
catholyte was circulated by a centrifugal pump, and the
anolyte (40% H3PO4), by a gas-lift system.
These results lead us to assume that hypophospho-
rous acid is reduced under such conditions. Under the
assumption that the currents due to the hydrogen
release and the reduction of H3PO2 follow the additivity
rule, the yield of the reduction process is about 30%. In
the conditions under consideration, sodium hypophos-
phite exhibits lower electrochemical activity (Fig. 3).
Phosphorous acid displays similar behavior (Fig. 4). In
the presence of hypophosphorous acid, we observe
slight depolarization at the copper cathode (Fig. 5).
0
1
–200
2
–400
3
–600
The voltammetry data obtained with the Pb cathode
is of most interest. To evaluate the stability of the
observed depolarization, we performed 20 sequential
cycles in the presence of 0.1M pH3PO2, with the poten-
tial varied from 0 to –2.0 V (Fig. 6). After each cycle,
the current was observed to gradually increase. After
the first ten cycles, a 10-min break was made, followed
by the next ten cycles, but the increase in current was
insignificant. As a result, a black spongy deposit was
produced on the Pb cathode. The observed increase in
current may be due to the following processes: the cath-
–800
–1000
–2000
–1500
–1000
–500
0
Ö, mV
Fig. 2. Voltammograms of a Pb cathode in 1 N H SO back-
ground electrolyte in the presence of H PO ; potential scan
from 0 to –2.0 V and to 0 at a rate of 0.2 V/s; (1) back-
ground, (2) background + 0.01 M H PO , (3) background +
0.1 M H PO .
2
4
3
2
3
2
3
2
INORGANIC MATERIALS Vol. 45 No. 12 2009

ELECTROCHEMICAL SYNTHESIS OF PHOSPHINE
1391
0
–200
–400
–600
3
1
2
–2000
–1500
–1000
–500
0
Ö, mV
Fig. 3. Voltammograms of a Pb cathode in 1 N H SO back-
Fig. 4. Voltammograms of a Pb cathode in 1 N H SO
2 4
2
4
ground electrolyte in the presence of NaH PO ; potential
background electrolyte in the presence of H PO ; potential
2
2
3 3
scan from 0 to –2.0 V and to 0 at a rate of 0.2 V/s; (1) back-
ground, (2) background + 0.01 M NaH PO , (3) back-
scan from 0 to –2.0 V and to 0 at a rate of 0.2 V/s; (1) back-
ground, (2) background + 0.01 M H PO , (3) background +
2
2
3
3
ground + 0.1 M NaH PO .
0.1 M H PO .
2
2
3 3
0.3 A/cm2. Typical results are presented in Table 3.
Unfortunately, we failed to reveal any general trends.
The current efficiency for phosphine ranged from 1.6 to
2.3%.
ode surface becomes activated, which either accelerates
the reduction of H3PO2 at the cathode or leads to cata-
lytic hydrogen release.
Similar measurements were performed with the
other electrode materials indicated above, but, in con-
In similar experiments with phosphorous acid (10 M
trast to Pb, none of them exhibited an activation effect. H3PO3; current density, 0.3 A/cm2), the current effi-
ciency for PH3 was 0.05% at 61–63°ë and within 0.5%
at 23–25°ë.
Given the earlier report [13] that phosphine can be
obtained with a good yield using a mercury cathode, we
studied the reduction of H3PO2 with the use of mercury.
We studied the reduction of hypophosphorous acid
The first series of experiments were performed using a with potassium and lithium amalgams. The amalgams
filter-press electrolyzer with horizontal electrodes were prepared directly during the experiment. The pro-
(platinum anode). The temperature was varied from 20 cess was run in the electrolyzer schematized in Fig. 1.
to 60°C, the hypophosphorous acid concentration from The outer beaker contained a concentrated alkali solu-
8 to 15 mol/l, and the current density from 0.15 to tion, and the inner beaker contained a hypophospho-
0
1
–500
–1000
3
2
–2000
–1000
–500
0
Ö, mV
Fig. 5. Voltammograms of a Cu cathode in 1 N H SO background electrolyte in the presence of H PO ; potential scan from 0 to
2
4
3
2
–1.5 V and to 0 at a rate of 0.2 V/s; (1) background, (2) background + 0.01 M H PO , (3) background + 0.1 M H PO
3
2
3
2.
INORGANIC MATERIALS Vol. 45 No. 12 2009

1392
TURYGIN et al.
0
–500
1
2
–1000
–1500
3
4
5
–2000
–1500
–1000
–500
0
Ö, mV
Fig. 6. Voltammograms of a Pb cathode in 1 NH SO background electrolyte in the presence of H PO ; potential scan from 0 to
2
4
3
2
−2.0 V and to 0 at a rate of 0.2 V/s. The first ten cycles (first stage) were followed by a 10-min break and then by the next ten cycles
(second stage): (1) background; (2) background + 0.1 M H PO , first cycle (first stage); (3) background + 0.1 M H PO , tenth cycle
3
2
3
2
(first stage); (4) background + 0.1 M H PO , first cycle (second stage); (5) background + 0.1 M H PO tenth cycle (second stage).
3
2
3
2,
rous acid solution. During electrolysis, the alkali metal
deposited on the mercury cathode, and the forming
amalgam was driven to the inner space by stirring,
where it reacted with the acid to form the corresponding
hypophosphite, hydrogen, and phosphine.
At 60–70°C, potassium amalgam reacted with a
8.06 M H3PO2 solution with a 0.7% current efficiency
for the preparation of the amalgam. The current effi-
ciency for lithium amalgam was 0.9%. The exact cur-
rent efficiency for this process cannot be evaluated
because the alkali metal deposited on the amalgam
cathode with a rather low yield. Given this, the current
efficiency for PH3 does not exceed the level reached
earlier by direct reduction of H3PO2 at a mercury cath-
ode (~2%).
Table 3. Electrochemical reduction of hypophosphorous ac-
id at a mercury cathode (Pt anode, 40% H3PO4 anolyte)
Current
Thus, our attempts at indirect electrochemical
reduction of hypophosphorous acid were unsuccessful.
Electrolysis conditions
efficiency, %
H3PO2
concentra-
No.
tion in the
The results obtained in the filter-press electrolyzer
with solid cathodes were inferior to those with a mer-
cury cathode, so the reduction of hypophosphorous
acid at a Pb cathode was studied in detail.
catholyte, M
current den- tempera-
PH3
H2
sity, A/cm2 ture, °C
Varying the electrolysis parameters in the same
ranges as with the mercury cathode, we obtained only
trace levels of phosphine. In view of this, conditions
were created for decreasing the degree of H3PO2 dissocia-
tion by reducing hypophosphorous acid in ethylene gly-
col, a water–glycol mixture, and water with an addition of
sulfuric acid. The results are presented in Table 4.
1
8.06
0.3
0.3
0.3
0.15
0.3
55–61 2.1
94
89
88
92
88
2
15.5
55–61 1.9
3
8.06
8.06
8.06
21–24 2.16
56–58 1.59
55–61 2.29
It can be seen that ethylene glycol markedly raises
the phosphine yield (from trace levels to 5.5%), the
effect being independent of the presence of water and
current density.
4
The addition of sulfuric acid, though less effective,
slightly increases the yield. It is of interest to note that
the phosphine yield has a tendency to rise with decreas-
ing current density, but even at a low current density the
current efficiency for PH3 is about 1%.
5*
* With the addition of 1% (C H ) NCl.
2
5 4
INORGANIC MATERIALS Vol. 45 No. 12 2009

ELECTROCHEMICAL SYNTHESIS OF PHOSPHINE
1393
Table 4. Electrochemical reduction of hypophosphorous acid at a Pb cathode (Pt anode, 40% H3PO4 anolyte)
Electrolysis conditions
Current efficiency, %
No.
Initial catholyte composition
current densi- temperature,
PH3
H2
ty, A/cm2
°C
1
2
11.3 M H3PO2 in H2O
0.15
0.30
19–20
60
Traces
5.5
100
94
5.7 M H3PO2 in a 2 : 1 mixture of C2H4(OH)2
and H2O
3
4
5
6
5.7 M H3PO2 in C2H4(OH)2
0.15
0.2
55
5.3
95
99
97
96
10.4 M H3PO2 + 4.5 M H2SO4 in H2O
21–24
21–24
21–24
0.23
0.56
1.03
0.1
0.05
4. Gordon, I., US Patent 3109785, 1963.
CONCLUSIONS
5. Miller, G.T. and Steingart, J., US Patent 3109786, 1963.
6. Thornton, P.D. and Gordon, I., US Patent 3109787,
1963.
7. Miller, G.T. and Steingart, J., US Patent 3109788, 1963.
8. Miller, G.T. and Steingart, J., US Patent 3109789, 1963.
9. Miller, G.T., US Patent 3109790, 1963.
10. Gordon, I., US Patent 3109791, 1963.
11. Gordon, I., US Patent 3109792, 1963.
12. Gordon, I., US Patent 3109793, 1963.
13. Gordon, I., US Patent 3109794, 1963.
14. Miller, G.T., UK Patent 1042391, 1963.
15. Miller, G.T., FGR Patent 1210424, 1966.
16. Miller, G.T., FGR Patent 1210425, 1966.
17. Miller, G.T., FGR Patent 1210426, 1966.
18. Tomilov, A.P and Chomutov, N.E, Phosphorus, Encyclo-
pedia of Electrochemistry of the Elements, Bard, A.J.,
Ed., New York: Marcel Dekker, 1975, vol. 3, pp. 1–37.
We studied the electrochemical reduction of the
lower phosphorus acids. The results demonstrate that
both hypophosphorous and phosphorous acids are dif-
ficult to reduce. The electrochemical reduction of these
acids occurs with low current efficiencies. The effi-
ciency can be raised by suppressing their dissociation.
The highest current efficiency for phosphine (5.5%) is
achieved with a Pb cathode in an ethylene glycol solu-
tion of hypophosphorous acid.
Most likely, the reduction involves undissociated
H3PO2 molecules in active form (II). To raise the phos-
phine yield, further research is needed with the aim of
increasing the concentration of the active form of the acid.
The depolarization of the Pb cathode in the presence
of hypophosphorous acid, evidenced by current–volt-
age data, is obviously due to catalytic hydrogen release.
19. Baudler, M. and Bongardt, A.S., Zur Kenntnis des
polarographischen Verhaltens der Salze von Mono- und
Diphosphorsäuren in Glicerin, Z. Anorg. Allg. Chem.,
1967, vol. 350, pp. 186–201.
ACKNOWLEDGMENTS
This work was supported by the Russian Foundation
for Basic Research, project no. 07-03-00343-a.
20. Vakhidov, R.S., Popov, V.I., and Starchenko, A.A., Peri-
odic Effects in Electrodeposition of Nickel–Phosphorus
Alloys, Elektrokhimiya, 1970, vol. 6, no. 11, pp. 1720–
1722.
REFERENCES
1. Zhigach, A.F. and Stasinevich, D.S., Khimiya gidridov
(The Chemistry of Hydrides), Leningrad: Khimiya,
1969.
21. Lurke, R., Z. Chem. Unter., 1986, vol. 3, p. 281.
22. Baudler, M. and Schellenberg, D., Elektrolytische Unter-
suchungen von Phosphorsäuren in wäβriger Lösung, Z.
Anorg. Allg. Chem., 1965, vol. 340, pp. 113–125.
23. Sundermeyer, W., FRG Patent 1 080 077, 1960.
24. Ayers, W.M., US Patent 5 158 656, 1002.
2. Devyatykh, G.G. and Zorin, L.D., Letuchie neorgan-
icheskie gidridy osoboi chistoty (Ultrapure Volatile Inor-
ganic Hydrides), Moscow: Nauka, 1974.
3. Shalashova, N.N., Smirnov, M.K., Turygin, V.V., et al.,
Electrolytic Formation of Phosphine from Red Phospho-
rus in Aqueous Solutions, Neorg. Mater., 2006, vol. 42,
25. Van Wazer, J.R., Phosphorus and Its Compounds,, New
York: Interscience, 1958.
no. 3, pp. 278–283 [Inorg. Mater. (Engl. Transl.), vol. 42, 26. Charlot, G., Les methodes de la chimie analytique, Paris:
no. 3, pp. 236–241].
Masson, 1961.
INORGANIC MATERIALS Vol. 45 No. 12 2009