A new approach to the synthesis of molybdenum phosphide via internal oxidation and reduction route

Article abstract of DOI:10.1016/j.jallcom.2008.05.048

A new method for the preparation of molybdenum phosphide was reported. Dispersed MoP with lamellar morphology was obtained from the decomposition of a mixed-salt precursor containing (NH₄)₄Mo₇O₂₄·2H₂O

Full text of DOI:10.1016/j.jallcom.2008.05.048

Journal of Alloys and Compounds 473 (2009) L10–L12
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Letter
A new approach to the synthesis of molybdenum phosphide via internal
oxidation and reduction route
Z.W. Yaoa,∗, Li Wangb, Haitao Dongb
a College of Chemical Engineering, Sichuan University, Chengdu 610065, PR China
b College of Chemical Engineering, Dalian University of Technology, Dalian 116024, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 April 2008
Received in revised form 12 May 2008
Accepted 13 May 2008
Available online 26 June 2008
A new method for the preparation of molybdenum phosphide was reported. Dispersed MoP with
lamellar morphology was obtained from the decomposition of a mixed-salt precursor containing
(NH4)4Mo7O24·2H2O, (NH4)2HPO4 and hexamethylenetetramine under argon. During the thermal
decomposition, reduction of molybdenum and phosphorous species by methane from the hex-
amethylenetetramine ligand occurred, with release of CO and CO2. Compared with traditional
H2-temperature-programmed reduction method, the method developed provided a simple one-step way
to molybdenum phosphide with relatively high surface area.
Keywords:
Inorganic materials
Chemical synthesis
Gas–solid reactions
Scanning electron microscopy
X-ray diffraction
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
metal phosphides were found to show relatively high surface areas,
suggesting that the presence of carbon in metal phosphides was
Transition metal phosphides are currently of great inter-
est in chemistry and materials sciences, because they exhibit
various interesting properties, including superconductivity, mag-
netrocaloric behavior, magnetoresistance and catalytic activity
[1–5]. In particular, recent studies have shown that MoP and
Ni2P are synthesized from ammonium molybdate and ammo-
nium phosphate by H2-temperature-programmed reduction (TPR)
method and show high activity for hydrodesulfurization (HDS) and
hydrodenitrogenation (HDN) of petroleum feedstocks [6,7]. Unfor-
tunately, the surface area of thus-prepared metal phosphides are
very low, because the P O bond is strong and its reduction requires
high temperature. Therefore, it is important to develop simple syn-
thetic approaches that yield bulk transition metal phosphides with
higher surface areas. Despite, a variety of methods have emerged
for the synthesis of these materials [8–12], the possibility of a range
of preparative routes being available, very few others have been
explored. Recently, Prins and co-workers reported that the bulk
Ni2P was synthesized via hydrogen reduction of the phosphate pre-
cursor in the presence of polymer surfactants and ethylene glycol
[13]. In addition, Burns et al. reported the application of methane
as a reductant in the preparation of bulk MoP [14]. Both binary
advantageous in preventing sintering of phosphides at high tem-
perature.
In recent studies, the nitrides and carbides of transition met-
als can be obtained on the use of hexamethylenetetramine (HMT)
as a reductant [15–18]. Based on the success with the HMT route
to synthesize metal nitrides and carbides, we decided to exam-
ine whether this simple procedure can also be used to produce
metal phosphides. In the present study, MoP was prepared in a
one-step synthesis, in which a mixed-salt precursor containing
(NH4)4Mo7O24·2H2O, (NH4)2HPO4 and HMT was directly con-
verted to phosphide under a flow of argon at 750 ◦C.
2. Experimental
An aqueous solution of (NH4)4Mo7O24·2H2O, (NH4)2HPO4 and HMT was mixed
with a mole ratio of 1:7:15; a few drops of NH3·2H2O were added in order to dissolve
the precipitate. The solution was then evaporated slowly to dryness, and a white
solid was obtained. After drying at 120 ◦C for 12 h, the solid was heated in a quartz
reactor under a flow of argon (50 mL/min). The temperature was increased linearly
at a rate of 10 ◦C/min and then kept at a given value for 2 h, followed by cooling to
room temperature (RT) under argon flow, and then passivated at RT in a stream of
1%O2/He. Three samples have been obtained from heating at 550, 650 and 750 ◦C,
designated, respectively, as sample-550, sample-650 and sample-750.
X-ray diffraction (XRD) examination was carried out on a X-ray diffractome-
ter (Rigaku D-Max Rotaflex) with Cu K␣ radiation and Ni filter. Scanning electron
microscopy (SEM) image was obtained using a KYKY 1000 B scanning electron
microscope with an energy dispersive X-ray spectroscopic (EDS) system. The BET
surface areas of passivated samples were measured on an ASAP 2010 instrument.
∗
Corresponding author. Tel.: +86 28 85405220; fax: +86 28 85461108.
E-mail address: mezhiwei@163.com (Z.W. Yao).
0925-8388/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2008.05.048

Z.W. Yao et al. / Journal of Alloys and Compounds 473 (2009) L10–L12
L11
Fig. 1. XRD patterns of the products of mixed-salt precursor thermal decomposition
in Ar flow at 550, 650 and 750 ◦C.
N
2 was used for standard five-point BET surface area measurements. The gases emit-
ted during the decomposition were monitored using a mass spectrometer (MS, HP
G1800A) and an infrared absorption spectrometer (IRAS, SICK-MAIHAK-S710). The
signals of m/e = 2, 16, 17, 18, and 28 and the formation rates of CO, CO2 and CH4 in
the gas phase effluent were continuously detected by MS and IRAS, respectively.
3. Results and discussion
The XRD patterns (Fig. 1) show that the sample-550 and
sample-650 were completely amorphous, whereas the sample-750
demonstrated the formation of MoP crystalline phase. The lines
at 27.9◦, 32.0◦, 42.9◦, 57.0◦, 64.7◦, 67.4◦, 74.0◦ and 85.5◦ corre-
sponded to the ꢀ0 0 1ꢁ, ꢀ1 0 0ꢁ, ꢀ1 0 1ꢁ, ꢀ1 1 0ꢁ, ꢀ1 1 1ꢁ, ꢀ2 0 0ꢁ, ꢀ2 0 1ꢁ
and ꢀ1 1 2ꢁ reflections of MoP, respectively. Subsequent BET analy-
sis of this sample yielded a surface area of 10 m2 g−1. Fig. 2 shows
the SEM image and EDS result of the sample-750. Observation of
Fig. 2a showed that the sample-750 consisted of submicrometer-
sized thin sheets. The EDS results for the sample-750 are shown
in Fig. 2b and give a P/Mo mole ratio of 0.95, which was a little
lower than theoretical value. Wang et al. reported that, although
the preparations were carried out with stoichiometric quantities of
metal and phosphorus, the final products tended to be metal-rich
due to the loss of phosphorus in the preparation processes [19].
Fig. 3a shows the MS of gaseous products from the decompo-
sition of the mixed-salt precursor during linear heating in argon.
Formation of gaseous products began 100 ◦C and was mostly com-
pleted at 800 ◦C. At low temperatures, the release of H2O and NH3
had coinciding maxima at 200 and 250 ◦C. Simultaneously, the for-
mation of NH3 decomposed via sequential dehydrogenation, giving
H2 and NHx (x = 0, 1, 2). With increasing temperature, the detection
of the intense H2 signal at 530 ◦C was due to the decomposition of
CH4. Additionally, we observed a very weak signal of CH4 (m/e = 16)
at 750 ◦C, which coincided with the signals of H2 (m/e = 2) and N2 or
CO (m/e = 28). As shown in Fig. 3b, the formation rates of CO, CO2 and
CH4 in the outflow gas are determined by IRAS. The CO formation
was observed with a weak peak at 300 ◦C, followed by the formation
of large amount of CO and low levels of CO2 at 750 ◦C. In addition,
we detected the CH4 formation with an intense peak at 260 ◦C but
there was no obvious peak at 750 ◦C, possibly due to the detec-
tion limit for IRAS analysis of CH4. The results suggested that the
m/e = 16 signal below 500 ◦C in the mass spectra (Fig. 3a) was partly
attributed to the product of CH4, which could react with a portion
of molybdenum and phosphorous species to give CO. During the
Fig. 2. (a) SEM image of MoP particle. (b) EDS result for the particle.
decomposition, reduction of the last compound occurred with the
simultaneous emission of CO and CO2 at higher temperature. Form
the results of XRD (Fig. 1), it can be observed that the minimum
temperature (750 ◦C) at which the molybdenum phosphide was
formed as the crystalline phase. Thus, it was reasonable to deduce
that the internal oxidation and reduction process of the mixed-
salt precursor was finally completed at 750 ◦C, which was in good
agreement with the phenomena observed by MS and IRAS. It was
worthy to note that the formation of CO and CO2 certainly argued for
oxidation of a carbonaceous species, but it was not possible to con-
firmative whether it was carbon or CH4 from HMT decomposition.
In view of point that the molybdenum and phosphorous species
could be reduced to MoP by carbon, carbon-supported molybde-
num phosphate with Mo loading of 30 wt.% and Mo/P ratio of 1/1
was prepared by pore-volume impregnation of a activated carbon
support. The carbon-supported precursor was heated and passi-
vated according to the procedure for the preparation of sample-750
to produce the carbon-supported sample. Subsequent XRD analysis
(not shown) of this sample yielded a mixture of MoOPO4 and MoO2
phases on carbon support. The results inferred that the MoP could
not be obtained from carbon-supported molybdenum phosphate
precursor via carbothermal reduction method at a temperature of
750 ◦C. In other words, the molybdenum and phosphorous species
were indeed reduced by CH4 from the HMT ligand occurred, to give
MoP phosphide with release of CO and CO2.
The ensemble of MS, IRAS and XRD analysis data allowed us
to conclude that thermal decomposition of the mixed-salt precur-
sor as an internal oxidation–reduction process in which methane
reduced molybdenum and phosphorous species and was elimi-

L12
Z.W. Yao et al. / Journal of Alloys and Compounds 473 (2009) L10–L12
simpler and more straightforward. Furthermore, the BET surface
area of 10 m2 g−1 for the material prepared by this new method was
larger than the value of 2 m2 g−1 for that prepared by TPR method
[20]. It was reported that a large amount of water produced in TPR
process, which can lead to hydrothermal sintering and low surface
area of the resulting catalysts [14]. On the contrary, in the present
synthesis route, it was apparent that CO and CO2 were the major
products, and no quantification of H2O was made overall reduc-
tion process. This was a point worthy of which suggested the use
of CH4 as reductant in phosphide synthesis procedure can adapt
to minimize hydrothermal sintering. Additionally, it was reason-
able to believe that a degree of carbon lay-down arised from HMT
route to MoP, and the presence of carbon in the metal phosphides
was reported to be advantageous in preventing high temperature
sintering of phosphide particles [13,14]. Therefore, the content of
HMT in the precursor had an influence on the products and should
be carefully optimized. The internal oxidation–reduction synthe-
sis route is expected to be a general method for the generation of
relatively high surface area phosphides.
4. Conclusion
A new one-step method was found for the synthesis of MoP.
Decomposition of the mixed-salt precursor provided a synthesis
route free of diffusion limitations. Such a simple and straightfor-
ward synthesis route in this paper, suggests that it may be possible
to further optimize the synthesis described for the production of
high surface area phosphides.
Acknowledgements
The work was supported by the National Natural Science Foun-
dation of China (No. 20573014) and the Natural Science Foundation
of Liao-Ning Province (20041072).
References
Fig. 3. Thermal decomposition of mixed-salt precursor at 10 ◦C/min followed by MS
and IRAS.
[1] L.E. DeLong, G.P. Meisner, Solid State Commun. 53 (1985) 119–123.
[2] O. Tegus, E. Bru¨ ck, K.H.J. Buschow, F.R. de Boer, Nature 415 (2002) 150–152.
[3] S.T. Oyama, J. Catal. 216 (2003) 343–352.
[4] J. Jiang, S.M. Kauzlarich, Chem. Mater. 18 (2006) 435–441.
[5] D.C.S. Souza, V. Pralong, A.J. Jacobson, L.F. Nazar, Science 296 (2002) 2012–
2015.
[6] S. Yang, R. Prins, Chem. Commun. (2005) 4178–4180.
[7] C. Stinner, R. Prins, Th. Weber, J. Catal. 191 (2000) 438–444.
[8] K.A. Gregg, S.C. Perera, G. Lawes, S. Shinozaki, S.L. Brock, Chem. Mater. 18 (2006)
879–886.
[9] K.L. Stamm, S.L. Brock, J. Alloys Compd. 453 (2008) 476–481.
[10] Y. Xie, H.L. Su, X.F. Qian, X.M. Liu, Y.T. Qian, J. Solid State Chem. 149 (2000)
88–91.
[11] C.M. Lukehart, S.B. Milne, S.R. Stock, Chem. Mater. 18 (1998) 903–908.
[12] K.L. Stamm, J.C. Garno, G. Liu, S.L. Brock, J. Am. Chem. Soc. 125 (2003)
4038–4039.
[13] S. Yang, C. Liang, R. Prins, J. Catal. 241 (2006) 465–469.
[14] S. Burns, J.S.J. Hargreaves, S.M. Hunter, Catal. Commun. 8 (2007) 931–935.
[15] P. Afanasiev, Inorg. Chem. 41 (2002) 5317–5319.
[16] H. Wang, W. Li, M. Zhang, Chem. Mater. 17 (2005) 3262–3267.
[17] S. Chouzier, P. Afanasiev, M. Vrinat, T. Cseri, M. Roy-Auberger, J. Solid State Chem.
179 (2006) 3314–3323.
[18] N. Fan, X. Ma, Z. Ju, J. Li, Mater. Res. Bull. 43 (2008) 1549–1554.
[19] X. Wang, P. Clark, S.T. Oyama, J. Catal. 208 (2002) 321–331.
[20] C. Stinner, R. Prins, Th. Weber, J. Catal. 202 (2001) 187–194.
nated as CO and CO2. It was observed earlier that the intramolecular
oxidation–reduction during the thermal decomposition of transi-
tion metals hexamethylenetetramine complexes can lead to the
formation of metal carbides and nitrides [15–17].
The possible formation mechanism of molybdenum phosphide
during the thermal decomposition process was proposed as fol-
lows: (1) HMT decomposed to CH4, N2 and C; (2) molybdenum and
phosphorous species reacted with CH4 to product the MoP particles
and the gases of CO, CO2 and H2.
Traditional metal phosphide catalysts were prepared using the
typical TPR method. The details were described elsewhere [20]. In
the TPR method, the precursor was first calcined at 500 ◦C for 5 h
and then reduced at a slow rate (1 ◦C/min) to 650 ◦C in a stream
of hydrogen (300 mL/min), whereas in this new method, the pre-
cursor was directly heated at a rate of 10 ◦C/min to 750 ◦C under
argon with a flow rate of 50 mL/min. Comparison of the two syn-
theses clearly indicated that the new method described here was