Tin phosphide coatings from the atmospheric pressure chemical vapour deposition of SnX4 (X = Cl or Br) and PRxH3-x (R = Cychex or phenyl)

Article abstract of DOI:10.1016/S0277-5387(02)01081-1

Atmospheric pressure chemical vapour deposition (APCVD) of tin phosphide thin films was achieved on glass substrates from the reaction of SnCl₄ or SnBr₄ with R_xPH_3-x (R = Cyc^hex or phenyl) at 500-600

Full text of DOI:10.1016/S0277-5387(02)01081-1

Polyhedron 21 (2002) 1943Á1947
/
www.elsevier.com/locate/poly
Tin phosphide coatings from the atmospheric pressure
chemical vapour deposition of SnX4 (XꢀCl or Br)
and PRxH3ꢁx (Rꢀ
Cychex or phenyl)
/
/
Russell Binions a, Christopher S. Blackman a, Clare J. Carmalt a, Shane A. O’Neill a,
Ivan P. Parkin a,*, Kieran Molloy b, Leonardo Apostilco b
a Department of Chemistry, Christopher Ingold Laboratories, University College London, 20 Gordon Street, London WC1H 0AJ, UK
b Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, UK
Received 11 March 2002; accepted 23 May 2002
Abstract
Atmospheric pressure chemical vapour deposition (APCVD) of tin phosphide thin films was achieved on glass substrates from the
reaction of SnCl4 or SnBr4 with Rx PH3ꢁx (Rꢀ 600 8C. These coatings show good uniformity and
Cychex or phenyl) at 500Á
/
/
surface coverage. They are reasonably adherent, passing the Scotch tape test. The films were largely opaque in appearance with
regions of birefringence. The films were X-ray amorphous. Scanning electron microscopy (SEM) showed surface morphologies
consistent with an island growth mechanism. X-ray photoelectron spectroscopy (XPS) binding energy shifts for SnP1.00 were 487.2
eV for Sn 3d5/2 and 133.6 eV for P 2p3/2. Energy dispersive X-ray analysis (EDXA) and electron probe studies gave elemental ratios
that were in agreement indicating slightly tin rich and stoichiometric films, with negligible chlorine or bromine incorporation
(SnP1.33
the films showed very poor reflectance (approximately 2%) and 5Á
measurements were in the range 70Á808, and showed little change after 60-min irradiation at 254 nm. # 2002 Elsevier Science Ltd.
All rights reserved.
Á
/
SnP0.40). No Raman scattering was observed. Sheet resistance measurements indicated the films were insulating. Optically
/20% total transmissions from 300 to 800 nm. Contact angle
/
Keywords: Tin phosphide; Thin film; APCVD
1. Introduction
The preparation of metal phosphides is a problem in
science and engineering at present [6]. Bulk metal
phosphides have been prepared from the reaction of a
metal with Ca3P2 [7], directly from the elements [8], from
the reaction of metal halides, by heating the metal in a
stream of phosphine [9], the reduction of the metal
oxides in the presence of phosphorus and iodine [10] and
by solid state metathesis reactions [11].
Formation of metal phosphide films has received little
attention, especially by the use of chemical vapour
deposition (CVD). To our knowledge no work has
been published on the formation of thin films of tin
phosphides. However, these phosphides as bulk materi-
als have been known for some time [12]. An investiga-
tion into the thermodynamic properties of the tin
phosphide system was published in 1974 [13]. Crystal
structures of bulk materials were published between
1957 and 1972 [14] and later refined [15]. Recently, tin
Transition metal phosphides have a variety of proper-
ties that could potentially find commercial applications.
They are metallic conductors, hard, refractory and in
some case resistant to oxidation [1]. Despite this they
have only found a limited use as catalysts [2] and as
diffusion barrier layers in semiconductors [3]. Whereas
the corresponding metal nitrides have many technolo-
gical applications [4] only the main group phosphides
InP and GaP have been extensively studied. A variety of
synthetic techniques have been developed concerning
the formation of thin films of these Group 13 phos-
phides [5].
* Corresponding author. Tel.: ꢂ
679-7463
E-mail address: i.p.parkin@ucl.ac.uk (I.P. Parkin).
/
44-207-679-4669; fax: ꢂ44-207-
/
0277-5387/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 2 7 7 - 5 3 8 7 ( 0 2 ) 0 1 0 8 1 - 1

1944
R. Binions et al. / Polyhedron 21 (2002) 1943Á1947
/
phosphides have received attention as potentially inter-
esting nanomaterials [16]. It is anticipated that such
nanomaterials may have interesting mechanical, optical
and electrical properties [17]. They are also known to be
catalysts of alkylation reactions [18].
Here we report the first instance of a tin phosphide
film prepared by dual source atmospheric pressure
chemical vapour deposition (APCVD).
diverting the nitrogen line through the bubbler and
hence to the reactor. Deposition experiments were timed
by use of a stopwatch and were conducted typically for 1
min. At the end of the deposition only nitrogen was
allowed to flow over the glass substrate until the
substrate was sufficiently cool to handle (approximately
60 8C). Cool, coated samples were handled and stored
in air. The large coated glass sample was cut into smaller
samples (1ꢃ
1 cm2), for analysis by X-ray photoelectron
/
spectroscopy (XPS), energy dispersive X-ray analysis
(EDXA), scanning electron microscopy (SEM), electron
probe, transmission/reflectance and UV absorption
2. Experimental
Nitrogen (99.99%) was obtained by BOC and used as
supplied. Coatings were obtained on SiO2 coated float
studies. Larger pieces of glass (4ꢃ
4 cm2) were used
for sheet resistance, X-ray powder diffraction, contact
/
glass. APCVD experiments were conducted on 150ꢃ
45ꢃ
3 mm3 pieces of glass using a flat bed cold walled
APCVD reactor. The glass was cleaned prior to use by
washing with petroleum ether (60Á80 8C) and isopro-
/
angle and Scotch tape tests.
X-ray powder diffraction patterns were measured on
/
a Siemens D5000 diffractometer using filtered (Cu Ka1,
˚
1.5406 A) radiation in the reflection mode using
/
lꢀ
/
panol and then dried in air. A graphite block containing
a Whatman cartridge heater was used to heat the glass
substrate. The temperature of the substrate was mon-
glancing angle incidence (1.58). SEM was obtained on a
Hitachi S570 instrument. EDXA was obtained using a
JEOL 35-CF instrument. Electron microprobe analysis
was obtained on a JEOL EMA and referenced against
Sn and P standards. X-ray photoelectron spectra were
recorded with a VG ESCALAB 220I XL instrument
using a focused (300 mm spot) monochromatic Al Ka
radiation at a pass energy of 20 eV. Scans were acquired
with steps of 50 meV. A flood gun was used to control
charging and the binding energies were referenced to an
adventitious C 1s peak at 284.6 eV. Depth profiling
measurements were obtained by using Ar beam sputter-
ing.
itored by a PtÁRh thermocouple. Independent thermo-
/
couple measurements indicated that temperature
gradients of up to 50 8C were observable at 600 8C
across the surface of the glass. The rig was designed so
that four independent gas lines could be used. All gas
handling lines, regulators and flow valves were made of
1
4
stainless steel and were ƒ internal diameter except for
the inlet to the mixing chamber and the exhaust line
1
2
from the apparatus that were ƒ in diameter. In these
experiments three gas lines were used. Gases came
directly from a cylinder and were preheated by passing
along 2-m lengths of stainless steel tubing, which were
curled and inserted inside a tube furnace. The tempera-
UVÁVis spectra were obtained using a Helios double
/
beam instrument. Reflectance and transmission spectra
were recorded between 300 and 1150 nm by a Zeiss
miniature spectrometer. Measurements were standar-
dised relative to an Rh mirror (reflectance) and air
(transmission). Raman spectra were acquired on a
tures of all the gas inlet lines were monitored by PtÁ
/
Rh
thermocouples and Eurotherm heat controllers.
Tin(IV) chloride (99.99%) was obtained from the
Aldrich Chemical Company and used without further
purification and was placed into a stainless steel
bubbler. The bubbler was heated to 35 8C by a heating
jacket and SnCl4 introduced into the gas streams by
passing hot nitrogen gas through the liquid. PCychexH2,
Renishaw Raman system 1000 using HeÁNe laser of
/
wavelength 632.8 nm. The Raman system was calibrated
against the emission lines of Ne. Contact angle experi-
ments were determined by measuring the spread of a 1 ml
droplet of water and applying a simple trigonometric
program. Electrical properties were determined by using
a four-probe device.
PCychex2H, PCychex and PhPH2 were purchased from
3
Strem Chemicals and used without further treatment.
The two components of the system were mixed by the
use of two concentric pipes of 14ƒ and 21ƒ diameter,
respectively. The inner pipe being 3 cm shorter than
the outer pipe. The concentric pipes were attached
directly to the mixing chamber of the coater. Gas flows
were adjusted using suitable regulators and flow con-
trollers. The exhaust from the reactor was vented
directly into the extraction system of a fume cupboard.
All of the apparatus was baked out with nitrogen at
150 8C for 30 min before use. Deposition experiments
were conducted by heating the horizontal bed reactor
and the bubblers to the desired temperatures before
3. Results
APCVD of tin phosphide films were achieved on glass
substrates from the dual source CVD reaction of SnCl4
or SnBr4 and a substituted phosphine (PCychexH2,
PCychex2H, PCychex or PhPH2). The process was
3
studied at different substrate temperatures and flow
rates of precursor (Table 1).
No film could be grown at substrate temperatures
below 400 8C, uniform films were obtained at 500 8C

R. Binions et al. / Polyhedron 21 (2002) 1943Á
/
1947
1945
Table 1
Experimental conditions and microanalytical results
Substrate
temperature
(8C)
Nitrogen flow through SnX4
bubbler (l minꢁ1) (SnX4)
[bubbler temperature (8C)]
Nitrogen flow through Rx PH3ꢁx
bubbler (l minꢁ1) (Rx PH3ꢁx
Nitrogen flow of
make up gas
(l minꢁ1
)
EDAX and
electron probe
analysis
)
[bubbler temperature (8C)]
500
500
500
500
500
550
600
500
500
600
600
600
600
600
0.5 (SnCl4) [35]
0.5 (SnCl4) [35]
0.2 (SnCl4) [35]
0.3 (SnCl4) [35]
0.8 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.3 (SnCl4) [35]
0.5 (SnBr4) [200]
0.4 (SnBr4) [200]
0.5 (PCychexH2) [120]
0.5 (PCychexH2) [120]
0.8 (PCychexH2) [120]
0.3 (PCychexH2) [120]
0.2 (PCychexH2) [120]
0.3 (PCychexH2) [120]
0.3 (PCychexH2) [120]
0.3 (PhPH2) [145]
1.4
1.8
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.7
0.2
0.8
0.8
SnP0.40
SnP0.66
SnP0.40
SnP1.00
Sn
SnP1.00
SnP1.00
SnP0.66
SnP1.33
SnP1.00
SnP0.66
SnP1.00
SnP1.00
SnP1.33
0.5 (PhPH2) [145]
0.5 (PhPH2) [145]
0.7 (PCych2ex H) [250]
1.8 (PCych2ex H) [250]
0.3 (PCych3ex ) [120]
0.4 (PCych3ex ) [120]
and above. The films grown from the dual source CVD
reaction of SnCl4 or SnBr4 and a substituted phosphine
(PCychexH2,Cychex2H, PCychex3 or PhPH2) were opaque
in colour, although those grown at higher temperatures
had a blue tint. The films appeared unreflective. The
films prepared at higher temperatures exhibited some
birefringence due to thickness effects. All of the films
passed the Scotch tape test. They could not be marked
with a wet towel but could be marked with a brass stylus
or steel scalpel. The films prepared at higher precursor
flow rates showed some ‘pin holing’ due to gas phase
nucleation. The films showed no optical change after
storage in air for 3 months.
Compositional analysis of the films was determined
by EDXA and electron probe measurements (sum-
marised in Table 1). The results show that the films
are free of chlorine (or bromine if SnBr4 was used) and
contain only tin and phosphorus. In all cases some
breakthrough to the underlying glass was observed. This
was corrected for based on a sample of plain glass.
These results show that it is possible to create films with
a selection of stoichiometries, ranging from SnP1.33 to
SnP0.40. The analysis showed that there was a slight
variation in composition across the surface of the glass
substrate, for example films with composition SnP1.33
had a composition of SnP1.00 at the extreme edges, this
correlates with the measured temperature gradient
across the glass substrate. Area maps indicate that
particles containing tin and phosphorus are made (Fig.
1), and that no oxygen is present in these particles.
SEM analysis of the films showed that the deposition
reaction proceeded via an island growth mechanism.
The majority of samples consisted of microspheres of tin
phosphide covering the majority of the surface. Two
representative samples are shown in Fig. 2.
Raman microscopy was used to investigate all of the
films. In all cases no Raman scattering due to the films
was observed. It is thought that tin phosphides are poor
Raman scatters, indeed a sample of Sn4P3 purchased
form Aldrich gave no Raman spectra. In the samples
prepared using PCychexH2 and PhPH2 no free carbon
was detected by Raman. This was not the case with films
prepared using PCychex2H and PCychex where free
3
carbon was present in the films. No Raman scattering
was observed from the film due to the presence of tin
oxides. In previous work we have shown that even trace
amounts of tin oxides (SnO2 or Sn3O4) give strong
Raman patterns [19].
The optical properties of the films were studied by
reflectance/transmission measurements and UV absorp-
tion between 400 and 1100 nm. All samples showed a
shift in the absorption edge relative to the plain glass
Fig. 1. EDXA area maps of tin phosphide particles. The lighter colours indicate a higher concentration of the specific element. The key on each
caption SE, secondary electron image; SnL, tin L X-ray line; PK, phosphorus K X-ray line; OK, oxygen K X-ray line. The numbers 255, 66, 19 and
32 indicate the uncorrected counts per second for each X-ray line.

1946
R. Binions et al. / Polyhedron 21 (2002) 1943Á1947
/
Fig. 2. Scanning electron microscope pictures of tin phosphide films formed from the APCVD reaction of tin tetrachloride and monocyclohexyl-
phosphine at 600 8C.
substrate. The films showed minimal reflectivity (2%),
less than the substrate (10%). The transmission spectra
were indicative of hazing; this is due to scattering by
particles in the film of a similar size to the wavelength of
the light.
cantly after irradiation with 254 nm UV light for 1 h.
This contact angle is similar to that of plain glass 508Á
758.
/
XPS measurements on a sample of SnP1.00 composi-
tion, revealed binding energy shifts of 487.2 and 133.6
eV for Sn 3d5/2 and P 2p3/2, respectively. These are in
reasonable agreement with previous values published in
4. Discussion
Reaction of SnCl4 (or SnBr4) and a substituted
phosphine (PCychexH2, PCychex2H, PCychex or PhPH2)
3
6
the literature concerning nanoscale Sn4P3 of 487.6 eV
under APCVD affords opaque films of tin phosphide on
glass substrates. The tin phosphide films were shown in
most cases to be tin rich, although it was possible to
make the stoichiometric compound tin monophosphide
when equimolar quantities of precursor were introduced
into the gas phase.
The films were found to be insulating. This is because
of grain boundary effects due to the island growth
mechanism.
for Sn 3d5/2 and 134.1 eV for P 2p3/2. A sample of
composition SnP0.40 revealed similar binding shifts of
487.2 and 133.8 eV for Sn 3d5/2 and P 2p3/2, respectively.
The XPS peak profiles for both samples (Fig. 3) are
consistent with one environment for both tin and
phosphorus. There is no evidence for the presence of
tin oxides from the XPS profiles.
The tin phosphide films were X-ray amorphous. Sheet
resistance measurements indicated that the films were
insulating. The films were shown to have contact angles
The mechanism of this reaction was not determined in
this study. It is likely that HCl is produced in the
reaction as well as CychexCl or PhCl (depending on the
of 708Á808. This contact angle did not change signifi-
/
Fig. 3. The XPS peak profiles for a SnP1.00 sample showing on the left, the phosphorus 2p3 region and on the right, the tin 3d5/2 and 3d3/2 region.

R. Binions et al. / Polyhedron 21 (2002) 1943Á
/
1947
1947
[2] K. Komaki, Japan Koka Tokkyo Kokon JP02150067.
[3] H. Su, Y. Xie, B. Li, X. Liu, Y. Qian, J. Solid State Chem. 146
(1999) 110.
phosphine precursor used) [20]. It is plausible that the
reaction occurs via the formation of a adduct as some
material was recovered from the exhaust which was not
starting material. Previous work also suggests that
similar deposition process occur via adducts [21].
These films show no Raman spectra. They shift the
absorption edge of, and undergo less transmission and
reflection than SiO2 coated glass due to scattering haze.
EDAX and electron probe results suggest that the
composition over the surface of the glass is similar.
SEM shows that whilst various morphologies are
possible, one predominates in samples of tin phosphide
films.
There is no variation of film stoichiometry with
different precursors. It was noted that distinct phases
(SnP1.33, SnP1.00, SnP0.66 and SnP0.40) were seen and no
intermediate phases were observed. Higher nitrogen
flow rates had the effect of lowering the phosphorus
content of the resultant films. Using di- and tri-
substituted phosphines required higher substrate tem-
peratures (600 8C) for films to form. Where films did
form at lower substrate temperatures (500 8C) with the
di- and tri-substituted phosphines they were lower in
phosphorus content than those formed at higher sub-
strate temperatures.
[4] (a) J.C. Bailer, H.J. Emeleus, R. Nyholm, A.F. Trotman-
Dickenson, Comprehensive Inorganic Chemistry, vol. 2, Perga-
mon, Oxford, 1973, p. 233;
(b) G.R. Gordon, S.R. Kurtz, Thin Solid Films 140 (1986) 277.
[5] A.N. Gleizes, Chem. Vapour Deposition 6 (2000) 155.
[6] V.B. Chernogorenko, S.V. Muchnik, K.A. Lynchak, Z.A. Kli-
mak, Mater. Res. Bull. 16 (1981) 1.
[7] (a) N. Schonberg, Acta Chem. Scand. 8 (1954) 226;
(b) T. Ludstrom, P.O. Snell, Acta Chem. Scad. 21 (1967) 1343.
[8] J.R. Van Wazer, Phosphorous and its Compounds, vol. 1,
Interscience, New York, 1958, p. 123.
[9] (a) R.L. Ripley, Less Common Metals 4 (1962) 496;
(b) W. Blitz, A. Rink, F. Wiechman, Z. Anorg. Chem. 238 (1938)
395.
[10] L.L. Vereikina, G. Samsanov, V. Zh. Neorgkhim. 5 (1960) 1888.
[11] (a) R.B. Kaner, R.E. Treece, J.A. Conklin, Inorg. Chem. 33
(1994) 5701;
(b) R.B. Kaner, R.M. Jacubinas, R.J. Jarvis, Inorg. Chem. 39
(2000) 3242;
(c) I.P. Parkin, A.L. Hector, J. Mater. Chem. 4 (1994) 229;
(d) I.P. Parkin, A.L. Hector, Z. Naturforsch., Teil B 49 (1994)
477;
(e) I.P. Parkin, J.C. Fitzmaurice, A.T. Rowley, J. Mater. Chem. 4
(1994) 285;
(f) I.P. Parkin, A.T. Rowley, J. Mater. Chem. 3 (1993) 689.
[12] A.C. Vivian, J. Inst. Met. 23 (1920) 325.
[13] (a) Vses. Konf. Kntalbkhim Intermet Soedin 2nd, 1974, p. 168;
(b) IZV Akad. Navk. SSSR Serziz 38 (1974) 562.
[14] (a) G. Katz, J.A. Kohn, J.D. Broder, Acta Crystallogr. 10 (1957)
607;
5. Conclusion
(b) O. Olofsson, Acta Chem. Scand. 21 (1967) 1659;
(c) P.C. Donohue, Inorg. Chem. 9 (1970) 335;
(d) O. Olofsson, Acta Chem. Scand. 24 (1970) 1153;
(e) J. Gullman, O. Olofsson, J. Solid State Chem. 5 (1972) 441.
[15] (a) Yu.B. Kuz’ma, S.I. Chikhrii, V.N. Davydov, Inorg. Mater. 35
(1999) 10;
Reaction of SnCl4 (or SnBr4) and a substituted
phosphine (PCychexH2, PCychex2H, PCychex
PhPH2) under APCVD affords opaque films of tin
phosphide on glass substrates. The films show good
surface coverage, adhesion and reasonable uniformity.
Tin phosphide films with a range of stoichiometries
or
3
(b) J. Gullman, J. Solid State Chem. 87 (1990) 202.
[16] Y. Xie, H. Su, B. Li, Y. Qian, Mater. Res. Bull. 35 (2000) 675.
[17] A.M. Rao, E. Richter, S. Bandow, B. Chase, P.C. Eklund, K.A.
Williams, S. Fang, K.R. Subbaswamy, M. Menon, A. Thess, R.E.
Smalley, G. Dresselhaus, M.S. Dresselhaus, Science 275 (1997)
187.
(SnP0.40ÁSnP1.33) could be obtained; these were all
/
opaque in appearance and electrical insulators.
[18] (a) R.L. Halm, K.M. Chadwick, B.R. Keyes, US Patent
4,946,980;
Acknowledgements
(b) R.L. Halm, K.M. Chadwick, B.R. Keyes, US Appl. 258,950;
(c) R.L. Halm, K.M. Chadwick, US Patent 4,985580.
[19] (a) L.S. Price, Ph.D. Thesis, University of London, 2001;
(b) T.G. Hibbert, M.F. Mahon, K.C. Molloy, I.P. Parkin, L.S.
Price, J. Mater. Chem. 11 (2001) 464.
The authors wish to thank the EPSRC for funding
and Pilkington for a CASE award to RB and for glass.
The EPSRC are also thanked for providing financial
assistance in the purchasing of the Raman spectrometer.
[20] (a) J.T. Scheper, K.C. Jayaratne, L.M. Liable-Sands, G.P.A. Yap,
A.L. Rheingold, C.H. Winter, Inorg. Chem. 38 (1999) 4354;
(b) T.S. Lewkebandara, J.W. Prosica, C.H. Winter, Chem. Mater.
7 (1995) 1053.
References
[21] C.S. Blackman, C.J. Carmalt, S.A. O’Neill, I.P. Parkin, L.
Apostilco, K.C. Molloy, J. Mater. Chem. 11 (2000) 2408.
[1] (a) K.A. Gingerich, Nature 200 (1963) 877;
(b) S. Motojima, T. Wakamatsu, K. Sugiyama, Less Common
Metals 82 (1981) 379.