In situ investigation of Pt100—xAux and Pt100—ySny nanoalloys
Ken A. Grant, Kelei M. Keryou and Paul A. Sermon*

Received 11th June 2007, Accepted 28th June 2007
First published as an Advance Article on the web 15th October 2007
DOI: 10.1039/b708810h

Dispersed sols of 1–10 nm sized Pt100—xAux and Pt100—ySny nanoalloys have been prepared separately at various x and y above and below the miscibility limit in the bulk metals. Pt100—xAux was derived from trisodium citrate reduction of aqueous solutions of H2PtCl6 and HAuCl4. Pt100—y–Sny was produced by (i) complexing Sn2+ with glucose at 323 K at pH 4 7, (ii)
neutralising this with H2PtCl6 addition and (iii) reducing the bimetallic precursor with glucose on raising the temperature to 373 K. For Pt100—xAux (where both metals were zero-valent) as x increased the average size of
nanoalloy particles increased. These particles adsorbed onto graphite, where the extent of hydrogen chemisorption at 298 K decreased by 67% at 9 at% Au. Pt/SnO2 nanoparticles (o3 nm in size) were adsorbed onto alumina.
The Pt interacted with and catalysed the reduction of SnO2, with some Pt100—ySny nanoalloy formation at about 673 K which even in the bulk occurs over a wider range of compositions than Pt–Au) and enhanced H2
chemisorption at 17–33 at% Sn. Nevertheless some Sn must remain in a positive oxidation state on the alumina surface. The ratio of rates of 2MP/ 3MP formation from MCP and n-hexane may be informative in chemically fingerprinting (and revealing fundamental differences in) these nanoalloy surfaces. The reasons for this are seen in terms of the surface structures on these two types of nanoalloy particles (i.e. the availability of contiguous asymmetric pairs of active surface atoms *, which, as expected, is found to pass through a maximum or decrease beyond specific values of x and y).

1. Introduction
Colloidal solutions of catalytic metals are readily prepared by reduction in solution of appropriate salts and as such are part of bottom-up nanotechnology. Some have seen similarities between their high catalytic activity and that of enzymes. Despite their instability,1 metal nanoparticles have many biological, chemical and physical applications.2 Therefore attention has now turned to the optimisation of the surface composition and properties of binary metallic nanoparticles; some can be modelled3– 9 (e.g. segregation,10–14 melting point,15,16 ordering17 and stability18), but others are best followed experimentally via high-resolution transmission electron microscopy,19 X-ray diffraction20 or catalytic/adsorptive probes.21–23 This is the subject of this Discussion.
Au catalyses CO oxidation at low temperature24 and, intriguingly, Pd–Au nanoparticles are 70 times more active23 than Pd alone and may be relevant to bimetallic nanoparticle catalysts for aqueous-phase trichloroethene hydro-

Chemistry, FHMS, University of Surrey, Guildford, Surrey, UK GU2 7XH

dechlorination (i.e. groundwater remediation). Here we have chosen Pt–Au21 to test because of their potential immiscibility and previous work.25
Monodisperse PtRu nanoalloy can be prepared from acac precursors using 1,2- hexadecanediol reducing agent deposited on carbon.26 Here PtSn nanoalloys have also been chosen because they have greater electro-oxidation activity towards ethanol than Pt–Ru/C27 which have a resistance to CO poisoning that is important in a fuel cell context and is not possessed by Pt alone. A Pt–Sn alloy may be formed on alumina above 0.6 wt%, but Pt and Sn28 species will need to be related to the surface chemistry of Pt and Sn in reforming catalysts.29 Using ethylene glycol reductant Pt50Sn50 (d = 2.7 nm) has been produced and supported on carbon;30 this nanoalloy appears to have a bright future in that it is a better electrocatalyst than PtRu/C or PtSnRu/C in CH3OH oxidation. Specifically, addition of Sn allows oxidation of ethanol at lower potentials than Pt alone.31 For 0.4–0.9 mg PtSn nanoparticles per cm2 of graphite, anodic stripping was used to determine the surface Pt : Sn ratio (3.6); this was higher than the interior ratio (2.5)32 possibly as a result of a surface stabilisation by tetraoctylammonium triethylhydroborate.
Here then we have selected Pt100—xAux and Pt100—ySny (the former less mutually miscible than the other, but the latter with greater ease of oxidation of one
component) as examples of nanoalloys to explore.
It seemed appropriate to try to ascertain the mode of sol-formation.33 In addition it seemed that in situ34 characterisation methods would also be needed under conditions of use. Active sites and surface properties of bimetallic solids have often been investigated using catalysis of carbon–hydrogen bonds35 (e.g. in n-hexane (nHex) and 2-methylpentane (2MP)) as a function of temperature at low (o0.133 mPa) and high ( Z 133 Pa) pressures in the absence and presence of hydrogen. Gaseous n-hexane can be in equilibrium with surface-held hexylidene (RC–CH2– CH2–CH2–CH2–CH3) and p-bound methylcyclopentane (MCP) over Pt–Pd. Turn- over frequencies (TOFs) and product selectivities for n-hexane (nHex)36 and methylcyclopentane (MCP) conversion at 623 K–773 K at a H2/MCP = 20 after pre-reduction in H2 to 773 K have been reported,37 although it was recognised that carbon accumulation and deactivation/reactivation occur.38 Ring opening of methylcyclopropane (an intermediate in n-hexane conversion) appeared to be highly sensitive to changes in surface structure of catalysts (as is hydrogenolysis39). Non- selective MCP ring opening will give 2MP/3MP = 2. n-Hexane (one of the simplest alkanes to be able to undergo skeletal rearrangement (hydrogenolysis, isomerisation, dehydrocyclisation and aromatisation) over Pt catalysts40) and methylcyclopentane (MCP, that converts to n-hexane, 3-methylpentane (3MP) or 2-methylpentane (2MP)41–43) were selected as the probe reactants here. In the past44 ethane hydro- genolysis was studied and found to decrease in rate on alloying of a low activity metal with a high activity one,45 while facile dehydrogenation46 changed less dramatically.

2. Experimental

One can extrapolate directly from the work on AuCl4— reduction to give Au nanoparticles undertaken by Faraday47 to Frens.48 Tri-sodium citrate is effective
at the solution boiling point developing a blue and then a red colour.49 Now more intricate core–shell Au–Cu nanocomposites50,51 and Au–Cu nanoparticles em-
bedded in SiO 52 have been prepared. Here Pt Au (0 o x o 100, where x
2 100—x x
was 0 and 9 and also 25, 50 and 100 at% Au) was prepared in pre-cleaned glassware by reducing a mixed aqueous (HPLC grade water) solutions (30 mg dm—3) of H2PtCl6 (Johnson Matthey) and HAuCl4 (Johnson Matthey) with a 1% trisodium
citrate solution and then stirring for 2–4 h at 373 K.53 The product sol was purified by addition of Amberlite MBI ion-exchange resin with stirring until constant

conductivity and pH were obtained. Such sols were adsorbed on graphite (Fluka; 99.9% purity; 11 m2 g—1), alumina (Degussa Al2O3–C; 100 m2 g—1), silica (Degussa Aerosil 200; 200 m2 g—1), titania (Degussa P25, 50 m2 g—1) and MgO (BDH AnalaR, 16 m2 g—1). Concentrations were deduced by atomic absorption after digestion.
Pt100—ySny (0 o y o 100, where y was 0, 17, 33, 43, 52 and 55 at%) was prepared
by dissolving SnCl2 (Aldrich; 99%) and NaOH in HPLC-grade water and then
complexing this with glucose (BDH, AnalaR). H2PtCl6 was then added and reduced while refluxing at 373 K. The product darkened from amber as the glucose polymerised, the SnCl2 hydrolysed and H2PtCl6 reduced. After cooling the sols were purified by addition of Amberlite MB1 resin, that was then removed by
filtering. These sols were also adsorbed to a level of 1 wt% Pt on alumina (Degussa Al2O3–C; 100 m2 g—1). The Sn2+–glucose interaction in solution (that occurred during sol synthesis) was followed at 373 K by polarographic (Metrohm Ploracord
626 with dropping Hg and Ag/AgCl electrodes) measurements of limiting currents in solutions under N2 (id in nA) and half-wave potentials (E1/2 in V that were 0.63 to
—0.65 V (anodic wave) and —1.09 to —1.15 V (cathodic wave) for Sn ). Clearly
Fig. 1 and Table 1 show that Sn2+ is consumed within 5 min at 373 K (and longer at lower temperatures) and a new electroactive species, presumably produced from the Sn2+–glucose complex, is seen (at —E1/2 = 1.64–1.68 V) to form to an increasing

Fig. 1 (a) Polarographs of Sn2+–glucose reaction at 348 K after 0, 1, 3, 5, 10 and 20 min (and in the absence of glucose (standard: std)). (b) % decrease in id peak at E1/2 = 0.63–0.67 V (relative to the standard value in the absence of glucose) at 323 K, 348 K and 373 K measured as a function of reaction time.

Table 1 Polarographic analysis of the Sn2+–glucose reaction at 373 K

t/min Sn2+ anodic wavea Sn2+ cathodic waveb Unknownc
0 3040 2500 0
0.2 1440 1200 660
1.0 500 280 1140
3.0 420 160 1840
5.0 0 0 2010
20.0 0 0 2540
a For Sn2+ E1/2 = —0.63 top —0.65 V. b For Sn2+ E1/2 = —1.09 to
—1.15 V. E1/2 = —1.64 to —1.68 V.

extent as the Sn2+ is lost. Hence, sol formation at 373 K was selected, rather than a slower process at a lower temperature (i.e. at 348 K a reaction time increased by a factor of 2 would have been required (see Fig. 1)).

Characterisation methods
Extents of hydrogen chemisorption (n) at 298 K were measured volumetrically for Pt100—xAux/C and Pt100—ySny/Al2O3 (pre-calcined in air at 673 K for 2 h, reduced in situ at 673 K in H2 for 2 h and finally outgassed to 0.1 mPa for 1 h) by extrapolation
of n–p isothermal data to p = 0 Pa. High-resolution transmission electron micro- scopy (TEM)54–56 and electron diffraction57 have characterised the structure, size and morphology of unsupported nanoparticles, including those of Pt–Sn;32 here a
Jeol 200CX25 was used.32 X-Ray photoelectron spectroscopy (XPS; Kratos E3500 using AlKa (1486.6 eV) normalised to Al2p at 74.7 eV) was applied to Pt100—ySny/ Al2O3 samples (pre-calcined in air at 673 K for 2 h and reduced in situ at 673 K in H2
for 2 h). Temperature-programmed reduction (TPR) using 5% H2/Ar flowing at 40 cm3 min—1 during programming from 103–900 K at 5 K min—1 was applied to Pt100—ySny/Al2O3 (also pre-calcined in air at 673 K for 2 h but not pre-reduced). AAS data for digested samples were used to calculate x and y on an atomic basis.

Catalytic methods
Surface atomic arrangements and ratios are always critical to nanoalloys; these have been assessed here using a catalytic probe.58 Catalytic activities and selectivities of Pt100—xAux/C were deduced using H2 : MCP = 19 : 1 or H2 : hexane = 18 : 1 with
H2 flow rates of 27 cm3 min—1 with catalysts that had been pre-reduced at 573 K for
1 h. Catalytic activities and selectivities were measured for Pt100—xSnx/Al2O3 using MCP (Fluka AG Chemicals) and n-hexane (Alfa Products) with catalyst samples
(0.05 g for MCP conversion and 0.25 g for n-hexane) precalcined (673 K; 2 h) and then pre-reduced (673 K; 2 h). H2 (30 cm3 min—1 for MCP and 90 cm3 min—1 for n- hexane) flowed through a saturator giving 5 kPa MCP (H2/MCP = 17.1) or 6.1 kPa
n-hexane (H2/n-hexane = 15.6). This reactant stream flowed over the catalyst samples as the catalyst temperature was controlled at 576–621 K (MCP) or 623– 663 K (n-hexane). In such tests all supports were entirely unreactive in both MCP and n-hexane conversions below 723 K. All products were analysed using gas chromatography with a flame ionisation detector.

3. Characterisation results
The micrograph in Fig. 2a shows the TEM-defined 1.9 nm Pt particles produced here by reaction with glucose were monodispersed, unlike the 2.3 nm particles produced by glucose–SnCl2 reaction that were aggregated (see Fig. 2b), in the same manner as

Fig. 2 Analysis of glucose-produced nanoparticles. (a) Pt sol after 40 min (scale bar = 50 nm);
(b) SnO2 sol after 45 min (scale bar = 50 nm); (c) electron diffraction of SnO2 sol (d = 0.335 nm (100%), 0.264 nm (80%), 0.236 nm (25%) and 0.176 nm (65%)) that compares well with cassiterite (0.340 nm, 0.266 nm, 0.230 nm and 0.178 nm).

were those produced previously59 on SnCl4 hydrolysis in the presence of HCl. The product in Fig. 2b gave electron diffraction patterns (see Fig. 2c) consistent with cassiterite SnO2. When Pt and SnO2 were produced they were combined in Pt/SnO2 aggregates that were at the end of the synthesis adsorbed onto Al2O3. In TPR Pt species reduced at 73–273 K and Sn species at 440–773 K. For TPR of SnO2 alone the temperature for the maximum rate of reduction (Tmax) was seen at 1051 K with
13.6 mmol H2 g—1 consumed overall, which corresponds to 103% reduction. Fig. 3 and Table 2 show TPR data for Pt100—ySnyO2y/Al2O3-Pt100—ySny/Al2O3. As y increases the contribution of reduction peaks above 273 K and associated with
Sny+ reduction increased and corresponding Tmax values increased from 595 K to 664 K. Further evidence of a Pt–Sn interaction is seen in the maximum H2 adsorption capacity (see Table 2) that was noted at intermediate y (17–33 at% Sn). For the moment it appears that Pt is reducing at much lower temperatures (o273 K) and then catalyses reduction of the Sny+ species to an extent that increases with lower Sn concentrations relative to the Pt (despite suggestions previously60 that Pt does not catalyse Sn4+ reduction). It may be that some Sn4+ reduces to Sn2+ (that is then stabilised by the alumina surface). Although, there is no clear evidence of the precise Sn oxidation state from XPS (see Fig. 3 and Table 2) or TPR (see Fig. 3 and Table 2 since Sn3d 5/2 was always at 486.2 0.2 eV and less than for SnO2), it is clear that the Pt is in close contact with the Sn and that, after 673 K reduction in H2, catalyst reduction is complete. Here the Pt-enhancement of Sn4+ reduction is affected by the [Sn]. Indeed it is likely that the low temperature TPR shoulder E480 K in Fig. 3 is associated with Pt-catalysed Sn0 formation which is especially evident at low [Sn]. Certainly XRD has found61 PtSn alloy formation.
Previously25 nanoalloy particle sizes were shown to increase as x increased. Table 2
confirms that as x increased the average size of particles seen in transmission electron microscopy (dTEM) increased,25 but in addition as x increased the extent of hydrogen chemisorption decreased by 67% at 9 at% Au (and dchem increased). Although both the Pt and the Au were zero-valent at all compositions, the Au may reduce faster than Pt and lead to some surface enrichment by Pt.

Fig. 3 X-Ray photoelectron spectra (XPS; left) and temperature programmed reduction (TPR; right) of alumina-supported Pt100—ySnyO2y/Pt100—ySny containing 1 wt% Pt and varying at% Sn relative to Pt: a: 17% Sn; b: 33% Sn; c: 43% Sn; d: 52% Sn and e: 55% Sn compared
with 1% Sn/Al2O3 (f).

Thus zero-valent Pt100—xAux nanoalloy particles (whose size increased with x) adsorbed onto graphite surfaces, and showed a level of hydrogen chemisorption at
298 K depressed by more than half when x = 9 at% Au. Pt/SnO2 nanoparticles (o3 nm in size) adsorbed onto alumina where after reduction at 673 K for 2 h was best
represented as Pt100—ySny/Al2O3 (despite some Sny+ remaining on the support surface) that showed enhanced levels of H2 chemisorption at y = 17–33 at% Sn. The surfaces
of such nanoalloy particles were now probed in situ in terms of their reactivity.

4. Characterisation through catalytic activity and selectivity
In terms of thermodynamic favourability of MCP conversion Fig. 4 shows that the reaction products are n-hexane (nHex) 4 2-methylpentane (2MP) 4 3-methylpen- tane (3MP).

Table 2 Average particle size determined by transmission electron microscopy (dTEM), extent of monolayer H2 adsorption (nH2), extent of H2 reduction and temperature of maximum rates of reduction in temperature-programmed reduction (TPR) and X-ray photoelectron spectro-
scopic (XPS) binding energies in Pt100—ySnyO2y/Pt100—ySny
dTEM/ nma mmol
H2/g cat % of H2 consumed for
Sny+ above 273 K Tmax/K for
most Sn redn Sn3d 5/2/ eVd
1.00% Pt 1.9 12.8 0.00 — —
1 wt% Pt with
17 at% Sn 14.9 10.3 595 K 486.1
33 at%Sn 14.6 49.6 614 K 486.1
43 at% Sn 9.1 43.8 621 K 486.4
52 at% Sn 67.1 634 K 486.2
55 at% Sn 2.3 67.8 664 K 486.4
a TEM after calcination at 673 K and reduction in H2 at 673 K. b Amount of H2 adsorption at 298 K, p = 0 kPa after pre-calcination and reduction in situ at 673 K and outgassing to
0.1 mPa. c The extent of H2 reduction of Pt and Sn species in samples pre-calcined at 673 K in air was at T o 273 K and T 4 273 K. The percentage of H2 consumed in Sn reduction above
273 K and the temperature (Tmax) of the dominant TPR peak for Sn reduction are shown. For SnO2 13.6 mmol H2 g—1 was consumed (corresponding to 103% reduction) with Tmax at 1051
K. d Sn3d 5/2 binding energy in samples pre-calcined in air at 673 K and then reduced in situ in
H2 at 673 K for 2 h. These should be compared with the values for Sn0 (484.9 eV), SnO2 (487.0 eV) and 1% Sn/Al2O3 (486.7 eV) and so it is difficult to differentiate 0/2+/4+ oxidation states of Sn.

Since below 723 K under present conditions C, Al2O3 and Sn/Al2O3 showed no activity in MCP conversion, reactions are assumed to be primarily initiated on the surfaces of the supported nanoalloy particles. Again Fig. 4 suggests that n-hexane conversion is more favourable to 2-methylpentane (2MP) than 3-methylpentane (3MP), but for the same reason the reaction is also thought to be associated with
(and therefore to probe) the Pt100—xAux and Pt100—ySny nanoparticle surfaces rather than the supports.
Fig. 5 shows that the conversion of MCP and n-hexane decreased with reaction time (when tested isothermally), presumably with the build-up of a carbonaceous deposit on these surfaces.
However, as a function of isothermal reaction time (see Fig. 6) the 2MP/3MP (R) product ratio
(i) increased over Pt100—xAux/C

Table 3 Average particle size of the sol Pt100—xAux seen in TEM (dTEM), extent
of monolayer adsorption of H2 at 298 K (n* ) and the average particle size
estimated from chemisorption (dchem)

n*H2/mmol H2 g—1 cat dchema/nm b/nm
Pt100Au0/C 2.55 8.0 2.3
Pt91Au9/C 0.66 31.1 4.8
a Measured after adsorption on Fluka graphite by hydrogen chemisorption at 298 K and extrapolation to p = 0 Pa. b Previously25 dTEM was measured for samples at different x (2.3 nm at x = 0; 4.8 nm at x = 10, 12.3 nm at x = 50; 14.3 nm at x = 90; 24.7 nm at x = 100).

Fig. 4 Inter-relationship of 3-methylpentane (3MP; DfH1 = —171.9 kJ mol—1 at 298 K), 2- methylpentane (2MP; DfH1 = —174.6 kJ mol—1 at 298 K), p- and multiply-bound species on surface atoms (*), n-hexane (nHex; DfH1 = —166.9 kJ mol—1 at 298 K) and methylcyclopentane (MCP; DfH1 = —106 kJ mol—1 at 298 K).

Fig. 5 Decrease in the relative conversion of (a) MCP over PtAu/C at 623 K and PtSn/Al2O3 at 603 K and (b) n-hexane over PtAu/C at 623 K and PtSn/Al2O3 at 643 K with increasing reaction time.

Fig. 6 2MP/3MP product ratio seen in (a) MCP conversion over Pt100—xAux/C at 623 K and Pt100—ySny/Al2O3 at 603 K and (b) n-hexane conversion over Pt100—xAux/C and Pt100—ySny/ Al2O3 at 643 K. Using n-hexane reactant MCP was also a significant product over Pt100—ySny/ Al2O3.

(ii) was close to constant on Pt100—ySny/Al2O3 from both MCP and n-hexane conversion. Of course MCP was another significant product from n-hexane, as were traces of C1—5 hydrogenolysis products (predominantly C3).
Fig. 7 suggests that the 2MP/3MP (R) product ratio decreased with increasing reaction temperature with both MCP and n-hexane reactants over either Pt100—xAux/ C or Pt100—ySny/Al2O3, but most especially with n-hexane over Pt100—xAux/C.
Consider then (in Fig. 8 and 9) what happens as y increases in Pt100—ySny/Al2O3.
With MCP, the initial activity decreases at 603 K as y increases. In parallel
the product 2MP/3PM ratio rises and then above y = 40 at% Sn drops. With n-hexane, first, the % conversion of n-hexane also drops as y increases at 643 K. Second, selectivity towards isomerisation to 2MP, 3MP and MCP is dominant until over 50 at% Sn is present, at which point dehydrogenation ( H2) becomes dominant. Third, the 2MP/3MP product ratio is constant at just below 2 until it drops at and above 50 at% added Sn. Therefore, beyond specific y some reactive surface atoms pairs are not available. Further, HRTEM is required to investigate this further.

Fig. 7 Effect of temperature (T) on the 2MP/3MP product ratio seen in (a) MCP and (b)
n-hexane conversion over Pt100—xAux/C and Pt100—ySny/Al2O3.

5. Discussion and conclusions
Here trisodium citrate produces zero-valent Pt100—xAux particles; as x increased the average size increased and the extent of hydrogen chemisorption decreased. These adsorbed onto graphite. Glucose produces composite aggregates of monodispersed
Pt particles (1.9 nm) and 2.3 nm SnO2 particles (similar to those seen previously59) that adsorbed onto alumina surfaces. Pt species reduced below 273 K and Sn species at 440–773 K (i.e. temperatures much lower than for SnO2 alone (1051 K) and so there was a good interaction between Pt and Sn components). Further evidence of this interaction was seen in the effect of y on Tmax values and H2 adsorption capacities (i.e. a maximum at y = 17–33 at% Sn), possibly with Pt0 formed at o273 K then catalysing reduction of the Sny+ species especially at lower [Sn]/[Pt] (despite previous suggestions60 to the contrary), possibly with a low temperature TPR shoulder suggesting some PtSn nanoalloy formation. However, some Sny+ must remain on the alumina surface.
Although MCP conversion to n-hexane, 2-methylpentane and 3-methylpentane is thermodynamically favourable, below 723 K no supports used were active. Gault envisaged that if there was equal probability of attack on all MCP bond positions in

Fig. 8 (a) Decrease in initial activity in MCP conversion at 603 K as the at% Sn added to 1 wt% Pt on Al2O3 increases. (b) 2MP/3MP product ratio in MCP ring opening at 603 K over catalysts containing 1 wt% Pt but varying at% Sn as Pt100—ySny/Al2O3.

the ring on the surface of nanoalloy particles then the product 2MP/3MP ratio (R) would be 2.
Over Pt100—xAux and Pt100—ySny nanoalloy particles the product R ratio for MCP conversion dropped as
(i) the reaction temperature rose,
(ii) as x or y increased (suggesting a similarity in these two nanoalloys) and was stable with time for PtSn, but rose for PtAu with catalyst restructuring (or deactivation).
2MP and 3MP are also favourable products of n-hexane conversion over metals. With n-hexane over Pt100—ySny/Al2O3 when held isothermally:
(i) initial conversion of n-hexane decreased as y increased
(ii) conversion decreased with increasing reaction time due to build-up of carbonaceous deposits
(iii) isomerisation was the dominant reaction over hydrogenolysis,dehydrogena- tion or isomerisation except at the highest levels of Sn addition (when it became replaced by dehydrogenation).

Fig. 9 (a) Decrease in initial activity in hexane conversion at 643 K as the at% Sn added to 1 wt% Pt on Al2O3 increases. (b) Variation in initial selectivity in hexane conversion over catalysts containing 1 wt% Pt but varying at% Sn as Pt100—ySny at 643 K to ring-opening
isomerisation in preference to hydrogenolysis to CH4–C5H12, dehydrogenation to methylcy-
clopentenes or aromatisation to benzene. (c) 2MP/3MP product ratio in n-hexane isomerisation at 643 K over catalysts containing 1 wt% Pt but varying at% Sn as Pt100—ySny/Al2O3 at 643 K.
(iv) 2MP/3MP selectivity within isomerisation decreased with reaction time with either y = 0 or 17 at% Sn (although the initial 2MP/3MP product ratio at 643 K was independent of y o 50 at% Sn and then decreased)
(v) the minor hydrogenolysis reactions produced predominantly C3 products irrespective of y (y o 40–50 at% Sn).
For n-hexane reaction over Pt100—xAux/C at 590–645 K

(i) the rate of conversion was higher for Pt/C than for Pt91Au9/C, although the activation energies (98 1 kJ mol—1) were similar, but
(ii) the 2MP/3MP product ratio is higher after gold addition to 9 at%, despite
falling with increasing temperature.
Many years ago catalytic probes were used to elucidate alloy structures (e.g. the rate of ethane hydrogenolysis was noted to drop faster on Cu addition to Ni than did the rate of dehydrogenation45). Here we have prepared two different nanoalloys by low energy bottom-up nanotechnology and then used adsorption and catalysis to probe their surface states. The reactive probes selected here (i.e. methylpentanes formation on MCP ring opening and n-hexane isomerisation) seem appropriate and useful.
It is known that ring opening of MCP to methylpentanes occurs over EuroPt-1 silica-supported Pt,62 where at 473 K and 8 o pH2 o 101 kPa the steady-state product ratio 2MP/3MP (R) E3.0. It seems that the products adsorb more weakly
on a vacant active site * than reactant MCP (i.e. C6H12+ * = C6H*12—2a + aH2, where 2a is the average number of H atoms lost per reactant MCP molecule).
In addition it has been suggested63 that a partially selective ring-opening might prevail at higher temperatures. Over ZSM-5 it converts predominantly to cyclohex- ane with hydride transfer64 while hexane isomers are formed more on metals,65 where larger Pt particles (i.e. dPt = 17 nm) can lead to more selective MCP hydrogenolysis (i.e. 85% selective MCP hydrogenolysis giving higher 2MP/3MP product ratios than 2) than smaller particles (i.e. when at dPt = 8.5 nm there was 0% selective MCP hydrogenolysis leading to 2MP/3MP = 2 (i.e. the statistical and thermodynamic equilibrium value66)).67 Not surprisingly then sulfation of Pt/ZrO2 does not raise the 2MP/3MP (R) product ratio in MCP conversion, even though this is 3.3–4.2 at 313–380 K over sulfated Pt/ZrO2; rather the ratio is thought to be defined by the nature of surface sites.68 Thus pairs of contiguous surface atoms (where one is electron deficient and an asymmetric site pair exists) are said68 to enhance rates of formation of 2MP. The drop in product R ratio at higher y must indicate a change in the availability of contiguous Pt atom * pairs at this surface composition; HRTEM is being used to investigate this further.
As with alumina-supported Pd–Re surfaces, it seems that the 2MP/3MP product ratio arising from n-hexane and MCP conversion is a useful ‘chemical fingerprint’ for such bimetallic catalysts.69 Certainly Pd–Pt surfaces have been studied using the 2MP/3MP ratio arising from conversion of cis or trans methyl-ethyl-cyclopro- pane (MECP) and MCP.70,71 Of course one could also consider 2MP/hexane product ratios for MCP conversion, but to enable both reactions to be followed in the same way only 2MP/3MP ratios in the products of MCP and n-hexane conversion were considered here.
The present results suggest that catalytic probes do provide another insight into the nature of nanoalloy surfaces.
1–5 nm bimetallic nanoparticles72 are of increasing importance. They have useful optical,73,74 catalytic,75 superparamagnetic,76 fuel cell77 and bio-78 properties and applications. It may be that catalytic probes will enhance this potential and allow their potential to be realised that much faster.

The authors thank EPSRC for support of K. A. G. through the provision of a studentship.

1 E. K. Rideal, Concepts in Catalysis, Academic Press, New York, 1968, p. 8.
2 D. I. Gittins and F. Caruso, ChemPhysChem, 2002, 3, 111.
3 S. H. Overbury, P. A. Bertrand and G. A. Somorjai, Chem. Rev., 1975, 75, 547.
4 F. G. Meng, H. S. Liu, L. B. Liu and Z. P. Jin, J. Alloys Compd., 2007, 431, 292.

5 L. O. Paz-Borbon, R. L. Johnston, G. Barcaro and A. Fortunelli, J. Phys. Chem. C, 2007,
111, 2936.
6 D. Cheng, S. Huang and W. Wang, Eur. Phys. J. D, 2006, 39, 41.
7 A. Rapallo, G. Rossi, R. Ferrando, A. Fortunelli, B. C. Curley, L. D. Lloyd, G. M. Tarbuck and R. L. J. Ohnston, J. Chem. Phys., 2005, 122, 194308.
8 S. Darby, T. V. Mortimer-Jones, R. L. Johnston and C. J. Roberts, J. Chem. Phys., 2002,
116, 1536.
9 X. D. Dai, Y. Kong and J. H. Li, Phys. Rev. B, 2007, 75, 104101.
10 S. Sahoo, G. Rollmann and P. Entel, Phase Transitions, 2006, 79, 693.
11 D. J. Cheng, W. C. Wang and S. P. Huang, J. Phys. Chem. B, 2006, 110, 16193.
12 (a) N. A. Zarkevich, T. L. Tan and D. D. Johnson, Phys. Rev. B, 2007, 75, 104203;
(b) V. S. K. Balagurusamy, R. Streitel, O. G. Shpyrko, P. S. Pershan, M. Meron and B. H. Lin, Phys. Rev. B: Condens. Matter, 2007, 75, 104209.
13 C. Massen, T. V. Mortimer-Jones and R. L. Johnston, J. Chem. Soc., Dalton Trans., 2002, 4375.
14 A. S. Shirinyan and M. Wautelet, Nanotechnology, 2004, 15, 1720.
15 D. J. Cheng, S. P. Huang and W. C. Wang, Phys. Rev. B: Condens. Matter, 2006, 74, 064117.
16 A. Aguado and J. M. Lopez, J. Chem. Theor. Comput., 2005, 1, 299.
17 N. T. Wilson and R. L. Johnston, J. Mater. Chem., 2002, 12, 2913.
18 L. D. Lloyd, R. L. Johnston, S. Salhi and N. T. Wilson, J. Mater. Chem., 2004
14, 1691.
19 L. D. Menard, H. P. Xu, S. P. Gao, R. D. Twesten, A. S. Harper, Y. Song, G. L. Wang, A.
D. Douglas, J. C. Yang, A. I. Frenkel, R. W. Murray and R. G. Nuzzo, J. Phys. Chem. B, 2006, 110, 14564.
20 S. Pal and G. De, J. Mater. Chem., 2007, 17, 493.
21 Y. B. Lou, M. M. Maye, L. Han, J. Luo and C. J. Zhong, Chem. Commun., 2001, 473.
22 T. Montanari, O. Marie, M. Daturi and G. Busca, Appl. Catal., B, 2007, 71, 216.
23 M. O. Nutt, K. N. Heck, P. Alvarez and M. S. Wong, Appl. Catal., B, 2006, 69, 115.
24 E. G. Szabo, A. Tompos, M. Hegedus, A. Szegedi and J. L. Margitfalvi, Appl. Catal., A, 2007, 320, 114.
25 P. A. Sermon, J. M. Thomas, K. Keryou and G. R. Millward, Angew. Chem., Int. Ed. Engl., 1987, 26, 918.
26 Y. H. Lee, G. Lee, J. H. Shim, S. Hwang, J. Kwak, K. Lee, H. Song and J. T. Park, Chem. Mater., 2006, 18, 4209.
27 E. Antolini, F. Colmati and E. R. Gonzalez, Electrochem. Commun., 2007, 9, 398.
28 R. Srinivasan, R. J. de Angelis and B. H. Davis, J. Catal., 1987, 106, 449.
29 H. Lieske and J. Volter, J. Catal., 1984, 90, 96.
30 A. O. Neto, R. R. Dias, M. M. Tusi, M. Linardi and E. V. Spinace, J. Power Sources, 2007,
166, 87.
31 C. Lamy, E. M. Belgsir and J. M. Leger, J. Appl. Electrochem., 2001, 31, 799.
32 D. R. Lycke and E. L. Gyenge, Electrochim. Acta, 2007, 52, 4287.
33 O. Trapp, Electrophoresis, 2007, 28, 691.
34 M. Bruncko, I. Anzel and A. Kneissl, Corros. Sci., 2007, 49, 1228.
35 G. A. Somorjai and A. L. Marsh, Philos. Trans. R. Soc. London, Ser. A, 2005, 363, 879.
36 Z. Paal, A. Wootsch, I. Bakos, S. Szabo, H. Sauer, U. Wild and R. Schlogl, Appl. Catal., A, 2006, 309, 1.
37 C. Dossi, A. Pozzi, S. Recchia, A. Fusi, R. Psaro and V. Dal Santo, J. Mol. Catal. A: Chem., 2003, 204, 465.
38 Z. Paal, A. Wootsch, R. Schlogl and U. Wild, Appl. Catal., A, 2005, 282, 135.
39 (a) S. M. Davis, F. Zaera and G. A. Somorjai, J. Am. Chem. Soc., 1982, 104, 7453; (b) J. A. Dalmon and G. A. Martin, J. Catal., 1999, 66, 214.
40 A. Wootsch and Z. Paal, J. Catal., 1999, 185, 192.
41 H. Matsumoto, Y. Saito and Y. Yoneda, J. Catal., 1970, 19, 101.
42 J. R. Anderson and Y. Shimoyam, Proc. 5th Int. Cong. Catal., 1972, 48.
43 (a) G. Maire, G. Plouidy, J. C. Prudhom and F. G. Gault, J. Catal., 1965, 4, 556; (b) J. R. Anderson, Adv. Catal., 1974, 23, 1.
44 J. H. Sinfelt, Catal. Rev. Sci. Eng., 1969, 3, 175.
45 J. H. Sinfelt, D. J. C. Yates and J. L. Carter, J. Catal., 1972, 24, 283.
46 M. Boudart, A. W. Aldag, L. D. Ptak and J. E. Benson, J. Catal., 1968, 11, 35.
47 M. Faraday, Philos. Trans. R. Soc. London, 1857, 147, 145. 48 G. Frens, Nat. Phys. Sci., 1973, 241, 20.
49 S. Pande, S. K. Ghosh, S. Praharaj, S. Panigrahi, S. Basu, S. Jana, A. Pal, T. Tsukuda and T. Pal, J. Phys. Chem. C, 2007, 111, 4596.

50 G. Barcaro, A. Fortunelli, G. Rossi, F. Nita and R. Ferrando, J. Phys. Chem. B, 2006, 110, 23197.
51 P. W. Zheng, X. W. Jiang, X. Zhang, W. Q. Zhang and L. Q. Shi, Langmuir, 2006, 22, 9393.
52 Y. Suchorski, J. Beben, A. Frac, V. K. Medvedev and H. Weiss, Surf. Interface Anal., 2007, 39, 161.
53 (a) J. Turkevich, P. C. Stevenson and J. Hillier, Discuss. Faraday Soc., 1951, 11, 55; (b) L.
G. Tejuca, K. Aika, S. Namba and J. Turkevich, J. Phys. Chem., 1977, 81, 1399; (c) D. N. Furlong, A. Launikonis, W. H. F. Sasse and J. V. Sanders, J. Chem. Soc., Faraday Trans. 1, 1984, 80, 571.
54 (a) M. J. Yacaman and J. M. Domingueze, J. Catal., 1980, 64, 213; (b) D. A. Jefferson, J.
M. Thomas, G. R. Millward, K. Tsuno, A. Harriman and R. D. Brydson, Nature, 1986,
323, 428.
55 R. T. K. Baker, Catal. Rev. Sci. Eng., 1979, 19, 161.
56 J. M. Thomas, Faraday Discuss., 1996, 105, 1.
57 J. M. Cowley and R. J. Plano, J. Catal., 1987, 108, 199.
58 S. M. Augustine and W. M. H. Sachtler, J. Catal., 1987, 106, 417.
59 M. Ocana, C. J. Serna and E. Matijevic, Colloid Polym. Sci., 1995, 273, 681–686. 60 R. Burch, J. Catal., 1981, 71, 348.
61 S. R. Adkins and B. H. Davis, J. Catal., 1984, 89, 371.
62 Y. P. Zhuang and A. Frennet, Appl. Catal., A, 1999, 177, 205. 63 F. G. Gault, Adv. Catal., 1981, 30, 1.
64 K. H. Lee and D. Farcasiu, J. Chem. Eng. Jpn., 2001, 34, 1557.
65 T. J. McCarthy, G. D. Lei and W. M. H. Sachtler, J. Catal., 1996, 159, 90.
66 B. A. Lerner, B. T. Carvill and W. M. H. Sachtler, J. Mol. Catal., 1992, 77, 99.
67 J. M. Dartigues, A. Chambellan and F. G. Gault, J. Am. Chem. Soc., 1976, 98, 856.
68 M. R. Smith, J. K. A. Clarke, G. Fitzsimons and J. J. Rooney, Appl. Catal., A, 1997, 165, 357.
69 W. Juszczyk and Z. Karpinski, Appl. Catal., A, 2001, 206, 67.
70 N. Gyorrfy, L. Toth, M. Bartok, J. Ocsko, U. Wild, R. Schlogl, D. Teschner and Z. Paal,
J. Mol. Catal. A: Chem., 2005, 238, 102.
71 A. Barrera, J. A. Montoya, M. Viniegra, J. Navarrete, G. Espinosa, A. Vargas, P. del Angel and G. Perez, Appl. Catal., A, 2005, 290, 97.
72 Metal Nanoparticles Synthesis Characterisation and Applications, ed. D. L. Feldheim and
C. A. Foss, Marcel Dekker, New York, 2002.
73 G. L. Hornyak, C. J. Patrissi, E. B. Oberhauser, C. R. Martin, J. C. Valmalette, L. Lemaire, J. Dutta and H. Hofmann, Nanostruct. Mater., 1997, 9, 571.
74 G. Mattei, G. Battaglin, E. Cattaruzza, C. Maurizio, P. Mazzoldi, C. Sada and B. F. Scremin, J. Non-Cryst. Solids, 2007, 353, 697.
75 I. Pastoriza-Santos, J. Perez-Juste, S. Carregal-Romero, P. Herves and L. M. Liz-Marzan,
Chem.–Asian J., 2006, 1, 730.
76 M. Mandal, S. Kundu, T. K. Sau, S. M. Yusuf and T. Pal, Chem. Mater., 2003, 15, 3710.
77 N. M. Galea, D. Knapp and T. Ziegler, J. Catal., 2007, 247, 20.
78 M. Noyong, K. Gloddek, J. Mayer, T. Weirich and U. Simon, J. Cluster Sci., 2007, 18, 193.PT-100