Surface segregation of Pt–Rh alloysN. Sano and T. Sakurai Citation: Journal of Vacuum Science & Technology A 8, 3421 (1990); doi: 10.1116/1.576525 View online: http://dx.doi.org/10.1116/1.576525 View Table of Contents: http://scitation.aip.org/content/avs/journal/jvsta/8/4?ver=pdfcov Published by the AVS: Science & Technology of Materials, Interfaces, and Processing Articles you may be interested in Entropydriven surface segregation of Pt in PtRh alloys J. Vac. Sci. Technol. A 5, 558 (1987); 10.1116/1.574672 Compositional variations in the near surface layers, an atomprobe study of cosegregation of sulfur in Pt–Rh andPt–Ir alloys J. Chem. Phys. 83, 388 (1985); 10.1063/1.449782 A nonmonotonic concentration depth profile of Pt–Rh alloys: A surface segregation study using the atomprobefield ion microscope J. Vac. Sci. Technol. A 3, 806 (1985); 10.1116/1.573315 Abstract: Surface composition of Pt/Rh alloys J. Vac. Sci. Technol. 16, 663 (1979); 10.1116/1.570051 NMR and Susceptibility Studies of PtRh Alloys AIP Conf. Proc. 18, 297 (1974); 10.1063/1.2947337

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 193.0.65.67 On: Fri, 28 Nov 2014 11:24:53

Surface segregation of Pt-Rh alloys N. Sano and T. Sakurai The InstituteforSolid State Physics, The UniversityofTokyo, 7-22-1 Roppongi, Tokyo 106, Japan

(Received 21 August 1989; accepted 5 March 1990)

An atom-probe field ion microscope (APFIM) has been employed to determine the surface composition and depth profiles of the (100) plane ofPt-20, 40 wt. % (32.1,55.8 at. %) Rh alloys. The enrichment of Pt at the top surface layer on annealing at 700·C (- 1000 K) was observed, agreeing with previous studies by Auger electron spectroscopy, ion scattering spectroscopy, and APFIM. On annealing below 600 ·C, however, a reversed surface segregation, Rh enrichment at the first surface, has been observed even without any impurity atoms such as S and P. Upon annealing up to 700·C in the presence of oxygen, we found no appreciable oxidation of the alloy samples, instead, there appeared chemisorbed oxygen forming an overlayer.

I. INTRODUCTION

Surface segregation of Pt-Rh alloys has been studied with many surface analytical techniques, such as Auger electron spectroscopy (AES), 1.2 ion scattering spectroscopy (ISS), 3

and atom-probe field ion microscope (APFIM). 4 The inter­est in Pt-Rh binary alloys has been enhanced by the fact that: they are industrially important autocatalysts, and their surface segregation behavior cannot be accounted for by the simple ideal solution model. It predicts Rh enrichment at the surface layer, but many studies have revealed Pt segregation at the top layer, instead of Rh. The contribution of lattice vibration entropy to the surface segregation has been pro­posed5

•6 and applied to the Pt-Rh system. 7

Using APFIM, Ren and Tsong8 have reported that Pt segregates at the (100) and (111) surfaces of high-purity Pt-44.8 at. % Rh alloys equilibrated at 700 ·C and that Rh segregation was observed in impurity sulfur-containing al­loys. They attributed the latter to cosegregation with S, which formed an overlayer even with a very small amount ( 100 ppm). 9 The influence of traces of impurities has been discussed by van Delft et aU

At high temperatures, an entropy term may play the dominant role for the surface segregation of Pt-Rh alloys. Thus, it might be natural to expect that below certain tem­peratures an enthalpy term overcomes the entropy effect and that Rh would surface segregate as predicted with ideal solu­tion models. We report here the surface segregation study of the (100) plane of Pt-20, 40 wt. % (32.1,55.8 at. %)Rh alloys annealed at 700· and below 6OO·C using a TOF (time-of-flight) APFIM. 1O

II. EXPERIMENTAL

High purity commercial Pt-20, 40 wt. % Rh wires 0.2 mm in diameter were used in this study. To remove impuri­ties, the wires were preannealed by Joule heatings at 1000·C for 1 h in 1 X 10- 4 Torr high purity O2 flow. Field ion mi­croscopy (FIM) tips were prepared from these wires byelec­trochemical etching in molten salt bath of NaN03 :

NaCI = 4: 1. FIM tips were once again preannealed around 800 ·C for 30 s in a main chamber of the APFIM (4 X 10 - 10

Torr) to smooth the tip surface prior to observing a field ion image. An FI image was observed with 2 X 10- 5 Torr heli­um as an imaging gas at 40 K and the (100) pole was adjust-

ed to a probe hole. While He gas was evacuated, the direct current (dc) high voltage was kept unchanged. It was re­duced to 50% of the imaging voltage when the tip tempera­ture was raised up to 140 K in order to pump out absorbed gas on a cold refrigerator and the sample holder. After cool­ing the tip down to 40 K, the dc voltage was turned down to zero and followed by annealing by Joule heating under 4X 10- 10 Torr ultra-high vacuum (UHV). The tip tem­perature during annealing at 700·C was measured with an optical pyrometer, and below 600 ·C, at which the FIM tip could not be seen in red, it was estimated by an extrapolation of the relationship between the tip temperature and electric power supplied to the tip, measured over the range of 700-1200 ·C. Quenching from the annealing temperature was done by merely turning off the current for Joule heating. Since the sample holder was kept at cryogenic temperature and the surface-ta-volume ratio of the FIM tip was very large, the quenching rate was considered to be fast enough to freeze equilibrated phase. On finishing the anneal, we soon started atom-probe analyses. It should be noted here that the FIM tip was not imaged after the annealing, thus, the tip surface was free from gas contamination and gas-induced preferential field evaporation. Annealing in the presence of oxygen was carried out similarly, except that the tempera­ture at the cold refrigerator was kept at 140 K in order to prevent oxygen from being condensed and pumped out easi­ly after the annealing.

The conditions of atom-probe analyses were; pulse frac­tion, 0.15~.2; tip temperature, 40 K; pressure, 4 X 10 - 10

Torr. These conditions had been confirmed by a preliminary study to give correct bulk composition within the accuracy of ±2%.

III. RESULTS In. the present atom-probe analyses, both platinum and

rhodIUm were detected doubly or triply charged ions. Sulfur was detected neither at the top surface layer nor in the bulk in contrast to the AP data by Ahmad and Too\\~,9 ",b.<l £e: ported that S formed an overlayer even with traces of S as 100 ppm. From the so-called ladder-step diagram (the num­ber of detected ions was plotted against the number of high voltage trigger pulses), the layer-by-Iayer composition depth profiles were determined. Figure 1 shows the ladder-

3421 J. Vac. Sci. Technol. A 8 (4), Jul/Aug 1990 0734-2101/901043421-04$01.00 @ 1990 American Vacuum Society 3421

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 193.0.65.67 On: Fri, 28 Nov 2014 11:24:53

3422 N. Sano and T. Sakurai: Surface segregation of Pt-Rh alloys

400 --------,---, Pt-40wt.%( 55.8 at.%1 Rh

.r::: 700"C 2 min

a: it '0 ... Q) .0 E :J Z

0 0 Number of pulses

400~-------------------------.~

.r::: a: -a: '0

~ E ::J Z

Pt-40wt.%155.8at.%IRh (100)

600"C 20 mIn

0'----________ . _____ . _______ _

o Number of pulses 1.8 x 106

FIG. 1. Ladder-step diagrams of the (100) plane of Pt-40 wt. % (55.8 at. %) Rh alloys annealed; (a) at 700 ·C for 2 min. (b) at (below) 6OO·C for 20 min. Vertical lines separate individual (100) planes.

step diagrams obtained for the (100) plane of Pt-40 wt. % Rh annealed at 700 ·C and below 600 ·C. As is shown in Fig. 1, approximately 40--70 ions were collected from each layer of the (100) plane field-evaporated. Thus, the statistical er­ror (the standard deviation 0') ofRh concentration at indi­vidual (100) planes, which is given by ~x( 1 - x)/n, where x is the atomic fraction of Rh and n is the number of ions collected from the each (100) plane, is ± 8 at. % at maxi-

Pt-40wt.%(55.8 at.%1 Rh (100) 80

~ 70 ! -IV

"- 50

c: 50 .Q - ! IV ... 40 -c:

CD u 30 c: 0 u ~ 20 • 700"C 2mln a:

10 • 600 "C 20 min

0 2 3 4 5 6 7 8 9

Number of layers

FIG. 2. Composition depth profiles of the (100) plane ofPt-40 wt. % (55.8 at. %) Rh alloys. The error bars show the statistical error (the maximum standard deviation ± u) of each datum point.

J. Vac. Sci. Technol. A, Vol. 8, No.4, Jull Aug 1990

c .2 .... tV ... .... C CD U C o U

.I: a:

80

70

60

50

40

30

20

10

Pt - 20wt.%\ 32.1 at.% I Rh

(100 )

• 700°c 1min

• 800°c 2min

1

Number of layers

3422

FIG. 3. Composition depth profiles of the (100) plane ofPt-20 wt. % (32.1 at. %) Rh alloys.

mum. Figure 2 is composition depth profiles with atomic layer resolution obtained from Fig. 1. Error bars at each datum point shown ± 0' calculated by the above formula. On annealing at 700 ·c, Pt is enriched at the first layer and depleted at the second layer. Annealing below 6OO'C re­versed the surface segregation, namely, Rh enrichment is observed at the top surface layer and its depletion is observed at the second layer. On both annealing temperatures, an os­cillatory behavior of composition appeared to exist over ten atomic layers. Similar oscillation has been found for Pt-Rh alloys by an APFIM study.8.9." Composition depth profiles of the (l00) plane ofPt-20 wt. % Rh are shown in Fig. 3.

Pt - 20 wt.%( 32.1 at.% 1 Rh (100 ) 80

~ 70 Without Annealing -IV

"-60

c 50 .Q -IV ... 40 -c: CD u 30 c: 0 u

20 ~ a:

10

0 2 3 4 5 6 7 8 9 10

Number of layers

FIG. 4. A composition depth profile of the (100) plane of the Pt-20 wt. % (32.1 at. %) Rh alloy without annealing.

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 193.0.65.67 On: Fri, 28 Nov 2014 11:24:53

3423 N. Sano and T. Sakurai: Surface segregation of Pt-Rh alloys

FIG. 5. Field ion images of a Pt-20 wt. % Rh alloy annealed at about 300 'C for 5 min in I X 10 - • Torr 0,; (a) before annealing, clean surface; (b) after annealing, the surface began to be imaged at lower applied voltage, where somehow disordered spots were observed along with the substrate clean surface atoms; (c) as the applied voltage increased, the disordered over­layer easily field evaporated; (d) clean surface was recovered.

Similar to the result of Pt-40 wt. % Rh, the reversion of the surface segregant with decreasing temperature was ob­served. Although we did not see clearly the Pt segregation on annealing at 700 ·C, Rh did segregate at the surface on lower annealing temperature. To il1ustrate the effect of annealing upon the composition depth profiles, we show the case for a Pt-20 wt. % Rh alloy without annealing in Fig. 4. Com­pared with the annealed samples shown in Figs. 2 and 3, the nonannealing sample has more uniform Rh distribution from the surface to over ten atomic layers, which substan­tiates the effect of annealing demonstrated in this study.

Figure 5 shows FI images ofPt-20 wt. % Rh annealed at 300 ·C for 5 min in I X lO - 6 Torr O2 • A clean surface [Fig. 5(a)] was covered with disordered spots after annealing in oxygen (b), and these spots were easily removed by field

i 20Rh - 300't: 10 min 1 x10-s Torr 02 0

Q) 200 Rh2+ -Q) '0 fI) 150 c .Q - 100

Pt2+ 0 ~ Q)

.J:l 50 H+ 3+ E u ~+i

Pt3+ ::J 1 z 0

0 50 100 150 200

Mass to charge ratio: mIn

FIG. 6. A mass spectrum ofa Pt-20wt. % Rhalloy annealed at about 300 'C for 10 min in I X 10 'Torr °2 ,

J. Vac. ScI. Technol. A, Vol. 8, No.4, JuliAug 1990

3423

FIG. 7. Field ion images of a Pt-40 wt. % Rh alloy annealed at 700'C for 2 min in I X 10 ' Torr 02; (a)-(el ehemisorbed oxygen started to be imaged below the best image voltage (BlV); (d) clean surface was recov­ered.

evaporation of a few layers (C), and the clean surface was recovered (d). From the series of these pictures, it seems that a thin overlayer consisting of chemisorbed oxygen was formed during the annealing in O2 , In Fig.· 6, a mass spec­trum of Pt-20 wt. % Rh annealed at 300'C for lO min in 1 X lO - 5 Torr O2 is shown. We see no oxides of Pt or Rh, but at the beginning of the probe, a few oxygen ions were detected. Figure 7 illustrates an oxygen overlayer formed in the annealing of Pt-40 wt. % Rh at 700'C for 2 min in I X lO - 5 Torr O2 . A clean surface was recovered by single layer field evaporation. Large facets were observed on the ( 1(0) and OlO) planes. This suggests that oxygen atoms were not bound tightly with the surface atoms.

IV. DISCUSSION

It has been reported so far that annealing at lOOO K ofPt­Rh alloys with any Rh concentration resulted in Pt segrega­tion at the top surface. Present AP study agrees with it, but upon annealing below 600·C surface enrichment ofRh was observed. A similar tendency has been obtained for Pt-lO wt. % Rh alloys, and is now under investigation. It is ques­tionable whether the tip surface composition was equilibrat­ed at 6OO·C for 20 min. Nevertheless, it is the experimental fact that Rh concentration increased at the top layer sub­stantially compared with nonannealing samples beyond the level of the statistical error .

Low temperature surface segregation of Pt-Rh alloys seems to be open to question. Below 1000 K, the \lOssibility ofRh enrichment at the top layer has been suggested by van Delft and Neiuwenhuys, I and Langeveld and Niemantsver­driet. 12 Langeveld and Niemantsverdriee have assumed that the contribution of the lattice vibration entropy to the surface segregation aSvib can be expressed using Debye tem-

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 193.0.65.67 On: Fri, 28 Nov 2014 11:24:53

3424 N. Sano and T. Sakurai: Surface segregation of Pt-Rh alloys

perature () as follows:

ASvib = [Svib (Pt,urf) - Svib (Ptbu1k )]

- [Svib (Rh,urf) - Svib (Rhbu,k ) ] ,

Svib = 3R [1 -In(() In]'

(1)

(2)

where R is the gas constant. Therefore, ASvib is rewritten as

tl.Svib = 3R {In ()( Ptbu1k ) . ()( Rhsurf ) }. (3) ()( Ptsurf ) . (Rhbu,k )

Substituting the corresponding () values of ()(Ptbu1k ) = 230 K, ()( Pt,urf) = 118 K,13 ()( Rhbu,k ) = 350 K, ()( Rhsurf ) = 200 K,14 into Eq. (3), we obtain tl.Svib = 2.7 J Imol K.

On the other hand, the enthalpy term for surface segrega­tion tl.H scgr is generally written was

tl.Hsegr = (Z"IZ) [tl.HsUb (Pt) - tl.Hsub (Rh» (4)

in which Z is the number of neighboring atoms in the bulk, Z" is the number of missing atoms at the surface, and tl.Hsub is the enthalpy for sublimation. Z"IZ is 1/3 for the (100) plane of face-centered-cubic (fcc) metals, and tl.H.ub (Pt) = 565.7 ± 1.3 kJ/mol, AHsub (Rh) = 557 ± 4 kJ/mol,15 thus by substituting those values into Eq. (4), we obtain tl.Hsegr = 2.9 ± 1.4 kJ Imol. If the value 2.9 kJ Imol is used for the enthalpy term in the free energy for surface segregation tl.Gscgr the entropy term T tl.Svib counterba­lances the enthalpy term at 1074 K, at which temperature tl.G.egr becomes zero. This suggests that at low temperatures Rh would surface segregate due to the dominant enthalpy term and that the segregant would be changed to Pt with increasing temperature, where the entropy term becomes dominant.

Our results are consistent with the predictions1.12 and are in good agreement with the abovementioned simple calcula­tion incorporating the lattice vibration entropy effect. How-

J. Vac. Sci. Technol. A, Vol. 8, No.4, Jul/Aug 1990

3424

ever, there are some shortcomings in this simple calculation, that is, the degree of surface segregation should be consider­ably reduced because of near-zero free energy for segrega­tion.

V. CONCLUSIONS

The surface composition and depth profiles of the (100) plane of Pt-20, 40 wt. % (32.1, 55.8 at. %) Rh alloys were investigated using the APFIM. Upon annealing at 700 ·C, Pt is enriched at the top surface layer, agreeing with previous studies, but below 600 ·C Rh has been found to segregate at the surface. No appreciable oxidation was found during an­nealing up to 700 ·C in 1 X 10 - 5 Torr O2 , It appeared that a thin overlayer was formed by chemisorbed oxygen.

'F. C. M. J. M. van Delft and B. E. Nieuwenhuys. Surf. Sci. 162. 538 (1985).

'F. c. M. J. M. van Delft, A. D. van LangeveJd, and B. E. Nieuwenhuys, Surf. Sci. 189/190,1129 (1987).

1 F. L. Williams and O. C. Nelson, App\. Surf. Sci. 3, 409 (1979). 4T. T. Tsong, D. M. Ren, and M. Ahmad, Phys. Rev. B 38, 7428 (1988). 5K. Hoshino, J. Phys. Soc. lpn. 50, 577 (1981). 6 A. Sakai and T. Sakurai, Surf. Sci. 138, 159 (1984). 7 A. D. van Langeveld and 1. W. Niemantsverdriet, l. Vac. Sci. Techno!. A 5,558 (1987).

• D. Ren and T. T. Tsong, Surf. Sci. 184, L439 ( 1987). "M. Ahmad and T. T. Tsong, 1. Chern. Phys. 83,388 (1985). lOT. Sakurai, A. Sakai, and H. W. Pickering, Advances in Electronics &

Electron Physics Suppl. 20 (Academic, New York, 1989), pp. 1-299. 11 M. Ahmad and T. T. Tsong, Surf. Sci. 149, L7 (1985). "A. D. van LangeveJd and 1. W. Niemantsverdriet, Surf. Sci. 178, 880

(1986). 13 H. B. Lyon and O. A. Somorjai, 1. Chern. Phys. 44, 3707 (1966). 14D. O. Castner, O. A. Somorjai, J. E. Black, D. Castiel, and R. F. Wallis,

Phys. Rev. B24, 1616 (1981). 15 Handbook ojChemistry and Physics, 64th ed. (Chemical Rubber, Cleve­

land,OH,1983-1984).

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 193.0.65.67 On: Fri, 28 Nov 2014 11:24:53