7 ADC-A09 290 IBM THOMAS J WATSON RESEARCH CENTER TORKTOWN HEIGHTS NT FIG 7/2THE INTERACTION OF SILICON WITH A PD (100) SURFWCE.(UIDEC 80 K CHRISTMANN, J DEMUTH N00014-77-C-0366

UNCLASSIFIED TR-14 NL

EN7

LEVEL'OFFICE OF NAVAL RESEARCH

iJ

Contract N00014-77-C0366 b

Task No. NR 056-123

TECHNICAL REPORT NO.14

THE INTERACTION OF SILICON WITH A Pd(100) SURFACE

by /

K. Christmann and J.E. Demuth 'J

Prepared for Publication

in

Proceedings of the 5th International Vacuum Congress

Cannes, France

IBM T.J. Watson Research Center

Yorktown Heights, NY 10598

December 9, 1980

Reproduction ii whole or in part is permitted for

any purpose of the United States Government

0D

* This document has been approved for public releaseJand sale; its distribution is unlimited

80 24 032

UNCLASSIFIEDSEiuRITY CLASSIFICATION OF TH4IS PAGE (W7hen Date Enitrd) __________________

REPORT DOCUMENTATION PAGE BEFORE COPEIGFORM- - 2.GOVT ACCESSION NO J. RZCIPICNT'S CATALOG NUMBER

Te J_/(~1Technical epart, i 4_____/1___

TITLE S. TYPE OF it

7. AUTHOR a. CONTRACT OR qgAhLTMim*etw()

~K T Christmann an J.E./eth5 Wq047-X-36

9. PERFORMING ORGANIZATION NAME AND ADDRESS.POGA LEET P~ET.TS

IBM T.J. Watson Research Center, P.O. Box218, Yorktown Heights, New York 10598

Ii. CONTROLLING OFFICE N AMC AND ADDRESS I2

Office of Naval Research N. UMUER Of PAGESChemistry Program Office

4. MO10NT 118 Za A N Y N Mz O R S(Id SECURITY CLASS. (01 this reo rt)

DS. ECLASSIFICATION, DOWNGRADINGSCHEDOULE

IS. DISTRIBUTION STATEMENT (of this Report)

Approved for Public Release; Distribution Unlimited.

17. DISTRIBUTION STATEMENT (of the obettedt entered in Stock 20. it differet from Report)

IS. SUPPLEMENTARY NOTYES

Preprint; to be published in Proceedings of the 5th InternationalVacuum Congress, in Cannes, France.

1S. KEY WORDS (Continue on reverse side if noceesswir nd identify by block number)

Sulicide, Palladium, Silicon, Chemisorption

211 ABSTRACT eContinue n reverse side if necessery mod Identify by block number)+The interaction of silicon with a clean Pd(100) surface wasstudied i~nL1VHV using LEED, AES and UV-potoemission. Depositionaround 75 C leads to disordered Si layers with little evidence ofbulk diffusion or compound formation. Slight warming to-15O'Ccauses the immediate development of a complex LEED pattern indica-tive of long-range ordering. A pronounced splitting of the Si LVVAuger line as well as the appearance of a photoemission level22.5 eVbelow E. is attributed to the formation of Palladium (Ifl.I

60 143 " 'M d UNCLASSIFIED

SECURITY. CLSI.C.M F rTI 41 11H ~

LNCLASS IFIED

SIECURITY CLASSIFICATION OF THIS pO4gCPhOV 004160"~rn) /

silicide, Pd2Si. Annealing at 250-500 C leads to decompositionof this silicide as indicated by the destruction of the complexLEED pattern and the disappearance of the Pd Si Auger lines.Above 600"C diffusion of silicon into the buik predominateswhich leads to a rapid depletion of silicon on the Pd surfaceas can be followed from both AES and UPS.

!I

!V

D ..... .. ....

Dist L .c[

I.

UNCWASIVILD86CUNITY Ct.A PICATtIO OF T;gs PA64 r-W - Do* -.- 0 -- mj

. ..

RC 8300 (36092) 6/2/80Surface Science 4 pages

The Inter..tion of silicon with a Pd(lO0) Surfacet

K. Christmann* and J.E. Demuth

IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598New York 10598

Abstract: The interaction of silicon with a clean Pd(100) surface was studied in UHV using

LEED, AES and UV-photoemission. Deposition around 75*C leads to disordered Si layers with

little evidence of bulk diffusion or compound formation. Slight warming to - 1500C causes the

immediate development of a complex LEED pattern indicative of long-range ordering. A

pronounced splitting of the Si LVV Auger line as well as the appearance of a photoemission level

- 2.5 eV below Ef is attributed to the formation of palladium (II) silicide, Pd2Si. Annealing at

250-500*C leads to decomposition of this silicide as indicated by the destruction of the complex

LEED pattern and the disappearance of the Pd2 Si Auger lines. Above 600*C diffusion of silicon

into the bulk predominates which leads to a rapid depletion of silicon on the Pd surface as can be

followed from both AES and UPS.

Permanent address: Institut fur Physikalische Chemie der Universitat

Munchen, D-8000 Munchen 2, West Germany

tWork supported in part by the Office of Naval Research.

1. Introduction

The reaction between silicon and metals is of considerable practical importance. Traces of Si mayeasily lead to embrittlement and fracture of many common metals (Fe, Al); whereas larger Siamounts result in silicide formation, important compounds for manufacturing semiconductor devicecontacts.

Early structural investigations of Si on metal surfaces utilizing LEED have been reported by Jona,e.g., for Al [11, Ni [2] and Be (3] single crystal faces. From these studies it could be inferred thatSi strongly reacts with the metal rather than merely forms a normal type of chemisorption bond.Quite recently, Cardillo and Becker [4] investigated the precipitation of Si impurities on a Pt(100)surface by means of LEED/AES and He atom diffraction. They showed that submonolayerconcentrations of silicon on this surface readily form Pt(II) silicide Pt 2Si with a surprisingly largelong-range order. Similar observations were made with the Pt( 11 l)/Si system [5]. As far astransition metal silicides are concerned, an extensive data body exists (61, in particular forpalladium(II)silicide, Pd2 Si, and some other silicides of Ni, Pt, Co, Rh, V, and Ti. AES studies onthese compounds by Roth et al [7] revealed a nice correlation between the Auger lineshape andthe bonding state character of the Si atoms in the silicide, in that silicides with a low heat offormation (e.g., Pd2Si) exhibit the least covalent bonding and give rise to a pronounced Si LVVAuger line splitting. Experimental (81 and theoretical studies (9] into the electronic structure ofPd2Si showed that this sillcide has properties very similar to a metal with low lying, filled 4-dstates.

In the following paper, we report on a combined LEED, AES and UPS study on the reaction ofsilicon with a clean Pd(100) surface with the aim to follow the silicide formation as a function oftemperature.

Permanent address: Institut fur Physikallsche Chemie der UniversitatMunchen, D-8000 Munchen 2, West Germany

t Work supported in part by the Office of Naval Research.

22. Experiment

The measurements were carried out in a turbo-molecular pumped stainless steel uhv system (basepressure below 8 x l0 t 11torr), equipped with four-grid LEED optics (Varian), quadrupole massfilter (UTI). a differentially pumped He discharge resonance lamp and a double pass cylindricalmirror analyser. AES was performed using the LEED gun as a source of high-energy electrons(typical beam current 20pA, energy 2keV) and the LEED optics as a RF type analyzer. Themodulation voltage was as low as 2 V ptp in order to give sufficient resolution of the Si LVVAuger transition region. UV-photoemission data were taken using electron counting technique.the pass energies employed corresponded to energy resolutions of 100 meV (He 1) and 250 meV(He I1). The photon beam was incident at an angle of - 750 with respect to the CMA axis and ~

400 with respect to the surface normal.

Silicon deposition was achieved using an electronically regulated evaporation source [10]; the Sideposited from a tungsten boat was of very high purity (99.999%). A fairly large Si source -substrate distance (- 50 mm) was chosen in order to provide a uniform deposition over the entirearea of the Pd crystal surface. The Pd crystal was prepared and cleaned by cycles of argonsputtering and oxygen treatments as is described in more detail elsewhere [I I ]. The temperaturewas recorded via a chromel-alumel thermocouple spot-welded onto the rear of the crystal.

3. Results and Discussion

a) LEED

In contrast to what is observed with Ni, Al, or Be[l-31, submonolayer deposition of Si on Pd(100)at room temperature does not lead to long-range ordering, i.e., no LEED extra spots are obtainedfrom the Si covered Pd(100) surface. However, even small amounts of Si cause the backgroundintensity to increase considerably until after deposition of - 2-3 Si layers the integral order Pdbeams disappear completely. Careful warming of this covered surface up to 10 C, produces acomplex LEED pattern (fig.1) with numerous sharp spots even at the lowest attainable electronenergies, thus indicating a large unit mesh with a surprisingly pronounced long-range order.At this temperature, the Si atomsalready react with the palladiumsurface to form palladium sUicide.As this is an exothermic process(AHf, Pd2 SimO.4S eV [7)), the ex-cess of the heat of formation givesthe Si atoms sufficient mobilitythat they can form a weli-orderedphase with pronounced long-rangeperiodicity. It should be noted thatCardillo and Becker [41 made al-most exactly the same observationswith a P( 100) surface reactingwith Si. From their He scatteringdata they came up with a structuralmodel of the surface suicide whichactually exclusively consisted ofPt 2 Si. From our AES data (cf. sect.3b) we have good reasons to be-

Fig. 1. L ED pattern formed by PdSi on a Pd(100) surface;E - 83 eV.

3

1 Pdlieve that the related silicide, PdSi. is formed on the Pd surfacewhich causes the observed complex LEED pattern. (This pat- dN(E) 2tern is more complicated than the (6,'2 x 6.2) R 453 pattern dE 3

obtained by Cardillo et al. for the Pt(fl)silicide). 4

Raising the temperature beyond 300*C irreversibly destroys the 5extra LEED pattern without evidence of other intermediateordered structures, until finally the clean Pd(100) pattern isI 6restored. As the Si does not desorb (all our attempts to get Si 7TD spectra failed) only rapid bulk diffusion of Si can accountfor this observation. 8

I f9b) Auger electron spectroscopy

Fig. 2 shows a family of Auger spectra taken after successive Si 10depositions at 75*C without further annealing. At very low Sicoverages (below -. I ml) small features appear at 79 and 86.5eV. Increasing Si coverages cause mainly the latter peak togrow, and after about half a monolayer of Si a shoulder appears //1around 90 eV. Further deposition into the monolayer range andabove makes this shoulder grow almost exclusively (with the Si86.5 eV structure remaining constant) until finally the typical 12

Auger spectrum of silicon is obtained. Simultaneously, the PdMNN transitions (main peak at 331 eV) lose their intensitycontinuously until they disappear after roughly 5 Si layers havebeen deposited. No attempt was made to analyze the Augerdata in terms of absolute Si concentrations because of the lackof a second independent measure for the number of Si atoms inthe surface.

If the Pd surface covered with Si in this way is carefully subject-ed to annealing around 150*C. the one huge Si LVV-line at 92 Pd ieV shrinks considerably, and after a couple of minutes a new I

spectrum (shown in fig. 2 as a dotted line) is obtained: The 86.5 / x 5)eV transition now is completely separated from the 92 eV line I

that actually has shifted down to 91.2 eV and has a distinct Isatellite around 96 eV. In addition to that, two other peaks are ' tobserved at 75 and 80.6 eV. A comparison of this spectrum [ Iwith that reported by Roth et al (7] for bulk PdSi clearly re- Iveals that these spectra are practically identical. With the occur-rence of these new Auger features we also observed the complex

LEED pattern as well as the development of a new UPS level

(cf. sect. 3a and c. respectively). Higher annealing temperatures i(above 200*C) again lead to alterations in the silicon Augerspectrum: the peak at 96 eV disappears, and the 91.2 eV linedecreases progressively, whereas the 86.5 eV transition gets the

predominant feature. At this stage, the Pd 2Si LEED pattern hasdisappeared completely, and the photoemission spectra more and M go 90 100 eV

more resemble the clean Pd spectrum (cf. sect. 3). Fig. 2. Sequence of AES curvesdisplaying the Si LVV transition

Even higher annealing up to 700"C drives the Si completely into energy region. The Si depositionthe bulk. and the Auger spectrum of the clean Pd finally is increases in going from spectrum 1restored. to 12.

4c) UV photoemission

Photoemission spectra were taken at T - 75*C with He [ radiation (21.2 eV) as a function ofsilicon deposition. As the deposition proceeds, a preferential attenuation of the d-band featuresnear the Fermi edge occurs (the Pd d-states at 2,3 eV below Ef are little affected). For - 5deposited Si layers we no longer observe d-band emission at Ef; new (weak) features appear at 3.2eV and 7 eV below Ef that are characteristic of clean Si. The total work function change thataccompanies the Si deposition was determined to be approximately -600 meV. This fairly smallvalue indicates little charge transfer in forming a silicon-palladium bond, actually electronic chargeis withdrawn from the Si atoms to fill Pd d states. This observation is in agreement with calcula-tions by Lang and Williams [12] for Si atoms interacting with a jellium surface. In contrast toAES, annealing at 1500C does not produce dramatic changes in the photoemission c-rves. Asingle, fairly broad level appears near the clean Pd emission maximum, i.e., at 2.5 eV below Ef,

along with a single shoulder around 4.2 eV below Ef. Photoemission studies by Rubloff et al[8,9,13] on Pd2 Si produced by subsequent Pd deposition on a clean Si( 11) surface showed that ingoing from Si to Pd2 Si the above mentioned maximum in the EDC continuously shifts from 3.5 eVto - 2.7 eV below Ef as the Pd 2Si is formed progressively. We therefore interprete the observedUPS curve as being caused by Pd 2Si precipated on a background of clean Pd, although the vicinityof the Pd2 Si and Pd density of states maximum render a clear distinction quite difficult. We alsonote that we have occasionally observed UPS features near 6.5 eV and 10.2 eV below Ef, whicharise from the presence of oxygen in the silicon layer. This oxygen competitively reacts with theSi to form Si0 2 at 150'C, and the observed photoemission levels correspond very well with UPSresults on the valence levels of amorphous sin 2 reported by DiStefano and Eastnan [ 14]. Higherannealing temperatures cause the Pd2Si photoemission maximum at 2.5 eV to shift to 2.3 eVwhere it is more characteristic of the Pd d-band emission of the clean surface.

4. Conclusions

We have shown that the reaction of silicon with a clean Pd(100) surface at 150 0 C leads to theformation of palladium(II)silicide at the interface. There was no evidence from both AES andUPS measurements for any further formation of a silicide with different stoichiometry, i.e., with alower Si : Pd ratio. At elevated temperatures (T > 5000C) the bulk diffusion of silicon causesPd2Si compound decomposition and the rapid depletion of Si on the palladium surface.

5. References

[1) F. Jona, J. AppL. Phys. 42. 2557 (1971)(21 F. Jona, J. Appi. Phys. 44, 351 (1973)[31 F. Jona, J. Appl. Phys. 44, 4240 (1973)[41 M.J. Cardillo and G.E. Becker, to be published[51 K. Christmann, unpublished result[6] K.N. Tu and J.W. Mayer, "Silicide Formation," Thin Film Reactions and Diffusion, ed. byJ.M. Poate, K.N. Tu and J.W. Mayer (Wiley New York, 1978)[71 J.A. Roth and C.R. Croweil, J. Vac. Sci. Technol. 15, 1317 (1978) [81.L. Freeouf, G.W.Rubloff, P.S. Ho and T.S. Kuan, Phys.Rev.Lett. 43, 1836 (1979)[91 P.S. Ho, G.W. Rubloff, .E. Lewis, V.L. Moruzzi and A.R. Williams, Phys.Rev. B, submitted[101 K. Christmann and G. Ert. Thin Solid Films 28, 2 (1975)[ I I R.J. Behm. K. Christmann, G. Ertl and M.A. van Hove, J.Chem.Phys, in press(121 N.D. Lang and A.R. Williams, Phys.Rev. B 18, 616 (1978)(131 P.S. Ho, T.Y. Yan, J.E. Lewis and G.W. Rubloff, to be published[141 T.H. DiStefano and D.E. Eastman, Phys.Rev.Lett. 27, 1560 (1971)

* sP172-3/A1 472:GAN:716:ddc*78u472-608

TECHNICAL REPORT DISTRIBUTION LIST, GEN

No. No.Copies Copies

Office of Naval Research U.S. Army Research OfficeAttn: Code 472 Attn: CRD-AA-IP800 North Quincy Street P.O. Box 1211Arlin-ton, Virginia 22217 2 Research Triangle Park, N.C. 27709 1

ONR Branch Office Naval Ocean Systems CenterAttn: Dr. George Sandoz Attn: Mr. Joe McCartney536 S. Clark Street San Diego, California 92152 1Chicago, Illinois 60605 1

Naval Weapons Center[

ON? Area Office Attn: Dr. A. B. Amster,Attn: Scientific Dept. Chemistry Division715 Broadway China Lake, California 93555 1New York, New York 10003 1

Naval Civil Engineering LaboratoryONR Western Regional Office Attn: Dr. R. W. Drisko1030 East Green Street Port Hueneme, California 93401 1Pasadena, California 91106 1

Department of Physics & ChemistryONR Eastern/Central Regional Office Naval Postgraduate SchoolAttn: Dr. L. H. Peebles Monterey, California 93940 1Building 114, Section D666 Sumner Street Dr. A. L. SlafkoskyBoston, fassachusetts 02210 1 Scientific Advisor

Commandant of the Marine CorpsDirector, Naval Research Laboratory (Code RD-i)Attn: Code 6100 Washington, D.C. 20380 1Washington, D.C. 20390 1

Office of Naval ResearchThe Assistant Secretary Attn: Dr. Richard S. !iller

of the Navy (RE&S) 800 N. Quincy StreetDepartment of the Navy Arlington, Virginia 22217 1Room 4E736, PentagonWashington, D.C. 20350 1 Naval Ship Research and Development

CenterCommander, Naval Air Systems Command Attn: Dr. G. Bosmajian, AppliedAttn: Code 310C (H. Rosenwasser) Chemistry DivisionDepartment of the Navy Annapolis, Maryland 21401Washington, D.C. 20360

Naval Ocean Systems CenterDefense Technical Information Center Attn: Dr. S. Yamamoto, MarineBuilding 5, Cameron Station Sciences DivisionAlexandria, Virginia 22314 12 San Diego, California 91232 1

Dr. Fred Saalfeld Mr. John BoyleChemistry Division, Code 6100 Materials BranchNaval Research Laboratory Naval Ship Engineering CenterWashington, D.C. 20375 1 Philadelphia, Pennsylvania 19112 1

- .-- '--

472:GAN:71b:ddc

S?472-3/A3 78u4 72-6 08

TEChNICAL REPORT DISTRIBUTION LIST, GEN

No.Copies

Dr. Rudolph J. Marcus

Office of Naval Research

Scientific Liaison Group

American Embassy

APO San Francisco 96503 ,

Mr. James KelleyDTNSP.DC Code 2803 20

Annapolis, Llaryland 21402

1tP

SP4.72-3/A17 472:GAN:716:ddc78u472-608

TECHNICAL REPORT DISTRIBUTION LIST, 056

No. No.Copies Copies

Dr. D. A. Vroom Dr. C. P. FlynnIRT Department of PhysicsP.O. Box 80817 University of IllinoisSan Diego, California 92138 1 Urbana, Illinois 61801

Dr. G. A. Somorjai Dr. W. KohnDepartment of Chemistry Department of PhysicsUniversity of California University of CaliforniaBerkeley, California 94720 1 (San Diego)

LaJolla, California 92037Dr. L. N. JarvisSurface Chemistry Division Dr. R. L. Park4555 Overlook Avenue, S.W. Director, Center ofWashington, D.C. 20375 1 Materials Research

University of MarylandDr. J. B. Hudson College Park, Maryland 20742Materials DivisionRensselaer Polytechnic Institute Dr. W. T. PeriaTroy, New York 12181 1 Electrical Engineering

DepartmentDr. John T. Yates University of MinnesotaSurface Chemistry Section Minneapolis, Minnesota 55455National Bureau of StandardsDepartment of Commerce Dr. Narkis TzoarWashington, D.C. 20234 . City University of New York

Convent Avenue at 138th Street

Dr. Theodore E. Madey New York, New York 10031Surface Chemistry SectionDepartment of Commerce Dr. Chia-wei WooNational Bureau of Standards Department of PhysicsWashington, D.C. 20234 1 Northwestern University

Evanston, Illinois 60201

Dr. J. M. WhiteDepartment of Chemistry Dr. D. C. MattisUniversity of Texas Polytechnic Institute ofAustin, Texas 78712 1 New York

333 Jay StreetDr. Keith 11. Johnson Brooklyn, New York 11201Department of Metallurgy and %aterials

Science Dr. Robert M. HexterMassachusetts Institute of Technology Department of ChemistryCambridge, Massachusetts 02139 1 University of Minnesota

Minneapolis, Minnesota 55455

J . E. DemuthSIBM >8 n Dr. R. P. Van DuyneThomas J. Wats earch Center Chemistry DepartmentP.O. Bo Northwestern University

own Heights, New York 1 Evanston, Illinois 60201

SP472-3/A19 472 :GA:: 716: d

78u472-6C8

TECHNICAL REPORT DISTRIBUTION LIST, 056

No. No.Copies Conies

Dr. M. G. Lagally Dr. J. OsteryoungDepartment of Metallurgical Chemistry Department

and Mining Engineering SUNY, BuffaloUniversity of Wisconsin Buffalo, New York 14214 1Madison, Wisconsin 53706 1

Dr. G. RubloffDr. Robert Gomer I.B.M.Department of Chemistry Thomas J. Watson Research CenterJames Franck Institute P. 0. Box 2185640 Ellis Avenue Yorktown Heights, New York 10598Chicago, Illinois 60637 1

Dr. J. A. GardnerDr. R. G. Wallis Department of PhysicsDepartment of Physics Oregon State University

University of California, Irvine Corvallis, Oregon 97331Irvine, California 92664 1

Dr. G. D. SteinDr. D. Ramaker Mechanical Engineering DepartmentChemistry Department Northwestern UniversityGeorge Washington University Evanston, Illinois 60201Washington, D.C. 20052 1

Dr. K. G. SpearsDr. P. Hansma Chemistry DepartmentChemistry Department Northwestern UniversityUniversity of California, Evanston, Illinois 60201

Santa BarbaraSanta Barbara, California 93106 1 Dr. R. W. Plummer

University of PennsylvaniaDr. P. Hendra Department of PhysicsChemistry Department Philadelphia, Pennsylvania 19104Southhampton UniversityEngland S09JNH 1 Dr. E. Yeager

Department of ChemistryProfessor P. Skell Case Western Reserve UniversityChemistry Department Cleveland, Ohio 41106 2Pennsylvania State UniversityUniversity Park, Pennsylvania 16802 1 Professor George H. Morrison

Cornell UniversityDr. J. C. Hemminger Department of ChemistryChemistry Department Ithaca, New York 14853 1University of California, IrvineIrvine, California 92717 1 Professor N. Winograd

Pennsylvania State UniversityDr. Martin Fleischmann Chemistry DepartmentDepartment of Chemistry University Park, Pennsylvania 16802 1Southampton UniversitySouthampton 509 5NH Professor Thomas F. GeorgeHampshire, England The University of Rochester

Chemistry DepartmentRochester, New York 14627 1

e!---------