420 Applicationsof SurfaceScienceII / 12 (1982)420—424North-HollandPublishingCompany

XPSSTUDIES OF THE PHOTODECOMPOSITION OF AgCI

J. SHARMA ‘~, P. DiBONA ** andD.A. WIEGAND

EnergeticMaterials Division, LCSWL,Army ArmanentR&D Command,Dover, NewJersey07801,

USA

Received8 June 1981

The decompositionof AgCI dueto ultravioletandX-ray irradiationwasinvestigatedby X-rayPhotoelectronSpectroscopy(XPS). Sampleswerepreparedfrom freshly madeAgCI powdereitherdirectly from the powderor by melting AgC1 onto a platinum substrate,the latter givingsignificantlyhigherXPS intensitiesthan theformer.The AgCI wasexposedonly to weak red lightprior to irradiation.Theresultsafter irradiation indicateda broadenedCl 2p doublet which wasresolvedintoandunbroadenedandunshiftedCl 2pdoublet anda Cl 2p doubletshiftedto a higherbindingenergy(2.6 eV). Similar resultswerefoundfor theCl 2s line. Theresultsthusindicatethepresenceof a lesselectro-negativeform of Cl on thesurfaceof AgCl after irradiation.The Ag lineswere neither broadenednor shifted within the precision of the measurements(0.2 eV). Withincreasingtime of irradiation the intensitiesof the Ag lines decreased.Possibleforms of thisphoto-inducedCl and therelationshipto thephotographicprocesswill bediscussed.

1. Introduction

Thechemicalshifts of the electroniclevels as observedin X-ray Photoelec-tron Spectroscopy(XPS) havebeenutilized to study photodecompositionofsilver chloride, which is an importantphotoesensitivematerial. Direct experi-mental evidencefor the generationof neutralchlorineas a step in the earlystageof the photolysisprocesshasbeenobserved.Thechlorine linesshowedashift towardshigherbinding energy.In the samplingdepthof the technique,which is about 15—20A for silver chloride, an appreciableamount of neutralchlorinehasbeenfound to build up while no new linesof silver wereobserved.

2. Experimental

The presentstudyhas beencarriedout in a Varian lEE-is instrument,inwhich about1.2 cm2 of the areaof the specimenprovidesa largecountingrate

* Presentaddress:Naval SurfaceWeaponsCenter,White Oak. Silver Spring.Maryland 20910,

USA.~ Presentaddress:Honeywell Inc., Hopkins, Minnesota55343, USA.

0378-5963/82/0000—0000/$02.75© 1982 North-Holland

J. Sharmaet a!. / XPSstudiesof photodecompositionof AgCI 421

e.g. 90,000c/s at 9 kV, 100 mA X-ray tubevoltage and current.The spectrawerecalibratedwith the Au 4f7/2 line assumedto be at 83.8 eV. Two methodsof sample preparationwere used. In some cases the pure silver chloride,exposedonly to red light, was mountedas powderon the cylindrical sampleholderwith the help of double-stickScotch tape.Also sampleswerepreparedby rolling a hollow platinumcylinder in a pool of silver chloridemeltedin aplatinumdish. Thelattermethodgavea cleartransparentfilm on the platinumsubstrate.The melt samplegavea narrowerline width of 1 eV correspondingto the instrumentalline width. In additionthesmoothsurfaceresultedin largercount rates. In most of the work the samplespreparedby the melt methodwereused.Bothkinds of samplegaveidentical results.

During most of the studies, ultraviolet irradiation of the specimenwascarriedout at roomtemperaturein thereactionor samplepreparationchamberof the instrument,usingan unfiltered low pressuremercuryvaporlamp, withthe samplein l0~2_l0_3mm of vacuum. In somecasesirradiationwasdonein air to avoid anyheatingof the specimendueto irradiationin vacuum.Thetwo modes of irradiation gave identical results. The vacuum in the mainchamberof the instrumentwas 10—6 mm or mercury.

The X-rays of the XPS instrumentalso causeda radiation effect on thechlorinespectraas the UV, but it wassmallerby oneorderof magnitudefor agiven exposuretime. ThereforeX-ray damagedid not interferewith the UVphotolysisreported.

3. Resultsanddiscussion

The Cl2P

1/2,3/2spectrumof unirradiatedsilver chlorideis shownin fig. la.On irradiation with UV for half an hour, the spectrumbroadenedtowardsthehigher binding energy side. With prolongedirradiation of 2 h, anotherset of

1/2,3/2 lines developedat 2.5 eV towardsthehigherbindingenergy side.Thespectrumof heavily irradiatedsilver chlorideis shownin fig. lb. Thedifferencespectrumbetweenthat of irradiatedandunirradiatedspectrumis shownin fig.2b; this figure indicatesclearly the developmentof another

2P

1/2,3/2 doublet.Broadeningtowardshigher binding energy side is also evidencedby the 3pspectrumin the valenceregion asshown in fig. 3. The2s spectrumof chlorineshoweda similar shift of 2.5 eV towardshigherbinding energyside.

The 3d and all other levels of silver accessiblewith Mg K radiationin theXPS instrumentwere closely followed but none of them showedany shift orbroadeningwith UV irradiation. Thesilver lines in silver chloride,during thepresentinvestigation,matchedwell with those of metal foil.

In the earlystageof irradiation it was observedthat the total areaof thenormal andshiftedCl ~ 1/2,3/2 levels remainedconstant.While the linesin thenew position increasedin intensity, the intensity of normal lines decreased.During later stageof continued irradiation, the lines in the new position

422 J. Sharmaeta!. / XPSstudiesofphotodecompositionofAgC!

Ag CL (a) UNIRRADIATED A9CLCL2p1/2,3/2 CL2p 1/2.3/2

I I (b) (a)DIFFERENCE CURVE UNIRRADIATED

~~ ( IRRADIATED F

2~l2 ~ ~BINDING ENERGY (eV) BINDING ENERGY (eVJ

Fig. I. The Cl 2p 1/2.3/2spectraof (a) unirradiatedand(b) irradiatedsilverchloridewith unfilteredlight from a low pressuremercuryvapor lamp.

Fig. 2. The Cl2P1/2.3/2 spectraof unirradiatedsilver chloride (a) and the differencecurve (b)

betweenthe spectraof irradiatedand unirradiatedsilver chloride, showing a shift of —~2.5eV,indicating thegenerationof neutral chlorine.

AgCLCL 3p 1/2, 3/2, /VALENCE REGION

UNIRRADIATED

~j (b),‘ IRRADIATED

~ /1

8 6 4 2BINDING ENERGY 1eV)

Fig. 3. The Cl 3p spectraof unirradiated(a) andirradiated(b) silver chlorideshowingbroadeningdue to thegenerationof neutralchlorine.

J. Sharmaet aL / XPSstudiesof photodecompositionof AgC! 423

reached a limiting value, and yet the lines in normal position kept ondiminishingdown to less thanone-tenthof their original value.

The intensityof silver lines monotonicallydecreasedduring the irradiationexperimentand it could be reducedto one-tenthof its original value. Sinceboththe chlorine(total) andsilver signalsdroppedin intensityas a function ofirradiation,it was interestingto watch the ratio of chlorineandsilver lines. Inthe initial stagethe ratio remainedconstant,but after prolongedexposuretheratio beganto drop dueto largerlossof chlorine.Theratio of Cl 2p andAg 3dlinesin unexposedsilver chloridewasaspredictedby cross-section.On merelykeepingthe specimenin the instrumentwithout irradiation, the intensitiesofsilver andchlorinelinesdid not decrease.

During the long runs of the experimentsthe is lines of contaminationcarbonand oxygen showedprogressiveincreaseswithout the developmentofany structureor shift. Thisindicatesthat the shiftsobservedin chlorinecannotbe due to any chargingof the sample by X-ray exposure.The increaseofcarbonandoxygen contaminationwasnot enoughto explain the decreaseofchlorineandsilver signalsby masking.

The shift of the chlorinelevels indicatesthat some chloride chlorine goesinto a higheroxidative state.By comparingthe shift of the 2p lineswith thecalculatedvaluesof Siegbahn[1], it can be concludedthat neutralchlorineisbeing generated.Thus the XPS study hasgiven direct evidenceof the genera-tion of neutralchlorinefrom Cl — of silver chloridephotolysis.Theresultsalsoshow that in the sampling depth of the specimenneutral chlorine lines ofintensity comparableto that of Cl — are observed.The fact that in the earlystagesthe signal due to total chlorine remainsconstantwhile the neutralchlorineincreasesat the cost of ionic chlorine,shows that during thesestagesno escapeof chlorine from the sampling depth takes place. The presentexperimentsdo not tell whether the neutralchlorineis in atomicor molecularform. The aboveresultsare very significant for the theory of photographicaction.They show thatchlorinehaslargestickingcoefficientin silver chloride,and doesnot easily diffuse out at leastin the early stagesof exposure.Thepresenceof the neutralchlorineandthe trappedelectronsnecessaryfor chargeneutrality,may be the essentialingredientsof “latent image.” The resultsofprolongedexposureindicatea lossof chlorineas observedby Luckey [2]. Theexistenceof neutralchlorine,probablyin interstitial positions,also meansthatit has the possibility of healingradiationdamageof small amount.That mayexplainwhy photographicfilms do not show significant effectsof cosmicraysor weak ionizing radiation during shelf life. Probablythe presenceof neutralchlorinehas acuring effect via “replacementmechanism”[3].

It is difficult to explain lack of observableshift of the silver lines. It iswell-known that silver ions changeto silver atomsduring photolysis.Eitheronehasto assumethatwhensilver atomsaredeposited,e.g.,at dislocations[4]and form colloidal particles, the electronic levels do not changefrom theirinitial state in silver chloride, at least within detection capability of the

424 J. Sharmaet a!. / XPSstudiesofphotodecompositionofAgCI

instrumentused (0.2 eV) or that the silver atoms diffuse out of the XPSsamplingdepth.Onelimitation of the presentwork hasbeenthat the degreeofphotolysis is not quantitivelycorrelatedto the exposurelevels.

After a particularamountof neutralchlorineis reached,further conversionof the chlorineion mustbe associatedwith the escapeof chlorineout of thesamplingdepthof the XPS instrument.Someof it could be escapinginto thevacuumsystemof the instrument[2], while the restof it might be goingdeeperinto the sample.The monotonicdecreaseof the silver signal can parly beexplainedby assumingthat silver is forming colloidal particlesleadingpartlyto a decreasein the surfacearea.Grain orientationmay alsobe takingplace.Itis possiblethat the silver as well is migratingout of the samplingdepthof theXPS as notedabove.Theincreasein the carbonandoxygenis understandablebecausethe vacuumwasonly 10—6 mm of mercury,which is a poor vacuumfor surfacework of this type. As pointed out before, the build-up of thecontaminationwas not enough to explain the decreasein the silver andchlorinesignalsby masking.Onepossiblecauseof the decreasein chlorineandsilver signalsis a changein surfacemorphology.

4. Summaryand conclusions

The presentwork shows that XPS can be used to follow photolysis inphotosensitivematerials, yielding useful information. The generationof freechlorine and its existencein the lattice may play a significant role in thephotographicprocess.

References

[I] K. Siegbahn,C. Nordling,A. Fahlman,R. Nordberg,K. Hamrin,J. Hedman,G. Johansson,T.Bergmark,SE. Karlsson,I. Lindgren and B. Lindberg, ESCA: Atomic, Molecular and SolidState StructureStudiedby Meansof ElectronSpectroscopy(Almquist andWicksells, Stock-holm, 1967)p. 82.

[2] G.W. Luckey,J. Chem. Phys.23 (1955) 882.[3] RH. Herz,The PhotographicAction of Ionizing Radiations(Wiley—Interscience,New York,

1969) p. 70.[4] J.W. Mitchell, Rept. Progr. Phys.20 (1957) 433.