Auger Electron Spectroscopy (AES)

1 Auger Electron Spectroscopy (AES) Prof. Paul K. ChuAuger Electron SpectroscopyIntroductionPrinciplesInstru mentationQualitative analysisQuantitative analysis Depth profilingMapping ExamplesTheAugerEffectisnamedafteritsdis coverer,PierreAuger, AugerTodayAugerelectronspectroscopyisapo werfulsurfaceanalyticaltooltoprobesurfac es,thinfilms, ( ),goodspatialsurfaceresolution(asgoodas1 0nm),periodictablecoverage(excepthydroge nandhelium),andreasonablesensitivity(100 ppmformostelements).Signals of Auger electronsDistribution of Energies of Emitted ElectronsAuger electronsElectron Beam -Sample InteractionPrimary Electron BeamAuger Electrons4-50 >Atomic No.

2 3 Characteristic X-rays> Atomic No. 4 Volume of Primary Excitation<1 -3 mmSample SurfaceBackscattered ElectronsSecondary ElectronsAuger , ,theAugerelectron, for Auger TransitionsKL1L2 Transition labelThe three symbols in the transition label correspond to the three energy levels involved in the of Auger ProcessesAuger Electron SpectrumElectron signal vselectron energyThe Auger peaks are obscure even using an expanded vertical Energy Spectrum of Pd metalDirect spectrumDifferential spectrum-enhances AES features/ removes backgroundPlottingd[E N(E)]/dE, [E N(E)] ,dN(E)/dE, ,LMM, , ,thekineticenergyoftheAugerelectronisequ altoEA-EB-EC,whereEA,EB, )1()(21)1()(21)( zEzEzEzEzEECCBBAABCFor approximation, the following equation is used to calculate the kinetic energy of Auger electrons (z: atomic number of the atom).

3 InitialFinalABCEEE Forexactcalculation,othereffectsshouldbe takenintoaccount, , of Kinetic Energy Auger PeaksAuger spectra in the differential distribution characteristic of the lightest elements (The principal peak is KL2,3L2,3and the relative intensities are not plotted to scale)b) Auger PeaksDifferentiated Auger spectra of chromium, manganese, and ironManganeseChemical Electron Intensity X-ray , , ,allelements(exceptHandHe) relative intensities of the Auger electrons also depend on the primary excitation relative intensities can be used for quantitative analysis10 keV Electron beam3 keV Electron beamKLLLMMMNNS urface Sensitivity of Auger Electron ( ), electronsBackgroundMNNThe term inelastic mean free path is usually used to describe the surface sensitivity of Auger electronsIoIoIsIsded -= edxedxexx Mean distance that an Electron travels without energy lossIo: number of electrons originating at a depth dIs.

4 Number of electrons that can escape the surface without energy loss95%ofthemeasuredAugerelectronsorigin atefromthedepthregionof3 Inelastic Mean Free Path Surface Sensitivity of AESAES spectraofNisurface:(a)afterArsputter-cle aning;(b) SourceElectron energy analyzerIon GunVacuum systemSample stageElectronic controlsComputerSoftwareElectron detectorVacuum SystemSample HandlingSample introductionSample stage with x, y, z, rotation, and ~ (LaB6) ,currentdensitiesof~100 >107 , (103to106A/cm2) of Field Emission (-V) Electron SpectrometerElectron Detector Electron ( Electron ,ion,highenergyneutral,orhighen ergyphoton)

5 Strikesthefirstdynode, Plate Electron Multiplier ArraysTheseplatesconsistoflargearraysofs mallchannelelectronmultipliersField Emission GunMicrochannel Plate DetectorCylindrical Mirror AnalyzerSampleIon beam for + :(1)surfacecleaning,(2)depthprofilingIon GunParameters CalibrationEnergyscaleiscalibratedusinga flat, (transmissionfunctionanddetectorefficien cy) spectrum of a unknown samplePrimary Electron beam energy: 3keVNiFeCrNi, Fe and Cr are preliminarily identifiedStandard spectrumStandard spectrumStandard spectrumFeCrFeNiN(E) vsEPeak-to-peak heightCommonly use peak to peak height of differentiated Auger peak.

6 For high resolution, use peak area of the original Auger peak, but this needs background areaNeed background subtractionQuantitative Analysisd[E N(E)]/dE vsENote: Sensitivities for these two methods are differentRDTFrNIIiiiPi cos)1(For a homogeneous sample, the measured Auger intensity is given byIi: Auger intensity for the ABC transition of element iIP: Primary Electron beam currentNi: Number of atoms of element i per unit volume i: Ionization cross section for A level of element i i: Auger transition probability for the ABC transition of element ir: Secondary ionization for the A level of element i by scattered electrons : Inelastic mean free path : Anger between the direction of Auger Electron and the sample normalF: Factor for analyzer solid angle of acceptanceT: Analyzer transmission functionD: Detector efficiencyR.

7 Surface roughness factorAuger Electron IntensitydDetectorEmpirical ApproachUse of relative sensitivity factors (Si)For two pure elements222111 SNISNI The relative sensitivity factors were measured for all elements under a certain excitation Electron beam energy NtoiiiaaNtoiiaaSISINNX11//Then, the atomic concentration of element a on a sample with N elements can be determined as Percentage atomic concentration = Xa 100%Theempiricalmethoddoesnotincludethem atrixeffectsofthesample,whichincludesthe inelasticmeanfreepath( ),thebackscatteringfactor(r), ,anerrorof15% ,asamplereportedwith80%Feand20% 12%.

8 Ifstandardsampleswiththesamematrixisused forthedeterminationofthesensitivityfacto rs,theerrorcanbeassmallas1%.FeCrFeNiPeak -to-peak height: SiCr peak at 529eV: peak at 703eV: peak at 848eV: ;%71%:%21% NiFeSimilarlyxCrDepth ProfilingToanalyzesamplesindepth, ,stopping,measuringrelevantportionsofthe Augerspectrum, + Auger signalSputtering timeAuger signal can be Auger peak area or peak-to-peak sputtering removes the materials systematicallyAuger analysis is performed on the newly exposed crater bottom and independent of the sputtering processAuger Depth , signalSputtering timeCalculationCalibration of Depth ScalePtTiInPSampleFor each element, an Auger peak is selected at each :MNN at 1697 eVTi:LMM at 418 eVC:KLL at 272 eVIn:MNN at 404 eVO.

9 LMM at 120 eVContamination at Semiconductor / Metallization Interface520 SiO2Si66 Depth Resolution Depends on Ion beam energy and incident angle Surface roughness of sample Type of materialsAr+Ion mixingIon MixingPreferential SputteringAr+Surface is rich in8000 AlSiSEM ImagesSputtering timePoor depth resolutionDifference in Chemical composition Crystal orientationSample RotationNo sample rotationWith sample rotationDepth Profile of a Ta/Si Multilayer SampleDepth resolution usually degrades with depthImprove depth resolution: Sample rotation during Ar+sputtering Low Ar+energy (less than 1 keV) Low incident angle01020304050607080901000 Sputter Time (min)IntensityTi in TiNTi in TiNNNTi in Metallic TiO in SiO2Si in SiO2O in TiNTi in Oxidized TiN0102030405060708090100010203040506070 8090100 Sputter Time (min)Atomic Concentration (%)Ti in TiNTi in TiNNNTi in Metallic TiO in SiO2Si in SiO2O in TiNTi in Oxidized TiNDepth Profiling of TiN/Ti/TiN on SiO2 TwoTiN(250 )/Ti(375 )/TiN(600 )

10 Auger MicroscopyAESA uger Electron SpectroscopySAMS canning Auger MicroscopySampleElectron focusing and scanningIon BeamThesameinstrumentcangiveSEMimages, Carbon100 kX SEM160 kX SEMAu islands on C160 kX Auger MapsComposite Auger imageSecondary Electron imagesScanning Auger MicroscopyImages and Auger elemental maps from a contact pad on a semiconductor device: (a)Adsorbed current image;(b)Secondary Electron image;(c)oxygen map (O KLL, 512 eV);(d)Silicon oxide map (Si LMM, 1625 eV, oxide peak);(e)Silicon map (Si LMM, 1625 eV, non-oxide peak);(f)aluminum map (Al LMM, 1400 eV).The Auger maps reveal that the oxide layer in the middle of the bond pad has been broken exposing a small area of aluminum due to bonding currentAuger Si map (non-oxide) Auger O mapSEM Secondary electronsAuger Si map (oxide) Auger Al mapAnalysis of bond padTwo chips were supplied for analysis, a chip from a contaminated wafer and one from a normal or good waferSecondary Electron Images (Box defines analyzed area for AES)Bond pad from normal chipBond pad from contaminated chip500100015002000 Kinetic Energy (eV)dExN(E)/dECOFAl500100015002000 Kinetic Energy (eV)dExN(E)