Professional Documents
Culture Documents
Bruker Tracerand Artax XRF Raw Spectrum Analysis User Guide Draft
Bruker Tracerand Artax XRF Raw Spectrum Analysis User Guide Draft
TABLEOFCONTENTS
TABLEOFCONTENTS ABSTRACT XRFTHEORY INSTRUMENTATION 1 3 4 6
6 8 8 9 9
9 11
11 12 ERROR!BOOKMARKNOTDEFINED. 15 16
17
17 18 19
THINFILMANALYSIS(BACKGROUNDCONTRIBUTION)
PHENOMENAINTHESAMPLE
ERROR!BOOKMARKNOTDEFINED.
20
28 29 30
41
43 48 49 50 51
ABSTRACT
WhileperformingXRFspectroscopy,threemainfactorscontributetotheanalyticalspectrum:interactions inthedetector,interferencewiththeinstrumentation,andphenomenainthesample.ThisuserguidetoEDXRF providesabasicoutliningofthephysicsinvolvedinXRFspectroscopy,anoverviewofthemaincomponentsofthe BrukerTRACeRandARTAXunits,aswellasadelineation,explanation,andresolutionoftheseveralphenomena includedinperformingXRFspectroscopy.Eachsectionincludesatextualexplanationofwhyitoccursandhowit affectsaspectrum,aswellasexamplespectrathatclearlyidentifythephenomena.Severalfieldapplicationsand examplesareprovided,aswellasanappendixwithadditionalinformation
XRFTHEORY
GeneralConceptBehindXRayFluorescenceSpectroscopy Every element has a characteristic electron structure. When inner shell electrons are ejected from an atom,electronsfromshellswithlessbindingenergyfilltheholesandmayreleasexrayradiationequivalenttothe difference in energy between the level the electrons came from to that which they went. The x ray radiation releasedduringthesetransitionsischaracteristictotheelementandhasaspecificenergy(2eV)dependingon thetransitionmadewithintheatom.Bybombardingasamplewithradiationthatexceedsthebindingenergyof theelectronsintheatomsofwhichthematerialiscomposedofanddetectingtheenergyandnumberofresultant characteristic x rays emitted from each element, it is possible to determine the composition and proportional concentrationsofthoseelements. Two common methods of Xray spectroscopy exist: Wavelength Dispersive XRF (WDXRF) and Energy Dispersive XRF(EDXRF).Themaindifferencebetweenthetwomethodsishowtheemittedxraysaremeasured;WDXRF usesananalyzingcrystaltodiffractthedifferentxraywavelengthsanddetectorsareplacedatthevariousangles tomeasurethenumberxraysdiffractedateachangle.Asingledetectormaybeusedtomeasureallthevarious energies if one moves the detector to cover all the angles, because each energy comes out of the crystal at a differentangle. Energydispersivexrayfluorescence(EDXRF)usesadetectorthatcollectsxraysofallenergiesandsortsouteach xrayenergybytheamountofelectronseachxrayknocksfreeinthedetectorlattice,typicallysilicon.Thenumber ofelectronsknockedfreedependsontheincomingxrayenergyandtheparticularinteractionthatthatxrayhas withthemateriallattice.Toaccuratelydeterminethexrayenergyalltheelectronsfromeacheventthatoccursin thedetectormustallbecollectedandconvertedultimatelytoadigitalsignal.Thusthedetectormeasuresonex rayatatime. BremsstrahlungRadiation In most xrf systems the beam of x rays incident on the sample are produced with a vacuum tube and createdbybombardingatarget(suchasRh,W,Cu,orMo)withhighlyacceleratedelectrons.ShowninFigure1,as theelectronspenetratethetargetatoms,theymayhavetheirdirectionchangedastheypassnearthenucleusof the target atoms causing a sudden deceleration and loss of kinetic energy. In this loss of kinetic energy the electron may emit an x ray with energy relatedtotheamountofenergylost.Asa Atomofthe Outboundelectron targetmaterial (deceleratedanddiverted) resultabroadspectrumofxrayenergies, known as a Bremsstrahlung continuum, is
Fastinbound electron
Figure1:DiagramoftheBremsstrahlungeffect
Figure2:Diagramofpossibleexcitationroutes
emittedfromthexraytubetarget.Thiscontinuumcanbeadjustedbytubehighvoltagesettings,beamfiltering andsecondarytargetstoallowonetofocusondetectionofspecificelementswithinthesample.Thiscapitalizes onthedifferentabsorptionedgesofeachelement.Theacceleratedelectronsalsocausethetargettofluoresce. Thesetargetcharacteristicxraysarealsoincidentonthesample,andmustbeconsideredduringdataanalysis ThePhotoelectricEffectandInnerShellIonization When an x ray interacts with an atom several different reactions can occur depending on the x rays energy.Lowenergyphotons,suchasalightwaves,canexciteandejectoutershellelectronsfromtheatomand exciteinnershellelectronstohigherenergylevels.Xrayscanejectinnershellelectronsfromtheatom,creatinga vacancy in the inner shell and put the atom in an unstable condition, providing the that incoming x ray has an energywhichexceedsthebindingenergyoftheelectronitinteractswith.Whenavacancyiscreated,theatom quickly relaxes (less than 107 s) by transitioning an electron from a higher shell to the vacancy. During this transitionanxraywiththeenergyequaltotheenergydifferenceofthetransitionmaybereleased.Thisxrayis knownasacharacteristicxrayandisspecifictothetransitionandelementinwhichitoccurs.Figure2illustrates the different interactions that can occur when x rays interact with the bound electrons of an atom. Once the characteristicxraysarecreated,theymayescapetheatomandthematrixmaterialatrandomanglesandavery smallfractionenterthedetector.Ifgeometry,densityandotherfactorsareknown,thenumberofxraysentering
Table1:Possibleelectronlocationsinanatom
the detector can be related to the type and numberofatomspresentinthesample. CharacteristicLines Theenergyofthexrayreleasedbytherelaxation of an ionized atom is dependent on the element from which it came, the location of the vacancy, and the electron that fills the vacancy. Electrons existinquantizedenergylevels;theyarenotfree to roam anywhere around the nucleus of the atom. A given electron is located in a given electronshellusingthequantumnumbersn,l,m, s,wherenindicatestheshell,lindicatesthesub shell,mlindicatestheenergyshiftwithinthesub
Table2:ExampleofpossibletransitionnotationsforaBariumatom.
shell,andmsindicatesthespinoftheelectron.Table1listssomeofthepossibilitiesforelectronlocationsinan atom. There are a couple ways of describing an x ray of a certaintransition. It canbe written inLine notation, Siegbahnnotation,ordescribedbythecharacteristicenergyorwavelengthconnectedthatxray.Thesenotations areshownbelowinTable2withanexampleofBariumelectrontransitions. IntheSiegbahnnotation,theGreek subscriptdenotestheprobabilityofthetransition(intensity),proceedingfromthemosttoleast(,,,etc.).
INSTRUMENTATION
TheTRACeRandARTAXarebothEDXRFunitswithSiliconbaseddetectors.TheTRACeRisahandheld unitcommerciallyofferedbyBrukerAXSandprovidesforquickandeasyqualitativeanalysisandchemistriesfor elements as low as Mg. The Tracer handheld XRF analyzer provides spectral analysis through PXRF analytical software. The instruments high sensitivity allows the user to identify the elements in a sample matrix, with concentrations as low as ppm. The PXRF software program provides qualitative and quantitative analysis, in addition to the voltage andcurrentcontrol of the Xraytube, which makes possiblea wider rangeof elemental analysis. The ARTAX is the first commercially available, portable microXRF spectrometer designed to meet the requirements for a spectroscopic analysis of unique and valuable objects on site, i.e. in archeometry and art history.ThesystemperformsasimultaneousmultielementanalysisintheelementrangefromNa(11)toU(92) and reaches a spatial resolution of down to 30 m. Both instrumentsallow one to utilize filters and secondary targettoadjusttheincidentxraybeaminbothenergydistributionandintensity.
EDXRFEQUIPMENT
TheARTAXandTRACeRsystemshaveseveralseparatecomponentsthatallservetheirownfunctionin the process of recording Xray fluorescence. The main components in terms of functionality are the Xray tube system, collimators, filters, detector and signal processing hardware and software. The ARTAX and the TRACeR Turbo are very similar in functionality. Both units employ energy dispersive technology and a Silicon based detector.Beingahandheldunit,theTRACeRisbatteryoperated,moreconvenient,buthasabeamspotsizeof3 by4mm(muchlargerthantheArtax).TheARTAXisportable;however,itisnotgenerallyusedinthefieldlikethe battery operated handheld TRACeR unit. Both units can use a variety of changeable filters, tube voltage and current settings making them uniquely capable of being configured to maximize their sensitivity to specific elementsofinterest. Figure4:Diagramofanendwindowxraytube XrayTube:Electronsaregeneratedandacceleratedtohigh speedsandthenbombardedatargetusuallycomposedof a pure metal (e.g. W, Mo, Cr, or Rh). Upon reaching the target the electrons either interact and ionize the target, creating characteristic xrays, or are decelerated upon nearing the nuclei, creating a Bremsstrahlung continuum. Figure 4 illustrates the production of the XRay beam.
Figure 5: Overlapping emission spectra of W and Mo target Xray tubes run at differentvoltages
Figure6:EmissionspectraofanRhtargetXraytube 6 runatthreedifferentvoltages
Filter:Afiltercanbeplacedbetweenthetubeandthesampletoremoveundesirablebackgroundradiationbelow a certain voltage. The level of radiation filtered out is dependent on the filter element composition and its thickness.Table3suggestssomefiltertypesforcertainapplicatonsandAppendixAprovidesalistofothersfor theTRACeRandARTAXunits,aswellasotherfilterinformation.Notebothunitsallowtheusertofabricateany filterorsecondarytargettheythinkisbestfortheirapplication.
Table3:AvailablefiltersthatcanbeusedwiththeARTAXandTRACeRunits
Collimator:Collimatorsareusuallycircularoraslitandrestrictthesizeorshapeofthesourcebeamforexciting smallareas.Collimatorsizesrangefrom12micronstoseveralmm.Figure7illustratesthegeneralfunctionofa collimator. Detector: The detector is used to convert incoming xrays into proportionally sized analog pulses that are then converted by a digitalpulseprocessingsystemtoinformationthatcanbereadby a computer and displayed on a spectrum. The resolution of the detector depends on the type and quality of the detector (see Figure8).
Figure7:Xraypassingthroughacollimator Figure 8: Graph comparing the resolution of several differenttypesofdetectors
TRACER
The TRACeR is a handheld EDXRF unit used for instant nondestructive elemental analysis anywhere, anytime. It can be used in a wide range of applications including elemental analysis in material research, archeological digs, museum artifact analysis, conservation and restoration, electric utility industry, engine assembly,airframeassembly,scrapindustry,metalproducers,foundries,andmaintenanceassessment,andmany otherapplications. Xraytube(typicallyAg,RhorRe Target) Upto45kVXrays 170eVSiPINor145eVSDD 13BeDetectorWindow IRSafetySensor Vacuumwindow Userselectablefilter/target Upto200kcps(SDD)
Figure10:SketchofTRACeRmodule
Figure9:FrontviewoftheTRACeRunit
VacuumPort
SIPINDIODE/SDD DETECTORPARAMETERS
Table4:SiPIN/SDDcharacteristics (seeAppendixBforexplanations)
Ag,Rh,orRe Selectable(SeeAppendixA) Variable,045kV Variable,060A Selectable 15,000 max cps (PIN), 150,000 (maxcps(SDD) AirorVacuum
1023PIN/2048SDD
Table5:TRACeRoperatingparameters
ARTAX
The ARTAX unit is a semiportable open beam EDXRF machine used for nondestructive elemental analysisofsurfacesandthespectralmappingofsurfaceareaswithinminutes.Itmaybeusedinmanyapplications such as in archeometry, art history, restoration, forensic sciences, process related quality control, and material sciences.AppendixCincludesacompletetableoftechnicalparametersfortheARTAXunit.
Mo, W, Rh, or Cr, Cu, TiXraytube Upto50kV Less than 155 eV resolution Si(Li) SDD Detector Collimator ChangeableFilter Morethan100kcps
Figure11:ARTAXmodelwithlabeledparts
SI(LI)SDDDETECTORPARAMETERS
XrayTube Mo,W,Rh,Cr,orCu Selectable(SeeAppendixA) 050kV 01000A Variable <50,000cps AirorHeliumFlush
4096
Table6:TableofSi(Li)detectorparameters
Table7:ARTAXoperatingparameters
SPECTRALINTERPRETATION
AlthoughthegoalofXRFspectroscopyisgenerallytoelementallyanalyzethesample,severalphenomena inherentfromxrayphysicsinvolvedcontributetothespectra.Theseinfluencesrequireinterpretationinorderto correctlyunderstandthedata(seeFigure12).Threemaininfluencescontributetotheoutputspectraofasample: interactionsinthedetector,xrayscontributedbytheanalysissystem,andxrayinteractionsinthesample.These interactionsaredelineatedbelowanddiscussedindetaillaterinthedocument. InteractionsintheDetector o SumpeaksInterpretationoftwoormorepulsesasone
o o o
xrayscontributedbyinteractionsintheanalysissystem o XraytubetargetcharacteristiclinesRayleighscatteredintothedetector o DetectorcanlinesIronandNickeltracepeaks o WindowlinesCalciumtracepeaks o CollimatorandinstrumentstructurelinesAluminumtracepeaks o ThinFilmAnalysisDetectionofsurfacebelowsample PhenomenaintheSample o Rayleighscattering(elasticcollisions)Nolossofxrayenergy o Comptonscattering(inelasticcollisions)Partiallossofxrayenergyinthesample o MatrixeffectsMisrepresentationduetosecondaryabsorption/excitation,densityeffects o BraggscatteringConstructiveinterferenceofXraysinlatticestructures
TheATRAXandTRACeRunitshavetheoptiontochangethevoltage,current,andfilterselection,andthe use of a secondary target to specify the most efficient parameters for a given sample. By selecting the correct combination of these parameters, the above phenomena can be isolated, identified, and/or corrected yielding valuableinformationaboutthesamplescharacter.
Figure12:AbstractviewofPhysicsinvolvedinXraySpectroscopy
10
INTERACTIONSINTHEDETECTOR
SUMPEAKS
Whentwoormorexraysenterthedetectorattheexactsametimetheyarereadandconvertedintoone pulsewithenergy(e.g.amplitude)equaltothetwopulsescombined.Sumpeaksappearonaspectrumwhenthis occursenoughtimestocreateavisiblepeak,asseeninFigure13andFigure14.Intheory,sumpeakscanappear inanycombinationofcharacteristicenergies,buttheyaremostcommonlyfoundasdoubleKK,KKandKK becausethehigherrateofocurranceofthesexraysleadstoahigherprobabilityofasumeventbeingrecorded. Althoughsumpeaksaresmall,theymaybemistakenastraceelementsandcausespectralinterferencewithother characteristicpeaks. ARTAXunit 25kV 500A MoTube Nofilter 600seconds 49574cps
MoKpeak(17.48keV)
CuK1peak(8.047keV)
11
Figure14:Linearscaleofabovespectrum
ESCAPEPEAKS
While most characteristic xrays entering the detectorareconvertedintopulseswhichareprocessedby the digital pulse processor, an incoming xray can excite andcausefluorescenceinanatominthedetector.Ifthe xrayenteringthedetectorhasanenergygreaterthanthe absorption edge of an element in the detector (for the ARTAX and the TRACeR: Silicon), then fluorescence in the detector may occur. Figure 15 shows the typical relationshipbetween incoming x rayenergy and resulting Si escape peak counts. The inbound x ray will lose the amountofenergyrequiredtofluorescethedetectoratom, leaving the x ray with an energy E=E inbound E Characteristic energyofdetector,thuscausingthedetectortoreadthexrayas havinganenergyofE.
Figure15:Typicalrelationshipbetweenescapepeakcountand parent energy peak. Parent peak energies closer to the absorptionedgeofsiliconwillcreatemoreescapecounts
Escapepeaksaremuchlessintensethanthecharacteristicpeaksfromwhichtheyarederived.Several escapepeakscanoccurinonespectrum,giventhatallcharacteristicenergiesabovetheabsorptionedgeofthe detector are capable of causing fluorescence. In the case of a Si based detector, escape peaks will appear approximately 1.74 keV lower than a characteristic peak because silicon has a K absorbtion edge. Error! Referencesourcenotfound.showstheescapepeakfromtheCuKpeak.Thisfigurealsoshowsdetectoredge effect,whichoccursatapproximately60%ofthetotalKparentpeakenergy.Escapepeakscanbeautomatically correctedbycomputeralgorithmsandsoftwarethatcalculatesandoutputsacorrecteddatacurve(seeFigure17 andFigure18).
12
SiEscapepeak (8.041.73=6.26keV) Detectoredgeeffects (8.04x.6=4.82keV) Cusumpeaks Cu K parent peak (8.04keV) Cu K parent peak (8.90keV)
Figure16:SpectrafromaTRACeRmodelshowingboththeSiescapepeakoffofaCuparentpeakandthedetectoredge effects.
13
Figure18:LinearscalespectrumofaNISTstandardC1122(seeAppendixD)withoutthecorrecteddatacurveshowstheappearanceofthe Cuescapepeak.
Figure17:Byapplyingthecorrecteddatacurve,theescapepeakisremovedandthetruecharacterofthesampleisshown.
14
HETEROGENEITY
After radiation enters the detector and converts to pulses, discrepancies between peaks due to similar energy levels may occur. Heterogeneity can occur with any combination of lines, including different elements characteristicpeaks,sumpeaks,satellitepeaks,escapepeaks,etc.Theresolutionofthedetectordeterminesthe amountofoverlapbetweensimilarpeaks.
x 1E3 Pulses 150
100
Zn Fe Cu
50
0 0 5 10 - keV 15 20 25
Figure19:Linearscalespectrumofabronzeingotshowingtheoverlapofcopperandzinccharacteristicpeaks
Zn K peaks overlap with the Cu K peak to form a shoulder on the Cu peak insteadoftwoseparatepeaks
15
HETEROGENEITY(CONTINUED)
x 1E3 Pulses 150
100
Pb
Fe
Cu Zn
Pb
50
15
20
25
4.0
The Si K escape peak off of the Cu K peak (6.31 keV) can be easily confused withtheIronKpeak(6.40keV).Figures 3133showtheaffectoverlappingpeaks canhaveonapparentsamplecharacter.
0.0 x 1E3 Pulses 2 4 6 8 - keV 10 12 14 3.0
Pb
Sn
Fe
Ni
Cu
Zn
Pb
2.0
1.0
4.0
Figure 21: Closeup linear scale spectrum without the corrected datacurve.TheCuescapepeakappearstobeaFepeak
3.0
Pb Sn Fe Ni Cu Zn Pb 2.0
The downside of using a corrected data curve is the possibility of removingordiminishingtruesample lines.
1.0
16
INTERFERENCEWITHINSTRUMENTATION
EQUIPMENTANDINSTRUMENTCONTRIBUTION
Astheincidentradiationtravelsfromthesourcetothesample,itmaycausefluorescenceinmaterialsin themachinewhichmaybedetectedandshownonthespectrum.Thetargetelementmaybedetected(seeFigure 23)inadditiontoiron,zinc,copper,andnickelinthetube,collimators,lens,etc(seeFigure24).Byaddingafilter inbetweenthetubeandthesample,muchofthisunwantedradiationcanberemovedfromthespectrum(see Figure26).
Pulses
100
WL1
80
60
WL1
W
WL1
40
WL3
20
0 0 5 10 - keV 15 20
Figure23:LinearscalespectrumofanAl203refractorywithtungstenpeaksfromxraytube
17
EQUIPMENTANDINSTRUMENTCONTRIBUTION(CONTINUED)
Pulses
600
400 Ca Zr Fe Zr W Pb Ca Fe W Pb Zr
200
Fe K and KPeaks
WL3
Figure25:Linearscalespectrumshowingtheappearanceofpeaksofelementsfoundinthecanandotherinstrumentation(Ni,Fe,Cu, Cr). By adding a filter much of the unwanted radiation is removed from the spectrum. (Taken from Bruker AXS presentation IntroductiontoXraySpectrometry)
WDXRF
WL1 WL2
18
THINFILMANALYSIS(BACKGROUNDCONTRIBUTION)
X rays with high energy have the ability to partially penetrate through the surface of a sample. This phenomenon is also found in the use of filters, where a thin layer of metal or substance is used to attenuate certain energies from the exciting x ray beam that is incident on thesample of radiation. Ifthe sample being testedisthinenoughfortheradiationtoentirelypenetratethrough,elementsinthesurfacebelowthesample maybefluorescedanddetected.Thefiguresbelow(seeFigure26,Figure27,andError!Referencesourcenot found.)showstheeffectoflayeringthinfilmsamplesandthedetectionoflowerlayers.
Figure26:Measuredseparately,samples1,2,and3shows11565,21077,and50522countsofPb,respectively.Whenthesamples arelayeredintheorder321thetotalcountsincreasesto55768(around5000abovethatofjustsample3),indicatingthepartial detectionofthelowerlayersofPb.Whenthesamplesarelayered123thetotalcountsincreasesto22757(nearly2xthatofjust sample1),indicatingthedetectionoflowerlayersisdiminishedbutstillaffectsthespectra
19
PHENOMENAINTHESAMPLE
RAYLEIGH(ELASTIC)SCATTERING
Incidentradiationfromthetubethatreachesthesampleiseitherabsorbedinthephotoelectriceffector reflectedandscattered.Whenanxrayreflectsofftheatomsofthesamplewithoutlosinganyenergyitiscalled Rayleigh(orelastic)scattering.Theenergyoftheoutboundxraywillbeequaltotheenergyoftheinboundxray, thus being detected as a source peak with the energy of the inbound x ray. The Rayleigh scatter peaks visible correspondwiththecharacteristicenergiesofthexraytubetargetelement.Figure29andFigure30showthe appearanceofRayleighpeaksinaspectrum.Rayleighscatterpeaksarecharacterizedbysharpshapesthatarethe sameasthexrayflourescencepeaksbecausetheyareproducedinthedetectorbysingleenergyxrays.
x 1E3 Pulses
2.5
2.0
0.5 1.5
Fe
Co
Ni
Cu
1.0
SiKescapepeak
0.0 4 6 8 - keV 10 12
Figure28:CloseuplinearscalespectrumofaNISTstandardC1251a(seeAppendixDforcomposition)showingthe appearanceofthetungstentargetRayleighscatterinspectra
TRACeRunit
Figure29:CloseuplinearscalespectrumshowingtheappearanceofRhcharacteristiclinesinaspectrumduetoRayleighscatter
20
RAYLEIGH(ELASTIC)SCATTERING(CONTINUED)
InadditiontoRayleighscatterpeaksatthecharacteristiclinesofthetargetinthetube,aBremsstrahlung radiationcurvemayappearinthebackgroundofaspectrumduetoRayleighandComptonscatterofallincidentX rays(Figure31). ARTAXunit 40kV 998A WTube 315mAlfilter 600seconds 1486cps
Figure30:LinearscalespectrumofahighpurityAl2O3 refractory withawelldefinedBremsstrahlungcontinuum
21
COMPTON(INELASTIC)SCATTERING
Incident radiation with sufficient energy to ionize an inner shell electron in an atom does not always cause fluorescence, but insteadcausesanexcitationwithoutlosingallofitsenergy(seeFigure 32).Intheseinteractions,calledComptonscattering,axraystrikesan atom and loses energy, causing the excitement of an innershell electron.Becausenovacancyiscreatedintheatom,nocharacteristic energy is released; however, the x ray will lose energy and be scattered in all directions (noted by the formula in Figure 32). Comptonscatterxraysappearasabroadpeakdefinedbytheangle between the incident beam and the detector for the target characteristicxraylowerinenergythanRayleighscatterpeaksbecause Figure 31: Compton scattering causes a shift in theyonlyloseasmallamountofenergyintheexcitationofanelectron. wavelengthoftheincidentxray Theyaregenerallybroadduetotheareaofthedetectorandtheareaof theexcitingbeam,andtheyoccurmoreofteninlowZelements.ComptonscatteringcanbeseeninFigure33.
22
Figure32:LinearscalespectrumshowingComptonscatteringofrhodiumparentpeaksonaTRACeRmodel.
MATRIXEFFECTS
Absorption: Any element that can absorb or scatter the incident xrays is capable of reabsorbing characteristic xrays of other elements (see Figure 34). After an atom undergoesthephotoelectriceffectandemitsa characteristic x ray, the x ray may be reabsorbed by another atom in the sample. When this happens, it causes a misrepresentation of the counts of elements detectedfromthesamplebyfailingtocounta x ray for the initial element. The expected number of x ray counts recorded by the detector will be lower than expected because someareabsorbedwithinthesample.
Figure33:Exampleofsecondaryabsorption.
SecondaryExcitation:Whenthecharacteristic radiation from one atom is reabsorbed by another atom and has sufficient energy to ionize the atom, it will cause fluorescence in the second atom, producing only the characteristic radiation of the second atom (see Figure 35). This can lead to a misrepresentation of elements by enhancing the appearance of elements through secondaryexcitation.
Figure34:Exampleofsecondaryexcitation
23
BRAGGSCATTERING
Many samples are composed of a periodic arrangement of atoms or molecules that create a crystal lattice.Whenamaterialexhibitsalatticestructure,severaldifferentlatticeplanescanexistorientedindifferent directions. All planes that are parallel to a given lattice plane are a set distance away from one another, as establishedbythecrystalstructure. Whentwoparallelincidentxraysstrikeapairofparallellatticeplanes,theraysarereflectedandcan interferewitheachother.ThisphenomenonisknownasBraggscattering,orthescatteringofincidentxraysdue tothecrystallatticeofasample.Braggslawstatesthattwowavesinterfereconstructivelywhenn=2dsin where:nisthereflectionorder,isthewavelengthoftheincidentxray,disthedistancebetweenthelattice planes,andistheangleofreflectionuponthecrystallattice.AccordingtoBraggslaw,agivenanglehasa specificsetofwavelengths(orenergies)thatcancauseconstructiveinterference.Sampleswithuniformcrystal structurewillexhibitnarrowpeaks,whileasamplewithanonuniformcrystallatticewillbebroader.Allthedata usedintheexamplesbelowaretakenwiththeArtaxsystem.ThesharperlinesinFigure37suggestthatthe sampleisofhigherquality(moreuniformcrystal)thanthatinFigure38.
Figure35:LinearscalespectrumofabluegemstoneshowingtheeffectofincidentangleonBraggpeaks
24
MultipleBraggpeaks
Figure36:Linearscalespectrumofagreengemstoneatangles0,10,and25,showingthechangesinBraggpeaksandtheireffectonother characteristicpeaks.
Bragg Scattering can interfere with peaks and causeamisrepresentation ofelementsinthesample. In this example, the peak is greatly enhanced by Bragg scattering in the green spectrum, whereas the pink spectrum peak is not exaggerated because theanglewaschanged.
25
BRAGGSCATTERING(NISTC1122EXAMPLES)
Angle 1 Angle 2 Angle3
Figure37:LinearscalespectrumofaNISTstandardC1122(seeAppendixD)samplewithnofilterexhibitingtheBraggpeaks
26
BRAGGPEAKS
BraggPeaks
Figure38:Braggpeaksappearindifferentintensitiesandpositionsbasedonthecrystalorientationofthesample.
BraggScattering(Gemstoneexamples)
BraggPeaks
Figure39:CertainBraggpeaksareamplifiedatspecificangles,whilediminishedatothers.
27
BraggPeaks
Figure40:Braggpeaksmayappearontopofotherpeaks,asseeninthered.
BraggPeaks
BraggPeaks
Figure41:TheredspectrumdisplaysnoBraggpeaks,whilethegreenandpinkhaveseveraldifferentpeaks
28
FIELDAPPLICATIONS
IDENTIFYINGGENUINEARTIFACTS(CHELSEABULLFINCHEXAMPLE)
WiththeincreasedportabilityandeaseofuseintheTRACeRandARTAXunits,severalapplicationscanbemadeinseveraldifferentareasofstudy. ThefollowingareafewexamplestakenfromapresentationbyDr.BruceKaiseronEDXRFapplications.
Figure42:OverlappingspectraofthetwodifferentpaintsprovethattheceramicBullfinchwasrestoredfromitsnaturalcondition.
29
IDENTIFYINGTRUEORIGINS(STONEWAREEXAMPLE)
ByUsingtheTRACeRandXRFspectroscopytechnology,historians can easily identify the origins of otherwise unknown artifacts. Here, English stoneware (blue spectrum) contains significantly moreironthanGermanstoneware(redspectrum).Byusingthis knowledge,historianscanidentifytwovisuallyidenticalpots.
Figure43:TrueoriginsofartifactscanbefoundusingelementalanalysisthroughXRFtechnology
30
Measuring Chlorine with the Tracer Measurement of Cl is very important as it is often involved in corrosion and degradation of artifacts in marine environments. Or in some cases is a key constituent of pigments or other coatings, or an issue in paper conservation. The following slides first depict how to make up thin film standards to determine the Cl surface content in micro grams per square centimeter. And then how to set the Bruker handheld xrf instrument up to measure levels as low as 10 micro grams per square and shows 2 applications. It should be noted the Cl analysis is very much a SURFACE ANALYSIS when using xrf, as the Cl atom emits only a 2.7 keV x ray. This low an energy x ray is not able to escape the sample unless the atom is very near the surface.
31
CreationofveryverythinfilmChlorineStandards
32
C
Pure iron 0 .0 Cl 0.99 Cl 4.93 Cl 9.87 Cl
Each standard was analyzed for 3minutes at 2 different voltage and currentsettings. A 0.001TitaniumfoilwasusedinbothcasestoeliminatetheRhLlinesandgenerateTix raystoexciteClefficiently.ThethinpaperstandardswerebackedbypureFetomimicCl corrosiononFe.Theaboveisaplotofthemeasurementsthatweretakenat8kVand35 micro amps. The peak at 2.6 keV is the Cl K xray. The peak at 2.95 keV is a constant amplitudeandisaresultofArKxraywhichisintheairinthepaper.Itisclearthesystem issensitivetoCldowntolevelsaslowas1microgram/squarecm
33
BrukerArtaxXrfScanofHunleyRivet
Theoperatingparametersare:
Spotsize0.070mm2(microfocustube) Samplinggrid0.070mm2 15kVtubexraytubevoltage Motubetarget 300microampsxraytubefilamentcurrent 60secondanalysistimeperpointwithHeliumflush Beamarmwaspointeddownbutcanbeorientedin anydirectionforanysizedobject. Systemtripodisonwheelsandcanbemovedquickly. AnalyticalsoftwarerunseasilyonanyWindowsXPsystem
34
Micrograms/cm2 of Chlorine
26-28 24-26 22-24 20-22 18-20 16-18 14-16 12-14 10-12 8-10 6-8 4-6 2-4 0-2
Start scan
mm
Machined area
mm
un Machined area
35
Fe Si escape peak
Scan56showingtypicalinstrumentresponse
36
BrukerTRACeROperatingParameters
40kVand10microamps .006Cu,.001Ti,.012Alfilters 180secdataacquisition
ThefollowingdataisanexampleofTraceranalysisdonebyJeffSpeakmanoftheSmithsonian
37
38
6000
5000
4000
3000
Fe
2000
alca Chivay CRG 0002 Ixtepeque MLZ 1019 Mono Glass MTN Otumba Pico de Orizaba Quispisisa Sierra de Pachuca Ucareo UNL-050_1 XMC 020 Yellowstone
Zr
1000
Zr
0 5 6
Rb F7 C
8
Nb Sr
14
10
11
12
13
15
16
17
18
19
20
39
alca Chivay CRG 0002 Ixtepeque MLZ 1019 Mono Glass MTN Otumba Pico de Orizaba Quispisisa
Rb Zr
Zr
Nb
C
0 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
40
MeasurementofToningAgentsonPhotographs
Michele
Use.006Cu,.001Ti,.012Alfilter Analyzeatphotograph
1970
Thewhiteareaisjustthepaperandmountingmaterials Thefilterusedremovesmostofthebackscatteredxrays
41
1500
Ba L
Note the red difference spectrum clearly shows that the image forming agent is only Ag
Ag
1000
Pd K
500
Ag L
Rh K Rh K Pd K inelastic
Ag Ba
Fe K
Cu
Sr
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
APPENDICES
APPENDIXA
FILTERSAVAILABLEFORTHEARTAXANDTRACERUNITS(NOTEUSERCANMAKEUPANYFILTERORSECONDARY
TARGETHECHOSESDEPENDINGONHISPARTICULARNEEDANDAPPLICATION )
SomeoftheFiltersavailablefortheARTAXunitfromBruker o 315mAl o 25mNi o 12.5mNi o 12.5mMo o 100mAl o 200mAl SomeoftheFiltersavailablefortheTRACeRunitfromBruker o Bluefilter(1milCu) o Yellowfilter(12milAl+1milTi) o Redfilter(12milAl+1milTi+1milCu)
EFFECTOFAFILTERONANXRAYTUBESPECTRUM
Figure44:Scatteredexcitationspectrumprovidedbyasilvertargetxraytubeoperatedat15kV.Theplotsshowtheeffectoftwo thicknessesofanalprimarybeamfilter:(1)unfiltered;(2)thinalfilter;(3)thickalfilter.
43
EXAMPLEOFANATTENUATIONGRAPHFORAFILTERCOMPOSEDOFVARIOUSTHICKNESSESOFFE
Figure45:Toattenuateahigherpercentageofradiation,aswellasahigherenergyradiation,thefilterthicknessmustincrease.
44
Figure46:Thisgraphismadetodeterminethefilterthicknessrequiredforoptimalresultsintestingacoppersample.
45
46
MeasurementofMg,Al,Si,P,Cl,S,K,Ca,V,Cr,andFe(andanyLandMlinesfortheelementsthatfallbetween 1.2and6.5keV) Tifilter 15to20kV 15to20microamps Vacuum Thesesettingsallowxraysfrom3to12keVtoreachthesample.InparticularthisdoesnotallowtheRhLlines fromthetubetoreachthesample.TheseRhLxrayswouldinterferewithClandSanalysis.Forexample,thisisa verygoodsetupformeasuringClonthesurfaceofFe. Measurementofmetals(TitoAgKlinesandtheWtoBiLines): 0.001Ti,.012AlFilter(yellow) 40kV 1.2to2.6microamps Novacuum Thesesettingsallowallthexraysfrom12keVto40keVtoreachthesamplethusefficientlyexcitingtheelements notedabove.Thesearethesettingsusedtocalibratethesystemforallmodernalloysofthoseelementsofthose listedinthetitleofthissection.ThereislittleornosensitivitytoelementsbelowCawiththesesettings. 1. 2. 3. 4. 1. 2. 3. 4.
47
48
APPENDIXB
Figure47:ExplanationofSiPINcharacteristics
49
APPENDIXC
50
APPENDIXD
BerylliumCopper C1122 PhosphorizedCu C1251a
Elem Sb Sn Ag Bi Pb Se Zn Cu Ni Co Fe Mn Cr Al P Si Mg S Be As Te Au
% .01 .005 .003 .01 97.45 .01 .22 .16 .004 .002 .17 .004 .17 1.75
Elem Sb Sn Ag Bi Pb Se Zn Cu Ni Co Fe Mn Cr Al P Si Mg S Cd As Te Au
% .0014 .0016 .0080 .00037 .00235 .0011 .0024 99.89 .00236 .00132 .0285 .00046 .0003 <.002 .0420 <.005 <.002 .0035 <.0003 .0016 .0016 .00155
51
APPENDIXE
Figure48:PeriodictableincludingtheKandLseriescharacteristicenergiesofalloftheelements
52
53