materials Article Positron Annihilation and Complementary Studies of Copper Sandblasted with Alumina Particles at Different Pressures PawełHorodek1,2,*,KrzysztofSiemek1,2 ID,JerzyDryzek1andMirosławWróbel3 1 InstituteofNuclearPhysicsPolishAcademyofSciences,PL-31342Krakow,Poland; [email protected](K.S.);[email protected](J.D.) 2 JointInstituteforNuclearResearch,6JoliotCurieStr.,141980Dubna,Russia 3 FacultyofMetalsEngineeringandIndustrialComputerScience,AGHUniversityofScienceandTechnology, 30MickiewiczaAve.,90-059Krakow,Poland;[email protected] * Correspondence:[email protected];Tel.:+7-(496-21)-64-479 Received:6October2017;Accepted:20November2017;Published:23November2017 Abstract: Positron annihilation spectroscopy and complementary methods were used to detect changesinducedbysandblastingofaluminaparticlesatdifferentpressuresvaryingfrom1to6bar inpurewell-annealedcopper. Thepositronlifetimemeasurementsrevealedexistenceofdislocations andvacancyclustersintheadjoinedsurfacelayer. Thepresenceofretainedaluminaparticlesinthe copperatthedepthbelow50µmwasfoundintheSEMpicturesandalsointheannihilationline shapeparameterprofilesmeasuredintheetchingexperiment. Theprofilesshowusthatthetotal depthofdamagedzonesinducedbysandblastingofaluminaparticlesrangesfrom140µmuptoca. 800µmanditdependsontheappliedpressure. Thework-hardeningoftheadjoinedsurfacelayer wasfoundinthemicrohardnessmeasurementsatthecross-sectionofthesandblastedsamples. Keywords: defects;copper;sandblasting;positronspectroscopy 1. Introduction Theprocessofbombardingasurfacewithsmallabrasiveparticlesiscalledsandblasting.Itis oftenusedinindustrytocleansurfacesofdifferentobjectsbyremovingpaints,impurities,orcorrosion products. It is the basic tool in eliminating oxides created during heating of alloys destined to be future implants in prosthetics [1]. Another important application of sandblasting is induction of nanocrystallization[2]. The most studied materials exposed to sandblasting are dental alloys [1], stainless steels [3], titanium[4],andtitaniumalloys[5]. Scanningelectronmicroscopy(SEM),atomicforcemicrocopy (AFM),X-raydiffraction(XRD),ormicrohardnessmeasurementsarecommonlyusedtoinvestigate themorphologyofsandblastedsurfacesorpropertiesofthenanocrystallizedsubsurfaceregions[6,7]. Studiesrevealedthatsandblastedsurfacesaresensitivetothestreampressure, size, andchemical compositionoftheappliedparticles. Fastparticleshittingthesurface,nexttochangesin,e.g.,roughness,generateanavalancheof stressesandplastic,aswellaselastic,deformationsbelowit. Theseleadtothecreationofanumber of different kinds of structural defects with unknown distributions and occupied depths. Studies are rarely performed in this context, although the presence of defects has a direct impact on the propertiesofthesubsurfaceregion[8]. Theiroccurrenceisrevealedinchangesofstrength,ductility, andmicrohardness[9,10]. Corrosioncanbealsoconnectedwithdefects[11]. Moreover,fasterwearof materialsstartsattheatomiclevelandcanaffectthelifetimeofthematerial[12]. Inthispaperwereportpositronannihilation(PAS)andcomplementarystudiesofpurecopper samples exposed to sandblasting of alumina particles at different pressures. Copper was chosen Materials2017,10,1343;doi:10.3390/ma10121343 www.mdpi.com/journal/materials Materials2017,10,1343 2of15 becauseitisametalwhosespecialphysical,chemical,electrical,andmechanicalpropertieslocateit onthethirdpositionamongthemostcommonlyusedmetalsintheworld. Thewidespectrumof applicationsandwell-knownstructuremakecopperaconvenientobjectforbasicstudies[13,14]. The experimental techniques were selected to provide wider discussion related to impact of sandblastingonchangesgeneratedonthesurfaceandintheregionbelowit. Positronannihilation spectroscopy(PAS)isthemainappliedmethod. Thisisasuitabletoolforthedetectionofopen-volume defects,suchasvacancies,theirclusters,dislocations,voids,etc. Itallowsustorecognizethetypeof introducedstructuraldefectsandtheirdepthprofile. ThesuccessfulapplicationofPASinstudiesof defectsinmaterialstreatedindifferentwaysiswelldocumentedintheliterature[15–18]. Althoughtherearemanystudiesofsandblastedcopperreportedintheliterature,PASinvestigations havenotyetbeendone.Erosionofcopperwasmonitoredwithsurfaceprofilemeasuringbyobservation oftheprofileproducedbyblastingwithaluminagritparticles30µmindiameter[19].Itwasdemonstrated thaterosiondependsonincidentangleofthesandstreamandincreasesfromtheminimumfornormal incidencetothemaximumforanglesintherangebetween10◦ and20◦. Thenitdecreasestozerofor anglesreducingto0◦. Usingelectronmicrographs,Milleretal.[20]showedthatslowertimedecayof Ramansignalsisobservedduetothegrowthofasurfaceoxidelayer.Inturn,theinitialRamansignal strengthsaresimilarincomparisonwithonlyetchedplates. Yuanetal.[21]studiedtheeffectofCu surfacerougheninginducedbysandblastingontheoxidationbehaviourofCu.Theydemonstratedthat increasingthesandblastingtreatmenttimereducestheoxidefilmgrowthandsuppressestheoxidefilm delaminationfromtheCusubstrate. Weintendtocharacterizesurfaceandsubsurfacelayerofsandblastedcopperdependingonthe pressureofairtransferringaluminaparticles.Themainfocusisonademonstrationofthedepthprofiles of structural defects concentration and evaluation of their kind/size. Additionally, SEM pictures, microhardness,androughnessparameterswerealsotestedtosupplementthestudies. 2. MaterialsandMethods 2.1. SamplePreparation Commercialpuritycopper(99.95wt%),producedbyKGHMPolskaMiedz,Warsaw,Polandand purchasedfromKBHAkord,Krakow,Poland,wastheobjectofstudies. Allsampleswithdimensions of10mm×10mm×5mmwerefirstlypolishedtoobtainmaximallysmoothsurfaces. Thenthey wereannealedat1000◦Cfor4hinavacuumof10−5 Torrandslowlycooledtoroomtemperature. Thisprocedureallowedustoprepareidenticalspecimens,withonlyresidualdefects,inconditions protectedfromoxidation. Twosamplesweresavedasreference,whileotherswereexposedtosurface treatment. SandblastingwasperformedusingaRenfertVarioBasicJetblaster,Hilzingen,Germany. TheabrasivematerialEdelkorundcontaining99.8%aluminumoxide(Al O )withasizeof250µm 2 3 wasapplied. Thesurfaceswereblastedfor60sunderpressuresof1–6bar(with1barsteps)with adistanceof10mmbetweenthesampleandtheperpendicularly-directednozzle. Thenozzlesize was1mm. Therewerealwaystwospecimenspreparedforeachemployedpressuretouseinpositron measurements. TheschemeofthesandblastingprocedureisshowninFigure1. Materials 2017, 10, 1343 3 of 15 Materials2017,10,1343 3of15 Figure1.Theschemeofthesandblastingprocedureofthestudiedsamples. Figure 1. The scheme of the sandblasting procedure of the studied samples. 2.2. PositronMeasurements 2.2. Positron Measurements Thepositronlifetime(LT)measurementswereperformedatInstituteofNuclearPhysicsPolish The positron lifetime (LT) measurements were performed at Institute of Nuclear Physics Polish AcademyofSciencesinKrakowusingafast-fastspectrometerbasedonBaF scintillators. Thetiming Arecsaodleumtioyn oefq Sucaielendcecsa .in2 5K0rpaks.oTwh eusisinotgo ap efa2s2tN-faaswt sitpheacntraocmtievtietyr boafs3e2dµ oCni eBnavFe22 lsocpinedtililnattoortws. oTh7eµ tmimthinicgk resolution equaled ca. 250 ps. The isotope 22Na with an activity of 32 µCi enveloped into two 7 µm kaptonfoilswasplacedbetweentwoidenticalsamples. Theanalysisofobtainedspectra,including thick kapton foils was placed between two identical samples. The analysis of obtained spectra, 106counts,wasprovidedwiththeLTprogram[22]. Thesourcecontribution,background,andfinite including 106 counts, was provided with the LT program [22]. The source contribution, background, timeresolutionweretakenintoaccountasadjustableparametersinthedeconvolutionprocedure. and finite time resolution were taken into account as adjustable parameters in the deconvolution Doppler broadening of the annihilation line (DB) measurements were performed at the Joint procedure. InstituteforNuclearResearchinDubnausinganencapsulated22Napositronsourcewithanactivity Doppler broadening of the annihilation line (DB) measurements were performed at the Joint of15µCi. Theisotopewasclosedinacoppercapsuleof5mmindiameterand7µmthicktitanium Institute for Nuclear Research in Dubna using an encapsulated 22Na positron source with an activity window. The exit of positrons was only possible through the window according to this geometry. of 15 µCi. The isotope was closed in a copper capsule of 5 mm in diameter and 7 µm thick titanium Thesourcewasplacedinfrontofthedetectorwiththewindowdirectedtothetop.Thesamplecovered window. The exit of positrons was only possible through the window according to this geometry. thewindowwiththestudiedsurface. Positronswereimplantedinalldirectionsgivinginformation The source was placed in front of the detector with the window directed to the top. The sample fromthesampleandthecapsule. AnychangesinPAScharacteristicscouldcomeonlyfromthesample. covered the window with the studied surface. Positrons were implanted in all directions giving Moredetailsofthemeasurementsetuparegivenin[23]. TheDBexperimentswereconductedusing information from the sample and the capsule. Any changes in PAS characteristics could come only an HPGe detector with an energy resolution of 1.20 keV for an energy of 511 keV. Each obtained from the sample. More details of the measurement setup are given in [23]. The DB experiments were spectrumwasanalyzedtocalculatetheannihilationlineshapeparametercalledtheSparameter. Itis conducted using an HPGe detector with an energy resolution of 1.20 keV for an energy of 511 keV. givenastheratiooftheareabelowthecentralpartofannihilationpeaktothetotalareaintherangeof Each obtained spectrum was analyzed to calculate the annihilation line shape parameter called the S thisline. Theenergyintervaltakenforcalculationisalwaysconstantwithinthewholemeasurement parameter. It is given as the ratio of the area below the central part of annihilation peak to the total session. Extraction is additionally provided without background. Usually the energy window is area in the range of this line. The energy interval taken for calculation is always constant within the fixedinthiswaytoobtainavalueoftheSparametercloseto0.5. Theextractionofthemeasured whole measurement session. Extraction is additionally provided without background. Usually the annihilationlineswasperformedusingtheSPcode[24]takingintoaccountfollowingenergywindow: energy window is fixed in this way to obtain a value of the S parameter close to 0.5. The extraction of |Eγ −511keV|<1.38keV.Suchawindowisattheannihilationlineinthelow-momentumelectron the measured annihilation lines was performed using the SP code [24] taking into account following region,andtheirgreaternumberisintheopenvolumedefects. Then,thevalueoftheSparameter energy window: |Eγ − 511 keV| < 1.38 keV. Such a window is at the annihilation line in the low- is extremely sensitive to the open volume defects which trap positrons. Briefly, the value of the S momentum electron region, and their greater number is in the open volume defects. Then, the value parameterincreaseswhentheirconcentrationincreases,however,thedependencyisnotlinearand of the S parameter is extremely sensitive to the open volume defects which trap positrons. Briefly, canalsobesensitivetothesizeandtypeofdefect(seeAppendixA). the value of the S parameter increases when their concentration increases, however, the dependency Positronsemitteddirectlyfrom22Naarecharacterizedbycontinuousenergyspectrumwiththe is not linear and can also be sensitive to the size and type of defect (see Appendix A). energyend-pointequalto545keV.Forthiscasethenumberofpositronsdecreasesexponentiallywith Positrons emitted directly from 22Na are characterized by continuous energy spectrum with the thedepthincreasefromtheentrancesurface. Thelinearabsorptioncoefficientfor22Napositronsin energy end-point equal to 545 keV. For this case the number of positrons decreases exponentially copperisabout348cm−1[25]. Thus,about63%ofprimaryemittedparticlesannihilateatthemean with the depth increase from the entrance surface. The linear absorption coefficient for 22Na positrons Materials 2017, 10, 1343 4 of 15 Materials2017,10,1343 4of15 in copper is about 348 cm−1 [25]. Thus, about 63% of primary emitted particles annihilate at the mean implantation depth, which is about 26 µm (or 99% at the depth up to ca. 119 µm). The total depth of itmhep dlaanmtaatgioend dlaeypetrh ,isw mhiucchhi shaigbhoeurt t2h6anµ mth(iso rv9a9lu%e,a tthtuhse, doebptathinuinpg ttohcea d.e1p1t9hµ dmep).eTnhdeentociteasl dofe pththe omfetahseudreadm aangneidhillaayteiorni scmhaurachctehriigshtiecrs trheaqnuitrheiss av asplueec,iathl tuesc,honbitqauine.i ngthedepthdependenciesofthe measIut rheadsa bneneinh iplartoiovnedc hina rmacatneryi setxicpserreimqueinretss tahsapt esceiqaulteencthianli qreume.oval of layers by chemical etching and mItehaassubreeemnepnrtosv oefd ainnnmihainlaytieoxnp ecrhimareancttesrtihsatitcsse mquaeknet iiatl preomssoibvlael toof ldaeyteercst btyhec hdeempitcha lpertocfhilien gina nadn macecausruarteem weanyts. Cofhaenmniichaill aetticohninchga draocetse rniostti cpsromdaukceei tanpyos dsiebfleecttso wdehtieccht tchoeudlde pdtihstpurrobfi tlheei ninaintiaalc cduerfaetcet wdeapyt.hC dhiestmriibcuatlioetnc h[1in5,g16d]o. eIns nooutr pstruoddiuecse thaen ysadmepfelcetss wwehriec hetcchoeudld ind ias t2u5r%b tshoeluitnioitnia lofd nefietrcitdde eapctihd danisdtr dibisuttililoend [w15a,t1e6r]. .AI nthoinu lrasyteurd oife asbtohuets 1a0m µpmle ws wase craereeftuchlleyd reinmaov2e5d% byso tlhuitsi omnixotfurnei.t rTidhee tahciicdknaensds dofi stthilele sdamwpaltee rw. Aast mhienalsauyreerdo ufsaibnogu at d10igµitmal wmaicsrcoa-srecrfuewlly wreitmh o±v2e µdmb yactchuisramcyix. ture. Thethicknessof thesamplewasmeasuredusingadigitalmicro-screwwith±2µmaccuracy. 2.3. Surface Characterization 2.3. SurfaceCharacterization Scanning electron microscopy (SEM) tests were performed at AGH University of Science and TechSncoalongnyin ign eKlercatkroown musicinrogs Hcoiptaych(Si ESM-35)0te0sNts SwEMere (bpye rHfoirtmacehdi LattdA. GToHkyUon, iJvaeprasnit)y, eoqfuSicpipenedce wanitdh TEeDcSh nNooloragny 9in86KBr-a1kSoPwS (ubsyi nTgheHrmitaoc hSiciSe-n3t5if0i0cN, WSaElMtha(mby, MHiAta, cUhSiALt)d. .ThToe kryoou,gJhanpeasns) ,weaqsu dipepteerdmwiniethd EfrDomS Nopotriacnal9 p8r6oBf-i1leSrPs Sob(btayinTehde rbmy oWSYciKenOt iNficT,9W30a0l t(hbaym V,eMecAo ,IUnsStAru)m.Tehnetsr,o Pulgaihnnveisesww, NasYd, eUteSrAm).i nTehde frrooumghonpetsisc aplarparmofielteerrss wobetraei nveedrifbieyd Win YfoKuOr inNdTe9p3e0n0d(ebnyt mVeeeacsourIenmsternutms feonrt es,acPhl asianmvipelwe., TNhYe ,chUoSsAen). Tarheea wroausg lhoncaetsesdp calorasem teot ethrse cweenrteer vaenrdifi ceodveirnedfo 4u mr min2d ienp ae nsidnegnlet tmeset.a Asunrye mweinndtsowfo rfiletearcihngs aomptpiolen. Tfohre ocbhtaoisneinnga rveaaluweas swlaosc antoetd acplposlieedto dtuhreincge nthteer taesntds. covered 4 mm2 in a single test. Any window filteringoptionforobtainingvalueswasnotappliedduringthetests. 2.4. Microhardness Tests 2.4. MicrohardnessTests Samples sandblasted at 1, 3, and 6 bar were chosen as the representative ones for the measSuarmempleensts oafn tdhbel amstiecdroahta1r,d3n,easnsd d6epbathr wpreorfeilceh. oTsoe nthaiss tahimer,e tphree ssaenmtaptlievse wonerees cfourt tuhseinmge aa sduiaremmoenndt osafwth peermpeicnrdoihcaurldarnleys sto dthepe tshurpfarocefi lteo. oTbotatihni sthaei cmro,stsh-seecstaimonp. lTehsewn,e trhee csuutrfuascien ogf tahed icarmososn-sdecstiaown pwearsp seinmduicltualnaerloyutsolyt hgerosuunrfda mceetcohoabntiacainllyth uescinrogs fsi-nsee-cgtrioanin.eTdh SeinC, sthanedsuparfpaecre wofitthh gercitrso susp-s teoc 4ti0o0n0 wanads sfiinmaulllyta npeooliusshleydg ruopu.n dCumteticnhga,n gicrainllydiunsgin, ganfidn ep-gorlaisinheindgS iwCesraen dcopnapdeurcwteidth agcrciotsrduipngto t4o0 0a0 apnrodcfiendaulrlye pdoedliischaetdedu fpo.r Cteusttsti bnyg ,eglercintrdoinn bga,caknsdcaptotelrisehdi ndgiffwraecrteiocno n[2d6u].c Stetrduearcsc’o drdevinicgesto anadp mroacteedriuarles wdeedreic uasteedd finor ttheisst spbryepealercattrioonn. bTachkes cmatitcerroehdadrdifnfreascst imonea[2su6]r.emSterunetsr s’wdeerve icceasrrainedd mouatte rwiailtsh wTeUreKuOsNed™i n2t5h0i0s pmraenpuarfaatcitounr.edT hbeym Wicilrsoohna rIndsntersusmmeenatss u- rAemn eInnsttsrwone rCeocmarpraiendy,o NutowrwitohoTdU, MKOAN, U™SA25. 0K0nmooapn’usf macetuthroedd bwyasW rielsaolinzeIdn swtriuthm ean ltosa—d Aofn 0I.n1s tNro (nHCKo0m.0p1)a.n Tye,sNtso arwbiodoedd, bMyA th,eU SreAco.Kmnmoeonpd’satmioenths oodf twhae sISreOa l4iz5e4d5 wstaitnhdaarloda. dof0.1N(HK0.01). TestsabidedbytherecommendationsoftheISO4545standard. DDuurriinngg mmeeaassuurreemmeennttss lloonnggeerr ddiiaaggoonnaall wwaass aallwwaayyss ppaarraalllleell ttoo aavveerraaggee ttrraaccee ooff tthhee bbllaasstteedd ppllaannee.. TThhee lleennggtthh oofft hthisisd diaigagononalaal nadndit sitds idstiasntacnecteo ttoh itshsihs aspheap(he i(nhF iing uFriegu2r)ew 2e)r ewmereea smuereadsueraecdh etiamche atifmteer iamftperr eimssipornesesxioecnu etxioenc.ution. Figure2.Aschematicillustrationofmicrohardnessmeasurementprocedure. Figure 2. A schematic illustration of microhardness measurement procedure. Materials2017,10,1343 5of15 Materials 2017, 10, 1343 5 of 15 3. ResultsandDiscussion 3. Results and Discussion 3.1. SEMandOpticalProfilers 3.1. SEM and Optical Profilers Asexpected,thesandblastingdefinitelymodifiesthemorphologyofthesurfacecanbeseenin As expected, the sandblasting definitely modifies the morphology of the surface can be seen in Figure3. Itshowsthesurfacesofreferenceandsandblastedat1and6barsamplesobservedbyoptical Figure 3. It shows the surfaces of reference and sandblasted at 1 and 6 bar samples observed by optical profilometryandSEM.Itisclearlyvisiblethatthefiguresdifferandthisprovestheappliedpressure profilometry and SEM. It is clearly visible that the figures differ and this proves the applied pressure influencestheroughness. Roughnessseemstobedependentonpressure. influences the roughness. Roughness seems to be dependent on pressure. Figure 3. SEM and optical images of reference and sandblasted copper specimens at 1 and 6 bar. Figure3.SEMandopticalimagesofreferenceandsandblastedcopperspecimensat1and6bar. Results of the quantitative examination of optical profilers for all studied samples were plotted Rine Fsuiglutsreo 4f tahs ethqeu raonutgihtanteivsse aevxearamgien Raat ipoanraomfeotpert ivcearlspusr opfirelessrusrfeo. rRaa lils stthued mieedans ahmeigphlet sexwperersespedlo tted inFiginu rmei4craosmtehteerrso aus gchalncueslasteadv eorvaegr ethRe epntairrea mmeetaesrurveedr sleunsgpthr eosrs aurreea.. RIn tihsrethe-edmimeeannsiohneaigl hcatseex ipt irse ssed a a inmigcirvoemn eatse: rsascalculatedovertheentiremeasuredlengthorarea. Inthree-dimensionalcaseitis givenas: 1 M N RRaa=MM1NN∑iM1∑jN1(cid:12)(cid:12)ZZjjii(cid:12)(cid:12),, (1) (1) i=1j=1 where M and N are number of data points in X and Y, and Z is the surface height relative to the mean whereMandNarenumberofdatapointsinXandY,andZisthesurfaceheightrelativetothemean plane. The dashed line represents the value of Ra = 0.4 µm for the reference sample. In turn, the Ra planep.aTrahmeedtears ihse hdiglhinere froerp sraensdebnltassttehde svpaecluimeeonfs Ranad= in0c.4reµasmesf woritthh perreessfeurree.n Aceddsaitmionpallel.y,I nthteu lrinne,atrh eRa paramteentdeernicsyh iisg mhearrkfeodr isna nthdeb rlaansgteed ofs papepcilmiede npsreasnsudreins.c Hreoawseesvewr,i tdhevpiraetisosnu frreo.mA tdhdisi ttiroennadl lbye,ltohwe 1li near tendency is marked in the range of applied pressures. However, deviation from this trend below 1bar andabove6barcannotbeexcluded. Theresultoftheleast-squaresfittotheobtainedpointsis presentedinFig. 4asthestraightline. Materials 2017, 10, 1343 6 of 15 bar and above 6 bar cannot be excluded. The result of the least-squares fit to the obtained points is presented in Fig. 4 as the straight line Materials 2017, 10, 1343 6 of 15 Materiablsa2r 0a1n7,d1 0a,b1o3v4e3 6 bar cannot be excluded. The result of the least-squares fit to the obtained points is6 of15 presented in Fig. 4 as the straight line Figure 4. Mean roughness of the sandblasted copper specimens in dependency on the blasting FigurFei4g.uMre e4a.n Mroeuang hrnoeusgshonfesths eosf atnhde bslaanstdebdlacsotepdp ecropsppeerc imspeencisminends eipne nddepeenncdyeonncyt hoenb tlhaset ibnlgasptirnegs sure. presspurrees.s uTrhee. Tshoeli dso lbidla cbkla clkin lein ree prerperseesnentst st hthee lliinneeaarr lleeaasstt--ssqquuaraerse sfi tf itto ttoh et hoeb toabinteadin peodi nptso ianntds athned the Thesolidblacklinerepresentsthelinearleast-squaresfittotheobtainedpointsandtheresultsaregiven arbesouvletr,sew saurhletes rg eairvpee ignsi vtahebeno pavbreeo,s vwseu,h rweerhiene rpbe aipsr . itsTh theh eep dprearesssshsueurdreeo iinnne bbraaerrp.. rTTehhseee nd dtasashtshhedee dvo anoleun reee porerfepRsreaenfsotesr ntththsee tvhraeelfu evera eolnuf cRee ao sffao mrR ap floer. the retfheer erenfceere snacme spalme.p le. PPrreesseePnnrteteesdedn iitnned FF iiingg uuFrirgeeu 55r,e, t t5hh,e eth SSeEE SMMEM iimm imaaggaegee oo offf a aa r rreeeppprrreeessseeennntttaaattitivivvee es sasamammpplpel leaed aadddidtdioiitntiioaolnnlyaa lrllleyyv rereaevlvse epaaollsus nppdooeuudnn ddeedd partspoafritms opfa icmtepdacatberda asibvreaspivaer tpicalretsic.leTsh. eThoeb toabintaeindeEdD ESDaSn aanlaylsyissiso fofc hcheemmicicaall ccoommppoossiittiioonn ccoonnffiirrmms sthe parts of impacted abrasive particles. The obtained EDS analysis of chemical composition confirms presenthcee porfetsheneceex opfe tchtee dexcpoemctepdo ncoemntpso,nsuencths,a ssuCchu a,sA Cl,ua, nAdl, Oan.dT hOu. sT,htuhse, tahlue maluinmainpaa rptaicrtliecslerse rmemaianina tthe the presence of the expected components, such as Cu, Al, and O. Thus, the alumina particles remain at the surface of the copper specimens. surfaceofthecopperspecimens. at the surface of the copper specimens. Figure 5. SEM image of a representative sandblasted copper sample, the arrows show the alumina particles. On the right, EDS analysis of chemical composition insets. Figure 5. SEM image of a representative sandblasted copper sample, the arrows show the alumina Figure5. SEMimageofarepresentativesandblastedcoppersample,thearrowsshowthealumina p articles. On the right, EDS analysis of chemical composition insets. particles.Ontheright,EDSanalysisofchemicalcompositioninsets. 3 .2. PositronLifetimeResults Thepositronlifetimespectra(LT)weremeasureddirectlyatthesurfacesofsandblastedsamples, asshowninFigure6. Inthisway, thegainedinformationcomesmainlyfromthedepthof26µm, Materials 2017, 10, 1343 7 of 15 3.2. Positron Lifetime Results Materials2017,10,1343 7of15 The positron lifetime spectra (LT) were measured directly at the surfaces of sandblasted samples, as shown in Figure 6. In this way, the gained information comes mainly from the depth of 26 µm, but slight contributions from the whole penetrated depth can also have an impact on it. Only a single butslightcontributionsfromthewholepenetrateddepthcanalsohaveanimpactonit. Onlyasingle lliiffeettiimmee ccoommppoonneenntt eeqquuaall ttoo 112200 ±± 11 ppss wwaass rreessoollvveedd ffrroomm tthhee LLTT ssppeeccttrruumm oobbttaaiinneedd ffoorr tthhee rreeffeerreennccee sample. It corresponds well with the positron lifetime in the well-annealed pure copper [27]. This sample.Itcorrespondswellwiththepositronlifetimeinthewell-annealedpurecopper[27].Thisvalue value ensures that samples were free of defects which could have been introduced during preparation. ensuresthatsampleswerefreeofdefectswhichcouldhavebeenintroducedduringpreparation. FFiigguurree 66.. DDeeppeennddeennccieiess oof fththe emmeaena npopsoistritorno nlifleifteimtime eτ ( a(a),),t hthee fifrirsstt lliiffeettiimmee ccoommppoonneenntt wwiitthh iittss iinntteennssiittyy ((bb)),, aanndd tthhee sseeccoonndd lliiffeettiimmee ccoommppoonneenntt wwiitthh iittss iinntteennssiittyy ((cc)) oonn pprreessssuurree oobbttaaiinneedd ffrroomm tthhee ppoossiittrroonn lliiffeettiimmee ssppeeccttrraa mmeeaassuurreedd ffoorr tthhee ccooppppeerr ssaammpplleess eexxppoosseedd ttoo ssaannddbbllaassttiinngg ffoorr 11 mmiinn wwiitthh aalluummiinnaa ppaarrttiicclleess.. As it was mentioned above, the alumina particles are present in adjoined surface layer. The SEM Asitwasmentionedabove,thealuminaparticlesarepresentinadjoinedsurfacelayer. TheSEM images of the samples’ surfaces, such as those shown in Figure 5, allowed us to roughly estimate that images of the samples’ surfaces, such as those shown in Figure 5, allowed us to roughly estimate about a 10–20% volume fraction is occupied by these particles on the basis of the stereology method that about a 10–20% volume fraction is occupied by these particles on the basis of the stereology [28]. Due to the differences in density between the copper matrix and alumina only about 2–5% of method[28]. Duetothedifferencesindensitybetweenthecoppermatrixandaluminaonlyabout positrons annihilate in these particles. Foster et al., reported a value of the mean positron lifetime in 2–5%ofpositronsannihilateintheseparticles. Fosteretal.,reportedavalueofthemeanpositron Alifle2Otim3 aet inroAoml O tematpreoroamturtee meqpuearla ttuor e15e0q upasl [t2o91].5 0Inp tsu[r2n9,] .NIongtuucrhni, Neto aglu. c[3h0i]e tnaolt.ic[e3d0] tnwooti cleifdettiwmoe 2 3 cliofemtipmoeneconmts,p eo.ng.e,n τt1s ,=e 1.g59., ±τ 8 =ps1,5 τ92 ±= 782p0s ±, τ30 =ps7,2 I02 =± 23.20%p sf,oIr p=ol2y.2c%rysftoarllpinoely Acrly2Ost3a. llineAl O . 1 2 2 2 3 In our LT spectra no long lifetime component is observed, however, in the deconvolution proceduretheLTCof150pswithanintensityof3%,asoriginatingfromannihilationinthealumina Materials2017,10,1343 8of15 particles,wasadded. TwolifetimecomponentswerealsoresolvedinallLTspectra. Weshouldnotice thatthecontributionoftheaddedcomponentof150psdidnotaffectthedeconvolutionprocedure, i.e.,nochangeintheX2valuewasobserved,whichindicatesthemarginalcontributionofthealumina particlestotheLTspectra. InFigure6athemeanpositronlifetimeτversuspressureisshown. Itisdefinedas: τ = I τ +I τ (2) 1 1 2 2 where I and τ are intensities of lifetimes respective resolved from the LT spectra and it gives 1,2 1,2 informationaboutthetotalchangesinthedefectstructureinthesamples. τincreaseswithpressure andsaturatesquicklyabove3barpressureofthealuminastream. Thevalueof τ (Figure6b), relatedtothefirstkindofdefectdetectedinthestudiedsamples, 1 variesfrom168±6psto193±1pswithintensitiesI between76±5%and91±1%. Thehighest 1 valuesofτ andI werenotedforthepressureofthesandstreamequalto3and4bar. Similarvalues 1 1 (164 ± 10 ps and 181 ± 10 ps) were noticed by Hinode et al. [31] in copper exposed to thickness reduction of 3% and 13%. In cold-rolled copper Lynn et al. [32] observed a steady increase of the meanpositronlifetimefrom157ps(forasinglecrystal)to173psasonegoestoprogressivelyfiner grain-sizedmaterial(finally169µm). CampilloRoblesetal.[33]reportedcalculatedpositronlifetime forsinglevacancyinawiderangeofperiodicelements.Inthecaseofcopper,positronlifetimewas foundtodependstronglyonappliedmodelandvariedfrom153psto200ps. Themeasuredvalue of the positron lifetime in a single vacancy defect equals 180 ps according to Schaefer et al. [34]. τ =166±2ps,observedbyCˇížeketal.[35],incoppersamplespreparedbyhigh-pressuretorsion 1 (HPT)wasinterpretedbytheauthorsaspositronstrappedatdislocations. Additionally,thelifetimeof 164pswasshownbyMcKee[36]asrepresentingpositronannihilationsatdislocations. Onthebasis oftheabovediscussiondislocationshavebeenintroducedinthesamplesandblastedat1barwhile, forspecimensprocessedwithhigherpressure,singlevacanciesoccur. Thevalueofτ ,depictedinFigure6c,variesfrom298±18psfor1barto427±23psregistered 2 for 4 bar with intensities I in the range between 9 ± 1% and 23 ± 5%, respectively. This reveals 2 thesecondtypeofdefectpresentinthestudiedsandblastedsamples. Accordingtocalculationsfor nanocrystalline Cu performed by Zhou et al. [37] τ values obtained by us can be interpreted as 2 anoccurrenceofvacancyclusterscontainingfromeightuptoca. 40vacancies. Positronlifetimesin therange290–320psobtainedinHPT-deformedCubyCˇížeketal.[38]weresimilarlyrecognized bytheseauthorsasclusterscontaining7–9vacancies. Thehighestvalueofτ appearstapressure 2 of 4 bar, however, taking into account the variation of presented points the plateau in the range 3–5 bar is possible. The τ increases with pressure to a maximum in the mentioned range while, 2 forapressureof6bar,thewell-markeddecreaseisnoted. Inthiscasetheinverseproportionbetween τ andI isrevealed. I declineswiththepressureachievingaminimumandstartstorise. Giventhis, 2 2 2 concentrationofgenerateddefectsdecreaseswiththepressureincrease,butonlyupto4bar. Above thisvaluethedefectconcentrationincreases. Consequently,thesizeofthevacancyclustersrepresented byτ expandsupto4barandreducesforhigherpressures. 2 Tosumup,twolifetimecomponentsprovetheexistenceoftwokindsofdefectsineachstudied sandblasted sample. In the case of blasting under the pressure of 1 bar they were recognized as dislocations and clusters of about eight vacancies. For other samples, next to single vacancies, largerclusterscontainingupwardsof40vacanciesappeared. 3.3. DefectDepthsProfiles LTspectroscopyisthemosteffectivePAStechnique,however,itistime-consuming. Forinstance, 12hwereneededtoobtainonlyoneLTspectrumwithasatisfyingstatistic. Forthisreasonthedefect depth profiles, requiring multiple measurements along with sequenced etching, were determined Materials2017,10,1343 9of15 Materials 2017, 10, 1343 9 of 15 uussiinngg DDBB ssppeeccttrroommeettrryy.. TThhiiss aapppprrooaacchh aalllloowweedd uuss ttoo rreeflfleecctt tthhee sshhaappee ooff mmeennttiioonneedd ddiissttrriibbuuttiioonnss pprreecciisseellyy,,w whhaatti issv viissiibblleei innF Fiigguurree7 7a asst thheeS Sp paarraammeeteterrv veerrssuusse etctchheeddd deeppthth.. FFiigguurree 77.. SS ppaarraammeetteerr ddeeppeennddeenntt oonn tthhee ddeepptthh ffoorr tthhee wweellll--aannnneeaalleedd ccooppppeerr ssaammpplleess ssaannddbbllaasstteedd wwiitthh aalluummiinnaa ppaarrttiicclleess aatt ddiiffffeerreenntt pprreessssuurreess.. TThhee ddrroopp lliinnee mmaarrkkss tthhee vvaalluuee ooff tthhee SS ppaarraammeetteerr oobbttaaiinneedd ffoorr ppuurree aalluummiinnaa.. TThhee ssoolliidd lliinneess rreepprreesseenntt tthhee bbeesstt fifittss ooff lliinneeaarr oorr eexxppoonneennttiiaall ddeeccaayy ffuunnccttiioonnss ttoo ddeeccrreeaassiinngg ppaarrttss ooff mmeeaassuurreedd pprroofifilleess.. The hatched region tags in Figure 7 represent the value of the S parameter for a reference, well- The hatched region tags in Figure 7 represent the value of the S parameter for a reference, annealed sample, i.e., for the “bulk”, where positrons annihilate in the delocalized state only well-annealedsample, i.e., forthe“bulk”, wherepositronsannihilateinthedelocalizedstateonly (certainly, after annealing, some residual defects remained, but they are neglected). We expect that (certainly,afterannealing,someresidualdefectsremained,buttheyareneglected). Weexpectthatthis this value will be reached at a depth where the interaction of processes at the surface during blasting valuewillbereachedatadepthwheretheinteractionofprocessesatthesurfaceduringblastingfade fade out. The measured points in Figure 7 are represented by black triangles. In this case constant out. ThemeasuredpointsinFigure7arerepresentedbyblacktriangles. Inthiscaseconstantvaluesof values of the S parameter shows us that sequenced etching is neutral for our measurement procedure. theSparametershowsusthatsequencedetchingisneutralforourmeasurementprocedure. The different shape and colour points represent the S parameter values for samples sandblasted ThedifferentshapeandcolourpointsrepresenttheSparametervaluesforsamplessandblastedat at different pressures, as shown in Figure 7. Presented profiles are placed above the bulk region differentpressures,asshowninFigure7. Presentedprofilesareplacedabovethebulkregionpointing pointing to the existence of the layer below the surface occupied by defects introduced by the totheexistenceofthelayerbelowthesurfaceoccupiedbydefectsintroducedbythesandblasting. sandblasting. The close surface region is characterized by the S parameter increasing with depth. This TheclosesurfaceregionischaracterizedbytheSparameterincreasingwithdepth. Thisfeatureis feature is clearly visible for samples treated at a pressure range of 2–6 bar. Then, after achieving the clearlyvisibleforsamplestreatedatapressurerangeof2–6bar. Then,afterachievingthemaximum maximum the S parameter decreases towards the value characteristic of the bulk region pointing to theSparameterdecreasestowardsthevaluecharacteristicofthebulkregionpointingtotheendof the end of the damaged zone. The unusual shape of the obtained profiles and connections between thedamagedzone. Theunusualshapeoftheobtainedprofilesandconnectionsbetweenthemshould them should be highlighted. We supposed rather that the S parameter will be decreased with depth behighlighted. WesupposedratherthattheSparameterwillbedecreasedwithdepthwithagiven with a given type of tendency saved for all samples. For this reason, discussion of a few aspects is type of tendency saved for all samples. For this reason, discussion of a few aspects is required in required in their analysis. theiranalysis. The characteristic decay of the S parameter values above ca. 50 µm is well understood. With the ThecharacteristicdecayoftheSparametervaluesaboveca. 50µmiswellunderstood. Withthe depth increase fewer defects are generated during sandblasting and their concentration is the highest depthincreasefewerdefectsaregeneratedduringsandblastingandtheirconcentrationisthehighest at the surface. The total depth of the damaged zone is extended from 140 to 780 µm and strictly at the surface. The total depth of the damaged zone is extended from 140 to 780 µm and strictly depends on the applied pressure. This is shown in Figure 8. dependsontheappliedpressure. ThisisshowninFigure8. Materials2017,10,1343 10of15 Materials 2017, 10, 1343 10 of 15 FFiigguurree 88.. TToottaall tthhiicckknneessss ooff tthhee ddaammaaggeedd zzoonneess oobbttaaiinneedd ffrroomm PPAASS mmeeaassuurreemmeennttss ddeeppeennddeenntt oonn tthhee pprreessssuurree aapppplliieedd iinn tthhee ssaannddbbllaassttiinngg ooff ccooppppeerr.. The maximum, at the depth of about 50 µm (Figure 7), is a new interesting feature which was Themaximum,atthedepthofabout50µm(Figure7),isanewinterestingfeaturewhichwas not observed in distributions registered for copper exposed to other treatment processes [16,17]. For not observed in distributions registered for copper exposed to other treatment processes [16,17]. analysis of the near surface region the depth up to 100 µm was additionally presented in Figure 9. Foranalysisofthenearsurfaceregionthedepthupto100µmwasadditionallypresentedinFigure9. On closer inspection this maximum also depends on the pressure, and it ranged from 20 to 50 µm for Oncloserinspectionthismaximumalsodependsonthepressure,anditrangedfrom20to50µmfor the increasing pressure. We accept the explanation that the alumina particles which are pushed into theincreasingpressure. Weaccepttheexplanationthatthealuminaparticleswhicharepushedinto the samples are responsible for this maximum. The S parameter value for the alumina, measured in thesamplesareresponsibleforthismaximum. TheSparametervalueforthealumina,measuredin independent experiments, is higher in comparison with its value obtained for the pure well-annealed independentexperiments,ishigherincomparisonwithitsvalueobtainedforthepurewell-annealed copper, but lower than the S parameter value for the deeply-deformed copper, after 80% thickness copper,butlowerthantheSparametervalueforthedeeply-deformedcopper,after80%thickness reduction. Thus, positrons which are implanted into the sandblasted copper sample with pushed reduction. Thus, positrons which are implanted into the sandblasted copper sample with pushed alumina particles can annihilate in deformed copper, but also in these particles. Finally, the measured aluminaparticlescanannihilateindeformedcopper,butalsointheseparticles. Finally,themeasured S parameter, which is averaged over all possible annihilation states, can be lower than its value for Sparameter,whichisaveragedoverallpossibleannihilationstates,canbelowerthanitsvalueforthe the deformed copper. The alumina particles are present exclusively in the layer ca. 50 µm below the deformedcopper. Thealuminaparticlesarepresentexclusivelyinthelayerca. 50µmbelowthelevel level where only deformed copper is extended. whereonlydeformedcopperisextended. Solid lines in Figure 7 represent the best fits of the linear and exponential decay functions to the SolidlinesinFigure7representthebestfitsofthelinearandexponentialdecayfunctionstothe parts of profiles in the regions between maximum and the end. In case of distributions representing partsofprofilesintheregionsbetweenmaximumandtheend. Incaseofdistributionsrepresenting blasting at 1–3 bar, the linear tendency was observed. In turn, the exponential character can be blastingat1–3bar,thelineartendencywasobserved.Inturn,theexponentialcharactercanbeassigned assigned to the rest of profiles. The former tendency was registered in machined copper [16], while totherestofprofiles.Theformertendencywasregisteredinmachinedcopper[16],whilethelatterwas the latter was noted in copper samples exposed to sliding and cutting [17]. Various types of S notedincoppersamplesexposedtoslidingandcutting[17].VarioustypesofSparameterdistributions parameter distributions from one treatment process were also obtained for austenitic stainless steel fromonetreatmentprocesswerealsoobtainedforausteniticstainlesssteelexposedtoslidingunder exposed to sliding under varied dry load conditions [39]. varieddryloadconditions[39].
Description: