UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

XPS and corrosion studies on zinc phosphate treated surfaces of aluminum Heung, Wai Fan 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0011.pdf [ 1.99MB ]
JSON: 831-1.0061692.json
JSON-LD: 831-1.0061692-ld.json
RDF/XML (Pretty): 831-1.0061692-rdf.xml
RDF/JSON: 831-1.0061692-rdf.json
Turtle: 831-1.0061692-turtle.txt
N-Triples: 831-1.0061692-rdf-ntriples.txt
Original Record: 831-1.0061692-source.json
Full Text

Full Text

XPS AIJD CORROSION STUDIESON ZINC PHOSPHATE TREATED SURFACES OF ALUMINUMbyWAJ FAN HEUNGB.Sc., University of British Columbia, 1991A THESIS SUBMITTED IN PARTIAL FULFULLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF CHEMISTRYWe accept this as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOctober 1993© Wai Fan Heung, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.__________________Department of_______________The University of British ColumbiaVancouver, CanadaDate r?7.’DE-6 (2/88)AbstractThe work in this thesis deals with an investigation of zinc phosphate (ZPO) treatedsurfaces of 7075-T6 aluminum alloy. X-ray photoelectron spectroscopy (XPS) andcorrosion tests are used to characterize the treated surfaces and to examine their corrosionprotection performance.The coating processes are performed by immersing polished alloy surfaces into 10wt. % of ZPO suspension in water adjusted at various pH conditions. Biased XPS isapplied to discriminate the physically trapped and chemically absorbed zinc compounds inthe ZPO treated surfaces. Ultrasonic rinsing in distilled water for the treated surfaces isfound important in the sample preparation procedure so as to remove any physicallytrapped compounds from the surfaces.Five different pH conditions (pH3.5, 5.0, 6.6, 10.5, and 13.0) are studied in thiswork. Acetic acid and sodium hydroxide solution are used to adjust the required pHs forthe ZPO solutions. The treated surfaces are studied by angle dependent XPS (ADXPS) toobtain chemical information at different probed depths. ZPO is found to be an effectivecoating compound for the 7075-T6 aluminum surface. At pH=6.6 and 3.5, the coatingsfound on the treated surfaces are believed to be respectively a ZnOx-AlOx mixed materialand a thin ZPO-like compound. In alkaline and weakly acidic conditions (pH=13.0, 10.5and 5.0), the coatings formed on the surfaces are enhanced and have stmctures with mixedZnOx-AlOx-ZPO materials.Weight loss measurements, atomic absorption spectrometry (AAS) and scanningelectron microscopy (SEM) are involved in the corrosion studies. 3.5 % NaCl solutionsare used as the corrosive environments. Surfaces before and after the corrosion tests are11compared to examine the corrosion protection performance of each surface. As judged byXPS, the surface prepared at pH=13.O is likely to provide the best corrosion controlamong the five treated surfaces. Dissolution of aluminum from the alloy is observed in thecorrosive environment. The coating is believed to play a role as a physical barrier tosuppress the corrosion attack on aluminum.111Table of ContentsAbstract.iiTable of Contents ivList of Tables viiList ofFigures viiiAcknowledgments xiChapter 1 Introduction 11.1 General 11.2 Chemical Conversion Coating 41.3 Surface Analytical Techniques Applied to Coating and CorrosionStudies 51.4 Aims of this Research 7Chapter 2 Experimental Methods 102.1 X-ray Photoelectron Spectroscopy 112.1.1 Introduction 112.1.2 Basic Concepts 12(A)Principles 12(B) Surface Sensitivity of XPS 212.1.3 Chemical Analyses with XPS 23(A) Qualitative Analysis 23(B) Quantitative Analysis 282.1.4 AngleDependentXPS 302.1.5 Bias Technique Applied on XPS 33iv2.1.6 Instrumentation of XPS .35(A) Ultra-High Vacuum 35(B) Excitation Source 38(C) Sample Handling 41(D) Electron Energy Analyzer 41(E) Detector 462.2 Scanning Electron Microscopy 482.2.1 Background 482.2.2 System Used 482.3 Weight Loss Measurements 522.4 Atomic Absorption Spectrometry 53Chapter 3 Coating and Corrosion Studies on ZPO treated Aluminum Surfaces 543.1 Coating Studies on Zinc Phosphate Treated Aluminum Surfaces 543.1.1 Sample Preparations 543.1.2 XPS Measurements 543.1.3 Results and Discussion 57(A)The Aluminum Control Panel (Sample A) 57(B) The Aluminum Panel Treated in Natural ZPO Solutionat pH=6.6 (Samples B and C) 60(C) The Aluminum Panel Treated in ZPO Solution atVarious pH Values (Samples D to G) 66(1) Strongly Alkaline Coating Solution(pH= 13.0; Sample D) 72(2) Moderately Alkaline Coating Solution(pH=1O.5; Sample E) 73V(3) Moderately Acidic Coating Solution(pH=5.0; sample F) 73(4) Strongly Acidic Coating Solution(pH=3.5; sample G) 753.2 Corrosion Studies on Zinc Phosphate Treated Aluminum Surfaces 763.2.1 Sample Preparations 763.2.2 Results and Discussion 80(A)Part I: Initial Studies 80(B) Part II: Further Studies 86Chapter 4 Concluding Remarks and Future Work 904.1 Concluding Remarks 904.2 Future Work 91References 93viList of TablesTable 1.1 Features of various surface analytical methods 8Table 2.1 A table of some characteristic X-ray lines 17Table 2.2 A summary of the atomic orbital nomenclature 25Table 2.3 The Auger parameters for zinc compounds 29Table 3.1 Sample descriptions 55Table 3.2 Elemental composition and atomic ratios for blank Al surface(sample A) with varying take-off angle 59Table 3.3 Elemental composition and atomic ratios for ZPO treated Alsurface (sample C) with varying take-off angle 64Table 3.4 A list of binding energies (in eV) of zinc, phosphorus on ZPOtreated aluminum surfaces and the ZPO reference compound 68Table 3.5 Atomic ratios for samples C, D, E, F, G, and the ZPO referencecompound with varying take-off angle 71Table 3.6 Zn/Al ratios for samples A, Al, C and Cl with varying take-offangle 81Table 3.7 Qualitative observations from XPS for samples after immersion inNaCl solution 87viiList of FiguresFigure 1.1 The natural oxide layer formed on an aluminum surface 2Figure 1.2 Difference between metallic and oxide type protective layers onaluminum 2Figure 1.3 Comparison between bulk and surface analysis 6Figure 2.1 A schematic diagram of the photoemission process 13Figure 2.2 The experimental setting for XPS 14Figure 2.3 (a) An illustration of X-rays produced from electronic transitions.(b) Approximated energy distribution of unmonochromatized Alradiation, showing the strong characteristic KcL and K13 lines andthe Bremsstrahlung radiation 16Figure 2.4 Reference levels for a metal sample and a spectrometer 18Figure 2.5 A schematic diagram of the Auger process 20Figure 2.6 The compilation by Seah and Dench of measurements of theinelastic mean free path, ?, for different elements 22Figure 2.7 XPS survey scan spectrum of copper, with an insert of the CuLMM Auger series 24Figure 2.8 Spin-orbit splitting for the Zn 2p photoelectron peaks 25Figure 2.9 High resolution XPS spectrum of the Al 2p photoelectron peakfrom an aluminum alloy with naturally formed oxide layer 27Figure 2.10 Illustration of the principle of angle dependent XPS (ADXPS) 31Figure 2.11 Theoretical angular dependent curves for a flat clean surface coatedwith an overlayer 31Figure 2.12 Zn 2P3/2 photoelectron peaks from a zinc phosphate treatedaluminum sample. Curve 1 is obtained from the grounded sample.viiiCurve 2 is obtained from the sample with -94 V biased potentialapplied on the sample and shifted back by 94 eV after measuring 34Figure 2.13 A schematic indication of the Leybold MAX 200 spectrometer (a)top view (b) side view 36Figure 2.14 Components of the MAX 200 analysis chamber 37Figure 2.15 The pumping system of the XPS spectrometer 39Figure 2.16 A schematic diagram of the Al and Mg dual anode X-ray source 40Figure 2.17 The MAX 200 system: sample handling assembly 42Figure 2.18 The analysis and transfer chambers of the spectrometer in MAX200 system 43Figure 2.19 A schematic diagram of a concentric hemispherical analyzer (CHA) 44Figure 2.20 (a) A cutaway view of a microchannel plate, (b) a schematicdiagram of a microchannel plate assembly 47Figure 2.21 A schematic diagram of a scanning electron microscope (SEM) 49Figure 2.22 The illuminating / imaging system of a SEM unit 50Figure 3.1 A flow chart summarizing the sample preparation procedure for thecoating process 56Figure 3.2 XPS survey spectrum for sample A, a blank Al surface 58Figure 3.3 XPS survey spectrum for sample C, ZPO treated Al surfaceprepared at natural pH 61Figure 3.4 High resolution Zn 2P3/2 spectra for: (a) sample C, 9 = 90°; (b)sample C, 0 = 30°; (c) sample B, 9 = 90°; (d) sample B, 0 = 30°; (e)sample B, 0 = 90° after bias potential; (f) sample B, 9 = 30° afterbias potential 62Figure 3.5 Proposed surface morphologies for (a) sample B, (b) sample C, (c)samples D & E, (d) sample F and (e) sample G 65ixFigure 3.6 XPS survey spectrum for sample D, ZPO treated Al surfaceprepared at pH=13.0 67Figure 3.7 pH effect on the atomic ratios of elements on the ZPO treated Alsurfaces (a) before and (b) after corrosion tests (values obtained attake-off angle equal 900) 70Figure 3.8 Sample preparation steps for Part I (initial studies) of corrosionstudies 77Figure 3.9 Sample preparation steps for Part II (further studies) of corrosionstudies 79Figure 3.10 SEM micrographs of: (a) sample A, blank aluminum; (b) sampleAl, sample A after corrosion; (c) sample C, ZPO treated aluminum;(d) sample Cl, sample C after corrosion 83Figure 3.11 Photographs taken from surfaces after the 5 hours corrosion tests,(a) sample A, blank Al; (b) sample A2, sample A after corrosion;(c) sample D, ZPO treated surface prepared at pH=1 3.0; (d) sampleD2, sample D after corrosion 89xAcknowledgmentsI would like to thank my supervisor, Professor K.A.R. Mitchell, for his sponsoringthis work and his helpful comments and advice on this thesis. I am especially grateful toDr. P.C. Wong, Professor Y.P. Yang and Professor M.Y. Zhou for their activeinvolvement, and helpful discussions in this work. I also thank Dr. Y.S. Li for teachingme the sample polishing procedure. I also acknowledge the staff from the electricalworkshop for maintaining the instruments in working order and the staff from themechanical workshop for preparing the necessities for this research. I also appreciate thehelp from the microscope laboratory in the Department of Metals and MaterialsEngineering for the SEM investigation. I also acknowledge the financial support providedby the Defence Research Establishment Pacific (DREP) and thank Dr. Terry Foster for hiscomments on this work.In addition, I am very grateful to my group members, Y.L. Leung, Y.M. Wang, W.Liu, K.C. Wong, D-T Vu Grimsby, Bernie Flinn and Harman Cheng for their “alwaysavailable” help in the laboratory, and my friends for their prayers and encouragement inthese two years. Finally, I would like to thank my parents and my brother for theirsupport and encouragement.xChapter 1 Introduction1.1 GeneralMetals have played an important role in the development of civilization, and inrecent times aluminum, copper, zinc, iron and steel have been very widely used. Amongthese, aluminum has a vital role in the light metal industry. Its applications in everyday lifeinclude household usage as well as uses in mechanical apparatus, the chemical and foodindustry, architecture, transportation and so forth [1.1].The main properties on which the applications of aluminum are based are its lowdensity of approximately 2.7 g/cm3 [1.21, its high mechanical strength achieved by alloyingand heat treatments, and the relatively high corrosion resistance of the metal. The chiefalloying constituents added to aluminum are copper, magnesium, silicon, manganese,nickel and zinc. Generally, the higher the purity of aluminum, and its alloys, the greater thecorrosion resistance [1.31. The excellent corrosion resistance of aluminum is due to itsaffinity for oxygen; a very thin oxide film covers the surface as soon as a freshly-cut pieceof the metal is exposed to the atmosphere [1.41.In general, this air-formed film is believed to be porous and amorphous, with theouter surface being a hydrated aluminum oxide (Figure 1.1). Its thickness ranges from 1 to3 nm. This firmly adhered oxide film is insoluble in water, and many other chemicals, andhelps protect the underlying metal from foreign attack. Breakdown of the oxide film canresult from mechanical rupture, or from chemical attack by anionsIPores in the oxide layerlevel of aluminum surfacebefore layer applicationmetallic deposit layerFigure 1.2 Difference between metallic and oxide type protective layers onaluminum.AluminumNatural oxide layerFigure 1.1 The natural oxide layer formed on an aluminum surface.oxide layer/2such as chloride ions. A consequence of the film breakdown is corrosion [1.5]. Corrosionis generally defined as a gradual wearing away of metal by a chemical or electrochemicaloxidizing process [1 .6]. In normal situations, repair of the oxide film is immediate and canbe accompanied by the oxygen reduction reaction or the hydrogen evolution reaction asshown here:Al —* + 3 e (Anodic reaction) (1.1)02 + 4e —> 202- (1.2)} (Cathodic reaction)2H0 + 2e —> 20W + H2 (1.3)Therefore, aluminum oxide or aluminum hydroxide can be formed. However, in thepresence of aggressive ions, this repassivation process is hindered , and soluble complexions, e.g. Al(OH)2C1,may be formed [1.7] with consequent dissolution and thinning ofthe metal component.A obvious way to prevent the aluminum corrosion is to avoid oxidizing oraggressive species coming in contact with the metal surface, on which the corrosion isalways initiated. This may be done by applying a coating layer to the surface. Theprotection capacity of a coating (i.e. its service life), depends both on its physical thicknessand on its chemical durability in a particular environment. The types of coating layersapplied on metal surfaces can be divided into two main categories: metallic coatings andoxide coatings, as shown in Figure 1.2 [1.8]. In the first type, a metal or metal alloy willbe sprayed, cladded or deposited on to the aluminum surface for corrosion protection.However, this layer sometimes suffers problems of mechanical strength and elongation,3and results in cracking. The oxide-type coating layer is usually used as a base for organicfinishes. Chemical methods (e.g. chemical conversion treatments), or electrolytic methods(e.g. anodizing), are applied to the metal to produce oxide-type coating layers whosethicknesses may vary from 2 to 100 nm, and the compositions may show different oxidesand other compounds, depending on the reaction conditions. In this thesis, chemicalconversion treatments will be emphasized.1.2 Chemical Conversion CoatingIn forming a chemical conversion coating, the metal surface is immersed into acoating solution, from which the coating compounds react and incorporate with the metalsurface. Although an oxide conversion coating can be used without further treatment, itsfunction is normally to act as an undercoat and a base for organic finishes [1.9, 1.101, inorder to improve the adhesion of paints and adhesives.The main attraction of finishes obtained by chemical conversion is the economyand the speed with which they can be produced. Compared with anodizing treatments,chemical conversion treatments require a plant with simple construction and relatively lowconsumption of electric power; also coatings can be produced rapidly on a large numberof articles. All these factors make for an economic process [1.9].Zinc chromate is one of the most extensively used and efficient coating compoundsfor aluminum [1.10, 1.111. This chromate coating forms a continuous layer which acts asa barrier to reduce the active surface area on base metal, and delay the transportation ofoxidizing and aggressive species. However, chromates have been found to cause irritationof the respiratory tract, and produce lung cancer in workers employed in chromium4manufacturing plants [1.12, 1.13]. Because of these toxicological concerns, newchromate-free coatings are being evaluated for effective corrosion control [1.12, 1.14-1.15]. Among several non-toxic and anti-corrosive coatings so far developed, users andmanufacturers have mainly focused their attention on zinc phosphate.Zinc phosphate is an established coating treatment for steel [1.16-1.19]. Thecoating on steel, resulting from the formation of a mixed iron phosphate and zinc ironphosphate coating layer, functions as an intermediate layer to improve the adhesion ofpaint, thereby increasing the corrosion resistance of the painted products [1.20, 1.211.The phosphate coatings are used in highly corrosive applications such as car bodies,household appliances, structural steel and the like [1 .22]. Because of its anti-corrosionproperty, zinc phosphate is being considered for application to aluminum as well.1.3 Surface Analytical Techniques Applied to Coating and CorrosionStudiesAnalytical investigations of a coating layer on a metal surface can give knowledgeabout the composition and structure of the layer, and possibly about the mechanism of thecoating process. Studies of the corrosion protection performance of a treated surfaceprovide an evaluation of the anti-corrosion ability of the coating layer. All this informationhelps to determine the best coating for a particular metal substrate. Both coating andcorrosion processes on a metal are initiated at the outermost surface layers. Frequently,the chemical and physical properties of the surface region of a solid are different fromthose in the bulk [1.23]. As a result, surface sensitive techniques are required in order toprovide the most detailed knowledge. Figure 1.3 [1.24] compares the bulk and the surfaceanalytical approaches. By definition, surface analysis is concerned with the analysis of the5Bulk analySiSFigure 1.3 Comparison between bulk and surface analysis.X 30000x 300000X 3000000Thin film analysisX 30000000Surface analysis6elemental composition, the chemical state, the surface structure and morphology of theoutermost atomic layers of a solid. The most commonly used surface analytical methodsinclude X-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES),and secondary ion mass spectroscopy (SIMS). Their major characteristics are listed onTable 1.1 and their sampling depths are especially in the 1 to 5 nm range.1.4 Aims of this ResearchThe motivation for the work in this thesis originated from a contract with theDefence Research Establishment Pacific (DREP) to study various corrosion protectiontreatments on aluminum alloys. The present investigation concentrates on the coating andcorrosion protection performance of zinc phosphate, as a possible replacement for zincchromate, on 7075-T6 aluminum alloy, which has long been used in the aircraft andautomobile industries because of its favourable strength-to-weight ratio and high durability[1.27].The primary goal of the work is to determine whether zinc phosphate can beeffectively coated (i.e. chemically absorbed) into the aluminum alloy. The treated surfacesare characterized by XPS, in order to investigate the compositions and morphologies ofthe coating layers, with a special emphasis on determining the amounts of zinc andphosphate present.The next objective is to establish optimal conditions for coating maximum amountsof zinc and phosphorus by adjusting the acidities and basicities of the coating solutions. Inaddition, XPS is used in attempts to learn about the coating mechanisms of zinc phosphate7Table 1.1 Features of various surface analytical methodsAnalytical methods Brief description Information RefX-ray photoelectron Core level photoelectrons Composition 1.24spectroscopy (XPS) emitted from surface atoms by and chemicalX-ray excitation stateAuger electron Study Auger electrons emitted Composition 1.24spectroscopy (AES) from surface atoms afterexcitation by X rays orenergetic electronsSecondary ion mass Mass analysis of sputtered ions Composition 1.25spectroscopy (SIMS) resulting from bombardment byenergetic primary ions8onto the aluminum alloy and the morphologies of the treated surfaces.The third objective of the study is to evaluate the corrosion protectionperformance of these treated surfaces. No paints are applied on these treated surfaces, sothat the corrosion protection ability of the zinc phosphate coating layer can be studiedwithout other influencing factors. A 3.5% sodium chloride solution is used to performimmersion corrosion tests for these surfaces [1.28], over periods ranging from 2 to 5hours. The surfaces are compared before and after the corrosion tests. In order to gainmaximum information on the process, other analytical methods are employed for theevaluation, including scanning electron microscopy (SEM), weight loss measurements andatomic absorption spectrometry (AAS).9Chapter 2 Experimental MethodsFour main methods are used in this work. They are X-ray photoelectronspectroscopy (XPS), scanning electron microscopy (SEM), weight loss measurements andatomic absorption spectrometry (AAS). XPS is used to characterize the sample surfacesboth after the coating processes and after the corrosion tests. Samples are prepared in anatmospheric environment and are introduced into the spectrometer for analysis. Detailsabout the principles and instrumentation of XPS will be discussed in the following section.The evaluation of the corrosion protection performance of the treated surfaces is done bySEM, weight loss measurements and AAS. Scanning electron micrographs at 4000magnification are taken from surfaces before and after corrosion tests. Weight lossmeasurements and AAS are employed to determine the corrosion rate of each sample andto detect the dissolution of aluminum in the corrosive environment. The details for thesetwo methods will also be discussed in this chapter.102.1 X-ray Photoelectron Spectroscopy2.1.1 IntroductionX-ray photoelectron spectroscopy (XPS) is generally regarded as an importanttechnique for surface characterization and analysis. This technique, also called ESCA(electron spectroscopy for chemical analysis), provides compositional analysis (except forhydrogen and helium), for the topmost layers of a solid surface. Chemical bonding stateinformation can also be provided [2.11.XPS has its origin in investigations of the photoelectric effect, which wasdiscovered by Hertz in 1887 [2.2]. He produced photoelectrons by illuminating matter byultraviolet light. Einstein, in 1905, established the relationship between the kinetic energyof a photoelectron, the binding energy in the solid, the radiation energy and the workfunction of a solid [2.3]. Before World War I, Moseley, Rawlison and Robinson carriedout experiments in this new field. After the war, Robinson and Maurice de Broguecontinued the work separately and gave rapid development for the subject. However, thatwork finished, without any major advance in resolution or sensitivity, at the outbreak ofWorld War II.Starting from the late 1940’s, decisive developments were underway in Sweden.Siegbahn developed the technique of 3-ray spectroscopy to high levels of precision andthen realized that X-rays could also be used for the excitation. improvements in resolvingpower were obtained by utilizing a double-focusing spectrometer. Later, Siegbahn’s groupsuccessfully observed photoelectron peaks and thus was able to measure the electronbinding energy more accurately. This group also observed the chemical shift effects incore-level binding energies, and continued to develop the whole field of electron11spectroscopy from 1955 to 1970. Commercial instruments started to appear in the mid1960’s. In 1972, Brundle and Roberts performed XPS studies on carefully preparedsurfaces under ultra-high vacuum, and that work established XPS as a surface analyticaltechnique [2.4].2.1.2 Basic Concepts(A) PrinciplesXPS depends on photoemission. The process is shown in Figure 2. 1. If a photonof energy ho entirely transfers its energy to an electron with binding energy Eb (where ho> Eb), for example in the core level of an atom, the kinetic energy (Ek) of an emittedphotoelectron will be approximated by:Ek ho - Eb (2.1)Since different elements have different sets of electronic binding energies, measurementsof the kinetic energies of photoelectrons from a sample can provide elementalidentification.An overview of the experimental setting for XPS is shown in Figure 2.2. X-rayswith fixed energy ho bombard onto a sample and result in the emission of photoelectrons.These photoelectrons are then directed into the electron analyzer for energy analysis andtheir intensities are measured by a detector. All these data are processed by a computersystem and an XPS spectrum, usually plotted as intensity against binding energy, is122shoEmitted photoelectronFigure 2.1 A schematic diagram of the photoemission process.13.Eleciron analyserFigure 2.2 The experimental setting for XPS.Ultra-high vacuumpump systemII IIX-ray source.—BE. —K.E.Samplehu[jtata processing unitphotoelectrons.P7Zz7 Detector1W14generated. As shown from the figure, XPS analysis is performed under a ultra-highvacuum system (section 2.1.6 (A)).The X-rays used in XPS are characteristic emission lines generated from an anodebombarded by high energy electrons emitted from a heated filament. The energies of thesecharacteristic lines are dependent on the electronic transitions within the anode atoms(Figure 2.3(a)) [2.5]. Besides these characteristic lines, a continuous spectrum, known asBremsstrahlung radiation, is also produced (Figure 2.3(b) [2.6]). In choosing a suitableanode material for XPS, the line width of the characteristic emission line is in majorconsideration. The line width determines the energy resolution of a spectrum (section2.1.6 (D)). Another consideration is that the anode material must be a good conductor,otherwise heat transfer from the point of impingement of the electrons to stimulate the Xrays will be insufficient and the anode might melt. Table 2.1 [2.7] lists the energies andthe widths of some characteristic X-ray lines of some materials. The most commonly usedX-ray lines in XPS are the Ka lines of Al and Mg which have line widths of 0.85 eV and0.70 eV and energies of 1486.6 eV and 1253.6 eV, respectively.For a metallic solid, the binding energy (Eb) of an electron is referenced to theFermi level, which is defined as the highest occupied energy level. The Fermi level isbelow the vacuum level by the work frmnction (Wm), i.e. the energy which must besupplied for an electron to escape from the metal (Figure 2.4). Therefore, the kineticenergy (Ek) of an electron emitted from a metal surface is:Ek=hu-Eb-Wffl (2.2)15>-I(I)zwI—zLU>I.LUFigure 2.3 (a) An illustration of X-rays produced from electronic transitions.(b) Approximated energy distribution of unmonochromatized Alradiation, showing the strong characteristic Kct and K13 lines andthe Bremsstrahlung radiation.NKaK/3LaMa(a)(b)—.‘ ra7IK15.0A 1.4 AI I I I4.0 2.0 1.6 1.0-8 AENERGY (Key)16Table 2.1 A table of some characteristic X-ray lines.Line Energy (eV) Width * (eV)YM( 132.3 0.47ZrM 151.4 0.77NbMC 171.4 1.21MgKcL 1253.6 0.70MKc 1486.6 0.85SiKc 1739.5 1.00YLct 1922.6 1.50ZrLa 2042.4 1.70* Defined as the fhll width at half maximum height of the line17Sample ear;Sample ILi SpectrometerSpectrometerho--spWm iFermi level ,j. Fermi levelFigure 2.4 Reference levels for a metal sample and a spectrometer.During XI’S measurement, a metallic sample is in electrical contact with thespectrometer. Consequently, their Fermi levels are equal. However, their vacuum levels,and so their work fhnctions, are not the same. The resultant kinetic energy (Ek’) of aphotoelectron measured in the spectrometer will be equal to:Ek’ = ho - Eb - W (2.3)where W, is the work function of the spectrometer. is determined by calibratingwith standard samples, and it is constant for different measurements as long as thespectrometer is not exposed to atmospheric pressure. A common reference photoelectronline for calibration is Au 4f712 at a binding energy of 84.0 eV.In addition to the photoelectrons emitted from the above process, Auger electronscan also be emitted due to relaxation of the energetic ions left after the photoemission. Inthe Auger process, shown schematically in Figure 2.5, an electron from the L1 shell fallsinto the inner K shell vacancy created by the initial X-ray irradiation; the energy released(EK - EL1) is transferred to an electron in the L23 shell, which is emitted as an Augerelectron. The final state of the atom is doubly ionized. The kinetic energy (E KL1L2,3)ofthis Auger electron is approximately equal to:E KL1L2,3 = EK - EL1 - E3 (2.4)and this kinetic energy is basically a function of the atomic energy levels involved, so thatmeasurements of the Auger energies from a surface give direct elemental identification. InXPS analysis, photon excited Auger electron spectra are frequently present, and can helpthe identification of chemical composition in the sample.19Auger electronL23 L23 • S S S •—O--•—-— L23L1K -0- KL1 0 L1KInitial state Excitation andemissionFigure 2.5 A schematic diagram of the Auger process.Final state20(B) Surface Sensitivity of XPSThe reason why XPS is a surface sensitive analytical technique is related to theinelastic mean free path of electrons in the low energy range (e.g. 2 keV or less) in solids.Electrons in this energy range traveling through a material have a relatively highprobability of experiencing inelastic collisions, therefore causing an energy loss. Theinelastic mean free path (IMFP, ) is defined as the average distance traveled by anelectron without losing energy.The intensity of electrons, I, traversed through a sample with thickness t, is equalto:I = J0 (1 - exp (-t / A)) (2.5)where J0 is the intensity of electrons originated from the bulk. It is observed that 63% ofthe electron intensity will emanate from a depth of A, 87% from a depth of 2A, and 95%from a depth of 3A. A common expression of the sampling depth in XPS is 3A for thenormal exit direction, that is the depth from which 95% of signal intensity arises.A is a fi.inction of the material and the electron energy. Calculations of A have beenmade [2.8-2.11], and the compilation of experimental values by Seah and Dench [2.121 isgiven in Figure 2.6, in terms of atomic monolayers as a function of electron kinetic energy.For photoelectrons with kinetic energy between 100 to 1000 eV, A typically ranges fromabout 2 to 8 monolayers. This ensures that the photoelectrons in a peak must originateclose to the surface in order that they can escape into the vacuum and be detected withoutloss in energy. Strong inelastic scattering therefore gives XPS its surface sensitivity.211 oco—I I I..100- .a)>.C ..o ..E.• •••..10 -•• \ •• • —.••. ..••.•: : • •• •• • • •••.••.1I••S.••..• • ••• . .••1- —•I ,,..... I I I1 10 100 1000Energy (eV)Figure 2.6 The compilation by Seth and Dench of measurements of theinelastic mean free path, , for different elements.222.1.3 Chemical Analyses with XPS(A) Qualitative AnalysisTo characterize the surface chemistry of a sample under investigation, the first stepto be taken is to identiiy’ the elements present. This is achieved by recording a survey orwide scan XPS spectrum. A typical survey scan spectrum of a copper sample excited byMgKc radiation is shown in Figure 2.7 [2.13]. Photoelectron peaks, arising from directexcitation of core-level electrons, are labeled with 2p 1/2’ 2P3/2 3 s, and 3p. In general, thecore photoelectron peaks are the narrowest features in a XPS spectrum. Auger electronpeaks in the LMM series are also indicated. Identification of these peaks can be madewith the aid of data tabulated in Handbooks [2.13, 2.14].In XPS spectra, peaks generally appear with an increasing background. In partthis is due to photoemission by Bremsstrahlung radiation. In addition, the presence of ahigh binding energy tail is a consequence of electrons suffering energy loss during theinteraction with the sample. This latter phenomenon causes the background in a XPSspectrum to have a stair-step appearance.Photoelectron core peaks are labelled with the principal quantum number, n, (e.g.values 1, 2, 3,...) and the angular momentum quantum number, 1, (e.g. values 0, 1, 2, 3,or respectively called s, p, d, or 0 (Table 2.2) [2.15]. A characteristic feature of the p, d,and f peaks is the spin-orbit splitting, for which Figure 2.8 [2.161 shows the example oftheZn 2p photoelectron peaks. The interaction of the spin and orbital electron motionsresults in two different energy levels, although this does not occur for s orbitals, which aresymmetrical. The coupling effect is designated in XPS spectra by the subscripts 1/2 and23Figure 2.7 XPS survey scan spectrum of copper, with an insert of the CuLMM Auger series.Copper, CuMg Kcz-J‘43cCu 3sBINDING ENERGY (eV)24Table 2.2 A summary of the atomic orbital nomenclature.Principal quantum number (n):Angular momentum quantum number (1):ci)1, 2, 3, 4, 5K,L,M,N,O0, 1, 2, 3s, p, d, fBINDING ENERGY (eV)Subscript: 1+1/2 or!- 1/2Example: 2 P1/2 2 P3/21065 1055 1045 1035 1025 1015Figure 2.8 Spin-orbit splitting for the Zn 2p photoelectron peaks.253/2 for p orbitals, 3/2 and 5/2 for the d orbitals, and 5/2 and 7/2 for f orbitals; the peakintensity ratios are 1:2, 2:3 and 3:4 respectively.Chemical bonding or chemical state information for a particular element can beobtained from high-resolution spectra. Figure 2.9 shows a high resolution spectrum of Al2p photoelectron peaks from an aluminum alloy with a naturally formed oxide layer. Thetwo Al 2p peaks result from different chemical environments; that at 71.5 eV is identifiedas metallic aluminum, while that at 74.1 eV is identified as aluminum oxide. This changein binding energy is a chemical shift effect, whereby the binding energies of the inner coreelectrons are influenced by changes in the valence electron environment. The higherbinding energy peak is assigned to the oxide because of the net positive charge in thosealuminum atoms which are bonded to oxygen. The reduced electrostatic shielding resultsin these Al 2p electrons being more strongly attracted to the nucleus, and so a higherbinding energy is observed for this Al 2p photoelectron peak.On the other hand, a possible error in identification of a peak position may resultfrom any electrostatic charging of a poor conductive sample. It is noteworthy thatcharging effects occur with Auger electron lines, as well as with photoelectron lines, andthe use of the Auger parameter (cL) [2.17] can help minimize possibilities for mistaking anartifically induced shift in binding energy from charging as being due to a chemical shiftdependent on chemical environment. The Auger parameter can be defined as:(2.6)where E is the kinetic energy of an appropiate Auger electron peak from a particularatom, and Ekp is the kinetic energy of a photoelectron from the same atom. Since thevalues of Eka and Ekp have a common reference energy, measured values of the Auger26zzFigure 2.9 High resolution XPS spectrum of the Al 2p photoelectron peakfrom an aluminum alloy with naturally formed oxide layer.77 7175 73BINDING ENERGY (cV)27parameter should be independent of any electrostatic charging of the sample. A table ofthis parameter for zinc compounds is shown in Table 2.3 [2.14]. Conventionally, the sumof the excitation source energy (ho) and the Auger parameter is used for comparison.From the table, it is shown that the measured binding energies of the Zn 2P3/2photoelectrons for zinc and zinc oxide are both at the value of about 1022 eV. Therefore,by only comparing the binding energies, it is impossible to distinguish their differences.However, if their Auger parameters are taken into consideration, the sums of theexcitation energy and Auger parameter for zinc (at 2013.9 eV) and zinc oxide (at 2010.0eV) provide a better identification for their chemical states.(B) Ouantitative AnalysisThe area of a photoelectron peak defines the peak intensity. For a homogeneoussample, the number of photoelectrons emitted per second (I) can be expressed by [2.18]:I = nf a9y?AT (2.7)where n = number of appropriate atoms per unit volume of the sample (cm3)f = the X-ray flux (cm2sec1)a = photoelectric cross-section for the atomic orbital (cm2)o = an angular efficiency factor for the instrumental arrangement based onthe angle between the photon path and detected electrony = efficiency in the photoelectric process for the formation ofphotoelectrons= inelastic mean free path of the photoelectrons in the sample (cm)28Table 2.3 The Auger parameters for zinc compounds.Compound 2p3/2’ L3M45+ c + hu*Zn 1021.7 992.2 2013.9ZnOx 1021.8 988.2 2010.0ZnF2 1022.4 986.7 2009.1ZnBr2 1023.2 987.5 2010.7‘ Binding energy of Zn2P3/2 photoelectrons (in eV)Kinetic energy of the Auger LMM electrons (in eV)* hu = 1253.6 eV (MgKcL excitation source)29A = area of a sample from which photoelectrons are detected (cm2)T = detection efficiency for electrons emitted from a sampleFor given instrumental conditions, and a given photoelectron peak, the factors f a, 8, y, 2,A and T can all be grouped into the atomic sensitivity factor S. Then equation 2.7 can bere-expressed as:n=I/S (2.8)and this allows the relative concentrations of the various constituents in a sample to bedetermined from:C, = n / n = (I / Sx) / ( I / Si) (2.9)Values of sensitivity factors are generally available for photoelectron peaks relative to theflourine is peak as standard. In the work reported in this thesis, using the MAX 200spectrometer, the sensitivity factors used were those provided by the manufacturer.2.1.4 Angle Dependent XPSAngle dependent XPS (ADXPS) can provide information on composition withdepth for a material. The surface enhanced composition is emphasized for small take-offangle 0, as illustrated in Figure 2.10. The sampling depth t (introduced on p. 21) can nowbe extended to 3 sin 9. Therefore, the surface sensitivity is greater for smaller take-offangles.30huDetectorForO2 <9k, t2 <t1Figure 2.10 Illustration of the principle of angle dependent XPS (ADXPS).substrateintensity(a) Clean surface orthick homogeneous overlayer(c) Thin overlayerFigure 2.11 Theoretical angular dependent curves for a flat clean surface coatedwith an overlayer.tiI1f4s \‘\\Detector3?,hooverlayerC(b) Substrate(d) Ratio of‘OL’100 900031In a practical situation, a sample is often coated with an overlayer on top of asubstrate material (Figure 2.11) [2.19]. In this case, the angular variations of intensitiesfrom the substrate, and the overlayer,‘OL’ with thickness d, are given by:‘SUB = I° exp (-d / sin 9) (2.10)‘OL= 1° [1 - exp (-d / ? sin 9)] (2.11)where and A.c, are the inelastic mean free paths of the substrate and overlayer materials,respectively. In the ideal situation, where the overlayer is homogeneous and flat, theabove equations lead to curves of the types shown in Figure 2.11. For a clean surface(Figure 2.11(a)), or for a surface with an overlayer with thickness larger than about 3the intensity will be angle independent. For a sample covered with a thin overlayer (Figure2.11(b) & (c)), the intensity from the substrate will increase with 0; while the intensityfrom the overlayer will decrease with 9. Their ratio (Figure 2.11(d)),‘OL’ ‘SUB’ shows amore rapid decay with increasing 0. Actually, this ratio, 10L’ is useful to determinethe thickness of an overlayer. If ? and 2c’ are equal, then,‘OL’ = ‘01) I° [exp (d / ? sin 9) - 1] (2.12)R(9) = K [exp (d / sin 9) - 1] (2.13)[R(9) / K + 1] = exp (d / ? sin 9) (2.14)ln[R(9)/K+1] = d/Xsin9 (2.15)where R(9)= ‘OL’ ‘SUB and K = ‘OLd” ‘SUB32Since the values of R(9) and K can be obtained experimentally, the thickness of theoverlayer, d, can be determined from the slope of a graph with in [R(9) I K + 1] plottedagainst 1/sin Bias Technique Applied on XPSIf a sample has some regions which are not in good electrical contact with thespectrometer, differential charging may occur, and this can give rise to misleading bindingenergy information about the chemical states of elements if care is not taken. This isbecause photoeiectrons emitted from these regions have to overcome an extra potentialresulting from this charging effect; thus the measured binding energy of thesephotolectrons will become higher. However, recent work by Leung et a! [2.201, followinga study by Pertsin and Pashunin [2.21], suggests that components at a polymer-metalinterface which are not in good electrical contact with the spectrometer can bedistingusihed by applying a bias potential. Figure 2.12 shows the Zn 2P312 photoelectronspectra from a zinc phosphate treated aluminum alloy surface. Curve I is obtained fromthe grounded sample. The presence of two components, at 1026.7 eV and 1023.3 eV, ofthis photolectron peak implies that either the treated sample has two chemical states ofzinc, or some areas of the sample suffered differential charging. Curve 2 shows the resultof shifting the spectrum back by 94 eV after measuring with negatively biased potential (-94 V) applied on the sample. Only the 1023.3 eV component is observed in curve 2, andit is present in both curves. This suggests that the chemical component giving rise to it isnot subject to differential charging, but that component at 1026.7 eV, which only appearsin curve 1, is not in good electrical contact with the metal and thus becomes changed bythe bias potential in an uncontrolled way so that it does not appear in curve 2.33Curve 1rJDzzFigure 2.12 Zn 2P3/2 photoelectron peaks from a zinc phosphate treatedaluminum sample. Curve 1 is obtained from the grounded sample.Curve 2 is obtained from the sample with -94 V biased potentialapplied on the sample and shifted back by 94 eV after measuring.Curve 2I I I1029 1027 1025 1023 1021BINDING ENERGY (eV)342.1.6 Instrumentation of XPSThe XPS measurements made in this work were obtained with a Leybold MAX200 spectrometer [2.221 which is shown schematically in Figure 2.13. This spectrometerconsists of three vacuum chambers, including a transfer chamber, an analysis chamber anda preparation chamber. The latter two are maintained for ultra high vacuum conditionsbut the analysis chamber (Figure 2.14) is where the XPS measurements are performed.Major components in the analysis chamber include a dual anode X-ray source, aconcentric hemispherical analyzer with a multichannel detector, and 5-motion manipulatorwith sample holding, heating and cooling. In addition, this chamber also contains an iongun for sputtering and ion scattering spectroscopy (ISS), an X-ray monochromator forincreased energy resolution, and an electron flood gun to aid studies on insulating samples.Further discussion of some of these components is included in the following sections. Thepreparation chamber was not used in the present work, and the transfer chamber is usedfor sample entry.(A) Ultra-High VacuumThe photoelectron spectrometer is operated in the ultra-high vacuum (UHV)[2.23] range of 10-8 to 10-10 torr. In such a vacuum condition, the emitted electrons canreach the energy analyzer without being significantly scattered by the residual gasmolecules. As mentioned before, most photoelectrons analyzed originate from theoutermost atomic layers. As a result, the technique is very sensitive to surfacecontamination. Since many experiments need an atomically clean sample surface, andsometimes very small amounts of contaminant can affect an experiment significantly, it isnecessary to operate under conditions in which the accumulation of contamination is35(a)A(b)A. Transfer chamberB. Sample manipulatorC. Analysis chamberD. Concentric hemispherical energy analyser (CHA)E. Preparation chamberFigure 2.13 A schematic indication of the Leybold MAX 200 spectrometer (a)top view (b) side view.DABC36Microchannel plates///Data processing l[urntFigure 2.14 Components of the MAX 200 analysis chamber.-S5%‘S\/,// Concentric hemisphericalenergy analyzer (CHA)Kinetic energy —Sample/I/Analysis chamber\“5.//37negligible during an experiment. At a pressure of 1 0 torr, it is possible for a monolayerof gas to be adsorbed onto a surface in about one fifth of a second, if every collisionsticks. Therefore, to ensure that no more than about 0.05 atomic layers of contaminantaccumulates during data acquisition (say 30 minutes), pressure as low as 10-10 torr isrequired.The chambers and the associated piping of the spectrometer are constructed fromstainless steel. The preparation, transfer and analysis chambers are initially rough pumpedto the 10-2 torr range by rotary pumps, then pumped to the 1 0 or 10-8 torr range by thetubomolecular pumps, finally UHV can be achieved in the analysis chambers by baking ataround 140 °C for 12 hours or more while pumping. Baking removes adsorbed gas fromthe chambers walls so that UHV can be achieved within a resonable time, after coolingback to room temperature. Figure 2.15 shows the pumping system used in the MAX 200spectrometer; an additional ion pump is used for the X-ray source system.For sample entry, the transfer chamber is vented with nitrogen at atmosphericpressure, and after closing, the chamber is pumped down to 10-8 torr. The samples arethen transfered into the analysis chamber. The whole sample transfering process isefficient and maintains the UHV condition in the analysis chamber.(B) Excitation SourceA dual anode [2.231 X-ray source was used in this research; it has aluminum andmagnesium deposited on different faces of the anode block. Thus, the excitation sourcecan be changed by simply switching the anode filaments. Typically, an acceleratingpotential of 15 kV is used for the production of MgKa and AlKct radiation, whilesimultaneouly, the anode is cooled by a deionized water system (Figure 2.16).38TurbomolecuiarpumpFigure 2.15 The pumping system of the XPS spectrometer.Turbo-molecularpumpPreparationchamberIon sputteringgunTransferchamberIonpumpRotaiy Rotaiy Rotaiypump pump pump39Cooling waterFilament 1 -. Filament2.z,,, ,fltZZI V )‘ZZfltZAnode face 1 / 4indoi/ Anode face 2(magnesium) (aluminum)Figure 2.16 A schematic diagram of the Al and Mg dual anode X-ray source.40An aluminum foil window, of thickness about 2 jim, is positioned between theanode and sample to prevent stray electrons, radiative heating, and contamination from theanode region reaching the sample. This window material is transparent to MgJCc and AlKct radiation, and is sufficiently thin that the X-ray flux is not significantly attenuated.(C) Sample HandlingIn the MAX 200 system, up to seven sample holders can be locked on to thesample magazine, prior to introducing into the transfer chamber. One sample holder istransfered at a time to the analysis chamber by the lock rod, and then locked on to themanipulator. The manipulator allows three linear degrees of movement for the sample (x,,y, and z), as well as two rotational variables, 8 and 4 , to give proper sample positioningand angle dependent measurements. Figures 2.17 and 2.18 show this sample handlingassembly.(D) Electron Energy AnalyzerIn the MAX 200 system, the photoelectron kinetic energies are measured with aconcentric hemispherical analyzer (CHA) [2.23], which is coupled with an entranceretarding lens system. Figure 2.19 [2.24] shows a schematic diagram of the Cl-IAcomponent. The CHA consists of two concentrical hemispheres of radii R1 (inner) and R2(outer). A deflecting potential iW is applied between the spheres so that the outer isnegative and the inner is positive. The entrance slit (with width w i) and exit slit (withwidth 02) are centered on the mean radius R0, where R0 (R1 + R2) / 2. For the idealsituation, where an electron (charge e, and kinetic energy E0), is travelling on the circularorbit of radius R0, the relationship between E0 and the41Figure 2.17 The MAX 200 system: sample handling assembly.Analysis chamberSamplemagazineTransfer chamberSampleholderLockrod42Electron analyserManipulatorFigure 2.18 The analysis and transfer chambers of the spectrometer in MAX200 system.MonochromatizedX-ray sourceAnalysisX-ray gunSampleholderTransfer‘: \Lock rod0IJSamplemagazine43Figure 2.19 A schematic diagram of a concentric hemispherical analyzer (CHA).44deflecting potential is:eAV E0 (R2 / R1 - R1 / R2) (2.16)to ensure the electrons pass through the exit slit of the analyzer. Suppose that thedivergence of electrons entering the analyzer from the ideal path is cc the relativeresolution of the concentric hemispherical analyzer is given by:spec I E = (i + 2) / 2 R0+2 (2.17)where is the energy resolution. Since o, w2, R0 and cc are restricted by thespectrometer construction, the relative resolution of the spectrometer varies with the passenergy E0. Therefore, the smaller the E0, the better the resolution.In order to perform a highly precise energy analysis under a constant resolutioncondition, electrons entering the analyzer are pre-retarded by the lens system to a fixedkinetic energy, called the pass energy (E0). The kinetic energy scanning is done during thepre-retardation process, where the applied voltage in the lens system keeps ramping so thekinetic energies of electrons are retarded to the energy equal to E0 and able to passthrough the analyzer. Electrons with higher or lower kinetic energy than the pass energywill be deflected and hit the walls of the analyzer; thus are unable to reach the detector.The experimental resolution for a photoelectron peak, defined as the full width athalf maximum (FWHM), in a XPS spectrum is actually determined by three basic factorsnamely AEsoce, the peak width for the X-ray source, the analyzer resolution andatom’ the natural width determined by the uncertainty principle for the lifetime of the45ionic core state. For the assumption that the peaks have a Gaussian line shape, themeasured half width is given by:tota1 =(2source + LS.E2spec + AE2atom)” (2.18)(E) DetectorElectrons which exit the hemispherical analyzer are detected by the microchannelplates (Figure 2.20 [2.251). Each electron impacting on this detector results in a electronpulse in a channel with a multiplication of around io. These pulses are transformed intoelectrical signals and processed by the computer. The MAX 200 system in our laboratoryis interfaced to a FTP 1000 based microprocessor using Data System DS 100 software andis connected to the display unit.46radiationChannels/ IMicorchannel plateMicrochannel plate V4///////////////////I!11i1Channels(a)(b)Electrons MetalOutput anodeFigure 2.20 (a) A cutaway view of a microchannel plate, (b) a schematicdiagram of a microchannel plate assembly.472.2 Scanning Electron Microscopy2.2.1 BackgroundBecause of its high resolution and extensive magnification range, scanning electronmicroscopy (SEM) [2.26] has unique capabilities for analyzing surfaces. Unlike thereflected light microscope, which forms an image from light reflected from a samplesurface, SEM uses electrons for surface imaging. Actually, the wavelength of theradiation source determines the resolution levels. Higher energy electrons have shorterwavelengths and are thus capable of generating higher-resolution information. Enhancedresolution permits higher magnification without loss of detail. Generally, the maximumresolution and magnification limits of conventional light microscopes are 2000A and 2000times; whereas the limits of conventional SEM are 40A and 75000 times, respectively.A basic SEM unit consists of four main components: the illuminating I imagingsystem, the information system, the display system and the vacuum system. Each of thesesystems and their relationships are discussed below. In our work, a Hitachi S-2300 modelscanning electron microscope is employed. Micrographs are taken at 4000 magnificationwith 5 kV accelerating voltage (Figure 2.21).2.2.2 System UsedThe function of the illuminating/imaging system is to produce and direct anelectron beam on to the sample. It comprises an electron gun and several magnetic lensesas indicated in Figure 2.22. The major components of the electron gun are the filament, anapertured shield and an anode. Electrons are produced from the filament by the48To pumping systemFigure 2.21 A schematic diagram of a scanning electron microscope (SEM).Electron gunObjective lensSpecimen positionStandard stage49ElectrodesL1 Anodefilament[.\\ NJApertureCondenser lensObjective lensSpecimenFigure 2.22 The illuminating / imaging system of a SEM unit.50thermionic effect, and an accelerating voltage is applied between the filament and anode.The apertured shield, or grid cap, is a slightly biased cylindrical cap which serves tocollimate the electrons from the filament and direct them to the anode. The electron beamis then focused from a diameter of about 25000A to 1 ooA by a series of magnetic lenses,and directed on to the sample. As a rule, the smaller the beam diameter, or the spot size,the higher the resolution.The information system consists of the sample, which releases the electron signalsresulting from interaction with the imaging beam, and the detectors, which recognize andanalyze the signals. The sample is mounted on a conductive substrate and secured withinthe sample stage of the microscope. The stage serves as an electrical pathway to ground,and is also equipped with several controls for linear movement. The electron signalsgenerated from the collision of the electron beam with the sample are detected by animaging detector.SEM images are displayed on the screen of a cathode-ray tube (CRT) and thescanning electron micrographs can be photographed for permanent record.The SEM chamber is operated under vacuum ( 1 o torr) to avoid scatteringbetween the electron beam and the residual gas. Also, this condition can help slowoxidation of the X-ray gun filament and limit contamination of the sample. The chamber ispumped by tubomolecular pumps backed by rotary pumps.512.3 Weight Loss MeasurementsWeight loss measurements [2.27] allow determination of the amount of corrosionoccurring for a sample in a specific environment. In practice, the measurements are doneby measuring the weight of the sample before corrosion (w1), then exposing the sample tothe corrosion environment, and then measuring the weight of the sample again aftercorrosion and cleaning of the sample surface (w2). The corrosion rate can be calculatedaccording to:Corrosion rate = (K x W)/(Ax Tx D) (2.19)where W = w1 - w2 = the total weight loss (g)A = surface area of the sample (cm2)T = time of exposure in the corrosion environment (hour)D = density of the sample (glcm2)K = a constant(= 2.78 x 106 x D if corrosion rate is expressed in p.g/m2s)In our work, the corrosion environment employed for the weight lossmeasurements is immersion in a 3.5% NaC1 solution for 2 hours at room temperature.After this treatment, the sample is chemically cleaned by immersing in concentrated nitricacid for 15 minutes at room temperature to remove any corrosion products from thesurface. The sample is then rinsed with distilled water, vacuum dried, and weighed. Allweighings were done on an analytical balance to 0.0001g.522.4 Atomic Absorption SpectrometryIn this work, atomic absorption spectrometry (AAS) [2.28] is applied to detect thedissolution of aluminum ions (A13+) from the sample during treatment in a corrosiveenvironment. This method is used as a supplement to evaluate the corrosion protectionperformance of the zinc phosphate treated aluminum alloy surface. The 3.5% NaC1solution which is used for the corrosion test is analyzed by AAS to detect the presence ofA13. The absorbance is measured at wavelength of 309.21 nm by a Perkin Elmer 305Aatomic absorption spectrometer using aN20/acetylene flame as the atomization source.Calibrations are done to correct the background by NaC1 interference. The sensitivity ofthe method is about 0.01 ppm.53Chapter 3 Coating and Corrosion Studies on ZPO treated AluminumSurfaces3.1 Coating Studies on Zinc Phosphate Treated Aluminum Surfaces3.1.1 Sample PreparationsThe samples studied in this part of the work are designated by labels A to G (Table3.1). A flow chart which summaries the whole sample preparation steps is shown inFigure 3.1. The 7075-T6 aluminum sheets, which contain about 6 % of zinc, 3 % ofmagnesium, 2 % of copper, 0.5 % of iron, 0.4 % of silicon, 0.2 % of titanium, 0.3 % ofmanganese and aluminum with the rest of the composition [3.11, were obtained from theDefence Research Establishment Pacific (DREP). They were cut into 1 cm x 1 cm panelsfor the experimental purpose. These panels were all polished to a mirror-like appearance(0.05i.tm finish) and then degreased with acetone and methanol. The coating processeswere done by suspending these test panels in 10 wt.% zinc phosphate (ZPO) suspension atdifferent pH conditions for 1 hour at room temperature. The acidity and basicity of thecoating solution were adjusted by acetic acid (HOAc) and sodium hydroxide solution(NaOH) respectively. After coating, some panels were ultrasonically rinsed in distilledwater for 1 mm., while some of them were just normally rinsed in distilled water. All thepanels were then rinsed with absolute ethanol and air-dried, followed by the XPS analyses.3.1.2 XPS MeasurementsXPS spectra were measured in a Leybold MAX 200 spectrometer at an operatingpressure of 4.5 x i0 torr. The unmonochromatized MgKa radiation source (1253.6 eV)54Table 3.1 Sample descriptions.Code in text DescriptionA 7075-T6 aluminum panel polished to 0.05 jim finish, followed bydegreasing with acetone and methanol.B Sample A treated in natural zinc phosphate (ZPO) solution(pH=6.6) without ultrasonic rinsing afterwards.C Sample A treated in natural zinc phosphate (ZPO) solution(pH=6.6) with ultrasonic rinsing afterwards.D Sample A treated in ZPO solution adjusted at pH=13.0 by NaOHwith ultrasonic rinsing afterwards.E Sample A treated in ZPO solution adjusted at pH10.5 by NaOHwith ultrasonic rinsing afterwards.F Sample A treated in ZPO solution adjusted at pH5.0 by HOAcwith ultrasonic rinsing afterwards.G Sample A treated in ZPO solution adjusted at pH=3.5 by HOAcwith ultrasonic rinsing afterwards.557075-T6 Aluminum surface (1 cm x 1 cm in area)1-Polished to 0.05 im finish(mirror-like appearance)4,Degreased with acetone and methanolTreated in 10 wt.% zinc phosphate solution for 1 hour4,Sample Sample Sample Sample Sample Sample SampleA B C D E F GBlank Immersed in Immersed in Immersed in Immersed in Immersed in Inunersed inAl natural natural pH13.0 pH=10.5 pH=5.0 pH=3.5zP0 zPO ZP0 zPO zPO zP0solution solution solution solution solution solution(pH=6.6) (pH=6.6) adjusted adjusted adjusted adjustedby NaOH by NaOH by HOAc by HOAcI F I I1- ‘I,4,. 4,Rinsed in distilled Ultrasonically rinsed in distilled water andwater and ethanol then rinsed in ethanol and air dried4, 4,4,XPS analysis Biased XPS XPS analysismeasurementsFigure 3.1 A flow chart summarizing the sample preparation procedure for thecoating process.56was operated at 15 kV and 20 mA. Survey spectra for use in qualitative analysis wereobtained with the pass energy of the hemispherical analyzer set at 192 eV; higher-resolution narrow scan spectra were measured for the Zn 2P3/2 0 is, C is, P 2p and Al2p core levels at a 48 eV pass energy. For the latter, integrated peak areas determinedafter background subtraction were taken to measure relative elemental amounts aftercorrection with the appropriate sensitivity factors provided by the manufacturer. Corelevel binding energies were referenced to the gold 4f712 binding energy at 84.0 eV.Narrow scan spectra were also measured for different values of the take-off angle (9). Anegative bias potential (-94 V) was applied to sample B and C during the biasedexperiments. The biased spectra obtained were shifted back by 94 eV for comparison withnon-biased spectra.3.1.3 Results and Discussion(A) The Aluminum Control Panel (Sample A)The XPS survey spectrum of sample A indicates the presence of oxygen, carbon,and aluminum on the surface (Figure 3.2), as identified from the 0 is, C is and Al 2pphotoelectron peaks. The carbon signals arise from airborne contamination or residualacetone and methanol in the degreasing procedure. The oxygen signals come from theoxygen component of various compounds on the surfaces, e.g. air-borne contamination,residual organic compounds and metal oxide. High resolution XPS spectrum of Al 2pphotoelectron peak (Figure 2.7, as discussed in section 2.1.3 (A) of Chapter 2) shows thepresence of both aluminum oxide (AlOx) at binding energy of 74.1 eV and metallicaluminum at binding energy of 71.5 eV. Quantitative analysis and angle dependent XPSare performed, and the results are shown in Table 3.2. It is found that the compositionratio of the oxide peak to the metallic peak of aluminum increases with decreasing the57L.o Is-IBINDING ENERGY (eV)Figure 3.2 XPS survey spectrum for sample A, a blank Al surface..4. C (Auger)0 (Auger) I. C Is.i Al 2sAl 2p1000 800 600 400 200 058Table 3.2 Elemental composition and atomic ratios for blank Al surface (sample A)with varying take-off angle, 0.0 ‘Elemental composition (%) Atomic ratio0 *M AlOx **M) AlOx/Al300 71.2 28.8 27.1 1.7 16.6450 70.5 29.5 27.6 1.9 14.960° 69.8 30.2 27.9 2.3 12.190° 70.2 29.8 26.6 3.2 8.3*M indicates the total Al content.* *Al() indicates the metallic Al content.Carbon was not taken into consideration since it is mainly from contamination.59take-off angle. This behaviour matches the conventional expectation that the oxide layersare on the top of the bulk metal [3.2]. Also, the thickness of the oxide layers areestimated at around 36 A [3.3], using the inelastic mean free path values of 23 A for AlOxand 32 A for metallic Al [3.4]. It is noteworthy that on this blank aluminum surface, noother alloy composition (e.g. zinc or magnesium) is detected.(B) The Aluminum Panel Treated in Natural ZPO Solution at pH=6.6 (Samples B andThe coating processes for both sample B and sample C are similar, in which theyare treated in natural zinc phosphate solution (with pH=6.6) for 1 hour at roomtemperature, except sample B without ultrasonic rinsing afterwards but sample C withultrasonic rinsing (Figure 3.1). Qualitatively, the XPS survey spectra for the two sampleslook the same, both showing the presence of zinc in addition to oxygen, carbon andaluminum. The spectrum for sample C is shown in Figure 3.3. In both samples, however,no phosphorus is detected.Biased and non-biased XPS measurements were performed on samples B and C.The high resolution spectra of Zn 2P3/2 photoelectron peaks for samples B and C at takeoff angles of 900 and 30° obtained from non-biased XPS measurements are shown inFigure 3.4 (a) to (d). It is clear that two forms of zinc are present in sample B, withbinding energies of 1026.7 eV and 1023.3 eV, but only the second one is well-establishedin sample C. With variation of the take-off angle, the intensity of the 1026.7 eV peak isenhanced at e = 30°, showing that this higher binding energy component is located at thetopmost region of the surface. Bias potential technique is applied to sample B and sampleC. The resulting spectra, after shifting the measured spectra back by 94 eV, are shown inFigure 3.4 (e) and (f). It is found that the negative bias potential technique has no effect60.lo IsIBINDING ENERGY (eV).l Zn (Auger).I Al 2sFigure 3.3 XPS survey spectrum for sample C, ZPO treated Al surfaceprepared at natural pH..1 Zn 2p 0 (Auger). C (Auger)‘C IsAl 2p1000 800 600 400 200 0610=900 0=30°ZJ%L(c) (d)I1029 1027 1025 1023 1021 1029 1027 1025 1023 1021BINDING ENERGY (eV)Figure 3.4 High resolution Zn 2P3/2 spectra for: (a) sample C, 0 = 900; (b)sample C, 0 = 30°; (c) sample B, 0 = 90°; (d) sample B, 8 = 30°; (e)sample B, 0 = 90° after bias potential; (f) sample B, 8 = 30° afterbias potential.62on the lower binding energy component, which still lies at 1023.3 eV after the measuredspectra are shifted back. On the contrary, the higher binding energy componentdisappears in the spectra, suggesting that it is not in good electrical contact with the metal[3.5). It is concluded that the 1026.7 eV component corresponding to zinc is weaklybonded or physically trapped at the surface; thereby it can be lost by ultrasonic rinsing, andit is not observed in sample C. The lower binding energy component (1023.3 eV), on theother hand, binds strongly at the surface. Therefore, it is not affected by the biased XPSmeasurements or the ultrasonic rinsing procedure. In conclusion, the ultrasonic rinsing indistilled water, which removes all the physically trapped species on the treated surfaces, isessential in the sample preparation procedure.For the chemically absorbed zinc in the surface, the sum of its Auger parameter (ce)and the excitation energy (1253.6 eV) equals 2010.0 eV, which is consistent with thecomponent arising from an oxide ZnOx [3.6]. Quantitative and angle dependent XPS aredone on sample C. The results are listed in Table 3.3. Comparing with the blank Alsurface, there is a significant increase in the amount of zinc in sample C. The Zn/Al ratioessentially shows an angle independent behaviour, suggesting that ZnOx forms a mixturewith the AlOx [3.7]. Figure 3.5(a) and (b) schematically indicates the morphologiesproposed for these ZPO treated alloy surfaces.The presence of zinc on the treated surface originates from the ZPO compound.ZPO is slightly soluble in water [3.8].Zn3(P04)2 —* 3 Zn + 2 PO43 = 9.1 x i0 (3.1)The zinc ions exist as a hydrated form, Zn(H2O)4, in solution and also hydrolyze inwater, giving a slightly acidic coating solution [3.9].63Table 3.3 Elemental composition and atomic ratios for ZPO treated Al surface(sample C) with varying take-off angle, 9.9 ‘Elemental composition (%) Atomic ratio0 Al Zn Zn/Al300 74.3 24.5 1.2 0.05450 72.3 26.0 1.7 0.06600 712 26.9 1.9 0.0790° 71.9 26.2 1.9 0.07‘‘Carbon was not taken into consideration since it is mainly from contamination.64(a) Sample B: ZPO treated surface withoutultrasonic rinsingPhysicallytrapped ZnOxZnOx-AIOX -._ -— —-I”S.-—‘I 4 4Al(b) Sample C: ZPO treated surface withultrasonic rinsing• ZnOx-AlOx-ZPOI ••.1(c) Sample D & E: ZPO treated surfacesprepared at pH= 13.0 &pH= 10.5• Al4(d) Sample F: ZPO treated surfaceprepared at pH5.0.4Al(e) Sample G: ZPO treated surfaceprepared at pH=3 5Figure 3.5 Proposed surface morphologies for (a) sample B, (b) sample C, (c)samples D & E, (d) sample F and (e) sample G.44-Zinc and phosphorusrich at the outersurface regionZnOx-AIOx-ZPOF 4 —.4%_. —4 4 44SzPo•:... AlOx :. L ;-•••14V 44 5.44 444.44 .465Zfl(H2O)42(aq + H20(l) —> Zfl(H2O)3(OH)(aq) + (H3O)(aq) (3.2)The hydrated zinc ions are believed to interact with the alloy surface, and incorporatedeeply into the aluminum oxide matrix, thereby forming a AlOx-ZnOx mixture.Phosphorus, though, may be adsorbed on the surface, but it is undetectable by XPS. Thisbehaviour provides a reference point with which to compare the samples prepared inalkaline and acidic coating solutions.(C) The Aluminum Panel Treated in ZPO Solution at Various pH Values (Samples Dto G)After the ZPO treatments, all the four samples (samples D, E, F and G) retain amirror-like appearance. Based on the discussion in Section B, these treated samples areultrasonically rinsed in distilled water to remove any physically trapped zinc component.XPS analyses are performed and their XPS spectra look similar, except that the intensityof the photoelectron peaks are different. The XPS survey spectrum of sample D is shownin Figure 3.6. It is observed that besides oxygen, carbon and aluminum, zinc andphosphorus are detected on these four samples. The high resolution XPS spectrum of Al2p peak shows that it contains both oxide and metallic components, lying at bindingenergies of 74.1 eV and 71.5 eV.The binding energies of zinc and phosphorus and the kinetic energies of zincL3M45 Auger lines measured on these surfaces are listed in Table 3.4. The table alsogives the values obtained from the ZPO reference compound for comparison. The bindingenergies of Zn 2P3/2 photoelectron peaks lie between 1023.1 eV and 1023.7 eV, and thesums of the Auger parameters of Zn and the excitation energy are in the range of 2010.166—a.Ia.IC’)Iz4’ Zn 2p,IZn3s J.AJ2s,, P 2pBINDING ENERGY (eV)BINDING ENERGY (eV)‘Zn3pFigure 3.6 XPS survey spectrum for sample D, ZPO treated Al surfaceprepared at pH=13.O..I P 2s Al 2p200 180 160 140 120 100 80 60‘J’OJs4’ C (Auger)4’ 0 (Auger) 4’ Zn (Auger)4’ Zn (Auger) I/1000 800 600 400 200 067Table 3.4 A list of binding energies (in eV) of zinc, phosphorus on ZPO treatedaluminum surfaces and the ZPO reference compound.Zn Auger’Sample Zn 2p3,2’ L3M45 cL+hu* P 2p C isZPO referencecompound 1023.7 988.6 2010.3 133.7 285.0C(atpH=6.6) 1023.3 987.1 2010.4 -- 285.0D(atpH13.0) 1023.4 986.9 2010.3 133.5 285.0E(atpH=10.5) 1023.5 987.0 2010.5 133.7 285.0F (at pH=5.0) 1023.6 987.0 2010.6 133.7 285.0G(atpH=3.5) 1023.1 987.0 2010.1 133.9 285.0‘‘ The values of the binding energies ofZn2P3I2 P 2p and C is photoelectrons (in eV)The values of the kinetic energy of the Zn AugerL3M45 electrons (in eV)* The excitation source energy (hu) equal to 1253.6 eV68to 2010.6 eV, consistent with a ZnOx [3.6], although very similar values (1023.7 eV and2010.3 eV repectively) were measured for the ZPO reference compound. Therefore, itwas not possible to discriminate between these forms in relation to zinc (i.e. this metal ispresent in the +2 oxidation state). The measured P 2p photoelectron peaks lie between133.5 eV and 133.9 eV, matching the value (133.7 eV) of the ZPO reference compound.Therefore, the chemical state of phosphorus on the treated surfaces is in the form ofphosphate (i.e. in the +5 oxidation state).Figure 3.7(a) shows the pH effect on the samples of the atomic ratios of zinc toaluminum and phosphorus to aluminum measured at8 = 90°. It is found that starting frompH=6.6 of the ZPO coating solution, the amounts of zinc and phosphorus on the alloysurface increase with the pH to 10.5 and 13.0. However, with decreasing pH from 6.6,the amount of coated zinc and phosphorus is increased at pH=5.0 and decreased atpH=3.5. Angle dependent XPS and quantitative analysis (Table 3.5) reveal thedistributions of zinc and phosphorus on the surfaces. Values of Zn/P ratios can be used tofollow the trends in composition, although it is not generally possible to relate to absolutecompound composition. For example, the ratio measured for the ZPO referencecompound is 0.88. Aside from any uncertainties in the atomic sensitivity factors,differences from the value of 1.5, predicted by the formulaZn3(P04)2,may be expected toarise from changes in surface composition (e.g. this ratio will vary markedly withinvolvement by HP04 andH2P04 ions near the surface of the reference material).The following subsections summarize the XPS observations for each coatingsolution and discuss in terms of the solution and interface chemistry that are likely to berelevant to the observations.69(a) Before corrosion0.140.05 0.05r—i 0.00 E] 0.00 IA2 G2 F2 C2 E2 D2(Blank) (pH=3.5) (pH=5.O) (p11=6.6) (pH=1O.5) (pH=13.O)Figure 3.7 pH effect on the atomic ratios of elements on the ZPO treated Al surfaces(a) before and (b) after corrosion tests (values obtained at take-off angleequal 900).0.520.300.07 D(pH=1O.5) (pH=13.0)0.35Zn/Al• P/Al0.04 0.050.00 0.00 I •;G F(p11=3.5) (p11=5.0)0.290.00A(Blank)0.03I—1 0.00(b) After corrosion0.020.0070Table 3.5 Atomic ratios for samples C, D, E, F, G, and the ZPO reference compoundwith varying take-off angle, 8.Sample 8 Zn/Al P/Al Zn/PZPO referencecompound 0.88C 300 0.05(at pH=6.6) 450 0.0660° 007900 0.07D 30° 0.42 0.10 4.2(atpH=13.0) 45° 0.49 0.15 3.360° 0.53 0.09 5.990° 0.52 0.08 6.5E 30° 0.39 0.09 4.3(atpH=10.5) 450 0.37 0.08 4.660° 0.34 0.06 5.790° 0.30 0.07 4.3F 30° 0.55 0.19 2.9(atpH=5.0) 450 0.53 0.19 2.860° 0.44 0.11 4.090° 0.35 0.07 5.0G 30° 0.03 0.10 0.3(atpH=3.5) 45° 0.03 0.06 0.560° 0.04 0.06 0.790° 0.04 0.05 0.871(1) Strongly Alkaline Coating Solution (pIF43.O; Sample D)XPS studies (Table 3.5) show that large amounts of zinc and phosphorus arepresent on this treated alloy surface. The enhanced value of the Zn/P ratio (comparedwith ZPO) indicates that the surface is zinc rich, relative to phosphorus, possibly bothZnOx and ZPO being involved in the coating. Angle dependent XPS shows that the Zn/Alratios vary slowly with the take-off angle, and the P/Al ratio is almost angle independent,implying that zinc and phosphorus are randomly distributed in the aluminum oxide mixture[3.7]. It is proposed that this coating is a mixed ZPO-AlOx-ZnOx material, as shown inFigure 3.5(c).The formation of this layer is the result of both the etching and coating processeson the surface. The hydroxide ions present in this strong alkaline solution will react withthe aluminum oxide on the alloy surface [3.9, 3.101.Al203(s) + 2 OW(aq) + 3 HO * 2 Al(OH)4(aq) (3.3)The aluminum oxide layer is thus etched out, and therefore zinc, one of the constituents ofthe alloy composition, is exposed to the surface region.On the other hand, the solubility of ZPO is enhanced because of the formation ofzinc hydroxide and zinc hydroxide complexes [3.101.Zn2(aq) + 2 OW(aq) —* Zn(OH)2(5) (3.4)Zn(OH)2(s) + OH(aq) _+ Zfl(OH)3(aq) (3.5)72Zn(OH)3(aq) + OH(aq) + Zfl(OH)42(aq (3.6)These complexes, together with the increased amount of phosphate ions in the solution,are believed to interact with the freshly etched alloy and incorporate into the aluminumoxide matrix. It is expected that both the etching and coating processes occurconcurrently, thus giving a ZPO-AlOx-ZnOx structure. Since the enhancement insolubility causes more zinc and phosphate ions in solution, larger amounts of zinc andphosphorus can be coated on the surface.(2) Moderately Alkaline Coating Solution (p11=10.5; Sample E)XPS studies indicate that the amounts of zinc and phosphorus on the alloy surfacetreated at this condition are less than that of sample D. XPS measured for different 0(Table 3.5) shows that the P/Al ratio is angle independent and the Zn/Al ratio slightlyvaries with the take-off angle, suggesting that the layer is a mixture of zinc, phosphorusand aluminum [3.7]. The Zn/P ratio indicates sample E is zinc rich, implying both ZnOxand ZPO are present. It is believed that the surface morphology of this surface is similarto that of sample D, i.e. a mixture ofZPO-AlOx-ZnOx structure (Figure 3.5(c)). Also, theetching and coating processes are similar in both cases. However, with lower alkalinityfor the coating solution at pH= 10.5, a relatively milder etching process is expected. Thus,smaller amounts of zinc and phosphorus are found on the treated surface.(3) Moderately Acidic Coating Solution (pH=5.O; sample F.XPS studies (Figure 3.7(a)) indicate that more zinc and phosphorus are on sampleF than on sample C. Angle-dependent XPS (Table 3.5) shows that both Zn/Al and P/Alratios were almost unchanged at small take-off angles, but they decrease for the large73values of 0. This suggests that zinc and phosphorus are mixed in the AlOx matrix [3.7],but with distributions that are relatively richer at the outer surface region. The Zn/P ratioof sample F is larger than that of the ZPO reference compound, implying that the coatinginvolves ZnOx and ZPO, and that etching and coating processes have occurred. Aproposed surface morphology is shown in Figure 3.5(d).At pH=5.0, the presence of H ions in the ZPO coating solution enhances thesolubility of ZPO because of the formation of hydrogen phosphate (HP042), dihydrogenphosphate (H2P04),and phosphoric acid (H3P04)[3.12, 3.13].P043(aq) + H(aq) _+ HP042(aq) K1 = 4.55 X 1012 (3.7)HP042(aq) + H(aq) _* H2PO4(aq) K2 = 1.61 X 1 0 (3.8)H2PO4(aq) + H(aq) _* H3P04(l) K3 = 133 (3.9)The presence of ZnOx suggests that the etching process, possibly by phosphoric acid, hasoccurred on the surface, and thus exposed zinc from the substrate. Also, the enhancedsolubility of ZPO in this medium ensures that the surface is exposed to Zn2 andphosphate ions (e.g. H2P04), which can be absorbed and incorporated into the AlOxmatrix. This is consistent with the formation of a ZnOx-AlOx-ZPO coating on the alloy.Insofar as the zinc and phosphorus compositions are greater at the outer surface region,relative to the inner coating region, the precipitation of zinc appears to occur relativelyfaster than the zinc build up from the etching process.74(4) Strongly Acidic Coating Solution (pH=3.5; sample G)XPS results (Figure 3.7(a)) indicate that the ZnJA1 and P/Al ratios have thesmallest values among the five treated alloy surfaces. However, it is found that the Zn/Pratio at 8=900 (Table 3.5) is closest to that of the ZPO reference compound. The Zn/Alratios appear constant with take-off angle; likewise the P/Al ratios are almost unchangedfrom 90° to 45°, although the ratio increases at 8 = 30°, suggesting that more phosphorusis at the outermost surface region. Figure 3.5(e) shows the proposed surface morphologyof sample G.The low zinc to phosphorus ratio on this treated surface may result from thedissolution of the ZnOx-AlOx structure. At pH=3.5, phosphoric acid is a majorcomponent in the coating solution. This acid possibly etches the surface and exposes zincfrom the substrate. However, relating to the observations, it is expected that the ZnOxAlOx structure is unstable under strong acidic condition and is thereby removed from thesurface. Moreover, the precipitate of ZPO is suppressed because of the strong acidicenvironment [3.14]. Consequently, the amounts of zinc and phosphorus on the treatedsurface are small.753.2 Corrosion Studies on Zinc Phosphate Treated Aluminum Surfaces3.2.1 Sample PreparationsThe corrosion studies were divided into two parts. Part I was an initial test onsamples A and C only; while Part II was a further study involving all the treated surfaces(samples A, C, D, E, F and G), which were prepared at different pH conditions.Figures 3.8 and 3.9 summarize the sample preparation procedures for Parts I andII of these studies. In Part I (a), XPS and scanning electron miscroscopy (SEM) analyseswere performed, as shown in Figure 3.8(a). Samples A and C, after the SEMexaminations, were immersed in 3.5 % NaC1 solution for 2 hours. After that, they wererinsed with distilled water and then air-dried. The resulting surfaces were designated assamples Al and Cl respectively. Both XPS and SEM studies followed.In Part I (b), weight loss and atomic absorption spectroscopy (AAS) analyses wereperformed, as shown in Figure 3.8(b), on blank 7075-T6 aluminum panels which werepolished to a 220 grit finish (including 2 faces and 4 edges). After polishing, the panelswere degreased with acetone and methanol. The coating process was done by suspendingthe test panels in 10 wt. % zinc phosphate solution (natural ZPO solution) for 1 hour atroom temperature. After the ZPO treatment, the panels were ultrasonically rinsed indistilled water for 1 minute, followed by rinsing in absolute ethanol and then air-dried.Weight loss measurements were performed on these treated surfaces and blank surfacesrespectively. After the tests, the 3.5% NaCI solution was analyzed by atomic absorptionspectrometry to detect the presence ofA13+ ions.76(a) XPS and SEM examinationsMirror-like Mirror-likesample A sample CSEM evaluation‘I..Immersed in 3.5% NaCI solutionfor 2 hours‘I,Rinsed in distilled water and air-driedSampleAl Sample Cl‘I•1-SEM evaluation‘I,4-XPS analysisFigure 3.8 Sample preparation steps for Part I (initial studies) of corrosionstudies.77(b) Weight loss and atomic absorption spectrometry analyses7075-T6 Al panel 7075-T6 Al panel220 grit finish 220 grit finish1-Degreased in acetone Degreased in acetoneand methanol and methanol1..Treated in natural ZPO solution‘I- followed by rinsing (as in sample C)1Weighed before corrosion test.1..Immersed in 3.5% NaC1 solution for 2 hours4.AAS analysis forRinsed in distilled water NaC1 solution todetect M31..Cleaned in nitric acid for 15 minutes“Rinsed in distilled water and vacuum driedWeighed after corrosion testFigure 3.8 Sample preparation steps for Part I (initial studies) of corrosionstudies.78Sample Sample Sample Sample Sample SampleA C D E F GBlank Al Treated in Treated in Treated in Treated in Treated inZPO solution ZPO solution ZPO solution ZPO solution ZPO solutionat pH=6.6 at pH=13.O at pH= 10.5 at pH=5.O at pH=3.5‘ifImmersed in 3.5% NaCl solution for 5 hours‘If‘I,.Rinsed in distilled water and air-dried‘I,1-Sample Sample Sample Sample Sample SampleA2 C2 D2 E2 F2 G2‘If‘ifXPS analysisFigure 3.9 Sample preparation steps for Part II (fhrther studies) of corrosionstudies.79In Part II of the corrosion studies, as shown in Figure 3.9, samples A, C, D, E, Fand G were immersed into 3.5% NaC1 solution for 5 hours, followed by distilled waterrinsing and air drying. The resulting surfaces were designated as A2, C2, D2, E2, F2 andG2 respectively. XPS analyses were performed on these surfaces.3.2.2 Results and Discussion(A) Part I: Initial StudiesSamples A and C after the corrosion test were analyzed by XPS. The atomic ratiosobtained are listed in Table 3.6. It is found that zinc is present in sample Al, which is thesurface of sample A after the corrosion test. High resolution XPS shows that this Zn2P312 peak lies at binding energy of 1023.3 eV, implying that zinc exists as a form of zincoxide (ZnOx). The presence of zinc in sample Al suggests that immersion in the NaC1solution exposes zinc from the bulk aluminum alloy. Angle dependent XPS measurements(Table 3.6) show that the Zn/Al ratio is independent of the take-off angle, indicating thatZnOx is distributed evenly within the aluminum oxide layer [3.7]. Only the AlOx atbinding energy of 74.1 eV is detected on sample Al. No metallic Al is found. Thissuggests the sample has a further oxidation of aluminum metal to AlOx in the salt solution.The Zn/Al ratios, measured at take-off angles of 90° to 45°, are similar for samplesC and Cl, as shown in Table 3.6. This suggests that immersion in the corrosive NaClsolution has no significant effect on the amount of zinc on the surface, and the surfacestructure created from ZPO treatment does not significantly change under the corrosiontest condition for 2 hours. At take-off angle equal to 30°, the Zn/Al ratio of sample Clincreases, probably resulting from the effect of the preferential removal of AlOx at theoutermost surface region and thus effectively building up ZnOx on the surface.80Table 3.6 Zn/Al ratios for samples A, Al, C and Cl with varying take-off angle, 0.Sample Sample Sample Sample0 A Al C Cl300 0.00 0.01 0.05 0.09450 0.00 0.02 0.06 0.0760° 0.00 0.02 0.07 0.0690° 0.00 0.02 0.07 0.0681Consequently, there is relatively more zinc present at the outer surface region than at theinner region.Figure 3.10 shows the SEM micrographs for samples A, Al, C and Clrespectively. Both samples A and C (Figure 3.10(a) and (c)), which are surfaces prior toexposure to 3.5% NaC1 solution, have a similar flat appearance. However, after exposureto 3.5% NaC1 solution, the appearance of a large corrosion area on sample Al suggeststhat serious corrosion attack has occurred on this untreated aluminum surface. On theother hand, the corrosion areas on sample Cl are much smaller and the sample has most ofthe surface unchanged.Weight loss measurements are used to determine the corrosion rates of the blankaluminum sample and the ZPO treated aluminum sample. After 2 hours exposure to 3.5%NaC1 solution at room temperature, the corrosion rate for the blank sample is found to be118 p.g.m2.s However, no detectable weight loss is observed for the treated surface.The results from atomic absorption spectrometry (AAS) are consistent with the weightloss measurements. About 0.2 ppm of Al is found in the 3.5% NaCl solution after the 2hours corrosion test for the blank Al sample. By contrast, the Al concentration in theNaCl solution, in which the surface treated with ZPO is immersed, is below the detectionlimit of 0.01 ppm. These observations, together with the SEM and XPS analyses, showthat a more serious corrosion attack occurs on the surface of sample A than on sample C.The common point observed in previous studies of CF attack on naturally formedaluminum-oxide covered aluminum is the dissolution of the metal [3.15-3.18], which isalso observed in the present studies. Nguyen and Foley [3.15, 3.17, 3.18] proposed thefollowing mechanism:82(a)(b)Figure 3.10 SEM micrographs of: (a) sample A, blank aluminum; (b) sampleAl, sample A after corrosion; (c) sample C, ZPO treated aluminum;(d) sample Cl, sample C after corrosion.I83(c)(d)Figure 3.10 SEM micrographs of: (a) sample A, blank aluminum; (b) sampleAl, sample A after corrosion; (c) sample C, ZPO treated aluminum;(d) sample Cl, sample C after corrosion.84Step I: Adsorption on the oxide filmCF (in the bulk solution) —> CF (adsorbed on A1203.H sites) (3.10)Step II: Chemical reactionM3(in lattice) + 2 CF + 2 °‘(aq) —* Al(OH)2Cl2(aq) (3.11)The product A1(OH)2C1 is a soluble complex which is proposed to difihise from reactionsites into solution. In the meantime, oxidation of the metallic Al occurs [3.17, 3.18]:Al + 3 HO —* Al(OH)3 + 3 H + 3 e (3.12)The results in the present work are consistent with this mechanism. Breakdown ofthe oxide film by CF causes a rough and corroded surface appearance according to SEMinvestigation. The aluminum complex dissolving from the solid (Equation 3.11)contributes to the aluminum detected by AAS. Equation 3.11 is also consistent with theXPS observation that no C1 is detected on the 7075-T6 aluminum alloy surface after thecorrosion test, which was first reported by Arnott et al [3.19]. Also, these authors failedto detect zinc on the blank surface but confirmed its presence after the corrosion test. Thedissolution of aluminum from the ZPO treated surface is greatly suppressed, indicatingthat the zinc may strengthen the surface structure and thereby slow down the aluminumdissolution.85(B) Part II: Further StudiesThe corrosion tests performed in this part of the study are extended from 2 hoursto 5 hours. Results from XPS are summarized in Table 3.7. It is found that phosphorus isdetected from sample D2, but not at all from the other surfaces. On the other hand, zinc,in the form of ZnOx, is found on all the six samples, where some surfaces have an increasein the amount of zinc compared with the surfaces before corrosion, but some have adecrease. After the 5 hour corrosion attack, the metallic component of Al 2pphotoelectron peak is detected on samples D2, E2, F2 and G2.Figure 3.7(b) (on p.70) summarizes the quantitative results of the corrosion tests.As in the case of sample Al, the presence of zinc on sample A2 confirms that the C1attack causes the dissolution of aluminum, thereby exposing zinc from the metal substrateto the surface region.The extended corrosion test on sample C indicates that the amount of zincdecreases from sample Cl to sample C2, implying that the zinc originally coated on thesurface dissolves into the solution during the corrosion process. Similar phenomena arealso found on samples D to G. In addition, their amounts of phosphorus decrease to anundetectable level, except that Sample D2 has some residual amount of phosphorus.These phenomena imply that in such an corrosive condition, the ZnOx-AlOx-ZPOstructure on the surface is damaged by the C1 attack, causing zinc and phosphorus to beremoved from the surface. However, among the treatment conditions studied in thiswork, the one prepared in ZPO solution at pH=13.0 has the largest amounts of zinc andphosphorus on the surface, and likely has the strongest structure to defend against thecorrosion attack.86Table 3.7 Qualitative observations from XPS for samples after immersion in NaC1solution.Sample Zinc* Phosphorus* Muminum’A2 Detected Not detected Metallic Al not detectedAmount increasedC2 Detected Not detected Metallic Al not detectedAmount decreasedD2 Detected Detected Metallic Al detectedAmount decreased Amount decreasedE2 Detected Not detected Metallic Al detectedAmount decreased Amount decreasedF2 Detected Not detected Metallic Al detectedAmount decreased Amount decreasedG2 Detected Not detected Metallic Al detectedAmount increased Amount decreased* The decrease or increase in the amount of zinc or phosphorus is compared with the samesurface before immersion in NaC1 solution.‘‘ AlOx always detected.87The detection of metallic Al on samples D2, E2 and F2 indicates the ZnOx-AlOxZPO layer on Al surface acts as a physical barrier [3.20], which is sacrificed in thecorrosion attack, in the corrosion attack, thus slowing down the dissolution and oxidationof aluminum. Figure 3.11 shows the contrast between the blank Al surface and the ZPOtreated surface prepared at pH13.0. Figures 3.11(a) and (c) are the surfaces of samplesA and C respectively. Before the corrosion test, they both have a shiny and silver-likeappearance. Figure 3.11(b) and (d) shows the two surfaces after corrosion test. Thesurface of sample A2 turns brown and rough, while the surface of sample D2 only changesto a slightly yellow color with retention of its shiny appearance. The protected surface isclearly less corroded than the blank aluminum.The significant increase of zinc content on sample G after corrosion presumblyrelates to its coating structure. The detection of metallic Al on sample G2 suggests thatthe coating on sample G is so ineffective, and the dissolution is so rapid that the oxide filmattains only limited thickness. It is believed that the zinc from the treatment process hasbeen removed and that the zinc detected on sample G2 comes from the alloy substrate.Possibly, the C1 ions induce the diffusion of zinc from the bulk to the surface region.88(a) (b)(c) (d)rFigure 3.11 Photographs taken from surfaces after the 5 hours corrosion tests, (a)sample A, blank Al; (b) sample A2, sample A after corrosion; (c) sample D,ZPO treated surface prepared at pH13.O; (d) sample D2, sample D aftercorrosion.89Chapter 4 Concluding Remarks and Future Work4.1 Concluding RemarksXPS studies on 7075-T6 aluminum surfaces treated in ZPO solution providevaluable information about the nature and the chemical properties of the coating [4.1].Through these studies, it is concluded that zinc phosphate can provide an effective coatingon aluminum alloy, although the results depend strongly on the coating pH. Also, biasedXPS analysis reveals that ultrasonic rinsing on the surface after the ZPO treatment isneeded to remove any physically adsorbed coating compounds from the surface, so as toensure that a well-established coating is obtained.The reactions occurring during the ZPO treatment are complex, because differentprocesses, e.g. coating, etching, oxidation and bulk diffusion, occur concurrently but indifferent proportions under the various pH environments. The coatings formed in ZPOsolution at pH=5.0, 10.5 and 13.0 have a mixed ZnOx-AlOx-ZPO structure, even thoughtheir atomic compositions and distributions are different. These variations are related tothe different rates of etching and precipitation processes occurring at the alloy surfaces.The coating formed in natural ZPO solution (pH=6.6) is a ZnOx-AlOx mixed material,while the coating formed in ZPO solution at pH=3.5 exists as a thin ZPO-like compound.Among these pH conditions, the largest amounts of zinc and phosphorus were detected(for 8 = 90°) on the surface treated at pH= 13.0.Corrosion studies show that the treated surfaces provide better corrosion controlthan the untreated one. The surface prepared at ZPO natural pH provides corrosionprotection in 3.5 % NaCI solution for 2 hours. However, this surface is not strong enough90to defend for the extended 5-hour corrosion test. Among the treated surfaces underinvestigation, the one prepared at pH=13.O is likely to give the best corrosion protection inthe experimental corrosive environment. The studies show that zinc and phosphorus onthe coating will dissolve and lose out from the surface under the C1 ions attack. Thus, thecoating acts as a physical barrier sacrificing itself to slow down the oxidation anddissolution of aluminum under the corrosion attack. In terms of the amounts of zinc andphosphorus in the coatings from different pHs, it is likely that the greater the amounts ofzinc and phosphorus on the coatings, the better is the corrosion control.4.2 Future WorkFuture reseach is suggested within the following approaches:(1) Testing of other phosphate compoundsPhosphate compounds with other cation components, e.g. calcium phosphate,should be studied in order to determine the most effective compounds for coating andcorrosion control on aluminum alloy materials. Also, these studies may be able to helpinvestigate the effects of both the cation and anion components of the compounds on thecoating process, and thus give a better understanding about the mechanism.(2) Testing of organosilanes and phosphates mixturesOrganosilanes are used as additives in paint primers to improve the adhesion ofpaint on metal surface. Previous studies from our laboratory [4.2, 4.3] on the coating oforganosilanes on aluminum indicate the presence of direct Si-O-Al bonding on the coatedsurface. It is suggested that the study of mixtures of silanes and phosphate compounds,91which is more close to the “real life” situation, should give a better understanding on theircombination effect for both coating and corrosion protection.92References[1.1] D.Altenpohl, “Aluminum Viewed from Within” (Aluminum-Verlag, Düsseldorf:1982) Chapter 1.[1.2] R.C.Weast, Ed., “Handbook of Chemistry and Physics” (CRC Press Inc.,Cleveland, 1977) p.B-85.[1.3] B.C.Craig, Ed., “Handbook of Corrosion Data” (ASM International, Metals Park,1989) p.16.[1.4] F.King, “Aluminum and its alloys” (Ellis Horwood Ltd., Chichester, 1987) Chapter5.[1.5] S.Wernick, R.Pinner, P.G.Sheasby “The Surface Treatment and Finishing ofAluminum and its Alloys. Vol. I” (Finishing Publications Ltd., Teddington, 1987)Chapter 1.[1.6] P.A.Schweitzer, Ed., “Corrosion and Corrosion Protection Handbook” (MarcelDekker Inc., New York, 1989) Chapter 1.[1.7] T.H.Nguyen and R.T.Foley, J. Electrochem. Soc. 127(1980)2563.[1.8] Ref [1.1] p.185.[1.9] Ref [1.5] ChapterS.[1.10] I. Suzuki, “Corrosion-Resistant Coatings Technology” (Marcel Dekker Inc., NewYork, 1989) Chapter 5.[1.111 K.Hatanaka, M.Fukui, Y,Mukai, K.Toyose, Kobelco Tech. Rev. 6(1989)28.[1.12] T.Foster, G.N. Blenkinsop, P.Blattler and M.Szandorowski, J. Coat. Tech.63(1991)91.[1.13] F.de.L.Fragata and J.E.Dopico, Surf Coat. Internat. 74(1991)92.[1.14] T.Foster, G.N.Blenkinsop, P.Blattler and M.Szandorowski, J. Coat. Tech.63(1991)101.[1.15] G.Adrian and A.Bittner, J. Coat. Tech. 58(1986)59.[1.16] W.J.Van Ooij and A.Sabata, Surf Coat. Tech. 39/40(1989)667.[1.17] G.N.Bhar, N.C.Debnath and S.Roy, Surf Coat. Tech. 35(1988)171.93[1.18] N.Satoh and T.Minami, Surf Coat. Tech. 34(1988)331.[1.191 A.Turuno, K.Toyose, H.Fujimoto, Kobelco Tech. Rev. 11(1991)14.[1.20] W.Rausch, “The Phosphating of Metals” (Finishing Publications Ltd., Teddington,1990) Chapter 3.[1.21] D.B.Freeman, “Phosphating and metal pre-treatment” (Industrial Press Inc., NewYork, 1986) Chapter 2.[1.221 F.J.Spaeth, Mod. Paint. Coat. 74(1984)49.[1.23] D.J.O’Conuor, B.A.Sexton and R.St.C.Smart, “Surface Analysis Methods inMaterials Science” (Springer-Verlag, Berlin, 1992) Chapter 1.[1.24] D.Briggs and M.P.Seah, Ed., “Practical Surface Analysis” (John Wiley and SonsLtd., West Sussex, 1990).[1.25] Ref [1.231 Chapter 5.[1.26] A.W.Czanderna and D.M.Hercules, Ed., “Ion Spectroscopies for SurfaceAnalysis” (Plenum Press, New York, 1991) Chapter 6.[1.27] R.B.Ross, Ed., “Metallic Materials Specification Handbook” (Chapmani & Hall,London, 1992).[1.28] P.E.Francis and T. S.Lee, Ed., “The Use of Synthetic Environments for CorrosionTesting, ASTM STP 970” (American Society for Testing and Materials,Philadelphia, 1988) p.217.[2.1] J.D.Andrade, “Surface and Interfacial Aspects of Biomedical Polymers. Vol. 1”(Plenum Press, New York, 1985) Chapter 5.[2.2] H.Hertz, Anna!. Phys. 31(1887)982.[2.3] A.Einstein, Annal. Phys. 17(1905)132.[2.4] C.R.Brundle and M.W.Roberts, Proc. R. Soc. A331(1972)383.[2.5] B.L.Gabriel, “SEM: A User’s Manual for Materials Science” (American Society forMetals, Metals Park, 1985) p.55.[2.6] Ref [2.1] p.114.[2.7] Ref [1.24] p.52.[2.8] S.Tanuma, C.J.Powell and D.R.Penn, Surf Interface Anal. 11(1988)577.94[2.9] S.Tanuma, C.J.Powell and D.R.Penn, Surf Interface Anal. 17(1991)911.[2.101 S.Tanuma, C.J.Powell and D.R.Penn, Surf Interface Anal. 17(1991)927.[2.11] D.R.Penn, J. Elect. Spect. Rd. Pheno. 9(1976)29.[2.12] M.P.Seah and W.A.Dench, Surf Interface Anal. 1(1979)2.[2.131 C.D.Wagner, W.M.Riggs, L.E.Davis, J.F.Moulder and G.E.Muilenberg, Ed.,“Handbook of X-ray Photoelectron Spectroscopy” (Perkin-Elmer Co., EdenPrairie, 1979) p.83.[2.14] D.Briggs, Ed., “Handbook of X-ray and Ultraviolet Photoelectron Spectroscopy”(Heyden & Son Ltd., London, 1977).[2.15] Ref [2.1] p.123.[2.16] Ref [2.13] p.84.[2.17] C.D.Wagner, Faraday Discuss. Chem. Soc. 60(1975)306.[2.18] Ref[2.13] p.21.[2.191 D.Briggs and M.P.Seah, “Practical Surface Analysis” (John Wiley and Sons,Chichester, 1983) p.134.[2.201 Y.L.Leung, M.Y.Zhou, P.C.Wong, K.A.R.Mitchell and T.Foster, Appi. Surf Sd.59(1992)23.[2.21] A.J.Pertsin and Yu.M.Pashunin, Appi. Surf Sci. 44(1990)171.[2.22] Max 200 User Manual (Leybold, KOln, Germany).[2.231 Ref [1.24] Chapter 2.[2.24] J.M.Walls, Ed., “Methods of Surface Analysis” (Cambridge University Press,Cambridge, 1989) Chapter 3.[2.251 F.A.White and G.M.Wood, “Mass Spectrometry Applications in Science andEngineering” (Wiley-Interscience, New York, 1986) p.138-142.[2.26] Ref [1.23] p.79-95.[2.27] 1985 Annual Book of ASTM Standards, Section 3, Metals Test Methods andAnalytical Procedures, p.88-93 and p.176-185.[2.28] D.A.Skoog, “Principles of Instrumental Analysis” (CBS College Publishing,Philadelphia, 1985)p.250-91.95[3.11 D.A.Jones, “Principles and Prevention of Corrosion” (Macmillan Publishing Co.,New York, 1992).[3.2] C.S.Fadley, R.J.Baird, W.Siekhaus, T.Novakov and S.A.L.Bergstrom, I. Electro.Spect. 4(1974)93.[3.3] J.Massies and J.P.Contour, J. App!. Phys. 58(1985)806.[3.4] Ref [2.12][3.5] Y.L.Leung, M.Y.Zhou, P.C.Wong, K.A.R.Mitchell and T.Foster, Appi. Surf Sci.59(1992)23.[3.6] Ref. [2.14] Chapter 7.[3.7] B.D.Ratner, T.A.Horbett, D.Shuttleworth and H.R.Thomas, J. Colloid Interf Sci.83(1981)630.[3.8] W.F.Linke, “Solubility: Inorganic and Metal-organic Compounds, Vol. II”(American Chemistry Society, 4th ed., 1965) p.1680.[3.9] F.A.Cotton and G.Wilkinson, “Advanced Inorganic Chemistry” (John Wiley &Sons Inc., New York, 1972) Chapter 9.[3.10] G.F.Liptrot, “Modern Inorganic Chemistry” (Bell & Hyman Ltd., London, 1983)Chapter 17.[3.11] Ref [3.10] Chapter25.[3.12] D.E.C.Corbridge, “Studies in Inorganic Chemistry 10: Phosphorus” (ElesvierScience Publishers B.V., Amsterdam, 1990) Chapter 3.[3.13] Ref [1.2].[3.14] F.J.Spaeth, Mod. Paint & Coat. 74(1984)49.[3.15] T.H.Nguyen and R.T.Foley, 3. Electrochem. Soc. 127(1980)2563.[3.16] Z.A.Foroulis and M.J.Thubrikar, J. Electrochem. Soc. 122(1975)1296.[3.17] T.H.Nguyen and R.T.Foley, J. Electrochem. Soc. 126(1979)1855.[3.18] R.T.Foley and T.H.Nguyen, J. Electrochem. Soc. 129(1982)464.[3.19] D.R.Arnott, N.E.Ryan, B.R.W.Hinton, B.A.Sexton and A.E.Hughes, Appl. SurfSci. 22/23 (1985)236.96[3.20] W.M.Morgans, “Outlines of Paint Technology” (Haisted Press, New York, 1990)Chapter 4.[4.1] W.F.Heung, Y.P.Yang, P.C.Wong, K.A.R.Mitchell and T.Foster, J. Mat. Sci.(accepted and in process).[4.2] Ref [3.5].[4.3] Y.L.Leung, Y.P.Yang, P.C.Wong, K.A.R.Mitchell and T.Foster, J. Mat. Sci. Lett.12(1993)844.97


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items