6063 aluminyum alaşımlarının korozyon direnci üzerine anodizasyon öncesi yüzey işlemlerinin ve AA anodizasyonla üretilmiş ince oksit filmlerinin etkisi
- Global styles
- Apa
- Bibtex
- Chicago Fullnote
- Help
Abstract
ÖZET Toz boya kaplı alüminyum alaşımlarına, boya altına iyi yapışma sağlanması amacıyla uygulanan kromatlanma işlemi insan sağlığı ve çevre açısında zehirli atıklar içermesi nedeniyle, bu işlem yerine ince (4-5um) anodik oksit film uygulaması umut vadeden işlemlerden biri olarak göze çarpmaktadır. Ancak literatürde ince anodik oksit filmlerinin korozyon özellikleri ile ilgili sistematik bir çalışma mevcut değildir. Bu nedenle, bu tez çalışmasının amacı ince (4-5 um) anodik oksit filmlerinin korozyon özelliklerini incelenmek olarak belirlenmiştir. Bu incelemede, toz boyalı alüminyum alaşımlarında ortaya çıkan en önemli bölgesel korozyon türlerinden olan iplikçik korozyonunu da etkileyen şu temel etmenler göz önüne alınmıştır: a) metallerarası faz partiküllerinin etkileri ve kontrolü, b) Çeşitli çözeltilerde ve farklı anodizasyon yöntemleriyle hazırlanmış oksit filmlerinin toplam korozyon direncine etki eden parametrelerin incelenmesi ve oksit filmlerinin korozyon direncinin geliştirilmesi. Bu çalışmada kullanılan malzeme 6063 alüminyum alaşımıdır. Metallerarası faz partiküllerinin kontrolü için anodizasyon öncesi uygulanan dağlama (% 5 NaOH (55°C)) ve dağlama ürünlerinin temizlenmesi (1/3 HN03 (Oda Sıc.)) işlem süreleri (3-1 5dk aralığında) değiştirilmiştir. Anodizasyon numuneleri % 10 H2S04, 20°C ve 25 V da, alternatif akım (AA) ve doğru akım (DA) anodizasyonu kullanılarak hazırlanmış ve anodik oksit filmlerinin direncine etkisinin incelenmesi için anodizasyon çözeltisine molibdat iyonu (5.7,8,40 g/l Na2Mo04) ilave edilmiştir. Deneysel çalışmalarda yüzeylerin korozyon direncini saptamak için elektrokimyasal empedans spektroskopisi (EIS), yüzeylerdeki elektrokimyasal olarak aktif metallerarası partiküllerin saptanması için bakır dekorasyonu, yüzeylerin incelenmesi için taramalı elektron mikroskobu (SEM) kullanılmıştır. Kostik dağlama işlem süresinin artmasıyla, alüminyum matris çözünürken AlFeSi partiküllerindeki alüminyum da çözünmüş ve bu, partiküllerde demir ve silis zenginleşmesine yol açmıştır. Bu yüzeylerin 0.1 N NaCI çözeltisindeki korozyon direnci artan işlem süresi ile azalmıştır. Dağlanmış yüzeylerin 1/3 HN03 içinde işlem süresinin artması, yüzeylerdeki AlFeSi partiküllerindeki demirin tercihli çözünmesine neden olmuş ve silisyumca zenginleşmiş partikül sayısı artmıştır. Nitrik asit ile muamele görmüş yüzeylerin 0.1 N NaCI içindeki korozyon direnci ise artan işlem süresiyle artmıştır. Bakır dekorasyonu AlFeSi partiküllerinin elektrokimyasal olarak aktif bölgelere yol açtığını, silisyumca zenginleşmiş partiküllerin aktif olmadığını göstermiştir. 4-5 xm kalınlığındaki ince anodik oksit filmlerinin, filmdeki partiküllerin tümünü kaplamadığı gözlenmiştir. Bakır dekorasyonu ile oksit filminde tamamen kaplanmamış AlFeSi partiküllerinin elektrokimyasal aktif bölgeleri oluşturduğu gözlenmiştir. Anodizasyon öncesi yüzey işlemleri, ince AA oksit filmlerinin 0.01 N HCI çözeltisindeki korozyon direncini etkilemiştir. Kostik dağlama işleminden sonra daha uzun süre nitrik asit içinde tutulmuş numune yüzeylerindeki oksit filmlerinin daha iyi korozyon direnci verdiği saptanmıştır. Bu numunelerde AlFeSi partiküllerinin çok daha az sayıda olduğu belirlenmiştir. Sülfürik asit çözeltilerine molibdat ilavesi ise bu çözeltilerden elde edilen AA oksit filmlerinin büyüme hızını arttırırken, 0.01 N HCI'de korozyon direncinin azalmasına yol açmıştır. Sülfürik asit çözeltisinde üretilmiş AA anodik filmleri, aynı şartlarda ve aynı kalınlıkta üretilmiş DA anodik oksit filmlerinden daha yüksek korozyon direnci göstermişlerdir. Ancak, AA anodizasyonunda trafonun iki ucundan elde edilen numuneler farklı hızlarda korozyona uğramışlardır. XI THE EFFECT OF PRETREATMENT AND THIN LAYER SULFURIC ACID AC ANODIZATION OF 6063 ALUMINUM ALLOYS ON ITS CORROSION RESISTANCE SUMMARY The industrial use of powder coated or painted aluminium alloys has been steadily increasing. The main purposes of powder coatings or paintings are to provide an appeal to the surface and also to improve its corrosion resistance. Aluminium alloys are chromate passivated prior to powder coating to provide adhe sion of powder coating to the substrate. Although, for a long, time it was accepted that chromate treatment eliminated filiform corrosion of painted surfaces, recent surveys have shown that there are increasing evidence of filiform corrosion and underfilm corrosion of chromated and painted or powder coated aluminium prod ucts in environments combining marine and industrial atmospheres. Hence, chro mate conversion coatings are not sufficient in preventing filiform corrosion com pletely. Additionally, hexavalent chromium is not safe for humans as well as for the environment and extensive work are carried out to replace it with safer substi tutes. Filiform corrosion (FFC) which occurs at the metal/coating interface affects deco rative appearance of coated materials. Due to an abrupt increase of filiform corro sion failures in Western Europe, recent studies have focused on the causes and remedies for it. The most important factors affecting filiform corrosion are adhesion between coating and substrate, defects in the coatings and corrosion resistance of the substrate or pretreated surface. The higher the corrosion resistance of the sub strate, the higher filiform corrosion resistance of the powder coated aluminium al loys. The formation of a thin oxide film produced by anodisation on aluminium alloys as an underlayer before powder coating may be a potential candidate in the future to substitute chromate pretreatment. It has been also reported that a thin layer of direct current (DC) anodised oxide film remarkably improved filiform corrosion re sistance of powder coated aluminium alloy substrate. Thin anodised layers provide both good adhesion to the coating and protection to aluminium alloys. It was also reported that a thin layer of DC anodised film was very effective in preventing the corrosion, especially at the propagation stage of FFC because of higher corrosion resistance of the oxide film than chromate coatings. However, chromate coating was more effective at the initiation stage since dissolved chromate ions dynamically passivate active sites during initiation stage of FFC. Molybdate ions has been widely exploited as a promising corrosion inhibitor, and a passivating agent to replace environmentally hazardous hexavalent chromium. These ions are specifically used in the prevention of several types of localised cor rosion, which develop in chloride containing media. Besides, addition of molybdate ions into sulphuric acid solutions during alternative current (AC) anodisation elimi nates yellowish coloration of the oxide films and intensive H2S odour, which are XIIdisadvantages of oxide films produced by AC anodisation for decorative purposes. It is known that anodised oxide films contain electrolyte anions in their chemical composition. In the case of AC anodisation in the molybdate containing solutions, the oxide films may contain also molybdate ions. Such an oxide film, in addition to the inherently higher corrosion performance of the oxide layer, may have passivat- ing properties during activation of the defects in the film. Therefore, it may be worth to examine corrosion properties of the oxide films produced in molybdate contain ing solutions. Preliminary studies on thin AC anodised oxide films showed that the oxide thick ness of 4-5n,m was not sufficient to cover all intermetallic particles in the oxide film. Intermetallic particles play an important role both in the corrosion of aluminium al loys and at the initiation stage of filiform corrosion of the coated aluminium alloys pretreated in only caustic solutions. In the case of thin anodised films, uncovered particles may also affect corrosion properties of the film and thus, filiform corrosion properties of the coated aluminium alloys. However, there is no systematic work on the effect of such particles on the corrosion properties of thin anodised films produced by neither AC nor DC methods. Therefore the aim of this thesis was to investigate the effect of surface pretreat- ments such as etching and desmutting on the corrosion performance of AC ano dised thin oxide films on commercial 6063 aluminium alloys and also to examine the effect of molybdate ion addition into the anodisation solution on the corrosion performance of the thin oxide layers. Examination of the corrosion properties of the thin anodised oxide films on 6063 aluminium alloys in dilute acidic chloride containing solutions, which are common solutions to inoculate filiform corrosion, may be very important from the points of scientific and technological view, if thin anodic oxide layers will be considered to replace chromate underlayer before powder coating; and this subject has not been investigated yet. Before starting the above study, a detailed investigation on the effects of anodisation pretreatments, such as caustic etching and desmutting in the nitric acid solutions, on the intermetallic particles that exist in 6063 aluminium alloys must be carried out. Besides, it would be worth while to examine the effect of molybdate ions addition into AA sulphuric acid anodisation solution, on both, AA anodisation process and corrosion resistance of AA oxide films. In the first part of this thesis, in order to investigate the possibility to eliminate in termetallic particles causing electrochemically active regions, the effect of anodi sation pretreatments such as alkaline etching and acidic desmutting treatments on the intermetallic particles that exist in 6063 aluminium alloy was examined. 6063 series aluminium alloy ( 0.158 % (w) Fe, 0.365 Si, 0.365 Mg) was used during the experiments since in architectural applications, it is the most widely used alu minium alloy for powder coatings. The alloy was extruded in strips of 1000x50x1 mm and submitted to T5 heat treatment. The experimental samples (50x40x1 mm in size) were cut from these strips. In order to investigate the effect of anodisation pretreatments on the intermetallic phase particles in 6063 aluminium alloy, the following experiments were conducted: First, mechanically polished samples were examined to determine original compo sition and distribution of the particles in the alloy. Then, two groups of as received samples were etched for 3 min., 5 min and 10 min. following degreasing in a solu- XIIItion containing 5% Na2C03+5% Na3P04 at 55°C for 5 min.. One group of samples were ultrasonically cleaned in distilled water after etching treatment for 3min., 5 min and 10 min.. Three groups of 5 minute-etched sample were desmutted for 3 min., 10 min. and 15 min.. Following each stage the samples were rinsed with distilled water and dried in cool air current. The solutions of 5% NaOH (55°C±2) and 1/3 HN03 (room temperature) were used for etching and desmutting treatments re spectively, since the most common industrially used etching and desmutting solu tions are sodium hydroxide and nitric acid solutions. Electrochemical impedance spectroscopy (EIS) is widely used technique to com pare various pretreatments before painting and coating since it allows to charac terise the electrochemical and corrosion behaviour of the pretreated surfaces in aqueous solutions. Therefore, this technique was applied to determine the corro sion resistance of the pretreated samples. After etching, etching+ultrasonic cleaning and desmutting treatments, one group of the samples were immediately submitted to EIS experiments in 0.1 N NaCI solutions at Ecorr and at room tem perature. The second group of etched, etched+ultrasonically cleaned and des mutted samples were examined with scanning electron microscope (SEM), equipped with X-ray EDS analysis. Third group of desmutted samples were copper decorated to identify active and nonactive particles on these surfaces. The copper decoration procedure was as follows: The sample was kept in 100 ml of 0.1 N NaCI solution for 15 min., then, 0.5 ml of Cu++ solution (35 g/l CuS04.5H20) was added into the beaker at end of 15 min. After holding the sample for 60 sec. in the solu tion with Cu++ ions, it was removed from the beaker and rinsed with distilled water, then dried in cool air current. Following decoration, the samples were examined with SEM. Two types of particle composition were determined on the mechanically polished sample: a) particles with Mg and Si, b) particles with Al, Fe and Si. They were in majority and their shapes were acicular with a length of 7-8 /im and width of 1-2 xm These particles were distributed homogeneously and contained two different Fe/Si ratio. Etching treatment in 5 % NaOH (55°C) compared to mechanically polished one has led to higher concentration of small particles at the smut layer that covered the etched surface. Mg-Si particles became more visible after etching and their con centration continued to increase with increasing etching time while smut layer be came thicker. When the smut layer was removed from the surface by ultrasonic cleaning in distilled water, no Mg-Si particle was detectable on the surface, indi cating dissolution of these particles during etching treatment and their accumula tion in the smut layer. After ultrasonic cleaning, it was observed that particle con centration on the clean surface did not remarkably change with increasing etching time. However, composition of AlFeSi particles differed from the one of mechani cally polished samples. Etching treatment in 5% NaOH at 55°C lead to not only selective dissolution of aluminium matrix but also selective dissolution of aluminium in the AlFeSi particles. The results of EIS experiments of as-received and etched and etched+ultrasonically cleaned samples showed that corrosion resistance of the as- received surfaces in 0.1 N NaCI solutions decreased with etching treatment in 5 % NaOH (55°C), possibly due to the removal of protective air formed oxide layer on the as-received samples. Furthermore, increasing etching time led to decrease in the corrosion resistance of the etched and etched+ultrasonically cleaned samples. XIVDecreasing corrosion resistance of the etched surfaces was attributed to iron en richment in the AlFeSi particles due to selective dissolution of aluminium in the par ticles. Smut layer on the etched samples was removed very easily by the desmutting treatment in 1/3 HN03 solutions. The particle population on the desmutted sam ples did not change compared to etched+ultrasonically cleaned samples. Further more, no remarkable change on the particle population was observed depending on desmutting time. However, SEM examination of the samples revealed that the number of AlFeSi particles decreased with increasing desmutting time while the number of Al-Si particles increased. Majority of the particles on the 15 min des mutted sample surfaces were composed of only Al-Si, which was probably the re mains from the selective dissolution of Fe in the ternary AlFeSi particles during desmutting. In order to determine electrochemically active particles, copper decoration tech nique was applied to the samples following desmutting treatment. In this tech nique, copper ions reduce at the cathodic regions and deposit as metallic copper, taking electrons produced by the dissolution of active sites. Copper deposited around all particles with Fe while no copper deposition was observed on or around the particles with Al and Si; that is, particles with Fe formed electrochemically active regions. Density of copper decorated sites was very high on the surfaces desmut ted for a short time (3 min.) and decreased with increasing desmutting time. In other words, density of particles containing Fe or electrochemically active regions decreased. These results imply that desmutting treatment not only removes the smut layer but also causes selective dissolution of Fe in Fe containing particles prominently displayed following etching treatment. EIS experiments revealed that desmutting treatment in 1/3 HN03 solution caused the corrosion resistance of 5 minute-etched samples to increase in 0.1 N NaCI solution. Furthermore, the surfaces became more resistant to chloride solutions with increasing desmutting time. Increased corrosion resistance may arise from the formation of passive film on the Al matrix in nitric acid solution. Also, the change in the number and the properties of intermetallic particles may contribute to the re sistance increase. Both etching and desmutting studies showed that it is possible to eliminate parti cles causing electrochemically active regions by controlling, at least, pretreatment time. The second part of the thesis consisted of anodisation studies. In this part, first the effect of etching and desmutting pretreatments on the corrosion behaviour of AC anodised thin oxide films was investigated. Secondly, the effect of addition of mo- lybdate ions into sulphuric acid solution on the corrosion behaviour of AC anodised thin oxide films as well as on the rate of anodisation was studied. Thirdly, corro sion resistance of AC and DC anodised thin oxide films were compared. AC and DC anodisation were conducted in 10% H2S04 (20°C) at a constant volt age of 25 V by using an autotransformer (10 A, 200 V and a frequency of 50 Hz from the main) as power supply and a DC power supply, respectively. All samples were anodised to produce 4-5 ^.m thick oxide films; then, rinsed in distilled water and dried in cool air current. After anodisation treatments, the following experi ments were conducted with the anodised samples: a) To compare corrosion per- xvformance of anodised samples during 8 day exposure to 0.01 N HCI solution which is widely used to inoculate filiform corrosion, EIS experiments were conducted at the end of 20 min, 1 h, 20 h, 24, h, 48 h and 8 days. All experiments were carried out at ECorr, with an AC potential amplitude of 10 mV and in the frequency range of 10`2-105 Hz. b) Electrochemical copper decoration technique was applied to the anodised samples to identify electrochemically active and non-active regions. Cop per decoration procedure was as follows. Decoration was carried out immediately following anodisation (without drying). The copper decoration solution contained 100 ppm Cu++ at pH=1 (fixed with HCI). Using the specimen as a cathode, ano dised aluminium alloy surfaces were decorated by applying a 2 V cell voltage for 20 sec, then the decorated samples were examined with SEM. Also, the samples anodised and exposed to 0.01 N HCI solution were examined with SEM. The AC anodised oxide films of 4-5j-m thickness were not sufficient to cover large intermetallic phase particles. However, the oxide film was grown coherently with these particles and no dissolution region around them was observed. Small parti cles containing Si were embedded completely in the film. When copper decoration was applied to all anodised samples, Fe was detected from the copper decorated spots, indicating that AlFeSi particles led to electrochemically active regions on the anodised films. However, Si containing particles were not copper decorated. To study the effect of anodisation pretreatments on the AC anodised thin oxide films, two types of sample were prepared : etched for longer time (10 min.)+desmutted for shorter time (5 min.) and etched for shorter time (5 min.)+desmutted for longer time(10 min.). EIS experiments conducted in 0.01 N HCI solution showed that the oxide film resistance of the anodised samples were affected by the surface treatment. When the desmutting time in 1/3 HN03 was kept longer, the surface resistance of anodised samples at the end of 20 minute exposure to 0.01 N HCL solution became higher. The oxide film resistance of ano dised surfaces became lower if desmutting time was shorter because of higher density of the Fe rich particles. On these surfaces, the density of copper decorated spots, indicating to the density of Fe containing particles, was higher than those desmutted for longer time. The results of these experiments revealed that uncovered Fe containing intermet allic particles in the thin anodised films affects film resistance; however their num ber can be controlled by optimising the surface pretreatments prior to anodisation. To investigate the effect of addition of molybdate ions to sulphuric acid solutions on the corrosion performance of AC anodised films, the samples were AC anodised in 10% H2SO4 solution containing 5.7, 6 and 40 g/l sodium molybdate. Addition of molybdate ions at all concentrations used in this study, was able to eliminate the evolution of sulphurous smell and to increase the anodisation rate. Several spots with a poor coherency to the oxide film were observed on the oxide film produced in molybdate containing sulphuric acid solutions, which were AlFeSi particles. These particles were copper decorated during copper decoration test. Besides, some other clusters with different structure than the oxide film were also determined. However, EDS analysis did not reveal any composition difference between the oxide film and these clusters and no copper decoration was deter mined on or around these clusters; however dissolution of the oxide film around the clusters was observed. XVIUsing different analytical techniques, it was not possible to determine clearly whether the oxide films produced in molybdate containing solutions contain molyb- date ions or molybdate compounds, since the results were contradictory. For ex ample, EDS analysis of the oxide films at an accelerating voltage of 30 kV re vealed some Mo (0.62-1.29 % (w)). However, the results of auger electron spec troscopy (AES) analysis revealed no molybdenum both at the surface and up to the depth of 24 nm from the surface; but minor Mo was determined with X-ray fluores cence spectroscopy, indicating that Mo exists in the film. EIS results of the oxide films exposed to 0.01 N HCI solutions during 8 days showed that the AA films produced in molybdate containing solutions had dissolved faster than the one produced in molybdate free solutions. Also these oxide films had higher copper decorated spots on the AlFeSi particles. Lower corrosion resis tance of the oxide films produced in molybdate containing solutions was attributed to the fact that: a) these oxide films had higher number of electrochemically active spots, which were clusters and AlFeSi particles with poor coherency with the oxide film, b) also, probably the oxide films produced in molybdate containing solutions had a different composition than that of produced in molybdate free sulphuric acid so lution; since molybdate ions eliminate sulphate reduction and sulphide and sul phur formation in the oxide films during cathodic cycle of AC anodisation. Corrosion performance of the AC and DC oxide films produce in 1 0% H2S04 solu tions were compared. EIS experiments conducted at the end of 20 min., 1, 20, 24 and 48 h. in 0.01 N HCI solutions revealed that AC anodised oxide films dissolved in a longer time than DC anodised films. In other words, The AC oxide films showed higher corrosion resistance than the DC oxide films in 0.01 N HCI solu tions. The AC oxide films with a thickness of 4-5 nm had lower density of uncovered Al FeSi particles than DC anodised oxide films with the same thickness. Also copper decoration tests showed that the AC oxide films revealed lower number of deco rated spots on the AlFeSi particles. The higher corrosion resistance of the AC anodised oxide films may be attributed to two properties of the film: a) the AC oxide films had lower density of electrochemically active AlFeSi particles because of their capability to cover most of AlFeSi particles, b) the chemical composition of AC oxide films is different than that of DC films. They contain sulphide up to % 40-60 of total sulphur together with sulphate in the oxide; but DC oxide films contain sulphur in the form of sulphate only. In the AC anodisation, the oxide films produced on the samples connected to two poles of the transformer showed different and random corrosion properties. The conclusions of this thesis are given as following: A) Anodisation pretreatment studies in caustic etching and acid desmutting solu tions: XVII1) Etching in 5% NaOH (55°C) increased the exposed surfaces of intermetallic particles on the aluminium surface by dissolving the Al matrix selectively and exposing the intermetallics. Etching dissolved the particles with Mg and Si. 2) Desmutting in 1/3 HN03 (R.T.) dissolved selectively Fe, leaving mostly Si be hind. Copper deposited around Fe containing particles, indicating their electro- chemically active behaviour; while no copper deposited on or around only Si containing particles. 3) The corrosion resistance (determined by EIS) in 0.1 N NaCI solutions of the etched samples decreased considerably with etching time while it increased with desmutting time. B) Anodisation studies 1) The oxide film thickness of about 4-5 ^m produced in molybdate containing and molybdate free 10% H2S04 solutions by using both AC and DC anodisation did not provide a film thick enough to cover all particles on the alloy surface. 2) On the AC anodised oxide films, copper deposited onto the protruding Fe con taining particles, indicating electrochemically active sites, while there was no deposition on or around the particles enriched in Si. 3) Desmutting and etching affected the resistance of AC anodised surfaces. The anodised samples desmutted for longer time (10 min.) exhibited higher corrosion resistance in 0.01 N HCI solution due to lower density of Fe rich particles caus ing electrochemically active sites. 4) During AC anodisation the presence of molybdate ions in 10% H2S04 solution increased the oxide film formation rate and prevented the evolution of sulphur ous smell. Oxide film formation rate increased with increasing molybdate ion concentration. 5) Molybdate addition to anodisation solution did not improve the corrosion resis tance of the oxide films to 0.01 N HCI solutions. The resistance of the films formed in molybdate containing sulphuric acid solutions decreased faster in 48 hour exposure to 0.01 N HCI solution than the resistance of the films formed in molybdate free sulphuric acid solution. Similarly, porous side of the oxide film produced in molybdate containing solutions dissolved faster than the porous side of the films produced in molybdate free solutions. 6) Molybdate containing sulphuric acid anodised oxide films had more active defect sites such as incoherent AlFeSi particle/oxide interface. 7) At the end of 8 day exposure to 0.01 N HCI solutions, the AC oxide films pro duced in both molybdate containing and molybdate free sulphuric acid solutions had dissolved completely. Shallower and larger pits were formed on the sub strate following dissolution of the AC oxide films produced in the molybdate containing solutions, while smaller and deeper pits had developed on the sub strates after dissolution of the oxide films produced in molybdate free solutions. 8) The AC films produced in 10% H2SO4 solution had lower density of copper decoration on AlFeSi particles than DC films produced in the same solution. In other words, the DC films with the same thickness of AC films had higher num ber of uncovered AlFeSi particles by the oxide film. 9) EIS results of the AC and DC films with the same thickness exposed to 0.01 N HCI solution for 8 days showed that the DC films dissolved faster. XVIII
Collections