• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Eurocorr 2004 - Part 2

Posted on: Wednesday, 8 June 2005, 03:00 CDT

Eurocorr 2004, the annual conference of the European Federation of Corrosion, took place in Nice, France on 12-16 September last year. Part 1 of this report was published in Corros. Eng. Sci. Technol., 2004, 39, (4).

Session 1: inhibitors

WP1 (inhibitors) was a large session chaired by G. Schmitt, the Chairman of EFC Working Party 1 (corrosion and scale inhibition) and F. Moran. One reason for its size was because a joint session was organised with WP 13 ( oil and gas) and the papers from the joint session have been included in the inhibitors group.

Inhibitors for oil and gas production

The joint session ran all day on the Tuesday and had 13 papers. The first paper was 'Review of testing methods and standards for oilfield corrosion inhibitors' by Sankara Papavinasam et al. from the Natural Resources Canada, Ottawa, Canada. Factors influencing the hydrodynamic parameters in laboratory testing were discussed. B. Kinsella et al. from Curtin University of Technology, Perth, Australia gave a paper on 'Predicting the adsorption properties of carbon dioxide corrosion inhibitors using a structure-activity relationship'. The performance of the inhibitors in curbing the corrosion of mild steel in a brine electrolyte was fitted to a Temkin adsorption isotherm. A thermodynamic analysis revealed that compounds chemisorb onto the steel electrode. In addition, molecular modelling was undertaken using PCSPARTAN Plus and HyperChem Professional. A four-parameter fit for both negative and positive charged molecules was found.

Eurocorr 2004 welcoming reception

A particularly interesting paper was 'Evaluation of local energy densities in disturbed flow: a new approach to characterise inhibitor efficiencies to mitigate erosion corrosion'. This was given by one of the chairmen, G. Schmitt from lserlohn University of Applied Sciences, lserlohn, Germany. Electrochemical current noise (ECN) had been measured at microelectrodes in order to quantify local flow intensities in terms of wall shear stresses in disturbed flow. Analysis of the noise data using Wavelet transform allowed quantification of the maximum interaction energies between the flowing liquid and the solid wall. This was done using a similar 'phasing' algorithm as is applied to simulate oceanic freak waves and allowed realistic correlations between flow intensities and fracture stresses of protective scales. The near-wall freak energy densities were found to be damped significantly by adding surface active compounds at concentrations above the critical micelle concentration. This additive effect was shown to relate to interactions of supramolecular aggregates (micelles) with near-wall microturbulences.

A contribution entitled oilfield chemicals synergistic effects on the corrosion rate of L-80 steel in sea and formation waters' was by A. Al Hashem et al, from Kuwait Institute for Scientific Research, Kuwait City. The corrosion rate in the injection sea water, formation water and a 50: 50 mixture of both waters under downhole conditions was investigated in the presence of commercially available corrosion inhibitor, scale inhibitor, and biocide products. The corrosion rates were measured by linear polarisation using the rotating cylinder electrode (1000 and 2000 rev min^sup - 1^) at 80 C. In sea water the corrosion scale inhibitor and biocide scale inhibitor combinations provided the best protection at both rotation speeds. In formation water, a higher corrosion rate of L- 80 carbon steel was seen accompanying higher shear forces. In the 50: 50 mix the corrosion scale inhibitors-biocide combination provided the best protection at both rotational speeds.

The next paper was The effect of pre-corrosion and steel microstructure on inhibitors performance in CO2 corrosion' by L. D. Paolinelli et al. from INTEMA in Mar del Plata, Argentina. A carbon steel with two different microstructures (ferrite-pearlite and tempered martensite) was tested in deoxygenated 5 wt-%NaCI, saturated with CO2 at 40C, pH 6, stirred at 100 rev min^sup -1^ and containing a commercial imidazoline-based product. The performance was studied by means of electrochemical measurements employing ac and dc techniques. The inhibitor was added after precorrosion periods of 24, 48 and 72 h. Surface film characteristics were studied using SEM. The efficiency of the inhibitor was superior for tempered martensite compared with ferrite-pearlite. Precorrosion decreases efficiency and its impact is also microstructure dependent.

A paper entitled The inhibitors of corrosion and hydrogen entry into carbon steel in the H^sub 2^S and CO2 media' was delivered by L. E. Tsygankova from Derzhavin State University in Tambov, Russia. Two Mannich's Phenol Bases were used: MPB 9 and MPB 9-20 in HCI solutions (0.005-0.1 mol L^sup -1^) saturated with CO2 (1-2 atm) and H^sub 2^S (50-400 mg L^sup -1^) separately and together. Both bases decreased hydrogen entry into the steel membrane. Another paper from Russia was 'Biotesting integral toxicity of corrosion inhibitors, biocides and oil hydrocarbons in oil and gas processing industry' by V. A. Chugunov et al. from the State Research Center for Applied Microbiology, Obolensk.

The country of Iran gave us 'Electrochemical studies of sodium metavanadate as corrosion inhibitor of carbon steel 1020 in CO2 and H^sub 2^S saturated DEA solutions' by Alireza Aghasadeghi et al. from NIOC in Tehran. The Tafel polarisation test method was used in 30 vol.-%DEA solutions saturated with CO2 and H^sub 2^S at 65C. The inhibitive mechanism of sodium metavanadate was found to be anodic and effective dosage of the inhibitor is within 0.03 to 0.05 wt-%. Surface observations indicated that the corrosion was general. This concluded the combined inhibitors and oil and gas session.

Corrosion and scale inhibitors in industrial environments

This session ran all day on the Wednesday. The first paper was 'Modern cooling water treatment' by Frank Muller et al. from Degussa Water Chemicals in Krefeld. The presenter summarised past cooling water treatment programmes (especially chemicals) on corrosion and scale inhibition, giving their advantages and disadvantages. The inhibitors were often P containing products with high zinc and molybdate contents and used in open recirculating systems. More environmentally friendly products which show active and passive corrosion inhibition and the development of antiscalants were also covered as were water saving regulations, health aspects and a short look at the specific problem of legionella pneumophila contamination.

The second paper was 'Electrochemical study of the behaviour in corrosive media of a new friendly environmental inhibitor' by P. Bommersbach et al. from INSA de Lyon-Bt, in Villeurbanne. An inhibitor suitable for limiting corrosion of cutting tool steel was looked at. Different tests were performed on mild steel (XC38) immersed in aerated water containing 100 ppm of each salt (NaCI, Na^sub 2^SO^sub 4^ and Na^sub 2^CO^sub 3^). The results indicated 3D film formation on the steel surface. However, no detail on the exact nature of the inhibitor was given.

Another paper by G. Schmitt et al. from lserlohn University of Applied Sciences was on 'Evaluation of inhibitor efficiencies in hydraulic fluids'. Again, use had been made of the electrochemical current noise (ECN) analysis method (CoulCount) between 'identical' crevice electrodes (either both steel or steel/bronze) to assess pitting corrosion inhibitors within a few days. This method was also shown to be applicable for monitoring purposes.

A talk entitled 'Influence of flow on the synergistic effect of an inhibitive mixture used for water treatment in cooling circuits' was given by Nathalie Ochoa from CIRIMAT, Toulouse, France. A carbon steel was tested in a 200 mg L^sup -1^ NaCI solution protected by a non-toxic multicomponent inhibitor composed of 50 mg L^sup -1^ fatty amines associated with 200 mg L^sup -1^ phosphonocarboxylic acid salts. The influence of flow and immersion time was investigated by electrochemical impedance spectroscopy with a rotating disc electrode. Equivalent circuits were used to fit the experimental data. It was found that the properties of the protective layer were dependent on the electrode rotation rate and on the immersion time and this was related to the influence of flow on the anodic kinetics.

'Synergistic inhibition of corrosion of carbon steel using a new phosphonate, Zn^sup 2+^ and heterocyclic compound' came from B. V. Appa Rao of the Department of Chemistry, Andhra Pradesh University, India. 50 ppm N-(2-hydroxyl ethyl) imino bis (methyl phosphonic acid) in combination with 50 ppm Zn^sup 2+^ gave an inhibition efficiency of 92% for carbon steel. In an attempt to reduce the concentration of Zn^sup 2+^ a second synergist, viz. 1,3-dimercapto 2-amino triazole was added, which is non-toxic in nature. This at 8 ppm with 30 ppm of phosphonate and 30 ppm of Zn^sup 2+^ gave an inhibition efficiency of 90%. This ternary combination is effective in acidic and neutral environments i.e., between pH 3 and pH 7. A second paper from the same group was 'Phosphonated glycine, a new corrosion inhibitor for carbon steel in aqueous environment'. The inhibition efficiency of 40 ppm N, N-bis (phosphonomethyl) glycine (BPMG) in combination with 30 ppm Zn^sup 2+^ for carbon steel in aqueous solution containing 60 ppm chloride reached 95% in the pH range of \1 to 8.

'Study of the corrosion inhibition mechanism of new rare earth cinnamate compounds' was given by Frdric Blin from Monash University, clayton Victoria, Australia. Electrochemical studies showed that less than 900 ppm of cerium and lanthanum cinnamates significantly inhibited corrosion. The chemistry and morphology of the inhibited surfaces in relation to the corrosion inhibition mechanism were discussed.

It was back to France again for '3,5-bis(n-tolyl)-4-amino-1,2,4- triazoles as corrosion inhibitors of mild steel in acidic media: electrochemical and quantum chemical studies' by Michel Lagrene et al, from ENSCL, Villeneuve d'Ascq, France. The corrosion inhibition of mild steel by some new synthesised triazole organic type compounds was investigated in IM HCI and 0.5M H^sub 2^SO^sub 4^ using weight loss measurements, electrochemical impedance spectroscopy and polarisation curves. The ability of the n-DTAT to inhibit the corrosion of mild steel was found to be dependent on the position of the CH^sub 3^ group on the tolyl substituent. Polarisation curves studies showed that 4-DTAT is a mixed type inhibitor in 1M HCI and a cathodic type in 0.5M H^sub 2^SO^sub 4^ and appears to function through the general adsorption mode following the Langmuir adsorption isotherm model. A rather similar paper was '3,5-Bis(n-Methoxyphenyl)-4-Amino-1,2,4-Triazoles as inhibitors of corrosion of mild steel in acidic media' from M. Traisnel ef al., also from ENSCL, Villeneuve d'Ascq, France. This was followed by the 'Synergistic effect of tolutriazole and sodium carboxylates on zinc corrosion in atmospheric conditions' by C. Georges et al. from the Universit Henri Poincar, Nancy, France. The need is to avoid 'wet storage stain' appearing on galvanised steel products during transport or storage. Investigations were focused on the action of non-toxic linear sodium monocarboxylates of general formula CH3(CH2)^sub n-2^ COONa, abbreviated as NaC^sub n^ where n=7 to 11, in aerated corrosion reference water (ASTM D1384-87) using stationary electrochemical methods (R^sub p^, E^sub corr^, i=f(E) curves). To improve the efficiency of carboxylates, triazol compounds such as tolutriazol (TTA) were added. The results showed that TTA had an effect on both the cathodic reaction and the pitting potential of zinc.

Another interesting paper was 'Copper local corrosion under the joint presence of alanine and inorganic activators, e.g. Cl- ions' by S. A. Kaluzhina ef al. from Voronezh State University, Voronezh, Russia. It was shown that in alkaline solution (pH 12) with a 1 :10 ratio of organic and inorganic activators, Cl- ions had the role of activator, but if the ratio was reversed amino acid was responsible for copper local activation. Then came 'Synergism in copper corrosion inhibition by sodium dodecylbenzenesulphonate and 2- mercaptobenzoimidazole' by M. G. Hosseini ef al, from the Tarbiat Modarres University, Tehran. Inhibition of the corrosion of copper in sulphuric acid solution had been studied using electrochemical impedance spectroscopy (EIS) and Tafel polarisation measurements. For (2-MBI), a monotonous increase in inhibition efficiency was observed as a function of concentration. For SDBS, however, an optimum in the inhibition efficiency had been observed which was ascribed to the formation of hemi-micellar aggregates that provoke inhibitor desorption from the metal/solution interface at higher concentration. This was followed by 'Effect of boron on the IGSCC behaviour of alloy 600 in a caustic solution' by Yongsun Via et al, from the Korea Atomic Energy Research Institute, Daejeon, South Korea. The intergranular stress corrosion cracking susceptibility was evaluated by slow strain rate tests (SSRT) and the electrochemical properties of the alloy in solutions containing several candidate inhibitors were measured using potentiodynamic polarisation measurements. In 40%NaOH, severe intergranular cracks were observed on the surface after the SSRT tests. However, boron- containing compounds like CeB6, NiB, and FeB2, significantly suppressed the IGSCC. Surface oxide examination by Auger electron microscopy showed that boron was deposited into the oxide.

Posters

Among the papers presented as posters, one that caught the eye of this reporter was The jet impingement cell: a valuable device for investigating CO2 corrosion' by Brian Kinsella ef al, of the Curtin University of Technology, Perth, WA 6845, Australia. This covered aspects of cell design and the determination of mass transfer and wall shear stress at electrodes used in the cell. The performance of different generic type corrosion inhibitors and their effect on the type of corrosion (i.e. uniform, pitting or crevice corrosion) was shown and discussed. Another noteworthy poster was 'Electrochemical study of corrosion inhibition of stainless steel in phosphoric medium' by K. Hnini et al., Laboratoire d'Electrochimie et de Bio corrosion, Beni- Mellal, Maroc. This work reported the corrosion inhibiting behaviour of the organic compound Methoxy-2-Allyl-4 Phenol (MAP) containing co-ordinating and conjugated groups, in three forms (natural, polymerised and chemically treated) using potentiodynamic polarisation, electrochemical impedance spectroscopy and SEM. The results revealed that this compound is an efficient inhibitor in all forms, but the most inhibition efficiency is obtained with the polymerised form. A final poster was 'Zinc sodium hypophosphite as a corrosion inhibitor' by Bhaskar V. Vaidyaa et al. from The Maharaja Sayajirao University of Vadodara, Gujarat (India). This showed how zinc sodium hypophosphite (ZSHP) was found to be a stable and effective inhibitor for mild steel, copper, and aluminum plates using weight loss and potentiodynamic techniques. Assessment of chemical oxygen demand and the total biological count were both done in the presence of inhibitor and found to be within acceptable limits.

Workshop 2: local probe techniques for corrosion research

This session ran from early on the Tuesday through to Wednesday lunchtime. Twenty six papers were presented. In the following report, selected papers are divided into groups depending on the main techniques used, e.g. scanning vibrating electrode technique (SVET), scanning Kelvin probe (SKP), and others.

Scanning vibrating electrode technique

The session started off with a contribution from UK, 'Application of scanning vibrating electrode technique (SVET) to the study of weld corrosion'. This work had been done by R. Akid et al. from Sheffield Hallam University. Welding induces major differences in both chemistry and metallurgy of the welded joint and SVET offers the opportunity to study these. The paper discussed both the influential parameters governing the SVET output signal, in terms of magnitude and resolution, and the application of the technique to a number of different weld configurations. Conventional dc electrochemical techniques and optical/SEM observation of the weld corrosion were also presented for comparison purposes.

The poster display

'Microstructural changes in zinc aluminium alloy galvanising as a function of processing parameters and their influence on corrosion' came from Jonathan Elvins et al. from the University of Wales, Swansea. The surface and cut edge corrosion performance of Galvan (Zn-4.5 wt-%AI) subjected to differing cooling rates had been investigated using SVET. For fast-cooled samples there had been an increase in corrosion activity on the surface but a decrease at the cut edge. In more slowly cooled samples higher intensity, long- lived anodes along the cut edge were seen and this was directly related to the increase in zinc dendrite size within the coating.

The next paper came from France and had looked at a quite different application in work by R. Bassguy et al. from the Laboratoire de Gnie Chimique in Toulouse, and was entitled 'Local analysis by SVET of the involvement of biological systems in aerobic biocorrosion'. In non-sterile natural environments, microorganisms colonise the surface of materials and grow to form a biofilm which can induce severe biocorrosion phenomena. One hypothesis to explain this is the presence of enzymes (as peroxidase, catalase, etc.) that catalyse oxygen reduction. Using different geometric arrangements the SVET confirmed that this mechanism predominates when pitting occurs in natural corrosion. These first results open new routes in the understanding of biocorrosion and help to define strategies to fight against it.

A paper from Holland by Laetitia Philippe et al. from the Netherlands Institute for Metals Research (NIMR), Delft was 'Use of SVET to study local effects in coatings on metals'. The first application was an investigation of the influence of aluminium alloy composition on anodised films. During the anodising process, the anodic current densities of the aluminium matrix and the second- phase material were clearly shown to be different, reflecting local anodic processes occurring at different rates. The second application looked at a scratch on an Intrinsically Conducting Polymer Film (ICPF) applied on a steel strip. The SVET was used to monitor the evolution in time of the damaged ICPF specimen in a corrosive environment compared with a silane coated steel sample. Monitoring the electrochemical behaviour of both polymer films enabled conductive/insulating properties of the coatings to be obtained. Another related paper was Observation of self-healing functions on the cut edge of galvanised steel using SVET and pH microscopy' by Kevin Ogle et al. from Arcelor R&D, Maizires ls Metz in France. The current distribution was measured on steel with a 25 m thick galvanised coating in sodium chloride. The distribution of pH was measured using liquid membrane glass capillary electrodes. A clear correlation was observed between the spatial pattern of precipitated corrosion products and the anodic/cathodic c\urrent and pH distribution. Zinc dissolution occurred at very localised points indicating that a pitting mechanism predominated. In a NaCI/(NH^sub 4^)^sub 2^S0^sub 4^ electrolyte the zinc surface was entirely active, only minimal corrosion product precipitation was visible, and no significant increase in pH was observed in the vicinity of the cathodic reactions. These results demonstrate how different mechanisms may predominate under different accelerated tests conditions.

Another French paper with Polish (University of Krakow) connections was 'In situ detection of the galvanic coupling during pitting: a combination of scanning vibrating electrode technique (SVET) and microcapillary electrochemical cell (MEC)' by Roland Oltra from Universit de Bourgogne, Dijon. Local electrochemical measurements had been made around localised corrosion sites in free corrosion conditions and these indicated actual galvanic coupling between the corroding sites and the passive surrounding surface. Complementary SVET measurements were able to image this in the case of stable propagation of a pitting site in stainless steel.

Finally in this group there was 'Role of Fe(ll)-containing compounds during pitting processes of carbon steels in NaHCO^sub 3^/ NaCI electrolytes' by M. Reffass, from Universit de La Rochelle, La Rochelle, France. Carbon steels are characterised by passive behaviour in various alkaline media, for instance 0.1 M NaHCO^sub 3^ aqueous solutions, but pitting can occur in the presence of cations. The dissolution of iron in the pits induces rapidly the precipitation of Fe(ll)-containing phases, for instance iron carbonate, iron hydroxide or Fe(II)-Fe(III) hydroxysalts known as 'green rusts'. The mechanisms leading to those various transient compounds were studied by coupling classical electrochemical methods (cyclic voltammetry and chronoamperometry) with electrochemical microprobes (SVET) and analysed by micro-Raman spectroscopy. The activity of a given pit, followed by SVET, reached a maximum before it decreased, in agreement with the evolution of the overall current density that was observed to stabilise, whereas the number of pits increased.

Scanning Kelvin probe

The first scanning Kelvin probe paper was from Norway entitled 'SKPFM applied on different conversion coated Al surfaces' by B. S. Tanem et al, working at SINTEF in Trondheim, Norway. Removal or passivation of cathodic intermetallic particles through pre- treatment should decrease the overall corrosion rate and pitting susceptibility of commercial aluminium alloys. To investigate this, SKPFM has been used to map the distribution of the Volta potential at and in the vicinity of cathodic particles on smooth and clean ultramicrotomed EN AW-6082 aluminium surfaces before and after the application of different conversion coatings. Application of a chromate free Ti-Zr based pretreatment resulted in a small reduction in the Volta potential difference between cathodic α- Al(Fe1Mn)Si particles and the aluminium matrix. Conversely, chromate conversion coating (CCC) reduced the Volta potential difference between the α-Al(Fe,Mn)Si phase and matrix to a minimum.

Back to UK for 'Crime and corrosion: fingerprint detection using a scanning Kelvin probe' from Geraint Williams et al. of the University of Wales, Swansea. This paper demonstrated on a range of metallic samples such as iron, copper, nickel, titanium and aluminium that fingerprint ridge deposits induce beat Volta potential variations and these can be spatially mapped using the SKP. Furthermore, these patterns can be mapped even when the sample has been coated with a polymer layer or heated (e.g. as might occur with a cartridge case). The explanation given was that electrolytes present in the sweat lead to electrochemical depassivation of the metal surface and this was confirmed by secondary ion mass spectrometry (SIMS).

Japan provided the next paper with H. Masuda from NIMS in lbaraki describing the 'Atmospheric corrosion of low alloy steels near seashore environment under high relative humidity condition'. Tests had been carried out on SM50B and 3%Ni steel either with constant deposition rate of sea salt particles or after a certain amount of sea salt particles had been attached on low alloy steels. The initiation of atmospheric corrosion was observed both by the super Kelvin force microscope (SKFM) and the colour laser microscope. An SKFM image was taken every hour to observe the detail of corrosion. The rate of corrosion of SM50B was faster than that of 3%Ni steel at the beginning, but the total amount of corrosion became the same with time. The in situ SKFM observation showed that in more negative potential areas of the part, attached sea salt particles spread rapidly once corrosion started. The increase of the area of the negative potential part was explained by the spread of a thin liquid film of FeCl^sub 3^.

It was back to France for 'Application of scanning Kelvin probe, XANES and FTIR microspectroscopy for investigating the mechanisms of filiform corrosion on coated aluminium alloys' by Nathalie LeBozec et al. from Institut de la Corrosion, Brest, France. This work also had the involvement of the Swedish Corrosion Institute (Dominic Thierry). SKP (spatial resolution about 50 m) was used to identify the anodic and cathodic sites on filaments formed on coated aluminium during atmospheric exposure. FTIR microspectroscopy allowed identification of the composition of corrosion products formed in the head and in the tail of the filaments. XANES measurements were performed using a high-resolution beam (e.g. 15- 35 m), on coated chromated aluminium alloys in order to study the mechanisms of inhibition of chromate on the filiform corrosion. X- ray maps of Cr(VI) and total chromium over the surface including the filament were obtained.

Holland and Germany contributed to the next paper delivered by F. Andreatta from Netherlands Institute for Metals Research (NIMR) in Delft on 'Use of the scanning Kelvin probe force microscopy (SKPFM) and microcapillary cell MIC) to investigate the local corrosion behaviour of 7xxx aluminium alloys'. AA7075, AA7349 and an experimental alloy with composition in the range of AA7449 were investigated. It was found that Cu- and Fe-rich intermetallics are cathodic relative to the matrix and are characterised by a high Volta potential difference with the matrix (strong galvanic coupling) the exact value being dependent on the composition both of the intermetallics and of the matrix. Moreover, it is affected by heat treatments performed on the alloy. As an example, solution heat- treated AA7075 showed a strong increase of the Volta potential difference between the intermetallics and the matrix compared to AA7075-T6.

Finally in this group there was a spectacular paper (excellent images) from Douglas Hansen with Hugh Isaacs, one of the founding fathers of these local electrochemical techniques, as a co-author. This was 'Detection of corrosion processes beneath applied paint coatings utilising the scanning Kelvin probe (SKP) technique' in work done at Princeton Applied Research, Oak Ridge TN and Brookhaven National Laboratory, Upton, NY, USA. SKP had been used to map out the work function of aluminium aircraft alloys modified with various surface treatments and different applied thicknesses of polymer coatings under such conditions. This appeared to pick up corroded areas (i.e. it still exhibited differences in Volta potential from that of a non-corroded sample) even when the sample had previously undergone corrosion but was now temporarily quiescent and had had multiple coatings of paint applied over the corroded area. The results indicated that the scanning Kelvin probe technique is capable of discriminating between areas of varying corrosion in a non-contact and non-destructive manner.

Scanning tunnelling microscopy/atom force microscopy and FTIR

A paper on 'Initial stages of the localised corrosion by pitting of passivated nickel surfaces studied by STM and AFM' was given by V. Maurice et al. from the Ecote Nationale Suprieure de Chimie de Paris. Corrosion tests were performed in 0-05M H^sub 2^SO^sub 4^ + 0.095M NaOH (pH 2-9) with or without NaCI (0-05M). In situ electrochemical STM was used to study the surface dynamics associated with the dissolution in the passive state. Dissolution was observed to be located at the step edges of the facetted hydroxylated NiO(111) surface of the passive film, leading to a 2D retraction of the terraces of the oxide (i.e. step flow). Some interesting observations were also made on the effect of potential and chloride on passivity breakdown.using the AFM in oscillating mode. The combination of these results with previous STM results indicated that chlorides facilitate the dissolution at the grain boundaries of the oxide and sustain the dissolution at the metal surface located at the bottom of the dissolving boundary.

A similar paper also from France was 'In situ AFM study of pitting corrosion and corrosion under strain on a 304L stainless steel'. By combining three techniques: atomic force microscopy; an electrochemical cell, and a traction plate on a sample surface exposed to a corrosive solution, the relationships between local geometrical defects (grain boundaries, dislocation lines, etc.) or chemical defects (inclusions) with the preferential sites of corrosion on the strained material had been determined. In situ exploration of a 304L strained surface revealed the first stages of the surface plastic evolutions like activation of sliding dislocations, shown by parallel steps of about 2 nm high in the same grain. secondary sliding plane systems were also noticeable for higher deformation rates.

It was over to Japan for the next paper on 'Etching processes of ZnO(0001) surface in solution studied by in situ STM' by Junji lnukai et al. from Tohoku University, Sendai. On the surface prepared by chemica\l etching in alkaline solutions atomically flat terrace step structures were consistently observed. Steps observed on the surface were composed of bi-layers with a step height of 0.26 nm. Atomic images obtained by in situ EC-STM revealed that the surface has a (1 x 1) structure with a hexagonal lattice. Etching was observed to proceed at step edges.

There were two posters associated with this workshop both describing use of infrared methods. 1An FTIR microscopic base imaging system: applications to corrosion layer characterisation' by J. Kasperek et al. from Universit de Rouen, Saint Etienne du Rouvray. To create natural patinas zinc and nickel samples were exposed in an urban site located at Rouen in France whereas an artificial patina was developed on copper substrate during a cyclic immersion test. Infrared microscopy and mapping techniques showed up the heterogeneous distribution of chemical species. Zinc surfaces were uniformly corroded but the distribution of corroded species was inhomogeneous with regard to carbonate and hydroxysulfate groups. The corrosion mechanism observed on nickel substrate is pitting and pit size, chemical mapping and 3D visualisation determined from SEM/ EDS and FTIR map were very close.

A second poster from the same university was 'Improvement of IR microspectroscopy for data analysis of corroded surface' by B. Lefez et al. The interpretations of reflection spectra are difficult for several reasons: first, there is a lack of reference data; secondly, the surfaces are often rough; and, thirdly, the thickness of the external layer can be the origin of physical effects like interference fringes. In order to study the reflection spectra, a theoretical calculation was undertaken with the infrared optical constants obtained by Kramers-Krnig analysis. In the case of copper corrosion, the absorption band was attributed and the two copper chlorides like atacamite and clinoatacamite were differentiated.

Other local techniques (LEIS, microelectrochemical cells, and Raman spectroscopy)

The first paper in this group was 'Study of the delamination of organic coatings by local electrochemical impedance spectroscopy' by Jean-Baptiste Jorcin et al. from the Centre Inter Universitaire de Recherche et d'Ingnierie des Matriaux in Toulouse. Delamination of epoxy-vinyl primers from steel occurred from artificial defects after ageing in a salt spray chamber for up to 50 days. LEIS was carried out with a Solartron device using a five electrode configuration. Mappings were performed at 5 kHz in a low conductivity medium (0.001 M NaCI) to optimise resolution. The initiation and the propagation of the delamination of the different systems had been clearly observed. The delaminated surface areas measured by visual observation after the removal of the coating were in good agreement with those determined by local electrochemical impedance mapping.

Back to UK for 'Modelling the local electrochemical impedance response from surface heterogeneities in organic coatings' by S. B. Lyon et al. from University of Manchester. To help deal with problems in interpretation of local EIS data where the coating is partially defective, a finite element model of a defective organic coating on a conducting metallic ground plan had been developed. This model was used to predict macroscopic and local impedance data and, hence, to study the degree of convolution of response as a function of defect size and probe location.

It was to France again for 'Contribution of local probes in the understanding of mechanical effects on localised corrosion' by Vincent Vignal et al. from Universit de Bourgogne, Dijon. The microcapillary electrochemical cell technique was used to assist understanding of the effects of mechanical stresses on the processes leading to pitting corrosion. Typical embrittlement can be observed after straining around MnS inclusions on resulphurised 316 stainless steels. Numerical simulation of the location of stress gradients before the local electrochemical analysis was shown to be possible and could be a very useful way to define the pitting susceptibility of microcracked areas during straining.

UK contributed again in the person of B. J. Connolly from the University of Birmingham. He described 'Localised corrosion measurements on aluminium alloys via a microelectrochemical cell. The work characterised the electrochemical reactivity as well as the resultant localised corrosion site morphology of friction stir welded (FSW) 2024 T351 aluminium alloys as a function of remote applied stress state in chloride-containing environments. It utilised the 'droplet cell' method which employs fine pipettes, resulting in a maximum lateral resolution on the order of 5-10 m. Specific FSW regions and therefore specific microstructures that lead to enhanced localised corrosion initiation and propagation were identified and the role of stress in modifying localised corrosion behaviour was examined.

A new way to study transition metal corrosion by in-situ micro- Raman spectroscopy was described by S. Joiret et al, from Universit Pierre et Marie Curie, Paris. A way to increase the Raman signal is to use surface enhanced Raman spectroscopy (SERS). A new route has been tested to do this, which consists in electrodepositing iron or nickel on top of a well defined Au layer presenting the capability of SERS activity. Passivity of iron had been studied in borate buffer solution, and pitting corrosion of the passive layer in the presence of chloride or sulphate ions had been analysed. Nickel pitting corrosion had also been tested.

Workshop 3: corrosion by hot gases and combustion products (with emphasis on high temperature corrosion in the chemical and petrochemical industry)

This was another large workshop, with two dozen or more oral presentations and a poster session featuring seven more contributions. It had been organised jointly by Michael Schtze of the Karl-Winnaker-lnstitute of DECHEMA eV in Frankfurt, the Chairman of EFC Working Party 3 (corrosion by hot gases and combustion products) and Franois Ropital, Chairman of EFC Working Party 15 (corrosion in the refinery industry).

It began with a notable keynote lecture by Jrgen Korkhaus, Head of the Materials Engineering Department at BASF AG in Ludwigshafen, Germany. Entitled 'Failure mechanisms and material degradation at high temperature in ammonia synthesis', this traced problems encountered with the Haber process since its inception nearly one hundred years ago to the present day. The earliest problems, with the pilot plant, were due to hydrogen attack of the carbon-steel reactors operating at temperatures approaching 500 C in gases containing hydrogen, nitrogen and ammonia. They were overcome by using a carbon free iron liner, resistant to hydrogen attack, inside the pressure bearing shell, into which holes were drilled to allow hydrogen diffusing through the liner to escape without attacking the carbon steel shell. Since then, process temperatures have increased progressively in the pursuit of higher operating efficiency. Over the past 50 years, the specific energy efficiency of ammonia plants has increased fourfold. Some sections of modern steam reforming ammonia plants now operate at temperatures of about 800C and the failure of primary reformer tubes is usually due to creep. BASF's most modern plant, in Antwerp has air pre-heaters made from alloy 800H, the outlet headers of which operate at ~600 C and developed cracks after only 2 years in service. These were a consequence of relaxation cracking caused by high residual stresses from fabrication and grain boundaries that were relatively weak compared with the grain interiors in which strengthening had occurred due to the precipitation of fine carbides during service. The solution here was to heat treat the alloy 800 H at about 950C to precipitate coarse carbides, thereby reducing the strength differential between the grain boundaries and the grains. Alternatively, use can be made of a low-carbon nitrogen alloyed austenitic stainless steel. Another increasing problem, which has affected waste heat recovery units behind the secondary reformer after as little as 9 months in service is metal dusting. The affected parts included the waste heat boiler and steam superheater. Process streams containing CO, CO2, H^sub 2^ and N^sub 2^ with carbon activities greater than 1 were in contact with alloy 800 H at design temperatures of 650 C. This attack is the result of severe carburisation, leading to pit-like damage due to metal disintegration and loss into the gas stream. Other alloys seem to offer better performance, including alloys 600, 690 and 671, but there are still problems in predicting when metal dusting is likely to occur and the need for further research was highlighted in the concluding remarks of the presentation.

Coke deposition, carburisation and metal dusting

The first batch of papers in the session that followed were concerned with coke deposition, carburisation and metal dusting. The first, by Franois Ropital of Institut Franais du Ptrole, highlighted the benefits to be derived in mechanistic studies of coking by the simultaneous use of two measurement techniques. His paper was entitled: 'Coupling Thermogravimetric and Acoustic Emission Measurements: Its Application to Study the Inhibition of Catalytic Coke Deposition'. The catalytic deposition of coke is a major problem on the reactor walls in refineries because it disrupts heat transfer and gives rise to pressure drops. Inhibitors can be injected into the atmosphere to curb the deposition process, but the correct quantity to be used must be established empirically. To assist study of the inhibition by sulphur additives of the catalytic decomposition of coke on an iron surface at about 650C in atmospheres containing various amounts of hydrocarbons and hydrogen, a thermogravimetric balance was equipped with an acoust\ic emission monitoring device. It was demonstrated that acoustic emission measurements can reliably indicate the effect of an addition of 3 ppm H^sub 2^S on the catalytic deposition of coke in a hydrogen- isobutane process stream. An initial inhibition efficiency of 91% was achieved by the H^sub 2^S addition, but this fell to 76% during subsequent exposure. On the basis of the good correlation between the measurements of mass change and acoustic emission, it was concluded that acoustic emission is a promising tool for use on site.

In a paper entitled 'Coke deposition and carburisation on alloys used in the petrochemical industry', Rolf Kichmmeiner of Schmidt and Clemens, Lindlar, Germany, discussed a new alloy that is under development, which provides increased resistance to coking and carburisation in petrochemical environments. He explained that the centrifugally cast high-nickel alloys now used rely on the formation of a dense, adherent chromia layer to protect them from carbon penetration and subsequent degradation at high temperatures. The new alloy relies instead on the formation of a dense, adherent alumina layer. The extreme high temperature stability of the alumina layer will allow maximum operating temperatures to be extended beyond 1150C.

Metal dusting was next main theme, beginning with a presentation by Mrs Jorun Albertsen of the Materials Technology Department of NTNU in Trondheim, Norway. This concerned 'Metallurgical investigations of metal dusting corrosion in plant exposed nickel based alloy 602CA'. Nicrofer 602CA is a nickel based superalloy (63Ni-25Cr-9-5Fe-2AI-0.1 Si-0.1Ti-0-1Mn-0.1Zr-0.2C, wt-%). A test piece ground to a 60-grit finish was exposed for 2 years to a syngas environment in the reforming section of a methanol plant at a temperature of 540 and pressure of 35 bar. The process gas contained 14-9%CO, 47-3%H^sub 2^, 5.7%CO2, 31-2%H2O and 0.90%CH^sub 4^ (vol.- %), with possible traces of Na and SiO. Following exposure, large M^sub 23^C^sub 6^ carbides were present within the grains and grain boundaries to a depth of about 20 m below the corroded surface, together with a limited number of sub-surface corrosion pits that were attributed to metal dusting. The high intrinsic resistance to metal dusting of this alloy was explained by the formation of a stable surface oxide layer, presumed to be corundum (Al^sub 2- x^Cr^sub x^O^sub 3^), which prevents carbon penetration deep into the bulk of the material. This observation complements other recent evidence that metal dusting in nickel based alloys occurs because of coupled carburisation and oxidation reactions.

The next paper also concerned metal dusting, this time of Fe-AI alloys at 600C in a CO-H^sub 2^-H2O gas mixture. Reinhard Bernst of the Max-Planck-lnstitut, Dsseldorf, Germany, discussed the metal dusting behaviour of binary iron-aluminium alloys with aluminium contents of between 5 at.-0Xo and 45 at.-% in gas mixtures containing 0-2%H2O and 15, 25 and 50 vol.%CO (a^sub c^ > 1). It was reported that with increasing aluminium content the rate of carburisation fell. This was attributed to the formation of a thin alumina scale on the high-Al alloys. Another interesting observation was that samples with ground surfaces carburised more rapidly than ones with polished surfaces. This was explained by the influence of deformation induced during sample preparation on the kinetics of the surface reactions.

The first day's proceedings were brought to a close with a paper by Helen Ackermann of the Oel-Wrme-lnstitut, Herzogenrath, Germany, entitled 'Metal dusting in low NO^sub x^ recirculation burners for fuel oil'. This highlighted a tendency towards metal dusting of flame tubes made from the Ni based Alloy 601 (Ni-23 wt-%Cr-14 wt- %Fe-14 wt-%AI) in low-NO^sub x^ burners when burning low sulphur fuel oils. Pits and holes characteristic of metal dusting developed at very high corrosion rates, of up to 1 mm in 500 h in areas of high hydrocarbon concentration, where tube temperatures of about 700C were measured. Ten low-sulphur fuel oils, with sulphur contents ranging from 5 to 72 mg kg^sup -1^ and one high-sulphur oil containing 1700 mg kg^sup -1^ were used to investigate the role of sulphur. The fuel with the lowest sulphur content caused no corrosion attack in the tests but otherwise the sulphur content did not affect metal dusting or the amount of sulphur adsorbed on the burner surface, which corresponded to a monolayer of sulphur atoms whether the sulphur content of the oil was 15 or 1700 mg kg^sup - 1^. On this basis it was concluded that the inhibiting effect of adsorbed sulphur on metal dusting should be equally effective for low sulphur fuel oil as for standard fuel oil. However, there was evidence that the tendency towards metal dusting increases as the polycyclic aromatic hydrocarbon content of the fuel oil increases.

On the following morning, Till Weber of DECHEMA eV, Frankfurt am Main, Germany, discussed 'Development of advanced metallic coatings resistant to corrosion in high temperature industrial atmospheres'. Based on the knowledge that ?-TiAI is highly resistant in reducing sulphidising atmospheres, a range of Ti-AI co-diffusion coatings were produced using a single step pack cementation process. These included various AI-Ti, AI-Si and AI-Ti-Si diffusion coatings that were applied to both austenitic steels and Ni base alloys. The coatings displayed excellent behaviour in reducing sulphidising atmospheres with high carbon contents at up to 700C and under metal dusting conditions at temperatures of 620 and 700. It is more difficult to use diffusion coatings with ferritic steels due to the effect of the high temperature (900C) treatments used during the pack cementation process on mechanical properties. Therefore, the high velocity oxy-fuel (HVOF) method was used to apply TiAI coatings on ferritic steels. This coating-substrate system displayed good mechanical stability under thermal cycling conditions because of the similarity of the thermal expansion coefficients of the coating and substrate.

Corrosion by highly aggressive environments

The next part of this session was devoted to corrosion by highly aggressive environments, and the first paper on this topic, entitled 'hot corrosion of coated single crystal superalloys' was presented by Nigel Simms of the Power Generation Technology Centre at Cranfield University, UK. Single crystal superalloys were originally developed for the hot gas path components of gas turbines using clean fuel but are now being used for industrial gas turbines that might burn dirtier fuels, possibly including gasified coal or biomass. Laboratory tests have, therefore, been conducted to establish the sensitivity of the single crystal materials to such exposure. The materials included CMSX-4 (Ni-6-5Cr-l OCo-5-6AI-4-9Ti- 6W-0-6Mo-6Ta-0-1C-2-9Re-(MHf) and SC^sup 2^-B (a proprietary ONERA alloy). They were tested both without coatings and with Pt-AI and Amdry 997 (Ni-8-5AI-20Cr-23Co-4Ta-0-6Y) coatings. The laboratory test conditions covered the range 50-500 vpm SOx and 0-500 ppm HCI in air with deposits of 4/1 mol.-% (Na/K)2SO4or 1/1 mol.-% (Na/ K)^sub 2^SO^sub 4^ at deposition fluxes of 1-5-15 g cm^sup -2^ hT1 at temperatures of 700C or 900 C for durations of 500 or 1000 h. The damage rates of the uncoated single crystal materials were too high for them to be used with confidence in gas turbines fuelled with gases derived from dirty fuels. Under the more severe test conditions the corrosion rates with Pt-AI coatings were also excessive, although the Amdry 997 coatings performed better.

The efficiency of electricity production by burning biomass is limited by corrosion of the superheater tubes caused by alkali chlorides in the deposits, explained Sung Chul Cha of the Max- Planck-lnstitut, Dsseldorf, Germany. His paper, entitled 'Local Reactions between NaCI and KCl particles and Metal Surfaces' described an investigation of the local reactions of NaCI particles with pure iron surfaces in comparison with the local reactions with KCI particles, which had been studied previously. Bare and pre- oxidised iron surfaces had particles of salt in the 0.05-1.5+ m size range deposited on them by thermophoresis prior to oxidation in a N^sub 2^-20 vol.%O2 mixture at 100C for 10 min. This oxidation treatment appeared to have no effect on the NaCl particles, but increasing the temperature to 300C and the time to 20 or 40 min caused the particles to spread out into a square shape on the metal surface, probably by melt formation in the presence of iron. This differed from the behaviour with KCl particles, which spread out in round form during oxidation at the higher temperature. The addition of 500 vppm HCl led to complete coverage of the iron surface with chloride.

The next paper, presented by Andreas Ruh, also from the Max- Planck-Institut in Dsseldorf, concerned the 'Influence of HCI and water vapour on the corrosion kinetics of Fe beneath molten ZnCl^sub 2^/KCI'. It addressed the acceleration of corrosion processes that occurs in waste-fired boilers, in which the presence of salts accelerates the corrosion processes, especially in the presence of corrosive species like HCI and H2O in the flue gases. The aim was to study the acceleration of heat exchanger tube corrosion by solid and molten salts formed by the condensation of aerosols within the flue gas. Thermogravimetric tests were conducted on pure Fe covered with a 50 mol.-%KCI/ 50 mol.-%ZnCl^sub 2^ salt mixture. These demonstrated enhanced mass gains for atmospheres containing HCI in comparison to similar tests without HCI. The most prominent difference between salt induced corrosion in atmospheres with and without HCI was the presence of an incubation phase in atmospheres without HCI. This was taken to be an indication of the time needed for the formation of chlorine, suggesting that the oxidation of ZnCI^sub 2^ to ZnO and \Cl^sub 2^ takes longer than the oxidation of HCI to H2O and Cl^sub 2^.

Sbastien Doublet of the Karl-Winnacker-lnstitut in Frankfurt was also concerned with the corrosive effect of chlorine, this time in waste incinerators where its presence may significantly reduce the lifetime of components. For example, PVC contains about 55% chlorine, which is released upon combustion, attacking structural materials at high temperature. In oxidising atmospheres, corrosion resistance is usually achieved by the formation of a protective surface oxide based on Al, Cr and Si. However, in the presence of chlorine the situation is much more complex. Quasi-stability diagrams have been established for each alloying element (i.e. Cr, Fe, Al, Mo, and Ni) where a partial pressure value (pMeCl^sub x^) of 10^sup -4^ bar is used as the critical vapour pressure above which significant evaporation of gaseous metal chlorides occurs. Sbastien Doublet's study had checked the validity of the quasi-stability diagrams in 'reducing' chloridising atmospheres with high chlorine (pCI^sub 2^ between 2 10^sup -2^ and 10^sup -4^ bar) and low oxygen contents (pO^sub 2^ between 10^sup -2^ and 10^sup -8^ bar). Tests had investigated nine different commercial alloys. The results showed that chromium is the least effective alloying element. Aluminium displayed better behaviour than chromium in chloride- containing atmospheres. Molybdenum offered the best results in the various chlorine-containing environments. Silicon also appeared to be a good alloying element, possibly due to its ability to form a dense SiO^sub 2^ layer. Differences between the thermodynamic calculations, which provide the basis for quasi-stability diagrams, and the experimental results were found for all the alloying elements. For example, chromium, which appears in the 'protective zone' for atmospheres containing 0.1 %Cl^sub 2^ appeared to be strongly attacked in all atmospheres. This discrepancy was attributed to an effect of the kinetics of metal chloride formation.

An interesting presentation by Ryutaro Fujisawa of Japan Chemical Innovation Institute in Sendai related to the 'Corrosion behaviour of Ni base alloys and type 316 stainless steel in supercritical water under alkaline conditions'. Supercritical water processes can be used in industrial applications, for example to decompose hazardous organic wastes. However, the corrosion of reactors and heat exchangers can be a major problem, which means that suitable structural materials must be selected for each particular process where supercritical water is used. So far, most studies of corrosion in supercritical water have used oxidising conditions. However, in some applications, such as the upgrading of low quality hydrocarbon resources, the environmental conditions are reducing. Therefore, a study to evaluate the corrosion rates of candidate materials for corrosion resistance in supercritical water environments under reducing conditions had been undertaken. Tests were conducted on five different materials, Type 316 austenitic stainless steel and four nickel based alloys (alloy 625, alloy C-276, Ni-45Cr-1 Mo and Ni-19Cr-19Mo). Corrosion tests were conducted in solutions of deionised water and NaOH at 400C and 25 MPa pressure for 50 h. In reducing supercritical water environments containing NaOH, the corrosion rate of Type 316 stainless steel was always greater than that of the nickel-based alloys, in which the presence of molybdenum was beneficial. Slow strain rate stress corrosion tests were also conducted in similar environments. Stress corrosion cracking did not occur in deionised water but the addition of 0-001 mol L^sup -1^ of HCI or 0-01 mol L^sup -1^ NaOH produced cracking of Type 316 stainless steel at 400C and 25 MPa pressure. The cracking susceptibility of the nickel-based alloys in supercritical water at 400C and 25 MPa pressure containing either HCI or NaOH decreased with increasing chromium content.

The last presentation in this section on highly aggressive environments was by Simon Rose of the University of Northumbria in Newcastle, UK. This discussed The performance of pack cementation coatings in air and 100% steam environments for supercritical steam turbines'. The use of metallic coatings can allow the operating temperature of the steam turbines used for power generation to be raised to 650C, with significant gains in thermal efficiency. In this study, tests were conducted on P92 HP turbine rotor steel both uncoated and with either a 26 m thick pack aluminised coating or a 12 m thick electroless nickel coating. Tests were conducted in laboratory air at 650C and in pure steam at the same temperature. The weight gains in the steam environment were up to seven times higher than those in air. In both environments, weight gains for uncoated material exceeded those for the coated materials by factors of about 10 times or more and the pack aluminised coating outperformed the electroless nickel one.

General high temperature corrosion issues

The third and final section of the workshop addressed general high temperature corrosion issues. It began with a contribution by Santu Datta of the University of Northumbria, who presented 'Nanoscale studies of the initial stages of degradation of diamond- like carbon (DLC) coatings'. These coatings are produced by chemical and vapour deposition processes and, structurally, they fall between graphite with sp^sup 2^ bonding and diamond with sp^sup 3^ bonding to give a good combination of hardness and toughness for tools and other wear-dominated applications. Unfortunately, at high temperatures DLC coatings degrade by transformation into graphite. This presentation reported nanoscale studies, utilising scanning tunnelling microscopy, scanning tunnelling spectroscopy and current imaging tunnelling spectroscopy to investigate the early stages of degradation.

The EFC bookstand

This was followed by a paper presented by Eric Chauveau of the Arcelor Group Research Centre in Ugine, France, about the high temperature corrosion of the austenitic stainless steel, low nickel austenitic stainless steel and nickel base alloys used for the conveyor belts in furnaces used for the heat treatment, brazing and sintering of metal parts. Cyclic carbonitriding tests were used to assess materials for conveyor belt applications in heat treatment furnaces. The atmosphere was 1 %NH^sub 3^-0-5%H2O-0-7%CH^sub 3^-4- 8%N^sub 2^-64%H^sub 2^ plus 29%CO at 920C. After 8 cycles of 20 h duration the results suggested that alloys 625, 811, 330Nb and 602CA are well suited for conveyor belts in carbonitriding atmospheres. For a sintering atmosphere (15%H^sub 2^-85%N^sub 2^O at 1130C), the best performance was displayed by alloys 625, 600 and 230. Finally, for continuous and cyclic oxidation in air at 980C, the best alloys were 625, 230, 602CA and HRl20. Clearly, the best grades overall for different conditions of chemical oxidation were 625 (63.5Ni-22.9Cr- 8.7Mo-3-6Nb + Ta) 602CA (62-7Ni-25-2Cr-2AI-0-16C-0.08Zr) and 230 (61.4Ni-21-8Cr-1-2Mo-0.016La-13-6W).

Martti Mkip of VTT Processes in Espoo, Finland, presented results from a study of The effect of CO2 on the corrosion rate in simulated combustion atmospheres'. This was relevant to both biomass and waste combustion processes. The results from laboratory, pilot and full scale testing of materials used for superheater tube materials in such applications were reported. Experiments were carried out on common ferritic and austenitic steels (X10, X20, 2-25Cr1Mo, AC66, Sanicro28, Esshette 1250, etc.). at 535 C in various simulated combustion atmospheres (22%H2O, 5%02, xCO^sub 2^ + N^sub 2^) where x varied from 0-25 vol.-%. It was found that the corrosion rates for the alloys increased with increasing CO2 content, especially for the ferritic steels.

Silicon has a high affinity for oxygen, and forms SiO^sub 2^ and/ or Fe^sub 2^SiO^sub 4^, usually as particles, filaments down the main grain boundaries, or a barrier layer at the base of the chromia scale. Rachel Pettersson of the Swedish Institute for Metals Research in Stockholm described a study to elucidate the effects of both silicon alloying on the oxidation behaviour of a Fe-20Cr-35Ni alloy in various types of oxidation experiments at 700 and 1000C. The alloys used had silicon contents ranging from 0-5 to 2-2%. It was established that the formation of a silicon oxide sub-layer beneath the main chromia scale reduces the chromia growth kinetics and suppresses breakaway oxidation related to iron oxide formation at 1000C. However, it also enhances scale spallation during cyclic oxidation tests. This dichotomy of beneficial and detrimental effects highlights the need to use testing methods that closely reflect service conditions for materials selection purposes.

A study of the surface chemistry of a Fe-15 at.-%Cr polycrystalline alloy after short-term heat treatments at 800C (the conventional heat treatment temperature for Fe-Cr alloys and stainless steels) was reported by Eungyeul Park of the Max-Planck- lnstitut, Dsseldorf, Germany. During heat treatment in air, the chemistry of the oxide layer on the surface transformed in the following sequence: Fe rich oxide [arrow right] Duplex oxide layer (Fe rich outer and Cr rich inner) [arrow right] Cr rich monolayer (M^sub 2^O^sub 3^). In the transformation, the reaction rate for the formation of the Cr rich oxide layer was controlled by the diffusion of cations. Heat treatments with annealing atmospheres (N^sub 2^ and N^sub 2^-5 vol.-%H^sub 2^) increased the Cr content at the surface, the inclusion of hydrogen increasing both the Cr content and the thickness of the surface and interface.

Valentin Rohr of the Karl-Winnacker-lnstitut in Frankfurt, Germany gave a talk about 'Diffusion coatings for the high temperature corrosion protection of 9-12%Cr steels'. These steels have been designed for applications at temperatures of up to 650\C are of interest for use as heat exchanger tubes in power stations. However, while their thermal conductivity and strength are suitable for the application, there is concern about their performance in corrosive environments at this temperature. Therefore, the possibility of protecting them with a suitable protective coating has been considered. Pack cementation was selected as an easy and inexpensive coating method, but could not be undertaken at above 650C without affecting the mechanical properties of the steel. For this reason, the coating process had to be carried out at 650C. Steels that had been aluminised by pack cementation at this temperature were tested in a simulated coal firing environment containing 14%CO-10%H^sub 2^O-1 %02-0-1 %SO^sub 2^-0-01%HCl, balance N^sub 2^ for 1000 h at 650C, and their performance was compared with that of samples with aluminide coatings deposited by fluidised bed chemical vapour deposition. Both coatings provided corrosion protection for up to 1000 h, although the aluminium content remained higher for the coating applied by pack cementation.

The penultimate presentation in the session was entitled Oxidation behaviour of Fe-AI alloys analysed using in and ex situ techniques'. It was given by Birgit Poter from the Max-Planck- Institut in Dusseldorf, Germany. She explained that the oxidation resistance of Fe-AI alloys containing more than 10%Al depends on the formation of a protective α-Al^sub 2^O^sub 3^ layer on the metal surface. However, at temperatures below 1000C metastable Al^sub 2^O^sub 3^ poiymorphs (γ, δ or θ), within which Al and O diffuse more quickly, grow preferentially, leading to higher oxidation rates. The initial stages of this phenomenon were studied using alloys containing 15-40 at.-%Al with defined oxygen fugacities at temperatures between 700 and 1000C. Ex situ experiments were conducted in an infrared furnace while in situ experiments were conducted within a field emission scanning electron microscope equipped with EDX and EBSD. Preliminary ex situ exposures of as cast materials with elongated grains and Al contents of 26 and 40 at-% to synthetic air at 800C for 3 min showed the formation kinetics of Al^sub 2^O^sub 3^ to be strongly dependent on the orientation of the grains. The orientation dependent diffusion of O, Al and Fe was then quantified by means of the in situ measurement techniques.

Finally, Professor Valery Belousov spoke about the 'Role of fused oxides in high temperature oxidation of metals'. This was concerned with the catastrophic oxidation of metals that sometimes occurs when their surfaces contact fused oxides. An investigation had been conducted of the accelerated oxidation of copper in contact with Bi^sub 2^O^sub 3^. It was found that super fast copper oxidation is caused by continuous dissolution of the Cu^sub 2^O protective oxide in the Bi^sub 2^O^sub 3^ melt and that oxidation occurs on a permanently bare metal surface due to a redox reaction.

Copyright Institute of Materials Mar 2005

Source: British Corrosion Journal

More News in this Category