Development of a model for the anodic behavior of T60 titanium in chlorinated and oxygenated aqueous media. Application to the specific conditions of hydrothermal oxidation (1 MPa

Development of a model for the anodic behavior of T60 titanium in chlorinated and oxygenated aqueous media. Application to the specific conditions of hydrothermal oxidation (1 MPa
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  De v elopment of a model for the anodic beha v ior of T60 titanium inchlorinated and oxygenated aqueous media. Application to thespecific conditions of hydrothermal oxidation (1 MPa B / pressure B / 30MPa, 20  8 C B / temperature B / 400  8 C) C. Frayret a , Th. Jaszay b , B. Lestienne a , M.H. Del v ille a, * a Institut de Chimie de la Matie`re Condense´e de Bordeaux (ICMCB)/CNRS-UPR 9048 87, A v enue du Dr Schweitzer, 33608 Pessac Cedex, France b Ecole Centrale de Nantes    /   Laboratoire de Me´canique des Fluides/CNRS-UMR 6598 1, Rue de la Noe¨ , BP 92101, 44321 Nantes Cedex 3, France Recei v ed 20 November 2002; recei v ed in re v ised form 30 January 2003; accepted 6 February 2003 Abstract This work e v aluates the anodic electrochemical beha v ior of titanium metal in hydrothermal oxidation conditions (up to 400  8 Cand 28 MPa) in chlorinated media in order to estimate the supercritical water oxidation reactors reliability for the treatment of lessthan 10% organic-waste waters. The titanium room temperature dissolution mechanism in chlorinated acidic medium (pH B / 0) isnot fundamentally modified by oxygen. Deduced from the ‘current    / potential’ and ‘ v alence    / potential’ cur v es, it is based on fourcrucial elementary steps leading to two branches: a so-called acti v e branch corresponding to a tri v alent dissolution (its effect isin v ersely proportional to the pH), and a passi v e branch (TiO 2  oxide formation with a  v ery limited tetra v alent dissolution). Inhydrothermal oxidation (pH  / 1), only the second branch is effecti v e. The titanium protection is directly related to the oxidestability in high pH systems. The mechanism model is expressed in terms of ‘current    / potential’ laws, which pro v ide kineticparameters using optimization calculations. The different elementary steps reaction rates were estimated as well as the e v olution of the reaction intermediates co v erage ratios with the potential. The quantification of each elementary step was performed tounderstand and/or orient the materials beha v ior according to different factors (pH, chloride ions contents, potentials . . . ). #  2003 Else v ier Science Ltd. All rights reserved. Keywords:  Titanium; Dissolution    / passi v ation model; Kinetic parameters; Anodic current    / potential cur v es; Supercritical conditions 1. Introduction In our pre v ious articles [1    / 3], we ha v e elaborated anexperimental set-up along with a protocol capable of measuring cathodic and anodic currents under steady-state and quasi steady-state conditions of metallicmaterials in supercritical aqueous media (pressure   / 23 MPa and temperature   / 373  8 C). These papersha v e shown that in oxygenated and chlorinated media(pH  / 1), titanium remains passi v e and is thereforepreferred to iron or nickel base alloys [4    / 7], for therealization of reactors using wet and supercritical wateroxidation processes of Cl  ions rich organic-waste [8    / 16]. Howe v er, these results, e v en if quantitati v e, remainglobal since they are obtained from the follow-up of therough current [3]. This current results from a reactionkinetics, in v ol v ing se v eral elementary steps. The deter-mination of the rate limiting step(s) is crucial for a betterunderstanding, a more fundamental analysis of theprocess and therefore a more tailored conditioning andbeha v ior of the material.The present work reports the results that we obtainedfor the reaction kinetics of the anodic dissolution of titanium in chlorinated oxygenated media under sub andsupercritical conditions. The goal is to  v erify if it ispossible to transpose to these drastic conditions, theresults pre v iously obtained in chlorinated oxygenatedmedia and in standard thermodynamic conditions [17    / 22]. * Corresponding author. Tel.:   / 33-556-8484-60; fax:   / 33-556-8427-61. E-mail address:  del v (M.H. Del v ille).Electrochimica Acta 48 (2003) 1685    / 1695www.else v$ - see front matter # 2003 Else v ier Science Ltd. All rights reserved.doi:10.1016/S0013-4686(03)00111-7  2. Investigation means Before specifying the present study conditions, webriefly recall the data already known on the anodicbeha v ior of titanium (T40 and T60 grades) [16    / 22] inoxygen-free concentrated hydrochloric acid media, un-der standard conditions (20  8 C B / T  B / 90  8 C). Thesteady-state current    / potential plots were carried out,polarizing a rotating-disk electrode, at pH B / 0. Theypresent a beha v ior characteristic of those obser v ed forpassi v able metal types with, starting from the corrosionpotential, an acti v ation peak followed by a passi v ityplateau. These profiles, along with the associated v alence law [19] and the presence of a Langmuiradsorption, can be simulated considering a reactionchain mechanism with a bifurcation and at least fourdeterminant steps. This chain is made of both an acti v eand a passi v e branch. The passi v e one is stronglydependent on the formation of the tetra v alent oxideTiO 2  [3,19,21,22]. According to the potential, the con-tribution of these branches e v ol v es: in the rising regionof the peak current, the acti v e dissolution to the  v alenceIII is predominant, in the passi v ity plateau region thepassi v e dissolution of titanium into  v alence IV go v ernsthe process. The kinetic parameters are e v aluated usingan optimization calculation based on the Nelder simplexmethod [16    / 19,21,23,24]. The adjustment between theexperimental and calculated profiles is a serious elementfor the model  v alidation. This is also confirmed by thedetermination of the acti v ation energies of the differentsteps, which ha v e  v alues corresponding to those emittedin the hypotheses for chemical and electrochemicalreactions. A four determinant-step model ga v e thereaction orders for two acti v e species in the medium(H  and Cl  ions) showing that the pH has a strongeffect whereas Cl  ions ha v e a minor influence on thereactions.As mentioned in Section 1, the aim of this work is to v erify whether this model is maintained in oxygenatedconcentrated hydrochloric acid media at low tempera-ture and if it is then applicable in higher pH media andunder hydrothermal oxidation conditions.  2.1. Experimental de v ice The experimental de v ice used for the electrochemicalacquisition in high pressure and high temperatureconditions was described in the pre v ious articles [1,2].The general scheme is presented on Fig. 1 with someinformation on technological choices and use modal-ities.The performed measurements are carried out in ausual 3-electrode potentiostatic configuration of the cell;howe v er, the use of chlorinated media (pH  / 1) and hightemperature and high pressure conditions (supercriticalconditions: 28 MPa and 400  8 C) implies specific adapta-tions. The essential aspects concern the chemical inert-ness of the cell and of its equipment (the cell material isTi 6 AL 4 V (also called TA6V)) as well as the electrodeswater tightness (obtained using a specific configurationdepending on the desired temperature and the epoxyglue nature) and electrical isolation (performed using aspecific PTFE sheathing) [1]. The choice of an internalreference electrode for potential measurements at hightemperature is particularly determinant. This internalreference electrode is a hydrogen dynamic electrode[19,25], which consists in a gal v anostatically controlledmicro electrolysis between two platinum-wires (with noincidence on the electrolytic medium). The electroche-mical cell is completely co v ered by a platinum sheathingfor  v ery low pH solutions (pH B / 1) [16]. For thechlorinated media studied under standard thermody-namic conditions, a con v entional glassware electrolyticcell and a double junction KCl saturated calomelelectrode (SCE) were used. In this specific situation,the working electrode can be used as a rotating-disk andthe anodic and cathodic compartments are separated bya glass-frit. In all cases, the auxiliary electrode is aplatinum-wire, and the working one is made of thestudied material, the useful part of this electrode being adisk with 1.5    / 5-mm diameter. Results are presentednormalized with respect to the SCE.  2.2. Experimental conditions The material is a non-alloyed titanium material,which was submitted, for 2 h, to a homogenizationannealing treatment (720  8 C) in an argon atmosphere. Aslow 2-h cooling process within a furnace followed thisfirst step. T60 impurity contents are gi v en in Table 1.The electrolytic media are prepared either withhydrochloric acid solutions, with potassium chlorideones or HCl    / KCl or HCl    / NaCl mixtures (Merckproducts of supra pure quality). These media may bedeaerated with argon. These solutions can be di v ided intwo types: those with pH B / 0 and those with pH E / 1.The first ones are used to understand the reactionmechanisms; the second ones are planed to simulatehydrothermal oxidation conditions.The pH B / 0 solutions are (i) 4.0 mol/l HCl solution,(ii) six HCl    / NaCl media at a constant pH of    / 0.2 andof   v ariable Cl  acti v ities (2.5 B / a Cl  B / 12.5) and (iii) fi v eNaCl    / HCl mixtures of constant acti v ity in Cl  ( a Cl  / of 12) [26]. The study of these different media showedthat the pH (respecti v ely chloride acti v ity) remainsconstant for a gi v en chloride acti v ity (respecti v ely pH).The compositions for pH E / 1 solutions are gi v en inTable 2 [2,3].The samples are sheathed according to the desiredtemperature conditions [1    / 3]. Their acti v e surface isthen polished using SiC abrasi v e paper up to grade 1200and rinsed in pure deionised water under sonication. C. Frayret et al. / Electrochimica Acta 48 (2003) 1685    /  1695 1686  After a delay corresponding to the time necessary toreach the open-circuit potential stability (corrosionpotential), titanium is polarized towards anodic poten-tials. The current    / potential cur v es are recorded startingfrom the corrosion potential towards these positi v e v alues using either a steady-state polarization (with 25mV increments) or a dynamic polarization at slowpotential sweep rate (0.17 mV/s), which approachespseudo-steady-state conditions in the context of thestudy. Different measurements ha v e shown that thedifference between these two polarization types is notsignificant especially under hydrothermal conditions [3].In addition, at low temperature, the use of rotating-diskelectrode as a working electrode shows that with norotation speed, the anodic dissolution of titanium isinsensiti v e to the presence of oxygen. The contributionof mass transport can, as well, be neglected in theabsence of oxygen [3]. Otherwise, the ohmic drop doesnot exceed a  v alue of 0.1 mV [2,3]. The major results aregi v en in the following section. 3. Results The most striking results are presented first consider-ing the  v ery acidic media (pH B / 0) and close to standardconditions (atmospheric pressure and temperatures  B / 100  8 C). A 4-mol/l concentration was chosen to in v esti-gate the influence of oxygen on the reaction mechanismsince it is an a v erage situation as compared to ourpre v ious studies done in a range of hydrochloric acidsolution of concentrations comprised between 1 and 8mol/l [17    / 22]. The quasi superposition of the ( I  ,  E  )steady-state polarization cur v es obtained on a fixed disk(Fig. 2), respecti v ely, in oxygen-free (Argon degassing)and non-deaerated solutions illustrates the minor influ-ence of this gas in the dissolution    / passi v ation mechan-ism of titanium once the corrosion potential isstabilized. This is, also, confirmed by the relati v elyconcomitant e v olution of the  v alence laws in bothdeaerated and oxygenated solutions (Fig. 2).The ( I     / V ) laws ( V , rotation speed of the disk) showthat as opposed to non-deaerated solutions, the titaniumdissolution in deaerated media is independent on anymass transport (constant steady-state current whate v erthe disk rotation speed). Howe v er, the extrapolation toa 0.0 rpm speed shows that on a fixed electrode, thedifference of currents between the two media is minor Fig. 1. Schematic design of the experimental set-up.Table 1Impurities of the T60 used in the experimentsElements Fe W, Mn C O N H Sippm 1800 1990 190 2000 195 60 60Table 2Characteristics of the media used at high pressure and high tempera-tureMedium pH measured at21  8 CComposition of the medium in H 2 O 2 and Cl  A 5.10 [KCl]  / 0.0292 mol/lB 4.77 H 2 O 2 /KCl: 3 wt.% H 2 O 2  and 0.0292mol/l Cl  C 1.71 [HCl]  / 0.0289 mol/lD 1.69 H 2 O 2 /HCl: 3 wt.% H 2 O 2  and 0.0289mol/l Cl  C. Frayret et al. / Electrochimica Acta 48 (2003) 1685    /  1695  1687  [3]. According to these data, the influence of thetemperature and of the  v arious constituents of themedia was studied in non-deaerated solutions in orderto illustrate and/or precise the model of the proposedmechanisms (Figs. 3    / 5).These cur v es show that the anodic dissolution currentof T60 in oxygenated acidic media (pH B / 0) is char-acteristic of a passi v able metal with an acti v ity peak anda passi v ity plateau [3]. The main contributions are dueto temperature and pH. The profiles preser v ation is anargument to consider that the dissolution mechanismremains the same. For media simulating hydrothermaloxidation (pH  / 1), the cur v e profiles are modified [3].As an example, results concerning three temperatureand pressure  v alues are presented in Figs. 6    / 8.They illustrate the fact that whate v er the hydrother-mal conditions at pH  / 1 and in highly oxidati v e media,it is not possible to induce the titanium acti v e dissolu-tion. All the ( I  ,  E  ) cur v es present a low amplitudeplateau ( 0 / 20  m A/cm 2 ), characteristic of the passi v enature of T60 material (absence of acti v ity peak). It ishowe v er, possible to obser v e that the supercritical fluid(400  8 C, 28 MPa) is less aggressi v e than the superior subcritical one, illustrating once more the modifications of the fluid properties at the critical point crossing. Themedium presents a strong tendency to become apolar(attenuation of its ionic character) [27] with the partialdisappearance of the H  ion incidence when reachingsupercritical conditions [3]. The passi v e nature of titanium is also illustrated by the measurement frommass loss, of the dissolution  v alence, which is experi-mentally equal to 4 (as an example see Fig. 9, mediumD, 350  8 C). Measurements were performed understeady-state polarization and the results presented hereare based on the a v erage of 3 attempts. The presence of  Fig. 3. Anodic steady-state current    / potential cur v es of T60 titanium( V  / 0 rpm) in non-deaerated HCl 4 mol/l (pH  /  / 0.64,  a H  /  / 4.4), atdifferent temperatures.Fig. 4. Anodic steady-state current    / potential cur v es of T60 titanium( V  / 0 rpm) in non-deaerated solutions with a chloride acti v ity  a Cl  /  / 12, at 21  8 C for different hydronium acti v ities  a H  : / Fig. 5. Anodic steady-state current potential cur v es for T60 titanium( V  / 0 rpm) in non-deaerated HCl    / KCl solutions at 21  8 C fordifferent chloride acti v ities  a Cl  : / Fig. 6. Anodic current    / potential cur v es of T60 titanium ( V  / 0 rpm)at 150  8 C and 10 MPa (inferior sub-critical domain), in the fourelectrolytic solutions A, B, C and D. Steady-state polarization.Fig. 2. Anodic steady-state current potential cur v es of T60 titanium( V  / 0 rpm) in HCl 4 mol/l at 21  8 C in the presence (  *  / ) and in theabsence ( I ) of oxygen and the corresponding  v alence laws. C. Frayret et al. / Electrochimica Acta 48 (2003) 1685    /  1695 1688  the tetra v alent form (TiO 2 ) was also underlined by thepresence of a purple-blue film on the electrode inagreement with other authors [28,29]. 4. Discussion and reaction model 4.1. Acidic media: basic model  We just showed (i) the absence of incidence of oxygenin terms of mass transport at a 0 rpm speed of arotating-disk, (ii) the superposition of the  v alence lawsin both aerated and oxygen-free media and of the ( I  ,  E  )cur v es profile. Considering these data, the electroche-mical beha v ior of titanium in deaerated acidic media canbe described by a reaction model with two electroche-mical steps leading to intermediate species which cane v ol v e by a bifurcation process in two different direc-tions as already mentioned for deaerated systems[16,21,30]. Indeed, both the slope in v ersion obser v edon the ( I  ,  E  ) cur v e peak and the concomitant e v olutionof the  v alence 3 to an intermediate one (between 3 and4), are major indications for a modification of thereactions chain, modification that is correlated to theappearance of   v alence 4. Consequently, the two-branchmodel de v eloped for oxygen-free media can as well beapplied in the case of aerated solutions. It will be used inthis paper with the aim of limiting the number of kineticparameters, in order to optimize the credibility of theirfurther numerical adjustment and get a rational exten-sion in the case of hydrothermal oxidation.The  v alues of dissolution  v alences confirm the solu-bilization of at least a tri v alent species [Ti(III) sol ] and atetra v alent one [Ti(IV) sol ]. The purpose is to determinethe elementary steps leading from the initial metallicstate Ti(0) to the final soluble species. These chainreactions were established [16,21,30] considering somesimplifying hypotheses: (i) the electrochemical steps aremono-electronic, (ii) all the elementary steps (bothchemical and electrochemical) are irre v ersible, e.g.,oriented towards the dissolution, (iii) the adsorption of the different species ( called adsorbates [Ti(X)] ads )follows the Langmuir isotherm beha v ior (Scheme 1)[31].Depending on their nature and their lifetime, in amajor part owing to the polarization conditions, theadsorbates more or less co v er the interface according toa co v erage ratio ( U ), which is expressed in surface units.This model supposes as well the chemical dissolution of Ti(III) ads  and Ti(IV) ads  as  v alence 3 and 4, respecti v ely(Fig. 2).Consequently, the simplest model able to explain thetitanium dissolution    / passi v ation presents six elemen-tary steps: four electrochemical ones and two chemicalones. If   K  i   is the rate of the elementary step  i   per surfaceunity, the reaction rate  v i   associated with the differentsteps of the model can be expressed as a function of theco v erage ratio  U i   according to: v 0  K  0 ( 1  U 1  U 2  U 3  U 4 ) (1) v 1  K  1 U 1  (2) v 2  K  2 U 2  (3) v 3  K  3 U 3  (4) v 4  K  4 U 3  (5) v 5  K  5 U 4  (6) For a steady-state, the co v erage ratios are fixed. Theformation rate of a product is then identical to itsdisappearance rate: Fig. 7. Anodic current    / potential cur v es of T60 titanium ( V  / 0 rpm)at 350  8 C and 28 MPa (superior sub-critical domain), in the fourelectrolytic solutions A, B, C and D. Polarization rate: 0.17 mV/s.Fig. 8. Anodic current    / potential cur v es of T60 titanium ( V  / 0 rpm)at 400  8 C and 28 MPa (super-critical domain), in the four electrolyticsolutions A, B, C and D. Polarization rate: 0.17 mV/s.Fig. 9. Dissolution  v alence ( k ) of T60 titanium ( V  / 0 rpm) as afunction of the potential at 350  8 C and 28 MPa in medium D. C. Frayret et al. / Electrochimica Acta 48 (2003) 1685    /  1695  1689
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks