Corrosion of Titanium and Ti-Alloys at High ...

5 downloads 0 Views 287KB Size Report
University of British Columbia. Department of Metals and Materials Engineering. 309 -6350 ... Superior crevice corrosion resistance vs. Ti-2, and ... (Titanium Metals Corporation). ..... Entropies and Heat Capacities above 200°”,J.Am.Chem.Soc.
Corrosion of Titanium and Ti-Alloys at High Temperatures and Pressures J.Vaughan, P. Reid, A. Alfantazi, D. Dreisinger and D. Tromans University of British Columbia Department of Metals and Materials Engineering 309 -6350 Stores Road Vancouver, B.C. Canada V6T 1Z4 M. Elboudjaini CANMET-MTL 568 Booth Street Ottawa, ON. Canada K1A 0G1 ABSTRACT In this study, pure Ti and select Ti alloys (ASTM Grade 1, 2, 7, 12 and 18), those typically employed as autoclave linings and internal components, were tested in sulphate media (30 gpL) at elevated temperatures (423-523 Kelvin) and pressures (700-4200 kPa), simulating conditions found during the high pressure acid leaching of nickel laterite ores. The DC electrochemical testing employed was cyclic polarization. Analysis of uncorroded select corroded samples was done using SEM, XPS and AES. Pitting of the samples was not apparent, the alloys adapted to the high temperature environment by thickening of the protective oxide layer.

INTRODUCTION The liner and many internal components of High Pressure Acid Leaching (HPAL) autoclaves used in the processing of nickel laterites, and other feedstock, are constructed from titanium and titanium alloys. These autoclaves are constructed using a steel shell for structural support with titanium internal liners (typically ASTM Grade 1) installed using explosion-bonding techniques [2]. The bulk process solution is oxidizing which leads to a strong protective film on the titanium metal. However, scales formed during batch processing often lead to localized corrosion problems in regions where the chemistry can deviate significantly from the bulk. The operating conditions involve acid concentrations of 30 to 100 gpL injected as pure H2SO4, temperatures ranging from 498 to 535 Kelvin and correspondingly high pressures up to 5200 kPa [1,2]. In this work the electrochemical tests were performed in a bench scale autoclave to simulate these conditions. XRD (x-ray diffraction), SEM (scanning electron microscope), XPS (x-ray photoelectron spectroscopy) and AES (auger electron spectroscopy) were used to characterize the corroded and uncorroded surfaces. ASTM Grade 1,2

7

12 18

Table I-Composition and Corrosion Characteristics of Ti-Alloys Alloy Corrosion Composition Highly corrosion resistant in oxidizing and mildly Unalloyed Ti reducing environments, including chlorides. Ti-1 used for autoclave liners. Most corrosion-resistant titanium alloy offering outstanding resistance to general and localized crevice Ti-0.15Pd corrosion in a wide range of oxidizing and reducing acid environments including chlorides. Superior crevice corrosion resistance vs. Ti-2, and excellent resistance under oxidizing to mildly reducing Ti-0.3-Mo-0.8Ni conditions, especially chlorides. Used for nozzles and valves. Ti-3Al-2.5VEnhanced corrosion resistance. Used for dynamic 0.05Pd components such as agitator blades. EXPERIMENTAL PROCEDURES

The titanium metals tested were ASTM grades (1,2,7,12,18) from TIMET (Titanium Metals Corporation). The samples were punched from cold rolled sheets that were reduced by about 50% and annealed, with the exception of Grade 18, which was not cold rolled. Each grade was characterized using XRD and all grades were confirmed to be alpha structured alloys (HCP). SEM micro photos confirmed grain sizes ranging from 4 to 20um, typical for industrial metals of similar history (see Figure 1).

Figure 1 - SEM Micrographs of the Four Alloys, Ti-2 (top left), Ti-7 (bottom left), Ti-12 (top right), Ti-18 (bottom right) Prior to testing, sample surfaces were ground to 600 grit on SiC paper, degreased with ethanol, and ultrasonically cleaned for 2 minutes. Grades 7, 12 and 18 were connected electrically by spot welding a platinum wire to the backside of the sample, while the unalloyed grades 1 and 2 were mounted in epoxy and connected by contact with a steel rod. The samples were air dried for at least one day before testing to ensure homogeneity between samples. Uncorroded samples for XPS and Auger characterization were prepared in the same manner as those for the polarization tests. The test cell was a laboratory scale autoclave from PARR with a glass liner (Figure 1). Potential measurements were compared to the ERE (External Reference Electrode), a pressurized Ag/AgCl system mounted above the autoclave. This electrode operates at system pressure and ambient temperature, which is maintained by an external copper-cooling coil with coolant being passed at 283 Kelvin.

8 9 10

6

7

11 12 4 2

5 3 1

1 Furnace 2 Autoclave body (Hastelloy-C) 3 Glass beaker 5 Working electrode 6 Gas outlet 7 Rupture pipe 8 Reference electrode 9 Cooling coil 10 Gas inlet 11 Pressure gauge 12 Counter electrode

Figure 2 - Schematic of the Autoclave Setup. The glass beaker contained 1.2 Liters (1.1 Liters for temperatures greater than 498) of 30 gpL H2SO4 solution. Working electrode attachments were covered with Teflon tape and a coiled platinum wire served for a counter electrode. The solution was de-aerated for 15 minutes by purging with N2 prior to increasing temperature. The open circuit potential was observed and recorded during warm-up. After the temperature and open circuit potential stabilized, the sample was cyclically polarized with a scan rate used of 5 mV/sec. Corroded samples were prepared by exposing the metal to 30 gpL H2SO4, heating them at 473 K in the autoclave for 3 hours then allowing them to cool overnight. Corroded sample contact with air was minimized by quickly rinsing the samples and storing them in isopropanol after opening the autoclave. RESULTS AND DISCUSSION The potential values obtained during the high temperature electrochemical tests were corrected for liquid junction potentials (LJP or ФI - ФII) and thermal junction potential (TJP). The LJP consisted mainly of the interaction between the 0.3 M sulfuric acid test solution and 1 M KCl salt bridge, a very small value was also included for the salt bridge and the Ag/AgCl reference electrode. Table II lists the TJP and LJP for the temperatures tested. The concentration of ions in the solution (ci) considered the temperature effect on the second dissociation constant of sulphuric acid HSO4-  SO4-- + H+ [3]. The values of ion mobility (ui) were replaced by ion diffusivities at infinite dilution [8] with a high temperature approximation using the T-dependent activation energy for self-diffusion of water [9]. The LJP between solution-I and solution-II were calculated using the Henderson equation (Equation 1) as follows [8]. R=8.314 J/mol*Kelvin, T is the temperature in Kelvin, F is 96500 Coulombs/mol. ФI - ФII =-(RT/F)*A*(ln(BI/BII)/( BI-BII))

(1)

Where, summations with respect to i and zi being the ionic charge, A = Σziui (ciI-ciII), BI = Σzi2uiciI, BII = Σzi2uiciII

(2,3 and 4)

The thermal junction potential occurred across the salt bridge, these values were approximated using known values for a 0.505 M KCl solution from D.D. Macdonald [10]. The reference electrode drifted slightly from test to test, the reference potential was measured compared to a SCE electrode before each test and considered in the conversion to NHE. Table II – Thermal and Liquid Junction Potential Corrections Temperature TJP [10] LJP (mV) Overall Correction (mV) (Kelvin) ∆ESHE(T) – ∆EOBS (mV) 423 87 -25.77 61.23 473 14 -28.56 -14.56 498 -24 -30.00 -54.00 523 -71 -31.48 -102.48 150

E (mV) vs NHE

100

Ti-18 (403-498 K)

50

0

-50

Ti-12 (405-498 K) -100

-150

-200 -5000

-4000

-3000

-2000

-1000

0

1000

2000

3000

4000

5000

Time (Seconds)

Figure 3 – Open Circuit Potential vs Time During the Autoclave Warm-up. Many factors affect the open circuit potential such as surface preparation, alloying elements, temperature and solution chemistry. The open circuit potential of the working electrode was measured before the polarization tests. Samples experienced a drop in potential during the warm-up at a temperature of about 410 K, see Figure 3. The temperature was stable at the test value for at least an hour before the cyclic polarization took place. All of the samples tested at 473 Kelvin or above settled to potentials between –0.05 and –0.30 V vs NHE. Figure 3 - Open circuit potential of Ti-Alloys during warmup and stabilization (ERE used), potential adjusted for LJP and TJP at 498 K.

Comment [JV1]: I would think the people reading this paper would already know this.

The thermodynamic stability of the oxide layer in water is also higher at higher temperatures as seen in the stability diagrams Figure 4. The entropy and heat capacity of the aqueous ionic substances and hydrated compounds have been calculated previously by Jantje Been [4] using Latimer’s method [5]. Thermodynamic data was also obtained from other sources [12,13]. High temperature thermodynamic data was extrapolated from these values using the Criss Cobble entropy correspondence principle [6,7]. TiO2++

2.0 1.5

E(V)

1.0

O2 Evolution

TiO++

0.5

TiO2.H2O HTiO3-

0.0 +++

Ti

-0.5

H2 Evolution

Ti(OH)3

-1.0

TiH2

-1.5 0

2

4

6

8

10

12

14

pH

TiO2

2.0

++

1.5

E(V)

1.0

TiO++

0.5

O2 Evolution

TiO2.H2O 0.0

HTiO3-

Ti+++

-0.5 H2 Evolution -1.0

TiH2

Ti(OH)3

-1.5 0

2

4

6

8

10

12

14

pH

Figure 4 – Stability Diagram for the Ti-H2O System at 298 K (top) and 498 K (bottom), Titanium Ion Activity is 1 x 10-6

1.5

Ti-7

Tem perature / Pressure (Kelvin / kPa)

1.0

473 / 1930 0.5

0.0

-0.5

-1.0 1.E -09

1 .5

1.E-08

1.E-07

1.E-06

1.E-05

T i-12

1.E -04

1.E-03

T e m p eratu re / P res s u re (K / k P a )

1 .0

E (V) vs NHE

498 / 2760 473 / 1930

0 .5

0 .0

-0 .5

-1 .0 1 .E -0 9

1 .E -0 8

1 .E -0 7

1 .E -0 6

1 .E -0 5

1.5

1 .E -0 4

1 .E -0 3

Temperature / Pressure (K / kPa)

Ti-18 1.0

498 / 2760 523 / 4240

0.5

0.0

-0.5

-1.0 1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

Current (A/cm^2)

Figure 5 – High Temperature Cyclic Polarization for Ti-Alloys, Ti-7, Ti-12, Ti-18 (Top to Bottom). Potentials corrected for LJP, TJP. The Scan Rate was 5 mV/sec.

The arrows on the Ti-7 curve indicate the scan direction beginning with a forward sweep (cathodic to anodic) and a reverse sweep (anodic to cathodic). High scan rates such as the one used for these experiments are known to exaggerate the passive currents as the potential is increases [11]. Ti-18 was exposed to higher temperatures and pressures than the other two alloys and both of its curves are nearly identical. A deviation from the passive current can be identified on all of the forward scans, the following table is a summary of where the potential begins to deviate. Table III – Apparent Trans-passive Potential Sample, Temperature / Pressure Apparent Trans-passive Potential (K / kPa) (V vs NHE) Ti-7, (473 / 1930) 0.5 Ti-12, (473 / 1930) 0.5 Ti-12, (498 / 2760) 0.2 Ti-18, (498 / 2760 and 523 / 4240) 0.2 The deviation from a passive potential on the forward scan is likely due to the oxidation of metal species that dissolved from the autoclave walls. A rough integration of the areas of deviation for the Ti-12 plots reveal that the charge is on the order of 600 µC at 473K and 1500 uC at 498 K. The concentration of iron in the solution after the tests was found to be on the order of 100 ppm while the nickel content was closer to 500 ppm. Other elements that were found on the corroded samples surfaces using XPS were W, Mo, Mg S, Ca, Na, Si and C. The autoclave body was made of Hastelloy-C that contains high amounts of Ni, Cr and Mo as well as some Fe, W, Co, Mn and Si . Possible oxidation reactions are MoS2 to MoS3, Cr++ to Cr+++ and Fe++ to Fe+++. The pH of sulphuric acid solutions increase with increasing temperature, the initial pH of the solution was about 0.55 at 298 Kelvin, this will correspond to a pH of 1 at 423 Kelvin and a pH of nearly 2 at 523 Kelvin [3]. Pitting was not identified on the polarized samples however the XPS analyses show that the oxide thickness of the uncorroded samples is on the order of 300-500 Å while samples exposed to high the temperature acid have a much greater oxide thickness between 5,000-10,000 Å. Surface roughness lead to the large error involved with these values which represent the depth at which the concentration of Ti is 67-75 atomic %. Auger Electron Spectroscopy depth profiles taken on well-defined small spots reduce the error associated with surface roughness values for the uncorroded the oxide thickness was 250-300 Å. The corroded oxide thickness for Ti-18 was found to be less than for Ti12, approximately 2200 and 4500 Å respectively. These values which represent the depth at which the concentration of Ti is 90 atomic %. The chemistry of the oxide film at the surface was composed of mainly Titanium (IV) associated with water Ti(OH)4 or TiO2.2H2O. For an example of the XPS spectra see Figure 6.

Figure 6 – XPS Spectra, Ti-12 Exposed to 30gpL H2SO4 at 473 K for 3 Hours.

CONCLUSIONS After electrochemical testing and analysis, it was found that these conditions did not produce localized corrosion of the samples. This is due to the increased stability of the oxide layer as predicted in the thermodynamic analysis and observed in terms of thickening using XPS and AES depth profiling. Other contributing factors are the presence of oxidizing inhibitor ions from the autoclave walls that were identified using XPS. The apparent trans-passive potential in the polarization results is attributed to the oxidation of these species. AES analysis reveals that the oxide film exposed to high temperature acid is significantly greater on Ti-12 than Ti-18.

ACKNOWLEDGEMENTS Many thanks to the staff at CANMET Materials Testing Laboratory, Ottawa, Ontario, for uninhibited assistance in providing expertise in high temperature corrosion testing, (Alex Doiron, Yves Lafreniere and Michael Attard from the Corrosion group) and surface analysis, (James Brown and Larry Galbraith and Jian Li from the Surface & Microstructural Analysis group). Thanks to Jim Grauman, with TIMET for continued support and guidance.

REFERENCES 1. J. S. Grauman and Tony Say, “TITANIUM for Hydrometallurgical Extraction Equipment,” Advanced Materials & Processes, March 2000. 25-29. 2. J.G. Banker, “Hydrometallurgical Applications of Titanium Clad Steel,” Report CLAD Metal Products Inc., 2000, Boulder, Colorado, U.S.A. 3. Staples, B. R. Holcomb, G. R. Cramer, S. D. “Calculation of pH for hightemperature sulfate solutions at high ionic strengths,” Corrosion (Houston, TX, United States) (1992), 48(1), 35-41. 4. J. Been “Titanium Corrosion in Alkaline Hydrogen Peroxide Environments” PhD Thesis, University of British Columbia, Vancouver, Canada. 1998. 5. Latimer, W.M. “The Oxidation States of Elements and their Potentials in Aqueous Solutions,” Prentice-Hall, Englewood Cliffs, N.J.:359-369, 1952. 6. Criss, C.M., Cobble, J.W. “ The Thermodynamic Properties of High Temperature Aqueous Solutions. IV. Entropies of the Ions up to 200° and the Correspondence Principle”,J.Am.Chem.Soc.,86:5385 (1964). 7. Criss, C.M., Cobble, J.W. “ The Thermodynamic Properties of High Temperature Aqueous Solutions. V. The calculation of Ionic heat Capacities up to 200° Entropies and Heat Capacities above 200°”,J.Am.Chem.Soc.,86:5385 (1964). 8. John Newman “Electrochemical Systems” Prentice-Hall, Inc. Englewood Cliffs, N.J. 1973. 9. Desmond Tromans “Modeling Oxygen Solubility in Water and Electrolyte Solutions” Ind. Eng. Chem. Res., Vol. 39, No. 3, 2000, 805-812. 10. D.D. Macdonald, Arthur C. Scott, Paul Wentreck “External Reference Electrodes for Use in High Temperature Aqueous Systems” J.Electrochem. Soc. June 1979, 908-911. 11. Kelly, Eugene J. “Electrochemical Behaviour of Titanium” Modern Aspects of Electrochem. 1982. 12. R.C. Weast, CRC Handbook of Chemistry and Physics, The Chemical Rubber Co. Cleveland, Ohio, 1973. 13. I. Barin, Thermochemical Data of Pure Substances, Weinheim, New York, 1995.