SOME ASPECTS OF REFRACTORY CERAMIC ELECTROLYTES FOR SULFUR DETECTION IN MOLTEN PIG IRON

SOME ASPECTS OF REFRACTORY
CERAMIC ELECTROLYTES FOR SULFUR

DETECTION IN MOLTEN PIG IRON

By

Mark A. Swetnam

Submitted in accordance with the requirements for the degree
of Doctor of Philosophy.

THE UNIVERSITY OF LEEDS
DEPT. OF MINING AND MINERAL
ENGINEERING

September 1999

The candidate confirms that the work submitted is his own and that appropriate
credit has been given where reference has been made to the work of others.

i

Abstract

This thesis describes the development and testing of an electrochemical sensor for
detection of sulfur in molten iron. The sensor is based around a ceramic electrolyte
comprising strontium ß-alumina, whose mobile species is Sr²⁺. As a result it is an
indirect sensor where the activity of the strontium is fixed by favourable reactions with
sulfur, present as a dissolved element in the molten iron and as a sulfide reference
electrode. The sensor EMF can be measured with a voltmeter and the sulfur content of
the iron determined using the well-known Nernst equation.

The electrolyte of choice, strontium ß-alumina, was studied in detail with the aid of
phase diagrams and XRD patterns. Unlike many other ß-aluminas, it was possible to
make dense strontium ß-alumina ceramic by conventional processing methods. Studies
showed that phase pure strontium ß-alumina begins to form at 1200^oC and exists over a
limited range of compositions. The presence of MgO as a dopant was essential to
stabilise the ß-alumina structure and prevent the formation of magnetoplumbite, a poor
ionic conducting phase.

The ionic conductivity of strontium ß-alumina was determined over a range of elevated
temperatures using ac impedance spectroscopy. Resistivity values and activation
energies for the polycrystalline were higher than published data regarding ionexchanged
crystals but still acceptable for sensor applications because of the high
temperatures involved. Work in low oxygen atmospheres suggests that strontium ß-
alumina may suffer from partial electronic conductivity in very reducing conditions.

The mechanical properties of strontium ß-alumina were studied in detail. Failure by
thermal shock was a major issue as the sensors were expected to survive direct
immersion into molten iron. Toughening mechanisms using zirconia additions were
implemented and the toughness / thermal shock resistance improved considerably.

The formation of electrolyte cups by ceramic injection moulding was studied in detail.
A suitable binder system for strontium ß-alumina was developed and the binder removal
/ sintering schedules were determined. A catalogue of defects was documented.

ii
A sulfur sensor based on strontium ß-alumina was designed and 160 probes tested in a
range of molten irons with varying sulfur content. The probes survived extremely well
and by optimising the design of the sensor thermal shock was virtually eliminated. The
data are tabulated and evaluated statistically. The sensors responded to changes in bath
sulfur and in general were in good agreement with combustion (LECO) analysis. The
probes were consistently less precise than LECO. However a direct comparison was
difficult because the sensor only responded to dissolved sulfur, not sulfur present in
inclusions.

The study concludes that it is possible to measure the sulfur content of molten pig iron
provided that the melt is saturated with carbon and also contains silicon; this ensures
that the oxygen activity of the melt is kept suppressed to prevent chemical interference.
Further work is required to prove the viability of such a sensor in a real-life
environment.

iii

Acknowledgements

The research described in this thesis was performed at Cookson Technology Centre,
Oxford, between October 1993 and July 1995, and at Vesuvius McDanel, Beaver Falls,
PA, USA, between August 1995 and October 1996. Supervision of my work was by Dr.
R. V. Kumar and Prof. D. J. Fray formerly of The Department of Mining and Mineral
Engineering, University of Leeds. The work described in this thesis is original, except
where reference has been made to the work of others.

I would like to thank Dr. Vasant Kumar and Professor Derek Fray for their continued
support and advice during this considerable period of my life. Other people who have
played a part in helping me during this time are:

My fellow workers at Cookson Technology Centre (1990 - 1996 R.I.P.). I would
particularly like to thank the ‘Sensors Team’ comprising Stef Witek, Alex Smeets,
Donald Kings, Andy Davies and Tim Jeffries. Much needed inspiration was supplied
by Dr. Lyn Holt to whom I am very grateful.

My colleagues at Vesuvius McDanel who not only had to put up with ‘The Brit’ but
assisted with the melt experiments in sweltering temperatures. Heartfelt thanks go to
Robin Sommers, Todd Kirkpatrick and Jason Stewart. Wisdom and guidance was
supplied by Dr. John Usher to whom I am very grateful. A special thanks to Dr. John
Dodsworth for moral support and encouragement, and for being the other ‘Brit’.

More recently I have received much support and help from my friends in the
Department of Materials Science and Metallurgy, Cambridge University. I would
particularly like to thank ‘The Bourgeois Cookery Club’ comprising Tony Cox, Jane
Freidina, Robert Copcutt and Francis Tailoka. Especially Tony who checked my
thermodynamic calculations.

A special thanks to my Mum and Dad, Elaine and Derek Swetnam, for support and
encouragement over the years. Especially Dad who has kindly proof read this thesis.
My Sister Ruth has also provided helpful advice for which I am grateful. My ‘other’
Mum and Dad Penny and Bert Higginbottom have been very supportive as well.

iv

Finally the biggest thank you of all goes to my family; my wife Pam and my wonderful
daughters Jenny and Catherine. Thank you Pam for putting up with me during all those
difficult times over the last seven years. I dedicate this thesis to you…..

Mark Swetnam
Department of Mining and Mineral Engineering
University of Leeds
September 1999

“An ounce of specificity is worth a pound of generality”
John Usher

vi

Nomenclature

l = Length
r = Radius
A = Area
t = Thickness
d = Diameter
π = PI (3.14159)
cm = Centimetre
m = Metre
g = Gramme
kg = Kilo gramme
tonne = 1000 kg
ln = Natural logarithm
Log = Base 10 logarithm (ln₁₀)
exp = Exponent
EMF = Electro-motive force (V)
R = Molar gas constant (8.3144 J.mol^-1.K^-1)
T = Absolute temperature (°K)
z = Number of charge units transferred
F = Faraday’s constant (9.6485 x 10⁴ C.mol^-1)

melt

P_S2

reference

P_S2
= Activity of dissolved sulfur in the melt.
= Partial pressure of sulfur in the reference electrode
G = Gibb’s free energy (J.mol^-1)
Δ = Prefix meaning ‘change in’
° = Suffix when used with free energy meaning ‘standard’
ΔG° = Standard free energy change (J.mol^-1)

h_s
f_s

j

Ε

e^j_s
r^j_s
= Henryan sulfur activity (1 wt. % alternate standard state)
= Interaction coefficient for elements present in iron
= dissolved element interacting with the sulfur
= Summation of all the relevant species j
= First order interaction coefficient
= Second order interaction coefficient

vii
ac = Alternating current
dc = Direct current
v(t) = Ac voltage
i(t) = Ac current
Ω = Ohm
f = Frequency (Hz)
φ = Phase difference between the voltage and the current
ω = 2πf (where f is the sinusoidal frequency in Hz)
Z = Impedance (ohms)
Z’ = Real (resistive) component of impedance
Z” = Imaginary (capacitive) component of impedance
ρ = Resistivity (Ω.cm)
σ = Conductivity (Siemens.cm^-1)

H_m

F
C
eV
E
λ
F
= Activation enthalpy for ionic migration
= Farad
= Capacitance value (F)
= Electron Volt
= Young’s Modulus (Gpa)
= Indented strength (Pa)
= Load (kN)
g = Gravitational acc
ⁿ. (m.s^-2)
m = Indent load (kg)

H_v

ν

Z_k
Z_u

= Vickers hardness (Gpa)
= Poisson’s Ratio
= Known Impedance
= Unknown Impedance

viii

Table of Contents

1. Introduction 1

1.1. Background to the Research 1
1.2. Sulfur in Iron and Steel 2

1.2.1. Current Desulfurisation Practice 3
1.2.2. Control of the Desulfurisation Process 4
1.2.3. Considerations and Uncertainties Arising During Desulfurisation 5
1.2.4. Desulfurisation Practice with a Direct Reading Probe 7
1.3. Sulfur Sensor Theory 9

1.3.1. The Electrolyte Material 9
1.3.1.1. ß-alumina 10
1.3.1.2. ß-alumina Selection 10

1.3.2. Principle of the Sulfur Sensor 11
1.3.2.1. Description 11
1.3.2.2. Basic Reactions 12

1.3.3. Other Criteria 14
1.4. Thermodynamic Calculations 15

1.4.1. The Activity of Sulfur in the Reference Electrode 15
1.4.2. The Activity of Sulfur in Pig Iron 17
1.4.3. The Overall Nernst Equation 20
1.4.4. Chemical Interference Effects 22
1.4.5. Chemical Phase Stability Diagrams 24
1.5. Conclusions 25

2. Literature Review 27

2.1. Introduction 27
2.2. Sodium ß-alumina 28

2.2.1. The Relative Proportion of ß and ß” Phases Present 33
2.2.2. Grain Size 33
2.2.3. Proximity to Theoretical Density 34
2.2.4. The Effects of Dopants. 35

2.2.4.1. To increase the number of Na⁺ ions 35
2.2.4.2. To stabilise the ß and ß” phases. 35
2.2.4.3. To act as a sintering aid 36
2.2.4.4. To act as toughening agents 36

2.2.5. Sodium ß-alumina Production 38
2.3. Strontium ß-alumina 38

2.3.1. Ion Exchanged Single Crystals 38
2.3.2. Polycrystalline Strontium ß-alumina 40
2.3.3. Direct Synthesis of Strontium ß-alumina 42
2.4. Solid Electrolyte Sensors 51

2.4.1. Zirconia Oxygen Sensors 51
2.4.2. Calcium / Magnesium Sulfide-Based Sensors 52
2.4.3. ß-alumina Sensors 54
2.4.4. Sensors in Steelmaking 57
2.4.5. Recent Developments 58

3. Sample Preparation 59

ix
3.1. Introduction 59
3.2. Sample Recipes 60
3.3. Sample Preparation 61
3.4. Particle Size Analysis 61
3.5. Thermo-gravimetric and Differential Thermal Analysis 63
3.6. Calcination and Binder Addition 65
3.7. Pressing and Sintering 66
3.8. Bulk Chemical Analysis 66

4. X-ray Diffraction Analysis 69

4.1. Introduction 69
4.2. Phase Analysis 69
4.3. Observations 73

4.3.1. Between 1000 and 1100^oC 73
4.3.2. At 1200^oC 74
4.3.3. At 1300^oC 76
4.3.4. At 1400 ^oC 77
4.3.5. At 1650 ^oC 79
4.4. Lattice Parameter Calculations 80

4.4.1. Method 80
4.4.2. Results 81
4.4.3. Discussion 81
4.5. X-ray Diffraction Conclusions 84

5. Scanning Electron Microscopy 89

5.1. Introduction 89
5.2. Discussion 89
5.3. Thermally-Etched Samples 97
5.4. SEM Conclusions 100

6. Ac Impedance Spectrometry 101

6.1. Introduction 101
6.2. Background Theory 101
6.3. Effect of Microstructure 104
6.4. Method of Analysis 104

6.4.1. Experimental Procedure 104
6.4.2. Analysis of Data 105
6.4.3. Results 106
6.4.4. Discussion 108
6.5. Ionic Conductivity: Grain Boundary Effects 112
6.6. Ionic Conductivity: Reduction Effects 115

6.6.1. Method of Analysis 117
6.6.2. Results and Conclusions 118

7. Density and Porosity 121

7.1. Introduction 121
7.2. Method of Analysis 122
7.3. Results 123
7.4. Discussion 123

8. Zirconia Toughening 129