oxidation potentials in iron and steel making
TRANSCRIPT
Oxidation Potentials in Iron and Steel Making
J. W. MATOUSEK1,2
1.—Englewood, CO, USA. 2.—e-mail: [email protected]
The state of oxidation of a pyrometallurgical process given by the partialpressure of oxygen and the temperature (the oxidation potential) is one of theimportant properties monitored and controlled in the smelting and refining ofiron and the nonferrous metals. Solid electrolyte sensors based on ZrO2 and areference electrode such as Cr/Cr2O3 to measure the oxygen pressure foundearly application in the steel industry, followed soon after in copper, nickel,lead, and zinc smelting. Similar devices are installed in automobile postcom-bustion/exhaust trains as part of emission control systems. The current dis-cussion reviews this technology as applied in the primary steps of iron andsteel making and refining.
INTRODUCTION
Previous papers in this series explored the con-cepts of oxidation potentials (given as the base-tenlogarithm of the equilibrium oxygen partial pres-sure in bars and the temperature) in studying andcontrolling pyrometallurgical extraction pro-cesses.1–3 Applications to copper, nickel, lead, andzinc smelting and metallurgical slags in generalwere examined. It is fitting to close these discus-sions with a return to where the technology be-gan—with the making of iron and steel.
TECHNOLOGY
Variations of Fig. 1 have been a central fixture inthe education of every extractive metallurgist sinceat least the 1951 publication of Basic Open HearthSteelmaking.4 The solid line represents theapproximate equilibrium at 1600�C between steelsof the carbon content given and atmospheres ofcarbon monoxide. The broken line indicates thedirection of the trend toward the conditions ob-served in production furnaces.
The measurement of carbon in steel has long beenan established practice. The direct measurement ofoxygen began to look practical with the 1960s pub-lications of Fisher and Ackermann,5,6 Fitterer7 andFruehan et al.,8 describing sensors based on solidelectrolytes such as stabilized zirconia (ZrO2) cou-pled with the reference electrodes Ni/NiO, Fe/FeO,Mo/MoO2, and Cr/Cr2O3; an example of the latter isgiven by Eq. 1. The difference in activities (concen-trations) between oxygen dissolved in iron and the
reference electrode generates an electromotive force(emf) that is related to the two activities by theNernst equation (Eq. 2). Since the 1970s, solid-electrolyte oxygen sensor use has been standardpractice in a variety of smelting and refining oper-ations; Fig. 2 illustrates the range over which theyhave found application.1
Cr;Cr2O3== ZrO2== O½ �Fe;Fe (1)
E ¼ �2:3RT=zF log p O2ð ÞFe= p O2ð Þref
� �(2)
where E is the cell emf in millivolts (mV), R is thegas constant, T is the absolute temperature, z is thecharge transfer (four in the case of O2), F is theFaraday constant (96,500 J/V mol), and log p(O2) isthe base-ten logarithm of the oxygen pressure inbars.
IRON AND STEELMAKING
Classical ironmaking begins with the blast fur-nace smelting of iron ore (magnetite, Fe3O4, orhematite, Fe2O3) with coke and fluxes. The product‘‘pig iron’’ contains approximately 4% carbon andvirtually no oxygen, as suggested by Fig. 1. The ironis made into steel with the removal of carbon (andother impurities) by oxidation in basic oxygen fur-naces (BOFs) or formerly in Bessemer converters(pig iron charge) and open hearth furnaces (pig ironand scrap charge). Modern electric arc furnaces(EAFs) start with scrap iron and steel, bypass theprimary smelting step, and proceed directly to
JOM, Vol. 65, No. 11, 2013
DOI: 10.1007/s11837-013-0713-1� 2013 TMS
1584 (Published online August 14, 2013)
melting followed by refining in ladle furnaces andcontinuous casting. These latter two operations arealso standard in BOF practice. It is in the refiningand casting that solid electrolyte oxygen sensorsplay an important role in quality control.
Figure 3 illustrates the range of products andoxygen levels in steels treated with various deoxi-dizers.9 The upper product is a ‘‘rimmed’’ steel. Theseterms, rimmed, semikilled, and killed, describe thebehavior of molten steel as it freezes in an ingot moldand the macrostructures of cross-sectioned ingots.10
For reference, from Eq. 3 at 1600�C the maximumsolubility of oxygen in pure, liquid iron is 2290 partsper million (ppm) at log p(O2) = �8.2.4,9,10
log O½ �ppm¼ 6:734� 6320=T at saturationð Þ (3)
THEORY AND PRACTICE
Equation 4 relates the oxygen content of steel andcell voltage of a Cr/Cr2O3 sensor.8 From this, theNernst equation, and tables of standard thermody-namic data, Eqs. 5–9 were written for 1600�C.Equation 8 would give the potential noted above,�8.2 at 2,290 ppm [O], as �7.9. The data used todevelop Eq. 8, however, did not extend to concen-trations greater than 1000 ppm [O]. As previously,the standard electrode potential for Cr2O3
(�1120 mV at 1600�C) in Eq. 10 was taken to bethat for Eq. 11—the opposite sign of the formationreaction.1 This was designated for emphasis in theoriginal paper as E��. At 1600�C and an oxygenconcentration of 40 ppm, the value of E in Eq. 10 is
Mo/MoO2 Reference, Lead and Zinc
1200 oC
Cr/Cr2O3 Reference, Iron, 1600 oCy = -92.87x - 1120
-800
-600
-400
-200
0
200
400
600
800
-16.0 -14.0 -12.0 -10.0 -8.0 -6.0 -4.0 -2.0 0.0
log pO2 (bar)
E, m
V
Ni/NiO Reference, Copper, 1150 oCy = -70.56x - 590
Air Reference, Copper, 1150 oCy = -70.56x - 48
Cr/Cr2O3 Reference, Lead and Zinc Smelting, 1200 oC
Fig. 2. Applications of solid electrolyte oxygen sensors in smelting and refining.
0 50 100 150 200 250 300
[O], ppm
Deo
xid
izer
Fe/Mn
Si/Mn, semi-killed
Si/Mn/Al, semi-killed
Si/Mn/Ca, semi-killed
Al, killed
Fig. 3. Oxygen contents of deoxidized steels.
0
100
200
300
400
500
600
0.0 0.5 1.0 1.5 2.0 2.5
wt% [C]
[O],
pp
m T = 1600 oC p(CO) = 1 bar%[C] x %[O] = 0.0020
Fig. 1. Oxygen and carbon dissolved in molten iron at 1600�C.
Oxidation Potentials in Iron and Steel Making 1585
zero and log p(O2) = �12.1. At oxygen concentra-tions less than this and oxidation potentials morenegative than �12.1, the Cr/Cr2O3 reference elec-trode is positive relative to the working electrode, asseen in Fig. 2.
Equations 4–9 are written for T = 1600�C =1873 K.
E ¼ �217 log O½ �ppmþ 348 (4)
log O½ �ppm¼ �0:0046E þ 1:604 (5)
E ¼ �92:87 log p O2ð ÞFe� 1120 (6)
log p O2ð ÞFe¼ �0:0108E� 12:06 (7)
log p O2ð ÞFe¼ 2:34 log O½ �ppm� 15:81 (8)
log O½ �ppm¼ 0:427 log p O2ð ÞFeþ 6:76 (9)
E ¼ E� � 2:3RT=zF log p O2ð ÞFe
E�
E��� �
1873K¼ �1120 mV
(10)
2=3 Cr2O3 ¼ 2=3 Cr þ O2
DG� ¼ 749:0� 0:169T; kJ=mol O2
(11)
Engineers new to this field should be prepared tofind in the technical literature every possible rear-rangement of terms in Eqs. 2 and 10 and everyconceivable choice of signs, within the equationsand those assigned to E and E�. The polarity of the
thermoelectric emf between the dissimilar metalcouple of reference and working electrodes mustalso be noted.
Equation 4 is shown graphically in Fig. 4. Thedata points were extracted from trade literature forsteelmaking operations.11
INDUSTRIAL APPLICATIONS
Table I summarizes typical composition rangesfor the slags of nonferrous and ferrous productionfurnaces: Cu(Ni) flash, Cu(Ni) Peirce-Smith con-verter, lead blast, iron blast, Bessemer converter,basic open hearth, basic oxygen, electric arc, andrefining ladles. The slags may be classified as ferro-silicates or calcium-ferro-silicates, containing vary-ing proportions of alumina and magnesia—gener-ally in the olivine family of minerals. Thethermodynamic properties of the various systemsare determined by the temperature, slag compo-nents, oxidation of ferrous to ferric iron, and ap-proach to equilibrium between the slag and metal ormatte. The ratio of ferric to ferrous iron (Fe+3/Fe+2)for a given temperature can be related to the oxi-dation potential. Equilibrium is approached in theflash, blast, open hearth, and electric smelting fur-naces, but less so in the converters and ladle refin-ing furnaces. The primary function of nonferrousslags is to absorb iron oxides in a form suitable fordiscard; that of ferrous slags is to absorb theimpurities oxidized during refining. Slag productionrelative to the quantity of metal (or matte) is higherin nonferrous smelting than in the production ofiron and steel. The times between metal and slagtaps in nonferrous smelting are measured inhours—minutes in the production of iron and steelexcept in the blast and open hearth furnaces.
Taking the carbon content of blast furnace pig iron tobe 4%, the oxidation potential can be estimated fromFig. 1 and Eq. 8 to be around�13 for a temperature of1600�C. This would correspond to a CO/CO2 ratio of1400 in the gases above the slag. But it is also inter-esting to note that blast furnace slags contain virtuallyno iron and to recall a suggestion made by Schuh-mann16 ‘‘… oxide slags without iron do not necessarilyeven have an oxygen activity’’; that is, they may nothave oxidation potentials. This statement could be re-placed by one that stresses that slags must contain oneof the transition metals in order for oxidation/reductionreactions to occur and, therefore, for there to be anoxidation potential—but it is still an intriguing point.
Richards13 in 1918, gave the composition of aBessemer steel blown from blast furnace pig iron as0.04% C, 0.02% Si, 0.01% Mn, and ‘‘probably lessthan 0.3% oxygen’’ (3000 ppm) with a temperatureof 1500�C. The maximum soluble oxygen at 1600�Cwas given above as 2290 ppm with log p(O2) = �8.2;at 1500, [O] = 1477 ppm and log p(O2) = �9.0.
Oxidation potentials in open hearth furnaces at1550�C to 1600�C were estimated indirectly to be
-400
-300
-200
-100
0
100
200
300
400
0 100 200 300 400 500 600 700
[O], ppm
E, m
V
Cr/Cr 2O3 Reference 1600 oCE = - 217 log [O] + 348
Fig. 4. EMF of solid electrolyte oxygen sensors and the oxygencontent of steel
Matousek1586
approximately �8.3 for steel oxidized with iron oreand �7.2 with oxygen lancing.10 It was noted in1960, in a comparison of the open hearth furnacewith basic oxygen and other processes then beingproposed, that operation of the open hearth was stillmore of an art than a science.17 The availability ofsolid electrolyte oxygen sensors might have changedthat, and it is unfortunate that a working openheath furnace has not been preserved to test thishypothesis.
Kawakami and co-workers14 examined the appli-cation of solid electrolyte sensors to the slag andmetal of an industrial LD (basic oxygen) converter;the slag composition is given in Table I. For anominal temperature of 1600�C, the oxidationpotentials of the slags averaged �8.1 and that of thesteels averaged �9.6. For the steels, the constant inthe Fig. 1 relationship, m = %[C] 9 %[O], wasapproximately 0.005.
Similar work on ladle furnace operations was re-ported by Riyahimalayeri et al.15 and Coletti and co-workers18,19. The first reference gives composition,Table I below, and temperatures (1487�C to 1546�C)of slags. The latter studies report 25 industrialoxidation potential measurements of slags and alu-minum-killed steels: log p(O2) = �10 for the slagsand �14 for the steels. This example and that abovefor the BOF support the suggestion made previously
of the probable nonequilibrium between metals andslags in these furnaces. From Eq. 8, the steel po-tential would suggest an oxygen concentration ofapproximately 6 ppm, in general agreement withFig. 3.
Oxygen control is also a quality concern in ironfoundry operations; applications of oxygen sensorsin this industry were reviewed by Mampaey et al.20
SUMMARY
Nominal temperatures and oxidation potentialsfor iron and steel making operations are summa-rized in Table II and compared with nonferroussmelting in Fig. 5. Solid electrolyte oxygen sensorshave become important quality-control tools in theladle refining and continuous casting of steel; how-ever, they have not been used to their full advan-tage. Data from active operations may be stored incompany files, but there seems to be no publishedcases of oxygen sensors being used to monitor ironand steel making from start to finish—the slag andpig iron from a blast furnace, the compositions ofgases within the furnace, the course of oxidation ina BOF, the composition of BOF exhaust gases as afunction of time, or the progression of refining andgas compositions of ladle furnaces. Missing are theresults of a coordinated effort to obtain measured
Table I. Compositions of nonferrous and ferrous metallurgical slags
Furnace %FeOx %SiO2 %CaO %MgO %Al2O3 %MnO
Cu(Ni) flash 50–60 25–35 3–5 3–5 5–10 –Cu(Ni) converter 65 25 3 3 4 –Pb(Zn) blast 25–35 20–25 15–25 – 3–5 –Iron blast12 <1 37 40 10 9 1Bessemer converter13a 10–25 10–25 40–50 1–5 +P2O5 5–10Bessemer converter13b 15–25 45–65 1–2 1–2 2–5 10–30Basic open hearth4 15 18 44 7 2 11Basic oxygen9 26 12 48 6 1 5Basic oxygen (LD)14 16–24 15–20 37–44 10–12 +P2O5 7–8Electric arc9 10–30 5–15 40–60 3–8 – 2–5Ladle refining15 – 2–5 50–60 3–8 30–38 –
aBasic (Thomas) process.bAcid process.
Table II. Nominal temperatures and oxidation potentials in iron and steel making
Operation Temperature (�C) log pO2 (bar)
Cu (Ni) flash smelting 1250 –7.5Cu(Ni) converting 1275 –7.5Pb(Zn) blast furnace smelting 1200 –10.5Fe blast furnace smelting—iron 1600 –13.0Bessemer converter—steel 1500 –9.0Basic open hearth furnace—steel 1600 –8.0Basic oxygen furnace—slag/steel 1600 –8.0/–10.0Electric arc furnace—steel 1600 –10.5Ladle refining—slag/steel 1525 –10.0/–14.0
Oxidation Potentials in Iron and Steel Making 1587
oxidation potentials across the full range of pro-cessing steps. To paraphrase the Chinese prov-erb—a few measurements would be worth athousand calculations or ten thousand guesses.
ACKNOWLEDGEMENTS
Appreciation is expressed to Professor GordonIrons, McMaster University, for guidance in theformative stage of this article.
REFERENCES
1. J.W. Matousek, JOM 62, 75 (2010).2. J.W. Matousek, JOM 63, 63 (2011).3. J.W. Matousek, JOM 64, 1314 (2012).4. W.O. Philbrook and M.B. Bever, eds., Basic Open Hearth
Steelmaking (New York: AIME, 1951).5. W.A. Fisher and W. Ackermann, Arch. f.d. Eisenhuttenwe-
sen Part I 36, 643 (1965).6. W.A. Fisher and W. Ackermann, Arch fd Eisenhuttenwesen
Part II 36, 695 (1965).7. G.R. Fitterer, J. Metals 18, 961 (1966).8. R.J. Fruehan, L.J. Martonik, and E.T. Turkdogan, Trans.
Met. Soc. AIME 245, 1501 (1969).
9. R.J. Fruehan, eds., The Making, Shaping and Treating ofSteel, Steelmaking and Refining (Pittsburgh: AISI SteelFoundation, 1998).
10. H.E. McGannon, eds., The Making, Shaping and Treating ofSteel (Pittsburgh: United States Steel, 1964).
11. Heraeus Electro-Nite (Celox), Langhome, PA, www.heraeus-electro-nite.com.
12. D.H. Wakelin, eds., The Making, Shaping and Treating ofSteel, Iron Making (Pittsburgh: AISI Steel Foundation, 1999).
13. J.W. Richards, Metallurgical Calculations (New York:McGraw Hill, 1918).
14. M. Kawakami, K.S. Goto, and M. Matsuoka, Metall. Trans.B 11B, 463 (1980).
15. K. Riyahimalayeri, P. Olund, and M. Selleby, Steel Res. Int.84, 136 (2013).
16. R. Schuhmann, Reinhardt Schuhmann International Sym-posium (Warrendale: TMS, 1986), p. 567.
17. E.F. Kurzinsky, Iron Steel Eng. 37, 65 (1960).18. B. Coletti, S. Smets, B. Blanpain, P. Wollants, J. Plessers, C.
Vercruyssen, and B. Gommers, Ironmaking Steelmaking 30,101 (2003).
19. S. Smets, J. Janssens, B. Coletti, J. Plessers, B. Blanpain,and P. Wollants, International Conference on Molten Slags,Fluxes and Salts (Johannesburg: SAIMM, 2004), p. 687.
20. F. Mampaey, D. Habets, J. Plessers, and F. Seutens, Int.Found. Res. 60, 2 (2008).
-11
-10
-9
-8
-7
-6
pO
2(b
ar)
QSL Oxidation (Smelting)Copper and Nickel Matte Smelting
QSL ReductionLead Blast Furnace
Zinc Slag-Fuming Furnace
Zinc Sulfide Fuming
Ausmelt Fce. --Slag Cleaning/Zinc Fuming
Basic Open Hearth Furnace
Basic Oxygen FurnaceElectric Arc Furnace
Bessemer Converter
-16
-15
-14
-13
-12
900 1000 1100 1200 1300 1400 1500 1600 1700
log
p
Temperature, °C
ISF Dicard SlagKIVCET, Coke Checker
Imperial Smelting FurnaceOff-Gas
Iron Blast Furnace
Ladle Furnace
Fig. 5. Oxidation potentials in pyrometallurgical operations.
Matousek1588