WWW.IRENG.ORG WWW.IRENG.ORGJULY 2013 ISSUE JULY 2013 ISSUE

20 21TECHNICAL PAPER TECHNICAL PAPER

Magnesia-Carbon Brick: a RetrospectiveRuth Engel

IntroductionThe steel industry is the biggest consumer of refractories and the main user of magnesia based ones. Consequently, improvements and new developments of magnesia based refractory types refl ect the demands of this industry.

The continual need to increase process vessel availability, improve life and decrease refractory lining cost has been the driver of many technological changes. The development of what today is understood as a magnesia-carbon brick, a resin bonded magnesia brick with graphite and often antioxidant additions, resulted from the quest to develop better refractories for the high wear areas of converters. This overview will cover the early days of magnesia-carbon brick and some of the developments that have taken place with this technology.

HistorySince, at least, the 1870s pressed basic raw materials, in particular burned dolomite, mixed with pitch have been available to line vessels for use in different processes.

The advent of the BOF in the late 1950s and early 60s was accompanied with developments in tar bonded dolomite, mag- dolomite and magnesia brick [1]. High strength burned MgO brick was introduced in the late 1960s and by the 1970s BOFs were being lined with pitch bonded magnesia brick and also burned and pitch impregnated MgO brick. The latter became the standard for the charge pad and other high wear areas which was the beginning of zoned linings [2]. Pitch bearing brick required some manufacturing fi ne tuning in order to facilitate their installation at the user’s site as excess pitch covered the brick bonding them together while in transit to the installation site, made them slippery to handle and modifi ed their dimensions. During this time period the expected converter life was in the 200 to 500 heat range.

By the late 1970s the need for better refractories became acute as a result of moving from ingot to continuous casting which increased the maximum temperature of the steel in the converter from the 1500oC - 1600oC range to 1650oC or 1700oC or even higher. In addition, changes in steelmaking processes added stresses to the vessels further increasing refractory wear.

Between 1975 and 1980 magnesia-carbon refractories were developed and started to be used in Japan (Figure 1) fi rst for electric arc furnace hot spots and shortly thereafter for BOFs. This entailed the replacement of pitch with resin and allowed for a dramatic increase in the carbon level as a result of graphite additions. Subsequently, antioxidants were added to the magnesia-carbon brick to protect the carbon from oxidation.

Figure 1: Annual production of MgO-C refractories [3]

The use of magnesia carbon refractories in the high wear areas of the BOF led to a signifi cant reduction in refractory consumption and those brick were soon being incorporated into other areas as well until the whole vessel was lined with them. Although refractories are an important component of the improvement in BOF life, other changes such as better temperature control, development and improved control of slag chemistry, implementation of slag splashing technology, improved refractory repair mixes, etc. have all contributed to the ever increasing BOF life. By 2008 a North American BOF was expected to be relined after 20,000 to 35,000 heats while a European one after 2,500 to 3,000 heats [4] and these numbers have almost doubled by today.

Magnesia Much confusion can arise from the term magnesia refractory. The American Society for Testing and Materials (ASTM) defi nes it as “a dead-burned refractory material consisting predominantly of crystalline magnesium oxide”, but that is not the terminology found in the literature. The terms magnesite, magnesia and MgO are used interchangeably to signify magnesium oxide. In addition, older publications use the term periclase for MgO, even though this is the name of the naturally occurring mineral of that composition. Over time different raw materials have been the source for the MgO used in refractories. These changes have been driven by technology, economics, availability and the expectation of continuously improving refractory properties. Today, most magnesia refractories contain sintered and/or fused grain, both being a synthetic product.

Sintered magnesia, also called dead burned magnesite, is produced by heating, sintering, a Mg bearing mineral or a synthetic compound to drive off volatile gasses. When the resultant product is further benefi ciated by melting in an electric arc furnace, it becomes fused

magnesia. The amount and chemistry of the impurities present affect the high temperature grain properties. Many papers contain sections discussing the properties of magnesia grain, the changes that have taken place in its production over time and the effect of impurities at high temperature [1,5,6, etc.]. It should be noted that there has been a continuous improvement in magnesia grain quality. This has resulted in the availability of bigger grains with higher density and the continuous lowering of the SiO2 and other impurities which lead to low melting point phases at high temperature [6]. The decrease in SiO2 can be observed in the changes in lime to silica ratios from 2.6, in the early days, to about 4 with improvements in technology and the current level of 10 or higher. Use of these improved grains has contributed to the production of brick with ever better physical properties and slag resistance.

CarbonCarbon is added to refractories because it is not wetted by slags. Traditionally, refractory chemical corrosion resistance was improved by decreasing the size or number of pores, but this could make them more sensitive to thermal shock damage. Consequently, methods were developed to incorporate carbon to fi ll the spaces between the magnesia, or other refractory oxide grains. The carbon then forms a fi lm blocking the slag penetration thereby minimizing the damage it causes the brick. The early carbon sources were tars which have been replaced by pitches or resins. The carbon level so achieved can be further increased with the addition of carbon black and graphite.

Pitch could be used to hold the refractory components together as in pitch bonded brick or forced into the brick’s pores as in burned pitch impregnated brick. The residual carbon contained after tempering and coking, was approximately 5 % for pitch bonded brick and 2.5 % or slightly higher, for burned impregnated brick. Contrast this with magnesia-carbon brick which can have 8 to 30% carbon while the most common range is between 10 and 20%. The carbon in burned impregnated brick also affected their porosity by decreasing it from approximately 18% to about 12%, but did not detrimentally affect the brick’s physical properties.

The graphite used in refractories is a naturally occurring fl ake material selected for its high temperature stability and chemical inertness. The important properties are fl ake size, carbon content and ash or, impurity level. A consequence of the fl ake shape is that during manufacture they preferentially align themselves leading to anisotropic brick properties. In particular, its thermal conductivity is higher in the long vs. short direction.

Magnesia-CarbonThe ground work for magnesia-carbon refractories was set, among others,

by Herron et.al. [8], who showed that the carbon in pitch impregnated burned magnesia brick prevented slag from penetrating the pores.

Dense zone formation, thought to be a characteristic of magnesia-carbon refractories, was known before these brick were developed [9, 10]. Work using laboratory generated and fi eld samples of pitch bonded and pitch impregnated magnesia and dolomite brick showed that an MgO dense zone formed at high temperature protecting the carbon and thereby preventing slag penetration. Figure 3 shows a part of a “relatively continuous dense magnesia zone between the carbon-bearing zone and the carbon-free zone” observed in a pitch impregnated brick fi red in the laboratory using conditions that simulated the BOF environment. The dense zone was thought to be the result of the reduction of the MgO in the presence of carbon and at steelmaking temperatures, which vaporized the Mg as a gas. There was a point at which the oxygen potential and temperature relationship was no longer stable leading to the precipitation of Mg vapor to form a magnesia dense zone [10].

Figure 3: Laboratory fi red pitch containing 95% MgO brick, showing “bridging” of MgO grains (x130) [10]

The concept of adding a metallic component to a refractory was already patented in 1935 [11], a long time before the technology was available to produce magnesia-carbon brick. In 1983 a patent for the addition of metallic Mg to a chemically bonded brick was issued. The idea was to increase the amount of Mg gas generated which should lead to a thicker dense zone [12]. This concept was reduced to practice, and a commercial product was successfully manufactured. To carry out production several hurdles had to be overcome, not the least of which was how to handle metallic Mg, a spontaneously fl ammable material in the presence of nitrogen, carbon dioxide, oxygen and/or water.

With the increased use of magnesia-carbon refractories came the realization that graphite, in the 10% and higher level, behaved in a completely different manner than the residual carbons from pitch or resins. Figure 4 shows that the higher C levels decreased slag penetration, increased the brick’s thermal conductivity leading to more effective cooling of the hot face, and reduced its modulus of elasticity (Young’s modulus) with the latter two improving its thermal shock resistance.

Figure 2: Magnesia carbon brick micrographs showing the graphite fl ake alignment: A) perpendicular and B) parallel to brick’s long dimension [7]

50000

100000

5 10 15Year

Prod

ucti

on (

T/an

o)

WWW.IRENG.ORG WWW.IRENG.ORGJULY 2013 ISSUE JULY 2013 ISSUE

22 23TECHNICAL PAPER TECHNICAL PAPER

Figure 4: Effect of graphite on slag resistance, thermal conductivity and Young’s modulus on the MgO-C brick [13]

To ensure the brick’s carbon would be available for a long time its rate of oxidation had to be reduced. The formation of the dense zone to protect it was already known and ways to increase its speed of formation and/or its stability were being investigated. To this end Al and/or Si metals were added to commercial magnesia-carbon brick, but a major problem remained, the laboratory determined dense zone could not be detected in post mortem brick [14]. It was speculated that the technique used for brick removal from actual fi eld installations somehow destroyed it or that slag penetration during cooling was to blame. By 1997, this problem was no longer an issue and dense zones from used brick were being analyzed [15].

Rymon-Lipinski [16] with his seminal work, provided the theoretical basis for determining the steps in which Mg, Al, and Si metals react to provide the carbon protection in a magnesia-carbon brick. Table 1 shows the expected reactions as a function of temperature and position within the BOF lining.

Zone I (CO2, CO, N2)

MgO + MA MgO + M2SMA = MgO Al2O3 MgO M2S = 2MgO SiO2(Spinel)

Zone II (CO, N2)

> 2,080 K: > 1,800 K:MgO + AlN + C MgO + M2S + C< 2,080 K: MgO + C < 1,800 K:MgO + MA + C MgO + SiC + C

Zone III (N2)

MgO + AlN + C > 1,600 K: > 1,700 K: MgO + Mg + C MgO + SiC + C < 1,600 K: < 1,700 K: MgO + Mg3N2 + C MgO + Si3N4 + C

Zone IV(Neutral Atmosphere)

>1,380 K: > 1,810 K:MgO + Mg + C MgO + M2S + Mg +C<1,380 K: MgO + Mg + C < 1,810 K:MgO + Al4C3 + C MgO + SiC + C

Zone V(Unchanged Material)

MgO + Al + C MgO + Mg + C MgO + Si + C

Table 1: Reaction products for Al, Mg and Si with the refractory at different temperatures in a BOF lining with Zone I representing the hot face and Zone V the as received brick [16].

Figure 5 shows the expected dense zone morphology of a metal containing magnesia-carbon brick. Compare this to the one shown in Figure 3 which is more of a bridging between grains. In the latter case the amount of contained carbon is considerably lower than that found in a modern magnesia-carbon brick which results in the MgO grains being very close to each thereby providing a scaffold for the newly precipitated material.

Figure 5: Dense zone formation in a magnesia-carbon refractory [17]

If the metal added to the brick was Al then the dense zone will consists of an MgAl spinel not pure MgO. Figure 6 shows the reaction products from the hot face to the cold face of a used magnesia-carbon brick with Al addition. Note, that in addition to the MgAl spinel, nitride and carbide phases also formed while the brick was in service.

Figure 6: Distribution of reaction products in used MgO-C brick [18]

Although much has been written about the addition of Al, Si and Mg to magnesia-carbon brick other additives have also been investigated. At some point, interest was shown in additions of glass, Mg-B, TiO2, borides and carbides [19, 20, 21]. Of these, B4C has been found to be a useful addition to magnesia-carbon brick. The microstructures resulting from the addition of different antioxidants to magnesia-carbon brick was published by Zhang et. al. [22] and a review of more current additives was assembled by Luz and Pandolfelli in 2007 [23].

Antioxidants also improve the strength of the brick as compared to similar ones without it. Figure 7 shows the effect of Al on the brick’s strength after fi ring to different temperatures [24].

Figure 7: Relative crushing strength of 15% magnesia-graphite samples with and without Al additions. B.O.= before oxidation [24]

At the same time that the early research on antioxidants was being carried out, magnesia and carbon as raw materials and in contact with each other were studied to establish the important parameters for achieving good refractory properties which would lead to increased life in actual applications [25].

DolomiteBaker, et. al. [10] showed evidence of a dense zone formation in pitch bonded dolomite, similar to that observed in magnesia brick. Following up on this work, studies with antioxidant additions showed that boron bearing compounds were more effective than metals in forming a dense zone, but their presence had a detrimental effect on slag corrosion [26] so it was not pursued.

SummaryAlthough magnesia-carbon refractories showed low wear in actual applications the move from the laboratory to the steel plant was slow until the rate of carbon oxidation could be controlled. Much research was carried out to determine its role and how to reduce its oxidation. This work traced the development of the early magnesia carbon brick and the role of antioxidants in the dense zone formation. Their role can be summarized as follows:

At low temperatures, less than around 1200oC, there is a decrease in the brick’s porosity

Above about 1400oC graphite will oxidize in the presence of MgO and both Mg and C will be lost from the refractory.

If the refractory is above about 1500oC for long periods of time and, in the presence of carbon, an Mg vapor forms which reacts to form the MgO or an MgAl -spinel dense zone at the brick’s hot face. This prevents the ingress of air or slag into the brick thereby “resisting” oxidation.

The commonly added antioxidants of Al and Si are used because they are low cost and afford effective protection.

Borides, in particular B4C, are added because they fi ll pores at intermediate temperatures thereby preventing air to react with the refractory components.

It is interesting to note that brick based on this technology are used with great success, not only in the BOFs, but also in most EAFs and ladles and the metal addition concepts were expanded to encompass Al2O3-SiC-C (ASC) brick, SENs, slide gate plates, etc.

Post script: the biggest challenge to writing this overview was acquiring copies of the cited papers and many others that I used for background. I would like to thank all the libraries which have kept journals and conference proceedings from years gone by.

ReferencesI would like to thank N. Sullivan for his critical reading of this manuscript.

[1] A.M. Garbers-Craig: Presidential address: How Cool are Refractory Materials?, The Journal of The Southern African Institute of Mining and Metallurgy, V 108, p. 1, 2008

[2] E. Ruh: International Ceramic Monographs, V 1, no. 2, pp. 772-93, 1994

[3] T. Hayashi: Recent development of Refractories Technology in Japan, First International Conf. on Refractories, Japan, 1983

[4] H.J. Junger, C. Jandl, J. Cappel: Relationships Between Basic Oxygen Furnace Maintenance Strategies and Steelmaking Productivity, Iron & Steel Tech.,V5, #11, 2008

[5] C. Alvarez, E. Criado, C. Baudin: Refractarios de magnesia-grafi to, Bol. Soc. Esp. Ceram. Vidri., V 31, #5, 1992

[6], R. A. Landy: Magnesia Refractories, Chapter 5 in Refractories Handbook, ed. C. Schacht, 2004

[7] W E Lee @ http://core.materials.ac.uk

[8] R.H. Herron, C.R. Beechan, R.C. Padfield: Slag attack on carbon-bearing basic refractories, Amer. Ceram. Soc. Bull, V 46, p 1163–1168, 1967

[9] B. Brezny, R. Landy: Microstructural and Chemical Changes of Pitch Impregnated Magnesite Brick under Reducing Conditions, Trans. and Journal of the British Ceramic Soc., V 71, p. 163, 1972

[10] B. H. Baker, B. Brezny, R.L. Shultz: Role of Carbon in Steel Plant Refractories, Am. Ceram. Soc. Bull., V 55, # 7, p 649, 1976

[11] US Patent #2,013,625, 1935

[12] US Patent #4,407,972, 1983

[13] T. Ayashi : Recent Trends in Japanese Refractory Technology, Transactions ISIJ, V21, 1981

[14] J.D. Smith, R.E. Moore: Thermochemical evaluation of Dense MgO Layer Formation, Proceedings UNITECR, 1993

[15] M. Karakus, J.D. Smith, R.E. Moore: Mineralogy of the Carbon Containing Steelmaking Refractories, Proceedings UNITECR, 1997

[16] T. Rymon-Lipinski: Reaktionen von Metallzusatzen in Magnesia-Kohlenstoffsteinen in einem Sauerstoffknoverter, Teil 1, p 1049 (47), Teil 2, p 1055 (53), Stahl u. Eisen, V. 108, # 22, 1988

[17] I. Strawbridge, D.G. Apostolopoulos: High Purity Magnesias and Graphites in Magnesia-Carbon Refractories, Stahl und Eisen, Special, 1994

[18] K. Ichikawa, H. Nishio, O. Nomura, Y. Hoshiyama “Suppression Effects of Aluminum on Oxidation of MgO-C Bricks, Taikabutso Overseas, V 15, #2, 1995

[19] K. Kurata, T. Matsui, K. Kono: Oxidizing Prevention of MgO-C Bricks for Converter, Taikabutso Overseas, V 11, #2, 1991

[20] T. Suruga: Effect of Mg-B Material Addition to MgO-C Bricks, Taikabutso Overseas, V 15, #2, 1995

[21] C.G. Aneziris, J. Hubalkova, R. Barabas: Microstructure evaluation of MgO–C refractories with TiO2- and Al-additions, J. of the European Ceramic Society, V 27, 2007

[22] S. Zhang, N.J. Marriott, W.E. Lee: Thermochemistry and Microstructures of MgO-C Refractories Containing Various Antioxidants, J. of the European Ceram. Soc.,V 21, p 1037, 2001

[23] A. P. Luz, V. C. Pandolfelli: Review Article: Performance of the Antioxidants in Carbon Containing Refractories”, Cerâmica, V 53, 2007

[24] M. Rigaud, P. Bombard, X. Li, B. Gueroult: Phase Evolution in Various Carbon-Bonded Basic Refratories, Proceedings UNITECR, 1993

[25] H. Ishii, M. Nagafune, I. Tsuchiya: Reaction between Magnesia and Carbon at High Temperature, Taikabutso Overseas, V 11, #2, June, 1991

[26] F. Ye, M. Rigaud, X. Zhong: Effects of Boron Bearing Additives on Oxidation and Corrosion Resistance of Doloma-Based Carbon Bonded Refractories, Proceedings UNITECR, 1997

MgO-A2O3

ALNA14C3

A1Without additive

With 4 wt% Al

6000

5000

7000

4000 0

2

4

6

8

10

12

50 10 15 20

Thermalconductivity

Graphite content (%)

Youn

gs m

odul

us (

kg/m

m2)

Dept

h of

sla

g pe

netr

atio

n (m

m)

Ther

mal

con

duct

ivity

(kca

l/m.h

r.oC)

at 1

000o

C

(mm)25

Young’s modulus

Depth of slagpenetration

0

10

20

30

40

50

60

MgO-A2O3

AlNAl4C3

Al

Without additiveWith 4 wt% Al

10

50

00 100 200 300

BackHot face

Rela

tive

inte

nsit

y of

X.R

.D p

eaks

(mm)

Graphite Content: 15 wt%Firing Time: 5 hours

Without additive

With 4 wt% Al

60

50

40

30

20

10

0B.O. 1000 1200

Rela

tive

Col

d Cr

ushi

ng S

tren

gth

(kgk

/cm

2)

Hot face - slag

Dense zone

Unreacted brick

Temperature oC