Determination of the intrinsic kinetics of iron oxide reduced by carbon monoxide in an isothermal differential micro-packed bed☆

Baolin Hou,Haiying Zhang,Hongzhong Li*,Qingshan Zhu*

State Key Laboratory of Multi-phase Complex Systems,Institute of Process Engineering,Chinese Academy of Sciences,Beijing 100190,China

ABSTRACT The intrinsic kinetics of iron oxide reduced by carbon monoxide is evaluated by a method of online measuring concentration of off-gas in an isothermal differential micro-packed bed.Under the condition of getting away from the in fluence of gas diffusion and gas–solid heat transfer and mass transfer,the reaction of Fe2O3 to Fe3O4,Fe3O4 to FeO and FeO to Fe in the process of single reaction can be clearly distinguished from each other,and the relevantactivation energy is characterized to be 75.4,74.4,and 84.0 kJ·mol-1,respectively.Therefore,the change ofsurface area in the reaction process due to losing oxygen could be easily calculated by combining it with pre-exponential parameters of Arrhenius equations.In conclusion,these kinetic parameters are verified by the experimental data for the process of ore reduced by carbon monoxide in a packed bed.

1.Introduction

The steel and iron industry is well known as the foundation of other manufacturers,which is very important for the development of human being society.In recentdecades,due to the shortage ofhigh-quality coking coals and the environmental problems,some researchers focused attentions on developing a new,environmentally friendly and economical process,for example the direct reduction process with hydrogen or carbon monoxide.However,because of the limitation of understanding the relevant intrinsic kinetics of iron oxide,people often conduct this process at the temperature higher than 1073 K to improve the rate of chemical reaction,and thus,the extensive energy will have to be consumed,which can partly account for the difficulty of commercial producing sponge iron by this process.Therefore,until now,the main method of producing iron is still the blast furnace.

There are two main factors involving in reducing the rate of direct reduction iron:one is the sponge iron sintered；another is the limitation of chemical equilibrium.In the process of FeO reduced to sponge iron,the sponge iron newly generated is very active and easily sintered,which must lead to the serious resistance of gas intraparticle diffusion,especially at the relative high reaction temperature.Additionally,due to the limitation of chemical thermodynamic,the equilibrium conversion of hydrogen and carbon monoxide is less than 30%in the reduction process of FeO[1].Therefore,the main methods for reducing energy consumption of the direct reduction of ore are to reduce the resistance of gas intraparticle diffusion and to improve the utilization ratio of reducing gas by employing an appropriate technical process.In our group's work[2],the bubbling fluidized bed reactortechnology with hydrogen and carbon monoxide as the fluidizing gas at the relatively low temperature was developed,and the several parallel fluidized beds were used to improve the utilization rate of gas.For example,the exhausting gas of fluidized bed for the reaction of FeO to Fe is used as the reducing gas for the reaction of Fe3O4to FeO or Fe2O3to Fe3O4.

However,in order to scale up and design the fluidized bed,the mechanism and kinetic parameters of Fe2O3reduced by hydrogen and carbon monoxide need to be studied.Although there are numerous works about this published literature[3,4,5,6,7],it is difficult to select an appropriate value due to the differences of activation energy reached at least one order of magnitude among the results reported by different authors[8,9].The main reasons leading to these differences can be expressed as follows:

(1)The in fluence of gas diffusion was not completely eliminated.In the case of thermal gravitation(TG)study,the balance requires a quiescent atmosphere to obtain accurate measurement of weight,and gas velocity should not be higher than 0.01 m·s-1in mostcases,which is too slow to eliminate the mass transfer resistance for a fast reaction like the iron oxide reduction.

(2)The reaction mechanism for reducing iron oxide varies with temperature.In-situ XRD[8,9]investigations showed the dependence of reduction route on various temperatures as follows:

Obviously,it is impossible to clearly distinguish each reaction separately using temperature program reduction(TPR)and TG techniques,especially when the heating-up speed is relatively high[10,11,12].

(3)In the literature,over 20 semi-empirical equations[13]have been proposed for the reduction reactions of iron oxides,which were employed to fit experimental data for exploring the reaction mechanisms.However,it is often found that one set of experimental data could be well fitted by the different empirical equations[14]when the least square method was used,but the Arrhenius parameters obtained were of significant difference.

(4)When the reduction process of Fe2O3happened,the surface area ofsolid would be gradually increased due to oxygen atomdepleted.In the literature,the pure Fe3O4or FeO solids were often employed to evaluate the kinetic parameters.However,correctly quantifying the change ofsolid's surface area in the process ofreducing Fe2O3is difficult,so it is impossible to accurately predict this process based on these available kinetic parameters.

Based on the above discussions,it is meaningful for developing a new method to accurately obtain the kinetic parameters of Fe2O3reduced by hydrogen and carbon monoxide.However,the corresponding kinetic parameters of Fe2O3reduced by hydrogen have been reported in our earlier work[15].In this work,the results for carbon monoxide will be detailed.Underthe conditions ofgetting away fromthe effectofmass transfer and heattransfer,the intrinsic kinetics of iron oxide reduced by carbon monoxide was determined in the isothermal differential micropacked bed.In this experiment,the reaction temperature has been set over 843 K to avoid the confusion of mechanism of reducing Fe2O3with carbon monoxide and ensure that the reduction proceeds from Eqs.(1)to(3)[16],for example,iron oxide(hematite)is reduced firstly to magnetite,then wustite,and finally to iron sponge.

In addition,through using an online process mass spectrum to measure the concentration ofexhausting gas,the three successfully different reaction steps like Eqs.(1)to(3)can be distinguished from each other.And then,the kinetic parameters are determined by fitting experimentaldata with the Arrhenius equation.In the conclusion,the experiments ofdirectly reducing ore with carbon monoxide were performed,and the experimental results are employed to validate our kinetic parameters.

2.Experiments

Iron oxide powder(purity＞99.9%,by mass,mean diameter 249 nm,BET surface area 3.9 m2·g-1)was boughtfrom Sinopharm ChemicalReagent Co.,China.To obtain iron oxide with different particle sizes,the iron oxide powder was granulated by mixing with a small amount of water,followed by heating at 873 K for 4 h.The heated granulation was then crushed and sieved to obtain iron oxide particles of desired size ranges.The iron oxide particles were diluted with γ-Al2O3of 0.1 mm by 1:1(by mass)to avoid the hotspotduring the reduction process.And in allexperiments,the mixture ofcarbon monoxide and argon(1:20,by volume)was used.

The experimentalsetup was illustrated in Fig.1,where the diameter of bed was 8 mm.Quartz wool of 3 mm thickness was placed on the top ofthe perorated plate to keep the iron oxidefine particulates from being carried out by gas.The mixture of 0.2 g iron oxide and γ-Al2O3was sandwiched between the top and bottom layers of 0.1 g γ-Al2O3,which was positioned on the top of quartz wool.A K-type thermal couple protected by a 2 mmdiameterquartz tube wasinserted into the iron oxide to measure the reaction temperature.

Fig.1.Scheme of packed bed.

Gas flow rate was controlled by a mass flow meter(Seven Star Co.,China).The quartz reactor andγ-Al2O3were firstly treated by hydrogen atthe 1073 Kfor 2 h to eliminate any measurementerrors.Before beginning each experiment,the sample was purged to desorb oxygen by argon for half an hour,and then heated to the specified temperatures in argon by the external electric heating coil.After the temperature was stabilized,the gas stream was switched to the mixture of carbon monoxide and argon,and meanwhile the off-gas composition was analyzed by a process mass spectrum(Ametek Process Instrument,USA)with the frequency of about 3 Hz.

3.Kinetic Models

The reaction order for reducing iron oxide with carbon monoxide has been discussed in literatures,and mostly recommending to be one order.Therefore,the reaction rates of iron oxide reduced by carbon monoxide were often expressed by the following equations[1,3,4]:

It is difficult to measure and predictβmand βw,but if the three reactions can be distinguished from each other,they can be evaluated together with kmCOand kwCObased on the experimental data.

The following equations can be obtained by integrating Eqs.(4),(5)and(6)when the composition of the gas phase remains constant:

The above equations can be verified when Δt is small enough so that the gas composition can be regarded as constant.

The total reduction degree of iron oxide is substituted into Eqs.(9),(10)and(11),and the following equations can be obtained:

These equations are valid just when the three reaction steps can be well distinguished,and Δt is sufficiently small.The reduction degree α,or the conversion of iron oxide,is defined as:

where,ugis the super ficial gas velocity,A is the cross area of section of reactor,cinand coutare the inlet and outlet gas concentrations respectively,mFe2O3is the mass of Fe2O3sample,and MFe2O3is the molar mass of Fe2O3.

Since the reaction speed ofhematite to magnetite is severalorders of magnitude faster than that of two others,so it is easy to distinguish this step from the other two,which was clearly indicated by the TG experiments[8,9].In our experiment,this also has been confirmed,and the results were shown in Fig.2.In these experiments,the flow rate of mixture gas was 200 ml·min-1,and the temperature is set at 973 K.As being shown in Fig.2,when the conversion α reaches 11%,there is a complete peak of carbon monoxide which implies the end of process from hematite to magnetite.When the total conversion reaches to 33%,there is significant change in the curve slope of carbon monoxide concentration,indicating that the reaction transition from magnetite–wustite to wustite–iron.Therefore,in this experiment,three steps can be distinguished from each other under the condition of isothermal temperature,and thus,Eqs.(12),(13)and(14)can be rewritten as follows:

Fig.2.Concentration of carbon monoxide,carbon dioxide and degree of reduction with time during reduction of iron oxide by carbon monoxide at the temperature 720°C.

Based on Eqs.(16),(17)and(18),only ifthe three reaction steps can be clearly distinguished from each other,the left side of Eqs.(16),(17)and(18)will keep constant,which may be further referred to in the following section.

4.Results and Discussions

4.1.Heat and mass transfer limitations

The experiments aboutthe in fluence of gas velocity and particle size on the reaction rate forhydrogen as reducing gas had been performed in our previous work[15].The reaction rate of carbon monoxide is lower than hydrogen under the same condition[8],so the in fluence of intraparticle diffusion and gas–solid mass transfer can be eliminated when the gas velocity reaches greater than 0.3 m·s-1and the particle size smaller than 0.045 mm.

The limitation of heat transfer was checked by the Mears criterion[17]:

where,ΔHris the reaction heat；R the idealgas constant,hgthe gas–solid heat transfer coefficient obtained as the following equation:

where,Cpis the heat capacity of gas,λgis the heat conductivity of gas.The calculated ΔhM,heatis lower than 0.05 in this experiment,suggesting that the heat transfer does not play an important role in the kinetic study.

4.2.Kinetic parameters

The difference of carbon monoxide concentration between inlet and outlet is shown at the various temperatures in Fig.3.In these experiments,0.2 g mixture of iron oxide of mean diameter 249 nm and γ-Al2O3with weightration 1:1 was used,and the mixture of carbon monoxide and argon was employed as the reducing gas.And when the temperature of reactor reached the setting point,the gas was transferred from argon to the mixture of carbon monoxide and argon.Meanwhile,mass spectrum began to characterize the concentration ofcarbon monoxide in off-gas.As illustrated in Fig.3,due to the transition from the purge argon to the mixture of carbon monoxide and argon,the difference of carbon monoxide concentration between inlet and outlet all firstly is increased to be a maximum value,and then decreases gradually.Therefore,in order to more accurately evaluate the intrinsic kinetic parameters,only those values after the blue dot line in Fig.3 were employed to obtain the corresponding kinetic parameters.

Fig.3.The rate of reaction under various temperatures for carbon monoxide.

To evaluate the kinetic parameters,at the right of Eq.(16)are plotted with the reduction degree in Fig.4.As being shown in Fig.4,when the reduction degree is very much close to zero,i.e.,the reaction just begin,due to the transition of gas,the values ofhappen to apparently change.However,with the reduction degree improved,the value ofcan be dealt as a constant.When the reduction degree approaches 11%,this value deviates signi ficantly from the constant value,which can be explained by a little product of Fe3O4further reduced to FeO by carbon monoxide,and this kind of trend becomes more and more clearwith the temperature further increased.In order to obtain the relative accurate kinetic parameters,those values of right side of Eq.(16)for the reduction degree between 2%and 6%were used.In Fig.5,the value of apparent activation energy is shown to be 75.4 and the pre-exponential parameter is calculated to be 1.5×102.

Fig.4.Plot of-3. against degree of reduction of iron oxide.

Fig.5.Kinetic parameters of reducing Fe2O3 to Fe3O4 by carbon monoxide.

In Fig.6,the differences of monoxide concentration between inlet and outlet for the reaction of Fe3O4reduced to FeO are shown atvarious temperatures.In these experiments,other conditions are kept the same with thatin Fig.3,exceptforthe reaction temperatures thatare relatively higher.As being shown in Fig.6,due to the fast reaction rate of Fe2O3reduced to Fe3O4by carbon monoxide,the gas–solid mass transfer dominates this reaction atthis temperature range,and therefore the effectof reaction temperature on the reaction rate of Fe2O3reduced to Fe3O4is not apparent,as being shown in Fig.6.However,with the reaction proceeding,the reaction goes into the stage of Fe3O4reduced to FeO,and meanwhile the intrinsic chemical kinetic becomes a controlling-step.And with the sample further reduced,a great deal of sponge iron would be appeared.In this stage,due to the high active sponge iron sintered,the gas intraparticle diffusion becomes the dominating factors.As a result,the in fluence of temperature on the reaction rate becomes marginal again,and this also can be validated by the SEM picture of products in Fig.7.As shown in Fig.7,the metallization ratio of all products is higherthan 90%,and the reaction time is 3 h.The productofFe for hydrogen is more easily sintered than for carbon monoxide at high temperature.

Fig.6.The reaction rate under the various temperatures for carbon monoxide.

Fig.8.Plot of- against degree of reduction of iron oxide.

Fig.7.SEM image ofiron oxide and as-reduced iron atvarious temperatures(a:iron from reduction at620 °C by CO,b:iron from reduction at620 °C by H2,c:iron from reduction at640 °C by CO,d:iron from reduction at 640 °C by H2,e:iron from reduction at 660 °C by CO,f:iron from reduction at 660 °C by H2).

Fig.9.Kinetics parameters of reducing Fe3O4 to FeO by carbon monoxide.

Fig.10.Plot of-Kr wCO S w against degree of reduction of iron oxide.

Fig.11.Kinetics parameters of reducing FeO to Fe by carbon monoxide.

In order to check the kinetic parameters in this paper,we compare the simulation results and experimental data for the process of directly reducing ore.In the experiments,a packed bed with ID 10 mm was employed as a reactor.Iron ore of 4 g from Australia[Fe 65%(by mass),surface area 1.59 m2·g-1]with a diameter of 0.15–0.22 mm was positioned on the quartz sand sintered plate,and heated to 873 K in the stream of pure argon.And then it was switched to the mixture of carbon monoxide and argon(1:1,by volume)with the total flow rate of400 ml·min-1.When the reaction wascarried outforthe desired time,the stream of reducing gas was switched to pure argon.And the sample was cooled to room temperature,followed by characterizing the reduction degree of the as-reduced sample by chemical titration method.Due to the high gas flow velocity,the change ofcarbon monoxide concentration is assumed to be neglected.In order to simply the calculation process,the in fluence of gas intraparticle diffusion is not considered.Based on the above assumptions,the comparison between simulation results based on the kinetic parameters evaluated in this paperand experimentaldata are shown in Fig.12.Because the in fluence of gas intraparticle diffusion is neglected,the reaction rate predicted with the developed kinetic parameters here overestimates the experimental data at the stage of the reduction degree lower than 11%,as being shown in Fig.12.With the reaction proceeding,the simulation results are kept consistent with the experimental data.However,when the solid conversion reaches to a certain value,the simulation results are higher than the experimental data,which is attributed to the wrong experimental results from carbon deposition on the surface of iron at temperature lower than 973 K.

Fig.12.Comparison between simulation and experimental data of reduction degree with time for reduction of ore by carbon monoxide(50%,by volume).

5.Conclusions

The method of online measuring concentration of off-gas by using mass spectrum for obtaining kinetic parameters of iron oxide reduction by carbon monoxide in an isothermal differential packed bed is reasonable.In a single reaction process,the three steps ofreaction,i.e.,Fe2O3to Fe3O4,Fe3O4to FeOand FeOto Fe,can be successfully distinguished from each other.Under the conditions of free resistance of gas intraparticle diffusion and gas–solid mass transfer,the values of activation energy together with the pre-exponential parameters were evaluated for every reaction.In addition,because the change of surface area of sample solid due to depleting oxygen in the reaction process has been treated together with the pre-exponential parameters,the values of Smand Swcan be directly replaced with Shin the expression of reaction rate for every reaction.Thus,the reaction rate can be expressed as follows:

In conclusion,the kinetic parameters developed in this paper are validated by the experiment of directly reducing ore by carbon monoxide.The good agreements show that the kinetic parameters and the calculating method in this paper are reasonable.

Nomenclature

nh,nm,nwmass number,mol

R ideal gas constant,J·mol-1·K-1

Rpradius of solid particle,m

Sh,Sm,Swsuper ficial area of solid,m2·mol-1

t time,s

T temperature,K

Subscript

h Fe2O3

m Fe3O4

w FeO