The Kinetics of the Nitrogen Reaction with Saturated Alloys*

By Fumitaka TSUKIHASHI** and Richard J. FR UEHAN * * *

Synopsis gen reaction and showed that and sulfur re- The kinetics of the nitrogen reaction with liquid Fe-C alloys was tarded the rate of the reaction. Later Inouye and investigated by using the isotope exchange technique. The effectof small Choh4) studied the rate of absorption of nitrogen at amounts of S, P, Pb, Sn, Bi and Te in Fe-Csat , alloys on the rate low nitrogen content in liquid Fe and Fe alloys constant of the interfacial nitrogen reaction was studied at l 450°C. using a sampling technique and found that at low The rate constant for Fe-Csat .-S alloys was in agreement with the sulfur and oxygen concentrations the rate of nitrogen previous resultsfor Fe-S alloys. It was also confirmedthat the influenceof absorption was half order with respect to nitrogen carbon in Fe-C-S alloys on the rate of the nitrogen reaction is negligible. The effectof tin on the rate constant is small even when the tin content is pressure. However, at high sulfur and oxygen levels the rate was proportional to the nitrogen pressure. as high as 0.8 wt%. However Pb, Bi and Te have significant effects on the rate constant and the addition of small amounts of these elementsde- Mori and Suzuki7) measured the rate of nitrogen re- creases the rate significantly. Phosphorus also decreasedthe rate constant. moval from liquid iron and found the rate to be According to the adsorption model, the interfacial reaction rate should second order with respect to dissolved nitrogen con- approach zero at high concentrationof surface active elements, but in the tent. Mowers and Pehlke8) investigated the rate of case of Fe-Csat.-P, Fe-Csat.-Pb and Fe-Csat.-Te alloys, there was solution of nitrogen in liquid Fe-Se and Fe-Te alloys evidenceof a residual rate at high concentrationsof P, Pb and Fe. using a modified Sieverts' apparatus at pressure of 1 atm and a temperature of 1 600°C. They demon- Key words: kinetics; nitrogen reaction rate; carbon saturated iron alloy strated selenium and tellurium present in low con- phosphorus; tin; lead; bismuth; tellurium; sulfur. centrations have a marked retarding effect on the rate of solution of nitrogen in liquid Fe alloys, Narita I. Introduction et a1.9)investigated the kinetics of nitrogen desorption In recent years the reaction of nitrogen with iron reaction of liquid iron and steel and found that the and steel has taken on added importance. Dissolved desorption reaction rate is second order. Ban-ya nitrogen can be harmful in several ways. For exam- et a1.12)also studied the rate of removal of nitrogen ple, it can be responsible for porosity or it may react from liquid iron investigating the influence of 0, S, to form nitrides which decrease the formability of the Si, C, Cr and Ni on the rate and they found that the steel. Nitrogen gas is used in many new ladle metal- rate is second order with respect to nitrogen content lurgical processes and subsequently can be absorbed and the surface active elements such as sulfur and into the metal. Once nitrogen is absorbed into the oxygen retard the nitrogen desorption. Amano et metal it is difficult to remove; vacuum degassing, at a1.14>studied the absorption and desorption reaction best, only removes 10 to 20 % of the nitrogen present. of nitrogen by levitation technique. They reported Therefore it is important to control the nitrogen ab- that the reaction is first order when there exists no sorption during all the steelmaking processes. Nitro- gas boundary layer and the reaction is between first gen can also be harmful in the production of cast iron. and second order when there are two boundary In particular, porosity in iron castings resulting from layers. Kadoguchi et al.17) studied the rate of ab- gas evolution can be a major problem. It may be sorption of injected nitrogen in molten iron and found possible to control the reaction of nitrogen with Fe- that the rate controlling steps are the chemical reac- Csat, alloys by the addition of surface active elements tion and the mass transfer in liquid phase. Kawa- which retard the rate. kami et al.18) studied absorption and desorption rates Several researches have been reported on the rate of nitrogen in molten iron by bottom injection. They of the reaction of nitrogen with liquid Fe alloys.l-23) found that nitrogen absorption rate from pure nitro- Pehlke and Elliott1'2) studied the thermodynamics and gen is controlled by the mass transfer in metal phase the kinetics of nitrogen reaction in liquid Fe and and the desorption of nitrogen by Ar-N2 mixture was Fe alloys using a constant pressure Sieverts' appara- controlled by the mass transfer in gas phase. tus. They found that the rate is first order with Fruehan and Martonik19) measured the rate of respect to the nitrogen content and at low sulfur and nitrogen absorption into and desorption from liquid oxygen levels the rate was controlled by mass transfer iron containing sulfur and oxygen by employing a in the liquid phase. They investigated the effect of modified Sieverts' technique with a highly sensitive Al, Cb, Cr, Ni, 0, S and W on the rate of the nitro- pressure transducer. They conclusively demonstrated

* Manuscript received on April 16, 1987; accepted in the final form on September 11, 1987. © 1987 ISIJ * * Formerly Carnegie-Mellon University . Now at Department of Metallurgy, Faculty of Engineering, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113. * * * Department of Metallurgical Engineering and Materials Science, Carnegie-Mellon University, Pittsburgh , PA 15213, U.S.A.

( 858 ) Research Article Transactions ISIJ, Vol. 27, 1987 (859) that the rate of nitrogen absorption into iron alloys the effect of S, P, Pb, Bi, Sn and Te on the nitrogen with surface active solute was second order with reaction with liquid Fe-Csat . alloys. The rate was respect to nitrogen concentration or first order with measured by using the isotope exchange technique, respect to nitrogen partial pressure and the rate was since the measured rate is not affected by liquid phase controlled by the dissociation of the nitrogen molecule mass transfer or the evolution of other gases from the on the surface. melt. In the later work, Fruehan and Martonik20~ mea- sured the effect of sulfur on the nitrogen reaction with Fe-Cr and Fe-Cr-Ni alloys. They found that chro- II. Experimental mium increases the rate. At low levels of sulfur con- As described in the papers of Glaws and Frue- tent absorption rate was controlled by liquid phase han,22,23~it is possible to employ both the isotope mass transfer of nitrogen, but at high sulfur levels the exchange and the Sieverts' technique in a single rate was controlled by mass transfer and the chemical apparatus. As Ishii et al.24~have pointed out, since rate on the surface in series. A correction was made CO gas is evolved from Fe-C alloys, the Sieverts' for the influence of liquid phase mass transfer on the technique was not suitable for the measurements in measured rates. These effects are negligible at high the present study. Therefore the isotope exchange sulfur concentrations, but become significant at inter- technique was employed to investigate the kinetics mediate sulfur contents. At low levels of sulfur this of the nitrogen reaction with liquid Fe-C-X solu- method of determining the intrinsic chemical reaction tions, where X is a solute element. The experimental rate constant, becomes very imprecise, since the mass apparatus and technique were previously described in transfer effects become significant. detail.23~ Metal samples weighing approximately 30 To measure the interfacial reaction rate and elimi- g were contained in a crucible which is high density nate the influence of liquid phase mass transfer, Byrne and high purity (99.8 %) alumina with 20 mm I.D., and Belton21 employed the isotope exchange tech- 22 mm O.D, and 33 mm height. Alumina crucible nique in their study of the kinetics of nitrogen reac- was set in the outer safety crucible which was slip cast tion with liquid Fe and Fe-S alloys. This technique lime-stabilized zirconia with approximately 25 mm permits a direct experimental determination of the I.D. and 50 mm height. The crucible was placed in rate of N2 dissociation on the liquid metal surface via a fused silica reaction chamber and carefully induc- measurement of the rate of exchange reaction tively heated to the experimental temperature. The 28N2+30N2 = 229N2 experimental temperature was mainly 1 450°C, and ...... (1) some experiments for Fe-C-S alloys were also con- The experiment is run under the condition of equi- ducted at 1 600°C. The sample temperature was librium between the gas at the interface and liquid monitored by a two-color ratio pyrometer, sighted bulk phase. This permits the measurement of inter- onto the metal surface. The pyrometer was calibrated facial chemical reaction rates that would normally be against the melting temperature of pure iron. Any considered too fast for the conventional kinetic ex- necessary temperature adjustments were made man- perimental technique for inductively stirred melts. ually via the power controls on the R.F. generator. Byrne and Belton21 found the exchange rate to be The desired metal surface temperatures were main- first order with respect to nitrogen partial pressure on tained within +5°C. high purity iron and the rate for Fe-S indicated that During heat up, a N2-H2 gas mixture was passed sulfur closely followed an ideal chemisorption model. over the sample at a combined flow rate of approxi- Recently Glaws and Fruehan22~ used both the mod- mately 200 ml/min STP. was used to ified Sieverts' technique and an improved isotope ex- suppress the effects of the few ppm of oxygen in the change technique in the same experimental apparatus isotopically labeled gas, and to reduce the sulfur con- to verify that N2 dissociation is the rate limiting ele- tent in metal which affects on the rate. In the case mentary step in the intrinsic chemical reaction mech- of experiments of Fe-C-S alloys, a N2-H2-H2S gas anism. This provided strong evidence that the dis- mixture was passed over the sample to maintain the sociation of the N2 molecule is the rate limiting step desired sulfur activity in the sample. The H2 or H2- in intrinsic chemical reaction mechanism of nitrogen H2S gas is approximately 15 % of total gas mixture. absorption into liquid Fe-S alloys. More recently Once the liquid metal temperature was reached, the Glaws and Fruehan23~ investigated the kinetics of the standard nitrogen gas was replaced by the labeled nitrogen reaction with liquid Fe-Cr alloys using the nitrogen gas, containing approximately 1 at% of the isotope exchange technique. They found that chro- isotope 30N2. The isotopically labeled nitrogen gas mium greatly enhances the rate of nitrogen dissocia- was jetted onto the surface of the liquid metal for a tion and absorption into liquid iron solutions. The sufficient time to allow for equilibrium to be achieved. mechanism is due to N2 dissociation being faster on Continuous gas analysis was possible through the use chromium surface sites than on iron sites. Due to of an in-line Dycor Precision Mass Analyzer. Peaks the faster nitrogen reaction rates encountered in liquid of mass numbers of 28, 29 and 30 amu were moni- Fe-Cr alloys, the influence of gas phase mass trans- tored continuously during each isotope exchange ex- port was taken into account in the determination of periment and over fifty 30N2/28N2ratio measurements the interfacial rate constant. were taken for each gas flow rate in order to insure It is the purpose of the present work to determine that accurate values for the exchange reaction rate

Research Article ( 860) Transactions ISIJ, Vol. 27, 1987 constant were obtained. The flow rate was 200, 250, 30F0: the initial fractions of 30N2 molecules 30F 300, 360 and 450 ml,lmin. Continuous monitoring of eq : the probability of forming a molecule of the gas analysis allowed for quick recognition of the 30N2, which is equal to the fraction of stabilization of the 30N2I28N2ratio for each new flow 30N2 molecules at isotope exchange equi- rate, thus affecting a minimal waste of the expensive librium. concentrated 30N2isotope gas. However, the influence of gas phase mass transfer on To prepare sample alloys, electrolytic iron was pre- the measured exchange rate was not considered in this melted in a graphite crucible and added each element rate equation. Glaws and Fruehan12~ showed that and mixed at experimental temperature. Samples of although the nitrogen reaction rate constants for sev- the original alloys were analyzed before and after each eral Fe-Cr-S alloys are fairly invariant with changes element for the relevant elements. Sulfur was ana- in gas flow rate, a correction for gas phase mass trans- lyzed by combustion method and phosphorus was fer effects on the observed exchange rate still may be determined by molybdenum blue photometric meth- necessary. The correction is small for slow exchange od. Sn, Pb, Bi and Te analysis was conducted by rates, but becomes important for the fast interfacial using atomic absorption spectrometry. reaction rates that are commonly encountered in In this study the effect of S, P, Sn, Pb, Bi and Te liquid iron alloys containing low concentrations of on the rate constant of nitrogen reaction with Fe- solute elements. Glaws and Fruehan23~ derived fol- Csat, alloys was investigated. In the case of lead lowing relationship and bismuth experiment, the lead or bismuth content 1 in metal would change because of vaporization of k= (4) 1 1 these metals. Therefore Ag-Bi or Ag-Pb alloys were + contained in the crucible with the Fe-C alloys kg ks to maintain a constant potential of these elements. where kg is given by Although the bismuth and lead contents in silver de- creased during experiments, the bismuth and lead k m g R T (5) contents in Fe`Csat, alloys did not change significant- ly. In the case of the tellurium experiment, since and m is the gas phase mass transfer coefficient for a the vapor pressure of tellurium is much higher than particular experimental geometry. Therefore, the in- that of lead and bismuth, almost all of tellurium in terfacial rate constant (Ice) can be determined from Fe -Csat. alloys evaporate during a long experiment. the measured exchange rate constant (k) and the rate Therefore, the measurement was started as soon as constant for gas phase mass transfer (kg) which is cal- possible after metal alloy melted and the measure- culated using the experimentally derived value for ment was conducted for only one flow rate for each m/ T for the specific experimental conditions. The experiment. The change of tellurium content in experimentally derived value of m/ T for the present metal with time was measured and the tellurium geometry was determined from the rate of oxidation content at the beginning of measurement was esti- of graphite in CO-CO2 gas mixtures under conditions mated. No significant phosphorus and tin losses were that the rate is controlled by gas phase mass trans- encountered. fer.26~ The flow rate of gas mixture was varied be- tween 250 and 600 ml/min. The value of m/T con- III. Rate Equation for the Isotope Exchange verted to nitrogen gas is approximately 2.2 X 10-3 cm/s K26)and m is 3.8 cm/s at 1 450°C. Reaction Surface active elements such as sulfur and oxygen Several authors21,25~ have derived rate equations retard the reaction on the surface. As shown by for the nitrogen exchange reaction. Byrne and Bel- Darken and Turkdogan27~ the apparent interfacial ton21~ derived the following rate equation rate constant, k, at high degree of surface coverage by V d 30F the surface active element is given by _ kA(30' _30 RT dt q ) ...... (2) k = kbO jai ...... (6) and integrated Eq. (2) to obtain exchange rate con- where, kb : the rate constant for the pure iron stant, k, (Eq. (3)), from their experimental data, ci : a constant 30F _ 30F ai : the activity of the surface active element. Q eq -k= In Belton28>derived the following equation assuming an ART 30`0 `30F (3) eq ideal Langmuir adsorption where, V a volume of reacting gas k = kbl(1 + Kiai) ...... (7) Q the volumetric flow rate of the com- bined gas (nitrogen+inerts) at the tem- where, Ki : the equilibrium adsorption coefficient. perature T Under conditions of a high concentration of a strong- A the interfacial area ly adsorbed solute, where Kiai>> 1, the interfacial rate R the gas constant constant is inversely proportional to the activity of 30F the fractions of 30N2 molecules in the the solute, consistent with Darken and Turkdogan's gas after reaction relation (Eq. (6)).

Research Article Transactions ISIJ, Vol. 27, 1987 (861)

Iv. Results measured by Byrne and Belton21) are also shown in All isotope exchange results have been corrected Fig. 2 for high purity iron. From the slope of line for gas phase mass transfer effects according to the in Fig. 2, the activation energy of nitrogen reaction method previously outlined.23) Figure 1 shows the was calculated to be 27 kcal/mol. This activation en- variation of rate constant, k5, in mol/cm2ms. atm with ergy is in good agreement with the activation energy activity of sulfur at 1 450 and 1 600°C for Fe-Csat. estimated by Byrne and Belton.21) Furthermore, the alloys. The present results at 1 600°C agree with sulfur content in Fe alloys does not have significant the results with Fe-S alloys obtained by Glaws and influence on the activation energy of nitrogen reac- Fruehan.22) The rate constants for the exchange tion. reaction increase linearly as the sulfur activity de- In Fig. 3 the effect of carbon on the rate constant creases, as previously shown with the Fe-S alloys.22) was shown for the metal in which the activity of sulfur As the metal is saturated with carbon, activity coeffi- was 0.11 with respect to 1 wt% as standard state at cient of sulfur of 6.329) with respect to 1 wt% as 1 600°C and compared with the results of Glaws and standard state was used for the calculation of activity Fruehan.22) The rate constant is independent of of sulfur. carbon concentration up to 5.4 % C which is the The rate constants shown in Fig. 1 for the sulfur solubility limit at 1 600°C. activities of 0.06 and 0.028 are plotted against the In Fig. 4, the effect of phosphorus on the inter- reciprocal of the temperature in Fig. 2. The results facial rate constant at 1 450°C is shown. The rate constant is 1.0 X 10-6 mol/cm2ms. atm at 0.03 wt% P, and decreases linearly with increasing phosphorus con- tent and then at 0.16 wt% P the rate constant be- comes 0.5 X 10-6 mol/cm2. s . atm. At high phosphorus concentration the rate constant becomes independent of phosphorus content indicating the presence of the residual rate. Figure 5 shows the effect of tin con-

Fig. 1. Nitrogen reaction rate constants in Fe-Csat.-S alloys at 1 450 and 1600°C.

Fig. 3. Effect of carbon in Fe-C-S alloys on the rate con- stant of nitrogen reaction at 1 600°C and the con- stant sulfur activity.

Fig. 2. Temperature dependence of the rate constant of Fig. 4. Effect of the phosphorus in Fe-Csat ,-P alloys on the nitrogen reaction with Fe-Csat,-S alloys. rate constant of the nitrogen reaction at 1450°C.

Research Article ( 862 ) Transactions ISIJ, Vol. 27, 1987

tent in Fe-Csat.-Sn alloys on the rate constant at estimated lead content in Fe-Csatt alloys becomes 1 450°C. Although the content of tin in metal is 0.07 wt%. The observed value of 0.078 wt% is very high such as 0.8 %, the change of rate constant therefore a reasonable value. is not significant. Very small amounts of bismuth has a strong in- The effect of lead on the rate constant at 1450°C, fluence on the rate constant with 56 ppm Bi reducing shown in Fig. 6, indicates a linear relation with the the rate constant to 4.4x 10-7 mol/cm2. s •atm. Figure inverse of the activity of lead, with pure lead as the 7 shows the effect of activity of bismuth, with pure standard state, at low lead concentrations in Fe- bismuth as standard state, on the rate constant. The Csat, alloys. The activity of lead was calculated from rate constant varies linearly with the inverse of the the lead content in Ag-Pb alloys using the thermo- activity of bismuth and there is no evidence of residual dynamic data of Ag-Pb alloys.30~ However at high rate. At present, it is not clear whether the behavior activities of lead, the rate constant deviates from the of the residual rate constant is present at higher bis- linear relationship. Small amounts of lead have a muth contents or not, because of the difficulty of remarkable effect on the rate constant. For example, running the experiment at high bismuth concentra- addition of 0.014 wt% Pb decreases the rate constant tions in Fe-Csat, alloys due to the vaporization of the by nearly an order of magnitude to 0.4x 10-6 mol/ bismuth. cm2.s.atm. In the case where the activity of lead is In the case of the Fe-Csat.-Te experiments, much unity, the liquid Fe-C alloy was in equilibrium with of the tellurium vaporized during the experiments at pure lead instead of an Ag-Pb alloy. The measured 1450°C; therefore the change of tellurium content in concentration of lead in the Fe alloy, 0.078 wt% Pb, the Fe-C alloy, during an experiment, was inves- is the solubility limit of lead in the Fe-Csat, alloy. tigated. Figure 8 shows the change of tellurium The solubility of lead in iron is 0.24 wt% at content with time; at about 50 min the metal begins 1550°C.31) Since the interaction coefficient of car- melting and the deposition of metal vapor was ob- bon on lead is 0.066 at 1 600°C,32) if it is assumed that this value is valid to high carbon contents, the

Fig. 7. Effect of the activity of bismuth in Fe-Csat .-Bi alloys on the rate constant of the nitrogen reaction Fig. 5. Effect of the tin in Fe-Csat .-Sn alloys on the rate at 1450°C. constant of the nitrogen reaction at 1 450°C.

Fig. 6. Effect of the activity of lead in Fe-Csat .-Pb alloys on the rate constant of the nitrogen reaction at Fig. 8. Change of tell urium content with time during an 1450°C. experiment at 1450°C.

Research Article Transactions ISIJ, Vol. 27, 1987 (863)

same value of gas phase mass transfer coefficient previously outlined is used to correct this extrapolated value for gas phase mass transfer, since the geometry and conditions of Byrne and Belton's experiments are similar to those of present study. The corrected rate constant is 2.4x 10-6 mol/cm2, s . atm, which is in good agreement with the observed rate constant in the present study. Glaws and Fruehan22~ found that the influence of carbon on the rate constant is negligible for a con- stant sulfur activity. This result is consistent with Byrne and Belton's measurements21 for high purity iron (sulfur content is 9 ppm) and Fe-C (carbon is 2.0 and 4.3 %) alloys at 1 650°C. In this study the

Fig. 9. Effect of the tellurium in Fe-Csat .-Te alloys on the rate constant was measured at 5.4 %, which is the rate constant of the nitrogen reaction at 1 450°C. solubility limit of carbon at 1600°C, and the results show that at high carbon content the effect of carbon on rate constant is still negligible as indicated in Fig. served on the wall of cooling chamber. Therefore 3. As Glaws and Fruehan have pointed out,22~this tellurium content is decreasing rapidly. Usually the phenomenon is consistent with the surface tension experiments were conducted between 90 and 120 min data of Kozakevitch et a1.33~who found that the sur- and the tellurium at the beginning of measurement face tension of liquid Fe-C-S solutions was a function was less than 0.02 wt%. Small amounts of tellurium of sulfur activity only; however, it is necessary to take have a great effect on the rate constant with 120 ppm into account the effect of carbon on the activity of Te reducing the rate constant to 2.8x 10-7 mol/cm2. s sulfur. atm. Figure 9 shows the effect of tellurium in Fe- Consideration of the relationship between nitrogen C-Te alloys on the rate constant at 1 450°C. The reaction rate data and reciprocal activity of solute tellurium content used in Fig. 9 is the final content elements may provide insight into whether solute ele- after the experiment. The rate constant is a linear ment affects the kinetics through an influence on the function of the inverse of the tellurium content at low adsorption behavior of solute element in liquid Fe-C tellurium contents. However, at high tellurium con- solutions. As discussed previously, at high activities tents, the rate constants deviate from the linear rela- of surface active elements the rate constant should tionship and the behavior of a residual rate constant approach zero. However, from the more accurate is observed. results of the modified Sieverts' experiments at high solute concentration, the presence of a residual rate at V. Discussion high sulfur contents was clearly indicated by Glaws Byrne and Belton21 summarized previous re- and Fruehan.22~ Studies on the decarburization of searches for the effect of sulfur on the rate of the liquid Fe-C-S alloys by Samn and Belton,34~Lee and nitrogen reaction and have corrected the rate con- Rao,35~and Hayer and Whiteway36~ provide evidence stant for the effect of liquid phase mass transfer. of the presence of a residual rate for the CO2 reaction Their results show that the corrected rate constant is on a liquid Fe-C surface at high sulfur levels. In proportional to the inverse of the sulfur content in the present study, a residual rate is found for Fe-C- the Fe alloy. Assuming a Langmuir adsorption Te, Fe-C-Pb and Fe-C-P alloys. On the other hand, isotherm for sulfur, they derived the relationship be- the residual rate is not found for Fe-C-Bi alloys. The tween the rate constant of nitrogen reaction and the residual rate constant of tellurium and lead are about activity of sulfur at 1 550 and 1 600°C as follows : 0.3 x 10-6 mol/cm2. s. atm. This rate constant is about 3 % of the rate constant for pure iron obtained by ka l.7 x 10-5/(1 + 130a5) mol/cm2•s.atm Byrne and Belton.21> The residual rate constant for at 16000C21) ...... (8) Fe-C-P alloys was slightly higher at 0.5 x 10-6 mol/ ka = l.4 x 10-5/(1+ 140as) mol/cm2m s • atm cm2 • s • atm. at 15500C21> ...... (9) Mowers and Pehlke8~ in their experiments with Fe-Te alloys showed no indication of residual rate. where, as : the activity of sulfur with 1 wt% as the The gas volume measurement system that they used standard state. was not as sensitive as the technique employed in pres- For example, the present result shows that the rate ent work and may not have been sensitive enough to constant is 2.3 x 10-6 mol/cm2ms . atm at the sulfur con- detect a slow residual rate. Tellurium has a stronger tent of 0.0030 % in Fe--Csat. alloys at 1450°C. The effect on the rate constant of the nitrogen reaction rate constant which was obtained by extrapolating than sulfur which is in the same group VIa as tel- Eqs. (8) and (9) to 1 450°C and using the activity lurium. This agrees with the results of Mowers and coefficient of sulfur in Fe-Csat, alloys is 2.2 x 10-6 mol/ Pehlke.8j This behavior is consistent with tellurium cm2. s . atm. However, the influence of gas phase mass being more surface active than sulfur in liquid iron. transfer was not considered in this rate constant. The There has been no work on the effect of P, Sn, Bi

Research Article ( 864) Transactions ISIJ, Vol. 27, 1987 and Pb. The present results indicate these elements In addition, the assistance and helpful discussions with decrease the rate constant of the nitrogen reaction; Dr. P. C. Glaws gratefully acknowledged. possibly indicating they are surface active and reduce the surface tension of carbon saturated iron melts. REFERENCES For the Fe-C-Pb alloys an Ag-Pb alloy was used to maintain the lead potential in the system; silver 1) R. D. Pehlke and J. F. Elliott: Trans. Metall. Soc. AIMS, has a small solubility in iron.37~ However, since the 218 (1960), 1088. solubility data in Fe-Csat 2) R. D. Pehlke and J. F. Elliott: Trans. Metall. Soc. AIMS, . alloys is not available an 227 (1963), 844. experiment was conducted to estimate the effect of 3) T. Choh and M. Inouye: Tetsu-to-Hagane, 53 (1967), 1393. silver using silver with a Fe alloy instead of an Ag- 4) M. Inouye and T. Choh: Trans. Iron Steel Inst. Jpn., 8 Pb alloy. The rate constant obtained, 1.0 x 10-6 molf (1968), 134. cm2• s •atm, at high silver activity is nearly equal to 5) T. Choh and M. Inouye: Tetsu-to-Hagane, 54 (1968), 19. the rate constant obtained at very low activity of 6) T. Choh, M. Okamura and M. Inouye: Tetsu-to-Hagane, solute elements such as Te, Pb and Bi. Therefore it 55 (1969), 1177. can be concluded that the effect of silver on the rate 7) K. Mori and K. Suzuki: Trans. Iron Steel Inst. Jpn., 10 constant is negligible. The activity of lead for the (1970), 232. Fe-C-Pb alloys is determined by the lead content in 8) R. G. Mowers and R. D. Pehlke: Metall. Trans., 1 (1970), Ag-Pb alloys. The thermodynamic data30~for Ag- 51. Pb alloys is available at 1 000°C, this data was ex- 9) K. Narita, S. Koyama and T. Makino: Tetsu-to-Hagane, 57 (1971), 2207. trapolated to 1450°C. Since the lead content in 10) T. Choh, F. Kuze and M. Inouye: Tetsu-to-Hagane, 59 Fe-Csat , alloys was analyzed, plotting the relation (1973), 372. between the activity of lead in Ag-Pb alloys and the 11) T. Choh, T. Takabe and M. Inouye : Tetsu-to-Hagane, 59 lead content in Fe-Csat, alloys, the activity coefficient (1973), 1914. of lead in Fe--Csat,alloys can be obtained; that value 12) S. Ban-ya, T. Shinohara, H. Tozaki and T. Fuwa: Tetsu- is 2.7 x 103. It is not confirmed whether the system to-Hagane, 60 (1974), 1443. is in equilibrium or not and thus it should be em- 13) T. Choh, T. Yamada and M. Inouye: Tetsu-to-Hagane, 62 phasized that above calculation is only an estimation. (1976), 334. In the same manner as the lead experiment, Ag- 14) K. Amano, K. Ito and H. Sakao : Tetsu-to-Hagane, 62 Bi alloys were used to maintain a constant bismuth (1976), 1179. activity. Therefore, activity of bismuth was calcu- 15) T. Choh, T. Moritani and M. Inouye: Tetsu-to-Hagane, 64 lated from the bismuth content in Ag-Bi alloys using (1978), 701. 16) T. Choh, T. Takabe and M. Inouye: Tetsu-to-Hagane, 67 the thermodynamic data for the Ag-Bi system.30) (1981), 2665. The calculated activity coefficient of bismuth in Fe- 17) K. Kadoguchi, M. Sano and K. Mori: Tetsu-to-Hagane, Csat, alloys is 2.4x 104; again it should be emphasized 71 (1985), 70. this is only an estimation. 18) M. Kawakami, K. Ito, M. Okuyama, T. Kikuchi and S. Sakase : Tetsu-to-Hagane, 73 (1987), 661. VI. Summary and Conclusions 19) R. J. Fruehan and L. J. Martonik: Metall Trans. B, 11B The isotope exchange technique was employed to (1980), 615. investigate the kinetics of nitrogen reaction with liquid 20) R. J. Fruehan and L. J. Martonik: Metall. Trans. B, 12B Fe-C alloys and the following results are obtained. (1981), 379. (1) The addition of small amounts of P, Pb, Bi 21) M. Byrne and G. R. Belton : Metall. Trans. B, 14B (1983), and Te to liquid Fe-Csat. solution decreases the rate 441. constant of the nitrogen reaction. However, the 22) P. C. Glaws and R.J. Fruehan: Metall. Trans. B, 16B (1985), 551. effect of tin content on the rate constant of nitrogen 23) P. C. Glaws and R. J. Fruehan: Metall. Trans. B, 17B is small. (1986), 317. (2) The effect of carbon on the rate constant for 24) F. Ishii, S. Ban-ya and T. Fuwa: Tetsu-to-Hagane, 68 a constant bulk sulfur activity is negligible in Fe-C-S (1982), 1551. alloys. 25) A. Ozaki: Isotope Studies of Heterogeneous Catalysis, (3) The presence of a residual reaction rate is ob- Kodansha Ltd, and Academic Press, Tokyo, (1977). served for Fe-C-Pb, Fe-C-P and Fe-C-Te alloys, but 26) P. C. Glaws: Ph.D. Thesis to Carnegie-Mellon University, not for Fe-C-Bi solutions over the composition range Pittsburgh, (1985), 66. investigated. 27) L. S. Darken and E. T. Turkdogan: Heterogeneous Ki- (4) The rate constant of nitrogen reaction for Fe- netics at Elevated Temperatures, ed. by G. R. Belton and C-S alloys are proportional to the inverse of sulfur W. L. Worrell, Plenum Press, New York, (1970), 25. activity at low sulfur activity and the rate constants 28) G. R. Belton: Metall. Trans. B, 7B (1976), 35. agree with previous measurements. 29) S. Ban-ya and J. Chipman: Trans. Metall. Soc. RIME, 245 (1969), 133 & 391. Acknowledgements 30) R. Hultgren, P. D. Desai, D. T. Hawkins, M. Gleiser and K. K. Kelly: Selected Values of The Thermodynamic The authors wish to thank The Center for Iron and Values of Binary Alloys, Am. Soc. Met., Metals Park, Steelmaking Research, Pittsburgh, PA, U.S.A., their (1973), 32-35 & 79-83. member companies and National Science Foundation 31) A. E. Lord and N. A. Parlee: Trans. Metall. Soc. RIME, (Grant CPE-8421112) for the support of this research. 218 (1960), 644.

Research Article Transactions ISIT, Vol. 27, 1987 ( 865)

32) G. K. Sigworth and J. F. Elliott: Met. Sci., 8 (1974), 298. 403. 33) P.. Kozakevitch, S. Chatel, G. Urbain and M. Sage: Rev. 36) M. Hayer and S. G. Whiteway: Can. Metall. Quart., 12 Metall., 52 (1955), 134. (1973), 35. 34) D. R. Samn and G. R. Belton: Metall. Trans. B, 9B (1978), 37) M. Hansen : Constitution of Binary Alloys, McGraw Hill 403. Book Co. Inc., New York, (1958), 20-21. 35) H. G. Lee and Y. K. Rao : Metall. Trans. B, 13B (1982),

Research Article