Study of the Aluminothermic Reduction of Niobium Pentoxide through Thermal Analysis Experiments and High Energy Milling Processing

Aluminothermic reduction of niobium pentoxide was studied through thermal analysis techniques such as differential thermal analysis (DTA) and thermogravimetry (TG) as well as through high energy milling processing. Reactants mixtures were composed by powders of Nb 2 O 5 and Al. In the case of DTA-TG experiments, different molar ratios Nb 2 O 5 :Al were heated in a dynamic atmosphere of synthetic air under controlled conditions. The high energy milling runs were carried out via SPEX vibratory mill under argon atmosphere and with milling power equal to 7:1 (ratio of mass of balls to mass of mixture) with 10 pct excess of Al over the stoichiometric mass of aluminum necessary. In both kinds of experiments, X ray diffraction was used in order to identify the products of reaction. From DTA-TG experiments, it was possible to determine the experimental value of the enthalpy change (–595.9 kJ.mol), which is near to the theoretical one. From the milling experiments, it was possible to verify the possibility of the occurance of aluminothermic reducion of niobium pentoxide via this kind of processing.


Introduction
Metallothermic reactions are very important for many refractory metals producing processes.The heat of reaction and the rate of reaction are essential for the understanding of the metallurgical operations involved.Since the aluminothermic reduction of metal oxides is both highly exothermic and rapid reactions, there are few attempts to study the kinetics of such a process [1][2][3] .
The aluminothermic reduction of Cr 2 O 3 4 , Fe 2 O 3 5,6 , MnO 2 6,7 and V 2 O 5   7   , have been studied using the simultaneous differential thermal analysis (DTA) -thermogravimetry (TG) techniques.In a DTA apparatus the heat losses cannot be accurately predicted, therefore it is impossible to estimate precisely the maximum temperature attainable by aluminothermic reacting mixtures in a DTA set-up.Therefore an excess of aluminium has been used in order to use the excess enthalpy [4][5][6][7] .
In terms of high energy milling, this is a process used for producing powders having a fine microstructural scale and/or as a technique for alloying difficult or normally incompatible materials.The capability of high energy milling process to produce powders with unique microstructures has been clearly established and this process is also inherently flexible [8][9][10] .
In high energy milling, a suitable powder charge (typically, a blend of elemental or preallyoed powders) is placed in a high-energy mill, along with a suitable grinding medium (typically, hardened steel balls).From a macroscopic viewpoint, the resultant powder develops through the repeated cold working and fracture of the powder particles with a final composition corresponding to the percentages of the respective constituents in the initial charge.This technique has been utilized to producing a wide range of materials including amorphous materials, intermetallic compounds, and solid solution alloys [9][10][11][12] .
In the case of a vibratory mill like SPEX, it is possible to produce small quantities of mechanically alloyed powder in relatively short times.In this system, the ball and powder charges are placed in a small jar which is agitated at a high frequency in a complex cycle which involves motion in three orthogonal directions [8][9][10] .
In the present work, aluminothermic reduction of niobium pentoxide was studied through thermal analysis techniques such as differential thermal analysis (DTA) and thermogravimetry (TG) as well as through high energy milling processing.Reactants mixtures were composed by powders of Nb 2 O 5 and Al.In the case of DTA-TG experiments, different molar ratios Nb 2 O 5 :Al were heated in a dynamic atmosphere of synthetic air under controlled conditions.The high energy milling runs were carried out via SPEX vibratory mill under argon atmosphere and with milling power equal to 7:1 (ratio of mass of balls to mass of mixture) with 10 pct excess of Al over the stoichiometric mass of aluminum necessary.In both kinds of experiments, X ray diffraction was used in order to identify the products of reaction.
In each run a sample weighing 40 mg was put into alumina cylindrical crucible and heated from room temperature up to 1200 °C in a dynamic atmosphere of synthetic air (O 2 + N 2 = 99.997vol.(%), 100 mL/min).Alumina powder was used as reference material.The adopted heating rate was 10 °C/min.
X ray diffraction was used in order to identify the products of reduction.

�.�. High energy milling experiments
The experiments were carried out via vibratory high energy milling.A SPEX CertPrep 8000 Mixer/Mill was used for milling under argon atmosphere (99,999 vol.(%) purity).A tool steel jar and chromium steel balls were used.The milling power ratio was 7:1 (ratio of mass of balls to mass of mixture) and runs were carried out for periods of 45, 75 e 120 minutes.For each milling, the reactant mixture was prepared from powders of Nb 2 O 5 (98.80 wt.(%) purity, average size 24.75 micrometers) and Al (99.70 wt.(%) purity, average size 41.48 micrometers), with 10 pct excess of Al over the stoichiometric mass of aluminum necessary for the reduction of the niobium oxide to metallic niobium.The temperature of the jar was monitored by means of type K thermocouple, in connection with a data collector.
X ray diffraction was used in order to identify the products of reduction.

�.1. Thermal analysis experiments (DTA�TG)
Figure 1 shows typical DTA-TG curves generated.On the DTA curve one can see two peaks as follows: a) the endothermic peak related to the melting of aluminium; and b) the exothermic peak related to the overall enthalpy change of the system.On the other hand, on the TG curve one can see the mass gain related to the oxidation of part of the excess of aluminium of the mixture.According to Sarangi, Sarangi, Ray and Misra 5 the exothermic DTA peak represent the total heat effect due to the following factors: a) heat of the aluminothermic reaction (exothermic with no gain of mass); b) heat of oxidation of some excess of aluminium (exothermic with gain of mass); and c) dissolution of the reduced metal (niobium) in excess of aluminium and formation of intermetallic compounds (of the Al-Nb system) (exothermic with no gain of mass).
The heats of formation at 298 K of the intermetallic compounds of the Al-Nb system (Nb 2 Al, Nb 3 Al and NbAl 3 ) are very small in comparison with the heat of reaction of aluminothermic reduction of Nb 2 O 5 .They are equal respectively to 13 -29.8kJ.mol -1 , -19.7 kJ.mol -1 and -49.4 kJ.mol -1 while the heat of reduction of Nb 2 O 5 with Al (Nb 2 O 5 (s) + 10/3 Al(s,l) = 2 Nb(s) + 5/3 Al 2 O 3 (s)) is -890.0kJ.mol -1 , at 25 °C (298 K) 14 .Therefore, the experimental heat value of the aluminothermic reduction of Nb 2 O 5 was obtained by subtracting the heat of oxidation of some excess of aluminium calculated from the gain of mass detected on the TG curve from the overall enthalpy change calculated from the DTA peak.The results of the experiments are summarised in Table 1 which contains: the molar ratio used in each experiment, mass of mixture, the DTA temperature of the exothermic peak, the calculated overall enthalpy change of the system, the mass gain, the enthalpy of oxidation of the aluminium and, finally, the experimentally determined enthalpy change associated to the reduction of Nb 2 O 5 with Al.
Figure 2 shows, for each tested molar ratio Nb 2 O 5 :Al, both the theoretical and the experimental values of the enthalpy change of the reduction, at the respective DTA peak temperature.The calculated values of the enthalpy of reduction presented on Figure 2 have been calculated from the sum of the change of enthalpy corresponding to the reduction at 298 K plus variation of the heat capacities of the products of reaction from 298 K up to the corresponding peak temperatures.
It must be noted that for molar ratios Nb 2 O 5 :Al equal or greater than 1:11 the experimental value of the enthalpy change is almost constant and very near to the theoretical value (approximately 6% less than the theoretical value).
Figure 3 shows the X ray diffraction pattern of the product reaction for the molar ratio Nb 2 O 5 :Al=1:11 which indicates the presence of   The presence of the intermetallic compound NbAl 3 is due to the fact that there is aluminium in excess over the stoichiometric quantity (3.3 more than the stoichiometric quantity).So, the reduced niobium reacts with the excess of aluminium leading to the formation of NbAl 3 .Sarangi et al. 7 have observed in the case of the reduction of V 2 O 5 and MnO 2 with Al the formation of V-Al and Mn-Al intermetallic compounds, respectively.Cintho et al. 8 have observed in the case of the reduction of Cr 2 O 3 with Al the formation of Cr-Al intermetallic compound.

�.�. High energy milling experiments
Figure 4 shows the X ray diffraction pattern of the mixture homogeneized during 1 hour (without milling).Figures 5, 6 and 7 show X ray diffraction patterns of the mixtures milled during 45, 75 and 120 minutes, respectively.
It must be noted in Figure 4 (mixture only homogeneized -not milled) and in Figure 5 (mixture milled for 45 minutes) that the peaks of the diffraction correspond only to the presence of the reactants Nb 2 O 5 and Al.However, in the diffraction pattern related to the mix-    Figure 8 shows the temperature evolution of the jar during the 120 minutes milling run.
From Figure 8, it is possible to verify an abrupt increase of the temperature and the existance of a characteristic milling time related to the beginning of reaction (about 75 minutes).

Conclusion
The following conclusions have been obtained from the present study: • For molar ratios Nb 2 O 5 :Al equal or greater than 1:11 the experimental value of the enthalpy change is almost constant (-595.9kJ.mol -1 ) and very near to the theoretical value (approximately 6% less than the theoretical value).The analysis of the product of reaction for the molar ratio Nb

Figure 1 .
Figure 1.DTA and TG curves for the mixture with molar ratio Nb 2 O 5 :Al=1:11.A (DTA curve): endothermic peak related to the melting of aluminium; B (DTA curve): exothermic peak related to the aluminothermic reduction of Nb 2 O 5 plus the oxidation of part of the aluminium in excess plus the formation of an intermetallic compound of Al-Nb system (NbAl 3 ); and C (TG curve): mass gain related to the oxidation of part of the aluminium in excess.

Figure 2 .
Figure 2. Theoretical and experimental values of the enthalpy change (kJ.mol -1 ) of the reduction for each tested molar ratio (Nb 2 O 5 :Al), at the respective DTA peak temperature.

Figure 3 .Figure 4 .
Figure 3. XRD pattern of reduction of Nb 2 O 5 with Al (molar ratio Nb 2 O 5 : Al=1:11).Anode: Cu, Generator tension: 40 kV, Generator current: 40 mA, Kα 1 = 1.54060Å. Phases were identified according to ICDD -International Centre for Diffraction Data.Position (Angle (2 )) 10 20 30 40 50 60 70 80 2 O 5 :Al=1:11 indicates the presence of NbAl 3 , Al 2 O 3 and unreacted Al and Nb 2 O 5 .• X ray diffraction patterns of the mixtures milled allowed to verify the possibility of the occurance of aluminothermic reduction of niobium pentoxide via high energy milling processing.For 120 minutes milling time, the reaction products were Al 2 O 3 and intermetallic compound Nb 3 Al, with the presence of some residual Al.

Figure 8 .
Figure 8. Temperature profile of the jar during milling.

Table 1 .
Consolidated results of the experiments.