稀有金属(英文版) 2015,34(01),28-33

Decomposition behavior of titanium hydride treated by surface oxidation

School of Materials and Metallurgy, Northeastern University

Engineering Technology Research Centre of Ministry of Education for Materials Advanced Preparation

摘 要：

Titanium hydride attracts more attention as foaming agent in the fabrication of cellular metal materials.In order to meet most aluminum casting alloy's melting properties, the heat treatment processes for Ti H2 particles were investigated in a rotating device. In the present work,the most two important dynamic parameters, the treating temperature and oxidation interval, were taken under consideration. The decomposition behavior of titanium hydride was measured by differential scanning calorimetry(DSC) and the residual hydrogen content, morphologies and phase conversion were also characterized by hydrogen determinator, scanning electron microscopy(SEM), and X-ray diffractometer(XRD), respectively. The results show that the effect of temperature on the formation of oxidation film and decomposition behavior of TiH2 is more significant than that of oxidation time. The onset temperature and peak value of TiH2 decomposition shift from left to right through elevating temperature and extending time.Heat treatment process for TiH2 at 500 °C between 1 and5 h in air is favorable for preparing aluminum foam.

收稿日期：27 April 2014

基金：financially supported by the National Natural Science Foundation of China (No. 51174060);the Science and Technology Department of Liaoning Province of China (No. 2013223004);

Decomposition behavior of titanium hydride treated by surface oxidation

Hong-Jie Luo Hao Lin Pei-Hong Chen Guang-Chun Yao

School of Materials and Metallurgy, Northeastern University

Engineering Technology Research Centre of Ministry of Education for Materials Advanced Preparation

Abstract：

Titanium hydride attracts more attention as foaming agent in the fabrication of cellular metal materials.In order to meet most aluminum casting alloy's melting properties, the heat treatment processes for TiH2 particles were investigated in a rotating device. In the present work,the most two important dynamic parameters, the treating temperature and oxidation interval, were taken under consideration. The decomposition behavior of titanium hydride was measured by differential scanning calorimetry(DSC) and the residual hydrogen content, morphologies and phase conversion were also characterized by hydrogen determinator, scanning electron microscopy(SEM), and X-ray diffractometer(XRD), respectively. The results show that the effect of temperature on the formation of oxidation film and decomposition behavior of TiH2 is more significant than that of oxidation time. The onset temperature and peak value of TiH2 decomposition shift from left to right through elevating temperature and extending time.Heat treatment process for TiH2 at 500 °C between 1 and5 h in air is favorable for preparing aluminum foam.

Keyword：

Titanium hydride; Oxidation treatment; Thermal decomposition; Aluminum foam;

Author： Hao Lin,e-mail: mumuhao5@sina.com;

Received： 27 April 2014

1 Introduction

As a hydrogen storage material, titanium hydride (Ti H₂) isextensively used in producing metal foams [1, 2]. However, the original decomposition temperature of Ti H₂isaround 400 °C reported by the Refs. [3, 4] and thisdecomposition characteristic mismatches the meltingpoints of aluminum, and most of its alloys are used as thematrixes to fabricate aluminum foams. As the aluminumfoam is prepared by powder metallurgy, the gas releasefrom Ti H₂far below the melting point of aluminum matrixcan cause the precursor to expand in solid state and theirregular structure induced by crack-like pores may beformed [5]. Similarly, if aluminum foam is fabricated bymelt foaming, hydrogen generated in early stage will bepartially enwrapped in the aluminum melt, not onlyresulting in the non-uniformity of the melt but alsoincreasing its bulk viscosity. The dispersion of Ti H₂powder and subsequent foaming process are difficult to becontrolled. Owing to the difficulty for selecting moresuitable blowing agent to prepare aluminum foam, moreattention was paid on the surface modification of Ti H₂particles. Although the several methods, such as hydrolysisprecipitation, powder coating, and oxidation in differentatmosphere or with positive voltage, were suggested to beeffective in modifying the surfaces of the blowing agent(Ti H₂, Ca CO₃, Zr H₂) [6–9], many researchers gave thesuggestion on modifying the Ti H₂particles by oxidation inair [10–14]. This method was proved to be useful in theformation of thick oxide layer without large amount ofhydrogen lost. Nevertheless, owing to the treatment in fanoven or chamber furnace, the agglomeration of Ti H₂particles is not avoided, especially at high temperature. Theoxidation on the surface of Ti H₂in static state is uneasy toensure the particles coated by oxide layer evenly. Unlikethe previous methods, the treatment of Ti H₂in this studywas carried out in a rotating device, which can provide auniformly oxidizing process and gain parameters fortreating Ti H₂particles, especially for preparing aluminumfoam in volume quantity. The decomposition dynamics ofTi H₂before and after oxidation in air, the surface morphologies as well as the phase conversion were alsoinvestigated in this paper.

2 Experimental

Titanium hydride (purity 98.6 %, particle size \74 lm)was used in the oxidation process and thermal analysis.Ti H₂particles were pretreated under air inside a steel rollerin a horizontal tube furnace. The schematic diagram fortreating Ti H₂in air is shown in Fig. 1. The oxidationtreatment device consists of Ti H₂container, transmission,and heating systems. A stainless steel tube can draw theoutside air into the stainless steel roller, which contains theTi H₂particles, and make Ti H₂oxidize. In order to intensifygas exchange, the surface of the steel tube was drilled a lotof holes with 2 mm in diameter. An employment ofdeceleration chain not only slows down the rotatingvelocity but also let the Ti H₂particles rotate. To ensurethat the Ti H₂particles can be moved to the top position ofthe steel roller instead of staying at the bottom, threedeclining baffles were fixed on its inner surface. Althoughthe horizontal tube furnace can give a relatively uniformtemperature field, the length of constant temperature zoneis only 120 mm with temperature difference within±2.5 °C. In order to avoid the temperature error exceedingthe above value, the size of stainless steel roller wasdetermined to be 90 mm in length and 75 mm in diameter.50 g Ti H₂was filled in the steel roller and then was oxidized in the device. The steel roller was driven by a motorand rotated at speed of 50 rámin^-1. A heating rate of50 °Cámin^-1was employed to elevate temperature to theconstant value quickly and then record the heating time.The thermal decomposition of both as-received and oxidized Ti H₂particles was measured by the differentialscanning calorimetry (DSC, Netasch STA409). About10 mg Ti H₂particles were added in alumina crucible andheated at a heating rate of 10 Kámin^-1under continuousargon flow of 50 mlámin^-1. The hydrogen content of pretreated Ti H₂particles was detected by a hydrogen determinator (RH-404). The morphologies of Ti H₂particleswere characterized by scanning electron microscopy (SEM,Shimadzu SSX-550) coupled with energy dispersivespectroscopy (EDS). The phase components of Ti H₂particles were determined by X-ray diffractometer (XRD,Panalytical PW3040/60) coupled with X’Pert Pro MPDanalysis system.

Fig. 1 Schematic diagram of TiH 2pretreatment device. (1) decelerationchain; (2) small gear; (3) electromotor; (4) temperature controller; (5)support; (6) axletree; (7) thermocouple; (8) stainless steel roller; (9)heated wire; (10) furnace shell; (11) insulating brick; (12) stainless steeltube

3 Results

3.1 Non-isothermal oxidation experiments

As oxidation of Ti H₂mainly concerns two dynamicparameters, i.e., temperature and time, the Ti H₂particleswere first treated at different temperatures in air for constant time and then measured by DSC under argon atmosphere. Figure 2 reveals the decomposition behaviors ofTi H₂particles before and after treatment. For the untreatedTi H₂, two peaks can be seen in the curve, indicating thattwo stages occur in the decomposition process. Hydrogenrelease of the first peak starts at 464 °C and reaches amaximum at 507 °C; while the second peak position islocated at 533 °C corresponding to starting point of513 °C. The Ti H₂particles treated in air not only eliminatethe first peak but also show significantly changes in boththe onset of the hydrogen release and peak temperaturevalue. The onset temperature of gas release shifts from 483,497, 527, 548 to 615 °C with the oxidation temperaturesraising from 400, 450, 500, 550 to 600 °C, respectively,and the corresponding peak temperature up to 547, 566,606, 615, and 666 °C. The position of the remaining peakis obviously shifted to the higher temperatures comparedwith that of the second peak of the untreated Ti H₂. Thehydrogen content measured simultaneously shows thatoxidation in air leads to continuous hydrogen lost from0.3 wt%, 0.5 wt%, 0.7 wt%, 0.9 wt% to 1.6 wt%, respectively. The rapid elevation of onset and peak temperaturesaccompanies the violent release of hydrogen.

Fig. 2 DSC analysis of untreated and oxidation treatment Ti H2particles

3.2 Isothermal oxidation experiments

The subsequent isothermal oxidation experiments werecarried out at a fixed temperature with treating time varying to ensure proper dehydrogen property and decomposition temperature of pretreated Ti H₂particles, which couldmatch the melt points of aluminum casting alloys. Theanalysis results of DSC and residual hydrogen content areshown in Fig. 3, Tables 1 and 2, respectively. Figure 3shows that the Ti H₂particles treated at 400, 450, and500 °C for different intervals have the analogous decomposition behaviors which are same as that of Ti H₂particlesshown in Fig. 2, namely the onset temperature and peakvalue position are shifting from left to right. As seen inFig. 3a, the increase of onset and peak temperatures is veryslow. Only 1 °C difference can be identified when theoxidation time is prolonged from 20 to 25 h (Table 1).Accordingly, the release of hydrogen is gently, all valuesare C 3.1 wt% (Table 2). It can be found in Fig. 3b that thepeak temperature rises up to 606, 608, and 612 °C whenTi H₂is treated for 10, 15, and 20 h, respectively, while thehydrogen content is also above 3.1 wt%. But the peakvalue drops down to 598 °C as the Ti H₂is treated for 25 h.Although the peak temperatures all exceed 600 °C inFig. 3c, the residual hydrogen content comes below2.8 wt% when treated for over 10 h. Likewise, the peakvalue does not increase as the treating time is up to 20 h. Itcan be also seen both in non-isothermal and isothermalexperiments that treatment in air changes the shape ofdecomposition curves of Ti H₂. The height of decomposition peaks decreases with oxidizing temperature increasingor treating time extending.

Table 1 Onset temperature and peak value of Ti H₂treated at 400,450, and 500 °C for different time (°C)  下载原图

Table 1 Onset temperature and peak value of Ti H₂treated at 400,450, and 500 °C for different time (°C)

4 Discussion

Comparing the non-isothermal with isothermal experimentresults, the effect of oxidation temperature on Ti H₂particles is more significant than that of oxidation time. Thereseems to be a limit value for peak temperature along withthe extension of oxidation time. For most aluminum alloys,their melt points are around 600 °C, so the peak temperature above 600 °C can be assumed as an assessed value forTi H₂particles to be treated. As the oxidation of Ti H₂treated for 25 h at 400 °C is not enough to form an oxidelayer with high decomposition temperature of over 600 °C,it is infeasible for Ti H₂to be treated at this temperature.Oxidation interval up to 20 h may be an optimum value forthe Ti H₂particles oxidized at 450 °C if a longer treatingtime is out of consideration. It is preferred that the treatingtime of Ti H₂particles is from 1 to 5 h at 500 °C because ofthe maximum of peak temperature of up to 613 °C and theremaining hydrogen content of over 3.0 wt%. In the isothermal experiments, the oxidation process nearly levelsoff when the oxidation time reaches a certain value. Thereason for the phenomena may be attributed to the baffle ofoxidation film. Owing to that the density coefficient (theratio of the volume of oxidation film generated to thevolume of metal consumed by forming the oxide) of titanium oxide is 1.75, which is calculated by assuming thegenerated oxide corresponding to rutile phase, the oxidation film of titanium becomes compact and the oxidationfilm grows in terms of a parabola according to Wagneroxidation theory. Supposing the Ti H₂particle can be seenas spherical grain, these particles would consist of theoutside oxidation film and inside titanium hydride. After aperiod of time, the hydrogen generated on the titaniumhydride is hardly to penetrate outward through the oxidefilm for the generally thickened oxide film. Similarly, thediffusion of oxygen inward through the film is also limited.When the diffusion process reaches equilibrium, thethickness of oxidation film fluctuates around some valueand the peak temperature tends to be constant. However,the oxidation film will exhibit different structural properties under different temperatures.

Fig. 3 DSC analysis of Ti H2particles treated at different temperatures for 5, 10, 15, 20, and 25 h: a 400 °C, b 450 °C, c 500 °C

Table 2 Residual hydrogen content of Ti H₂treated at 400, 450, and500 °C for different time (wt%)  下载原图

Table 2 Residual hydrogen content of Ti H₂treated at 400, 450, and500 °C for different time (wt%)

Figure 4 shows the morphologies of Ti H₂particlesbefore and after treatment. The oxidation film around theTi H₂particle is clearly observed compared with the Ti H₂particle untreated in Fig. 4a. The film shown on the surfaceof Ti H₂particle in Fig. 4b is coarse and the crystals protrude upward. This film structure is inhomogeneous inthickness and many defects may be concealed in it,whereas the film shown in Fig. 4c is smooth and uniform,the protective effect of this compact film is obvious andimpels the peak temperature to 627 °C. It is plausible thatincreasing oxidation temperature moderately, such as500 °C, is favorable for obtaining an effective oxide layer,but the oxidation time should be controlled in a shorterrange. If the Ti H₂particles are treated at a relatively lowertemperature, such as 450 °C, prolonging the oxidationinterval excessively weakens the effectiveness of oxidationfilm, and even causes the side effect like that the peaktemperature falls down to 598 °C with time up to 25 h. Theside effect may also be ascribed to the mechanical abrasionof oxide film and exposure of inner defects through a longterm of rolling in the condition of oxidation reaction lyingin an equilibrium state.

The sensitivity of oxidation film on temperature andtime can be also reflected by XRD. The XRD patterns forTi H₂samples, untreated and pretreated, are shown inFig. 5. The unmarked peaks correspond to the Ti H₂or thesubstoichiometric Ti H_xcompounds. As seen in Fig. 5,there is no significant change in the XRD patterns forpretreatment at 450 °C for 1 h. Until the treating time lastsup to 20 h, titanium oxides can be observed. The resultsillustrate that not only the structure of oxidation filmformed under 450 °C is uneven but also its quantity isinsufficient. While the oxidation temperature is over500 °C, titanium oxides, such as Ti₃O and Ti O₂, are clearlydetected, indicating the quickly oxidation reaction andformation of massive oxides on the Ti H₂surface withprolonging the oxidation time. The crystal structure ofTi₃O and Ti O₂(rutile) was testified to be hexagonal andtetragonal [15], respectively. Regardless of the formationand quantity of oxide products, Fig. 5 shows that all thehydride peaks, except the untreated Ti H₂, shift to thehigher angles in a varying degree. The untreated Ti H₂isidentified as Ti H_1.97with fcc crystal structure, which corresponds to hydrogen content of 3.94 wt%. Substoichiometric hydride, which is determined as Ti H_1.5with crystalstructure of d-phase [16], appears in all the shifting XRDspectra. The presence of Ti H_1.5indicates that the hydrogenloses along with the temperature elevating and timeextending. According to Figs. 2 and 3, the hydride withoutthe first peak means that if the stoichiometry of hydridechanges from Ti H_1.97to Ti H_1.5completely, nearly 25 %hydrogen might escape by heat treatment. The results lay afoundation for the oxidation treatment of Ti H₂particles inair with the residue hydrogen content of over 3.0 wt% andprove that the previous obtained information is reasonable.

Fig. 4 SEM images and corresponding EDS spectra of Ti H2particles before and after oxidation treatment: a untreated, b treated at 450 °C for25 h, and c treated at 500 °C for 15 h

Fig. 5 XRD patterns of Ti H2untreated and pretreated at differentconditions

The FORMGRIP process for metallic foam productionis presented in Ref. [17], which puts the pretreated titaniumhydride powder into the Al–9Si/Si Cp composite melt toobtain the precursor materials, and then the precursor wasremelted in a die to prepare aluminum foam. The process isvalidated to be flexibility in design of foam structures. Thefoamable precursors were also prepared by mixing thepretreated Ti H₂powder with aluminum, silicon, and copperpowders in proportion of Al Si₆Cu₄. The foaming resultsshow that the decomposition of pretreated Ti H₂is delayedand more uniform distributions of rounder pores are found[15]. In order to verify the foaming capacity of the Ti H₂particles pretreated at 500 °C for 5 h, these powders of1.2 wt%, as well as the untreated Ti H₂, were, respectively,added into the Al Ca₃melt, then formed at 690 °C for4 min. The prompt release of hydrogen stored in untreatedTi H₂leads to a certain expansion of aluminum melt andsustained explosion caused by releasing the hydrogen. Nosimilar phenomenon occurs in the process of pretreatedTi H₂addition. Although no significant difference in porestructure occurs, the foamed block using the pretreatedTi H₂as foaming agent is nearly 20 % higher than that ofuntreated one. This result indicates that the foamingcapacity of the pretreated Ti H₂is not affected by the disappearance of the first peak held by the untreated Ti H₂.Instead, the delay of hydrogen release is beneficial to thedispersion of Ti H₂particles and full use of the hydrogen.

5 Conclusion

DSC result and hydrogen content analysis illustrate that thefirst peak of untreated Ti H₂can be eliminated by oxidizingin air above 400 °C. The effect of temperature on theformation of oxidation film and decomposition behavior ofTi H₂is more significant than that of oxidation time. Theonset temperature and peak value of Ti H₂decompositioncan shift from left to right through elevating temperatureand extending time. The parameter for treating Ti H₂at500 °C between 1 and 5 h is preferred in consideration ofraising peak temperature above 600 °C and keepingresidual hydrogen content over 3.0 wt%. The similar resultcan be realized by treating Ti H₂at 450 °C, but the oxidation interval lasts between 10 and 20 h.

SEM and XRD results indicate that there are differentcrystal structures of oxidation film formed at differentconditions. The crystal morphology formed under 450 °Cis coarse and inhomogeneous in thickness, while oxidationfilm generated above 500 °C is compact and uniform. Thehydrogen evolution process follows by changing stoichiometry of the hydride. About 25 % hydrogen escapingfrom the untreated Ti H₂corresponds to the stoichiometryfrom Ti H_1.97to Ti H_1.5. The delay of hydrogen evolutioncontributes to the dispersion of Ti H₂particles in moltenAl Ca₃alloy and full utilization of hydrogen in preparingaluminum foam.