Abstract

Electrostatic levitation combined with laser heating is becoming a mature technique that has been used for several fundamental and applied studies in fluid and materials sciences (synthesis, property determination, solidification studies, atomic dynamic studies, etc.). This is attributable to the numerous processing conditions (containerless, wide heating temperature range, cooling rates, atmospheric compositions, etc.) that levitation and radiative heating offer, as well as to the variety of diagnostics tools that can be used. In this paper, we describe the facility, highlighting the combined advantages of electrostatic levitation and laser processing. The various capabilities of the facility are discussed and are exemplified with the measurements of the density of selected iron-nickel alloys taken over the liquid phase.

1. Introduction

Noncontact processing of materials is required when dealing with corrosive materials or when there is a need to reach and maintain samples at high temperatures for times long enough to complete the processing. This is a stringent requirement for materials (e.g., refractory metals and ceramics) with melting temperatures higher than those of crucibles. In particular, containerless conditions are desirable to allow a host of investigations (thermophysical property measurements, high energy beam interaction (neutron, synchrotron, etc.)), solidification studies, synthesizing new phases, to name but a few. Moreover, processing without containers avoids any physical contact which can contaminate or change the shape of a sample. By doing so, data analysis is also simplified and it is possible to obtain properties of a material in its purest form. In addition, this allows to reach and maintain a material under undercooled conditions (below melting temperature). This limits heteregeneous nucleation and eases vitrification. Similarly, the reduced gravity conditions prevailing on orbiting platform (space shuttle, International Space Station, etc.), in sounding rockets, in airborne laboratory or in drop shaft, all require some sort of positioning system to maintain a processed material subject to residual forces in a specific location [16].

Electrostatic levitation is very well suited to meet the above requirements for the noncontact processing of materials. In particular, the method does not intrinsically heat the samples and does not deform them. Furthermore, it allows the processing of conducting and nonconducting materials, solid or liquid, while offering a wide field of view of the processed materials. It can operate under high vacuum or atmospheric pressure, it can control the rotation of samples, and can be equipped so that controlled induced oscillation and nucleation are possible in a processed material.

Most of the studies in materials science do require some sort of heating. In the context of contactless processing, radiative heating, in particular, laser heating, is very attractive because it allows precisely controlled and fast heating and cooling. Moreover, the laser can impart momentum to a processed material through photon momentum transfer. By selecting the emission wavelength, the laser source allows to target specific materials and also permits to avoid spectral interference (e.g., with cameras, pyrometer).

Using laser heating with an electrostatic levitator combines their own merits with the additional advantages of having a processing facility in which heating and levitation controls are independent. Furthermore, fast heating and fast radiative cooling are possible when the materials are processed under high vacuum. In addition, the laser beam can access specific region of a levitated sample, perform localized heating, generate micrograving (e.g., spherical IC), or photo-induce interactions (physical, chemical, etc.). In such a hybrid facility, the field of view of the sample is not blocked by a heating device such as coils.

In this paper, we first describe a hybrid processing facility that combines electrostatic levitation and laser heating, highlighting its capability in terms of environment, heating, and cooling. The fabrication as well as the processing of a man-made analogue of asteroid materials is presented. Representative density data for a few iron-nickel alloys are then reported. These data could be of help for numerical modeling and simulation of early formation, differentiation, and disruption of asteroids.

2. Experimental Methods and Procedures

2.1. Electrostatic Levitation

The measurements reported in this paper have been carried out using an electrostatic levitation furnace developed by JAXA [7, 8]. The apparatus described here was based on a design by Rhim et al. [2] but modified in areas of sample handling, charging, electrode design and configuration, feedback implementation, levitation initiation, imaging, and heating configuration [711] without which the described experiments would have been difficult to perform. Figure 1 illustrates schematically the electrostatic levitation furnace. The facility consisted of a stainless steel chamber which was typically evacuated to ~10−5 Pa before sample processing was initiated. A similar facility can sustain a pressurized atmosphere of up to 1 MPa with a variety of pure or mixture of gases (air, N2, O2, Kr, etc.).

In this work, Fe-Ni binary alloy samples were prepared by arc melting measured quantities of solid pieces of the component metal in a water-cooled copper hearth under Ar atmosphere. The metals used for the alloy preparation were high purity iron (99.995% mass) and nickel (99.99% mass) from Nilaco, Tokyo, Japan. Once a large piece of material was made, it was cut into small pieces that were arc melted once again into ca. 2 mm diameter spheroid samples. For this study, the emphasis was given on alloys with three compositions, Fe50Ni50, Fe75Ni25, and Fe25Ni75 because M-class asteroids (Tholen taxonomic scheme) [12, 13] and part of classes Q and S asteroids [14] are believed to be constituted of such metallic alloys. Analysis of iron meteorites originating from the M-class asteroids revealed composition of taenite (Fe80−35Ni20−65) or of kamacite (Fe95−90Ni5−10) [15]. Data of thermophysical properties of Fe-Ni alloys would provide valuable information for numerical modeling of asteroid impact, formation, and evolution (e.g., melting, differentiation) [16]. Therefore, this would help to understand the primeval processes in the asteroid belt and give insights about the formation of the solar system. Moreover, the data could be of interest to elucidate issues related to the core of the Earth.

To handle the samples, a sliding cartridge containing ten individual tungsten pedestals was used. This helped to circumvent any cross-talk contamination problems between distinct samples and permitted the handling of soft, porous, or submillimeter materials [7, 8]. To perform a processing experiment, one of the 10 samples located in the cartridge was brought up to the level of a bottom electrode. The sample was then charged by thermionic emission and was levitated between two horizontal electrodes, a concave bottom electrode (30 mm dia.), and a flat top electrode (10 mm dia.), both having a through hole. An electrical field of ~12 kV field was applied between the two electrodes, and a feedback control software was activated. The top electrode was gimbaled with three micrometer screws, allowing accurate electrode balancing and spacing (10 mm between both electrodes). These electrodes were mainly effective for vertical position sample control. However, the conical electrical field distribution of the electrode configuration provided a restoring force towards the center, thus contributing to the horizontal position control as well. In addition, four secondary spherical electrodes with applied positive or negative potentials, distributed around and at the height of the levitated sample, were used to further control the horizontal position. Since the electrostatic scheme could not produce a potential minimum, a feedback position control system was necessary. Position sensing was achieved with a set of orthogonally disposed He-Ne lasers and the associated position detectors. The position information, obtained by the shadow of the He-Ne laser illuminated sample, was fed to a computer that input, new values of , , and to a high voltage amplifier at a rate of 720 Hz so that a prefixed position could be maintained. The lower electrode was also surrounded by four coils that were used as a stator to generate a horizontal and rotating magnetic field. The magnetic field was used for sample rotation control [17]. In addition, an ac voltage could be superimposed on the levitation voltage from the top electrode to excite drop oscillations for the surface tension and viscosity experiments.

The sample was observed by three charged-coupled-device (CCD) cameras (Figure 1). One camera offered a general view of both the electrode assembly and the sample. Two high resolution CCD cameras, equipped with telephoto objectives in conjunction with high intensity UV or visible background lights allowed sample perimeter and surface features to be analyzed [9]. In addition to the CCD cameras, each telephoto objective was equipped with a half-mirror, an interference filter (He-Ne emission line), and a detector. One of the detectors, coupled with a monochromator slit, was used to determine the oscillating drop amplitude from the shadow of an He-Ne laser backlit sample and was dedicated to the measurement of the sample oscillation [18]. The other sensor was used for sample rotation rate measurement by detecting the reflected He-Ne laser beam from its surface [17].

Once the sample was levitated, the next step in the experiments consisted in melting and resolidifying a levitated sample to confirm pyrometer calibration and alignment. The sphericity of the sample was confirmed by software analysis by comparing its shape with that of a calibration sphere.

2.2. Laser Processing

Sample heating was performed using the radiation coming from computer controlled CO2 and Nd:YAG lasers (Figure 1). To avoid the destabilizing effect of photon and evaporative anisotropic-induced forces [19] and to maintain good position stability, for materials processed under vacuum, while observing the constraint of the chamber layout, a flattened tetrahedral laser heating configuration was implemented (Figure 2) [10, 11]. Specifically, three focused beams (50 W each) of CO2 lasers (Synrad, 10.6 μm emission) in a same plane, separated by 120 degrees, hit the specimen. In addition, the focused beam of a 500 W Nd:YAG laser (Lee Laser, 1.064 μm emission) (equipped with a 0–90% power attenuator) coming from the top can be used when additional power is needed (e.g., for refractory materials such as Os, Re, and W). Heating with the most powerful laser along the vertical was the best configuration as the strongest field and the fastest feedback were also along this axis. Besides the exceptional 3D sample position stability it provided, this multibeam heating configuration helped to control sample rotation and improved temperature homogeneity.

2.3. Density Measurements

In this paper, one of the capabilities of the facility is exemplified with the measurements of the density of selected iron-nickel alloys. The density was determined by simultaneously recording the temperature and magnified images of a spherical sample illuminated from behind with an ultraviolet source [9]. Upon closing the shutters of all lasers, the sample cooled and even undercooled, yielding data over a large temperature range. The sample area was extracted from each digitized video images and matched to a temperature profile. These images were calibrated by levitating a sphere of precisely known diameter under identical conditions. Since the sample was axi-symmetric and because its mass was known, the density could be found for each temperature by dividing the mass by the volume.

2.4. Other Processing Capabilities

Several capabilities of combining electrostatic levitation and laser heating are briefly described. The interested reader can consult the given references for more information.

2.4.1. Surface Tension and Viscosity Measurements

In addition to its usefulness for industrial processes (e.g., bubble migration, refining, casting, welding), the knowledge of the surface tension and viscosity and their temperature dependences is important in fundamental studies (e.g., atomic dynamics, surface physics) and when designing new high performance alloys [20]. The surface tension and the viscosity can be determined with the current facility by studying the behavior of the sample oscillation about its equilibrium shape [21]. In this technique [18, 22], an oscillation is induced to the levitated sample by superimposing a small sinusoidal electric field on the levitation field. The transient signal that followed the termination of the excitation field is detected and analyzed. This is done several times at a given temperature and repeated on a large temperature range. The surface tension can then be found using the characteristic oscillation frequency of the signal after correcting for nonuniform surface charge distribution [23, 24]. Similarly, the viscosity is obtained using the decay time of the same signal [24].

2.4.2. Vapor Pressure Measurements

The sample area variation in time could be accurately measured for a given temperature using the UV imaging technique, knowing the elapsed time between the start and the end of an experiment. Since the sample evaporated isotropically, the effusion Knudsen method [31] could be utilized assuming that the effective area of effusion was the surface of the sample at a given time. Therefore, the vapor pressure could be found using the measured rate of evaporation in vacuo, the gas constant, the sample temperature, and the molecular weight of the material [32].

2.4.3. Atomic Structure Characterization by Neutron Scattering Experiments

Vacuum electrostatic levitation furnace was also developed for the structural study of materials above the melting point as well as in undercooled phase by neutron scattering [33, 34] and synchrotron radiation [35]. Preliminary experiments performed with a solid alumina sample did not reveal any sharp peaks in the background data coming from the furnace materials which is an advantage compared to other methods [3638]. The observed diffraction peak intensities and location were identified as those derived from the mirror indices of hexagonal structure of alumina and were in complete agreement with those reported in the literature [39]. Experiments to investigate the characterization of the atomic structure and dynamic of liquid boron by synchrotron radiation have also been initiated [35].

2.4.4. Processing and Solidification from Supercooled Phase

Besides elemental metals, several oxides (e.g., Nd-CaAl2O4, Y3Al5O12, Al2O3, some slags, BiFeO3, BaTiO3) have been processed using either the pressurized or vacuum facilities. It was even possible to vitrify several compositions from the undercooled state [40, 41]. The containerless solidification behavior of Nd2Fe14B was also studied at different cooling rates and the effects of undercooling depth on the microstructure and magnetization were investigated [42]. For the sample solidified from the deepest undercooling level, a fine microstructure was obtained and the higher magnetization was revealed. Solidification of barium titanate (BaTiO3) from deep undercooled phases revealed transparent materials that consisted of micrometer-size particles and a single crystal phase exhibiting a giant dielectric constant and weak temperature dependence [43].

3. Results

3.1. Heating and Cooling

Here we report representative data obtained with the combined capabilities of electrostatic levitation and laser heating.

The radiance temperature was measured by two single-color automatic pyrometers (Chino Corp, Model IR-CS 2S CG and Chino Corp, Model IR-AP: 0.96 μm and 0.98 μm working wavelength, 10 Hz and 120 Hz acquisition rates, ±50 K) and was calibrated using the known melting temperature of the binary alloys. Calibration to the true temperature was performed from Planck’s law after adjusting the melting plateau to correspond to the known melting point of each binary alloy (Fe50Ni50, 1450°C, Fe75Ni25, 1475°C, and Fe25Ni75, 1460°C) [44]. The thermal history for three compositions: Fe25Ni75, Fe50Ni50, and Fe75Ni25 is illustrated in Figure 2. Typical profiles for cooling samples of three alloy compositions exhibiting undercooling (from 260 degrees for Fe25Ni75 to 310 degrees for Fe75Ni25) and recalescence (sudden temperature rise due to the release of the latent heat of fusion of undercooled samples upon solidification) are shown in Figure 2. For example, in Figure 2(b), a 2.7 mm diameter sample (18.42 mg) was heated up to 1550°C. At time ~10 s, the laser power was deliberately turned off and the sample cooled down to the undercooled phase before solidifying as evidenced by recalescence.

Although recent measurements with a fast polarimeter showed variations in the emissivity in the liquid phase for several metals [44], the lack of data for the undercooled binary alloys prompted us to assume that the emissivity remained constant over the whole undercooled range. Optical microscopy and electron probe microanalysis were performed to ensure that voids or oxide or nitride layers were not present on the surface or in the bulk of the solidified samples and that the materials did not undergo incongruent evaporation during processing.

3.2. Density Measurements

The density measurements of liquid binary alloys were taken over a large temperature range, covering the undercooled region by a few hundred degrees. The density data are presented in Figure 3 for three compositions: Fe25Ni75, Fe50Ni50, and Fe75Ni25. For all compositions, the density showed a linear behavior as a function of temperature and can be, respectively, fitted bywhere is the melting temperature for each binary alloy. In these measurements, the uncertainty was estimated to be less than 2% from the resolution of the video grabbing capability (640 × 480 pixels) and from the uncertainty in mass measurement (±0.0001 g). A detailed uncertainty analysis can be found elsewhere [45]. To our knowledge, these measurements are the first to cover such a large temperature span, including the undercooled region. No value for the temperature coefficient was found in the literature for comparison.

The data reported in the open literature were taken for only one temperature. The literature data, whenever necessary, were interpolated to match our compositions and to enable comparison. From Table 1, it can be seen that our data are, within experimental uncertainties (<2%), identical to those reported in the literature [2530]. The merit of the electrostatic levitation is to allow such measurements over a large temperature span.

4. Summary and Conclusions

The combination of electrostatic levitation and laser heating is a wonderful approach for the processing, characterization, and study of materials as exemplified in this paper with Fe-Ni alloys. Furthermore, it shows promise for the synthesis of novel materials by controlling the resulting phases through appropriate cooling.

Property measurements would nonetheless benefit from a few improvements and additions to the current techniques. The density measurement technique is quite mature but a further improvement might allow to resolve the changes in the liquid density due to the structural changes occurring in some liquid phase transitions of glass-forming materials. The use of vacuum ultraviolet or X-ray sources as a background to the sample would offer improved imaging, in particular for samples near 4000 K and when performing experiments under very high pressure. The surface tension and viscosity data obtained so far are remarkable in a sense that they provided measurements unknown before, and that, all the way to tungsten (>4000 K). However, large scatters still appear in the viscosity data (even when the feedback frequency of the levitation system is diminished [46]) due to the motion of the sample upon exciting the drop oscillation as the feedback control system tends to bring back a perturbed sample to its original position. Moreover, since the electrostatic scheme does not input any heat, a high temperature sample in vacuum experienced pure radiative cooling when the heating sources are shut off. From the simplification of the energy equation governing the cooling, the ratio of constant pressure heat capacity and hemispherical total emissivity could be determined. An independent way to measure the hemispherical total emissivity is currently being pursued [47].

Over the last few years, a strong focus was put on the thermophysical property measurements of refractory materials, in particular elemental metals (e.g., Ta, Os, Re, W). From now on, emphasis will gradually shift towards the property measurements of industrial ceramic and metallic alloys.

Although electrostatic levitation combined with laser heating has shown wide applications and is still in its infancy, some materials are nonetheless difficult to process on the ground. Reduced gravity conditions offered by an orbital platform would allow noncontact positioning of these materials or larger samples while providing an hydrodynamically quiet environment. Similarly, to better understand the effect of gravity on processing, solidification, and vitrification to name but of few, the Japan Aerospace Exploration Agency has pursued, for already several years, the development of an electrostatic levitation facility with laser heating for the International Space Station. The commissioning should occur in the 2013-2014 time frame.

Because electrostatic levitation combined with laser heating offers a variety of processing conditions (large range of pressures (~10−5 Pa to 1 MPa) and gaseous compositions (air, N2, O2, Kr, etc.), controlled heating (up to 800 W), cooling, and solidification), the conditions of outer space or those of planetary atmospheres can be simulated. As such, the technique could be used in processing man-made analogues of the constituents of meteorites, asteroids, transneptunian objects, or comets to better understand body collisions, meteoritic impacts, impact debris, or bolide atmospheric entry.

Acknowledgments

This work was partially supported by a Grant-in-Aid for Scientific Research (B) from the Japan Society for the Promotion of Science. One of the authors (P. F. Paradis) wishes to acknowledge the Institut national d’optique (Quebec) for short-term sabbatical leaves.