The Dubon Group

Like the related technique of pulsed laser deposition (PLD), pulsed laser melting (PLM) uses a pulsed, high-power laser for material modification. However, in PLM the laser beam impinges directly onto the sample whereas in PLD it is directed onto a target which is ablated onto the sample substrate. Our group typically uses PLM to grow epitaxial semiconductor alloys from ion-implanted substrates in a process we abbreviate as II-PLM. Figure 1 gives a schematic of the two-step II-PLM process. We have successfully used II-PLM to produce films of many difficult-to-produce compound semiconductor alloys: ferromagnetic III-V semiconductors like Ga_1-xMn_xAs and Ga_1-xMn_xP as well as highly-mismatched alloys such as GaAs_1-xN_x and Zn_1-xMn_xTe_1-yO_y.

Figure 1 - Schematic of the ion implantation and pulsed-laser melting (II-PLM) process used by our group to synthesize semiconductor alloy films. A semiconductor wafer is first implanted with the alloying species and then irradiated with a single pulse from an excimer laser. The laser energy is converted to heat, which melts through the implant-damaged layer into the underlying substrate. Epitaxial solidification proceeds as heat is extracted into the substrate, resulting in a single-crystalline alloy film on the original substrate.

PLM was originally investigated as a technique for activating extraordinary concentrations (up to 10²¹ /cm³ !) of dopants in semiconductors. In the early years (1970's and 1980's), it was known as "pulsed-laser annealing" as the laser melting took the place of conventional furnace annealing or rapid thermal annealing (RTA) for repairing implant damage and activating the implanted dopants. Laser irradiation in the sub-melting regime can be used to drive solid phase epitaxy (SPE), the process in which the amorphized implanted layer is converted directly into a crystalline epitaxial film. It should be remembered, however, that PLM is qualitatively different than any of the other techniques mentioned herein because it involves passing through a liquid phase. This fact restricts to which materials systems II-PLM can be successfully applied - films derived from Si, Ge and GaAs can be easily synthesized, while alloys based on GaN or InN can not because they do not form a stable liquid phase. A rule of thumb - if the substrate material of interest can be grown by Czochralski, then PLM has a good chance of success.

This section takes a more detailed view of the II-PLM process. The synthesis of III-V ferromagnetic semiconductor films is used as an illustrative example of how the II-PLM process can be applied to a materials system in which film synthesis is challenging.

Unlike many material processing methods, both ion implantation and PLM occur far enough from thermodynamic equilibrium to be governed by kinetics. Ion implantation produces non-uniform spatial distributions of implanted species which are quenched in due to near-negligible diffusion distances at room temperature, while PLM exploits the transient regime of heat flow following a rapid deposition of energy by the laser. These departures from thermodynamic equilibrium are what are exploited to form homogeneous III-Mn-V ferromagnetic semiconductor alloys, which require Mn concentrations 2-3 orders of magnitude higher than equilibrium solubility limits.

In ion implantation, ions of the desired species are accelerated by an electric potential to energies typically in the range 10-1000 keV and are directed onto the semiconductor substrate. The aggregate effect of implanting many ions (typical doses 10¹⁴-10¹⁶ ions/cm²) is a concentration depth profile resembling a Gaussian within the material which can be characterized by its center (projected ion range) and width (straggle), both of which scale (nearly linearly, in the current range of interest) with incident energy. Typical parameters for the implantation of Mn into GaAs and GaP used in our work are: range 200-600 Å, straggle 100-300 Å, peak concentration ~1-4x10²¹ /cm³ (5-20 at% of Ga sublattice sites). When the implanted ions come to rest, they may or may not inhabit a substitutional lattice site, so post implantation annealing of some form is necessary to allow them to diffuse onto one. Also, the injection of ions into the substrate does not come without cost; significant structural damage is produced in the crystal due to the formation of large numbers of vacancies and interstitials per ion. These can recombine, notably in the end-of-range region, to form extended defects such as stacking faults and dislocation loops. At the high Mn ion doses (10¹⁵-10¹⁶ /cm²) and low energies (30-100 keV) used in our work, the long-range order in the crystal is completely destroyed within a depth (very) roughly approximated by the ion range resulting in an amorphous layer. This structural disorder must be removed by an annealing process which occurs far from thermodynamic equilibrium in order to achieve substitutional, ferromagnetic III-Mn-V alloys.

The characteristics of pulsed laser melting make it ideal for synthesizing epitaxial films of ferromagnetic III-Mn-V alloys. In our work to date, a single pulse (~30 ns) from a KrF excimer laser (λ=248 nm, E_photon = 5 eV, fluence 0.2-0.5 J/cm²) is directed onto an implanted substrate after being spatially homogenized. The transient melting and solidification process has been well understood in terms of 1-D heat flow into the substrate. The absorption length for UV photons in amorphous GaAs is approximately 15 nm, thus most of the laser energy is deposited in the near surface region and converted to heat. As the heat flows into the substrate, the liquid / solid interface moves through the ion damaged region into the underlying substrate, which seeds single-crystalline epitaxial solidification as the front returns to the surface. The short times and small distances typical in II-PLM result in truly unparalleled solidification characteristics - solidification velocities of a few m/s and cooling rates of 10⁹-10¹⁰ K/s - rates exceeding those in splat quenching! This high velocity results in a departure from local equilibrium the interface known as solute trapping, whereby impurity atoms are essentially buried in the growing crystal at concentrations far in excess of the solid solubility because the solidification rate exceeds the diffusive velocity. In the case of Mn in GaAs, the equilibrium solubility limit at room temperature appears to be in the range of 10¹⁸ - 10¹⁹ /cm³, while we routinely produce films with Mn concentrations in the 10²¹ /cm³ range thus demonstrating how the solute trapping resulting from PLM allows us to investigate metastable alloy systems.