Laser Induced Fabrication of Nothing

Experimental procedures

Experiments were conducted to obtain B doped samples using same experimental setup of Figure ‎4 1 with replacing Phosphoric acid with Boric acid solution (2%). The dopant liquid was prepared by dissolving 2 gm of Boric Acid H3BO3 in 1000 gm of distillated water for 45 min then using ultrasonic for 1 min to ensure full solubility since 2% is the maximum solubility that can be obtained at room temperature. The liquid absorption coefficient and reflectance were also calculated by applying same experiment of ‎ Figure ‎4 2 and using Equation ‎4 1 to Equation ‎4 4 for calculations.

The experiment resulted in an absorption coefficient of approximately 0.27763 cm -1 which confirm low absorbance of Boric acid solution. This means that almost 94.5% of incident power is transmitted in case of (z=2 mm) dopant depth above substrate. This confirm the validity of the liquid to be used as a dopant source without affecting laser irradiated power since no significant losses occurs during transmission through the acid itself.

After immersing the diamond substrate in Boric acid 2%, ArF excimer laser irradiations were carried out at different fluences, repetition rates and number of laser shots as show in in the irradiation matrix of Figure ‎4 8. Electrical conductivity, Raman Measurement and Cathodoluminescence (CL) measurements were conducted for the sample after irradiation.

After acid treatment to remove conductivity for the sample surface, measurements were conducted to test electrical behavior of the irradiated areas. No electrodes were deposited, prope heads were directly contacted with irradiated spot. I-V measurements showed clear resistance decrease with increasing number of laser shots as shown in Figure ‎4 9, similarly to the experimental results using the phosphoric acid liquid. Electrical resistivity further decreased with increasing temperature as well which evidently indicate that the surface layer generated by irradiation is semiconducting and conductivity is not resulting from any residual graphite on surface.

Activation Energy

The activation energy for boron at low boron levels is reported to be about 0.37 eV [ ] and decreasing to almost zero demonstrating metallic conductivity at very high doping levels [ ]. It was also previously found that the experimentally detected boron activation energy in the samples decreased with an increase of B/C ratio [ ]. While the decrease in the activation energy increases the number of charge carriers at reasonable temperature, the mobility of the charge carriers decreases, limiting the conductivity and utility in some device structures and this problem still facing CVD diamond [7].in our experiment the activation energy was estimated from the temperature dependence of the electrical conduction via the Arrhenius plot as shown in Figure ‎4 10.

The activation energy estimated from the temperature dependence of the electrical conduction decreases with increasing the shot numbers, which might be associated with an increase in the B/C ratio. The linearity to some extend agrees with result obtained from the Arrhenius plot for a natural boron contained diamond and found it to be linear with a MSE of 1 meV [7]. The electrical response of boron doped epitaxial films can be quite complicated and type of conductivity still under investigation.

Raman mapping measurements

In order to investigate the influence of structural defects on the Raman spectra, confocal Raman measurements were performed as shown in Figure ‎4 11 : Typical Raman spectra measured at irradiated areas a) 5 j/cm2, 5 Hz, 1000 shots b) 5 j/cm2, 20 Hz , 50 shots c) non irradiated background. .Raman spectroscopy is a powerful non-destructive method to evaluate the quality of diamond. The diamond Raman peak is sensitive to the lattice structure, strain, and impurities [ , ]. The crystalline defects could be observed using analysis based on full-width-at-half-maximum (FWHM) and peak-shift mapping [ , , ]. For p+ diamond, the diamond Raman peak is broadened and shifted toward a lower wavenumber with increasing boron amount [ , ].

The Raman spectra were measured using Nanophoton RAMAN-touch confocal Raman microscope system with a Nd:YAG laser source operated at a wavelength of 532 nm and laser power of 5.5 W. All Raman spectra were measured before any acid treatment to the sample and at room temperature with fixed exposure time of 0.5 Sec and fixed integration number. For the Raman mapping analysis, the peak position and FWHM values were used to create the mapping images as shown in Figure ‎4 12.

It is known, that the first-order diamond peak at 1333 cm–1 (the zone-center optical phonon) usually decreases in intensity, downshifts in wavenumber, and broadens asymmetrically with an increase in the boron content [ ]. Raman showed single-peak centered at 1333 cm-1 which is due to diamond, but none peaks due to graphite or a-C. Mapping showed amplitude modifications depending on the doping conditions while Peak shift and FWHM mapping were kept intact. Intensity change could be referred to many factors:

• loss of crystallinity due to lattice distortion.
• electron-electron scattering contribution which increase with charge concentration.

Defect increase in crystal

Additionally, there could be other possibilities for intensity change will be further investigated with other measurements. It has been reported that the Raman peak position is shifted owing to the lattice strain generated around dislocations. The observed shift was not larger than +/-1.0 cm-1 from the diamond peak position of 1333 cm-1, which corresponds to compressive/tensile stress of tens or hundreds of MPa.

Large lattice strain of the order of GPa is typically observed in poly- or nanocrystalline diamonds, but rarely generated in single-crystal diamond [ ] and this could be the reason for not observing any peak shit. the asymmetric diamond Raman peak due to Fano-effect could not be observed which might be owning to lightly doping. according to previous studies an asymmetric shape appeared with CVD doped diamond owing to the Fano-effect caused by heavily boron doping accompanied with a small peak appearance at lower Raman shift. this was attributed to the local vibration modes of the boron pairs, but in our technique we are using different dopant and different incorporation method from CVD diamond.

Cathodoluminescence (CL) measurements

Cathodoluminescence (CL) spectroscopy was used to study exciton recombination. CL is a useful tool to investigate electronic states in diamond due to such lattice imperfections as impurities, point defects, and dislocations [ ]. appearance of CL bands with energies less than the band-gap energy (5.47 eV) in CVD diamond indicates an existence of energy states characteristic of the lattice imperfections such as, a very broad visible band (band A), is ordinary observed in diamond, originating from dislocations and is sometimes decorated by boron or some other impurities [ ]. Conversely, diamond without such CL bands means that it contains only a low density of lattice imperfections [ ]. Previously reported result regarding CVD diamond stated that the minimum energy of an electron-hole pair in diamond, the band gap, is 5.49 eV. The energy that binds

the electron to the hole in the free exciton is 0.08 eV; the minimum energy of a free exciton is thus 5.41 eV. Because diamond has an indirect band gap, luminescence from electron-hole pairs and free excitons requires the emission of one or more phonons for each photon. The most intense free-exciton luminescence line, at 5.27 eV, requires the emission of a 0.14-eV transverse-optic (TO) phonon. The neutral boron acceptor in lightly boron-doped diamond gives rise to a bound-exciton state with a binding energy 0.05 eV greater than the binding energy of the free exciton, and a total energy above the ground state of 5.36 eV. The TO phonon line of the acceptor-bound exciton in lightly boron-doped diamond

thus occurs at 5.22 eV [ ]. As shown in Figure ‎4 13 Cathodoluminescence (CL) measurements were conducted after H2 plasma surface treatment to avoid charging-up. bound-exciton (BE) due to B atoms that are substitutionally incorporated into diamond lattices could be clearly observed at the irradiated area. Yet it’s intensity is too small to confirm doping, still similar result obtained for lightly doped CVD diamond [ ].

Summary and future perspectives

Laser beam was partially characterized to understand it’s behavior and effect on diamond substrate. Generated heat upon irradiation was found to increase with increasing fluence. Melting depth and heat distribution was also greatly affected by laser fluence change. Exprimentally measured reflectivity proved that melting duration is approximately same as laser pulse duration and reflectivity increase with increasing fluence which indicate if partial or full ablation occurred upon irradiation.

Combining this result and result from simulation ,we believe that rise in the ambient temperature of the substrate, the fluence of the laser beam, laser pulse duration and ambient pressure controls the total energy required to melt the film and increase the melting duration to facilitate doping, so performing the expriment in heated vaccumed chamber and increasing pulse duration is expected to enhance the process greatly.

Singlecrystalline diamond samples were immersed in Phosphoric acid / Boric acid and irradiated at different fluences, number of shots and different repetition rates. Substrate was optically examined upon irradiation and no significant damages were observed. Samples was cleaned to remove graphitic layer generated and electrical behavior of irradiated area was tested and conductivity was confirmed at laser parameters. Electrical conductivity enhanced with increasing Fluence, number of shots, repetition rate and temperature which indicated that the area generated by irradiation is semiconducting.

Depth profile as well confirmed phosphorus incorporation in diamond up to thickness of 30 nm. Boron doped samples shows similar electrical behavior to phosphorus doped samples. Different measurements like Raman and CL measurements were conducted and showed possibility of boron incorporation to diamond.

Till now It was demonstrated that laser-induce doping is probably applicable to the surface doping of diamond and certainly increases the electrical conductivity of the surface, which might be beneficial to facilitating the formation of Ohmic contacts. Additionally, this method has a potential for exploring new dopant elements that have not experimentally investigated due to difficulties in the doping during the deposition so far.

It is just the beginning of the research at present, to clarify the doping mechanism and study the doping quality including damages in diamond, further systematic experiments are indispensable from this. The laser-induced doping certainly increases the electrical conductivity of the surface, which might be beneficial to facilitating the formation of Ohmic contacts between n-type diamond and electrodes. This method has a potential for exploring new dopant elements that have not experimentally investigated due to difficulties in the doping during the deposition so far.

Currently we are conducting more analysis such as X-ray photoemission spectroscopy, Near-edge X-ray absorption fine-structure spectroscopy, SIMS and Electrical measurements such as mobility and conduction type will be applied to the evaluation. We will widely investigate the doping method from this.

Fully understanding and controlling the process and achieving p-type and n-type diamond film with controllable concentration is our target for now. Next step will be successful ohmic contact fabrication for both p and n type and Fabrication of first p-n diamond homojunction by excimer laser irradiation. Once succeeded in this we can proceed for next step of transistor fabrication out of singlecrystalline diamond fabricated by excimer laser irradiations.