Ultrafast laser micro-processing finds application in various industries such as consumer electronics, biomedicine, aerospace, information technology, energy, and novel materials. Therefore, accurately detecting temperature information during ultrafast laser micro-processing is crucial for optimizing the laser pulse width and repetition rate to ensure precise process guidance. Several temperature measurement methods have been proposed, including those based on the Langmuir probe volt-ampere curve, Boltzmann spectrum, Saha-Boltzmann spectrum, and spatial deconvolution. However, each of these methods exhibits certain limitations. This study aims to verify the effectiveness of using laser-induced blackbody radiation to measure temperature during ultrafast laser micro-processing. Furthermore, this method is applied to analyze the surface temperature of silicon wafers and copper plates induced by single-pulse laser irradiation on a nanosecond time scale.

Silicon wafers and copper plates, representing semiconductors and metals, respectively, were chosen for transient temperature measurement during ultrafast laser surface micro-processing by blackbody radiation. An experimental setup was devised to measure the blackbody radiation spectrum generated by laser machining on a nanosecond scale, comprising a time calibration system to ensure zero measurement time. A special sequence of single-pulse processing was designed to sequentially present the spectral results, forming rows corresponding to the processing steps, while the time-resolved spectrum was arranged in a processing column. The spectra were acquired using an intensified charge-coupled device, and a comprehensive analysis of the spectral information was performed. The measured spectra were fitted to a Planck curve using the least-squares method. A finite difference method for heat diffusion was employed to model the temperature decay. Two parameters, the average temperature at 0 ns and the relaxation time, were used to quantitatively describe the process.

A time-resolved spectrum excited by femtosecond laser irradiation on the surface of a silicon wafer is examined and specifically presented at intervals of 1 ns, 2 ns, 4 ns, and 8 ns (Fig. 3). Similarly, the spectrum on the copper plate surface is observed at 6 ns, 14 ns, 22 ns, and 30 ns (Fig. 4). The laser-induced plasma emits spectra encompassing three distinct components: second harmonic generation, excited atomic spectrum, and blackbody radiation spectrum. Planck fitting of the femtosecond laser-excited blackbody radiation is performed from 0 ns to 13 ns at 1 ns intervals to measure the temperature of the silicon wafer (Fig. 5), while the range is extended for copper plate from 0 ns to 40 ns at 2 ns intervals (Fig. 6). The temperature measured on the surface of the silicon wafer exhibits a relative error, which increases from 13% at 2 ns to 25% at 12 ns [Fig. 7(a)]. Similarly, the relative error in temperature measured on the surface of the copper plate increases from 13% at 2 ns to 37% at 40 ns [Fig. 7(b)]. The temperature trend follows the exponential decay predicted by the finite difference method for heat diffusion. The fitted exponential curves reveal average temperatures of 231000 K and 226000 K centered around time zero with relaxation times of 4.41 ns and 2.97 ns for silicon and copper, respectively (Fig. 8).

The temperature of the laser-induced plasma on the surface of the silicon wafer is measured at 1 ns, ranging from laser excitation to 13 ns, while for the copper plate, measurements are taken at 2 ns intervals, spanning from laser excitation to 40 ns. These temperature measurements are conducted using the blackbody radiation spectrum. The finite difference method for heat diffusion accurately predicts the observed temperature decay. The average temperatures of silicon wafer and copper plate centered around time zero are 231000 K and 226000 K, accompanied by relaxation time of 4.41 ns and 2.97 ns, respectively. These findings validate the applicability of using blackbody radiation to ascertain the material temperatures during femtosecond-pulse laser processing. Nevertheless, further advancements are required in several areas, including the universal applicability of this method to materials containing complex components, the implementation of a vacuum environment for enhanced processing, and the theoretical verification of temperature equilibrium within the plasma particle system.