Ion implantation offers several significant advantages, including precise control over the amount of ion implantation, minimal horizontal diffusion, targeted ion implantation at desired positions, the ability to conduct the process at low temperatures, and excellent reproducibility. These benefits make ion implantation an essential step in high-tech semiconductor manufacturing, despite the need for subsequent heat treatment due to the complex equipment structure and potential damage to the lattice of the sample [1,2]. When constructing a semiconductor circuit on a silicon (Si) substrate using ion implantation, it is crucial to ensure that the ions are distributed precisely at the desired locations. This precision is essential when ions must be implanted at low energy to achieve a shallow depth distribution within the sample. The accuracy of this process can be influenced by several factors, including the type of ions used, energy of the implantation ions, dose of the implantation ions, temperature, and tilt/rotation settings during implantation [2–5].

In the specific scenario of implanting boron ions (B-ions) into Si using tungsten trioxide thin films (WO₃-TF), the optimization of Bion implantation can be predicted by analyzing the depth profile results obtained from crystal transport of ions in matter (Crystal-TRIM) simulations and the hole concentration results derived from the hole concentration program. These predictive analyses are critical for finetuning the implantation process to achieve desired electrical characteristics in semiconductor devices.

2.1. Crystal TRIM

Crystal-TRIM was developed by Dr. Poselt in Germany in 1991 and has been supplemented annually [6]. Crystal-TRIM allows ions to be implanted with crystallinity in the sample, whereas TRIM performs ion implantation with amorphous samples. The program was created to enable 1 and 2 dimension depth profiling, and three types of samples are available: Si, germanium (Ge), and diamond [6–11].

2.2. Hole concentration program

Based on previous studies, we intended to create a function between the depth profile and hole concentration profile as influenced by the dose of B-ion implantation. Table I summarizes the relationship between the dose and hole, allowing for the calculation of hole concentration based on experimental studies. The conditions under which B-ions form the shape and structure of the crystal inside the Si were determined using X-ray photoelectron spectroscopy. In addition, the hole distribution was analyzed through hole measurements, and the conditions for obtaining a usable voltage for the device were determined [12–15]. Figure 1 illustrates the hole concentration profile obtained through hole measurements, and Fig. 2 presents the hole concentration profile obtained from the hole concentration program. The two figures appear extremely similar; therefore, it is possible to infer a hole-generated area within the Si during B-ion implantation.

Figure 1. Depth profile of hole concentration in as-implanted sample. A high concentration of holes is generated at a depth where the boron concentration is the highest.

Figure 2. Calculation of hole concentration program from depth profile.

Table 1 . Relation between dose and hole depth..

B con. more than 2.0 × 1021		α=1, σ=200, Rp= Rp	Hole = (con. × 2/12) × (60/100) × Hi (x) Hi(x) = a ${\text{exp}}^{-\frac{{\left(x-R_p\right)}^2}{2{\sigma }^2}}$
B con. less than 2.0 × 1021	Dose more than 1.0 × 1016	α=1, σ=5, Rp= depth
	Dose less than 1.0 × 1016	α=0.01, σ=5, Rp= Rp

2.3. Simulation of B-ion implantation into Si with WO₃-TF at 35, 50, and 65 keV

Si with <100> crystallinity was simulated. B was ion-implanted with an energy of 35 keV, tilt/rotation was at 0∘/0∘, and the temperature was 300 K, which is room temperature. During ion implantation, a depth profile was obtained by changing the dose at a constant rate. The depth profile was obtained by dividing the dose into five values in the ultrahigh dose area: 1.0 × 10¹⁶, 2.0 × 10¹⁶, 3.0 × 10¹⁶, 5.0 × 10¹⁶, and 1.0 × 10¹⁷ cm⁻². There is an essential parameter for calculating the electronic cross-section during boron ion implantation. The Cel and C_el(110) values used in the equations related to the interaction between the grids were set to 1.25 and 1.00, respectively, and the cacc and ccrit values were set to 0.1 and 1.0, respectively. For the thickness of the WO₃-TF, values of 40, 80, 120, 160, and 200 nm were used. WO₃ is a chemical compound containing one atom of W (atomic number 74) and three atoms of O (atomic number 8).

The simulation was performed under the same conditions as for the case where B-ion implantation was simulated at 35 keV, and the initial energy values were changed to 50 and 65 keV. In this case, thicknesses of 80, 120, and 160 nm were considered for the WO₃-TF. This was because the effects at 40 and 200 nm were insignificant from the simulation performed at 35 keV. Therefore, the state value of the simulation was applied in the same manner, and only the ion implantation energy was changed.

Figure 3 presents the Crystal-TRIM simulation results for B-ion implantation at an energy level of 35 keV. Given that not all simulation outcomes were essential for the analysis, we focused on the results indicative of the optimized values. In this scenario, the WO₃-TF had a thickness of 40 nm, and the distribution of B-ions did not extend toward the surface beyond a depth of 250 nm. This observation suggests that the energy loss that occurred as the ions passed through the WO₃-TF layer significantly affected their penetration depth during implantation. Figure 4 illustrates the Crystal-TRIM results for Bion implantation at 50 keV. The thickness of the WO₃-TF was 80 nm. These results indicate that B-ions penetrated deeper than 600 nm into the Si(100) layer.

Figure 3. Crystal-TRIM simulation results for ultrahigh dose of B-ion implantation into WO₃-TF (40 nm)-Si(100) at 35 keV.

Figure 4. Crystal-TRIM simulation results for ultrahigh dose of B-ion implantation into WO₃-TF (80 nm)-Si(100) at 50 keV.

At the highest implantation dose of 1.0 × 10¹⁷ cm⁻², there was a predicted potential for hole generation within the 80–260 nm range, suggesting the formation of a hole generation area approximately 180 nm thick. Figure 5 illustrates the results of the Crystal-TRIM simulation for B-ion implantation at 65 keV with a 120-nm thick WO₃-TF. The simulation results demonstrate two prominent peaks in the distribution, indicating the likelihood of hole formation within the 120–320 nm range at the highest dose of 1.0 × 10¹⁷ cm⁻². Additionally, the efficiency of hole generation in the Si(100) layer is expected to be superior at 65 keV than at 50 keV.

Figure 5. Crystal-TRIM simulation results for ultrahigh dose of B-ion implantation into WO₃-TF (120 nm)-Si(100) at 65 keV.

Figures 6–8 illustrate the hole concentration program results at 35, 50, and 65 keV, respectively. Figure 6 presents the hole concentration graph for a 40-nm thick WO₃-TF. The effective electrical conductivity is shown for the case where the dose ion-implanted into Si(100) was 1.0 × 10¹⁷ cm⁻²; the area was measured to be 50–170 nm². Importantly, it formed at the interface between WO₃-TF and Si(100) where the area was the widest. Figure 7 shows the hole concentration graph corresponding to an 80-mm thick WO₃-TF. The effective electrical conductivity is shown for the case where the dose ion-implanted into the Si(100) was 1.0 × 10¹⁷ cm⁻²; the area was measured to be 90–220 nm². It formed at the interface between WO₃-TF and Si(100), and the area was the widest at 130 nm. This can be correlated between the initial energy value of ion implantation and the optimized thickness of the WO₃-TF. Figure 8 illustrates the hole concentration graph corresponding to a 120-mm thick WO₃-TF. The effective electrical conductivity is shown for the case where the dose ion-implanted into the Si(100) was 1.0×1017 cm⁻²; the area was measured to be 130–270 nm². The hole generation section was 140 nm wide and distributed around the interface between WO₃-TF and Si(100); thus, it was the most optimized.

Figure 6. Hole concentration program results for ultrahigh dose of B-ion implantation into WO₃-TF (40 nm)-Si(100) at 35 keV.

Figure 7. Hole concentration program results for ultrahigh dose of B-ion implantation into WO₃-TF (80 nm)-Si(100) at 50 keV.

Figure 8. Hole concentration program results for ultrahigh dose of B-ion implantation into WO₃-TF (120 nm)-Si(100) at 65 keV.

Table II summarizes the area where the holes were generated for the various thicknesses of the WO₃-TF and the initial ion implantation energy at a dose of 1.0 × 10¹⁷ cm⁻². A value optimized for the initial ion implantation energy may be selected according to the thickness of the WO₃-TF.

Table 2 . Hole generation area corresponding to a dose of 1.0 × 10¹⁷ cm⁻²..

Initial energy WO3-TF	35 keV	50 keV	65 keV
40 nm	50–170 nm (120 nm)	-	-
80 nm	80–170 nm (90 nm)	90–220 nm (130 nm)	130–280 nm (150 nm)
120 nm	120–130 nm (10 nm)	120–210nm (90 nm)	130–270 nm (140 nm)
160 nm	-	160 nm	160–240 nm (80 nm)
200 nm	-	-	-

Crystal-TRIM simulations were performed to determine the relationship between WO₃-TF thickness and initial ion implantation energy. Holes could be stably generated only when the dose was 1.0 × 10¹⁷ cm⁻², even at initial ion implantation energies of 50 and 65 keV. The hole generation area was determined for a dose above 1.0 × 10¹⁷ cm⁻². Crystal-TRIM simulations make it possible to predict the area of the holes generated in Si and to determine the conditions for obtaining an optimized device. The optimization conditions vary depending on the characteristics of the thin film type. Crystal-TRIM continuously accumulates data in parallel with experiments even now from the simulation of various components [16–18]. If the optimized conditions are predicted using Crystal-TRIM and hole concentration programs, and data on the experimental characteristics of different elements are analyzed to improve the hole concentration program, it can contribute significantly to the development of various semiconductor devices.

This study was supported by the National Research Foundation of Korea (Grant number: RS-2023-00249361).

The authors declare no conflicts of interest.