Revolutionizing Semiconductor Manufacturing: New Quadrupole SIMS Method Enhances Rapid Thermal Anneal Monitoring

In a new study published in the Journal of Vacuum Science & Technology B, lead author Dr. Zhixiong Jiang, principal engineer at NXP Semiconductors, and colleagues have unveiled a novel method for monitoring rapid thermal anneal (RTA) using secondary ion mass spectrometry (SIMS). This innovative approach promises to significantly improve the accuracy and repeatability of RTA monitoring, a critical process in semiconductor device manufacturing.

Understanding Rapid Thermal Anneal

Rapid thermal anneal (RTA) is a vital step in semiconductor fabrication, typically conducted at temperatures exceeding 900 °C. This process activates dopants in silicon (Si), creating specific dopant distributions essential for device performance. Precisely controlling RTA settings, such as temperature and duration, is crucial for achieving optimal device yield and functionality.

Traditionally, sheet resistance (Rs) measurement has been the standard method for RTA monitoring. This involves implanting a wafer with certain species, annealing it, and measuring its sheet resistance with a four-point probe instrument. Despite its high throughput and general repeatability, Rs measurement is prone to errors from wafer aging effects, such as dopant migration and native oxide growth, as well as variations in test wafers from ion implants. Additionally, Rs measurement lacks the sensitivity to detect slight variations in anneal temperatures, particularly at temperatures above 1000 °C.

The Role of Secondary Ion Mass Spectrometry (SIMS)

SIMS has long been recognized for its ability to analyze dopant depth profiles in silicon wafers. It can reveal differences in dopant distribution among wafers annealed under various conditions. However, traditional SIMS methods have not been precise enough for routine RTA monitoring, especially for detecting subtle changes in temperature around 1000 °C or higher.

Innovative Approach for Enhanced Monitoring

Jiang and his team at NXP Semiconductors developed a new method that leverages SIMS for routine RTA monitoring with unprecedented precision. By focusing on the peak concentration of boron (11B) in the dopant depth profiles and normalizing this peak concentration to a reference wafer from the same batch, the researchers achieved remarkable sensitivity (Figure 1). This approach eliminated the impact of wafer aging and lot-to-lot variations, which are common issues in Rs measurement.

The team conducted SIMS analyses using a CAMECA Quadrupole (Quad) SIMS 4550, employing a primary beam of O2+ at 500 eV and an incidence angle of 30°. The sputter rate was approximately 0.1 nm/s, and to improve the signal-to-noise ratio, they used a 3-second counting time for 11B+ and averaged the highest three data points. The reference wafer was a silicon wafer implanted with 80 keV 11B to a dose of 2.0 × 10^14 at/cm2.

The results were impressive: the normalized peak concentration method achieved an error margin of less than ±0.5%, enabling the detection of temperature changes as small as 1 °C in anneal temperatures exceeding 1000 °C. This sensitivity was further enhanced by averaging multiple SIMS analyses.

Impact on Semiconductor Manufacturing

"SIMS is a powerful inspection technique thanks to its unique combination of great detection sensitivity and excellent depth resolution. The work presented in this publication is one of many examples of how SIMS provides indispensable support to semiconductor process control in the industry," said Dr. Jiang.?

Implementing this Quad SIMS-based monitoring method has already demonstrated significant improvements in device yield. This method ensures better control over dopant activation and distribution by providing more accurate and repeatable monitoring of RTA processes, leading to more consistent and reliable semiconductor devices (Figure 2).

The study by Jiang et al. at NPX represents a significant advancement in semiconductor manufacturing technology. Using Quad SIMS for precise RTA monitoring not only overcomes the limitations of traditional Rs measurement but also enhances the quality and yield of semiconductor devices. This innovative approach is set to become a standard practice in the industry, paving the way for more efficient and effective semiconductor production.

The CAMECA Quad SIMS 4550

Use of the CAMECA Quad SIMS 4550 in the study is significant for several reasons:

1. High Sensitivity and Precision: The CAMECA SIMS 4550 provides exceptional sensitivity and precision in detecting secondary ions, which is essential for accurately measuring the depth profiles of dopants in silicon wafers. This capability is critical for monitoring the fine details of RTA processes.
2. Enhanced Signal-to-Noise Ratio: The study utilized a primary O2+ beam at 500 eV and an incidence angle of 30°, significantly boosting the signal-to-noise ratio. These settings ensure that the boron peak concentration measurements are accurate and dependable.
3. Stable Sputtering Rate: Maintaining a sputtering rate of approximately 0.1 nm/s, the?SIMS 4550 ensures consistent and repeatable measurements. This stability is critical for detecting minor changes in anneal temperatures and is crucial for optimizing semiconductor device performance and yield.
4. Accurate Data Collection: Employing a 3-second counting time for 11B+ and averaging the top three data points enhances measurement accuracy. This technique minimizes random noise, leading to more reliable data outcomes.
5. Quantitative Depth Profiling: The instrument's ability to conduct quantitative depth profiling of boron allows for the normalization of peak concentrations to a reference wafer. This process improves the long-term repeatability and precision of RTA monitoring, leading to better overall process control.

Overall, using the CAMECA Quad SIMS 4550 allowed the researchers to achieve the necessary sensitivity, precision, and repeatability required for effective RTA monitoring, ultimately leading to improvements in semiconductor device yield.

For More Information

For more details about the study by Dr. Jiang and the team at NPX, refer to the original article: Z. X. Jiang et al., "Monitoring of rapid thermal anneal with secondary ion mass spectrometry," J. Vac. Sci. Technol. B 42, 034007 (2024); doi: 10.1116/6.0003599. Published online 20 May 2024.

For more information about the CAMECA Quad SIMS 4550, visit Quadrupole SIMS Dopant Depth Profiling and Thin Layer Analysis in Semiconductors.