永井研究室では最先端のAPTやFIB、TEMといった機器を扱いながらさまざまな研究を行っています

H. Takamizawa¹, K. Inoue^1,2, Y. Shimizu¹, T. Toyama¹, Y. Nagai¹, N. Okada³, M. Kato³, H. Uchida³, F. Yano⁴, T. Tsunomura⁴, A. Nishida⁴ and T. Mogami⁴


Acknowledgments:
New Energy and Industrial Technology Development Organization ( NEDO )
Grants-in-Aid for Scientific Research of the Ministry of Education, Culture, Sport, Science and Technology, Japan ( No. 17002009, 20760212, and 21246142)

The characteristic variability of metal-oxide-semiconductor field-effect transistors (MOSFETs) is becoming a bottleneck in achieving device scaling. The major factor of characteristic variability is believed to be random dopant fluctuation (RDF) in the channel. In this work to understand the dopant behavior in MOSFET structure directly, laser-assisted atom probe tomography (laser-APT) is applied to gate patterned MOSFET structure samples. We found boron density in a channel underneath the gate was locally higher than in neighboring channel region in n-type MOSFET.

Purpose of this work
- To obtain the three dimensional elemental maps of MOSFET which has gate patterned sample, and analyze the dopant distribution in gate and channel regions.
- To investigate the effect of fabricating the gate pattern and source/drain extension on channel dopant distribution.

Sample preparation
Following three-type samples of n- and p- type MOSFETs are prepared in order to study the dopant distribution in gate patterned MOSFET structures.

1. Gate patterned MOS structures with channel, source/drain extension and Halo implantation
2. Gate patterned MOS structures with channel and source/drain extension implantation without halo (to study the effect of source/drain extension implantation on the channel dopant distribution)
3. Only gate patterned structures with channel implantation

Table.1 A list of prepared samples


Needle specimens for APT are fabricated by FIB/SEM dual beam systems.

1. Selection of the analysis region.
2. Scraping out around the site.
3. Lift out using manipulator.
4. Mounting the sample on micro flattop tip.
5. Sharpening the specimen using annular milling patterns.
6. Needle specimen for APT measurement.

[Remarks]

• Region ① and ②
  Segregation of P atoms on grain boundaries was observed.
• 　Region ③
  We observed extension of source/drain and halo structures clearly, and a high concentration of B atoms was observed at the edge of source/drain extension.
• 　Region ④ and ⑤
  Distribution of B atoms seems to uniform in the gate.
• 　Region ⑥
  We observed extension of source/drain and slightly observed P atoms implanted as halo.

n-MOS(ii),(iii), p-MOS(ii),(iii)

• [n-MOSFET]
  -B density in the channel region underneath the gate was locally higher than that in neighboring channel region. This tendency is more remarkable in the samples without source/drain extension.
• [p-MOSFET]
  -There was no significant difference between the samples with and withoutsource/drain extension.