伴隨著製程微縮技術與量子物理的蓬勃發展，對於能在微小區域進行物理量的量測手法需求日益增加，本研究主要是以頻帶激發以及機器學習的方式，試圖增進掃描探針顯微技術，並以其中的壓電力顯微鏡與磁力顯微鏡為研究目標。
在頻帶激發部分，我們利用掃頻波結合壓電力顯微鏡，改善在單頻掃描過程中，因共振頻率改變導致的量測誤差，並成功架設基於頻帶激發的壓電力掃描顯微鏡、去靜電鐵電光譜與接觸式表面電位等量測技術。利用基於頻帶激發的鐵電翻轉光譜陣列作為素材，進行機器學習分析，發現在鐵酸鉍71度鐵電疇壁附近，其鐵電翻轉光譜可能有鐵彈性貢獻。
本研究除了利用機器學習解析鐵電材料，也試圖改善磁力顯微鏡。我們使用在輕敲模式下的磁性探針力曲線掃描作為素材，經過機器學習與主成分分析其在磁性樣品上的掃描，發現其分類結果與磁性分布相同。此外，可以根據上述機器學習分析之結果，區分出絕對磁吸力與磁斥力，這是以往磁力顯微鏡所無法達成的。藉由主成分分析的輔助，也可以更加確定利用機器學習的結果為磁性貢獻。並且，此分析方法也成功於磁帶、兩種不同厚度的鈷鐵氧晶體與無磁性的鈦酸鍶上，得到正確的磁性分布。

With the vigorous development of process miniaturization technology, there is an increasing demand for measurement techniques that can measure physical quantities in small areas. This research is mainly based on frequency band excitation and machine learning to try to improve scanning probe microscopy technology.
In the frequency band excitation part, we use the frequency sweep wave combined with the piezoelectric force microscope to minimize the measurement error caused by the resonance frequency shifting, and successfully set up the piezoelectric force scanning microscope, switching spectrum and contact Kelvin probe force microscopy, which are based on the frequency band excitation. A band excitation-based ferroelectric switching spectrum array is treated as a dataset for machine learning analysis, and result shows that the ferroelectric flip spectrum may have a ferroelastic contribution at near the 71-degree ferroelectric domain wall of bismuth ferrite.
In addition to analyze ferroelectric materials by machine learning, the another goal of this research is to improve the magnetic force microscope. We use the magnetic probe force curve scan in tapping mode as the material, and scan it on the magnetic sample through machine learning and principal component analysis. We found that the classification result is the same as the magnetic distribution. Furthermore, the absolute magnetic attraction and magnetic repulsion can be distinguished based on the results of the above-mentioned machine learning analysis, which cannot be achieved by the previous magnetic force microscope. With the aid of principal component analysis, it can also be more certain that the results of machine learning contribute to magnetic properties. In addition, this analysis method is also successful on magnetic tape, two different thicknesses of cobalt ferrite crystals and non-magnetic strontium titanate.

1. Jesse, S., Baddorf, A. P. & Kalinin, S. V. Switching spectroscopy piezoresponse force microscopy of ferroelectric materials. Appl. Phys. Lett. 88, 3, doi:10.1063/1.2172216 (2006).
2. Balke, N. et al. Exploring Local Electrostatic Effects with Scanning Probe Microscopy: Implications for Piezoresponse Force Microscopy and Triboelectricity. ACS Nano 8, 10229-10236, doi:10.1021/nn505176a (2014).
3. Jesse, S. & Kalinin, S. V. Band excitation in scanning probe microscopy: sines of change. J. Phys. D-Appl. Phys. 44, 16, doi:10.1088/0022-3727/44/46/464006 (2011).
4. Agar, J. C. et al. Machine Detection of Enhanced Electromechanical Energy Conversion in PbZr0.2Ti0.8O3 Thin Films. Adv. Mater. 30, 9, doi:10.1002/adma.201800701 (2018).
5. B.D.Cullity, S.R.Stock, Prentice Hall, “Elements of X-ray Diffraction “,New Jersey(2001)
6. Berdard Jaffe, William R. Cook and Hans Jaffe, “Piezoelectric Ceramics”, Academic Press Inc., London, (1970).
7. Toshio Mitsui, Itaru Tatsuzaki and Eiji Nakamura, “An introduction to the physics of ferroelectrics”, Gordon and Breach Science Publishers, New York, (1976).
8. 謝孟勳, (110)取向之鋯鈦酸鉛薄膜的鐵電/鐵彈結構分析與操控, 成功大學, 碩士論文, (2019)
9. William D. Callister, Jr., “Material Science And Engineering”, John Wiley & Sons Inc., (London).
10. 蔡宜哲, 以光調控鈷鐵氧體磊晶薄膜之磁性, 成功大學, 碩士論文, (2018)
11. R. Liithi, H. Haefke, K.-P. Meyer, E. Meyer, L. Howald, and H.-J. Gijntherodt, “Surface and domain structures of ferroelectric crystals studied with scanning force microscopy”, J. Appl. Phys., 74, 12, (1993).
12. 王洸富, 屏蔽電荷對180度域壁成核動態機制之影響”, 成功大學, 碩士論文, (2010).
13. 曾賢德、果尚志,“奈米電性之掃描探針量測技術”, 物理雙月刊(廿五卷五期), (2003).
14. Magnetic Force Microscopy (MFM) Applicable to Dimension™ Series and MultiMode™ Systems, Digital Instruments, 1996.
15. https://dotblogs.com.tw/dragon229/2013/02/04/89919
16. https://stats.stackexchange.com/questions/259806/anomaly-detection-using-pca-reconstruction-error