A new systematic and quantitative approach to characterization of surface nanostructures using fuzzy logic.

详细信息   

• 作者：Al-mousa ; Amjed A.
• 学历：Ph.D.
• 年：2010
• 导师：Rahman, Mahmudur,eadvisor
• 毕业院校：Santa Clara University
• ISBN：9781109759389
• CBH：3407270
• Country：USA
• 语种：English
• FileSize：2440694
• Pages：155

文摘

Thin films are essential constituents of modern electronic devices and have a multitude of applications in such devices. The impact of the surface morphology of thin films on the device characteristics where these films are used has generated substantial attention to advanced film characterization techniques. In this work, we present a new approach to characterize surface nanostructures of thin films by focusing on isolating nanostructures and extracting quantitative information, such as the shape and size of the structures. This methodology is applicable to any Scanning Probe Microscopy SPM) data, such as Atomic Force Microscopy AFM) data which we are presenting here. The methodology starts by compensating the AFM data for some specific classes of measurement artifacts. After that, the methodology employs two distinct techniques The first, which we call the overlay technique, proceeds by systematically processing the raster data that constitute the scanning probe image in both vertical and horizontal directions. It then proceeds by classifying points in each direction separately. Finally, the results from both the horizontal and the vertical subsets are overlaid, where a final decision on each surface point is made. The second technique, based on fuzzy logic, relies on a Fuzzy Inference Engine FIE) to classify the surface points. Once classified, these points are clustered into surface structures. The latter technique also includes a mechanism which can consistently distinguish crowded surfaces from those with sparsely distributed structures and then tune the fuzzy technique system uniquely for that surface. Both techniques have been applied to characterize organic semiconductor thin films of pentacene on different substrates. Also, we present a case study to demonstrate the effectiveness of our methodology to identify quantitatively particle sizes of two specimens of gold nanoparticles of different nominal dimensions dispersed on a mica surface. A comparison with other techniques like: thresholding, watershed and edge detection is presented next. Finally, we present a systematic study of the fuzzy logic technique by experimenting with synthetic data. These results are discussed and compared along with the challenges of the two techniques.