Cross-sectional imaging of semiconductor devices using nanometer scale point contacts

dc.contributor.authorNxumalo, Jochonia Normanen_US
dc.date.accessioned2007-05-17T12:36:32Z
dc.date.available2007-05-17T12:36:32Z
dc.date.issued1998-05-01T00:00:00Zen_US
dc.degree.disciplineElectrical and Computer Engineeringen_US
dc.degree.levelDoctor of Philosophy (Ph.D.)en_US
dc.description.abstractIn order to study semiconductor dopant profiles a novel technique based on localized contact resistance measurements has been designed and implemented. Scanning Resistance Microscopy (SRM) is a nanometer scale technique for performing two-dimensional pn junction delineation and carrier profiling on semiconductor surfaces. It utilizes a sharp conducting probe tip that is biased and raster scanned in contact with the sample surface while an ohmic back contact provides the circuit return path. By monitoring the current flowing across the tip-surface interface while scanning the sample beneath the probe this technique performs localized resistance measurements over a semicanductor surface. These measurements are used to delineate between regions of different doping type and magnitude with spatial high resolution (20 nm). SPM technology is used to regulate the contact force between the tip and the surface during a scan in order to attain high spatial resolution consistently. Simulations of the junction formed between SRM tips and silicon substrates suggest that when the tip loading pressure is kept low enough ($\leq$5 GPa) the resistance across the junction is dominated by the contact resistance rather than series spreading resistance. Furthermore, I-V spectra taken over uniformly doped silicon substrates in contact with conducting diamond tips demonstrate that these junctions are rectifying. These results show stronger rectification when the diamond tip makes contact to p-silicon compared to n-silicon. Capacitance derivative measurements performed on the same samples also indicate a similar asymmetry. By studying the I-V spectrum and comparing with band models for a similar heterojunction used by other researchers we have proposed an energy band model that explains SRM data quite well. In this thesis we present experimental results obtained from imaging cross-sections of a wide variety of semiconductor devices. These results demonstrate the capability of SRM to perform two-dimensional imaging on device cross-sections providing simultaneously surface topographic and resistance profiles. We also demonstrate tha by using conducting diamond (instead of metal) tips the SRM spatial resolution is improved significantly and resistance profiles of MOSFET cross-sections reveal source drain extensions that are a result of the presence of lightly doped drains. A technique for characterizing semiconductor devices in their normal operation mode has been investigated. Normal operation of a sectioned device is verified by comparing its I-V characteristics before and after sectioning. Imaging is done with a conductive tip that is attached to a high input impedance voltage sensing circuit and used to probe local surface potentials on the device cross-section. This is perhaps the most direct method of evaluating the impact of semiconducror process variation because the potential profiles reflect the activity inside the device during normal operation. Preliminary results demonstrate the ability for this technique to perform 2D surface potential mapping and hence the capability to study the behavior of MOSFET channel formation under different bias condition.en_US
dc.format.extent8610556 bytes
dc.format.extent184 bytes
dc.format.mimetypeapplication/pdf
dc.format.mimetypetext/plain
dc.identifier.urihttp://hdl.handle.net/1993/1440
dc.language.isoengen_US
dc.rightsopen accessen_US
dc.titleCross-sectional imaging of semiconductor devices using nanometer scale point contactsen_US
dc.typedoctoral thesisen_US
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
NQ32010.pdf
Size:
8.21 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
184 B
Format:
Plain Text
Description: