Non-invasive near-field THz imaging using a single pixel detector
Stantchev, Rayko Ivanov
Thesis or dissertation
University of Exeter
Creative Commons Attribution License 4.0 (CC BY)
The terehertz radiation potentially has many interesting applications. From air port security, non-destructive evaluations of electronics and space shuttle panels, to non-ionizing photon energies with the potential to detect cancer growths and quality control of pharmaceutical tables, the list of potential applications is vast as shown in chapter 1. However, there is a lack of cheap, robust and efficient THz sources, detectors and modulators. Further, the long wavelengths render micron sized details unseeable with far-field imaging techniques. This has rendered most imaging applications unusable in the real world. This thesis is based around demonstrating an imaging technique that uses a near-field THz modulator to obtain sub-wavelength images. There are five distinct experimental demonstrations that show the full capacity of the imaging technique developed here. Chapter 2 gives an outline of the background physics knowledge needed to understand the entirety of the thesis. An outline of the mathematics used for modellingis given in the latter part of the chapter as well. Chapter 3 gives a background on the THz generation and detection techniques used in our THz-TDS system, optical rectification and electro-optic sampling in ZnTe. Further more, our system is capable of photoexciting a sample in conjunction to it being probed with a THz pulse. For the most part, we photoexcite a silicon wafer in order to use its photoconductive properties to modulate our THz pulse. Our photoexcitation pulse is spatially modulated, via a digital micromirror device, which in turn spatially modulates our THz pulse. This patterned THz pulse can then be used with a single-element detector to perform imaging. How to do this and the type of patterns needed is described in the latter part of chapter 3. Chapter 4 is the first demonstration that photo-induced conductivity in silicon can be used to manipulate evanescent THz fields for sub-wavelength imaging. For this, we imaged a 1D sub-wavelength slit and were able to obtain the slit profile with 65μm (λ/6 at 0.75T Hz) resolution. Chapter 5 demonstrates what limits the resolution in our imaging system. Namely, the distance which the patterned THz pulse propagates to the object from where itwas spatially modulated. We demonstrate 9μm (λ/45 at 0.75T Hz) resolution using an ultra-thin (6μm) silicon wafer. At such sub-wavelength scales polarization becomes an important factor. We show how one can use polarization in order to detect 8μm breaks in a circuit board hidden by 115μm of silicon. Chapter 6 concerns itself with showing how noise affects our images. Further more, our imaging system is compatible with compressed sensing where one can obtain an image using fewer measurements than the number of pixels. We investigate how different under-sampling techniques perform in our system. Note under-sampling at sub-wavelength resolutions, as is done here, is rather unusual and is of yet to be demonstrated for other part of the electro-magnetic spectrum. Chapter 7 shows that one does not need to photoexcite silicon. One can in principle illuminate any material, hence we photoexcite graphene with our spatially modulated optical pulses. This allows us to obtain the THz photoconductive response of our graphene sample with sub-wavelength resolution (75μm ≈ λ/5 at 0.75T Hz). We compare our results with Raman spectra maps. We find a clear correlation between THz photoconductivity and carrier concentration (extracted from Raman). Chapter 8 exploits the full capacity of our imaging system by performing hyper-spectral near-field THz imaging on a biological sample. For this, in our imaged field of view, we measured the full temporal trace of our THz pulse at a sub-wavelength spatial resolution. This has allowed us to extract the frequency dependent permittivity of our biological sample, articular cartilage, over our spectral range (0.2-2T Hz). We find the permittivity to change on a sub-wavelength scale in correlation with changes in the structure of our sample. However, the permittivity extraction procedures that have been developed make a far-field approximation. We mathematically show the presence of the THz near-fields to render the long wavelength spectral parts of our extracted permittivity to be wrong. Chapter 9 is where we conclude and point out the main problem that needs to be addressed in order to make the measurements presented here more accessible to others. Namely, the cost of the laser system powering the THz-TDS and how to further reduce the acquisition time.
QinetiQ iCase award 12440575
PhD in Physics