Transition metal oxides based tunable plasmonic materials for biological applications

Plasmon resonance is commonly seen in noble metal systems, showing enhanced absorption and scattering resonances in specific wavelengths that can be utilized in many fields, including photovoltaics, surface-enhanced spectroscopies, sensors and possibly optical communications. However, the plasmonic responses of such systems are permanently locked in and cannot be actively controlled once they have been engineered through their inherent parameters and specific morphologies. Such a critical challenge can be potentially solved by introducing a new class of plasmonic materials based on degenerately doped semiconductors, in which their free charge carrier concentrations and plasmon resonance frequencies can be tuned by the concentrations of dopants in the host. Transition metal oxides (TMOs) have been identified as suitable hosts for degenerate doping using various synthesis methods. Their plasmon resonance wavelengths have been reported to appear in infrared (IR), near-infrared (NIR) or even visible light range. However, plasmonic properties of some current doped TMOs have many limitations: (a) tuning the plasmonic is limited due to the insufficient capabilities to accommodate many different levels of free charge carriers; and (b) the tuning process is not reversible because of the intrinsic properties of the doping processes, which makes the concept of "active plasmonic control" challenging to achieve. The author of this thesis hypothesizes that the incorporation of TMOs with intercalatable capability and high surface affinity to charged particles can introduce the possibility of "active plasmonic control". Therefore, the functional nano-plasmon based devices can be reversibly tuned or switched. The author demonstrates this concept by using molybdenum trioxide (MoO3) as the host TMO and precisely intercalated H+ ions through the hydrothermal method. These H+ ions are eventually accommodated within the MoO3 crystal structure via the formation of weak bonds with oxygen atoms. It is observed that the concomitantly brought electrons in the process of doping behave similarly to Drude-model-like free electrons, which enter the electronic band structure of the host and transform it from semiconducting to quasi-metallic depending on the doping level. In addition, the morphologies change from nano-belt to single-unit cell thick nanoflakes upon the incorporation dopants. The accordant plasmon resonance behavior was observed to shift from Terahertz to NIR and finally within the visible light range. The achieved degenerately doped HxMoO3 (0 < x < 2) presents great potential in sensing applications due to the facile H+ intercalation/de-intercalation movement in a biochemical event. In this thesis, the author proves the ultra-high sensitivity of HxMoO3 towards biochemistry reaction events by using a representative enzymatic glucose sensing model. The facile H+ ions provide exceptional sensitivity and fast kinetics to charge perturbations during enzymatic oxidation. The optimum sensing response is found at H1.55MoO3, achieving a detection limit of 2 nM glucose at the visible light range, when the biosensing platform is adapted into an LED-photodetector setup. The performance is superior to all previously reported plasmonic enzymatic glucose sensors, providing a great opportunity in developing high-performance biosensors. After investigating the tunable plasmonic behavior of HxMoO3 with ultra-high sensitivity towards biochemical events, the author further explores the possibility of combining degenerately doped HxMoO3 with other sensing platforms to expand their sensing capabilities. The author has pioneered the integration of mass-producible silicon photonics platform with HxMoO3 for ion sensing capability. In particular, the author utilizes a silicon-based micro-ring resonator (MRR) coupled with a two-dimensional (2D) re-stacked layers of NIR H0.3MoO3. When the 2D plasmonic H0.3MoO3 layer interacts with ions from the environment, it is hypothesized that a strong change in the refractive index results in a shift in the MRR resonance wavelength and simultaneously the alteration of plasmonic absorption leads to the modulation of light transmission power within the MRR. Therefore, such a novel ion sensor has a dual-sensing output in respect to the light magnitude and phase. The author demonstrates the proof-of-concept of such a hypothesis via a typical pH sensing model. The H0.3MoO3-MRR sensor is exposed to the solutions at different pH values across from 1~13, showing an up to 7 orders improvement on sensitivity per unit area compared to those of other reported optical pH sensors. The platform offers great potential for ultra-sensitive and robust measurement of changes in the ionic environment, generating new modalities for on-chip chemical sensors in the micro/nano-scale. In the final stage of this PhD thesis, the author investigates the possibility of using the degenerately doped HxMoO3 for detecting cancer cells. There is enormous research interest in the label-free detection at a single cancer cell level from grouped biological samples. The distinct oxidative stress behaviors and Raman fingerprints of different cell lines guide a significant pathway of cellular signaling. Here the author realizes a HxMoO3 based label-free dual-functional detection of individual THP-1 cancer cells immobilized in a microfluidic chamber. The cells are incubated and coated with the degenerately doped plasmonic HxMoO3. The concentration-optimized glucose buffer solution is then utilized as the drug for inducing the oxidative stress efflux from the cells gradually. The author demonstrates this unique cancer cell metabolism behavior by monitoring the plasmonic absorption variation from a single THP-1 cancer cell in real-time for at least 60 min, while the healthy human white blood cells (WBCs) are used as the control. A almost 20 times difference on the optical absorbance is observed from the isolated single THP-1 cancer cell compared with those of WBCs. Furthermore, the plasmonic HxMoO3 featured with surface-enhanced Raman scattering for label-free individual cancer cell detection has been also demonstrated. The author confirms the Raman signal enhancement factor of up to three orders of magnitude. To detect and differentiate individual THP-1 cancer cells and WBC, the Raman signal data are obtained from the HxMoO3 coated cell surface and processed using the analytical algorithm principle component analysis-support vector machines (PCA-SVM), which shows up to 98% successful rate in determining cancer cell from samples with unknown mixed ratio of healthy and cancer cells. In summary, the author successfully demonstrates several significant discoveries in the course of this PhD research, developing concepts based on plasmonic degenerately doped MoO3 and revealing its extraordinary prospective in biological sensing at the transducing platforms of open-path optics, integrated photonics and vibrational Raman spectroscopy. It is expected that the outcomes of this PhD research will contribute to the development of high-performance optical biosensors for the aspects of clinical diagnosis and human health monitoring.