Investigations into femtosecond laser hybrid manufacturing for 3D power sources

• Main Author: Tham, Nicholas Cheng Yang
• Other Authors: Murukeshan Vadakke Matham
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Materials::Organic/Polymer electronics; Engineering::Manufacturing::Flexible manufacturing systems
• Online Access: https://hdl.handle.net/10356/152012
• Institution: Nanyang Technological University

The changing energy needs of a new generation of smart wearables and Internet-of-Things (IoT) devices present a growing challenge that cannot be met by existing technologies. These small and compact devices come with tight form factors and energy requirements demanding higher performance and site-specific energy sources such as micro-batteries (MB) and micro-supercapacitors (MSCs). Current MBs and MSCs, however, do not yet meet these energy requirements. The 2D design of these power sources limits either the amount of energy stored (energy capacity) or the rate of energy release (power capacity). Surpassing these limitations require moving beyond these 2D designs, made possible by exploiting the third dimension. Compared to 2D MBs and MSCs, a combination of high power and energy capacities for MBs and MSCs can be achieved with 3D structures. However, the fabrication of 3D structures for MBs and MSCs faces major challenges. Existing fabrication strategies, which include 3D structures formed through template-less nanostructure growth or in-situ material deposition are still limited by high costs and complexity. Therefore, there is a definite need for cost-effective and scalable alternatives to fabrication. Two such alternatives are femtosecond direct laser writing (fsDLW) and femtosecond laser interference lithography (fsLIL). fsDLW is an umbrella term where femtosecond (10-15 s) laser pulses are focused and scanned across a material surface to alter, deposit, or otherwise remove material through unique laser-material interaction phenomena. These phenomena allow for unprecedented control over in-situ material manipulation to form unique 3D features at different scales. One of these 3D features is randomly arranged 3D porous graphene (PG) flakes. PG can be formed directly from fsDLW in a single step, eliminating costs associated with multiple process steps. fsDLW PG therefore vastly simplifies the fabrication of high-performance 3D MSCs and MBs. fsLIL, in contrast, is a patterning process where femtosecond pulses interfere to form a variety of regular periodic interference pattern to pattern different materials. One area of research in this thesis is the direct patterning of optically transparent substrate materials such as PDMS with fsLIL. High aspect ratio structures can thus be formed directly in the bulk of optically transparent materials, enabling high aspect ratio 3D micro-batteries without the need for complex and expensive photolithographic steps. Therefore, this thesis will explore how high energy and power density 3D MB and MSC structures could be fabricated using fsDLW and fsLIL. These developments are supported by in-depth analysis of formation mechanisms and material characterisation for 3D MBs and MSCs. The first major contribution of the research work presented in this thesis is the research and development of instrumentation, methodologies, and set-ups to achieve better stability and control over fsDLW and fsLIL. Improvements in femtosecond laser stability are first achieved through the in-line monitoring of femtosecond laser pulses, such as pulse repetition rate, fluence and number of pulses during fabrication, reducing the impact of laser fluctuations and drift on fsDLW and fsLIL. These improvements then form the foundation for development of a fsDLW methodology to uniformly fabricate PG. This methodology consists of three key steps. First, in the design phase, acceleration and deceleration ‘wings’ are added to designs to reduce the impact of uneven fluence from mechanical movement of the galvano-scanner mirrors. Next, during the sample preparation phase, a temporary adhesive layer is developed to minimise any curling or warping of precursor materials during fsDLW. Lastly during the set-up preparation phase, a method to repeatably control sample height is developed. Through these three key steps, uniform and repeatable PG is demonstrated. The yield and repeatability of PG fabrication is thus improved, minimising the influence of unwanted factors during fabrication. This methodology is demonstrated with a custom lab-built fsDLW platform consisting of a femtosecond laser source coupled with in-line monitoring and a two axis galvano-scanner. A fsLIL platform incorporating the same in-line monitoring is developed in parallel. Strict design considerations due to the low temporal coherence of femtosecond pulses and the requirement for high fluence patterning are considered. From these design considerations, a two beam fsLIL platform is developed for low angle (< 10°) large pitch (> 1 μm) periodic patterning in 1D. This set-up includes the development of a method to quickly calibrate the fsLIL set-up for patterning. By doing so, optical path difference (OPD) is kept within the coherence length of the femtosecond pulses for patterning. Together, these improvements in instrumentation, methodology and set-up enable consistent, repeatable, and controlled fsDLW and fsLIL, which is critical for any deeper research. Predicting the formation of PG from fsDLW is critical to optimise PG for use in 3D power sources. Therefore, the next step in this investigation is the modelling of femtosecond laser-material interaction to predict PG formation. An analytical temperature model is developed to predict the temperature rise induced by impinging femtosecond pulses in a material’s cross-section. This model considers laser parameters and changing material properties with temperature. Predicting PG formation is achieved from a predefined top surface boundary temperature (T_grphtn = 1775 K) taken to be the minimum temperature required for PG formation. The temperature model is experimentally validated by comparing the predicted formation of PG with actual PG formation with the fsDLW platform and methodology discussed previously. Material changes in irradiated regions are identified through Raman spectroscopic analysis and optical characterisation. Experimental validation reveals that the analytical model can accurately predict and function as a proxy for the formation of PG above 1775 K. Subsequently, these optimised PG structures are demonstrated in 3D MBs. The performance of PG as an active anode for 3D MB is then characterised and analysed. An initial energy capacity of 0.37 mA h cm-2 is observed at 2 C discharge rate, and even at 17 C, up to 0.03 mA h cm-2 is still maintained for 1000 cycles at close to 100% Coulombic efficiency. These results showcase the potential for PG formed from fsDLW in 3D MBs. The model developed in this research work lays the groundwork for a subsequent exploration into in-situ integration of MSCs with fsDLW and PG. Integrating site-specific 3D power sources into thermally sensitive materials such as paper becomes possible using the same temperature model. With some minor improvements, the same model can be used to predict the temperature rise at both the top and bottom surfaces of PI tape. A thermally localised laser graphitisation (LLG) regime can then be identified to achieve in-situ PG integration. Parameters for LLG are determined from the model by first predefining a lower boundary for thermal damage at this bottom surface (348 K) in combination with the same surface boundary temperature (1775 K). With these parameters, localised PG (LPG) from PI can be fabricated in-situ without causing damage to the electronics substrate below from excessive heating. Thermochromic paper, which changes colour above a defined temperature is used to validate the model. LLG is critical to integrate site-specific power sources into thermally sensitive electronics. LPG MSCs are demonstrated and characterised on paper together with other useful LPG sensors for humidity. LPG MSCs demonstrate high capacitive performance, with 0.428 mF cm-2 at scan rates of 10 mV s-1, and even an increase in scan rates to 10,000 mV s-1 only reduces performance by two-thirds, falling in line with reported PG MSCs. LPG humidity sensors are found to repeatably detect humidity across a range of 4.2% to 7.64% for up to three cycles, with a maximum deviation of σ = 0.64. Together, these results demonstrate LPG’s usefulness for high performance, self-powered graphene paper electronics. Lastly, the bulk structuring of optically transparent materials through fsLIL for 3D MBs and MSCs is studied. The investigation also targeted towards the fabrication of 3D structures within such materials. The potential ability of these structures is demonstrated for alternate applications in optics. To demonstrate these alternatives 3D volume phase gratings are holographically patterned in unmodified optically transparent polydimethylsiloxane (PDMS). The previously discussed fsLIL platform and methodology is used to form Δn of 1.95 × 10-4 in unmodified PDMS with 1 μm periodicity across a thickness of 180 μm. These VPGs are then demonstrated as active beam steering units and tuneable spectroscopic optical elements. Results from this research work will be critical for further research into the direct patterning of substrate materials for 3D power sources. The major findings, original contributions and future work proposed in this thesis are envisaged to contribute towards the cost-effective and scalable fabrication of high performance and high-capacity 3D power sources for the next generation of smart wearable devices.