Trapped ion qubits are a promising technology for building large-scale quantum computers, and numerous high-fidelity quantum logic gates have been demonstrated with this platform. However, most trapped ion quantum gates require the resolution of specific motional sidebands of trapped ions in the adiabatic regime. Thus, the gate speeds cannot exceed the MHz secular frequencies of trapped ions, limiting the practical gate implementations to timescales of ∼ μs. One approach to overcome the intrinsic gate speed limits is fast entangling gates with ultrafast pulsed lasers, also known as fast gates. Fast gates are a non-adiabatic approach, based on state-dependent photon recoil kicks, also known as state-dependent kicks, from a sequence of counter-propagating, resonant, ultrafast Rabi π-pulse pairs applied to two trapped ions. Each π-pulse pair induces an absorption and stimulated emission of a photon, resulting in 2¯hk momentum kicks applied to the ions, with the best efficiency of the momentum transfers possible when using single resonant π-pulses. The spatial displacements of the ions, precisely controlled by the series of the momentum kicks, are the fundamental mechanism for the gates to work as fast controlled-phase gates, whose gate speeds are no longer limited by the secular frequencies of the ions, and thus sub-microsecond gate speeds are feasible. During my research, I demonstrated a key component of fast gates, which is an ultrafast coherent excitation of a single 171Yb+ ion across the 2S1/2 - 2P1/2 transition with a single, near-resonant picosecond pulse at 369.52 nm, with a maximum population transfer of 94(1)%. To achieve the result, the pulsed laser system was vastly modified to accurately tune the central frequency of the pulsed laser across the atomic resonance of 171Yb+. I also devised a novel method to constantly monitor and stabilise the central frequency with a high-resolution diffraction grating spectrometer. The main limiting factors towards the higher-accuracy population transfer with a single pulse are several different types of perturbations, including laser intensity and frequency fluctuations and other experimental noises. I thoroughly characterised the laser properties and noises that could deteriorate the gate fidelity and proposed solutions to stabilise the fluctuations to achieve a higher population transfer efficiency. Based on the noise measurements, I investigated the impact of non-ideal imperfect π-pulses on gate performances and estimated realistic gate fidelity for some gate schemes that could provide a usable gate fidelity in our system. Finally, I conclude this thesis by considering a future architecture with new methods to realise higher-accuracy coherent population transfer, more robust against the laser noises, to achieve sub-microsecond fully entangled two-qubit phase gates.