The primary objective of neuromorphic electronic devices is the implementation of neural networks that replicate the memory and learning functions of biological synapses. To exploit the advantages of electrolyte gate synaptic transistors operating like biological synapses, we engineered electric double-layer transistors (EDLTs) using phosphorus-doped silicate glass (PSG). To investigate the effects of phosphorus on the EDL and synaptic behavior, undoped silicate spin-on-glass-based transistors were fabricated as a control group. Initially, we measured the frequency-dependent capacitance and double-sweep transfer curves for the metal-oxide-semiconductor (MOS) capacitors and MOS field-effect transistors. Subsequently, we analyzed the excitatory post-synaptic currents (EPSCs), including pre-synaptic single spikes, double spikes, and frequency variations. The capacitance and hysteresis window characteristics of the PSG for synaptic operations were verified. To assess the specific synaptic operational characteristics of PSG-EDLTs, we examined EPSCs based on the spike number and established synaptic weights in potentiation and depression (P/D) in relation to pre-synaptic variables. Normalizing the P/D results, we extracted the parameter values for the nonlinearity factor, asymmetric ratio, and dynamic range based on the pre-synaptic variables, revealing the trade-off relationships among them. Finally, based on artificial neural network simulations, we verified the high-recognition rate of PSG-EDLTs for handwritten digits. These results suggest that phosphorus-based EDLTs are beneficial for implementing high-performance artificial synaptic hardware.