Xiyuan Liu

Monitoring toxic heavy metals in complex, real water sources is essential for safe water supplies. Current sensors lack the sensitivity, selectivity, and limits of detection (LOD) needed to detect heavy metals such as cadmium (Cd²⁺) in real water samples with minimal preparation. Here, we demonstrate a novel ternary nanocomposite-coated carbon electrochemical sensor that detects Cd²⁺ with the highest sensitivity (14.71 ± 0.3, 6.94 ± 0.8 μA·ppb⁻¹·cm⁻²) and the lowest LODs (0.06 ± 0.02, 0.61 ± 0.10 ppb) in spiked deionized and lake waters, respectively. Sensor fabrication optimization included electrochemical polishing of carbon screen-printed electrodes, surface modification with a 1 mg/mL bismuth-reduced graphene–molybdenum disulfide nanocomposite ink, and rapid heating and cooling at 75 °C for 3 min. These steps, along with square wave anodic stripping voltammetry (SWASV) deposition potential identification, produced distinct detection peaks and high reproducibility (8% RSD). We further investigated the SWASV parameters, specifically frequency (5 to 50 Hz), amplitude (1 to 100 mV), and step potential (5 to 70 mV), using a fast causal inference algorithm, which resulted in a 233-fold increase in Cd²⁺ sensitivity. We found that the LOD and the full width at half maximum of the Cd²⁺ peak are dependent on frequency, while sensitivity and peak resolution depend on amplitude and step potential. Under optimal detection conditions of 35 Hz, 60 mV amplitude, and a 10 mV step potential, the sensor showed adequate sensitivity and LODs, even in the presence of 10-fold concentrations of Zn²⁺, As³⁺, and Cu²⁺ as interferents, confirming its suitability for real-world water-quality field testing.

The Part-Of-Speech tagging is widely used in the natural language process. There are many statistical approaches in this area. The most popular one is Hidden Markov Model. In this paper, an alternative approach, linear-chain Conditional Random Fields, is introduced. The Conditional Random Fields is a factor graph approach that can naturally incorporate arbitrary, non-independent features of the input without conditional independence among the features or distributional assumptions of inputs. This paper applied the Conditional Random Fields for the car review word Part-Of-Speech tagging and then the feature extraction, which can be used as an input to an opinion mining system. To reduce the computational time, we also proposed applying the Limited-memory BFGS algorithm to train the Conditional Random Fields. Furthermore, this paper evaluated the Conditional Random Fields and the classical graph approach using the car review dataset to demonstrate that the Conditional Random Fields have a more robust result with a smaller training dataset.