Connected Chemistry

The technology geek at the heart of an analytical scientist is naturally driven to automate repetitive tasks performed in laboratories (1). As early as the 1950s, Leonard Skeggs introduced one of the first automated analytical techniques, continuous flow analysis (CFA) (2). In this technique, chemical reagents are introduced to the sample plugs, which move along a tube, separated with air bubbles. Chemical reactions occur as the train of plugs advances from the tubing inlet toward a detector. With this advance, variability of the analytical results became independent of sample processing by human analysts. Later, to address the limitations of CFA, flow-injection analysis was introduced (3), eventually leading to the development of microfluidic systems (lab-on-a-chip) (4).

Today, automated systems can perform high-throughput chemical analysis, as well as record and process experimental data more accurately and efficiently (5). With the introduction of artificial intelligence to chemistry research, analytical chemists can now obtain deeper insights from the colossal amounts of data recorded by the laboratory equipment (6). Algorithms assist in mass spectral interpretation and prediction, chromatographic peak picking, structural elucidation of molecules, reaction product prediction, experimental planning, and many more (7). Moreover, the open-source tools now available, such as microcontrollers, single-board computers, 3D- printing and software accessories provide great fodder to an analytical chemist who is hungry for innovation (1) (8) (9) (10). Here at the Urban Laboratory at the National Tsing Hua University, Taiwan, we make use of these inexpensive open-source tools to automate our analytical procedures.

For example, in one study, continuous and automated dilution of complex samples was achieved using an Arduino-based control unit by plug-volume modulation, a technique involving continuous introduction of short plugs of a sample, separated with short plugs of a solvent (11). Automated one-shot multipoint calibration of analytical detectors is possible with this simple setup. Another project led to the development of a dual robotic arm “production line” for sample processing and introduction to a mass spectrometer (12). The robotic arms allow the analysis of multiple samples without human interaction. Moreover, a system built around a miniature single- board computer was used to determine optimum sample flow rates for mass spectrometric analysis (13).

The analytical chemist’s dream of a “smart laboratory” is not far from reality. The “Internet of Chemical Things” (15) can now connect, interact and exchange data among various laboratory equipment – keeping analytical chemists, chemical manufacturers and technology developers interconnected.