A pinboard by
Zheng Li

Postdoc, University of Illinois at Urbana-Champaign


Developed a polymer-based colorimetric sensor array for the detection of explosives and biomarkers

Developing a sensitive, rapid and inexpensive sensor for in-situ detection of hazardous analytes has become an urgent need in many aspects, from security screening to food inspection and to health monitoring. The use of colorimetric sensor arrays has proven to be a fast and effective method for liquid or gas analysis where the specificity derives from the pattern of response from sets of cross-reactive and chemoresponsive dyes. Colorimetric sensor arrays have seen successful applications in differentiating diverse families of analytes, ranging from single compounds to composite mixtures, including industrial toxins, foods and beverages, bacteria and fungi. 
As a continuation of this work, a hand-held reader equipped with a color contact image scanner (CCIS) was integrated with a 40-element colorimetric sensor array to target a broad range of home made explosives (HMEs), including various nitro-compounds and two peroxide-based explosives, triacetone triperoxide (TATP) and hexamethylene triperoxide diamine (HMTD). New sensors incorporated include ones that use Fenton reagent (Fe(II) acetate), redox-sensitive dyes (e.g., toluidine and o-dianisidine), transition metal-containing chromogens for nitro-compounds, and optimized Brady’s and Schiff test for ketones. Using multivariate analysis, successful classification results were achieved to group 15 generic explosives into 12 separate groups with the accuracy >99%, as well as 9 variations of TATP and 3 variations of HMTD into 12 separate clusters with the accuracy >96%.
Possible applications of the sensor arrays in medical diagnosis were also explored. As an example, I have successfully quantified several biomarkers in the simulated human urine and blood samples, including trimethylamine (TMA), trimethylamine N-oxide (TMAO) and creatinine with relevance to point-of-care evaluation of renal function and diagnosis of a metabolic disease, trimethylaminuria (TMAU). I made use of porous sol-gel formulations to obtain better sensor response to biomolecules and ideal hydrophobicity of the matrices to minimize the dissolution of the dyes during liquid sensing. Detection limits in aqueous phase are ~2 μM for TMA, ~4 μM for TMAO and ~10 μM for creatinine, all of which are well below the diagnostically significant levels. Sensors show great selectivity towards three targeted analytes against other possible interferents (e.g., amino analogues), indicating its promising applications in medical diagnosis.