Surface engineering of zinc phthalocyanine organic thin-film transistors results in part-per-billion sensitivity towards cannabinoid vapor

Phthalocyanines (Pcs) and their derived metal complexes (MPcs) are macrocyclic organic compounds with a variety of interesting spectral and electronic properties. First reported in 1907, Pcs were highlighted for their excellent stability and brilliant color, finding extensive use as dyes and pigments, however today, Pcs are more often deposited as charge transport layers within organic photovoltaics and organic thin-film transistors (OTFTs). Pc-based OTFTs have been demonstrated as sensors for a variety of liquid and gas sensing applications, including our groups’ previous demonstration of ratiometric detection and differentiation of Δ⁹-tetrahydrocannbinol (THC) and cannabidiol (CBD).The dominant cannabinoids in Cannabis sativa consumer products, THC and CBD are used for therapeutic and recreational purposes, however, as THC and CBD elicit different pharmacological effects, accurate, low-cost identification is of interest to industry, law enforcement, and consumers. Commercially, identification can be accomplished with higher order lab techniques, though they can be limited in throughput and impractical for companies or individuals with limited resources. In our previous works, we examined a variety of single-use Pc-based OTFT sensors and established that the observed sensing responses were due to a combination of electrochemical interactions and physical effects on thin-film crystallinity. Ultimately, we established a relationship between OTFT sensing characteristics and analyte induced physical thin-film effects, which highlighted the importance of sensor material selection.

An advantage of Pcs as a semiconducting sensor layer is the ease by which modifications can be made to the central metal (Figure 1) which allows tuning of the electronic, colorimetric, and sensor properties. Thus, we began by examining the OTFT-sensitivity of a series of copper (Cu) and zinc (Zn) Pcs with varied degrees of peripheral fluorination, and by extracting electronic characteristics from the transfer curves, established ZnPc as the most THC-sensitive material.

Figure 1. Effects of THC vapor on Pc OTFT electrical characteristics and XRD spectra. (a.) CuPc, F₁₆-CuPc, ZnPc, F₄-ZnPc, or F₁₆-ZnPc OTFTs were exposed to 4 ppm THC vapor over a period of 90 seconds. Boxed regions represent a sign change from the median response. Characteristic transfer and XRD spectra demonstrate sampled regions. Dashed lines represent regions of the curve within which slope was measured. (b.) General Pc structure and (c.) a table of the Pc’s studied. Mobility was calculated from the saturation region of the transfer curves while defect density and voltage threshold were estimated from the subthreshold slope, with average values taken from 20 devices.

With our most sensitive material, we sought to further increase OTFT sensor performance by altering Pc deposition conditions to achieve different surface nanostructures (Figure 2a). Examining the electronic properties (Figure 2b,c) we found that the “low” crystallinity ZnPc surface produced the largest response which was similarly reflected in the X-ray diffraction data (Figure 2d,e).

Figure 2. Effect of surface morphology on ZnPc OTFT sensitivity to THC vapor. (a) AFM images, transfer data, and XRD spectra of ZnPc OTFTs with varying degrees of crystallinity (b, d) pre- and (c, e) post- exposure to 400 ppb THC vapor over 90 seconds. Low crystallinity thin-films were deposited at a rate of 1 Å/s and 25°C, medium (med) at 0.05 Å/s and 25°C, high at 0.2 Å/s and 140°C, and very high at 0.2 Å/s and 180°C with a pre-deposited monolayer of p-sexiphenyl (p-6P). Scale bars represent 500 nm.

By reducing the thin-film thickness of the least crystalline films, we further increased OTFT sensitivity and demonstrated a response to THC vapor in concentrations as low as 40 ppb. We also examined the effects of crystal polymorph, finding varied analyte induced physical thin-film effects with altered surface crystal phases. Finally, we examined OTFT sensor response to THC vapor in real-time, allowing a deeper interpretation of the interactions between analyte and sensor surface. Through film engineering of Pc-based OTFT sensors we achieved a 100x increase in sensitivity over our previously developed CuPc-based devices.

In a broader context, sensor development, particularly sensors for small organic molecules such as cannabinoids, is a core component to improve preventative medicine. The future of healthcare, an increased focus on preventative medicine dramatically improves patient outcomes while reducing the societal costs of disease, however, there is a present need for low-cost, highly sensitive devices. This work illustrates the importance of not only material selection, but also thin-film nanostructures, thickness, and polymorphism for improving OTFT sensor implementations.