Laser-induced graphene trending in biosensors: understanding electrode shelf-life of this highly porous material

Laser-induced graphene (LIG) has received much attention in recent years as a possible transducer material for electroanalytical sensors. Its simplicity of fabrication and good electrochemical performance are typically highlighted. However, we found that unmodified and untreated LIG electrodes had a limited shelf-life for certain electroanalytical applications, likely due to the adsorption of adventitious hydrocarbons from the storage environment. Electrode responses did not change immediately after exposure to ambient conditions but over longer periods of time, probably due to the immense specific surface area of the LIG material. LIG shelf-life is seldomly discussed prominently in the literature, yet overall trends for solutions to this challenge can be identified. Such findings from the literature regarding the long-term storage stability of LIG electrodes, pure and modified, are discussed here along with explanations for likely protective mechanisms. Specifically, applying a protective coating on LIG electrodes after manufacture is possibly the easiest method to preserve electrode functionality and should be identified as a trend for well-performing LIG electrodes in the future. Furthermore, suggested influences of the accompanying LIG microstructure/morphology on electrode characteristics are evaluated. Supplementary Information The online version contains supplementary material available at 10.1007/s00216-023-05082-y.


Experimental methods in reference to chapter 2 Reagents
All water used in this study for the preparation of any solutions and for contact angle measurements was of purity grade 1 obtained from a Millipore water system (resistivity > 18 MΩ cm @ 25 °C, TOC < 10 ppb).All chemicals were obtained in analytical grade and used without further purification.

Sample fabrication and storage
LIG electrodes (see photo in Fig. S 1A) were prepared with a flatbed CO2 laser cutter machine (VLS2.30with Pmax = 30 W, λ = 10.6 µm, Universal Laser Systems Inc., USA) using polyimide sheets (Kapton HN500, 125 µm thickness, DuPont, USA) as substrate which were placed in focus of a 125 µm diameter beam (1/e²).Substrates were cut into pieces of suitable size and secured in place with masking tape to avoid movement and rippling.The electrode design in Fig. S 1B was used for voltammetry and, as seen in the sketch, active electrode area, current collectors and contacts to socket all consisted of LIG.For contact angle measurement, an array of six squares with 1 cm length was used, as depicted in Fig. S 1C.The machine settings were 1% power, 10% speed and 1000 by 1000 pulses per inch, unless stated otherwise.In order to electrically insulate current collectors and to confine the size of the electrode area, commercial transparent nail-polish was applied manually (see green area in Fig. S 1B).In comparison to adhesive tape as delimiter, nail-polish was found more resistant to leaking and more convenient to apply.However, the nail-polish is also a likely source of hydrocarbon contaminants which may become significant over long storage times.When comparing storage environments, electrodes were kept without nail-polish and only insulated after being taken out of storage, while allowing 35 min of drying time before measurement.
When the presence of nail-polish was investigated as a factor, glass flasks were used exclusively as storage containers and nail-polish was let to dry for 35 min under ambient conditions before samples were put in.Here, all flasks with samples were kept at a constant 21 °C in a dark cabinet.Over the course of experiments, the initially chosen commercial nail-polish (np#1) suddenly became unavailable, forcing us to switch to another seemingly similar brand (np#2).Later investigation showed that the choice of nail-polish brand had strongly influenced capacitance over storage time.A list of components of np#1 and np#2 can be found in Tab.S 1.

Measurements
Static water contact angles (WCA) were measured optically via the sessile drop method on an OCA 25 system with corresponding software (DataPhysics Instruments GmbH, Filderstadt, Germany).No environmental chamber was used.A volume of 3 µL DI water was suspended from an automatic syringe, then transferred to the sample by vertical movement of the stage, followed by quick acquisition of an image series.To handle LIG samples, they were fixed onto glass slides with double sided adhesive tape and measurements were done timely after removal from storage.LIG is very porous and the contact angles on a hypothetical flat surface of the same chemistry would probably differ from the ones observed here, which are not suitable for further calculations.However, the data obviously still allow conclusions about wettability and can be used to directly compare between samples of a similar porous nature.
For all electrochemical measurements, LIG electrodes were horizontally plugged into a three-contact socket which was fastened to a movable stage and connected to a PalmSens4 potentiostat (PalmSens BV, Netherlands).The material of the LIG contact areas was found robust enough for reproducible electrical connection even when repeatedly re-plugging them into this particular sliding socket, that modification with e.g.silver paste or sticky copper tape was not found necessary and omitted.In every measurement, a priming volume of 50 µL was pipetted center onto the WE, then gently removed with a paper wick and replaced with another 50 µL of the same solution for measurement.The stage was then lifted to contact the droplet with a positionally fixed Ag/AgCl reference electrode (3 M KCl, BASinc., USA), while the LIG pattern designated as "RE" was not used.
After measurement, LIG electrodes were pulled from the socket and discarded.A photo of the electrochemical setup is provided in Fig. S 3. Since the construction of (bio)analytical sensors is our main use-case for LIG electrodes, we were interested in the change in signal from faradaic electrode reactions vs electrode storage-time.As redox probes, we chose potassium(III)hexacyanoferrate (K3[Fe(CN)6]), hexammineruthenium(III)chloride ([Ru(NH3)6]Cl3) and acetaminophen to cover different modes of charge transfer.Stock solutions were prepared in PBS at 10 mM concentration and kept in the fridge at 4 °C in falcon tubes over the course of experiments, a maximum time of 3 months.Working solutions of 1 mM were prepared on each day of measurement by removing a portion of stock, letting it assume ambient temperature and mixing 100 µL stock with 900 µL PBS in PP tubes, followed by mixing and repeated once more for 0.1 mM solutions.A pair of two consecutive CVs were recorded at 50 mV s -1 and 1 mV step, starting in cathodic direction for K3[Fe(CN)6] and [Ru(NH3)]Cl3 and starting in anodic direction for acetaminophen.Each electrode was used for only one measurement and then discarded.

Data evaluation
Figures were created with R/ggplot2.The discussion of Fig. 1 and Fig. 2 in the manuscript mentions the differences in magnitude between the means of the earliest and latest dataset of each group.Doublesided unequal variances t-tests were calculated with R and reported in the form of "t(DF)=T-STATISTIC, P-VALUE".

Fig
Fig. S 1:Photo (A) and sketch (B) of electrode for voltammetric measurements and perspective on sketch of contact angle measurement (C) Immediately after fabrication, samples were stored in one of the following ways (see Fig. S 2): on a small paper shelf above a lab bench, in a segmented, non airtight, polypropylene (PP) box (Hünersdorff GmbH, Ludwigsburg, Germany), fixed with tape on a paper sheet inside a drawer, or inside roundbottom flasks made of borosilicate glass.The flasks were previously cleaned thoroughly with acetone and glass stoppers were fixed without grease.Wide-necked flasks were used to allow for insertion of the samples without bending.All samples were kept in a non-air conditioned chemical laboratory at temperatures between 15 and 28 °C.

Fig. S 2 :
Fig. S 2: Photos of LIG samples stored in different locations/containers

Fig. S 3 :
Fig. S 3: Photo of electrochemical cellThe capacitance was determined by cyclic voltammetry (CV) in phosphate buffered saline (PBS) of pH 7.4 (8.1 mM Na2HPO4, 1.9 mM KH2PO4, 2.7 mM KCl, 137 mM NaCl).Cyclic voltammograms were collected at different scan-rates () between 10 and 400 mV s-1 in a 100 mV-wide potential window centered on the open-circuit potential, the value of which was estimated from the endpoint of a 30 s measurement.The capacitive current ( ) of each directional scan was averaged from the roughly stable second half and 10 s pauses were kept between scans.The slope of  vs  yields the averaged capacitance of the electrode according to  = (see Fig. S 4 for representative data).