Mesoporous carbon materials are one of the most appealing materials for the fabrication of bioelectrochemical devices such as biosensors and biofuel cells [1–3]. These materials combine high conductivity, high surface area, and excellent biocompatibility, they are excellent for electrodes and matrices for enzyme immobilization. Yang et al. reported an increased temperature and pH stability when glucose oxidase was immobilized on ordered mesoporous carbon . Among the different types of mesoporous carbon materials is MgO-templated carbon (MgOC), and it works excellently as its pore size can be controlled by controlling the size of the MgO template [5, 6]. The effect of the pore size of MgOC on the direct electrochemistry has been investigated for D-fructose dehydrogenase  and bilirubin oxidase (BOD) [8, 9]. Furthermore, biofuel cells (BFCs) fabricated with MgOC ink-modified carbon cloth had a high-power output of 2 mW cm-2  and 4.3 mW cm-2  with glucose dehydrogenase (GDH) and lactate oxidase (LOx) as enzymes, respectively.
A MgOC ink is also the first step in fabricating a screen-printed MgOC electrode. The conductive carbon material in screen-printing inks needs to be dispersed evenly under shear stress applied during printing. An uneven dispersion might lead to a partially brittle electrode (where too little binder is present) and/or a partially increased resistance (where too much binder is present). A higher dispersion can also lead to a higher degree of porosity, as clumping becomes less likely. Small quantities of additives can improve the dispersion of ink without interfering with the conductivity, and thus the quality and reproducibility of the printed electrode. However, although biocompatible and sustainable materials, such as carboxymethyl cellulose (CMC), have been used as dispersants for carbon materials , dispersants have not been considered for MgOC inks for screen-printing.
Screen-printed electrodes are promising for the fabrication of wearable biosensors, especially for healthcare applications [13–15]. Wearable biosensors are receiving significant attention in recent years owing to the trend of a more personalized, real-time healthcare management of patients, as well as a more data-driven, closer monitoring of the physical condition of high-performance professionals, such as athletes and firefighters. Similarly, wearable BFCs are also receiving considerable attention, both as energy harvesters and self-powered sensors [16–19]. As energy harvesters, wearable BFCs collect energy from glucose or lactate contained in bodily fluids to power small devices. Wearable BFCs as self-powered sensors utilize the fact that the power collected from glucose or lactate at any time depends on the concentration of the respective fuel. Self-powered sensors do not require an energy source for the sensing device. Some examples of wearable biosensors and BFCs are integrated into the nose-pad of eyeglasses , microfluidics fabricated from a soft material [21, 22], fabricated on thin flexible film [22, 23], tattoo-type , textile-based [25, 26], and paper-based [27–29].
Paper-based devices also integrate the wicking effect of paper and can work with small sample volumes. del Torno-de Román et al. utilized paper as a fuel delivery system and achieved a power density of up to 37.5 µW cm− 2 with 5 mM glucose . Lau et al. used filter paper for fuel delivery and carbon fiber or carbon nanotube paper for the bioelectrodes and achieved a power density of 35.5 µW cm− 2 with cascade-type 4-electron oxidation of ethanol and 26.9 µW cm− 2 with formate, formaldehyde, and methanol as fuel and three cascade enzymes . Rewatkar et al. also used filter paper for fuel delivery and multiwall carbon nanotube paper for the bioelectrodes and achieved a power density of 46.4 µW cm− 2 with 30 mM glucose as fuel in a 4-cell-series configuration .
Our group has developed several BFCs with electrodes directly printed on Japanese paper. Using Ketjenblack as electrode material and glucose oxidase as anode-enzyme, we achieved a power density of 0.12 mW cm− 2 . Using MgOC as electrode material and lactate oxidase (Lox) as an enzyme, we achieved a power density of 0.113 mW cm− 2 . Using GDH as an enzyme and improving immobilization, we achieved a power density of 0.12 mW cm− 2 . These studies focused mainly on the anode performance. However, with a high-performing anode, the focus needs to shift to improving the cathode, especially in the case of self-powered biosensors, which need to be anode-limited in their performance.
A popular enzyme for constructing biocathode is bilirubin oxidase (BOD). One advantage of this enzyme is its capability for direct electron transfer (DET) [34–36]. As with all DET-type enzyme electrodes, the orientation of the enzyme on the electrode surface is crucial. Compared to a flat surface, a mesoporous surface structure increases the chances of the active site of a randomly oriented enzyme being within DET distance ; a directed orientation would increase the performance of a DET-type biocathode. Lalaoui et al. achieved an ordered immobilization of BOD on carbon nanotubes by utilizing protoporphyrin IX as a “guide” for binding the enzyme . Al-Lolage et al. engineered BOD to have cysteine at a specific site and used that cysteine for a directed, covalent immobilization .
In this study, we used two approaches for improving the performance of screen-printed, paper-based biofuel cells, especially the biocathode. We considered the addition of carboxymethyl cellulose (CMC) as a dispersant to the MgOC ink and investigated its rheological effect. Focusing on the biocathode, we considered bilirubin as a “guide” for immobilizing BOD in an oriented manner.