Nafion®-coated mesoporous Pd film toward remarkably enhanced detection of lactic acid

Mesoporous metal films can detect biomarkers with high sensitivity. Further coating the mesoporous metal with polymers enhances sensing selectivity by favoring specific biomarkers against other interferents. In the present study, we report the fabrication of a Nafion®-coated mesoporous Pd film to filtrate interferents present in sweat during non-invasive biosensing. By using a Nafion®-coated mesoporous Pd film, lactic acid, a metabolite present in sweat, can be successfully detected with high sensitivity.


Introduction
Non-invasive sensing of biomarkers present in sweat, urine and saliva, for early diagnosis of various diseases and health management, is a research topic which has recently attracted considerable attention. 1 Compared with invasive sensing such as blood collection, non-invasive sensing can reduce damage to and discomfort of the patients, which makes them highly coveted; therefore numerous reports on non-invasive bio-sensing have been released. In particular, sweat can be easily collected from patients and contains various biomarkers which, if detected in time, can contribute to early diagnosis of diseases. For example, tattoo-type, sheet-type and small integrated device-type sensors have previously been reported. 2 These reported devices use enzymes to detect biomarkers from sweat without having to ltrate any interferents. Despite the presence of interfering species in the medium, enzymes can form enzyme-substrate complexes with specic biomarkers, resulting in high selectivity. However, these sensors have some disadvantages in terms of cost and stability (requiring suitable pH and temperature to prevent degradation), which make their practical use less attractive. On the other hand, non-enzymatic sensors, such as mesoporous metal lms and nanoparticles, have recently been reported as a promising alternative. 3 Compared with enzymatic sensors, their selectivity is still relatively low because they cannot form complexes with biomarkers. Under the application of an optimal voltage, however, the inuence of interferents can be reduced. In addition, high sensitivity can be achieved because of the large surface area reacting with the biomarkers.
The typical biomarkers in sweat are lactic acid, glucose, interleukin and other proteins. 4 Among these biomarkers, lactic acid is one of the metabolites whose concentration in sweat is different before and aer exercise. 5 Furthermore, the relationship between the lactic acid content and the metabolic activity is well-known. 6 Therefore, if a biosensor could detect and monitor the exact content of lactic acid in sweat, it could be used for health management purposes.
In this study, we aim to detect biomarkers from sweat aer ltrating out the interferents with a Naon®-coated mesoporous Pd lm. Naon® is a polymer which is typically used for ion-exchange. 7 Mesoporous Pd lms can be fabricated in nonionic solution consisting of P123 surfactant and PdCl 2 metal precursor. The resulting large surface area is suitable for applications in biosensor. Bare mesoporous Pd lms have already a strong potential for biosensing, but further coating the lm with a Naon® layer with mesoporous roughness is critical to remove interferents present in sweat.
Although Naon® has been utilized as a part of electrode materials for biosensing, 8 the hybrid structure consisting of a mesoporous metal combined with a Naon® layer has never been reported. Composites made of mesoporous materials and polymers have previously been reported for drug delivery and photochemical applications, 9 but not for electrochemical biosensing. Integrating ltration and detection in one functional system using mesoporous metal lms is a novel approach. Filtration generally needs complicated process, but Naon® coating is simple and can be used in a "disposable device" technology. In this study, aer ltration by Naon®, lactic acid is monitored by chronoamperometry under voltage which is tuned to minimize the inuence of interferents such as glucose, urea, ammonia and ethanol present in sweat. To benchmark the sensitivity of our Naon®-coated mesoporous Pd lm, bare mesoporous Pd, Naon®-coated nonporous Pd and bare nonporous Pd lms are also compared.

Materials synthesis
The mesoporous Pd lm was prepared according to a method we reported previously. 10 At rst, 80 mg of P123 (poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide), PEO 20 -PPO 70 -PEO 20 ) was dissolved in 2 mL water, followed by the slow addition of 2 mL PdCl 2 (80 mM) under constant stirring. The solution was then sonicated continuously for 20 min. The three-electrode system used to prepare the mesoporous Pd lms consisted of an Ag/AgCl reference electrode, a Pt counter electrode and an Au-coated Si substrate (3 mm Â 3 mm) as working electrode. The electrodeposition was performed at an applied potential of 0.0 V for 600 s. Aer deposition, the Aucoated Si substrate became black. The working electrode was withdrawn quickly from the system and washed with water to remove the remaining P123. The mesoporous Pd lm was coated by dropping 10 mL of Naon® peruorinated resin solution which was then dried under a lamp. These steps were repeated 12 times.

Materials characterization
Scanning electron microscope (SEM) imaging of the mesoporous Pd lms were obtained using a Hitachi FESEM SU-8000 microscope at an accelerating voltage of 5 kV. Wide-angle X-ray diffraction (XRD) patterns were acquired with a SmartLab Rigaku (Cu Ka radiation; operating voltage 40 kV and current 30 mA). Low-angle XRD patterns were obtained using a Rigaku NANO-Viewer (Cu Ka radiation) equipped with a camera length of 700 mm, an operation voltage of 40 kV and a current of 30 mA. All electrochemical measurements were recorded with a CHI 842B electrochemical analyzer (CHI Instruments, USA).

Results and discussion
Both the top-surface and cross-section of the obtained lms are imaged by SEM aer carefully removing the surfactants ( Fig. 1a and S1 †). The SEM images reveal that the pores are uniformly distributed over the entire lm. The periodicity of the pore organization can be determined by low-angle XRD. The clear peak centered at 0.62 corresponds to a pore-to-pore distance of ca. 14.2 nm (Fig. S2 †). From the high resolution SEM image shown in Fig. S3, † the average pore diameter is measured to be ca. 10.9 nm (calculated from over 200 pores). The average wall thickness is estimated to be ca. 3.3 nm which is consistent with the average wall thickness measured from SEM (Fig. S3 †). In the absence of nonionic surfactant, large Pd crystals are formed (Fig. 1b). The cross-sectional SEM image of Naon®-coated mesoporous Pd lm is shown in Fig. 1c. The thickness of the mesoporous Pd lm is approximately 600 nm, from which the growth rate is calculated to be about 100 nm min À1 . Fig. 1d shows the schematic structure of the Naon®-coated mesoporous Pd lm. From Fig. 1c, uniform Naon® coating was conrmed with a thickness about 1.5 mm. Naon® is an ionexchange membrane containing anionic sulphonate groups which can combine with cationic species. The XRD pattern of mesoporous Pd lm contains diffraction peaks located at 40.18 , 46.72 , 68.16 , 82.18 and 86.82 , which can be assigned to the (111), (200), (220), (311) and (222) diffraction planes of Pd with a fcc structure, respectively (JCPDS card no. 05-0681, Fig. S4 †).
The electrochemical response of the mesoporous Pd lms, with and without Naon® coating were measured in NaCl solution (0.4%) to mimic the typical composition of human sweat. The CV measured in NaCl solution is shown in Fig. S5 † and the area under the CV curve for the Naon®-coated lm was smaller than that for the uncoated lm. This can be explained as the anionic Naon® layer has abundant sulphonate groups as well as anionic ions such as chloride, which cannot easily penetrate the membrane. For the bare mesoporous Pd lm, the peak observed at around 0.25 V is due to the oxidation of Pd. For reference, Pd in KCl solution is oxidized into K 2 PdCl 4 with an oxidation peak observed at around 0.25 V. 11 Furthermore, the smooth peaks spanning from À0.2 V to À0.6 V are also reported in the reference and can be ascribed the reduction of Pd salt into Pd. 11 These peaks cannot be observed when the mesoporous lm is coated with Naon® which prevents Pd from redox reaction. Previous works reported that protein adhesion is hindered by the presence of Naon® coating. 12 Only small molecules such as lactate and glucose can be transported through this polymer, 13 which is why good selectivity is expected for lactic acid sensing.
The electrochemical surface area (ECSA) of the catalyst lms were investigated by carrying out CV in 0.5 M H 2 SO 4 with a scan rate of 50 mV s À1 between À0.2 and 1.2 V (vs. Ag/AgCl) (Fig. 2a). The ECSA is estimated by calculating the charge associated with PdO reduction between the potential of 0.35 and 0.70 V (vs. Ag/AgCl) from the negative potential scan of the CV. By assuming that the conversion factor for an oxide monolayer reduction is 420 mC cm À2 on a smooth Pd surface, the massnormalized ECSA for mesoporous Pd lms is calculated to be 37.9 m 2 g À1 , which is 10 times higher than that of the nonporous Pd lms (3.56 m 2 g À1 ). As shown in Fig. 2b, CVs are measured in NaCl solution containing lactic acid (10 mM) by using mesoporous and nonporous Pd lms as the working electrode. Anodic current and cathodic peaks can be observed, corresponding respectively to the oxidation of lactic acid and reduction of pyruvic acid. The electro-oxidation of lactic acid into pyruvic acid is likely to occur at the surface of the Pt electrode, as previously reported. 14 Because of a high ECSA, the mesoporous lm displays larger current peaks and superior catalytic activity towards the oxido-reduction reaction of lactic acid and pyruvic acid compared to that of the nonporous lm. Both the dependence of the ECSA and anodic peak current density by lactic acid oxidation on the deposition time are shown in Fig. S6. † Since the size of lactic acid molecules is much bigger than that of H + ion, the anodic peak current density due to oxidation of lactic acid becomes saturated when the lm thickness is larger than 600 nm. Therefore, we used the mesoporous Pd lm prepared for 600 s for further electrochemical study.
From Fig. 2b, it can be observed that the oxidation of lactic acid and reduction of pyruvic acid on mesoporous Pd lm take place between À0.5 V and 0.5 V (vs. Ag/AgCl). Therefore, the chronoamperometry is measured under the same voltage window to minimize the inuence of glucose, urea, ammonia or ethanol. These interferents, present in human sweat, 15 and cannot be removed by the Naon® coating because they are too small to be ltrated. The current response due to lactic acid and other interferents are compared in Fig. 3a, thus conrming that the contribution from the interferents becomes negligible under an applied potential of 0.4 V (vs. Ag/AgCl).
Therefore, chronoamperometry was measured at each addition of 1 mM of lactic acid and the results are shown in Fig. 3b. To evaluate the sensitivity of the Naon®-coated mesoporous Pd lm to lactic acid, bare mesoporous Pd, Naon®-coated nonporous Pd and bare nonporous Pd lms were also tested. The limit of detection (LOD) is calculated to be 0.34 mM (for the Naon®-coated mesoporous Pd lm), 0.48 mM (for the bare mesoporous Pd lm), 0.75 mM (for the Naon®-coated nonporous Pd lm), and 1.5 mM (for the bare nonporous Pd lm). The LOD using Naon®-coated mesoporous Pd lm (0.34 mM) is superior to what was previously reported (0.5 mM). 16 The relationship between the concentration of lactic acid and the current is shown in Fig. 3c. The mesoporous Pd lm shows a higher sensitivity than the nonporous Pd lm, due to a higher ECSA. Interestingly, the Naon®-coated mesoporous Pd lm shows a higher sensitivity than the bare mesoporous Pd lm. To discuss this result, adsorption of lactic acid on both electrodes in NaCl solution should be considered. In 0.4% NaCl solution, the pH is almost 7.0 and lactic acid releases H + ion because the pK a is 3.86. Therefore, negatively charged lactic acid ions cannot approach the Naon® surface due to electrostatic repulsion. It would then  be expected that the bare mesoporous Pd lm shows higher sensitivity. This is due to chloride ions being critical interferents during the sensing of lactic acid. When Naon® is present, the chloride ions are removed. We tested the repeatability of this sensor using the relationship between the amperometric response and the lactic acid concentration. Even aer addition of 15 mM, the linearity is maintained showcasing high repeatability (Fig. S7 †).

Conclusion
We successfully prepared a Naon®-coated mesoporous Pd lm by simply dropping Naon® peruorinated resin solution onto a mesoporous Pd lm. The uniform Naon® coating can prevent Pd from redox reaction in NaCl solution. Furthermore, by applying an appropriate external potential, serious interferents present in human sweat, such as glucose, urea, ammonia and ethanol, can be minimized so the lactic acid concentration can be determined precisely. Such non-invasive electrochemical sensing based on Naon®-coated mesoporous metal lm is promising for daily diagnosis of various diseases and health management.

Conflicts of interest
There are no conicts to declare.