MXene-Graphene Oxide Heterostructured Films for Enhanced Metasurface Plasmonic Biosensing in Continuous Glucose Monitoring

Abstract

Non-invasive biosensors have attracted attention for their potential to obtain continuous, real-time physiological information through measurements of biochemical markers, such as one of the most important-glucose, in biological fluids. Although some optical sensing materials are used in non-invasive devices for continuous glucose monitoring (CGM), surface or localized plasmon sensing material are seldom applied in CGM owing to modest sensitivity and bulk sensing apparatus. Herein, a metasurface (MGMSPR) biosensor based on the metasurface plasmon resonance chip modified with heterostructured Ti₃C₂ MXene-Graphene oxide (MG) is reported, which potentially enables ultra-sensitive glucose detection. The sensor consists of a dual-channel microfluidic device integrated with silver mirror enhanced MGMSPR chips. Not only does it promote the entry of glucose oxidase (GOD) into the internal pores and enhance the stable fixation of GOD in the membrane, but also the integration of MG material provides a high specific surface area and unique electronic properties, thereby significantly enhancing the sensitivity of the MGMSPR sensor. The detection limit of MGMSPR biosensor is 106.8 µM. This pioneering approach opens new avenues for monitoring physiological parameters and process analytical technology on an optical platform, providing continuous health monitoring and production process control through optical sensors.

Keywords: MXene; glucose monitoring; heterostructured material; metasurface plasmonic sensor; microfluidics.

Figure 1

Scheme of the real‐time monitoring MGMSPR glucose quantitative analysis system. A) Principle of the MGMSPR chip to enhance the metasurface sensor by cavity reflection with 2D material. B) Schematic of a reflective dual‐channel sensor for an optical platform. The GOD on the surface of the test chip of the MGMSPR microfluidic platform reacts with the incoming glucose, while no reaction occurs on the reference chip. C) Workflow of an automatic microfluidic device for detecting glucose analytes in sweat and through PAT.

Figure 2

Performance of SME chip sensors. A) Photograph for a metasurface chip. B) Top view SEM image of metasurface chip. C) Cross‐section of metasurface chip SEM image. D) Schematic diagram of preparation of SME chip. E) Relationship between silver deposition thickness and sensitivity. F) Relative reflection spectra of the 9 nm Ti + 90 nm Au chip without SME influence in the 500–700 nm range, with characteristic wavelengths at 605 and 660 nm. G) Relative reflection spectra of the 9 nm Ti + 90 nm Au chip with SME influence in the 500–700 nm range, with characteristic wavelengths at 605 and 660 nm. H) The 4PL regression was fitted to the differential wavelength of sucrose concentration on 9 nm Ti + 90 nm Au chip. I) Relative reflection spectra of 9 nm Ti + 70 nm Ag + 20 nm Au chips without SME influence in the 500–700 nm range, with characteristic wavelengths at 628 and 650 nm. J) Relative reflection spectra of 9 nm Ti + 70 nm Ag + 20 nm Au chips with SME influence in the 500–700 nm range, with characteristic wavelengths at 628 and 650 nm. K) The 4PL regression was fitted to the differential wavelength of sucrose concentration on 9 nm Ti + 70 nm Ag + 20 nm Au chip. The error bars are defined as mean ± SD (n = 3).

Figure 3

Preparation and performance verification of MGMSPR sensor. A) Preparation of MG films. B) Preparation of MGMSPR sensor. C) TEM of MXene film. D) TEM of GO film. E) TEM of the MG hybrid film. F) SEM of different concentrations of MG hybrid films modified to metasurface. G) Investigation of the correlation between the MG Ratio and MetaSPR sensitivity. H) Analysis of relative response in the 4PL curve for detecting sucrose concentrations. I) Real‐time monitoring of IgG binding curves. J) Wavelength shifts in the 4PL curve for detecting various concentrations of IgG. K) Real‐time monitoring of the association and dissociation interactions between Rapamycin and FKBP12 molecules via MGMSPR sensor. The error bars are defined as mean ± SD (n = 3).

Figure 4

Construction of the MGMSPR glucose chip sensor. A) The principle of MGMSPR sensor for detecting glucose. B) Comparison of glucose detection efficiency using three different methods of GOD modification. C) Comparison of Nafion film modification capabilities at different concentrations. D) Comparison of MG ratio capabilities at different concentrations. E) The differential spectra of glucose detection at various concentrations using the MGMSPR biosensor. F) The relative response of the MGMSPR biosensor as a function of glucose concentration. The error bars are defined as mean ± SD (n = 3).

Figure 5

The MGMSPR sensor detects glucose in sweat and PAT. A) Schematic diagram for constructing the MGMSPR sensor ligand and monitoring glucose. B) Photograph of a microfluidic biochip card based on a dual‐channel MGMSPR sensor. C) Schematic diagram for monitoring glucose levels in sweat and PAT. D) Glucose curves in artificial sweat monitored by MGMSPR biosensor at different concentrations. E) The relative response of the MGMSPR biosensor as the 4PL regression for the glucose concentration in artificial sweat. F) Detection of glucose concentration in artificial sweat using the commercialized meter and MGMSPR‐based biosensor. G) Selectivity validation of the MGMSPR glucose sensor (a: 110 µm glucose, b: 100 µmM lactate, c: 10 µMm ascorbic acid, d: 59 µm uric acid). The glucose sensor responds only to changes in glucose concentration. H) Non‐specific validation of the MGMSPR glucose sensor (testing against seven analytes: glucose, lactate, uric acid, Dopamine, NaCl, KCl, sample number = 5). I) Detection of glucose concentrations during PAT using commercially available glucose devices and MGMSPR‐based biosensors. J) Changes in the response of the MGMSPR platform to the number of repetitions are shown. Figure 5E,F,I,J error bars defined as mean ± SD (n = 3).