Recent advances in biofluid detection with micro/nanostructured bioelectronic devices

Hu Li ac, Shaochun Gu b, Qianmin Zhang b, Enming Song c, Tairong Kuang b, Feng Chen *b, Xinge Yu *c and Lingqian Chang *ad
aBeijing Advanced Innovation Center for Biomedical Engineering, School of Biological Science and Medical Engineering, Beihang University, Beijing, 100191, P. R. China. E-mail:
bDepartment of Material Science and Engineering, Zhejiang University of Technology, Zhejiang, 310014, P. R. China. E-mail:
cDepartment of Biomedical Engineering, City University of Hong Kong, Hong Kong, China. E-mail:
dSchool of Biomedical Engineering, Research and Engineering Center of Biomedical Materials, Anhui Medical University, Hefei 230032, P. R. China

Received 19th October 2020 , Accepted 12th December 2020

First published on 15th December 2020


Most biofluids contain a wide variety of biochemical components that are closely related to human health. Analyzing biofluids, such as sweat and tears, may deepen our understanding in pathophysiologic conditions associated with human body, while providing a variety of useful information for the diagnosis and treatment of disorders and disease. Emerging classes of micro/nanostructured bioelectronic devices for biofluid detection represent a recent breakthrough development of critical importance in this context, including traditional biosensors (TBS) and micro/nanostructured biosensors (MNBS). Related biosensors are not restricted to flexible and wearable devices; solid devices are also involved here. This article is a timely overview of recent technical advances in this field, with an emphasis on the new insights of constituent materials, design architectures and detection methods of MNBS that support the necessary levels of biocompatibility, device functionality, and stable operation for component analysis. An additional section discusses and analyzes the existing challenges, possible solutions and future development of MNBS for detecting biofluids.

image file: d0nr07478k-p1.tif

Hu Li

Hu Li is a Postdoctoral Fellow at the Department of Biomedical Engineering, City University of Hong Kong, Hong Kong, China. He received his Ph.D. from Beihang University, China, in 2020, and his Bachelor's degree from Tianjin Polytechnic University, China, in 2014. His research interests are focused on nanogenerators, self-powered sensors, flexible electronics and bioelectronics.

image file: d0nr07478k-p2.tif

Shaochun Gu

Shaochun Gu is a Master at the Department of Materials Science and Engineering, Zhejiang University of Technology, China. She received her Bachelor's degree from East China Jiao Tong University, China. Her research area is focused on the modification of polysaccharides, and their photolithography process.

image file: d0nr07478k-p3.tif

Feng Chen

Dr Feng Chen is currently an Associate Professor and Senior Research Chemist at Zhejiang University of Technology, China. He obtained his B.A. (2003) and Ph.D. degree (2008) from Zhejiang University, China. His current research areas are in the synthesis of biomacromolecules, researching the micro/nano-scale patterning technique of biomacromolecules, and targeting the biosensors, drug delivery and disease diagnosis. He is also interested in the fundamental aspects of the processing of polymers, composites and polymeric nanocomposites.

image file: d0nr07478k-p4.tif

Xinge Yu

Xinge Yu is an Assistant Professor of Biomedical Engineering at CityU. He finished his Ph.D. research of printable flexible electronics at Northwestern University (NU) and UESTC in 2015. From 2015 to 2018, he was a postdoctoral associate in the Center for Bio-Integrated Electronics at NU, and the Department of Materials Science and Engineering at the University of Illinois at Urbana-Champaign. His research focuses on developing skin-integrated electronics and bioelectronics, and he conducts multidisciplinary research addressing challenges in practical applications.

image file: d0nr07478k-p5.tif

Lingqian Chang

Prof. Lingqian Chang obtained his Ph.D. in Biomedical Engineering from Ohio State University, followed by postdoc training in CCNE nanoscale center at Northwestern University. He used to be an assistant professor at the University of North Texas. Currently he is a full professor at Beihang University, and founded The Institute of Single Cell Engineering. His research is mainly focused on cellular micro-/nano-technologies, aiming to design novel nanochips and nanosensors for gene detection and cell therapy in live cells. He has published more than 50 peer-reviewed papers, 1 book, 3 book chapters and hold 5 China Patents and 3 US Patents.

1 Introduction

Many physiological indices of the human body can be obtained by various biosensors to evaluate the physiological status.1–5 Biochemical detection is an effective tool to judge the human health status in our daily life. Traditional detection methods rely on collecting blood samples to measure component parameters in blood for a definite diagnosis, such as blood urea, creatinine, glucose, Na+, K+, and Ca2+.6 This process usually brings pain to patients, consumes time for component analysis, and presents the risk of infection. Meanwhile, the consumables and medical expenses needed for repeated testing will also cause an economic burden and psychological pressure on patients.7–10 Benefiting from modern biomedical and bioelectronic technology, many alternative tools are in development for addressing these types of issues or others.

Biofluids (i.e., biological body fluids) are secreted by organisms and contain various biochemical components that relate to the human health condition. The real-time monitoring of biofluid detection can collect physiological information of importance. These biofluids often refer to interstitial fluid (ISF), sweat, saliva, tears and urine.11 Biofluid detection can not only reduce the burden of frequent blood collection in patients that suffer from chronic diseases (such as diabetes), but also provide long-term health monitoring for the human body, especially in fitness and athletics.12 Therefore, the study of biofluids is of great significance for the advancement of medical treatment and daily health.

As examples of detecting pH and ion concentration in biofluids (such as Na+ and K+) and conventional detection technology (such as venipuncture for blood collection), finger blood collection can provide high-precision measurement in clinical practice. However, it showed limited ability in detecting low-concentrated molecules, for example, glucose, lactic acid, uric acid and drugs. Among them, the glucose level in body fluids is the most widely studied parameter. Up to now, relevant commercial products mainly depend on implantable sensors, where the product category is less with poor selectivity. The main reason is that the concentration of components in ISF is similar to that of blood, but lower in other bio-fluids (Table 1). As the ultimate solution, advances in bioelectronic devices are highly necessary, with high sensitivity/accuracy and low detection limit. Unfortunately, traditional sensors fail to meet these requirements, mostly due to the interference of multiple components.13

Table 1 Comparison of several components in blood plasma, ISF, saliva, sweat, tear and urine
  Na+ K+ Lactate Uric acid Glucose Cortisol (Unbound) Drugs (Unbound)
Molecular weight 23 39 90 168 180 362 Hundreds of daltons
Oil affinity Very low Very low Very low Very low Low High Often high
Blood plasma 135–145 mM 3.5–5 mM 0.5–10 mM 155–428 mM 3.9–6.2 mM Tens of nmol Related to the dose
ISF Similar to plasma Similar to plasma Similar to plasma Similar to plasma Similar to plasma Similar to plasma Similar to plasma
Saliva Tens of mmol Tens of mmol Tenths to ones of mmol ∼1% of plasma ∼1% of plasma Similar to plasma Similar to plasma
Sweat Tens of mmol 5–15 mM 5–10 mM ∼1% of plasma ∼1% of plasma Similar to plasma Similar to plasma
Tear 120–165 mM 20–42 mM 2–5 mM ∼1% of plasma ∼1% of plasma Similar to plasma Similar to plasma
Urine Tens of mmol Tens of mmol ∼1% of plasma ∼1% of plasma Similar to plasma Similar to plasma

Micro/nano structure designs have provided feasible solutions for the above outlined issues. They usually have a large relative area, and specific optical and electrical properties that serve as biosensor platforms for biological and biomedical applications.14 This technical progress greatly reduces the sample dimension for biofluid detection, even at microliters or less, with high accuracy and stability.15 These advanced biosensors have made biofluid detection rapid, accurate and noninvasive. Given this point, we provide a timely review of the current progress of micro-/nano-biosensors (MNBS) in biofluid detection (Fig. 1). The first section introduces the application scenarios of traditional biosensors (TBS) in biofluid detection. The subsequent section summarizes the recent advances of MNBS for detecting various biofluids, such as sweat, urine, saliva, tear and ISF. At length, the existing challenges, possible solutions and future prospect of MNBS for detecting biofluids are discussed. Overall, this review focuses on a new insight in the field of biofluid detection.

image file: d0nr07478k-f1.tif
Fig. 1 Overview of biofluid detection of various analytes reflecting health conditions. Various MNBS were developed to detect five types of biofluids (i.e., sweat, urine, saliva, tear and ISF) and obtain the corresponding contents of the associated analytes (i.e., Na+/K+, glucose, uric acid, drug, lactate, cortisol).

2 Bio-fluid detection with traditional biosensors (TBS)

2.1 Interstitial fluid (ISF)

ISF is present in the interstitial space of tissue cells, including most of the dermis, and also surrounds the salivary and sweat glands. Most of ISF is gelatinous and exists in the interstitial tissue cells. The gelatinous ISF cannot flow freely, but exchanges various substances within the blood; the results of which lead to a composition similar to that of plasma, in addition to lower protein content. In 2019, Heo et al.16 summarized the development of subcutaneous implantable sensors for continuously monitoring ISF glucose. One of his former research studies highlighted the use of glucose-responsive fluorescent hydrogel fibers implanted in humans or animals, combined with wireless transdermal transport for long-term glucose monitoring (Fig. 2a). They found that the combination of polyethylene glycol (PEG) and polyacrylamide (PAM) gel fibers (length, 5 mm) can effectively improve the biocompatibility and provide stable operation in a safe fashion up to 140 days in a male mouse body (weight, 21 g–26 g), eliminating the hassle and pain required for frequent implantation of the implantable sensor.17 In addition, there are some other minimally invasive detection technologies, such as iontophoresis18 and micro-dialysis.19 Compared with traditional methods, these technologies have the advantages of less harm to the human body and short-term real-time measurement, but at the same time, they require high device performance (i.e., safety, biocompatibility, stability) and induce the potential risk of infection. In this case, non-invasive testing is more attractive. In 2001, the Food and Drug Administration (FDA) approved Cygnus's hand-held glucose detection system that uses reverse electrophoresis to force glucose molecules to penetrate the skin surface via an electric current, subsequently leveraging the enzyme-catalyzed conversion and electrochemical methods to detect the glucose.20 Although there is no direct damage to the skin, reverse electrophoresis can cause irritation or unknown risks of damage. The ISF in five body fluids is the closest to the plasma component, but it is difficult to collect in the gel state, with limitation on its study and application. In contrast, sweat, saliva, tears and urine are relatively easy to collect.
image file: d0nr07478k-f2.tif
Fig. 2 Biofluid detection with a traditional biosensor (TBS) on different body positions. (a) An interstitial glucose sensor implanted in an animal's ear.17 (b) Transparent soft contact lens tear glucose sensor.38 (c) Tear sensor integrated on the nose pad of the spectacle frame.41 (d) A mouth-guard biosensor for monitoring salivary glucose.47 (e) Mouth-guard devices with electrochemical sensors to measure the salivary uric acid concentration.46 (f) Sweat sensors on the wrist, chest, waist and head.21,24 (a) was reproduced with permission from National Academy of Sciences, ref. 17, Copyright 2011; (b) was reproduced with permission from The American Association for the Advancement of Science (Open Access), ref. 38. Copyright 2018. (c), (d) and (e) were reproduced from ref. 41. Copyright 2019, ref. 57. Copyright 2016 and ref. 46, Copyright 2015, respectively, with permission from Elsevier. (f) was reproduced with permission from Springer Nature, ref. 24. Copyright 2016.

2.2 Sweat

Sweat is widely studied for its ready availability and noninvasive detection. Sweat glands are all over the body, and normal people evaporate about 600–700 ml sweat within a day. Perspiration contains a large amount of water, and a small amount of electrolyte, glucose, lactic acid and other substances. The most advanced wearable device has been able to measure the Na+, K+, pH in sweat and skin temperature indicators. They are often worn on the wrist, chest, waist and head (Fig. 2f). In 2016, Caldara et al. studied electronic devices for monitoring sweat pH based on intelligent fabrics. The device was assembled with cotton fabric treated with a pH sensing agent via wireless interface and a small electronic device (50 mm × 47 mm × 15 mm), which can be worn around the waist to detect sweat in motion.21 Other components, such as glucose, lactate, uric acid,22,23 ethanol, and drugs in sweat are also studied. In 2016, Gao et al.24 reported a fully integrated wearable sensor array for multiplexing in situ perspiration analysis. This smart wristband can realize the simultaneous selective detection of various components in sweat. The integration of five sensors (glucose, lactate, Na+, K+, temperature) on a mechanically flexible PET (polyethylene terephthalate) substrate, where FPCB (flexible printed circuit board) technology was exploited to incorporate the critical signal conditioning, processing, and wireless transmission functionalities. Experiments demonstrated that the results of detection were consistent with the non-in situ detection. A lactate sweat sensor and an electrocardio sensor system were integrated on a flexible patch consisting of a lactate amperometric sensor and an electrocardiogram sensor, which were attached to the fourth rib of the rib cage, for simultaneous ECG (Electrocardiograph) monitoring and lactate concentration detection of sweat.25 The experiment indicated that these two sensors can maintain good independence and detect separately without interfering with each other. Beyond these studies, there are also some reports for detecting other ingredients. For example, M. Gamella et al.26 proposed a method in 2014 to measure in situ sweat ethanol. The biodevice consisted of a bienzyme composite graphite-Teflon electrode, an Ag/AgCl reference electrode and an auxiliary Pt (platinum) electrode. The three electrodes are submerged in phosphate buffer solution (PBS, 0.05 M, pH 7.4). This biodevice had the advantage of detecting liquid ethanol directly from sweat rather than gas, compared with traditional alcohol detectors, ensuring the accuracy of the sensor.

Sweat sensing is sensitive to human activity and ambient temperature, so its stability and data reliability still need to be further improved. In practical application, sweat sensing is easily affected by multiple factors (e.g., temperature, exercise, physiological status).27–29 The analytes in body fluids are different due to the individual differences and varieties of physiological and pathological status.30 For example, sweat in normal people has a low glucose concentration (less than 0.2 mM), while sweat in diabetic patients has a glucose concentration of 0.28–1.11 mM.31 Ono et al. analyzed the sweat glucose concentration of patients with specific dermatitis, and found that specific dermatitis would cause an increase of the sweat glucose concentration.32 The uncertainty of sweat analyte concentration leads to a wide range of pH fluctuations (4.0–6.8).33 The pH value increases with the increase of the Na+ concentration and sweat rate.34,35 Overall, whether individual physiological or pathological differences lead to the uncertainty of the sweat glucose content or the fluctuation of the pH value, the final result directly or indirectly affects the reliability and effectiveness of the sweat glucose sensors used for monitoring or diagnosis.36

Therefore, it is necessary and urgent to establish a robust and meaningful data connection between the level of certain content (for instance, glucose) in sweat and in the human body under real-world clinical circumstances, especially considering the individual differences and varieties of physiological and pathological status. Wearable sweat detectors can be used as medical devices and sports medicine only if these problems are resolved.

2.3 Tear

Similarly, tears are produced by the lacrimal glands (reflex tears) and para-lacrimal gland (resting/basal tears) with the effects of lubrication and sterilization. There are many elements, such as lysozyme, immunoglobulin, sugar, inorganic salts, in tears. There have been numerous reports of tear sensors, particularly contact lens sensors,37,38 which can serve as a continuous, minimally invasive detection device based on the relationship between blood and glucose, sodium, potassium and chloride concentrations in tears. A flexible transparent smart contact lens glucose sensor (Fig. 2b) has been reported by Park et al. in 2018.38 The flexible substrate (elastofilcon A) and rigid island design (a photocurable optical polymer, 50 μm thick) can protect the rigid electronic devices, and a new tear sensor has been fabricated combining LED pixel and wireless transmission technology. The contact lens sensor leveraged the enzyme sensor to analyze the glucose concentration in tears, and thereby set a standard value according to the glucose concentration in tears of fasting diabetic patients in advance. If the detection was higher than this standard value (0.9 mM), the LED will turn off. Such design still has room for improvement. For example, the security and biocompatibility of the product need to be taken into account. It was worth mentioning that the standard value (0.9 mM) in this work was just set by the authors, the mean tear glucose value in clinic vary in a range. According to the clinical study between 121 diabetic and nondiabetic subjects, the mean values of diabetic and nondiabetic tear glucose were 0.35 ± 0.04 mM and 0.16 ± 0.03 mM, respectively.39 On the other hand, the reported values for tear glucose in normal individuals ranged from 0 to 3.6 mM (65 mg dL−1) for normal individuals, whereas concentrations as high as 4.7 mM (84 mg dL−1) were obtained for patients with diabetes mellitus.40 The values are also different when the tears were obtained by different methods, such as mechanical stimulation, chemical and non-contact stimulation, and non-stimulated tears.40

In order to avoid the potential harm of in situ tear detection, Sempionatto et al.,41 in 2019, reported an electrochemical enzyme biosensor that can be integrated into the nasal bridge pad of the eye (Fig. 2c). This biosensor was made up of a polycarbonate membrane, an adhesive spacer, a paper outlet and electrochemical biosensor. Such wearable electronic devices collected, stored, and analyzed the alcohol content of tears by using a microfluidic electronic device integrated into the nasal bridge pad of the glasses. Although their experiments showed that the sensor's data were relatively accurate, the research was still in its infancy stage, where a future challenge focuses on the long-term stability and measurement accuracy of the sensor.

2.4 Saliva

Saliva contains many different enzymes, electrolytes and other components. It has the function of decomposing food and immune sterilization. There has been a lot of controversy about saliva testing in glucose. Some previously reported research studies have shown no significant correlation between the salivary glucose levels42 and serum glucose concentration, while other research studies demonstrated a clear correlation.43–45 According to the later view, saliva was thought to be a useful tool for noninvasive, supplemental testing for diabetes. Moreover, saliva can be an alternative and used to detect pepsin, cortisol and glucose to help diagnose certain diseases by an in situ method, due to the risk of swallowing with small sensors and foreign body response with larger ones. Meanwhile, the mouth-guard biosensor provided another solution for this problem. The design and preparation of denture sensors for detecting analysis in saliva have attracted much attention from researchers in recent years. In 2015, Kim et al. designed a biosensor for detecting salivary uric acid based on a silk-screen electrode modified by uricase and a tiny electronic device, which were integrated into the mouth-guard (Fig. 2e) with high sensitivity, good selectivity, and achieved continuously real-time monitoring. It can also be customized for other expanded functions, with high application prospects.46 Another similar work is a mouth-guard saliva glucose sensor, which is made by fixing the working electrode (0.20 mm2 Pt) and reference electrode (4.0 mm2 Ag/AgCl) on the mouth-guard made by an enzyme membrane (Fig. 2d). By integrating with the wireless measurement system, the long-term real-time monitoring of oral saliva glucose concentration can be achieved (exceeding 5 hours).47 Experiments showed that all of these sensors had good accuracy and were reusable for multiple times, with great potential for monitoring chronic diseases, such as diabetes, parotitis, and periodontal disease. Saliva was also involved in the detection of carboxymethyl lysine,48 methamphetamine,49 the active ingredient of marijuana,50 and others.

2.5 Urine

Urine collection for biochemical tests of human health has many clinical applications, such as to diagnose the health status of the liver, kidney and gallbladder. Urine contains many components, such as glucose, uric acid, bacteria, urine protein, and others. The detection of the components can reflect corresponding organ lesions. The clinical tests are convenient and inexpensive, but they still require the use of bulky urine analysis equipment. As a result, in recent years, emerging classes of detection devices focus on portable design and minimized structures, with a large portion combined with smartphones that allow users to receive data and monitor health in a wireless fashion.51 Paper-based urine biosensors have been designed to measure the concentration of gold ions in urine, in conjunction with a smartphone-based fluorescence diagnostic system.52

3 Biofluid detection with micro/nano-structured biosensors (MNBS)

Following the development of materials science and manufacture technology, a variety of novel biosensors with different functions have been proposed, such as hybrid sensors that can measure both electrocardiograms and lactic acid.25 However, biosensors based on the biofluids still face common challenges, in the aspects of material biocompatibility and stability, accuracy, reliability, power supply, data collection, processing and transmission.53 For implantable biosensors, biocompatible and stable materials are highly necessary. For the purpose of solving the issues with simplicity, a promising approach is to construct micro/nanostructured devices that offer the capability to reduce the detection amount of biofluids to μL, and thereby to improve the reliability in the measurement time.13 In this context, micro-/nano-materials and micro-/nano-structures have been applied to bioelectronic devices for biofluid detection.

As shown in Table 2, there are many types of nanomaterials that can be used for biofluid detection, including metals and metal oxides, carbon materials, other inorganics, nanocomposites and micro-patterns. These nanostructures include nanoparticles, nanowires, nanotubes, nanospheres, nanoflakes, nanoribbons, microneedles and microfluidic structures. These nanostrutures contribute to improving the biosensing performance because of their high surface-to-volume ratio, and conductive or semiconductive property. The diversity of materials and structures provided more options for biosensor fabrication.

Table 2 Nanomaterials and nanostructures used for biofluid detection
Type Material and structure Ref.
Notes: NP, nanoparticle; NW, nanowire; NOS, nanosphere; NF, nanoflake; NR, nanoribbon; GO, graphene oxide; rGO, reduced graphene oxide.
Metals and metal oxides Au NP, Ag NP, Ag NOS, ZnO NF, ZnO NR, Co3N NW, HfO2 NP, Ti3C2Tx, Co3O4, NiO NP 32, 39, 43, 75–77 and 80–81
Carbon Graphene, GO, rGO, CNT, SWCNT, MWCNT 36–37, 41, 52–54, 60 and 67–72
Other inorganics Si NW, Si NR 34
Nanocomposites Graphene/Ag NW, Ag-rGO, Fe3O4/GO/MIP, IrO2@NiO core–shell NWs, CuO/GO/CNF, Au/rGO/AuPt NP, rGO-ZnO 36, 55–56, 58, 61 and 70
Micro-pattern PANI-Au hybrid nanostructure, micro needle, microfluidic structure 38, 44–48 and 51

3.1 ISF

As the body fluid whose composition is closest to blood, tissue fluid is the most promising body fluid to replace blood, with the successful extraction of tissue fluid. In this respect, future efforts include minimally invasive percutaneous detection. The applications in this area mainly focus on nanostructured platforms of implantable detection devices. In 2010, Yuen et al.54 reported a surface-enhanced Raman spectroscopy (SERs) implantable glucose sensor. Silver film over nanosphere (AgFON) surfaces were functionalized with a mixed self-assembled monolayer (SAM), and implanted subcutaneously in a Sprague-Dawley rat. The sensor solved the problem of metabolites, and therefore improved the drug sensitivity in the body. The sensor served as an in vivo implant in animal models, such as mice, to measure glucose levels. Tests showed that the device performance remained stable for 17 days.55 Biodegradable elastomers (POC, (poly(1,8-octanediol-co-citrate)), Mg (magnesium, 300 nm thick) and silicon nanofilms/nanoribons (100 nm or 300 nm thick) were used by Suk-Won Hwang et al. to construct biodegradable sensors to obtain implantable devices with good biocompatibility that is completely dissolvable in biofluids, where the final product was non-toxic and harmless.56 The biodegradability in this research was demonstrated by submerging the device in phosphate buffer solution (PBS, 0.1 M, pH 4–10) for 12 h at 37 °C. Electrocardiograms (ECG) and electromyograms (EMG) were recorded by attaching the sensors on the chest and right forearm of a volunteer within minutes. Considering the high requirements of the medical device, the biocompatibility evaluation of the sensors in this research needs to be further improved according to assays given in ISO 10993 “Biological evaluation of medical devices” standards, such as in vitro irritation test and in vivo irritation test. To make the functionality of the implantable devices last longer, the researchers found that nanoscale transistors constructed by metal silicide alloys (TiSi2) instead of monocrystalline silicon could extend the life of implantable devices by over 20 times.57 Both strategies are applicable as tissue fluid sensors for subcutaneous implants. However, as mentioned above, the minimally invasive detection of the implanted device still has great defects compared with the non-invasive detection. Therefore, future studies will focus on the non-invasive detection of sweat, saliva, tears and urine.

3.2 Sweat detection

There are two main ways to detect sweat on the skin surface using micro/nanostructured devices: (1) wearable electrochemical devices made of nanomaterials, or to design micro needle structures on them; (2) soft, skin-integrated microfluidic systems for collection and colorimetric chemical analysis. As an example, a sweat glucose sensor based on an rGO (reduced graphene oxide) nanocomposite electrode consisted of a flexible polyimide substrate, rGO, gold and platinum alloy nanoparticles, a chitosan-glucose oxidase composite, and a water-resistant Nafion layer.58 The sensor showed a high sensitivity (48 μA mM−1 cm−2), a rapid response (20 s) and an outstanding linearity (0.99). In addition, the results of the sweat samples were accurate, and the detection ranges from 0 to 2.4 mM, covering the widest glucose range in human sweat.

In 2016, Lee et al.59 reported a graphene-based sweat glucose sensing device integrating five sensors (temperature, humidity, pH, glucose and vibration) on a serpentine gold mesh and gold-doped graphene. Users can see the real-time sweat glucose concentration value and change curve on the phone. In order to prove the accuracy of the patch, the change curve of the glucose concentration at each time period measured by different methods (blood glucose, sweat glucose concentration and the diabetes patch) was also plotted (Fig. 3c). Due to the excellent electrochemical and mechanical properties of graphene itself, these properties were amplified by gold nanoparticle doping and other operations. Consequently, the device was highly sensitive and accurate, with a minimum detection limit of 20 μL. The pH sensor can also perform the in situ correction of the pH-dependent glucose concentration to ensure the accuracy of the measurement results. Before long, they designed temperature-controlled hydrogel-based transdermal drug delivery microneedles for the device, as well as wearable sweat glucose monitors.60 Transdermal drug delivery microneedles consisted of two PCM loaded drugs with phase change temperatures of 38 °C and 43 °C, wrapped in microneedles made of hyaluronic acid, and then coated by PCM to prevent dissolution in contact with body fluids. After combining the microneedle array with the sensor patch, a two-stage drug delivery can be realized (Fig. 3e).

image file: d0nr07478k-f3.tif
Fig. 3 Sweat detection with various micro/nanostructured biosensors (MNBS). (a) A laser-engraved graphene-based wearable biosensor for detecting uric acid and tyrosine.61 (b) Near-field communication (NFC) coil for wireless measurements of the skin temperature.69 (c) Schematic illustration of the diabetes patch, which is composed of the sweat-control (i and ii), sensing (iii–vii) and therapy (viii–x) components.59 (d) A non-enzyme sensor for CNTs silicone patch apply to the skin.63 (e) Photos of wearable sweat patches, disposable sweat monitoring strips and transdermal drug delivery devices.60 (f) The microfluidic patch is applied to the skin to detect various components of sweat.70 (a) and (c) were reproduced from ref. 61. Copyright 2019 and ref. 59. Copyright 2016, with permission from Springer Nature; (b), (e) and (f) were reproduced from ref. 69. Copyright 2019, ref. 60. Copyright 2017 and ref. 70. Copyright 2020, with permission from The American Association for the Advancement of Science (Open Access). (d) was reproduced with permission from American Chemical Society, ref. 63. Copyright 2018.

Alternatively, Y. Yang et al.,61 in 2020, reported a laser-engraved graphene-based chemical sensor (LEG-CS) for detecting low concentrations of uric acid (UA) and tyrosine (Tyr) (size: 2 cm × 2 cm) (Fig. 3a). The detection range for UA and Tyr were 20 μM to 80 μM and 50 μA to 200 μA, respectively. The LEG-CS showed high sensitivities for UA and Tyr at physiological concentrations, i.e., 3.5 μA μM−1 cm−2 and 0.61 μA μM−1 cm−2, respectively, and low limits of detection of 0.74 μM respectively, respectively. The whole device on the polyimide layer consisted of a microfluidic module and LEG-based chemical and physical sensors for detecting sweat UA, Tyr, temperature, heart rate and respiration rate. On the other hand, non-enzyme sensors have also attracted significant attention because of their simple preparation and convenient operation.62 For example, Oh et al. in 2018 reported a carbon nanotube (CNT)-based silicone-patch sweat glucose sensor with sensitivities of 10.89 μA mM−1 cm−2 and 71.44 mV pH−1 for glucose and pH.63 Depositing the CNT layer by layer on gold nano-sheet resulted in stretchable electrodes. Then, cobaltous tungstate (CoWO4)/CNT and polyaniline/CNT nanocomposites were modified on the electrode to detect glucose and pH. The diameter of the electrode pad was 4 mm, the size of the substrate was 1.5 cm × 1.8 cm × 100 μm. The sweat glucose sensor can be attached to the sweaty skin surface to monitor and map the sweat glucose concentration before and after eating and exercise. Its mechanical stability reached up to 30% stretching, and air stability for 10 days. The curve changed in line with the general rule (Fig. 3d). However, non-enzyme sensors present essential defects in selectivity and stability. In a study, the authors fabricated a Pt-graphite electrode with screen printing technology, and prepared the non-enzyme and enzyme sensor. By detecting the body's sweat glucose, the authors found that the enzyme sensor showed more promising results, with high selectivity and stability. Low sensitivity, narrow linear range and low stability limit the use of non-enzyme sensors. Nevertheless, modification of the graphite electrode with graphene oxide and Pt can effectively improve the sensitivity of the non-enzyme sensor.64

In another example, the non-enzyme glucose electrochemical sensor fabricated by the cobalt nitride nanowire array electrode material on the titanium network has high sensitivity and selectivity, good stability and reproducibility.65 Although these wearable devices allowed for the biochemical detection of sweat outside of labs and clinics, they required an associated power supply and additional data collection, data transmission, among others. Specifically, the skin-mounted microfluidic system offers attractive capabilities. The Rogers team66–69 has conducted abundant research studies in this field. The lithography process was used to obtain microstructured channels on PDMS to collect and store small amounts of sweat, and then colorimetric changes in response to the markers through embedded chemical analysis to obtain sweat secretion, sweat loss, pH, chloride, glucose, lactate, and other data.66 A superabsorbent polymer valve and colorimetric sensing reagent were two important parts of this equipment. In one study, an optimized colorimetric approach was developed to stabilize the color development by designing a superabsorbent polymer water drive valve that selectively isolated individual sweat deposits. The accuracy and long-term stability in this work surpass those of previously reported microfluidic devices by orders of magnitude.67 A poly(styrene-isoprene-styrene) (SIS) encapsulated waterproof, electronics-enabled, epidermal microfluidic device for sweat collection, biomarker analysis, and thermography in aquatic settings was reported to fill the void left by an underwater sweat analysis device. The microchannels of the device have depths of 220 μm microchannel serpentine geometries with 40 turns. Each microchannel just needed 1.5 μL of sweat for detection, the total volume for the whole device was just about 60 μL (Fig. 3b).69

In 2020, Y. Song et al. reported a wireless self-powered sweat sensor without battery. Authors designed a freestanding triboelectric nanogenerator (TENG, 5.78 cm × 3.78 cm) to harvest mechanical energy from human motion based on a flexible printed circuit board. The power density of TENG reached up to 416 mW m−2, and it can power the multiplexed sweat biosensor, and collect and transmit data to users by bluetooth. Both pH and Na+ sensors showed wide detection ranges from 4 to 8 and 12.5 to 200 mM, respectively. They displayed outstanding selectivity, repeatability and stability when detect relevant analytes (Fig. 3f).70

The Gao team also conducted some similar research studies on the soft microfluidic platform. They combined an electrochemical sensor and sweat rate sensor based on current impedance into a microfluidic channel to make a wearable sweat sensing patch to effectively analyze sweat secretion.71 In spite of the many advantages, the susceptibility of sweat to contamination and evaporation caused the sweat analysis to be inadequate in practical applications. For example, sweat secretion and composition were usually influenced by diet, environment and exercise. These factors serve as key features for practical application.

The secretion and composition of sweat will change with human movement, age, diet and disease. Due to the diversity of components in sweat, a glucose sensor is required to have certain selectivity. In laboratory studies, the buffer solution similar to human sweat is usually configured for testing to prove the sensor's selectivity to interfering components.

To obtain accurate results in practical application, the following aspects could be considered:

(1) Data calibration: External factors (e.g., temperature and pH) have an obvious influence on glucose sensing; the test data can be calibrated in factory or by users using a standard solution to improve the reliability and accuracy of the data.

(2) Package. Sweat is easy influenced by contamination and evaporation; a reliable package to sensor can protect the sweat chamber or detection region of sensor from contamination and evaporation.

(3) Developing targeted sensor. For a specific target (e.g., glucose), its content is different in normal people, obese people and diabetic people, even old and young. Therefore, it is a reasonable option to develop a targeted sensor for special groups, or develop multifunctional sensors that can meet multiple requirements. For different people, to design different sensing devices may provide a better service for consumers.

(4) Human trials. Test the practical performance of developed sweat sensors on a lot of normal people and diabetic patients by noninvasive methods, and obtain statistical data before and after exercise and diet. Comparing the differences of the glucose concentration with commercial glucose meters, and calibrate the values by coefficient.

3.3 Saliva detection

To seek the convenience and accuracy of detection, saliva sensors usually focus on the teeth. Two strategies are commonly adopted for this purpose: (1) directly integrated onto the teeth; (2) integrated into the braces, or collecting samples for non-in situ detection. In 2012, Mannoor et al.72 reported a device for detecting saliva bacteria in tooth enamel. Graphene was printed on water-soluble silk fibroin film (50k fibroin to enable an intimate biological transfer of graphene nano-sensors onto biological materials, including tooth enamel. Subsequently, the sensor (size, 1 cm × 2 cm) consisted of a resonant circuit, and interdigitated capacitive electrodes were integrated into the teeth for the remote monitoring of respiratory and saliva bacteria detection. The sensor can detect bacterial cells with 1 μL of solutions containing different concentrations of bacteria (103–108 CFU ml−1) (Fig. 4a). Since the in situ saliva sensor was inconvenient and uncomfortable to wear for a relatively long time, studies were conducted on non-in situ detection. A new localized surface plasmon resonance (LSPR) substrate composed of polyaniline (PANI)-gold hybrid nanostructures as an optical sensor facilitates the monitoring the pH of saliva samples. It can also modify different sensitive materials to detect different components.73
image file: d0nr07478k-f4.tif
Fig. 4 Saliva detection with various micro/nanostructured biosensors (MNBS). (a) Graphene-based wireless saliva sensor integrated onto tooth enamel for bacterial detection in saliva.72 (b) Schematic diagram of the stepwise fabrication of the biosensor.76 (c) Fabrication steps of a BSA/anti-CYFRA-21-1/APTES/nHfO2/ITO platform for oral cancer detection.82 (d) Fabrication process and constituents of the reagent layer of the working electrode.75 (e) Design principle and photos of a microfluidic biosensor.85 (a) and (b) were reproduced from ref. 72. Copyright 2012 and ref. 76. Copyright 2020, respectively, with permission from Springer Nature; (c) and (e) were reproduced from ref. 82. Copyright 2016 and ref. 85. Copyright 2017, respectively, with permission from Elsevier; (d) was reproduced with permission from John Wiley and Sons, ref. 75. Copyright 2016.

In 2016, Du et al.74 developed a nanostructured sensor based on single-walled carbon nanotubes (SWCNTs) for salivary glucose sensing. The screen-printed sensor chip was covered with SWCNTs at first. Subsequently, a chitosan/gold nanoparticles/glucose oxidase (CS/GNp/GOD) thin film was deposited in a layer-by-layer (LBL) fashion on a functional electrode (Pt). Compared with conventional CNTs, SWCNTs can transfer electrons directly, which improved the sensitivity of the sensor. The LBL self-assembly was a simple technology and had good uniformity, stability and sensitivity. Experiments demonstrated a low working potential in this sensor, with high sensitivity and accuracy, serving as reliable saliva glucose detection. Liu et al.75 in 2016 developed a disposable saliva biosensor. The work electrode was modified by an Os complex, MWCNT (multiwalled carbon nanotube) and HRP (horseradish peroxidase) (Fig. 4d). The sensor was highly sensitive to monitor the concentration of saliva glucose, and can achieve two hours of continuous monitoring with good stability. The catalytic current of the biosensor showed a linear dependence upon glucose bulk concentration from 0.05 mM to 1.5 mM (R = 0.998). The limit of detection was 0.003 mM. By adding ascorbic acid, uric acid, lactic acid as the interference, the authors verified the selectivity of the sensor. Another salivary uric acid sensor similarly modified CNTs and uricase for screening the printed electrodes (Fig. 4b). The CNTs and uricase can detect uric acid concentration in saliva samples. The experimental data were consistent with the results of clinical analysis.76

Next is a sensor based on graphene materials. In 2013, Ye et al.77 reported a novel non-enzymatic amperometric glucose sensor based on a copper oxide nanoneedle/graphene/carbon nanofiber modified glass carbon electrode. These nanostructures intercross to form an electronic network that greatly enhanced the activity of glucose detection. The sensor presented a rapid response (<2 s), a very low detection limit (0.1 μM) and linear range (1 μA–5.3 mM). Tang et al.78 constructed a novel electrochemical sensor for interleukin-8 (IL-8) detection based on Fe3O4/graphene oxide (GO)/molecularly imprinted polymer (MIP) nanoparticles. Graphene and GO generally perform well in saliva tests, especially glucose tests. Graphene-based glucose sensors provide powerful capabilities of monitoring the blood glucose concentration in diabetic patients.79 In addition, reduced graphene oxide (rGO) nanocomposites were often used in saliva tests. A biosensor was fabricated by Ag-rGO nanometer complex and modified with myoglobin to detect luteinizing hormone and follicle stimulating hormone in saliva from children.80 Another biosensor based on GO and chitosan was in development for salivary glucose detection.81

Nano-structured metal oxide is a special non-enzyme sensor material, which greatly improves the sensitivity and accuracy of sensor detection. In 2016, Kumar et al. reported a saliva sensor for oral cancer detection. The biosensor showed a high sensitivity (9.28 μA mL ng−1 cm−2), wide linear detection range (2–18 ng mL−1) and rapid response speed (15 min). Nanostructured hafnium oxide (nHfO2) was deposited on indium tin oxide (ITO)-coated glass, followed with 3-amino propyl triethoxy silane (APTES) on the surface of an antibody (anti-CYFRA-21-1). Bovine serum albumin (BSA) serves to block the specific loci of the electrode surface. The detailed preparation process of the immunoelectrode is in Fig. 4c.82 By combining electrospinning and chemical bath deposition, Wang et al. fabricated IrO2 (iridic oxide) conductive core and NiO (nickel oxide)-based nanoscale peel, and then designed a layered shell core of IrO2@NiO NWs. Experiments showed that this shell core structure had a good non-enzyme detection effect on glucose.83 One and two-dimensional zinc oxide nanomaterials, as non-enzymatic saliva cortisol sensor materials, have a low surface area and surface charge density mismatched with bulk materials. Experimental studies showed that the sensor had a good correlation with the data measured by traditional methods.84 Similar to sweat, microfluidic sensors have emerged for saliva detection, but they needed to be combined with highly sensitive organic optical detectors for protein markers in saliva.85 The microfluidics biosensor was composed of an optical element made from a polythiophene-C70 heterojunction with optical activity. It was used in the detection of protein markers in human saliva in conjunction with a highly sensitive organic photodetector (OPDs) (Fig. 4e).

3.4 Tear detection

Studies of tear detection mostly focus on contact lens sensors. The associated mechanical performance of the device platform is poor on the basis of a rigid electronic platform. To solve this problem, researchers used a “rigid island” design to protect the rigid devices, and obtained good mechanical performance. The contact lens glucose sensor adopts such method. Silver nanowires served as antennas and interconnections.38 In contrast, flexible materials such as graphene and CNTs are preferable here for contact lens sensors. In 2017, Kim et al.86 developed a wireless contact lens sensor for measuring glucose and intraocular pressure. A graphene/silver nanowire (Ag NW) hybrid structure was formed in a random network of silver nanowires, which improved electrical and mechanical properties without sacrificing transparency (Fig. 5a). The hybrid materials could be used as a stretchable transparent source/drain electrode for field effect transistors (FETs), with graphene as the device channels. The resistance change was lower than 10% because of the large elasticity of graphene and silver nanowires. The hybrid resistance was kept stable (ΔR < 6%) even after 5000 cycles of stretching and relaxation.
image file: d0nr07478k-f5.tif
Fig. 5 Tear detection with various micro/nanostructured biosensors (MNBS). (a) Wireless contact lens glucose sensor based on graphene/AgNW hybrid.86 (b) Microfluidic systems embedded in commercial contact lenses to prepare sensors to detect pH, glucose, protein, and nitrite ions in tears.90 (c) Flexible, wearable microfluidic contact lens with capillary networks for glucose, urea and chloride detection in tears.89 (a) was reproduced with permission from Springer Nature, ref. 86. Copyright 2017 (Open Access); (b) was reproduced with permission from Springer Nature, ref. 90. Copyright 2020; (c) was reproduced with permission from Elsevier, ref. 89. Copyright 2020.

It is worth noting that nanotoxicology is an inevitable topic when researchers apply nanomaterials (e.g., graphene, CNT, Ag NW) in clinical applications. Materials usually show nanotoxicology when their sizes turn into nanometers (less than 100 nm).87,88 The high surface-to-volume ratio can make them reactive. Taking advantage of their small size, nanomaterials are hypothesized to penetrate cellular membranes, biological barriers and tissues more efficiently than larger sized materials.118,119 Once nanomaterials are released from humoral detection devices, they are likely to cause great damage to biological cells or tissues. To avoid this problem, possible solutions include:

(1) Adopt an appropriate package strategy to prevent the release of nanomaterials; (2) Adopt a non-in situ body fluid detection method (samples are collected for detection) to avoid direct contact between the human body and detection equipment; (3) Replace the toxic nanomaterials by suitable substitute materials; (4) Modify the size, surface or morphology of nanomaterials to improve their physical and chemical performance. (5) Study the mechanism and toxic effects of nanomaterials on organisms, and provide a safe time limit for the storage and use of the devices.

In 2020, a representative study has recently emerged, which used carbon dioxide (CO2) laser ablation technology to embed a microfluidic system into a contact lens to fabricate a tear sensor for detecting pH, glucose, protein and nitrite ions.89 The microchannel consisted of a central ring with four branches. The sensor was placed in a micro-cavity at the end of the branch (Fig. 5b). The micro-runner device itself was flexible as embedded in contact lenses, serving as a sensor.

In 2020, Yang et al.90 also used microfluidic technology and developed a biocompatible, hydrophilic, UV-curable material that can be used to make contact lenses with capillary networks. Such design entrances made tears spontaneously enter the capillary networks and reservoirs (Fig. 5c). Many studies have been conducted on the contact lens sensors, which can basically meet the requirements in terms of softness and comfort. Studies showed that many properties of contact lens (thickness, size, shape, air permeability, water content) will influence its comfort level when patients wear them. Here, we cited the comments from H. A. Ketelson et al. that “Two important elements contributing to the overall performance of contact lens wear are the wettability and lubrication properties of the lens surfaces. These properties can have a significant impact on lens comfort. As the eyelid sweeps over the ocular surface-about 15 times a minute, lens wearers will be relatively unaware of the presence of the lens and they will be more comfortable if: (1) the lens has the ability to maintain moisture at the surface (wetting) and (2) the lens remains lubricated so that the mechanical friction and the associated pressures from the eyelid do not impact the lens and cornea negatively”.91

In clinical practice, researchers usually characterize the basic properties of contact lens in vitro at first. The characterization parameters include the modulus, sessile-drop contact angle, coefficient of friction, water content, base curve, center thickness and UV filtration of contact lens. Then, the contact lens will be inserted and contacted with eyes. Contact lens dry eye symptoms and bulbar redness were assessed using the 8-item Contact Lens Dry Eye Questionnaire (CLDEQ-8). The perceived vision quality and subjective lens comfort will also be rated at insertion, mid-day and end of the day using Visual Analog Scales (VAS).92,93

On the other hand, a limitation remains on the difficulty to achieve sufficient transparency. Proper material selection and nanostructure design can improve the transparency to some extent. The advantage of microfluidic analysis with nanotechnology is that it can store and detect a small amount of tears, which is worth further exploration. A recently reported microfluidics electrochemical detector can integrate into the nasal bridge pad of the eye to stimulate the tear production. After collecting the tears, the levels of glucose and vitamin in the tears can be measured.41 This design avoided the possible damage caused by direct contact with the eyes, and the tears were exported through the microchannel for detection, which was safer and required slightly less materials.

3.5 Urine detection

There have been a lot of research studies on the new nanostructured urine sensor, with various carbon materials, nano-metal oxides and composite materials. Nanocarbons (NCS) are commonly used to modify the interface of electrodes, and there have been many studies on urine detection. A uricase/carboxymethyl cellulose dispersible CNT/gold thin film biosensor has been developed for the uric acid component in human urine and blood samples.94 As a surfactant, carboxymethyl cellulose enabled the CNT to be well dispersed in aqueous solution with excellent sensor performance, and a high sensitivity (233 μA mM−1 cm−2) at low potential (0.35 V). In 2016, Afkhami et al.95 studied a nano-gold/MWCNTs modified sensor electrode for the detection of sodium polychlorophenate. In the experiments, the concentrations of sodium diclofenate in urine and drug samples were measured, and the performance of the sensor before and after modification was compared. The modified sensor showed excellent analytical performance.

Ramonas et al.96 in 2019 developed a highly sensitive and rapid urine glycerol sensor prepared by CNTs-modified graphite electrode and redox medium tetrathiopentyl immobilized pseudomonas alcohol dehydrogenase. The integration scheme and chemical reaction principle of the glycerine sensor are shown in Fig. 6f. This biosensor showed a high sensitivity to glycerol (29.2 ± 0.9 μA mM−1 cm−2), low detection limit (18 μM). There are also many studies on graphene and reduced graphene oxide. Zhang et al.97 fabricated a simple glassy carbon electrode modified by reduced graphene oxide (rGO) and zinc oxide composite material. The sensor with these electrodes can be used for the detection of urine ascorbic acid (AA), dopamine (DA) and uric acid (UA). In 2018, Asif et al.98 designed and prepared a kind of rGO material with double hydroxide nanocrystals superimposed on a layer. The preparation method is shown in Fig. 6a. These materials construct sensors that show high performance in detecting dopamine (DA), ascorbic acid (AA), and (uric acid) UA in urine. The limit of detection of DA, AA and UA were 0.1 nM, 13.5 nM and 0.9 nM. In 2017, Baccarin et al.99 designed and prepared a rGO-carbon black-chitosan film (rGO-CB-CTS) composite by oxidization of graphene into GO, and reaction of GO and CB in the chitosan film (Fig. 6b). The composite material was used to modify the glassy carbon electrode to make a urine sensor to simultaneously determine dopamine and Acetaminophen. For dopamine (DA) and paracetamol (PAR), the authors obtained linear analytical curves from 3.2 × 10−6 to 3.2 × 10−5 mol L−1 and from 2.8 × 10−6 to 1.9 × 10−5 mol L−1, respectively. The limit of detection of DA and PAR were 2.0 × 10−7 and 5.3 × 10−8 mol L−1, respectively. In 2016, Ghanbari et al.100 fabricated a novel uric acid sensor electrode. It was modified layer by layer on a glassy carbon electrode with rGO, polyaniline and zinc oxide nanomaterials (Fig. 6c). The figure also showed the structure diagram of the modified electrode surface. DA and UA tests proved the sensor with a good sensing effect. The detection limit of DA and UA were 0.8 nM and 0.042 μM, respectively. In 2018, Ji et al.101 demonstrated a sensor for detecting AA, DA, UA using a screen-printed electrode modified with rGO and gold nanoparticles (Fig. 6e, left). Specifically, the sensor system, which consisted of sensors, detectors and smartphones, was more convenient and cheaper than traditional electrochemical workstations (Fig. 6e, right).

image file: d0nr07478k-f6.tif
Fig. 6 Urine detection with various micro/nanostructured biosensors (MNBS). (a) Preparation of Zn-NiAl LDH/rGO superlattice composites.98 (b) Schematic illustration of the rGO synthesis and the rGO-CB-CTS/GCE sensor fabrication process.99 (c) Preparation of ZnO/PANI/RGO/GCE.100 (d) One-step preparation of GOx-HRP-Cu3(PO4)2 hybrid nanoflowers and the principle of E. coli detection.105 (e) The modification process of the sensor electrode and the composition diagram of the detection system.101 (f) A principal design scheme of the biosensor (left) and a catalytic oxidation of glycerol (right).96 (a) was reproduced with permission from Springer, ref. 98. Copyright 2019; (b), (c), (d), (e) and (f) were reproduced from ref. 99. Copyright 2017, ref. 100. Copyright 2016, ref. 105. Copyright 2018, ref. 101. Copyright 2018 and ref. 96. Copyright 2019, respectively, with permission from Elsevier.

Metal oxide nanomaterials are also commonly used to modify sensor electrodes. For example, urine creatine biosensors consist of a two-dimensional layered nanomaterial Ti3C2TX immobilization enzyme on the electrode surfaces.102 Urine biosensors were modified with mesoporous Co3O4 nanometer tablets for the simultaneous determination of glutamate and uric acid.103,104 In 2018, Li et al.105 fabricated a special inorganic–organic hybrid nanomaterial with a special flower-like structure. The solution containing copper sulfate was added to PBS (phosphate buffered solution) containing glucose oxidase and horseradish peroxidase, incubated at room temperature for 12 h, and then centrifuged and dried to obtain the organic–inorganic hybrid nanoflower (GOx-HRP-Cu3(PO4)2). GOx and HRP were abundant in the hybrid nanomaterials, which can catalyze the redox reaction. As an example, the diagram showed the detection principle of this sensor (Fig. 6d). AMP magainin I, serving as a capturing probe, was fixed on the electrode surface. In the case of viable bacteria being present, the nanophage binds to AMP magainin I and therefore, triggered a reaction in the glucose solution (Fig. 6d). Similar flower-shaped structures were also found in zinc oxide nanoscale flowers.100,106 Besides these attributes, nanozyme based electrochemical sensors can be also effective biosensors.107,108

4 Conclusion and perspectives

This review summarizes the latest progress of TBS and MNBS in detecting biofluids, with highlights on the engineered materials, design constructs, integration schemes and application scenarios that are highly essential to the development of micro-nanostructured devices for biofluid component analysis. Compared with TBS, MNBS showed great advantages in many aspects, such as portability, comfortability, non-invasiveness and sensitivity. Additionally, some types of MNBS can be integrated with flexible materials, and thereby wearable on the body. These MNBS were successfully applied for the sensitive detection of ISF, sweat, saliva, tears and urine. In practical use for biofluid detection, they can effectively improve the accuracy of detection data, reduce the amount of liquid required for detection and shorten the response time, which endow MNBS with great application prospects in medical treatment, sports detection and other aspects.

As shown in Table 3, a comparison was made between TBS and MNBS about their difference in the sensitivity, detection range, response time and detection limit. Taking the glucose detection as an example, the detection sensitivity of MNBS is generally higher than that of TBS. The response time of MNBS (seconds) is generally shorter than that of TBS (tens of seconds). In terms of the minimum detection limit, MNBS also showed superior values in comparison with TBS. From these results, MNBS showed a great application potential in future biofluid detection.

Table 3 Performance comparison of MNBS and TBS
Type Electrode Glucose source Sensitivity (μA mM−1) Detection range (mM) Response time (s) Detection limit (μM) Ref.
TBS Ag/AgCl Sweat 2.35 0–0.2 50 24
Pt Saliva 0.05–1.0 60–180 50 47
Metals Tear 240 0.1–0.6 20 10 113
Microfluidic thread Tear 0.42 0.075–7.5 20 22.2 114
Pt Saliva ≈0.47 0.00175–10 180 5.34 115
MNBS Co3N NW/TM Glucose 3325.6 0.00015–2.5 5 0.05 65
Co3O4 nanofibers Glucose 1440 0.005–12 2 0.08 116
Au/rGO/AuPtNP Sweat 48 0–2.4 20 5 58
CNT/AuNS Sweat 10.89 0–0.3 10 1.3 63
Nafion/CuO Tear 850 0.003–0.7 12 s 2.99 117
PEDOT/GP-hybrid Sweat 1.0 0.01–0.7 10 59
Graphene/Ag NW Tear 0.001–10 1 86

Although a series of significant advances have been made in MNBS for biofluid detection, there is still spacious room for their further improvement. The main concerns are summarized as follows:

4.1 Signal credibility

In practical usage, the amount of detected biofluids on the body surface is very small with a range of microliter to milliliter, and the analyte concentration is usually low. Sometimes, researchers have to measure biofluid (e.g., sweat) in motion mode. These factors will greatly lower the signal credibility in this context. To meet these requirements, improvement should be made in biomarkers, device structure, sample chamber and detection technology.

4.2 Miniaturization

Both TBS and MNBS encounter the limitation on the mismatch between the device dimension and targeted bio-tissues. Such issue will reduce portability and comfortability, increase structure and operation complexity, or even bring non-negligible damage to the detection position, such as eyes. Meanwhile, the continuous improvement of material science and micro-nano machining technology miniaturize the biosensors. On the other hand, miniature dimension means reducing the measurement liquid amount, which put forward a higher requirement on sensitivity and accuracy. Therefore, the miniaturized biosensor with expected sensitivity and accuracy will be a development target in the future.

4.3 Multifunctionality

Single function will obviously limit the biosensor's application scenes and lower their conveniences. Each biofluid usually contains multiple analytes. A simple, single device is preferable to achieve simultaneous detection with multifunctionality, which will greatly lower the cost in practical application and provide conveniences for users. In the future, the integration of sensing units with the ability of detecting multiple analytes will be a trend for biofluid detection.

4.4 Disposability

Disposable testing equipment is very attractive for POCT (point-of-care testing), because they do not need to consider external environment interference and other factors. For disposable equipment, cost is the issue. At present, carbon-based materials (diamond, carbon nanotubes and graphene) are the central materials for the preparation of disposable micro/nano-electronic systems. However, the preparation of standard micro/nanostructured materials often requires higher costs. The cost of the sensors can be controlled by producing smaller devices on a larger substrate.109 Finally, of course, the power supply and data transmission issues must be addressed.

4.5 Smart healthcare

The development of materials science, microfluidic fabrication and electronic technique has greatly promoted the progress of body fluid detection technology. Nanostructured materials have made a significant contribution to the development of wearable sensing devices. The addition of metal nanoparticles, CNTs, semiconductor nanofilms and other materials has improved the sensing performance of the electrochemical and optical biosensors. In the coming years, the use of nanomaterials or the design of nanostructures, particularly 3D micro/nanostructures, remains a promising research field for biosensors. The future goal should be in the development of wearable devices that can conduct real-time and reliable health monitoring, which can not only be used for clinical detection and treatment, but also be applicable to daily health monitoring and exercise monitoring.

At present, most of the medical devices designed to collect and analyze the contents from ISF, sweat, saliva and tears still stay at the laboratory level. There is a long way to realize their clinical application.6,110,111 Taking the glucose sensor as an example, there are only a few devices that are approved by the FDA (Food and Drug Administration), e.g., MiniMed 670G Guardian Sensor and G4 Platinum produced by Medtronic, G6 Mobile produced by Dexcom, FreeStyle Libre produced by Abbott.16,112 For the newly-designed biofluid sensors, there are many factors that should be further considered in clinical usage as follows (4.6–4.8):

4.6 Accuracy

In actual usage, the response of TBS or MNBS can be influenced by external factors from sunlight, humidity and temperature in unpredictable ways. Meanwhile, pH, ionic concentration and the chemical component of biofluids, disease/physiological state can also cause a negative impact on sensors. In order to guarantee a high accuracy, factory calibration or user calibration against reference standards is necessary after a long-term and repeated operation.

4.7 New materials and package

Time variances existed at the interfaces between different components and target tissues when the body generated movement, which brings mechanical failure and measurement error. In this respect, new interface materials and packaging strategies should be made for sensor components. The sensor components must directly contact with target biofluids, the supporting electronic components should be kept away from them.

4.8 Biocompatibility

Biocompatibility is a key parameter to make TBS or MNBS safely attach onto the skin surface, or in a body. A long-term surface attachment or in vivo implantation is usually necessary for continuous monitoring of biofluid contents (e.g., glucose). In these or broader contexts, skin irritation, inflammation, tissue damage and severe foreign body responses are obviously undesired. To obtain the desired biocompatibility,120 many factors should be taken into comprehensive consideration, including but not limited to: device shape, size, texture, material, surgery and properties of the biointerface of biofluid sensors. Reasonable structure design, proper material selection and precise surgery operation will boost the clinical application.

Conflicts of interest

The authors declare that they have no competing interests.


This work was supported by the Beijing Advanced Innovation Center for Biomedical Engineering, National Natural Science Foundation of China (Grant No. 32071407 and 62003023), the 111 Project (No. B13003), City University of Hong Kong (Grant No. 9610423 and 9667199).

Notes and references

  1. Q. Zheng, Q. Tang, Z. L. Wang and Z. Li, Nat. Rev. Cardiol., 2020, 1–15 Search PubMed.
  2. D. J. Jiang, B. J. Shi, H. Ouyang, Y. B. Fan, Z. L. Wang and Z. Li, ACS Nano, 2020, 14, 6436–6448 CrossRef CAS.
  3. Y. Zou, P. Tan, B. Shi, H. Ouyang, D. Jiang, Z. Liu, H. Li, M. Yu, C. Wang, X. Qu, L. Zhao, Y. Fan, Z. L. Wang and Z. Li, Nat. Commun., 2019, 10, 2695 CrossRef.
  4. H. Ouyang, Z. Liu, N. Li, B. Shi, Y. Zou, F. Xie, Y. Ma, Z. Li, H. Li, Q. Zheng, X. Qu, Y. Fan, Z. L. Wang and Z. Li, Nat. Commun., 2019, 10, 1821 CrossRef.
  5. H. Li, X. Zhang, L. Zhao, D. J. Jiang, L. Xu, Z. Liu, Y. Wu, K. Hu, M. Zhang, J. Wang, Y. Fan and Z. Li, Nano-Micro Lett., 2020, 12, 50 CrossRef CAS.
  6. J. Zhao, H. Guo, J. Li, A. J. Bandodkar and J. A. Rogers, TrAC, Trends Anal. Chem., 2019, 1(6), 559–571 CrossRef CAS.
  7. X. Qu, X. Ma, B. Shi, H. Li, L. Zheng, C. Wang, Z. Liu, Y. Fan, X. Chen, Z. Li and Z. L. Wang, Adv. Funct. Mater., 2020, 2006612 Search PubMed.
  8. J. Sun, A. Yang, C. Zhao, F. Liu and Z. Li, Sci. Bull., 2019, 64, 1336–1347 CrossRef.
  9. B. Shi, Z. Liu, Q. Zheng, J. Meng, H. Ouyang, Y. Zou, D. Jiang, X. Qu, M. Yu, L. Zhao, Y. Fan, Z. L. Wang and Z. Li, ACS Nano, 2019, 13, 6017–6024 CrossRef CAS.
  10. Y. Shi, R. Liu, L. He, H. Feng, Y. Li and Z. Li, Smart Mater. Med., 2020, 1, 131–147 CrossRef.
  11. J. Heikenfeld, A. Jajack, B. Feldman, S. W. Granger, S. Gaitonde, G. Begtrup and B. A. Katchman, Nat. Biotechnol., 2019, 37, 407–419 CrossRef CAS.
  12. G. Matzeu, L. Florea and D. Diamond, Sens. Actuators, B, 2015, 211, 403–418 CrossRef CAS.
  13. M. Dervisevic, M. Alba, B. Prieto-Simon and N. H. Voelcker, Nano Today, 2020, 30, 100828 CrossRef CAS.
  14. G. Maduraiveeran, M. Sasidharan and V. Ganesan, Biosens. Bioelectron., 2018, 103, 113–129 CrossRef CAS.
  15. Z. Farka, T. Jurik, D. Kovar, L. Trnkova and P. Skladal, Chem. Rev., 2017, 117, 9973–10042 CrossRef CAS.
  16. Y. J. Heo and S.-H. Kim, Appl. Sci., 2019, 9, 2158 CrossRef CAS.
  17. Y. J. Heo, H. Shibata, T. Okitsu, T. Kawanishi and S. Takeuchi, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 13399–13403 CrossRef CAS.
  18. T. D. La Count, A. Jajack, J. Heikenfeld and G. B. Kasting, J. Pharm. Sci., 2019, 108, 364–371 CrossRef CAS.
  19. S. Delbeck, T. Vahlsing, S. Leonhardt, G. Steiner and H. M. Heise, Anal. Bioanal. Chem., 2019, 411, 63–77 CrossRef CAS.
  20. G. Rao, R. H. Guy, P. Glikfeld, W. R. LaCourse, L. Leung, J. Tamada, R. O. Potts and N. Azimi, Pharm. Res.-Dordr., 1995, 12, 1869–1873 CrossRef CAS.
  21. M. Caldara, C. Colleoni, E. Guido, V. Re and G. Rosace, Sens. Actuators, B, 2016, 222, 213–220 CrossRef CAS.
  22. H. Li, L. Zhao, J. Meng, C. Pan, Y. Zhang, Y. Zhang, Z. Liu, Y. Zou, Y. Fan, Z. L. Wang and Z. Li, Nano Today, 2020, 33, 100873 CrossRef CAS.
  23. J. R. Windmiller, A. J. Bandodkar, S. Parkhomovsky and J. Wang, The Analyst, 2012, 137, 1570 RSC.
  24. W. Gao, S. Emaminejad, H. Y. Y. Nyein, S. Challa, K. Chen, A. Peck, H. M. Fahad, H. Ota, H. Shiraki, D. Kiriya, D. Lien, G. A. Brooks, R. W. Davis and A. Javey, Nature, 2016, 529, 509 CrossRef CAS.
  25. S. Imani, A. J. Bandodkar, A. M. V. Mohan, R. Kumar, S. Yu, J. Wang and P. P. Mercier, Nat. Commun., 2016, 7, 11650 CrossRef CAS.
  26. M. Gamella, S. Campuzano, J. Manso, G. Gonzlez De Rivera, F. Lopez-Colino, A. J. Reviejo and J. M. Pingarron, Anal. Chim. Acta, 2014, 806, 1–7 CrossRef CAS.
  27. B. M. Kuehn, J. Am. Med. Assoc., 2016, 316(3), 255–257 CrossRef.
  28. P. C. Ferreira, V. N. Ataide, C. L. S. Chagas, L. Angnes, W. K. T. Coltro, T. R. L. C. Paixão and W. R. de Araujoet, TrAC, Trends Anal. Chem., 2019, 119, 115622 CrossRef CAS.
  29. J. Heikenfeld, Electroanalysis, 2016, 28, 1242–1249 CrossRef CAS.
  30. D. Bruen, C. Delaney, L. Florea and D. Diamond, Sensors, 2017, 17(8), 1866 CrossRef.
  31. A. M. V. Mohan, V. Rajendran, R. K. Mishra and M. Jayaraman, TrAC, Trends Anal. Chem., 2020, 131, 116024 CrossRef CAS.
  32. E. Ono, H. Murota, Y. Mori, Y. Yoshioka, Y. Nomura, T. Munetsugu, H. Yokozeki and I. Katayama, PLoS One, 2018, 13(4), e0195960 CrossRef.
  33. D. Morris, S. Coyle, Y. Wu, K. T. Lau, G. Wallace and D. Diamond, Sens. Actuators, B, 2009, 139(1), 231–236 CrossRef CAS.
  34. M. J. Patterson, S. D. R. Galloway and M. A. Nimmo, Exp. Physiol., 2000, 85(6), 869–875 CrossRef CAS.
  35. D. Granger, M. Marsolais, J. Burry and R. Laprade, Am. J. Physiol.: Cell Physiol., 2003, 285(5), C1047–C1058 CrossRef CAS.
  36. S. R. Corrie, J. W. Coffey, J. Islam, K. A. Markey and M. A. F. Kendall, Analyst, 2015, 140(13), 4350–4364 RSC.
  37. N. M. Farandos, A. K. Yetisen, M. J. Monteiro, C. R. Lowe and S. H. Yun, Adv. Healthcare Mater., 2015, 4, 792–810 CrossRef CAS.
  38. J. Park, J. Kim, S. Kim, W. H. Cheong, J. Jang, Y. Park, K. Na, Y. Kim, J. H. Heo, C. Y. Lee, J. H. Lee, F. Bien and J. Park, Sci. Adv., 2018, 4, eaap9841 CrossRef.
  39. J. D. Lane, D. M. Krumholz, R. A. Sack and C. Morris, Curr. Eye Res., 2006, 31, 895–901 CrossRef CAS.
  40. J. T. Baca, D. N. Finegold and S. A. Asher, Clin. Sci., 2007, 5, 280–293 Search PubMed.
  41. J. R. Sempionatto, L. C. Brazaca, L. Garcia-Carmona, G. Bolat, A. S. Campbell, A. Martin, G. Tang, R. Shah, R. K. Mishra, J. Kim, V. Zucolotto, A. Escarpa and J. Wang, Biosens. Bioelectron., 2019, 137, 161–170 CrossRef CAS.
  42. R. Sashikumar and R. Kannan, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endodontol., 2010, 109, 706–711 CrossRef.
  43. S. Gupta, S. V. Sandhu, H. Bansal and D. Sharma, J. Diabetes Sci. Technol., 2014, 9, 91–96 CrossRef.
  44. B. N. V. S. Satish, P. Srikala, B. Maharudrappa, S. M. Awanti, P. Kumar and D. Hugar, J. Int. Oral Health, 2014, 6, 114–117 CAS.
  45. V. Kadashetti, R. Baad, N. Malik, K. M. Shivakumar, N. Vibhute, U. Belgaumi, S. Gugawad and R. C. Pramod, Rom. J. Intern. Med., 2015, 53, 248–252 Search PubMed.
  46. J. Kim, S. Imani, W. R. de Araujo, J. Warchall, G. Valdés-Ramírez, T. R. L. C. Paixão, P. P. Mercier and J. Wang, Biosens. Bioelectron., 2015, 74, 1061–1068 CrossRef CAS.
  47. T. Arakawa, Y. Kuroki, H. Nitta, P. Chouhan, K. Toma, S. Sawada, S. Takeuchi, T. Sekita, K. Akiyoshi, S. Minakuchi and K. Mitsubayashi, Biosens. Bioelectron., 2016, 84, 106–111 CrossRef CAS.
  48. B. Ciui, M. Tertis, C. N. Feurdean, A. Ilea, R. Sandulescu, J. Wang and C. Cristea, Sens. Actuators, B, 2019, 281, 399–407 CrossRef CAS.
  49. C. Bartlett, S. Taylor, C. Fernandez, C. Wanklyn, D. Burton, E. Enston, A. Raniczkowska, M. Black and L. Murphy, Chem. Cent. J., 2016, 10, 1–9 CrossRef.
  50. C. Wanklyn, D. Burton, E. Enston, C. Bartlett, S. Taylor, A. Raniczkowska, M. Black and L. Murphy, Chem. Cent. J., 2016, 10, 1–11 CrossRef.
  51. J. Liu, Z. Geng, Z. Fan, J. Liu and H. Chen, Biosens. Bioelectron., 2019, 132, 17–37 CrossRef CAS.
  52. K. Guo, P. Chen, C. Lin, C. Chen, I. Lee and Y. Yeh, ACS Sens., 2018, 3, 744–748 CrossRef CAS.
  53. A. J. Bandodkar, I. Jeerapan and J. Wang, ACS Sens., 2016, 1, 464–482 CrossRef CAS.
  54. J. M. Yuen, N. C. Shah, J. T. Walsh, M. R. Glucksberg and R. P. Van Duyne, Anal. Chem., 2010, 82, 8382–8385 CrossRef CAS.
  55. K. Ma, J. M. Yuen, N. C. Shah, J. T. Walsh, M. R. Glucksberg and R. P. Van Duyne, Anal. Chem., 2011, 83, 9146–9152 CrossRef CAS.
  56. S. Hwang, C. H. Lee, H. Cheng, J. Jeong, S. Kang, J. Kim, J. Shin, J. Yang, Z. Liu, G. A. Ameer, Y. Huang and J. A. Rogers, Nano Lett., 2015, 15, 2801–2808 CrossRef CAS.
  57. J. Li, R. Li, H. Du, Y. Zhong, Y. Chen, K. Nan, S. M. Won, J. Zhang, Y. Huang and J. A. Rogers, ACS Nano, 2018, 13, 660–670 CrossRef.
  58. X. Xuan, H. S. Yoon and J. Y. Park, Biosens. Bioelectron., 2018, 109, 75–82 CrossRef CAS.
  59. H. Lee, T. K. Choi, Y. B. Lee, H. R. Cho, R. Ghaffari, L. Wang, H. J. Choi, T. D. Chung, N. Lu, T. Hyeon, S. H. Choi and D. Kim, Nat. Nanotechnol., 2016, 11, 566–572 CrossRef CAS.
  60. H. Lee, C. Song, Y. S. Hong, M. S. Kim, H. R. Cho, T. Kang, K. Shin, S. H. Choi, T. Hyeon and D. Kim, Sci. Adv., 2017, 3(3), e1601314 CrossRef.
  61. Y. Yang, Y. Song, X. Bo, J. Min, O. S. Pak, L. Zhu, M. Wang, J. Tu, A. Koganm, H. Zhang, T. K. Hsiai, Z. Li and W. Gao, Nat. Biotechnol., 2020, 38(2), 217–224 CrossRef CAS.
  62. D. Hwang, S. Lee, M. Seo and T. D. Chung, Anal. Chim. Acta, 2018, 1033, 1–34 CrossRef CAS.
  63. S. Y. Oh, S. Y. Hong, Y. R. Jeong, J. Yun, H. Park, S. W. Jin, G. Lee, J. H. Oh, H. Lee, S. Lee and J. S. Ha, ACS Appl. Mater. Interfaces, 2018, 10, 13729–13740 CrossRef CAS.
  64. A. Abellán-Llobregat, I. Jeerapan, A. Bandodkar, L. Vidal, A. Canals, J. Wang and E. Morallón, Biosens. Bioelectron., 2017, 91, 885–891 CrossRef.
  65. F. Xie, X. Cao, F. Qu, A. M. Asiri and X. Sun, Sens. Actuators, B, 2018, 255, 1254–1261 CrossRef CAS.
  66. A. Koh, D. Kang, Y. Xue, S. Lee, R. M. Pielak, J. Kim, T. Hwang, S. Min, A. Banks, P. Bastien, M. C. Manco, L. Wang, K. R. Ammann, K. Jang, P. Won, S. Han, R. Ghaffari, U. Paik, M. J. Slepian, G. Balooch, Y. Huang and J. A. Rogers, Sci. Transl. Med., 2016, 8(366), 366ra165 CrossRef.
  67. S. B. Kim, Y. Zhang, S. M. Won, A. J. Bandodkar, Y. Sekine, Y. Xue, J. Koo, S. W. Harshman, J. A. Martin, J. M. Park, T. R. Ray, K. E. Crawford, K. T. Lee, J. Choi, R. L. Pitsch, C. C. Grigsby, A. J. Strang, Y. Y. Chen, S. Xu, J. Kim, A. Koh, J. S. Ha, Y. Huang, S. W. Kim and J. A. Rogers, Small, 2018, 14(12), 1703334 CrossRef.
  68. A. J. Bandodkar, P. Gutruf, J. Choi, K. Lee, Y. Sekine, J. T. Reeder, W. J. Jeang, A. J. Aranyosi, S. P. Lee, J. B. Model, R. Ghaffari, C. J. Su, J. P. Leshock, T. Ray, A. Verrillo, K. Thomas, V. Krishnamurthi, S. Han, J. Kim, S. Krishnan, T. Hang and J. A. Rogers, Sci. Adv., 2019, 5, eaav3294 CrossRef.
  69. J. T. Reeder, J. Choi, Y. Xue, P. Gutruf, J. Hanson, M. Liu, T. Ray, A. J. Bandodkar, R. Avila, W. Xia, S. Krishnan, S. Xu, K. Barnes, M. Pahnke, R. Ghaffari, Y. Huang and J. A. Rogers, Sci. Adv., 2019, 5, eaau6356 CrossRef.
  70. Y. Song, J. Min, Y. Yu, H. Wang, Y. Yang, H. Zhang and W. Gao, Sci. Adv., 2020, 6, eaay9842 CrossRef CAS.
  71. H. Nyein, L. C. Tai, Q. P. Ngo, M. Chao, G. B. Zhang, W. Gao, M. Bariya, J. Bullock, H. Kim, H. M. Fahad and A. Javey, ACS Sens., 2018, 3, 944–952 CrossRef CAS.
  72. M. S. Mannoor, H. Tao, J. D. Clayton, A. Sengupta, D. L. Kaplan, R. R. Naik, N. Verma, F. G. Omenetto and M. C. McAlpine, Nat. Commun., 2012, 3, 763 CrossRef.
  73. C. Luo, Y. Wang, X. Li, X. Jiang, P. Gao, K. Sun, J. Zhou, Z. Zhang and Q. Jiang, Nanomaterials, 2017, 7, 67 CrossRef.
  74. Y. Du, W. Zhang and M. L. Zhang, Biosensors, 2016, 6, 10 CrossRef.
  75. J. Liu, S. Sun, H. Shang, J. Lai and L. Zhang, Electroanalysis, 2016, 28, 2016–2021 CrossRef CAS.
  76. W. Shi, J. Li, J. Wu, Q. Wei, C. Chen, N. Bao, C. Yu and H. Gu, Anal. Bioanal. Chem., 2020, 412(26), 7275–7283 CrossRef CAS.
  77. D. Ye, G. Liang, H. Li, J. Luo, S. Zhang, H. Chen and J. Kong, Talanta, 2013, 116, 223–230 CrossRef CAS.
  78. P. Tang, H. Zhang, J. Huo and X. Lin, Anal. Methods, 2015, 7, 7784–7791 RSC.
  79. V. K. L. Shanbhag and K. S. Prasad, Anal. Methods, 2016, 8, 6255–6259 RSC.
  80. L. A. D. Gugoasa, R. Stefan-van Staden, J. F. van Staden, M. Coros and S. M. Pruneanu, Anal. Bioanal. Chem., 2020, 412, 5191–5202 CrossRef CAS.
  81. A. Rabti, W. Argoubi and N. Raouafi, Microchim. Acta, 2016, 183, 1227–1233 CrossRef CAS.
  82. S. Kumar, S. Kumar, S. Tiwari, S. Augustine, S. Srivastava, B. K. Yadav and B. D. Malhotra, Sens. Actuators, B, 2016, 235, 1–10 CrossRef CAS.
  83. J. Wang, L. Xu, Y. Lu, K. Sheng, W. Liu, C. Chen, Y. Li, B. Dong and H. Song, Anal. Chem., 2016, 88, 12346–12353 CrossRef CAS.
  84. P. K. Vabbina, A. Kaushik, N. Pokhrel, S. Bhansali and N. Pala, Biosens. Bioelectron., 2015, 63, 124–130 CrossRef CAS.
  85. T. Dong and N. M. M. Pires, Biosens. Bioelectron., 2017, 94, 321–327 CrossRef CAS.
  86. J. Kim, M. Kim, M. Lee, K. Kim, S. Ji, Y. Kim, J. Park, K. Na, K. Bae, H. K. Kim, F. Bien, C. Y. Lee and J. Park, Nat. Commun., 2017, 8, 14997 CrossRef.
  87. A. V. Singh, P. Laux, A. Luch, C. Sudrik, S. Wiehr, A.-M. Wild, G. Santomauro, J. Bill and M. Sitti, Toxicol. Mech. Methods, 2019, 29(5), 378–387 CrossRef CAS.
  88. H. F. Krug and P. Wick, Angew. Chem., Int. Ed., 2011, 50, 1260–1278 CrossRef CAS.
  89. R. Moreddu, J. S. Wolffsohn, D. Vigolo and A. K. Yetisen, Sens. Actuators, B, 2020, 317, 128183 CrossRef CAS.
  90. X. Yang, H. Yao, G. Zhao, G. A. Ameer, W. Sun, J. Yang and S. Mi, J. Mater. Sci., 2020, 55, 9551–9561 CAS.
  91. H. A. Ketelson, S. S. Perry, W. G. Sawyer and J. T. Jacob, Contact Lens Spectrum, 2011, 26, 30–36 Search PubMed.
  92. M. Vidal-Rohr, J. S. Wolffsohn, L. N. Davies and A. Cerviño, Contact Lens Anterior, 2018, 41(1), 117–121 CrossRef.
  93. U. Stahl, M. D. P. Willcox, T. Naduvilath and F. Stapleton, Optom. Vis. Sci., 2009, 86(7), 857–867 CrossRef.
  94. T. Fukuda, H. Muguruma, H. Iwasa, T. Tanaka, A. Hiratsuka, T. Shimizu, K. Tsuji and T. Kishimoto, Anal. Biochem., 2020, 590, 113533 CrossRef CAS.
  95. A. Afkhami, A. Bahiraei and T. Madrakian, Mater. Sci. Eng., C, 2016, 59, 168–176 CrossRef CAS.
  96. E. Ramonas, D. Ratautas, M. Dagys, R. Meškys and J. Kulys, Talanta, 2019, 200, 333–339 CrossRef CAS.
  97. X. Zhang, Y. Zhang and L. Ma, Sens. Actuators, B, 2016, 227, 488–496 CrossRef CAS.
  98. M. Asif, A. Aziz, H. Wang, Z. Wang, W. Wang, M. Ajmal, F. Xiao, X. Chen and H. Liu, Microchim. Acta, 2019, 186, 61 CrossRef.
  99. M. Baccarin, F. A. Santos, F. C. Vicentini, V. Zucolotto, B. C. Janegitz and O. Fatibello-Filho, J. Electroanal. Chem., 2017, 799, 436–443 CrossRef CAS.
  100. K. Ghanbari and M. Moloudi, Anal. Biochem., 2016, 512, 91–102 CrossRef CAS.
  101. D. Ji, Z. Liu, L. Liu, S. S. Low, Y. Lu, X. Yu, L. Zhu, C. Li and Q. Liu, Biosens. Bioelectron., 2018, 119, 55–62 CrossRef CAS.
  102. S. Hroncekova, T. Bertok, M. Hires, E. Jane, L. Lorencova, A. Vikartovska, A. Tanvir, P. Kasak and J. Tkac, Processes, 2020, 8, 580 CrossRef CAS.
  103. M. M. Hussain, M. M. Rahman, A. M. Asiri and M. R. Awual, RSC Adv., 2016, 6, 80511–80521 RSC.
  104. N. Lavanya, E. Fazio, F. Neri, A. Bonavita, S. G. Leonardi, G. Neri and C. Sekar, J. Electroanal. Chem., 2016, 770, 23–32 CrossRef CAS.
  105. Y. Li, G. Xie, J. Qiu, D. Zhou, D. Gou, Y. Tao, Y. Li and H. Chen, Sens. Actuators, B, 2018, 258, 803–812 CrossRef CAS.
  106. M. Govindasamy, S. Wang, B. Subramanian, R. J. Ramalingam, H. Al-lohedan and A. Sathiyan, Ultrason. Sonochem., 2019, 58, 104622 CrossRef CAS.
  107. M. Khairy, H. A. Ayoub and C. E. Banks, Food Chem., 2018, 255, 104–111 CrossRef CAS.
  108. M. N. Karim, S. R. Anderson, S. Singh, R. Ramanathan and V. Bansal, Biosens. Bioelectron., 2018, 110, 8–15 CrossRef CAS.
  109. C. Dincer, R. Bruch, E. Costa-Rama, M. T. Fernandez-Abedu, A. Merkoci, A. Manz, G. A. Urban and F. Gueder, Adv. Mater., 2019, 31, 1806739 CrossRef.
  110. Y. Yang and W. Gao, Chem. Soc. Rev., 2019, 48(6), 1465–1491 RSC.
  111. A. J. Bandodkar, W. J. Jeang, R. Ghaffari and J. A. Rogers, Annu. Rev. Anal. Chem., 2019, 12, 1–22 CrossRef.
  112. H. K. Akturk and S. Garg, Curr. Opin. Endocrinol., 2019, 26(2), 84–89 CrossRef.
  113. H. Yao, A. J. Shum, M. Cowan, I. Lähdesmäki and B. A. Parviz, Biosens. Bioelectron., 2011, 26, 3290–3296 CrossRef CAS.
  114. D. Agustini, M. F. Bergamini and L. H. Marcolino-Junior, Biosens. Bioelectron., 2017, 98, 161–167 CrossRef CAS.
  115. T. Arakawa, K. Tomoto, H. Nitta, K. Toma, S. Takeuchi, T. Sekita, S. Minakuchi and K. Mitsubayashi, Anal. Chem., 2020, 92(18), 12201–12207 CrossRef CAS.
  116. C. Guo, X. Zhang, H. Huo, C. Xu and X. Han, Analyst, 2013, 138, 6727–6731 RSC.
  117. A. Romeo, A. Moya, T. S. Leung, G. Gabriel, R. Villa and S. Sánchez, Appl. Mater. Today, 2018, 10, 133–141 CrossRef.
  118. H. Li, H. Ouyang, M. Yu, N. Wu, X. Wang, W. Jiang, Z. Liu, J. Tian, Y. Jin, H. Feng, Y. Fan and Z. Li, Small, 2017, 13(8), 1603642 CrossRef.
  119. H. Li, L. Zhao, W. Zhu, X. Qu, C. Wang, R. Liu, Y. Fan and Z. Li, ACS Omega, 2019, 4, 16209–16216 CrossRef CAS.
  120. H. Li, C. Zhao, X. Wang, J. Meng, Y. Zou, S. Noreen, L. Zhao, Z. Liu, H. Ouyang, P. Tan, M. Yu, Y. Fan, Z. L. Wang and Z. Li, Adv. Sci., 2019, 6(6), 1801625 CrossRef.


These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2021