A HiPAD Integrated with rGO/MWCNTs Nano‐Circuit Heater for Visual Point‐of‐Care Testing of SARS‐CoV‐2

Abstract

Nowadays, the main obstacle for further miniaturization and integration of nucleic acids point‐of‐care testing devices is the lack of low‐cost and high‐performance heating materials for supporting reliable nucleic acids amplification. Herein, reduced graphene oxide hybridized multi‐walled carbon nanotubes nano‐circuit integrated into an ingenious paper‐based heater is developed, which is integrated into a paper‐based analytical device (named HiPAD). The coronavirus disease 2019 (COVID‐19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) is still raging across the world. As a proof of concept, the HiPAD is utilized to visually detect the SARS‐CoV‐2 N gene using colored loop‐mediated isothermal amplification reaction. This HiPAD costing a few dollars has comparable detection performance to traditional nucleic acids amplifier costing thousands of dollars. The detection range is from 25 to 2.5 × 10¹⁰ copies mL⁻¹ in 45 min. The detection limit of 25 copies mL⁻¹ is 40 times more sensitive than 1000 copies mL⁻¹ in conventional real‐time PCR instruments. The disposable paper‐based chip could also avoid potential secondary transmission of COVID‐19 by convenient incineration to guarantee biosafety. The HiPAD or easily expanded M‐HiPAD (for multiplex detection) has great potential for pathogen diagnostics in resource‐limited settings.

rGO/MWCNTs are first developed into rGO/MWCNTs nano‐circuit heaters as cheap, easy‐to‐use, and flexible electrothermal modules for nucleic acid testing point‐of‐care testing devices. The miniaturized rGO/MWCNTs heater integrated paper‐based analytical device can visually detect SARS‐CoV‐2 with 25 copies mL⁻¹ of detection limit based on colored LAMP reaction in 45 min.

1. Introduction

Point‐of‐care testing (POCT) devices in in vitro diagnostics fields, with a rapid and reliable quantification of biomarkers or pathogens, have attained increasing significance no matter in resource‐rich areas or in resource‐limited settings.^{[ 1 ]} Tremendous technologies, such as microfluidics, lab‐on‐a‐chip devices and lateral flow immunoassays,^{[ 2 ]} have been applied in POCT devices for infectious viruses like HBV,^{[ 3 ]} Zika,^{[ 4 ]} Ebola,^{[ 5 ]} and SARS‐CoV‐2.^{[ 6 ]} However, POCT devices have limited applications in resource‐limited settings owing to complex sample procedures and expensive signal detection equipment.^{[ 7 ]} Therefore, developing low‐cost and easy‐operated POCT devices remains challenge in clinical diagnostics in resource‐limited settings.

Proteins (antigens or antibodies) and nucleic acids (DNA or RNA) are two primary categories of targets for POCT.^{[ 8 ]} Numerous commercial POCT devices are widely applied to detect proteins in serum,^{[ 9 ]} saliva,^{[ 10 ]} and urine,^{[ 11 ]} which are based on the principle of specific antigen–antibody reaction. However, lacking of signal amplification in such POCT devices leads to low detection sensitivity that impedes their applications in clinical early diagnosis. Meanwhile, nucleic acid testing (NAT) based on central signal amplification can detect low levels even at single copy of nucleic acids for early diagnosis.^{[ 12 ]} The NAT involves various polymerase chain reaction (PCR) (like classical PCR, real‐time PCR, multiplex PCR, and nest PCR)^{[ 13 ]} and isothermal amplification methods (such as loop‐mediated isothermal amplification (LAMP), rolling circle amplification (RCA), strand displacement amplification (SDA), and recombinase polymerase amplification (RPA)).^{[ 14 ]} RCA, SDA, and RPA often are carried out at 37 °C, while LAMP is at 60 °C. Basically, nonspecific amplification of LAMP is less than those isothermal methods at 37 °C. Generally, the above mentioned NAT needs independent professional instruments to carry out different steps of sample preparation, nucleic acid amplification and signal detection.^{[ 14a ]} Recently, these steps have been fully or partially integrated into one single NAT device, which can facilitate reducing false positives and improving detection efficiency.^{[ 15 ]} For POCT devices of NAT, one of the major challenges is the miniaturization and integration of multifunctional modules, including nucleic acids extraction, purification, and detection, and temperature modules for nucleic acids amplification. In particular, low‐cost paper‐based devices represent a future development tendency of miniaturized and integrated NAT devices.^{[ 16 ]} Whereas most existing devices need external expensive heaters to complete nucleic acids amplification. Thereby, miniaturization and integration of temperature modules are very important for paper‐based NAT devices.^{[ 17 ]} The paper‐based heater is a promising alternative of miniaturized and integrated electrothermal modules for its portability and low cost. Nevertheless, underwhelming electrothermal performance of existing paper‐based heaters mostly made by expensive screen printers is one of disadvantages impeding the development of paper‐based POCT devices, which needs to be resolved in the future.^{[ 18 ]} Additionally, portable and low‐cost batteries can also be supplied to the paper‐based heaters in POCT devices especially in resource‐limited settings.

Currently, for almost all nucleic acids amplifiers that worth at least several thousand dollars, the commonest temperature modules are electrothermal tubes and Peltier semiconductor heating block,^{[ 19 ]} which are indeed difficult to be miniaturized. Besides, other attempts to miniaturize temperature modules composed of aluminum or copper blocks are also unsatisfying because of complicated process and expensive cost.^{[ 20 ]} Thereby, novel thermal materials with simple fabrication process are essential to miniaturize POCT devices for NAT.

Electrothermal materials including metallic electrothermal materials and non‐metallic electrothermal materials have been applied in industry,^{[ 21 ]} aerospace,^{[ 22 ]} and manufacturing.^{[ 23 ]} Despite their superior electrical and thermal performances, metallic electrothermal materials, such as nichrome^{[ 23 ]} and ferro‐aluminum alloy,^{[ 23 ]} are not suitable for integration into POCT devices due to high prices and fabrication complexity. Non‐metallic electrothermal materials, such as carbon nanotubes,^{[ 24 ]} graphene,^{[ 25 ]} and carbon fibers,^{[ 26 ]} are flexible and easy to process. Among the electrothermal carbon materials, the cheap MWCNTs with high electrothermal conversion^{[ 27 ]} have been applied to smart textiles^{[ 28 ]} and electric cable deicing.^{[ 29 ]} Reduced graphene oxide (rGO) has better electrical and thermal conductivity than graphene oxide (GO),^{[ 30 ]} which have been used in electrothermal heaters,^{[ 31 ]} sensors,^{[ 32 ]} electromechanical actuators.^{[ 33 ]} Lately, the mixture of carbon nanotubes and GO or rGO have been reported to restack and reinforce their electrical and thermal conductivity,^{[ 34 ]} because rGO can be an enhanced conductive dispersant when mixed with carbon nanotubes.^{[ 35 ]} Yet, it has been not reported that hybridized rGO/MWCNTs are developed into electrothermal modules for POCT devices.

Herein, we fabricated a rGO/MWCNTs nano‐circuit electrothermal heater by a simple suction filtration and integrated it into a POCT paper‐based analytical device (named HiPAD) to visually detect nucleic acids by colored LAMP reaction^{[ 36 ]} (Figure  1 ). COVID‐19 pandemic caused by SARS‐CoV‐2 infection as a significant public health problem remains prevalent worldwide and its vaccines is in development.^{[ 37 ]} The POCT devices for rapid detection of SARS‐CoV‐2 are quite in demand, which facilitates to effectively prevent the spread.^{[ 38 ]} Therefore, in this study, SARS‐CoV‐2 was utilized to evaluate the performance of the HiPAD. Based on such rGO/MWNCTs nano‐circuit electrothermal modules, after increasing reaction chambers, the HiPAD can be upgraded for synchronously multiplex POCT in the near future.

Figure 1.

rGO/MWCNTs nano‐circuit heater integrated paper‐based analytical device (HiPAD) is applied for visual nucleic acids POCT. a) Schematic of rGO/MWCNTs nano‐circuit. b) The fabrication and integration of HiPAD. c) The principle of visual detection using HNB as a color indicator of LAMP. dNMP is deoxynucleoside monophosphate. PPi⁻ is pyrophosphate ion. dNTP is composed of dNMP and PPi⁻.

2. Results and Discussion

The electrothermal property of the rGO/MWCNTs determines the final detection performance of the HiPAD. rGO was successfully obtained by hydrothermal treatment of GO at 100 °C for 24 h as characterized by Raman spectra in Figure S1, Supporting Information. The maximum absorption wavelength of obtained rGO was ≈240 nm (Figure S2, Supporting Information). Before vacuum filtering, the rGO/MWCNTs mixed solution was sufficiently ultrasonic mixed and dispersed to ensure the uniformity of batches and reduce manual influence on dispersity before filtering. During vacuum (−0.1 MPa) filtering, microscopically, the process of deposition and stack of rGO and MWCNTs was disordered and irregular; however, macroscopically, the process of different batches was the same, which resulted in the same distribution of rGO and MWCNTs approximately. Hence, the method of preparation had good repeatability. Next, we firstly observed the surface morphology of the prepared rGO/MWCNTs heater. As shown in Figure  2a–e, after suction filtration, different ratios of rGO/MWCNTs were compacted tightly onto filter papers to form stable and robust rGO/MWCNTs films (Figure S3, Supporting Information). From SEM pattern, the films surface gradually became flat with increasing dosage ratio of rGO. Meanwhile, quantitatively, the surface gray scale values (pseudo‐color) of films diminished gradually. Detailedly in Figure 2f, the cross‐contacting structure of rGO sheets and MWCNTs formed nano‐circuits for electron proximity transfer imparting itself a great heat‐conductivity. Mean gray scale values also decreased as the rGO ratio increased in Figure 2g. Then, the relative intrinsic resistance and resistivity were measured in Figure 2h,i, respectively. 5% rGO/MWCNTs heaters had minimum resistance and resistivity. Herein, the rGO plays a role of electronic conduction adhesive for MWCNTs. MWNCTs can be considered as the multi‐layer rolled‐up 2D graphene composed of sp²‐bonded carbon atoms.^{[ 39 ]} A small amount of (5%) of rGO can connect and conduct irregularly curled MWCNTs to form high conductive nano‐circuits (Figure 2f) through π–π stacking and van der Waals force.^{[ 34 ]} Furthermore, by means of the π–π stacking and van der Waals force, the 2D rGO compressed and flattened the fluffy curled 3D MWCNTs (Figure 2a–e). But the excessive addition of rGO (over 10%) agglomerated and impeded electron transport, which resulted in less conductive of the hybridized rGO/MWCNTs. Moreover, the amount of rGO added was negatively correlated with the electrical conductivity of hybridized rGO/MWCNTs (Figure 2h,i).

Figure 2.

The surface morphology and electrothermal properties of rGO/MWCNTs nano‐circuit heater. The SEM pattern and surface gray scale value (pseudo‐color) of rGO/MWCNTs heaters with various rGO dosage ratio (wt./wt.) from a) 0%, b) 5%, c) 10%, d) 15% to e) 20%. f) The SEM pattern of rGO/MWCNTs nano‐circuit. g) The mean surface gray scale value of rGO/MWCNTs heaters with various rGO dosage ratio from 0%, 5%, 10%, 15%, to 20%. h) The percentage change in resistance relative to 0% rGO/MWCNTs heater. i) The percentage change in resistivity compared to 0% rGO/MWCNTs heater. j) At room temperature, the maximum surface temperatures of rGO/MWCNTs heaters with rGO dosage ratio from 0%, 5%, 10%, 15% to 20% at different voltages (1 V to 5 V). k) The infrared thermal images and temperature distribution on the surface of 5% rGO/MWCNTs heater at voltages from 1 to 5 V. l) The infrared thermal images and average temperatures of the surface in the paper‐based chamber over time heated by 5% rGO/MWCNTs heater at 5 V.

The optimum temperature range of LAMP reaction was 60–65 °C. Thus, we need to implement an appropriate voltage to rGO/MWCNTs heaters. Five kinds of rGO/MWCNTs, namely 0–20% rGO/MWCNTs heaters, were electrified at 1 V to 5 V (Figure 2j). The surface temperatures of rGO/MWCNTs heaters rapidly (less than 10 s) raised to equilibrium temperatures with voltage increasing. When 5 V of voltage was applied to the 5% rGO/MWCNTs heater, rGO/MWCNTs heaters would reach over 70 °C. In addition, the infrared thermal images of 5% rGO/MWCNTs heater at different voltages displayed uniform temperature distribution in a larger region than LAMP reaction zone (Figure 2h). Under the same conditions, we also observed that the paper‐based chambers were quickly heated and achieved a stable temperature at ≈60 °C in 1 min (Figure 2i), which satisfied the demand of LAMP reaction temperature. The infrared thermal images proved the electrothermal stability and uniformity of the 5% rGO/MWCNTs heater over a long period of time (from 1 to 60 min) during LAMP reactions. After heating for 60 min under voltage of 5 V, we observed the SEM pattern and measured the resistance and resistivity of the 5% rGO/MWCNTs heater (Figure S4, Supporting Information). We found that the morphology and the electrical performance didn't change obviously (P > 0.05). Therefore, the 5% rGO/MWCNTs heater was selected as an optimal electrothermal module for subsequent LAMP reactions in the HiPAD.

Based on the energy conservation law, the temperature of rGO/MWCNTs heaters could reach an equilibrium when it became balanced among the power of conduction, convection, and radiation.^{[ 28a ]} Accordingly, the electricity‐thermal radiation conversion efficiency of the rGO/MWCNTs heater could be calculated by the following equation based on Stefan Boltzmann law and Joule's law:^{[ 40 ]}

$$\begin{array}{c}{\eta }_r=\frac{S\sigma \left(T_r^4-T_0^4\right)}{P}\times 100\%\end{array}$$ (1)

$$\begin{array}{c}P=\frac{U^2}{R}\end{array}$$ (2)

$$\begin{array}{c}{\eta }_j=1-{\eta }_r\end{array}$$ (3)

where η_r is the theoretical electricity‐thermal radiation conversion efficiency, S is the area of the radiant surface (0.030 mm × 0.027 mm), σ is the Stefan Boltzmann constant (5.67 × 10⁻⁸ W m⁻² K⁻⁴)), T _r is the surface mean radiation temperature (70 °C), T ₀ is the environment temperature (25 °C), U is the voltage (5 V), R is the resistance value of the 5% rGO/MWCNTs heater (16.5 Ω), P is the electric power, η_j is theoretical electrothermal conversion efficiency. After calculation, 5% rGO/MWCNTs heater exhibited a radiant electrothermal conversion efficiency of 18.07% and a conductive electrothermal conversion efficiency of 81.93%.

The infrared thermal imaging is only suitable for surface temperature analysis of paper‐based chambers. In order to explore inner thermal property of the paper‐based chamber in HiPAD, we adopted SolidWorks Premium 2015, a 3D modeling and simulation software, to simulate and analyze the inner thermal distribution of the HiPAD heated by 5% rGO/MWCNTs heater. The major simulation parameters were set up as followed. The 3D model was constructed approximating the physical entity (HiPAD) as shown in Figure S5, Supporting Information. Joint contact was set between modules. Simulation ambient temperature was set at room temperature of 25 °C (298.15 K), with 5 W m⁻² K⁻¹) of natural air convective heat transfer coefficient. The coefficients of thermal conductivity of the electrodes plate, the rGO/MWCNTs heater, the paper‐based chamber, the liquid in the chamber and the sealing glass were respectively set as 0.16, 168, 0.05, 0.61, and 0.75 W m⁻¹ K⁻¹ according to their materials property, respectively. Additionally, the accuracy of simulation model is highly dependent on the nature of the mesh and it increases as the mesh size becomes extra finer using the method of finite element analysis. The software is able to evaluate physical effects according to meshes. At the same time, if the mesh becomes finer, the computational time increases. Hence, the mesh size should be optimized considering the accuracy and the computational time. Herein, the size of the mesh was optimized as 1.0563 ± 0.0528 mm with high quality as shown in Figure  3a, and the Jacobian ratio used to assess the mesh quality of each mesh was close to 1 as shown in Figure S6, Supporting Information. Generally, a Jacobian ratio of 40 or less is acceptable based on stochastic studies.

Figure 3.

The analysis of thermal distribution of the HiPAD heated by rGO/MWCNTs heater simulated by SolidWorks Premium 2015. a) Fine mesh model of the HiPAD using finite element analysis. The temperatures of the rGO/MWCNTs heater are set at b) 25.0 °C (1 V), c) 28.7 °C (2 V), d) 38.2 °C (3 V), e) 51.8 °C (4 V), and f) 70.0 °C (5 V) to simulate the temperatures of liquid in the paper‐based chamber of HiPAD, respectively.

The integral temperature distribution of the HiPAD was simulated by setting the temperature of the 5% rGO/MWCNTs heater at 25.0 (1 V), 28.7 (2 V), 38.2 (3 V), 51.8 (4 V), and 70.0 °C (5 V), respectively, according to Figure 2k. Then, the surface temperature and the temperature of the liquid in the chamber of the HiPAD was simulated and calculated for analysis. From Figure 3b–f, it was clear that the mean temperature of the surface and the liquid in the chamber increased with the temperature of rGO/MWCNTs heater. When rGO/MWCNTs heater was at 70 °C, the mean surface temperature would reach at ≈62.5 °C and the liquid would reach at ≈64.5 °C (Figure 3f), which approximated the actual experimental results (Figure 2i). From side view, uniform temperature distribution was observed in the chamber, which basically met the temperature condition of LAMP reaction.

The LAMP reaction was carried out in the paper‐based chamber in chip of HiPAD (Figure  4a,b), which should be sealed well to hold LAMP reaction mixture preventing from liquid volatilization and contamination. We used Whatman grade 1 chromatography paper to fabricate the chamber. The microstructure of Whatman grade 1 chromatography paper was directly observed by optical microscopy. It was clear that the paper had interwoven fibrous structure (Figure 4c). In addition, the paper was highly hydrophilic (contact angle θ ≈ 0°). After wax immersion, the paper filled with wax became hydrophobic (static contact angle θ = 91.69° ± 0.79°) (Figure 4d) enabling confining LAMP mixture in the chamber (Figure 4e). To test its sealing property, we added 25 µL deionized water into the chamber, then sealed it and put it into an oven at temperature of 60 °C or heated it by the 5% rGO/MWCNTs heater at voltage of 5 V ≈45 min. We found that the relative weight loss of water was less than 5% after 45 min at 60 °C (Figure 4f). Therefore, the chamber could be sealed well to carry out LAMP reaction.

Figure 4.

The characterization of the HiPAD. a) The 3D model of the HiPAD. b) The real prototype of the HiPAD. c) The fiber structure of super‐hydrophilic (contact angle θ ≈ 0°) paper‐based chamber in chip made by Whatman grade 1 chromatography paper observed by optical microscopy. d) The structure of hydrophobic (static contact angle θ = 91.69° ± 0.79°) waxed paper‐based chip made by Whatman grade 1 chromatography paper observed by optical microscopy. e) The top view of the paper‐based chip of the HiPAD. f) Sealing evaluation of the paper‐based chamber. The average relative weight loss of deionized water is less than 5% for 45 min at 60 °C. (*P < 0.05, n = 3).

For POCT, it is better to easily distinguish positive and negative results by obvious color difference with naked eyes. Thereby, we optimized the visual indicator for LAMP reaction from four common dyes including SYBR Green I, GelRed, GeneFinder, and hydroxynaphthol blue (HNB). As shown in Figure S7, Supporting Information, we found that the color difference between negative and positive samples was difficult to distinguish in staining with SYBR Green I and GelRed. And GeneFinder inhibited LAMP reaction resulting in no color change except additional post‐staining after reaction. Nevertheless, the color difference was distinct between negative control and positive samples in staining with HNB. Hence, HNB was employed as an optimal indicator for visual detection of SARS‐CoV‐2 in HiPAD.

SARS‐CoV‐2 was chosen as a target to evaluate the detection performance of the HiPAD or M‐HiPAD (multiplex HiPAD) (Figure  5a). To verify the sensitivity, SARS‐CoV‐2 N gene template was serially diluted from 2.5 × 10¹⁰–2.5 copies mL⁻¹ (per 25 µL reaction) at tenfold intervals for LAMP reactions. LAMP mixture containing HNB as a color indicator was added into the chambers. Then, the HiPAD was connected to a 5 V Li⁺‐battery for 45 min, the colors of all chambers with SARS‐CoV‐2 displayed clear dusty‐blue, which differed from the color of purple in the negative chambers without SARS‐CoV‐2 (Figure 5b). In order to confirm the results, all LAMP products were extracted out and electrophoresed in 1% agarose gel. Evidently, all positive lanes showed classic ladder‐like patterns that were a unique characteristic of LAMP reaction (Figure 5b).^{[ 14b ]} Additionally, we used ImageJ software to quantify the color intensity of each reaction chamber. Relatively quantitatively, the color intensity decreased with the serial dilution of concentration in each reaction chamber (Figure 5d). In summary, the HiPAD for SARS‐CoV‐2 had a low detection limit of 25 copies mL⁻¹ and a wide detection ranged from 25 to 2.5 × 10¹⁰ copies mL⁻¹ in a short time of 45 min. In comparison with conventional real‐time PCR devices with 1000 copies mL⁻¹ of detection limit in 2.5 h,^{[ 41 ]} the HiPAD had remarkable advantages of high sensitivity and shorter detection time and was more suitable for POCT in undeveloped areas.

Figure 5.

The visual detection sensitivity and specificity in HiPAD or M‐HiPAD. a) The model of HiPAD and M‐HiPAD. b) Visual detection of SARS‐CoV‐2. Color change is observed by naked eyes in paper‐based chambers with 5 mm diameters in the HiPAD. Chamber N indicates a negative control, while chamber 1 to 11 represent positive samples, with SARS‐CoV‐2 N gene template concentration of 2.5 × 10¹⁰–2.5 copies mL⁻¹ (tenfold serial dilution). c) Gel electrophoresis pattern of LAMP products in 1% agarose gel. The classic ladder‐like patterns verify successful positive LAMP reactions in chamber 1 to 10 in comparison with negative LAMP reaction in chamber N. Lane M is DNA size marker. Lane N, and 1 to 11 are from (b). d) Colorimetric determination of color change in (b) using ImageJ software. Specificity assay of visual detection of e) SARS‐CoV‐2, f) H1N1 virus, g) HBV, h) ASFV, and i) VSV in M‐HiPAD. All Chamber 1 to 5 in five M‐HiPADs, respectively, contained the LAMP primers of SARS‐CoV‐2, H1N1, HBV, ASFV and VSV. Color change is observed by naked eyes in paper‐based chambers with 5 mm diameters in the M‐HiPAD. The concentration of each virus genome is 1 ng µL⁻¹.

To evaluate the specificity of visual detection in the M‐HiPAD, we contrastively detected five viruses namely SARS‐CoV‐2, H1N1 virus, hepatitis B virus (HBV), African swine fever virus (ASFV), and vesicular stomatitis virus (VSV). The LAMP or RT‐LAMP reactions were carried out in a M‐HiPAD with six chambers for multiplexed simultaneous detection. Experimentally, only the chamber containing SARS‐CoV‐2 could be detected, exhibiting a distinct dusty‐blue (Figure 5e‐1) and notable ladder‐like patterns (Figure S8a‐1, Supporting Information). Nevertheless, the colors of chamber 2 to 5 containing other viruses’ genome were all identical with the negative control group of purple. Similarly, the detection results and validation by gel electrophoresis of H1N1, HBV, ASFV, and VSV all resembled the results of SARS‐CoV‐2 (Figure 5f–i and Figure S8b–e, Supporting Information). The results manifested that visual detection of viruses’ genome in the M‐HiPAD possessed a high specificity.

Currently, global medical waste due to SARS‐CoV‐2 detection has been increasing. These materials are mostly made of plastics, and these used medical plastic wastes are hard to be disposed. The paper‐based chip in the HiPAD is suitable for incineration disposal to avoid potential secondary transmission of COVID‐19;^{[ 42 ]} other modules like the sealing glass, heater, the electrode plate and the battery can be reusable. Besides, in our future work, sample processing will be integrated into the HiPAD to achieve real‐sample‐in‐answer‐out POCT, reducing contamination caused by the testing process. Therefore, to some extent, HiPAD will reduce medical waste pollution and avoid virus spread caused by used medical waste in this severe COVID‐19 outbreak.

3. Conclusion

In summary, rGO/MWCNTs, as a desired electrothermal material can be developed into a paper‐based heater. Furthermore, we successfully designed and fabricated a paper‐based analytical device integrated with a rGO/MWCNTs heater (HiPAD) for visual SARS‐CoV‐2 detection by colored LAMP reaction. The rGO/MWCNTs heater had high reliability and robust stability and rapidly supplied enough electrothermal energy for LAMP reaction. For SARS‐CoV‐2 detection in the HiPAD, the detection range was from 25–2.5 × 10¹⁰ copies mL⁻¹ with 25 copies mL⁻¹ as detection limit in 45 min. Additionally, with slight modification, the single‐chamber HiPAD can be easily expanded to M‐HiPAD (multiplex HiPAD) for high throughput multiplex detection. The simplified HiPAD is very suitable for POCT in resource‐limited settings globally. More importantly, HiPAD has addressed a major challenge in in vitro diagnostics field by miniaturizing and integrating an efficient yet cost effective portable heater into a POCT device. Of note, this is the first documented report that rGO/MWCNTs were successfully developed into an electrothermal module in nucleic acids amplification equipment, which distinctly differed from conventional Peltier semiconductor temperature modules. The novel HiPAD or easily expanded M‐HiPAD integrated with rGO/MWCNTs heaters has great potential in clinical POCT with much lower cost and lower medical waste burden, which is particularly meaningful especially in the undeveloped areas.

4. Experimental Section

Fabrication of Paper‐Based rGO/MWCNTs Heater

A paper‐based heater is fabricated by depositing rGO hybridized MWCNTs onto a general filter paper to supply adequate temperature for LAMP reaction. GO (10–20 µm lateral size) was synthesized using a modified Hummers method according to our previous study.^{[ 43 ]} To obtain rGO, 20 mL prepared GO suspension (2 mg mL⁻¹) was sealed in a teflon reactor and then sustainably heated in a muffle furnace at 100 °C for 24 h. After that, the rGO suspension was characterized by Raman spectrometer (Laser Confocal Raman Microspectrometer, Finder Vista, Zolix) and UV–Vis absorption spectrometer (Cary 60 UV–Vis Spectrophotometer, Agilent Technologies). MWCNTs (10–20 µm of length, 50 nm of diameter) was purchased from XFNANO Company, Nanjing, China. Firstly, 100 mL mixed rGO/MWCNTs dispersion containing 15 mg of rGO/MWCNTs and 5 mg SDBS (Sodium Dodecyl Benzene Sulfonate, Sinopharm Chemical Reagent Co., Ltd, China) as a surfactant was ultrasonically treated at 37 °C for 1 h. For the 15 mg of rGO/MWCNTs in the above dispersion, the weight percent of rGO was 0%, 5%, 10%, 15%, and 20%, respectively. After ultrasonic treatment, the prepared rGO/MWCNTs dispersion was vacuum filtrated on a Millipore Nylon filter paper (0.22 µm pore size, Millipore). Then, the Nylon@rGO/MWCNTs composite paper was dried at 60 °C for 2 h in an oven. And the dried Nylon@rGO/MWCNTs paper was cut into a proper size square sheet to be a rGO/MWCNTs heater. Finally, the resistivity and resistance of rGO/MWCNTs heaters were respectively measured using a double testing digital four‐probe tester (ST2263, Suzhou Jingge Electronic Co., Ltd, China) and a digital multimeter (DM3068, RIGOL technologies, Inc.). An infrared thermal imager (TESTO 885‐1) was used to measure the surface temperature.

Fabrication of HiPAD

A complete HiPAD was composed of five functional modules. The five modules consisted of a battery, electrodes plate, a 5% rGO/MWCNTs heater, a paper‐based chamber and a sealing glass. For POCT, the electric power of a Li⁺‐battery was converted to thermal energy generated by the 5% rGO/MWCNTs heater. Electrodes plate was made on a plastic sheet by screen printing using silver chloride conductive paste. The 5% rGO/MWCNTs heater contacted with electrodes plate was used to heat the chamber. The paper‐based chamber was specially designed for LAMP reaction. The paper‐based chamber was fabricated using Whatman grade 1 chromatography paper (GE Healthcare Life Sciences) in a square with side length of 24 mm. Then, the Whatman grade 1 chromatography paper was quickly immersed in melted sliced paraffin, pulled out, and dried to become hydrophobic waxed paper. The chamber mainly contained three parts, from bottom to top namely, a waxed paper as a base layer, a five‐layer piled‐up waxed paper with a hole of diameter of 5 mm in the middle to hold LAMP reaction mixture, a sealing glass to seal the chamber on the top layer. Each layer was adhered by solid gum or optical adhesive to form a sealed reaction chamber preventing from liquids volatilization and contamination. Especially, the sealing glass was adhered onto the chambers by adhesive tapes; meanwhile, the tape can prevent the contact between the glass and the reactants to avoid contamination. Finally, these five modules were assembled to construct the HiPAD. The multiplex HiPAD had six chambers (M‐HiPAD, the diameter of each chamber was 5 mm).

LAMP Assay for Visually Detecting SARS‐COV‐2 in the HiPAD

LAMP reaction was carried out in the chamber of prepared HiPAD. 25 µL mixture reacted at 60 °C for 45 min, which is the time of nucleic acids amplification not including the time of sample preparation. Of note, the sample preparation step could be versatile and take distinct scale of additional time.^{[ 44 ]} We used MolPure Viral DNA/RNA Kit (Yeasen Biotech Co. Ltd., China) to extract nucleic acids in an hour. The mixture contained 5 µL betaine (5 m) (Sigma‐Aldrich Ltd.), 1 µL MgSO₄ (150 mm) (Sinopharm Chemical Reagent Co., Ltd, China), 3.5 µL dNTPs (10 mm) (Takara), 2.5 µL 10 × Bsm buffer (200 mm Tris‐HCl, 100 mm KCl, 100 mm (NH₄)₂SO₄, 20 mm MgSO₄, 1% (v v⁻¹) Tween 20, pH = 8.8) (Thermo Fisher Scientific), 1 µL Bsm DNA polymerase large fragment (8 U µL⁻¹, Thermo Fisher Scientific), 1 µL dye, 4 µL H₂O (Milli‐Q), 1 µL SARS‐COV‐2 N gene DNA template, 6 µL primers mix (2 µL FIP (20 µm), 2 µL BIP (20 µm), 0.5 µL F3 (10 µm), 0.5 µL B3 (10 µm), 0.5 µL LF (10 µm), 0.5 µL LB (10 µm)) (synthesized by Invitrogen, Thermo Fisher Scientific). The mixture was added into the chamber then sealed.

Primers were carefully designed to amplify based on high conserved SARS‐CoV‐2 N gene sequences from the National Center for Biotechnology Information nucleotide sequence database (GenBank)^{[ 45 ]} as shown in Table S1 in Supporting Information. These samples approximated real clinical samples and therefore were used to demonstrate the practical feasibility and performance of the HiPAD.

For visual detection and verification of LAMP products, several dyes were added into the LAMP reaction mixture as described previously.^{[ 46 ]} We chose four dyes to stain LAMP products, namely 1 µL SYBR Green I (25 ×) (YEASEN Biotech Co. Ltd.), 1 µL GelRed (25 ×) (Sangon Biotech Co. Ltd., Shanghai), 1 µL GeneFinder (1000 ×) (ZEESAN Biotech, Xiamen, China), and 1 µL HNB (3 mm) (Macklin Biochemical Technology Co. Ltd., Shanghai, China). Among these dyes, an appropriate dye with obvious color change was optimized by in‐tube LAMP reaction.

The Sensitivity Analysis

The detection limit is important for diagnosis and therapy. SARS‐COV‐2 N gene template from 2.5 × 10¹⁰ to 2.5 copies mL⁻¹ in tenfold serial dilutions were tested in each LAMP reaction. LAMP reaction without SARS‐COV‐2 was regarded as a negative control.

LAMP amplicons were confirmed by gel electrophoresis. 5 µL of each LAMP product was loaded into each lane of a 1% agarose gel. Electrophoresis of the amplified DNA fragment was carried out in Tris‐borate‐ethylenediaminetetraacetic acid buffer (pH 8.0) at a constant voltage of 100 V. DNA molecular mass markers (100 bp) were used to estimate the sizes of the various amplified products. After electrophoresis, the images were captured and analyzed with a UV illumination (Tanon 1200 gel imaging system, Tanon Science & Technology Co., Ltd., Shanghai, China).

The Specificity Analysis

Specificity and repeatability are the key considerations for practical clinical applications. Five viruses were used to assess the specificity in the M‐HiPAD. The five viruses were SARS‐CoV‐2 N gene template, H1N1 virus RNA genome,^{[ 47 ]} HBV plasmid DNA,^{[ 48 ]} ASFV B646L DNA template,^{[ 49 ]} and VSV RNA genome,^{[ 50 ]} respectively. These primers are listed in Table 1. SARS‐CoV‐2, HBV, and ASFV were detected by LAMP reaction. H1N1 and VSV were detected by RT‐LAMP reaction by adding reverse transcriptase (SuperScript IV, Thermo Fisher Scientific). The concentration of each virus genome was 1 ng µL⁻¹. Each reaction was carried out in three replicates to guarantee accurate reproducibility. All LAMP products were analyzed by 1% agarose gel electrophoresis.

Statistical Analysis

Comparison for two groups was performed using student‐t test. All numerical results were reported as means ± S.D. *P value <0.05 was considered as significant. Error bars represent the standard deviation value of each test with three replicates.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Li S., Huang S., Ke Y., Chen H., Dang J., Huang C., Liu W., Cui D., Wang J., Zhi X., Ding X., A HiPAD Integrated with rGO/MWCNTs Nano‐Circuit Heater for Visual Point‐of‐Care Testing of SARS‐CoV‐2. Adv. Funct. Mater. 2021, 31, 2100801. 10.1002/adfm.202100801

Data Availability Statement

Data available on request from the authors.