In order to prepare the materials of electrodes, the RGO-CNT complexes were uniformly covered on the textile matrix based on cellulose fibers (TMC) substrate by drop coating method. After that, LDH and Lac were applied dropwise to the RGO-CNT/TMC, thereby preparing the LDH/RGO-CNT/TMC and Lac/RGO-CNT/TMC. The whole preparation process and the pictures of the raw TMC and the RGO-CNT/TMC were showed in Fig. 1. It was obvious from the figure that the electrode color was darker after the addition of RGO-CNT complex, mainly due to the attachment of RGO-CNT.
3.1 Morphology of RGO-CNT/TMC
TMC showed a smooth fibrous porous structure under SEM images, after RGO-CNT complexes decoration didn't change the fibrous porous structure of the original TMC, but it made the fabric surface rougher. SEM analysis only reflects the surface structure of the material used for electrodes; it was also essential to explore the chemical structure on the surface of the materials;( additional data are given in Online Resource 1)
3.2 Chemical structure of RGO-CNT/TMC
To study the RGO-CNT/TMC furtherly, the chemical structure was investigated by surface infrared spectroscopy (IR), X-ray photoelectron spectroscopy (XPS), and Raman spectroscopy (RS), compared with raw TMC.
First, the IR spectra of the raw TMC and RGO-CNT/TMC were investigated. hydroxyl, alkyl, aldehyde, ether and C = C bonds were present on the surface of the TMC. Figure 3g showed the possible chemical structure of the original TMC, and the main chemical groups on RGO-CNT/TMC were hydrophilic hydroxyl and carboxyl groups. The functional groups of RGO-CNT/TMC were then investigated by XPS, and the results showed the presence of hydroxyl and carboxyl groups on the RGO-CNT complexes, which was consistent with the results of IR spectroscopy. Finally, the Raman spectra of pristine TMC and RGO-CNT/TMC were characterized, and the presence of D, G and 2D peaks also implied that the modification of TMC substrate by RGO-CNT complexes was successful. The results of their Raman spectra were consistent with those of IR and XPS;( additional data are given in Online Resource 2)
3.3 Electron transfer resistance of the electrodes
To explore the electrochemical properties of the electrodes, electrochemical impedance spectroscopy (EIS) was used to analyze the electron transfer resistance (Rct) by measuring the diameter of the semicircle in higher frequency portion of EIS curves[27]. Figure 4a showed representative EIS curves of raw TMC, RGO-CNT/TMC, LDH/RGO-CNT/TMC, and Lac/RGO-CNT/TMC electrodes.
The results show that the Rct of RGO-CNT/TMC electrode (Fig. 4a curve b, 0.49 kΩ) were much smaller than that of raw TMC electrode (Fig. 4a curve a, 1.92 kΩ). Because that the high conductive RGO-CNT complexes were decorated on TMC substrate. The RGO-CNT complexes were outstanding materials to decrease the Rct of non-conductive substrate materials. The small Rct value of the RGO-CNT/TMC electrodes meant great electron transfer capability. Therefore, the RGO-CNT/TMC electrodes was a good indicator for fabricating high-performance enzymatic electrodes.
The Rct of LDH/RGO-CNT/TMC and Lac/RGO-CNT/TMC electrodes (Fig. 4a curve c and d) were increased to 0.72 kΩ and 0.66 kΩ, respectively, after decorated with enzymes and PEGDGE on RGO-CNT/TMC electrode (Fig. 4a curve b, 0.49 kΩ), suggesting that the non-conductive protein shell of Lac and LDH and the space steric hindrance of PEGDGE hindering the electron transfer. The increase of the Rct showed the success of the coating method and the inevitable resistance to electron transfer caused by enzymes and polymers.
In sum, the high conductive RGO-CNT complexes were outstanding materials to decrease the Rct of non-conductive substrate materials. However, with the process of coating enzymes and polymers, the Rct increase unavoidably.
3.4 Electroactive surface area of the electrodes
In order to investigate the electrochemical effects of RGO-CNT complexes to TMC, electroactive surface area of raw TMC and RGO-CNT/TMC electrodes were measured by chronocoulometry, respectively. This experiment was based on the following Eq. (1)[28]:
$$Q=(2nFA{{D}_{o}}^{\frac{1}{2}}\bullet {\pi }^{-\frac{1}{2}}\bullet {C}_{o})\bullet {t}^{1/2}$$
1
In the equation, Q was the absolute value of the charge passed; n was the number of electrons in the reaction; F was the Faraday constant; Do was the diffusion coefficient of potassium ferrocyanide; Co was the initial molar concentration of the electroactive species in solution; A was the electroactive surface area of the electrode[28]. According to the above equation, Q was linearly related to t1/2, because the other parameters in the equation remained almost constant under the same electrochemical measurement conditions[29]. The Q-t1/2 relationship for different electrodes could be plotted to study the electroactive surface area of different electrodes.
Figure 4b showed the relationship between the Q and t1/2. The electroactive surface area of RGO-CNT/TMC electrodes (Fig. 4b curve b) was significantly increased compared with that of raw TMC (Fig. 4b curve a), indicating that the RGO-CNT complexes expanded the electroactive surface area of raw TMC. The nano structure of the RGO-CNT complexes could enlarge the area of the TMC, and laid a solid foundation to greater enzymatic loading density, which had strong relationship with electrocatalytic efficiency of the bioelectrodes. In short, the nano structure of RGO-CNT complexes significantly enhanced the electroactive surface area of the electrodes.
3.5 Performances of LDH/RGO-CNT/TMC bioanode
In order to study the electrochemical characteristics of the LDH/RGO-CNT/TMC bioanode, cyclic voltammetry (CV) was carried out in the air-saturated 0.1 M phosphate buffer pH 7.0 PBS containing 2 mM NADH and 50 mM lactate (Fig. 5a). The CVs of the raw TMC (Fig. 5a curve a) and RGO-CNT/TMC (Fig. 5a curve b) didn’t had any peaks, suggesting that they didn’t had reaction with NADH and lactate. Moreover, the double-layer charging and discharging current of RGO-CNT/TMC was stronger than the raw TMC, due to the growth of surface area of RGO-CNT/TMC.
In compared with the non-enzymatic electrode, the LDH/RGO-CNT/TMC electrode showed larger current caused by oxidation and reduction reaction with NADH and lactate (Fig. 5a curve c). Nicotinamide adenine dinucleotide (NADH) was the cofactor of LDH. LDH converted lactate to pyruvate, while NAD+ was reduced to NADH, which was then oxidized by the modified electrode to regenerate NAD+. The reaction equation was as follows (2) and (3)[30] :
$$Lactate+LDH\left({NAD}^{+}\right)\to LDH\left(NADH\right)+Pyruvate+{H}^{+}$$
2
$$LDH\left(NADH\right)\to LDH\left({NAD}^{+}\right)+2{e}^{-}+{H}^{+}$$
3
The enzymatic loading density (Γ ) on the electrodes surface was a key factor in determining the performances of bio-electrodes. Therefore, the G value of the LDH/RGO-CNT/TMC bioanode were calculated according to the following Eq. (4)[28]:
Γ = Q/nFA (4)
where Q was the charge obtained by integrating the anode peak (Fig. 5A curve c) at a low scan rate (0.01 mV s− 1); n was the number of electrons transferred in the reaction; F was the Faraday constant, and A was the geometric area of the electrode (4.5 cm2). The value of Γ was calculated to be 3.59×10− 8 mol cm− 2. RGO-CNT/TMC had a higher enzyme loading capacity compared with previous reports[31, 32]. The high loading capacity of the bioanode was attributed to the porous structure of the RGO-CNT complexes and the large electroactive surface area due to the three-dimensional structure of the TMC substrate.
To investigated the electrocatalytic performance of the LDH/RGO-CNT/TMC electrode for lactate oxidation, the electrochemical response of the LDH/RGO-CNT/TMC electrode was measured at lactate concentrations from 0 mM to 100 mM (Fig. 5b). With increasing lactate concentration, the number of electrons consumed by oxygen during the enzymatic reaction increases leaded to an increase in the oxidation and reduction current [33].When the lactate concentration exceeded 80 mM (inset of Fig. 5b), indicating that the lactate reached saturation. Increased lactate concentration would improve the rate of the reaction in Eqs. (2) and (3), which enlarge the electrocatalytic current, thus indirectly enabling the quantitative determination of lactate.
When the concentration of lactate was high, the slope of the straight line increases first and then tends to be flat, which conforms to the characteristics of Michaelis Menten kinetics[34]. Michaelis–Menten constant (\({K}_{m}^{app}\)) was an important parameter to reveal the kinetics of enzyme substrate reaction[35].This parameter was calculated by Lineweaver Burk Eq. (5)[36] .
Where Iss was the current at steady state, C was the concentration at steady state, and Imax was the maximum current measured under saturated conditions. The enzyme activity value of LDH/RGO-CNT/TMC electrode for lactate was calculated as 1.46 mM. Compared with previous works, the \({K}_{m}^{app}\) in this work was lower [37]. The lower constant verified that LDH loaded on RGO-CNT/TMC electrode had high enzymatic activity, which was due to the excellent enzyme embedding method and good biocompatibility of PEGDGE.
In conclusion, the LDH/RGO-CNT/TMC biocathode had a large enzyme loading density, high electrocatalytic efficiency, and high enzymatic activity. Therefore, it was a good choice for the preparation of lactate self-powered biosensors.
Figure 5 (a) CV responses of (a) raw TMC, (b) RGO-CNT/TMC, and (c) LDH/RGO-CNT/TMC in air-saturated 0.1 M PBS (pH 7.0) containing 2 mM NADH and 50 mM lactate at scan rate of 0.01 V s− 1. (b) Recording CV curves of LDH/RGO-CNT/TMC bioanode in air-saturated PBS (pH 7.0) containing 2 mM NADH at a scan rate of 0.01 V s− 1 during successive addition of lactate from 0 mM to 100 mM; inset: The peak current for lactate oxidation vs the concentration of lactate.
3.6 Performances of Lac/RGO-CNT/TMC biocathode
To investigate the electrocatalysis performance of the Lac/RGO-CNT/TMC biocathode, the CV responses of the biocathode with different concentration of oxygen was compared, and the results were showed in Fig. 6. When oxygen was saturated in the electrolyte (Fig. 6 curve b), the reduction current of the Lac/RGO-CNT/TMC electrode became larger and the oxidation current decreased, compared with the response of the same electrode in the air-saturated electrolyte. The whole CV curve moved downward, indicating that the biocathode with Lac as the electro-catalyst could catalyze oxygen reduction reaction effectively. The high electrocatalytic efficiency of the biocathode resulted from the increased electron transfer resistance caused by decoration of RGO-CNT complexes.
According to our previous research, without oxygen in electrolyte, the direct redox of multicopper active sites in Lac would dominte on the surface of the biocathode[38]. The relative reaction was showed in Eq. 6 [36, 38]:
Lac(Ox) + 4e−+4H+↔Lac(Red) (6)
Without oxygen, the reduction of Lac (Red) and oxidation of Lac (Ox) reaction rates (Eq. 6) was equal and balanced[36]. In comparison, after increasing the concentration of oxygen in the electrolyte, the reaction between oxygen and Lac (Red) (Eq. 7) accelerates the generation of Lac (Ox). More Lac (Ox) means faster electrons transfer from biocathode to active sites in Lac (Eq. 8), leading to a larger absolute value of reduction current. On the other hand, oxygen limits the quantity of Lac (Red), resulting in the reverse reaction rates of Eq. 6 decrease. Thus, the decline of the oxidation current of Lac occurs in the presence of oxygen. The relative reaction was showed in Eqs. 6 and 7 [39]:
O2 + Lac(Red)→Lac(Ox) + 2H2O (7)
Lac(Ox) + 4e−+4H+→Lac(Red) (8)
In short, the Lac/RGO-CNT/TMC biocathode had a high electrocatalytic efficiency for oxygen reduction reaction.
3.7 Performances of the self-powered biosensors
Due to the excellent electrochemical activity of the bioelectrode, the LDH/RGO-CNT/TMC bioanode and Lac/RGO-CNT/TMC biocathode were assembled into a stretchable and bendable lactate self-powered biosensor (Fig. 7d). The bioelectrode performance was tested in the state of stretching 43.3% and bending 180° with the help of external forces, as showed in Fig. 7, where the biosensor operated in air-saturated 0.1 M PBS (pH 7.0) containing 2 mM NADH and various concentrations of lactate.
Figure 8a showed the polarization curves of LSV in 0–10 mM lactate solution. In order to investigate the performance of the lactate sensor, the power density profile of the biosensor was measured at variations of lactate concentration from 0 mM to 10 mM (Fig. 8b), and when 10 mM lactate was added, the maximum output power was 18.46 µW cm− 2, which can be used as a biosensor fuel. Power-based sensing in sensing lactate was investigated by recording the current voltage when lactate was gradually added. The showed the (Fig. 8b) showed the power density profiles of the biosensor measured under different conditions with lactate concentrations ranging from 0 mM to 10 mM. The functioning of the biosensor was investigated by recording the highest power output density after the progressive addition of lactate. The calibration curve (Fig. 8c) showed a linear response (R2 = 0.994) of the lactate sensor in the range of 0 ~ 10 mM lactate with a sensitivity of 3.16 µW mM− 1 cm− 2. The limit of detection (LOD) calculated from the calibration curve was 9.49 µM (S/N = 3). The linear range, detection limit and sensitivity were compared with the self-powered lactate biosensor reported in Table 1, and the biosensor was found to be promising for lactate detection
In addition, bendability and stretchability of the flexible textile matrix enzymatic bioelectrodes was characterized. Figure 7b and 7c showed the mechanical flexibility of the bioelectrodes in bending and stretching [2]. Figure 8d showed the mechanical stability of the bioelectrodes by measuring their current density in 0.1 M PBS (pH 7.0) with 2 mM lactate and 2 mM NADH after repetitively bending and stretching. The bioelectrodes were bent to 180° and stretched to 43.3%. The original current density could still maintain more than 95% after 300 cycles. The high mechanical flexibility of the bioelectrodes was made possible by the flexible TMC. This property demonstrated the potential good flexibility of the biosensors prepared by these flexible TMC electrodes.
Table 1
Comparison of sensing performance of self-powered lactate biosensors
The main feature of the
biosensors
|
Output signal
|
Linear range (mM)
|
LOD(µM)
|
Reference
|
Lactate biosensors using sensing membranes
|
Current
|
0.1–0.8
|
60
|
[40]
|
New quasi-direct electron transfer type sensors
|
-
|
0–1
|
9.9
|
[41]
|
Electrochemical biosensors for lactate detection
|
Current
|
0–2
|
-
|
[42]
|
Smartphone-based electrochemical lactate biosensor.
|
|
0.05-10
|
9.1
|
[43]
|
Electronic skin stretchable
self-powered biosensor
|
Voltage
|
0–10
|
-
|
[2]
|
Nanocomposites for detection of lactate in human serum and urine
|
Current
|
100–5000
|
11.6
|
[44]
|
Self-powered lactate biosensors based on flexible textile matrix enzymatic electrodes
|
Power
|
1–10
|
9.49
|
This work
|
To test the anti-interference of the lactate self-powered biosensor, the effects of interfering substances in sweat, such as NaCl, KCl and urea, were studied by cyclic voltammetry. As showed in Fig. 8E, after the addition of 10 mM NaCl, 20 mM KCl and 0.05 mM [19]Urea to the lactate solution. The effect of interfering substances to the current responses were kept within 8.0%, implying the excellent anti-interference ability due to the selectivity of the LDH in lactate biosensor. Thus, the obtained biosensor was promising for the detection of lactate.
Figure 8 (a) Polarization curves at 0–10 mM lactate solution by LSV at low scan rate of 0.01 V s− 1 (b) Power density curves of self-powered biosensors at lactate concentrations; (c) Plot of power density versus lactate concentration (R2 = 0.994) calibration curve; (d) Stability of the bioelectrodes under tensile and bending conditions; (e) Anti-interference of the biosensor to 2 mM lactate in 0.1 M PBS (pH 7.0) containing 2 mM NADH, 10 mM NaCl, 20 mM KCl and 0.05 mM Urea; (f) Long-term storage stability of 30 days in 4°C
To investigate the stability of the biosensor in long-term storage, the performance was obtained by measuring the biosensor current signal every 6 days for 30 consecutive days, as showed in Fig. 8F. The performance of the biosensor can still be maintained at 83% during 30 consecutive tests, which showed that the device was stable. The biosensor relies mainly on lactate dehydrogenase to measure lactate. During storage, the activity of the enzyme decreases, which led to a decrease in electron transfer efficiency and a decrease in sensing performance.
3.8 Determination of lactate in real sweat sample
In order to test the actual performance of the lactate biosensor for use in real sweat, we integrated the sensor into clothing, as shown in Fig. 7e. We performed experiments with lactate recovery in the clothing. As shown in Table 2, the lactate recovery ranged from 99.66–108.43% with a relative standard deviation (RSD) of less than 4.65%. Therefore, the lactate self-powered biosensor has good application prospects for detecting lactate content in real sweat.
Table 2
Determination of lactate content in real sweat
Samples
|
Theory (mM)
|
Detected (mM)
|
Recovery (%)
|
RSD (%)
|
1
|
0.1
|
0.1508
|
105.31
|
4.64
|
0.1
|
0.238
|
108.43
|
0.1
|
0.1537
|
106.45
|
2
|
1
|
0.9900
|
99.7
|
3.02
|
1
|
1.04859
|
100.28
|
1
|
1.1198
|
102.61
|
3
|
2
|
2.08
|
100.41
|
1.48
|
2
|
1.936
|
99.66
|
2
|
2.085
|
101.60
|