Graphene Field-Effect Transistors for the Sensitive and Selective Detection of Escherichia coli Using Pyrene-Tagged DNA Aptamer
石墨烯场效应晶体管，用于使用芘标记的DNA适配体灵敏和选择性地检测大肠杆菌

2.1 Design and Fabrication of the Pyrene-Tagged Aptamer-Modified G-FETs
2.1 芘标记适配体修饰G-FET的设计与制备

The G-FET contains two metal electrodes, namely, the source and drain electrodes, and the graphene layer serves as the conducting channel bridging the contact electrodes. Si wafer with a 300 nm oxidation layer was used as the substrate. The patterned graphene (120 µm × 100 µm) was assembled with the FET structure. Then the source and drain electrodes were insulated by SU-8 photoresist. Details on graphene patterning and G-FET fabrication can be found in the Supporting Information. The active sensing area of graphene was about 100 µm (w) × 100 µm (l). The G-FET was functionalized with the PTDA to realize the desired specific detection. As shown in Figure 1, the as-prepared G-FET was incubated with the PTDA (0.93 × 10⁻⁶ m in deionized water) for 4 h at room temperature (25 °C). This particular aptamer was proved to possess high affinity and specificity toward E. coli 8739, which was ascribed to the specific binding between the aptamer and an out membrane protein or the lipopolysaccharide of E. coli.20 In this study, it was coupled into the G-FET to develop a new biosensor for electrical detection of bacteria. To anchor the PTDA onto the surface of graphene, the pyrene tag (pyrene phosphoramidite) was coupled into the 5′-end of the DNA aptamer with a spacer of four thymines (T) during the DNA aptamer synthesis process. The pyrene tag was successfully characterized by ¹H NMR, ¹³C NMR, ³¹P NMR, and ESI-MS, and the PTDA was confirmed with MALDI-TOF MS. (Detailed information is given in the Supporting Information.)
G-FET包含两个金属电极，即源极和漏极，石墨烯层用作桥接接触电极的导电通道。采用具有300 nm氧化层的硅晶片作为衬底。将图案化石墨烯（120 μm × 100 μm）与FET结构组装在一起。然后用SU-8光刻胶对源极和漏极进行绝缘。有关石墨烯图案化和 G-FET 制造的详细信息，请参阅支持信息。石墨烯的活性传感面积约为100 μm （w） ×100 μm （l）。G-FET与PTDA一起功能化，以实现所需的特定检测。如图1所示，将制备的G-FET与PTDA（在去离子水中0.93×10 ⁻⁶ m）在室温（25°C）下孵育4小时。这种特殊的适配体被证明对大肠杆菌 8739 具有高亲和力和特异性，这归因于适配体与大肠杆菌的外膜蛋白或脂多糖之间的特异性结合。20 在这项研究中，它被耦合到 G-FET 中，以开发一种用于细菌电检测的新型生物传感器。为了将PTDA锚定在石墨烯表面，在DNA适配体合成过程中，将芘标签（芘亚磷酰胺）偶联到DNA适配体的5'端，并带有四个胸腺嘧啶（T）的间隔物。用 ¹ H NMR、C NMR、 ^{13 31} P NMR和ESI-MS成功表征了芘标签，并用MALDI-TOF MS.确认了PTDA。

2.2 Theoretical Model and Optimization
2.2 理论模型与优化

Understanding the change of the carrier density in the probe-modified graphene due to the attachment of E. coli is the key to improving the sensing performance of the G-FETs. The adsorption of charged targets in a solution-gated G-FET can cause an increase or decrease of the source–drain current (ΔI_ds) during real-time electrical monitoring. In our study, the binding of negatively charged E. coli induces holes in the graphene channel, causing the device current to increase. The current modulation is expressed as a function of the change in the carrier density (Δn) in the graphene channel, which is proportional to the number of targets (N) attached on the surface of graphene21, 22
了解探针修饰石墨烯中由于大肠杆菌附着而导致的载流子密度的变化是提高G-FET传感性能的关键。在实时电气监控期间，溶液门控G-FET中带电目标的吸附会导致源漏电流（ΔI _ds ）的增加或减少。在我们的研究中，带负电荷的大肠杆菌的结合会在石墨烯通道中诱导空穴，导致器件电流增加。电流调制表示为石墨烯通道中载流子密度（Δn）变化的函数，该变化与附着在石墨烯表面的靶数（N）成正比21,22

(1)

where w is the width of the graphene channel; l is the length of the graphene channel; e is elementary charge (1.602 × 10⁻¹⁹ C); µ is the carrier mobility; V_ds is the source–drain voltage. Both V_ds and w/l ratio have been known to affect the electronic properties of G-FETs.23 First, a low V_ds allows low power operation of the G-FETs, which ensures no chemical reactions on the surface of the graphene channel. Second, more wrinkles and vestiges are possible on the graphene channel with large w/l ratios, which deteriorate the carrier mobility of the G-FETs. Besides, more contamination can be introduced to the graphene sample with large size.24 Both w/l ratio and V_ds were kept as constants for the real-time electrical monitoring in this study and the carrier mobility µ was chosen as the most important parameter to assess the sensing performance. The carrier mobility, which is a measure of the carrier transport in a material, is an essential characteristic of the G-FET and high carrier mobility suggests that the carriers (hole or electron) can be collected by the electrodes efficiently. The carrier mobility is determined by the gate voltage when other conditions are kept constant, which is given by25
其中w为石墨烯通道的宽度;l为石墨烯通道的长度;e 是基本电荷 （1.602 × 10 ⁻¹⁹ C）;μ是载波的移动性;V _ds 是源漏电压。众所周知，V _ds 和 w/l 比都会影响 G-FET 的电子特性。23 首先，低 V _ds 允许 G-FET 的低功耗运行，从而确保石墨烯通道表面不会发生化学反应。其次，在w/l比较大的石墨烯通道上可能出现更多的褶皱和残留物，这降低了G-FET的载流子迁移率。此外，大尺寸的石墨烯样品可能会被引入更多的污染。24 本研究将w/l比和V _ds 作为实时电监测的常数，并将载流子迁移率μ作为评估传感性能的最重要参数。载流子迁移率是材料中载流子输运的量度，是G-FET的基本特征，高载流子迁移率表明电极可以有效地收集载流子（空穴或电子）。当其他条件保持不变时，载流子迁移率由栅极电压决定，栅极电压由25给出

(2)

where C_TG is the total gate capacitance, which is defined by the concentration of the electrolyte.26 The electrolyte with a suitable concentration makes the probe–target interaction unscreened. The detailed carrier mobility calculations can be found in the Supporting Information. In this study, the gate voltage that gives the largest carrier mobility is chosen to obtain the electrical response of the graphene devices to E. coli.
其中 C _TG 是总栅极电容，由电解质的浓度定义。26 具有适当浓度的电解质使探针与靶标的相互作用未被筛选。详细的运营商移动性计算可以在支持信息中找到。在这项研究中，选择具有最大载流子迁移率的栅极电压来获得石墨烯器件对大肠杆菌的电响应。

2.3 Characterizations of the PTDA Functionalization
2.3 PTDA功能化的表征

The functionalization of the G-FETs with the PTDA was characterized by electrical measurements. The G-FETs were fabricated on the Si chip and each chip consisted of four single G-FET devices as shown in Figure 2a. The electrical measurements were recorded on a Keysight B1500 semiconductor analysis system assembled with a micromanipulation stage. To obtain the transfer characteristics of the APG-FETs, the measurements were performed with a V_ds of 0.1 V to avoid possible chemical reaction on the surface of the graphene channel.23 In Figure 2b, a representative G-FET exhibits ambipolar characteristics (black line) and the voltage at the Dirac point (V_DP) is 0.24 V, which indicates that the G-FETs operate in the p-type region. After the PTDA functionalization, the V_DP shifts slightly from 0.24 to 0.225 V (red line) because the PTDA is negatively charged DNA chain, which donates electrons to graphene causing the n-doping effect on graphene. This phenomenon is consistent with previous observations.27 The transfer characteristics of ten graphene transistors for the PTDA functionalization were measured and the reproducibility of the devices was analyzed as given in the Supporting Information. The small fluctuations of V_DP (0.24 ± 0.003 V) for these G-FETs suggest excellent repeatability of the fabrication of the graphene devices. The V_DP of these G-FETs after the PTDA treatment is 0.225 ± 0.004 V, which implies the excellent repeatability of the PTDA functionalization.
G-FET与PTDA的功能化通过电气测量进行了表征。G-FET在Si芯片上制造，每个芯片由四个单个G-FET器件组成，如图2a所示。电气测量结果记录在装有显微操作台的 Keysight B1500 半导体分析系统上。为了获得APG-FET的传输特性，在0.1 V的V _ds 下进行测量，以避免石墨烯通道表面可能发生的化学反应。23 在图2b中，具有代表性的G-FET表现出双极性特性（黑线），狄拉克点（V _DP ）处的电压为0.24 V，这表明G-FET工作在p型区域。PTDA功能化后，V _DP 从0.24V略微偏移到0.225V（红线），因为PTDA是带负电的DNA链，它向石墨烯提供电子，从而对石墨烯产生n掺杂效应。这种现象与之前的观察结果一致。27 测量了用于PTDA功能化的10个石墨烯晶体管的转移特性，并分析了这些器件的可重复性，如支持信息中所示。这些G-FET的V _DP （0.24±0.003 V）的小波动表明石墨烯器件的制造具有出色的可重复性。这些G-FET在PTDA处理后的V _DP 为0.225±0.004 V，这意味着PTDA功能化具有出色的可重复性。

Atomic force microscopy (AFM) was employed to examine the PTDA functionalization on graphene as well. The AFM scanning images and the thickness of graphene before and after the PTDA functionalization are shown in Figure 2c,d. The thickness of the bare graphene is estimated to be 1.2 nm, indicating a monolayer graphene. This value is larger than the theoretical thickness of graphene (0.34 nm), which is due to some adsorbed molecules on the graphene surface or on the interface between graphene and the substrate during the transfer process.28 After the surface functionalization of graphene, this value increases to 4.8 nm. The height of the PTDA is 3.6 nm, which is consistent with the size of the aptamers in other studies.29, 30 The surface roughness of graphene (R_a) exhibits a significant increase from 0.766 to 2.11 nm after the PTDA functionalization, which is attributed to the flexibility of the aptamer molecule. This observation further proves the attachment of the aptamer on the graphene surface.
原子力显微镜（AFM）也用于研究石墨烯上的PTDA功能化。PTDA功能化前后的AFM扫描图像和石墨烯厚度如图2c，d所示。裸石墨烯的厚度估计为1.2纳米，表明是单层石墨烯。该值大于石墨烯的理论厚度（0.34 nm），这是由于在转移过程中石墨烯表面或石墨烯与衬底之间的界面上吸附了一些分子。28石墨烯表面功能化后，该值增加到4.8nm。PTDA的高度为3.6 nm，与其他研究中适配体的大小一致。29， 30 PTDA官能化后，石墨烯（R _a ）的表面粗糙度从0.766 nm显著提高到2.11 nm，这归因于适配体分子的柔韧性。这一观察结果进一步证明了适配体在石墨烯表面的附着。

2.4 Electrical Detection of E. coli and Gate-Dependent Carrier Mobility
2.4 大肠杆菌的电检测和门依赖性载流子迁移率

We then investigated the electrical performance of the APG-FETs for E. coli detection. In this study, the gate voltage was applied to the graphene channel via an Ag/AgCl electrode, which was immersed into the electrolyte. As shown in Figure 3, the addition of high concentration of E. coli (10⁶ CFU mL⁻¹) causes a significant right shift of the transfer characteristics. The current response ranges from 2% to 13% (detailed current response calculation can be found in the Supporting Information). The current response of 13% to E. coli (10⁶ CFU mL⁻¹) at V_g = 0.04 V is one of the largest responses compared with previous studies. Antibody-modified G-FET exhibited a response of 12% upon exposure to E. coli (10⁵ CFU mL⁻¹).9 Antibody-modified reduced graphene oxide FET exhibited a response of 7.3% to E. coli (10⁵ CFU mL⁻¹).10 In another study, the antimicrobial peptides-modified G-FET exhibited a larger response of 20% to much higher E. coli concentration (10⁷ CFU mL⁻¹).31 These comparisons indicate that our biosensor using aptamer and G-FET can be used for E. coli detection with good current response. The V_DP exhibits a right shift from 0.225 to 0.27 V, implying a p-doping effect of E. coli on graphene. Ten APG-FET devices with similar electrical properties (V_DP = 0.225 ± 0.004 V) were exposed to E. coli (10⁶ CFU mL⁻¹), and the V_DP exhibited obvious and consistent right shift to 0.27 ± 0.004 V as shown in Figure S10 (Supporting Information). The right shift of V_DP is explained by the E. coli-induced electrostatic gating mechanism (Figure 4a,b). The surface of E. coli is negatively charged in 0.01 × phosphate buffered saline (PBS) buffer (pH = 7.2) due to the deprotonation of both carboxylates and phosphates.32 Since graphene is electrically grounded, the negatively charged E. coli imposes an external electric field with direction toward the electrolyte. The external electric field shifts the Fermi level of graphene downward. Therefore, when the negatively charged E. coli is added and captured by the aptamer PTDA, the hole carrier density in graphene is increased. As a result, the right shift of the Dirac point (p-doping) is observed. It is noteworthy that the aptamer undergoes a conformational change33, 34 when the aptamer binds with E. coli as shown in Figure 4c. The E. coli-induced conformational change brings the negatively charged E. coli close to graphene surface, increasing the hole carrier density efficiently in graphene. Therefore, a significant increase of the electrical current is observed when the APG-FET is exposed to E. coli. The possible structure of this DNA aptamer that binds with E. coli is predicted using Mfold.35 As shown in Figure 4d, we speculate that the particular structure in red color may play a key role in binding with E. coli because it shares the same sequence of other aptamers that show high affinity toward E. coli in the previous study.20 (Detailed analysis of the aptamers in binding with E. coli can be found in Figure S13 in the Supporting Information.)
然后，我们研究了用于大肠杆菌检测的APG-FET的电气性能。在这项研究中，通过Ag/AgCl电极将栅极电压施加到石墨烯通道上，并将其浸入电解质中。如图3所示，添加高浓度的大肠杆菌（10 ⁶ CFU mL ⁻¹ ）会导致转移特性显着右移。电流响应范围从 2% 到 13%（详细的电流响应计算可在支持信息中找到）。与以前的研究相比，在V _g = 0.04 V时，目前对大肠杆菌（10 ⁶ CFU mL ⁻¹ ）的反应为13%，这是最大的反应之一。抗体修饰的 G-FET 在暴露于大肠杆菌 （10 ⁵ CFU mL ⁻¹ ） 时表现出 12% 的反应。9 抗体修饰的还原氧化石墨烯 FET 对大肠杆菌 （10 ⁵ CFU mL ⁻¹ ） 的反应为 7.3%。10 在另一项研究中，抗菌肽修饰的 G-FET 对更高的大肠杆菌浓度 （10 ⁷ CFU mL ⁻¹ ） 表现出 20% 的更大响应。31 这些比较表明，我们使用适配体和 G-FET 的生物传感器可用于大肠杆菌检测，具有良好的电流响应。V _DP 表现出从0.225到0.27 V的右移，这意味着大肠杆菌对石墨烯的p掺杂效应。将10个具有相似电气特性（V _DP = 0.225 ± 0.004 V）的APG-FET器件暴露于大肠杆菌（10 ⁶ CFU mL ⁻¹ ），V _DP 表现出明显且一致的右移至0.27±0.004 V，如图S10所示（支持信息）。V _DP 的右移由大肠杆菌诱导的静电门控机制解释（图4a，b）。大肠杆菌表面带负电荷为 0.01 ×磷酸盐缓冲盐水 （PBS） 缓冲液 （pH = 7.2），因为羧酸盐和磷酸盐都去质子化。32 由于石墨烯是电接地的，带负电的大肠杆菌施加了一个朝向电解质的外部电场。外部电场使石墨烯的费米能级向下移动。因此，当带负电荷的大肠杆菌被加入并被适配体PTDA捕获时，石墨烯中的空穴载流子密度增加。结果，观察到狄拉克点（p掺杂）的右移。值得注意的是，当适配体与大肠杆菌结合时，适配体会发生构象变化33、34，如图4c所示。 大肠杆菌诱导的构象变化使带负电荷的大肠杆菌靠近石墨烯表面，有效地增加了石墨烯中的空穴载流子密度。因此，当APG-FET暴露于大肠杆菌时，观察到电流显着增加。使用 Mfold 预测了这种与大肠杆菌结合的 DNA 适配体的可能结构。35 如图 4d 所示，我们推测红色的特定结构可能在与大肠杆菌结合中起关键作用，因为它与先前研究中对大肠杆菌表现出高亲和力的其他适配体序列相同。20 （与大肠杆菌结合的适配体的详细分析可在支持信息的图S13中找到。)

To investigate the gate-dependent response of the APG-FETs, the changes in the I_ds caused by the addition of E. coli are summarized by subtracting the black line in Figure 3 from the red line. The relationship between the current change of the graphene device and the gate voltage is plotted in Figure 5a. The results show that the sign and the magnitude of the response are gate dependent. In the left branch (the region marked in red), the additional holes induced by E. coli enhance the conductivity of the graphene devices, leading to the positive response. While in the right branch (the region marked in blue), the E. coli-induced holes counteract the electrons in graphene, causing the negative response. The two maximum responses are noted at the gate voltage of 0.04 V in the left branch and 0.59 V in the right branch.
为了研究APG-FET的栅极依赖性响应，通过从红线中减去图3中的黑线来总结添加大肠杆菌引起的I _ds 的变化。石墨烯器件的电流变化与栅极电压之间的关系如图5a所示。结果表明，符号和响应的大小与门相关。在左分支（标记为红色的区域）中，大肠杆菌诱导的额外孔增强了石墨烯器件的电导率，导致积极响应。在右分支（蓝色标记的区域）中，大肠杆菌诱导的空穴抵消了石墨烯中的电子，导致负响应。在左支路的栅极电压为0.04 V，在右支路的栅极电压为0.59 V时，注意到两个最大响应。

We speculate that the gate-dependent response is related to the carrier mobility of graphene, as the electrical response is proven to be proportional to the carrier mobility based on Equation 1.20 The gate-dependent carrier mobility was extracted from Equation 2, as shown in Figure 5b. The positive and negative values are defined for hole and electron, respectively. Obviously, the magnitude of the carrier mobility is gate-dependent and a maximum hole (electron) mobility of 1030.99 cm² V⁻¹ s⁻¹ (459.21 cm² V⁻¹ s⁻¹) is obtained at the gate voltage of 0.04 V (0.595 V), which matches well with the gate voltages at the maximum electrical response. The detailed carrier mobility calculations can be found in the supporting information. Therefore, the device performance can be optimized by controlling the gate voltage as seen from the results above.
我们推测栅极相关响应与石墨烯的载流子迁移率有关，因为基于公式1的电响应被证明与载流子迁移率成正比。20 从公式2中提取了与栅极相关的载流子迁移率，如图5b所示。正值和负值分别针对空穴和电子定义。显然，载流子迁移率的大小与栅极有关，在0.04 V（0.595 V）的栅极电压下，最大空穴（电子）迁移率为1030.99 cm ² V s（459.21 cm ² V ^{−1 −1} s ^{−1 −1} ），这与最大电响应下的栅极电压匹配良好。详细的运营商移动性计算可以在支持信息中找到。因此，从上述结果可以看出，可以通过控制栅极电压来优化器件性能。

2.5 Real-Time Current Monitoring of the APG-FETs for E. coli Detection and Quantitative Analysis
2.5 用于大肠杆菌检测和定量分析的APG-FET的实时电流监测

To quantitatively evaluate the concentration-dependent response of the biosensors toward the bacteria, the real-time current recording of the as-prepared APG-FETs was performed with successive addition of E. coli at different concentrations. The real-time measurements were performed with a V_ds of 0.05 V to avoid possible oxidation reaction on the graphene surface. The 0.01 × PBS buffer was used as the electrolyte and the graphene devices were gated at the optimal gate voltage of 0.04 V. The crude bacterial culturing solution was directly diluted with 0.01 × PBS buffer to different concentrations. Since the bacterial culture medium and the PBS buffer are involved in the bacterial detection, the electrical signals of the APG-FETs upon addition of these two solutions were investigated as the control group. As shown in Figure 6a (black line), negligible changes in the current of the APG-FETs are seen after the addition of the culture medium and the PBS buffer, which indicates that both these solutions have no obvious effect on the devices. Then we measured the electrical current of the APG-FET devices exposed to E. coli at different concentrations of 10², 10³, 10⁴, 10⁵, and 10⁶ CFU mL⁻¹. As shown in Figure 6a (red line), the APG-FETs show a concentration-dependent response. The source–drain current I_ds exhibits a stepwise increase at each concentration. The detection limit of the APG-FETs is 10² CFU mL⁻¹, which is comparable with previous studies as shown in Table 1. The exposure of the graphene devices to this concentration of E. coli causes an obvious increase in device current within about a minute (72 s) as shown in the inset of Figure 6a; this is faster than some of the reported methods for bacterial detection such as the flow cytometry, the polymerization chain reaction (PCR) test, fluorescence method, and antibody-modified G-FET, as shown in Table 1. Besides, the sensitivity at the detection limit (10² CFU mL⁻¹) was about 2.5 × 10⁻³ µA per (CFU mL⁻¹). This value is at the same order when compared with previous study as shown in Table 1. These results further indicate that our biosensor using the aptamer and G-FET is a promising platform for bacterial detection.
为了定量评估生物传感器对细菌的浓度依赖性响应，连续添加不同浓度的大肠杆菌对制备的APG-FETs进行实时电流记录。在0.05 V的V _ds 下进行实时测量，以避免石墨烯表面可能发生的氧化反应。使用0.01 × PBS缓冲液作为电解质，并在0.04 V的最佳栅极电压下对石墨烯器件进行门控。将粗菌培养液直接用0.01×PBS缓冲液稀释至不同浓度。由于细菌培养基和PBS缓冲液参与细菌检测，因此以对照组为对照组，研究了APG-FETs加入这两种溶液后的电信号。如图6a（黑线）所示，在加入培养基和PBS缓冲液后，APG-FET的电流变化可以忽略不计，这表明这两种溶液对器件没有明显影响。然后，我们测量了暴露于大肠杆菌的APG-FET器件在10 ² 、10 ³ 、10 ⁴ 、10 ⁵ 和10 ⁶ CFU mL ⁻¹ 的不同浓度下的电流。如图6a（红线）所示，APG-FET显示出浓度依赖性响应。源极-漏极电流I _ds 在每种浓度下呈阶梯式增加。APG-FETs的检出限为10 ² CFU mL ⁻¹ ，与以前的研究相当，如表1所示。石墨烯器件暴露于这种浓度的E。 大肠杆菌在大约一分钟（72 秒）内导致器件电流明显增加，如图 6a 的插图所示;如表1所示，这比一些报道的细菌检测方法（如流式细胞术、聚合链反应（PCR）测试、荧光法和抗体修饰的G-FET更快。此外，检测限 （10 ² CFU mL ⁻¹ ） 的灵敏度约为 2.5 × 10 ⁻³ μA /（CFU mL ⁻¹ ）。与表1所示的先前研究相比，该值处于相同的顺序。这些结果进一步表明，我们使用适配体和G-FET的生物传感器是一个很有前途的细菌检测平台。

The source–drain current I_ds increases significantly with E. coli concentration from 10² to 10⁵ CFU mL⁻¹, and then the signal saturates gradually for concentrations above 10⁵ CFU mL⁻¹. The maximum response is obtained (the maximum net change in source–drain current ΔI_{ds, max} = 1.686 µA) at an E. coli concentration of 10⁶ CFU mL⁻¹. We idealize that the whole surface of the graphene is occupied by E. coli and all of these E. coli cells contribute to the increase of the device current. The active sensing area of graphene is about 100 µm × 100 µm and the size of a single E. coli is about 1.5 µm × 0.5 µm.39 Therefore, the largest number of E. coli reaching the surface of the aptamer-modified graphene is estimated to be ≈13 333. Based on Equation 1, the smallest number of carriers (holes) induced per E. coli is calculated as 1531. The detailed carrier density evaluation can be found in the Supporting Information.
当大肠杆菌浓度从 10 ^{2 5} 到 10 CFU mL ⁻¹ 时，源漏电流 I _ds 显着增加，然后当浓度高于 10 ⁵ CFU mL ⁻¹ 时，信号逐渐饱和。在大肠杆菌浓度为10 ⁶ CFU mL ⁻¹ 时获得最大响应（源漏电流ΔI _{ds, max} = 1.686 μA的最大净变化）。我们理想地认为石墨烯的整个表面都被大肠杆菌占据，所有这些大肠杆菌细胞都有助于器件电流的增加。石墨烯的活性感应区域约为100μm×100μm，单个大肠杆菌的大小约为1.5μm×0.5μm。 39因此，到达适配体修饰石墨烯表面的大肠杆菌数量估计为≈13 333。根据公式1，每个大肠杆菌诱导的最小载体（孔）数计算为1531。详细的载流子密度评估可在支持信息中找到。

We then investigated the affinity of the aptamer toward E. coli by estimating the dissociation constant. The relationship between the electrical current of the APG-FET devices and the concentration of E. coli can be interpreted by the Langmuir adsorption isotherm40
然后，我们通过估计解离常数来研究适配体对大肠杆菌的亲和力。APG-FET器件的电流与大肠杆菌浓度之间的关系可以用Langmuir吸附等温线40来解释

(3)

where ΔI_ds is the net change of the source–drain current caused by E. coli, ΔI_{ds, max} is the maximum net change in the source–drain current caused by E. coli, C_{E. coli} is the E. coli concentration, and K_D is the equilibrium dissociation constant that can interpret the binding affinity of the probe–target complex. In Figure 6b, K_D is estimated as 700 CFU mL⁻¹ from the fitting curve using Equation 3. This value is comparable to that of antibody-bacteria binding in previous studies (K_D = 10^2.6 CFU mL⁻¹),41 indicating that the aptamer used in this study has a high affinity toward E. coli.
式中，ΔI _ds 为大肠杆菌引起的源漏电流净变化，ΔI _{ds, max} 为大肠杆菌引起的源漏电流的最大净变化，C _{E. coli} 为大肠杆菌浓度，K _D 为平衡解离常数，可解释探针-靶复合物的结合亲和力。在图6b中，使用公式3从拟合曲线中估计K _D 为700 CFU mL ⁻¹ 。该值与先前研究中的抗体-细菌结合值相当（K _D = 10 ^2.6 CFU mL ⁻¹ ），41表明本研究中使用的适配体对大肠杆菌具有高亲和力。

2.6 Selectivity and Stability
2.6 选择性和稳定性

Selectivity is an important indicator to weigh the performance of biosensors. To investigate the specificity of APG-FETs for E. coli detection, we tested the electrical response of APG-FETs to different bacteria including E. coli 8739, E. coli K12 ER2925, Streptococcus mutans, and Staphylococcus saprophyticus. The electrical response in this study is defined as ΔI_ds/I_ds where ΔI_ds and I_ds are the change in the source–drain current caused by the addition of bacteria and the source–drain current of the device before the addition, respectively. The selectivity evaluation of APG-FETs toward different bacteria is summarized in Figure 7a. The total response to E. coli 8739 (13.06%) is significantly larger than that to E. coli K12 ER2925 (2.14%), S. mutans (−0.83%), and S. saprophyticus (−0.69%) when the concentrations of these bacteria are kept the same. The response of APG-FETs to the other three bacteria is weak. Exposure to E. coli K12 ER2925 results in a slight current increase, which is attributed to the possibility that the binding site of the aptamer may be less expressed on the surface of E. coli K12 ER2925 than that on the surface of E. coli 8739. The current decreases of the device caused by S. mutans and S. saprophyticus are because these two are Gram-positive bacteria; the positively charged bacterial cell wall in PBS buffer (pH = 7.2) could cause the decrease of hole density and hence the device current. These results further indicate that the response of APG-FETs is caused by the aptamer-bacteria binding and the aptamer-modified G-FETs can be used as the platform for bacteria detection with high selectivity.
选择性是衡量生物传感器性能的重要指标。为了研究APG-FETs对大肠杆菌检测的特异性，我们测试了APG-FETs对不同细菌的电响应，包括大肠杆菌8739、大肠杆菌K12 ER2925、变形链球菌和腐生葡萄球菌。本研究中的电响应定义为 ΔI _ds /I _ds ，其中 ΔI _ds 和 I _ds 分别是添加细菌引起的源漏电流的变化和添加前器件的源漏电流。APG-FET对不同细菌的选择性评估总结如图7a所示。当这些细菌的浓度保持不变时，对大肠杆菌8739的总反应（13.06%）显著大于对大肠杆菌K12 ER2925（2.14%）、变形链球菌（-0.83%）和腐生链球菌（-0.69%）的总反应。APG-FETs对其他3种细菌的响应较弱。暴露于大肠杆菌 K12 ER2925 导致电流略有增加，这归因于适配体的结合位点在大肠杆菌 K12 ER2925 表面的表达可能低于在大肠杆菌 8739 表面的表达。由变形链球菌和腐生链球菌引起的装置电流下降是因为这两种细菌是革兰氏阳性菌;PBS缓冲液（pH = 7.2）中带正电的细菌细胞壁可能导致空穴密度降低，从而降低器件电流。这些结果进一步表明，APG-FETs的响应是由适配体-细菌结合引起的，适配体修饰的G-FETs可以作为高选择性细菌检测的平台。

Stability is another concern to evaluate the quality of a biosensor. To explore the stability of E. coli biosensors, the APG-FETs were stored in PBS buffer solution (pH = 7.2) at three representative temperatures: 4 °C (common temperature for sample storage), 25 °C (room temperature), and 40 °C (high temperature). For each temperature, three APG-FETs were used to investigate the electrical response upon exposure to E. coli solution in each week. As shown in Figure 7b, no significant changes of the response of all APG-FETs to bacteria were observed when the APG-FETs were kept at 4 °C (12.95 ± 0.19%) for six weeks. The same phenomena were observed for 25 °C (12.97 ± 0.19%) and 40 °C (12.89 ± 0.16%). Moreover, the response from the APG-FET devices, which were kept for the same time at different temperatures, shows only a small fluctuation from 12.82 ± 0.20% to 13.09 ± 0.14%. All of these results suggest excellent stability of our APG-FET biosensors. We attribute this stability to the inherent stability of the aptamer when it was stored for a long time (six weeks) at these temperatures.
稳定性是评估生物传感器质量的另一个问题。为了探索大肠杆菌生物传感器的稳定性，将APG-FET储存在PBS缓冲溶液（pH = 7.2）中，在3个代表性温度下储存：4 °C（样品储存的常用温度）、25 °C（室温）和40 °C（高温）。对于每个温度，使用三个 APG-FET 来研究每周暴露于大肠杆菌溶液时的电响应。如图7b所示，当APG-FET在4°C（12.95±0.19%）下保持6周时，未观察到所有APG-FET对细菌的反应发生显着变化。在25°C（12.97±0.19%）和40°C（12.89±0.16%）下观察到相同的现象。此外，在不同温度下保持相同时间的APG-FET器件的响应仅显示出从12.82±0.20%到13.09±0.14%的小幅波动。所有这些结果都表明，我们的APG-FET生物传感器具有出色的稳定性。我们将这种稳定性归因于适配体在这些温度下长时间（六周）储存时的固有稳定性。