3  Measurement formats for pathogen detection
In addition to a physical device composed of an integrated transduction element and biorecognition element, an electrochemical biosensor-based assay for pathogen detection potentially involves processing steps associated with sample preparation and complementary physical systems for biosensor housing and sample handling. The associated protocols for sample preparation and sample handling are often referred to as the measurement format. Several important considerations regarding the measurement format for pathogen detection applications can be considered and vary based on the assay design, the biosensor performance (e.g., sensitivity and LOD), the volume, material properties, and composition of the pathogen-containing sample, and the application. For example, the use of DNA-based assays for pathogen detection typically requires sample preparation steps associated with the extraction of genetic material. Similarly, the use of a label-based biosensing approach requires sample preparation steps associated with target labeling. In cases where the concentration of target species in the sample is below the biosensor's LOD, pre-concentration steps may be required. Applications to process monitoring, such in bioreactor or tissue-chip monitoring, may require flow-based sample handling formats. We next discuss the measurement formats associated with pathogen detection in terms of sample preparation and sample handling.

3.1  Sample preparation: Filtration and pre-concentration
Sample preparation steps have various purposes, including concentrating or amplifying the target species through separation and growth processes, reducing the concentration of background inhibitory species, and reducing the heterogeneity of the sample's composition and properties (Zourob et al. 2008). We next discuss sample filtration and pre-concentration techniques.

3.1.1  Sample filtration
Generally, sample filtration relies on the principle of size discrepancy between the target pathogen and background species. Membranes, fibers, and channels have been used in size-selective sample filtration processes for biosensing applications. Biorecognition elements are commonly used to assist the separation process when the target species exhibits similar properties to background species or the matrix. For example, biorecognition elements that exhibit affinity to a broad group of pathogens, such as lectins, have been used in pre-concentration steps for pathogen detection (Zourob et al. 2008). Bacteria typically exhibit a net negative charge at physiological pH (7.4) because of an abundance of lipopolysaccharides or teichoic acids on the cell membrane (Gram-negative bacteria and Gram-positive bacteria, respectively) (Silhavy et al. 2010). This physical property of cell-based pathogens is leveraged in biofiltration processes, for example, using electropositive filters (Altintas et al. 2015). While the majority of the aforementioned separation processes involve manual handling steps, sample filtration processes are now being integrated with microfluidic-based biosensing platforms (Song et al. 2013). For example, Chand and Neethirajan incorporated an integrated sample filtration technique using silica microbeads for the detection of norovirus in spiked blood samples (Chand and Neethirajan 2017).

3.1.2  Centrifugal separation
Centrifugation can be used as a density gradient-based separation principle for concentrating target pathogens within a sample. In cases where the target species exhibits similar density to background species, the approach is often implemented with antibody-functionalized beads. This technique is commonly employed in applications requiring pathogen detection in complex matrices (e.g., body fluids). Centrifugation-based separation techniques can also potentially be applied to microfluidic-based biosensing platforms. For example, Lee et al. utilized centrifugal microfluidics to process a whole blood sample for subsequent analysis using ELISA (Lee et al. 2009), suggesting that this approach could be extended to electrochemical biosensor-based assays for pathogen detection.

3.1.3  Broth enrichment
Broth enrichment is a technique used to increase the concentration of target species in the sample through growth or replication of target species prior to measurement, thereby increasing the number present for detection. The technique is commonly used in food safety applications. For example, Liebana et al. enriched S. typhimurium-spiked milk samples in Luria broth (LB) for 8 h to improve the assay LOD from 7.5 × 103 CFU/mL for the 50-min enriched sample to 0.108 CFU/mL (Liebana et al. 2009). Salam et al. enriched fresh chicken samples in enrichment buffer peptone for 18–24 h to recover injured S. typhimurium cells for detection via chronoamperometry (Salam and Tothill, 2009). While enrichment can be a useful sample preparation step when the target concentration is below the biosensor's LOD, it is inherently limited to viable and cultural organisms. Further, analysis of the results obtained from multiple samples should consider potential differences in the growth rate of bacteria across different samples. It is important to note that the need for sample enrichment significantly increases the TTR and impedes rapid and real-time detection.

3.1.4  Magnetic separation
The separation of the target species from a sample using magnetic beads has become a commonly used sample preparation approach in pathogen detection applications. Target pre-concentration via magnetic bead-based separation processes typically involves the binding of antibody-functionalized magnetic beads to the target species. The bead-target complexes are subsequently separated from the solution by externally-applied magnetic fields. Magnetic-assisted separation processes are useful when the target species exhibits similar properties to other analytes or background species in the sample, such as those with similar size, density, or chemical properties (Chen et al. 2017). The bead-target complexes are then introduced directly to the biosensor to enable quantification of the target pathogen that was present in the initial sample. As shown in Table 2, magnetic bead-based separation processes have been extensively used for pathogen detection as well as general substrates for traditional immunoassays. Such assays have been used to detect a variety of pathogens, including bacteria, such as E. coli (Chan et al. 2013 ) and Bacillus anthracis (B. anthracis) (Pal and Alocilja, 2009), and viruses, such as bovine viral diarrhea virus (Luo et al. 2010) and human influenza A virus (Shen et al. 2012). In addition to serving as a separation agent, magnetic beads also serve as labels.

3.2  Sample handling formats
The sample handling format is highly influenced by the biosensor application. As discussed in further detail in the following sections, pathogens are present in liquid and solid matrices and on surfaces (e.g., of biomedical devices). In addition, pathogens can be aerosolized, which is a significant mode of disease transmission associated with viral pathogens (e.g., influenza and COVID-19). Sample handling formats can be generally classified as droplet-, flow-, or surface-based.
Droplet formats involve sampling from a larger volume of potentially pathogen-containing material or fluid. The sample droplet is subsequently analyzed by deposition on a functionalized transducer or transferred to a fluidic delivery system. For example, Cheng et al. created an electrochemical biosensor based on a nanoporous alumina electrode tip capable of analyzing 5 μL of dengue virus-containing solutions (Cheng et al. 2012). Droplet formats are simplistic sample handling formats and have the advantage of being performed by unskilled users. While dropletformats have been extensively used with colorimetric biosensors, they have also been adapted for electrochemical biosensors. For example, commercially-available blood glucose meters use a droplet format (Vashist et al. 2011). Examples of low-cost, paper-based, or disposable electrochemical biosensors for pathogen detection that utilize droplet formats are provided in Table 1. For example, Zhao et al. created a screen-printed graphite-based electrode for electrochemical detection of Vibrio parahaemolyticus (V. parahaemolyticus) based on 5 μL samples (Zhao et al. 2007). However, while droplet formats minimize the technical and methodological barriers to measurement, such as eliminating the need for physical systems associated with biosensor housing and sample handling, they can exhibit measurement challenges associated with mass transport and target sampling limitations.
One of the most critical considerations associated with application of droplet formats to pathogen detection is sampling, specifically if sufficient sampling has been performed on the system for which bioanalytical information is desired (e.g., a human, a food source, or source of drinking water). For example, the rationale that the bioanalytical characteristics of a droplet represent that of the bulk system is sound only in a well-mixed system, specifically, a system that exhibits a uniform spatial distribution of species (i.e., concentration profile). We note that while this is typically the case for samples acquired from closed, convective systems, such as body fluids, it should be challenged when considering open systems that exhibit complex flow profiles or regions of static fluid. For example, groundwater systems (e.g., aquifers), rivers, and lakes have been reported to have complex flow profiles (Ji, 2017; Zhang et al. 1996). Thus, the sampling approach should be considered when examining droplet formats for food and water safety applications. In addition to a consideration of system mixing, one should also consider the potential measurement pitfalls when analyzing samples that contain dilute levels of highly infectious pathogens, such as the potential for false-negative results.
Flow formats involve the detection of target species in the presence of flow fields. Flow formats include continuously-stirred systems (e.g., continuously-stirred tank bioreactors), flow cells, and microfluidics. Flow formats have the advantage of exposing the biosensor to target-containing samples in a controlled and repeatable fashion and the benefit of driving exposure of the functionalized biosensor to target species via convective mass transfer mechanisms. Flow formatsare also typically compatible with large sample volumes (liters). Flow cells are typically fabricated via milling and extrusion processes using materials such as Teflon or Plexiglas. They have the advantage of accommodating a variety of biosensor form factors, such as rigid three-dimensional biosensors. In addition to flow cells, flow formats are commonly achieved using microfluidic devices. While microfluidic devices are typically used with biosensors that exhibit thin two-dimensional form factors, such as planar electrodes, they offer various measurement advantages. Unlike flow cells, which are typically fabricated from machinable polymers, microfluidics are typically fabricated using polydimethylsiloxane (PDMS) and polymethyl methacrylate (PMMA) given their low cost and compatibility with microfabrication approaches. One advantage of microfluidic devices is their ability to perform integrated sample preparation steps, which eliminates the need for additional steps in the sample-to-result process (Sin et al. 2014). For example, microfluidic formats for pathogen detection using electrochemical biosensors have demonstrated fluid pumping, valving, and mixing of small sample volumes (Rivet et al. 2011). An example of a microfluidic format created by Dastider et al. for detection of S. typhimurium is shown in Fig. 4a (Dastider et al. 2015).
Detection in the presence of flow fields requires high stability of immobilized biorecognition elements (Bard and Faulkner, 2000). The effect of flow characteristics on biosensor collection rates is an important consideration, especially when considering micro- and nano-scale transducers with microfluidic formats (Squires et al. 2008). For example, emerging nanostructured electrodes, such as functionalized nanoporous membranes, have been shown to achieve high stability in microfluidic devices (Joung et al. 2013; Tan et al. 2011). A detailed discussion on the relationship between device dimensions, flow characteristics, achievable target collection rates, and equilibrium measurement times has been provided elsewhere (Squires et al. 2008). It is paramount for interpreting biosensor response that users understand the interplay between mass transport of target molecules (both diffusive and convective mechanisms) and reaction at the biosensor surface (i.e., binding of target species to immobilized biorecognition elements). Such fundamental understanding can also be employed in biosensor and experiment design to create improved assay outcomes, such as reducing TTR or improving measurement confidence.
While the presence of pathogens on the surfaces of objects can be analyzed using droplet- and flow-based sample handling formats using material transfer processes, such as swabbing, in situ pathogen detection on the object surfaces is a vital measurement capability for medical diagnostic, infection control, and food safety applications. Surface-based measurement formats typically require biosensors with flexible or conforming (i.e., form-fitting) form factors. For example, Mannoor et al. detected the presence of pathogenic species directly on teeth using a flexible graphene-based biosensor (Mannoor et al. 2012). Further discussion of surface-based pathogen detection applications are provided in the following sections.
The sample handling format often provides insight into the biosensor's reusability. Biosensors within the aforementioned measurement formats can be broadly classified as single- or multi-use biosensors. Single-use biosensors are unable to monitor the analyte concentration continuously or upon regeneration, while multiple-use biosensors can be repeatedly recalibrated (Thévenot et al. 2001). For example, droplet-based low-cost, disposable biosensors for water safety are typically single-use, while biosensors for process monitoring applications can be recalibrated to characterize multiple samples and facilitate continuous monitoring. The ability to regenerate biosensor surfaces following pathogen detection (i.e., remove selectively-bound pathogens) is a significant technical barrier limiting progress in multiple-use biosensors, and industrial applications thereof, and is discussed further in the following sections.

3.3  Electrochemical methods for pathogen detection using electrochemical biosensors
Various electrochemical methods can be performed using functionalized electrodes to enable pathogen detection (Bard and Faulkner, 2000). These methods differ in electrode configuration, applied signals, measured signals, mass transport regimes, binding information provided (Thévenot et al. 2001), and target size-selectivity (Amiri et al. 2018). Electrochemical methods used for pathogen detection can be classified as potentiometric, amperometric, conductometric, impedimetric, or ion-charge/field-effect, which often signify the measured signal (Thévenot et al. 2001). The applied signals may be constant or time-varying. The result of the electrochemical method may require analysis of the output signal's transient response, steady-steady response, or a combination of both. A detailed discussion of the aforementioned electrochemical methods has been provided elsewhere (Bard and Faulkner, 2000). Here, we briefly review the most recent methods employed for pathogen detection using electrochemical biosensors.

3.3.1  Potentiometry
Potentiometric methods, also referred to as controlled-current methods, are those in which an electrical potential is measured in response to an applied current (Bard and Faulkner, 2000). The applied current is typically low amplitude. An advantage of controlled-current methods is the ability to use low-cost measurement instrumentation relative to that required for controlled-potential methods.
Hai et al. used potentiometry with a conductive polymer-based biosensor to detect human influenza A virus (H1N1) at a LOD of 0.013 HAU (Hai et al. 2017). Hernandez et al. used potentiometry with a carbon-rod modified electrode that contained reduced graphene oxide to detect S. aureus at a single CFU/mL (Hernandez et al. 2014). Boehm et al. detected E. coli via potentiometry utilizing a Pt wire electrode (Boehm et al. 2007). Further studies utilizing potentiometric sensing approaches are listed in Table 1, Table 2.

3.3.2  Voltammetry
Voltammetric methods, also referred to as controlled-potential methods, are those in which a current is measured in response to an applied electrical potential that drives redox reactions (Bard and Faulkner, 2000). The measured current is indicative of electron transfer within the sample and at the electrode surface, and thus, the concentration of the analyte. In chronoamperometry, the electrical potential at the working electrode is applied in steps, and the resulting current is measured as a function of time. The applied electrical potential can also be held constant or varied with time as the current is measured.
Although various types of biosensors are compatible with voltammetry-based measurements, field-effect transistor (FET)-based biosensors often utilize amperometric-based methods for pathogen detection (Huang et al. 2011; Liu et al. 2013). FET biosensors detect pathogens via measured changes in source-drain channel conductivity that arise from the electric field of the sample environment. This is achieved by immobilizing biorecognition elements on the metal or polymer gate electrode of the device. He et al. showed that FETs based on PEDOT:PSS organic electrochemical transistor electrodes enabled the detection of E. coli in KCl solutions using Pt and Ag/AgCl gate electrodes (He et al. 2012). Wu et al. used a graphene-based FET to detect E. coli in nutrient broth diluted with phosphate buffered saline solution with amperometry using a Ag/AgCl gate electrode (Wu et al. 2016).
Further examples of amperometric sensing include the detection of human influenza A virus by Singh et al. using a reduced graphene oxide-based electrode and chronoamperometry using Fe(CN)6 3 - /4- at a LOD of 0.5 plaque-forming units (PFU)/mL (Singh et al. 2017b). Lee and Jun utilized wire-based electrodes for amperometric detection of E. coli and S. aureus (Lee and Jun 2016). A detailed list of studies that utilize amperometric methods for pathogen detection is provided in Table 1, Table 2

3.3.2.1  Linear sweep and cyclic voltammetry
Linear sweep voltammetry (LSV) methods are those in which a current is measured in response to an applied electrical potential that is swept at a constant rate across a range of electrical potentials (Bard and Faulkner, 2000). Cyclic voltammetry (CV) is a commonly used linear-sweep method in which the electrical potential is swept in both the forward and reverse directions in partial cycles, full cycles, or a series of cycles. CV is one of the most widely used voltammetric methods for pathogen detection.
Hong et al. used sweep voltammetry to detect norovirus in a sample solution with Fe(CN)6 3 - /4- extracted from lettuce (Hong et al. 2015). A typical CV response using Fe(CN)6 3 - /4- associated with pathogen detection is shown in Fig. 5 a for various concentrations of E. coli binding to a polymer composite electrode (Güner et al. 2017). A detailed overview of pathogen detection studies based on CV is provided in Table 1, Table 2.
Fig. 5 Typical responses associated with the common electrochemical methods used for pathogen detection. a) Cyclic voltammetry (CV) data using Fe(CN)63-/4- for varying concentrations of E. coli (Güner et al. 2017). b) Differential pulse voltammetry (DPV) data using Fe(CN)63-/4- for varying concentrations of S. aureus (Bhardwaj et al. 2017). c) Electrochemical impedance spectroscopy (EIS) in 100 mM LiClO4 solution in the form of a Nyquist plot and corresponding equivalent circuit model associated with biorecognition element immobilization and detection of S. typhimurium (Sheikhzadeh et al. 2016). d) Conductometry data for varying concentrations of B. subtilis (Yoo et al. 2017).

3.3.2.2  Pulse voltammetry
Pulse voltammetry is a type of voltammetry in which the electrical potential is applied in pulses. The technique has the advantage of improved speed and sensitivity relative to traditional voltammetric techniques (Bard and Faulkner, 2000; Molina and González, 2016). In staircase voltammetry, the electrical potential is pulsed in a series of stair steps and the current is measured following each step change, which reduces the effect of capacitive charging on the current signal. Square wave voltammetry (SWV) is a type of staircase voltammetry that applies a symmetric square-wave pulse superimposed on a staircase potential waveform. The forward pulse of the waveform coincides with the staircase step. In differential pulse voltammetry (DPV), the electrical potential is scanned with a series of fixed amplitude pulses and superimposed on a changing base potential. The current is measured before the pulse application and again at the end of the pulse, which allows for the decay of the nonfaradaic current (Scott, 2016).
For example, Iqbal et al. used SWV with AuNP-modified carbon electrodes for detection of C. parvum in samples taken from fruit (Iqbal et al. 2015). Kitajima et al. also used SWV with Au microelectrodes to detect norovirus at a LOD of 10 PFU/mL (Kitajima et al. 2016). Cheng et al. used DPV and a nanostructured alumina electrode for detection of dengue type 2 virus with a LOD of 1 PFU/mL (Cheng et al. 2012). As shown in Fig. 5b, Bhardwaj et al. used DPV with a carbon-based electrode to detect S. aureus (Bhardwaj et al. 2017). Additional studies that utilize pulse voltammetry methods forpathogen detection are listed in Table 1, Table 2.

3.3.2.3  Stripping voltammetry
Many of the previously described voltammetric methods can be modified to include a step that pre-concentrates the target on the electrode surface. Subsequently, the pre-concentrated target is stripped from the surface by application of an electrical potential. In anodic stripping voltammetry (ASV), a negative potential is used to pre-concentrate metal ions onto the electrode surface. These ions are then stripped from the surface by applied positive potentials. Although most commonly used to detect trace amounts of metals, this method has been adapted for pathogen detection by electrocatalytically coating metallic labels on bound targets for oxidative stripping and subsequently measuring the current response (Abbaspour et al. 2015).
Chen et al. used stripping voltammetry with a polymer-CNT composite-based electrode to detect E. coli at a LOD of 13 CFU/mL (Chen et al. 2014). In that study, the biosensor was first incubated with E. coli. Silica-coated Ag nanoparticles conjugated with anti-E.coli were subsequently introduced to the system, inducing a binding reaction between the bacteria and the nanoparticles. After rinsing non-specifically bound particles, acid was introduced to dissolve Ag(s), and the resulting Ag+-rich solution was characterized using DPV. Viswanathan et al. used ASV with screen-printed composite electrodes for multiplexed detection of Campylobacter, S. typhimurium, and E. coli with a LOD of 400 cells/mL, 400 cells/mL, and 800 cells/mL, respectively (Viswanathan et al. 2012). In that study, antibody-functionalized nanocrystalline bioconjugates were first introduced to biosensor-bound bacteria, the specifically bound particles were dissolved with acid, and the ions were then stripped using a square-wave voltammetric waveform. Additional studies using stripping voltammetry for electrochemical detection of pathogens can be found in Table 1, Table 2.

3.3.3  Electrochemical impedance spectroscopy
The aforementioned electrochemical methods involved responses based on step changes or continuous sweeps in the applied current or voltage that drove the electrode to a condition far from equilibrium. Alternatively, frequency response methods, often referred to as impedance-based or impedimetric methods, are based on frequency response analysis (i.e., the response of the system to periodic applied current or potential waveforms at either a fixed frequency or over a range of frequencies) (Bard and Faulkner, 2000). This provides several advantages, including measurement over a wide range of times and frequencies and high precision in time-averaged responses. We next discuss impedance-based electrochemical methods for detection of pathogens using electrochemical biosensors.
In EIS the impedance and phase angle of the system are measured as a function of the frequency of the applied electrical potential. EIS is a diverse electrochemical method, which can be done as a faradaic or non-faradaic process, and enables the study of intrinsic material properties, experiment-specific processes, or biorecognition events at the electrode surface. EIS is often performed using an applied low-amplitude sinusoidal electrical potential and a three-electrode configuration. Equivalent circuit models are commonly fit to experimental impedance and phase angle data to interpret the electrochemical process in terms of passive circuit elements, such as resistors and capacitors. For example, the electric double layer is typically modeled as a capacitive element, while the resistance to faradaic charge transfer at the electrode-electrolyte interface is represented as a resistor, often referred to as the charge transfer resistance. Additional circuit elements, such as constant-phase or Warburg elements, can also be included to represent other features of the electrochemical cell and process, such transport characteristics of the species at the electrode-electrolyte interface. The Randles model is a commonly used equivalent circuit for interpretation of biosensor EIS data. The circuit consists of an electrolyte resistance in series with a parallel combination of the double-layer capacitance with the charge transfer resistance and the Warburg impedance element (Randles, 1947). Variations of this model have been formulated for a variety of biosensing studies. For example, the equivalent circuit model and associated Nyquist plot for electrochemical detection of S. typhimurium using EIS with a poly(pyrrole-co-3-carboxyl-pyrrole) copolymer supported aptamer can be found in Fig. 5c (Sheikhzadeh et al. 2016). The equivalent circuit model consists of the solution resistance, charge transfer resistance at the copolymer-aptamer/electrolyte interface, and constant phase element for the charge capacitance at the copolymer-aptamer/electrolyte interface (Sheikhzadeh et al. 2016).
While the impedance can be measured across a range of frequencies and interpreted using equivalent circuit models that describe impedance response over a wide frequency range, fixed-frequency measurements are also useful for biosensing applications. Fixed-frequency measurements are typically based on the identification of single frequencies or small frequency ranges in the impedance spectra that are most sensitive to molecular binding events. Fixed-frequency approaches have the advantage of increasing the sampling frequency of the biosensor. As a result, impedance-based electrochemical methods generate biosensor responses in terms of changes in the measured physical quantities (e.g., changes in impedance) or calculated equivalent circuit elements (e.g., double-layer capacitance or charge-transfer resistance).
As shown in Table 1, Table 2, EIS is one of the most commonly used methods for electrochemical detection of pathogens. For example, Zarei et al. used EIS with an Au nanoparticle-modified carbon-based electrode for detection of Shigella dysenteriae (S. dysenteriae) at a LOD of 1 CFU/mL (Zarei et al. 2018). Primiceri et al. used EIS with Au interdigitated microelectrode arrays and Fe(CN)6 3 - /4- to detect L. monocytogenes at a LOD of 5 CFU/mL (Primiceri et al. 2016). Andrade et al. used EIS with a CNT-based electrode for multiplexed detection of E. coli, B. subtilis, and Enterococcus faecalis (Andrade et al. 2015).
Redox reactions at the electrode-electrolyte interface are typically established using a redox probe. Owing to its high reversibility, the Fe(CN)6 3 - /4- redox couple has been widely investigated as an electrochemical probe for biosensing applications and is regarded as a standard model for highly reversible electrochemical reactions (Daum and Enke, 1969). While useful electrochemical probes, redox reactions may also affect the electrode and immobilized biorecognition elements. For example, redox reactions associated with the Fe(CN)6 3 - /4- probe can cause etching of Au electrodes due to the presence of CN− ions when using the redox couple for EIS measurements (Vogt et al. 2016). This observation warrants further investigation, particularly in the context of establishing the effects on biosensor repeatability and reusability. The use of alternative redox probes or electrode materials may mitigate such effects. For example, ferrocene and ferrocenemethanol have also been used as redox probes for pathogen detection. Ruthenium(III)/ruthenium(II) (Schrattenecker et al. 2019) and immobilized quinone pairs (Piro et al. 2013) are also potentially useful alternatives.
Biosensors that use impedance-based methods and whose impedance response can be modeled using equivalent circuit models can be used to calculate the capacitance of the electric double layer. The double-layer capacitance is recognized to be sensitive to the structure of the electrode, the characteristics and concentration of analytes at the electrode surface and in the electrolyte, and the characteristics of the electrolyte (Lisdat and Schäfer, 2008). As a capacitor, the double-layer is not only dependent on the dielectric material but also the thickness of the dielectric layer. Importantly, both characteristics could be affected by molecular binding events on an electrode. For example, when a target analyte binds to an immobilized biorecognition element, counter ions around the electrode surface are displaced, leading to a change in the capacitance (Berggren et al. 2001). The capacitance can be determined from the reactive component of the impedance or by fitting of an equivalent circuit model (Barsoukov and Macdonald, 2018).
Idil et al. used the capacitive response of a MIP electrode for the detection of E. coli (Idil et al. 2017). Jantra et al. similarly used the capacitive response of an Au rod electrode for the detection of E. coli (Jantra et al. 2011). Luka et al. used the capacitive response of an Au interdigitated microelectrode array based on equivalent circuit analysis for the detection of C. parvum (Luka et al. 2019). See Table 1, Table 2 for a detailed list of studies that have used the capacitive response of an electrochemical biosensor for pathogen detection.

3.3.4  Conductometry
Conductometry methods are those in which the conductivity of the sample solution is monitored using a low-amplitude alternating electrical potential (Dzyadevych and Jaffrezic-Renault, 2014). The principle relies on conductivity change in the sample via the production or consumption of charged species. The measurement has the advantage of not requiring a reference electrode and can be used to detect both electroactive and electroinactive analytes (Jaffrezic-Renault and Dzyadevych, 2008; Narayan, 2016). Given the method can be performed using a two-electrode configuration, conductometric biosensors can be easily miniaturized. In addition, they are less vulnerable to many types of interference due to their differential measurement mode (Jaffrezic-Renault and Dzyadevych, 2008).
As shown in Fig. 5d, Yoo et al. used a conductometric biosensor with CNT-based electrodes for the detection of B. subtilis (Yoo et al. 2017). Mannoor et al. used a previously described conductometric biosensor to detect S. aureus and Helicobacter pylori on tooth enamel (Mannoor et al. 2012). Shen et al. detected two strains of human influenza A virus (H1N1 and H3N2) using conductometry with a silicon nanowire array at a LOD of 29 viruses/μL (Shen et al. 2012). Additional studies that have examined the use of conductometric biosensors for pathogen detection can be found in Table 1, Table 2.

3.4  Secondary binding approaches
Electrochemical biosensors would ideally produce sensitive and selective results using label-free protocols. However, secondary binding reactions are sometimes required to facilitate the robust detection of pathogens that lack initial labels depending on the biosensor characteristics and measurement demands. Secondary binding steps can facilitate target labeling, biosensor signal amplification, and verification of target binding. Secondary binding steps provide useful in situ controls and can increase sensitivity, LOD, dynamic range, and measurement confidence (e.g., verification of target binding). Secondary binding steps also provide opportunities for acquiring additional bioanalytical information about the target species.
Here, we classify assays that use secondary binding steps as labeled approaches in Table 1, Table 2 regardless of if the primary binding step produced a response. There is, however, a more subtle distinction if binding of the secondary species is used for amplification or verification purposes as previously discussed. Labels often include a biorecognition element-enzyme or -nanoparticle conjugate. In electrochemical biosensing applications, such labels often serve the purpose of altering the material properties or transport processes of the electrode-electrolyte interface, often by inducing a secondary reaction. Secondary binding of optically-active nanomaterials to captured targets can also enable the use of optical transducers for simultaneous detection or bioanalysis. Enzymes are among the most commonly used secondary binding species for label-based pathogen detection. As shown in Table 2, electrochemical biosensors for pathogen detection that employ enzymes are commonly performed as a sandwich assay format. A schematic of secondary binding steps for biosensor amplification based on the binding of HRP-antibody conjugates is shown in Fig. 6 a (Kokkinos et al. 2016). Hong et al. used HRP-labeled secondary antibodies to amplify the CV and EIS responses of a concanavalin A-functionalized nanostructured Au electrode to detect norovirus (Hong et al. 2015). Gayathri et al. used an HRP-antibody conjugate to induce an enzyme-assisted reduction reaction with an immobilized thionine-antibody receptor in an H2O2 system for detection of E. coli down to 50 CFU/mL using a sandwich assay format (Gayathri et al. 2016). Xu et al. used glucose oxidase and monoclonal anti-S. typhimurium to functionalize magnetic bead labels for separation and detection of S. typhimurium on an Au IDAM using EIS and glucose to catalyze the reaction that exhibited a linear working range of 102 to 106 CFU/mL (Xu et al. 2016b).
Fig. 6 Highlight of secondary binding and signal amplification approaches utilized in electrochemical biosensor-based pathogen detection. a) Four amplification approaches associated with the secondary binding of enzyme-labeled secondary antibodies: (A) electron transfer mediation; (B) nanostructuring of surface for increased rate of charge transfer kinetics; (C) conversion of electrochemically inactive substrate into a detectable electroactive product; (D) catalysis of oxidation of glucose for production of hydrogen peroxide for electrochemical detection (Kokkinos et al. 2016). b) Signal amplification via non-selective binding of AuNPs to bound bacterial target (E. coli) (Wan et al. 2016).
In addition to enzymes, secondary binding of nanoparticles has also been used for pathogen detection. As shown in Fig. 6b, Wan et al. utilized non-functionalized AuNPs to amplify the EIS response of an antibody-immobilized planar Au electrode to E. coli detection (Wan et al. 2016). A detailed overview of studies that employ enzymes and nanoparticles is provided in Table 2. We remind the reader that while secondary binding steps are useful techniques, assays that avoid secondary binding steps have advantages for bioprocess monitoring and control applications, as they avoid the addition of reagents to a process that may compromise product quality).