Advances in Electrochemical Aptasensors Based on Carbon Nanomaterials

Carbon nanomaterials offer unique opportunities for the assembling of electrochemical aptasensors due to their high electroconductivity, redox activity, compatibility with biochemical receptors and broad possibilities of functionalization and combination with other auxiliary reagents. In this review, the progress in the development of electrochemical aptasensors based on carbon nanomaterials in 2016–2020 is considered with particular emphasis on the role of carbon materials in aptamer immobilization and signal generation. The synthesis and properties of carbon nanotubes, graphene materials, carbon nitride, carbon black particles and fullerene are described and their implementation in the electrochemical biosensors are summarized. Examples of electrochemical aptasensors are classified in accordance with the content of the surface layer and signal measurement mode. In conclusion, the drawbacks and future prospects of carbon nanomaterials’ application in electrochemical aptasensors are briefly discussed.


Introduction
There is growing interest in the development of fast and reliable sensor devices intended for the determination of various species demanded in medicine [1], food quality [2] and environmental contamination monitoring [3]. The necessity in the sensors is due to many of their functions influencing health and life quality of population, e.g., early diagnostics of various diseases [4], chemical contamination detection [3,5] and better control of the agriculture production and foodstuffs [6][7][8].
The importance of the sensors in medicine is confirmed by impressive progress in point-of-care diagnostic devices [9][10][11] and lab-on-chip platforms [12][13][14] that combine the steps of sampling, sample treatment, separation and analysis. The continuous extension of the number of species to be analyzed in air, water and soil and the decrease in their limited threshold concentrations straighten the requirements of selectivity and sensitivity of the sensor-based analysis that exceed those formulated for conventional routine analytical instrumentation. In the last decade, many sensors have been successfully introduced into everyday life. Their application accelerated decision making in many areas sensitive to the information on the chemical content of the environment. Nevertheless, the number of chemicals under routine chemical monitoring remains quite narrow. This is due to various reasons, including the unsatisfactory analytical performance of portable analytical equipment, their insufficient mobility, and necessity in time and labor-consuming sample treatment, etc. In many cases, conventional than graphene. The reduction of GO results in the partial recovery of polyaromatic system and conductivity, but the product of such a treatment differs from ideal graphene by a significant number of defects and functional groups retained in its structure. For this reason, this product is commonly named as reduced graphene oxide (rGO). The structures of graphene, GO and rGO are outlined in Figure 1. Surprisingly, it has been found that rGO is more suitable than "pure" graphene in the assembly of aptasensors due to higher hydrophilicity and easier attachment of functional groups required for the covalent immobilization of the aptamers or for binding to the transducer/support surface. rGO has found numerous applications in electrochemical sensors and biosensors. In some cases [57][58][59][60], rGO particles are used as labels or redox indicators. Due to its size and flat aromatic structure, such labels act as carriers for Au nanoparticles and small redox active molecules. The examples of such application of graphene are given below. This results in amplification of the signal because one analyte molecule bonded follows the attachment of a multiply higher number of redox indicators. Meanwhile, the electroconductivity of the rGO particle ions is mostly sufficient for the electric wiring of the species involved in redox conversion on the electrode interface, including redox centers of oxidoreductases and DNA molecules.
The insertion of heteroatoms in the graphene structure (doping) is another way to regulate the defects and hence electrocatalytic activity of graphene. Thus, fluorine, boron, phosphorus and nitrogen have been introduced by a number of methods including calcination [61], liquid-phase reaction [62], CVD [63] and electrochemical treatment [64]. Such doped graphene materials have found application in the supercapacitors, batteries and photovoltaic devices but some examples of their use in aptasensors assembly exist and will be discussed later on. In addition, graphene is easily combined with inorganic and metallic composites that can be adjusted by the selection of the components choice and ratio to reach maximal synergetic effect, mechanical and chemical stability and compatibility with both the support and biochemical components of biosensors [65].
Carbon nanotubes (CNTs) are rolled up cylinders of graphene sheets. The CNTs can contain one tube (single-walled carbon nanotubes, SWCNTs) and several concentric tubes (multiwalled carbon nanotubes, MWCNTs) separated by 0.34 nm [66]. The diameter of the SWCNTs is commonly 0.4 -2 nm and that of MWCNTs is 2-100 nm, depending on the synthesis conditions [67]. As synthesized, CNTs are capped with fullerene-like hemispheres from both ends. In the following oxidation normally used prior to CNT application in the sensor assembly, these caps are removed and the tubes become shorter. SWNTs show the surface area up to 1600 m 2 /g [68]. As-produced MWNTs give the surface area between 200 and 400 m 2 /g. Chemical treatment with KOH increases this value to about 1000 m 2 /g [69].
CNTs are synthesized by arc discharge, laser ablation, and CVD. In arc discharge, carbon atoms are sublimated from solid precursors at 4000-6000 K [70]. This method does not require metallic catalysts, whose residues can alter the MWNTs' properties. In laser ablation, laser pulses ablate carbon target heated to 1200 o C. In laser plume, the temperature can reach 4000 K [71]. Both methods produce the CNTs with minimal defects on the sidewalls of the nanotubes. Amorphous carbon and graphitic particles are removed by controlled oxidation. In CVD, ethylene, acetylene or carbon monooxide is heated and passed through the tube furnace with metal catalyst at 550-700 o C. Carbon atoms are precipitated on the catalyst particles to form tubular carbon solids in sp 2 structure. Thus, commercially available HiPCO nanotubes are obtained with high-pressure CVD with Ni catalyst and CO as a carbon source. Surprisingly, it has been found that rGO is more suitable than "pure" graphene in the assembly of aptasensors due to higher hydrophilicity and easier attachment of functional groups required for the covalent immobilization of the aptamers or for binding to the transducer/support surface. rGO has found numerous applications in electrochemical sensors and biosensors. In some cases [57][58][59][60], rGO particles are used as labels or redox indicators. Due to its size and flat aromatic structure, such labels act as carriers for Au nanoparticles and small redox active molecules. The examples of such application of graphene are given below. This results in amplification of the signal because one analyte molecule bonded follows the attachment of a multiply higher number of redox indicators. Meanwhile, the electroconductivity of the rGO particle ions is mostly sufficient for the electric wiring of the species involved in redox conversion on the electrode interface, including redox centers of oxidoreductases and DNA molecules.
The insertion of heteroatoms in the graphene structure (doping) is another way to regulate the defects and hence electrocatalytic activity of graphene. Thus, fluorine, boron, phosphorus and nitrogen have been introduced by a number of methods including calcination [61], liquid-phase reaction [62], CVD [63] and electrochemical treatment [64]. Such doped graphene materials have found application in the supercapacitors, batteries and photovoltaic devices but some examples of their use in aptasensors assembly exist and will be discussed later on. In addition, graphene is easily combined with inorganic and metallic composites that can be adjusted by the selection of the components choice and ratio to reach maximal synergetic effect, mechanical and chemical stability and compatibility with both the support and biochemical components of biosensors [65].
Carbon nanotubes (CNTs) are rolled up cylinders of graphene sheets. The CNTs can contain one tube (single-walled carbon nanotubes, SWCNTs) and several concentric tubes (multiwalled carbon nanotubes, MWCNTs) separated by 0.34 nm [66]. The diameter of the SWCNTs is commonly 0.4-2 nm and that of MWCNTs is 2-100 nm, depending on the synthesis conditions [67]. As synthesized, CNTs are capped with fullerene-like hemispheres from both ends. In the following oxidation normally used prior to CNT application in the sensor assembly, these caps are removed and the tubes become shorter. SWNTs show the surface area up to 1600 m 2 /g [68]. As-produced MWNTs give the surface area between 200 and 400 m 2 /g. Chemical treatment with KOH increases this value to about 1000 m 2 /g [69].
CNTs are synthesized by arc discharge, laser ablation, and CVD. In arc discharge, carbon atoms are sublimated from solid precursors at 4000-6000 K [70]. This method does not require metallic catalysts, whose residues can alter the MWNTs' properties. In laser ablation, laser pulses ablate carbon target heated to 1200 • C. In laser plume, the temperature can reach 4000 K [71]. Both methods produce the CNTs with minimal defects on the sidewalls of the nanotubes. Amorphous carbon and graphitic particles are removed by controlled oxidation. In CVD, ethylene, acetylene or carbon monooxide is heated and passed through the tube furnace with metal catalyst at 550-700 • C. Carbon atoms are precipitated on the catalyst particles to form tubular carbon solids in sp 2 structure. Thus, commercially available HiPCO nanotubes are obtained with high-pressure CVD with Ni catalyst and CO as a carbon source.
Comparing with flat graphene, π-electron clouds in SWNTs exhibit significant curvature that results in partial σ-π hybridization. The C−C bond length is elongated from 1.41 Å (graphene) to 1.44 Å. Curvature leads to the dependence of electric conductivity and reactivity of the CNTs on their diameter and chirality expressed by the chirality angle (θ). It can be defined as a function of a pair of integers (n, m), θ = tan −1 [3 1/2 m/(m + 2n)]. Here, the pair (n, m) ascribes the number of lattice units in the repeating part of a graphene sheet in the SWCNT [72]. SWNTs can exert metallic (n = m), semi-metallic (n − m = 3k, where k is an integer) or semi-conducting properties (n − m 3k). For commonly used names of the SWNTs geometry, zigzag is characterized by θ = 0 • and armchair configuration with θ = 30 • (Figure 2). Ǻ. Curvature leads to the dependence of electric conductivity and reactivity of the CNTs on their diameter and chirality expressed by the chirality angle (θ). It can be defined as a function of a pair of integers (n, m), θ = tan -1 [3 1/2 m/(m+2n)]. Here, the pair (n, m) ascribes the number of lattice units in the repeating part of a graphene sheet in the SWCNT [72]. SWNTs can exert metallic (n = m), semi-metallic (n-m = 3k, where k is an integer) or semi-conducting properties (n -m≠ 3k). For commonly used names of the SWNTs geometry, zigzag is characterized by θ = 0 o and armchair configuration with θ = 30 o (Figure 2).
MWCNTs exhibit a wide diversity of the structure, which is highly affected by defects of graphene rolls. Considering relative position of inner layers, several morphological variations can be counted like hollow tube, bamboo-like MWCNTs and herringbone nanotubes. In them, graphene sheets are placed parallel to the main axis (hollow tube) or offset at a slight angle to the axis (herringbone). Bamboo-like MWCNTs are similar to herringbone structure, but inert tubes are periodically closed off into compartments like in the bamboo plant. The defects of the MWCNTs walls are decorated with oxygen-containing functional groups involved in redox catalysis of the electron transfer [73,74]. (5,5) Armchair (9,0) Zigzag Graphitic carbon nitride g-C3N4 ( Figure 3) is obtained by the polymerization of nitrogen rich precursors (cyanamide, urea, thiourea, melamine) performed by physical vapor deposition, CVD, solvothermal method or solid-state reaction [75]. The above reactions result in the formation of a non-volatile product, which can be transferred into the various nano-and microstructures, including nano flacks and microrods by exfoliation conducted under partial oxidation, sonication or chemical treatment. One of the common isothermal methods involves treatment of the g-C3N4 monolith in the nitrogen atmosphere with chloride salts (the so-called isothermal method [76]). As a result, negatively charged particles well dispersed in aqueous solutions are obtained. Dispersed products were also obtained by a cathodic reduction of carbon nitride [77]. Similar to the graphene, the g-C3N4 crystal structure can be ascribed as a hexagonal framework of the sp 2 hybridized carbon and nitrogen atoms. Two structural isomers of g-C3N4 exist, i.e., one derived from s-1,3,5-triazine core and another one from s-heptazine (1,3,4,6,7,9,9b-heptaazaphenalene). Appropriate subunits are connected with planar tertiary amino groups. The g-C3N4 exhibits semiconducting properties and the band gap (2.7 eV) makes it suitable for photocatalytic and electrochemiluminescence (ECL) processes [78,79]. MWCNTs exhibit a wide diversity of the structure, which is highly affected by defects of graphene rolls. Considering relative position of inner layers, several morphological variations can be counted like hollow tube, bamboo-like MWCNTs and herringbone nanotubes. In them, graphene sheets are placed parallel to the main axis (hollow tube) or offset at a slight angle to the axis (herringbone). Bamboo-like MWCNTs are similar to herringbone structure, but inert tubes are periodically closed off into compartments like in the bamboo plant. The defects of the MWCNTs walls are decorated with oxygen-containing functional groups involved in redox catalysis of the electron transfer [73,74].
Graphitic carbon nitride g-C 3 N 4 ( Figure 3) is obtained by the polymerization of nitrogen rich precursors (cyanamide, urea, thiourea, melamine) performed by physical vapor deposition, CVD, solvothermal method or solid-state reaction [75]. The above reactions result in the formation of a non-volatile product, which can be transferred into the various nano-and microstructures, including nano flacks and microrods by exfoliation conducted under partial oxidation, sonication or chemical treatment. One of the common isothermal methods involves treatment of the g-C 3 N 4 monolith in the nitrogen atmosphere with chloride salts (the so-called isothermal method [76]). As a result, negatively charged particles well dispersed in aqueous solutions are obtained. Dispersed products were also obtained by a cathodic reduction of carbon nitride [77]. Similar to the graphene, the g-C 3 N 4 crystal structure can be ascribed as a hexagonal framework of the sp 2 hybridized carbon and nitrogen atoms. Two structural isomers of g-C 3 N 4 exist, i.e., one derived from s-1,3,5-triazine core and another one from s-heptazine (1,3,4,6,7,9,9b-heptaazaphenalene). Appropriate subunits are connected with planar tertiary amino groups. The g-C 3 N 4 exhibits semiconducting properties and the band gap (2.7 eV) makes it suitable for photocatalytic and electrochemiluminescence (ECL) processes [78,79]. Chemosensors 2019, 7, x FOR PEER REVIEW 6 of 42 g-C3N4 on s-heptazine core Carbon black (CB) is mostly obtained by combustion of petroleum products They consist of the spherical nanoparticles with a diameter from 3 to 100 nm, which are often bonded to each other to form aggregates. The CB surface area can vary from 15 to 1000 m 2 /g, and the electroconductivity ((1.0-2)×10 3 Sm -1 [80]) is higher than that of CNTs (0.02 -0.25 Sm -1 [81]). The CB nanoparticles have a high concentration of oxygen-containing surface groups that can be involved in the covalent attachment of many biochemical receptors. Together with low cost and high accessibility, these properties make CBs very attractive for use in voltammetric biosensors [82]. Some other materials obtained in a similar manner should also be mentioned. These materials include acetylene black, which is formed in acetylene combustion and is applied in the modification of electrodes as transducers of the DNA [83] and enzyme [84] biosensors.
Fullerenes are allotropic modifications of the carbon in the form of symmetrical clusters. In them, carbon atoms in the sp 2 hybridization state are bonded to each other on the surface of a sphere at the vertices of pentagons and hexagons [85]. C60 fullerene is most intensively investigated ( Figure 4). It consists of 20 hexagons and 12 pentagons with a diameter of 0.7 nm. Fullerenes are synthesized by electric arc discharge [86], electron beam ablation [87] and sputtering [88]. However, most of these methods cannot provide sufficient quantities of the product with a regular structure. Fullerenes are rather easily modified with many functionalized compounds via defects of the structure or intermediate formation of epoxy rings. Besides, fullerenes can interact with CNTs and graphene materials by hydrophobic interactions. Meanwhile fullerenes exert efficient electron exchange ability and can mediate electron transfer in the assembly of various electrochemical sensors and biosensors with electrocatalytic response [89,90]. Carbon dots (CDs) are zero-dimension materials based on graphitic cores, which are classified in accordance with the size, structure, source and specific properties [91]. Most of them are applied in optic methods of analysis and energy storage devices. Besides CD classes, carbon nanodot, carbon quantum dot, and graphene quantum dot classes are mentioned [92][93][94]. Thus, in electrochemical sensors, graphene quantum dots are frequently utilized. They consist of up to ten single atom layers with obvious graphene lattice and have a lateral dimension of about 100 nm [95]. Meanwhile, such graphene quantum dots are more hydrophilic than graphene and have a higher electrocatalytic activity due to the predominant participation of edges in the electron transfer. Graphene quantum dots are synthesized by the electrochemical [96] or ultrasonication exfoliation [97] or by pyrolysis of small organic molecules such as trisodium citrate (hydrothermal method [98]). Carbon nanodots are Carbon black (CB) is mostly obtained by combustion of petroleum products They consist of the spherical nanoparticles with a diameter from 3 to 100 nm, which are often bonded to each other to form aggregates. The CB surface area can vary from 15 to 1000 m 2 /g, and the electroconductivity ((1.0-2) × 10 3 Sm −1 [80]) is higher than that of CNTs (0.02-0.25 Sm −1 [81]). The CB nanoparticles have a high concentration of oxygen-containing surface groups that can be involved in the covalent attachment of many biochemical receptors. Together with low cost and high accessibility, these properties make CBs very attractive for use in voltammetric biosensors [82]. Some other materials obtained in a similar manner should also be mentioned. These materials include acetylene black, which is formed in acetylene combustion and is applied in the modification of electrodes as transducers of the DNA [83] and enzyme [84] biosensors.
Fullerenes are allotropic modifications of the carbon in the form of symmetrical clusters. In them, carbon atoms in the sp 2 hybridization state are bonded to each other on the surface of a sphere at the vertices of pentagons and hexagons [85]. C 60 fullerene is most intensively investigated ( Figure 4). It consists of 20 hexagons and 12 pentagons with a diameter of 0.7 nm. Fullerenes are synthesized by electric arc discharge [86], electron beam ablation [87] and sputtering [88]. However, most of these methods cannot provide sufficient quantities of the product with a regular structure. Fullerenes are rather easily modified with many functionalized compounds via defects of the structure or intermediate formation of epoxy rings. Besides, fullerenes can interact with CNTs and graphene materials by hydrophobic interactions. Meanwhile fullerenes exert efficient electron exchange ability and can mediate electron transfer in the assembly of various electrochemical sensors and biosensors with electrocatalytic response [89,90].  Carbon black (CB) is mostly obtained by combustion of petroleum products They consist of the spherical nanoparticles with a diameter from 3 to 100 nm, which are often bonded to each other to form aggregates. The CB surface area can vary from 15 to 1000 m 2 /g, and the electroconductivity ((1.0-2)×10 3 Sm -1 [80]) is higher than that of CNTs (0.02 -0.25 Sm -1 [81]). The CB nanoparticles have a high concentration of oxygen-containing surface groups that can be involved in the covalent attachment of many biochemical receptors. Together with low cost and high accessibility, these properties make CBs very attractive for use in voltammetric biosensors [82]. Some other materials obtained in a similar manner should also be mentioned. These materials include acetylene black, which is formed in acetylene combustion and is applied in the modification of electrodes as transducers of the DNA [83] and enzyme [84] biosensors.
Fullerenes are allotropic modifications of the carbon in the form of symmetrical clusters. In them, carbon atoms in the sp 2 hybridization state are bonded to each other on the surface of a sphere at the vertices of pentagons and hexagons [85]. C60 fullerene is most intensively investigated ( Figure 4). It consists of 20 hexagons and 12 pentagons with a diameter of 0.7 nm. Fullerenes are synthesized by electric arc discharge [86], electron beam ablation [87] and sputtering [88]. However, most of these methods cannot provide sufficient quantities of the product with a regular structure. Fullerenes are rather easily modified with many functionalized compounds via defects of the structure or intermediate formation of epoxy rings. Besides, fullerenes can interact with CNTs and graphene materials by hydrophobic interactions. Meanwhile fullerenes exert efficient electron exchange ability and can mediate electron tran Carbon dots (CDs) are zero-dimension materials based on graphitic cores, which are classified in accordance with the size, structure, source and specific properties [91]. Most of them are applied in optic methods of analysis and energy storage devices. Besides CD classes, carbon nanodot, carbon quantum dot, and graphene quantum dot classes are mentioned [92][93][94]. Thus, in electrochemical sensors, graphene quantum dots are frequently utilized. They consist of up to ten single atom layers with obvious graphene lattice and have a lateral dimension of about 100 nm [95]. Meanwhile, such graphene quantum dots are more hydrophilic than graphene and have a higher electrocatalytic activity due to the predominant participation of edges in the electron transfer. Graphene quantum dots are synthesized by the electrochemical [96] or ultrasonication exfoliation [97] or by pyrolysis of small organic molecules such as trisodium citrate (hydrothermal method [98]). Carbon nanodots are Carbon dots (CDs) are zero-dimension materials based on graphitic cores, which are classified in accordance with the size, structure, source and specific properties [91]. Most of them are applied in optic methods of analysis and energy storage devices. Besides CD classes, carbon nanodot, carbon quantum dot, and graphene quantum dot classes are mentioned [92][93][94]. Thus, in electrochemical sensors, graphene quantum dots are frequently utilized. They consist of up to ten single atom layers with obvious graphene lattice and have a lateral dimension of about 100 nm [95]. Meanwhile, such graphene quantum dots are more hydrophilic than graphene and have a higher electrocatalytic activity due to the predominant participation of edges in the electron transfer. Graphene quantum dots are synthesized by the electrochemical [96] or ultrasonication exfoliation [97] or by pyrolysis of small organic molecules such as trisodium citrate (hydrothermal method [98]). Carbon nanodots are smaller (spherical particles of about 10 nm in diameter). They are mostly used in the electrochemical sensor assembly after functionalization with redox labels or receptors. Carbon nanodots are obtained by the methods described above for the CNTs synthesis. Among others, hydrothermal carbonization is currently considered the most suitable method due to its low cost, environment-and user-friendly protocol and possibility of mass production of the product [99]. In this method, organic precursors (proteins, glucose, citric acid, or chitosan) are heated in reactor. Nitrogen doping and the synthesis of the composites with inorganic components can be performed by addition of appropriate materials to the same solution. The additives can also serve as solid templates for the formation of porous particles with carbonic shell and compact core. Thus, the use of proteins allows for the synthesis of carbon dots containing up to 8 wt.% of nitrogen [100]. The carbonization of biomass can be effectively promoted by addition of Fe 2+ ions [101]. Electrochemical properties of carbon nanodots and their application for electrochemical sensing have been recently reviewed in [102].

Aptamer Immobiization
The immobilization of aptamers should provide reliable sensitive signal and its minimal drift during the aptasensor storage and operation period. Similarly, to other biosensors, all the aptamer immobilization protocols can be divided into three main groups, i.e., physical adsorption, chemical (covalent) and affinity immobilization. Their advantages and drawbacks are briefly described below. It should be noted that only methods described for assembling aptasensors based on carbon nanomaterials are mentioned and the variety of all the immobilization protocols is certainly wider.
Physical adsorption is based on weak multiple interactions involving electrostatic, donor-acceptor and hydrophobic interactions, as well as van der Waals and hydrogen bonds. The aptamers commonly bear negative charge related to the phosphate residues of the backbone and hence they can be rather easily accumulated on positively charge carriers. For this purpose, some polyelectrolytes, such as poly(ethylene imine) (PEI) [103,104], poly(diallyldimethylammonium chloride) (PDDA) or polyaniline [105], are used as a kind of molecular "glue" combining oppositely charged components of the surface layer (e.g., carboxylated carbonaceous materials and aptamers). Electrostatic assembling is governed by the charge density and flexibility of the charge carriers and is also affected by small ions present in the solution as a supporting electrolyte. As for electrochemical aptasensors, electrostatic accumulation can be improved by preliminary polarization of the electrode or its partial oxidation resulted in formation of carbonyl and carboxylate groups on the surface of electrode or the nanomaterials deposited on its surface [106]. In some cases, such oxidative treatment is performed on the preliminary stages of electrode assembling for cleaning or suspending of the carbon nanomaterials with no respect of the signal transduction. Physical adsorption is the simplest immobilization method and is performed by drop casting of the aptamer solution on the pretreated surface. Excessive quantities of reactants are removed by washing. The addition of oppositely charged polyelectrolytes can result in the formation of multilayered complexes, where the aptamer layer is additionally protected from desorption by polyelectrolyte deposited on its surface (layer-by-layer immobilization) [107]. This protocol makes it possible to vary the charge and roughness of the layer by selection of appropriate polyelectrolytes. PEI and PDDA are mostly used in these complexes as cationic species.
Physical adsorption provides reproducible characteristics of aptasensors due to self-assembling and self-organization principles. The results depend on the quality of underlying surface and immobilization conditions (ionic strength, pH, buffer composition, temperature). On the other hand, non-covalent immobilization is reversible and does not guarantee from the losses of aptamer molecules if sharp changes of the electrolyte content or of the solution pH takes place. In the case of graphene and, to a lesser extent, CNTs, hydrophobic interactions of polyaromatic lattices can be used for the immobilization of aptamers via the π-staking of aromatic systems [108]. For this purpose, pyrene fragments are first covalently attached to the terminal groups of aptamers. Physical adsorption is compatible with microfluidic techniques and paper based electrochemical sensors.
Chemical (covalent) immobilization includes the formation of covalent bonds between the aptamer and carrier. For this purpose, terminal functional groups are introduced in the aptamer assembly. If Au or Ag nanoparticles or a bare golden electrode are used as carriers, the thiol terminal group spontaneously forms the bonds Au(Ag)-SH, providing strong binding of the aptamer to the electrode [109,110]. The primary amino group is another common modifier compatible with glutaraldehyde and carbodiimide binding. In the former case, the reaction takes place between two amino groups with the formation of the Schiff base that can be then reduced by NaBH 4 to a more stable amino group ( Figure 5). purpose, pyrene fragments are first covalently attached to the terminal groups of aptamers. Physical adsorption is compatible with microfluidic techniques and paper based electrochemical sensors. Chemical (covalent) immobilization includes the formation of covalent bonds between the aptamer and carrier. For this purpose, terminal functional groups are introduced in the aptamer assembly. If Au or Ag nanoparticles or a bare golden electrode are used as carriers, the thiol terminal group spontaneously forms the bonds Au(Ag)-SH, providing strong binding of the aptamer to the electrode [109,110]. The primary amino group is another common modifier compatible with glutaraldehyde and carbodiimide binding. In the former case, the reaction takes place between two amino groups with the formation of the Schiff base that can be then reduced by NaBH4 to a more stable amino group ( Figure 5). Chitosan or ethylene diamine can be the counterparts of glutaraldehyde binding. Excessive aldehyde groups are removed by treatment with glycine. The reaction is performed at ambient temperature with glutaraldehyde solution or vapors.
Carbodiimide binding accelerates the reaction between amino and carboxylic groups to the formation of the amide bond. Commonly, carboxylic groups are present on the surface of carbonaceous materials, and the amino group is introduced in the assembly of an aptasensor sequence. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS) are common reactants applied for carbodiimide binding [111] (Figure 6). The protocol assumes preliminary treatment (activation) of the carrier with consecutive addition of aminated aptamer. The reaction is not complicated with oligomerization of the reagent or formation of multiple covalent bonds preventing access of an analyte. Introduction of carboxylate function on the electrode interface can be also performed using diazonium salt generated in situ. In this method, p-nitrobenzoic acid is first electrochemically reduced in acidic media to amino derivative, which is transformed into diazonium salt by sodium nitrite. The subsequent cathodic reaction results in the formation of phenylcarboxylic radicals that are attached to the surface of screen-printed electrode, CNTs or graphite paste electrode (electrografting protocol) [112,113]. Easy handling and formation of dense monolayers with reproducible predicable structure offer dense packing of the aptamers required for many protocols of the signal generation. The mechanism of generation of diazonium salt and its reaction with CNTs is outlined in Figure 7. Chitosan or ethylene diamine can be the counterparts of glutaraldehyde binding. Excessive aldehyde groups are removed by treatment with glycine. The reaction is performed at ambient temperature with glutaraldehyde solution or vapors.
Carbodiimide binding accelerates the reaction between amino and carboxylic groups to the formation of the amide bond. Commonly, carboxylic groups are present on the surface of carbonaceous materials, and the amino group is introduced in the assembly of an aptasensor sequence. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS) are common reactants applied for carbodiimide binding [111] (Figure 6). The protocol assumes preliminary treatment (activation) of the carrier with consecutive addition of aminated aptamer. The reaction is not complicated with oligomerization of the reagent or formation of multiple covalent bonds preventing access of an analyte. purpose, pyrene fragments are first covalently attached to the terminal groups of aptamers. Physical adsorption is compatible with microfluidic techniques and paper based electrochemical sensors. Chemical (covalent) immobilization includes the formation of covalent bonds between the aptamer and carrier. For this purpose, terminal functional groups are introduced in the aptamer assembly. If Au or Ag nanoparticles or a bare golden electrode are used as carriers, the thiol terminal group spontaneously forms the bonds Au(Ag)-SH, providing strong binding of the aptamer to the electrode [109,110]. The primary amino group is another common modifier compatible with glutaraldehyde and carbodiimide binding. In the former case, the reaction takes place between two amino groups with the formation of the Schiff base that can be then reduced by NaBH4 to a more stable amino group ( Figure 5). Chitosan or ethylene diamine can be the counterparts of glutaraldehyde binding. Excessive aldehyde groups are removed by treatment with glycine. The reaction is performed at ambient temperature with glutaraldehyde solution or vapors.
Carbodiimide binding accelerates the reaction between amino and carboxylic groups to the formation of the amide bond. Commonly, carboxylic groups are present on the surface of carbonaceous materials, and the amino group is introduced in the assembly of an aptasensor sequence. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS) are common reactants applied for carbodiimide binding [111] (Figure 6). The protocol assumes preliminary treatment (activation) of the carrier with consecutive addition of aminated aptamer. The reaction is not complicated with oligomerization of the reagent or formation of multiple covalent bonds preventing access of an analyte. Introduction of carboxylate function on the electrode interface can be also performed using diazonium salt generated in situ. In this method, p-nitrobenzoic acid is first electrochemically reduced in acidic media to amino derivative, which is transformed into diazonium salt by sodium nitrite. The subsequent cathodic reaction results in the formation of phenylcarboxylic radicals that are attached to the surface of screen-printed electrode, CNTs or graphite paste electrode (electrografting protocol) [112,113]. Easy handling and formation of dense monolayers with reproducible predicable structure offer dense packing of the aptamers required for many protocols of the signal generation. The mechanism of generation of diazonium salt and its reaction with CNTs is outlined in Figure 7. Introduction of carboxylate function on the electrode interface can be also performed using diazonium salt generated in situ. In this method, p-nitrobenzoic acid is first electrochemically reduced in acidic media to amino derivative, which is transformed into diazonium salt by sodium nitrite. The subsequent cathodic reaction results in the formation of phenylcarboxylic radicals that are attached to the surface of screen-printed electrode, CNTs or graphite paste electrode (electrografting protocol) [112,113]. Easy handling and formation of dense monolayers with reproducible predicable structure offer dense packing of the aptamers required for many protocols of the signal generation. The mechanism of generation of diazonium salt and its reaction with CNTs is outlined in Figure 7. Contrary to physical immobilization, covalent binding is site specific and provides a certain orientation of the aptamer against the underlying surface. However, the rigidity of the binding site structure and irreversible character of the immobilization require additional measures to avoid spatial limitation of an analyte access and recognition. Thus, long-chain spacers consisting of thymidine fragments (polyT), methylene or oxyethylene units significantly decrease the steric hindrance of the analyte recognition and binding [114]. Besides, spacer selection offers new opportunities in tuning the hydrophobicity and flexibility of the spacer in accordance with the nature of the underlying layer [115].
Affinity immobilization assumes application of native receptors with a very high efficiency of recognition. The avidin (streptavidin)-biotin pair is the most popular example of such interactions. The protein part of this pair (avidin or streptavidin) can bind up to four biotin molecules. The latter ones can be easily attached to an aptamer or to another protein molecule via the carboxylic group ( Figure 8). This allows for assembling branched complexes on the solid support or electrode surface and employs detection schemes similar to those of the ELISA assay or DNA hybridization analysis. Due to high affinity (dissociation constant KD ~ 10 -14-10 -15 M [116]), the avidin-biotin binding is comparable to that of antigen-antibody interaction, with higher stability to high temperature and hydrolysis.
Another affinity immobilization protocol has been proposed for label-free methods of the signal detection. In them, the auxiliary DNA sequence, complementary to the part of the aptamer (capturing DNA strand), is covalently attached to the electrode via carbodiimide binding or the Au-SH interaction described above for covalent immobilization [117,118]. The non-complementary part of aptamer can bind to an analyte. This commonly results in the folding of the linear aptamer conformation and increased density of the surface layer. Alternatively, the aptamer-analyte reaction results in the dissociation of the DNA-aptamer helix followed by the release of the aptamer-analyte complex from the electrode interface to the solution (displacement protocol). Details of signal generation in both cases are considered below in the Section 3.2.

Aptasensor Assembling
The immobilization of aptamers is of great importance for aptasensor assembling, but it is not the only step of the aptasensor assembling. In the case of using carbon nanomaterials, some additional steps are required prior to or together with the aptamer attachment.

•
Synthesis or preliminary treatment of carbon nanomaterials Contrary to physical immobilization, covalent binding is site specific and provides a certain orientation of the aptamer against the underlying surface. However, the rigidity of the binding site structure and irreversible character of the immobilization require additional measures to avoid spatial limitation of an analyte access and recognition. Thus, long-chain spacers consisting of thymidine fragments (polyT), methylene or oxyethylene units significantly decrease the steric hindrance of the analyte recognition and binding [114]. Besides, spacer selection offers new opportunities in tuning the hydrophobicity and flexibility of the spacer in accordance with the nature of the underlying layer [115].
Affinity immobilization assumes application of native receptors with a very high efficiency of recognition. The avidin (streptavidin)-biotin pair is the most popular example of such interactions. The protein part of this pair (avidin or streptavidin) can bind up to four biotin molecules. The latter ones can be easily attached to an aptamer or to another protein molecule via the carboxylic group ( Figure 8). Contrary to physical immobilization, covalent binding is site specific and provides a certain orientation of the aptamer against the underlying surface. However, the rigidity of the binding site structure and irreversible character of the immobilization require additional measures to avoid spatial limitation of an analyte access and recognition. Thus, long-chain spacers consisting of thymidine fragments (polyT), methylene or oxyethylene units significantly decrease the steric hindrance of the analyte recognition and binding [114]. Besides, spacer selection offers new opportunities in tuning the hydrophobicity and flexibility of the spacer in accordance with the nature of the underlying layer [115].
Affinity immobilization assumes application of native receptors with a very high efficiency of recognition. The avidin (streptavidin)-biotin pair is the most popular example of such interactions. The protein part of this pair (avidin or streptavidin) can bind up to four biotin molecules. The latter ones can be easily attached to an aptamer or to another protein molecule via the carboxylic group ( Figure 8). This allows for assembling branched complexes on the solid support or electrode surface and employs detection schemes similar to those of the ELISA assay or DNA hybridization analysis. Due to high affinity (dissociation constant KD ~ 10 -14-10 -15 M [116]), the avidin-biotin binding is comparable to that of antigen-antibody interaction, with higher stability to high temperature and hydrolysis.
Another affinity immobilization protocol has been proposed for label-free methods of the signal detection. In them, the auxiliary DNA sequence, complementary to the part of the aptamer (capturing DNA strand), is covalently attached to the electrode via carbodiimide binding or the Au-SH interaction described above for covalent immobilization [117,118]. The non-complementary part of aptamer can bind to an analyte. This commonly results in the folding of the linear aptamer conformation and increased density of the surface layer. Alternatively, the aptamer-analyte reaction results in the dissociation of the DNA-aptamer helix followed by the release of the aptamer-analyte complex from the electrode interface to the solution (displacement protocol). Details of signal generation in both cases are considered below in the Section 3.2.

Aptasensor Assembling
The immobilization of aptamers is of great importance for aptasensor assembling, but it is not the only step of the aptasensor assembling. In the case of using carbon nanomaterials, some additional steps are required prior to or together with the aptamer attachment.

•
Synthesis or preliminary treatment of carbon nanomaterials This allows for assembling branched complexes on the solid support or electrode surface and employs detection schemes similar to those of the ELISA assay or DNA hybridization analysis. Due to high affinity (dissociation constant K D~1 0 −14 -10 −15 M [116]), the avidin-biotin binding is comparable to that of antigen-antibody interaction, with higher stability to high temperature and hydrolysis.
Another affinity immobilization protocol has been proposed for label-free methods of the signal detection. In them, the auxiliary DNA sequence, complementary to the part of the aptamer (capturing DNA strand), is covalently attached to the electrode via carbodiimide binding or the Au-SH interaction described above for covalent immobilization [117,118]. The non-complementary part of aptamer can bind to an analyte. This commonly results in the folding of the linear aptamer conformation and increased density of the surface layer. Alternatively, the aptamer-analyte reaction results in the dissociation of the DNA-aptamer helix followed by the release of the aptamer-analyte complex from the electrode interface to the solution (displacement protocol). Details of signal generation in both cases are considered below in the Section 3.2.

Aptasensor Assembling
The immobilization of aptamers is of great importance for aptasensor assembling, but it is not the only step of the aptasensor assembling. In the case of using carbon nanomaterials, some additional steps are required prior to or together with the aptamer attachment.

•
Synthesis or preliminary treatment of carbon nanomaterials • Electrode cleaning and treatment • Aptamer modification/immobilization • Labels/indicators implementation At present, most of the carbon nanomaterials used in the assembly of electrochemical aptasensors are commercially available. In many cases, they are already modified in a manner most convenient for their use in aptasensors design. Thus, CNTs are pre-oxidized to remove fullerene hemispheres, metal catalyst residues and amorphous carbon inclusions. Only carbon dots are still prepared by the research team that developing the aptasensor. This might be referred to a rather simple and inexpensive route of their synthesis and limited availability of appropriate products to be used as they are. The GO and related graphene materials take an intermediate position: the GO suspensions are commercially available, but their reduction is often performed or adapted within the study of aptasensors. On another hand, Hummers' method of graphite oxidation to graphite oxide and then to GO [119,120] is also applied for GO synthesis instead of commercial product use. As a result, many articles devoted to the GO and rGO in the aptasensor assembly contain very similar photographs illustrating FTIR spectra or scanning electron microscopy (SEM) images of the primary coatings that should convince reader in chemical nature and structure of 'home made' GO (rGO) particles [121][122][123][124][125]. If the carbon nanomaterials are intended for the increased roughness of the electrode active area and bigger quantities of aptamer immobilized, such attention to the peculiarities of the carbon particle structure seems exhausting and actually does not contribute to better understanding of the aptasensor behavior. The situation is opposite if the same nanoparticles serve as redox labels or carriers for such labels. In this case, shape, size, charge and durability of the particles, both initial and modified with aptamers and redox labels, are of critical importance for the aptasensor performance.
Electrodes applied for aptasensor assembling are presented by a glassy carbon electrode (GCE), an Au electrode and a number of screen-printed electrodes with carbon and gold as an electroconductive layer. Their preliminary treatment does not dramatically differ from that described for other biosensors and electrochemical sensors. All the manipulations are directed to the removal of impurities, the polishing of the surface and the final production of the functional groups, which are available for the following immobilization or adsorption of other components. In the case of GCE, partial oxidation with the formation of oxygen-containing functional groups takes place during mechanical polishing and can be activated by electrochemical protocols (polarization, multiple cycling in strong acids etc.). Au electrodes should be thoroughly cleaned to provide a repeatable surface for covalent attachment of thiolated species. Deposition of carbon nanomaterials as well as their combination with Au (Ag) nanoparticles chemically or electrochemically deposited onto the primary modifier correct mechanical defects of the bare electrode and make milder the requirements to its treatment. This does not mean that primary cleaning becomes exhausting because such stages predetermine the efficiency of the nanoparticles loading and holding on a flat even surface. The choice of primary electrode (GCE vs. Au etc.) depends rather on the protocol of aptamer immobilization (availability of thiolated and aminated aptamers) than on specific electrochemical reactions that commonly require low working potential and do not assume specific interactions with underlying surface.
Contrary to carbon nanomaterials, aptamers bearing specific functional groups are not synthesized but mostly ordered and bought from appropriate suppliers. In the past years, few works devoted to the optimization of the aptamer sequence toward certain analytes were published in the framework of electrochemical aptasensors design [126,127]. However, commercial aptamers should also be pre-treated depending on their structure and/or immobilization protocol. Most often, they are slowly heated and/or treated with chemical reductant for mild breaking of the internal S-S bonds and de-hybridization of random parts of their sequences. After that, they are slowly cooled to ambient temperature and drop-casted on the modifier surface, where left to react with them to for regular coating.
The inclusion of specific functions responsible for aptamer signal can be performed at any stage of the aptasensor assembling. It should be performed in a manner providing close proximity of aptamer and redox active parts of the layer. In some cases, the formation of appropriate hybrids can be performed in one step. Thus, the formation of polymerized components can be assisted with reduction of GO [128,129].
In most cases, the formation of the layers during the loading of certain components and aptamer immobilization is monitored by conventional tools including SEM and atomic force microscopy (AFM) images, electrochemical impedance spectroscopy (EIS) and direct current (DC) cyclic voltammetry. In addition to the visualization of the surface changes with SEM and transmission electron microscopy (TEM), EIS and DC data make it possible to assess the permeability of the surface layer and the rate of the electron transfer. Ferri/ferrocyanide ions [Fe(CN) 6 ] 3-/4taken in a 1:1 molar ratio are mostly used in these experiments. This redox indicator exerts reversible behavior in electron exchange reaction, which is pH independent and does not respond on the dissolved oxygen in solution. It is important that the charge of the [Fe(CN) 6 ] 3-/4ions does not allow its accumulating on the aptamers because of the multiple phosphate anions present in their structure and bearing the same charge explaining repulsion of the indicator from the aptamer. Previously, it was established that the application of the [Fe(CN) 6 ] 3-/4redox indicator has some limitations in EIS measurements performed on Au electrodes [130]. However, no evidence of the reaction between the indicator and gold was reported within the present review area.

Signal Measurement Modes
In spite of a variety of the components implemented in the surface layer of the aptasensors, there are a limited number of signal measurement protocols, which are similar for many affinity biosensors assuming reversible binding of an analyte on the transducer surface: • Measurement of the changes in the permeability of the surface layer using diffusionally free redox indicator. The examples of particular aptasensors using the measurement protocols mentioned will be given in the following sections devoted to the applications of carbon nanomaterials. Here, general description of the protocols is presented.
Permeability assessment. The first approach mentioned is also called the label-free detection mode. In such an approach, the transfer of the redox indicator is monitored by the current of its redox rection at the electrode. Differential pulse (DPV) or square-wave (SWV) voltammetry are used for this purpose because of their higher sensitivity against the DC mode. The principal scheme of the measurement is presented in Figure 9. aptamer and redox active parts of the layer. In some cases, the formation of appropriate hybrids can be performed in one step. Thus, the formation of polymerized components can be assisted with reduction of GO [128,129]. In most cases, the formation of the layers during the loading of certain components and aptamer immobilization is monitored by conventional tools including SEM and atomic force microscopy (AFM) images, electrochemical impedance spectroscopy (EIS) and direct current (DC) cyclic voltammetry. In addition to the visualization of the surface changes with SEM and transmission electron microscopy (TEM), EIS and DC data make it possible to assess the permeability of the surface layer and the rate of the electron transfer. Ferri/ferrocyanide ions [Fe(CN)6] 3-/4-taken in a 1:1 molar ratio are mostly used in these experiments. This redox indicator exerts reversible behavior in electron exchange reaction, which is pH independent and does not respond on the dissolved oxygen in solution. It is important that the charge of the [Fe(CN)6] 3-/4-ions does not allow its accumulating on the aptamers because of the multiple phosphate anions present in their structure and bearing the same charge explaining repulsion of the indicator from the aptamer. Previously, it was established that the application of the [Fe(CN)6] 3-/4-redox indicator has some limitations in EIS measurements performed on Au electrodes [130]. However, no evidence of the reaction between the indicator and gold was reported within the present review area.

Signal Measurement Modes
In spite of a variety of the components implemented in the surface layer of the aptasensors, there are a limited number of signal measurement protocols, which are similar for many affinity biosensors assuming reversible binding of an analyte on the transducer surface: • Measurement of the changes in the permeability of the surface layer using diffusionally free redox indicator. The examples of particular aptasensors using the measurement protocols mentioned will be given in the following sections devoted to the applications of carbon nanomaterials. Here, general description of the protocols is presented.
Permeability assessment. The first approach mentioned is also called the label-free detection mode. In such an approach, the transfer of the redox indicator is monitored by the current of its redox rection at the electrode. Differential pulse (DPV) or square-wave (SWV) voltammetry are used for this purpose because of their higher sensitivity against the DC mode. The principal scheme of the measurement is presented in Figure 9.  It should be noted that the signal of [Fe(CN) 6 ] 3-/4redox indicator always decays after the binding of the analyte molecule. In this respect, the choice of DPV (SWV) voltammetry becomes not obvious. Indeed, the range of the changes recorded is limited by the value of the initial current obtained prior to the contact of the aptasensor with the analyte. In this case, the possibility of the decay quantification depends rather on the accuracy of the peak current detection than on the value of this signal. Assuming that the common deviation of the above value typical for a solid-state electrochemical sensor is equal to 2%, this gives a resolution of the possible measurements of 50 independent values (100%/2%) with no respect to the concentrations varied. Nevertheless, the approach to the use of the most sensitive techniques (DPV and SWV) dominates and in many publications preliminary study on the conditions for the signal measurement are performed with cyclic voltammetry (DC mode), allowing for much better visualization of the results but the calibration graph is obtained using PDV or SWV modes.
The decrease in the current unambiguously reflects the binding of an analyte to the aptamer attached via terminal functional group to the electrode surface. two mechanisms of such influence are considered, i.e., (i) decrease in the diffusion rate caused by the implementation of additional non-conductive agent (analyte molecules) in the surface layer and (ii) increased electrostatic repulsion between the anionic redox indicator and negatively charged phosphate groups of the aptamer sequence. Although most of the analytes determined with aptasensors belong to small molecules, their influence on the permeability of the surface layer is increased to the changes in the conformation of the aptamer sequence. Thus, guanine-rich fragments of the aptamer molecule can reversibly fold to so called G4 quadruplex ( Figure 10) that significantly increases density of the surface layer and its charge [131,132]. G4 quadruplexes are stabilized by central cation (normally, K + ). For this reason, electrochemical aptasensors are often sensitive to the electrolyte composition and namely potassium content [133].
Chemosensors 2019, 7, x FOR PEER REVIEW 12 of 42 obtained prior to the contact of the aptasensor with the analyte. In this case, the possibility of the decay quantification depends rather on the accuracy of the peak current detection than on the value of this signal. Assuming that the common deviation of the above value typical for a solid-state electrochemical sensor is equal to 2%, this gives a resolution of the possible measurements of 50 independent values (100%/2%) with no respect to the concentrations varied. Nevertheless, the approach to the use of the most sensitive techniques (DPV and SWV) dominates and in many publications preliminary study on the conditions for the signal measurement are performed with cyclic voltammetry (DC mode), allowing for much better visualization of the results but the calibration graph is obtained using PDV or SWV modes.
The decrease in the current unambiguously reflects the binding of an analyte to the aptamer attached via terminal functional group to the electrode surface. two mechanisms of such influence are considered, i.e., (i) decrease in the diffusion rate caused by the implementation of additional non-conductive agent (analyte molecules) in the surface layer and (ii) increased electrostatic repulsion between the anionic redox indicator and negatively charged phosphate groups of the aptamer sequence. Although most of the analytes determined with aptasensors belong to small molecules, their influence on the permeability of the surface layer is increased to the changes in the conformation of the aptamer sequence. Thus, guanine-rich fragments of the aptamer molecule can reversibly fold to so called G4 quadruplex ( Figure 10) that significantly increases density of the surface layer and its charge [131,132]. G4 quadruplexes are stabilized by central cation (normally, K + ). For this reason, electrochemical aptasensors are often sensitive to the electrolyte composition and namely potassium content [133]. Pinhole aptamers with pseudo-circle areas subdivided with self-hybridized double-stranded items are another example of the aptamers, dramatically changing their conformation in an analyte recognition. Being synthesized for the so-called E-sensors [134], they contain two terminal groups applied for attaching to the electrode and for the binding label to be detected. Transfer from pinhole to linear structure increases the distance between the label and electrode as decreases the signal ( Figure 11). Methylene blue and ferrocene are mostly used in such aptamers. Although their application in combination with carbon nanomaterials is rather rare, such an approach offers opportunities in creating label-and reagent-free aptasensors with one-step measurement protocol.
Both mechanisms affect charge transfer resistance measured by EIS in the presence of the same redox indicator. Measurement is performed at the equilibrium potential assessed as a half sum of the Pinhole aptamers with pseudo-circle areas subdivided with self-hybridized double-stranded items are another example of the aptamers, dramatically changing their conformation in an analyte recognition. Being synthesized for the so-called E-sensors [134], they contain two terminal groups applied for attaching to the electrode and for the binding label to be detected. Transfer from pinhole to linear structure increases the distance between the label and electrode as decreases the signal ( Figure 11).
Chemosensors 2019, 7, x FOR PEER REVIEW 12 of 42 obtained prior to the contact of the aptasensor with the analyte. In this case, the possibility of the decay quantification depends rather on the accuracy of the peak current detection than on the value of this signal. Assuming that the common deviation of the above value typical for a solid-state electrochemical sensor is equal to 2%, this gives a resolution of the possible measurements of 50 independent values (100%/2%) with no respect to the concentrations varied. Nevertheless, the approach to the use of the most sensitive techniques (DPV and SWV) dominates and in many publications preliminary study on the conditions for the signal measurement are performed with cyclic voltammetry (DC mode), allowing for much better visualization of the results but the calibration graph is obtained using PDV or SWV modes.
The decrease in the current unambiguously reflects the binding of an analyte to the aptamer attached via terminal functional group to the electrode surface. two mechanisms of such influence are considered, i.e., (i) decrease in the diffusion rate caused by the implementation of additional non-conductive agent (analyte molecules) in the surface layer and (ii) increased electrostatic repulsion between the anionic redox indicator and negatively charged phosphate groups of the aptamer sequence. Although most of the analytes determined with aptasensors belong to small molecules, their influence on the permeability of the surface layer is increased to the changes in the conformation of the aptamer sequence. Thus, guanine-rich fragments of the aptamer molecule can reversibly fold to so called G4 quadruplex ( Figure 10) that significantly increases density of the surface layer and its charge [131,132]. G4 quadruplexes are stabilized by central cation (normally, K + ). For this reason, electrochemical aptasensors are often sensitive to the electrolyte composition and namely potassium content [133]. Pinhole aptamers with pseudo-circle areas subdivided with self-hybridized double-stranded items are another example of the aptamers, dramatically changing their conformation in an analyte recognition. Being synthesized for the so-called E-sensors [134], they contain two terminal groups applied for attaching to the electrode and for the binding label to be detected. Transfer from pinhole to linear structure increases the distance between the label and electrode as decreases the signal ( Figure 11). Methylene blue and ferrocene are mostly used in such aptamers. Although their application in combination with carbon nanomaterials is rather rare, such an approach offers opportunities in creating label-and reagent-free aptasensors with one-step measurement protocol.
Both mechanisms affect charge transfer resistance measured by EIS in the presence of the same redox indicator. Measurement is performed at the equilibrium potential assessed as a half sum of the Methylene blue and ferrocene are mostly used in such aptamers. Although their application in combination with carbon nanomaterials is rather rare, such an approach offers opportunities in creating label-and reagent-free aptasensors with one-step measurement protocol.
Both mechanisms affect charge transfer resistance measured by EIS in the presence of the same redox indicator. Measurement is performed at the equilibrium potential assessed as a half sum of the peak potentials on a cyclic voltammogram of the [Fe(CN) 6 ] 3-/4redox pair. Its shift with the deposition of the layer components is interpreted in terms of the charge and electroconductivity of the compounds deposited. The sensitivity of the EIS measurement is commonly higher than that of voltammetry but the choice of equivalent circuit for data fitting and EIS parameters calculation can be a source of possible errors and false conclusions. This is especially true if the interface of the aptamer-analyte interaction is far from the plane assumed in most common Randles circuit applied. The higher the roughness is the more complicated is the circuit and the more parameters are introduced to meet real behavior of an aptasensor.
Assessment of the permeability of the aptasensor layer and its changes with analyte concentration offer strict requirements to the morphology of the aptamer layer. Ideally, monolayer with regular structure and no defects with naked electrode "holes" is desirable. The use of golden bare electrode including EIS and surface plasmon resonance (SPR) measurements are near this ideal. In the case of the deposition of carbon nanomaterials and for other primary transducers applied, the interface is additionally treated with the substances preventing the non-specific adsorption of reactants and analyte is necessary. Bovine serum albumin for GCE and thiolated alcohols for the Au electrode are mentioned at this stage of aptasensor assembling [135][136][137][138][139][140].
The detection of label signals or redox indicators remained in the aptasensor surface layer. Although permeability assessment is a general approach popular for all the affinity biosensors, including those based on immunoreactions and single-stranded DNA probes, it leaves occasions for false decisions due to non-specific adsorption and possible defects in the surface layer. This might be crucial, especially in real sample analysis. Alternative approaches to the signal measurement utilize more specific substances. The use of covalently attached labels and redox probes specifically interacting with DNA/aptamer molecules is one of such solutions becoming popular in the past decade. Ferrocene and Methylene blue are mostly applied as such redox active substances. They can be covalently attached to the aptamer sequence like in the case of E-sensor described above [134]. Meanwhile, Methylene blue is often used in the soluble form due to its ability to interact with singleand double-stranded DNAs with appropriate changes of its redox activity. Possible protocols combine the concept of E-sensor with pinhole aptamer and displacement mode that is often used in affinity biosensors. The reaction schemes are outlined in Figure 12.
In these aptasensor, auxiliary single-stranded DNA (capturing DNA) is used to attach the aptamer to the electrode surface by a hybridization step. The double-stranded product of hybridization is saturated with Methylene blue molecules, which can both intercalate the DNA helix and adsorb in the minor grooves on the DNA surface. Its signal is not as high because of the shielding of the molecules intercalated DNA. After the addition of the analyte molecules, the formation of aptamer-analyte complex shifts the hybridization equilibrium and liberates auxiliary DNA remaining attached to the electrode. After the removal of all dissolved components, the Methylene blue signal decreases against its primary value because of the partial leaching of dye molecules from the surface layer and their substitution with the analyte molecules ( Figure 12A). In the second protocol, auxiliary DNA with terminal redox label contains complementary fragments capable of self-hybridization and the formation of pinhole configuration. However, its reaction with complementary aptamer prevents this conformation and the redox label is placed too far from the electrode for a fast electron transfer. The removal of the aptamer binding to analyte molecules restores the pseudo-circle configuration of the auxiliary DNA and makes the label positioned in the proximity of the electrode surface ( Figure 12B). In the second case, ferrocene label is used together with Methylene blue.
Aptamer can also be modified with redox label. In this case, it first attaches to the electrode surface by hybridization with a partially complementary DNA strand (capturing DNA) and then released in the solution in the complex with an analyte ( Figure 12C).
Signal amplification strategies. Although carbon nanomaterials increase the signal due to the enhancement of the surface area for aptamer immobilization and due to the multiplication of the redox labels covalently attached to their surface, an additional increase in sensitivity can be reached by the implementation of catalytic cycles based on electro-or enzymatic catalysis or biochemical approaches to the DNA amplification. Enzymes as labels in aptasensors employ the strategy well known in ELISA and sandwich immunoassay [141,142]. Here, peroxidase and its natural (hemin) and artificial analogs (Pt catalysts, metal-organic framework particles) are introduced in the aptamer-analyte complex and then monitored by the catalytic current related to H 2 O 2 reduction/oxidation. The similarity is stressed by the possibility of joining both approaches and assemble a hybrid complex on the surface with aptamer attached to the electrode and secondary (signaling) antibody to the analyte molecule. Enzyme is then immobilized onto the hybrid complex by avidin-biotin binding [143].
The detection of label signals or redox indicators remained in the aptasensor surface layer. Although permeability assessment is a general approach popular for all the affinity biosensors, including those based on immunoreactions and single-stranded DNA probes, it leaves occasions for false decisions due to non-specific adsorption and possible defects in the surface layer. This might be crucial, especially in real sample analysis. Alternative approaches to the signal measurement utilize more specific substances. The use of covalently attached labels and redox probes specifically interacting with DNA/aptamer molecules is one of such solutions becoming popular in the past decade. Ferrocene and Methylene blue are mostly applied as such redox active substances. They can be covalently attached to the aptamer sequence like in the case of E-sensor described above [134]. Meanwhile, Methylene blue is often used in the soluble form due to its ability to interact with singleand double-stranded DNAs with appropriate changes of its redox activity. Possible protocols combine the concept of E-sensor with pinhole aptamer and displacement mode that is often used in affinity biosensors. The reaction schemes are outlined in Figure 12. With regard to the amplification of the DNA quantities at the electrode interface, such approaches offer superior sensitivity and selectivity with the limit of detections (LOD) on atto-and femtomolar levels. In them, auxiliary DNA strands are involved in the polymerization cycle with formation of a long DNA sequences consisting of partially complementary strands, those formation assisted with exonuclease is triggered by target recognition of an analyte molecule. First developed for fluorescence detection [144], such systems are rather complex and rarely use carbon nanomaterials due to requirements of site-specific immobilization and careful arrangement of all the reactants on the transducer interface. In addition, some reactions are performed in bulk solution so that this protocol is not exactly a biosensor but biosensing device.
In another example, Y-DNA, containing three strands partially complementary to each other, was applied for assembling the G4 quadruples/hemin complex from two pinhole aptamers and another one immobilized on the electrode surface. The catalytic cycle of the analyte release resulted in the amplification of the quantities of the complex mimicking peroxidase activity in the presence of hydrogen peroxide. This improved the electrochemical performance of the biosensor due to the oxidation of Fe(II) in hemin [145].
Other detection schemes. The above description does not exhaust the variety of the methods for monitoring aptamer-analyte interactions described in the aptasensors using carbon nanomaterials. Thus, the intrinsic redox activity of electroactive polymers such as polyaniline and polypyrrole changes with analyte binding due to variation of the charge distribution and counter ion access within the polymer layer [105]. In some cases, aptamer acts as a specific sorbent and accumulates analyte molecules for their following oxidation on electrode. Although such approaches show lower sensitivity against main schemes described above, they offer reliable results due to control of the processes on the electrode by working potential and its shift with the analyte addition. Then, doped MCNTs and graphene nanosheets have been described in the assembly of field effect transistors (FET) [146][147][148]. Appropriate aptamers can be immobilized on their surface by hydrophobic interactions. In such devices, aptamer folding caused by target interaction results in significant changes of the charge near the gate area of the FET and hence the currents flow between source and drain. This can find analytical applications though the range of concentrations determined could be rather narrow.

Carbon Nanotubes
The analytical performance of electrochemical aptasensors with the CNTs in their assembly is summarized in Table 1 for last five years (2016-2020). Some other examples of application of the CNTs in combination with other carbonaceous nanomaterials are considered below in the sections devoted to the second component of the mixture (graphene or g-C 3 N 4 ).
As mentioned above, CNTs play a rather passive role in the assembly of electrochemical aptasensors. Though they can exert electrocatalytic activity, which is attributed to the edges of the graphene rolls and sidewall defects; it is rarely used for signal amplification. In most aptasensors described for the determination of small molecules, CNTs are used as supports for the immobilization of aptamers and auxiliary reagents required for signal detection. It should be noted that CNTs themselves have electric conductivity typical rather for semiconductors than for metals. Thus, their deposition on the electrode does not obligatory follow improved parameters of the electron transfer. Increased currents recorded by voltammetric techniques (DC voltammetry, DPV, SWV, etc.) are mostly related to the increased surface of the working electrode. This effect is quantified by the peak currents of the redox probe (commonly [Fe(CN) 6 ] 3-/4-) in accordance with the Randles-Sevcik equation. In many aptasensors described, the authors used additional materials with a higher conductivity, such as Au or Pt nanoparticles or polyaniline, that compensate for the modest conductivity of the CNTs and provide better shape of the peaks and lower contribution of the resistance to the electron transfer at the electrode interface. However, even non-conductive materials such as ferrites or metal oxides, which cannot participate in the chain of the electron transfer often improve the characteristics of electrochemical aptasensors with CNTs due to improved mechanical properties of the surface layer and stabilization of the porous 3D structure of the CNTs net promising both high permeability for the charge transfer carriers and electric wiring of the aptamers with electrode. The authors announce a synergetic effect of such materials but few of them provided the results obtained with applied auxiliary non-conductive compounds with no CNTs. To some extent, the tradition of application of metal oxides appeared from other examples of aptasensing that employed fluorescence signal (carbon and metal nanodots, graphene nanoflakes) or magnetic separation of biochemical reactants in conventional immunoassay (ferromagnetic species). Explaining the need in such auxiliary components of the CNTs layer, the authors refer to their strong mechanical, magnetic, semiconductor and other properties [135,[149][150][151] that make sense in other protocols of signal measurement and are then transferred to the area of "pure" electrochemical analysis. In many cases, such a strategy works, but the results should depend on the size distribution of both CNTs and no-conductive electrochemical inactive components.

17β-Estradiol
Origami paper folded planar electrode with microfluidic channels modified with aminated SWCNTs, Au nanoparticles and New Methylene Blue redox indicator. Aptamer immobilized via terminal SH group DPV signal of New Methylene Blue redox indicator 10-500 pg/mL, LOD 5 mg/mL [163] GCE covered with MWCNTs with adsorbed thionine and Au nanoparticles, thiolated aptamer immobilized via Au-SH binding DPV measurement of thionine and estradiol oxidation peaks 12 pM-60 nM, LOD 1.5 pM [164] Furaneol Silver electrode covered with monolayer of cysteamine followed by CNTs immobilization and covalent attachment of capturing DNA strand hybridized with the aptamer nearing Methylene blue as redox indicator Displacement protocol with SWV determination of decreasing signal of Methylene blue oxidation 1 fM-35 µM, LOD 0.557 fM [165] Ibuprofen GCE covered with MWCNTs treated with terephtalaldehyde. Aminated capturing DNA strand is immobilized by carbodiimide binding and complementary aptamer by hybridization.
Aptamer is saturated with the MB molecules, analyte released MB from the surface layer so that its signal recorded with DPV decreases with ibuprofen concentration    Urea GCE covered with MWCNTs with electrochemically deposited Au nanoparticles. Thiolated aptamer with urea immobilized on the Au by Au-SH binding. Then, dopamine was electropolymerized and urea removed EIS measurements of charge transfer resistance 0.005-0.1 nM, 1.00-500 nM, LOD 900 fM [189] Zearalenon Au electrode covered with PEI-MoS 2 -MWCNTs composite and Pt@Au nanoparticles followed by thiolated aptamer immobilization Toluidine blue (redox indicator) signal measured by cyclic voltammetry 0.5 pg/mL-50 ng/mL, LOD 0.17 pg/mL [190] This might result in a higher deviation of the signal and the necessity to use relative parameters such as the sight of the signal against blank experiment or against its maximal deviation within the analyte concentration range considered. The same can be referred to the use of surfactants and polyelectrolytes, which simplify the dispersion of CNTS and improve the stability of appropriate suspensions, especially those used as redox indicator carriers.
Au nanoparticles obtained electrochemically or chemically on the CNTs support play anchoring role for immobilization of the aptamers with terminal SH-group. They increase the density of the aptamers on the electrode surface against bare gold and in some cases result in more regular coating required in the label-free impedimetric detection of the signal. It should be noted that the use of CNTs as templates in the synthesis of metal nanoparticles improves their morphology and makes even the size distribution against the same process on bare electrode. This is another argument in favor of the use of CNTs prior to metal deposition.

Graphene Materials
As was mentioned above, rGO is mainly used for the aptasensor assembling. Its following reduction by cathodic polarization or chemical treatment retains a certain number of functional groups for aptamer immobilization. It should also be noted that graphene itself is rather hydrophobic and easily loses its excellent properties due to reverse amalgamation to graphite-like nanostructures. For this reason, the selection of solvents for suspending graphene materials and electrolyte content can significantly alter the aptasensor performance. Doping graphene with heteroatoms improves its properties in the aptasensor assembly. Together with partial oxidation and amination, this makes it possible to change main characteristics of the surface charge and durability of the coating. Graphene materials are easily combined with other carbon nanomaterials that are retained in the composites by π-π stacking interactions of aromatic systems. In such complexes, the individual characteristics of additives, as a rule, retained and can be used for the following immobilization of aptamer via Au-SH interactions or carbodiimide binding. Aminated and carboxylated graphene particles are also implemented in polyelectrolyte complexes. Interestingly, the adsorption of organic dyes onto graphene sheets is quite effective and a reliable way to implement them as redox labels together with aptamers or auxiliary DNA strands. Separation of the reactants (dyes on graphene and aptamer on Au nanoparticles) does not alter the sensitivity of the measurement by making aptasensor assembling easier. Graphene materials can also serve as templates for electropolymerization. In addition to the electric wiring and providing immobilization support, the products of electropolymerization make the interface more hydrophilic and improve the reproducibility of EIS measurements based on the transfer of small multiply charged [Fe(CN) 6 ] 3-/4on the graphene interface. Contrary to CNT-based aptasensors, the signal of aptasensors based on graphene materials is less sensitive to non-specific adsorption of the real sample components. The necessity to block bare surface of the electrode or to treat them with bovine serum albumin (BSA) or surfactants is not obligatory and can be often omitted. The application of graphene materials in the assembly of electrochemical aptasensors is summarized in Table 2 for the period from 2016 to 2020.

Other Carbon Nanomaterials
The examples of the application of g-carbon nitride, carbon black, fullerene and related materials in electrochemical aptasensors are presented in Table 3. g-C 3 N 4 particles are mostly obtained from melamine though there are many other nitrogen containing organic compounds applicable as their sources. The structure, electroconductivity and purity of the product highly depend on the conditions of pyrolysis and calcination and vary from one article to another one. This variability offers additional possibilities to affect the aptasensor performance but is still rarely used compared with CNTs or graphene materials. Carbon nitride has never been used alone in the aptasensor assembly. It is involved in various composite materials that mostly include metal oxides, graphene nanosheets, Au and Ag nanoparticles.  [197] Myoglobin Indium-tin oxide (ITO) electrode modified with rGO modified with poly(ethylene imine) and electrostatically accumulated aptamer DPV reduction peak referred to the myoglobin redox activity 0.001-1000 ng/mL, LOD 0.97 pg/mL [198] Ochratoxin A     Tobramycin Au electrode covered with suspension of carbon nanospheres embedded with SnO x @TiO 2 obtained by calcination of the mixture of bovine serum albumin, Na 2 SnO 3 and butyl titanate. Aptamer immobilized by electrostatic adsorption EIS measurements of the charge transfer resistance shift 0.01-5 ng/mL, LOD 6.7 pg/mL [228] Metal oxides can be obtained "in one pot" with g-C 3 N 4 particles by mixing appropriate precursors and their heating in autoclave. As a sign of the 'green chemistry' approach, biomasses and agriculture wastes can be used [229].
Other carbon nanomaterials mentioned here are also applied mostly as auxiliary components of the surface layer required for the control of the density, mesoporous structure and roughness of the coating. They are always combined with other modifiers and are not used alone. Due to their own size being much lower than that of CNTs or graphene nanosheets, carbon nitride, carbon dots and carbon black do not predetermine the properties of the surface layer but improve some of the existing ones, such as permeability for small ions or increased adsorption capacity. To some extent, carbon black and carbon dots can be considered as a less expensive alternative to graphene and CNTs, although, at present, economic factors do not play a significant role in the biosensors production. The cost of carbon materials always remains much lower than that of biochemicals required (aptamers, capturing DNA strands etc.). Except CDs and fullerene, they are rarely applied in microsensors because of diversity of properties and irregular content of many industrial samples. On the other hand, easy synthesis of complex multicomponent composites by the hydrothermal method offers opportunities for extending the number of such compounds available for aptasensor development.
It is important to note that the sensitivity of most of the aptasensors based on carbon nanomaterials is extremely high and meets most requirements of medical diagnostics and environmental monitoring. Moreover, in some cases concerning the determination of organophosphate pesticide, the concentrations determined are so small that it is hard to find the analytical problems corresponding to their value. Indeed, the detection of profenofos on the level of 0.025 ng/mL [174] or malathion on the level of 0.1 fM [168] does not have sense in terms of food safety or contamination assessment. To some extent, this also refers to the determination of some metabolites, such as trypthophan and urea. One of the possible reasons to reach lower and lower LOD values is to allow for a higher dilution of the samples and hence suppress matrix interference. However, in other cases, especially for cancer biomarker detection, the extremely high sensitivity of the aptasensrs described seems an important advantage due to opportunities of early diagnosis of cancer diseases and hopes for more successful therapy of such patients. Regarding relative influence of different modifiers on the analytical performance of aptasensors, only a few analytes have been determined with various content of the surface layer. Some of such works (see determination of thrombin and CEA) were performed in the same research groups with rather insignificant changes of the measurement conditions. It can be concluded from the comparison of these results that the use of carbon nanomaterials in combination with Au nanoparticles or other catalytic systems provides better performance against 'pure' carbon nanomaterial. Then, the biochemical amplification mechanism also increased the sensitivity of the assay, even to a higher extent than the inclusion of additional inorganic components into the aptasensor assembly. It should be noted that such systems, being very sensitive, require much more reagents and extra stages so that the duration of the measurement cycle stands to about one hour. Together with the requirements of thermosetting, this makes such measurement protocols incompatible with the biosensor idea assuming portable instrumentation with fast and rather simple measurement protocol. Such a complication of the bioassay is rational if biomarkers crucially important for danger disease diagnostics are detected. Regarding signal measurement mode, EIS shows a much higher sensitivity of the response against voltammetry though the use of DPV technique can decrease this difference to acceptable levels. Meanwhile, EIS requires careful control of non-specific adsorption and can provide positive false in the diagnostics of some biomarkers.

Conclusions
During the past decades, aptamers have emerged from the laboratories and have become a conventional tool of biomedical assay, especially for the detection of disease biomarkers, drug residues and environmental pollutants. Similar to the conventional immunoassay technique, aptamers are also reversibly bonded to target species, but they are more stable toward hydrolysis and oxidative damage and can be easily modified for implementation in biosensor assembly. The number of aptamers selected for the detection of particular species is growing dramatically. However, their use in biosensors as recognition elements has some limitations related to the introduction of aptamers in the biosensors and achievement of the highest reliable signal toward a specific analyte. It should be mentioned that most of the species detected are small against the aptamer structure so that its binding results in rather modest changes of the interface conditions. To some extent, the sensitivity problem is solved by the use of aptamers changing its conformation (G 4 quadruplexes and pinhole sequences). Nevertheless, their use is especially sensitive to immobilization conditions. Taking into account the sub-nanomolar levels of analytes being detected, all above makes the development of aptasensors challenging and requires new solutions in aptasensor development and their signal optimization.
Carbon nanomaterials are considered a universal solution of many problems in biosensor design. The variety of their structure, size and morphology, unique electrical, optical and mechanical properties, and broad possibilities of functionalization and assembling in complex 3D structures make carbon nanomaterials very attractive for application in aptasensors. Numerous examples presented in the review show the diversity of approaches and achievements as a success story of aptasensors design. Many stages of the aptasensor assembling have been investigated in detail and make the process predictable and understandable. Sensitivity of detection and general usability of aptasensors become most important criteria in comparison with other analogs.
The following trends in the application of carbon nanomaterials in the electrochemical aptasensors can be emphasized:

1.
Transfer from general label-free measurement protocols to more specific modes with increased reliability of the results, especially in real sample assay.

2.
Synthesis and application of more complex and sophisticated materials that combine advantages of various components, e.g., electropolymerized materials, molecularly imprinted polymers, polyelectrolyte complexes on the support of carbon nanomaterials. 3.
Using the simplest and most reliable approaches to the assembling of the surface layer, especially form microfluidics and microsensor devices based on carbon paper and origami constructs.

4.
Varying redox labels/indicators as a route to higher robustness of the signal, especially in multicomponent samples.
In spite of diversity of publications, some problems in the design and application of aptasensors should be mentioned.

1.
Few works are devoted to the multiplex assay and simultaneous determination of several analytes. Meanwhile, this might be critical for the implementation of aptasensors in diagnostic tools or environmental monitoring stations.

2.
Most of the works do not consider the problems related to the changes in the aptasensor performance during the storage/operation period. Real sample assay is often presented by very simple tasks (analysis of drinking water, spiked samples of biological fluids etc.).

3.
The assessment of metrological characteristics is often formal and assumes extrapolation of calibration graphs by several orders of magnitude of concentration (LOD calculation). The uncertainty in the preparation of solutions with extra-low concentration of an analyte (below 1 pM) is not considered or taken into account.

4.
Most of the newly proposed aptasensors require labor-and time-consuming operation and multistep synthesis of necessary reactants. One-step measurement is the main goal and main advantage of biosensors as an alternative to conventional laboratory instrumentation. 5.
The number of aptamers used in aptasensor assembly remains limited. This might result from present tasks in appropriate areas of their application, but needs further extension, meaning future progress in medicine, the food industry and environmental pollution.