An Overview of Carbon Nanotubes and Graphene for Biosensing Applications
- 4.5k Downloads
With the development of carbon nanomaterials in recent years, there has been an explosion of interests in using carbon nanotubes (CNTs) and graphene for developing new biosensors. It is believed that employing CNTs and graphene as sensor components can make sensors more reliable, accurate, and fast due to their remarkable properties. Depending on the types of target molecular, different strategies can be applied to design sensor device. This review article summarized the important progress in developing CNT- and graphene-based electrochemical biosensors, field-effect transistor biosensors, and optical biosensors. Although CNTs and graphene have led to some groundbreaking discoveries, challenges are still remained and the state-of-the-art sensors are far from a practical application. As a conclusion, future effort has to be made through an interdisciplinary platform, including materials science, biology, and electric engineering.
KeywordsBiosensor Carbon nanotubes (CNTs) Graphene
A biosensor refers to a sensing device that transfers a biological event to a measurable signal. It usually consists of a biological recognition element and a transducer to translate the biology information to the measurable signal . To be specific, the biological recognition element in a biosensor must be selective to a certain biomolecule, biology process, or chemical reaction. Depending on the types of the recognition elements, the biosensors are able to detect a wide range of biological elements, such as nucleic acids, antibodies, enzymes, bacteria, and viruses . To date, the biosensors have been tested for their usage in food, environmental, and human samples . The biological recognition elements often are immobilized onto the surface of transducer with high bioactive for targeting. The attachment methods include adsorption, encapsulation, entrapment, covalent binding, and cross-linking . The interaction between the recognition element and the target will then be monitored and further converted to a readable signal like current. Depending on the interaction models, different types of transducers can be utilized to convert the recognition events into a digital signal that is proportional to the presence and the amount of the target. The most common transducing methods include electrochemical, optical, piezoelectric, and magnetic. Among them, electrochemical, electrical, and optical techniques are very popular due to the fast response and the flexibility in employing recognition elements .
A great effort has been given in the past few years in the worldwide range on developing novel biosensors with high sensitivity and selectivity. The recent, fast development of nanomaterials has made a profound influence on the development of biosensors. The application of nanomaterials has been given to all technical components of biosensors from recognition components to signal processers. When the material’s size is reduced to nanoscale, the interesting changes in chemical and physical properties are happened due to two principal factors: surface effect and quantum effect. The surface to volume ratio of nanomaterials increases dramatically compared to their bulk form and is able to improve the sensitivity of biosensors through increasing the interface for recognition element allocation. The quantum confinement phenomenon can lead to an increase in the band-gap energy and a blue shift in light emission with decreasing size. As a result, the electrical and optical properties of nanomaterials become size and shape dependent. These essential features of nanomaterials make it possible to turn chemical and physical properties to specific biosensor applications by controlling their size, shape, and chemical composition .
The world of nanomaterials is huge and consists of various materials with different nature, size, shape, composition, chemistry, etc. For biosensors, nanomaterials like carbon nanotubes (CNTs) and graphene are popular and at the forefront of the research [6, 7]. These two are also the most representatives in the big family of carbon nanomaterials . To date, CNTs and graphene have been widely studied for biosensor applications due to their unique chemical and physical properties . In this review article, we give a brief review on the recent developments of CNT- and graphene-based biosensors, aiming to provide a comprehensive introduction to researchers who are new to this field. The article also gives a brief perspective summary on the challenges of these biosensors toward the practical application.
2 Carbon Nanotube-Based Biosensors
CNTs have had a profound impact on a wide range of applications because of their unique electronic, chemical, and mechanical properties . CNTs are made of cylinders of sp 2-hybridized carbon atoms with several nanometers in diameter and many microns in length. There are two classes of CNTs, single walled carbon nanotubes (SWNTs) and multiwalled carbon nanotubes (MWNTs). SWNTs can be considered as one rolled-up graphene sheet, while MWNTs are concentric tubes separated by about 0.34 nm of two or more rolled-up graphene sheets. SWNTs have very unique electrical properties, depending on the chirality of the wrap, and they can behave as either metals or semiconductors [11, 12]. Recent studies have established the fact that several intriguing properties of CNTs, such as their nanodimensions and graphitic surface chemistry , make them extremely attractive for new types of electrochemical, electric, and optical biosensors .
2.1 Pre-functionalization of CNTs
It is known that one of the biggest barriers for developing CNTs-based biosensors is the dispersion issue caused by the high surface energy of CNTs. It results in difficulties in handling CNTs in a controlled way, and most solvents cannot suspend CNTs well. In order to overcome this deficiency, CNTs are usually functionalized with polymer and small molecules to render the surface compatibility to solvents and bioenvironments for further biosensing applications . Surface functionalization can be made through covalent and noncovalent bonding. For covalent bonding approach, the most common one is the oxidation of CNTs with an acid such as nitric acid and the mixture of nitric acid and sulfuric acid . Depending on the types of the employed oxidizing agents, carboxyl or hydroxyl groups are introduced onto the ends and the sidewalls of CNTs during the oxidation . These groups lead the reduction of the van der Waals interactions between CNTs and enable further modifications to covalently connect with other molecules, like long alkyl chains, polymeric molecules, dendrimers, nucleic acids, and enzymes . Compared to the oxidation approach, the direct covalent functionalization gives stronger influence on chemical and physical properties of CNTs and provides opportunities for further CNTs-associated applications . In nonplanar π-conjugated carbon framework, two factors govern the chemical reactivity of the sidewalls of CNTs: (1) curvature-induced pyramidalization at the individual carbon atoms and (2) misalignment of π-orbital between adjacent carbon atoms [18, 19, 20]. Some highly reactive species (like halogens, radicals, carbenes, or nitrenes) are the ideal reagents for covalent functionalization of the sidewalls . These groups can be bonded onto π-conjugated carbon structures of the CNTs through a series of addition reactions as introduced in the article by Balasubramanian and Burghard .
Modification via 1, 3-dipolar cycloaddition is another widely used type of covalent sidewall functionalization of CNTs . The attachment of 1, 3-dipolar cycloaddition of azomethine-ylide onto the graphite sidewall of CNTs is generated by condensation of an aldehydes and an α-amino acids . A pyrrolidine ring was formed on the CNTs surface through the reaction between C=C bond and azomethine-ylide . Functional groups introduced via above methods enable CNTs soluble in aqueous or organic solvents and open the possibility for the further fabrication of CNT-based biosensors .
Compared to covalent functionalization, noncovalent functionalization of CNTs keeps the structure of CNTs intact and thus retains their physical properties . Noncovalent interactions include electrostatic interaction, π–π stacking, van der Waals force, and hydrophobic or hydrophilic interactions are efficient methods for the immobilization of biomolecules onto CNTs surface [23, 24]. Chen et al. reported the noncovalent functionalization of CNTs with certain aromatic molecules through π–π stacking . A biofunctional molecule, 1-pyrenebutanoic acid, succinimidyl ester was found to strongly interact with the basal plane of graphite on the sidewall of SWNTs via π–π stacking. The anchored succinimidyl ester on the CNTs surface could be used to attach DNA or proteins through the formation of amide bonds . Some other biochemical active molecular with amine groups, such as streptavidin and ferritin, has been immobilized onto CNTs using above approach as well [25, 26]. Similarly, many biocompatible polymers can be wrapped or physically adsorbed onto the surface of CNTs by π–π stacking. O’Connell et al.  wrapped SWNTs with polystyrene sulfonate (PSS) and polyvinyl pyrrolidone (PVP) to render them reversible solubility in water. Furthermore, fluorescein-polyethylene glycol (Fluor-PEG) has been found able to be attached onto SWNTs through strong π–π interactions by Nakayama-Ratchford et al. The finite fluorescence intensity of fluorescein-PEG/SWNTs can be used in biosensor and biomedical imaging . Chitosan (CHI), as a biopolymer with good film-forming ability, has been widely used in the detection of various biological molecules through the formation of a special CHI–CNTs system . Using surfactants to wrap around the surface of CNTs is another strategy to noncovalently modify CNTs. Water-soluble surfactants, like sodium dodecyl sulfate (SDS) and cetyltrimethylammonium bromide (CTAB), can be applied to improve solubility and stability of CNTs in various suspensions .
2.2 CNT-Based Electrochemical Biosensors
Electrochemical biosensor is a two- or three-electrodes electrochemical cell, which can transfer a biological event to electrochemical signal. They often contain a biological recognition element on the electrode which reacts with the analyte and then produce electrochemical signal . CNT-based electrochemical biosensors play an important role in CNT-based biosensors because of their intrinsic advantages such as high sensitivity, fast response, easy operation, and favorable portability. Based on the method of the recognition process, CNT-based electrochemical biosensors can be divided into biocatalytic sensors and bioaffinity sensors. Biocatalytic sensors use the biological recognition element (e.g., enzyme) that can produce electroactive species, while bioaffinity sensors monitor a binding event between the biological recognition element and the analyte [32, 33]. CNT-based enzymatic electrochemical biosensors and CNT-based bioaffinity electrochemical sensors will be reviewed in details.
2.2.1 CNT-Based Enzymatic Electrochemical Biosensors
Enzymatic biosensors that combine electrochemical technology with specificity of enzyme have provided great opportunities for strategies in the early diagnosis . The direct electron transfer between the redox-active center of enzyme and the electrode without mediators is critical to the development of enzymatic biosensor. However, because the active centers of enzymes are surrounded by a thick protein layer and located deeply in hydrophobic cavity of molecules, the direct electrochemistry of enzyme is very difficult [35, 36]. Therefore, the use of an electrical connector is required to enhance the transportation of electrons. CNTs, with their small size, extraordinary electrochemical properties, and high specific surface area, have been widely used to promote electron transfer between the electrode and the redox center of enzyme . During the past few years, there have been many reports of CNT-based enzymatic biosensor for the detection of clinically important analytes through the electrochemical reactions catalyzed by various enzymes , such as glucose oxidase (GOx) , horse radish peroxidase (HRP) , lactate oxide , malate dehydrogenase (MDH) , and so on. One of the major challenges for the design of CNT-based enzymatic biosensor is how to achieve stable attachment of enzyme while still retaining their bioactivity. According to the different architectures, there are four main types of CNT-derived enzyme electrodes as discussed in the following.
184.108.40.206 CNT Paste Enzyme Electrodes
The first application of CNTs as electrode was reported by Britto et al. . A carbon nanotube paste electrode (CNPE) was constructed by using bromoform as binder to mix with carbon nanotubes, and better performance of electrochemical oxidation toward dopamine was observed on CNPE than other carbon electrodes . In a similar manner, CNTs have been mixed with mineral oil for glucose detection by adding GOx into the composite material. A detection limit of 0.6 mM was obtained with the CNPE containing 10 wt% GOx .
220.127.116.11 CNT-Modified Electrodes with Immobilized Enzymes
In most cases, CNT-based enzymatic biosensors were fabricated by modifying electrodes with CNTs and enzymes via different approaches . Similar to the functionalization of CNTs, methods for linking enzyme onto CNTs include noncovalent and covalent interaction. Noncovalent approach can preserve the structural integrity and properties of enzyme as well as provide high surface loading of enzyme . However, the interaction between enzyme and CNTs is not strong; thus, the immobilized enzyme may be gradually lost during the use. This limitation can be overcome by adsorbing enzymes onto polymer or nanoparticles-modified CNTs. Cai and Chen  dispersed CNTs in the solution of CTAB and then mixed with graphene oxide (GO). Nafion was used as a binder to hold the GOx/CNTs mixture on the electrode. The promotion effect of CNTs on the direct transfer of glucose oxidase which immobilized on CNTs was observed. In our previous work, bamboo-shaped carbon nanotube/chitosan film has been used for the immobilization of horseradish peroxidase (HRP) and related bioelectrochemical studies. The results indicated that immobilized HRP in the film shows excellent bioelectrocatalytic activity toward H2O2 . As a further example, a biosensor for glucose detection has been obtained by the deposition of Pt nanoparticles onto Nafion-containing GOx/CNTs film. The designed glucose biosensor achieved a fast response time of 3 s and a low detection limit of 0.5 µM .
Vertically aligned CNTs are another type of architecture for electrode modification. Vertically aligned CNTs coupled with enzyme on their tips facilitate rapid electron transfer compared to randomly distributed CNTs. It is because the CNTs tips have more activity sites than the sidewalls and the electrons directly transfer along the vertical direction of the tube . Patolsky et al. reported a structural alignment of GOx onto the edge of CNTs that are linked to a gold electrode surface. Flavin adenine dinucleotide (FAD) was first covalently attached onto the edge of CNTs, and then GOx was electrically linked onto the immobilized FAD. The CNTs were used as electrical connectors between the enzyme redox and the electrode. The electrons are transported along distances greater than 150 nm, and the rate of electron transport is controlled by the length of the CNTs .
Compared with noncovalent enzyme adsorption, covalent conjugation provides durable attachment to prevent enzyme leakage. Ruhal et al. designed an amperometric malic acid biosensor by covalently immobilizing malate dehydrogenase (MDH) on MWNT-coated screen-printed carbon electrode using standard water-soluble coupling agents 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxy-sulfo-succinimide (sulfo-NHS). The detection limit of malic acid was 60–120 µM, and the response time was 60 s . In the work reported by Yu et al., vertically aligned SWNTs were initially assembled on ordinary pyrolytic graphite electrodes. Iron heme proteins horse heart myoglobin (Mb) and HRP were covalently attached onto the ends of the SWNTs via amide linkages, respectively. The detection limits toward H2O2 were found to be 70 nM for SWNT/Mb and 50 nM for SWNT/HRP. The authors suggested that vertically aligned SWNTs behaved electrically similar to a metal, conducting electrons from the external circuit to the redox sites of the enzymes .
18.104.22.168 CNT Forest Electrodes with Immobilized Enzymes
CNT forest electrodes refer to the use of vertically aligned carbon nanotube arrays as a sole conductive component instead of modifying it onto another electrode surface. In this case, a CNT array is grown directly on a substrate surface. Besides the general advantages of vertically aligned CNTs as mentioned above, the structure and morphology control of the tubes during the synthesis step provides more possibilities for diversifying the electrode design.
2.2.2 CNT-Based Bioaffinity Electrochemical Biosensors
Bioaffinity sensors, such as DNA biosensors and immunosensors, are based on the recognition and specific binding which happens between two biomolecules. One of the two biomolecules is initially bonded onto the transducer and will be used to capture the target analyte during the detection. Bioaffinity electrochemical sensors collect the measurable electrochemical signal produced by the molecular recognition. CNT-based DNA electrochemical sensors and CNT-based immunosensors will be discussed in the following .
22.214.171.124 CNT-Based DNA Electrochemical Sensors
DNA biosensor, based on DNA–DNA hybridization, is of considerable recent interest due to its simplicity, speed, and economical assay for the diagnosis of genetic and infectious diseases and for the detection of genome mutation . When it comes to electrochemical biosensing, a single-stranded DNA (ssDNA) is attached onto an electrode for sensing complementary DNA. An electronic single is directly given by electrochemical reactions caused by the DNA hybridization. However, it is difficult to collect sensitive electrochemical signals for the DNA electrochemical sensor-based electrochemical oxidation of nucleobases (primarily purine) . There are two main reasons: (1) The electrochemical oxidation of purine occurs at high potentials and is characterized by a low electron transfer rate; (2) the peak current is too small to be investigated on classic electrode unless using mercury-based electrode. In order to solve these problems, electroactive indicators such as a cationic metal complex or intercalating organic compound have been used to improve the electrochemical response in DNA electrochemical biosensor. Some other indicator-free designs involve the attachment of the redox group onto the target DNA . With the development of nanomaterials, many researches have demonstrated that the performance of this type of biosensor can benefit from the use of CNTs [58, 59].
DNA oligonucleotides can be immobilized onto the CNT-based electrode through physical absorption . However, covalent attachment plays more important role in CNT-based DNA electrochemical sensors. Cai et al. first reported the use of CNTs to fabricate an electrochemical DNA biosensor for the specific DNA detection. An oligonucleotide probe with amino group at its 5′-phosphate end (NH2-ssDNA) was covalently bonded onto the CNTs–COOH-modified glassy carbon electrode (GCE) surface. CNT-modified electrode allows fast electron transfer between electrode and the redox intercalator daunomycin. The DPV measurements were taken from 0.00 to +0.60 V (vs. SCE), and the detection sensitivity achieved 1.0 × 10−10 mol L−1 for complementary oligonucleotide . In another similar protocol, ssDNA was covalently immobilized to the CNTs-COOH-modified electrode surface and Mn(II) complexes were used as DNA intercalator. ssDNA fragment could be selectively detected with a detection limit of 1.4 × 10−10 mol L−1 . As we mentioned before, aligned CNTs arrays exhibit quick electron transfer speed, and the use of these nice CNT structures offers promising prospect in fabricating sensitive DNA electrochemical sensors. He et al. demonstrated an effective method to prepare sensitive aligned CNT-based DNA electrochemical sensor. In their protocol, specific DNA sequences were covalently coupled on the tips and sidewalls of plasma-activated aligned CNTs for sensing complementary DNA and/or target DNA chains of specific sequences. The CV results showed that the sensitivity of the DNA electrochemical sensors was 11.36 ng mL−1. They concluded that aligned CNTs have implications for advancing the device-level applications of CNT-DNA chips . An ultrasensitive DNA electrochemical sensor based on vertically aligned CNTs embedded in SiO2 was reported by Jun et al. Primary amine-terminated oligonucleotides were coupled with terminal –COOH groups on the ends of the CNTs arrays with the assistant of EDC and sulfo-NHS. Ru(bpy) 3 2+ mediators were employed to amplify signal for the detection of target DNA. From CV and AC voltammetry (ACV) data, a detection limit lower than a few attomoles of oligonucleotide targets was found .
As an important technology in electrochemistry, impedance spectra also have been utilized to observe DNA hybridization without using hybridization marker or intercalator. Xu et al. presented a composite material of polypyrrole (PPy)- and MWNT-based label-free DNA electrochemical sensor by using impedance spectra as detection single. The composite film was electropolymerized onto the electrode in the presence of MWNTs-COOH. Similar to the protocol as mentioned above, ssDNA was covalently coupled with PPy/MWNTs-COOH-modified electrode. A decrease in impedance was observed after the DNA hybridization reaction. It is because that electron transfer resistance of double-stranded DNA is lower than that of ssDNA. In this work, a detection limit of 5 × 10−12 mol L−1 was achieved for the detection of complementary DNA sequence . Another similar work based on SWNTs was reported by Weber et al., and instead of using conductive polymers, they modified electrode with a mixture of dimethylformamide (DMF) and SWNTs-COOH. This impedance DNA sensor was found to be able to sense complementary target DNA concentration at 1 × 10−9 mol L−1 .
126.96.36.199 CNT-Based Electrochemical Immunosensors
For immunosensors using amperometric method, most of them need enzyme or other electroactive labels to provide electrochemical singles. However, the antibody–antigen interaction can be directly detected by impedance spectroscopy without any labels. Hafaiedh et al. reported an electrochemical impedance immunosensors for sensing IgG. The interaction of goat anti-rabbit IgG with different concentration IgG on MWNT-coated electrode was monitored by impedance spectroscopy. The detection limit was found to be 10 pg mL−1 .
2.3 CNT-Based Field-Effect Transistor (FET) Biosensors
In order to achieve biocompatible interaction between CNTs and living cells, Sudibya et al. presented a strategy to functionalize SWNTs with bioactive sugar moieties for the detection of dynamic biomolecular release from these cells. N-acetyl-d-glucosamine (GlcNAc), d-glucose (Glc), or d-mannose (Man) was anchored onto the nanotube by either a pyrene or a lipid tail. The direct adhesion and growth of PC-12 cells on these glycosylated CNTs networks were examined. They found that GlcNAc-functionalized SWNT nets had better performance than others. Therefore, a GlcNAc-functionalized SWNT-net-based FET biosensor was proposed for the real-time detection of regulated secretion (or exocytosis) of PC12 cells. The influx of Ca2+ ion solution in Ca2+ ion channels opened by membrane depolarization triggered the actions of fusion of vesicles. Upon the fusion of vesicles, catecholamine molecules inside the vesicle were released onto the narrow interface between the cells and the SWNTs net and then interacted with them by π–π stacking. The conductance of nanotube was highly sensitive to the electrochemical perturbations at the surface induced by these interacting molecules. So, the triggered catecholamine molecules released from PC12 cells can be continuously monitored through the changes in current flowed on the surface of nanotubes .
2.4 CNT-Based Optical Biosensors
The unique optical properties of CNTs have aroused widespread concerns in development of biosensors during the past few years. Semiconducting SWNTs can act as quenchers for the fluorophores and can display distinctive near-infrared (NIR, wavelength ∼0.8–2 µm) photoluminescence arising from the band-gap fluorescence . Optical biosensors based on these properties have been reported by many research groups. Yang et al. reported a self-assembled quenched complex of fluorescent ssDNA and SWNTs as an efficient molecular beacon (MB) to fluorescently detect single nucleotide variations in DNA. In this design, one end of the ssDNA was labeled with a fluorophore and then assembled onto the surface of SWNTs through π-stacking interactions. Here, the SWNTs act as both nanoscaffold and nanoquencher. If the target DNA is not present in the sample, the fluorescently labeled ssDNA-SWNT complexes completely quench the fluorescence. In the presence of the target DNA, the competitive binding of the target and the SWNTs with the ssDNA suppresses the fluorescence quenching, and hence a fluorescence signal was observed. This approach can be extended to design a variety of fluorescent biological probes with detection limits in the nanomolar range .
SWNTs also can be utilized as NIR fluorescent tags for selective probing and imaging cells. In the work reported by Welsher et al., polyethyleneglycol (PEG)-modified SWNTs are conjugated to Rituxan antibodies to selectively recognize CD20 cell surface receptor on B cells with little nonspecific binding to negative T cells and Herceptin antibodies to recognize HER2/neu-positive breast cancer cells. The selective SWNT antibody binding to cells was imaged by detecting intrinsic NIR photoluminescence of the nanotubes .
Electrochemiluminescent (ECL) is a luminescence process that is produced by electrochemical reactions in solution. A CNTs microwell array-based ECL biosensor for detection of cancer biomarkers was developed by Sardesai et al. SWNTs forests were coated onto each of microwells on a pyrolytic graphite (PG) chip. Cancer biomarker prostate-specific antigen capture antibodies (PSA-Ab1) and interleukin-6 capture antibodies (IL-6-Ab1) were covalently coupled onto the SWNTs forest for capturing protein analytes. Silica nanoparticles containing [Ru(bpy)3]2+ and secondary antibodies (RuBPY-silica-Ab2) were used as the signal-amplified element. [Ru(bpy)3]2+ labels produce ECL in a multistep electrocatalytic redox reactions with a suitable sacrificial reductant such as tripropylamine (TprA) to yield photoexcited [Ru(bpy) 3 2+ ]* that emits light at 610 nm. ECL light intensity was integrated by the CCD camera. The detection limit for PSA was 1 pg mL−1 and for IL-6 was 0.25 pg mL−1 in serum. These SWNT forest arrays can be used to interfacing with microfluidic for simultaneous detection of different types of proteins .
3 Graphene-Based Biosensors
Graphene, a 2D carbon material with one-atom thickness, has become one of the hottest research topics in the field of biosensors. Similar to CNTs, sp 2-bonded carbon atoms in graphene are closely packed in a honeycomb lattice structure. Owing to their unusual structure, graphene and its derivatives exhibit several extraordinary properties including high thermal conductivity, tunable optical property, high planar surface, superior elasticity, and mechanical strength . In addition, many research results have revealed that graphene and its derivatives possess remarkable electronic properties, such as a high quantum Hall effect at room temperature , an ambipolar electric field effect along with ballistic conduction of charge carriers , electron–hole symmetry, and internal degrees of freedom . These notable properties make graphene an attractive candidate for the development of the new generation of biosensors with outstanding performance [94, 95]. Currently, there are several physical and chemical methods for producing graphene and graphene-related materials, such as mechanical exfoliation of bulk graphite, chemical vapor deposition (CVD) of hydrocarbons on metal substrates, and chemical or thermal exfoliation of graphite oxide to graphene oxide sheet . Among them, chemical or thermal exfoliation has attracted much attention because of easy operation and high yields of bulk-quantity graphene and graphene oxide. Based on different reduction approaches, thermally reduced graphene oxide (TRGO), chemically reduced graphene oxide (CRGO), and electrochemically reduced graphene oxide (ERGO) can be produced . Although made with different methods, all graphene products made by these methods contain large amounts of defects and oxygen-bearing groups. The presence of these defects can be beneficial to the further functionalization of chemical or biomolecules .
To employ CNTs or graphene in biosensors, chemical functionalization has to be made to enable a biocompatible surface, which is able to be further conjugated with other molecules. The sensing performance is highly dependent on the functionalization process. As we mentioned, CNTs can be chemically functionalized by both covalent and noncovalent methods. Chemical functionalization is also of great importance to graphene to render high water solubility and biocompatibility; for example, diazonium chemistry has been used to graft various functional groups onto graphene surface , and many aromatic molecules can be physically absorbed onto graphene surface through π–π stacking . After the pre-functionalization, graphene is much easier to be modified by other chemical and biological compounds for further sensing applications [99, 100]. Graphene-based electrochemical biosensors, field-effect transistors, and optical biosensors have greatly stimulated research interests in the past few years and will be reviewed in the following part.
3.1 Graphene-Based Electrochemical Biosensors
It has been reported that heterogeneous electron transfer of sp 2 carbons occurs at the edges and defects not at the basal plan of graphene sheets [101, 102]. Oxygen containing groups on GO surface and the super high surface area provided by its 2D structure can enhance the electron transfer rate and are advantageous for improving the sensitivity of electrochemical biosensors [96, 103]. Graphene-based electrochemical biosensors thus are of great interest to researchers in recent years.
3.1.1 Graphene-Based Enzymatic Electrochemical Biosensors
As we discussed in the above, CNTs have exhibited excellent performance in enhancing the direct electron transfer between various enzymes and electrodes. The ongoing research shows that graphene-related materials can also promote the electron transfer between enzymes and electrode due to their extraordinary electron transport property and high surface area [94, 104]. The common approaches for immobilization of enzyme onto graphene materials are similar to that onto CNTs surface. Enzymes can be noncovalently or covalently attached onto the electrode modified with graphene materials. In the work reported by Shan et al., GO was exfoliated from graphite. PVP-protected GO was then prepared and found to have good electrochemical reduction toward H2O2 and O2. A carboxyl-terminated ionic liquid polyethylenimine-functionalized ionic liquid (PFIL) was applied to be covalently attached onto PVP-protected GO for the immobilization of glucose oxidase. Counter anions were allowed to be exchanged with negatively charged GOx. The amperometric response of graphene-GOx-PFIL biosensor is linear against the concentrations of glucose ranging from 2 to 14 mM . Liu et al. reported a glucose biosensor through the covalent attachment between carboxyl acid groups on GO sheets and the amine groups of GOx. The amperometric responses at the covalently linked GOx-GO enzyme electrode were examined, and the sensitivity was found to be 8.045 mA cm−2 M−1 . It has been reported that doping with heteroatoms such as nitrogen is an excellent method to provide pathways for efficient electron transfer processes. Prathish et al. have developed and compared various modified electrodes with basic or acidic functionalized graphene and nitrogen-doped graphene (NG) as well as with composite materials of NG, conducting polymer poly(3,4-ethylenedioxythiophene) (PEDOT) and the redox polymer poly(PNR). Among these electrodes, the N-graphene-modified GCE has shown higher sensitivities than the other electrodes for the detection of cofactors present in oxidase and dehydrogenase-based enzymes: nicotinamide adenine dinucleotide (NADH) and flavin adenine dinucleotide (FAD), due to its n-type semiconductor like properties . The layer-by-layer technique has been used by Barsan et al. for the construction of a glucose biosensor based on multilayer films containing chitosan biopolymer, poly(styrene sulfonate) (PSS), GOx, and nitrogen-doped graphene. GOx and NG were dispersed in the biocompatible positively charged polymer chitosan (chi + (NG + GOx)) and coated onto a gold electrode surface. Then the negatively charged PSS, PSS−, was assembled onto the electrode. This process was repeated for further bilayers. They found that the presence of N-doped graphene decreases the charge transfer resistance of the assembly and provides a film matrix that presents significant charge separation. The biosensor operates at a low potential of −0.2 V versus Ag/AgCl, exhibiting a high sensitivity of 10.5 μA (cm−2 mM−1) and a detection limit of 64 μM . Zhang et al. compared the loading amount of enzymes on GO and chemically reduced graphene oxide (CRGO). They found that the enzyme can be adsorbed onto CRGO directly with a tenfold higher loading than that on GO. The immobilizations of HRP and oxalate oxidase (O x O x ) on CRGO were reported. The maximum enzyme loadings were found to reach 1.3 and 12 mg mg−1 for HRP and O x O x , respectively. The results indicate that CRGO is a potential promising substrate for efficient enzyme immobilization .
Besides the graphene-based enzymatic biosensors, the research on graphene-based nonenzymatic biosensors has been carried out worldwide. Wang et al. reported a nonenzymatic glucose biosensor based on Pd nanoparticle-modified graphene oxide (Pd NPs/GO). A linear range of 0.2–10 mM with high selectively was achieved . A novel structure of Au nanoparticle-decorated three-dimensional (3D) porous reduced graphene oxide-single-walled carbon nanotube (rGO-SWCNT-Au) nanocomposites was proposed by Luo et al. for sensing glucose. Their method could achieve the simultaneous reduction of GO and HAuCl4. The prepared rGO-SWCNT-Au nanocomposites were utilized to monitor glucose and revealed a wide linear range and a low detection limit . Bimetallic nanoparticles also have been used for enzymeless biosensors, and Luo et al. developed a PdCu nanoparticle-decorated 3D graphene hydrogel-based biosensor for the detection of glucose. The synthesized nanohybrids showed a high sensitivity of 48 μA (mg mM)−1 with a wide linear range up to 18 nM toward glucose oxidation .
3.1.2 Graphene-Based Bioaffinity Electrochemical Biosensors
188.8.131.52 Graphene-Based DNA Electrochemical Sensors
Graphene’s large 2D aromatic surface is an ideal substrate for adsorption of biomolecules as well as a burgeoning support to disperse and stabilize metal, metal oxide, and semiconductor nanomaterials . Similar to CNT-based DNA biosensors, in order to collect sensitive current signal, electroactive indicators often used to provide and improve the electrochemical response. Zhu et al. reported a DNA electrochemical biosensor based on thionine–graphene nanocomposite. Thionine–graphene was covalently modified onto a gold electrode surface with the use of l-cysteine (Cys) and cross-linker glutaraldehyde (GA). Then, the probe ssDNA was covalently grafted onto the electrode with the presence of GA. The hybridization procedure was monitored by DVP analysis using an electroactive intercalator daunomycin on the electrode. The complementary oligonucleotide could be quantified in a wide range of 1.0 × 10−12 to 1.0 × 10−7 M with a good linearity and a low detection limit of 1.26 × 10−13 M . Using graphene as signal amplification element is also a very common strategy in graphene-based DNA biosensors. Wang et al. presented a DNA biosensor based on using graphene-supported ferric porphyrin as a peroxidase mimic. A biotinylated molecular beacon (MB) was coated on a gold nanoparticles single-walled carbon nanohorn (AuNPs–SWCNH) composite-modified glassy carbon electrode. In the presence of target DNA, the MB probe “opened” and the biotin group moved away from the electrode surface. Then, a porphyrin-based HRP mimic as a trace label was introduced through the specific interaction between streptavidin on the trace label and biotin group on the electrode. In their work, an iron porphyrins iron (III) meso-tetrakis (N-methylpyridinium-4-yl) porphyrin (FeTMPyP) was used to mimic the biological function of HRP. Due to the 2D structure of graphene, FeTMPyP could be attached on the both sides of graphene and led to a high loading amount. They found that this method showed a detection limit down to attomolar levels, much lower than that obtained with a HRP-based trace label .
Graphene-based DNA impedimetric sensor has also attracted a significant amount of attention due to their advantage of being indicator free. Bonanni et al. have demonstrated a label-free electrochemical impedance DNA biosensor for the detection of DNA hybridization and polymorphism. They also compared the performance of sensors with different graphene layers. Different graphene platforms were modified with hairpin-DNA (hpDNA) probes for the sensitive detection of a single nucleotide polymorphism (SNP) correlated with the development of Alzheimer’s disease. The negatively charged hpDNA probes were immobilized on the graphene and caused an increase in electron transfer resistance. After hybridization with wide type target, the electron transfer resistance decreased due to the partial release of the hpDNA probes from the electrode surface. The less resistance decrease was observed for the hybridization with mutant, and no decrease was observed with the noncomplementary sequence. They also found that sensors with four layers of graphene provided the best sensitivity and were employed for the detection of SNPs. A higher sensitivity was found for as-prepared impedance DNA biosensor compared to that obtained with similar platform using fluorescence methods . In the work reported by Zhang et al., DNA probe was grafted onto the tryptamine (TRA)-functionalized reduced graphene oxide (TRA–RGO) through the cross-linker GA. DNA hybridization reaction is monitored by EIS. The as-prepared electrochemical impedance DNA biosensor shows excellent reproducibility and a detection limit of 5.2 × 10−13 M .
184.108.40.206 Graphene-Based Electrochemical Immunosensors
As one of the most promising nanomaterials, graphene has been widely used in electrochemical immunosensors. An electrochemical biosensing platform based on graphene for analysis of vascular endothelial growth factor receptor 2 (VEGFR 2) was proposed by Wei et al. In their work, a sandwich-type assay format based on thionine-functionalized chitosan-reduced graphene oxide (CHI-RGO) electrode was developed to detect VEGFR 2 in cell lysates directly. Capture antibodies were grafted onto RGO through the cross-linker GA for the adsorption of VEGFR 2, followed by the immobilization of biotinylated detection antibody. A bioconjugate of antibody and streptavidin-HRP was then used as enzyme labels to provide current signals. The detection limit for VEGFR 2 was found to be 0.28 pM . Liu et al. presented a graphene-assisted dual amplification strategy for the fabrication of sensitive amperometric immunosensor for detection of human IgG (HIgG). In their protocol, a composite of poly(diallyldimethylammonium chloride)-functionalized graphene nanosheets (PDDA-G) and gold nanoparticles (AuNPs/PDDA-G) was firstly coated on the electrode, followed by the immobilization of probe antibodies. The target HIgG antigen was then assembled onto the electrode through antibody–antigen interactions. In order to achieve a high sensitivity for the detection of HIgG, the authors applied HRP–detection antibody/AuNPs/PDDA–EGO bioconjugation to fabricate a sandwich-type assay format. The proposed methods showed excellent analytical performance for the detection of HIgG ranging from 0.1 to 200 ng mL−1 with a detection limit of 0.05 ng mL−1 .
The label-free graphene-based impedimetric immunosensors have been developed by many researchers. An electrochemical impedance-based biosensor for determination of bovine interleukin-4 (bov-IL-4) was proposed by Chen et al. Reduced graphene oxide (RGO)/chitosan composites were modified on glassy carbon electrode for the attachment of monoclonal bov-IL-4 antibody. The resultant electrochemical impedance is linearly proportional to its logarithmic concentration in the range from 0.1 to 50 ng mL−1, and the detection limit is 80 pg mL−1 . Mishra et al. reported a novel electrochemical impedimetric immunosensors for the quantitative detection of human cardiac myoglobin (Ag-cMb). In their protocol, 3-mercaptopropionic acid (MPA)-capped ZnS nanocrystals (ZnS(MPA)) were anchored on RGO sheets through a cross-linker and deposited onto silane-modified indium-tin-oxide (ITO) glass plate. ZnS nanocrystals allow greater affinity with the target biomolecules and provide a favorable environment to retain the activity of the antibodies. The protein antibody, Ab-cMb, was then covalently coupled onto ZnS–RGO nanocomposite to capture antigen Ag-cMb. They found that the ZnO–RGO hybrid-modified electrode exhibited a higher sensitivity of about 2.5-fold higher than that of the bare RGO-modified electrode. The proposed immunosensors exhibited a linear electrochemical impedance response to Ag-cMb from 10 ng mL−1 to 1 mg mL−1 with a sensitivity of 177.56 Ω cm2 per decade .
3.2 Graphene-Based Field-Effect Transistor (FET) Biosensors
3.3 Graphene-Based Optical Biosensors
Graphene also can be used as SERS substrate to enhance Raman signals of absorbed biomolecules. He et al. reported a graphene-based SERS platform for multiplex DNA detection. Dye-labeled DNAs were assembled on gold-decorated graphene-SiO2/Si substrate. Due to the double effect of gold nanoparticles and graphene, the Raman signals of dye were dramatically enhanced. This platform exhibited extraordinarily high sensitivity and a detection limit of 10 pM for DNA. Moreover, it is possible to detect simultaneously more types of target DNAs on the proposed substrate with single-laser excitation .
Graphene has been also utilized to fabricate electrogenerated chemiluminescence (ECL) immunosensor. Xu et al. reported a graphene-based sandwich-type ECL biosensor for the detection of cancer marker prostate-specific antigen (PSA). The cathodic ECL behavior of luminol at a positive potential with a strong light emission on a graphene-modified electrode was presented in this work. A signal amplification strategy from enzyme–antibody-conjugated gold nanorods was developed. The graphene–chitosan-modified glassy carbon electrode was prepared for the immobilization of PAS capture antibodies to immobilize PSA. The gold nanorods were used to carry secondary antibody (Ab2) and GOx, which further amplified the ECL signal of luminol in the presence of glucose and oxygen. A linear relationship between ECL signals and the concentrations of PSA was obtained in the range from 10 pg mL−1 to 8 ng mL−1. The detection limit of PSA was 8 pg mL−1 .
4 Conclusions and Perspectives
Compared to CNTs, graphene has a much younger history. With the development of graphene in biological and biomedical areas in recent years, a natural question came up is that which one is better for making biosensors. As two popular carbon nanomaterials, CNTs and graphene share some properties in common, including excellent mechanic, electronic, and thermal properties. Graphene can be massively produced at a low cost and does not have metallic impurities which often present in CNTs . Graphene is much more amenable in large-scale architectures than high aspect ratio, one-dimensional CNTs. Due to the defect-rich property, graphene can be easily functionalized chemically to generate functional groups on the 2D plane. These advantages provide graphene more opportunities in biomedical imaging and therapy applications and offer more sensitive response as well . In addition to the 2D nature, graphene has high electron mobility, making it a promising material in electronic biosensors. Graphene shows light absorption from UV to NIR regions and efficient fluorescence quenching properties. However, graphene does not have different chiral forms; thus, some optical properties caused by different chiral forms are unique for CNTs . In applications in selective probing and imaging of living cells, graphene may need external labels, while CNTs do not need these.
Currently, it is a worldwide interest in employing carbon nanomaterials like CNTs and graphene in biosensors. It has been demonstrated by researchers that CNTs and graphene can offer new sensing protocols and also improve the detection limit over the existing protocols. The sensors using CNTs and/or graphene are faster in response, cheaper in cost, and smaller in size. However, great challenges remained in biosensors with CNTs and graphene. First, both CNTs and graphene need surface functionalization. The surface properties determine the subsequent performance in sensors. The established chemical approaches for surface functionalization are still far from being applicable and highly reproducible. The reliable and simple methods have to be developed for functionalizing CNTs and graphene in future. Second, biosensors are a class of devices that are able to translate the biological information to other information, like electric and optical signals. The integration of the sensing bed with the signal translator and receiver is another factor determining the efficiency of the sensor. The translation must be fast, reliable, and readable. Third, detecting more than one target molecules is still a great challenge in biosensor development. The medical devices usually require multifunctionality. Fourth, the sensors with strong anti-interference ability are highly desired. The real bioenvironment is much more complicated than the laboratory-built environment. Various biomolecules with similar properties often interfere with the detection of target molecules. To overcome this challenge, the effort must be made on developing highly specific targeting agents and these agents must be robust in various bioenvironments. To solve this problem and to develop CNT and graphene-based biosensors toward a practical application, efforts have to be made in not only materials science and technology, but also electric engineering and medical science.
- 28.N. Nakayama-Ratchford, S. Bangsaruntip, X. Sun, K. Welsher, H. Dai, Noncovalent functionalization of carbon nanotubes by fluorescein—polyethylene glycol: supramolecular conjugates with pH-dependent absorbance and fluorescence. J. Am. Chem. Soc. 129(9), 2448–2449 (2007). doi: 10.1021/ja068684j CrossRefGoogle Scholar
- 54.A.T. Ezhil Vilian, R. Madhu, S.-M. Chen, V. Veeramani, M. Sivakumar, Y.S. Huh, Y.-K. Han, Facile synthesis of MnO2/carbon nanotubes decorated with a nanocomposite of Pt nanoparticles as a new platform for the electrochemical detection of catechin in red wine and green tea samples. J. Mater. Chem. B 3(30), 6285–6292 (2015). doi: 10.1039/C5TB00508F CrossRefGoogle Scholar
- 104.Q. Zhang, Y. Qiao, F. Hao, L. Zhang, S. Wu, Y. Li, J. Li, X.-M. Song, Fabrication of a biocompatible and conductive platform based on a single-stranded DNA/graphene nanocomposite for direct electrochemistry and electrocatalysis. Chem. Eur. J. 16(27), 8133–8139 (2010). doi: 10.1002/chem.201000684 CrossRefGoogle Scholar
- 111.Y. Luo, F.-Y. Kong, C. Li, J.-J. Shi, W.-X. Lv, W. Wang, One-pot preparation of reduced graphene oxide-carbon nanotube decorated with Au nanoparticles based on protein for non-enzymatic electrochemical sensing of glucose. Sens. Actuator B 234, 625–632 (2016). doi: 10.1016/j.snb.2016.05.046 CrossRefGoogle Scholar
- 123.S.K. Mishra, A.K. Srivastava, D. Kumar, A.M. Biradar, Microstructural and electrochemical impedance characterization of bio-functionalized ultrafine ZnS nanocrystals-reduced graphene oxide hybrid for immunosensor applications. Nanoscale 5(21), 10494–10503 (2013). doi: 10.1039/c3nr02575f CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.