Review article: Graphene in biosensing

Biosensing is paramount for improving the quality of human life. Biosensors and biosensing protocols are able to detect a wide range of compounds, sensitively and selectively, with applications in security, health care for point-of-care analyses of diseases, and environmental safety. Here, we describe biosensors…

Biosensing is paramount for improving the quality of human life. Biosensors and biosensing protocols are able to detect a wide range of compounds, sensitively and selectively, with applications in security, health care for point-of-care analyses of diseases, and environmental safety. Here, we describe biosensors and biosensing systems employing graphene. Graphene is a zero-gap semiconductor material, which is electroactive and transparent. Because of its interesting properties, graphene has found its way into a wide variety of biosensing schemes. It has been used as a transducer in bio-field-effect transistors, electrochemical biosensors, impedance biosensors, electrochemiluminescence, and fluorescence biosensors, as well as biomolecular labels. In our review, we describe the application of graphene for enzymatic biosensing, DNA sensing, and immunosensing. We compare different techniques and present our views on the future development of the field.

Martin Pumera
Division of Chemistry & Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore 637371
E-mail: pumera@ntu.edu.sg

Graphene is a one-atom-thick material consisting of sp2-bonded carbon with a honeycomb structure. It resembles a large polyaromatic molecule of semi-infinite size1, 2. In the past five years, graphene-based nanomaterials have been the focus of a vast amount of attention. The interesting and exciting properties of single-layer graphene sheets, such as high mechanical strength3, high elasticity and thermal conductivity, demonstration of the room temperature quantum Hall effect, very high roomtemperature electron mobility2, tunable optical properties4,5, and a tunable band gap 6 have excited the scientific community especially in the areas of materials, physics, and chemistry. Different, but similarly fascinating properties are exhibited by double-, few-, and multilayer graphene. Because graphene is a conductive yet transparent material, with a low cost and low environmental impact, it is an ideal material for the construction of sensors and biosensor-based devices in various transduction modes, from  electrical and electrochemical transduction to optical transduction.

The detection of biologically active molecules is of critical importance from a biomedical, environmental, and security point of view. Such detection can be carried out by biosensor or by bioanalytical protocols. A chemical sensor is a device that quantitatively or semiquantitatively converts information about the presence of a chemical species to an analytically useful signal7. Sensors consist of two elements: a receptor and a transducer (see Fig. 1). A receptor can be any organic or inorganic material with (preferably) a specific interaction with one analyte or group of analytes. In the case of biosensors, the recognition element is a biomolecule. The second key element of the sensing platform is the transducer, which converts chemical information into a measurable signal. Bioanalytical protocols usually include more than one processing step. In this review, we will describe biosensors and bioanalytical systems that utilize graphene as a key component.

Fundamentals

There are different kinds of graphene-based nanomaterial and their type is closely related to the method of production. Graphene can be produced in many ways; by chemical vapor deposition (CVD) growth, mechanical exfoliation of graphite, or exfoliation of graphite oxide 8. Neither CVD-produced graphene nor mechanically  exfoliated graphene contain large quantities of defects or functionalities. However, bulk quantities of graphene- based nanomaterials are typically prepared by different methods, such as the thermal exfoliation of graphite oxide 9 which leads to a material called thermally reduced graphene (TRGO) or, for example, sono-assisted exfoliation of graphite oxide to graphene oxide (GO) 10. Graphene oxide can be further reduced chemically or electrochemically. Thermally reduced graphene oxide (TRGO) contains large amounts of defects and significantly differs from pristine graphene, which has a perfect honeycomb lattice structure 11. The presence of the defects is not disadvantageous. To the contrary, it is well known that heterogeneous electron transfer in the electrochemistry of sp2 carbons occurs at the edges and defects, and not at the basal plane of graphene sheets12. Graphene oxide has a structure that is not fully planar because the sp2 carbon network is heavily damaged. It contains large amounts of oxygen-containing groups, which can be beneficial to the functionalization through the action of the biomolecules for biorecognition events during biosensing 10. Graphene oxide can be chemically or electrochemically reduced. Such products have a partly restored sp2 lattice that also contains some degree of oxygen-bearing groups. The products are typically referred to as chemically reduced graphene oxide (CRGO) or electrochemically reduced graphene oxide (ERGO) 13. The structures of graphite oxide, graphene oxide, TRGO, and CRGO are shown in Fig 2. Ruoff et al. suggested that graphite oxide, graphene oxide, TRGO, ERGO, and CRGO should be termed chemically modified graphene s8.  Therefore, one could have a large graphene “toolbox” to choose the right type of graphene for the right application and transduction mechanism. When compared to carbon nanotubes (for reviews on carbon nanotube biosensors, see e.g.14-16), it is clear that the structural differences, such as tubes vs. sheets, will play a major role in the nanoarchitectonic design of biosensors. From a practical point of view, it is important to note that CNTs are grown from metallic particles which might affect the response of the sensor, while graphene is often synthesized in different ways. In the following text, we will discuss how graphene-based nanomaterials are used in different transduction systems.

 

Graphene in bio-field-effect transistors

Field-effect transistors (FET) have received a great deal of interest in the area of biosensing as they can provide full electronic detection that is fully integrated into the electronic chips produced by semiconductor companies. Therefore, it is not only academia that is fascinated by these devices, but there is a strong interest (and investment) from industry as well17. Field-effect transistor-based biosensors rely on biorecognition events between molecules at the gate of the FET18,19. Upon biorecognition between the probe and target biomolecules, the electric charge distribution changes the charge carrier density at the biorecognition layer and thus, the conductivity of the channel between the source and drain. Graphene is an ideal material for the construction of FET biosensors because it is a zero-band gap semiconductor, and the band gap can be tuned by surface modification20. As mentioned above,
FET transistors are ideal for sensing charged molecules. Therefore, graphene-based FETs can be employed for DNA sensing since DNA has a charged phosphate backbone21. Chen, Li, and co-workers 22 demonstrated that large-scale chemical vapor deposition (CVD) grown single- and few-layer graphene films are highly sensitive to DNA hybridization (Fig. 3). The shift in gate voltage was found to be sufficiently large for detection even at a concentration of 10 nM of single stranded DNA (ssDNA). The addition of gold nanoparticles to the probe surface led to an extension of the linearity of the response to 500 nM as this increased the amount of probe DNA immobilized on the FET surface. Tamanaha et al.23 used reduced graphene oxide modified with DNA for real-time detection of ssDNA with a detection limit of 10 nM.

It is apparent that the approach of modifying graphene sheets with nanoparticles to increase the number of probe biomolecules at the FET gate is beneficial in terms of the linearity of the response. Such an approach is not only used for sensing DNA but also for immunosensing. Thermally reduced graphene oxide sheets were suspended over the gold electrodes of an FET. TRGO sheets were then modified with gold nanoparticle (Au NP)-antibody ( anti- immunoglobulin G) conjugates (see Fig. 4) 24 . This functions as a specific recognition site for the detection of immunoglobulin G (IgG). When a biorecognition event occurs, significant changes occur in the electrical characteristics of the FET which are used as an analytical signal. Yang and Gong used a very simple system that did not employ an FET but only conductivity measurements25. Graphene was percolated with anti-prostate-specific
antigen and the detection of the prostate cancer marker was followed by measuring changes in the resistance of the graphene film. An aptamer based FET assay was also developed for the detection of IgE26.

Graphene impedimetric biosensors

Electrochemical impedance (EIS) platforms provide very high sensitivity for biosensing. We were one of the first groups to develop a graphene platform for the detection of DNA hybridization and polymorphism using electrochemical impedance spectroscopy as a detection technique 27. We compared the performance of three different graphene platforms, and showed how different numbers of graphene sheets can affect detection (Fig. 5). We found that few-layer graphene provided the best sensitivity and we employed this platform for the detection of a single nucleotide polymorphism. A higher sensitivity was obtained with impedimetric detection compared to that obtained with a similar platform using fluorescence methods. We believe that the graphenebased strategy presented here could be used in the development of an analytical device for point-of-care diagnostic tests and for very sensitive detection of SNPs correlated with various diseases. A number of schemes for the immobilization of single-stranded DNA on graphene surfaces can be used, e.g., a chemical bond between the carboxylic group of graphene sheets and NH2-modified ssDNA 28. Upon hybridization with target DNA, the conformation of ssDNA changes from a “lying” structure to a “standing” double helix. This change of DNA conformation and distribution of charges at the surface of the electrode ( note that the backbone of DNA is negatively charged due to the presence of phosphate groups) leads to changes in impedance of the electrode surface and to a measurable analytical signal. Hu et al.28 were able to determine low concentrations of the HIV-1 gene.

Graphene in electrochemical biosensors

Electrochemical detection is highly sensitive to electroactive molecules. In addition to sensitivity (which is also a property of electrical detection), it also offers detection selectivity as different molecules can be oxidized/reduced at different potentials. Graphene is an excellent conductor of electrical charge. Heterogeneous electron transfer (the
transfer of electrons between graphene and the molecule in the solution necessary for the oxidation/reduction of said molecule) occurs at the edges of the graphene or at defects in the basal plane.

Thus, the high surface area of graphene facilitates large amounts of defects and thus, electroactive sites29. Graphene has been employed in many schemes for sensing glucose30. This is reflected by the fact that electrochemistry is paramount to sensing glucose for diabetic patients (reflected by the multibillion USD glucosensing market) 31. The glucose oxidase enzyme is used as a biorecognition element: glucose oxidase oxidizes glucose to gluconic acid and shuffles electrons into the oxygen which is dissolved in the solution, and then reduced to hydrogen peroxide. Hydrogen peroxide is typically detected electrochemically. However, in several examples, direct electron transfer from the enzyme (without the need of O2 as an electron acceptor) is possible, making this an analytically valuable sign al31. Ultrathin multilayer graphene platelets (also called graphite nanoplatelets) have been used as a transducing material for the biosensing of glucose32. In another example, it was shown that N-doped graphene provides significantly enhanced oxidation currents for the enzymatic detection of glucose, compared to ordinary graphene materials (Fig. 6) 33. Direct electron transfer from glucose oxidase has been reported by several authors. Glucose oxidase enzyme was immobilized on a Nafion polymer film with graphene nanoplatelets 34. It was demonstrated that such simple non-covalent bonding enhances the redox current of a ferrocyanide solution and leads to a lowering of the overpotential of hydrogen perox ide34. In another example, direct electron transfer was observed in graphene/ionic liquid/glucose oxidase systems35. Other schemes for enzymatic detection of glucose have been reported using Au NPs/graphene/chitosan composites36. Another biosensor was developed for the detection of pesticides37 and hydrogen peroxide (using horseradish peroxidase38 or hemoglobin39).

The electrochemical detection of DNA has also attracted a significant amount of attention. It is possible to detect DNA recognition (hybridization) directly using the oxidative signals of DNA bases or by using electroactive labels40. Direct detection has an advantage because it is label free, but offers poorer sensitivity than label-based DNA assays. In addition, on traditional carbon materials, such as glassy carbon and graphite, the adenine (A) and guanine (G) bases give analytically useful signals while cytosine (C) and thymine (T) do not provide well-resolved signals. It was shown by Dong et al. 41 that chemically reduced graphene oxide provides well-resolved signals of all four A,G,C,T bases (Fig. 7, left panel) with higher sensitivity than with graphite. This is attributed to the high defect density of the CRGO and thus superior electrochemical performance when compared to graphite. This feature was used for the label-free detection of single mutation polymorphism (Fig. 7, right panel). In a similar manner, Loh et al.42 compared the electrochemistry of DNA bases on epitaxially grown graphene (meaning the graphene was grown on a substrate where the basal plane is exposed to the solution; it resembles basal plane pyrolytic graphite with the difference being that the number of graphene layers is smaller). Electrochemical oxidation of pristine epitaxially grown graphene led to the creation of defects on its surface (this has previously been shown with the walls of carbon nanotubes43) and leads to a significantly higher response. It was suggested that electrochemically oxidized graphene or graphene with large amounts of defects could be used for highly sensitive electrochemical sensing. This is consistent with a previous publica tion41. Loh et al. applied electrochemically oxidized graphene to discriminate between ssDNA and hybridized DNA. We have also shown that large amounts of defects are beneficial for the electrochemical detection of DNA using stacked graphene platelet nanofibers44. These nanofibers are the direct
opposites of carbon nanotubes because they consist of perpendicularly stacked graphene sheets along the c-axis, exhibiting exclusively electrochemically active edges (with the exception of terminal basal planes). Such nanofibers also provide significantly enhanced signals for all four DNA bases when compared to graphite, glassy carbon, or pure
carbon nanotubes.

Graphene has also been used for electrochemical immunosensing. In immunosensing, the direct electrochemical detection of antibodyantigen recognition is usually not possible and electrochemically active labels must typically be used. There are two strategies in which graphene can be used. First, graphene can be used as an electrode surface for sensitive detection of a label 45. This case was employed for the graphene-enhanced detection of α-fetoprotein, which is a cancer biomarker. Graphene sheets were modified with antibodies, then the α-fetoprotein was added and consequently secondary antibodies loaded with microspheres bearing horseradish peroxidase enzyme as a sensitive label (Fig. 8, left panel). The second approach employs graphene as a label-bearing nanocarrier 46. More specifically, a gold nanoparticle electrode was modified with a probe antibody, to which phosphorylated protein p53 was entrapped. The secondary antibody was conjugated with graphene oxide and horseradish peroxidase to generate large amounts of electroactive molecules and thus a larger signal (Fig. 8, right panel).

Electrochemiluminiscence (ECL) or electrochemically generated luminescence is a type of chemilumininscence where one or more reactants are generated electrochemically. ECL is highly sensitive and is used for the detection of thrombin in the presence of interference on graphene platforms47.

Fluorescence

Fluorescence is a highly sensitive platform for biomolecular detection. Graphene is applied in various roles as a substrate in fluorescence quenching detection schemes. For example, the quenching principle was used for the aptamer-based detection of thrombin 48. Aptamer specific to thrombin was labeled with fluorescent dye. Graphene was used as a substrate for the nonspecific adsorption of the fluorescent dye-labeled aptamer for non-covalent assembly. In such a configuration, graphene quenched the fluorescence signal due to a transfer of fluorescence resonance energy from dye to graphene. The addition of thrombin results in the formation of quadruplex-thrombin complexes, which have a weak affinity to graphene. The change in conformation leads to a configuration where the dye is no longer in contact with the graphene sheet and thus the fluorescence is no longer quenched (Fig. 9a). This scheme is very sensitive, with detection limits as low as 31 pM. The very high surface area of graphene oxide also enables the application of a similar fluorescence quenching scheme, this time for DNA detection based on the hybridization of the complementary DNA strands, allowing for multiple detection of three different strands 49. This leads to a multicolor sensor for detecting multiple DNA targets in a single solution (Fig. 9b). Fluorescence detection can be used for virus detection in the format of a graphene microarray 50. A rotavirus-specific antibody was chemically linked via –NH2 groups to –COOH groups present on graphene oxide. After binding of the rotavirus to a chemically attached antibody, a secondary antibody linked to gold nanoparticles was added as a label. The gold nanoparticles act to quench the fluorescence of the graphene oxide (Fig. 10).

 

Conclusion and outlook


To summarize, we have described how biosensors can benefit from graphene as a transducing material. We have discussed the fundamental differences of the different types of graphene and their influence on applications in biosensing. Because graphene is a zero-gap semiconductor and an electroactive and transparent material, there are many possibilities for its application as a transducer or label in biosensing schemes. We have discussed enzymatic biosensors, genosensors, and immunosensors based on different transduction approaches, such as field-effect transistors, electrochemistry and electrochemiluminiscence, impedance, and fluorescence measurements. Although it would be beneficial to have side-by-side comparison of CNTs and graphene-based biosensors, such reports are very scarce. It is known that graphene nanoribbons, which have the dimensional restriction of a graphene sheet of tens of nanometers, exhibit significantly different electrical properties from large-sheet graphene. This is yet to be integrated in bio- FET design and promises even higher sensitivity and selectivity for FET devices. It is also expected that zigzag and armchair graphene edges will exhibit different electrochemical properties and, although fabrication of
purely zigzag or armchair edges is difficult, a breakthrough in this area is expected. Because the field is still young, it is expected to branch out into many applications to meet the needs of society in the areas of safety, enhanced health care, and a clean environment.

 

REFERENCES
1. Fitzer, E., et al., Pure Appl Chem (1995) 67, 473.
2. Novoselov, K. S., et al., Science (2004) 306, 666.
3. Bunch, J. S., et al., Science (2007) 315, 490.
4. Nair, R. R., et al., Science (2008) 320, 1308.
5. Shi, Y., et al., Small (2009) 5, 2005.
6. Geim, A. K., and Novoselov, K. S., Nat Mater (2007) 6, 183.
7. Hulanicki, A., et al. Pure Appl Chem (1991) 63, 1247.
8. Dreyer, D.R., et al., Angew Chem Int Ed (2010) 49, 9336.
9. Szabo, T., et al., Chem Mater (2006) 18, 2740.
10. Dreyer, D. R., et al., Chem Soc Rev (2010) 39, 228.
11. Gomez-Navarro, C., et al., Nano Lett (2010) 10, 1144.
12. Davis, T. J., et al. Angew Chem Int Ed (2005) 44, 5251.
13. Gao, W., et al. Nat Chem (2009) 1, 403
14. Pumera, M., Chem Eur J (2009) 15, 4970.
15. Trojanowicz, M., Trends Anal Chem (2006) 25, 480.
16. Scida, K., et al. Anal Chim Acta (2011) 691, 6.
17. Biotech Week, Dec 29 2010, “Hitachi, Ltd. details research in bioelectronics.”
18. Ahn, J.-H., et al., Nano Lett (2010) 10, 2934.
19. Ohno, Y, et al. Nano Lett (2009) 9, 3318.
20. Wallace, P. R., Phys Rev (1947) 71, 622.
21. Mohanty, N. and Berry, V. Nano Lett (2008) 8, 4469.
22. Dong, X., et al. Adv Mater (2010) 22, 1649.
23. Stine, R., et al., Adv Mater (2010) 22, 5297.
24. Mao, S., et al., Adv Mater (2010) 22, 3521.
25. Yanga, M., and Gong, S., Chem Commun (2010) 46, 5796.
26. Ohno, Y., et al., J Am Chem Soc (2010) 132, 18012.
27. Bonanni, A., and Pumera M., ACS Nano (2011) 5, 2356.
28. Hu, Y., et al., Chem Commun (2011) 47, 1743.
29. Pumera, M., Chem Soc Rev (2010) 39, 4146.
30. Pumera, M., et al. Trends Anal Chem (2010) 29, 954.
31. Wilson, G. S., and Hu, Y., Chem Rev (2000) 100, 2693.
32. Lu, J., et al., Chem Mater (2007) 19, 6240.
33. Wang, Y., et al., ACS Nano (2010) 4, 1790.
34. Fu, C., et al., Electrochem Commun (2009) 11, 997.
35. Shan, C., et al., Anal Chem (2009) 81, 2378.
36. Shan, C., et al, Biosens Bioelectronics (2010) 25, 1070.
37. Choi, B. G., et al. ACS Nano (2010) 4, 2910.
38. Lu, Q., et al, Talanta (2010) 82, 1344.
39. Xu, H., et al., Talanta (2010) 81, 334.
40. Palecek, E., and Fojta, M., Anal Chem (2001) 73, 74A.
41. Zhou, M., et al., Anal Chem (2009) 81, 5603.
42. Lim, C. X., et al. Anal Chem (2010) 82, 7387.
43. Pumera, M., et al., Chem Asian J (2008) 3, 2046.
44. Ambrosi, A., and Pumera, M., Phys Chem Chem Phys (2010) 12, 8943.
45. Du, D., et al., Anal Chem (2010) 82, 2989.
46. Du, D., et al., Anal Chem (2011) 83, 746.
47. Guo, Y., et al., Chem Commun (2011) 47, 725.
48. Chang, H., et al., Anal Chem (2010) 82, 2341.
49. He, S., et al. Adv Funct Mater (2010) 20, 453.
50. Jung, J. H., et al., Angew Chem Int Ed (2010) 49, 5708.