×
![]( "Buchcover ISBN 9783941650794")
Graphene Transistors for Biosensing and Bioelectronics
von Lucas Heinrich HessA major challenge in the field of bioelectronics is the advancement of neural prostheses that allow restoring damaged abilities such as hearing[Wil91] and vision,[Hum03] or can help to find solutions for treating motor disabilities[Jac06] or brain pathologies such as Parkinson’s disease.[Pah97] An essential step to achieve such goals is the development of suitable interfaces between the biological system, such as the cochlea in the ear or the retina in the eye, and electronic devices. The critical part of these interfaces is the efficient signal transfer from the electronics to the nervous systems, and vice versa. The communication with electronic sensors, such as microphones or cameras, or actuators can be achieved easily with standard microprocessor technol- ogy.
Today’s commercially available technologies are mostly based on micro-electrode arrays (MEAs) using silicon or metals. [Wil91, Hum03, Zre02] MEA-based devices are being applied for restoring hearing and vision, or for the treatment of neural disorders. In spite of the great advances achieved in the last years, these implants still suffer from a suboptimal performance, notably an insufficient long-term stability. In addition, MEA devices have an intrinsic poor spatial resolution, which is related to the rather large electrode impedance.[Sha07]
To overcome some of these limitations, arrays of field effect transistors (FETs) have also been developed for recording the electrical activity of nerve cells and tissue.[Off97, Vas98] If submicron FETs are used for signal recording, an increased spatial resolution can be combined with a considerably enhanced signal-to-noise ratio due to the possibility of intrinsic amplification. The maturity of semiconductor technology facilitates the production of high density FET structures, which can be used to record cell activity with unprecedented spatial resolution.[Pat06]
However, several drawbacks are associated with the presently used FETs platforms, which are not only related to an imperfect technology but are also inherent to the materials that are employed. In the harsh biological environment, many materials show a poor stability. For example, most semiconductors, including silicon are subject to stability problems in aqueous environments,[Bou84, Cab91] whereas others such as gallium arsenide are even toxic.[Web84] This has not only a negative impact on the performance of the sensor but can also induce damage to the surrounding tissue. Besides these chemically induced effects, the living tissue around the implant can also be damaged mechanically. Rigid and sharp prostheses usually induce scarring in the vicinity of the implant. This scar tissue not only loses its original function, but also introduces a barrier between the healthy tissue and the implant making the device inoperative. Although it is possible to fabricate flexible bioelectronic devices from classic semiconductors[Kim08] or organic semiconductors,[For04] the sensitivity of such devices is strongly limited by the poor electronic properties of such materials linked to the low crystal quality.
Thus, it is a major challenge to develop sensors from a chemically stable material providing a high sensitivity based on good electronic properties and allowing at the same time the fabrication of flexible devices. Graphene complies with all of these requirements.[Sch12] Consisting of only a monolayer of sp2-hybridized carbon atoms, graphene offers the mechanical stability to fabricate flexible devices and shows a very high chemical stability even in harsh biological environments due the strong C-C bond, which is one of the strongest in nature. As will be discussed later, this stability results in an excellent cytocompatibility.
Furthermore, the particular band structure of graphene results in outstanding elec- tronic properties.[Gei07] Charge carriers close to the charge neutrality point, also referred to as Dirac point, behave like quasi-relativistic particles, which largely reduces scattering and gives rise to extremely high charge carrier mobilities. Even at near- room temperature, values of more than 105 cm2V−1s−1 [Bol08] can be reached with exfoliated graphene, outnumbering other semiconductors which have been so-far used for similar applications. Furthermore, the chemical properties of graphene allow the fabrication of electrolyte-gated transistors without a solid dielectric.[Che10, Dan10] At the graphene-electrolyte interface, ions accumulate to screen the charges in the graphene and at its surface, forming an electrical double-layer. Together with the original charges, this charged ionic layer acts as a capacitor at the interface. As a result, interfacial capacitances of several µFcm−2 can be obtained, which is almost one order of magnitude higher than for traditional semiconductors using a stable dielectric.[Vas97] Consequently, graphene shows significantly higher transconductive sensitivities than other materials.[Hes11a] Moreover, the fact that graphene is only one atomic layer thick and can be transferred to almost any arbitrary substrate – including thin polymer films – allows the fabrication of high performance flexible transistors.[Eda08, Kim09]
Motivated by this combination of properties, such as high sensitivity, low noise, and stability in aqueous electrolytes, graphene solution-gated field-effect transistors (SGFETs) have been used for the detection of biologically relevant molecules and biomolecules such as glucose,[Kwa12] nucleic acids,[Moh08] proteins[Ohn10, Che12] and growth factors.[Kwo12] Furthermore, graphene SGFETs can been employed to detect enzyme activity[Hes14b] or to register action potentials from cells.[CK10, Hes11b, Hes13, Hes14a]
In this thesis, graphene SGFETs will be discussed both on a fundamental and an applied level. First, a short introduction to graphene will be provided in Chapter 2, discussing the basic properties and the electronic band structure of this material. The growth of high quality graphene by chemical vapor deposition (CVD) will be explained in this chapter and the characterization of the grown layers by Raman spectroscopy and scanning electron microscopy will also be shown. For the application of graphene transistors as biosensors in physiological environments, a detailed understanding of the graphene-electrolyte interface is necessary. Chapter 3 will focus on this interface and will introduce a basic model to describe it. This model will be complemented by numerical simulations and experimental data from electrochemical measurements. Based on this fundamental information, graphene solution-gated field- effect transistors will be explained in Chapter 4. First, details on the device fabrication will be provided followed by their basic characterization and an explanation of their working principle. The characterization will be completed by Hall effect experiments and the measurement of low-frequency noise. The performance of graphene SGFETs will be compared to transistors made from other common semiconductors such silicon or hydrogen-terminated diamond. To demonstrate the potential of graphene SGFETs for biochemical sensing, Chapter 5 will provide a discussion on the sensitivity towards the electrolyte’s composition. Experiments on the ion and pH sensitivity of the devices will be shown and explained based on models of the graphene-electrolyte interface. In the second part of Chapter 5, the modification of graphene transistors with polymer brushes will be discussed. To this end, graphene transistors were modified to exhibit an improved pH sensitivity and to detect the activity of surface-bound enzymes. In the final chapter of this thesis, the suitability of graphene as a material for the bioelectronic coupling between cells and electronics will be demonstrated, starting with the cytocompatibility of graphene. With the help of genetically-modified human embryonic kidney cells (HEK293), the electrical and electrochemical coupling between graphene transistors and cells will be analyzed. The coupling will be discussed with the help of the established point-contact model. This model will be extended to account for the specific coupling between cells and graphene SGFETs. In particular, the accumulation of ions between the cell and the transistor will be estimated and the influence of this accumulation on the transistor signal will be analyzed. Furthermore, recordings of spontaneous action action potentials from cardiomyocytes are presented and compared qualitatively to simulations.
Today’s commercially available technologies are mostly based on micro-electrode arrays (MEAs) using silicon or metals. [Wil91, Hum03, Zre02] MEA-based devices are being applied for restoring hearing and vision, or for the treatment of neural disorders. In spite of the great advances achieved in the last years, these implants still suffer from a suboptimal performance, notably an insufficient long-term stability. In addition, MEA devices have an intrinsic poor spatial resolution, which is related to the rather large electrode impedance.[Sha07]
To overcome some of these limitations, arrays of field effect transistors (FETs) have also been developed for recording the electrical activity of nerve cells and tissue.[Off97, Vas98] If submicron FETs are used for signal recording, an increased spatial resolution can be combined with a considerably enhanced signal-to-noise ratio due to the possibility of intrinsic amplification. The maturity of semiconductor technology facilitates the production of high density FET structures, which can be used to record cell activity with unprecedented spatial resolution.[Pat06]
However, several drawbacks are associated with the presently used FETs platforms, which are not only related to an imperfect technology but are also inherent to the materials that are employed. In the harsh biological environment, many materials show a poor stability. For example, most semiconductors, including silicon are subject to stability problems in aqueous environments,[Bou84, Cab91] whereas others such as gallium arsenide are even toxic.[Web84] This has not only a negative impact on the performance of the sensor but can also induce damage to the surrounding tissue. Besides these chemically induced effects, the living tissue around the implant can also be damaged mechanically. Rigid and sharp prostheses usually induce scarring in the vicinity of the implant. This scar tissue not only loses its original function, but also introduces a barrier between the healthy tissue and the implant making the device inoperative. Although it is possible to fabricate flexible bioelectronic devices from classic semiconductors[Kim08] or organic semiconductors,[For04] the sensitivity of such devices is strongly limited by the poor electronic properties of such materials linked to the low crystal quality.
Thus, it is a major challenge to develop sensors from a chemically stable material providing a high sensitivity based on good electronic properties and allowing at the same time the fabrication of flexible devices. Graphene complies with all of these requirements.[Sch12] Consisting of only a monolayer of sp2-hybridized carbon atoms, graphene offers the mechanical stability to fabricate flexible devices and shows a very high chemical stability even in harsh biological environments due the strong C-C bond, which is one of the strongest in nature. As will be discussed later, this stability results in an excellent cytocompatibility.
Furthermore, the particular band structure of graphene results in outstanding elec- tronic properties.[Gei07] Charge carriers close to the charge neutrality point, also referred to as Dirac point, behave like quasi-relativistic particles, which largely reduces scattering and gives rise to extremely high charge carrier mobilities. Even at near- room temperature, values of more than 105 cm2V−1s−1 [Bol08] can be reached with exfoliated graphene, outnumbering other semiconductors which have been so-far used for similar applications. Furthermore, the chemical properties of graphene allow the fabrication of electrolyte-gated transistors without a solid dielectric.[Che10, Dan10] At the graphene-electrolyte interface, ions accumulate to screen the charges in the graphene and at its surface, forming an electrical double-layer. Together with the original charges, this charged ionic layer acts as a capacitor at the interface. As a result, interfacial capacitances of several µFcm−2 can be obtained, which is almost one order of magnitude higher than for traditional semiconductors using a stable dielectric.[Vas97] Consequently, graphene shows significantly higher transconductive sensitivities than other materials.[Hes11a] Moreover, the fact that graphene is only one atomic layer thick and can be transferred to almost any arbitrary substrate – including thin polymer films – allows the fabrication of high performance flexible transistors.[Eda08, Kim09]
Motivated by this combination of properties, such as high sensitivity, low noise, and stability in aqueous electrolytes, graphene solution-gated field-effect transistors (SGFETs) have been used for the detection of biologically relevant molecules and biomolecules such as glucose,[Kwa12] nucleic acids,[Moh08] proteins[Ohn10, Che12] and growth factors.[Kwo12] Furthermore, graphene SGFETs can been employed to detect enzyme activity[Hes14b] or to register action potentials from cells.[CK10, Hes11b, Hes13, Hes14a]
In this thesis, graphene SGFETs will be discussed both on a fundamental and an applied level. First, a short introduction to graphene will be provided in Chapter 2, discussing the basic properties and the electronic band structure of this material. The growth of high quality graphene by chemical vapor deposition (CVD) will be explained in this chapter and the characterization of the grown layers by Raman spectroscopy and scanning electron microscopy will also be shown. For the application of graphene transistors as biosensors in physiological environments, a detailed understanding of the graphene-electrolyte interface is necessary. Chapter 3 will focus on this interface and will introduce a basic model to describe it. This model will be complemented by numerical simulations and experimental data from electrochemical measurements. Based on this fundamental information, graphene solution-gated field- effect transistors will be explained in Chapter 4. First, details on the device fabrication will be provided followed by their basic characterization and an explanation of their working principle. The characterization will be completed by Hall effect experiments and the measurement of low-frequency noise. The performance of graphene SGFETs will be compared to transistors made from other common semiconductors such silicon or hydrogen-terminated diamond. To demonstrate the potential of graphene SGFETs for biochemical sensing, Chapter 5 will provide a discussion on the sensitivity towards the electrolyte’s composition. Experiments on the ion and pH sensitivity of the devices will be shown and explained based on models of the graphene-electrolyte interface. In the second part of Chapter 5, the modification of graphene transistors with polymer brushes will be discussed. To this end, graphene transistors were modified to exhibit an improved pH sensitivity and to detect the activity of surface-bound enzymes. In the final chapter of this thesis, the suitability of graphene as a material for the bioelectronic coupling between cells and electronics will be demonstrated, starting with the cytocompatibility of graphene. With the help of genetically-modified human embryonic kidney cells (HEK293), the electrical and electrochemical coupling between graphene transistors and cells will be analyzed. The coupling will be discussed with the help of the established point-contact model. This model will be extended to account for the specific coupling between cells and graphene SGFETs. In particular, the accumulation of ions between the cell and the transistor will be estimated and the influence of this accumulation on the transistor signal will be analyzed. Furthermore, recordings of spontaneous action action potentials from cardiomyocytes are presented and compared qualitatively to simulations.