Patent Application
20230080531
A sequence listing containing SEQ ID NOs: 1-18 is provided herewith in a computer-readable .txt file and is specifically incorporated by reference.
Provided herein are sensors and related methods for electrical detection of material in an ionic solution using a field-effect transistor (FET) having a bent and curved (“crumpled”) channel layer positioned between source and drain electrodes. The invention is particularly useful for increasing electric detection of biomolecules and can be used in a variety of applications, including as a research and/or diagnostic tool. One specific use is for detecting molecules such as biomolecules such as proteins, small molecules and nucleic acid, including from an unprocessed biological sample.
Many techniques are available for detecting biological materials such as nucleic acids, including optical, electrical and mechanical. An issue with many of those techniques, however, relates to sensitivity, with a need for substantial pre-processing to increase the concentration of the material in the solution. Electrically-based techniques are particularly promising in that they can achieve high sensitivity. See, e.g., U.S. Pat. No. 9,376,713 (“Label free detection of nucleic acid amplification”), U.S. Pat. No. 10,527,579 (“Label free analyte detection by electronic desalting and field effect transistors”), and U.S. Pat. No. 10,175,195 (“Nanopore Sensors for Biomolecular Characterization”); U.S. Pub. Nos. 2017/0022546 (“Detection and Qualification of Methylation in DNA”), 2019/0011349 (“Label-Free Characterization Of Particles Suspended In A Fluid”).
A fundamental issue associated with electronically detecting molecules in a charged solution relates to Debye screening from counter ions in solution, also referred to as “Debye shielding” or “Debye length”. The charge of molecules in a solution can be masked by the charge of ions in solution, so that molecules can only be reliably electrically detected if they are close to the sensor surface. Outside the Debye length, charges are electrically screened. Accordingly, an increase in Debye length can provide a reduced screening effect and a corresponding more sensitive electrical detection of charged target materials, including biomolecules. The invention provided herein addresses this by effectively increasing the
Debye length for a FET used in a sensor to electrically sense a target molecule in a solution, thereby increasing sensor sensitivity.
There is also a need in the art to sensitively and reliably detect DNA amplification products. Enzymatic DNA amplification-based approaches involving intercalating DNA-binding fluorescent dyes and expensive optical detectors are the gold standard for nucleic acid detection. As components of a simplified and miniaturized system, conventional silicon-based ion sensitive field effect transistors (ISFETs) that measure decrease in pH due to generation of pyrophosphates during DNA amplification have been previously reported. Provided herein are methods and systems for selective adsorption of only single stranded DNA molecules on a FET channel surface, such as a graphene surface, to detect enzymatic DNA amplification by sensing the consumption of single stranded DNA primers, including in an isothermal amplification reaction.
Provided herein are sensors and related methods for sensitively and accurately detecting target molecules in an ionic solution, specifically including biosensors for detecting a target biomolecule. The sensors and methods use a specially configured FET having a crumpled channel layer. The crumpled geometry can effectively increase the Debye length extending from the channel surface for biomolecules in a physiological (e.g., ionic) solution. This provides a platform for detecting extremely small amounts of biomolecules, such as DNA, RNA, microRNA, proteins, small molecules and the like in an ionic or charge-containing solution, such as unprocessed biological fluid or biologically-relevant solutions that maintain biomolecule integrity (e.g., buffers).
The biosensor comprises a field effect transistor (FET). The FET has a source electrode and a drain electrode, wherein the source and drain electrodes are separated from each other by an electrode separation distance. The biosensor is compatible with a range of electrode separation distances. A channel layer is positioned between the source electrode and the drain electrode, wherein the channel layer has a crumpled geometry. The crumpled geometry is particularly relevant for increasing the Debye length at the channel layer surface, thereby increasing detection efficiency of the system by better accommodating charge shielding associated with detection in an ionic sample. A sample reservoir is in fluidic contact with the channel layer, wherein the sample reservoir is configured to hold a sample solution. The sample reservoir may be as simple as corresponding to the volume formed between the crumpled channel and the adjacent source and drain electrodes, particularly if microfluidics are used to precisely deliver fluid sample to the crumpled channel. For a larger volume sample, the sample reservoir may be formed from reservoir walls, such as by a polymer or non-polymer material, so long as there is not an adverse impact on the crumpled channel electronic measurement. A gate electrode is configured to electrically contact the sample solution in the sample reservoir. The crumpled geometry of the channel increases a detection limit of the biosensor to charged molecules in an ionic solution.
The channel layer may be formed of a two-dimensional layer of material selected from the group consisting of: graphene, doped silicon, silicene, ultra-thin metal, germanane, MoS2, and dichalcogenides. Preferably, the channel layer is a layer of graphene.
A support substrate layer may support the electrodes and channel layer. For example, the channel layer and electrodes may be deposited on the support substrate layer, and the support substrate layer shrinkably deformed to generate crumples in the overlying channel layer. “Support” is used broadly to refer to a material having continuous or discontinuous contact and/or continuous or discontinuous bonding with the supported component(s).
To further increase sensitivity, the biosensor may further comprise a probe anchored to the channel by a linker molecule configured to selectively bind a target molecule. For example, the probe can be selected from the group consisting of: a polynucleotide, or a peptide nucleic acid (PNA) probe, an aptamer, a protein, an antibody, and a capture agent, wherein the probe has a sequence selected to specifically bind a target molecule. For a nucleotide sequence, the probe may have a complementary sequence to at least a portion of the target molecule target sequence. Depending on the desired specificity, the length and sequence complementary homology are accordingly selected as known in the art. In this manner, the sensors and methods provided herein are compatible with any number of target sequences.
The biosensors provided herein are so sensitive that they can be configured to reliably detect a single target molecule, such as having a limit of detection that corresponds to about 1 to 1000 total nucleic acids in a sample solution that is introduced to the biosensor, including in a sample solution that is between about 1 μL and 1 mL. The sensitivity may also be described in terms of a concentration of a target molecule, with a sensitivity as low as atto- to zepto-molar range, such as a limit of detection of a biomolecule concentration of between 0.1 aM to 100 aM (corresponding to about 100-1000 DNA molecules in a 50 μL sample volume). As described herein, this high sensitivity is achieved by the crumpled geometry of the channel layer between source and drain electrodes, including from parameters associated with the channel layer, such as the material, amplitude and periodicity.
The crumpled geometry configuration, including periodicity and amplitude of bending, may be selected to provide a bandgap for an exponential change in a source drain current from a small number of charges.
The channel may have a thickness of between 1 nm to 2 mm and a surface area of between 1 mm2 to 100 mm2. Again, the particular geometry is selected depending on various parameters associated with the application of interest. For example, expected low concentration of target molecule in the sample may inform a smaller separation distance between the electrodes. Higher concentrations may have a larger surface area channel. The sample solution, including ionic strength, also informs selection, as well as how the sample is fluidically delivered to the channel. In other words, the invention tolerates a range of channel geometry, with a range of thicknesses, surface area, length and width.
During use with the sample solution, a Debye length at the surface of the crumpled geometry is greater than a Debye length of an equivalent channel having a flat geometry. Because the Debye length is solution-dependent, the increase may be described in terms of a percent increase. For example, the crumpled geometry may be described as providing an at least 10%, 20%, 50% or 100% increase in Debye length compared to an equivalent uncrumpled or flat geometry for an equivalent sample.
The biosensor is compatible with any fluid sample. For example, the sample may comprise a biomolecule selected from the group consisting of a protein, a DNA sequence, an RNA sequence, and any fragments thereof, and the biological solution is unprocessed whole blood, plasma, saliva or sputum. “Unprocessed” refers to there is not any active steps to concentrate the target molecule.
The crumpled geometry may correspond to a multi-axial deformation or a uniaxial deformation of the channel layer. For example, an underlying support substrate may be corresponding deformed to shrink in a preferred direction, thereby resulting in spatially aligned peaks in the channel. Alternatively, an underlying support substrate may be corresponding deformed to shrink in all directions, thereby resulting in non-spatially aligned peaks in the channel.
The crumpled geometry may have an average periodicity ranging from between 1 nm and 100 nm and an average amplitude ranging from between 1 nm and 100 nm.
The channel layer may be in continuous contact or discontinuous contact with a support substrate layer.
The biosensor electrode separation distance may be between 1 μm and 5 cm.
The support substrate can be formed of a material capable of undergoing a shrinkage transformation to thereby crumple the channel layer that is supported by the support substrate. The shrinkage may be by an applied force that is subsequently removed, to cause a strained support layer to decrease in size and relax to an unstrained dimension, thereby crumpling the channel bonded to the support layer, including discretely bonded. The shrinkage may be induced by a temperature change.
Also provided herein are methods of using any of the biosensors described herein. For example, the method may be for detecting a biomolecule by the steps of providing the biosensor and introducing the sample solution to the sample reservoir channel layer. A gate voltage is applied to the sample solution. A FET electrical parameter is monitored, wherein a change in the FET electrical parameter corresponds to presence of the biomolecule in the sample solution. Examples of FET electrical parameters include, capacitance, electric potential, resistivity, current and the like.
The crumpled geometry can be selected to provide an increase in sensitivity is by a factor of at least 1000× compared to a conventional sensitivity for an equivalent planar channel geometry. The increase in sensitivity can be further achieved by providing a microfluidic system to precisely position the sample at the channel surface, and also by use of probes that selectively bind the target molecule. Accordingly, any of the methods may further comprise the step of anchoring a probe to the channel, wherein the probe has a capture sequence configured to selectively bind a target biomolecule of interest.
In an embodiment, the Debye length for a flat FET channel may be less than the thickness of the source and drain electrodes. The geometry of the crumple may be selected to extend the Debye length that is greater than the source and drain electrode thickness, thereby effectively increasing the distance at which a charged molecule can be reliably detected more into the bulk solution volume.
The method may be configured for micro RNA (miRNA) detection, cell-free DNA (cfDNA) detection, protein detection using an antibody capture agent, or small molecule detecting using an aptamer capture molecule. The method is compatible for use with a sample solution corresponding to an unamplified sample.
The method may be configured for detecting DNA, RNA, including miRNA, protein, or a small molecule circulating in plasma or whole blood.
Also provided are specially configured FETs and related methods useful in a range of applications, including for determining presence or absence of a target polynucleotide in a sample. In particular, the FET channel is specially configured so that ssDNA preferably interacts with the channel surface relative to dsDNA. In this manner, an amplification solution plus sample that may contain an amplifiable target polynucleotide of an amplified polynucleotide sequence, is introduced to the FET, and whether or not a target polynucleotide is present in the solution is determined by monitoring a FET parameter whose output is dependent on the amount of primer in the amplification solution. As target polynucleotide is amplified, the ssDNA primer is incorporated into an amplified dsDNA target, and so less ssDNA primer is available to interact with the FET surface. This is reflected by a change in a FET parameter from the original “baseline” value where all ssDNA primer is available to interact with the FET. As more ssDNA primer is “incorporated” into the amplified target, there is a greater deviation or difference from the initial baseline FET parameter. In this manner, the amount of target polynucleotide in the original sample can be determined. In contrast, if there is no polynucleotide target in the sample, the FET parameter is unchanged as the amount of ssDNA primer in the amplification solution is unchanged. In a particularly useful embodiment, the FET's described herein may be incorporated into the biosensors described herein, including for a FET having a crumpled channel material, such as a semiconductor material.
The FET preferably has a crumpled channel layer that is formed of a semiconductor material, including graphene. See, e.g., U.S. Pat. App. No. 62/982,801 filed Feb. 28, 2020 titled “ULTRASENSITIVE BIOSENSOR USING BENT AND CURVED FIELD EFFECT TRANSISTOR BY DEBYE LENGTH MODULATION” (Atty Ref. 338264: 7-20P US), which is specifically incorporated by reference for the devices disclosed therein, including crumpled or bent and curved FET, and related methods.
Any semiconductor material that can interact with the aromatic rings of a nucleotide molecule via pi-pi (π-π) interaction is suitable. One example is graphene, where the hexagonal cells of graphene and an aromatic ring of the ssDNA form noncovalent (e.g., “stacking”) π-π interaction that is significantly stronger than for a corresponding dsDNA. Other examples include, but are not limited to, MoS2; dichalcogenides and silicene.
In an aspect, provided herein is a method of detecting amplification of a target polynucleotide. The method may comprise the steps of: providing a field effect transistor (FET) having a crumpled semiconductor material channel that is configured to form a π-π interaction with single stranded DNA; conducting an amplification reaction in an amplification solution comprising single stranded DNA (ssDNA) primers to obtain an amplified solution; contacting the FET with the amplified solution; and electrically detecting with the FET an amount of ssDNA primer in the amplified solution, thereby detecting amplification.
The semiconductor material is a two-dimensional layer selected from the group consisting of graphene; MoS2; dichalcogenides and silicene. The two-dimensional aspect refers to a layer configuration to provide adequate receiving surface area for interactions with ssDNA. The material in the FET for interacting with ssDNA is preferably crumpled graphene, referred herein as a gFET.
The step of conducting the amplification reaction may occur prior to the contacting step or may occur simultaneously with the contacting step.
The method may further comprise the steps of: detecting a FET electrical parameter prior to the step of conducting the amplification reaction to obtain a baseline FET electrical parameter value; detecting the FET electrical parameter after the step of conducting the amplification reaction to obtain a post-amplification FET electrical parameter value; comparing the baseline and the post-amplification FET electrical parameter values; and identifying presence of the target polynucleotide for a statistically significant difference between the baseline and the post-amplification FET electrical parameter values. The concept of “statistical significance” is a recognition that there is some inherent variability in experimental systems, including in the biological context, including arising from noise or random perturbance in the system, including degradation of ssDNA in the sample for reasons unrelated to amplification. Depending on the application of interest, including starting materials, the amount of change in the FET electrical parameter may be set to a value that will encompass these variations, such as greater than a 5%, greater than 10% or greater than 25% difference, corresponding to an at least 5%, at least 10% or at least 25% decrease in ssDNA primer.
The FET electrical parameter may be a change in current at a fixed voltage or a Dirac point shift voltage.
The electrically detecting step may occur periodically or continuously during the step of conducting the amplification reaction.
The method is compatible with any of a range of initial starting concentration of the target polynucleotide, including as low as 8×10−21 M in the amplification solution. The actual detection sensitivity is affected by a number of parameters, including the primer sequence (e.g., homology to target sequence), length of time of amplification, solution composition or purity (e.g., amount of ssDNA that is not the ssDNA primer), and the like.
The negative and positive target polynucleotide amplification solutions can be distinguished from each other at a target polynucleotide detection limit of between 4×10−21 and 1×10−18 M.
As described hereinbelow, the ssDNA primer binds to a surface of the crumpled material (e g , graphene) by noncovalent π-π interaction between hexagonal cells of a crumpled graphene and an aromatic ring of the ssDNA and amplified dsDNA does not bind to the crumpled graphene as strongly as ssDNA due to π-π stacking of aromatic rings in the dsDNA.
The ssDNA primers may be provided at a concentration so that after 60-90 minutes of amplification, at least 90% of all available ssDNA primers have been incorporated into amplified dsDNA.
The methods provided herein are compatible without labels and/or without surface-functionalization. This makes the methods, and related systems that implement the method, robust, reliable, cost-effective and particularly well suited with point-of-care devices. Unlike conventional amplification assays, where amplified target is visualized, such as with a fluorescent probe, the instant amplified target is electronically detected. This avoids need for labels and associated optical components.
The methods provided herein are compatible with an amplification reaction that is a loop mediated isothermal amplification (LAMP or RT-LAMP) reaction or a polymerase chain reaction (PCR). A LAMP reaction is preferred as it avoids the need for continuous thermal cycling over given number of amplification cycles.
The target polynucleotide may be present in the amplification reaction so that ssDNA primers are incorporated into amplified double stranded DNA (dsDNA) amplification product and identification of target polynucleotide comprises identifying a change in a FET electrical parameter measured during the electrically detecting step, including a decrease in a Dirac point shift, including a change in FET electrical parameter that is statistically significant from a baseline value.
Similarly, the target polynucleotide may be absent from the amplification reaction so that ssDNA primers are not incorporated into an amplified double stranded DNA (dsDNA) amplification product and identification of no target polynucleotide comprises identifying a no change condition in a FET electrical parameter measured during the electrically detecting step, including a no change in a Dirac point shift, including a not statistically significant change in the FET electrical parameter.
The methods and systems provided herein are compatible with a range of applications. For example, the target polynucleotide may be associated with: a pathogen; a disease condition; a genetic mutation; or a gene of interest. The pathogen may be viral or bacterial.
The methods, being label free, robust and sensitive, may be incorporated in a point-of-care device for rapid target polynucleotide identification in less than 1 hour, including as fast as 10 minutes or less.
Also provided herein are systems that implement any of the methods described herein.
Provided are systems for detecting a target polynucleotide in a sample solution comprising: a FET having: a source electrode; a drain electrode, wherein the source and drain electrodes are separated from each other by an electrode separation distance; a channel layer between the source electrode and the drain electrode, wherein the channel layer comprises a two-dimensional crumpled semiconductor material; a channel layer receiving surface that forms part of a sample reservoir, wherein the sample reservoir is configured to hold an amplifiable sample solution; an electrical detector electrically connected to the FET. See, e.g., U.S. Pat. App. No. 62/982,801 filed Feb. 28, 2020 for various FET configurations having a crumpled material therein, which is specifically incorporated by reference herein. The amplifiable sample solution comprises ssDNA primers and amplification reagents to amplify the target polynucleotide, wherein during use the amplifiable sample solution contacts the channel layer receiving surface of the sample reservoir. The ssDNA primers during use bind to the channel layer receiving surface by a noncovalent π-π interaction between the crumpled semiconductor material and an aromatic ring of the ssDNA at a higher affinity than dsDNA, and the electrical detector is configured to detect a level of ssDNA primer by detection of a change in a FET electrical parameter.
The crumpled semiconductor material is selected from the group consisting of: graphene; MoS2; dichalcogenides and silicene.
The FET electrical parameter may be a Dirac point shift voltage or a change in current at a fixed voltage, wherein the magnitude of the change varies with amount of ssDNA probe in the amplifiable sample solution.
The devices (biosensors) and methods provided herein are compatible with a range of to-be-detected targets or analytes, including charged or uncharged targets or analytes. For example, for a device that is a biosensor, the biosensor may detect a biomolecule, such as a polynucleotide, protein, antibody, polypeptide, and fragments thereof, as well as small molecules. The target may be a biomarker, including a biomarker associated with an infectious agent, thereby providing a diagnostic platform. For example, COVID-19 viral proteins are detectable, including capsid and spike protein, indicating the biosensors provided herein are useful for diagnosing infection with an infectious agent, including a virus or a bacteria. Uncharged molecules may also be detected, as illustrated by the experimental results for dopamine detection. Proteins, in general, are detected, as evidenced by the experimental results for IL-6 Accordingly, the biosensors provided herein are a useful platform for detecting a wide range of biomolecules.
Without wishing to be bound by any particular theory, there may be discussion herein of beliefs or understandings of underlying principles relating to the devices and methods disclosed herein. It is recognized that regardless of the ultimate correctness of any mechanistic explanation or hypothesis, an embodiment of the invention can nonetheless be operative and useful.
whnere q(x), A and ε0 are the net charge of the system (ions, DNA and water) in z, surface area of the bottom of the trench and vacuum dielectric constant, respectively.
The Dirac point shift due to a bandgap opening in 10−7% of the crumpled graphene after adding 2 aM complementary DNA is ˜5 mV and the corresponding I-V curve is plotted in
In the following description, numerous specific details of the devices, device components and methods of the present invention are set forth in order to provide a thorough explanation of the precise nature of the invention. It will be apparent, however, to those of skill in the art that the invention can be practiced without these specific details.
In general, the terms and phrases used herein have their art-recognized meaning, which can be found by reference to standard texts, journal references and contexts known to those skilled in the art. The following definitions are provided to clarify their specific use in the context of the invention.
“Field effect transistor” (FET) refers herein to a transistor having a sensor that detects changes in an electric field in and around the sensor. One unique feature of the instant FETs is that the channel is formed of a layer (e.g., 2-D) that is in a bent and curved configuration, having peaks and troughs with a periodicity between peaks, that provides an increase in sensitivity. FETs are also referred herein generally as ion-sensitive FETs (ISFETs) to emphasize the FET is sensitive to ions in the sample. Provided herein are devices and methods that effectively extend the Debye length to minimize adverse impact of ions in the sample, so as to increase FET sensitivity and accuracy. Any of the FETs provided herein may correspond to an ISFET.
“FET electrical parameter” refers to an electrically-measured parameter such as current, voltage, impedance or a parameter calculated therefrom, that reflects a target molecule interaction with the FET sensor, including a ssDNA primer. The FET electrical parameter may be a Dirac point shift voltage.
“Physiological level” of salts refers to a solution that is isotonic relative to a biological material, so that the biological analyte does not adversely swell or shrink under osmotic pressure.
“Minimally processed” refers to a biological sample that may be provided to the devices herein without any complex processing steps and so may be suitable directly with a biological sample. One example of a process that is considered minimal is sample dilution by introduction of a physiologically-compatible fluid, such as a solution isotonic to biological materials suspended within the sample, such as PBS or equivalents thereof.
“Analyte” is used interchangeably with “target” and refers to a material suspended in a fluid to be detected by the FET sensor. In an aspect, the analyte is a biological material and so is suspended within a fluid having a relatively high ionic strength that is compatible with the biological material. As specifically exemplified herein the analyte or target may be a biomolecule, including a biomarker or a polynucleotide. The systems are exemplified herein for polynucleotides, proteins and small molecules. The biomolecule may be charged or uncharged. Particularly relevant biomolecules include those having a net charge. The devices and methods described herein, however, may be used to detect molecules that do not have a net charge. For example, any molecule that alters the electric potential underlying the bent channel. Crumpled biosensors can detect uncharged molecules using various underlying mechanisms, including: (1) An uncharged molecule binds to a charged receptor (protein, antibody, aptamer made of DNA, RNA, LNA) attached on the curved 2D sensor. The binding of the neutral molecule changes the conformation of the receptor molecule altering the charge state and the charged mirrored in the sensor layer. Due to the curved, bent and crumpled 2D layers, the sensitivity is enhanced as compared to a flat surface; (2) A molecule that has a net zero charge can have an internal dipole where the length of the dipole is on the order of the Debye length over the curved 2D sensor layer. In this case, the charge mirrored or imaged is perturbed and is detected in the sensor surface. In this manner, any net charged or even uncharged molecule such as a single nucleotide or amino acid, can be probed in the cavities within the crumpled, curved 2D sensor layer. These nanoscale cavities are ion free, and even water molecules can be excluded, so that even individual single bases or amino acids are detectable due to the different internal net charged state.
“Functionalized” is used broadly to refer to processing of the sensor or sensing surface to facilitate interaction or binding between a target and the sensing surface. The processing is dependent on the analyte being measured. For example, a capture agent, such as an antibody, a receptor, a polynucleotide, a polypeptide, or other target-specific material may be attached to the sensing surface to provide target-specific binding. The target-specific binding is determined by those of skill in the art based on binding kinetics between a capture agent and binding region of the target. For nucleic acid molecules, a sequence complementary to the binding region is used. For a protein, an antibody that specifically binds the protein may be used.
“Crumpled” refers to a spatially-varying height of the channel layer. In other words, the 2-D “flat” layer of material has a 3-D geometry provided by the bends and curves in the layer. The crumples may be formed as described in U.S. Pat. No. 9,908,285 (Three-Dimensional Texturing of Two-Dimensional Materials) by SungWoo Nam et al., which is specifically incorporated by reference herein, including for crumpled graphene/graphite.
A “π-π interaction” refers to the interaction between a ssDNA and a material of the FET and is also referred to as π-stacking or π-π stacking. It is a noncovalent interaction between aromatic rings. The interactions herein are much stronger for ssDNA than dsDNA, so that the electrical parameter from a FET having the material is used to assess how much ssDNA is present in the fluid. In this manner, determination of target polynucleotide is determined.
“Target polynucleotide” refers to a polynucleotide that is desired to be detected. As used herein, target polynucleotide is used broadly to refer to any polynucleotide sequence that is capable of being amplified by an amplification reaction using single stranded DNA (ssDNA) primers such that, in the presence of the target polynucleotide the ssDNA primers are incorporated into the amplified product, which is double stranded DNA (dsDNA). An example of a target polynucleotide is a molecule that is capable of being amplified, so that a target polynucleotide may be exponentially amplified by a technique such as PCR or an isothermal technique such as LAMP. “Template” refers to the nucleic acid that is to be amplified by PCR. The target may be DNA. In an aspect, the target to be amplified corresponds to a nucleic acid sequence that is uniquely identified with a specific organism. For example, to detect a bacterial pathogen, a target polynucleotide that is a contiguous nucleic acid sequence unique to that pathogen is selected. The polynucleotide sequence informs the selection of primers that provides specific amplification of the target polynucleotide. For example, the primer may target a portion of the polynucleotide sequence itself, or may target a flanking sequence to the target polynucleotide. Selection of primers based on the desired target sequence is known in the art, with primers generally about 15 to 30 nucleotides in length, having a high sequence “homology” to a target sequence (e.g., corresponding to G-C and A-T base pairing), with forward and reverse primers separated by suitable distance to ensure forward and reverse primers are incorporated into the amplified dsDNA amplification product.
“Amplification reaction” refers to the biological sample (in which a one or more target polynucleotide detection is desired) plus amplification solution. Accordingly, an “amplification solution” includes a buffer fluid comprising an amplification enzyme and primer(s) for nucleic acid amplification of a target polynucleotide sequence. The materials required may correspond to those necessary to perform a PCR as known in the art. Examples of such materials include primers, enzymes such as DNA polymerases (e.g., Taq. polymerase), reverse transcriptases, dNTP, nucleases, salts (MgCl2) and PCR/LAMP buffers to facilitate effective PCR and LAMP. In an aspect, the amplification solution contains target polynucleotide, and more specifically a nucleic acid sequence that contains the target polynucleotide as well as flanking sequences. The solution may contain nucleic acid material from a biological cell, such as nucleic acid material from a lysed cell. In an aspect, the solution does not contain target polynucleotide, in which case the absence of target polynucleotide in the system will lead to no detection of target. The amplified product refers to the nucleic acid sequence that is produced as a result of the amplification reaction process, and corresponds to double stranded polynucleotides.
A solution or sample that “may contain target polynucleotide” refers to a sample that is introduced to the device in which it is desired to determine whether or not it contains target polynucleotide. For example, in point-of-care diagnostics or assays it is often desired to determine whether any of the DNA in the sample contains template. Similarly, in assays for the detection of a food borne pathogen, the target polynucleotide may correspond to a sequence that uniquely identifies the pathogen of interest. For rapid screening of multiple pathogens, the target polynucleotide may correspond to multiple target polynucleotides, with each target polynucleotide identifying a specific pathogen. Accordingly, the method may rapidly confirm there is no food borne pathogen, but if there is measureable detection, subsequent tests may be performed, if desired, to identify the specific pathogen.
“Isothermal amplification” refers to methods that can amplify nucleic acids exponentially, similarly to PCR amplification, but at a constant temperature, thereby avoiding need for thermocyclers. Examples include, but are not limited to, LAMP (loop-mediated isothermal amplification), HAD (helicase-dependent amplification), RCA (rolling circle amplification), MDA (multiple displacement amplification), WGA (whole gene amplification) and RPA (recombinase polymerase amplification).
The methods and devices provided are characterized as a platform-level technology in that the invention can be used with any target polynucleotide and, therefore, any primers, as known in the art of amplification reactions, such as polymerase chain reaction (PCR) and LAMP.
“Point-of-care” refers to tests performed on a sample obtained from a patient, wherein the diagnosis is provided at the time of the test. For example, tissue may be obtained directly from the patient, and introduced to any of the systems described herein, and a diagnostic result generated with the test result provided to the patient.
The invention can be further understood by the following non-limiting examples.
Field-effect transistor (FET)-based biosensors allow label-free detection of biomolecules by measuring their intrinsic charges. The detection limit of these sensors is determined by the Debye screening of the charges from counter ions in solution. Here, we use FETs with a deformed monolayer graphene channel for the detection of nucleic acids. These devices with even millimeter scale channels show an ultra-high sensitivity detection in buffer and human serum sample down to 600 zM and 20 aM, respectively, which are ˜18 and ˜600 nucleic acid molecules. Computational simulations reveal that the nanoscale deformations can form ‘electrical hot spots’ in the sensing channel which reduce the charge screening at the concave regions. Moreover, the deformed graphene could exhibit a band-gap, allowing an exponential change in the source-drain current from small numbers of charges. Collectively, these phenomena allow for ultrasensitive electronic biomolecular detection in millimeter scale structures.
All-electrical detection of biomolecules and specifically nucleic acids are of great interest for gene-expression investigations1, pharmacogenomics2, drug discovery3, and molecular diagnostics4-6. These methods also offer considerable promise for forensics7, environmental monitoring8 and global personalized medicine9. In particular, field effect transistor (FET) based detection of nucleic acids has drawn great attention as label-free and highly sensitive biomolecular sensing platform which can be readily integrated with other electronic components such as data analyzers and signal transducers. 2D materials such as graphene are attractive due to their unique properties such as ambipolar field effect, high carrier mobility, biocompatibility, mechanical strength, and flexibility10. 2D materials intrinsically exhibit high sensitivity in detection of charged biomolecules due to their ultimate thinness and extremely high surface to volume ratio. Compatibility to the conventional CMOS fabrication process is another potential advantage of using 2D materials, which carbon nanotube, Si-nanowire, nanoparticles do not have. Especially, large area graphene, which is grown through chemical vapor deposition (CVD) method, has been utilized in electrical systems, such as FET device for bio-sensing5 including detection of pH11, microorganisms12, blood sugar12, and more specifically, proteins at concentrations of 10 fM12,13, and nucleic acids (DNA or RNA) at the 100 fM concentrations5,14. There are a few reports that showed DNA and RNA detection at aM level, however, had significant level of background noise and lacked robust controls15,16. Further sensitivity would be highly desirable for detection of very rare molecules such as micro RNA (miRNA) or cell-free DNA (cfDNA) from unamplified samples17.
It is important to detect DNA/RNA such as miRNA circulating in serum or plasma with high ionic strength. Such detection could enable liquid biopsy, which can replace invasive tumor-tissue biopsies in many diagnostic applications. The existing approaches to monitor cancer-related miRNA is based on the polymerase chain reaction (PCR)18. Unfortunately, PCR is susceptible to interference by the inhibitory factors in biological samples, therefore not suitable to analyze miRNA directly from blood or serum samples19. Moreover, the result can be misinterpreted by bias and artifact due to the amplification efficiency of different sequences20. Therefore, there is an urgent need to develop amplification- and purification-free method to directly detect miRNA from biological samples such as serum.
One of the major hurdles to lower the detection limit of FET-based biosensor is shielding of the molecule charge by the counter ions in solution (termed Debye shielding)21,22. Outside the Debye length, which is less than 1 nm in physiological solutions, the charges are electrically screened. An increase in the Debye length can result in reduced screening effect and allow for a more sensitive electrical detection of charged biomolecules. While methods have been proposed to overcome this intrinsic limitation of FET biosensors21,22, these have focused on detecting biomolecules which are larger than the Debye length itself. None of the reports has tried to overcome the concentration limit of detection by modulating the Debye screening. Moreover, none of the works have modulated Debye screening in clinical solution such as serum or plasma21,22.
Computational reports have also predicted that the curved morphology of sensing materials can affect the Debye length (or volume), which can increase in concave regions of a nanowire sensor23. Thus, we hypothesized that if the surface of the sensing channel can be curved or bent at the micro- and nanometer scale, the Debye length could be modulated resulting in higher sensitivity. Previous works have shown that 3-dimensional ‘crumpled’ graphene can be created by deformations at the micro and nano-scale on pre-strained thermoplastics by relieving the stress and inducing buckle delamination of the graphene24. This approach can be used to engineer curving and bending of 2D materials and thin films. Several applications have been investigated using the mechanically-tunable crumpled graphene such as stretchable photosensors25, nanoplasmonic sensor26 and strain gauges27, and the in-plane strain can change the electronic properties of graphene, opening a band gap with a 1% stretch28,29. Moreover, crested 2D materials FETs recently showed large increase in the mobility as compared to standard devices30. It has also been separately shown that flat 2D semiconducting material such as MoS2 can be 10 times more sensitive than flat semi-metallic 2D graphene for biosensing applications31. Hence, we hypothesize that nanoscale bending of 2D graphene in 3-dimensions could result in high sensitivity due to modulation of the Debye length (or volume), and possibly due to strain induced band-gap opening in the graphene channels. Such deformed graphene (curved or bent 2D) layers have not yet been used for biosensor applications.
Here, we report the use of these deformed and bent (crumpled) graphene FET based electrical biosensors for ultra-sensitive detection of DNA/RNA molecules down to 600 zM of limit of detection (LOD) on millimeter scale structures. To the best of our knowledge, this is the highest sensitivity reported so far using any electronic biosensor for detection of DNA. Because of the simple fabrication process, the presented approach has several benefits over structures such as nanoribbons or nanopores32. The process does not require electron beam lithography to fabricate the nano-confined devices. The realization of the bent and crumpled graphene is achieved by macroscopic manipulation of the 2D layer and the resulting ‘nano’-sized features exhibit the superior sensing performance, while allowing facile fabrication and reproducibility. We demonstrate detection of 22-mer single and double stranded molecules by adsorption and hybridization experiments, respectively. We also showed that the target miRNA (let-7b) and a DNA probe hybridization was detectable as low as 600 molecules in 50 μL of buffer and undiluted human serum directly without amplification. The performance of the sensor was further enhanced using peptide nucleic acid (PNA) probe, which showed 600 zM of LOD, ˜18 molecules in an hour of incubation time. We show via molecular dynamics simulations the formation of electrical ‘hot spots’ at the nanoscale crumpled graphene regions where the Debye length can increase and also result in a local high potential due to the charge of DNA/RNA. Furthermore, the bending of the graphene monolayer at these hot spots can result in opening of a bandgap and provide for an exponential increase in the conductivity in vicinity of the biomolecules. These effects combined can allow for a measurable current change in millimeter scale channels even with 600 zM concentration of the target molecules.
Results: Characterization of Deformed Graphene FET Biosensor
The scheme for the graphene FET DNA sensing is illustrated in
The morphology of the flat and crumpled graphene was characterized by scanning electron microscope (SEM) and atomic force microscope (AFM). Disorganized herringbone-like structures were observed (
Then the flat and crumpled graphene FET sensor were characterized with liquid gate. Graphene FET generally shows intrinsic p-type behavior due to negatively charged impurities underneath the graphene sheet which are trapped during the transfer process34. The conductance of the graphene channel was modulated by liquid-gate potential applied to the solution reservoir. The accumulated ions modulate the conductance of the graphene channel by either p- or n-doping effects because of the ambipolar characteristics of graphene. The ambipolar transport characteristics of the FETs are illustrated in
Performance comparison of crumpled and flat graphene devices: The devices were then used for pH sensing and when the H+ ion concentration changes, the current through the transistor will change accordingly (
We then examined electrical sensing of DNA in fluid. Consistent with prior reports5,14, we also observed that the physical contact of ssDNA strands (let-7b sequence, Table 1) imposed n-type doping effects on flat graphene, resulting in a negative shift of Dirac point as shown in
Next, we investigated DNA hybridization to measure the sensitivity of the FET biosensor. Probe DNA molecules were immobilized as reported previously (
0.122 ± 0.007 V
0.072 ± 0.002 V
We also repeated the hybridization tests using PNA probe. It has been reported that PNA probe showed one order of higher sensitivity as PNA does not have the negative charges originated from phosphate backbone of DNA. Moreover, PNA does not need addition of NH2 functional group to react with the linker (Pyrenebutanoic acid succinimidyl ester) (
Taking into account convection-diffusion-reaction considerations38, evaporation induced convection and surface roughness effects on molecular absorption can facilitate the transport of nucleic acids to the graphene surface, reducing the diffusion-reaction time and result in high-sensitivity detection39-41. While about 35% of initial volume was evaporated in lhr in our experimental set up (
We also performed quantification of DNA attached on graphene with radioactive labeling to see if there was a difference in the density of attached DNA between the flat and the crumpled graphene surface. The relative signals from flat and crumpled graphene were similar to conclude that this high sensitivity of the crumpled graphene FET sensor was not from a difference in density of the attached molecules (
To demonstrate the capability of realistic applications of the platform, we performed miRNA detection spiked in undiluted human serum, which is not only highly ionic but also a complex mixture of biological components. In the same testing time (1 hour), we measured a clear Dirac point shift, but about half of the earlier measurements in PBS. We measured shifts at 20 aM compared to the negative control tests as shown in
Origin of Dirac Shift: To investigate the phenomena underlying the experimental results and the effect of ionic screening of DNA molecules, we studied the electrostatics and charge distribution of DNA and ions near flat and deformed graphene surfaces using molecular dynamics (MD) simulations (see the methods section for simulation details). The presence of the unscreened charges (acting as dopants) carried by DNA molecules near the surface of the graphene produces long range electrostatic potential leading to a change in the charge carrier density (Δn) of graphene and therefore a shift in Dirac point (ΔVD) given by45
where CT is the total gate capacitance, and Δn is directly proportional to the charge density of the unscreened DNA molecules (NDNAunscreend) adsorbed on the graphene surface45. The counter-ion screening of the DNA molecules lowers the net charge of adsorbed DNA and affects the detection sensitivity. As shown in
The interaction energies show that the adsorption of DNA to graphene in the concave region is the strongest. The calculated energies for the concave, convex, and across cases are −532.187 kcal mol−1, −467.484 kcal mol−1, and −416.308 kcal mol−1, respectively (
The detection sensitivity of DNA molecules is defined by the degree to which the DNA molecules are screened by ions present in the solution23. When the DNA molecules are isolated from the surface due to the ions present in the solution, the detection sensitivity can be significantly lower. We simulated the structure of ions and DNA relative to each other at the graphene interface for the four different configurations (
F is the Faraday constant, z is the normal distance from graphene surface (z=0 on graphene) and σ is the graphene surface charge density. As shown in
The bending used in
However, the observed Dirac point shift for aM concentrations on crumpled graphene cannot be explained merely by the low ionic screening of adsorbed DNA charges (assuming the maximum limit for NDNAunscreend). Therefore, there must be other factors that change the charge carrier density of graphene (n) to explain our measured detection of aM concentrations of DNA. Sarkar et al. showed that the existence of a bandgap in a single-layer MoS2 leads to a higher sensitivity of charge detection compared to its graphene counterpart31. Here, we hypothesize the creation of a band-gap in bent graphene and next calculated the bandgap (Eg) for flat and crumped graphene in the absence and presence of DNA bases using density functional theory (DFT) and GW methods. Typically, GW methods are more accurate for graphene than DFT.
As n is inversely related to μ, the change in n (and the corresponding Dirac point shift) due to bandgap widening upon addition of DNA can be obtained48. To achieve the observed Dirac point shift for aM concentrations used in our experiments, the change in bandgap from 0.4224 eV (crumpled graphene with no DNA) to 1.7641 eV (crumpled graphene with base A) must occur in at least ˜10−7% of the area of our graphene sensor (see the
Effect of nanoscale deformation on EDL structure: To further investigate the validity of the Debye length modulation, we measured the capacitance between the graphene and the liquid electrolyte23. The decreased screening for the crumpled graphene is seen once again (
To determine if the crumpled graphene device is capable of detecting biomolecules outside the normal Debye length at a certain buffer ionic concentration, we also varied the distance of the double strand (probe+target) DNA from the surface by a distance of 3 nt51 (
The devices can be miniaturized to micro- or nano-sized sensor in an array format. There are many fabrication and integration process challenges associated with the miniaturization. Some of these include; (i) maintaining a high quality crumpled surface if lithography is to be performed after shrinking, (ii) performing lithography first to form smaller flat graphene islands and then performing the heating or local shrinkage to cause crumpling while keep those smaller islands attached to the underlying surface, (iii) integration of silicon FETs at each pixel for row and column addressing to form larger arrays in a silicon substrate, etc.
We have demonstrated nucleic acid molecule detection on crumpled graphene FET electrical biosensor with unprecedented sensitivity using DNA and PNA as a probe. DNA adsorption and hybridization tests were demonstrated using cancer-related biomarker miRNA let-7b sequence in buffer and in human serum. The limit of detection was found to be 600 zM for crumpled graphene FET biosensor and 2 pM for flat graphene. We show via experiments and simulations that the nanoscale bending and deformations increases the Debye length in ionic solution to decrease the screening of the DNA/RNA molecules, thus contributing to the dramatic enhancement of sensitivity as compared to flat graphene FET sensors. To explain the results, Molecular Dynamics simulations revealed generation of large electrical potential due to DNA molecules in the nanoscale crevices and deformed regions that exclude ions, compounded by the formation of electrical band gap in the deformed graphene regions. These attributes coupled with increased molecular residence time due to increased roughness can result in ‘electrical hotspots’ allowing for a change in local conductivity, hence allowing atto-molar detection of DNA in millimeter scale sensors. We demonstrate the applicability of the technology by target molecules detection in undiluted human serum for cancer related miRNA. This technology can open opportunities for the development of more reliable and efficient diagnostic tools, and electrical point-of-care and implantable biosensors for early detection of biomolecules for various human diseases.
Methods:
Graphene synthesis: Monolayer graphene was grown on a copper foil via chemical vapor deposition (CVD). Before placing into the CVD furnace, copper foil was degreased with acetone and IPA, followed by nitrogen blow drying. The foil was then annealed at elevated temperature for 3 hours, while 50 sccm of hydrogen (H2) gas flow continuously. Monolayer graphene growth was initiated when 50 sccm hydrogen and 100 sccm methane (CH4) were introduced to the CVD chamber at annealing temperature of 1030° C., and the synthesis came to its end when two reaction gases were turned off. CVD furnace was then cooled down with flow of argon (Ar) gas, which completes the monolayer graphene synthesis process. To compare the performance among different source of graphene, graphene was also purchased from Grolltex and Graphenesquare. All the graphene showed the same LOD.
Fabrication of Graphene FET: After graphene was synthesized, the graphene/Cu foil was spin-coated with Poly(methyl methacrylate) (PMMA). Undesired graphene formed at the back side of copper foil was removed by oxygen plasma etching. The sample was cut into 1 mm×15 mm pieces with scissors or a razor blade. PMMA serves as a supporting layer while copper foil was etched by floating on 0.15 M sodium persulfate for about 5 h. The PMMA/graphene was rinsed by moving from the sodium persulfate solution to deionized (DI) water. The PMMA/graphene was then transferred onto a polystyrene substrate. The PMMA layer was removed by soaking in acetic acid for 5 min. The sample was annealed at 110° C. for 2 h to shrink the polystyrene substrate into ¼ of the original size and crumple the graphene. To fabricate transistor, conducting silver paste was used as source and drain electrodes at both ends of the graphene. Silicone rubber (DOWSIL™ 3140 RTV Coating) was used to insulate source and drain electrodes from liquid and construct a solution reservoir.
Immobilization of Probe: Pyrenebutanoic acid succinimidyl ester (PASE) (20 mM) in dimethyl sulfoxide (DMSO) was treated on the graphene overnight and rinsed with pure DMSO, ethanol and DI water; 50 μM of probe DNA solution was added on PASE-modified graphene for 3 hours. The graphene FET with probe DNA functionalization was rinsed with 1× PBS. 200 mM ethanolamine solution was treated for 1 hour to saturate the possibly unreacted amino group on PASE and rinsed with ethanol and 1× PBS solution. The volume of all treated chemicals and samples was 50 μL.
Target DNA or RNA incubation: The target DNA or RNA incubation was conducted by dropping complementary and negative control strands with concentrations that are indicated in the legends in
Electrical Measurements: I-V curves and resistance were measured in a semiconductor parameter analyzer equipped with a probe station. Ag/AgCl electrode was used to apply gate voltage (Vgs) to the 0.1× and 1× PBS buffer solution. For the human serum test, the tests were performed in in serum. In case of DNA absorption, the graphene chips were incubated in PBS overnight because of wettability issue of hydrophobic graphene. For the serum test, the graphene chips were incubated in serum overnight and the blank measurements were repeated till there was no shift only with serum. Then target and NC RNA in serum was treated on the chip. Vg was swept from −0.5 to 1 V and drain—source voltage (Vds) was picked between 0.03 and 0.1 V. Drain-source current (Ids) was measured at an assigned Vds.
Capacitance measurements: The capacitance measurements of flat and crumpled graphene on PS substrate were carried out using a CS 350 potentiostat (Contest, China) with three electrodes, including reference, counter, and working electrode. Here, silver chloride (Ag/AgCl), platinum (Pt), and the surface of the graphene channel on PS substrate were used for reference, counter, and working electrode, respectively. Cyclic voltammetry (CV) was chosen for the characterization method, while all three electrodes were immersed in 1× PBS solution for the measurement. Normalized total capacitance by measured graphene area (CT) based on CV is shown in
MD Simulations Methods: Molecular dynamics simulations were performed using the LAMMPS package52. The systems were generated by the visual molecular dynamics (VMD)53. To study the effect of the crumples on the EDL formation near graphene, different simulations with different graphene surface topologies were created (
Before starting the equilibrium and production simulations, the energy of each system was minimized for 15,000 steps. The equilibrium simulations were then performed in NPT ensemble for 2 ns at a pressure of 1 atm and a temperature of 300 K. This ensures that the water reaches its equilibrium density. The systems were further equilibrated in NVT ensemble for another 2 ns at 300 K. Temperature was maintained at 300 K by using the Nosè-Hoover thermostat57,58 with a time constant of 0.1 ps. The production simulations were carried out in NVT ensemble for 10 ns. Trajectories of particles were dumped every picosecond to study the structure of DNA and ions near the graphene sheet.
DFT and GW Methods: The optimized geometry of the adsorbed DNA nucleobases on flat and crumpled graphene are obtained using density functional theory (DFT) calculations. All DFT calculations are performed using the Vienna Ab initio simulation package (VASP)59-61. The local density approximation (LDA)62 functional of Ceperley-Alder is employed based on the projector augmented wave (PAW) method60. The cutoff energy for the plane wave basis set is 550 eV for all calculations. All ionic positions are optimized by a conjugate gradient method until the forces on each ion are less than 0.01 eV/A. The size of flat graphene and crumpled graphene model is about 12 Å×12 Å and 8.9 Å×12 Å, respectively. A vacuum separation of 40 Å between graphene and its periodic replicas is employed to eliminate the interaction between them. For accurate calculations of the electronic structures, the Brillouin zone is sampled using 18×18×1 k-point grid. The lattice parameter of flat graphene is computed to be 2.445 Å which is consistent with previous theoretical and experimental results63,64.
Single-shot approximation of GW (G0W0)65 was performed to obtain the band structures using VASP59-61 Each system has been initialized using DFT with the LDA66 exchange-correlation functional, energy cutoff of 400 eV, and Gaussian smearing of 0.05 eV. The number of total bands is set to 512 in all structures to ensure a significant number of empty bands is available as required by GW method. The energy cutoff of the response function and the number of frequency grid points are respectively set to 90 eV and 32 to control the high memory demand by the calculation of GW. Gamma-centered k-points are selected to be 6×1×1 for structures without a DNA base and 4×1×2 for structures with a DNA base. Finally, the vacuum space is maintained large enough (>20 Å) in the periodic directions to avoid unphysical interactions between the periodic images.
References corresponding to Example 1:
1. Gilliland, G., Perrin, S., Blanchard, K. & Bunn, H. F. Analysis of cytokine mRNA and DNA: detection and quantitation by competitive polymerase chain reaction. Proc. Natl. Acad. Sci. 87, 2725 LP-2729 (1990).
2. Inouye, M., Ikeda, R., Takase, M., Tsuri, T. & Chiba, J. Single-nucleotide polymorphism detection with “wire-like” DNA probes that display quasi “on-off” digital action. Proc. Natl. Acad. Sci. U.S.A. 102, 11606 LP-11610 (2005).
3. Schweitzer, B. et al. Immunoassays with rolling circle DNA amplification: A versatile platform for ultrasensitive antigen detection. Proc. Natl. Acad. Sci. 97, 10113 LP-10119 (2000).
4. Star, A. et al. Label-free detection of DNA hybridization using carbon nanotube network field-effect transistors. Proc. Natl. Acad. Sci. United States Am. 103, 921-926 (2006).
5. Hwang, M. T. et al. Highly specific SNP detection using 2D graphene electronics and DNA strand displacement. Proc. Natl. Acad. Sci. 113, 7088-7093 (2016).
6. Ping, J., Vishnubhotla, R., Vrudhula, A. & Johnson, A. T. C. Scalable Production of High-Sensitivity, Label-Free DNA Biosensors Based on Back-Gated Graphene Field Effect Transistors. ACS Nano 10, 8700-8704 (2016).
7. Sobrino, B., Brion, M. & Carracedo, A. SNPs in forensic genetics: a review on SNP typing methodologies. Forensic Sci. Int 154, 181-194 (2005).
8. Bohmann, K et al. Environmental DNA for wildlife biology and biodiversity monitoring. Trends Ecol. Evol. 29, 358-367 (2014).
9. Gonzalez-Angulo, A. M., Hennessy, B. T. J. & Mills, G. B. Future of Personalized Medicine in Oncology: A Systems Biology Approach. J. Clin. Oncol. 28, 2777-2783 (2010).
10. Rao, C. N. R., Gopalakrishnan, K. & Maitra, U. Comparative Study of Potential Applications of Graphene, MoS2, and Other Two-Dimensional Materials in Energy Devices, Sensors, and Related Areas. ACS Appl. Mater. Interfaces 7, 7809-7832 (2015).
11. Duarte-Guevara, C. et al. Characterization of a 1024×1024 DG-BioFET platform. Sensors Actuators B Chem. 250, 100-110 (2017).
12. Lin, Z el al. Carbon Nanomaterial-Based Biosensors: A Review of Design and Applications. IEEE Nanotechnology Magazine 13, 4-14 (2019).
13. Andoy, N. M., Filipiak, M. S., Vetter, D., Gutiérrez-Sanz, Ó. & Tarasov, A. Graphene-Based Electronic Immunosensor with Femtomolar Detection Limit in Whole Serum. Adv. Mater. Technol. 0, 1800186 (2018).
14. Hwang, M. T. et al. DNA Nanotweezers and Graphene Transistor Enable Label-Free Genotyping. Adv. Mater. 30, 1802440 (2018).
15. Campos, R. et al. Attomolar Label-Free Detection of DNA Hybridization with Electrolyte-Gated Graphene Field-Effect Transistors. ACS Sens. 4, 286-293 (2019).
16. Tian, M. et al. RNA Detection Based on Graphene Field-Effect Transistor Biosensor. Advances in Condensed Matter Physics, 8146765 (2018).
17. Kelley, S. O. Advancing Ultrasensitive Molecular and Cellular Analysis Methods to Speed and Simplify the Diagnosis of Disease. Acc. Chem. Res. 50, 503-507 (2017).
18. White A. K. et al. High-throughput microfluidic single-cell RT-qPCR. Proc. Natl. Acad. Sci. 108, 13999 LP-14004 (2011).
19. Costa M. C., Leitão A. L., Enguita F. J. MicroRNA Profiling in Plasma or Serum Using Quantitative RT-PCR. In: Alvarez M., Nourbakhsh M. (eds) RNA Mapping. Methods in Molecular Biology (Methods and Protocols). 1181 (Humana Press,2014).
20. Schwarzenbach, H., Hoon, D. S. B. & Pantel, K. Cell-free nucleic acids as biomarkers in cancer patients. Nature Rev. Cancer 11, 426-437 (2011).
21. Kulkarni, G. S. & Zhong, Z. Detection beyond the Debye Screening Length in a High-Frequency Nanoelectronic Biosensor. Nano Lett. 12, 719-723 (2012).
22. Gao, N. et al. Specific detection of biomolecules in physiological solutions using graphene transistor biosensors. Proc. Natl. Acad. Sci. 113, 14633 LP-14638 (2016).
23. Shoorideh, K. & Chui, C. O. On the origin of enhanced sensitivity in nanoscale FET-based biosensors. Proc. Natl. Acad. Sci. 111, 5111 LP-5116 (2014).
24. Wang, M. C. et al. Heterogeneous, Three-Dimensional Texturing of Graphene. Nano Lett. 15, 1829-1835 (2015).
25. Kang, P., Wang, M. C., Knapp, P. M. & Nam, S. Crumpled Graphene Photodetector with Enhanced, Strain-Tunable, and Wavelength-Selective Photoresponsivity. Adv. Mater. 28, 4639-4645 (2016).
26. Leem, J., Wang, M. C., Kang, P. & Nam, S. Mechanically Self-Assembled, Three-Dimensional Graphene—Gold Hybrid Nanostructures for Advanced Nanoplasmonic Sensors. Nano Lett. 15, 7684-7690 (2015).
27. Lee, M.-S. et al. High-Performance, Transparent, and Stretchable Electrodes Using Graphene-Metal Nanowire Hybrid Structures. Nano Lett. 13, 2814-2821 (2013).
28. Ni, Z. H. et al. Uniaxial Strain on Graphene: Raman Spectroscopy Study and Band-Gap Opening. ACS Nano 2, 2301-2305 (2008).
29. Choi, S.-M., Jhi, S.-H. & Son, Y.-W. Effects of strain on electronic properties of graphene. Phys. Rev. B 81, 81407 (2010).
30. Liu, T. el al. Crested two-dimensional transistors. Nature Nanotechnology 14, 223-226 (2019)
31. Sarkar, D. et al. MoS2 Field-Effect Transistor for Next-Generation Label-Free Biosensors. ACS Nano 8, 3992-4003 (2014).
32. Heerema, S., Dekker, C. Graphene nanodevices for DNA sequencing. Nature Nanotech 11, 127-136 (2016) .
33. Ma, R.-S. et al. Direct measurements of conductivity and mobility in millimeter-sized single-crystalline graphene via van der Pauw geometry. Chin. Phys. B 26, 066801 (2017).
34. Shi, Y el al. Effective doping of single-layer graphene from underlying SiO2 substrates. Phys. Rev. B 79, 115402 (2009).
35. Chen, T.-Y. et al. Label-free detection of DNA hybridization using transistors based on CVD grown graphene. Biosens. Bioelectron. 41, 103-109 (2013).
36. Adam, S., Hwang, E. H., Galitski, V. M. & Das Sarma, S. A self-consistent theory for graphene transport. Proc. Natl. Acad. Sci. 104, 18392 LP-18397 (2007).
37. Cai, B. et al. Ultrasensitive Label-Free Detection of PNA—DNA Hybridization by Reduced Graphene Oxide Field-Effect Transistor Biosensor. ACS Nano 8, 2632-2638 (2014)
38. Squirres, T. M. el al. Making it stick: convection, reaction and diffusion in surface-based biosensors. Nature Biotechnology 26, 417-426 (2008)
39. Barmi, M. R. et al. Convective Flows in Evaporating Sessile Droplets. J. Phys. Chem. B 118, 2414-2421 (2014).
40. Ljung, A. et al. Evaporation of a sessile water droplet subjected to forced convection in humid environment. Drying Technology 37, 129-138 (2019).
41. Stellwagen, E. et al. Determining the electrophoretic mobility and translational diffusion coefficients of DNA molecules in free solution. ELECTROPHORESIS 23, 2794-2803 (2002).
42. Ruggeri, F et al. Entropic Trapping of a Singly Charged Molecule in Solution. Nano Lett 18, 3773-3779 (2018).
43. Rechendorff, K. et al. Enhancement of Protein Adsorption Induced by Surface Roughness. Langmuir 22, 10885-10888 (2006)
44. Wang, Y. et al. Quantification of distinct let-7 microRNA family members by a modified stem-loop RT-qPCR. Mol. Med. Rep. 17, 3690-3696 (2018).
45. Dontschuk, N. et al. A graphene field-effect transistor as a molecule-specific probe of DNA nucleobases. Nat. Commun. 6, 6563 (2015). https://doi.org/10.1038/ncomms7563
46. Wu, G. et al. Graphene Field-Effect Transistors for the Sensitive and Selective Detection of Escherichia coli Using Pyrene-Tagged DNA Aptamer. Adv. Healthc. Mater. 6, 1700736 (2017).
47. Chen, J.-H. et al. Charged-impurity scattering in graphene. Nat. Phys. 4, 377 (2008).
48. Wang, J., Zhao, R., Yang, M., Liu, Z. & Liu, Z. Inverse relationship between carrier mobility and bandgap in graphene. J. Chem. Phys. 138, 84701 (2013).
49. Fang, J., Vandenberghe, W. G. & Fischetti, M. V. Microscopic dielectric permittivities of graphene nanoribbons and graphene. Phys. Rev. B 94, 45318 (2016).
50. Kim, J.-Y. et al. Flexible and Transparent Dielectric Film with a High Dielectric Constant Using Chemical Vapor Deposition-Grown Graphene Interlayer. ACS Nano 8, 269-274 (2014).
51. Zhang, G.-J. et al. DNA Sensing by Silicon Nanowire: Charge Layer Distance Dependence. Nano Lett. 8, 1066-1070 (2008).
52. Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput Phys. 117, 1-19 (1995).
53. Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 14, 33-38 (1996)
54. Hockney, R. W. & Eastwood, J. W. Computer Simulation Using Particles. Computer Simulation Using Particles (McGraw-Hill, 1981).
55. Wu, Y & Alum, N. R. Graphitic Carbon-Water Nonbonded Interaction Parameters. J. Phys. Chem. B 117, 8802-8813 (2013).
56. MacKerell, A. D. & Banavali, N. K. All-atom empirical force field for nucleic acids: II. Application to molecular dynamics simulations of DNA and RNA in solution. J. Comput. Chem. 21, 105-120 (2000).
57. Nosé, S. A unified formulation of the constant temperature molecular dynamics methods. J. Chem. Phys. 81, 511-519 (1984).
58. Hoover, W. G. Canonical dynamics: Equilibrium phase-space distributions. Phys. Rev. A 31, 1695-1697 (1985).
59. Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6, 15-50 (1996).
60. Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169-11186 (1996).
61. Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 1758-1775 (1999).
62. Ceperley, D. M. & Alder, B. J. Ground State of the Electron Gas by a Stochastic Method. Phys. Rev. Lett. 45, 566-569 (1980).
63. Lee, J.-H., Choi, Y.-K., Kim, H.-J., Scheicher, R. H. & Cho, J.-H. Physisorption of DNA Nucleobases on h-BN and Graphene: vdW-Corrected DFT Calculations. J. Phys. Chem. C 117, 13435-13441 (2013).
64. Karpan, V. M. et al. Graphite and Graphene as Perfect Spin Filters. Phys. Rev. Lett. 99, 176602 (2007).
65. Shishkin, M. & Kresse, G. Self-consistent GW calculations for semiconductors and insulators. Phys. Rev. B 75, 235102 (2007).
66. Kohn, W. & Sham, L. J. Self-Consistent Equations Including Exchange and Correlation Effects. Phys. Rev. 140, A1133-A1138 (1965).
Supplementary Information (for Example 1)
Note for Dirac voltage point determination: The measurements were repeated 6 times every two minutes for each data point (each concentration of target nucleic acids) and the Dirac point was confirmed to be stabilized when at least the last two measurements of Dirac points were same. The device was rinsed with fresh PBS every 3 measurements. Once the device and the Dirac point became stable, the Dirac points were same over many repeats of measurements.
The Sips model (J Chem Phys 16, 490 (1948), ACS Nano 10, 8700 (2016)) is adapted to fit the DNA hybridization specifically bound on the flat and crumpled graphene surface. Single stranded DNA target molecules, Let-7b, are specifically bound on the other complimentary single stranded DNA, 22-mer probe, molecules. They form duplex DNA molecules. The Sips model is the best fit to describe the relation of the shifted Dirac point voltage responding to the absorbed DNA concentration on the crumpled graphene surface to the target concentration (C) dissolved in a buffer,
where A is the maximum value of Dirac point shift with all probe sites occupied, Ka is the equilibrium dissociation constant, and a is the characteristic parameter of the Gaussian distribution of DNA-binding energies on the graphene surface.
The Sips model is commonly applied to describe a statistical distribution of the molecular (or gas) adsorption energies on a solid surface especially when the adsorption energies of the binding sites are heterogeneous other than homogeneous.
The A value corresponds to the saturation value of Dirac point voltage when all probe molecules are fully occupied. The saturation voltage for the crumpled graphene device is about 0.122 V which is about 1.7 times larger than 0.072 V for flat. The saturation voltage becomes larger when the graphene is crumpled, in a good agreement with the capacitance simulation results. i.e., the voltage is inversely proportional to the capacitance and the capacitance of the crumpled is reduced in a half of the value of the flat.
The association constant a characterizes the energy distribution of the DNA adsorption isotherm on the surface. The a value is in a range from 0 to 1. When a=1, the Equation turns into the Langmuir adsorption isotherm, in which all the DNA-binding sites have the same binding energy. If a decreases, the distribution curve shows a transition from a steep slope to a low slope as the target concentration increases. For the limit of a=0, it results in a constant value. From the crumpled graphene results, a=0.2 reflects the broader
DNA-binding energy distribution on the crumpled as compared with 0.436 for the flat case. It assumes that the broad energy distribution must be correlated with heterogeneous DNA-binding sites, such as the deep valleys, the slopes, or the peaks of the crumpled surface.
The dissociation constant Ka is a strong relationship with the binding DNA length. The Ka value decreases exponentially as the adsorbed molecular size increases (ref. ACS Nano 2016 10 8700). From the data set, two Ka values are very similar between the crumpled and flat devices.
As shown in
Consideration of convection-diffusion-reaction model: Taking into account convection-diffusion-reaction considerations, evaporation induced convection and surface roughness effect on molecular absorption may facilitate the transport of nucleic acids to the graphene surface, reducing the diffusion-reaction time significantly and contributing to the high-sensitivity detection. Also, the surface roughness of crumpled graphene may influence the molecular reaction process, compared to a flat graphene
In general, while a small volume of water droplet is placed on a solid substrate at a room temperature, the water droplet evaporates and drives convention flow. The convection flows down from the top surface of the droplet to its bottom solid-water surface and then flows up from the edge to the top. The convection flow rates depend on temperature and humidity. Due to convective flow, the molecules at the central region move along downstream to the bottom surface and spread over the surface in a radial direction. The flow speed varies inside the droplet. The speed is highest at the top central region and becomes much slower on the bottom region (where diffusion wins). The typical value ranges from 0.1 to 100 um/s at room temperature (J. Phys. Chem. B 118, 2414-2421 (2014)). In other words, the molecules approach to the surface fast at the central region and slow down on the surface where diffusive transport is dominant (Drying Technology 37, 129-138 (2019)).
Considering that target molecules are across r=3 mm, radius of water droplet, the diffusion time is proportional to r/2D, D is diffusion constant, and convective time is proportional to the evaporation rate, Q, is about 0.33 μl/min inside the probe station. Peclet number, Pe, can characterize the mass transport as diffusion-limited or reaction-limited (Nature Biotechnology 26, 417-426 (2008)). For convection and diffusion (D=100 μm2/s) (ELECTROPHORESIS 23, 2794-2803 (2002)), If Pe=diffusive time/convective time is >>1, reaction limit dominates the system while if Pe<<1, diffusion limit dominates In our case, Pe is ˜10, which means that chemical reaction occurs slowly while target molecules are supplied relatively fast to the surface. In the reaction-limited, the binding reaction is the major time obstacle to collect all target molecules on the sensor area.
By applying Langmuir kinetics to assume target bindings on the reactive surface, the target concentration can be estimated by:
where Pm is the probe concentration, [T] target concentration, [TP] target-probe complex concentration, kon is an associate rate constant, koff is a dissociation rate constant. For chemical reactions when all molecules can diffuse to the surface, the equation above can be solved:
where KD=koff/kon is the equilibrium dissociation constant. Given parameters, KD=1 pM (from Sip's fitting results), Pm=1×103/um2 (ACS Nano 10, 8700 (2016)), kon is 106 M−1 s−1, and Koff=10−5 s−1, as shown in the plot of [TP(t)] vs time, the equilibrium time reaches to ˜105 seconds (16 hours) for sub pico-molar target concentration. Since the equilibrium time decreases as the concentration increases, the graphene sensor is limited to pico-molar sensitivity within one-hour equilibrium state (Nano Lett. 5, 803 (2005)). Before reaching the equilibrium state, the one-hour practical incubation time indicated by arrow, about 16.5% of target DNA concentration molecules can be bound on the surface. Graphene sensors are experimentally able to detect them (Nano let. 18, 3509 (2018)).
The surface roughness of crumpled graphene may also influence the molecular adsorption process, compared to a flat graphene. Crumpled graphene forms randomly oriented valleys-and-peaks surface. The RMS roughness is about 500nm between valleys and 300 nm for their depth (Nano Lett. 15, 7684-7690 (2015)). When the molecular size is larger than 500 nm, conformational entropic trap holds the molecules inside the valley. On the other hand, when the size is much smaller, the molecules move “relatively” freely inside and outside of the rough surface. The molecules face increased interactions, such as Van der Waals, electrostatic force, and hydrophobicity. Thus, they stay in the valleys for longer than typical diffusion time scale (Nano Lett. 18, 3773-3779 (2018)). Ruggeri et al, claimed the well depth of 330 nm gives up to 5 kBT configurational free energy barrier, W, and the molecular residence time is proportional to exp(W/kBT). The molecules stay longer by 102× of diffusion time scale. Practically, molecular adsorption rate is increased with increasing surface roughness (Langmuir 22, 10885-10888 (2006)).
Therefore, collectively evaporation induced convection and surface roughness effect on molecular absorption possibly facilitate the transportation of nucleic acids to the graphene surface and may reduce the diffusion-reaction time significantly and contributed to the high-sensitivity detection.
DNA Adsorption to the Graphene Surface
The interaction energies between the DNA and crumpled graphene are calculated for different configurations of DNA. The equilibrium energies for the DNA in the concave, convex, across and flat regions are −532.187 kcal mol−1, −467.484 kcal mol−1, −416.308 kcal mol−1 and −500.383 kcal mol−1, respectively.
We investigated the DNA adsorption onto the graphene by plotting the area per nucleotide (packing) for the concentrations used in this study in
Dirac point shift mechanisms for different concentrations: We studied the Dirac point shift for a range of different concentrations to identify the dominant mechanisms by which the shift takes place. First, we excluded the effect of the band gap and calculated the shift solely based on the charge transfer from the unscreened DNA molecules using
By matching the experimental ΔVD, charge transfer (NDNAunscreened) can be extracted. The ratio of the transferred charge to the total available DNA charge in the solution is plotted for different concentrations in
Modeling of Dirac point shift for band gap opening:
The shift is directly obtained from
without including any effect from bandgap. However, for the ultralow concentrations (e.g., 2 aM), the shifts are too small to be noticed without including the effects from bandgap opening (
where e is the elementary charge, and ρ is the resistivity. Wang et al.1 showed that electronic mobility (μ) in graphene decreases with increasing bandgap where mobility is obtained empirically from
(Eg in eV and μ in cm2 V−1 s−1) Since the bandgap changes are local, the global (macroscopic) change in the charge carrier density is estimated by Δnglobal=b Δnbangap, where b is the area fraction of the affected regions. For the 2 aM concentration, assuming a bandgap change from 0.4224 eV to 1.7641 eV in 10−7% of the crumpled graphene where each DNA nucleotide affects at least an area of 39.9 nm2 (see the calculation below in the next section), the shift is noticeable and is estimated to be ˜5 mV
as shown in
Crumpling graphene by introducing 1D periodic ripples is found to produce a bandgap opening2-4. The opening of bandgap is attributed to the change in graphene curvature introducing quantum confinement with distinct electronic structures compared to the pristine/flat graphene3. In pristine graphene, the C—C bond length is ˜1.41 Å for all carbon atoms with an angle of 120° which results in sp2 hybridization. When graphene is crumpled, the bond length and the angles vary across the graphene. The optimized structure that we obtained using DFT shows that the C—C bond length has a value of 1.41-1.55 Å depending on the local curvature. In addition, the angles between the carbon atoms is found to be either 120° or 110°. The bond length of 1.55 Å and angle of 110° resemble the structure for sp3 hybridized C atoms5,6. Thus, crumpled graphene contains sp3 and sp2 hybridization between C atoms. This is expected since the C atoms are not in the same plane due to crumpling. Further, the partial density of states of the pristine graphene (see
Calculation of area needed for low number of molecules
DNA Orientations in Ab Inito Calculations:
In orientation 1 and orientation 2 (see
Since GW is computationally expensive, we only performed GW for orientation 1 on all the graphene surfaces and orientation 2 on the crumpled armchair graphene. Comparing the DFT and GW bandgaps of crumpled graphene, we note that the DFT bandgaps for orientations 1 and 2 are almost identical. However, the GW bandgaps are different for different orientations. To understand the interactions of the different orientations, we computed the interfacial charge density. The charge density difference of the system (i.e., graphene and nucleobase system), Δρs, is defined as
Δρs=ρs−ρg−ρn
where ρs is the charge density of the full system including both the crumpled graphene and the nucleobase in the unit cell, ρg is the charge density obtained by simulating the crumpled graphene without the nucleobase, and ρn is the charge density obtained by simulating the nucleobase without the crumpled graphene. Δρ_s represents the interfacial charge density due to the adsorption of the nucleobase onto the crumpled graphene. As shown in
Raman Spectroscopy Analysis on VHS Substrate:
As shown in the
1. Transfer graphene on a PDMS stamp
2. Transfer graphene onto VHB substrate.
a. For Flat sample: (VHB tape was put on a slide glass because the soft VHB nature made hard the contact printing process.)
b. For crumpled sample: 100% prestrain was applied in x- and y-axes to apply same amount of prestrain as crumpled graphene on PS substrate.
The results are shown in
Based on 2D/G intensity ratio (
References for the Supplementary Information
1. Wang, J. Y., Zhao, R. Q., Yang, M. M., Liu, Z. F. & Liu, Z. R. Inverse relationship between carrier mobility and bandgap in graphene. J. Chem. Phys. 138, doi:10.1063/1.4792142 (2013).
2. Bai K K, Zhou Y, Zheng H, Meng L, Peng H, Liu Z, Nie J C, He L. Creating one-dimensional nanoscale periodic ripples in a continuous mosaic graphene monolayer. Physical review letters. 2014 Aug. 18; 113(8):086102.
3. Lee J K, Yamazaki S, Yun H, Park J, Kennedy G P, Kim G T, Pietzsch O, Wiesendanger R, Lee S, Hong S, Dettlaff-Weglikowska U. Modification of electrical properties of graphene by substrate-induced nanomodulation. Nano letters. 2013 Jul. 19; 13(8):3494-500.
4. Lim H, Jung J, Ruoff RS, Kim Y. Structurally driven one-dimensional electron confinement in sub-5-nm graphene nanowrinkles. Nature communications. 2015 Oct. 23; 6:8601.
5. Sofo J O, Chaudhari A S, Barber G D. Graphane: A two-dimensional hydrocarbon. Physical Review B. 2007 Apr. 10; 75(15):153401.
6. Boukhvalov D W, Katsnelson M I, Lichtenstein A I. Hydrogen on graphene: Electronic structure, total energy, structural distortions and magnetism from first-principles calculations. Physical Review B. 2008 Jan. 22; 77(3):035427.
7. Abrasonis G, Gago R, Vinnichenko M, Kreissig U, Kolitsch A, Möller W. Sixfold ring clustering in s p 2-dominated carbon and carbon nitride thin films: A Raman spectroscopy study. Physical Review B. 2006 Mar. 24; 73(12):125427.
8. Zhang Y, Shang J, Fu W, Zeng L, Tang T, Cai Y. A sp 2+ sp 3 hybridized carbon allotrope transformed from AB stacking graphyne and THD-graphene. AIP Advances. 2018 Jan. 30; 8(1):015028.
9. Miglio A, Heinrich C P, Tremel W, Hautier G, Zeier W G. Local bonding influence on the band edge and band gap formation in quaternary chalcopyrites. Advanced Science. 2017 September; 4(9):1700080.
10. Jaffe J E, Zunger A. Theory of the band-gap anomaly in AB C 2 chalcopyrite semiconductors. Physical Review B. 1984 Feb. 15; 29(4):1882.
11. Khoa D Q, Nguyen C V, Bui L M, Phuc H V, Hoi B D, Hieu NV, Nha V Q, Huynh N, Nhan L C, Hieu N N. Opening a band gap in graphene by C—C bond alternation: a tight binding approach. Materials Research Express. 2019 Jan. 9; 6(4):045605.
12. Prasanna R, Gold-Parker A, Leijtens T, Conings B, Babayigit A, Boyen H G, Toney M F, McGehee M D. Band gap tuning via lattice contraction and octahedral tilting in perovskite materials for photovoltaics. Journal of the American Chemical Society. 2017 Aug. 4; 139(32):11117-24.
13. Peng X, Velasquez S. Strain modulated band gap of edge passivated armchair graphene nanoribbons. Applied Physics Letters. 2011 Jan. 10; 98(2):023112.
14. Park C, Atalla V, Smith S, Yoon M. Understanding the Charge Transfer at the
Interface of Electron Donors and Acceptors: TTF-TCNQ as an Example. ACS applied materials & interfaces. 2017 Aug. 2; 9(32):27266-72.
15. Fernandez A C, Castellani N J. Dipole moment effects in dopamine/N-doped-graphene systems. Surface Science. 2020 Mar. 1; 693:121546.
16. Hybertsen MS, Louie SG. Electron correlation in semiconductors and insulators: Band gaps and quasiparticle energies. Physical Review B. 1986 Oct. 15; 34(8):5390.
17. Johnson K A, Ashcroft N W. Corrections to density-functional theory band gaps. Physical Review B. 1998 Dec. 15; 58(23):15548.
The biosensors, FETs, systems and methods described herein can be used for the sensitive detection of an amplification product, such as polynucleotide activation. For example, this example relates to a high sensitivity graphene field effect transistor (gFET) for detection of DNA amplification.
Any of the biosensors described previously, including as illustrated in
In this Example, we use Bst polymerase in a Loop Mediated Isothermal Amplification (LAMP) reaction combined with target specific primers and crumpled graphene field effect transistors (gFET) to electrically detect the amplification by sensing the reduction in primers. Graphene is known to adsorb single stranded DNA due to noncovalent π-π bonds, but not double stranded DNA. Other materials that mediate a similar reaction can be used in place of graphene, as described herein. Our approach does not require any surface functionalization and allows the detection of primer concentrations at the endpoint of reactions. We chose crumpled gFET over the conventional flat gFET sensors due to their superior sensitivity as recently demonstrated (see, e.g., U.S. Pat App. No. 62/982,801 filed Feb. 28, 2020 titled “ULTRASENSITIVE BIOSENSOR USING BENT AND CURVED FIELD EFFECT TRANSISTOR BY DEBYE LENGTH MODULATION” (Atty Ref 338264: 7-20P US)). We were able to detect the endpoint of amplification reaction with starting concentrations down to 8 zeptomolar in 90 minutes including the time of amplification and detection. With its high sensitivity and small footprint, our platform will help bring complex lab based diagnostic and genotyping amplification assays to the point-of-care.
Rapid and accurate detection of infection causing pathogens such as E. coli and others remain a challenge in healthcare1. State of the art for sensitive and specific detection of pathogens usually relies on their genomic DNA amplification using techniques such as Polymerase Chain Reaction (PCR) or Loop Mediated Isothermal Amplification (LAMP) which require expensive optical readout instrumentation for measuring the fluorescence signal from intercalating DNA binding dyes2. Towards optics-free platforms, approaches using carbon-nanotube and silicon nanowire field effect transistors for electrical and label-free DNA ‘hybridization-based’ detection have been previously reported3,4,5,6. However, these DNA hybridization based techniques often have a lower a limit-of-detection (LOD) (picomolar to femtomolar range)6, as compared to the LOD of optics based DNA amplification techniques (low attomolar range)7,8. CMOS compatible ion sensitive field effect transistors (ISFET) arrays have also been used for hybridization-based DNA detection but with limit of detection usually in nanomolar range9. ISFETs have also been used for label-free electrical detection of DNA amplification by detecting pH changes during an amplification reaction10,11. However, these platforms often require special low buffer capacity reactions for detecting signals and have inferior performance (LOD>10 aM10) as compared to optical readout systems.
Towards electrical detection of biomolecules, graphene Field-effect transistor-based biosensors (gFET) offer many potential advantages such as large surface-to-volume ratio, high carrier mobility and low cost12. Several gFET based platforms offering high sensitivity, low cost, and high throughput detection using diverse sensing methods such as electrochemical13, back-gated G-FETs14, and liquid-gated G-FETs15-18 have been reported in the literatures. The semiconductor-dielectric interface inside the conventional ISFET is not accessible for biomolecule functionalization and usually the analyte in the sample will be adsorbed/attached to the gate oxide and will modulate the capacitance and electric field across it16. In such devices, the gate oxide should be thin to better modulate the electric signal and increase the sensitivity of the device while also being thick enough to reduce the gate leakage current and increase signal to noise ratio19. In contrast, graphene electrolyte-gated FET biosensors can overcome these drawbacks as the transistor channel is formed by a single two-dimensional (2-D), one-atom-thick carbon layer, which can be left accessible for direct functionalization or adsorption with biomolecules. Hence, the local gating effect is much more effective than conventional devices20-26. Many studies have demonstrated the use of gFET sensor for biomolecules detection, including protein, DNA, and bacteria27. To enable specific binding, GFETs can be functionalized with single-stranded probe DNA for detection of specific target DNA through DNA-DNA hybridization with complementary sequences. Using this DNA hybridization based approach, gFETs typically offer a limit of detection (LOD) ranging from 100 pM to 1 fM15,17,28-30. A further improvement of LOD, up to 25 aM, was recently reported by optimizing the bio-FET channel and using a large-area in-plane gate surrounding the graphene channel that allows a uniform distribution of potential inside the water droplet and a uniform gating field19.
The net electrostatic effect of a charged molecule in the solution containing different ions is measured in terms of Debye length with characteristic thickness of less than 1 nm in physiological solutions19. Outside the Debye length, charge carriers are increasingly electrically screened. By increasing the Debye length, the sensitivity of the G-FETs to detect target DNA can be enhanced because more sequence length of DNA strand is within the Debye length and thus more electric charge is induced near the graphene surface. This will result in a higher change in the electric conductance of the graphene channel. Previous computational studies have shown that curved morphologies, such as the concave regions of nanowire sensors, can affect the Debye length31. Studies have also reported that crumpled graphene, which has concave and convex deformations at the micro and nano scale can be fabricated using pre-strained thermoplastics and relieving stress to induce buckle delamination of graphene32. The mechanically tunable crumpled graphene has already been explored in several applications such as stretchable photosensors33 and strain gauges34. Its application in biosensing has also been recently reported with hybridization based-DNA detection with improved sensitivity36. The sensitivity enhancement in crumpled graphene in this recent study was attributed to the nanoscale deformations, specifically the concave regions, that decrease the charge screening of the nucleic acid molecules by increasing the Debye length in the ionic solution36. In addition, the crumpled graphene could form a band gap in the deformed regions further allowing for an exponential source-drain current change from a small number of charges36. However, despite its improved sensitivity, this hybridization-based method still requires surface functionalization of the graphene channel and thus can lead to complications in fabrication process.
In this example, we show that crumpled graphene FETs can be used to detect physiosorbed single stranded DNA (ssDNA) molecules (as compared to double stranded DNA product) on its surface and use this for detecting enzymatic amplification by monitoring the reduction in primer (ssDNA) concentration in a reaction. Unlike double stranded DNA (dsDNA), ssDNA can be strongly adsorbed on the graphene FET surface through noncovalent stacking interaction between the hexagonal cells of graphene and the aromatic ring structure of unpaired nucleobases37-40. In our platform, we use this discrimination power of crumpled graphene coupled with primer (ssDNA) consumption in enzymatic loop mediated isothermal amplification(LAMP)41 to detect E. coli DNA down to zeptomolar (zM) concentrations in end-point LAMP reactions. gFET Signal generated from primers is reduced only if the specific target is present and amplification occurs where primers are consumed during amplification and become a part of formed dsDNA. In contrast, the dsDNA produced in the amplification does not produce any significant shift in the Dirac voltage. LAMP was chosen as our assay reaction as it uses a robust strand-displacement Bst polymerase and six sequence specific primers compared to the 2 used in PCR42. Moreover, since it is an isothermal reaction, the instrumentation demands are easier. LAMP is also known to be highly specific and sensitive and hence offers single molecule (attomolar) sensitivity for detecting target DNA43. We expect our platform, with its electrical, label-free and surface modification-free detection of enzymatic amplification, will allow translation of complex and sensitive lab-based amplification assays to truly point-of-care and small footprint detection devices.
Results and discussion: Process overview and device characterization for physisorption of ssDNA and dsDNA: The approach for using crumpled graphene FET sensor for detection of LAMP reaction is illustrated in
Our crumpled graphene FET biosensor fabrication process was adapted from previously published protocol32. Briefly, a 2 mm×14 mm graphene channel was transferred onto a thermoplastic polystyrene substrate and annealed at 110° C. for 4 hours. During the annealing process, the underlying pre-strained thermoplastic substrate shrinks and results in the buckling and crumpling of graphene channel. For the flat graphene FET used in this study, this annealing step was omitted. Finally, the source and drain metal electrodes are formed and an ionic solution reservoir was created around the graphene channel and gate voltage is applied directly to the top of the ionic solution placed in the reservoir in the device. The black regions on
Next, we investigated the physical adsorption of dsDNA on crumpled graphene FET biosensors by using a similar protocol as above and
AFM characterization of physical adsorption of molecules on graphene surface: Structural features of the flat graphene surface with physiosorbed molecules were characterized using an atomic force microscope (AFM) for different test cases and the results are shown in
Attomolar E. coli DNA detection using crumpled gFET biosensors: To test the ability of our crumpled gFET biosensor to distinguish positive amplification as compared to LAMP reactions where no target was present, we first performed real-time LAMP reactions for E. coli DNA with Evagreen dye to confirm that DNA amplification has occurred.
Zeptomolar E. coli DNA detection using crumpled gFET biosensors: In
We demonstrate detection of enzymatic DNA amplification using crumpled graphene FET with detection limits down to Zeptomolar target concentrations in the starting sample. Our platform deploys the evolutionary sensitivity and robustness of Bst polymerase combined with target specific primers, while using crumpled graphene FET to electrically detect the amplification by sensing the reduction in the primer molecules. We also characterize the physisorption of ssDNA, dsDNA, and LAMP reaction mix, pre- and post-amplification, to show lower physisorption of molecules (lower roughness in AFM) post-amplification due to reduced or consumed primers. Crumpled graphene FET provided better and higher sensitivity and Dirac voltage shifts in comparison to the flat counterpart and hence were chosen for our study. For our enzymatic reactions, we chose Loop Mediated Isothermal
Amplification (LAMP) as it only needs a constant temperature for performing reactions and will allow easy translation into point-of-care and small footprint devices in the future. Moreover, since Bst polymerase has been applied to direct detection from complex matrices such as blood or saliva and requires specificity of 6 unique target specific primers, our platform will be more suitable for direct processing of complex samples in the future and will be superior to direct hybridization based approaches which rely on a single sequence/primer specificity. Compared to the conventional silicon based ISFET based approaches which sense changes in pH, require low buffer capacity reaction, and have a reported a detection limit of >10 aM, our platform can sense target down to zeptomolar concentrations with electrical sensing possible directly in 1× PBS. Due to these reasons, the platforms provided herein allow translation of complex lab-based diagnostic and genotyping amplification assays to truly point of care and bedside platforms.
References for Example 2:
1. Sinha, M. et al. Emerging technologies for molecular diagnosis of sepsis. Clinical Microbiology Reviews 31, (2018).
2. Peker, N., Couto, N., Sinha, B. & Rossen, J. W. Diagnosis of bloodstream infections from positive blood cultures and directly from blood samples: recent developments in molecular approaches. Clinical Microbiology and Infection 24, 944-955 (2018).
3. Star, A. et al. Label-free detection of DNA hybridization using carbon nanotube network field-effect transistors. Proc. Natl. Acad. Sci. United States Am. 103, 921-926 (2006).
4. Dong, X. et al. Label-Free Electronic Detection of DNA Using Simple Double-Walled Carbon Nanotube Resistors. J. Phys. Chem. C 112, 9891-9895 (2008).
5. Zhang, G. J. et al. DNA sensing by silicon nanowire: Charge layer distance dependence. Nano Lett. 8, 1066-1070 (2008).
6. Dong, X. et al. Electrical Detection of Femtomolar DNA via Gold-Nanoparticle Enhancement in Carbon-Nanotube-Network Field-Effect Transistors. Adv. Mater. 20, 2389-2393 (2008).
7. Liu, Y. et al. Detection of 12 Common Food-Borne Bacterial Pathogens by TaqMan Real-Time PCR Using a Single Set of Reaction Conditions. Front. Microbiol. 10, (2019).
8. Itahashi, M., Higaki, S., Fukuda, M. & Shimomura, Y. Detection and quantification of pathogenic bacteria and fungi using real-time polymerase chain reaction by cycling probe in patients with corneal ulcer. Arch. Ophthalmol. 128, 535-540 (2010).
9. Xu, G., Abbott, J. & Ham, D. Optimization of CMOS-ISFET-based biomolecular sensing: Analysis and demonstration in DNA detection. IEEE Trans. Electron Devices 63, 3249-3256 (2016).
10. Toumazou, C. et al. Simultaneous DNA amplification and detection using a pH-sensing semiconductor system. Nat. Methods 10, 641-646 (2013).
11. Duarte-Guevara, C. et al. On-chip electrical detection of parallel loop-mediated isothermal amplification with DG-BioFETs for the detection of foodborne bacterial pathogens. RSC Adv. 6, 103872-103887 (2016).
12. Graphene Field Effect Transistors for Biomedical Applications: Current Status and Future Prospects. Diagnostics 7, 45 (2017).
13. Rasheed, P. A. & Sandhyarani, N. Graphene-DNA electrochemical sensor for the sensitive detection of BRCA1 gene. Sensors Actuators, B Chem. 204, 777-782 (2014).
14. Ping, J., Vishnubhotla, R., Vrudhula, A. & Johnson, A. T. C. Scalable Production of High-Sensitivity, Label-Free DNA Biosensors Based on Back-Gated Graphene Field Effect Transistors. ACS Nano 10, 8700-8704 (2016).
15. Xu, G. et al. Electrophoretic and field-effect graphene for all-electrical DNA array technology. Nat. Commun. 5, (2014).
16. Hajian, R. et al. Detection of unamplified target genes via CRISPR-Cas9 immobilized on a graphene field-effect transistor. Nat. Biomed. Eng. 3, 427-437 (2019).
17. Hwang, M. T. et al. DNA Nanotweezers and Graphene Transistor Enable Label-Free Genotyping. Adv. Mater. 30, 1802440 (2018).
18. Hwang, M. T. et al. Highly specific SNP detection using 2D graphene electronics and DNA strand displacement. Proc. Natl. Acad. Sci. 113, 7088-7093 (2016).
19. Campos, R. et al. Attomolar Label-Free Detection of DNA Hybridization with Electrolyte-Gated Graphene Field-Effect Transistors. ACS Sensors 4, 286-293 (2019).
20. Justino, C. I. L., Gomes, A. R., Freitas, A. C., Duarte, A. C. & Rocha-Santos, T. A. P. Graphene based sensors and biosensors. TrAC - Trends in Analytical Chemistry 91, 53-66 (2017).
21. Liu, Y., Dong, X. & Chen, P. Biological and chemical sensors based on graphene materials. Chem. Soc. Rev. 41, 2283-2307 (2012).
22. Yang, W. et al. Carbon Nanomaterials in Biosensors: Should You Use Nanotubes or Graphene? Angew. Chemie Int. Ed. 49, 2114-2138 (2010).
23. Geim, A. K. & Novoselov, K. S. The rise of graphene. Nat. Mater. 6, 183-191 (2007)
24. Mao, S., Chang, J., Zhou, G. & Chen, J. Nanomaterial-enabled Rapid Detection of Water Contaminants. Small 11, 5336-5359 (2015).
25. Mao, S., Lu, G. & Chen, J. Nanocarbon-based gas sensors: Progress and challenges. J. Mater. Chem. A 2, 5573-5579 (2014).
26. Mao, S. Graphene Field-Effect Transistor Sensors. in Graphene Bioelectronics 113-132 (Elsevier Inc., 2018). doi:10.1016/B978-0-12-813349-1.00005-6
27. Zhan, B. et al. Graphene Field-Effect Transistor and Its Application for Electronic Sensing. Small n/a-n/a (2014). doi:10.1002/sm11.201400463
28. Chen, T. Y. et al. Label-free detection of DNA hybridization using transistors based on CVD grown graphene. Biosens. Bioelectron. 41, 103-109 (2013).
29. Cai, B. et al. Ultrasensitive label-free detection of PNA-DNA hybridization by reduced graphene oxide field-effect transistor biosensor. ACS Nano 8, 2632-2638 (2014).
30. Zheng, C. et al. Fabrication of Ultrasensitive Field-Effect Transistor DNA Biosensors by a Directional Transfer Technique Based on CVD-Grown Graphene. ACS Appl. Mater. Interfaces 7, 16953-16959 (2015).
31. Shoorideh, K. & Chui, C. 0. On the origin of enhanced sensitivity in nanoscale FET-based biosensors. Proc. Natl. Acad. Sci. U. S. A. 111, 5111-5116 (2014).
32. Wang, M. C. et al. Heterogeneous, Three-Dimensional Texturing of Graphene. Nano Lett. 15, 1829-1835 (2015).
33. Kang, P., Wang, M. C., Knapp, P. M. & Nam, S. Crumpled Graphene Photodetector with Enhanced, Strain-Tunable, and Wavelength-Selective Photoresponsivity. Adv. Mater. 28, 4639-4645 (2016).
34. Leem, J., Wang, M. C., Kang, P. & Nam, S. Mechanically Self-Assembled, Three-Dimensional Graphene-Gold Hybrid Nanostructures for Advanced Nanoplasmonic Sensors. Nano Lett. 15, 7684-7690 (2015).
35. Hwang, M. T. et al. Ultrasensitive detection of nucleic acids using deformed graphene channel field effect biosensors. Nat. Commun (UNDER Rev.
36. Hwang, M. T. et al. Ultrasensitive detection of nucleic acids using deformed graphene channel field effect biosensors. Nat. Commun. 11, 1543 (2020).
37. Li, M. et al. Pt/single-stranded DNA/graphene nanocomposite with improved catalytic activity and CO tolerance. J. Mater. Chem. A 3, 10353-10359 (2015).
38. Wei, G., Li, Q., Steckbeck, S. & Ciacchi, L. C. Direct force measurements on peeling heteropolymer ssDNA from a graphite surface using single-molecule force spectroscopy. Phys. Chem. Chem. Phys. 16, 3995-4001 (2014).
39. Banerjee, S. et al. Slowing DNA transport using graphene-DNA interactions. Adv. Funct. Mater. 25, 936-946 (2015).
40. Huang, P.-J. & Liu, J. Separation of Short Single- and Double-Stranded DNA Based on Their Adsorption Kinetics Difference on Graphene Oxide. Nanomaterials 3, 221-228 (2013).
41. Jansson, L. & Hedman, J. Challenging the proposed causes of the PCR plateau phase. Biomol. Detect. Quantif. 17, (2019).
42. Notomi, T. Loop-mediated isothermal amplification. Nippon rinsho. Japanese journal of clinical medicine 65, 957-961 (2007).
43. Ganguli, A. et al. Hands-free smartphone-based diagnostics for simultaneous detection of Zika , Chikungunya , and Dengue at point-of-care. 1-13 (2017). doi:10.1007/s10544-017-0209-9
44. Notomi, T. et al. Loop-mediated isothermal amplification of DNA. Nucleic Acids Res. 28, E63 (2000).
45. Wang, F., Jiang, L. & Ge, B. Loop-mediated isothermal amplification assays for detecting Shiga toxin-producing Escherichia coli in ground beef and human stools. J. Clin. Microbiol. 50, 91-97 (2012).
Materials and Methods:
Fabrication of Graphene FET: Graphene grown on copper film substrate was obtained from Grolltex and graphenesquare. Poly(methyl methacrylate) (PMMA) was obtained from Microchem. The high-quality graphene film on a grown copper film substrate was transferred and attached to a glass slide with double-sided adhesive. The top side of the graphene was spin-coated at 2000 rpm for 1 min, with a PMMA layer for protection from downstream copper etching as well as support for the graphene layer post etching. After, the spin coated PMMA/graphene sample was removed from the glass slide and the back side of the graphene was removed by oxygen plasma etching (at 100W for 13 min). Post etching, the sample was cut into 2 mm×14 mm pieces. Then, PMMA/graphene layer was delaminated from the copper film substrate using 0.1 M sodium persulfate (copper etchant) (Sigma Aldrich) for about 5 hours. The etched sample was incubated in deionized (DI) water overnight. The PMMA/graphene sample was then transferred onto a polystyrene sheet, after which the PMMA layer was removed by acetic acid for 10 min. The graphene on polystyrene sample was finally rinsed with DI water and dried with compressed air.
To fabricate the transistor, conducting silver paste (Ted Pella, Inc.) was used as source and drain electrodes at the two ends of the graphene on polystyrene. Silicone rubber (Dow Corning) was used to insulate source and drain electrodes from liquid and construct a solution reservoir for the LAMP sample or PBS. Before using for electrical measurement, the FET device is incubated with 1× PBS overnight.
DNA and Bacteria: Genomic DNA of Escherichia coli (O157:H7), NR-4629, was obtained through BEI Resources. These genomic DNA vials were aliquoted and stored at −80 C. Appropriate stock volumes were used either for diluted to the right concentration in buffer or water.
Primer Sequences
All primer sequences for the LAMP reactions were synthesized by Integrated DNA Technology (IDT). Primer sequences for E. coli eae gene were obtained from Wang et al45. Single-stranded DNA samples used in control measurements were the FIP primers (37 bases). The double stranded DNA sample contains complementary strands. See for the sequences of primers and DNA strands.
LAMP Reactions: LAMP assays were designed to target the eae gene for E. coli. The LAMP assay consisted of the following components: 1× final concentration of the isothermal amplification buffer (New England Biolabs), 1.025 mmol L−1 each of deoxy-ribonucleoside triphosphates (dNTPs), 4 mmol L−1 of MgSO4 (New England Biolabs), and 0.29 mol L−1 of Betaine (Sigma-Aldrich). These individual components were stored according to manufacturer instructions and a mix including all components was created fresh prior to each reaction. In addition to the buffer components, 0.47 U μL−1 Bst 2.0 WarmStart
DNA Polymerase (New England Biolabs), 1 mg/ml BSA (New England Biolabs), and 0.74× EvaGreen (Biotium), a double-stranded DNA (dsDNA) intercalating dye, was included in the reaction. 1× primer concentration in each reaction consisted of 0.15 μM of F3 and B3, 1.17 μM FIP and BIP, and 0.59 μM of LoopF and LoopB primers. 0.1× primer concentration in each reaction consisted of 0.015 μM of F3 and B3, 0.117 μM FIP and BIP, and 0.059 μM of LoopF and LoopB primers. 1 uL of template in water was added to make the final reaction volume 16 uL. For negative assays, water was substituted for the template. Ten-fold serial dilutions in water of the template were done as necessary for the LAMP assays.
All the LAMP reactions were carried out in 0.2 mL PCR reaction tubes in an Eppendorf Mastercycler® realplex Real-Time PCR System. The tubes were incubated at 65° C. for 60 min in the thermocycler, and fluorescent data were recorded every 1 min. Fluorescence data were recorded after each cycle of the reaction. Three repeats were done for each reaction.
Amplified products for each concentration were pooled together for electrical measurements. Unamplified products for negative control were also pooled together for electrical measurements.
The Zeptomolar reactions were designed for starting sample DNA concentrations of 8 zM, 40 zM, and 400 ZM. For 8 zM sample concentration of template DNA, 1.14 uL of DNA (2.91 copies) was added to 571.98 uL of water. For 40 zM sample concentration of DNA, 1.43 uL of DNA (3.63 copies) was added to 143 uL of water. For 400 zM sample concentration of DNA, 14.3 uL of DNA (36.3 copies) was added to 143 uL of water. After these starting samples were prepared, the LAMP assay at concentrations mentioned above were added to make the total reaction volume 1000 uL (for 8 zM reaction) or 250 uL (for 40 and 400 zM reactions). The reactions were carried out in 1.5 mL Eppendorf tubes in a digital heater at 65° C. for 60 min (Thermo Fisher Scientific).
Amplification Data Analysis: The off-chip raw fluorescence curves and amplification threshold bar graphs were analyzed using a MATLAB script and plotted using GraphPad Prism. The threshold time for each curve was taken as the time required for each curve to reach 10% of the total intensity. The amplification threshold bar graphs are show a mean of 3 samples.
Electrical Measurements and Analysis: I-V curves of gFET were measured in a semiconductor parameter analyzer equipped with a probe station (Company). Post overnight incubation with PBS (Thermo Fisher Scientific), the device was rinsed 3 times with 1× PBS, and finally, 50 uL of 1× PBS was added for measurement. Silver wire was used as an electrode, which applied gate voltage (Vg) to the PBS buffer solution. Vg was swept from 0 to 1 V in steps of 2 mV and drain—source voltage (Vds) was adjusted between 0.03 and 0.1 V depending on the resistance of the device. Drain-source current (Ids) for all sample measurements on the specific device was obtained at a constant Vds. Sweeping the Vg, Ids measurements were collected, and the Dirac voltage was found. The Dirac voltage is the minimum point in the resulting I-V curve of the gFET device.
Before measuring the LAMP samples, measurements of PBS solution were consistently done every 1 hr to find a stable Dirac voltage.
Measurement of DNA Samples: Before measuring the amplified and unamplified DNA samples from the LAMP reactions, each sample was diluted with 1× PBS (ranging from 1:100 to 1:1E5 dilution). In the cases for ssDNA and dsDNA adsorption tests, these DNA sample concentrations were luM in PBS.
Before measuring the amplified DNA samples, negative and positive control (pre-LAMP reaction) samples were electrically measured. Negative control sample was added on the device and incubated for 15 min, after which the device was rinsed three times with 1× PBS. After, 50 uL of 1× PBS was added on the device and sequential electrical measurements were conducted. The final measurement at which no changes in Dirac voltage was seen was taken as the final Dirac point, using which the Dirac Point Shift was calculated. The difference in Dirac voltage (or Dirac Point Shift) between the average and the Dirac voltage before addition of the sample was calculated. Thereafter, positive controls and amplified DNA samples were measured in the same protocol, and Dirac point shift due to physical adsorption of each sample was calculated from PBS to PBS measurements. After amplified DNA samples were measured, negative control samples were again measured to confirm that there is no change in the Dirac voltage.
Atomic Force Microscopy (AFM) for DNA and Graphene surface visualization: AFM images were acquired for all sample types to visualize the graphene surface morphology due to adsorbed biomolecules. All AFM images were acquired using a tapping mode Asylum Research MFP-3D AFM. Silicon cantilevers with a spring constant of 42 N/m (PPP-NCHR; Nanosensor) were used for imaging in air. The samples were incubated 15 min on the flat graphene, then rinsed with 1× PBS three times and DI water. The samples were dried using low pressure N2 gas. The MFP-3D AFM system software was used for analyzing image data. Max and min in the figures indicate the maximum and minimum height measured on the graphene surface, respectively. RMS, is the root mean square roughness of the graphene surface.
This example illustrates use of the devices and methods described herein to detect a variety of biomarkers.
Universal platforms for biomolecular analysis using label free sensing modalities can address a range of important diagnostic challenges in infectious disease, cancer, and other important areas. Electrical field effect sensors are an important class of devices that can enable point of care sensing by probing the intrinsic charge in biological entities. Use of graphene and especially curved or crumpled graphene for this application is especially promising. We have previously reported the lowest limit of detection (LoD) on electrical label-free field effect-based sensors using single or double stranded DNA molecules on the crumpled graphene FET platform. Here, we report field effect transistor-based biosensing of other important biomarkers including small molecules and proteins. We systematically evaluated and optimized the performance of these devices by studying the effect of the crumpling ratio on electrical double layer (EDL) formation and bandgap opening on the graphene. We show that a small and electroneutral molecule dopamine can be captured by an aptamer and the conformation change of the probe molecule induced electrical signal changes in the sensor. Three different kinds of proteins were captured with specific antibodies including interleukin-6 (IL-6) and viral proteins. All tested biomarkers were detectable with the highest sensitivity reported on a label-free electrical platform. Significantly, two COVID-19 related proteins, nucleocapsid (N-) and spike (S-) proteins antigens were successfully detected in PBS with extremely low LoDs. This label-free electrical antigen tests addresses the challenge of rapid, point of care diagnostics for infectious disease and other important clinical conditions.
An all electrical biosensor is of great interest, especially in the pandemic situation as these devices can be used for clinical diagnosis, point-of-care testing, and on-site detection[1]. The COVID-19 pandemic, amongst other healthcare challenges, has pointed attention to the importance of rapid and effective diagnosis of diseases[2].
FET-based biosensors allow label-free and highly sensitive biomolecular detection on integrated lab-on-a-chip systems. Detection of pH, nucleic acids, proteins and other biomarkers have been reported using FET biosensor with many different channel materials such as conventional silicon, nanowire and 2D materials. Among those, graphene is an attractive material due to its unique mechanical, chemical and electrical properties. CVD-grown graphene exhibits superior sensitivity in charge-based detection of biomolecules owing to its single-atom thickness and ultimate aspect ratio.
As illustrated in previous examples, a crumpled graphene FET-based biosensor is capable of detecting nucleic acids with ultra-high sensitivity, up to 18 molecules in 50 μL[3]. In this example, we demonstrate that the crumpled graphene can be universally used for different analytes and biomarkers across a range of sizes at an unprecedented concentration level (
Importantly, we successful detected COVID-19 related proteins, spike (S-) and nucleocapsid (n-) proteins with extremely low LODs.
Molecular dynamics (MD) Simulations: We investigate the effect of RNA and ionic species near flat and crumpled graphene surfaces to predict and explain the enhanced sensitivity of molecular detection from our experimental observations. As shown in
where F is the Faraday constant, z is the normal distance from graphene surface (z=0 on graphene) and σ is the surface charge density. As shown, because of the confined nature of the concave crumpled region, the ionic layer forms farther away from the concave graphene surface compared to the case of flat graphene. As the degree of crumpling increases, more ions are excluded from the graphene interface and the ionic screening takes place at a longer distance away from the graphene. With less of the RNA charge screened in highly crumpled graphene sheets, the RNA detection is enhanced. Hence, it should be noted that performance of crumpled graphene can be correlated with the crumpling ratio and systemically optimized for improved performance. The crumpling ratio was controlled by varying the annealing temperature (110˜115° C.) and time (10˜120 minutes).
To investigate the electronic structure of the deformed graphene, we perform density functional theory (DFT) calculations using the local density approximation (LDA)[34] and Perdew-Burke-Ernzerh (PBE)[35] functionals. Both LDA and PBE functionals produce bandgap values that are close to each other and show a local bandgap opening in the deformed graphene over all the crumpling range (10-60%) studied here. While the bandgap exhibits an oscillatory behavior as a function of the crumpling percentage, the overall bandgap opening upon crumpling has been shown to play an important role in increasing the graphene detection sensitivity[3].
Surface Characterization: Surface characterization of the crumpled graphene with different ratios was performed with AFM and SEM imaging, contact angle measurements, EDL capacitance measurements and Raman spectroscopy analysis (
Hydrophobicity heavily influences the formation of the EDL and it is directly related to Debye screening effect thus affect the sensitivity of the charge-based sensors[27]. The hydrophobicity of crumpled graphene samples with four different crumpling ratios is compared by contact angle measurements. It has already been shown that the hydrophobicity of MoS2 sheets on the pre-strained polystyrene substrate can be tuned by different crumpling ratio with the similar experimental schematics with this work[26]. In the previous study on MoS2, higher crumpling ratio provided higher hydrophobicity. Similar to this result, 60% crumpled graphene showed the highest hydrophobicity with the largest contact angle while 15% crumpling showed the smallest contact angle. EDL capacitance measurements results were also consistent with the contact angle data (
Because of increased hydrophobicity, crumpled graphene devices need ‘wetting process’ by repeating the I-V measurements over hours without adding target molecules. When the Dirac points are identical for at least two hours, the actual biomarker detection experiments can be initiated. We found that devices with higher crumpling required a longer time for the wetting process. Most of devices with 50% crumpled ratios required up to 4 hours of wetting process, while flat devices did not need any prior incubation time for wetting. When the crumpling ratio was 60%, some devices needed more than one day of wetting and even after the Dirac point being stable, 25% of the devices yielded reliable and consistent sensing results. These issues could be attributed to irreversible bonding of some ions with graphene with prolonged exposure to water[28]. Hence, increase in sensitivity of devices is also correlated with decreasing yield and we concluded that 50%-55% is an optimal crumpling ratio when considering the trade-offs between sensitivity, reproducibility and yield.
It is known that strain effect can open the bandgap of graphene. We have shown that the strain induced by crumpling process may locally open the bandgap of the graphene. The strain can be quantified by analyzing Raman G and 2D mode[23],[29]. It has been reported that the strain modulation of graphene by nanoscale substrate curvatures can be characterized by measuring shifts of G and 2D peaks[30]. Also, the rippled graphene can behave as a semiconductor due to the strain[31]. Moreover, larger strain may open larger bandgap[32]. Therefore, if increasing the crumpling ratio generates larger strain, the bandgap opening effect can be larger with higher crumpling ratio and this would contribute to larger exponential changes in the current from a small numbers of charges[33].
Typical raw spectra Raman peaks of crumpled graphene with different crumpling ratios are shown in
To clarify the physical distinction of highly crumpled graphene, 40% and 60% of crumpled graphene samples were imaged by atomic force microscopy (AFM) (
DNA absorption through 7E-7E stacking on the graphene FET was also investigated with varied crumpling ratios and electrical measurements (
Detection of Small Molecule Dopamine: We first examine the detection of an uncharged small molecule, dopamine, which is a neurotransmitter associated with motivational salience, can serve as a biomarker for the onset and progression of diverse diseases such as schizophrenia, Parkinson's disease, and several cancers such as neuroblastoma and pheochromocytoma[4]. Conventional monitoring methods of upregulation, downregulation or imbalance of dopamine are challenged by rigorous sample preparation to achieve the desired specificity and sensitivity. Precise detection of dopamine can be closely related to early diagnosis of neurological diseases, function tests of dopaminergic neurons derived from various stem cell sources and toxicity assessments of drugs[36]. FET-based dopamine detection can solve those limitations. Dopamine is electrically neutral however, a previous study showed that the neutral molecules can be detectable with conformational change of aptamer probes[9]. Graphene- and its nanocomposite-based electrical or electrochemical sensing of dopamine have been demonstrated previously however, as dopamine is electroneutral, the detections relied on redox or oxidation process of analytes, which lack specificity and sensitivity[5],[6],[7]. A few studies utilized aptamer as a probe molecule for aptamer detection and one report demonstrated Indium oxide FET-based dopamine detection at fM range of concentration with conformation change of aptamer probe when capturing the target molecules, which does not need complicated sample preparation process[8],[9]. Precise detection of dopamine can be closely related to early diagnosis of neurological diseases, function tests of dopaminergic neurons derived from various stem cell sources and toxicity assessments of drugs[36].
An aptamer is a single stranded DNA or RNA, when its conformation is changed, the distance of the overall charge from the negative backbone to the active channel surface can be changed and its electrical properties can be modulated. In this previous report where the same aptamer was used on an organic FET biosensor, the aptamer, which is negatively charged, became closer to the active channel surface when capturing dopamine[9]. This situation can be analogous to DNA capturing with probe molecules and it has been reported that the IV curve of the proposed graphene FET biosensor shifts to left when DNA capturing happens[3]. As seen in
Detection of IL-6 Protein: Larger size molecules, proteins such as interleukin-6 (IL-6) and viral protein were also detected on the crumpled graphene FET biosensor with the highest sensitivity ever reported herein[1],[10],[11]. IL-6 protein is a multipotent cytokine that plays an important role in immune responses, inflammation, bone metabolism, reproduction, arthritis, aging and neoplasia[12]. Monitoring the level of IL-6 protein can help diagnose many inflammatory diseases and cancers including sepsis. Concentration level of IL-6 in blood can grow up to ˜nM range[13]. However, several reports proposed advanced diagnostics in non-invasive manners from saliva, sweat or urine, and those solutions contain much lower concentration of IL-6, thus require highly sensitive sensing platform[14],[15].
We examine the detection of Interleukin 6 (IL-6) as shown in
Detection of SARS CoV-2 Antigens: There are three kinds of commercially available tests for COVID-19. These are; (i) detection of viral RNA genes based on nucleic acid amplification techniques are the gold standard for confirmation of COVID-19, (ii) Antigen tests which target the proteins on or inside the virus. The antigen tests do not need amplification and hence can significantly reduce the turnaround time but are inherently less sensitive as compared to the RNA tests, and (iii) antibody tests from blood samples which detect presence of antibodies and confirm immunity from the infection. There is an opportunity to improve the sensitivity of the antigen tests and use of label free electric sensors could play a role in this area. For SARS-CoV-2, the spike protein or the nucleocapsid protein can be captured with specific antibodies anchored on surfaces. Several antigen tests are available or in development by companies including Quidel, OraSure, Iceni Diagnostics, and E25Bio. These antigen tests are typically based on enzyme-linked immunoassay (ELISA) and have an optical readout. New tests typically require fluorescence labelling or nanoparticle anchoring to enhance the output signal. There are reports which have raised concerns about the performance of available antigen tests[16],[17]. For example, an analysis suggest that each 10-fold increase in LoD is expected to increase the false negative rate by 13%, missing an additional one in eight infected patients[17].
The crumpled graphene FET-based antigen test can be important as its sensitivity can be superior to ELISA and it does not require any labelling[18]. Such sensors can attempt to capture whole virus or the proteins after the virus lysis, the later can be safer from a user perspective. Recent reports have demonstrated label-free COVID-19 related protein detection in human nasopharyngeal swab specimens on graphene FET sensors[1].
We immobilized N- and S-protein antibodies on the graphene channels as reported previously[1]. Different kinds of coronavirus-related N-proteins detections were reported using nanowire FET in the pM concentrations[37],[38], however an improvement in sensitivity desired if antgen tests were to replace the RNA molecular detection tests[17]. Moreover, COVID-19 N-protein has not been tested on any kinds of FET based biosensor before this report.
The isoelectric point of the N-protein is pH˜10 and thus the overall charge would be positive in 1× PBS. However, previous studies showed that some parts of the SARS N-protein, of which the isoelectric point is also ˜10, are negatively charged. It might be possible that the positive charges on active channel surface attract the negatively charged regions of N-protein captured on the surface, thus leading to conductance change as a result of the local negative charges of the antigens at the surface[37],[38]. This result was supported by another literature with computational simulations, which shows that N-protein has an asymmetric charge distribution. Depends on its ‘up’ or ‘down’ orientation, the protein would generate different charge signals at pH=7.4. It was concluded that, with ‘up orientation’, the N-protein is expected as negative charges for Debye screening lengths up to 3 nm[40]. Even though these previous results were based on SARS N-protein and not the SARS-CoV-2 N-protein, their isoelectric points and molecular weights are very similar, i.e. pH˜10 and 46 kD, respectively. Furthermore, it was recently confirmed that many aspects of SARS N-protein and COVID-19 N-protein are similar and they show more than 90% sequence similarity[41]. Thus it is reasonable that electrical charge effect of COVID-19 n-protein on graphene surface could be either positive or negative depends on the orientation[42],[43].
The S-protein was also tested on the crumpled graphene FET (
In summary, detection of several important disease-related biomarkers including dopamine, IL-6 protein and viral proteins were demonstrated on the crumpled graphene FET biosensor. The effect of various crumpling ratio on the detection sensitivity was studied guided and confirmed by computational simulations, Raman spectroscopy analysis, and EDL capacitance measurements. The optimized crumpled graphene FET biosensors were able to detect various size of biomarkers with unprecedented sensitivities. Importantly, we also demonstrated SARS CoV-2 antigen test by detecting S- and N-proteins with the crumpled graphene FET biosensor in a label-free format that could result in a diagnostic tool with small footprint. This platform can overcome the limitation of currently available antigen tests owing to its superior sensitivity due to modulation of the Debye layer on the crumpled graphene biosensors. In the future, this platform can be highly multiplexed and can target many multiple biomarkers on a chip and have a significant impact on the diagnostics market.
Methods
MD simulations: Molecular dynamics simulations were performed using the LAMMPS package3. The systems were generated by the Visual Molecular Dynamics (VMD)4. As mentioned in the manuscript, different graphene topologies with crumpling degrees of 0% (flat), 10%, 30%, 50% and 70% are used. Each simulation box consists of a single-layer graphene sheet, a single-stranded COVID-19 RNA, water and ions. The COVID-19 RNA segment has 20 bases with a sequence of GAC CCC AAA ATC AGC GAA AT (SEQ ID NO:18). A concentration of 1.2M Sodium chloride is considered. Depending on the degrees of crumpling, the simulation systems contain 20,000 to 80,000 atoms. Periodic boundary conditions are applied in the x and y directions (projected plane of graphene lies in the xy plane). The systems are non-periodic in z direction. The CHARMM forcefield5 is used. The Lennard-Jones (LJ) potential with a cutoff distance of 1.2 nm is employed. The long—range electrostatic interactions are calculated using the PPPM6.
First, the energy of each system was minimized for 15,000 steps. Equilibrium simulations were then performed in NPT ensemble for 2 ns at a pressure of 1 atm and a temperature of 300 K. The NPT simulation ensures that the water reaches its equilibrium density. The systems were further equilibrated in NVT ensemble for another 2ns at 300 K. Temperature was maintained at 300 K by using the Nose-Hoover thermostat7,8 with a time constant of 0.1 ps. The carbon atoms of graphene are kept frozen. The production simulations were carried out in NVT ensemble for 10 ns. Trajectories of atoms were dumped every picosecond to obtain the structure of DNA and ions near graphene sheets.
Atomic Force Microscopy imaging: AFM images were recorded using an ASYLUM RESEARCH MFP-3D AFM SYSTEM (Asylum Research, Santa Barbara, Calif.).
Raman Spectroscopy imaging: The Raman spectra and imaging of the crumpled graphene were recorded using a Nanophoton RAMAN-11 laser confocal microscope (Nanophoton, Osaka, Japan). A 532 nm diode laser was used for excitation.The excitation power was 0.1 mW with 3 s exposure time and 3 times averaging for point and mapping, respectively. In Raman mapping mode, region of interest was 2.2 by 2.2 um in x, y axis with 200 nm/pixel resolution with NA 0.9 100× Plan Fluor objective lens. The grating was 600 gr/mm. The wave number range covered was 700-2900 cm−1. The wavenumber shift compensation was −8.4 cm−1 after initial Ne-sample calibration. The Raman signals were detected by a Peltier cooled CCD camera at −70° C.
Example 3 References:
All references throughout this application, for example patent documents including issued or granted patents or equivalents; patent application publications; and non-patent literature documents or other source material; are hereby incorporated by reference herein in their entireties, as though individually incorporated by reference, to the extent each reference is at least partially not inconsistent with the disclosure in this application (for example, a reference that is partially inconsistent is incorporated by reference except for the partially inconsistent portion of the reference).
The terms and expressions which have been employed herein are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments, exemplary embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims. The specific embodiments provided herein are examples of useful embodiments of the present invention and it will be apparent to one skilled in the art that the present invention may be carried out using a large number of variations of the devices, device components, methods steps set forth in the present description. As will be obvious to one of skill in the art, methods and devices useful for the present methods can include a large number of optional composition and processing elements and steps.
As used herein and in the appended claims, the singular forms “a”, “an”, and “the” include plural reference unless the context clearly dictates otherwise. Thus, for example, reference to “a cell” includes a plurality of such cells and equivalents thereof known to those skilled in the art. As well, the terms “a” (or “an”), “one or more” and “at least one” can be used interchangeably herein. It is also to be noted that the terms “comprising”, “including”, and “having” can be used interchangeably. The expression “of any of claims XX-YY” (wherein XX and YY refer to claim numbers) is intended to provide a multiple dependent claim in the alternative form, and in some embodiments is interchangeable with the expression “as in any one of claims XX-YY.”
When a group of substituents is disclosed herein, it is understood that all individual members of that group and all subgroups, are disclosed separately. When a Markush group or other grouping is used herein, all individual members of the group and all combinations and subcombinations possible of the group are intended to be individually included in the disclosure.
Every device, system, formulation, combination of components, or method described or exemplified herein can be used to practice the invention, unless otherwise stated.
Whenever a range is given in the specification, for example, a size range, a volume range, a number range, a temperature range, a time range, or a composition or concentration range, all intermediate ranges and subranges, as well as all individual values included in the ranges given are intended to be included in the disclosure. It will be understood that any subranges or individual values in a range or subrange that are included in the description herein can be excluded from the claims herein.
All patents and publications mentioned in the specification are indicative of the levels of skill of those skilled in the art to which the invention pertains. References cited herein are incorporated by reference herein in their entirety to indicate the state of the art as of their publication or filing date and it is intended that this information can be employed herein, if needed, to exclude specific embodiments that are in the prior art. For example, when composition of matter are claimed, it should be understood that compounds known and available in the art prior to Applicant's invention, including compounds for which an enabling disclosure is provided in the references cited herein, are not intended to be included in the composition of matter claims herein.
As used herein, “comprising” is synonymous with “including,” “containing,” or “characterized by,” and is inclusive or open-ended and does not exclude additional, unrecited elements or method steps. As used herein, “consisting of” excludes any element, step, or ingredient not specified in the claim element. As used herein, “consisting essentially of” does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim. In each instance herein any of the terms “comprising”, “consisting essentially of” and “consisting of” may be replaced with either of the other two terms. The invention illustratively described herein suitably may be practiced in the absence of any element or elements, limitation or limitations which is not specifically disclosed herein.
One of ordinary skill in the art will appreciate that starting materials, biological materials, reagents, synthetic methods, purification methods, analytical methods, assay methods, and biological methods other than those specifically exemplified can be employed in the practice of the invention without resort to undue experimentation. All art-known functional equivalents, of any such materials and methods are intended to be included in this invention. The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention that in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.
This application claims the benefit of and priority to U.S. provisional patent application Nos. 62/982,801, filed Feb. 28, 2020; 63/029,136 filed May 22, 2020; and 63/054,039 filed Jul. 20, 2020, each of which is incorporated by reference herein in its entirety, except to the extent inconsistent herewith.
This invention was made with government support under Award Number DMR-1720633 awarded by the National Science Foundation. The government has certain rights in the invention.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2021/020006 | 2/26/2021 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 62982801 | Feb 2020 | US | |
| 63029136 | May 2020 | US | |
| 63054039 | Jul 2020 | US |
