WO2016205190A1

WO2016205190A1 - Affinity nanosensor for detection of low-charge and low-molecular-weight molecules

Info

Publication number: WO2016205190A1
Application number: PCT/US2016/037362
Authority: WO
Inventors: Qiao Lin; Yibo Zhu; Junyi SHANG; Zhixing Zhang
Original assignee: The Trustees Of Columbia University In The City Of New York
Priority date: 2015-06-16
Filing date: 2016-06-14
Publication date: 2016-12-22

Abstract

Techniques for detecting and measuring low-charge, low-molecular -weight analytes using a MEMS affinity sensor is provided. The sensor can use one or more materials with a surface-immobilized synthetic polymer that is sensitive to and can bind with an analyte. Techniques include graphene-based affinity binding and hydrogel-based affinity binding. Binding by the analyte to the polymer induces measurable changes in the electrical properties of the material for detection of target molecules while reducing the need for movable, mechanical components in sensor design.

Description

AFFINITY NANOSENSOR FOR DETECTION OF LOW-CHARGE AND LOW-MOLECULAR-WEIGHT MOLECULES

STATEMENT REGARDING FEDERALLY FUNDED RESEARCH

This invention was made with government support under 1DP3 DK101085-01 awarded by the National Institutes of Health. The U.S. government has certain rights in this invention.

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to U.S. Provisional Application Serial No. 62/180,484, filed on June 16, 2015, and U.S. Provisional Application Serial No. 62/188,281, filed on July 2, 2015, each of which is hereby incorporated by reference in its entirety.

BACKGROUND

Diabetes affects millions of people worldwide, including in the United States. Diabetes creates abnormal blood sugar levels, such as hyperglycemia (abnormally high blood sugar level) and hypoglycemia (abnormally low blood sugar level), which can require afflicted individuals to monitor and regularly measure their blood glucose levels. Certain approaches to blood-based glucose monitoring can involve extracting blood such as by intermittent finger-stick testing, or continuous glucose monitoring, and can have drawbacks. For example, apart from the invasive and sometime painful extraction of blood, finger-stick glucose monitoring can miss abnormal blood excursions. Continuous glucose monitoring, while able to monitor glucose levels throughout the day via electrochemical detection, can suffer from interferences from electroactive chemicals. MEMS (micro-electro-mechanical systems) technology can enable innovative subcutaneously implanted sensors to measure concentrations of certain molecules, such as glucose, in interstitial fluid (ISF). Such sensors can employ affinity binding between the target molecules and a sensing material to achieve high accuracy and stability. Affinity sensors that are based on dielectric measurements have been used in applications such as detecting or quantifying biochemical targets under excitations at various frequencies. However, such microsensors utilizing dielectric measurements can lack sufficient sensitivity for the detection of certain low-charged and low-molecular-weight molecules.

Additionally, MEMS devices can be limited by complicated and inefficient designs or slow time responses. For example, affinity glucose sensing has been implemented using optical, mechanical, and electrical methods on conventional or microscale platforms typically requiring complex sensor structures such as moving mechanical components or physical barriers. Semi-permeable membranes or other physical barriers or mechanically movable structures can increase complexity of the devices and limit the reliability of the devices.

Graphene is a single a single atom thick two-dimensional nanomaterial with honeycomb lattice of carbon. While graphene can be attractive functional nanomaterial in sensors that allow highly sensitive detection of chemical and biological analytes, analytes and like molecules detectable by such graphene-based sensors can be highly charged or strong electron donors or acceptors that can induce carrier doping in graphene for field effect transistor (FET)-based measurements. In particular, graphene can form a conducting channel in field effect transistors (FETs), allowing sensitive electrically based detection of gas molecules, physiological parameters of liquids (e.g., pH level) and biological molecules (e.g., proteins) in solution. Glucose, however, is an uncharged, low-molecular-weight molecule. Sensitive detection of glucose has been accomplished within graphene FET-based enzymatic sensors. Unfortunately, due to the irreversible, consumptive nature of the enzyme-catalyzed electrochemical reactions of glucose, as well as undesirable byproducts (e.g., hydrogen peroxide) generated in the reactions of enzymes and glucose, certain enzyme-based sensors can suffer from limitations in stability and accuracy when operating in physiological environments.

Subcutaneous detection of certain low-charged molecules, low- molecular weight molecules, which can be important for applications such as glucose monitoring, thus still remains a challenge. Accordingly, a need exists for an accurate, subcutaneous sensor reliable for continuous monitoring of low-charged molecules in a physiological environment.

SUMMARY

The disclosed subject matter provides a synthetic polymer- functionalized affinity -based nanosensor for detection of low-charge, low-molecular- weight molecules, such as, for example and without limitation, glucose. The disclosed subject matter can utilize a material functionalized with a synthetic polymer monolayer derivatized with a boronic acid group whose reversible complexation with the target low-charge, low-molecular-weight molecule (e.g., glucose) generates a detectable signal.

In one aspect of the disclosed subject matter, a graphene-based affinity nanosensor is provided. The binding of the polymer monolayer with glucose on the graphene surface of the device can induce changes in the carrier density and mobility in the graphene, which can cause an increase in charge on the graphene. Thus, the binding can offer a high detection sensitivity of the target molecule and/or analyte.

For example and without limitation, an affinity nanosensor for detection of low-charge, low-molecular-weight molecules includes a solution-gated field effect transistor, which can enable reliable monitoring of a target molecules in a sample solution. The solution-gated field effect transistor can include a silicon substrate, a source electrode disposed on the silicon substrate, and a drain electrode disposed on the silicon substrate. The graphene can be a graphene sheet, which can be disposed between the source electrode and drain electrode, and can connect the source and drain electrodes. The solution-gated field effect transistor can further include graphene functionalized with a synthetic polymer monolayer, which is disposed between the source electrode and drain electrode on the silicon substrate. The functionalized graphene forming a conducting channel of the solution-gated field effect transistor, and the synthetic polymer monolayer being responsive to a first analyte. The affinity nanosensor can also include a reference electrode disposed between the source and drain electrodes, and an electrical double layer at the interface of the graphene and solution comprising a gate capacitor.

Additionally, or alternatively, the silicon substrate can be an oxidized silicon substrate wafer. The source electrode and drain electrode can be gold electrodes. The synthetic polymer monomer can include a boronic acid. The graphene can be functionalized with the synthetic polymer monolayer via π-π stacking interactions. The first analyte can be glucose. The reference electrode can include silver chloride. Furthermore, the gate capacitor can be affected by varying concentrations of the first analyte in the sample solution.

The graphene-based affinity nanosensor can include any or all of the features described herein.

According to another aspect of the disclosed subject matter, methods of fabricating an affinity nanosensor for detecting low-charge, low-molecular-weight molecules are provided. An example method includes providing a silicon substrate wafer having a uniform thickness throughout, the wafer having oppositely disposed top and bottom faces; and providing a first gold portion on the top face of the first material, and a second gold portion separated from the first gold portion by a channel region. The method can further include transferring graphene onto the top face of the wafer in the channel region to connect the first and second gold portions, and functionalizing the graphene with a synthetic polymer monolayer, the synthetic polymer monolayer being sensitive to a first analyte. The method can also include mounting a conductive wire within the channel region.

Additionally, or alternatively, the providing the first and second gold portions can include etching the first gold portion and the second gold portion to the top face of the wafer. The transferring the graphene can include coupling graphene to the top face of the wafer via chemical vapor deposition. Furthermore, the functionalizing can include immersing at least the graphene transferred onto the wafer in a solution comprising boronic acid for at least four hours at room temperature, and washing the graphene transferred onto the wafer using methanol. The immersing can include coupling a pyrene-1 -boronic acid to the graphene via π-π stacking interactions. The mounting the conductive wire within the channel region can include providing a silver wire mounted on a positioner to serve as a gate electrode.

In another aspect of the disclosed subject matter, a hydrogel-based affinity nanosensor is provided. For example, an affinity nanosensor can include a parallel plate transducer, a synthetic hydrogel disposed between a first plate and a second plate of the parallel plate transducer, the hydrogel being responsive to a first analyte, and a temperature sensor located below the first and second plates.

Additionally, or alternatively, the first and second plates of the parallel plate transducer each further include a sensing electrode. The sensing electrode can be formed of gold. At least one of the first plate and second plate of the parallel plate transducer can be perforated and passivated within a perforated diaphragm. The at least one perforated plate and at least one perforated diaphragm can be supported by at least one micropost. The synthetic hydrogel of the affinity nanosensor can further include a synthetic copolymer including boronic acid. The first analyte can be glucose.

The hydrogel-based affinity nanosensor can include any or all of the features described herein.

According to yet aspect of the disclosed subject matter, an example method for fabricating an affinity nanosensor for detecting low-charge, low- molecular-weight molecules includes providing a silicon substrate wafer having a uniform thickness throughout and having oppositely disposed top and bottom faces, providing a first electrode on the top face of the wafer, and providing a second electrode, spaced a first distance over the first electrode, above the top face of the wafer, the second electrode being supported above the top face of the wafer by at least one micropost. The method can also include preparing a hydrogel functionalized with a polymer responsive to a first analyte, and filling the hydrogel between the first and second electrodes.

The method can also include providing the second electrode with one or more perforations, and separating the second electrode from the hydrogel by a perforated diaphragm. The first analyte can be glucose. Furthermore, the preparing the functionalized hydrogel can include synthesizing the hydrogel in situ via polymerization of the hydrogel with a boronic acid, and gelating the hydrogel between the first and second electrodes.

As herein disclosed, a sample solution containing such low-charge, low molecular-weight molecules can be a bodily fluid, a non-bodily fluid, or a laboratory sample. The bodily fluid can be, for example and without limitation, tears, blood, saliva, mucus, ISF (interstitial fluid), amniotic fluid, lymphatic fluid, pericardial fluid, peritoneal fluid, pleural fluid, sweat, or other bodily fluid of a subject.

The accompanying drawings, which are incorporated in and constitute part of this specification, are included to illustrate and provide a further understanding of the disclosed subject matter. Together with the description, the drawings serve to explain the principles of the disclosed subject matter.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 depicts glucose sensing via affinity binding in accordance with an exemplary embodiment of the disclosed subject matter.

FIG. 2A depicts a schematic representation of a graphene-based affinity nanosensor in accordance with an exemplary embodiment of the disclosed subject matter.

FIG. 2B depicts another schematic representation of a graphene-based affinity nanosensor in accordance with an exemplary embodiment of the disclosed subject matter.

FIG. 3 depicts a synthesis of the pyrene-terminated glucose-sensing polymer PAPBA and its coupling to graphene via π-π stacking interactions in accordance with an exemplary embodiment of the disclosed subject matter.

FIGS. 4A-4C depict fabrication of the nanosensor in accordance with an exemplary embodiment. FIG. 4A depicts patterning of drain and source electrodes. FIG. 4B depicts transfer of graphene onto an oxide-coated silicon substrate. FIG. 4C depicts bonding of the PDMS microchannel to the nanosensor chip. FIG. 5 depicts the graphene conducting channel connecting the source and drain electrodes in accordance with an exemplary embodiment.

FIG. 6 depicts a measurement setup of the nanosensor in accordance with an exemplary embodiment.

FIG. 7 depicts the Raman spectrum of graphene in accordance with an exemplary embodiment. The G and 2D bands are indicative of the graphene consisting of a single layer of carbon atoms.

FIGS. 8A-8B depict AFM images of graphene before and after functionalization in accordance with an exemplary embodiment. FIG. 8A depicts graphene before functionalization. FIG. 8B depicts graphene after functionalization with a PAPBA polymer.

FIG. 9 depicts transfer characteristics measured before (dashed line) and after (solid line) functionalization of graphene with the PAPBA polymer. The left shift of the Dirac point indicates that the graphene was n-doped due to the attachment of the polymer molecules.

FIG. 10 depicts transfer characteristics in different glucose solutions at varying glucose concentrations in accordance with an exemplary embodiment. In response to increases in the glucose concentration, the Dirac point position, Fcs.Dirac shifted to higher gate voltages and the transconductance descreased from 100 to 20 μ$.

FIG. 11 depicts control experiments using pristine graphene without functionalization of the polymer. The change in the Dirac point position and transconductance is insignificant compared to the embodiment of FIG. 9, and indicates that the changes in carrier mobility and density of FIG. 9 were caused by the glucose-polymer binding. FIG. 12 depicts coupling of boronic acid and graphene via π-π stacking interactions between the pyrene group and graphene in accordance with an exemplary embodiment.

FIG. 13 depicts formation of a glucose-boronate ester at a physiological pH of 7.4 in accordance with an exemplary embodiment.

FIGS. 14A-14B depict properties of the graphene before and after functionalization in accordance with an exemplary embodiment. FIG. 14A depicts transfer characteristics of pristine graphene and PBA-functionalized graphene, and transfer characteristics of the pristine graphene exposed to glucose solutions (0.1 mM to 25 mM). Transfer characteristics after rinsing with PBA solution show that FNP shifted from 0.33 V to 0.575 V. FIG. 14B depicts the Raman spectra of the graphene before and after exposure to PBA solution. Signature peaks of the boronic acid and the graphene-pyrene interaction were observed after immersing in PBA solution.

FIG. 15 depicts transfer characteristics measured when the nanosensor was exposed to glucose solutions (concentration ranging from 2 μιη to 25 mM) in accordance with an exemplary embodiment. The curve shifted to the left as a result of the increase in the glucose concentration, i.e., a monotonic decrease of 7_Ds at VGS ⁼ 0.4 V.

FIG. 16 depicts neutral point shift ratio AFNP,G/AJ¾>,B as a function of glucose concentration in accordance with an exemplary embodiment. Glucose concentration is on a logarithmic scale. The inset of FIG. 16 depicts a fit to the Hill- Langmuir equation, yielding an equilibrium dissociation constant (Κ_Ό) of 38.6 μηι.

FIG. 17A-17B depict synthetic glucose-affinity hydrogel sensing in accordance with an exemplary embodiment. FIG. 17A depicts the reversible affinity binding of PHEAA-ra«-PAAPBA integrated hydrogel to glucose. FIG. 17B depicts hydrogel embedded in a capacitive transducer.

FIG. 18 depicts a schematic of a sensor chip with coplanar electrodes in accordance with an exemplary embodiment.

FIGS. 19A-19C depict fabrication of a sensor chip in accordance with an exemplary embodiment. FIG. 19A depicts deposition of a gold/chrome layer on a substrate. FIG. 19B depicts gold patterning. FIG. 19C depicts hydrogel integration.

FIGS. 20A-20D depict an impedance/voltage transformation circuit driven by a sinusoidal input from a function generator connected to a sensor in accordance with an exemplary embodiment. FIG. 20A depicts a sensor chip before hydrogel integration. FIG. 20B depicts a sensor after hydrogel integration. FIG.

20C depicts a measurement setup in accordance with the above. FIG. 20D depicts an impedance/voltage transformation circuit.

FIGS. 21A-21D depict a hydrogel's dielectric relaxation in accordance with an exemplary embodiment. FIG. 21A depicts an effective capacitance as a function of frequency response without glucose. FIG. 21B depicts an effective capacitance as a function of frequency response with glucose. FIG. 21C depicts an effective resistance as a function of frequency response without glucose. FIG. 21D depicts an effective resistance as a function of frequency response with glucose.

FIG. 22 depicts a time-resolved effective capacitance at 30 kHz in response to changes in glucose concentration in accordance with an exemplary embodiment.

FIGS. 23A-23B depict a sensor's output as a function of glucose concentration in accordance with an exemplary embodiment. FIG. 23A depicts a sensor's effective capacitance as a function of glucose concentration. FIG. 23B depicts a sensor's effective resistance as a function of glucose concentration.

FIG. 24 depicts a hydrogel-based microsensor in accordance with an exemplary embodiment.

FIGS. 25A-25B depicts schematics of an affinity microsensor in accordance with an exemplary embodiment. FIG. 25A depicts a top view of a schematic of an affinity microsensor. FIG. 25B depicts a side view of a schematic of an affinity microsensor.

FIGS. 26A-26D depict sensor chip fabrication in accordance with an exemplary embodiment. FIG. 26A depicts standard fabrication procedures. FIG. 26B depicts an image of a fabricated capacitive transducer. FIG. 26C depicts hydrogel integration in a capacitive transducer. FIG. 26D depicts and image of a hydrogel-integrated sensor chip.

FIGS. 27A-27D depict an experimental setup for testing a sensor in accordance with an exemplary embodiment. FIG. 27A depicts a schematic of a testing setup. FIG. 27B depicts an image of a testing setup. FIG. 27C depicts an example setup. FIG. 27D depicts a capacitance/voltage transformation circuit.

FIGS. 28A-28B depict measurements of glucose concentration using a microsensor in accordance with an exemplary embodiment. FIG. 28A depicts dependence of the effective capacitance on measurement frequency. FIG. 28B depicts dependence of the effective capacitance on glucose concentration. Note that exemplary effective capacitance values depicted herein are averages of triplicate measurements, and standard errors are shown as error in FIG. 28B.

FIGS. 29A-29B depict dependence of effective capacitance on glucose concentration in accordance with an exemplary embodiment. FIG. 29A depicts a ratio of interferent-induced effective capacitance change to glucose-induced capacitance change (concentration: 90 mg/dL for glucose and each of the interferents including fructose, galactose, ascorbic acid, and lactate). FIG. 29B depicts sensor response to glucose when bornoic acid components are absent in the hydrogel. Note the bias voltage frequency of the embodiments depicted in FIGS. 29A-29B are 30 kHz.

FIG. 30 depicts the time-resolved device response to time-varying glucose concentration with a bias voltage frequency of 30 kHz in accordance with an exemplary embodiment.

The accompanying figures, where like reference numerals refer to identical or functionally similar elements throughout separate views, serve to further illustrate various embodiments and to explain various principles and advantages all in accordance with the disclosed subject matter. For purpose of explanation and illustration, and not limitation, exemplary aspects and embodiments of the device are shown in FIGS. 1-30.

DETAILED DESCRIPTION

As used herein, the terms "device," "sensor," and "nanosensor" are interchangeable and used here as a reference to low-charge, low-molecular- weight affinity nanosensor herein disclosed. Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art to which the disclosed subject matter belongs. Although methods and materials similar or equivalent to those described herein can be used in its practice, suitable methods and materials are described below.

It is to be noted that the term "a" entity or "an" entity refers to one or more of that entity. As such, the terms "a", "an", "one or more", and "at least one" can be used interchangeably herein. The terms "comprising," "including," and "having" can also be used interchangeably. In addition, the terms "amount" and "level" are also interchangeable and can be used to describe a concentration or a specific quantity. Furthermore, the term "selected from the group consisting of refers to one or more members of the group in the list that follows, including mixtures (i.e., combinations) of two or more members.

The term "about" or "approximately" means within an acceptable error range for the particular value as determined by one of ordinary skill in the art, which will depend in part on how the value is measured or determined, i.e., the limitations of the measurement system. For example, "about" can mean within 3 or more than 3 standard deviations, per the practice in the art. Alternatively, "about" can mean a range of up to +/-20%, up to +/-10%, up to +1-5%, or alternatively up to +/-1% of a given value. Alternatively, with respect to biological systems or processes, the term can mean within an order of magnitude, preferably within 5 -fold, and more preferably within 2-fold, of a value.

As used herein, the term "analyte" is a broad term and is used in its ordinary sense and includes, without limitation, any chemical species the presence or concentration of which is sought in material sample by the sensors and systems disclosed herein. For example, the analyte(s) include, but not are limited to, glucose, ethanol, insulin, water, carbon dioxide, blood oxygen, cholesterol, bilirubin, ketones, fatty acids, lipoproteins, albumin, urea, creatinine, white blood cells, red blood cells, hemoglobin, oxygenated hemoglobin, carboxyhemoglobin, organic molecules, inorganic molecules, pharmaceuticals, cytochrome, various proteins and chromophores, microcalcifications, electrolytes, sodium, potassium, chloride, bicarbonate, and hormones. In one embodiment, the analyte is glucose. In various embodiments, the analytes can be other metabolites, such as lactate, fatty acids, cysteines and homocysteines.

The presently disclosed subject matter provides an affinity -based nanosensor. As disclosed herein, affinity binding is employed for monitoring and measuring low-charge and/or low-molecular-weight analytes. Affinity binding can be specific and reversible. In other words, affinity binding can be specific when analytes or target molecules bind with analyte-specific, or target molecule-specific receptors, which do not bind with interferents that may also come into contact with such receptor. Additionally, or alternatively, affinity binding can be reversible when the analytes and/or target molecules can be released from the receptor, as depicted in exemplary FIG. 1. Such reversible affinity binding can avoid consumption of the analytes and/or target molecule and result in low drift, stable and accurate measurement. For example, and without limitation, the affinity interaction of a glucose responsive polymer (receptor), such as boronic acid, with glucose (analyte), can eliminate interference of electroactive species when used in MEMS sensing technology.

In one aspect of the presently disclosed subject matter, a graphene- based nanosensor for affinity -based detection of low-charge molecules is provided. A low-charge molecule can be a molecule, or analyte, that is substantially uncharged, such as, for example, glucose and other sugars. Additionally, or alternatively, the graphene-based nanosensor herein disclosed can accomplish affinity-based detection of low-molecular-weight molecules. A low-molecular-weight molecule, or analyte, can be, for example and without limitation, a molecule or organic compound that can regulate a biological process with a size on the order of 10^"9 m, such as glucose. By way of example, and not limitation, the nanosensor can be configured as a solution- gated graphene-based field effect transistor (GFET), as depicted in exemplary FIGS. 2A-2B. A solution-gated GFET can detect analytes by transducing the binding of such analytes at the graphene surface to a change in current-voltage relationships between source and drain electrodes. The graphene 101 can be the conducting channel, formed between the two electrodes, representing the source 102 and the drain 103, on an insulating substrate 104, as illustrated in exemplary FIGS. 2A-2B. A layer of synthetic polymer, such as a glucose responsive polymer, including for example, a boronic acid attached to pyrene, can be coupled to the graphene surface via π-π stacking interactions at 105. Such π-π stacking interactions are generally depicted in exemplary FIG. 3. A conducting material 107, such as a silver wire, can be inserted within the microchannel 106 above the functionalized graphene to serve as the gate electrode 107 of the GFET described herein. The microchannel 106 can provide a path for a solution containing the target molecules and/or analytes to come into contact with immobilized glucose responsive polymer 105.

During operation, a voltage can be applied between the drain 103 and source 102 electrodes to generate a current in the graphene that can be measured. A bias voltage can be applied to the gate electrode 107. Further to the above, and as disclosed herein, when no voltage is applied to the gate electrode 107, the resistance along the graphene microchannel 106 can be about zero. However, when a bias voltage is applied to the gate electrode, the resistance within the microchannel 106 can increase, and can result in a functional dependence of the drain-source current on the gate electrode voltage, which can represent the transfer characteristics of the GFET. Thus, the transfer characteristics of the GFET can be affected by glucose, or other target molecule and/or analyte, contacting and binding with the glucose- responsive polymer in the microchannel, which can be measured to determine the concentration of glucose or other target molecule and/or analyte.

For example, and without limitation, upon contacting the graphene- functionalized microchannel 106, analytes can cause detectable changes in the electrical properties of the graphene as a result of the binding of boronic acid moieties of the polymer in FIGS. 2A-2B. Such polymer-glucose binding can change the position of the Dirac point, or the value of the gate voltage at which charge carriers neutralize and the drain-source current achieves its minimum. Additionally, or alternatively, the carrier density of the solution can vary as a result of the electron exchanges between the graphene and the target solution. For example, and not limitation, cyclic esters of boronic acid can form as a result of the binding of boronic acid groups to glucose molecules, which causes an overall ionization equilibrium shift from neutral/insoluble boronic acid moieties to anionic/hydrophilic boronate (see, e.g., FIG. 13). Thus, the charge density in the solution can change, which can in turn change the carrier density and alter the Fermi level of the graphene. Example results disclosed herein demonstrate that the detection of glucose in a concentration range of 0 to 200 mg/dL can be measured with a sensitivity of approximately 2.5 mV/(mg/dL), indicating a potential for blood glucose monitoring and control in diabetes care.

Additionally, and further to the above, due to its vanishing bandgap and high mobility, the bipolar transfer characteristics of graphene can exhibit definitive shifts upon glucose-boronic acid binding. Such shifts can reflect affinity binding-induced charge transfer to graphene, or changes in the electrostatic potential in the immediate proximity of the graphene, thereby allowing for insights into the underlying physiochemical mechanisms for affinity glucose recognition on the nanomaterial. The small size of the graphene as the transduction element can allow miniaturization of the sensor dimensions.

Additionally, and further to the above, the polymer functionalization of the graphene, as illustrated in exemplary FIGS. 2A-2B, or in other words, the coupling of graphene with boronic acid via stable chemical bonding, such as for example and without limitation, via π-π stacking interactions, can reduce the need for mechanical movable structures or physical barriers such as semipermeable membranes commonly used in existing affinity glucose sensors. Such techniques and configurations of the presently disclosed subject matter therefore simplify the device design and can enable a consistent, rapidly responsive measurement for noninvasive glucose monitoring. For example, the presently disclosed subject matter can enable wearable glucose monitoring devices to be realized, such as, in one non-limiting illustration, by integrating such sensors with contact lenses to detect glucose concentration in tears.

In another aspect of the disclosed subject matter, an affinity nanosensor can measure the concentration of low-charge target molecules or analytes and/or low-molecular-weight target molecules or analytes, such as glucose, via the dielectric response of a hydrogel embedded in a MEMS capacitive transducer. Such techniques can accomplish detection and, additionally or alternatively, monitoring of an analyte by transducing the binding of the analyte with receptors, such as functional groups, in the hydrogel to changes in the dielectric properties of the hydrogel. In this manner, for example, changes in the dielectric properties of the hydrogel can be measured using a MEMS capacitive transducer. The hydrogel-based affinity nanosensor can eliminate the irreversible consumption of the target molecules as well as the interference of electroactive species, and provide stable, nontoxic material amiable for implantation.

For example, and without limitation, a hydrogel can be directly immobilized onto the surface of the transducer via in situ polymerization and can be stable over time, thereby reducing the use of a semipermeable membrane, or other mechanical barriers and moving parts otherwise required to hold glucose sensitive material, that can be found in existing sensors, including existing CGM (continuous glucose monitoring) sensors, which often results in device complexity, among other drawbacks. Further, the hydrogel-based nanosensor can employ non-reactive equilibrium binding between glucose and the synthetic hydrogel, and can avoid irreversible glucose consumption.

Additionally, and further to the above, a synthetic hydrogel-based affinity glucose sensor and a MEMS differential dielectric transducer can be integrated to create a novel, miniaturized affinity CGM (continuous glucose monitoring) device with high levels of stability and accuracy, as depicted in exemplary FIGS. 17A-17B. As herein disclosed, the hydrogel 501 can be disposed between a first electrode 502 and a second electrode 503, and can be synthetically prepared, non-toxic and polymerized in situ in the device. Thus, the hydrogel and electrodes can represent, for example and without limitation, a dielectric disposed between two capacitor plates and represented by an effective capacitance and an effective resistance. Reversible affinity binding of glucose 504 with a boronic acid group, or other suitable glucose responsive polymer, such as PHEAA-ra«-PAAPBA, 505 in the hydrogel 501 can change the dielectric properties of the hydrogel 501, such as for example, the permittivity of the hydrogel. The permittivity of the hydrogel can represent the ability of the hydrogel to store electrical energy in an electric field {e.g., effective capacitance). The permittivity can then be measured using a MEMS capacitive transducer 506 to determine glucose concentration, or concentration of the target molecule 504.

Example results herein disclosed demonstrate that in a practical glucose concentration range such as for example 0-500 mg/dL and with a resolution of 0.35 mg/dL or better, the hydrogel-based affinity nanosensor of the presently disclosed subject matter can exhibit a repeatable and reversible response, and can be useful for CGM (continuing glucose monitoring).

Further details of device structure, fabrication, and operation procedures of the disclosed subject matter can be found in the following Examples, which are provided for illustration purpose only, and not for limitation.

EXAMPLE 1

This Example demonstrates a graphene-based affinity glucose nanosensor configured using the above-described techniques. Specifically, a synthetic polymer-functionalized graphene nanosesnor for affinity -based, label -free detection of low-charge, low-molecular-weight molecules was configured. As disclosed herein, the graphene is functionalized with a synthetic polymer monolayer derivatized with a boronic acid group, which is illustrated by way of example and without limitation in FIGS. 2A-2B. The synthetic polymer monolayer can thus exhibit reversible complexation with glucose to generate a detectable signal. For example and without limitation, the binding of the polymer monolayer with glucose on the graphene surface disposed between source and drain electrodes, as shown in exemplary FIGS. 2A-2B, can induce changes in the carrier density and mobility in the bulk of the graphene, thereby offering a high detection sensitivity of the glucose. Similar sensitivity can be obtained of other like analytes and/or target molecules. The small size of the graphene as the transduction element in the sensor of the instant Example allows miniaturization of the sensor dimensions. Moreover, the polymer functionalization of the graphene can reduce the need for physical barriers such as semipermeable membranes commonly used in existing sensors, thereby simplifying the device design and enabling rapidly responsive measurements for reliable glucose monitoring.

Devices of Example 1 can be configured as solution-gated graphene- based field effect transistors (GFET) whereby the graphene can be the conducting channel, formed between two gold electrodes (i.e., a source electrode and a drain electrode) on an insulating substrate surface, as shown in exemplary FIGS. 2A-2B. In particular, and by way of example without limitation, the monolayer of the synthetic glucose responsive polymer can be pyrene-terminated poly(3- acrylamidophenylboronic acid) (py-PAPBA). The py-PAPBA can be attached to the graphene surface via π-π stacking interactions depicted in exemplary FIG. 3. A target solution, such as a glucose solution, in phosphate buffered saline (PBS) can be held directly above the polymer-functionalized graphene in a polydimethylsiloxane (PDMS) microchannel, with an Ag/AgCl electrode inserted into the solution to serve as a gate electrode. In this manner, an electrical double layer (EDL) can form at the interface of the graphene and solution, and can serve as the gate dielectric layer.

During operation, under the control of a voltage applied between the gate and source electrodes (FGS), a bias voltage applied between the drain and source electrodes (Vus) can generate a current through the graphene (drain-source current, T_DS) that can be measured. This yields transfer characteristics of the GFET, i.e., the functional dependence of 7_Ds on GS, and can allow glucose concentration to be determined due to the binding of boronic acid moieties of the polymer PAPBA, which changes the electrical properties of the graphene.

For example and without limitation, the polymer-glucose binding can change the position of the Dirac point (FGs,Dirac), or, in other words, the value of the gate voltage at which the charge carriers neutralize and the drain-source current, I_D, achieves its minimum. Cyclic esters of boronic acid can form as a result of the binding of boronic acid groups to glucose molecules, which causes the overall ionization equilibrium to shift from neutral/insoluble boronic acid moieties to anionic/hydrophilic boronate. Thus, the carrier density can vary because of the electron exchanges between the graphene and the solution when the charge density in the solution changes. This can alter the Fermi level of the graphene, thereby shifting the Dirac point position.

Additionally, and further to the above, the polymer-glucose binding can also change the transconductance, g_m, i.e., the drain-source current change rate with respect to the gate voltage (S/_DS/^FGS), ^m the linear region of the GFET transfer characteristics. The charged polymer molecules on the graphene surface can be considered charged impurities, and induce electron scattering that can degrade the carrier mobility, μ, of the graphene. For example, and without limitation, this can decrease the transconductance according to:

where W and L are respectively the width and length of the graphene conducting channel, and Q is the gate capacitance per unit area.

As disclosed herein and further to the above, the graphene-based low- molecular-weight affinity nanosensor of Example 1 was fabricated using micro and nanofabrication methods on an oxidized silicon wafer, as illustrated in FIGS. 4A-4C. Other known fabrication techniques are also contemplated by the presently disclosed subject matter. After cleaning by piranha, or other suitable cleansing mixtures and/or protocols to remove metals and organic contamination, a layer of 5/45 nm Cr/Au can be deposited using thermal evaporation. A layer of photoresist can then be spin- coated on top of an Au layer and baked at 115°C for 1 minute. Other suitable bake temperatures and times are within the contemplated scope of Example 1. Photolithography can then be used to pattern the gate electrode, and the wafer can then be developed and etched in gold and chrome etchant sequentially. Graphene synthesized via chemical vapor deposition (CVD) on a copper sheet can be transferred onto the substrate following an established protocol to cover the source and drain electrodes, as depicted in exemplary FIG. 5.

Next, to perform PAPBA polymer functionalization in accordance with Example 1, the graphene and the underlying substrate can be immersed in a solution of pyrene-terminated polymer (py-PAPBA/methanol 3% w/v) for 4 hours at room temperature, and then washed thoroughly using methanol. Other suitable immersion times and temperatures are within the contemplated scope of Example 1. During testing of the nanosensor, described in additional detail below, glucose solution can be placed directly above the graphene and held in a PDMS open microchannel (-2.5 μΐ_^ in volume), which can be fabricated using soft lithography and reversibly bonded to the sensor device. An Ag/AgCl reference electrode 107 can be inserted into the solution above the graphene 101 to serve as the gate electrode for application of a gate voltage, as depicted in exemplary FIG. 6.

Raman spectroscopy can be used to test the sensor of the instant Example. Other techniques for observing and/or determining like molecular characteristics can also be employed to test the nanosensor of the presently disclosed subject matter. As illustrated by way of example and without limitation in FIG. 7, Raman spectroscopy verified that single-layer graphene was used in the device. The G band at approximately 1580 cm^"1 in the Raman spectrum, characteristic of the planar geometry of sp² bonded carbon, indicated that the material was graphene. Moreover, the sharp and symmetric 2D band at approximately 2685 cm^"1 indicated that the graphene consisted of a single layer of carbon atoms.

Polymer functionalization of the graphene was verified using Atomic Force Microscope (AFM) imaging. As shown in the exemplary AFM images of FIGS. 8A-8B, there was an increase (-10 nm) in the apparent height of the graphene sheet, suggesting the successful grafting of the polymer molecules. The polymer functionalization of the graphene of the sensor in the instant Example was also verified by measurement of the GFET (graphene-FET) transfer characteristics. For example and without limitation, the shape of the /DS- ^GS curve was similar before and after the functionalization protocol, while the Dirac point position, Fcs.Dirac ^was found to have shifted from 0.22 V to 0.18 V, as depicted in exemplary FIG. 9. The lack of change in shape of the /DS- ^GS curve suggested that changes in the carrier mobility in the graphene were insignificant. Such results are consistent with the polymer, and in particular the boronic acid moieties, being electrically neutral such as would cause little electron scattering to change the graphene' s carrier mobility. The shift in cs.Dirac can be attributed to ^-doping (i.e., electron doping) of the graphene, and was consistent with the surface-attached polymer inducing electron transfer from the solution to the graphene. It can therefore be concluded that the graphene successfully functionalized with the PAPBA polymer in the instant Example.

Testing of the graphene' s transfer characteristics at varying glucose concentrations (e.g., 0 mg/dL, 50 mg/dL, 100 mg/dL, and 200 mg/dL) revealed that the transfer characteristics changed consistently as the binding of glucose to the boronic acid shifted the electrically neutral boronic acid groups to anionic boronate esters, as depicted herein at reference nos.

in exemplary FIG. 10. In response to increases in the glucose concentration in Example 1, the Dirac point position, Gs,Dirac, shifted to higher gate voltages with a sensitivity of approximately 2.5 mV/(mg/dL), while the /DS-^GS curve broadened in shape. The shift in ^Gs^irac indicated that the graphene was p-doped, or hole-doped, which can be attributed to changes in the amount of electric charge on the EDL gate capacitor due to the formation of anionic boronate esters. The broadening of the transfer characteristic curve reflected a decrease in transconductance from 100 to 20 μ8. According to Equation (1), this can correspond to a decrease in the carrier mobility as a consequence of the polymer-glucose binding, as the negatively charged boronate esters, in the role of charged impurities, can cause electron scattering.

The influence of potential contributors or interferents other than polymer-glucose binding was also tested in Example 1. Control measurements were performed on pristine graphene that was not functionalized with the PAPBA polymer. It was observed that neither the Dirac point position nor the transconductance changed as the glucose concentration was varied from 60 to 200 mg/dL, as depicted in exemplary FIG. 11, indicating that when not functionalized with the PAPBA polymer, there is a negligible response of the graphene to glucose concentration changes. Thus, the response of the polymer-functionalized nanosensor to the changes in glucose concentration resulted from the glucose-polymer binding of the presently disclosed subject matter. EXAMPLE 2

This Example demonstrates a graphene-based affinity glucose sensor configured using the above-described techniques. Specifically, an atomically thin graphene-based affinity glucose nanosensor was configured as a solution-gated graphene field effect transistor (GFET), as illustrated in exemplary FIGS. 2A-2B. The graphene, serving in Example 2 as the conducting channel, was functionalized with pyrene-l-boronic acid (PBA) via π-π stacking interactions, as shown by way of example and not limitation in FIG. 12. The sensor was fabricated using known micro and nanofabrication techniques. A polydimethylsiloxane (PDMS)-based open well (-20 μΐ,) was bonded to the substrate, and glucose solution was placed into the well. An Ag/AgCl reference electrode, mounted on a three-axis positioner, was inserted into the solution to serve as the gate electrode. An electrical double layer (EDL) formed at the interface of the graphene and solution served as the gate capacitor. Binding of glucose and the boronic acid formed a glucose-boronate ester complex, as shown in exemplary FIG. 13, inducing changes in the electric conductance of the graphene, which was measured to determine the glucose concentration.

Because the capacitance of the double layer can be influenced by the solution composition in a solution-gated FET, prior to any chemical functionalization of the graphene, the fluctuations of the EDL capacitance that can be attributed to changes in the glucose concentration were examined in the instant Example. For example, and without limitation, glucose was dissolved in phosphate buffered solution (pH 7.4) to obtain desired concentrations (2 μΜ to 25 mM). The same solutions were used in all of the subsequent experiments of the instant Example for purpose of consistency, and not limitation. Without any chemical functionalization of graphene, the transfer characteristics (source-drain current 7_DS as a function of gate voltage VQS) measured at the different glucose concentrations were approximately identical, as depicted in exemplary FIG. 14A. The results suggested that glucose, at the selected concentration range, did not either interact with graphene or vary the capacitance of the EDL.

As herein disclosed, and further to the above, the nanosensor of

Example 2 was then immersed in PBA (pyrene-l-boronic acid) solution for 4 hours at room temperature, followed by sequentially rinsing in acetonitrile, isopropanol and deionized water to remove free PBA. Prior to chemical functionalization, in the Raman spectrum of the graphene, depicted in exemplary FIG. 14B, at the channel region, the ratio of the intensity of the 2D band to the G band {I2DIIG) was 2.5, and the full width at half maximum (FWHM) of the 2D band was -27 from Lorentz fitting, both of which are evidence of monolayer graphene in addition to the color contrast observed under microscope shown in exemplary FIG. 5. The Raman spectrum of the PBA solution-rinsed graphene exhibited signature peaks of BOH bending (1286 cm^" B-0 stretching (1378 cm^"1), and G-band splitting (1574, 1595, 1613 cm^"1) due to the graphene-pyrene π-π stacking interaction. Also, the 2D band was measured as shifting to a higher wavenumber (from 2685 to 2692 cm^"1), which was considered as a result of chemical doping. The measured transfer characteristics, depicted in FIG. 14A, also verified the chemically induced p-type doping, represented by the increase of the neutral point voltage Κ_Νρ (the gate voltage at which 7_DS attains its minimum) from 0.33 V to 0.575 V. These observed characteristics of the boronic acid as well as the graphene-pyrene interaction confirm that the PBA molecules successfully immobilize on the graphene.

After functionalization, it was confirmed that replenishment of sample solution to the nanosensor did not interrupt the pyrene-graphene coupling. As depicted in exemplary FIG. 15, the nanosensor of Example 2 was then tested by exposure to glucose solution at different concentrations. The transfer characteristics curve was found to shift to the left. For example, and with reference again to exemplary FIG. 15, the shift was approximately 0.1 15 V as the glucose concentration increased from 0 to 25 mM. This suggests that the binding of glucose and boronic acid generated n-type doping to graphene. As the estimated transconductance (i.e., the slope of linear sections of the transfer characteristics curve) did not change significantly, the carrier mobility of the graphene was believed to be approximately constant. Rather, changes in the carrier concentration of graphene was considered the main contributor to the observed shift of FNP. Measurements using butyric-acid functionalized graphene were also performed to serve as control. Variations in the source-drain current 7_Ds with the glucose concentration at a fixed gate voltage VGS were also examined. It was observed that TDS decreased monotonically with glucose concentration when VQS was lower than the neutral point voltage FNP, and this trend was reversed when VQS ^> ^NP, as depicted in FIG. 15, which can be attributed to the shift of the transfer characteristics. Using this observed dependence of 7_Ds on the glucose concentration, it can be estimated that, with a noise level of approximately 17 nA for TDS, the resolution of the nanosensors for glucose measurements herein described in Example 2 was approximately 0.46 μΜ, appropriate for monitoring glucose in human bodily fluids, such as for example, saliva and tears.

Additionally, the change of Κ_Νρ before and after PBA functionalization, denoted herein as ΔΚΝΡ,Β, and the further changes of after the graphene was exposed to glucose, denoted herein as AJ¾>,G, were further studied. As herein disclosed, Δ¾Ρ,Β = ΔΚΝΡ,Β ^~ Δ¾ρ,ρ, and AJ¾>,G = ΔΚΝΡ,Β ^~ AJ¾>,G, where ΔΚΝΡ,Ρ and the ΔΚΝΡ,Β are the neutral point voltages measured in fresh buffer for pristine graphene and PBA-functionalized graphene, respectively; AJ¾>,G is the neutral point voltage for PBA-functionalized graphene measured in glucose solution. Both ΔΚΝΡ,Β and AJ¾>,G varied from sensor to sensor, e.g. , because of artifacts such as organic residue left on graphene from the fabrication process. These artifacts can cause a device-to-device disparity in chemical functionalization of graphene, and hence in the doping level at a given glucose concentration. At a given concentration, it was observed, and depicted in exemplary FIG. 15, that the ratio ΔΚΝΡ,Ο ΔΚΝΡ,Β did not vary significantly from sensor to sensor, with a variation of less than 6% for the three nanosensors tested in the instant Example. It can be noted that ΔΚΝΡ,Β is the shift of FNP caused by functionalization of boronic acid and VNP,G is by glucose- boronic acid binding, therefore ΔΚΝΡ,Ο ΔΚΝΡ,Β can be regarded as a measure of the fraction of boronic acid that is occupied by glucose. Since the present Example was conducted under conditions of constant temperature and pH, the fraction of boronic acid that binds to glucose is solely dependent on the glucose concentration. This suggests that ΔΓΝΡ,Ο ΔΓΝΡ,Β should be a function of glucose only, and independent of the sensor or the order in which the sample solution was added. The measured dependence of this ratio on glucose concentration followed the Hill-Langmuir equation for equilibrium ligand-receptor binding. For example, and as depicted in exemplary FIG. 16, a least squares fit yielded an equilibrium dissociation constant (Κ_Ό) of 38.6 μΜ, which is appropriate for practical glucose sensing applications.

To explain the observed p-type doping due to the attachment of PBA, depicted in FIG. 8A, it can be noted that while pyrene group is electron-rich and not expected to induce p-doping, boronic acid is electron deficient and its electron- withdrawing nature can induce p-doping in the graphene. This is supported by experiments in which immobilization of electron-rich groups on graphene, such as for example, butyric acid, a carboxylic acid, can result in n-type doping in the graphene. Further, the observed n-type doping due to the boronic acid-glucose binding can be attributed to an increase in the local electrostatic potential in the proximity of graphene. This electrostatic potential increase can result from the formation of boronate, which increasing the electron donating ability of boronic acid while weakening its electron-withdrawing ability.

In the instant Example, a graphene-based nanosensor for affinity-based detection of low-charge, low-molecular-weight molecules and/or analytes, such as glucose, was configured and tested. Similar analytes are within the scope of Example 2, as well as the other Examples of the presently disclosed subject matter. The nanosensor of Example 2 employed a GFET (graphene field-effect transistor) in which graphene was functionalized with boronic acid groups for glucose recognition. The boronic acid was attached to the graphene via the interaction between graphene and pyrene groups, thereby allowing sensitive detection of electrically neutral glucose molecules. Testing results from Example 2 demonstrate that the nanosensor herein disclosed can measure glucose in a practically relevant range of 2 μΜ to 25 mM, with a resolution of 0.46 μΜ. The observed shifts of the transfer characteristics strongly indicate that recognition of glucose can result from the formation of glucose-boronate ester, which can reduce the boronic acid-induced p-type doping in the graphene. For practical clinical applications, the nanosensor of Example 2, representative of the presently disclosed subject matter, without limitation, can be highly miniaturized without the use of mechanical moving parts or physical barriers, and thus of practical utility in glucose monitoring applications.

In addition to the embodiments of Examples 1 and 2, it is contemplated that graphene modified using other attachment groups, such as for example and without limitation, 9-anthracene-boronic acid, instead of PBA, can obtain consistent results.

EXAMPLE 3

This Example demonstrates a hydrogel-based affinity glucose nanosensor configured using the above-described techniques. Specifically, the MEMS affinity of the instant Example can consist of a sensor chip integrated with a hydrogel of a synthetic copolymer 505, such as, for example and without limitation, poly(N-hydroxy ethyl acrylamide-ra«-3-acrylamidophenylboronic acid) (PHEAA-ra«- PAAPBA), as depicted in exemplary FIG. 17A-17B. The sensor chip can further include coplanar electrodes 502, 503 for impedance sensing, and a thermistor sensor 507 for closed-loop temperature control, as depicted in exemplary FIG. 18. Glucose can bind reversibly to phenylboronic acid moieties to form strong cyclic boronate ester bonds, which can induce changes in the hydrogel' s dielectric properties. Such changes in the dielectric properties of the hydrogel can be measured to determine the glucose concentration, or other such target molecule or analyte.

To fabricate the MEMS affinity nanosensor herein disclosed in Example 3, reference can be made to exemplary FIGS. 19A-19C. With reference to FIG. 19A, for example and without limitation, a chrome (Cr)/gold (Au) film (5/50 nm) can be deposited on a Si0₂ substrate by thermal evaporation and patterned to form the coplanar electrodes (502, 503) (1 mm x 1 mm) and thermistor sensor 507 as in exemplary FIGS. 19B-20A. The hydrogel 501 can be synthesized in situ on the sensor, as shown in exemplary FIGS. 19C-20B. A mixed prepolymer solution containing hydrogel components (AAPBA, HEAA, TEGDA, and AAPH) can be spin- coated on the sensor chip and allowed to gelate in situ and covalently attach to the substrate at 70 °C, or other suitable temperature. During testing of the sensor herein disclosed in Example 3, the sensor was immersed in glucose solution. The device was connected to an impedance/voltage transformation circuit 701 driven by sinusoidal input from a function generator 702, such as for example and without limitation, an Agilent 33220A, and as shown in exemplary FIGS. 20C-20D. Other such suitable function generators are within the scope of the presently disclosed subject matter. The sinusoidal input from the function generator 702 imposed an AC field between the electrodes (502, 503), which caused permittivity to be manifested in hydrogel polarization. Such experiments were conducted at frequencies below 100 kHz allowed by a lock-in amplifier 703 used to detect the amplitude and phase shift of the output voltage from the circuit.

First, in testing the sensor of Example 3, the sensor's hydrogel thickness was characterized using vertical scanning interference microscopy. Other suitable methods and techniques can be used to characterize hydrogel thickness. Next, the sensor's frequency-dependent effective capacitance and effective resistance at selected glucose concentrations were tested. Further, the sensor's time response to changes in glucose concentration within 40-300 mg/dL was assessed to demonstrate the potential of the sensor for realizing real-time monitoring of glucose concentrations. Finally, the output of the sensor under physiologically relevant ranges of glucose, such as, for example and without limitation, 0-500 mg/dL, was tested to evaluate the repeatability of the device for stable CGM (continuous glucose monitoring) applications.

The frequency responses of the nanosensor of Example 3, with an 8- μπι hydrogel was tested at varying bias voltage frequency, denoted herein as and ranging from 5-90 kHz, and glucose concentration, denoted herein as c_gi_ucose, and ranging from 40-300 mg/dL. At a given c_gi_UCOse in the instant Example, the effective capacitance (C) and resistance (R) both decreased with frequency as a result of the hydrogel's dielectric relaxation, as illustrated by way of example and not limitation in FIGS. 21A-21D. At all frequencies tested in Example 3, C and R both increased with glucose concentrations. Thus, the frequency-dependent sensor response suggests that glucose-induced dielectric change at a fixed excitation frequency can be measured. In response to such changes in c_gi_UCOse within 40-300 mg/dL, the effective capacitance of the sensor at a given / (e.g., 30 kHz) rapidly reached steady state, for example, within 22 seconds, as depicted in exemplary FIG. 22. Also at / = 30 kHz, C and R both increased monotonically with c_gi_UCOse in the range of 0-500 mg/dL, as depicted in exemplary FIGS. 23A-23B. This trend indicates that the hydrogel-glucose binding of the presently disclosed subject matter influences the polarization behavior of the hydrogel. For example and without limitation, a number of polarization mechanisms such as dipole reorientation, ionic polarization, interfacial polarization, and counterion diffusion, can combine to contribute to the glucose-dependent impedance change of the nanosensor herein disclosed. The sensor's glucose measurement resolution was determined to be 0.47 mg/dL (for capacitance) and 0.27 mg/dL (for resistance).

In the instant Example, the sensor can be used for CGM monitoring. The sensor can include coplanar electrodes and small hydrogel thickness. Additionally, the sensor disclosed herein permits repeatable and rapid measurements of glucose concentration. The effective capacitance (C) and effective resistance (R) both increasing monotonically with glucose concentration (c_gi_ucose) in clinically relevant glucose concentrations of 0-500 mg/dL, also demonstrating the applicability of such sensors for ISF (interstitial fluid) glucose monitoring.

EXAMPLE 4 This Example provides a hydrogel-based affinity sensor configured using the above-described techniques. Specifically, a MEMS affinity nanosensor can utilize a synthetic glucose-sensitive hydrogel, which can be constructed, for example and without limitation, by N-3-aciylamidophenylboronic acid (AAPBA) glucose- binding motifs, and acryl N-Hydroxyethyl acrylamide (HEAA) with a tunable hydrophilicity, and which uses tetraethyleneglycol diacrylate (TEGDA) as the cross- linker 508 and 2,2'-Azobis (2-methylpropionamidine) dihydrochloride (AAPH) as the one-step free radical polymerization initiator. In the instant Example, when glucose 504 binds reversibly to the phenylboronic acid moieties in the AAPBA segments to form strong cyclic boronate ester bonds, a change in the dielectric properties of the hydrogel 501 occurs and can be measured to determine the glucose concentration.

By way of example, and not limitation, the dielectric properties of the hydrogel can be represented by the complex permittivity:

ε = ε — ιε (2) where the real permittivity ε ' represents the ability of the hydrogel to store electric energy, while the imaginary permittivity ε" is related to the dissipation of energy. In the instant Example, when the gap between the electrodes of a parallel- plate transducer is filled with the hydrogel, as shown in exemplary FIG. 24, the transducer can be represented by a capacitor having an effective capacitance denoted herein as C_x, and resistor having an effective resistance denoted herein as R_x, connected in series. Correspondingly, the real and imaginary parts of the complex permittivity can be related to the these parameters by the following equations:

ε' = C_x/C₀ (3)

where o can be the capacitance when the electrode gap is in vacuum. The interactions of the hydrogel with glucose as disclosed herein can cause changes in the hydrogel' s composition and conformation, and thus changes in its dielectric properties ε ' and ε". Thus, the transducer of the instant Example can experience changes in its effective capacitance C_x and effective resistance R_x, which can be measured to determine glucose concentration.

Further to the above, and as disclosed herein, the transducer of the instant Example can be enabled by MEMS technology and can use a pair of parallel electrodes (510, 511) sandwiching the hydrogel 501, as shown by way of example and not limitation, in FIGS. 24 and 25A-25B. The upper electrode 510 can be perforated to allow passage of glucose molecules 504, and can be passivated within a perforated diaphragm to avoid direct contact with the hydrogel 501. The perforated electrode 510 and diaphragm can be supported by microposts 512 so that they do not collapse onto the lower electrode 511 on the substrate. In this Example, glucose molecules 504 can reversibly bind with the hydrogel 501, thereby changing the hydrogel' s complex permittivity. Such changes can occur in both the real and imaginary parts of the complex permittivity, which can be used to determine glucose concentration. For example, and without limitation, the real permittivity can be interrogated via measurement of the capacitance between the electrodes (510, 511) to determine the glucose concentration.

To fabricate the exemplary MEMS capacitive transducer of the instant Example, a chrome (Cr)/gold (Au) film (5/100 nm) can be deposited by thermal evaporation and patterned to form a lower electrode 511 (500 μπι x 500 μπι) on a Si0₂-coated wafer. The patterned gold electrode can then be passivated, for example, with Parylene (1 μπι). Subsequently, a sacrificial layer (5 μπι), such as for example an S1818, and an additional layer of Parylene (1.5 μιη) can be deposited. Another Cr/Au (5/100 nm) film can then be patterned to form the upper electrode 510 and passivated by another Parylene layer or like chemical vapor. An SU-8 layer can then be patterned to form a channel and anti-collapse microposts 512 between the electrodes (510, 511). The Parylene diaphragm can be patterned with reactive ion etching (RTE) to form perforation holes to allow glucose permeation through one or more glucose permeation holes 510a. The sacrificial photoresist layer can then be removed with acetone to release the diaphragm. Such fabrication process is illustrated by way of example without limitation in FIGS. 26A-26B.

In the instant Example, the hydrogel can be prepared in situ in the capacitive transducer. First, a mixture of the hydrogel components (AAPBA, FIEAA, TEGDA, and AAPH) in solution can be deoxygenated by nitrogen gas for 30 minutes, or other such suitable time, and then injected into the sensor, filling the gap between the parallel electrodes. The sensor can then be placed in a nitrogen environment and heated for four hours at 70°C, or other such suitable time and temperature, as shown by way of example and not limitation in FIG. 26D. The hydrogel-integrated sensor can then be rinsed with water and ethanol or other suitable washing solution to remove unreacted monomer and reagents.

The hydrogel, as herein disclosed, can be synthesize via free radical polymerization with AAPBA and HEAA monomers. An HEAA to AAPBA molar ratio can be approximately equal to 9, or approximately 10% AAPBA in all the monomers. A solution consisting of AAPBA (1.1% w/v), HEAA (5.5% v/v), TEGDA (0.8% v/v), and AAPH (0.16% w/v) in distilled water can then be prepared for polymerization. A stock solution (0.1 M) of glucose can be prepared by dissolving D-(+)-glucose (0.9 g) in distilled water to 50 mL. In the instant Example, glucose solution at varying concentrations (40, 70, 90, 180, 300, and 500 mg/dL) can be prepared by diluting the stock solution.

During testing of the nanosensor of the instant Example, the device was placed in an acrylic test cell (2 mL in volume) filled with glucose solution, as shown in exemplary FIGS. 27A-27D. The sensor was connected to a capacitance/voltage transformation circuit 701 driven by a sinusoidal input from a function generator 702, such as an Agilent 33220A, which imposed an AC field on the electrodes (510, 511) of the device to induce a glucose concentration-dependent change in the permittivity of the hydrogel 501. The resulting changes in the effective capacitance C_x of the capacitance/voltage transformation circuit 701 can be determined by measuring the output voltage (U_out) from a given input AC voltage (Ui„). All experiments of the instant Example were conducted at frequencies in a range of 1 to 100 kHz as allowed by a lock-in amplifier 703, such as Stanford Research Systems SR844, used in output voltage measurements.

First, the sensor of the presently disclosed subject matter of the instant

Example was tested to investigate its response to different glucose concentrations under bias voltages of different frequencies, as depicted in exemplary FIGS. 28A- 28B. At each of a series of physiologically relevant glucose concentrations, such as for example and without limitation, 0-500 mg/dL, the effective capacitance of the sensor, and hence the permittivity of the hydrogel, decreased with increasing frequency over the entire frequency range tested, ranging from 1-100 kHz, as depicted in exemplary FIG. 28A. This trend demonstrates dielectric relaxation of the hydrogel, in which the dielectric properties of the hydrogel have a momentary delay with respect to a changing electric field. The dielectric properties of the hydrogel in an electric field can be influenced by a number of mechanisms of polarization, or in other words, can be influenced by a shift of electrical charges from their equilibrium positions under the influence of an electric field, including for example, electronic polarization, ionic polarization, dipolar polarization, counterion polarization, and interfacial polarization. Electronic polarization and ionic polarization can involve the distortion of electron clouds with nucleus and the stretching of atomic bonds, while counterion polarization and dipolar polarization reflect redistribution of ions and reorientation of electrical dipoles.

In the instant Example, at a given frequency, the effective capacitance of the hydrogel increased consistently with glucose concentration in the entire range tested of 0-500 mg/dL, which is reflected in the sensor's frequency response depicted in exemplary FIG. 28A. The response is plotted versus the glucose concentration in exemplary FIG. 28B. For example, and without limitation, at 30 kHz, the effective capacitance increased from 16.2 pH to 24.8 pF as the glucose concentration increased from 0 mg/dL to 500 mg/dL. This indicates that the binding between the hydrogel and the glucose influences the polarization of the hydrogel, which can include changes in the hydrogel' s structural conformations, permanent dipole moments, elastic resistance to the dipole rearrangement in the electric field, and EDL (electric double layer ) characteristics. Such effects, resulting from the configuration of the presently disclosed subject matter, combine to result in the glucose concentration dependence of the hydrogel' s dielectric properties, explaining the observed variation of the device's effective capacitance with glucose concentration.

With reference to exemplary FIG. 28B, the dependence of the effective capacitance on glucose concentration is generally nonlinear over the full glucose concentration range tested (0-500 mg/dL). Thus, in practical applications, a calibration curve represented by a lookup chart, nonlinear equation, or other suitable representation, can be used to determine the glucose concentration from a measured effective capacitance value. In moderately smaller glucose concentration ranges, however, this relationship can become more linear. For example, and without limitation, the effective capacitance was approximately linear with glucose concentration from 40-300 mg/dL (R =0.993), a range that is relevant to CGM (continuous glucose monitoring) needs. In this range, a calibrated linear equation can hence be adequate for the determination of glucose concentration from measurement results.

As described herein and further to the above, the sensor was next tested by conducting experiments in triplicates to examine the ability of the sensor herein disclosed to measure glucose concentrations in a repeatable manner and with adequate sensitivity, as depicted in exemplary FIG. 28B. At all glucose concentrations of the instant Example, the standard error in the effective capacitance was less than 0.91 pF (2.3%), indicating excellent repeatability. In addition, at all of the measurement frequencies used in the instant Example, the resolution and range of glucose measurement resolution were found to be appropriate for CGM. Considering, for example and without limitation, a 30 kHz frequency, the sensitivity of the sensor was approximately 15 fF (mg/dL)^"1 in the glucose concentration range of 0-40 mg/dL. With a capacitance measurement resolution of 3 fF, as allowed by the measurement setup herein disclosed, the resolution for glucose concentration measurement of the sensor was correspondingly estimated to be 0.2 mg/dL. At a signal-to-noise ratio of 3, such resolution yielded a detection limit of 0.6 mg/dL, which is well below the physiologically relevant glucose concentration range (typically greater than 40 mg/dL). For glucose concentrations within 40-300 mg/dL, the sensitivity was approximately 23 fF (mg/dL)^"1, corresponding to an estimated resolution of 0.12 mg/dL. At higher glucose concentrations, such as for example, 300-500 mg/dL, the nonlinear sensor response can experience a gradual declination in sensitivity and resolution (respectively to 8.4 fF(mg/dL)^_1 and 0.35 mg/dL at 500 mg/dL) as an increasingly small number of binding sites remained available in the hydrogel. These sensor characteristics, appropriate for practical applications, are comparable to those of commercially available electrochemical sensors, such as for example, 1 mg/dL over a glucose concentration range from 0-400 mg/dL, or 500 mg/dL) as well as other research-stage boronic acid-based affinity sensors, such as for example, 0.3 mg/dL over a range from 0-300 mg/dL or 540 mg/dL, for CGM.

Additionally, and further to the above, the response to glucose as compared to the response of potential interferents of the hydrogel-based sensor herein disclosed was tested. For example, and without limitation, nonspecific molecules exist in ISF (interstitial fluid) and can interact with boronic acid, which is the glucose sensitive component of the hydrogel of the sensors herein disclosed. Such molecules can include fructose (~ 1.8 mg/dL), galactose (-1.8 mg/dL), lactate (~9 mg/dL), and ascorbic acid (-1.32 mg/dL). The hydrogel-based sensor of the presently disclosed subject matter, subjected to interaction with such molecules, resulted in a response lower than the response to glucose molecules. For example, at the same concentration of 90 mg/dL, the effective capacitance change at 30 kHz due to fructose, galactose, lactate, and ascorbic acid was 17%, 38%, 32%, and 28%, respectively, as depicted in exemplary FIG. 29A. The effective capacitance C, of the instant Example can be calculated according to:

AC = C - C₀ (5) where C is the effective capacitance at a given glucose (or interferent) concentration, and o is the effective capacitance in the absence of glucose and interferents. Considering that the physiological concentrations of the potential interferents were about one order of magnitude lower than that of glucose, the sensor of the instant Example was determined to be sufficiently selective for measurements of glucose in interstitial fluid for CGM applications.

While boronic acid can bind to diol-containing molecules, the selective response of the sensor to glucose over the potential interferents can be attributed to the unique binding behavior between boronic acid and glucose. At a 1 : 1 ratio, boronic acid in fact binds more strongly to fructose than glucose. However, with a high concentration of boronic acid units, which is the case in the instant Example, boronic acid can bind with glucose at a 2: 1 ratio. The 2: 1 binding between glucose and boronic acid units can play a major role in the sensor response by causing additional crosslinking of the hydrogel that can lead to the augmentation of elastic resistance to electric field-induced dipole reorientation. The rather insignificant device response to the exemplary potential interferents (fructose, galactose, ascorbic acid, and lactate) can alternatively be attributed to a lack of the 2: 1 binding mode.

To gain further insight into the principle and operation of the sensor herein disclosed, the role of boronic acid in glucose concentration was also tested. For example and without limitation, using hydrogels with and without AAPBA content, the dependence of the effective capacitance on glucose concentration can be obtained. Such technique is illustrated by way of example and not limitation in FIG. 29B. It can be seen that when using an AAPBA-free hydrogel, the sensor can exhibit negligible changes in the effective capacitance in response to glucose concentration changes. This contrasts with the strong glucose-induced response of the sensor when it was equipped with the 10%-AAPBA hydrogel, indicating for example that boronic acid moieties in the hydrogel were responsible for the desired recognition of glucose by the sensor.

Finally, the sensor of Example 4 was tested with time-resolved glucose concentration measurements to assess its ability to track glucose concentration changes in a consistent and reversible manner. For example and without limitation, the measured sensor output at 30 kHz varied from 16.5 pF at 0 mg/dL to 24.8 pF at 500 mg/dL, as depicted in exemplary FIG. 30. In particular, when the sensor was exposed to a glucose concentration after experiencing another sample that was either higher or lower in concentration, virtually the same effective capacitance value was consistently obtained. For example, the effective capacitance at 40 mg/dL over the two periods, from 20 to 38 minutes and from 321 to 341 minutes in this Example, were respectively 16.87 pF and 16.73 pF, agreeing within 0.8%. Similarly, the reversibility was within 3.4% and 1.3% for the measurement data at glucose concentrations of 180 and 300 mg/dL, respectively. This indicates that because the binding between glucose and boronic acid moieties in the hydrogel, the sensor possesses excellent reversibility in response to glucose concentration changes. The time constant of the response, in other words the time for the sensor to reach 63 % of the steady state response, was approximately 16 minutes. This time constant was attributable to the relatively large thickness of the hydrogel (-200 μιη), through which glucose molecules diffuse to interact with the capacitive transducer. As the glucose diffusion time decreases with the square of the hydrogel thickness, thinner hydrogels can be used to effectively obtain more rapid time responses.

Accordingly, a hydrogel-based affinity glucose nanosensor that measures glucose concentration through dielectric transduction was provided and tested in Example 4 of the presently disclosed subject matter. The sensor can consist of a pair of thin-film parallel capacitive electrodes sandwiching a synthetic hydrogel. Glucose molecules permeate into the hydrogel through electrode perforations, and bind reversibly to boronic acid moieties of the hydrogel. Such binding can induce changes in the dielectric polarization behavior, and hence the complex permittivity, of the hydrogel. Thus, the effective capacitance between the electrodes, which is directly related to the real part of the complex permittivity, can be measured to determine glucose concentration. The use of an in situ polymerized hydrogel in Example 4 can simplify the design of the sensor, facilitate its miniaturization and robust operation, and can improve the tolerance of the device to biofouling, for example, when implanted subcutaneously. Testing of the sensor, herein described by way of example and not limitation, showed that the effective capacitance of the sensor, in a measurement frequency range of 1-100 kHz, responded consistently to glucose concentration changes ranging from 0 to 500 mg/dL. At a given frequency, the effective capacitance increased consistently with glucose concentration, suggesting that the affinity binding between the glucose and boronic acid moieties caused the real permittivity of the hydrogel to increase. At 30 kHz, the measurement resolution of the sensor was estimated to be 0.2, 0.12, and 0.35 mg/dL in the glucose concentration ranges of 0-40, 40-300, and 300-500 mg/dL, respectively. When subjected to time varying glucose concentration changes in the full 0-500 mg/dL range, the microsensor response was consistent and reversible. The time constant of this response was approximately 16 minutes, which can be improved by using thinner hydrogels for reduced diffusion distances.

The presently disclosed subject matter is not to be limited in scope by the specific Examples disclosed herein. Indeed, various modifications of the disclosed subject matter in addition to those described herein will become apparent to those skilled in the art from the foregoing description and accompanying figures. Such modifications are intended to fall within the scope of the disclosed subject matter.

Claims

1. An affinity nanosensor for detection of low-charge, low-molecular-weight molecules comprising:

a solution-gated field effect transistor, comprising

a silicon substrate,

a source electrode disposed on the silicon substrate, a drain electrode disposed on the silicon substrate,

graphene functionalized with a synthetic polymer monolayer disposed between the source electrode and drain electrode on the silicon substrate, the functionalized graphene comprising a conducting channel of the solution-based field effect transistor, and the synthetic polymer monolayer being responsive to a first analyte, and

a reference electrode disposed between the source and drain electrodes; and

an electrical double layer at the interface of the graphene and solution comprising a gate capacitor.

2. The affinity nanosensor of claim 1, wherein the silicon substrate comprises an oxidized silicon substrate wafer.

3. The affinity nanosensor of claim 1, wherein the source electrode and drain electrode further comprise gold electrodes.

4. The affinity nanosensor of claim 1, wherein the synthetic polymer monomer further comprises a boronic acid.

5. The affinity nanosensor of claim 1, wherein the graphene comprises graphene functionalized with the synthetic polymer monolayer via π-π stacking interactions.

6. The affinity nanosensor of claim 1, wherein the first analyte comprises glucose.

7. The affinity nanosensor of claim 1, wherein the reference electrode further comprises silver chloride.

8. The affinity nanosensor of claim 1, wherein the gate capacitor comprises a capacitor adapted to be affected by varying concentrations of the first analyte.

9. A method of fabricating an affinity nanosensor for detecting low-charge, low- molecular-weight molecules, comprising:

providing a silicon substrate wafer having a uniform thickness throughout, the wafer having oppositely disposed top and bottom faces;

providing a first gold portion on the top face of the first material, and a second gold portion separated from the first gold portion by a channel region;

transferring graphene onto the top face of the wafer in the channel region;

functionalizing the graphene with a synthetic polymer monolayer, the synthetic polymer monolayer being sensitive to a first analyte; and

mounting a conductive wire within the channel region.

10. The method of fabricating an affinity nanosensor of claim 9, wherein the providing the first and second gold portions comprises etching the first gold portion and the second gold portion to the top face of the wafer.

11. The method of fabricating an affinity nanosensor of claim 9, wherein the transferring the graphene further comprises coupling graphene to the top face of the wafer via chemical vapor deposition.

12. The method of fabricating an affinity nanosensor of claim 9, wherein the functionalizing further comprises:

immersing at least the graphene transferred onto the wafer in a solution comprising a boronic acid for at least four hours at room temperature; and

washing the graphene transferred onto the wafer using methanol.

13. The method of fabricating an affinity nanosensor of claim 12, wherein the immersing further comprises coupling a pyrene-1 -boronic acid to the graphene via π-π stacking interactions.

14. The method of fabricating an affinity nanosensor of claim 9, wherein mounting the conductive wire within the channel region comprises providing a silver wire mounted on a positioner to serve as a gate electrode.

15. An affinity nanosensor for detection of low-charge, low-molecular-weight molecules comprising:

a parallel plate transducer;

a synthetic hydrogel disposed between a first plate and a second plate of the parallel plate transducer, the hydrogel being responsive to a first analyte; and

a temperature sensor located below the first and second plates.

16. The affinity nanosensor of claim 15, wherein the first and second plates of the parallel plate transducer further each further comprise a sensing electrode comprising gold.

17. The affinity nanosensor of claim 16, wherein at least one of the first plate and second plate of the parallel plate transducer comprises a perforated plate passivated within a perforated diaphragm, and wherein the at least one perforated plate and at least one perforated diaphragm are supported by at least one micropost.

18. The affinity nanosensor of claim 15, wherein the synthetic hydrogel further comprises a synthetic copolymer comprising a boronic acid.

19. The affinity nanosensor of claim 15, wherein the first analyte comprises glucose.

20. A method of fabricating an affinity nanosensor for detecting low-charge, low- molecular-weight molecules, comprising:

providing a silicon substrate wafer having a uniform thickness throughout and having oppositely disposed top and bottom faces;

providing a first electrode on the top face of the wafer; providing a second electrode, spaced a first distance over the first electrode, above the top face of the wafer, the second electrode being supported above the top face of the wafer by at least one micropost;

preparing a hydrogel functionalized with a polymer responsive to a first analyte; and

filling the hydrogel between the first and second electrodes, the hydrogel being.

21. The method of fabricating an affinity nanosensor of claim 20, further comprising perforating the second electrode with one or more permeation holes.

22. The method of fabricating an affinity nanosensor of claim 20, wherein the first analyte comprises glucose.

23. The method of fabricating an affinity nanosensor of claim 20, wherein the preparing the functionalized hydrogel further comprises synthesizing the hydrogel in situ via polymerization of the hydrogel with a boronic acid, and gelating the hydrogel between the first and second electrodes.