Carbon nanofiber sensor for non-enzymatic glucose detection and methods of glucose detection using such carbon nanofiber sensor

Abstract

A general methodology for the development of sensitive and selective sensors that can achieve a low cost detection of glucose without using enzymes is disclosed. The method uses carbon nanofiber (CNF) array electrodes for the electrochemical detection of glucose. CNFs grown by plasma enhanced chemical vapor deposition (PECVD) with diameters ranging from 13-160 nm and a height of approximately one micrometer are preferred. The CNFs have a sensitivity of 2.7 μA/mM cm.sup.2 and detection limit of 2 mM. Also provided are methods of preparing the CNF sensors and kit components. Methods of using such CNF sensors for detecting target agents, particularly glucose, are also provided.

Claims

1. A non-enzymatic glucose sensor, comprising an electrode having carbon nanofibers on a transition-metal coated silicon substrate, wherein the carbon nanofibers after growth are not further modified.

2. The sensor of claim 1 wherein the carbon nanofibers have diameters between 13 to 160 nm.

3. The sensor of claim 2 wherein the diameter of the carbon nanofibers is between 60 to 70 nm.

4. The sensor of claim 2, wherein the carbon nanofibers have a density of 40 fibers/μm.sup.2.

5. The sensor of claim 1, wherein the electrode surface area is at least 100 mm.sup.2.

6. The sensor of claim 1 wherein the electrode capacitance is 0.2 μc/mV.

7. A method for detecting glucose in a sample, comprising: contacting one or more sensors of claim 1 with a sample under conditions sufficient to allow the sample to contact the sensors; and detecting glucose, wherein detection of glucose indicates the presence of glucose in the sample.

8. The method of claim 7 wherein the electrode is comprised of carbon nanofibers having diameters between 13 to 160 nm.

9. The method of claim 7 wherein an amount of glucose in the sample is between 10 to 2,000 mg/dL.

10. The method of claim 7 wherein the sample is selected from the group consisting of saliva, blood or urine.

11. A kit comprising: one or more non-enzymatic glucose sensors of claim 1 and at least one table for correlating detected glucose level.

12. A method for detecting hydrogen peroxide in a sample, comprising: contacting one or more sensors of claim 1 with a sample under conditions sufficient to allow the sample to bind to the sensor; detecting hydrogen peroxide, wherein detection of hydrogen peroxide indicates the presence of hydrogen peroxide in the sample.

13. A non-enzymatic hydrogen peroxide sensor, comprising an electrode having carbon nanofibers on a transition-metal and metal oxide coated silicon substrate, wherein the carbon nanofibers after growth are not further modified.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The patent or application contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings will be provided by the Office upon request and payment of the necessary fee.

(2) FIG. 1 illustrates the steps taken to fabricate CNFs.

(3) FIG. 2 provides a Glassy Carbon Electrode.

(4) FIG. 3 illustrates a Carbon Nanofiber Electrode.

(5) FIG. 4 provides a FESEM micrograph of glassy carbon.

(6) FIG. 5 provides FESEM micrographs of CNFs.

(7) FIG. 6 provides FESEM Micrographs of CNFs with increasing magnification. FIG. 6A is ×10,000 magnification; FIG. 6B is ×20,000 magnification; and FIG. 6C is ×35,000 magnification.

(8) FIG. 7 illustrates a CV Plot for glassy carbon electrode.

(9) FIG. 8 illustrates a CV Plot for CNF electrode.

(10) FIG. 9 illustrates a CV Data for GC electrode with no redox couple.

(11) FIG. 10 provides CV Data for CNFs electrode with no redox couple.

(12) FIG. 11 provides an amperometric study of .sub.H2O2 for glassy carbon.

(13) FIG. 12 provides an amperometric study of .sub.H2O2 for CNFs.

(14) FIG. 13 provides a comparison of CNFs and GC for additions of hydrogen peroxide.

(15) FIG. 14 illustrates the optimization of the applied voltage (electrode potential).

(16) FIG. 15 provides cyclic voltammograms for the GC electrode.

(17) FIG. 16 illustrates the results of an amperometric study of GC current response to glucose additions.

(18) FIG. 17 provides the results of a amperometric study of CNFs current response to glucose additions.

(19) FIG. 18 provides detector signals (or charge) versus the glucose concentrations.

(20) FIG. 19 provides a distribution map of where the nickel is presented in the CNF electrode.

(21) FIG. 20 provides a SEM Image of CNFs and the distribution map for the carbon and nickel content.

(22) FIG. 21 illustrates the CV data for CNFs electrode, showing no indication of nickel redox/oxidation peaks.

(23) FIG. 22 illustrates the results of the amperometric study of CNFs electrodes current response to interfering reagents.

(24) FIG. 23 provides a schematic diagram of a glucose sensor.

(25) FIG. 24 provides an embodiment of the glucose sensor.

DETAILED DESCRIPTION OF THE INVENTION

(26) The present disclosure describes sensor/devices, methods, systems, and kits for sensing glucose levels in a fluid using glucose sensors and the fabrication of the same. Specifically, the present disclosure contemplates a highly sensitive non-enzymatic glucose sensing device and a method for fabricating the same.

(27) In one embodiment, the glucose sensor comprises one or more carbon nanofibers and an electrode. The non-enzyme sensors of the invention avoid problems associated with use of conventional enzyme sensors, including temperature sensitivity, degradation of the enzyme leading to short shelf life, high costs, and restrictions on adaptability to field use.

(28) In one embodiment the carbon nanofiber sensor is implantable in an intravenous device, such as an implantable needle-type device or port, suitable for real-time monitoring. Such intravenous device may include any central venous catheter, pulmonary artery catheter, venal probes, peripheral IV catheter, Swan-Ganz catheter, or any other blood management systems.

(29) The present invention incudes a combination of providing on an insulating silicon wafer base plate/substrate a combination of materials, including a thermal oxide on the base plate, a chromium layer, a nickel layer, and a coating of carbon nanofibers. The use of this combination, and in particular the use of carbon nanofibers, allows for the electrochemical oxidization of glucose. In one embodiment, the sensor to detect the electrochemical oxidization is a potentiostat. The potentiostate may be connected to a computer using software installed from the potentiostat vendor.

(30) A representation of the process in making one configuration of the fabrication of the non-enzyme glucose sensor of the invention is shown in FIG. 1. The active material is provided and deposited on an electrode surface or base plate, particularly carbon nanofibers. In this process, the depositing is the carbon nanofibers on one or more transition-metal or metal oxide materials. The transition metal may be titanium and SiO.sub.2 is the metal oxide.

(31) The base plate or substrate can be made of a variety of materials, such as plastics, ceramics, polymers or silicon wafers based upon the application of the electrode. Such selection of the base plate material should take into consideration the end application of the glucose sensor device, for example if the device is implanted in a patient or utilized in an automated detection system. The base plate can be a silicon wafer with a thickness between the ranges of 100 μm to 1000 μm, between 160 μm to 500 μm, or between 200 μm to 300 μm.

(32) Referring to FIG. 1, Methods of fabrication the CNF glucose sensor comprise providing a base plate at step 1. The base plate is fabricated of a material that can be used for high temperature processing, such as graphene. At step 1, one or more layers of thermal oxide is applied to the base plate, with an average thickness of ˜500 nm. The layer of thermal oxide can be applied between 100 nm-1000 nm thick. The resulting oxide-covered silicon wafer is used as the base or substrate (platform) for the growth of CNFs. At step [300], the titanium layer is deposited over the oxide-covered base using an electron beam (e-beam) deposition. At step 2, a chromium layer is deposited over the oxide-covered base. Other techniques for the deposition of the titanium layer, including printing methods or drop casting may be utilized. The layer of titanium can be between 50 nm-500 nm and more preferably is about 200 nm. The chromium layer is approximately 200 nm thick. At step 3, a layer of nickel is deposited on the silicon base. The nickel layer is applied in a range between 20-60 nm. Preferably, the nickel layer is approximately 30 nm thick. The nickel layer is deposited on the substrate using printing techniques know in the art. For example, the nickel layer is applied using e-beam or UV lithography. The nickel layer can be applied in a pattern of dots that could range in size from 100 nm to over 1 μm. At step 4, a layer of copper or zinc is deposited on the silicon base. At step 4, using direct current plasma enhanced chemical vapor deposition (PECVD), a CNFs forest array is grown from the nickel catalyst. The CNFs array is created using a feeder gas C.sub.2H.sub.2 feedstock and NH.sub.3 diluent. The CNFs array may be created using a feeder gas feedstock and diluent at a pressure of ˜4.0 torr and a plasma power of 410.

(33) Method of preparing a CNFs electrode sensor may include an additional step of depositing one or more transition-metals or metal oxides, wherein the transition-metal may be selected from the group consisting of nickel, copper, copper oxide, platinum and zinc.

(34) An actual configuration of the non-enzyme CNFs glucose sensor is shown in FIG. 24. This figure includes images of wires that lead to a potentiostat. The potentiostat may be connected to a computer with software installed from the potentiostat vendor; this combination serves as the data acquisition and display mechanism.

(35) In FIG. 23, a schematic figure of the complete configuration of the glucose sensor is provided, such that the glucose sensor comprises at least one electrode, a working electrode (we), a reference electrode (re), a counter electrode (ce). An energy source is provided as voltage (V) and the energy current (i) is also shown. The individual sensors are then incorporated into a suitable device for testing the glucose level in a fluid. A completed CNFs electrode as prepared using the process presented above is shown in FIG. 3.

(36) The carbon nanofibers using the process provided above, are vertically aligned and range in size between 13 to 160 nm in diameter. The carbon nanofibers may have a diameter between 60 to 70 nm. The sensor may be comprised of carbon nanofibers having a density of 40 fibers/μm.sup.2. The carbon nanofibers may have a height of about one micron.

(37) The non-enzyme sensor is also characterized by having a CNF electrode surface area of at least 100 mm.sup.2. Further, the sensor of the present invention has a capacitance of 0.2 μc/mV. In various embodiments of the invention, it is preferable for the sensor to have a capacitance of at least 0.2 μc/mV.

(38) One or more CNF sensors are contacted with a sample under conditions sufficient to allow the sample to bind to the sensor, wherein detection of glucose indicated the presence of glucose in the sample. The CNF sensors are comprised of nanofibers with diameters between 12 to 160 nm as shown in FIG. 5. The nanofibers may have a density of about 40 fibers/μm.sup.2.

(39) One or more electrodes of the sensor are connected to a glucometer that senses the glucose level in blood. The non-enzymatic glucose sensors of the invention are stable at a wider range of temperature while maintaining a high degree of selectivity and sensitivity. The non-enzymatic CNF electrode sensors of the present invention may be tested using electrochemical techniques such as amperometry to determine a response current proportional to any analyte in a solution. The amperometric measurements using embodiments of the invention can be performed using instruments and readout electronics known in the art for such analysis.

(40) When using CNF sensors of the invention to detect glucose in a body fluid sample from a mammal, one or more CNF sensors are contacted with an aqueous or fluid sample obtained from the mammal. The sensor may be contacted with any bodily fluid sample, including saliva, blood or urine.

(41) The invention also includes a kit comprising: one or more CNF sensors and at least one table for correlating the detected glucose level.

EXAMPLES

(42) The various experiments described herein illustrate the production and characterization of the CNF electrode sensor and non-enzymatic glucose device. These experiments also provide various characterization methods. Further, these experiments demonstrate the electrochemical systems and three electrode systems used in the development of the glucose device. Finally, two electrochemical techniques, cyclic voltammetry (CV) and amperometry also demonstrate the improved sensitivity of the sensors of the present invention.

Example 1: Preparation of CNF Electrode Sensor Using PECVD and Characterization of the Sensor

(43) An exemplary sensor according to the various embodiments of the invention is illustrated in FIG. 1 was prepared. Specifically, the electrode was an array of vertically aligned carbon nanofibers (CNFs) fabricated using plasma enhanced chemical vapor deposition (PECVD).

(44) FIG. 1 shows an example of the steps used in the fabrication of the CNF electrode. As illustrated in FIG. 1, the first step is to use a 4 inch in diameter silicon wafer. The thickness of the silicon wafer can be between the ranges of 100 μm to 1000 μm, even more standard is 160 to 500 μm, typical is 200-300 μm and is covered with thermal oxide, with an average thickness of ˜500 nm. The layer of thermal oxide can be applied between 100-1000 nm thick. This oxide-covered silicon wafer is used as the base or substrate (platform) for the growth of CNFs. Next, an electron beam (e-beam) deposition is used to deposit a chromium layer to the oxide-covered silicon base. The layer of chromium can be between 20-60 nm, with the preferred thickness of approximately 20 nm and most preferred thickness of approximately 30 nm. A layer of nickel used as the catalyst is then deposited on the chromium coated silicon wafer. The nickel layer is approximately 30 nm thick. This nickel film could be patterned in dots that could range from 100 nm to over 1 μm using e-beam and UV lithography, respectively (Koehne et al. 2004). Finally, a CNFs forest array is grown from the nickel catalyst using direct current plasma enhanced chemical vapor deposition (PECVD) in a custom made chamber using a feeder gas C.sub.2H.sub.2 feedstock (22.5 sccm) and NH.sub.3 diluent (80 sccm) at a pressure of ˜4.0 torr and a plasma power of 410 W (Arumugam et al. 2010). Scanning electron micrographs were taken for structural characterization of the CNFs.

(45) Structural and Morphological Characterization Methodology

(46) The CNF electrode sensors were characterized using SEM and EDS. The SEM instrument was a field emission scanning electron microscope (FESEM) (S4800, Hitachi, Pleasanton, Calif.). A field-emission cathode located in the electron gun of an SEM provides narrow probing beams at low or high electron energy, which results in enhanced spatial resolution with minimal sample charging and damage relative to standard SEMs (Field Emission Scanning Electron Microscopy 2015). In this method, the surface on the material being analyzed is bombarded with electrons, which results in a micrograph of the material at high magnetization.

(47) EDS was another technique used with SEM. This technique identifies the elemental composition of the material being imaged by the SEM for all elements whose atomic number is greater than boron with the ability of detection at concentrations of 0.1% (Energy Dispersive Spectroscopy 2015). This method utilizes x-rays to determine the chemical composition of the material. The sample is bombarded with an electron beam from an SEM, which scans across the material surface. This results in excitation of x-rays that are emitted from the sample material. These excited x-rays are specific for each element. The energy of each x-ray photon produced corresponds and relates to a specific element (Energy Dispersive Spectroscopy 2015). The x-rays are sorted and plotted based on their energy then identified and labeled for a specific element (Energy Dispersive Spectroscopy 2015). Distribution maps of the composition of the materials are also produced based on the compositions of each element in the material. This technique was used to determine the presence of nickel in the CNFs sample.

(48) Electrochemical Experimentation

(49) A three electrode system consisting of a working electrode, a reference electrode, and a counter electrode was used for electrochemical testing. The working electrode is the electrode that is sensitive to the analyte's concentration (Harvey 2006). FIGS. 2 and 3 show the working electrodes used for this experimentation, specifically a glassy carbon electrode (GCE) (SPI Supplies) and the CNF electrode, respectively. The reference electrode is the electrode whose potential remains constant, specifically a saturated calomel electrode was used (Princeton Applied Research) that has a potential of +0.2444 V at 25° C. (Harvey 2006). The counter electrode, an electrode that completes the electrical circuit, used in this study was a platinum wire (Sigma Aldrich).

(50) Cyclic Voltammetry

(51) CV was one of the electrochemical techniques used for characterization of the electrodes used in this study. The testing was a three-electrode system, which consisted of a working electrode, a reference electrode, and a counter electrode. The electrodes were connected to a potentiostat. All tests were conducted using a CH Instruments Electrochemical Analyzer potentiostat. The working electrodes were glassy carbon and CNFs. The reference electrode was a saturated calomel (mercury chloride saturated KCl). The counter electrode was platinum wire.

(52) The reagent used for the CV characterization was either 1 mM potassium hexa-cyanoferrate (III) (K.sub.3[Fe(CN).sub.6]) in 1 M potassium chloride or 10 mM phosphate buffered saline (PBS) solution. The testing parameters for CV were as follows: 1) potential range=−0.3 V to 0.7 V, 2) scan rate=0.1 v/s, 3) number of sweeps=6, and 4) sensitivity=10.sup.−1-10.sup.−3 A. The data collected using this technique was current in amperes versus potential in volts. These plots have two discreet peaks, one is referred to as the redox peak and the other is the oxidation peak. These peaks correspond to the reduction and oxidation of the reagent solutions in the forward and reverse of the reaction. The electrode capacitance was also calculated using the CV diagram.

(53) For the CV experiment, an initial voltage of −0.3 V is applied to the working electrode. The corresponding current is recorded. The applied voltage is increased at a rate of 0.1 V until the maximum voltage of 0.7 V is reached. Then, the voltage is decreased at a rate of 0.1 V until the initial voltage of −0.3 V is reached. This process is repeated 6 times.

(54) Amperometry

(55) Amperometry is an electrochemical technique in which a constant potential is applied to an electrode and the current is measured. The same three-electrode system as used for CV testing was also used for the amperometry experiments.

(56) Several experiments were conducted using amperometry. For studies conducted to determine the electrode response to repeated additions of hydrogen peroxide, the three electrode system was configured, the solvent, PBS, was added to the fluid cell; the stir plate was turned on, and 30 μL of 30 vol % hydrogen peroxide (Sigma Aldrich) was added to the cell at 90 s. The testing parameters for amperometry were as follows: 1) potential=−0.2 V, 2) run time=300 s, and 3) sensitivity=10.sup.−3-10.sup.−6 A. This test was designed so that the addition of the concentrated hydrogen peroxide would create 100 mM, 10 mM, or 1 mM hydrogen peroxide solutions.

(57) In the glucose detection experiments, the following process was used: 2940 μL of 10 mM PBS was micropipetted into the fluid liquid cell that was placed on a stir plate with a stir bar for continuous mixing, the reference electrode tip was placed in the 10 mM PBS solution and the counter electrode was placed in the solution. The duration of the test was 900 s and every 180 s a 60 μL aliquot glucose was added to the fluid cell. The glucose solutions were prepared using 10 mM PBS and the concentrations used were 0.1 M, 0.25 M, 0.5 M, and 1 M.

(58) In the selectivity studies, the initials steps were the same as for the glucose detection studies. After the counter electrode was placed in solution, the duration of the test was 900 s. During the experiment, 60 μL of uric acid, ascorbic acid, dopamine and glucose were added at 180 s, 360 s, 540 s and 720 s, respectively.

(59) Electrode Characterization Experimentation

(60) Both the carbon nanofiber electrodes and the glass carbon were tested to determine the physical appearance of the electrodes using scanning electron microscopy and to conduct electrochemical techniques such as cyclic voltammetry and amperometry to determine electrochemical properties of these electrodes using the following methodologies.

(61) Morphological Characterization

(62) Scanning Electron Microscopy (SEM) was conducted using a field emission scanning electron microscope (FESEM) (S4800, Hitachi, Pleasanton, Calif.). GC, which is known as vitreous carbon or glass-like carbon, was selected as one of the electrodes because of its properties, which include low electrical resistance, a broad range of potentials, and good biocompatibility (Lewis et al. 1963). This material is non-graphitizing carbon with amorphous structure and has properties similar to ceramic/glass and graphite (Cowlard and Lewis 1967). FIG. 4 shows the FESEM micrographs for the GC electrode, which shows a flat two dimensional structure. FIGS. 5 and 6 show FESEM micrographs of carbon nanofibers. In FIG. 5, the inset picture shows a high magnification (×50,000) FESEM micrograph of the CNFs after a CV test. It was observed that the nanofibers coalesce after removal of fluid. This observation was consistent with the literature on CNFs (Nguyen-Vu et al. 2006). FIG. 6(a) is a lower magnification (×10,000) of the CNFs. The micrograph shows an array of vertically aligned fibers with varying diameters and slightly variable heights. FIG. 6(b) utilizes a magnification of (×20,000) and shows approximately 40 fibers/μm.sup.2, while FIG. 6(c), an even higher magnification image (×35,000), allows the average fiber diameter to be determined using image analysis revealing a diameter range of 13-160 nm and an average diameter of ˜66 nm and height of one micron.

(63) Electrochemical Activity

(64) Cyclic voltammetry (CV) studies were conducted using the experimental setup previously above. The reagents used in this study were 1 mM ferrocenemethanol in 10 mM phosphate buffered saline (PBS), 2 mM potassium ferricyanide (III) (K.sub.3[Fe(CN).sub.6]) in 1 M potassium chloride (KCl), and 2 mM potassium hexacyanoferrate (II) trihydrate (K.sub.4[Fe(CN).sub.6]) in 1 M KCl. The input settings for the program were the following: 1) initial potential=−0.3V, 2) high potential=0.7 V, 3) low potential=−0.3 V, 4) scan rate=0.1 v/s, 5) number sweeps=6, and 7) sensitivity=10.sup.−1 or 10.sup.−6 A. FIGS. 23 and 24 show the CV graphs for the GCE and the CNF electrode, respectively. The current for the CNF electrode is significantly higher than that for the GCE for the same voltage parameters and suggests that the capacitance for the CNFs is much higher. Two discreet peaks are observed that corresponded to redox and oxidation peaks of a redox/oxidation reaction that occurs during the test. From this data, the peak separation (ΔEp) is calculated using the following equation:
ΔE.sub.p=E.sub.reduction−E.sub.oxidation
ΔE.sub.p is peak separation. Ereduction is the reduction peak. Eoxidation is the oxidation peak. For an ideal reversible reaction, ΔE.sub.p is approximately 59 mv/n at 25° C., where n is the number of electrons being transferred (Mabbot 1983). FIG. 7 provides the ΔE.sub.p for glassy carbon electrode and the CNF electrode were 69 mV and 194 mV, respectively.
Capacitance Characterization

(65) Capacitance is a measure of the ability to store a charge and is a property that has several implications as it relates to electrodes for solution electrochemistry. First, the capacitance has a negative effect on the response time of an electrode such that a higher capacitance corresponds to a longer response time for double layer charging. This characteristic, for the CNFs, is offset by the size of the electrodes because the response time scales inversely with size. For CV experiments, a higher capacitance corresponds to a higher current signal according to the following equation that is applicable when there is no redox couple present to transfer electrons between the solution and the electrode (Nguyen-Vu et al. 2006): C.sub.o=Δi/2v where C.sub.0 is the specific capacitance, Δi is the difference in current between positive and negative potential cycles, v is the scan rate.

(66) CV experiments were conducted using 10 mM PBS solution to determine the capacitance of both the CNF and GC electrodes at a scan rate of 100 mV/s. FIG. 15 provides CV data for the GC electrode, which has a calculated capacitance of ˜0.00625 μC/mV. FIG. 16 provides CV data for the CNF array, which has a calculated capacitance of ˜0.2 μC/mV. The results clearly demonstrate that the capacitance is a factor of 30 greater for the CNF electrodes in comparison to GC electrodes.

(67) Surface Area Determination

(68) As previously mentioned the GC electrode was imaged using SEM and appeared as a flat structure. However, the CNF array was observed to have a three-dimensional structure, which consists of an array of nanofibers on the electrode surface. One important property of the working electrodes is the surface area. Several methods can be used to determine the surface area of a working electrode, including drop weight (or volume), capacitance ratio, Parson-Zobel plot, hydrogen adsorption from solution, oxygen adsorption from solution, underpotential deposition of metals, voltammetry, negative adsorption, ion-exchange capacity, adsorption of probe molecules from solution, mass transfer, adsorption of probe from gas phase, x-ray diffraction, porosity, and microscopy. With great improvements made in microscopy, however, SEM is a good option to determine the surface area. SEM was particularly applicable in the determination of the surface area because the CNF electrodes are vertically aligned and the tops are circular in nature.

(69) Because of the advances in electron microscopy and the vertical alignment of the CNFs, SEM was used to determine the area of the CNF electrode. The area was calculated based on the area exposed by a 4 mm o-ring. The surface area of the CNFs was approximated by assuming a fiber height of 1 μm. Image analysis was then conducted to evaluate the total perimeter represented by the circumference of the CNFs over a measured area. The total perimeter was multiplied by the fiber length to determine the total surface area of the fibers over the corresponding flat area analyzed in the SEM. Finally, the area was scaled to approximate the total electrode area provided by the CNFs, which was approximately 100 mm.sup.2. This represents an approximate order of magnitude increase over the area for the glassy carbon, which was approximately 10 mm.sup.2.

(70) Electrode Response to Hydrogen Peroxide

(71) Amperometric studies were conducted using the previously discussed experimental setup discussed in Example 1. The reagents used in this study were 10 mM PBS solution and 30 vol % hydrogen peroxide. The input settings for the program were the following: potential=−0.2 V, run time=300 s, and sensitivity=10.sup.−3 A.

(72) For the hydrogen peroxide studies, an aliquot of hydrogen peroxide was added at 90 s and run for 5 min. FIG. 11 illustrates the one step study comparing 1 mM, 10 mM, and 100 mM concentrations for the glassy carbon. FIG. 28 demonstrates the one step study comparing 1 mM, 10 mM, and 100 mM concentrations for the CNFs. The results for both electrodes show a current response to the addition of hydrogen peroxide, which is in agreement with the following reaction:
H.sub.2O.sub.2 (aq).fwdarw.2H+(aq)+O2(aq)+2e−
In this reaction, hydrogen peroxide is oxidized giving up two electrons. This was observed by an increase in the current in the amperometric data for both electrodes. In spite of having a higher capacitance, however, the response time of the CNFs is much faster. Its response is an instant current response that was observed after the addition of the peroxide. This improved response is due to the nanoscale size of the carbon nanofibers. In contrast, the GC electrode did not respond until after 10 s.

(73) FIGS. 11 and 12 demonstrate that the more concentrated the hydrogen peroxide solution being added to the cell the larger the current response. FIG. 13 shows the electrode comparisons between CNFs and GC electrodes for the additions of hydrogen peroxide at 90 s. The results clearly show that the CNF responses to hydrogen peroxide were much larger than those of the GC, corresponding to sensitivities of 1900 μA/M and 6.3 μA/M, respectively. Moreover, with the exception of ΔEp, all aspects of the characterization, including the capacitance, surface area and hydrogen peroxide response, indicate the CNF electrode has improved performance over the more traditional GCE for solution phase electrochemical detection of glucose.

Example 2: Non-Enzymatic Detection of Glucose

(74) Carbon nanofiber electrodes and glass carbon electrode were tested in a series of experiments to compare the selectivity of both electrodes and to demonstrate the non-enzymatic detection of glucose using carbon nanofibers.

(75) Experimental Methodology:

(76) An amperometric electrochemical technique was used for glucose detection. Also, before every experiment, a CV was run to ensure a proper electrical connection of the three electrode system discussed above.

(77) The following is a brief description of the experimental procedure:

(78) First, 2,940 μL of 10 mM phosphate buffered saline (PBS) was micropipetted into the liquid cell (as provided above) and stirred continuously; the calomel (mercury chloride saturated potassium chloride) reference electrode tip was placed in the 10 mM PBS solution. Second, a platinum wire, used as the counter electrode, was placed in the solution. A voltage was applied to the working electrode to begin the experiment.

(79) The duration of the experiment was 900 s and every 180 s a 60 μL aliquot of glucose was added to the fluid cell. The glucose solutions were prepared using 10 mM PBS and 0.1 M, 0.25 M, 0.5 M, and 1 M glucose concentrations.

(80) Determination of the Applied Voltage

(81) The applied voltage was optimized for the amperometric studies. The following voltages evaluated: −900 mV, −800 mV, −500 mV, −200 mV, 100 mV and 400 mV. As shown in FIG. 14, at approximately −500 mV, the signal strength plateaus and does not increase significantly even though the absolute magnitude of the potential increases to −900 mV. Consequently, −500 mV was chosen for the study to maximize the signal while not choosing a higher absolute magnitude of potential that does not provide any GC electrodes served as a benchmark carbon material for the investigation of CNFs. FIG. 15 shows three cycles of CV data for the GC electrode with a capacitance of ˜0.00625 μC/mV. These results establish that the CNF electrode is stable over multiple cycles and a good electrical connection is achieved. FIG. 16 shows the amperometric results of the glucose addition at 180 s periods for the GC electrode. These results clearly show that the GC electrode showed no current response with the addition of glucose solution.

(82) Carbon NanoFiber Detection

(83) As previously stated, the GC electrode was used as a benchmark for the study of the primary CNF electrode. Although the GC electrode showed no response to glucose, the CNF electrode successfully detected glucose and the signal response was linear. Nonetheless, several points of interest are provided to describe certain challenges.

(84) Detection Demonstrated

(85) As shown in FIG. 17, the results of the current response of the CNFs to the addition of glucose at stock concentrations of 0.1, 0.25, 0.5 and 1 M are provided. In FIG. 17, raw signals from the sensor are provided. The results show an instantaneous and sharp increase in current with the addition of glucose. The results suggest that the electrode is stable and the experiment is reproducible. The intensity of the peaks increases as the glucose concentration increases suggesting that the electrode not only detects the presence of glucose but also responds such that it can be calibrated to determine the glucose concentration.

(86) FIG. 19 shows the detector signal versus the glucose concentration. This graph was prepared by first calculating the concentration of glucose in the cell based on the concentration and volume of the stock solution added to the test cell as well as the final volume of liquid in the cell. These values represent the x-axis coordinates for the data shown in FIG. 19. Next, the area under the signal peak was integrated to determine the charge (μC) corresponding to a particular dose of glucose. These values represent the y-axis coordinates for the data shown. The graph shows a good linear relationship between the detector signal and the glucose concentration. Although a linear signal is observed, the uncertainty in the measurement is fairly large at low concentrations. This is attributed to the variability that remains with the improved dosing method. Additionally, use of adequate mixing creates additional sources of noise in the signal that ultimately increases uncertainty with the measurement.

(87) Detection Mechanism

(88) Two mechanisms were considered to explain the non-enzymatic detection of glucose: nickel catalysis and increased reaction rate due to the high surface area of the CNFs. As previously described, the CNFs were grown from a nickel catalyst. This is important because, in the non-enzymatic glucose detection literature (Dhara et al. 2014; Kiani et al. 2014; Shervedani et al. 2014; Tarlani et al. 2014), there are examples of testing platforms that contain nickel in the electrode that have detected glucose without the aid of the glucose oxidase enzyme. The explanation is that the reaction is catalyzed by a nickel redox reaction. For this reason, nickel catalysis was the first mechanism considered. To evaluate this mechanism, EDS studies were conducted to confirm the presence of nickel in the electrode. The EDS results shown in Table 1 and graphically in FIG. 19 indicate that the materials present in the sample are as follows: 68.51% carbon, 28.62% silicon, 1.60% nickel, 0.71% oxygen and 0.56% titanium. FIG. 20 provides a distribution map which illustrates where the nickel is present in the material; this is observed by the bright dots. The mapping distribution was completed for carbon as well and the presence of carbon was observed by the intense bright dots. Thus, the presence of nickel in the electrode was confirmed.

(89) TABLE-US-00001 TABLE 1 EDS Composition of CNFs Material App Intensity Weight % Element Conc. Correlation Weight % Sigma Atomic % C 19.12 0.3315 68.51 0.35 83.81 O 0.18 0.2994 0.71 0.15 0.65 Si 25.16 1.0441 28.62 0.31 14.97 Ti 0.37 0.7881 0.56 0.04 0.17 Ni 1.07 0.7978 1.60 0.08 0.4 Totals 100

(90) Next, the nickel redox reaction was evaluated using cyclic voltammetry. The results reflected in FIG. 21 clearly show no indication of redox/oxidation peaks for the presence of nickel.

(91) Without direct evidence of the nickel catalysis mechanism, the second mechanism of a faster reaction rate facilitated by a high surface area was evaluated to explain the ability of this material to oxidize glucose without the aid of the enzyme glucose oxidase. The surface area of the GC electrode is 12.5 mm.sup.2 whereas the surface area of the CNF electrode is 105 mm.sup.2. The glucose oxidation is thermodynamically favorable, but it is typically too slow to observe without a biological catalyst such as glucose oxidase. In the case of the CNF sensor, the evidence supports that the area is large enough to help overcome the slow reaction rate in the absence of a catalyst.

(92) Sensitivity Determination

(93) The electrode sensitivity was calculated using the following equation (Dhara et al. 2014; Kiani et al. 2014; Shervedani et al. 2014; Tarlani et al. 2014): Sensitivity=m/A, where m=the slope from the graph of the detector signal versus the glucose concentration and A=the two dimensional surface area of the CNF electrodes.

(94) Based on the above equation, the sensitivity of the CNFs is ˜795.2 μC/mM-cm.sup.2. However, a peak current was observed in contrast to the limiting current that is often observed in studies from the literature. Integrating under the peak allowed the signal to be correlated to glucose concentration, but the sensitivity has different units than what is typical in the literature. To address this lack of consistency, in addition to the aforementioned sensitivity calculation, the peak current value was substituted for the limiting current to calculate sensitivity in comparable units. Using this approach, the sensitivity of the CNFs is 2.7 μA/mM cm.sup.2.

(95) Detection Limit

(96) The following equation was used to determine the signal to noise ratio (Ripp, 1996):

(97) $\left(\begin{array}{c}S\\ N\end{array}\right)=X_{\mathrm{AVE}}/s$
where X.sub.AVE=average of either the calculated concentrations or analytical signals, s=standard deviation.

(98) The lowest concentration of glucose solution added to the testing cell that showed a current was 0.1 M corresponding to 2 mM after being added to the testing cell. Accordingly, the detection limit is 2 mM (S/N=1).

(99) Selectivity Experimentation

(100) For the selectivity experiments, three known interfering reagents, uric acid (UA), ascorbic acid (AA), and dopamine (DA) were evaluated. These reagents are known to have similar electroactivities to glucose, which causes interference in the detection of glucose (Kani et al., 2014). However, the concentration of glucose in the human blood is approximately 30 times the concentration of AA, UA and DA (Wang et al. 2009). Thus, the interfering agents were added in quantities with physiological relevance. FIG. 22 shows the selectivity results where 0.1 M of AA (120 s), UA (240 s), DA (360 s), and 1 M glucose solution (480 s) were added. The results showed that the interfering reagents showed no current response but a signal was observed with the addition of glucose solution, which suggests that the electrode is selective for glucose in the presence of interfering compounds at concentrations that are physiologically relevant.

(101) The various preferred embodiments and experiments having thus been described, those skilled in the art will readily appreciate that various modifications and variations can be made to the aforementioned preferred embodiments without departing from the spirit and scope of the invention. The invention thus will only be limited to the claims as ultimately granted.