Novel 1D/2D KWO/Ti₃C₂T_x Nanocomposite-Based Acetone Sensor for Diabetes Prevention and Monitoring

Abstract

The acetone content in the exhaled breath of individuals as a biomarker of diabetes has become widely studied as a non-invasive means of quantifying blood glucose levels. This calls for development of sensors for the quantitative analysis of trace concentration of acetone, which is presents in the human exhaled breath. Traditional gas detection systems, such as the Gas Chromatography/Mass Spectrometry and several types of chemiresistive sensors are currently being used for this purpose. However, these systems are known to have limitations of size, cost, response time, operating conditions, and consistent accuracy. An ideal breath acetone sensor should provide solutions to overcome the above limitations and provide good stability and reliability. It should be a simple and portable detection system of good sensitivity, selectivity that is low in terms of both cost and power consumption. To achieve this goal, in this paper, we report a new sensing nanomaterial made by nanocomposite, 1D KWO (K₂W₇O₂₂) nanorods/2D Ti₃C₂T_x nanosheets, as the key component to design an acetone sensor. The preliminary result exhibits that the new nanocomposite has an improved response to acetone, with 10 times higher sensitivity comparing to KWO-based sensor, much better tolerance of humidity interference and enhanced stability for multiple months. By comparing with other nanomaterials: Ti₃C₂, KWO, and KWO/Ti₃C₂T_x nanocomposites with variable ratio of KWO and Ti₃C₂T_x from 1:1, 1:2, 1:5, 2:1, 4:1, and 9:1, the initial results confirm the potential of the novel KWO/Ti₃C₂ (2:1) nanocomposite to be an excellent sensing material for application in sensitive and selective detection of breath acetone for diabetics health care and prevention.

1. Introduction

The number of people living with diabetes in the world has continued to increase, with nearly 463 million adults between ages 20–79, as reported by the International Diabetes Federation (IDF) [1]. Future projections estimate this number can be increased to about 700 million people by 2045 [1]. This means that if urgent actions are not taken, diabetes will become a global health emergency to impact on mortality rates and economic productivity. Frequent monitoring of blood glucose levels is a crucial step in the management of diabetes and extensively prevents exacerbation of the disease. In recent years, a variety of apparatuses based on blood glucose measurement have been introduced into practice (i.e., blood sensors and urinalysis), although some are still under their developing stages, such as breath-based sensors.

Studies indicate that the blood glucose level is closely related to the concentration of acetone in exhaled breath [2]. Breath acetone or dimethyl ketone is a volatile organic compound and known for its rotten apple smell, as described by physicians [3], which is physiologically produced in the body as one of the three by-products of ketoacidosis—when the liver breaks down fats to produce energy in response to the body’s inability to utilize glucose [3,4]. Given this reason, elevated level of acetone in the exhaled breath can be used as the biomarker to diagnose diabetes. Generally, the concentration of breath acetone levels for healthy person is between 0.4 ppm (parts-per-million) to 0.9 ppm while 1.8 ppm and above for diabetics [5]. Individuals with acetone levels between 0.9 and 1.8 are considered to be at risk of developing diabetes in the future [5,6,7]. This paper reports a new portable sensor device based a novel sensing nanomaterial, KWO/Ti₃C₂T_x nanocomposites, for sensitive and selective detection of trace concentration acetone, as a potential simple, low-cost, portable, non-invasive, and user-friendly tool to help diabetics monitoring their daily health condition and preventing the complications caused by this disease.

Several techniques have been developed to determine acetone concentration in exhaled breath. One of the conventional methods used for quantifying breath acetone is Gas Chromatograph (GC) paired with a Flame Ion Detection (FID) or Mass Spectrometry (MS) incorporated [8,9]. Such techniques can provide accurate measurements of breath acetone but are bulky, heavy, expensive, complicated, and time-consuming [10]. Real-time detection of acetone in exhaled breath has been made possible with the use of proton transfer reaction-mass spectrometry (PTR-MS) and the selected ion flow-mass spectrometry (SIFT-MS). However, they are still costly, bulky and require frequent calibration [11]. In order to overcome above limitations associated with size, cost, and complexity of these quantitative analytical techniques, researchers started to investigate new sensing techniques by using metal oxide semiconductors (MOS) to detect breath acetone. In particular, the nanostructured MOS-based chemiresistive sensors are now widely used for gas sensors due to their attractive electrical and material properties which enable detection of target gases [12,13,14,15,16,17,18,19,20,21,22] with enhanced sensitivity, fast response, low cost and capability of miniaturization. The sensing mechanism of chemiresistive response works by changing resistance of the sensing material in response to surface adsorption and desorption of gas molecules. This interaction leads to the formation of an electron-depleted region. That is, the gas detection in semiconductor-based sensors is related to ionosorbed surface oxygen and target analyte gas that causes a change in the conductivity of the semiconductor materials [23,24].

Table 1. Selected Metal Oxide Semiconductors as acetone sensor.

Material	Principle of Operation Device Type	Lowest Concentration Detected (ppm)	Response Time	Operation Temperature
ZnO-CuO [13]	Resistance change Core-hollow cube	0.04 ppm	5.59 s for 0.5 ppm	200 °C
In2O3 [14]	Resistance change Nanowire	25 ppm	~10 s (in N2)	400 °C
InN [15]	Resistance change Thin Films	0.4 ppm	150 s for 10 ppm (in air)	200 °C
GaN [16]	Resistance change thin Films	500 ppm	10 s for 1000 ppm (in air)	350 °C
WO3 [16]	Resistance change Nanoparticles	0.2 ppm	~3.5 m	400 °C
WO3 Fibers w/Pt [17]	Resistance change Nanoparticles	0.12 ppm	5 min (in air)	300 °C
In/WO3-SnO2 [18]	Resistance change thin films	Verify	Verify	200 °C
ZnO [19]	Resistance change thin film	100 ppm	30 s	200 °C
Fe2O3 [20]	Resistance change Thin Film	500 ppm	33 s (in air)	275 °C
TiO2 [21]	Resistance change Thin Film	1 ppm	10 s (in air)	500 °C
K2W7O22 [22]	Resistance change Thin Film	0.5 ppm	~30 s *	25 °C

* our work.

lists current MOS-based chemiresistive sensors which have been reported to detect acetone. These semiconductors include Tungsten (VI) oxide (WO₃), Tin(IV) Oxide (SnO₂), Zinc oxide (ZnO), Titanium dioxide (TiO₂), Iron (III) oxide(Fe₂O₃), Indium (III) oxide (In₂O₃), Copper (II) oxide (CuO), ZnO-CuO, and KWO (our work) [13,14,15,16,17,18,19,20,21,22].

The performance of these sensing materials can be improved either through modification of sensor device configuration/structure or combination with other nanomaterials to form nanocomposites in order to facilitate interaction with gas acetone molecules. For example, ZnO-CuO core-hollow cube nanostructures-based p-n heterojunction chemiresistive sensor exhibits a remarkable response to acetone at 200 °C. The lowest concentration can be detected down to 0.04 ppm with high selectivity and excellent stability for up to 40 days [13]. However, all these sensors are operating at an elevated temperature at least (200 °C), which ultimately increase the power requirements and overall cost of the devices. In contrast, the as-synthesized K₂W₇O₂₂ (KWO), a new functionalized semiconductor nanomaterial in our group, has been validated to be able to operate at room temperature with sensitive and selective response to acetone [22]. This is because KWO-based sensing mechanism is different than traditional MOS-based chemiresistive sensors, due to its unique material and surface properties. Detailed discussion of KWO sensing acetone can be found somewhere else [22,25,26,27,28,29,30]. The brief summary in this paper focuses on the p-type semiconductor property, high surface area and porous morphology, and specifically the room-temperature ferroelectric property. All these structural and material properties enable nanostructured KWO to effectively interact with the high dipole-moment compound, acetone, thereby causing a significant change of the resistance [22,25,26,27,28,29,30]. However, the results indicate that water vapor can cause a strong interference and baseline shift to weaken KWO sensing performance [30]. We need to find a solution which can effectively address the water vapor interference and baseline shift in KWO-based sensors.

Currently, our group has successfully synthesized a new 2-dimentional (2D) nanomaterial, Ti₃C₂T_x (T_x stands for OH⁻, –O, and F⁻¹ surface terminated groups). MXene, named from the general formula of M_n+1X_n (n = 2, M = Ti, and X = C) [31, 32]. Due to the unique multi-layered structure with extreme large surface and interface area, metallic conductivity, and flexible surface functionality, 2D MXene Ti₃C₂T_x, attracts a lot of attention in the field of biomedical sensing and energy storage with high signal to noise ratio [31,32,33,34,35,36,37]. Motivated by these findings, here, we aim to explore the potential of using 1D KWO and 2D Ti₃C₂T_x nanocomposite as the new sensing material for acetone detection which is low cost, highly sensitive and selective, low noise, low interference from water vapor, and room-temperature operation. The details are described below.

2. Experiments

2.1. Material Synthesis

The KWO sensor nanomaterial was synthesized using the hydrothermal method which has been described in our previous report [22]. Briefly, a precursor solution containing Na₂WO₄·2H₂O (95%, Alfa Aesar, Haverhill, MA, USA), oxalic acid dihydrate (>99%, VWR, Radnor, PA, USA), K₂SO₄ (>99%, VWR, Radnor, PA, USA) and HCl (36%–38%, Aqua Solutions Inc. Deer Park, TX, USA) was made. This solution was then put into a 30 ml autoclave for synthesis. KWO samples were grown at 225 °C for 24 h. Sensor materials used for acetone sensing tests were made by blade coating and then annealed at 350 °C. X-ray diffraction (XRD) was obtained using a Bruker AXS D8 Discover to study as-synthesized KWO crystalline structure (Bruker, Billerica, MA, USA). Sensors were prepared by coating a paste made from KWO and ethanol on glass substrates. A diffraction pattern was gathered from a 2-Theta of 5° to 90°. Scanning electron microscopy (SEM) image was obtained using a JEOL JEM-2100 high-resolution analytical TEM (JEOL Ltd., Tokyo, Japan).

Pure Ti₃C₂T_x MXene was synthesized based on an Hydrofluoric (HF) etching method [38,39]. Our group has recently developed a “hot etching” method which can make high-yield, high purity of Ti₃C₂T_x MXene [40]. In detail, 1) a precursor, so-called Ti₃AlC₂ MAX phase was synthesized by first ball milling TiC, Ti, and Al powders for 5 h. The TiC:Ti:Al molar ratio was 2:1:1.2, respectively. Excess Al powder was used due to reporting of an Al deficiency even when used in proper molar ratio due to unknown reasons. The resulting powder is then pressed into a pellet and sintered at 1350 °C for 4 h under argon flow. The collected pellet is then milled back into powder and sieved through a 160-mesh sieve. The MAX powder was then collected and used in the high temperature etching experiments. A total of 0.5 g of MAX phase was etched using HF acid in a 25 ml Teflon lined autoclave at temperature of 150 °C for five hours in a Thermolyne furnace. HF concentration of 5%wt was used to remove Al-layers from the MAX phase. Materials were then sonicated for 1 hour using a sonicating bath and collected via centrifuge. All materials were then dried overnight at 65 °C in a drying oven.

The nanocomposite film made by KWO and Ti₃C₂T_x MXene was formed with variable material ratios of KWO/Ti₃C₂T_x–1:1, 1:2, 1:5, 2:1, 4:1, and 9:1 using electrostatic self-assembly technique. In detail, amounts of as-synthesized KWO nanorods were dispersed into distilled water and sonicated for 1h. Then, 10mg of as-synthesized Ti₃C₂T_x nanosheets were added to 10 mL distilled water and sonicated for 1h. Thereafter, the homogeneous KWO solution was poured into the Ti₃C₂T_x solution. After that, the mixture was stirred vigorously for 12 h. Finally, the obtained KWO/Ti₃C₂T_xpowder was washed three times with distilled water and dried at 80 °C for 24 h [41]; Finally, eight as-synthesized nanomaterials including Ti₃C₂T_x, KWO, and KWO/Ti₃C₂T_x (1:1, 1:2, 1:5, 2:1, 4:1, and 9:1) were made into thin films using drop-casting method on the gold-electrode patterned glass substrates.

2.2. Sensing Test System

For a typical chemiresistive device, the sensitivity, as described in equation (1), is calculated about the resistance change once the sensing film exposed to the analyte gas. A couple of factors have previously been reported on KWO sensing acetone to significantly influence device performance, such as the shift of baseline resistance of the film and interference of relative humidity (RH) [22,26,30]. Hence, the sensing system for this study, was focusing on how to improve the performance of as-reported acetone sensor with less interference of humidity and long-term stability.

The acetone vapor and RH can be generated and controlled through an OVG-4 vapor generator and humidity generator, OHG-4 (Owlstone Inc., Westport, CT, USA), as shown in Figure 1. The vapor generator was designed to generate trace concentrations of gases precisely. For example, using OVG-4 vapor generator, we can obtain variable concentrations of acetone vapor ranging from 0.1 ppm to 5.6 ppm with controlled RH levels from 8% up to 85%. The sensing test has been conducted to detect variable concentration of acetone at a certain RH under the room temperature. The change in resistance across the film was measured using the Keithly electrometer. RH and temperature were monitored using a commercialized sensor. The concentrations of acetone vapor used in this study is 2.85 ppm with 10%, 50% and 70% RH conditions, respectively, at the room temperature. To conduct the experiment, baseline resistance of the sensors at room RH (between 20%–40%) was measured and recorded, as R1. The generated acetone and water vapor are delivered through carrier gas from OVG-4 and OHG into the sensing chamber.

The sensitivity of the sensor films to acetone at a given RH value can be calculated as the percentage change in resistance of the sensing material with and without exposure of acetone:

$$S\%= \left(\frac{R3-R2}{R1}\right)\times 100$$

(1)

Here, S is sensitivity (%) to acetone, R1 is the baseline resistance (MΩ) at room RH, R2 is the resistance (MΩ) at the controlled RH without acetone in-flow and R3 is the resistance (MΩ) at the same controlled RH with acetone in-flow.

3. Results and Discussion

3.1. Characterization

The microstructures and morphologies of the as-synthesized nanomaterials: KWO, Ti₃C₂T_x, KWO/Ti₃C₂T_x nanocomposites have been investigated by SEM. As shown in Figure 2a, the SEM image of KWO clearly shows a three-dimensional mesh of randomly orientated and interconnected nanorods with several µm in length and 20–50 nm in diameter. The SEM image of pristine Ti₃C₂T_x MXene (Figure 2b) exhibits multilayered structure and accordion-like morphology. Figure 2c–f show the SEM images of the KWO/Ti₃C₂T_x nanocomposites at the ratio 2:1, 1:1, 4:1, and 9:1, which all indicate a rougher surface due to the decoration of the KWO nanorods on the Ti₃C₂T_x nanosheets and illustrate the successful construction of the 1D/2D KWO/Ti₃C₂T_x hybrids. This could suggest a strongly coupled interaction between KWO and Ti₃C₂T_x.

The XRD patterns of the as-synthesized samples (KWO, Ti₃C₂T_x MXene, and KWO/Ti₃C₂T_x nanocomposites) are shown in Figure 3. The peak at 2θ = 6.2° is corresponding to the (002) facet of Ti₃C₂T_x. While the peaks at 2θ = 23°, 28° correspond to (002) and (200) of KWO. In our previous reports [22,23,24], the relative peaks (002)/(200) play an important role to realize the selective detection of acetone. The XRD spectrum of KWO/Ti₃C₂T_x (2:1) nanocomposite has shown the strongest relative peak at (002). In addition, nanocomposites show higher crystalline structure while the ratio of KWO/Ti₃C₂T_x is higher. Beside these, the much weaker intensity of MXene observed in XRD indicates that the content of the Ti₃C₂T_x in the composite is very low. The low content of Ti₃C₂T_x in the nanocomposites means Ti₃C₂T_x will not cause dramatic change of material properties, e.g., p-type semiconducting, and ferroelectricity, but KWO has interacted with Ti₃C₂T_x in the nanocomposites and caused the change of their crystal facets.

3.2. Sensing Tests

The sensing performance of as-fabricated device to detect acetone is based on a room-temperature chemiresistive response. That is, the detection and sensing response of acetone largely depend on sensing material’s chemical, electrical and structural properties. These properties play an important role to affect charge transfer between sensing material and detected gas, and electrons/holes diffusion within the nanocomposite. This results in a change in resistance of sensing material [38,39]. As shown in Figure 4, comparing to other nanomaterials, the KWO/Ti₃C₂T_x (2:1) nanocomposite exhibits the highest sensitivity (almost 10 times higher than the response of KWO) to 2.86 ppm acetone at the same sensing condition, RH = 20% and room temperature. Although the detailed sensing mechanism is still under investigation, the improved sensing performance of KWO/Ti₃C₂T_x (2:1) nanocomposite can be considered due to the contributions of (1) the unique surface and electric properties of Ti₃C₂T_x nanosheets [38,42], which can provide better signal to noise ratio and higher sensitivity, (2) the room temperature of ferroelectric property of KWO [22], and (3) the interaction between KWO and Ti₃C₂T_x makes stronger (002) facet. Here, 2D multilayered Ti₃C₂T_x nanosheets not only provide extreme large surface and interface area for the absorption of acetone molecules but also the rich surface functional groups, T_x (–O, –OH, and –F). These can be the active sites for the nucleation and growth of the KWO nanorods to enhance the interfacial interactions between KWO and Ti₃C₂T_x and expedite charge transfer and transportation.

To evaluate the tolerance of humidity, the KWO/Ti₃C₂T_x (2:1) nanocomposite-based sensor has been tested under different level of relative humidity (RH) at the room temperature. The result is shown in Figure 5. The nanocomposite, KWO/Ti₃C₂T_x (2:1) nanocomposite shows relative stable baseline resistance (red color) while RH is from 10% up to 86%. Considering the less influence of RH but much higher sensitivity to acetone, KWO/Ti₃C₂T_x (2:1) nanocomposite shows the best sensing response to acetone at room temperature. The selectivity of KWO/Ti₃C₂T_x (2:1) nanocomposite-based sensor can be evaluated via referring to the selectivity of KWO-based sensors, which has been studied in other work [24].

The potential sensing mechanism of KWO/Ti₃C₂T_x nanocomposite (2:1) to detect acetone, can be explained: (1) 1D KWO nanorods is a p-type semiconductor [22]. Although the nanocomposite has changed the crystal structure of the material, the Hall Effect measurement reveals that KWO/Ti₃C₂T_x nanocomposites still are p-type semiconducting properties. The XRD spectra further clearly show that the content of Ti₃C₂T_x is much lower comparing to the content of KWO in nanocomposites. Therefore, the property of material semiconducting in nanocomposites has no change. This means the majority carriers in nanocomposites are still holes. Once electron rich compounds, such as acetone, absorb onto KWO/Ti₃C₂T_x, acetone molecules will donate electrons and then combine with holes to result in the increase of resistance of the sensing materials. (2) The KWO room-temperature ferroelectric property is still the major reason to cause an effective charge transfer between KWO and acetone and realize selective detection of acetone [22,27,28,29]. (3) The high electrical conductivity of Ti₃C₂T_x MXene and the interfacial interaction between Ti₃C₂T_xand KWO not only can lower the signal noise, but also can further enhance the charge transfer within the material. In other words, Ti₃C₂T_x MXenes have been used as an efficient charge transfer material for biosensors. Herein, it is believed they play a similar role in the KWO/Ti₃C₂T_x material sensing acetone. While they likely do not contribute much to the interaction with acetone, the Ti₃C₂T_x MXenes help to lower the baseline resistance of the sensor film and provide a much higher signal to noise ratio. In other words, the Ti₃C₂T_x MXene works as a type of electron collector which can help KWO more efficiently transfer charges while interacting with acetone. More study into the mechanism at contact between the two materials should be studied in order to fortify this theory.

In a word, the KWO/Ti₃C₂T_x (2:1) nanocomposite is a new functionalized sensing material with great potential for application in the breath acetone sensor device. This device can deliver a low-cost, non-invasive, high accuracy, and portable device, with less interference of humidity, for early and rapid diabetes detection and long-term health monitoring with extreme high signal-to-noise ratio (SNR).

4. Conclusions

In summary, a new functionalized nanocomposite made by 1D nanorods, KWO, and 2D nanosheets, Ti₃C₂T_x nanosheets, with ratio of 2:1, has been synthesized and used to detect acetone, the breath biomarker of diabetes. The experimental results exhibit significant improvement of sensing response of KWO/Ti₃C₂T_x (2:1) nanocomposite to acetone. The sensitivity to detect acetone at room temperature is almost 10 times higher than the sensitivity of as-reported nanostructured KWO. In particular, the new nanocomposite, 1D/2D KWO/Ti₃C₂T_x shows much less interference with humidity with extreme high signal-to-noise ratio (SNR) and capability of miniaturization and integration. A proposed sensing mechanism based on the p-type semiconducting property, ferroelectricity, extremely large surface, and much better conductivity has been discussed. Although more systematic understanding of the sensing mechanism is still under investigation, the initial study indicates KWO/Ti₃C₂T_x (2:1) nanocomposite is a very promising new sensing material for application in acetone detection and diabetes care.