Fabrication of aligned carbon nanotubes and graphene materials for biosensor applications

oncentrations. 6 Figure 2.8: TEM images of two VA-CNTs growth in the same CVD condition in two cases: a) without water vapor; b) with water vapor The SEM results (fig. 2.7) show that the addition of water vapor during CVD has significantly altered the length, diameter and growth rate of CNTs. The length of the CNTs increased from 6.5 μm to 40.5 μm corresponding without and with water vapor, respectively (corresponding to the growth rate of CNTs increased from 200 nm/min to 1330 nm/min). In addition, the density of CNTs also increased and CNTs became more uniform when water vapor was introduced during the CVD process. Figure 2.8 is a TEM image of two samples of VA-CNTs synthesized without water vapor (Fig. 2.8a) and with water vapor at 60 sccm during CVD process (Fig. 2.8b). The TEM images can clearly distinguish amorphous carbon and structural defects of carbon nanotubes. During CNTs growth, under certain conditions, CNTs are formed in a bamboo structure (as bamboo crops) which are considered undesirable structural defects. Unlike the CNTs grown with H2O vapor, the CNTs have a hollow structure, straight, thin tube, small diameter and uniform. The analysis of Raman spectra evidences increased graphitization of CNTs samples with addition of water vapor during CVD process. In addition, the effect of water vapor flow rate on the growth rate and properties of the VA-CNTs was also investigated. We have grown VA-CNTs using 0.033 g.mL-1 CoFe1.5O4 (M3) catalyst, in the same CVD condition: at 750°C, Ar/H2/C2H2 = 300/100/30 sccm, and CVD time 30 min, with different water vapor flow rates: 20 sccm, 40 sccm, 60 sccm 7 and 80 sccm. The SEM results (Fig. 2.10) show that at 60 sccm of water vapor flow rate the length, the density of CNTs is highest and the orentiation of CNTs is also best among four water vapor flow rate investigated. 2.2.3.3. Influence of the ratio of catalyst components In this study, we have grown VA-CNTs using four cobalt ferrit samples M1, M2, M3, M4 at different ratios of Co2+:Fe3+ = x:y precursors, in the same 0.033 g.mL-1 concenstration and in the same CVD condition. The results of SEM (fig. 2.12) show that adding the Co2+ component to the catalyst mixture plays a very important role in improving the growth rate, length, density or yield of the VA–CNTs. The highest length of the VA-CNTs was 128.3 ± a) b ) c ) d ) Figure 2.10: SEM images of VA-CNTs samples VA-CNTs growth from CoFe1,5O4 (M1) 0.033 g.mL-1catalyst in the same CVD condition with with different water vapor flow rate Figure 2.12: SEM images of VA-CNTs growth from 04 catalysts with component ratios Co2+:Fe3+ = x : y: a) x:y =0:3, b) x:y =1:2, c) x:y =1:1.5, d) x:y =1:1, respectively, in the same CVD condition. 8 5.5 μm on the (M3) catalyst with a Co2+:Fe3+ ratio of 1:1.5 (corresponding to a 40% added Co2+) which is significantly higher than that of the non-Co2+ (M1) catalyst sample. The density of VA-CNTs that are grown on the (M3) catalyst are also much higher than those of others VA-CNTs. This is explained by the differences in physical properties such as transition temperature, melting temperature, and this causes the metal particles to separate, reducing the diffusion and agglomeration of small catalyst particles into large particles of catalyst particles at a high temperature in CVD conditions. This makes the catalyst particles disperse more evenly than the surface of the substrate and keeping the initial small size of the catalyst particle, facilitating the formation and development of the VA-CNT material. However, if too much Co2+ content was added, which meant that the proportion of Fe3+ was reduced, the height and density of CNTs decreased (fig. 2.12d), which reduced the yield of VA- CNTs. 2.3. Fabrication of horizontally aligned CNTs (HA-CNTs) 2.3.1. Preparation of substrate and catalyst material We used a silicon wafer with a 90 nm thick-SiO2 as substrate for the catalysts to fabricate HA-CNTs. We used FeCl3.6H2O salts as catalyst precursors. The salt was dispersed with deionized water at different concentrations of 0.1M, 0.01M, and 0.001M. The solution was then deposited on the clean silicon substrate by spin-coating at a spin rate of 6000 rpm. Figure 2.18: Fabrication process HA- CNTs material by thermal CVD method. 9 2.3.2. Fabrication process of HA-CNTs The procedure and steps for fabrication of HA-CNTs by thermal CVD method are divided into 4 stages as described in fig. 2.18. 2.3.3. Fabrication results of HA-CNTs 2.3.3.1. Influence of catalyst solution concentration We have growth HA-CNTs using a solution of FeCl3 with different solution concentrations: 0.001M, 0.01M, and 0.1M under the same CVD conditions: the growth temperature of 1000°C, Ar/C2H5OH:H2 of 20:30 sccm, and CVD time of 60 min. Figure 2.20 shows the SEM of HA-CNTs grown using FeCl3 with different concentrations of solution: 0.001M, 0.01M, 0.1M. The SEM results show that the density of CNTs increases with increasing the concentration of FeCl3 catalyst solution from 0.001 M to 0.01 M. However, when the solution concentration was too high (fig. 2.20c), the CNTs were not straight, overlapping, coils and quickly end the long growing process. By counting the number of CNTs at different distances (1 mm, 5 mm, 10 mm, and 15 mm) from the catalytic, we can plot the density distribution of CNTs by the length of the substrate corresponding to different catalyst solutions. The difference in density and length of the HA-CNTs is due to the difference in size of the catalyst particles when we change the concentration of the catalyst solution. In this study, the FeCl3 catalyst solution concentration of 0.01 M was appropriate. HA-CNTs were formed at this concentration with high density and good orientation. Figure 2.20: SEM images of HA-CNTs grown using FeCl3 with different solution concentrations: a) 0.001M, b) 0.01M, c) 0.1M. 10 2.3.3.2. Influence of CVD temperature The HA-CNTs were grown at four different temperatures from 850°C to 1000°C with growth conditions: synthesis time of 60 min, Ar/C2H5OH:H2 = 20:30 sccm. The SEM images (fig. 2.23) indicates that the density of CNTs increases with increasing temperature from 850°C to 950°C. The explanation is that a higher growth temperatures promote a higher density of nanotube nucleations resulting in a higher density of ulralong nanotubes. However, when the temperature was too high, other carbon products, such as amorphous carbon begin to deposit and cover the catalyst particles, and affecting on the germination and growth processes of CNTs. Under our experiment condition, the optimum temperature for the growth process of HA-CNTs was 950oC. 2.3.3.3. Influence of carbon source flow rate We study the influence of ethanol flow rate on the CNTs density and alignment. The results of SEM images (fig. 2.24) show that the CNTs density increases with increasing ethanol flow rate. The highest CNTs density achieves when the ethanol/Ar flow rate was 40 sccm (~150 tubes/mm). However, at this flow rate of ethanol, the density of CNTs decreases rapidly and the ratio of CNTs extending all the length of substrate was low (~25/150 tubes). For others cases, the ethanol flow rate Figure 2.23: SEM images of HA-CNTs grown at four different temperatures: a) 850oC, b) 900oC, c) 950oC, d) 1000oC 11 was 30 sccm for the highest density of CNTs (~80 tubes/mm), CNTs have good orientation, high purity and the ratio of CNTs extending all the length of substrate is about 30/80. As a result, the optimum flow rate of ethanol/Ar to grow the HA-CNTs was 30 sccm. 2.3.4. Growth mechanism and structure of HA-CNTs materials To demonstrate the growth mechanism and aligned along the gas streamlines of HA-CNTs in fast- heating CVD method, we proceeded to grow HA-CNTs on a SiO2/Si substrate with a slit of 60 μm width and directly growth HA- CNTs on field effect transistor (FET) electrodes consisting of 19 pairs of S-D electrodes (source-drain) with an electrode spacing of 30 μm and the total thickness of the metal layers of the electrode is 188 μm (Cr/Pt = 8/180 μm). The catalyst used to grow HA-CNTs in this case is 0.01M FeCl3 solution with the CVD conditions were optimized: temperature Figure 2.25: a,b) Optical microscopy and SEM image of a SiO2/Si substrate with a slit c) SEM image of HA-CNTs SiO2/Si substrate with a gap of 60 μm Figure 2.24: SEM images of HA-CNTs grown using 0,01 M FeCl3 with different flow rate of ethanol/Ar: a) 10 sccm, b) 20 sccm, c) 30 sccm, d) 40 sccm. 12 950oC, 60 minutes CVD time and gas Ar/ethanol: H2O flow rate = 30/30 sccm. The results of SEM images (fig. 2.25 and fig. 2.26) shows that the HA-CNTs grown accross the slit and accross the rough surface of FET electrode. The structure of the HA- CNTs materials was determined by image analysis of HRTEM, Raman spectrum. The results of the HRTEM analysis (fig. 2.28c) and the Raman spectrum (fig. 2.29) show that CNTs have a diameter of 1.5 nm and 70% of HA- CNTs are double wall CNTs (DWCNTs), 30% the remaining are single wall (SWCNTs) and about 50% of them are semiconductor. Figure 2.26: SEM image describes structure of FET electrode and HA-CNTs grown on the FET Figure 2.28: a)Illustration of the experimental setup of HA-CNTs growing on the TEM grid, b) SEM images and c)HRTEM image of a HA- CNT on the TEM grid. Figure 2.29: Raman spectrum of HA-CNTs. 13 CHAPTER 3: FABRICATION OF GRAPHENE MATERIAL BY THERMAL CHEMICAL VAPOR DEPOSITION METHOD 3.1. Thermal CVD system used for fabrication of graphene The thermal CVD system used to fabricate graphene films is also the thermal CVD system used to fabricate oriented CNTs materials but has improved the vacuum system. 3.2. Prepare catalyst material Catalyst material for synthesis of graphene films is poly-crystalline Cu foils (25 m-thick) of 99.8% purity (Alfa-Aesar). Cu foils were cut into small pieces of 2-5 cm2 and cleaned before growing graphene. 3.3. Fabrication process of graphene material on Cu substrate The procedure and steps for fabrication of graphene by thermal CVD method are divided into four stages as described in fig. 3.2. 3.4. Fabrication results of graphene films on the Cu substrate 3.4.1. Influence of surface morphology of Cu substrate To study the influence of surface morphology of Cu substrate to the quality of graphene films, we compared the quality of graphene films grown on the Cu substrate treated by two different methods: 5% HNO3 acid for 10 minutes and treatment by the electrochemical polishing method using 85% H3PO4 acid at 1.9 V for 15 minutes. Results of the measurement and calculation are obtained from Raman spectra (fig. 3.8) Figure 3.2: Fabrication process graphene material by thermal CVD method atmospheric pressure condition 14 of graphene show that the graphene films were fabricated on Cu substrate which was treated the surface by the electrochemical polishing method having the highest quality and the lowest number of layers (about 2 layers) in the three graphene materials investigated. The quality of graphene is shown via values: I2D/IG = 1.28 is the highest, ID/IG = 0.18 is the lowest, the full width at half maximum (FWHM) = 38.05 cm-1 is the lowest, and the position of 2D = 2731.68 is the lowest. Therefore, electrochemical polishing method was chosen to treat the surface of Cu before synthesizing graphene films in all subsequent analyzes. 3.4.2. Influence of the CVD temperature To study the influence of the growth temperature on the graphene quality, we used five electropolished Cu samples and grown at different temperatures from 850oC to 1030oC with the same CVD condition: CH4 as a carbon source, a synthesis time of 30 minutes, Ar/H2/CH4 flow rates of 1000/300/20 sccm. Raman spectra in Fig. 3.10 show that no graphene growth occurs at 8000C. With a higher temperature 8000C, the appears the graphene character peaks inclusion D peak, G peak, 2D peak at the positions 1370 cm-1, 1590 cm-1 and 2734 cm-1 respectively. Thus, fabricate graphene on the Cu substrate surface with the CH4 catalyst source gas which requests the temperature CVD satisfy the condition Figure 3.8: Raman spectra of graphene films on Cu substrate in three cases: a) before treatment, b) after treatment by HNO3 5% and c) after treatment by the electrochemical polishing method. 15 Hình 3.10: Raman spectrum of graphene films on Cu substrate grown at different temperatures from 850oC to 1030oC with the same CVD condition higher 8500C. The quality factor (number of layers, uniformity, defect, impurity) of the graphene film was identified via 2D peak position, FWHM, I2D/ID and ID/IG. The results show that the CVD temperature 10000C is the suitable for fabricating the graphene fims on the Cu substrate with the CH4 source gas. 3.4.3. Influence of hydrocarbon source flow rate To study the influence of the hydrocarbon source flow rate on the graphene quality, we grown graphene at different flow rates of CH4: 0.5 sccm; 2 sccm; 5 sccm; 10 sccm; 20 sccm and 30 sccm, with the same CVD condition: the growth temperature of 1000°C, a synthesis time of 30 minutes, Ar/H2 flow rates of 1000/300 sccm. Fig. 3.13, and fig. 3.15 present Raman spectra and HRTEM images of graphene films on the Cu Figure 3.13: a) Raman spectra and b-e) Lorentzian lineshape of 2D band of graphene films on the Cu substrate with at different flow rate of CH4 16 substrate with at different flow rate of CH4, respectively. The analyst results shown that the number of layers and quality of graphene films were grately affected by CH4 flow rates. High quality single layer graphene films can be manufactured under atmospheric pressure conditions if the CH4 flow rate is low enough. The number of graphene layers incresed and quality of graphene decresed when the CH4 flow rate was too high. According to our experimental results, the CH4 flow rate of 5 to 10 sccm was the optimum to obtain 1-2 layers of graphene films with high uniform and good quality. 3.4.4. Influence of pressure Pressure has been revealed as key factor during graphene growth. To study the influence of the pressure on the graphene quality, we compared the quality of graphene films grown on the Cu substrate grown by two different conditions: The first grown in atmospheric pressure (APCVD) at 1000°C, a CVD time of 30 minutes, Ar/H2/CH4 flow rates of 1000/300/10 sccm and the second grown in low pressure (LPCVD) at 1000oC, pressure of the reactor of 60 torr, a synthesis time of 30 minutes, H2/CH4 flow rates of 20/0.3 sccm. We also study influence of the reactor vacuum levels on the graphene quality by changing pressure of the reactor from 80 torr to 20 torr. Results of the measurement and calculation are obtained from Raman mapping spectra (fig. 3.16 and fig. 3.17) show that the graphene films synthesized by LPCVD method were higher quality and more uniform than those synthesized by APCVD. The Figure 3.15: HRTEM images of graphene films on the Cu substrate with flow rate of CH4: a) 10 sccm, b) and c) 30 sccm 17 quality and uniformity of the graphene films increases with decreasing the pressure in the reaction chamber. This because of the sublimation of Cu at lower pressure, decrease the number of the sharp structures, thereby making the Cu surface smoother. That is the cause lead to increase the quality and uniformity of graphene films. The lower pressure of the reactor, the lower the density of impurities and residual oxygen. That is also a reason why the graphene films quality increases. Using LPCVD method with pressure of 60 torr, CVD temperature of 1000oC, synthesis time of 30 minutes, H2/CH4 flow rates of 20/0.3 sccm, graphene films is formed with a maximum area about 10 cm2 with a high uniformity and less structural defects. About 70% of the graphene film area is monolayer, 30% of the remaining area is bilayers. Figure 3.16: Raman spectra of graphene films on the Cu substrate were synthesized by APCVD and LPCVD methods Figure 3.17: Raman spectra of graphene films on the Cu substrate were synthesized with differrent pressure of the reaction chamber. 18 CHAPTER 4 ENZYME-GrISFET SENSOR FOR TRACE-DETECTION OF HERBICIDE ATRAZINE 4.1. Basis of the graphene material selection for fabricating the enzyme-GrISFET sensor In this section, we present the basis of the graphene material selection for fabricating the enzyme-GrISFET sensor, including: the synthesis technology, the material properties, the mobility of the carrier charges of the graphene conductive channel and the effective surface area of the material. 4.2. Fabrication of the enzyme-GrISFET sensor The procedure for fabrication of the enzyme-GrISFET sensor as described in fig. 4.3. Fig. 4.9 is an optical microscopy of the enzyme- GrISFET sensor. 4.3. Application of enzyme-GrISFET sensor for detection of pesticide residue atrazine The atrazine detection in solution was performed via the competitive inhibition mechanism of itself with respect to the catalytic activity of the enzyme urease. Under the inhibition of atrazine, the Figure 4.9: Optical microscopy of the enzyme-GrISFET sensor Hình 4.3: Fabrication process of the enzyme-GrISFET sensor Figure 4.10: Atrazine detection mechanism of the enzyme-GrISFET sensor. 19 catalytic ability for the urea substance hydrolysis reaction of the urease enzyme was decreased which lead to the NH4+ ion concentration and the OH- ion concentration generated by the hydrolysis reaction decreases, those are the causes which decrease the p (n) doping effect into the conductivity channel. This leads to the change in the transmission characteristic of the sensor, namely the potential position changes at Dirac point (V0) in the horizontal direction as well as changing the current-out signal intensity ΔIds (fig. 4.10). 4.4. Results and discussion 4.4.1. The structural morphology of the enzyme-GrISFET sensor The surface morphology of the enzyme-GrISFET sensor was considered by the (FESEM) after synthesis (fig. 4.14). 4.4.2. Determining the saturated concentration of urea substance for the enzyme-GrISFET sensor Saturated concentration of urea substrate is the concentration which the urease enzyme membrane reacted 100% of its activity, and then the output current intensity Ids of the sensor does not change despite the concentration of urea substrate increase. Fig. 4.16 describes the Ids-Vg characteristics of enzyme- Figure 4.16: The characteristic Ids-Vg of enzyme-GrISFET sensor for Vds = 1 V, Vg in the 0 V - 3 V range, step 0.05 V with the urea substance concentration range of 5 - 35 mM Figure 4.14: The optical and SEM images of the electrode after graphene were transferred 20 GrISFET sensor for Vds = 1 V, Vg in the 0 V - 3 V ranger with the step 0.05 V and the urea concentration range of 5 mM to 35 mM. Observing Fig. 4.16a image, we can recognize that when increases the urea substance concentration from 5 mM to 30 mM, the current-out signal intensity ΔIds doubly increases from 0.15 mA to 0.30 mA and the position of Dirac point (Vo) was shifted towards higher negative values. However, continuously increasing the substance concentration to 35 mM, the value ΔIds hardly increases as well as there is no the movement of the V0 point. 4.4.3. The response characteristic of the enzyme-GrISFET sensor In this part, the author discusses in detail the installation steps, setting parameters to determine the transmission characteristic Ids - Vds, Ids - Vg of the enzyme-GrISFET sensor and how to determine, calculate the sensor's parameters such as leakage current, transconductance, capacitance, electron and hole mobility in the graphene channel. 4.4.4. The effect of the fabrication process on the output signal In this part, the author presents in detail the results of investigating the influence of several factors on the operation of the sensor such as immobilize enzyme temperature and enzyme immobilize time. The results show that the enzyme immobilize temperature from 30oC to 40oC and the enzyme immobilize time from 40 to 60 minutes are appropriate and the performance of the sensor is the best. From the study results, the authors selected a enzyme immobilize temperature of 30°C and a enzyme immobilize time of 60 minutes to immobilize the urease enzyme on the interdigitated electrodes using glutaraldehyde (GA) vapor as cross-linking agent in fabrication the enzyme-GrISFET sensor for the detection of atrazine herbicide residues will be discussed in detail in the next part of this thesis. 4.4.5. Enzyme-GrISFET sensor application for detection of pesticide residue atrazine 4.4.5.1. Sensor characteristics which are inhibited by atrazine To investigate the sensor characteristics by inhibition of atrazine, we 21 processed to measure their electrical characterization processed in the solution containing atrazine with concentration of 2 10-2 ppb. Observing the fig.4.22 we found that the variations of the drain- source current ΔIds of the sensor significantly decreases from 304 μA to 136 μA and the Dirac point of graphene Vo shifts to higher positive values from 0.75 V to 1.25 V when it is inhibited by atrazine with the concentration 2  10-2 ppb. This result was also observed in other report. The working principle of our atrazine sensors is based on the inhibition of atrazine toward urease in the hydrolysis reaction of urea substance. The hydrolysis of urea substance provided ions (NH4+, OH−) that would be adsorped onto electrode surfaces, then led to the increase in charge density and mobility on graphene layer. When atrazine is introduced, it acts as an inhibitor to reduce activity of enzyme, lowers the concentrations of ions at the electrodes and subsequently decreases current signal as also as shift of Vo. 4.4.5.2 The repeatability of the enzyme-GrISFET sensor. Fig. 4.23: The six times of measurement results of the characteristics Ids-Vg, with Vg from 0 V to 3 V, Vds = 1V at the atrazine concentration Catz = 2  10-4 ppb. Fig. 4.22. The characteristic transmission Ids-Vg of the GrISFET sensor with Vg from 0 V to 3 V step 0,5 V, Vds = 1V, in two cases (before, after) that is inhibited by the atrazine 2  10-2 ppb. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 6.4 6.5 6.6 6.7 6.8  I d si = 1 36 ( A )  I d so = 3 04 ( A ) ATZ = 0 ATZ = 2 x 10 -2 ppb I d s ( m A ) Vg (V) Vo = 1.25 V ATZ = 2 x 10 -2 ppb Vo = 0.75 V ATZ = 0 22 To consider the repeatability of the sensor, we continuously repeated the six times measurement and comparing the results after the measurements. From these measurements we can calculate the standard deviation of sensor Sy = 9.2 at the atrazine concentration of 2  10-4 ppb. This is an important value in calculating the limit of detector (LOD) of the sensor. 4.4.5.3 Detection litmit of the GrISFET biosensor To determine the LOD of the enzyme- GrISFET sensor for the detection of the herbicide atrazine in solution, we use saturated urea solution (30 mM) and various carbaryl solutions with concentrations ranging from 2  10-4 ppb to 20 ppb were prepared in deionized water. For each measurement, the sensors were incubated for 30 mins at room temperature in the herbicide solution and then their electrical characterization Ids - Vg was immediately processed in urea solution. The electrical characteristics Ids - Vg of the sensor is measured with the voltage of the Vg is sweeped from 0 V to 3 V, step of 0,05 V, the voltage applied to the drain – the source Vds is 1V. The current responses of atrazine enzyme-GrISFET sensor at different sample concentrations are shown in Fig. 4.25. We found that the atrazine level can be monitored by either position or intensity of Dirac point. When the concentration of atrazine increased from 2  10-4 ppb to 20 ppb, the position of Dirac point was shifted towards higher positive values from 0.1 1 6.4 6.5 6.6 6.7 ATZ = 0 ATZ = 2 x 10 -4 ppb ATZ = 2 x 10 -3 ppb ATZ = 2 x 10