Designing Polymer Coated Carbon Nanotube Detectors for Alkane Vapors

Alkane vapors are difficult to be detected in oil pipelines until a leak exceeds 1% of the total flow. Yet, an alkane leak can cause severe security, environmental, and human health problems if it is not identified or addressed. Detecting low concentrations of alkanes will lead to the detection of small leaks in oil pipelines. However, this requires a sensor, which can detect a substance that is inert at room temperature. Polymer coated carbon nanotubes provide that sensor. Polymers, when coating a network of carbon nanotubes, can serve as a tunnel barrier between adjacent carbon nanotubes. In a sensor, adsorption causes this barrier to swell increasing the resistance of the network. This enables the detection of charge transferred inactive species such as alkanes. In this research, design rules were applied to the polymer to make it a suitable host for alkanes by varying the functional end groups of the polymer to enable detection of specific alkanes. The sensors produced were able to detect various vapors at low concentrations with stronger responses to certain vapors and exhibiting almost no response to water.

ABSTRACT Alkane vapors are difficult to be detected in oil pipelines until a leak exceeds 1% of the total flow. Yet, an alkane leak can cause severe security, environmental, and human health problems if it is not identified or addressed. Detecting low concentrations of alkanes will lead to the detection of small leaks in oil pipelines. However, this requires a sensor, which can detect a substance that is inert at room temperature. Polymer coated carbon nanotubes provide that sensor. Polymers, when coating a network of carbon nanotubes, can serve as a tunnel barrier between adjacent carbon nanotubes. In a sensor, adsorption causes this barrier to swell increasing the resistance of the network. This enables the detection of charge transferred inactive species such as alkanes. In this research, design rules were applied to the polymer to make it a suitable host for alkanes by varying the functional end groups of the polymer to enable detection of specific alkanes. The sensors produced were able to detect various vapors at low concentrations with stronger responses to certain vapors and exhibiting almost no response to water. TABLE OF CONTENTS ABSTRACT .............................................................................................................................................................. ii ACKNOWLEDGEMENTS .................................................................................................................................... 1 1. INTRODUCTION ................................................................................................................................................ 2 2. LITERATURE REVIEW ................................................................................................................................... 4 2.1 CURRENT RESEARCH ............................................................................................................................. 4 2.2 CARBON NANOTUBE GAS SENSOR RESEARCH ..................................................................... 4 2.3 OLIGOMER-COATED CARBON NANOTUBE SENSORS RESEARCH ............................. 5 2.4 NANOFIBRIL ALKANE VAPOR DETECTORS RESEARCH .................................................. 7 2.5 CHALLENGES AND DIRECTIONS ..................................................................................................... 8 2.6 RESULTS OF ALKANE VAPOR SENSORS RESEARCH .......................................................... 9 3. PRODUCT DESIGN AND SPECIFICATIONS ..................................................................................... 11 4. EXPERIMENTAL PROCEDURES ............................................................................................................ 13 4.1 LISTS OF MATERIALS .......................................................................................................................... 13 4.2 POLYMER/ CNT SOLUTION SYNTHESIS ................................................................................... 13 4.3 UV-VIS TESTING PROCEDURES ..................................................................................................... 15 4.4 SENSOR FABRICATION PROCEDURES ...................................................................................... 15 4. 5 VAPOR TESTING PROCEDURCES ................................................................................................ 16 5. RESULTS AND DISCUSSION.................................................................................................................... 18 5.1 POLYMER SYNTHESIS RESULTS/ DISCUSSION ................................................................... 18 5.2 UV-VIS TEST RESULTS/ DISCUSSION ......................................................................................... 18 5.3 SENSOR FABRICATION RESULTS/ DISCUSSION ................................................................. 20 6. FUTURE WORK ............................................................................................................................................... 26 7. BUSINESS, SOCIAL AND ETHICAL CONSIDERATIONS .......................................................... 26 8. SUMMARY AND CONCLUSIONS .......................................................................................................... 28 9. REFERENCES .................................................................................................................................................... 30 iii 1 ACKNOWLEDGEMENTS I would like to thank Dr. Ling Zang for his support on this project. I would also like to thank Ben Bunes and Yaqiong Zhang for their guidance on this project. I appreciate the amount of time Yaqiong Zhang invested in training us in different experimental testing procedures. I also appreciate her insight on every step of this project. Lastly, I would like to thank the Undergraduate Research Opportunities Program (UROP) and Vaporsens for providing funds for this project. 2 1. INTRODUCTION Alkanes are hydrocarbon chains that contain between 2 to 16 carbons chains. These are commonly used as fuels because of their ability to combust readily in the presence of oxygen, which is naturally available everywhere. This combustion releases energy that is used to power devices and appliances. Methane, a single carbon chain, is the main component in natural gas that is used in stove gas, while eight or longer carbon links are present in petroleum. The energy used from the combustion of alkanes has improved the lives of many, but if left unchecked alkanes can cause great harm. Because of potential harm, alkanes pose a risk to security, the environment, and human health. Due to the susceptibility of alkanes to combust under oxygen, explosives using alkanes can be easily constructed by anyone. Alkanes have also been proven toxic to the human nervous system [1]. However, detection of alkanes is difficult due to their chemical properties. Alkanes are odorless, colorless, and inert at room temperature. In order to mitigate these dangers, a technology for detecting alkanes at low concentrations is required. Previous methods for detecting alkanes used chemoresistive materials that catalyze the oxidative dehydrogenation of alkanes [2]. While demonstrating a good response to low concentrations of alkanes, methods using these materials are limited to light alkanes (methane to butane) and require high working temperatures (300-600 °C). Another method uses thin films of Au-dodecanethiol core/shell nanoparticles deposited via Langmuir-Schafer method [3]. This method proved to be effective at detecting decane 3 due to its similar structure with the gold nanoparticles, but it was not as effective at detecting alkanes whose structure differed from decane. Alternatively, a new approach based on carbon nanotubes (CNTs) was introduced in 2003 [4]. CNTs in a network are conductive via quantum tunneling of electrons from one CNT to another. Adsorption of gas occurs in the interstitial space between CNTs, which increases the tunneling barrier of the electrons due to the increase in tunneling distance. The change in conductivity can be precisely measured from a change in current. Inspired by this concept, a sensor for detecting alkane vapors was designed by using CNTs in a network placed on an interdigitated electrode. The concept of this design can be seen in Figure 1. The CNTs are coated with a polymer that will cause adsorption of similar structured alkanes to the CNTs occurring at room temperature. An array of sensors is used to distinguish between other alkane vapors, resulting in high selectivity. Figure 1. Schematic of CNTs coated with a polymer allow holes to transfer at junctions. The alkane vapor adheres to the polymer causing reduced mobility of holes at junctions [8]. 4 2. LITERATURE REVIEW 2.1 CURRENT RESEARCH Carbon nanotubes have been an area of great interest because of their unique properties, electronic structure, and susceptibility to chemical reactions. There are two types of carbon nanotubes: single-walled carbon nanotubes (SWNT) and multiple-walled carbon nanotubes (MWNT). Research has been done with oligomer-coated carbon nanotube sensors and nanofibril alkane vapor detectors. Research has proven that CNTs react easily to gases such as NH3, NO2, H2, CH4, CO, SO2, H2S, and O2 [5]. Therefore, many researchers are trying to exploit these sensitivities in order to develop new sensor technologies, such as CNT-based sensors for gas- and vapor-phase analytes [5]. 2.2 CARBON NANOTUBE GAS SENSOR RESEARCH In 2000, Douglas R. Kauffman and Alexander Star provided evidence on the potential of CNT-based gas sensors from the response of nanotube field-effect transistor (NTFET) devices to NO2 and NH3 gases. NTFET devices demonstrated unique responses to NH3 and NO2 through chemical gating of the SWNT. Both gases showed a shift in the transfer characteristic gate voltage of approximately -4V and +4V. Kauffman and Star were able to slightly detect NH3 and NO2 gases with CNT-polymer composites. SWNT NTFET devices, functionalized with the polymer poly (m-aminobenzene sulfonic acid) (PABS), exhibited significant n-type sensitivity (5 ppb) to NH3. PABS became deprotonated when exposed to NH3, which resulted in a hole depletion from the SWNT and a reduction in the overall conductance of the SWNT-PABS system. SWNT-PABS devices were used to reach detection limits of 100 ppb NH3 and 20 ppb NO2 with short 5 response times of a few minutes and total recovery. More research with SWNT-PABS was done to prove that the SWNT- poly(ethylene imine) (PEI) and SWNT- polymer polyaniline (PANI) systems accounted for the faster and more complete recovery of the most sensitive CNT-based sensor for NO2 and NH3. In addition, when CNT-based sensors were tested in industrial, environmental, personal safety, medical, and military scenarios, they displayed a sensitivity to the following gases: H2, CH4, CO, SO2, H2S and, O2. CNT-based gas sensors have proven to detect many gases; therefore, these can be manipulated to detect alkane gases as well. 2.3 OLIGOMER-COATED CARBON NANOTUBE SENSORS RESEARCH Zhang et al. experimented with using CNT sensors coated with a carbazolylethynylene oligomer to detect trace amounts of explosive materials such as 4Nitrotoluene (NT), 2,4,6-trinitrotoluene (TNT), and 2,4-dinitrotoluene (DNT). The sensing abilities were enhanced due to the high porosity in the sensor, which increased the surface area available for the analyte to interact [1]. The CNTs were coated with a noncovalent material to maintain their electron transport properties and to improve the selectivity of the explosive compounds. Generally, the CNT network conducts electrons via quantum tunneling at the junctions [6]. The insulating Tg-Car oligomer creates a barrier for this tunneling, and swelling of the oligomer due to the presence of analyte results in a decrease in conduction caused by the increased tunneling distance at the junctions. A current can be passed through the CNT network and monitor any slight changes in relative conductivity. 6 CNT sensors can be made to be highly selective. Compared to other common chemical reagents, the detection of the targeted explosive compound NT was relatively high. When uncoated CNTs were compared with Tg-Car/CNT, the Tg-Car/CNT showed a decrease in conductivity, while the uncoated CNT exhibited an increase in conductivity under the same conditions. This is due to uncoated CNT being a p-type material. Holes are the main charge carriers of p-type materials. Therefore, conductivity increases when electron-withdrawing analytes are present. NT, DNT, and TNT have similar chemical structures and properties, so being able to differentiate which compound is being detected is quite useful. Sensor arrays are a possible solution to this. Since uncoated CNT and Tg-Car/CNT have different results when detecting NT, DNT, and TNT, an array was built that incorporated the data from both of these sensors. The conductance changes of the Tg-Car sensor and the uncoated CNT sensor in response to NT, DNT, and TNT are shown in Figure 2. 7 Figure 2. The conductance change of the Tg-Car/CNT sensor and the uncoated CNT sensor in response to NT, DNT, and TNT [1]. Clusters of data points for NT, DNT, and TNT were in clearly defined regions separated from each other [1]. This demonstrates that CNT sensors in an array could be made to differentiate between similar compounds. 2.4 NANOFIBRIL ALKANE VAPOR DETECTORS RESEARCH Wang et al. developed a nanofibril composite that detects alkanes using a photoinduced charge transfer (PCT) based approach. PCT is a process in which photons are used to excite electrons, causing an electron transfer from the donor to the acceptor site [7]. A long alkyl-substituted arylene-ethynylene tetracycle (ACTC) was selected as 8 the donor and perylene-3,4,9,10-tetracarboxylic-3,4,9,10-diimide (PTCDI) as the acceptor [8]. The ACTC and PTCDI were layered on top of each other. These formed a continuous nanofibril network that is porous to allow interaction with alkane gases. The alkanes can be absorbed between the ACTC and PTCDI, causing the distance between donor and acceptor to increase and resulting in a decrease in PCT efficiency. Findings showed that in the presence of several alkane gases, when the photocurrent of the film was measured over time, each alkane (n-hexane, n-octane, n-decane, and n-dodecane) produced dramatically different response profiles. Both the amount of photocurrent that decreased and the recovery times varied. Principle component analysis (PCA), which was used to quantitatively analyze the response profiles of the four alkanes, showed that clustering of the same alkane was apparent and distinct from other clusters. This means that it is possible to distinguish between different alkanes. However, ACTC degrades too quickly to be stable enough for real world applications. 2.5 CHALLENGES AND DIRECTIONS It is important for gas sensors to be able to detect trace amounts of alkane vapors, as these pose a great risk to society in terms of security, the environment, and health. However, due to alkanes’ inherent properties, such as colorlessness, odorlessness, and inertness at room temperature, they are difficult to detect. There are also many kinds of alkanes, depending on the number of carbon atoms in the molecule, which results in difficulties in distinguishing one alkane from another. CNTs can be a possible solution due to how easily they react with other gases as well as their unique mechanical and 9 electrical properties. Because sensors made with CNTs can be made to be highly selective, CNTs are an ideal material to use when designing an alkane vapor detector [9]. Several advances have been made using CNTs to detect vapors, but little progress has occurred in the detection of alkanes, largely due to their inertness at room temperature. This project focused on the design of a polymer coated carbon nanotube detector for alkane vapors. The goal is to detect alkane vapors in concentration as low as 1% of the saturated vapor. 2.6 RESULTS OF ALKANE VAPOR SENSORS RESEARCH After researching prior work on CNTs based detectors, it was decided that four specific chain groups of regioregular poly(3-hexylthiophene-2,5-diyl)(P3HT) would be used to develop the detectors. P3HT is a regioregular semiconducting polymer and is very useful because of its regular end-to-end arrangements of side chains. Table 1 shows the chemical structures and formulas of the polymers selected. Within the red highlighted circles are the different functional groups that make each polymer chemically different and selective towards a specific alkane vapor. 11 3. PRODUCT DESIGN AND SPECIFICATIONS This project designed an array of sensors that can be used for detection of alkane vapors. The design of this vapor testing system can be seen in Figure 3. In the technical design, four sensors were made. Each was designed to detect a specific alkane vapor more efficiently than other vapors. In each sensor, a network of carbon nanotubes was coated with a specific polymer corresponding to the alkane vapor to be detected and each network was placed on an interdigitated electrode (IDE), which was coated with the same polymer/CNT solution. Efficient detection of a specific alkane vapor (ability to select for a certain alkane over others) depended on the similarity between the chemical structure of each alkane vapor and the functional group of the polymer that was being used to detect, causing the vapor to adhere to the polymer. This resulted in a decrease in current measured because of the increase in the tunneling barrier of the carbon nanotube network. Six electrodes were attached to a circuit with an ammeter to measure the change in current. The circuit contained a chip carrier, which contained the four sensors. Small changes to the tunneling barrier had a great effect on the overall conductivity of the network. This means measurement of conductivity can be made to be very accurate. Therefore, this method is ideal for detecting small concentrations of alkane vapors. The electrodes enabled the sensors to detect hexane, dodecane, acetone, ethyl acetate, hydrazine, methanol, and ethanol. The measureable specification that was required when designing the sensor was the ability to detect alkane vapors at 1% of the saturated vapor. Currently, other devices 12 that can detect alkane vapors use alkanethiol encapsulated gold nanoparticles as chemiresistors and can only detect alkane vapors at 10,000 ppm [3]. Thus, these are not capable of detecting concentrations as low as 1% of the saturated vapor. The sensor also needed to exhibit selectivity and to differentiate between the alkane vapors present. By using P3HT coated CNT devices, it should be able to determine which alkane vapors are present in the environment: hexane, dodecane, acetone, ethyl acetate, hydrazine, methanol, and ethanol. The sensor should detect alkane vapors only and not result in a false positive if exposed to other vapors such as water. Additional design specifications include the following. The devices need to show results within seconds of being exposed to the vapor. They must operate at room temperature and still be effective at detecting alkane vapors within the range of 15°C30°C. They should also be able to detect alkane vapors for one hundred or more cycles without being replaced. Figure 3. The sensor testing diagram [1]. 13 4. EXPERIMENTAL PROCEDURES Polymer coated carbon nanotube detectors were synthesized, fabricated, and then tested to determine the sensors’ response when exposed to various analytes. The materials used and the experimental procedures for this project are described below. 4.1 LISTS OF MATERIALS The following materials were purchased for this experiment: four chain groups of regioregular poly(3- hexylthiophene-2,5-diyl), 4020, 4027, 4030, and 4033; carbon nanotubes (single walled); chloroform (CHCI3) and dimethyl sulfoxide (DMSO). 4.2 POLYMER/ CNT SOLUTION SYNTHESIS Each regioregular poly (3-hexylthiophene-2,5-diyl)( P3HT) was synthesized in the following steps: 1. Dissolve Carbon Nanotubes with Chloroform or DMSO 5.0 mg of carbon nanotubes were added to 5.0 mL of chloroform in a test tube and was ultrasonicated for 120 minutes in a water bath at room temperature. This solution was used to dilute and dissolve the two side chains groups of P3HT, 4020, and 4027. Also, 5.0 mg of carbon nanotubes were added to 5.0 mL of DMSO Figure 4. CNT dissolved in DMSO in a test tube and was ultrasonicated for 120 minutes in a water bath at room temperature. This solution was used to dilute and dissolve the two other side chains groups of P3HT, 4030, and 4033. 15 4.3 UV-VIS TESTING PROCEDURES UV-Vis tests were performed on both the diluted pure polymer solutions and the diluted CNT/ polymer mixed solutions. The following procedures were performed for each polymer solution (pure and mixed) separately: 1. Wash two cuvettes with a cuvette cleaning solution and then with acetone. 2. Ultrasonicate the diluted solutions at room temperature for 30 minutes. 3. Pipette 3.0 mL of chloroform or DMSO into each cuvette. Note: This is the baseline test for calibration using chloroform or DMSO. 4. Pour out the chloroform or the DMSO from one of the cuvettes. 5. Pipette 2.0 mL of the diluted 4020, 4027, 4030 or 4033 coated CNT solution into the empty cuvette. Note: The diluted solutions must be ultrasonicated before use. 6. Place the cuvette holding the diluted solution in the UV-Vis machine and run a test. 7. Repeat steps 4-5 for the other diluted polymer solutions. 4.4 SENSOR FABRICATION PROCEDURES 1. Clean 5-10 chips with 3 different cleaning reagents (acetone, methanol, and isopropanol). This means that each chip is cleaned 3 times, once with each reagent. 2. Dry the chips and place them on a hot plate for 5 minutes at 80°C. 3. Grab a glass slide and place a thin piece of double sided tape on the glass slide. 4. Place all the cleaned chips on the glass slide and then label the chips. 5. Drop-cast 1.0-4.0 µL (1.0 µL at a time) of the diluted CNT/P3HT solution on each 16 chip. Note: the diluted CNT/P3HT solutions should be ultrasonicated for 30-60 minutes before use. 6. Place the chips on a hot plate for 5-10 minutes at 80°C. Note: This is done to remove chloroform or DMSO residue on the device. 7. Check the resistance of each chip with an ohmmeter. The desired resistance of each chip is 10.0 KΩ -1000.0 KΩ. Note: If the resistance of a chip is not in the desired range, repeat step 5 until the desired resistance is achieved. 4.5 VAPOR TESTING PROCEDURCES 1. Place a double-sided adhesive on a ceramic chip carrier. Then place the sensor device on the carrier and wire-bond the sensors into the chip carrier. Note: Wirebonding was done by Yaqiong Zhang, senior project PhD student. Figure 6. A picture of the ceramic chip carrier with the chips before wire-bonding. 2. Place the chip carrier in a breadboard and cover the chip carrier with a Teflon 17 enclosure, then connect the sensors. 3. Dilute the analyte to 1%. Obtain a glass syringe and infuse 40 mL of the saturated headspace of the analyte diluted to 1%. 4. Program the syringe pump (NE-1000 Pump system) to infuse into 1% of each analyte for 20 seconds, pause for 40 seconds, repeat for 3 cycles, and then pause for 60 seconds. This program was repeated for each analyte (hexane, dodecane, acetone, ethyl acetate, hydrazine, methanol, ethanol). Table 2 summarizes the program steps for hexane. The procedure was followed for the other 6 analytes. Concentration 1% of Hexane diluted to 1% 2% of Hexane diluted to 1% 4% of Hexane diluted to 1% 8% of Hexane diluted to 1% Step 1 Infuse 1.01 mL/min for 20 seconds Infuse 1.01 mL/min for 20 seconds Infuse 1.01 mL/min for 20 seconds Infuse 1.01 mL/min for 20 seconds Step 2 Pause for 40 seconds Step 3 Repeat step 1 and 2 two times Step 4 Pause for 60 seconds Pause for 40 seconds Repeat step 1 and 2 two times Pause for 60 seconds Pause for 40 seconds Repeat step 1 and 2 two times Pause for 60 seconds Pause for 40 seconds Repeat step 1 and 2 two times Pause for 60 seconds Table 2. Brief summary of the programmed steps for Hexane. The same procedure was followed for the other 6 analytes. 18 5. RESULTS AND DISCUSSION 5.1 POLYMER SYNTHESIS RESULTS/ DISCUSSION Initially, polymers 4020, 4027, 4030, and 4033 were diluted only in chloroform. Polymer 4020 and 4027 were successfully dissolved in the solution. However, 4030 and 4033 were not able to dissolve in chloroform even after being heated and ultrasonicated for 3 hours. After further literature review, dimethyl sulfoxide (DMSO) was identified as an ideal solvent to dissolve polymer 4030 and 4033 [10]. Polymers 4030 and 4033 were successfully dissolved in DMSO within 1 hour of ultrasonicating. 5.2 UV-VIS TEST RESULTS/ DISCUSSION UV- Vis tests were performed for both pure and mixed (polymer/ CNT) polymer solutions of 4020 and 4027, diluted in chloroform. This was done to determine the presence of polymer coated CNT. The results of these UV-Vis tests can be seen Figure 7 and Figure 8. Pure polymer solutions 4020 and 4027 dissolved in chloroform have shown to have peak absorption around 450 nm. The mixed polymers solution 4020 and 4027 dissolved in chloroform had a smaller peak of absorption around 450 nm, but exhibited a slight peak around 600 nm. This peak at 600 nm that is not present in the pure solutions indicates the presence of CNTs. The UV-Vis results from these two mixed solutions indicate that the polymers 4020 and 4027 are in fact bonded with the single-walled CNTs. 19 Figure 7. UV-Vis absorption of polymer 4020 and 4027 dissolved in chloroform. The same UV-Vis testing was conducted for pure polymer solutions 4030 and 4033 dissolved in DMSO and mixed polymer solutions 4030 and 4033 dissolved in DMSO. The data collected from UV-Vis tests on these polymers is shown in Figure 8. Pure polymer solutions 4030 and 4033 dissolved in DMSO had peak absorption at 460 nm and 500 nm respectively. In contrast, both mixed polymer solutions have very strong peak absorption around 300 nm with very slight peaks at 475 nm and 525 nm for 4030 and 4033 respectively. The peak at 300 nm that is not present for the pure polymers confirms the presence of CNT in the solvent, indicating that in these solutions the polymers are bonded with the single-walled CNTs. Therefore, the UV-Vis results indicated that all four polymers bonded with the single-walled CNTs. 20 Figure 8. UV-Vis absorption of polymer 4030 and 4033 dissolved in dimethyl sulfoxide (DMSO). 5.3 SENSOR FABRICATION RESULTS/ DISCUSSION Three sets of sensors were fabricated using the diluted polymer/ CNT solutions. The sensors were then tested in order to conduct vapor sensor testing three separate times. The resistances for each chip functionalized with a polymer and the number of drops of the polymer solution used during drop casting is recorded in Table 3. The desired resistance range for each chip was 10.0 kΩ-1000.0 kΩ. 1.0 µL of the solution was drop casted each time and then measured until observed to be within the ideal range. Any chip that needed more than 4 µL of solution was discarded because too much polymer can oversaturate the CNTs and degrade their ability to conduct holes. 21 Chip # 1 2 3 4 Polymer 4020 4027 4030 4033 Drops 2.0 µL 2.0 µL 1.0 µL 2.0 µL Resistance 34.5 kΩ 6 kΩ 66 kΩ 278 kΩ Table 2. Measured resistances for each polymer 5.4 ALKANE VAPOR TEST RESULTS/ DISCUSSION The performance of the sensors was evaluated by testing it on an array of alkane vapors. These included vapors for which a response was desired and vapors for which no response was desired. The strength of the sensor’s response to an alkane vapor or other vapors is indicated by the relative change in conductivity of the sensor before and after being exposed to the vapor. Figure 9 shows the sensory response of set 3 polymers 4020 and 4027 exposed to hexane at 17.22 ppm. The 17.22 ppm level of saturation meets the desired specification for the project. The polymers were exposed to the vapors for 20-second intervals and then not exposed for 40 seconds to recover between intervals. According to Figure 9, both polymers show a distinct response to the hexane, which begins at 60 seconds, 120 seconds, and 180 seconds. 22 Figure 9. Real-Time Sensor response to hexane of 4020 and 4027 at 17.22 ppm. Exposure time was 20 s and recovery time was 40 s. The ability to sense other common reagents such as ethyl acetate, acetone, methanol, ethanol, hydrazine and water was tested. Figure 10 shows the change in conductance of polymer coated CNTs with respect to these common reagents at 1% of saturated vapor. The polymers 4020 and 4027 showed a response to hexane and hydrazine, and a strong response to ethyl acetate, methanol, and ethanol. Acetone exhibited varied responses so detection of acetone is not conclusive. The polymers showed almost no response to water vapor, which is ideal since water vapor is fairly common, and one of the design specifications was for the device to not show a response unless detecting a desired vapor. There is a poor response to dodecane. This is due to polymers 4020 and 4027 not having a chemical structure similar dodecane. A similar structured functional group attached to P3HT would have made dodecane adhere to the CNTs. 4020 and 4027 have a short polymer chains whereas dodecane is made up of long polymer chains. 23 Figure 10. Change in conductivity in response to exposure of common reagents for polymers 4020 and 4027. Figure 11 depicts the change in conductance for set 1 polymers 4020, 4027, 4030, and 4033 in response to a diluted saturated vapor of hydrazine, ethanol, acetone, water, and then hydrazine again after the other reagents were tested. This set of polymers showed a stronger response to hydrazine compared to the other reagents. While not as strong, after the other vapors were tested there was still a noticeable response to hydrazine with the sensors 4030 and 4027. 4033 and 4020 still exhibited a similar response to hydrazine. This shows that the sensors are capable of undergoing multiple tests and the results indicate that the level of degradation does not prevent from using a sensor multiple times. 24 Figure 11. Change in conductivity in response to exposure of common reagents for Set 1. A second set of polymer sensors were synthesized and left in an open-sealed container for a week. Figure 12 depicts the change in conductance for the set 2 polymers 4020, 4027, 4030, and 4033 in response to diluted saturated vapor of hydrazine, ethanol, acetone, water, and then hydrazine after the other reagents were tested. The average response is much weaker than set 1 with the exception to 4030 detecting hydrazine at nearly twice the magnitude as set 1. However, 4030 exhibited a weak response to hydrazine after the other reagents were tested. Set 2 provides evidence that the sensors are not capable of conducting after a week has passed. This may be due to the degradation of the CNT network structure caused by the solvent evaporating. This may also be due to other vapors in the atmosphere adhering to the surface of the CNTs over time, leaving the ability of the alkane vapors to adhere to the surface almost non-existent. 25 Figure 12. Change in conductivity in response to exposure of common reagents for Set 2 after exposure to open air for a 1 week. 26 6. FUTURE WORK While there has been great progress in detecting and sensing alkane vapors, there are many aspects that can be improved in the future. The device was able to detect alkane vapors at low concentrations. However, the exact ideal resistance can be found to detect alkanes more effectively. A resistance that results in an extremely strong response without sensing unwanted vapors would be the most reliable. This can be achieved by testing chips with various resistances and determining the range of resistance that provides the most ideal conditions for sensing. One major problem with these sensor devices is that the chips only lasted for a few days due to the degradation of the polymer coating on the CNTs. This was observed from the increase in resistance of the chips over time. This could be prevented by developing a protective coating for the polymer, which would shield it from the environment and could then be removed as needed without damaging the system. A possible heat treatment to stabilize the polymer coating could also be another solution. The polymers tested do not have a long enough shelf life to be a commercially viable product. However, based on the findings, they did not degrade rapidly enough to not be used for multiple vapors in multiple tests. It may be possible to identify polymers, which can be commercially viable, but this would require multiple trials. The selectivity of this technology could also be improved. The devices made were able to detect various vapors, but future work could be done to develop devices that only respond to a specific chemical vapor. This could be achieved by testing other polymers used to coat the CNTs. Finding more polymers whose end groups have a similar structure to the targeted alkane would be a good place to start. 27 7. BUSINESS, SOCIAL AND ETHICAL CONSIDERATIONS Carbon nanotubes are ideal nanomaterials for technology because of their small size and unique properties. Polymer coated CNT based detectors can be highly beneficial for applications used in daily life because, unlike other sensors, CNT based detectors can detect extremely low concentrations of gases used for heating and cooking, while also detecting vapors from gasoline. These detectors also have military applications in detecting alkane-based explosives for security. This technology appears to be safe and reliable as well as cost-effective, raising no red flags for economic restraints. Ethically, the widespread use of this technology could keep us safe from gas leaks due to wear as well as from intentional damage that could be connected to terrorist attacks. Therefore, it makes sense to pursue development of this technology. The social and environmental impact is minimal because these sensors do not create waste or toxins. The actual chip is 1/2 cm x 1/2 cm, so only a relatively small amount of material is required to produce the chips. 28 8. SUMMARY AND CONCLUSIONS This project set out to design a technology that could be used for detection of alkane vapors as well as other common reagents. The goal was to design a sensor that displayed selectivity towards specific vapors, as well as detect those alkane vapors at low concentrations. This project is significant as alkanes are commonly used, highly combustible in the presence of oxygen, and are difficult to detect due to being colorless, odorless, and inert at room temperature. Common uses of alkanes are cooking and heating gas, gasoline, and as fuel for military projects making safety a top priority. The design specifications for the sensor included cost-effectiveness, ease of manufacture, lack of toxic or environmentally threatening side effects, and a long shelf life, in addition to being able to sense very low levels of alkanes. The sensors are lowcost to produce due to small amount of material required. Also, using these sensors is a simpler procedure than other sensors in the past as only a small current is required and a computer could monitor the change in response. The device itself is small, so it is easy to store and to transport. There appears to be no environmentally harmful or toxic side effects. The shelf life is not as long as desired, lasting less than a week at most. Three of the four design specifications were met. A polymer/CNT sensor was successfully designed using a very simple drop casting method that allows for alkane vapors to adhere to the polymer/CNT causing a change in conductivity. UV-Vis testing shows that the CNTs used for sensing indeed contained polymer coatings. The sensor exhibited detection of saturated vapors of dodecane, hexane, hydrazine, ethyl acetate, methanol, ethanol, and acetone, all diluted 29 with dry air to 1% of the saturated vapor. The sensor did not show a false positive in the presence of water vapor. Polymers 4020 and 4027 were more successful at sensing hexane due to its shorter polymer chains and further development needs to be done to effectively detect longer polymer chains such as dodecane. 30 9. REFERENCES [1] C. Wang, B. R. Bunes, M. Xu, N. Wu, X. Yang, D. E. Gross, and L. Zang, “Interfacial Donor–Acceptor Nanofibril Composites for Selective Alkane Vapor Detection,” ACS Sensors, vol. 1, no. 5, pp. 552–559, May 2016. [2] D. Koziej, N. Bârsan, V. Hoffmann, J. Szuber, and U. Weimar, “Complementary phenomenological and spectroscopic studies of propane sensing with tin dioxide based sensors,” Sensors and Actuators B: Chemical, vol. 108, no. 1-2, pp. 75–83, Jul. 2005. [3] H. Alqahtani, M. Sugden, D. Puzzovio, L. Hague, N. Mullin, T. Richardson, and M. Grell, “Highly sensitive alkane odour sensors based on functionalised gold nanoparticles,” Sensors and Actuators B: Chemical, vol. 160, no. 1, pp. 399–404, Dec. 2011. [4] S. Chopra, K. Mcguire, N. Gothard, A. M. Rao, and A. Pham, “Selective gas detection using a carbon nanotube sensor,” Applied Physics Letters, vol. 83, no. 11, pp. 2280–2282, Sep. 2003. [5] D. E. R. Kauffman and A. Star, “Carbon Nanotube Gas and Vapor Sensors,” Angewandte Chemie International Edition, vol. 47, no. 35, pp. 6550–6570, Aug. 2008. [6] F. H. Gojny, M. H. Wichmann, B. Fiedler, I. A. Kinloch, W. Bauhofer, A. H. Windle, and K. Schulte, “Evaluation and identification of electrical and thermal conduction 31 mechanisms in carbon nanotube/epoxy composites,” Polymer, vol. 47, no. 6, pp. 2036– 2045, Mar. 2006. [7] T. M. Clarke and J. R. Durrant, “Charge Photogeneration in Organic Solar Cells,” Chemical Reviews, vol. 110, no. 11, pp. 6736–6767, Sep. 2010. [8] Y. Zhang, M. Xu, B. R. Bunes, N. Wu, D. E. Gross, J. S. Moore, and L. Zang, “Oligomer-Coated Carbon Nanotube Chemiresistive Sensors for Selective Detection of Nitroaromatic Explosives,” ACS Applied Materials & Interfaces, vol. 7, no. 14, pp. 7471–7475, Mar. 2015. [9] L. Niu, Y. Luo, and Z. Li, “A highly selective chemical gas sensor based on functionalization of multi-walled carbon nanotubes with poly(ethylene glycol),” Sensors and Actuators B: Chemical, vol. 126, no. 2, pp. 361–367, Oct. 2007. [10] M. Thomas, B. J. Worfolk, D. A. Rider, M. T. Taschuk, J. M. Buriak, and M. J. Brett, “C60Fullerene Nanocolumns–Polythiophene Heterojunctions for Inverted Organic Photovoltaic Cells,” ACS Applied Materials & Interfaces, vol. 3, no. 6, pp. 1887–1894, Jun. 2011.