• No results found

Bioinspired smell sensor to trace pheromone released by the European spruce bark beetle

N/A
N/A
Protected

Academic year: 2021

Share "Bioinspired smell sensor to trace pheromone released by the European spruce bark beetle"

Copied!
51
0
0

Loading.... (view fulltext now)

Full text

(1)

Linköping University | Department of Physics, Chemistry and Biology Master thesis, 30 hp | Engineering Biology Spring term 2020 | LITH-IFM-A-EX--20/3839--SE

Bioinspired smell sensor to trace pheromone released by the

European spruce bark beetle

Isac Cederquist Examiner: Jens Eriksson Supervisor: Marius Rodner

(2)

Abstract

Forests have as a of late become increasingly plagued with bark beetle infestations as a result of climate change. The damage caused by tree killing bark beetles has within recent years seen a substantial increase. Detecting and removing infested trees at an early stage is an essential part of mitigating the spread of and the damage caused by the beetle. Today, the most common way of early detection is visual detection by forestry personnel. However, this is time consuming with highly variable results. In this thesis a novel approach to tracing the European spruce bark beetle through pheromone detection is investigated. With this approach, the antennae of the beetle were paired with an epitaxial graphene chip in order to create a bioinspired smell sensor. Tests were conducted on the sensor in order to investigate how the resistance changed over the chip as a result of the sensor being exposed to the pheromone 2-methyl-3-buten-2-ol. As a result of the tests, a corelation between exposing the sensor to pheromone and an increase of the resistance over the graphene chip was noted. However, more tests need to be conducted in order to draw any definite conclusions about the efficacy of the sensor in its current form. Additionally there are opportunities to investigate further optimization alternatives regarding the design of the sensor.

(3)

Abbreviations

VOC: Volatile organic compound FET: Field-effect transistor PPB: Parts per billion OR: Odorant receptor

ORN: Olfactory receptor neuron OBP: Odorant binding proteins SiC: Silicon Carbide

AM: Additive manufacturing FDM: Fused deposition modelling VAT: Vat polymerization

SLA: Stereolithography DLP: Digital light processing CAD: Computer-aided design CVD: Chemical vapour deposition PVD: Physical vapor deposition MFC: Mass flow controller GMA: Gas mixing apparatus RH: Relative humidity

EPROM: Erasable programmable read-only memory SEM: Scanning electron microscopy

EDX: Energy-dispersive X-ray spectroscopy RR: Relative response

(4)

Table of Contents

1. INTRODUCTION ...1 1.1 MOTIVATION ... 1 1.2 PURPOSE ... 3 1.3 AIMS ... 3 1.4 DELIMITATIONS ... 3 2. THEORY ...4

2.1 BIOINSPIRED SMELL SENSOR ... 4

2.2 GRAPHENE ... 5

2.3 3DPRINTING AND AUTODESK FUSION 360 ... 6

2.3.1 3D Printing Techniques ... 6

2.3.2 Autodesk Fusion 360 ... 7

2.4 SPUTTER DEPOSITION ... 8

2.5 CONDUCTIVE GEL ... 9

2.6 SCANNING ELECTRON MICROSCOPY ... 9

2.7 BEAGLEBONE ... 10

2.8 ARDUINO AND RELAY ... 11

3. METHOD ... 12

3.1 BEETLE ACQUISITION AND MANAGEMENT ... 12

3.2 CREATING A BIOINSPIRED SMELL SENSOR ... 14

3.2.1 Contact Deposition on Epitaxial Graphene Chip ... 14

3.2.2 Sensor Housing Design in Autodesk Fusion 360 ... 14

3.2.3 Sensor Housing Printing with Form 3 ... 15

3.2.4 Conductive Gel ... 16

3.2.5 Sensor Assembly ... 17

3.3 CREATING A CONTROLLED LAB ENVIRONMENT ... 19

3.3.1 Flow Chamber Design in Autodesk Fusion 360 ... 19

3.3.2 Flow Chamber Printing with Form 3 ... 21

3.3.3 Mass Flow Controllers and Lab Setup ... 24

3.4 EXPERIMENTAL PROCEDURE ... 26

3.5 SCANNING ELECTRON MICROSCOPY ANALYSIS ... 28

4. RESULTS AND DISCUSSION ... 30

4.1 DATA FROM SOURCEMETER MEASUREMENTS... 30

4.2 DATA FROM BEAGLEBONE MEASUREMENTS ... 32

(5)

5. CONCLUSION ... 37

6. ACKNOWLEDGEMENTS ... 38

7. REFERENCES ... 39

(6)

1. Introduction

The following section aims to justify the existence of this thesis and the work that went into it. Additionally, the chapter will present the aims of the thesis as well as its delimitations.

1.1 Motivation

Forests are an essential resource for humanity which not only provide important ecosystems but are also a source of economical property. However, this resource has as of late become increasingly threatened as a result of climate change which also favors the development and lifecycle of bark beetles (1). There are around 6000 confirmed bark beetle species. Among the species there are only a few who have the capacity to attack and kill fully healthy trees. One of these tree killing species is the spruce bark beetle, Ips typographus. The beetle is most occurring in Europe where Norwegian spruce (Picea abies L. Karst) forests are commonly grown and considered an imperative resource. In this region the beetle is an infamous pest known to be one of the most devastating forces to large man-made spruce forests (2)(3).

As previously mentioned, the change in climate has had a substantial effect on the frequency and severity of pest outbreaks resulting in long-lasting effects on both ecosystems and regional economies (1). Several climatic events such as drought and storms have occurred in recent years which has resulted in lasting consequences on forest ecosystems. Droughts affecting tree vigor and storms increasing the frequency of felled trees have created optimal conditions for bark beetles which has culminated in large scale bark beetle outbreaks in many European forests (4). As a result of large-scale pest outbreaks the affected forest owners are forced to cut and harvest affected trees to prevent further spread in an act of sanitation felling. This is also a way for forest owners to salvage the income from the timber that has been killed by the beetles before it loses all its value. However, when this type of forced large scale wood cutting happens the market gets flooded with timber which in turn results in plummeting timber prices (1)(5). The most recent example of this occurred in the Czech

Republic where trees cut as a result of bark beetle infestation increased from 5.3 million m3 to 18

million m3 between the years 2017 and 2018. This caused a tremendous decline in the timber prices

(7)

One distinguishing feature among tree killing beetles is the utilization of aggregation pheromones. The pheromone is released to attract both male and female conspecifics to a weakened or healthy tree in order to initiate a colonization (2). The pheromone released by the beetles partly consists of the chemical compound 2-methyl-3-buten-2-ol which travels through the wind as an odor plume. When the pheromone molecules bind to the odorant receptors located mainly on the antennae of the beetle a neuron signal is sent to the antennal lobe in the brain of the insect. This neuron signal is usually an excitation which is an increase in the frequency that action potentials are fired (6).

The knowledge regarding the spruce bark beetle’s way of chemically communicating and coordinating colonization of trees is today used in two ways to mitigate damages caused by the beetle. The first of these is the installation of pheromone traps. These are constructions rigged with pheromone bait which lures nearby bark beetles in and traps the insects. This method was widely used towards the

end of the 20th century in an attempt to catch a large part of the bark beetle population during an

infestation. However, the method has since been deemed unreliable due to the bark beetle’s tendency to disperse. Thus, this method is primarily used today on a smaller scale as a means for forest owners to monitor the severity of the local beetle infestation (1)(7). The second way is to utilize detection dogs. These dogs are trained to detect several of the aggregation pheromones used by the beetle and are able to identify trees infested by the bark beetle within an hour after an initiated attack. The detection dogs can also identify infested trees from over 100 meters away, as opposed to traditional human visual detection which requires a proximity of under one meter for newly infested trees (8). While detection dogs can assist in recognizing infested trees at an increased range, the dogs mostly used during special circumstances and therefore visual detection by humans is the most common procedure (1).

Visual detection is typically done by forestry personnel and made possible by boring dust found around the bottom of the tree from beetles entering the trunk. Visual detection of newly infested trees is an essential part of mitigating the damage caused by the beetle and makes early sanitation felling possible. However, this type of visual detection is both variable in efficacy and time consuming (1).

(8)

1.2 Purpose

This qualitative thesis covers a novel approach to tracing the European spruce bark beetle, Ips

Typographus, through pheromone detection. The purpose is to create a bioinspired smell sensor

which will fill a niche in the methods used to mitigate damage by the beetle. Going into this thesis work, there are no known earlier works with sensors with bark beetle antennae integrated as receptors. Thus, this thesis is a novel attempt of tracing the spruce bark beetle through the help of a bioinspired smell sensor. The aim is to develop a sensor that will provide a quantifiable means of identifying infested trees at an early stage and assist in the visual detection process. Due to the novelty of this approach the main contribution of this thesis is to design as well as investigate the efficacy and the viability of a bioinspired smell sensor for pheromone detection towards this aim.

1.3 Aims

In order to fulfil the purpose of the thesis the following aims were formulated:

• Design and create a bioinspired smell sensor for pheromone detection.

• Design and create a controlled lab environment to test the bioinspired smell sensor. • Analyze the data recorded from the controlled lab environment.

1.4 Delimitations

The primary restricting factor of the work is a limited project timeframe of 20 weeks in total. This in combination with the novelty of the approach and the implementation of a bioinspired smell sensor in this context lead to a primary focus on the construction and design of said sensor. Additionally, the previously mentioned restricting factors resulted in a data collection process exclusively from a controlled lab environment despite the intention of applying this technology in field experiments eventually.

(9)

2. Theory

This section aims to give the reader a better understanding of the theory behind the methods and instruments used in this thesis.

2.1 Bioinspired Smell Sensor

As a product of evolution insect antennae have developed into instruments with the capacity to detect volatile organic compounds (VOCs) with high specificity. The unique capability of the antennae to detect specific chemical signals in the surrounding environment is higher than almost all existing measuring devices which has resulted in insect antennae-based sensors becoming an increasingly active area of research (9). Additionally, there are many potential areas of application for the technology such as plant protection, environmental monitoring and agricultural production. For example, a bioinspired smell sensor for smoke detection was created utilizing the antennae from the jewel beetle as sensitive element. Antennae-based smell sensors have also been integrated with field-effect transistors (FETs) to enable detection of VOCs on trace levels as low as in the parts per billion (PPB) range (10). That being said the integration of insect antennae with technical transduction devices has a wide range of design options and areas of application (9). However, no matter the design or the area of application the structure and electrophysiology of the insect antennae is a central part.

Insects are dependent on their ability to detect VOCs in order to find sources of food or mating partners. Both plants and animals excrete multiple VOCs that insects pick up and utilize as olfactory clues in their environment. VOCs travel from their source through the air in a complex odor plume to be picked up by odorant receptors (ORs) located primarily in the cell membrane of the olfactory receptor neuron (ORN) on the antennae of the insect. The ORNs in turn are housed within the olfactory sensilla (6). There are multiple different sensilla specific for different signals in the environment such as chemical, mechanical, thermo and humid signals. Once exposed to a VOC, the olfaction process is initiated by allowing the VOC inside the olfactory sensilla through pore tubules. The VOC then arrives at the sensilla lymph. Here, the VOC is captured by odorant binding proteins (OBPs). To avoid degradation of the VOC by enzymes, the VOC is formed into a hydrophobic volatile soluble. The reformed VOC is then transferred through the aqueous sensillum lymph before reaching the outer membrane of dendrites. Thereafter, the OBPs which bind VOCs enable a direct interaction between VOCs and ORs by either releasing VOCs near ORs or directly interacting with the ORs of the cellular membrane. As a result of this interaction between VOCs and ORs, a series of intracellular biochemical reactions is initiated. These reactions result in the opening of ion channels located on the

(10)

cellular membrane which in turn results in cell membrane potential changes. If a large enough quantity of VOCs causes enough ion channels to open simultaneously there is a possibility that the membrane potential changes exceed the threshold of the cells. As a result of this threshold being exceeded, action potentials are generated (9). Conclusively, chemical stimuli in the form of VOCs generate electrical signals as a result of membrane potential changes.

The fundamental principle of the bioinspired smell sensor is to combine the unique sensing features of the insect antennae with a transducer. That way, the signals generated from the VOCs in the environment can be registered and allow for highly specific detection of chemical compounds (9). One material that can be used as a transducer is graphene.

2.2 Graphene

Graphene was discovered in 2004 and has the structure of a two-dimensional honeycomb sheet which consists of sp2 bonded carbon atoms (11). The material is well known for its electrochemical

properties, potentially being the world’s thinnest electrode material with both high electrical conductivity and large surface area (12). The unique electrochemical properties and structure makes graphene a well suitable fit as a transducer in sensor technologies. Sensors generally consist of two components. These are a receptor and a transducer. The purpose of the receptor is to interact with a molecule of interest, for example a VOC, while the transducer is the sensor component which converts the chemical information from molecule of interest into a measurable signal (13).

Graphene can be grown in a multitude of ways. One of these is growing the material on silicon carbide (SiC). SiC is a crystal containing multiple layers of tetrahedrally bonded Si-C atom pairs and has two surface terminations. One of these is the C-face (carbon atoms) termination. The termination on the C-face allows for a fast but inhomogeneous and multi-layered growth of graphene. The other is the Si-face (Si atoms) termination, which is typically used for applications where high quality graphene is needed e.g. graphene intended for sensor application. The reason behind this is that the growth on the Si-face is slower which results in a more controlled growth process. In order for graphene to form, SiC is heated to high temperatures which results in Si atoms being refined, leaving a carbon-rich layer beneath. It is this layer which eventually forms graphene. The first graphene layer that forms has an insulating effect, also called zero-layer graphene. This layer can also be called a buffer layer where approximately 30% of the carbon atoms remain covalently bound to the SiC. Once this layer has been formed, an additional graphene layer is created as a freestanding conducting monolayer. Being only <

(11)

4 Å apart, the buffer layer can act as an electron donor to the graphene layer (14). A schematic of epitaxial graphene grown on SiC can be found in Figure 1.

Figure 1. An illustration of epitaxial graphene grown on SiC. Figure used with permission from (14)

This is how epitaxial graphene on SiC is created. In this thesis, the graphene growth was performed on hexagonal polytype 4H-SiC (0001) on-axis Si-terminated substrates. The sensor created for this thesis aims to measure the change in resistance over an epitaxial graphene chip as a result of exposing the receptors, i.e. severed antennae, to the VOC 2-methyl-3-buten-2-ol. In order to assemble the sensor in a satisfactory manner as well as expose the sensor to the VOC in a controlled environment, 3D printing components was an essential part of the work.

2.3 3D Printing and Autodesk Fusion 360

The purpose of this section is to give the reader a better understanding of different 3D printing techniques and the design software Autodesk Fusion 360.

2.3.1

3D Printing Techniques

Since the start of the 21st century 3D printing technology has changed the landscape and the nature of prototyping and manufacturing substantially. The general principle of a 3D printer is that it allows the user to create a 3D object by having a printer depose several layers of a given material on top of each other. This technique of creating an object by multilayer deposition falls under the label additive manufacturing (AM). There is a wide range of materials available depending on the type of 3D printing that is required for a certain task. The most common techniques today are forms of resin printing called fused deposition modelling (FDM) and vat polymerization (VAT). These are both AM technologies despite being very different technologies. The FDM printer consists of a polymeric

(12)

filament which is extruded as a result of being heated to an appropriate temperature. Then the extruded polymeric filament is deposited in a thin layer on a designated building plate. After this the remaining layers are placed on top of the previous deposited layer in order to create the desired object. This allows for a cheap and simple printing process at the cost of a rougher surface of the final object when compared to other AM techniques. Utilizing VAT techniques with reduced layer thickness allows for smoother and more detailed prints. Here a liquid resin, typically acrylate, is used which is photopolymerized close to room temperature. The VAT printers typically are made of a mobile build platform which is submerged in a resin tank. This build platform is located precisely with regards to the depth of the resin tank and leaves exactly the required amount of resin to cure a single layer. Then, a light source enables the polymerization of one layer at a time as the build platform moves out of the tank step by step (15). An overview of each respective technique can be found in Figure 2.

Figure 2. A basic schematic of the VAT printing principle (a)) and FDM printing principle (b))

VAT printers can be further divided into stereolithography (SLA) and digital light processing (DLP) printers. SLA utilizes a laser which focuses the beam and cures the resin at the correct planar coordinates on the build plate with the help of mirrors and is the most detailed AM process available (16). DLP printers on the other hand utilize a digital projector screen which projects and cures all the points of the layer at the same time (15).

2.3.2

Autodesk Fusion 360

3D printing has revolutionized the engineering world and has allowed teams and private persons to print test parts for projects at a trivial cost. In order to create custom made 3D models from a 3D printer the use of advanced computer-aided design (CAD) software is required. Creating 3D models through a CAD software can be divided into two categories. These are parametric modelling and mesh

(13)

modelling. With a mesh modelling software there are similarities to sculpturing in real life. Here 3D models are created through pushing and pulling a virtual mesh which is often utilized to create animations or video game characters. Here, precise dimensions are not as important. With a parametric modelling software, parameters and dimensions are a fundamental part of creating a 3D model. Shapes and objects are defined by their dimensions and their coordinates in the 3D design landscape. This type of modelling is often used by engineers since specifying the exact shape and size of the different features of a part is an essential part of the design process (17).

In this thesis creating airtight components was an essential part in order to carry out tests in a controlled environment. Therefore, the Form 3 SLA printer from Formlabs was used to print a sensor housing and a flow chamber designed in the parametric CAD software Autodesk Fusion 360.

2.4 Sputter deposition

Two frequently used techniques for depositing a thin film of a material of interest (target) on another already existing surface (substrate) are physical vapor deposition (PVD) and chemical vapor deposition (CVD). When using PVD a solid film of the material of interest is produced on an existing surface through a mechanical, thermal or electromechanical process. In the case of CVD a liquid or gaseous precursor of the material of interest is transformed on an existing surface through a chemical process (14).

Magnetron sputter deposition is one example of a PVD technique. Here an inert gas such as Ar can be ionized by an energy source and used as a sputter gas to be directed at a sputter target. As a result of directing the ionized sputter gas at the sputter target, the target is bombarded and eject target atoms and secondary electrons. A magnet is used to restrict sputter electrons close to the target surface which increases the ionization rate at the target surface. The target atoms then move and adhere to the substrate and creates a thin film of the material which the target consists of. Magnetron sputter deposition can be used as a so-called soft deposition technique in which cases the atoms of the material of interest is presented to the surface with low kinetic energy (14). In this case a soft magnetron sputter deposition technique was used to create electrodes in Ti/Au on the epitaxial graphene chip without damaging the graphene. The sputtering process was a key part in creating the sensor and making it compatible with data collecting instruments and is illustrated in Figure 3.

(14)

Figure 3. An illustration of the Ar beam being directed at the target (e.g.Ti/Au) resulting in target atoms adhering to the substrate (e.g. epitaxial graphene chip). Figure used with permission from

(14)

2.5 Conductive gel

When attempting to measure biopotentials from a biological surface, utilizing a liquid or solid gel can provide a continuous conductive path between the surface and the sensing element. In order for a gel to gain an electrical current flow it generally contains salt in the form of KCl or NaCl. A gel containing a high concentration of salt can reliably act as a conducting element between the surface and the sensing element (18).

In this thesis, a conductive agarose gel was utilized with the intent of carrying signals from the receptors (i.e. severed antennae) to the graphene transducer. Additionally, the gel was used as an attempt to shield the graphene transducer from unwanted background noise that might occur in the controlled lab environment e.g. background gas.

2.6 Scanning Electron Microscopy

Scanning Electron Microscopy (SEM) is a way of combining a large depth of field with high-resolution imaging. This is made possible as a result of the electrons ability to be focused utilizing electrostatic and electromagnetic lenses and due to the short wavelength of the electrons. Additionally, as a result of the strong interactions between electrons and matter data can be generated about matter down to a nanoscopic level. Especially the ability to visualise the topography of bulk specimens has proven to be valuable. When analysing metallic materials the SEM process is generally relatively straightforward. The first step of analysing a material of interest is achieving high vacuum in the SEM

column which is usually around 10-3-10-5 Pa. As a result of this requirement, the material of interest

(15)

specimens, foods, foams etc. must be prepared to eliminate any potential volatile substances. Additionally, SEM bombards the sample of interest with high-energy electrons. This bombardment can result in a rapid generation of negative charge if the sample is not electrically conductive. However, if the sample is electrically conductive the negative charge can be dispersed over a grounded specimen holder where the sample of interest is mounted. Since metallic samples are both electrically conductive and lacking volatile compounds, they can easily be imaged with a SEM (19).

In order to gain information about the elemental composition of the specimen of interest, SEM can be paired with techniques such as Energy-dispersive X-ray spectroscopy (EDX). Here, X-ray photons emitted from a specimen as a result of the SEM process are collected through a solid-state detector. The X-ray photons have characteristic energies depending on electronic transitions among the atoms of the specimen that is being investigated. These X-ray photons are recorded and can then be used to create an energy dispersive spectrum giving information about the chemical composition of the specimen (20).

In this project, SEM was used to identify the damages caused to the sputtered gold electrodes created on an epitaxial graphene sensor which rendered the sensor useless. The data generated from the SEM was an essential part to confirm what damage had been done to the electrodes on the chip and update the sensor design accordingly to prevent future issues.

2.7 BeagleBone

BeagleBone is a hardware platform offering a wide range of so-called Beagle boards. These boards are compact Linux computing platforms which allows the user to build software applications for a low cost on low level electronic circuits. The computing capability of the boards make them suitable platforms for a vast range of projects which can benefit from custom coded software (21).

For this thesis, the Beagle board “BeagleBone Black Industrial” was programmed and utilized in order to measure the resistance over the graphene transducer over time. Originally a SourceMeter, which is a more accurate measuring instrument, was used to register the resistance over the epitaxial graphene chip. However, since the setup including the SourceMeter ceased to work during the project the BeagleBone was set up as a workaround. The majority of the data was recorded from the BeagleBone setup. It should be noted that the BeagleBone is inherently noisier and since the epitaxial graphene chip allows for detection of very small changes, pairing it with a BeagleBone will cripple the potential of the sensitive material.

(16)

2.8 Arduino and Relay

An Arduino is at its heart a small scale computer on a chip, containing all the regular features such as a processor, RAM memory and a small storage capability in the form of erasable programmable read-only memory (EPROM). Paired with multiple options for electrical connections, the Arduino board is an excellent tool for controlling electronic devices and is a compatible tool to many custom built electronic projects (22).

In this project, an Arduino Uno board was programmed and paired to a relay. Pairing this equipment allowed for regulating the voltage applied to the 2-way valve on a set time interval, resulting in the closing and opening of the valve in the experimental setup explained previously.

(17)

3. Method

This section aims to give the reader a detailed explanation about the steps that were performed in order to fulfil the first two aims that were formulated in the introduction of the report. Additionally, the experimental procedure to generate the data presented in the results section is explained.

3.1 Beetle Acquisition and Management

An essential part of this project revolved around acquiring European spruce bark beetles in order to retrieve the antennae which serve as the receptor in the sensor. Since previous research on severed antennae has estimated a lifetime of approximately 2 hours with a notable exponential decay over time (23) it was imperative to have access to a source of living bark beetles. This was made possible through the help the senior lecturer Charlotte Norrman and Södra Skogsägarna. Beetles were obtained by emptying pheromone traps around Hjulsbro, Linköping which contained a varying amount of beetles. The contents of the traps were then emptied in a glass jar and transported to Linköping University. This process is illustrated in Figure 4. The glass jar was filled with pieces of bark and sprayed with a water spray bottle in order to increase the survivability of the beetles.

Figure 4. A depiction of the acquisition of bark beetles. A pheromone trap (a)) was located in the woods. The trap compartment could then be opened (b) c)) and emptied of its contents into a glass

jar (d))

Once the beetles had been successfully transported to Linköping University, the isolation of the antennae could begin. First, a sorting process took place where live beetles were separated from deceased ones and put in a petri dish sealed with a lid. Once this was done the acquisition of antennae from living beetles could begin. This was done under a microscope using a scalpel to separate the head

(18)

of the beetle from the body before isolating the antennae of the beetle with the help of fine tweezers as illustrated in Figure 5.

Figure 5. The figure illustrates a spruce bark beetle (a)) under a microscope and a severed spruce bark beetle antenna (b)) held by a fine tweezer

The beetles were about 4 mm long and relatively immobile. Separating the beetle’s head from its body was a simple process with the help of a scalpel. However, isolating the antennae of the beetle could be slightly more finical mostly due to the small size of the antennae. In order to simplify the removal process, two tweezers were used. One was used to pin the carapace of the beetle to the bottom of the petri dish while the other was used to remove the antennae. The antennae were then gathered on a separate petri dish until a number of antennae had successfully been removed. In the tests including antennae they were always removed and integrated in the sensor as the last step before conducting the test due to the relatively short lifetime of the antennae.

(19)

3.2 Creating a Bioinspired Smell Sensor

This subsection aims to give the reader all necessary information about the work associated with creating a bioinspired graphene sensor.

3.2.1

Contact Deposition on Epitaxial Graphene Chip

First, an epitaxial graphene 4″ wafer was diced into 7×7 mm squares. In order to create contact electrodes in gold on the graphene chip, sputter deposition was used. This took place in a clean room. In order to ensure there were no photoresist residue on the graphene chip, it was carefully cleaned with acetone and ethanol. The photoresist was originally placed on the graphene wafer before being cut into its designated shape in order to prevent the graphene surface from being damaged. Then the diced graphene was dried with a nitrogen gas gun before being baked in the oven at 100°C for ten minutes. The chip was then covered with a mask only leaving two rectangular fields with the dimensions 6×1.3 mm open where the electrodes were to be created in the sputtering process. The chip and mask were fixated with a thin adhesive copper tape before being placed on a big sample holder where chip and mask were fixated with screws. Then the turbo pump used to alter the pressure

was stopped in order to flush the load chamber with N2. The sample holder was then inserted into the

chamber and the turbo pump was started again. The load chamber was then pumped down to approximately the same pressure as the deposition chamber. Once the same pressure was achieved in both chambers, a valve was opened between the chambers in order to transfer the sample to the deposition chamber. First, a Ti layer was deposited onto the rectangular fields. This was done introducing an Ar flow to the target before igniting plasma. Then a sputtering cleaning process took place for about 30 seconds. After this a shutter was opened and the Ti was deposed on the rectangular fields for around 30 seconds. The shutter was then closed before turning off the plasma and the Ar flow. The same procedure was then repeated but for Au deposition which took a total of approximately 10 minutes. The sample was then moved back into the load chamber before closing the valve between load and deposition chamber. Then the turbo pump was stopped and the load

chamber was once again flushed with N2. The sample holder was then removed before starting the

turbo pump again. As a last step the sample was removed from the sample holder and a gold sputtered epitaxial graphene chip had been created.

3.2.2

Sensor Housing Design in Autodesk Fusion 360

In the parametric CAD software Autodesk Fusion 360, a sensor housing was created with the purpose to hold the graphene chip in place. The design was composed of two separate parts illustrated in Figure 6. One of the parts was the sensor housing bottom which was a 10×10×1.5 mm square with

(20)

skewed edges for easier removal from the build plate. The sensor housing bottom also had a 7×7×1 mm square cutout on the inside. This part of the design was made with the dimensions of the epitaxial graphene chip in mind with the intent to hold it in place in the sensor hosing. The other part was the sensor housing top. This was a 6.9×6.9×1.5 mm rectangle with beveled edges for easier removal from the build plate and insertion into the sensor housing bottom. The sensor housing top also had one 6×1.4×1.5 mm rectangular cutout on either side to make space for the sputtered golden contact electrodes on the graphene chip. Additionally, a 6×2×1.5mm rectangular cutout was made in the center of the sensor housing top, intended to leave an exposed area of graphene in the center where the conductive agarose gel and antennae could be assembled onto. The CAD file was then saved in an STL format to be uploaded to the Form 3 printer.

Figure 6. An illustration of the sensor housing bottom (a)) and the sensor housing top (b)) in CAD

3.2.3

Sensor Housing Printing with Form 3

When the STL file had been uploaded to the Form 3 SLA printer the sensor housing top and bottom were printed with a layer resolution of 25 µm which took approximately 30-45 minutes. Once the printer had finalized curing the resin for the print, the build plate was removed from the printer and

(21)

cleaned. Ethanol was used for this, being a suitable substitute for isopropyl alcohol which dissolves any liquid resin and is the recommended liquid for cleaning Formlabs SLA printers (24). The printed parts where then gently removed from the build plate using a small metal spatula and placed in a plastic container filled with enough ethanol to keep the parts submerged. Then the build plate was put back in place on the printer. Thereafter the plastic container was placed in an ultrasonic bath for 15-20 minutes in order to remove any uncured resin which finalized the printing process. The results from a print can be seen in Figure 7.

Figure 7. The figure illustrates the sensor housing bottom (a)), the gold sputtered epitaxial graphene chip (b)) and a version of the sensor housing top (c))

3.2.4

Conductive Gel

The purpose of the conductive gel was partly to shield the graphene chip from external signals with the hope to pick up clear, undisturbed signals from the receptors, in this case the antennae of the beetle. In order to create the gel, 5.844g NaCl was dissolved in 100 ml de-ionized water in order to achieve a 1 molar electrolyte solution. To create a gel which matched the 1.5 mm height of the middle rectangular part of the sensor housing top, a petri dish with a diameter of 8.6 cm was used. In order to achieve a gel thickness of 1.5mm, 8.71 ml of the 1 molar electrolyte solution was added to a small glass cup. Then, 0.261g of agarose was added to the solution in the cup to obtain an agarose gel with the rigidness of 3%. Then the glass cup was stirred and microwaved on medium effect for approximately 20 seconds to cure the gel. The viscous liquid was then poured into the petri dish with

(22)

a diameter of 8.6 cm which was kept in room temperature in an airtight bag to prevent the gel from drying.

3.2.5

Sensor Assembly

Once the previously described steps were completed, the sensor could successfully be assembled. First four plastic coated copper wires were soldered to a 25×25 mm hole chip to create the base for the sensor. Then the epitaxial graphene chip was placed in the sensor housing top. These parts were then fixated to the hole chip with a silver-based epoxy glue. In order to connect the gold sputtered electrodes on the graphene chip to the soldered copper wires on the hole chip, multiple fine gold bond wires were carefully placed in pools of silver-based epoxy glue on both components. This was done in order created a conductive link between the chip and the copper wires since soldering the copper wires directly to the chip would ruin the graphene. As a result of linking the chip with the copper wires through fine gold bond wires, it was made compatible with various measuring equipment. This assembly can be seen in Figure 8.

Figure 8. An illustration of the hole chip with soldered copper wires (a)) and the same hole chip with the sensor housing mounted on top (b)). In the sensor housing the graphene chip is located with thin

(23)

In order to cure the silver-based epoxy and thereby finalize the sensor component, the sensor seen in Figure 8 b) was baked in an oven at around 100 °C for two hours. The sensor seen in Figure 8 b) was the second out of two. The first of these sensors is illustrated in Figure 9. Unfortunately, this sensor suffered damage on one of the electrodes as a result of potential corrosion from liquid leaking from the gel in the middle section to said electrode. In order to avoid this from happening to the sensor illustrated in Figure 8 b), a barrier of resin was cured with a UV-laser around the edge of the sensor housing top.

Some tests were performed on the chip which required a piece of conductive gel which was sometimes also paired with severed beetle antennae. The procedure was then to cut out a piece of the conductive gel and place it over the graphene in the middle section of the sensor. Then, antennae from bark beetles were carefully removed from a number of beetles to ensure that not all of the antennae applied to the gel were damaged and therefore limited in their performance to produce an electrical current based on their interaction with potential VOC. The result of this procedure can be seen in Figure 9.

(24)

3.3 Creating a Controlled Lab Environment

This subsection aims to explain how a controlled lab environment was achieved to expose the sensor to the VOC in a controlled manner.

3.3.1

Flow Chamber Design in Autodesk Fusion 360

The flow chamber component was designed with the parametric CAD software Autodesk Fusion 360. The purpose of this component was to create an airtight environment around the sensor in order to expose the sensor to background gas and the VOC. Initially, this part was designed with a tube sticking out on either side with the intent to attach the plastic wires leading the gas from the MFCs to the sensor. The initial flow chamber CAD design can be seen in Figure 10.

Figure 10. A figure illustrating the initial flow chamber design in Autodesk Fusion 360

(25)

Figure 11. A figure illustrating the initial design of the flow chamber with the flow chamber top placed on top of the flow chamber bottom with gas tubes attached on either side

Multiple tests were done with this flow chamber implemented before it was realized that this flow chamber was not air-tight. This rendered the results from the tests useless and an updated design of the flow chamber was made according to Figure 12. The flow chamber consisted of two parts. One of these parts was the flow chamber bottom which was a 50×50×4 mm square with a 37×37×2.5 mm cutout in the center. Additionally, four holes with a diameter of 2 mm to lead the plastic covered copper wires out from the flow chamber to the bottom of the flow chamber bottom. Here four half circle cutouts with a diameter of 3 mm were made in order to make space for the wires.

The other part was the flow chamber top. This was a 50×50×22.5 mm square with beveled edges for easier removal from the build plate. In the middle of the flow chamber top, a 36.9×36.9×2.5 mm square was extruded in order to fit snugly into the flow chamber bottom. Additionally, a 33.9×33.9×23.5 mm square was cut out in the center of the flow chamber top in order to leave space for the sensor and to create a controlled airtight environment where gas could flow past the sensor. Finally, a circular cut out with a diameter of 5.5 mm was made through the housing top in order to

(26)

screw metal components making it compatible with existing lab equipment for airtight systems. The CAD file was then saved in an STL format to be uploaded to the Form 3 printer.

Figure 12. An illustration of the updated design of the flow chamber top (a)) and the flow chamber bottom (b))

3.3.2

Flow Chamber Printing with Form 3

When the STL file had been uploaded to the Form 3 STL printer the flow chamber top and bottom were printed with a layer resolution of 25 µm which took approximately four hours. A picture of a flow chamber attached to the build plate can be seen in the Figure 13.

(27)

Figure 13. The components of a flow chamber recently cured and attached to the build plate of the Form 3 printer

Once the printer had finalized curing the resin for the print, the build plate was removed from the printer and cleaned with ethanol. The printed parts where then gently removed from the build plate using a small metal spatula and placed in a plastic container filled with enough ethanol to keep the parts submerged. Then the build plate was put back in place on the printer. Thereafter the plastic container was then placed in an ultrasonic bath for 15-20 minutes in order to remove any uncured resin which finalized the printing process of the flow chamber. The flow chamber parts were then spray-painted in black in order to eliminate the risk of having light sources influencing the resistance recordings of the chip. The results from a spray-painted print can be seen in Figure 14.

(28)

Figure 14. Both flow chamber components newly printed and spray-painted

Once the print had dried, threads were used to make the holes compatible with screwable bolts which were designed to create an airtight connection between the flow chamber top and the gas tubes of the system. Additionally, Teflon tape was wrapped around the extruded section of the top and the bolts to ensure that an airtight design was achieved which can be seen in Figure 15.

Figure 15. The top of the flow chamber turned upside down with teflon tape attached to both the extruded section and the bolts on either side

(29)

The assembled sensor could then be placed in the housing by inserting the four copper wires into the holes of the flow chamber bottom and filling the half circle cutouts with glue to lock the wires in place as well as sealing off the holes to ensure that the design was air-tight which was also tested with the help of a flow meter.

3.3.3

Mass Flow Controllers and Lab Setup

Mass flow controllers (MFCs) were used in order to expose the sensor to a well-known gas flow in known concentrations. Traditionally, these are also connected to a gas mixing apparatus (GMA) (14) which allows for automatic time regulated adjustments to the gas flow through the MFC and in turn the sensor. However, due to technical difficulties the GMA was not available which was circumvented by using two free-standing MFCs which allowed a well-known gas flow to reach the sensor. The gas

stream of pure nitrogen gas (N2) from one of the MFCs with a flow of 80 ml/min was directed through

a flask of water called a bubbler in order to achieve 80% relative humidity (RH) which increased the longevity of the conductive gel applied to the sensor. The other MFC provided a gas stream of pure

oxygen (O2) with a flow of 20 ml/min. The two separate streams were then joined for a total gas flow

of 100 ml/min with 80% RH. The gas stream then reached a 2-way valve which controlled whether the stream passed straight to the sensor or if the stream passed through a separate chamber. When collecting data, this chamber was either empty or containing a pheromone infused cloth. In the case of the chamber being empty the sensor would only be exposed the gas stream throughout the entire experiment. In the other case where the chamber contained a pheromone infused cloth the sensor would be exposed to the VOC 2-methyl-3-buten-2-ol for a set period of time. A schematic illustration of the setup that was used to achieve a controlled lab environment can be found in Figure 16 below.

Figure 16. The figure presents a basic schematic of the experimental setup from MFCs on the left to Waste on the right

(30)

The schematic illustrated in Figure 16 is shown in its real form in Figure 17.

Figure 17. The complete lab setup including all necessary components in order to conduct the experiment

A more detailed breakdown of the components can be found in Figure 18. Here, a) 1 illustrates the bubbler which was used to reach 80% relative humidity and a) 2 illustrates the MFCs providing a total

gas flow of 100 ml/min consisting of 80% N2 and 20% O2. Further, b) 3 is the power source connected

to the valves, b) 4 is the Arduino connected to b) 5, both allowing the valves found in c) to direct the flow of the gas stream on a time-based interval. In c), the gas flow either takes the path of the valve found at c) 7.1 going directly to the sensor, or it takes the path along the c) 7.2 valves, going past the pheromone chamber seen at c) 6. Here, depending on if the chamber has been filled with a pheromone infused cloth, the sensor would be exposed to the VOC or only to the background gas flow throughout the current test cycle. Finally, d) 8 illustrates the assembled graphene sensor inside the

(31)

flow chamber with wires going to the BeagleBone in e) 9 which allowed for the recording of the resistance over the chip.

Figure 18. A figure illustrating the different components of the lab setup

3.4 Experimental Procedure

In order to investigate the efficacy of the bioinspired smell sensor for pheromone detection it is essential to perform tests on the sensor in a controlled lab environment. Therefore, initial test of the resistance over the graphene chip were performed using a SourceMeter connected to a computer. Here, 3V were applied from the SourceMeter between the electrodes on the chip. The resistance was then measured over the chip with a pheromone infused cloth in the pheromone chamber and a conductive gel on top of the graphene. The pheromone cloth had been ordered from the internet under the name Typosan P306 and is typically used as bait in pheromone traps. In this case however, the cloth was cut to fit the pheromone chamber as illustrated in Figure 19.

(32)

Figure 19. The figure illustrates the package ordered from the internet (a), b)). The cloth was then cut into smaller cubes (c)) before being placed in the pheromone chamber (d))

Tests were ran either with antennae integrated into the conductive gel or without integrated antennae. In the case where antennae were integrated, eight were used in order to minimize the risk that all of the integrated antennae were damaged in the removal process. During the tests, the sensor was initially exposed to 30 minutes of background gas in an attempt to make the sensor stabilize. Thereafter the sensor was exposed to five 15-minute cycles where the gas path was altered by the 2-way valve automatically. During five of these minutes the gas flow was directed through the pheromone chamber and the remaining ten minutes the gas flow went straight to the sensor exposing it to pure background gas. The reasoning behind letting the background gas exposure be twice as long as the exposure to pheromone was to not overexpose the antennae to pheromone and let the receptors rest a longer period of time between exposures. The data that was plotted and analyzed was the last four cycles since the first cycle often is an adjustment cycle and the first 30 minutes were meant to stabilize the resistance readings. Unfortunately, the server required to collect data from the computer and SourceMeter setup broke down which made it impossible to quantify the measurements with this setup to a satisfactory manner.

After the server breakdown, the remaining tests were done as described below in order to gather more data about the resistance readings over the chip. The parameters that were tested here was measuring the resistance with or without conductive gel, with or without pheromone and with or without antennae integrated into the gel. During the tests, 0.5V were applied from a BeagleBone Black Industrial to the sensor. Unfortunately, this setup also ran into difficulties collecting data since the BeagleBone randomly would stop collecting data which resulted in the measurements with antennae included being interrupted early and thus rendered useless. Additionally, the sensor that had previously been assembled ceased to work due to a problem which occurred at one of the gold sputtered electrodes. Therefore, a new sensor had to be assembled in order to run more tests. By this

(33)

time the season for the bark beetles was slowly coming to an end and the time assigned for the project had already been exceeded. Despite this, a new BeagleBone setup was created to run a slightly shorter test cycle in order to at least gather some changing the gel and pheromone parameters. The tests that were carried out on the sensor initially exposed the sensor to 30 minutes of background gas. After this exposure, four ten-minute cycles took place where the gas flow was altered by the 2-way valve every five minutes. The gas was either directed directly to the sensor or to the pheromone chamber which was either empty or containing a pheromone infused cloth. The data that was analyzed were the three last ten-minute cycles following the same reasoning as described above.

The relative response (RR) has been calculated between the start and the finish of the gas being routed through the pheromone chamber. The purpose of this is to explore how the sensor is affected by pheromone exposure depending on the previously mentioned parameters. The resistance in all of the tests was sampled with a frequency of one recording per second.

3.5 Scanning Electron Microscopy Analysis

As mentioned previously, damage was caused to one of the Au electrodes on the sensor. In an attempt to get more knowledge about the damage that had been caused, SEM was used. In Figure 20 the damaged electrode is illustrated by a) 1 where b) illustrates the sensor being broken apart before being analyzed in the SEM.

Figure 20. A figure illustrating the damaged Au electrode on the graphene chip (a)) and the top of the broken sensor being separated from the bottom in preparation of SEM analysis (b))

(34)

The SEM analysis took place in a clean room. In Figure 21, a) 1 and 2 illustrate the broken sensor housing top and bottom respectively attached to grounded sample holders. The sample holders were then placed on a metal disk illustrated in b) and fixated with screws. The metal disk was then inserted into the SEM chamber where a vacuum was gradually built up. Once high vacuum ≈10 Torr had been achieved, the sample could be analyzed by focusing the electron cannon on the sample of interest.

Figure 21. The broken sensor (a) 1 being the sensor housing top and 2 being the epitaxial graphene chip) placed on grounded specimen holders before being placed on a metal disk (b)) going into the SEM

(35)

4. Results and Discussion

This section aims to present the results generated from tests described in the experimental procedure above.

4.1 Data from SourceMeter Measurements

In the following subsection the data from the SourceMeter measurements is presented.

Figure 22. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The green sections of the plot indicate the time periods where the gas flow

was directed through the pheromone chamber. The red sections of the plot indicate the time periods where the gas flow was directed straight to the sensor

(36)

Figure 23. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The green sections of the plot indicate the time periods where the gas flow

was directed through the pheromone chamber. The red sections of the plot indicate the time periods where the gas flow was directed straight to the sensor

By looking at the results presented in Figure 22 and Figure 23, it becomes apparent that there is a strong corelation between pheromone exposure and a change in resistance in this particular test setup. There is a clear response when the gas flow is altered to take the path through the filled pheromone chamber. To better judge the difference in resistance achieved in each respective case, the RR was calculated. The results of the RR calculations can be found in Table 1.

Table 1. The table lists the RR calculated between the start and end of each respective pheromone exposure illustrated in green in Figure 22 and Figure 23

Interval RR with antennae RR no antennae

0-300s 0.75% 2.36%

900-1200s 0.97% 1.65%

1800-2100s 1.15% 1.46%

(37)

The RR that was calculated and listed for comparison in Table 1 shows that the RR is higher without antennae than with antennae although the trends with increased exposure are opposite. While this is a reason to question the efficacy of the sensor in its current form, more tests need to be done in order to give a definitive answer to this. However, it should be noted that the general downward slope of the graph associated with the background exposure seen in Figure 22 is steeper than that of Figure 23. This might be an indication that the stabilization of the graphene was ongoing in Figure 22. That being said, RR is based on having a stable point of reference. In the previously mentioned figures, the resistance reading indicates that a stable point likely had not been achieved. Therefore, more tests should be conducted for a chance at drawing any definitive conclusion regarding how the RR is affected by integrating antennae into the sensor.

Additionally, in both Figure 22 and Figure 23, a similar pattern was seen as a result of pheromone exposure in both figures despite no antennae being integrated in the latter. This might be an indication that the conductive gel is not entirely sealing off the graphene section in the middle of the chip. It could also mean that the pheromone in the surrounding likely permeates the gel and reacts with the chip. This implies there might be reason to investigate alternative conductive and shielding materials for the connection between receptor and transistor for the current design of the sensor.

4.2 Data from BeagleBone Measurements

(38)

Figure 24. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

periods where the gas flow was directed straight to the sensor. A total of four measurements were done with this setup where the remaining plots can be found in the Appendix Figure A1 – Figure A3

Figure 25. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

periods where the gas flow was directed straight to the sensor. A total of four measurements were done with this setup where the remaining plots can be found in the Appendix Figure A4 – Figure A6

(39)

Figure 26. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The green sections of the plot indicate the time periods where the gas flow

was directed through the pheromone chamber. The red sections of the plot indicate the time periods where the gas flow was directed straight to the sensor. A total of four measurements were done with this setup where the remaining plots can be found in the Appendix Figure A7 – Figure A9

Figure 27. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The green sections of the plot indicate the time periods where the gas flow

(40)

periods where the gas flow was directed straight to the sensor. A total of four measurements were done with this setup where the remaining plots can be found in the Appendix Figure A10 – Figure

A12

The figures illustrated in this section and their corresponding figures located in the Appendix sections were for the most part inconclusive. The setup used to produce the data in this section using the BeagleBone differ greatly from that of the previous section using the SourceMeter. Not only is the recording equipment changed but also an entirely new sensor was assembled for the BeagleBone measurements. Based on the inconsistent results that were achieved from the measurements with the BeagleBone, it is difficult to arrive to any reliable conclusions based on the work done in this thesis.

4.3 Scanning Electron Microscopy Results

In the following subsection the data acquired from the SEM is presented.

Figure 28. An illustration of a gradual zoom ranging from 65-32k x zoom of the damaged gold sputtered electrode on the broken graphene chip illustrated in Figure SEM a) 2. Here, the outline of

(41)

the graphene chip is illustrated as well as the damaged electrode on top of the graphene chip (a)). A gradual zoom of the damaged electrode is illustrated (b)-d)) down to a micro level

Table 2. The table contains data acquired from the EDX feature in the SEM. The EDX provided information regarding the elemental composition of the damaged gold sputtered electrode illustrated in Figure SEMRESULTS c). The Wt% is the relative concentration of a certain element in

the image Element Wt% Wt% Sigma C 22.54 0.70 O 1.42 0.26 Na 10.98 0.20 Si 0.31 0.05 Cl 21.90 0.27 Ag 42.86 0.49 Total: 100.00

According to Table 2, levels of Na and Cl are detected at the electrodes which confirms a migration from the gel to the electrodes in the initial sensor design. This migration likely caused a corrosion which severely affected the sensors ability to accurately read the resistance over the graphene chip. Additionally, despite the electrode being created out of Ti and Au, none of these elements were detected by the EDX which is surprising. Instead a high amount of Ag is detected, which can be explained by the silver-based epoxy glue which was used to adhere the fine gold bond wires to the electrodes.

Finally, based on the descriptive subsections following sections 2 and 3, the replicability should be sufficient for others to reproduce the experiments performed in this thesis and conduct the same study elsewhere. With regards to the ethics revolving this thesis, is it correct to kill innocent beetles? Since this beetle is considered an invasive pest and that the beetles caught in the pheromone traps would head for a certain death either way, one can argue that the killing of the beetles for this purpose is justified. Additionally, laws and regulations regarding insects are non-existent, however the question whether taking a life no matter how small is important to raise. With regards to another type

(42)

of ethics, the results in the thesis have been presented with full transparency and without bias according to research ethics praxis.

5. Conclusion

It is safe to say that the creation of a sensor with the help of 3D printing and an epitaxial graphene chip is possible. The data presented in section 4 is in one way or another proof that the sensor works on a conceptual level. Additionally, achieving a controlled lab environment with the help of 3D printing custom components is also possible. However, in order to draw any conclusion regarding whether or not the bioinspired smell sensor created in this thesis is an efficient pheromone detector requires more successful tests of exposing the sensor to pheromone while antennae are integrated in the conductive gel are needed.

Finally, there are no known earlier works with sensors and receptors in the form of bark beetle antennae to compare this particular work to. However, there is evidence that pairing antennae with a transducer has yielded satisfying results in other cases. Despite the uncertainty of the results presented in this particular thesis, the data presented in section 4.1 serves as a proof-of-concept of a novel sensor solution to a severe socio-economical and environmental problem.

(43)

6. Acknowledgements

Finally, I would like to extend my gratitude to the following people who have been an immense help during this work.

Jens Eriksson: Thank you for taking on this obscure project and for offering your guidance and expertise during this journey.

Marius Rodner: A special thanks to you for putting up with my questions and taking time out of your day to supervise this project which got quite drawn out. Your help and expertise has been invaluable.

Charlotte Norrman: For being the number one spruce bark beetle cab driver in Sweden, thank you for transporting hundreds of beetles from your summer house all the way to Linköping. Without you, the project would not have started and without your support no one knows how it would have ended.

Mike Andersson: A big thanks to Mike for being the ever-present omniscient person on LiU and in the gas lab. Thank you for taking the time to help me even though you had no obligation to do so.

Jan Ybrahim: Thanks for good times and almost setting fire to the lab. In all seriousness, thank you for sharing your knowledge regarding sensors, graph plotting, Arduino programming and much more. Half of the contents of this project probably came together because of your guidance and assistance. Thank you for all the help.

Guillem Domenech: Thank you for giving valuable feedback regarding oral presentations and especially for giving me a hand with the SEM, that thing sure is hard to operate.

Anke Suska: For your expertise and guidance regarding the creation of conductive gels, thank you!

Håkan Rönnberg: For allowing me to stroll around in your forest and empty your bark beetle traps, thank you so much. I hope you have an easier time with this pest the coming years despite the evidence saying otherwise.

(44)

7. References

1. Hlásny T, Krokene P, Liebhold A, Montagné-Huck C, Müller J, Qin H, et al. Living with bark

beetles: impacts, outlook and management options. In.

2. Öhrn P. The spruce bark beetle Ips typographus in a changing climate-Effects of weather

conditions on the biology of Ips typographus.

3. Hlásny T, Barka I, Roessiger J, Ladislav Kulla •, Trombik J, Sarvašová Z, et al. Conversion of

Norway spruce forests in the face of climate change: a case study in Central Europe.

4. de Groot M, Ogris N. Short-term forecasting of bark beetle outbreaks on two economically

important conifer tree species. For Ecol Manage [Internet]. 2019 Oct 15. Available from: https://doi.org/10.1016/j.foreco.2019.117495

5. Holmes TP. Price and Welfare Effects of Catastrophic Forest Damage From Southern Pine

Beetle Epidemics. Vol. 37, Forest Science. 1991.

6. Andersson MN. Olfaction in the Spruce Bark Beetle , Ips typographus. Sciences-New York.

2011.

7. Duelli P, Zahradnik P, Knizek M, Kalinova B. Migration in spruce bark beetles (Ips typographus

L.) and the efficiency of pheromone traps. J Appl Entomol. 1997;121(6):297–303.

8. Johansson A, Birgersson G, Schlyter F. Using synthetic semiochemicals to train canines to

detect bark beetle–infested trees. Ann For Sci. 2019;76(2).

9. Hsia KJ, Wu C, Liu Q. Bioinspired smell and taste sensors. Bioinspired Smell Tast Sensors.

2015;1–328.

10. Schütz S, Weissbecker B, Hummel HE, Apel KH, Schmitz H, Bleckmann H. Insect antenna as a

smoke detector. Nature. 1999 Mar 25;398(6725):298–9.

11. Pumera M. Graphene-based nanomaterials and their electrochemistry. Chem Soc Rev

[Internet]. 2010 Oct 19;39(11):4146–57. Available from: www.rsc.org/csr

12. Raj MA, John SA. Graphene-modified electrochemical sensors [Internet]. Graphene-Based

Electrochemical Sensors for Biomolecules: A Volume in Micro and Nano Technologies. Elsevier Inc.; 2018. 1–41 p. Available from: http://dx.doi.org/10.1016/B978-0-12-815394-9.00001-7

13. Peña-Bahamonde J, Nguyen HN, Fanourakis SK, Rodrigues DF. Recent advances in

graphene-based biosensor technology with applications in life sciences [Internet]. Vol. 16, Journal of Nanobiotechnology. BioMed Central Ltd.; 2018. p. 75. Available from:

https://doi.org/10.1186/s12951-018-0400-z

14. Rodner M. Towards a versatile gas sensing platform with epitaxial graphene [Internet].

(45)

15. Heidt B, Rogosic R, Bonni S, Passariello-Jansen J, Dimech D, Lowdon JW, et al. The

Liberalization of Microfluidics: Form 2 Benchtop 3D Printing as an Affordable Alternative to Established Manufacturing Methods. Phys status solidi [Internet]. 2020 Jul

8;217(13):1900935. Available from:

https://onlinelibrary.wiley.com/doi/abs/10.1002/pssa.201900935

16. Military T. 3D Printing 3D Printing. Http://WwwGlobalviewGr [Internet].

2020;578(February):1–14. Available from: http://www.globalview.gr/2016/06/30/62949/

17. Coward C. A Beginner’s Guide to 3D Modeling : A Guide to Autodesk Fusion 360. No Starch

Press [Internet]. 2019; Available from:

https://search-ebscohost-com.e.bibl.liu.se/login.aspx?direct=true&AuthType=ip,uid&db=cat00115a&AN=lkp.1078328& lang=sv&site=eds-live&scope=site

18. Lee J, Kr D, Chan H, Kr J, Koh JS. (12) United States Patent. 2016;2(12).

19. Stokes D. Principles and Practice of Variable Pressure / Environmental Scanning Electron

Microscopy (VP-ESEM) : Environmental Scanning Electron Microscopy (VP-ESEM). John Wiley Sons, Inc [Internet]. 2008;30–46. Available from:

https://ebookcentral.proquest.com/lib/linkoping-ebooks/reader.action?docID=406511

20. Herzing AA, Watanabe M, Edwards JK, Conte M, Tang Z-R, Hutchings GJ, et al. Energy

dispersive X-ray spectroscopy of bimetallic nanoparticles in an aberration corrected scanning transmission electron microscopew. Available from: www.rsc.org/faraday_d

21. Molloy D. Exploring BeagleBone : Tools and Techniques for Building with Embedded Linux.

John Wiley Sons; ProQuest Eb Cent [Internet]. 2019; Available from:

https://ebookcentral.proquest.com/lib/linkoping-ebooks/detail.action?docID=5625435

22. Monk S. Programming Arduino: Getting Started with Sketches [Internet]. 2016. Available

from:

https://www-accessengineeringlibrary-com.e.bibl.liu.se/binary/mheaeworks/e0000c8662a3f911/5f8f29b5a0b68756f74dbacc92016 a6e8957e2057abd452b5d4f235fe6df6841/this-is-arduino.pdf

23. Martinez, D., Arhidi, L., Demondion, E., Masson, J. B., Lucas P. Using Insect

Electroantennogram Sensors on Autonomous Robots for Olfactory Searches. 2014;

(46)

8. Appendix

Figure A1. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

periods where the gas flow was directed straight to the sensor

Figure A2. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

(47)

Figure A3. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

periods where the gas flow was directed straight to the sensor

Figure A4. The figure illustrates the recorded resistance over the graphene sensor on the y-axis and the time on the x-axis. The white sections of the plot indicate the time periods where the gas flow was directed through the empty pheromone chamber. The red sections of the plot indicate the time

References

Related documents

Generally, a transition from primary raw materials to recycled materials, along with a change to renewable energy, are the most important actions to reduce greenhouse gas emissions

För att uppskatta den totala effekten av reformerna måste dock hänsyn tas till såväl samt- liga priseffekter som sammansättningseffekter, till följd av ökad försäljningsandel

The increasing availability of data and attention to services has increased the understanding of the contribution of services to innovation and productivity in

Generella styrmedel kan ha varit mindre verksamma än man har trott De generella styrmedlen, till skillnad från de specifika styrmedlen, har kommit att användas i större

Närmare 90 procent av de statliga medlen (intäkter och utgifter) för näringslivets klimatomställning går till generella styrmedel, det vill säga styrmedel som påverkar

På många små orter i gles- och landsbygder, där varken några nya apotek eller försälj- ningsställen för receptfria läkemedel har tillkommit, är nätet av

Detta projekt utvecklar policymixen för strategin Smart industri (Näringsdepartementet, 2016a). En av anledningarna till en stark avgränsning är att analysen bygger på djupa

While firms that receive Almi loans often are extremely small, they have borrowed money with the intent to grow the firm, which should ensure that these firm have growth ambitions even