Novel Planar Nanodevices for Chemical Sensing Applications
Disclaimer: This dissertation has been submitted by a student. This is not an example of the work written by our professional dissertation writers. You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
In recent years, planar electronic nanodevices have attracted much attention due to their simple architecture, ease of fabrication and low cost of manufacture. Such devices address a wide variety of applications in printed and plastic electronics industry. Using this approach a new type of sensor, which is sensitive to different chemicals, has been developed and reported here.
By exploiting the unique characteristics of semiconductor asymmetric nanochannels, a highly selective and sensitive planar nano-transistor based chemical sensor has been realised which can discriminate between wide range of chemical compounds in the ambient atmosphere. The active part of the sensor device was fabricated in a single nanolithography step and was tested using variety of chemicals including polar protic, polar-aprotic and nonpolar solvents. The sensing results showed that, all three solvent categories have exhibited unique chemical signature which could be identified with increased or decreased drain current depending on the analyte used.
A significant rise in transistor drain current was observed when the device was exposed to polar aprotic solvents compared to polar protic and nonpolar ones. Further it has been noticed that the exposure of the device to polar protic solvents which has hydroxyl (–OH) functional groups in their molecular frame work has shown very high hysteresis in current voltage measurements. In contrast, the device has exhibited very little hysteresis when exposed to polar aprotic and non-polar solvents with later being the minimum of all.
The effect of solvent’s polarity on the sensor’s drain current in terms of adsorption and desorption processes has been studied and reported here. Also the effects of water molecules in ambient air and hydroxyl groups on the device hysteresis behaviour have been investigated. As the gas sensing properties of the sensor are related to the chemisorption of gaseous species at its surface, a detailed understanding of the charge transfer in a chemisorption process is very important; hence most of the discussions in this report focus on explaining this complex phenomenon with a special emphasis on the role of surface states during sensing process.
All the measurements were performed at room temperature and the responses were found to be very fast, reversible and reproducible over many cycles of vapour exposure and suggested the stability of the device to be very high. The simple, low-cost, multi-chemical sensing device described in this work could be useful for a variety of applications, such as environmental monitoring, sensing in chemical processing plants, and gas detection for counter-terrorism.
Nanofabrication and Characterisation
Recent advancements in the area of micro/nanofabrication have created a unique opportunity to manufacture nanometer-sized structures with absolute precision that has wide range of applications ranging from electronic, optical, chemical and biological fields. (Springer Handbook of Nanotechnology
Bhushan, Bharat (Ed.) 2nd rev. and extended ed., 2007, XLIV, 1916 p. 1593 illus. in color. With CD-ROM., Hardcover
This chapter will introduce two of such major top-down fabrication techniques namely photolithography and e-beam lithography followed by a brief description on atomic force microscopy and scanning electron microscopes which have been used in this project to fabricate and image the planar nanosensors reported in chapter 5.
In semiconductor processing area-patterning techniques are very important. Lithography is a process of transferring patterns from medium to the other ( ampere a. tseng, kuan chen, chii d. chen, and kung j. ma ieee transactions on electronics packaging manufacturing, electron beam lithography in nanoscale fabrication: recent developmentvol 26, no 2, april 2003 pp 141-149). These transferred patterns are then subjected to a development process that selectively removes either the exposed or unexposed resist depending on the resist nature. The positive resist removes the exposed part where as unexposed resist is developed away using negative resists as shown in the figure 4.1. The exposure systems may be any of these; ultraviolet light rays, X-rays, ion beams or electron beams. But this section focuses on the systems using ultraviolet and electron beams as their source.
Photolithography is the most common patterning method, by which the shape and critical dimensions of a semiconductor device are transformed onto the surface of the wafer (got from lecture notes titled photolithography sly). This is the technique used to define the mesa structures and metallic contacts of the device described in this thesis. A photo sensitive resist is spun on to the substrate and exposed through a mask which transfers the patterns on the sample by means of UV light. Then the sample is developed to get the desired pattern as shown below.
Figure 4.1. Typical photolithography process. The substrate (A) is ¯rst coated by
photoresist (B) and then exposed by UV radiation through a mask (C). The latent image is either removed (D) or ¯xed (E) by a developer solution. Source : M. J. Madou, Fundamentals of microfabrication, 2nd ed., CRC Press (2002), p. 19.
4.2.2 Metal film deposition
In order to perform electrical measurements on the device, we need to define the metal patterns, through which it can be connected to the electrical probe station. So two contacts are formed, ohmic and schottky contacts through a process called lift off as shown in the figure 4.2. The GaAs substrate is coated with photoresist and the patterns are defined by photolithography. First the metal film is thermally evaporated and the unwanted metal laying on the resist is lifted off by dissolving the photoresist in acetone. To facilitate the ‘lift off’ of technique, photoresist edges with undercut profiles are desirable. This can be achieved by the treatment of photoresist with chlorobenzene before the UV exposure. Chlorobenzene swells inside the photoresist and makes its “skin” harder. After the exposure and the development, the profile of the photoresist edges forms an undercut [M. J. Madou, Fundamentals of microfabrication, 2nd ed., CRC Press (2002), p. 19. and M. Hatzakis, B. J. Canavello, and J. M. Shaw, IBM J. Res. Develop. 24, 452 (1980).,], as shown in Fig. 3.3E.
Source fundamentals of micro fabrication book
Figure 4.2. Typical lift-of process. The substrate (A) is coated by photoresist (B) and then prebaked to partially dry the solvents (C). A dip in chlorobenzene follows to make the photoresist skin harder. (D) UV exposure through the mask. The edges of the patterns developed into the photoresist after such process show a typical undercut pro¯le (E). The metal is evaporated onto the sample, forming a thin ¯lm (F). The unwanted metal is then lifted o® by dissolving the remaining photoresist in a solvent (G).
Ohmic contacts (obeys Ohm’s law, linear I-V)
They are essentially formed by a metal layer deposited on a highly-doped semiconductor. Because of the high-doping concentration a very thin Schottky barrier is formed, and the charge carriers, namely electrons and holes, can easily tunnel through. The substrate used in this research work consists of semiconductor heterostructures in which a two-dimensional electron gas (2DEG) was confined between undoped GaAs and doped AlGaAs layers. (R. Williams, Modern GaAs processing methods, Artech House (1990), Chapter 11.)
The choice of metals for any given application will depend on conductivity, thermal stability, adhesion, nature of electrical contact with semiconductor (work function/barrier height), and ease of patterning. (got it grom sly lecture notes) A thin layer (~ 45-50 nm) of Au/Ge/Ni alloy which is the most common scheme for making alloyed ohmic contacts to n-type GaAs is used for this work and was evaporated onto the substrate surface at temperatures higher than 360°C. In this alloy, the germanium diffuses into the GaAs and acts as a dopant, while nickel acts as a wetting layer and also assists the diffusion of Ge into the GaAs.
Schottky contacts (rectifying, diode like I-V)
Depositing a metal film on an undoped, or lightly n-doped, semiconductor whose electronic affinity is lower than the work function of the metal, will form a thick schottky barrier which is typically several hundreds of meV high, and the thermal energy gained by the electrons, about 26 meV at room temperature, is too low to permit thermionic emission over the barrier. (R. Williams, Modern GaAs processing methods, Artech House (1990), Chapter 12). When a bias is applied to the metal, the height of the energy barrier seen by the electrons injected from the metal into the semiconductor does not change, being fixed by the metal work function and the electronic affinity of the semiconductor.
On the other hand, the barrier seen by the electrons injected from the semiconductor into the metal is increased/decreased by a negative/positive bias. This mechanism is responsible for the well-known rectifying effect observed in Schottky junctions [V. L. Rideout, Thin Solid Films 48, 261 (1978).  A. M. Cowley and S. M. Sze, J. Appl. Phys. 36, 3212 (1965).]. At negative biases, the Schottky junction essentially behaves like a capacitor: in substrates with embedded 2DEGs, it can be utilised as a gate electrode to modulate the 2DEG carrier concentration, e.g., for the fabrication of field-effect transistors.
4.2.2 Electron beam lithography (EBL)
One of the modern approaches in dealing with nanoscale structures is e-beam lithography in which, electrons are accelerated by very high voltage, typically of 10s of kV and then focused onto a layer of polymer to create very fine patterns. EBL provides much higher resolution and more precise than photolithography or x-ray lithography: patterns with feature sizes well below 20 nm can be achieved in modern systems. EBL does not require the fabrication of masks as in the photolithographic process. There are two methods to expose e-beam on to the substrate surface (Rainer waser (Ed.) nanoelectronics and information technology, WILEY-VCH chapter 9, pp 234-236) 2005.
Direct writing is the most common EBL approach and used for fabrication of the device reported here. In this approach, a beam of electrons directly impinges on the resist to form the pattern in a serial fashion. As shown in the figure 4.6, a direct writing system consists of a source of electrons, a focusing optics set, a blanker to turn on and off, a deflection system for moving the beam, and a stage for holding the substrate. Where as projection printing is used to project entire pattern simultaneously on to the wafer and can be divided into two ways; SACLPEL (scattering with angular limitation in projection electron beam lithography) and PREVAIL (projection reduction exposure with variable axis immersion lenses). However we will only concentrate on direct writing technique.
Fig 4 dose test patterns of an array of self switching diodes (SSDs) fabricated using e-beam direct write.
Figure 4.1. Simplified structure of a SEM column. The blue lines show the trajectory of the electrons.
4.2.4 E-beam process and proximity efect
To perform electron beam lithography, PMMA (polymethyl methacrilate) resist was used which can be chemically changed under exposure to the electron beam. Final resolution of patterns in the e-beam resist and their eventual transfer into the substrate can be affected due to the imperfections in electron optics, the magnetic environment interaction, the overall thermal stability, the interaction between the beam and the substrate all play an equally important role in determining the ultimate system performance. When the electron beam strike the polymer film or any solid material, it losses energy via elastic and inelastic collisions collectively know as electron scattering. Elastic collisions change the direction of electron scattering, where as inelastic collisions lead to energy loss.
As the electrons penetrate though the resist into the substrate, some of them undergo large angle scattering leading to undesired exposure that form backscattering. This causes additional exposure in the resist and is known as proximity effect. The magnitude of electron scattering depends on the density of the resist and substrate as well as the velocity of the electrons or the accelerating voltage (guozhong cao, nanostructures and nanomaterials, imperial college press, 2004, pp 280-300). (m.a. McCord and m.j.rooks, handbook of microlithoghraphy, micromachininbg and microfabrication, p.rai-choudary, Ed. Bellingham, WA:SPIE Optical engineering, 1997, ch 2, pp 139-249). The proximity effect is more severe in dense patterns, particularly when the separation between adjacent structures is less than 1μm. Since the amount of backscattered electrons depends on the substrate material, a dose calibration is necessary each time different substrates and resist thicknesses are used.
Electron Scattering in Resist and Substrate The scattered electrons also expose the resist!
Electrons, resist and substrates
The smallest thing you can write with the ebeam depends on a large number of factors. These are the spot size used, the type of resist used, the thickness of the resist, the density of the features and the substrate material.
When electrons are used to expose a pattern in resist it is not a simple process. Electrons enter the resist and hit the atoms of the resist, these will either forward scatter or back scatter. Backscattered electrons from the resist will leave the resist and, in general, do not contribute to the resist exposure, forward scattered electrons continue into the resist and contribute to the exposure. The thicker the resist the larger the forward scattering and the lower the resolution. High energy electrons (in our case 100kV) will go through the resist and deep into the substrate. Here they will again get scattered and will forward and backscatter.
In this case the forward scattered electrons will be moving away from the resist and don’t contribute to the exposure, backscattered electrons from the substrate have a large contribution to the exposure. The higher the energy of the incoming electrons the deeper they will penetrate into the resist and hence the contribution to the resist exposure will be reduced. In the figure below you can see that going from 10kV to 20kV increases the penetration depth of the electrons from 1µm to around 6µm. At 100kV the penetration depth in Silicon is around 100µm
Figure schematic diagram of inter proximity effect and intra proximity effect
The smallest feature sizes that can be achieved are when the features are isolated from one another. As you make your features closer together the backscatter from the neighbouring features will all contribute to the exposure and it will become harder to find the correct dose to correctly expose all your features. This is call proximity effects. There are 2 main effects of this; inter-proximity and intra-proximity.
With inter-proximity when two features are close together the electrons from the exposure of on shape contributes to the dose of the neighbouring pattern. The larger and closer the features the worse this effect.
With intra-proximity the dose in the centre of the pattern is larger than at the edges, and especially the corners. This is simply a geometric effect as there are less electrons contributing to the dose in the corners of the shape.
The electrons need a path to ground. If you are using a conducting (or semi-conducting) substrate the contact with the holder is sufficient to provide a conducting path. If you are using an insulating substrate (fused glass, quartz) you will need to provide a conductive path for the electrons. This is normally done by evaporating a metal layer on top of the or underneath the resist. Aluminium or Chrome is are often good choices as they can often be easily be removed without effecting the resist, but you should check the chemical compatibility of your process with the removal procedure.
Performing a Meaningful Dose Test
Exposing a pattern correctly usually requires performing a preliminary test exposure referred to as a dose test. In this test, the pattern is repeated several times on a test substrate. Each repetition is performed at a different dose or set of doses creating a matrix of different exposure conditions. Once the pattern is developed and pattern transfer has been performed the correct dose can be obtained through inspection in a suitable inspection tool (scanning electron microscopy, atomic force microscope, optical microscope, etc). There are several issues which can impact the usefullness of a dose test. Here are some guidelines:
Use the same type of substrate.
If there are films present on the surface of the substrate us a substrate with the identical film stack.
For large arrays of features, shooting the entire array as a test is not an efficient use of time. However, reducing the size of the array to an unrealistically small extent can give incorrent results during the test due to differences in the proximity effect.
· Expose your patterns so that they are easy to locate. For example, do not expose a test pattern consisting of a 500 micron x 500 micron array of 50 nm squares in the middle of a 150 mm wafer. You will probably never find them. Including some locating features (large lines or a box surounding the pattern) can help tremendously. If you are exposing an array of patterns use as small of a repeat vector as possible. This will make locating the entire array easier and minimize the chances of getting lost when travelling in between adjacent elements of the array.
As an electron from the writing beam strikes the surface of a substrate it undergoes various scattering events losing energy and causing the generation of secondary electrons. The energy range of most secondary electrons falls between 1 and 50 eV. Secondary electrons that are close to the substrate/resist interface are actually responsible for the bulk of the actual resist exposure process. While their range in resist is only a few nanometers they create what is known as the proximity effect. Simply put, the proximity effect is the change in feature size of pattern features as a consequence of nonuniform exposure.
While the dose from the primary beam may be uniform across an entire pattern, the contribution of secondary electrons from the substrate may differ depending on pattern geometry. Two adjacent features will contribute a background dose of secondary electrons to each other resulting in a higher effective dose. This causes a broadening of the exposed features. This is particularly apparent with dense features (e.g. gratings). Consequently, dense arrays of features may require significantly less dose from the primary beam to print correctly.
Pattern size can also be adjusted to compensate for this effect. For example, 100 nm lines 100 nm apart are typically drawn in CAD as 90 nm lines 110 nm apart to get them to print correctly. This strategy stops working at the edges and corners of patterns. This sometimes requires the the creation of dummy patterns or devices outside of the primary pattern region to get the main features of interest to print correctly. One common practice is to draw a box around the pattern to normalize the dose in the primary pattern region.
4.3 Imaging nanostructures
Characterisation and manipulation of individual nanostructures requires not only extreme sensitivity and accuracy, but also atomic-level resolution that leads to various microscopes that will play a central role in characterisation and measurements of nanostructured materials (guozhong cao, nanostructures and nanomaterials, imperial college press, 2004, pp 280-300). Nevertheless, when we think of microscopes, we think of optical or electron microscopes that can image an object by focusing electromagnetic radiation, such as photons or electrons, on its surface and gives the image with very high magnifications.
However, the images obtained with these microscopes can only provide the information in the plane horizontal to the surface of the object and do not give any information in vertical dimensions of object’s surface height and depth. This section deals with the imaging of surface topography and surface property measurements of planar sensor using AFM and SEM techniques, which can provide us with all necessary information in both horizontal and vertical planes. (www.afmuniversity.org/pdf/Chapter_1_.pdf pp 1-16)
4.3.1 Atomic force microscopy (AFM)
AFM is a very high-resolution type of microscope from the family of scanning probe microscopy (SPM) with the resolutions thousand times the better than optical diffraction limit (http://en.wikipedia.org/wiki/Atomic_force_microscope). Unlike traditional microscopes, AFM does not rely on electromagnetic radiation to create an image. AFM is a mechanical imaging instrument that measures the three dimensional topography as well as physical properties of a surface with a sharpened probe. ((www.afmuniversity.org/pdf/Chapter_1_.pdf pp 1-16)
AFM Basic principles
It consists of very sharp tip attached to cantilever and is positioned close enough to the surface such that it can interact with the atomic/molecular forces associated with the surface. Then a collimated laser beam focuses onto the cantilever, which scans across the surface such that the forces between the probes remain constant. An image of the surface is then produced by monitoring the precise motion of the probe that can sense the movements as tiny as 0.1 nm. Such high resolution allows to image even single atoms, which are typically 0.5 nm apart in a crystal. Normally the probe is scanned in a raster-like pattern as shown in the figure 4. ((www.afmuniversity.org/pdf/Chapter_1_.pdf pp 1-16)
Source : http://www.afmuniversity.org/index.cgi?CONTENT_ID=33
AFM probe: Cantilever and Tip
AFM is a force sensor with a sharp tip used to probe the surface. When the tip at the end of the cantilever interacts with the surface, the cantilever bends, and consequently beam path also changes, causing the amount of light in the two photo-detector sections to change. Thus, the electronic output of the force sensor is proportional to the force between the tip and the sample. Tips used for probing the surface is usually made of silicon that have a radius of about 10-20 nm and can be coated by silicon nitride to make them harder, or by noble metals, such as gold and platinum, to locally probe electrical quantities or to induce chemical modifications.
Optical detection and Piezo electric scanner
In order to detect the cantilever movements, when the AFM is operating in ambient conditions, optical detection is used. Reflected light from the focused laser beam is collected by a photodiode and the cantilever deflection and torsion are detected as a change in the photocurrents of the photodiode elements, as shown in Fig. 4.
In the typical AFM configuration the tip is kept still, and the imaging is performed by moving the sample with piezoelectric scanner also referred as piezo tube as shown in the figure 4b. By controlling the bias of one inner and four outer electrodes the piezotube can be moved in three dimensions. This photosensitive detector measures the change in optical beam position and the change in cantilever height.
Feedback control is used in AFM for maintaining a ¬xed relationship, or force, between the probe and the surface. According to the mode used, the feedback loop can be controlled either by the cantilever deflection (contact mode) or by the amplitude of the cantilever oscillation (dynamic modes). The typical feedback system used in contact mode is shown in Fig. 3.11. The feedback control operates by measuring the force between the surface and probe, then controlling a piezoelectric ceramic that establishes the relative position of the probe and surface. Feedback control is used in many applications; Figure 2-4 illustrates the use of feedback control in an oven. Section 2.3 has a more
AFM modes: Tip – sample interactions
Depending on separation between tip and the sample a variety of forces can be measured by AFM. At shorter distances van der Waals forces are predominant. Where as these forces become negligible if the tip-sample distance increases. Forces like electrostatic attraction or repulsion, current induced or static magnetic interactions comes into play at these larger separations. The tip-surface forces (approx.) is given by the following equation
Fa = - ΔU = 12 B/Z13 – 6A/Z7 attractive
B and A are coffecients depend upon the surfaces involved.detectable forces for an AFM 1 nN in the contact regime and 1 pN in the noncontact regime (theory 10-18 N)
(r. wiesendanger, “ chapter 11. future sensors.” In h.meixner, r. jones, eds vol 8: micro and nanosensor technology /trends in sensor markets. )
Based on these interactions, AFM usually has two operational modes; contact mode and dynamic mode. Depending on resonant frequency shift of tip-sample, dynamic mode is further divided into tapping mode and non-contact mode. Imaging for this work was carried out in tapping mode.
Also called as repulsive-static mode, in which, the tip rides on the sample in close with the sample surface (low k). The force produced in the feedback loop is frictional force; hence, the tip might interact with the sample surface.
Also called as attractive-dynamic mode, in which the tip hovers 5-15 nm away from the sample surface. The force generated in the feedback loop is typically van der Waals forces. Applied force (dependent on height z) changes the cantilever oscillation frequency.
Figure: AFM Measurement in the figure PSPD represents photosensitive detector.
Also called repulsive-dynamic mode, in which the AFM tip taps the surface as it maps the height z. This type of mode eliminates the hysteresis due to the tip sticking on the sample. Also using this method there is less likely to damage the sample.
Scanning electron microscopy
Scanning electron microscopy is also one of the major techniques for imaging the nanostructures. Although AFM gives high-resolution images with absolute precision, it takes much of time to scan and image the surface area of the sample. Where by SEM can provide an alternative to AFM, which is very fat at imaging the samples in both horizontal and vertical directions.
These schematics show the ray traces for two probe-forming lens focusing conditions: small working distance (left) and large working distance (right). Both conditions have the same condenser lens strength and aperture size. However, as the sample is moved further from the lens, the following occurs:
the working distance S is increased
the demagnification decreases
the spot size increases
the divergence angle alpha is decreased
The decrease in demagnification is obtained when the lens current is decreased, which in turn increases the focal length f of the lens. The resolution of the specimen is decreased with an increased working distance, because the spot size is increased. Conversely, the depth of field is increased with an increased working distance, because the divergence angle is smaller.
Comparison between AFM and SEM
The AFM is more often compared with the electron beam techniques such as the SEM or TEM. With an AFM, if the probe is good, a good image is measured. (www.afmuniversity.org/pdf/Chapter_1_.pdf pp 1-16) the following comparison between AFM and SEM gives a fair idea of the capabilities for applications
A comparison of the some of the major factors follows:
FIGURE 1-8 Both the AFM and SEM measure topography. However, both types of microscopes can measure other surface physical properties. The SEM is good for measuring chemical composition and the AFM is good for measuring mechanical properties of surfaces.
This chapter has covered the main processing and imaging techniques used for fabrication of nanosensor reported in chapter 5. Patterning of metal contacts and mesa structures on to the substrate using photolithography have been discussed in detail. The mechanism for the thin film deposition of Au/Ge/Ni alloy for forming ohmic and schottky contacts have been presented followed by a brief discussion of wet etching for undercut profiles.
e-beam lithography which can overcome the resolution limitation in photolithography has been introduced with a description of its basic elements followed a discussion on proximity effect. So overall, this chapter provides the reader with fundamental knowledge to understand the basic fabrication and characterisation process of which serves as a tool for better understanding the fabrication of planar nanodevices discussed in next chapter (i.e chapter 5).
The evolution of semiconductor industry has brought a revolutionary change in the way we live today. Right from the invention of germanium transistor in 1947 to the latest sensation graphene transistor, the world has seen some of the spectacular breakthroughs that the human kind had ever imagined few decades ago.
In the last fifteen years, more than twelve noble prizes have been awarded for the research based in the field of nanotechnology.
1.1 Sensors and sensor science
Life without sensors and sensing would be like an opera without singer or a violin without strings. Such life does not exist. Sensors and sensing, on the contrary, are basic properties of life that are responsible for the closed loop real time control of what is going on inside and how it reacts to the outside situation. From bacteria to plants and animals to human beings, all living organisms use their sensing organs for orientation and communication, for monitoring the environment and for their survival.
(sensors and sensing in biology and engineering, springer wien newyork, 2003, friedrich g. barth, joseph a.c. Humphrey, timothy w.secomb pp3-34 chapter1 and 2.)
Digital systems however complex and intelligent they are, must receive information from the outside world. Sensors act as an interface between various physical values and electronic circuits that ‘understand’ only a language of moving electrical charges. In other words, sensors are eyes, ears, and noses of silicon chips. Some sensors are relatively simple and some are complex, which operate on fundamental basic principles. Understanding of these devices generally requires an interdisciplinary background in fields such as physics electronics, chemistry etc. Thus, sensors research has brought a unique team of chemists, biologists, physicists & electronic engineers, together on one platform, thus making it a truly interdisciplinary field.
1.1.1 The term ‘Sensor’
In this ever-changing world, sensors are becoming ubiquitous in our daily lives and play an important role in this process. Since the early 1990s, semiconductor industry has seen a tremendous growth in the development of variety of sensors. The technological trends in this field have made electronic products not only smaller and sleeker, but also more interactive and powerful. These sensors with their improving performance–cost ratio will be the key components for the future nanoelectronic devices.(http://www.frost.com/prod/servlet/market-insighttop.pag?docid=140061375)
The word sensor is derived from the Latin word sentire, which means, “to perceive”. A sensor is often defined as a “device that receives and responds to a signal or stimulus” (‘Semiconductor sensors’ S.M.SZE john wiley and sons, inc 1994 chapter 8 pp 383-414)
The purpose of the sensor is to respond to some kind of an input physical property (stimulus) and convert it into an electrical signal which is compatible with electronic circuits. In the other word sensor is a translator of nonelectrical value to electrical value. The sensors output signal may be in the form of voltage, current or charge. These may be further described in terms of amplitude, frequency and phase. Therefore, a sensor has input properties (of any kind) and electrical output properties.
(Jacob Fraden, ‘AIP Handbook of modern sensors’ American institute of physics publishers, 1993 pp 1-16. chapter 1 and 2)
1.1.2 Sensors Characterisation
Depending on type of sensing (eg. Chemical, biological etc) the requirement for sensor performance factor varies. For example, detection of trace explosive requires extremely high sensitivity and low limits of detection (LODs) because of the relatively small number of molecules that can be collected as a result of their very low vapour pressures. High selectivity is essential in order to have an acceptable rate of false positives. In general, any sensor should be readily reversible at room temperature to facilitate continuous operation. These sensors should also have fast detection and regeneration times for efficient operation. Usually the performance of any sensor can be judged using the below mentioned parameters…
Table: Definition of performance parameters for sensors
Sensitivity is defined as change in sensor response curve to the change in analyte concentration.
It is the ability of a sensor, to respond primarily to only one desired species in the presence of others species.
The ability of a sensor to reproduce the output readings when the same analyte* is applied for different cycles of vapour exposure.
The amount of time a sensor takes to recognise the analyte when the analyte flow is turned on..
The maximum difference in sensor characteristics curve for increasing and decreasing analyte concentration.
It is the measure of a sensor that can detect even the smallest variation, when the analyte concentration is changing continuously.
Limit of detection (LOD)
Lowest analyte concentration value that can be detected.
The analyte concentration from LOD to maximum concentration that can be reliably detected.
The range where the sensor response is in direct proportion to the analyte concentration.
*The term analyte is used in analytical chemistry to define the species being analysed in a sample. An analyte can be an atom, ion or molecule in a solution, a solid or in the gas phase.
A sensor satisfying all the above performance parameters is called a true sensor, which is highly impractical due to various limitations.
1.1.3 Types of sensor systems
Classification of sensors schemes range from very simple to the complex, because of their involvement in various fields ranging from biology to astronomy. Nevertheless, depending on their operating principle, we can divide sensors into three general categories, namely physical, chemical and biological sensors.
Physical sensor is device that provides the information about a physical property of the system like, measuring the distance, mass, temperature, pressure, magnetic field and force that do not have any chemical interface.
The IUPAC (International Union of Pure and Applied Chemistry) definition of biosensor is defined as a “device that uses specific biochemical reactions mediated by isolated enzymes, immunosystems, tissues, organelles or whole cells to detect chemical compounds usually by electrical, thermal or optical signals.”
Chemical sensors rely on a particular chemical reaction for their response. It is this group, which is discussed in this thesis. The other two types were not focused here.
1.1.4 What is a chemical sensor? Why do we need chemical sensors?
According to the R.W. Cattrall’s book definition, “A chemical sensor is a device that responds to a particular analyte in a selective way through a chemical reaction and can be used for the qualitative or quantitative determination of analyte” (chemical sensors, Robert w.cattrall, oxford chemistry primers, 1997, chapter 1, pp-1-3)
There is escalating need to monitor our environment in real time due to the increasing concerns of pollution, our health and safety. Sensing of volatile organic compounds (VOCs) is an important issue. Because these volatile compounds evaporate rapidly, their vapours are hazardous to human health. Chemical sensors allow us to monitor and detect all these abnormalities in the surrounding environment without human presence. These sensors can provide real-time detection to reduce the impact of threats and can be deployed remotely so that direct contact with toxic agents can be avoided; they also cover a large area, to increase detection efficiency. They can detect which substances are present and in what quantity.
Chemical sensors work on the principle of adsorption and desorption mechanisms, which will be discussed in detail in chapter 2. During the chemical sensing mechanism, the analyte molecules form a physical/chemical bond with the sensing substance. As a result, charge transfer occurs between the molecule and the substance that changes the properties of adsorbent. This change could involve a change in conductivity, change in resistivity or capacitance.
Characteristics of chemical sensors
The field of chemical sensor research has brought together a unique team of chemists, physicists, biologists, and electrical engineers onto one platform and thus is a truly interdisciplinary field. In the field of analytical chemistry, chemical sensors has been adapted and now largely considered as a significant sub-discipline. (Chemical sensors: an introduction for scientists and engineers By Peter Gründler, chapter 1, pp 1-13, springer) Where as in physics, sensing of chemical compounds has always been an active area of research.
There are two obvious sources for the formation of sensor science as an independent ¬eld. One of these sources is development of microtechnologies, which stimulated a demand for sensing organs. The second source is a consequence of the evolution of analytical chemistry, which brought about a growing need for mobile analyses and their instrumentation. Below figure showing, two sources in the development of chemical sensors.
Figure 1.2: attempts to outline the formation of sensor science as a bona ¬de branch of science.
As a science, chemistry from the very beginning required information about chemical composition. In other words, analytical chemistry comprises one of the earliest foundations of chemical science; it is as old as general chemistry. (Chemical sensors: an introduction for scientists and engineers By Peter Gründler, chapter 1, pp 1-13, springer) Where as in physics, sensing of chemical compounds has always been an active area of research.
Although chemical sensor comprises of many fields they must have the following characteristics. A chemical sensor should:
• Transform chemical quantities into electrical signals,
• Respond rapidly,
• Maintain their activity over a long period,
• Be small,
• Be cheap,
• Be speci¬c, i.e. they should respond exclusively to one analyte, or at least be selective to a group of analytes. The above list could be extended with, e.g., low detection limit, or a high sensitivity. This means that even the low concentration values should be detected.
Because of the desire to monitor everything around us, there is a high input of resources in developing sensors for multiple applications. The fruitful result of all this research would one day provide us with the powerful, portable and smart sensor device that could monitor anything we wish.
The anticipated applications of chemical sensors are in, health care industry, homeland security and defence, information technology (IT) industry, environment etc. The micro-chemical sensors that have in use until now did not live up to their expectations that cannot completely satisfy the demands. Only by means of system approaches, satisfactory results can be achieved. Thus, a novel approach is needed to develop/enhance the performance of these sensors for realising the above mentioned practical applications. Sensors based on nanoscience, however, offer a clear path to the development of nano-sensors that satisfy these criteria.
1.2 Nanosensor Technology: Innovation for tomorrow’s world
The advent of nanotechnology has revolutionised this world so much that the word ‘Nano’ has become one of the buzzwords across the globe, as the research related to it, is slowly moving towards commercialisation across many fields. Since, nanotechnology comprises of many areas from physics to electronics, chemistry to material science, and biology to medicine, it is very difficult to precisely define the term nanotechnology. However, in general, it can be defined as the manipulation of the matter between the scales of 1-100 nm. Here, one nanometer is 0.000000001 meters and about the size of three atoms. At this scale, the properties that appear on the macro scale may not necessarily apply at nano-level.
Since the sensing measurements of sensors have to deal at atomic and molecular scale interactions, they have obvious sensing applications in the field of nanotechnology. Sensors constructed at the molecular scale would be extremely sensitive, selective, and responsive and thus, the impact of nanotechnology on sensors research is huge. Hence, the sensors fabricated using nanotechnology are called as nanosensors.
1.2.1 Classification of Nanosensors
Nanoscience deals with new physical or chemical properties of matter at the nanoscale, and new sensor devices being built on nanoscale take advantage of these phenomena. Depending on the important characteristics like use of sensing elements and area of applications nanosensors can be classified into six different types as shown in the figure
Figure: Nanosensors classification depending on its application
1.3.2 Nanochemical sensors
Wolfgang Goepel at Montana state university has first introduced the concept of “nanochemical sensor” in early nineties. Since then they have grown strength to strength in moving towards more practical applications. In fact, in the last 10 years almost 2000 papers referring to “chemical nanosensors” have been published and the field seems to be one of the most immediate and promising sectors for the application of nanotechnologies . This rapid growth can be attributed largely to recent advances in nanotechnologies that enabled the synthesis and engineering of materials to realize devices that exhibit functionalities specifically originated by their nanostate .
3. Chemical AND nano∗ AND sensor∗, 1998–2007, http://apps.isiknowledge.com/.
4. Nanotechnology: Basic Information, http://es.epa.gov/ncer/nano/questions/index.html.
A nano-chemical sensor can be defined as an electronic device, consisting of a substrate and a sensing element and relies on at least one of the physical and chemical properties typical at the nanoscale for its operation. It operates as any other chemical sensor: charge transfer occurs between molecules and an “active” material, resulting in an electrical signal depending on the molecule type and polarity. However, unlike macroscopic sensors, nanochemical sensors can take advantage of the merging of three different features:
() the quantum confinement,
() the surface-to-volume ratio, S/V, and
() the nanoparticle morphology, (the word nanoparticle, is here used to describe, in general, any kind of structure with at least one of its dimension in the nanorange. Sometimes expressions such as nanowire, nanodot, or nanotube will be used, when a more precise reference to the morphology is required).
These properties improve the sensitive material behaviour of room temperature operating devices that can provide sensitivities down to the single molecule level. (Conductometric Gas NanosensorsGirolamo Di Francia, Brigida Alfano, and Vera La Ferrara Journal of Sensors Volume 2009, 20th april 2009 pp 1-18)
There are various approaches for sensing chemicals at nano-level. They are, Receptor-based detection, Receptor-free detection, Nanowire and Nanotube platform, Electronic nose approach, Nanomechanical sensor platform. Among all five approaches, chemical sensing based on ‘electronic nose approach’ is of special interest due to its high potential commercial applications and is discussed here.
Electronic nose approach
Of all five human senses, sensing of smell is perhaps the most interesting part for sensor industry. Nose is an extremely sensitive and selective organ that can easily distinguish many different odours and can give a fair idea of ‘quantity’ detected down to very low detection limits. Replicating such kind of ultra sensitive system artificially is very challenging and could be rather complex. Therefore, any experiments regarding this artificial approach, first needs to understand the basic sensing mechanism of the ‘human nose’.
In human nasal olfactory system, nostrils collect the odour molecules that are sensed by the olfactory membrane and generate an electrical response in the primary nerve cell, which is then passed to the secondary nerve system as shown in the figure. The secondary nerve sends this electrical signal to brain for interpretation. Thus, the brain acts as central processing unit, turning the signal into a sensation which we call smell.
Although, our nose can distinguish between hundreds of smells, it fails to detect the absolute gas concentrations or odourless gases. Therefore, there is demand for gas sensing devices that can support human nose for many safety applications to protect our environment. So, a technique similar to human nose approach has been used for the development of an ‘artificial’ nose called electronic nose or simply e-nose.
Figure: a) diagram-showing mammalian olfaction system as a sensor, in which the olfactory membrane is the biological recognition element; primary, secondary nerve cells are sensing and data processing elements, and the brain is the pattern recognition element. b) Schematic of a sensor showing the component parts, i.e. analyte, sensing element on the substrate and computer generated data and pattern recognition systems.
Since the mid 1980s electronic-noses has been gaining increasing interest for detecting and recognising complex odours. “Electronic nose can be defined as an artificial vapour analyzing system, which comprises an array of electronic chemical sensors that provide a recognizable image of specific vapour mixtures (fragrances) containing possibly hundreds of different chemical species”. Just like a ‘human nose sensing mechanism’, the information in the e-nose concept is also transmitted and processed in electrical form.
However, this transmission is through the transport of electrons. Here, the sample and the sensing element act as an olfactory membrane, that detects the chemical species and transmits this chemical signal to ‘conditioning and data-pre-processing’ unit, which converts the received chemical signal to electrical signal. This electrical signal is then transferred to the pattern recognition system, where it transforms the electrical signal to a recognisable pattern that we can perceive depending on the type of sensing element used. In this work, a new type of e-nose based on planar electronics has been introduced and was reported in chapter 5, which can potentially become a strong contender for future simple and inexpensive electronic-noses.
1.3.3 Global nanosensor market trends and forecasts
As nanosensors related applications, are attracting the interest of research groups around the world, governments and research organisation in many countries have now realised the potentials of it and are investing millions of dollars in nanosensor related research. As a result, sensor market has taken a rapid phase of development in the last few years. The global sensor market was valued at US$ 32.5 billion in 1998 and reached US$ 42.2 billion in 2003. By 2008, it has crossed US$ 50 billion and is expected to reach US$61.4 billion by 2010. Western Europe itself constitutes about 32 per cent of the total predictions, which is the largest geographical market sector in the world. The UK market is expected to be the second largest in Europe, after Germany.
Major growth will occur in many applications, product sectors and technologies like, automotive and environmental sensors; gas, chemical and biosensors. (Title: Global sensor markets, Journal: Sensor Review, Year: Jun 2005, Volume: 25, Issue: 2) This indicates that there is an exciting period of growth for semiconductor sensors, which are being integrated into an increasingly diverse array of applications in consumer electronic products and this trend is expected to continue for the years to come.. (http://www.bccresearch.com/report/NAN035A.html December 2004, Analyst: Andrew McWilliams)
The market estimate for nanosensors by BCC (Business communications company) report provides the information on existing and future nanosensor technologies and applications, and assesses their commercial potential. BCC Inc has reported that, in 2003, the nanosensors market mainly consisted of nanochemical sensors (chiefly ultrasensitive gas sensors), nanobiosensors (nano-LC systems). In 2004, total global nanosensor sales was at $190 million and are expected to rise at (AAGR) of 25.5% to $592 million by 2009. Brief market surveys on some of the nanosensors are reported below.
Figure: Graphical chart showing the market for nanosensors from the year 2003 to 2009 that shows, chemical and biosensors have seen tremendous growth in last 2-3 years when compared to nano-thermal, motion and radiation sensors.
According to them, nano-chemical and nano-biosensors will grow significantly through 2009, at an average annual growth rate of 53.1% and 32.9%, respectively where as nano-motion and nano-radiation sensors are not expected to account for any large portion of the market, while nano-thermal sensors are not projected to achieve any commercial sales during the period. (http://www.bccresearch.com/report/NAN035A.html December 2004, Analyst: Andrew McWilliams)
This chapter has covered the general background about the sensors. A brief introduction to sensors has been presented first by defining and characterising them, followed by a brief explanation on chemical sensors and their role in sensing various harmful species in the environment. The impact of nanotechnology on sensors with a comprehensive overview of the main concepts behind the development of nanosensors has also been presented in the last section, which describes about various classes of nanosensors with a special emphasis on nanochemical sensors.
Various approaches in chemical sensing has been presented with a focus on electronic nose concept. Stastical survey by business communication company Inc on the global market for nanosensors was discussed in the last part of the second section. Thus, the first chapter has covered all necessary topics on principles of sensors which will be helpful in dealing with the chapters 3 and 5.
The main aim of this project was to design and fabricate a chemical sensor that can detect various gas species in ambient conditions with high precision. Both the theoretical and experimental parts of the project were presented in the report, which is organised into six different chapters. The report begins with an introductory chapter that describes about the general idea of sensors and their importance in our day-to-day life, principles of chemical sensors followed by nanosensors and their classification. Chapter 2 and 3 discusses about the physical chemistry of solid surfaces and the surface chemistry of III-V semiconductors and their importance in chemical sensing respectively.
Chapter 4 then gives a brief understanding of the nanofabrication and characterisation techniques used for this project. Chapter 5, is the most vital part of the report. The first half of the chapter is an experimental part that discusses about planar nanosensor design, fabrication and electrical measurements followed by the sensing results of the sensor to different chemicals. Where as, second part of the chapter is an explanation part that has the detailed discussions on the hysteresis, which revolves around the effect of hydroxyl groups in the sensor responses. Finally, report ends with a chapter on conclusion and outlook of the device described here.
3. Surface Chemistry of III−V Semiconductors and Implications for Chemical Sensing
In the last chapter, we have studied the basic sorption process on solid surfaces (metals and non-metals). This chapter deals with the sorption process on semiconductor surfaces. But, before going into to the into the chemistry of semiconductor surfaces ,we shall first see the basic structure of semiconductors, especially GaAs/AlGaAs that has bee used for this work and then move on to the physical chemistry of semiconductors for chemical sensing. First section will focus on the general properties of heterostructures with semiconductors composed of more than one material and would cover the basic growth mechanisms of those heterostructures. Second section introduces the GaAs/AlGaAs based system, which provides a two-dimensional electron gas channel that has been widely used in mesoscopic experiments. Later part of the section mainly focuses on the adsorption and desorption process on III-V Semiconductor surfaces.
3.2 Physics of heterostructures
Heterostructures are the basic building blocks of the most of the advanced semiconductor devices used today. III-V compound semiconductor heterostructures are widely used for high frequency electronic and optical applications. One of the main advantage in using heterostructures is that they offer precise control over the states and motions of charge carriers in semiconductors.( http://www.utdallas.edu/~frensley/technical/hetphys/node2.html#SECTION00020000000000000000) a heterostructure can be defined as an semiconductor structure in which the chemical composition changes with position. I.e. the structure of semiconductor changes when two dissimilar semiconductors are grown together on top of the other. The abrupt changes taking place in the conduction and valance band energies on going from one semiconductor to the other is highly important in determining the electronic and optical properties of hetetojunction. Figure 2 shows the values of band gap vs lattice constant for common III-V semiconductors.
Figure: the band gap values for common III-V semiconductors against their lattice constants. As can be seen, the materials span a range from 0.17 eV (InSb) to 2.45 eV (AlP. At the high band gap region (> 1eV), the binary compounds become indirect and is indicated by dotted lines in the diagram above.
Figure is from lecture notes semiconductor device physics prof missous
While growing the heterostructures, the epitaxy of materials needs to have equal lattice constants of the two (or more) constituent semiconductors. Unfortunately, there are only few pairs of semiconductors that have equal/close lattice constants.. Nevertheless, materials can be grown even with variation in lattice constants, provided the difference in lattice constant value should be in acceptable level called critical thickness. For example, when the two different materials are brought in contact with each other, due to the variation in their lattice constants, the atoms at the interface of the substrate adjusts to match with the lattice constant of the other material. This causes the epilayer to be under strain. Further, if we increase the strain energy the thickness of the strained layer also increases. Therefore, one can only grow a certain critical thickness of epilayer before dislocations are generated and the structures formed by such strained layer are called pseudomorphic structures.
The critical thickness can be calculated by using the below formula.  J. Singh, Electronic and Optoelectronic Properties of Semiconductor Structure, Cambridge University Press (2003).  J. H. van Der Merwe, J. Appl. Phys. 34, 117 (1963).
 J. H. van Der Merwe, J. Appl. Phys. 34, 123 (1963).
 H. Nobuhara et al., IEEE Phot. Tech. Lett. 5, 961 (1993).
 C. S. Whelan et al., Microwave Journal, Euro-Global Edition 44, 110 (2001).
Heterojunction band line-ups
The term heterojunction is usually referred to the interfaces in which, traps play a negligible role. The heterojunction band line up can be any of the three variants shown in the Figure below
As shown in the figure, type 1 heterostructures. When the different band gap materials are sandwiched, the one with narrow gap semiconductors falls within the larger band gap semiconductors. Many systems such as GaAs-AlGaAs, InGaAs-AlInAs etc come under this type. Where as in type 2 hetero-interfaces forms the basis of a new class of optical devices based on AlGaAs interfaces (or even AlAs). In this case the two band gaps are staggered with respect to each other. In the broken gap case, or type 3, the conduction band edge of InAs falls below the valence band edge of GaSb (by an amount of ~ 60 meV)  J. H. Davies, The physics of low-dimensional semiconductors: an introduction, Cambridge University Press (1998).  L. De Caro et al., Solid State Comm. 108, 599 (1998).  J. C. Mikkelsen Jr., and J. B. Boyce, Phys. Rev. Lett. 49, 1412 (1982).
Figure: a) Pseudomorphic layer showing the interface between the two materials with acceptable critical thickness. b) Figure showing the latticed matched heterostructure without any dislocations.
3.2.1 Heterostructures growth
Heterostructures are mainly fabricated using epitaxial techniques. During these structures growth, it is very important to match the crystal lattice parameters of the material being grown with that of the base material. The most widely spread epitaxial methods for growing semiconductor heterostructures are MBE (Molecular Beam Epitaxy) and MOCVD (Metal Organic Chemical Vapour Deposition). Ever since it had evolved, MBE has become a popular technique for growing majority of III-V semiconductors a well as metals and insulators. It can produce excellent high quality layers with good control of thickness, doping and composition. (the physics of low-dimensional semicoductors, john h. davies, Cambridge university press, 1998, chapter 3, pp 80-90.)
Figure: Simplified schematic arrangement showing the sources and substrate in a conventional MBE system to form GaAS and related compounds by molecular evaporation on to the substrate.
In MBE, the elements that compose of heterostructures, such as, Ga and As are deposited onto the heating substrate to form epitaxial layers under high vacuum pressure (5X106) as shown in the figure. At this low pressure, the mean free path of molecules between collisions is much longer than the width of the chamber. This is known as the Knudsen regime, and the crucibles are known as Knudsen cells. The growth begins when shutters open, and the flux rates are controlled by the temperature of the crucibles.
During the growth process, the substrate temperature is carefully monitored. If the temperature is too low, the crystal will not gain enough thermal energy to relax defects; where as, if the temperature is too high, the elements will tend to diffuse, reducing the sharpness of interfaces. In order to minimise variations of the layer thicknesses along the entire substrate, the substrate holder normally rotates during the growth. MBE uses an evaporation method for molecular impingement rate, i.e. the rate of molecules impinging on a unit area of the substrate per unit time. The advantages of using MBE is, high degree of control and reproducibility. Although MOCVD is also more sophisticated technique for growing heterostructures, handling the poisonous gases and cost of maintenance limits them using for practical applications.
3.3 Two-dimensional electron gas (2DEG)
A thin layer of two-dimensional conducting layer called two-dimensional electron gas (2-DEG in short) forms between two different semiconductor material band-gaps (e.g. GaAs and AlGaAs materials), which is the most important low-dimensional system for electronic transport. To understand why this 2DEG isformed, we need to consider the valance band and conduction band structures in Z-direction when we first bring these layers into contact (see the fig).
The Fermi energy Ef of wider-band-gap AlGaAs (n-doped) layer will be higher than the lower-band-gap GaAs (undoped) layer. Therefore, the electrons from the n-doped AlGaAs spill out of it and fall into the GaAs layer to occupy lower energy states leaving positively charged spaces. Because of this space charge, an electrostatic potential is generated causing the bands to bend as shown in the figure. At equilibrium condition, the electron density is sharp near the GaAs-AlGaAs interface-giving rise to a thin ”triangular” shape quantum well conducting layer called two-dimensional electron gas that confines the electrons in the perpendicular direction to the layers. (electronic transport in mesoscopic systems, supriyo data, Cambridge university press, 2007, chapter 1, pp 1-7). Hence, a 2DEG can be defined as a gas of electrons free to move in two dimensions, but tightly confined in the third dimension. This tight confinement leads to quantised energy levels for motion in that direction, which can be ignored in most cases. Thus, the electron appears to be two-dimensional sheet embedded in a 3D world.
The carrier concentration in a 2DEG typically ranges from 2x1011/cm2 to 2x1012/cm2 and can be depleted by applying a negative gate voltage. This kind of structure is used to fabricate field effect transistors like modulation doped field effect transistor (MODFET) or high electron mobility transistor (HEMT) for terahertz based electronic devices (Drummond t.j., masselink w.t and morkoc h, 1986, ‘modulation-doped GaAs/(Al,Ga)As heterojunction field effect transistors: MODFET’s’ proc, IEEE, 74, 773.
(Melloch M R, 1993, molecular beam epitaxy for high electron mobility Modulation doped two-dimensional gases’. Thin solid films, 231, 74.) and is also very important to study the fractional and quantum hall effects.
The extremely low scattering rates make the 2DEG in GaAs very special with ulrtra high mobilities around 1,000,000 cm2Vs’ at low temperatures. In fact, with the specially grown state-of-the-art heterostructures, mobilities’ of 30,000,000
Cite This Dissertation
To export a reference to this article please select a referencing stye below: