Project topics and materials
Get Complete Project Material File(s) Now! »
The theory behind scanning tunneling microscopy
Essentially, the starting point of SPM was the invention of STM in 1981 by Gerd Binnig and Heinrich Rohrer. This technique was firstly utilized by Binning, Rohrer, Gerber, and Weibel.[173], [174] Since then, it has led to the creation of various new probe microscopes (e.g. AFM) and has been a widely used tool in surface science. This is to be expected, as these instruments are able to provide a topography with atomic resolution,[175] but also information on the local electronic structure.[176] The operation mechanism is based on the quantum mechanical tunnelling effect. It makes use of a sharp metallic tip that is positioned just a few angstroms above a conductive surface whilst applying a voltage. Here, the potential barrier between the tip and the surface is larger than the available electron energy so that no current can flow based on classical mechanics. However, quantum mechanically, the overlap of the electron wave functions of the sample and the STM tip allows a tunnelling current to occur. For a better comprehension of this phenomenon, in the following a brief overview of the main concepts is given. The simplest form of the one-dimensional potential barrier scenario (illustrated in the top part of the Figure 2.2) is the most convenient way to understand the concept of quantum mechanical tunneling.[177] Here, an electron having some energy E is moving in the potential U(z) and can be described by the wavefunction Ψ(z) following the time-independent Schrödinger’s equation in one dimension[178]: Where m is the electron mass, ħ is the reduced Planck’s constant and U(z) is the height of the potential barrier. Now, since the regions I and III are outside the potential barrier, where the potential U = 0, by the scattering theory the solution for Eq.2.1 will be
The lock-in amplifier
There are many potential noise sources in the STM systems, such as mechanical, thermal, amplifier induced and most commonly, flicker noise (1/f).[185] The last mentioned is present in most of the electronic devices and it results from slow resistance fluctuations caused by temperature variations in the system. In order to increase signal to noise ratio, but also obtain dI/dV curves while avoiding numerical differentiation, lock-in is used. The basic idea is to measure the derivative of the I-Vs, rather than compute it and in such a way obtain better signal-to-noise ratio. How this is accomplished is schematically presented in Figure 2.4. The STM tip is held stable at the set distance d from the sample.
The bias voltage is ramped within a chosen range while the tunneling current is recorded. During the measurements, a small high-frequency (around 500Hz-10kHz) sinusoidal signal, Vmod sin(ωt) is superimposed to the tunneling voltage. As a result, the tunneling current will oscillate at the same frequency. Next up, the tunneling current, along with the noise is converted to a voltage by transimpedance amplifier. Multiplying it with the reference signal will give a response at the sum and at the difference frequencies. Hence, only the signal oscillating at the same frequency as reference one will give a sinusoidal output, whereas everything else will be oscillating at the average value of zero. By tuning the phase with a phase shifter, one can effectively filter out only useful quantity of the signal by having it in the mixing product, which will be pushed to the dc value, while the noise will still have the ac character. amplifier. Using the low-pass filter, only the dc component is extracted from which the amplitude of the modulated current (sensitive to the slope of the I-V curve) can be measured. This is significant, since for small modulations, the tunneling current can be expanded in the Taylor series:
STM and STS on the Omicron LT-STM
The single-probe low-temperature STM system (LT-STM) developed by Omicron GmbH, that was used during this thesis is an instrument of the Physics group at IEMN and is located at the Insitut Supérieur d’Électronique et du Numérique (ISEN) in Lille, France.
The whole system is kept in ultra-high vacuum (UHV) and comprises of two main chambers, analysis and preparation ones, as it can be seen in Figure 2.5. To reach and maintain the UHV conditions ( ̴ 10- 11 Torr), a set of different pumping tools is required. Aside from mechanical pumping this system relies on ion getter pumps and titanium sublimation pumps (TSP). Both chambers can be pumped separately since they are isolated from each other with a gate valve. In fact, the easiest way to describe the system layout and operation, whilst staying in the context of this thesis, is to follow the path of some archetypal colloidal nanocrystal (NC) and selective area grown nanowire (NW) samples from their arrival to their first STM image and STS curve. Before the measurements, both samples are protected from atmospheric conditions in their own respective ways. The sample NC is a suspension of ligand-passivated nanocrystals and the NW sample is aone-centimeter squared piece of wafer with nanowires capped with a thin layer of arsenic. The starting point of any STM measurements, is to have a high-quality probe, which can be reliably done by electro-chemical etching of a tungsten wire. In practice, the wire is dipped below the surface of a sodium hydroxide (NaOH) solution and the voltage is applied between it and the cathode. After 10-20 minutes the lower part of the wire falls off, while the leftover part is sharp enough to be used as the STM tip. Both NW and NC samples are stored and transported in a way to prevent oxidation and contamination. If not stored inside the UHV systems, semiconductor nanowire samples are usually held in highly portable vacuum containers (Figure 2.5a), while air-tight vials containing liquid suspensions of colloidal nanocrystals are preserved in a cold and dark environment. The next step is to prepare the samples. The NW sample will be cut to size, whereas for NC sample the suspension will be drop-casted on chosen substrates and dried with nitrogen flux. Subsequently, both samples will be mounted on the sample holders. Depending on the type of the sample and the measurements needed to be done, different configurations of the sample holders can be used (simple plate, bridged, spring loaded, angled etc.). Once mounted the following step is their introduction to the UHV conditions without breaking the vacuum of the whole LT-STM system. This is done through the load-lock compartment located on the side of the preparation chamber (layout in the inset of Figure 2.5). Both samples are placed on a short transfer rod and the compartment is sealed shut by screwing it tightly with a copper gasket at the joint. The follow-up bake-out procedure is done at 120 °C for several hours to remove any residual water from the inner chamber surfaces, but also the samples. After that, the valve between the load-lock and the preparation chamber can be opened and the samples can be safely transferred from one to the other. The preparation chamber contains all the necessary tools for the sample cleaning and other treatments, such as five-slotted storage bay, quadrupole mass analyzer (QMS), sputtering gun, evaporation stage, heating elements, thermocouple, and various electrical connections which allow for the sample polarization or for its isolated heating. All the maneuvering of the samples in the preparation chamber is done either with the transfer rod or with a manipulator with four degrees of freedom.
Locating and identifying the nanowires on the surface.
As mentioned before STM is an instrument designed for the samples with the surface roughness reaching no more than couple of dozens of nanometers. In that regard, most of the epitaxially grown nanowires are exceptionally large objects, by STM standards. Therefore, when acquiring STM images special care must be taken for inevitable and sudden height variations which could cause a tip to crash. The usual protocol which is used for these kinds of samples involves first finding and then, climbing the nanowire. Finding the nanowires in the absence of SEM, entails for the initial scanning to be done on a large scale with an exceptionally low speed, setpoint current, resolution, and with a fast feedback loop. Once a sudden change in height is detected, the scan is stopped and rotated in a direction parallel to the nanowire. Climbing the nanowire is done mostly with the same parameters with the only optional improvement being the resolution. When the tip has reached the top of the nanowire, the scanning parameters can be changed to favor the quality of the image (e.g. lower bias voltage, higher setpoint current, stiffer feedback loop, smaller scan area, increased number of pixels, etc.). From the noise levels and drift traces in the STM images one can know when the tip has reached the adequate level of thermal and mechanical stability (see Figure 2.7). Only then, the scanning tunneling spectroscopy can be performed. However, for selective area grown anowire samples, only second half of this protocol is applicable. Because the distribution of the nanowires will be dictated by the initial patterning of the mask, finding a nanowire should be in theory, simpler. However, this is not true since the mask itself presents another more severe problem. Due to its non-conductive nature it renders the possibility of having a successful blind automatic approach next to impossible. In fact, this is more an instrumental challenge, rather than the one for the operator.
Table of contents :
Abstract
Acknowledgements
Contents
Introduction
Chapter 1. State of the art
1.1 Semiconductors
1.2 Quantum structures
1.3 The appeal of one-dimensionality in nanotechnology
1.4 A broad definition of nanowires
1.5 What are the trends?
1.6 Heterointerfaces with the emphases on the 1D nanostructures
1.7 Planar 1D heterostructures by means of selective-area epitaxy
1.8 Quantum structures by means of colloidal synthesis
1.9 Take-away points and ensuing challenges
Chapter 2. Characterization techniques
2.1 Introduction to scanning probe microscopy
2.2 The theory behind scanning tunneling microscopy
2.3 The theory behind scanning tunneling spectroscopy
2.3.1 The lock-in amplifier
2.4 STM and STS on the Omicron LT-STM
2.4.1 STM and STS on selective area grown nanowires
2.4.1.1 Locating and identifying the nanowires on the surface.
2.4.1.2 Performing STS measurements on semiconductor nanowires.
2.4.2.2 STM and STS on colloidal nanocrystals
2.4.2.2 Performing STS measurements on colloidal nanocrystals.
2.5 The Omicron VT four-probe STM
2.6 Complementary techniques
2.6.1 Atomic force microscopy
2.6.2 Optical spectroscopy
Chapter 3. Morphology and electronic structure of SAG InGaAs nanowires
3.1 Introduction
3.2 Sample preparation
3.3 Nanowire morphology
3.3.1 SEM measurements
3.3.2 AFM measurements
3.3.3 STM measurements
3.3.4 Diffusion kinetics model
3.4 Electronic properties of heterostructures
3.4.1 STS measurements of the In0.53Ga0.47As cross-shaped nano-heterostructure
3.4.1.1 In0.53Ga0.47As / InP band alignment from STS
3.4.2 Charge transport in the In0.53Ga0.47As cross-shaped nano-heterostructure
3.4.2.1 Two-probe configuration
3.4.2.2 Four-probe configuration
3.5Conclusions
Chapter 4. Quantum confinement and band alignment in CdSe NPL-based heterostructures
4.1Introduction
4.2 Growth of CdSe NPLs
4.3 Bare cadmium-selenide nanoplatelets
4.3.1 STM investigation of the structure morphology
4.3.1.1 LT-STM on 5.5 ML-thick NPLs
4.3.1.2 LT-STM on 7.5 ML-thick NPLs
4.3.2 STS study of the electronic properties
4.3.2.1 LT-STS on individual flat-lying NPLs
4.3.2.2 LT-STS on stacked up NPLs
4.4 Extensive “pre-STM” study of CdSe/CdS Core/crown NPLs
4.4.1 Influence of the crown size on optical properties
4.4.2 CdSe exciton fine structure
4.4.3 Recombination dynamics of the exciton complex
4.4.4 Optical properties of CdSe/CdS core/crown NPLs in high magnetic field
4.5 Conclusions
Conclusion and future work
Bibliography
