Defended Dissertations in 2016

Irina Černiukė
Author: Irina Černiukė
Dissertation title: Fabrication and investigation of the heterostructures based on manganites and organic semiconductors
Fields of science: Technological Sciences, Materials Engineering (08T)
Scientific supervisor: Prof. dr. Bonifacas Vengalis
Defence of the dissertation: 2016-02-29
Great attention is currently focussed on the development of new fields of electronics such as spin-dependent electronics (spintronics) and molecular electronics. Spintronics is a new rapidly developing research area exploiting spin-dependent electronic transport in various device structures based on magnetic materials while incorporation of hybrid organic - inorganic structures and even single molecules into electronics circuits is the main goal of molecular electronics.
Of key importance for the development of spintronics is the group of ferromagnetic (FM)   materials   exhibiting   spin-polarized   carriers   and   known   as   half   metallic ferromagnets. The group includes  colossal magnetoresistance manganites such as Lao,67Cao,33Mn03 (LCMO), Lao,67Sro,nMn03 (LSMO), magnetite (Fe304), and other FM oxides [1,2]. Attempts are undertaken to prepare various multilayered device structures composed of ferromagnetic oxides, metallic, semiconducting and dielectric thin films. High magnetoresistance values and fast magnetic response make the heterostructures promising for fabrication of magnetic field sensors, magnetic random access memory elements (MRAM), magnetic field-driven optical devices and other devices.
Semiconductor p-n junction  and Schottky diode (metal-semiconductor junction)  are of key importance for Si-based electronics. One can expect that the combination of ferromagnetic oxides and organic semiconductors (OSC) could provide new electrical and magnetic functionalities into a single integrated device. Therefore elaboration of device structures containing FM oxides and OSC materials such as magnetic p-n, p-i-n diode structures as well as FMII/FM tunnelling heterostructures containing barrier (I) from dielectric or OSC materials could be a basic issue for novel devices. The main advantage of organic materials is relatively simple and low temperature  processing of their thin films through inexpensive techniques such as thennal evaporation,  drop­ casting and spin coating. Furthennore, OSC materials are composed of light elements and demonstrate unusually large values of spin relaxation time. It means that spin­ polarized carriers could be directly injected into OSC from ferromagnetic electrodes or generated via photoexcitation by a circularly polarized light. This provides an interesting possibility for the development of new magnetic field-driven optoelectronic devices [3].
Though several hybrid device structures containing OSC and ferromagnetic materials have been reported and some applications of their properties have been pointed out, more attention needs to be focused on both preparation and complex investigation of heterostructures based on magnetic oxides as manganites and organic compounds. More attention could be paid also to integration of the ferromagnetic oxides into Siwbased electronic circuits. There is also a great need to elucidate possible mechanisms of electrical transport at the interfaces of the heterostructures.
In this work, thin films and heterostructures based on La0,67Sr0,33Mn03 (LSMO) and Lao,67Cao,33Mn03 (LCMO) manganites and several OSC compounds were grown by various methods on crystalline MgO, n-SrTi03<Nb> (STON), n-Si and p-Si substrates. Structural quality of the prepared films has been investigated. Major attention was paid to electrical and magnetoresistive properties of both manganite films and the preparedp­ and p-i-n diode structures.
Simonas Kičas
Author: Simonas Kičas
Dissertation title: Fabrication, characterization and application of dispersive gradient refractive index multilayered systems
Fields of science: Technological  sciences, Material engineering (08 T)
Scientific supervisor: dr. Ramutis Drazdys
Defence of the dissertation: 2016-09-12


The first picture of a moving body was made in order to analyze the movement of objects too fast for the human eye to be clearly identified. In the famous article by J.D.B Stillman and E. Muybridge a running horse was the object. An example much closer to nowadays' realities would be pump-probe spectroscopy, which was introduced in the 19th century for exploring shock waves. So, in order to analyze short processes which change rapidly in time, we need as short as possible light pulses, as they define time the resolution of our experiments or setups. Laser sources exceeding long pulses were a breakthrough [1, 2]. Only several decades had passed until the first ultrashort pulses were demonstrated using passive-mode synchronization in dye lasers [3]. Since then, ultrashort pulse systems became a universal tool to make not only scientific research but were also used in industrial, medical and other applications. Pulses exceeding several optical cycles were used to monitor molecular dynamics, to make the micromachining of various materials [4].

One of the main characteristics of the short pulse is a broad spectral range. All these spectral components in the pulse have to be in a specific manner or, as it is called, "in phase" to form a transform limited pulse [5]. This means that the pulse is then travelling in a dispersive medium, and phase changes have to be compensated to maintain it short. Several different methods have been proposed over the years. Most of them are based on diffraction gratings [6], prism pairs [7], dispersive mirror [8-11) or a combination of these methods [5, 12]. Each of the aforementioned methods has its benefits and drawbacks. The grating and prism systems can create large amounts of negative dispersion, but they are sensible to alignment errors, have a low efficiency and cannot compensate a higher-order dispersion in a wide spectral range. Meanwhile, dispersive mirrors can efficiently compensate a higher-order dispersion, are easy to align, but can create a limited amount of dispersion [13]. However, today dispersive mirrors are most widely used when it comes to pulse compression, mostly because of the alignment advantage and flexibility to have a varying dispersion in a wide spectral range [14-16]. Since the first demonstration of a chirped mirror in 1994 [9], a lot of different designs have been proposed  to modify the classical one: At present, it is possible to distinguish at least five most popular lypes of these mirrors: double ehirped [17-19], Brewster [20, 21 ], complementary pair [16, 22, 23], high dispersion (HDM) [II, 14, 24], BASIC [25] mirrors. Each type has its own benefits, but in general they all are ttying to solve several problems related to chirped mirrors: 1) GVD dispersion oscillations caused by interference at the interface between the air and the mirror surface; 2) a large GVD characteristic sensitivity to deposition errors; 3) the amount ofGVD introduced in one bounce from such a mirror.

Along with the above mentioned tasks, chirped mirrors have to withstand the requirements of the ever-growing need for a higher power of ultrashort systems. Recently created laser oscillators and amplifiers  based on the fiber [26, 27] or the thin disk technology [27] can generate sub-picosecond pulses with the energy of up to tens of d and the repetition rate close to 10 MHz, thus reaching hundreds of watts as an average power or several gigawatts as the peak power. The laser-induced damage mechanisms at these power levels are unavoidable. One of the weakest elements in respect of LIDT in ultrafast systems is optical coatings. The resistance of optical elements to laser exposure was one of the main topics after the first laser pulses had been demonstrated. Laser damage studies in the nanosecond regime were published in variety of papers and are still an interesting topic for both science and industry. After femtosecond pulses had been demonstrated, optical resistance in this field became an interesting issue, too. However, most of studies in this field are carried out at the so called 1 on-1 regime, or using pulse trails at a kHz repetition rate. It should be noted that the laser resistance of dispersive coatings is a rather new field, and only several papers on this topic have been published recently.
Rasa Godliauskienė
Author: Rasa Godliauskienė
Dissertation title: Research of persistent organic pollutants in food and feed using mass spectrometry method
Fields of science: Physical sciences, Chemistry (03P)
Scientific supervisor: dr. Evaldas Naujalis
Defence of the dissertation: 2016-10-21
Various groups of persistent organic pollutants such as polychlorinated dibenzo-p-dioxins (PCDD), polychlorinated dibenzofurans (PCDF), dioxin like polychlorinated biphenyls (DL-PCB), non dioxin like polychlorinated biphenyls (non DL-PCB), and polybrominated diphenyl ethers (PBDE) are known as xenobiotics, characterized by not only bioaccumulative features but also toxicological, carcinogenic, neurotoxicological, and teratogenic influence. Undoubtedly, food contaminated by these kind of pollutants may cause a range of various illnesses (porphyria, chloracne, dermatitis, disorders of both immune and neurological systems) including different types of cancers which in fact are prevailing. The group ofPCDD/PCDF contains 210 isomers where 2,3,7,8-tetrachlorodibenzo-p-dioxin (tetraCDD) can be distinguished as the most toxic one. Polychlorinated biphenyls (PCB) is a category of 209  organic compounds which differ in two benzene rings connected by C-C link. 209 brominated diphenyl ethers are aromatic compounds in which l-10 hydrogen atoms are replaced by bromine. Polybrominated diphenyl ethers (PBDE) considering bromine atoms from 2 to 10 are used in commercial fire retardant and it is divided into tri (3), tetra (4), penta (5), hexa (6), hepta (7), octa (8), nona (9)  and decabrom ether. The latter, which are incorporated into polymers as fire retardants reduce the risk of ignition. Increasing environmental pollution was caused by the chemical stability of POP. For this reason, industrial production of these compounds was gradually rejected. However, the pollution remains to be a huge problem even nowadays.
Since these persistent organic pollutants may provoke negative effects on human beings even with extremely low concentrations like l0"12g/g or l0"15g/g, in this case precise, sensitive and reliable methods of analysis are necessary in order to identify these negative effects. For the last 30 years, gas chromatography mass spectrometry method (GC-MS) has been used for the qualitative and quantitative identification  of dioxins and furans. Nevertheless, because of decreasing regulated maximum allowable concentration, the methods of high-resolution  mass spectometry analysis became irreplaceable due to both high sensitivity  and ability to identify low concentrations.
Due to the toxicity of low concentrations of persistent organic pollutants, not only complex instrumental analysis is required, but also thorough methods of sample preparation. In recent years, more than one instrument which is able to automatize a part of sample preparation procedure has been invented. It reduces the duration of analysis, and the level of pollution and increases the sensitivity. Nevertheless, the manual preparation method which has been used for a long time still holds its position comparing with the latter since it meets all the requirements of European Commission for POP analysis.
Even  90%  of  pollutants  get  into  the  human  body  together  with  nutrition, especially with the food of animal origin. Therefore, it is of great importance to estimate and control the tolerated daily limit of POP. Moreover, large quantities which have been accumulated during several decades are stored in sediments of the bottom of the sea which is essential for Lithuania which possesses the seashore of the Baltic sea.
The aim of gathered scientific research in this doctoral thesis is to examine the possibilities and to establish an identification procedure of persistent organic pollutants (PCDD/PCDF, DL-PCB, non DL-PCB and PBDE) assigned to the analysis of food and feed with the help of high-resolution gas chromatography mass spectrometry method.
In this dissertation the purpose of summaried research work is to develope, optimise and validate  the persistent organic pollutants (PCDD/PCDF, DL-PCB, non DL-PCB and PBDE) method of analysis in food and feedingstuffs using high resolution gas chromatography-mass spectrometry method.
Jolita Jablonskienė
Author: Jolita Jablonskienė
Dissertation title: Formation  and studies of grapi-iene- metal particles nanocomposites and their application  for fuel cells
Fields of science: Physical sciences, chemistry (03 P)
Scientific supervisor: dr. Loreta Tamašauskaitė-Tamašiūnaitė
Defence of the dissertation: 2016-09-23
Hydrogen energy is no longer a novelty in the history of our days. Fuel cells, which use hydrogen fuel, have been already investigated quite a long time. Fuel cells are alternative source of energy where chemical energy is directly converted to electricity due to the occurring catalytic electrochemical reactions. Principle of fuel cells operation is similar to that of conventional batteries, but the main difference is that they do not consume and they do not need recharging. They produce both electricity and heat only when the fuel  is supplied to them. Fuel cells are promising environmentally clean energy sources. During the last decades fuel cells is one of the most extensively developing fields of research and the scientific achievements and technical solutions already allow to use their prototypes to fulfil different needs. A number of pilot car models operated by electricity produced by fuel cells (eg., General Motors, Toyota, Honda, etc.) have been created. Fuel cells are also applied as power sources for local power plants for houses (Ballard, Siemens) as well as for power sources for portable electronic devices (mobile phones, computers, audio-video equipment, etc.). Fuel cell research is also intensively carried out in Lithuania.
Although platinum is well known as one of the best catalyst used in fuel cells, however, the use of a noble metal as the anode material is limited by its high price. In recent years, various Pt-based catalysts with non-noble metals are extensively created with the aim to decrease the costs of catalysts and to improve their activity. So, the main aim is to prepare an efficient catalyst for fuel cells as well as to create fast, simple and cost-effective technologies for their fabrication. For this purpose, the promising microwave synthesis method for the one-pot synthesis of metallic nanostructures in solutions has been chosen. This method allows to deposit catalysts with a small particle size and high distribution on the support and with tailored properties.
From another point view, to obtain an effective catalyst the proper support is needed for deposition of Pt or other metals with a small particle size, natTOW size distribution, and excellent  dispersion  on the support.  Recently, graphene  as catalyst  support  has  incurred intense  interest  in  fuel  cell  applications  due to  its  unique, outstanding  physicochemical properties, such as an extremely high specific surface area (2600 m2g-1), a  superior electronic conductivity, a high surface to volume ratio and a high stability. The combination of metal nanoparticles and graphene opens up new possibilities for design of the next generation catalysts having more pronounced  catalytic activity.
Karolis Viskontas
Author: Karolis Viskontas
Dissertation title: Fabrication  and characterization  of low­ dimensional nanomaterial  saturable absorbers for mode-locking  of fiber lasers
Fields of science: Technological Sciences, Material Engineering (08T)
Scientific supervisor: dr. Kęstutis Regelskis
Defence of the dissertation: 2016-09-22
Here we introduce 20, 10 and OD nanomatcrial saturable absorbers for self start mode­ locking of fiber lasers. The locking of longitudinal  modes in a resonator is the only wny to produce ultra-short (<10 ps) pulses. By utilizing this technique for the fiber lasers  and choosing a particular cavity configuration it is possible to tunc pulse duration  from many picoseconds down  to several hundreds of  femtoscconds. These ultra-short pulses are particularly important for femtochemistry [1], spectroscopy [2], two-photon microscopy and optical metrology [3]. However, it is extremely important to use the correct  saturable absorber for different cavity configuration and only low-dimensional materials have the appropriate characteristics to sclf start and hold .stable long-term operation.
Semiconductor saturable absorber mirrors (SESAMs) arc well known 20 nanomaterial based structures suitable for mode locking of both the solid state and of the fiber lasers. However, the reliability of these devices depends on growth  technology,  materials  and cavity design. For fiber lasers, SESAMs must posse...:;s higher nonlinearity than for solid state lasers. For this reason it is necessary to increase the number of active quantum well layers and usc a higher  field enhancement factor for the device. Both cases  influence the damage thresholds of the SESAM. Therefore, we perfmmed long term (>5000 h) optical degradation tests of the high modulation depth satumble absorber. Also, we present results of experiments with the reliability of saturable absorbers utilized in different fiber laser configurations. Along with the results, we introduce a growth methodology to fabricate the low saturation fluence and high modulation depth SESAM, which is resistant  to  intense optical radiation. Also, we investigate the influence of different kinds of defects on the quantum well region.
SESAM reflects  the great  potential  of  low dimensional  structures  to  mode lock  ultra­ - short pulse fiber lasers. In order to compare the nonlinear optical properties of 20, ID and OD nanomateria\s, we also investigated single-wall  carbon nanotube ( 1D) and PbSe quantum dot (OD) saturable absorbers. Unlike the  SESAM,  many  low  dimensional structures exhibit naturally fast recovery of optically excited carriers. Also, the saturation fluence  of  the  device  decreases  with  reducing  the  dimensionality  of  absorber  material. Another advantage of ID and OD semiconducting nanoparticles is the possibility to disperse them in a liquid environment. This feature enables us to fabricate a fiber integrated saturable absorber. As a consequence, the robust monolithic structure of an all-fiber laser with potentially trouble-free, long-term performance provides us with  the opportunity  to  use them for medical or industrial applications (4]. However, the nonlinearity strongly depends on the size of the nanoparticle and laser wavelength. Therefore, it is necessary to choose a nanomaterial with the first excitonic state at the operating wavelength of a fiber laser.
Irmantas Ratas
Author: Irmantas Ratas
Dissertation title: Algorithms for inhibition and desynchronization of neural systems
Fields of science: Physical Sciences, Physics (02P)
Scientific supervisor: prof. habil. dr. Kęstutis Pyragas
Defence of the dissertation: 2016-01-22
The dissertation deals with excitable am\ oscillatory neural systems. Although we use simplified mathematical models for real neural networks, we believe that the new insights made in this thesis might lead to novel therapeutic approaches in neuroscience.
The dissertation  consists of four chapters.  Three of these chapters present  an originalmaterial. In the first chapter motivation, definitions and used models arc introduced. In the second chapter, homogeneous high-frequency stimulation effect on a pulse propagation in a single axon is analysed. We consider two cases - unmyelinated and myelinated axons. As a basis for an axon model we chose the FitzHugh-Nagumo equations. Seeking for analytical results we first apply an averaging  method  and separate the neuron  dynamics into slow and fast components. As a result we derive averaged equations for the slow component that do not  contain a high-frequency term. Then we analyse the travelling waves in the averaged system via an asymptotic pulse construction method.  Such an approach allow us to obtain the depemlence of the parameters of the travelling pulse on the amplitude of high­ frequency stimulation current in an analytkal form. It was estimated, that a homogeneous high-frequency stimulation  is an effective tool to control  pulse propagation  in the a.xon us stimulation parameter regulates pulse  width and speed. It was established that high stimulation intensities can suppres pulse propagation. In the case of myelinated axons and small stimulation intensity, pulse enhancement is possible. Mechanism of this behaviour is related to the dependence of the excitability parameter  on the stimulation intensity. The validity of the results arc rested hy numerical simulations.
The  third  chapter  deals with  synchrony  suptn·c_ssion  in  ensemhlcs  of  globally  coupled oscillators. Firstly we propose an algorithm for a complicated control  situation,  when  an output. signal  is small  in  comparison  to  a  stimulated  signal  so  that  reliable  simultaneous registration and stimulation  of the system  is not. possible  and the standard feedback cont­ rol algorithms cannot be applied. Sud1 a situation is t.ypical for neuronal systems. Our al).!;orithm is haM'cl on the act-and-wait. control, which assumes a separation in time of the registration  and stimulation stages.  In the registration (wait) stage, the mean field of the free osillatory system is recorded in a memory and in the stimulation  (act) stage it is fed back to the system.  The periodic repetition of these two stages can effectively destroy the mutual synchronization in ensembles of globally coupled oscillators. It was shown, that the incoherent state stability domains of the act-and-wait algorithm in the parameter  plane of the feedback delay time τ and gain P, have a resonance structure; they are located at the values of τmT/2, where is the characteristic period of oscillations of the mean field of synchronized oscillators in the absence of control and is a non-negative integer number. Also a modified version of the algorithm taking into account the charge-balanced require ment has been considered. The charge balance requirement docs not destroy the resonance structure of the incoherent state stability domains in the (τ, P) plane, but the algorithm stops to work for small values of 7. The efficiency of our algorithm is demonstrated ana­ lytically and numerically for globally coupled Stuart-Landau oscillators, and synaptically all-to-all coupled FitzHugh-Nagumo as well as Hodgkin-Huxley neurons.
In the la.'lt chapter the problem of controlling synchrony in bistable networks, which possess the coexisting coherent and incoherent states, both being stable for the same values of the parameters, were considered. The specific asymptotic state of such a system depends on the initial conditions. The control problem that we formulate here is as follmvs. We assume that a bistable network is initially in the stable coherent state and our aim is to design a particular time-dependent perturbation which enables us to switch the system to the stable incoherent state. Such a problem can be motivated, e.g., by a possible control of epilepsy, which is modelled hy a histahk neural network in which the disease and healthy states arc associated with the stable coherent am\ incoherent states, respectively.
The act-and-wait and multisitc coordinated reset stimulation methods were suggested for switching the bistable networks from the stable coherent state to the stable incoherent. state. It is shown, that act-and-wait algorithm is efficient for the globally coupled populations of ost"illators, while multisite coordinates reset stimulation is effective for global coupling and complex scale-free topology networks.
Linas Minkevičius
Author: Linas Minkevičius
Dissertation title: Terahertz imaging arrays for room temperature operation
Fields of science: Physical Sciences, Physics (02 P)
Scientific supervisor: prof. dr. Vincas Tamošiūnas
Defence of the dissertation: 2016-06-20
Tcrahertz radiation  range  is  between  electromagnetic  wave  spectra  from   100 GHz and 10 THz (1 THz = 1012 Hz) or wavelength from 3 mm to 30 μm [1]. This radiation has a feature  to penetrate  through  dielectric,  textile,  plastic,  paper  and  other  materials.  It can be used as a nondestructive  way  to  inspect  objects  in  customs  and  airports  [2]; to  locate the defects  in solar cells  panels  [IP 5] and  even  as  an  auxiliary  way  in  cancer diagnostics in medicine [3].
All  of these  applications  require  different  terahertz  imaging  systems  solutions. Particular attention goes to fast sensors and sensor technologies which have the potential for real-time imaging while maintaining a high dynamic range. Furthermore, it is highly desirable to be able to operate the devices at room temperature (RT). Examples of sensors fulfilling these requirements are Schottky detectors [4], microbolometers arrays [5,6] and semiconductor field-effect transistors (FETs) [7]. An important role during the development of imaging systems goes to simplicity and reliability ofTHz detectors. Planar asymmetrically shaped- also called bow-tie (BT) diodes because of their general shape should be mentioned. Asymmetric shape GaAs/AlGaAs sensors are robust and electrostatic resistant. However, direct  application of such detectors in THz imaging systems are difficult due to their low sensitivity at room temperature [8].
Several steps should be done in order to increase  the detector sensitivity. First, altemative materials could be employed, such as InGaAs due to its high electrons mobility at room temperature (9]. Also, the design of the sensor antenna should be considered. Moreover, detector ability to operate in heterodyne mode [1OJ could increase the dynamic range of imaging systems. Finally, modern compact diffractive optics components, suitable for detector arrays should be considered. To complete all above-mentioned processes, complex interaction studies of sensors and diffractive focusing elements should be performed.
Thus, resolving these challenges, widely used relative slow spectroscopic imaging systems in future could be changed with much faster system consisting of sensors array, working in heterodyne mode. Such solutions can find the applications in material inspection or ·solar cells quality control in fast production line.
Rimvydas Venckevičius
Author: Rimvydas Venckevičius
Dissertation title: Compact  spectroscopic  terahertz  imaging solutions using GaAs/AlGaAs and InGaAs semiconductor  nanostructures
Fields of science: Physical Sciences, Physics (02P)
Scientific supervisor: prof. habil. dr. Gintaras Valušis
Defence of the dissertation: 2016-06-20
Terahertz (THz) radiation corresponds to electromagnetic spectrum lying between microwaves and infrared at frequency range from 0.1 to 30 THz (1 THz = 1012 Hz ~4 meV). Most organic materials, chemical compositions, biological agents exhibit spectral fingerprints in this region. However, THz range is rather complicated to implement into practice because operation principles for devices are neither truly electronic, nor truly optical.  Progress in development of coherent THz Time-Domain (TDS) spectroscopy systems in last decade allowed to exploit this frequency range in various scientific as well as numerous practical applications. Although coherent systems exhibits very large dynamic range, they are not very well suited for fast and compact THz imaging systems due to necessity to measure THz impulse waveform. That is why non-coherent THz imaging systems with fast operation are considered for practical applications (such as security systems, quality control, etc.). Possible option would be to avoid coherent detection, but this would require new more powerful emitters and amplifiers and/or more sensitive detectors. Therefore, beam shape of THz emitter and its influence on image quality is also rather relevant, especially when dimensions of emitter is decreased below wavelength and additional beam shaping may add some power loss.
As for modern room temperature THz emitters, Quantum Cascade Lasers (QCL) with intracavity difference-frequency generation are now of particular interest. However, these devices operate at higher frequencies (~3 THz), but for practical needs, longer penetration depth is needed, hence, frequencies below 1 THz are of preference.
Invention of Molecular Beam Epitaxy technology stimulated huge scientific interest in artificial materials, having desirable properties defined rather by materials layer thickness than properties itself. One of the exciting options in materials engineering was semiconductor superlattices (SL) as a medium to excite Bloch oscillations (BO) as it was suggested by Esaki-Tsu in 1971.They were discovered in optical experiments, however, still neither scientific works revealing signatures of electrically pumped BO nor Bloch lasing are published.
Special technological requirements are considered for spectroscopic THz imaging systems essential for practical applications allowing one direct determination of material composition without measuring full spectrum. For real-time operation fast response broadband THz detectors or cameras with a few color pixels are required. Recent development of THz detectors resulted with some high performance sensors, such as Schottky diodes and field effect transistors, but complicated technology determines high prices of such devices. THz sensors displaying good sensitivity, fast operation speed, resistant to electrostatic noise, suitable for array fabrication, and manufactured by industrial low cost technology remains an important issue. Also, the preference would be devoted to planar technology solutions.
This doctoral thesis concentrates on finding, investigating, and applying new solutions for fast, compact, relatively easy to use, room-temperature spectroscopic imaging systems. Comprehensive complex study was carried out involving all main devices of THz imaging systems including sources, amplifiers, optical components and detectors. Main results of this work are distributed throughout 4 main chapters.
In the 1st chapter the results of experimental study on THz surface emitter based on p-i-n-i-type AlGaAs/GaAs heterostructure designed by A. Reklaitis are presented. The 2nd chapter is dedicated to consistent investigation of AlGaAs/GaAs superlattice diode as a medium with strong homogeneous electric field required for the manifestation of stable and modulated Bloch gain. In the 3rd chapter experimental study revealing how usage of multimode laser source influences quality of THz images. Spectroscopic THz imaging system employing TEM01 laser mode was applied for the THz imaging of silicon solar cells. The last 4th chapter of the thesis is dedicated to the investigation and application of new innovative non-coherent THz detector solutions. Using spectroscopic imaging the principal component analysis of explosive simulators based on sucrose and tartaric acid was carried out. In such experiment broadband bow-tie InGaAs diodes and resonant antenna-coupled field effect transistors were used as THz sensors for the first time. As for THz imaging systems with optoelectronic emitters, usually used in coherent THz time-domain systems, non-coherent sensitive low noise sensors based on titanium microbolometers were applied for spectroscopic imaging of samples made of lactic acid and tartaric acid compounds.
Ernesta Meinorė
Author: Ernesta Meinorė
Dissertation title: Variation of the physical and chemical properties of submicron atmospheric aerosol particles
Fields of science: Physical Sciences, Physics (02P)
Scientific supervisor: dr. Kęstutis Kvietkus
Defence of the dissertation: 2016-07-01
Science   of   atmospheric   aerosol   particles   was   commonly   acknowledged as independent discipline  from  the  mid  of XIXth century, but nowadays  there are stili a substantial uncertainties in  detennination  of  the  total  effect  of  atmospheric  aerosol particles  on the global  climate change and the local pollution.  Therefore, continuous measurements  are mandatory:  i) to better understand  ongoing processes;  ii) to predict possible consequences; iii) to maintain and implement goals of sustainable growth. This dissertation   is   dedicated   to   obtain   supplementary information   physical-chemical properties (PMI) and their influence on the particle evolution.
Data of the chemical composition, concentration, size distribution of semi-volatile submicron atmospheric aerosol particles  are collected at the three measurement  sites: Vilnius- an urban site, Rūgšteliškis - a rural site, Preila- a coastal marine site. Analysis indicates that organics is the most abundant  constituent  that accounts on the average from 68.8% to 77.1% of the total measured volatile and semi-volatile PMl at rural and background   sites,   respectively.   Theoretical   analysis   of   ammonium   sulfate   and ammonium nitrate contribution to the total of ammonium shows that ammonium nitrate out of all nitrates (on the average 61 ± 9% at Vilnius site, 61 ± 5% at the Rūgšteliškis and 80 ± 3 % at the Preila site) made a greater contribution to the total of ammonium than that of ammonium sulfate out of all sulfates (37 ± 5 %, 44 ± 2% and 36 ± 2 %for the Vilnius, Rūgšteliškis and Preila sites, respectively).
Positive Matrix Factorization (PMF) tool is applied to organics component for the origin apportionment. Three factors HOA (Hydrocarbon-like Organic Aerosol), BBOA (Biomass Burning Organic Aerosols) and SV-OOA (Semi-Volatile Oxygenated Organic Aerosols) are found at the Vilnius site; three factors of BBOA, SV-OOA and LV-OOA (Low-Volatility Oxygenated Organic Aerosols) are indicated at the Rūgšteliškis site and two factors of LV-OOA and SV-OOA (64% and 36% respectively) are identified at the Preila  site. Constituents of  organic  component  are  recognized  to  be relatively  more oxygenated in the SV-OOA factor, meaning that they arc also more aged at the rural sites. This finding supports the presumption that long-range transport air masses have a great impact on physical properties and overall burden of ambient aerosol particles at the remote sites.
Density of aerosol particulate matter within the particle itself is analyzed from the data collected at the urban site in Siauliai near the main street with intensive traffic. Field measurement is performed with SMPS and Q-AMS that were deployed to collect the data. Peculiarities  of obtained data were taken into consideration and therefore, corrections were applied accordingly, in order to combine data from both instruments. However, results of combined SMPS and Q-AMS data analysis stipulate that density of the particles with the diameter below I 00 nm is relatively too high when compared with the density sum of all components with the diameter above 100 nm (less than 1 g cm-3).
Initially it was presumed  that density for particles  in the size range above  I 00 nm should exceed 1 g cm-3 contrary to that of obtained results. Further analysis was conducted and it was found that deployed instruments have assigned pmiiclcs according to the size en·oneously; hence a correction on the diameter calculation was upplicd and the modeling in the density-size field was conducted. Results imply that it is erroneous to consider the same size particles to be homogeneous. Particles arc emitted from different sources; therefore the particulate matter density of the same size particles may vary. It is concluded that fresh-formed particles form aggregates and their density is less than 1 g cm-3 in  the  size range  less  than  100 nm.  Besides,  the  value  of the  median  density decreases, whereas  a standard  deviation  of the density  increases  with  the decreasing diameter of the aerosol particles.
Data from captured volcanic pollutant episodes from 24 May until 29 May 2011 at Institute of Physics, Vilnius are evaluated in order to study the extent of particle neutralization. Two episodes are selected to analyze the extent of spontaneous neutralization. The first episode is unique and enables evaluating the extent of acidic sulfates neutralization with ammonia in the atmosphere. Conducted calculations indicate that acidic sulfates on volcanic origin aerosol particles can be fully neutralized with atmospheric ammonia when their size is of about tenths of nanometers less than 100 nm, whereas  the  larger  particles  have  a  lower  extent  of  neutralization.  Therefore,  the neutralization of acidic sulfates with ammonia in volcanic origin aerosol particles can be estimated only based on the size of aerosol particles. It is also shown that ammonia is deposited on the surface area of acidic sulfate containing aerosol particles during the transfer with the air masses to the measurement site. A rough estimation of the ammonia flux onto the aerosol particle surface has been made using data of Episode 1; the value is estimated  to be  from  30 µg m-2h-1 to 55 µg m-2 h-1 (in case of particle  being  dry, the  flux would be 74 µg m-2h-1).