Polarization-sensitive photoresponse of metal-semiconductor nanocomposite film

(1)

uef.fi

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND Dissertations in Forestry and Natural Sciences

ISBN 978-952-61-3670-7 ISSN 1798-5668

Dissertations in Forestry and Natural Sciences

DISSERTATIONS | ALEKSANDR SAUSHIN | POLARIZATION-SENSITIVE PHOTORESPONSE OF METAL-SEMICONDUCTOR... | No 411

ALEKSANDR SAUSHIN

Polarization-sensitive photoresponse of metal-semiconductor

nanocomposite film

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND

The Thesis is devoted to the investigation of the polarization sensitive photoresponse of silver-

palladium (Ag/Pd) nanocomposite films in the ultraviolet, visible, and infrared spectral ranges

at nanosecond and femtosecond excitation.

Based on results it was shown that the photocurrent in Ag/Pd films can be described

by photon drag and surface photogalvanic effects. It was shown that polarization sensitive

photoresponse of Ag/Pd films is of great interest for photonics and optoelectronics, in particular, for the development of new types of

polarization analyzers.

ALEKSANDR SAUSHIN

(2)

(3)

POLARIZATION-SENSITIVE PHOTORE- SPONSE OF METAL-SEMICONDUCTOR

NANOCOMPOSITE FILM

(4)

(5)

Aleksandr Saushin

POLARIZATION-SENSITIVE PHOTORE- SPONSE OF METAL-SEMICONDUCTOR

NANOCOMPOSITE FILM

Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences

No 411

University of Eastern Finland Joensuu

2020

Academic dissertation

To be presented by permission of the Faculty of Science and Forestry for public examination in the Auditorium M100 in the Metria Building at the University of Eastern Finland, Joensuu, on December, 04, 2020, at

12 o’clock noon

(6)

Grano Oy Jyväskylä, 2020

Editors: Pertti Pasanen, Raine Kortet, Jukka Tuomela, Matti Tedre

Distribution: University of Eastern Finland / Sales of publications www.uef.fi/kirjasto

ISBN: 978-952-61-3670-7 (Print) ISBN: 978-952-61-3671-4 (PDF)

ISSNL: 1798-5668 ISSN: 1798-5668 ISSN: 1798-5676 (PDF)

(7)

Author’s address: Aleksandr Saushin

University of Eastern Finland

Department of Physics and Mathematics Institute of Photonics

P.O. Box 111

80101 JOENSUU, FINLAND email: alekssau@uef.fi Supervisors: Professor Yuri Svirko

University of Eastern Finland

Department of Physics and Mathematics Institute of Photonics

P.O. Box 111

80101 JOENSUU, FINLAND email: yuri.svirko@uef.fi Professor Gennady Mikheev Institute of Mechanics

Udmurt Federal Research Center of the UB RAS Laser Laboratory

426067 IZHEVSK, RUSSIA email: Mikheev@udman.ru Reviewers: Professor Sergey Tarasenko

Ioffe Physical-Technical Institute of the RAS Centre of Nanoheterostructure Physics 194021 SANKT-PETERSBURG, RUSSIA email: tarasenko@coherent.ioffe.ru Professor Teruya Ishihara

Tohoku University Department of Physics 980-8578 SENDAI, JAPAN email: t-ishihara@tohoku.ac.jp Opponent: Professor Zhipei Sun

Aalto University

Department of Electronics and Nanoengineering P.O. Box 11000

02150 ESPOO, FINLAND email: zhipei.sun@aalto.fi

(8)

(9)

7 Saushin, Aleksandr

Polarization-sensitive photoresponse of metal-semiconductor nanocomposite film Joensuu: University of Eastern Finland, 2020

Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences 2020; 411 ISBN: 978-952-61-3670-7 (print)

ISSNL: 1798-5668 ISSN: 1798-5668

ISBN: 978-952-61-3671-4 (PDF) ISSN: 1798-5676 (PDF)

ABSTRACT

The Thesis is devoted to the investigation of the polarization sensitive photoresponse of silver-palladium (Ag/Pd) nanocomposite films, which are widely used in microelectronics, in the ultraviolet, visible, and infrared spectral ranges at nanosecond and femtosecond excitation. By using X-ray diffraction, X-ray photoelectron spectroscopy, Raman spectroscopy and thermodynamic modeling, it was shown that the phase composition of the Ag/Pd films depends on fabrication temperature, which determines electronic and photovoltaic properties of the films.

Based on results of experiments we found that the photocurrent in Ag/Pd films originates from the photon drag and surface photogalvanic effects. The obtained results allowed us to obtain nonlinear susceptibilities of the Ag/Pd nanocomposites and to investigate their dependence on the fabrication conditions. It was shown that polarization sensitive photoresponse of Ag/Pd films is of great interest for photonics and optoelectronics, in particular, for the development of fundamentally new types of polarization analyzers.

Universal Decimal Classification: 535.215, 535.51, 538.975, 539.23, 620.3, 621.315.5 Library of Congress Subject Headings: Semiconductors; Semiconductor films; Thin films;

Nanocomposites (Materials); Microfabrication; Silver; Palladium; Photovoltaic effect;

Photoelectricity; Photons; Polarization (Light); Photonics; Nanophotonics

Yleinen suomalainen ontologia: puolijohteet; ohutkalvot; komposiitit; nanomateriaalit;

hopea; palladium; polarisaatio (aaltoliike); fotonit; fotoniikka

(10)

8

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my supervisors Professor Yuri Svirko and Professor Gennady Mikheev, for their general scientific guidance, invaluable contribution to my development as researcher, comprehensive support, fruitful discussion of research results, assistance in understanding the essence of the observed phenomena and great help with thesis preparation.

Great thanks to all my coauthors from Institute of Mechanics of Udmurt Federal Research Center of the UB RAS, University of Eastern Finland and Physical- Technical Institute of Udmurt Federal Research Center of the UB RAS for fruitful collaboration and significant contribution to the experiments and the preparation of papers.

I am grateful to my colleagues from Laser Laboratory of the Institute of Mechanics of Udmurt Federal Research Center of the UB RAS and from Departament of Physics and Mathematics of University of Eastern Finland for good working atmosphere.

Finally, I would like to thank my parents Sergei and Svetlana and my wife Julia for their love and involvement in my life.

Joensuu, 19th September 2020 Aleksandr Saushin

(11)

9

LIST OF ABBREVIATIONS

Ag/Pd silver-palladium

Ag-Pd silver-palladium solid solution CPDE circular photon drag effect CPGE circular photogalvanic effect FHG fourth harmonic generation HIC hybrid integrated curcuit

KDP potassium dihydrogen phosphate crystal KH²PO⁴ KTA potassium titanyl arsenate KTiOAsO⁴

KTP potassium titanyl phosphate KTiOPO4

LPGE linear photogalvanic effect OPG optical parametric generator OPO optical parametric ocillator PDE photon drag effect

PGE photogalvanic effect RC resistor–capacitor

SEM scanning electron microscope SHG second harmonic generation SPGE surface photogalvanic effect THG third harmonic generation XPS X-ray photoelectron spectroscopy

(12)

10

(13)

11

LIST OF ORIGINAL PUBLICATIONS

This thesis is based on data presented in the following articles, referred to by the Roman Numerals I-X.

I A. S. Saushin, R. G. Zonov, E. V. Aleksandrovich, K. G. Mikheev, R. Ali, V. V. Vanyukov, G. M. Mikheev (2019). Influence of electrochemical hydrogenation on the circular photocurrent in the Ag/Pd nanocomposite.

Physica status solidi b, 1800671.

II G. M. Mikheev, A. S. Saushin, V. M. Styapshin, Yu. P. Svirko (2018).

Interplay of the photon drag and the surface photogalvanic effects in the metal-semiconductor nanocomposite. Scientific Reports, 8, 8644.

III A. S. Saushin, K. G. Mikheev, V. M. Styapshin, G. M. Mikheev (2017). Direct measurement of the circular photocurrent in the Ag/Pd nanocomposites.

Journal of Nanophotonics, 11, 032508.

IV G. M. Mikheev, A. S. Saushin, V. V. Vanyukov, K. G. Mikheev, Yu. P. Svirko (2017). Femtosecond circular photon drag effect in the Ag/Pd nanocomposite.

Nanoscale Research Letters, 12.

V A. S. Saushin, R. G. Zonov, K. G. Mikheev, E. V. Aleksandrovich, G. M. Mikheev (2016). The Influence of PdO Content on Circular Photocurrent in Resistive Ag/Pd Films. Technical Physics Letters, 42, 963–966.

VI K. G. Mikheev, A. S. Saushin, R. G. Zonov, A. G. Nasibulin, G. M. Mikheev (2016). Photon-drag in single-walled carbon nanotube and silver-palladium films: the effect of polarization. Journal of Nanophotonics, 10, 012505.

VII A. S. Saushin, R. G. Zonov, K. G. Mikheev, R. R. Shamshetdinov, G. M. Mikheev (2016). Polarization-sensitive photocurrent in the resistive Ag/Pd films. Journal of Physics: Conference Series, 741, 012093.

VIII G. M. Mikheev, A. S. Saushin, V. V. Vanyukov (2015). Helicity-dependent photocurrent in the resistive Ag/Pd films excited by IR laser radiation.

Quantum Electronics, 45, 635-639.

IX G. M. Mikheev, A. S. Saushin, O. Yu. Goncharov, G. A. Dorofeev, F. Z. Gil’mutdinov, R. G. Zonov (2014). Effect of the Burning Temperature on the Phase Composition, Photovoltaic Response, and Electrical Properties of Ag/Pd Resistive Films. Physics of the Solid State, 56, 2286–2293.

(14)

12

X G. M. Mikheev, A. S. Saushin, R. G. Zonov, V. M. Styapshin (2014). Spectral dependence of circular photocurrent in silver–palladium resistive films.

Technical Physics Letters, 40, 424-428.

In addition to these publications, the author has other papers marked in the bibliography [1–6] which are not included in the Thesis.

(15)

13

AUTRHOR’S CONTRIBUTION

The publications selected in this dissertation are original research papers on photovoltaic effect in silver-palladium films. The papers I-X present results obtained by the team members stated in the list of authors. The main ideas of the papers were created in productive discussions of all team members. In all papers, the author has been involved in analyzing the experimental data and participated in preparation of the manuscripts. The experimental results on measuring of photocurrent pulses at different polarization in papers VI and VII were partially obtained by the author. The experimental measurements of photocurrent pulses in papers III and X were mainly performed by the author. The author has contributed to the measurements processing in papers I, II, III, IV, VI, VII, VIII and X. In paper III, the author has performed main calculations. The papers have been completed with significant co-operation with co-authors.

(16)

14

1 THE MAIN MECHANISMS OF THE POLARI- ZATION SENSITIVE PHOTORESPONSE

Irradiation of the metallic and semiconductor materials with laser pulses results in the generation of a photocurrent, which depends on the polarization and angle of incidence of the laser beam. On the microscopic level, this photocurrent originates from the sensitivity of the elementary processes of the elastic and non-elastic electron-photon scattering to the photon momentum and the symmetry of the crystal- line field [7]. In semiconductors without inversion center, this sensitivity manifests itself as the photogalvanic effect (PGE) [8]. When the electron-hole pairs due to light absorption are created in the proximity of the semiconductor surface, the diffuse scattering of the photogenerated carriers on the surface results in the generation of the surface current which is manifestation of the surface photogalvanic effect (SPGE) [9]. In the media with free carriers, the transfer of momenta of incident photons to these carriers in the process of elastic scattering manifests itself as the photon drag effect (PDE) [10,11], which can be observed in non-centrosymmetric and centrosymmetric media for arbitrary polarized light beam [12].

The finite conductivity of metals and semiconductors makes registration of the above effects a difficult task because the photocurrent is short-circuited in the bulk of the sample. However, if the sample is a film with a surface resistivity of several tens of Ω/□, the photocurrent can be visualized using a conventional broadband oscilloscope.

Figure 1.1. Experimental configurations showing the plane of incidence of the radiation and the arrangement of contacts to measure (a) the longitudinal and (b) the transverse photocurrents.

(20)

18

In the film materials, it is possible to measure photocurrents propagating both along and perpendicular to the plane of incidence of the excitation laser beam.

These photocurrents are correspondingly referred to as longitudinal (J^x) and trans- versal (J^y) photocurrents (see Figure 1.1). The transversal photocurrent can consist of linear and circular contributions, which vanish for the circular and linearly polarized beams, respectively. The measurement of the longitudinal and transverse photocurrents on the incidence angle and polarization state of the excitation beam allows us to reveal the mechanism responsible for the photoresponse.

In this Chapter, we provide overview on the major mechanisms responsible for the polarization sensitive photoresponse with emphasis on film materials.

1.1 PHOTOGALVANIC EFFECT

The photogalvanic effect (PGE) manifests itself as a generation of the stationary electric current in a homogeneous medium during its uniform irradiation with light. PGE is due to the asymmetry of the elementary processes of generation, recombination, and scattering of charge carriers, i.e. it is not associated with inhomogeneity in the irradiation pattern or a temperature gradient.

Figure 1.2. Illustration of the asymmetry of elastic scattering on a wedge (adapted from [8]).

A simple example of the asymmetry of elementary electronic processes is the elastic scattering of a particle by a wedge-shaped potential (Figure 1.2) [8]. It can be seen that for particles located to the right of the wedge during elastic scattering, the probability of the transition of a particle from a state with momentum p to a state with momentum p' is equal to the probability of a transition from state –p' to state -p. At the same time, for particles on the right side of the wedge, it cannot be said that the transition from the –p state to -p' is equally probable the transition from p to p'. Thus, for particles whose momentum is directed in opposite directions, scattering will occur in different ways. If variable electric field acts on the particles,

(21)

19 the field vector of which is oriented vertically, and its amplitude changes according to E = E0cos(ωt). Then the momentum of the elastically scattered electrons will be directed to the left, i.e. an electric current will appear. A similar situation is observed for particles undergoing photoexcitation and recombination in an asymmetric potential well.

As one can see from the example, the fluxes of electrons with opposite momenta are not compensated, instead, the particles reversed in time are compensated.

Therefore, almost any non-centrosymmetry leads to a current.

Observation of the PGE is possible in the crystal without inversion centre, i.e. in those belonging to the following point groups: C1, C2, Cs, C2v, C4, C4v, C3, C3v, C6v, C3h, D3h, Td, D2, D4, D2d, D3, D6, S4, T, O.

In the general case, the PGE can be described by the following constitutive equation [8]:

𝑗_𝑖=𝜒_𝑖𝑘𝑙^(LPGE)𝐸_𝑘𝐸_𝑙^∗+𝑖𝜒_𝑖𝑘^(CPGE)�𝐸×𝐸^∗�

𝑘, (1.1)

where χ^(LPGE) is the tensor of the third rank, χ^(CPGE) is the pseudotensor of the second rank. It can be seen from equation (1.1) that the first and second terms are maximal for the linearly and circular polarized excitation beams, respectively. The second term exists only at elliptical polarized light. Accordingly, the χ^(LPGE) tensor is responsible for the linear photogalvanic effect (LPGE), and χ^(CPGE) is responsible for the circular photogalvanic effect (CPGE). In the general case, CPGE is associated with the conversion of the angular momentum of photons into the translational motion of electrons [8,13]. The nature of LPGE is more complex and associated with the processes of dissipation.

The basis of the LPGE theory was established in Ref. [14]. The mechanism of LPGE can be most clearly demonstrated by a simplified model shown in Figure 1.3 for laser beam polarized along x-axis. If the electron bandstructure of the material has no plane of symmetry perpendicular to the x-axis, each act of photon absorption will be accompanied with generation of an photoelectron having a momentum determined by the excitation conditions, for example, the frequency of light ω. In this case, the momentum distribution of particles will be proportional to

~cos2(ψ) , where ψ the angle between the electron quasimomentum p and electric fields of the light wave E [9,15,16]. Under normal conditions, at a normal incidence of radiation, the momentum of particles moving to the right will be compensated by the momentum of particles moving to the left. However, due to the asymmetric structure of the sample, the excitation of particles moving to the right will be more likely than the excitation of particles moving to the left (Figure 1.3, inset on the right). In addition, recombination of particles moving to the left is more likely than particles moving to the right (Figure 1.3, insert on the left). Thus, due to the asymmetric structure of the sample, the momentum of particles moving to the right is not compensated, and a photocurrent will appear in the sample in the direction of the x axis.

(22)

20

Figure 1.3. Illustration of LPGE using an example of a sample containing asymmetric potential barriers in the energy structure.

In the general case, LPGE is associated with asymmetry due to electron scattering processes in the crystal lattice lacking inversion centre. A simple example may be asymmetry due to elastic collisions of electrons with impurities of the lattice [8]. In addition, asymmetry can be caused by the scattering of electrons and holes by phonons, excitons, and other disturbances of the lattice periodicity.

There are two mechanisms of LPGE. In the first mechanism, the photocurrent is generated due to the absorption of light on free charge carriers, followed by their scattering by phonons. This is ballistic mechanism of the LPGE. The second mechanism is caused by the displacement of charge carriers in real space during quantum transitions. This is the shift mechanism of the LPGE. Both mechanisms are considered in more detail in Ref. [17].

For the first time, LPGE was discovered on tellurium crystals in the study of optical rectification [18]. Pulse laser radiation with a wavelength of 10.6 μm and a peak power of 3.5 kW was directed to the crystal. The pulse duration was 200 ns.

For experiments, the irradiated ends of the crystal were treated with a solution of chromium trioxide in hydrochloric acid, then washed in hydrochloric acid, and then deionized in water. The photovoltage was measured in two mutually perpendicular directions, perpendicular to the direction of light propagation. The crystal was mounted so that it could rotate 360° around the axis of the laser beam.

The registered photo emf showed a strong dependence on the polarization of light, and the authors identified it as a manifestation of optical rectification. However, it was later established that the photovoltage in tellurium can be constant for a long time, therefore, the observed phenomenon cannot be a consequence of the transitional process [8]. For the same reasons, it can be said that LPGE was detected in GaP crystals [19], although the authors of the work also explained the appearance of the photocurrent by the effect of optical rectification.

(23)

21 Very often the crystal symmetry permits simultaneous observation of the LPGE and CPGE. For example, the both effects were observed in bismuth silicate [20], lithium niobate [21], quantum wells p-SiGe [22], InGaAs/AlGaAs [23], InAs/AlGaSb and GaAs/AlGaAs [24,25], and indium nitride films [26], as well as in a two- dimensional electron gas in the MgZnO/ZnO structure [27]. CPGE is a conversion of the angular momentum of photons into the translational motion of free charge carriers. Two microscopic mechanisms of CPGE have been studied: 1) the spin orientation of the carriers by light is accompanied by their directed motion due to the spin-orbit coupling [13]; 2) a purely orbital mechanism, when the photocurrent arises as a result of the interference of various contributions to absorption by free electrons [28]. One can visualize CPGE by considering mechanical systems that convert rotational motion to translational. There are two types of such systems, the first type contains a screw mechanism, such as in an airplane with a propeller, and the second one contains systems that have contact between a circular and flat surface, for example, a car wheel and a road. The electronic analog of screw systems is described by the diagonal components of the χ^(CPGE) tensor, in this case the motion of the charge carriers will occur parallel to the propagation of the exciting radiation.

Accordingly, the wheel analog is described by non-diagonal components of the χ^(CPGE), in this case the direction of motion of the charge carriers is determined by their ratio. The circular photocurrent is propotional to the degree of circular polarization P which is determined by the angle φ between the plane of incidence and the optical axis of the quarter-wave plate, P = sin(2φ) (see, for example, [13]).

For the first time, CPGE in tellurium was predicted in [29], where the photocurrent was calculated for interband and intraband light absorption. It is interesting that the developed theory permits the inverse effect, i.e. according to [29], an electric current in gyrotropic crystals can result in the emission of circularly polarized light. The experimental discovery of CPGE in tellurium crystals has been reported in [30]. A pulsed CO2 laser with a power of 3 kW and a pulse duration of 100 ns was used as a radiation source. The polarization of radiation was determined by the angle φ between the optical axis of the quarter-wave plate and the initial plane of incidence. The samples were tellurium cylinders 0.8 cm long with a cross section of 3.1·10^–2 cm². To register the photoresponse, ring contacts were applied to the cylinders near the ends. As the theory predicted, at a temperature of 300 K the magnitude of the photovoltage was proportional to the degree of circular polarization of the radiation. Figure 1.4 shows the dependence of the photoresponse on the degree of circular polarization. It can be seen that the photovoltage due to the CPGE varies according to ~sin(2φ) and reaches its maximum values at circular polarization.

(24)

22

Figure 1.4. The dependences of the CPGE photovoltage on the degree of circular polarization from [30] in a tellurium crystal excited by the radiation of a CO2 laser at 1) T = 300 K, 2) T=150 K. Inset shows scheme of experiment.

The dependence of the photocurrent on the external electric field applied to the Pb⁵Ge³O¹¹ crystal was studied in [31]. Like other ferroelectrics, this crystal has the property of switching spontaneous polarization under the influence of an electric field and, thereby, transforming from a left crystal to a right one. In the ferroelectric phase, Pb⁵Ge³O¹¹ belongs to the C³ group, while above the transition temperature this crystal is a representative of the C^3h group. For experiments, a plate cut from Pb⁵Ge³O¹¹ with translucent electrodes was used. The radiation source was a helium- cadmium laser with a wavelength of 440 nm and a power of 5 mW. An important feature of the experiments is that, before measurements, the sample was heated to a temperature above the Curie point of 177° C, and then gradually cooled to room temperature. Immediately after cooling, under laser irradiation, the photocurrent was zero. This was explained by the fact that the crystal domains were oppositely directed, and on average the arising currents were mutually compensated.

However, after exposure of the sample for 10 minutes under the influence of a constant electric field, a photocurrent appeared. The sign of the photocurrent changed at field strength of 16 and -20 kV/cm. The difference in the field strength, as well as the small dark current in the sample was attributed by the authors to a small birefringence of the sample.

CPGE induced by a magnetic field was discovered in Ref. [32]. The experiments were carried out on p-GaAs(Zn) crystals having a parallelepiped shape of 7 × 4 × 1.5 mm in the temperature range 78–300 K. The radiation source was a 5 kW CO2

laser. The radiation hit the sample normally. In the absence of a magnetic field, a photocurrent was practically absent, but when it was turned on a photo-emf pulse was observed. The photo-emf pulse shape almost completely repeated the shape of exciting laser pulse. The amplitude of the detected pulses also depended on the

(25)

23 orientation of the quarter-wave plate as sin(2φ), while the sign of the photo-emf changed when the direction of the external magnetic field reversed. It was shown that the nature of the effect is associated with the processes of dissipation, and although it manifests itself as CPGE, its mechanism is more similar to LPGE and can have ballistic and shift components like LPGE. It is also interesting that this effect was not observed in n-GaAs crystals.

In Ref. [21] the authors paid attention to an important problem in detecting a circular photocurrent associated with spatial oscillations of the photovoltaic current. In most crystals, birefringence occurs when light is propagating through a crystal because the phase difference between the ordinary and extraordinary waves changes periodically leading to oscillations of the photocurrent [8]. An exception is the case when radiation propagates along the optical axis of a uniaxial crystal (crystals of classes D², D³, D⁴, D⁶, C³, C⁴, C⁶) or when circular polarization corresponds to the eigenoscillation mode (crystals of classes T and O). It was usually proposed to use samples in the form of thin films or to measure using a periodic system of electrodes to detect spatially oscillating currents [33]. In photorefractive crystals, the detection of such currents is possible by recording diffraction gratings using waves with orthogonal polarization [34]. Instead, it was proposed to use electrodes located on the surface of the crystal perpendicular to the direction of wave propagation. With sufficiently uniform illumination of the entire space between the electrodes, the photocurrent will be determined only by a thin near-surface region. Such a photocurrent will depend only on the state of polarization of the incident radiation, and there will be no spatial oscillations of the photocurrent.

An experimental verification of this idea was performed in the same work on a lithium niobate crystal LiNbO³ [21]. In the experiments the authors obtained a sinusoidal dependence of the photocurrent on the phase shift between the ordinary and extraordinary waves, which indicated the direct detection of a circular photocurrent in lithium niobate without spatial oscillations.

Noteworthy work is Ref. [35], in which the author, on the basis of a phenomenological analysis, predicts the CPGE in gyrotropic isotropic liquids. In that work, it is considered the photocurrent in isotropic non-centrosymmetric media of the limiting symmetry class ∞∞, to which chiral liquids belong, which arise during the passage of elliptically polarized light through them. The resulting expression for the transverse current density was proportional to sin(2φ).

According to Ref. [8], the presence of such a dependence is one of the main signatures of CPGE. Thus, in that work, chiral liquids are proposed as a new photorefractive material.

It is interesting to note Ref. [36]. A feature of this publication is that experiments where the photocurrent was excited in a sample under the influence of uniaxial deformation were carried out. An InN crystal grown on a sapphire substrate with a GaN buffer layer was used as a sample. In the absence of mechanical stress,

(26)

24

experiments revealed that a photocurrent can be divided into two parts: the current arising due to the spin splitting of energy levels, which can arise due to bulk or structural asymmetry [37]; and the current arising due to the asymmetry of scattering by impurities, called in the work inhomogeneous. As mechanical stress was applied, the photocurrent increased linearly. The authors show that with increasing mechanical stress, the inhomogeneous photocurrent decreases, while the current associated with structural asymmetry increases. This indicates that the greatest contribution to the photocurrent in InN is made by the current due to structural asymmetry. Thus, the spin-dependent photocurrent associated mainly with structural asymmetry was demonstrated in the work. It was shown that the degree of structural asymmetry can be changed, for example, by application of mechanical stress.

Recently much attention has been paid to the study of topological insulators. A topological insulator is a material that is an insulator in volume and is a conductor on the surface due to the spin-split metal states. A number of materials that can be considered topological insulators were predicted in [38]. Topological insulators can have such interesting effects as magnetoelectric polarizability [39], magnetic monopole behavior [40], and CPGE. The theoretical study of CPGE in topological insulators was the subject of papers [41–43]. In addition to theoretical work, an experimental study of CPGE was carried out. An example of such studies can be found in Ref. [44,45], where a Bi²Se³ crystal was used as a topological insulator.

Thus, the observation of the PGE photocurrent is possible in the presence of asymmetry, bulk or structural. It can be observed in bulk non-centrosymmetric crystals or in presence of an external field. Structural asymmetry can be associated with crystal surface (surface photogalvanic effect). Experimentally, the PGE is observed in the form of a photocurrent depending on the polarization of the exciting radiation. Thus, the momentum and angular momentum of photons can be transferred to charge carriers, which leads to the appearance of an electric current, and this has been observed in many different materials, both with linearly and circularly polarized excitation beams.

1.2 SURFACE PHOTOGALVANIC EFFECT

The surface photogalvanic effect (SPGE) manifests itself as generation of the photocurrent in the subsurface layers of semiconductor under irradiation with the intense laser beam. [9,46,47]. The microscopic mechanism of the SPGE can be understood if one considers photoexcitation of the semiconductor with p-polarized laser beam at obliquely incidence. If the photon energy is bigger than the semiconductor bandgap, the probability of interband transitions will be proportional to (𝑬 ⋅ 𝒑)²∝ 𝑐𝑜𝑠²𝜓. Since for the p-polarized laser beam E lies in the XZ plane (see Figure 1.5), the probability of the movement of electrons from the

(27)

25 surface and forward the surface will be the same, i.e. the number of photoexcited electrons moving in towards and outwards the surface will be the same. However , in a subsurface layer with a thickness less than the average mean free path of an electron, the scattering losses of electrons moving out of the surface will be lower than those for electrons moving towards the surface because the surface provides additional scattering channel. If the scattering on the surface is diffuse, i.e. if the x- component of the electron moment changes after reflection from the surface, the electrons moving to the right in Figure 1.5, will lose momentum faster than the electrons moving to the left because they experience scattering only inside the material. Such an imbalance will lead to a directed flow of electrons in the direction opposite to the x axis. Accordingly, the density of the surface longitudinal photocurrent due to SPGE can be estimated from the expression [9]:

𝑗_𝑥^{(𝑆𝑃𝐺𝐸)}∝ 𝐴𝑝𝑜𝑙 𝑒𝐼

ℏ𝜔 𝛬𝛬sin𝛼cos𝛼, (1.2)

where 0 ≤ A^pol ≤ 1 (A^pol = 0 and A^pol = 1 corresponds to specular and diffuse scattering, respectively), 𝛬𝛬 is the mean free path of electrons, ω is the light frequency, e is the electron charge, I is the intensity, α is the angle of radiation incidence. SPGE photocurrent disappears at specular reflection of electrons from the surface and at excitation by s-polarized radiation. Therefore, the SPGE requires the anisotropy of the photoexcited carriers distribution in the momentum space and prevalence of the diffuse scattering of the carriers on the surface. In another words, one may say that the SPGE photocurrent is proportional to the probability of the diffuse scattering of photoexcited carriers.

Figure 1.5. Schematic illustration of SPGE excited by p-polarized radiation. As a result of the interaction of light with matter, electrons acquire momentum with a probability that is proportional to cos(2ψ), where ψ is the angle between the electric field E and the electron quasimomentum p. Electrons directed to the surface lose momentum faster than electrons directed from the surface, their momentum becomes unbalanced and an electric current occurs (adapted from [48]).

(28)

26

An experimental study of SPGE began in Ref. [46],where the SPGE photocurrent studied in epitaxial layers of gallium arsenide n-GaAs grown on semi-insulating substrates under excitation with linearly polarized nanosecond pulses. To record the SPGE photocurrent, parallel indium strips, which acted as contacts, were deposited on the sample surface by annealing. The temperature of the sample during the experiments was 1.6 and 4.2 K. It has been shown that the SPGE photocurrent depends on the polarization of the excitation beam (see Figure 1.6a) as 𝐽 ∝⋅ 𝑐𝑜𝑠(Φ), where Φ is the angle between the plane of incidence and the polarization plane. It had been found [9] that the SPGE photocurrent is an odd function of the incidence angle α (Figure 1.6b).

The theory of the SPGE have been developed in [9]. In particular it has been demonstrated that SPGE is forbidden for s-polarized excitation beam. It is worth noting that in Ref. [9], the photocurrent at the s-polarized excitation beam was non- zero due to inhomogeneity of illumination, non- ideality of sample and contribution other effects such as PDE and PGE. Nevertheless Refs. [9,46] properly describe the dependence of the SPGE photocurrent on the polarization and angle of incidence of the excitation beam.

Figure 1.6. Typical dependences of the SPGE photovoltage on a) the angle of orientation of the plane of polarization from [46], b) the angle of incidence of the exciting radiation from [9].

Inset shows scheme of experiments.

A quantitative description of the SPGE in GaAs have been developed taking into account the structure of bands, momentum scattering mechanisms, and Coulomb interaction in Ref. [9]. The theoretical results well correlated to the results of measurements of the angular, polarization, and spectral characteristics of the photoresponse. It has been found in particular that the spectrum of the SPGE photocurrent have features associated with the partial relaxation of the momentum of electrons and holes upon scattering on the surface. The SPGE photocurrent changes sign and decreases with increasing frequency of the exciting laser pulse,

(29)

27 tends to zero at high frequencies. The authors explained the observed discrepancies between theory and experiment by the influence of the subsurface electric field.

The theory of SPGE have been developed further in Ref. [47], which was devoted to the theoretical analysis of SPGE in metals. The authors considered the possibility of generating a surface photocurrent in metals during interband transitions. It was shown that the SPGE photocurrent in metals is maximum at the maximum surface roughness and depends on the excitation wavelength, temperature, and the ratio of the sample thickness to the electron mean free path. It was shown that in the region of the absorption edge, the SPGE photocurrent depends on the number of charge carriers generated near the Fermi surface;

however, with a further increase in the frequency, the photocurrent should decrease like a damped oscillator. The polarization and orientation dependences of the SPGE photocurrent should be determined by the crystal lattice of the sample.

According to theoretical estimates [47], the SPGE photocurrent should be of the order of 10^–6 A·cm/W, while in pure metals at low temperatures, the SPGE photocurrent can be as high as 0.1 A·cm/W in the vicinity of the absorption edge.

However, the developed theory assumed the spherical Fermi surface, i.e. its predictions may be valid only for alkali and noble (gold, copper and silver) metals.

SPGE in copper have been observed in Ref. [15]. For this purpose, a 5 × 5 × 1 mm sample was cut from a copper crystal with a purity of 99.999%, one of the sides of which was polished. SPGE was observed by measuring the magnetic flux appearing in the sample with short-circuited lead electrodes (a superconducting quantum interferometer was used). To this end, the coils with the aid of which the SPGE photocurrent was measured, were located near the sample. An argon laser (radiation wavelength λ = 514.5 nm) was used as a light source, the radiation of which passed through a quartz window into a chamber with a sample cooled to a temperature of 4.2 K. The angle between the wave vector and the normal to the surface of the sample was 30°.

The experiments in Ref. [15] showed that the SPGE photocurrent directed parallel to plane of light incidence changes according to the law ~cos(2Φ), which corresponds to the results obtained in Ref. [9,46]. A polarization-independent additive to the photocurrent was also present, but in this case it is provided by the theory described in the same work. The photocurrent flowing in the direction perpendicular to the plane of incidence depended on the polarization as ~sin(2Φ);

according to the developed theory, a polarization-independent background should not appear, although it was observed experimentally. Such a discrepancy between theory and experiment can be attributed to the fact that the laser radiation incident on the sample was slightly elliptically polarized due to local mechanical stresses in the quartz window, resulting of low temperatures of up to 4.2 K.

Thus, in the above studies, it was experimentally and theoretically shown that SPGE is a polarization-orientationally dependent effect. This effect is inherent to all solids of sufficiently large thickness. In thin films, photoexcited charge carriers

(30)

28

reach the opposite surface of the film, i.e. two subsurface currents compensate one another leading to zero photoresponse. It is worth noting that the SPGE studies are rather sparse because it usually accompanied with the PDE and PGE (see, for example, [15,49]).

1.3 PHOTON DRAG EFFECT

Similarly to the PGE and SPGE, the photon drag effect (PDE) is also manifested as the generation of an electric current under irradiation with a light pulse. However, the mechanism of the PDE differs significantly from the PGE. Specifically, the PDE originates from the momentum transfer from the incident photons to free conduction electrons [10,50]. In semiconductors, the PDE is associated with the momentum-selective interband transitions and with different carrier mobility at the corresponding levels [51–53].

The PDE is the second-order nonlinear optical phenomenon that can be phenomenologically described by the following constitutive equation (see, for example, [54]):

𝑗_𝑖^{(𝑃𝐷𝐸)}∝ 𝜒_{𝑖𝑘𝑙𝑚}^{(𝑃𝐷𝐸)}𝑘𝑘𝐸𝑙𝐸𝑚∗, (1.3)

where χ^(PDE) is the tensor of the fourth rank; k is the wave vector; E is the vector of the electric field of the light wave, k, l, m are the indices corresponding to the selected coordinate system.

In the isotropic medium the PDE photocurrent can be estimated in terms of light pressure according to the formula [7]:

𝑗^{(𝑃𝐷𝐸)}∝ 𝑒𝑡𝑝ℏ𝑘 𝑀^∗

𝐴𝑎𝑏𝑠𝐼

ℏ𝜔 , (1.4)

where tp is the electron relaxation time, M* is the effective mass of the electron, and A^abs is the relative absorption. It is worth noting that a similar expression can be obtained from the equations of conservation of energy and momentum [7].

The electron drag by photons has first described in Ref. [55]. However, the developed model is applicable only if the photon energy is much higher than the temperature of the crystal lattice. The transitions between two subbands have been also considered.

Ref. [10] reports PDE in p-type germanium excited by a pulsed CO² laser with a peak power of 2 kW. An interesting feature of the photocurrent is a change of the current sign when the temperature changes from room temperature to the temperature of liquid nitrogen. The obtained experimental regularities were well described by the developed theory. The PDE in germanium was also investigated in Ref. [50], which also showed that the CO2 laser radiation can transmit a sufficient pulse to produce a photocurrent in a four-centimeter rectangular rod. Studies of PDE in germanium continued in Refs. [54,56–58].

(31)

29 PDE has also been studied in other materials. For example, the theory of the PDE photocurrent at optical interband transitions in tellurium has been developed in Ref. [59]. The theory predicts the temperature dependence of the PDE photocurrent. The experimental study of PDE in tellurium [60] involved complex measurements were performed to calculate and interpret all components of the χ^(PDE) tensor in a tellurium single crystal under irradiation with the CO2 laser. Three mechanisms of the appearance of the photocurrent were identified in addition to PDE including optical rectification and structural imperfections of the crystal. The experimental results obtained were consistent with the theory of PDE.

Experimental studies of PDE were also carried out in bismuth [61,62], two- dimensional electron gas [51,63–66], gallium arsenide [9,46,67,68], ZnTe single crystals [69], silver [70] and gold nanostructures [70–74], graphene [75]. As a result of experimental and theoretical studies, the properties of the PDE photocurrent in 2D materials have been studied. In particular, it was shown that the PDE photocurrent linearly depends on the intensity of the exciting radiation (see, for example, [66]). It has been also found that the dependence of the photocurrent on the angle of incidence is an odd function (for example, [70]). Experiments had shown that the transverse photocurrent has sin(2Φ) dependence on the polarization plane azimuth Φ of the excitation beam (see, for example, [54,60]). One can see that other mechanisms (e.g. SPGE and PGE) show similar polarization dependences of photocurrent.

In addition to the momentum of photons, their angular momentum can also be transferred to charge carriers. This manifests itself in the dependence of the photocurrent on degree of circular polarization of the excitation beam and often referred to as the circular photon drag effect (CPDE). One of the first works devoted to the study of the CPDE was [76], where it was pointed out that it is possible to generate a CPDE photocurrent in a tellurium crystal at a CO² laser irradiation. It is worth noting that in the classical approximation, the CPDE can be considered as the Hall effect in the field of a light electromagnetic wave [49]. Phenomenologically, the CPDE photocurrent can be described by the following constitutive equation [77]:

𝑗_𝑖^{(𝑃𝐷𝐸)}∝ 𝑖𝜒_𝑖𝑘𝑙^{(𝑃𝐷𝐸)}[𝐸𝐸^∗]_𝑙𝑘_𝑘, (1.5)

where χ^(CPDE) is the pseudotensor of the third rank, which is odd under time inversion. Numerous experimental and theoretical studies have shown that the CPDE photocurrent is directly proportional to the degree of circular polarization.

For example, in Ref. [78], the observation of a circular photocurrent in graphene was reported. A photocurrent was generated in a graphene monolayer upon oblique irradiation with a CO² laser. The polarization was controlled by a rotating a quarter-waveplate, and the longitudinal and transverse photocurrents were measured. Figure 1.7 shows the obtained dependences of the photocurrent on the angle of incidence of radiation. It can be seen that the photocurrent dependences are odd functions of the incident angle, i.e. at normal incidence, the photocurrent is equal to zero. The dependences of the CPGE on angle φ are presented in Figure 1.8.

(32)

30

Figure 1.7. Typical dependences of the a) transverse and b) longitudinal PDE photocurrents on the angle of incidence of the exciting radiation. The dependences were obtained in Ref [78] on graphene layers excited by the radiation of a CO2 laser.

Figure 1.8. Typical dependences of the longitudinal and transverse PDE photocurrent on angle φ. The dependences were obtained in Ref. [78] on graphene layers excited by the radiation of a CO2 laser.

In Ref. [78] the experimental dependence of the transverse photocurrent on rotation angle φ of the quarter-wave plate can be described as

𝐽𝑦=𝐽𝑦,𝑐𝑖𝑟𝑐sin 2𝜑+𝐽𝑦,𝑙𝑖𝑛sin 4𝜑+𝐽𝑦,𝑐𝑜𝑛𝑠, (1.6) where J^y,circ and J^y,lin are responsible for the circular and linear contributions, respectively, J^y,const is the constant component of the photocurrent, which was

(33)

31 observed in some measurements. The longitudinal photocurrent is described by the following equation:

𝐽_𝑥=𝐽_{𝑥,𝑐𝑖𝑟𝑐}cos 4𝜑+𝐽_{𝑥,𝑐𝑜𝑛𝑠}. (1.6) Based on the results obtained, the authors conclude that the photocurrent in graphene is due to the circular Hall effect (CPDE). In addition, it was found that in graphene on a SiC substrate the photocurrent exhibits resonance at frequencies that coincide with longitudinal optical phonons in SiC. It should be noted that PDE can also result in the shown in Figure 1.8 dependences, which are typical for the photocurrent having both linear and circular contributions.

The Ref. [79] is devoted to the study of CPDE in quantum wells grown in zinc blende in the (110) direction. The generation of the photocurrent was caused by transitions between subbands induced by infrared and terahertz radiation. It is reported that the circular photocurrent at normal incidence was due to CPGE, but at an oblique incidence, it changed sign, which indicates the generation of a circular photocurrent by the mechanisms of CPGE and CPDE. In addition to the experiments, in Ref. [79] a microscopic theory was developed based on the sensitivity of the spin orientation to the wave vector and the subsequent asymmetric spin relaxation.

In addition, the possibility of generating a CPDE photocurrent in bulk tellurium [11], two-dimensional metal photonic crystals [80], and nanoporous gold films [81]

has been demonstrated. The angular and polarization dependences of the PDE photocurrent obey the same regularities as the SPGE and PGE photocurrents.

1.4 CONCLUSIONS TO CHAPTER 1

Thus, PGE, SPGE, and PDE result in the generation of the photocurrent. While the mechanisms of the PGE, SPGE, and PDE photocurrents are significantly different, the polarization and angular dependences of the photocurrent for each mechanism are difficult to distinguish. Therefore, in order to reveal mechanism of the photoresponse of the Ag/Pd films a set of experiments is required. One of the most important tasks is to study the phase composition and properties of the studied Ag/Pd films. A significant role in the generation of the photocurrent under the indicated mechanisms is played by quantum transitions between energy levels;

accordingly, it is necessary to study the angular and polarization dependences of the photocurrent in Ag/Pd films with different phase compositions in a wide spectral range.

(34)

32

2 FABRICATION AND THE PROPERTIES OF SILVER-PALLADIUM FILMS

The chapter is devoted to the fabrication and study of the physical properties of Ag/Pd films. The fabrication technology of Ag/Pd films by resistive paste annealing is described. The results of thermodynamic calculations of the equilibrium composition formed in the Pd – Ag2O system in air during annealing, i.e. in the functional component of the initial paste are presented. The chapter also contains the results of studies of the films by scanning electron microscopy, the results of a study of the electrical properties of the obtained films, as well as study of the films phase composition by X-ray diffractometry and photoelectron spectroscopy. In addition, the relative content of PdO in Ag/Pd films obtained at different temperatures was estimated by Raman spectroscopy.

2.1 SILVER-PALLADIUM FILMS

2.1.1 Using of Ag/Pd films in microelectronics

The Ag/Pd films studied in this work are used in microelectronics for fabrication of thick-film resistors, which are widely used in electronic circuits. Film resistors are made by coating an insulating substrate with a resistive layer.

Thin film resistors are made by spraying a conductive material onto a dielectric substrate. The thickness of the conductor and its conductivity determine the resulting resistance. The thin-film resistors are affected by high temperatures, their resistivity may change with time due to oxidation of the resistive layer. The smaller the thickness, the less stable thin-film resistors parameters, which are also sensitive to electrostatic discharge and environment [82].

Thick-film resistors, which have a thickness larger than ten micrometers, are made by applying resistive pastes on the dielectric substrate, followed by drying and annealing. Thick-film resistors have a wider range of possible resistance values that makes them more versatile and cheaper than thin-films ones. The resistance of a thick film resistor depends not only on the cross-sectional area of the conductive layer, but also on the composition of the initial resistive paste. The resistive track of the thick-film component is formed by particles of a conductive material in a glassy matrix. The heating of the resistor by electric current may results in weaken contacts between conductive particles in the matrix and hence in increasing of the resistance. However, the glass coating that forms in the process of annealing makes thick-film resistors are more stable to environment impacts. Thick-film resistors are used in almost all types of devices [82].

(35)

33 2.1.2 Thick Film Technology

Fabrication of the Ag/Pd films studied in this Thesis was performed by using the widely known technology, which was developed for fabrication of passive elements of hybrid integrated circuits (HIC). HICs consist of passive elements (resistors, capacitors, etc.) and miniature open-frame active devices (transistors, diodes, etc.) situated on a common dielectric substrate. The HIC fabrication, which is described in a variety of sources, for example [83–85], is based on the formation of more than ten micrometers thick layers of different materials on the surface of the dielectric substrate, which are used as conductive paths, contact pads, resistors, capacitors, and so on.

HIC dielectric substrates are used also for heat removal and can be made from glass, sitall or ceramics. If it is necessary to remove a large amount of heat, metal substrates coated with a dielectric are used. As a rule, ceramics are used as the basis for thick-film HIC because of its mechanical strength, thermal conductivity, thermal and chemical durability.

The fabrication of HIC begins with the formation of passive elements on the surface of the dielectric substrate. First a conductive paste is applied through a mesh stencil having open and closed areas. The composition of the paste depends on the type and characteristics of the elements to be formed. As a rule, paste contains the functional, structural and technological components. The functional component is responsible for the electrical characteristics of the fabricated element.

For fabrication of thick-film resistors, the metals Ag, Au, Pt, Pd, In, Os, Ro, alloys Pt-Au, Ag-Pd, Pd-Au, and multicomponent Pd-PdO-Ag systems are usually used.

The structural component that determines the mechanical properties of the element in most cases is a finely dispersed glass powder. The technological component, a mixture of organic substances and solvent, provides the necessary viscosity and surface tension to allow the paste to easily penetrate through the stencils and to remain on the substrate without spreading. The technological component is evaporated and completely vanishes in the process of annealing.

Application of the paste can be done by contact and non-contact methods. In the non-contact method, the paste is applied on the substrate, which is placed on some distance from the mesh stencil. Moving the squeegee, the paste through the holes in the stencil is transferred to the substrate copying the holes pattern of the stencil.

Print quality depends on the speed of movement and pressure of the squeegee, the size of the gap, the tension of the stencil and the properties of the paste.

In the contact stencil printing method, the stencil is placed on the substrate without a gap (Figure 2.1). The separation of the substrate from the stencil is carried out by vertical movement without slipping to avoid smearing the print of the paste.

The contact method allows to apply the paste by spraying using a spray gun. The accuracy of the print with the contact method is higher than with the contactless.

(36)

34

Figure 2.1. Scheme of the contact method of applying the paste.

After applying the paste, it is dried and annealed. Drying is necessary to remove volatile components (solvent) from the paste at a temperature of from 80 to 150° C for 10-15 minutes. IR radiation heats entire paste layer providing uniform drying without the formation of a crust on the surface.

Annealing is carried out in conveyor-type furnaces of continuous operation with a gradual increase in temperature to a maximum, and subsequent gradual cooling.

Furnace may contain infrared dryer to combine these operations.

At the beginning of the annealing, the organic binder burns out when the temperature does not exceed 400° C. In the temperature range from 500 to 1000° C, the particles of the conductive materials melt together to form conductive bridges and sinter them with glass and the substrate. The rate of temperature change also plays an important role. For HIC in which passive elements are arranged in several layers, it is necessary to select pastes so that the annealing temperature of each subsequent layer is lower than that of the previous one in order to avoid changing its parameters.

After heat treatment, a protective coating of glass with a low softening temperature is formed on the resulting board. After applying and annealing all the layers of the passive part of the circuit, the film elements are adjusted, the active components are mounted, the terminals are installed and all elements are sealed.

2.1.3 Thermodynamic modeling of the equilibrium composition of the func- tional component of Ag/Pd films

The Ag/Pd films studied in this work were formed on a alumina-ceramic substrate using rectangular stencils. For the production of films, we used pastes PR5 and PR50, which can be used to produce a resistive film with a surface resistance of 5 and 50 Ohm/□, respectively. The composition of these pastes is given in Table 2.1.

Since, the technological component of the paste completely burns out at high

(37)

35 temperature, in Table 2.1, only functional and structural components are shown as 100%. Let the annealing temperature Tann will be the maximum temperature reached in the furnace during the heat treatment of the film. The equilibrium composition formed in the Pd-Ag²O system in air, that is, in the functional component of the initial paste was evaluated. To this end the method described in Ref. [86] was used. Thermodynamic calculations were carried out according to the algorithm [87]. The equilibrium compositions formed in the system under consideration were estimated for the temperature range T^ann = 473–1173 K at an atmospheric pressure of 0.1 MPa and the composition of the gas phase, approximately corresponding to the composition of the air — N54O14. The content of the system components was described in terms of the mass fraction of palladium, g = M¹/(M¹+M²), where M¹ and M² are the masses of Pd and Ag²O, respectively. The data on thermodynamic properties of components formed in the system, given in the handbook [88] and contained in the ASTRA program database were used. It is known that annealing of pastes based on Ag and Pd is usually carried out in the temperature range from 773 to 973 K; therefore, for calculating the equilibrium composition in the functional component in air, the temperature was taken equal to T^ann = 873 K.

Figure 2.2. Dependences of the equilibrium contents of the components (1 - Ag, 2 - Pd, 3 - PdO, 4 - Ag2O) a) on the mass fraction of palladium g in the initial paste at T = 873 K and b) on the temperature T for g corresponding to the experimental compositions PR50 (adapted from [89]).

Figure 2.2a presents the results of calculations of the equilibrium content of Gⁱ components in mole fractions, where the indices i = 1, 2, 3, 4 correspond to Ag, Pd, PdO, Ag2O in the Pd-Ag2O-air system depending on the mass fraction of palladium g in the initial mixture (in the functional component of the paste) at Tann = 873 K.

From this figure it is clearly seen that in a wide range of variation of g, the content of metallic silver G¹ decreases linearly (Figure 2.2a, curve 1). At the same time, an

(38)

36

increase in the molar fraction of metallic palladium in the G² film occurs stepwise with a sharp increase at g ~ 0.6 and g ~ 0.8 (Figure 2.2a, curve 2). With an increase in the Pd content in the initial mixture, G³ (the content of PdO oxide) also increases, reaching a maximum at g = 0.56 (Figure 2.2a, curve 3). The indicated maximum content of palladium oxide in the equilibrium mixture corresponds to the composition of paste PR50. The second maximum of the PdO content is observed at g = 0.75, and with a further increase in the fraction of Pd in the initial mixture, the content of PdO in the obtained films decreases. Thus, the dependence of G³ on g has a complex form with two maxima at g ~ 0.56 and g ~ 0.75. The equilibrium content of Ag2O (Figure 2.2a, curve 4) is insignificant, and for g > 0.6 its mass fraction does not exceed 0.001.

Table 2.1. Composition of PR5 and PR50 pastes Name of

the paste

The composition of the functional component, (wt. %)

The composition of the structural component, (wt. %)

The composition of the technological component, (wt. %)

PR5 Ag2O (36)

Pd (15.3) glass (48,7) organic binder (25) PR50 Ag2O (24.2 - 22.22)

Pd (30.8 - 22.56) glass (45 - 55.21) organic binder (25)

The equilibrium content of the components was also found for other temperatures in the range T^ann = 473–1173 K. It was shown that with increasing temperature the PdO content decreases at all concentrations g. Figure 2.2b shows the temperature dependences of the content of components in the Pd-Ag2O system corresponding to the experimental compositions of PR50. It can be seen from the figure that, with increasing temperature, the content of palladium metal in the films increases due to the dissociation of palladium oxide. Silver oxide behaves similarly, the decomposition of which with increasing temperature leads to an increase in the content of metallic silver. It should be noted that the established regularities are characteristic not only for compositions PR50 and PR5, but also for compositions with a different ratio of the starting components g.

To assess the influence of structural and technological components on the composition of the films, a thermodynamic analysis was carried out in which the composition of the paste took into account not only functional components (silver oxide Ag²O and palladium Pd), but also glass and organic binder components. It was shown that the constitutions of the structural and technological component do not affect the patterns of transformations in the functional component. However, new phase components appear in the film. For example, at a annealing temperature T^ann = 878 K, as a result of the decomposition of hydrocarbons, pure carbon should be expected in the film, and the dissociation of lead oxide in the glass will result in the formation of pure lead. In addition, the formation of silicates Al2SiO5 and, below 473 K, lead aluminate PbAl2O4 is possible.

Polarization-sensitive photoresponse of metal-semiconductor nanocomposite film

uef.fi

Dissertations in Forestry and Natural Sciences

ALEKSANDR SAUSHIN

Polarization-sensitive photoresponse of metal-semiconductor

nanocomposite film

ALEKSANDR SAUSHIN

POLARIZATION-SENSITIVE PHOTORE- SPONSE OF METAL-SEMICONDUCTOR

NANOCOMPOSITE FILM

Aleksandr Saushin

POLARIZATION-SENSITIVE PHOTORE- SPONSE OF METAL-SEMICONDUCTOR

NANOCOMPOSITE FILM

ABSTRACT

ACKNOWLEDGEMENTS

LIST OF ABBREVIATIONS

LIST OF ORIGINAL PUBLICATIONS

AUTRHOR’S CONTRIBUTION

CONTENTS

1 THE MAIN MECHANISMS OF THE POLARI- ZATION SENSITIVE PHOTORESPONSE

1.1 PHOTOGALVANIC EFFECT

1.2 SURFACE PHOTOGALVANIC EFFECT

1.3 PHOTON DRAG EFFECT

1.4 CONCLUSIONS TO CHAPTER 1

2 FABRICATION AND THE PROPERTIES OF SILVER-PALLADIUM FILMS

2.1 SILVER-PALLADIUM FILMS