Kelvin probe force microscopy of the nanoscale electrical surface potential barrier of metal/semiconductor interfaces in ambient atmosphere

  1. Petr Knotek1ORCID Logo,
  2. Tomáš Plecháček2ORCID Logo,
  3. Jan Smolík1ORCID Logo,
  4. Petr Kutálek2,§,
  5. Filip Dvořák3ORCID Logo,
  6. Milan Vlček4,
  7. Jiří Navrátil5 and
  8. Čestmír Drašar5ORCID Logo

1University of Pardubice, Faculty of Chemical Technology, Department of General and Inorganic Chemistry, Studentská 573, 532 10 Pardubice, Czech Republic
2University of Pardubice, Faculty of Chemical Technology, Joint Laboratory of Solid State Chemistry, Studentská 84, 532 10 Pardubice, Czech Republic
3University of Pardubice, Faculty of Chemical Technology, Center of Materials and Nanotechnologies, Studentská 95, 532 10, Pardubice, Czech Republic

4Institute of Macromolecular Chemistry of the Academy of Sciences of the Czech Republic v.v.i., Heyrovsky sq. 2, 12006 Prague, Czech Republic
5University of Pardubice, Faculty of Chemical Technology, Institute of Applied Physics and Mathematics, Studentská 95, 532 10, Pardubice, Czech Republic

  1. Corresponding author email

§ Tel.: +420 466036155; fax: +420 466036011

Associate Editor: T. Glatzel
Beilstein J. Nanotechnol. 2019, 10, 1401–1411.
Received 04 Apr 2019, Accepted 18 Jun 2019, Published 15 Jul 2019


This study deals with the preparation and characterization of metallic nanoinclusions on the surface of semiconducting Bi2Se3 that could be used for an enhancement of the efficiency of thermoelectric materials. We used Au forming a 1D alloy through diffusion (point nanoinclusion) and Mo forming thermodynamically stable layered MoSe2 nanosheets through the reaction with the Bi2Se3. The Schottky barrier formed by the 1D and 2D nanoinclusions was characterized by means of atomic force microscopy (AFM). We used Kelvin probe force microscopy (KPFM) in ambient atmosphere at the nanoscale and compared the results to those of ultraviolet photoelectron spectroscopy (UPS) in UHV at the macroscale. The existence of the Schottky barrier was demonstrated at +120 meV for the Mo layer and −80 meV for the Au layer reflecting the formation of MoSe2 and Au/Bi2Se3 alloy, respectively. The results of both methods (KPFM and UPS) were in good agreement. We revealed that long-time exposure (tens of seconds) to the electrical field leads to deep oxidation and the formation of perturbations greater than 1 µm in height, which hinder the IV measurements.

Keywords: Kelvin probe atomic force microscope; nanoinclusion; Schottky barrier; thermoelectric materials; work function


Increasing energy demand and the negative effects of current energy technologies on the environment lead to increased interest in alternative energy technologies. Thermoelectric (TE) devices, utilizing TE phenomena, i.e., the Seebeck and the Peltier effect, can be considered solid-state heat engines. The former enables direct waste heat conversion into electrical energy very elegantly (no moving parts, long-term stability, maintenance-free, silent). Unfortunately, one of the limiting factors of this technology is still poor conversion efficiency. The conversion efficiency of the material could be expressed in terms of figure-of-merit, ZT, defined as dimensionless quantity ZT = S2·σ·T/κ , where S is the thermopower (Seebeck coefficient), σ is the electrical conductivity, T is the absolute temperature and κ is the thermal conductivity. There were many concepts for enhancement of ZT values, e.g., advanced TE bulk compounds of more complex crystal structures [1], low-dimensional material systems [2] or nanostructured advanced bulk materials [3]. Recently the concept of multi-phase nanocomposites (NCs), i.e., an incorporation of second-phase nanoinclusions/nanoparticals (NIs/NPs) into the bulk semiconducting matrices has been proposed [4-6] and supported in theoretical works [7-9]. Such NIs or NPs are able to enhance ZT via a reduction of the thermal conductivity by phonon scattering [10-14], by modulated carrier doping or by the carrier energy filtering effect [15-17]. The potential energy barrier connected with a metallic NP (Schottky barrier) or a multi-phase interface could scatter low-energy electrons more effectively than high-energy electrons [18]. This, in turn, results in an enhancement of the Seebeck coefficient with virtually no harm to other transport parameters.

The characterization of NIs or NPs in TE materials is realized most frequently by the different modes of atomic force microscopy (AFM): i) by comparing the conductivity/resistivity (CAFM) or IV curves measurement in the direct-contact of the conductive tip and the material [19-21]; ii) by mapping of the different surface contact potential values by Kelvin probe force microscopy (KPFM) in the semicontact mode [19,22-25], or iii) by measuring the differences in thermal conductivity by scanning thermal microscopy (SThM) [19,20,26]. Shape, size, homogeneity of distribution and chemical composition as the basic characteristics are studied by using electron microscopy (SEM+TEM), although there is a high probability that the NPs diffuse into the TE matrix due to the e-beam interaction. Hence, these characterizations are also performed by using scanning tunneling microscopy (STM) [27,28] or by using AFM in the semicontact mode. The latter enables a describtion not only of the topography (size and shape) but also a detection of the changes in density, stiffness and adhesion of NPs [20,21,24,29,30].

In the present study we demonstrate that the Schottky barrier (surface contact potential) value and the polarity can be controlled by the barrier-forming metal NPs (Au, Mo) and can reflect their different chemical behavior with the Bi2Se3 matrix. These metals were selected due to the different interaction with the matrix, as Au can diffuse to the Bi2Se3 forming Au/Bi2Se3 alloy and NIs in the shape of a semispherical defect, whereas Mo reacts with the matrix and forms thermodynamically more stable layered MoSe2 in the shape of nanosheets.


Single crystalline Bi2Se3 samples were grown by heating stoichiometric mixtures of the pure elements obtained from Sigma-Aldrich, i.e., 5N Bi and 5N Se. The crystal growth was performed by cooling in a horizontal furnace from 1073 K to 823 K at a rate of 6 K per hour. The crystals were then annealed at 823 K for 350 h and quenched in air. This free-melt crystallization (FMC) procedure produces single crystals of 10–20 mm in length, 3–6 mm in width and up to 3 mm in thickness [31]. The surface of freshly cleaved layered Bi2Se3 single crystals was used as a substrate. The samples for the AFM measurement, Au nanoparticles (NPs) and thin films of gold or molybdenum, were prepared via DC sputtering in a SEM Coating System (Bio-Rad) in Ar atmosphere (p ≈ 20 Pa, I = 18 mA, U = 1.4 kV) from pure metal sheets (Mo 4N, Au 4N). For preparing separated Au nanoparticles under the same sputtering conditions, however, a stainless steel mask (system of 100 × 500 μm2 holes, 200 holes/cm2) between the substrate and target had to be used to decrease the plasma intensity.

The topography, phase shift image, Kelvin probe force microscopy and IV characteristics were measured by the AFM SolverPro M, Nt-MDT (Russia) with a resolution of 512 × 512 pixels. The HA_NC tips (resonant frequency 140 kHz, force constant 3.5 N/m) were used for measuring the metal layer thickness by the scratch method and phase contrast [30,32], while conductive NSG01/TiN tips (150 kHz, 5.1 N/m) were used for Kelvin probe force measurements and contact CSG-10/Pt tips (22 kHz, 0.11 N/m) for recording IV characteristics [33,34] for 10 replicas from an identical point without delay; see the corresponding references above for other experimental details. The set-point was stabilized in the range of 40–50% and the scanning frequency of 0.5 Hz was used, if not mentioned otherwise. A separation of the topographical signal and the VCPD (the contact potential difference) measurement was achieved by the modulation of the VAC at a frequency higher than the bandwidth of the topography feedback system. The topography was measured by the oscillation at the first resonance frequency of the AFM tip, and VCPD was measured by the amplitude of the oscillation at the second resonance frequency of the AFM tip [35,36]. It also has to be noted that all measurements were carried out at room temperature.

Photodiffusion of the metal into Bi2Se3 was enabled by means of a Ronchi ruling (non-transparent Cr lines on the SiO2 substrate) with a density of 2000 lines per millimeter (period 500 nm) as a photomask. The material was illuminated with a UV light source LC08 (Hamamatsu) emitting at wavelengths of 310 and 360 nm; total intensity 0.8 W/cm2 for 60 s [37,38]. The e-beam diffusion was realized by a SEM JEOL JSM 5500-LV in an of area 8 × 10 µm2 with an acceleration voltage of 20 kV (which results in a penetration depth of the electrons of 1 µm into Bi2Se3) for a period of maximally 360 s.

The work function (WF) was determined by utraviolet photoelectron spectroscopy (UPS) using a helium gas discharge source with He I radiation (hν = 21.22 eV). All UPS measurements were performed in an UHV apparatus (ESCA 2SR, Scienta-Omicron) with a base pressure below 1 × 10−9 mbar. The spectra were calibrated using the Fermi edge of sputter-cleaned Au as reference. During the measurements, a bias of −10 V was applied to the sample in order to cut off secondary electrons generated in the analyzer. The work function of the sample was calculated as WF = hν Ecut-off, where Ecut-off was determined from the intersection of the linear extrapolation of the secondary-electron cut-off (SECO) with the background. All samples were sputtered with argon ions using a scanning focused ion beam source in order to remove surface contaminants. A monoatomic argon ion source was utilized with energy of 2 keV, ion current 10 µA, raster area 1 × 1 mm2 and sputtering time 30 s.

Results and Discussion

Separated metal nanoparticles on the substrate

In TE materials the NIs applicable for an increase of their profitability/efficiency should be present as 1D materials or 2D layers. The aim of the characterization of the NIs is the proof of their presence, the description of their shape and of their electrical behavior in contrast to the unaffected matrix material. We used Bi2Se3 as a standard material because of its well-defined layered structure with sub-nanometer roughness similar to mica or highly oriented pyrolytic graphite (HOPG). The electrical conductivity and mechanical stiffness of Bi2Se3 allow for the measurement with high-intensity electrical fields (10 V/30 nm) without damaging the topmost layer.

The Au particles were dc-sputtered onto the Bi2Se3 substrate and the surface was immediately measured by means of AFM. The particles exhibited a height ranging from 5 to 70 nm. The height distribution (with the maximal population at ca. 40 nm) and the FWHM did not change within 72 h after preparation (Figure 1A and Figure 1D). The stability of the histograms over 3 days led us to believe that during this time interval there is only negligible diffusion of the Au NPs into the Bi2Se3 substrate. The material of the particles was mechanically different as observed on the phase shift image (Figure 1B) due to local changes in adhesion, density, and stiffness [30]. This fact demonstrates the difference between the NPs and the substrate, regardless of any artifacts of the sample preparation/measurement.


Figure 1: (A) Topographical and (B) phase-shift image, and (C) surface contact potential map of Au nanoparticles on Bi2Se3 (as prepared); (D) histogram of the Au NP height distribution for the as-prepared sample and after 72 h.

KPFM is a double-pass measurement technique, where the lift height between the first (topographical) and second (electrical) pass is an important parameter for the spatial resolution [36,39,40]. The tested lift heights were 10, 30 and 50 nm, and the optimal value was found to be 30 nm (Figure 1C). The KPFM image has a good contrast for a lift height of 30 nm and particles with a height less than 10 nm were detected in the map of the surface contact potential. A higher lift height led to a vanishing of the contrast and a lower value resulted in a decrease in reproducibility (data not shown). The results are in a good agreement with the KPFM contrast of nanodiamonds on a Si substrate, where the contrast even changed for the lower lift height (see Figure 10.2 in [41]). The atypical conductive TiN coating of the tip (instead of Pt, Au Pt/Ir or Pt/Cr tips [33,41-43]) was used because of the higher mechanical stability of the tips and in order to eliminate diffusion and other chemical interactions of the tip coating with the samples. The measured difference of the surface contact potential (Figure 1C) between Au particles and Bi2Se3 was not identical for all particles depended on the height of the particles. The height, i.e., the vertical distance between the top of the NP and the substrate and the contrast of the surface contact potential of Au particles (Figure 2A, B) were measured within the same line profile. Au NPs of greater height showed lower values of the surface contact potential according to the non-linear curve limited to a Δ surface contact potential between −50 and −60 mV. This observation can be attributed to quantum size effects [36,44], which are accompanied by a change in the charge transfer from substrate to NPs [45,46].


Figure 2: (A, B) Determination of the surface contact potential decrease and the height of the Au NPs; (C) surface contact potential as a function of the Au NP height.

Metal layers on Bi2Se3

The dependence of the surface contact potential on the size of the separated Au nanoparticles theoretically allows for an optimization of the efficiency of TE materials. Our efforts to prepare separated Mo NPs on the Bi2Se3 surface was not successful. Because Mo has significantly higher melting/boiling points (bulk Mo 2623/4639 °C, bulk Au 1064/2856 °C [47]) dc sputtering yielded homogeneous layers of Mo instead of separated NPs. In addition, the preparation from colloidal metallic NPs failed as the solvents or stabilizing agents as polyphosphates or vinylpyrrolidones [48] interacted with the uppermost Bi2Se3 layer, or residua of the organic solvents that were not evaporated under the vacuum changed the electrical behavior of the nanocomposite.

An alternative way to study Schottky barriers at the nanoscale is to prepare a metallic layer on the Bi2Se3 surface and characterize such a nanocomposite. IV spectroscopy is typically used for the determination of Schottky barriers [25,49,50]. The setup was tested on bulk Au and on a 107 nm Au layer on the Bi2Se3 (for both cases the contact electrode and conductive tip were located on the same material). We observed a higher conductivity of the bulk material (quantum size effect) (Figure 3, compare curves a) and b)) and a small photodiode effect of Au/Bi2Se3 (−100 mV) due to the AFM illumination (Figure 3, see the intersection of curve b with the x-axis). We also measured Au layers with a thickness of 53 nm and 19 nm (before and after e-beam forced diffusion), and bare Bi2Se3 (contact placed on the 107 nm Au layer), see Figure S1, Supporting Information File 1. The curves for each Au thickness at the same area with the best reproducibility are shown in Figure 3. The potential cut-in barrier strongly depends on the Au thickness with values of 1.5 V for the 53 nm Au layer and 3.0 V for 19 nm Au (see Figure 3; curves c, d)). No potential barrier was measurable on Bi2Se3 up to 6 V.


Figure 3: Single-point IV curves performed during the mechanical contact with the Pt-coated conductive tip on a) bulk Au; b) 107 nm Au layer (both contacts on the same layer); c) 53 nm Au layer; d) and e) 19 nm Au layer before and after e-beam diffusion; f) cleaved Bi2Se3. Curves c–f) were measured with the ohmic contact on the 107 nm Au layer (see Figure S1, Supporting Information File 1).

There is a systematic increase in the potential cut-in barrier for 53 nm Au/Bi2Se3 upon cycling (Figure 4A). The cut-in barrier was less than 1 V during the first measurement, and above 6 V for the tenth measurement at the same place. The topography also changed significantly after 10 cycles of IV measurements. Newly formed perturbations were observed with a height of over 1 µm and a FWHM of 2 µm (Figure 4B). The changes in the electrical behavior as well as the growth of the perturbations can be explained by the growth of an isolating oxide layer by anodic oxidation. This has also been observed for Si [51,52], Ti, Ni or Al [53-55], carbon [56] and organic–inorganic compounds [57]. We suggest, in our case, that OH or O/O2 ions diffuse through our polycrystalline Au layer with many defects (grain boundaries, dislocations) forced by the strong electric field (8 V/53 nm of Au, i.e., ca. 1.5 × 108 V/m) [35,36]. At the Au–Bi2Se3 interface the oxidation of Bi2Se3 can occur accompanied by the formation of non-conductive bismuth oxides (optical Eg = 3.31 eV for a 60 nm BiOx film [58]). Typically, the thickness of surface oxide layers is much lower (of the order of nanometers for SiO2 on Si [51,52]). In our case it is increased by the electric field and good oxygen mobility in BiOx due to the formation of charged Bi vacancies [59,60].


Figure 4: (A) Change of the IV characteristics for ten repetitions of a single-point and (B) induced changes in the topography and (C) height profiles across a perturbation and a non-affected area (horizontal lines in (B)).

This assumption is in good agreement with the observed increase of the cut-in potential barrier upon thinning of the Au layer (see Figure 3) as a reminiscence of the diffusion limitation of oxygen/hydroxy anions due to the electric field through the polycrystalline Au layer. Bare Bi2Se3 had no diffusion limits and the IV curves are not measurable as electrons immediately form isolating BiOx on the surface. The abovementioned facts, tentatively summarized as IV characteristics, are strongly affected by the oxidation of Bi2Se3. The perturbation with a height of more than 1 µm could be easily formed by the oxidation of a highly O-mobile material, e.g., BiOx. This limits the usage of IV measurements in the static electrical contact of the Schottky barrier in ambient atmosphere to stable materials.

KPFM is an alternative way for mapping the changes in the surface contact potential [19,20,26] and reduces force and time of the tip–sample interaction, which in turn reduces the oxidation of the material. The NIs were embedded by local Au diffusion or the reaction of the Mo layer with the Bi2Se3 substrate. To enhance the diffusion/reaction rate, we used e-beam irradiation. An area of the 8 × 10 µm2 was locally irradiated by an electron beam for different periods of time (75 s; 180 s and 360 s). This resulted in a nanocomposite material consisting of the metal layer locally reacted with Bi2Se3 support and the original metal layer on the support surface. We were able to compare metallic materials with different thicknesses and determine the highest contrast in the surface contact potential reducing the problem of the diffused interface.

The e-beam-irradiated area exhibits a different SEM signal due to the changes of the chemical composition (Au diffusing into Bi2Se3) and changes in the topography (Figure 5). The same NIs, detected by the AFM (Figure 5), exhibited the topographical features of a few tens of nanometers (10 nm in the center and almost 40 nm at the edges) and changes in the phase contrast, both due to the density decrease of the Au alloy upon reaction with the substrate. All samples were irradiated for a period of 360 s. KPFM demonstrated strong variations of the surface contact potential of NIs (measured in the center of the irradiated area) with respect to the non-irradiated area of the Au layer (Figure 6). It should be noted, that the edges of the irradiated area are topographically elevated and exhibit a decreased surface potential because of the higher electron irradiation dose at the turning points of the e-beam. A higher dose led to a higher local temperature, which implies an increased formation of Au alloy and rims. The comparable so-called “Marangoni effect” was observed in laser-irradiated polymers as described by Lyutakov and co-workers [61].


Figure 5: (A) SEM image of the Au layer (33 nm in thickness) on Bi2Se3 irradiated by the electron beam for different periods of time; (B) topographical profile across the irradiated area taken from (C); (C) topographical AFM and (D) and phase-contrast AFM image of the e-beam-irradiated (360 s) Au layer (33 nm in thickness) on Bi2Se3.


Figure 6: Maps of the surface contact potential measured by means of KPFM for the e-beam formed NIs on the Au layer/Bi2Se3 for different values of Au layer thickness: (A) 12 nm, (B) 20 nm (B), and (C) 33 nm; (D) surface potential contrast between irradiated and non-irradiated parts of the metal layers as a function of the layer thickness.

The NIs formed on the thinnest Au (12 nm) exhibited a surface potential signal slightly above the detection limit of KPFM (Figure 6). With increasing thickness of the Au layer there is a higher contrast (Figure 6). The maximum value of the surface potential difference is between −80 and −90 mV and is connected with the diffusion of Au into Bi2Se3. Although the dependence and the values are alike for 1D particles (Figure 2) and for layers (Figure 6), it is difficult to compare them due to the different “substrate” (Bi2Se3 for 1D particles and non-diffused Au for the layers).

The Mo layers on Bi2Se3 reacted to the e-beam irradiation in a different way. The irradiated part of the Mo layer expanded less strongly than the Au layer (5 nm at the center, 20 nm at the edge, see Figure 7 profiles) and the contrast in the surface contact potential increased. The most intensive contrast (109 mV) was detected for NIs in the 9 nm thick Mo layer, whereas almost negligible surface contact potential difference (−4 mV) was detected for the 63 nm thick Mo layer.


Figure 7: (A) Topographical and (B) surface contact potential profiles of the 9 nm irradiated Mo layer; maps of the surface contact potential measured by the KPFM for the e-beam formed NIs on the Mo layer/Bi2Se3 for Mo thickness values of (C) 9 nm, (D) 17 nm, and (E) 63 nm.

The results of the surface contact potential mapping at the nanoscale were correlated to the macroscopic work functions (WF) of freshly cleaved Bi2Se3 and Au or Mo layers (both 30 nm in thickness). Work functions were measured by surface-sensitive ultraviolet photoelectron spectroscopy (UPS) (Figure 8) employing He I irradiation (hν = 21.22 eV) with an information depth up to 3 nm. The work function was determined from UPS spectra as an intersection between the background and the linear extrapolation of the secondary-electron cut-off. We obtained WFBi2Se3 = 5.22 eV, WFAu/Bi2Se3 = 4.79 eV, WFMo/Bi2Se3 = 5.98 eV for Ar+-sputtered samples. The WF increased by 760 meV with the Mo layer, whereas the WF decreased by −430 meV after sputtering the Au layer onto Bi2Se3.

We can sum up above results as follows:

The Au layer on the Bi2Se3 surface exhibited lower work function (WF) values than the freshly cleaved Bi2Se3 surface, measured by means of the UPS (4.79 eV vs 5.22 eV). This is in good agreement with the decrease of the surface contact potential of the separated Au particles on Bi2Se3 detected by the KPFM (limited to −40 mV, see Figure 2C).

The local irradiation of the Au layer on Bi2Se3 (Ar+ ions before UPS or SEM electrons during formation of NIs for KPFM measurements) led to a reaction with a substrate, most likely the incorporation of interstitial Au atoms into the Bi2Se3 matrix and Au alloy/compound formation. This material displayed lower density (topographical expansion, Figure 5) and the creation of the defect states (tentatively connected to the shoulder of the UPS edge, see Figure 8, black curve near 17 eV as well as a decrease in the surface contact potential of the Au NIs, see Figure 6).

In the case of Mo the situation is more complex. We were not able to prepare separated Mo 1D particles due to the higher melting temperature of Mo compared to the Au. The thinnest Mo layer led to the highest contrast in the surface contact potential of the NIs (see Figure 7) due to the formation of a thermodynamically more stable MoSe2 layer. The increase in Mo layer thickness led to the presence of the metallic Mo, which is not stable in air and is oxidized within 1 h at RT [62]. The presence of the MoOx layer is reflected by the high value of the WF (5.98 eV in this work, Mo 4.3 eV and MoSe2 5.65 eV as reported in [63]), which fits better to MoOx layer after the thermal evaporation (6.8 eV) and after 1 h oxidation at air at RT (5.6 eV). Partly oxygen-depleted MoOx exhibited a WF of 6.36 eV [62]. In our case, non-reacted Mo on the stable MoSe2 remains reactive (layered MoSe2 forms a barrier for further Mo diffusion) and Mo reacts with oxygen to produce MoOx. The UPS information is obtained from the topmost 2–3 nm, i.e., from the oxidized MoOx layer (30 nm of Mo was sputtered on the Bi2Se3 for UPS measurements). The 9 nm Mo layer after SEM electron irradiation led to the formation of MoSe2 (through the reduction of MoOx in vacuum to Mo and subsequent reaction with Bi2Se3) and these NIs revealed the highest contrast in contact potential (Figure 7). When the thickness of the Mo layer (17 and 63 nm, see Figure 7) is increased, not all Mo reacts to MoSe2 and there is a cover layer of MoOx. If the MoOx layer is irradiated by electrons under the same conditions as the thinnest Mo layer, there will be a partial oxygen depletion as well [62]. However, the long distance from the Bi2Se3 substrate hinders a reaction (the sample exhibits a sandwich structure of reactive Bi2Se3/isolating MoSe2/isolating Mo and MoOx/reactive oxygen-depleted MoOx). This is further reflected in a low contrast in KPFM measurements, as only non-irradiated and re-oxidized (due to ambient atmosphere) MoOx layers are measured. Even if MoSe2 is formed at the MoOx/Bi2Se3 interface, its presence is hidden for both UPS (low information depth) and KPFM (electrons prefer a more conductive path in the Mo/MoOx layer).


Figure 8: UPS data for Bi2Se3; Mo or Au layers on the top of the Bi2Se3 cleaned by the Ar+ sputtering excited by the He I (hν = 21.22 eV).

The applicability of the partial diffusion of the Au layers was tested by the photodiffusion of the 20 nm Au layer into Bi2Se3 by UV source through a Ronchi ruling [64,65]. The illuminated Au–Bi2Se3 exhibited a periodical pattern of the surface potential with an amplitude of 20 mV (Figure 9). This value is slightly lower in contrast to the NIs obtained by e-beam irradiation of the 20 nm Au layer (decrease 29 mV), and could be connected with the partial lateral diffusion of the heat or Au atoms/ions. In this way, the possibility to prepare a material with patterned surface contact potential at the macroscopic scale was demonstrated.


Figure 9: (A) Changes of the surface potential in the pattern after UV photodiffusion of Au (20 nm) into Bi2Se3 and (B) three line profiles.


We demonstrated local changes of the work function/surface contact potential as a consequence of metal nanoinclusions on the surface of Bi2Se3. These Schottky barriers can be realized in the form of separated metal nanoparticles through dc sputtering, or through the local reaction/diffusion of a metallic layer irradiated with electrons or UV photons. The value of the work function barrier is measurable by KPFM at the nanoscale even in ambient atmosphere and, most importantly, it strongly depends on the size of the metal particles. The results are in good agreement with the macroscopic UPS results (1 mm2) obtained in UHV. The investigation of the Schottky barrier in ambient atmosphere is limited to noble metals only, as the Mo nanolayer reacts with oxygen forming MoOx at RT within a few hours. We therefore conclude that KPFM has the following limitations: i) Only oxidization-resistant metals/semiconductor nanoinclusions can be used. Transition metals react with air and, thus, vacuum equipment is required for the preparation, transport and measurement. ii) The conductive AFM tip has to be coated with a mechanically hard and non-reacting layer as TiN. Typical Au or Pt coatings need to be avoided because the coating might diffuse into the material forced by the electric field during the interaction. iii) The IV characteristics can be affected by oxidation and perturbations with a height above 1 µm can be easily formed by oxidation of a highly O-mobile material such as BiOx.

Supporting Information

Supporting Information File 1: Scheme of electric circuit.
Format: PDF Size: 243.5 KB Download


The financial support from the Grant Agency of the Czech Republic (GA CR), projects No. 16-07711S and No. 19-11814S is greatly appreciated. The work was supported from European Regional Development Fund-Project "Modernization and upgrade of the CEMNAT (No. CZ.02.1.01/0.0/0.0/16_013/0001829)".


  1. Slack, G. A. CRC Handbook of Thermoelectrics; CRC Press: Boca Raton, FL, U.S.A., 1995; p 407.
    Return to citation in text: [1]
  2. Hicks, L. D.; Dresselhaus, M. S. Phys. Rev. B 1993, 47, 12727–12731. doi:10.1103/physrevb.47.12727
    Return to citation in text: [1]
  3. Liu, W.; Yan, X.; Chen, G.; Ren, Z. Nano Energy 2012, 1, 42–56. doi:10.1016/j.nanoen.2011.10.001
    Return to citation in text: [1]
  4. Li, H.; Tang, X.; Zhang, Q.; Uher, C. Appl. Phys. Lett. 2009, 94, 102114. doi:10.1063/1.3099804
    Return to citation in text: [1]
  5. Han, M.-K.; Ahn, K.; Kim, H.; Rhyee, J.-S.; Kim, S.-J. J. Mater. Chem. 2011, 21, 11365–11370. doi:10.1039/c1jm10163c
    Return to citation in text: [1]
  6. Liu, D.-W.; Li, J.-F.; Chen, C.; Zhang, B.-P. J. Electron. Mater. 2011, 40, 992–998. doi:10.1007/s11664-010-1476-x
    Return to citation in text: [1]
  7. Faleev, S. V.; Léonard, F. Phys. Rev. B 2008, 77, 214304–214312. doi:10.1103/physrevb.77.214304
    Return to citation in text: [1]
  8. Zebarjadi, M.; Esfarjani, K.; Shakouri, A.; Bahk, J. H.; Bian, Z. X.; Zeng, G.; Bowers, J.; Lu, H.; Zide, J.; Gossard, A. Appl. Phys. Lett. 2009, 94, 202105. doi:10.1063/1.3132057
    Return to citation in text: [1]
  9. Narducci, D.; Selezneva, E.; Cerofolini, G.; Frabboni, S.; Ottaviani, G. J. Solid State Chem. 2012, 193, 19–25. doi:10.1016/j.jssc.2012.03.032
    Return to citation in text: [1]
  10. Kulbachinskii, V. A.; Kytin, V. G.; Popov, M. Y.; Buga, S. G.; Stepanov, P. B.; Blank, V. D. J. Solid State Chem. 2012, 193, 64–70. doi:10.1016/j.jssc.2012.03.065
    Return to citation in text: [1]
  11. Popov, M.; Buga, S.; Vysikaylo, P.; Stepanov, P.; Skok, V.; Medvedev, V.; Tatyanin, E.; Denisov, V.; Kirichenko, A.; Aksenenkov, V.; Blank, V. Phys. Status Solidi A 2011, 208, 2783–2789. doi:10.1002/pssa.201127075
    Return to citation in text: [1]
  12. Park, D.-H.; Kim, M.-Y.; Oh, T.-S. Curr. Appl. Phys. 2011, 11, S41–S45. doi:10.1016/j.cap.2011.07.007
    Return to citation in text: [1]
  13. Li, F.; Huang, X.; Sun, Z.; Ding, J.; Jiang, J.; Jiang, W.; Chen, L. J. Alloys Compd. 2011, 509, 4769–4773. doi:10.1016/j.jallcom.2011.01.155
    Return to citation in text: [1]
  14. Li, J.-F.; Liu, J. Phys. Status Solidi A 2006, 203, 3768–3773. doi:10.1002/pssa.200622011
    Return to citation in text: [1]
  15. Xiong, Z.; Chen, X.; Huang, X.; Bai, S.; Chen, L. Acta Mater. 2010, 58, 3995–4002. doi:10.1016/j.actamat.2010.03.025
    Return to citation in text: [1]
  16. Dou, Y. C.; Qin, X. Y.; Li, D.; Li, L. L.; Zou, T. H.; Wang, Q. Q. J. Appl. Phys. 2013, 114, 044906. doi:10.1063/1.4817074
    Return to citation in text: [1]
  17. Zou, T. H.; Qin, X. Y.; Li, D.; Ren, B. J.; Sun, G. L.; Dou, Y. C.; Li, Y. Y.; Li, L. L.; Zhang, J.; Xin, H. X. J. Appl. Phys. 2014, 115, 053710. doi:10.1063/1.4864220
    Return to citation in text: [1]
  18. Ko, D.-K.; Kang, Y.; Murray, C. B. Nano Lett. 2011, 11, 2841–2844. doi:10.1021/nl2012246
    Return to citation in text: [1]
  19. Agarwal, K.; Kaushik, V.; Varandani, D.; Dhar, A.; Mehta, B. R. J. Alloys Compd. 2016, 681, 394–401. doi:10.1016/j.jallcom.2016.04.161
    Return to citation in text: [1] [2] [3] [4]
  20. Agarwal, K.; Mehta, B. R. J. Appl. Phys. 2014, 116, 083518. doi:10.1063/1.4894145
    Return to citation in text: [1] [2] [3] [4]
  21. Singh, B.; Mehta, B. R.; Varandani, D.; Govind; Narita, A.; Feng, X.; Müllen, K. J. Appl. Phys. 2013, 113, 203706. doi:10.1063/1.4807411
    Return to citation in text: [1] [2]
  22. Miwa, K.; Salleh, F.; Ikeda, H. Makara J. Technol. 2013, 17, 17–20. doi:10.7454/mst.v17i1.1922
    Return to citation in text: [1]
  23. Muñoz-Rojo, M.; Caballero-Calero, O.; Martín-González, M. Appl. Phys. Lett. 2013, 103, 183905. doi:10.1063/1.4826684
    Return to citation in text: [1]
  24. Olaya, D.; Hurtado-Morales, M.; Gómez, D.; Castañeda-Uribe, O. A.; Juang, Z.-Y.; Hernández, Y. 2D Mater. 2017, 5, 011004. doi:10.1088/2053-1583/aa90d8
    Return to citation in text: [1] [2]
  25. Durcan, C. A.; Balsano, R.; LaBella, V. P. J. Appl. Phys. 2014, 116, 023705. doi:10.1063/1.4889851
    Return to citation in text: [1] [2]
  26. Vanis, J.; Zelinka, J.; Zeipl, R.; Jelinek, M.; Kocourek, T.; Remsa, J.; Navratil, J. J. Electron. Mater. 2016, 45, 1734–1739. doi:10.1007/s11664-015-4193-7
    Return to citation in text: [1] [2]
  27. Klimovskikh, I. I.; Sostina, D.; Petukhov, A.; Rybkin, A. G.; Eremeev, S. V.; Chulkov, E. V.; Tereshchenko, O. E.; Kokh, K. A.; Shikin, A. M. Sci. Rep. 2017, 7, 45797. doi:10.1038/srep45797
    Return to citation in text: [1]
  28. Yamasaka, S.; Nakamura, Y.; Ueda, T.; Takeuchi, S.; Sakai, A. Sci. Rep. 2015, 5, 14490. doi:10.1038/srep14490
    Return to citation in text: [1]
  29. Bala, M.; Gupta, S.; Tripathi, T. S.; Varma, S.; Tripathi, S. K.; Asokan, K.; Avasthi, D. K. RSC Adv. 2015, 5, 25887–25895. doi:10.1039/c5ra01000d
    Return to citation in text: [1]
  30. Knotek, P.; Chanova, E.; Rypacek, F. Mater. Sci. Eng., C 2013, 33, 1963–1968. doi:10.1016/j.msec.2013.01.006
    Return to citation in text: [1] [2] [3]
  31. Cermak, P.; Ruleova, P.; Holy, V.; Prokleska, J.; Kucek, V.; Palka, K.; Benes, L.; Drasar, C. J. Solid State Chem. 2018, 258, 768–775. doi:10.1016/j.jssc.2017.12.009
    Return to citation in text: [1]
  32. Knotek, P.; Tichý, L. Thin Solid Films 2009, 517, 1837–1840. doi:10.1016/j.tsf.2008.09.041
    Return to citation in text: [1]
  33. Knotek, P.; Tasseva, J.; Petkov, K.; Kincl, M.; Tichy, L. Thin Solid Films 2009, 517, 5943–5947. doi:10.1016/j.tsf.2009.04.038
    Return to citation in text: [1] [2]
  34. Das, S.; Zazpe, R.; Prikryl, J.; Knotek, P.; Krbal, M.; Sopha, H.; Podzemna, V.; Macak, J. M. Electrochim. Acta 2016, 213, 452–459. doi:10.1016/j.electacta.2016.07.135
    Return to citation in text: [1]
  35. Glatzel, T.; Sadewasser, S.; Lux-Steiner, M. C. Appl. Surf. Sci. 2003, 210, 84–89. doi:10.1016/s0169-4332(02)01484-8
    Return to citation in text: [1] [2]
  36. Melitz, W.; Shen, J.; Kummel, A. C.; Lee, S. Surf. Sci. Rep. 2011, 66, 1–27. doi:10.1016/j.surfrep.2010.10.001
    Return to citation in text: [1] [2] [3] [4]
  37. Knotek, P.; Tichy, L.; Kutalek, P. Thin Solid Films 2015, 594, 67–73. doi:10.1016/j.tsf.2015.09.055
    Return to citation in text: [1]
  38. Knotek, P.; Vlcek, M.; Kincl, M.; Tichy, L. Thin Solid Films 2012, 520, 5472–5478. doi:10.1016/j.tsf.2012.03.116
    Return to citation in text: [1]
  39. Axt, A.; Hermes, I. M.; Bergmann, V. W.; Tausendpfund, N.; Weber, S. A. L. Beilstein J. Nanotechnol. 2018, 9, 1809–1819. doi:10.3762/bjnano.9.172
    Return to citation in text: [1]
  40. Sadewasser, S.; Nicoara, N.; Solares, S. D. Beilstein J. Nanotechnol. 2018, 9, 1272–1281. doi:10.3762/bjnano.9.119
    Return to citation in text: [1]
  41. Rezek, B.; Stehlik, S. Chapter 10 - Surface potential of nanodiamonds investigated by KPFM. In Nanodiamonds; Arnault, J.-C., Ed.; Elsevier: Amsterdam, Netherlands, 2017; pp 273–300. doi:10.1016/b978-0-32-343029-6.00011-8
    Return to citation in text: [1] [2]
  42. Watanabe, T.; Fujihira, M. Ultramicroscopy 2009, 109, 1035–1039. doi:10.1016/j.ultramic.2009.03.047
    Return to citation in text: [1]
  43. Zeng, T.-W.; Hsu, F.-C.; Tu, Y.-C.; Lin, T.-H.; Su, W.-F. Chem. Phys. Lett. 2009, 479, 105–108. doi:10.1016/j.cplett.2009.07.104
    Return to citation in text: [1]
  44. Yamauchi, T.; Tabuchi, M.; Nakamura, A. Appl. Phys. Lett. 2004, 84, 3834–3836. doi:10.1063/1.1745110
    Return to citation in text: [1]
  45. Sasahara, A.; Pang, C. L.; Onishi, H. J. Phys. Chem. B 2006, 110, 17584–17588. doi:10.1021/jp063665h
    Return to citation in text: [1]
  46. Sasahara, A.; Hiehata, K.; Onishi, H. Catal. Surv. Asia 2009, 13, 9–15. doi:10.1007/s10563-009-9062-z
    Return to citation in text: [1]
  47. Lide, D. R. Handbook of Chemistry and Physics, 84th ed.; CRC Press: Boca Raton, FL, U.S.A., 2003.
    Return to citation in text: [1]
  48. Nanomaterials: Synthesis, Properties and Applications; J. W. Arrowsmith, Ltd.: Bristol, United Kingdom, 1996.
    Return to citation in text: [1]
  49. Lee, H.; Keun Lee, Y.; Nghia Van, T.; Young Park, J. Appl. Phys. Lett. 2013, 103, 173103. doi:10.1063/1.4826140
    Return to citation in text: [1]
  50. Tedesco, J. L.; Rowe, J. E.; Nemanich, R. J. J. Appl. Phys. 2009, 105, 083721. doi:10.1063/1.3100212
    Return to citation in text: [1]
  51. Avouris, P.; Martel, R.; Hertel, T.; Sandstrom, R. Appl. Phys. A: Mater. Sci. Process. 1998, 66, S659–S667. doi:10.1007/s003390051218
    Return to citation in text: [1] [2]
  52. Tello, M.; Garcı́a, R. Appl. Phys. Lett. 2001, 79, 424–426. doi:10.1063/1.1385582
    Return to citation in text: [1] [2]
  53. Irmer, B.; Kehrle, M.; Lorenz, H.; Kotthaus, J. P. Appl. Phys. Lett. 1997, 71, 1733–1735. doi:10.1063/1.120019
    Return to citation in text: [1]
  54. Fang, T.-H.; Chen, K.-J. Mater. Trans. 2007, 48, 471–475. doi:10.2320/matertrans.48.471
    Return to citation in text: [1]
  55. Shingubara, S.; Murakami, Y.; Morimoto, K.; Takahagi, T. Surf. Sci. 2003, 532–535, 317–323. doi:10.1016/s0039-6028(03)00433-3
    Return to citation in text: [1]
  56. Chuang, M.-C.; Chien, H.-M.; Chain, Y.-H.; Chi, G.-C.; Lee, S.-W.; Woon, W. Y. Carbon 2013, 54, 336–342. doi:10.1016/j.carbon.2012.11.045
    Return to citation in text: [1]
  57. Baumgärtel, T.; von Borczyskowski, C.; Graaf, H. Beilstein J. Nanotechnol. 2013, 4, 218–226. doi:10.3762/bjnano.4.22
    Return to citation in text: [1]
  58. Leontie, L.; Caraman, M.; Alexe, M.; Harnagea, C. Surf. Sci. 2002, 507-510, 480–485. doi:10.1016/s0039-6028(02)01289-x
    Return to citation in text: [1]
  59. Zhong, G.; Wang, Y.; Dai, Z.; Wang, J.; Zeng, Z. Phys. Status Solidi B 2009, 246, 97–101. doi:10.1002/pssb.200844271
    Return to citation in text: [1]
  60. Torruella, P.; Coll, C.; Martín, G.; López-Conesa, L.; Vila, M.; Díaz-Guerra, C.; Varela, M.; Ruiz-González, M. L.; Piqueras, J.; Peiró, F.; Estradé, S. J. Phys. Chem. C 2017, 121, 24809–24815. doi:10.1021/acs.jpcc.7b06310
    Return to citation in text: [1]
  61. Lyutakov, O.; Huttel, I.; Siegel, J.; Švorčík, V. Appl. Phys. Lett. 2009, 95, 173103. doi:10.1063/1.3254210
    Return to citation in text: [1]
  62. Irfan, I.; James Turinske, A.; Bao, Z.; Gao, Y. Appl. Phys. Lett. 2012, 101, 093305. doi:10.1063/1.4748978
    Return to citation in text: [1] [2] [3]
  63. Keszler, D. A.; Wager, J. F. Novel Materials Development for Polycrystalline Thin-Film Solar Cells: Final Subcontract Report, 26 July 2004--15 June 2008; Office of Scientific and Technical Information (OSTI), 2008; pp 1–16. doi:10.2172/942065
    Return to citation in text: [1]
  64. Balakrishnan, M.; Kozicki, M. N.; Poweleit, C. D.; Bhagat, S.; Alford, T. L.; Mitkova, M. J. Non-Cryst. Solids 2007, 353, 1454–1459. doi:10.1016/j.jnoncrysol.2006.09.071
    Return to citation in text: [1]
  65. Adam, J.-L.; Zhang, X., Eds. Woodhead Publishing Series in Electronic and Optical Materials; Elsevier: Amsterdam, Netherlands, 2014.
    Return to citation in text: [1]

© 2019 Knotek et al.; licensee Beilstein-Institut.
This is an Open Access article under the terms of the Creative Commons Attribution License ( Please note that the reuse, redistribution and reproduction in particular requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Nanotechnology terms and conditions: (

Back to Article List

Other Beilstein-Institut Open Science Activities

Keep Informed

RSS Feed

Subscribe to our Latest Articles RSS Feed.


Follow the Beilstein-Institut


Twitter: @BeilsteinInst