Novel low-macroscopic-field emission cathodes for electron probe spectroscopy systems

Electron probe methods comprise a series of various techniques that are commonly being used for composition and morphology analysis of structures,^1–3 materials,⁴ and devices.⁵ Among the most widespread and established ones is electron probe energy-dispersive x-ray spectroscopy (EDS), which is a common method utilized for performing the elemental composition investigation in a wide range of applied fields.^6–8 The main drawback of EDS is its very limited (due to a well-known beam scattering effect inside the sample) spatial resolution amounting to the values of Δd ≈ 1 μm (see Refs. 9 and 10, for example).

Modern electron probe microanalysis systems also include a scanning electron microscope (SEM), which allows mapping and imaging of the sample surface by scanning it with an electron beam, followed by displaying the received (with the use of an electron detector) signal on a screen. In this case, it is possible to register both the topography and the elemental composition of a particular phase (that is, its average atomic number).

Usually, Δd ≈ 1 μm turns out to be sufficient for dealing with routine tasks, but in a number of cases, methods with much higher spatial, energy, and temporal resolving capacity^11,12 are required, since it is often necessary to carry out a quantitative analysis of low-dimensional regions. One such technique currently utilized for thin-film investigations is the characteristic electron energy loss spectroscopy (EELS). This method is based on a fundamental difference between the elastic and the inelastic electron scattering process that takes place during the interaction of a primary electron beam and the electrons from atomic shells of a sample.

Elastically scattered electrons experience strong deviations from their original trajectory and, therefore, are not recorded by EELS detectors, which collect inelastically scattered electrons from a region with dimensions comparable to the diameter of the initial beam. Thus, the attainable value of the spatial resolution can be as low as several nanometers.¹² The utilization of field cathodes acting as primary electron sources has already made it possible to overcome the issues of insufficient electron beam brightness allowing one to perform quantitative analysis of the elemental composition in nanoregions.

A modern EEL spectroscopy system is capable of selecting only those inelastically scattered electrons that have undergone a single characteristic energy loss (i.e., collided with electrons from the atomic shells of the investigated object once) and discriminating them in accordance with their energy.

When further improvement of electron probe methods' capabilities in general and EELS, in particular, is considered, a number of research areas arise. These include ensuring the energy stability of the primary electron beam, the development and implementation of narrow energy window electron detectors, and, finally, the creation of new primary electron sources characterized by even higher probe current densities as well as narrower electron distribution functions. The last two objectives are closely related to the problem of enhancing the EELS energy and temporal resolving capacity.

In this context, the design of novel devices characterized by an even higher accuracy of nanoscale elemental and phase composition analysis, capable of trace elements determination as well as the investigation of nonstationary processes by methods of time-resolved spectroscopy,^13,14 is crucial.

One of the approaches to increasing the resolving capacity is the utilization of the so-called direct detection sensors, which in many cases possess a number of advantages over classical indirect detection sensors, including higher energy resolution values and a relatively low noise level.¹¹ This approach has been successfully used in imaging and diffractometry systems.

Despite the advances in the development and application of direct detection sensors, fundamental factors limiting the EELS energy resolution are not analyzed thoroughly enough.

Emitted electron energy spectrum dispersion [or electron energy distribution function (EEDF) width] is one of those factors currently relevant to a wide range of electron probe microscopy methods limiting a further increase in both their energy resolving capacity as well as detection threshold.^11,12

One of the most effective ways to cope with this constraint is the introduction of monochromators. To date, commercially available monochromators for transmission electron microscopy methods allow one to achieve energy resolutions of less than 0.1 eV. This result was successfully used for EELS systems (see Refs. 15–17, as an example). However, the introduction of monochromators leads to a beam current density (and thus, the brightness of a charge carrier source) decrease, simultaneously negatively affecting the cross section of an electron probe. While the second phenomenon can be relatively easily dealt with via the implementation of a modern focusing system, the first one is inevitable in any monochromated electron probe system due to its fundamental nature.

This is especially important in the case of inner-shell energy loss studies (the so-called core-loss spectroscopy), since the analysis of the electronic levels' fine structure as well as the detection of chemical elements' trace concentrations requires high electron beam current density values (otherwise it becomes nearly impossible to read low-intensity signals). The monochromated source energy resolution of several meV turns out to be excessive because in the case under consideration, electron energy loss spectrum is dominated by the effects of spectral line broadening due to the finite lifetime of states (the so-called lifetime broadening) and solid-state effects.^14,16

Thus, the use of monochromators in EELS systems is by no means the ultimate solution, and an appropriate method should be carefully chosen based on the goals of a particular analysis procedure by various EELS techniques.¹⁷ 

In practice, the energy resolution of a system depends not only on the energy dispersion of the source electrons but also on the voltage stability parameter of the microscope,^12,14 the dispersion of the spectrometer itself, and on the parasitic electromagnetic field. In addition, evidence exists that the characteristics of a system with a monochromator degrade due to the presence of source energy instabilities.¹⁸ These instabilities have a bad effect on all types of electron sources, and, therefore, the actual energy resolution of a system can be much lower than its source could otherwise provide. These risks need to be taken into account when conducting analytical research, although currently, there are a number of software methods for correcting such drift (or the so-called low-frequency instability), which make it possible to obtain energy resolution values up to 0.27 eV when using cathodes based on field emission electron guns without monochromators.¹⁵ The implementation of high-speed data acquisition to utilize the available energy resolution of the electron source¹⁵ also leads to some improvements, allowing for partial mitigation of energy instabilities when conducting research at a high frequency.

It should be also noted that a fully-fledged theory that unambiguously relates the parameters and characteristics of a source to the electron distribution function (or energy dispersion) of emitted electrons, regardless of its type (for example, CFEG—cold field emission cathode, Schottky emitter, heated cathode,⁹ etc.), material (lanthanum hexaboride, thoriated tungsten, carbon), and spatial configuration (pointed, planar cathodes) still not existing, while the processing of experimental data on emission from cathodes based on a new type of nanostructured materials requires the interpretation of several exotic physical effects.¹⁹ 

For example, insufficient attention is currently being paid to quantum phenomena in low-dimensional systems [for example, the formation of two-electron (2e) states with negative correlation energy].¹⁹ This fact is all the more important since the analysis of optimization problems for both modern instrumentation and materials,^20,21 the design, manufacture, implementation of devices, and prototypes,²² and diagnostics (including the development of local micro- and nanodiagnostic systems,²³ as well as control and communication systems^24–27), indicates the presence of uncertainty and only the relative possibility of its reduction on the basis of reliable quantitative optimizing information in the form of physical models,²³ describing the relationship between the recorded signals and interaction processes of various physical fields with matter. Despite this ambiguity, there is an extensive layer of experimental data that could allow one to derive empirical dependencies with sufficient accuracy for dealing with both the traditional set of routine tasks (for example, elemental analysis) as well as less trivial problems such as phase analysis at nanograin interfaces.

As already mentioned, one of the most important factors contributing to the energy resolution of an EELS system as a whole appears to be the EEDF width. Thus, the investigation of the possibilities of its further reduction is crucial for the improvement of electron microscopy methods.

The creation of novel low-macroscopic-field emission (LMFE) cathodes that exhibit lower threshold extraction voltage and surface work function values while also reducing the EEDF width and maintaining (or even increasing) the emission current density is one of the potential solutions to this issue. Thus, the aim of our research is to increase the energy and temporal resolutions of EEL spectroscopy systems for the study and characterization of natural and artificial objects by developing a cathode prototype based on the LMFE physical effect. Although the exact mechanism of this phenomenon is still a matter of debate, its detailed description and new LMFE physical models can be found in Refs. 19 and 28–31.

LMFE is undoubtedly some form of field-induced emission, but its nature has not yet been studied thoroughly. In recent years, the LMFE from carbon-based films³² has attracted much attention, but earlier examples of both field and LMF emission from thin dielectric films³³ also included phenomena that are rather difficult to substantiate. In fact, the phenomenon of LMFE is (under appropriate conditions) a common property of thin (50–300 nm) nanostructured films.

According to Ref. 34, several types of possible LMFE structures may exist. The most common one is characterized by the presence of conductive particles in an insulating matrix, which can be described as a leaky dielectric. In these structures, filamentous conducting channels exist (or are formed by activation) between the substrate and the particles, between the particles themselves, or between the conducting particles near the surface and vacuum.³⁴ 

In addition to carbon-based films, nanostructured materials with pronounced field emission include oxide films and semiconductor inclusions on the high-voltage electrode surfaces;³³ composite materials with small conductive or semiconductor inclusions deliberately dispersed in resins, glasses, or other dielectrics; small clusters of gold atoms embedded in an amorphous carbon (a-C) matrix. It is assumed³¹ that field emission from such materials is associated with the formation of a conducting nanostructure inside the film, near or directly at the film/vacuum interface. Local field enhancement at the tip of such nanostructure takes place, facilitating the field emission process. In some cases, the emission mechanism may also be thermally activated.

As a result of our earlier studies, an LMFE emitter material based on porous silicon functionalized with silver-doped C₆₀ fullerene structures was created.¹⁹ The prototype outperforms current and promising analogs in terms of the threshold field strength (F_th). The height of the electron emission potential barrier for the emitter material as well as the values of the local near-surface field strength were estimated using the developed LMFE model,¹⁹ which is based on Anderson's theory of two-electron localized states with negative correlation energy.^35,36 It may be expected that this effect is shared by a broad variety of emitting carbon structures, including graphene-like layers, carbon nanotubes, as well as layers of C₆₀ fullerene molecular aggregates.

It has already been demonstrated¹⁵ that under certain operating conditions (i.e., lower extraction voltage values) regular field emission cathodes show promising results when emitted electron spectrum width is concerned. The decrease in dispersion amounted to 0.07 eV for the corresponding extraction voltage reduction from 5 to 3.5 kV. The main drawback of this technique is that the brightness of the electron sources operating under such conditions was insufficient for conducting core-loss investigations (same as in the case of monochromated EELS systems^15,16) due to a dramatic decrease in the electron beam current density by several orders of magnitude.

The problem of electron work function lowering has long stood before leading scientific groups across the world and, until recently, remained virtually unresolved. In fact, it is due to the development of new surface modification technologies utilizing carbon nanostructures that it is now possible to achieve the values of the effective potential barrier for electrons less than 1 eV. This result was obtained via the modification process of the anode surface with layers of graphene intercalated by cesium atoms.³⁷ This technique has already been successfully utilized to implement a prototype thermionic energy converter characterized by a significant increase in efficiency under laboratory conditions.^37,38

It should be also noted that the proper control of any vacuum electronic device requires one to perform the charged particles' distribution function analysis. Since the plasma of such devices is localized in a very small area and is characterized by an extremely high density of charged particles in strong fields, such studies are, in turn, impossible without the methodology for accurate distribution function measurements under conditions that correspond to their actual operating modes. In this instance, the basic processes are diverse, and the distribution functions themselves are anisotropic. Hence it follows that the use of traditional approaches to the analysis of the properties of both the electrodes themselves and the plasma within the device is not possible. In this regard, novel probe diagnostics techniques for charged particle plasmas with an arbitrary anisotropy degree are developed and tested, while the studies on processes taking place both within the volume and immediately in the near-electrode areas are carried out.^39,40

Thus, the development and implementation of LMFE electron sources for EEL systems is relevant to ensure a significant reduction in both the threshold field strength and EEDF width values, leading to a further increase in the efficiency of both elemental analysis for the purpose of trace concentration measurements, as well the phase composition investigation with subnanometer spatial resolution (which is especially important in the field of metal science, for example, when mapping the composition of carbide phases at the grain interface).

As already noted, low-macroscopic-field cathodes for novel electron guns should demonstrate high current density values and low energy dispersion of emitted charge carriers. The main idea proposed in this study is to create a LMFE material via the functionalization process of a matrix (semiconductor) surface with an adjusted aspect ratio governed by its synthesis parameters.

During the study, a novel material based on a porous silicon matrix functionalized with C₆₀ fullerene carbon nanostructures (the so-called “solgel prototype”) was developed and investigated. Porous silicon was chosen as a matrix due to its unique properties, including a developed surface and the ability to create a substrate with the desired shape, structure, and morphology.^4,41

The synthesis and functionalization processes were carried out at the Department of Micro- and nanomaterials (ETU “LETI”) using the solgel synthesis technological method also developed there.^42,43 Its advantages comprise the high purity degree, the ease of obtaining multicomponent layers, and the ability to control the structure of acquired materials on various scales, allowing one to create hierarchical nanostructured objects.^4,13 Purity is ensured through the use of molecular precursors, the purification of which is straightforward if standard methods of rectification and chromatography are being utilized. The relative simplicity of a synthesized specimen doping procedure and the ability to control its structure are both directly conditioned by main solgel method reactions—hydrolysis and polycondensation. The course of these reactions can be controlled by changing the acidity of the medium. Because the hydrolysis and polycondensation processes are well understood, tetraethoxysilane has been widely employed as a precursor. The C₆₀-OH(Ag/Cu) water solution (same as in Ref. 19) was used as a precursor for the formation of the surface functionalization layer. Then, the required volume of tetraethoxysilane was added to the solution. The resulting solution matured for 3 days. After maturation, the solution was applied to a substrate and centrifuged (3000 rpm) for 15 s. Next, annealing was carried out at a temperature of 300 °С for 30 min. Porous silicon matrices were obtained via the process described in Refs. 19 and 44. Monocrystalline silicon wafers with (100) crystallographic orientation of KDB-12 brand were used as a starting material. The substrates were preliminarily chemically cleaned of contaminants by sequential washing in an ultrasonic bath in acetone, isopropyl alcohol, and distilled water.

Scanning electron microscopy of the functionalized emitter surface was carried out using a FIB-SEM TESCAN S9251G double-beam scanning electron-ion microscope (4th gen).

The emission characteristics measurements were carried out using the same setup as in Ref. 19. The device includes a signal generator, a high-voltage power supply unit, a digital oscilloscope, and a vacuum chamber containing an electrode structure. A more detailed description of its capabilities and emission measurement parameters can also be found in Ref. 19. The processing and analysis of emission characteristics were carried out via the Fowler–Nordheim linearization technique.^19,31

The charge carrier source of a standard electron probe system is an electron gun typically based on a thermionic, thermo-field, or field emission cathode. The emitted electrons are accelerated by the field formed in the channel (the negative potential values applied to the cathode are usually as high as 100 kV) and pass through a hole in the anode plate, which is held at zero voltage (ground potential).

The focusing of the beam is carried out by using electromagnetic lenses, under the influence of which a narrow electron probe is formed, which arrives at the surface of a thin sample. Electrons passing through the test sample undergo elastic and inelastic scattering processes. Inelastically scattered electrons practically do not change their trajectory and, if the source is characterized by sufficient brightness (current density in the initial beam), they are recorded by detection sensors^11,12 with a certain signal-to-noise ratio (SNR) and detective quantum efficiency (DQE). The setup operates in a high vacuum mode due to the need to reduce the influence of electron scattering effects directly in the beam as well as to avoid rapid degradation of the charge carrier source.

The goal set in the framework of this study is of a complex nature, so it is necessary to provide a theoretical justification for increasing the resolving capacity of characteristic electron energy loss spectroscopy methods. As noted in the introductory section, the main advantage of EEL spectroscopy is its spatial resolution (Δd < 10 nm). Nevertheless, when conducting a local analysis of the oxidation state as well as the type of chemical bonds in the sample, the energy resolution of an EEL system (ΔE) is of particular importance.

As already noted, a substantial increase in ΔE can be achieved by utilizing direct detection sensors that require the use of a thinning process on their substrate back side, which, in turn, makes it possible to register each electron entering the sensor with a minimum probability of backscattering, significantly reducing the degree of signal delocalization and, thus, contributing to both the detector LSF narrowing and an increase in the overall signal-to-noise ratio. One of the most successful examples of increasing the EEL system resolution through the utilization of direct detection sensors is presented in Ref. 11, where the zero-loss peak FWHM value measured for a direct detection sensor turned out to be 1 eV lower than the corresponding value for a system with an indirect detection sensor. Since our focus is to investigate the influence of source electron energy dispersion separately, we assume that the detector LSF equals a delta function (the case of an ideal detector), meaning that E_s and ZLP are interchangeable in the course of this study.

Electron energy loss spectroscopy is one of the most promising methods for analyzing and mapping the composition of multiphase rocks and materials (including oxide heterostructures^45,46), for which any variations in the valence of elements and stoichiometric composition at the phase boundary lead to radical changes in properties and characteristics, which ultimately determine the possibility of their use in the development and creation of functional electronics devices. Even though, unlike soft condensed media, such as polymers, colloidal systems, liquid crystals, etc., solids are relatively resistant to short-term electron beam exposure, the effects of samples' structure degradation (both physical and electronic) under prolonged exposure still remain a serious problem for this class of analytical methods.⁴⁷ It is for this reason that an increase in the SNR and, consequently, the DQE, via the use of larger electron beam exposure duration values is detrimental for any samples under consideration.

As in the case of energy resolving capacity, the solution to this problem requires a comprehensive approach that comprises the utilization of direct detection sensors¹¹ as well as the creation of a new type of primary carrier source, characterized by low energy dispersion values.

Thus, ensuring a reduction in the duration of data collection when measuring the EEL spectra via an increase in the temporal resolution of the system is one of the primary goals, which is also extremely important for maintaining a high spatial resolution of the EELS methods, since it could allow one to reduce the probability of random sample displacement.

It should be noted that EELS SNR (and DQE) is a function of both the electron dose and the energy frequency of the signal; consequently, for such modifications of EELS as the characteristic electron energy loss near the absorption edge spectroscopy (ELNES), the energy resolution is of decisive importance. Therefore, one should consider not only the low-frequency detective quantum efficiency but also the high-frequency one.

The successful use of a direct detection sensor¹¹ has already made it possible to reduce the temporal resolution Δt and the detection threshold of chemical elements with trace concentrations by a factor of 2.2 and by 45%, respectively; however, according to Ref. 11, an emission current density growth will also contribute to the increase in efficiency of the dynamic systems' investigations (i.e., characterized by the presence of physical or chemical processes, the duration of which is too short to be analyzed using conventional EELS).

As already noted, the energy dispersion of charge carriers (hence, the width of the ZLP function and, thus, the energy resolution of an EELS system) depends on the extraction voltage (V_e) value of a particular field emission cathode gun.¹⁵ In this case, an experimentally and theoretically confirmed reduction in ZLP FWHM is observed with decreasing V_e.^10,48 At the same time, a sharp drop in the emitted electron current density values accompanied by a decrease in the electron gun brightness occurs. This, in turn, leads to a reduction in SNR and, thus, the energy, spatial, and temporal resolution of the system. In that regard, it is necessary to develop a new generation of field emission cathodes, demonstrating lower values of the extraction voltage while maintaining or even enhancing the emission current density.

The results of scanning electron microscopy of the synthesized solgel emitter prototype are shown in Figs. 1(a) and 1(b).

The surface of the sample is uniformly covered with two arrays of island-type nanosized structures. The first one is characterized by a lateral size of about 30–50 nm, the second one of less than 10 nm.

It should be noted that initially this sample demonstrated threshold extraction voltage values V_th < 1.5 kV, while the current values reached 2 mA. During the first two measurement cycles, this prototype demonstrated steady emission characteristics, however, due to the rapid degradation of the porous gel matrix containing the doped Ag and Cu molecular structures based on fullerene C₆₀, a sharp increase in threshold voltage values occurred, accompanied by a decrease in current density.

The emission characteristics of the stabilized sample indicate that the current strength depends exponentially on the voltage applied between the electrodes. V_th for the solgel prototype stabilized at 2200 V (see Fig. 2), still surpassing the results achievable when utilizing classical planar semiconductor materials (including silicon) as field emission cathodes at comparable V_e values but being inferior to the prototype acquired earlier via the use of the patented dynamic impregnation technology,^19,44 where V_th was as low as 1100 V (see Fig. 3).

The functional connection in the modified coordinates [see Figs. 2(b) and 3(b)] is close to linear (Pearson correlation coefficient ranged from −0.92 to −0.9, depending on the measurement cycle), thus indicating the presence of a field-induced emission mechanism.

It is well known that field emission cathodes are superior to their analogs both in terms of their charge carriers' current density and energy dispersion (more than 0.5 eV for thermal-field Richardson–Schottky cathodes at comparable source brightness values). Today, several models of the emitted electrons energy distribution (mainly based on the classical Fowler–Nordheim theory^10,49,50) describing the dependence of its shape on the emission parameters (including the extraction voltage and temperature) are widely used.

For the following analysis, we decided to focus on the LMFE emitter, which was obtained and investigated as a result of cooperation with the ETU LETI “Theranostics” R&D laboratory (supervisor Yu. M. Spivak), SPbPU “Self-organizing high-temperature nanostructures” research laboratory (supervisor P. G. Gabdullin), the Ioffe Institute, and TESCAN CIS specialists. This prototype is based on porous silicon functionalized with C₆₀ fullerene carbon structures acquired via the patented dynamic impregnation technique (see Refs. 19 and 44 for details) and, as already mentioned, is demonstrating better overall emission resource.

The exact methodology of emission characteristics analysis is described extensively in Ref. 19. The obtained results signify the advantage of the prototype in terms of the attainable emission current density and the values of the threshold field strength over modern and some of the perspective analogs, including metal field emitter array (FEA) cathodes (which, despite the relative simplicity of manufacture and theoretical description, are limited from the standpoint of practical application because of insufficient compatibility with CMOS technologies), carbon nanotubes, characterized by low emission current values due to the technological complexity of manufacturing a large-area regular array of those,^51,52 as well as field emitters based on monocrystalline silicon, the emission properties of which are limited by high values of the macroscopic inter-electrode (IE) threshold field strength (more than 1 V/nm).⁵³ 

To describe the relationship of the recorded parameters (including the energy stability of the beam, the emission current density in the beam, and the threshold voltage) with the physical processes occurring in the material under study, our new model¹⁹ of the LMF emission phenomenon based on the theory of two-electron localized states with negative Hubbard correlation energy was used. This model makes it possible to characterize the most important parameters of the prototype, including the effective potential barrier height, as well as both macroscopic and local field threshold values.

Despite the relative simplicity and versatility of the solgel technique the respective cathode still has a lot of room for improvement, mainly when the emission characteristics stability is considered. Furthermore, the analysis performed via the aforementioned Fowler–Nordheim linearization technique (see Figs. 2 and 3) shows that the F–N slope for the solgel emitter (–22 890) vastly surpasses the corresponding value for the dynamically impregnated prototype (–8504) resulting in a much larger effective potential barrier height (0.75 vs 0.36 eV, respectively).

The result of source electrons' energy distribution spectra physical modeling for the chosen emitter is presented in Fig. 4. One can see from the figure that a standard metal (gold) cathode with a field enhancement factor of γ ≃ 100 and the effective potential barrier for electrons W = 5.3 eV is characterized by the electron energy spectrum FWHM of 0.38 eV.

In terms of the emission current density (j_max = 6.28 × 10³ μA/cm² for the IEF strength F_max = 7.5 V/μm), a prototype cathode with a low potential barrier for electrons value (W = 0.36 eV¹⁹) and a field enhancement factor of γ ≃ 100 appears to be superior to the classical metal emitter for both the identical values of macroscopic IE field strength as well as the values one order of magnitude lower. In this scenario, the maximum attainable field emission current density from emitters based on carbon nanotubes^51,54,55 is more than 1.2 times lower than that of the proposed prototype.

The electron energy spectrum FWHM value for the obtained prototype is 0.17 eV (Fig. 4). Due to the achieved decrease of 0.21 eV in the ZLP function peak (1), a 20% energy resolution gain can be provided in comparison with the classical field emission cathode.

A more robust and informative way for comparing systems with different electron sources in terms of energy resolution is their modulation transfer functions' (MTF) analysis.¹¹ Since MTF is acquired via the calculation of the ZLP Fourier transform, it allows one to compare the EEL systems' response to harmonic (sinusoidal) signals of varying energy frequency (f_e, eV⁻¹). A modulation transfer function of unity is expected for a system with ideal (i.e., delta function) ZLP response. Figure 5 contains the MTFs for both the reference and the investigated cathodes as a function of energy frequency.

One can clearly see from the figure that the utilization of LMFE provides a substantial increase in energy resolution regardless of the electron detection system type.

According to the presented model (5), a growth in the emission current density from the surface due to the utilization of the developed LMFE cathode based on composite nanostructures leads to an increase in the EELS temporal resolution by 17% compared to promising systems with electron sources based on carbon nanotubes.^11,12

From the results of the analysis carried out according to Ref. 12, it is also clear that the detection threshold strongly depends on the current density of the electron source (6). The performed calculation shows that the use of the developed LMFE cathode makes it possible to reduce the detection threshold of chemical elements with trace concentrations by 9% in comparison with the systems using electron guns based on carbon nanotubes as charge carrier sources.

This paper proposes a solution to pressing scientific and technical problems of the design, implementation, investigation, and analysis of novel low-macroscopic-field emission cathodes. The emission characteristics and surface structure of obtained prototypes are investigated. Their utilization in EEL spectroscopy systems for increasing the resolution as well as reducing the detection threshold of chemical elements with trace concentrations is discussed. The obtained scientific results can be summarized as follows:

1. The possibility and expediency of using a LMFE cathode based on a porous matrix functionalized with C₆₀-based carbon nanostructures to increase the resolution and reduce the detection threshold for trace concentration analysis in EEL spectroscopy systems have been substantiated.

2. It is possible to achieve an increase in the EELS system energy resolution by at least 20% via the use of the developed novel LMFE cathode due to the narrowing of the source electrons' energy distribution spectra while maintaining high values of the emission current density in comparison with classical systems using electron guns based on golden field emission cathodes as charge carrier sources.

3. An increase in temporal resolution by 17% and a decrease in the detection threshold of chemical elements with trace concentrations by 9% are attainable using the developed LMFE cathode, characterized by a higher emission current density at comparable or lower voltage values than those for promising systems with electron sources based on carbon nanotubes.

4. Even though the solgel technology provides a relatively quick and facile way of synthesizing an island-like⁵⁷ cathode surface demonstrating promising emission characteristics, the stability of this prototype leaves much to be desired. Additional research is required before making a clear statement on the perspective of utilizing such structures as reliable electron sources.

The authors thank the head of ETU “LETI” “Theranostics” laboratory associate professor Spivak Yu. M. for invaluable cooperation in the field of porous silicon matrix synthesis and composite materials realization, associate professor Bobkov A. A. (ETU “LETI” “Theranostics” laboratory) for the solgel material prototypes he provided, associate professor Gerasimov V. I. (Saint Petersburg State Institute of Technology) for the provided aqueous solutions of doped carbon structures based on C₆₀ fullerene, Somov P. A. (“Tescan CIS”) for his help in analyzing the samples by means of electron microscopy, Bizyaev I. S., researcher at the “Self-organizing high-temperature nanostructures” laboratory for his help in measuring and processing the emission characteristics of cathode prototypes, as well as Moshnikov V. A., professor of Micro-and Nanoelectronics Department ETU “LETI” for constructive and effective advice. This research was funded by the Russian Science Foundation (Grant No. 21-19-00139).

The authors have no conflicts to disclose.

R. Smerdov: Conceptualization (lead); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (equal); Validation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (equal). A. Mustafaev: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (lead); Investigation (equal); Methodology (equal); Project administration (lead); Resources (lead); Supervision (lead); Validation (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available within the article.