*Article* **Electron-Beam Sintering of Al2O3-Cr-Based Composites Using a Forevacuum Electron Source**

**Aleksandr Klimov 1,\* , Ilya Bakeev <sup>1</sup> , Anna Dolgova <sup>1</sup> , Efim Oks 1,2, Van Tu Tran <sup>1</sup> and Aleksey Zenin <sup>1</sup>**


**Abstract:** We describe our studies of the influence of Cr content in an Al2O<sup>3</sup> -Cr composite on its thermal and electrical conductivity properties during and after electron-beam sintering in the forevacuum range of pressure. Sintering was carried out using a plasma-cathode forevacuumpressure electron source of an original design, capable of processing non-conducting materials directly. It is shown that the chromium content affects the efficiency of the beam power transfer to the irradiated composite. The efficiency decreases with increasing chromium content. Measurement of the composite's coefficient of thermal conductivity, in the temperature range 50–400 ◦C, shows that it varies almost linearly from 25 W/(m·K) to 68 W/(m·K) as the Cr content in the composite increases from 25% to 75% wt. The electrical conductivity properties after sintering exhibit a non-linear behavior. The conduction activation energy Ea, measured via the dependence of the current through composites of different compositions, is slightly lower than the Al2O<sup>3</sup> band-gap. The addition of metallic Cr results in a disproportionate decrease in Ea, almost by an order of magnitude, from 6.9 eV to 0.68 eV. By varying the chromium content, it is possible to form a material with thermal and electrical conductivities controllable over a wide range.

**Keywords:** pressureless sintering; composite ceramics; electron beam; electron-beam irradiation; sintering; conduction activation energy; Al2O<sup>3</sup> -Cr; thermal conductivity; electrical conductivity; forevacuum pressure region

### **1. Introduction**

Metal–ceramic composites offer numerous advantages. They are materials of high hardness and mechanical strength that can operate at temperatures above 1000 ◦C [1]. Wide commercial use of composites is possible, contingent upon the availability of rapid production methods and a good understanding of their thermal and electrical characteristics. Of the various kinds of ceramic materials, Al2O3-based ceramic is one of the most widely used, possessing high strength, high hardness, and excellent thermal resistance [2–5]. However, its high brittleness restricts its possible applications. The introduction of malleable metal phases to ceramics is an effective method of reducing its brittleness. Metal–ceramic composites obtained in this way acquire not only low brittleness but also often new electrical, optical, magnetic, and thermal properties [6–11]. In the work described here, for the metallic component we select Cr, which has a high melting point (1863 ◦C [12]), high-temperature oxidation resistance, and high-temperature plasticity [13,14]. These properties may be useful in creating thermally, chemically, and mechanically strong metal–ceramic materials. Among the main areas of application of such materials are the coating of jet nozzles and the protective coating of gas furnaces, crucibles, heat shields, etc.

The main methods used for forming composite materials can be divided into two types. The first is selective sintering by an electron or laser beam [15,16], and the second includes hot pressing, [17,18], cold pressing, microwave, thermal [19], and spark plasma

**Citation:** Klimov, A.; Bakeev, I.; Dolgova, A.; Oks, E.; Tran, V.T.; Zenin, A. Electron-Beam Sintering of Al2O3-Cr-Based Composites Using a Forevacuum Electron Source. *Ceramics* **2022**, *5*, 748–760. https:// doi.org/10.3390/ceramics5040054

Academic Editor: Amirhossein Pakseresht

Received: 12 September 2022 Accepted: 11 October 2022 Published: 14 October 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

sintering [20–22]. The sintering of powders involves issues with controlling the heating rate and the uniformity of target heating. Furnace heating and sintering require time-consuming exposure at high temperatures and have low energy efficiency. When a target is irradiated by laser radiation, heating of the material starts from the surface and proceeds to the core unevenly [23]. This temperature gradient leads to uneven distribution of grain size and of density with depth [24]. These problems can be resolved by lowering the heating rate, thereby increasing the overall processing time. Additionally, the efficiency of the laser energy transfer to the target depends on its optical properties, which restricts the range of materials that can be used for sintering.

In the case of target heating by an electron beam, the target optical properties are not important. However, a new problem arises, which is removing the electrical charge carried by the electron beam onto the dielectric target. The charged surface causes electron beam defocusing and diminishes the efficiency of the beam power transfer to the irradiated target [25,26]. This issue can be resolved by use of a forevacuum-pressure plasma-cathode electron source for the e-beam irradiation. Such sources generate electron beams at a pressure of 1–100 Pa, and the negative surface charge is compensated by the positive ion flux from the beam plasma created during beam propagation at such elevated pressure values [27]. We have previously shown [28–30] that the use of electron beams generated by forevacuum plasma electron sources is a useful approach for sintering ceramic compacts. In the work described here, we have used this method for electron-beam sintering of Al2O3-Cr-based composites and have explored the thermal and electrical properties of the Al2O3-Cr composites produced.

#### **2. Materials and Methods**

We used commercially available Al2O<sup>3</sup> powders with particle size 10–30 µm, and Cr powder with particle size 50 µm.

The main parameters pertaining to the experiment are shown in Table 1 [31–34].


**Table 1.** Material parameters at 20 ◦C.

The composites for sintering were produced by mixing ceramic and metal powders in various mass proportions. The sample and the mixture used to produce it were assigned the same nomenclature label. The proportions are shown in Table 2.



Pellets 3 ± 0.1 mm thick and 10 ± 0.1 mm in diameter were formed from the mixtures by uniaxial pressing. The composites were processed in a vacuum chamber equipped with necessary pumping equipment and manipulators; see Figure 1a. A forevacuum plasmacathode electron source [35] was used for composite heating. A beam-focusing system of a special design allowed an electron beam of 0.6 mm in diameter to be generated in the vacuum chamber, under forevacuum conditions at a pressure of 30 Pa (helium). For

sintering, the composite of given composition was placed in the vacuum chamber on a graphite holder of special design. The holder assembly consisted of a graphite crucible with mounting and bracing that minimized heat loss to its fastening elements. The composite heating efficiency was improved by placing a heat-reflecting shield around the graphite holder to reduce heat transfer to the chamber walls by thermal radiation. After installing the composite and evacuating the chamber to a working pressure of 3.0 Pa, the electron source was turned on. Sintering was performed as follows. First, a smooth heating of the composite at constant electron beam energy of 15 keV, by slowly increasing the beam current from 10 to 100 mA, depending of the composite composition; then, exposure to the beam at a constant temperature of 1400 ◦C for 10 minutes; next, cooling by reducing the beam current, followed by turning off the electron source and further cooling in the vacuum chamber for 20 minutes. Since the area of the irradiated surface was much greater than the cross-sectional area of the electron beam at its point of incidence on the composite, the electron beam was rapidly scanned over the composite surface. The scanning was performed using a magnetic deflecting system controlled by a sweep circuit. The scanning frequency was 100 Hz, and the scanned area was 1 <sup>×</sup> 1 cm<sup>2</sup> . elements. The composite heating efficiency was improved by placing a heat-reflecting shield around the graphite holder to reduce heat transfer to the chamber walls by thermal radiation. After installing the composite and evacuating the chamber to a working pressure of 3.0 Pa, the electron source was turned on. Sintering was performed as follows. First, a smooth heating of the composite at constant electron beam energy of 15 keV, by slowly increasing the beam current from 10 to 100 mA, depending of the composite composition; then, exposure to the beam at a constant temperature of 1400 °С for 10 minutes; next, cooling by reducing the beam current, followed by turning off the electron source and further cooling in the vacuum chamber for 20 minutes. Since the area of the irradiated surface was much greater than the cross-sectional area of the electron beam at its point of incidence on the composite, the electron beam was rapidly scanned over the composite surface. The scanning was performed using a magnetic deflecting system controlled by a sweep circuit. The scanning frequency was 100 Hz, and the scanned area was 1 × 1 cm<sup>2</sup> .

plasma-cathode electron source [35] was used for composite heating. A beam-focusing system of a special design allowed an electron beam of 0.6 mm in diameter to be generated in the vacuum chamber, under forevacuum conditions at a pressure of 30 Pa (helium). For sintering, the composite of given composition was placed in the vacuum chamber on a graphite holder of special design. The holder assembly consisted of a graphite crucible with mounting and bracing that minimized heat loss to its fastening

*Ceramics* **2022**, *5,* FOR PEER REVIEW 3

**Figure 1.** Schematic diagram of the experimental setup (**a**) and electrical heater (**b**).

**Figure 1.** Schematic diagram of the experimental setup (**a**) and electrical heater (**b**). The composite surface during sintering was monitored remotely by a RAYTEK 1MH (Raytek Corp., Santa Cruz, CA, USA) infrared pyrometer with measurement range 550– 3000 °С. A tungsten–rhenium thermocouple was used to measure the temperature of the composite non-irradiated side. The thermocouple and the composite surface were in tight The composite surface during sintering was monitored remotely by a RAYTEK 1MH (Raytek Corp., Santa Cruz, CA, USA) infrared pyrometer with measurement range 550–3000 ◦C. A tungsten–rhenium thermocouple was used to measure the temperature of the composite non-irradiated side. The thermocouple and the composite surface were in tight contact with each other. The thermocouple and the pyrometer were calibrated by heating of a thin copper plate. The difference in readings did not exceed 10 ◦C.

contact with each other. The thermocouple and the pyrometer were calibrated by heating of a thin copper plate. The difference in readings did not exceed 10 °С. Current through the composite during its irradiation was measured by replacing the thermocouple with a flat metal electrode (not shown in Figure 1). The sample was placed Current through the composite during its irradiation was measured by replacing the thermocouple with a flat metal electrode (not shown in Figure 1). The sample was placed on this electrode on the irradiated side. The current through the sample was measured by a True-RMS Multimeter 289 (Fluke Corp., Everett, WA, USA), with one sensing wire connected to the electrode and the other grounded.

on this electrode on the irradiated side. The current through the sample was measured by a True-RMS Multimeter 289 (Fluke Corp., Everett, WA, USA), with one sensing wire connected to the electrode and the other grounded. In order to investigate the thermal properties of the composites after sintering, a In order to investigate the thermal properties of the composites after sintering, a model of an electrical heater was made; see Figure 1b. The heater could heat composite materials up to 400 ◦C in a controlled manner. In addition, it was possible to measure the heat flow through the samples and to measure the temperature of the irradiated and non-irradiated surfaces. The temperature was measured using standard chromel–copel thermocouples.

model of an electrical heater was made; see Figure 1b. The heater could heat composite

Sample microstructure and elemental composition were studied using a JEOL JSM-7500FA (JEOL Ltd., Freising, Germany) scanning electron microscope, equipped with a set of add-on units for energy dispersive elemental analysis (EDS) and electron backscatter diffraction (EBSD) (Bruker Nano GmbH, Berlin, Germany). We used the facilities at the TPU Center for Sharing Use, "Nanomaterials and Nanotechnologies", supported by the Ministry of Education and Science of Russia under grant number 075-15-2021-710. with a set of add-on units for energy dispersive elemental analysis (EDS) and electron backscatter diffraction (EBSD) (Bruker Nano GmbH). We used the facilities at the TPU Center for Sharing Use, "Nanomaterials and Nanotechnologies", supported by the Ministry of Education and Science of Russia under grant number 075-15-2021-710. **3. Experimental Results and Analysis**

materials up to 400 °C in a controlled manner. In addition, it was possible to measure the heat flow through the samples and to measure the temperature of the irradiated and non-irradiated surfaces. The temperature was measured using standard chromel–copel

Sample microstructure and elemental composition were studied using a JEOL JSM-7500FA (JEOL Ltd., Freising, Germany) scanning electron microscope, equipped

#### **3. Experimental Results and Analysis** *3.1. Electron-Beam Sintering*

*Ceramics* **2022**, *5,* FOR PEER REVIEW 4

#### *3.1. Electron-Beam Sintering* Focused electron-beam heating in the forevacuum medium enabled us to heat the

thermocouples.

Focused electron-beam heating in the forevacuum medium enabled us to heat the composite surfaces quite easily to temperatures of 800–900 ◦C. The temperature growth rate at constant beam power of 40 W/min was 80 ◦C/min; see Figure 2. composite surfaces quite easily to temperatures of 800–900 °С. The temperature growth rate at constant beam power of 40 W/min was 80 °С/min; see Figure 2.

**Figure 2.** Dependencies of the electron beam power 1 and the temperature of irradiated 2 and **Figure 2.** Dependencies of the electron beam power 1 and the temperature of irradiated 2 and non-irradiated 3 surfaces of the 75A composite, as a function of time.

non-irradiated 3 surfaces of the 75А composite, as a function of time.

The sample temperature growth rate decreased to 26–30 °С/min, depending on the chromium content in the composite. The temperature-increase rate for zero chromium content is 30 °С/min, and the addition of chromium decreases the temperature growth rate to 26 °С/min. This change in the rate of temperature rise is not significant, but is still noticeable. Clearly the thermal parameters, especially the coefficient of thermal conductivity, affect the composite heating. Since chromium has a greater coefficient of thermal conductivity than aluminum oxide, 94 W/(m·K) and, on average, 30 W/(m·K), respectively, the addition of chromium results in greater heat transfer to the graphite crucible The sample temperature growth rate decreased to 26–30 ◦C/min, depending on the chromium content in the composite. The temperature-increase rate for zero chromium content is 30 ◦C/min, and the addition of chromium decreases the temperature growth rate to 26 ◦C/min. This change in the rate of temperature rise is not significant, but is still noticeable. Clearly the thermal parameters, especially the coefficient of thermal conductivity, affect the composite heating. Since chromium has a greater coefficient of thermal conductivity than aluminum oxide, 94 W/(m·K) and, on average, 30 W/(m·K), respectively, the addition of chromium results in greater heat transfer to the graphite crucible and the holder elements, thereby reducing the composite heating. This circumstance should be taken into account when automating and optimizing the process.

#### and the holder elements, thereby reducing the composite heating. This circumstance *3.2. Microstructure and Parameters*

should be taken into account when automating and optimizing the process. After sintering, the samples were cut in half along a diameter and were polished. In order to remove the products of grinding, they were then rinsed in alcohol and distilled water in an ultrasonic bath for 20 minutes. The microstructures of 75A, 50A, and 25A composites are shown in Figure 3. The samples display homogeneous areas with good compaction, as well as pores of various sizes formed at grain boundaries. EDX analysis revealed that the gray areas correspond to Al2O<sup>3</sup> ceramic matrix and the lighter areas are Cr. The size of Al2O<sup>3</sup> grains is 20–40 µm, and the size of Cr grains is 20 to 80 µm. With the increase in Cr content, the Cr grain size increases due to smaller grains amalgamating into bigger grains. compaction, as well as pores of various sizes formed at grain boundaries. EDX analysis revealed that the gray areas correspond to Al2O<sup>3</sup> ceramic matrix and the lighter areas are Cr. The size of Al2O<sup>3</sup> grains is 20–40 µm, and the size of Cr grains is 20 to 80 µm. With the increase in Cr content, the Cr grain size increases due to smaller grains amalgamating into bigger grains. revealed that the gray areas correspond to Al2O<sup>3</sup> ceramic matrix and the lighter areas are Cr. The size of Al2O<sup>3</sup> grains is 20–40 µm, and the size of Cr grains is 20 to 80 µm. With the increase in Cr content, the Cr grain size increases due to smaller grains amalgamating into bigger grains.

After sintering, the samples were cut in half along a diameter and were polished. In order to remove the products of grinding, they were then rinsed in alcohol and distilled water in an ultrasonic bath for 20 minutes. The microstructures of 75А, 50А, and 25А composites are shown in Figure 3. The samples display homogeneous areas with good

After sintering, the samples were cut in half along a diameter and were polished. In order to remove the products of grinding, they were then rinsed in alcohol and distilled water in an ultrasonic bath for 20 minutes. The microstructures of 75А, 50А, and 25А composites are shown in Figure 3. The samples display homogeneous areas with good compaction, as well as pores of various sizes formed at grain boundaries. EDX analysis

*Ceramics* **2022**, *5,* FOR PEER REVIEW 5

*Ceramics* **2022**, *5,* FOR PEER REVIEW 5

*3.2. Microstructure and Parameters*

*3.2. Microstructure and Parameters*

**Figure 3.** SEM image of composite microstructure for samples (**a**) 75A, (**b**) 50A, and (**c**) 25A. **Figure 3.** SEM image of composite microstructure for samples (**a**) 75A, (**b**) 50A, and (**c**) 25A. **Figure 3.** SEM image of composite microstructure for samples (**a**) 75A, (**b**) 50A, and (**c**) 25A.

The elemental composition of the selected areas, obtained by energy-dispersive analysis, is shown in Figure 4. The increase in Cr content corresponds to its content in the initial mixture of powders used for sintering. The elemental composition of the selected areas, obtained by energy-dispersive analysis, is shown in Figure 4. The increase in Cr content corresponds to its content in the initial mixture of powders used for sintering. The elemental composition of the selected areas, obtained by energy-dispersive analysis, is shown in Figure 4. The increase in Cr content corresponds to its content in the initial mixture of powders used for sintering.

**Figure 4.** Contents of O, Al, and Cr in the cross sections of 75А, 50А, and 25А composites. **Figure 4.** Contents of O, Al, and Cr in the cross sections of 75A, 50A, and 25A composites.

Composite parameters before and after electron-beam sintering are shown in Table 3. **Table 3.** Composite parameters before and after sintering. Composite parameters before and after electron-beam sintering are shown in Table 3. Composite parameters before and after electron-beam sintering are shown in Table 3. **Table 3.** Composite parameters before and after sintering.


The maximum increase in density after irradiation, 21%, was for the composite containing 75% aluminum oxide. The minimum increase in density, 9%, was for the 25A sample, with the lowest content of Al2O3. It is apparent that since aluminum oxide has greater shrinkage in the course of sintering, the corresponding samples with greater content of it must have smaller geometric dimensions and, hence, a higher density after sintering. Compared with the sintering of similar composites using the hot-pressing method [36], the porosity value in this work turned out to be higher. This difference may be due to the peculiarity of the electron beam method—sintering without applying pressure. sample, with the lowest content of Al2O3. It is apparent that since aluminum oxide has greater shrinkage in the course of sintering, the corresponding samples with greater content of it must have smaller geometric dimensions and, hence, a higher density after sintering. Compared with the sintering of similar composites using the hot-pressing method [36], the porosity value in this work turned out to be higher. This difference may be due to the peculiarity of the electron beam method—sintering without applying pressure. The mass of all composites changes (decreases) after sintering; see Table 3. A possi-

The maximum increase in density after irradiation, 21%, was for the composite containing 75% aluminum oxide. The minimum increase in density, 9%, was for the 25A

Diameter d, mm before 10.3 10.32 10.23 10.30

after 9.32 9.54 9.69 9.86

before 1.7 1.77 2.13 2.68 after 2.02 2.14 2.43 2.92

*Ceramics* **2022**, *5,* FOR PEER REVIEW 6

Density ρ, g/cm<sup>3</sup>

The mass of all composites changes (decreases) after sintering; see Table 3. A possible reason could be mass evaporation during sintering. However, the sintering temperature of 1400 ◦C is not high enough for melting and evaporating the composite components; see Table 1. To further explore the possibility of mass loss by evaporation, we conducted experiments to study the composition of coatings on witness substrates. Substrates in the form of flat steel disks were placed at a distance of 5 cm from the sintered composite, as shown in Figure 1. After sintering, the sample surface elemental composition was studied by energy-dispersive analysis. According to the composition measurements for three compacts, the substrate coatings contained elemental constituents of the composites, namely chromium and aluminum; see Figure 5. ble reason could be mass evaporation during sintering. However, the sintering temperature of 1400 °С is not high enough for melting and evaporating the composite components; see Table 1. To further explore the possibility of mass loss by evaporation, we conducted experiments to study the composition of coatings on witness substrates. Substrates in the form of flat steel disks were placed at a distance of 5 cm from the sintered composite, as shown in Figure 1. After sintering, the sample surface elemental composition was studied by energy-dispersive analysis. According to the composition measurements for three compacts, the substrate coatings contained elemental constituents of the composites, namely chromium and aluminum; see Figure 5.

**Figure 5.** Contents of Al and Cr in the witness substrate coatings: 75A, 50A, and 25A. **Figure 5.** Contents of Al and Cr in the witness substrate coatings: 75A, 50A, and 25A.

As shown in Figure 5, the coatings contain a significant amount of the composite elements. Thus, for the composite with 75% aluminum oxide (75A), the substrate coating contains over 13% at. aluminum. The witness substrates used for the other composites with a lower aluminum oxide content demonstrate a considerable decrease in aluminum in the coating. For the 50A composite, the aluminum content on the substrate is less than for 75A, by almost a factor of 10. At the same time, the chromium content in the substrate coating increases almost proportionally to the increase in chromium content in the sintered composite—from 1.26% for composite 75А with 25% Cr to 3.56% for composite 25A with 75% Cr. Evidently, these elements appear on the substrate due to evaporation from the composite surface. Despite the rather low temperature, according to pyrometer As shown in Figure 5, the coatings contain a significant amount of the composite elements. Thus, for the composite with 75% aluminum oxide (75A), the substrate coating contains over 13% at. aluminum. The witness substrates used for the other composites with a lower aluminum oxide content demonstrate a considerable decrease in aluminum in the coating. For the 50A composite, the aluminum content on the substrate is less than for 75A, by almost a factor of 10. At the same time, the chromium content in the substrate coating increases almost proportionally to the increase in chromium content in the sintered composite—from 1.26% for composite 75A with 25% Cr to 3.56% for composite 25A with 75% Cr. Evidently, these elements appear on the substrate due to evaporation from the composite surface. Despite the rather low temperature, according to pyrometer readings, for the evaporation of these elements to occur, such an effect can occur during electronbeam sintering of ceramics employing an electron beam deflecting system. When scanning the ceramic surface, the high power-density electron beam can cause local heating of the surface at the impact point; the beam cross-section at the impact point is less than 1 mm<sup>2</sup> . Since the pyrometer measures the mean value of the composite surface temperature over an area of about 2.5 cm<sup>2</sup> , the local temperature increase at the beam impact point is not

readily registered. Additionally, mass loss from the composite surface can occur due to evaporation of low-melting-point impurities with a content in the aluminum oxide powder used in the experiments that can be as high as 5%. minum oxide powder used in the experiments that can be as high as 5%. Another possible mechanism for heating to the evaporation temperatures of Al2O<sup>3</sup> and Cr could be heating due to the current through the composite bulk, as is the case for

readings, for the evaporation of these elements to occur, such an effect can occur during electron-beam sintering of ceramics employing an electron beam deflecting system. When scanning the ceramic surface, the high power-density electron beam can cause local heating of the surface at the impact point; the beam cross-section at the impact point is

impact point is not readily registered. Additionally, mass loss from the composite surface can occur due to evaporation of low-melting-point impurities with a content in the alu-

. Since the pyrometer measures the mean value of the composite surface

, the local temperature increase at the beam

*Ceramics* **2022**, *5,* FOR PEER REVIEW 7

temperature over an area of about 2.5 cm<sup>2</sup>

less than 1 mm<sup>2</sup>

Another possible mechanism for heating to the evaporation temperatures of Al2O<sup>3</sup> and Cr could be heating due to the current through the composite bulk, as is the case for the flashsintering technique. In this technique, a constant electric field of 100–150 V/m is established between the two opposite surfaces of the sintered sample, which is simultaneously heated to a high temperature. The electrical conduction, arising as a result of the temperature increase, leads to Joule heating of the sample. This significantly reduces the ceramic sintering time. the flash-sintering technique. In this technique, a constant electric field of 100–150 V/m is established between the two opposite surfaces of the sintered sample, which is simultaneously heated to a high temperature. The electrical conduction, arising as a result of the temperature increase, leads to Joule heating of the sample. This significantly reduces the ceramic sintering time.

Measurement of the current through the composite show that it can reach several milliamperes; see Figure 6. Measurement of the current through the composite show that it can reach several milliamperes; see Figure 6.

**Figure 6.** Temperature dependence of current through composites during electron-beam sintering: (1) composite 75A; (2) composite 50A; (3) composite 25A. **Figure 6.** Temperature dependence of current through composites during electron-beam sintering: (1) composite 75A; (2) composite 50A; (3) composite 25A.

The current increases with increasing chromium content, which is directly related to the increase in the composite electrical conductivity. The addition of a metal, as a more electrically conducting material, increases the overall electrical conductivity of the com-The current increases with increasing chromium content, which is directly related to the increase in the composite electrical conductivity. The addition of a metal, as a more electrically conducting material, increases the overall electrical conductivity of the composite. The dependence of the coefficient of electrical conductivity is well-known to be exponential with temperature:

$$\gamma = \gamma\_0 \cdot \exp\left(-\frac{\Delta E\_d}{kT}\right) \tag{1}$$

*E a*

(1)

exp

 

where

0 *kT* = − *γ*<sup>0</sup> is an electrical conductivity of the conductor/dielectric, S/m;

where *γ*<sup>0</sup> is a temperature-independent coefficient determined by the properties of the conductor/dielectric, S/m;

*γ<sup>0</sup>* is an electrical conductivity of the conductor/dielectric, S/m; *k* is Boltzmann's constant, J/K;

*γ<sup>0</sup>* is a temperature-independent coefficient determined by the properties of the *T* is the temperature of the irradiated composite surface, K; and

conductor/dielectric, S/m; *E<sup>a</sup>* is the conduction activation energy, eV.

*k* is Boltzmann's constant, J/K; *Т* is the temperature of the irradiated composite surface, K; and *E<sup>а</sup>* is the conduction activation energy, eV. The conduction current in semiconductors, of which aluminum oxide may be referred to as a particular kind, depends on the coefficient of electrical conductivity, as well as on the field strength in the semiconductor. Assuming that the field strength is determined by the difference of potentials between the irradiated and non-irradiated surfaces, one can write an equation for evaluating the value of the conduction current flowing through the composite as a function of temperature during electron-beam irradiation: *S* is the composite base area, m<sup>2</sup> ; Δ*φ* is the difference of potentials between the irradiated and non-irradiates surfaces

*I S*

*Ceramics* **2022**, *5,* FOR PEER REVIEW 8

$$I\_{\gamma} = \frac{\Delta\varphi}{h} \cdot \mathbf{S} \cdot \gamma\_0 \cdot \exp\left(-\frac{E\_a}{kT}\right) \tag{2}$$

The conduction current in semiconductors, of which aluminum oxide may be referred to as a particular kind, depends on the coefficient of electrical conductivity, as well as on the field strength in the semiconductor. Assuming that the field strength is determined by the difference of potentials between the irradiated and non-irradiated surfaces, one can write an equation for evaluating the value of the conduction current flowing through the composite as a function of temperature during electron-beam irradiation:

> 0 exp

*h kT*

 = −

*E a*

(2)

where sites with different elemental contents. Assuming that the temperature dependence of the

where

*S* is the composite base area, m<sup>2</sup> ; potential of the composite surface is not as strong as that of the coefficient of electrical

∆*ϕ* is the difference of potentials between the irradiated and non-irradiates surfaces of the composite, V; and conductivity and plotting the graphs ( ) 1 ln *I f T* <sup>=</sup> using the experimental data of

*h* is the composite thickness.

of the composite, V; and

Expression (2) allows the conduction activation energy to be estimated for composites with different elemental contents. Assuming that the temperature dependence of the potential of the composite surface is not as strong as that of the coefficient of electrical conductivity and plotting the graphs ln(*I*) = *f* 1 *T* using the experimental data of Figure 4, one can determine the activation energy from the slope of the straight lines obtained. The dependencies ln(*I*) = *f* 1 *T* , plotted for the three composites over the temperature range 1000–1400 ◦C, where a noticeable increase in current is observed, are shown in Figure 7. As can be seen, the experimental points of the logarithm of the current from the inverse temperature fit into a linear dependence. From one perspective, this serves as an argument in favor of the chosen mechanism for increasing electrical conductivity with increasing temperature and the correctness of choosing Formula (1) for theoretical estimates. Figure 4, one can determine the activation energy from the slope of the straight lines obtained. The dependencies ( ) 1 ln *I f T* <sup>=</sup> , plotted for the three composites over the temperature range 1000–1400 °С, where a noticeable increase in current is observed, are shown in Figure 7. As can be seen, the experimental points of the logarithm of the current from the inverse temperature fit into a linear dependence. From one perspective, this serves as an argument in favor of the chosen mechanism for increasing electrical conductivity with increasing temperature and the correctness of choosing formula 1 for theoretical estimates.

**Figure 7.** Temperature dependencies ( ) 1 ln *I f T* <sup>=</sup> for composites of different compositions: (1) **Figure 7.** Temperature dependencies ln(*I*) = *f* 1 *T* for composites of different compositions: (1) composite 75A; (2) composite 50A; (3) composite 25A.

composite 75A; (2) composite 50A; (3) composite 25A. The fact that the experimental data fit well to straight lines indirectly indicates that the assumption not to take into account the change in potential on the composite surface over the given temperature range is correct.

The fact that the experimental data fit well to straight lines indirectly indicates that the assumption not to take into account the change in potential on the composite surface The calculated values of the conduction activation energy *E<sup>a</sup>* are given in Table 4.

over the given temperature range is correct.


**Table 4.** Conduction activation energy for composites of different compositions.

The obtained values of *E<sup>a</sup>* for the Cr-containing composites are slightly lower than the Al2O<sup>3</sup> band gap, which is predictable, since it was a metal that was added to the composite. The addition of metallic Cr results in a disproportional decrease in *Ea*. Thus, for the composite with 25% chromium content, the conduction activation energy, or band gap, decreases compared to pure Al2O3, from an average value of 6.9 eV (see Table 2) to 3.1 eV, i.e., by more than a factor of two. The addition of 75% Cr leads to a further decrease in *Ea*, down to 0.68 eV, which is almost by an order of magnitude.

As shown in [37], the electrical conductivity of aluminum oxide-based composites can be adjusted, when reinforced with conductive or semi-conductive phases (such as silicon carbide, for example), added in an amount at which they penetrate into an insulating aluminum oxide matrix. After sintering, such a composite can be used in many industries. The main factors affecting the electrical properties of composites with reinforced semiconductor phases are the volume fraction of SiC and the content of other impurities. The addition of SiC improves the electrical conductivity, which increases with an increase in the volume fraction of SiC [37]. Thus, in a composite with 20 vol.% SiC, the conductivity of 4.05 <sup>×</sup> <sup>10</sup>−<sup>2</sup> <sup>S</sup>·m−<sup>1</sup> was measured, which is an increase of four orders of magnitude compared to the reference monolithic alumina (7.80 <sup>×</sup> <sup>10</sup>−<sup>6</sup> <sup>S</sup>·m−<sup>1</sup> ).

A rather strong dependence of the composite electrical conductivity on the content of metal phase has been observed [38], when adding Ti to an Al2O<sup>3</sup> ceramic matrix. Specific electrical resistance, with the addition of 20% vol. Ti, decreases from 10<sup>12</sup> Ohm·m to <sup>10</sup>−2–10−<sup>3</sup> Ohm·m, and the fall is rather abrupt. The authors have explained this by the formation of conducting paths through the composite bulk, due to the melting of fine grains of Ti. In the present work, the changes are not so severe, which may be related to pressureless sintering. Chromium grains combine without forming conducting paths throughout the composite volume; see Figure 3.

The electrical parameters of the Al2O3-Cr composite can be controlled over a fairly wide range.

#### *3.3. Composite Thermal Conductivity*

Thermal conductivity is an important parameter in such applications of Al2O<sup>3</sup> ceramics as high-temperature structural components, refractories, gas burners, wear parts, and cutting tools. To reduce thermal shock, the thermal conductivity of the composite in all these applications should be as high as possible. It can be expected that Cr particles improve the thermal conductivity of Al2O3-based composites due to the inherent high thermal conductivity of Cr.

To measure the thermal conductivity, the sintered composites were placed in a heating device with a fixed heater temperature T<sup>1</sup> and a temperature T<sup>2</sup> on the composite side not subject to irradiation; see Figure 1b. The coefficient of thermal conductivity *λ* was determined using the expression:

$$
\lambda = \frac{\mathbb{Q} \cdot h}{\Delta T \cdot \mathbb{S}} \tag{3}
$$

where *λ* is the coefficient of thermal conductivity, W/(m·K);

*Q* is the heat flux through the composite, W/m<sup>2</sup> ;

∆*T* is the difference of temperatures: T<sup>1</sup> − T2, ◦C; and

*S* is the composite surface area, m<sup>2</sup> .

The obtained value of the coefficient of thermal conductivity corresponds to the average temperature ∆*T*/2.

The measured thermal conductivity over the temperature range 50–400 ◦C is shown in Figure 8. As seen, the coefficient of thermal conductivity decreases with increasing

temperature for composites of any composition, as reported in the literature [39]. The general pattern here is as follows: the thermal conductivity of ceramics of a crystalline structure, especially an oxide, with an increase in temperature, as a rule, drops significantly [40]. This is based on the idea of heat transfer in solid non-metallic bodies by thermal elastic waves—phonons. The thermal conductivity of the composite is closely related to their microstructure and depends on the free path length of the phonons: the degree of disturbance of the harmonic oscillations of heat waves during their passage through a given substance. Phonons are also known to interact with lattice defects, grain boundaries, and other microstructure defects. The presence of a metallic phase in the form of chromium inclusions leads to a higher porosity value, characteristic of composites and, as a result, negatively affect thermal conductivity. The resulting internal stresses in composites also lead to a decrease in thermal conductivity [41]. However, despite these negative factors, the thermal conductivity of the composite increases with an increase in the chromium content and is still higher than that of pure aluminum oxide. thermal elastic waves—phonons. The thermal conductivity of the composite is closely related to their microstructure and depends on the free path length of the phonons: the degree of disturbance of the harmonic oscillations of heat waves during their passage through a given substance. Phonons are also known to interact with lattice defects, grain boundaries, and other microstructure defects. The presence of a metallic phase in the form of chromium inclusions leads to a higher porosity value, characteristic of composites and, as a result, negatively affect thermal conductivity. The resulting internal stresses in composites also lead to a decrease in thermal conductivity [41]. However, despite these negative factors, the thermal conductivity of the composite increases with an increase in the chromium content and is still higher than that of pure aluminum oxide.

*Ceramics* **2022**, *5,* FOR PEER REVIEW 10

erage temperature Δ*T*/2.

The obtained value of the coefficient of thermal conductivity corresponds to the av-

The measured thermal conductivity over the temperature range 50–400 °С is shown

in Figure 8. As seen, the coefficient of thermal conductivity decreases with increasing temperature for composites of any composition, as reported in the literature [39]. The general pattern here is as follows: the thermal conductivity of ceramics of a crystalline structure, especially an oxide, with an increase in temperature, as a rule, drops significantly [40]. This is based on the idea of heat transfer in solid non-metallic bodies by

**Figure 8.** Temperature dependence of thermal conductivity for composites of different compositions: (1) composite 75A; (2) composite 50A; (3) composite 25A. **Figure 8.** Temperature dependence of thermal conductivity for composites of different compositions: (1) composite 75A; (2) composite 50A; (3) composite 25A.

The coefficient of thermal conductivity of the composite with 75% content of Al2O<sup>3</sup> is 25 W/m·K and does not differ significantly from that of pure Al2O<sup>3</sup> at the same temperature [42]. The values of thermal conductivity measured at room temperature are somewhat lower than the data given in the literature [43] and are measured for mono-cast Al2O<sup>3</sup> (28–30 W/m K). A possible reason is the greater porosity of the materials obtained in this work. With the addition of Cr, the composite conductivity rises almost proportionally to the content of Cr. Thus, for the composite with 75% content of Cr, i.e., three The coefficient of thermal conductivity of the composite with 75% content of Al2O<sup>3</sup> is 25 W/m·K and does not differ significantly from that of pure Al2O<sup>3</sup> at the same temperature [42]. The values of thermal conductivity measured at room temperature are somewhat lower than the data given in the literature [43] and are measured for mono-cast Al2O<sup>3</sup> (28–30 W/m·K). A possible reason is the greater porosity of the materials obtained in this work. With the addition of Cr, the composite conductivity rises almost proportionally to the content of Cr. Thus, for the composite with 75% content of Cr, i.e., three times as much compared to that in the 25% Cr composite, the conductivity increases from 25 to 68 W/m·K. For both Al2O<sup>3</sup> and Cr, the thermal conductivity decreases with temperature [44], as is reflected in Figure 8. Compared with the data of [45], the thermal conductivity of the composite remains at a high level and does not decrease to values below 15 W/m·K.

#### times as much compared to that in the 25% Cr composite, the conductivity increases from **4. Conclusions**

15 W/m·K.

**4. Conclusions**

25 to 68 W/m·K. For both Al2O<sup>3</sup> and Cr, the thermal conductivity decreases with temperature [44], as is reflected in Figure 8. Compared with the data of [45], the thermal conductivity of the composite remains at a high level and does not decrease to values below Electron-beam irradiation allows Al2O3-Cr-based composites to be sintered at a temperature of 1400 ◦C. The complete sintering cycle, including heating and cooling, is no longer than 50 minutes. By varying the Cr content, one can change the electrical and thermal conductivity properties of the composite. In this case, the thermal conductivity in

Electron-beam irradiation allows Al2O3-Cr-based composites to be sintered at a

temperature of 1400 °С. The complete sintering cycle, including heating and cooling, is no longer than 50 minutes. By varying the Cr content, one can change the electrical and thermal conductivity properties of the composite. In this case, the thermal conductivity in

the temperature range of 20–400 ◦C varies directly proportionally to the Cr content and inversely proportionally to temperature. The thermal conductivity increases from 25 to 68 W/m·K, when the Cr content increases from 25% to 75%, and decreases with increasing temperature, especially for composites with higher Al2O<sup>3</sup> content.

The electrical conductivity properties, illustrated by the current through the composite and the conduction activation energy, depend on the Cr content nonlinearly. The addition of 75% Cr to the composite decreases the Al2O3-Cr activation by an order of magnitude compared to that of Al2O3, from 6.9 to 0.68 eV. At the same time, the addition of 50% Cr reduces this energy only by a factor of two, to 2.5 eV. In general, by varying the chromium content, it is possible to produce materials with values of electrical conductivity controllable over orders of magnitude and thermal conductivity controllable within range limits differing by almost a factor of two.

**Author Contributions:** Conceptualization, A.K. and E.O.; methodology, A.Z.; software, A.D.; validation, A.K., A.Z., and I.B.; formal analysis, A.K.; investigation, A.Z. and V.T.T.; resources, E.O.; data curation, A.K.; writing—original draft preparation, A.K.; writing—review and editing, E.O.; visualization, I.B.; supervision, E.O.; project administration, A.K.; funding acquisition, A.K. and V.T.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work was supported by grant FEWM-2020-0038 from the Ministry of Science and Higher Education of the Russian Federation, and a study of coatings composition was funded by research project 20-38-90184 from RFBR.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** Special thanks to Ian Brown (Berkeley Lab) for the English correction and helpful discussion.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**


**Yury Yushkov 1,2 , Efim Oks 1,2, Andrey Kazakov <sup>2</sup> , Andrey Tyunkov 1,2 and Denis Zolotukhin 2,\***

> 1 Institute of High Current Electronics SB RAS, 634055 Tomsk, Russia


**Abstract:** In this study, fore-vacuum plasma electron beam sources were used to deposit a few micronthick boron coatings on A284 and ZrNb1 alloys and modify their surfaces. The coating deposition rate with a continuous 1 kW electron beam that evaporated the boron target at a distance of 10 cm was 0.5 µm/min, and the boron coating density was 2.2 g/cm<sup>3</sup> . Based on the comparison of data on the mass-to-charge composition, beam plasma density, and coating parameters, the contribution of the plasma phase of the evaporated material to the growth of coatings was greater than that of the vapor phase. Using the scanning electron and atomic force microscopy techniques, surface modification by repeated electron beam pulses with electron energies of 8 and 6 keV and a beam power per pulse of 2 J/cm<sup>2</sup> and 2.25 J/cm<sup>2</sup> , respectively, transformed a relatively smooth coating surface into a hilly structure. Based on a structural phase analysis of coatings using synchrotron radiation, it was concluded that the formation of the hilly coating structure was due to surface melting under the repeated action of electron beam pulses. The microhardness, adhesion, and wear resistance of coatings were measured, and their corrosion tests are presented herein. The pure boron coatings obtained and studied are expected to be of use in various applications.

**Keywords:** boron coatings; electron beam deposition; fore-vacuum electron source; film properties; electron beam modification

### **1. Introduction**

Boron-based coatings are promising protective surface coatings [1]. They are used to harden the surface of parts and structural materials in the field of mechanical engineering [2]. They are characterized by high hardness, resistance to wear [3], corrosion [4], and high thermal stability [5]. For example, wurtzite boron nitride has comparable hardness to natural diamond [6]. Pure boron thin films are used as materials in electronic [7] and optical devices [8] as well as in protective layers of thermonuclear installations [9]. Moreover, boron coatings of monoisotopic compositions find applications in the nuclear industry; for example, <sup>10</sup>B-based coatings are promising as burnable neutron absorbers in nuclear reactors [10] and as absorber coatings of neutron detectors [11]; they are also used for the delivery of <sup>10</sup>B atoms deposited on the surface of nanosized particles to the malignant neoplasm during boron neutron capture therapy [12], while <sup>11</sup>B-based coatings are used for the <sup>11</sup>B aneutronic fusion of protons and <sup>11</sup>B atomic nuclei, which may be an alternative to deuterium and tritium fusion [13]. Hence, developing techniques to deposit boron coatings, investigating the properties of boron coatings, and revealing the interrelationship between these properties and conditions that brought about their formation are crucial.

Boriding is an industrial technique that has been used widely for decades to develop boron-containing layers on metal and alloy surfaces. In this process, boride atoms are diffused into the surface of a metal component, resulting in the formation of metal borides in the surface layer, thereby increasing the surface hardness and wear resistance. Conventional boriding can be achieved in a solid, liquid, or gaseous [14,15] medium [16,17].

**Citation:** Yushkov, Y.; Oks, E.; Kazakov, A.; Tyunkov, A.; Zolotukhin, D. Electron-Beam Synthesis and Modification and Properties of Boron Coatings on Alloy Surfaces. *Ceramics* **2022**, *5*, 706–720. https://doi.org/10.3390/ ceramics5040051

Academic Editors: Amirhossein Pakseresht and Kamalan Kirubaharan Amirtharaj Mosas

Received: 14 September 2022 Accepted: 8 October 2022 Published: 10 October 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

However, there are disadvantages to using it, including the high energy consumption required for heating parts and electrolysis of boron-containing media, the long duration of surface diffusive saturation with boron, the use of hazardous and toxic substances, and environmental pollution.

Vacuum-plasma methods, such as magnetron sputtering [18,19] or cathodic arc deposition [20–22], can be used as alternatives for creating boron surface layers owing to their eco-friendliness because the equipment of this type generates a boron flux from a consumable boron-containing solid cathode during an electric discharge in a vacuum chamber. Moreover, there is no heating requirement for the surface because the main process involves the deposition of boron atoms or ions onto the surface. Another advantage of these methods over conventional boriding is the short process duration, which is determined by the flux intensity. The maximum flux intensity is limited by thermal stability in the discharge of the consumed boron-containing cathode. At maximum discharge parameters, the deposition rate of boron-containing coatings at a characteristic distance of 10 cm from the cathode can reach approximately 20–30 nm/min [19] for magnetron sputtering and approximately 100 nm/min for arc deposition [22].

A method for depositing boron coating has been proposed based on the electron beam evaporation of a boron solid target using an electron beam at a fore-vacuum pressure [23]. Briefly, a boron or boron-containing ceramic target is heated locally by an electron beam to the melting temperature and the melted surface area undergoes an intense evaporation. The flux of boron vapor, partially ionized by the beam, deposits onto the substrate surface forming a boron-containing coating [24]. The maximum temperature of the boroncontaining target is limited in this case not by its thermal resistance, as in magnetron or vacuum arc deposition of boron coatings [18–22] but by the temperature of intense boiling of boron, at which an unwanted flux of droplets occurs because of the splashing of the molten target material. Thus, the boron coating deposition rate under this method is much higher than that under magnetron sputtering or arc deposition and can reach up to 1 µm/min [24], which provides a higher coating production performance. This work aims to further develop this method using an additional action of a wide-aperture electron beam in the fore-vacuum on the deposited boron coating. The characteristics and properties of boron coatings are studied by surface analysis, which includes structural phase analysis with synchrotron radiation generated by the VEEP-2 electron storage ring based in the Siberian Center for Synchrotron and Terahertz Radiation at G. I. Budker Institute of Nuclear Physics, SB of RAS [25].

#### **2. Experimental Setup and Diagnostics**

Figure 1 illustrates the setup for depositing boron coatings by electron-beam heating and evaporation of a crystal boron target. The target was evaporated using a fore-vacuum plasma electron source operating in continuous mode [26]. In such a source, electrons are extracted from a hollow cathode glow discharge plasma. The discharge current varies from 100 to 500 mA, while the discharge burning voltage, depending on the discharge current and gas pressure, varies from 200 to 500 V. At the maximum discharge current, the source is capable of generating electron beams with a current of up to 200 mA and energy of up to 20 keV, while the beam can be focused on the target surface by the magnetic field of the focusing system to a diameter of 3 mm. The electron beam current is controlled by the discharge current. The average power density in the electron beam focal spot on the target surface is 500 W/mm<sup>2</sup> , which is several times greater than the radiation power density on the sun's surface (approximately 60 W/mm<sup>2</sup> ). Hence, the source is capable of melting and evaporating the surface of a target composed of any refractory material.

A 4-mm-thick, 1 × 1 cm2 boron plate target fabricated using the hot pressing of 99.6%

pure boron crystals 1–10 µm in size was placed at the bottom of the vacuum chamber on

a carbon crucible. The average density of the target due to the pores between crystals was

1.2 g/cm3, which is two times less than the density of amorphous boron. Because the

pores occupied 50% of the target volume, the effective surface-to-volume ratio for the

target material was about 3 × 103 cm-1 and the total effective surface area of the target was

about 0.1 m2. The electron beam toward the boron target was transported through the

vacuum chamber filled with 99.9% pure nitrogen at a pressure of 10 Pa. An ISP-500C

helical mechanical oil-free fore-vacuum pump with a pumping rate of 500 *l*/min was used

to maintain the vacuum in the chamber. Prior to the experiment, the vacuum chamber

**Figure 1.** Schematic of the electron-beam synthesis of boron-based coatings.

**Figure 1.** Schematic of the electron-beam synthesis of boron-based coatings. Because boron is a wide-gap semiconductor, its specific resistivity is as high as 1 MOhm × cm at a normal temperature of 20 °С [27], which is insufficient for the complete drain of the electron beam charge. Therefore, at the initial heating of the boron target, its negative surface charge introduced by the electron beam was neutralized by a flux of positive ions from the beam plasma [28]. As the beam heated up the boron target, its A 4-mm-thick, 1 <sup>×</sup> 1 cm<sup>2</sup> boron plate target fabricated using the hot pressing of 99.6% pure boron crystals 1–10 µm in size was placed at the bottom of the vacuum chamber on a carbon crucible. The average density of the target due to the pores between crystals was 1.2 g/cm<sup>3</sup> , which is two times less than the density of amorphous boron. Because the pores occupied 50% of the target volume, the effective surface-to-volume ratio for the target material was about 3 <sup>×</sup> <sup>10</sup><sup>3</sup> cm−<sup>1</sup> and the total effective surface area of the target was about 0.1 m<sup>2</sup> . The electron beam toward the boron target was transported through the vacuum chamber filled with 99.9% pure nitrogen at a pressure of 10 Pa. An ISP-500C helical mechanical oil-free fore-vacuum pump with a pumping rate of 500 *l*/min was used to maintain the vacuum in the chamber. Prior to the experiment, the vacuum chamber was evacuated to a residual pressure of 1 Pa.

specific resistivity decreased to 0.1 Ohm × cm at a temperature of about 750 °С. At this Because boron is a wide-gap semiconductor, its specific resistivity is as high as 1 MOhm × cm at a normal temperature of 20 ◦C; [27], which is insufficient for the complete

resistivity, the effect of the target surface charging by the beam becomes immaterial be-

two stages, the parameters of which were determined experimentally [24]. Initially, the

electron beam with a current of 80 mA at an accelerating voltage of up to 5 kV heated up

the target to a temperature of approximately 900 °С for 70 s. Subsequently, the acceler-

ating voltage was increased slowly up to 9 kV for 1 min; in this case, there was an ob-

served local melting of the target surface at the focal spot of the beam. Target evaporation

occurred and the boron coating was deposited on the sample surface at a temperature of

the melting pool of approximately 2400 °С, and after increasing the accelerating voltage

To prevent the boron target from thermal shock destruction, the target was heated in

to 10 kV and the beam current to 100 mA.

drain of the electron beam charge. Therefore, at the initial heating of the boron target, its negative surface charge introduced by the electron beam was neutralized by a flux of positive ions from the beam plasma [28]. As the beam heated up the boron target, its specific resistivity decreased to 0.1 Ohm × cm at a temperature of about 750 ◦C. At this resistivity, the effect of the target surface charging by the beam becomes immaterial because of the electron beam current running through the boron target volume.

To prevent the boron target from thermal shock destruction, the target was heated in two stages, the parameters of which were determined experimentally [24]. Initially, the electron beam with a current of 80 mA at an accelerating voltage of up to 5 kV heated up the target to a temperature of approximately 900 ◦C for 70 s. Subsequently, the accelerating voltage was increased slowly up to 9 kV for 1 min; in this case, there was an observed local melting of the target surface at the focal spot of the beam. Target evaporation occurred and the boron coating was deposited on the sample surface at a temperature of the melting pool of approximately 2400 ◦C, and after increasing the accelerating voltage to 10 kV and the beam current to 100 mA.

The samples with polished surfaces on which boron coatings were applied were disks of A284 steel and a special reactor alloy ZrNb1 with a hexagonal crystal lattice; they had a 2 cm diameter and 0.5 cm thickness. The samples were placed at an angle of 30◦ relative to the electron beam propagation axis. The sample surfaces were normally oriented to the center of the evaporated target. A high-performance Raytek-MM1MH optical pyrometer was used to monitor the temperature of the sample surface during the deposition process; the temperature did not exceed 500 ◦C. Plasma composition during the coating was controlled with a quadrupole mass analyzer based on an RGA-300 residual gas analyzer. We have produced several dozen samples with boron coating thicknesses of 1–7 µm throughout the course of our research. Herein, for comparison, we analyze coatings fabricated under the same conditions, such as a distance of 10 cm from the beam-heated boron target to the sample surface and an overall coating deposition time of 6 min and 10 s. Considering that the first two heating stages under which the deposition did not occur took 1 min and 10 s, the direct deposition time was 5 min.

The deposition rate of boron coatings by a 1 kW electron beam was measured with an MII-4 interferometer and an MNL-1 interference microscope-profilometer. The thickness of the deposited boron coatings was about 2.1 µm. Thus, under the experimental conditions, the coating growth rate was about 0.5 µm/min. The weight gain because of boron coating deposition, measured with a VL-220M analytical balance with a precision of ±10 µg, was 1.45 mg. Considering a sample surface area of 3.13 cm<sup>2</sup> , the density of boron coatings was estimated to be *<sup>ρ</sup>* <sup>≈</sup> 2.2 g/cm<sup>3</sup> , which is close to the density of crystalline boron of 2.34 g/cm<sup>3</sup> .

Figure 2 shows a setup for the modification of boron coatings obtained with a pulsed electron beam. The pulsed electron beam was generated by a fore-vacuum plasma-cathode electron source based on an arc discharge [29]. The electron source was mounted on the vacuum chamber, which was evacuated with an ISP-500C pump. The working gas was helium at a pressure of 10 Pa. The samples were placed on a movable grounded holder, which allowed several samples to be sequentially irradiated after the chamber evacuation. To prevent the sample surface from beam exposure, a protective stainless-steel screen was applied on top of one of the samples. The treatment of boron-coated samples was performed in two regimes through a series of 300 pulses with a repetition rate of 2 pulses per second (p.p.s.). The length of each pulse was τ*<sup>e</sup>* = 500 µs. The beam diameter was 7 cm. In regime 1, the treatment was performed at a beam current amplitude of 20 A and an electron energy of 8 keV; in regime 2, it was performed at a beam current amplitude of 30 A and an electron energy of 6 keV. The electron current density *j<sup>e</sup>* on the boron coating surface and the energy per pulse *P<sup>e</sup>* were 0.5 A/cm<sup>2</sup> and 2 J/cm<sup>2</sup> for regime 1, and 0.75 A/cm<sup>2</sup> and 2.25 J/cm<sup>2</sup> for regime 2, respectively. The pulsed beam parameters for modifying the boron coatings were selected empirically: on the one hand, they should lead to a noticeable

change in the surface morphology; on the other hand, they should be soft enough to prevent the coating from cracking, or its evaporation.

**Figure 2.** Setup for the modification of boron coatings with a pulsed electron beam. **Figure 2.** Setup for the modification of boron coatings with a pulsed electron beam.

The morphology of the final coatings was studied using a Hitachi S3400N scanning electron microscope and a Solver P47 atomic force microscope. Elemental composition was analyzed with a Bruker X'Flash 5010 energy dispersive spectrometer. The surface hardness of coatings was determined using the micro-Vickers technique. A square-section diamond indenter with a dihedral angle of 136° acted on the sample surface at various points with a constant load of 100 g while the penetration depth and indentation area were recorded. The phase composition of coatings was measured at the The morphology of the final coatings was studied using a Hitachi S3400N scanning electron microscope and a Solver P47 atomic force microscope. Elemental composition was analyzed with a Bruker X'Flash 5010 energy dispersive spectrometer. The surface hardness of coatings was determined using the micro-Vickers technique. A square-section diamond indenter with a dihedral angle of 136◦ acted on the sample surface at various points with a constant load of 100 g while the penetration depth and indentation area were recorded. The phase composition of coatings was measured at the Synchrotron Radiation Station for High-Precision X-ray Diffraction Studies of Materials (also called the "Anomalous Scattering" station) on beamline No. 2 of the VEPP-3M electron storage ring at the Siberian Synchrotron Radiation Center (Budker Institute of Nuclear Physics, SB of RAS).

Synchrotron Radiation Station for High-Precision X-ray Diffraction Studies of Materials (also called the "Anomalous Scattering" station) on beamline No. 2 of the VEPP-3M electron storage ring at the Siberian Synchrotron Radiation Center (Budker Institute of Nuclear Physics, SB of RAS). For the evaluation of the adhesive properties of the coatings, the scratch method was used. The method consisted of applying a force *F* (linearly growing with time) to the diamond indenter with its simultaneous uniform displacement along the coating surface. At the critical load *Fc*, the coating begins to break down. The critical load *F<sup>c</sup>* is determined using the sensors of acoustic emission and friction force, indenter immersion depth, indenter loading force, and optical microscopy.

For the evaluation of the adhesive properties of the coatings, the scratch method was used. The method consisted of applying a force *F* (linearly growing with time) to the diamond indenter with its simultaneous uniform displacement along the coating surface. Adhesion can be characterized by a parameter *G* (specific peel work). The calculation formula connecting the parameter *G* with the critical lateral load *F<sup>c</sup>* at the beginning of the film detachment from the substrate is as follows:

$$\mathbf{G} = (F\_{\mathcal{L}}) \mathbf{2}d / \pi(r\_{\mathcal{L}}) \mathbf{4}E\_{IT}$$

using the sensors of acoustic emission and friction force, indenter immersion depth, in-

tion formula connecting the parameter *G* with the critical lateral load *Fc* at the beginning

( )2 ( )4 , *G F d r E c c IT*

The wear resistance of the obtained coatings was measured using a Pinon Disc and

where *d* is the film thickness; *rc* is the radius of the contact spot at the moment of peeling; *EIT* is the Young's modulus of the substrate material. The determination of adhesion was

Oscillating TRIBO tester (France) using the "ball on disk" method. The sample surface was pressed by a tungsten carbide spherical tip with a load of 2 N. The coefficient of friction was determined by measuring the deflection of the lever. The wear rate was

*V* = 2π*RA*/*FL*,

Adhesion can be characterized by a parameter *G* (specific peel work). The calcula-

carried out using a Micro-Scratch Tester MST-S-AX-0000 device.

denter loading force, and optical microscopy.

calculated by the formula:

where *d* is the film thickness; *r<sup>c</sup>* is the radius of the contact spot at the moment of peeling; *EIT* is the Young's modulus of the substrate material. The determination of adhesion was carried out using a Micro-Scratch Tester MST-S-AX-0000 device.

The wear resistance of the obtained coatings was measured using a Pinon Disc and Oscillating TRIBO tester (France) using the "ball on disk" method. The sample surface was pressed by a tungsten carbide spherical tip with a load of 2 N. The coefficient of friction was determined by measuring the deflection of the lever. The wear rate was calculated by the formula:

$$V = 2\pi R A / FL\_\star$$

where *R* is the track radius, µm; *A* is the cross-sectional area of the wear groove, µm<sup>2</sup> ; *F* is the value of the applied load, N; *L* is the distance traveled by the ball, m.

#### **3. Results and Discussion**

During the target heating and coating deposition, the ion composition of the plasma was measured with a quadrupole mass spectrometer based on an RGA-300 residual gas analyzer [30]. In our previous work [31], at an electric beam power of 0.4 kW, which corresponds to the preheating of the target at a beam current of 80 mA and an electron energy of 5 keV, the density of such a plasma is about 3 <sup>×</sup> <sup>10</sup><sup>10</sup> cm−<sup>3</sup> . Initially, at the target heating, in addition to nitrogen ions, a significant number of ions, up to 50% of the total ions, were produced through water vapor ionization processes: HO<sup>+</sup> , H2O<sup>+</sup> , and water dissociation products O<sup>+</sup> , H<sup>3</sup> + , H<sup>2</sup> + , and H<sup>+</sup> . The sources of water molecules were the walls of the vacuum chamber with a surface area of approximately 1 m<sup>2</sup> , the overall effective surface of the boron crystalline target with dimensions 1 × 1 × 0.4 cm, and the total area of about 0.1 m<sup>2</sup> . The appearance of water molecules on the walls and in the target was due to their exposure to the atmosphere prior to the experiments. With the electron beam power increasing to 0.7 kW, the surface temperature of the boron target at the beam focal spot reached 2100 ◦C; this started the melting process, and the plasma spectrum recorded traces of <sup>10</sup>B <sup>+</sup> and <sup>11</sup>B + ions. At this temperature, water molecules apparently evaporated from the target and the heated walls, resulting in the appearance of peaks of water vapor ions and their derivatives.

A further increase in the electron beam power of up to 1 kW led to an increase in the target temperature at the focal spot to 2400 ◦C. Consequently, a brightly glowing melt area about 4 mm in diameter formed on the target surface, from which an intense evaporation of boron occurred. In this case, the plasma density increased to approximately 1.6 <sup>×</sup> <sup>10</sup><sup>11</sup> cm−<sup>3</sup> . The fraction of boron ions in this plasma, evaluated by the height of its ion peaks in the spectrometer signals, was approximately 75%, and their total density in the plasma was about 1.2 <sup>×</sup> <sup>10</sup><sup>11</sup> cm−<sup>3</sup> . The ratio of <sup>10</sup>B + to <sup>11</sup>B + isotopes in the plasma at a beam power of 1 kW was 1:4, which is close to their natural ratio. Thus, the concentrations of <sup>10</sup>B <sup>+</sup> and <sup>11</sup>B + in the beam plasma were 2.4 <sup>×</sup> <sup>10</sup><sup>10</sup> and 9.6 <sup>×</sup> <sup>10</sup><sup>10</sup> cm−<sup>3</sup> , respectively.

Because boron coatings are formed from two-phase states, namely, plasma and boron vapor, it is important to know the contribution of each phase in coating formation. Because the thickness, elemental composition, and specific density of coatings were determined using independent methods, one can demonstrate that to form such boron coatings, the flux density of atomic and ion boron onto the sample surface must be approximately 9.6 <sup>×</sup> <sup>10</sup><sup>16</sup> cm−<sup>2</sup> s −1 . The speed of boron isotope ions of mass *M<sup>i</sup>* that leave the beam plasma with plasma electron temperature *T<sup>e</sup>* , in eV units, is determined by the ambipolar speed of sound *v<sup>i</sup>* = √ *eTe*/*M<sup>i</sup>* . At *<sup>T</sup><sup>e</sup>* <sup>≈</sup> 3 eV, this speed is *<sup>v</sup>*10B <sup>=</sup> 5.4 <sup>×</sup> <sup>10</sup><sup>5</sup> cm/s for <sup>10</sup>B + ions and *<sup>v</sup>*11B <sup>=</sup> 5.2 <sup>×</sup> <sup>10</sup><sup>5</sup> cm/s for <sup>11</sup>B + ions. Because only singly charged ions were registered in the beam plasma, the flux density of <sup>10</sup>B + ions on the sample surface equaled the product of their density in the plasma and their speed from plasma, which is 1.3 <sup>×</sup> <sup>10</sup><sup>16</sup> cm−<sup>2</sup> s −1 . This value was 5.0 <sup>×</sup> <sup>10</sup><sup>16</sup> cm−<sup>2</sup> s −1 for the <sup>11</sup>B + isotope. Thus, the flux density of all boron ions on the sample surface was around 6.3 <sup>×</sup> <sup>10</sup><sup>16</sup> cm−<sup>2</sup> s −1 . Comparing this value with the above estimate of the total boron particle flux both in ionized and neutral states, it was concluded that the contribution of the plasma phase to the formation of boron coatings is about 65 at.%, which exceeds the 35 at.% contribution of boron vapor.

Figure 3 shows the scanning electron microscopy (SEM) images of the surfaces of the obtained boron coatings. The coatings do not contain any defects, pores, or cracks, indicative of coating uniformity and smoothness. The boron coating deposited without the electron beam comprises small tightly packed segments 30–150 nm in size, while that deposited with a pulsed electron beam (regime 1) has a structure with discernible round hills on the surface. The characteristic side of the hill base in the image field is approximately 1–3 µm. *Ceramics* **2022**, *5,* FOR PEER REVIEW 7

**Figure 3.** Scanning electron microscopy (SEM) images of boron coatings on sample surfaces (**a**) after deposition on the substrate and (**b**) after processing with electron beam pulses in regime 1 at a peak current of 20 A and an electron energy of 8 keV. **Figure 3.** Scanning electron microscopy (SEM) images of boron coatings on sample surfaces (**a**) after deposition on the substrate and (**b**) after processing with electron beam pulses in regime 1 at a peak current of 20 A and an electron energy of 8 keV.

Figure 4a shows the elemental composition of the coating, comprising 95.5 at.% boron with small admixtures of carbon (3.7 at.%) and oxygen (0.9 at.%). Analysis of the elemental composition of coatings modified and unmodified by the electron beam showed that beam treatment did not practically change the coating composition. The carbon content in the coating is caused by the presence of this element in the substrate material (steel A284). Figure 4b,с shows the coating profiles measured using a Solver P47 SEM before and after treatment with a pulsed electron beam in regime 2. For a better visual perception of the beam effect, the scale of the height axes in Figure 4b,с is two orders of magnitude smaller than the scales of width and length, and the columnar structure of the surface in Figure 4с is actually a landscape of gently sloping hills. Hence, the deposition of coating forms a relatively smooth surface with nonuniformities a few fractions of a micrometer in size, while subsequent treatment by a pulsed electron beam leads to the formation of gentle hills with a height of up to 1.5 µm. In contrast to SEM, the use of atomic force microscopy (AFM) to determine the surface relief may screen small nonuniformities adjacent to the larger ones with a larger height. Thus, Figure 4b shows mostly hills with a base diameter of 1–10 µm although their structure qualitatively matches the surface structure in Figure 3b with discernible smaller hills. The change in Figure 4a shows the elemental composition of the coating, comprising 95.5 at.% boron with small admixtures of carbon (3.7 at.%) and oxygen (0.9 at.%). Analysis of the elemental composition of coatings modified and unmodified by the electron beam showed that beam treatment did not practically change the coating composition. The carbon content in the coating is caused by the presence of this element in the substrate material (steel A284). Figure 4b,c shows the coating profiles measured using a Solver P47 SEM before and after treatment with a pulsed electron beam in regime 2. For a better visual perception of the beam effect, the scale of the height axes in Figure 4b,c is two orders of magnitude smaller than the scales of width and length, and the columnar structure of the surface in Figure 4c is actually a landscape of gently sloping hills. Hence, the deposition of coating forms a relatively smooth surface with nonuniformities a few fractions of a micrometer in size, while subsequent treatment by a pulsed electron beam leads to the formation of gentle hills with a height of up to 1.5 µm. In contrast to SEM, the use of atomic force microscopy (AFM) to determine the surface relief may screen small nonuniformities adjacent to the larger ones with a larger height. Thus, Figure 4b shows mostly hills with a base diameter of 1–10 µm although their structure qualitatively matches the surface structure in Figure 3b with discernible smaller hills. The change in energy during the pulse action on the coating does not significantly affect the picture of the surface; however, the size and height of the hills in regime 1 are approximately 20% greater than those in regime 2.

energy during the pulse action on the coating does not significantly affect the picture of the surface; however, the size and height of the hills in regime 1 are approximately 20%

greater than those in regime 2.

than 15% can be estimated by

and regime 2 are shown in Figure 5.

of the boron coating.

*Ceramics* **2022**, *5,* FOR PEER REVIEW 8

**Figure 4.** (**a**) Elemental composition of the obtained coatings. The surface profiles of (**b**) boron-based coating and (**c**) boron-based coating after pulse modification in regime 2 measured under a Solver P47 atomic force microscope. **Figure 4.** (**a**) Elemental composition of the obtained coatings. The surface profiles of (**b**) boron-based coating and (**c**) boron-based coating after pulse modification in regime 2 measured under a Solver P47 atomic force microscope.

Accelerated electrons in a solid are decelerated in the layer of their maximal range of penetration. Based on previous works [32], this range *Re* in micron units at an electron beam energy *Ee* = 0.5–10 keV in a solid with density ρ (in g/cm3) with an accuracy of better Accelerated electrons in a solid are decelerated in the layer of their maximal range of penetration. Based on previous works [32], this range *R<sup>e</sup>* in micron units at an electron beam energy *E<sup>e</sup>* = 0.5–10 keV in a solid with density ρ (in g/cm<sup>3</sup> ) with an accuracy of better than 15% can be estimated by

$$R\_{\ell} = 9 \times 10^{-2} \text{ } \text{\$\rho\$}^{-0.8} E\_{\ell}^{1.3} \text{\$} \tag{1}$$

 = 9 × 10 ρ. . (1) For boron coating ρ ≈ 2.2 g/cm3 and electron beam energy *Ee* = 8 keV, the value of *Re* For boron coating <sup>ρ</sup> <sup>≈</sup> 2.2 g/cm<sup>3</sup> and electron beam energy *<sup>E</sup><sup>e</sup>* = 8 keV, the value of *<sup>R</sup><sup>e</sup>* is 0.72 µm for regime 1, while for *E<sup>e</sup>* = 6 keV, the *R<sup>e</sup>* value is 0.49 µm for regime 2. Thus, for both regimes, the electron energy is released in the layer whose thickness is less than that of the boron coating.

is 0.72 µm for regime 1, while for *Ee* = 6 keV, the *Re* value is 0.49 µm for regime 2. Thus, for both regimes, the electron energy is released in the layer whose thickness is less than that The dependence of energy density distribution *Q* absorbed from the beam of accelerated electrons by the boron coating at depth *x* can be determined by [33]

$$Q(\mathbf{x}) = (\mathbf{E}\_{\mathbf{t}}/\mathbf{R}\_{\mathbf{t}})(1 - \mathbf{x}/\mathbf{R}\_{\mathbf{t}})^{5/4}(\mathfrak{Z} - \mathfrak{Z}\exp(-(\mathbf{Z} + 8/4) \times (\mathbf{x}/\mathbf{R}\_{\mathbf{t}}))(\mathbf{j}\_{\mathbf{t}}\mathfrak{r}\_{\mathbf{t}}) \tag{2}$$

erated electrons by the boron coating at depth *x* can be determined by [33] () = (⁄)(1 − ⁄)⁄(3 − 2exp (−(Z + 8⁄4) × (⁄))(τ) (2) where *j<sup>e</sup>* and τ*<sup>e</sup>* are the densities of the electron beam current on the sample surface and the pulse duration, respectively, and Z = 5 is the number of electrons in a boron atom. When

When evaluating (), it is convenient to substitute in the first factor in Expression (2), the electron energy in eV and in cm, and the remaining factors should use in µm. In this case, the *Q* value is expressed in J/cm3 and the depth *x* is expressed in µm. The dependences of the energy density distribution for the electron beam in both regime 1

The boron temperature profile with depth *х* will approximately match the profile of

(), provided that the condition ατ ≪ е holds true, where α ≈ 0.1 cm2/s is the thermal diffusivity of the boron coating. At experimental electron energies, this condition is strictly satisfied for submicrosecond electron beam pulses, when the heat due to the place.

evaluating (*x*), it is convenient to substitute in the first factor in Expression (2), the electron energy *E<sup>e</sup>* in eV and *R<sup>e</sup>* in cm, and the remaining factors should use *R<sup>e</sup>* in µm. In this case, the *Q* value is expressed in J/cm<sup>3</sup> and the depth *x* is expressed in µm. The dependences of the energy density distribution for the electron beam in both regime 1 and regime 2 are shown in Figure 5. other thing that can be noted is that the pulsed electron beam treatment may have an effect on the structure of the deposited coating at different depths. Although the study of such effects is an interesting task, it was outside of the scope of the current research focused on the study of the surface properties of the coating.

*Ceramics* **2022**, *5,* FOR PEER REVIEW 9

energy imparted from the beam at a depth *Re* does not have time to propagate deep into the surface of the solid. However, even though this condition is not satisfied in our experimental conditions due to much longer pulse width (hundreds of microseconds), anyway, as follows from the dependences in Figure 5, about 70% of the released energy of accelerated electrons and, therefore, the most intense heating of the coating material occurs at a depth of 0.5*Re*, which corresponds to 17% of the overall coating thickness at *Ee* = 8 keV and about 12% at *Ee* = 6 keV. That means that a significant part of electron beam energy can be deposited even at much longer pulses. Moreover, because the experimental processing of coatings was performed using a series of 300 pulses for 150 s, it is possible that the gradual heating of the coating surface during this period of time took

The surface hilly structure appeared due to the beam action, which is apparently

related to cyclic temperature effects. The energy density of the pulsed electron beam on the coating surface is about 2 J/cm2 per pulse, while the thermal conductivity coefficient of boron is an order of magnitude less than those of the majority of metals. Thus, the surface of the boron coating at the depth with the maximum release of the beam energy can heat up to a temperature at which the thin surface layer of the coating begins to melt. A similar relief of the surface of TiNi alloy was observed during repeated treatment by an electron beam with an electron energy of 20 keV and a beam power per pulse of 4 J/cm2 [34]. Under the beam action, a hilly structure was formed; the number of pulses (128) in this case was comparable to that in our experiments (300 pulses). From a prior study [32], such surface relief may be associated with the development of instabilities on the

and cooling of a thin surface layer. Similar effects seem to take place in our case too. An-

**Figure 5.** Distribution of the energy density *Q* absorbed by the boron coating from the beam of accelerated electrons versus depth for two regimes. The beam parameters are τ = 500 ms, regime 1: = 0.5 A/cm2, *Ee* = 8 keV, regime 2: = 0.75 A/cm2, *Ee* = 6 keV. **Figure 5.** Distribution of the energy density *Q* absorbed by the boron coating from the beam of accelerated electrons versus depth for two regimes. The beam parameters are τ*e* = 500 ms, regime 1: *j<sup>e</sup>* =0.5 A/cm<sup>2</sup> , *E<sup>e</sup>* = 8 keV, regime 2: *j<sup>e</sup>* =0.75 A/cm<sup>2</sup> , *Ee* = 6 keV.

The boron temperature profile with depth *x* will approximately match the profile of *Q*(*x*), provided that the condition √ ατ*<sup>e</sup> <sup>R</sup><sup>e</sup>* holds true, where <sup>α</sup> <sup>≈</sup> 0.1 cm2/s is the thermal diffusivity of the boron coating. At experimental electron energies, this condition is strictly satisfied for submicrosecond electron beam pulses, when the heat due to the energy imparted from the beam at a depth *R<sup>e</sup>* does not have time to propagate deep into the surface of the solid. However, even though this condition is not satisfied in our experimental conditions due to much longer pulse width (hundreds of microseconds), anyway, as follows from the dependences in Figure 5, about 70% of the released energy of accelerated electrons and, therefore, the most intense heating of the coating material occurs at a depth of 0.5*R<sup>e</sup>* , which corresponds to 17% of the overall coating thickness at *E<sup>e</sup>* = 8 keV and about 12% at *E<sup>e</sup>* = 6 keV. That means that a significant part of electron beam energy can be deposited even at much longer pulses. Moreover, because the experimental processing of coatings was performed using a series of 300 pulses for 150 s, it is possible that the gradual heating of the coating surface during this period of time took place.

The surface hilly structure appeared due to the beam action, which is apparently related to cyclic temperature effects. The energy density of the pulsed electron beam on the coating surface is about 2 J/cm<sup>2</sup> per pulse, while the thermal conductivity coefficient of boron is an order of magnitude less than those of the majority of metals. Thus, the surface of the boron coating at the depth with the maximum release of the beam energy can heat up to a temperature at which the thin surface layer of the coating begins to melt. A similar relief of the surface of TiNi alloy was observed during repeated treatment by an electron beam with an electron energy of 20 keV and a beam power per pulse of 4 J/cm<sup>2</sup> [34]. Under the beam action, a hilly structure was formed; the number of pulses (128) in this case was comparable to that in our experiments (300 pulses). From a prior study [32], such surface relief may be associated with the development of instabilities on the melt/vapor phase interface under repeated action of the electron beam and cyclic melting and cooling of a thin surface layer. Similar effects seem to take place in our case too. Another thing that can be noted is that the pulsed electron beam treatment may have an effect on the structure of the deposited coating at different depths. Although the study of such effects is an interesting task, it was outside of the scope of the current research focused on the study of the surface properties of the coating.

The structural phase properties of boron coatings were studied using synchrotron radiation on beamline 2 of the VEPP-3 electron storage ring at the Anomalous Scattering station. In the X-ray diffraction patterns of boron coatings treated and untreated with the electron beam, reflections of low intensity and considerable width are observed. Among them, there are two remarkable reflections that correspond to interplanar distances of 0.27 and 0.24 nm (areas 1, 2, 3, and 4 in the inset of Figure 6). These reflections are associated with ultrafine crystals of nonstoichiometric boron nitride. Notably, these reflections decrease for regime 2, which may be indicative of the partial destruction of crystals by the beam with a high power density. However, the X-ray patterns do not show any reflections that can be associated with the crystal structure of boron, as is the case for the hexagonal crystal structure of the zirconium alloy. Thus, the hilly surface of the boron coating after the beam treatment does not exhibit a pronounced inner crystalline structure, and the hills themselves are not specific crystalline formations. This fact again verifies the nature of their formation as a result of repeated melting and cooling of the coating surface at the depth of the maximum beam energy release during the cyclic beam action.

Figure 7 shows the microhardness results of a steel substrate, a crystalline boron target, boron coatings, and boron coatings treated with a pulsed electron beam (regime 1, regime 2). The microhardness of the crystalline boron target is approximately three times less than that of the boron coating. This is apparently due to the density of the target (1.2 g/cm<sup>3</sup> ) being lower than the measured density of the coating (2.2 g/cm<sup>3</sup> ). The microhardness of the boron coating is 12 ± 0.35 GPa, while additional surface modification with a pulsed beam further increases the microhardness, up to 15.5 ± 0.45 GPa.

The adhesion measurements of the samples with boron-based coatings showed that the pulsed beam treatment of coatings did not affect the adhesion value between the boron coating and the sample surface. This is because the beam mainly affected the surface layers of the coating as adhesion is the interface property of the coating–substrate boundary on which the effect of beam treatment was weak. Figure 8 shows the typical micrographs of the surface under different loads *F* exerted on a diamond indenter with a radius of 100 µm. As the pressure on the indenter increases, it begins to submerge into the coating. This is accompanied by an increase in the coefficient of friction, indicating the growing resistance of the sample to the indenter movement. At a load of 6 N exerted on the coating, the coefficient of friction begins to fluctuate, which is indicative of the destruction of the surface structure. In the micrograph of the 6 N load, the start of the local film peeling can be seen. A further increase in the load on the indenter leads to increased fluctuations in the coefficient of friction and in the submergence depth, which is indicative of the film peeling off the substrate. The maximum load on the indenter was 30 N; nevertheless, it sufficed to

completely peel off the coating. Traces of the coating remained on the substrate surface. Thus, it can be estimated that the specific peel work of the coating was about 100 J/m<sup>2</sup> . tions decrease for regime 2, which may be indicative of the partial destruction of crystals by the beam with a high power density. However, the X-ray patterns do not show any

The structural phase properties of boron coatings were studied using synchrotron

radiation on beamline 2 of the VEPP-3 electron storage ring at the Anomalous Scattering

station. In the X-ray diffraction patterns of boron coatings treated and untreated with the

electron beam, reflections of low intensity and considerable width are observed. Among

them, there are two remarkable reflections that correspond to interplanar distances of

ciated with ultrafine crystals of nonstoichiometric boron nitride. Notably, these reflec-

*Ceramics* **2022**, *5,* FOR PEER REVIEW 10

The wear resistance of the A284 steel samples with deposited boron coatings and the samples with the same coatings modified by a pulsed electron beam was also measured. The wear rate of the original sample was 6 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mm3/N·m, while that of the boron-based coating sample was significantly lower and equal to 0.8 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mm3/N·m. The wear rate of the coating modified by the pulsed electron beam was 1.3 <sup>×</sup> <sup>10</sup>−<sup>4</sup> mm3/N·m, lower than that for the uncoated samples but higher than that for the boron coating without beam treatment. Thus, the boron-based coating increases the surface wear resistance by a factor of 7.5, but the coating modification by a pulsed electron beam rolls it back by about 60%. Meanwhile, even a surface with boron coating modified by the beam has a wear resistance 4.5 times higher than that without coating. In our opinion, the reduced wear resistance of the boron coating after modification is associated with an increase in the coefficient of friction due to the formation of the surface hilly structure. reflections that can be associated with the crystal structure of boron, as is the case for the hexagonal crystal structure of the zirconium alloy. Thus, the hilly surface of the boron coating after the beam treatment does not exhibit a pronounced inner crystalline structure, and the hills themselves are not specific crystalline formations. This fact again verifies the nature of their formation as a result of repeated melting and cooling of the coating surface at the depth of the maximum beam energy release during the cyclic beam action.

**Figure 6.** X-ray patterns of the surfaces of boron coatings on samples of the ZrNb1 alloy with a a pulsed beam further increases the microhardness, up to 15.5 <sup>±</sup> 0.45 GPa. **Figure 6.** X-ray patterns of the surfaces of boron coatings on samples of the ZrNb1 alloy with a hexagonal crystal lattice. The inset shows the regions where the reflections of crystalline boron appear: regions 1, 2, 3, and 4. The wavelength of synchrotron radiation is 0.154 nm.

hexagonal crystal lattice. The inset shows the regions where the reflections of crystalline boron

regime 2). The microhardness of the crystalline boron target is approximately three times

less than that of the boron coating. This is apparently due to the density of the target (1.2

g/cm3) being lower than the measured density of the coating (2.2 g/cm3). The micro-

hardness of the boron coating is 12 ± 0.35 GPa, while additional surface modification with

Figure 7 shows the microhardness results of a steel substrate, a crystalline boron

appear: regions 1, 2, 3, and 4. The wavelength of synchrotron radiation is 0.154 nm.

**Figure 7.** Microhardness of the steel substrate, the boron target, and fabricated boron coatings. **Figure 7.** Microhardness of the steel substrate, the boron target, and fabricated boron coatings.

**Figure 8.** Micrographs of the boron coating surface under different loads exerted on a diamond indenter with a radius of 100 µm (regime 2), taken with an optical microscope; indenter loads: 1–3 N; 2–6 N; 3–9 N; 4–30 N. **Figure 8.** Micrographs of the boron coating surface under different loads exerted on a diamond indenter with a radius of 100 µm (regime 2), taken with an optical microscope; indenter loads: 1–3 N; 2–6 N; 3–9 N; 4–30 N.

The wear resistance of the A284 steel samples with deposited boron coatings and the samples with the same coatings modified by a pulsed electron beam was also measured. The wear rate of the original sample was 6 × 10−4 mm3/N·m, while that of the boron-based Additionally, a corrosion rapid test was performed. Uncoated and boron-coated steel samples were placed in 25 wt.% saturated aqueous NaCl solution and exposed at 70 ◦C for 200 h. We tested 2 samples without coatings and 10 samples with deposited boron coatings,

coating sample was significantly lower and equal to 0.8 × 10−4 mm3/N·m. The wear rate of the coating modified by the pulsed electron beam was 1.3 × 10−4 mm3/N·m, lower than that for the uncoated samples but higher than that for the boron coating without beam

of 7.5, but the coating modification by a pulsed electron beam rolls it back by about 60%. Meanwhile, even a surface with boron coating modified by the beam has a wear resistance 4.5 times higher than that without coating. In our opinion, the reduced wear resistance of the boron coating after modification is associated with an increase in the coef-

Additionally, a corrosion rapid test was performed. Uncoated and boron-coated steel samples were placed in 25 wt.% saturated aqueous NaCl solution and exposed at 70 °С for 200 h. We tested 2 samples without coatings and 10 samples with deposited boron coatings, including those treated with a pulsed electron beam. Uncoated samples bore traces of pitting corrosion, distinctly seen in Figure 9a. Signs of corrosion were not noticeable on the surface of all the boron-coated samples. As an example, we include here pictures of the sample surface with boron coating treated with a pulsed electron beam (regime 2) and the sample kept in the solution. This verifies the high corrosion re-

ficient of friction due to the formation of the surface hilly structure.

sistance of boron coatings and the absence of slits and cracks in them.

including those treated with a pulsed electron beam. Uncoated samples bore traces of pitting corrosion, distinctly seen in Figure 9a. Signs of corrosion were not noticeable on the surface of all the boron-coated samples. As an example, we include here pictures of the sample surface with boron coating treated with a pulsed electron beam (regime 2) and the sample kept in the solution. This verifies the high corrosion resistance of boron coatings and the absence of slits and cracks in them. *Ceramics* **2022**, *5,* FOR PEER REVIEW 13

**Figure 9.** Photographs of A284 steel sample surfaces after a 200-h exposure in 25 wt.% NaCl aqueous solution: (**A**) uncoated sample, (**B**) sample with boron coating. **Figure 9.** Photographs of A284 steel sample surfaces after a 200-h exposure in 25 wt.% NaCl aqueous solution: (**A**) uncoated sample, (**B**) sample with boron coating.

#### **4. Conclusions**

**4. Conclusions** Electron-beam evaporation was used on boron targets to fabricate coatings on the surface of A284 steel and ZrNb1 alloy with a thickness of a few microns and a uniform structure. The deposition rate of boron coating at a power of 1 kW of a continuous beam of a fore-vacuum source was about 0.5 µm/min. The coating density of 2.2 g/cm3 was close to the density of crystalline boron. On comparing the data on the mass-to-charge Electron-beam evaporation was used on boron targets to fabricate coatings on the surface of A284 steel and ZrNb1 alloy with a thickness of a few microns and a uniform structure. The deposition rate of boron coating at a power of 1 kW of a continuous beam of a fore-vacuum source was about 0.5 µm/min. The coating density of 2.2 g/cm<sup>3</sup> was close to the density of crystalline boron. On comparing the data on the mass-to-charge composition, beam plasma density, and coating parameters, it was concluded that the contribution of the plasma phase to the growth of the boron coating exceeded that of the vapor phase.

composition, beam plasma density, and coating parameters, it was concluded that the contribution of the plasma phase to the growth of the boron coating exceeded that of the vapor phase. The boron coatings were modified by a pulsed beam with electron energies of 8 and 6 keV and powers of 2 and 2.2 J/cm2 for a duration of 300 pulses at a repetition rate of 2 p.p.s. Using SEM and AFM, it was determined that such modifications yielded a considerable change in the surface morphology, forming a hilly structure with a characteristic size of the hill base of 1–10 µm and a height of 0.2–1.5 µm. Analysis of the absorbed beam energy showed that at a boron coating thickness of about 2 µm, 70% of the electron energy was released at a depth constituting less than 20% of this thickness. Structural phase analysis of coatings using synchrotron radiation showed that boron, both in The boron coatings were modified by a pulsed beam with electron energies of 8 and 6 keV and powers of 2 and 2.2 J/cm<sup>2</sup> for a duration of 300 pulses at a repetition rate of 2 p.p.s. Using SEM and AFM, it was determined that such modifications yielded a considerable change in the surface morphology, forming a hilly structure with a characteristic size of the hill base of 1–10 µm and a height of 0.2–1.5 µm. Analysis of the absorbed beam energy showed that at a boron coating thickness of about 2 µm, 70% of the electron energy was released at a depth constituting less than 20% of this thickness. Structural phase analysis of coatings using synchrotron radiation showed that boron, both in modified and unmodified coatings, was present in the form of amorphous or ultrafine phases. Based on the bulk of the data obtained, it was concluded that the formation of the hilly surface structure was not caused by crystallization due to the electron beam but was associated with cyclic melting and cooling of the surface under the action of repetitive electron beam pulses.

modified and unmodified coatings, was present in the form of amorphous or ultrafine phases. Based on the bulk of the data obtained, it was concluded that the formation of the hilly surface structure was not caused by crystallization due to the electron beam but was associated with cyclic melting and cooling of the surface under the action of repetitive electron beam pulses. Research on mechanical properties of the boron coatings showed that modification of its surface by the electron beam improved the coating hardness from 12 ± 0.35 to 15.5 ± 0.45 GPa. Based on the study of the coating–surface adhesion, it was shown that the specific peel work of the coating was about 100 J/m<sup>2</sup> . The measured wear resistance of boron coatings was of the order of 10−<sup>4</sup> mm3/N·m, which exceeded the wear resistance of

Research on mechanical properties of the boron coatings showed that modification of its surface by the electron beam improved the coating hardness from 12 ± 0.35 to 15.5

boron coatings was of the order of 10−4 mm3/N·m, which exceeded the wear resistance of the A284 steel substrate manyfold. Based on the tests conducted in saturated salt solu-

**Author Contributions:** Conceptualization, Y.Y. and E.O.; methodology, Y.Y. and A.T.; investigation, Y.Y, A.T. and A.K.; resources, E.O.; data curation, A.K.; writing—original draft preparation, Y.Y.; writing—review and editing, D.Z.; visualization, D.Z.; supervision, E.O.; project administra-

tion, we demonstrated that the deposited coatings have anti-corrosive properties.

the A284 steel substrate manyfold. Based on the tests conducted in saturated salt solution, we demonstrated that the deposited coatings have anti-corrosive properties.

**Author Contributions:** Conceptualization, Y.Y. and E.O.; methodology, Y.Y. and A.T.; investigation, Y.Y, A.T. and A.K.; resources, E.O.; data curation, A.K.; writing—original draft preparation, Y.Y.; writing—review and editing, D.Z.; visualization, D.Z.; supervision, E.O.; project administration, Y.Y.; funding acquisition, E.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was conducted with financial support from the Russian Federation represented by the Ministry of Science and Higher Education under Project No. 075-15-2021-1348 within the framework of event No. 2.1.6.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data are available upon reasonable request.

**Acknowledgments:** The authors are grateful to Georgy Yushkov (Institute of High Current Electronics SB RAS, Tomsk), the father of one of the co-authors of this article (Yu. Yushkov), for advice in the analysis of the results and preparing the manuscript as well as to Ian Brown (Berkeley Lab) for useful discussions and English language improvement.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

