Development of alloy coatings by electron beam physical vapour deposition method

 Introduction

Electron beam physical vapour deposition (EB-PVD), a derivative of EB melting method, is a process in which a high-energy electron beam focused onto the target(s) in a vacuum chamber is being used to melt the component(s) to create a vapour cloud. The vapour cloud condenses on the determined substrates, developing a film coating the substrates. External heating is a common method used to improve the metallurgical bonding between the film and substrate and to control the film microstructure. EB-PVD is a sightline process; thus, to have an even coating thickness on complex parts such as turbine blades, continuous rotation in the vapour cloud is required during the coating process [1]. Some advantages of EB-PVD include high evaporation rates compared to other PVD techniques (up to 150 µm min⁻¹ with evaporation rate of 10–15 kg h⁻¹), a possibility of controlling the chemical composition variations over the depth of the coatings as well as developing functionally graded coatings, a possibility of achieving a dense, smooth and homogenous coating, having low impurity and providing a good control over the microstructure of the deposited layer by changing parameters of the process. In addition, a broad range of metallic and ceramic materials including elements with extremely low vapour pressure such as Mo, W, Re and C can be evaporated by this method [1-3]. However, the complexity and expense of the method can be considered as demerits of this process [4].

 Physical vapour deposition process

The coating process by PVD includes an atom-by-atom basis deposition from the vapour phase with four important steps as follows:

Generation of the vapour phase flux by a physical process through either evaporation or sputtering.

There are different methods used to increase the target's temperature during evaporation, such as resistive heating, electron and laser beam heating, arc and electrical induction heating [2,5].

EB-PVD components

An EB-PVD unit has four main components, including electron beam gun system, ceramic or copper-based water-cooled crucible containing the determined materials for evaporation, the substrate, which is intended to be coated, and the vacuum chamber, to apply different types of coatings [1]. A photograph of a vacuum coating chamber by PVD method can be seen in Figure 1. The EB gun can be either self-accelerated or work-accelerated; moreover, with the help of magnetic lenses, electron beams can be directed in different angles as presented in Figure 2 [7]. Details of several crucible types and some common feeding systems used in EB-PVD are shown in Figure 3.

OPEN IN VIEWER

Figure 1

Photograph of EB-PVD chamber with six EB guns shown by numbers and three feeding sources, A–C, the substrate is shown with an arrow [8].

OPEN IN VIEWER

Figure 2

Schematic layout of evaporation sources with different EB paths [7].

OPEN IN VIEWER

Figure 3

Some types of crucibles and frequent feeding systems used in EB-PVD process [7].

Deposition of alloys in EB-PVD process

In this method, when the evaporant material warms up, its vapour pressure rises in a way that at a critical temperature will be higher than the atmospheric pressure generating gas flow. Figure 4 shows variations of saturated vapour pressure with temperature for some elements. As per Clausius–Clapeyron equations, the vapour pressure increases with increasing temperature for all elements. Usually, owing to the difference in vapour pressure of elements, maintaining the stoichiometry of the alloy during the evaporation process is difficult [2,9-13].

OPEN IN VIEWER

Figure 4

Changes in vapour pressure versus evaporation temperature for some metallic elements [2].

Compounds and alloys evaporate like pure elements, but in the application, the chemical composition of the coating obtained from them differs from the initial stoichiometry of such evaporant resources. This is because the vapour pressure of elements at different temperatures is unequal leading to different evaporation rates. One method to resolve this problem is to use several targets and EB guns, and through adjusting the beam dwell time, frequency and beam current on each target. This way, the capabilities of the system increase. Homogenisation and transfer of the vapour cloud can also be done by an inert jet gas known as direct vapour deposition as presented in Figure 5 [14,15]. For instance, Yu et al. [14] proved that by setting the dwell time to be 39:61 for Al:Ni system deposition of homogeneous β-NiAl is possible. In addition, as illustrated in Figure 6, changing the EB current changes the evaporation rate of elements providing the possibility of achieving the desired chemical composition [15].

OPEN IN VIEWER

Figure 5

Evaporation of separate feeding targets mixed in the vapour state and transferred by helium gas to reach the substrate known as EB-DVD process [14].

OPEN IN VIEWER

Figure 6

Variations of evaporation rate of (a) aluminium, (b) nickel and (c) platinum with increasing the electron beam current. The corresponding feed rates were determined to keep a constant melt and vapour cloud [15].

The ideal evaporation rate in vacuum, J, in different temperatures can be calculated from the Langmuir's equation [9,11,13,16] as follows: (1) where R_g is the ideal gas constant, T is evaporation temperature, M is the molar mass of the evaporant(s) and is its vapour pressure.

As evaporation rate is a non-linear function of temperature, it is extremely dependent on its fluctuations. Such fluctuations are due to changes in the beam power, dwell spot size and frequency as well as the ability of the evaporant to dissipate energy, or heat, from its surface in forms of radiation, convection and conduction. Furthermore, Marangoni shear and buoyancy are the dominant flows in the molten material while Lorentz forces are negligible here [16].

As aforementioned, the use of the electron beam provides the possibility of depositing multi-component alloy coatings with different compositions [17]. EB-PVD is widely used to develop coatings for turbine blades, multilayer coatings and other applications possessing excellent merits such as high precision in thickness, composition homogeneity, low contamination levels and high evaporation rates. Both one and two evaporation sources are common for producing alloy coatings; however, using two evaporation sources, alloying is directly dependent on the vapour streams and is limited to a specific area in front of the evaporant sources. To achieve the proper mixing of components in the vapour cloud, in addition to decreasing the evaporation rate, the distance between evaporant sources and substrate should be further compared and adjusted with cases when only one evaporant target is being used. As a result, one target is being used in many cases due to superiority in geometry, simplicity and the possibility of depositing at a higher rate [10,18].

Owing to the difference in saturated vapour pressures of each element as well as interactions among them, a selective evaporation occurs during the EB-PVD process of one multi-component target source. This leads to either enrichment or dilution of the molten pool out of an element. In addition, interaction among different elements is dependent on the temperature and concentration of the elements. That is why, when depositing alloys, continuous feeding of the molten pool is required. After a transition time, according to the law of conservation of mass, the chemical composition of the coating will be the same as the evaporant source. A study of the quantitative effect of these interactions and a measurement of activity coefficients of the components are required to explain the behaviours, something which is also unknown for the alloys. For this reason, experimental results regarding transition time duration or the effect of evaporation parameters on the chemical composition of coatings can be useful. These two parameters play an important role in depositing alloy coatings in EB-PVD process [10,17].

In the case of multi-component alloys having one source of evaporant, the total vapour pressure is the sum of the partial vapour pressures of each component [17]: (2) where (3) in which N_{i 2} is the molar concentration of component i in the molten pool, K_1i and K_2i are constants specific to each material, T is the temperature and is the activity coefficient of component i which depends on temperature, concentration and type of the material. To simplify the calculations, it is possible to utilise some approximations, therefore, knowing all constants might not be required. To do so, the following assumptions should be considered:

Evaporation parameters like EB intensity, molten pool volume, vacuum level, chamber temperature as well as molten pool surface temperature are all constant.

Generally, the process of evaporation of alloys is defined as follows: at the beginning of the evaporation process, a change in the composition of the molten pool occurs, since the component with higher vapour pressure has a higher evaporation rate; consequently, its concentration decreases resulting in a decrease in its evaporation rate after some time. After a transition time, equilibrium is achieved, and concentrations stabilise. The duration of this transition time is mainly dependent on the evaporation temperature, total evaporation rate, molten pool volume, difference in the vapour pressure of involving components and concentration of alloying elements. Higher evaporation rates and lower difference in the vapour pressure of the components lead to a significant reduction in transition time. These hurdles are being considered as limitations of using one source of the evaporant target when depositing alloys. In most applications, the transition time is in the range of several minutes to several hours; however, by knowing the amount of concentration variation through the passage of time for an alloy in the molten pool, and producing the corresponding alloy as the primary source, the transition time can be reduced [10,17,19].

Figure 7 depicts the calculated evaporation rates and concentration variations over time for FeCr18Ni12 in the molten pool and the deposited layer. As it is shown, concentrations have reached a constant level after a while. Nickel, due to its low vapour pressure, reaches equilibrium after a long transition time. It is physically reasonable that increasing the activity coefficient of a component leads to a lower transition time. Moreover, a higher activity coefficient of a component accelerates its evaporation rate. In this regard, the effect of the molten pool volume and vapour-emitting surface area is remarkably lower than the notable impact of temperature and activity coefficient [17]. So far, some studies have been done on the evaporation of binary alloys and several relations have been presented as follows.

OPEN IN VIEWER

Figure 7

Calculated amounts in evaporation of FeCr18Ni12 alloy over time: (a) evaporation rates, (b) composition of molten pool (c) composition of the deposited layer [17].

As per the Langmuir equation, the evaporation rate of component i in a multi-component solution is being calculated as: (4) where (5) and is the evaporation coefficient, which for ideal evaporation equals to 1.  is molar concentration and is the activity coefficient of component i in the melted pool. T is evaporation temperature and is the mole mass of component i. Obviously, the evaporation rate of each component is a function of activity, saturated vapour pressure and molar concentration of that component in the melt [10,20].

 Miedema Model

Owing to the limitations in experimentation procedures of activity coefficients in infinite diluted solutions, we can use the semi-empirical Miedema model to obtain activity coefficients: (12) where (13) in which is the electron density (the number of electrons in the unit volume in volts), V is the mole volume, R is the gas constant, is temperature, is the melting temperature of component i, f and u are functions of materials’ constant like electronegativity and electron density. Table 1 lists some of these constants for some elements [10,22,24].

OPEN IN VIEWER

Table 1

The values of some parameters to be used in the Miedema model for some elements [22].

Element	φ (V)	V2/3 (cm2)	Tm (K)
Al	4.2	4.6	933
Si	4.7	4.2	1685
Mg	3.45	5.8	922
Ni	5.2	3.5	1728
Cr	4.65	3.7	2136
Cu	4.55	3.7	1357

By inserting the coefficient of activity obtained from the Wilson relation in the Langmuir equation, and by knowing the evaporation temperature, concentration variations of the elements in the molten pool and deposited layer are justifiable and correspond to experimental results. Ch. Metzner et al. [17] and Y. Xin et al. [10] showed this correspondence as illustrated in Figures 8 and 9. The required transition time, in the experiment shown in Figure 9, to reach the complete compositional balance is 2 h which is time consuming and results in wasting a lot of material. Therefore, parameters of the process should be optimised to have a shorter transition time.

OPEN IN VIEWER

Figure 8

Experimental measurements and calculated values in evaporation of FeCr18Ni12 over time; (a) composition of molten pool, (b) composition of deposited coating film [17].

OPEN IN VIEWER

Figure 9

Concentration variations in the evaporation of Ni28Fe18Cr6Al alloy over time; (a) composition of molten pool, (b) composition of vapour cloud [10].

Conclusion

Deposition of multi-component coatings with homogenous chemical composition via EB-PVD is a process which requires continuous feeding of the evaporant components with different rates based on their various vapour pressures to reach equilibrium after a transition time. To this end, molecular considerations including activity coefficient, mole volume, electronegativity and electron density, etc. ought to be considered.