Secondary electron emission serves as the foundation for a broad range of vacuum electronic devices and instrumentation, from particle detectors and multipliers to high-power amplifiers. While secondary yields of at least 3–4 are required in practical applications, the emitter stability can be compromised by surface dynamics during operation. As a result, the range of practical emitter materials is limited. The development of new emitter materials with high yield and robust operation would advance the state-of-the-art and enable new device concepts and applications. In this Perspective article, I first present an analysis of the secondary emission process, with an emphasis on the influence of material properties. From this analysis, ultra-wide bandgap (UWBG) semiconductors and oxides emerge as superior emitter candidates owing to exceptional surface and transport properties that enable a very high yield of low-energy electrons with narrow energy spread. Importantly, exciting advances are being made in the development of promising UWBG semiconductors such as diamond, cubic boron nitride (c-BN), and aluminum nitride (AlN), as well as UWBG oxides with improved conductivity and crystallinity. These advances are enabled by epitaxial growth techniques that provide control over the electronic properties critical to secondary electron emission, while advanced theoretical tools provide guidance to optimize these properties. Presently, H-terminated diamond offers the greatest opportunity because of its thermally stable negative electron affinity (NEA). In fact, an electron amplifier under development exploits the high yield from this NEA surface, while more robust NEA diamond surfaces are demonstrated with potential for high yields in a range of device applications. Although c-BN and AlN are less mature, they provide opportunities to design novel heterostructures that can enhance the yield further.
Secondary electron emission is produced when a beam of energetic electrons impacts a solid surface. As a result, it occurs to some degree in any vacuum electronic device that employs an electron beam. The secondary electrons are the low-energy portion of the emitted electron energy distribution produced by the primary beam (as shown in Fig. 1). This spectrum extends from the primary beam energy (Eo) (elastic peak) down to 0 eV (vacuum-level cutoff). True secondary electrons are defined as those having kinetic energy KE ≤ 50 eV, but most secondary electrons actually have much lower energy (< ∼5 eV). This low-energy characteristic can be understood by considering the scattering dynamics of the primary beam. While some of the primary electrons are elastically reflected, most of these energetic electrons penetrate into the material and interact inelastically with atoms and electrons in the crystal lattice. Through repeated inelastic scattering, the primary beam loses most of its energy and is eventually absorbed in the material. However, some of the primary electrons and excited electrons are deflected back toward the surface during the scattering process and are emitted into vacuum. Backscattered electrons that suffer only a few inelastic scattering events are emitted with relatively high energy; and importantly, those that undergo only a single-scattering interaction (e.g., Auger electrons, x-ray photoelectrons) can provide information about fundamental excitations that are characteristic of the specific elemental composition. However, most backscattered electrons undergo multiple inelastic scattering interactions as they diffuse to the surface, and as a result, they lose such element-specific information. In the case of secondary electrons, most of the kinetic energy is lost during the diffusion process and, consequently, they are emitted with energies just above the vacuum level (Evac), as shown in Fig. 1. In fact, the secondary-emission cutoff serves as an important indicator of Evac in the energy spectrum. Although secondary electrons do not carry element-specific information the way that Auger electrons and x-ray photoelectrons do, secondary electron emission is, nonetheless, very sensitive to the material properties (e.g., crystallinity, band structure, surface barrier), and opportunities to exploit such material dependencies will be a focus of this article.
As long as a primary electron beam is used in a vacuum electronic device, secondary electron emission cannot be avoided. However, it affects the performance of vacuum electronic devices and instrumentation in various ways that depend on the specific device design and operating conditions. In some applications, secondary electrons are very useful and can even be exploited, while in other cases, the secondary electrons are detrimental and must be avoided. In scientific analysis, for example, scanning electron microscopes use a dedicated secondary electron detector to obtain important imaging information from secondary electrons scattered from the target material.2 On the other hand, the low-energy background signal produced by secondary electrons can interfere with the spectral analysis in electron spectroscopic techniques. In device applications, a high secondary yield is critical to the successful operation of electron multipliers,2 photo multipliers and detectors,3,4 and night vision/image intensifiers, as well as higher-power vacuum electronic devices such as magnetrons and crossed-field amplifiers.1,5 In such cases, higher secondary yield can enable higher signal gain, higher signal-to-noise ratio, higher efficiency, and lower power consumption of the device. Conversely, secondary electrons can degrade the performance of devices by producing stray signal and noise in analyzers and collectors and through insulator charging and micro-gap/multipactor discharge;6 and in microwave and millimeter wave power tubes, secondary electron emission in depressed collectors can reduce the energy conversion efficiency.6 As a result, surface coatings are often used in analyzers and collectors to suppress secondary emission, as well as on grids and tube walls to prevent RF vacuum breakdown.
In vacuum electronic applications that are strongly affected by secondary electron emission, the materials must be carefully selected to provide enhanced secondary emission or to suppress the emission. However, the operating environment and device configuration can place additional constraints on the material selection. For example, robust insulator materials such as MgO and BeO are commonly used as secondary emission cathodes in magnetrons and crossed-field amplifiers that operate under harsh conditions.5 In contrast, fragile monolayer coatings are often deposited on semiconductor elements in electron/photon multipliers and detectors to provide very high secondary yields. Conversely, graphite and other carbon coatings are commonly applied to the surfaces of an instrument or a device to suppress the secondary emission throughout the electron transport region. To choose the appropriate material for each application and to optimize the device design, it is crucial to have a thorough understanding of the secondary electron emission process, and this is especially true when considering new opportunities to advance the field. In this article, I will focus on the development of new materials that have the potential to improve existing device capabilities and enable new device designs. As such, I will first present a brief review of the secondary emission process with a focus on the material properties that are critical to high secondary electron yield. I will then discuss exciting opportunities to achieve greatly enhanced secondary emission from promising UWBG materials under development and to achieve new electronic devices by exploiting the secondary emission process in these materials.
A. Secondary electron emission: Three-step process
Secondary electron emission is a complex dynamical process2,7,8 that is difficult to model due to the numerous scattering interactions that occur in the material. Instead, secondary electron emission is characterized experimentally by measuring the secondary electron yield (δ), which is defined as the ratio of the secondary electron current (Is) to the primary beam current (Io). As such, the secondary yield represents the current gain. However, the secondary yield is not a constant of the material but, instead, depends very strongly on the primary beam parameters and is especially sensitive to Eo. As a result, the secondary yield is typically plotted as a function of Eo, as shown in Fig. 2, and the resulting yield curve has a characteristic bell shape for most materials.9 In addition, the secondary yield can depend strongly on the angle of incidence10 as well as on material properties, such as conductivity, crystallinity, and surface properties.
To understand these material and beam dependencies, it is necessary to consider the three-step process (Fig. 3) that governs secondary electron emission: (1) secondary electron generation, (2) transport to the surface, and (3) escape into vacuum.5 In the generation step, primary electrons excite bound electrons into higher energy states in the conduction band, thereby creating the secondary electrons. In the transport step, secondary electrons diffuse through the bulk to reach the surface, while losing energy along the way. In the escape step, those secondary electrons that reach the surface are emitted into vacuum if they have sufficient kinetic energy to overcome the surface barrier. In the case of a metal, this surface barrier is the work function ϕ (i.e., ϕ = Evac − EF, where EF is the Fermi level), while in a semiconductor or insulator, the surface barrier is, instead, the electron affinity χ (i.e., χ = Evac − Ec, where Ec is the conduction band minimum). It should be noted that this three-step process also applies to electron emission produced by the impact of sufficiently energetic beams of ions or photons. The generation step is different for each type of impacting particle due to the different interactions that occur, but once the secondary electrons are generated, the transport and emission steps are essentially the same in all cases.
While conceptually straightforward, this three-step process is actually quite complex, with each step depending strongly on the primary beam and/or material properties. Fortunately, advanced secondary emission modeling and simulation capabilities have been developed over the years,11,12 most notably within the electron microscopy field to provide an accurate interpretation of the secondary emission signal in scanning electron microscopes.2,13 Such modeling first considers the primary electron beam impact and penetration into the material. For example, the range formula developed by Kanaya and Okayama,14
can be interpreted as the radius of the interaction region up to the electron stopping point in the film. From this formula, the range RK−O of the primary beam is seen to increase strongly with increasing Eo and more so in materials with higher atomic mass A and lower atomic number Z and density ρ. Therefore, K–O range plots are useful for providing a first-order estimate of the electron penetration depth (Dp) as Eo is varied. However, the K–O range can only be interpreted as an upper limit on Dp since most electron trajectories will not be perpendicular to the surface but will extend laterally as well because of scattering-induced momentum changes. For a more detailed picture of the electron penetration process, two-dimensional Monte Carlo simulations are used to examine the energy-dependent evolution of the electron distribution in a material.13,15
Because of the Eo-dependent range of the primary electron beam, it follows that the secondary emission process has a strong Eo-dependence as well. As Eo increases, the primary electrons penetrate deeper into the material, where they generate secondary electrons. Thus, the generation depth (Dg) increases with Eo (assuming Dg ∼ Dp). In addition, the primary electrons have more kinetic energy available for electron–electron scattering and, therefore, the number of generated electrons increases. Once the secondary electrons are generated, the subsequent transport step depends indirectly on Eo through a longer transport distance. However, this increased transport can lead to a greater loss of electron energy and current through scattering, thermalization, and electron capture, and consequently, fewer secondary electrons may reach the surface with sufficient kinetic energy to overcome the energy barrier. Clearly, the transport and escape processes are both affected strongly by the material properties, such as crystal quality, band structure, and surface barrier. Thus, Eo determines how deeply the secondary electrons are generated (i.e., Dg), but the material properties determine the depth from which they can be transported to and emitted at the surface, i.e., the “escape” depth.
The electron escape depth (Desc) is a useful metric for assessing the influence of different material properties on the secondary electron yield, and it can be deduced from the shape of the yield curve.9 As shown in Fig. 2, the yield curve typically exhibits three regimes that form a bell shape. For low values of Eo, Dp is small so that Dg < Desc and secondary electrons can readily escape. Thus, the yield increases with increasing Eo as the number of generated electrons increases. For high values of Eo, Dp is sufficiently large such that Dg > Desc. Thus, the yield decreases with increasing Eo since more electrons are captured in the material during increasingly longer transport. In between these two regimes, the yield reaches a maximum at an energy Eo = Emax for which the condition is Dg ∼ Desc. Using this simple model and the K–O range equation, the shape of the yield curve can be analyzed to deduce an upper bound on Desc.
B. Influence of material properties
Finally, it is important to understand how the material properties influence the secondary emission process so that appropriate materials can be selected for specific applications. If high secondary yields are required, the material must have sufficiently high transport efficiency for electrons to reach the surface. Therefore, the crystal quality and band structure are critically important. Additionally, the surface barrier must be sufficiently low to provide efficient emission into vacuum and, therefore, the surface electronic structure plays a critical role. Finally, the material must have sufficient electrical conductivity to replenish the emitted charge. While other device-specific requirements may be placed on the emitter material, as a general rule, we can apply these three conditions to metals, semiconductors, and insulators to determine their suitability for different secondary emission applications.
In spite of their high electrical conductivity, metals are relatively poor secondary emitters. As depicted in Fig. 3, primary electrons excite bound electrons into higher energy states in the conduction band. However, these excited (secondary) electrons scatter strongly with other conduction electrons, and this inelastic scattering leads to rapid energy loss during transport. Consequently, very few secondary electrons reach the surface with sufficient energy to overcome the large work function (typically 4–5 eV), and those electrons that are emitted typically come from very close to the surface (∼ several nm). So although metals are conductive, the strong scattering among conduction electrons leads to poor transport efficiency, while the high work function leads to poor emission efficiency. As a result, the maximum secondary yield is only ∼1–1.5 for most metals. In spite of their low yield, it should be noted that metal emitters are used in some applications because of their relatively stable and robust emission characteristics. In particular, secondary yields are relatively insensitive to changes in the work function at metal surfaces and, thus, metal emitters can operate in harsh environments. However, there are device applications where even the low yields produced by metals can be a problem. In such cases, the device components are often coated with graphite or other carbon-based materials because of their extremely low secondary yield (<1). These low yields are due to the relatively high beam penetration and, hence, long transport distance, in elements with a low atomic number and density, such as carbon.
In the case of semiconductors and insulators, i.e., materials with a bandgap, the secondary emission characteristics can vary from very low to extremely high. In these materials, secondary electrons are produced when primary electrons excite valence electrons into the conduction band via impact ionization (Fig. 3). In this process, the internal electron gain is proportional to Eo and inversely proportional to the bandgap (Eg) (where the bandgap dependence reflects the average amount of energy spent creating an electron–hole pair).16 In most semiconductors, relatively few conduction electrons are present to scatter with and, therefore, the excited (secondary) electrons do not undergo the rapid energy loss that occurs in metals. Instead, the secondary electrons lose energy through two other scattering interactions as they diffuse through the conduction band: scattering with valence electrons (impact ionization) and scattering with atomic lattice vibrations (i.e., phonons).
Impact ionization (i.e., electron–electron scattering) is the dominant energy-loss mechanism for energetic secondary electrons since they lose an amount of energy ΔE ≥ Eg during the excitation process.17,18 However, secondary electrons must have an excess kinetic energy of at least this amount in order for this scattering channel to be available. Once the electron energy drops below this threshold energy (where Eth = Ec + Eg), the electron–electron scattering channel is turned off and the dominant energy-loss mechanism becomes electron–phonon scattering. Such scattering involves much smaller energy loss (typically ≤0.1 eV) and, consequently, the secondary electrons slowly thermalize to the bottom of the conduction band as they diffuse through the material. As a result, they can travel longer distances compared to transport in metals, and they reach the surface in a quasi-thermalized distribution near the bottom of the conduction band.18 It should be noted, however, that this process is dependent on the size of the bandgap. Specifically, Eth increases with increasing Eg and, therefore, secondary electrons have more kinetic energy available during the diffusion process to the surface. This larger kinetic energy enables longer transport distances and greater thermalization of the secondary electron distribution.
The question now is whether these low-energy secondary electrons can escape into vacuum. In the case of very wide bandgap insulators (e.g., with Eg ∼5–10 eV), the electron affinity is typically <1 eV19 and, therefore, at least some of these low-energy electrons can be emitted. In fact, high secondary yields have been measured from insulators such as NaCl and single-crystal MgO (with δmax ∼7 and 25, respectively),19,20 as well as thin-film metal oxides such as MgO, BeO, and Al2O3 (with δmax ∼ 3–4 under practical conditions).3,19 However, insulators suffer from poor conductivity and, therefore, high yields are difficult to maintain without the use of pulsed primary beams and thin-film emitter structures. In the case of most semiconductors, the electron affinity is typically ∼3–4 eV (although, importantly, very wide bandgap semiconductors tend to have smaller electron affinities). As a result, few secondary electrons have sufficient kinetic energy to overcome the large surface barrier and those that can escape have a very short inelastic mean free path (∼few nm). Consequently, the secondary yields from most semiconductors are similar to those obtained from metals.
C. Negative electron affinity (NEA)
However, some semiconductor surfaces can be treated to significantly reduce the surface barrier, thereby making it possible to emit large numbers of secondary electrons. This is typically done through the adsorption of an electropositive element (e.g., Cs and other alkali metals), either alone or with oxygen, that forms a dipole layer on the surface that lowers Evac relative to Ec.21 In some cases, Evac can be lowered to a position below Ec such that the electron affinity becomes negative. [Note: such metal-coated surfaces actually form an effective NEA, i.e., Ec(surf) < Evac < Ec(bulk), due to strong band bending induced by charge transfer from the electropositive element.] In fact, surfaces with low or negative electron affinity are produced in this way for many electron/photon multipliers and photocathodes, including cesiated GaAs.3,4 In secondary emission studies, very high secondary yields (>100) have been reported on cesiated surfaces of Si, GaP, and GaAs owing to the excellent transport properties combined with the NEA surface.21 However, these cesiated surfaces are fragile and require ultrahigh vacuum (UHV) conditions and even in situ repair. Nonetheless, these material systems demonstrate that, when sufficient conductivity is present, semiconductors and insulators can be excellent secondary emitters due to efficient transport and emission processes.
As discussed in Sec. II B, wider bandgap materials are advantageous for secondary emission applications owing to the longer transport distances and lower surface barriers. In fact, excellent secondary emission characteristics have been demonstrated from the ultra-wide bandgap (UWBG) semiconductor diamond (with Eg = 5.5 eV) due to its exceptional transport and surface properties. Most notably, the H-terminated surface of single-crystal diamond has a very low NEA of −1.3 eV22,23 that is thermally stable (up to 800 °C) and can even survive exposure to atmosphere, unlike surfaces treated with monolayer (ML) alkali-metal coatings. This remarkable surface property is a result of the high dipole density formed by the polar C–H surface bonds, and it produces a true NEA at the surface with Evac < Ec(surf). The existence of this true NEA surface was first reported in 1979,18 and many subsequent studies have provided important insight into the remarkable surface properties and electron emission characteristics of diamond,24–26 including reports of very high secondary yields of ∼30–100 from both single-crystal and polycrystalline B-doped diamond.27–32
In the case of secondary emission studies, diamond has also provided an opportunity to verify the transport model for wide bandgap materials due to the ability to completely remove the surface barrier, in which case the emitted electron energy distribution should correspond to the internal distribution reaching the surface. Figure 4 shows the secondary yield curves and energy distribution curves (EDCs) measured from the bare, H-terminated, and cesiated surfaces of diamond, i.e., C(100).31 While a maximum yield of 3 was measured at the bare surface, extremely high yields of 60 and 132 were measured from the H-terminated and cesiated surfaces, respectively. It can also be seen that the high yields correspond to the emergence of sharp, low-energy emission peaks in the EDCs following a large decrease in the electron affinity at the H-terminated and cesiated surfaces. As such, the narrow emission peaks confirm that the secondary electrons have undergone extensive thermalization prior to reaching the NEA surface but can still be emitted due to the absence of a surface barrier.
To gain further insight into the transport process, information can be deduced from an analysis of the yield curves. As discussed earlier, typical yield curves exhibit a bell shape, with the maximum corresponding to the position where Desc ∼ Dg. In Fig. 4(a), this bell-shaped yield curve is observed for the bare surface, but the yields for the hydrogen-terminated and cesiated surfaces are still rising at Eo = 3 keV. By estimating Dg (at Emax) from the K–O range formula, it can be deduced that Desc ∼ Dg ∼ 10 nm at the bare diamond surface, but at the NEA surfaces, we can only estimate a lower limit of . To obtain a better estimate of Desc for the NEA surfaces, the measured yield data were compared to calculated yield curves.21,31 From this analysis, it was found that Desc ≥ 5 μm in the C(100) sample, thereby confirming the excellent electron transport properties of diamond. Similarly, the very short Desc measured at the bare diamond surface reveals the consequence of even a modest surface barrier and thereby demonstrates the extreme importance of an NEA to achieving very high secondary electron yields.
The secondary emission properties of diamond are truly exceptional, and they represent an opportunity for great advances in secondary emission applications. In particular, the ability to form a thermally stable NEA surface with H termination rather than alkali metal absorption is extremely important since it broadens the scope of potential device applications. However, in practical secondary emission applications, the emission surface is often subjected to prolonged exposure to energetic electron beams that can remove surface adsorbates (including hydrogen) and can lead to surface damage that is difficult to repair.29,33 This highlights the fundamental challenge in achieving high secondary electron yields from semiconductors and insulators, namely, the simultaneous need for a robust, low-barrier surface and an electrically conductive bulk. In UWBG semiconductors with adsorbate-induced NEA surfaces (e.g., H-terminated diamond), the secondary electron yield is primarily limited by surface modification produced by the beam. Conversely, at robust UWBG insulator surfaces with low electron affinity (e.g., BeO and MgO), the secondary electron yield is primarily limited by electrical charging in the bulk. In Sec. III, I will discuss several opportunities to overcome these challenges.
Opportunities in secondary electron emission primarily lie in the development of new UWBG materials with the potential for higher secondary electron yield and more robust performance. The availability of such materials will open the door to new device concepts and will significantly expand the capabilities of existing device technology. At present, diamond offers the greatest opportunity for advanced secondary emission applications since its potential has not yet been fully realized. Recent research efforts have investigated new NEA surface terminations on diamond that offer the potential for more robust secondary emission, while exciting device concepts are being explored using H-terminated diamond. In the long term, other materials under development may offer a range of capabilities for different secondary emission applications. Toward this end, research is underway to develop promising new UWBG semiconductors such as cubic boron nitride (c-BN) and aluminum nitride (AlN), and opportunities exist to explore novel heterostructures that exploit the secondary emission properties of such materials. In addition, advanced deposition techniques can be exploited to develop UWBG metal oxides with increased conductivity and crystallinity.
A. Robust NEA diamond surfaces
Diamond has exceptional properties that make it an extremely attractive candidate for secondary electron emission applications.34 While the unique NEA property of H-terminated surfaces has already been discussed, diamond also has excellent electrical, thermal, and mechanical properties. Specifically, diamond has high electron (and hole) mobility for efficient transport to the surface, and it can be controllably doped with boron to provide p-type electrical conductivity. Additionally, such p-type doping produces downward band bending at the surface that can enhance the emission efficiency. Moreover, diamond has the highest known thermal conductivity of any material as well as high mechanical strength, both of which are important for practical device fabrication and operation. However, extended exposure to energetic electron beams can produce electron-stimulated desorption of H atoms from the surface, and the continued removal of hydrogen results in a gradual increase in the electron affinity.29,33 Additionally, the H-terminated surface will degrade in poor vacuum conditions from slow oxidation in the presence of an excitation source such as electrons (as well as x rays and UV light).35 This gradual oxidation will result in a shift from negative to positive electron affinity, and the secondary emission performance will become compromised over time.
Hydrogen-terminated diamond also suffers from the electronic degradation of the surface due to p-type surface transfer doping that takes place spontaneously in air.36–39 In this process, charge transfer from atmospheric species results in a large internal electric field that produces upward band bending and, hence, a barrier to electron emission (even though the NEA property remains intact). This surface transfer doping phenomenon has actually attracted great interest in diamond as an active electronic device for high power RF electronics due to a two-dimensional hole gas that is formed ∼5 nm below the surface. However, the upward band bending destroys the secondary electron emission capabilities by preventing low-energy electrons from reaching the NEA surface. Fortunately, the surface adsorbates can be removed (along with the internal field and upward band bending) by modest heating (∼400 °C) in a UHV environment,40 but this limits the use of diamond to those secondary emission applications with heating capabilities. If, however, more stable NEA surfaces were available, this would enable the development of more versatile secondary emission devices as well as provide new opportunities to employ secondary emitters in a variety of operating environments.
To this end, exciting progress has been made in recent years toward the development of robust NEA surfaces on diamond. In particular, electronic structure calculations have been used to explore the chemical and electronic properties of light metals incorporated into an oxygenated diamond surface,41,42 and subsequent experimental studies have successfully verified the predictions.43,44 The oxygenated surface was chosen because, in the case of diamond, a native oxide does not form. As a result, a true O-terminated surface (with positive electron affinity) can be prepared, thereby allowing for precise control over the surface dipole formation (and reduced electron affinity) following adsorption by metal atoms. The research efforts have focused on light alkali and alkaline-earth metal atoms such as Li and Mg since heavy metals like Cs cannot form ordered, high-coverage surface layers on the diamond lattice and are thermally unstable. In particular, a size-dependent model developed from DFT calculations predicts that the adsorption of smaller alkali/alkaline earth metals (M) leads to a significant surface dipole resulting from the transfer of charge across M–O–C complexes, whereas the adsorption of the larger metals results in conventional dipole formation between the ionic adsorbate and a negatively charged surface.42 While all adsorbates were predicted to produce a true NEA on the O-terminated surface, the Li- and Mg-adsorbed surfaces were predicted to also have high adsorption energies, thereby suggesting the potential for more robust NEA surfaces.
These theoretical predictions were confirmed by the first successful demonstration of an NEA surface on O-terminated diamond following Li adsorption and thermal activation.43 Photoelectron spectroscopy indicated that Li atoms bond to the O-terminated regions of the diamond, while subsequent annealing to temperatures above 600 °C induces a structural change to a surface with a large surface dipole. The existence of an NEA was confirmed by total photoyield measurements (Fig. 5) (where photoyield is the number of emitted photoelectrons per incident photon). When an NEA is present, there is a clear onset in photoyield at Ephoton = Eg, and such an onset is observed at 5.5 eV (i.e., Eg of diamond). The NEA surface was found to be stable in air, as evidenced by the nominal change in photoyield upon exposure to atmosphere (Fig. 5). Importantly, this indicates that p-type surface transfer doping (and the associated upward band bending) does not occur, which is critical to achieving a high secondary electron yield.
While the lithiated diamond surface possesses a stable and robust NEA surface, there are several practical drawbacks including the high-temperature activation process and the inconvenience of working with Li. As an alternative, Mg adsorption has been investigated on the O-terminated diamond surface since computational studies predicted similar bonding chemistry and electron affinity for 0.5 ML of Mg compared to 1 ML of Li. In fact, an NEA of −2.0 eV was produced immediately upon Mg adsorption without any required activation process.44 This surprising result demonstrated that the two adsorbates have different adsorption properties on O-terminated diamond. While thermal energy is required to form the necessary Li–O–C complex, Mg atoms are able to form Mg–O–C complexes and their associated dipole without thermal assistance. Fortunately, the Mg-adsorbed surface is similar to the lithiated surface in being resistant to surface transfer doping in air. As a result, the NEA remains stable in air and water as well as at high temperature. Thus, NEA surfaces produced on O-terminated diamond following Mg adsorption are very promising for practical secondary emission applications due to their straightforward preparation and handling procedures as well as their robustness under different environmental conditions.
Additional theoretical studies have examined a broader range of adsorbate species, including transition metals, on bare, O-terminated, and N-terminated diamond surfaces.39 At bare surfaces, metals that form carbides (e.g., Ti, V, Al) are favored due to the smaller metal–carbon bond distance that improves both the NEA and thermal stability of the surface layer while also being more resistant to oxidation. For example, a DFT study examined different Ti structures on diamond and found that sub-ML Ti coverage on the bare surface created a stable titanium carbide with NEA as low as −1.6 eV. While these predictions still need to be experimentally verified, an earlier study reported consistent findings.45 More recently, a very exciting ab initio study has been reported that examined the scandium (Sc) adsorption of up to 1 ML on the bare, O-terminated, and N-terminated surfaces of diamond and found that Sc adsorption is very energetically favorable.46 In fact, most of the Sc-adsorbed surfaces displayed a substantial NEA, with the most negative values (as low as −3.73 eV) predicted for 0.25 ML of Sc (as shown in Fig. 6). More generally, the DFT calculations found that Sc adsorbs most strongly on the O-terminated surface owing to adsorption energies that are much larger than those for even a hydrogenated diamond, which suggests that very high thermal stability may be possible. Because of the high adsorption energies for Sc, NEA surfaces may be easier to prepare using Sc compared to other metal adsorbates since the precise coverage is not critical. Overall, the theoretical studies suggest that Sc adsorbed onto bare and O-terminated diamond surfaces may be the best candidates yet for electron emission applications, but experimental studies need to verify these predictions and demonstrate stable and robust NEA surface characteristics.
Finally, recent DFT calculations were performed for monolayer c-BN on diamond, and the electron affinity was examined for different surface terminations.47 In the case of the C–B–N terminated diamond (111) surface (i.e., with outermost N atoms), 1 ML of adsorbed H (corresponding to N–H bonds) results in an NEA of −4.00 eV. This is the largest NEA of any reported diamond surface, and it could provide very interesting secondary emission opportunities. However, such surfaces need to be experimentally synthesized and studied to characterize the surface electronic structure and secondary emission properties.
B. Transmission device: Diamond electron current amplifier
The development of robust NEA surfaces will be extremely beneficial to the field of secondary electron emission and can lead to improved device performance as well as new device applications. In the meantime, one approach to avoid surface damage during the secondary emission process is to use a transmission configuration as a way to protect the emitting surface from the impacting beam. In fact, such transmission schemes are used in some existing multiplier and detector technologies.3 However, the demonstration of extremely high secondary yields from diamond, together with long diffusion lengths projected from the analysis, provides an opportunity to explore other device applications that may benefit from such current multiplication capabilities. One exciting possibility is to use a diamond transmission stage to multiply the beam current produced by an electron source in a vacuum electronic device.48,49 Primary electron beams are used in a broad range of vacuum electronic device technology, including high-power microwave and millimeter-wave generation, accelerator technology, electron beam lithography, coherent x-ray sources, and ultrafast electron spectroscopy and microscopy. These technologies face constant demands to provide increasing levels of performance (e.g., higher power/frequency/resolution) from smaller, more efficient, more robust device components. Some mm-wave amplifier applications require high electron-beam current in increasingly smaller beams (e.g., current densities up to 100 A/cm2 from sub-mm size beams),50 while accelerator applications require pulsed electron beams with high charge and high beam quality (i.e., low emittance and narrow energy spread).49 Unfortunately, existing thermionic and field emitter technologies are constrained by intrinsic and/or practical limitations that prevent them from providing the beam characteristics required for such advanced applications. Fortunately, there is an opportunity to address this deficiency by exploiting the secondary electron multiplication capabilities of diamond to achieve the required beam enhancement.
Specifically, a current amplifier stage (based on diamond technology) can be used in conjunction with an existing cathode to produce higher-brightness electron beams.48,49 In this scheme, the diamond current amplifier serves to multiply the beam current produced by a primary cathode (e.g., thermionic, photo-, or field emitter) while also converting the current into a cold electron distribution having low energy spread. For example, such a current amplifier could be used in vacuum electronic amplifiers (as shown in Fig. 7) to provide the high current density and high beam quality required for higher-frequency (i.e., mm-wave) operation without the need for extreme beam compression. Importantly, the transmission configuration protects the emitting NEA surface from damage caused by beam impact. As a result, higher beam energies can be used to generate higher current and higher electron gain without the risk of surface damage. However, the transport process must be evaluated to ensure sufficient transmission efficiency, especially since relatively thick (∼1–100 μm) diamond films are required for structural integrity. Such long transport distances are in stark contrast to the relatively short transport distances in reflection measurements taken at low beam energies (e.g., ∼20–100 nm at 1–3 keV). Although the secondary electron generation and escape processes are identical in reflection and transmission configurations, the longer transport distance in the transmission configuration increases the probability of current loss caused by electron scattering, capture, and recombination.
Previous studies have evaluated secondary electron transmission through thin (0.1–10 μm) polycrystalline51,52 and single-crystal48 diamond films using a range of beam energies (1–20 keV) to generate the secondary electrons. In the case of polycrystalline diamond, electrons were transmitted through the conduction band over distances up to ∼1 μm and were emitted with a narrow energy distribution (FWHM ∼ 0.65 eV). However, low transmission gain (<5) was measured due, in part, to grain-boundary scattering during the diffusion-limited transport process. In the case of high-purity diamond (100), electrons were transmitted after traveling up to 8 μm through the film, and extremely narrow emission peaks were observed (FWHM ∼ 0.35 eV). These observations confirmed the improved transport efficiency predicted for single-crystal diamond due to the absence of grain boundaries. They also confirmed the narrow energy spread expected from the complete thermalization of the electron distribution during transport through many micrometers of diamond. This is seen in the EDCs taken in transmission and reflection configurations (Fig. 8), where notably, the high-energy tail present in the reflected EDC is completely absent in the transmitted EDC. However, the transmission gain was comparable to that measured from the polycrystalline diamond, which suggests that diffusion-limited transport is the limiting factor in achieving higher transmission gain. Consequently, an electric field is required inside the high-purity single-crystal diamond to sweep the generated secondary electrons to the emitting surface and thereby achieve the high gain needed for current amplification.
In fact, experimental studies have confirmed this by demonstrating internal transmission gains of over 200 when a field is present,49 as well as stable emission gains of up to 140 from the H-terminated surface.53 In these studies, a diamond electron amplifier (DEA) (Fig. 9) was characterized by directing electrons emitted from a thermionic cathode (with Eo = 0–10 keV) onto the metalized back surface of high-purity, single-crystal diamond wafers with a thickness of 0.15–0.3 mm. The DEA was connected to a high-voltage (HV) pulse generator that applied a potential gradient across the wafer with respect to an anode located downstream of the wafer (not shown). Transmitted electron current was measured at the front surface, with peak current densities >0.4 A/cm2 and average current densities of >0.1 A/cm2 demonstrated. While the gain increased with increasing Eo, the maximum gain plateaued at ≈1 MV/m field gradient at all values of Eo. In emission studies at the hydrogenated surface, the DEA properties were found to be stable, and it was demonstrated that repeated, extended exposure to air did not degrade its performance. From these and other experiments, a model was developed for the DEA that is being further investigated.54 In this model, primary electrons thermalize rapidly (in ∼100 fs) and create secondary electron/hole pairs, with the high gain of the device being possible because each primary high-energy electron produces a few hundred secondary electrons. A portion of the secondary electrons emerge from the hydrogenated NEA surface of the DEA with kinetic energy ∼0.1 eV and enter an acceleration cavity (not shown) where they are accelerated. The holes travel back to the metalized surface to enable the release of electrons from the HV pulse generator, thereby replenishing the emitted electrons.
Besides amplifying the primary beam current produced by the thermionic cathode, the DEA also produces electrons with a lower emittance and smaller energy spread than the primary electrons. In other words, it converts a relatively poor-quality electron beam from the thermionic cathode into a high-quality beam with higher charge and inherently low emittance. Such beams are especially important for accelerator applications where emittance and beam quality are critical, and a demonstration program is currently in progress at Brookhaven National Laboratory and Stony Brook University.54 In this program, researchers will construct and validate an electron beam prototype source based on a DEA. The DEA prototype features a chopper/buncher scheme for creating electron bunches that are sent into a diamond wafer located inside a single-gap quarter-wave resonator (QWR). Each primary electron will generate over 200 secondary electrons within the diamond wafer, and these secondary electrons will be accelerated to 100 keV by the QWR. Modeling analysis indicates that the total rms emittance (which is limited by space charge effects) will be <1 μm, and this can be improved with a more aggressive design. The chopper/buncher scheme provides an important additional benefit by using a double-bunch technique to automatically remove charge buildup on the diamond wafer.
Although this demonstration project uses an experimental scheme that is specific to accelerator environments, it will establish the ability of a thin film NEA diamond to amplify electron current in a practical device configuration. As such, it can provide guidance for the design and development of other device prototypes that employ an electron current amplifier. For example, extremely compact designs can be envisioned, such as by placing a field emission cathode (single tip or array) directly behind an NEA diamond film, which can then multiply the relatively low beam current typically produced by field emitters while also decreasing the energy spread and emittance. In such a scenario, the diamond film would require metalized back and front surfaces such that a bias could be applied across the film to generate an internal field. However, the metalized front surface would need to be appropriately patterned so as to retain exposed NEA regions for electron emission. Additionally, other schemes can be used to optimize the field profile inside the diamond such as through the use of appropriately doped layers at the back and front sides of the diamond film.55
C. Epitaxial UWBG semiconductors: c-BN and AlN
While research is underway to develop more robust NEA surface terminations on diamond, it would be extremely advantageous to have secondary emitter materials with a very low or negative electron affinity at a robust bare surface. Such surfaces would be less sensitive to impact by electron (or other particle) beams or to environmental contamination since there would be no adsorbed layer to desorb or react at the surface. As a result, more stable emission and longer operating life would be possible than with adsorbate-covered emitters. Fortunately, there are new opportunities to explore other promising UWBG semiconductors with the potential for robust low-electron-affinity surfaces. As discussed earlier, the electron affinity tends to decrease with increasing bandgap and, therefore, UWBG semiconductors with larger bandgaps than diamond have the potential for low or negative electron affinity as well. Among the UWBG materials of interest are c-BN and AlN, with Eg = 6.4 and 6.2 eV, respectively.
The development of these UWBG semiconductors (along with diamond) has received great attention in the past several years due to the demand for higher power and higher frequency electronic device performance that cannot be met by existing wide bandgap semiconductors (e.g., SiC and GaN with Eg = 3.3 and 3.4 eV). With their much larger bandgaps, these UWBG semiconductors can provide the high breakdown fields and high voltage operation required by advanced high-power electronics under development. Of particular note, a substantial research investment has been made in recent years by the Department of Defense and Department of Energy toward the epitaxial growth and fabrication of UWBG semiconductor films and device structures that can enable the transformation to next-generation electronic systems56 and electricity grids.57 With these new UWBG materials beginning to emerge, there will be a growing opportunity to explore their surface and bulk electronic properties and to evaluate their potential for secondary electron emission.
In principle, these UWBG semiconductors should be excellent secondary electron emitters for the reasons discussed earlier, namely, efficient electron transport through the conduction band and the trend toward smaller electron affinity as Eg increases. In fact, insulators with very large bandgap (i.e., MgO and BeO with Eg ∼8 and 10 eV, respectively) are commonly used as secondary emission cathodes. However, these insulator-based cathodes are often prepared using fabrication approaches such as thermal oxidation that lack the control needed to produce high crystalline quality or provide good electrical conductivity. In contrast, epitaxial growth techniques such as molecular beam epitaxy (MBE) and chemical vapor deposition (CVD) provide the ability to grow high-quality, single-crystal films needed for efficient transport, and they also permit the controlled incorporation of dopants needed for conductivity. Additionally, films can be grown with different surface terminations so that the surface properties can be evaluated in a controlled manner. As such, systematic studies can be designed to characterize these new UWBG materials as a function of the bulk and surface properties, thereby making it possible to identify promising secondary emission candidates and to optimize the material as needed for a high secondary yield.
Among the UWBG materials under development, c-BN is an especially promising secondary emitter candidate because of its similarities to diamond. In particular, it is isoelectronic with diamond and has a small lattice mismatch (1.4%) to diamond.34 Because of this compatibility with diamond, c-BN is being epitaxially grown on diamond substrates using MBE and CVD techniques,58,59 thereby providing opportunities to fabricate c-BN/diamond heterostructures. In addition, c-BN can be successfully doped with Si and Be for n-type and p-type character, respectively, thereby providing the electrical conductivity needed to avoid emitter charging while also enabling the design of tailored films. Finally, with a bandgap that is almost 1 eV larger than that of diamond, c-BN has the potential for low electron affinity at the bare surface and possibly similar NEA characteristics as diamond. Yet, in spite of these promising characteristics, there have been few experimental studies of c-BN due to the difficulty in synthesizing the cubic phase of the material. In this regard, BN is also similar to C in possessing both sp2-bonded phases (e.g., h-BN, graphite) and sp3-bonded phases (e.g., c-BN, diamond), with the hexagonal phase (h-BN) being more stable than the cubic phase (c-BN). As such, the growth of c-BN films is not a straightforward matter, and various techniques are employed to initiate and facilitate growth.
Nonetheless, several previous studies have examined the surface electronic properties of c-BN films. In one early study, an NEA was verified on single-crystal and polycrystalline c-BN (synthesized by high-temperature high-pressure methods) following H plasma treatment.60 Using photoemission techniques, the NEA at the hydrogenated surface was found to be thermally stable up to 800 °C. The NEA was removed upon heating above 1200 °C, but it was regenerated upon re-exposure to atomic H. More recent studies have revealed similar behavior from c-BN deposited by plasma-enhanced CVD on silicon substrates.61 Specifically, photoelectron spectroscopy confirmed the presence of an NEA on the as-deposited c-BN film as well as after H plasma treatment and annealing at 780 °C. Subsequent modeling suggested that N–H bonding is responsible for the NEA condition, which is consistent with the recent DFT calculations for monolayer c-BN on diamond mentioned earlier that also predicted N–H bonding at the NEA surface. Hence, such DFT studies could help guide in the growth of NEA c-BN on diamond, especially with respect to an optimized surface termination.
Given the unique properties of c-BN and diamond, the successful growth of NEA c-BN films on diamond could provide very interesting secondary emission opportunities. However, the epitaxial growth of c-BN is still immature, and c-BN material is not readily available for study. Fortunately, this situation should be changing soon with the recent investment in c-BN development, and greater opportunities will be available to evaluate the secondary emission characteristics of c-BN. In fact, some progress is already being made. Specifically, a thin layer of vanadium oxide on the c-BN surface has been found to possess an NEA,62 and such metal oxide layers may provide opportunities for exploring robust surface treatments for c-BN.
Finally, AlN is another promising UWBG semiconductor that has potential use in secondary electron emission applications. More generally, AlN is part of the AlxGa1−xN alloy system that extends from GaN to AlN with direct bandgaps of 3.4–6.2 eV, respectively. These alloy materials are of great interest to the electronic device community due to their high breakdown fields, high electron mobilities and velocities, and n-type doping capabilities. However, much of the research has focused on alloys with a lower Al content for reduced misfit strain in optoelectronic and electronic applications. As a result, the electronic properties of AlN surfaces have not been extensively studied to date. In particular, measurements of the electron affinity of AlN are relatively scarce and cover a wide range of values including an NEA as well as positive values from 0.25 to 1.9 eV.63–65 This large uncertainty is likely due to the chemically active surface of AlN that is vulnerable to oxygen contamination, as well as the possible growth-dependent surface properties of the different samples. However, advanced theoretical tools are now available to predict material and electronic properties, and a new theoretical study has recently reported an electron affinity of 0.6 eV,66 consistent with the lower value of 0.25 eV that was previously measured. With a small but positive electron affinity, AlN would be unable to provide the high secondary electron yields that H-terminated diamond (and possibly c-BN) can produce, but it may yet be a promising alternative to conventional metal oxide cathodes due to the superior crystalline properties of epitaxially grown films. In addition, it may be possible to further reduce the already-low electron affinity through surface modification.
At present, material challenges, such as controlled doping, limit the potential for secondary electron emission from AlN. While n-type doping is possible using Si donors, AlN can be unintentionally n-type doped due to impurities (such as oxygen). This n-type doping leads to upward band bending at the surface that impedes secondary electron transport to the surface. This behavior was observed previously in secondary electron yield measurements from an AlN surface in which the secondary yield increased only modestly (from 3 to 7) following the cesiation of the bare surface, even though the cesiated surface was found to have an NEA.67 On the other hand, p-type doping is advantageous in secondary emitter materials since downward band bending accelerates low-energy electrons toward the surface. However, AlN has been difficult to dope with p-type acceptors due to issues with compensation and self-trapping. Fortunately, significant progress has been made to address these doping issues in recent years. In fact, a new study has reported n-type and p-type doping in AlN and the realization of a pn AlN diode.68 With an increased focus on AlN for both high-power electronics and optoelectronics technologies, research efforts will be directed heavily toward addressing carrier generation and transport issues since these are critically important to device development.
Of course, the surface properties are also critical to secondary electron emission, and a specific challenge is the presence of surface states on the AlN surface that make it chemically reactive, especially to oxygen. However, this is a critical issue for device-related research efforts as well since an understanding of and control over the surface termination and surface electronic properties of AlN are critical for the growth of high-quality heterostructures. In fact, a new study has reported the successful use of an aluminum-assisted cleaning technique during the homoepitaxial MBE growth of AlN that produces smooth surfaces with no noticeable defects as are found in the starting AlN growth templates.69 While the electronic properties of these (and other) AlN surfaces need to be fully characterized, the potential exists to begin exploring the secondary emission capabilities of AlN and other UWBG materials that are being reproducibly grown using controlled epitaxial growth techniques.
While c-BN and AlN have many excellent properties, including the potential for low or negative electron affinity, these materials are still relatively immature. However, with the increasing availability of epitaxial AlN and c-BN films, there will be an opportunity to experimentally evaluate the secondary emission characteristics as a function of growth conditions and surface termination. In addition, these materials offer the possibility of heterostructure fabrication such that new device concepts can be envisioned. For example, both n- and p-type doping capabilities exist for c-BN and have recently been reported for AlN, which makes possible the design of p–i–n diode structures with controlled internal fields that can assist during electron transport to the surface. Similarly, bandgap engineering is possible with AlxGa1−xN alloys that may enable the design of novel emitter structures. While the presence of an NEA (or very small electron affinity) is the key attribute needed for high secondary electron yields, the electron transport properties are also critically important to the emission process. As such, these materials offer opportunities to develop novel ways to control or enhance the electron transport process to the surface.
D. Epitaxial UWBG oxides
While UWBG semiconductors offer the potential transport and emission properties needed for high secondary yields, these materials are not yet available for device applications. In the meantime, there are opportunities to improve the secondary emission performance of UWBG oxides that are currently used in many practical devices. For example, metal oxides such as BeO, Al2O3, and MgO are used as coatings on conducting electrodes in commercial dynodes (e.g., in electron/photon multipliers) and as secondary emitter materials in microwave tubes (e.g., crossed-field devices) because of their high yields (∼3–4). More recently, MgO has demonstrated excellent performance as a secondary emission cathode in a newly developed flat electron emission lamp.70 However, in all these applications, higher secondary yield would greatly improve the device performance, such as by enabling lower-voltage operation and thereby reducing the device power consumption and improving the efficiency; or by producing higher gain and thereby increasing the signal and reducing noise in the device. However, these insulator materials are hindered by electrical charging and poor crystalline quality that prevents higher yields from being achieved.
As discussed earlier, insulators with a very wide bandgap benefit from a low electron affinity that allows a larger portion of the internal electron distribution to be emitted compared to metals and smaller bandgap semiconductors. However, these UWBG insulators, unlike their UWBG semiconductor cousins, do not have sufficient electrical conductivity to replenish the emitted electrons when the secondary yield exceeds 1 (i.e., net flow of electrons out of the material). As a result, the emitter charges up and an internal field is established that reduces the secondary yield to 1. In practical applications, different approaches have been used to provide electrical conductivity to insulator emitter materials, including the use of a very thin insulator layer on a metal substrate (which supplies the conductivity) and the incorporation of metal crystallites in the insulator bulk.19 However, each of these approaches presents challenges. In the case of oxidized Be, only a 10–20 Å-thick layer of BeO is formed and, consequently, the thin emitting layer is slowly eroded in a harsh environment and must be replenished by the use of an oxygen source. In the case of Ag–Mg or Cu–Be alloys, the size and distribution of Ag or Cu crystallites in the composite materials are difficult to control, and such materials suffer from localized charging problems. The secondary emission from UWBG insulators can also be limited by poor electron transport in low-quality polycrystalline materials, such as through hopping conduction rather than conduction-band diffusion as well as by scattering and capture by defects and traps.
While the issue of conductivity may be challenging to overcome, there are advanced deposition techniques available today that provide opportunities to improve the conductivity of UWBG oxide films through controlled doping and to improve the electron transport properties through higher crystalline quality. In fact, a recent study demonstrated that the improved microstructure and electrical conductivity of a thin MgO/Au composite film deposited onto a thin Au buffer layer resulted in higher secondary electron yield (i.e., δmax ∼ 11) and reduced charging.71 For thicker films, epitaxial layers can be deposited on conductive substrates using MBE and CVD growth techniques to produce metal oxide layers with controlled thickness, composition, and purity as well as high crystalline quality. MBE growth techniques are ideal for investigating different oxide compounds in a systematic and controlled manner, and new classes of functional materials (such as Perovskites) are being developed that provide the ability to tune the electronic and material properties of the oxide. While dedicated research is needed to investigate the doping possibilities of oxide layers, the sharp and intimate interface produced at the substrate–oxide heterojunction should improve the electrical conductivity.
For thinner films, deposition techniques such as atomic layer deposition (ALD) and atomic layer epitaxy (ALE) offer great opportunities to explore metal oxide systems with fine control over film thickness and composition. Moreover, ALD produces conformal coatings, which can be of great practical benefit when depositing on irregularly shaped electrodes or emitting surfaces. For example, ALD has been used to synthesize Al2O3 and MgO films as potential secondary emissive layers in the channels of micro-channel plates. Studies investigated the effects of film thickness and surface chemical composition on the secondary yield, and a yield as high as ∼7 was reported from a 20-nm-thick MgO film without significant charging issues.72 In addition, many binary and ternary oxides have been synthesized using ALD (such as MgAl2O4),3 thereby demonstrating the potential versatility of such deposition techniques and providing an opportunity to explore new secondary emitter material candidates.
Secondary electron emission is critical to a wide range of vacuum electronic devices and instrumentation that operate under very different conditions, ranging from sensitive particle detectors to high-power crossed-field devices. Yet, although the device applications are quite disparate, all would benefit greatly from higher secondary yield and more robust emitter performance than is currently available with existing secondary emission materials. A higher yield would provide a higher gain, greater sensitivity, and improved signal-to-noise ratio, while also enabling lower voltage operation for increased efficiency and reduced power consumption. Moreover, if high-yield emitters were robust under different environmental conditions, new device applications and novel device concepts would be enabled, along with longer device lifetime. However, it is an extremely challenging task to accomplish either goal, let alone both. First, most materials are configured such that the surface energy barrier prevents all but the most energetic secondary electrons from escaping into vacuum, thereby making it difficult to substantially increase the yield. As a result, applications that require a high secondary yield often use monolayer coatings of alkali metals to lower the surface barrier, but these reactive coatings require extremely controlled vacuum conditions. Even with such care, the secondary emission process itself can damage the surface through the electron bombardment of the fragile adsorbate layer. Conversely, applications that require robust performance often use UWBG metal oxides with inherently low energy barriers at the bare surface. However, a mechanism is required to continuously replenish the emitted electrons and, thus, the secondary yield is limited by charging issues caused by the poor electrical conductivity of the metal oxides.
Fortunately, there are exciting opportunities to overcome these challenges through the development of new classes of secondary emitter materials that can simultaneously provide high yield and robust performance. Specifically, the UWBG semiconductors diamond, c-BN, and AlN are newly emerging electronic materials that have unrivaled potential for secondary emission applications due to their large bandgaps, inherently low or negative electron affinity, and excellent carrier transport properties. Importantly, these opportunities are possible due to advances in epitaxial growth techniques such as MBE and CVD that are being used to explore such extreme material systems, along with advanced theoretical techniques such as DFT that can predict the material and electronic properties and thereby provide critical guidance in the design of high-performance secondary emission materials.
Among secondary emitter candidates, diamond is unique in having a true NEA at the H-terminated surface, and it has demonstrated exceptional secondary emission capabilities that are superior to those of other secondary emission materials in use today. Importantly, extremely high yields have been achieved from a stable NEA surface without the use of fragile alkali-metal surface coatings, while energy distribution measurements confirm the efficient emission of low-energy thermalized electrons in the absence of a surface barrier. These NEA diamond surfaces can be used in applications that operate under good vacuum conditions and low-intensity beam impact or in transmission mode. In fact, a demonstration project is already underway to use diamond as an electron amplifier for primary beam sources (e.g., thermionic, photo, field emitters) in order to generate high-brightness electron beams with exceptional beam quality. Additionally, promising research is underway to develop robust NEA surface terminations on diamond to enable use in a broader range of secondary emission applications, and successful candidates have already been identified with demonstrated thermal stability in vacuum as well as in air (and even water!).
New opportunities are also emerging to explore the secondary emission characteristics of other promising UWBG semiconductors such as c-BN and AlN that possess many of the same qualities as diamond. Importantly, these materials offer the possibility of heterostructure fabrication such that new device concepts can be envisioned. While the development of these materials is still in the early stages, rapid progress is being made in epitaxial growth research to achieve the superior electronic properties needed for both advanced electronic device capabilities and enhanced secondary emission performance. In the meantime, substantial improvements in UWBG oxides can be realized by exploiting the superior synthesis capabilities of epitaxial growth techniques and atomic layer deposition. Improved secondary emission has been demonstrated from metal oxide layers prepared using advanced deposition techniques, but the issue of increased electrical conductivity through impurity doping has been relatively unexplored. As such, this represents a critical area of research with the potential to greatly impact the field of secondary electron emission.
This work was supported by the Office of Naval Research. J. E. Yater wishes to thank D. S. Katzer for many helpful discussions.
Distribution Statement A. Approved for public release. Distribution is unlimited.
Conflict of Interest
The authors have no conflicts to disclose.
J. E. Yater: Conceptualization (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (lead).
Data sharing is not applicable to this article as no new data were created or analyzed in this study.