Figure 1 illustrates the schematic model of the SEY suppression method using a micro-porous surface. The regular and non-regular structures artificially constructed on the metal surface are approximated by cylindrical wells and are defined by three primary parameters, e.g., the well radius R, the well depth H, and the distance between adjacent wells D. The figure indicates that the electron trajectory in a single well is bent and multi-generation electrons can be absorbed by the well.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. A schematic diagram of the SEY suppression method using a micro-porous surface characterized by a cylinder of height H and radius R.

When the radius of the well R is far smaller than the wavelength of the propagating wave λ, i.e., R ≪ λ, the electron uniformly moves to the wall of the well while the electromagnetic wave can be neglected in the well. Based on the Monte Carlo method and a phenomenological statistical model of the SEY, electron trajectories can be tracked, and the emission probability from the well can be calculated. Then, the SEY of the well is calculated and defined as δ_w. Combined with the SEY of the flat surface, δ_f, the SEY of the micro-porous surface is given by

Through the Monte Carlo simulation with the cylinder well model of certain height and radius,^[15] the SEY according to different incident energies and angles are recorded. The actual maximum yield δ_max0 of the porous cylindrical surface is fitted by the function

Thus, micro-porous structures can be designed to obtain a maximum decrease of the SEY as shown in Fig. 2.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. Variations of maximum yield of porous surfaces with porosity P and height-to-width ratios AR.

It can be concluded from Fig. 2 that the deeper and denser the pores, the smaller the maximum SEY is. Correspondingly, these pores on the surface are likely to suppress multipaction in microwave components due to the correlation between SEY and multipaction. Artificial structure such as the grooved surface is no longer applicable for microwave components due to the change of the electric performance. Furthermore, the increase of insertion loss due to the roughening of the surface is another challenge that should be suppressed for practical components.

At first, micro-cylindrical structures scaled from nanometer to micrometer on the metal surfaces are proposed for multipaction suppression when applying a field with the wavelength of centimeter scale (f = 300 MHz∼ 30 GHz). The secondary electrons are efficiently trapped in the well and it is conjectured that the insertion loss of the component increases little while R < 0.0001λ. For the approximation calculation of the loss due to the micro-structures, a resonant cavity is utilized and its quality factor, Q₀, is measured. Replacing the top plate with one of a porous surface, the surface impedance is calculated from

Moreover, the photolithography pattern method and the directly chemical etching method are proposed for the surface treatment. The dielectric loss caused by plating dielectric foam on the metal is avoided.

Regular and irregular micro-porous surfaces are fabricated on Ag-plated aluminum samples for validating the SEY suppression method. The regular surfaces are realized with a predefined shape using a traditional photolithography pattern process, and the direct chemical etching method is used for realizing the irregular micro-porous surfaces.^[15]

All Ag-plated samples are initially ultrasonically cleaned to remove dust and contamination. Figure 3 shows scanning electron microscopy (SEM) images of three different types of samples, i.e., the original flat Ag-coated sample, the photolithography pattern processed sample, and the directly chemical etched samples. A group of samples with irregular micro-pore arrays using ferric nitrate (Tianjin Yongsheng Fine Chemical Co., Tianjin, China, analytical reagents) in an aqueous solution with a mass fraction of 20% for tens of seconds at 50 °C and another group of samples with regular micro-pore arrays using traditional photolithography are fabricated. The parameters characterizing the physical micro-structures of the pores are measured and recorded in Tables 1 and 2.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. SEM images of samples with different surface treatments: (a) Ag-coated; (b) photolithography patterned; (c) directly chemical etched.

Table 1.

Table 1.

Table 1. SEY of samples under different conditions of photolithography pattern. . Samples AR P Calculated δmax Measured δmax Rs/mΩ 25-s etching 0.30 45% 1.56 1.56 69.3 35-s etching 0.36 57% 1.52 1.38 139.5 45-s etching 0.36 44% 1.53 1.47 104.7 50-s etching 0.44 26% 1.55 1.65 83.6

Table 1. SEY of samples under different conditions of photolithography pattern. .

Table 2.

Table 2.

Table 2. SEY of samples under different etching conditions. . Samples AR P Calculated (δmax/E1)/eV Measured (δmax/E1)/eV Rs/mΩ 5-s etching 0.75 83% 1.30 / 137 1.37 / 137 105.1 10-s etching 0.93 94% 1.20 / 167 1.20 / 200 114.3 20-s etching 1.17 93% 1.13 / 193 1.25 / 160 159.2 40-s etching 1.38 83% 1.15 / 187 1.17 / 200 112.4 60-s etching 1.34 80% 1.17 / 177 1.34 / 150 110.8

Table 2. SEY of samples under different etching conditions. .

Correspondingly, it is reasonable to obtain the following conclusions:

Using theoretical analysis and calculation, the micro-structures on the metal surface can be ‘designed’ to suppress the secondary electron emission phenomenon and the suppression effect is quantified for the first time.

It should be noted that the plating thickness of silver is 10 μm which is thick enough to guarantee that the aluminum base is not naked after the surface treatment processing.

Fabrication of a porous surface on the practical components brought a new challenge to the surface treatment technology compared with that on the samples. Corners, steps, bent and wedged waveguides made it difficult to construct a uniform porous surface with a predesigned size. In addition, the large physical sizes of practical components exacerbate the problem. To solve these problems, some practical efforts have been made. On the one hand, the spin processor (Chemat Technology, Spin Master 100) burdened larger load is utilized for the photolithography patterned processing and the etching method is optimized. On the other hand, the electromagnetic analysis is performed on the components to find the sensitive area where the field density is strongest and multipaction is most likely to occur.

Parameter optimization shows that larger P and A_R lead to a better SEY suppression effect. To obtain a compromise between processing limitations and parameter optimization in practice, samples with diameters of about 20 μm and depths of about 8 μm were chosen for multipaction suppression in metal components.

The total SEY consists of two different populations: the true secondary electrons and the backscattered secondary electrons. We measure the SEY using the conventional sample-current method.^[16] The SEY measurement is conducted by the apparatus shown in Fig. 4. The measurement apparatus is comprised of three major systems: ultra-high vacuum chamber, test system, and data processing system.^[16,17] The test system is composed of a low energy electron gun, sample platform, and pico-ammeter. The electron gun is a thermionic electron gun, which is integrated in the electron spectrometer (DESA150, Staib Instruments, Germany).

Fig. 4.

Figure Option ViewDownloadNew Window

Fig. 4. Configuration of the SEY measurement platform, 1: analysis chamber; 2: loading chamber; 3: sample plate entry; 4: sample transfer plate; 5: rack and pinion transfer mechanism; 6: sample plate travel; 7: four-dimensional manipulator; 8: ion gun; 9: electrostatic energy analyzer; 10: electron gun; 11: residual gas analyzer; 12: gate valve; 13: ion pump and Ti sublimation pump; 14: molecule pump and mechanical pump.

The SEY test results of witness samples of practical components are shown in Fig. 5. The largest feasibly obtainable porosity on the components is about 57%. It can be seen that the curves of the photolithography patterned and directly chemical etched sample surfaces demonstrate a good SEY suppression effect over the entire energy range considered. The directly chemical etched sample achieves the smallest SEY. This type of micro-structured surface and etching method can be extended to the SEY suppressions of other metals. Thus, the micro-porous surface offers an alternative approach to SEY suppression in relevant areas such as multipaction effects in satellite payloads and electron cloud effects in accelerators.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. The SEY characteristics of three different types of surfaces on practical components.

Based on the theoretical analysis, the physical dimensions of micro-pores on the surface are designed to realize the optimal suppression effect of SEY in practical components. However, the calculated SEE parameters are not accurate enough for the multipaction simulation nor for threshold analysis neither. Thus, combining with the practical SEY properties of material with concrete morphology,^[26,27] a quantitative relationship between the surface treatment and discharge threshold is determined and simulated through the EM-PIC method.

The analytical model established by Li et al.^[18] is used for curve fitting of the SEY results obtained in our simulation and analysis method. As shown in Fig. 6, the parameters of the mathematical model, including δ_max0, E_max0, and E₁ can be extracted from the measured data.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. Typical SEY curve of a metallic material.

The values of δ_max0, E_max0, and E₁ for the silver-plated sample, photolithography pattern processed sample, and directly chemical etched sample are fitted with the results listed in Table 3. According to the previous research, the lower δ_max0 as well as higher E_max0 and E₁, the multipactor threshold would become higher, while the exact threshold depends on the structure of the microwave component and electromagnetic frequency. Therefore, as shown in Table 3, both the photolithography processing method and direct chemical etching method can suppress multipaction. The SEY can be obtained for an arbitrary incident angle through interpolation, and the fitting model is used for the multipaction simulation.

Table 3.

Table 3.

Table 3. SEY parameters of three samples. . Surface treatment method δmax0 Emax0/eV E1/eV Silver plating 2.17 285 29.76 Photolithography processing 1.70 380 35.36 Direct chemical etching 1.58 450 37.17

Table 3. SEY parameters of three samples. .

As shown in Fig. 7, a C-band impedance transformer is used to validate the effectiveness of suppressing the SEY by micro-porous surface treatment and the improvement of the multipaction threshold. The base material is Ag coated aluminum alloy. A three-dimensional and self-consistent numerical simulation method of the multipactor effect in an arbitrary microwave passive component is adopted for threshold analysis.^[18,19] The multipaction thresholds of the transformer obtained through simulations and experimental investigation are listed in Table 4.

Fig. 7.

	Figure Option ViewDownloadNew Window
	Fig. 7. Three-dimensional models of the impedance transformer: (a) dimensions and (b) geometry.

Table 4.

Table 4.

Table 4. Multipaction thresholds of a C-band impedance transformer. . Surface treatmentmethod Measuredthreshold/W Simulationthreshold/W Insertionloss Silver plating 2100 2219 0.13 Photolithography processing 5500 5625 0.15 Direct chemical etching 7200 7375 0.15

Table 4. Multipaction thresholds of a C-band impedance transformer. .

Figures 8 and 9 show the patterned transformers using photolithography processing and direct chemical etching, respectively. Among the three different types of surface, the directly chemically etched sample surface demonstrate the smallest yield and the highest threshold. Compared with for the Ag-plated component, the value of δ_max0 for the directly chemical etched sample is reduced from 2.17 to 1.58, and the threshold power P_threshold increases from 2100 W to 7200 W. In addition the value of E_max0 increases from 285 eV to 450 eV. The simulation results and experimental results both show that micro-porous surfaces can effectively increase the multipaction threshold of microwave components. The multipaction thresholds of microwave components using regular and irregular micro-porous surfaces are about 4.18 dB and 5.35 dB higher than those using the Ag-plated surface, respectively.

Fig. 8.

	Figure Option ViewDownloadNew Window
	Fig. 8. Photographs of the photolithography patterned impedance transformer: (a) component and (b) witness sample.

Fig. 9.

	Figure Option ViewDownloadNew Window
	Fig. 9. Photographs of the directly chemically etched impedance transformer: (a) component and (b) witness sample.

In addition, the insertion loss of the processed component increases from 0.13 dB to only 0.15 dB in the transmission band. It can therefore be reasonably conjectured that the pore size is sufficiently small compared with the electrical dimensions of the components, and it is effectively ignored by the electromagnetic radiation. While the macro-periodic structures for plasma discharge suppression in dielectric windows are not feasible for microwave components due to the resulting dielectric and power losses, these two surface treatment methods show great potential in anti-multipaction design in the space industry.