Effects of ion source etching on the impedance performances of quartz component

doi:10.21203/rs.3.rs-3794405/v1

Quartz element (i.e., active quartz crystal oscillator or resonator) is coated on the quartz crystal with an Ag/Ni double-layer film. An ion source etching system is used to etch the Ag film surface of the quartz crystal through masks of different sizes. Three groups of masks were adopted for quartz crystal etching, the sizes of which are A (1.30×1.10 mm²), B (1.70×1.30 mm²), and C (2.42×1.62 mm²) with the remaining area not being etched. During the etching process, a frequency counter was connected to monitor the load resonance frequency (FL, MHz), load resonant frequency difference (FLD, ppm), resistance at series resonant frequency (RR, Ω, related to impedance performance). After ion source etching, the quartz element (mask A) corresponds to FLD (-385.5 ~ -256.8 ppm), and load resonant impedance RR ranges from 8.7 to 9.1 Ω, representing a 6.32% average impedance reduction. Using masks C, the quartz element corresponds to FLD (-282.9 ~ -241.1 ppm), and load resonant impedance RR ranges from 10.0 to 21.1 Ω. The average impedance divergence is 13.04 Ω. It is speculated that the size of the etching area of the quartz crystal using mask B or mask C was too large, and some remaining zones have been etched improperly, which led to an increase in impedance and also caused crystal frequency instability or oscillation failure. Precise etching and mask selection when producing quartz components is essential to maintain their functionality and stability.

Quartz element

Coating

Ion source

Mask selection

Quartz crystals exhibit the piezoelectric effect, and the stability of their oscillation frequency is primarily determined by parameters such as the cut angle, crystal thickness, and quality factor of the crystal. A crystal oscillator is an electronic circuit that utilizes the mechanical resonance of a piezoelectric crystal to generate a precise electrical signal at a specific frequency [1]. Conventional manufacturing processes involve cutting and grinding, which require exact mechanical machining techniques to control the production quality of quartz crystal components. The oscillation frequency can be adjusted by controlling the thin film's thickness on the quartz crystal's surface [2]. A versatile load sensor is crucial as it captures diverse biosignals across a spectrum of load ranges. Murozaki et al. [3] reported a load sensor is introduced, utilizing a quartz crystal resonator (QCR) with an extensive measurement range spanning from 0.4 mN to 600 N, and its temperature sensitivity concerning load has been enhanced to -7 Hz/°C (།18 mN/°C). A design approach for this sensor is suggested to enhance the load measurement range, involving the mitigation of QCR buckling and the utilization of a thinner QCR.

Silver (Ag) nanostructures have attracted significant research attention due to their distinctive optical properties characterized by robust resonant electromagnetic interaction with light. This property allows for adjusting surface plasmon resonance across a broad optical spectrum ranging from 300 to 1000 nm [4]. Ag films find extensive applications, including but not limited to chemical sensors, biological sensors, surface-enhanced Raman spectroscopy, surface-enhanced fluorescence spectroscopy and photovoltaics [5]. The augmentation of infrared active vibrational modes in Ag films on silicon substrates is explained as a coupling mechanism between molecular vibrational modes and intensified light fields close to the surface [6]. The close arrangement of Ag nanoparticles leads to a more pronounced enhancement of electromagnetic coupling between them, which may be attributable to the robust confinement of the local electric field [7].

In this study, a bi-layer film composed of Ag (180 ~ 200 nm) and Ni (~ 40 nm) was coated on the quartz crystal, namely Ag/Ni/quartz crystal (see Figs. 1 and 2). Figure 3 shows the quartz component, namely active quartz crystal oscillator, or resonator. The Ni layer serves as an intermediate layer to enhance the adhesion of the Ag film coating. An ion source etching system is used to etch the Ag film surface of the quartz crystal through masks of different sizes. Three groups of masks (see Fig. 4) were adopted for quartz crystal etching, the sizes of which are A (1.30 mm × 1.10 mm), B (1.70 mm × 1.30 mm), and C (2.42 mm × 1.62 mm) with the remaining area not being etched. These masks are called masks A, B, and C respectively. During the etching process, a frequency counter was connected to monitor the load resonance frequency (FL, MHz) of the crystal and the resistance at series resonant frequency (RR, Ω).

The plasma generation (ionized gas) requires extremely high temperatures (around 10,000°C). However, artificial plasma, typically called low-temperature plasma, is generated at room temperature. The key to creating artificial plasma is to provide energy that allows the outer electrons in atoms to overcome the binding energy of the atomic nucleus. The reactor chamber for generating plasma includes several essential elements: 1. Two opposing electrodes. 2. High-voltage direct current or high-frequency alternating current power supply. 3. Vacuum system.

Argon [8] gas contains a small number of free electrons. When a voltage is applied to the two opposing electrodes, the free electrons are accelerated in the electric field and gain enough energy. When these electrons collide with argon atoms, they can remove an outer electron from the atom, resulting in a positively charged argon ion. The removed and original free electrons continue to be accelerated by the electric field and further collide with other argon atoms. This process leads to a cascade effect, producing a plasma. The plasma exists between the two electrodes and is divided into two regions:

The glowing central region is called the luminous zone. The color of the light emitted depends on the gas used to generate the plasma.

The region where there is no glow, between the luminous zone and the electrodes, is called the dark region or sheath.

The artificial plasma is created by providing energy to free electrons in a gas, resulting in ionization and plasma formation in the reactor chamber, with different regions having distinct characteristics and properties [9–12]. The frequency of quartz components is adjusted by controlling the thickness of the quartz crystal thin film, which, in turn, changes the oscillation frequency. When the thin film on the quartz crystal is more delicate, it can oscillate at higher frequencies. However, the demand for high-frequency communication applications has grown, necessitating breakthroughs in providing clock-controlled quartz components. In addition to the limitations posed by miniaturization, quartz components must be smaller. However, traditional manufacturing processes for miniaturized quartz crystals face challenges such as increased crystal impedance leading to reduced oscillation stability and increased frequency sensitivity, which can result in frequency shifts in the oscillation frequency.

To achieve precise frequency adjustment in quartz components, ion etching removes excess silver film thickness from the chip surface [13–15]. Initial film deposition typically uses a sputtering method to mass-produce quartz crystals in the quartz component industry's manufacturing process. Due to batch production processes, achieving precise film thickness control is challenging, resulting in frequencies typically controlled within the range of -1500 to -4000 ppm (parts per million), where ppm represents the deviation of the loaded resonance frequency relative to the standard frequency. Subsequently, each product is individually etched to ensure its final frequency meets customer requirements. Therefore, etching technology is used to obtain accurate clock signals. The relationship between frequency and impedance is determined by crystal design [16–17], and improper etching during the production process can increase impedance. Avoiding this increase is a significant challenge in the industry.

This research uses ion source etching equipment (Showa Shinku CO., LTD, SFE-7051CM) to process quartz components. This machine is designed for thin-film etching to fine-tune the frequency of quartz components. The ion source etching process involves argon (Ar) gas discharge, electron beam collision with gas atoms (or molecules), and charged ion beams causing the ejection (etching) of material from the working surface. Due to the interaction between ions with a specific energy and the thin film material, it generates electrons and X-ray radiation. Therefore, avoiding direct exposure in areas such as electrode legs and non-etched regions is essential. Additionally, during the etching process, direct eye contact should be avoided. Due to the difference in the power of the ion source during the etching process, each unit (quartz crystal) underwent two different etching rates, i.e., a faster etching rate (about 800 ~ 1200 ppm/sec) first and then a slower etching rate (about 60 ~ 80 ppm/sec) was used. The etching time was about 2 to 5 seconds (deviation may occur due to different quartz crystal frequencies).

Measurement frequency equipment and test fixtures, as shown in Fig. 5. The quartz components from various samples are placed on the testing platform, and measurements are taken using test probes. The testing is performed with the help of the S&A 250B Network Analyzer from Saunders & Associates, utilizing its testing software. Data is recorded for measurements taken before and after the etching process on the quartz crystals of the products, and the results are analyzed.

The main monitoring reference data for this measurement include standard frequency (series resonant frequency, FR), load capacitance (CL), load resonance frequency (FL, MHz), load resonant frequency difference (FLD, ppm, see Eq. 1), resistance at series resonant frequency (RR, Ω, related to impedance performance). The RR variation is measured using the Saunders & Associates 250B (NETWORK ANALYZER testing software).

load resonant frequency different (FLD)=\(\frac{[\text{F}\text{L}\left(\text{M}\text{H}\text{z}\right)-\text{F}\text{R}\left(\text{M}\text{H}\text{z}\right)]}{\text{F}\text{R}\left(\text{M}\text{H}\text{z}\right)}\times {10}^{6}\frac{1}{{10}^{6}} \left(\text{p}\text{p}\text{m}\right)\) (1)

The specifications of the quartz component are as follows: FR = 38.880000 MHz, with CL = 12 pF. The chip size and electrode size are the same, and three (Groups A, B, C) different mask sizes are used, as shown in Table 1. The masks are placed on the surface of the quartz component, as depicted in Fig. 6. Group A mask size is close to the electrode size, preventing non-ion source etched zones and electrode legs from being etched. Group B mask size is slightly larger than the electrode size, and non-ion source etched zones and electrode legs may be etched. Group C mask size is larger than the electrode size, resulting in many non-ion source etched zones and electrode legs being etched.

Table 1

Sample grouping (Groups A, B, C).
Sample Grouping/No.		FR (MHz)	Chip size (mm²)	Electrode size (mm²)	Mask size (mm²)
A	1 ~ 5	38.880000	5.026 × 2.511	1.25 × 1.00	A (1.30 × 1.10)
B	6 ~ 10	38.880000	5.026 × 2.511	1.25 × 1.00	B (1.70 × 1.30)
C	11 ~ 15	38.880000	5.026 × 2.511	1.25 × 1.00	C (2.42 × 1.62)

The primary purpose of ion source etching is to process the thin film thickness. Both sides of the quartz crystal are deposited with an Ag/Ni/quartz crystal bi-layer film. The mask is placed on the front side, i.e., the lower layer, and ion source etching is performed from the bottom to the top, only one side. The overall thickness of the film determines the frequency since only one side is etched. Adjusting the thickness of the Ag film achieves the fine-tuning of the quartz components' frequency. The measurements include the FL (MHz), FLD (ppm), and RR impedance (Ω) of the quartz crystal element after ion etching. Table 2 lists the measurement data of quartz components before and after ion etching of Groups A, B, and C.

Before ion source etching, the quartz crystal elements (Groups A, No. 1 to No. 5) exhibit different FL (38.745099 ~ 38.777337 MHz), FLD (-3469.7 ~ -2640.5 ppm), and load resonant impedance RR (8.4 ~ 10.5 Ω), with an average impedance of 9.48 Ω. After ion source etching, the surface of the quartz crystal with Ag thin film (Groups A, No. 1 to No. 5) corresponds to different FL (38.865013 ~ 38.870015 MHz), FLD (-385.5 ~ -256.8 ppm), and load resonant impedance RR as follows: No. 1 = 8.7 Ω, No. 2 = 8.9 Ω, No. 3 = 8.9 Ω, No. 4 = 9.1 Ω, and No. 5 = 8.8 Ω. The load resonant impedance RR ranges from 8.7 to 9.1 Ω, with an average impedance convergence of 8.88 Ω, representing a 6.32% average impedance reduction.

Before ion source etching, the quartz crystal elements (Groups B, No. 6 to No. 10) exhibit different FL (38.723784 ~ 38.791838 MHz), FLD (-4017.9 ~ -2267.6 ppm), and load resonant impedance RR (8.7 ~ 13.2 Ω), with an average impedance of 10.0 Ω. The mask size of Group B is slightly larger than Group A's, so some electrode zones will be etched. After ion source etching, the surface of the quartz crystal with Ag thin film (Groups B, No. 6 to No. 10) corresponds to different FL (38.868439 ~ 38.870232 MHz), FLD (-297.4 ~ -251.2 ppm), and load resonant impedance RR as follows: No. 6 = 9.6 Ω, No. 7 = 9.8 Ω, No. 8 = 9.7 Ω, No. 9 = 9.8 Ω, and No. 10 = 9.7 Ω. The average impedance convergence is 9.74 Ω, representing a 2.6% average impedance reduction.

Before ion source etching, the quartz crystal elements (Groups C, No. 11 to No. 15) exhibit different FL (38.767371 ~ 38.794739 MHz), FLD (-2896.8 ~ -2192.9 ppm), and load resonant impedance RR (9.1 ~ 14.5 Ω), with an average impedance of 11.22 Ω. The mask size of group C is more significant than that of groups A and B, so more electrode zones will be etched. After ion source etching, the surface of the quartz crystal with Ag thin film (Groups C, No. 11 to No. 15) corresponds to different FL (38.868999 ~ 38.870627 MHz), FLD (-282.9 ~ -241.1 ppm), and load resonant impedance RR as follows: No. 11 = 12.2 Ω, No. 12 = 21.1 Ω, No. 13 = 11.0 Ω, No. 14 = 10.9 Ω, and No. 15 = 10.0 Ω. The average impedance divergence is 13.04 Ω, an increase of up to 46.8% compared to the standard impedance value (~ 8.88 Ω). The data reveals that the larger mask size in Groups A, B, and C results in exposure to the quartz crystals non-etched zone, causing a rapid increase in load resonant impedance RR after etching. This leads to the instability of the quartz crystal frequency or failure to oscillate (i.e., producing defective products). This result aligns with Eq. (2), where improper etching of the quartz crystal electrode feet and non-ion-etched areas leads to a decrease in the cross-sectional area (A), resulting in an undesired increase in resistance.

R = ρ (L/A) (2)

Where:

R is the electrical resistance (in ohms, Ω).

ρ is the resistivity of the material (in ohm-meters, Ω·m).

L is the length of the conductor (in meters, m).

A is the cross-sectional area of the conductor (in square meters, m²)

Table 2

Measurement data for the quartz component before and after ion etching.
Group A	FR (MHz)	FL (MHz)	FLD (Eq. 1) (ppm)	RR (Ω)
before
No.1	38.880000	38.766991	-2906.6	9.1
No.2	38.880000	38.777337	-2640.5	10.0
No.3	38.880000	38.760936	-3062.3	9.4
No.4	38.880000	38.745099	-3469.7	8.4
No.5	38.880000	38.747338	-3412.1	10.5
after
No.1	38.880000	38.869569	-268.3	8.7
No.2	38.880000	38.865056	-384.4	8.9
No.3	38.880000	38.869238	-276.8	8.9
No.4	38.880000	38.870015	-256.8	9.1
No.5	38.880000	38.865013	-385.5	8.8
Group B
before
No.6	38.880000	38.743611	-3507.9	9.7
No.7	38.880000	38.723784	-4017.9	13.2
No.8	38.880000	38.779794	-2577.3	8.7
No.9	38.880000	38.791838	-2267.6	9.1
No.10	38.880000	38.757885	-3140.8	9.3
after
No.6	38.880000	38.870232	-251.2	9.6
No.7	38.880000	38.869946	-258.6	9.8
No.8	38.880000	38.868439	-297.4	9.7
No.9	38.880000	38.869881	-260.3	9.8
No.10	38.880000	38.869163	-278.7	9.7
Group C
before
No.11	38.880000	38.767371	-2896.8	10.5
No.12	38.880000	38.771422	-2792.7	9.8
No.13	38.880000	38.794739	-2192.9	14.5
No.14	38.880000	38.772951	-2753.3	9.1
No.15	38.880000	38.783256	-2488.3	12.2
after
No.11	38.880000	38.869775	-263.0	12.2
No.12	38.880000	38.870627	-241.1	21.1
No.13	38.880000	38.869044	-281.8	11.0
No.14	38.880000	38.869690	-265.2	10.9
No.15	38.880000	38.868999	-282.9	10.0

Adjusting the film thickness in the electrode area of the quartz crystal is crucial for achieving accurate crystal frequency. Traditionally, a vacuum deposition method has been commonly used in earlier processes to apply a thin Ag film to the surface of the quartz crystal electrode zone. This practice increases the electrode zone's thickness and decreases the quartz crystal's resonant frequency until the desired crystal frequency is attained. In modern technology, Ag/Ni double-layer films are deposited on the upper and lower surfaces of the quartz crystal electrode area, and the mask is placed on the lower layer. Ion etching is performed to obtain the precise crystal frequency of the quartz element. Due to adjustments in the Ag film thickness occurring only on one side through etching, the thickness of the Ag film on the upper and lower sides becomes uneven. This lack of symmetry can expedite the aging of the quartz component. Consequently, only minimal etching of one side of the Ag film is permissible.

This research highlights the importance of selecting an appropriate mask size during the ion source etching. Employing a vast mask size, representing an oversized etching zone, leads to the etching of electrode legs in the quartz crystal electrode zone, causing a swift rise in the equivalent impedance of the quartz component. Consequently, abnormal changes in crystal frequency (or characteristic curves) occur, contributing to an escalation in product defect rates. In this investigation, mask A (1.30×1.10 mm²) ensured the complete preservation of the electrode legs in the quartz crystal electrode area (i.e., no etching), showing a minimal increase in impedance values. This facilitates the production of products of high quality and excellent stability.

Author Contribution

Code availability: Not applicable.Author contribution: Chih-Fu Yang, Chih-Tsung Chang conceptualization, Yuan-Nan Tsai, Chih-Lung Lin experimentation, Chih-Lung Lin, Chun-Yao Hsu, characterization, Chih-Fu Yang validation, Yuan-Nan Tsai, Chih-Lung Lin analysis, Chih-Fu Yang, Chun-Yao Hsu writing - original draft.

Acknowledgments

The authors gratefully acknowledge the support of the Ministry of Science and Technology of the Republic of China through Grant Nos. NSTC 111-2622-E-262-002.

Disclosure statement

No potential conflict of interest was reported by the authors.

Gobato, R., Gobato, M.R.R., Heidari, A.: Calculation by UFF method of frequencies and vibrational temperatures of the unit cell of the rhodochrosite crystal. Int. J. Adv. Chem. 7, 77–81 (2019)
Tang, Y., Ao, D., Li, W., Zu, X., Li, S., Fu, Y.Q.: NH₃ sensing property and mechanisms of quartz surface acoustic wave sensors deposited with SiO₂, TiO₂, and SiO₂-TiO₂ composite films. Sens. Actuators B. 254, 1165–1173 (2018)
Murozaki, Y., Sakuma, S., Arai, F.: Improvement of the measurement range and temperature characteristics of a load sensor using a quartz crystal resonator with all crystal layer components. Sensors. 17, 1067–1080 (2017)
Rycenga, M., Cobley, C.M., Zeng, J., Li, W., Moran, C.H., Zhang, Q., Qin, D., Xia, Y.: Controlling the synthesis and assembly of silver nanostructures for plasmonic applications. Chem. Rev. 111, 3669–3712 (2011)
Zhou, H., Yang, D., Ivleva, N.P., Mircescu, N.E., Niessner, R., Haisch, C.: SERS detection of bacteria in water by in situ coating with Ag nanoparticles. Anal. Chem. 86, 1525–1533 (2014)
Dayal, G., Chin, X.Y., Soci, C., Singh, R.: High-Q plasmonic fano resonance for multiband surface-enhanced infrared absorption of molecular vibrational sensing. Adv. Opt. Mater. 5, 1600559–1600567 (2017)
Pal, A.K., Mohan, D.B.: Structural, morphological and optical properties of Ag–AgO thin films with the effect of increasing film thickness and annealing temperature. Opt. Mater. 48, 121–132 (2015)
Gaens, W.V., Bogaerts, A.: Kinetic modelling for an atmospheric pressure argon plasma jet in humid air. J. Phys. D. 46, 275201–275254 (2013)
Rao, B.A., Iqbal, M.Y., Sreedhar, B.: Electrochemical and surface analytical studies of the self-assembled monolayer of 5-methoxy-2-(octadecylthio) benzimidazole in corrosion protection of copper. Electrochim. Acta. 55, 620–631 (2010)
Vogt, S., Su, Q., Gutierrez-Sanchez, C., Noll, G.: Critical view on electrochemical impedance spectroscopy using the ferri/ferrocyanide redox couple at gold electrodes. Anal. Chem. 88, 4383–4390 (2016)
Liang, J., Huang, J., Ding, S., Ueda, T.: Flow-injection-based miniaturized quartz crystal microbalance. Sens. Mater. 25, 519–526 (2013)
Zhu, H., Zhang, Y., Ye, L., Dang, Z., Xu, R., Yan, B.: High-efficiency excitation of spoof surface plasmon polaritons through rectangular waveguide using dipole antenna. IEEE Trans. Antennas Propag. 70, 3899–3903 (2022)
Kao, P., Tadigadapa, S.: Micromachined quartz resonator based infrared detector array. Sens. Actuators, A. 149, 189–192 (2009)
Lin, Y.C., Ono, T., Esashi, M.: Fabrication and characterization of micromachined quartz-crystal cantilever for force sensing. J. Micromech. Microeng. 15, 2426–2432 (2005)
Rabe, J., Buttgenbach, S., Zimmermann, B., Hauptmann, P.: Design, manufacturing, and characterization of high-frequency thickness-shear mode resonators, In Proceedings of the IEEE/EIA International Frequency Control Symposium and Exhibition 106–112 IEEE. (Cat. No. 00CH37052) (2000)
Kondo, T., Lee, S., Honda, K., Kawai, T.: Conductive diamond hollow fiber membranes. Electrochem. Commun. 11, 1688–1691 (2009)
Kawata, H., Yasuda, M., Hirai, Y.: Sheath voltage estimation for inductively coupled plasma etcher by impedance analysis. Jpn. J. Appl. Phys. 47, 6914–6916 (2008)

No competing interests reported.

Effects of ion source etching on the impedance performances of quartz component

Status:

Version 1

Abstract

Figures

1. Introduction

2. Related works

3. Experimental procedures

4. Results and discussion

5. Conclusions

Declarations

Author Contribution

Acknowledgments

Disclosure statement

References

Additional Declarations

Status:

Version 1