This paper presents the design, fabrication, and characterization of cantilever-type resonators with a novel stacked structure. Aluminum nitride is adopted as the material for both the structural layer and the piezoelectric layer; this simplifies the fabrication process and improves the quality factor of the resonator. Both in-plane and out-of-plane flexural modes were investigated. The effect of the structural dimensions and electrode patterns on the resonator’s performance were also studied. Finite-element simulations and experiments examining anchor loss and thermoelastic damping, which are the main loss mechanisms affecting the quality factor of these resonators, were carried out. The optimal structural dimensions and electrode patterns of the cantilever-type resonators are presented. A quality factor of 7922 with a motional impedance of 88.52 kΩ and a quality factor of 8851 with a motional impedance of 67.03 kΩ were achieved for the in-plane and out-of-plane flexural-mode resonators, respectively. The proposed resonator design will contribute to the development of high-performance devices such as accelerometers, gyroscopes, and pressure sensors.

HIGHLIGHTS

  • A novel structure for cantilever resonators with AlN structural layers is proposed, which helps to achieve higher performance.

  • The effect of geometric design and electrode patterns on the Q value and the optimization methods were studied.

  • The Q value reached up to 8000 after optimization, which is the highest among similar resonators in previous work.

Piezoelectric micromachined cantilevers have been widely used in various applications, including biosensors,1–3 chemical sensors,1,4,5 mass sensors,6 atomic force microscopes,7 energy harvesters,8,9 and oscillators.10 Recently, aluminum nitride (AlN) has become the most popular thin-film piezoelectric material for vibrating microelectromechanical systems (MEMS) devices because of its high acoustic velocity, low temperature coefficient of elasticity, high thermal conductivity, and good compatibility with complementary metal–oxide–semiconductor (CMOS) devices.11 As a piezoelectric resonator, AlN provides a balance between performance and manufacturability when compared with other thin-film piezoelectric materials such as lead zirconate titanate (PZT), zinc oxide (ZnO), lithium niobate (LiNbO3), and lithium tantalate (LiTaO3). Additionally, AlN has a high melting point of 2470 °C and maintains most of its mechanical strength up to very high temperatures,12 which enables AlN-based MEMS devices to withstand harsh environments.

Four main types of piezoelectric micromachined cantilever structure have been reported to date, and these are shown in Fig. 1. The most common structure is a silicon-based piezoelectric cantilever, which shows a relatively high quality factor (Q).13 Considering the conductivity of silicon, for this kind of resonator, either a silicon structural layer is directly used as the bottom electrode,14 as shown in Fig. 1(a), or an insulating layer needs to be added between the bottom metal electrode and the silicon structural layer,15 as shown in Fig. 1(b). However, increasing the number of material interfaces in the cantilever structure tends to increase the thermoelastic loss and interface dissipation,16 which degrades Q. The fabrication of the two types of resonator shown in Figs. 1(a) and 1(b) is usually based on silicon-on-insulator (SOI) or cavity-SOI technology, which leads to high costs. In addition, the processing flows of Si-based resonators are usually complex. A thin-film piezoelectric resonator,17–19 which is simply composed of a single piezoelectric layer sandwiched between two electrodes, as shown in Fig. 1(c), is another kind of cantilever structure that has a relatively low Q value. Resonators with this structure can excite only the in-plane flexural mode, rather the out-of-plane flexural mode, which limits its application scenarios.

FIG. 1.

Schematic diagrams of stacks for four types of piezoelectric micromachined cantilever.

FIG. 1.

Schematic diagrams of stacks for four types of piezoelectric micromachined cantilever.

Close modal

Figure 1(d) presents a unimorph structure, in which AlN is used as both the passive layer and the piezoelectric layer.20,21 Compared with silicon-based piezoelectric resonators, this structure can be fabricated using a relatively simple surface micromachining technique, which is helpful for realizing batch fabrication of MEMS devices and CMOS circuits. The use of AlN as a structural layer can achieve a Q value comparable to that of silicon and can contribute to realizing wafer-level MEMS–CMOS integration by direct fabrication of the resonators on the electronic-circuit wafer rather than using a hybrid multi-chip process. Fabian et al. completed the fabrication of triple-beam tuning-fork accelerometers using such a structure, and they completed the integration of a system-on-chip design.20 Yuan et al. designed and fabricated a tuning fork with such an AlN-based structure.21 However, there has been little research on this kind of cantilever to establish its potential for high performance.

In this work, AlN-based cantilever-type resonators vibrating in the in-plane and out-of-plane flexural modes were designed, fabricated, and characterized. The effects of anchor loss, thermoelastic damping (TED), and air damping on the performance of these resonators were studied. Devices with different structural dimensions and electrode patterns were investigated, and a set of design guidelines is proposed to achieve AlN-based cantilever resonators with good performance. These devices are likely to find applications in resonant sensors.

An AlN-based unimorph microcantilever, in which AlN is used as the material for both the passive layer and the piezoelectric layer, can stimulate both in-plane and out-of-plane flexural modes with different electrode patterns, as shown in Fig. 2. To excite the out-of-plane flexural mode, voltages with different polarities are applied to the top and bottom electrodes. For in-plane flexural modes, the top electrode is patterned into two separate stripes that are connected to either the signal port or the ground port, while the bottom electrode is floating, as shown in Fig. 2(a). The key dimensions of the basic device are listed in Table I, in which TS, TP, and TE are the thicknesses of the structural layer, piezoelectric layer, and electrodes of the resonator, respectively.

FIG. 2.

Top view and cross section of (a) in-plane and (c) out-of-plane flexural-mode cantilever-type resonators, and corresponding simulated displacements of (b) in-plane and (d) out-of-plane flexural modes.

FIG. 2.

Top view and cross section of (a) in-plane and (c) out-of-plane flexural-mode cantilever-type resonators, and corresponding simulated displacements of (b) in-plane and (d) out-of-plane flexural modes.

Close modal
TABLE I.

Key dimensions of the basic devices.

ParameterIn plane (μm)Out of plane (μm)
TS 2.50 
TP 1.00 
TE 0.15 
WC 50 50 
LC 250 300 
LE 248 298 
WE (out of plane) ⋯ 44 
WEL, WER (in plane) 20, 20 ⋯ 
WG (in plane) ⋯ 
ParameterIn plane (μm)Out of plane (μm)
TS 2.50 
TP 1.00 
TE 0.15 
WC 50 50 
LC 250 300 
LE 248 298 
WE (out of plane) ⋯ 44 
WEL, WER (in plane) 20, 20 ⋯ 
WG (in plane) ⋯ 
The frequency of a cantilever resonator vibrating in a flexural mode is given by
ω=E12ρaL2βn,
(1)
where ω is the angular frequency, E is the Young’s modulus, ρ is the density, L is the length of the cantilever, a is the width of the cantilever in the vibrating direction, and βn is the nth mode coefficient, which is 1.875 for the first mode. Since the width of the cantilever is much larger than its thickness in this work, the resonant frequencies of the in-plane flexural modes are much higher than those of the out-of-plane flexural modes.

To investigate the relationship between the electrode design and performance of the in-plane and out-of-plane modes, we designed AlN-based microcantilevers with different electrode patterns. For in-plane flexural-mode resonators, as listed in Table II, in each set, one of either the electrode length, electrode gap width, or electrode ratio was set as variable, while the other dimensions were kept consistent with the counterparts shown in the in-plane column of Table I. For the out-of-plane flexural-mode resonators, either the electrode length or electrode width was set as variable while the other dimension was kept consistent with the counterparts shown in the out-of-plane column of Table III. The resonant frequencies of the in-plane flexural-mode cantilever resonator range from 600 kHz to 1.3 MHz, while those of the out-of-plane flexural-mode cantilever resonator range from 50 to 500 kHz, as calculated using Eq. (1).

TABLE II.

Electrode design variables for the in-plane flexural-mode resonators.

LEWGWEL, WER
50 4.4, 35.6 
100 10 8, 32 
150 15 13.3, 26.6 
200 20 20, 20 
248 25 26.6, 13.3 
 30 32, 8 
  35.6, 4.4 
LEWGWEL, WER
50 4.4, 35.6 
100 10 8, 32 
150 15 13.3, 26.6 
200 20 20, 20 
248 25 26.6, 13.3 
 30 32, 8 
  35.6, 4.4 
TABLE III.

Electrode design variables of the out-of-plane flexural-mode resonators.

WELE
14 50 
24 100 
34 150 
44 200 
 250 
 298 
WELE
14 50 
24 100 
34 150 
44 200 
 250 
 298 
The quality factor Q is one of the most important parameters of a resonator, and it dominates the sensitivity and signal-to-noise ratios of MEMS sensors. The total Q is determined by several energy-dissipation mechanisms, and it is defined as
1Q=1QTED+1Qanc+1Qair+1Qother,
(2)
where QTED, Qanc, Qair, and Qother are the quality factors related to TED, anchor loss, air damping, and other loss mechanisms including material loss, dielectric loss, etc., which can be neglected, respectively.

It has been reported that TED, anchor loss and air damping are the dominant energy-dissipation mechanisms for piezoelectric resonators.22 Air damping is determined by the air pressure, so it can be neglected in vacuum. Therefore, in this work, only TED and anchor loss were simulated and analyzed using a three-dimensional finite-element method (FEM). The physical fields of heat transfer in solids and structural mechanics were coupled to calculate QTED. A perfectly matched layer was adopted to absorb the leaked mechanical waves to derive Qanc.23 The total Q value of the resonator was then calculated using Eq. (2) with the simulated QTED and Qanc values.

Resonators with a 2.5-μm-thick AlN structural layer and a 1-μm-thick piezoelectric layer were simulated to investigate the effects of cantilever length and width on Qanc, QTED, and Q. Figure 3(a) shows the results for the in-plane flexural-mode resonators. With a fixed length, the greater the width, the higher the Qanc value. Additionally, Qanc increases as the length increases. The situation is more complex for QTED. When the width of the cantilever is small, such as 10 μm, QTED increases as the length increases. However, when the width is larger, such as 30 or 50 μm, QTED decreases as the length increases. Thus, for a cantilever with a 10-μm width, the total Q value increases as the length of the cantilever increases, while for cantilevers with 30- and 50-μm widths, the total Q value first increases, reaches a peak at a certain length, and then decreases. In the simulation results for the out-of-plane flexural mode, as illustrated in Fig. 3(b), as the length increases, Qanc, QTED, and the total Q value are all increased. A wider cantilever contributes to a higher Qanc value, but QTED is almost unaffected by the width of the cantilever. Thus, for a fixed length, the smaller the width, the higher the total Q value.

FIG. 3.

Effects of the cantilever length and width on Qanc, QTED, and Q for a 2.5-μm-thick structural layer and a 1-μm-thick piezoelectric layer. The unit of WC is μm. Results for (a) the in-plane flexural-mode resonator and (b) the out-of-plane flexural-mode resonator.

FIG. 3.

Effects of the cantilever length and width on Qanc, QTED, and Q for a 2.5-μm-thick structural layer and a 1-μm-thick piezoelectric layer. The unit of WC is μm. Results for (a) the in-plane flexural-mode resonator and (b) the out-of-plane flexural-mode resonator.

Close modal

Cantilever-type resonators with three different stacked structures were also simulated to investigate the effect of the structure on performance: a Si-based cantilever consisting of a 1-μm-thick AlN piezoelectric layer and a 2.5-μm-thick Si structural layer; a resonator referred to as AlN-based01, consisting of a 1-μm-thick AlN piezoelectric layer and a 2.5-μm-thick AlN structural layer; and a resonator referred to as AlN-based02, consisting of a 0.5-μm-thick AlN piezoelectric layer and a 3-μm-thick AlN structural layer, which has the same total thickness as AlN-based01. The simulation results are illustrated in Fig. 4. As can be seen, the Si-based resonator shows better performance in the in-plane flexural mode, while the AlN-based resonators have better performance in the out-of-plane flexural mode. In addition, there is no difference in the performance between AlN-based01 and AlN-based02.

FIG. 4.

Effects of the cantilever length and width on Qanc, QTED, and Q for different structural layers. Results for (a) the in-plane flexural-mode resonator and (b) the out-of-plane flexural-mode resonator.

FIG. 4.

Effects of the cantilever length and width on Qanc, QTED, and Q for different structural layers. Results for (a) the in-plane flexural-mode resonator and (b) the out-of-plane flexural-mode resonator.

Close modal

The stacked structures of the devices with the above two vibration modes are the same, so they can be fabricated on the same wafer simultaneously. The process flow is illustrated in Fig. 5. As shown in Fig. 5(a), the process begins with the etching of an air cavity directly on a silicon wafer by reactive ion etching. Then, the cavity is filled with a sacrificial layer (phosphosilicate glass, PSG), which is planarized by chemical mechanical polishing. After that, 2.5 μm AlN is deposited as a structural layer, and 150 nm molybdenum is then deposited and patterned as the bottom electrode [Fig. 5(b)]. The electrode is patterned into two parts, one is used as the bottom electrode of the resonator, and the other is used as the pad for testing. Next, 1 μm AlN is deposited as a piezoelectric layer, and 150 nm molybdenum is deposited [Fig. 5(c)]. Then, the top molybdenum layer is patterned as the top electrode [Fig. 5(d)]. The next step is to remove the AlN above the testing pad [Fig. 5(e)]. After that, 3.5 μm AlN is etched to form the cantilever [Fig. 5(f)]. Then, 0.3 μm gold is deposited and patterned with a lift-off process to connect the top electrodes and the pads [Fig. 5(g)]. Finally, the sacrificial layer is etched using diluted hydrofluoric acid solution, and the cantilever is left suspended [Fig. 5(h)].

FIG. 5.

Stages of the fabrication process.

FIG. 5.

Stages of the fabrication process.

Close modal

Optical microscope and scanning electron microscope (SEM) images of the fabricated in-plane flexural-mode cantilever and out-of-plane flexural-mode cantilever are illustrated in Fig. 6. To prevent the free ends of the arms from touching the bottom of the cavity, the stress in the AlN film is regulated during processing so that there is an internal compressive stress in the upper part of the film, which results in an upward curling, as shown in Figs. 6(c) and 6(d). The surface topology of the resonators was characterized using an Olympus LEXT OLS4500 laser confocal microscope, and the results are shown in Fig. 7. The maximum height differences of a 250-μm-length cantilever resonator and a 300-μm-length cantilever resonator were found to be ∼9 and 12 μm, respectively.

FIG. 6.

Optical microscope images of (a) in-plane and (b) out-of-plane flexural-mode cantilever resonators, and corresponding SEM images of (c) in-plane and (d) out-of-plane flexural-mode cantilever resonators.

FIG. 6.

Optical microscope images of (a) in-plane and (b) out-of-plane flexural-mode cantilever resonators, and corresponding SEM images of (c) in-plane and (d) out-of-plane flexural-mode cantilever resonators.

Close modal
FIG. 7.

Laser confocal microscope characterization results of cantilever resonators: (a) in-plane flexural-mode resonator; (b) out-of-plane flexural-mode resonator.

FIG. 7.

Laser confocal microscope characterization results of cantilever resonators: (a) in-plane flexural-mode resonator; (b) out-of-plane flexural-mode resonator.

Close modal

The electrical performance of the resonators was measured in a vacuum probe station using an impedance analyzer (E4990A). The vacuum probe station was equipped with a turbomolecular pump and a rotary vane pump. This allowed us to adjust the pressure from atmosphere down to high vacuum.

Typical electrical responses for the basic devices in vacuum and the corresponding curves fitted using the modified Butterworth–Van-Dyke (MBVD) model24 are displayed in Fig. 8. The in-plane flexural-mode resonator had a Q of 4615 at its series resonant frequency of 1089.7 kHz. The out-of-plane flexural-mode resonator presented better performance, and the Q at its series resonant frequency (Qs) was 7979. In addition to Q, the effective electromechanical coupling coefficient (kteff2) and Rs are also important parameters for evaluating the performance of the resonator, as they have a significant impact on the bandwidth and power consumption of MEMS systems. The Rs values of the in-plane and out-of-plane flexural-mode resonators were found to be 85.4 and 138.4 kΩ, respectively. The corresponding kteff2 values were 0.11% and 0.29%, respectively. A further study was carried out to reveal the influence of the geometric parameters on the performance of these different resonators.

FIG. 8.

Electrical responses in vacuum and fitted curves of the (a) in-plane and (b) out-of-plane flexural-mode resonator. (c) The MBVD model and extracted parameters.

FIG. 8.

Electrical responses in vacuum and fitted curves of the (a) in-plane and (b) out-of-plane flexural-mode resonator. (c) The MBVD model and extracted parameters.

Close modal

A set of experimental results for in-plane flexural-mode resonators with different electrode patterns according to Table II are shown in Fig. 9. The figure shows the relationships between Qs, fs, Rs, and kteff2 and the top electrode pattern. In Fig. 9(a), the Qs value reaches a peak of 7580 when the value of WG/WE is 1, that is, WG = WEL = WER; in this case, the highest kteff2 and the lowest Rs can also be obtained. As the WG value increases further, the performance of the resonator starts to degrade. In Fig. 9(b), the fluctuation of Qs is limited as LE/LC is greater than 0.4, whereas kteff2 first increases and then decreases, while Rs first decreases and then increases. By jointly considering these three performance-evaluation parameters, the optimum LE/LC for excellent performance is 0.6. In Fig. 9(c), the curves of the performance-evaluation parameters were verified to be symmetric. A ratio of 4:1 between the left and right electrodes leads to the highest Qs value. For WER/WEL values between 1/4 and 4, the kteff2 value of the resonator first increases significantly and then decreases, whereas Rs first decreases and then increases. Considering the symmetry of the structure, a better overall performance can be obtained at WER/WEL = 1.

FIG. 9.

Influence of different (a) WG, (b) LE, and (c) WEL/WER values on the performance of the in-plane flexural-mode cantilever resonator.

FIG. 9.

Influence of different (a) WG, (b) LE, and (c) WEL/WER values on the performance of the in-plane flexural-mode cantilever resonator.

Close modal

In contrast, changes in the electrode gap and the width ratio of the left and right electrodes basically have no effect on the resonant frequency, as presented in Figs. 9(a) and 9(c). However, due to the mass loading effect, the resonant frequency will decrease with the increasing electrode coverage in the length direction, as shown in Fig. 9(b).

The effect of varying the cantilever length on the performance of resonators with different widths is shown in Fig. 10. The resonant frequency is inversely related to LC and positively related to WC, which is consistent with Eq. (1). The Qs values of the resonators with different widths both reach a peak at a certain length and then decrease as the length of the resonator is increased further. The trends shown in the experimental results are consistent with the FEM simulation results. For a constant resonant frequency, a wider cantilever has a higher Qs value.

FIG. 10.

Influence of the width and length on the performance of the in-plane flexural-mode cantilever resonator.

FIG. 10.

Influence of the width and length on the performance of the in-plane flexural-mode cantilever resonator.

Close modal

The performance of out-of-plane flexural-mode resonators with different design dimensions, including the length of the cantilever and the width and length of the electrode, was also studied. Figure 11 shows how the electrode pattern affects the performance of the resonator. In Fig. 11(a), Qs first decreases and then increases, while Rs decreases and kteff2 increases with increasing WE. As shown in Fig. 11(b), Qs fluctuates only slightly when LE/LC is greater than 1/3, whereas kteff2 first increases and then decreases, while Rs decreases and tends to saturate. In the LE/LC range 0.50–0.83, both kteff2 and Rs can reach better values simultaneously.

FIG. 11.

Influence of different (a) WE and (b) LE values on the performance of the out-of-plane flexural-mode cantilever resonator.

FIG. 11.

Influence of different (a) WE and (b) LE values on the performance of the out-of-plane flexural-mode cantilever resonator.

Close modal

The effect of the length of the cantilever on the performance of out-of-plane flexural-mode resonators of different widths was also studied. The results are shown in Fig. 12. The resonant frequencies of the cantilever resonators vibrating in the out-of-plane flexural mode only depend on the length. Therefore, for a given length, the resonant frequencies are almost the same for both widths of resonator. The trends of Qs shown in Fig. 12 are similar to those in Fig. 10, but they are different from the simulation results shown in Fig. 3(b). We attribute this phenomenon to the residual stress of the AlN. As noted earlier, the residual stress in the AlN films causes the cantilever to warp upward, which influences the strain distribution of the resonator and the trend of QTED. For the in-plane flexural mode, the warping direction is perpendicular to the direction of vibration, which has little effect on QTED, so the trend is consistent with the FEM simulation results shown in Fig. 3(a). The best length-to-thickness ratio in terms of resonator performance is in the range 57 to 86. As for the in-plane flexural mode, for a constant resonant frequency, a wider cantilever has a higher Qs value.

FIG. 12.

Influence of the width and length of the out-of-plane flexural-mode cantilever resonator on its performance.

FIG. 12.

Influence of the width and length of the out-of-plane flexural-mode cantilever resonator on its performance.

Close modal

Another difference between the in-plane and out-of-plane flexural vibration modes is their sensitivity to the level of vacuum. Figure 13 shows the performance of in-plane and out-of-plane flexural-mode resonators with different lengths; the widths of the cantilevers were all 50 μm. It can be seen that the out-of-plane flexural-mode cantilever resonator is more sensitive to pressure than the in-plane flexural-mode cantilever resonator. For the in-plane flexural mode, the Q value can stabilize at a relatively high level under pressures below 100 Pa, which can be easily achieved by wafer-level vacuum packaging. For the out-of-plane flexural mode, the pressure needs to be less than 1 Pa to keep the Q value high enough. Therefore, the out-of-plane device is more suitable for use as an indicator of the vacuum level.

FIG. 13.

Dependence of quality factor on pressure for resonators with different designs.

FIG. 13.

Dependence of quality factor on pressure for resonators with different designs.

Close modal

Table IV presents a comparison of the performance of cantilever resonators reported in previous studies. The data in the table show that the quality factor of a cantilever resonator using PZT as the piezoelectric material is generally limited due to the high acoustic loss of the PZT material; using AlN as the piezoelectric material improves the quality factor. Compared with Si-based cantilever resonators, the AlN-based cantilever resonators presented in this paper achieve higher quality factors for both the in-plane and out-of-plane flexural modes.

TABLE IV.

Comparison of cantilever-type resonators tested in vacuum presented in previous articles.

ArticleStructurePiezoelectric materialVibration modeResonant frequency (kHz)Quality factor
Reference 25  Silicon-based PZT Out of plane 202 1400 
Reference 15  Silicon-based PZT Out of plane ⋯ 1887 
Reference 26  Silicon-based AlN Out of plane and in plane Out of plane: 7 Out of plane: ∼7500 
In plane: 96 In plane: ∼1700 
Reference 27  Silicon-based AlN Out of plane 48.5 4328 
This work AlN-based AlN Out of plane and in-plane Out of plane: 58.16 Out of plane: 8851 
In plane: 522.5 In plane: 7922 
ArticleStructurePiezoelectric materialVibration modeResonant frequency (kHz)Quality factor
Reference 25  Silicon-based PZT Out of plane 202 1400 
Reference 15  Silicon-based PZT Out of plane ⋯ 1887 
Reference 26  Silicon-based AlN Out of plane and in plane Out of plane: 7 Out of plane: ∼7500 
In plane: 96 In plane: ∼1700 
Reference 27  Silicon-based AlN Out of plane 48.5 4328 
This work AlN-based AlN Out of plane and in-plane Out of plane: 58.16 Out of plane: 8851 
In plane: 522.5 In plane: 7922 

A new type of cantilever resonator was designed and manufactured with a new stacked structure. This has the advantages of high Q values, low processing difficulty, and low cost. The proposed cantilever resonator design can be used in gyroscopes, accelerometers, and pressure sensors. For the in-plane flexural mode, the resonator has the best performance when the values of WG/WE, LE/LC, and WEL/WER are 1.0, 0.6, and 1.0, respectively. The highest Q achieved was 7922 with a resonance frequency of 522.5 kHz. For the out-of-plane flexural mode, better performance can be achieved when the values of WE/WC and LE/LC are 1.00 and 0.50–0.83, respectively. The highest Q achieved was 8851 with a resonance frequency of 58.16 kHz. For both the in-plane and out-of-plane flexural modes, a wider cantilever can achieve a higher Q value. The best length-to-thickness ratio for the out-of-plane flexural-mode resonator is in the range 57 to 86. These results demonstrate the importance of geometrical optimization for achieving a high quality factor in micromachined AlN cantilevers in vacuum. Further research can investigate the practical implications of these findings and explore the potential of AlN-based resonators.

This work was supported in part by the National Key Research and Development Program of China (Grant No. 2020YFB2008800) and in part by the Nanchang Institute for Microtechnology of Tianjin University. The authors would like to thank Quanning Li, Xuejiao Chen, Bohua Liu, and Chongling Sun, who are engineers of the M/NEMS laboratory in Tianjin University, for their help in the fabrication of our devices.

The authors have no conflicts to disclose.

The data that support the findings of this study are available within the article.

1.
Lavrik
NV
,
Sepaniak
MJ
,
Datskos
PG
.
Cantilever transducers as a platform for chemical and biological sensors
.
Rev Sci Instrum
2004
;
75
(
7
):
2229
2253
.
2.
Singh
AD
,
Patrikar
RM
.
Design and fabrication of PDMS-based electrostatically actuated MEMS cantilever beam
.
Micro Nano Lett
2020
;
15
(
5
):
302
307
.
3.
Raiteri
R
,
Grattarola
M
,
Butt
H-J
,
Skládal
P
.
Micromechanical cantilever-based biosensors
.
Sens Actuators B
2001
;
79
(
2-3
):
115
126
.
4.
Lochon
F
,
Dufour
I
,
Rebière
D
.
An alternative solution to improve sensitivity of resonant microcantilever chemical sensors: Comparison between using high-order modes and reducing dimensions
.
Sens Actuators B
2005
;
108
(
1-2
):
979
985
.
5.
Lange
D
,
Hagleitner
C
,
Hierlemann
A
, et al
Complementary metal oxide semiconductor cantilever arrays on a single chip: Mass-sensitive detection of volatile organic compounds
.
Anal Chem
2002
;
74
(
13
):
3084
3095
.
6.
Ivaldi
P
,
Abergel
J
,
Matheny
MH
, et al
50 nm thick AlN film-based piezoelectric cantilevers for gravimetric detection
.
J Micromech Microeng
2011
;
21
(
8
):
085023
.
7.
Binnig
G
,
Quate
CF
,
Gerber
C
.
Atomic force microscope
.
Phys Rev Lett
1986
;
56
(
9
):
930
933
.
8.
Rua Taborda
MI
,
Elissalde
C
,
Chung
UC
, et al
Key features in the development of unimorph stainless steel cantilever with screen-printed PZT dedicated to energy harvesting applications
.
Int J Appl Ceram Technol
2020
;
17
(
6
):
2533
2544
.
9.
Shen
D
,
Park
J-H
,
Noh
JH
, et al
Micromachined PZT cantilever based on SOI structure for low frequency vibration energy harvesting
.
Sens Actuators A
2009
;
154
(
1
):
103
108
.
10.
Lee
SS
,
White
RM
.
Self-excited piezoelectric cantilever oscillators
.
Sens Actuators A
1996
;
52
(
1-3
):
41
45
.
11.
Shelton
S
,
Chan
ML
,
Park
H
, et al
CMOS-compatible AlN piezoelectric micromachined ultrasonic transducers
.
2009 IEEE International Ultrasonics Symposium
.
IEEE
.
2009
. pp.
402
405
.
12.
Kim
T
,
Kim
J
,
Dalmau
R
, et al
High-temperature electromechanical characterization of AlN single crystals
.
IEEE Trans Ultrason Ferroelectrics Freq Control
2015
;
62
(
10
):
1880
1887
.
13.
Lu
J
,
Ikehara
T
,
Zhang
Y
, et al
High quality factor silicon cantilever driven by piezoelectric thin film actuator for resonant based mass detection
.
Microsyst Technol
2009
;
15
(
8
):
1163
1169
.
14.
Isarakorn
D
,
Sambri
A
,
Janphuang
P
, et al
Epitaxial piezoelectric MEMS on silicon
.
J Micromech Microeng
2010
;
20
(
5
):
055008
.
15.
Lu
J
,
Ikehara
T
,
Zhang
Y
, et al
High quality factor silicon cantilever transduced by piezoelectric lead zirconate titanate film for mass sensing applications
.
Jpn J Appl Phys
2007
;
46
(
12
):
7643
7647
.
16.
Frangi
A
,
Cremonesi
M
,
Jaakkola
A
,
T.
Pensala
.
Analysis of anchor and interface losses in piezoelectric MEMS resonators
.
Sens Actuators A
2013
;
190
:
127
135
.
17.
Isobe
A
,
Asai
K
,
Matsumoto
H
, et al
4F-2 AlN tuning fork type resonators used with MEMS technologies
.
2006 IEEE Ultrasonics Symposium
.
IEEE
.
2006
. pp.
560
563
.
18.
Abdul
B
,
Mastronardi
VM
,
Qualtieri
A
, et al
Design, fabrication and characterization of piezoelectric cantilever MEMS for underwater application
.
Micro Nano Eng
2020
;
7
:
100050
.
19.
Vigevani
G
,
Goericke
FT
,
Izyumin
II
, et al
Electrode design and coupling optimization of Aluminum Nitride DETF
.
2011 IEEE International Ultrasonics Symposium
.
IEEE
.
2011
. pp.
1731
1734
.
20.
Goericke
F
,
Mansukhani
K
,
Yamamoto
K
, et al
Experimentally validated aluminum nitride based pressure, temperature and 3-axis acceleration sensors integrated on a single chip
.
2014 IEEE 27th International Conference on Micro Electro Mechanical Systems (MEMS)
.
IEEE
.
2014
. pp.
729
732
.
21.
Yuan
Y
,
Yang
Q
,
Li
H
, et al
High quality factor and low motional impedance aluminum nitride tuning fork resonators vibrating in in-plane flexural mode
.
IEEE Electron Device Lett
2022
;
43
(
10
):
1744
1747
.
22.
Chen
W
,
Jia
W
,
Xiao
Y
,
G.
Wu
.
Design, modeling and characterization of high-performance bulk-mode piezoelectric MEMS resonators
.
J Microelectromech Syst
2022
;
31
(
3
):
318
327
.
23.
Qiu
H
,
Ababneh
A
,
Feili
D
, et al
Analysis of intrinsic damping in vibrating piezoelectric microcantilevers
.
Microsyst Technol
2016
;
22
(
8
):
2017
2025
.
24.
Larson
JD
,
Bradley
PD
,
Wartenberg
S
, et al
Modified Butterworth-Van Dyke circuit for FBAR resonators and automated measurement system
.
2000 IEEE Ultrasonics Symposium. Proceedings of the An International Symposium (Cat. No.00CH37121)
.
IEEE
.
2000
. pp.
863
868
. vol.
1
.
25.
Liu
Z
,
Chen
J
,
Zou
X
.
Modeling the piezoelectric cantilever resonator with different width layers
.
Sensors
2020
;
21
(
1
);
87
.
26.
Qiu
H-C
,
Feili
D
,
Wu
X-Z
,
H.
Seidel
.
Resonance-mode effect on piezoelectric microcantilever performance in air, with a focus on the torsional modes
.
Chin Phys B
2014
;
23
(
2
):
027701
.
27.
Fischeneder
M
,
Oposich
M
,
Schneider
M
,
U.
Schmid
.
Tuneable Q-factor of MEMS cantilevers with integrated piezoelectric thin films
.
Sensors
2018
;
18
(
11
):
3842
.

Shuai Shi received a B.S. degree from China University of Mining and Technology, China, in 2020. He is currently a student at the State Key Laboratory of Precision Measuring Technology and Instruments. His research field is the design and fabrication of MEMS resonators.

Qingrui Yang received B.S. and Ph.D. degrees in instrument science and technology from Tianjin University, China, in 2011 and 2018, respectively. She is currently an assistant professor at the State Key Laboratory of Precision Measurement Technology and Instruments, Tianjin University. Her research interests focus on the design, fabrication, and application of piezoelectric MEMS resonators, filters, and sensors.

Yi Yuan received a B.S. degree in instrument science and technology from Tianjin University, China, in 2018. He is currently pursuing a Ph.D. with the School of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China. His current research interests include the design and fabrication of MEMS piezoelectric resonators.

Haolin Li received a B.S. degree in mechanical design, manufacturing, and automation from China University of Petroleum in 2019. He is currently pursuing a Ph.D. in instrument science and technology at Tianjin University. His current research interests include MEMS resonators and sensors.

Pengfei Niu is currently an Associate Professor with the State Key Laboratory of Precision Measuring Technology and Instruments, Tianjin University. He is working on acoustic microelectromechanical systems (MEMS) platforms, wearable electrochemical sensing devices, and their coupling for health diagnosis applications.

Wenlan Guo received her M.S. in food science from Tianjin University in 2013. She is currently an engineer at the School of Precision Instruments and Opto-Electronics Engineering, Tianjin University. Her research interests include physical vapor deposition using various materials, and semiconductor package processing.

Chen Sun received his B.S. degree in optical information science and technology from Nanjing University of Science and Technology in 2006, and his M.S. degree in optics from Tianjin University in 2015. He is currently an engineer at the School of Precision Instruments and Opto-Electronics Engineering, Tianjin University. His research interests include wet processing of silicon-based materials and structures, photo mechanics, and digital image processing.

Wei Pang received a B.S. degree from Tsinghua University, China, in 2001 and M.S. and Ph.D. degrees in electrical engineering from University of Southern California, USA, in 2005 and 2006, respectively. He has been a full professor with Tianjin University (TJU), China, since 2009. At TJU, he set up the TJU MEMS Laboratory, a 15 000-square-foot facility completely dedicated to MEMS devices. This facility has a clean room (∼7000 sq. ft), a CAD lab, and a measurement/test/metrology lab. Over the past 12 years in TJU, he has launched major research efforts in a variety of areas, including resonators/filters/oscillators, acoustofluidics, ultrasound imaging and TOF sensors, pressure sensors, gas and particle sensors, uncooled infrared sensors, sweat sensing, droplet ejectors, microphones, and microspeakers. He has authored or coauthored more than 110 SCI journal papers, 70 international conference papers, and has 350 granted and pending patents in the MEMS field. He pioneered the commercialization of piezoelectric MEMS in China, developed over 400 BAW filter products (in frequency ranges between 600 MHz and 10 GHz, covering fractional bandwidth ranges between 0.01% and >20%), and has shipped hundreds of millions of aluminum nitride MEMS devices to the world’s respected OEMs.