Open Access
Add to BibSonomy
Abstract
Confocal microscopes and two-photon microscopes are powerful tools for early cancer diagnosis because of their high-resolution 3D imaging capability, but applying them for clinical use in internal organs is hindered by the lack of axially tunable lens modules with small size, high image quality and large tuning range. This paper reports a compact MEMS lens scanner that has the potential to overcome this limitation. The MEMS lens scanner consists of a MEMS microstage and a microlens. The MEMS microstage is based on a unique serpentine inverted-series-connected (ISC) electrothermal bimorph actuator design. The microlens is an aspheric glass lens to ensure optical quality. The MEMS microstage has been fabricated and the lens scanner has been successfully assembled. The entire lens scanner is circular with an outer diameter of 4.4 mm and a clear optical aperture of 1.8 mm. Experiments show that the tunable range reaches over 200 µm at only 10.5 V and the stiffness of the microstage is 6.2 N/m. Depth scan imaging by the MEMS lens scanner has also been demonstrated with a 2.2 µm resolution, only limited by the available resolution target.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Laser scanning confocal microscopy (LSCM) and two-photon microscopy (TPM) are two of the powerful optical biopsy techniques that have been widely used in biomedical research including endoscopic early cancer diagnosis and head-mounted in vivo brain imaging thanks to their radiation-free, high-resolution, cross-sectional imaging capabilities [1–5]. Axially tunable lenses are among them since both LSCM and TPM rely on moving a sharply focused laser spot axially to acquire image slices in different depths. Endoscopy and head-mounting place stringent size and weight constraints on the optomechanical components that can be used to build LSCM and TPM.
A typical laser scanning engine for a miniature confocal or two-photon microscopic probe is illustrated in Fig. 1, which includes a transverse x/y scanning mirror and an axial z scanning lens. Galvo mirrors are usually used to realize the x/y scan, which are being replaced by MEMS mirrors for miniaturization [5–7]. For axial scanning, motorized stages are typically used in bench-top LSCM and TPM systems [8,9], which are not applicable in either endoscopy or head-mounted monitoring. Various methods have been proposed to miniaturize axial lens scanners. One of them is to control the divergence of the laser beam so that the focal length is tunable [10]. Devices such as deformed mirrors, spatial light modulators, variable-focus lenses and acoustic optical deflectors can be used for this purpose [10–13], but varying the focal length means varying the numerical aperture (NA). Thus, the imaging resolution will be affected significantly.
Another method is to move the objective lens directly by a microstage. Unlike the beam divergence control, directly moving the lens maintains a constant NA and thus a constant axial resolution, which is desired. Both the lateral and axial resolutions of LSCM and TPM are highly relative to the inverse of NA [14,15]. Therefore, a high NA of at least 0.5 is needed. The working distance is also important, which should be about 1 mm. This means the clear lens aperture should be about 1 mm. MEMS technology can be employed to fabricate microstages for making miniature tunable lens scanners. MEMS microstages have been demonstrated with electrostatic, magnetic, and piezoelectric actuation. For instance, Kwon et al. reported an electrostatic vertical comb drive microlens scanner which achieved a 55 µm range on resonance [16], where the loaded lens was a 300-µm-diameter polymer lens. Bargiel et al. presented another electrostatic microlens scanner based on parallel-plate actuation [17], which generated a 38 µm displacement on resonance at 160 V with a bonded 300-µm-diameter glass ball microlens. The travel ranges of these electrostatic lens scanner are usually lower than 100 µm. At the same time, the microlenses typically have smaller clear apertures as the stiffnesses of these electrostatic actuators are relatively small. Siu et al. reported a magnetically actuated MEMS lens scanner that achieved a larger scan range of 125 µm with a larger lens diameter of 1.6 mm [18], but the lens was made of PDMS and a varying external magnetic field was needed to perform axial scanning. In addition, Michael and Kwok reported a piezoelectrically driven lens scanner with a 600-µm-diameter glass ball lens [19], but a maximum displacement of only 22 µm was achieved. Thus, these reported MEMS lens scanners all encounter one or more issues of small travel range, high driving voltage, low stiffness, low lens quality, or small clear aperture.
On the other hand, electrothermal actuators are known for achieving large displacement at low voltage and small area [20–24]. For example, Wu and Xie demonstrated a 880-µm-tunable-range microlens scanner (clear aperture: 0.5 mm) based on a lateral-shift-free electrothermal bimorph actuator [25], where a glass rod lens with a diameter of 1 mm was loaded and the resultant resonance frequency was only 79 Hz. Later, Liu et al. developed another electrothermal MEMS lens scanner (clear aperture: 2 mm) loaded with a plano convex glass lens with a larger diameter of 2.4 mm and achieved a maximum travel range of 400 µm [26], but the resonance frequency of the scanner was only about 24 Hz due to its low stiffness (0.23N/m), which can be applied only in well-controlled conditions.
In this work, we propose a microlens scanner based on a unique design of serpentine inverted-series-connected (SISC) electrothermal bimorph actuator. This new lens scanner has high stiffness and large clear aperture so that a high-quality microlens can be loaded. This lens scanner can simultaneously achieve large displacement, high fill factor, high optical quality, and low driving voltage. A similar lens scanner which achieved 120 µm with a mounted glass lens was first reported in [27], but its design including layer thicknesses and length ratio between bimorphs was not optimized and the design methodology was not given. This paper presents the design analysis, fabrication, and characterization of the MEMS lens scanner in detail as well as the depth tuning experimental results.
2. MEMS lens scanner design
The proposed MEMS lens scanner design is illustrated in Fig. 2, where a microstage is loaded with a lens on its ring-shaped frame. The microstage consists of a ring-shaped frame and 8 sets of bimorph actuators. The opening of the ring-shaped frame provides the lens with a large clear aperture (1.8 mm in diameter). The bimorph actuators support and move the ring-shaped frame (thus the lens) in the vertical direction. The lens is a plano-convex aspheric glass lens with a diameter of 2.4 mm and a NA of 0.58. It weighs about 8 mg. The entire chip is circular for minimizing its footprint and better fitting into an imaging probe which is typically circular too. The diameter of the chip is 4.4 mm.
This actuator is based on electrothermal bimorphs. Each bimorph consists of two thin layers with different coefficients of thermal expansion (CTEs). A cantilevered bimorph beam can convert a small in-plane strain difference between the two layers into a large out-of-plane displacement. As shown in Fig. 3(a), an as-fabricated bimorph curls into an arc shape whose radius of curvature is determined by the residual stresses in the bimorph. Aluminum and silicon dioxide (Al/SiO2), are used here as the two bimorph constitutive materials, whose large CTE difference is used to maximize the curling. As the temperature increases by Joule heating, the bimorph’s radius of curvature increases and its tip drops. The relation between the arc angle and the radius of curvature, ρ, at a given temperature, T, can be expressed as
where l is the length of the bimorph. Thus, when the temperature is changed from the room temperature T0 to an elevated temperature T1, the change of the tip angle of the bimorph is given by
Fig. 3. (a) Schematic of a cantilevered bimorph. (b) Side view of an ISC bimorph actuator. (c) Side view of a DISC bimorph actuator. (d) SEM of a MEMS microstage based on DISC bimorph actuators.
For a bimorph composed of only two layers, Eq. (2) can be expressed as [28]
where t is the bimorph thickness, $\Delta \alpha $ is the CTE difference of the two materials, and $\beta $ is the actuation coefficient related to the thickness and biaxial Young’s moduli of the two materials. To generate Joule heating, a thin Pt layer, which has a stable and relatively large electrical resistivity compared to other metals, has been added between the two functional bimorph layers. As shown in Fig. 3(a), the simple cantilever bimorph has a large initial tilt angle and suffers lateral shift during actuation.To overcome the tilting of a simple bimorph, an inverted-series-connected (ISC) bimorph actuator design, as illustrated in Fig. 3(b), was proposed [20,29], in which a non-inverted bimorph and an inverted bimorph is connected in series but with flipped layer order. With two same ISC structures folded into a double ISC (DISC) structure, a pure out-of-plane motion can be obtained at its loading point, as illustrated in Fig. 3(c), where the lateral shift at the free end is irrelevant. An SEM image of a MEMS microstage with eight DISC actuators is shown in Fig. 3(d), where the 1-mm-diameter platform is 160 µm above the top surface of the silicon substrate. This MEMS microstage has achieved a maximum piston displacement of 160 µm. This microstage was designed and fabricated by the authors as a reference. However, the total stiffness of all the bimorph actuators combined was only 1 N/m, which would lead to a very low resonance frequency of 56 Hz if an 8-mg lens would be mounted. Thus, the ISC actuator design must be optimized.
For simplicity, the stiffness of one DISC actuator shown in Fig. 3(c) can be treated as a simple clamped-guided beam. The stiffness of a DISC actuator can be increased by a3 times if its length is reduced by a, but its travel range will decrease by a2. It is challenging to achieve high stiffness and large travel simultaneously.
In order to overcome this tradeoff, a new serpentine ISC (SISC) actuator design is proposed (Fig. 4(a)), where three ISC structures are connected in series. The first and the third ISC are identical, while the second ISC is slightly shorter due to the practical layout consideration. As shown in Fig. 4(b), each of the three ISC segments is an S-shaped beam with the tilts of both ends compensated. One advantage of this SISC structure is that the anchor and loading points are at its two opposite ends, facilitating parallel configuration. Four such SISC structures can form one set of bimorph actuators, as shown in Fig. 4(c). The proposed MEMS stage consists of eight sets of such bimorph actuators, i.e., thirty-two SISC actuators, as shown in Fig. 2.
Fig. 4. The SISC actuator design. (a) 3D model. (b) Side view. (c) An actuator block consisting of four parallel SISC actuators and one small silicon beam. (d) Side view of the ISC actuator with all the layers except Cr adhesion layers.
The design of the SISC actuators starts from the basic ISC structure (Fig. 3(b)). As shown in Fig. 4(d), each ISC consists of three segments: a non-inverted Al/SiO2 bimorph (NI), an overlapped SiO2/Al/SiO2 trimorph (OV), and an inverted SiO2/Al bimorph (IV). The residual stresses in the Al and SiO2 layers cause the NI and IV segments to initially curl towards the Al layer after releasing, while the OV segment stays flat due to the stress compensation of its top and bottom SiO2 layers.
To generate vertical displacement without bending on its central platform, each ISC structure must eliminate its tilt angle, i.e., θ = 0. Table 1 lists the structure parameters used in this design. All layers are assumed to be isotropic and uniform. The bimorph width is set as 25 µm, which is limited by the practical releasing process. Although wider bimorphs can be used to increase the stiffness of the microstage, the increased releasing time during the isotropic dry etching (see Fig. 7(j)) will result in undercutting too much silicon (Si) underneath the central platform, making the central platform fragile.
Table 1. ISC Structure Parameters and Physical Properties
As shown in Eq. (3), once the materials are chosen for the bimorphs, to maximize the angular displacement is simply maximizing β. It has been found that β reaches its maximum value of 1.5 when the thickness ratio of the two bimorph layers is the square root of the inverse of their Young’s modulus ratio if only the two bimorph layers are considered for the bimorphs [30]. However, in reality there are always several more layers needed for heating, electrical insulation, and passivation, in addition to the two bimorph layers. As depicted in Fig. 4(d), (a) bottom thin SiO2 layer adhesion is used to promote the adhesion between Pt and Si substrate, and another thin SiO2 layer is deposited on the Pt layer for electrical isolation. In addition, a thin Cr layer is used as the adhesion layer for Pt and SiO2.
Thus, according to Eq. (2), the radius of curvature, ρ, of a multimorph must be known, which is readily derived from the strain continuity equation as [31]:
As indicated by Eq. (9), the angular displacement is exactly inversely proportional to $\hat{\rho }\; $ for given bimorph length and temperature change. Plugging the values listed in Table 1, we can plot the change of $1/\hat{\rho }$ versus the Al thickness with the 1st SiO2 fixed at 1.3 µm for the NI bimorph, as shown in Fig. 5 (solid line), where $1/\hat{\rho }$ reaches the minimum at the Al thickness of around 0.76 µm. The trend of the solid line in Fig. 5 shows that the $1/\hat{\rho }$ of the NI bimorph does not decrease significantly after 0.76 µm, so 1.2 µm is chosen as the Al thickness, resulting in a decrease in $1/\hat{\rho }$ less than 7% but around three times increase of the bending rigidity. As can be seen on the solid line in Fig. 5, at this chosen Al thickness,
Fig. 5. Numerically calculated $1/\hat{\rho }$ of the NI bimorph with regard to the thickness of Al (solid line), and of the IV bimorph at 1.2-µm-thick Al with regard to the thickness of the 2nd SiO2 (dashed line).
The $1/\hat{\rho }\; $ of the IV bimorph is also calculated using the same fixed values shown in Table 1 and the newly determined Al thickness of 1.2 µm. The result is plotted as the dashed line in Fig. 5, where $1/\hat{\rho }\; $gradually changes from negative to positive and reaches a peak at the 2nd SiO2 thickness of 1.58 µm. Thus, to keep the positively large enough and the IV and NI bimorphs balanced, the thickness of the 2nd SiO2 is chosen as 1.3 µm, correspondingly as shown on the dashed line in Fig. 5,
As can be seen from Eqs. (11) and (12), the NI and IV bimorphs bend in the opposite directions, providing a means of cancelling the tip tilt angle. Thus, assuming both NI and IV bimorphs have the same temperature change upon actuation, according to Eq. (9), the optimal length ratio is given by
As shown in Fig. 3(b), there exists a lateral shift with three ISC structures In order to eliminate this lateral shift, a completely symmetrical ISC actuator set, as shown in Fig. 3(d), is designed, where two SISC actuators are attached to each side of a thick Si slab and these two SISC actuators are mirrored to ensure there are no lateral shift. The lengths of the three ISC segments of each SISC actuator are listed in Table 2. The central Si slab has a width of 140 µm for high rigidity for heavy duty loading. This actuator set can be fully confined in an area of 1024 × 220 µm2.
Table 2. ISC Length Properties
A 3D model of the proposed MEMS lens scanner is shown in Fig. 2, where the MEMS microstage consists of eight actuator sets that are evenly distributed along the perimeter of a central ring-shaped platform with a 1.8-mm clear aperture. The radial width of the ring-shaped platform is 0.5 mm. The microstage layout with multiple concentric actuator sets not only ensures high stiffness, but also maximizes the clear aperture while minimizing the die area. With eight distributed actuator sets, the tilt and lateral shift of this microstage are greatly reduced even when the loaded lens is not perfectly aligned. The electrical connections of the SISC actuators are realized by the embedded Pt thin layer and Al on the silicon substrate.
The microstage design is simulated in COMSOL by using the analytically determined structural parameters. For simplicity, the thin Cr layer is omitted. Figure 6(a) shows the result of the simulated initial state, indicating the initial elevation is around 225 µm and the initial tilt angle of each ISC structure is well compensated. The microstage yields a 225 µm vertical displacement at a temperature rise of ΔT = 210 K, as shown in Fig. 5(b).
Fig. 6. Simulation results. (a) The initial elevation due to residual stresses (225 µm above the substrate). (b) After a uniform temperature change (ΔT=210 K) on all actuators
3. MEMS microstage fabrication
A combination of surface and bulk micromachining that generally follows the previously reported process for MEMS mirrors [20] is used for the device fabrication. The SOI wafer employed in the fabrication has a 25 µm-thick device layer, a 2 µm-thick buried oxide (BOX) layer, and a 400 µm-thick handle layer. At first, a 1.3µm SiO2 layer is deposited with plasma-enhanced chemical vapor deposition (PECVD) and wet etched to form the bottom layer of the NI bimorphs (Fig. 7(a)). After that, a 0.1 µm-thick SiO2 layer is deposited on top primarily for promoting the adhesion on Si, followed by sputtering and lifting off a Cr/Pt/Cr layer to form the embedded resistors for Joule heating (Fig. 7(b)). Another 0.1 µm-thick SiO2 layer is deposited as an electrical isolation layer and patterned by reactive ion etching (RIE) to form vias, followed by the sputter and lift-off of a 1.2µm-thick Al layer to form the top layer of the NI bimorphs and the bottom layer of the inverted (downward) IV bimorphs (Fig. 7(c)). Next, a final 1.3 µm-thick SiO2 layer is deposited via PECVD and patterned by RIE to form the top layer of the IV bimorphs (Fig. 7(d)). In addition, a thin Al layer with a thickness of 0.2 µm is sputtered and patterned on the top of the central platform for extra protection for later DRIE (Fig. 7(e)). So far, all the bimorphs, frames, and the clear aperture have been defined.
Fig. 7. Fabrication process. (a) PECVD 1st SiO2 and wet etch. (b) Pt sputtering and lift-off. (c) Al sputtering and lift-off. (d) PECVD 2nd SiO2 and wet etch. (e) Al (for protection) sputtering and lift-off. (f) Backside Al2O3 sputtering and wet etch. (g) Carrier wafer bonding. (h) Backside DRIE Si and RIE BOX. (i) Frontside anisotropic DRIE. (j) Frontside isotropic DRIE.
Next, the SOI wafer is flipped over. A 100 nm-thick Al2O3 layer is sputtered and etched using diluted HF (Fig. 7(f)). Later, a thick layer of photoresist is spun on top of the wafer, and a carrier wafer is then bonded on it with a gentle pressure (Fig. 7(g)). After baking, DRIE is performed to etch the Si down to the BOX layer and then RIE is used to remove the BOX layer (Fig. 7(h)). Next, individual die are separated from the carrier wafer by soaking in acetone. Then the separated dies are flipped and bonded on another carrier Si wafer with photoresist or thermal tape for final releasing. The device releasing includes two steps. First, a DRIE anisotropic etching is used to etch down the exposed device layer by about 15 µm (Fig. 7(i)). Second, a DRIE isotropic etching is immediately followed to etch away all the Si left between and under the bimorphs.
Note that the releasing step must be strictly timed. Otherwise the Si near the bimorph anchor points and central ring will be over-etched, which will seriously reduce the lens carrying capability of this MEMS microstage. Thus, the bimorph widths cannot be too large. Through multiple trials, the maximum bimorph width is set at 25 µm. After releasing, the central frame moves upwards due to the residual stresses in the bimorph material layers.
4. MEMS microstage characterization
A scanning electron micrograph (SEM) of a fabricated device is shown in Fig. 8(a), where the footprint of the device is circular with a diameter of 4.4 mm. The central ring-shaped frame has an outer diameter of 2.8 mm and a clear aperture with a diameter of 1.8 mm. The central ring is exactly concentric with the chip outer frame, without any rotation or tilt observed. The elevation of the central ring-shaped platform is 257 µm above the substrate after release. Before using it, the device is gone through a burn-in process in which a biased sinewave voltage of Vpp = 1 V is applied to all the SISC bimorph actuators for 40 hours. After the burn-in process, the elevation of the central platform of the microstage is stabilized at 218 µm. The change of the elevation is due to the relaxation and redistribution of the residual stresses in the bimorphs of the SISC bimorph actuators. As shown in Fig. 8(b), all the ISC structures are in S shape after releasing. Round hinges are used to connect all ISC structures for the purpose of mitigating the stresses during motion. In addition, the round hinges and the overlap parts of the ISC structures are composed of all of the material layers deposited for enhancing the stiffness. Since they are much shorter than NI and IV bimorphs, they are treated as rigid beams and their curvatures are omitted in the above analytical part.
Fig. 8. SEMs of a fabricated microstage. (a) The whole view. (b) The close-up view of one half of an actuator set, where the silicon beam is 257 µm above the outer silicon frame.
The quasi-static response of the piston motion is measured with a microscope (Olympus BX51). As shown in Fig. 9(a), maximum displacement of 210 µm is obtained at a DC voltage of 5.25 V applied to all eight actuator sets. The frequency response has also been characterized. A position-sensitive detector (PSD) is employed for tracking the displacement of the central ring. Figure 10 plots the frequency response at a biased sinusoidal voltage signal of 1.5 + 0.5cos(2πft) Volts. The first peak in Fig. 10 corresponds to the piston resonant frequency at 1.064 kHz while the other two higher-frequency peaks are the tip-tilt resonant modes.
Fig. 9. The measured quasi-static piston displacement of the fabricated MEMS microstage without and with a lens, and lateral shift with the lens. The measurement errors were ± 1.0 µm.
Fig. 10. The frequency response of the MEMS microstage with and without a lens under a swept frequency AC voltage, 1.5 + 0.5cos(2πft) Volts.
5. MEMS lens scanner assembling and imaging experiment
The MEMS microstage shown in Fig. 8(a) is developed to load a lens. To ensure high optical quality for potential applications in microendoscopy, an aspheric glass microlens with a NA of 0.58 is used here. The diameter of the microlens is 2.4 mm, so it fits well on the central ring-shaped platform of the MEMS microstage. Figure 11 shows two photos of an assembled MEMS lens scanner, where the plano-convex lens can be glued on the central platform with either the flat side or convex side down. The central platform of the MEMS microstage is lowered by 12.5 µm after the 8-mg microlens is loaded. Thus the z-axis stiffness of the MEMS microstage is about 6.3 N/m, a 27 times increase compared to the previous work [26]. The measured quasi-static response of this assembled MEMS lens scanner is plotted as the cross dots in Fig. 9, showing only small difference from that of the MEMS microstage without a lens. The maximum travel range of the assembled MEMS lens scanner drops only a few percent from 210 µm to 205 µm. The lateral shift of the MEMS lens scanner’s central platform is measured and the experimental data is plotted on Fig. 9, where the maximum lateral shift is less than 3 µm, which is due to the discrepancies of the microactuators caused by the fabrication variations. The tilt effect of the MEMS lens scanner is not observable in the imaging experiment below, but will be investigated in the future when the application system is sensitive to the lens tilt. The measured piston resonant frequency drops to about 140 Hz due to the added mass as shown in Fig. 10, but it is still 5 times greater than that of the previous design [26].
Fig. 11. Photos of a MEMS lens scanner with a lens mounted. Left: The convex surface of the lens facing upward. Right: The convex surface of the lens facing downward.
The above measurement results show great promise of this MEMS lens scanner for applications in optical endomicroscopy. Thus, a simple imaging experiment has been performed to verify the axial scan ability of the assembled lens scanner. As shown in Fig. 12, the MEMS lens scanner is placed about 3.25 mm under the focal plane of a microscope and a sample (two 1951 USAF resolution test targets stack) is placed about 3 mm under the MEMS lens scanner. The opaque lines on the two resolution test targets are facing each other. The gap between the two resolution targets is kept at 170 µm. A back illumination is used to enhance the light flux towards the microscope. The images of the opaque lines on the targets could be captured through the camera mounted on top of the microscope.
Figure 13 shows the pictures of the MEMS lens scanner being actuated at three different voltages of 0 V, 5 V, and 8 V and the corresponding microscopic images of the sample patterns. As shown in Fig. 13, the lens scanner is first at the focus of the bottom resolution target (Fig. 13(a)), and then focused in between the two resolution targets (Fig. 13(b)), and finally tuned to the top resolution target (Fig. 13(c)). The 2.2 µm lines on both targets are clearly resolved, showing that the MEMS lens scanner has high-quality depth scan ability. The 2.2 µm lines are the narrowest lines on the targets.
Fig. 13. Microscopic images with regard to the MEMS lens scanner at different actuation voltages of (a) 0 V, (b) 5 V, and (c) 8 V, corresponding to a displacement of 0 µm, 77 µm, and 151 µm, respectively. Scale bar: 25 µm. The inset in (a): A complete Group 7 image taken with the lens, in which Element 6 (linewidth 2.2 µm) is clearly delineated.
6. Discussion and conclusion
As the measurement results presented in Sections 4 and 5 indicate, the MEMS microstage is capable of lifting an 8-mg lens by as far as 205 µm, which is suitable for confocal microscopy and two-photon microscopy imaging applications. If a greater imaging depth is needed, the electrothermal bimorph actuator design can be optimized to increase the travel range. The bimorph length can be easily increased to increase the piston displacement. For example, a 20% increase of the bimorph length will increase the travel range by more than 40% but at the price of a decreased stiffnesses. In this case, a way of arranging more SISC bimorph actuators must be figured out or using bimorph materials with greater Young’s moduli.
The resonant frequency of the MEMS microstage loaded with lens reaches 140 Hz, which is applicable to endomicroscopic imaging. If a higher resonant frequency is required such as in head-mounting TPM of freely moving animals, stiffer electrothermal bimorphs and/or lighter lenses may be employed. For stiffness enhancement, one method is through increasing the layer thicknesses of the bimorphs; for example, a 26% increase in the bimorph thickness will double its stiffness. Another method to increase the stiffness it to choose the bimorph materials with higher Young’s moduli; for example, the stiffness of a Cu/W bimorph is about three times of that of an Al/SiO2 bimorph with the same width and length. Cu/W bimorphs have been successfully developed for MEMS mirrors [21,23]. For lighter lenses, high-index plastic aspheric lenses or thin flat lenses such as Fresnel lenses [33,34] and metamaterial lenses [35,36] may be employed.
In summary, a MEMS microstage based on a new SISC bimorph actuator design has been fabricated and characterized. With a high-NA aspheric glass lens loaded on the MEMS microstage, a compact lens scanner with large axial tuning range, high stiffness, low drive voltage, and high image quality has been successfully demonstrated. An analytical model based on multimorph beam bending theory has been developed to take into account the layers in addition to the two main layers in bimorph beams. This model can be used to optimize the thicknesses of the layers in the bimorph beams and the length ratio of the NI and IV segments. More effort is ongoing to integrate this MEMS lens scanner into a TPM probe that can be mounted on the head of a freely moving animal for real time neural monitoring.
Funding
National Science Foundation (1512531); National Institutes of Health (R01EB020601).
Acknowledgements
The authors thank the staff of the UF Research Service Centers for their assistance in the microfabrication.
Disclosures
The authors declare no conflicts of interest.
References
1. R. D. Frostig, In vivo optical imaging of brain function (CRC press, 2009).
2. B. R. Masters, Confocal microscopy and multiphoton excitation microscopy: the genesis of live cell imaging (SPIE press, 2006).
3. H. Neumann, R. Kiesslich, M. B. Wallace, and M. F. Neurath, “Confocal laser endomicroscopy: technical advances and clinical applications,” Gastroenterology 139(2), 388–392.e2 (2010). [CrossRef]
4. F. Helmchen, W. Denk, and J. N. Kerr, “Miniaturization of two-photon microscopy for imaging in freely moving animals,” Cold Spring Harbor Protocols 2013, pdb. top078147 (2013).
5. W. Zong, R. Wu, M. Li, Y. Hu, Y. Li, J. Li, H. Rong, H. Wu, Y. Xu, and Y. Lu, “Fast high-resolution miniature two-photon microscopy for brain imaging in freely behaving mice,” Nat. Methods 14(7), 713–719 (2017). [CrossRef]
6. W. Piyawattanametha, E. D. Cocker, L. D. Burns, R. P. Barretto, J. C. Jung, H. Ra, O. Solgaard, and M. J. Schnitzer, “In vivo brain imaging using a portable 2.9 g two-photon microscope based on a microelectromechanical systems scanning mirror,” Opt. Lett. 34(15), 2309–2311 (2009). [CrossRef]
7. S. Tang, W. Jung, D. T. McCormick, T. Xie, J. Su, Y.-C. Ahn, B. J. Tromberg, and Z. Chen, “Design and implementation of fiber-based multiphoton endoscopy with microelectromechanical systems scanning,” J. Biomed. Opt. 14(3), 034005 (2009). [CrossRef]
8. N. Callamaras and I. Parker, “Construction of a confocal microscope for real-time xy and xz imaging,” Cell Calcium 26(6), 271–279 (1999). [CrossRef]
9. K. H. Kim, C. Buehler, and P. T. So, “High-speed, two-photon scanning microscope,” Appl. Opt. 38(28), 6004–6009 (1999). [CrossRef]
10. N. Ji, J. Freeman, and S. L. Smith, “Technologies for imaging neural activity in large volumes,” Nat. Neurosci. 19(9), 1154–1164 (2016). [CrossRef]
11. C. C. Archer-Zhang, W. B. Foster, R. D. Downey, C. L. Arrasmith, and D. L. Dickensheets, “Dynamic performance of microelectromechanical systems deformable mirrors for use in an active/adaptive two-photon microscope,” J. Biomed. Opt. 21(12), 121507 (2016). [CrossRef]
12. B. N. Ozbay, J. T. Losacco, R. Cormack, R. Weir, V. M. Bright, J. T. Gopinath, D. Restrepo, and E. A. Gibson, “Miniaturized fiber-coupled confocal fluorescence microscope with an electrowetting variable focus lens using no moving parts,” Opt. Lett. 40(11), 2553–2556 (2015). [CrossRef]
13. W. J. Shain, N. A. Vickers, B. B. Goldberg, T. Bifano, and J. Mertz, “Extended depth-of-field microscopy with a high-speed deformable mirror,” Opt. Lett. 42(5), 995–998 (2017). [CrossRef]
14. W. R. Zipfel, R. M. Williams, and W. W. Webb, “Nonlinear magic: multiphoton microscopy in the biosciences,” Nat. Biotechnol. 21(11), 1369–1377 (2003). [CrossRef]
15. C. L. Smith, “Basic confocal microscopy,” Curr. Protoc. Neurosci. 56(1), 1–18 (2011). [CrossRef]
16. S. Kwon, V. Milanovic, and L. P. Lee, “Large-displacement vertical microlens scanner with low driving voltage,” IEEE Photonics Technol. Lett. 14(11), 1572–1574 (2002). [CrossRef]
17. S. Bargiel, C. Gorecki, T. Verdot, K. Laszczyk, J. Albero, and L. El Fissi, “Electrostatically driven optical Z-axis scanner with thermally bonded glass microlens,” Procedia Eng. 5, 762–765 (2010). [CrossRef]
18. H. Zeng and M. Chiao, “Magnetically actuated MEMS microlens scanner for in vivo medical imaging,” Opt. Express 15(18), 11154–11166 (2007). [CrossRef]
19. A. Michael and C. Kwok, “Piezoelectric micro-lens actuator,” Sens. Actuators, A 236, 116–129 (2015). [CrossRef]
20. K. Jia, S. Pal, and H. Xie, “An electrothermal tip–tilt–piston micromirror based on folded dual S-shaped bimorphs,” J. Microelectromech. Syst. 18(5), 1004–1015 (2009). [CrossRef]
21. L. Zhou, X. Zhang, and H. Xie, “An electrothermal Cu/W bimorph tip-tilt-piston MEMS mirror with high reliability,” Micromachines 10(5), 323 (2019). [CrossRef]
22. H. Xie, S. Todd, A. Jain, and G. K. Fedder, “Single-crystal silicon based electrothermal MEMS mirrors for biomedical imaging applications,” in MEMS/NEMS (Springer, 2006), pp. 1429–1471.
23. X. Zhang, L. Zhou, and H. Xie, “A fast, large-stroke electrothermal MEMS mirror based on Cu/W bimorph,” Micromachines 6(12), 1876–1889 (2015). [CrossRef]
24. S. Pal, L. Zhou, X. Zhang, and H. Xie, “Electrothermally actuated MEMS mirrors: Design, modeling, and applications,” in Optical MEMS, Nanophotonics, and Their Applications (CRC Press, 2017), pp. 173–200.
25. L. Wu and H. Xie, “A millimeter-tunable-range microlens for endoscopic biomedical imaging applications,” IEEE J. Quantum Electron. 46(9), 1237–1244 (2010). [CrossRef]
26. L. Liu, E. Wang, X. Zhang, W. Liang, X. Li, and H. Xie, “MEMS-based 3D confocal scanning microendoscope using MEMS scanners for both lateral and axial scan,” Sens. Actuators, A 215, 89–95 (2014). [CrossRef]
27. L. Zhou, X. Yu, and H. Xie, “A robust compact lens scanner with large tunable range,” in 2020 IEEE 33rd International Conference on Micro Electro Mechanical Systems (MEMS) (IEEE, 2020), pp. 1149–1152.
28. N. Hasan, A. Banerjee, H. Kim, and C. H. Mastrangelo, “Tunable-focus lens for adaptive eyeglasses,” Opt. Express 25(2), 1221–1233 (2017). [CrossRef]
29. S. T. Todd, A. Jain, H. Qu, and H. Xie, “A multi-degree-of-freedom micromirror utilizing inverted-series-connected bimorph actuators,” J. Opt. A: Pure Appl. Opt. 8(7), S352–S359 (2006). [CrossRef]
30. S. T. Todd and H. Xie, “An electrothermomechanical lumped element model of an electrothermal bimorph actuator,” J. Microelectromech. Syst. 17(1), 213–225 (2008). [CrossRef]
31. E. Garcia and N. Lobontiu, “Induced-strain multimorphs for microscale sensory actuation design,” Smart Mater. Struct. 13(4), 725–732 (2004). [CrossRef]
32. L. Wu and H. Xie, “A large vertical displacement electrothermal bimorph microactuator with very small lateral shift,” Sens. Actuators, A 145-146, 371–379 (2008). [CrossRef]
33. H.-Y. Lu, C.-E. Lu, Z.-R. Huang, S. Lin, S.-C. Lo, R. Chen, and W. Fang, “Fabrication and integration of binary phased Fresnel lens and micro linear actuator for IR laser beam scanning application,” in 2019 20th International Conference on Solid-State Sensors, Actuators and Microsystems & Eurosensors XXXIII (TRANSDUCERS & EUROSENSORS XXXIII) (IEEE, 2019), pp. 1584–1587.
34. M. J. Moghimi, J. Fernandes, A. Kanhere, and H. Jiang, “Micro-Fresnel-zone-plate array on flexible substrate for large field-of-view and focus scanning,” Sci. Rep. 5(1), 15861 (2015). [CrossRef]
35. E. Arbabi, A. Arbabi, S. M. Kamali, Y. Horie, M. Faraji-Dana, and A. Faraon, “MEMS-tunable dielectric metasurface lens,” Nat. Commun. 9(1), 812–819 (2018). [CrossRef]
36. T. Roy, S. Zhang, I. W. Jung, M. Troccoli, F. Capasso, and D. Lopez, “Dynamic metasurface lens based on MEMS technology,” APL Photonics 3(2), 021302 (2018). [CrossRef]
