2. Design

cells, blood vessels and tissues, for applications in micro-assembly of Microelectromechanical

Different actuators were investigated due to the significant role in the MEMS configuration. The actuation methods include mainly the electrostatic, electromagnetic, piezoelectric and electrothermal principles. Each actuation approaches have their proper disadvantages and benefits in agreement with the designed purpose. The actuators are usually integrated with MEMS for the necessary need of energy conversion, motion generation and force production [1–3]. The V-shaped actuators are widely used for grippers, micro-valves, micro-pumps and other devices. V-shaped electrothermal actuators have the advantages of generating a large force (up to several 100 mN), the simple structure design, a lower dive voltage and a large deformation. Que et al. [2] developed single and cascaded V-shaped electrothermal actuators and present the experimental results. Shen and Chen [3] present an analytical model for cascaded V-shaped actuators bringing a complete description of the mechanical performance. Usually, materials such as silicon, polysilicon or aluminum are used as the structural material

A variety of microgrippers have been studied using the SU-8 based electrothermal actuators designed on different configurations such as, U-shape or V-shape. This is proving the interest in the bio-micro-manipulation domain [1, 4–23]. SU-8 is a highly crosslinked epoxy-type photo-patternable polymer which has been used extensively as the preferred polymer material for fabrication of biocompatible structures. The SU-8 polymer has a relatively large coefficient of thermal expansion (CTE) of 52 ppm, good mechanical strength with a modulus of elasticity of around 4.02 GPa and good thermal stability with a glass transition temperature of 210C [15], which make it a good polymer material for fabrication of electrothermal actuators. The polymer V-shaped actuators are preferred for the better performance in aqueous medium [4]. Different processing technologies were investigated and realized in order to fabricate reliable microgripper with reduces out-of-plane displacement [17–22]. Usually two or three material

In this chapter, we report a complete work regarding the design, numerical simulation results, fabrication process and the experimental investigations of an SU-8 polymeric microgripper. The design is based on the cascaded V-shaped electrothermal actuators. The SU-8 microgripper can be used for micro-robotics and bio-manipulation and assembly applications. The microgripper was numerically investigated using the coupled electro-thermo-mechanical simulations based on finite element method (FEM) and using the Coventorware 2014 software in order to confirm the performance of the microgripper. To fabricate the microgripper, a sandwich structure actuator with three layers was used. Two kings of fabrication processes were presented in order to improve the structure functionality. As structural material of the microgripper, the SU-8 biocompatible polymer was used during the fabrication process. The metallic micro-heaters were encapsulated in the polymeric actuation structures of the microgrippers to reduce the undesirable outof-plane displacement of the gripper tips, the mechanical stress and to improve the thermal efficiency and the electrical isolation of the structure. Experimental testing and characterizations have been performed to determine the openings and the temperatures of the microgripper tips as function of electrical current. A comparison between the simulation results and the measure-

Systems (MEMS) and MOEMS components (lenses, fibers) and in micro-robotics.

of such actuators.

26 Actuators

ments were also presented.

layers are utilized to compose a sandwich structure.

The SU-8 microgripper was designed in a previous work using the principle of the cascaded Vshaped electrothermal actuators [22]. The gripper was designed with two initial opening of 50 μm and 100 μm, respectively (Figure 1). When the gripper structure is electro-thermally actuated the arms and the jaws will close and will be able to handgrip a micro-object. The total length of the gripper arms used to grasp an object is of 920 μm. The arms were designed with a width of 20 μm [22]. A metallic micro-heater is implanted between two SU-8 layers. The heater lines have a width of 10 μm and were designed first, to be fabricated using Cr/Au/Cr materials and second, to be fabricated using only the gold. Usually, a chromium thin layer or other adhesion layers are used to improve the connection between the gold metal and the polymer. On the other hand, it was reported that the deposition of the SU-8 polymer over the gold metal do not need an adhesion layer [23].

The optimized design consists of symmetrically disposed of three material layers. A metallic layer for the heater is implanted between two SU-8 based structure layers having the same thickness, as described previously [18–22]. The thicknesses of the Cr/Au/Cr films were 10 nm/300 nm/10 nm. The thickness of the gold layer is 100–300 nm. For each SU-8 layer we obtained a thickness of 9 μm. The details of the fabrication process where using the Cr/Au/Cr films have been reported also previously in [21] and when using only gold in [24] but for other gripper designs.

The proposed microgripper in this work was designed symmetrically with encapsulated metallic micro-heaters in the structural material of the grippers, the SU-8 polymer, in order to reduce the undesirable out-of-plane displacement of the gripper, to obtain the electrical isolation of the heaters and to reduce the mechanical stress that can occur in the structure [22].

Figure 1. Schematic design of the SU-8 microgripper [22].

#### 3. Finite element simulation

In order to check the performance of the microgripper, finite element simulations were performed. The microgripper with the initial opening of 50 μm was numerically investigated.

Coupled electro-thermo-mechanical simulations were completed using the MemMech simulator from the Coventorware 2014 software tool. A simplified 3D microgripper model (Figure 2) was meshed using hexahedral elements (Extruded bricks). The number of volume elements was optimized choosing the proper size of the mesh elements using the Split and Merge algorithm. The thicknesses of the SU-8 layers and the gold layer are 18 μm and 300 nm, respectively.

The materials properties and the surface boundary conditions were set for the simulations (Table 1). The initial temperature of the whole structure and the temperature of the environment were considered to be T0 = 27C, with respect to the Coventorware settings requirements for such kind of analyses. The radiation losses from the device are negligible in comparison with the heat loss by convection to the surrounding media [4], since the maximum temperature reached in the microgripper, in order to operate, is lower than 800C. The air convection coefficient was set to 250 W/m<sup>2</sup> K [4].

The Young's modulus of the SU-8 was measured with the nanoindentation technique and was set in simulations for a value of 4.6 GPa (Figure 3). The indentation tests have been carried out using A G200 Nano Indenter from Agilent Technologies (Keysight Technologies).

The thermal coefficient of expansion was fixed at 52 (ppm/o C) and the thermal conductivity at 2 X 105 pW/μmK. For the gold layer we used a Young's modulus of 77 GPa reported for thin films.

The TCR, temperature coefficient of resistance, was measured for the Cr/Au/Cr thin films and obtained the value 0.001569/C. This value is significantly smaller than the value of 0.0034– 0.0037/C used for the bulk gold material. For the thin gold film the TCR was measured at 0.00314/C.

Electrical conductivity of the gold layer was set as function of the temperature using the

Property SU-8 Au Young's Modulus (E) [GPa] 4.6 77 Poisson ratio (ν) 0.22 0.35

TCE Coeff. of Thermal Expansion (α) [1/K] <sup>52</sup> � <sup>10</sup>�<sup>6</sup> 14.1 � <sup>10</sup>�<sup>6</sup> (300 K)

An SU-8 Microgripper Based on the Cascaded V-Shaped Electrothermal Actuators: Design, Fabrication, Simulation…

http://dx.doi.org/10.5772/intechopen.75544

29

Electric Cond. [pS/μm] Conf. Eq. (1) and (2)

Thermal Conductivity (λ) [pW/(μm�K)] 0.2 � <sup>10</sup><sup>6</sup> <sup>297</sup> � <sup>10</sup><sup>6</sup>

SpecificHeat(pJ/kgK) 1.2 � <sup>10</sup><sup>15</sup> 12.87 � <sup>10</sup><sup>15</sup> TCR [/�C] — 0.001569

Softening point [�C] 210

Table 1. The materials properties used in simulations.

where r(T) is the resistivity as function of the temperature, r<sup>0</sup> is the resistivity at T0, ε is the

The simulation results regarding the temperatures distribution reached in the microgripper when it is actuated were visually compared with the thermal measurements realized using an IR camera SC5000 from FLIR system (Figure 4). The thermal measurements show that the distribution of the temperatures in the microgripper has a similar map with the simulation results. The temperature values at the tips remains near initial temperature of the environment.

rð Þ¼ T r<sup>0</sup> ½ � 1 þ ε ð Þ T � T0 (1)

σð Þ¼ T 1=rð Þ T (2)

Eqs. (1) and (2):

TCR of the gold and σ is the electrical conductivity.

Figure 3. Young's modulus vs displacement into surface of SU-8 layer.

Figure 2. The simplified 3D model of the microgripper with encapsulated heaters in SU-8 polymer: The layers sandwich structure used in FEM simulations (Coventorware 2014) [22].

An SU-8 Microgripper Based on the Cascaded V-Shaped Electrothermal Actuators: Design, Fabrication, Simulation… http://dx.doi.org/10.5772/intechopen.75544 29


Table 1. The materials properties used in simulations.

Coupled electro-thermo-mechanical simulations were completed using the MemMech simulator from the Coventorware 2014 software tool. A simplified 3D microgripper model (Figure 2) was meshed using hexahedral elements (Extruded bricks). The number of volume elements was optimized choosing the proper size of the mesh elements using the Split and Merge algorithm. The thicknesses of the SU-8 layers and the gold layer are 18 μm and 300 nm,

The materials properties and the surface boundary conditions were set for the simulations (Table 1). The initial temperature of the whole structure and the temperature of the environment were considered to be T0 = 27C, with respect to the Coventorware settings requirements for such kind of analyses. The radiation losses from the device are negligible in comparison with the heat loss by convection to the surrounding media [4], since the maximum temperature reached in the microgripper, in order to operate, is lower than 800C. The air convection

The Young's modulus of the SU-8 was measured with the nanoindentation technique and was set in simulations for a value of 4.6 GPa (Figure 3). The indentation tests have been carried out

X 105 pW/μmK. For the gold layer we used a Young's modulus of 77 GPa reported for thin films.

The TCR, temperature coefficient of resistance, was measured for the Cr/Au/Cr thin films and obtained the value 0.001569/C. This value is significantly smaller than the value of 0.0034– 0.0037/C used for the bulk gold material. For the thin gold film the TCR was measured at

Figure 2. The simplified 3D model of the microgripper with encapsulated heaters in SU-8 polymer: The layers sandwich

C) and the thermal conductivity at 2

using A G200 Nano Indenter from Agilent Technologies (Keysight Technologies).

The thermal coefficient of expansion was fixed at 52 (ppm/o

structure used in FEM simulations (Coventorware 2014) [22].

respectively.

28 Actuators

0.00314/C.

coefficient was set to 250 W/m<sup>2</sup> K [4].

Figure 3. Young's modulus vs displacement into surface of SU-8 layer.

Electrical conductivity of the gold layer was set as function of the temperature using the Eqs. (1) and (2):

$$
\rho(\mathbf{T}) = \rho\_0 \left[ 1 + \varepsilon \left( \mathbf{T} - \mathbf{T}\_0 \right) \right] \tag{1}
$$

$$
\sigma(\mathbf{T}) = \mathbf{1}/\rho(\mathbf{T})\tag{2}
$$

where r(T) is the resistivity as function of the temperature, r<sup>0</sup> is the resistivity at T0, ε is the TCR of the gold and σ is the electrical conductivity.

The simulation results regarding the temperatures distribution reached in the microgripper when it is actuated were visually compared with the thermal measurements realized using an IR camera SC5000 from FLIR system (Figure 4). The thermal measurements show that the distribution of the temperatures in the microgripper has a similar map with the simulation results. The temperature values at the tips remains near initial temperature of the environment.

we consider the microgripper with 100 μm the initial opening then the manipulating size is from

An SU-8 Microgripper Based on the Cascaded V-Shaped Electrothermal Actuators: Design, Fabrication, Simulation…

http://dx.doi.org/10.5772/intechopen.75544

31

The fabrication of the microgrippers is based on a three mask process. The OmniCoat stripper

A silicon wafer of any orientation was used after a typical chemical cleaning. A thin layer of 40 nm thickness of Omnicoat was deposited on the silicon wafer by spin-coating and baked at 200C on a hotplate for 1 minute. Then the SU-8 2015 (MicroChem) polymer was deposited on the wafer using a spinner at 4000 rpm in order to obtain a thickness of 9 μm. The wafer was soft-baked at 65C and at 95C for 1 minute and 3 minutes, respectively. The SU-8 layer was then patterned using the first mask in order to obtain the microgripper configuration. After the exposure, the wafer was post-baked at 65C and at 95C for 1 minute and 2 minutes, respectively and then developed using mr-Dev 600 developer. The polymer structure was hardbaked at 185C for 15 minutes in order to complete cross-linking of the SU-8 polymer. The metal layer consists of a sandwich of Cr/Au/Cr films of 10 nm/300 nm/10 nm thicknesses. The metals were evaporated and the heater and pads were obtained using a lift-off process based on AZ4562 photoresist. The second SU-8 2015 layer was obtained using the same settings as for the first layer. In this step the access to the metallic pads was created using the third mask for SU-8. The final thermal process of the polymer in this step was the hard-baking at 195C for 30 minutes for cross-linking of the SU-8 polymer. To release the microgripper structures the Omnicoat layer was developed (Figure 6). The SU-8 and the metallic layers are well patterned (Figure 7). Figure 8 (a) shows an optical picture of the fabricated microgripper before releasing.

A thin layer of Omnicoat was deposited on a silicon wafer by spin-coating as in the first fabrication process. The SU-8 polymer was deposited on the wafer using a spinner in order to

Figure 6. Schematic cross section of the actuator arm after fabrication and release.

50 to 90 μm. A temperature change with only 5C is observed at the microgripper tips.

(MicroChem) is used in order to completely release the final structures [21–24].

4. Fabrication and characterization

4.1. First fabrication process

4.2. Second fabrication process

Figure 4. Simulations and thermal measurements results: (a) FEM coupled electro-thermo-mechanical simulations results of the temperatures distribution in the microgripper at 22 mA (Coventorware 2014 simulation) [22]; (b) the radiation distribution in the microgripper (IR thermography measurements).

Figure 5. FEM coupled electro-thermo-mechanical simulations results: (a) the in-plane deflections at 22 mA [22]; (b) the out-of-plane displacements of the tips vs. electrical current (Coventorware 2014 simulation).

The simulated in-plane and the out-of-plane deflections of the microgripper tips were presented in order to evaluate the opening and the displacements of the gripper arms (Figure 5). The simulation results demonstrate that the tips deflect no more than 0.12 μm in the out-of-plane direction (Figure 5(b)).

The simulation results show that the gripper can work up to a temperature of 165C for a complete closing tips. The gripper can continue to work up to 205C in order to obtain a higher displacement or a higher pressure on the griped micro-object. The results indicate that the polymeric micromanipulators can work at low operation temperatures of the tips and with high in plane displacement. A displacement of 25 μm for each microgripper polymeric arm was obtained at a temperature value of 165C and for a current value of 25 mA. At the tips the temperatures remain of 30–35C close to the settings performed for the initial temperature of the media. The capable manipulating size range of the simulated microgripper is from 1 to 50 μm. If we consider the microgripper with 100 μm the initial opening then the manipulating size is from 50 to 90 μm. A temperature change with only 5C is observed at the microgripper tips.
