2.1

Base metallization process

The process modifications and applications in this study are based off of a low-cost, large-scale, micropattern transfer process recently developed by our group^[6]. A simplified description of the ‘base’ process is shown in Figure 1. A clear transparency film designed for high temperature applications is used as a substrate for an electrically conductive copper paint seed layer, which is deposited using an airbrush (a). Dry film photoresist is laminated on top and patterned using standard lithographic techniques (b). Copper is electroplated through the photoresist mask and the remaining photoresist is stripped, resulting in copper metal microstructures on top of the conductive seed (c). The copper films are 70μm thick. The metal transfer step (d) is distinguished by an infrared assisted transfer process, with the copper structures initially embedded in uncured PDMS. After heating, the PDMS is cured and the copper structures are delaminated from the seed, remaining embedded in the PDMS. This method does not require chemical processing. However, if unwanted seed layer residue is transferred to the copper, it can be rinsed away with methanol if desired. This process, as is shown in the following sections, can be modified to be used as a stamping process and in applications requiring aligned layers and integration with NCPs.

Figure 1.

Base fabrication process. A transparency film is airbrushed with a conductive copper paint layer (a) before photoresist is deposited and patterned (b). Copper is electroplated through the photoresist, which is then stripped before liquid PDMS is poured over the assembly (d). A baking procedure transfers the copper to PDMS and a PDMS encapsulation layer is added (e).

2.2

Modified process: stamping on stretchable fabric

In order to deposit metal layers onto unconventional surfaces, the role of PDMS is converted from that of a substrate to an adhesive layer. In this manner, copper patterns can be ‘stamped’ onto the target substrate with PDMS acting to facilitate delamination from the seed and deposition onto the substrate. This modified process is shown in Figure 2. In (a), the electroplated copper is shown on the sacrificial seed layer, the same as in Figure 1 (c). The difference is in (b), where a thin layer of PDMS, thick enough to encapsulate the electroplated copper, is deposited onto the stamp. Then, the copper stamp is placed on top of the target substrate (c) and the infrared assisted transfer proceeds as before. The result is a copper pattern deposited directly onto the fabric (d). During the transfer and before the PDMS is fully cured, liquid PDMS seeps into the voids of the fabric creating a fully integrated PDMS-fabric hybrid layer with exceptional adhesion.

Figure 2.

A modified process for stamping conductive layers onto wearable fabrics. A thin layer of PDMS is deposited onto a metal pattern stamp (a, b) before being placed on top of the target fabric (c). A baking procedure releases the metal into the PDMS, and cures the PDMS as it integrates with the fabric.

In this study, the target substrate is the stretchable fabric known as Luon™, produced by Lululemon Athletica. This fabric is used due to its popularity and well-known stretchable properties. The deposited curvilinear pattern shown in Figure 3 has been studied previously by our group in stretchable applications, and accommodates the most strain among all tested structures^[23]. Each segment in the curve has a 300μm trace width, an arc angle of 270°, and a radius of curvature of 1.37mm.

Figure 3.

A stretchable interconnect for wearable electronics fabricated using a new stamping process with PDMS integrated directly into the fabric substrate.

2.3

Modified process: aligned layers and NCP integration

The process is further modified to demonstrate the ability to produce aligned and stacked metallization layers and the incorporation of NCPs. This stacking and alignment process is demonstrated in Figure 4. In (a) and (b), an Fe-PDMS NCP post is fabricated on a PDMS substrate using a micromolding process^[24]. In (c) and (d), the electroplated copper stamp (discussed in the previous section) is modified with an alignment hole to facilitate alignment between layers using the corresponding Fe-PDMS alignment post. This alignment hole is laser etched before electroplating the metallization layer. The stamping of a single layer using the alignment post is illustrated in (e) and (f), while (g) shows the result of stacking four aligned metal layers.

Figure 4.

A process for fabricating aligned metallization layer with an integrated NCP feature. The Fe-PDMS NCP alignment post is fabricated via micromolding (a, b), and the metallized stamp is laser etched with a corresponding alignment hole (c, d). A single layer is transferred and aligned via stamping (e, f), which results in a multi-layer device when repeated with successive stamping layers (g).

As mentioned earlier, a single coil in this actuator incorporates finned features to improve heat dissipation and allow for larger currents and magnetic field generation. This design is shown in Figure 5. The trace width, w, and spacing between fins are both 500μm, while the radius, r, is 2.3mm. A relatively large 500μm trace width is chosen in order to reduce the mechanical stress in the coil during magnetic field generation^[20], while the radius is calculated according to the radius-to-trace width ratio that is used in our previous study^[23]. The laser-etched alignment hole in the transparency substrate must have a slightly smaller radius than the electroplated coil (Figure 4(d)), and is designed with 80% of the coil radius to allow a degree of tolerance. The corresponding Fe-PDMS post (Figure 4(b)) must similarly have a smaller radius than the alignment hole, and is also designed to be 80% smaller than the alignment hole.

Figure 5.

Single layer planar coils for magnetic field generation. A ‘clean’ design (a) in contrast with a ‘finned’ design (b) that has incorporated heat sinking features to accommodate larger currents, thus enabling increased magnetic field generation

The completed electromagnetic actuator has four metallization layers and is shown in Figure 6. Note that successive layers are rotated 180° to minimize asymmetries in the design. During the transfer step (Figure 4(e)), pressure is applied to the stamp via a 130g mass placed on top of the stamp during the transfer. Before testing, it is not known what the resulting thickness of PDMS will be between layers, or if electrical isolation is maintained between layers.

Figure 6.

A four-layer electromagnetic actuator incorporating an Fe-PDMS NCP post for increased magnetic field generation and to facilitate alignment between layers.

3.1

Stamping on stretchable fabric for wearable electronics

After fabricating the wearable device in Figure 3, electrical connections are established via copper tape with a conductive adhesive backing prior to testing. The connections are reinforced with a silver-based paint to ensure reliable contact. Before stretching, the structure is confirmed to be highly conductive and has a measured resistance of 1.1Ω. Stretching measurements are collected using a hand-operated stretching apparatus capable of stretching in 250μm increments. In this case, a strain of 21.9% is applied before the device experiences conductive failure. This is comparable to the previous result of 18.3% +/- 3% strain before failure found using the same structure with the base process^[23].

In addition to testing with Luon™, we deposit the metal patterns onto various fabrics including a stretchable crepe fabric (95% polyester, 5% spandex), two non-stretchable drapery linings (100% cotton, and 70% cotton-30% polyester, respectively), and a stretchable constellation knit (100% mixed fibers). This success is promising, and suggests that the process will succeed in many wearable fabrics with ease. These experiments demonstrate a versatile new method for stamping metal layers onto unconventional substrates such as stretchable fabrics, with initial results indicating that this process is well-suited to stretchable and wearable electronics. Conductive patterns are transferred directly to wearable fabrics without losing integrity or becoming less effective compared to those fabricated using the base process.

3.2

Aligned and stacked layers for a high current electromagnetic actuator

Before fabricating the multi-layer actuator, we first simulate the finned and clean coil designs from Figure 5 to confirm that the fins effectively dissipate heat and allow for larger currents, and therefore, larger induced magnetic fields. Previous studies have shown that this approach is indeed effective^[20][21][22], however our goal is to determine the effect on the geometry used in this study. We simulate nine different single-layer coil geometries, with coil conductor widths ranging from 100μm to 300μm and an inner radius ranging from 0.866mm to 2.66mm. Each of the nine geometries is given an increasing applied voltage in a series of steady-state simulations until the maximum temperature within the conductor reaches the melting point of copper. This signals device failure, and the maximum applied voltage for the geometry has been reached. This series of simulations is performed for both the clean and the finned designs. The average increase in applied voltage that the finned designs are able to accommodate is found to be 3.49 times that of the clean designs, thus confirming the enhanced performance in finned designs. An example of the resulting current densities found during simulation is shown in Figure 7, where the finned design (b) is able to accommodate a 0.05V voltage input compared to 0.016V in the clean design (a), resulting in an average current density of 3.6x10⁸ A/m² compared to 1.3x10⁸ A/m².

Figure 7.

Typical current density distribution at steady state for two designs of an example coil geometry (r = 0.886mm, w = 100μm). Within the conducting coil domain, the fin design (b) shows an average current density of 3.6x10⁸ A/m² at 0.05V applied, while the clean design (a) shows an average current density of 1.3x10⁸ A/m² at 0.016V applied.

After simulating the performance of a single layer, we fabricate the multi-layer device as described earlier. To assess the success of the fabrication process, we first test the conductivity of each coil layer and confirm electrical isolation between layers using handheld digital multimeter (DMM). Each layer is found to be effectively isolated, with an open circuit load detected when measuring across different layers. Within the same layer, a resistance of 0.0Ω is measured end to end. The metallization layers are, therefore, well-isolated and highly conductive, and the resistance cannot be measured effectively with the DMM.

We also verify the effectiveness of the alignment method used in this process. Figure 8 shows a single-layer stamp and alignment hole (a), the Fe-PDMS alignment post (b), and the gap g₁ between the top metallization layer and the alignment hole in the assembled device (c). This gap is easily identified because of the residual traces of seed layer that have been transferred to the PDMS during the curing process. This residual seed layer is undesired, but is found to be effectively non-conducive and does not result in electrical shorts across conducting patterns, even between traces spaced as near as 50μm. The gap between the Fe-PDMS alignment post and the stamp alignment hole, g₂, is also easily visible (c). Both gaps are relatively even around the alignment features, indicating an effective alignment process.

Figure 8.

A single-layer stamp with alignment hole (a) and Fe-PDMS alignment post (b). The single-layer stamps are transferred successively to create a well-aligned multi-layer device (c). The gap between the top metallization layer and the alignment hole, g₁, and the gap between the Fe-PDMS alignment post and the stamp alignment hole, g₂, are labelled (c).

Finally, as mentioned earlier, we place a 130g mass on top of the stamp during the transfer step in the fabrication process, which results in an unknown PDMS thickness between metallization layers. In order to determine the PDMS thickness, we measure the height of the Fe-PDMS substrate before transferring any metallization layers and after all four layers have been transferred. The resulting average PDMS thickness between layers is found to be 246μm. Ideally, this thickness would be reduced in order to keep the conductive layers as close as possible while maintaining electrical isolation. This would result in stronger magnetic field generation above the conductor, as the intensity of the magnetic field decreases as the distance to the coil increases.