MEA Fabrication#

We customize our MEAs using standard cleanroom micro-fabrication techniques. The process is mainly based on photolithography and metal lift-off. On average, the overall fabrication time for each batch of MEA (typically 2~5 wafers are processed per batch) is 2~3 days. Through this process, we have successfully demonstrated fabrication of a range of MEA layouts, with variable electrode sizes from 5 to 30 microns in diameter.

An overview of our fabrication process can be found in the paper [1], where detailed characterizations and demonstrations of our MEAs are also presented. An in-depth fabrication protocol is presented below.

Photolithography#

  1. HMDS Coating

Place fresh/cleaned wafer flat in the oven. HMDS (Hexamethyldisilazane) vapor will be deposited from top, so make sure that nothing is covering the top surface. HMDS is used to enhance the adhesion of the photoresist (PR).

Machine: Yield Engineering HMDS Vapor Prime Oven.

Recipe: run the default recipe for a full cycle.

  1. Spin coating of PR

Spin coat PR right after the deposition of HMDS. Apply AZ5214E PR to cover the area on the wafer where you want the pattern to be.

Machine: Headway PWM32 Wafer Spin Coater

Spin coating recipe:

step 1 - 500 rpm for 5s, ramping up speed 500rpm/s

step 2 - 3000 rpm for 30s, ramping up speed 1000rpm/s

step 3 - terminate with ramping down speed 1000rpm/s

This recipe will create a PR layer of ~1.6um. After the spin coating, to ensure the stability of the PR layer during exposure, apply soft bake at 60 deg C for 2 mins, followed by 1 min of hard bake at 110 deg C.

  1. Maskless Photolithography

The PR coated wafter is then proceeded for photolithography using Heidelberg MLA150, which is a high-speed direct-write photolithography machine that bypasses the need of a dedicated photomask for each design. Comparing with the conventional, masked photolithography, this process greatly reduce the fabrication time, and allows a quick prototyping of custom MEA devices.

Machine: Heidelberg MLA150 Maskless Aligner

Exposure recipe: 375nm laser source, 210 mJ/cm^2 exposure dose.

NOTE: The minimum feature size allowed by Heidelberg MLA150 is 1 micron.

Develop right after the exposure.

Develop recipe: 30s development in 250mL, 1:4 diluted AZ400k developer. Gently swirl around the beaker during development. Dispose the used solution and use fresh developer for each wafer.

After the development, rinse the developed wafered with DI water to remove the excess solution, and air dry with Nitrogen gun. Then examine the pattern under microscope. Check whether the pattern is complete, edges are sharp, and traces have uniform width. If critical defects are found in the pattern, it is likely due to nonuniform coating of PR. In this case, clean the wafer with Acetone and IPA and start the coating procedures again. Lastly, post bake of the wafter is highly recommended. This is found to enhance the stability of the PR layer, and also to greatly reduce the metal lift off time (need to confirm on this point). Post bake is done at 120 deg C for 50s.

Constitutive material deposition and lift-off#

  1. Deposition

Subsequent to the photolithography, we deposit the constitutive material of the MEA onto the exposed wafer. Most of our MEAs are made of two metal layers including 15nm Ti and 85nm Pt. An alternative process is also explored to allow the deposition of ITO (indium tin oxide) to make transparent MEAs.

Machine for Metal deposition: AJA sputter coater 3

Metal recipe:

step 1 - Ti strike (process to establish the plasma)

step 2 - 15nm Ti deposition at 3mT and 200W, with a deposition rate about 0.5 angstrom/s.

step 3 - Pt strike

step 4 - 85nm Pt deposition at 3mT and 50W , deposition rate about 0.67 angstrom/s

NOTE: The deposition rate can change drastically after each maintenance/calibration done by the lab staff. Please always look up to the chart hung next to the machine for the latest numbers and update the recipe accordingly.

Machine for ITO deposition: AJA sputter coater 2

ITO recipe: airflow with Ar and 3% O2. Deposition at room temperature, with pressure of 5mT and power of 80W.

NOTE: Different parameter settings in ITO deposition process can lead to totally different electrical and optical properties of the film. Literature also suggests that this process is very machine dependent. Parameters listed above lead to good quality samples in our fabrication setting. A systematic parameter optimization may still be needed to refine the process.

(ONLY for ITO) Following the ITO deposition, samples need to go through an additional annealing process to enhance the conductivity and transparency.

Machine for annealing: Metal deposition: AJA sputter coater 3

Annealing recipe: run sputter coater 3 without deposition at 450 deg C for 1 hour.

  1. Lift-off Despite different constitutive materials, the same lift-off process should be performed as soon as possible after the deposition. The longer the wait time in between two processes, the harder it gets to remove the unwanted materials. Lift-off can be done by submerging the wafer into acetone for about 20mins.

NOTE:

  1. Acetone is highly evaporative, so use aluminum foil to cover the beaker during the process.

  2. Sonication will greatly promote the lift-off, but should be applied carefully to avoid damage to the pattern. Empirically, we sonicate the beaker about every 5 mins during the lift-off, for about 10s.

  3. Spray acetone over the sample will help to remove some of the hanging pieces.

  4. Warming the acetone (to about 40 deg C) is also suggested to speed up the process.

  5. If there are small pieces still attaching to the wafer, soak cleanroom wipes with acetone to softly wipe it down.

Rinse the sample after the lift-off with DI water and dry it with nitrogen gun. Inspect the wafter under microscope to check the material attachment and confirm that there are no unwanted pieces shortening two channels.

Passivation#

  1. Plasma-Enhanced Chemical Vapor Deposition (PECVD)

With the MEA layout patterned onto the wafter, we then applied a thin layer of silicon-nitride to encapsulate the entire pattern through PECVD.

Machine: Oxford PECVD

Si3N4 recipe: The machine runs deposition recipe by cycles. Each cycle will deposit ~4nm. We run 125 cycles, targeting a total thickness of 500nm.

NOTE: 1.The quality of the deposition is machine/recipe/deposition-rate dependent. Low-stress recipe is recommended for better film stability. 2. SiO2 is often used as passivation materials as well. However, both literature and our experience suggest that Si3N4 performs better against diffusion and detachment when the sample is used for long-term contact with fluid.

Color change of the wafer should be observed after the deposition. A rough estimate of the thickness can be done by looking up the color in the standard color spectrum of Si3N4 on glass. A precise thickness measurement can be done through a profilometer.

  1. 2nd Photolithography

Perform 2nd photolithography after the passivation layer deposition. The goal of this process is to create a layer of PR with openings only on the electrodes and contact pads. Therefore, the Si3N4 on these locations will be etched away in the following process to allow electrical signal detection (electrodes) and connection (contact pads). PR will serves as a mask here to protect the passivation layer elsewhere.

Machine: Heidelberg MLA150 Maskless Aligner

Exposure recipe:

step 1 - align the 2nd layer in the design file on top of the deposited MEA pattern. 4 alignment markers should suffice.

step 2 - expose with 375nm laser source, 210 mJ/cm^2 dose.

Follow the same procedure in the first photolithography for developing and post-baking. It’s critical to check the quality of the alignment after the exposure. You should see a perfect overlap of the 2nd layer on top of the MEA pattern, so that openings (regions having no PR) should only be on the electrodes and contact pads.

  1. Dry etching

The exposed wafer will then be sent for reactive ion etching (RIE) to remove the passivation layer on top of the electrodes and contact pads.

Machine: Oxford Freon RIE

Recipe:

Use 30sccm CF4 as the etching gas and run recipe at 35mT, 90W for 18mins.

To examine the etching condition, use a multimeter to check if the contact pads are completely exposed - placing both probes on top of one contact pad (but no touching each other directly) to see if shortening can be detected. If not, means some SiN still remain on the pad, so additional etching is needed. Check a couple of contact pads at different locations to ensure the etching uniformity. Meanwhile, try not to over-etch, as the PR layer will also be etched by CF4, but at a slower rate. Excessive etching time/power can rip away the PR and damage the passivation layer everywhere.

Cleaning, dicing and post-treatment#

At this point, the main surface structure of the MEA is finished but a set of post-processing procedures is still needed before the MEA can be deployed for cell-seeding.

  1. Cleaning

Clean the MEA with acetone to remove the remaining PR.

  1. Dicing

Depending on the sizes of your wafer and MEA pattern, dicing might be needed. We have explored two main ways to dice the wafer - manually using a wafer cutter or through a wafer dicing machine. Manual cut is quick, but comes at a risk that the wafer may be damaged during cutting and all the processes done before ended up being pieces of broken glasses. Therefore, machine dicing is highly recommended.

Machine: ADT 7122 Wafer Dicer

Recipe: following the machine manual for dicing hard substrates (glasses).

  1. Glue culture well

Glue the culture well (for holding culture media) using Kwik-Sil, a bio-compatible silicone adhesive.

  1. Pt deposition

Platinum-black deposition is recommended to reduce the electrode impedance. Follow a separate protocol to prepare the chloroplatinic acid and preform the deposition. Impedance should be measured before and after the deposition to characterize the electrode property. A impedance below 200kOhm is typical after deposition and ideal for signal detection. Finally, after cleaning the excessive chloroplatinic acid solution, the MEA should be ready for cell seeding.

P.S.

  1. It is always recommended to clean the wafer with acetone and IPA before every major step during the process, or whenever dust/fiber is spotted on the sample.

  2. If the entire process cannot be finished within a couple of days, good stopping points are: after the lift-off (before passivation), after passivation (before 2nd lithography) or after etching.

Acknowledgement: Our fabrication is facilitated by the Materials Research Laboratory and Holonyak Micro & Nanotechnology Lab at University of Illinois at Urbana-Champaign.

Reference: [1] Zhang, Dou, Kim, Upadhyay, Havert, Kang, Kazemi, Huang, Aydin, Huang, Rahman, Ellis-Mohr, Noblet, Lim, Chung, Gritton, Saif, Kong, Beggs, Gazzola. ‘Mind in Vitro’ platforms: Versatile, scalable, robust and open solutions to interfacing with living neurons. bioRxiv, 2023, pp.2023-08