In order to automate the output of a microfluidic device, we have divided our goals into three main components:
To learn more about these components, explore the tabs below!
- Microfluidics: A microfluidic chip designed to execute a desired biological experiment.
- Hardware: A low-cost, do-it-yourself active XY-translational stage that selectively dispenses the output of the microfluidic chip to a 96-well plate, as well as automated control syringes.
- Software: A software interface that will allow the user to detail the parameters of the experiment run on the chip; the specific location per output on the 96-well plate; and the amount of each output dispensed.
To learn more about these components, explore the tabs below!
Microfluidic chips are engineered devices that manipulate fluid flow through microchannels, which are typically molded or engraved into a material. Due to the small size of the channels, turbulent flow is minimized, allowing scientists to accurately direct fluids through the chip. The most common method of fabricating microfluidic devices is soft lithography, which creates channels by casting a polymer, such as polydimethylsiloxane (PDMS), onto a mold and curing the polymer. This, however, involves high costs due to the need to create a master mold for every chip design and the materials needed to create and cure the PDMS chip. In order to make microfluidics more accessible to synthetic biologists, we have adopted the Boston University CIDAR Lab's low-cost manufacturing protocol for microfluidic devices.
BU CIDAR Lab has created a manufacturing workflow, MakerFluidics, which allows scientists to create microfluidic devices by milling microchannels into polycarbonate using a desktop CNC mill.
These microfluidic devices are comprised of two layers of polycarbonate, the flow layer and control layer, which sandwich a thin layer of PDMS. The polycarbonate layers contain etched features that allow for the manipulation of fluid through the device and the PDMS allows for the actuation of valves, which selectively direct fluid flow. Typically, the fabrication of these devices ranges from one to two hours for milling, depending on the complexity of the design, and another hour for assembly. Since these microfluidic devices are made from polycarbonate and PDMS, both of which are affordable and widely available, and require only a desktop CNC mill to fabricate, scientists are able to rapidly prototype low-cost microfluidic chips without sinking time and money into conventional manufacturing methods.
- TERRA Adapter
The TERRA Adapter serves to increase the number of and types of microfluidic chips that are compatible with our system. This makes the system more user-friendly, as scientists then don’t need to design a chip that both suits the experiment and is capable of selecting for the proper output to be dispensed.
The TERRA Adapter is a scalable, predesigned chip capable of selectively dispensing the output of the microfluidic chip into the intended lab vessel. For droplet based experiments, for every output, the TERRA Adapter has two valves and a set of four pressure regulator. For continuous flow experiments, there are two valves for every output and an extra output that serves as a flush mechanism for cleaning the nozzle between outputs.
Click here to learn more about the TERRA Adapter:
To ensure TERRA is low-cost and accessible to synthetic biologists, we have designed a hardware system consisting primarily of 3D-printed parts and inexpensive components. All of the CAD and STL files are available to the public through GitHub.
- XY-Translational Stage
In order to dispense the output of microfluidic chips into specific locations, we decided to incorporate a low-cost, DIY XY-translational stage. The XY-stage enables translational motion along the X and Y axes, allowing the system to move a vessel, such as a 96-well plate, to a predetermined location.
The XY-stage consists of 3D-printed supports, stepper motors, timing belts, pulleys, bearings, and stainless steel rods. The stepper motors are connected to a timing belt and pulley system, which converts the rotational motion of the motor shaft to translational motion of the timing belt. The timing belts are then attached to supports that travel linearly along steel rods using ball bearings. This enables the main support for a 96-well plate to move in both the X- and Y-direction.
Click here to learn more about the XY-stage:
- Syringe Pumps
We have adopted the design of syringe pumps from the 2016 BostonU_HW Team, Neptune. Neptune originally utilized these pumps to input fluid to microfluidic devices, but we have integrated them to TERRA in order to actuate control valves on microfluidic devices.
Actuating valves of CIDAR Lab microfluidic devices typically requires manually applying a negative pressure to control ports and the PDMS layer using syringes. This then requires the scientist to stay by the microfluidic chip throughout the entire experiment and allows for human error. By utilizing syringe pumps, TERRA allows scientists to automate experimental processes and reduces error.
The syringe pumps are comprised of 3D-printed supports and servo motors. The 3D-printed supports include mounts for servo motors and syringes, as well as an arm that attaches the servo motors to the syringes. The arm allows for the conversion of rotational motion of the motor to translational motion of the syringe plunger. This system enables scientists to automate actuation of valves, thereby controlling the selection of output fluids from a microfluidic device.
Click here to learn more about how we integrated and improved on Neptune's syringe pumps:
TERRA includes both front-end and back-end software to collect data, process data, and control hardware accordingly. The front-end software, the UI, is where the user will directly interact with TERRA while the back-end software is responsible for controlling the XY Plane and Control Syringes.
- User Interface
The UI is an intuitive web application that enables the user to interact with TERRA. The interface is designed so that it takes the necessary experiment information from the user and sends it to the device. The user will enter how many outputs their microfluidic chip has, the flow rate on the microfluidic chip, and properties about the fluid itself (density, continuous vs droplet, etc.). Given this information we are able to use our TERRA droplet predictive model to calculate a dispense time. The user can then select which wells each output should be dispensed into. This information is sent to the Arduino and enters the back-end software components. The video below demonstrates how an experimenter would use the UI.
The UI is able to facilitate communication between the user and TERRA through the use of Peripheral Manager, a daemon that runs on the workstation and facilitates communication between cloud-based applications and hardware connected to the workstation. This communication link uses WebSockets, a communication protocol that enables two way communication between the browser and server. Peripheral Manager takes advantage of this protocol for serial communication with the Arduino.
- Back-end code
- The back-end software runs both the XY Plane and Control Syringes. The approach in designing the back-end was to treat information in the context of a series of different outputs coming off of the microfluidic chip. By creating an “Output” class, we created different methods that were responsible for assigning each output a set of dispensing coordinates and pair of corresponding control syringes that open and close the output. We are then able to access each output to get the necessary information to control the hardware. The control syringe and xy-plane software are detailed further in the Software section of our Wiki.