skip to main content


Title: Machine learning enables design automation of microfluidic flow-focusing droplet generation
Abstract

Droplet-based microfluidic devices hold immense potential in becoming inexpensive alternatives to existing screening platforms across life science applications, such as enzyme discovery and early cancer detection. However, the lack of a predictive understanding of droplet generation makes engineering a droplet-based platform an iterative and resource-intensive process. We present a web-based tool, DAFD, that predicts the performance and enables design automation of flow-focusing droplet generators. We capitalize on machine learning algorithms to predict the droplet diameter and rate with a mean absolute error of less than 10μm and 20 Hz. This tool delivers a user-specified performance within 4.2% and 11.5% of the desired diameter and rate. We demonstrate that DAFD can be extended by the community to support additional fluid combinations, without requiring extensive machine learning knowledge or large-scale data-sets. This tool will reduce the need for microfluidic expertise and design iterations and facilitate adoption of microfluidics in life sciences.

 
more » « less
NSF-PAR ID:
10208579
Author(s) / Creator(s):
; ; ; ; ; ; ;
Publisher / Repository:
Nature Publishing Group
Date Published:
Journal Name:
Nature Communications
Volume:
12
Issue:
1
ISSN:
2041-1723
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    We present a new, robust three dimensional microfabrication method for highly parallel microfluidics, to improve the throughput of on-chip material synthesis by allowing parallel and simultaneous operation of many replicate devices on a single chip. Recently, parallelized microfluidic chips fabricated in Silicon and glass have been developed to increase the throughput of microfluidic materials synthesis to an industrially relevant scale. These parallelized microfluidic chips require large arrays (>10,000) of Through Silicon Vias (TSVs) to deliver fluid from delivery channels to the parallelized devices. Ideally, these TSVs should have a small footprint to allow a high density of features to be packed into a single chip, have channels on both sides of the wafer, and at the same time minimize debris generation and wafer warping to enable permanent bonding of the device to glass. Because of these requirements and challenges, previous approaches cannot be easily applied to produce three dimensional microfluidic chips with a large array of TSVs. To address these issues, in this paper we report a fabrication strategy for the robust fabrication of three-dimensional Silicon microfluidic chips consisting of a dense array of TSVs, designed specifically for highly parallelized microfluidics. In particular, we have developed a two-layer TSV design that allows small diameter vias (d < 20 µm) without sacrificing the mechanical stability of the chip and a patterned SiO2etch-stop layer to replace the use of carrier wafers in Deep Reactive Ion Etching (DRIE). Our microfabrication strategy allows >50,000 (d = 15 µm) TSVs to be fabricated on a single 4” wafer, using only conventional semiconductor fabrication equipment, with 100% yield (M = 16 chips) compared to 30% using previous approaches. We demonstrated the utility of these fabrication strategies by developing a chip that incorporates 20,160 flow focusing droplet generators onto a single 4” Silicon wafer, representing a 100% increase in the total number of droplet generators than previously reported. To demonstrate the utility of this chip for generating pharmaceutical microparticle formulations, we generated 5–9 µm polycaprolactone particles with a CV < 5% at a rate as high as 60 g/hr (>1 trillion particles/hour).

     
    more » « less
  2. Abstract

    We present analysis using a citizen science campaign to improve the cosmological measures from the Hobby–Eberly Telescope Dark Energy Experiment (HETDEX). The goal of HETDEX is to measure the Hubble expansion rate,H(z), and angular diameter distance,DA(z), atz= 2.4, each to percent-level accuracy. This accuracy is determined primarily from the total number of detected Lyαemitters (LAEs), the false positive rate due to noise, and the contamination due to [Oii] emitting galaxies. This paper presents the citizen science project, Dark Energy Explorers (https://www.zooniverse.org/projects/erinmc/dark-energy-explorers), with the goal of increasing the number of LAEs and decreasing the number of false positives due to noise and the [Oii] galaxies. Initial analysis shows that citizen science is an efficient and effective tool for classification most accurately done by the human eye, especially in combination with unsupervised machine learning. Three aspects from the citizen science campaign that have the most impact are (1) identifying individual problems with detections, (2) providing a clean sample with 100% visual identification above a signal-to-noise cut, and (3) providing labels for machine-learning efforts. Since the end of 2022, Dark Energy Explorers has collected over three and a half million classifications by 11,000 volunteers in over 85 different countries around the world. By incorporating the results of the Dark Energy Explorers, we expect to improve the accuracy on theDA(z) andH(z) parameters atz= 2.″4 by 10%–30%. While the primary goal is to improve on HETDEX, Dark Energy Explorers has already proven to be a uniquely powerful tool for science advancement and increasing accessibility to science worldwide.

     
    more » « less
  3. Abstract

    Advances in microfluidic technologies rely on engineered 3D flow patterns to manipulate samples at the microscale. However, current methods for mapping flows only provide limited 3D and temporal resolutions or require highly specialized optical set-ups. Here, we present a simple defocusing approach based on brightfield microscopy and open-source software to map micro-flows in 3D at high spatial and temporal resolution. Our workflow is both integrated in ImageJ and modular. We track seed particles in 2D before classifying their Z-position using a reference library. We compare the performance of a traditional cross-correlation method and a deep learning model in performing the classification step. We validate our method on three highly relevant microfluidic examples: a channel step expansion and displacement structures as single-phase flow examples, and droplet microfluidics as a two-phase flow example. First, we elucidate how displacement structures efficiently shift large particles across streamlines. Second, we reveal novel recirculation structures and folding patterns in the internal flow of microfluidic droplets. Our simple and widely accessible brightfield technique generates high-resolution flow maps and it will address the increasing demand for controlling fluids at the microscale by supporting the efficient design of novel microfluidic structures.

     
    more » « less
  4. null (Ed.)
    A multifunctional microfluidic platform combining on-demand aqueous-phase droplet generation, multi-droplet storage, and controlled merging of droplets selected from a storage library in a single integrated microfluidic device is described. A unique aspect of the technology is a microfluidic trap design comprising a droplet trap chamber and lateral bypass channels integrated with a microvalve that supports the capture and merger of multiple droplets over a wide range of individual droplet sizes. A storage unit comprising an array of microfluidic traps operates in a first-in first-out manner, allowing droplets stored within the library to be analyzed before sequentially delivering selected droplets to a downstream merging zone, while shunting other droplets to waste. Performance of the microfluidic trap is investigated for variations in bypass/chamber hydrodynamic resistance ratio, micro-chamber geometry, trapped droplet volume, and overall flow rate. The integrated microfluidic platform is then utilized to demonstrate the operational steps necessary for cell-based assays requiring the isolation of defined cell populations with single cell resolution, including encapsulation of individual cells within an aqueous-phase droplet carrier, screening or incubation of the immobilized cell-encapsulated droplets, and generation of controlled combinations of individual cells through the sequential droplet merging process. Beyond its utility for cell analysis, the presented platform represents a versatile approach to robust droplet generation, storage, and merging for use in a wide range of droplet-based microfluidics applications. 
    more » « less
  5. Abstract

    Condensation is ubiquitous in nature and industry. Heterogeneous condensation on surfaces is typified by the continuous cycle of droplet nucleation, growth, and departure. Central to the mechanistic understanding of the thermofluidic processes governing condensation is the rapid and high‐fidelity extraction of interpretable physical descriptors from the highly transient droplet population. However, extracting quantifiable measures out of dynamic objects with conventional imaging technologies poses a challenge to researchers. Here, an intelligent vision‐based framework is demonstrated that unites classical thermofluidic imaging techniques with deep learning to fundamentally address this challenge. The deep learning framework can autonomously harness physical descriptors and quantify thermal performance at extreme spatio‐temporal resolutions of 300 nm and 200 ms, respectively. The data‐centric analysis conclusively shows that contrary to classical understanding, the overall condensation performance is governed by a key tradeoff between heat transfer rate per individual droplet and droplet population density. The vision‐based approach presents a powerful tool for the study of not only phase‐change processes but also any nucleation‐based process within and beyond the thermal science community through the harnessing of big data.

     
    more » « less