The study of yeast cell morphology requires consistent identification of cell cycle phases based on cell bud size. A computer-based image processing algorithm is designed to automatically classify microscopic images of yeast cells in a microfluidic channel environment. The images were enhanced to reduce background noise, and a robust segmentation algorithm is developed to extract geometrical features including compactness, axis ratio, and bud size. The features are then used for classification, and the accuracy of various machine-learning classifiers is compared. The linear support vector machine, distance-based classification, and k-nearest-neighbor algorithm were the classifiers used in this experiment. The performance of the system under various illumination and focusing conditions were also tested. The results suggest it is possible to automatically classify yeast cells based on their morphological characteristics with noisy and low-contrast images.

This is the first part of a two-part study on the generation of droplets at a microfluidic T-junction operating in the transition regime where confinement of the droplet creates a large squeezing pressure that influences droplet formation. In this regime, the operation of the T-junction depends on the geometry of the intersection (height-to-width ratio, inlet width ratio), capillary number, flow ratio, and viscosity ratio of the two phases. Here in paper I we presented our experimental observations through the analysis of high-speed videos of the droplet formation process. Various parameters are tracked during the formation cycle such as the shape of the droplet (penetration depth and neck), interdroplet spacing, production rate, and flow of both phases across several T-junction designs and flow conditions. Generally, the formation process is defined by a two-stage model consisting of an initial filling stage followed by a necking stage. However, video evidence suggests the inclusion of a third stage, which we term the lag stage, at the beginning of the formation process that accounts for the retraction of the interface back into the injection channel after detachment. Based on the observations made in this paper, a model is developed to describe the formation process in paper II, which can be used to understand the design and operation of T-junction generators in the transition regime.

This is the second part of a two-part study on the generation of droplets at a microfluidic T-junction operating in the transition regime. In the preceding paper [Phys. Rev. E 85, 016322 (2012)], we presented our experimental observations of droplet formation and decomposed the process into three sequential stages defined as the lag, filling, and necking stages. Here we develop a model that describes the performance of microfluidic T-junction generators working in the squeezing to transition regimes where confinement of the droplet dominates the formation process. The model incorporates a detailed geometric description of the drop shape during the formation process combined with a force balance and necking criteria to define the droplet size, production rate, and spacing. The model inherently captures the influence of the intersection geometry, including the channel width ratio and height-to-width ratio, capillary number, and flow ratio, on the performance of the generator. The model is validated by comparing it to speed videos of the formation process for several T-junction geometries across a range of capillary numbers and viscosity ratios.

The conventional fabrication methods for microfluidic devices require cleanroom processes that are costly and time-consuming. We present a novel, facile, and low-cost method for rapid fabrication of polydimethylsiloxane (PDMS) molds and devices. The method consists of three main fabrication steps: female mold (FM), male mold (MM), and chip fabrication. We use a CO ₂ laser cutter to pattern a thin, spin-coated PDMS layer for FM fabrication. We then obtain reusable PDMS MM from the FM using PDMS/PDMS casting. Finally, a second casting step is used to replicate PDMS devices from the MM. Demolding of one PDMS layer from another is carried out without any potentially hazardous chemical surface treatment. We have successfully demonstrated that this novel method allows fabrication of microfluidic molds and devices with precise dimensions (thickness, width, length) using a single material, PDMS, which is very common across microfluidic laboratories. The whole process, from idea to device testing, can be completed in 1.5 h in a standard laboratory.