Research

Vision and Overview

Our group's research program, straddling the fields of nanotechnology and soft-matter/biophysics, is devoted to developing the field of nanofluidics. Nanofluidic devices are networks of fluid-filled structures on a chip, such as nanochannels and nanopores, with dimensions on order of 1-100 nm. Devices with such small dimensions have radically different properties than their macro world analogs: new physics arises as molecular length-scales interact with the imposed “device” confinement, leading to the ability to analyze and manipulate single biomolecules and bioparticles without need for molecular amplification. Simultaneously, cells and viruses are inherently highly structured and confined environments. Artificial nanofluidic structures can model nanoconfinement in biology, enabling confinement to be probed over a wide-parameter space and placing results from in vivo experiments in a global, physical context.

Our research has three core axis:

Development of new types of nanofluidic systems and technology
Application of nanofluidics to individualized analysis of biomolecules and bioparticles
Elucidating the physics and biophysical implications of molecular confinement

Synergistic feedbacks between the three axis are key: enhanced understanding of the physics leads to new device concepts and optimized design while the need to develop single molecule tools with specialized properties leads to probing of new physical regimes.

Some Ongoing Projects

Nanofluidic Pneumatic Control System

We have developed a nanofluidic device that uses pressure driven deflection of a thin membrane lid to trap single macromolecules and nanoparticles in nanoscale compartments (called "molecule in a box"):

Left: upon applying pressure to membrane, membrane deflects, driving macromolecules into nanoscale corrugations where they remain confined.

Top: schematic of pneumatic chip and 3D printed chuck for mounting chip on microscope

We use this approach to explore interactions of macromolecules in confined volumes as an in vitro model for how multiple macromolecules (e.g. bacterial nucleoid, secondary chromosomes, plasmids) interact and self-organize inside bacteria.

DNA in a Box. Fluorescently labeled DNA molecules (one chain colored red, the other green) interact in a nanocavity with increasing anisotropy (ellipticity). In circular cavities, the chains undergo random rotation around the cavity center; in strongly elliptical cavities the chains occupy distinct poles of the ellipse with intermittent pole switching events (from Z. Liu et al., Nat. Com., 2022).

Some Related Publications

X. Capaldi and Z. Liu et al, Probing the organization and dynamics of two DNA chains trapped in a nanofluidic cavity, Soft-Matter 14, 8455 (2018)
Z. Liu et al, Confinement Anisotropy Drives Polar Organization of two DNA Molecules Interacting in a Nanoscale Cavity, Nat. Comm. 13, 4358 (2022)
Z. Liu et al, Characterizing interaction of multiple nanocavity confined plasmids in presence of large DNA model nucleoid, Soft Matter 19, 6545-6555 (2023)

Nanofluidic Technologies for Analysis of Single Extracellular Vesicles

(with Sara Mahshid in Bioengineering)

Extracellular vesicles (EVs) are the body’s ‘Amazon-like’ delivery system—nanoscale, lipid-bilayer packages that carry molecular cargo such as proteins and RNA between cells. As EVs circulate in the bloodstream, reflect their single of cell of origin, and carry disease-related markers, EVs provide a powerful combination of single-cell resolution, diagnostic relevance, and accessibility. Yet, harnessing EVs for diagnostics remains challenging—a consequence of their nanoscale size, biochemical complexity, and extreme heterogeneity. To fully realize the biomedical potential of EV based diagnostics, in this project we develop nanofluidic tools that can perform analysis of individual EVs, obtaining single EV resolved information over large ensembles of EVs. Working in a cancer context, we focus on two specific technology approaches: (1) EV confinement in biochemical reactors for targeted isothermal amplification and (2) performing surface enhanced Raman spectroscopy (SERS) of single EVs in nanophotonic cavity arrays. The latter approach can obtain a SERS-based fingerprint of a given EV that can be harnessed via machine learning to identify cancer and specific cancer mutations/subgroups.

MoSERS Spectral Fingerprinting of Individual EVs. (a) EV containing samples are introduced in device. (b) Single EVs are entrapped in the plasmonic nanocavities with one EV/cavity. (c) The Ag/ZnO/MoS2 design leads to strong SERS enhancement of cavity trapped EVs. (d) SERS spectra of individual EVs are acquired and then (e) classified via machine learning.

Some Related Publications

M. Jalali et al, MoS2-Plasmonic Nanocavities for Raman Spectra of Single Extracellular Vesicles Reveal Molecular Progression in Glioblastoma, ACS Nano 17, 12052-12071 (2023)
I. I. Hosseini et al, Nanofluidics for Simultaneous Size and Charge Profiling of Extracellular Vesicles, Nano Letters 21, 4895 (2021)
I. I. Hosseini et al, Tunable nanofluidic device for digital nucleic acid analysis, Nanoscale 16, 9583 (2024)

Dual Nanopore Device for DNA Feature Mapping and Enhanced Nanostructure Analysis

In a nanopore device, a charged nanoanalyte (e.g. macromolecule, nanoparticle or nanostructure) is driven by an electric field though a small pore in a membrane separating two electrolyte reservoirs. The passing molecule modulates the local ionic flow through the pore, and hence electric current, creating a dynamic single analyte sensing mechanism.

Nanopores offer a purely electrical readout obviating need for fluorescent labeling and reducing assay cost. Solid-state pores, as opposed to protein-based pores, can be made in a larger size range (>10 nm) suitable for interrogating dsDNA with bound molecular features (e.g. antibodies, transcription factors). In addition, solid-state pores can be used to perform electrically based characterization of complex solutions at level of single analytes (one example: to benchmark nanostructure self-assembly in DNA nanotechnology applications).

Top: cartoon of labeled dsDNA translocating through pore (not to scale!)

Bottom: cartoon of current trace for translocating labeled dsDNA

We have a unique dual nanopore technology featuring not just one but two closely separated pores. Coupled to fast active electronics (Field Programmable Gate Array, FPGA), this enables scanning the same analyte multiple times at the two pores, or actually capturing and translocating a single DNA chain at both pores simultaneously. The ability to perform repeated translocations can greatly increase the information extracted from a given single analyte by increasing the number of independent measurements that can be performed on the analyte. In addition, by trapping a dsDNA molecule in a “tug-of-war” mode, with balanced force applied at each pore, the dual-pore can achieve translocation control, slowing down translocation while suppressing folding.

Top: DNA molecule co-captured by two pores undergoing "tug-of-war" control. The pores can independently biased; in tug-of-war mode an opposing bias is applied at the pores, so that the translocation speed is determined by the differential bias. The signal at each pore is determined by the absolute bias. As current can be measured separately at each pore, the time-of-flight of DNA bound tags between the pores can be assessed to obtain in situ measurements of translocation speed.

Bottom: Schematic tug-of-war chip; U-shaped microchannels are etched in a glass substrate, bonded to a thin nitride membrane, and pores are drilled through the membrane at point of closest approach of the microchannels.

(a) Schematic of molecule in dual-pore tug of war; (b) end escapes through pore 1; (c) end escapes through pore 2. (d) Current measured at pore 1 (I1) and pore 2 (I2) showing a molecule in tug-of-war with current blockade (decrease) present at both pores. (e) The molecule end first escapes from pore 1 (leading to increase in current at pore 1) and then escapes from pore 2 (leading to increase in current at pore 2 at later time, due to time taken for molecule end to traverse the distance between the pores).

In collaboration with Deborah Fygenson (Dept. of Physics, University of Santa Barbara), we used our dual-pore system to quantify solutions of very similar cylindrical DNA origami nanostructures, using the dual-pore system to perform repeated translocations through and between the pores.

(W. Dong et al, ACS Nano, 2025)

(a) Schematic showing multiple scans of a single DNA nanostructure at the two pores: (i) first translocation at pore 1 (pore 1 sensing), (ii) second translocation at pore 1 (pore 1 resensing), (iii) structure crosses pore-to-pore separation, (iv) first translocation at pore 2 (pore 2 sensing), (v) second translocation at pore 2 (pore 2 resensing). (b) Current traces at pore 1 and pore 2 over entire resensing process. (c) Magnified view of current trace showing blockade at pore 1 due to first pore 1 translocation, (d) pore 1 resensing and pore 2 sensing (indicating the time-of-flight (TOF) between the pores) and (e) pore 2 resensing.

When coupled to machine learning based classification, the ability to perform repeated independent measurements of a single nanostructure, by increasing the dimensionality of the measurement, enables increased discrimination capability. We also demonstrated that quantification of the time-of-flight between pores, coupled to modeling of transport dynamics, enabled determination of structure size.

(a) Closely related cylindrical DNA Origami structures. (b) Confusion matrix for classification of three structures generated via random forest algorithm trained on dual-pore resensing results. We achieved an accuracy of 83% at 30 kHz bandwidth by using 8 parameters generated via pore 1 sensing, resensing and pore 2 sensing (three scans total, 8 parameters include blockade depth and dwell time at pore 1 and 2, plus TOF and recapture time at pore 1). (c) Accuracy achieved as a function of scan number; single pore sensing achieves an accuracy of only around 50%, demonstrating the power of incorporating multiple independent measurements via resensing.

Some Related Publications

W. Dong et al, DNA nanostructures characterized via dual nanopore devices, ACS Nano, 2025
A. Rand et al, Electronic Mapping of a Bacterial Genome with Dual Solid-State Nanopores and Active Single Molecule Control, ACS Nano 16, 5258-5273 (2022)
X. Liu et al, Flossing DNA in a Dual Nanopore Device, Small 16, 1905379 (2020)
X. Liu et al, Controlling DNA Tug-of-War in a Dual Nanopore Device, Small 15, 1901704 (2019)
Y. Zhang et al, Single Molecule Resensing Using a Two-Pore Device, Small 14, 1801890 (2018)

Google Sites

Report abuse