Research | Dr. Shankar Dutt

01Method

How a nanopore measurement works

A solid-state nanopore is a hole a few nanometres wide in a membrane of silicon nitride or silicon dioxide. I make the membranes and pores, record the current through them as molecules pass, and analyse the events with software and machine-learning models written for the purpose.

01
Open-pore current
A membrane with a single pore separates two reservoirs of salt solution. A voltage drives ions through the pore, giving a steady current.
02
Capture
A charged molecule near the pore is pulled in by the electric field that the voltage creates.
03
Blockade
While the molecule is inside, it displaces ions and the current drops. The drop lasts as long as the molecule stays in the pore.
04
Analysis
Each event is described by its depth, duration and shape. Distributions over many events carry information about size, shape, charge and conformation.

Fig. 5The four stages of a resistive-pulse measurement.Illustrative schematic, not to scale.

Resistive Pulse Sensing

Interactive · illustrative values

Applied Voltage100 mV

Fig. 6Resistive-pulse sensing: how a translocating molecule changes the ionic current.Interactive explainer with illustrative values, not experimental data.

Ion Current Rectification

Interactive · illustrative values

Applied Voltage+80 mV

Fig. 7Ion current rectification in a conical nanopore.Interactive explainer with illustrative values, not experimental data.

What the method offers

Label-free
No fluorescent tags or chemical modification: molecules are measured in their native state.
Single-molecule
Each event is one molecule, so heterogeneity that averages away in bulk methods remains visible.
Real time
Events are recorded as they happen, which allows dynamic processes to be followed.
Manufacturable
Solid-state membranes are made with semiconductor processes and can be produced at wafer scale.
Robust and tunable
Silicon nitride and silicon dioxide are chemically stable, and pore size and shape can be controlled.
Small samples
Measurements need only small sample volumes, which matters when samples are scarce.

Definitions

02Established results

What has been shown

Three results with a peer-reviewed paper behind each, set out as question, approach, result, limitations and contribution.

2.1Published result

Telling proteins apart by their electrical signatures

Can a nanopore distinguish proteins of similar size without any labels?

Why it matters: Proteins are the most informative markers of disease, but most assays need antibodies or labels chosen in advance. Reading proteins directly, one molecule at a time, could make assays faster and cheaper.
Approach: High-bandwidth recordings through solid-state nanopores; event extraction; features describing each event's depth, duration and shape; supervised machine-learning classifiers trained on labelled events.
Demonstrated result: Label-free identification of similarly sized proteins reached F-values up to 88.7%, with specificity of 96.4%.
Limitations and status: The result concerns purified proteins in buffer under laboratory conditions. It is not a clinical test; extending it to complex samples such as blood is the aim of current work.
My contribution: First author.
Collaborators: H. Shao, B. I. Karawdeniya, Y. M. Bandara, E. Daskalaki, H. Suominen, P. Kluth (ANU).
Paper: Small Methods 7(11), 2300676 (2023)
Related projects: PhD and Honours projects in machine learning for nanopore signals.

Fig. 2The features of a single blockade event used to tell proteins apart: depth, duration and shape.Illustrative redraw of the analysis concept, not experimental data, 2023, doi:10.1002/smtd.202300676.

2.2Published result

Watching an RNA change shape

Does a disease-linked mutation change how a transfer RNA folds?

Why it matters: Misfolded RNAs are implicated in neurodegeneration, and RNA is an emerging drug target. A label-free way to see individual conformations complements ensemble techniques that average them away.
Approach: Translocation of wild-type and mutant tRNA through solid-state nanopores, and analysis of blockade levels and dwell times across many single-molecule events.
Demonstrated result: Single-molecule detection of metastable conformers of a neuron-specific tRNA, whose mutation causes neurodegeneration in mice; these states are invisible to ensemble methods.
Limitations and status: Conformations are inferred from current signatures under the experimental conditions studied; the work does not yet measure drug effects, which remain a direction for future work.
My contribution: First author. The first solid-state nanopore study of RNA conformational dynamics.
Collaborators: L. B. Lai, R. Mehta, B. I. Karawdeniya, Y. M. N. D. Y. Bandara, A. J. Clulow, S. Glatt, V. Gopalan, P. Kluth.
Paper: Nucleic Acids Research 54(2), gkaf1411 (2026)
Related projects: Projects on RNA structure and nanopore spectroscopy.

Fig. 3Two conformations of a tRNA and the distinct current levels they produce.Illustrative schematic of the concept, not experimental data, 2026, doi:10.1093/nar/gkaf1411.

2.3Published result

Membranes that last for millions of events

Can a solid-state pore stay stable long enough to collect the data machine learning needs?

Why it matters: Pore lifetime limits how much data one experiment can yield. Thin, durable membranes improve signal and make large, consistent datasets possible.
Approach: Wafer-scale fabrication of silicon nitride membranes, pore formation, electrical characterisation and long translocation runs with DNA and proteins.
Demonstrated result: Membranes with an effective thickness of about 3 nm supported more than 500,000 DNA and 1.8 million protein translocations through a single pore.
Limitations and status: Event counts refer to the conditions and pores reported in the paper; lifetime in other buffers or with complex samples may differ.
My contribution: First author; featured on the cover of Analytical Chemistry.
Collaborators: B. I. Karawdeniya, Y. M. Bandara, N. Afrin, P. Kluth (ANU).
Paper: Analytical Chemistry 95(13), 5754–5763 (2023)
Related projects: Fabrication and characterisation projects; two patent applications cover related methods.

Cover of Analytical Chemistry, volume 95, number 13, 4 April 2023: DNA strands passing through a pore in a thin membrane between two electrodes. — Fig. 4Cover of Analytical Chemistry 95(13), 4 April 2023, featuring the ultra-thin silicon nitride membrane study.Journal cover artwork, 2023, doi:10.1021/acs.analchem.3c00023.Cover image designed by Mia Kluth; © American Chemical Society.

03In development

Work in progress, and what it is not yet

Applications of the method that are funded, proposed or exploratory. Each states what has, and has not, been demonstrated.

3.1In development
A blood test for early signs of Alzheimer's disease
With Thaum Pty Ltd, I am developing nanopore detection of blood-based biomarkers of neurodegeneration, such as neurofilament light chain (NfL), building on the protein-identification work.
The aim is a rapid, inexpensive and non-invasive test. It has not yet been demonstrated in patient samples; the published results concern purified proteins.
Funded: Australia's Economic Accelerator Ignite grant, AUD 198,767 (2025), as lead entrepreneur.
Thaum Pty Ltd · Industry partner
3.2Proposal submitted
Heavy metals in water
Nanopores functionalised with DNAzymes could report specific metal ions. With ACT water and health agencies we are studying metal release into water after bushfires and floods, and with Rio Tinto, heavy-metal ion sensing.
The aim is field-deployable monitoring. No sensing results are published yet.
Proposals submitted to the Disaster Ready Fund (round 3) and the AEA Ignite program (round 2); a PhD project on DNAzyme-functionalised pores is under way.
ACT Office of Water, Icon Water and ACT Health · Government partner
Rio Tinto · Industry partner
3.3Exploratory
Cardiac biomarkers
The same measurement and classification approach could apply to cardiac markers such as troponin, where a rapid point-of-care test would shorten diagnosis of heart attack.
This is a goal, not a demonstrated capability.
A direction for future work; no results to report.
3.4In development
Signal-analysis software
Custom analysis software for solid-state nanopore data: event extraction, feature calculation and classification, including a parallel extraction framework published in 2024.
Published benchmark: event extraction up to 1120× faster than existing tools, under the conditions reported in Small Methods (2024).
In use by the Advanced Materials Group; public release planned.

04Materials and methods

The physics underneath

Research on how pores form and how ions move through them, which the sensing work depends on, and a secondary line on protective coatings.

Published result

Ion tracks and pore shape

Swift heavy ions leave narrow damage tracks in solids; etching them produces pores. Using small-angle X-ray scattering at the Australian Synchrotron, I measure track and pore geometry directly, including the first quantitative characterisation of pore shape in track-etched polycarbonate.

All papers: Ion tracks & SAXS

Published result

Ion transport in conical nanopores

Conical pores in silicon dioxide with tunable surface charge rectify ionic current. We study how pore geometry and electrolyte properties control transport, with applications in molecular separation and osmotic power.

All papers: Ion transport

In development

Coatings for carbon composites in space

A secondary line of work with New Frontier Technologies: thin coatings, applied with ultrasonic-assisted deposition, that protect carbon-fibre composites against radiation and atomic oxygen in orbit.

Industry collaboration; covered by ANU Science as “Creating sunscreen for satellites” (2025).

Working on a question these methods could answer?

I welcome collaborations on measurement, fabrication and analysis, and enquiries from prospective students.

Propose a collaboration Student projects

How a nanopore measurement works

Open-pore current

Capture

Blockade

Analysis