The Electrochemical Society hosted Prof. John A. Rogers’ live online webinar, “Microfluidic Systems for the Skin: Quantitative Sensing of Biomarkers in Sweat,” on June 23, 2021. Below are answers to questions posed during the presentation.
NOTE: Registration is required to view the webinar.
Professor John A. Rogers is the Louis Simpson and Kimberly Querrey Professor of Materials Science and Engineering, Biomedical Engineering, and Neurological Surgery at Northwestern University, with affiliate appointments in Mechanical Engineering, Electrical and Computer Engineering, and Chemistry. He is also Director of Northwestern’s recently endowed Querrey Simpson Institute for Bioelectronics. Rogers completed an SM in physics and chemistry in 1992, and PhD in physical chemistry in 1995, at the Massachusetts Institute of Technology. He was a Junior Fellow in the Harvard University Society of Fellows from 1995 to 1997; worked at Bell Labs from 1997 to 2002; then served on the faculty of the University of Illinois for 13 years. Rogers received many important awards including a MacArthur Fellowship and membership in the National Academies of Engineering, Sciences, Medicine, Inventors, and the American Academy of Arts and Sciences. Rogers has published more than 750 papers, is a co-inventor on more than 100 patents, and co-founded several successful technology companies.
Q&A
Can this technology be used on dogs?
Perhaps, but unless the fur is shaved away, it may cause significant practical difficulties in establishing interfaces to the skin.
How about sensitivity and cross talk? Will an electrochemical-based system work better than colorimetric?
Yes, sensitivity and selectivity are two main considerations. Colorimetric chemistries can achieve both, at least for certain biomarkers.
Is there a potential problem of cross talk between channels and are analyses downstream affected by analyses upstream?
No, the sweat mixes only diffusively at these slow flow rates and small channel geometries. Volumes of sweat in various different reservoirs do not mix over timescales of relevance. Inertial effects are negligible.
If the channels fill up, but the patch is not removed, what happens?
Then the sweat passes through the outlet port onto the surface of the skin.
Can you comment on challenges/limits to detecting larger bio markers—hormones, cytokines, etc., in sweat using these types of colorimetric or hybrid devices?
Selectivity is the main challenge. Currently, we use ex situ analysis techniques for hormones, cytokines, and certain metabolites and vitamins. Aptamers, antibodies, enzymes, and related approaches appear to have some promise for in situ evaluation—using colorimetric or electrochemical techniques.
Are there plans to combine sweat microfluidics with other sensors like EMG (electromyography), PPG (photoplethysmograph), etc.?
Yes, we’ve published on these types of combined systems, and we will publish further detailed studies in the near future.
How can colorimetric analysis using digital techniques compensate for the environmental light changes in real environments?
We use color calibration markings on the devices themselves to allow calibration for different lighting conditions.
How is quantitative data generated as the concentration of analytes changes with the volume of sweat?
With colorimetric approaches, we use collections of reservoirs designed to fill in a time sequential manner, via strategically located capillary burst valves and/or superabsorbent polymer valves. For exact timing information, we have published approaches that use electrochemical cells activated by sweat, where a subsequent measurement of voltage can be used to determine the time of filling. The most versatile approach is to incorporate active electronics and battery systems directly into the devices.
Are these skin-based micro-devices re-usable?
Yes.
Can this technology be extended to detection of pathogens in the air?
Perhaps. We have done a bit of exploratory work in this direction.
Is this manufacturing process soft lithography?
Yes, but only for the research grade microfluidic devices. For commercial devices with Gatorade, we use a proprietary reel-to-reel process based on molding, laser cutting, screen printing, etc. We use techniques in transfer printing for the heterogeneous integration necessary to produce the soft electronics platforms.
Have you considered measuring cortisol in sweat, as a stress biomarker?
Yes, we published a paper recently in PNAS (Proceedings of the National Academy of Sciences of the United States of America) on this topic.
What are the best solid support materials that can be used as the sensing platform (colorimetric and etc.)?
We like silicones, but other soft elastomeric polymers can be considered.
Are you considering expanding this research to other parts of Africa, e.g. Liberia, Sierra Leone, etc.?
Yes, we are also in South Africa; Mexico, as well. We launch in countries where we have strong partnerships, e.g., Gates (Bill & Melinda Gates Foundation) and Save the Children.
Do you ever need to filter a contaminant (e.g., ions) from fluid to analyze it?
Not to this point—but this is an option that can be considered.
What type of polymer do you use as matrix and why do you need more stretchable materials?
For certain applications we use hydrogels as skin adhesives. We like various formulations of silicones, but we have also used poly (styrene-isoprene-styrene). The Gatorade devices are made of low cost, non-elastomeric but medical grade plastics.
I assume these ultimately saturate and can’t be used for time varying measurement, correct?
Yes, the devices can fill completely, but in general, the designs are such that this does not occur under normal circumstances. Yes, they can be used for time dependent measurements, per an answer to a previous question.
In terms of plasmonic, have you tried to add a light source into the sensors?
Yes, we included a small LED indicator in one case that we published in Nature Communications. Plasmonic or other photonic approaches to signal amplification are interesting to consider. We have active programs in those directions.
What type of computational modelling and simulation is deployed/implemented for the overall project?
We (through collaborations with Prof. Y. Huang here at Northwestern University) model every aspect of the devices—circuit layouts, EM properties, mechanical properties, adhesive characteristics, and microfluidic behaviors. We use modeling to guide design choices.
What about signal delay/loss and electromagnetic interference/artifacts during data acquisition?
Interference can be important for wireless communication, but it’s not a problem with optimized designs (per modeling question above).
Has it received MDR 2017/2045 approval?
We are in the middle of the US FDA (Food and Drug Administration) approval process.
Could you elaborate on biocompatibility of such microfluidic skin sensors (biomaterials use and systemic/adverse reactions induced to some category)?
For skin contacting components of the devices, we use established materials that are already approved for use in bandages and other medical components.
Do you think it is feasible to develop such technology for COVID-19 detection?
Our devices were deployed on health care workers and patients across the medical complex here in Chicago starting in April 2020 and continuing through until about two months ago, for purposes of early symptom detection and syndromic tracking—a total of 500 people and 40,000 hours of recordings or so. We published some of the results in a recent paper in PNAS.
Can electrochemical sensors induce micro-shocks leading to injury to the user group?
Yes, possibly, but not with the designs adopted in our devices.
Epifluidic chloride vs. macroduct chloride?
Same. We published detailed results in a paper in Science Translational Medicine a couple of months ago.
What about signal delay/loss and electromagnetic interference/artifacts during data acquisition?
Not an issue with computationally optimized designs.
What is the minimum sweat volume requirement to obtain an accurate reading on a colorimetric platform?
Hard to say because we haven’t studied this specific question, but probably a volume corresponding to a reservoir with a diameter and depth of ~0.1 mm.
At what flow rate does sweat wick through microfluidic channels?
Though passive perspiration or exercise? Typically exercise, or sweat induced in a sauna or warm water shower. The rate depends on many factors—cross sectional dimensions of the channels, sizes of the inlets, rate of sweating, etc.
Are there any biomarkers where differentiating between left- and right-handed (chiral) molecules would be useful?
Yes, perhaps.
