Creating self-cleaning surfaces with smart materials

Andreea Danielescu
3 min readSep 6, 2023

We’ve talked before on this blog about our vision for a fut ure with smart materials, and how we were beginning to bring that future to life within Accenture Labs . We’re making progress through several efforts that combine in-house research with academic partnerships. We’re excited to share an update on one area of focus: self-cleaning materials.

S urfaces in many public spaces are cleaned infrequentl y, and thorough cleaning is time-consuming; Metro Transit for Minneapolis/St. Paul reported that cleaning a transit bus takes about three hours from start to finish . This is a key problem in the challenge of helping people safely return to offices and other high-traffic spaces.

For now, most cleaning efforts involve multiple crews of people revisiting high-touch surfaces with disinfectants on a regular schedule. Even then, some surfaces resist easy cleaning, like fabrics that are still used on some public transit vehicles ‘ seats. But what if surfaces could clean themselves?

In fact, they can. Researchers were inspired by the cleaning potential of ultraviolet light, which disinfects surfaces thoroughly and has been used to sterilize surfaces in medical settings for years. But the wavelength of ultraviolet light used for cleaning is harmful to humans, and can only be used in spaces where no humans are present. That rules it out as a viable approach for most public spaces. Instead, researchers examined other light sources with cleaning potential — including visible light.

It turns out you don’t need ultraviolet light to clean surfaces. Visible light can also do the job when you combine it with a second step : treating a surface with a nanoparticle coating . To make the coating, a small concentration of manganese is introduced to t itanium dioxide (TiO 2 ) nanoparticles , creating manganese-doped TiO 2 . Titanium dioxide is a naturally occurring , biologically inert material that is widely available, and has been in use in paints, inks, and similar applications for decades. Manganese-doped TiO 2 can be safely used to coat surfaces ranging from metals to fabrics without changing the feel or function of the surfaces. When visual light hits the coated surface, it kicks off a chemical reaction that attacks organic compounds — destroying everything from food crumbs to bacteria.

This kind of cleaning is slower than what ultraviol et light can achieve, but it has a big advantage over UV: cleaning can occur continuously. As long as light keeps hitting the treated surface, the surface can continue self-cleaning.

So how can we make sure that visible light keeps hitting the surface? We’re partnering with Advanced Functional Fabrics of America (AFFOA) to develop tech that will let us embed light-emitting diodes — LEDs — directly into fibers. With this approach, a n LED- embedded fabric could supply the light that’s needed to kick off its own self-cleaning. For hard, non-fabric surfaces like railing, we’re considering other approaches; metal railings , for example, could have micro-perforations that let light from LEDs inside the railing disperse across the surface; or railings could be made of transparent materials that let light refract through the whole railing.

W e’re currently testing the effectiveness of TiO 2 nanocoating , using E .coli to determine antibacterial effectiveness. We’re also testing the approach of embedding LEDs into fabric . Based on our results, we’ll continue to refine these approaches for durable, affordable self-cleaning materials — which will no doubt become a key element in the future reimagination and design of public spaces. Stay tuned for more updates on this and other projects from our Future Technologies group!

We would like to thank the MIT Media Lab, the University of Colorado Boulder, and the Boston Forge for their partnership in this work. For more information about our smart materials research and our Future Technologies R&D group, contact Andreea Danielescu and Alex Kass.

Originally published at on September 29, 2020



Andreea Danielescu

Passionate about the the intersection of HCI x AI, Conversational interfaces, AR/VR, and gesture.