Touchscreen Surface Warfare— Physics and Chemistry of Antimicrobial Behavior of Ion-Exchanged Silver in Glass

Home , Antimicrobial surface

bulletin cover story Touchscreen surface warfare— Physics and chemistry of antimicrobial behavior of ion-exchanged silver in glass

Glass composition, silver concentration, and ambient conditions control antimicrobial efficacy on ion-exchanged glass surfaces and could reduce bacteria population on glass touchscreens.

By C. Kosik Williams, N. F. Borrelli, W. Senaratne, Y. Wei, and O. Petzold

he ubiquitous use of smart- Tphones, tablets, bank ATMs, and other touchscreens has increased scientific and public concerns associ- ated with bacteria on the surfaces of these devices. Engineering antimicrobial functionality onto the surface of hand- held electronic devices—for example, the cover glass of smartphones—and can address these concerns via bacteria sup- pression. Doing so requires selecting the appropriate antimicrobial agent, developing an understanding of how to incorporate it into the glass, and how the antimicrobial effect is rendered. Extensive literature addresses the role of silver in materials as an antimicrobial agent.1-3 (We adopt the generally accepted finding that the Ag1+ ion is the antimicrobial agent. References to silver will be understood to mean the ion unless otherwise noted.) Here we limit our attention to the antimicrobial phenomenon as it occurs in the particular situation of silver-containing glasses. We will focus exclusively on the physical mechanism of the interaction between glass and microbe, rather than the actual biochem- ical process by which silver causes death.4 The efficacy of antimicrobial action depends on opti- mization of factors such as disposition of the silver in glass (Credit: Corning Inc.) and influence of the ambient environment—in other words, how the silver releases from the glass and is brought The background display shows green-stained E. coli bacteria from the surface of a glass similar to the type used for elec- into contact with the microbe. This article discusses glass tronic device touchscreens. composition, a method of incorporating silver into glass,

20 www.ceramics.org | American Ceramic Society Bulletin, Vol. 93, No. 4 effective concentration of silver on the surface, and effect of ambient temperature and relative humidity. An overall quantitative kinetic model predicting these factors is developed.

Ion-exchange—Pushing ‘silver bullets’ into glass O concentration (wt%) O concentration Molten salt baths for ion-exchange 2 (IOX) were prepared from reagent Ag grade salts: NaNO3, KNO3, and AgNO3 in baths inside a furnace with temperature capability ranging from 300°C to 450°C. We used AgNO –alkali 3 Depth (µm) nitrate mixed baths rather than pure (Credit: Corning Inc.) Figure 1. Electron microprobe profile after silver-ion exchange in a 50 wt% AgNO ·50 AgNO3 for reasons of cost, stability 3 against decomposition, and ability to wt% KNO3 bath for 5 min at 350°C. control the desired silver concentra- composition.8,9 The reduced silver 50 wt% AgNO ·50 wt% KNO bath. tion. Selection of the alkali bath to use 3 3 has two adverse effects: It reduces the However, the relationship of the depends on the specific glass composi- amount of active Ag1+, and it induces steady-state concentration of Ag1+ in tion so as to maintain silver as the only color in the glass, which is an undesired the molten bath [e.g., xAgNO ·(1 – x) diffusing species. 3 trait for certain applications. MNO , where M is an alkali] to that Several measurement techniques 3 One way to incorporate silver in in the glass is not straightforward. It lend themselves well to characterizing glass is to add it as a salt to the batch depends on glass composition and various aspects of the Ag1+ disposi- composition. A more efficient method bath composition, that is, whether the tion. Electron microprobe (EMP), X is to IOX Ag1+ for the alkali using a sil- IOX of silver is for lithium, sodium, or ray photoluminescence spectroscopy ver-containing molten salt bath.10 IOX potassium. The statistical mechanical (XPS), and secondary ion mass spec- costs less because only Ag1+ near the approach used by Araujo7 provides the troscopy (SIMS) techniques measure surface participates in the antimicro- basis for developing a model for IOX in the depth of ions after the IOX pro- bial activity. It has the further advan- this system. We used Araujo’s approach cess. Inductively coupled plasma/mass tages of lessening the coloring effect to calculate silver mole fraction in the spectroscopy (ICP/MS) quantifies the and applying to a variety of available glass as a function of the mole fraction amount of silver leached from the glass commercial glasses, which might be of silver in the bath for a typical alkali into a fixed amount of water for a given of interest for their desirable thermo- aluminoborosilicate glass. It is important time. In addition, ultraviolet–visible mechanical properties. However, not to recognize that this relationship is (UV–VIS) spectroscopy measures the all commercial glass compositions are highly nonlinear and depends on bath relative amounts of Ag1+ in glass from suitable, because good antimicrobial constituents and glass composition. the shift in the UV edge. behavior requires other properties, too, The base glass compositions were as will be explained. alkali aluminoborosilicates, and the Creating a ‘kill zone’—Role of the effective species of silver was Ag1+. The hydrated surface layer glass composition selected must allow Measuring and modeling silver To interpret antimicrobial behav- maximal Ag1+ to be incorporated into concentration profile ior requires understanding the role of the glass, while maintaining the desired The antibacterial action produced ambient environment. Any reaction 1+ glass properties, which are determined by silver ions is a “surface effect.” of the microbe with Ag occurs on largely by the specific application.5 Therefore, a quantitative knowledge of the surface, and there must be a liq- 1+ 2 1+ In general, Ag1+ behaves in oxide surface Ag concentration, in µg/cm uid vehicle to support both the Ag 2 glasses as a network modifier in a or ions/cm , is critical in ascertain- and the microbe to create a reaction similar manner to Na+ or K+.6,7 One ing the effectiveness of antimicrobial zone. This vehicle is the ubiquitous approach to maximize the Ag1+ con- action. Silver in the bulk glass below hydrated layer described in several centration is to melt glass with a high the surface plays no antimicrobial role, references as the hydrogen-bonded alkali concentration and then substi- because it has no access to bacteria on (chemically adsorbed) and physi- 11,12 tute Ag1+ for the alkali. Unfortunately, the surface. Figure 1 shows the EMP sorbed water layers. 13 at high concentrations, Ag1+ tends to profile of silver-ion concentration, Saliba et al., measured water layer expressed as Ag O wt%, versus depth thickness (number of water layers) as a reduce to metal. The extent to which 2 this happens is a function of the glass for IOX glass samples prepared from a function of relative humidity using an

American Ceramic Society Bulletin, Vol. 93, No. 4 | www.ceramics.org 21 Touchscreen surface warfare—Physics and chemistry of antimicrobial behavior . . .

Killing germs—A kinetic model of ent humidity on the antimicrobial antimicrobial function efficacy, we modeled a representative N N To be useful for touchscreen devices, example of silver concentration in Ag B the antimicrobial property must acti- glass. We then use the hydrated layer vate within a practical timeframe. thickness as determined by relative A general kinetic model can help humidity.13 Using the model to pre- understand the antimicrobial results dict antimicrobial efficacy as a func- irrespective of the type of test used, in tion of relative humidity, Figure 4 particular the role of ambient humid- shows that efficacy drops rapidly with ity. The proposed model assumes silver decreasing relative humidity. interacts with the microbe only in the The span of time over which anti- hydrated layer, where both species are microbial function needs to be effec- sufficiently mobile. We can ignore dif- tive varies widely depending on the fusion within the layer, although one application. For example, smartphones could easily add the computational have an average lifespan of 21 months, terms if required. The model is essen- whereas bank automated teller tially a kinetic model and accounts for machines could be in service for up to the amount of Ag1+ at the glass surface 10 years. An advantage of the kinetic model is that it can be used to predict Glass Water Bug and in the hydrated layer, and how it (Credit: Corning Inc.) interacts with bacteria. Figure 3 shows a the lifetime efficacy of glass antimicro- Figure 2. Schematic representation of the typical output of numerical simulation bial activity. Figure 5 shows antimi- interaction of Ag1+ in the hydrated layer using the mathematical model for fac- crobial efficacy of a silver-containing with a microbe. NAg represents the con- tors described in the previous sections antimicrobial glass versus a standard centration of silver that exchanges into to predict the anticipated antimicrobial cover glass without antimicrobial func- the hydrated layer, and Q is the partition property of the glass surface. In other tionality. The number of Escherichia function from Eq. (3). N is the concentra- B words, will silver introduced into the coli (E. coli) bacteria, made visible tion of bacteria, and k is the bacterial with a fluorescent dye, deposited on growth or death rate constant and varies glass ultimately lead to antimicrobial the glass surfaces clearly declined on according to species. activity? More importantly, what factors control antimicrobial efficacy? silver-containing antimicrobial glass infrared (IR) measurement of water on The kinetic model incorporates within two hours, while no perceptible a glass surface under controlled relative various factors (Ag1+ in glass and change was observed on standard glass humidity conditions. the hydrated layer, thickness of the within the same timeframe. Our understanding of the phenom- layer, and interaction of the bacteria Antimicrobial lifetime refers to the enon is that moisture on the surface with Ag+1 in the layer) in a quantita- viability of Ag1+ over an extended peri- equilibrated from ambient air provides tive way to understand antimicrobial od of use. Standardized test protocols the mediating role. We further assume activity. To estimate the role of ambi- evaluate antimicrobial surfaces in terms that Ag1+ near or at the glass surface comes into equilibrium with silver in the hydrated layer, where the interaction with bacteria occurs (Figure 2). We estimated how much Ag1+

from glass equilibrates with that in ) 2 the hydrated layer using a “leaching” approach, where we soaked silver- containing glass in deionized water and measured the amount of silver in the leach solution using ICP/MS detection. We also had recourse to the Araujo chemical potential model7 to relate the amount of silver in the hydrated layer Bacteria (CFU/mL) concentration to that in glass. We now consider the (ions/cm concentration Silver “bath” as the silver-containing glass and the “glass” as the hydrated layer. The Time (min) goal was to ascertain how glass compo- (Credit: Corning Inc.) sition influences phenomena relating to Figure 3. Solution for numerical simulation of the kinetic model for silver-ion concentrations and bacteria levels over a range of times. In this case, relative humidity was 80%. the hydrated layer.

22 www.ceramics.org | American Ceramic Society Bulletin, Vol. 93, No.4 that end, we modeled the amount of silver in glass from an IOX bath and, ultimately, how much silver ends up on the glass surface in the hydrated layer. We determined that explicit understanding of the role of the hydrated layer and how it relates to the ambient environment is critical to evaluate the specific antimicrobial efficacy. Our model also allowed us to estimate the

Predicted bacterial log kill lifetime of antimicrobial activity as a function of Ag1+in glass.

Acknowledgements The authors thank Joseph Schroeder Relative humidity and Advanced Materials and Processing (Credit: Corning Inc.) Figure 4. Kinetic model prediction for antimicrobial efficacy as a function of relative laboratory for sample preparation; humidity. Each "log kill" represents a 90% kill rate, i.e, 10% survival rate. Karl Koch for assistance with the kinetic model; Benjamin Hanson (EMP); Elzbieta Bakowska (ICP/MS) of “bio-burden.”14 There are several ver- Summary in Characterization and Materials sions of bio-burden. What is common Glasses containing silver ions intro- Processing for various characteriza- to all is the application of inoculum duced by IOX can produce a viable tions; and Gary Calabrese and Prantik for a given extended period of time, antimicrobial surface if care is taken Mazumder for helpful discussions and followed by a series of rinses or wipes to understand the physical and chemi- suggestions for the paper. and then another inoculum exposure cal processes that contribute to overall and antimicrobial test. We can use the antimicrobial efficacy. To find an About the authors kinetic model to estimate antimicrobial optimum set of conditions to maximize C. Kosik Williams, N. F. Borrelli, lifetime, including the mathematical antimicrobial efficacy, we must under- W. Senaratne, Y. Wei, and O. Petzold term that regenerates Ag1+ from glass. stand each process in some detail. To are researchers in the Science and Technology Division of Corning International Inc., Corning, NY 14830. Contact Kosik Williams at kosikwilca@ Corning.com.

References 1L. Esteban-Tejada, F. Malpartida, A. Esteban-Cubillo, C. Pecharroman, and J.S. Moya, “The antibacterial and antifungal activity of a soda–lime glass containing silver nanoparticles,” Nanotechnology, 20, 85103 (2009). 2J.S. Kim, E. Kuk, K. Yu, J.-H. Kim, S.J. Park, H.J. Lee, S.H. Kim, Y.K. Park, Y.H. Park, C.-Y. Hwang, Y.-K. Kim, Y.-S. Lee, D.H. Jeong, and M.-H. Cho, “Antimicrobial effects of silver nanoparticles,” Nanomed: Nanotechnol., Biol., Med., 3, 95 (2007). 3Z. Xiu, Q. Zhang, H.L. Puppala, V.L. Colvin, and P.J. Alvarez, “Negligible par- ticle-specific antibacterial activity of silver

(Credit: Corning Inc.) nanoparticles,” Nanolett., 12, 4271 (2012). 4R.B. Thurman and C.P. Gerba, “The Figure 5. Viability of E. coli bacteria, labeld with a fluorescent dye, on standard and molecular mechanisms of copper and silver silver-containing antimicrobial cover glasses after 30 min and 2 h. After 4 h the number ion disinfection of bacteria and viruses,” of bacteria colony-forming units observed fell by more than six orders of magnitude. Crit. Rev. Environ. Control, 18, 295 (1989).

American Ceramic Society Bulletin, Vol. 93, No. 4 | www.ceramics.org 23 Touchscreen surface warfare—Physics and chemistry of antimicrobial behavior . . .

5N. Borrelli, D. Morse, W. Senaratne, Y. Wei, and F. Verrier, “Coated, antimicrobial, chemically strengthened glass and method of making,” US Pat. Application No. 20120034435, 2011. April 28-30, 2015 6H.H. Garfinkel, Phys. Chem., 78, 4175 (1968). Cleveland, Ohio 7R.J. Araujo, S. Litkitvanichkul, and D.C. Allen, “Ion exchange equilibria between The manufacturing tradeshow for advanced glass and molten salts,” J. Non-Cryst. Solids, 318, 262 (2003). ceramic materials and technologies 8S.E. Paje, M.A. Garcia, J. Llopis, and M.A. Villegas, “Optical spectroscopy of silver ion- exchanged As-doped glass,” J. Non-Cryst. 2015 Solids, 318, 239 (2003). exhibition 9S.E. Paje, M.A. Garcia, M.A Villegas, and & sponsorship J. Llopis, “Optical properties of silver ion- opportunities exchanged antimony-doped glass,” J. Non- now open Cryst. Solids, 278, 128 (2000). 10R. Araujo, “Colorless glasses containing ion-exchanged silver,” Appl. Opt., 31, 5221 (1992). 11R.K. Iler, The chemistry of silica. Wiley, New York, 1979. 12L.T. Zhuravlev, “The surface chemistry of amorphous silica: Zhuravlev model”; pp. 1–38 in Colloids and Surfaces A: Physicochemical and Engineering Aspects, Vol. 173. Elsevier, New York, 2000. 13N.A. Saliba, H. Yang, and B.J. Finlayson- Pitts, “Reaction of gaseous nitric oxide with nitric acid on silica surfaces in the presence of water at room temperature,” J. Phys. Chem. A, 105, 10339 (2000). 14“Test method for efficacy of copper alloy surfaces as a sanitizer,” US Environmental Protection Agency, Washington, D.C., 2010. n Ceramics Expo Ceramics Expo offers a comprehensive marketplace for ceramic materials and component manufacturing. The event provides a shop floor for all equipment, products and services used throughout the ceramic supply chain.

Contact our team today to find out how you can ﬁnd your participate. Email us at [email protected] or call us toll free: +1 855 436 8683.

Founding Partner vendors Ceramics Expo establishes a crucial with marketplace for ceramic ceramicSOURCE manufacturing and supply chain products Charlie Spahr, Executive Director, The American ceramicsource.org Ceramic Society

24 www.ceramicsexpousa.comwww.ceramics.org | American Ceramic Society Bulletin, Vol. 93, No.4