Optical Response of PDMS Surface Diffraction Gratings under Exposure to Volatile Organic Compounds

Abstract

Monitoring volatile organic compounds (VOCs) in indoor air is significantly gaining importance due to their adverse effects on human health. Among the diverse detection methods is optical sensing, which employs materials sensitive to the presence of gases in the environment. In this work, we investigate polydimethylsiloxane (PDMS), one of the materials utilized for gas sensing, in a novel transducer: a surface relief diffraction grating. Upon adsorption of the volatile analyte, the PDMS grating swells, and its refractive index changes; both effects lead to increased diffraction efficiency in the first diffraction order. Hence, the possibility of VOC detection emerges from the measurement of the optical power transmitted or diffracted by the grating. Here, we investigated responses of PDMS gratings with varying surface profile properties upon exposure to VOCs with different polarities, i.e., ethanol, n-butanol, toluene, chloroform, and m-xylene, and compared their response in the context of the Hansen theory of solubility. We also studied the response of the grating with a 530 nm deep surface profile to different concentrations of m-xylene, showing a sensitivity and limit of detection of 0.017 μW/ppm and 186 ppm, respectively. Structures in the PDMS were obtained as copies of sinusoidal surface gratings fabricated holographically in acrylamide photopolymer and revealed good sensing repeatability, reversibility, and a fast response time. The proposed sensing technique can be directly adopted as a simple method for VOC detection or can be further improved by implementing a functional coating to significantly enhance the sensitivity and selectivity of the device.

Figure 1

Hansen interaction sphere for PDMS with the VOCs examined in this study.

Figure 2

Dependence of diffraction efficiency in the zeroth and first diffraction orders on height of the sinusoidal PDMS surface grating for refractive index modulation Δn = 0.41 and wavelength of incident light λ = 658 nm. Diffraction efficiency is independent of the grating’s spatial frequency.

Figure 3

Change of diffraction efficiency upon modulation of refractive index and PDMS swelling for gratings with an initial depth of (a) 120 and (b) 530 nm. Surface relief amplitude change is due to swelling of the material.

Figure 4

Setup for VOC exposure. The investigated grating is placed in a small sample chamber, connected to the gas development container with the analyte, vacuum pump, and ambient air. The optical power from the 658 nm laser, diffracted into the zeroth and first orders, is collected with photodetectors.

Figure 5

Atomic force microscopy image of the grating with a 490 nm height and a 8 μm period. The grating is a PDMS copy of the sinusoidal surface relief grating fabricated in acrylamide photopolymer.

Figure 6

Optical power change in the first order of diffraction for gratings with different depths of the surface profile after exposure to VOCs. The change is determined as the difference in power measured when the sample, after exposure to vacuum, was exposed to ambient air and to the analyte. Error bars were estimated based on the mechanical stability of the sensor, and when more than one measurement cycle was performed, the standard deviation from all responses was added.

Figure 7

Optical power shifts observed in the (a, b) zeroth and (c, d) first orders of diffraction during cycles of exposure to vacuum and sample exposure to ambient air or vapor toluene. The plots in (a, c) depict the response of the grating with a 120 nm depth, while the plots in (b, d) were obtained with the grating with a 530 nm depth.

Figure 8

Cycles of evacuating and exposure to air and m-xylene of grating with a 530 nm deep surface profile, monitored in the first diffraction order. Concentration of m-xylene was 275 ppm.

Figure 9

Optical response of grating with a 530 nm depth to varying concentrations of m-xylene. The linear fit was extrapolated to calculate the limit of detection. Error bars are the standard deviation estimated from mechanical instability during repeating vacuum–air cycles, which is 1.44 μW.