Plasmonic Sensing Studies of a Gas-Phase Cystic Fibrosis Marker in Moisture Laden Air - PubMed (original) (raw)
Plasmonic Sensing Studies of a Gas-Phase Cystic Fibrosis Marker in Moisture Laden Air
Libin Sun et al. Sensors (Basel). 2021.
Abstract
A plasmonic sensing platform was developed as a noninvasive method to monitor gas-phase biomarkers related to cystic fibrosis (CF). The nanohole array (NHA) sensing platform is based on localized surface plasmon resonance (LSPR) and offers a rapid data acquisition capability. Among the numerous gas-phase biomarkers that can be used to assess the lung health of CF patients, acetaldehyde was selected for this investigation. Previous research with diverse types of sensing platforms, with materials ranging from metal oxides to 2-D materials, detected gas-phase acetaldehyde with the lowest detection limit at the µmol/mol (parts-per-million (ppm)) level. In contrast, this work presents a plasmonic sensing platform that can approach the nmol/mol (parts-per-billion (ppb)) level, which covers the required concentration range needed to monitor the status of lung infection and find pulmonary exacerbations. During the experimental measurements made by a spectrometer and by a smartphone, the sensing examination was initially performed in a dry air background and then with high relative humidity (RH) as an interferent, which is relevant to exhaled breath. At a room temperature of 23.1 °C, the lowest detection limit for the investigated plasmonic sensing platform under dry air and 72% RH conditions are 250 nmol/mol (ppb) and 1000 nmol/mol (ppb), respectively.
Keywords: acetaldehyde; cystic fibrosis; humidity; image processing; localized surface plasmon resonance; plasmonic sensing.
Conflict of interest statement
The authors declare no conflict of interest.
Figures
Figure 1
Schematic of nanohole array fabrication. Fabrication steps include: (a) LPCVD depositing silicon nitride layer; (b) EBL and RIE patterning NHA structures; (c) Mask aligner and RIE patterning membrane window; KOH wet etching silicon to create silicon nitride membrane; (d) E-beam evaporator depositing Ti/Au layer; (e) Coating Cu-BTC MOF layer.
Figure 2
Sensing setup. (a) Diagram of the sensing system consisting of a spectrometer, a smartphone, microscope, NHA sensors, MFCs, and dew point generator. (b) SEM image of the bare NHA structure and an image of a single sensor chip versus a quarter. The chip had 4 NHA sectors, each with an area of 300 µm × 300 µm.
Figure 3
(a) Peak wavelength and intensity shift response of the NHA sensor when exposed to varying acetaldehyde concentrations in a dry air background. (b) The NHA sensor response to acetaldehyde for 250 nmol/mol, 500 nmol/mol, and 1 µmol/mol at a wavelength of 583.5 nm. The dotted boxes represent the input periods of the target analyte at each concentration. To help with visualization, a trend curve generated with the Savitzky–Golay method was overlaid. (c) Normalized intensity shifts of NHA sensors when exposed to varying acetaldehyde concentrations. Error bars along the _y_-axis represent ±1 standard deviation for measurement results captured in 5 different testing trials on 4 sensor chips.
Figure 4
Peak wavelength and intensity shifts of the NHA sensor while the RH varies from 0% to 72% with (a) 0 µmol/mol acetaldehyde and (b) 1 µmol/mol acetaldehyde. (c) Responses, shown as difference curves, were obtained by subtracting from the response spectrum of the NHA to 1 µmol/mol of acetaldehyde in air (at each RH level), the response spectrum of 0 µmol/mol of acetaldehyde in air at the corresponding RH levels. (d) Summary of peak wavelength shifts vs. RH levels of each condition in (a,b). For high RH levels (≥40%), the sensor shows linear responses vs. RH levels. The estimated upper RH limit for detecting 1 µmol/mol acetaldehyde is about 95% RH based on the linear curve fitting with 95% confidence interval. The error bars represent ±1 standard deviation for measurement results captured from 4 different testing trials on 2 sensors.
Figure 5
(a) Peak wavelength and intensity shifts of the NHA when exposed to air with 0 µmol/mol acetaldehyde and with 1 µmol/mol of acetaldehyde, both in backgrounds with 60% RH and 72% RH. (b) Response, shown as a difference curve, was obtained by subtracting the spectrum of the NHA in 60% RH air from the spectrum in 72% RH air. (c) Response, shown as a difference curve, was obtained by subtracting the spectrum of the NHA in 0 µmol/mol of acetaldehyde in air with 72% RH from the spectrum in 1 µmol/mol of acetaldehyde in air with 72% RH. Error bars along the _y_-axis represent ±1 standard deviation for measurement results captured in 3 different testing trials from 4 sensor chips. (d) Comparison of the difference curves generated for (b,c).
Figure 6
NHA color variation. (a–c) show raw images of a single NHA when exposed to air with 0% RH, air with 72% RH, and 1 µmol/mol of acetaldehyde in air with 72% RH, respectively. All images were captured by a CMOS imaging device. (d,e) HSV plot of the NHA responses to air with varying RH levels (0%, 60%, and 72%) and to 1 µmol/mol of acetaldehyde in air with varying RH levels (60% and 72%). Error bars along the H-axis, S-axis, and V-axis represent ±1 standard deviation for H, S, and V values of each pixel in each image, respectively. Ellipses on the HS-plane that cover each dataset were generated with 95% confidence.
Figure 7
(a) Comparing NHA sensors’ responses to varying concentrations of acetaldehyde (0 to 1 µmol/mol) and varying RH levels (0%, 20% 40%, 60%, and 72%). (b) Comparison of NHA sensors’ responses to varying RH levels with and without 1 µmol/mol acetaldehyde. (c) Comparison of trend curves in Figure S5d–f. The error bars along the _x_-axis and _y_-axis represent ±1 standard deviation of H and S for data collected from 4 different testing trials on 2 sensor chips, respectively. Please note that for relatively small error values, the data labels may obscure the error bars. (d) PCA result of NHA sensors’ responses to varying RH levels with and without 1 µmol/mol acetaldehyde. The first 2 PCs account for 97.2% (≥95%) of the total variation. Ellipses that cover each dataset were generated with 95% confidence.
References
- Layeghifard M., Li H., Wang P.W., Donaldson S.L., Coburn B., Clark S.T., Caballero J.D., Zhang Y., Tullis D.E., Yau Y.C.W., et al. Microbiome Networks and Change-Point Analysis Reveal Key Community Changes Associated with Cystic Fibrosis Pulmonary Exacerbations. NPJ Biofilms Microbiomes. 2019;5:4. doi: 10.1038/s41522-018-0077-y. -DOI -PMC -PubMed