Gold nanorod arrays enable highly sensitive bacterial detection via surface-enhanced infrared absorption (SEIRA) spectroscopy


Dizaji A. N. , ŞİMŞEK ÖZEK N. , YILMAZ A. , AYSİN F. , YILMAZ M.

COLLOIDS AND SURFACES B-BIOINTERFACES, vol.206, 2021 (Journal Indexed in SCI) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 206
  • Publication Date: 2021
  • Doi Number: 10.1016/j.colsurfb.2021.111939
  • Title of Journal : COLLOIDS AND SURFACES B-BIOINTERFACES
  • Keywords: Surface-enhanced infrared absorption, spectroscopy (SEIRA), Bacterial detection, Machine learning analysis, Gold nanorod arrays, IN-SITU, NUCLEIC-ACIDS, CHEMOMETRICS, IR, IDENTIFICATION, NANOPARTICLES, OPTIMIZATION

Abstract

Infrared (IR) spectroscopy is a unique and powerful method in the identification, characterization, and classification of chemical and biological molecules. However, the low absorbance of biological molecules has arisen as a major bottleneck and inhibits the application of IR in practical applications. To overcome this limitation, in the last four decades, surface-enhanced IR absorption (SEIRA) spectroscopy has been proposed and has become the focus of interest in various applications. In this study, for the first time, we proposed the employment of 3D anisotropic gold nanorod arrays (GNAs) as a highly active SEIRA platform in bacterial detection. For this, GNA platforms were fabricated through an oblique angle deposition (OAD) approach by using a physical vapor deposition (PVD) system. OAD of gold at proper deposition angle (10 degrees) created closely-packed and columnar gold nanorod structures onto the glass slides in a well-controlled manner. GNA platform was tested as a SEIRA system in three different species of bacteria (Escherichia coli, Staphylococcus aureus, and Bacillus subtilis) by collecting IR spectra of each bacteria from different parts of GNA. The employment of GNA provided robust IR spectra with high reproducibility and signal-to-noise ratio. For the comparison, IR spectra of each bacteria were collected from aluminum foil and a smooth gold surface (SGS). No or very low IR spectra were observed in comparison to the GNA platform for these substrates. Unsupervised (PCA, HCA) and supervised (SIMCA, LDA, and SVM classification) machine learning analysis of bacteria spectra obtained from GNA substrate indicated that all bacteria samples can be detected and identified without using a label-containing biosensor, in a fast and simple manner.