Fully integrated optofluidic SERS platform for real-time and continuous characterization of airborne microorganisms

Highlights

• A fully integrated optofluidic SERS platform was developed for monitoring bioaerosols.

• Airborne particles were collected continuously into a liquid suspension by inertia.

• Collection efficiency as high as ~ 99.6% was achieved at a particle diameter of 1 μm.

• Total bacteria aerosol concentrations were obtained in real time using the SERS platform.

Abstract

An optofluidic, surface-enhanced Raman spectroscopy (SERS) platform was developed to detect airborne microorganisms, continuously and in real time. The platform consists of an on-chip analysis system integrated with an aerosol sampler and Raman spectrometer. A stratified two-phase flow, consisting of the sampled air stream and a stream of collection medium, is formed in the curved channel. The inertia of collected particles, such as bacterial cells, carries them across the phase boundary in the curved channel such that they impact the liquid stream directly. The collection efficiency of the microchannel was evaluated using different-sized standard polystyrene-latex particles. A collection efficiency of 99.6% was attained for particles with an average aerodynamic diameter of 1 μm, a typical size for bacterial aerosols, by optimizing the flow rates of the sample air and liquid medium. A silver colloid in the collection medium was used as the SERS adsorbent. After passing through a serpentine mixing channel, bacterial particles were detected by SERS in real time using the custom Raman spectroscopy system. The detection system was evaluated with five test bacteria: S. epidermidis, M. luteus, E. hirae, B. subtilis, and E. coli. The concentration of airborne S. epidermidis corresponded to a Raman peak at 732.5 cm⁻¹. The limit of detection was approximately 10² CFU/mL and the total bacterial aerosol concentration was determined in real time based on the ratio of sampling air to SERS colloid.

Introduction

Bioaerosols such as viruses, bacteria, and fungal spores are involved in several human health problems (Burge, 1995; Douwes et al., 2003). Due to their small size (~0.3–10 μm), airborne bacteria can remain suspended in air for a long time, allowing them to spread over a wide area within a short period.

Effective collection of fine particles is a prerequisite for rapid detection of microorganisms in air. Various aerosol sampling methods, including gravitational sedimentation, inertial force-driven impaction, centrifugal collection, and air filtration, have been investigated for the collection of bioaerosols (Choi et al. 2015b, 2017a; Kulkarni et al., 2011). Unlike abiotic airborne particle sampling, maintaining the viability of microbial particles is essential when collecting microorganisms. Variables such as humidity, sampling media, and temperature can significantly affect the biological properties of collected bioaerosols. Microorganisms collected under harsh conditions, such as through impaction into a dry medium, may be physically damaged and/or metabolically stressed; this can negatively affect their viability and genomic characteristics (King and McFarland, 2012).

In addition to viability concerns, collection based on direct impaction onto a solid substrate is usually accompanied by sample extraction and suspension in a suitable buffer solution prior to analysis. Direct aerosol-into-liquid impaction avoids the need for sample extraction and resuspension, without degrading bioaerosol viability (Jung et al., 2009; King and McFarland, 2012; Lighthart and Tong, 1998). Therefore, direct collection of bioaerosols into a liquid medium is preferable to collecting on a solid plate, e.g., using an impactor or filtration (Choi et al., 2017b). Wet sampling processes, such as wet cyclone and BioSampler (SKC Inc., Eighty Four, PA, USA) systems, have been used for collecting airborne particles into a liquid medium (Cho et al., 2020; Jung et al., 2009; King and McFarland, 2012; Lighthart and Tong, 1998). However, the transfer of suspended samples to the analysis system remains a bottleneck for real-time monitoring systems (Tan et al., 2011).

Colony counting is a representative method for quantifying bioaerosol concentrations. However, this method is based on traditional cell culturing processes that often require more than 24 h for sufficient microbial growth. Colony counting techniques are also unsuitable for many species (Chen and Li, 2005). Another popular method for determining microbial concentrations is microscopy with a hemocytometer. Although both colony counting and microscopy methods are relatively simple, the processes are time-consuming and dependent on operator proficiency.

Bioaerosols can be qualitatively detected and analyzed using nucleic acid assays and immunological methods. Polymerase chain reaction (PCR) and enzyme-linked immunosorbent assays (ELISAs) are widely used methods that boast high sensitivity and specificity (Alvarez et al., 1995; Wu et al., 2012). However, these methods require at least four to five steps, including sample pretreatment, which should be performed by well-trained operators; thus, they are unsuitable for real-time monitoring of bioaerosols in the field (Fronczek and Yoon, 2015).

Several optical sensor-based techniques have been employed for real-time monitoring of biological particles. Auto-fluorescence from metabolites and various components of living cells can be excited by ultraviolet (UV) light (Bao et al., 2008). Laser-induced fluorescence (LIF) particle detectors are widely used for instant detection of hazardous biomaterials (Kaye et al., 2000; Lee et al., 2010a; Pinnick et al., 1995). However, such laser-based systems typically require a very precise setup with expensive and bulky optical components, which represents a significant obstacle to the development of integrated, portable analysis systems. Furthermore, UV-LIF systems, i.e., ultraviolet aerodynamic particle sizers (APSs) and aerosol fluorescence sensors, often overestimate the concentration of airborne biomaterials due to the fluorescence of non-biological particles (Jung and Lee, 2013; Pinnick et al., 1999).

Recently, portable microfluidic-based devices for high-sensitivity detection of microorganisms and fine particles have been developed to overcome the limitations of the conventional techniques described above (Bhagat et al., 2010; Chung et al., 2013; Kim et al., 2009; Song et al., 2011). As a cell counting method, single-cell detection based on flow cytometry and target aptamer-conjugated fluorescent nanoparticles was developed for real-time cell counting (Chung et al., 2015). Choi et al., 2015a, Choi et al., 2015b reported a micro-optofluidic platform for quantitative analysis of bioaerosols (Choi et al., 2015a). However, studies for simultaneous sampling and analysis using a microchip-based platform appear to be in the infant stage.

Raman spectroscopy is a label-free, non-destructive method that employs a laser to excite vibrational modes in the sample material. Raman scattering provides an intrinsic fingerprint as a function of the chemical composition of the sample, and is used widely for molecular and cellular analyses. However, compared to Rayleigh scattering, Raman scattering is intrinsically weak, yielding approximately 1 photon per 10⁶ incident photons (Chen and Choo, 2008). Here, we demonstrate a micro-optofluidic platform for real-time detection and quantitative analysis of airborne microorganisms using a detector based on Raman spectroscopy. To overcome the low intensity of conventional Raman scattering, we exploited the plasmonic resonance effects of silver nanoparticles (AgNPs) and SERS to amplify the Raman signal by a factor up to ~ 10¹¹ (Moskovits, 1978).

Our optofluidic SERS platform involves the following steps: (1) sampling airborne microorganisms directly into a SERS-active colloid, (2) mixing of the sampled particles and AgNPs in a microchannel, allowing the AgNPs to bind to the cell wall of any bacterial particles, and (3) real-time detection and analysis of microorganisms based on the SERS signal. A simple curved channel was used to collect airborne particles by inertia and drag forces. The efficiency of the particle sampler was determined using standard polystyrene-latex (PSL) particles and five different microorganisms: S. epidermidis, M. luteus, E. hirae, B. subtilis, and E. coli. The optofluidic SERS platform ensures simultaneous airborne particle collection and mixing with SERS-active nanoparticles in a single microfluidic channel. A custom Raman system was used to obtain SERS spectra from the bacteria. S. epidermidis concentrations from 10² to 10⁵ colony forming units (CFU)/mL were correlated to the SERS intensity of specific peaks. Real-time detection and quantification of S. epidermidis were performed in our laboratory using the airborne bacteria generation system. This paper introduces a strategy for real-time, highly sensitive SERS analyses with an integrated bioaerosol collector.