Devices and methods involving dielectrophoresis for bacterial identification in wastewater

Abstract

In certain examples, an apparatus comprising: a chamber to provide wastewater; a power supply to provide electric field energy: and an optically-responsive material in the wastewater to respond to the field energy. The optically-responsive material (e g., nanomaterial(s) and / or metasurface(s)) are designed to respond to the field energy by causing light signals that scatter from targeted bacteria to be highlighted so as to distinguish the light signals from light due to non-targeted contaminants. In more-specific examples: electrokinetic trapping increases the highlighting for differentiating targeted bacteria; the targeted bacteria is label-free bacteria; the material is optically resonant; the electric field energy is dielectrophoresis field energy; and / or the optically-responsive material include particles each of which is characterized, in part, by a physical structure having one or more subwavelength dimensions, by causing light to concentrate on membranes of the targeted bacteria.

Description

STFD.468PCT (S24-426) 1DEVICES AND METHODS INVOLVING DIELECTROPHORESIS FOR BACTERIAL IDENTIFICATION IN WASTEWATERFEDERALLY-SPONSORED RESEARCH AND DEVELOPMENT

[0001] This invention was made with Government support under contract Al 152072 awarded by the National Institutes of Health. The Government has certain rights in the invention.BACKGROUND(0002] Aspects of the present disclosure are related generally to the field of detecting certain species of contaminants in wastewater, and as may be exemplified by uses in dielectrophoresis.

[0003] In exemplary' contexts, aspects of the present disclosure are directed to overcoming previously-known approaches to identifying different types of bacteria in wastewater rapidly, for monitoring community-level health and providing early warnings of infection outbreaks. The existence and concentrations of pathogen targets in wastewater correlate with disease occurrence in the communities contributing to the wastewater. As such, wastewater-based epidemiology (WBE) has become a valuable tool for understanding the epidemiology of various diseases and informing clinical decision-making and public health response to outbreaks. Thus far, most WBE applications have focused on viral diseases, but WBE monitoring of bacterial pathogens is likely to become increasingly important in coming years; in particular, antibiotic-resistant bacteria are an increasing threat to public health, expected to claim 10 million lives per year by 2050. The ability to accurately monitor bacterial infections at a population level and to identify the start of outbreaks may be a critical w ay to inform public health response and clinical decision-making. Identifying bacteria using nucleic-acid detection techniques is challenging to scale up to the dozens of potential bacterial species that can exist in a community.

[0004] An alternative approach is culturing bacteria which can take weeks, and therefore is not an option during disease outbreaks, especially when a large number of samples needs screening w ithin a short time, or as in underdeveloped regions, where the necessary' infrastructure and equipment may not be accessible. In addition, wastewater is also host to complex metals, ions, biomolecules, and microorganisms, while the target bacteria represent a very small proportion of the (bio)chemicals present in w astew ater, posing challenges to bacterial identification.STFD.468PCT (S24-426) 2

[0005] These and other matters have presented challenges and inadequacies to identifying bacteria in these and other contexts.SUMMARY OF VARIOUS ASPECTS AND EXAMPLES

[0006] Vanous examples / embodiments presented by the present disclosure are directed to issues such as those addressed above and / or others which may become apparent from the following disclosure. For example, some of these disclosed aspects are directed to methods and devices that use or leverage from application of electric field energy in electrophoresis in combination with spectroscopy as applied to wastewater for the detection of selected wastewater contaminants.

[0007] In one specific example, an apparatus comprising: a chamber to provide wastewater; a power supply to provide electric field energy; and an optically -responsive material in the wastewater to respond to the field energy’. The optically-responsive material (e.g., nanomaterial(s) and / or metasurface(s)) are designed to respond to the field energy by causing light signals that scatter from targeted bacteria to be highlighted so as to distinguish the light signals from light due to non-targeted contaminants. In more specific examples, such a chamber may be implemented as a fully- or partially-enclosing fluid retainer such as in any one or any combination of the following forms which are not limited: a well, a microfluidic channel, and outlet or pathway to facilitate the flow of the fluid (e.g., selected portions of wastewater) and / or to facilitate the suspension of fluid flow.

[0008] Another specific example is directed to a method involving field energy' and wastewater. This example method comprises use of a chamber to hold wastewater, in which there are presumably various types of contaminants. While using the chamber, field energy' is applied toyvards optically-responsive material, yvhich is part of the wastewater, and the optically-responsive material responds to the field energy by causing light signals that scatter from targeted bacteria to be highlighted, relative to light signals from non-targeted contaminants, in the wastewater.

[0009] In certain other examples which may also build on the above-discussed aspects, such methods and apparatus may include other one or more other aspects as exemplified by the following: use of an electrokinetic trap that increases the highlighting of light signals for differentiating the targeted bacteria; the targeted bacteria being label-free bacteria; the optically-responsive material being optically resonant; and / or the optically-responsive material including particles each of yvhich is characterized, in part, by a physical structure having one or more subwavelength dimensions, by causing light to concentrate onSTFD.468PCT (S24-426) 3membranes of the targeted bacteria. Further, in certain more specific examples, the field energy from the power supply is dielectrophoresis (DEP), which generates a non-uniform electric field and is advantageous in such examples for the Raman signals. By using non-uniform electric fields (e.g., DEP-type), the clinically relevant bacteria particles are polarized and so as to pose directional forces on the bacteria with specific sizes and dielectric properties. This form of DEP therefore drives bacterial movement to the detection point against their Brownian motions, leading to increased Raman signal relative to wastewater background.(0010] Various aspects and examples according to the present disclosure are directed to issues such as those addressed above and / or others which may become apparent from the following disclosure involving exemplary embodiments to rapidly detect bacteria in wastewater for bacterial WBE by using one or more methods involving use or analysis of the electromagnetic spectrum such as by employing spectroscopy equipment (e.g., Raman spectroscopy).

[0011] In other specific example embodiments, aspects of the present disclosure are directed to a method and / or an apparatus involving a bacteria-identification enrichment platform that uses Raman-based machine-learning spectroscopy with an integration of one or more of the following features carried out within the platform to allow bacteria identification (e.g., label-free) in wastewater with high efficiency and surprising levels of sensitivity. These features are: (1) optically resonant nanomaterials and metasurfaces to amplify the light scattering from bacteria; (2) external electric fields to selectively or collectively displace and enrich bacteria to the detection region: and (3) machine learning algorithms to allow for rapid data interpretation. For example, a Raman-based machine-learning spectroscopy type of platform, configured with its related methodology including these three features, can be integrated into real-time monitoring systems, thereby providing continuous detection and tracking of bacterial pathogens in wastewater, which can be important for early warning systems in public health and industrial applications.

[0012] In another example, the platform is entirely label -free, which eliminates the need for (bio)chemical reagents to bind with specific bacteria. It is also generalizable to diverse types of bacteria. The DEP (dielectrophoresis) enrichment and optical approach is rapid, only taking minutes to complete. By using CMOS-compatible fabrication of a DEP-compatible electrode and potentially surface-enhanced Raman scattering (SERS) substrates on a sensor chip, the aspects of the present disclosure enable patterning of the sensor chip on a large scale and at low costs.STFD.468PCT (S24-426) 4

[0013] The above discussion is not intended to describe each aspect embodiment or every implementation of the present disclosure. The figures and detailed description that follow also exemplify various embodiments.BRIEF DESCRIPTION OF FIGURES

[0014] Various example embodiments, including experimental examples, may be more completely understood in consideration of the following detailed description and in connection with the accompanying drawings, each in accordance with the present disclosure, in which:

[0015] FIGs. 1 A-1E illustrate certain exemplary aspects, with FIG. 1 A showing an example platform showing spectra data being collected from liquid samples in a chamber, FIG. IB as an electron micrograph for exemplary optically-responsive material, FIG. 1C as a schematic illustrating SERS performance of optically-responsive material in such liquid wells, FIG. ID as a cryoelectron micrograph of electrostatic binding of optically-responsive material to a targeted bacteria, and FIG. IE as an image of SERS spectra of four model species collected in formalin and wastewater;

[0016] FIGs. 2A, 2B, 2C, 2D. 2E, 2F, 2G and 2H, also according to certain exemplary- aspects, are respective cryoelectron micrographs of exemplary optically-responsive material (for example, gold nanorods such as AuNRs) bound to: S epidermidis in formalin (FIG. 2A), S. epidermidis in wastewater (FIG. 2B), S. aureus in formalin (FIG. 2C), (d) S. aureus in wastewater (FIG. 2D), S. marcescens in formalin (FIG. 2E), S. marcescens in wastewater (FIG. 2F), E. coli in formalin (FIG. 2G). and E. coli in wastewater (FIG. 2H);

[0017] FIGs. 3 A, 3B, 3C and 3D, according to certain exemplary' aspects, form a stacked area chart depicting the intensities of selected bacterial peaks in wastewater depending on concentration of the particular optically-responsive material (in this example, AuNR);

[0018] FIG. 4A is a confusion matrix indicating significant accuracy of bacterial classification using an example platform, and FIG. 4B is a related graph also indicating significant accuracy of a support vector machine when the test set is perturbed at various wavenumbers; and

[0019] FIGs. 5A-5C depict SERS performance aspects for various model species collected in wastewater, with: FIG. 5A as a set of four graphs illustrating SERS performance of exemplary optically-responsive material for four model species collected in wastewater at varying bacterial concentrations, and FIGs. 5B and 5C as respective graphs indicating accuracy aspects relating to the model species of FIG. 5 A.STFD.468PCT (S24-426) 5

[0020] While various embodiments discussed herein are amenable to modifications and alternative forms, aspects thereof have been show i by way of example in the drawings and will be described in detail. It should be understood, how ever, that the intention is not to limit the disclosure to the particular embodiments described. On the contrary7, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure including aspects defined in the claims. In addition, the term “example” as used throughout this application is only by w ay of illustration, and not limitation.DETAILED DESCRIPTION

[0021] Aspects of the present disclosure are believed to be applicable to a variety of different types of apparatuses (e.g., systems, platforms, optical equipment, electrical circuits and / or components) involving Raman spectroscopy for detection of contaminants in wastewater. While the present disclosure is not necessarily limited to such aspects, an understanding of specific examples in the following description may be understood from discussion in such specific contexts.

[0022] Accordingly, in the following description various specific details are set forth to describe specific examples presented herein. It should be apparent to one skilled in the art, however, that one or more other examples and / or variations of these examples may be practiced without all the specific details given below. In other instances, well known features have not been described in detail so as not to obscure the description of the examples herein. For ease of illustration, the same connotation and / or reference numerals may be used in different diagrams to refer to the same elements or additional instances of the same element. Further, unless otherw ise indicated ranges (of any, and all metrics) are merely exemplary7of “approximate ranges” or levels of improvement, w herein such terms are understood to vary the bound(s) of the range or level(s) (e.g., using improved and / or degraded material- or circuit-based design parameters) by a degree of anywhere from 10-to-20 percent (or in some instances) from 5-35 percent, and, in the context of comparison to an improvement over a previously-reported effort, by a degree of improvement of 20 percent or greater. Also, although aspects and features may in some cases be described in individual figures, it will be appreciated that features from one figure or embodiment can be combined with features of another figure or embodiment even though the combination is not explicitly show n or explicitly described as a combination.

[0023] Exemplary aspects of the present disclosure are related to apparatuses and methods involving field energy7and wastewater. In one specific example, such a methodSTFD.468PCT (S24-426) 6comprises use of a chamber to hold wastewater, in which there are presumably various types of contaminants. While using the chamber, the field energy is applied towards optically-responsive material, which is part of the wastewater, and the optically-responsive material responds to the field energy by causing light signals that scatter from targeted bacteria to be highlighted, relative to light signals from non-targeted contaminants, in the wastewater.

[0024] Consistent with the above aspects, such a manufactured device or method of such manufacture may involve aspects presented and claimed in U.S. Provisional Application Serial No. 63 / 721,319 filed on November 15, 2024 (STFD.468P1 / S24-426) with Appendices A-C, to which priority is claimed. To the extent permitted, such subject matter is incorporated by reference in its entirety generally and to the extent that further aspects and examples (such as experimental and / more-detailed embodiments) may be useful to supplement and / or clarify.

[0025] As noted above, certain exemplary' aspects of the present disclosure may be used in connection with the above-disclosed aspects to improve and / or optimize the detection process and / or its efficacies. For example, electrokinetic trapping may be used to increase the highlighting of the light signals for differentiating the targeted bacteria, the targeted bacteria may be detected without any labeling of the bacteria (e.g., without external fluorescent dyes, stains and / or molecular tags), and the optically-responsive material is optically resonant.

[0026] The electrokinetic trap may be used alone or in combination with patterned microelectrodes to facilitate the detection. The patterned microelectrodes may be located at or near (e.g., immediately adjacent) the detection region, which may be an electrode and / or a substrate surface, used as optional part(s) of the apparatus. In this way, the electric field effect significantly increases the concentration and selectivity for bacteria on the Raman detection region (e.g., surfaces) by attracting the targeted bacteria to the sensor surface selectively or collectively, thereby promoting target signals from bacterial cells over other non-targeted contaminants in the wastewater. Further, certain more-specific examples may use non-uniform (DEP type) electric fields, which polarizes the clinically relevant bacteria particle and pose directional forces on the bacteria with specific sizes and dielectric properties. This form of DEP therefore drives bacterial movement to the detection point against their Brownian motions, leading to increased Raman signal relative to wastewater background.

[0027] In certain applications and also according to the present disclosure, such exemplary aspects address a significant unmet need in the field of wastewater-based epidemiology. Although wastewater-based epidemiology has been widely used for surveillance of viral diseases, wastewater-based epidemiology has not been used to a similar extent for detection of specific bacteria, for example, that may lead to bacterial diseases. InSTFD.468PCT (S24-426) 7part, this is due to difficulties in distinguishing pathogenic from non-pathogenic bacteria when attempting to use nucleic-acid-based methods such as PCR (Polymerase Chain Reaction) and sequencing which often require burdensome primers or probes specific to each bacterial species, thereby limiting their ability to detect a broad range of species in a single test. Also, many existing bacterial identification methods, including PCR, rely on labeling techniques, such as fluorescent tagging or antibody binding, which require specific reagents and additional sample preparation steps. Further, virulence of a pathogen can be determined by phenotypic signatures (such as protein expression levels and post-translational modifications) which are not accessible from genomic or transcriptomic analysis alone, and a priori knowledge of pathogen genomic sequences (e.g., as with PCR and related label-required approaches) is typically incompatible with emerging antibiotic resistant strains.

[0028] In certain specific examples, aspects of the present disclosure are implemented to realize a label-free WBE bacteria detection platform that eliminates the need for (bio)chemical reagents and simplifies the sample preparation process, thereby reducing costs and improving the speed and scalability of bacterial detection in wastewater. The generalizable platform can detect diverse types of bacteria in wastewater without needing species-specific reagents, and a filter; e.g., Raman spectroscopy tools and / or AI / ML-based computer circuitry (artificial-intelligence and / or machine learning) can be used to identify multiple bacterial species from their Raman fingerprints, making the platform adaptable to different bacterial communities.

[0029] According to certain more specific examples and consistent with the above aspects, the present disclosure is directed to a method to distinguish, detect and / or identify bacteria that is targeted among a variety of contaminants in wastewater. The method includes, while a chamber is used to hold the wastewater, applying field energy towards optically -responsive material, which is part of the wastewater, and the optically-responsive material responding to the field energy by light signals scattering from the targeted bacteria being highlighted, such that the highlighted light signals are distinguishable relative to light signals reflected and / or scattered from non-targeted contaminants.

[0030] In the above examples and others according to the present disclosure, the optically-responsive material may include particles that are characterized, in part, by a physical structure having one or more subwavelength dimensions, by causing light to concentrate on membranes of the targeted bacteria. Via the Raman spectroscopy, the light scatters (e.g., off of the bacteria membranes) and provides for the capture of fingerprint-like identification of the targeted bacteria due to their molecular and physical structure. ToSTFD.468PCT (S24-426) 8improve the intrinsically low Raman scattering efficiency and signal-to-noise ratios on biological samples and also according to exemplary aspects of the present disclosure, such as resonant nanostructures and / or resonant metasurfaces with patterned nanoantennas; these structures may be implemented by using a sensor chip, which is integrated into a microfluidic chamber which facilitates the flow of wastewater across the chip for identifying bacterial species and their mixtures (e.g., in experimental examples, several (representative) modeled bacterial species and their mixtures are readily identified).

[0031] In such specific examples, the subwavelength nanomaterials are used to concentrate light on membranes of the targeted bacteria, and this enhances surface-enhanced Raman scattering (SERS). For example, Raman signal from bacteria is enhanced in wastewater using plasmonic gold nanorods (AuNRs) as such subwavelength nanomaterials. The AuNRs (at one or more of various concentrations) electrostatically bind to the bacterial surface, as may be readily confirmed using cryoelectron microscopy. Various (e.g., four) clinically-relevant bacterial species and AuNRs are spiked into filtered wastewater, with the AuNR concentrations varying to maximize signal. From this spiked wastewater, a multitude of spectra (e.g., 540) are collected from each species (e.g., at 109cells / mL), and such species are then identified using, for example, Raman spectrometry tools and / or by training a machine learning model to identify the species. In such experimental examples, these species are readily identified with more than 87% accuracy. Other specific experimental examples, according to the present disclosure, demonstrate an environmentally-realistic limit of detection of 104cells / mL.

[0032] Accordingly, such results confirm that for detection of selectable ("‘targeted") bacteria in wastewater, spectrally-responsive materials, such as nanomaterial and / or metasurface materials, may be utilized to respond to the applied field energy, with light signals scattering from the targeted bacteria being highlighted to distinguish the targeted bacteria from non-targeted contaminants. In more-specific applications, this approach and such related experimental results may be used as a SERS platform implemented as a practical and efficient approach (e.g., via scalable and label free) for such detection of selectable bacteria in w astew ater.

[0033] In certain other example embodiments (which also may be combined with the above-characterized embodiments), such a platform may also be configured specifically with the third of the above three aspects and integrated with machine learning algorithms that can discern between the most common bacterial species in wastewater and validate the system on real w astew ater. Machine learning (ML) methods including convolutional neural networksSTFD.468PCT (S24-426) 9(CNNs) and use Raman fingerprints to discern the bacterial species in wastewater with over 96% overall accuracy. ML methods also interpret Raman spectra and identify the spectral features that the model uses for classification by perturbing the spectra in the test set at individual wavenumbers and assessing the effect of these perturbations on classification accuracy, to ensure the features are biologically relevant. In more specific ML-based examples, experimental efforts leading to the present disclosure have revealed that Raman spectroscopy (with its use of a monochromatic light source to measure vibrational energy levels based on inelastically-scattered photons) is particularly suitable for wastewater analysis, in part because it involves a visible excitation and detection wavelengths that manifest minimal absorption in aqueous samples for the spectroscopy measurements of different species and strains of bacteria which possess unique Raman “fingerprints”, allowing Raman identification of clinically -relevant information at the species and subspecies level. In such ML-based examples . a priori data can be used to equip a library (or multiple libraries with species being categorized) of Raman spectra of bacteria in water, in addition to bacteria isolated from sputum and bacteria mixed with red blood cells in an Ethylenediaminetetraacetic acid solution. The set of one or more libraries (hereinafter “library”) can be used as a basis for the ML algorithm which can be used not only to identify bacteria by comparing to the data in a model corresponding to detailed aspects of the fingerprints in the library, but also to recognize and categorize new uncommon bacteria species based on attributes of such new species that do not correlate with attributes of known species previously recorded in the library7, which evolves as the platform used (alone or in combination with other such WBE platforms) on wastewater samples.

[0034] In yet other examples, aspects of the present disclosure are directed to apparatuses and / or methods having, depending on the implementation, certain applications and advantages. In terms of exemplary applications, rapid, label-free bacterial identification in wastewater using such a Raman-based machine-learning spectroscopy platform can drive the field of bacterial WBE (wastewater-based epidemiology), which can address urgent global challenges in understanding the epidemiology of infectious diseases and informing early warning of outbreaks. As examples, aspects of the present disclosure can be implemented at least as part of products and services with applications for government agencies, research institutions, healthcare providers, and industrial sectors. Some of these applications are: wastewater monitoring systems (e.g., which may meet unsatisfied needs of previously-known methods for detecting bacterial pathogens in wastewater, which methods are either too slow (culturing) or too difficult to scale (nucleic acid detection), particularly forSTFD.468PCT (S24-426) 10identifying multiple bacterial species); a rapid, label-free detection method that provides actionable data for public health monitoring; as a product and / or sereice, the Raman-based machine-learning spectroscopy type of platform can be implemented for automated wastewater monitoring systems deployed by municipalities, environmental agencies and utilities to track bacterial pathogens in wastewater (e.g.. to provide real-time data on bacterial contamination, helping to inform public health responses of outbreaks and monitoring the spread of antibiotic-resistant bacteria).[00351 In connection with experimental and / or proof-of-concept efforts leading to the present disclosure, specific examples have successfully established that the chamber containing wastewater can be used with the applied electric field energy to distinguish the light signals from light due to non-targeted contaminants, and as realized by use of an optically-responsive material in the wastewater to respond to the field energy. The optically -responsive material are designed to respond to the field energy by causing light signals that scatter from targeted bacteria to be highlighted so as indicate the light signals scattered from the targeted bacteria, relative to other light ensuing from non-targeted contaminants.

[0036] To help in understanding important aspects relating to such approaches, FIG. 1A through FIG. IE depict related aspects and data corresponding to these experimental efforts. As an overview, FIG. 1 A is a representative platform (or apparatus) including a chamber in the form of microfluidic wells that contains wastewater being used with the electric field energy to distinguish light signals scattered (and highlighted) from targeted bacteria species from detected light due to non-targeted contaminants. Raman spectra is collected from the liquid samples in the microfluidic wells. FIGs. IB and 1C are, respectively, an electron micrograph of AuNRs used for SERS, and a schematic illustrating that in the liquid wells, AuNRs selectively enhance scattered signals from bacteria over a complex mixture of background components, such as colloids, proteins, and nucleic acids. FIG. ID is a ci oelectron micrograph of electrostatic binding of AuNRs to E. coli surface, and FIG. IE is a pair of graphs showing SERS spectra, plotting spectra intensity versus Raman shift, of four model species.

[0037] In a more-specific example of this type, such experimental efforts have realized that bacterial SERS can be achieved in a filter-sterilized wastewater matrix, for bacterial WBE. In this regard, bacterial SERS in wastewater generates Raman peaks which are comparable to those observed in formalin; that cryoelectron microscopy (cryoEM) reveals AuNR binding to the bacterial surface in this matrix; that a high AuNR concentration of 150 pg / mL is suitable for SERS, enhancing signal rather than quenching it; that bacterial SERSSTFD.468PCT (S24-426) 11spectra collected in filtered wastewater are sufficiently robust for species identification; and that combined with ML, this platform can detect bacteria in wastewater at concentrations as low as 104cells / mL. These results indicate that Raman spectroscopy can be used, as disclosed via various examples herein, for bacterial WBE by measuring the distinct Raman scattering from bacteria in such wastewater. It is noted that AuNRs are used in these efforts as an example type of optically-responsive material, and that alternative material types may manifest similar optical responsiveness by changing their optical properties (e.g., as absorption, emission, or refractive index) when exposed to light. For example, as alternatives to AuNRs and other Au material, such alternative material types (used alone or in any combination (including in combination with Au material) include photochromic molecules, photoresponsive polymers, photonic crystals, plasmonic nanostructures (e.g., silver nanoparticles and / or gold nanoparticles other than AuNRs), liquid crystals, and luminescent nanomaterials, among others.

[0038] In this experimental example, the AuNRs focus the incident field onto the analytes, with reported enhancement factors ranging from 103-109. This technique enables rapid, sensitive Raman spectroscopy. As opposed to a system wherein Raman signal is developed from a secondary reporter increased in wastewaters with higher estrogen concentrations, this example achieves reporter-free detection of bacteria in wastewater, by utilizing electrostatic interaction between bacteria, which are generally negatively-charged, and AuNRs, which is synthesized with positively-charged ligands such as cetrimonium bromide (CTAB). See FIG. IB and FIG. 1C. Experiments have shown that in deionized (DI) water, electrostatic interactions will allow AuNRs to bind selectively to the cell surface of almost any bacteria (thereby enhancing bacterial Raman over the background medium’s Raman signature), but with the experimental efforts of the present disclosure the samples are as complex as wastew ater, which contains a mixture of a variety of biomolecules and waste products, The AuNRs are used to enhance bacterial Raman signal over the background medium, enabling label-free identification of bacteria in wastewater. See FIG. ID. This is a significant development over previous bacterial SERS work: unlike in previous literature that has primarily characterized bacteria in controlled buffers and matrices, the experimental efforts of the present disclosure are associated with wastewater in which the bacterial signal is in a complex mixture of unknown molecules, which can contribute to Raman background, and affect the pH and ionic strength of the solution in ways that interfere with AuNR binding.

[0039] With these experimental efforts of the present disclosure, the feasibility of WBE using SERS and ML is established and, as a proof of concept, four bacterial species areSTFD.468PCT (S24-426) 12considered: Staphylococcus epidermidis (S. epidermidis). Staphylococcus aureus (S. aureus), Serratia marcescens (S. marcescens), and Escherichia coli (E. coli). With these species being spiked into filter-sterilized wastewater with CTAB-functionalized AuNRs, these efforts demonstrate SERS from this spiked wastewater with an acquisition time of only 10 seconds. Additionally, transmission cryoelectron microscopy (cryoEM) is used to confirm AuNR binding to the cell surface through strong electrostatic interactions. Further, the relationship between AuNR concentration and bacterial Raman signal in wastewater demonstrates that the AuNRs have a primarily enhancing, rather than quenching effect among the tested concentrations. Additionally, such efforts show that bacterial SERS spectra collected in wastewater can be used to accurately predict bacterial species: a support vector machine (SVM) was trained to classify them with 87% accuracy, and that such methodology according to such examples of the present disclosure can be implemented to detect the presence of bacteria in wastewater at environmentally -relev ant concentrations as low as 104cells / mL.

[0040] More specific aspects of these experimental example efforts are described as follows. The AuNRs were synthesized using a seed grow th method, with the AuNRs having structural aspects designed to have a longitudinal surface plasmon resonance (LSPR) at approximately 700 nm with dimensions of approximately 75 nm by 30 nm. designed for SERS under a 785 nm laser excitation, as in FIGs. 1 A and IB. Clinically-relevant pathogens were used (S. epidermidis, S. aureus, S. marcescens, and E. coli) as model organisms to determine the effects of wastewater on AuNR enhancement of bacterial Raman spectra. Among such pathogens, S. marcescens, S. aureus, and S. epidermidis have respective surface charge densities 4, 29, and 102 times greater than E. coli and have stronger electrostatic binding to AuNRs in deionized water. The bacterial SERS were tested in a wastew ater matrix, which was prepared from the liquid portion of wastewater from a local w astew ater treatment plant. These efforts filter-sterilized the wastewater through a 0.22 pm pore-size filter to remove indigenous bacteria, in addition to other cells and particles, while preserving the colloids, dissolved ions, biomolecules, and various dissolved molecules in wastewater. Samples were then formalin treated for biosafety. For each experiment, this wastewater matrix was spiked with one of the model species and mixed 1 : 1 with AuNRs for a final concentration of 150 pg / mL AuNRs and 109cells / mL, which is approximately equal to the total bacterial concentration in untreated wastewater. As a control, samples w ere also prepared in formalin-treated DI water.STFD.468PCT (S24-426) 13

[0041] The performance-related left and right graphs of FIG. IE (using a platform as in FIG. 1A), shows that Raman spectra were collected from each species of bacteria with and without AuNRs in formalin and filter-sterilized wastewater. The left graph corresponds to the four model species being collected in formalin, and the right graph corresponds to the four model species being collected in wastewater. The plotted grey lines in the plot indicate negative control without AuNRs in an example involving SERS spectra that are averaged over a minimum of 216 samples and unenhanced spectra are averaged over a minimum of 36 samples. Shaded regions (mostly apparent in the upper portion of the left graph of FIG. IE) indicate a standard deviation error of one.

[0042] More specifically in connection with the plots of FIG. IE, the samples were transferred to a liquid well for Raman analysis, acquired over 10 seconds with a 785 nm laser at a pow er of 11 mW and a laser spot size of 2 pm. A minimum of 216 spectra per species were collected with AuNRs (and 36 spectra were collected without AuNRs). In samples without AuNRs, no bacterial Raman peaks were observed as shown in FIG. IE, consistent with liquid Raman results with similarly low exposure times and power densities. In samples with cells and AuNRs in formalin, strong Raman peaks are visible as also in FIG. IE. These peaks overlap strongly with dozens of wavenumbers known to be associated with biomolecules on the bacterial surface. These features include particularly strong peaks at 760 cm-1(adenine); 851 cm'1(thymine); 1,040 cm'1(aryl); 1,125 cm'1(phosphate); and 1,599 cm'1(carboxyl). All of these major peaks are conserved across the four species, which is expected: differences betw een the Raman spectra of different bacterial species tend to be subtle and difficult to discern by the human eye. Despite these similarities, it has been previously demonstrated that ML techniques can be used to classify between bacterial species with shared Raman peaks.

[0043] In samples w ith bacteria and AuNRs in a filtered w astew ater matrix, strong peaks are still present, with some differences from those collected in formalin. A peak at 670 cm'1is only present in bacterial spectra collected in wastewater. This band is characteristic of the C-S stretching vibration of the thiol group in cysteine, which is known to covalently bond to noble metal nanoparticles. Therefore, the data indicates that this peak is from fecal proteins bonding to the AuNRs through their cysteine residues. Despite this new' thiol peak, spectra collected in wastewater are otherwise qualitatively similar to those collected in formalin. All major peaks are conserved, including the biological peaks at 760 cm'1, 851 cm'1, 1,040 cm'1, 1,125 cm'1, and 1,599 cm'1that were previously highlighted.STFD.468PCT (S24-426) 14

[0044] Despite these qualitative similarities, the spectra collected in filtered wastewater have varied peak intensities from those collected in formalin. Some, but not all, of these peaks decrease in intensity in wastewater compared to formalin. The greatest decrease was observed for the 760 cm-1peak, which decreases by 76-88% depending on species. The 1,040 cm'1peak also decreases across all species by 51-78%. The 851 cm'1. 1,125 cm'1, and 1,599 cm'1peaks, however, do not display a statistically-significant change in intensity in formalin compared to wastewater, indicating that wastewater does not strongly affect the intensities of these peaks (as indicated by other experimental efforts). Moreover, the peaks that do decrease in intensity are still clearly observable. Overall, SERS is achievable from diverse bacteria in a wastewater matrix.

[0045] To validate that these results in filter-sterilized wastewater spiked with bacteria are representative of unfiltered wastewater, SERS were additionally performed on a sample of unfiltered wastewater with no bacteria spiked in. The unfiltered wastewater displayed strong Raman signal at biologically-relevant wavenumbers. Moreover, scanning electron microscopy of a dried sample of this mixture reveals that 100% of indigenous bacteria colocalize with AuNRs, indicating that AuNRs bind to the surface of indigenous bacteria in raw wastewater. Therefore, the Raman signal observed from unfiltered wastewater is the result of AuNR enhancement of signals from indigenous bacteria, indicating that AuNRs enhance bacterial Raman signal even in matrices far more complex than the model system as exemplified in certain of the embodiments of the present disclosure.

[0046] As indicated in FIGs. 2A-2H, cryoEM images (cryoelectron micrographs) were collected of samples containing 109cells / mL and 150 pg / mL AuNRs in both formalin and filtered wastewater to further validate that bactenal Raman signal in filtered wastewater was the result of AuNR binding to cells in this matrix. In samples in formalin, multiple dozens of AuNRs are bound to the surface of the average cell for all four species as depicted in FIG. 2A (S. epidermidis in formalin), FIG. 2C (S. aureus in formalin), FIG. 2E ( . marcescens in formalin), and FIG. 2G (A. coli in formalin). In formalin-treated wastewater, cryoEM images still show AuNR binding to all four species as depicted in FIG. 2B (S. epidermidis in wastewater), FIG. 2D (A aureus in wastewater), FIG. 2F (S. marcescens in wastewater), and FIG. 2H (E. coli in wastewater).

[0047] To more quantitatively assess this binding to cells, these samples were centrifuged to remove all cells and bound AuNRs, but not to remove the unbound AuNRs. The concentration of gold in the supernatant of each of these samples was quantified using inductively coupled plasma optical emission spectroscopy, and the decrease in concentrationSTFD.468PCT (S24-426) 15for each sample relative to a control was used to quantify AuNR binding. For . epidermidis, S. aureus, and S. marcescens, >95% AuNR binding was observed in both formalin and filtered wastewater. For / / . coli, on the other hand, only 59% binding was observed in formalin and 13% binding was observed in filtered wastewater (as indicated by other experimental efforts). While binding to E. coll is lower than to other species, it is notable that cryoEM still shows many AuNRs bound to the typical E. coli cell, thereby confirming that the binding to this species is still sufficient to achieve SERS.; 0048 j Among these four model species, E. coli (FIG. 2H) has by far the lowest surface charge density. Thus, the results indicate an intuitive relationship between surface charge density and electrostatic interactions. Moreover, the decreased AuNR binding to E. coli in filtered wastewater suggests that weaker electrostatic interactions can be disrupted by the ionic strength of the w astewater environment. Thus, wastewater can decrease AuNR binding to the cell surface, but AuNR binding is never eliminated.

[0049] FIGs. 3A, 3B, 3C and 3D are respective plots providing for a comparison of SERS-related intensify data involving each of the modeled four species. Collectively, FIGs. 3A-3D form a stacked area chart depicting the intensities of selected bacterial peaks in wastewater depending on concentration of the particular optically -responsive material (in this example, AuNR). Each colored band indicates the intensity of one peak, and the total height of the chart indicates the sum of all five peak intensities. Error bars indicate standard deviation of 1599 cm'1peak, which is representative of standard deviation for all peaks.

[0050] In order to maximize Raman signal, next the concentration of AuNRs was varied, from 10 pg / mL to 150 pg / mL. The mean intensities of the 760 cm1, 851 cm1, 1,040 cm1, 1,125 cm'1, and 1,599 cm'1peaks (which respectively correspond to adenine, thymine, aryl, phosphate, and carboxyl) for each species were compared. Across all species, the greatest signal was generally observed at 150 pg / mL. There is a local maximum at AuNR concentrations of 50 pg / mL, compared to concentrations of 100 pg / mL. In these intermediate concentration regimes, the quenching outcompetes enhancement (and / or the AuNRs bind less stably to bacteria or are more prone to aggregation in certain concentration regimes). Overall, the data for these specific example efforts indicate that a concentration of 150 pg / mL AuNRs enhances Raman signal more than it diminishes the Raman signal, thereby making it suitable for bacterial SERS in wastewater systems.

[0051] FIGs. 4 A and 4B illustrate accuracy attributes for validation of the bacterial identification efforts, using the platform and related aspects exemplified by FIG. 1A. FIG. 4A is a confusion matrix showing 87% accuracy of bacterial classification using an SVM trainedSTFD.468PCT (S24-426) 16on bacterial SERS spectra collected in wastewater, and FIG. 4B is a graph showing accuracy of SVM when the four-species test set is perturbed at various wavenumbers. Sharp decreases can be seen with perturbation at bands associated with (i) adenine, (ii) thymine, (iii) aryl, (iv) phosphate, and (v) carboxyl. In developing this data, ML was used to classify the Raman spectra and assess the utility of SERS for bacterial identification in wastewater. From each bacterial species, 540 spectra were collected in filtered wastewater with 150 pg / mL AuNRs and analyzed using an SVM. Based on this set up, the ML model was trained on the bacterial Raman spectra collected in wastewater from 725 cm-1to 1800 cm'1. This example range was selected because it excludes the non-bacterial 670 cm thiol peak from wastewater, which could lead to overfitting. To avoid overfitting, the model was trained on the top 25 principal components of the spectra, which account for 37% of variance.

[0052] The ML-based model was validated using k-folds cross-validation. As indicated in the highlighted entries of FIG. 4A, this model identified S. epidermidis with 84% accuracy (top-left entry), S. aureus with 86% accuracy, S. marcescens with 85% accuracy, and E. coli with 96% accuracy (bottom-right entry). Overall, the model predicted bacterial species with over 87% accuracy (average of approximately 88%). To ensure that this accurate ML classification was based on biologically relevant Raman peaks, relevant wavenumbers were identified using an iterative perturbation method. For each iteration, all spectra in the test set were perturbed at a specific wavenumber, and the effect of that perturbation on accuracy was recorded. Perturbation of a wavenumber relevant to classification can be expected to cause a strong decrease in accuracy, so this method identified these relevant wavenumbers. From these experimental efforts, the wavenumbers were found to be relevant to classification accuracy that strongly overlapped with known bacterial Raman peaks. For example, perturbation at 760 cm'1(adenine) decreases accuracy7by 13%, perturbation at 851 cm'1(thymine) decreases accuracy by 11%, perturbation at 1,040 cm'1(ary l) decrease accuracy by 8%, perturbation at 1,125 cm'1(phosphate) decreases accuracy by 18%, and perturbation at 1,599 cm'1(carboxyl) decreases accuracy by 31%, while perturbation at non-biological wavenumbers generally decreased accuracy by <5% as in FIG. 4B. Thus, the 87% classification accuracy is based on Raman spectra differences stemming from the different chemical compositions of these bacterial species.

[0053] The concentrations of clinically relevant species in raw wastewater span many orders of magnitude. While the total concentration of bacteria in wastewater has been reported on the order of 109cells / mL, the concentrations of individual species tend to be far lower (e.g., reports indicate E. coli concentrations in wastewater ranging from an order of 102STFD.468PCT (S24-426) 17cells / mLn to order of 103cells / mL. To assess the limit of detection (LoD) for bacterial SERS in wastewater, samples were prepared in a fdtered wastewater matrix with 150 pg / mL AuNRs and bacteria at concentrations ranging from 0 cells / mL to 109cells / mL.

[0005] FIGs. 5A, 5B and 5C are provided to show the results of this assessment, with FIG. 5A corresponding to these four model species as collected in wastewater at varying bacterial concentrations, FIG. 5B shows ROC (Receiver Operating Characteristic) analysis of KNN (K-Nearest Neighbors) classifiers trained to detect presence of bacteria at varying concentrations. Red dotted line indicates the expected result of random guessing, and FIG. 5C shows the sensitivity, selectivity, and accuracy of KNN classifiers. Shaded regions indicate concentrations previously reported for total bacteria in wastewater and E. coli in wastewater.

[0055] At each concentration, 756 Raman spectra were collected, including a minimum of 108 spectra from each species (FIG. 5 A). At these lower concentrations, biological peaks are still present in the spectra. This result is likely due to the wastewater matrix containing biomolecules similar to those on the bactenal cell surface. However, the intensities of many peaks decrease at concentrations below 109cells / mL, indicating a strong contribution to these peaks from the bacterial surface.

[0056] To test the limit of detection, a k-nearest neighbors (KNN) classifier was used. This model was trained on a projection of the spectra using the top 3 eigenvectors used to calculate the principal components of a dataset containing all spectra with 109cells / mL and all spectra with no bacteria. The model was then assessed by calculating its receiver operating characteristic (ROC), sensitivity, and selectivity. At concentrations of 104-107cells / mL, ROC analysis consistently displays an area under curve (AUC) in the range of 0.65-0.75. At 108cells / mL, AUC increases to 0.83±0.02. At 109cells / mL, this value reaches 0.96±0.01 (FIG. 5B). Similarly, from 104-l 07cells / mL, sensitivity, selectivity, and accuracy all fall approximately in the 55-70% range; at 108cells / mL, sensitivity, selectivity7, and accuracy respectively increase to 81±3%, 75±3%, and 77±2%; and at 109cells / mL, sensitivity, selectivity, and accuracy all are 90±2% (FIG. 5C). The threshold concentration is around 107cells / mL, above which related experimentation shows accuracy increases significantly. This threshold matches the concentration above which a dense layer of cells forms on the bottom of the well. Above the threshold concentration, cells are believed to be present in the laser spot for most spectral acquisitions, so the spectra contain information on the cell surface itself; below the threshold, the presence of bacteria can be detected based on Raman signal from molecules that cells not in the laser spot release into their surrounding medium.STFD.468PCT (S24-426) 18

[0057] It is notable that the presence of bactena could be detected with >50% accuracy, even at low concentrations at which it is highly improbable for bacteria to be in the laser spot for any given acquisition. This result indicates that there is potential for future Raman-based bacterial detection systems that do not require direct irradiation of the cells themselves. Together, these results suggest that SERS is capable of detecting bacteria at environmentally-relevant concentrations. For real-time operation of such a wastewater system, methods (e.g., filtering involving chemical, optical, electrical forces and / or AI / ML iterative feedback control) may be implemented to isolate signals from individual species in raw wastewater, as raw wastewater may contain broad classes of bacteria, protozoa and / or other organic and inorganic particles.

[0058] For such experimental and / or proof-of-concept examples according to the present disclosure, the above-disclosed data shows surprising and unexpected results, and as may be applied to address such unmet needs for scalable wastewater monitoring of bacterial pathogens. Consistent with the type of approach disclosed with FIG. 1A and with the representative set of examples and for superior performance relative to an SVM, further aspects and exemplary embodiments of the present disclosure are directed to ML training datasets that include larger number (e.g., hundreds) of species and tens of thousands of spectra per species, with the expanded dataset enabling SERS monitoring of a fuller breadth of the bacterial pathogens present in wastewater and with sufficient data per class for implementation of deep learning ML models. Additionally, work is needed to diversify the wastewater background training data (e.g., from a variety of geographic locations), and the bacterial population (e.g. multiple strains of each species, varied culture conditions). There is also potential for synthetic data augmentation techniques to diversify data. Such diversity would prevent potential overfitting to a specific set of conditions.

[0059] Yet further aspects and exemplary' embodiments of the present disclosure are directed to aspects that improve identification accuracy, such as by a platform that is (similar to the platform of FIG. 1A and) configured to collect spectra from few-to-single cells in wastewater, so as to allow characterization of wastewater on a cell-by-cell basis. More specific aspects of such a cell-detection platform, may include: integration of bioprinting and microfluidics with Raman spectroscopy, for realizing higher throughput for single cell isolation; and / or enrichment techniques, including but not limited to improvements in dielectrophoresis, for concentrating bacteria near electrodes (e.g., as detection surfaces and / or detection regions and with certain systems according to the present disclosure including such electrodes), to maximize the corresponding Raman signal and even at low bacterialSTFD.468PCT (S24-426) 19concentrations. In one such specific example system according to such aspects of the present disclosure, a Raman system (as in FIG. 1 A) is configured to combine bioprinting and / or microfluidics with enrichments sufficient for sensitively generating single-cell spectra.(00601 Various specific example embodiments of the present disclosure are useful to enable implementations to address more specific needs of entities such as public health organizations, pharmaceutical and research organizations. As examples, these include entities serving disease surveillance and prevention such as organizations in underdeveloped regions, and such entities that need tools to study bacterial pathogens, including antibiotic-resistant strains, and their interactions in various environments (e.g., wastewater, clinical settings). As previously-known methods are relatively slow, among other issues, more rapid screening is necessary for developing new treatments, and the Raman-based machine-learning spectroscopy type of platform can be adapted to address such issues (e.g., as part of laboratory instruments for pharmaceutical companies and academic researchers working on microbiome studies and pathogen identification, thereby enabling rapid identification and characterization of bacteria in various complex liquid samples). In more detailed implementations, a platform or system according to the present disclosure is implemented as a portable diagnostic tool for use in clinics, hospitals, or mobile health units. With such a tool implemented according to the present disclosure, on-site bacterial identification is enabled in wastewater and other liquid samples, thereby providing early warnings of infection. Certain example aspects and embodiments of the present disclosure can also be implemented to address needs in industrial and agricultural settings, such as monitoring for bacterial contamination in wastewater which is essential for environmental compliance and operational safety, particularly in sectors like food processing, pharmaceuticals, and agriculture. As there is a lack of methods for on-site, rapid pathogen tracking, the Raman-based machine-learning spectroscopy type of platform can be also adapted to address these needs, for example, with the platform implemented as a monitoring system for bacterial contamination in industrial and agricultural wastewater to enable industries to detect harmful bacterial species rapidly, ensuring compliance with regulations and preventing potential health hazards.

[0061] Again depending on the implementation, the spectroscopy types of platform disclosed herein can provide significant improvements not only over more conventional (e.g., nucleic-acid-based) methods but also over Mass Spectrometry’ (e.g., MALDI-TOF MS) which identifies bacterial species based on the mass of cellular proteins. Limitations of such Mass Spectrometry include requiring complex instrumentation.STFD.468PCT (S24-426) 20

[0062] In yet further examples, implementations of the present disclosure involve realtime (including near real-time) monitoring, which can also provide improvements over existing methods. Certain existing technologies, such as culturing or molecular methods, are batch processes that cannot provide continuous, real-time data.

[0063] In view of the above, the skilled artisan would appreciate that the present disclosure describes and / or illustrates aspects useful for implementing the claimed disclosure by way of various structures circuits or circuitry which may be illustrated as or using terms such as blocks, modules, device, system, unit, controller, optical elements and / or other circuit-related depictions and materials and / or layers (e.g., substrates, optically -responsive material such as metasurface(s), nanoantennas, nanomaterials), that are semiconductive, conductive, insulative, and / or have other attributes such as being metallic or semi- metallic. Such circuits (or circuitry ), materials and the like may be used together with other elements to exemplify how certain embodiments may be carried out in the form of structures, steps, functions, operations, activities, etc. As other examples, where the Specification may make reference to: comprising, including and / or having, such terms are to be taken as being open ended and may be interchangeable; a noun as being either plural or singular where one or more individual elements are often used to form a set of one or more elements (e g., metasurface(s), or circuit versus circuitry), in context the noun may be taken to generally refer to either the plural or the singular; and where a type of structure is delineated by adjectives such as “first” and “second”, such adjectives are not used to connote any description of the structure or to provide any substantive meaning.

[0064] In certain of the embodiments discussed herein, one or more modules are discrete logic circuits or programmable logic circuits configured and arranged for implementing these operations / activities (e.g., ML and / or Al), as may be carried out in the approaches shown in the figures of the Appendices of the underlying provisional application. Also, certain specific examples relating to the above-described aspects may be directed to a computer program product (e.g.. non-volatile memory’ device), which includes a machine or computer-readable medium having stored thereon instructions which may be executed by a computer (or other electronic device) to perform these operations / activities. In certain such CPU-related embodiments, a programmable circuit may be used as one or more computer (data-processing) circuits, including memory circuitry for storing and accessing a program to be executed as a set (or sets) of instructions (and / or to be used as configuration data to define hoyv the programmable circuit is to perform), and an algorithm or process as described above is used by the programmable circuit to perform the related steps, functions, operations,STFD.468PCT (S24-426) 21activities, etc. Depending on the application, the instructions (and / or configuration data) can be configured for implementation in logic circuitry, with the instructions (whether characterized in the form of object code, firmware or software) stored in and accessible from a memory' (circuit).[0065 j Based upon the above discussion and illustrations, those skilled in the art will readily recognize that various modifications and changes may be made to the various embodiments without strictly following the exemplary embodiments and applications illustrated and described herein. For example, methods as exemplified in the Figures may involve steps carried out in various orders, with one or more aspects of the embodiments herein retained, or may involve fewer or more steps. Such modifications do not depart from the true spirit and scope of various aspects of the disclosure, including aspects set forth in the claims.

Claims

STFD.468PCT (S24-426) 22What is Claimed:

1. An apparatus comprising:a chamber to provide wastewater;a power supply to provide electric field energy: andan optically-responsive material, from among one or more of at least one nanomaterial and at least one metasurface, to respond to the field energy from the power supply by causing light signals that scatter from targeted bacteria to be highlighted, relative to light signals from non-targeted contaminants, in the wastewater.

2. The apparatus of claim 1, wherein the power supply is to generate a non-uniform electric field and cause particles in the targeted bacteria to become polarized, and to pose directional forces on the targeted bacteria with specific sizes and dielectric properties.

3. The apparatus of claim 1, wherein the optically-responsive material is optically resonant.

4. The apparatus of claim 1, wherein the targeted bacteria is label-free bacteria.

5. The apparatus of claim 1, further including an electrokinetic trap to facilitate or increase the light signals that scatter from targeted bacteria concentration and selectivity, as indicated by the targeted bacteria to be highlighted, for differentiating the targeted bacteria.

6. The apparatus of claim 1, further including a filter to identify or differentiate the light signals without relying on one or more labels, from among external fluorescent dyes, stains and molecular tags.

7. The apparatus of claim 1, wherein the electric field energy' is to increase concentration and selectivity for the targeted bacteria by attracting bacteria to a region selectively for one or more specific bacterial species.STFD.468PCT (S24-426) 238. The apparatus of claim 1, wherein the electric field energy is to increase concentration and selectivity for the targeted bacteria by attracting bacteria to a region collectively for one or more bacterial species relative to or against other species from among chemical species and bacterial species.

9. The apparatus of claim 1, wherein the electric field energy is to increase concentration and selectivity for the targeted bacteria by attracting bacteria by displacing and enriching the targeted bacteria sufficiently proximal a sensor surface for detection.

10. The apparatus of claim 1, wherein the electric field energy is to present application of non-uniform electric fields to drive bacterial movement towards a point of detection, based on Brownian motions of the targeted bacteria, by amplifying differentiation of the light signals, relative to the non-targeted contaminants in the wastewater, for increased Raman spectroscopy detection.

11. The apparatus of claim 1, further including a filter to identify or differentiate the light signals in response to response to the field energy, thereby differentiating the targeted bacteria from the non-targeted contaminants in the wastewater, and wherein the field energy is dielectrophoresis-type field energy.

12. The apparatus of claim 1, wherein the targeted bacteria is label-free bacteria, and the optically -responsive material is optically resonant, and the field energy is dielectrophoresistype field energy.

13. The apparatus of claim 1, wherein the optically-responsive material include particles each of which is characterized, in part, by a physical structure having one or more subwavelength dimensions, by causing light to concentrate on membranes of the targeted bacteria.

14. The apparatus of claim 1, wherein the optically-responsive material includes optically resonant patterned nanoantennas to improve Raman scattering efficiency and signal-to-noise ratios on biological samples obtained from a collection region affected by the wastewater.STFD.468PCT (S24-426) 2415. The apparatus of claim 1, wherein the optically-responsive material includes a plurality of nanoantennas to increase or facilitate surface-enhanced Raman scattering.

16. A method comprising:using a chamber to hold wastewater.applying field energy towards optically-responsive material, as part of the wastewater in the chamber; andcausing optically-responsive material, from among one or more of at least one nanomaterial and at least one metasurface, to respond to the field energy by causing light signals that scatter from targeted bacteria to be highlighted, relative to light signals from non-targeted contaminants, in the wastewater.

17. The method of claim 16, further including, after said causing light signals that scatter from targeted bacteria to be highlighted, discerning whether the targeted bacteria are associated with expected spectral features, indicated by external data as corresponding to certain of the targeted bacteria, by perturbing spectra as part of the field energy, and by¬ assessing effects of the perturbing spectra, relative to the external data.

18. The method of claim 16, wherein the optically-responsive material is optically resonant.

19. The method of claim 16, further including differentiating or identifying, via a filter, without relying on one or more labels, from among external fluorescent dyes, stains and molecular tags.

20. The method of claim 16, further including differentiating or identifying by using at least one of an artificial-intelligence computer and one or more Raman spectroscopy tools.

Images