Accurate virus identification with interpretable Raman signatures by machine learning.

Rapid identification of newly emerging or circulating viruses is an important first step toward managing the public health response to potential outbreaks. A portable virus capture device, coupled with label-free Raman spectroscopy, holds the promise of fast detection by rapidly obtaining the Raman signature of a virus followed by a machine learning (ML) approach applied to recognize the virus based on its Raman spectrum, which is used as a fingerprint. We present such an ML approach for analyzing Raman spectra of human and avian viruses. A convolutional neural network (CNN) classifier specifically designed for spectral data achieves very high accuracy for a variety of virus type or subtype identification tasks. In particular, it achieves 99% accuracy for classifying influenza virus type A versus type B, 96% accuracy for classifying four subtypes of influenza A, 95% accuracy for differentiating enveloped and nonenveloped viruses, and 99% accuracy for differentiating avian coronavirus (infectious bronchitis virus [IBV]) from other avian viruses. Furthermore, interpretation of neural net responses in the trained CNN model using a full-gradient algorithm highlights Raman spectral ranges that are most important to virus identification. By correlating ML-selected salient Raman ranges with the signature ranges of known biomolecules and chemical functional groups­for example, amide, amino acid, and carboxylic acid­we verify that our ML model effectively recognizes the Raman signatures of proteins, lipids, and other vital functional groups present in different viruses and uses a weighted combination of these signatures to identify viruses.

Angstrom-Size Defect Creation and Ionic Transport through Pores in Single-Layer MoS2.

Atomic-defect engineering in thin membranes provides opportunities for ionic and molecular filtration and analysis. While molecular-dynamics (MD) calculations have been used to model conductance through atomic vacancies, corresponding experiments are lacking. We create sub-nanometer vacancies in suspended single-layer molybdenum disulfide (MoS2) via Ga+ ion irradiation, producing membranes containing ∼300 to 1200 pores with average and maximum diameters of ∼0.5 and ∼1 nm, respectively. Vacancies exhibit missing Mo and S atoms, as shown by aberration-corrected scanning transmission electron microscopy (AC-STEM). The longitudinal acoustic band and defect-related photoluminescence were observed in Raman and photoluminescence spectroscopy, respectively. As the irradiation dose is increased, the median vacancy area remains roughly constant, while the number of vacancies (pores) increases. Ionic current versus voltage is nonlinear and conductance is comparable to that of ∼1 nm diameter single MoS2 pores, proving that the smaller pores in the distribution display negligible conductance. Consistently, MD simulations show that pores with diameters <0.6 nm are almost impermeable to ionic flow. Atomic pore structure and geometry, studied by AC-STEM, are critical in the sub-nanometer regime in which the pores are not circular and the diameter is not well-defined. This study lays the foundation for future experiments to probe transport in large distributions of angstrom-size pores.

Ultrasensitive gas detection of large-area boron-doped graphene.

Heteroatom doping is an efficient way to modify the chemical and electronic properties of graphene. In particular, boron doping is expected to induce a p-type (boron)-conducting behavior to pristine (nondoped) graphene, which could lead to diverse applications. However, the experimental progress on atomic scale visualization and sensing properties of large-area boron-doped graphene (BG) sheets is still very scarce. This work describes the controlled growth of centimeter size, high-crystallinity BG sheets. Scanning tunneling microscopy and spectroscopy are used to visualize the atomic structure and the local density of states around boron dopants. It is confirmed that BG behaves as a p-type conductor and a unique croissant-like feature is frequently observed within the BG lattice, which is caused by the presence of boron-carbon trimers embedded within the hexagonal lattice. More interestingly, it is demonstrated for the first time that BG exhibits unique sensing capabilities when detecting toxic gases, such as NO2 and NH3, being able to detect extremely low concentrations (e.g., parts per trillion, parts per billion). This work envisions that other attractive applications could now be explored based on as-synthesized BG.

Ultrashort optical pulse characterization using WS2 monolayers.

We demonstrate the application of two-dimensional materials for ultrashort optical pulse characterization. Monolayer transition metal dichalcogenides, such as tungsten disulfide (WS2), possess extraordinarily large second-order nonlinear susceptibility, and due to their atomic thickness, have relaxed phase-matching requirements and, hence, an inherently wide bandwidth. Synthesized monolayer WS2 triangular islands were used to characterize ultrashort optical pulses at the focal point of an objective lens through second-harmonic generation collinear frequency-resolved optical gating.

Extraordinary room-temperature photoluminescence in triangular WS2 monolayers.

Individual monolayers of metal dichalcogenides are atomically thin two-dimensional crystals with attractive physical properties different from those of their bulk counterparts. Here we describe the direct synthesis of WS2 monolayers with triangular morphologies and strong room-temperature photoluminescence (PL). The Raman response as well as the luminescence as a function of the number of S-W-S layers is also reported. The PL weakens with increasing number of layers due to a transition from direct band gap in a monolayer to indirect gap in multilayers. The edges of WS2 monolayers exhibit PL signals with extraordinary intensity, around 25 times stronger than that at the platelet's center. The structure and chemical composition of the platelet edges appear to be critical for PL enhancement.

Rapid and label-free influenza A virus subtyping using surface-enhanced Raman spectroscopy with incident-wavelength analysis.

Early virus identification is a key component of both patient treatment and epidemiological monitoring. In the case of influenza A virus infections, where the detection of subtypes associated with bird flu in humans could lead to a pandemic, rapid subtype-level identification is important. Surface-enhanced Raman spectroscopy coupled with machine learning can be used to rapidly detect and identify viruses in a label-free manner. As there is a range of available excitation wavelengths for performing Raman spectroscopy, we must choose the best one to permit discrimination between highly similar subtypes of a virus. We show that the spectra produced by influenza A subtypes H1N1 and H3N2 exhibit a higher degree of dissimilarity when using 785 nm excitation wavelength in comparison with 532 nm excitation wavelength. Furthermore, the cross-validated area under the curve (AUC) for identification was higher for the 785 nm excitation, reaching 0.95 as compared to 0.86 for 532 nm. Ultimately, this study suggests that exciting with a 785 nm wavelength is better able to differentiate two closely related influenza viruses and likely can extend to other closely related pathogens.

Effective plant virus enrichment using carbon nanotubes and microfluidics.

Plant virus detection and identification in crops is a pillar for disease management, import of crop material, production of clean stock plants and basic plant virology studies. In this report, we present a platform for the enrichment and isolation of known or unknown viruses. This platform is based on carbon nanotube arrays inside a microfluidic device that can be a solution for the identification of low titer viruses from plants. Using our microfluidic devices, we achieved enrichment of two economically important viruses, the orthotospovirus, tomato spotted wilt orthotospovirus (TSWV) and the potyvirus, zucchini yellow mosaic virus (ZYMV). The carbon nanotube arrays integrated in these microfluidic devices are capable of trapping viruses discriminated by their size; the virus rich arrays can be then analyzed by common downstream techniques including immunoassays, PCR, HTS and electron microscopy. This procedure offers a simple to operate and portable sample preparation device capable of trapping viruses from raw plant extracts while reducing the host contamination.

Engineering Vacancies for the Creation of Antisite Defects in Chemical Vapor Deposition Grown Monolayer MoS2 and WS2 via Proton Irradiation.

It is critical to understand the laws of quantum mechanics in transformative technologies for computation and quantum information science applications to enable the ongoing second quantum revolution calls. Recently, spin qubits based on point defects have gained great attention, since these qubits can be initiated, selectively controlled, and read out with high precision at ambient temperature. The major challenge in these systems is controllably generating multiqubit systems while properly coupling the defects. To address this issue, we began by tackling the engineering challenges these systems present and understanding the fundamentals of defects. In this regard, we controllably generate defects in MoS2 and WS2 monolayers and tune their physicochemical properties via proton irradiation. We quantitatively discovered that the proton energy could modulate the defects' density and nature; higher defect densities were seen with lower proton irradiation energies. Three distinct defect types were observed: vacancies, antisites, and adatoms. In particular, the creation and manipulation of antisite defects provides an alternative way to create and pattern spin qubits based on point defects. Our results demonstrate that altering the particle irradiation energy can regulate the formation of defects, which can be utilized to modify the properties of 2D materials and create reliable electronic devices.

Substrate-Induced Changes on the Optical Properties of Single-Layer WS2.

Among the most studied semiconducting transition metal dichalcogenides (TMDCs), WS2 showed several advantages in comparison to their counterparts, such as a higher quantum yield, which is an important feature for quantum emission and lasing purposes. We studied transferred monolayers of WS2 on a drilled Si3N4 substrate in order to have insights about on how such heterostructure behaves from the Raman and photoluminescence (PL) measurements point of view. Our experimental findings showed that the Si3N4 substrate influences the optical properties of single-layer WS2. Beyond that, seeking to shed light on the causes of the PL quenching observed experimentally, we developed density functional theory (DFT) based calculations to study the thermodynamic stability of the heterojunction through quantum molecular dynamics (QMD) simulations as well as the electronic alignment of the energy levels in both materials. Our analysis showed that along with strain, a charge transfer mechanism plays an important role for the PL decrease.

Understanding the Excitation Wavelength Dependence and Thermal Stability of the SARS-CoV-2 Receptor-Binding Domain Using Surface-Enhanced Raman Scattering and Machine Learning.

COVID-19 has cost millions of lives worldwide. The constant mutation of SARS-CoV-2 calls for thorough research to facilitate the development of variant surveillance. In this work, we studied the fundamental properties related to the optical identification of the receptor-binding domain (RBD) of SARS-CoV-2 spike protein, a key component of viral infection. The Raman modes of the SARS-CoV-2 RBD were captured by surface-enhanced Raman spectroscopy (SERS) using gold nanoparticles (AuNPs). The observed Raman enhancement strongly depends on the excitation wavelength as a result of the aggregation of AuNPs. The characteristic Raman spectra of RBDs from SARS-CoV-2 and MERS-CoV were analyzed by principal component analysis that reveals the role of secondary structures in the SERS process, which is corroborated with the thermal stability under laser heating. We can easily distinguish the Raman spectra of two RBDs using machine learning algorithms with accuracy, precision, recall, and F1 scores all over 95%. Our work provides an in-depth understanding of the SARS-CoV-2 RBD and paves the way toward rapid analysis and discrimination of complex proteins of infectious viruses and other biomolecules.

Interaction of gases with monolayer WS2: an in situ spectroscopy study.

The optical and electronic properties of two-dimensional (2D) materials can be tuned through physical and chemical adsorption of gases. They are also ideal sensor platforms, where charge transfer from the adsorbate can induce a measurable change in the electrical resistance within a device configuration. While 2D materials-based gas sensors exhibit high sensitivity, questions exist regarding the direction of charge transfer and the role of lattice defects during sensing. Here we measured the dynamics of adsorption of NO2 and NH3 on monolayer WS2 using in situ photoluminescence (PL) and resonance Raman spectroscopy. Experiments were conducted across a temperature range of 25-250 °C and gas concentrations between 5-650 ppm. The PL emission energies blue- and red-shifted when exposed to NO2 and NH3, respectively, and the magnitude of the shift depended on the gas concentration as well as the temperature down to the lowest concentration of 5 ppm. Analysis of the adsorption kinetics revealed an exponential increase in the intensities of the trion peaks with temperature, with apparent activation energies similar to barriers for migration of sulfur vacancies in the WS2 lattice. The corresponding Resonance Raman spectra allowed the simultaneous measurement of the defect-induced LA mode. A positive correlation between the defect densities and the shifts in the PL emission energies establish lattice defects such as sulfur vacancies as the preferential sites for gas adsorption. Moreover, an increase in defect densities with temperature in the presence of NO2 and NH3 suggests that these gases may also play a role in the creation of lattice defects. Our study provides key mechanistic insights into gas adsorption on monolayer WS2, and highlights the potential for future development of spectroscopy-based gas sensors based on 2D materials.

Quantification and Healing of Defects in Atomically Thin Molybdenum Disulfide: Beyond the Controlled Creation of Atomic Defects.

Atomically thin 2D materials provide an opportunity to investigate the atomic-scale details of defects introduced by particle irradiation. Once the atomic configuration of defects and their spatial distribution are revealed, the details of the mesoscopic phenomena can be unveiled. In this work, we created atomically small defects by controlled irradiation of gallium ions with doses ranging from 4.94 × 1012 to 4.00 × 1014 ions/cm2 on monolayer molybdenum disulfide (MoS2) crystals. The optical signatures of defects, such as the evolution of defect-activated LA-bands and a broadening of the first-order (E' and A'1) modes, can be studied by Raman spectroscopy. High-resolution scanning transmission electron microscopy (HR-STEM) analysis revealed that most defects are vacancies of few-molybdenum atoms with surrounding sulfur atoms (VxMo+yS) at a low ion dose. When increasing the ion dose, the atomic vacancies merge and form nanometer-sized holes. Utilizing HR-STEM and image analysis, we propose the estimation of the finite crystal length (Lfc) via the careful quantification of 0D defects in 2D systems through the formula Lfc = 4.41/ηion, where ηion corresponds to the ion dose. Combining HR-STEM and Raman spectroscopy, the formula to calculate Lfc from Raman features, I(LA)/I(A'1) = 5.09/Lfc2, is obtained. We have also demonstrated an effective route to healing the ion irradiation-induced atomic vacancies by annealing defective MoS2 in a hydrogen disulfide (H2S) atmosphere. The H2S annealing improved the crystal quality of MoS2 with Lfc greater than the calculated size of the A exciton wave function, which leads to a partial recovery of the photoluminescence signal after its quenching by ion irradiation.

Confined Crack Propagation in MoS2 Monolayers by Creating Atomic Vacancies.

In two-dimensional crystals, fractures propagate easily, thus restricting their mechanical reliability. This work demonstrates that controlled defect creation constitutes an effective approach to avoid catastrophic failure in MoS2 monolayers. A systematic study of fracture mechanics in MoS2 monolayers as a function of the density of atomic vacancies, created by ion irradiation, is reported. Pristine and irradiated materials were studied by atomic force microscopy, high-resolution scanning transmission electron microscopy, and Raman spectroscopy. By inducing ruptures through nanoindentations, we determine the strength and length of the propagated cracks within MoS2 atom-thick membranes as a function of the density and type of the atomic vacancies. We find that a 0.15% atomic vacancy induces a decrease of 40% in strength with respect to that of pristine samples. In contrast, while tear holes in pristine 2D membranes span several microns, they are restricted to a few nanometers in the presence of atomic and nanometer-sized vacancies, thus increasing the material's fracture toughness.

Defect creation in WSe2 with a microsecond photoluminescence lifetime by focused ion beam irradiation.

Defect engineering is important for tailoring the electronic and optical properties of two-dimensional materials, and the capability of generating defects of certain types at specific locations is meaningful for potential applications such as optoelectronics and quantum photonics. In this work, atomic defects are created in single-layer WSe2 using focused ion beam (FIB) irradiation, with defect densities spanning many orders of magnitude. The influences of defects are systematically characterized. Raman spectroscopy can only discern defects in WSe2 for a FIB dose higher than 1 × 1013 cm-2, which causes blue shifts of both A'1 and E' modes. Photoluminescence (PL) of WSe2 is more sensitive to defects. At cryogenic temperature, the low-energy PL induced by defects can be revealed, which shows redshifts and broadenings with increased FIB doses. Similar Raman shifts and PL spectrum changes are observed for the WSe2 film grown by chemical vapor deposition (CVD). A four microsecond-long lifetime is observed in the PL dynamics and is three orders of magnitude longer than the often observed delocalized exciton lifetime and becomes more dominant for WSe2 with increasing FIB doses. The ultra-long lifetime of PL in single-layer WSe2 is consistent with first-principles calculation results considering the creation of both chalcogen and metal vacancies by FIB, and can be valuable for photo-catalytic reactions, valleytronics and quantum light emissions owing to the longer carrier separation/manipulation time.

Atomically Thin Layers of Graphene and Hexagonal Boron Nitride Made by Solvent Exfoliation of Their Phosphoric Acid Intercalation Compounds.

The development of scalable and reliable techniques for the production of the atomically thin layers of graphene and hexagonal boron nitride (h-BN) in bulk quantities could make these materials a powerful platform for devices and composites that impact a wide variety of technologies (Nature 2012, 490, 192-200). To date a number of practical exfoliation methods have been reported that are based on sonicating or stirring powdered graphite or h-BN in common solvents. However, the products of these experiments consist mainly of few-layer sheets and contain only a small fraction of monolayers. A possible reason for this is that splitting the crystals into monolayers starts from solvent intercalation, which must overcome the substantial interlayer cohesive energy (120-720 mJ/m2) of the van der Waals solids. Here we show that the yield of the atomically thin layers can be increased to near unity when stage-1 intercalation compounds of phosphoric acid are used as starting materials. The exfoliation to predominantly monolayers was achieved by stirring them in medium polarity organic solvents that can form hydrogen bonds. The exfoliation process does not disrupt the sp2 π-system of graphene and is gentle enough to allow the preparation of graphene and h-BN monolayers that are tens of microns in their lateral dimensions.

Optical identification of sulfur vacancies: Bound excitons at the edges of monolayer tungsten disulfide.

Defects play a significant role in tailoring the optical properties of two-dimensional materials. Optical signatures of defect-bound excitons are important tools to probe defective regions and thus interrogate the optical quality of as-grown semiconducting monolayer materials. We have performed a systematic study of defect-bound excitons using photoluminescence (PL) spectroscopy combined with atomically resolved scanning electron microscopy and first-principles calculations. Spatially resolved PL spectroscopy at low temperatures revealed bound excitons that were present only on the edges of monolayer tungsten disulfide and not in the interior. Optical pumping of the bound excitons was sublinear, confirming their bound nature. Atomic-resolution images reveal that the areal density of monosulfur vacancies is much larger near the edges (0.92 ± 0.45 nm-2) than in the interior (0.33 ± 0.11 nm-2). Temperature-dependent PL measurements found a thermal activation energy of ~36 meV; surprisingly, this is much smaller than the bound-exciton binding energy of ~300 meV. We show that this apparent inconsistency is related to a thermal dissociation of the bound exciton that liberates the neutral excitons from negatively charged point defects. First-principles calculations confirm that sulfur monovacancies introduce midgap states that host optical transitions with finite matrix elements, with emission energies ranging from 200 to 400 meV below the neutral-exciton emission line. These results demonstrate that bound-exciton emission induced by monosulfur vacancies is concentrated near the edges of as-grown monolayer tungsten disulfide.

Low-temperature Synthesis of Heterostructures of Transition Metal Dichalcogenide Alloys (WxMo1-xS2) and Graphene with Superior Catalytic Performance for Hydrogen Evolution.

Large-area (∼cm2) films of vertical heterostructures formed by alternating graphene and transition-metal dichalcogenide (TMD) alloys are obtained by wet chemical routes followed by a thermal treatment at low temperature. In particular, we synthesized stacked graphene and WxMo1-xS2 alloy phases that were used as hydrogen evolution catalysts. We observed a Tafel slope of 38.7 mV dec-1 and 96 mV onset potential (at current density of 10 mA cm-2) when the heterostructure alloy was annealed at 300 °C. These results indicate that heterostructures formed by graphene and W0.4Mo0.6S2 alloys are far more efficient than WS2 and MoS2 by at least a factor of 2, and they are superior compared to other reported TMD systems. This strategy offers a cheap and low temperature synthesis alternative able to replace Pt in the hydrogen evolution reaction (HER). Furthermore, the catalytic activity of the alloy is stable over time, i.e., the catalytic activity does not experience a significant change even after 1000 cycles. Using density functional theory calculations, we found that this enhanced hydrogen evolution in the WxMo1-xS2 alloys is mainly due to the lower energy barrier created by a favorable overlap of the d-orbitals from the transition metals and the s-orbitals of H2; with the lowest energy barrier occurring for the W0.4Mo0.6S2 alloy. Thus, it is now possible to further improve the performance of the "inert" TMD basal plane via metal alloying, in addition to the previously reported strategies such as creation of point defects, vacancies and edges. The synthesis of graphene/W0.4Mo0.6S2 produced at relatively low temperatures is scalable and could be used as an effective low cost Pt-free catalyst.

Tunable and label-free virus enrichment for ultrasensitive virus detection using carbon nanotube arrays.

Viral infectious diseases can erupt unpredictably, spread rapidly, and ravage mass populations. Although established methods, such as polymerase chain reaction, virus isolation, and next-generation sequencing have been used to detect viruses, field samples with low virus count pose major challenges in virus surveillance and discovery. We report a unique carbon nanotube size-tunable enrichment microdevice (CNT-STEM) that efficiently enriches and concentrates viruses collected from field samples. The channel sidewall in the microdevice was made by growing arrays of vertically aligned nitrogen-doped multiwalled CNTs, where the intertubular distance between CNTs could be engineered in the range of 17 to 325 nm to accurately match the size of different viruses. The CNT-STEM significantly improves detection limits and virus isolation rates by at least 100 times. Using this device, we successfully identified an emerging avian influenza virus strain [A/duck/PA/02099/2012(H11N9)] and a novel virus strain (IBDV/turkey/PA/00924/14). Our unique method demonstrates the early detection of emerging viruses and the discovery of new viruses directly from field samples, thus creating a universal platform for effectively remediating viral infectious diseases.

Ultrasensitive molecular sensor using N-doped graphene through enhanced Raman scattering.

As a novel and efficient surface analysis technique, graphene-enhanced Raman scattering (GERS) has attracted increasing research attention in recent years. In particular, chemically doped graphene exhibits improved GERS effects when compared with pristine graphene for certain dyes, and it can be used to efficiently detect trace amounts of molecules. However, the GERS mechanism remains an open question. We present a comprehensive study on the GERS effect of pristine graphene and nitrogen-doped graphene. By controlling nitrogen doping, the Fermi level (E F) of graphene shifts, and if this shift aligns with the lowest unoccupied molecular orbital (LUMO) of a molecule, charge transfer is enhanced, thus significantly amplifying the molecule's vibrational Raman modes. We confirmed these findings using different organic fluorescent molecules: rhodamine B, crystal violet, and methylene blue. The Raman signals from these dye molecules can be detected even for concentrations as low as 10(-11) M, thus providing outstanding molecular sensing capabilities. To explain our results, these nitrogen-doped graphene-molecule systems were modeled using dispersion-corrected density functional theory. Furthermore, we demonstrated that it is possible to determine the gaps between the highest occupied and the lowest unoccupied molecular orbitals (HOMO-LUMO) of different molecules when different laser excitations are used. Our simulated Raman spectra of the molecules also suggest that the measured Raman shifts come from the dyes that have an extra electron. This work demonstrates that nitrogen-doped graphene has enormous potential as a substrate when detecting low concentrations of molecules and could also allow for an effective identification of their HOMO-LUMO gaps.