Five near-infrared-emissive graphene quantum dots for multiplex bioimaging.

Due to high tissue penetration depth and low autofluorescence backgrounds, near-infrared (NIR) fluorescence imaging has recently become an advantageous diagnostic technique used in a variety of fields. However, most of the NIR fluorophores do not have therapeutic delivery capabilities, exhibit low photostabilities, and raise toxicity concerns. To address these issues, we developed and tested five types of biocompatible graphene quantum dots (GQDs) exhibiting spectrally-separated fluorescence in the NIR range of 928-1053 nm with NIR excitation. Their optical properties in the NIR are attributed to either rare-earth metal dopants (Ho-NGQDs, Yb-NGQDs, Nd-NGQDs) or defect-states (nitrogen doped GQDS (NGQDs), reduced graphene oxides) as verified by Hartree-Fock calculations. Moderate up to 1.34% quantum yields of these GQDs are well-compensated by their remarkable >4 h photostability. At the biocompatible concentrations of up to 0.5-2 mg ml-1 GQDs successfully internalize into HEK-293 cells and enable in vitro imaging in the visible and NIR. Tested all together in HEK-293 cells five GQD types enable simultaneous multiplex imaging in the NIR-I and NIR-II shown for the first time in this work for GQD platforms. Substantial photostability, spectrally-separated NIR emission, and high biocompatibility of five GQD types developed here suggest their promising potential in multianalyte testing and multiwavelength bioimaging of combination therapies.

In Vitro Prostate Cancer Treatment via CRISPR-Cas9 Gene Editing Facilitated by Polyethyleneimine-Derived Graphene Quantum Dots.

CRISPR-Cas9 is a programmable gene editing tool with a promising potential for cancer gene therapy. This therapeutic function is enabled in the present work via the non-covalent delivery of CRISPR ribonucleic protein (RNP) by cationic glucosamine/PEI-derived graphene quantum dots (PEI-GQD) that aid in overcoming physiological barriers and tracking genes of interest. PEI-GQD/RNP complex targeting the TP53 mutation overexpressed in ~50% of cancers successfully produces its double-stranded breaks in solution and in PC3 prostate cancer cells. Restoring this cancer "suicide" gene can promote cellular repair pathways and lead to cancer cell apoptosis. Its repair to the healthy form performed by simultaneous PEI-GQD delivery of CRISPR RNP and a gene repair template leads to a successful therapeutic outcome: 40% apoptotic cancer cell death, while having no effect on non-cancerous HeK293 cells. The translocation of PEI-GQD/RNP complex into PC3 cell cytoplasm is tracked via GQD intrinsic fluorescence, while EGFP-tagged RNP is detected in the cell nucleus, showing the successful detachment of the gene editing tool upon internalization. Using GQDs as non-viral delivery and imaging agents for CRISPR-Cas9 RNP sets the stage for image-guided cancer-specific gene therapy.

Doped Graphene Quantum Dots as Biocompatible Radical Scavenging Agents.

Oxidative stress is proven to be a leading factor in a multitude of adverse conditions, from Alzheimer's disease to cancer. Thus, developing effective radical scavenging agents to eliminate reactive oxygen species (ROS) driving many oxidative processes has become critical. In addition to conventional antioxidants, nanoscale structures and metal-organic complexes have recently shown promising potential for radical scavenging. To design an optimal nanoscale ROS scavenging agent, we have synthesized ten types of biocompatible graphene quantum dots (GQDs) augmented with various metal dopants. The radical scavenging abilities of these novel metal-doped GQD structures were, for the first time, assessed via the DPPH, KMnO4, and RHB (Rhodamine B protectant) assays. While all metal-doped GQDs consistently demonstrate antioxidant properties higher than the undoped cores, aluminum-doped GQDs exhibit 60-95% radical scavenging ability of ascorbic acid positive control. Tm-doped GQDs match the radical scavenging properties of ascorbic acid in the KMnO4 assay. All doped GQD structures possess fluorescence imaging capabilities that enable their tracking in vitro, ensuring their successful cellular internalization. Given such multifunctionality, biocompatible doped GQD antioxidants can become prospective candidates for multimodal therapeutics, including the reduction of ROS with concomitant imaging and therapeutic delivery to cancer tumors.

Cancer Therapeutic siRNA Delivery and Imaging by Nitrogen- and Neodymium-Doped Graphene Quantum Dots.

While small interfering RNA (siRNA) technology has become a powerful tool that can enable cancer-specific gene therapy, its translation to the clinic is still hampered by the inability of the genes alone to cell transfection, poor siRNA stability in blood, and the lack of delivery tracking capabilities. Recently, graphene quantum dots (GQDs) have emerged as a novel platform allowing targeted drug delivery and fluorescence image tracking in visible and near-infrared regions. These capabilities can aid in overcoming primary obstacles to siRNA therapeutics. Here, for the first time, we utilize biocompatible nitrogen- and neodymium-doped graphene quantum dots (NGQDs and Nd-NGQDs, respectively) for the delivery of Kirsten rat sarcoma virus (KRAS) and epidermal growth factor receptor (EGFR) siRNA effective against a variety of cancer types. GQDs loaded with siRNA noncovalently facilitate successful siRNA transfection into HeLa cells, confirmed by confocal fluorescence microscopy at biocompatible GQD concentrations of 375 µg/mL. While the GQD platform provides visible fluorescence tracking, Nd doping enables deeper-tissue near-infrared fluorescence imaging suitable for both in vitro and in vivo applications. The therapeutic efficacy of the GQD/siRNA complex is verified by successful protein knockdown in HeLa cells at nanomolar siEGFR and siKRAS concentrations. A range of GQD/siRNA loading ratios and payloads are tested to ultimately provide substantial inhibition of protein expression down to 31-45%, comparable with conventional Lipofectamine-mediated delivery. This demonstrates the promising potential of GQDs for the nontoxic delivery of siRNA and genes in general, complemented by multiwavelength image tracking.

Automated Approach to In Vitro Image-Guided Photothermal Therapy with Top-Down and Bottom-Up-Synthesized Graphene Quantum Dots.

Graphene-based materials have been the subject of interest for photothermal therapy due to their high light-to-heat conversion efficiency. Based on recent studies, graphene quantum dots (GQDs) are expected to possess advantageous photothermal properties and facilitate fluorescence image-tracking in the visible and near-infrared (NIR), while surpassing other graphene-based materials in their biocompatibility. Several GQD structures including reduced graphene quantum dots (RGQDs) derived from reduced graphene oxide via top-down oxidation and hyaluronic acid graphene quantum dots (HGQDs) hydrothermally bottom-up synthesized from molecular hyaluronic acid were employed to test these capabilities in the present work. These GQDs possess substantial NIR absorption and fluorescence throughout the visible and NIR beneficial for in vivo imaging while being biocompatible at up to 1.7 mg/mL concentrations. In aqueous suspensions, RGQDs and HGQDs irradiated with a low power (0.9 W/cm2) 808 nm NIR laser facilitate a temperature increase up to 47.0 °C, which is sufficient for cancer tumor ablation. In vitro photothermal experiments sampling multiple conditions directly in the 96-well plate were performed using an automated simultaneous irradiation/measurement system developed on the basis of a 3D printer. In this study, HGQDs and RGQDs facilitated the heating of HeLa cancer cells up to 54.5 °C, leading to the drastic inhibition of cell viability from over 80% down to 22.9%. GQD's fluorescence in the visible and NIR traces their successful internalization into HeLa cells maximized at 20 h suggesting both extracellular and intracellular photothermal treatment capabilities. The combination of the photothermal and imaging modalities tested in vitro makes the GQDs developed in this work prospective agents for cancer theragnostics.

Dual-Mode Fluorescence/Ultrasound Imaging with Biocompatible Metal-Doped Graphene Quantum Dots.

Sonography offers many advantages over standard methods of diagnostic imaging due to its non-invasiveness, substantial tissue penetration depth, and low cost. The benefits of ultrasound imaging call for the development of ultrasound-trackable drug delivery vehicles that can address a variety of therapeutic targets. One disadvantage of the technique is the lack of high-precision imaging, which can be circumvented by complementing ultrasound contrast agents with visible and, especially, near-infrared (NIR) fluorophores. In this work, we, for the first time, develop a variety of lightly metal-doped (iron oxide, silver, thulium, neodymium, cerium oxide, cerium chloride, and molybdenum disulfide) nitrogen-containing graphene quantum dots (NGQDs) that demonstrate high-contrast properties in the ultrasound brightness mode and exhibit visible and/or near-infrared fluorescence imaging capabilities. NGQDs synthesized from glucosamine precursors with only a few percent metal doping do not introduce additional toxicity in vitro, yielding over 80% cell viability up to 2 mg/mL doses. Their small (<50 nm) sizes warrant effective cell internalization, while oxygen-containing surface functional groups decorating their surfaces render NGQDs water soluble and allow for the attachment of therapeutics and targeting agents. Utilizing visible and/or NIR fluorescence, we demonstrate that metal-doped NGQDs experience maximum accumulation within the HEK-293 cells 6-12 h after treatment. The successful 10-fold ultrasound signal enhancement is observed at 0.5-1.6 mg/mL for most metal-doped NGQDs in the vascular phantom, agarose gel, and animal tissue. A combination of non-invasive ultrasound imaging with capabilities of high-precision fluorescence tracking makes these metal-doped NGQDs a viable agent for a variety of theragnostic applications.

Detection of Pancreatic Cancer miRNA with Biocompatible Nitrogen-Doped Graphene Quantum Dots.

Early-stage pancreatic cancer remains challenging to detect, leading to a poor five-year patient survival rate. This obstacle necessitates the development of early detection approaches based on novel technologies and materials. In this work, the presence of a specific pancreatic cancer-derived miRNA (pre-miR-132) is detected using the fluorescence properties of biocompatible nitrogen-doped graphene quantum dots (NGQDs) synthesized using a bottom-up approach from a single glucosamine precursor. The sensor platform is comprised of slightly positively charged (1.14 ± 0.36 mV) NGQDs bound via π-π stacking and/or electrostatic interactions to the negatively charged (-22.4 ± 6.00 mV) bait ssDNA; together, they form a complex with a 20 nm average size. The NGQDs' fluorescence distinguishes specific single-stranded DNA sequences due to bait-target complementarity, discriminating them from random control sequences with sensitivity in the micromolar range. Furthermore, this targetability can also detect the stem and loop portions of pre-miR-132, adding to the practicality of the biosensor. This non-invasive approach allows cancer-specific miRNA detection to facilitate early diagnosis of various forms of cancer.

Nanoporous Silicon with Graphene-like Coating for Pseudocapacitor Application.

This paper presents the results of studies of the nanoporous silicon structure, both with different pore depths (up to 180 µm) and with layers in which a graphene-like coating was synthesized on the inner surface of the pores. The nanoporous layers were characterized by SEM as well as IR and Raman spectroscopy. Cyclic voltammetry and galvanostatic charge-discharge data in 3 M H2SO4 are presented as well as the results of the cyclic stability of these characteristics for the nanoporous structure. It was found that the degree of electrolyte pre-impregnation significantly affected the electrochemical processes, and the capacitance values depended on the depth (thickness) of the nanoporous layer. Increasing the thickness of the porous layer led to an increase in area-normalized pseudocapacity and was limited only by the mechanical strength of the structure. Performance improvement was also achieved by synthesis of the graphene-like layer in the volume of the nanoporous structure. The electrodes (composite materials) proposed in the work showed one of the best capacitive characteristics (87 mF/cm2 with 100% capacity retention after 15,000 cycles) in comparison with the data reported in the literature at present.

Using mathematical modeling to estimate time-independent cancer chemotherapy efficacy parameters.

One of the primary cancer treatment modalities is chemotherapy. Unfortunately, traditional anti-cancer drugs are often not selective and cause damage to healthy cells, leading to serious side effects for patients. For this reason more targeted therapeutics and drug delivery methods are being developed. The effectiveness of new treatments is initially determined via in vitro cell viability assays, which determine the IC 50  of the drug. However, these assays are known to result in estimates of IC 50  that depend on the measurement time, possibly resulting in over- or under-estimation of the IC 50 . Here, we test the possibility of using cell growth curves and fitting of mathematical models to determine the IC 50  as well as the maximum efficacy of a drug ( ε max ). We measured cell growth of MCF-7 and HeLa cells in the presence of different concentrations of doxorubicin and fit the data with a logistic growth model that incorporates the effect of the drug. This method leads to measurement time-independent estimates of IC 50  and ε max , but we find that ε max  is not identifiable. Further refinement of this methodology is needed to produce uniquely identifiable parameter estimates.

Effect of Different Tool Probe Profiles on Material Flow of Al-Mg-Cu Alloy Joined by Friction Stir Welding.

Friction Stir Welding (FSW) was utilized to butt-join 2024-T4 aluminum alloy plates of 1.9 mm thickness, using tools with conical and tapered hexagonal probe profiles. The characteristic effects of FSW using tools with tapered hexagonal probe profiles include an increase in the heat input and a significant modification of material flow, which have a positive effect on the metallurgical characteristics and mechanical performance of the weld. The differences in mechanical properties were interpreted through macrostructural changes and mechanical properties of the welded joints, which were supported by numerical simulation results on temperature distribution and material flow. The material flow resulting from the tapered hexagonal probe was more complicated than that of the conical probe. If in the first case, the dynamic viscosity and strain rate are homogeneously distributed around the probe, but in the case of the tapered hexagonal probe tool, the zones with maximum values of strain rates and minimum values of dynamic viscosity are located along the six tapered edges of the probe.

Gas Metal Arc Welding Modes in Wire Arc Additive Manufacturing of Ti-6Al-4V.

In wire arc additive manufacturing of Ti-alloy parts (Ti-WAAM) gas metal arc welding (GMAW) can be applied for complex parts printing. However, due to the specific properties of Ti, GMAW of Ti-alloys is complicated. In this work, three different types of metal transfer modes during Ti-WAAM were investigated: Cold Metal Transfer, controlled short circuiting metal transfer, and self-regulated metal transfer at a direct current with a negative electrode. Metal transfer modes were studied using captured waveform and high-speed video analysis. Using these modes, three walls were manufactured; the geometry preservation stability was estimated and compared using effective wall width calculation, the microstructure was analyzed using optical microscopy. Transfer process data showed that arc wandering depends not only on cathode spot instabilities, but also on anode processing properties. Microstructure analysis showed that each produced wall consists of phases and structures inherent for Ti-WAAM. α-basketweave in the center of and α-colony on the grain boundary of epitaxially grown ß-grains were found with heat affected zone bands along the height of the walls, so that the microstructure did not depend on metal transfer dramatically. However, the geometry preservation stability was higher in the wall, produced with controlled short circuiting metal transfer.

Graphene quantum dot formulation for cancer imaging and redox-based drug delivery.

This work develops a new multifunctional biocompatible anticancer nanoformulation to provide targeted image-guided cancer-selective therapeutics. It consists of three active covalently bound components: (1) biocompatible nitrogen-doped graphene quantum dots (GQDs) as a multifunctional delivery and imaging platform, (2) hyaluronic acid (HA) unit targeted to the CD44 receptors on a variety of cancer cells, and (3) oxidative stress-based cancer-selective ferrocene (Fc) therapeutic. The biocompatible GQD platform synthesized from glucosamine exhibits high-yield intrinsic fluorescence. It is utilized for tracking Fc-GQD-HA formulation in vitro indicating internalization enhancement in HeLa cells targeted by the HA over non-cancer HEK-293 cells not overexpressing CD44 receptor. Fc-GQD-HA, non-toxic at 1 mg/mL to HEK-293 cells, induces cytotoxic response in HeLa enhanced over time, while therapeutic ROS generation by Fc-GQD-HA is ~3 times greater than that of Fc alone. This outlines the targeted delivery, imaging, and cancer-specific treatment capabilities of the new Fc-GQD-HA formulation enabling desired cancer-focused nanotherapeutic approach.

Graphene Quantum Dots as Intracellular Imaging-Based Temperature Sensors.

Non-invasive temperature sensing is necessary to analyze biological processes occurring in the human body, including cellular enzyme activity, protein expression, and ion regulation. To probe temperature-sensitive processes at the nanoscale, novel luminescence nanothermometers are developed based on graphene quantum dots (GQDs) synthesized via top-down (RGQDs) and bottom-up (N-GQDs) approaches from reduced graphene oxide and glucosamine precursors, respectively. Because of their small 3-6 nm size, non-invasive optical sensitivity to temperature change, and high biocompatibility, GQDs enable biologically safe sub-cellular resolution sensing. Both GQD types exhibit temperature-sensitive yet photostable fluorescence in the visible and near-infrared for RGQDs, utilized as a sensing mechanism in this work. Distinctive linear and reversible fluorescence quenching by up to 19.3% is observed for the visible and near-infrared GQD emission in aqueous suspension from 25 °C to 49 °C. A more pronounced trend is observed with GQD nanothermometers internalized into the cytoplasm of HeLa cells as they are tested in vitro from 25 °C to 45 °C with over 40% quenching response. Our findings suggest that the temperature-dependent fluorescence quenching of bottom-up and top-down-synthesized GQDs studied in this work can serve as non-invasive reversible/photostable deterministic mechanisms for temperature sensing in microscopic sub-cellular biological environments.

Impact of Impulses on Microstructural Evolution and Mechanical Performance of Al-Mg-Si Alloy Joined by Impulse Friction Stir Welding.

Impulse Friction Stir Welding (IFSW) was utilized to join 6082-T6 alloy plates at various impulse frequencies. A distinctive feature of IFSW is the generation of mechanical impulses that enhances the forging action of the tool, and thereby, alters the weld microstructure. The microstructural evolution in the Stir Zone (SZ) with special focus on the strengthening precipitation behavior, and overall mechanical properties of the IFSW joints have been investigated. It was demonstrated that the strengthening ß″ precipitates reprecipitated in the SZ of the IFSW joints during natural aging. In contrast, no precipitates were found in the SZ of the Friction Stir Welding (FSW) weld. Partial reversion of ß″ after IFSW is supposed to occur due to more developed subgrain network and higher dislocation density introduced by impulses that accelerated precipitation kinetics. Dynamic recrystallisation was facilitated by impulses resulting in a fine, homogeneous structure. There was no significant difference between the microhardness in the SZ, tensile and yield strength of the FSW and IFSW joints. However, the application of impulses demonstrated the smoothing of the hardness reduction in the transition region at the advancing side. The shift of the fracture location from the Heat-Affected Zone (HAZ) by FSW to the SZ as well as higher elongation of the joints by IFSW of lower frequencies could be related to the grain refinement and the change of the grain orientation.

Effects of Doxorubicin Delivery by Nitrogen-Doped Graphene Quantum Dots on Cancer Cell Growth: Experimental Study and Mathematical Modeling.

With 18 million new cases diagnosed each year worldwide, cancer strongly impacts both science and society. Current models of cancer cell growth and therapeutic efficacy in vitro are time-dependent and often do not consider the Emax value (the maximum reduction in the growth rate), leading to inconsistencies in the obtained IC50 (concentration of the drug at half maximum effect). In this work, we introduce a new dual experimental/modeling approach to model HeLa and MCF-7 cancer cell growth and assess the efficacy of doxorubicin chemotherapeutics, whether alone or delivered by novel nitrogen-doped graphene quantum dots (N-GQDs). These biocompatible/biodegradable nanoparticles were used for the first time in this work for the delivery and fluorescence tracking of doxorubicin, ultimately decreasing its IC50 by over 1.5 and allowing for the use of up to 10 times lower doses of the drug to achieve the same therapeutic effect. Based on the experimental in vitro studies with nanomaterial-delivered chemotherapy, we also developed a method of cancer cell growth modeling that (1) includes an Emax value, which is often not characterized, and (2), most importantly, is measurement time-independent. This will allow for the more consistent assessment of the efficiency of anti-cancer drugs and nanomaterial-delivered formulations, as well as efficacy improvements of nanomaterial delivery.

Experimental and theoretical inquiry into optical properties of graphene derivatives.

Graphene oxide (GO), a functional derivative of graphene, is a promising nanomaterial for a variety of optoelectronic applications as it exhibits fluorescence and maintains many of graphene's beneficial physical properties. although other graphene derivatives are chemically plausible and may serve to the benefit of the aforementioned applications, GO remains the one heavily used. the nature of optical behavior of other graphene derivatives has yet to be fully understood and studied. in this work we develop a variety of graphene derivatives and characterize their optical properties concomitantly suggesting a unified model for optical emission in graphene derivatives. in this process we examine the influence of different functional groups on the surface of graphene on its optoelectronic properties. mildly oxidized graphene (oxo-g1), nitrated graphene, arylated graphene, brominated graphene, and fluorinated graphene are obtained and characterized via TEM and EDX, FTIR and fluorescence spectroscopies with the latter indicating a potential band gap-derived fluorescence from each of the materials. this suggests that optical properties of graphene derivatives have minimal functional group dependence and are manifested by the localized environments within the flakes. this is confirmed by the hyperchem theoretical modeling of all aforementioned graphene derivatives indicating a similar electronic configuration for all, assessed by the pm3 semi-empirical approach. this work can further serve to describe and predict optical properties of similar graphene-based structures and promote graphene derivatives other than GO for utilization in research and industry.

Manganese-nitrogen and gadolinium-nitrogen Co-doped graphene quantum dots as bimodal magnetic resonance and fluorescence imaging nanoprobes.

Graphene quantum dots (GQDs) are unique derivatives of graphene that show promise in multiple biomedical applications as biosensors, bioimaging agents, and drug/gene delivery vehicles. Their ease in functionalization, biocompatibility, and intrinsic fluorescence enable those modalities. However, GQDs lack deep tissue magnetic resonance imaging (MRI) capabilities desirable for diagnostics. Considering that the drawbacks of MRI contrast agent toxicity are still poorly addressed, we develop novel Mn2+ or Gd3+ doped nitrogen-containing graphene quantum dots (NGQDs) to equip the GQDs with MRI capabilities and at the same time render contrast agents biocompatible. Water-soluble biocompatible Mn-NGQDs and Gd-NGQDs synthesized via single-step microwave-assisted scalable hydrothermal reaction enable dual MRI and fluorescence modalities. These quasi-spherical 3.9-6.6 nm average-sized structures possess highly crystalline graphitic lattice structure with 0.24 and 0.53 atomic % for Mn2+ and Gd3+ doping. This structure ensures high in vitro biocompatibility of up to 1.3 mg ml-1 and 1.5 mg ml-1 for Mn-NGQDs and Gd-NGQDs, respectively, and effective internalization in HEK-293 cells traced by intrinsic NGQD fluorescence. As MRI contrast agents with considerably low Gd and Mn content, Mn-NGQDs exhibit substantial transverse/longitudinal relaxivity (r 2/r 1) ratios of 11.190, showing potential as dual-mode longitudinal or transverse relaxation time (T 1 or T 2) contrast agents, while Gd-NGQDs possess r 2/r 1 of 1.148 with high r 1 of 9.546 mM-1 s-1 compared to commercial contrast agents, suggesting their potential as T1 contrast agents. Compared to other nanoplatforms, these novel Mn2+ and Gd3+ doped NGQDs not only provide scalable biocompatible alternatives as T1/T2 and T1 contrast agents but also enable in vitro intrinsic fluorescence imaging.

Rare-Earth-Doped Cerium Oxide Nanocubes for Biomedical Near-Infrared and Magnetic Resonance Imaging.

Near-infrared (NIR) fluorescence provides a new avenue for biomedical fluorescence imaging that allows for the tracking of fluorophore through several centimeters of biological tissue. However, such fluorophores are rare and, due to accumulation-derived toxicity, are often restricted from clinical applications. Deep tissue imaging not only provided by near-infrared fluorophores but also conventionally carried out by magnetic resonance imaging (MRI) or computed tomography (CT) is also hampered by the toxicity of the contrast agents. This work offers a biocompatible imaging solution: cerium oxide (CeO2) nanocubes doped with ytterbium or neodymium, and co-doped with gadolinium, showing simultaneous potential for near-infrared (NIR) fluorescence and magnetic resonance imaging (MRI) applications. A synthetic process described in this work allows for the stable incorporation of ytterbium or neodymium, both possessing emissive transitions in the NIR. As a biocompatible nanomaterial, the CeO2 nanocubes act as an ideal host material for doping, minimizing lanthanide fluorescence self-quenching as well as any potential toxicity associated with the dopants. The uptake of nanocubes by HeLa cells maximized at 12 h was monitored by hyperspectral imaging of the ytterbium or neodymium NIR emission, indicating the capacity of the lanthanide-doped nanocubes for in vitro and a potential for in vivo fluorescence imaging. The co-doped nanocubes demonstrate no significant loss of NIR emission intensity upon co-doping with 2 atomic % gadolinium and exhibit magnetic susceptibilities in the range of known negative contrast agents. However, a small increase to 6 atomic % gadolinium significantly affects the magnetic susceptibility ratio (r2/r1), shifting closer to the positive contrast range and suggesting the potential use of the CeO2 nanocube matrix doped with selected rare-earth ions as a tunable MRI contrast agent with NIR imaging capabilities.

Electric field quenching of graphene oxide photoluminescence.

With the advent of graphene, there has been an interest in utilizing this material and its derivative, graphene oxide (GO) for novel applications in nanodevices such as bio and gas sensors, solid-state supercapacitors and solar cells. Although GO exhibits lower conductivity and structural stability, it possesses an energy band gap that enables fluorescence emission in the visible/near infrared leading to a plethora of optoelectronic applications. In order to allow fine-tuning of its optical properties in the device geometry, new physical techniques are required that, unlike existing chemical approaches, yield substantial alteration of GO structure. Such a desired new technique is one that is electronically controlled and leads to reversible changes in GO optoelectronic properties. In this work, we for the first time investigate the methods to controllably alter the optical response of GO with the electric field and provide theoretical modeling of the electric field-induced changes. Field-dependent GO emission is studied in bulk GO/polyvinylpyrrolidone films with up to 6% reversible decrease under 1.6 V µm-1 electric fields. On an individual flake level, a more substantial over 50% quenching is achieved for select GO flakes in a polymeric matrix between interdigitated microelectrodes subject to two orders of magnitude higher fields. This effect is modeled on a single exciton level by utilizing Wentzel, Kremer, and Brillouin approximation for electron escape from the exciton potential well. In an aqueous suspension at low fields, GO flakes exhibit electrophoretic migration, indicating a degree of charge separation and a possibility of manipulating GO materials on a single-flake level to assemble electric field-controlled microelectronics. As a result of this work, we suggest the potential of varying the optical and electronic properties of GO via the electric field for the advancement and control over its optoelectronic device applications.

Convolutional Neural Networks Can Predict Retinal Differentiation in Retinal Organoids.

We have developed a deep learning-based computer algorithm to recognize and predict retinal differentiation in stem cell-derived organoids based on bright-field imaging. The three-dimensional "organoid" approach for the differentiation of pluripotent stem cells (PSC) into retinal and other neural tissues has become a major in vitro strategy to recapitulate development. We decided to develop a universal, robust, and non-invasive method to assess retinal differentiation that would not require chemical probes or reporter gene expression. We hypothesized that basic-contrast bright-field (BF) images contain sufficient information on tissue specification, and it is possible to extract this data using convolutional neural networks (CNNs). Retina-specific Rx-green fluorescent protein mouse embryonic reporter stem cells have been used for all of the differentiation experiments in this work. The BF images of organoids have been taken on day 5 and fluorescent on day 9. To train the CNN, we utilized a transfer learning approach: ImageNet pre-trained ResNet50v2, VGG19, Xception, and DenseNet121 CNNs had been trained on labeled BF images of the organoids, divided into two categories (retina and non-retina), based on the fluorescent reporter gene expression. The best-performing classifier with ResNet50v2 architecture showed a receiver operating characteristic-area under the curve score of 0.91 on a test dataset. A comparison of the best-performing CNN with the human-based classifier showed that the CNN algorithm performs better than the expert in predicting organoid fate (84% vs. 67 ± 6% of correct predictions, respectively), confirming our original hypothesis. Overall, we have demonstrated that the computer algorithm can successfully recognize and predict retinal differentiation in organoids before the onset of reporter gene expression. This is the first demonstration of CNN's ability to classify stem cell-derived tissue in vitro.