skip to main content

Title: Release, detection and toxicity of fragments generated during artificial accelerated weathering of CdSe/ZnS and CdSe quantum dot polymer composites
Next generation displays and lighting applications are increasingly using inorganic quantum dots (QDs) embedded in polymer matrices to impart bright and tunable emission properties. The toxicity of some heavy metals present in commercial QDs ( e.g. cadmium) has, however, raised concerns about the potential for QDs embedded in polymer matrices to be released during the manufacture, use, and end-of-life phases of the material. One important potential release scenario that polymer composites can experience in the environment is photochemically induced matrix degradation. This process is not well understood at the molecular level. To study this process, the effect of an artificially accelerated weathering process on QD–polymer nanocomposites has been explored by subjecting CdSe and CdSe/ZnS QDs embedded in poly(methyl methacrylate) (PMMA) to UVC irradiation in aqueous media. Significant matrix degradation of QD–PMMA was observed along with measurable mass loss, yellowing of the nanocomposites, and a loss of QD fluorescence. While ICP-MS identified the release of ions, confocal laser scanning microscopy and dark-field hyperspectral imaging were shown to be effective analytical techniques for revealing that QD-containing polymer fragments were also released into aqueous media due to matrix degradation. Viability experiments, which were conducted with Shewanella oneidensis MR-1, showed a statistically significant decrease more » in bacterial viability when the bacteria were exposed to highly degraded QD-containing polymer fragments. Results from this study highlight the need to quantify not only the extent of nanoparticle release from a polymer nanocomposite but also to determine the form of the released nanoparticles ( e.g. ions or polymer fragments). « less
Authors:
; ; ; ; ; ; ; ;
Award ID(s):
1503408
Publication Date:
NSF-PAR ID:
10060327
Journal Name:
Environmental Science: Nano
ISSN:
2051-8153
Sponsoring Org:
National Science Foundation
More Like this
  1. We demonstrate that colloidal quantum dots of CdSe and CdSe/ZnS are detected during the photooxidation of MeOH, under broad spectrum illumination (250 mW/cm 2 ). The stepwise photocurrent vs. time response corresponds to single entities adsorbing to the Pt electrode surface irreversibly. The adsorption/desorption of the QDs and the nature of the single entities is discussed. In suspensions, the QDs behave differently depending on the solvent used to suspend the materials. For MeOH, CdSe is not as stable as CdSe/ZnS under constant illumination. The photocurrent expected for single QDs is discussed. The value of the observed photocurrents, > 1 pA is due to the formation of agglomerates consistent with the collision frequency and suspension stability. The observed frequency of collisions for the stepwise photocurrents is smaller than the diffusion-limited cases expected for single QDs colliding with the electrode surface. Dynamic light scattering and scanning electron microscopy studies support the detection of aggregates. The results indicate that the ZnS layer on the CdSe/ZnS material facilitates the detection of single entities by increasing the stability of the nanomaterial. The rate of hole transfer from the QD aggregates to MeOH outcompetes the dissolution of the CdSe core under certain conditions of electron injection to themore »Pt electrode and in colloidal suspensions of CdSe/ZnS.« less
  2. Colloidal semiconductor nanocrystals (NCs) represent a promising class of nanomaterials for lasing applications. Currently, one of the key challenges facing the development of high-performance NC optical gain media lies in enhancing the lifetime of biexciton populations. This usually requires the employment of charge-delocalizing particle architectures, such as core/shell NCs, nanorods, and nanoplatelets. Here, we report on a two-dimensional nanoshell quantum dot (QD) morphology that enables a strong delocalization of photoinduced charges, leading to enhanced biexciton lifetimes and low lasing thresholds. A unique combination of a large exciton volume and a smoothed potential gradient across interfaces of the reported CdS bulk /CdSe/CdS shell (core/shell/shell) nanoshell QDs results in strong suppression of Auger processes, which was manifested in this work though the observation of stable amplified stimulated emission (ASE) at low pump fluences. An extensive charge delocalization in nanoshell QDs was confirmed by transient absorption measurements, showing that the presence of a bulk-size core in CdS bulk /CdSe/CdS shell QDs reduces exciton–exciton interactions. Overall, present findings demonstrate unique advantages of the nanoshell QD architecture as a promising optical gain medium in solid-state lighting and lasing applications.
  3. The use of nanoparticle-in-matrix composites is a common motif among a broad range of nanoscience applications and is of particular interest to the thermal sciences community. To explore this morphological theme, we create crystalline inorganic composites with nanoparticle volume fractions ranging from 0 to ∼100% using solution-phase processing. We synthesize these composites by mixing colloidal CdSe nanocrystals and In 2 Se 3 metal–chalcogenide complex (MCC) precursor in the solution-phase and then thermally transform the MCC precursor into a crystalline In 2 Se 3 matrix. We find rich structural and chemical interactions between the CdSe nanocrystals and the In 2 Se 3 matrix, including alterations in In 2 Se 3 grain size and orientation as well as the formation of a ternary phase, CdIn 2 Se 4 . The average thermal conductivities of the 100% In 2 Se 3 and ∼100% CdSe composites are 0.32 and 0.53 W m −1 K −1 , respectively. These thermal conductivities are remarkably low for inorganic crystalline materials and are comparable to amorphous polymers. With the exception of the ∼100% CdSe samples, the thermal conductivities of these nanocomposites are insensitive to CdSe volume fraction and are ∼0.3 W m −1 K −1 in all cases.more »We attribute this insensitivity to competing effects that arise from structural morphology changes during composite formation. This insensitivity to CdSe volume fraction also suggests that very low thermal conductivities can be reliably achieved using this solution-phase route to nanocomposites.« less
  4. InAs quantum dots (QDs) embedded into a waveguiding GaAs semiconductor matrix may produce scintillation detectors with exceptional speed and yield, making them valuable for nuclear security, medical imaging, and high energy physics applications. In this work, we developed thick (~25um) epitaxial heterostructres with high luminescence efficiency composed of self-assembled nano-engineered InAs QDs grown by molecular beam epitaxy. The bulk GaAs acts as a stopping material for incident particles and as a waveguide when layer-transferred onto a low-index substrate. Waveguiding and self-absorption (<1cm-1) were studied using photoluminescence with scanning laser excitation and modeled with ray optics approximation and geometrical coupling of high-index waveguide to a collection fiber. Scintillating signals from alpha-particles were analyzed with an external photodiode (PD) and an integrated PD which provided an improved optical coupling. The mean charge collected by the integrated PD corresponded to 5×1e4 photoelectrons per 1 MeV of deposited energy, or ~20% of the theoretically achievable light yield. Combined with the previously measured QD scintillation time of 0.3-0.6 ns, this makes the InAs/GaAs QD heterostructures the fastest high yield scintillation material reported.
  5. Obeid, Iyad Selesnick (Ed.)
    The Temple University Hospital EEG Corpus (TUEG) [1] is the largest publicly available EEG corpus of its type and currently has over 5,000 subscribers (we currently average 35 new subscribers a week). Several valuable subsets of this corpus have been developed including the Temple University Hospital EEG Seizure Corpus (TUSZ) [2] and the Temple University Hospital EEG Artifact Corpus (TUAR) [3]. TUSZ contains manually annotated seizure events and has been widely used to develop seizure detection and prediction technology [4]. TUAR contains manually annotated artifacts and has been used to improve machine learning performance on seizure detection tasks [5]. In this poster, we will discuss recent improvements made to both corpora that are creating opportunities to improve machine learning performance. Two major concerns that were raised when v1.5.2 of TUSZ was released for the Neureka 2020 Epilepsy Challenge were: (1) the subjects contained in the training, development (validation) and blind evaluation sets were not mutually exclusive, and (2) high frequency seizures were not accurately annotated in all files. Regarding (1), there were 50 subjects in dev, 50 subjects in eval, and 592 subjects in train. There was one subject common to dev and eval, five subjects common to dev andmore »train, and 13 subjects common between eval and train. Though this does not substantially influence performance for the current generation of technology, it could be a problem down the line as technology improves. Therefore, we have rebuilt the partitions of the data so that this overlap was removed. This required augmenting the evaluation and development data sets with new subjects that had not been previously annotated so that the size of these subsets remained approximately the same. Since these annotations were done by a new group of annotators, special care was taken to make sure the new annotators followed the same practices as the previous generations of annotators. Part of our quality control process was to have the new annotators review all previous annotations. This rigorous training coupled with a strict quality control process where annotators review a significant amount of each other’s work ensured that there is high interrater agreement between the two groups (kappa statistic greater than 0.8) [6]. In the process of reviewing this data, we also decided to split long files into a series of smaller segments to facilitate processing of the data. Some subscribers found it difficult to process long files using Python code, which tends to be very memory intensive. We also found it inefficient to manipulate these long files in our annotation tool. In this release, the maximum duration of any single file is limited to 60 mins. This increased the number of edf files in the dev set from 1012 to 1832. Regarding (2), as part of discussions of several issues raised by a few subscribers, we discovered some files only had low frequency epileptiform events annotated (defined as events that ranged in frequency from 2.5 Hz to 3 Hz), while others had events annotated that contained significant frequency content above 3 Hz. Though there were not many files that had this type of activity, it was enough of a concern to necessitate reviewing the entire corpus. An example of an epileptiform seizure event with frequency content higher than 3 Hz is shown in Figure 1. Annotating these additional events slightly increased the number of seizure events. In v1.5.2, there were 673 seizures, while in v1.5.3 there are 1239 events. One of the fertile areas for technology improvements is artifact reduction. Artifacts and slowing constitute the two major error modalities in seizure detection [3]. This was a major reason we developed TUAR. It can be used to evaluate artifact detection and suppression technology as well as multimodal background models that explicitly model artifacts. An issue with TUAR was the practicality of the annotation tags used when there are multiple simultaneous events. An example of such an event is shown in Figure 2. In this section of the file, there is an overlap of eye movement, electrode artifact, and muscle artifact events. We previously annotated such events using a convention that included annotating background along with any artifact that is present. The artifacts present would either be annotated with a single tag (e.g., MUSC) or a coupled artifact tag (e.g., MUSC+ELEC). When multiple channels have background, the tags become crowded and difficult to identify. This is one reason we now support a hierarchical annotation format using XML – annotations can be arbitrarily complex and support overlaps in time. Our annotators also reviewed specific eye movement artifacts (e.g., eye flutter, eyeblinks). Eye movements are often mistaken as seizures due to their similar morphology [7][8]. We have improved our understanding of ocular events and it has allowed us to annotate artifacts in the corpus more carefully. In this poster, we will present statistics on the newest releases of these corpora and discuss the impact these improvements have had on machine learning research. We will compare TUSZ v1.5.3 and TUAR v2.0.0 with previous versions of these corpora. We will release v1.5.3 of TUSZ and v2.0.0 of TUAR in Fall 2021 prior to the symposium. ACKNOWLEDGMENTS Research reported in this publication was most recently supported by the National Science Foundation’s Industrial Innovation and Partnerships (IIP) Research Experience for Undergraduates award number 1827565. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the official views of any of these organizations. REFERENCES [1] I. Obeid and J. Picone, “The Temple University Hospital EEG Data Corpus,” in Augmentation of Brain Function: Facts, Fiction and Controversy. Volume I: Brain-Machine Interfaces, 1st ed., vol. 10, M. A. Lebedev, Ed. Lausanne, Switzerland: Frontiers Media S.A., 2016, pp. 394 398. https://doi.org/10.3389/fnins.2016.00196. [2] V. Shah et al., “The Temple University Hospital Seizure Detection Corpus,” Frontiers in Neuroinformatics, vol. 12, pp. 1–6, 2018. https://doi.org/10.3389/fninf.2018.00083. [3] A. Hamid et, al., “The Temple University Artifact Corpus: An Annotated Corpus of EEG Artifacts.” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2020, pp. 1-3. https://ieeexplore.ieee.org/document/9353647. [4] Y. Roy, R. Iskander, and J. Picone, “The NeurekaTM 2020 Epilepsy Challenge,” NeuroTechX, 2020. [Online]. Available: https://neureka-challenge.com/. [Accessed: 01-Dec-2021]. [5] S. Rahman, A. Hamid, D. Ochal, I. Obeid, and J. Picone, “Improving the Quality of the TUSZ Corpus,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2020, pp. 1–5. https://ieeexplore.ieee.org/document/9353635. [6] V. Shah, E. von Weltin, T. Ahsan, I. Obeid, and J. Picone, “On the Use of Non-Experts for Generation of High-Quality Annotations of Seizure Events,” Available: https://www.isip.picone press.com/publications/unpublished/journals/2019/elsevier_cn/ira. [Accessed: 01-Dec-2021]. [7] D. Ochal, S. Rahman, S. Ferrell, T. Elseify, I. Obeid, and J. Picone, “The Temple University Hospital EEG Corpus: Annotation Guidelines,” Philadelphia, Pennsylvania, USA, 2020. https://www.isip.piconepress.com/publications/reports/2020/tuh_eeg/annotations/. [8] D. Strayhorn, “The Atlas of Adult Electroencephalography,” EEG Atlas Online, 2014. [Online]. Availabl« less