Bridging Nanotechnology and Neurobiology: Voltage-Sensing and Photothermal Control of Neural Activity Using Semiconductor and Metallic Nanocrystals

Bridging Nanotechnology and Neurobiology: Voltage-Sensing and Photothermal Control of Neural Activity Using Semiconductor and Metallic Nanocrystals Thesis by Zuzanna Lawera for the degree of Doctor of Philosophy in Physics Universidad del País Vasco Euskal Herriko Unibertsitatea Supervised by Marek Grzelczak and Rafael Yuste Donostia, November 2025 (cc) 2025 Zuzanna Lawera (cc by-sa 4.0) 1 Abstract Driven by the need for minimally invasive, high-resolution neural interfaces, this dissertation bridges nanotechnology and neurobiology by developing nanoscale tools for optical detection and modulation of neural activity. With a focus on semiconductor quantum dots and plasmonic gold nanocrystals, this dissertation presents efforts to develop complementary platforms to "Read" and "Write" neural signals. Designed to either report changes in membrane potential or deliver light-driven stimuli, these nanomaterials offer functional advantages at dimensions inaccessible to conventional bioelectronic approaches. Chapter 1 provides an interdisciplinary literature review, introducing the NanoNeuro, outlining how quantum dots and plasmonic nanoparticles can interact with neural systems and highlighting key challenges such as targeted delivery, biocompatibility, and integration with living cells. It provides an overview of the current state of the art and situates the objective of the presented research within this framework. Chapter 2 describes an experimental platform of genetically engineered “self-spiking” HEK293 cells that fire spontaneous, synchronized action potentials. This in vitro workhorse model, validated via calcium imaging, provided a simplified yet consistent testbed for evaluating nanoscale voltage-sensitive probes. At the heart of the dissertation, Chapter 3 details the development of a quantum dot-based voltage sensor. Spherical quantum well nanocrystals were synthesized and characterized using spectroscopic and electron microscopy techniques to confirm their physicochemical properties. To adapt the nanocrystals for biological environments, their surfaces were engineered for aqueous stability and incorporated into cell membranes using fusogenic liposomes. Their ability to uniformly label the plasma membrane was assessed by fluorescence microscopy in selfspiking HEK293 monolayers. Although the quantum dots remained photostable, fluorescence changes during action potentials remained below the threshold required for reliable detection. The absence of clear spike-correlated signals underscored limitations in sensitivity and temporal resolution of camera, pointing to the need for brighter probes, faster imaging, and improved signal-to-noise performance. In parallel, Chapter 4 investigates photothermal modulation of neurons with plasmonic nanocrystals. Gold nanoparticles of controlled size and morphology (spheres and bipyramids) were synthesized and characterized by spectroscopy and transmission electron microscopy to 2 determine their optical properties and structural uniformity. Following surface functionalization to ensure biological compatibility, nanocrystals were introduced into cell cultures (SH-SY5Y and cortex primary neurons). Their association with cells and internalization was assessed using confocal fluorescence microscopy and electron microscopy. Upon optical stimulation, the nanoparticles converted incident light into localized heating. In a proof-of-concept experiment performed in brain slices, this effect was sufficient to elicit neuronal firing: single action potentials were induced at moderate excitation power, and bursts of activity at higher intensities. These findings support the feasibility of plasmonic nanocrystals as optically addressable actuators for controlled neuromodulation. Chapter 5 summarizes the main findings and reflects on technological limitations and future directions of the work. The experimental research presented in this dissertation was carried out by the author at the Centro de Física de Materiales in Donostia, Spain (Marek Grzelczak’s group), and complemented by research stays at the Neurotechnology Center at Columbia University in New York (Rafael Yuste’s group), the Department of Chemistry at Columbia University (Jonathan Owen’s group), and the FluoroNanoTools Laboratory at the Biofisika Institute in Bilbao, Spain (Monica Carril’s group). These research stays formed the foundation of the experimental work described in this thesis. The broader collaborative framework extended beyond these four laboratories and included members of the IKUR Basque NanoNeuro Network (B3N), reinforcing the interdisciplinary nature of the project. Although some outcomes, particularly in voltage sensing, did not fully meet initial expectations, the work contributed to a deeper understanding of nanomaterial-cell interactions and clarified key design parameters for future tool development. The insights gained provide a foundation for more targeted and informed advancement of neural interfaces based on nanocrystals and nanoparticles. 3 Table of Contents CHAPTER 1 – Nano Neuro Objective ....................................................................................................................................... 8 1. Neurobiology ......................................................................................................................... 9 1.1. Neural Networks and Brain Function ............................................................................ 9 1.2. Current Methods of Bioimaging .................................................................................. 14 1.3. Experimental Models of the Human Nervous System ................................................ 18 2. Nanotechnology .................................................................................................................. 22 2.1. Semiconductor Nanocrystals....................................................................................... 24 2.2. Plasmonic Nanoparticles ............................................................................................. 28 3. Nano Neuro ......................................................................................................................... 34 4. References for Chapter 1 .................................................................................................... 47 CHAPTER 2 – Self-spiking HEK Cells 1. Creation and Maintenance of the Cell Line ......................................................................... 62 1.1. Cell Culture .................................................................................................................. 65 2. Calcium Imaging of Spiking HEK Cells .................................................................................. 67 2.1. Fluorescence Microscope Setup ................................................................................. 67 2.2. Results and Discussion ................................................................................................ 70 3. References for Chapter 2 .................................................................................................... 78 CHAPTER 3 – Quantum Dots for Detecting Neuronal Activity 1. Spherical Quantum Wells .................................................................................................... 82 1.1. Overview of Quantum-dot-based Platforms ............................................................... 86 1.2. Synthetic Approaches .................................................................................................. 88 1.3. Surface Chemistry on CdS/CdSe/CdS Nanocrystals .................................................... 94 2. Strategies for Water Transfer .............................................................................................. 96 2.1. OCG Micelles ............................................................................................................... 97 2.2. Lipid Vesicles ............................................................................................................. 102 3. Cell Membrane Insertion .................................................................................................. 120 3.1. OCG Micelles ............................................................................................................. 120 4 3.2. Fusosomes – Fusogenic Lipid Vesicles ...................................................................... 127 4. Recordings of Spiking HEK Cells with Spherical Quantum Dots ........................................ 153 5. Chapter Summary.............................................................................................................. 187 6. References for Chapter 3 .................................................................................................. 188 CHAPTER 4 – Plasmonic Nanocrystals for Photothermal Modulation of Neuronal Activity 1. Synthesis of Gold Nanocrystals ......................................................................................... 196 1.1. Synthesis of Citrate-Stabilized Gold Nanoparticles (Au NPs@citrate) ...................... 197 1.2. Synthesis of Size-Controlled Spherical Gold Nanoparticles (Au NPs@CTAB) ........... 199 1.3. Seed-Mediated Synthesis of Gold Bipyramids .......................................................... 202 2. Biofunctionalization of Gold Nanocrystals ........................................................................ 205 3. Gold Nanocrystals in Neurobiological Research ............................................................... 209 3.1. Physicochemical Characterization of NPs in Biological Media .................................. 209 3.2. In Vitro Internalization Studies.................................................................................. 214 3.3. In Vitro Optical Modulation of Neuronal Activity with Au NPs ................................. 220 4. References for Chapter 4 .................................................................................................. 223 CHAPTER 5 – Summary and Outlook 1. Summary of Major Findings and Conclusions ................................................................... 228 2. Methodological Refinement for Future Studies ............................................................... 229 3. Closing Remarks and Field Outlook ................................................................................... 238 4. References for Chapter 5 .................................................................................................. 239 APPENDIX 1. List of Abbreviations .......................................................................................................... 242 2. Supplementary Material for Chapter #2 ........................................................................... 244 2.1. Materials ................................................................................................................... 244 2.1.1. Reagent Preparation - Recipes .............................................................................. 245 2.2. Methods .................................................................................................................... 246 2.2.1. Cell Culture Protocols ............................................................................................ 246 2.2.2. Calcium Imaging .................................................................................................... 247 2.2.3. Fluorescence Microscope Setup ........................................................................... 247 5 2.3. Additional Figures...................................................................................................... 249 3. Supplementary Material for Chapter #3 ........................................................................... 252 3.1. Materials ................................................................................................................... 252 3.1.1. Reagent Preparation - Recipes .............................................................................. 252 3.2. Methods .................................................................................................................... 254 3.2.1. Instrumentation .................................................................................................... 254 3.2.2. CryoTEM ................................................................................................................ 254 3.2.3. Confocal Microscope ............................................................................................. 254 3.2.4. Fluorescence Microscope ...................................................................................... 255 3.3. Additional Figures...................................................................................................... 256 4. Supplementary Material for Chapter #4 ........................................................................... 263 4.1. Materials ................................................................................................................... 263 4.2. Methods .................................................................................................................... 263 4.2.1. Protocols................................................................................................................ 263 4.2.2. Instumentation ...................................................................................................... 266 4.2.3. Confocal Imaging of Fixed Cells ............................................................................. 266 4.2.4. Transmission Electron Microscopy (TEM) of Fixed Cells ....................................... 267 4.3. Additional Figures...................................................................................................... 267 5. Data and Code Repository ................................................................................................. 272 6. Documentation of Digital Tools ........................................................................................ 272 6.1. Data Analysis ............................................................................................................. 272 6.2. Figure Preparation ..................................................................................................... 272 6.3. Assisted Editing and Literature Search ...................................................................... 272 Acknowledgments ..................................................................................................................... 273 6 7 Chapter 1 Nano Neuro Chapter 1. Nano Neuro 8 Objective The human brain, the most complex organ known, remains one of the greatest scientific challenges. To unravel the processes that govern behavior and cognition, it is essential to deepen our understanding of their primary driver: the central nervous system. Neurobiology, which investigates the structure and function of neural systems, has yielded foundational insights into perception, memory, and disease, catalyzing breakthroughs in the diagnosis and treatment of neurological disorders. The impact of these discoveries extends far beyond medicine, shaping fields as diverse as artificial intelligence and public policy. Yet, as our scientific questions grow increasingly precise, so too must the tools we employ.1 The computational capabilities of brain is driven by nanoscale machinery: ion channels just a few nanometres wide that gate millisecond electrical spikes, and synapses that compress thousands of molecular components into volumes smaller than a femtolitre.2,3 Conventional tools such as electrodes and fluorescent dyes lack the temporal resolution and specificity to fully access or modulate these microscopic dynamics, leaving key biological processes obscured. This gap marks the emergence of Nano Neuro, a field at the intersection of nanoscience and neurobiology, which offers unprecedented opportunities for precision neural interfacing. Leveraging the distinctive electronic, optical, and biochemical properties of nanomaterials, researchers can now sense, manipulate, and study neural circuits at scales previously inaccessible.1 The promise of Nano Neuro is profound: to reveal hidden patterns of neural activity, enable localized neuromodulation, and drive innovation in both therapeutic intervention and neural enhancement. Nonetheless, the field faces formidable challenges, including the need for precise targeting, long-term biocompatibility, and minimally invasive integration with living tissue. Chapter 1 establishes the conceptual and technical foundation for introducing Nano Neuro. It begins with a review of essential neurobiological principles, followed by an introduction to nanotechnology and the materials most relevant to neural applications. With this groundwork in place, we examine the convergence of these disciplines, highlighting current technologies, landmark studies, and emerging directions. The goal is to situate the work presented in this dissertation within its broader scientific context and highlight the foundational questions it seeks to address: reading and writing neural activity. Chapter 1. Nano Neuro 15 bioimaging and electrophysiological methods that continue to expand our ability to observe and interpret neural activity. As understanding the brain requires not only identifying its components but also recording and visualizing their dynamic interactions, the development of imaging technologies has become central to neuroscience – transforming it from a descriptive discipline into one capable of real-time, multiscale interrogation of neural activity.30 Current methods for recording brain activity fall into two main categories: electrical and optical techniques, each with distinct advantages in temporal and spatial resolution (Figure 1.3).31 Electrical approaches directly measure neuronal voltage changes. At the single-cell level, patchclamp recording offers high temporal precision and is used to study ion channel activity.32 For population-level recordings, multielectrode arrays (MEAs) capture signals from dozens to hundreds of neurons simultaneously,33,34 while high-density probes such as Neuropixels allow recordings from thousands of sites across brain regions in behaving animals.35,36 Optical methods rely on translating neural signals into light, using either chemical dyes or genetically encoded indicators. Calcium imaging, one of the most widely used techniques, detects intracellular calcium transients associated with neuronal firing.37 Genetically encoded calcium indicators (e.g., GCaMP variants) enable long-term imaging in specific cell types with cellular resolution.37 Voltage imaging, though technically more challenging, uses indicators such as ASAP (Accelerated Sensor of Action Potential) or Archon to directly track changes in membrane potential with improved temporal fidelity.38–40 Additionally, fluorescent biosensors have been developed to visualize neurotransmitter release (e.g., iGluSnFR for glutamate) and intracellular signaling pathways.41–44 These techniques can be integrated with advanced microscopy methods, to achieve deep tissue access and high-resolution recordings. On the systems level, functional magnetic resonance imaging (fMRI), electroencephalography (EEG) and magnetoencephalography (MEG) provide non-invasive measurements of brain activity with whole-brain coverage, offering important context for linking cellular activity to behavior.45,46 Chapter 1. Nano Neuro 16 Figure 1.3 Overview of electrical and optical neuronal recording techniques. Electrical methods (top): Patch-clamp – high‐precision, single‐cell measurements of ion channel and membrane dynamics; Multielectrode arrays (MEAs) – simultaneous recordings from tens to hundreds of neurons in vitro or ex vivo; Neuropixels probes – high‐density, thousands‐site in vivo recordings across brain regions. Optical methods (bottom): Chemical indicators (e.g., Ca²⁺ dyes) – report intracellular calcium transients linked to spiking; Genetically encoded indicators – cell type-specific fluorescence reporters for calcium (GCaMP), membrane voltage (ASAP/Archon), and neurotransmitter release (e.g., iGluSnFR). Calcium Imaging Calcium imaging is a central technique in modern neuroscience, providing means to monitor neuronal activity by visualizing intracellular calcium dynamics that are tightly coupled to action potential firing and a vast array of cellular processes.47 Cells at rest maintain a very low intracellular calcium concentration of around 50-100 nM, but this level can transiently increase C C C C C C C C C C C C Electrode C C C C C Synapse Molecular Cage Depolariza on Ca2 Chapter 1. Nano Neuro 17 to approximately 1 µM upon activation, with localized "hot-spots" near active channels potentially reaching tens of micromolar.48 The diversity of cellular processes regulated by calcium arises from the versatility of its signaling dynamics, ranging in speed, amplitude, and spatial or temporal distribution. As a result, the choice of calcium imaging tools must be carefully matched to the specific characteristics of the process under study, as different indicators vary in their kinetics, sensitivity, and spatial resolution.48 The underlying principle of the technique involves detecting free intracellular calcium ions through their binding to sensor molecules that undergo a measurable change in fluorescence upon calcium binding.49 Calcium indicators are broadly classified into two types: synthetic small-molecule dyes (chemical indicators) and genetically encoded calcium indicators (GECIs).50 GECIs, such as the commonly used GCaMP family, are protein-based sensors that can be introduced via viral vectors or expressed in transgenic animals. The basic design often involves a fluorescent protein fused to the calcium-binding protein calmodulin and a calmodulin-binding peptide (M13).31 They enable long-term imaging in specific cell types, offering excellent spatial resolution and compatibility with in vivo microscopy.31,51 Their main limitations include slower kinetics compared to chemical dyes and potential buffering effects on intracellular calcium, which can affect the fidelity of spike detection – particularly during high-frequency activity.49 Additionally, delivery via viral transduction or in utero electroporation is invasive and can lead to heterogeneous cellular labeling or tissue damage.49 Chemical calcium indicators offer a valuable alternative to genetically encoded sensors, particularly in experiments that demand rapid kinetics and high temporal resolution.52 These synthetic fluorescent dyes are typically derived from calcium chelators such as EGTA or BAPTA, chemically modified to include fluorescent reporter groups.47 Their fast-binding kinetics and strong fluorescence responses provide higher signal-to-noise ratios, making them well-suited for detecting brief calcium transients, such as those associated with high-frequency neuronal firing or localized events in dendritic spines. Unlike GECIs, chemical indicators do not require genetic manipulation; instead, they are introduced into cells as membrane-permeable acetoxymethyl (AM) esters, which are activated by intracellular esterases to release the active dye.50 Early examples such as Fura-2 allowed for ratiometric calcium measurements, while singlewavelength indicators like Fluo-4 and Calcium Green-1 became widely used due to their strong fluorescence changes upon calcium binding and compatibility with standard excitation wavelengths.47 Although this method allows for rapid and widespread labeling, it lacks cell-type specificity and can lead to variability in loading, dye extrusion, or compartmentalization. Despite these limitations, chemical calcium indicators remain essential for acute preparations where Chapter 1. Nano Neuro 18 high-speed, high-resolution imaging is required, including studies in brain slices and at subcellular compartments such as synaptic terminals.51 These indicators comprise a large toolkit offering a broad range of calcium affinities and excitation/emission spectra, enabling researchers to tailor imaging protocols to specific experimental needs, from lowto high-calcium environments and across various optical setups. While such advances continue to expand the capabilities of calcium imaging, the focus of this dissertation is on chemical calcium imaging using Calcium Green-1, chosen for its sensitivity, fast kinetics, and suitability for high-resolution imaging in vitro. 1.3. Experimental Models of the Human Nervous System The human nervous system is complex, with billions of neurons interconnected by trillions of synapses.53 Direct study of this system in humans is constrained by significant ethical and technical challenges, limiting the extent to which we can experimentally manipulate or observe neural processes in vivo.54 To address these constraints, researchers employ model systems that capture essential features of the nervous system while offering greater accessibility and experimental control. Many features of neurons and networks are conserved across species. Fundamental properties of neurons and networks observed in humans are often mirrored in model organisms such as non-human primates,55–57 rodents (mice, rats and squirrels),58–60 zebrafish,61 fruit fly62,63 and even hydrozoans like Hydra.64,65 Relying on animal stand-ins to explore human physiology is a tradition rooted in ancient Greek medicine and carried into modern research.66 Each occupies a distinct investigative niche: primates enable analyses of higher-order cognition and fine motor control; rodents provide exceptional genetic tractability for cortical and behavioral studies; zebrafish allow whole-brain optical access throughout development; and the simple diffuse nerve net of Hydra offers a minimalist framework for exploring the evolutionary origins of coordinated neural activity.54,67–69 Beyond evolutionary inquiry, model systems serve as essential platforms for translational research.70 Many neurological disorders – ranging from neurodegenerative diseases like Alzheimer's and Parkinson's to conditions involving aberrant excitability such as epilepsy – cannot be fully understood through human clinical data alone. An extensive repertoire of experimental models, from cultured neuronal cells to genetically modified animals, provides the foundational tools for exploring the structure and function of the human nervous system. These models not only offer insight into basic biological questions but also facilitate the development Chapter 1. Nano Neuro 19 of diagnostic and therapeutic strategies, forming a critical bridge between experimental neuroscience and clinical application (Figure 1.4).57,64 One way to study the nervous system is to isolate its building blocks – the cells – and examine them in controlled laboratory conditions (in vitro). This ranges from immortal cell lines to actual neurons kept alive outside the body. These models let researchers observe molecular and cellular processes up close, free from the complexity of a whole organism.70,71 Immortalized cell lines are cells that can grow indefinitely in the lab. Though they are not neurons, cell lines like HEK293 (human embryonic kidney cells) or HeLa (human cell line derived from cancer cells) are invaluable for neurobiological research.72 Scientists often use these cells as host systems to study specific neural proteins. For example, HEK293 cells are widely used to produce and analyze neuronal ion channels or receptors by inserting the relevant genes.73 While these cell lines do not fire impulses or form networks like real neurons, they offer a highthroughput, easy-to-maintain platform to dissect molecular details. By contrast, primary neurons are nerve cells taken directly from an animal’s brain or spinal cord and grown in a dish. A common example is rodent hippocampal neurons cultured from newborn rats or mice. These primary neuron cultures do form synapses and electrical activity, essentially creating a miniature neural network in the dish. Researchers use them to study synaptic communication, development of neural connections, and the effects of drugs or genetic changes on neuron behavior. Because they are real neurons, primary cultures more closely mimic brain function than cell lines. The downside is that primary neurons are delicate and have a finite lifespan in vitro. Nonetheless, they have been a workhorse for studying mechanisms of neurodevelopment and neuroplasticity at the cellular level (for example, uncovering how synapses strengthen or weaken during learning-like processes).70 Extending the elemental power of primary cultures to the network scale, brain-slice preparations maintain authentic synaptic architecture in a fully manipulable ex-vivo setting. Introduced by Henry McIlwain in the 1950s,74 this preparation – maintained in precisely defined artificial cerebrospinal fluid – proved its worth early on by demonstrating normal resting potentials, intact synaptic transmission, and stimulation-evoked metabolic shifts. Today, hippocampal slices remain the benchmark for dissecting long-term potentiation and other forms of synaptic plasticity, whereas cortical slices support real-time imaging of network oscillations and high-throughput pharmacological screening.75 Their relative “quietude,” combined with unobstructed optical and electrode access to dendrites and axons, enables mechanistic analyses – from ion-channel kinetics to circuit-level disease models – that are impractical in vivo. Brain- Chapter 1. Nano Neuro 20 slice work thus occupies a pivotal middle ground between cell cultures and whole-animal studies, offering a cost-effective, ethically favorable platform that continues to drive modern neurobiology. A revolutionary advance in the last decade has been the use of stem cells to model the human brain. Stem cells (especially induced pluripotent stem cells, iPSCs, made from adult human cells) can be guided to develop into neural cells. From these, scientists can grow 3D brain organoids – tiny, pea-sized blobs of tissue that self-organize into layers and cell types reminiscent of a developing brain. These organoids are not full brains, but they can mimic specific aspects of the human nervous system, such as forming cortical-like structures or even primitive eye cups in a dish. This approach allows us to study human neural development and disease in vitro. For instance, researchers have created patient-specific brain organoids to investigate microcephaly (a developmental brain disorder) and found that certain genetic mutations or viruses (like Zika virus) can stunt organoid growth, reflecting the patient’s condition. Organoids thus serve as experimental avatars of a human brain, enabling tests of drug treatments or genetic corrections on patient-derived tissue. There are also structures called assembloids, where multiple organoids (e.g., cortical and spinal cord, or brain and muscle) are fused to study interactions between regions, and even experiments transplanting human organoid tissue into animal brains (to observe how human neurons integrate in vivo). While still a developing area, these stem cell–derived models are a major step toward bridging the gap between animal models and the human brain itself, especially for understanding human-specific neurodevelopmental disorders and evolution.76,77 Chapter 1. Nano Neuro 21 Figure 1.4 Spectrum of experimental models used in neuroscience. This chart organizes model systems by biological complexity (vertical axis) and research focus, ranging from basic research to clinical applications (horizontal axis). Models are categorized as in vitro or in vivo. In vitro systems include immortalized cell lines, primary neurons, stem-cell–derived preparations, and brain slices obtained from animals. In vivo models span from simple invertebrates (e.g., Hydra, fruit fly, cuttlefish) to aquatic vertebrates (e.g., zebrafish), rodents, non-human primates, and humans. These are selected examples from a broader array of experimental models used in neuroscience. C C C C C C C C C C Chapter 1. Nano Neuro 22 2. Nanotechnology Since the devil is said to dwell in the details, modern materials science looks downward – toward atoms, molecules, and nanoscale structures. Nanotechnology is the interdisciplinary field that focuses on the design, synthesis, and characterization of structures typically ranging from 1 to 100 nanometers. At this scale, materials often display properties that differ significantly from their bulk counterparts. These differences arise primarily from two factors: (1) Surface effects, due to a high surface area-to-volume ratio, where a large proportion of atoms reside on the surface with fewer neighboring atoms. This can reduce cohesive energy, lower melting points (e.g., 2.5 nm gold particles melt at much lower temperatures than bulk gold), and enhance chemical reactivity. Catalytic activity, for instance, often increases at the nanoscale due to the abundance of active surface sites.78,79 (2) Quantum effects, which become relevant when particle dimensions approach the characteristic length scales of electrons (e.g., the exciton Bohr radius). In such cases, energy levels become quantized, leading to new optical, electronic, and magnetic behaviors. For example, while bulk platinum is non-magnetic, 2 nm clusters can exhibit magnetism. Similarly, semiconductor nanoparticles show size-dependent bandgaps, resulting in a blue shift in optical absorption.78,79 Together, these surface and quantum effects give rise to the unique mechanical, thermal, electronic, optical, and catalytic properties.80 Nanomaterials are commonly classified by their dimensionality, according to the number of degrees of freedom in the particle momentum. While nanoparticles – structures with all three dimensions in the 1–100 nm range – are considered zero-dimensional (0D) due to their roughly isotropic shape, several other forms of nanomaterials exist.78,81 One-dimensional (1D) category of nanomaterials includes nanotubes (e.g., carbon nanotubes), nanorods, nanowires, nanofibers, and nanohorns. Electrons in these structures are typically confined in two dimensions, allowing free movement along the length. Two-dimensional (2D) nanomaterials have nanoscale thickness but extended lateral dimensions. Examples include nanosheets, nanofilms, nanolayers, and graphene-derived materials.23 These structures confine electrons in one dimension. Three-dimensional (3D) nanomaterials refer to bulk assemblies or structures that are not confined to the nanoscale in any single dimension but are composed of nanoscale building blocks (e.g., arrays of nanoparticles, nanowires, or nanotubes).78,82 Chapter 1. Nano Neuro 23 Nanomaterials can be synthesized using two general strategies: top-down and bottom-up approaches.79,80 Top-down methods start with bulk materials that are physically or mechanically reduced to the nanoscale, using techniques such as high-energy ball milling, lithography, laser ablation, or arc discharge.82,83 These methods are well-suited for large-scale production but often result in broader size distributions and potential structural defects. In contrast, bottomup approaches assemble nanostructures atom-by-atom or molecule-by-molecule through chemical or biological processes.79 Examples include chemical vapor deposition (CVD), sol-gel synthesis, co-precipitation, and colloidal chemical reduction. Bottom-up techniques generally offer better control over particle size, shape, and crystallinity. In many cases, a combination of both approaches is employed to optimize structure and functionality, depending on the material system and application.84,85 As this thesis centers on colloidal nanoparticles, the subsequent sections will explore this class of nanomaterials in more detail. Nanoparticles can be classified in various ways; one useful approach is to group them based on their chemical composition and the dominant physical phenomena that govern their behavior at the nanoscale: Semiconductor quantum dots (QDs) are semiconductor nanocrystals, typically below 10 nm in diameter, often composed from CdSe, CdTe, or InP. QDs are used in biological imaging, lightemitting devices, solar cells, and quantum electronics.79,86,87 Plasmonic metal nanoparticles, commonly made of gold (Au) or silver (Ag), but also non-noble metals such as copper (Cu), aluminum (Al) or nickel (Ni), exhibit localized surface plasmon resonance (LSPR) – the collective oscillation of conduction electrons induced by incident light. This leads to strong, sizeand shape-dependent optical absorption and scattering. Common applications are biosensing, diagnostics, catalysis, and surface-enhanced spectroscopy.84,88,89 Magnetic nanoparticles, such as iron oxides (Fe3O4, Fe2O3) or FePt alloys, can exhibit superparamagnetism at sizes below the single-domain limit. These particles respond strongly to external magnetic fields but exhibit no remanent magnetization, making them ideal for use in MRI contrast agents, magnetic hyperthermia, drug delivery, and magnetic separation technologies.90,91 Carbon-based nanomaterials include several structurally distinct systems. Fullerenes (e.g., C60) are spherical carbon cages with high electron affinity. Carbon nanotubes (CNTs) are rolled graphene sheets with exceptional mechanical strength and electrical conductivity, available in singleor multi-walled forms. Carbon quantum dots (CQDs) and carbon dots (CDs) are quasispherical particles below 10 nm, with sizeor defect-dependent photoluminescence. These Chapter 1. Nano Neuro 24 materials are applied in bioimaging, sensing, energy storage, and environmental technologies.78,82,92 Ceramic nanoparticles, such as titanium dioxide (TiO₂) and zinc oxide (ZnO), are inorganic, nonmetallic materials with high thermal stability and chemical resistance. Depending on synthesis conditions, they may be amorphous or crystalline, dense or porous. They are widely used in catalysis, biomedical coatings, photonics, and energy conversion devices.93,94 Lipid-based nanoparticles – including liposomes and solid lipid nanoparticles (SLNs) – are typically spherical and range from 10 to 1000 nm. Their amphiphilic composition enables encapsulation and controlled release of therapeutic agents, making them central to drug delivery, imaging, and diagnostic systems.90,95,96 Polymeric nanoparticles are formed from synthetic or natural polymers and range from 1 to 1000 nm. They can carry active substances either on their surface or within a polymer matrix. A notable subclass, conjugated polymer nanoparticles (CPNs) or polymer dots (Pdots), displays high brightness and tunable photophysical properties, making them suitable for imaging and biosensing.90,97 Nanoparticles exhibit a broad diversity in shape, size, and internal structure – ranging from spherical, rod-like, cylindrical, tubular, and hollow-core geometries to chiral forms; they may be crystalline (singleor multi-domain), amorphous, uniform, or composed of multiple layers – reflecting a rapidly expanding library of materials that continues to grow as synthesis methods advance.78,98 2.1. Semiconductor Nanocrystals Over the past decade, quantum dot (QD) research has evolved from fundamental studies in nanochemistry into a broad technology platform supporting advances in displays, lighting, photovoltaics, quantum information systems, and bioanalytics. The field began in the early 1980s, when Ekimov and Brus independently observed that semiconductor nanocrystals exhibit size-dependent optical properties due to quantum confinement. In 1993, Bawendi and co-workers introduced a reliable synthetic route producing highly uniform colloidal QDs with tunable emission, which catalyzed rapid progress in the field. Continued innovations in composition, shape, and surface chemistry have since enabled QDs with high brightness, stability, and spectral control. In recognition of these contributions, Ekimov, Brus, and Bawendi Chapter 1. Nano Neuro 31 the synthesis of diverse anisotropic morphologies, including high-aspect-ratio rods, bipyramids, and nanostars with sharp protrusions, whose structural features give rise to narrow spectral linewidths and tunable optical properties.124 More complex architectures, including nanoshells and nanocages, are typically synthesized via templating approaches, such as gold deposition on dielectric cores (e.g., silica) or galvanic replacement reactions using silver templates.127,128 A key advantage of gold nanocrystals lies in their amenability to surface modification.121,141 Their strong affinity for thiol groups allows straightforward functionalization with polymers, biomolecules, or other targeting ligands.127,129,142,143 For biological applications, polyethylene glycol (PEG) is commonly grafted to enhance colloidal stability and reduce non-specific interactions with biological components.127,131 Specificity can be further introduced through conjugation with antibodies, peptides, or nucleic acids138,144. In many cases, the removal or exchange of toxic surfactants (such as CTAB) is necessary prior to in vivo use.145,146 Ligand exchange protocols, combined with surface characterization techniques like ζ-potential measurements, FTIR (Fourier‐Transform Infrared Spectroscopy), or XPS (X-ray Photoelectron Spectroscopy), are essential to confirm the chemical composition and stability of the final nanostructure. These features make gold nanocrystals highly versatile for biomedical and sensing applications.130,131,138 In photothermal therapy, NIR-resonant particles – such as nanorods, shells, or branched structures – efficiently convert light into localized heat to ablate targeted tissue with minimal invasiveness.146,147 In photoacoustic imaging, pulsed laser excitation of AuNPs generates strong ultrasonic signals for deep-tissue imaging.121,122 Surface-enhanced Raman scattering (SERS) exploits local field enhancements for ultra-sensitive molecular detection, while simple aggregation-based assays enable rapid, colorimetric biosensing.124,128,138 Furthermore, their integration into lab-on-chip platforms allows real-time, label-free detection of analytes through LSPR shifts induced by biomolecular interactions at the nanoparticle surface.129,138 Looking ahead, continued progress in shape-controlled synthesis, hybrid design (e.g., multilayer core-shell structures or Janus structures , with two chemically distinct surfaces), and tailored light-matter interactions is expanding the functional landscape of gold nanocrystals.127,135,146 While challenges such as scalability, batch-to-batch reproducibility, and long-term biocompatibility persist, the synergistic development of synthetic strategies, surface engineering, and application-specific performance continues to drive innovation in areas spanning nanomedicine,148 diagnostics,128,149 catalysis,130,143 and photonic devices150,151. Chapter 1. Nano Neuro 32 Photothermal Effect The photothermal effect is a direct nanoscale conversion of absorbed light into heat. When a plasmonic nanostructure is illuminated at LSPR, the incident electromagnetic energy is absorbed, causing the collective oscillation of free conduction electrons.147 According to the principle of energy conservation, this absorbed optical energy must subsequently be converted to other forms.152 The excited plasmon decays non-radiatively within femtoseconds (typically ~100 fs), generating energetic charge carriers called "hot electrons" and "hot holes" through a process known as Landau damping.129,153,154 These hot electrons possess kinetic energies higher than their equilibrium thermal values and initially do not follow a thermalized Fermi-Dirac distribution. Within approximately 10 fs, hot electrons lose energy through electron-electron scattering, redistributing their energy across the broader electron gas and thereby thermalizing into a Fermi-Dirac distribution.123,147 Subsequently, this hot electron gas relaxes by transferring its energy to the ionic lattice of the plasmonic nanoparticle through electron-phonon interactions. For gold, this electron-phonon relaxation occurs over a characteristic timescale of around 1.7 ps, resulting in an increased lattice temperature of the nanoparticle.121 From another perspective, this heat generation can be viewed as analogous to Joule heating, involving energy dissipation within the metallic structure. Finally, heat diffuses outward from the nanoparticle into the surrounding medium (e.g., liquid or glass), leading to an elevated temperature in its immediate environment. This external heat diffusion occurs over relatively longer timescales, typically ranging from approximately 100 ps to a few nanoseconds, depending on factors such as nanoparticle size and the thermal conductivity of the surrounding medium.155 Figure 1.6 Photothermal conversion in plasmonic nanoparticles: (1) Light absorption at the localized surface plasmon resonance. (2) Non-radiative plasmon decay (~100 fs) generates hot electrons and holes. (3) Electron-electron scattering (~10 fs) thermalizes the carrier distribution. (4) Electron-phonon coupling (~1.7 ps for Au) transfers energy to the lattice, raising nanoparticle temperature. (5) Phonon-phonon heat dissipation into the surrounding medium occurs over ≈100 ps to a few ns. Chapter 1. Nano Neuro 33 The efficiency of the photothermal effect in plasmonic nanomaterials depends on nanoparticle size and morphology determining the balance between absorption and scattering.121 Small nanocrystals absorb incident radiation, converting it efficiently into localized heat. Furthermore, the geometry of nanoparticles dictates the internal spatial distribution of generated heat: elongated shapes facilitate more uniform heating across their volume, whereas spheres primarily heat at the particle surface. 156 Additionally, the illumination conditions, particularly wavelength, irradiation intensity, and illumination mode, are decisive. Tailoring the excitation wavelength to coincide with the nanoparticle plasmon resonance maximizes absorption-driven heat generation and can be adjusted, for example, to match the biologically transparent NIR-I (650–950 nm) or NIR-II (1000–1350 nm) windows.141,157,158 The nature of illumination, continuous-wave versus pulsed, also shapes thermal dynamics. Under CW illumination, the system reaches a steady-state temperature distribution, with the heat dissipating into the surrounding medium according to a radial profile that decays as 1 𝑟 ⁄, where 𝑟 is the distance from the heat source. In contrast, pulsed illumination, particularly with ultrashort (femtosecond) pulses, confines heat more tightly in space and time. The resulting temperature decays more steeply with distance, scaling as 1 𝑟3 ⁄, and can give rise to additional effects such as acoustic wave generation or structural modifications of the nanoparticle.121,159 In addition, heat dissipation rates depend on the surrounding medium’s thermal conductivity, influencing nanoparticle internal temperature uniformity and environmental heating dynamics.129,147,160 Collective photothermal effects emerge from interparticle interactions, amplifying and homogenizing temperature increases across nanoparticle ensembles.120,159 Finally, modifying the nanoparticle surface chemistry via coatings or functional groups not only enhances colloidal stability and biocompatibility but also can significantly boost optical absorption and overall photothermal conversion efficiency. This enhancement arises through several mechanisms: increasing the local refractive index, introducing broadband-absorbing layers, and forming junctions that intensify the local electromagnetic field. Surface coatings such as titanium oxide, melanin, polydopamine, and graphene oxide have been shown to increase laser absorption in plasmonic nanomaterials.122 In steady-state CW illumination, the temperature rise experienced by a plasmonic nanoparticle is ultimately set by the balance between the optical power it absorbs and the ability of the surrounding medium to conduct that heat. Expressing the absorbed power as 𝑄 = σ𝑎𝑏𝑠 ∙I and solving the stationary heat-diffusion equation gives the canonical relation: δ𝑇𝑁𝑃 = 𝑄 4𝜋𝜅𝑠𝑅𝑒𝑞 (1.2) Chapter 1. Nano Neuro 34 where 𝑅𝑒𝑞 is a thermal radius (Laplace radius) that characterizes how the particle geometry influences heat diffusion and 𝜅𝑠 the thermal conductivity of the host medium.147 To simplify comparison across different materials and shapes, the optical term 𝜎𝑎𝑏𝑠 can be factored out into the dimensionless Joule number: 𝐽𝑜 = 𝜎𝑎𝑏𝑠 λ𝑟𝑒𝑓 2𝜋𝑉 (1.3) figure of merit that reports the light-to-heat conversion capability per unit volume at an arbitrary reference wavelength λ𝑟𝑒𝑓 (commonly 1240 nm). Because absorption cross-section 𝜎𝑎𝑏𝑠 scales with the local plasmonic field enhancement, 𝐽𝑜 is increasing as the anisotropy of a material increases (the bigger aspect ratio, the higher 𝐽𝑜 ). For example, in gold nanorods, increasing the aspect ratio from 1 to about 8 can increase 𝐽𝑜 by nearly two orders of magnitude due to the stronger and red-shifted longitudinal surface plasmon resonance. By substituting 𝐽𝑜 into the expression for δ𝑇𝑁𝑃 makes explicit how geometry (through aspect ratio), material permittivity and the ambient thermal conductance collectively dictate the temperature jump that drives the “nanoscale heat-engine” described above. With the physical principles and technical capabilities of plasmonic heating established, we now explore how these can be adapted to the challenges of neural modulation, where precision, compatibility, and responsiveness are paramount. 3. Nano Neuro Since the beginning of the 21st century, nanotechnology has steadily advanced from a speculative frontier into a practical and versatile toolkit for neuroscience.161 In 2000, Mattson and colleagues demonstrated that functionalized multi-walled carbon nanotubes (MWCNTs) could support neuronal adhesion and neurite outgrowth, establishing the first clear evidence that engineered nanomaterials could interface directly with neural tissue.162 Within a few years, researchers began modifying the surface chemistry of carbon nanotubes to selectively promote neural differentiation and synaptogenesis, while also improving conductivity and biocompatibility. By 2005, studies showed that purified MWCNTs enhanced the efficiency of electrical signal propagation in cultured networks, laying the groundwork for their integration into low-impedance neural probes.163 In 2007, aligned carbon nanofiber arrays were successfully used for stable extracellular recordings and targeted synaptic stimulation, illustrating the potential of nanoscale electrodes for chronic neural interfacing.164 Since then carbon nanotubes Chapter 1. Nano Neuro 35 are intensively studied as a tool that could boost the performance of microelectronic interfaces with brain tissue.163 Beyond carbon allotropes, two-dimensional graphene and its derivatives have become foundational materials for transparent, flexible neural interfaces.163 Monolayer and nanoporous graphene sheets are capable of modulating ion-channel activity, delivering photothermal or photoelectrical stimulation, and – when patterned into micro-transistors – recording cortical dynamics across a broad frequency spectrum, from infraslow oscillations to high-gamma bursts. Their optical transparency preserves the imaging window, enabling multimodal investigations. Notably, large-area reduced graphene oxide (rGO) electrocorticography foils are now undergoing first-in-human trials for intraoperative brain mapping, underscoring their translational potential.163 To address the mechanical mismatch between rigid implants and the soft neural tissue, that may lead to immunological response and cell damage, recent research has turned to conductive polymer hydrogels. These are semi-interpenetrating networks composed of conjugated polymers (like PEDOT:PSS, polypyrrole, or polythiophene) combined with PEG or gelatin-methacryloyl (GelMA) matrices. Such composites exhibit brain-like elastic moduli (≤10 kPa) while maintaining high charge-injection capacities (>1 mA·cm-2) and excellent long-term electrochemical stability. These hydrogels are not only robust under cyclic stress and sonication but are also compatible with advanced fabrication techniques such as 3D printing and two-photon polymerization, enabling microscale patterning of bioactive cues for guided axon growth and reduced tissue inflammation.163 Metallic nanostructures also contribute significantly to the neurotechnology landscape. Even though we will explore plasmonic nanostructures – especially gold nanoparticles for neuromodulation – in more detail shortly, it is worth noting now that other metal-based nanomaterials have already played a significant role in advancing neurotechnology. For example, silver nanowire networks, either free-standing or embedded in ultrathin, flexible Parylene-C polymer films, have been developed as low-impedance electrocorticography arrays. These arrays are mechanically compliant, can be folded or injected, and conform closely to the brain’s surface, allowing stable signal acquisition with minimal tissue disturbance. In parallel, superparamagnetic iron-oxide nanoparticles (SPIONs) have enabled wireless neural stimulation. When exposed to alternating magnetic fields, these particles can either generate localized heating (magnetothermal effect) or produce mechanical torque. Under magnetic stimulation, SPIONs oscillate or rotate and apply mechanical forces to the nearby cell membrane; this membrane strain opens Piezo-family mechanosensitive ion channels, allowing sub-millisecond Chapter 1. Nano Neuro 36 control of neuronal firing entirely wirelessly – without the need for implanted electrodes or optical fibers. Proof-of-concept studies have demonstrated that such magnetically controlled particles can modulate neuronal activity and behavior in freely moving animal models, highlighting their potential as minimally invasive, remotely controlled neuromodulation tools.163 Before diving into the core focus of this dissertation – plasmonic nanoparticles (with particular emphasis on gold) and semiconductor quantum dots – it is important to first consider several critical factors that must be addressed when integrating nanotechnology into biological research. First aspect to consider is the blood-brain barrier (BBB) that severely restricts nanoparticle delivery to the central nervous system. BBB is a highly selective membrane barrier located in the central nervous system that plays a crucial role in maintaining brain balance and protecting it from harmful substances and pathogens.165 Delivery of nanomaterials requires employing strategies as: passive diffusion with long-circulating or ultrasmall particles (around 2 nm diameter); temporary BBB disruption through focused ultrasound (often combined with microbubbles), osmotic agents such as mannitol, or nitric oxide-releasing nanoparticles; targeted delivery using surface-modified lipid carriers, receptor-targeted ligands, cationic vesicles, apolipoprotein adsorption, or receptor-mediated transcytosis; and magnetic guidance of iron-oxide or magnetoelectric nanoparticles using external magnetic fields to enhance brain uptake in a minimally invasive manner.165 Second, once in biological fluids, nanoparticles rapidly acquire a “protein corona”.166 In this spontaneous process layer of biomolecules, primarily proteins, are adsorbed on the surface of nanomaterial driven by various physicochemical interactions between proteins and the nanomaterial surface. Protein corona alters aggregation and biodistribution, cellular uptake, or potential toxicity of nanoparticles.163 Understanding and controlling corona composition is essential for predictable in vivo behavior. Strategies to minimize undesired corona formation include surface modification with polymers such as PEG or zwitterionic ligands, coating nanoparticles with natural cell membranes to mimic biological identity, and removing contaminants that promote non-specific interactions.163 Alternatively, the protein corona can be deliberately engineered – for example, through ligand functionalization or controlled protein adsorption – to enhance selective uptake, facilitate receptor-mediated transport, and reduce off-target effects.166 Third, cellular endocytosis, the process by which cells immerse extracellular material into membrane-bound vesicles, is a double-edged sword for nanoparticle applications.167–169 On one Chapter 1. Nano Neuro 37 hand, it enables intracellular delivery of cargos such as biosensors, drugs, or imaging agents, and underlies the design of nanomedicines that exploit receptor-mediated or adsorptive uptake pathways to reach specific organelles or cell types. Typical endocytic uptake begins within minutes – clathrin-mediated pit formation and vesicle scission occur within 15-60 seconds, and caveolae-mediated processes take about 5-10 minutes – while full internalization, sorting within endosomes, and transport to lysosomes can take up to one or two hours.170 When optimizing nanoparticles for desired application it is crucial to take into consideration the uptake kinetics and properly tune the size, charge, and surface of nanomaterial.171 For neuromodulation strategies that activate thermosensitive ion channels through photothermal heating, endocytic uptake must be suppressed so that nanoparticles remain at the membrane in close proximity to the channels. Fourth, the unique reactivity of nanomaterials raises neurotoxic risks – oxidative stress, inflammation, barrier disruption, and direct membrane damage – that depend sensitively on composition, coating, and dose. This highlights the need for rigorous and physiologically relevant toxicity assessments, particularly when considering in vivo use and future clinical translation.170 Finally, slow clearance and long-term retention of non-degradable particles threaten chronic accumulation and inflammation. PEG-coated QDs have been shown to persist in bone marrow and lymph nodes for several months following intravenous injection in animal models.172 While this prolonged visibility makes them valuable as long-term imaging probes, it also raises important questions about the potential long-term effects and safety of nanomaterial accumulation in vivo. On the other hand, magnetoelectric nanoparticles were detectable in cortical tissue for about 24 hours, with their stimulatory effects dissipating within three days – suggesting that they are likely cleared from the brain within that period. These findings suggest that nanoparticle retention is highly dependent on both the type of nanomaterial and the specific tissue to which it is delivered, underscoring the importance of thorough in vivo studies and the development of biodegradable materials to ensure long-term safety and clinical viability. The development of biodegradable cores and immunologically inert coatings is imperative, and addressing these challenges is key to harnessing nanotechnology safely and effectively in neuroscience.170,173 Chapter 1. Nano Neuro 38 Semiconductor Nanoparticles in Neurobiology – State of the Art Semiconductor nanocrystals quickly captured the interest of biologists, not only for their exceptional brightness and photostability as fluorescent labels, but also for their potential as tools for neuromodulation. While quantum dot-mediated neuronal stimulation has been explored experimentally,174,175 this thesis will not address that direction. Instead, the focus here is on their optical detection properties. This section first considers their role in visualization, with a later discussion addressing their neuromodulatory potential. A key advantage that distinguishes QDs from conventional dyes is not only their higher quantum yield – often reaching 65-95% compared to 30-50% in standard fluorophores – but also their significantly larger one-, two-, and three-photon absorption cross-sections. These properties help overcome the long-standing trade-off between imaging depth and spatial resolution.176 When discussing the use of semiconductor nanocrystals in deep brain imaging, it is important to distinguish between two main modalities: structural imaging and functional imaging. Structural imaging aims to reveal anatomical features such as vasculature, fiber tracts, and cellular organization. In contrast, functional imaging captures dynamic physiological processes like neural activity, membrane potential fluctuations, or neurotransmitter release. The application of QDs in deep brain imaging began in the early 2000s, with pioneering studies that demonstrated their effectiveness in penetrating deep tissue and providing stable, high-resolution signals. In 2002, Dubertret et al. achieved the first in vivo visualization of QDs by encapsulating them in phospholipid micelles.177 Early breakthroughs began in 2003, when Dahan et al. demonstrated real-time tracking of glycine receptors in living neurons using single QDs, followed by Larson et al. that successfully used water-soluble quantum dots for in vivo multiphoton fluorescence imaging of labeled blood vescels.178–180 The exploration of QDs for deep brain imaging gained significant momentum in the 2010s, with main focus on harnessing QDs as tools for voltage sensing.176 The following discussion examines key developments that have established QDs as a prominent tool in both structural and functional neuroimaging. In structural imaging applications, the high quantum yield and narrow, tunable emission spectra of QDs significantly enhance signal-to-background ratios, particularly at long excitation wavelengths (1.3-2.2 µm), where tissue scattering is minimized. Commercially available probes such as Qtracker655, when excited at 1700 nm, have enabled three-photon imaging to depths of up to 2100 µm in the mouse brain – outperforming conventional fluorophores like Texas Red dextran, which is typically limited to around 1340 µm. Custom-engineered heterostructures, Chapter 1. Nano Neuro 39 including CdSe/CdS/ZnS QDs, have demonstrated imaging depths of 850 µm through an intact skull and 1550 µm following craniotomy. QDs have been employed across multiphoton imaging modalities, supporting two-, three-, and even four-photon fluorescence detection.181 Among these, three-photon excitation offers the greatest penetration depth with reduced risk of photodamage. Additionally, the broad spectral tunability of QDs facilitates multicolor visualization of vasculature, neurons, and microglia.176 Most structural applications have focused on the mouse brain vasculature, with particularly detailed imaging of deep subcortical regions such as the hippocampus. Advances in ligand design are further expanding the potential of QDs, enabling more targeted labeling strategies and improved traversal of the blood-brain barrier.176 Functionally, QDs extend these optical advantages to real-time physiology. Their resistance to photobleaching allows sustained excitation, supporting hemodynamic measurements at significant depths. For example, commercially available Qtracker800 has been used to quantify blood-flow velocities up to 600 µm deep, while Qtracker655 enables detailed vascular imaging at depths reaching 1600 µm. More importantly, the quantum-confined Stark effect and engineered energyor charge-transfer (FRET and ET) constructs equip QDs with millisecondscale voltage sensitivity.176,179 In 2013, Marshall and Schnitzer proposed the use of QDs as fluorescent voltage indicators, leveraging the quantum-confined Stark effect (QCSE) to detect changes in membrane potential.179 This foundational concept was subsequently advanced by the research groups of Shimon Weiss and James B. Delehanty through the development of membrane-inserting QD nanosensors.163 Both simulations and in vivo experiments demonstrated that single action potentials could be resolved using significant larger optical response than with conventional voltage-sensitive dyes or genetically encoded voltage indicators (GEVIs). Notably, Delehanty’s group further introduced bioconjugated QD-fullerene platforms that enabled the detection of electrically evoked cortical responses through charge transfer mechanism (Figure 1.7 A).182 In parallel, the groups of Rafael Yuste and Jonathan Owen183 developed quartz nanopipettes with ultra-narrow tip diameters (15–30 nm), coated with QDs to facilitate two-photon visualization.183 These nanopipettes functioned as minimally invasive nanoelectrodes for intracellular voltage recordings and were successfully targeted to dendritic spines of hippocampal neurons in both cultured cells and acute cortical slices. This approach was further refined in a follow-up study184 where a flexible version of the nanopipette was introduced and deployed in vivo.184 Using cranial microprisms, the authors were able to access and record from Chapter 1. Nano Neuro 40 visually identified pyramidal neurons and interneurons in both anesthetized and awake, head-fixed mice. Complementing these advances, S. Weiss, K. Park, and colleagues (2018) demonstrated that rod-shaped QDs functionalized with specific peptides could spontaneously insert into lipid membranes and detect single-particle voltage responses in model systems (Figure 1.7 C).185 These results provided experimental confirmation of prior theoretical work by the same group.112,186 Building on this work, S. Weiss together with A. Ludwig and collaborators reported in 2020 that type-II ZnSe/CdS seeded nanorods (NRs), when functionalized with a lipid mixture derived from brain tissue, could spontaneously incorporate into neuronal plasma membranes.187 This insertion was driven by the intrinsic compatibility of the nanorod coating with the native lipid composition of the membrane. Once embedded, the nanorods, operating through QCSE, were able to sense local changes in membrane potential. This functionality was demonstrated across multiple systems, including self-spiking and patched HEK293 cells, as well as primary cortical neurons – all without the need for genetic modification.187,188 Another recent study by Qiangbin Wang’s group189 demonstrated that glutathione-capped CdSe/ZnS quantum dots can localize to neuronal membranes and function as voltage sensors (Figure 1.7 B). The sensing mechanism relies on Förster resonance energy transfer (FRET), with dipicrylamine (DPA) acting as an acceptor within the lipid bilayer. Changes in membrane potential alter the position of DPA relative to the QD surface: during depolarization or hyperpolarization, DPA migrates within the membrane, thereby modulating the donor–acceptor distance and, consequently, the fluorescence intensity of QDs.189 Caglar et al. (2019) demonstrated that InP/ZnS QDs could detect subthreshold voltage fluctuations in live cells with greater sensitivity than calcium-based indicators, which is a significant progress to offer improved biocompatibility.190 Chapter 1. Nano Neuro 47 4. References for Chapter 1 (1) Garcia-Etxarri, A.; Yuste, R. Time for NanoNeuro. Nat. Methods 2021. https://doi.org/10.1038/s41592-021-01270-9. (2) Langthaler, S.; Lozanović Šajić, J.; Rienmüller, T.; Weinberg, S. H.; et al. Ion Channel Modeling beyond State of the Art: A Comparison with a System Theory-Based Model of the Shaker-Related Voltage-Gated Potassium Channel Kv1.1. Cells 2022, 11 (2), 239. https://doi.org/10.3390/cells11020239. (3) Hübel, N.; Dahlem, M. A. Dynamics from Seconds to Hours in Hodgkin-Huxley Model with Time-Dependent Ion Concentrations and Buffer Reservoirs. PLoS Comput. Biol. 2014, 10 (12), e1003941. https://doi.org/10.1371/journal.pcbi.1003941. (4) Yuste, R. Lectures in Neuroscience; Columbia University Press, 2023. (5) Lyons, D. A.; Nicola J. Allen. Glia as Architects of Central Nervous System Formation and Function. Science 2018, 362 (6411), 181–185. https://doi.org/10.1126/science.aat0473. (6) Stuart, G. J.; Spruston, N. Dendritic Integration: 60 Years of Progress. Nat. Neurosci. 2015, 18 (12), 1713–1721. https://doi.org/10.1038/nn.4157. (7) Fields, R. D. A New Mechanism of Nervous System Plasticity: Activity-Dependent Myelination. Nat. Rev. Neurosci. 2015, 16 (12), 756–767. https://doi.org/10.1038/nrn4023. (8) Eric Kandel, John D. Koester, Sarah H. Mack, Steven Siegelbaum. Principles of Neural Science, Sixth Edition; McGraw-Hill Education, 2021. (9) Kole, M. H. P.; Stuart, G. J. Signal Processing in the Axon Initial Segment. Neuron 2012, 73 (2), 235–247. https://doi.org/10.1016/j.neuron.2012.01.007. (10) Hwang, S.; Chang, J.; Oh, M.-H.; Lee, J.-H.; et al. Impact of the Sub-Resting Membrane Potential on Accurate Inference in Spiking Neural Networks. Sci. Rep. 2020, 10 (1), 3515. https://doi.org/10.1038/s41598-020-60572-8. (11) Bean, B. P. The Action Potential in Mammalian Central Neurons. Nat. Rev. Neurosci. 2007, 8 (6), 451–465. https://doi.org/10.1038/nrn2148. (12) Geoffrey M Cooper. The Cell, 2nd ed.; Sinauer Associates, 2000. (13) Yoo, M.; Yang, Y.-S.; Rah, J.-C.; Choi, J. H. Different Resting Membrane Potentials in Posterior Parietal Cortex and Prefrontal Cortex in the View of Recurrent Synaptic Strengths and Neural Network Dynamics. Front. Cell. Neurosci. 2023, 17. https://doi.org/10.3389/fncel.2023.1153970. (14) Gerstner, W.; Kistler, W. M.; Naud, R.; Paninski, L. Neuronal Dynamics: From Single Neurons to Networks and Models of Cognition; Cambridge University Press, 2014. (15) David Cardozo. An intuitive approach to understanding the resting membrane potential. https://doi.org/10.1152/advan.00049.2016. (16) Melendy, R. F. A Single Differential Equation Description of Membrane Properties Underlying the Action Potential and the Axon Electric Field. J. Electr. Bioimpedance 2018, 9 (1), 106–114. https://doi.org/10.2478/joeb-2018-0015. (17) Kennedy, M. B. Synaptic Signaling in Learning and Memory. Cold Spring Harb. Perspect. Biol. 2016, 8 (2), a016824. https://doi.org/10.1101/cshperspect.a016824. Chapter 1. Nano Neuro 48 (18) Citri, A.; Malenka, R. C. Synaptic Plasticity: Multiple Forms, Functions, and Mechanisms. Neuropsychopharmacology 2008, 33 (1), 18–41. https://doi.org/10.1038/sj.npp.1301559. (19) Goda, Y.; Davis, G. W. Mechanisms of Synapse Assembly and Disassembly. Neuron 2003, 40 (2), 243–264. https://doi.org/10.1016/S0896-6273(03)00608-1. (20) Caterina, M. J.; Schumacher, M. A.; Tominaga, M.; Rosen, T. A.; et al. The Capsaicin Receptor: A Heat-Activated Ion Channel in the Pain Pathway. Nature 1997, 389 (6653), 816–824. https://doi.org/10.1038/39807. (21) McKemy, D. D.; Neuhausser, W. M.; Julius, D. Identification of a Cold Receptor Reveals a General Role for TRP Channels in Thermosensation. Nature 2002, 416 (6876), 52–58. https://doi.org/10.1038/nature719. (22) Coste, B.; Mathur, J.; Schmidt, M.; Earley, T. J.; et al. Piezo1 and Piezo2 Are Essential Components of Distinct Mechanically Activated Cation Channels. Science 2010, 330 (6000), 55–60. https://doi.org/10.1126/science.1193270. (23) Baez, D.; Raddatz, N.; Ferreira, G.; Gonzalez, C.; et al. Chapter Three - Gating of Thermally Activated Channels. In Current Topics in Membranes; Islas, L. D., Qin, F., Eds.; Thermal Sensors; Academic Press, 2014; Vol. 74, pp 51–87. https://doi.org/10.1016/B978-0-12800181-3.00003-8. (24) Lamas, J. A.; Rueda-Ruzafa, L.; Herrera-Pérez, S. Ion Channels and Thermosensitivity: TRP, TREK, or Both? Int. J. Mol. Sci. 2019, 20 (10), 2371. https://doi.org/10.3390/ijms20102371. (25) Uchida, K. TRPM3, TRPM4, and TRPM5 as Thermo-Sensitive Channels. J. Physiol. Sci. 2024, 74 (1), 43. https://doi.org/10.1186/s12576-024-00937-0. (26) Luan, S.; Williams, I.; Nikolic, K.; Constandinou, T. G. Neuromodulation: Present and Emerging Methods. Front. Neuroengineering 2014, 7. https://doi.org/10.3389/fneng.2014.00027. (27) York, G. K.; Steinberg, D. A. Chapter 3 Neurology in Ancient Egypt. In Handbook of Clinical Neurology; Aminoff, M. J., Boller, F., Swaab, D. F., Eds.; History of Neurology; Elsevier, 2009; Vol. 95, pp 29–36. https://doi.org/10.1016/S0072-9752(08)02103-9. (28) Keshavan, M. S.; (Michael) Song, S. H.; Zhang, Y.; Lizano, P. Neuroscience in Pictures:1. History of Psychiatric Neuroscience. Asian J. Psychiatry 2024, 92, 103869. https://doi.org/10.1016/j.ajp.2023.103869. (29) Glickstein, M. Golgi and Cajal: The Neuron Doctrine and the 100th Anniversary of the 1906 Nobel Prize. Curr. Biol. 2006, 16 (5), R147–R151. https://doi.org/10.1016/j.cub.2006.02.053. (30) Advanced Imaging Methods in Neuroscience; Malva, J. O., Valero, J., Castelo-Branco, M., Roebroeck, A., Eds.; Frontiers Research Topics; Frontiers Media SA, 2022. https://doi.org/10.3389/978-2-88971-725-5. (31) Roth, R. H.; Ding, J. B. From Neurons to Cognition: Technologies for Precise Recording of Neural Activity Underlying Behavior. BME Front. 2020, 2020, 7190517. https://doi.org/10.34133/2020/7190517. (32) Ghovanloo, M.-R.; Dib-Hajj, S. D.; Waxman, S. G. The Evolution of Patch-Clamp Electrophysiology: Robotic, Multiplex, and Dynamic. Mol. Pharmacol. 2025, 107 (1), 100001. https://doi.org/10.1124/molpharm.124.000954. Chapter 1. Nano Neuro 49 (33) Ferrea, E.; Maccione, A.; Medrihan, L.; Nieus, T.; et al. Large-Scale, High-Resolution Electrophysiological Imaging of Field Potentials in Brain Slices with Microelectronic Multielectrode Arrays. Front. Neural Circuits 2012, 6, 80. https://doi.org/10.3389/fncir.2012.00080. (34) Soscia, D. A.; Lam, D.; Tooker, A. C.; Enright, H. A.; et al. A Flexible 3-Dimensional Microelectrode Array for in Vitro Brain Models. Lab. Chip 2020, 20 (5), 901–911. https://doi.org/10.1039/C9LC01148J. (35) Jun, J. J.; Steinmetz, N. A.; Siegle, J. H.; Denman, D. J.; et al. Fully Integrated Silicon Probes for High-Density Recording of Neural Activity. Nature 2017, 551 (7679), 232–236. https://doi.org/10.1038/nature24636. (36) Durand, S.; Heller, G. R.; Ramirez, T. K.; Luviano, J. A.; et al. Acute Head-Fixed Recordings in Awake Mice with Multiple Neuropixels Probes. Nat. Protoc. 2023, 18 (2), 424–457. https://doi.org/10.1038/s41596-022-00768-6. (37) Two-Photon Calcium Imaging of Neuronal Activity. Nat. Rev. Methods Primer 2022, 2 (1), 1–1. https://doi.org/10.1038/s43586-022-00160-4. (38) Bando, Y.; Sakamoto, M.; Kim, S.; Ayzenshtat, I.; et al. Comparative Evaluation of Genetically Encoded Voltage Indicators. Cell Rep. 2019, 26 (3), 802-813.e4. https://doi.org/10.1016/j.celrep.2018.12.088. (39) Miller, E. W. Small Molecule Fluorescent Voltage Indicators for Studying Membrane Potential. Curr. Opin. Chem. Biol. 2016, 33, 74–80. https://doi.org/10.1016/j.cbpa.2016.06.003. (40) Peng, L.; Xu, Y.; Zou, P. Genetically-Encoded Voltage Indicators. Chin. Chem. Lett. 2017, 28 (10), 1925–1928. https://doi.org/10.1016/j.cclet.2017.09.037. (41) Wright, E. C.; Scott, E.; Tian, L. Applications of Functional Neurotransmitter Release Imaging with Genetically Encoded Sensors in Psychiatric Research. Neuropsychopharmacology 2025, 50 (1), 269–273. https://doi.org/10.1038/s41386-02401903-5. (42) Aggarwal, A.; Liu, R.; Chen, Y.; Ralowicz, A. J.; et al. Glutamate Indicators with Improved Activation Kinetics and Localization for Imaging Synaptic Transmission. Nat. Methods 2023, 20 (6), 925–934. https://doi.org/10.1038/s41592-023-01863-6. (43) Bade, A.; Yadav, P.; Zhang, L.; Naidu Bypaneni, R.; et al. Imaging Neurotransmitters with Small-Molecule Fluorescent Probes. Angew. Chem. Int. Ed. 2024, 63 (34), e202406401. https://doi.org/10.1002/anie.202406401. (44) Szyszka, Ł.; Górecki, M.; Cmoch, P.; Jarosz, S. Fluorescent Molecular Cages with Sucrose and Cyclotriveratrylene Units for the Selective Recognition of Choline and Acetylcholine. J. Org. Chem. 2021, 86 (7), 5129–5141. https://doi.org/10.1021/acs.joc.1c00019. (45) Babiloni, C.; Pizzella, V.; Gratta, C. D.; Ferretti, A.; et al. Chapter 5 Fundamentals of Electroencefalography, Magnetoencefalography, and Functional Magnetic Resonance Imaging. In International Review of Neurobiology; Academic Press, 2009; Vol. 86, pp 67– 80. https://doi.org/10.1016/S0074-7742(09)86005-4. (46) Babaeeghazvini, P.; Rueda-Delgado, L. M.; Gooijers, J.; Swinnen, S. P.; et al. Brain Structural and Functional Connectivity: A Review of Combined Works of Diffusion Magnetic Resonance Imaging and Electro-Encephalography. Front. Hum. Neurosci. 2021, 15. https://doi.org/10.3389/fnhum.2021.721206. Chapter 1. Nano Neuro 50 (47) Thomas, D.; Tovey, S. C.; Collins, T. J.; Bootman, M. D.; et al. A Comparison of Fluorescent Ca2+indicator Properties and Their Use in Measuring Elementary and Global Ca2+signals. Cell Calcium 2000, 28 (4), 213–223. https://doi.org/10.1054/ceca.2000.0152. (48) Berridge, M. J.; Lipp, P.; Bootman, M. D. The Versatility and Universality of Calcium Signalling. Nat. Rev. Mol. Cell Biol. 2000, 1 (1), 11–21. https://doi.org/10.1038/35036035. (49) Grienberger, C.; Konnerth, A. Imaging Calcium in Neurons. Neuron 2012, 73 (5), 862–885. https://doi.org/10.1016/j.neuron.2012.02.011. (50) Paredes, R. M.; Etzler, J. C.; Watts, L. T.; Zheng, W.; et al. Chemical Calcium Indicators. Methods San Diego Calif 2008, 46 (3), 143–151. https://doi.org/10.1016/j.ymeth.2008.09.025. (51) Mertes, N.; Busch, M.; Huppertz, M.-C.; Hacker, C. N.; et al. Fluorescent and Bioluminescent Calcium Indicators with Tuneable Colors and Affinities. J. Am. Chem. Soc. 2022, 144 (15), 6928–6935. https://doi.org/10.1021/jacs.2c01465. (52) Bootman, M. D. Calcium Signaling. Cold Spring Harb. Perspect. Biol. 2012, 4 (7), a011171. https://doi.org/10.1101/cshperspect.a011171. (53) Lent, R.; Azevedo, F. A. C.; Andrade-Moraes, C. H.; Pinto, A. V. O. How Many Neurons Do You Have? Some Dogmas of Quantitative Neuroscience under Revision. Eur. J. Neurosci. 2012, 35 (1), 1–9. https://doi.org/10.1111/j.1460-9568.2011.07923.x. (54) Neziri, S.; Köseoğlu, A. E.; Deniz Köseoğlu, G.; Özgültekin, B.; et al. Animal Models in Neuroscience with Alternative Approaches: Evolutionary, Biomedical, and Ethical Perspectives. Anim. Models Exp. Med. 2024, 7 (6), 868–880. https://doi.org/10.1002/ame2.12487. (55) Lear, A.; Baker, S. N.; Clarke, H. F.; Roberts, A. C.; et al. Understanding Them to Understand Ourselves: The Importance of NHP Research for Translational Neuroscience. Curr. Res. Neurobiol. 2022, 3, 100049. https://doi.org/10.1016/j.crneur.2022.100049. (56) Capitanio, J. P.; Emborg, M. E. Contributions of Non-Human Primates to Neuroscience Research. The Lancet 2008, 371 (9618), 1126–1135. https://doi.org/10.1016/S01406736(08)60489-4. (57) Higo, N. Non-Human Primate Models to Explore the Adaptive Mechanisms After Stroke. Front. Syst. Neurosci. 2021, 15. https://doi.org/10.3389/fnsys.2021.760311. (58) Dennis, E. J.; Hady, A. E.; Michaiel, A.; Clemens, A.; et al. Systems Neuroscience of Natural Behaviors in Rodents. J. Neurosci. 2021, 41 (5), 911–919. https://doi.org/10.1523/JNEUROSCI.1877-20.2020. (59) Feketa, V. V.; Bagriantsev, S. N.; Gracheva, E. O. Ground Squirrels – Experts in Thermoregulatory Adaptation. Trends Neurosci. 2023, 46 (7), 505–507. https://doi.org/10.1016/j.tins.2023.04.008. (60) Mohr, S. M.; Dai Pra, R.; Platt, M. P.; Feketa, V. V.; et al. Hypothalamic Hormone Deficiency Enables Physiological Anorexia in Ground Squirrels during Hibernation. Nat. Commun. 2024, 15 (1), 5803. https://doi.org/10.1038/s41467-024-49996-2. (61) Doszyn, O.; Dulski, T.; Zmorzynska, J. Diving into the Zebrafish Brain: Exploring Neuroscience Frontiers with Genetic Tools, Imaging Techniques, and Behavioral Insights. Front. Mol. Neurosci. 2024, 17, 1358844. https://doi.org/10.3389/fnmol.2024.1358844. (62) Quadros-Mennella, P. S.; Lucin, K. M.; White, R. E. What Can the Common Fruit Fly Teach Us about Stroke?: Lessons Learned from the Hypoxic Tolerant Drosophila Melanogaster. Front. Cell. Neurosci. 2024, 18. https://doi.org/10.3389/fncel.2024.1347980. Chapter 1. Nano Neuro 51 (63) Moulin, T. C.; Covill, L. E.; Itskov, P. M.; Williams, M. J.; et al. Rodent and Fly Models in Behavioral Neuroscience: An Evaluation of Methodological Advances, Comparative Research, and Future Perspectives. Neurosci. Biobehav. Rev. 2021, 120, 1–12. https://doi.org/10.1016/j.neubiorev.2020.11.014. (64) Lages, Y. V.; McNaughton, N. Non-Human Contributions to Personality Neuroscience – from Fish through Primates. An Introduction to the Special Issue. Personal. Neurosci. 2022, 5, e11. https://doi.org/10.1017/pen.2022.4. (65) Hanson, A.; Reme, R.; Telerman, N.; Yamamoto, W.; et al. Automatic Monitoring of Neural Activity with Single-Cell Resolution in Behaving Hydra. Sci. Rep. 2024, 14 (1), 5083. https://doi.org/10.1038/s41598-024-55608-2. (66) Soto Veliz, D.; Lin, K.-L.; Sahlgren, C. Organ-on-a-Chip Technologies for Biomedical Research and Drug Development: A Focus on the Vasculature. Smart Med. 2023, 2 (1), e20220030. https://doi.org/10.1002/SMMD.20220030. (67) Hanson, A. On Being a Hydra with, and without, a Nervous System: What Do Neurons Add? Anim. Cogn. 2023, 26 (6), 1799–1816. https://doi.org/10.1007/s10071-023-018168. (68) Kanwisher, N. Animal Models of the Human Brain: Successes, Limitations, and Alternatives. Curr. Opin. Neurobiol. 2025, 90, 102969. https://doi.org/10.1016/j.conb.2024.102969. (69) Ethical Issues in Behavioral Neuroscience; Lee, G., Illes, J., Ohl, F., Eds.; Current Topics in Behavioral Neurosciences; Springer Berlin Heidelberg: Berlin, Heidelberg, 2015; Vol. 19. https://doi.org/10.1007/978-3-662-44866-3. (70) Ghiasvand, K.; Amirfazli, M.; Moghimi, P.; Safari, F.; et al. The Role of Neuron-like Cell Lines and Primary Neuron Cell Models in Unraveling the Complexity of Neurodegenerative Diseases: A Comprehensive Review. Mol. Biol. Rep. 2024, 51 (1), 1024. https://doi.org/10.1007/s11033-024-09964-x. (71) Zhang, J.; Yang, H.; Wu, J.; Zhang, D.; et al. Recent Progresses in Novel in Vitro Models of Primary Neurons: A Biomaterial Perspective. Front. Bioeng. Biotechnol. 2022, 10. https://doi.org/10.3389/fbioe.2022.953031. (72) Obinata, M. The Immortalized Cell Lines with Differentiation Potentials: Their Establishment and Possible Application. Cancer Sci. 2007, 98 (3), 275–283. https://doi.org/10.1111/j.1349-7006.2007.00399.x. (73) He, B.; Soderlund, D. M. Human Embryonic Kidney (HEK293) Cells Express Endogenous Voltage-Gated Sodium Currents and Nav1.7 Sodium Channels. Neurosci. Lett. 2010, 469 (2), 268–272. https://doi.org/10.1016/j.neulet.2009.12.012. (74) Collingridge, G. L. The Brain Slice Preparation: A Tribute to the Pioneer Henry Mcllfwain. (75) Khurana, S.; Li, W.-K. Baptisms of Fire or Death Knells for Acute-Slice Physiology in the Age of ‘Omics’ and Light? Rev. Neurosci. 2013, 24 (5), 527–536. https://doi.org/10.1515/revneuro-2013-0028. (76) Pașca, S. P.; Arlotta, P.; Bateup, H. S.; Camp, J. G.; et al. A Framework for Neural Organoids, Assembloids and Transplantation Studies. Nature 2025, 639 (8054), 315–320. https://doi.org/10.1038/s41586-024-08487-6. (77) Kim, Y.; Kim, I.; Shin, K. A New Era of Stem Cell and Developmental Biology: From Blastoids to Synthetic Embryos and Beyond. Exp. Mol. Med. 2023, 55 (10), 2127–2137. https://doi.org/10.1038/s12276-023-01097-8. Chapter 1. Nano Neuro 52 (78) Joudeh, N.; Linke, D. Nanoparticle Classification, Physicochemical Properties, Characterization, and Applications: A Comprehensive Review for Biologists. J. Nanobiotechnology 2022, 20 (1), 262. https://doi.org/10.1186/s12951-022-01477-8. (79) Sumanth Kumar, D.; Jai Kumar, B.; Mahesh, H. M. In Quantum Nanostructures (QDs): An Overview; Elsevier Ltd., 2018. https://doi.org/10.1016/b978-0-08-101975-7.00003-8. (80) Altammar, K. A. A Review on Nanoparticles: Characteristics, Synthesis, Applications, and Challenges. Front. Microbiol. 2023, 14. https://doi.org/10.3389/fmicb.2023.1155622. (81) Kolahalam, L. A.; Kasi Viswanath, I. V.; Diwakar, B. S.; Govindh, B.; et al. Review on Nanomaterials: Synthesis and Applications. Mater. Today Proc. 2019, 18, 2182–2190. https://doi.org/10.1016/j.matpr.2019.07.371. (82) Algar, W. R.; Massey, M.; Rees, K.; Higgins, R.; et al. Photoluminescent Nanoparticles for Chemical and Biological Analysis and Imaging. Chem. Rev. 2021, 121 (15), 9243–9358. https://doi.org/10.1021/acs.chemrev.0c01176. (83) Zhang, J.; Zhang, S.; Zhang, Y.; Al-Hartomy, O. A.; et al. Colloidal Quantum Dots: Synthesis, Composition, Structure, and Emerging Optoelectronic Applications. Laser Photonics Rev. 2023, 17 (3), 2200551. https://doi.org/10.1002/lpor.202200551. (84) Park, J.; Joo, J.; Soon, G. K.; Jang, Y.; et al. Synthesis of Monodisperse Spherical Nanocrystals. Angew. Chem. - Int. Ed. 2007, 46 (25), 4630–4660. https://doi.org/10.1002/anie.200603148. (85) Murray, C. B.; Norris, D. J.; Bawendi, M. G. Synthesis and Characterization of Nearly Monodisperse CdE (E = S, Se, Te) Semiconductor Nanocrystallites. J. Am. Chem. Soc. 1993, 115 (19), 8706–8715. https://doi.org/10.1021/ja00072a025. (86) Battaglia, D.; Li, J. J.; Wang, Y.; Peng, X. Colloidal Two‐Dimensional Systems: CdSe Quantum Shells and Wells. Angew. Chem. Int. Ed. 2003, 42 (41), 5035–5039. https://doi.org/10.1002/anie.200352120. (87) Kagan, C. R.; Lifshitz, E.; Sargent, E. H.; Talapin, D. V. Building Devices from Colloidal Quantum Dots. https://doi.org/10.1126/science.aac5523. (88) Rasch, M. R.; Rossinyol, E.; Hueso, J. L.; Goodfellow, B. W.; et al. Hydrophobic Gold Nanoparticle Self-Assembly with Phosphatidylcholine Lipid: Membrane-Loaded and Janus Vesicles. Nano Lett. 2010, 10 (9), 3733–3739. https://doi.org/10.1021/nl102387n. (89) Sayed, M.; Yu, J.; Liu, G.; Jaroniec, M. Non-Noble Plasmonic Metal-Based Photocatalysts. Chem. Rev. 2022, 122 (11), 10484–10537. https://doi.org/10.1021/acs.chemrev.1c00473. (90) Khan, I.; Saeed, K.; Khan, I. Nanoparticles: Properties, Applications and Toxicities. Arab. J. Chem. 2019, 12 (7), 908–931. https://doi.org/10.1016/j.arabjc.2017.05.011. (91) Mittal, A.; Roy, I.; Gandhi, S. Magnetic Nanoparticles: An Overview for Biomedical Applications. Magnetochemistry 2022, 8 (9), 107. https://doi.org/10.3390/magnetochemistry8090107. (92) Ayanda, O. S.; Mmuoegbulam, A. O.; Okezie, O.; Durumin Iya, N. I.; et al. Recent Progress in Carbon-Based Nanomaterials: Critical Review. J. Nanoparticle Res. 2024, 26 (5), 106. https://doi.org/10.1007/s11051-024-06006-2. (93) Marinho, T.; Costa, P.; Lizundia, E.; Costa, C. M.; et al. Ceramic Nanoparticles and Carbon Nanotubes Reinforced Thermoplastic Materials for Piezocapacitive Sensing Applications. Compos. Sci. Technol. 2019, 183, 107804. https://doi.org/10.1016/j.compscitech.2019.107804. Chapter 1. Nano Neuro 53 (94) Wang, N.; Thameem Dheen, S.; Fuh, J. Y. H.; Senthil Kumar, A. A Review of MultiFunctional Ceramic Nanoparticles in 3D Printed Bone Tissue Engineering. Bioprinting 2021, 23, e00146. https://doi.org/10.1016/j.bprint.2021.e00146. (95) Xu, Y.; Fourniols, T.; Labrak, Y.; Préat, V.; et al. Surface Modification of Lipid-Based Nanoparticles. ACS Nano 2022, 16 (5), 7168–7196. https://doi.org/10.1021/acsnano.2c02347. (96) Puri, A.; Loomis, K.; Smith, B.; Lee, J.-H.; et al. Lipid-Based Nanoparticles as Pharmaceutical Drug Carriers: From Concepts to Clinic. Crit. Rev. Ther. Drug Carr. Syst. 2009, 26 (6). https://doi.org/10.1615/CritRevTherDrugCarrierSyst.v26.i6.10. (97) Zielińska, A.; Carreiró, F.; Oliveira, A. M.; Neves, A.; et al. Polymeric Nanoparticles: Production, Characterization, Toxicology and Ecotoxicology. Molecules 2020, 25 (16), 3731. https://doi.org/10.3390/molecules25163731. (98) Anu Mary Ealia, S.; Saravanakumar, M. P. A Review on the Classification, Characterisation, Synthesis of Nanoparticles and Their Application. IOP Conf. Ser. Mater. Sci. Eng. 2017, 263 (3), 032019. https://doi.org/10.1088/1757-899X/263/3/032019. (99) Ann Fernholm; Peter Brzezinski; Heiner Linke; Johan Åqvist. They Added Colour to Nanotechnology. 2023. (100) Quesada-González, D.; Merkoçi, A. Quantum Dots for Biosensing: Classification and Applications. Biosens. Bioelectron. 2025, 273, 117180. https://doi.org/10.1016/j.bios.2025.117180. (101) Bera, D.; Qian, L.; Tseng, T.-K.; Holloway, P. H. Quantum Dots and Their Multimodal Applications: A Review. Materials 2010, 3 (4), 2260–2345. https://doi.org/10.3390/ma3042260. (102) Bae, W. K.; Padilha, L. A.; Park, Y.-S.; McDaniel, H.; et al. Controlled Alloying of the Core– Shell Interface in CdSe/CdS Quantum Dots for Suppression of Auger Recombination. ACS Nano 2013, 7 (4), 3411–3419. https://doi.org/10.1021/nn4002825. (103) Yong, K. T.; Sahoo, Y.; Swihart, M. T.; Prasad, P. N. Shape Control of CdS Nanocrystals in One-Pot Synthesis. J. Phys. Chem. C 2007, 111 (6), 2447–2458. https://doi.org/10.1021/jp066392z. (104) Lawera, Z.; Parzyszek, S.; Pociecha, D.; Lewandowski, W. Small CdS Nanorods via Sacrificial Synthesis on Perovskite Nanocrystals – Synthesis and Hierarchical Assembly. J. Mater. Chem. C 2024, 12 (16), 5793–5800. https://doi.org/10.1039/D3TC03556E. (105) Lim, S. J.; Smith, A.; Nie, S. The More Exotic Shapes of Semiconductor Nanocrystals: Emerging Applications in Bioimaging. Curr. Opin. Chem. Eng. 2014, 4, 137–143. https://doi.org/10.1016/j.coche.2014.01.013. (106) García de Arquer, F. P.; Talapin, D. V.; Klimov, V. I.; Arakawa, Y.; et al. Semiconductor Quantum Dots: Technological Progress and Future Challenges. Science 2021, 373 (6555), eaaz8541. https://doi.org/10.1126/science.aaz8541. (107) Chae, W. S.; Shin, H. W.; Lee, E. S.; Shin, E. J.; et al. Excitation Dynamics in Anisotropic Nanostructures of Star-Shaped CdS. J. Phys. Chem. B 2005, 109 (13), 6204–6209. https://doi.org/10.1021/jp044402v. (108) Ghasemi, M.; Hao, M.; Xiao, M.; Chen, P.; et al. Lead-Free Metal-Halide Double Perovskites: From Optoelectronic Properties to Applications; 2021. https://doi.org/10.1515/nanoph-2020-0548. Chapter 1. Nano Neuro 54 (109) Almeida, G.; Ubbink, R. F.; Stam, M.; du Fossé, I.; et al. InP Colloidal Quantum Dots for Visible and Near-Infrared Photonics. Nat. Rev. Mater. 2023, 8 (11), 742–758. https://doi.org/10.1038/s41578-023-00596-4. (110) Wen, G. W.; Lin, J. Y.; Jiang, H. X.; Chen, Z. Quantum-Confined Stark Effects in Semiconductor Quantum Dots. Phys. Rev. B 1995, 52 (8), 5913–5922. https://doi.org/10.1103/PhysRevB.52.5913. (111) Empedocles, S. A.; Bawendi, M. G. Quantum-Confined Stark Effect in Single CdSe Nanocrystallite Quantum Dots. Science 1997, 278 (5346), 2114–2117. https://doi.org/10.1126/science.278.5346.2114. (112) Park, K.; Deutsch, Z.; Li, J. J.; Oron, D.; et al. Single Molecule Quantum-Confined Stark Effect Measurements of Semiconductor Nanoparticles at Room Temperature. ACS Nano 2012, 6 (11), 10013–10023. https://doi.org/10.1021/nn303719m. (113) Kuo, Y.; Li, J.; Michalet, X.; Chizhik, A.; et al. Characterizing the Quantum-Confined Stark Effect in Semiconductor Quantum Dots and Nanorods for Single-Molecule Electrophysiology. ACS Photonics 2018, 5 (12), 4788–4800. https://doi.org/10.1021/acsphotonics.8b00617. (114) Chen, C.; Yan, J.-Y.; Babin, H.-G.; Wang, J.; et al. Wavelength-Tunable High-Fidelity Entangled Photon Sources Enabled by Dual Stark Effects. Nat. Commun. 2024, 15 (1), 5792. https://doi.org/10.1038/s41467-024-50062-0. (115) Walters, G.; Wei, M.; Voznyy, O.; Quintero-Bermudez, R.; et al. The Quantum-Confined Stark Effect in Layered Hybrid Perovskites Mediated by Orientational Polarizability of Confined Dipoles. Nat. Commun. 2018, 9 (1), 4214. https://doi.org/10.1038/s41467-01806746-5. (116) Miller, D. A. B.; Chemla, D. S.; Damen, T. C.; Gossard, A. C.; et al. Electric Field Dependence of Optical Absorption near the Band Gap of Quantum-Well Structures. Phys. Rev. B 1985, 32 (2), 1043–1060. https://doi.org/10.1103/PhysRevB.32.1043. (117) Efros, A. L.; Nesbitt, D. J. Origin and Control of Blinking in Quantum Dots. Nat. Nanotechnol. 2016, 11 (8), 661–671. https://doi.org/10.1038/nnano.2016.140. (118) Bae, W. K.; Park, Y.-S.; Lim, J.; Lee, D.; et al. Controlling the Influence of Auger Recombination on the Performance of Quantum-Dot Light-Emitting Diodes. Nat. Commun. 2013, 4 (1), 2661. https://doi.org/10.1038/ncomms3661. (119) Cragg, G. E.; Efros, A. L. Suppression of Auger Processes in Confined Structures. Nano Lett. 2010, 10 (1), 313–317. https://doi.org/10.1021/nl903592h. (120) Baffou, G.; Cichos, F.; Quidant, R. Applications and Challenges of Thermoplasmonics. Nat. Mater. 2020, 19 (9), 946–958. https://doi.org/10.1038/s41563-020-0740-6. (121) Baffou, G.; Quidant, R. Thermo-Plasmonics: Using Metallic Nanostructures as NanoSources of Heat. Laser Photonics Rev. 2013, 7 (2), 171–187. https://doi.org/10.1002/lpor.201200003. (122) Yan, T.; Su, M.; Wang, Z.; Zhang, J. Second Near‐Infrared Plasmonic Nanomaterials for Photoacoustic Imaging and Photothermal Therapy. Small 2023, 19 (30), 2300539. https://doi.org/10.1002/smll.202300539. (123) Indhu, A. R.; Keerthana, L.; Dharmalingam, G. Plasmonic Nanotechnology for Photothermal Applications – an Evaluation. Beilstein J. Nanotechnol. 2023, 14, 380–419. https://doi.org/10.3762/bjnano.14.33. Chapter 1. Nano Neuro 55 (124) Kant, K.; Beeram, R.; Cao, Y.; Santos, P. S. S. dos; et al. Plasmonic Nanoparticle Sensors: Current Progress, Challenges, and Future Prospects. Nanoscale Horiz. 2024, 9 (12), 2085– 2166. https://doi.org/10.1039/D4NH00226A. (125) Aslam, U.; Rao, V. G.; Chavez, S.; Linic, S. Catalytic Conversion of Solar to Chemical Energy on Plasmonic Metal Nanostructures. Nat. Catal. 2018, 1 (9), 656–665. https://doi.org/10.1038/s41929-018-0138-x. (126) Zhan, C.; Chen, X.-J.; Yi, J.; Li, J.-F.; et al. From Plasmon-Enhanced Molecular Spectroscopy to Plasmon-Mediated Chemical Reactions. Nat. Rev. Chem. 2018, 2 (9), 216–230. https://doi.org/10.1038/s41570-018-0031-9. (127) Zare, I.; Yaraki, M. T.; Speranza, G.; Najafabadi, A. H.; et al. Gold Nanostructures: Synthesis, Properties, and Neurological Applications. Chem. Soc. Rev. 2022, 51 (7), 2601– 2680. https://doi.org/10.1039/D1CS01111A. (128) Loiseau, A.; Asila, V.; Boitel-Aullen, G.; Lam, M.; et al. Silver-Based Plasmonic Nanoparticles for and Their Use in Biosensing. Biosensors 2019, 9 (2), 78. https://doi.org/10.3390/bios9020078. (129) Zheng, J.; Cheng, X.; Zhang, H.; Bai, X.; et al. Gold Nanorods: The Most Versatile Plasmonic Nanoparticles. Chem. Rev. 2021, 121 (21), 13342–13453. https://doi.org/10.1021/acs.chemrev.1c00422. (130) Hammami, I.; Alabdallah, N. M.; Jomaa, A. A.; Kamoun, M. Gold Nanoparticles: Synthesis Properties and Applications. J. King Saud Univ. - Sci. 2021, 33 (7), 101560. https://doi.org/10.1016/j.jksus.2021.101560. (131) Amina, S. J.; Guo, B. A Review on the Synthesis and Functionalization of Gold Nanoparticles as a Drug Delivery Vehicle. Int. J. Nanomedicine 2020, Volume 15, 9823– 9857. https://doi.org/10.2147/IJN.S279094. (132) Goldmann, C.; Li, X.; Kociak, M.; Constantin, D.; et al. Longitudinal and Transversal Directed Overgrowth of Pentatwinned Silver Nanorods with Tunable Optical Properties. J. Phys. Chem. C 2022, 126 (28), 11667–11673. https://doi.org/10.1021/acs.jpcc.2c02846. (133) Lee, J.-H.; Gibson, K. J.; Chen, G.; Weizmann, Y. Bipyramid-Templated Synthesis of Monodisperse Anisotropic Gold Nanocrystals. Nat. Commun. 2015, 6 (1), 7571. https://doi.org/10.1038/ncomms8571. (134) Liu, M.; Guyot-Sionnest, P. Mechanism of Silver(I)-Assisted Growth of Gold Nanorods and Bipyramids. J. Phys. Chem. B 2005, 109 (47), 22192–22200. https://doi.org/10.1021/jp054808n. (135) Sánchez-Iglesias, A.; Grzelczak, M. Expanding Chemical Space in the Synthesis of Gold Bipyramids. Small 2025, 21 (2), 2407735. https://doi.org/10.1002/smll.202407735. (136) Becerril-Castro, I. B.; Calderon, I.; Pazos-Perez, N.; Guerrini, L.; et al. Gold Nanostars: Synthesis, Optical and SERS Analytical Properties. Anal. Sens. 2022, 2 (3), e202200005. https://doi.org/10.1002/anse.202200005. (137) Topete, A.; Varela, A.; Navarro-Real, M.; Rial, R.; et al. Revisiting Gold Nanoshells as Multifunctional Biomedical Nanotools. Coord. Chem. Rev. 2025, 523, 216250. https://doi.org/10.1016/j.ccr.2024.216250. (138) Anik, M. I.; Mahmud, N.; Al Masud, A.; Hasan, M. Gold Nanoparticles (GNPs) in Biomedical and Clinical Applications: A Review. Nano Sel. 2022, 3 (4), 792–828. https://doi.org/10.1002/nano.202100255. Chapter 1. Nano Neuro 56 (139) Herizchi, R.; Abbasi ,Elham; Milani ,Morteza; and Akbarzadeh, A. Current Methods for Synthesis of Gold Nanoparticles. Artif. Cells Nanomedicine Biotechnol. 2016, 44 (2), 596– 602. https://doi.org/10.3109/21691401.2014.971807. (140) Sánchez-Iglesias, A.; Winckelmans, N.; Altantzis, T.; Bals, S.; et al. High-Yield Seeded Growth of Monodisperse Pentatwinned Gold Nanoparticles through Thermally Induced Seed Twinning. J. Am. Chem. Soc. 2017, 139 (1), 107–110. https://doi.org/10.1021/jacs.6b12143. (141) Liu, X.; Atwater, M.; Wang, J.; Huo, Q. Extinction Coefficient of Gold Nanoparticles with Different Sizes and Different Capping Ligands. Colloids Surf. B Biointerfaces 2007, 58 (1), 3–7. https://doi.org/10.1016/j.colsurfb.2006.08.005. (142) Jumbo-Nogales, A.; Rao, A.; Olejniczak, A.; Grzelczak, M.; et al. Unveiling the Synergy of Coupled Gold Nanoparticles and J-Aggregates in Plexcitonic Systems for Enhanced Photochemical Applications. Nanomaterials 2024, 14 (1), 35. https://doi.org/10.3390/nano14010035. (143) Rogolino, A.; Claes, N.; Cizaurre, J.; Marauri, A.; et al. Metal–Polymer Heterojunction in Colloidal-Phase Plasmonic Catalysis. J. Phys. Chem. Lett. 2022, 13 (10), 2264–2272. https://doi.org/10.1021/acs.jpclett.1c04242. (144) Rao, A.; Iglesias, A. S.; Grzelczak, M. Choreographing Oscillatory Hydrodynamics with DNA-Coated Gold Nanoparticles. J. Am. Chem. Soc. 2024, 146 (27), 18236–18240. https://doi.org/10.1021/jacs.4c06868. (145) Mehtala, J. G.; Zemlyanov, D. Y.; Max, J. P.; Kadasala, N.; et al. Citrate-Stabilized Gold Nanorods. Langmuir 2014, 30 (46), 13727–13730. https://doi.org/10.1021/la5029542. (146) Zhang, Y.; Li ,Yawen; Liao ,Wei; Peng ,Wenzao; et al. Citrate-Stabilized Gold NanorodsDirected Osteogenic Differentiation of Multiple Cells. Int. J. Nanomedicine 2021, 16, 2789–2801. https://doi.org/10.2147/IJN.S299515. (147) Baffou, G.; Quidant, R. Nanoplasmonics for Chemistry. Chem. Soc. Rev. 2014, 43 (11), 3898. https://doi.org/10.1039/c3cs60364d. (148) Bayda, S.; Adeel, M.; Tuccinardi, T.; Cordani, M.; et al. The History of Nanoscience and Nanotechnology: From Chemical-Physical Applications to Nanomedicine. Molecules 2020, 25 (1), 1–15. https://doi.org/10.3390/molecules25010112. (149) Oliveira, B. B.; Ferreira, D.; Fernandes, A. R.; Baptista, P. V. Engineering Gold Nanoparticles for Molecular Diagnostics and Biosensing. WIREs Nanomedicine Nanobiotechnology 2023, 15 (1), e1836. https://doi.org/10.1002/wnan.1836. (150) Olejniczak, A.; Lawera, Z.; Zapata-Herrera, M.; Chuvilin, A.; et al. On-Demand Reversible Switching of the Emission Mode of Individual Semiconductor Quantum Emitters Using Plasmonic Metasurfaces. APL Photonics 2024, 9 (1), 016107. https://doi.org/10.1063/5.0170535. (151) Ullah, I.; Guo, J.; Wang, C.; Liu, Z.; et al. Design and Analysis of High-Efficiency Perovskite Solar Cell Using the Controllable Photonic Structure and Plasmonic Nanoparticles. J. Alloys Compd. 2023, 960, 170994. https://doi.org/10.1016/j.jallcom.2023.170994. (152) Cui, X.; Ruan, Q.; Zhuo, X.; Xia, X.; et al. Photothermal Nanomaterials: A Powerful Lightto-Heat Converter. Chem. Rev. 2023, 123 (11), 6891–6952. https://doi.org/10.1021/acs.chemrev.3c00159. Chapter 2. Self-spiking HEK Cells 63 proportion of the culture surface covered by cells. Transferring cells to a new vessel at this stage prevents overgrowth and maintains favorable conditions for continued replication. Between passages, the culture medium may be partially or fully replaced to restore nutrients and remove waste products.6 These standard practices help ensure stable cell physiology and reproducible experimental outcomes. As an immortalized line, HEK293 cells can be propagated indefinitely, though to minimize genomic drift-such as chromosomal rearrangements and gene copy number changes-cultures are commonly maintained for no more than twenty passages.3,5 Several genetically modified derivatives of the original HEK293 cell line have been developed over the years to meet specific experimental and industrial demands. A particularly innovative adaptation of the HEK293 cell line was developed by Adam E. Cohen’s group at Harvard University to create a reliable and accessible model of electrical excitability through the stable expression of defined ion channels, known as spiking HEK cells (Table 2.2). This engineered system enables high-throughput screening and detailed characterization of fluorescent voltage indicators, while also supporting a broad range of applications including ion channel research, all-optical electrophysiology, compound screening, toxicity evaluation, and studies of electroporation dynamics. In our research, it provides a central platform for evaluating voltagesensitive nanomaterials within a simplified, scalable, and well-controlled experimental context.7,8 Spiking HEK cells are genetically modified to stably express one of two key ion channels: NaV1.3 or NaV1.5 - voltage-gated sodium channels that initiate action potentials, and Kir2.1, an inward rectifier potassium channel that stabilizes the resting membrane potential near physiological values. The inclusion of Kir2.1 reduces the resting potential to approximately -66 mV, allowing the cells to exhibit repetitive action potential firing upon depolarization. In some versions of the system, Kir2.1 expression is placed under the control of a doxycycline-inducible promoter, enabling the generation of both spiking and non-spiking variants from the same line for comparative studies. Kir2.1 channel might be additionally tagged with cyan fluorescent protein (CFP) as a fluorescent tag, which enables optical assessment which cells express the channel.9 When cultured as a monolayer at high confluency (90-95%), spiking HEK cells form a functionally integrated network through intercellular coupling mediated by endogenous gap junctions, so called connexin-45. This electrical connectivity supports spontaneous, rhythmic action potentials that propagate uniformly across the culture in a coordinated, metronome-like pattern.7 In addition to their intrinsic activity, action potentials can be externally triggered using electrical stimulation. In certain variants of the cell line, the optogenetic protein CheRiff is additionally expressed, enabling precise, light-induced activation of action potentials through Chapter 2. Self-spiking HEK Cells 64 blue light stimulation. These electrophysiological properties make spiking HEK cells particularly well suited for optical electrophysiology, including the evaluation of genetically encoded voltage indicators (GEVIs) and voltage-sensitive dyes (VSDs). The system also enables high-throughput screening of pharmacological agents targeting ion channels, as well as phenotypic assays relevant to neurological and cardiac disorders. It is especially valuable for identifying modulators of excitability, assessing potential neurotoxicity or cardiotoxicity, and investigating disease mechanisms through all-optical control and monitoring of membrane voltage. While spiking HEK cells do not fully replicate the biophysical properties of neurons-due to longer action potentials and variable repolarization kinetics-they offer a robust, scalable, and experimentally accessible model for studying excitable cell behavior in a controlled setting.7–9 Table 2.2 Spiking HEK293 cell line variants with their ion channel payloads, inducible systems, sources, and key practical notes. Commercially available lines were developed in the Adam E. Cohen laboratory and are distributed by ATCC (https://www.atcc.org/). Cell line Payload Inducible system Source Practical notes Spiking HEK (Original, Not catalogued) Constitutive: •NaV1.3 •Kir2.1 No fluoro tags None (constitutive expression) A. Cohen lab; not commercially available Less stable; Kir2.1 expression lost after a few passages. tet-on Spiking HEK (ATCC CRL-3479) Constitutive: •NaV1.5 Inducible: • Kir2.1-CFP rtTA-Tet-On; induce with 1-2 µg mL⁻¹ doxycycline for 24-36 h Commercial (ATCC) HEK cells do not tolerate sustained Kir2.1-maintain a master plate without doxycycline and induce only plates destined for assays. CheRiff-eGFP tet-on Spiking HEK (ATCC CRL-3480) Constitutive: •NaV1.5 •CheRiff-eGFP Inducible: •Kir2.1-CFP rtTA-Tet-On; induce with 1-2 µg mL⁻¹ doxycycline for 24-36 h Commercial (ATCC) Adds optical stimulation capability via CheRiff. Chapter 2. Self-spiking HEK Cells 65 1.1. Cell Culture Standard culture conditions for adherent HEK293 cells involve maintaining the cells at 37 °C in a humidified atmosphere with 5% carbon dioxide (CO2).6 Cells are typically grown in Dulbecco’s Modified Eagle Medium (DMEM), usually combined with nutrient mixture, and supplemented with FBS, L-glutamine, and antibiotics.6 Antibiotics in cell culture serve two main purposes: they prevent bacterial contamination, commonly through the use of agents such as penicillin and streptomycin, and they function as selective agents to isolate or maintain genetically modified cells that carry antibiotic resistance genes. For example, compounds such as tetracycline, geneticin, puromycin, and blasticidin are routinely used to select cells that have been successfully transfected with genes of interest.4,7 Cultures are usually passaged two to three times per week. As cells grow and metabolize, they gradually acidify the surrounding medium. This shift is easily monitored by the color change of DMEM containing phenol red, which transitions from red to orange or yellow as the pH dropsoften serving as a visual cue that the culture is due for passaging or medium replacement (workflow illustrated on Figure 2.9). Passaging begins by removing the old culture medium from the flask. If the cell monolayer is well attached and stable, the cells can be gently rinsed with phosphate-buffered saline (PBS) to wash away dead cells and waste. To detach cells from the surface, a mild enzymatic treatment with trypsin is used.5 After a short incubation (typically 1-3 minutes at 37 °C), the cells begin to round up and lift from the surface. At this point, the enzyme is neutralized by adding fresh DMEM10. The suspension is pipetted gently to break up any clumps, then transferred to a sterile centrifuge tube. The cell suspension is centrifuged to pellet the cells, forming a visible puck at the bottom of the tube. The supernatant is carefully discarded, and the cells are resuspended in fresh medium, again pipetting up and down to ensure a uniform, single-cell suspension. Controlling the seeding density is important for experiments like spiking monolayer cultures for imaging, and this requires a precise cell count using a hemocytometer. Optionally, trypan blue dye can be added to discriminate between viable and non-viable cells, as the dye penetrates only the membranes of dead cells. Based on the calculated cell concentration, the appropriate number of cells is seeded into new culture flasks and experimental sample plates, using fresh DMEM10 supplemented with antibiotics to maintain selectiveness and ensure consistent gene expression where relevant. In some cases, culture vessels or sample dishes may be coated with poly-D-lysine-a synthetic, positively charged polypeptide that enhances cell adhesion and supports growth on the culture surface.3,7 Once seeded, all vessels are clearly labeled and placed Chapter 2. Self-spiking HEK Cells 66 back into the incubator to allow the cells to reattach and continue growing under standard conditions. A detailed protocol, including the complete media recipe, exact reagent volumes, and the step-by-step cell counting procedure, is provided in the Appendix (Section 2.2.1 Cell Culture Protocols). Figure 2.9 Schematic overview of the HEK293 cell passaging workflow. Steps include reagent preparation, cell detachment with trypsin, centrifugation and resuspension in fresh medium, cell counting with trypan blue exclusion, and reseeding into new culture vessels for continued incubation. Con uency 80 Trypsin Discard media and wash cells with PBS Dead cell Viable cell DMEM Incuba on 37°C 5 CO2 Discard supernatant Redisperse in DMEM 1:1 Trypan blue Cell solu on 37°C 15 min Hemocytometer Chapter 2. Self-spiking HEK Cells 67 2. Calcium Imaging of Spiking HEK Cells Fluorescence serves as a fundamental indicator of neural structure and dynamics, forming the basis for numerous imaging techniques employed by neurobiologists (see Chapter 1.1.2). By leveraging fluorescent signals, it is possible to achieve high-resolution imaging of neurons and their functional activities, thereby enabling exploration of cellular mechanisms underlying neural activity.10 The choice of fluorescence microscope plays a pivotal role in determining the spatial and temporal resolution achievable, and neurobiological studies frequently employ wide-field, confocal, two-photon, or multiphoton microscopes, each suited to particular experimental requirements and imaging depths. While confocal microscopy offers excellent optical sectioning for moderate-depth samples, two-photon microscopy has become a mainstay in imaging neural activity deep within scattering tissues, such as intact brain preparations.11,12 With these considerations in mind, this chapter describes the use of fluorescence microscopy to study calcium dynamics in self-spiking HEK cells. The imaging setup employs standard epi-fluorescence illumination and camera-based detection, which allows monitoring of intracellular calcium changes in response to electrical activity. This straightforward configuration was selected to support the development of an approach that can be easily adopted in typical biological laboratory environments without requiring specialized instrumentation. 2.1. Fluorescence Microscope Setup Fluorescence microscopy visualizes biological samples by detecting emitted fluorescence light from molecules called fluorophores. Fluorophores absorb excitation light at a shorter wavelength and emit fluorescence at a longer wavelength; this shift, known as the Stokes shift, allows clear separation of emitted fluorescence from excitation illumination.13 A typical fluorescence microscope employs an epi-fluorescence setup, where the same objective lens serves both to deliver excitation light and to collect emitted fluorescence (Figure 2.10 A).14 Objectives with a high numerical aperture (NA) are preferred, as they increase light-gathering efficiency and improve spatial resolution. Immersion objectives-used with a medium such as water, oil, or glycerol between the lens and the specimen-can further enhance image quality by minimizing refractive index mismatches and reducing optical distortions.13 Central element of such a fluorescent setup is the filtering cube, containing three essential components: an excitation filter that transmits only specific wavelengths required to excite the fluorophores, a dichroic mirror placed at a 45-degree angle that reflects excitation wavelengths onto the sample and transmits emitted fluorescence towards the detector, and an emission (barrier) filter that selectively allows only the longer-wavelength fluorescence signal to reach Chapter 2. Self-spiking HEK Cells 68 the detector, blocking residual excitation light.11,13 These filters must be carefully matched to the spectral properties of the fluorophore used. Illumination is typically provided by a mercury lamp, which emits intense, broad-spectrum light with strong peaks strong peaks in the UV and visible spectrum. To regulate illumination intensity and minimize photobleaching or phototoxicity, neutral density (ND) filters can be introduced into the optical path to reduce the amount of excitation light reaching the sample.11 Fluorescence signals are typically recorded using sensitive digital devices such as CCDs (ChargeCoupled Device), EMCCDs (Electron-Multiplying CCD), or sCMOS (Scientific Complementary Metal-Oxide-Semiconductor) cameras.10,11,13,14 Camera selection should be guided by experimental needs, weighing factors such as signal sensitivity, temporal and spatial resolution, as well as cost and accessibility. Prior to data collection, key recording parameters such as camera exposure time, frame rate (frames per second, fps), pixel binning, and the size of the recorded area must be carefully considered. Shorter exposures enable higher frame rates and improved temporal resolution but may require increased excitation intensity or higher detector sensitivity to preserve signal quality. Binning, which combines adjacent pixels into larger units, enhances the signal-to-noise ratio at the cost of spatial resolution, and can support faster acquisition by reducing data volume. Similarly, reducing the recorded field of view (FOV) allows faster readout and higher frame rates. These parameters must be tuned to achieve the appropriate balance between temporal and spatial resolution, signal quality, data size, and sample viability, based on the specific goals and constraints of the experiment.13,14 Chapter 2. Self-spiking HEK Cells 69 Figure 2.10 Epi-fluorescence microscopy setup used for calcium imaging experiments. (A) Optical diagram of the fluorescence light path, showing key components including the mercury arc lamp, excitation and emission filters, dichroic mirror, immersive objective, and detector. The system is equipped with a motorized filter cube turret, allowing rapid switching between multiple filter cubes during experiments. (B) Photograph of the fluorescence microscope system located at the Neurotechnology Center, Columbia University. The microscope used for the majority of measurements presented in this dissertation was an upright fixed stage microscope system located in the Yuste Laboratory at the Neurotechnology Center, Columbia University. This setup, internally referred to as the “North Rig”, was configured for single-channel calcium imaging using Calcium Green-1, with FITC filter cube matched to the dye’s excitation and emission characteristics (Figure 2.10 B). Excitation light from a mercury lamp was spectrally filtered through a 475/85 nm band-pass excitation filter and directed onto the specimen via a 500 nm long-pass dichroic mirror and a water-immersion objective. The same objective lens collected the emitted fluorescence, which then passed through a 535/50 (passing approximately 510-560 nm wavelengths) emission filter before being detected by a sCMOS Hamamatsu camera mounted on the microscope’s trinocular port.11,13 An immersion objective was used to minimize refractive index mismatch between the objective lens, the coverslip, and the aqueous medium covering the sample, thereby improving light collection efficiency and preserving spatial resolution. Acquisition parameters-including exposure time, pixel binning, and frame rate-were optimized under continuous illumination to capture rapid calcium transients with minimal signal loss. Recordings were performed at 50, 100, and 200 ms exposure times; an exposure of 100 ms (corresponding to 10 frames per second) provided a suitable Chapter 2. Self-spiking HEK Cells 70 balance between temporal resolution and fluorescence signal brightness for detecting spiking events. In addition to fluorescence illumination, a secondary white light source was positioned beneath the specimen, enabling the acquisition of bright-field images aligned with the corresponding fluorescence fields of view. The microscope is equipped with a filter cube turret, allowing rapid switching between filter sets during experiments to support imaging of different fluorophores. 2.2. Results and Discussion The genetically modified HEK cells, described by Park and Cohen (2013), spontaneously generate electrical spikes when cultured as a monolayer at high confluence (80-95%). This activity emerges as coordinated depolarization events, with reported spike frequencies around 3 Hz and propagation velocities of approximately 2 cm/s. These spikes are characterized by a resting membrane potential near -66 mV, peak depolarizations reaching +34 mV, and rise times of approximately 3 ms, reflecting stable and repeatable action potential dynamics across the culture.7 For imaging, monolayers were stained with the chemical calcium indicator Calcium Green-1 and recorded in Tyrode’s solution, a physiological extracellular buffer with defined ionic composition; detailed protocols for both are provided in the Appendix (Section 2, Supplementary Material for Chapter #2). To determine the optimal time window for reliable recordings, we monitored cultures at days 3, 4, and 5 post-seeding (Figure 2.11). Examples shown for each day illustrate representative recordings from qualitative analysis of three culture plates per time point, with several regions examined in each dish. On day 3, the monolayer was often incomplete, and spontaneous spiking was not detected. Even in more confluent regions, cells appeared insufficiently mature to support the ion channel expression necessary for robust spiking. In these areas, we observed slow calcium transients, but not the fast, periodic signals characteristic of electrical spikes. By day 4, the cultures consistently formed dense and uniform monolayers, and spiking activity became clear, rhythmic, and reproducible across samples. Estimating dominant frequency of the calcium signals in presented sample yielded peak value of 0.26 Hz, as determined using the Welch method. On day 5, signs of overgrowth were evident, with cells growing in multiple layers. Under these conditions, spiking activity was reduced or absent, and when present, it was often unreliable and not synchronized. The corresponding Welch spectrum of the day 5 trace identify a dominant frequency at 0.26 Hz, but the power spectral density of the peak was approximately Chapter 2. Self-spiking HEK Cells 71 threefold lower compared to day 4. These observations define a narrow experimental window, with day 4 representing the most favorable conditions for capturing spontaneous, synchronized spiking events in this system. We next examined whether the observed activity reflected coordinated behavior across the monolayer rather than independent cellular responses (Figure 2.12). Fluorescence was recorded from a 166.4 × 166.4 µm field of view at 10 frames per second, using 100 ms exposures. With a reported wave propagation speed of approximately 2 cm/s, a spike would traverse the field in about 10 ms, well below the temporal resolution of our acquisition. Under these conditions, propagating spikes are expected to appear as simultaneous events across the field, enabling synchronized activity to be captured as a uniform signal. The averaged fluorescence trace of whole FOV (Figure 2.12 B) shows regular calcium transients with consistent shape and timing. A heatmap of 500 individual cell traces (Figure 2.12C) reveals a uniform pattern: nearly all cells within the field exhibit aligned spiking, no significant variation in timing. This synchrony is further illustrated in selected single-cell traces (Figure 2.12 D), which follow the same rhythm and phase. These results confirm that the culture behaves as a synchronized system, with cells spiking collectively at regular intervals. Chapter 2. Self-spiking HEK Cells 72 Figure 2.11 Spontaneous spiking activity in Calcium Green-1-labeled cells monolayers at different days post-seeding. Top: Fluorescence images from days 3, 4, and 5. Bottom: Average fluorescence traces over 60 seconds for whole field of view (full frame average). Day 3 monolayers are incomplete and show no spiking. On day 4, uniform monolayers start to exhibit periodic spiking. By day 5, cultures are overgrown and activity is diminished or absent. On the right, power spectral density plots from Welch analysis illustrate the frequency components corresponding to each trace. Day 3 Day 4 Day 5 50 m50 m50 m Chapter 2. Self-spiking HEK Cells 79 (16) Même, W.; Ezan, P.; Venance, L.; Glowinski, J.; et al. ATP-Induced Inhibition of Gap Junctional Communication Is Enhanced by Interleukin-1 β Treatment in Cultured Astrocytes. Neuroscience 2004, 126 (1), 95–104. https://doi.org/10.1016/j.neuroscience.2004.03.031. Chapter 2. Self-spiking HEK Cells 80 81 Chapter 3 Quantum Dots for Detecting Neuronal Activity Chapter 3. Quantum Dots for Detecting Neuronal Activity 82 1. Spherical Quantum Wells Spherical quantum wells (SQWs) represent an advanced category of colloidal semiconductor nanocrystals that feature a multilayered architecture, commonly referred to as core/well/shell heterostructures. These structures consist of a central semiconductor core surrounded by two distinct shell layers, each contributing to the overall quantum confinement and optical behavior of the material.1,2 One of the most extensively investigated configurations is the CdS/CdSe/CdS structure, composed of a CdS core (2.7 eV bulk bandgap), a middle CdSe layer functioning as the quantum well (1.75 eV bulk bandgap), and an outer CdS shell.3,4 In this architecture, the intermediate CdSe layer acts as a three-dimensionally confined region for hole wavefunctions, while electron wavefunctions are delocalized over the entire volume of the nanocrystal, which leads to a quasi-type-II band alignment. This unique spatial distribution of charge carriers is central to the distinctive optical and electronic properties of SQWs, distinguishing them from both 2D planar quantum wells and 0D systems such as traditional quantum dots (QDs) or conventional core/shell quantum dots (CSQDs). Strong confinement within the spherical CdSe layer contributes to high photochemical stability and quantum yields approaching unity.1 Moreover, the strong confinement leads to wellseparated electronic states and altered recombination dynamics, including suppressed Auger recombination and extended carrier lifetimes.5–7 In particular, thick-shell CdS/CdSe/CdS SQWs have achieved near-unity photoluminescence quantum yield (PLQY) at room temperature (Figure 3.16 A). This high efficiency results from reduced lattice mismatch between the CdSe well and CdS layers, where coherent strain suppresses misfit defect formation. These structures also exhibit strongly reduced blinking, spending up to 90 of the time in the emissive “on” state.1,3,8 Another noteworthy property of SQWs is their enhanced multiexciton emission, especially in large-core structures. SQWs with larger core diameters have shown high biexciton quantum yields, sometimes approaching 100%. This efficiency arises from reduced Auger recombination rates and increased effective exciton volumes (Figure 3.16 B). Furthermore, the well-defined spherical quantum confinement allows higher biexciton binding energies compared to purely spherical core nanocrystals.4,7 Although the CdS/CdSe/CdS system is extensively studied – benefiting from the well-established understanding of CdSe and CdS quantum dots and core/shell structures – other material combinations, including ZnS/CdS/ZnS,9 InP/ZnSe/ZnS (cadmium free alternative)10, and CdSe/CdS/ZnS11 have also been investigated. These alternative architectures aim to tailor exciton dynamics, carrier confinement, and band alignments to meet the demands of specific technological applications. Chapter 3. Quantum Dots for Detecting Neuronal Activity 83 Figure 3.16 Optical performance enhancement in semiconductor nanocrystal heterostructures through spherical-well geometry and interfacial alloying. (A) Comparison of photoluminescence quantum yield (PLQY) as a function of CdS shell thickness for CdS/CdSe/CdS spherical quantum well (SQW) nanocrystals and conventional CdSe/CdS core/shell structures. SQW nanocrystals achieve near-unity PLQY and suppressed blinking at shell thicknesses greater than 5 nm, attributed to coherent strain in the CdSe layer, defect-free outer shell growth, and reduced hole trapping at the surface. (B) Photoluminescence decay dynamics for conventional CdSe/CdS core/shell quantum dots with an abrupt interface and core/alloy/shell structures incorporating a CdSeₓS₁₋ₓ alloy interlayer. The presence of the alloyed interface slows multiexciton decay, indicating suppressed Auger recombination due to a smoother hole confinement potential. Panel A reprinted with permission from B. G. Jeong, Y.-S. Park, J. H. Chang, et al., ACS Nano, 2016, 10, 10, 9297–9305. Copyright © 2016 American Chemical Society. Panel B reprinted with permission from W. K. Bae, L. A. Padilha, Y.-S. Park, et al., ACS Nano, 2013, 7, 4, 3411–3419. Copyright © 2013 American Chemical Society. Chapter 3. Quantum Dots for Detecting Neuronal Activity 84 Due to these superior photophysical properties, SQWs have been identified as highly promising candidates for a wide range of advanced optoelectronic applications. In quantum light sources, their suppressed blinking and near-unity photoluminescence quantum yield (PLQY) make them excellent single-photon emitters for quantum cryptography, with antibunching sustained even at telecom wavelengths when integrated into photonic cavities.3,8,12 In light-emitting devices, SQWs have improved microcavity laser performance and enabled flexible displays with broad color gamut coverage and extended lifetimes.4,7 They also show promise in radiation detection, with fast X-ray response times that outperform traditional scintillators, opening the possibility for new generation of radiation detection.8,13 Among these diverse applications, bioimaging stands out as a particularly promising area. Quantum dots, including SQWs, are actively researched as fluorescent probes for biosensing applications. Due to their brightness, photostability, and the ability to be functionalized for targeting specific cellular structures they are widely used for bio-labeling and cellular imaging.14 Their bright and tunable emission can be tailored as preferred for particular application. If needed, it can be engineered into the near-infrared region, facilitating deep-tissue imaging with minimal photodamage and reduced background autofluorescence. Additionally, their enhanced two-photon absorption cross-sections enable high-resolution imaging using near-infrared excitation, which is advantageous for in vivo applications.4,7,15 In summary, SQWs constitute a major advancement in colloidal nanocrystal technology, effectively combining beneficial aspects from planar quantum wells and classical nanocrystals. By strategically managing exciton dynamics and band alignment through material and structural optimization, SQWs significantly outperform conventional semiconductor nanostructures (Table 3.3), providing new opportunities across multiple scientific and industrial domains. Chapter 3. Quantum Dots for Detecting Neuronal Activity 85 Table 3.3 Comparison of conventional quantum dots and spherical quantum well (SQW) nanocrystals across structural, optical, and functional properties. Aspect Quantum Dots ingle component or thin core/shell) Spherical Quantum Wells (core/well/shell) Reference Dimensional confinement Confinement in all three dimensions; band-gap rises and the DOS becomes discrete as the diameter decreases Confined radially, with free carrier motion around the well; results in quasi-2D shell states governed mainly by well thickness, not particle diameter 6 Battaglia et al. Angew. Chem. 2003 Energy-level structure, exciton wavefunctions 1S-1P energy spacing set by core radius; electron & hole wave-functions overlap strongly (type-I) unless thick graded shells are added Quasi-type-II: hole in CdSe, electron in CdS, weak overlap; gap set by CdSe thickness – each added monolayer red-shifts, size largely irrelevant 5 Peng et al. APL 2005 6 Battaglia & Peng Angew. Chem. 2003 Optical properties (absorption& PL) PL FWHM 30-40 nm (broader due to size distribution and phonon interactions) PL FWHM narrower, often ~20 nm, color tunning by altering well thickness 1 Jeong et al. ACS Nano 2016 Absorption onset is broader, gradual rise, caused mainly by size dispersion and variations in electron structure Absorption onset is steep and well-defined, from consistent well thickness and quantized 2Dlike energy levels Stokes shift moderate (~20 meV), varies by structure. PLQY 50-70 after passivation Stokes shift larger (30-60 meV) due to quasi-type-II alignment, PLQY ~ 100% Carrier dynamics (radiative & nonradiative lifetimes) Radiative lifetimes 5-20 ns; Auger recombination of multiexcitons is fast (0.1 ns); blinking and gain loss Exciton lifetime similar to QDs, but multiexciton lifetime extended to >1 ns (“giant” suppression); biexciton PLQY 80100% 4 Nagamine et al. ACS Photonics 2020 ; 7 Marder et al. ACS Mater. Lett. 2023 Response to external electric fields Nearly co-located carriers, the Stark response is largely quadratic and red-shifts stay modest (0-2 meV/10 kV cm-1) Build in radial dipole makes thinwell SQWs respond linearly (3-6 meV/10 kV cm-1 for 1-2 ML wells), but thick CdS barrier screens modest external fields. 1 Jeong et al. ACS Nano 2016 High-order nonlinear optics TPA cross-section σ₂ 103-104 GM; enhanced in anisotropic structures SQWs show σ₂ up to 1.8 × 102 GM/nm3 and low ASE thresholds (~10 µJ/cm2); biexciton binding tunable (-130 to -50 meV) 8 Xiang et al. Nanomaterials 2024 Applications demonstrated QD-LEDs, photodetectors, LSCs, single-photon sources. Low-threshold solutionprocessed lasers, optical amplifiers, single-photon emitters, down-converters for SSL; quantum-light sources 16 García-de-Arquer et al. Science 2021; 3 Allemand et al. Nanotechnology 2022 Key challenges Auger losses & blinking, Surface-trap passivation Precise control of well thickness at ML level, Scale-up 17 Cragg et al. Nano Lett. 2010; 18 Giansante et al. JPCL 2017 Chapter 3. Quantum Dots for Detecting Neuronal Activity 86 The study introduced in this PhD dissertation focuses on the CdS/CdSe/CdS configuration as the primary structural model, based on both theoretical predictions and experimental findings that highlight its high optical efficiency and strong sensitivity to external electric fields.5,8,19 These attributes position SQWs as promising candidates for sensing neuronal activity, where changes in membrane voltage could modulate their optical emission.20 Such voltage sensitivity offers a potential pathway for non-invasive neural imaging with high spatial and temporal resolution.21 While experimental validation is still underway, emerging studies suggest that SQWs could play a pivotal role in the development of advanced neuroimaging tools and brain-computer interface technologies. 1.1. Overview of Quantum-dot-based Platforms Before diving into the experimental protocols and performance analysis, it is important to provide a clear overview of the quantum dot (QD) systems explored in this study. These nanocrystal architectures form the basis for all subsequent investigations presented in this chapter. Although the rationale behind the selection and development of each system will be discussed in detail in later sections, introducing the full set of materials here will help orient the reader and provide a consistent reference point throughout the text. Figure 3.17 summarizes the six QDs-based systems, identified as Systems I-VI, that were included in the course of study presented in this dissertation. These systems span a range of core-shell geometries and surface chemistries, allowing for a systematic examination of how structural variations influence both optical performance and biological interactions. System I is a commercially available CdSe/ZnS core-shell nanocrystal capped with tri-n-octylphosphine oxide (TOPO), serving as a widely used reference standard. Systems II, III, and IV share a CdS/CdSe/CdS spherical quantum well (SQW) configuration but differ in their surface coordination strategies: cadmium oleate (system II), zinc oleate (system III), and a combination of cadmium fluoride and octylamine (system IV). System V builds on the structure of III by incorporating an additional ZnS outer shell, enabling evaluation of the effects of extended passivation. System VI, on the other hand, modifies the ligand environment by replacing linear oleate with a branched 2-hexyldecanoate, providing insights into how ligand geometry impacts colloidal behavior and biological interactions. Chapter 3. Quantum Dots for Detecting Neuronal Activity 87 Figure 3.17 Overview of quantum dot systems investigated throughout this study. Six distinct nanocrystal architectures were evaluated, differing in core-shell structure and surface ligand composition. System I represents a commercially available CdSe/ZnS core-shell quantum dot capped with tri-n-octylphosphine oxide (TOPO), serving as a reference material. Systems II-VI are custom-synthesized CdS/CdSe/CdS or CdS/CdSe/CdS/ZnS spherical quantum wells (SQWs), each modified with a distinct surface coordination environment: cadmium oleate (II), zinc oleate (III and V), cadmium fluoride and octylamine (IV), and a branched 2-hexyldecanoate ligand (VI). This materials library enables a systematic investigation into how nanocrystal architecture and surface chemistry influence optical behavior, colloidal stability, and membrane fusion efficiency. All systems were designed and prepared with reproducible protocols, ensuring consistent core size and quantum well thickness. This uniformity enables the evaluation of the influence of surface chemistry and shell composition on the physical of quantum dots. The sections that follow will first detail the synthetic strategies used for each custom-made nanocrystal system (3.1.3 Surface Chemistry on CdS/CdSe/CdS Nanocrystals), followed by investigations of water transfer (3.2 Strategies for Water Transfer), and their interaction with biological material (3.3 Cell Membrane Insertion). From this point forward, each system will be referred to by its assigned Roman numeral, with Figure 3.17 serving as a visual reference throughout the chapter. Chapter 3. Quantum Dots for Detecting Neuronal Activity 88 1.2. Synthetic Approaches Synthesis of layered CdS/CdSe/CdS nanocrystals CdS/CdSe/CdS nanocrystals with layered radial composition were synthesized under an inert argon atmosphere using a combination of single and sequential injection strategies at 240 °C. The method relied on the simultaneous injection of sulfide and selenide precursors into a cadmium oleate and oleic acid solution to initiate the growth of the CdS/CdSe core/shell structure, followed by the sequential growth of an outer CdS shell. The chalcogen precursor mixture consisted of thiourea22 and selenourea23 derivatives, which served as sulfur and selenium sources and enabled precise control of nucleation and growth. As reported by the group of Jonathan Owen, these precursors release metal-chalcogen monomers, which serve as building blocks for nanocrystals, at tunable rates, allowing control over the number of nuclei formed and thus the final size of the nanocrystals. Unlike traditional methods that stop the reaction early to control size, often at the expense of yield, this approach allows the reaction to proceed to full completion, producing nanocrystals with uniform size, well-defined composition, and high efficiency. Furthermore, by adjusting the reactivity of the precursors, the method can be tailored to work with different metal sources, enabling the synthesis of a wide range of metal sulfide and selenide nanocrystals. Large-scale synthesis of air-stable N,N′-disubstituted and N,N,N′-trisubstituted thioureas can be efficiently achieved in a single-step reaction, in most cases performed in a room temperature. This process involves the direct coupling of readily available substituted isothiocyanates with primary or secondary amines, resulting in high reaction efficiency.22 Detailed protocols can be found in the Appendix (Section 3.1.1). Small CdS/CdSe/CdS quantum wells were synthesized using a sequential injection strategy under an inert argon atmosphere, inside a glovebox. For the initial CdS/CdSe formation, a solution of cadmium oleate (0.244 g, 0.36 mmol), oleic acid (0.204 g, 0.72 mmol), and hexadecane (28.46 mL, 22 g) was loaded into a 100 mL three-neck round-bottom flask equipped with a magnetic stir bar, connected to the Schlenk line under Argon and heated to 240 °C (Figure 3.18, panel A). In parallel, a chalcogen precursor mixture was prepared by dissolving N-hexyl-N′- dodecyl thiourea (0.0496 g, 0.15 mmol) and 1,3-diethy imidazolidine selone (0.0308 g, 0.15 mmol) in tetraglyme (0.8 g, 0.848 mL). The chalcogen precursor solution was swiftly injected into the hot cadmium mixture, and the reaction was maintained at 240 °C for 60 minutes (Figure 3.18, panel B). The reaction of this fast-converting thiourea (conversion rate constant kconv = 0.03 s-1)22 and slower-converting selenourea (kconv = 0.001 s-1)23 with cadmium Chapter 3. Quantum Dots for Detecting Neuronal Activity 95 Ligand exchange for Zn(2-hexyldecanoate)2 Ligand exchange using Zn(2-hexyldecanoate)2 was performed following the same procedure as described for Zn(oleate)2, with the only modification being the substitution of zinc oleate with an equimolar amount of Zn(2-hexyldecanoate)2. All other reaction conditions and purification steps remained unchanged. Ligand exchange – fluorination Cd-oleate-stabilized SQWs (0.100 µmol) were dried under vacuum and redispersed in octadecene (15 mL). Octylamine (0.0259 g, 0.0331 mL, 0.2 mmol) and benzoyl fluoride (0.0248 g, 0.0205 mL, 0.2 mmol) were added to the dispersion, and the reaction mixture was stirred at room temperature for 1 hour. The resulting nanocrystals were purified by precipitation with acetone and centrifugation. Shelling with additional ZnS layer To obtain CdS/CdSe/CdS/ZnS multilayer structure, Cd-oleate-stabilized CdS/CdSe/CdS SQWs were prepared as previously described, with the CdS shelling step modified to use half the reagent quantities and half the reaction time. Following CdS shell growth, a ZnS shelling solution was prepared by dissolving zinc oleate (0.3393 g, 0.357 mL, 0.54 mmol), 2-hexyldecanoic acid (0.1385 g, 0.1556 mL, 0.54 mmol), and N-hexyl-N′,N′-dibutyl thiourea (0.1472 g, 0.54 mmol) in octadecene (14.2 g, 18.0 mL). This mixture was injected into the reaction via syringe pump over 2 hours at a rate of 9 mL/h. The resulting nanocrystals were purified by four sequential precipitations using a 5:1 v/v methyl acetate to toluene mixture. Chapter 3. Quantum Dots for Detecting Neuronal Activity 96 2. Strategies for Water Transfer In order for spherical quantum wells to be applicable in biological systems, their transition from organic solvents to water or water-based buffers is critical. After the synthesis, nanocrystals are typically stabilized with organic molecules (e.g., trioctylphosphine oxide, oleic acid), which makes them soluble in non-polar organic solvents such as hexane, toluene or chloroform and ensure colloidal stability. This hydrophobic surface chemistry also drives the initial association with the cell membrane: by the classic hydrophobic effect, alkyl chains on the QD surface minimize unfavorable water contacts by partitioning into the lipid bilayer’s hydrophobic core.24 Bar et al. demonstrate on simplified supported lipid bilayers that negatively charged QDs first dock onto the membrane via electrostatic attraction and then penetrate at high‐curvature or defect sites – where hydrophobic ligand chains insert into the lipid tails – supporting a two‐step, defect-mediated mechanism for membrane embedding.24 Lipophilic-coated quantum dots inherently partition into the hydrophobic core of the membrane rather than staying in the surrounding aqueous phase. Yet, to enable efficient cellular delivery – and, ultimately, reliable voltage sensing – these nanoparticles must first feature colloidal stability in physiological media. To transfer QDs into aqueous solution, several strategies have been reported in the literature. Ligand exchange replaces native hydrophobic ligands with hydrophilic ones (such as thiols or multidentate ligands).25,26 This method is relatively straightforward but can disturb the QD surface and often results in significant loss of photoluminescence quantum yield (QY) and reduced colloidal stability. In contrast, encapsulation techniques preserve the original ligand shell by wrapping the hydrophobic QD with an amphiphilic coating. These include polymer or surfactant micelle encapsulation and coating with biomolecules like lipids or proteins. Encapsulation is generally more complex but tends to better retain the QDs’ optical properties. For example, amphiphilic polymer coatings maintain the native hydrocarbon ligand environment, yielding water-dispersible QDs with minimal changes in absorption or emission spectra.27,28 Similarly, protein-based coatings (e.g., embedding QDs in bovine serum albumin or polymer-protein hybrid shell) can confer water solubility while keeping the original ligand in place, preserving fluorescence and providing biological interfaces.29–31 Each approach differs in complexity and effectiveness. Ligand exchange produces smaller hydrodynamic diameters but risks diminished photostability, whereas polymer or lipid encapsulation yields larger bioconjugates with excellent photophysical retention. Chapter 3. Quantum Dots for Detecting Neuronal Activity 97 Among methods that maintain the native hydrophobic coating on QDs, two prominent strategies are encapsulation in octyl glucoside (OCG) micelles and lipid vesicles (liposomes). These approaches avoid direct chemical modification of the QD surface, thereby aiming to preserve the original ligand passivation and the QD’s photophysical properties. 2.1. OCG Micelles n-Octyl-β-D-glucoside (OCG) is a small non-ionic detergent with a hydrophobic octyl tail and a hydrophilic glucose head (Figure 3.7 A). It is widely used in membrane biochemistry due to its ability to gently solubilize lipid bilayers while maintaining the structural and functional integrity of membrane-associated proteins. OCG is particularly effective for isolating functional membrane proteins, as it disrupts the membrane environment without denaturation and can be readily removed through dialysis or dilution.32 Additionally, OCG is employed to induce controlled transitions between lipid vesicles and micelles by modulating its concentration near the critical micelle concentration.33,34 This property is valuable for studying membrane curvature, lipid packing, and the energetics of bilayer assembly.32 In supported lipid bilayer systems, OCG facilitates the precise solubilization and reformation of membranes on solid substrates, which is essential for techniques such as atomic force microscopy and other surfacebased analyses. Although the encapsulation of quantum dots within OG micelles has not been previously reported, a related study by Rasch et al. demonstrated the successful aqueous phase transfer of small gold nanoparticles using OCG molecules, resulting in the formation of nanoparticle-vesicle hybrid structures.35 Since previous studies have shown that TOPO-capped CdSe quantum dots can be effectively transferred to aqueous media through encapsulation with surfactants forming micellar structures,36 and given the widespread use of OCG in membrane biochemistry, we aimed to investigate its potential for membrane staining. OCG’s ability to solubilize hydrophobic environments while preserving molecular integrity, along with its established role in modulating vesicle-micelle transitions, makes it a strong candidate for encapsulating hydrophobic quantum dots without altering their native surface chemistry. In this study, we evaluate the use of OCG micelles for aqueous phase transfer of QDs and assess their utility for trans-membrane delivery in biological systems. Chapter 3. Quantum Dots for Detecting Neuronal Activity 98 Encapsulation protocol Data shown in this subchapter, for SQWs encapsulated within OCG micelles, are based on the sample described in a previous section (Figure 3.19). Prior to conducting any experiments, it was essential to determine the concentration of spherical quantum dots in solution. This was accomplished using the Beer-Lambert law, expressed as A = εcl, where A is the absorbance, ε is the molar extinction coefficient, c is the concentration, and l is the optical path length. In our calculations, we utilized the absorbance value at a wavelength of 350 nm and assumed an extinction coefficient of 1.7, which aligns with reported values for CdSe quantum dots of comparable size and optical properties.37 Figure 3.22 Encapsulation of CdS/CdSe/CdS spherical quantum wells (SQWs) using the amphiphilic surfactant β-D-octyl glucoside (OCG). (A) Chemical structure and schematic representation of OCG, highlighting its amphiphilic nature with a hydrophilic glucose head (blue) and a lipophilic alkyl tail (yellow). (B) Schematic illustration of the SQW encapsulation process. Hydrophobic SQWs are phase-transferred from dichloromethane (DCM) to water via OCG-assisted encapsulation, forming stable SQWs@OCG micelles upon DCM evaporation at 50 °C. To optimize the protocol, several key parameters were systematically investigated and adjusted. These included the choice of organic solvent, the water-to-solvent ratio, the OCG-to-SQW ratio, stirring speed and temperature. Multiple phase transfer strategies were evaluated, with the scheme presented above in Figure 3.22 representing the most reproducible method identified during the study. The following section focuses on the influence of OCG concentration on the quality of the micelles obtained. In a 4 mL glass vial equipped with a magnetic stirrer, 500 μL of OCG solution in dichloromethane (DCM) at the desired concentration was mixed with 100 μL of SQWs redispersed in DCM (0.06 μM). The mixture was stirred at room temperature for at least 5 minutes. Subsequently, 250 μL of water was added under vigorous stirring, and the system was stirred for an additional 5 minutes at room temperature. The solution was then heated to 50 °C until the DCM had fully Chapter 3. Quantum Dots for Detecting Neuronal Activity 99 evaporated. Finally, an additional 250 μL of water was introduced, and the resulting dispersion was subjected to further optical characterization. Characterization The concentration of β-D-octyl glucoside (OCG) used during encapsulation had a pronounced effect on the physical and optical quality of the resulting SQW formulations. Above its critical micelle concentration (CMC, ~20-25 mM), OCG molecules spontaneously assemble into micelles in aqueous media. When hydrophobic SQWs are introduced into an aqueous OCG solution, the alkyl chains of OCG interact with the hydrophobic ligand shell of the SQWs via the hydrophobic effect, while the sugar headgroups orient toward the surrounding water, resulting in the formation of a micellar colloid. The encapsulation process can be carried out under mild conditions, involving moderate stirring and gentle heating, which makes it straightforward to implement while minimizing the risk of altering the nanocrystal surface chemistry. As shown in Figure 3.23, samples prepared with 25 mM (sample 1), 50 mM (sample 2), and 100 mM (sample 3) OCG display progressively greater turbidity and light scattering, particularly at the highest concentration. Dynamic light scattering (DLS) measurements support this observation: sample 1 exhibits small micelles with a hydrodynamic diameter of 35.2 ± 7.5 nm and a low polydispersity index, indicating uniform size. In contrast, higher OCG concentrations yield much larger and more polydisperse aggregates, with hydrodynamic diameters of 260.8 ± 126.0 nm and 347.2 ± 146.4 nm for the samples 2 and 3, respectively (Figure 3.23 B). Photophysical characterization confirmed that micellization preserves the optical properties of the SQWs. The corresponding UV-Vis spectra of sample 1 shows pronounced scattering, while lower-concentration samples show minimal deviation from baseline, suggesting the presence of well-dispersed micelles or particles smaller than 100 nm (Figure 3.23 C). Emission spectra reveal a concentration-dependent blue shift in photoluminescence maximum from 529 nm (sample 1, 25 mM) to 526 nm (sample 2, 50 mM) and 523 nm (sample 3, 100 mM) (Figure 3.23 D). All micellar samples are therefore slightly red shifted relative to the same SQWs dispersed in toluene (520 nm, Figure 3.19). Chapter 3. Quantum Dots for Detecting Neuronal Activity 100 Figure 3.23 Visual appearance and optical characterization of spherical quantum wells (SQWs) encapsulated with increasing concentrations of β-D-octyl glucoside (OCG). (A) Photographs of samples containing 25 mM, 50 mM, and 100 mM OCG under ambient white light (top) and UV illumination (bottom). Samples with 25 mM and 50 mM OCG (samples 1 and 2) are optically clear with a slight yellow tint, while the 100 mM sample (sample 3) appears increasingly turbid due to light scattering. (B) Dynamic light scattering (DLS) analysis shows hydrodynamic diameters of 35.2 ± 7.5 nm, 260.8 ± 126.0 nm, and 347.2 ± 146.4 nm for the 25 mM, 50 mM, and 100 mM samples, respectively. (C) UV-Vis extinction spectra exhibit a concentration-dependent increase in baseline intensity, consistent with enhanced scattering in the 100 mM sample. (D) Emission spectra of the same samples display a blue shift in the photoluminescence maximum with increasing OCG concentration. Despite the differences in the absorption spectra of sample 1 and sample 3, their photoluminescence lifetimes are notably similar (Figure 3.24). Both samples exhibit a bi-exponential decay profile, with components of 44.0 ± 1.0 ns and 6.0 ± 0.22 ns for sample 1, and 43.0 ± 1.1 ns and 9.1 ± 0.31 ns for sample 3. These correspond to average fluorescence lifetimes of 18.09 ns and 23.69 ns, respectively. Compared to SQWs dispersed in toluene (τ1 = 44.0 ± 2.1 ns, τ2 = 13.0 ± 0.79 ns), both encapsulated samples exhibit a shortened average lifetimes, more pronounced at the lowest OCG concentration. Interestingly, the sample with the lowest OCG concentration and smallest hydrodynamic diameter displays the shortest fluorescence lifetime. In smaller micelles, a higher surface-to-volume ratio places a greater fraction of SQWs near the micelle-water interface, increasing exposure to OCG headgroups and the aqueous phase. This proximity may enhance surfaceor trap-mediated nonradiative relaxation, affecting the faster τ2 component while leaving the long-lived component largely unchanged. This is a possible contributing factor, alongside other micellization-related effects Chapter 3. Quantum Dots for Detecting Neuronal Activity 101 such as changes in local dielectric environment, interfacial strain, or surface ligand rearrangement. Figure 3.24. Photoluminescence lifetime analysis of encapsulated CdS/CdSe/CdS SQWs. (A) PL decay curve for sample 1, fitted with a biexponential model yielding lifetimes of τ₁ = 44.0 ± 1.0 ns and τ₂ = 6.0 ± 0.22 ns, corresponding to an average lifetime of 18.4 ns. (B) PL decay of sample 3, showing τ₁ = 43.0 ± 1.1 ns and τ₂ = 9.1 ± 0.31 ns, resulting in a longer average lifetime of 23.4 ns. The increased τ₂ component in sample 3 suggests improved passivation or altered surface dynamics. Transmission electron microscopy analysis of the CdS/CdSe/CdS SQWs encapsulated in OCG micelles reveals that the micelles exhibit a spherical morphology with diameters ranging from approximately 20 to 100 nm (Figure 3.25 A). Higher-magnification imaging confirms that each micelle contains multiple SQWs densely packed within the micellar core (Figure 3.25 B). Statistical analysis of 250 micelles yields an average diameter of 44.3 ± 11.6 nm (Figure 3.25 C), which is larger than the average hydrodynamic diameter of 35.2 ± 7.5 nm determined by dynamic light scattering (Figure 3.23). The shift in distribution maximum between TEM and DLS measurements may result from population mismatch due to sample polydispersity or from drying-induced deformation of soft micelles during grid preparation. Overall, the OCG micelle encapsulation method is an effective strategy for preserving the photophysical properties of nanocrystals during phase transfer, providing a rapid and practical route to obtain water-dispersible QDs while maintaining the original spectral features. Chapter 3. Quantum Dots for Detecting Neuronal Activity 102 Figure 3.25 TEM analysis of CdS/CdSe/CdS spherical quantum wells encapsulated within OCG micelles. (A) Low-magnification TEM image showing SQWs encapsulated within micelles, with overall micelle sizes ranging from 20 to 100 nm. (B) Higher magnification image of the selected area in (A) reveals that individual micelles enclose multiple SQWs. (C) Histogram of micelle size distribution obtained from the analysis of 250 micelles based on TEM images. The average micelle diameter is 44.3 ± 11.6 nm. 2.2. Lipid Vesicles Another widely employed strategy for transferring hydrophobic nanomaterials into aqueous environments involves their encapsulation within lipid bilayer vesicles, commonly referred to as liposomes. This method enables the transport of lipophilic, ligand-coated nanoparticles through water-based media while preserving their native surface chemistry. Liposome formation reflects a general organizational principle observed in biological systems, where amphiphilic molecules spontaneously arrange into bilayer structures to form cell membranes, transport vesicles, and compartmentalized reaction spaces. Amphiphilic lipids contain polar headgroups and non-polar tails, and their assembly is driven primarily by the hydrophobic effect, supported by van der Waals forces, electrostatic interactions, and hydrogen bonding. Although each of these non-covalent interactions is individually weak, their combined effect yields stable bilayer membranes. Kinetic factors, such as mixing rate, can influence the assembly pathway, potentially leading to structural variations in membrane thickness or vesicle morphology. Owing to their architecture and composition, liposomes can encapsulate both hydrophilic species in the aqueous core and hydrophobic species in the bilayer. These properties have enabled their widespread use as model systems for studying membrane-associated phenomena and as carriers for a broad range of substances, including therapeutic drugs and nanomaterials.38,39 Encapsulation within liposomes can reduce nanoparticle cytotoxicity and improve stability, facilitating biomedical applications. 38,40 Examples include gold nanoparticles for stimuli-responsive release, 41 silver nanoparticles for antimicrobial treatments,42,43 zinc 50 nm BA 200 nm C Chapter 3. Quantum Dots for Detecting Neuronal Activity 103 oxide44,45 and cerium oxide46 nanoparticles for cancer therapy, titanium dioxide nanotubes for sustained drug delivery, and superparamagnetic iron oxide nanoparticles in magnetoliposomes for hyperthermia, targeted chemotherapy, and magnetic resonance imaging (MRI).47–50 Liposomes have also been used to encapsulate semiconductor quantum dots for bioimaging,51,52 and hybrid systems combining liposomes with silica, polymers, or graphene-based materials have been developed to enhance encapsulation efficiency and enable stimuli-responsive release.53,54 Typically, the nanoparticles are co-dissolved with phospholipids in an organic solvent; after solvent evaporation and hydration, the lipids spontaneously self-assemble into bilayer vesicles, with the hydrophobic nanocrystals becoming embedded within the nonpolar region of the membrane. We conducted a series of experiments to optimize the synthesis of phospholipid vesicles incorporating quantum dots within their lipid bilayers. The phospholipids used were selected based on their structural and functional properties, which support both stable vesicle formation and the potential for membrane fusion – an aspect relevant to our envisioned biomedical applications. A detailed justification for the lipid composition is provided in Section 3.2, Fusosomes – Fusogenic Lipid Vesicles (page 127). However, it is worth noting here that the molecular characteristics of these lipids were chosen to enhance the interaction of the vesicles with biological membranes, which is central to the aim of this study. Preparation of lipid vesicles Vesicles can be classified as unilamellar, bilamellar, or multilamellar based on the number of lipid bilayers they contain. Phospholipids, depending on their molecular size and threedimensional arrangement, commonly form either small unilamellar vesicles or giant unilamellar vesicles.55 To meet the central objectives of the present work, the goal was to produce small, unilamellar vesicles with minimal size variability. The lipid composition used in these experiments was adapted from the work of G. Gopalkrishnan et al.56 A lipid solution was prepared in chloroform at a total concentration of 1 mM, comprising 5% of the cationic lipid DOTAP, 0.5% DPPE-PEG2000, and 74.5% DMPC (see molecular structures in Figure 3.26 A) , and was mixed with 3 × 10-7 mmol of quantum dots. Once the mixture was thoroughly homogenized, the solvent was evaporated under a stream of nitrogen, and the resulting lipid film was further dried under vacuum for 2 hours. Following dehydration, 1 mL of PBS buffer was added to the vial, and the solution was vortexed to Chapter 3. Quantum Dots for Detecting Neuronal Activity 104 redisperse the quantum dots. A magnetic stirrer was introduced to ensure thorough mixing. To facilitate complete dispersion, the vial was intermittently vortexed and briefly immersed in a hot water bath. After obtaining a homogeneous mixture of lipids and quantum dots in PBS, a freeze-thaw cycle was performed by alternately placing the sample vial in a liquid nitrogen bath (~-200 °C) and a water bath maintained at 50 °C (Figure 3.26 B). This technique is closely tied to the lipids' phase transition temperature, often referred to as the melting temperature (Tm) – the point at which lipid molecules shift from a solid-like gel phase to a fluid-like liquid crystalline state. This transition plays a critical role in liposome preparation, as it influences membrane stability and permeability. Below the Tm, lipid bilayers are tightly packed with restricted molecular mobility; above the Tm, they become more fluid, allowing lateral diffusion of lipid molecules within the membrane. Repeated freeze-thaw cycles, commonly employed to transform multilamellar vesicles into unilamellar ones, leverage this phase behavior. During freezing, ice crystal formation disrupts the bilayers, breaking larger vesicles into smaller fragments; upon thawing, these fragments reassemble into unilamellar structures. The efficiency of this process depends on several factors, including lipid composition, ionic strength of the buffer, and the Tm, which is especially critical during the thawing phase when membrane fluidity is required for vesicle fusion. Properly optimizing these parameters leads to more uniform and stable vesicle populations, thereby improving the reliability and reproducibility of vesicle-based systems.57,58 DOTAP exhibits a transition temperature below 5 °C, DMPC has a well-characterized transition temperature of approximately 24 °C, and DSPE undergoes a phase transition at around 74 °C; while PEGylation may slightly alter this value, it is expected to remain relatively high.59 All values are taken from data provided by Avanti Polar Lipids (https://avantiresearch.com/). To calculate the transition temperature of a mixed phospholipid system, we can use a weighted average approach based on the composition and individual transition temperatures of the lipids. Tm(mixture) = (5 °C × 0.05) + (74 °C × 0.005) + (24°C × 0.745) = 18.5 °C (3.1) Therefore, the estimated transition temperature of the mixed phospholipid system is approximately 18.5 °C. It is important to note that this is a simplified calculation and the actual transition behavior of mixed lipid systems can be more complex. Factors such as lipid-lipid interactions, domain formation, and the presence of PEGylated lipids can affect the overall phase behavior. More accurate determinations would require experimental measurements using techniques such as differential scanning calorimetry (DSC) or nanoplasmonic sensing (NPS) as described in the literature.57,60 Chapter 3. Quantum Dots for Detecting Neuronal Activity 111 Stability study The visual inspection of the solutions containing vesicles and QDs revealed slight haze under white light Figure 3.34 and bright fluorescence under UV illumination (λex = 365 nm) indicating homogenous distribution of quantum dots (Figure 3.19). Figure 3.34 Visual comparison of vesicle solutions at three different extrusion stages under (A) white light and (B) UV irradiation. To evaluate long-term stability of vesicle, we monitored the change of hydrodynamic diameter of samples extruded through 100 nm (Extrusion IV), 400 nm (Extrusion III), and 800 nm (Extrusion II) membranes over a five-day period. The results are presented in Figure 3.35 which includes extinction spectra (A, B, C), emission spectra (D, E, F), and dynamic light scattering (DLS) data (G, H, I) for each vesicle size. The extinction spectra (Figure 3.35 A, B, C) remain largely unchanged throughout the observation period, suggesting structural integrity are well preserved. In the emission spectra (Figure 3.35 D, E, F), the fluorescence profile remains consistent over five days; however, a gradual decrease in intensity is observed. This decline may result from factors such as vesicle aggregation, partial degradation of quantum dots in the aqueous environment, or minor sample dilution during repeated handling. Dynamic light scattering (DLS) measurements of the vesicles from Extrusion IV, Extrusion III, and Extrusion II, which represent progressively larger average vesicle sizes, were performed over a five-day period to provide insights into their stability. For the smallest vesicles from Extrusion IV (Figure 3.35 G) the Z-average diameter decreased from 116.4 nm on day 1 to 109.4 nm on day 5 (-6%), accompanied by a modest increase in the polydispersity index from 0.129 to 0.158 (Figure 3.36). Vesicles extruded through 400 nm membranes (Extrusion III, Figure 3.35 H) showed no notable change after 5 days of incubation, with differences below 2%: the Z-average diameter decreased from 203.8 nm to 201.2 nm, while the PdI showed a slight drop from 0.216 to 0.210 (Figure WHITE LIGHT UV‐LIGHT Extrusion II (800 nm) A B Extrusion III (400 nm) Extrusion IV (100 nm) Extrusion II (800 nm) Extrusion III (400 nm) Extrusion IV (100 nm) Chapter 3. Quantum Dots for Detecting Neuronal Activity 112 3.36). The largest vesicles from Extrusion II, obtained using an 800 nm-pore membrane (Figure 3.35 I), exhibited an increase in Z-average diameter from 256.8 nm to 269.6 nm (+5%), accompanied by a rise in PdI from 0.247 to 0.270 (Figure 3.36). The DLS profile broadened significantly by day 5, indicating increased heterogeneity. These results suggest a tendency toward vesicle fusion or aggregation over time, highlighting the comparatively lower structural stability of larger vesicles during prolonged storage. Vesicles subjected to three or four extrusion cycles through smaller-pore membranes (400 nm and 100 nm) maintained relatively stable diameters over the first two days, with changes below 1% (116.4 to 115.3 nm and 203.8 to 203.7 nm, respectively). However, all three vesicle populations exhibited measurable size or PdI variations over five days, indicating that the dispersions are most reliable when used immediately after preparation or within a short time frame, preferably within 24 hours. Chapter 3. Quantum Dots for Detecting Neuronal Activity 113 Figure 3.35 Stability assessment of vesicles extruded through membranes of 100 nm (Extrusion IV; A, D, G), 400 nm (Extrusion III; B, E, H), and 800 nm (Extrusion II; C, F, I) over a five-day period. Extinction spectra (A, B, C) show minimal changes. Emission spectra (D, E, F) reveal a gradual decrease in fluorescence intensity, possibly due to aggregation or QD degradation. DLS measurements (G, H, I) indicate a slight increase in vesicle size, particularly in the 800 nm samples, suggesting some degree of aggregation or fusion over time. A B C D E F G H I Chapter 3. Quantum Dots for Detecting Neuronal Activity 114 Figure 3.36 Summary of DLS measurement of vesicles extruded through 100 nm (Extrusion IV), 400 nm (Extrusion III), and 800 nm (Extrusion II) membranes over a five-day period. The table summarizes Z-average diameters and polydispersity indices obtained from dynamic light scattering measurements on days 1, 2, and 5. The graph illustrates temporal changes in average vesicle diameter for the three samples, showing minor variations for Extrusion IV and Extrusion III vesicles and an increase Extrusion II vesicles. Fluorescence lifetime measurements were conducted to gain deeper insight into the photophysical behavior of the vesicle samples over time. Measurements were performed on vesicles obtained after extrusion through 100 nm-, 400 nm-, and 800 nm-pore membranes on days 1, 2, and 5. The resulting fluorescence decay curves were fitted using a triple exponential model, a commonly applied approach for describing complex, multi-component decay processes (Appendix, Figure A.2). The functional form of the model is shown below: 𝑓(𝑡)= 𝐻 + 𝐴1𝑒{−𝑡 𝜏1}+ 𝐴2𝑒{−𝑡 𝜏2}+ 𝐴3𝑒{−𝑡 𝜏3} (3.2) Here, f(t) represents the fluorescence intensity at time t; H is the baseline offset; A1, A2, and A3 are the amplitudes of the three decay components; and 𝜏1, 𝜏2, 𝜏3 are the corresponding lifetimes. This model accounts for multiple emissive species or environments contributing to the overall decay profile. Although the fit quality showed minor deviations beyond 100 ns in some cases, this had little effect on the analysis, as the majority of photon counts occurred within the first 40 ns – the time window most critical for accurate lifetime determination. A comparative analysis of the fitted decay curves is shown in the following Figure 3.37, highlighting changes in the average fluorescence lifetimes (𝜏ₐᵥ), calculated as: 𝜏 av =𝐴1𝜏1+𝐴2𝜏2+𝐴3𝜏3 𝐴1+𝐴2+𝐴3 (3.3) Chapter 3. Quantum Dots for Detecting Neuronal Activity 115 across the different days of measurement. The model incorporates two exponential decay components, each characterized by a distinct lifetime 𝜏1, 𝜏2, and 𝜏3. The pre-exponential factors, A1, A2, A3, represent the relative contributions of these two decay processes to the overall fluorescence intensity. The average fluorescence lifetime of vesicles extruded through 100 nm membranes dropped markedly over time, from 2.1 ± 3.0 ns on day 1 to just 0.5 ± 0.2 ns by day 5, indicating a substantial rise in non-radiative decay processes. This rapid decrease aligns with mechanisms such as trap-assisted recombination and Auger-type interactions, both of which can become more prominent when become surface-charged or develop new trap states.61 A likely contributing factor is curvature-induced stress in the lipid membrane: at higher curvature, as seen in smaller vesicles, the lipid bilayer may exhibit packing defects and transient permeability to water, oxygen, and ions. This increased permeability can facilitate oxidative damage or trion formation, both of which degrade photophysical performance over time.62,63 Even though bilayer thickness and lipid composition are identical across all batches, the smaller radius amplifies surface-to-volume ratio and membrane tension, accelerating these degradation pathways. By contrast, vesicles extruded through 400 nm and 800 nm membranes, whose hydrodynamic diameters are around 200 nm, maintained stable lifetimes (Figure 3.37). The modestly larger radius lowers curvature stress, reduces defect density, and limits solvent/ion ingress, thereby suppressing the creation of surface traps and keeping Auger recombination rates near their initial values. Collectively, the results reveal a curvature-dependent photostability: smaller vesicles experience faster lifetime erosion owing to curvature-induced membrane leakiness, while slightly larger vesicles (< 200 nm) preserve QD optical integrity over several days. For experiments requiring consistent optical output, vesicles should ideally be used shortly after preparation; however, extrusion through membranes ≥ 400 nm and keeping the diameter of a vesicle slightly larger presents a straightforward strategy for improving formulation robustness. Chapter 3. Quantum Dots for Detecting Neuronal Activity 116 Figure 3.37 Time-resolved fluorescence decay analysis of lipid vesicles with different diameters: 100 nm (A, B), 400 nm (C, D), and 800 nm (E, F), measured on Day 1, Day 2, and Day 5 post-synthesis. Panels A, C, and E present the fitted fluorescence decay curves showing how fluorescence lifetimes evolve over time for each vesicle size. The corresponding average lifetimes are summarized in panels B, D, and F, respectively. Notably, 100 nm vesicles exhibit a marked decrease in fluorescence lifetime over time, indicating reduced photophysical stability, while larger vesicles (400 nm and 800 nm) maintain more consistent lifetimes over the five-day period. This suggests that vesicle size plays a role in the temporal stability of encapsulated fluorophores or quantum dots. PLQY in the presence of lipids The initial phase of our work on optimizing lipid vesicle preparation was conducted using commercially available CdSe/ZnS core/shell quantum dots (QDs) functionalized with trioctylphosphine oxide (TOPO). Given their low cost and wide availability, these QDs served as an ideal system for preliminary protocol development and optimization of the experimental workflow. Throughout these early experiments, we observed that the fluorescence of the A B C D E F AVERAGE LIFETIME 100 nm vesicles: Day 1: 2.1 ns ± 3.0 ns Day 2: 0.9 ns ± 1.0 ns Day 5: 0.5 ns ± 0.2ns Day 5, 100 nm: 0.5 ns ± 0.2 ns Average life me for Day 5, 400 nm: 1.8 ns ± 2.2 ns Average life me for Day 5, 800 nm: 1.8 ns ± 2.8 ns 100 nm: 0.5 ns ±0.2 ns Average life me for Day 5, 400 nm: 1.8 ns ± 2.2 ns Average life me for Day 5, 800 nm: 1.8 ns ± 2.8 ns AVERAGE LIFETIME 400 nm vesicles: Day 1: 1.5 ns ± 2.3 ns Day 2: 1.9 ns ± 2.7 ns Day 5: 1.8 ns ± 2.2 ns AVERAGE LIFETIME 800 nm vesicles: Day 1: 1.6 ns ± 2.6 ns Day 2: 1.9 ns ± 3.0 ns Day 5: 1.8 ns ± 2.8ns Chapter 3. Quantum Dots for Detecting Neuronal Activity 117 commercial QDs remained stable in lipid environments, with any minor decrease attributed primarily to material loss during the extrusion process. However, achieving the broader objectives of this research required transitioning to our material of choice: CdS/CdSe/CdS spherical quantum wells (SQWs), synthesized in Jonathan Owen’s laboratory. These more sensitive nanostructures are better suited for future applications due to their unique optical and electronic properties. The incorporation of SQWs into lipid systems revealed displayed a sharp decline in photoluminescence almost immediately upon mixing with lipids in non-polar solvents. This effect was especially pronounced in the presence of cationic lipids like DOTAP, suggesting that surface polarization might be triggering enhanced non-radiative recombination pathways. To preserve fluorescence, surface passivation must be improved to mitigate quenching effects. Several approaches for controlling surface chemistry were implemented. SQWs functionalized with Cd-oleate were evaluated first. This system, however, proved to be highly vulnerable to lipid-induced fluorescence loss, with PLQY values dropping from 31.6% in DCM to just 2.2% and 0.25% in the presence of DOTAP and DSPE-PEG, respectively (Figure 3.38). We then explored fluorinated passivation using CdF2 combined with octylamine. Although these nanoparticles exhibited a high initial PLQY of 94.0% in DCM, they too experienced quenching in lipidic environments, highlighting the insufficient photostability of this passivation strategy (Figure 3.39). The surface treatment with Zn-oleate ligands showed improved resistance to quenching, with PLQY values remaining relatively stable at 51% in DCM in the presence of DMPC and dropping to 10-12% in DOTAP and DSPE-PEG solutions. This effect can be attributed to the stronger affinity of Zn-oleate to anionic sulfur compared to Cd-oleate, which leads to more robust binding of Zn-oleate ligands to the nanocrystal surface and more effective passivation of surface states.64 We further introduced a ZnS shell to form a robust core/shell structure, which markedly enhanced the optical stability of the SQWs. ZnS is particularly effective because undercoordinated sulfur atoms at the nanocrystal surface act as trap sites that promote nonradiative recombination; overcoating with ZnS suppresses these defect-related pathways and passivates surface states, thereby stabilizing the emission. In addition, the wider bandgap of ZnS relative to CdS or CdSe increases carrier confinement within the CdSe core and mitigates Auger recombination, further improving both photoluminescence efficiency and stability.11,65 As a result, these ZnS-shelled SQWs stabilized with Zn-oleate ligands maintained high PLQY values (~54%) across all lipid formulations tested, including DMPC, DOTAP, and DSPE-PEG, and retained a notable 29% PLQY even in aqueous PBS vesicle suspensions (Figure 3.40). This final system Chapter 3. Quantum Dots for Detecting Neuronal Activity 118 demonstrated that with the right combination of inorganic shell and surface ligands, SQWs can be effectively shielded from environmental perturbations, paving the way for their reliable use in biological applications. Extinction and emission spectra corresponding to each studied system (System I – IV) are provided in the Appendix (Section 3.3, Figure A.3). Figure 3.38 Photoluminescence response of CdS/CdSe/CdS spherical quantum wells (SQWs) functionalized with Cd-oleate upon exposure to lipid environments. (A) Schematic of SQW surface functionalization with cadmium oleate ligands. (B) Photoluminescence quantum yield (PLQY) measurements of SQWs in various media, demonstrating significant fluorescence quenching upon mixing with lipid components, particularly DOTAP and DSPE-PEG. (C) Summary of PLQY values corresponding to the samples shown in panel B. (D) Digital images of SQW samples under UV-light illumination, visually confirming the fluorescence loss in lipid-containing environments. Chapter 3. Quantum Dots for Detecting Neuronal Activity 119 Figure 3.39 Photoluminescence of CdS/CdSe/CdS functionalized with CdF2 and octylamine in lipid environments. (A) Illustration of SQW surface passivation (B) PLQY measurements showing a sharp decline in emission upon exposure to lipid components, particularly DOTAP and DSPE-PEG, and near-total quenching in PBS vesicles. (C) PLQY values from B. (D) Digital images of SQW samples under UV irradiation. Figure 3.40 Photoluminescence stability of passivated CdS/CdSe/CdS/ZnS stabilized with Zn-oleate in lipid environments. (A) Schematic representation of the nanocrystal featuring a ZnS outer shell. (B) PLQY measurements showing consistent emission in lipids, with only a modest decrease observed in PBS. (C) PLQY values for each condition as shown in B. (D) Digital images of the SQW samples under UV-light irradiation, confirm fluorescence retention in all lipid media tested, and underscoring the effectiveness of the ZnS/Zn-oleate surface passivation strategy. Chapter 3. Quantum Dots for Detecting Neuronal Activity 120 3. Cell Membrane Insertion The next step was to assess their ability of SQWs to interface with biological membranes – an essential requirement for their intended use in detecting neuronal activity. This section focuses on the membrane insertion capabilities of SQWs delivered via the two strategies previously developed: OCG micelles and lipid vesicles. The primary objective was to evaluate whether these delivery systems could successfully embed SQWs into cell membranes in a manner compatible with their photoluminescent function and biological stability. To validate membrane insertion in a biologically relevant context, we employed cultured HEK cells as a model system. Human embryonic kidney (HEK293-T) cells provide a robust and wellcharacterized platform for initial biological validation. In some experiments, unmodified HEK293-T cells were used to study fundamental membrane interactions. In others, we utilized self-spiking HEK cells transfected with voltage-gated sodium and potassium channels, providing a dynamic model for testing activity-dependent fluorescence responses. Details of the cell line used are provided in the description of each experimental result. The experiments presented in the following subsections aim to assess both the membrane integration of SQWs and their functional responsiveness under conditions of real-time neuronal activity, establishing a foundation for their future use in all-optical electrophysiology, as detailed later in Subchapter 4, Recordings of Spiking HEK Cells with Spherical Quantum Dots (page 153). 3.1. OCG Micelles We began by evaluating the membrane insertion capabilities of SQWs encapsulated within OCG micelles. The central question was whether these micelles could bring SQWs into close apposition with the plasma membrane of live cells without compromising cell viability – critical prerequisites for optical sensing applications. Initial experiments were conducted using a specialized HEK cell line obtained from ATCC (CRL‑3479), originally developed by Adam Cohen’s laboratory.66 This Tet-On “spiking” HEK cell line is genetically engineered to constitutively express the voltage-gated sodium channel NaV1.5, while the inward-rectifier potassium channel Kir2.1 is placed under doxycyclineinducible control. The Kir2.1 channel is also fused to cyan fluorescent protein (CFP), which emits fluorescence with an excitation peak around 456 nm and an emission peak near 480 nm. Upon addition of doxycycline (typically 1-2 µg/mL), Kir2.1‑CFP is robustly expressed, leading to membrane hyperpolarization. In combination with NaV1.5, this enables the cells to generate Chapter 3. Quantum Dots for Detecting Neuronal Activity 127 The unusually bright cell in the lower-right corner is most likely attributable to spectral bleedthrough from CFP, as previously noted. Across multiple replicates and experimental conditions, this punctate, extracellular accumulation pattern was consistently observed, even when micelle formulations were supplemented with additives such as Lipofectamine or Pluronic F-127 – neither of which altered the distribution or promoted membrane integration. These results suggest that although the purified OCG micelles are biocompatible and capable of approaching the cell surface, their structural and chemical properties are insufficient to support efficient insertion of SQWs into the lipid bilayer. This limitation prompted a strategic shift toward alternative delivery platforms, specifically lipid vesicles, which offer a more membrane-compatible interface for SQW incorporation, as described in the following sections. 3.2. Fusosomes – Fusogenic Lipid Vesicles In an earlier chapter (2.2 Lipid Vesicles, page 102), when introducing the preparation of lipid vesicles and the encapsulation of quantum dots, the specific choice of lipids was deliberately left without detailed justification. The vesicles described so far were not designed merely as carriers for solubilizing hydrophobic quantum dots in aqueous media. Their formulation was driven by enabling them a fusion with biological membranes. Through careful selection of lipid components, a conventional liposome can be transformed into a fusosome – a vesicle with the intrinsic ability to merge with cellular membranes and deliver its contents directly into the membrane or the cytosol. Fusogenic liposomes, or fusosomes, are lipid-based nanocarriers engineered to merge directly with cellular membranes, enabling efficient cytosolic delivery of therapeutic or diagnostic cargo. These systems often incorporate viral fusion peptides or synthetic ligands to improve target specificity and enhance membrane fusion efficiency. Membrane fusion itself is a fundamental biological process in which two lipid bilayers merge to form a continuous membrane, allowing the contents of previously distinct compartments to mix. This mechanism underlies essential cellular activities, including endocytosis, exocytosis, synaptic vesicle release, and intracellular vesicle trafficking. Fusion proceeds through a series of orchestrated steps: membrane recognition, docking, local bilayer destabilization, hemifusion (outer leaflet merging), and finally full fusion, characterized by pore formation and content mixing. Chapter 3. Quantum Dots for Detecting Neuronal Activity 128 General frameworks for designing fusosomes should consider few crucial elements: 1. Building-Block Lipids • Structural scaffold: A zwitterionic phosphatidylcholine (PC) lipid (e.g. POPC) provides a stable bilayer “skeleton” that maintains vesicle integrity and fluidity under physiological conditions.67 • Curvature modulators: Cone-shaped lipids such as phosphatidylethanolamine (PE) or high-curvature POPC enrich the bilayer with the necessary negative curvature, helping to stabilize the structure of vesicle and maintain membrane integrity.67 • Membrane stiffeners: Cholesterol intercalates between phospholipids, increasing bilayer order and rigidity to prevent premature cargo leakage without abolishing fusogenicity (membrane-fusion capacity).67 • Charge anchors: Cationic lipids (e.g. DOTAP) enhance electrostatic attraction to the typically anionic cell surface, promoting vesicle docking and fusion initiation.68,69 2. Fusogenic Triggers • Aromatic lipids: Lipids bearing aromatic headgroups (e.g. fluorescent BODIPYor rhodamine-tagged PE) intercalate into membranes, locally perturbing lipid packing and reducing the activation energy for stalk formation; they also serve as built-in tracers for imaging. 68,70,71 • Conical lipids: In addition to their role as key curvature-inducing building blocks in the bilayer, their pronounced cone shape actively drives fusion by creating local membrane strain that stabilizes hemifusion intermediates and promotes pore opening. • Fusogenic peptides: Short peptides derived from viral fusion proteins insert into and destabilize target bilayers; studies show they induce liposome shrinkage and local lipid rearrangements that prime membranes for fusion.72,73 3. Application Specific Surface Modifications: • Antifouling zwitterions: Inclusion of additional zwitterionic lipids (e.g. DOPC) resists serum-protein adsorption, preserving fusogenic activity in complex media.74 • PEGylation: Grafting PEG chains (via DSPE-PEG) further reduces nonspecific interactions, enhances colloidal stability in serum, and extends circulation time in vivo, with minimal impact on fusion when optimized at low molar ratios.56,74 Chapter 3. Quantum Dots for Detecting Neuronal Activity 129 • Targeting and reporting: Introduce reactive headgroups for antibody or ligand conjugation,69,70,72,75, fluorescent lipids for imaging,70,71 or stimuli-responsive lipids for controlled cargo release52. • Incorporate lipids bearing hydrophobic anchors (e.g., cholesterol or tocopherol) to secure fusosomes onto target membranes or solid supports.76 By balancing these elements – structural lipids, active fusogens, and adaptive surface modifications – fusosomes can be rationally tuned to achieve rapid, efficient, and specific membrane fusion for cytosolic delivery of drugs, nucleic acids, proteins, or imaging agents.77 The fusosome composition in this study consisted of DMPC (1,2-dimyristoyl-sn-glycero-3phosphocholine), DOTAP (1,2-dioleoyl-3-trimethylammonium-propane), and DPPE-PEG2000 (1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-methoxy(polyethylene glycol)-2000).56 At physiological temperature (37 °C), DMPC exists in a fluid, liquid-crystalline phase (Tm ≈ 24 °C), providing a flexible and stable bilayer structure that supports the dynamic membrane rearrangements required for fusion. The addition of 5 mol% DOTAP, a cationic lipid, introduces a positive surface charge that facilitates strong electrostatic interactions with negatively charged cellular membranes. This enhances membrane binding and lowers the energetic barrier to initial bilayer destabilization and hemifusion.56,70,77 An important component is DPPE-PEG2000 that plays a multifaceted role. While PEGylation is traditionally associated with providing steric stabilization and reducing aggregation in serum-containing environments, at low concentrations it also supports fusion.56,76 The PEG chains help exclude interfacial water, reducing hydration repulsion between bilayers – a phenomenon often referred to as “PEG-induced dehydration”.78,79 In addition, the bulky PEG headgroups introduce lateral pressure at the membrane interface, generating packing defects and localized curvature that can serve as nucleation sites for lipid mixing.80 This effect is particularly beneficial when used alongside fusogenic or cationic lipids, as it promotes closer membrane apposition without completely shielding the surface charge. Together, this carefully balanced combination of lipids ensures both colloidal stability and high fusogenic potential, enabling the vesicles to function effectively as fusosomes for direct membrane interaction and intracellular delivery.56 Chapter 3. Quantum Dots for Detecting Neuronal Activity 130 Figure 3.45 Schematic representation of fusosome structure and fusion mechanism. (A) Cross-sectional view of a fusosome composed of 74.5 mol% DMPC, 5 mol% DOTAP, and 0.5 mol% DPPE-PEG2000, with nanocrystals embedded between the lipid bilayers. The bilayer architecture is stabilized by DMPC, which exists in a fluid, liquid-crystalline phase at physiological temperature (37 °C). Surface-exposed DOTAP introduces a net positive charge, enhancing electrostatic interactions with negatively charged cellular membranes. PEGylated lipids (depicted with extended blue chains) contribute steric stabilization and promote membrane fusion by facilitating interfacial dehydration and inducing packing defects. (B) Upon contact with a negatively charged cellular membrane, the positively charged fusosome engages in electrostatic interactions, resulting in spontaneous membrane fusion. This process is driven by reduced hydration repulsion and membrane curvature stress induced by the PEGylated lipids, promoting lipid mixing and direct bilayer merger. The embedded nanocrystals are transferred directly into the target membrane, where they remain localized within the energetically favorable hydrophobic environment of the lipid tails. In order to confirm the fusogenic capability of the liposome formulation, we performed a series of in-vitro experiments using liposomes labeled with the lipophilic fluorescent dye DiI (1,1'-Dioctadecyl-3,3,3',3'-Tetramethylindocarbocyanine Perchlorate). DiI is a highly hydrophobic, cationic indocarbocyanine dye that anchors into lipid bilayers via its two long C18 alkyl chains and serves as a highly sensitive probe for membrane fusion. Notably, DiI exhibits minimal fluorescence in aqueous environments due to aggregation and intramolecular flexibility, which favor non-radiative decay. Upon insertion into a lipid bilayer, the dye becomes brightly fluorescent as its chromophore is stabilized within the ordered, low-polarity membrane environment. This transition disrupts aggregate formation, restricts torsional motion of the polymethine chain, and significantly enhances its quantum yield – making fluorescence effectively gated by membrane embedding. The chemical structure of the dye is shown in Figure 3.46 A, and its spectroscopic profile after incorporation into vesicle membranes (Figure 3.46 B) reveals a distinct absorption peak near 550 nm and an emission centered at 570 nm. This environment-induced fluorescence makes DiI an ideal tracer for detecting lipid mixing and vesicle-cell fusion events. In this set of experiments, we varied amounts of PEGylated lipid (DSPE-PEG2000) and kept constant DMPC and DOTAP. To assess how PEG density influences membrane fusion efficiency, we systematically adjusted the PEG content to 0.5 mol%, 1.0 mol%, 2.5 mol%, and 5.0 mol%. This concentration range was established based on previous findings by Gopalakrishnan et al.56 Chapter 3. Quantum Dots for Detecting Neuronal Activity 131 To visualize fusion events, we added 10 mol% of the fluorescent lipid DiI to each formulation. The remaining 90 mol% was used to preserve the original DMPC/DOTAP ratio (approximately 74.5:25), with the specified amount of PEG-lipid incorporated into this fraction. While the addition of DiI slightly reduced the absolute percentages of the other components, their relative proportions remained consistent across samples. This design ensured that any differences in fusion behavior could be attributed to the PEG content, without introducing variability from changes in the core lipid composition. Panels C and D (Figure 3.46) display cuvettes containing the DiI-labeled vesicle suspensions under daylight and UV illumination, respectively. In white light, all samples look nearly identical, showing only a faint pink tint from the incorporated DiI, whereas UV excitation reveals the expected uniform orange-red fluorescence across every formulation. Figure 3.46 Characterization of DiI-labeled control liposomes used to evaluate vesicle-cell fusion. (A) Chemical structure of DiI [1,1ʹ-dioctadecyl-3,3,3ʹ,3ʹ-tetramethylindocarbocyanine perchlorate], a lipophilic cationic carbocyanine bearing two C18 alkyl tails that anchor stably into lipid bilayers. (B) Absorption (dashed blue) and photoluminescence (solid purple) spectra of DiI-labeleld liposomes. The dye, embedded within the vesicle membrane, exhibits a strong, narrow absorption peak centered at 550 nm and a fluorescence emission maximum at 570 nm, resulting in intense red-orange fluorescence. (C, D) Photographs of liposome suspensions containing 10 mol% DiI and increasing fractions of DSPE-PEG2000 (0.5, 1.0, 2.5, and 5.0 mol%). Images were taken under ambient white light (C) and UV illumination (D). Uniform bright-red fluorescence confirms successful incorporation of DiI in all formulations, enabling visual tracking of vesicle fusion in subsequent cell assays. Fusion experiments were performed on spiking HEK293 monolayers, and the extent of membrane staining was evaluated via fluorescence microscopy and quantitative intensity profiling (Figure 3.47). At low PEG concentrations (0.5 mol% and 1.0 mol%), DiI fluorescence appeared sparse and punctate, with intensity profiles characterized by narrow, isolated peaks 1 2 1 2 hite light i ill mina on n itrogen il tain 1,1 ‐Dioctadecyl‐3,3,3 ,3 ‐ Tetramethylindocarbocyanine Perchlorate AB C D Chapter 3. Quantum Dots for Detecting Neuronal Activity 132 and overall lower signal levels. This contrasts with the higher PEG conditions, reflecting more robust and continuous membrane labeling. This pattern is indicative of limited lipid mixing, suggesting incomplete hemifusion without substantial membrane incorporation. In contrast, at higher PEG concentrations (2.5 mol% and 5.0 mol%), the intensity profiles showed a continuous and uniform distribution of DiI along the cell periphery. The signal remained elevated across the scan, with minimal fluctuations toward lower intensity values. These trends paralleled the qualitative differences observed in the fluorescence images, confirming that vesicle fusion efficiency increases in a PEG-dependent manner. Across all conditions, bright-field images confirmed intact cell morphology, indicating no cytotoxic effects. Together, these data validate the functional integrity and fusogenic design of the liposomes proposed here. Figure 3.47 PEG-dependent fusion of DiI-labeled control liposomes with HEK293 cells. Bright-field (top row) and corresponding fluorescence (middle row) microscopy images acquired after 30 minutes of incubation with liposomes containing 0.5, 1.0, 2.5, or 5.0 mol% DSPE-PEG2000. All liposomes were labeled with 10 mol% DiI to track lipid mixing with the plasma membrane. Intensity profiles (bottom row) across the dashed purple lines reveal sparse, low-intensity, punctate signal at 0.5% and 1.0% PEG, and progressively stronger, more continuous fluorescence at 2.5% and 5.0% PEG, consistent with efficient membrane labeling. 1.0 PEG0.5 PEG 2.5 PEG 5.0 PEG 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m Chapter 3. Quantum Dots for Detecting Neuronal Activity 133 Given the high efficiency of membrane fusion observed with DiI-labeled vesicles containing 5 mol% PEG-lipid, we next examined the minimum incubation time required to achieve detectable lipid mixing at the cell surface. This is an important experimental parameter since the fusion time constrains the time window needed to obtain representative data. That is, one needs to obtain fast fusogenesis that prolongate as long as possible to assure constant conditions for analysis of spikes dynamics. We tested durations range from 5 to 30 min to assess the kinetics of fusion under optimized conditions. As shown in Figure 3.48, membrane-associated DiI fluorescence was already detectable after just 5 minutes, with a continuous outline marking the cell periphery. However, the signal at this early time point appeared noticeably dimmer compared to longer incubations. By 10 and 15 minutes, the fluorescence intensity increased compared to the 5-minute time point and closely resembled the pattern observed at 30 minutes, although with slightly lower overall brightness. This suggests that membrane fusion initiates rapidly and progressively intensifies over time. Bright-field images confirmed preserved cell morphology throughout the time course, indicating that rapid fusion does not compromise membrane integrity. These results demonstrate that fusion occurs within minutes under optimal conditions, and confirm that 30-minute time window is sufficient to ensure complete and uniform membrane labeling. In the following sections of this work, the terms vesicle, liposome, and fusosome will be used interchangeably, each referring specifically to the DMPC/DOTAP/DPPE-PEG2000-based vesicular structures described above (unless specified otherwise). Chapter 3. Quantum Dots for Detecting Neuronal Activity 134 Figure 3.48 Time-course analysis of membrane fusion using DiI-labeled liposomes (5 mol% PEG). Brightfield (top row) and corresponding fluorescence (middle row) images of HEK293 cells incubated with DiI-labeled vesicles for 5, 10, 15, 30 minutes. A continuous fluorescent outline is visible as early as 5 minutes, indicating rapid initiation of lipid mixing. Intensity profiles (bottom row) show progressive increases in signal amplitude and uniformity over time, with the 30-minute trace reaching the highest and most consistent values, reflecting efficient and widespread membrane labeling. Cell morphology remains preserved at all time points, confirming that fusion proceeds without compromising membrane integrity. Membrane staining Building on the six nanocrystal systems introduced in 3.1.1 Overview of Quantum-dot-based Platforms and illustrated in Figure 3.17 (page 87), this section examines their interaction with cell membranes. The series includes one commercially available CdSe/ZnS reference (System I) and four custom-designed spherical quantum wells (Systems III-VI), each developed iteratively in response to experimental challenges encountered throughout the study. With aqueous transfer procedures in place, the next step is to evaluate how effectively these quantum dots associate with fusogenic vesicles. The membrane-staining results assessment of labeling performance and serve as a functional presented here provide a comparative, qualitative prelude to the fusion behavior explored. 5 min 100 m 100 m 30 min 100 m 100 m 15 min 100 m 100 m 10 min 100 m 100 m Chapter 3. Quantum Dots for Detecting Neuronal Activity 135 System I To evaluate the fusion capacity of the formulated fusosomes, we first examined vesicles incorporating commercially available CdSe/ZnS quantum dots capped with TOPO – referred to as System I (Figure 3.49, Panels A and B) – on HEK293 cell monolayers. Confocal fluorescence microscopy was used to monitor vesicle-cell interactions in real time, with the goal of determining whether fusion with the plasma membrane occurred and whether the quantum dots became incorporated into the membrane structure. The emission spectrum of the quantum dots showed a peak at 542 nm (Figure 3.49 B), enabling clear visualization of membraneassociated fluorescence. These experiments served as an initial validation of vesicle-cell fusion and provided a basis for evaluating the membrane integration of the quantum dots. To systematically explore the variables influencing fusion efficiency, we investigated a range of parameters: cell seeding density (which affects monolayer confluency at the time of imaging), cell age (number of days post-seeding), the medium used during fusion (growth medium vs. PBS), fusion duration, time spent outside the incubator, and the volume of fusosomes suspension applied. A critical challenge during imaging was the presence of cellular autofluorescence, which can obscure weak signals from membrane-bound quantum dots. This endogenous background signal arises from naturally fluorescent biomolecules such as NADH, flavins, and aromatic amino acids, and tends to intensify as cells age or undergo stress. The cellular autofluorescence can contribute background signals that may be mistakenly interpreted as probe-derived emission. Therefore, we carefully assessed imaging parameters to minimize such interference and ensure accurate evaluation of fusosome-mediated membrane staining. Under the optimized conditions established through parameter screening, the most consistent and robust membrane labelling was achieved approximately 25-30 minutes after vesicle application. As shown in Figure 3.49 (Panels C-E), confocal fluorescence imaging of HEK293 cell monolayers (80% confluent, 24 hours post-seeding) incubated with quantum dot-loaded liposomes revealed a distinct green fluorescence outlining the cell periphery, accompanied by a limited number of bright intracellular puncta. For these experiments, 500 µL of liposome suspension – DMPC, DOTAP, and DSPE-PEG2000 at a 74.5:25:0.5 molar ratio and extruded through a 100 nm membrane – was added directly to 2 mL of complete culture medium (DMEM supplemented with 10 FBS) in each well. This gentle delivery strategy was chosen to preserve cell viability and minimize stress by avoiding abrupt Chapter 3. Quantum Dots for Detecting Neuronal Activity 136 changes to the environment. Cells were incubated with the liposomes for 10-30 minutes, with optimal membrane-associated signal observed consistently at the 25-30-minute time point. Imaging was performed using a Leica TCS SP5 confocal microscope with 458 nm laser excitation set to 30% of maximum power. Emission was collected using standard settings optimized for membrane-associated fluorescence. Full acquisition parameters are provided in the Appendix (Section 3.2.3). The resulting images, presented on Figure 3.49, exhibit a largely continuous membrane-localized fluorescence signal, strongly suggestive of quantum dot integration into or association with the plasma membrane. The occasional presence of bright intracellular spots may arise from fused vesicle aggregates or local QD accumulation. However, the lack of extensive cytoplasmic or perinuclear fluorescence argues against large-scale internalization as the dominant mechanism. The intensity plot profiles corresponding to the fluorescence micrographs within yellow rectangles (Fig. 3.31, panels C and D) provide further support for the membrane-associated localization of quantum dots. These profile scans show distinct peaks in fluorescence intensity at positions aligning with the cell periphery, indicating a higher concentration of fluorescent signal at the plasma membrane compared to the cell interior. The dual-peak pattern observed across individual cell profiles is consistent with fluorescence arising from left and right side of the membrane. The distance between the peaks corresponds to the diameter of HEK-293 cells, typically 10–15 µm, further confirming their assignment to cell boundaries. Background fluorescence remained stable at approximately 10 a.u., whereas membrane-associated regions reproducibly reached values twoto three-fold higher. Quantification across a population of approximately 30 cells confirmed this observation: statistics of fluorescence intensities measured at the cell outline and in the cytoplasm showed clearly separated distributions, with a calculated membrane-to-cytoplasm ratio (MCR) of 2.79 (Fig. 3.31 E). Overall, these spatial intensity profiles align with the interpretation that quantum dot-labeled vesicles preferentially associate with, or integrate into, the plasma membrane rather than being internalized. While definitive distinction between fusion, adsorption, and endocytosis would require colocalization with membrane or endosomal markers, the overall fluorescence pattern demonstrates effective membrane staining and confirms that the liposome formulation reached and interacted extensively with the plasma membrane under the tested conditions. Chapter 3. Quantum Dots for Detecting Neuronal Activity 143 Figure 3.52 Representative results for System III (CdS/CdSe/CdS SQWs) delivered using vesicles containing 5 mol% PEG-lipid. (A) Schematic of SQW structure with Zn-oleate ligands; lipid mix: DMPC 75%, DOTAP 25%, DSPE-PEG2000 5%. (B) UV-Vis (dashed blue) and PL (solid purple) spectra; λem = 530 nm. (C-D) Fluorescence remains punctate and non-membrane-associated despite increased PEG concentration. (E-F) Intensity profiles from Areas #1 and #2 (C and D respectively) show sparse, narrow peaks near control levels, confirming lack of membrane labeling. C C e C n leate Zn(COO C17H33)2 (COO C17H33)2 C A B C D 100 m 100 m 100 m 100 m 100 m 100 m Bright Field BF Em Emission Bright Field BF Em Emission C 2 E F Chapter 3. Quantum Dots for Detecting Neuronal Activity 144 Replacing the oleate shell with a mixed cadmium-fluoride/octylamine passivation layer (System IV) produced colloidal SQWs that retained the archetypal CdS/CdSe/CdS architecture (Figure 3.53 A) while emitting a narrow green band centered at 541 nm (Figure 3.53 B). The new ligand set improved the interaction between the nanocrystal-loaded vesicles and the plasma membrane. In both presented regions (Figure 3.53, panels C and D), the fluorescence channel revealed pattern coincident with the cell boundaries seen in bright-field mode. Only a few discrete intracytoplasmic spots were detected, and extracellular aggregates were largely absent. The merged BF + Em overlays further confirm that the photoluminescent signal is confined to the cell perimeter, indicating that the liposome cargoreachedthe outer leaflet of the HEK293 membrane. The absence of diffuse cytosolic glow argues against vesicle endocytosis, whereas the uniform rim suggests lipid-to-lipid fusion rather than adsorption. Chapter 3. Quantum Dots for Detecting Neuronal Activity 145 Figure 3.53 Representative membrane staining with System IV liposomes loaded with CdS/CdSe/CdS SQWs capped by CdF2/octylamine. (A) Schematic of SQW structure and ligand composition. Lipid mix: DMPC 74.5%, DOTAP 25%, DSPE-PEG2000 0.5%. (B) Optical spectra showing UV-Vis (dashed blue) and PL (solid purple), λem = 541 nm. (C-D) Fluorescence microscopy of HEK293 cells after 30 min incubation shows uniform membrane-associated signal. (E-F) Intensity profiles reveal consistent signal above baseline with regularly spaced peaks, indicating effective membrane localization. CdF2 NH(octyl) C C C e C C 2 ct lamine A B C D 100 m 100 m 100 m 100 m 100 m 100 m Bright Field BF Em Emission Bright Field BF Em Emission C 2 E F Chapter 3. Quantum Dots for Detecting Neuronal Activity 146 This interpretation is further supported by the fluorescence intensity profiles shown in panels E and F, corresponding to the emission regions from Areas #1 and #2. In both traces, the signal remains above the baseline obtained from an unstained control (Appendix, Figure A.4), with a series of peaks indicative of fluorescence along the cell periphery. This pattern stands in contrast to the profiles observed for System III, which exhibited irregular spikes corresponding to puncta and showed no indication of membrane association. The broader peaks observed for System IV suggest homogeneous distribution of the fluorophores and association with the plasma membrane across different regions of the sample. These spatial profiles, in agreement with the imaging data, support the conclusion that the CdF2/octylamine surface chemistry facilitates vesicle-membrane interaction while maintaining photoluminescence for imaging at 100 ms exposure. Compared to the discontinuous signal seen in System III, System IV achieves membrane-specific labeling, highlighting the importance of surface ligand composition in directing vesicle behavior at the cellular interface. However, while the signal is detectable under current conditions, the overall fluorescence intensity remains below the level required for highspeed imaging of neural dynamics. Our goal is to develop a labeling system that not only enables accurate membrane targeting but also exceeds the brightness of conventional indicators, allowing for shorter exposure times and higher frame rates. To enhance signal intensity while maintaining resistance to lipid-induced quenching, we subsequently explored System V, which features a CdS/CdSe/CdS core-well structure further passivated with an outer ZnS shell. As described in Section 3.2.2 PLQY in the presence of lipids, this additional layer significantly improves optical performance, yielding a PLQY of approximately 50% in non-polar lipid-rich solutions and retaining around 30% in aqueous media – making it the most photostable and efficient formulation within our current nanocrystal library. Despite this favorable photophysical profile, membrane staining using System V was less efficient than that achieved with the CdF2/octylamine-capped SQWs of System IV. Also, no effect of PEG concentration was observed on the quality of staining. Fluorescent microscopy of samples revealed a punctate fluorescence pattern distributed sparsely throughout the sample, with no continuous membrane-associated signal, regardless of PEG amount. Emissive spots were frequently observed away from the cell periphery, either adjacent to or detached from the membrane, suggesting inefficient delivery of nanocrystals to the plasma membrane and an absence of clear evidence for membrane fusion. Several factors may contribute to this reduced fusogenic behavior. The addition of the ZnS shell increases the overall particle size, which may elevate the energetic barrier for lipid mixing and bilayer incorporation. Furthermore, the long Chapter 3. Quantum Dots for Detecting Neuronal Activity 147 and conformationally rigid chain of Zn-oleate ligands may create a steric barrier that limits close contact between the SQWs and the cell membrane, potentially hindering effective mixing with the acyl chains of the lipid bilayer. These results underscore the delicate balance between optical optimization and membrane interaction, highlighting the need to tailor both surface chemistry and structural parameters to achieve efficient and bright membrane labeling. Chapter 3. Quantum Dots for Detecting Neuronal Activity 148 Figure 3.54 Fluorescence microscopy of System V: CdS/CdSe/CdS/ZnS SQWs with Zn-oleate ligands. (A) Schematic of nanocrystal structure; lipid formulation: DMPC 74.5%, DOTAP 25%, DSPE-PEG2000 0.5%. (B) UV-Vis (dashed blue) and PL (solid purple) spectra in non-polar solvent; λem = 489 nm. (C-D) Imaging of HEK293 cells shows sparse spots with no clear membrane-associated signal. (E-F) Intensity profiles across blue rectangles in C and D reveal weak, irregular peaks near baseline, indicating poor membrane localization. Zn(COO C17H33)2 n (COO C17H33)2 C C e C n n leate A B C D 100 m 100 m 100 m 100 m 100 m 100 m Bright Field BF Em Emission Bright Field BF Em Emission C 2 E F Chapter 3. Quantum Dots for Detecting Neuronal Activity 149 Figure 3.55 Representative results for System V (CdS/CdSe/CdS/ZnS SQWs) delivered using vesicles containing 5 mol% PEG-lipid. (A) SQW structure with Zn-oleate ligands and outer ZnS shell; lipid mix: DMPC 75%, DOTAP 25%, DSPE-PEG2000 5%. (B) UV-Vis (dashed blue) and PL (solid purple) spectra; λem = 489 nm. (C-D) Bright-field and fluorescence images show sparse, non-membrane-associated puncta despite increased PEG content. (E-F) Intensity profiles from Areas #1 and #2 (C and D respectively) remain near baseline, confirming lack of membrane staining. Zn(COO C17H33)2 n (COO C17H33)2 C C e C n n leate A B C D 100 m 100 m 100 m 100 m 100 m 100 m Bright Field BF Em Emission Bright Field BF Em Emission C 2 E F Chapter 3. Quantum Dots for Detecting Neuronal Activity 150 System VI To further optimize the surface chemistry of QD we performed a ligand exchange in which zinc bis(2-hexyldecanoate) [Zn(2-hexyldecanoate)2] was used in place of Zn(oleate)2 as the surface capping agent. In this configuration, Zn2+ ions coordinated by branched 2-hexyldecanoate ligands bind to surface chalcogenide (S2-) sites, preserving electronic passivation while forming a more sterically bulky and loosely packed ligand shell. Compared to the linear C18 oleate, the secondary-alkyl branching of 2-hexyldecanoate disrupts efficient chain packing, reduces inter-ligand van der Waals interactions, and increases conformational entropy at the nanocrystal interface. These properties have been linked to significantly enhanced solubility, reduced aggregation, and improved film formation, as demonstrated by Yang et al., who described such ligands as “entropic” in their behavior.81 Based on this, we hypothesized that the branched-ligand shell would facilitate lipid penetration and improve compatibility with the plasma membrane. To test this, SQW-loaded vesicles were prepared using 5 mol% DSPEPEG2000 – an optimized formulation previously shown to support rapid and efficient fusion in DiI-labeled system – and evaluated for their ability to achieve uniform membrane staining. The staining performance of System VI is presented in Figure 3.56. Panels A and B show the nanocrystal architecture and optical characterization, with the emission peak centered at approximately 495 nm. The lower panels display representative fluorescence and corresponding bright-field images acquired after a 30-minute incubation with HEK293 cells. Each column depicts a distinct field of view from separate samples, including two independent liposome batches prepared and imaged on different days. The consistency across these measurements confirms the reproducibility of the observed outcome. In all samples, bright, continuous fluorescence clearly delineates the cell perimeter, indicating efficient delivery of SQWs to the plasma membrane via vesicle fusion (Figure 3.56). In each of the three analyzed regions, the fluorescence signal remains consistently elevated above the unstained control baseline (yellow trace; Appendix, Figure A.4) across nearly the full scan length, dipping only at locations corresponding to spaces between neighboring cells. The profiles exhibit broad, gradually rising and falling signal plateaus, in contrast to the narrow, irregular spikes characteristic of Systems III and V, confirming a uniform distribution of SQWs along the plasma membrane. Signal intensity and pattern are highly consistent across Areas #1, #2, and #3, underscoring the reproducibility of the staining achieved with independently prepared liposome batches on three independent samples. The corresponding bright-field images confirm the integrity and confluency of the HEK293 monolayer, indicating that both the fusogenic vesicle formulation and the modified SQWs are well-tolerated by the cells. Chapter 3. Quantum Dots for Detecting Neuronal Activity 151 System VI yielded visibly stronger fluorescence and markedly higher membrane-to-background contrast. Notably, it supported shorter exposure times (as low as 30 ms), enabling frame rates up to ~33 Hz – beyond the capability of System IV under the same conditions. This represents an important step toward high temporal resolution, real-time imaging such as neuronal activity imaging, where temporal resolution is critical. Still, the achieved brightness remains below the potential of quantum dot-based probes, suggesting further optimization of ligand chemistry and nanocrystal architecture is needed to fully realize their performance in fast, high-resolution cellular imaging. Chapter 3. Quantum Dots for Detecting Neuronal Activity 152 Figure 3.56 Representative membrane staining results for System VI: CdS/CdSe/CdS SQWs capped with branched 2-hexyldecanoate ligands. (A) Schematic of SQW structure; vesicles contained DMPC 75%, DOTAP 25%, DSPE-PEG2000 5%. (B) UV-Vis (dashed blue) and PL (solid purple) spectra; λem = 495 nm. (C-H) Fluorescence and bright-field images of HEK293 cells after 30 min incubation show strong, continuous membrane labeling. (I-K) Intensity profiles from Areas #1-3 confirm uniform fluorescence above baseline, supporting efficient and reproducible delivery to the plasma membrane. A B - n C C e C 2 e l ecanoate C 2 100 m 100 m 100 m 100 m 100 m 100 m Bright Field Emission Bright Field Bright Field Emission Emission C D E F G H I J K APPENDIX 255 Emission was detected in a single channel via PMT 2, configured at 750 V gain, with 8-bit digitization and a linear LUT (gamma = 1.0). The image acquisition was performed at a zoom factor of 2.00, resulting in a final sampling resolution of 0.1217 µm per pixel in the lateral (X-Y) plane across a 512 × 512 pixel raster, yielding a total field of view of 62.29 µm × 62.29 µm. The acquisition was limited to a single optical plane, with Z-step size set to zero and the axial scanner deactivated. These parameters ensured Nyquist-sufficient lateral sampling and reliable optical sectioning for subsequent intensity-based image quantification. 3.2.4. Fluorescence Microscope The fluorescence microscope setup used in this study was identical to the configuration described in the supplementary materials for Chapter 2. All imaging parameters, optical components, and acquisition settings remained consistent unless otherwise noted in the main text. APPENDIX 256 3.3. Additional Figures Stability Study - Fluorescence Lifetime Traces Shown below are the original fluorescence decay curves recorded for vesicles extruded through 100 nm, 400 nm, and 800 nm membranes on days 1, 2, and 5 (Figure A.102). These traces were used for triple exponential fitting, with averaged lifetime values presented in the main text (Figure 3.22, Chapter 3, Section 2.2). Figure A.102 Raw fluorescence lifetime decay curves for vesicles of different sizes—100 nm (A, D, G), 400 nm (B, E, H), and 800 nm (C, F, I) - measured on Day 1 (A-C), Day 2 (D-F), and Day 5 (G-I). The overlaid yellow lines represent the fitted triple exponential decays. A B C D E F G H I APPENDIX 257 Optical characterization of SQWs systems used for lipid encapsulation As discussed in Chapter 3, Section 2.2, photoluminescence quantum yield (PLQY) measurements were performed for selected quantum well systems in the presence of lipid vesicles. Shown below are the corresponding absorption and emission spectra for all synthesized systems. Figure A.103 Normalized absorption (dashed blue lines) and photoluminescence (solid purple lines) spectra of the investigated systems: (A) System I – CdSe/ZnS@TOPO, (B) System II – SQWs@Cd-oleate, (C) System III – SQWs@Zn-oleate, (D) System IV – SQWs@Cd-F2 and octylamine, (E) System V – SQWs@ZnS@Zn-oleate, and (F) System VI – SQWs@branched ligand. Emission maxima (λem) are indicated in each panel. AB CD EF APPENDIX 258 Membrane staining experiments – control sample The image below shows the non-stained HEK293 cell sample used as a baseline reference in membrane staining experiments discussed in Chapter 3, Section 3.2. This control was imaged under identical conditions as stained samples and served as a background reference for the intensity profiles shown in Figures 3.36-3.41. Figure A.104 Bright-field (A) and GFP channel (B) images of non-stained HEK293 cells used as a control. Panel (C) shows the fluorescence intensity profile extracted along the indicated region in the GFP channel image. The trace represents baseline autofluorescence in the absence of any membrane dye or SQW staining. Spiking HEK Cells Recordings – Traces Comparison To additionally support the results presented in Chapter 3, Section 3.2, representative fluorescence traces are shown for HEK cells stained with SQWs (System VI), for cells treated with TTX, and for a non-stained control (baseline). These traces serve as the basis for evaluating signal dynamics. Raw baseline-corrected and filtered traces are presented both in offset view (Figure A.105) and overlaid for direct comparison (Figure A.106). 100 m100 m Bright Field GFP channel ABC APPENDIX 259 Figure A.105 Representative fluorescence traces for SQW-labeled HEK cells, TTX-treated cells, and nonstained control. (A) Baseline-corrected raw traces.(B) Filtered traces (bandpass), shown with vertical offset. Figure A.106 Same traces as above overlaid for direct comparison. (A) Baseline-corrected raw traces. (B) Bandpass-filtered traces, highlighting differences between spiking, TTX-treated, and control conditions. A B Comparison plots A B APPENDIX 260 Trace averaging analysis To assess the presence of periodic activity in the recorded fluorescence signals, a trial-averaging approach was applied. Based on the dominant frequency detected in the calcium signal, each trace was divided into equal-length segments corresponding to the expected spike period. In the absence of an external synchronization trigger, the segmentation start point was varied using a series of offsets. For each offset value, the trace was divided accordingly, and the resulting segments were averaged. This sliding-window approach allowed identification of the most favorable alignment for periodic features. Consistent, well-aligned transients were observed in calcium recordings. In contrast, traces from SQW-labeled and non-stained control cells lacked periodic structure, and their averaged signals appeared random and indistinguishable. The analysis was performed on both raw (Figure A.107) and filtered traces (Figure A.108). APPENDIX 261 Figure A.107 Trace averaging for different offsets applied to raw (baseline-corrected, unfiltered) recordings. (A) Calcium trace shows consistent spike-aligned transients across offsets. (B) SQW-labeled signal appears noisy and non-periodic. (C) Non-stained control resembles SQW trace, indicating lack of structured signal. Yellow lines indicate averaged response; grey lines represent individual segments. A B C APPENDIX 262 Figure A.108 Trace averaging for different offsets applied to processed (baseline-corrected and filtered) recordings. (A) Calcium trace shows robust, well-aligned periodic signal across offsets. (B) Non-stained control and (C) SQW-labeled traces remain inconsistent, with averaged signals lacking clear periodic features. Yellow lines indicate averaged response; grey lines represent individual segments. A B C APPENDIX 263 4. Supplementary Material for Chapter #4 4.1. Materials Gold (III) chloride trihydrate (HAuCl4 ∙ 3H2O), trisodium citrate dihydrate (C6H5Na3O7 ∙ 2H2O) were purchased from Sigma Aldrich. Hexadecyltrimethylammonium bromide (CTAB) and hexadecyltrimethylammonium chloride (CTAC, 25 wt% solution), sodium hypochlorite solution (NaClO), sodium borohydride (NaBH4), sodium hydroxide (NaOH), silver nitrate (AgNO3) citric acid, ascorbic acid, hydrochloric acid solution (HCl) were purchased form Sigma Aldrich and use d without further purification. Avidin, NeutrAvidin, NeutrAvidin-FITC conjugate and NeutrAvidin-DyLight™ 633 conjugate along with EZ-Link™ NHS-PEG12-Biotin (No-Weigh™ Format) were purchased from ThermoFisher Scientific (Molecular Probes™). Cell culture reagents, including DMEM/F-12, Neurobasal Plus, B-27 Plus supplements, GlutaMAX, B-27, Penicillin-Streptomycin (10,000 U/mL), poly-D-lysine, and Geltrex™ were purchased from Gibco. Texas Red™-phalloidin (by Invitrogen) and DAPI used for staining were purchased from Thermo Fisher Scientific. 4.2. Methods 4.2.1. Protocols Turkevich Synthesis of small Au NPs@citrate Glassware was freshly cleaned with aqua regia (3:1 v/v HCl : HNO3), rinsed with ultrapure water (18.2 MΩ cm) and dried at 120°C. An aqueous solution of chloroauric acid (HAuCl4·3H2O, 500 mL, 0.50 mM Au) was transferred to a 1 L flask equipped with a magnetic stirring bar. The suspension was heated to the boiling point of the solution (≈100 °C) while stirring at 700 rpm. In parallel, a 1% w/v trisodium citrate dihydrate solution (25 mL, 38.8 mM) was pre-heated to ~ 95°C. Once the gold solution reached boiling point, the hot citrate solution was introduced in one quick shot. Heating was maintained for 15 min; during this period the colour evolved from colourless to dim bright blue/purple shade to wine-red. After 15 min the heating plate was turned off and the flask was allowed to cool to room temperature under continued stirring. The final dispersion volume was 525 mL, giving an Au0 concentration of 0.48 mM and a citrate: Au molar ratio of 5:1. Samples were stored in vials at 4°C. APPENDIX 264 Tailored size 30 nm spherical Au NPs@CTAB Gold nanospheres (30 nm) were synthesized via a three-step seed-mediated growth method. In step one, 4.7 mL of 0.1 M CTAB was preheated at 35°C for 5 min, followed by addition of 25 µL of 0.05 M HAuCl4. After another 5 min incubation at 35°C, 0.3 mL of freshly prepared 0.01 M ice-cold NaBH4 was injected in one shot under vigorous magnetic stirring. Stirring continued for 2 min, followed by incubation for 30 min at 27°C. The seed solution (Au0 ≈ 0.25 mM) should have an absorbance of 0.6 at 400 nm. Immediately, 72 mL of 0.1 M CTAC was mixed with 360 µL of 0.05 M HAuCl4 and stirred mildly at room temperature for 10 min. Then, 27 mL of 0.1 M ascorbic acid was rapidly added, followed by immediate injection of 0.6 mL seed solution. The reaction proceeded for 60 min under mild stirring at room temperature, yielding 12 nm Au NPs (Au0 ≈ 0.225 mM, A400 ≈ 0.54). The product was centrifuged at 8500 rpm for 10 min and redispersed in 5 mL water. The gold concentration was adjusted to 2.8 mM. For the final growth, 0.518 mL of the 12 nm seed solution was added to 100 mL of 15 mM BDAC pre-incubated at 35°C for 10 min. Then 0.5 mL of 0.1 M ascorbic acid was injected, followed by 0.5 mL of 0.05 M HAuCl4. The solution was stirred for 30 min at 35°C. The resulting 30 nm Au NPs had Au0 ≈ 0.26 mM and A400 ≈ 0.624. Oxidative shape correction was performed by adding 360 µL of 1% v/v NaClO dropwise and stirring 10 min at 35 °C. Then, 100 µL of 0.05 M HAuCl4 was added dropwise, followed by a 30 min incubation at 35°C. The nanoparticles were centrifuged twice at 14 500 rpm for 10 min each and redispersed in 10 mM CTAB to reach a final gold concentration of 0.4 mM. Seed-Mediated Synthesis of Gold Bipyramids Gold seeds were prepared by mixing 10.0 mL of 50 mM CTAC, 0.05 mL of 50 mM HAuCl4 and 0.05 mL of 1 M citric acid in a 20 mL glass vial and stirring at 20°C. Freshly prepared 0.25 mL of 25 mM NaBH4 was then injected under vigorous stirring, resulting in an immediate color change from yellow to brownish; the absorbance at 400 nm was recorded (A400 ≈ 0.6 for a 0.25 mM gold concentration) to ensure complete reduction. The vial was sealed and the seed solution aged by heating at 80 °C with gentle stirring for 90 min, during which the color evolved from brown to red. For bipyramid growth, 100 mL of 100 mM CTAB was heated to 30°C and sequentially mixed with 5.0 mL of 10 mM HAuCl4, 1.0 mL of 10 mM AgNO3, 2.0 mL of 1 M HCl and 0.8 mL of 100 mM ascorbic acid. Under vigorous stirring at 30°C, 1.1 mL of the aged seed solution was rapidly injected and the mixture was left undisturbed at 30 °C for 2 h. The resulting gold bipyramids APPENDIX 271 Figure A.112 Confocal imaging of primary cortical (CTX) neurons incubated with FITC-labeled gold bipyramids (BPs@NeutrAv*FITC) at various time points to monitor nanoparticle internalization. Each row corresponds to a different incubation time: 0.5 h, 1 h, 6 h, and 24 h, with a control sample (no BPs) at the bottom. Columns show: (left) green fluorescence channel (FITC), (middle) merged fluorescence and brightfield, and (right) bright-field alone. o additional nuclear or cytoskeletal stains were applied for these observations. h 1h 2 h h Control o 50 m50 m50 m 50 m50 m50 m 50 m50 m50 m 50 m50 m50 m 50 m50 m50 m APPENDIX 272 5. Data and Code Repository All datasets, analysis scripts, and supplementary files associated with this thesis are available in an online repository at https://saco.csic.es/s/WoC52KNgGsWw6WC. The repository contains raw and processed experimental data, code used for statistical analyses and figure generation, and additional documentation that supports the results presented in the main text. The structure of the repository is organized by chapters, enabling direct access to the underlying data and analysis pipeline. QR code to the repository: 6. Documentation of Digital Tools 6.1. Data Analysis Data analysis and visualization were carried out using Python code executed in JupyterLab, OriginLab software, and ImageJ/Fiji for microscopy image processing. Custom scripts are included in the accompanying online repository to ensure reproducibility of the results. 6.2. Figure Preparation Unless otherwise indicated, all figures and graphical elements included in this thesis were prepared by the author. Figure assembly was performed using Microsoft PowerPoint. Individual elements were created or processed using the following software tools (in alphabetical order): BioRender, ChemDraw, ChatGPT (OpenAI), GIMP, Inkscape, Jmol, and Pexels. 6.3. Assisted Editing and Literature Search Software provided by OpenAI (ChatGPT models) was used to support phrasing, language corrections, and general writing assistance. Perplexity and NotebookLM were used as search engines to identify and organize relevant sources. Zotero was used for managing references and citation formatting. All scientific content, data interpretation, and conclusions were developed by the author. 273 Acknowledgments I would like to acknowledge the opportunities and resources provided by my supervisors, Marek Grzelczak and Rafael Yuste, which made this research possible. I am also grateful to Jonathan Owen and Mónica Carril for welcoming me into their laboratories and for their involvement in the project. Access to these research environments shaped both the progress of this work and my scientific growth. These four years have been a demanding journey, but I have always considered myself very lucky. Wherever I go, I find extraordinary people who accompany me on my path. The starting point for my PhD was Donostia. A big shout-out to my AAA support group – Adam, Anish, and Alba. Your encouragement, guidance, and understanding meant a great deal to me. Your insights and advice helped shape the research, and even from a distance, you remained actively involved during the writing phase, helping me navigate moments of uncertainty. Thank you for everything. That gratitude extends to the fourth A – Ane Escobar. Your positive energy was a real support. Thank you for your biological insights, for recognizing the effort I was putting in, and for your encouraging and motivating feedback. Adam! Dziunia – Twoja obecność i wsparcie wybiega dużo dalej niż mury CFM. Dziękuję za każdą porcję gnocchi po trudnym dniu, wspólne wieczory filmowe, pływanie, górskie wycieczki, wspólne zajawki i memy – dużo memów. You “Make Life Harder Easier”. Alba – you are my Ecuadorian twin, and I am so glad we met. Thank you for all the help navigating both scientific and administrative challenges. Thank you for being my translator and my Spanish voice whenever needed. For listening, for understanding, and for bringing color – not only to my hair. Kateryna, Adam, Mattia, Paschalis – thank you for making our apartment to feel like HOME. The fact is that we always could find there exactly what we needed. Our house was many things (a hotel, a cinema, a restaurant, a bakery, an arena to fight insects) – but most importantly, it was a place I always felt comfortable coming back. Jorge – for your great, universal advice (“SHUMP, Zuzanna, SHUPM!”) and swimming adventures. August – for making “Bilbao for Vote and More” truly “More”. Lorenzo – for carbonara lesson and short, yet meaningful, presence. Jozef – for the good throws and your calm attitude. Đorđe – for crosswords and your Eastern-European sarcasm. ACKNOWLEDGMENTS 274 The thesis writing process was closely observed by my new office mates – Quimey, Asier, Victor, Rubén, Martin, and Juli. Thank you for the pastries, the venting chats, uplifting stories and for reminding me what day of the week it was. A big thank-you to Raúl, Anto, Sabine, Nathaniel, Caro, Phuong, Idoia, and everyone else who made Donosti a good place to be. Another important part of this journey took place in New York. Noah – thank you million for being the best adventure seeker. Thanks to you I started seeing NY in a different light and appreciated its perks. For all the escapades, quests, hidden gems, quizzes, laughs, late nights in the lab but also scientific guidance, discussions and sharing your scientific curiosity – thank you. Teresa – only you can truly understand how traveling back and forth could be the hardest, yet most meaningful part of this PhD experience. Thank you for being a great companion through it all. You are the person connecting both worlds – all happy moments or difficulties shared with you always felt well understood. Hakim – for your friendship. I value all your expertise, help with the setup and scientific input but also all the discussions that happened along. Thank you for the museums, the concerts, ballet performances, and for making the lab a better place. Justin – for all the fun you were introducing every day. For being positive even during tough times, finding joy regardless, for being “a mountain of a man”. Jojo – for being a moral compass, and reminding me what is truly important. For swimming, hiking, all the chats, rooftop hangouts and reality checks. For making me a part of your very colorful world. Boris – Thank you for being such a great scientific support, for all the suggestions and tips. Thank you for celebrating all my wins with me, and for patiently helping me understand the reasons behind the setbacks. Thank you for all the coffees, lunch breaks and late cinema evenings as a fellow A-lister. You brought fresh energy when the lab began to change. Thank you, Tzitzi and Victor H., for your patient guidance, especially at the beginning of my biological research. You both helped a lot when nothing seemed to make sense. Darik and Ellie – thank you for being great friends. Together with Catherin and Beans, you made me feel like part of the local life and community. To fully understand Washington Heights, you have to live it! We survived the Dumpster Fire, and it looks like we’re both still standing. It was a pleasure being a silent witness to your practical jokes on Noah. Ł a z lona i Miszon dzięki za wspaniałe towarzystwo na finiszu w Nowm Jorku. Dzięki Wam odkryłam nowe aspekty tego miasta i nowe pasje – bird watching w Central Parku to ostatnia rzecz jakiej się po sobie spodziewałam. Zadbaliście o to żebym nie zaginęła w odmętach laboratorium i dawaliście poczucie normalności, przypominając co jest ważne. Wspaniale było mieć Was za sąsiadów. Dziękuję! ACKNOWLEDGMENTS 275 A big thank-you to Bereket, Dan and Galder for your help with chemistry aspects, and to all members of the Owen Lab for welcoming me into your space and making me feel a part of it. To Bella, Samu, Alex, Wataru, Jesus, Tabata, Doug, Mei with the gang (Bill, Bobby and Mattie), and everyone else who made New York much more than “just a research stay” – thank you. Dziękuję całemu mojemu rozbudowanemu systemowi wsparcia, który pozostaje w mocy mimo upływu lat i dzielących nas kilometrów: Santorkowi, Marcie, Gregorowi z Dr ż ną, Zofce, Anecie, rz ś o i i Piotrowi, Multipolkom i Brudasom+ (Adamowi, Zetasowi, Andrzejowi, Puchatowi i Koziarowi – wraz ze wszystkimi „plusami”). And last but not least, I want to thank my amazing family, for whom I am always enough. To my wonderful and supportive parents Katarzyna and Leszek, my siblings Agata and Wiktor. And to Paschalis, for being my greatest support on day-to-day basis. Without you, writing this thesis would be so much harder. You helped me to keep my focus while ensuring proper rest. Thank you for grounding me, for all your care, and your loving support. DZIĘKUJĘ!