Stefan Zauscher

The rational design of mesoscale superlattices with prescribed architectures remains a fundamental challenge in materials science. While self-assembly of patchy colloidal particles offers a powerful route to such structures, the combinatorial complexity linking patch number, placement, and interaction properties to emergent assembly architecture has hindered systematic inverse design. Here, we present a discrete neural-adjoint framework that enables direct optimization of building-block design for target superlattice formation. Leveraging a Gumbel–Softmax relaxation, our approach renders discrete patch configurations differentiable, allowing efficient gradient-based exploration of an otherwise intractable design space. Applied to a model system of self-assembling two-dimensional tiles, our method discovers patch designs that robustly assemble into complex superlattices, including multiple Archimedean tilings as well as previously unrealized Pythagorean and gyrated truncated hexagonal tilings. These results establish differentiable inverse design as a powerful approach for programming self-assembly, opening new avenues for the discovery of architected materials with tailored structure and function.

Machine learning (ML) is transforming materials research, yet potential for biopolymer discovery remains constrained by fragmented data and nonstandardized reporting. Biopolymers differ significantly from synthetic polymers, requiring specialized approaches to represent their biosynthetic origins, hierarchical structures, and application-specific metrics. In this Perspective, we identify three core challenges limiting biopolymer representation: information encoding, data quality, and data sharing. We describe the most pressing issues and propose commensurate approaches to address each key challenge. Recommendations include the design and adoption of biopolymer-specific fingerprinting and representation frameworks, development of hybrid human-large language model (LLM) data extraction strategies, and expanding Findable, Accessible, Interoperable, Reusable (FAIR)-compliant repositories. We propose a robust foundation to define interoperable, high-quality data sets that capture the full context of biopolymer materials. Standardized metadata, shared ontologies, and community-driven infrastructure would enable scalable, reproducible workflows and accelerate the ML-driven development of biopolymers.

Tailoring the properties of 2D materials through precise control of both local arrangement and long-range order on surfaces remains a central challenge in materials science. While surface-assisted assembly of symmetric, non-interacting DNA origami nanostructures offers a facile route to 2D superlattices, existing approaches typically yield close-packed arrays with limited control over inter-origami spacing. Here, we present a tunable and reproducible strategy for regulating the lateral spacing of DNA origami during self-assembly into macroscopic 2D lattice patterns. By controlling the growth of uniform single-stranded polynucleotide brushes from the surface of DNA origami, we modulate their effective geometry by introducing longer-ranged entropic repulsion, which enables precise and adjustable control over inter-origami distances across macroscopic areas. Using integrated experiments and simulations, we demonstrate how systematic variation of brush length, surface adsorption strength, and brush density can lead to tunable surface patterns across different origami shapes. Overall, this straightforward approach advances the field of DNA-templated nanofabrication by providing highly programmable templates with precise spatial control. This platform offers a robust foundation for the future integration of functional nanomaterials and the development of organized nanostructures.

ABSTRACT Tailoring the properties of 2D materials through precise control of both local arrangement and long‐range order on surfaces remains a central challenge in materials science. While surface‐assisted assembly of symmetric, non‐interacting DNA origami nanostructures offers a facile route to 2D superlattices, existing approaches typically yield close‐packed arrays with limited control over inter‐origami spacing. Here, we present a tunable and reproducible strategy for regulating the lateral spacing of DNA origami during self‐assembly into macroscopic 2D lattice patterns. By controlling the growth of uniform single‐stranded polynucleotide brushes from the surface of DNA origami, we modulate their effective geometry by introducing longer‐ranged entropic repulsion, which enables precise and adjustable control over inter‐origami distances across macroscopic areas. Using integrated experiments and simulations, we demonstrate how systematic variation of brush length, surface adsorption strength, and brush density can lead to tunable surface patterns across different origami shapes. Overall, this straightforward approach advances the field of DNA‐templated nanofabrication by providing highly programmable templates with precise spatial control. This platform offers a robust foundation for the future integration of functional nanomaterials and the development of organized nanostructures.

The rational design of mesoscale superlattices with prescribed architectures remains a fundamental challenge in materials science. While self-assembly of patchy colloidal particles offers a powerful route to such structures, the combinatorial complexity linking patch number, placement, and interaction properties to emergent assembly architecture has hindered systematic inverse design. Here, we present a discrete neural-adjoint framework that enables direct optimization of building-block design for target superlattice formation. Leveraging a Gumbel–Softmax relaxation, our approach renders discrete patch configurations differentiable, allowing efficient gradient-based exploration of an otherwise intractable design space. Applied to a model system of self-assembling two-dimensional tiles, our method discovers patch designs that robustly assemble into complex superlattices, including multiple Archimedean tilings as well as previously unrealized Pythagorean and gyrated truncated hexagonal tilings. These results establish differentiable inverse design as a powerful approach for programming self-assembly, opening new avenues for the discovery of architected materials with tailored structure and function.

Understanding the mechanical properties of polymer nanocomposite materials is essential for industrial use. Particularly, the determination of the polymer modulus at the nanofiller–polymer interphase is important for optimizing the interfacial mechanical properties. Nanoindentation via Atomic Force Microscopy (AFM) is well-established for measuring the modulus of the interphase region with nanoscale spatial resolution. However, indentation into heterogeneous materials presents a confounding issue often referred to as the “substrate effect”, i.e., the structural stress field caused by the rigid body is convoluted with the actual modulus of the interphase region. While finite element analysis (FEA)-based methods can be used to deconvolute the interphase modulus from measured apparent modulus–distance profiles, the experimental validation of this method is still needed. Here, we provide this validation using AFM nanoindentation on a layered model composite that consists of three layers with different moduli to recapitulate the properties of the matrix, the filler, and the interphase of real polymer nanocomposites. By systematically varying the thickness of the “artificial” interphase layer and the AFM probe radius, we obtain modulus–distance profiles over a wide range of indentation conditions. We validate a method to deconvolute the substrate effect using an empirically derived master curve obtained from FEA analysis. Furthermore, we showed that the effect of the artificial interphase on modulus– distance profiles can be distinguished only if the interphase layer is thick enough compared to the contact radius of the probe. Finally, we established an innovative and quantitative framework to predict the interphase thickness from mechanical nanoindentation measurements and discussed the lower, practical limit for interphase thickness determination. In summary, we provide a broadly applicable method to extract interphase mechanical properties of multiphase soft materials and practical guidelines for choosing optimal characterization conditions.

Abstract Acoustic gravimetric biosensors attract attention due to their simplicity, robustness, and low cost. However, a prevailing challenge in these sensors is dissipation which manifests in a low quality factor ( Q -factor), which limits their sensitivity and accuracy. To mitigate dissipation of acoustic sensors in liquid environments we introduce an innovative approach in which we combine microfluidic channels with gravimetric sensors. To implement this novel paradigm we chose the quartz crystal microbalance (QCM) as our model system, owing to its wide applicability in biosensing and the relevance of its operating principles to other types of acoustic sensors. We postulate that the crucial determinant for enhancing performance lies in the ratio between the width of the microfluidic channels and the wavelength of the pressure wave generated by the oscillating channel side walls driven by the QCM. Our hypothesis is supported by finite element analysis (FEA) and dimensional studies, which revealed two key factors that affect device performance: (1) the ratio of the channel width to the pressure wavelength ( $$W/{\lambda }_{{\rm {p}}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>W</mml:mi> <mml:mo>/</mml:mo> <mml:msub> <mml:mrow> <mml:mi>λ</mml:mi> </mml:mrow> <mml:mrow> <mml:mi>p</mml:mi> </mml:mrow> </mml:msub> </mml:mrow> </mml:math> ) and (2) the ratio of the channel height to the shear evanescent wavelength ( $$H/{\lambda }_{{\rm {s}}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>H</mml:mi> <mml:mo>/</mml:mo> <mml:msub> <mml:mrow> <mml:mi>λ</mml:mi> </mml:mrow> <mml:mrow> <mml:mi>s</mml:mi> </mml:mrow> </mml:msub> </mml:mrow> </mml:math> ). To validate our hypothesis, we fabricated a microfluidic QCM (µ-QCM) and demonstrated a remarkable 10-fold improvement in its dissipation when compared to conventional QCM. The novel microfluidic approach offers several additional advantages, such as direct data interpretation, reduced volume requirement for sample liquids, and simplified temperature control, augmenting the sensor’s overall performance. By fostering increased sensitivity, accuracy, and ease of operation, our novel paradigm unlocks new possibilities for advancing gravimetric technologies, potentially for biosensing applications.

Understanding the mechanical properties of polymer nanocomposite materials is essential for industrial use. Particularly, the determination of the polymer modulus at the nanofiller-polymer interphase is important for optimizing the interfacial mechanical properties. Nanoindentation via Atomic Force Microscopy (AFM) is well established for measuring the modulus of the interphase region with nanoscale spatial resolution. However, indentation into heterogeneous materials presents a confounding issue often referred to as the "substrate effect", i.e., the structural stress field caused by the rigid body is convoluted with the actual modulus of the interphase region. While finite element analysis (FEA)-based methods can be used to deconvolute the interphase modulus from measured apparent modulus-distance profiles, the experimental validation of this method is still needed. Here, we provide this validation using AFM nanoindentation on a layered model composite which consists of three layers with different moduli to recapitulate the properties of the matrix, the filler, and the interphase of real polymer nanocomposites. By systematically varying the thickness of the “artificial” interphase layer and the AFM probe radius, we obtain modulus - distance profiles over a wide range of indentation conditions. We validate a method to deconvolute the substrate effect using an empirically derived master curve obtained from FEA analysis. Furthermore, we showed that the effect of the artificial interphase on modulus - distance profiles can be distinguished only if the interphase layer is thick enough compared to the contact radius of the probe. Finally, we established an innovative and quantitative framework to predict the interphase thickness from mechanical nanoindentation measurements, and we discussed the lower, practical limit for interphase thickness determination. In summary, we provide a broadly applicable method to extract interphase mechanical properties of multiphase soft materials, and practical guidelines for choosing optimal characterization conditions.

Modern materials design strategies take advantage of the increasing amount of materials property data available and increasingly complex algorithms to take advantage of that data. However, viscoelastic materials resist this trend towards increased data rates due to the very material time-dependence that should be measured. Therefore, viscoelasticity measurements present a roadblock for data collection in an important aspect of material design. For thermorheologically simple materials, time-temperature superposition (TTS) has for many years provided a method to accelerate relaxation spectrum measurements relative to, for example, very long creep experiments. However, TTS itself currently faces a speed limit originating in the common logarithmic discrete frequency sweep (DFS) mode of operation. In DFS, measurement time is proportional (by a factor much greater than one) to the lowest frequency of measurement. This state of affairs has not improved for TTS for half a century or more. We utilize recent work in experimental rheometry on windowed chirps to collect three decades of complex modulus data simultaneously. Furthermore, the three-decade chirp response is recorded in the time it would take to collect only the lowest frequency data point of a traditional DFS experiment, resulting in a 500% speedup in data collection. In BOTTS, we superpose several isothermal chirp responses to produce a master curve in a fraction of the time of the traditional DFS-TTS technique. The chirp responses have good, albeit nontrivial, signal-to-noise properties. We use linear error propagation and a noise-weighted least squares approach to automatically incorporate all the data in a reliable shifting method. Using model thermoset polymers, we show DFS-TTS and BOTTS results are comparable, and therefore BOTTS data represent a first step towards a fast, reproducible method for master curve generation from existing rheological measurement instruments.

electrostatic interactions. We surmised that the electrostatic interactions between the BPL-modified HA molecules (HA-BPL) and the cartilage facilitate localization of the HA molecules to the cartilage surface. The number of BPL molecules on the HA backbone was varied to determine the optimal grafting density for cartilage binding and HA localization. Collectively, our results show that our HA-BPL molecules adhered readily to cartilage and were effective as a lubricant in cartilage-on-cartilage shear measurements where the modified HA molecules significantly reduce the coefficient of friction compared to phosphate-buffered saline or HA alone. This proof-of-concept study shows how the incorporation of cartilage adhering moieties, such as cationic molecules, can be used to enhance cartilage binding and lubrication properties of HA.

Modern materials design strategies take advantage of the increasing amount of materials property data available and increasingly complex algorithms to take advantage of those data. However, viscoelastic materials resist this trend towards increased data rates due to their inherent time-dependent properties. Therefore, viscoelasticity measurements present a roadblock for data collection in an important aspect of material design. For thermorheologically simple (TRS) materials, time-temperature superposition (TTS) made relaxation spectrum measurements faster relative to, for example, very long creep experiments. However, TTS itself currently faces a speed limit originating in the common logarithmic discrete frequency sweep (DFS) mode of operation. In DFS, the measurement time is proportional (by a factor much greater than one) to the lowest frequency of measurement. This state of affairs has not improved for TTS for half a century or more. We utilize recent work in experimental rheometry on windowed chirps to collect three decades of complex modulus data simultaneously, resulting in a ∼500% increase in data collection. In BOTTS, we superpose several isothermal chirp responses to produce a master curve in a fraction of time required by the traditional DFS-TTS technique. The chirp responses have good, albeit nontrivial, signal-to-noise properties. We use linear error propagation and a noise-weighted least squares approach to automatically incorporate all the data into a reliable shifting method. Using model thermoset polymers, we show that DFS-TTS and BOTTS results are comparable, and therefore BOTTS data represent a first step towards a faster method for master curve generation from unmodified rheological measurement instruments.

DNA origami has been used as biotemplates for growing a range of inorganic materials to create novel organic-inorganic hybrid nanomaterials. Recently, the solution-based silicification of DNA has been used to grow thin silica shells on DNA origami. However, the silicification reaction is sensitive to the reaction conditions and often results in uncontrolled DNA origami aggregation, especially when growth of thicker silica layers is desired. Here, we investigated how site-specifically placed polynucleotide brushes influence the silicification of DNA origami. Our experiments showed that long DNA brushes, in the form of single- or double-stranded DNA, significantly suppress the aggregation of DNA origami during the silicification process. Furthermore, we found that double-stranded DNA brushes selectively promote silica growth on DNA origami surfaces. These observations were supported and explained by coarse-grained molecular dynamics simulations. This work provides new insights into our understanding of the silicification process on DNA and provides a powerful toolset for the development of novel DNA-based organic-inorganic nanomaterials.

We report a targeted prodrug delivery platform that can deliver a cytostatic nucleobase analog with high drug loading. We chose fluorouracil (5FU), a drug used to treat various cancers, whose active metabolite 5-fluorodeoxyuridine monophosphate (5-FdUMP) is the antineoplastic agent. We use terminal deoxynucleotidyl transferase (TdT) to polymerize 5-fluorodeoxyuridine triphosphate (5-FdUTP) onto the 3'-end of an aptamer. We find that (i) addition of hydrophobic, unnatural nucleotides at the 3'-end of the 5-FdU polynucleotide by TdT leads to their spontaneous self-assembly into nuclease resistant micelles, (ii) aptamers presented on the micelle corona retain specificity for their cognate receptor on tumor cells, and (iii) the micelles deliver 5FU to tumor cells and exhibit greater cytotoxicity than the free drug. The modular design of our platform, consisting of a targeting moiety, a polynucleotide drug, and a self-assembly domain, can be adapted to encompass a range of polymerizable therapeutic nucleotides and targeting units.

Joint diseases, such as osteoarthritis, often require delivery of drugs to chondrocytes residing within the cartilage. However, intra-articular delivery of drugs to cartilage remains a challenge due to their rapid clearance within the joint. This problem is further exacerbated by the dense and negatively charged cartilage extracellular matrix (ECM). Cationic nanocarriers that form reversible electrostatic interactions with the anionic ECM can be an effective approach to overcome the electrostatic barrier presented by cartilage tissue. For an effective therapeutic outcome, the nanocarriers need to penetrate, accumulate, and be retained within the cartilage tissue. Nanocarriers that adhere quickly to cartilage tissue after intra-articular administration, transport through cartilage, and remain within its full thickness are crucial to the therapeutic outcome. To this end, we used ring-opening polymerization to synthesize branched poly(l-lysine) (BPL) cationic nanocarriers with varying numbers of poly(lysine) branches, surface charge, and functional groups, while maintaining similar hydrodynamic diameters. Our results show that the multivalent BPL molecules, including those that are highly branched (i.e., generation two), can readily adhere and transport through the full thickness of cartilage, healthy and degenerated, with prolonged intra-cartilage retention. Intra-articular injection of the BPL molecules in mouse knee joint explants and rat knee joints showed their localization and retention. In summary, this study describes an approach to design nanocarriers with varying charge and abundant functional groups while maintaining similar hydrodynamic diameters to aid the delivery of macromolecules to negatively charged tissues.

DNA nanotechnology provides an approach to create precise, tunable, and biocompatible nanostructures for biomedical applications. However, the stability of these structures is severely compromised in biological milieu due to their fast degradation by nucleases. Recently, we showed how enzymatic polymerization could be harnessed to grow polynucleotide brushes of tunable length and location on the surface of DNA origami nanostructures, which greatly enhances their nuclease stability. Here, we report on strategies that allow for both spatial and temporal control over polymerization through activatable initiation, cleavage, and regeneration of polynucleotide brushes using restriction enzymes. The ability to site-specifically decorate DNA origami nanostructures with polynucleotide brushes in a spatiotemporally controlled way provides access to "smart" functionalized DNA architectures with potential applications in drug delivery and supramolecular assembly.

Abstract DNA nanotechnology provides an approach to create precise, tunable, and biocompatible nanostructures for biomedical applications. However, the stability of these structures is severely compromised in biological milieu due to their fast degradation by nucleases. Recently, we showed how enzymatic polymerization could be harnessed to grow polynucleotide brushes of tunable length and location on the surface of DNA origami nanostructures, which greatly enhances their nuclease stability. Here, we report on strategies that allow for both spatial and temporal control over polymerization through activatable initiation, cleavage, and regeneration of polynucleotide brushes using restriction enzymes. The ability to site‐specifically decorate DNA origami nanostructures with polynucleotide brushes in a spatiotemporally controlled way provides access to “smart” functionalized DNA architectures with potential applications in drug delivery and supramolecular assembly.

Cells throughout the human body detect mechanical forces. While it is known that the rapid (millisecond) detection of mechanical forces is mediated by force-gated ion channels, a detailed quantitative understanding of cells as sensors of mechanical energy is still lacking. Here, we combine atomic force microscopy with patch-clamp electrophysiology to determine the physical limits of cells expressing the force-gated ion channels (FGICs) Piezo1, Piezo2, TREK1, and TRAAK. We find that, depending on the ion channel expressed, cells can function either as proportional or nonlinear transducers of mechanical energy and detect mechanical energies as little as ~100 fJ, with a resolution of up to ~1 fJ. These specific energetic values depend on cell size, channel density, and cytoskeletal architecture. We also make the surprising discovery that cells can transduce forces either nearly instantaneously (<1 ms) or with a substantial time delay (~10 ms). Using a chimeric experimental approach and simulations, we show how such delays can emerge from channel-intrinsic properties and the slow diffusion of tension in the membrane. Overall, our experiments reveal the capabilities and limits of cellular mechanosensing and provide insights into molecular mechanisms that different cell types may employ to specialize for their distinct physiological roles.