pH-Controlled enzymatic computing for digital circuits and neural networks.

Unconventional computing paradigms explore new methods for processing information beyond the capabilities of traditional electronic architectures. In this work, we present our approach to digital computation through enzymatic reactions in chemically buffered environments. A key aspect of this approach is its reliance on pH-sensitive enzymatic reactions, with the direction of the reaction controlled by maintaining pH levels within a specific range. When the pH crosses a defined threshold, the reaction moves forward and vice versa, akin to the switching action of electronic switches in digital circuits. The binary signals (0 and 1) are encoded as different concentrations of strong acids or bases, offering a bio-inspired method for computation. The final readout is done using UV-vis spectroscopy after applying detection reactions to indicate whether the output is 1 (indicated by the presence of the enzymatic reaction's product) or 0 (indicated by the absence of the enzymatic reaction's product). We build and evaluate a set of digital circuits in the lab using our proposed methodology to model the circuits using chemical reactions. In addition, we demonstrate the implementation of a neural network classifier using our framework.

CMOS-Integrated Low-Noise Junction Field-Effect Transistors for Bioelectronic Applications.

In this work, we present a CMOS-integrated low-noise junction field-effect transistor (JFET) developed in a standard 0.18 pm CMOS process. These JFETs reduce input-referred flicker noise power by more than a factor of 10 when compared to equally sized n-channel MOS devices by eliminating oxide interfaces in contact with the channel. We show that this improvement in device performance translates into a factor-of-10 reduction in the input-referred noise of integrated CMOS operational amplifiers when JFET devices are used at the input, significant for many applications in bioelectronics.

Measurement of DNA Translocation Dynamics in a Solid-State Nanopore at 100 ns Temporal Resolution.

Despite the potential for nanopores to be a platform for high-bandwidth study of single-molecule systems, ionic current measurements through nanopores have been limited in their temporal resolution by noise arising from poorly optimized measurement electronics and large parasitic capacitances in the nanopore membranes. Here, we present a complementary metal-oxide-semiconductor (CMOS) nanopore (CNP) amplifier capable of low noise recordings at an unprecedented 10 MHz bandwidth. When integrated with state-of-the-art solid-state nanopores in silicon nitride membranes, we achieve an SNR of greater than 10 for ssDNA translocations at a measurement bandwidth of 5 MHz, which represents the fastest ion current recordings through nanopores reported to date. We observe transient features in ssDNA translocation events that are as short as 200 ns, which are hidden even at bandwidths as high as 1 MHz. These features offer further insights into the translocation kinetics of molecules entering and exiting the pore. This platform highlights the advantages of high-bandwidth translocation measurements made possible by integrating nanopores and custom-designed electronics.

Statistical inference for nanopore sequencing with a biased random walk model.

Nanopore sequencing promises long read-lengths and single-molecule resolution, but the stochastic motion of the DNA molecule inside the pore is, as of this writing, a barrier to high accuracy reads. We develop a method of statistical inference that explicitly accounts for this error, and demonstrate that high accuracy (>99%) sequence inference is feasible even under highly diffusive motion by using a hidden Markov model to jointly analyze multiple stochastic reads. Using this model, we place bounds on achievable inference accuracy under a range of experimental parameters.

Integrated nanopore sensing platform with sub-microsecond temporal resolution.

Nanopore sensors have attracted considerable interest for high-throughput sensing of individual nucleic acids and proteins without the need for chemical labels or complex optics. A prevailing problem in nanopore applications is that the transport kinetics of single biomolecules are often faster than the measurement time resolution. Methods to slow down biomolecular transport can be troublesome and are at odds with the natural goal of high-throughput sensing. Here we introduce a low-noise measurement platform that integrates a complementary metal-oxide semiconductor (CMOS) preamplifier with solid-state nanopores in thin silicon nitride membranes. With this platform we achieved a signal-to-noise ratio exceeding five at a bandwidth of 1 MHz, which to our knowledge is the highest bandwidth nanopore recording to date. We demonstrate transient signals as brief as 1 µs from short DNA molecules as well as current signatures during molecular passage events that shed light on submolecular DNA configurations in small nanopores.

High-bandwidth protein analysis using solid-state nanopores.

High-bandwidth measurements of the ion current through hafnium oxide and silicon nitride nanopores allow the analysis of sub-30 kD protein molecules with unprecedented time resolution and detection efficiency. Measured capture rates suggest that at moderate transmembrane bias values, a substantial fraction of protein translocation events are detected. Our dwell-time resolution of 2.5 µs enables translocation time distributions to be fit to a first-passage time distribution derived from a 1D diffusion-drift model. The fits yield drift velocities that scale linearly with voltage, consistent with an electrophoretic process. Further, protein diffusion constants (D) are lower than the bulk diffusion constants (D0) by a factor of ~50, and are voltage-independent in the regime tested. We reason that deviations of D from D0 are a result of confinement-driven pore/protein interactions, previously observed in porous systems. A straightforward Kramers model for this inhibited diffusion points to 9- to 12-kJ/mol interactions of the proteins with the nanopore. Reduction of µ and D are found to be material-dependent. Comparison of current-blockage levels of each protein yields volumetric information for the two proteins that is in good agreement with dynamic light scattering measurements. Finally, detection of a protein-protein complex is achieved.

Single ion channel recordings with CMOS-anchored lipid membranes.

We present single-ion-channel recordings performed with biomimetic lipid membranes which are directly attached to the surface of a complementary metal-oxide-semiconductor (CMOS) preamplifier chip. With this system we resolve single-channel currents from several types of bacterial ion channels, including fluctuations of a single alamethicin channel at a bandwidth of 1 MHz which represent the fastest single-ion-channel recordings reported to date. The platform is also used for high-resolution α-hemolysin nanopore recordings. These results illustrate the high signal fidelity, fine temporal resolution, small geometry, and multiplexed integration which can be achieved by leveraging integrated semiconductor platforms for advanced ion channel interfaces.

Repurposing Waste Chemicals for Sustainable and Durable Molecular Data Storage.

Molecular data storage offers the intriguing possibility of higher theoretical density and longer lifetimes than today's electronic memory devices. Some demonstrations have used deoxyribonucleic acid (DNA), but bottlenecks in nucleic acid synthesis continue to make DNA data storage orders of magnitude more expensive than electronic storage media. Additionally, despite its potential for long-term storage, DNA faces durability challenges from environmental degradation. In this work, we demonstrate nongenomic molecular data storage using molecular libraries redirected from chemical waste streams. This approach requires no synthetic effort and can be implemented by using molecules that have a minimal associated cost. While the technique is agnostic about the exact molecular content of its inputs, we confirmed that some sources contained poly fluoroalkyl substances (PFAS), which persist for long periods in the natural environment and could offer extremely durable information storage as well as environmental benefits. These demonstrations provide a perspective on some of the valuable possibilities for nongenomic molecular information systems.

Electrical Capacitance Tomography of Cell Cultures on a CMOS Microelectrode Array.

Electrical capacitance tomography (ECT) can be used to predict information about the interior volume of an object based on measured capacitance at its boundaries. Here, we present a microscale capacitance tomography system with a spatial resolution of 10 microns using an active CMOS microelectrode array. We introduce a deep learning model for reconstructing 3-D volumes of cell cultures using the boundary capacitance measurements acquired from the sensor array, which is trained using a multi-objective loss function that combines a pixel-wise loss function, a distribution-based loss function, and a region-based loss function to improve model's reconstruction accuracy. The multi-objective loss function enhances the model's reconstruction accuracy by 3.2% compared to training only with a pixel-wise loss function. Compared to baseline computational methods, our model achieves an average of 4.6% improvement on the datasets evaluated. We demonstrate our approach on experimental datasets of bacterial biofilms, showcasing the system's ability to resolve microscopic spatial features of cell cultures in three dimensions. Microscale capacitance tomography can be a low-cost, low-power, label-free tool for 3-D imaging of biological samples.

Microscale 3-D Capacitance Tomography with a CMOS Sensor Array.

Electrical capacitance tomography (ECT) is a non-optical imaging technique in which a map of the interior permittivity of a volume is estimated by making capacitance measurements at its boundary and solving an inverse problem. While previous ECT demonstrations have often been at centimeter scales, ECT is not limited to macroscopic systems. In this paper, we demonstrate ECT imaging of polymer microspheres and bacterial biofilms using a CMOS microelectrode array, achieving spatial resolution of 10 microns. Additionally, we propose a deep learning architecture and an improved multi-objective training scheme for reconstructing out-of-plane permittivity maps from the sensor measurements. Experimental results show that the proposed approach is able to resolve microscopic 3-D structures, achieving 91.5% prediction accuracy on the microsphere dataset and 82.7% on the biofilm dataset, including an average of 4.6% improvement over baseline computational methods.

Digital circuits and neural networks based on acid-base chemistry implemented by robotic fluid handling.

Acid-base reactions are ubiquitous, easy to prepare, and execute without sophisticated equipment. Acids and bases are also inherently complementary and naturally map to a universal representation of "0" and "1." Here, we propose how to leverage acids, bases, and their reactions to encode binary information and perform information processing based upon the majority and negation operations. These operations form a functionally complete set that we use to implement more complex computations such as digital circuits and neural networks. We present the building blocks needed to build complete digital circuits using acids and bases for dual-rail encoding data values as complementary pairs, including a set of primitive logic functions that are widely applicable to molecular computation. We demonstrate how to implement neural network classifiers and some classes of digital circuits with acid-base reactions orchestrated by a robotic fluid handling device. We validate the neural network experimentally on a number of images with different formats, resulting in a perfect match to the in-silico classifier. Additionally, the simulation of our acid-base classifier matches the results of the in-silico classifier with approximately 99% similarity.

Super-Resolution Electrochemical Impedance Imaging With a 512 × 256 CMOS Sensor Array.

Super-resolution imaging is a family of techniques in which multiple lower-resolution images can be merged to produce a single image at higher resolution. While super-resolution is often applied to optical systems, it can also be used with other imaging modalities. Here we demonstrate a 512 × 256 CMOS sensor array for micro-scale super-resolution electrochemical impedance spectroscopy (SR-EIS) imaging. The system is implemented in standard 180 nm CMOS technology with a 10 µm × 10 µm pixel size. The sensor array is designed to measure the mutual capacitance between programmable sets of pixel pairs. Multiple spatially-resolved impedance images can then be computationally combined to generate a super-resolution impedance image. We use finite-element electrostatic simulations to support the proposed measurement approach and discuss straightforward algorithms for super-resolution image reconstruction. We present experimental measurements of sub-cellular permittivity distribution within single green algae cells, showing the sensor's capability to produce microscale impedance images with sub-pixel resolution.

Programmable Electrochemical Stimulation on a Large-Scale CMOS Microelectrode Array.

In this paper we present spatio-temporally controlled electrochemical stimulation of aqueous samples using an integrated CMOS microelectrode array with 131,072 pixels. We demonstrate programmable gold electrodeposition in arbitrary spatial patterns, controllable electrolysis to produce microscale hydrogen bubbles, and spatially targeted electrochemical pH modulation. Dense spatially-addressable electrochemical stimulation is important for a wide range of bioelectronics applications.

Secret messaging with endogenous chemistry.

Data encoded in molecules offers opportunities for secret messaging and extreme information density. Here, we explore how the same chemical and physical dimensions used to encode molecular information can expose molecular messages to detection and manipulation. To address these vulnerabilities, we write data using an object's pre-existing surface chemistry in ways that are indistinguishable from the original substrate. While it is simple to embed chemical information onto common objects (covers) using routine steganographic permutation, chemically embedded covers are found to be resistant to detection by sophisticated analytical tools. Using Turbo codes for efficient digital error correction, we demonstrate recovery of secret keys hidden in the pre-existing chemistry of American one dollar bills. These demonstrations highlight ways to improve security in other molecular domains, and show how the chemical fingerprints of common objects can be harnessed for data storage and communication.

Leveraging autocatalytic reactions for chemical domain image classification.

Autocatalysis is fundamental to many biological processes, and kinetic models of autocatalytic reactions have mathematical forms similar to activation functions used in artificial neural networks. Inspired by these similarities, we use an autocatalytic reaction, the copper-catalyzed azide-alkyne cycloaddition, to perform digital image recognition tasks. Images are encoded in the concentration of a catalyst across an array of liquid samples, and the classification is performed with a sequence of automated fluid transfers. The outputs of the operations are monitored using UV-vis spectroscopy. The growing interest in molecular information storage suggests that methods for computing in chemistry will become increasingly important for querying and manipulating molecular memory.

Principles of Information Storage in Small-Molecule Mixtures.

Molecular data systems have the potential to store information at dramatically higher density than existing electronic media. Some of the first experimental demonstrations of this idea have used DNA, but nature also uses a wide diversity of smaller non-polymeric molecules to preserve, process, and transmit information. In this paper, we present a general framework for quantifying chemical memory, which is not limited to polymers and extends to mixtures of molecules of all types. We show that the theoretical limit for molecular information is two orders of magnitude denser by mass than DNA, although this comes with different practical constraints on total capacity. We experimentally demonstrate kilobyte-scale information storage in mixtures of small synthetic molecules, and we consider some of the new perspectives that will be necessary to harness the information capacity available from the vast non-genomic chemical space.

Multicomponent molecular memory.

Multicomponent reactions enable the synthesis of large molecular libraries from relatively few inputs. This scalability has led to the broad adoption of these reactions by the pharmaceutical industry. Here, we employ the four-component Ugi reaction to demonstrate that multicomponent reactions can provide a basis for large-scale molecular data storage. Using this combinatorial chemistry we encode more than 1.8 million bits of art historical images, including a Cubist drawing by Picasso. Digital data is written using robotically synthesized libraries of Ugi products, and the files are read back using mass spectrometry. We combine sparse mixture mapping with supervised learning to achieve bit error rates as low as 0.11% for single reads, without library purification. In addition to improved scaling of non-biological molecular data storage, these demonstrations offer an information-centric perspective on the high-throughput synthesis and screening of small-molecule libraries.

Encoding information in synthetic metabolomes.

Biomolecular information systems offer exciting potential advantages and opportunities to complement conventional semiconductor technologies. Much attention has been paid to information-encoding polymers, but small molecules also play important roles in biochemical information systems. Downstream from DNA, the metabolome is an information-rich molecular system with diverse chemical dimensions which could be harnessed for information storage and processing. As a proof of principle of small-molecule postgenomic data storage, here we demonstrate a workflow for representing abstract data in synthetic mixtures of metabolites. Our approach leverages robotic liquid handling for writing digital information into chemical mixtures, and mass spectrometry for extracting the data. We present several kilobyte-scale image datasets stored in synthetic metabolomes, which can be decoded with accuracy exceeding 99% using multi-mass logistic regression. Cumulatively, >100,000 bits of digital image data was written into metabolomes. These early demonstrations provide insight into some of the benefits and limitations of small-molecule chemical information systems.

Quasi-reference electrodes in confined electrochemical cells can result in in situ production of metallic nanoparticles.

Nanoscale working electrodes and miniaturized electroanalytical devices are valuable platforms to probe molecular phenomena and perform chemical analyses. However, the inherent close distance of metallic electrodes integrated into a small volume of electrolyte can complicate classical electroanalytical techniques. In this study, we use a scanning nanopipette contact probe as a model miniaturized electrochemical cell to demonstrate measurable side effects of the reaction occurring at a quasi-reference electrode. We provide evidence for in situ generation of nanoparticles in the absence of any electroactive species and we critically analyze the origin, nucleation, dissolution and dynamic behavior of these nanoparticles as they appear at the working electrode. It is crucial to recognize the implications of using quasi-reference electrodes in confined electrochemical cells, in order to accurately interpret the results of nanoscale electrochemical experiments.

In Situ Nanopore Fabrication and Single-Molecule Sensing with Microscale Liquid Contacts.

In this article, we introduce a flexible technique for high-throughput solid-state nanopore analysis of single biomolecules. By confining the electrolyte to a micron-scale liquid meniscus at the tip of a glass micropipette, we enable automation and reuse of a single solid-state membrane chip for measurements with hundreds of distinct nanopores per day. In addition to overcoming important experimental bottlenecks, the microscale liquid contact dramatically reduces device capacitance, which is a key limiting factor to the speed and fidelity of solid-state nanopore sensor recordings.