Search

A fully integrated near-microphone keyword spotting (KWS) chip is proposed to directly interact with a passive microphone and achieve submicrowatt power for the Internet of Things (IoT) devices. First, an on-chip analog frontend (AFE) is designed to avoid the inclusion of power-intensive off-chip active microphones. Second, a real-point serial fast Fourier transform (FFT)-based Mel-frequency cepstral coefficient (MFCC) feature extractor, cooperating with a genetic algorithm (GA) optimized bit-width quantization, is specifically customized to reduce the MFCC power by 67.4%. Finally, a binarized temporal depthwise separable CNN (TDSCN) is proposed, featuring hardware optimization through a parallel adder tree (PAT)-based PE with near-memory computing. This results in a 78.9% reduction in computation as compared to the traditional depthwise separable convolutional neural networks (CNNs). Fabricated in a 28-nm CMOS process, the proposed KWS chip consumes the lowest power of 0.61 $\mu \text{W}$ at 0.36-V neural network (NN), 0.9-V AFE, and 8-KHz frequency, while keeping 95.8% accuracy for two-KWS on Google speech command dataset (GSCD).

Developing a general, highly efficient, and enantioselective catalytic method for the synthesis of chiral alcohols is still a formidable challenge. We report in this article the asymmetric transfer hydrogenation (ATH) of N-methyliminodiacetyl (MIDA) acylboronates as a general substrate-independent entry to enantioenriched secondary alcohols. ATH of acyl-MIDA-boronates with (het)aryl, alkyl, alkynyl, alkenyl, and carbonyl substituents delivers a variety of enantioenriched α-boryl alcohols. The latter are used in a range of stereospecific transformations based on the boron moiety, enabling the synthesis of carbinols with two closely related α-substituents, which cannot be obtained with high enantioselectivities using direct asymmetric hydrogenation methods, such as the (R)-cloperastine intermediate. Computational studies illustrate that the BMIDA group is a privileged enantioselectivity-directing group in Noyori–Ikariya ATH compared to the conventionally used aryl and alkynyl groups due to the favorable CH–O attractive electrostatic interaction between the η6-arene-CH of the catalyst and the σ-bonded oxygen atoms in BMIDA. The work expands the domain of conventional ATH and shows its huge potential in addressing challenges in symmetric synthesis.

As well-known, the excessive discharge of heavy-metal mercury not only destroys the ecological environment, bust also leads to severe damage of human health after ingestion via drinking and bioaccumulation of food chains, and mercury ion (Hg2+) is designated as one of most prevalent toxic metal ions in drinking water. Thus, the high-performance monitoring of mercury pollution is necessary. Functional nucleic acids have been widely used as recognition probes in biochemical sensing. In this work, a carbazole derivative, ethyl-4-[3,6-bis(1-methyl-4-vinylpyridium iodine)-9H-carbazol −9-yl)] butanoate (EBCB), has been synthesized and found as a target-lighted DNA fluorescent indicator. As a proof-of-concept, Hg2+detection was carried out based on EBCB and Hg2+-mediated conformation transformation of a designed DNA probe. By comparison with conventional nucleic acid indicators, EBCB held excellent advantages, such as minimal background interference and maximal sensitivity. Outstanding detection capabilities were displayed, especially including simple operation (add-and-read manner), ultrarapidity (30 s), and low detection limit (0.82 nM). Furthermore, based on these advantages, the potential for high-performance screening of mercury antagonists was also demonstrated by the fluorescence change of EBCB. Therefore, we believe that this work is meaningful in pollution monitoring, environment restoration and emergency treatment, and may pave a way to apply EBCB as an ideal signal transducer for development of high-performance sensing strategies.