A biomimetic membrane allows lithium ion separation by electrodialysis

http://english.cas.cn/newsroom/research_news/chem/202504/t20250427_1042154.shtml

https://www.nature.com/articles/s41467-025-59188-1

A research team led by Prof. GAO Jun from the CAS Qingdao Institute of Bioenergy and Bioprocess Technology (QIBEBT) , in collaboration with researchers from Qingdao University, has developed an innovative membrane that mimics biological ion channels to achieve highly selective lithium ion separation from complex brines. Lithium, which is essential for batteries and clean energy technologies, is often found in low concentrations alongside high levels of sodium, potassium, magnesium, and calcium ions.

Inspired by biological ion channels, the team designed a sulfonic acid-functionalized covalent organic framework (COF)—r-TpPa-SO₃H. The membrane’s randomly oriented nanocrystalline structure creates ultra-narrow, winding channels that can differentiate ions based on size and hydration energy. This unique structure enables an unconventional reverse-sieving mechanism that allows the selective passage of Na⁺, K⁺, and even divalent ions like Mg²⁺ and Ca²⁺ under an electric field while effectively blocking hydrated Li⁺ ions.

In laboratory tests, the membrane demonstrated remarkable Na⁺/Li⁺ and K⁺/Li⁺ selectivity, comparable to that of biological ion channels. Its performance remained stable in complex solutions, including real salt-lake brines. Under electrodialysis conditions, the membrane consistently removed major interfering ions, resulting in a lithium-enriched solution ready for downstream processing.

more insights

A humanoid robot training center in Qingdao improves robot efficiency

https://en.people.cn/n3/2026/0324/c90000-20439477.html

At the Qingdao Humanoid Robot Data Training Center in Laoshan District of Qingdao, humanoid robots are trained for jobs such as intelligent industrial manufacturing, smart home, and commercial services. Data collectors here control robots to complete specific tasks like logistics sorting, supermarket restocking, kitchen operations, and component assembly. Through thousands of repetitions and trials, massive amounts of motion data are generated, endowing robots with a smarter “intelligent brain,” and helping humanoid robots enter all walks of life to serve thousands of households.

RamEx: high-throughput microbial ramanome analyses with accurate quality assessment

https://www.cas.cn/syky/202602/t20260226_5102870.shtml

https://doi.org/10.1186/s40168-026-02339-3

A research team at the CAS Qingdao Institute of Bioenergy and Bioprocess Technology has developed RamEx, an integrated analysis framework for Ramanome big data. This platform, tailored to the characteristics of Raman spectroscopy data, establishes a one-stop workflow from data reading and standardized preprocessing to downstream data mining, centered on automated quality control algorithms and efficient parallel computing processes. It also demonstrates a systematic analysis of microbial metabolomical heterogeneity and metabolic pattern differentiation at the single-cell level.

Raman genomics deep analysis can track the dynamic changes in the composition of macromolecules such as lipids, proteins, and nucleic acids in different cells, thus revealing the differentiation and succession patterns of microbial metabolic states at the population scale with single-cell precision. This provides new research ideas and technical pathways for understanding the functional organization and environmental adaptation mechanisms of complex communities.

RamEx removes spectral noise in Raman-based microbial cell sorting

http://english.cas.cn/newsroom/research-news/202602/t20260224_1151116.shtml

https://link.springer.com/article/10.1186/s40168-026-02339-3

Scientists from the CAS Qingdao Institute of Bioenergy and Bioprocess Technology have developed a novel computational tool, RamEx, designed to resolve the computational bottleneck in high-throughput microbial Ramanomics.

RamEx streamlines the full Ramanomic analysis pipeline, from data preprocessing and automated quality control to advanced data mining. An Iterative Convolutional Outlier Detection (ICOD) algorithm tackles spectral noise in an unsupervised manner to dynamically identify and eliminate spectral artifacts, ensuring high-quality input for downstream analysis.

The platform’s performance was validated using diverse datasets, including pathogenic bacteria, probiotics, and yeast fermentation systems. Notably, RamEx successfully captured phenotypic heterogeneity in genetically identical yeast cells by detecting subtle metabolic fluctuations and tracking the dynamic accumulation of intracellular macromolecules, including lipids, proteins, and nucleic acids.

Back to …

青岛 Qingdao
a hub for S&T