Defects, which are commonly in metal organic frameworks (MOFs), are closely related to the performance of materials in various applications. Unlike other MOFs where metal ion clusters are usually 4-, 5-, or 6-connected to the organic linkers, one secondary building unit (SBU) of UiO-66 is coordinated by 12 Zr 6 clusters via 12 benzen-1,4-dicarboxlate (BDC) linkers. Therefore, the integrity of the structure can be well maintained after linker or even cluster missing. So far, many methods have been reported on the defect engineering of UiO-66 including adjusting the synthesis conditions (temperature, Zr/linker ratio and choice of Zr precursor), addition of modulators, thermal activation/dehydration, linker modification and metal cation substitution. Various techniques have been used and developed to characterize the existence and concentration of defects, though each technique has its limitations. The formation of defects not only changes the pore structure, but also brings beneficial changes in thermal, electronic, catalytic and adsorbing abilities; thus improved performance can be achieved when defective UiO-66 is used as Lewis and/or Bronsted acids, photocatalysts, adsorbents, electrodes or porous support. In this review, a comprehensive review of defect engineering for UiO-66 including their preparations, characterizations, applications, and then the challenges and outlook are discussed, aiming to provide some designing knowledge for the synthesis of defective UiO-66 with high-performance and promote the wide application of UiO-66 in various fields.

No abstract

No abstract

In this report, we summarize the takeaways from the first NeurIPS 2021 NetHack Challenge. Participants were tasked with developing a program or agent that can win (i.e., 'ascend' in) the popular dungeon-crawler game of NetHack by interacting with the NetHack Learning Environment (NLE), a scalable, procedurally generated, and challenging Gym environment for reinforcement learning (RL). The challenge showcased communitydriven progress in AI with many diverse approaches significantly beating the previously best results on NetHack. Furthermore, it served as a direct comparison between neural (e.g., deep RL) and symbolic AI, as well as hybrid systems, demonstrating that on NetHack symbolic bots currently outperform deep RL by a large margin. Lastly, no agent got close to winning the game, illustrating NetHack's suitability as a long-term benchmark for AI research.

Reinforcement learning (RL) agents are particularly hard to train when rewards are sparse. One common solution is to use intrinsic rewards to encourage agents to explore their environment. However, recent intrinsic exploration methods often use state-based novelty measures which reward low-level exploration and may not scale to domains requiring more abstract skills. Instead, we explore natural language as a general medium for highlighting relevant abstractions in an environment. Unlike previous work, we evaluate whether language can improve over existing exploration methods by directly extending (and comparing to) competitive intrinsic exploration baselines: AMIGo (Campero et al., 2021) and Nov-elD (Zhang et al., 2021). These language-based variants outperform their non-linguistic forms by 45-85% across 13 challenging tasks from the Min-iGrid and MiniHack environment suites.