---
格式版本: 2
标题: "PACE-SIMS: Checkpoint-Gated Autonomous SIMS Characterization with AI-Agent Quality Control"
原文链接: "https://arxiv.org/abs/2608.12277"
发布日期: "2026-08-12"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Wed, 12 Aug 2026 17:17:23 UTC (7,325 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-13T18:31:06+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-13T18:24:32+08:00"
入库时间: "2026-08-13T10:31:06.366Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=AI&searchtype=all"
匹配关键词:
  - "AI"
  - "delivery"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 0
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "资料为材料科学领域AI质控论文，与超节点、AI Rack、机柜级AI基础设施等核心主题完全无关，属明显无关。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-13T18:31:12+08:00"
AI主题相关性: 0
AI来源权威性: 0
AI新颖性: 0
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 0
AI摘要: "PACE-SIMS提出一种由AI代理自主执行飞行时间二次离子质谱(SIMS)表征的工作流，研究者仅设定科学问题与质量标准，代理在检查点自动判断、修正或重试。"
AI摘要模型: "ali-deepseek-v4-flash"
AI摘要时间: "2026-09-07T03:39:55.861Z"
采集批次: "2026年8月13日18点24分27秒"
采集批次ID: "20260813-182427-1564e572"
去重键: "https://arxiv.org/abs/2608.12277"
---

## Condensed Matter > Materials Science

## Title:PACE-SIMS: Checkpoint-Gated Autonomous SIMS Characterization with AI-Agent Quality Control

Authors:[Anton V Ievlev](https://arxiv.org/search/cond-mat?searchtype=author&query=Ievlev,+A+V), [Heather Hare](https://arxiv.org/search/cond-mat?searchtype=author&query=Hare,+H), [Yiyang Li](https://arxiv.org/search/cond-mat?searchtype=author&query=Li,+Y), [Sergei V Kalinin](https://arxiv.org/search/cond-mat?searchtype=author&query=Kalinin,+S+V)

[View PDF](https://arxiv.org/pdf/2608.12277)

> Abstract:Time-of-flight secondary ion mass spectrometry (ToF-SIMS) is widely used for local chemical investigations across a broad range of materials and systems. However, its operation is expensive in expert time: a trained researcher must supervise acquisition throughout, dynamically tuning parameters throughout the experiment, often across a campaign spanning multiple days. Here, we present PACE-SIMS, an agentic workflow which runs a SIMS study as a human-AI collaboration. Here, the researcher specifies the scientific questions and quality requirements, and an AI agent builds the plan and, after approval, executes it autonomously, pausing at checkpoints to judge each measurement and to correct, retry, or escalate. To validate the approach, we applied it to a study of chemical composition in 18O-enriched WOx films. During this blind randomized, two-polarity study (8.1 hours, 35 measurements) the agent made three unscripted corrections a fixed script would have missed, and all four predictions held against the sealed ground truth. The same run returned transferable measurement science, including a composition calibration, a 5.3% isotope-reading offset between the two ion polarities, and the deposition's tracer-delivery mechanism, from less than two hours of researcher attention. The developed agentic architecture is not specific to SIMS and can be applied to other analytical techniques, with primary target being the destructive measurements for which optimization methods are poorly applicable.

| Subjects: | Materials Science (cond-mat.mtrl-sci) |
| --- | --- |
| Cite as: | [arXiv:2608.12277](https://arxiv.org/abs/2608.12277) \[cond-mat.mtrl-sci\] |
|  | (or [arXiv:2608.12277v1](https://arxiv.org/abs/2608.12277v1) \[cond-mat.mtrl-sci\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.12277](https://doi.org/10.48550/arXiv.2608.12277) |

## Submission history

From: Anton Ievlev \[[view email](https://arxiv.org/show-email/65cc8c1f/2608.12277)\]  
**\[v1\]** Wed, 12 Aug 2026 17:17:23 UTC (7,325 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.12277) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))
