返回技能列表

interpret-ir-spectrum

pjt222
更新于 2 days ago
6 次查看
17
2
17
在 GitHub 上查看
其他general

关于

This skill systematically analyzes infrared spectra to identify functional groups in a sample. It interprets both the diagnostic (4000-1500 cm⁻¹) and fingerprint (1500-400 cm⁻¹) regions, accounting for hydrogen bonding effects. Developers can use it for initial compound screening, as it compiles a functional group inventory with confidence levels.

快速安装

Claude Code

推荐
主要方式
npx skills add pjt222/agent-almanac -a claude-code
插件命令备选方式
/plugin add https://github.com/pjt222/agent-almanac
Git 克隆备选方式
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/interpret-ir-spectrum

在 Claude Code 中复制并粘贴此命令以安装该技能

技能文档

IRスペクトルの解釈

Analyze infrared absorption spectra to identify functional groups, assess hydrogen bonding, and compile a comprehensive inventory of structural features present in the sample.

使用タイミング

  • Identifying functional groups in an unknown compound as a first screening step
  • Confirming the presence or absence of specific functional groups (e.g., verifying a reaction converted an alcohol to a ketone)
  • Monitoring reaction progress by tracking the appearance or disappearance of characteristic absorptions
  • Distinguishing between similar compounds that differ in functional group content
  • Complementing NMR and mass spectrometry data with vibrational information

入力

  • 必須: IR spectrum data (absorption frequencies in cm-1 with intensities, either as %Transmittance or Absorbance plot)
  • 必須: Sample preparation method (KBr pellet, ATR, Nujol mull, thin film, solution cell)
  • 任意: Molecular formula or expected compound class
  • 任意: Known structural fragments from other spectroscopic data
  • 任意: Instrument parameters (resolution, scan range, detector type)

手順

ステップ1: Establish Spectrum Quality and Format

Verify that the spectrum is suitable for interpretation before analyzing peaks:

  1. Check y-axis format: Determine whether the spectrum is plotted in %Transmittance (%T, peaks point down) or Absorbance (A, peaks point up). All subsequent analysis assumes consistent convention.
  2. Verify wavenumber range: Confirm the spectrum covers at least 4000--400 cm-1 for a standard mid-IR analysis. Note any truncation.
  3. Assess baseline: A good baseline should be relatively flat and near 100%T (or 0 Absorbance) in regions with no absorption. Sloping or noisy baselines reduce reliability.
  4. Check resolution: Adjacent peaks separated by less than the instrumental resolution cannot be distinguished. Typical FTIR resolution is 4 cm-1.
  5. Identify preparation artifacts: KBr pellets may show a broad O-H band from absorbed moisture (~3400 cm-1). Nujol mulls obscure C-H stretches. ATR spectra show intensity distortion at low wavenumbers. Note any artifacts that limit interpretation.

期待結果: Spectrum confirmed as suitable for analysis, with format, range, and artifacts documented.

失敗時: If the spectrum has severe baseline problems, saturation (flat-bottomed peaks from too-concentrated samples), or preparation artifacts obscuring critical regions, note the limitation and flag affected spectral regions as unreliable.

ステップ2: Scan the Diagnostic Region (4000--1500 cm-1)

Systematically analyze the high-frequency region where most functional groups produce characteristic absorptions:

  1. O-H stretches (3200--3600 cm-1): Look for broad absorptions. A sharp peak near 3600 cm-1 indicates free O-H; a broad band centered at 3200--3400 cm-1 indicates hydrogen-bonded O-H (alcohols, carboxylic acids, water).
  2. N-H stretches (3300--3500 cm-1): Primary amines show two peaks (symmetric and asymmetric stretch); secondary amines show one peak. These are typically sharper than O-H bands.
  3. C-H stretches (2800--3300 cm-1):
Frequency (cm-1)Assignment
3300sp C-H (alkyne, sharp)
3000--3100sp2 C-H (aromatic, vinyl)
2850--3000sp3 C-H (alkyl, multiple peaks)
2700--2850Aldehyde C-H (two peaks from Fermi resonance)
  1. Triple-bond region (2000--2300 cm-1):
Frequency (cm-1)AssignmentNotes
2100--2260C triple-bond CWeak or absent if symmetric
2200--2260C triple-bond NMedium to strong
~2350CO2Atmospheric artifact, disregard
  1. Carbonyl region (1650--1800 cm-1) -- the most diagnostic single region in IR:
Frequency (cm-1)Assignment
1800--1830, 1740--1770Acid anhydride (two C=O stretches)
1770--1780Acid chloride
1735--1750Ester
1700--1725Carboxylic acid
1705--1720Aldehyde
1705--1720Ketone
1680--1700Conjugated ketone / alpha-beta unsaturated
1630--1690Amide (amide I band)
  1. C=C and C=N stretches (1600--1680 cm-1): Alkene C=C appears at 1620--1680 cm-1 (weak to medium). Aromatic C=C shows multiple peaks near 1450--1600 cm-1. C=N (imine) appears at 1620--1660 cm-1.

期待結果: All absorptions in the diagnostic region identified, with functional group assignments and confidence levels (strong, tentative, absent).

失敗時: If the carbonyl region is obscured (e.g., water absorption in KBr, atmospheric CO2), note the gap. If an expected functional group absorption is absent, confirm with a second preparation method before concluding it is truly absent.

ステップ3: Analyze the Fingerprint Region (1500--400 cm-1)

Examine the lower-frequency region for confirmatory and structural detail:

  1. C-O stretches (1000--1300 cm-1): Ethers, esters, alcohols, and carboxylic acids produce strong C-O stretching absorptions. Esters show a characteristic strong band near 1000--1100 cm-1 in addition to the carbonyl.
  2. C-N stretches (1000--1250 cm-1): Amines and amides; overlap with C-O makes assignment tentative without other evidence.
  3. C-F, C-Cl, C-Br stretches:
Frequency (cm-1)Assignment
1000--1400C-F (strong)
600--800C-Cl
500--680C-Br
  1. Aromatic substitution pattern (700--900 cm-1): Out-of-plane C-H bending reveals substitution:
Frequency (cm-1)Pattern
730--770Mono-substituted (+ 690--710)
735--770Ortho-disubstituted
750--810, 860--900Meta-disubstituted
790--840Para-disubstituted
  1. Overall fingerprint comparison: The fingerprint region is unique to each compound. If a reference spectrum is available, overlay and compare this region for identity confirmation.

期待結果: Confirmatory assignments for functional groups identified in Step 2, plus additional structural detail (substitution patterns, C-O/C-N assignments).

失敗時: The fingerprint region is inherently complex and overlapping. If assignments are ambiguous, flag them as tentative and rely on the diagnostic region and other spectroscopic data for final conclusions.

ステップ4: Assess Hydrogen Bonding and Intermolecular Effects

Evaluate how sample state and intermolecular interactions affect the spectrum:

  1. Hydrogen bonding broadening: Compare the width and position of O-H and N-H bands. Free O-H is sharp and near 3600 cm-1; hydrogen-bonded O-H is broad and shifted to 3200--3400 cm-1. Carboxylic acid dimers show a very broad O-H from 2500--3300 cm-1.
  2. Concentration and state effects: Solution spectra at different concentrations can distinguish intramolecular (concentration-independent) from intermolecular (concentration-dependent) hydrogen bonds.
  3. Fermi resonance: Two overlapping bands can interact to split into a doublet. The classic example is the aldehyde C-H pair near 2720 and 2820 cm-1. Recognize Fermi resonance to avoid misassigning extra peaks as separate functional groups.
  4. Solid-state effects: KBr pellets and Nujol mulls reflect solid-state packing, which broadens bands and can shift frequencies by 10--20 cm-1 relative to solution spectra. ATR spectra are closest to the neat liquid state.

期待結果: Hydrogen bonding state characterized, preparation-method artifacts accounted for, and any anomalous band shapes explained.

失敗時: If hydrogen bonding effects cannot be resolved (e.g., overlapping O-H and N-H bands), note the ambiguity. A D2O exchange experiment or variable-temperature study can help, but these require additional data.

ステップ5: Compile Functional Group Inventory

Assemble all findings into a structured report:

  1. List confirmed functional groups: Groups with strong, unambiguous absorptions in the diagnostic region (e.g., sharp C=O at 1715 cm-1 = ketone or aldehyde).
  2. List tentative assignments: Groups with weaker evidence or overlapping absorptions that could be explained by more than one functional group.
  3. List absent functional groups: Groups whose characteristic strong absorptions are clearly missing from the spectrum (e.g., no broad O-H band means no free alcohol or carboxylic acid).
  4. Note discrepancies: Any absorptions that do not fit the proposed functional group set, or expected absorptions that are missing.
  5. Cross-reference: Compare the IR-derived functional group inventory with information from other techniques (NMR, MS, UV-Vis) if available.

期待結果: A complete functional group inventory categorized by confidence level, with specific frequencies and intensities cited as evidence for each assignment.

失敗時: If the inventory is incomplete or contradictory, identify which additional experiments (ATR vs. KBr comparison, variable concentration, D2O exchange) would resolve the ambiguities.

バリデーション

  • Spectrum quality assessed (baseline, resolution, artifacts, y-axis format)
  • Solvent, preparation-method, and atmospheric artifacts identified and excluded
  • All absorptions in the diagnostic region (4000--1500 cm-1) assigned or flagged
  • Carbonyl region analyzed with specific sub-type assignment where possible
  • Fingerprint region examined for confirmatory evidence
  • Hydrogen bonding effects evaluated and their influence on peak shape/position documented
  • Functional group inventory compiled with confidence levels
  • Absent functional groups explicitly noted (negative evidence is informative)
  • Assignments cross-referenced with other available spectroscopic data

よくある落とし穴

  • Ignoring preparation artifacts: KBr moisture (broad 3400 cm-1), Nujol C-H (2850--2950 cm-1), and ATR intensity distortion at low wavenumbers all mimic or obscure real sample absorptions. Always consider the preparation method.
  • Over-interpreting the fingerprint region: The region below 1500 cm-1 is complex and overlapping. Use it for confirmation, not primary identification. Avoid assigning every peak.
  • Confusing atmospheric CO2 with sample peaks: The sharp doublet near 2350 cm-1 is almost always atmospheric CO2, not a sample absorption. Background subtraction should remove it, but verify.
  • Neglecting band intensity and width: A strong, broad absorption has different diagnostic value than a weak, sharp peak at the same frequency. Report intensity (strong/medium/weak) and shape (sharp/broad) alongside frequency.
  • Single-peak assignments: Never identify a functional group from a single absorption alone. Carbonyl groups, for example, should be supported by additional bands (C-O for esters, N-H for amides, C-H for aldehydes).
  • Assuming absence from weak absorption: Some functional groups produce inherently weak IR absorptions (symmetric C=C, triple bonds in symmetric alkynes). Absence of a peak does not always mean absence of the group.

関連スキル

  • interpret-nmr-spectrum -- determine detailed connectivity and hydrogen environments
  • interpret-mass-spectrum -- establish molecular formula and fragmentation pattern
  • interpret-uv-vis-spectrum -- characterize chromophores complementing IR functional group data
  • interpret-raman-spectrum -- obtain complementary vibrational data for IR-inactive modes
  • plan-spectroscopic-analysis -- select and sequence spectroscopic techniques before data acquisition

GitHub 仓库

pjt222/agent-almanac
路径: i18n/ja/skills/interpret-ir-spectrum
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

相关推荐技能

llamaguard

其他

LlamaGuard是Meta推出的7-8B参数内容审核模型,专门用于过滤LLM的输入和输出内容。它能检测六大安全风险类别(暴力/仇恨、性内容、武器、违禁品、自残、犯罪计划),准确率达94-95%。开发者可通过HuggingFace、vLLM或Sagemaker快速部署,并能与NeMo Guardrails集成实现自动化安全防护。

查看技能

cost-optimization

其他

这个Claude Skill帮助开发者优化云成本,通过资源调整、标记策略和预留实例来降低AWS、Azure和GCP的开支。它适用于减少云支出、分析基础设施成本或实施成本治理策略的场景。关键功能包括提供成本可视化、资源规模调整指导和定价模型优化建议。

查看技能

quantizing-models-bitsandbytes

其他

这个Skill使用bitsandbytes库量化大语言模型,能在GPU内存有限时通过8位或4位量化减少50-75%内存占用,同时保持精度损失最小。它支持INT8、NF4、FP4等多种量化格式,可与HuggingFace Transformers无缝集成,适用于需要部署更大模型或加速推理的场景。还提供QLoRA训练和8位优化器支持,让开发者能轻松实现高效模型压缩。

查看技能

dispatching-parallel-agents

其他

该Skill用于并行处理3个以上无依赖关系的独立故障,可为每个问题域分派专属Claude代理同时执行调查修复。它通过并发处理多个独立问题显著提升故障排查效率,特别适用于测试文件、子系统等无共享状态的场景。

查看技能