返回技能列表

interpret-ir-spectrum

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

关于

This Claude Skill analyzes infrared spectra to identify functional groups by examining diagnostic and fingerprint regions. It detects hydrogen bonding effects and compiles a functional group inventory with confidence levels. Use it for initial compound screening or to confirm structural features in unknown samples.

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

Interpret IR Spectrum

Read IR absorption spectra. Identify functional groups. Check hydrogen bonding. Compile inventory of structural features in sample.

When Use

  • Identify functional groups in unknown compound as first screen
  • Confirm presence or absence of specific functional groups (e.g., verify reaction turned alcohol to ketone)
  • Monitor reaction progress by tracking appearance/disappearance of characteristic absorptions
  • Distinguish between similar compounds differing in functional group content
  • Complement NMR + mass spec data with vibrational info

Inputs

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

Steps

Step 1: Check Spectrum Quality and Format

Verify spectrum suitable for interpretation before analyzing peaks:

  1. Check y-axis format: %Transmittance (%T, peaks point down) or Absorbance (A, peaks point up)? All analysis assumes consistent convention
  2. Verify wavenumber range: Covers at least 4000-400 cm-1 for standard mid-IR? Note any truncation
  3. Assess baseline: Good baseline relatively flat, near 100%T (or 0 Absorbance) in regions with no absorption. Sloping or noisy baselines cut reliability
  4. Check resolution: Adjacent peaks closer than instrumental resolution cannot be distinguished. Typical FTIR resolution = 4 cm-1
  5. Identify prep artifacts: KBr pellets may show broad O-H band from moisture (~3400 cm-1). Nujol mulls obscure C-H stretches. ATR spectra show intensity distortion at low wavenumbers. Note any artifacts limiting interpretation

Got: Spectrum confirmed suitable. Format, range, artifacts documented.

If fail: Severe baseline problems, saturation (flat-bottomed peaks from too-concentrated samples), or prep artifacts obscuring critical regions? Note limitation. Flag affected spectral regions unreliable.

Step 2: Scan Diagnostic Region (4000-1500 cm-1)

Systematic analysis of high-frequency region where most functional groups make characteristic absorptions:

  1. O-H stretches (3200-3600 cm-1): Look for broad absorptions. Sharp peak near 3600 cm-1 = free O-H. Broad band centered at 3200-3400 cm-1 = hydrogen-bonded O-H (alcohols, carboxylic acids, water)
  2. N-H stretches (3300-3500 cm-1): Primary amines show two peaks (symmetric + asymmetric). Secondary amines show one. 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) — 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 = 1620-1680 cm-1 (weak to medium). Aromatic C=C = multiple peaks near 1450-1600 cm-1. C=N (imine) = 1620-1660 cm-1

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

If fail: Carbonyl region obscured (e.g., water absorption in KBr, atmospheric CO2)? Note gap. Expected functional group absorption absent? Confirm with second prep method before concluding truly absent.

Step 3: Analyze Fingerprint Region (1500-400 cm-1)

Examine lower-frequency region for confirmatory and structural detail:

  1. C-O stretches (1000-1300 cm-1): Ethers, esters, alcohols, carboxylic acids make strong C-O stretching. Esters show characteristic strong band near 1000-1100 cm-1 in addition to 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 compare: Fingerprint region unique to each compound. Reference spectrum available? Overlay and compare this region for identity confirmation

Got: Confirmatory assignments for functional groups from Step 2, plus additional structural detail (substitution patterns, C-O/C-N assignments).

If fail: Fingerprint region inherently complex and overlapping. Assignments ambiguous? Flag as tentative. Rely on diagnostic region and other spectroscopic data for final conclusions.

Step 4: Assess Hydrogen Bonding and Intermolecular Effects

Evaluate how sample state and intermolecular interactions affect spectrum:

  1. Hydrogen bonding broadening: Compare width and position of O-H and N-H bands. Free O-H sharp, near 3600 cm-1. Hydrogen-bonded O-H broad, shifted to 3200-3400 cm-1. Carboxylic acid dimers show very broad O-H from 2500-3300 cm-1
  2. Concentration and state effects: Solution spectra at different concentrations distinguish intramolecular (concentration-independent) from intermolecular (concentration-dependent) hydrogen bonds
  3. Fermi resonance: Two overlapping bands interact to split into doublet. Classic example: 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. Broadens bands. Shifts frequencies 10-20 cm-1 relative to solution spectra. ATR spectra closest to neat liquid state

Got: Hydrogen bonding state characterized. Prep-method artifacts accounted for. Any anomalous band shapes explained.

If fail: Hydrogen bonding effects cannot be resolved (e.g., overlapping O-H and N-H bands)? Note ambiguity. D2O exchange experiment or variable-temperature study can help, but need additional data.

Step 5: Compile Functional Group Inventory

Assemble findings into structured report:

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

Got: Complete functional group inventory by confidence level, with specific frequencies and intensities cited as evidence for each assignment.

If fail: Inventory incomplete or contradictory? Identify which additional experiments (ATR vs. KBr compare, variable concentration, D2O exchange) would resolve ambiguities.

Checks

  • Spectrum quality assessed (baseline, resolution, artifacts, y-axis format)
  • Solvent, prep-method, atmospheric artifacts identified and excluded
  • All absorptions in 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, influence on peak shape/position documented
  • Functional group inventory compiled with confidence levels
  • Absent functional groups explicitly noted (negative evidence informative)
  • Assignments cross-referenced with other available spectroscopic data

Pitfalls

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

See Also

  • 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 — 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/caveman/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