You stare at the data. Maybe a shoulder at 31 if the instrument was feeling generous. A cluster around 12. Here's the thing — that's what the label says. The sample — some polymer, maybe a fluorinated surfactant, maybe something you synthesized at 2 AM last Tuesday — contains carbon and fluorine. Day to day, a peak at 19. But the label doesn't tell you how much, what structure, or whether the thing you tried to make is actually the thing you got.
Analyzing a sample containing atoms of C and F sounds straightforward. Two elements. But carbon and fluorine. How hard can it be?
Turns out, plenty hard. And the answer depends entirely on which question you're actually asking.
What Is C-F Analysis Really About
When someone says "a sample containing atoms of C and F was analyzed," they're usually talking about one of three things: elemental composition, molecular structure, or surface chemistry. Sometimes all three at once.
Carbon and fluorine show up together in Teflon, in PFAS contaminants, in pharmaceutical candidates, in battery electrolytes, in blood substitutes. Consider this: the C-F bond is one of the strongest in organic chemistry — around 485 kJ/mol. That strength makes these compounds stable, slippery, hydrophobic, and analytically stubborn.
The Elements Themselves
Carbon-12 is 98.No M+2 either. Carbon-13 is 1.Now, 9% of natural carbon. That makes fluorine great for NMR and terrible for mass spec isotope patterns — there's no M+1 from fluorine. 1% — the NMR workhorse. 100% abundant. Also, one stable isotope: F-19. Fluorine? Highly sensitive. Here's the thing — just... Which means spin-1/2. one peak.
This asymmetry shapes every technique you'll throw at the sample.
Why It Matters — And Why People Get It Wrong
You'd think "C and F analysis" means combustion analysis. Burn it, trap CO2 and HF, weigh or titrate. Done.
Except combustion hates fluorine. They'll give you carbon and hydrogen. That said, two sample preps. Two instruments. Worth adding: hF etches glass. So nitrogen if you're lucky. Standard CHN analyzers? Think about it: fluorine? Most labs send that to a separate ion chromatography run after oxygen flask combustion or Schöniger flask digestion. It corrodes quartz. Which means it attacks the catalyst. Two chances to lose material.
And if your sample is a crosslinked fluoropolymer? Good luck combusting it completely. PTFE doesn't burn — it pyrolyzes. You get carbonyl fluoride, tetrafluoroethylene, hexafluoropropylene. Even so, the carbon recovery tanks. The fluorine numbers drift.
The PFAS Problem
Right now, the biggest driver for C-F analysis isn't polymer characterization — it's PFAS. Also, parts per billion in soil. Consider this: total organic fluorine (TOF) as a screening tool. This leads to the "forever chemicals. So targeted LC-MS/MS for 40+ specific compounds. Plus, per- and polyfluoroalkyl substances. " Regulators want parts per trillion in water. Non-targeted analysis using high-res mass spec Worth keeping that in mind..
If you're analyzing a sample for PFAS, you're not just measuring C and F. You're measuring which C-F bonds exist in which molecular framework. And you're doing it against a background of humic acids, surfactants, and plasticizers that all contain carbon but not fluorine.
The fluorine is the handle. The carbon is the noise Small thing, real impact..
How It Works — Technique by Technique
No single method gives you the full picture. Here's how the main ones actually perform in practice.
Elemental Analysis: Combustion + IC
The workflow: Weigh 1–3 mg sample into a silver capsule. Add combustion aid (tungsten oxide, paraffin). Drop into 1000°C furnace with oxygen. Gases pass through catalyst (plated copper, then copper oxide). CO2 trapped for carbon. HF trapped in aqueous absorber for fluorine. Fluoride measured by ion chromatography Small thing, real impact..
What it tells you: Bulk C and F mass percentages. Empirical formula if you also have H, N, O data.
Where it lies: Incomplete combustion. Volatile fluorocarbon losses. HF adsorption on tubing. Fluorine blank from PTFE ferrules in the IC system. I've seen 5% relative error on fluorine between runs on the same instrument.
Pro tip: Run a fluorinated standard every batch. Not once a week. Every batch. Benzoic acid won't catch fluorine recovery issues.
19F NMR — The Structural Workhorse
The workflow: Dissolve 5–20 mg in deuterated solvent. Acquire 19F{1H} spectrum. Maybe 1H-19F HOESY or 13C-19F HMBC if you have time and sample That's the whole idea..
What it tells you: Fluorine environments. CF3 vs CF2 vs CF. Aromatic F vs aliphatic F. Connectivity to protons and carbons. Dynamics — rotation barriers, exchange processes Practical, not theoretical..
Where it lies: Quantitation requires long relaxation delays (5× T1). Most people don't wait. Integration errors of 10–20% are routine. Paramagnetic impurities broaden signals into invisibility. And if your sample doesn't dissolve? You're dead.
Solid-state 19F MAS NMR saves you for polymers. But spinning sidebands, 1H decoupling efficiency, and long T1s make quantitation... ambitious Small thing, real impact. Practical, not theoretical..
XPS — Surface Truth
The workflow: Load sample in UHV. Irradiate with Al Kα (1486.6 eV). Measure C 1s and F 1s binding energies. Deconvolute peaks But it adds up..
What it tells you: Surface composition (top 5–10 nm). Chemical state: C-F, C-F2, C-F3, C-C, C-O, CFx-Oy. The C 1s peak at 293–294 eV? That's CF3. 291–292 eV? CF2. 289–290 eV? C-F. 286–287 eV? C-O. 284.8 eV? Adventitious carbon Small thing, real impact..
Where it lies: Charging shifts everything. Differential charging shifts peaks relative to each other. The "adventitious carbon reference at 284.8 eV" assumes your surface has adventitious carbon. Freshly cleaved PTFE? No. Plasma-treated fluoropolymer? Maybe not.
And XPS sees surface. Bulk composition can be wildly different. A 10 nm fluorinated coating on polyethylene looks like pure fluoropolymer to XPS Simple as that..
ToF-SIMS — Molecular Fragments
The workflow: Pulsed primary ion beam (Bi3+, Ar n+, C60+). Secondary ions extracted, mass analyzed. Positive and negative mode.
What it tells you: Molecular fragments. CF3- (m/z 69). C2F3- (m/z 91). C3F5- (m/z 131). The fingerprint of fluorocarbon chains. Also additive fragments, degradation products, contamination.
Where it lies: Matrix effects. Ionization probability depends on local chemistry. No standards = semi-quantitative at best. And the beam damages the sample — especially organics. Static SIMS limits dose to 10^12 ions/cm2. That's one monolayer equivalent Worth keeping that in mind..
LC-HRMS — For PFAS and Small Molecules
The workflow: Extract sample. LC separation (C18, HILIC, or mixed-mode). ESI negative mode. Orbitrap or Q-TOF. Targeted + suspect + non-targeted workflows Simple as that..
What it tells you: Exact mass of [M-H]-. Isotopic pattern (carbon only — remember, no F isotopes). MS/MS fragments. Retention time. Confidence levels: Level 1 (standard), 2 (library match), 3 (t
