All four membranes share identical chemistry — 1100 equivalent weight perfluorosulfonic acid — so thickness and manufacturing method are your only real variables. Nafion 211 (25 µm) and 212 (51 µm) are dispersion-cast and thin: lowest resistance, highest gas crossover. Nafion 115 (127 µm) and 117 (183 µm) are extruded and thick: more voltage loss, far better gas separation and pressure tolerance. The right choice is the thinnest membrane that keeps crossover under your system's limit at its lowest operating current density — not its rated one. This guide shows you how to calculate that number.

What the Nafion part numbers actually mean

The digits encode equivalent weight, thickness and — less obviously — how the film was made.

• Extruded series (115, 117, 1110): the first two digits are equivalent weight divided by 100. So 1100 g of dry polymer per mole of sulfonic acid group, which works out to roughly 0.90 meq/g of acid capacity. The final digit is thickness in mils (thousandths of an inch). Nafion 117 = 1100 EW, 7 mil.
• Dispersion-cast series (211, 212, historically NR-211 and NR-212): the final digit is still thick in mils, but the film is cast from a Nafion dispersion rather than melt-extruded. Same polymer, different morphology, different mechanical behaviour.

Most comparison guides stop at thickness. The extruded-versus-cast split is what decides how the membrane behaves in your hands.

Nafion membrane comparison table: the numbers that matter

Property	N211	N212	N115	N117
Thickness	25.4 µm (1 mil)	50.8 µm (2 mil)	127 µm (5 mil)	183 µm (7 mil)
Process	Dispersion cast	Dispersion cast	Extruded	Extruded
Basis weight	50 g/m²	100 g/m²	250 g/m²	360 g/m²
Equivalent weight	1100	1100	1100	1100
Acid capacity	0.92 meq/g min	0.92 meq/g min	0.90 meq/g min	0.90 meq/g min
Conductivity	0.10 S/cm min	0.10 S/cm min	0.10 S/cm min	0.10 S/cm min
Calculated ASR	25 mΩ·cm²	51 mΩ·cm²	127 mΩ·cm²	183 mΩ·cm²
Ohmic loss @ 1 A/cm²	25 mV	51 mV	127 mV	183 mV
Relative H₂ crossover	Highest (7x)	High (3.6x)	Low (1.4x)	Lowest (1x)
In-plane swelling, 23 °C water	~10%	~10%	~10%	~10%
In-plane swelling, 100 °C water	~15%	~15%	~15%	~15%
Anisotropy (MD vs TD)	Near-isotropic	Near-isotropic	Anisotropic	Anisotropic
Best-fit duty	Low-ΔP fuel cells	General bench PEMFC	PEM electrolysis	High-ΔP, VRFB

ASR above is calculated as thickness ÷ 0.10 S/cm, the datasheet conductivity minimum at full hydration. Measured high-frequency resistance in a running cell is always higher — contact resistance and sub-saturated operation add to it. Treat these as the floor, not the expectation.

The trade-off in one line: ohmic loss versus gas crossover

Every extra mil of Nafion costs voltage and buys gas separation. That is the whole decision.

Run an electrolyser at 2 A/cm² on Nafion 117 and the membrane alone burns roughly 366 mV — about 20% of a 1.8 V cell, and so about 20% of that stack's electricity bill. Swap to N115 and you recover ~112 mV; swap to N212 and you recover ~264 mV, but you have multiplied hydrogen crossover by roughly 3.6 and lost most of your mechanical margin.

Crossover scales roughly inversely with thickness:

• Approximate H₂ crossover current at 80 °C, fully humidified, 1 bar: ~2 mA/cm² (N211), ~1 mA/cm² (N212), ~0.4 mA/cm² (N115), ~0.3 mA/cm² (N117).
• Crossover rises with differential pressure and with temperature, and it does not care what current density you are running. That asymmetry is what causes most real failures.

Almost every membrane-selection mistake in electrolysis comes from sizing on the rated point instead of the turndown point. Worked Example 1 below shows what that costs.

Extruded versus dispersion-cast: four consequences most guides skip

1. Anisotropy. Extruded N115 and N117 are drawn during production, so machine direction (MD) and transverse direction (TD) differ measurably in tensile strength and swelling. Cast N211 and N212 are near-isotropic. Die-cutting large sheets to tight tolerances? Mark the MD first.
2. Swelling magnitude. All four gain roughly 10% in-plane going from 50% RH to room-temperature water, and around 15% to boiling water. On a 200 mm sheet that is 20–30 mm — more than enough to buckle a membrane trapped inside a gasketed cell that was assembled dry.
3. Handling margin. N211 ships on a backing film for a reason. At 25 µm it creases if you look at it wrong, and a crease becomes a pinhole. N117 tolerates being handled like a sheet of paper.
4. Residual stress. Extruded film carries frozen-in stress from the draw and shrinks on the first hot press or thermal cycle. Cast film does this far less, which is why post-press N212 dimensions are more predictable than N115.

How to choose a Nafion membrane by application

PEM fuel cells

N212 is the sensible bench default; N211 when you are chasing power density. Thick membranes fail here for a reason that is easy to miss: cathode-generated water has to back-diffuse to the anode to keep it hydrated, and thick film obstructs that. Run N115 or N117 at high current density on dry anode feed and you will fight anode dry-out a thinner membrane would have solved by itself.

PEM water electrolysis

N115 is the workhorse; N117 when differential pressure is high or turndown is deep. Electrolysis flips the fuel-cell logic: the membrane becomes a structural component holding tens of bar of hydrogen against liquid water, and crossover becomes a safety limit rather than an efficiency line item.

Vanadium redox flow batteries

N117 gives the lowest vanadium crossover and the best coulombic efficiency of the four; N115 and N212 trade capacity retention for voltage efficiency. The honest caveat: Nafion is a cation-exchange membrane and vanadium ions cross it regardless of thickness. It is the incumbent here, not the optimum.

Sensors, humidifiers and bench electrochemistry

N117 and N115 win here because mechanical robustness beats resistance. Nobody is counting millivolts in a humidity sensor or a two-compartment cell, but a torn membrane ends the experiment. For gas drying specifically, use Nafion tubing rather than sheet.

When none of these four is the right answer

Reinforced grades — Nafion XL (around 27.5 µm with a support layer), Nafion HP, and composites such as Gore-Select — give thin-film resistance with thick-film mechanical life. If your system needs 10,000+ hours or wide RH cycling, unreinforced Nafion is a research material rather than a product choice; commercial automotive stacks moved to reinforced composites at 8–20 µm years ago.

Worked example 1: the crossover limit that sets your minimum current density

This is the calculation that catches teams out on pressurised electrolysers. Hydrogen-in-oxygen at the anode has a hard ceiling of 4 vol% — the lower flammability limit — and operators typically alarm at 2 vol% for margin.

1. Take your membrane's 1 bar crossover figure. N115: roughly 0.4 mA/cm² equivalent.
2. Scale for differential pressure. Permeation is roughly proportional to partial pressure. At 30 bar, that is ~12 mA/cm².
3. Compare against oxygen production. Oxygen is generated at i/4F; crossed hydrogen arrives at i_cross/2F. So vol% H₂ in O₂ ≈ 2 × (i_cross / i) × 100.
4. Run the numbers. At 2 A/cm² you get ~1.2 vol% — comfortable. At 1 A/cm², ~2.4 vol% — alarming. At 0.5 A/cm², ~4.8 vol% — above the LFL.
5. Act on it. Your minimum safe current density on N115 at 30 bar is somewhere near 1 A/cm². Below that you need N117, a lower operating pressure, or a recombination catalyst layer.

Worked example 2: building a 25 cm² MEA without wrecking the membrane

A repeatable bench procedure, and the places people lose membranes.

1. Pretreat. One hour each at 80 °C in 3 wt% H₂O₂ (strips organics), DI water, 0.5–1 M H₂SO₄ (protonates fully and strips metal cations), then two DI rinses. Store submerged in DI water.
2. Cut after treatment, not before. A 50 × 50 mm dry N117 coupon becomes roughly 55 × 55 mm after boiling. Cut a dry piece to final size and it will not fit the gasket window.
3. Blot, do not dry. Surface-dry with lint-free wipes. Hot-pressing a saturated membrane produces steam blisters.
4. Hot press. 130–140 °C for 2–3 minutes at around 3 MPa — on 25 cm² that is roughly 7.5 kN, or about 0.77 tonne on the press gauge. Drop to ~1–2 MPa and 90 seconds for N211, which creeps and thins under standard settings.
5. Check open-circuit voltage before trusting any data. A healthy H₂/air cell at 80 °C holds above ~0.95 V. Below ~0.90 V means crossover or a pinhole, and every polarisation curve after that is fiction.

Two mistakes specific to this procedure: over-torquing the cell, which extrudes membrane into the flow channels, and pressing above ~150 °C, which degrades the polymer rather than bonding it.

Five handling mistakes that scrap membranes

• Storing pretreated membrane dry. It embrittles and will not rehydrate to the same dimensions. Keep it wet, or in the original backing.
• Using metal tools. Stainless tweezers leach iron into the sulfonic acid sites. Use PTFE-tipped or plastic tools throughout.
• Assembling dry, running wet. The membrane swells ~10% inside a fixed gasket window and buckles. Size the window for the hydrated dimension.
• Reusing a membrane after a dry-out event. Nafion cycled to fully dry under compression retains microcracks that never show visually.
• Assuming thickness tolerance is tight. Extruded film varies meaningfully across a roll, which shows up as HFR scatter between nominally identical cells.

Limitations and where this guidance stops

Everything above assumes an acid-form, fully hydrated membrane below about 90 °C. Past that, Nafion dehydrates, conductivity collapses and none of the ASR figures hold — that regime belongs to PBI/phosphoric acid systems. None of these grades suits alkaline systems either, where you need an anion exchange membrane.

Two commercial caveats. Nafion is a PFAS material, and the ECHA universal PFAS restriction proposal submitted in January 2023 covers fluoropolymers, with fuel cell and electrolyser membranes under active derogation discussion — irrelevant to bench work, material to a ten-year product plan. And cost does not scale with thickness: dispersion casting carries its own process cost, so N211 is nowhere near one-seventh the price of N117.

Frequently asked questions

What is the difference between Nafion 115 and 117?

Thickness only — both are extruded 1100 EW film. Nafion 115 is 127 µm, Nafion 117 is 183 µm. N117 has about 44% more area specific resistance (183 vs 127 mΩ·cm²) but roughly 30% less gas crossover. Choose N117 for high differential pressure or deep turndown, N115 for everything else in electrolysis.

Is Nafion 212 the same as NR-212?

Yes. NR-212 was the original DuPont designation for the dispersion-cast 2 mil film; Chemours now markets it as Nafion 212. Specifications are unchanged: 50.8 µm, 100 g/m², 1100 equivalent weight. If a supplier lists both, they are the same product under old and new naming.

Why not always use the thinnest membrane?

Because gas crossover scales inversely with thickness and does not depend on current density. A thin membrane that is fine at rated load can exceed the 4 vol% hydrogen-in-oxygen flammability limit at low load. Thin film also has less mechanical margin against differential pressure, creep under compression and pinhole formation.

Do I need to pretreat Nafion before use?

For quantitative electrochemistry, yes. The standard sequence is one hour each at 80 °C in 3% H₂O₂, DI water, 0.5–1 M H₂SO₄, then DI water. This removes organic residues, converts the membrane fully to acid form and strips metal cations. Skipping it typically costs measurable conductivity and adds run-to-run scatter.

Which Nafion membrane is best for a PEM electrolyser?

Nafion 115 for most balance-of-plant designs operating up to roughly 30 bar with limited turndown. Move to Nafion 117 if you operate above that pressure, need to run below about 1 A/cm², or cannot fit a recombination catalyst. Reinforced grades are worth evaluating for anything targeting a long commercial lifetime.

How much does membrane thickness cost me in efficiency?

At 1 A/cm² the membrane's own ohmic loss is roughly 25 mV (N211), 51 mV (N212), 127 mV (N115) and 183 mV (N117). On a 1.8 V electrolyser cell, moving from N117 to N115 recovers about 3% of cell voltage — and at 2 A/cm² that doubles to about 6%.

Conclusion

One polymer, four thicknesses, two manufacturing routes. Thickness sets the ohmic penalty; thickness and pressure together set the crossover risk; the manufacturing route sets how the film handles and swells. Thin cast film (211, 212) for fuel cells and performance work; thick extruded film (115, 117) for pressure, turndown and mechanical life.