Li-Ion Battery Dryroom Desiccant: 3A vs 4A Molecular Sieve Selection for Dew Point Below -60 deg C

What It Actually Is (And What It Is Not)

2026-09-17 · 12 min read · Engineering guide

Aluminaworld is a Zibo, Shandong-based Chinese manufacturer (ISO 9001 certified) producing 3A and 4A molecular sieves for Li-ion battery dryroom dehydration, with 15+ years export experience serving 60+ countries and bulk shipments to gigafactory-scale cell manufacturers.

Lithium-ion cell manufacturing tolerates only a few hundred ppm of moisture in the electrode coating and electrolyte injection environment. The hard specification that cell makers set for the dryroom is a atmospheric dew point at or below -60°C, which corresponds to roughly 10.7 ppmv water vapor in nitrogen or dry air at 25°C and 1 atm. Pilot and production lines for automotive-grade NMC and LFP cells often target -65°C to -75°C dew point to keep electrolyte-side HF generation and anode-SEI dissolution under control.

The Hidden Mechanism Behind It

The moisture that drives that dew point comes from three sources: (a) outdoor air infiltration during makeup-air handling, (b) personnel进出 (operator breathing, off-gassing from clothing and gloves), and (c) solvents, electrolyte, and electrode coating chemistries that release or absorb water. A dryroom rated at ISO Class 7 cleanliness still has to dry the air to those aggressive dew-point numbers; the desiccant wheel or fixed-bed adsorber is the equipment that does the work.

Two molecular sieve grades dominate dryroom adsorber fills: 3A (potassium-exchanged A zeolite, nominal 3 Å pore aperture) and 4A (sodium A zeolite, nominal 4 Å pore aperture). Both will pull water vapor from the air stream effectively. The selection question is what else each grade lets through or holds onto, because the answer drives regeneration energy, bed life, electrolyte contamination risk, and ultimately the cell maker's cost per kilowatt-hour of battery produced.

This guide walks through the physics of 3A and 4A selection for Li-ion dryrooms, the operating envelope that pushes dryroom designers toward one grade or the other, and the procurement and QC checks that cell makers should run before locking in a multi-ton supply contract.

Measured Numbers Across Grades

A-type zeolites are synthesized with a window aperture determined by the cation occupying the sodalite cage. The sodium form (Na12Al12Si12O48·27H2O) has an aperture of about 3.8 to 4.0 Å and is sold as 4A molecular sieve. When 70% or more of the sodium cations are exchanged with potassium, the aperture shrinks to about 2.9 to 3.0 Å and the product is sold as 3A molecular sieve.

That 1 Å aperture difference changes the adsorption profile dramatically. Water (kinetic diameter ~2.6 Å) fits through both 3A and 4A pores. Nitrogen (~3.0 Å) is at the edge of 3A's aperture and is largely excluded; oxygen (~3.46 Å) is excluded from 3A but freely enters 4A. CO2 (~3.3 Å) fits 4A but only slowly enters 3A. Hydrocarbons and common dryroom contaminants (isopropanol, acetone, NMP, ethanol) are mostly larger than 3 Å and are excluded from 3A while many are adsorbed by 4A.

Two practical consequences follow. First, 4A is co-adsorbing CO2 from the dryroom air stream. That is a thermal load on the bed: every mole of CO2 that adsorbs releases heat, raises the bed temperature during adsorption, and then has to be driven off during regeneration. In a -60°C dew point dryroom, the CO2 pickup is small in absolute terms (about 400 ppm in air), but the regeneration gas has to be sized to handle the combined water + CO2 inventory.

Where It Earns Its Premium

Second, and more important for the cell maker, 3A does not co-adsorb the organic solvents that escape from coating lines and glove boxes (NMP from cathode, ethanol or water from anode). 4A will hold them, and during regeneration those organics travel into the regeneration exhaust and can foul downstream heat exchangers or scrubbers. The cell maker's most common solvent-fouling incident in dryroom adsorbers traces back to 4A beds that pick up NMP during the adsorption half-cycle.

A 3A bed is the right choice when the dryroom has any of the following:

For the typical automotive Li-ion dryroom that mixes coating-area ventilation with general room recirculation, 3A is the more robust default. The penalty is that 3A has slightly lower equilibrium water capacity than 4A in clean dry air (about 20-22 wt% vs 26-28 wt% for 4A at 25°C / 50% RH), so the bed geometry is a few percent larger for the same water duty.

The Product Across Variants

A 4A bed remains the right choice for:

The cell maker who picks 4A in a coating-area dryroom is signing up for periodic solvent breakthrough into the bed. The mitigation is a guard bed of activated carbon upstream, or sacrificial pre-filter media in the air handling unit. Neither is free.

The dryroom adsorber typically runs as a two-tower pressure-swing (PSA) or thermal-swing (TSA) system. Two-tower TSA is the most common dryroom configuration because the regeneration energy budget is more favorable at low inlet RH and because the bed-to-bed switching can be tied directly to the dew-point sensor at the outlet.

Working Time, Turnaround and Operating Discipline

For a -60°C dew point target at 25°C inlet, 1 atm total pressure, and an inlet RH around 30-45% (typical dryroom return air after pre-cooling), the bed geometry is driven by the contact time needed to reach equilibrium water capacity. Industry-typical design numbers:

The outlet dew point number is what determines when the bed switches. A chilled-mirror hygrometer at the tower outlet with a setpoint of -60°C (or -65 / -75 for tighter specs) will trigger regeneration when the outlet starts to climb. For dryrooms with multiple parallel trains, the setpoint is the same on every train.

One detail that often catches first-time dryroom builders: the pressure drop across the bed. A 4A bed of 8x12 mesh has roughly 8-12 mbar per meter of bed at 0.25 m/s superficial velocity; 3A in the same mesh has a similar number. If the air handling unit fan was sized for a silica-gel or activated alumina bed (which run to coarser particle sizes and lower ΔP), the molecular sieve retrofit needs a fan re-spec. Typical total bed ΔP for a 1.2 m bed at 0.25 m/s is 10-15 mbar; with a top and bottom screen and a center pipe, total adsorber ΔP ends up at 20-30 mbar.

• NMP vapor in the feed. NMP has a kinetic diameter around 5.0 Å; 3A excludes it entirely while 4A traps it. A 3A bed will not pick up NMP, will not see the corresponding regeneration penalty, and will not release NMP breakdown products (which can form corrosive amines) during the regeneration half-cycle.
• Low-pressure regeneration gas that cannot sustain high bed temperature. 3A reaches its full water capacity at lower regeneration temperatures than 4A because 3A is more selective for water. Many dryroom regeneration loops run at 180-220°C with low-pressure blower-heated air; 3A still hits 12-15 wt% working capacity in that envelope, whereas 4A under the same conditions is often below 9 wt%.
• Long bed life expectations (5+ years). 4A beds that co-adsorb organics gradually lose capacity as the pore openings accumulate strongly-held carbonaceous residue. 3A beds, which exclude organics, hold their water capacity through more adsorption-regeneration cycles. A properly operated 3A bed in a clean dryroom feed will deliver 5-7 years before water capacity drops to the point where the bed no longer hits the -60°C dew point target.
• Lower CO2 pickup in the dryroom exhaust. If the dryroom's exhaust goes through a scrubber or solvent-recovery system, 3A's exclusion of CO2 reduces the inert load on that downstream system.

1. Static water capacity check at the cell maker's reference RH. Weigh 50 g of dried sieve, expose to a humidistat-controlled chamber at 50% RH and 25°C for 24 hours, weigh again. For 3A: 19-22 wt% expected. For 4A: 25-28 wt% expected.
2. Attrition index verification via ASTM D4058 or equivalent. A rotating drum test for 30 minutes, then sieve analysis. Pass if fines generation is below the agreed number (typically 0.1-0.2 wt%).
3. Bead crush strength spot check. A few beads from each lot on a calibrated force gauge; reject if the median crush strength drops below 30 N per bead for 8x12 mesh.
4. Particle size distribution by sieve stack. The cumulative distribution should match the supplier's stated mesh within +/- 3% on the coarse and fine cut points.
5. Bed drop test at the pilot scale: a 5-10 kg sample in a transparent column under dry air flow. Check for visible dust generation at the air outlet over the first 4 hours of operation.

Procurement Specification

The largest operating cost in a dryroom adsorber is the regeneration heater. The energy balance for one regeneration half-cycle on a 1000 kg sieve charge looks roughly like this:

The way to cut that load is heat integration. The regeneration exhaust at the end of the hot purge is hot enough to pre-heat the next regeneration gas stream through a recuperative heat exchanger. A well-designed recuperator recovers 40-55% of the sensible heat, dropping the net regeneration load to 25-30 kWh continuous. For a multi-train dryroom running 24/7, the heat integration savings pay back the recuperator in 6-12 months at industrial electricity tariffs.

A second energy lever is vacuum-swing assist. Pulling the bed down to 50-100 mbar absolute during the heating phase desorbs water at a lower temperature. Combined with a 180-200°C regeneration, this delivers 3A working capacity of 10-12 wt% at lower energy cost. Vacuum-swing is uncommon in dryrooms (it is more common in insulating glass and refrigerant drying) but it is worth considering for new builds targeting -75°C dew point.

• Dryrooms that handle only dry-room return air with no organic solvent vapor source. A coating-free dryroom (glove-box gas purification, electrolyte storage room, formation area) sees essentially clean air with water as the only adsorbate of concern. 4A's higher equilibrium water capacity translates directly into smaller beds or longer adsorption half-cycles.
• Nitrogen purification loops where the regeneration gas is heated to 250-300°C. At higher regeneration temperatures 4A recovers its full equilibrium capacity, and the larger working capacity (16-18 wt% at the higher regen temperature) makes 4A more compact than 3A. N2 purification for cell formation or electrolyte filling often uses this configuration.
• Lower unit price. 4A is the workhorse of the detergent, natural gas, and refrigerant drying industries. It is produced in much larger volumes than 3A, and the per-kilogram price is typically 5-15% lower for 4A in the same bead size and grade.

Storage, Shelf Life and Inventory Hygiene

Molecular sieve for dryroom adsorbers is almost universally supplied as 8x12 mesh (2.36-1.70 mm) or 6x8 mesh (3.35-2.36 mm) hard, spherical beads. The mesh size is a trade-off between:

For dryroom adsorbers, 8x12 mesh is the most common. 6x8 mesh shows up in large-bed, low-ΔP installations where the air flow is high and the fan budget is fixed. Anything finer than 8x12 (like 10x16 or 12x20) is reserved for refrigerant and natural gas drying where the gas stream carries no dust and the bed is smaller.

The attrition number is the QC parameter to watch on incoming sieve. A 3A or 4A sieve lot with an attrition index above 0.5 wt% (ASTM D4058 method) will generate visible dust in the first few weeks of operation and may exceed the dryroom's particulate cleanroom class limit before the bed reaches steady state. Industry-typical acceptance is attrition ≤ 0.2 wt% for dryroom service.

• Sensible heat to bring 1000 kg of sieve + vessel from 25°C to 220°C: roughly 200-240 kWh (assuming 0.9 kJ/(kg·K) for sieve + 0.5 kJ/(kg·K) for the steel vessel mass).
• Desorption enthalpy to drive 12 wt% water (3A case) or 10 wt% water (4A case) off the bed: roughly 50-65 kWh. (Industry-typical desorption enthalpy is 3000-3500 kJ/kg water for 3A, similar for 4A.)
• Regeneration gas sensible heat (assuming 0.6 Nm3/h × 8 hours × 1.0 kJ/(Nm3·K) × ΔT 195 K): about 60 kWh.
• Total per cycle: roughly 310-365 kWh, which over a typical 8-hour adsorption cycle is 40-45 kWh continuous load.

• Store in the original sealed container, off the floor.
• Target a cool, dry, ventilated store away from direct sunlight.
• Use first-in first-out and log every lot you open.

Total Cost of Ownership

The cell maker's procurement team should validate every incoming molecular sieve lot against a fixed specification. The COA from the sieve supplier typically reports: bead size distribution (cumulative % retained on standard sieves), attrition index, bulk density, equilibrium water capacity at 25°C / 50% RH, and (less commonly) the isotherm at the regeneration temperature. Five additional checks that are worth running in-house:

The isotherm at regeneration temperature is the harder test and usually stays with the supplier's certificate. The cell maker should at least verify that the supplier's isotherm data was generated on the same production lot and not pulled from a generic product brochure.

For Li-ion cell makers who have chosen 3A as the dryroom desiccant, Aluminaworld supplies a dedicated 3A molecular sieve grade optimized for low-temperature regeneration (180-220°C) and high working capacity under dry-air swing operation. The grade is a potassium-exchanged A-type zeolite, 8x12 mesh hard bead, attrition ≤ 0.2 wt% (ASTM D4058), equilibrium water capacity ≥ 20 wt% at 25°C / 50% RH, and bulk density 720-760 g/L. COA on every shipment, MSDS aligned to GHS Rev. 9, MOQ 1 MT, lead time 15-25 days, free 1 kg sample available on request.

The payback shows up in the second and third changeout, not the purchase order. Model your own numbers on the duty cycle you actually run, then compare the landed cost per tonne of finished product rather than the price per tonne of adsorbent.

Related Products & Technical Resources

Mixing is not recommended for a working bed. The two grades have different equilibrium water capacities at any given temperature, so a mixed bed develops two distinct mass-transfer zones and the bed effectively behaves as two shorter beds in series. This leads to unpredictable breakthrough timing and makes the outlet dew point sensor harder to interpret. The only legitimate mix is a thin layer of 3A on top of 4A in a custom configuration designed to reject co-adsorbing organics in the upper layer; this is rare and should be specified by the dryroom designer with full isotherm data, not improvised at the operator level.

A 3A bed regenerated at 200-220°C and switched on outlet dew point will deliver an outlet dew point in the range of -65 to -75°C in steady state, depending on the inlet RH, the regeneration gas flow, and the bed height. Pushing the regeneration temperature to 250-280°C tightens the achievable dew point by another 5-10 K. For a sustained -75°C dew point target, design the adsorber for 250°C regeneration with a heat-integrated recuperator.

A clean-feed 3A bed will deliver 5-7 years of service before the working capacity drops to the point where the bed can no longer hit the -60°C outlet setpoint in the available adsorption half-cycle. Bed life is shortened by organic solvent breakthrough, oil contamination from upstream fans, and exposure to liquid water (which causes bead fracture on re-adsorption). The first sign of bed aging is a slowly creeping regeneration frequency at constant inlet conditions, which the dryroom controller should track as a trend.

• <a href="/products/activated-alumina.html" class="text-brand underline hover:text-blue-700">Activated Alumina</a> - technical datasheet, available grades and typical applications.
• <a href="/products/3a-molecular-sieve.html" class="text-brand underline hover:text-blue-700">Molecular Sieve 3A</a> - technical datasheet, available grades and typical applications.
• <a href="/products/molecular-sieve-5a.html" class="text-brand underline hover:text-blue-700">Molecular Sieve 5A</a> - technical datasheet, available grades and typical applications.
• <a href="/products/13x-molecular-sieve.html" class="text-brand underline hover:text-blue-700">Molecular Sieve 13X</a> - technical datasheet, available grades and typical applications.
• <a href="/products/zsm-5-zeolite.html" class="text-brand underline hover:text-blue-700">ZSM-5 Zeolite</a> - technical datasheet, available grades and typical applications.
• <a href="/products/pseudo-boehmite.html" class="text-brand underline hover:text-blue-700">Pseudo Boehmite</a> - technical datasheet, available grades and typical applications.
• <a href="/products/catalyst-carrier-alumina.html" class="text-brand underline hover:text-blue-700">Alumina Catalyst Carrier</a> - technical datasheet, available grades and typical applications.
• <a href="/products/aluminium-hydroxide.html" class="text-brand underline hover:text-blue-700">Aluminium Hydroxide</a> - technical datasheet, available grades and typical applications.
• <a href="/blog/3a-molecular-sieve-polyol-dehydration.html" class="text-brand underline hover:text-blue-700">3A Molecular Sieve Polyol Dehydration</a> - companion engineering article on the same duty.
• <a href="/blog/3a-molecular-sieve-refrigerant-drying.html" class="text-brand underline hover:text-blue-700">3A Molecular Sieve Refrigerant Drying</a> - companion engineering article on the same duty.
• <a href="/blog/acid-gas-breakthrough-molecular-sieve-13x-h2s-so2-poisoning-reversibility-thermal-swing.html" class="text-brand underline hover:text-blue-700">Acid Gas Breakthrough Molecular Sieve 13X H2S So2 Poisoning Reversibility Thermal Swing</a> - companion engineering article on the same duty.

Next Steps — Get a Sample or Talk to Our Engineers

Industry-typical regeneration gas flow is 500-1000 Nm3/h for a 1000 kg sieve charge, which gives a superficial velocity in the regeneration direction of about 0.15-0.30 m/s. Lower flow rates risk incomplete regeneration; higher flow rates waste heating energy and can fluidize the bed. The exact number depends on the bed diameter and the regeneration heater capacity.

Yes. CO2 (3.3 &Aring; kinetic diameter) is at the edge of 3A's aperture and is effectively excluded from adsorption at typical dryroom operating temperatures (20-40&deg;C). The 4A grade co-adsorbs CO2 along with water, which contributes to the regeneration heat duty. In a 4A dryroom, the regeneration gas flow has to be sized for the combined water + CO2 inventory.

8x12 mesh (2.36-1.70 mm) is the standard for dryroom adsorbers. 6x8 mesh (3.35-2.36 mm) is acceptable for very large beds where the air flow per square meter of bed cross-section is high and the fan budget is constrained. Anything finer (10x16, 12x20) is reserved for gas-phase drying in non-dusty service (refrigerant, natural gas). Specify attrition &le; 0.2 wt% and bead crush strength &ge; 30 N per bead as the incoming QC limits.

Yes. Nitrogen regeneration is the standard for any adsorber where the dryroom cannot tolerate oxygen ingress into the regeneration loop, or where the regeneration loop is shared with a nitrogen-purification system. The heat capacity of nitrogen is slightly lower than air, so the regeneration gas flow has to be increased by about 5-10% to deliver the same heat input. Most dryroom-grade 3A grades from Aluminaworld are qualified for both air and nitrogen regeneration.

The two reliable indicators are (a) the regeneration frequency starts to climb at constant inlet conditions (the bed can hold less water per cycle), and (b) the outlet dew point starts to drift upward toward the setpoint (the bed can no longer reach the target dryness within the adsorption half-cycle). A weekly trend of the regeneration cycle count and the outlet dew point histogram is the standard monitoring approach. When either trend crosses the threshold, schedule a bed change at the next planned shutdown.

Molecular sieve dust is chemically inert (synthetic zeolite, aluminosilicate) and non-flammable, but it is a particulate that the dryroom's HEPA filtration is designed to capture. A sieve lot with high attrition generates visible dust at the air outlet in the first days of operation and can blind the HEPA filters prematurely. This is why the incoming attrition QC (ASTM D4058, target &le; 0.2 wt%) is worth running on every shipment rather than trusting the supplier's COA.

ASTM D4058 covers the attrition test method. ISO 2076 covers chemical analysis of zeolites. The dryroom classification itself follows ISO 14644 (cleanrooms and associated controlled environments). For the dryroom dew point verification, the chilled-mirror hygrometer reference is typically calibrated against a national standard (NIST in the US, NIM in China, PTB in Germany). The sieve supplier should provide COA aligned to ASTM D4058 and ISO 2076; the dryroom builder is responsible for ISO 14644 conformance.

For Li-ion cell makers evaluating 3A vs 4A for a new dryroom or an adsorber retrofit, the next concrete steps are:

Potassium-exchanged A-type zeolite, 8x12 mesh hard bead, attrition &le; 0.2 wt%, equilibrium H2O &ge; 20 wt%, 200&deg;C regeneration, 5-7 year bed life in dryroom service.

WhatsApp: +86 133 2522 2240

Email: barry@okkinc.com

Free 1 kg sample available on request. MOQ 1 MT. Lead time 15-25 days. COA + MSDS on every shipment. ISO 9001 certified, exported to 60+ countries.

Zibo, Shandong, China Manufacturer of molecular sieve, activated alumina, alumina powder, ATH, pseudo-boehmite and tabular alumina.

We hold stock on hand and ship within 15-25 days of order confirmation. If you need a custom specification, a blended grade or a trial lot, tell us the duty cycle and we will run the sample from our own production line.

Related Articles

• Desiccant Bed Channelling: How to Diagnose Maldistribution from Outlet Dew-Point Sawtooth Traces - AluminaWorld
• Lithium Battery Electrode Coating Oven Solvent Recovery: Activated Alumina vs 3A vs 4A for NMP, DMF and Ethanol
• Molecular Sieve 4A for Natural Gas Pipeline Dew Point Control: Joule-Thomson Cooling, Hydrate Prevention, and ISO 18453 Compliance