316L Stainless Steel Frits
Precision-machined high-purity 316L stainless steel frits, used to retain the packing at each end of a column. Filtration ratings of 1 / 2 / 3 μm, diameters Ø4.7–310 mm, thicknesses 1.5 / 3.0 / 5.0 mm, with 16 stocked part numbers covering analytical, semi-preparative and industrial preparative columns; resistant to organic solvents and to strong acids and bases.
Precision-machined from high-purity 316L stainless steel to retain the packing at each end of a column. Filtration ratings of 1 μm, 2 μm or 3 μm, diameters from Ø4.7 mm to Ø310 mm, thicknesses of 1.5 / 3.0 / 5.0 mm, in 16 stocked part numbers. Resistant to organic solvents and to strong acids and bases, and suitable for self-packed analytical, semi-preparative and preparative columns and industrial DAC columns.
The 16 part numbers below are stocked items, defined bydiameter × thickness × filtration rating; diameters span Ø4.7 – 310 mm, from analytical columns through to industrial DAC preparative columns.Click a part number to add it to your inquiry; where several ratings are offered at the same diameter, choose according to the particle size of the medium —selection criteria are in Chapter 2。
| Part number | Size Ø × thickness | Rating |
|---|---|---|
| 00011-006 | Ø4.7 × 1.5 | 1 μm |
| 00011-033 | Ø4.7 × 1.5 | 2 μm |
| 00011-008 | Ø10.5 × 3.0 | 1 μm |
| 00011-014 | Ø10.5 × 3.0 | 2 μm |
| 00011-005 | Ø21.9 × 3.0 | 1 μm |
| 00011-015 | Ø21.9 × 3.0 | 2 μm |
| Part number | Size Ø × thickness | Rating |
|---|---|---|
| 00011-020 | Ø30 × 3.0 | 1 μm |
| 00011-007 | Ø30 × 3.0 | 3 μm |
| 00011-001 | Ø43 × 3.0 | 3 μm |
| 00011-022 | Ø56 × 3.0 | 1 μm |
| 00011-002 | Ø56 × 3.0 | 3 μm |
| Part number | Size Ø × thickness | Rating |
|---|---|---|
| 00011-003 | Ø63 × 3.0 | 3 μm |
| 00011-009 | Ø93 × 3.0 | 3 μm |
| 00011-010 | Ø127 × 5.0 | 3 μm |
| 00011-018 | Ø200 × 5.0 | 3 μm |
| 00011-027 | Ø310 × 5.0 | 3 μm |
Specifications are subject to the physical product and the supply batch; diameters, thicknesses and ratings outside the table can be made to drawing — please state these when requesting a quotationColumn ID, particle size and packing pressure。
Circular Frits (Bed Supports) — An Illustrated Anatomy
Organic-phase high-pressure preparative DAC columns · key parameters · sintered plate versus woven mesh · materials and applications
The parameters focus on organic-phase high-pressure silica DAC (10–20 µm media, 70–100 bar column pressure, 2 µm class frits); aqueous process columns appear only for comparison.
A Circular Frit, from Whole Disc to Microstructure
The figure below breaks a circular frit into three scales: the actual size of the whole disc, the cross-section of the five-layer mesh stack, and a true-scale comparison between a 2 µm pore and a 10–20 µm packing particle.
Eight Parameters a Frit Specification Must State
A frit performs three functions that conflict with one another, so every parameter is a compromise between them.
① Retention(holding back the packing) calls forsmall；② Distribution(turning one stream into flow across the full cross-section) calls forLowlow flow resistance and large pores; ③ Load bearing(withstanding compression and fluid thrust) calls for athick, stiff plate。
Small pores block easily and raise pressure drop; large pores let packing through and distribute poorly; thick and stiff means more dead volume and harder cleaning.
In practice, most cases of "we fitted a new frit and it still doesn't work" come from treating these three as one selection criterion.
Eight parameters
| Parameter | Typical value for organic-phase DAC | How to read the number | Selection and acceptance points |
|---|---|---|---|
| Nominal pore size | 2 µm | The most commonly misread item. For woven mesh the pore size is ageometric quantity— the diameter of a sphere that just passes through. For a sintered powder plate it is anefficiency rating— a mean flow pore size, or 98% retention of a test dust in a single pass. They are not the same physical quantity. | Two products both rated 2 µmare not directly interchangeable; a quotation must ask "what definition of pore size is this", and request thepore size distribution(the upper limit in particular). |
| pore size distribution | Must be requested; usually not on the datasheet | Datasheets usually give a mean value;what actually determines leakage is the upper limit of the distribution. If the retention layer's pore variation exceeds the smallest bead, packing particles will pass straight through. | Only two qualitative criteria can be cited: the pore size distribution must lie below thesmallest particlein the packing's size distribution, and the pore size must be smaller than the packing'snumber-average particle size。 |
| Outside diameter Ø | Ø50 – Ø1200 mm | Must match the column bore and bepress-fittedto give the lowest dead volume. Any gap at the rim creates edge bypass (channelling). | One-piece seamless discs reach about Ø1150 mm; larger sizes must be segmented, sofrits for columns above Ø1200 always have joints。 |
| Thickness | 0.7 – 1.6 mm (about 0.028″ / 0.062″) | Determines stiffness, and alsodead volume and extra-column band broadening. The way to improve it is to make it thinner or to curve it, not to make it thicker. | Consider thickness and permissible deflection together;the maximum permissible differential pressure and permissible deflection must be requested from the supplier in writing(they are generally absent from public documentation). |
| Porosity | About 21–52% for sintered metal | At the same pore size, higher porosity means lower pressure drop and higher throughput, but lower mechanical strength. | Write it into the purchase specification alongside pore size; sintered plastic parts usually state pore size and porosity together. |
| Permeability | α / β coefficients per ISO 4022 | The only parameter usable for quantitative acceptance. The bubble point standard (ISO 4003) itself states that it "should be regarded as a quality control test, not as a means of classifying filter media or determining exact pore size". | Using a bubble point to "verify that the nominal 2 µm is genuine" does not stand up methodologically; the bubble point can only answer whether there are obvious large pores or perforations. |
| Directionality | Present, but often unmarked | The fine face of a multilayer structure faces one way or the other, and the difference in life can be threefold. None of the three classes of manufacturer marks the upstream face in public documentation. | A drawing with the orientation marked must be requested; it must not be inferred from common sense. Fitting it the wrong way round raises no immediate alarm — it merely shortens life and changes the flow field. |
| Material grade | 316L | "Stainless steel" alone is not enough. Sintered and wrought parts corrode differently, so the grade must be stated and a material certificate requested. | Steel parts to EN 10204 3.1 certificate, with wetted and pressure-bearing parts traceable tobatch level。 |
Red line 1: the ratio rule does not exist
"Frit pore size = 1/2 or 1/3 of the particle size"has no traceable primary source and should not be written into a controlled document。
Working backwards from publicly documented real pairings gives a ratio range of 1/1.5 ～ 1/10, which is not a rule.
| Application | Frit pore size | Medium particle size | Back-calculated ratio |
|---|---|---|---|
| Organic-phase DAC | 2 µm | ≥10 µm silica | ~1/5 |
| Aqueous process column | 10 µm | 30–70 µm | 1/3–1/7 |
| Analytical column | 2 µm | 3–5 µm | 1/1.5–1/2.5 |
A workable alternative: instead of a ratio, keep apairing log— "column type X + medium Y = frit Z", each line noting the source (manufacturer recommendation or own validation) and recording the observed leakage and pressure drop for that pairing.
A pairing table accumulates, is traceable and can be audited; a ratio cannot.
Red line 2: normal pressure ≠ sound frit
An inlet frit that ispartially blocked does not necessarily raise pressure, yet is enough to distribute the sample unevenly — showing up astailing on every peakorsplitting on every peak。
The only quantitative relationship in the published research: at 90% distributor blockage, the volume variance rises about 62-fold; beyond70% blockage there is pronounced asymmetric tailing or additional peaks.
The order of investigation should bepeak shape → frit, not the pressure gauge alone.
Sintered Plate versus Woven Metal Mesh
- The pore is not a "hole" but a tortuous gap between particles
- Retention efficiency varies with flow rate, particle size and concentration
- Local density variation → dead zones
- Irregular pore shapes retain liquid → cleaning and validation are harder
- Sintered 316L is markedly less pitting-resistant than wrought 316L
- Finer pores form concentration cells more readily → worse crevice corrosion
- Pore size is fixed uniquely by the weave geometry
- The fine layer gives the rating and the coarse layers the strength — typically 5–6 layers diffusion sintered into one piece
- Regular channels → better distribution and lower pressure drop
- Contamination collects on the surface → backflushing and cleaning are effective
- Structural rigidity suppresses edge bypass
- The cost is directionality
Nine comparisons
| Criterion | Sintered metal powder plate | Multilayer woven metal mesh (sintered mesh) | What it means in practice |
|---|---|---|---|
| Method of manufacture | Metal powder into a die → pressed → sintered at high temperature into a porous body | Layers of different mesh counts stacked → vacuum diffusion sintered into one piece | The first is a sponge; the second is several sieves welded together |
| Filtration mechanism | Depth filtration: the channels are appreciably larger than the target particle, which is retained by lodging in them | Surface filtration: particles are held on the outer face by the finest layer | The retention efficiency of depth filtrationvaries with flow rate, particle size and concentration— which is where "it never leaks, but it leaked while packing" usually comes from |
| Physical meaning of pore size | An efficiency rating (mean flow pore size / 98% retention) | A geometric quantity (the diameter of a sphere that just passes) | The twocannot be converted directly, so definitions must be aligned before comparing quotations |
| Flow distribution | Tortuous channels with local density variation; the flow profile may deteriorate | Regular channels, more uniform distribution and lower pressure drop | Directly affects efficiency and peak symmetry |
| Tendency to block | Blocks more readily, and blocksinside the pores— backflushing is largely ineffective | Contamination collects on thesurface— backflushing and cleaning are effective | Backflush success is strongly structure-dependent; it is not a matter of operator skill |
| Cleanability | Irregular pores retain liquid, and cleaning ishard to verify | Easier to clean thoroughly | In a GMP setting, "not cleanable" and "not demonstrably clean" are equally serious |
| Mechanical strength and deflection | A single porous body whose stiffness depends on thickness | Coarse support layers carry the load, so deflection resistance can be designed in | Deflection unloads the bed locally and overcompacts it elsewhere → an immediate change in pressure |
| Directionality | Relatively insensitive | Sensitive— the fine face orientation governs life (up to a threefold difference) | A drawing with orientation marked must be requested and recorded in the assembly record |
| Where each is used | Where very high stiffness or special forming (shaped, variable thickness) is needed | The mainstream choice for industrial DAC | One media manufacturer states the position directly in a published DAC packing guide |
A manufacturer-level statement that can be cited
From a media manufacturer's published DAC packing guide:
"the best frits are thewoven meshtype, which distribute flow better and are easier to clean;sintered types block more readily, the flow profile may deteriorate, and they are not as easy to clean as woven types。」
The same document carries another line of equal importance:if a new frit is not used, be sure it has been cleaned thoroughly — a partially blocked frit will inevitably distribute flow unevenly and distort peak shape.
Two further constructions plus three shaped variants
| Type | Key characteristics |
|---|---|
| ③ Fine mesh + coarse support mesh | Fine mesh of 5–10 µm over coarse mesh of about 800 µm, with themesh welded or fused around the circumferenceto eliminate edge bypass |
| ④ Sintered plastic plate (PE / UHMWPE / PP / sintered PTFE) | Corrosion resistant and the first choice at low pH and high salt;hydrophobic, so it must be wetted thoroughly when dry; some manufacturers warn thatair cannot pass through a PE sintered plate, and improper venting may burst it |
| Sintered metal fibre plate | With the fine face away from the packing, life can reach that of a conventional powder plate multiplied by threefold |
| Variable thickness / variable porosity plate | The outer zone is more open, to compensate for lower velocity near the wall |
| Lenticular plate | Thicker at the centre, butthe face towards the bed must be flatso the packing stays level |
Materials
Material selection is determined by three constraints at once — the chemistry of the medium in contact (pH, salt, halides, solvent), the pressure rating of the column, and whether metal ions entering the packing can be tolerated.
| Materials | Typical form | Where it suits and why | Limitations and cautions |
|---|---|---|---|
| 316L stainless steel | Multilayer woven mesh / sintered powder plate | The dominant choice for organic-phase high-pressure DAC. Good chemical resistance and life, taking 70–100 bar packing pressure and organic solvents; in aqueous service it suitslow salt and pH > 5。 | Sintered parts aremarkedly less pitting-resistant than wrought; finer pores form concentration cells more readily and worsen crevice corrosion. High salt, low pH and halide-containing systems carry high risk, and rust particles or leached metal ions entering the packingcan write off a whole column of medium。 |
| Duplex steel(S32205 and similar) | Large-diameter support rings / structural parts | Stronger and more corrosion resistant than 316L | Expensive; used mainly for support rings rather than the filtration layer itself |
| Titanium / Hastelloy | Sintered plate / woven mesh | An upgrade for halide-containing and strongly acidic systems | High price and long lead time; only warranted once 316L has been shown to corrode |
| PE / UHMWPE | Sintered plastic plate | Corrosion resistant andthe recommended material at low pH and high salt; designed specifically where corroding metal column parts is a risk | Hydrophobic, so when dry it must be wetted thoroughly with >80% alcohol(the solvent is flammable and the work must be done in a classified area);air cannot pass through it, and improper venting may burst it; at higher flow rates it may dislodge or bulge.Its pressure rating and solvent compatibility rule it out for organic-phase high-pressure DAC. |
| PP | Woven mesh / moulded distributor | Acid and alkali resistant, low cost | Limited temperature and solvent resistance; not suitable for organic-phase high-pressure service |
| PTFE (sintered) | Sintered plastic plate | The best chemical inertness | Low mechanical strength and pronounced creep; deforms readily under pressure |
| PA / PEEK | Woven mesh / support ring | Common in aqueous process columns, with complete biocompatibility documentation | Pore sizes are generally 10–54 µm, which does not match the 2 µm requirement of organic-phase DAC |
A structural conflict that governs the spares budget
Because the medium is only 10–20 µm, organic-phase DACmustuse a 2 µm class frit — and 2 µm is precisely the size at whichcrevice corrosion risk is highest.
At the same time, an organic-phase columncannotswitch to plastic frits to avoid corrosion as an aqueous column can — neither the pressure rating nor solvent compatibility allows it.
Worse, the cheap and effective mainstay that aqueous process columns rely on to restore throughput — 1 M NaOH cleaning— is entirely closed off on silica DAC (silica generally has a pH ceiling of 8, and high pH rapidly dissolves and degrades the silica support until the bed blocks).
→ That leaves silica DAC frits three options: solvent backflushing (roughly one in three succeeds), offline acid cleaning plus passivation, or replacement. This is not a matter of execution; it is dictated by chemistry.
Documentation required for compliance
| Item | Requirement |
|---|---|
| Material certificate for steel parts | EN 10204 3.1 grade |
| Polymer parts | EN 10204 2.1 / 2.2 |
| Biopharmaceutical applications | USP Class VI (USP <88>), 21 CFR 177 (nylon 177.1500 / olefin polymers 177.1520 / perfluorocarbon resins 177.1550), statement of freedom from animal-derived material |
| Traceability | Pressure-bearing and wetted parts should be traceable tobatch level |
| Surface roughness | Ra characterization does not apply to a porous sintered body; requiring an ASME BPE SF grade on a frit serves no practical purpose and is not worth pursuing |
Position in the Column and Applications
- Verify theposition, part number, pore size and material— four items
- Inspect visually for blockage, discoloration, deformation and rim marks
- For every packing,use new support and filter mesh; always replace after sterilization
- Fill with liquid firstand purge residual air from the frit— leak testing a column containing air is an explosion hazard
(A note: photograph everything as it comes apart — a large column may only be opened once a year)
There is only one point of interception
Compared with seals, frit problems have only one point of interception. A seal problem still gets one more chance at the empty-column pressure hold step; a frit problem essentially does not —
a pressure hold tests whether it leaks, not whetherthe pore size is wrong, whether it is blocked, or whether it will deflect。
The only interception point for the frit is step ① of the seven-step packing procedure, hardware preparation: verify the four items, inspect visually, fit new parts, wet and vent. Let that step through and the remaining six cannot recover it.
Three representative applications
Ordering and acceptance
Six Questions Every Enquiry Must Ask
| # | Question | Point |
|---|---|---|
| ① | Construction and definition of pore size | Are the upper and lower pieces sintered powder plate or multilayer woven mesh? Is thedefinitionof the nominal pore size a geometric quantity, a mean flow pore size, or 98% retention? What is theupper limit of the pore size distribution? — without this, two quotations are not comparable |
| ② | Permissible differential pressure and deflection | What is the maximum permissible differential pressure in the forward andreversedirections? Please supply thedifferential pressure–deflection curve. This data determineswhether backflushing is possible and how high the packing flow rate can go. It is generally missing from public documentation and must be requested in writing; until an answer arrives, keep backflushing and high flow rates conservative |
| ③ | Directionality and drawings | For a multilayer frit,is there an upstream face requirement? Please supply adrawing with the orientation marked. Fitting it the wrong way round raises no immediate alarm — it merely shortens life and changes the flow field |
| ④ | Fit tolerances and clamping torque | Outside diameter to column borefit tolerance、flatness tolerance, permissible edge gap, andspecified clamping ring torque. Without recording the torque, pressure drop baselines across successive rebuilds are not comparable |
| ⑤ | Cleaning methods and the prohibited list | For this model,is backflushing permitted, and at what maximum flow rate? Areultrasonics, acid cleaning or alkali cleaning permitted? What is the？list of prohibited cleaning agents? — particularly important for silica DAC, where the alkali route is closed off by chemistry |
| ⑥ | Replacement criteria (with actual figures) | Bypercentage rise in pressure drop above baseline, bycumulative batches or volume, or by calendar? Please give actual figures. Note that"replace the frit annually, at most every two years" has no published manufacturer source; manufacturers work to event-driven logic |
Acceptance: three low-cost checks (internal methods, not standard conformity methods)
| Method | What it can answer | What it cannot answer |
|---|---|---|
| Single-piece pressure drop baseline(based on the ISO 4022 permeability concept) | Record for a new piece, re-measure on removal, and compare against the history —this is the most valuable of the three | The inherent ±20% uncertainty in the pressure reading must be allowed for |
| Spot bubble point test | Whether this piecehas obvious large pores or perforations— a reject screen only | cannotVerifying that the supplier's nominal 2 µm is genuine; the standard itself limits its use to quality control |
| Visual inspection and weighing | Discoloration, rust, pitting, deformation, rim marks and breaks | The degree of blockage inside the pores |
PM and replacement: event-driven with a calendar backstop
| Trigger | Action | Basis |
|---|---|---|
| Whenever the column has to be repacked | Inspect / clean / replace | Event-driven (manufacturer basis) |
| After autoclaving | Replace immediately | Event-driven (manufacturer basis) |
| Before every packing | Visual inspection of each batch | Internal control |
| Full inspection (including seals and hydraulic oil) | Suggested every 6 months | An internal limit; label it as such |
| Pressure drop above baseline reaches the internal limit | Clean or replace | Must be calibrated from this column's own data |
| Three successive replacement intervals at the same position clearly shortening | Stop replacing parts and go back to root cause | Escalation signal |
Four rules for spares management
- Write the specification down to the— four items— position + part number + pore size and its definition type + material grade
- The upper and lower frits aretwo part numbers, so the log must be kept on both position and part number
- Safety stock is essential— stocked items ship in about 3 days, non-stocked items can take 4–16 weeks, and one blockage can mean 1–4 months of lost production
- Order as a set with theseals, since the filter mesh often has a U-seal pressed directly onto it and physically comes apart in the same operation
Six items missing from public documentation, to be requested from the supplier
| # | Gap | Suggested action |
|---|---|---|
| 1 | A fixed ratio between frit pore size and particle size (1/2, 1/3 and so on) | Replace with a pairing log; do not write it into a controlled document |
| 2 | A calendar replacement interval for frits (such as "annually, at most every two years") | Use event-driven triggers with a calendar backstop |
| 3 | Maximum permissible differential pressure and deflection, and the differential pressure–deflection curve | Request from the supplier in writing; until an answer arrives, keep backflushing and high flow rates conservative |
| 4 | Frit pore size and material for DAC columns above Ø450 | Absent from public documentation; ask the supplier directly |
| 5 | Criteria for reading "rising intercept versus rising slope" on a pressure drop curve | An inference rather than a documented conclusion; verify it yourself |
| 6 | The upstream face requirement for multilayer frits | Request a drawing with the orientation marked |
Items marked "no public source found" (the fixed ratio between frit pore size and particle size, the calendar replacement interval, and the maximum permissible differential pressure and deflection) must be filled in from measured data or a written answer from the supplier before being written into a controlled document.