>
15+ Years Manufacturer| ISO 9001 Certified| Exported to 60+ Countries| MOQ 1 MT| Lead Time 15-25 Days| Free Sample
· 15 min read · Applications

Color Change in Molecular Sieve: A Diagnostic Pathway for Pink, Brown and Black Beds

Aluminaworld is a Zibo, Shandong-based Chinese manufacturer (ISO 9001 certified) producing activated alumina, molecular sieves (3A, 4A, 5A, 13X), pseudo-boehmite and calcined alumina, supplying gas-processing, air-separation and petrochemical customers in 60+ countries with technical datasheets, COA and MSDS on every shipment.

This article is written for the process or maintenance engineer who receives the alarm, opens the adsorbent vessel and has to decide within hours whether to change out the bed, clean the feed, adjust a regeneration temperature, or simply relabel the grade. If you are a purchasing engineer who needs the specification language for a purchase order, go straight to section 11. Everything before it is written to be useful at the vessel, not at the trade show.

1. What pink, brown and black actually mean

A commercial molecular sieve pellet should be white to cream. The zeolite crystal itself is effectively colourless — type 3A, 4A, 5A and 13X frameworks with sodium or potassium cations are optically white and remain so when properly made. Every colour you observe in a bed or a drum sample is therefore extrinsic: deposited, adsorbed, carried in or created after the pellet left the kiln.

That reframes the problem. "Discoloured molecular sieve" is a symptom with at least four unrelated root causes that respond to opposite interventions. A pink bed in a hydrogen PSA unit may be deliberate, functional cobalt-phthalocyanine impregnation needing no action. A black bed in the same unit is usually coke that destroyed pore volume and will not regenerate. A brown bed may be compressor oil you burn off with one thermal swing, or iron-bearing binder that never comes out. Treating all three as "replace" is expensive; treating all three as "regenerate" is worse.

This article is a decision path, not a reference. It runs from what you see in your loading bay, through four cheap on-site checks, to the laboratory methods that confirm the mechanism, then to what each colour costs you in working capacity and what the recovery route is for each. The organising principle is simple: colour is a fingerprint, not a grade. Establish where the colour came from before deciding what to do about it.

2. The chemistry behind each transition

2.1 Pink and salmon: cobalt phthalocyanine or iron

Two very different mechanisms produce a pink or salmon pellet, and they are trivially separated by asking the supplier one question.

The first is intentional impregnation. In hydrogen purification PSA, the 13X bed is routinely loaded with cobalt(II) phthalocyanine, a chemisorption catalyst that binds trace CO, methane and methanol through a coordinated metal-organic complex. It is dispersed on the external pellet surface at loadings suppliers typically quote at 0.5–3 wt%, giving a uniform salmon-pink pellet. The trade-off is a water-capacity loss of roughly 8–12% against unloaded 13X, in exchange for holding CO below the 0.1 ppm level a downstream amine or copper guard requires.

The second is trace iron. Fe₂O₃ from the kaolin or clay binder, or from wear in forming and kiln equipment, disperses as sub-micron particles through the pellet. Because the pellet is an ionic white solid, even a small iron oxide loading turns it visibly tan or salmon. Good-quality 13X specifications put iron oxide well below 0.05 wt%, tightest at 0.03 wt% or lower. Iron is not only a colour problem: it occupies exchange sites, depresses total exchange capacity and accelerates hydrothermal framework degradation during steam exposure. Unspecified pink is a defect, not a feature.

The distinction matters operationally. Cobalt-phthalocyanine pink is uniform through the pellet body, does not bleed into water and does not transfer to a white cloth. Iron pink tends to concentrate where the binder is richest and can show a mottled or speckled appearance under a loupe.

2.2 Brown and tan: adsorbed organics, oxidised binder or degraded framework

Brown is the most common complaint and the most frequently misdiagnosed, because three distinct mechanisms end up in the same colour band.

The dominant industrial mechanism is adsorption of heavy hydrocarbons. Compressor oil carryover, process condensate and heavier feed fractions adsorb on the external surface and, in a bed, are slowly catalytically cracked into carbon. An oil loading of only 0.1–0.5 wt% turns white beads visibly amber, and continued exposure pushes amber toward opaque brown. Because the colour is adsorbed rather than structural, it responds to heat and purge.

The second mechanism is intrinsic binder colour. Some binder systems, particularly those using iron-bearing clays or deliberately coloured tracer binders, produce pellets that are tan or buff from the moment of manufacture. Such pellets darken further on high-temperature regeneration because the binder's organic component chars. The colour here is permanent and the pellet's thermal stability is already compromised.

The third is hydrothermal framework degradation. Extended exposure to liquid water and steam at elevated temperature progressively destroys the crystalline framework, converting it to amorphous aluminosilicate gel. The material turns opaque, milky and often tan to brown, loses its crystalline signature in X-ray diffraction, and loses most of its adsorption capacity. This is fatal and is frequently misread as "the sieve is just dirty".

2.3 Black and grey: coke, burnt binder and metal contamination

Black is the serious end of the sequence and usually means the bed is finished.

Coke is the most common cause. The strong-acid Brønsted sites of 13X and 5A catalyse condensation and cracking of adsorbed hydrocarbons, and if regeneration is pushed high — typically above about 350 °C on a hydrocarbon-contaminated bed — the carbon is fixed rather than removed. Carbon blocks pore mouths permanently; the pellet loses water and CO₂ capacity, and the bed goes from brown to matte black with a sooty texture that rubs off on a glove.

Burnt binder is the second cause, and more common than plant engineers expect. Kaolin binders contain organic deflocculants and binding aids. A regeneration fire, air admitted into a hot regeneration loop, or a prolonged excursion above the specified bed temperature chars the binder and turns the pellet grey to black. The pore structure may still be intact, but mechanical strength is gone and the pellet generates fines in service.

Grey and metallic silver are rarer but worth naming: they suggest metal contamination — brass or bronze fines from a worn valve seat, stainless particles from a failed seal face, or a grey anti-caking additive.

One final black scenario is entirely mechanical: carbon black or soot ingress from an upstream process. Soot deposits as a fine powder on the pellet surface, is not adsorbed into the pore structure, and can largely be removed by screening and washing. It looks identical to coke in the vessel but behaves completely differently under regeneration.

3. The on-site decision tree

Before any laboratory work, run these four checks in order. They cost nothing but an hour and they will usually narrow the field to one mechanism.

Step 1 — Establish provenance. Where did you first see the colour — at incoming inspection, after the first regeneration, or only after months in service? Colour present on arrival is supply-side; the supplier's retained sample settles it in a day. Colour appearing only in service is process-side and points to feed contamination or a regeneration fault.

Step 2 — Establish distribution. Is the colour uniform across every pellet, or is it graded? A uniform colour across the whole vessel points to bulk supply or binder. A gradient that darkens toward the inlet end points to a feed contaminant entering with the process stream. A dark layer confined to the top of the bed points to an upstream vessel or line that is venting directly into the adsorber — a broken dip pipe, a failed check valve or a leaking relief.

Step 3 — The water-bleed test. Take five pellets, add 20 mL of deionised water in a clear tube, shake 30 seconds. Colour diffusing into the water indicates a soluble contaminant — iron salts, a dye, a surfactant or a corrosion inhibitor. Colour staying on the pellet with clear water indicates an insoluble deposit: oil, coke, carbon black or framework degradation.

Step 4 — The rub test. Rub a pellet vigorously against a white cotton swab. A brown or black streak means a surface deposit that is transferring, which points to oil, soot or a loosely adhering carbon. A pink streak means a surface-bound colourant or a soluble species. No transfer at all means the colour is inside the pellet body and the cause is in the manufacturing or in the framework.

CheckWhat you observeMechanism indicated
Water bleedWater stays clearOil, coke, soot, framework damage
Water bleedWater clouds and coloursSoluble iron salt, dye, corrosion inhibitor
Rub testHeavy streak on swabSurface deposit: oil, soot, loose carbon
Rub testFaint pink streakSurface-bound colourant or soluble species
Rub testNo transferIntrinsic: binder colour, Fe in matrix, framework damage
Spatial gradientDarker at inlet endFeed-borne contamination entering the bed
Spatial gradientDark only at bed topDirect gas ingress from upstream vessel or line

4. Confirming the mechanism in the laboratory

Once the on-site checks narrow the field, four laboratory methods confirm it. These are the methods worth naming in an incoming inspection instruction, because they distinguish causes rather than merely measure colour.

Thermogravimetric analysis is the most informative test here. A 20–50 mg sample is heated in a defined atmosphere and mass loss is attributed to moisture, then organic load, then framework decomposition. An organic load of several percent coming off between 200 and 400 °C identifies adsorbed oil or heavy hydrocarbons. A sudden loss at 600 °C and above identifies a burnt binder. Comparing against a retained incoming sample from the same supplier gives a direct before-and-after number.

ICP-OES or XRF elemental analysis resolves the pink question definitively. Reporting iron, cobalt, nickel, copper, sodium, calcium and magnesium on an oxide basis gives you: a cobalt figure that confirms intentional phthalocyanine impregnation, an iron figure that quantifies contamination, and a sodium figure that confirms the cation exchange grade. An iron content in the 0.05–0.2 wt% range alongside a pink pellet is a supply-side deviation, full stop.

X-ray diffraction answers the framework question. A sound 13X pellet shows the characteristic zeolite diffraction pattern with high crystallinity. A degraded pellet shows diffuse amorphous background in place of the crystalline peaks. Diffraction will not detect light adsorbed oil — it detects the destructive case, and it is essential to run it before you commit to an aggressive high-temperature recovery that might destroy an otherwise sound crystal structure.

Quantitative adsorption capacity closes the commercial loop. Run the test appropriate to the grade: ASTM D3464 for equilibrium capacity of 13X, ASTM D3616 for 3A, 4A and 5A, and ASTM D3159 for water content of 3A, 4A and 5A. Pair the capacity number with the colour reading and you have a defensible acceptance or rejection argument that a colour swatch alone can never give you.

MethodStandard / basisDiscriminatesDecision it supports
Thermogravimetric analysisMass loss vs temperatureOrganic load, coke, binder charIs heat treatment recovery viable?
ICP-OES / XRFOxide-basis elementalFe, Co, Ni, Cu, NaSupplier deviation vs CoPc grade
X-ray diffractionCrystallinity indexAmorphous framework damageRecover or replace the bed
Equilibrium capacityASTM D3464 / D3616 / D3159Working capacity retainedAccept, blend or reject the lot
ColourimetryASTM D2244 / ISO 11664-4Objective L*a*b* driftTrend monitoring, lot acceptance
Attrition & crushASTM D4058 / D7084Mechanical integrity after heatSafe to return to service?
Particle sizeASTM D2180 / ISO 1338Fines generation, attritionPressure drop and dust risk

5. What each colour is costing you in capacity

Colour is a symptom; capacity loss is the cost. These are industry-typical figures for the change-out justification — replace them with your own measured values before committing to a purchase order.

For a 13X bed, typical industrial working capacity on a water-and-CO₂ duty is quoted in the region of 14–16 wt% at standard conditions with a 4–6 bar regeneration. A pink, cobalt-impregnated 13X loses roughly 8–12% of that figure, so a bed specified at 15 wt% will realistically deliver 13.3–13.8 wt%. That is not a fault; it is the price of the CO capability and it should be in your design basis from day one.

Oil carryover is the most common recoverable loss. A 0.2–0.5 wt% oil loading typically costs 10–20% of water capacity and poisons the hydrocarbon co-adsorption a 13X bed performs in PSA service. Coke is the worst case: once carbon blocks the pore mouths, loss frequently exceeds 40% and no regeneration restores it, because the damage is chemical rather than physical.

Hydrothermal framework degradation is a different category, because the loss is progressive and irreversible. A 13X bed exposed to liquid water at high temperature can lose 30–50% of capacity over a few hundred hours of severe exposure. This is why hydrogen PSA design codes require the regeneration gas to stay above the product dew point at all times, and why a bypass or interlock failure is treated as a major incident rather than an inconvenience.

CauseTypical loadingCapacity impactReversible?
Cobalt phthalocyanine (intentional)0.5–3 wt% CoPc−8–12% on 13XBy design — not reversible
Trace iron in matrix0.05–0.2 wt% Fe₂O₃−5–15%No — supply deviation
Compressor oil carryover0.1–0.5 wt% oil−10–20%Yes — thermal swing
Heavy hydrocarbon / coke2–10 wt% carbon−30–50%+Only if coke is soft and recent
Burnt binderSurface char−10–25% plus attritionNo — mechanical loss
Hydrothermal framework damageAmorphous gel−30–50%No — replace the bed
Carbon black / soot ingressSurface powder−5–15%Yes — screen and wash

6. Recovery routes by mechanism

6.1 Recovering an oil or light-hydrocarbon fouled bed

This is the case where recovery works and you should not replace the bed. The sequence is a temperature swing with inert or low-hydrogen purge, hot enough to desorb the oil but below the coking threshold. For 13X in hydrogen service, heat the regeneration loop to 200–250 °C under nitrogen or hydrogen purge, hold 4–8 hours, then cool under purge. The key is not to let oil sit in the bed at high temperature with no flow — that combination is precisely what converts recoverable brown into unrecoverable black.

Verify thermogravimetrically before and after. If organic load falls from, say, 3.2 wt% to under 0.3 wt% and the colour returns to white, the bed is recovered and you have saved the change-out. If the load falls but the pellet stays amber, the colour was never oil — go back to the decision tree in section 3.

6.2 Recovering a soot-fouled bed

Soot and carbon black are surface deposits, not pore fillings, and respond to physical rather than thermal treatment. Screen off the top of the bed; where the deposit is significant, lift the bed, wash with hot detergent solution, rinse thoroughly and re-screen to remove broken pellets generated in handling. Before returning it to service, run attrition and crush strength. A washed and re-screened bed is mechanically weaker than a new one, because screening itself generates fines, and fines in a PSA bed mean channeling.

6.3 Why a black bed is usually not recoverable

Once the pellet is matte black with a sooty texture, the carbon is either inside the pore structure or strongly bound to the binder. There is one qualified exception: if the black material is genuinely soot that arrived as a powder and the pellet underneath is still white, screening and washing will work. The rub test from section 3 separates the two in five minutes — a sooty streak on the swab means surface powder, a clean swab with a still-black pellet means carbon in the body.

For a genuinely carbon-loaded 13X bed in hydrogen service, regeneration at 300–350 °C under purge removes soft, recently deposited carbon but will not restore blocked pore mouths. Compare post-regeneration capacity against the incoming certificate: below about 85% of certified working capacity, the bed is finished and should be changed out. Running a degraded 13X bed in PSA service costs product purity and cycle stability long before it costs capacity.

7. Species-specific behaviour and the standards that cover it

7.1 Where discolouration shows up first

Feed gas composition largely determines which grade discolours first. A 3A bed on instrument air sees essentially nothing but water and stays white for years unless compressor oil gets past the upstream coalescer, in which case it turns amber within weeks. A 4A bed on process air sees light hydrocarbons and takes on a tan cast. A 5A bed in hydrogen purification sees trace CO and CO₂ plus any compressor oil, and is most often supplied in the pink phthalocyanine form. A 13X bed in a PSA unit sees the worst of everything — water, CO₂, hydrocarbons, and ammonia slipping through from an upstream synthesis loop — so it turns brown first and black fastest.

Different grades behave differently under the same discolouration, and the standards defining each grade are the ones your incoming inspection should reference.

3A molecular sieve, used for the final water stage in air drying, is the most colour-sensitive grade in service because it runs in a clean, low-hydrocarbon stream. Any brown discolouration of a 3A bed almost always points to compressor oil or hydrocarbon slip, and it is the grade where ASTM D3159 water content is most valuable. 4A behaves similarly but sees more hydrocarbon. 5A in hydrogen purification is the classic candidate for intentional cobalt phthalocyanine impregnation, and pink-to-salmon on a 5A pellet in a hydrogen PSA unit is expected, not a defect. 13X shows the widest colour range because it sees the most demanding feeds.

For activated alumina the comparable references are EN 12904 (evaluation for water treatment) and EN 12915 (evaluation for gas drying). For molecular sieve gas drying, ASTM D3616 covers equilibrium capacity of 3A, 4A and 5A, and ASTM D3464 covers 13X. Particle size is ASTM D2180 or ISO 1338; attrition index for 13X and 5A is ASTM D4058, with bulk crush strength by ASTM D7084.

Colour is best specified objectively rather than descriptively. Instead of "white to cream, no discolouration", write an acceptable L*a*b* range under a defined illuminant, with a retained reference panel. ISO 11664-4 and ASTM D2244 give the method. This single change turns a subjective dispute into a measurable acceptance criterion, and it belongs in every adsorbent purchase order you write from now on.

8. Preventing the discolouration in the first place

Every discolouration investigation arrives at the same conclusion: the colour was preventable. Four upstream measures account for the overwhelming majority of cases.

Protect the bed from oil. Compressor oil carryover is the most common cause of brown beds and the cheapest to fix. Fit a coalescer and carbon adsorber ahead of the adsorber inlet, keep the coalescer within its service interval, and monitor its differential pressure rather than changing it on a calendar. A coalescer past capacity passes oil in bursts, and a burst of oil is what kills a bed.

Keep water out of the regeneration system in the off-nominal case. The hydrothermal damage pathway begins when liquid water reaches a hot bed. Interlock the regeneration heater to the product dew point, and confirm the interlock by testing it, not by reading the logic. A regeneration loop that admits liquid water during a shutdown and then re-heats the bed will destroy 13X in a single event.

Control regeneration temperature against coking, not throughput. The classic mistake is raising temperature to shorten cycle time or recover an overloaded bed. This is precisely how recoverable brown becomes unrecoverable black. If the bed cannot be properly regenerated at the specified temperature, the answer is more regeneration flow or a larger bed, not a hotter one.

Screen the incoming material. Ask the supplier for the colour specification and the iron content on the COA, and check both at incoming inspection against a retained sample. If a supplier's product arrives pink without a documented CoPc specification, you have a supply-side deviation and you should reject it before it is loaded, not discover it eighteen months later.

9. Sampling a discoloured bed: getting a sample you can act on

Most discolouration investigations fail not because the diagnosis was hard but because the sample was wrong. A discoloured bed is inherently a graded medium, and a sample taken from a bucket at the loading bay tells you nothing about the distribution inside the vessel. The sampling procedure therefore matters more than the laboratory method.

9.1 Take a graded sample, not a grab sample

Sample at three elevations — top, middle and bottom — and keep them separate and labelled. The gradient is itself diagnostic. A bottom sample darker than the top indicates a feed contaminant carried forward by the process stream. A top sample darker than the bottom indicates direct gas ingress from the vessel above, a piping or valve problem rather than an adsorbent quality problem. A uniform sample across all three elevations points upstream of the bed entirely, to the supplier or the manufacturing process.

9.2 Handle the sample correctly

Two handling mistakes corrupt the result before the laboratory sees it. First, moisture pickup: 13X and 5A are strongly hygroscopic and re-absorb atmospheric water in an open container, so a sample left in a jar overnight shows a water capacity unrelated to the colour. Seal it in a moisture-tight container immediately and record the sealing time. Second, cross-contamination: never handle a fresh incoming sample with the same bucket or glove used on a discoloured one. Trace oil or iron transfers readily.

9.3 Establish the chain of custody

Whatever the outcome, the record has to hold up later. Reference ASTM E300 for sample handling and chain of custody, and pair it with a sealed reference portion retained from the same delivery. That sample is the only thing that proves, months later, that the pink pellet came from the supplier rather than your own process. Our reference procedure is set out in the field sampling guide linked at the end of this article.

10. Frequently asked questions

Is pink molecular sieve always a defect?

No. Pink is expected on a 5A or 13X grade deliberately impregnated with cobalt(II) phthalocyanine for CO removal in hydrogen purification. The colour is uniform and does not transfer to water or a swab. Pink that is mottled, bleeds into water, or arrives with no documented cobalt specification is a defect — usually trace iron from the binder, typically 0.05–0.2 wt% Fe₂O₃. Request an ICP-OES iron result before accepting.

Can a brown or black molecular sieve bed be regenerated back to working condition?

It depends on the mechanism, which is why the four on-site checks come first. A brown bed fouled with compressor oil or light hydrocarbons is usually recoverable with a 200–250 °C temperature swing under nitrogen or hydrogen purge for 4–8 hours. A black bed with carbon inside the pellet is not — loss exceeds 40%. A grey or milky bed showing amorphous framework damage is also not recoverable. Let TGA decide: below about 0.3 wt% organic load and capacity within 85% of certified means recoverable; otherwise change it out.

What temperature causes the pink-to-brown-to-black progression?

Two mechanisms drive it. Surface adsorption of oil and hydrocarbons begins at ordinary service temperatures, with amber appearing as loading passes roughly 0.1–0.5 wt%. The brown-to-black transition is catalytic cracking: the strong-acid sites of 13X and 5A condense hydrocarbons into carbon, accelerating sharply above about 350 °C. Regenerating a hydrocarbon-fouled bed at high temperature without enough purge flow is the most common way to convert recoverable brown into unrecoverable black. Stay at or below the specified temperature and add flow or bed size instead.

Is a discoloured bed automatically a rejected lot at incoming inspection?

Not automatically, but it should never pass unexamined. Discolouration on arrival is a supply-side question, and the supplier's retained sample from the same batch resolves it within a day. Record an objective colour reading by ASTM D2244 or ISO 11664-4, request the ICP-OES iron and cobalt figures, run the equilibrium capacity test for the grade (ASTM D3464 for 13X, ASTM D3616 for 3A, 4A and 5A, ASTM D3159 for water content), and compare against the published specification.

Does a light-coloured bed still prove the feed gas is clean?

No, and this is a common misreading. A white bed does not prove a clean feed; it may simply not have been contaminated yet. More importantly, a brown bed does not tell you the feed is dirty either, because the same colour can arise from an upstream failure that has already been fixed. The reliable indicators of feed quality are upstream of the adsorber: compressor oil carryover measured at the bed inlet, hydrocarbon content on the feed analyser, and the differential pressure trend across the inlet coalescer. Colour is a diagnostic of the adsorbent; the feed instruments are the diagnostic of the process.

What is the difference between coke and carbon black in a discoloured bed?

Coke forms inside the adsorbent by catalytic condensation of hydrocarbons on the acidic sites of 13X and 5A. It is hard, bonded and fills the pore mouths; it does not wash off and does not come off in a thermal swing once hardened. Carbon black or soot arrives as fine powder on the external pellet surface. The rub test separates them in seconds: coke leaves a faint mark, soot a heavy black streak. Soot is recoverable by screening and washing; coke generally is not.

Can I blend a partially discoloured lot with fresh sieve to save money?

Only with care and with data. Blending is reasonable when the discolouration is cosmetic and measured capacity of the used material is within about 90% of the certified value of the new material. It is not reasonable when the mechanism is framework degradation, when mechanical strength was compromised by binder char, or after a thermal excursion. Before any blend, run attrition by ASTM D4058 and bulk crush strength by ASTM D7084, because fines from a weakened bed cause channeling downstream.

How much capacity does a pink cobalt-impregnated 13X bed actually lose?

Industry-typical figures put the water-capacity penalty at roughly 8–12% relative to unloaded 13X. A grade specified at 15 wt% equilibrium water capacity realistically delivers 13.3–13.8 wt% in service, and the certificate should state the capacity of the impregnated grade, not the unloaded parent. This is not a defect and not recoverable — it is the price of the CO capability. If your design basis assumed the unloaded figure, your bed sizing is wrong by 8–12%.

My bed is discoloured only in the top layer. What should I check first?

Start with the upstream vessels and piping that vent into the adsorber. A dark layer confined to the bed top is the signature of direct gas ingress: a broken dip pipe, a failed non-return check valve allowing product-side backflow, or a leaking relief valve passing contaminated gas straight to the adsorber. Check the inlet nozzle and the internal distribution hardware on top of the bed first. In most top-only cases the adsorbent is a symptom and the piping is the cause.

Which standard should I cite in the purchase order for colour and quality?

Cite four things. For performance, the equilibrium capacity method for the grade: ASTM D3464 for 13X, ASTM D3616 for 3A, 4A and 5A, ASTM D3159 for water content on 3A, 4A and 5A. For mechanical integrity, ASTM D4058 attrition and ASTM D7084 bulk crush strength. For particle size, ASTM D2180 or ISO 1338. For colour, an objective L*a*b* range under a stated illuminant by ASTM D2244 or ISO 11664-4, plus iron content on an oxide basis on the COA.

Is a milky or opaque pellet different from a brown pellet?

Yes, and it is the most serious presentation. A milky, opaque, almost translucent pellet indicates the crystalline framework has partially collapsed to amorphous aluminosilicate gel, usually after prolonged exposure to liquid water at elevated temperature. This is hydrothermal degradation, it is irreversible, and capacity losses of 30–50% are typical. X-ray diffraction shows loss of the zeolite peaks and a diffuse amorphous background. Such a bed must be changed out.

Does discolouration affect activated alumina the same way it affects molecular sieve?

The colour presentations are similar but the mechanisms differ, because the two families have very different surface chemistry. Activated alumina is a transitional alumina with 2–10 nm mesopores and a strongly hydrophilic surface that adsorbs oil and water readily. It rarely shows binder pink because it is not made with a clay binder, and shows brown and black mainly from oil carryover and thermal decomposition of surface hydroxyls. It regenerates more readily than a zeolite because there is no crystalline framework to destroy.

Next Steps

Send us the service history of the discoloured bed — grade, service dates, feed composition, regeneration temperature and purge flow, and whether the colour was present on arrival or appeared in service — and we will return a diagnostic assessment with the laboratory methods we recommend, a replacement grade proposal, a 1–5 kg sample for bench comparison against your own data, and COA and MSDS documentation with the trial lot. Typical response within 24 hours.

WhatsApp +86 133 2522 2240 Email barry@aluminaworld.com
Molecular Sieve from Aluminaworld
Featured Product

3A / 4A / 5A / 13X Molecular Sieve — Specified Purity for Drying and Gas Purification

Zeolite 3A, 4A, 5A and 13X supplied with COA stating equilibrium capacity, iron on oxide basis and particle size — so colour and capacity are verified on arrival, not discovered in service.

View Product Details →

Related technical reading