The Gold Standard of Zeolite Catalysis | Why Diffusion Can Change an Apparent Activation Energy

Let k(T) be a local first-order reaction coefficient at absolute temperature T. A rate measured per unit volume of porous catalyst can be smaller than k times the surface concentration because reactant concentration falls inside the specimen. We will describe that reduction by an effectiveness factor η.

Define Kobs = kη. This is the coefficient multiplying a fixed, specified surface concentration in our volume-averaged rate. It is not automatically the intrinsic k. Before interpreting a temperature plot, state which coefficient the observations actually determine and which concentration supplies the rate’s reference.

The distinction matters because a measured rate can also change when surface concentration, active-site inventory or the measurement response changes. Our first model holds those complications outside the account. We focus on one question: what does internal diffusion alone do to the temperature slope when local chemistry and diffusion each have specified temperature dependence? The answer remains conditional on that deliberately small contract.

Contents · Next chapter

Consider an isothermal slab whose centre plane is at x = zero and exposed surface at x = L. L is a half-thickness, not the full slab thickness. Assume symmetry at the centre, a fixed reactant concentration Cs at the surface and a constant effective diffusivity D throughout the slab at each temperature.

Assume local first-order consumption kC, with no external film resistance, heat gradient, changing pore structure or deactivation. The surface concentration used to normalise the rate is known and fixed for the comparison. Those assumptions produce a tractable reaction–diffusion problem; they do not describe every catalytic particle.

A real specimen may require a different geometry or additional transport and thermal relationships. State that difference before importing a slab equation into a sphere, a crystal network or a reacting bed. The educational purpose is to trace one observation from a defined concentration profile, so a learner can identify which part of the inference would change when the physical model changes.

Contents · Previous chapter · Next chapter

At steady state in the stipulated slab, D d²C/dx² − kC = zero. The boundary conditions are dC/dx = zero at the centre and C = Cs at x = L. The first condition expresses symmetry; the second supplies the surface state rather than assuming that every interior site experiences it.

Define φ = L√(k/D). It is dimensionless because k has inverse-time units and D has length-squared-per-time units. A large φ can arise from a longer diffusion path, a larger local reaction coefficient or a smaller diffusivity. Its value is a comparison of processes, not an independently measured chemical rate constant.

The concentration profile is C(x)/Cs = cosh(φx/L)/coshφ. Differentiate it to verify the governing equation and both boundary conditions. The profile is smallest at the centre and reaches the supplied surface value at x = L. A correct solution should preserve that physical direction before it is used to calculate an average rate.

Contents · Previous chapter · Next chapter

The volume-averaged consumption rate for this symmetric slab is k times the average concentration. Divide that rate by kCs, the hypothetical rate if the entire slab experienced the surface concentration. The resulting effectiveness factor is η = (1/L)∫[C(x)/Cs]dx from zero to L.

Integrating the hyperbolic-cosine profile gives η = tanhφ/φ. This relationship uses our first-order, isothermal slab assumptions. It is not a universal effectiveness formula for all particle shapes or reaction orders. The definition of L also belongs to the result.

For positive φ, η lies between zero and one. When φ is small it approaches one, while for large φ it approaches 1/φ. These limits help check the calculation. A smaller effectiveness factor does not mean that each surviving active site has become chemically slower. It means that the stipulated specimen uses less of the hypothetical uniform-concentration rate because the internal concentration profile is different.

Contents · Previous chapter · Next chapter

Choose a fictional reference temperature T0 = 350 K. At that state, stipulate k0 = 1 per second, D0 = 1 square micrometre per second and L = 1 micrometre. Then φ0 = one, η0 = tanh(1), approximately 0.76159, and Kobs is approximately 0.76159 per second.

The local coefficient remains one per second. The observed coefficient is smaller because the internal concentrations are below Cs. These figures are original model values, not adsorption, cracking or conversion measurements for a supplied catalyst.

An independent check uses the centre concentration: C(0)/Cs = 1/cosh(1), approximately 0.64805. The average concentration ratio should lie between that centre value and the surface value one. The calculated η does. Checking the profile and its average separately helps locate an error in a boundary condition or thickness convention before it becomes an apparently plausible fitted activation energy.

Contents · Previous chapter · Next chapter

Assume k(T) = k0 exp[−Ek/R(1/T − 1/T0)] and D(T) = D0 exp[−ED/R(1/T − 1/T0)]. Ek and ED are constant parameters of this teaching model. We use R = 8.314 joules per mole per kelvin, so the energies must be expressed in joules per mole inside these exponentials.

The reference forms make k(T0) = k0 and D(T0) = D0 exactly. They also avoid introducing unexplained prefactors. If Ek and ED are positive, both k and D increase with temperature, but they need not increase by the same factor.

Real effective diffusion can have a more complicated temperature dependence. Adsorption and changing coverage can also affect a measured rate. Here we stipulate the two Arrhenius forms so that their competing effects can be derived explicitly. The model’s value is that it identifies what a measured temperature slope would mean under those assumptions, while keeping other possible explanations available for a real investigation.

Contents · Previous chapter · Next chapter

Combining the two laws gives φ(T) = φ0 exp[−(Ek − ED)/(2R)(1/T − 1/T0)]. Geometry is fixed, so the difference between the two energy parameters determines how reaction and diffusion compete as temperature changes.

If Ek exceeds ED, φ rises with temperature: local reaction accelerates more strongly than the diffusive supply represented by D. Effectiveness then decreases, even though diffusivity itself is increasing. There is no contradiction; the relevant question is their relative temperature response.

If Ek is smaller than ED, φ instead decreases as temperature rises. If the two energy parameters are equal, φ stays constant and η contributes no additional temperature slope. These cases show why a statement such as diffusion always becomes less limiting at a higher temperature is incomplete. State which rates are changing and compare their laws, rather than infer the direction of a combined response from one parameter alone.

Contents · Previous chapter · Next chapter

Define the local apparent activation energy as Eapp = −R dlnKobs/d(1/T), with geometry and the rate’s concentration basis held fixed. It is a derivative at a specified state. A finite-range fitted slope is a related but different quantity that we will consider later.

A logarithm can be understood as ln(Kobs/Kref) for a fixed positive reference coefficient carrying the same units. Changing that fixed reference adds a constant to the plotted logarithm and leaves its slope unchanged. Mixing different rate bases at different temperatures would not have that harmless effect.

The definition also shows why the observable matters. If one plots a total rate whose surface concentration or active amount varies with temperature, their changes contribute to the slope. The resulting number should not be described as a clean estimate of the kη coefficient without the corresponding normalisation. Establish the measured quantity before attaching a chemical interpretation to its fitted gradient.

Contents · Previous chapter · Next chapter

Because Kobs = kη, lnKobs = lnk + lnη. The first contribution gives Ek under the stipulated Arrhenius law. The second contribution carries the temperature dependence of the concentration profile through φ.

For η = tanhφ/φ, differentiation gives dlnη/dlnφ = 2φ/sinh(2φ) − 1. Meanwhile dlnφ/d(1/T) = −(Ek − ED)/(2R). Applying the chain rule combines these relationships without treating η as constant when temperature changes.

The exact result is Eapp = (Ek + ED)/2 + [(Ek − ED)/2][2φ/sinh(2φ)]. Every term has energy-per-mole units, while the bracketed multiplier is dimensionless. This is the central model result. It connects a local temperature slope to reaction, diffusion and the current degree of internal limitation rather than assigning every measured slope directly to a molecular reaction step.

Contents · Previous chapter · Next chapter

As φ tends to zero, sinh(2φ) approaches 2φ, so the bracketed multiplier approaches one. The expression returns Eapp approaching Ek. This is consistent with η approaching one and Kobs approaching the local coefficient.

As φ becomes large, 2φ/sinh(2φ) tends to zero. The apparent energy approaches (Ek + ED)/2. If diffusion’s activation parameter is neglected, that limit reduces to Ek/2. The familiar half-energy statement therefore has an additional assumption about diffusion’s temperature response.

These limits should be checked before evaluating intermediate cases. They also prevent an approximate regime statement from becoming a universal rule. A specimen with finite φ need not sit exactly at either limit. A measured slope near a limiting value is compatible with the model but does not by itself prove that the real system follows all of its geometry, concentration and thermal assumptions.

Contents · Previous chapter · Next chapter

Take Ek = 80 and ED = 20 kilojoules per mole, while retaining the reference geometry and coefficients that give φ = one at T0. The half-sum is 50 kJ/mol and the half-difference is 30 kJ/mol.

At φ = one, 2φ/sinh(2φ) is approximately 0.55144. Therefore Eapp is approximately 50 + 30 × 0.55144 = 66.54 kJ/mol. The intrinsic parameter is still 80. The calculated apparent value is smaller because effectiveness changes with temperature around that state.

Do not interpret 66.54 as a new activation barrier that replaced the original chemical parameter inside each site. It is the temperature sensitivity of the combined coefficient kη. The distinction is useful when comparing catalysts or particle forms: an observed change in slope can reflect a changed transport contribution even if the stipulated local chemical law has not changed.

Contents · Previous chapter · Next chapter

At the same reference temperature, increase the half-thickness from one to three micrometres while leaving k0 and D0 unchanged. The new φ is three. Its effectiveness is tanh(3)/3, approximately 0.33168, and its observed coefficient is approximately 0.33168 per second.

The multiplier 2φ/sinh(2φ) is now approximately 0.02975. With the same energy parameters, Eapp is approximately 50.89 kJ/mol. Geometry changed both the apparent coefficient and its temperature sensitivity, while the intrinsic k law remained the same.

The comparison is volume-normalised under the stated model. A thicker physical specimen also contains more material if its area is fixed, so total production would need its own size account. Preserve that denominator before comparing two real outputs. A rate per volume, a rate per mass and a total reactor rate should not be exchanged merely because each is described as catalyst activity.

Contents · Previous chapter · Next chapter

Reduce the reference half-thickness instead to 0.1 micrometre, retaining the same local coefficients. Then φ = 0.1, η is close to one, and the multiplier 2φ/sinh(2φ) is close to its small-φ limit.

The apparent energy is therefore close to 80 kJ/mol, although it is not exactly equal for finite φ. Shortening the model diffusion path makes the measured coefficient a closer reflection of local chemistry. It does not change the intrinsic chemical energy by itself.

This is a mathematical comparison, not an instruction to prepare, crush or operate a catalyst. Changing real particle size can alter more than one property, including accessibility, external area, heat transfer and the relevant normalisation. The useful experimental reasoning is to ask which alternatives a geometry comparison would need to control before interpreting a change as evidence about internal diffusion alone.

Contents · Previous chapter · Next chapter

Our first examples had Ek greater than ED. Now reverse the energy ordering in another fictional model: Ek = 20 and ED = 80 kJ/mol. The small-φ limit of Eapp is 20, while the large-φ limit is 50.

An effectiveness factor can remain below one at every state while contributing a temperature dependence that raises the apparent energy above Ek. There is no contradiction between a multiplicative reduction in a coefficient and an increase in its logarithmic temperature sensitivity. The value of a function and the slope of its logarithm are different quantities.

This example limits the claim that diffusion must always halve or lower an apparent activation energy. Under specified common assumptions it often lowers the slope relative to a more strongly activated local reaction, but the general result should be read from the chosen k and D laws. Explain the parameter ordering instead of using a universal arrow of change.

Contents · Previous chapter · Next chapter

A two-temperature apparent energy can be defined as Esec = −R[lnKobs(T2) − lnKobs(T1)]/[1/T2 − 1/T1]. It is the slope across the supplied inverse-temperature interval, rather than the derivative at one arbitrarily chosen temperature.

If η changes appreciably across the interval, the Arrhenius plot can curve. A straight-line fit can then provide a summary over that range without identifying one constant intrinsic parameter. Its value depends on the interval, weighting and observations used in the fit.

For exact model values, the secant slope lies between local derivative values attained inside the interval when the usual continuity conditions hold. That mathematical fact does not say that the relevant temperature is the arithmetic midpoint. Nor does a visually straight short segment prove the absence of diffusion effects. Preserve the fitted range and observation model alongside the reported slope.

Contents · Previous chapter · Next chapter

The reciprocal temperature in the Arrhenius relationship is reciprocal kelvin. Substituting Celsius numbers changes the mathematical variable and does not merely move the graph’s origin. Convert the temperature before taking its reciprocal.

The energy unit must also match R. Using 80 as if it were joules per mole when it represents 80 kilojoules per mole reduces the exponent by a factor of one thousand. That can make a strongly temperature-dependent model appear nearly constant.

A useful checking record includes T, 1/T, k, D, φ, η and Kobs for each fictional state. Compute the dimensionless quantities from their defined inputs rather than copying a column from another geometry. Then reconstruct any fitted slope from the same Kobs basis. These checks locate a unit or state-definition error before it is interpreted as an unexpected catalytic mechanism.

Contents · Previous chapter · Next chapter

Our apparent coefficient was defined using a fixed, known surface concentration. If the actual observed rate is r = Cs(T)Kobs(T), its logarithmic slope also contains the temperature dependence of Cs. A changing gas density, adsorption state or external transfer condition can therefore alter a rate-based Arrhenius plot.

That extra contribution is not part of the internal slab result derived here. It should be modelled or removed through a justified normalisation rather than quietly attributed to Ek or ED. The same warning applies if the eligible catalyst amount changes between observations.

This is why careful reporting names the rate, concentration basis and specimen state together. An exact derivative of the wrong observable can be calculated perfectly and still answer a different question. Before discussing whether an apparent energy reveals a reaction barrier, identify which factors were allowed to vary in the recorded experiment and which were actually held fixed.

Contents · Previous chapter · Next chapter

The slab model fixes its surface concentration and excludes an external film resistance. If the bulk fluid cannot supply reactant to the surface sufficiently quickly, a second relationship is needed between bulk concentration, surface concentration and transfer rate.

The existing immobilised-enzyme lesson works through that different coupled boundary-supply problem and shows why bulk concentration need not be the concentration at a catalytic surface. Its numerical model is not the internal temperature-slope calculation developed here.

A real catalyst can experience both restrictions. A measured response to mixing or a geometry change may help investigate them, but the observation does not automatically isolate one resistance. Keep the proposed mechanism and the evidence supporting it separate. The value of the present exact model is that it gives a checkable internal-diffusion alternative, not that it makes every other explanation unnecessary.

Contents · Previous chapter · Next chapter

We assumed one temperature throughout the slab at each state. If reaction heats or cools the specimen unevenly, local k and D can vary spatially as well as between experiments. The concentration-only equation and its simple effectiveness expression no longer provide the whole account.

A measured external temperature may then differ from the temperatures experienced at catalytic sites. An apparent Arrhenius slope can reflect that observation difference, thermal feedback and transport in addition to the local chemical law.

Do not repair such a discrepancy by selecting a more dramatic activation energy from the same isothermal formula. State which added energy balance and measurement information the revised model would require. A model becomes more useful when its boundary is visible enough to identify the missing relationship, rather than being stretched to explain every unexpected rate or temperature trend.

Contents · Previous chapter · Next chapter

Suppose a measured apparent value is about 50 kJ/mol. Our large-φ example with Ek = 80 and ED = 20 approaches that value. Many other parameter choices or physical mechanisms can also produce a similar slope.

The number therefore does not uniquely identify either intrinsic energy or the presence of this particular diffusion regime. The inverse problem needs additional information: a transport model, geometry, concentration basis and evidence capable of distinguishing candidate explanations.

A more precise slope estimate can narrow uncertainty in the observed summary without resolving its physical nonuniqueness. Choose the next evidence for the missing relationship. In a paper exercise, constructing two parameter sets with the same limiting half-sum makes that point explicit. In an actual investigation, the corresponding evidence must concern the material, transport and reaction rather than only repeat the same fitted temperature range.

Contents · Previous chapter · Next chapter

Cornelia Breitkopf’s 2008 lectures at the Fritz Haber Institute discuss transport, effectiveness factors and the distinction between apparent and intrinsic kinetic parameters. The slides include slab effectiveness and limiting Arrhenius-regime reasoning.

Our numerical energies, reference coefficients, geometry comparisons and finite-slope questions are original teaching constructions. We explicitly allow an Arrhenius contribution from D rather than treating every diffusion coefficient as temperature independent. The exact intermediate expression follows from the model and chain rule shown here.

Read the institutional diffusion lecture for that methodological context. Source route checked on 24 October 2026. Its discussion of several transport regimes should not be reduced to a universal half-energy rule for every zeolite, reaction, particle shape or temperature range.

Contents · Previous chapter · Next chapter

Consider the claim that a larger specimen has a lower measured activation energy, so its catalytic sites must have a lower intrinsic reaction barrier. The conclusion selects a chemical explanation before transport and observation effects have been separated.

Our one- and three-micrometre examples reproduce that direction of change with identical local k and D laws. The measured coefficient’s temperature slope changes because the concentration profile contributes differently. This counterexample is enough to show that the original inference is not generally valid.

A repaired comparison states the observed rate basis, temperature interval, geometry and model alternatives. It can say that the result is compatible with transport influence without declaring that transport has been uniquely established. A strong correction preserves the measured fact while identifying which additional evidence would be needed before assigning the difference to local chemistry or to a particular resistance.

Contents · Previous chapter · Next chapter

Use a new fictional first-order isothermal slab model with Ek = 60 and ED = 10 kJ/mol. At the state of interest, φ = two. Calculate its exact local apparent activation energy and its large-φ limiting value.

Then halve L while holding the local k and D values at that temperature fixed. Predict how φ and η change, and calculate the new local apparent energy. State whether the intrinsic chemical parameter changed.

Finally decide whether a local energy inferred from the slab model is automatically the same as a slope fitted across an unspecified temperature interval. Explain what information is missing. Work from the half-sum, half-difference and hyperbolic multiplier rather than copying the preceding example’s numerical answer. The changed parameters test the structure of the argument, not only recall of the original result.

Contents · Previous chapter · Next chapter

The half-sum is 35 and the half-difference is 25 kJ/mol. At φ = two, the multiplier is 4/sinh(4), approximately 0.14657. The exact apparent energy is therefore approximately 38.66 kJ/mol, while the large-φ limit is 35.

Halving L changes φ from two to one. The multiplier becomes approximately 0.55144 and the apparent energy becomes approximately 48.79 kJ/mol. Effectiveness rises from tanh(2)/2 to tanh(1), making the observed coefficient closer to local kinetics under the model.

The intrinsic Ek remains 60. A different geometry changed the combined response, not the stipulated molecular parameter. A finite-range fit still needs its temperature interval and observation model. The transfer is successful when those conclusions remain separate rather than being compressed into an unexplained statement that the smaller specimen is simply more active.

Contents · Previous chapter · Next chapter

For a classroom investigation, begin with supplied k and D laws and compare predicted concentration profiles, effectiveness and slopes. Then remove one known parameter and ask which observations would distinguish the remaining candidate models.

For an actual scientific interpretation, relevant evidence can include appropriate geometry comparisons, independent transport information, surface-state definition and checks on thermal and external-transfer assumptions. Those are questions for a controlled study, not a universal procedure supplied by this article.

The evidence should address the source of nonuniqueness. Another copy of the same apparent slope may not distinguish Ek from ED or isolate a resistance. A smaller uncertainty bar on one summary is useful, but it should not be described as an identification of every parameter that can influence it. The model teaches a habit of connecting each additional observation with the relationship it actually supplies.

Contents · Previous chapter · Next chapter

A derivative describes the slope at one state. To see how actual pairs of temperatures enter the account, keep the same invented slab and use 340, 350 and 360 kelvin. Retain T0 = 350 K, k0 = one per second, D0 = one square micrometre per second, L = one micrometre, Ek = 80 kJ/mol and ED = 20 kJ/mol. Use R = 8.31446 joules per mole per kelvin with energy values converted to joules.

At 340 K, the stipulated laws give k about 0.44550 per second and D about 0.81698 square micrometres per second. Their ratio gives φ about 0.73845. Then η is about 0.85071 and Kobs is about 0.37899 per second. The surface-normalised coefficient is smaller than k, even though effectiveness is closer to one than at the reference temperature.

At 360 K, k is about 2.14606 per second and D about 1.21035 square micrometres per second. Now φ is about 1.33158, η about 0.65309 and Kobs about 1.40156 per second. The observed coefficient increased with temperature while the effectiveness decreased. Keep both statements: a shrinking fraction of a growing hypothetical rate can still produce a larger measured rate.

The table below collects the original model outputs. The digits are calculation aids, not a claim of experimental precision. If a reader obtains a negative effectiveness or a coefficient larger than k in this particular consuming slab, the first repair is the arithmetic or the model transcription.

Contents · Previous chapter · Next chapter

Use the two endpoint coefficients in Esec = −R ln[Kobs(360)/Kobs(340)] divided by (1/360 − 1/340). The logarithm is positive because the coefficient increased; the reciprocal-temperature difference is negative. The leading minus sign therefore produces a positive energy. The result is about 66.548 kJ/mol.

That value happens to be close to the local 350 K derivative of about 66.54 kJ/mol. This agreement belongs to the selected interval and model. It does not establish that the full curve is perfectly straight or that every temperature interval would return the same estimate. The local φ changes through the interval, so the exact derivative changes too.

An average gradient can conceal that change. To investigate curvature, compare more than one interval using consistent rate normalisation and justified uncertainty. Merely printing more decimal places on a two-point slope adds no evidence about linearity. Nor does drawing a straight line through exactly two points test whether the underlying relationship is straight.

For a changed calculation, leave the two temperatures fixed but use L = three micrometres in both states. Recompute φ and η before finding the endpoint slope. Do not reuse the small slab’s effectiveness values. This is a useful independent exercise because it tests whether the learner sees geometry inside the temperature-dependent coefficient rather than as an optional label attached after fitting.

Contents · Previous chapter · Next chapter

Suppose the only supplied information is an apparent energy close to 50 kJ/mol. One explanation in a strongly limited slab is Ek = 80 and ED = 20. Another is Ek = 70 and ED = 30. Their half-sums match, although their local chemical parameters differ. A nearly unrestricted specimen with Ek near 50 could also give a slope near that number under its own model.

These are alternative conditional explanations, not evidence that the specimens share a mechanism. A single slope lacks enough independent information to choose between them. The observed coefficient magnitude can add a constraint, but unknown reference coefficients, geometry or site normalisation may introduce further freedoms. Count the unknown relationships before claiming uniqueness.

An independently supported diffusion law, a justified size series or a separately established local kinetic measurement could make the inference more specific. Each addition needs its own assumptions. Changing crystal size can also change exposed surfaces or material history; a size comparison becomes a clean transport test only when the other relevant differences have been accounted for.

A useful reporting sentence is: this apparent slope is consistent with the specified reaction–diffusion model over the stated temperature interval, but it does not separately identify the intrinsic reaction and diffusion energy parameters. That sentence preserves the information gained without turning an underdetermined fit into a mechanism claim. Learning to state the remaining ambiguity is part of solving the problem, rather than an apology for an incomplete calculation.

Contents · Previous chapter · Next chapter

Does an apparent activation energy equal the intrinsic one?

Only under appropriate assumptions. In this model the equality is approached when internal concentration gradients become negligible, while other observation and concentration factors are held fixed.

Does internal diffusion always halve the energy?

No. The large-φ limit here is the half-sum of the reaction and diffusion activation parameters. The half-Ek statement requires an additional approximation about D, and intermediate states need not equal either limit.

Can particle geometry change an observed slope?

Yes, within this stipulated model geometry changes φ and the temperature dependence of effectiveness. The intrinsic local energy parameter can remain unchanged.

Does one slope prove a diffusion mechanism?

No. Different parameters and physical effects can produce similar observed slopes. A mechanism claim needs additional evidence suited to the competing explanations.

Contents · Previous chapter · Next chapter

For repair, practise distinguishing k, D, φ and η with their units and definitions. For stabilisation, derive the average from the concentration profile and check the two limiting energies. For extension, compare a local derivative with finite-range slopes or introduce a separately specified surface-concentration law.

Return to the Zeolite Catalysis root for material identity, active sites, reaction outcomes and transport boundaries. Use the immobilised-enzyme transport lesson for the separate external-supply and apparent-kinetics account, keeping its biochemical observable distinct.

For the denominator of a catalytic rate, continue with site-count bounds when probe access is incomplete. For a separate pressure-dependent surface effect, use a shared-site rate that can fall as one reactant increases. Their response equations concern different observations; neither substitutes for the temperature-dependent slab balance.

Continue through Learning That Transfers into Life and How X Works. A useful final explanation names the observed coefficient, geometry, concentration basis and model assumptions before interpreting its temperature slope. The learner should be able to identify which missing relationship prevents a measured apparent value from being assigned directly to a local chemical barrier.

Contents · Previous chapter · Return to Zeolite Catalysis