Two probe readings can look like a complete count of a zeolite’s internal and external acid sites. They may instead count two different accessible portions of a specified site population. A probe that cannot enter the internal pores may also fail to reach some sites classified as external. Subtracting its reading from a total can then assign too many sites to the interior.
The short answer is to write a response equation for each probe before subtracting signals. We will build an original fictional two-compartment model, show exactly what two measurements leave unidentified, derive useful bounds and test an additional measurement. Every quantity is a teaching value. This is a lesson in interpreting accessibility-dependent counts, not a laboratory protocol or a claim about a named zeolite.
eduKateSG · Zeolite Catalysis
Find your next learning step
Choose the question closest to your task, or follow the original worked example in order. All teaching chapters remain open.
ROUTE 1 · CHAPTERS 1–6
Define the response
Separate chemical eligibility, access and measured amounts.
ROUTE 3 · CHAPTERS 13–18
Test the assumptions
Examine uncertainty, incomplete access and catalytic denominators.
ROUTE 4 · CHAPTERS 19–26
Practise and compare
Solve a changed record and audit mass, time and consistency.
ROUTE 5 · CHAPTERS 27–28
Questions and onward routes
Check the interpretation and continue through verified guides.
Chapter contents
Define the response · Chapters 1–6
1. Start with the site population being counted
2. Turn a signal into a response equation
3. Why internal exclusion is not complete external access
4. Work a fictional pair of readings
Constrain the partition · Chapters 7–12
7. Add a justified lower bound on accessibility
8. Compare three interpretations side by side
9. Distinguish calibration uncertainty from inaccessible sites
10. Keep chemical class consistent across probes
11. Add a genuinely independent third response
12. Why repeating the total does not add a new partition equation
Test the assumptions · Chapters 13–18
13. Test sensitivity when response weightings nearly match
14. Allow uncertainty in the supplied readings
15. What changes if the small probe is also incomplete
16. Add an independent inventory cap carefully
17. Do not turn a site count into turnover frequency automatically
18. Use the published methodological example for its proper role
Practise and compare · Chapters 19–26
19. Try a changed-record problem independently
20. Check the changed-record solution
21. Build a measurement and interpretation ledger
22. Compare fractions and counts on the same mass basis
23. Recognise when a difference contains more than one effect
24. Keep a failed consistency check visible
Questions and onward routes · Chapters 27–28
CHAPTER 1 OF 28 · Define the response
1. Start with the site population being counted
Let I denote the amount of a defined class of internal sites and E the amount of the same class classified as external. For the first model, both are nonnegative and measured in micromoles per gram of dry specimen. Classification and mass basis remain fixed throughout the comparison. The class might be a carefully specified acid-site response; it is not automatically every aluminium atom or every site that operates in a catalytic reaction.
A location label needs an operational definition. Sites on an outer crystal surface, at pore mouths and in wider voids may not all behave alike. Our two compartments simplify that complexity so we can examine the inverse problem clearly. A real experiment must justify whether those compartments are useful for its own question.
Keep three questions separate: which chemical class is eligible, where eligible sites are classified, and which eligible sites the probe actually reports. A measurement can answer the last question very well while leaving the first two conditional. Defining the population is therefore a scientific step, not a heading to fill after a calculation has been completed.
An instrument signal is not itself an amount of sites. Suppose calibration has already converted each supplied signal into an amount-equivalent reading on the same dry-mass basis. Assume a one-equivalent response for each counted site in the defined class, with backgrounds and cross-signals handled as stipulated. Those are model inputs, not properties to assume for every probe.
Call the small-probe reading S. Initially stipulate complete access to both compartments and equal response per eligible site, so S = I + E. Call the large-probe reading B. Stipulate exclusion from the internal compartment but allow incomplete access to the external compartment: B = aE, where a is positive and at most one.
The fraction a combines the particular incomplete accessibility represented in this model. It is not a universal molecular-size correction or a reaction rate constant. If calibration, stoichiometry or chemical response also differs, they need additional justified terms. The simple equations let us see precisely which conclusion depends on which assumption before more realistic complications are introduced.
CHAPTER 3 OF 28 · Define the response
3. Why internal exclusion is not complete external access
Imagine that a large visitor cannot enter a building. Counting the people that visitor meets outside does not count every person classified as outside: some may be behind a screen or on another side. The analogy helps distinguish exclusion from completeness, but it supplies no quantitative probe calibration.
In the site model, exclusion justifies the absence of I from B = aE. It does not establish a = one. An external site can remain unreported under a particular probe, exposure or signal definition. Likewise, a small probe’s molecular dimensions alone do not prove that its reading equals I + E under every condition.
The most common shortcut joins two separate claims into one: the large probe does not enter internal pores, therefore its reading equals all external sites. The second statement requires its own evidence. When that evidence is missing, retain a rather than silently replacing it with one. The algebra will then show whether a useful range is still possible.
Suppose S = 100 and B = 30 micromoles per gram. The complete-small-probe model gives I + E = 100, while the second response gives aE = 30. There are three unknowns, I, E and a, but only two supplied equations.
If a = one, E = 30 and I = 70. If a = one-half, E = 60 and I = 40. Both reproduce exactly the same two readings. At a = 0.3, E = 100 and I = zero, which also reproduces them. Nothing in the two readings alone selects the first explanation.
These examples are deliberately simple enough to check by substitution. For each proposed allocation, add I and E to recover S, then multiply E by a to recover B. Verification against both equations is stronger than choosing an allocation because it looks plausible. A third plausible pair that fails one response equation is not another admissible explanation of this record.
Because a is at most one, B = aE cannot exceed E. Thus E is at least 30. Because I is nonnegative and I + E = 100, E cannot exceed 100. The combined feasible interval is therefore 30 ≤ E ≤ 100, with I = 100 − E.
The corresponding internal interval is zero ≤ I ≤ 70. These two ranges are linked. I and E cannot independently take their upper endpoints at the same time because their sum must remain 100. A valid uncertainty account preserves that relationship rather than reporting two unrelated boxes.
The same constraints give a = 30/E, so 0.3 ≤ a ≤ one. Notice that the measurements have themselves excluded accessibility fractions below 0.3 under this model. A claimed fraction of 0.2 would require E = 150, leaving a negative I. That conflict identifies an inconsistent assumption or record; it does not justify accepting a negative site population.
CHAPTER 6 OF 28 · Define the response
6. Use bounds without presenting them as estimates
A bound states which values have not been excluded by the stipulated information. It is not a probability distribution, a confidence interval or a statement that the midpoint is most likely. Nothing in our two exact readings makes E = 65 preferable to E = 40 or E = 90 when a is allowed to vary consistently.
The model does establish that at least thirty of the hundred amount-equivalents belong to the external compartment. It also establishes that at most seventy belong to the interior. Those are useful conclusions, even though a unique partition remains unidentified.
A careful summary is: under complete small-probe access and internally excluded but incompletely accessible large-probe response, the internal amount lies between zero and seventy micromoles per gram. Keep the assumptions next to the bound. Removing them changes the claim. In particular, calling seventy the measured internal count would discard every other feasible partition without supplying the additional evidence needed to do so.
CHAPTER 7 OF 28 · Constrain the partition
7. Add a justified lower bound on accessibility
Suppose independent evidence supports a ≥ 0.5 for this specimen and measurement definition. Retain the upper limit a ≤ one. Then E = 30/a lies between 30 and 60, so I lies between 40 and 70. The new information has narrowed the partition without pretending that a is exactly known.
The direction of the inequality matters. Increasing the accessible fraction means fewer external sites are needed to generate the same reading. Dividing by a lower bound therefore produces an upper bound on E. A quick substitution at both endpoints is an effective way to catch a reversed interval.
The independent evidence must support the fraction for the same relevant conditions. A fraction inferred from a different specimen, preparation or chemical class is not automatically transferable. If its applicability is uncertain, report the result conditionally rather than treating the narrowed interval as directly measured. The calculation is straightforward; deciding which accessibility constraint is justified is the substantive evidence question.
CHAPTER 8 OF 28 · Constrain the partition
8. Compare three interpretations side by side
The table follows the same exact pair S = 100 and B = 30. All amounts use micromoles per gram of dry specimen. The difference between rows is the added assumption about a, not a change in the readings.
| Information about large-probe access | External amount E | Internal amount I | Interpretation |
|---|---|---|---|
| Only zero < a ≤ one | 30 to 100 | zero to 70 | A family of partitions remains feasible |
| Independently supported 0.5 ≤ a ≤ one | 30 to 60 | 40 to 70 | An additional constraint narrows the family |
| Independently established a = one | exactly 30 | exactly 70 | Unique partition within this response model |
Read each row with I + E = 100 still enforced. The first row does not permit I = 70 and E = 100 together. Nor does the last row show that complete access was established by subtraction. It shows what follows if completeness has been established separately.
CHAPTER 9 OF 28 · Constrain the partition
9. Distinguish calibration uncertainty from inaccessible sites
Suppose B is a calibrated signal with an uncertain conversion factor. A weaker signal could then reflect fewer accessed sites, a smaller response per accessed site, or both. The single factor a can only be called an accessibility fraction if the response conversion has already been justified.
This distinction is especially relevant when comparing different probe molecules or signal bands. A raw peak area from one probe should not be subtracted directly from a raw peak area from another merely because both are described as acid-site signals. Their dimensions, calibration and chemical assignments must support the subtraction.
For the teaching problem, the calibrated amount-equivalent readings are supplied, so the algebra begins after that step. For an evidence review, ask how each reading was converted, which background was removed and which uncertainty belongs to that conversion. A useful measurement ledger records both the instrument observable and the amount assigned to it. Keeping both columns prevents a model assumption from disappearing inside an apparently direct site count.
CHAPTER 10 OF 28 · Constrain the partition
10. Keep chemical class consistent across probes
A site may be described through proton donation, electron-pair acceptance, a structural environment or an adsorbed probe’s spectral response. Those descriptions are related through specific chemistry and measurement assumptions; they are not interchangeable counting labels.
Our equations require that S and B refer to the same eligible class on a compatible amount basis. If one reading combines two classes while another reports only one, S − B need not isolate a location. It can mix a chemical-class difference with an accessibility difference.
Before partitioning, write a sentence defining eligibility for both responses. If that sentence cannot be made consistent, expand the response model rather than forcing the old subtraction. Additional chemical populations may require additional unknown amounts and independent measurements. This is why a well-labelled spectral observation can be more useful than a seemingly decisive total that conceals which species contributed. The learner’s aim is to make the inference traceable, not to turn every visible signal into the same kind of site.
CHAPTER 11 OF 28 · Constrain the partition
11. Add a genuinely independent third response
Return to S = 100 and B = 30. Suppose another calibrated response is supplied as C = 0.5I + E, with C = 80. The coefficients are established independently for the defined compartments; they are fictional teaching inputs, not universal probe properties.
Subtracting C from S gives 0.5I = 20, so I = 40 and E = 60. The original large-probe reading then gives a = 30/60 = 0.5. Substitution checks all three: 40 + 60 = 100, half of 40 plus 60 = 80, and half of 60 = 30.
The extra response helps because its weighting of I relative to E differs from the first response. It is not simply another total of the same form. The exercise shows what kind of information can identify a partition: an independently known response relationship that distinguishes the compartments. Merely naming a third instrument does not guarantee that it supplies that relationship.
CHAPTER 12 OF 28 · Constrain the partition
12. Why repeating the total does not add a new partition equation
Suppose the third reading instead equals twice the first, C = 2I + 2E = 200. It is consistent with every allocation already allowed by S = 100. Dividing it by two returns exactly the old equation, so it supplies no new partition information in the exact model.
Repeated measurements can still help assess repeatability or reduce some kinds of random uncertainty. Their value should not be confused with adding an independent response direction. Many precise copies of I + E cannot separate I from E without another justified constraint.
This is an accessible example of rank in an inverse problem. A reader need not begin with matrix terminology: ask whether the new equation treats the two unknown amounts in a different proportion. If it does not, the same trade-off remains invisible. The distinction transfers to mixture analysis, surveys and sensor calibration, but the response coefficients in each application must be established for that application rather than borrowed from this fictional chemistry example.
CHAPTER 13 OF 28 · Test the assumptions
13. Test sensitivity when response weightings nearly match
For a general third response C = cI + E, subtracting it from S gives I = (S − C)/(1 − c), provided c is known and differs from one. If c is close to one, a small difference between S and C is divided by a small number.
For example, c = 0.99 and S − C = 0.4 give I = 40. If the measured difference were 0.5 instead, the inferred I would be 50. A change of 0.1 amount-equivalents has changed the inferred internal amount by ten under these assumptions.
The equations can therefore identify a unique exact solution while the solution remains sensitive to measurement or calibration uncertainty. Uniqueness and stability are different questions. A useful evidence plan asks both whether the response directions differ and whether their difference is large enough to distinguish the quantities at the available precision. That does not select a real probe automatically; it tells us what to evaluate before calling a proposed extra measurement decisive.
CHAPTER 14 OF 28 · Test the assumptions
14. Allow uncertainty in the supplied readings
Suppose the amount-equivalent readings are intervals, S between 98 and 102 and B between 29 and 31. Retain complete small-probe access and a between 0.5 and one. A conservative marginal range is E between 29 and 62, obtained from B/a at the allowed extremes.
The corresponding conservative internal range is 36 to 73: the smallest total minus the largest external amount, and the largest total minus the smallest external amount. These endpoints are feasible here because the example independently allows those combinations. If the calibration errors are linked, a joint uncertainty model could exclude some endpoint combinations.
Do not label these deterministic input ranges as statistical confidence levels. They are conditional bounds constructed from supplied intervals. Also retain I + E = S within its allowed range, rather than allowing any pair from the two marginal intervals. A record of which uncertainties are shared is often more useful than an extra decimal place on the final subtraction.
CHAPTER 15 OF 28 · Test the assumptions
15. What changes if the small probe is also incomplete
Now relax the assumption that the small probe reaches every internal site. Write S = bI + E, where b is an unknown internal-access fraction between zero and one, while keeping B = aE. This is a new model, with an additional unknown response fraction.
The original upper bound I ≤ S − B no longer follows. With S = 100 and B = 30, take E = 40 and a = 0.75. Then bI = 60. Choosing I = 120 and b = 0.5 reproduces both readings, despite an internal amount greater than seventy.
If b can approach zero while its product with I remains sixty, the readings alone supply no finite upper bound on I in this simplified model. A separately known total eligible amount or a positive lower bound on b could restore a useful bound. The earlier interval was correct under its stated completeness assumption; it becomes incorrect when quoted after that assumption has been removed.
CHAPTER 16 OF 28 · Test the assumptions
16. Add an independent inventory cap carefully
Suppose a separately justified chemical inventory establishes I + E ≤ 150 for the same eligible class and mass basis. In the incomplete-small-probe model, this cap rules out arbitrarily large internal amounts. But it does not automatically establish equality or determine how much of that inventory is active in a catalytic reaction.
Inventory constraints can come from different kinds of evidence, and their relationship to the defined site class must be explained. A bulk elemental count, for example, is not automatically a one-for-one accessible acid-site count under all material states. Treating a bulk number as a cap may require a different argument from treating it as the exact total.
The practical lesson is to name the direction and strength of the added statement. An upper limit, a lower limit and an equality each constrain the feasible set differently. Replacing a cautious inventory bound with an equality can create apparent precision that the evidence did not provide. Good bookkeeping preserves the original wording of the evidence as well as its numerical value.
CHAPTER 17 OF 28 · Test the assumptions
17. Do not turn a site count into turnover frequency automatically
A catalytic rate divided by a site amount has inverse-time units. Those units alone do not establish that the denominator counts the sites carrying out the observed reaction under its conditions. Probe eligibility and catalytic participation are separate claims.
Suppose a supplied rate is sixty micromoles per gram per minute. Dividing by E = 30 gives two per minute, while dividing by E = 60 gives one per minute. Both denominator choices can fit the same original S and B when accessibility is unknown. The apparent per-site activity changes because the inferred denominator changes.
Even after E is identified, calling this a turnover frequency needs evidence connecting that compartment and chemical class to the reaction. Multiple populations, changing coverage or different working states can complicate the connection. In a learner’s calculation, report the ratio with its explicit denominator first. That makes clear what was computed and which additional interpretation remains conditional.
CHAPTER 18 OF 28 · Test the assumptions
18. Use the published methodological example for its proper role
The University of Manchester research record for Bräuer and colleagues’ 2017 study describes a DRIFTS-based quantitative approach to hydroxyl acid species in zeolites, using pyridine and collidine to distinguish accessibility-related populations alongside other characterisation. It is a useful primary research record showing why probe choice, location and independent chemical information belong together.
Our I, E, a and b equations and all numerical records are original teaching constructions. They do not reproduce that paper’s specimens, calibration or results. In particular, the incomplete-access fractions used here are not reported estimates for its probes. The record supports the methodological context, while the conditional bounds follow from the equations stated in this article.
Read the institutional record of the zeolite acid-site characterisation study with that distinction in mind. The source was checked on 24 October 2026. Before applying a published method, consult its complete experimental contract and establish whether the specimen and response assumptions match the new question.
CHAPTER 19 OF 28 · Practise and compare
19. Try a changed-record problem independently
A new fictional specimen gives S = 120 and B = 24 micromoles per gram. The small response is complete for the defined class; the large response excludes I. Independent information places a between 0.4 and 0.8. Find the feasible external and internal intervals and the external fraction of the total.
Then suppose a third response has known form C = 0.25I + E and value 66. Find I, E and a. Check that the resulting accessibility falls within the independently supplied range. Explain why a third reading with form C = 3I + 3E would not serve the same identification purpose.
Pause before reading the next chapter. Write the response equations, identify which quantities are unknown and preserve the direction of division by a. This problem changes the readings, accessibility range and third-response weighting, so a remembered answer from the first specimen will not suffice. The transferable method is to reconstruct the constraints and test the candidate partition against every supplied equation.
Since E = 24/a, the accessibility limits give E between thirty and sixty. With I + E = 120, I lies between sixty and ninety. The external fraction E/S therefore lies between one-quarter and one-half, with the fractions linked to the same partitions.
For the third response, subtract C from S: 0.75I = 54, so I = 72. Then E = 48 and a = 24/48 = 0.5. The checks are 72 + 48 = 120, one-quarter of 72 plus 48 = 66, and half of 48 = 24. All amounts and the response fraction satisfy their allowed domains.
The alternative C = 3I + 3E would equal 360 for every allowed partition. It would check consistency of a total response but would not distinguish the compartments. If your solution produced the right total but a outside 0.4 to 0.8, the next step would be to identify the inconsistent constraint or arithmetic. Do not discard an independently supplied range just to retain a preferred partition.
CHAPTER 21 OF 28 · Practise and compare
21. Build a measurement and interpretation ledger
A useful ledger begins with the specimen identity, preparation state, dry-mass convention and eligible chemical class. For each supplied response, record the raw observable, calibration relationship, amount-equivalent reading, access assumptions and uncertainty. Then write the equations that connect those readings to the unknown compartment amounts.
Mark which coefficients were supplied independently and which were inferred from the same data. A coefficient fitted from S and B cannot then be counted as an independent third constraint on S and B. That circularity is easy to miss when results are spread across several plots or instruments.
Finally, separate three outcomes: a uniquely identified partition within the model, a bounded family of partitions, or an unresolved quantity without a finite bound. Each is a legitimate result when supported. The ledger should make it possible for another reader to reproduce the substitutions and identify the assumption responsible for each limit. A clean account often changes the next question from how to obtain another number to which missing response relationship would actually reduce the ambiguity.
CHAPTER 22 OF 28 · Practise and compare
22. Compare fractions and counts on the same mass basis
Suppose the same specimen is reported per two grams rather than per gram. The original S = 100 and B = 30 per gram become total readings of 200 and sixty micromoles for that supplied mass. The external amount for the two-gram portion lies between sixty and two hundred micromoles under the unrestricted-accessibility model. Dividing back by two returns the original per-gram bounds.
The external fraction E/(I + E) does not change under this common mass scaling. It remains between 0.3 and one. An accessibility fraction is also unchanged because both B and E scale together. These cancellations work only when numerator and denominator refer to the same represented specimen and mass convention.
A wet-mass reading and a dry-mass reading cannot be reconciled by assuming the labels are cosmetic. A separately established conversion would be needed. Nor can a total amount from one specimen be divided by a mass-normalised response from another to form a meaningful fraction. Units are an efficient first check, but specimen identity and preparation state must match too.
CHAPTER 23 OF 28 · Practise and compare
23. Recognise when a difference contains more than one effect
Suppose two specimens give the same S = 100, while their large-probe readings differ: thirty for specimen A and forty for specimen B. Under complete external access, one might infer that B has ten more external amount-equivalents. With unknown access fractions, that inference is not unique.
For example, both specimens can have E = sixty and I = forty. The readings then follow from a = 0.5 for A and a = two-thirds for B. Alternatively, they can share a = one-half while having E = sixty and eighty respectively. Both accounts fit the supplied totals and large-probe readings, but they describe different changes.
The comparison therefore establishes a change in the reported accessible response, while leaving its allocation between total external amount and accessibility conditional. An independently supported common a would make the partition comparison stronger. Do not infer that common fraction merely because the same probe name appears in both records. The model asks whether the relevant response relationship is actually transferable between the two specimens.
CHAPTER 24 OF 28 · Practise and compare
24. Keep a failed consistency check visible
Return to the exact first model with S = 100 and B = thirty. Suppose an independent claim sets a = 0.2. Substitution requires E = 150 and I = −50, which violates the nonnegative domain. The three statements cannot all hold within the model.
A useful report keeps the contradiction visible and names its possible location: the reading, the calibration, the access claim or the response structure could require review. It does not identify which one failed from the algebra alone. Replacing the negative answer with zero would silently change at least one supplied constraint and conceal the discrepancy.
This is also a reason to preserve exact and approximate inputs distinctly. If the readings have uncertainty ranges, test whether any admissible combination restores a nonnegative partition. If none does, the inconsistency remains under those ranges. If some do, report the feasible subset rather than selecting a convenient value without explanation. A domain check can be more informative than an impressive-looking fitted count.
A sequence of probe readings during exposure can reveal that the recorded response is still changing. It does not automatically reveal the eventual total or separate an inaccessible population from a slowly approached one. Those interpretations require a time-response model and evidence that its extrapolation is justified.
For instance, a supplied reading of thirty at one time could be a complete response to thirty eligible sites or a partial response to sixty. A second reading that rises to forty rules out a constant thirty response over that interval, but it still does not establish the final amount. Different approach laws can agree over a short observed period and differ later.
The proper next step depends on the question. If the aim is an equilibrium amount, establish the relevant endpoint criterion and response relationship. If the aim is a rate of approach, define the time-dependent observable and its parameters. Neither can be obtained simply by renaming the last available signal a total. This temporal distinction complements the spatial accessibility bounds without supplying an unearned numerical correction to them.
CHAPTER 26 OF 28 · Practise and compare
26. Repair, stabilise and extend the reasoning
If subtraction gives a negative amount, return to the response equations and compatible units before changing the chemistry story. Check whether both signals were calibrated on the same basis and whether the probe populations were meant to match. Negative values can reveal an inconsistent model or record rather than a surprising physical site population.
If the algebra is correct but the interpretation is unstable, construct two different partitions that reproduce the observations. This makes the missing information visible. Then add a justified accessibility bound and see exactly how the family narrows. The exercise stabilises the distinction between a measured response and an inferred amount.
For extension, ask whether a proposed extra measurement changes the relative response to I and E, whether its coefficients are independently known and whether the distinction remains robust to uncertainty. Avoid adding complexity merely to obtain a unique numerical answer. The aim is a better-supported inference whose limits can be explained in ordinary language as well as in symbols.
CHAPTER 27 OF 28 · Questions and onward routes
27. Questions that help keep the count honest
Can the large-probe reading be called the external count?
Only when the response model and evidence support complete external reporting for the defined class and conditions. Internal exclusion alone does not establish that completeness. In our first two-reading example, B is a lower bound on E because a is at most one, while equality is a separate conditional case.
Does a narrow range prove that the site model is correct?
No. A narrow conditional range can follow from strong assumptions even when those assumptions have not been validated for a specimen. Check the response definitions and independent evidence before interpreting numerical precision. The bounds describe the stipulated model’s feasible set, not a guarantee that every real complication has been included.
Are internal sites necessarily inaccessible during catalysis?
No. Probe access concerns a particular molecule and measurement state. Catalytic reactants, products and working conditions can differ. Our location partition does not determine transport rates, catalytic participation or lifetime. Those need their own relationships and evidence rather than being inferred from the word internal.
What is the smallest useful next measurement?
There is no universal instrument answer. In the exact two-compartment problem, a justified access fraction or an independent response weighting can identify the partition. With uncertainty, the new relationship must also be sufficiently informative. Start with the unresolved variable and ask which evidence directly constrains it.
CHAPTER 28 OF 28 · Questions and onward routes
28. Continue through the relevant reading route
For the chemical foundation, return to Zeolite Catalysis and its explanations of acid character, access, diffusion and characterisation. Keep a structural inventory, a probe response and a working catalytic population as distinct quantities until evidence connects them.
For an adjacent rate problem, read immobilised enzymes, diffusion and apparent reaction rates. It considers how transport changes an observed rate; the present article concerns which site amounts an accessibility-dependent response can identify. The shared habit is to define the observation before interpreting its denominator or slope.
For the distinction between an amount and its approach in time, read equilibrium loading and the approach rate in a linear-driving-force model. For mass-basis and chemical-inventory reasoning, compare ideal charge capacity across hydrated and exchanged forms. Neither relationship by itself supplies the acid-site access fractions used here.
For the rate interpretation after defining a site population, read shared-site coverage and a declining two-reactant rate. For a different transport contribution to a measured kinetic summary, use diffusion and apparent activation energy. A probe partition, a coverage product and a temperature slope each need their own observation model.
Continue through Learning That Transfers into Life and How X Works. A successful explanation states what each probe reports, which access assumptions are supplied, what range or partition follows, and which independent evidence would change that conclusion.