Hand-considered hypotheses, 401–500

Surfaces, and what happens on them. Adsorption, heterogeneous catalysis, thin films, buried interfaces and electrodes — a hundred statements about how these are measured and how they behave. Each was written on its own and given its own logical form, printed with the entry so that structural repetition would be visible if it were there. None has been tested. Numbers appear only as estimates, marked as such; no measurement is quoted and no source is implied.

Adsorption and the caloric record

Misattribution
H401. Heats of adsorption measured on a supported metal at low loading are credited to the metal, yet at that loading the calorimeter is mostly sensing the support's hydroxyl pairs, which bind the probe more strongly than the metal does. Run the same metal at the same dispersion on a support with a matched point of zero charge but a different hydroxyl density: if the integral heat moves by more than the metal's own contribution, the metal term has at last been separated. A null result leaves the support unproven rather than disproved, since two supports can differ in more than hydroxyl density.
Hidden term
H402. Isosteric heats extracted from two isotherms by the Clausius–Clapeyron route carry a term from the probe gas's non-ideality at the pressure of the second isotherm, a few kJ/mol at a few bar (estimate) — the same size as the effects under discussion. Computing the heat from fugacity rather than partial pressure should leave the values unchanged, while a change of padding gas at fixed partial pressure would expose any instrumental term masquerading as thermodynamics.
Bound
H403. Microporous adsorbents cannot raise a Henry constant without limit as pores narrow: once a pore admits fewer than about two adsorbate diameters the low-pressure uptake is capped by pore volume over liquid molar volume, and the constant must turn over. Values that keep climbing into that regime are being read off the isotherm's knee; three temperatures should make the Arrhenius plot visibly concave.
Threshold
H404. Commensurate monolayers on graphite-like surfaces lose registry abruptly at a coverage set by the adsorbate's own lattice mismatch, because the loss is domain-wall unbinding rather than smooth compression. Helium diffraction across that coverage at fixed temperature resolves the commensurate peak's disappearance to a narrow interval — a few per cent (estimate) — where a compression model predicts a slow drift.
Reversal
H405. Depending on the bond, adsorption can expand a surface or contract it, and which one happens is decided by whether the adsorbate–substrate bond is stiffer in shear than the bond it replaces — not by the adsorbate's size. Dose one surface with two adsorbates of equal radius but different bond stiffness while watching wafer curvature: the sign should flip at matched coverage. Should both adsorbates bend the wafer the same way, the shear argument fails and the expansion is merely steric.
Count
H406. From N resolved desorption peaks one can extract at most N−1 independent statements about site heterogeneity, since position and pre-exponential factor trade along a line that is nearly degenerate. A heating-rate series separates the possibilities: heterogeneity walks the peaks along the compensation line, while a coverage-dependent barrier on one site walks them across it.
Scale argument
H407. Below a cluster of roughly a hundred atoms (estimate) there is no facet for a binding energy to be a property of, and the measured size-dependence is really a dependence on edge-atom fraction. Mass-selected clusters of known atom count, measured in one apparatus, separate the two because count and edge fraction no longer vary together.
Identity
H408. Two desorption peaks long catalogued as two sites may be one site at two coverages: the adsorbate sits in the bridge position when dilute and is pushed on top when crowded, and the same binding energy then yields two peak temperatures. A coverage series at constant heating rate should show the high-temperature peak's population saturating exactly where the site switches.
Artefact mimics effect
H409. What looks like negative adsorption on a microporous solid at high pressure is buoyancy: the balance weighs the solid plus the gas it displaces, and above a few bar (estimate) the displaced gas outweighs the adsorbed gas. A non-adsorbing reference of matched skeletal volume, run in the same basket through the same pressure ramp, should reproduce the negative branch exactly. The reference's own adsorption has to be measured, not assumed, or the correction overshoots.
Inversion of standard practice
H410. Baking an adsorbent at the highest temperature it tolerates is the usual activation, and for hydroxyl-bearing surfaces it is self-defeating: neighbouring hydroxyls condense into bridges that no longer bind the probe. Two activations ending at the same vacuum, one at 150 °C and one at 300 °C (estimate), compared at one measurement temperature, decide whether the gentler one holds more capacity.
Unit trap
H411. Two rates for the same powder can differ by whatever area was used to divide them, and when that area comes from a probe whose cross-section is wrong for a curved surface, the ratio acquires a factor that reads as promotion. Normalising both loadings by chemisorbed site count instead of by nitrogen-adsorption area should collapse the rates; failure to collapse means the denominator, not the chemistry, was varying.
Selection effect
H412. Published adsorption-energy distributions are truncated by the pump-down that precedes every measurement, since species bound more weakly than roughly 40 kJ/mol (estimate) leave before the first spectrum is taken. Dosing at 10⁻⁶ mbar (estimate) and cooling before evacuation should restore the missing low-energy tail, and the tail's position should track how long the chamber was pumped. The restored tail's weight then bounds how much of the literature's "strong-site" population is evacuation.

Surface structure and preparation

Misattribution
H413. Reconstructions that appear on annealing are credited to the surface's own thermodynamics, yet on many crystals it is a bulk impurity arriving at the surface that stabilises the new periodicity, which is then a contamination phase wearing a structural name. Two slices from one boule, one gettered beforehand by a high-temperature oxygen treatment, will differ in symmetry after identical anneals if the impurity is doing it.
Hidden term
H414. Step energies inferred from island shapes contain an elastic term from film–substrate misfit, and on a strained film that term can exceed the chemical step energy, so the measured anisotropy belongs to the misfit. Islands grown at equal temperature and supersaturation on a substrate matched to within 0.1% and on one mismatched by 2% (estimates) separate them, since only the second carries the elastic term.
Identity
H415. Two labels, (1×2) and missing-row, have been attached to what may be one structure averaged differently: a fit assuming a single domain reports the row absent, a fit over two domains reports it at half occupancy. A spot profile that resolves the two orientations separately should return one answer for both samples.
Threshold
H416. Step bunching sets in only above a critical miscut, below which the elastic repulsion between steps beats the attraction that would pair them, so the critical angle measures that repulsion directly. Vicinal cuts spanning 0.2° to 3° (estimate), annealed identically, should show the transition in the shape of the terrace-width distribution rather than in its mean. Broadening with anneal time would mean kinetics, not an elastic threshold.
Reversal
H417. Charge compensation is said to stabilise polar films, but below a few unit cells (estimate) the compensation cannot be completed and the unreconstructed form becomes the stable one, making the crossover a thickness rather than a material constant. Growing cell by cell and measuring symmetry after each should place the reconstruction's onset at a thickness that moves with the substrate's screening.
Count
H418. Curvature measurements of a film's stress yield at most two independent elastic constants however many curvatures are taken, because the third combination enters only through the Poisson ratio, which curvature cannot see. Films deliberately made with different Poisson ratios, measured on one cantilever, should leave the two recoverable constants fixed while the third absorbs all the variation.
Scale argument
H419. A probe whose coherence length exceeds the domain size reports an average, so a surface that diffracts sharply can still be a mosaic of small ordered patches. Changing the coherence length — electron energy in diffraction, beam size in scattering — should broaden the spot for the mosaic and leave it untouched for the genuinely ordered surface.
Artefact mimics effect
H420. Half-order spots appearing after a gentle anneal can be multiple scattering through a domain boundary rather than a new periodicity, and they fade when the azimuth is turned a few degrees. A dynamical calculation including boundary scattering reproduces the spot's energy dependence; a true reconstruction's spots survive every azimuth.
Inversion of standard practice
H421. Cleaning by sputtering and annealing is judged finished when the diffraction pattern sharpens, and it can sharpen precisely because residual carbon has segregated into a well-ordered overlayer. Compare a standard-cycle sample with one annealed in oxygen at the same temperature by Auger at one take-off angle: the sharper pattern should belong to the dirtier surface. One take-off angle suffices; a survey scan averages the two surfaces into false agreement.
Boundary-condition ownership
H422. Chlorine-free is not a property a surface owns; the bulk reservoir decides it, and at the temperatures where the surface looks clean the bulk is still delivering chlorine at a rate comparable to the experiment's own timescale. Two samples pre-treated to set two bulk activities, then quenched to one measurement temperature, should differ in surface coverage at equal nominal cleanliness.
Null space
H423. Registry of an incommensurate adlayer is not observable: sliding the whole overlayer by a fraction of a lattice constant changes no measured intensity, so reported registries are gauge choices. Driving a commensurate–incommensurate transition with a small temperature change breaks the degeneracy, and the registry inferred above the transition should then predict the one found below it. Disagreement below the transition would mean the overlayer was never in registry at all.
Two-clock race
H424. Hold an alloy at temperature and it will either oxidise or segregate its solute, and which one wins is a race between oxygen arrival and solute diffusion that the same temperature can lose either way depending on pressure alone. Ramping pressure at fixed temperature, instead of the usual temperature ramp at fixed pressure, should locate the crossover as a pressure threshold that shifts with the alloy's diffusivity.

Kinetics, sites and selectivity

Misattribution
H425. Sixty kilojoules per mole (estimate) read off an Arrhenius plot is commonly taken as the barrier of the rate-limiting step, but a number that size can equally be a heat of adsorption multiplied by a coverage that falls with temperature, and the two are indistinguishable in a straight line. Modulating the reactant's partial pressure at fixed temperature and measuring the phase lag of the rate separates them, because a coverage term lags and an intrinsic barrier does not.
Hidden term
H426. High conversion in a plug-flow reactor drags in an axial dispersion term whose size follows the bed's length-to-diameter ratio rather than the catalyst, so two laboratories with different beds derive different kinetics from one powder. Holding contact time fixed while changing bed length at constant flow should move the apparent rate constant by the dispersion term alone.
Bound
H427. Selectivity in a consecutive reaction A→B→C cannot exceed a ceiling fixed by the ratio of the two rate constants and the conversion reached, whatever the catalyst is said to do at zero conversion. Plotting B's yield against conversion for a family of catalysts should produce a common envelope, and a point above it means either a measurement error or a different network. The envelope's shape, not its height, is the invariant worth comparing between laboratories.
Threshold
H428. Spectators begin to flip the sign of the reaction order once they occupy roughly half a monolayer (estimate), because the sites that would re-adsorb the product are the ones they take. The flip appears as a cusp in rate against coverage at fixed temperature, and its position should not move when the reactant's pressure is changed — which is what tells you the spectators are doing it.
Reversal
H429. Adding potassium to an iron catalyst promotes at low temperature and poisons at high temperature: the alkali donates charge that helps the first step and blocks the sites the later steps need. A temperature series at fixed potassium coverage should show the enhancement crossing zero at a temperature that moves with loading, one experiment rather than a survey. Cooling back through the same range should retrace the crossing if the potassium is not redistributing.
Count
H430. Data at a single temperature and pressure can identify a rate and one order, and every remaining parameter of a proposed mechanism is a prior wearing the costume of a fitted value. A global fit over a temperature and pressure grid, reported as a confidence ellipsoid instead of a best fit, would show which parameters the data actually hold.
Artefact mimics effect
H431. A period that tracks the controller's settings is not a surface instability: thermal lag between sample and thermocouple can close a feedback loop with the furnace and put oscillations into the measured rate that have no chemical content. Welding the thermocouple to the sample and, in a second run, suspending it in the gas phase at the same setpoint should preserve or destroy the period accordingly. Removing the thermocouple altogether is the crudest version of the test and the most convincing.
Scale argument
H432. Structure sensitivity is usually plotted against particle diameter, but the length that matters is that diameter over the distance an adsorbate travels in one turnover; below about 3 nm (estimate) the two are comparable and shape begins to outweigh facet. Shape-controlled particles at one size, rather than a size series, are the experiment that separates them.
Identity
H433. Two mechanisms with separate names, Langmuir–Hinshelwood and Mars–van Krevelen, collapse into one rate law whenever the lattice oxygen exchanges faster than a turnover, and the names then describe bookkeeping rather than chemistry. Switching 16O2 for 18O2 at steady state decides it: a response time equal to the turnover time makes the lattice a spectator, a much longer one makes it the reactant.
Inversion of standard practice
H434. Bed dilution is meant to suppress hot spots, and in an exothermic reversible reaction it works against the equilibrium by cooling the inlet, where the equilibrium is most favourable — so the diluted bed converts less at the same inlet temperature and catalyst mass. Concentrating the catalyst where the equilibrium permits it should beat dilution in a two-zone bed of identical total mass.
Unit trap
H435. Turnover frequencies per exposed atom assume one probe molecule per surface atom, and on many metals hydrogen titrates closer to two (estimate), so the number is high by that factor and a bimetallic "promotion" can be nothing but the change in stoichiometry. Titrating each sample with two adsorbates and counting both, instead of importing a handbook ratio, should make the promotion disappear or survive. Whatever the second adsorbate's stoichiometry turns out to be, it must be measured on the same surface.
Selection effect
H436. Calling a reaction structure-insensitive is a statement about the sample set, not about the chemistry: it means the samples on hand happened to have similar site densities. One reaction on samples whose site densities differ by an order of magnitude, made by a single route, should either reveal the sensitivity or show the label was an artefact of narrow sampling.
Boundary-condition ownership
H437. The hydrogen order a reactor reports can belong to the gas-side boundary layer rather than to the surface, because at high pressure a transport limit is first order in hydrogen whatever the catalyst does. Raising the linear velocity at fixed partial pressures should move the measured order toward the intrinsic one, and the direction of that move says which regime the old number came from. Were the order immovable with velocity, transport would be dead as an explanation.
Derived quantity
H438. Dispersion, the divisor inside every turnover frequency, is inferred from a chemisorption uptake through an assumed stoichiometry, so the uncertainty in the rate is that assumption's uncertainty and not a random error bar. Chemisorption, microscopy counting and a catalytic titration applied to one sample give three dispersions, and their spread is the honest bound on the turnover frequency.
Null space
H439. The compensation effect — activation energy trading linearly against the logarithm of the pre-exponential factor — is a direction in which an Arrhenius fit cannot see, and any dataset with a narrow temperature window will display it. Doubling the window (estimate) should flatten the correlation if it is a fitting artefact and leave its slope untouched if it is real.
Two-clock race
H440. Light-off is a race between the rate at which the bed heats and the rate at which the reaction consumes the heat, and the coupling makes the outcome non-monotonic: an over-active catalyst lights earlier and then overheats into a regime where it converts less. Ramping at several rates with inlet composition fixed tells you which clock dominates by the direction the light-off temperature moves.
Hysteresis asymmetry
H441. Ascending and descending light-off curves are not mirror images, and the gap is information: the descending branch sees a surface reduced by the reaction, the ascending branch an oxidised one. Holding between the two branches and stepping the composition assigns the hysteresis to one surface state or the other. Slowing the ramp closes the gap if heat transfer, rather than the surface, owns it.
Aliasing
H442. Sampling the rate once a minute while the furnace cycles every thirty seconds (estimate) folds a thermal oscillation into an apparent chemical one at a period that belongs to the controller. Sampling at ten times the controller frequency, or logging the setpoint beside the rate, should move the apparent oscillation onto the controller's period or erase it.

Thin films and growth

Misattribution
H443. Compressive stress in a sputtered film is put down to atomic peening, yet at the pressures where peening is invoked the substrate is also being heated, and the thermal term does not vanish with pressure. Depositing at one pressure onto substrates held at two temperatures separates the two, since peening scales with pressure and the thermal term does not.
Hidden term
H444. Thickness read from a quartz crystal in humid air includes the water adsorbed on the film's own surface, and in a porous or hydrated film that mass can be a sizeable fraction of the first monolayer (estimate). Repeating the measurement in dry carrier gas and across a humidity series should give a dry intercept and a slope equal to the film's water uptake.
Bound
H445. How much strain an epitaxial film stores before relaxing is bounded by the areal density of misfit dislocations the interface can nucleate, not by the film's elastic limit, so the critical thickness follows the substrate's terrace width. Films of one misfit grown on substrates of different miscut should relax at different thicknesses if steps are the nucleation sites.
Threshold
H446. Conductance arrives only at coalescence, and the threshold thickness for a given growth mode is set by the islands' aspect ratio rather than by the metal's conductivity. In-situ conductance against thickness, taken at two deposition rates on one substrate, should place the threshold at the same island density if aspect ratio rules it. Rate-scaling of the threshold would implicate kinetics instead of geometry.
Reversal
H447. Wetting can reverse with thickness: below a critical value the interface energy dominates and the film is stable, above it the film's own surface energy makes dewetting favourable and annealing breaks the film into holes. A thickness series annealed together and imaged at one magnification splits into intact films below that thickness and holes above it.
Count
H448. One reflectivity curve at one wavelength carries a phase and an amplitude, so it can fix at most two independent thickness parameters however many layers the stack has. Measuring at three wavelengths, where the layers' indices differ, raises the count — and the count can be checked by fitting synthetic data whose thicknesses are known.
Scale argument
H449. Once terraces are a few tens of nanometres wide (estimate), the step-edge barrier rather than the surface diffusion coefficient controls growth, and a temperature series taken in that regime measures the barrier's temperature dependence dressed as diffusion. Growing on vicinal substrates whose terrace widths span a decade, at fixed temperature, should change the growth mode where the terrace width meets the diffusion length. Independence from growth rate is the test: the crossover width should not care how fast the film is made.
Artefact mimics effect
H450. Oscillations in a reflection signal are read as layer-by-layer growth, yet a surface that is merely smoothing during growth produces the same damped wave, because the signal tracks step density and not layer completion. Following the oscillations with an independent step-density measurement, or with diffraction, separates growth from annealing.
Inversion of standard practice
H451. Slow growth is the standard prescription for smooth films, and where roughness comes from residual-gas incorporation the prescription inverts: faster growth leaves less time per layer for the impurity to adsorb and can come out smoother. Two films grown an order of magnitude apart in rate at one base pressure, compared by one roughness metric, decide it. Roughness that keeps falling at still higher rates means something other than impurity uptake is at work.
Unit trap
H452. Monolayer, in the growth literature, means at least three things — a completed atomic layer, a thickness equal to one lattice constant, and an arbitrary calibration of the source's flux — and the three are used interchangeably. A sample grown to "one monolayer" and measured by a technique that counts atoms rather than thickness reveals which definition the calibration encodes, and the discrepancy is the sticking coefficient.
Selection effect
H453. Films that survive transfer through air are the ones that get reported, and air passivates some orientations and oxidises others, so a reported orientation dependence may be conditioned on survival. Measuring the same films in vacuum straight after growth should either preserve the dependence or show it belonged to the passivation. The comparison holds only at one probe and one spot; two spots measure two surfaces.

Buried interfaces and contacts

Misattribution
H454. Adhesion numbers from a pull test often describe the capillary force of the water layer between film and probe, and between 30% and 70% relative humidity (estimate) the number can move by a factor of two. Repeating the test in dry nitrogen and across a humidity series on one sample should give a straight line whose intercept is the adhesion.
Hidden term
H455. A Schottky barrier extracted from current–voltage data includes a dipole created at the interface during deposition, which the ideal-junction model does not contain. Growing one metal on one semiconductor at two deposition rates, while measuring both the barrier and the interface's composition, tests whether the barrier follows the dipole or the work function.
Identity
H456. Contact resistance from a transfer-length structure and interface resistance from a four-point measurement are one quantity under two boundary conditions, and the difference between them is the current-crowding length rather than a second resistance. Structures of several contact widths, measured both ways, should collapse onto one curve against width over crowding length.
Threshold
H457. Ohmic behaviour survives only below a critical oxide thickness, under which tunnelling carries the current and the barrier hardly matters; the threshold is a few nanometres (estimate) and depends on the oxide's electron affinity. Contacts on oxides grown to a thickness series, measured at one voltage, trace an exponential onset and a roll-over at the threshold. Killing the tunnelling picture would take a threshold that tracked the metal's work function instead of the oxide's affinity.
Reversal
H458. Annealing a contact lowers its resistance by intermixing and then raises it by voiding, and because the two processes have different activation energies the optimum depends on anneal time as well as temperature. An isothermal series with resistance logged continuously should show a minimum that moves with temperature, which is what distinguishes two mechanisms from one optimum.
Count
H459. Modelled as a slab, a grain boundary has at most two independent channel densities however many dislocations it contains, because a measurement sees only the channels parallel and perpendicular to it. Boundaries of several misorientations, measured in both directions, should fall on a two-parameter family if that is all there is. A third parameter would appear as systematic scatter about that family, which the two-direction measurement would catch.
Boundary-condition ownership
H460. Fermi-level pinning at a metal–semiconductor interface belongs to the interface's own defects rather than to the metal's work function, which is why the barrier barely notices which metal is used. Introducing a controlled defect density from the semiconductor side before metallisation should move the barrier in proportion to that density and not to the metal.
Null space
H461. Specular reflectivity cannot separate a conformal roughness from a graded density, because both damp the curve in the same way, and a fit that reports one has chosen a point in a null space. Off-specular scattering, sensitive to lateral correlation but blind to the gradient, breaks the degeneracy and should yield a roughness the specular fit could not. An off-specular roughness smaller than the specular fit's is the degeneracy made visible.
Artefact mimics effect
H462. Interdiffusion seen in a sputter depth profile is partly the sputtering's own mixing, and its apparent width then scales with the primary ion's energy rather than with annealing time. Profiles taken at two ion energies on one annealed sample separate the two: the width tracks energy if the artefact is the cause, and holds if the diffusion is real.

Electrochemical surfaces

Misattribution
H463. Current passed by a metal oxide at a given overpotential is taken as its intrinsic activity, while part of that current is the oxide being reduced and the two cannot be told apart in a steady-state polarisation curve. A potential-step series with the current integrated over time, compared against the film's known reduction charge, separates the faradaic current from the film's own reduction.
Hidden term
H464. Charging current in a porous electrode's voltammogram carries the pseudo-capacitance of the electrode's own surface redox couples, so the capacitance extracted from a scan-rate plot is a sum of two terms with different physics. Measuring across several scan rates and several potential windows should move the capacitive term only when a couple lies inside the window.
Bound
H465. How much current a porous electrode can deliver is bounded by ion transport in the pore rather than by the catalyst's intrinsic rate, and the bound is met at a current density that depends on pore length and not on chemistry. Two thicknesses of one catalyst at one potential should give the thinner electrode a higher current per mass, which inverts the usual instinct that more loading is better.
Threshold
H466. A pH exists above which hydrogen evolution on an oxide switches from water reduction to proton reduction, and the switch shows up as a change in Tafel slope at a pH that moves with the buffer's capacity. Buffers of different capacity at one pH should move the apparent switch, which identifies it as a local-pH effect rather than a change of surface chemistry. If capacity were the only lever, the switch would sit still when the buffer's diffusion coefficient changes.
Reversal
H467. Tafel slopes can fall with rising coverage while the mechanism stays put, because the coverage's own potential dependence adds a term of opposite sign to the kinetic one. Reading the coverage with an independent optical probe at matched current should show the slope tracking coverage rather than mechanism.
Count
H468. A single Tafel slope admits only a few mechanisms: at room temperature, 30, 40 and 60 mV/dec (estimates) each map onto specific rate-determining steps, and a slope between them is an average over a distribution rather than a new step. Current-interruption measurements at several currents, which recover the true surface overpotential, should collapse the distribution onto one of the discrete values. A slope that holds one intermediate value across a tenfold change in current is a mechanism, not an average.
Scale argument
H469. Inside pores narrower than the double layer — a few nanometres at moderate ionic strength (estimate) — a local electrolyte potential is not defined, so surface area and current density stop being separable quantities. Electrodes spanning that pore size at fixed chemistry give a capacitance per area that departs from the geometric expectation exactly where the double layers begin to overlap.
Artefact mimics effect
H470. Oxygen-evolution activity on a supposedly stable catalyst can be the metal dissolving and redepositing, with the corrosion current counted as catalysis. Measuring dissolved metal in the electrolyte alongside the current, over a series of catalysts, should correlate activity with dissolution if the current is corrosion.
Inversion of standard practice
H471. Holding a potential and watching the current is the standard stability test and the one least able to find degradation, because a catalyst that passivates under the hold looks stable and then fails on the first excursion. Cycling between the operating window's limits at the same charge passed should degrade what the hold did not, and the two tests belong side by side. Passing the hold and failing the cycle is an instruction about which test to report.
Unit trap
H472. Geometric area, catalyst mass and electrochemically active area give three current densities for one material that can differ by two orders of magnitude (estimate), and each has been written as "mA/cm²" in print. Reporting one dataset under all three normalisations, with the roughness factor stated, shows whether a claimed improvement survives the change of denominator.
Selection effect
H473. Lifetimes in the literature are censored: a run is stopped when the current has fallen by a fixed percentage, a criterion chosen after the fact, so what is reported is a truncated distribution rather than a measured one. Running a subset to a much lower current, or to a fixed charge, should lengthen the tail and show whether decay is a power law or an exponential. Only extrapolating the tail back to the stopping criterion makes the two literatures comparable.
Boundary-condition ownership
H474. Local pH at an electrode belongs to the buffer's transport rather than to the bulk's value, and one catalyst can appear to change mechanism between two buffers of different capacity. A rotating-disk series at fixed bulk pH and varying rotation rate should shift the apparent mechanism only if local pH is in control.
Two-clock race
H475. Catalyst or sacrificial anode is decided by which is faster, and the winner can change with potential inside one experiment. Measuring dissolution per turnover rather than current at several potentials locates the stability limit where that ratio crosses unity. Its potential dependence, not its value at one potential, is the number worth quoting.
Hysteresis asymmetry
H476. Forward and reverse scans diverge when the surface reconstructs, and the size of the divergence measures the reconstruction's kinetics rather than the redox couple's. Triangular scans at one rate but different vertex dwell times should alter the asymmetry only if a slow surface process is behind it.
Aliasing
H477. Mains pickup can be folded into an impedance feature at a frequency that moves with the excitation's resolution, because the potentiostat's sampling clock and its excitation are not necessarily synchronised. Changing the excitation frequency by a few per cent, or the sampling rate, moves the artefact and leaves a real feature where it was.

The model surface and the working one

Misattribution
H478. The pressure gap is often a temperature gap: the vacuum experiment runs where the adsorbate is stable and the reactor where it is not, so the two disagree about chemistry rather than about pressure. One apparatus varying pressure and temperature independently at fixed adsorbate chemical potential makes the rate a function of that potential and of neither variable alone.
Hidden term
H479. Image-charge screening is part of every binding energy measured on a metal, belongs to the substrate's electron density rather than to the bond, and shifts each adsorbate by an amount that depends on its height. One adsorbate on two substrates of different screening, at independently measured heights, reproduces the image term's shift rather than chemistry's.
Bound
H480. Ninety per cent of an XPS signal can come from a depth of a few nanometres (estimate) at the kinetic energies in common use, so a "surface" enrichment measured this way is an average over several atomic layers. Sweeping the take-off angle through its widest range should change the apparent enrichment, and that change is the bound made visible.
Threshold
H481. No clean metal surface survives above a threshold oxygen chemical potential, and the threshold is a line in pressure and temperature rather than a pressure, crossing the usual working range near 10⁻⁶ mbar (estimate). Approaching one chemical potential along two paths, hot and low-pressure versus cool and high-pressure, tests whether the line is thermodynamic or kinetic. A line that shifts with heating rate is kinetic, and "clean surface" then becomes a statement about the ramp.
Reversal
H482. Potassium promotes at low coverage and poisons at high, and the coverage where it turns is set by the alkali's own ordering transition rather than by a smooth electronic optimum. Sweeping coverage with the ordering measured by diffraction in the same chamber should put the reversal at that transition.
Count
H483. Infrared bands of adsorbed CO number at most as many as there are distinct geometries, and the assignments to "sites" are routinely over-read: a shift of a few wavenumbers (estimate) can be coverage or ensemble rather than a new site. Isotopic dilution, which removes dipole coupling without touching the sites, should leave the band count unchanged if the assignments hold. Bands that merge under dilution were one band split by coupling, and the site count drops with them.
Scale argument
H484. Ambient-pressure XPS looks through a gas layer whose mean free path is comparable to the analyser's sampling depth, so the surface it reports is not the one the reactant meets. Raising the pressure at fixed sample temperature should change the apparent composition by attenuation alone, and that change can be computed without any model of the surface.
Artefact mimics effect
H485. Motion of an adsorbate under the tip is sometimes the tip dragging it, and the displacement then looks like diffusion with a drift. Imaging the same area at two bias polarities, or with a different tip, should reverse or abolish the drift if the field is doing it.
Identity
H486. The phase that operando X-ray absorption calls active and the phase in the reactor are the same only if the beam has not made it; flux can reduce or oxidise a catalyst, and the species seen is then the beam's product. Two fluxes with extrapolation to zero should give a phase fraction that reactor conditions can be compared against.
Inversion of standard practice
H487. Catalysts are reduced before testing because the reduced phase is assumed to be the active one, and for a reaction run in oxidising conditions that preparation guarantees a transient: the first hours are the oxidation of the catalyst's own surface. A pre-oxidised and a pre-reduced sample of one material, run through the same time-on-stream series, should converge only if the transient is the preparation's. If convergence appears, its time is the preparation's half-life and is worth measuring directly.
Unit trap
H488. Leak rates quoted as mbar·L/s, Pa·m³/s and sccm convert by a factor near ten (estimate) that gets dropped in practice, so a "clean" system's impurity flux can be wrong by that factor. Calibrating one leak against a gas of known composition and reading the pressure rise in a single unit system removes the ambiguity.
Selection effect
H489. Argon sputtering decides which surfaces get studied, and it removes the lighter component preferentially, so the compositions in the literature are the ones sputtering happens to leave. A crystal cleaved or grown in situ, compared with the sputtered surface of the same material, should differ in composition by the preparation's own bias.
Null space
H490. Coordination number and disorder are degenerate in an extended-absorption fine-structure fit, because both damp the same oscillations, and a "nanoparticle" coordination number is one point in that null space. Two temperatures, where disorder changes and coordination does not, should break the degeneracy.

Deactivation and the long run

Misattribution
H491. Activity lost over a hundred hours is chalked up to sintering, and the same loss appears when the support's surface hydroxyls have been consumed, because the support was co-catalysing the reaction all along. Regenerating with water below the sintering threshold should restore activity if the loss was chemical and leave it flat if it was morphological. A support whose hydroxyls cannot be restored should show the loss as irreversible, which is the cleaner half of the test.
Hidden term
H492. Sintering rates carry a term from the mismatch between the support's thermal expansion and the holder's, which strains the particles each cycle and helps them detach. The same powder on two holders of different expansion, through one temperature programme, should sinter at different rates only if that term is real.
Bound
H493. Dispersion has a ceiling at a given loading, set by the chemical potential of a metal atom on the support against the bulk, and below roughly a nanometre (estimate) atoms dissolve into the support instead of forming particles. Loading series measured by chemisorption and by microscopy should diverge exactly where dissolution starts.
Threshold
H494. Ripening reverses on a small support: when the support's area is low enough that particles are never far from each other, the support's own surface diffusion can redisperse metal faster than ripening coarsens it. Supports of several areas at one loading, aged identically, put the sign change at an area estimable from the particle density. No sign change at any support area would put the effect under the detection limit rather than refute it.
Reversal
H495. Steam usually ages a catalyst and sometimes redisperses it, because steam volatilises the metal as a hydroxide that decomposes elsewhere on the support. A steam series at two temperatures, with the metal's mass balance measured, places redispersion below the temperature where the volatile species is stable.
Count
H496. One decay curve supports at most two modes, each contributing a time constant and the curve's information living in its curvature, so claims of three modes need three independent observables. Activity, coke content and dispersion measured together on the same samples raise the count, and synthetic curves with known modes test whether the fit can recover them.
Artefact mimics effect
H497. Decay in a small reactor can be the reactor's own dead volume releasing a trace impurity slowly into the feed, a poisoning front the vessel manufactured. Two reactors of different dead volume, one feed, one catalyst, should change the decay's shape only if the vessel is the source.
Boundary-condition ownership
H498. Measured lifetimes can belong to the reactor wall, which catalyses a side reaction that eats the feed or the product, leaving the catalyst to take the blame for the wall's chemistry. A reactor lined with an inert material against one lined with the catalyst's own material should give lifetimes that differ by the wall's contribution.
Two-clock race
H499. Recovery from sulfur under hydrogen is a race between the poison arriving at the surface and being removed from it, and the steady coverage sits where neither process alone would put it. Stepping the feed's sulfur concentration at constant hydrogen pressure drives the coverage to a new steady value on a timescale that is the geometric mean of the two rates.
Inversion of standard practice
H500. The denominator decides the ranking: normalise one dataset by mass, by area, by site count and by turnover, and four catalysts can reorder three times. Publishing all four rankings for one benchmark reaction in one laboratory would turn a habit into a test, and the reversals would be the result.

Status

H401–H500. One hundred hypotheses, each written separately and each tagged with the form of its argument rather than its subject. The tags are there so that a reader can check whether the variety is structural or only lexical, and so that a repeated form is visible as a repeated form.

What these share is only the shape of a usable test: a comparison in which the suspected cause and the suspected confounder move independently. None has been tested, and several would probably fail — the ordinary fate of a hypothesis, and not a defect of the list.

If one of these is worth an afternoon, it is H402 — that part of an isosteric heat is the probe gas behaving as a gas, and that fugacity instead of pressure would settle it. Two isotherms and a virial coefficient are enough.

Incoprea Research · hand-considered hypotheses 401–500 · 3 November 2026 · ~/Desktop/Nov3/