Skip to content

Boltzmann–Matano Analysis

Infer concentration-dependent interdiffusivity from a diffusion-couple profile using a mass-balanced reference plane.

Version
v2 · 2026-10-03 · History
Domain-specific #
13021
Domain group
Natural Sciences
Origin domain
Chemistry & Materials Science
Subdomain
Diffusion Analysis → Chemistry & Materials Science
Aliases
Matano–Boltzmann analysis, Boltzmann–Matano method

Core Idea

Boltzmann–Matano analysis estimates how an interdiffusion coefficient varies with concentration from a measured one-dimensional diffusion-couple profile at a known time. Two regions begin with different compositions; after diffusion, a spatial concentration curve replaces the initial step. Under suitable Fickian, binary and material-balance assumptions, the Boltzmann similarity coordinate x/√t reduces the diffusion description, while the Matano plane supplies the spatial reference at which transferred amounts balance. The method combines the profile's local slope with an integrated profile area to infer D(c) at selected compositions.[1][2]

The value is inverse: rather than assume one constant D and predict a profile, it uses an observed profile to estimate a concentration-dependent coefficient. The inference is conditional. Composition-dependent molar volume, multicomponent cross-diffusion, non-Fickian transport, uncertain boundary conditions and noisy profile derivatives can require modified analyses or make the classical result unreliable.[1][2]

Structural Signature

Sig role-phrases:

  • Diffusion couple — Adjacent regions initially have distinct concentrations or compositions.
  • One-dimensional measured profile — The evolved concentration is measured against position at a known elapsed time.
  • Similarity premise — For the appropriate step-couple Fickian problem, position scales with √t.
  • Matano plane — A mass-balance plane defines a meaningful zero for the spatial coordinate; it need not be the visible original join or geometric midpoint.[1]
  • Slope and integrated area — A derivative near a chosen concentration and a cumulative concentration-position integral jointly determine an estimate of D there.
  • Composition-dependent output — Repeating the evaluation across suitable concentrations produces a curve D(c), not automatically separate intrinsic or tracer diffusivities.[1]

What It Is Not

  • Not a constant-diffusivity forward fit. An error-function curve with a single D predicts or fits one profile; Boltzmann–Matano inversion aims to recover variation in D across composition.
  • Not the arbitrary midpoint of a sample. The Matano reference is chosen from a transferred-amount balance.[1]
  • Not an unconditional single-profile solution for every mixture. Binary/pseudo-binary and molar-volume assumptions matter, and a fully multicomponent problem may require multiple coefficients.[1]
  • Not an exact reading of noisy measurements. Differentiation and profile fitting can change inferred coefficients, especially at concentration extremes.[1]
  • Not a measurement of intrinsic or tracer diffusion by itself. It estimates an interdiffusion or mutual coefficient in the stated formulation.

Scope of Application

The method is used in diffusion-couple studies when a composition gradient is experimentally available. In an Fe–Ga alloy study, researchers measured a profile with electron-probe microanalysis and evaluated how interdiffusivity varied with composition; their paper also examines the instability caused by profile-fitting choices. In a distinct physical-chemistry setting, Kellö and colleagues used wedge interferometry at a polyvinyl acetate–solvent interface and calculated mutual diffusion coefficients for different compositions.[1][2]

These settings share the inverse relation but not every material assumption. The polymer paper explicitly states a one-dimensional nonstationary Fickian model and a no-volume-change condition for its formula. The alloy paper states that the simple atomic-fraction expression assumes constant or approximately ideal molar-volume behavior. It would be unsafe to transfer the formula into a strongly swelling or multicomponent system without checking those premises.[1][2]

Clarity

Place two different compositions together and allow exchange for time t. A line scan yields composition c(x). The Matano plane is positioned so the excess material transported toward one side balances the corresponding deficit on the other. At a chosen composition c*, the classical binary expression uses both the local inverse slope dx/dc and the accumulated signed area between an end composition and c*, scaled by 1/(2t). The signs and concentration conventions must be kept consistent so the physically inferred diffusivity is positive where the assumptions hold.[1]

The reduction is informative because an asymmetric profile can encode a coefficient that changes with composition. Yet two profile fits that look equally plausible can have different derivatives, especially in sparse endpoints. Therefore the output curve is an inference from data plus model, not a direct instrument reading.[1]

Manages Complexity

An evolving diffusion equation has a field that varies with both position and time. Similarity scaling turns a suitable step-couple problem into a profile relation, and the Matano plane fixes the integration reference. This compresses a continuum of local composition-dependent behavior into an estimable curve from an experimental snapshot.[1][2]

The compression also exposes a tradeoff: it gains a rich D(c) estimate from limited data while making slope estimation and mass-balance conventions load-bearing. A concentration profile without a reliable time, monotone composition interval, suitable mass balance or stable derivative cannot be promoted into a precise D(c) simply by naming the method.

Abstract Reasoning

Suppose two concentrations mix over a distance interval. If one assumed a constant diffusivity, an error-function form would have a fixed scale. If the observed profile has composition-dependent spreading, the Boltzmann–Matano construction evaluates an integral and local slope at several values of c*. Comparing those values is how the profile is turned into a varying coefficient rather than a single fitted number.[1]

Now suppose the measured tail is noisy. The profile area may change only slightly, while the local derivative changes substantially after a different smoothing fit. The Fe–Ga study demonstrates why apparently good profile fits can yield different coefficient curves, particularly near end members. The diagnostic is stability of the inferred coefficient under defensible fits, not visual smoothness of one chosen curve.[1]

Knowledge Transfer

The method transfers from alloy interdiffusion to polymer–solvent mutual diffusion because both can provide a one-dimensional concentration profile and a time scale. What transfers is the inversion architecture: similarity, mass-balance reference, slope and area. Material-specific concentration definitions, swelling or molar-volume behavior, phase changes and multicomponent coupling do not transfer automatically.[1][2]

Examples

Fe–Ga alloy interdiffusion

Kavakbasi and collaborators measured Fe–Ga diffusion-couple composition profiles by electron-probe microanalysis. They located a Matano plane and compared methods for fitting the profile before deriving composition-dependent interdiffusivity. Their focus on strong composition dependence makes this an example of the inverse method and of its derivative sensitivity.[1]

Mapped back: Couple → Fe–Ga end members; profile → EPMA composition line scan; time → annealing interval; reference → mass-balanced Matano plane; inversion → D(c) across suitable compositions; limit → fitted slope sensitivity.

Polyvinyl acetate–solvent interface

Kellö and colleagues measured concentration profiles for polyvinyl acetate in contact with solvents using wedge interferometry. They calculated mutual diffusion coefficients at different compositions using the Matano–Boltzmann method under a stated one-dimensional, no-volume-change formulation. It is not an alloy, but it has the same inference roles.[2]

Mapped back: Couple → polymer and solvent regions; profile → interferometric concentration variation across the swollen interface; time → nonstationary diffusion interval; reference → material-balance coordinate; inversion → mutual D(c); limit → volume assumption must be checked.

Structural Tensions

  • Local coefficient detail versus profile-fit stability. One measured diffusion profile can yield composition-dependent D(c), including sharp changes, but the inversion differentiates a fitted profile. A smoother fit suppresses experimental scatter and can stabilize the derivative, yet may erase a real steep concentration dependence; a more flexible fit preserves local detail but can convert noise into spurious extrema and unstable endpoint values. Kavakbasi and colleagues explicitly propose an improved fit for strongly composition-dependent Fe–Ga because standard inversion is sensitive to measured-profile uncertainty. Diagnostic: Do the inferred D(c) features survive plausible fit choices and independent error estimates, especially at the ends?[1]

Binary or defensible pseudo-binary, one-dimensional Fickian transport and material-volume premises are conditions for this classical formula, not the other side of that fitting tradeoff. A polymer–solvent profile must be assessed under its own experiment's assumptions rather than inheriting an alloy's volume behavior.[1][2]

Structural–Framed Character

Boltzmann–Matano analysis is a domain-specific inverse method. The abstraction is neither one diffusion experiment nor a generic appeal to scaling: it binds an initial composition contrast, an evolved spatial profile, a time scale, a mass-balance plane and a slope–area inversion. The method can be instantiated in different materials only when its governing assumptions remain defensible.

The inverse-profile relation is structural, but the classical method is also a human-devised analytical practice: choosing concentration coordinates, the mass-balanced Matano plane and acceptable binary/Fickian assumptions determines whether the computation is valid. It is not an evaluative classification; fit quality and uncertainty are judgments about an application, not its identity. Its institutional origin lies in diffusion research and its subsequent laboratory use across material systems, not in a rule that any institution can declare true by convention: the cited alloy and polymer studies each have to test the material and measurement assumptions for themselves.[1][2] “Inverse analysis” travels across domains, yet merely importing that phrase does not reproduce this diffusion-specific slope–area construction. Its character: a methodologically framed structural inversion whose mathematical skeleton is portable in principle but whose named realization belongs to diffusion analysis.

Structural Core vs. Domain Accent

Skeletal relation. Infer a varying transport coefficient from an evolved field profile by similarity reduction and a conserved-balance reference.

Domain-bound condition. Fickian interdiffusion, concentrations, diffusivity units, diffusion-couple boundary conditions and Matano mass balance make this the named method.

Prime bar. Inverse-profile reasoning recurs elsewhere, but this exact concentration-dependent diffusion construction has not been established as a cross-domain prime.

Parent check. Diffusion is the process being measured; fitting and integration are method components, not verified strict parents of the composite method. The live comparison found no checked inverse-diffusion genus. Whether a general profile-inversion prime should exist is a future identity question, not a relationship claimed here.

Diffusion is a physical process; curve fitting and integration are components of the method. None is a strict parent of Boltzmann–Matano analysis merely from topical similarity. With no broader inverse-diffusion-analysis genus identified, this method stands as a root; a closer genus could warrant a different relation.

Neighborhood in Abstraction Space

Boltzmann–Matano Analysis sits in a sparse region of the domain-specific corpus (81st percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2551 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-08

Not to Be Confused With

Constant-D error-function solution assumes one coefficient and solves forward. Sauer–Freise/de Broeder approaches use related profile integrals but alter reference/normalization conventions; the Fe–Ga paper compares them rather than treating them as exact synonyms. Intrinsic diffusivity of an individual component and tracer diffusivity require additional relations or experiments. The classical Matano output is a concentration-dependent interdiffusion or mutual coefficient in its stated model.[1][2]

References

[1] B. Tas Kavakbasi, I. S. Golovin, A. Paul and S. V. Divinski, “On the analysis of composition profiles in binary diffusion couples: systems with a strong compositional dependence of the interdiffusion coefficient” (2017), original Fe–Ga study, Introduction equations (1)–(2), experimental methods and results checked. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t

[2] V. Kellö et al., “Study of interfacial diffusion in the system solid poly(vinyl acetate)–solvent by the Matano–Boltzmann method”, Chemical Papers 33 (1979), pp.347–356, original study, abstract, equations (4)–(5) and results checked. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j