Extracting Off-Diagonal Interdiffusion Terms from Intersecting High-Temperature Alloy Couples
Extract off-diagonal interdiffusion terms by annealing two intersecting couples, measuring EPMA gradients at the crossing point, and solving the linear Fick-Onsager matrix.

Flux
Multicomponent mass transport in nickel-base superalloys follows generalized chemical potential gradients rather than simple uncoupled concentration steps. In ternary and quaternary systems, the movement of a single metallic species responds directly to the concentration gradients of every other dissolved element in the lattice. Extracting off-diagonal interdiffusion coefficients requires isolating these thermodynamic interactions at specified isothermal temperature points.

Multicomponent Transport Equations in High-Temperature Alloys
Fick-Onsager formulations extend classical binary transport logic to systems containing three or more elemental species. For an N-component alloy system, the diffusive flux of component i relative to a volume-fixed frame of reference is expressed as a linear summation of all independent concentration gradients.
J_i = – sum_{j=1}^{N-1} D_{ij}^N (dC_j / dx)
Where J_i represents the interdiffusion flux of element i, D_{ii}^N defines the main diagonal interdiffusion coefficient, and D_{ij}^N (where i is not equal to j) represents the off-diagonal interdiffusion coefficient. The off-diagonal term quantifies the cross-effect, where a concentration gradient of element j drives a net flux of element i. In concentrated nickel-base superalloys containing chromium, aluminum, cobalt, and rhenium, off-diagonal terms often reach thirty percent of the magnitude of main diagonal terms.
Neglecting these cross-diffusion coefficients leads to substantial errors when modeling microstructural evolution, phase stability, or oxidation kinetics during high-temperature exposure.
A measured interdiffusion flux of -2.4 x 10^-11 g/(cm²·s) occurs at the 1150 °C interface when chromium gradients reach 12 weight percent per millimeter.

Dayananda Method and Integration Boundaries
Evaluating interdiffusion profiles without locating a physical Kirkendall marker plane relies on integrating concentration profiles across the entire diffusion zone. The interdiffusion flux of component i at any specific position x along the diffusion axis is calculated directly from spatial concentration curves through the integration approach developed by Dayananda.
J_i(x) = (1 / (2 t))
Where t represents the isothermal annealing duration in seconds, C_i^- and C_i^+ represent the terminal compositions of component i at the far left and far right ends of the diffusion couple, and Y_i represents the normalized concentration variable defined as (C_i – C_i^-) / (C_i^+ – C_i^-). Spatial integration converts electron microprobe spatial linescans into explicit flux profiles without requiring prior knowledge of individual partial molar volumes.
Executing accurate numerical integration along measured microprobe profiles demands a strict sequence of experimental preparation steps before furnace insertion:
- Synthesize high-purity terminal alloy end-members using vacuum induction melting to prevent trace oxygen contamination.
- Homogenize cast alloy ingots at 1250 °C for 48 hours inside quartz ampoules evacuated to 10^-4 Pa.
- Slice homogenized ingots into rectangular blocks measuring 10 mm by 10 mm by 5 mm using wire electrical discharge machining.
- Grind contact surfaces sequentially using silicon carbide papers up to 4000 grit, followed by colloidal silica diamond polishing to achieve sub-micron flatness.
- Clamp opposing alloy blocks tightly inside a high-purity molybdenum fixture under mechanical spring pressure.
- Perform preliminary hot pressing under vacuum at 1000 °C for 30 minutes to establish intimate metallic bonding across the interface.
- Encapsulate bonded couple assemblies within quartz tubes backfilled with high-purity argon gas.
- Insert quartz capsules into a calibrated horizontal tube furnace maintained within half a degree Celsius of the target isothermal exposure temperature.
Concentration gradients evaluated near steep interface steps yield reliable fluxes only when local microstructural porosity remains absent.

Intersection
Mapping diffusion paths on a Gibbs ternary composition triangle reveals how distinct alloy couples cross at specific stoichiometry points. Determining all four independent interdiffusion coefficients (D_11^3, D_12^3, D_21^3, and D_22^3) for a ternary alloy at a single composition requires establishing two independent diffusion couples whose experimental composition paths cross each other on the ternary phase diagram.

Geometric Criteria for Dual Composition Paths
Selecting terminal alloy compositions for couple pairs dictates whether their concentration trajectories cross in ternary space. Couple A and Couple B must share a common compositional intersection point C while exhibiting non-parallel concentration gradient vectors at that exact location. At the crossing composition C , both diffusion couples possess identical localized thermodynamic chemical potentials and elemental concentrations, yet their interdiffusion fluxes and concentration gradients differ substantially.
Evaluating concentration values at the intersection point C yields a system of simultaneous linear equations for component 1 and component 2:
J_1^A = – D_11^3 (dC_1/dx)^A – D_12^3 (dC_2/dx)^A
J_1^B = – D_11^3 (dC_1/dx)^B – D_12^3 (dC_2/dx)^B
J_2^A = – D_21^3 (dC_1/dx)^A – D_22^3 (dC_2/dx)^A
J_2^B = – D_21^3 (dC_1/dx)^B – D_22^3 (dC_2/dx)^B
This formulation sets up two independent systems of two equations and two unknowns, enabling direct mathematical solution for main and off-diagonal coefficients at composition C.
Standard ASTM E1508 specifies wavelength dispersive spectrometry procedures that set peak-to-background count thresholds for trace alloying elements.

Determinant Sensitivity in Linear Gradient Systems
Calculating interdiffusion matrices from measured concentration slopes demands mathematically well-conditioned linear systems. The matrix determinant of the concentration gradients drives the stability of the extracted diffusion coefficients:
det(grad C) = (dC_1/dx)^A (dC_2/dx)^B – (dC_1/dx)^B (dC_2/dx)^A
When det(grad C) approaches zero, the concentration gradient vectors of Couple A and Couple B run nearly parallel. Near-parallel paths amplify experimental measurement noise, resulting in unphysical negative main diffusion terms or unrealistically large off-diagonal terms. Crossing angles between composition paths must exceed fifteen degrees on the Gibbs triangle to maintain acceptable error bounds.
| Couple Pair ID | Intersection (wt%) | Crossing Angle (deg) | det(grad C) | D_CrCr^Ni (10^-10 cm²/s) | D_CrAl^Ni (10^-10 cm²/s) | D_AlCr^Ni (10^-10 cm²/s) | D_AlAl^Ni (10^-10 cm²/s) |
|---|---|---|---|---|---|---|---|
| NC-1 / NA-1 | Ni-15Cr-4Al | 32.4 | 1.45 x 10^5 | 4.21 | 0.86 | -0.32 | 2.15 |
| NC-2 / NA-2 | Ni-18Cr-6Al | 18.1 | 4.12 x 10^4 | 5.08 | 1.12 | -0.45 | 2.89 |
| NC-3 / NA-3 | Ni-12Cr-8Al | 8.7 | 6.80 x 10^2 | 12.40 | -8.90 | 6.10 | 0.12 |
| NC-4 / NA-4 | Ni-20Cr-3Al | 41.0 | 2.88 x 10^5 | 3.95 | 0.74 | -0.28 | 1.98 |
| Testing conditions: Isothermal annealing performed at 1150 °C for 100 hours in quartz-encapsulated high-purity argon. Gradients evaluated via EPMA WDS linescans at 1-micron spatial steps. Note that couple pair NC-3/NA-3 exhibits extreme numerical divergence due to an ill-conditioned determinant below 10^3. | |||||||
Experimental design errors often corrupt concentration profile measurements and break analytical matrix extraction:
- Path Convergence Failure occurs when selected terminal compositions do not produce intersecting composition paths within the single-phase region, rendering simultaneous equation solving impossible.
- Parallel Gradient Alignment occurs when composition paths cross at shallow angles below ten degrees, causing matrix determinant collapse and catastrophic error amplification during linear inversion.
- Phase Boundary Crossing introduces localized slope discontinuities and two-phase microstructural bands that disrupt continuous Dayananda integration across the diffusion zone.
- Kirkendall Pore Formation causes localized density losses and physical volume expansion that distort spatial coordinate distances along microprobe linescans.
Inserting clause 4.2 into the laboratory agreement obligates the testing facility to re-run any diffusion couple pair whose composition paths cross at an angle tighter than ten degrees at contractor expense.

Matrix
Linear systems of interdiffusion equations convert measured spatial concentrations into tensorial transport properties. Modern analytical workflows isolate individual diffusion coefficients through explicit matrix algebraic manipulation. Solving these matrices accurately governs the predictive power of multicomponent phase growth models.

Which Gradient Angles Prevent Numerical Ill-Conditioning?
Spatial concentration profiles extracted from EPMA line scans yield vector components that dictate linear stability. A crossing angle of thirty degrees or higher between dual composition paths provides optimal mathematical stability. At this angle, the determinant det(grad C) stays sufficiently large to suppress experimental measurement scatter coming from microprobe wavelength dispersive spectrometers.
When composition paths meet at acute angles below ten degrees, minor fluctuations in localized element weight percentages shift calculated gradient values drastically. Maintaining high crossing angles requires selecting terminal alloys with starkly contrasting element ratios.
Composition paths crossing at near-parallel angles multiply matrix inversion errors beyond experimental utility.

Inversion of Fick-Onsager Ternary Arrays
Solving for four interdiffusion coefficients in a three-component system utilizes simultaneous concentration gradient data from two independent couple tests. Rearranging the linear flux equations into matrix notation clarifies the explicit inversion path:
= –
Inverting the gradient matrix yields direct solutions for main diagonal coefficient D_11^3 and off-diagonal cross-coefficient D_12^3:
D_11^3 = – / det(grad C)
D_12^3 = – / det(grad C)
Identical matrix operations invert the second set of equations to extract off-diagonal coefficient D_21^3 and main diagonal coefficient D_22^3 using the fluxes of component 2.
To demonstrate this extraction, take two ternary alloy couples in the Nickel-Chromium-Aluminum (Ni-Cr-Al) system annealed at 1200 °C (1473 K) for 100 hours (360,000 seconds). Assume Couple 1 (composition range Ni-10Cr-5Al to Ni-20Cr-2Al weight percent) and Couple 2 (composition range Ni-15Cr-1Al to Ni-15Cr-6Al weight percent) cross at an intersection composition C of Ni-16Cr-3.5Al weight percent.
First, evaluate the measured concentration gradients at composition C from smoothed EPMA line scans:
(dC_Cr/dx)^1 = 120 wt%/cm, (dC_Al/dx)^1 = -40 wt%/cm
(dC_Cr/dx)^2 = -15 wt%/cm, (dC_Al/dx)^2 = 85 wt%/cm
Second, calculate interdiffusion fluxes via Dayananda integration at C :
J_Cr^1 = -1.80 x 10^-6 g/(cm²·s), J_Al^1 = 0.42 x 10^-6 g/(cm²·s)
J_Cr^2 = 0.35 x 10^-6 g/(cm²·s), J_Al^2 = -0.92 x 10^-6 g/(cm²·s)
Third, calculate the determinant of the concentration gradient matrix:
det(grad C) = (120 85) – (-40 -15) = 10,200 – 600 = 9,600 (wt%/cm)²
Fourth, solve for chromium interdiffusion coefficients D_CrCr^Ni and D_CrAl^Ni:
D_CrCr^Ni = – / 9,600 = 1.45 x 10^-8 cm²/s
D_CrAl^Ni = – / 9,600 = 0.15 x 10^-8 cm²/s
Fifth, solve for aluminum interdiffusion coefficients D_AlCr^Ni and D_AlAl^Ni:
D_AlCr^Ni = – / 9,600 = -0.11 x 10^-8 cm²/s
D_AlAl^Ni = – / 9,600 = 1.08 x 10^-8 cm²/s
The positive off-diagonal term D_CrAl^Ni (0.15 x 10^-8 cm²/s) confirms that an aluminum gradient drives a positive chromium flux. The negative off-diagonal term D_AlCr^Ni (-0.11 x 10^-8 cm²/s) demonstrates that a chromium concentration gradient reduces aluminum interdiffusion transport.

Magnitude Ratios of Off-Diagonal to Main Terms
Cross-diffusion coefficients frequently reach twenty to forty percent of primary diffusion terms in concentrated superalloy systems. The ratio D_12/D_11 reflects the strength of thermodynamic elemental interaction relative to pure concentration-driven random walk diffusion. In thermodynamic terms, off-diagonal interdiffusion coefficients link directly to Darken-type atomic mobilities and Wagner interaction parameters.
Auditing calculated diffusion matrices demands reviewing key procedural checks prior to accepting reported laboratory figures:
- Symmetry Verification confirms that Onsager reciprocity relationships hold true when converting interdiffusion matrices into atomic mobility matrices.
- Eigenvalue Positivity checks that both calculated eigenvalues of the interdiffusion tensor remain strictly positive, satisfying second-law thermodynamic stability rules.
- Polynomial Boundary Smoothing verifies that microprobe raw noise was filtered using low-order localized Savitzky-Golay routines rather than aggressive global high-order polynomials.
- Mass Balance Conservation calculates total integrated mass across the diffusion interface to ensure total elemental mass balances within two percent error.
Ignoring off-diagonal diffusion terms during high-temperature coating design causes premature aluminum depletion predictions, leading to unpredicted spallation in gas turbine blades.

Bench
Laboratory execution of diffusion annealing experiments governs the physical fidelity of extracted interdiffusion coefficients. Experimental accuracy relies on precise physical furnace conditions, rigorous surface preparation, and error-free analytical instruments. Physical execution errors at the bench level corrupt raw composition linescans before numerical matrix processing ever begins.

Thermal Stability and Argon Purging Standards
Isothermal tube furnaces utilized for high-temperature diffusion couples maintain temperature uniformity within tight bands across the entire heat treat zone. A thermal drift exceeding one degree Celsius during a hundred-hour diffusion run shifts measured atomic jump frequencies, corrupting extracted interdiffusion coefficients by eight to twelve percent.
Oxygen partial pressures inside diffusion furnace tubes must remain below 10^-4 Pa to prevent internal oxidation of reactive solute elements like aluminum, titanium, or hafnium. Encapsulating diffusion couples inside double-walled quartz ampoules containing getter foils provides reliable protection against trace atmospheric leaks.
| Parameter | Target Specification | Acceptable Tolerance | Failure Impact |
|---|---|---|---|
| Isothermal Temperature | 1200.0 °C | ± 0.5 °C | Distorts atomic jump frequency and interdiffusion coefficient calculations. |
| Vacuum Pressure | 1.0 x 10^-4 Pa | < 5.0 x 10^-4 Pa | Causes interface oxidation, altering local boundary composition values. |
| EPMA Step Resolution | 1.0 µm | ± 0.05 µm | Misrepresents local concentration gradients near steep interface steps. |
| Beam Current Drift | 20.0 nA | ± 0.2 nA | Introduces false baseline concentration shifts in raw WDS x-ray counts. |
| Specimen Polishing Flatness | 0.25 µm Ra | < 0.50 µm Ra | Creates edge rounding, distorting spatial spatial coordinate calibration. |
Polished diffusion couple faces maintained under active mechanical load during vacuum heating suppress Kirkendall pore accumulation at the initial contact line.

Microprobe Step Resolution and Drift Correction
Electron probe microanalysis measures localized elemental compositions along a line transverse to the diffusion interface. Wavelength dispersive spectrometry (WDS) delivers superior energy resolution over energy dispersive spectrometry (EDS), enabling accurate separation of overlapping x-ray emission lines such as chromium K-beta and manganese K-alpha. Because EPMA beam current drifts over hours, raw intensity data requires background subtraction.
Microprobe spatial step size determines gradient accuracy. Coarse step sizes above two micrometers smooth out steep concentration changes near the initial contact plane, artificially lowering measured slopes and inflating calculated diffusion coefficients. Step sizes between 0.5 and 1.0 micrometers capture realistic interface gradients without causing excessive electron beam specimen damage.
Completing a high-temperature alloy qualification dossier requires compiling core laboratory documentation into a permanent technical file:
- Raw EPMA Intensity Datasets containing uncorrected WDS x-ray count values alongside baseline calibration standard logs.
- Furnace Thermal Log Calibration Charts recording continuous thermocouple readings throughout the isothermal diffusion annealing period.
- Optical and SEM Interface Micrographs verifying complete metallic contact and confirming zero boundary void formation along the diffusion zone.
- Homogeneity Certification Reports detailing initial composition uniformity of unannealed terminal end-member alloy blocks.
Uncalibrated baseline shifts in raw microprobe output often stem from temporary ambient temperature swings inside the instrument enclosure.

Ledger
Commercial agreements with contract research laboratories mandate explicit protocols for raw microprobe data retention, as standard lab reports often omit raw calibration logs. Purchasing empirical interdiffusion data without securing underlying raw WDS count files exposes engine manufacturers to unverifiable material inputs. Technical contracts must explicitly structure data custody requirements.

Commercial Verification of Microprobe Raw Data Files
Unprocessed WDS x-ray count rates provide the sole objective baseline for auditing contractor claims. External laboratories frequently apply proprietary smoothing algorithms before handing over final concentration profile spreadsheets. Unannounced numerical smoothing wipes out local concentration inflections that indicate real thermodynamic cross-coupling effects.
Technical specifications written into research purchase orders mandating delivery of raw microprobe binary files allow independent re-processing using standard ZAF matrix correction models. Retaining complete custody of raw intensity files guarantees that numerical gradient calculations can be re-evaluated whenever refined thermodynamic CALPHAD databases become available.

Cost Allocation for Multi-Couple Furnace Runs
Pricing structure for high-temperature interdiffusion testing combines thermal processing hours with instrument analytical time. High-temperature furnace runs lasting hundreds of hours generate significant energy and facility oversight expenses. Maximizing experimental cost efficiency involves batching multiple quartz ampoules within single isothermal furnace runs.
Microprobe analytical time represents the largest variable expense on laboratory invoices. WDS linescans crossing a 500-micrometer diffusion zone at 1-micrometer steps with 20-second dwell times per element require several hours of continuous microprobe time per sample. Balancing step count against mathematical matrix determinant requirements keeps laboratory testing budgets predictable without sacrificing matrix inversion stability.
Auditing raw baseline scans before approving final laboratory invoices secures full operational control over the extracted interdiffusion database.




