Lilienfeld 1926 Attribute the Modulation Mechanism
# Attribute the Modulation Mechanism: Path B Must Be Ruled Out Before Anything Is Attributed to the Film
The 1925 protocol used frequency response, temperature dependence, and polarity-history memory to discriminate genuine field-effect modulation from competing mechanisms like ion transport. This construction adds a prior question those three tests must answer before they can be trusted: does the fracture-margin leakage path (Path B, first identified in Step 8) have its own frequency and temperature signature that could imitate, or mask, the signature of genuine field-effect modulation in the film itself? Mechanism attribution here is a two-stage process — rule out Path B's contribution first, then apply the 1925 discrimination tests to what remains.
Path B must first be independently characterized at the test frequencies planned for mechanism discrimination. Step 8 already measured Path B's magnitude at zero signal frequency (the DC plateau after bias settling). Before running the frequency-response discrimination test, this device must be checked for whether Path B itself has frequency dependence — moisture-assisted ionic conduction along a fracture margin often shows its own corner frequency, distinct from both the pure capacitive feedthrough corner (set by C_coupling) and the genuine field-effect corner. If Path B's corner coincides with or sits near the expected field-effect corner, the measured total cannot be cleanly attributed by frequency shape alone, and a separate suppression or subtraction method is required.
Temperature dependence must distinguish three activation energies, not two. The 1925 protocol compared the activation energy extracted from ln(G_avail) vs. 1/T against the baseline conductance g(0)'s own activation energy, to tell field effect from ion transport. This construction must add a third comparison: Path B's own temperature dependence, since moisture-assisted conduction along a fracture margin typically has a distinctly different (often lower) activation energy than either the semiconductor film's intrinsic conduction or genuine field-effect modulation.
Trapped charge at the fracture margin can produce its own polarity-history memory, separate from the film's. The 1925 protocol tested whether sustained +V or −V conditioning left a retained zero-bias gain shift in the film, indicating film-level memory. This construction must run a parallel test on the control-electrode leakage current itself: if sustained conditioning leaves a retained shift in zero-bias I₁₃ (not just in the film's gain), that indicates charge trapping specifically at the fracture margin — a construction-specific memory effect the 1925 flat-electrode geometry had no equivalent pathway to produce.
| Step | Process operation | Input | Output | Specification | Constraint |
|---|---|---|---|---|---|
| 13.1 | Measure Path B's own frequency dependence at planned discrimination test frequencies | Stability-confirmed device from Step 12 | Path B frequency response curve | Measured independently, isolating I₁₃ response across the frequency range planned for Step 13.3 | Skipping this risks attributing Path B's frequency shape to the film's own mechanism |
| 13.2 | Compare Path B's corner frequency against the predicted field-effect corner from C_coupling | Path B curve from 13.1, C_coupling from Steps 1/3 | Overlap assessment | If corners are well-separated, proceed; if they overlap, flag for suppression method in 13.3 | Proceeding without this check risks an unresolvable ambiguity in the measured total |
| 13.3 | Measure frequency response of G_avail across the test range, applying Path B suppression if flagged in 13.2 | Overlap assessment from 13.2 | Field-effect frequency response | Compared against the capacitive-only prediction, matching the 1925 discrimination logic | An unsuppressed overlap produces a response shape that cannot be cleanly attributed |
| 13.4 | Measure temperature dependence of g(0), G_avail, and Path B's plateau I₁₃, each independently | Suppressed/clean responses from 13.3 | Three Arrhenius datasets | Each fit separately to extract E_a,film, E_a,field, and E_a,leak | Combining these into one fit obscures which mechanism explains the observed temperature behavior |
| 13.5 | Compare the three activation energies; confirm E_a,field is distinct from both E_a,film and E_a,leak | Three E_a values from 13.4 | Mechanism attribution for temperature dependence | E_a,field should not coincide with E_a,leak; coincidence invalidates attribution to genuine field effect | A match between E_a,field and E_a,leak means the "field effect" signal may actually be Path B's temperature dependence |
| 13.6 | Apply sustained +V conditioning, measure zero-bias g(0) and zero-bias I₁₃ after returning to zero | Temperature-characterized device from 13.5 | Post-conditioning film and leakage readings | Both quantities measured, not just film gain as in the 1925 protocol | Measuring only film gain misses fracture-margin-specific memory entirely |
| 13.7 | Repeat with sustained −V conditioning; compare both quantities' retained shifts across polarities | Readings from 13.6 | Polarity-history memory record, both channels | Film memory and Path B memory reported and compared independently | A combined memory reading cannot distinguish which mechanism is responsible for any observed retention |
| 13.8 | Synthesize final mechanism attribution, explicitly stating whether Path B was ruled out or required suppression | All prior results | Complete mechanism attribution record | Attribution conclusion includes the Path B ruling as a stated precondition, not an implicit assumption | Omitting the Path B precondition leaves the attribution unfalsifiable against this construction's own known confound |
Step 13 does not discover a new mechanism; it certifies that the mechanism attributed to the film is not actually Path B wearing the field effect's clothing. Every discrimination test the 1925 protocol used — frequency response, temperature dependence, polarity-history memory — remains valid here, but only after this construction's specific leakage pathway is shown to be either negligible or cleanly separable at the test conditions used. A mechanism attribution that skips this precondition is attributing a result to the film that may, in whole or in part, belong to the fracture margin instead.