Organic semiconductor magnetoresistance is attracting interest for magnetic sensing because organic devices can respond to fields across a broad range. For engineers and technical decision-makers, however, a useful sensor requires more than a measurable current change: its response must be understood and characterized. Research on PEDOT:PSS diodes shows how combining high-field magnetoresistance measurements with electrically detected magnetic resonance (EDMR) can clarify the spin processes behind the response.
Why spin dynamics matter to organic magnetoresistance
In organic semiconductors, magnetic-field-dependent changes in electrical or optical behavior are commonly linked to the spin dynamics of charge carriers. At lower fields, interactions between carrier spins and nearby hydrogen nuclear spins create fluctuating local magnetic fields. At sufficiently high fields, even the relatively weak spin-orbit coupling typical of organic materials can become important. At very low temperatures, thermal spin polarization can also influence spin statistics.
The high-field mechanism examined in PEDOT:PSS involves weakly bound charge-carrier pairs. Each carrier has a Landé g-factor, which determines how its spin precesses in an external magnetic field. If the two carriers have slightly different g-factors, their precession rates differ. This mismatch can drive mixing between singlet and triplet spin states, changing the likelihood that a pair dissociates into mobile carriers or recombines into a more strongly bound state. Either pathway can affect the measured device current.
The size of the g-factor difference, often denoted Δg, is associated with differences in the carriers’ spin-orbit coupling. Because the effect depends on both magnetic-field strength and the lifetime of a carrier pair, it provides a route to a field-dependent magnetoconductance signal.
Why magnetoresistance data alone can be ambiguous
Phenomenological models describe the high-field response with a Lorentzian-like dependence. Its characteristic width depends on the product of the g-factor spread and the carrier-pair lifetime. This creates a challenge: fitting a magnetoresistance curve does not independently reveal either quantity. A good fit may reproduce the measured response while leaving the microscopic parameters uncertain.
The difficulty is especially clear when the device response does not reach saturation within the available magnetic-field range. In the reported PEDOT:PSS measurements, the field was swept up to ±7 T, but the curve’s baseline was not fully apparent. Different assumed values for the response width could therefore fit the measured range well, even though their predicted behavior diverged at higher fields.
This is a practical lesson for sensor characterization: a close fit over a limited range is not, by itself, proof that the underlying parameters have been uniquely determined. The measurable window and the shape of the response both affect how confidently a model can be interpreted.
Using EDMR to measure carrier g-factors independently
EDMR provides a complementary measurement. The device is exposed to microwave radiation in an external magnetic field, and the current change at spin resonance is recorded. The resonance positions provide information about the charge carriers’ g-factors. Unlike a magnetoresistance-only fit, this approach offers an independent constraint on Δg.
Measurements at multiple microwave frequencies are important. At lower frequencies, unresolved hyperfine interactions can dominate the resonance line shape and obscure small differences between the carriers’ g-factors. At higher frequencies, the effects associated with spin-orbit coupling become more distinguishable. The study used multifrequency EDMR, including measurements at high microwave frequencies, on PEDOT:PSS devices under conditions comparable to those used for magnetoresistance measurements.
Using spectra to estimate Δg requires attention to whether the g-factors of the two carriers are correlated. The difference between their average g-values and the average pairwise difference need not be the same. Accounting for this distinction changes the inferred g-factor spread and, in turn, the carrier-pair lifetime estimated from the magnetoresistance model.
What the combined measurements reveal
The measured current response contained contributions across different field ranges. Very small and intermediate-field effects were present below a few hundred millitesla, while the high-field contribution became dominant above a few hundred millitesla. A model combining these components described the response across the measured range.
To assess the high-field model, the researchers examined how fit quality changed when the assumed response width was varied and when portions of the field range were excluded. The full dataset showed a meaningful minimum in the fit-quality analysis near the fitted width of about 10 T. That minimum became less distinct as the available range narrowed. This supports the interpretation that the full measurement captures the beginning of the curve’s baseline or inflection, although the authors note that uncertainty in the estimate remains substantial.
Combining the magnetoresistance-derived width with the independently measured g-factor spread yielded carrier-pair lifetime estimates ranging from about 0.4 to 5.1 nanoseconds, depending on assumptions about pair correlation. The range illustrates why an independent measurement of Δg matters: uncertainty in the g-factor distribution can substantially affect the lifetime inferred from transport data.
The estimated lifetimes were shorter than spin-related times measured through other methods in PEDOT:PSS. One possible explanation is that steady-state magnetoresistance and pulsed EDMR emphasize different populations of carrier pairs. Longer-lived, more stable pairs may contribute more strongly to a resonance signal, while shorter-lived, weakly bound pairs may dominate the nonequilibrium current response.
Implications for organic magnetic sensors
For developers of magnetoresistance-based sensors, the central implication is methodological. Understanding a device’s field response requires measuring both its transport behavior and the spin-related material parameters that shape it. EDMR can constrain g-factor differences independently, while magnetoresistance measurements establish how the device current varies with field. Together, these measurements provide a stronger basis for interpreting model parameters than curve fitting alone.
PEDOT:PSS is of interest in this context because the study demonstrates magnetic-field-dependent responses over a wide span of field strengths, with prior work cited by the authors extending sensitivity into the nanotesla range. The results support the potential of organic magnetoresistive devices as part of a broad-range sensing approach; they do not, on their own, establish a complete sensor specification or guarantee performance in a particular industrial environment.
Conclusion
High-field organic semiconductor magnetoresistance in PEDOT:PSS is linked to field-dependent singlet-triplet mixing caused by small differences in carrier g-factors. Since magnetoresistance fitting alone cannot cleanly separate g-factor spread from pair lifetime, complementary EDMR measurements are essential for a more physically grounded interpretation. For industrial sensing development, the practical takeaway is to validate the response across the relevant field range and independently constrain the spin parameters before relying on model-derived device behavior.