Robust absolute magnetometry with organic thin-film devices

Abstract

Magnetic field sensors based on organic thin-film materials have attracted considerable interest in recent years as they can be manufactured at very low cost and on flexible substrates. However, the technological relevance of such magnetoresistive sensors is limited owing to their narrow magnetic field ranges (∼30 mT) and the continuous calibration required to compensate temperature fluctuations and material degradation. Conversely, magnetic resonance (MR)-based sensors, which utilize fundamental physical relationships for extremely precise measurements of fields, are usually large and expensive. Here we demonstrate an organic magnetic resonance-based magnetometer, employing spin-dependent electronic transitions in an organic diode, which combines the low-cost thin-film fabrication and integration properties of organic electronics with the precision of a MR-based sensor. We show that the device never requires calibration, operates over large temperature and magnetic field ranges, is robust against materials degradation and allows for absolute sensitivities of <50 nT Hz(-1/2).

Figure 1. Device concept of an organic semiconductor MR-based magnetometer (MRM).

(a) The device consists of an organic diode structure (inset, the layers are not to scale), which is located above two mutually perpendicular striplines required for on-chip spin resonant excitation and field modulation. Electron and hole polarons are injected from opposite sides into the diode structure and recombine spin dependently in the organic semiconductor. (b,c) The magnetic field response of a DC current (no modulation) in a bipolar MEH-PPV diode as a function of magnetic field as RF radiation (200 MHz in (b); 50 MHz in (c) is applied. Reductions in the current are seen when MR conditions are satisfied. These are more pronounced when the applied field B₀>ΔB_Hyp where MR-induced spin mixing dominates. (d,e) Schematic illustration of the origin of resistance changes owing to spin mixing induced by the local hyperfine fields (ΔB_Hyp) and owing to MR excitation. All measurements were performed at room temperature.

Figure 2. Calibration of the MRM.

(a) Resonance spectrum for 350 MHz radiation, measured using pulsed resonant excitation. Note that the current change after the excitation pulse was detectable for 1 ms undergoing a quenching/enhancement transient that is known for spin-dependent pair processes. The data presented here were measured 20 μs after the pulse excitation to maximize signal to noise. The spin resonance used for the MRM device is the narrow (blue) component of the spectrum. The red component represents the wide peak, and the green curve is the fit consisting of the sum of red and blue curves. (b) Plot of the peak magnetic field where maximal MR-induced current change is measured as a function of the applied excitation frequency, following a linear relationship (note that the error of the data points is below the size of the symbols). A linear fit of the above data yields a gyromagnetic ratio γ=28.03(4) GHz T⁻¹ and a corresponding g-factor g=2.0026(4). Thus, the electrically detectable electron gyromagnetic ratio can be used as an absolute magnetic field standard.

Figure 3. Robustness and sensitivity limits of the magnetometer device.

(a) The current–voltage characteristics of a device at 5 K before (solid line) and after (dashed line) intentional significant degradation in air. (b) The gyromagnetic ratio, γ, measured as a function of temperature and degradation. The closed points derive from pristine devices and the open circles (with coloured error bars) from two of these devices after degradation. The error bars are upper estimates obtained from the fits of the individual spectra. The grey bar represents the s.d. obtained from all data points. The red solid line gives the temperature average of all data. Within this range, neither the change in temperature nor degradation of the materials impact the reproducibility of the gyromagnetic ratio. (c) Plot of the MR peak width (left axis) as well as the resulting field resolution (right axis) as a function of the externally applied magnetic field. The data belong to both the left and right axis. At low magnetic fields, the hyperfine interaction dominates the resonance widths while spin-orbit contributions (which cause g-factor inhomogeneities) dominate at high magnetic fields leading to spectral broadening. The dashed red line represents a fit of the data with both hyperfine field strength and spin-orbit (g-factor) distribution. The dashed black line shows the spin-orbit-induced broadening obtained from the fit of the data.

Figure 4. An integrated absolute magnetic field sensor.

The magnetic field-modulated current change in an integrated device as sketched in Fig. 1a as a function of stripline frequency in a static magnetic field of 8.93 mT. A small modulation field of 0.05 mT is applied to the static field at a frequency of 6 kHz via the second stripline (labelled B_mod in Fig. 1a) to enable lock-in detection. The data show the presence of two Gaussian resonances (red and blue curves), with the narrow resonance (blue fit) being significantly more pronounced in the lock-in-detected derivative spectrum. Note that the presence of the broader of the two resonance lines (red) does not compromise the measurement as both resonances exhibit identical gyromagnetic ratios. The green curve represents the fit of the data using the sum of the two Gaussians.