Single-Pass MFM for Multi-Channel Magnetometry

• Single-pass MFM is an advanced scanning probe technique that simultaneously acquires topographic and magnetic data to quantify stray field gradients.
• It utilizes innovative excitation schemes, including amplitude-modulation, frequency-modulation, and differential force detection, to capture both out-of-plane and in-plane field components.
• The method enhances throughput, minimizes artifacts, and enables precise 3D vector field reconstruction for detailed nanoscale magnetometry.

Single-pass Magnetic Force Microscopy (MFM) is an advanced scanning probe technique that enables spatially-resolved, quantitative measurement of magnetic stray field gradients with true simultaneous acquisition of topographic and magnetic signals. In contrast to conventional two-pass MFM, which sequentially records topography and magnetic phase contrast, single-pass implementations exploit simultaneous or differential mechanical excitation, frequency demodulation, or magnetization modulation to acquire multi-channel force data in one scan. Recent advances have enabled three-dimensional vector field reconstruction by combining out-of-plane and in-plane force-gradient measurements in a single pass, significantly mitigating drift, enhancing throughput, and expanding the accessible physics—such as the detection of in-plane-vortex and skyrmion textures. Methodologies include amplitude-modulation and frequency-modulation schemes on various cantilever platforms, including custom split-electrode actuators and qPlus sensors, as well as differential force detection with switchable tip magnetization (Schmidt et al., 2023, Schneiderbauer et al., 2011, Misra et al., 2024).

1. Underlying Principles and Motivation

Traditional MFM relies on a cantilever with a magnetic tip oscillating vertically above a magnetic sample. This setup is predominantly sensitive to the vertical component of the sample’s stray field gradient, limiting information on in-plane magnetic structures. Single-pass MFM aims to overcome these limitations by enabling:

• Simultaneous measurement of multiple field components (e.g., vertical and lateral).
• Differential detection schemes to separate magnetic forces from van der Waals/electrostatic backgrounds.
• Real-time magnetic contrast acquisition without spatial/temporal separation from topographic mapping.

The need to detect in-plane stray fields arises from the growing research interest in topologically non-trivial magnetic systems, including skyrmions and vortices, where the distinction between out-of-plane and in-plane magnetization winding is crucial (Schmidt et al., 2023).

2. Instrumentation and Excitation Schemes

Split-Electrode Piezo Actuator (Three-Dimensional Single-Pass MFM)

A custom tip holder integrates a piezoelectric actuator split into two electrically independent top-electrodes (E_V and E_L) sharing a common ground. E_V is addressed by the built-in AFM flexural oscillator, while E_L is driven by an external lock-in amplifier to independently excite the torsional (lateral) mode. The mechanical configuration results in predominant out-of-plane motion for E_V and torsion about the cantilever’s long axis for E_L. Standard PPP-MFMR silicon cantilevers (k_V ≈ 2.8 N/m, f_V1 ≈ 70 kHz) are mounted for dual-mode operation. Precise alignment allows rotation of the torsional axis for mapping different in-plane directions (Schmidt et al., 2023).

qPlus Sensor Frequency Modulation

A single-prong qPlus tuning fork with high stiffness (k = 1800 N/m, f₀ ≈ 31 kHz, Q ≈ 2,000) enables frequency-modulation detection of atomic and magnetic forces in the same scan. Magnetic field sensitivity in the milli-Hertz regime is achieved by large oscillation amplitudes (A ≈ 100 nm) and low deflection noise, bridging the sensitivity gap between atomic-scale and long-range dipole-dipole interactions (Schneiderbauer et al., 2011).

Switchable Tip Magnetization (Differential Single-Pass MFM)

An “inverted” MFM geometry mounts the sample on a force-sensing cantilever while the tip is affixed to a miniaturized electromagnet. By rapidly flipping the tip magnetization (square-wave modulation at several hundred Hz), the resulting periodic reversal of the magnetic force gradient encodes the pure magnetic contribution in well-defined sidebands of the cantilever resonance spectrum. This approach achieves single-pass, differential extraction of magnetic, electrostatic, and topographic signals (Misra et al., 2024).

3. Signal Generation and Detection

Phase–Gradient Relations (Amplitude-Modulation Schemes)

The cantilever-tip is modeled as a rigid point dipole $\vec{m} = m\hat{e}_z$ in the local stray field $\vec{H}(\vec{r})$. The key phase relations for amplitude-modulation MFM are:

• Vertical (flexural) mode: $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$
• Lateral (torsional) mode: $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$

Each phase channel is sensitive to the corresponding second derivative of the stray field along vertical or lateral axes. Careful frequency spacing of drive tones ensures negligible cross-talk (Schmidt et al., 2023).

Frequency-Modulation Readout

In FM-MFM, the resonance frequency shift $\Delta f$ is proportional to the period-averaged magnetic force gradient:

$$\Delta f = \frac{f_0}{2k} \langle k_{\mathrm{mag}}(z) \rangle_t, \qquad k_{\mathrm{mag}} = \frac{\partial^2 U_m}{\partial z^2}$$

Separation of atomic versus magnetic contributions is achieved through:

• Dual-eigenmode excitation with different oscillation amplitudes.
• Distance-dependent deconvolution of $\Delta f(z)$.
• Spatial-frequency filtering of $\Delta f(x, y)$ images (Schneiderbauer et al., 2011).

Sideband Demodulation (Differential MFM)

Periodic flipping of the tip moment at frequency $f_p$ (modulation amplitude sufficient to switch magnetization) produces frequency sidebands at $f_c \pm n f_p$ in the cantilever’s response. Lock-in detection at these sidebands directly isolates the time-varying magnetic force gradient,

$\vec{H}(\vec{r})$0

where $\vec{H}(\vec{r})$1 is the local magnetic stiffness (Misra et al., 2024).

4. Sensitivity, Resolution, and Calibration

Sensitivity

MFM Implementation	Typical Force Gradient Sensitivity (N/m)	Notable Parameters
Split-Electrode Single-Pass	$\vec{H}(\vec{r})$2 (vertical), $\vec{H}(\vec{r})$3–$\vec{H}(\vec{r})$4 (lateral)	$\vec{H}(\vec{r})$5 N/m, $\vec{H}(\vec{r})$6
qPlus FM-MFM	$\vec{H}(\vec{r})$7	$\vec{H}(\vec{r})$8 N/m, $\vec{H}(\vec{r})$9
Differential (Switchable Tip)	$$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$0 per $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$1	$$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$2 N/m, $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$3

The phase and frequency noise floors are dictated by mechanical stiffness, detector noise, and environmental factors. For split-electrode MFM, noise-limited force gradient floors are $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$4 N/m (vertical) and $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$5 N/m (lateral), yielding detectable stray-field gradients of $$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$6–$$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$7 A/m·μm$$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$8 (Schmidt et al., 2023).

Spatial Resolution

• Vertical: Set by lift height ($$\delta\phi_V \simeq \mu_0 m (Q_V / k_V) \, \partial^2 H_z/\partial z^2$$9) and tip radius ($$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$0), typically $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$1.
• Lateral (single-pass, split-electrode): For the torsional channel, the resolution is $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$2.
• qPlus FM-MFM: $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$3 nm domain resolution; atomic ($$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$4 nm) with small amplitude.

Calibration of mechanical constants ($$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$5, $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$6) utilizes thermal noise or Sader’s method; $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$7-factors are determined from resonance or ring-down measurements (Schmidt et al., 2023, Schneiderbauer et al., 2011).

5. Scanning Protocols and Data Processing

Experimental Protocol

• Mount standard magnetic cantilever (e.g., PPP-MFMR) in split-electrode holder.
• Drive flexural and torsional modes on separate electrodes (frequencies $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$8, $$\delta\phi_L \simeq \mu_0 m (Q_L / k_L) \, \partial^2 H_z/\partial x^2$$9; $\Delta f$0 nm, $\Delta f$1 nm).
• Scan in lift mode: record topography, then scan at $\Delta f$2 nm with both channels active, acquiring $\Delta f$3 and $\Delta f$4 simultaneously (Schmidt et al., 2023).

Reconstruction Algorithms

• Post-process demodulated signals: $\Delta f$5, $\Delta f$6.
• Fourier-invert second derivatives to reconstruct 3D $\Delta f$7.
• Combine with multiple sample/cantilever azimuths to map full vector field, integrating derivatives to reconstruct $\Delta f$8.
• Use digital mixers, finite impulse response filtering, and FFTs for sideband-based signal extraction (Schmidt et al., 2023, Misra et al., 2024).

6. Comparative Analysis with Two-Pass MFM

Single-pass MFM offers several critical advantages:

• Simultaneity: Both out-of-plane and in-plane channels are acquired in the same scan, eliminating drift and environmental discrepancies inherent in two-pass approaches.
• Throughput: Acquisition time is reduced (single line-scan time), and real-time feedback is possible.
• Artifact Suppression: Topographic-magnetic crosstalk is suppressed; genuine magnetic features are consistently present in both signal channels, in contrast with purely topographic artifacts (Schmidt et al., 2023).

Limitations include lower SNR in in-plane (torsional) channels (typically 2–5× worse than vertical amplitude), increased experimental complexity (custom hardware and phase control), and the requirement for multiple orientations to complete a 3D map. For sideband/differential approaches, SNR is reduced by additional technical noise, and maximal switching rate is limited by coil inductance and potential thermal drift from Joule heating (Misra et al., 2024).

7. Future Directions and Practical Considerations

Improvements in hardware—e.g., high-$\Delta f$9 trampoline or membrane-based force sensors—may yield three orders of magnitude gain in moment sensitivity, approaching the detection threshold for single-electron or nuclear spins (Misra et al., 2024). Microfabricated write-head coils promise more efficient, lower-dissipation tip magnetization switching. Limitations on switching rates, SNR, and spurious stray fields from actuators and tip mounting can be mitigated with further development. A plausible implication is the extension of single-pass MFM to quantum sensing regimes or real-time imaging in dynamic magnetic field environments.

In summary, single-pass MFM unifies multidimensional force-gradient sensing, rapid data acquisition, and robust artifact suppression, extending the capabilities of scanning probe magnetometry for fundamental and applied studies of nanoscale magnetism (Schmidt et al., 2023, Schneiderbauer et al., 2011, Misra et al., 2024).