Compositional and Magnetic Characterisation of Oblique Co and Fe Nanowire Structures Fabricated Using Focused Electron Beam Induced Deposition

This study demonstrates that focused electron beam induced deposition (FEBID) of cobalt and iron nanowires results in reduced metal content and magnetic induction at oblique growth angles due to non-uniform growth dynamics, but these variations can be mitigated by optimizing beam parameters such as using low voltage and high current to fabricate structures with consistent composition across angles from 0° to 60°.

The Gist

Imagine you are an architect trying to build a tiny, 3D bridge out of pure metal using a high-tech "pen" that draws with electrons instead of ink. This pen is called Focused Electron Beam Induced Deposition (FEBID). It works by shooting a beam of electrons at a surface while spraying a special gas. The electrons hit the gas, breaking it apart so that metal atoms stick to the surface, building up a structure layer by layer.

The problem the scientists in this paper faced is like trying to draw a perfect, straight line while walking sideways. When the electron beam stays still, it builds a tall, straight tower (a vertical nanowire) that is very pure and strong. But, to build a 3D bridge or an arch, the beam has to move. As the beam moves to create an angle, the "ink" (the metal) starts to get mixed with "dirt" (carbon and oxygen contaminants), making the structure weaker and less magnetic.

Here is the story of how they solved this, explained simply:

The Problem: The "Moving Pen" Effect

Think of the electron beam as a spotlight.

• When the spotlight is still (Vertical wires): It shines intensely on one spot. The gas breaks apart cleanly, leaving behind almost pure metal. The result is a shiny, strong, magnetic wire.
• When the spotlight moves (Oblique/Angled wires): As the beam travels to draw a curve or an angle, it spends less time on any single spot. It's like trying to paint a wall while walking; the paint gets thinner and messier. The beam also hits the structure from different angles, causing the metal to mix with leftover gas molecules. The result is a wire that is "diluted" with non-magnetic junk, making it a poor conductor of magnetism.

The Experiment: Testing 41 Different "Drawings"

The researchers built 41 tiny wires made of Cobalt (Co) and Iron (Fe). They drew them at different angles, from straight up (0°) to lying flat (90°). They wanted to see exactly how much the "purity" dropped as the angle increased and if they could fix it by changing the settings on their electron pen.

They tested three main "knobs" on their machine:

1. Voltage (The Power): How hard the electrons hit.
2. Current (The Intensity): How many electrons are in the beam.
3. Gas (The Ink): Whether they used Cobalt gas or Iron gas.

The Discovery: Finding the "Sweet Spot"

They found that the "moving pen" problem wasn't the same for every setting.

• High Voltage (30 kV): This was like using a very powerful, wide spotlight. When the beam moved, it spread out too much, hitting the sides of the wire and creating a very messy, oval-shaped wire with lots of impurities. The metal content dropped significantly as the angle increased.
• Low Voltage (5 kV) + High Current: This was the winning combination. Think of this as a dimmer, but very concentrated, laser-like beam. By using a lower voltage, the electrons didn't penetrate as deeply or spread out as much. By cranking up the current, they ensured there were enough electrons to break the gas molecules apart efficiently, even while the beam was moving.

The Iron vs. Cobalt Difference:
They also found that Iron was a more "cooperative" material than Cobalt. When they used the Iron gas, the wire stayed pure and round even at steeper angles. The Cobalt wire, however, got messy and oval-shaped much faster as the angle increased.

The Result: A Stronger 3D Bridge

By using the Low Voltage (5 kV), High Current, and Iron gas, they managed to build angled wires that stayed almost as pure and magnetic as the straight ones, at least up to a 60-degree angle.

They also used a special microscope technique (like a super-powered X-ray vision) to look inside the wires. They saw that when the wires were pure, they acted like strong magnets. But when the wires were "diluted" with impurities (because the beam moved too fast or the settings were wrong), the magnetic strength dropped. It's like a team of runners: if everyone is fit (pure metal), they run fast together. If many are tired or injured (impurities), the whole team slows down.

The Bottom Line

The paper concludes that you can build complex 3D magnetic shapes (like bridges or arches for future computer chips) without them falling apart or losing their magnetic power, if you tune your electron beam correctly. Specifically, you need to use a "gentle but intense" beam (low voltage, high current) and the right type of gas (Iron). This keeps the "ink" pure even when you are drawing at an angle, ensuring the tiny 3D structures work exactly as intended.

Technical Summary

Technical Summary: Compositional and Magnetic Characterisation of Oblique Co and Fe Nanowire Structures Fabricated Using Focused Electron Beam Induced Deposition

Problem Statement
Focused Electron Beam Induced Deposition (FEBID) is a critical additive manufacturing technique for creating 3D ferromagnetic nanostructures, such as those required for spintronic devices (e.g., magnetic racetrack memory and artificial neural networks). While FEBID allows for the fabrication of complex 3D geometries, including bridges and arches, a significant challenge arises when translating the electron beam to grow structures at oblique angles. The variation in growth dynamics and electron beam interaction volumes during translation leads to non-uniform atomic composition within the deposited nanostructures. Specifically, the purity and magnetic properties of FEBID materials are known to depend on the electron beam orientation. However, the specific relationship between the growth angle, metal content, and magnetic induction in oblique ferromagnetic nanowires (NWs) remains under-characterized, hindering the reliable design of bespoke 3D magnetic devices.

Methodology
The study employed a systematic approach to fabricate and characterize 41 Cobalt (Co) and Iron (Fe) nanowire structures with growth angles ($\theta$) relative to the optic axis ranging from $0^\circ$ (vertical) to $90^\circ$.

• Fabrication: Samples were created using a dual-beam Thermo Fisher Scientific HELIOS Xe plasma FIB-SEM. Two precursor gases were utilized: $Co_2(CO)_8$ and $Fe_2(CO)_9$. Deposition parameters were varied across nine datasets, including different accelerating voltages (5 kV and 30 kV), beam currents, and precursor types. To isolate the effects of beam translation, reference samples were deposited with a stationary beam, while oblique samples were created by translating the beam at controlled rates.
• Structural Characterization: Scanning Electron Microscopy (SEM) was used to measure NW dimensions at various tilt angles to determine cross-sectional ellipticity.
• Compositional Analysis: Scanning Transmission Electron Microscopy (STEM) coupled with Electron Energy Loss Spectroscopy (EELS) was performed to map elemental composition. The analysis focused on the central 20 nm band of the NWs to quantify metal content (Co or Fe atom %) while accounting for the radially layered structure (metal-rich core, carbon/oxygen-rich shell).
• Magnetic Characterization: Off-axis electron holography was conducted in a Lorentz mode TEM to reconstruct magnetic phase shifts and calculate the magnetic induction ($B_y$) parallel to the NW axis. Samples were saturated using an external magnetic field to isolate the magnetic contribution from the electrostatic potential.
• Simulation: Micromagnetic simulations (Mumax3) were performed on cylindrical models to verify that the observed NWs (diameters < 160 nm) would exhibit uniform uniaxial magnetization, validating the use of thickness-averaged induction measurements.

Key Results

1. Composition vs. Growth Angle: A clear negative correlation was observed between the growth angle ($\theta$) and metal content. As the beam translation speed increased (increasing $\theta$), the metal purity decreased.
   • At 30 kV, the reduction in metal content was significant, reaching up to 0.4% per degree of growth angle for Co NWs.
   • At 5 kV, the reduction was minimized. The Fe$_2$(CO)$_9$ precursor at 5 kV demonstrated the highest stability, with a purity loss of only 0.1% per degree.
2. Magnetic Induction: The magnetic induction ($B_y$) within the NWs decreased proportionally with the loss of metal content.
   • For Co NWs at 5 kV, $B_y$ decreased by approximately 2 mT per degree of $\theta$.
   • For Fe NWs at 30 kV, the decrease was more pronounced at 7 mT per degree.
   • The study confirmed that the reduction in magnetic signal is directly linked to the reduced ferromagnetic volume and increased non-magnetic contaminants (C, O) caused by the oblique growth geometry.
3. Cross-Sectional Geometry: The cross-sectional shape of the NWs became increasingly elliptical at higher accelerating voltages (30 kV) and larger growth angles due to the elongation of the electron interaction volume. At 5 kV, the interaction volume remained localized, preserving a more circular cross-section.
4. Optimization: The combination of the lowest viable accelerating voltage (5 kV), the highest viable beam current (2.8 nA), and the $Fe_2(CO)_9$ precursor yielded the most uniform oblique structures. Under these conditions, metal content remained above 75% for growth angles up to $60^\circ$.

Significance and Claims
The paper claims to provide essential characterization data linking the geometric growth angle of FEBID nanowires to their compositional and magnetic properties. The authors assert that:

• Uniformity is Achievable: By carefully selecting deposition parameters (specifically low voltage and high current), it is possible to fabricate oblique ferromagnetic NWs with consistent metal content and magnetic induction up to $60^\circ$ from the vertical axis.
• Parameter Tuning: The study demonstrates that the "trade-off" between beam energy (interaction volume) and beam current (intensity) can be managed to mitigate the non-uniformity inherent in 3D FEBID printing.
• Material Specificity: The choice of precursor gas significantly impacts stability; $Fe_2(CO)_9$ was found to be more robust against translation-induced purity loss than $Co_2(CO)_8$ under the tested conditions.
• Implications for 3D Spintronics: Ensuring uniform composition is critical for the reliable operation of 3D spintronic devices. The findings suggest that current FEBID algorithms for 3D printing can be refined to account for these angle-dependent variations, enabling the fabrication of complex magnetic architectures (such as helical lattices or 3D data storage) with predictable magnetic behavior.

The authors conclude that while oblique FEBID structures inherently suffer from reduced purity compared to vertical ones, this degradation can be minimized to a level suitable for functional device prototyping through the optimization of beam parameters and precursor selection.