Dropping test

Targeting a first demonstration of the sensors capability to detect magnetic nanoparticles (MNPs), a Nanoscience Bioforce system (Figure 3.6) was employed to spot micron-sized droplets of solution on specific positions. The instrument consists in several parts. A holder of 14 𝑐𝑚 x 14 𝑐𝑚, which can accommodate the sample, is deputed to perform the in-plan movements with micrometric resolution, in order to align the substrate with the spotting column, that is made by an AFM-like cantilever having a hole at its end and connected to a microchannel for solution’s sample delivery.

The distance from the surface is measured at high sensitivity through a laser which is reflected to the cantilever and is collected by a 4-quadrant photodetector: in this way it is possible to operate in safety conditions avoiding breaking the tip. The column movement is driven by a micrometric actuator. Furthermore, a reflective microscope system allows to visualize the system and the tip with the ability to memorize two different focusing positions. All the system is shielded by a glass box equipped with a module for controlling the humidity, which is a key parameter to perform the liquid deposition and should be optimized for any different solution and substrate.

While approaching the surface, the cantilever bends and there is a droplet formation, which holds on the substrate when the cantilever is pulled up. Its dimension depends on the environmental conditions like temperature and humidity, but also on the speed of moving down (that influences the force of bending) and the deposition time at minimal distance. This

60 procedure has been optimized on silicon-based substrates with Ruthenium deposited on top, to mimic the release on magnetoresistive structures to obtain the spotting of droplets having 10 𝜇𝑚 diameter with few percentages of deviations and high reproducibility.

Figure 3.6: micron-sized droplet of MNPs solution release on PHE based sample.

Specifically, for the experiments, a solution of magnetic nanoparticles (provided by University of Santiago de Compostela, a MADIA project partner) was spotted. The best results were obtained with 10 𝑛𝑚 multicore magnetite coated by another 10 𝑛𝑚 of Silica with spherical shape. The concentration has been optimized by looking for a compromise: at high concentration, there were some problems related to particles’ sticking within neighbors, obtaining clusters which blocked the on-tip channels. On the other hand, at very low concentration, reproducibility was poor in terms of the number of particles’ in different droplets. An equilibrium was reached at about 𝑚𝑔/𝐿, assisting the formation with a constant magnetic field provided by permanent magnets placed below the sample, to counteract particles’ attraction from the tip holder (made by magnetic material). The protocol started with the characterization of the magnetoresistive structures in a static magnetic field, to obtain a

61 baseline for the signal to be compared to the one eventually changed by the presence of the magnetic moments of nanoparticles.

Figure 3.7: Droplet released on GMR (a) and PHE (b) based samples.

Then, we spotted a standard (10 𝜇𝑚 of diameter containing 10 particles) droplet on a GMR or PHE based samples (Figure 3.7). We didn’t use TMR sensors since the layout performed on that kind of structures did not allow us to spot a droplet in the surface and contact them at the same time. Furthermore, for MTJs, the vertical transport, so the only configuration possible, lets the particles to be farer from the sensing layer than the other series of sensors, so mitigating the bigger signal excursion and sensitivity. After spotting, the number of particles on top was checked by Scanning Electron Microscopy (SEM) imaging, then Energy Dispersed X-Ray (EDX) analysis confirmed their composition.

Any sample has been characterized again in a static magnetic field with the same measurement parameters used before the spotting. Considering that the x-axis reports only the magnetic field due to the Helmholtz coils, obtained after a calibration with a standard Hall teslameter followed the fitting of the measured field versus the current provided by its power supply, a shift in the response curve represents a clear indication of magnetic particles’ sensing by the magnetoresistive structures developed and fabricated within the project.

Figure 3.8 reports the shift obtained with GMR based sensors with three small clusters on top of the sensing region. Notably, the shape of the signal is exactly the same as it was before and it results stable after repeated measurements, since the other loop performed, named nps_2 in Figure 3.8 (b) can be superposed on the first one (nps_1).

(a) (b)

62

Figure 3.8: (a) SEM image of a GMR junction (MNPs highlighted), (b) sensor response shift obtained by repeating the characterization after micron-sized droplet spotting on the sensing surface.

Figure 3.9 shows the results for PHE samples, with an appreciable translation obtained by a single small cluster of magnetic moments located on the sensing surface and, also in this case, the same shape as in absence of them.

This proof of concept shows that the structures fabricated are sensitive enough to sense this few magnetic moments, which generates a small magnetic field, near their surface.

After the measurement, the surface of all the samples has been investigated again through SEM imaging ad EDX, to be sure that particles’ number was unaffected by the measurement, for example because of the attraction due to the coils: this confirmed the hypothesis of the very strong interaction between surface and the Silica shell.

Figure 3.9: (a) SEM image of a PHE sample (MNPs highlighted), (b) sensor response shift obtained by repeating the characterization after micron-sized droplet spotting on the sensing surface.

(a) (b) (a) (b)

63