Development of ultralight nanocellulose magnets using ultrasonic agitation

Nanocellulose has been extensively studied in recent years due to its excellent mechanical properties, wide variety of potential uses in composite materials, and the overall abundance and renewability of its starting product, cellulose.^1–5 Nanocellulose has most commonly been synthesized using chemical means such as 2,2,6,6-tetramethylpiperidine 1-oxyl-oxidation and acid/enzymatic hydrolysis.^6,7 Lesser used methods such as mechanical cryo-crushing and ultrasonication have occasionally been utilized as pretreaters, but rarely are the sole means in which cellulose is broken down into nanocellulose.¹ However, using refined cellulose as a starting material, it is quite possible to create nanocellulose with properties matching chemical methods using only ultrasonication.^1,8 This allows one to obtain the desired physical properties of nanocellulose without the waste products inherent in other methodologies.

The most commonly cited methods for creating aerogels are vacuum freeze drying and the use of supercritical CO₂ drying on nanocellulose suspensions. While having some advantages, CO₂ drying is a slower, more complex, and more expensive method.^9,10 Nanocellulose aerogels have been successfully created with a structurally robust yet highly porous, foamlike structure of extremely low density and high surface area. Densities of 0.03 g/cm³ have been achieved in pure aerogels, which were approximately 98% air.^6,11 Nanocellulose aerogels provide a strong, porous network which can be easily used to incorporate other materials in forming ultralight composites. Some composites that have been created are oil absorbing, hydrophobic aerogels along with polyvinyl alcohol, silver nanoparticle, or montmorillonite composite aerogels, all with varying potential uses.^1,11–15 Other research has shown that magnetic nanocellulose composite aerogels can be created by adding ferromagnetic nanoparticles to the aerogel after its formation. This process creates a light weight, magnetic aerogel, which responds to a simple, household magnet.¹⁶ These magnetic composites can achieve densities similar to that of pure nanocellulose aerogels, depending on how much magnetic material is added.^16,17

In this work, we have utilized only ultrasonic agitation to form ultralight magnetic composites from finely ground commercially available AlNiCo powders and microcrystalline cellulose. Previous studies in which magnetic nanocellulose aerogels were created incorporated magnetic nanoparticles after aerogel formation^16,17 or used preformed magnetic nanoparticles.¹⁸ Results from this simplified process are comparable to other means with regard to structural stability, density, and magnetic character. As with other cellulose based magnetic aerogels, these composites can be compressed to form a magnetic paper. Alternately, the magnetic nanocellulose suspensions can be air dried to form a hard and sturdy solid magnetic cellulose material. These magnetic materials could be used in microfluidic devices, electronic actuators, and in other areas where weight is a critical constraint and magnetism is desired.¹⁷ 

Commercial AlNiCo powders were purchased from various scientific education equipment distributors. Microcrystalline cellulose (99.99% purity) was purchased from Alfa Aesar. The suspensions were prepared using distilled water.

AlNiCo powders were used as purchased or after grinding with a ceramic mortar and pestle for 10 min. Initial grinding was found to improve both incorporation of AlNiCo and homogeneity. Without pregrinding, there would often be residual AlNiCo powder left at the bottom of the beaker, likely due to larger sized particles that precipitated out quickly from the suspension. It was essential that the magnetic powder was not magnetized before ultrasonication. Magnetized powder formed clumps and would not fully disperse even after long periods of agitation by the ultrasonic probe.

The suspensions were created using a two-step process using a Sonics Vibra-Cell 20 KHz, 500W Ultrasonic Probe with a ¼ in. diameter removable Titanium tip. Suspensions were formed using 100 ml of distilled water in a 400 ml Berzelius beaker. The beaker was cooled using external copper coils to maintain a temperature of 20 °C at which the optimum power transfer from the ultrasonic probe was achieved. The probe was run at 80% amplitude which resulted in a 16 ± 2% power transfer.

In the first step, AlNiCo powder, either as-bought or finely ground with a ceramic mortar and pestle, was added to the water. The powder was dispersed into the distilled water via ultrasonic agitation for 30 min. It was found that using longer times did not show noticeable improvements.

The microcrystalline cellulose was then added to the dispersion. Ultrasonic agitation was subsequently performed at the same power settings for another 90 min. At this point, the cellulose was broken down into nanocellulose, forming a thick, viscous suspension. Again, longer times did not show significant improvements. In fact, longer ultrasonic times increased the amount of titanium impurities left in the sample from ultrasonic probe wear. Probe wear could be detected both in a color change in the sample (normally white for pure nanocellulose) and through chemical analysis using energy dispersive x-ray (EDX) spectroscopy.

Investigations included adding various amounts of cellulose and AlNiCo powders to the distilled water to form the suspension. It was found that between 2 and 10 g of cellulose per 100 ml of water gave rise to a thick, viscous suspension that appeared to have a complete breakdown into nanocellulose. If too much cellulose powder was added, the solution appeared to become too viscous before the breakdown was complete. If too little cellulose was used, the suspension remained watery so that when the magnetic powder was included it was not clear if it integrated homogeneously.

Various ratios of AlNiCo to cellulose were also explored. It was found that up to a 1:1 mass ratio of AlNiCo to cellulose resulted in a reliable and homogeneous integration into the nanocellulose suspension. If too much AlNiCo powder was used, the process was less reliable and at times some of the AlNiCo powder remained as residue on the bottom of the beaker or formed small clumps in the suspension. Except where noted, the results shown in this study were from samples formed using 100 ml of distilled water, 5 g of microcrystalline cellulose, and 5 g of AlNiCo powder. At these concentrations, the resultant light gray suspension was very thick, viscous, and appeared almost entirely homogeneous.

There were two important caveats for the ultrasonication process. First, temperature was very important. When cooling was not used, the water would heat up during the process and power transfer would be reduced. Cooling also prevented evaporation, which was insignificant when the system was held at a temperature of 20 °C. Second, probe wear was very important. During the ultrasonication process, the cavitation which breaks up the material into smaller particles also breaks down the titanium ultrasonic probe, forming small pits. Not only does this probe wear reduce power transfer and lead to titanium impurities, but it also can severely impede the inclusion of the magnetic powder. Being magnetic, the AlNiCo powder had a tendency to clump together, especially in the pits formed on the ultrasonic probe. It was easy to find AlNiCo powder residue stuck to the probe tip when it showed significant signs of wear. Incorporation of AlNiCo and homogeneity was greatly improved by using new probe tips or ones that at least showed few signs of pitting. Accumulation of AlNiCo powder on the tip could also be detected by a reduction in the power transfer as measured by the sonication instrument. During the 90 min of ultrasonication used to break down the cellulose, the beaker was also shifted about every 30 min. This motion helped to break up any AlNiCo that might be stuck to the probe, maintained optimum power transfer from the probe to the suspension, and also improved homogeneity.

Once formed, the magnetic nanocellulose suspension could be air dried to form a hard solid, applied to various surfaces to form thin films or sheets, or freeze dried to form aerogels. The suspension and all such solids were magnetic, and the solids could be readily magnetized. If the suspension in liquid form was exposed to a magnet, the AlNiCo powder would be attracted to it, destroying homogeneity as the AlNiCo accumulated on the sides of the beaker.

Aerogels were created using a standard vacuum freeze drying process. The suspension was first transferred to a beaker and subsequently frozen in liquid nitrogen. It was found that 1 h was sufficient to completely freeze samples with volumes of 20–50 ml. The frozen sample was then attached to a Labconco FreeZone 4.5 l freeze dryer for at least 48 h to ensure all water was sublimated. The resultant aerogels had a foamlike texture and did not differ much in their physical characteristics, except for color, between suspensions containing different concentrations of AlNiCo.

The density of the aerogels was determined by cutting the samples into small cubes using a scalpel and then measuring both mass and volume. Sample homogeneity was checked with both optical microscopy and scanning electron microscopy (SEM) using a Tescan Vega II electron microscope. To prevent charging, it was necessary to first flatten the aerogels and then deposit at least 5 nm of gold (Cressington 108 Auto Sputter Coater) onto their surface for charge dissipation. Chemical analysis was performed within the SEM using a Bruker Quantax EDX system. Particle size analysis was performed using gwyddion software. Magnetic measurements were done with a Quantum Design Physical Property Measurement System.

It was obvious from visual inspection that pregrinding of the AlNiCo powder led to more complete and more homogeneous incorporation of the AlNiCo powder into the nanocellulose suspension. Figure 1 shows a picture comparing aerogels formed with pure cellulose, as-bought AlNiCo/cellulose, and AlNiCo/cellulose where the AlNiCo was initially ground using a mortar and pestle. Samples using preground AlNiCo were overall darker in appearance and did not contain the visible clumps as seen in samples using as-bought AlNiCo powder.

Aerogels formed using 5 g of cellulose (1.5 g/cm³) in 100 ml of water resulted in an overall density of 0.06 g/cm³ indicating the aerogel is 97% air by volume, as expected from the water to cellulose ratio of 96.8%. For the samples with 5 g of cellulose and 5 g of as-purchased AlNiCo (6.8 g/cm³) powders, densities were 0.12 and 0.17 g/cm³ for AlNiCo sonication times of 30 and 45 min, respectively. Longer sonication times had no effect on either the density or the appearance of the sample in which small but visible clumping could be detected by eye. When the AlNiCo powders were first finely ground before sonication, the density of the composite aerogels was measured to be 0.21 g/cm³ with no gains to be had by increasing sonication times beyond 30 min. If the AlNiCo powders were fully incorporated and simply filled in open gaps of the nanocellulose aerogel, the expected density would be 0.096 g/cm³. Therefore, the incorporation of the magnetic powders must induce the cellulose to bind more closely together, resulting in a denser matrix than the pure aerogels. The density measurements are also in line with magnetic aerogels formed in the methods using the prefabricated ferromagnetic nanoparticles noted previously.^16,17 The combination of visual evidence and density measurements indicates that the as-bought powder contained particles too large to be readily dispersed using ultrasonic agitation alone, but fine powders could be used to have a good effect.

Solids, films, and aerogels formed from the AlNiCo/nanocellulose suspensions were all magnetic. It was also a simple matter to flatten an aerogel to form a magnetic paper which retained its shape. Figure 2 shows hysteresis curves for an aerogel formed using preground AlNiCo powder. Hysteresis measurements were performed with a Quantum Design Physical Property Measurement System. Cycles were completed at two temperatures: 298 and 10 K. At each temperature, the magnetization was measured at fixed values of field. Each measurement was taken after the field was allowed to stabilize for 2 min. The field was then changed, and the measurement process repeated.

The magnetization of the sample is consistent with 5 g of pure AlNiCo, and there was no significant difference between as-bought and preground AlNiCo composites. This indicates that the incorporation into the cellulose aerogel did not have a significant impact on the magnetic properties of the system. The saturation magnetization increases with decreasing temperature, as expected for a standard ferromagnet. The coercivity of approximately 1300 Oe is consistent with higher coercivity grades of AlNiCo.

Figure 3 shows the SEM and EDX results taken from pure and AlNiCo containing, flattened aerogels. The microstructure was nearly the same across all samples, and it was not possible to definitively assign any given structure to an AlNiCo particle. The red shaded areas shown in the figure indicate areas with a significant amount of iron, indicating the presence of AlNiCo. As can be seen, incorporation of AlNiCo was not completely homogeneous at the microscopic scale. Particle size analysis of EDX measurements across many samples did not reveal any significant differences between samples made from preground or as-bought AlNiCo powder. However, the visual evidence and density measurements show that the aerogels derived from preground AlNiCo have better homogeneity and incorporation. This, and the lack of readily identifiable AlNiCo structures by EDX, indicates the majority of the AlNiCo incorporated into the aerogel lies outside the detection range of the EDX either finely dispersed submicron sized particles or, for the case of the as-bought powder, remains as millimeter scale particles. The lateral resolution of the EDX measurements is at best 1 μm, so particles much smaller than this would not be individually resolved.

We were able to create magnetic aerogels formed of a composite of nanocellulose and AlNiCo alloys using only physical processes. The aerogels were magnetic in nature and could readily be magnetized to have a permanent polarization. Optimal results were obtained from samples in which the AlNiCo powder was first ground with a mortar and pestle, reducing the particle size to facilitate dispersion and further breakdown using ultrasonic agitation. The physical properties, except for a darker color and increased density, of the AlNiCo bearing aerogels were extremely similar to pure nanocellulose aerogels derived in the same fashion.

The most significant challenge for this technique was an agglomeration of the AlNiCo into larger clumps during the sonication process. If the AlNiCo powder had a net magnetization, for example, from having been exposed to a permanent magnet before the sonication process, the powder would remain clumped together and be difficult or impossible to disperse to form a homogeneous suspension. It was also important to take into account wear of the ultrasonic probe as pits in the probe formed by the cavitation process acted as sites within which the AlNiCo would cluster and hence not incorporate well into the suspension. However, these technical issues could be readily addressed by appropriately demagnetizing the AlNiCo powder, using fresh sonic probe tips and minimizing overall sonication times to limit the effects of wear, and occasionally swirling the beaker during breaks in the sonication process after the cellulose is added.

The simplicity of this technique, in which a dispersion of the AlNiCo is first created, then homogeneously incorporated into a nanocellulose suspension through further sonication, makes it easy to extend to the creation of other types of nanocellulose composite materials. The process is free of any chemicals aside from cellulose, water, and the material to be added to form a composite, making it environmentally safe and limiting possible contamination vectors. Most importantly, this synthesis route shows that industrial grade powders can be readily incorporated into nanocellulose aerogels without needing to convert the materials into the nanoparticle form. Furthermore, as sonication is a readily accessible technique that can be incorporated into batch processing, it is possible to utilize this process to create large amounts of composite materials in a feasible manner.

T.E.K. and P.M.S. acknowledge funding from NSF Grant No. DMR-1206530. T.E.K. additionally acknowledges funding from NSF Grant No. DMR-1410496. D.B. acknowledges funding from the federal work study program.