In the past few years, magnetic nanostructures have aroused considerable interests due to their exceptional properties and potential applications in catalysis, sensors, high-density magnetic records, bio-imaging, and many other domains 123456. Because most physical and chemical properties of these nanoparticles depend on their sizes and shapes, the shape control of nanostructures has become a new and interesting research field 789. In particular, one-dimensional (1D) magnetic nanostructures have become one of the focused points 10111213.

Recently, magnetic field-assisted hydrothermal process has been proven to be an efficient way for directing the growth of 1D magnetic material 1415161718. However, previous investigations indicated that the surfactants such as polyvinyl pyrrolidone, hexadecylamine, and cetyltrimethyl ammonium bromide were still necessary, and the diameters of the fabricated nanochains were usually larger than 100 nm 15181920. Template-free magnetic field-induced synthesis of well-aligned Ni nanochains with smaller diameters has rarely been reported 14.

Hence, in this work, we proposed a magnetic field-induced solution-phase synthesis and self-assembly approach of aligned Ni nanochains using environment-friendly reagents. The growth mechanism and the magnetic anisotropy of the Ni nanochains have been discussed in terms of the applied magnetic field.

Detailed results of characterization for sample 1 are presented as follows. Figure 2a, 2b shows typical morphologies of the as-prepared samples obtained without and with magnetic field of 5 kOe at 150°C, respectively. Curly interlaced net-like chain structures synthesized without magnetic field are vivid in Fig. 2a, with diameters of 35 ± 5 nm and lengths of up to 500 nm. Figure 2b shows the sample (sample 1) synthesized under an external 5-kOe magnetic field. The nanochains are axially aligned and relatively uniform in size (33 ± 3 nm in diameter, and 1–2 μm in length). The morphologies are obviously different without and with the presence of applied field, as revealed by Fig. 2a, 2b, attributed to the field-driven alignment effect. For the axially aligned sample shown in Fig. 2b, the random orientation of the freshly synthesized nanochains is mainly caused by the stirring and rinsing processes in the experiment. These axially aligned Ni nanochains can be orderly reoriented by drying a droplet of sample solution, which is prepared by dispersing the nanochains in ethanol, over the silicon substrate in the presence of an applied field (20 kOe), as shown in Fig. 2c. This reveals the dipolar properties of the whole nanochain along the chain axis. For each individual nanochain, the local direction of the nanoparticles deviates from the axial direction with a small angular distribution, which is expected to be of a Gaussian type 18. The crystal phase of all the samples was examined by using X-ray diffraction technique. A typical X-ray diffraction result of sample 1 is shown in Fig. 2d. The observed X-ray reflections can be well indexed with lattice parameters of face-centered cubic (f.c.c.) Ni (JCPDF No. 04-0850).

The chemical compositions of the as-synthesized sample were characterized by EDS. Figure 3a is for the EDS spectrum of sample 1. It reveals that the as-synthesized nanochains are metallic Ni, except that the carbon and copper peaks from the carbon film and copper grid. XAS investigation further supports the metallic nature of the sample. The spectra have been normalized at far end somewhere without subtraction of the step function 21, as shown in Fig. 3b. The L_{2, 3}-edge XAS spectrum of the nanochains is consistent with those of pure metallic Ni foil except that the L_{2, 3} peak intensities in our experiment are ~30% stronger. Hence, the nanochains have more 3d holes than the metallic Ni foils, which is attributable to the small size effect 22.

To characterize the microstructure, we carried out TEM and HRTEM investigations. Figure 4a shows a typical bright-field TEM image of a single nanochain (sample 1) corresponding to the nanochains shown in Fig. 2b. The single particle and the joint part of two adjacent nanoparticles, which are marked with frames “I” and “II” in Fig. 4a, are further investigated by HRTEM and are shown in Fig. 4b, 4c, respectively. Lattice fringes throughout the entire HRTEM images reveal the crystalline nature of the nanoparticle. The interplanar spacing in Fig. 4b, 4c is measured to be 0.203 and 0.176 nm, which agree well with those between {111} and {200} planes of Ni crystal, respectively. Detailed investigations reveal that each individual nanoparticle essentially consists of several single-crystalline grains with a typical size of 10–20 nm. The observed size of the grain agrees well with the estimated value about 14 nm from the XRD data using Scherrer’s equation. The selected area electron diffraction (SAED) pattern of the as-prepared nanochains (inset of Fig. 4b) confirms that the synthesized nanomaterial is metallic Ni with f.c.c. structure.

As illustrated in Fig. 5, a three-step model is proposed to explain the growth mechanism of the well-aligned Ni nanochains. In the first step, the primary Ni nanocrystallites nucleate and grow. Then, they aggregate to form nanoparticles, the size of which depends on the reaction temperature, hence is dictated mainly by the thermal agitation energy. Finally, these nanoparticles reach the size of thermal equilibrium and are self-assembled into nanochains by the alignment effect in the applied magnetic field. In this formation process, the competition between the magnetic field energy and the thermal agitation energy of the nanoparticles plays an important role. Without the magnetic field, the dendrite structure is observed, as shown in Fig. 2a. The thermal agitation energy dominates the thermodynamic driving forces and results in a structure of random alignment 23.

The in-plane field-dependent magnetization of the samples was measured by a SQUID magnetometer, with the applied field parallel and perpendicular to the direction of the aligned nanochains on the silicon substrate. The hysteresis loops measured at 5 K are plotted in Fig. 6, for (a) sample 1 with diameter of 33 ± 3 nm, (b) sample 2 with diameter of 78 ± 5 nm, and (c) sample 3 with diameter of 120 ± 12 nm. The diamagnetic background of the Si-substrate, with the susceptibility χ = −1.6 × 10⁻⁸ emu/Oe, as shown in the upper left inset of Fig. 6a, has been subtracted. There is obvious difference between the two hysteresis curves, i.e., the tilting of the open loops, attributed to the shape anisotropy 2425. To estimate the anisotropy by the expression, K = μ₀M_SH_A/2 26, the saturation magnetization of the bulk value of Ni, M_S = 485 emu/cm³, is adopted, and the anisotropy field H_A ~ 1.5 kOe is determined in Fig. 6a, at which the curves of the parallel and perpendicular loops bifurcate. The anisotropy is then obtained as K ~ 3.6 × 10⁴ J/m³. It is larger than the bulk value of magneto-crystalline anisotropy, ~4.5 × 10³ J/m³, by one order of magnitude. Obviously, the shape anisotropy dominates the magneto-crystalline anisotropy in this case. The demagnetization factor of the shape anisotropy is determined as ∆N ~ 0.24 according to the expression K = ½ μ₀∆NM_S². For samples 2 and 3, the anisotropy fields H_A are determined as ~1.5 and 2.0 kOe in Fig. 6b, 6c, respectively. The corresponding demagnetization factors thus obtained are ∆N ~ 0.24 and 0.32. These values are slightly larger than that of 50-nm nickel nanochains in a powder collection 27, ∆N ~ 0.19, and is about the same as that of nickel nanowires 28, with the size of about 12–50 nm, ∆N = 0.23–0.36.

For the axially aligned nanochains dispersed over the Si-substrate as shown in Fig. 2c, an interesting feature is observed. The nanochains appear to bundle up, 5 or 6 in each bundle. This is an indication that there exists intra-chain dipolar interaction which is likely to be of antiferromagnetic ordering, as revealed by the work of Bliznyuk et al. 29. This is an interesting issue to study. However, it is difficult to show evidence for this property by the magnetization measurements alone in the present work.

A facile and effective field-induced method has been employed to prepare highly uniform, size-controllable, well-aligned Ni nanochains without any surface active agent. The formation can be explained by the interactions of magnetic dipoles in the presence of applied magnetic field. In-plane M(H) curves with the applied field parallel and perpendicular to the direction of the aligned nanochains indicate that the shape anisotropy is a significant factor for the magnetization reversal. The greatly enhanced magnetic anisotropy of the well-aligned magnetic nanochains implies their potential applications in the fields of magnetic records, magnetic switches, etc.