Monodisperse face-centred tetragonal (fct) FePt nanoparticles with high magnetic anisotropy and, therefore, high coercivity have been prepared via a new heat treatment route. The as-synthesized face-centred cubic FePt nanoparticles were mixed with salt powders and annealed at 700˚C. The salts were then removed from the particles by washing the samples in water. Monodisperse fct FePt particles were recovered with the particle size and shape being retained. Coercivity of the isolated particles up to 30 kOe at room temperature has been obtained.

To transfer face-centered-cubic ͑fcc͒ FePt nanoparticles to the face-centered-tetragonal ͑fct͒ phase with high magnetic anisotropy, heat treatments are necessary. The heat treatments lead to agglomeration and sintering of the nanoparticles. To prevent the particles from sintering, salts as the separating media ͑matrix͒ have been used for annealing the nanoparticles in our experiments. The fcc nanoparticles produced by chemical synthesis were mixed with NaCl powders. The mixture was then annealed in forming gas ͑93% H 2 +7%Ar͒ in different conditions to complete the fcc to fct phase transition. After the annealing, the salt was washed out by water and monodisperse fct FePt nanoparticles were obtained. Detailed studies on the effect of the NaCl-to-FePt weight ratios ͑from 1:1 to 400:1͒ have been performed. It was found that a suitable NaCl-to-FePt ratio is the key to obtain monodisperse fct FePt nanoparticles. A higher NaCl-to-FePt ratio is needed for larger particles when the annealing conditions are the same. Increased annealing temperature and time should be accompanied by a higher NaCl-to-FePt ratio. Magnetic measurements show very high coercivity ͑up to 30 kOe͒ of the monodispersed fct nanoparticles by the salt-matrix annealing. 1 The chemically synthesized FePt nanoparticles, however, are of face-centered-cubic ͑fcc͒ phase without magnetic anisotropy. To transfer FePt nanoparticles from fcc phase to face-centered-tetragonal ͑fct͒ phase, heat treatments above 600°C are necessary, which undesirably lead to sintering of these nanoparticles.Since 2000, great efforts have been made to produce monodisperse fct FePt nanoparticles 2-8 driven by potential applications of the magnetically anisotropic nanoparticles in high-density recording media and high-performance nanocomposite magnets. Recently, we obtained monodisperse fct FePt nanoparticles with retained size and shape by using salts as the annealing separating media. 9 The salts can be completely removed after the annealing just by washing the samples in water. High coercivity up to 30 kOe of the fct particles has been obtained. In this paper we report detailed results in controlling the particle morphology and properties by adjusting the salt-to-FePt particle ratio. EXPERIMENTThe fcc FePt nanoparticles with size of 4, 8, and 15 nm were synthesized by chemical solution methods.1,10-13 Sodium chloride ͑NaCl͒ was selected as a separating media in this investigation due to its chemical stability and high solubility in water. NaCl was first ball milled for 24 h to reduce the particle size. The ball-milled NaCl powder was then dispersed in hexane and mixed with hexane dispersion of assynthesized fcc FePt nanoparticles. The mixture was stirred until all the solvent evaporates. Then the mixture was annealed in forming gas ͑93% H 2 +7%Ar͒ in different conditions to complete the fcc to fct transition. The annealed powders were washed in de-ionized water and centrifuged for several times to remove all the NaCl.9 Different NaCl-toFePt weight ratios from 1:1 to 400:1 were t...

FePt nanoparticles have great application potential in advanced magnetic materials such as ultrahigh-density recording media and high-performance permanent magnets. Size and chemical-ordering effects and their influence on the magnetic properties of the title nanoparticles are characterized by TEM, XRD, and magnetic measurements. It is shown that the long-range chemical ordering parameter, Curie temperature, and saturation magnetization drop significantly upon decreasing the particle size from 15 to 4 nm. -(RONG, C.-B.; LI, D.; NANDWANA, V.; POUDYAL, N.; DING, Y.; WANG, Z. L.; ZENG, H.; LIU*, J. P.; Adv.

Chemically ordered FePt nanoparticles were obtained by high temperature annealing a mixture of FePt particles with NaCl. After the NaCl was removed with de-ionized water, the transformed FePt nanoparticles were redispersed in cyclohexanone. X-ray diffraction patterns clearly show the L1 0 phase. Scherrer analysis indicates that the average particle size is about 8 nm, which is close to the transmission electron microscopy ͑TEM͒ statistical results. The coercivity ranges from 16 kOe to more than 34 kOe from room temperature down to 10 K. High resolution TEM images reveal that most of the FePt particles were fully transformed into the L1 0 phase, except for a small fraction of particles which were partially chemically ordered. Nano-energy dispersive spectroscopy measurements on the individual particles show that the composition of the fully transformed particles is close to 50/ 50, while the composition of the partially transformed particles is far from equiatomic. TEM images and electron diffraction patterns indicate c-axis alignment for a monolayer of L1 0 FePt particles formed by drying a dilute dispersion on copper grids under a magnetic field. For thick samples dried under a magnetic field, the degree of easy axis alignment is not as high as predicted due to strong interactions between particles.

In precision agriculture, the nitrogen level is significantly important for establishing phenotype, quality and yield of crops. It cannot be achieved in the future without appropriate nitrogen fertilizer application. Moreover, a convenient and real-time advance technology for nitrogen nutrition diagnosis of crops is a prerequisite for an efficient and reasonable nitrogen-fertilizer management system. With the development of research on plant phenotype and artificial intelligence technology in agriculture, deep learning has demonstrated a great potential in agriculture for recognizing nondestructive nitrogen nutrition diagnosis in plants by automation and high throughput at a low cost. To build a nitrogen nutrient-diagnosis model, muskmelons were cultivated under different nitrogen levels in a greenhouse. The digital images of canopy leaves and the environmental factors (light and temperature) during the growth period of muskmelons were tracked and analyzed. The nitrogen concentrations of the plants were measured, we successfully constructed and trained machine-learning- and deep-learning models based on the traditional backpropagation neural network (BPNN), the emerging convolution neural network (CNN), the deep convolution neural network (DCNN) and the long short-term memory (LSTM) for the nitrogen nutrition diagnosis of muskmelon. The adjusted determination coefficient (R2) and mean square error (MSE) between the predicted values and measured values of nitrogen concentration were adopted to evaluate the models’ accuracy. The values were R2 = 0.567 and MSE = 0.429 for BPNN model; R2 = 0.376 and MSE = 0.628 for CNN model; R2 = 0.686 and MSE = 0.355 for deep convolution neural network (DCNN) model; and R2 = 0.904 and MSE = 0.123 for the hybrid model DCNN–LSTM. Therefore, DCNN–LSTM shows the highest accuracy in predicting the nitrogen content of muskmelon. Our findings highlight a base for achieving a convenient, precise and intelligent diagnosis of nitrogen nutrition in muskmelon.