Introduction

The objective of this work is to prepare ferromagnetic 3d metal nanoparticles and investigate their magnetic properties, especially ultrafast demagnetization dynamics.

Ultrafast demagnetization has been observed in many systems; however, most of the observations have been made in bulk or thin films and few have been reported on ferromagnetic 3d metal nanoparticles. Metal nanoparticles are easily oxidized in ambient atmospheres, resulting in deterioration of the ferromagnetic properties. This instability makes the observation in metal nanoparticles difficult. Ferromagnetic metal nanoparticles have shape- or size-dependent magnetism. Fabrication of stable metal nanoparticles is, therefore, required for better understanding of ferromagnetism.

In this work, I propose a novel method for fabricating Co nanorods with various size and Ag-Co matchsticks, both of which are embedded in an anatase TiO2 matrix, thereby being stabilized. These Co nanostructures are prepared utilizing self-assembly process by pulsed laser deposition (PLD).

Ag-Co matchsticks were prepared in order to investigate the effect of localized surface plasmon resonance (LSPR) on demagnetization process. Remarkably enough, Ag-Co nanomatchsticks grew nearly in parallel to the film normal. These oriented Ag-Co nanomatchsticks are advantageous when LSPR of different modes are studied. Although there are many reports on such hybrid matchstick-like structures, most of the structures were fabricated by solution-growth technique, so that they were randomly oriented.

Finally, time-resolved Faraday effect measurements were performed on these Co nanorods and Ag-Co nanomatchsticks to observe ultrafast demagnetization. Enhanced demagnetization was observed in Ag-Co nanomatchsticks. This enhanced demagnetization is probably due to LSPR of Ag.

Experimental

Epitaxial thin films of (001)-oriented anatase TiO2 containing Co nanorods (Co:TiO2) and those containing Ag-Co nanomatches (Agx,Co:TiO2) were prepared on LaSrAlO4 (LSAO) (001) single-crystalline substrates by pulsed laser deposition (PLD). A Kr-F excimer laser (wavelength=248nm) operated with a laser fluence of 2 J cm-2 pulse-1 and a repetition rate of 2 Hz was used for ablation.

To obtain the crystallized anatase TiO2 phase in a relatively low temperature range, I used a seed-layer of pure TiO2, as described below. Sintered pellets of pure TiO2 ,a mixture of TiO2 and CoO (molar ratio Ti:Co=95:5),and a mixture of TiO2, CoO, and Ag2O (molar ratio Ti:Co:Ag=95:5:x) were used as target materials for the anatase TiO2 seed layers, Co:TiO2 films, and Agx,Co:TiO2 films, respectively.

The crystallinity and crystallographic orientation of the obtained films were evaluated by X-ray diffraction (XRD). The surface topography was characterized by atomic force microscopy (AFM). The sizes and distribution of the Co nanorods were examined by transmission electron microscopy (TEM) with the aid of energy dispersive X-ray analysis (TEM-EDX), scanning TEM (STEM), and high-angle annular dark field STEM (HAADF-STEM). The crystal structure of the Co nanorods was determined from nanobeam electron diffraction patterns. Magnetic properties were measured by a superconducting quantum interference device (SQUID) magnetometer (MPMS XL; Quantum Design, San Diego, CA, USA) and a UV-vis magneto-optical spectrometer (BH-M800UV-KC-KF; Neoark Corp., Tokyo, Japan). To measure ultrafast demagnetization dynamics, time-resolved Faraday effect measurement using pump-probe technique was performed for these films. A Ti:sapphire laser operating at 100 kHz was used. The fundamental wavelength was 800nm and the pulse duration was 220 fs. The wavelength of probe pulses was set to 800nm and that of pump pulses, which were generated by frequency doubling, was 400nm. An external magnetic field of 9 kOe was applied perpendicular to the film surfaces. All measurements were performed at room temperature.

Preparation of Co:TiO2 nanocomposite films

It was found that the Co:TiO2 films directly deposited on LSAO substrate do not show good crystallinity particularly at low temperatures. To obtain high-crystallinity anatase TiO2 phase over a wider temperature range, I used a seed-layer of pure TiO2, as follows.

An anatase TiO2 seed layer was first deposited on an LSAO substrate at Ts=650 °C and PO2=5×10-3 Torr, and then a Co:TiO2 thin film was grown at different substrate temperatures in the range Ts=200-400 °C and at an oxygen pressure PO2=1.0×10-6 Torr. The thicknesses of the TiO2 seed layer and the Co:TiO2 film were set at 5nm and 28nm, respectively.

According to magnetization measurement, Co concentrations in Co:TiO2 are estimated to be 4.3%, 3.7%, and 3.3% for Ts=250 °C, 300 °C, and 350 °C, respectively. These values are close to the Co content of the Co:TiO2 target (5%), implying that in these films Co exists mostly as nanoparticles of fcc Co metal, that is, the Co nanoparticles do not undergo oxidation even though they are embedded in the oxide matrix.

In the films grown at Ts=200 °C and 400 °C, the calculated fcc-Co contents are as low as ~0.7%. At Ts=400 °C, it is likely that Co reacts with O or Ti to form nonmagnetic oxides such as CoO and CoTiO3, resulting in the decrease in magnetization.

TEM results for the Co:TiO2/TiO2 films grown at Ts=250 °C, 300 °C, and 350 °C are summarized in figure 1. Plane-view and cross-sectional TEM images in figures 1(a)-(c) exhibit well-defined cylindrical nanostructures with long-axes nearly perpendicular to the film planes. The nanostructures are imaged with brighter contrast, suggesting that they are rich in a heavier element, Co. Nanobeam electron diffraction measurements at the Co rich regions revealed that they are Co single crystals with an fcc structure. These results clearly indicate that fcc Co nanorods are embedded in the TiO2 matrix.

The average diameter and height of the Co nanorods estimated from the TEM images are 9nm and 10nm, 10nm and 22nm, and 16nm and 17nm for the films grown at Ts=250 °C, 300 °C, and 350 °C, respectively, yielding the average volume of individual Co nanorods as 640nm3,1700nm3, and 3400nm3, respectively. The systematic increase of the average volume with increasing Ts is consistent with a kinetic picture that Co atoms become more mobile at higher temperatures.

In this nanocomposite fabrication technique, it is essential to use an anatase TiO2 seed layer, which lowers the crystallization temperature of anatase TiO2. In a Ts region of 250-350 °C, both high crystallinity of anatase TiO2 matrix and moderate condensation of Co clusters cab be realized at the same time. In this Ts region, furthermore, the shape and size of fcc-Co nanoparticles embedded in anatase TiO2 matrix is controllable by Ts.

Preparation of AgCo:TiO2 nanocomposite films

The anatase TiO2 seed layers were prepared in the same manner as that for Co:TiO2. Then a Agx,Co:TiO2 thin film was grown at a substrate temperature Ts=300 °C and at an oxygen pressure PO2=1.0×10-6 Torr. The thicknesses of the anatase TiO2 seed layer and the Agx,Co:TiO2 film were set to 5nm and 30nm, 5nm and 38nm, and 6nm and 36nm for x=5, 10, 20, respectively.

XRD measurements confirmed the anatase TiO2 phase with (001) orientation for all x.

TEM results for the Ag20Co:TiO2 films are shown in Fig. 2. A plane-view STEM-EDX image in Fig. 2 exhibits Co nanoparticles with an almost regular size embedded in the TiO2 matrix. The average diameter of the Co nanoparticles is 5nm. The molar ratio of Ti:Co:Ag is calculated to be 95:3.6:14 using the image. From magnetization measurements, the Co molar ratio is estimated to be rTi : RCo=95:3.9. These values are close to that of the Ag20,Co:TiO2 target (95:5:20). This indicates that in the present Agx,Co:TiO2/TiO2 films, Co exists mostly as nanoparticles of fcc-Co metal.

When Ag was co-doped into Co:TiO2 films, interestingly, Ag and Co atoms were completely separated in the film-growing direction and characteristic Ag-Co "nano-matchstick" structures, each of which consists of Co match head and Ag matchstick, were formed in the films, though Ag and Co atoms were deposited simultaneously.

Magneto-optical property

Faraday ellipticity spectra and absorption spectra of the films are depicted in figure 3(a) and (b). The Faraday ellipticity spectra of the Co:TiO2/TiO2 films and the AgxCo:TiO2/TiO2 films have a similar shape. In the absorption spectra, however, absorption peaks around 450nm resulting from Ag LSPR are found in AgxCo:TiO2/TiO2 films and the intensity of the peaks increases with increasing x.

Magnetization dynamics induced by optical pulses

Time-resolved Faraday measurements were performed on the Co:TiO2/TiO2 films and the AgxCo:TiO2/TiO2 films. Figure 4 shows normalized time-dependent differential Faraday ellipticity (Δη/η, η denotes saturation values of Faraday ellipticity in Figure 3(a)) of these films measured under pump fluency of 0.06 mJ cm-2. Ultrafast demagnetization was observed. The amplitudes of the demagnetization (maximum value of the-η/η) are plotted in figure 5.

As shown in Figure 5(a), The amplitude of the Co:TiO2/TiO2 films (Ts=250, 300, and 350℃) and the Ag5Co:TiO2/TiO2 film exhibit similar Δη/η. For the AgxCo:TiO2/TiO2 (x=5, 10, 20) films, however, the amplitude of Δη/η increases with increasing x. (Figure 5(b))

The possible reasons for this are as follows: 1) different shape of the Co nanoparticles in the films; and 2) different Ag content in the films.

The Co:TiO2/TiO2 films (Ts=250, 300, and 350℃) and the Ag5Co:TiO2/TiO2 film contain Co nanoparticles in different sizes. Especially, the Ag5Co:TiO2/TiO2 film contains much smaller Co nanosperes whose size is nearly within the superparamagnetic limit. However, the Δη/η values of the Co:TiO2/TiO2 films (Ts=250, 300, and 350℃) and the Ag5Co:TiO2/TiO2 film exhibit similar behavior. This means that the shape of Co nanoparticles does not affect Δη/η so much.

Therefore, the difference in Δη/η is considered to be caused by difference in Ag content. As shown in figure 3(b), the intensity of Ag LSPR at wavelength of 400nm increases with increasing x. The Co nanoparticles absorb energy of pump pulses with wavelength of 400nm efficiently through the Ag LSPR. This results in enhanced demagnetization in the AgxCo:TiO2/TiO2 (x=10, 20) films. As shown in figure 3(b), the Ag5Co:TiO2/TiO2 film shows only small absorbance, compared to the Co:TiO2/TiO2 films (Ts=250, 300, and 350℃), at wavelength of 400nm. Accordingly, the magnitude of the demagnetization in the Ag5Co:TiO2/TiO2 film is as small as that in the Co:TiO2/TiO2 films (Ts=250, 300, and 350℃).

Summary

Nanocomposite thin films consisting of an anatase TiO2 matrix with fcc-Co nanorods (Co:TiO2/TiO2 film) and those of an anatase TiO2 matrix with Ag-Co nanomatches (Agx,Co:TiO2/TiO2 film) were successfully prepared. These films were prepared using self-assembly by pulsed laser deposition (PLD). The size of Co nanorods in Co:TiO2/TiO2 film is controlled by growth temperature. The Ag-Co nanomatches in Agx,Co:TiO2/TiO2 films were oriented and show localized surface plasmon resonance (LSPR) of Ag. The intensity of LSPR increases with increasing Ag content x. Ti has larger oxygen affinity than that of Co, therefore, an anatase TiO2 matrix protects Co nanoparticles from oxidation and chemically stable.

Time-resolved Faraday effect measurements were performed on these Co nanorods and Ag-Co nanomatchsticks to observe ultrafast demagnetization. Differential Faraday ellipticity values (Δη/η) of the Co:TiO2/TiO2 films and the AgxCo:TiO2 (x=5) film exhibit similar behavior. This means that the shape of Co nanoparticles does not affect Δη/η so much. However, enhanced demagnetization amplitude was found in the AgxCo:TiO2/TiO2 (x=10, 20) films. This phenomenon can be understood by ultrafast energy transfer from Ag LSPR to electron system or spin system of the AgxCo:TiO2/TiO2 films. The investigation also showed that demagnetization recovery process of the AgxCo:TiO2/TiO2 films depends on Ag content x.

Figure 1 Planar and Cross-sectional TEM images, nanobeam electron diffraction patterns, and the shapes of each Co nanoparticles of Co:TiO2/TiO2 films (from left to right) prepared at Ts=(a) 250℃, (b) 300℃, and (c) 350℃.

Figure 2 STEM-EDX images of Ag20Co:TiO2 film. A planar image of Co K (White) (top) and a cross-sectional image of Co K (Gray) and Ag L (White) (middle), and a structure consists of a Co sphere and an Ag wire (bottom).

Figure 3 (a) Faraday ellipticity spectra and (b) absorption spectraof Co:TiO2 films and AgxCo:TiO2 films.

Figure 4 Magnetization dynamics of the Co:TiO2 (300℃) film and the Ag20Co:TiO2.

Figure 5 Demagnetization amplitudes of (a) Co:TiO2/TiO2 films and (b) Agx,Co:TiO2/TiO2 films

強磁性ナノ粒子は、バルクや薄膜とは異なった物性を示すことが知られている。特に、磁気異方性エネルギーが粒径に依存し、また、表面とコアで異なる磁気異方性を持つ可能性があると言う点で特異的である。本論文が取り扱うのは、Coナノ粒子の形状、及びCoナノ粒子のAg-Coハイブリッドナノ粒子化が、超高速消磁過程に与える効果である。

本論文は以下の5章より構成されている。

第1章は序論であり、本論文の背景および目的が述べられている。本論文で取り扱う強磁性ナノ粒子の諸物性を、バルクおよび薄膜との対比を通じて概観している。また、様々な系における超高速消磁過程測定の結果に触れている。Coナノ粒子の形状が磁性に及ぼす効果を評価する為には、Coナノ粒子の酸化や凝集を防ぎ安定化することが必要である。Coナノ粒子を、その酸化を防ぐ役割を果たすアナターゼ型TiO2母物質中に単分散させる事は、Coナノ粒子の磁気特性を評価する上で非常に優れた方法であると指摘している。また、本研究で用いられる薄膜成長手法では、薄膜成長過程における相分離を利用しているため、Coナノ粒子を大気に曝すこと無くアナターゼ型TiO2母物質中に内包させうる点を強調している。この方法の優位性を、これまでの金属ナノ粒子の作製方法との対比を通じて論じている。

第2章は、LaAlSrO4(001)単結晶基板上に作製したCoナノロッド内包アナターゼ型TiO2薄膜(Co:TiO2)の成長に関して述べている。薄膜成長パラメータとして基板温度(Ts)に着目し、金属Coナノロッドがアナターゼ型TiO2母物質中に内包されるTsの範囲を明らかにしている。また、Tsにより、Coナノロッドの形状を制御できることを示している。さらに、最適条件下で得られたCoナノロッドはfcc-Co単結晶である事を確認している。

第3章は、LaAlSrO4(001)単結晶基板上に作製したAg-Coハイブリッドナノ粒子内包アナターゼ型TiO2薄膜(Ag,Co:TiO2)の成長に関して述べている。Ag,Co:TiO2系の場合、薄膜成長過程における相分離の結果、金属Coナノ球及び金属AgナノワイヤからなるAg-Coハイブリッドナノ粒子が、アナターゼ型TiO2母物質中に内包されることを明らかにしている。Ag-Coハイブリッドナノ粒子は母物質中で配向しており、これは、固相成長において配向したハイブリッド構造を得た初めての例であると指摘している。また、これらAg,Co:TiO2試料は、波長400nm近傍に局在表面プラズモン共鳴による吸収を持ち、その吸収強度はAg含有量の増加に伴って増大すると述べている。

第4章では、第2章及び第3章の方法で作製したCo:TiO2及びAg,Co:TiO2に対し、ポンププローブ法を用いた時間分解Faraday効果測定を行った結果について述べている。ポンプ光波長をAgの局在表面プラズモン共鳴波長に合わせて400nmに設定することで、高いシグナル強度を得ることに成功している。また、試料中のAg量が多い程、超高速消磁の大きさが増大する事を示し、Agの局在表面プラズモンが消磁の増幅に関与している事を明らかにしている。一方、Coナノ粒子の形状は、超高速消磁の大きさや緩和過程に大きな影響を与えない事を述べている。

第5章は結論と総括である。

以上のように、本論文は、安定なCoナノ粒子及びAg-Coハイブリッドナノ粒子の作製方法を提案するに留まらず、局在表面プラズモン共鳴を示す金属ナノ粒子と強磁性金属ナノ粒子からなるハイブリッドナノ構造における超高速消磁の結果を初めて報告したものである。また、プラズモンからスピンへの超高速エネルギー移動という新たな現象を見出している。これらの成果は理学の発展に大きく寄与する成果であり、博士(理学)に値する。なお本論文は複数の研究者との共同研究であるが、論文提出者が主体となって行ったものであり、論文提出者の寄与は十分であると判断する。

したがって、博士(理学)の学位を授与できると認める。