No abstract

Koide et al. Reply: In their Comment [1] on our Letter [2], Bucher et al. criticize our interpretation of magnetic circular x-ray dichroism (MCXD) data in terms of isolated ferromagnetic clusters with a blocking temperature above 300 K and question our proposal for roomtemperature nanoscale magnetic bits. We admit their doubts to be natural but disagree with their central argument based mainly on their paper [3].Bucher et al. claim that the room-temperature transition from the superparamagnetic(SPM)-toferromagnetic(FM) state at a diameter of about 7 nm with increasing cluster size is due to a formation of large clusters by coalescence [1]. We first point out selfcontradiction in their argument in the Comment [1] and Ref. [3]. Namely, their beautiful STM pictures showed no topographic difference between clusters with coverages of 1.4 and 1.6 ML [3]. More importantly, those clusters are separated by small gaps even along the shortest direction, despite a drastic change in Kerr-effect signals across a 1.5-ML coverage at 300 K [3]. A serious assumption was made in [3] that the gaps between clusters are so small that they are coupled magnetically along rows. This magnetic coupling between isolated clusters is the very point we claimed for the origin of a magnetic transition at a coverage of 1:1 ML in [2]. Such abrupt magnetic transitions are difficult to explain by percolation but can be accounted for by magnetic interaction between separated clusters, as we claimed in [2]. This is because coalescence proceeds progressively, whereas a magnetic order-disorder transition can occur abruptly due to the essentially short-range nature of magnetic interaction.The morphology of our clusters [2,4,5] is different from that of Bucher et al. [1,3]. This is due to a difference in Au(111) surface preparation. Their clusters are more uniform in size than ours, while the distance between clusters is more isotropic in our clusters than in theirs. Our clusters had L 12:5=2 nm with d 6:5 nm [4,5], sharply contrasting with L 13 nm in [1,3], where d and L are the nearest distance and the distance between adjacent rows, respectively [3]. As shown in Fig. 1, our clusters maintain particle nature without forming uniform films even at a 2-ML coverage, owing to the smaller difference of anisotropic spacings. This is essentially different from the result of a Monte Carlo simulation in [3].Based on the results of [6], Bucher et al. estimated the blocking temperature of Co columns with 4.2 nm in diameter and 8 nm in height to be 300 K and claim that their magnetic moments cannot be reversed independently [1]. We agree with their estimate and argument, since the 8-nm height is 20 times as high as 2 ML 0:4 nm, which results in 400 times stronger dipoledipole interaction. Thus, nonreversibility of such magnetic moments is quite natural but clusters with such a small diameter and a much larger height are not our concern. We point out the possibility that ferromagnetic clusters with a diameter of 7:2 nm just above the SPMto-FM transition region could ...

The spin, in-plane and out-of-plane orbital and magnetic dipole moments of almost purely interfacial Co atoms were directly determined for Au/2-monolayer Co nanoclusters/Au(111) by angle-dependent magnetic circular x-ray dichroism (MCXD) measurements. The field- and temperature-dependent MCXD evidences a ferromagnetic(FM)-to-superparamagnetic phase transition in single-domain clusters with decreasing size. The interfacial moments are remarkably enhanced as compared with bulk values, verifying theoretical predictions. The FM clusters show strong perpendicular magnetic anisotropy, providing promise of applications for nanoscale magnetic bits.

Soft x-ray magnetic circular dichroism (XMCD) measurements at the Co L 2,3 edges of Co doped rutile TiO 2 at room temperature have revealed clear multiplet features characteristic of ferromagnetic Co 2+ ions coordinated by O 2− ions, being in sharp contrast to the featureless XMCD spectrum of Co metal or metallic clusters. The absorption and XMCD spectra agree well with a full atomicmultiplet calculation for the Co 2+ high-spin state in the D 2h -symmetry crystal field at the Ti site in rutile TiO 2 . The results indicate that the ferromagnetism arises from the Co 2+ ions substituting the Ti 4+ ions.

No abstract