鋁/氧化鋁/釔穿隧接點之二雜質近藤效應

Abstract

我們量測五個不同室溫電阻值的鋁/氧化鋁/釔穿隧接點的微分電導G(V,T)行為，量測溫度範圍約35-2.2 K，我們只專注於討論其中三個電阻值較小的穿隧接點(樣品編號A、B與C)。當溫度低於約30 K後，零偏壓電導G(0,T)不再反映出絕緣性行為，取而代之地表現出反常行為：當溫度~30-4 K，G(0,T) 隨溫度降低而上升，隨後當溫度低於~4 K， 轉變為隨溫度降低而下降， 對溫度為一非單調行為。 其中，當溫度~23-11 K (高溫區域)，G(0,T)為-logT的依存，且G(V,T)符合Appelbaum在弱耦合區域近藤效應理論的描述。當溫度降低至~17 K，其對溫度的依存開始過渡至-T^(1/2)，在溫度~17-8 K (中溫區域)G(V,T)會遵守一類似二通道近藤效應的scaling law，我們將此歸因於近藤效應與雜質間反鐵磁耦合之競爭所導致的現象，即我們的樣品反映出二雜質近藤模型的物理架構，其中的雜質為自旋1/2的釔原子(4d1)。我們認為可能是在樣品製程中，少量的釔原子座落或是輕微擴散入氧化鋁/釔介面，導致在介面存在侷域自旋1/2的磁矩。當溫度低於~4 K， 隨溫度降低而下降，此現象對應到二雜質近藤模型，我們的穿隧接點的基態是落在侷域自旋單重相，為單重態。 另外，我們在~0.25 K外加磁場1-4 T量測穿隧接點之Zeeman分裂行為。由Zeeman分裂隨外加磁場的變化，初步分析的結果顯示其基態為三重態。此結果與我們之前的結論不相符。我們到目前為止仍不了解如何解釋此現象，這部分還需再做更進一步的分析與討論。

We have measured the differential conductance G(V,T) in five Al/AlOx/Y tunnel junctions from 35 to 2.2 K, and all of them possess different room temperature resistance. We focus on three of them with smaller resistance (sample# A, B, and C). As the temperature below ~30 K, the zero bias conductance G(0,T) behavior crosses over from an insulating to the anomalies: G(0,T) increases with the temperature decreasing between ~30 and 4 K, and G(0,T) decreases as the temperature decreases lower then ~4 K. G(0,T) presents a non-monotonic behavior with the temperature. Between ~23 and 11 K (high temperature regime), G(0,T) is -logT dependence, and G(V,T) can be described by the Appelbaum’s weak coupling Kondo effect theory. As the temperature decreases to ~17 K, G(0,T) starts crossover to -T^(1/2) dependence, and G(V,T) obeys a scaling law between ~17 and 8 K (intermediate temperature regime), which is similar to the two-channel Kondo scaling. We attribute this behavior to a competition between Kondo effect and interimpurity antiferromagnetic coupling, i.e. the two-impurity Kondo physics is observed in our junction system, and the impurities are the spin-1/2 Y atoms (4d1). We think that a few Y atoms may situate at or slightly diffuse into AlOx/Y interface during fabrication process, leading to some localized spin-1/2 moments present at the interface. As the temperature decreases lower then ~4 K, G(0,T) decreases with decreasing temperature. According to the two-impurity Kondo model, this implies that the ground state of our system is the local spin singlet phase, it is an interimpurity singlet state. Furthermore, we have measured Zeeman splitting in the magnetic field between 1 and 4 T at ~0.25 K. The behavior of Zeeman splitting with variable magnetic field seems to imply that the ground state is triplet. This result is not consistent with the previous conclusion. We still do not understand how to explain it thus far, more analysis and discussion is needed.