11.03.2014 Views

Itinerant Spin Dynamics in Structures of ... - Jacobs University

Itinerant Spin Dynamics in Structures of ... - Jacobs University

Itinerant Spin Dynamics in Structures of ... - Jacobs University

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

List <strong>of</strong> Figures 111<br />

3.3 Exemplification <strong>of</strong> the second term <strong>in</strong> Eq.(3.4): Interference <strong>of</strong> electrons<br />

travel<strong>in</strong>g <strong>in</strong> the opposite direction along the same path causes an enhanced<br />

back-scatter<strong>in</strong>g, the WL effect. (a) Closed electron paths enclose a magnetic<br />

fluxfromanexternalmagneticfield, <strong>in</strong>dicatedastheredarrow, break<strong>in</strong>gtime<br />

reversal symmetry, break<strong>in</strong>g constructive <strong>in</strong>terference. (b) The entanglement<br />

<strong>of</strong> sp<strong>in</strong> and charge by SO <strong>in</strong>teraction causes the sp<strong>in</strong> to precess <strong>in</strong>between<br />

two scatterers around an axis which changes with the momentum vector <strong>of</strong><br />

the it<strong>in</strong>erant electron. This effective field can cause WAL. . . . . . . . . . . 30<br />

3.4 2D spectrum <strong>of</strong> H c , K i = Q i /Q SO . The physical mean<strong>in</strong>g <strong>of</strong> the gaps <strong>in</strong> the<br />

triplet modes is more comprehensible if H c is related to sp<strong>in</strong> diffusion, where<br />

the gaps appear as sp<strong>in</strong> relaxation rates. . . . . . . . . . . . . . . . . . . . . 38<br />

3.5 Persistent sp<strong>in</strong> helix solution <strong>of</strong> the sp<strong>in</strong>-diffusion equation for equal magnitude<br />

<strong>of</strong> l<strong>in</strong>ear Rashba and l<strong>in</strong>ear Dresselhaus coupl<strong>in</strong>g, Eq.(3.68). . . . . . 42<br />

3.6 Longpersist<strong>in</strong>gsp<strong>in</strong>helixsolution<strong>of</strong>thesp<strong>in</strong>-diffusionequation<strong>in</strong>aquantum<br />

wirewhosewidthW issmaller thanthesp<strong>in</strong>precessionlengthL SO forvary<strong>in</strong>g<br />

ratio <strong>of</strong> l<strong>in</strong>ear Rashba α 2 = αs<strong>in</strong>ϕ and l<strong>in</strong>ear Dresselhaus coupl<strong>in</strong>g, α 1 =<br />

αcosϕ, Eq.(3.82), for fixed α and L SO = π/m e α. . . . . . . . . . . . . . . . 46<br />

3.7 Dispersion <strong>of</strong> the triplet Cooperon modes for different dimensionless wire<br />

units Q SO W: (a) Q SO W = 2, (b) Q SO W = 8, (c) Q SO W = 12, (d) Q SO W =<br />

30, plotted as function <strong>of</strong> K x = Q x /Q SO . For Q SO W ≫ 3, E {t0,0} and E {t−,0}<br />

evolve <strong>in</strong>to degenerate branches for large K x . (For Q SO W = 30, not all<br />

high-energy branches are shown.) . . . . . . . . . . . . . . . . . . . . . . . . 48<br />

3.8 Probability density <strong>of</strong> the Cooperon eigenmodes <strong>in</strong> the wire for Q SO W/π =<br />

30. (a) 3D plot, (b) density plot for one <strong>of</strong> the two lowest branches, show<strong>in</strong>g<br />

their edge mode character. (c) 3D plot and (d) density plot <strong>of</strong> the density <strong>of</strong><br />

the third lowest mode, which shows bulk character. . . . . . . . . . . . . . . 49<br />

3.9 Absolute m<strong>in</strong>ima <strong>of</strong> the lowest eigenmodes E {t0,0} , E {t−,0} , and E {t+,0} plotted<br />

as function <strong>of</strong> Q SO W/π = 2W/L SO . We note that the m<strong>in</strong>imum <strong>of</strong> E {t−,0}<br />

is located at ±K x ≠ 0. For comparison, the solution <strong>of</strong> the zero-mode approximation<br />

E t0 is shown. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50<br />

3.10 Lowest eigenvalues at K x = 0 plotted aga<strong>in</strong>st Q SO W/π. For comparison, the<br />

global m<strong>in</strong>imum <strong>of</strong> the Cooperon spectrum for Q SO W 9 is plotted, F 3 .<br />

Curves F 1 [n] are given by 7/16+ ( (n/(Q SO W/π)) √ 15/4 ) 2<br />

, n ∈ N. F2 shows<br />

the energy m<strong>in</strong>imum <strong>of</strong> the 2D case, F 2 ≡ F 1 [n = 0]. Vertical dotted l<strong>in</strong>es<br />

<strong>in</strong>dicate the widths at which the lowest two branches degenerate at K x = 0.<br />

They aregiven by n/( √ 15/4); consider that the wave vector for them<strong>in</strong>imum<br />

<strong>of</strong> the E T− mode is ( √ 15/4)Q SO . . . . . . . . . . . . . . . . . . . . . . . . . 51<br />

3.11 The quantum conductivity correction <strong>in</strong> units <strong>of</strong> 2e 2 /2π as function <strong>of</strong> magnetic<br />

field B (scaled with bulk relaxation field H s ), and the wire width W<br />

scaled with 1/Q SO for pure Rashba coupl<strong>in</strong>g and cut<strong>of</strong>fs 1/D e Q 2 τ SO ϕ = 0.08,<br />

1/D e Q 2 SOτ = 4: Comparison <strong>of</strong> thezero-mode calculation (grid without shad<strong>in</strong>g)<br />

to the exact diagonalization where the lowest 21 triplet branches and<br />

seven s<strong>in</strong>glet branches were taken <strong>in</strong>to account. . . . . . . . . . . . . . . . . 52

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!