dedup-isc-ft-v107-score
float64
0.32
1
uid
stringlengths
32
32
text
stringlengths
0
32.5k
paper_id
stringlengths
1
14
original_image_filename
stringlengths
5
222
0.484794
00b0fe456182408dab9f94cfa89c4f88
Solid triangles show temperature dependencies of the out of plane quasiparticle conductivity σ _c (a) and in-plane quasiparticle conductivity σ_ab (b). Below T_c they extracted from the JFF experiment on BSCCO long stack #4 and above T_c they represent the dc normal-state conductivities of whiskers measured independently on samples from the same batch. Open circles correspond to the σ _c data from Ref.LatPRL99, obtained on small mesas in zero field, open squares correspond to 14.4 GHz microwave data for σ_ab from Ref. SigmaQBSCCO obtained on BSCCO epitaxial films. Solid lines in both plots are just guides to the eye. Inset in the upper figure shows the low-temperature part of σ_c(T) plotted versus T^2.
cond
mat0307450-sigmas1.jpg
0.422692
8a98f33ac25944478f5b914b3533bf54
The average spike rate for minute for the monitored potion of the network: 67 neurons. The vertical lines signal the borders of states with a given concentration of tetradotoxin. The numbers between two consecutive vertical lines denote the concentration of tetrodotoxin in nano moles, characterizing the corresponding state. The label REF indicates the network state chosen as the reference for this experiment. In our case the reference state is 30 minutes of native (no toxicant added) activity where the average spike rate for minute is fairly stable.
cond
mat0307451-fig1.jpg
0.474568
bf2b9c5d88bb454e8f30f3ef6f43c8c6
A sample of our method of data processing. This is the number of spikes per bin, during the first minute of activity of the neural network. During this period of time no tetrodotoxin is present in the neural network.
cond
mat0307451-fig2.jpg
0.422012
818d35319560462984f479b92a195e77
The CASSANDRA coefficient (triangles) C_t for λ = 10, as a function of time, compared to a properly rescaled average spike rate for minute (squares). Note that the index j of the prescription of Eq. (<ref>) is replaced by the corresponding time t, which can be thought of as continuous. The meaning of the vertical lines is the sams as that of Fig. 1.
cond
mat0307451-fig3.jpg
0.476754
d16daca3ca614a8ab9544a659e651fc6
Comparison of two different neural network activities. The bottom frame refers to the time region betwen 90 and 95 minutes with the concentration of 1 nanomole. The top frame refers to the time region between 135 and 140 minutes, with a concentration of 2 nanomoles.
cond
mat0307451-fig4.jpg
0.507333
d4fb2f1f0aa9487797efa73de70125ec
Fermi surfaces of Eq. (<ref>) for hole doping (dashed line, μ=-1.76, x=0.48) and electron doping (solid line, μ = -0.4, x=-0.15). The hot spots are shown by open and solid circles. The radius of the circles σ=0.1 represents the width of the interaction (<ref>) in the momentum space.
cond
mat0307454-fig1.jpg
0.498423
ca3cdca968af4326805de320d226bb9d
The pairing potential Δ at T=0 vs. the angle φ on the Fermi surface, shown by the dashed line for x=0.37 (μ=-1.6t_0), the solid line for x=-0.1 (μ=-0.6t_0), and the dotted line for x=-0.25 (μ=0). The main panel represents the d_x^2-y^2 state, and the upper inset the chiral p-wave state. The angle φ_ hs indicates position of the hot spot 1. The lower inset shows the phase θ of Δ(φ)=|Δ|e^iθ for the chiral p-wave pairing.
cond
mat0307454-fig2.jpg
0.488821
3a5e3bc51e7449fe9f1b9f6517464abf
Dependence of various quantities on doping x. Panel (a): the hot spot angle φ_ hs, panel (b): the transition temperature T_c, panel (c): the condensation energy F, and panel (d): the maximal gap Δ_ max. The solid and dashed lines correspond to the d_x^2-y^2 and to the chiral p-wave pairings.
cond
mat0307454-fig3.jpg
0.517091
b91e8d2cd94f463eaeb98e9d95d25e72
Solid lines: Superconducting phase diagram of electron-doped cuprates vs. doping x calculated on the basis of Fig. <ref>. The vertical dashed line is a guide for eye.
cond
mat0307454-fig4.jpg
0.392762
5faea6223790495ca820e010e65bd34f
Energy barrier as a function of j.
cond
mat0307457-EnergyBH.jpg
0.519463
b907e49f2008406aad480cb332054866
Relaxation rate (in logarithmic scale) of the two-spin system in the case of intermediate-to-high damping. The horizontal line is the Néel-Brown result with the barrier height taken twice that of a single spin.
cond
mat0307457-Gh01t01a07.jpg
0.423793
6ac36344810a4063a71b45f6ebb10880
Energyscape from Eq. (<ref>) (φ=0) for j=0.25, 0.8, 1.5 and h=0.1. The arrows indicate the switching paths.
cond
mat0307457-enscape32.jpg
0.49625
ac017bc85a1a424e9cfb58b1e7b49c3a
Left: Dependence h_z(a) at h_x=0 on the Stoner-Wohlfarth curve. Right: The Stoner-Wohlfarth curves at different temperatures, a≡ 2χ_/χ_⊥=0 (T=0), 0.3, 0.5, 2/3, 0.835, 1, 2.
cond
mat0307457-hz_astroid_tsw.jpg
0.395926
d51849af43bd47b4b57536afce1b8f49
Magnetic energy in the case of longitudinal field for some values of the reduced field h. Upon increasing the field the number of potential wells changes from two to one.
cond
mat0307457-sw.jpg
0.439828
c2aa692c85e34fed9edfd6ab7921d61d
Left: (numerical) hysteresis loops for different values of ψ increasing inwards: ψ =0,60^∘,85^∘,90^∘, for a 3^3 particle with uniaxial anisotropy and strong exchange coupling. Right: (numerical in squares and analytical in full line) SW astroid for the same particle.
cond
mat0307457-swt1.jpg
0.470369
225552910c434cbf854875157e3aaccf
E(J) at T_0=2.5 K, for 0.1 T ≤ B ≤ 0.5 T (B_c2 = 0.65 ± 0.03 T) increasing as indicated by the arrow, measured by sweeping the applied current up. The dashed line represents E_n=ρ_n J. Inset: Measured magnetoresistivity (solid line) and the LO ρ_f (open symbols) plotted using g_1 and g_2 as explained in the text.
cond
mat0307463-ffisubfig1.jpg
0.458298
686261f56fce433380c0d7ebc87ba449
E(J) at T_0 = 2.5 K (full lines) measured at an increasing applied current, for B=0.1 T (a) and B=0.4 T (b). The dashed lines are plots of Eq. <ref> with the appropriate choices of the parameters, as discussed in the text. Inset to (a): Extracted E_i(B) (circles), illustrating the validity of the LO theory with a critical velocity u_i=E_i/B independent of magnetic field.
cond
mat0307463-ffisubfig2.jpg
0.415799
cfccbb47ae204607aeddcde009cbe255
(a) E(J) at T_0=1.1 K (solid lines), for 0.6 T ≤ B ≤ 2.2 T (B_c2 = 3.0 ± 0.1 T) increasing as indicated by the arrow. The measurements correspond to an ascending applied current. The dashed line shows E_n = ρ_n J. The dotted lines are plots of Eq. <ref> for 0.6 T and 0.9 T. (b) E vs. b^* calculated from the measured E(J) and Eq. <ref> using α=3. The vertical scale is the same as in (a) and the arrow points again in the direction of increasing B.
cond
mat0307463-ffisubfig3.jpg
0.424122
6fdb37b9acbe43219dce4fd3512a216a
Scaling plot of the nonequilibrium reduced magnetic field b^* vs. 1 - E/E_c calculated from Eq. <ref> using α = 3, where E_c(B) is plotted against equilibrium (1 - b) in the inset. The error bars of E_c indicate how much E_c varies if the scaling is performed using α between 2 and 4.
cond
mat0307463-ffisubfig4.jpg
0.388455
e1084a782cf843ab8c704defca55aea4
Pictures of the rotating gas taken along the rotation axis after 18 ms time of flight. We indicate in each picture the stirring frequency during the second stirring phase (/2π=65.6 Hz). The real vertical size of each image is 306 μm.
cond
mat0307464-figure1.jpg
0.518354
c1ff43717fcb40898214c902a9dab9e8
Optical thickness of the atom cloud after time-of-flight for /2π=66 Hz. (a) Radial distribution in the xy plane of Fig. <ref>e. Continuous line: fit using the Thomas-Fermi distribution (<ref>). (b) Distribution along the z axis averaged over |x|<20 μm (imaging beam propagating along y).
cond
mat0307464-figure2.jpg
0.48075
5b74a562ad2d4efcb8b5c1e910e255a0
Effective rotation frequency as a function of the stirring frequency . : values deduced from the fit using the Thomas-Fermi distribution (<ref>). ∘: values obtained by measuring the vortex density in the TOF pictures. ▴: values obtained using surface wave spectroscopy. The bars indicate standard deviations.
cond
mat0307464-figure3.jpg
0.410177
cc4c1d9cc36f454e8d8b5e4f9cf05406
Spatial dependence of the magnetization in the whole system . Here γ_F/γ_S =0.5, γ̅_F=γ_F/ξ_0=0.1 (ξ_0=√(D_S/2T_c)), J/T_c=15 and d_F/ξ _0=1. Inset: Schematic view of the inverse proximity effect in a S/F system (for discussion see text).
cond
mat0307468-fig.jpg
0.528761
062d48ff71c641d5a7842b6c49da45c5
Temperature dependence of δ M_S/M_F0 (lower curves) and δ M_F/M_F0 (upper curves) for the following values of γ̅_F=γ _F/ξ_0 (ξ_0=√(D_S/2T_c)): γ̅_F=0.1 (solid line), γ̅_F=0.3 (dashed line), and γ̅ _F=0.5 (dot-dashed line). Here γ_F/γ_S =0.5, J/T_c=20, d_F/ξ _0=0.1 and λ=3 (see [18])
cond
mat0307468-mag_contributionsil.jpg
0.601623
053a7bce755e48d38ede595ae7ac77e0
Temperature dependence of the magnetic susceptibility of the randomly orientated polycrystalline samples of κ-(ET)_2Cu_2(CN)_3 and κ-(ET)_2Cu[N(CN)_2]Cl [9]. The core diamagnetic susceptibility is already subtracted. The solid and dotted line represent the result of the series expansion of triangular-lattice Heisenberg model using [6,6] and [7,7] Padé approximants, respectively, with J = 250 K. The low-temperature data of κ-(ET)_2Cu_2(CN)_3 below 30 K are expanded in the inset.
cond
mat0307483-fig_2.jpg
0.484598
21b9791a74174db2a00154cf38c6f2bb
(a) ^1H NMR absorption spectra for single crystals of κ-(ET)_2Cu_2(CN)_3 and κ-(ET)_2Cu[N(CN)_2]Cl [9] under the magnetic field perpendicular to the conducting planes.
cond
mat0307483-fig_3.jpg
0.445271
ba7145d29a4b430faf28c70a13ef4859
^1H nuclear spin-lattice relaxation rate T_1^-1 above 1 K for a single crystal (open circles) and a polycrystalline sample (closed circles) of κ-(ET)_2Cu_2(CN)_3 and a single crystal of κ-(ET)_2Cu[N(CN)_2]Cl (open diamonds) [9]. The inset shows the data down to 32 mK in logarithm scales.
cond
mat0307483-fig_4.jpg
0.416745
0b1e894388ce4849a82717dce5df5982
The MFPT(γ) as a function of barrier fluctuation rate γ for linear potentials with H_+=8, H_-=0 (upper panel), H_+=8, H_-=4 (middle panel) H_+=8, H_-=-8 (lower panel) are plotted along with asymptotic lines. The additive noise is of the Lévy–Smirnoff type (left panel) (α=0.5, β=1, μ=0) and of the Cauchy type (right panel) (α=1, β=0, μ=0), respectively. Different symbols correspond to varying intensity of fluctuations: σ=.5 (+), σ=1/√(2) (×) and σ=1 (∗). Numerical results have been obtained by use of Eq. (<ref>) with the time step dt=10^-5. Noises were generated according to Eqs. (<ref>) and (<ref>). In order to gain sufficient statistics averaging of Eq. (<ref>) over 10^3 realizations has been performed. Asymptotic lines have been evaluated for reference static potential profiles with the time step dt=10^-5 and averaged over 10^4 realizations. Error bars have been estimated by use of standard techniques for the MC data analysis <cit.>. They represent standard deviation from the mean and remain within the symbol size. Lines have been drawn to guide the eye.
cond
mat0307484-levyplot.jpg
0.472506
8939c336acda42e7b86d232085e0c421
A model potential studied in the paper. The barrier height fluctuates dichotomously between the values H_±. A particle starts its diffusive motion at a reflecting boundary x=0 and continues until the absorption at x=1.
cond
mat0307484-model.jpg
0.543369
a984afd288b74b45a8de0cf1fac3e53e
Example of the procedure for extending the DE method to the analysis of data represented by matrixes. The small black squares represent some of the many ways of locating a square of size 2 inside a matrix of dimension 9 × 9.
cond
mat0307485-fig1.jpg
0.380469
f6db05e6dce04dbfa89fd1bba33bf834
Results of the single color DE image analysis. The entropy S(t) as a function of time t in a logarithmic scale. For each image the results relative to the red, green and blue component of the image are plotted. The square indicates the results relative to the biotic target, the circles those relative to the matrix surrounding the biotic target, the diamonds those relative to the a-biotic tartget and the triangles those of the matrix surrounding the a-biotic target. Finally, the full and the dashed lines indicate a straight line, in logarithmic time scale, of slope 0.7 and 0.77, respectively.
cond
mat0307485-fig2.jpg
0.440215
09eb1979c6354aca9c3d68d57790e304
Results of the global DE image analysis. The entropy S(t) as a function of time t in a logarithmic scale. The empty squares and circles indicate the results relative to the biotic target and its surrounding matrix, while the black ones the results relative to the a-biotic target and its surrounding. Finally the full line and dashed lines indicate a straight line, in logarithmic time scale, of slope 2 and 2.2, respectively.
cond
mat0307485-fig3.jpg
0.46447
83fabc17beb44782ae3360c882e6b2d1
Left panel: Alloy configurations for various concentrations x≤ 3 (the x>3 cases are obtained by exchanging closed and open circles). For each x≤3 configuration, the value of the ground-state spin S_g is reported. Right panel: Energy difference (in units of 10^-4t ) between the lowest S≤2 eigenstates and the ground state as function of ϵ_B.
cond
mat0307486-Fig1.jpg
0.516403
60eb15cd6ec34842a73cdf63ee894865
Temperature dependence of the average f-susceptibility for different alloy concentrations. The inset shows the low-temperature behavior.
cond
mat0307486-Fig2.jpg
0.389157
e670610570d1439899e361d967b61a2b
A-atoms: Local Kondo Temperatures (in K) vs the local f-c spin correlation function, for different configurations and two different values of t. The closed circles refer to the x=0 case and the lines are a guide to the eye.
cond
mat0307486-Fig3.jpg
0.45041
89dfb33217db4aa1a98bebd6246c14c1
Left panel: Configuration-averaged local f-c (top) and n.n. f-f spin correlations (bottom) for the A atoms as function of t. The solid line refers to the homogenous x=0 case. Right panel: Zero-temperature t vs x phase diagram for the nanocluster. Black (gray) circles denote the Kondo (RKKY) regime. The white circles and the dashed contour delimit the FM region. The horizontal stripes denote the non-magnetic (NM), mixed valence (MV) and local moment (LM) behavior of the B-atoms.
cond
mat0307486-Fig4.jpg
0.50165
5018381c940549b1bbd7c23d5529d090
(a) States in antiferromagnetic superlattices (example N=6) with increasing field: F ferromagnetic; FM1/2 ferrimagnetic collinear - such phases may be energetically degenerate, but they own different (meta)-stability limits; SF spin-flop states; C asymmetric canted; AF antiferromagnetic). (b) Example of evolution of state with field H for Mills model in low anisotropy case: rotation angles θ_i (i odd) against easy-axis n || field H (for i even θ_i = - θ_N-i+1). Phases of type C and FM1/2 may occur only at intermediate and higher anisotropy.
cond
mat0307494-Fig1UKR.jpg
0.438744
321b3bdcfcc543aa8bec3cd08e62d750
(color) Example of evolution of magnetization (continuous lines and left scales) and lowest eigenvalue of stability matrices e_0 (dotted, right scales) for Mills model (N=16 and K=0.5). Black curves: magnetization of equilibrium states, color curves for various “canted” phases C1 … C4 and the (reentrant) spin-flop state “SF”. Arrows mark phase transitions. Inset gives full range of field H from antiferromagnetic (AF) to ferromagnetic (F) state (half-logarithmic plot for e_0(H)) - details are magnified in main figure.
cond
mat0307494-Fig3UKR.jpg
0.477948
c872b101f08b41d78270f02561423118
(a) Mass 31 depth profile of δ-doped samples grown at 250^∘C and room temperature, respectively, determined by SIMS using a 5.5 keV Cs^+ primary ion energy in an ATOMIKA system. The inset shows the high mass resolution CAMECA SIMS measurement. (b) Mass 31 depth profile of the δ-doped sample grown at 250^∘C using 2, 5.5, and 15 keV Cs^+ primary ion energies in an ATOMIKA system.
cond
mat0307495-Figure1_SIMS_small.jpg
0.485193
f09603e7786d466783c05b4f2f41328c
(a) Differentiated Auger electron spectroscopy spectrum of the clean Si surface and the surface after PH_3 saturation dosing and annealing, 5 ML Si growth at room temperature, and various annealing steps for 1 min each. (b) Phosphorus density determined from integrated Auger electron spectra normalized to the phosphorus density of the saturation dosed surface for Si encapsulation at 250^∘C and RT.
cond
mat0307495-Figure2_Auger_small.jpg
0.422964
963175d0606946a5b7b0f678c31531b0
Filled state STM images showing the clean Si(001) surface (a) and the surface after phosphorus incorporation (b)/Si overgrowth (c)/annealing at various temperatures (d) - (f) from two experiments: 5 ML Si growth at 255^∘C (left column) and room temperature (right column). STM images taken after various annealing steps show an increasing density of bright features at the surface which are Si-P heterodimers. Image size of individual STM images is 50 × 25 nm^2.
cond
mat0307495-Figure3_STM_small.jpg
0.431607
94f8be2fa5ff4c6aaf17088e823b36e8
Filled state STM images showing the Si(001) surface after Si overgrowth (a) and annealing at various temperatures (b) - (d) from two experiments: 5 ML (left column) and 10 ML Si growth at RT (right column). Image size of individual STM images is 50 × 25 nm^2.
cond
mat0307495-Figure4_STM_small.jpg
0.465454
227c9f13cd994f2890c8a787ba351901
Relative density of P atoms at the Si(001) surface after Si encapsulation and various annealing steps. The surface P density increases during annealing at temperatures up to ∼600^∘C (a) and decreases due to P desorption from the surface at annealing temperatures ≳ 650^∘C (b). The density relative to the initial coverage was determined from STM images. The lines are guide to the eye.
cond
mat0307495-Figure5_Pdensity_small.jpg
0.449725
6cc4eeccbedd4d8b8579b3e55484f2eb
Comparison of phosphorus segregation length data with experimental data from the literature. Note that all data were corrected for a growth rate R = 0.045 Å/s using the relation ∼ 1/√R <cit.>. Also note that experimental data of Friess and Nützel are from the same group and differ significantly from Hobart's results. The grey area denotes the range of values expected from literature results.
cond
mat0307495-Figure6_comparison_small.jpg
0.432042
7a38ef05a7d745469746e95ab2df8932
The overdamped deterministic motion ẋ(t) = - V'(x(t)) in the above shown potential induces a partitioning of the state space = ℝ_1 into the domains of attraction 𝒟_ of the three attractors =1,2,3 at the potential minima. Here the locations of the local potential maxima coincide with the boundaries between the domains of attraction.
cond
mat0307498-semifig1.jpg
0.426418
e1380076a59240e1bd1333d5893d71b8
The asymptotic exponential growth of the frequency independent ratio E/Ø is shown for the driving strength A=0.12. The straight line corresponds to a barrier height 0.97 × 10^-3.
cond
mat0307498-semifig3.jpg
0.512228
4f81faca839b4e43935b13359fa91127
The response of the average position on the driving force for the driving strength A=0.05 and driving frequency Ø = 10^-5 strongly depends on the magnitude of the noise. For β = 35 the noise is relatively large such that the average position closely follows its adiabatic value and therefore shows almost no hysteresis; for β = 48 it stays behind the force, and for β = 75 it is hardly influenced by the force. The hysteresis curves are traced in the mathematically positive sense.
cond
mat0307498-semifig4.jpg
0.466029
c14109cabb374f6a95cf94cb3fb97954
The overlap order parameter as a function of time for the SNA versus the GFA approach for α=0.1, T=0.1 for retrieval behaviour m_0 =0.4 and spin-glass behaviour m_0 =0.2. The thin line represents the stationary limit for retrieval.
cond
mat0307499-snagfa.jpg
0.428206
328268cfed6b47ca90e46aa6ab3b9c4c
Hole Andreev reflection vs two-electron emission: (a) When μ_S is used as the reference chemical potential both in S and N, a typical scattering process at the NS tunneling interface for eV≡μ_S-μ_N>0 is that of incoming holes with energies between μ_S and μ_S+eV that are most often normally reflected and only rarely Andreev reflected. (b) If, alternatively, μ_N is used to define quasiparticles in N, the many incoming or outgoing holes are viewed as a vacuum of quasiparticles. The outgoing electron generated in a rare Andreev reflection event appears now as a spontaneously emitted electron above μ_S. Such an event causes an outgoing hole state to be empty. This is now perceived as the emission of a second electron with energy between μ_N and μ_S. Tracking the spin and the momentum component parallel to the interface leads to the picture of two electrons emitted in a spin singlet state with opposite parallel momenta.
cond
mat0307500-2e-vs-Andreev.jpg
0.43229
4aa537a6ff1d4dbe89ffa29d18d860c4
Angular correlation profile (in arbitrary units) of the conditional probability distribution P(Ω|Ω_0) that, in a given tunneling event, an electrons goes into Ω if the other electron has gone into Ω_0. Here we plot P(Ω|Ω_0) as a function of Ω for fixed Ω_0≡(θ_0,ϕ_0)=(π/4,0). From top-left to bottom-right the radii are: R=3, 1, 0.5, 0.01λ_F. Observe that, as R increases, the angular dependence of the second electron tends to be the conjugate of the first one, i.e. the distribution becomes peaked around Ω=(π/4,π). Note also that, for small R, P(Ω|Ω_0) becomes ∝cos^2θ regardless of Ω_0.
cond
mat0307500-3D-2x2.jpg
0.491223
c79cc399a13745be82955d86b8eea862
Angular dependence of the normalized tunnel current ranging from U_0/E_F→∞ for the outer (cos^5θ) curve to U_0/E_F=1.1 for the inner one. Finite barriers have a width w=5λ_F. Observe how the angular distribution focalizes around the perpendicular direction as the barrier hight decreases.
cond
mat0307500-Dep-angular.jpg
0.462776
e6c3f55c6cc5444d9de53a91041b3076
Current density through a two-point interface stemming from non-locally entangled electron pairs, as a function of the distance between points. The dashed line corresponds to the current obtained using an energy-independent hopping approximation whereas in the solid line the correct momentum dependence has been taken into account.
cond
mat0307500-I2puntos.jpg
0.516266
8a17b743a86b42368ded129465ff6fe6
Radial dependence of the normalized NS Andreev tunnel current through a circular interface of radius R for different barrier heights. J(R)≡ I(R)/π R^2 and J_∞ is the current density in the thermodynamic limit. Finite barriers have a width w=5λ_F. Everywhere Δ/E_F≪ 1 is taken. Dots correspond to numerically exact results. Solid lines are computed with an approximation described in Appendix B which becomes exact for R/λ_F ≫ 1. The inset magnifies results for small R.
cond
mat0307500-IvsR.jpg
0.459638
038456d89a1a4ea9a4bf8cfd71bf8946
Schematic lateral view of the NS tunneling structures studied in this paper: (a) circular interface of arbitrary radius R and (b) interface made of two small holes at a distance r from each other. The rest of the NS interface is assumed to be opaque.
cond
mat0307500-SN-Structures.jpg
0.397087
d0704cf7967b4939a7741384d021015d
Dynamics of the x-component of the magnetisation
cond
mat0307503-1262_hollinger_fig2.jpg
0.465831
789a5d68bf3e4daaa291a8142c4a3d8b
Frequency-spectrum of the z-component of the magnetisation
cond
mat0307503-1262_hollinger_fig3.jpg
0.424402
d41a6720ae264df893ccf77e6d8fb9f2
Close-ups of the region where the reversal curves drop show the time-dependence of the effect for the sample of Fig.<ref>(a). Essentially, when there is more pausing before or during measurement, the magnetization drops farther. (a) The normal reversal curves are taken by decreasing the field in 0.1 second steps, then increasing and averaging for 0.13s at each field point. (b) Averaging for 3s at each field point gives avalanches more time to proceed and the magnetization drops further. (c) Equilibrating for 60s prior to taking data on the reversal curve eliminates the effect but puts the magnetization lower at a given field than (a) and (b). (d) As (b) but the field sweep time prior to reversal is also slowed to 3s.
cond
mat0307507-CP1.jpg
0.443425
f896d53b89ca4ccd9a0817a503092030
(a) Descending half of hysteresis loops with and without cooperative rules for two different temperatures, T=0.04 (thicker lines) and T=0.1 (thinner lines). The portion for increasing H is symmetric. The broad curves (solid lines) are the results when all the spins act independently; the sharp curves (dotted lines) show the effect of the nucleation barrier established by our cooperative rules. (b) A sequence of descending half hysteresis loops at different temperatures (T=0.03, 0.1, and 0.3 from left to right). Notice the overall sharpening at higher T and the much broader tail at lower T. (c) Calculated first-order reversal curves at T=0.1. Shown here are reversals at M=0.9, 0.5 and 0, but the curves essentially all overlap because in all three, even following reversal, magnetization continues to drop nearly to M=-1 before stabilizing. (d) For T=0.05, first-order reversal curves are more reasonable. Note that the magnetization does drop following reversal. We find that the amount of this drop decreases at lower temperatures or for more broad distributions of H_k.
cond
mat0307507-effect_temps.jpg
0.427319
e7437b8f812d4cfa8dfcce428be504ae
Major hysteresis loops (envelope) and first order reversal curves for two samples provided by B. Terris at IBM. Sample (a) shows sudden sharp drop in major hysteresis loop, while sample (b) shows a similar drop as the magnetization begins to fall, but then has a broad tail.
cond
mat0307507-forcs.jpg
0.418091
89716ac873094a67809187a810a15b8e
Left, a group of seven spins as shown, where each of the seven attempts (small arrows) to flip down in the same step, is the smallest group of spins that can stably flip down (right) in our rules. Thus our rules provide a nucleation barrier to formation of a domain of down spins near M=1, and similarly when near M=-1 for formation of a domain of up spins.
cond
mat0307507-rules.jpg
0.473381
0db199d7e93a4ef5aea4a3fd62a253b2
Descending half of the major hysteresis loops for our Co-Pt multilayer samples as measured at two different temperatures in a SQUID magnetometer. Curves with most negative coercivity are 5 K; the other are at 300K. The 5K curves also show the ascending half of the major loop. (a) The same sample as Fig. <ref>(a) shows a significantly larger coercivity at lower temperature (leftmost curve) and a broader tail, while sample (b) shows a slightly broader tail (leftmost curve) but little change in coercivity.
cond
mat0307507-squid.jpg
0.414093
8e060310efc24acaa95015dbf2951f75
Schematic representation of our view of the phase diagram. In the undoped limit the staggered flux state develops into the Néel state once gauge fluctuations are included. The pseudogap phase can be thought of as a fluctuating phase between the staggered flux state and the d-wave superconductor. The SU(2) theory allows us to smoothly connect these fluctuations. In this theory different states may be represented by a three-dimensional quantization axis with arrows pointing to the north and south poles representing staggered flux order and arrows in the equator representing the d-wave superconductor. These arrows are schematically shown in the figure. As temperature is lowered, the staggered flux regions (arrows pointing to north or south pole) are localized to form cores of hc/2e vortices. Eventually these vortices disappear via the Kosterlitz-Thouless transition to the d-wave superconducting state. There is a broad range in temperatures above this transition where vortices may give rise to Nernst effects.<cit.> This is sketched in the schematic phase diagram.
cond
mat0307508-Fig.2.jpg
0.489709
b2fef644b1044caaa8a9c53b526c1e61
The condensation energy per site vs. doping as estimated from the difference between the energy of the projected staggered flux state and the projected d-wave superconductor.<cit.> Data are shown for a variety of sample sizes.
cond
mat0307508-Fig.3.jpg
0.556972
45815a36c7c5437cb53863d25293b72f
The functional integrals which give (a) ⟨χ|ψ⟩, (b) ⟨χ|B|ψ⟩, (c) ⟨χ|B^-1|ψ⟩.
cond
mat0307511-B-B2-calc.jpg
0.47111
078a0b8c95924c9fb3cd72a64a73d613
In the top panel, the Jones-Wenzl projector P_n is written in terms of P_n-1. In the bottom two panels, it is shown how P_n annihilates a turnaround on the last two curves.
cond
mat0307511-J-W-proj.jpg
0.485159
76f753876c604989a801cfb4c672df69
The condition imposed on states in the Hilbert space of SU(2)_2×SU(2)_2 Chern-Simons theory by the Jones-Wenzl projector P_3.
cond
mat0307511-P_3-v4.jpg
0.466286
b9682be572564c878874bb2014dcb75b
A typical cluster configuration for the Potts model is shown by dashed lines. Spins belonging to the same cluster take the same value, which must be summed over the q possible values, as described in the text. Clusters can be represented by loops on the surrounding lattice, shown by solid lines.
cond
mat0307511-Potts_loops.jpg
0.430953
1195acd529ff48a9915f3760817f994b
In the top panel, three elements of TL_5 are depicted. In the bottom panel, the generators 1, e_1, e_2, …, e_n-1 are shown.
cond
mat0307511-T-L-eg.jpg
0.54276
be3dfde5213c4af691f237531a1cebb6
The trace of an arbitrary element (the shaded region) of the Temperley-Lieb algebra.
cond
mat0307511-T-L-trace.jpg
0.418483
95ba0540629a42cd9e2dbd77a3c9b1c5
The result of braiding two (1,1) quasiparticles. (1,1) quasiparticles are the type which are displayed in the lower part of fig. <ref>, with the plus sign. They involve a superposition of 2 pictures, hence two quasiparticles involve a superposition of four pictures. A clockwise interchange (second panel) leads, after surgery (third panel), to a phase π/2.
cond
mat0307511-braiding_d=-1.jpg
0.574825
e809977515e1477994a4873728edb1f0
α' and α” are the two ways of resolving the crossing in α
cond
mat0307511-cross-resolutions.jpg
0.523788
ccef57177ce2467396d25ccee778e18e
The second constraint, which assigns a -1 to the operation of cutting and rejoining of two parallel strands requires that a contractible loop change the wavefunction by -1.
cond
mat0307511-d-P_2.jpg
0.487536
99b2fb12a9c4423ba739a6dbbebe8a32
The fusion of two particles. After the first step, the contractible loop can be shrunk directly or after applying the surgery operation. The same result is obtained either way (either 1 or 3 minus signs accrue). Strictly speaking, the particles are the superpositions of fig. <ref>, so this picture must be superposed with 3 others; the result is the same.
cond
mat0307511-fusion_d=-1.jpg
0.491343
c1b5df96830e42c4bfdb4e8c2c1ce5dc
A complete, linearly independent (but not orthogonal) set of 5 states of two n=1 quasiparticles in the k=2 theory.
cond
mat0307511-k=2-four-qps.jpg
0.466744
a8bfc61e1a6e4f6ba29017f198c26103
The 9 quasiparticle species (including the vacuum, or `trivial particle') of the k=2 theory. See text for an explanation.
cond
mat0307511-k=2-qps.jpg
0.471104
1c8cd855ceff46bf96c7250fd15f62a7
The 9 ground states of the k=2 theory on the torus. Opposite sides of the square are identified, as indicated by the arrows.
cond
mat0307511-k=2-torus-Hilbert.jpg
0.477343
a4e32a7283c5469e817ae431c11cb5d8
The four species of quasiparticles (including the vacuum as the trivial quasiparticle) in U(1)_2×U(1)_2 Chern-Simons theory.
cond
mat0307511-m=2-qps.jpg
0.474387
8883089abf464944ab8301de182b0392
A surgery operation of this type changes the wavefunction by a factor of -1. The shaded area represents an arbitrary multi-curve which completes the 1-manifold shown.
cond
mat0307511-p_2-rel.jpg
0.463466
23f560e096ac41c19a28ff98b58e4b9e
Curves which terminate at the quasiparticle (i.e. the inner boundary) or the outer boundary of the annulus must do so at preferred points. The darker lines represent the boundary and the quasiparticle at the origin, while the lighter line represents a curve.
cond
mat0307511-puncture-term.jpg
0.476223
0fc8ae19523c4f70a90a84d62b5fa2ee
When α and γ intersect, α∪_ Rγ is defined as shown above.
cond
mat0307511-r-union.jpg
0.468419
96173ceb8c96498bb257ea24e1e3a12d
α⋆γ must be defined so that α⋆γ=α⋆γ'
cond
mat0307511-reidemeister.jpg
0.546734
6af1eb841fca4247b22ec20ba9dc0435
As a shorthand, we will denote the constraints satisfied by the physical Hilbert space as shown. In the top panel, we have the relations of U(1)_2×U(1)_2 Chern-Simons theory, while in the bottom panel, we have the relations of Z_2 gauge theory.
cond
mat0307511-shorthand.jpg
0.590332
e823c0eb0bb54c2ba0091647359077ed
The operation T_1T_2T^-1_1T^-1_2 results in a phase e^2π i/m because it is equivalent to the braiding operation on the right.
cond
mat0307511-torus2.jpg
0.579291
9c908d19511b47ae8da29dd12360cbe8
By superimposing a dimer configuration (dark lines) on a reference dimer configurations (dotted lines), it can be represented by the depicted multi-loop which consists of alternating solid and dotted lines.
cond
mat0307511-transition-graph.jpg
0.593505
c230fc411b6e45159b56b7777414bd68
S_z basis states of s=1/2 spins on the triangular lattice can be represented in terms of loops on the honeycomb lattice.
cond
mat0307511-triangular-domain.jpg
0.461228
c301d90ca71b40488053161cfedc3f9b
The E-T phase diagram of PZN-8%PT determined by Ohwada et al. <cit.> The notations C, T, M, and R refer to cubic, tetragonal, monoclinic, and rhombohedral phases, respectively. The arrows point in the scan direction, while the length of the arrows correspond to the scan range. The circles denote the transition temperatures determined from each scan.
cond
mat0307513-fig1.jpg
0.441065
607bf7302da04342bc82bcabd8ed4f68
Temperature dependence of the neutron diffuse scattering at Q⃗ = (0.04,0,3) in zero field (solid circles) and E = 3kV/cm // [001] (open circles). The zero-field Bragg intensity measured at (002) is shown as a function of temperature by the solid triangles. These data were taken by Ohwada et al. on a different single crystal of PZN-8%PT.
cond
mat0307513-fig2.jpg
0.538408
27c01e847997478abb429f60cb1a6f0c
Comparison of the diffuse scattering intensities measured near (300) (solid symbols) and (003) (open symbols) as a function of temperature in a field of 2kV/cm applied along [001].
cond
mat0307513-fig3.jpg
0.474653
c1eb6d02641b4be9845a33b134b61d54
A log-scale comparison of the diffuse scattering intensities measured transverse to the (300) (open circles) and (003) (solid circles) reciprocal lattice positions at 400 K (tetragonal phase) and 500 K (cubic phase). The top panels show data in zero field, and the bottom panels show data in an electric field of 2 kV/cm (bottom panels) applied along [001]. The dashed line in the lower left panel represents the instrumental transverse q-resolution.
cond
mat0307513-fig4.jpg
0.453409
58d4f48fe7f1457998eb4177552c7ad0
Plot of the diffuse scattering intensity measured in the tetragonal phase at (3,0,-0.06) at 450 K as a function of increasing field strength applied along the [001] direction. The open and solid circles represent independent measurements. The solid square corresponds to the diffuse scattering intensity at 2 kV/cm at (0.06,0,3) for comparison.
cond
mat0307513-fig5.jpg
0.437545
be32f57fccc649878c256e9080c6d211
Top panel: Phase-shifted PNR embedded in a non-polar cubic lattice for T > T_c. Middle panel: Below T_c the lattice achieves a polarization of its own, but the phase shift remains. Bottom panel: An applied electric field removes shifts along E, while preserving those that are orthogonal.
cond
mat0307513-fig6.jpg
0.381729
4706a96f4feb4b2a8265854ac8a3ecc7
Numerical results for a 4+2 particle chain with asymmetric mass-distribution. The particle masses are 100, 1.7, 1.4, 1.9, 1.3 and 100. The left-to-right current J_10=|b_2(1,0)|^2 is plotted as a function of the nonlinearity ϵ. The nonreciprocity, J_10-J_01, is shown for 7 different values of ϵ, and compared with the phase averaged current J_11 with both sources on (right vertical scale). The two quantities are equal in the perturbative state, but not in the non-perturbative state.
cond
mat0307514-fig1.jpg
0.457291
c6b5e62b9d84485c91c2de0447b41692
A schematic view of the splitting of the five-fold orbitals under cubic crystal field. The transition metal is located at the center of the oxygen octahedron.
cond
mat0307515-harrisfig1small.jpg
0.46511
fa8e826802eb41ae907c0b08db3f345b
Spin and orbital configurations for a cube of eight sites. The thick lines indicate singlet spin states (dimers) and X and Y indicate the orbital states of the electrons. (a) Schematic illustration of the quantum numbers associated with occupation of α orbitals in an α-plane. (Similar spin quantum numbers are not shown.) (b) and (c) Dominant configurations in the ground state wavefunction. Hopping between x and y orbital states is only allowed along z-direction bonds and these bonds are labeled in panel (c). (d) A subdominant configuration in the ground state which is obtained from (c) by allowing the interchange of two (X and Y) electrons along the z-axis bond labeled “d” while each electron retains its membership in its original spin singlets (even though the position of the electron has changed). If |0⟩ denotes configuration (c), then configuration (b) is | 0⟩≡ abcd|0 ⟩ and (d) is d|0 ⟩ in the notation of Table I. We also show (e) and (f) states with the same quantum numbers but which are not coupled to the manifold of 16 states we analyze. These two states are actually eigenstates of the Hamiltonian with energy -4ϵ.
cond
mat0307515-harrisfig3small.jpg
0.480582
3f911afda3644d22a37fdb47ad530c24
Temperature dependence of resistivity at B=0. The closed symbols are for the Si/SiGe sample and the open symbols for a Si-MOS sample used in Ref. Okamoto99. The dotted lines mark T=0.5T_F or T=T_F. The contribution of phonon-scattering to ρ in the Si/SiGe sample is calculated based on Ref. Paul and subtracted from the experimental data (dashed lines).
cond
mat0307521-fig1.jpg
0.451367
7199dc3bd7af4da6aa9633d17c792edf
Resistivity at 0.36 K (except the dashed line at 3.6 K) as a function of the in-plane magnetic field for different electron concentrations N_s (10^15m^-2). The closed symbols are for the current orientation of j ⊥ B and the open symbols for j ∥ B. The arrows indicate the critical magnetic field for the full spin polarization.
cond
mat0307521-fig2.jpg
0.391727
f94c26bc520f4b699fdaa1842b933a82
(a) Shubnikov-de Haas oscillations at T=0.36 K and N_s =1.00 × 10^15 m^-2 for different B_tot. (b) The spin polarization obtained from B_⊥ at the ρ_xx minima.
cond
mat0307521-fig3.jpg
0.434325
f9ee8c5f218d44fe936d323b812c6c1d
r_s dependence of α= g_FL m_FL / g_b m_b. The closed and open circles are determined from B_c in the magnetoresistance shown in Fig. <ref> using B_c = 2 πħ^2 N_s / μ_B g_v g_FL m_FL. The solid line shows the experimental data on Si-MOS structures with κ=7.7. <cit.> N_s is also indicated for the silicon quantum well (κ = 11.5) and for Si-MOS structures (κ= 7.7). The dashed line is after the correction of κ (see text).
cond
mat0307521-fig4.jpg
0.507324
65526d48fc8d47f5ab9ed4ac109656ec
ρ vs T/T_F in in-plane magnetic fields above B_c: B_||=9 T for N_s =1.00 and 1.03 × 10^15m^-2, and B_||=11 T for N_s =1.34 and 1.69 × 10^15m^-2. The spin polarization P is indicated for each N_s. The contribution of phonon-scattering is calculated based on Ref. Paul and subtracted from the experimental data (dashed lines).
cond
mat0307521-fig5.jpg
0.470773
6745657371394ccb9119e8d15f10e651
Schematic phase diagram for T-dependence of ρ in low-resistivity and strongly-correlated 2D systems.
cond
mat0307521-fig6.jpg