arXiv subjects

Yanfei Yang

Publications and source records attributed to Yanfei Yang.

15 recordsLinked to original sources

arXiv2024

Hybrid LSMR algorithms for large-scale general-form regularization

The hybrid LSMR algorithm is proposed for large-scale general-form regularization. It is based on a Krylov subspace projection method where the matrix $A$ is first projected onto a subspace, typically a Krylov subspace, which is implemented via the Golub-Kahan bidiagonalization process applied to $A$, with starting vector $b$. Then a regularization term is employed to the projections. Finally, an iterative algorithm is exploited to solve a least squares problem with constraints. The resulting algorithms are called the {hybrid LSMR algorithm}. At every step, we exploit LSQR algorithm to solve the inner least squares problem, which is proven to become better conditioned as the number of $k$ increases, so that the LSQR algorithm converges faster. We prove how to select the stopping tolerances for LSQR in order to guarantee that the regularized solution obtained by iteratively computing the inner least squares problems and the one obtained by exactly computing the inner least squares problems have the same accuracy. Numerical experiments illustrate that the best regularized solution by the hybrid LSMR algorithm is as accurate as that by JBDQR which is a joint bidiagonalization based algorithm.

math.NA↗

arXiv2024

Constructing 100 M{\Omega} and 1 G{\Omega} Resistance Standards via Star-Mesh Transformations

A recent mathematical framework for optimizing resistor networks to achieve values in the M{\Omega} through G{\Omega} levels was employed for two specific cases. Objectives here include proof of concept and identification of possible apparatus limitations for future experiments involving graphene-based quantum Hall array resistance standards. Using fractal-like, or recursive, features of the framework allows one to calculate and implement network designs with substantially lower-valued resistors. The cases of 100 M{\Omega} and 1 G{\Omega} demonstrate that, theoretically, one would not need more than 100 quantum Hall elements to achieve these high resistances.

cond-mat.mes-hall↗

arXiv2023

Fractal-like star-mesh transformations using graphene quantum Hall arrays

A mathematical approach is adopted for optimizing the number of total device elements required for obtaining high effective quantized resistances in graphene-based quantum Hall array devices. This work explores an analytical extension to the use of star-mesh transformations such that fractal-like, or recursive, device designs can yield high enough resistances (like 1 E{\Omega}, arguably the highest resistance with meaningful applicability) while still being feasible to build with modern fabrication techniques. Epitaxial graphene elements are tested, whose quantized Hall resistance at the nu=2 plateau (R_H = 12906.4 {\Omega}) becomes the building block for larger effective, quantized resistances. It is demonstrated that, mathematically, one would not need more than 200 elements to achieve the highest pertinent resistances

cond-mat.mes-hall↗

arXiv2023

Star-Mesh Quantized Hall Array Resistance Devices

Advances in the development of graphene-based technology have enabled improvements in DC resistance metrology. Devices made from epitaxially grown graphene have replaced the GaAs-based counterparts, leading to an easier and more accessible realization of the ohm. By optimizing the scale of the growth, it has become possible to fabricate quantized Hall array resistance standards (QHARS) with nominal values between 1 k{\Omega} and 1.29 M{\Omega}. One of these QHARS device designs accommodates a value of about 1.01 M{\Omega}, which made it an ideal candidate to pursue a proof-of-concept that graphene-based QHARS devices are suitable for forming wye-delta resistance networks. In this work, the 1.01 M{\Omega} array output nearly 20.6 M{\Omega} due to the wye-delta transformation, which itself is a special case of star-mesh transformations. These mathematical equivalence principles allow one to extend the QHR to the 100 M{\Omega} and 10 G{\Omega} resistance levels with fewer array elements than would be necessary for a single array with many more elements in series. The 1.01 M{\Omega} device shows promise that the wye-delta transformation can shorten the calibration chain, and, more importantly, provide a chain with a more direct line to the quantum SI.

cond-mat.mes-hall↗

arXiv2023

Hybrid CGME and TCGME algorithms for large-scale general-form regularization

Two new hybrid algorithms are proposed for large-scale linear discrete ill-posed problems in general-form regularization. They are both based on Krylov subspace inner-outer iterative algorithms. At each iteration, they need to solve a linear least squares problem, which is the inner least squares problem. It is proved that inner linear least squares problems, solved by LSQR, become better conditioned as k increases, so LSQR converges faster. We also prove how to choose the stopping tolerance for LSQR to guarantee that the computed and exact best regularized solutions have the same accuracy. Numerical experiments are provided to demonstrate the effectiveness and efficiency of our new hybrid algorithms, along with comparisons to the existing algorithm.

math.NA↗

arXiv2022

Designs for programmable quantum resistance standards based on epitaxial graphene p-n junctions

We report the fabrication and measurement of top gated epitaxial graphene p-n junctions where exfoliated hexagonal boron nitride (h-BN) is used as the gate dielectric. The four-terminal longitudinal resistance across a single junction is well quantized at the von Klitzing constant R_K with a relative uncertainty of 10-7. After the exploration of numerous parameter spaces, we summarize the conditions upon which these devices could function as potential resistance standards. Furthermore, we offer designs of programmable electrical resistance standards over six orders of magnitude by using external gating.

cond-mat.mes-hall↗

arXiv2018

A Joint Bidiagonalization Based Algorithm for Large Scale Linear Discrete Ill-posed Problems in General-Form Regularization

Based on the joint bidiagonalization process of a large matrix pair $\{A,L\}$, we propose and develop an iterative regularization algorithm for the large scale linear discrete ill-posed problems in general-form regularization: $\min\|Lx\| \ \mbox{{\rm subject to}} \ x\in\mathcal{S} = \{x|\ \|Ax-b\|\leq \tau\|e\|\}$ with a Gaussian white noise $e$ and $\tau>1$ slightly, where $L$ is a regularization matrix. Our algorithm is different from the hybrid one proposed by Kilmer {\em et al.}, which is based on the same process but solves the general-form Tikhonov regularization problem: $\min_x\left\{\|Ax-b\|^2+\lambda^2\|Lx\|^2\right\}$. We prove that the iterates take the form of attractive filtered generalized singular value decomposition (GSVD) expansions, where the filters are given explicitly. This result and the analysis on it show that the method must have the desired semi-convergence property and get insight into the regularizing effects of the method. We use the L-curve criterion or the discrepancy principle to determine $k^*$. The algorithm is simple and effective, and numerical experiments illustrate that it often computes more accurate regularized solutions than the hybrid one.

math.NA↗

arXiv2018

Confocal laser scanning microscopy: A tool for rapid optical characterization of 2D materials

Confocal laser scanning microscopy (CLSM) is a non-destructive, highly-efficient optical characterization method for large-area analysis of graphene on different substrates, which can be applied in ambient air, does not require additional sample preparation, and is insusceptible to surface charging and surface contamination. CLSM leverages optical properties of graphene and provides greatly enhanced optical contrast and mapping of thickness down to a single layer. We demonstrate the effectiveness of CLSM by measuring mechanically exfoliated and chemical vapor deposition graphene on Si/SiO2, and epitaxial graphene on SiC. In the case of graphene on Si/SiO2, both CLSM intensity and height mapping is powerful for analysis of 1-5 layers of graphene. For epitaxial graphene on SiC substrates, the CLSM intensity allows us to distinguish features such as dense, parallel 150 nm wide ribbons of graphene (associated with the early stages of the growth process) and large regions covered by the interfacial layer and 1-3 layers of graphene. In both cases, CLSM data shows excellent correlation with conventional optical microscopy, atomic force microscopy, Kelvin probe force microscopy, conductive atomic force microscopy, scanning electron microscopy and Raman mapping, with a greatly reduced acquisition time. We demonstrate that CLSM is an indispensable tool for rapid analysis of mass-produced graphene and is equally relevant to other 2D materials.

physics.app-ph↗

arXiv2017

Rapid characterization of wafer-scale 2D material: Epitaxial graphene and graphene nanoribbons on SiC

We demonstrate that the confocal laser scanning microscopy (CLSM) provides a non-destructive, highly-efficient characterization method for large-area epitaxial graphene and graphene nanostructures on SiC substrates, which can be applied in ambient air without sample preparation and is insusceptible to surface charging or surface contamination. Based on the variation of reflected intensity from regions covered by interfacial layer, single layer, bilayer, or few layer graphene, and through the correlation to the results from Raman spectroscopy and SPM, CLSM images with a high resolution (around 150 nm) reveal that the intensity contrast has distinct feature for undergrown graphene (mixing of dense, parallel graphene nanoribbons and interfacial layer), continuous graphene, and overgrown graphene. Moreover, CLSM has a real acquisition time hundreds of times faster per unit area than the supplementary characterization methods. We believe that the confocal laser scanning microscope will be an indispensable tool for mass-produced epitaxial graphene or applicable 2D materials.

physics.app-ph↗

arXiv2017

Part-per-million quantization and current-induced breakdown of the quantum anomalous Hall effect

In the quantum anomalous Hall effect, quantized Hall resistance and vanishing longitudinal resistivity are predicted to result from the presence of dissipationless, chiral edge states and an insulating 2D bulk, without requiring an external magnetic field. Here, we explore the potential of this effect in magnetic topological insulator thin films for metrological applications. Using a cryogenic current comparator system, we measure quantization of the Hall resistance to within one part per million and longitudinal resistivity under 10 m$Ω$ per square at zero magnetic field. Increasing the current density past a critical value leads to a breakdown of the quantized, low-dissipation state, which we attribute to electron heating in bulk current flow. We further investigate the pre-breakdown regime by measuring transport dependence on temperature, current, and geometry, and find evidence for bulk dissipation, including thermal activation and possible variable-range hopping.

cond-mat.mes-hall↗

arXiv2017

Modified Truncated Randomized Singular Value Decomposition (MTRSVD) Algorithms for Large Scale Discrete Ill-posed Problems with General-Form Regularization

In this paper, we propose new randomization based algorithms for large scale linear discrete ill-posed problems with general-form regularization: ${\min} \|Lx\|$ subject to ${\min} \|Ax - b\|$, where $L$ is a regularization matrix. Our algorithms are inspired by the modified truncated singular value decomposition (MTSVD) method, which suits only for small to medium scale problems, and randomized SVD (RSVD) algorithms that generate good low rank approximations to $A$. We use rank-$k$ truncated randomized SVD (TRSVD) approximations to $A$ by truncating the rank-$(k+q)$ RSVD approximations to $A$, where $q$ is an oversampling parameter. The resulting algorithms are called modified TRSVD (MTRSVD) methods. At every step, we use the LSQR algorithm to solve the resulting inner least squares problem, which is proved to become better conditioned as $k$ increases so that LSQR converges faster. We present sharp bounds for the approximation accuracy of the RSVDs and TRSVDs for severely, moderately and mildly ill-posed problems, and substantially improve a known basic bound for TRSVD approximations. We prove how to choose the stopping tolerance for LSQR in order to guarantee that the computed and exact best regularized solutions have the same accuracy. Numerical experiments illustrate that the best regularized solutions by MTRSVD are as accurate as the ones by the truncated generalized singular value decomposition (TGSVD) algorithm, and at least as accurate as those by some existing truncated randomized generalized singular value decomposition (TRGSVD) algorithms.

math.NA↗

arXiv2016

Epitaxial graphene homogeneity and quantum Hall effect in millimeter-scale devices

Quantized magnetotransport is observed in 5.6 x 5.6 mm^2 epitaxial graphene devices, grown using highly constrained sublimation on the Si-face of SiC(0001) at high temperature (1900 °C). The precise quantized Hall resistance of Rxy = h/2e^2 is maintained up to record level of critical current Ixx = 0.72 mA at T = 3.1 K and 9 T in a device where Raman microscopy reveals low and homogeneous strain. Adsorption-induced molecular doping in a second device reduced the carrier concentration close to the Dirac point(n ~ 1E10 (1/cm^2)), where mobility of 43700 cm^2/Vs is measured over an area of 10 mm^2. Atomic force, confocal optical, and Raman microscopies are used to characterize the large-scale devices, and reveal improved SiC terrace topography and the structure of the graphene layer. Our results show that the structural uniformity of epitaxial graphene produced by face-to-graphite processing contributes to millimeter-scale transport homogeneity, and will prove useful for scientific and commercial applications.

cond-mat.mtrl-sci↗

arXiv2016

Floating up of the zero-energy Landau level in monolayer epitaxial graphene

We report on magneto-transport measurements on low-density, large-area monolayer epitaxial graphene devices grown on SiC. We show that the zero-energy Landau level (LL) in monolayer graphene, which is predicted to be magnetic field ($B$)-independent, can float up above the Fermi energy at low $B$. This is supported by the temperature ($T$)-driven flow diagram approximated by the semi-circle law as well as the $T$-independent point in the Hall conductivity $σ_{xy}$ near $e^2/h$. Our experimental data are in sharp contrast to conventional understanding of the zeroth LL and metallic-like behavior in pristine graphene prepared by mechanical exfoliation at low $T$. This surprising result can be ascribed to substrate-induced sublattice symmetry breaking which splits the degeneracy of the zeroth Landau level. Our finding provides a unified picture regarding the metallic behavior in pristine graphene prepared by mechanical exfoliation, and the insulating behavior and the insulator-quantum Hall transition in monolayer epitaxial graphene.

cond-mat.dis-nn↗

arXiv2014

New rigorous perturbation bounds for the generalized Cholesky factorization

Some new rigorous perturbation bounds for the generalized Cholesky factorization with normwise or componentwise perturbations in the given matrix are obtained, where the componentwise perturbation has the form of backward rounding error for the generalized Cholesky factorization algorithm. These bounds can be much tighter than some existing ones while the conditions for them to hold are simple and moderate.

math.NA↗

arXiv2014

Low Carrier Density Epitaxial Graphene Devices On SiC

Monolayer epitaxial graphene (EG) grown on hexagonal Si-terminated SiC substrates is intrinsically electron-doped (carrier density is about 10^13 cm^(-2)). We demonstrate a clean device fabrication process using a precious-metal protective layer, and show that etching with aqua regia results in p-type (hole) molecular doping of our un-gated, contamination-free EG. Devices fabricated by this simple process can reach a carrier density in the range of 10^10 cm^(-2) to 10^11 cm^(-2) with mobility about 8000 cm^2/V/s or higher. In a moderately doped device with a carrier density n = 2.4 x 10^11 cm^(-2) and mobility = 5200 cm^2/V/s, we observe highly developed quantized Hall resistance plateaus with filing factor of 2 at magnetic field strengths of less than 4 T. Doping concentrations can be restored to higher levels by heat treatment in Ar, while devices with both p-type and n-type majority carriers tend to drift toward lower carrier concentrations in ambient air.

cond-mat.mes-hall↗