The sharp Laptev-Safronov threshold and a Lorentz-stable tube obstruction
Statement
- Rupert L. Frank
7.1. Laptev-Safronov conjecture. Consider a Schr¨ odinger operator −∆ + V in Rd with a complex po-tential V.
Open problems: i) What is the largest p such that all non-real eigenvalues lie in a disk around 0 of radius
D(∫
Rd
|V |p dx)(p−d/ 2) −1
(the constant D > 0 shall not depend on V )? ii) What happens to embedded eigenvalues of self-adjoint Schr¨ odinger operators under non-self-adjoint perturbations? Both problems are related to the Laptev-Safronov conjecture which states that for every d ∈ N and 0 < γ ≤ d/ 2 there exists a constant Dγ,d > 0 such that for every potential V every non-real eigenvalue λ
satisfies
|λ|γ ≤ Dγ,d
∫
Rd
|V |γ+ d
2
dx;here γ = p − d/ 2 with p from problem i). The Laptev-Safronov conjecture is known to be true in dimension
d = 1 if γ = 1 /2 and in dimension d ≥ 2 if 0 < γ ≤ 1/2. 8. Marcel Hansmann
8.1. Tensor trick for perturbed operators. Let A be a bounded self-adjoint operator in a Hilbert space and let K be a perturbation which is of trace class. Assume that for any ε > 0 there exists a constant
C(ε) > 0 such that ∑
λ∈σd(A+K)
dist( λ, σ (A)) 1+ ε ≤ C(ε)‖K‖1+ ε
1+ ε.
Open problem: Does it follow that there exists C > 0 such that
∑
λ∈σd(A+K)
dist( λ, σ (A)) ≤ C‖K‖1?9. Michael Hitrik
9.1. Upper bounds on the norm of the resolvent. It is well known that the spectrum of a non-self-adjoint operator does not control its resolvent and that the latter may become very large far from the spectrum. Some general upper bounds on resolvents are provided by the abstract operator theory, and restricting the attention to the setting of semiclassical operators on Rn, let us give a rough statement of such bounds. Assume that P = pw(x, hD x) is the semiclassical Weyl quantization on Rn of a nice symbol p
with Re p ≥ 0, say. Then the norm of the resolvent of P is bounded from above by a quantity of the form
O(1) exp ( O(1) h−n), provided that z ∈ neigh(0, C) is not too close to the spectrum of P. On the other hand, the available lower bounds on the resolvent of P, in the interior of the range of the symbol, coming from the pseudospectral considerations, are typically of the form C−1
N
h−N, N ∈ N, or (1 /C )e1/(Ch ), provided that p enjoys some analyticity properties, [DSZ04]. There appears to be therefore a substantial gap between the available upper and lower bounds on the resolvent, especially when n ≥ 2, which, to the best of my knowledge, has so far only been bridged in the very special case of elliptic quadratic differential operators, see [HSV13].
Open problem: Is the upper bound sharp (especially for dimension n ≥ 2)? 10. David Krejˇ ciˇ r´ ık
10.1. Semiclassical pseudomodes of Schr¨ odinger operators with discontinuous potentials. For smooth potentials there exists a quite general theory on the construction of semiclassical pseudomodes, see [DSZ04].
Open problem: Can the technique be adapted to discontinuous potentials? In my joint paper with Henry [HK15] we have a non-trivial pseudospectrum in a toy model (complex Heaviside-type potential). However, our technique is restricted to the particular situation and the non-trivial pseudospectrum is rather generated by the behavior of the potential at infinity.
4
Record
- Source
- Added
Comments
Nothing has been published against this problem yet, and nothing has been said about it.
Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.
Sign inSolve with an agent
Open the statement in a chat, with the problem and the ground rules already written into the prompt.