School of Mathematics and Statistics The University of ... · 2020 · 09 · 30 Jules Lamers · New...
Transcript of School of Mathematics and Statistics The University of ... · 2020 · 09 · 30 Jules Lamers · New...
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 1/18
How coordinate Bethe ansatzworks for Inozemtsev model
Jules Lamers
School of Mathematics and Statistics
The University of Melbourne
joint work with Rob Klabbers (Nordita)
building on [V. I. Inozemtsev, ’90–’99]
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 2/18
Exactly solvablelong-range spin chains
anisotropic(elliptic)
partially isotropic(trig)
isotropic(rational)
nearest-neighbour(contact)
intermediate(elliptic)
long range(trigonometric)
Heisenberg XXXInozemtsev
Haldane–Shastry
q-deformed Haldane–Shastry
q-deformed Inozemtsev ?(some version of)
Heisenberg XXZ
(some version of)Heisenberg XYZ
anisotropic Inozemtsev ??
anisotropic Haldane–Shastry ??
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 3/18
Inozemtsev’s elliptic spin chainAnatomy
long-range pairwise
elliptic pair potential
[Inozemtsev ’90, ’95]
periods
[Haldane ’88][Shastry ’88][Inozemtsev ’92]
[Heisenberg ’28]
spin exchange
Weierstrass
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 4/18
Inozemtsev’s elliptic spin chainExactly solvable interpolation
●●
●●
●●●●●●
●●●●
●●
●●
●●
●●
●●
●●
●●●●
●●
●●●●
●●
●●
●●
●●
0
2
4
6
8
10
0
π2
6
π2
3
π2
2
2π2
3
5π2
6
π2
κ
E EHeisenberg● Exactly solvable
up to solving Bethe-ansatz equations
● Quantum integrable
Haldane–Shastry ● Exactly solvable
in closed form(with Jacks)
● Quantum integrable
● Yangian symmetry
Inozemtsev● Exactly solvable
spectrum up to solving BAE● Quantum integrable?
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 5/18
Inozemtsev’s elliptic spin chainSome highlights
● Spectrum for M=2 using Hermite’s solution of Lamé equation
● Spectrum in hyperbolic limitvia connection to hyperbolic Calogero–Sutherland
● Spectrum for general Mvia connection to elliptic Calogero–Sutherland
● Asymptotic Yangian symmetry
● Thermodynamic Bethe ansatz
● Proposal for higher Hamiltonians
● Guest appearance in AdS/CFT
[Inozemtsev ’90]
[Inozemtsev ’92]using [Chalykh Veselov ’90]
[Inozemtsev ’95, ’99]using [Felder Varchenko ’95]
[Inozemtsev ’96]
[Serban Staudacher ’04]
[Dittrich Inozemtsev ’97][Klabbers ’16]
[Ha Haldane ’93][De La Rosa Gomez et al ’16]
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 6/18
Our goal
Understand Inozemtsev’s solution and its limits
●●
●●
●●●●●●
●●●●
●●
●●
●●
●●
●●
●●
●●●●
●●
●●●●
●●
●●
●●
●●
0
2
4
6
8
10
0
π2
6
π2
3
π2
2
2π2
3
5π2
6
π2
κ
E E
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 7/18
StrategyGeneral considerations
● Isotropy ( -invariance): fix , focus on highest weight
● Use coordinate basis (so will have to ensure cyclicity)
● Homogeneity (translational invariance) determinesDispersion relation
Heisenberg Haldane–Shastry
reciprocal periods [Inozemtsev ’90][Klabbers JL]
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 8/18
Strategy Extended coordinate Bethe ansatz
For general seek wave functions of the form
Assumptions● [technical] has simple poles at equal arguments
● double quasiperiodicity
Cyclicity of requires Bethe-ansatz equations
plane wave
[Inozemtsev ’95][Klabbers JL]
depends on positions!
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 9/18
Connection with eCSResult of extended CBA
If for quantum elliptic Calogero–Sutherland
(for with simple poles at equal arguments)and if BAE
Then as long as we identify [→ next slide]
has
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 10/18
Connection with eCSComments
● rapidities
● is additive iff is so (which doesn’t seem to be the case)
● will have simple poles or double zeroes at equal arguments; latter would be better in HS limit [but see later]
eCS wave function ( )plane wavequasimomenta
[Klabbers JL]
[Klabbers JL]
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 11/18
Two magnonsFrom Lamé to Inozemtsev
● LaméHermite
● Quasiperiodicity fixes in terms of
● Inozemtsev
provided BAEscattering phase
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 12/18
Two magnonsRationalisation & completeness
● Eigenvectors with all have highest weight
● Together the BAE
give an equation elliptic in
● In elliptic coordinates the latter gives a polynomial equation
– Very efficient for numerical solutions
– Can count number of solutions: matches # highest weight
● Energy function is elliptic in too
● Full spectral problem becomes rational in elliptic coordinates
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 13/18
Two magnonsHeisenberg limit
● We have
so
● Energy becomes additive
● BAE
remains finite
becomes Bethe’s wave function
[Klabbers JL]
Bethe’s equation
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 14/18
Two magnonsRecap of Heisenberg
○○○○○○○○○○○○○○○○○○○○○○
1520○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
304050100
●●
●●
●●
●●
●●
●●●●
●●●●●●●●●●●●●●●●●●●●●●
6.125
6.25
6.5
7
89
10
1520
●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●304050100
0 π
8
π
4
3π
8
π
2
0
1
2
3
Re pm
Impm
○○
○○
2021
○○ ○○○○○○○○
2223242526
●●
●●
2021
●●
Lcr
●●
22
0.41 0.42 0.43 0.44 0.45 0.46 0.47
-0.05
0.00
0.05
‘critical length’
π
2π
3π
2
ptot
2
4
6
8
E
descendants
scattering
cf ◦ free (φ=0)
bound
unbound
exceptional
π
2π
3π
2
ptot
2
4
6
8
E
descendants
scattering
cf ◦ free (φ=0)
bound
unboundπ
2π
3π
2
ptot
2
4
6
8
E
descendants
scattering
cf ◦ free (φ=0)
bound
unbound
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 15/18
0.5 1 1.5 2 2.5 3 ∞0
10
20
30
40
50
60
70
410
21.9
61.3
κ
L
n=5
n=4
n=3
n=2
Two magnonsInozemtsev: critical loci
π
2π
3π
2
ptot
2
4
6
8
E
descendants
scattering
cf ◦ free (φ=0)
bound
unbound (beyond cr,r)
exceptional
bound (beyond cr,i)π
2π
3π
2
ptot
2
4
6
8
E
descendants
scattering
cf ◦ free (φ=0)
bound
unbound (beyond cr,r)
exceptional
bound (beyond cr,i)
[Klabbers JL]
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 16/18
Two magnonsHaldane–Shastry limit (I)
π
2π
3π
2
ptot
2
4
6
8
10
E
descendants
motifsπ
2π
3π
2
ptot
2
4
6
8
10
E
descendants
motifs
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 17/18
Two magnons Haldane–Shastry limit (II)
● We have
so
with ‘evaluation’
● BAE become trivial
● Energy becomes (strictly) additive
[Klabbers JL]
Scattering states:Schur expansion of
Jack (upon ev)
2020 · 09 · 30 Jules Lamers · New connections · UQ/zoom 18/18
●●
●●
●●●●●●
●●●●
●●
●●
●●
●●
●●
●●
●●●●
●●
●●●●
●●
●●
●●
●●
0
2
4
6
8
10
0
π2
6
π2
3
π2
2
2π2
3
5π2
6
π2
κ
E E
● Solution via relation with elliptic Calogero–Sutherland
● ‘Quasi-additive’ energy
● limits in detail, recover known results
● Spectral problem rationalises
● Open questions
– ?
– Quantum integrability (R, etc) ?
– XXZ-like (q-)analogue ?
Conclusion