Journal of Nonlinear Mathematical Physics

Volume 27, Issue 1, October 2019, Pages 12 - 16

Remarks on the mass spectrum of two-dimensional Toda lattice of E8 type

Authors
Askold M. Perelomov
Institute of Theoretical and Experimental Physics, 117259 Moscow, Russia,aperelomo@gmail.com
Received 10 March 2019, Accepted 12 April 2019, Available Online 25 October 2019.
DOI
10.1080/14029251.2020.1683961How to use a DOI?
Abstract

A simple procedure for obtaining the mass spectrum of 2-dimensional Toda lattice of E8 type is given.

Copyright
© 2020 The Authors. Published by Atlantis and Taylor & Francis
Open Access
This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/).

1. Introduction. Basics

Let us recall several definitions; for more details, see the book [10].

Let 𝔤 be a simple Lie algebra of rank l, let R+ (resp. R) be the set of its positive (resp. negative) roots, and {α1,...,αl} be the set of simple roots. Let W be the Weyl group of the root system R acting in the space V = ℝl, let (·,·) be the non-degenerate W-invariant bilinear form in V, δ=1jlnjαj be the highest root, α0 = −δ, and h=1+1jlnj be the Coxeter number.

The 2-dimensional Toda lattice is 2-dimensional relativistic field theory describing l interacting scalar fields. The 2-dimensional Toda lattice is related with 𝔰𝔩 (n).

In the paper [8], the 2-dimensional Toda lattice was generalized for the case of any simple finite-dimensional Lie algebra 𝔤; it was shown that the generalized construction has remarkable integrability properties. This is a relativistic system with Lagrangian

L=12μμφU(φ),whereμ=0,1andφ=φ(x0,x1)isanldimensionalvector., where μ = 0, 1 and ϕ = ϕ(x0, x1) is an l-dimensional vector.

The potential U(φ) is constructed using the set of simple roots {αj}j=0l of the simple Lie algebra 𝔤of rank l:

U(φ)=0jlexp(2αj,φ).

In [8], the mass spectrum of scalar fields was found for all simple Lie algebras, except for the most complicated case 𝔤 = E8. For this algebra only numerical result was given.

In this note I describe two simple methods for obtaining the mass spectrum in the E8 case. Note that both methods work also for any other finite-dimensional simple Lie algebra.

The numbering of simple roots of the Lie algebra E8 is given on the Dynkin diagram:

The Dynkin diagram for the Lie algebra E8.

For this numbering, the highest root δ has the form

δ=2α1+3α2+4α3+5α4+6α5+4α6+2α7+3α8.

Observe that in 1989 A.B. Zamolodchikov discovered, using conformal theory, that this system appears also in the Ising model with nonzero magnetic field and explicitly calculated the mass spectrum, see [13]. The four mass ratios are equal to the “golden ratio”

r=5+12=2cos(π5)=1.6180339887

This remarkable property is related to the fact that the Coxeter number h = 30 of the Lie algebra E8 is divisible by 5.

In 2010, Zamolodchikov’s theory was experimentally confirmed for 1-dimensional Ising ferro-magnet (cobalt niobate) near its critical point [2].

2. Method 1

As it was shown in papers [1, 5] the masses of particles are proportional to the components of a special eigenvector of the matrix A = 2I − C, where C is the Cartan matrix of 𝔤. This eigenvector is called the Perron–Frobenius vector, see [6, 12]. For 𝔤= E8, we have

A=(0100000010100000010100000010100000010101000010100000010000001000)

The characteristic equation of this matrix is

x87x6+14x48x2+1=0,
and its roots are
xj=2cos(ajθ),whereθ=πh,andh=30istheCoxeternumber;
the numbers aj ∈ {1, 7, 11, 13, 17, 19, 23, 29} for 1 ≤ j ≤ 8 are called the exponents of E8. Note that they have no common divisors with the Coxeter number.

Note also that x5 = −x4, x6 = −x3, x7 = −x2, x8 = −x1. Let us give the expressions of the xj in terms of radicals (these expressions might be used in calculations):

x1=127+5+30+65,x2=127+530+65,x3=1275+3065,x4=12753065.

The matrix A has nonnegative elements and according to the Perron–Frobenius theorem [6, 12] it has a unique eigenvector (the Perron-Frobenius eigenvectors)

u=(u1,u2,u3,u4,u5,u6,u7,u8)
all coordinates of which are positive. This eigenvector corresponds to the maximal eigenvalue λ = 2cos(θ) and we have
uA=λu,
or, in more details,
u2=λu1,u1+u3=λu2,u2+u4=λu3,u3+u5=λu4,u4+u6+u8=λu5,u5+u7=λu6,u6=λu7,u5=λu8.

Solving the system of these equations, and fixing u1 = 2sin(θ), we obtain:

u=(2sin(θ),2sin(2θ),2sin(3θ),2sin(4θ),2sin(5θ),sin(2θ)sin(3θ),sin(θ)sin(3θ),sin(θ)sin(2θ)),(2.1)
or, approximately,
u=(0.2091;0.4158;0.6180;0.8135;1;0.6728;0.3383;0.5028).

Note that from eq. (2.1) it follows that (recall that θ=π30)

u7u1=r,u6u2=r,u5u3=r,u4u8=r,wherer=1+52=2cos(π5).(2.2)

This is a very nice solution, because these expressions for uj can be written immediately just by looking at the Dynkin diagram of E8.

Observe that for any simple Lie algebra the eigenvector corresponding to the maximal eigenvalue can also be written just by looking at the corresponding Dynkin diagram.

Let me also give expressions for some trigonometric quantities in terms of radicals (and use this occasion to correct a typo in the definition of H3 on p. 382 of [9], where ε=2cos(π3) should be ε=2cos(π5)

2cos(π5)=1+52=r,2sin(π5)=552,2cos(π10)=5+522sin(π10)=352,2cos(π15)=129+5+23552,2sin(π15)=127523552,2cos(π30)=127+5+235+52,2cos(π30)=1295235+52.

3. Method 2

In the paper [8] it was shown that the squares of masses are eigenvalues of the 8 × 8 matrix whose elements are

Ba,b=0jlnjαjaαjb,wheren0=1,
and where quantities αja are coordinates of the vector αj, and the nj for j > 0 are coordinates of the vector δ=1jlnjαj.

For the Lie algebra E8, the characteristic polynomial P of this matrix is

P=x860x7+1440x618000x5+127440x4518400x3+1166400x21296000x+518400.

In the paper [1], it was observed that P = P1 P2, where

P1=x430x3+240x2720x+720,P2=x430x3+300x21080x+720.

It is easy to check that the roots of polynomial P1 (resp. P2) are

m12,m32,m42,m62(resp.m22,m52,m72,m82).

Note that

u2u5u7u8=u1u3u4u6,andmj2=Muj2.(3.1)

The quantity M=23sin(6θ)sin(θ) can be found from the equation

M4(u2u5u7u8)2=720.

So, formula (3.1) gives a relation between methods 1 and 2.

Let me give also the explicit expression for quantities mj2 in terms of radicals:

m52=1215+35+625+115,m42=1215+35+655,m72=1215+35625+115,m62=1215+35655,m82=121535+625115,m32=121535+655,m22=121535625115,m12=12153565=5.

4. Conclusion

The remarkable property of the system under consideration is that the four mass ratios in (2.2) are equal to the “golden ratio”.

This is one more phenomenon of many in which the golden ratio appears. The golden ratio has a very long history, see e.g., the book [4, Ch. 11]. The first book on this topic, “Divina Proportione”, illustrated by Leonardo da Vinci, was published by Italian mathematician Luca Paccioli in 1509 [11].

Concluding, I would like to give here a quotation of the outstanding astronomer and mathematician Johannes Kepler [7]: “Geometry has two treasures: one of them is the Pythagorean theorem, and the other is dividing the segment in average and extreme respect ... The first can be compared to the measure of gold; the second is more like a gem”.

Acknowledgments

I am thankful to D. Leites who improved my English in this article.

References

[4]H.S.M. Coxeter, Introduction to Geometry, Cambridge Univ. Press, 1961.
[6]G. Frobenius, Über Matrizen aus nicht negativen Elementen, Sitzungsber. Konigl. Preuss. Akad. Wiss, 1912, pp. 456-477.
[7]J. Kepler, Misterium Cosmographicum, 1596. For full text scan in Latin, see https://www.e-rara.ch/doi/10.3931/e-rara-445, 181 pp.
[11]L. Paccioli, Divina Proportione, 1509. For the English translation of the full text of the original 1498 edition, see https://www.scribd.com/document/244035060/tennenbaum-pacioli-divine-proportion-pdf
Journal
Journal of Nonlinear Mathematical Physics
Volume-Issue
27 - 1
Pages
12 - 16
Publication Date
2019/10/25
ISSN (Online)
1776-0852
ISSN (Print)
1402-9251
DOI
10.1080/14029251.2020.1683961How to use a DOI?
Copyright
© 2020 The Authors. Published by Atlantis and Taylor & Francis
Open Access
This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/).

Cite this article

TY  - JOUR
AU  - Askold M. Perelomov
PY  - 2019
DA  - 2019/10/25
TI  - Remarks on the mass spectrum of two-dimensional Toda lattice of E₈ type
JO  - Journal of Nonlinear Mathematical Physics
SP  - 12
EP  - 16
VL  - 27
IS  - 1
SN  - 1776-0852
UR  - https://doi.org/10.1080/14029251.2020.1683961
DO  - 10.1080/14029251.2020.1683961
ID  - Perelomov2019
ER  -