Quantum transfermatrices for the sausage model
JHE
Quantum transfermatrices for the sausage model
Vladimir V. Bazhanov 0 1 2 5
Gleb A. Kotousov 0 1 2 3 5
Sergei L. Lukyanov 0 1 2 3 4
Australian National University 0 1 2
0 grable Models , Sigma Models
1 Piscataway , NJ 088550849 , U.S.A
2 Canberra , ACT 2601 , Australia
3 NHETC, Department of Physics and Astronomy, Rutgers University
4 L.D. Landau Institute for Theoretical Physics
5 Department of Theoretical Physics, Research School of Physics and Engineering
In this work we revisit the problem of the quantization of the twodimensional O(3) nonlinear sigma model and its oneparameter integrable deformation  the sausage model. Our consideration is based on the socalled ODE/IQFT correspondence, a variant of the Quantum Inverse Scattering Method. The approach allowed us to explore the integrable structures underlying the quantum O(3)/sausage model. Among the obtained results is a system of nonlinear integral equations for the computation of the vacuum eigenvalues of the quantum transfermatrices.
Field Theories in Lower Dimensions; Integrable Field Theories; Lattice Inte

HJEP01(28)
1 Introduction
2 Chiral transfermatrices for the cigar
2.1
Wilson loops for the long sausage
2.2 De nition and basic properties of the chiral transfermatrices
2.3 Basic facts about the quantum cigar
3 Chiral transfermatrix for Zn parafermions
3.1
Bosonization of Zn parafermions
3.2 Discretization of the chiral transfermatrix
4 Spectrum of the chiral transfermatrix
4.1
4.2
4.3
4.4
5.1
5.2
5.3
Operators ( ) ODE/IQFT correspondence for the vacuum eigenvalues ODE/IQFT correspondence for the full spectrum
Operators ( ) and ( ) 4.5 NLIE for the vacuum eigenvalues
5 Integrable structures in the sausage
Basic facts about the quantum sausage
NLIE for the kvacuum eigenvalues in the sausage model
A, B and T
5.4 ODE/IQFT for the sausage model
6 Discussion
A Scaling behaviour of discretized chiral transfermatrices
B Scaling behaviour of Bethe roots for the vacuum state
C Derivation of NLIE
D Modi ed NLIE for oscillating asymptotics E Twoparticle Smatrix for the sausage model
{ 1 {
Introduction
Integrability is a traditional area of mathematical physics having a long history. The notion
is relatively well understood in
nite dimensional systems: it requires the system to have
exactly n isolating and commuting integrals of motion where n is the number of degrees
of freedom. In the context of 1+1 dimensional eld theory, where the continuous number
of degrees of freedom makes the traditional de nition insu cient, a suitable paradigm of
integrability was also discovered. The key ingredient in this case is a Lie algebravalued
world sheet connection involving an analytic spectral parameter, whose atness condition
is equivalent to the classical equations of motion. Since the Wilson loops
T = Tr P exp
Z
C
A
remain unchanged under continuous deformations of the integration contour (see gure 1),
they generate an in nite family of conserved quantities which can be used to solve the eld
theory within the framework of the inverse scattering method. At the end of the seventies
the Quantum Inverse Scattering Method (QISM) was proposed [1]. The approach was
inspired by the pioneering works of Baxter on lattice statistical systems [2, 3] and based on
the study of the common spectrum of the transfermatrices (T operators)  the quantum
counterpart of the classical Wilson loops. The original formulation of the QISM was
restricted to the socalled \ultralocal" models where the elementary transport matrices M n =
P exp Rxxnn+1 A commute for di erent segments of the discretized path. The study of the
algebra of the ultralocal operators M n led to the discovery of new remarkable mathematical
structures collectively known today as the YangBaxter algebras. The most studied class of
integrable models is the one where the YangBaxter algebra of the ultralocal operators M n
admits a
nitedimensional representation. In this case the discretized quantum system
can be interpreted as an exactly soluble statistical model whose solution can be obtained
by means of the Bethe ansatz method. The solution of the continuous QFT is achieved by
taking a proper scaling limit. An archetype of this scenario is the sineGordon model, while
the corresponding statistical system is known as the inhomogeneous 6vertex model [4].
However the QISM fails when it is applied to many interesting systems including
classically integrable Nonlinear Sigma Models (NLSM). The origin of the problem can be
illustrated by the O(3) NLSM governed by the Lagrangian
relations [5, 6]
[ta; tb] = i "abc tc :
{ 2 {
(1.1)
where the 3dimensional vector n is subject to the constraint n
n = 1. In this case
the lightcone components of the at connection, A
21 (At
Ax), are given by the
1
2
A
=
n = nata 2 su(2) ;
@x) while n is built from the components of the unit vector n and ta:
x
x + R
where C(i)(x) are local elds built from na and @tna and take values in the tensor product
su(2)
su(2). Because of the derivative of the function, relations of this type are usually
referred to as nonultralocal. The presence of such terms creates serious problems with the
rigorous proof of Poisson commutativity of the conserved charges (1.1) for di erent values
of the spectral parameter. This, in turn, hampers the rstprinciples quantization of the
model.
The problem with nonultralocality is a general issue for integrable NLSM and is a
major obstacle which prevents the incorporation of these models into the framework of the
QISM. This was rst observed in attempts to quantize the principal chiral eld and O(N )
models. In the case of the O(4)model (i.e., the model (1.2) with n 2 S3), Polyakov and
Wiegmann [7] and later Faddeev and Reshetikhin [8] managed to bypass the problem with
nonultralocality. In both works, the NLSM was replaced by a di erent model satisfying
the ultralocality condition; Polyakov and Wiegmann considered a model with Nf fermion
avors, whereas Faddeev and Reshetikhin focused on a certain spinS chain. They studied
the thermodynamics using the Bethe ansatz technique and gained valuable results for the
O(4) sigma model through the large Nf and S ! 1 limit, respectively. Both limiting
procedures yielded the same system of Thermodynamic Bethe Ansatz Equations (TBA),
which was then justi ed by a comparison with perturbative calculations and the exact
results from the Smatrix bootstrap [9]. In the next 30 years an enormous amount of TBA
systems were discovered. However, in spite of signi cant achievements in their study, the
original problem of the construction of quantum transfermatrices (including the calculation
of their spectrum) has remained unsolved even for the O(3) and O(4) NLSM.
{ 3 {
HJEP01(28)
found that such a gauge also exists for the theory described by the Lagrangian
This theory also belongs to the class of NLSM, and, assuming 0 <
< 1, the
corresponding target space is topologically the twosphere. As
! 0 one can neglect the second
term in the denominator which results in the O(3) Lagrangian multiplied by . Hence, the
model (1.6) is the one parameter deformation of the O(3) sigma model. It is colloquially
known as the \sausage model" since for
! 1 the target manifold can be pictured as a
long sausage with length / log( 11+ ) (see gure 2). The sausage model was introduced in
the work [14] where strong evidence for its quantum integrability was presented. Its zero
curvature representation was found in ref. [15]. Remarkably, with the proper gauge
transformation, the at connection can be chosen to satisfy the ultralocal Poisson bracket relations
A (xj ) ; A (x0j 0) =
A (xj )
A (x0j 0); r( = 0) (x
x0)
1
(1.6)
(1.7)
(1.8)
(1.9)
Here
A (xj ) ; A (x0j 0) = 0 :
r( ) =
1
1
1It is worth noting that a transition to an ultralocal gauge has previously been utilised for some other
It is straightforward to check now that the formulae2
a( ) = f ( ) t3 + i g( ) t1 ;
f ( ) =
g( ) =
(1
(1
(1
)
)
)
p
) canonically conjugate to (Q; Q ) are given by the relations
tanh 12 (Q + Q ) . Let us also introduce the shortcut notations
3
=
=
2(
1
1
3
and the eld independent element of sl(2; C)
HJEP01(28)
de ne the at connection, i.e.,
satisfying the ultralocality conditions (1.7).
Having an explicit formula for the ultralocal at connection we may turn to the
construction of the timeindependent Wilson loops. In fact their construction requires some
assumptions to be made. First of all in writing eq. (1.1) it is assumed that the connection
is single valued on the spacetime cylinder. In the case under consideration this can be
achieved by imposing periodic boundary conditions on the unit vector n. Such boundary
conditions look natural for the O(3) NLSM, since they preserve the global invariance of
the Lagrangian. In the case of the sausage model, the O(3) symmetry is broken down to
U(1) and we will consider more general boundary conditions
n3(t; x + R) = n3(t; x) ;
n (t; x + R) = e 2 ik n (t; x)
(n
n1
in2)
(1.16)
2Notice that the connection in the ultralocal gauge takes values in the complexi cation of the Lie algebra
su(2). This should not worry us because only reality conditions imposed on gauge invariant quantities are
a subject of interest and importance. In what follows we'll prefer to use the CartanWeyl generators
h = 2t3; e
= t1
it2:
instead of ta (1.4).
[h; e ] = 2 e ;
[e+; e ] = h
{ 5 {
which depend on the twist parameter k
k + 1. In terms of the real and imaginary part
of the complex eld Q,
Q =
+ i ;
the twisted boundary conditions are equivalently given by
(t; x + R) = (t; x) ;
(t; x + R) = (t; x) + 2 k :
As it follows from eq. (1.14) the connection is not single valued on the spacetime cylinder
now,
charges
A (t; x + R) = ei kh A (t; x) e i kh ;
(1.19)
however with a little modi cation one can still introduce the time independent conserved
Tj ( ) = Trhe i kh M j ( ) i ;
Z
C
M j = j P exp
dx Ax :
(1.20)
Notice that the de nition of the Wilson loop requires a choice of the representation of the
Lie algebra and the subscript j in (1.20) labels the representations j of sl(2). In our case we
will focus on the nite dimensional irreps, so that in the standard convention j = 12 ; 1; : : : .
The ultralocal Poisson structure (1.7) implies that the monodromy matrix generates
the classical YangBaxter Poisson algebra:
M j ( ) ; M j0 ( 0)
=
M j ( )
M j0 ( 0) ; rjj0 ( = 0) ;
where rjj0 = j
j0 [ r ]: This, supplemented by the easily established property
implies the in nite set of relations
e i kh
e i kh ; r( = 0) = 0 ;
Tj ( ); Tj0 ( 0) = 0 :
Thus, we have arrived at formulae (1.21) and (1.23) which are key ingredients in the
Hamiltonian approach to the inverse scattering method [16, 17].
When faced with the problem of quantizing the sausage NLSM, a simple idea that may
come to mind is to discretize the pathordered integral onto N small segments. Due to
ultralocality, the N elementary transport matrices j hP exp Rxxnn+1 dx Axi satisfy the same
type of Poisson bracket relation as (1.21) and Poisson commute for di erent segments.
These relations can be formally quantized leading to a certain quantum YangBaxter
algebra. The major problem now is to construct a suitable representation of this abstract
algebraic structure. In the case under consideration, the representation is, in all
likelihood, in nite dimensional even for nite N . At this moment, it is not clear for us how to
construct and handle such representations, let alone take the scaling limit with N ! 1.
In this work, we will try to avoid discretization as much as we can and mostly
follow the socalled BLZ approach  the variant of the QISM developed in the series of
works [11, 18, 19]. For integrable Conformal Field Theories (CFT), it was demonstrated
{ 6 {
(1.17)
(1.18)
(1.21)
(1.22)
(1.23)
that the T operators can be constructed without any discretization procedure. Later it
was observed that many deep properties of representations of YangBaxter algebras in
integrable CFT can be encoded in the monodromies of certain linear Ordinary Di erential
Equations (ODE) [20{26]. These results were extended to massive Integrable Quantum
Field Theories (IQFT) [27] (for recent developments, see also refs. [28{35]). The general
relation of this type will be referred to in the paper as the ODE/IQFT correspondence.
Broadly speaking, the ODE/IQFT correspondence means that for a given IQFT the
eigenvalues of the quantum T operators are identi ed with certain connection coe cients
for the system of equations,
D( )
= 0 ;
D( )
which is found to be a function of the original spectral parameter from the
quantum theory. The system of ODE can be then interpreted as an auxiliary linear
problem, whose compatibility condition, [D( ); D( )] = 0, coincides with the zerocurvature
representation for some classically integrable eld theory. Thus the ODE/IQFT
correspondence reduces the calculation of the spectrum of quantum transfermatrices to a certain
problem in the theory of classical integrable equations. The latter can be e ectively treated
by the inverse scattering transform method. This makes the ODE/IQFT correspondence
a very powerful tool. In particular, it gives a practical way to make progress in the
conceptual long standing problem of the quantization of integrable NLSM. The ultimate goal
of this work is to demonstrate this for the case of the quantum sausage model.
2
Chiral transfermatrices for the cigar
The BLZ approach [11, 18, 19] begins with an analysis of the RG xed point which controls
the ultraviolet behaviour of the integrable QFT. With this in mind, let's take a quick look
at the sausage NLSM (1.6). In the traditional pathintegral quantization, the model should
be equipped with a UV cuto
. A consistent removal of the UV divergences requires that
the \bare" coupling in the Lagrangian (1.6) be given a certain dependence on the cuto
momentum, i.e.,
= ( ). To the rst perturbative order the RG ow equation is given by [14]
(2.1)
(2.2)
where ~ stands for the (dimensionless) Planck constant. Integrating this equation leads to
where
changing
= ~ + O(~2). The energy scale E is an RG invariant (i.e., it's kept xed with
), so that
! 1 as
! 1. Having in mind the quantization of the model, this
simple analysis shows that the classical eld theory (1.6) deserves special attention when
is close to one.
1 +
(1
2) + O(~2) ;
= (E = ) ;
{ 7 {
t
0
0
R
x
HJEP01(28)
replaced by an integration contour along the characteristics: x
arrow) and x+ = t0 + R with t0 < x
< t0
R (blue arrow).
= t0 with 0 < x+ < t0 + R (red
2.1
Wilson loops for the long sausage
Consider the conserved charges Tj ( ) (1.20) for 1
1. If the integration contour
appearing in the de nition (1.20) is chosen at the time slice t = t0, Tj ( ) are expressed in
terms of the pair of real variables ( ; ) (1.17) and the corresponding canonically conjugate
momenta. We can use the magic of the zerocurvature representation to reexpress Tj ( ) in
terms of the values of the elds (x; t) and (x; t) on the characteristics x+
t + x = t0 + R
and x
t
x = t0,
+(x+) = jx =t0 ;
(x ) = jx+=t0+R ;
+(x+) =
(x ) =
jx =t0
jx+=t0+R :
Indeed, the original integration along the time slice t = t0 in (1.20) can be replaced by
the pathordered integral over the contour glued from two lightcone segments as shown in
gure 3. Denoting the lightcone values of the connection as
A+(x+) = A+(t; x)jx =t0 ;
A (x ) = A (t; x)jx+=t0+R ;
and taking into account the boundary conditions (1.18), one can rewrite eq. (1.20) in the
form
where
Tj
Tr e i kh M j
= Tr
Z t0 R
t0
j (M) ;
This formula is a convenient starting point for the quantization procedure, though one
should have in mind that the lightcone elds are complicated functionals of the canonical
set at t = t0. In general, they are not independent variables and their Poisson brackets
(2.3)
(2.4)
(2.5)
(2.6)
{ 8 {
Here we use the notations
Lj( +) = j P exp
j( +)
1
i +
2
+
where
and
Notice that the rmatrix in the classical YangBaxter Poisson algebra (1.21) depends only
on the ratio of the spectral parameters and is independent of . Hence, we can conclude that
+li!m1xed
Tj = Trh Lj( +) e P1h i :
P1 =
+(t0 + R)
+(t0) :
where the variables
are de ned by the relations
HJEP01(28)
+
=
+=
and (h; e+; e ) are the CartanWeyl generators (see footnote 2 on page 5). Then the
connection component A+(x+) only depends on the parameter
+, while A (x ) only
depends on
assume that
. According to (2.8) at least one of the ( +;
) vanishes as
! 0 and + is kept xed. Then, it is easy to see that
! 1. Let us
are not known explicitly. Fortunately, we can overcome this problem when considering
the limit
quantities
+
lightcone elds
a(
1) (1.12) in the form
and
! 1. In this limit the at connection (1.14) simpli es considerably, since the
12 @xQ are then linearly expressed through the
. It is convenient to write the constant Lie algebra elements
a
Lj( +) ; Lj0( 0+) = Lj( +)
Lj0( 0+) ; rjj0( += 0+) ;
and also that the j are in involution:
f j( +); j0( 0+)g = 0 :
For 1
1 the target space of the sausage model consists of two cigars glued
together, whose tips are separated by a distance / log( 11+ ) (see
gure 2). In the
! 1
limit, the sausage is broken down into two disjoint halfin nite Hamilton's cigars [36]. Let
us focus on one of the cigars, say the left one. The Lagrangian of the NLSM which has this
target space follows from the sausage Lagrangian (1.6): one should express the sausage
Lagrangian in terms of the elds
and , perform the constant shift
and nally take the limit
! 1. As a result one nds [37, 38]
!
L =
e
2
2(1 + e2 )
(2.15)
This suggests that the j introduced through the limiting procedure (2.9), can be
interpreted as conserved quantities in the cigar NLSM (2.15). The
elds
and
which
appear in the construction of Lj (see eqs. (2.11), (2.10)), should be now understood as the
lightcone values of the elds in the cigar NLSM, in the case when the eld
becomes large,
i.e., in the asymptotically
at domain of the target manifold. The cigar equations of
mo0
0, whose general solution can be written as
+(x+) +
(x ) and
+(x+) +
(x ). Taking the time slice t = t0 where the eld
becomes large, one nds
1
2
1
2
f
(x); 0 (y)g =
(x
y) ; f 0 (x); 0 (y)g =
0(x
y)
(2.16)
f 0 (x); 0 (y)g = f 0 (x); 0 (y)g = f
(x); 0 (y)g = f
(x); 0 (y)g = 0 ;
where the prime stands for the derivative w.r.t. the argument. Recall that the j are time
independent charges, and their value does not depend on the choice of t0. Hence, we come
to the conclusion that the matrix Lj (2.10), built from the elds +(x), +(x) satisfying the
Poisson bracket relations (2.16), will obey the \rmatrix" Poisson bracket algebra (2.13).
The following comment is in order here. It is not di cult to see that the matrix Lj
does not change under the constant shifts + !
relations (2.16) can be lifted to
+ + const so that the Poisson bracket
f +(x); +(y)g =
(x
y) ;
1
4
(2.17)
where (x) = 2n + 1 for nR < x < (n + 1)R; n 2 Z. The later are consistent with the
quasiperiodicity condition (2.12).
2.2
De nition and basic properties of the chiral transfermatrices
We may now turn to the problem of quantization. The quantum counterpart of the
conserved charges (2.9) will be referred to as the chiral transfermatrices and in what follows,
will be denoted by the same symbol j. Their construction almost identically follows the
steps elaborated in refs. [11, 19] in the context of the quantum KdV theory. Here we
present them very brie y, referring the reader to those works for detailed explanations.
First of all we should \quantize" the Lie algebra sl(2), so that h, e
are understood
now as the generators of the quantum universal enveloping algebra Uq sl(2) :
[h; e ] =
2 e ;
[e+; e ] =
(2.18)
q
h
q
q h
q 1 ;
i~
where q = e 2 . Consequently the symbol j will stand for the (2j + 1)dimensional
representation of the quantum algebra. Instead of the Planck constant ~, for convenience we
will use the parameter n:
matrix
~
2
n
;
q = e n :
i
The quantum counterpart to Lj is the following (2j + 1)
(2j + 1) operator valued
Lj ( +) = j P exp i +
dx
V + q 2h e+ + V
q 2h e
e
P1h :
(2.20)
where c are some constants and
with
V (x) =
'+(x) = Q1 +
#+(x) = Q2 +
R
R
2 x p
n P1 + i X am e 2 Rim x
2 x p
n + 2 P2 + i X bm e 2 Rim x
;
Fp (\Fock space") be the highest weight module of the Heisenberg
algebra (2.23) with the highest weight vector j p i de ned by the equations
P1 j p i =
pn1 j p i ;
P2 j p i =
p2
n + 2 j p i :
V (x) :
Fp1;p2 7! Fp1 i;p2 ;
and therefore the matrix elements of Lj ( ) are operators in
sion (2.20) contains the ordered exponential which can be formally written in terms of a
1
m=
1Fp1+im;p2 . The
expres(2.19)
(2.21)
(2.22)
(2.23)
(2.24)
(2.25)
(2.27)
(2.28)
It is easy to see that power series in as
Lj ( +) = j
where
However, since
1
X (i +)m Z
m=0
dxm
dx1 K(xm)
K(x1) e
P1h ; (2.26)
t0+R>xm>:::x1>t0
K(x) = V + q 2h e+ + V
q 2h e :
V (x2)V (x1) x2!x1+0
a
commutation relations
2) c+c , the integrals in (2.26) diverge. As explained in [19], the
V 1 (x1)V 2 (x2) = q2 1 2 V 2 (x2)V 1 (x1) ;
x2 > x1
( 1;2 =
)
(2.29)
allow one to reexpress the integrals in (2.26) in terms of two basic contour integrals
X0 =
q
1
This procedure yields an unambiguous de nition of the ordered exponential in (2.20) for
n 6= 2; 4; 6 : : : The case of even n needs some special attention and we will return to it later.
Notice that the above analytical regularization of the Pordered exponential (2.20) is
not applicable at n ! 1. This limit can be studied within the renormalized
perturbation theory in ~ = 2n . As usual, the perturbative expansion will involve the counterterms
depending on the UV regulator. At ~ = 0 the counterterms give rise to anomalous
contributions which remain nite in the limit where the UV regulator goes away. This makes the
classical limit not entirely straightforward, however if the anomalous terms are properly
taken into account, the formula (2.20) precisely reduces to its classical version (2.10) in the
limit ~ ! 0. The relevant calculations will be presented elsewhere.
The operator valued matrices Lj (2.20) are designed in such a way that, for arbitrary
chosen constants c and t0, they obey the quantum YangBaxter algebra
Rjj0 ( 0+= +) L( +)
where the matrix Rjj0 ( ) is the trigonometric solution to the YangBaxter equation which
acts in the space j
j0 . In particular
R 1 1 ( ) = BBB
2 2
1
q 1
0
0
The proof of eq. (2.31) follows that from the work [19].
The chiral transfermatrices, de ned similar to (2.9),
satisfy the commutativity condition
as a simple consequence of (2.31). Notice that the chiral transfermatrices act inside a single
Fock space, whereas the same is not true for an arbitrary element of Lj ( ). Furthermore,
the Fock space Fp naturally splits into the nite dimensional \level subspaces"
Fp =
L1=0 Fp(L) :
L Fp
(L) = L Fp(L) ;
where the grading operator is given by
h
j ( +) = Tr Lj ( +) e
P1h i ;
[ j ( +); j0 ( 0+) ] = 0 ;
1
X
m=1
L = 2
a mam + b mbm :
(2.32)
(2.33)
(2.34)
(2.35)
(2.36)
Using the relation,
one can show (see appendix C from [19]) that the j ( +) commute with the grading
operator, and therefore, act invariantly in each nitedimensional level subspace:
j ( +) :
Fp 7! Fp(L) :
(L)
( +) y = (
+) :
The Fock space Fp can be equipped with an inner product consistent with the Hermiticity
conditions aym = a m ; bym = b m imposed on the Heisenberg operators (2.23). It is not
di cult to show that for real p21; p22 and 2 , ( +) is a Hermitian operator and
+
Notice that the commutativity with the grading operators can be interpreted as the
independence of the chiral transfermatrix on the arbitrary chosen constant t0. It turns out
that they further do not depend on the constants c appearing in the de nition of the vertex
operators V
(2.21). Also, a simple dimensional analysis shows that the spectral parameter
+ and R occur in the chiral transfermatrix through the combination
2+R n2 only. It is
convenient to introduce a dimensionless spectral parameter
by means of the relation
2 =
converges in the whole complex plane of 2 and de nes an entire function with an essential
singularity at
2 = 1. The asymptotic expansion near the essential singularity is of
primary interest. It can be written as
( ) = exp
2
sin( 2n ) (
2) n2
~ i (
2
n
) 2(n+2) ;
j
qj+ 21
1 ( ) = j+ 12
2
q
j + j 12
qj+1 ;
supplemented by the condition 0 = 1. In what follows, we will mostly focus on the
fundamental transfermatrix and use the notation
2
1 . The integrable structures associated
with the commuting family of operators j ( ) were already studied in the context of the
socalled paperclip model  an integrable model with boundary interaction [43]. Here for
convenience we make a short summary of some basic properties of the operator ( ).
For arbitrary complex p = (p1; p2), the operator ( ) 2 End Fp is an entire function
of 2 in the sense that all its matrix elements and eigenvalues are entire functions of this
variable. Thus the power series
and treat the chiral transfermatrices as functions of this variable rather than the
dimensionful +.
socalled fusion relation [39{41]
The chiral transfermatrices are not independent operators for di erent values of j =
where ~ is a formal power series of the form
~(~)
2 cos
2 p2
1
m=1
+
X ~tm ~2m :
(2.44)
This asymptotic expansion can be applied for arbitrary complex p = (p1; p2) and n 6=
2; 4; 6 : : : . Furthermore, in the case n
1 it holds true for j arg(
2
)j < .
The expansion coe cients in (2.42) and (2.44) form two in nite sets of mutually
commuting operators. Using the terminology of the work [18], we will refer to tm
~
tm
and
Remarkably, the formal power series ~(~) can be written in a form similar to (2.33).
1
m=1 as the nonlocal and dual nonlocal Integrals of Motions (IM), respectively.
1
m=1
Namely [43],
+
1
2
i 2) are the conventional Pauli matrices and the vertex operators
(x) =
'0+(x) +
is equal to 1 + n+12 and we assume here that they are normalized
(2.47)
(2.48)
(2.49)
where 3
;
in such a way that
t1(p1; p2) =
and
t~1(p1; p2) =
(2.46)
Because of the divergencies, the path ordered exponential in (2.45) should be understood
in the same manner as (2.20), i.e., the formal expansion in a power series of ~
+ should
be rewritten in terms of the basic contour integrals similar to (2.30). With this analytical
2
regularization the r.h.s. of eq. (2.45) becomes a formal power series in ~2+R n+2 with
unambiguously de ned expansion coe cients. Up to a factor similar to that in (2.40), this
combination can be identi ed with ~2 in eq. (2.44):
~2 =
For future reference we present here explicit formulae for the \vacuum" eigenvalues of the
operators t1 and ~t1 corresponding to the highest weight vector j p i 2 Fp (2.24):
( 12 + n1 )
(1 + n1 )
n + 2
n
2
+
4p22
1 + 4p21
2
( 1+2ip1 ) (
n
1 2ip1 )
n
n + 2 n+2
2
p
(
1
2
(1
n+12 )
n+12 )
n
n + 4
1
4p21
4p22
2
( 1n+2p22 ) ( 1n++2p22 ) :
An e cient integral representation for calculating the vacuum eigenvalue t~2(p1; p2) can be
found in appendix A of ref. [43].
For even n, the chiral transfermatrices require some careful handling. In this case
( ) can be de ned through the limiting procedure
( )jn=2l = lim exp
!0
4
2l
( ) n=2l+
so that the asymptotic formula (2.43) should be substituted by
(l = 1; 2; 3 : : :) ;
(2.50)
( )jn=2l = exp 2 n log(
2
)
~ i (
2
) 2(l+1)
l
n=2l
:
(2.51)
The formulae (2.43), (2.44) are not valid for positive real 2. In order to describe the
asymptotic behaviour for 2
through the relation
1
IM (2.44) by the set g~m m=1 which are algebraically expressed in terms of the former
! +1, it is convenient to substitute the set of dual nonlocal
2 cos(2 P2) +
X ~tm zm = 2 cos(2 P2) exp
1
X g~m zm
:
m=1
1,
as
2
! +1 ;
nm
z n+2 + O(z 1)
(2.52)
(2.53)
(2.54)
Then, for arbitrary complex p = (p1; p2), n 6= 2; 4; 6 : : : , n
( ) = 4 cos
eH( 2) cos G( 2)
where
1
m=1
2 p2
n + 2
n
2
1
X
m=1
H(z)
2 cot
G(z)
n
2 z 2 +
n
z 2 +
g~m sin
1
X
m=1
g~m cos
2 m
n + 2
2 m
n + 2
nm
z n+2 + O(z 1) :
For even n, the rst term in the formal power series H(z) should be replaced by 2z n2 log(z).
2.3
Basic facts about the quantum cigar
In the previous subsection, we described the formal algebraic construction of the chiral
transfermatrices. Here we brie y discuss how j ( ) can be understood as operators in the
quantum cigar NLSM (for more details on the quantum cigar see, e.g., ref. [44]).
The cigar NLSM was introduced before at the classical level by means of the
Lagrangian (2.15). In the classical eld theory, it is natural to consider the following
scattering problem. Suppose that at t !
1 we are given the eld con guration within the
asymptotically at domain of the target manifold, i.e.,
(t; x)jt! 1
(t; x)jt! 1
0
(in) +
0
(in) +
4
R
2
R
1
P (in) t +
(kx + k~t) +
X
m6=0
i
m
X
m6=0
i
m
a(min) e 2 Rim (t+x) + a(min) e 2 Rim (t x) (2.55)
b(min) e 2 Rim (t+x) + b(min) e 2 Rim (t x) :
P (in)
1
the string approaches the tip, scatters and then escapes back to the at region. After the scattering
process the zero mode momentum changes sign.
In writing this equation, we took into account the boundary conditions (1.18). Also,
the constant k~ is the conserved charge for the Noether U(1)current associated with the
Lagrangian (2.15). The set, A
(in) = f (0in); P (in); 0(in); k~; a(min); b(min)g, can be interpreted
1
as a classical \instate" for a string propagating on the target manifold (see gure 4). The
nontrivial interaction occurs at some nite time when the elds take values in the vicinity of
the tip of the cigar. After scattering at the tip, as t ! +1, the eld con guration returns
to the asymptotically at domain and takes the same form as in the r.h.s. of (2.55) with the
instate A
(in) replaced by the outstate A
(out) = f 0
(out); P1(out);
(out); k~; a(mout); b(mout) .3
0
g
The classical scattering problem can be formulated as the problem of nding the canonical
transformation which maps A
(in) to A
in nite sets of left and rightcurrents [45], i.e.,
(out). It turns out that the theory possesses two
HJEP01(28)
(2.56)
(2.57)
(2.58)
k(in) 2 Z.
(s = 2; 3; : : :) ;
so that the classical dynamics of the elds are strongly constrained. In particular, the
magnitude of the zeromode momentum remains unchanged after the scattering (see gure 4),
1
P (out) =
P (in) :
1
Consider now the quantum theory. First of all we note that the value of the U(1)
charge is quantized so that (n + 2) k~ = m 2 Z. Thus the space of states of the quantum
theory splits into orthogonal subspaces Hk;m labeled by the twist parameter k and the
integer m. The quantum theory still possesses the chiral currents satisfying eqs. (2.56). As
a result, Hk;m can be decomposed into the highest weight irreps of the W algebra, W
W
generated by the elds Ws and Ws [45]. Let Vh (Vh) be the highest weight representation
of the chiral W algebra W (W) labeled by the highest weight h (h). Then, schematically,
The highest weight h can be chosen to be a pair of numbers ( ; w), where
coincides
with the conformal dimensions of the highest weight vector, while w is the eigenvalue of
3Strictly speaking, the winding number k is only conserved modulo an integer, i.e., k(out)
Here we ignore this and assume that k(out) = k(in) 2 ( 12 ; + 12 ].
the dimensionless conserved charge R2 R0R dx W3(x), and similar for h. Let us rst focus
on the \left" component Vh in the tensor product Vh
Vh.
It should be clear that the quantum counterpart to the left components of the
inasymptotic elds (2.55) can be identi ed with the elds '+ and #+ given by (2.22). Since
the quantum
elds Ws are chiral currents, i.e. Ws(t; x) = Ws(t + x), they can be expressed
in terms of the asymptotic elds '+ and #+. Indeed, for given s, Ws is a certain orders
homogeneous polynomial with constant coe cients w.r.t. the elds '0+, #0+ and their higher
derivatives (in other words, any monomials appearing within Ws contains exactly s
derivative symbols). This implies that the Fock space Fp, which is the space of representation
for the elds '0+, #0+, possesses the structure of the highest weight representation of the
chiral W algebra. It turns out that for real p, the Fock space Fp coincides with irrep Vh
as a linear space, provided that h = ( ; w) is related to p = (p1; p2) as follows
4n
p21 +
3n + 2
3(n + 2) 2
p
2
2n + 1
12
:
Z
p1<0
In fact, one can use these formulae to conveniently parameterize the highest weight h by the
pair (p1; p2): Vh
Vp1; p2 . With this notation, a more accurate version of eq. (2.58) reads as
1
2
1
2
p2 =
m + (n + 2) k ;
p2 =
m
(n + 2) k :
The direct integral in (2.60) does not include the domain with positive p1, since, as follows
from eqs. (2.59), Vp1; p2
V p1; p2 .
A basis of inasymptotic states in Hk;m is formed by
a(inm)1 : : : a(in)
mN
a(inm)1 : : : a(in)
mN
b(inm)1 : : : b(in)
mM
b(inm)1 : : : b(in)
mM j vac i
and can be identi ed with the states from the tensor product of the Fock space Fp1; p2
a m1 : : : a mN a m1 : : : a mN b m1 : : : b mM b m1 : : : b mM j p1; p2 i
j p1; p2 i :
(2.63)
Similarly for the outstates, one has
a(omut1) : : : a(omutN) a
(omut1) : : : a(omutN) b
(omut1) : : : b(omutM) b(omut1) : : : b(omutM) j vac i
a m1 : : : a mN a m1 : : : a mN b m1 : : : b mM b m1 : : : b mM j
p1; p2 i
j
p1; p2 i :
Usually, the Smatrix is introduced as a unitary operator which relates the in and
outasymptotic bases. In the case under consideration, the Smatrix can be interpreted as the
intertwiner acting between the Fock spaces:
S^ :
Fp1; p2
Fp1; p2 7! F p1; p2
F p1; p2 :
(2.59)
(2.60)
(2.61)
(2.62)
(2.64)
(2.65)
It turns out that the operator S^ has the following structure
S^ = S0(p) S^L
S^R ;
(L) (L)
where S^L intertwines the level subspaces, S^L : Fp1; p2 7! F p1; p2 , and is normalized by
the condition S^L j p1; p2 i = j
p1; p2 i, and similarly for S^R.
construction of the operators S^L;R is a straightforward algebraic task. The more delicate
For a given level `, the
problem is
nding the overall scalar factor S0(p). It was obtained in the minisuperspace
approximation in ref. [46]. The exact form of S0(p) has been known since the unpublished
work of the Zamolodchikov brothers [47].4
Returning to the chiral transfermatrices, let us note that these operators should act
in the Hilbert space of the quantum cigar, and therefore their action should commute with
the intertwiner S^:5
S^ ( ) = ( ) S^ :
In practice, this condition implies that all matrix elements of the (dual) nonlocal IM in the
basis of Fock states (2.63) are even functions of p1 (for illustration see eqs. (2.48), (2.49)).
The quantum cigar also possesses an in nite set of the socalled local IM acting in Hk;m.
To get some feeling for these operators, we need to remind ourselves of an important feature
of the model. Namely, it admits an equivalent \dual" description in terms of the socalled
sineLiouville model. The dual Lagrangian is given by [47]
(2.66)
(2.67)
(2.68)
(2.69)
(2.70)
can
L
(dual) =
1
4
pn' cos
p
n + 2 # ;
with the sineLiouville elds satisfying the boundary conditions
'(t; x + R) = '(t; x) ;
#(t; x + R) = #(t; x) + p
2 m
n + 2
:
Notice that the \coupling" M is a somewhat fake parameter of the Lagrangian  by
an additive shift ' 7! ' + const the value of M can be chosen to be any real number.
Nevertheless, it is convenient to keep it unspeci ed.
To understand the relation between the elds in the NLSM and its dual description,
let us take the \zeromode" of the eld '
'0 =
Z R dx
0
R
'(x) ;
potential term in the action (2.68) can be neglected and p'n
and consider the region '0 ! +1 in con guration space. In this asymptotic domain, the
#
+ const, while pn+2
be identi ed with ~  the eld from the cigar NLSM de ned by the relation J =
where J stands for the Noether U(1)current.
4Although Zamolodchikov's notes have never been published, they were broadly distributed within the
scienti c community.
5The intertwiner S^ should not to be confused with the so called \re ection" operator R^ : Fp1; p2
Fp1; p2 7! F(Lp)1; p2
ers ^L : Fp1; p2 7! F
(L)
Fp1; p2 , and [R^; ( )] = 0. Note that R^ = ^ S^ where ^ = ^L ^R and the chiral
intertwinp1; p2 are de ned by the conditions ^Lam =
am ^L; ^Lbm = +bm ^L; ^L j p1; p2 i =
j p1; p2 i, and similar for ^R (see, e.g., [44]).
is 1 =
R
2
s 1 Z R dx
It turns out that for any even s = 2j there exists a local density (de ned modulo the
addition of a total derivative and an overall multiplicative constant) such that i2j 1 is an
integral of motion and satis es the commutativity conditions
[ i2j 1; ( ) ] = [ i2j 1; i2j0 1 ] = 0 :
These operators are referred to as the (chiral) local IM. They were studied in ref. [42],
where the explicit form for the rst local IM and their vacuum eigenvalues, i2j 1(p1; p2)
for j = 1; 2; 3, can be found. Here we only note that for any j = 1; 2; : : :
the numerical coe cients Cl(ms) can be written as
where the dots stand for monomials which include higher derivatives of @+' and @+# and
Cl(mj) = C2j 1
( 2)j+1(2j
2)! (n + 2)( 12
(j + 1)!
j) l ( n)( 12
l! m!
j) m ( n)l 1 (n + 2)m 1
:
Here (a)m = Qim=01(a + i) is the Pochhammer symbol. The overall normalization constant
C2j 1 is usually set to
C2j 1 =
2 3j (j + 1)! n(n + 2)
(n + 2)( 12
j) j ( n)( 12
j) j
:
The twist parameter k has a natural interpretation in the dual description  it can be
identi ed with the socalled quasimomentum. The sineLiouville Lagrangian is invariant
2
under the transformation # 7! # + pn+2 . Due to this periodicity, the space of states of
the theory with the boundary conditions (2.69), splits on the orthogonal subspaces Hk;m
such that for any state j A i 2 Hk;m, the corresponding wave functional
A[ '(x); #(x) ]
transforms as
2
A '(x); #(x) + p
= e2 ik
A[ '(x); #(x) ] :
higher derivatives. All such elds are periodic in x, so that one can introduce the integral,
HJEP01(28)
(2.71)
(2.72)
(2.73)
(2.74)
(2.75)
(2.76)
P2j =
X
l+m=j
3
Chiral transfermatrix for Zn parafermions
While quantizing the sausage model within the BLZ approach, we have run into the problem
of nding the spectrum of the chiral transfermatrices for the cigar NLSM. As it has been
explained, we can consider ( ) as an operator acting in the Fock space Fp with real
p = (p1; p2). From the formal point of view, the same spectral problem can be posed for
any complex values of p. Notice, that for real p2, 2 and pure imaginary p1, the operator
One of the most e ective methods for the calculation of the spectrum of commuting
families of operators including the transfermatrices in integrable quantum
eld theory is
based on the ODE/IQFT correspondence. From our study of the parafermionic
transfermatrix, we proposed the ODE counterpart in the correspondence for the cigar NLSM. It
turns out to be identical to that which was introduced earlier in the context of the socalled
paperclip model in the works [42, 43]. Based on the results of these papers, we derived
nonlinear integral equations for determining the vacuum eigenvalues of the chiral
transfermatrix which work both for the cigar and the parafermionic regimes. We believe that this
might be a good starting point for applying the powerful fermionic methods [92{96] to the
sausage/O(3) NLSM.
In refs. [27, 58], a conceptual explanation was given of how the ODE/IQFT
correspondence for integrable conformal eld theory can be generalized to the massive IQFT.
Following this route, we extended the ODE/IQFT correspondence from the cigar to the
sausage NLSM. With the correspondence one can uncover the basic integrable structures
by studying the properties of the connection coe cients of the ordinary di erential
equations. The main result of this paper is the list of properties (i){(x) from section 5.3 for
the commuting families of operators in the sausage model which includes the quantum
transfermatrix. The technical result that deserves to be mentioned is the system of NLIE
which describes the vacuum eigenvalues of the commuting families of operators. Among
other things, it allows one to calculate the kvacuum energies of the sausage/O(3) NLSM.
There are many results in the literature concerning the energy spectrum of the O(3)
sigma model in the sector with k = 0 [76, 97{99]. In ref. [14] a system of TBA equations
was proposed which allows one to calculate the ground state energy for k = 0 and integer
values of the dimensionless coupling n
3 of the sausage model. Recently Ahn, Balog
and Ravanini [70] transformed this system of TBA to a system of three non linear integral
equations which, it is a rmed, works for any real positive n. Their main assumption
is a periodicity condition for the Qfunction given by eq. (3.16) from that paper. In our
investigations, we did not nd any trace of a Qfunction satisfying such a strong periodic
condition. Nevertheless, the numerical results presented in
gure 2 from that paper seem
to be in agreement with the data obtained from the solution of our NLIE (5.30), (5.31)
with k = 0 and n = 1. This situation needs to be clari ed.
Let us brie y touch on some problems which have not been discussed in the text
but are directly related to the subject of this work. We did not make any mention of
the sausage model with the topological term equal to
which is also expected to be an
integrable QFT [14, 100]. Another closely related model is the socalled 3D sausage model
introduced by Fateev in [80]. In the work [15], classical integrability was established for
a fourparameter family of NLSM with torsion which includes the 3D sausage as a two
parameter subfamily. We believe that extending the ODE/IQFT approach to these models
will be useful, both as a step in the development of the method, and in terms of applications.
Some results in this direction have already been obtained in ref. [81]. There are also the
remarkable works [101{103] on toroidal algebras, which are deeply connected to this eld.
All the models mentioned above are based on the sl(2)algebra and its associated
integrable structures. Since the work of Klimcik [104] there has been increasing interest
2
3
4
5
0
p
2
48
0:0546105
0:0661040
0:0683646
n
6
7
8
9
10
0:0658731
0:0613178
0:0561029
0:0509101
0:0460445
in \deformed" integrable NLSM associated with higher rank Lie algebras [105{108]. The
rst principle quantization of such theories seems to be a very interesting problem. In
the recent works [109, 110], an important step in this direction was taken where a one
parameter deformation was found of the set of \circular brane" local integrals of motion
introduced in ref. [111]. This o ers the possibility for the quantization of the deformed
O(N ) NLSM along the lines of this work.
Perhaps the main motivation for studying NLSM is based on the fact that certain
types of SUSY sigma models are at the heart of the celebrated AdS/CFT correspondence,
and integrability is an important possibility. In particular, the NLSM associated with the
AdS side of the correspondence for N = 4 SUSY Yang Mills theory was argued to be
integrable [
112, 113
]. As brie y discussed in the introduction, the study of the rst principles
quantization of the NLSM by traditional techniques has proven to be di cult. A similar
situation exists with sigma models on supergroups and superspaces, which are expected to
provide theoretical descriptions of condensed matter systems with disorder [114]. That is
where one is most tempted to try the power of the ODE/IQFT approach.
Acknowledgments
PHY1404056.
S. L.'s gratitude goes to A. B. Zamolodchikov for sharing his insights and support.
The research of G. K. and S. L. is supported by the NSF under grant number
NSFA
Scaling behaviour of discretized chiral transfermatrices
To investigate the scaling behaviour of T (N)( ) (3.39){(3.44), we conducted numerical
work for integer n when the discretized operator is a
nite dimensional matrix that can be
diagonalized by means of the Bethe ansatz (see appendix B for details). We focused only
on the vacuum eigenvalue in the sector Hj(Nm2) and considered the cases with n = 2; 3; : : : ; 6
and all admissible values of j; m (3.14). Let (vac)( ) be the vacuum eigenvalue of the chiral
transfermatrix in the parafermionic subspace Vj
the vacuum eigenvalue of T
scaling limit of the discretized operator. To estimate numerical values of (vac)( ) we used
data obtained for a set of nite N and then performed a certain interpolation procedure to
N = 1. The results were compared with predictions coming from the properties of (vac)( )
(m). We expect that it can be obtained from
(N)( ) by using the formula (3.50) which explicitly describes the
root #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0:4818860
1:4891566
2:4919329
3:4935044
4:4946127
5:4955294
6:4963870
7:4972634
8:4982121
9:4992734
10:500481
11:501864
12:503448
13:505258
14:507318
0:4818829
1:4891424
2:4918863
3:4933890
4:4943769
5:4951073
6:4956974
7:4962107
8:4966857
9:4971480
10:497616
11:498105
12:498625
13:499187
14:499799
0:4818820
1:4891392
2:4918769
3:4933666
4:4943321
5:4950277
6:4955682
7:4960140
8:4964010
9:4967523
10:497084
11:497406
12:497729
13:498060
14:498404
0:4818814
1:4891372
2:4918715
3:4933541
4:4943074
5:4949844
6:4954981
7:4959077
8:4962476
9:4965392
10:496797
11:497031
12:497248
13:497455
14:497655
0:4818809
1:4891359
2:491868
3:493348
4:494295
5:494962
6:495463
7:495854
8:496171
9:496432
10:49665
11:49684
12:49701
13:49715
14:49728
0:47349
1:48725
2:49093
3:49276
4:49387
5:49464
6:49521
7:49564
8:49599
9:49628
10:49652
11:49672
12:49690
13:49705
14:49719
HJEP01(28)
g~1(p1; p2) =
t~1(p1; p2)
2 cos 2n+p22 ;
of the discretized operator T (N)( ) for n = 4, j = m = 0. The column \N = 1" was obtained by
interpolating the results for
nite N . The entries in the last column were calculated by using the
asymptotic formula (A.2) truncated at the rst nonzero term in the series.
discussed in section 2.2, specialized to the values p1 = 2i m and p2 = j + 12 . Agreement was
found in all cases considered. In this appendix, some of our numerical work is presented.
Let fulgl1=1 be the set of zeroes of (vac)( ) considered as a function of 2
. From the
numerical data it was found that all the zeroes are simple, real, positive, and accumulate
towards 2 = 1 with the leading asymptotic behaviour
ul
1 n
2
2
8
< l
: l
2j < n2 , this is consistent with the asymptotically exact formula,
n
ul2 +
1
2
1
X g~m
m=1
i
2
which can be easily derived from eqs. (2.53){(2.54). Knowledge of the coe cients g~m
allows us to compute systematic corrections to the leading asymptotic behaviour (A.1). As
it follows from eq. (2.52), the rst coe cient is
with t~1(p1; p2)  vacuum eigenvalue of ~t1  given by eq. (2.49). Notice that for p2 =
j + 12 = n+42 (neven), the denominator in (A.3) is zero so that (A.2) is no longer valid.
(A.1)
(A.3)
1.5 large (+ 2) asymptotics N = 1
1.0
asymptotic following from eq. (2.53). On the right panel, ~(vac) =
(vac) exp 2 (
2) 32 is plotted
and compared with the large (
2) asymptotic derived from eqs. (2.43), (2.44). The scaling function
was numerically estimated by interpolating to N = 1 the data for N = 500; 1000; 2000; 4000.
Also when j = m = 0, g~1 vanishes, but for this case the second term in the sum in (A.2)
is known, since
g~2 0;
1
2
t~2 0; 12
2 cos n+2
of T
and numerical values of t~2(0; 12 ) were calculated in ref. [43] and are reproduced in table 4.
Truncating the series in (A.2) at the rst non vanishing term, we calculated the corrections
to the leading asymptotic (A.1). This was compared to the zeroes of the vacuum eigenvalue
(N)( ) for increasing N . In all cases good agreement was observed. As an example, in
table 5 the results for n = 4, j = m = 0 are shown.
As 2
1, the asymptotic behaviour of (vac) is dictated by eqs. (2.43), (2.44).
Truncating the sum in (2.44) at the rst nonzero term and substituting ~tj by its vacuum
eigenvalue, we compared this to the results of the N = 1 interpolation. The agreement
was good considering that the interpolation procedure becomes rapidly less e cient for
increasing values of (
2). Figure 9 shows a plot of the estimated scaling function versus
the asymptotics for n = 3 and 2j = m = 1.
Another check that can be made is to consider the Taylor expansion of (vac)( ) at zero
following from formulae (2.42) and (2.48). The coe cient t1(p1; p2) (p1 = 2i m; p2 = j + 12 )
can be compared to the corresponding term in the vacuum eigenvalue of the discretized
operator:
T
(N;vac)( ) = 2 cos
+ t(1N) 2 + O( 4) :
Note that t(1N) is a divergent quantity for large N and must be regularized. According to
eq. (3.50), for n > 2, the following limit exists and converges to t1:
i
2
t1
n
cos( mn )
cos( n )
:
(A.4)
m
n
1
t(N;reg) =
0:54474
0:43807
0:54519
0:44710
0:54542
0:45357
0:54553
0:45818
0:5456440
0:469649
N = 1
0:8630048
0:419808
0:5456445
0:469446
eq. (2.48)
0:8630049
0:419632
to the expression for t1 2i m; j + 12 given by eq. (2.48). The column \N = 1" was obtained by
We compared the value of t1 2i m; j + 12
t(N;reg) and found good agreement for n = 3; 4 : : : ; 6 and all the allowed values of j; m. A
1
given by eq. (2.48) to the numerical values of
few cases are presented in table 6.
Finally, let us mention that for n = 2, analytic expressions exist for both (vac) and
the vacuum eigenvalue of T (N). In the case j = m = 0,
N
Y
m=1
T
(N;vac)( ) = 2
1
2 cot
2N
m
1
2
;
and using the formula (3.50), the scaling limit can be taken explicitly to yield
(vac)( ) =
e 2 2
2
2
1
2
It is easy to verify that this is consistent with the properties of the chiral transfermatrix
discussed in section 2.2. For n = 2 and 2j = m = 1, the discretized operator turns out to
be zero for any N and hence, ( ) = 0.
B
Scaling behaviour of Bethe roots for the vacuum state
space Hj(Nm2) . Recall that Hj(Nm2) denotes the eigenspace of the matrix
In this appendix we will consider the vacuum eigenvalue of the matrices Z ( ) in the
H(N) (Z) (3.41), (3.45)
having eigenvalue !j m2 , where j and m are restricted as in (3.14). Our considerations are
entirely based on the properties of Z ( ) (i){(v) listed in section 4.1.
Let Z
( )( ) be the eigenvalue corresponding to a common eigenvector j i of the
commuting family Z ( ). Using the analytical conditions (iv) and
!
symmetry (v), it
can be written in the form,
Z
(n 1)N 2j m
Y
i=1
1
i
(n
odd)
(B.1)
Z+( )( ) = B(N; ) m
Z
Q type relations (iii), it follows that the overall coe cient B(N; ) (depending
on the state j i) is the same for both Z+
is that the roots satisfy the following Bethe ansatz equations:
( ) and Z
( ). Another consequence of this relation
Y
i=1
n2N j m2 vi
Y
i=1
(n 22)YN j m2 wi
i=1
vi
wi
i + q 1 l =
i + q+1 l
q2m
q+ 21 l
q 21 l
2N
q 2 wl =
q+2 wl
q 2 vl =
q+2 vl
q2m
q2m
1
1
q+1 vl
q 1 vl
2N
(n
even)
(B.2)
(n
odd)
(B.3)
(n
even)
(B.4)
Similar equations for the FateevZamolodchikov spin chain (3.47) with periodic boundary
conditions were previously derived in the works [115] and [116] for odd and even n,
respectively. Notice that the constant B(N; ) in (B.1), (B.2) is determined (up to an overall sign)
by the quantum Wronskian type relations (ii).
The Bethe ansatz equations are valid for all integer n
2 and j; m restricted to (3.14),
except for 2j = m = n2 (n even) which requires special attention. In this case, for certain
sectors of H0
(N) a signi cant simpli cation occurs; Z
( ) vanishes so that the T
Q type relations (iii) become trivial and the quantum Wronskian type relations (ii) can be used to obtain much simpler equations for the roots. For instance, for the vacuum eigenvalue,
Z
(vac)( ) = 0 and Z+
(vac) is given explicitly by
Z+
(vac)( ) = 2
p
N
n
2
N 1
Y
l=1
1 + n cot
l
2N
n
2
;
;
2j = m =
n
even : (B.5)
Recall that the vacuum is de ned as the lowest energy state of the FateevZamolodchikov
spin chain Hamiltonian (3.47), (3.48), which commutes with both Z+( ) and Z ( ) for
any .
We studied the solutions to the Bethe ansatz equations corresponding to the low energy
states j i of the FateevZamolodchikov spin chain. It was found that the roots accumulate
×× 0.5 ●
1.5
1.0
1.0
1.5
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
1.5
● 0.5
● 0.5
0.5
functions of 2 are shown for n = 6, 2j = 3; m = 1 and N = 8.
(vac) (circles) and Z
(vac) (crosses) as
HJEP01(28)
along the rays given by (see gure 10)
arg( ) =
arg( 2) =
arg( 2) =
2
2
n
p ;
p ;
p ; p = 1; 3; : : : ; n
( i
p = 1; 3; : : : ; n
1 (vi
p = 2; 4; : : : ; n
(wi
roots)
roots)
roots)
In the scaling limit most of the roots become densely packed along the rays. However we
observed that at the edges of the distribution, the roots exhibit a certain scaling behaviour.
In particular, at the edge next to zero of the locus labeled by the integer p, with index i
enumerating the roots ordered by increasing absolute value, the following limits exist
iN!x1ed
lim N n1 i(;Np; ) ;
iN!x1ed
lim N n2 vi(;Np ; ) ;
iN!x1ed
lim N n wi;p
2 (N; ) :
Here we temporarily exhibit the dependence of the roots on N and the state j i
the scaling limit can be de ned for the coe cient B(N; ) in formulae (B.1), (B.2):
N!1
B
m
( ) = s lim ( =N ) n B(N; ) :
Keeping N
nite, consider the logarithm of the r.h.s. of eqs. (4.1) and (4.2) for a given
eigenvalue. With Z
( ) of the form (B.1), (B.2) it is straightforward to
nd their Taylor
series at
= 0. In the case of odd n, the expansion coe cients are given by
M m(N) =
Mn(N) =
M m(N) =
1
m
nN
1
m
N
X
i
N
m
n
i
n
X
i
i
m +
n +
X
i
2
n
i
m
(n
1) log
( 1)m N
m
cos 2n
N e
(m < n)
(m > n)
. Also,
(B.6)
(B.7)
For even n,
Vm(N) =
W m(N) =
V n(N) =
W n(N) =
2
2
Vm(N) =
1
1
2
nN
2
nN
1
N
N
i
N
2m
2m
2m
X vi m
i
X w m +
i
cos
N
n
n
X vi 2 + 2 log
N e
n
X w 2
i
(n
2) log
N e
2
X vi m ;
W m(N) =
1
N
2m
X w m
i
It is expected that the following limits exist,
M m( ) = slim M m(N)
N!1
(n
odd)
N!1
N!1
Vm( ) = slim Vm(N) ; W m( ) = slim W m(N) (n
even)
and coincide with the expansion coe cients in
e n of the CFT eigenvalues of log :
log ( )( ) = log B( ) +
X ( 1)m M m( ) e mn
log +( )( ) = log B( ) +
X Vm( ) e 2mn
log ( )( ) = log B( ) +
+
W m( ) e 2mn :
n
m
m
n
2
e
e
1
m=1
1
X
m=1
(n
(n
odd)
even)
Recall that the symbol \slim" stands for the scaling limit which is applied for low
energy eigenstates only. For numerical checks, we focused only on the vacuum of the
FateevZamolodchikov spin chain (3.47), (3.48). Our numerical work con rmed the existence of
the limits (B.9) for n = 3; 4; : : : ; 6 and all admissible values of j and m (3.14). Since a few
of the expansion coe cients in (4.35) are available in explicit form, we have the following
analytical predictions for some of the limits in (B.9).
Let f0;1 = f0;1(p1; p2) be de ned by eq. (4.33) and (x)
(x)= (1
x). Then for
n > 2 one has (here the superscript \(vac)" in the notation for the coe cients (B.9) is
(B.8)
HJEP01(28)
exact
Numerical values of the coe cients (B.7), (B.8) for the vacuum of the
FateevZamolodchikov spin chain (3.47), (3.48). The column N = 1 was obtained by interpolating the
niteN data. The last column lists the exact predictions given in (B.10), (B.11).
HJEP01(28)
W n (0; j) = 4 log
E
log(n) + 8 (1 + j)
1 +
+ 2 E
The numerical data agreed with these explicit formulae. This is shown, for a few cases, in
(m = 1; 3; : : : ; n
2
2m )
m = 0; 1; 2; : : : ;
n
2
n
2
n
2
n
2
2
n
2
(B.10)
(B.11)
n
2
2
1 n
1 n
4
4
1
1
Mm(m; j) = 0
M2(m; j) =
n n
1
f0;1
V1
W1
2
V n (0; j)
n
n
2
2
2
2
2
; j
; j
f0;1
Mn(0; j) = 2 log
log(n) + 4 (1 + j)
1 +
+ E
Vm(m; j) = Wm(m; j)
m = 1; 2; : : : ;
m
V1(m; j) = W1(m; j) =
f0;1
; j +
m = 0; 1; 2; : : : ;
2
1)
; j +
1
2
+
1
2
n n
2
2
e
2
1
2n
; j +
1
2
E
1
2
i m
2
i(n
i(n
i(n
4
2
n
4
4
4
n
2)
2)
; j +
; j +
i m
2
1
2
1
2
1
2
n n
2
2
+ n n
1
2
1
2
1
n
1
n
1.626210
N = 1
N = 1
N = 1
Zamolodchikov spin chain (3.47), (3.48). The column \N = 1" contains the results of numerical
interpolation from the nite N data. The analytical expression for Bs(m) is given by (4.34).
As was already mentioned, the constant B(N; ) 2 can be found using the quantum
Wronskian type relations (ii) from section 4.1. The r.h.s. of these relations is proportional
to the lattice shift operator P(N) (3.49) whose eigenvalues are pure phases (3.52). By
explicit diagonalization of Z
for small N we found that
1
arg B(N; )
1
nN
s) s + n (L
L) + s
(mod 2) ;
(B.12)
12 m and L, L are nonnegative integers depending on the state j i. For
the vacuum state L = L = 0, and the overall sign of the limit B(vac) (B.6) is ( 1)s.
This coincides with the sign factor in Bs(m) (4.34). For large values of N , when direct
diagonalization becomes impossible, we veri ed by means of the Bethe ansatz that the
m
absolute value of ( =N ) n B(N;vac) converges to ( 1)s Bs(m) (see table 8).
Recall that 2j = m = n2 with even n is a special case. Using eq. (B.5) the scaling
functions can be found explicitly,
e 2
1 + 2e
;
(vac)( ) = 0 :
This formula can be applied for n = 2. For the remaining n = 2 case, j = m = 0, it is easy
1
2 cot
1)
Z
(vac)( ) = p
2 ;
to show that for nite N
Z+
(vac)( ) = p
2
N
Y
m=1
so that the scaling functions are given by
2
12 + 2e
2
e
4N
2e
;
C
Derivation of NLIE
In this appendix we sketch some technical details in the derivation of the system of
NLIE (4.94).
(vac)( ) = p
and
are
real,
then
Hermiticity
tions (4.12), (4.62), (4.67) and the periodicity (4.60) imply that
= e i e
e i p2
e+i p2
( + i ) +
+
:
e i e
e i p2
( ) +
e+i p2 +( )
= e+i e
e i p2 +( )
e+i p2
( ) +
+
;
(C.1)
e+i e
e i p2
( + i ) +
e+i p2
i n
i n
+
i n
2
+
i n
2
i n
2
=
i n
2
i n
2
=
i n
2
Due to the analyticity of the operators
( ), these relations should be satis ed
for any complex . Let us introduce the shortcut notations
+( )
( )
B0 =
; B1 =
i 2n )
and
= exp 2 i e . Then (C.1) can be rewritten as
B0 = U
1 +
1 +
B1 = U
1 +
1 +
A+
A
:
Solving these equations w.r.t. A+ and A , one nds
1
2
A+ =
1
B0B1 U 1
1
2
B0
1
2
B1
and similar for A . This formula, combined with the quantum Wronskian relation (4.63)
Together with the periodicity condition
( + i ) =
i ) the last equation implies
A+
2
A+
i
2
+ i2
2i sin 2n+p22
2
U 1
+
i
2
As it follows from the quantum Wronskian relation (4.68):
U 1
i
2
U 1
i
2
=
2i e 2 ip2 sin 2n+p22
+( + i (n 1)
2
+
+ i (n+1) :
2
B0
B1
1
B0
( )
( + i )
2i sin 2n+p22
;
+( ) +( + i )
2i U( ) sin 2 p2
n+2
+( )
( + i ) +
( ) +( + i )
4i sin 2n+p22
written in the form
leads to
A+( ) =
cos 2 e
i
2
:
U 1
:
This can be substituted into the previous formula, yielding eq. (4.76) with the subscript
\+". Of course the formula is valid for the \ " case also.
Let us now take a closer look at the second equation in (C.1) specialized to the
eigenvalues corresponding to a common eigenvector j i. Suppose j is a zero of +( )( ). As follows
from the quantum Wronskian relation (4.63), ( )( j ) 6= 0, and therefore we conclude that
i n
2
e i (e j +p2) +( )
j
ei (e j +p2) +( )
j
i +
i n
2
;
which can be equivalently written in the form (4.81).
As was mentioned in the main body of the text, the zeroes of the entire periodic
function
(vac)( + i ) are simple, located on the lines =m( ) =
+
(2m + 1); m 2
Z, and accumulate toward <e( ) ! +1. Also, assuming that the parameters p1 and p2
are restricted as in cases (b), (c) from section 4.5, it is expected that the entire function
(vac)( ) does not have any zeroes within the strip j=m( )j < 2 (n + 2). Therefore, as
+
follows from the de nition (4.77), "(vac)( ) is an analytic function for j=m( )j <
where
it has the leading asymptotic behaviour (4.79) at <e( ) ! +1. Combining this analytic
information with the \quantization condition" (4.82) for the zeroes of
(vac)( ) and the
+
asymptotic behaviour see eq. (4.74)
log +(vac)( ) =
2 e
k
+ o(1)
as
<e( ) ! +1
=m( ) <
;
with k = n2+p22 , it is a simple exercise (see however appendix D) to derive a dispersiontype
2L(vac)( 0
i ) + i "(vac)( 0
i )
4 e 0 i + 2 k
:
relation
log
(C.2)
(C.3)
(C.4)
(vac)
Z 1 d 0
1
2
(vac)
1
1 + e2 2 0+2i
i
2
= 2 e
2k
Here
2 (0; 2 ) is an arbitrary constant and the notation
L(vac)( ) = log 1 + exp
i"(vac)( )
The next important property employed in the derivation of the system of integral
equations (4.94){(4.96) is that "(vac)( ) can be written in terms of the Fourier integral
"(vac)( ) = 4 e
2 k +
ei
"~( ) :
Z
d
2
Notice that the existence of the Fourier transform is ensured by the asymptotic
behaviour (4.79) at
! +1, and formulae (4.89), (4.90) for
1. One can expect
that the function "~( ) decays su ciently fast as
1, so that the integral in (C.4)
converges for any
in the strip of analyticity j=m( )j < . It is not di cult to see now
log
(vac)
2k + i
i (n + 1)
2
Z
R+i0
(vac)
i (n + 1)
cosh( (n+1) )
sinh( n2 )
2
2
and also that the imaginary part of the function (C.3) with
having in nitesimally small
negative imaginary part, can be represented by the convergent integral
=m L(vac)(
i0) =
L~( ) :
Z
R+i0
Similarly for the function !( ) (4.87) with
real, one has
Z
R+i0
2
!(vac)( ) = 4 e +
!~( ) ;
log 1 + e !(vac)( )
=
M~ ( ) :
The remaining part of the derivation of the NLIE consists of straightforward manipulations
with the Fourier images "~, L~, !~, M~ . Finally, going back to functions of the variable , one
derives the system of integral equations (4.94){(4.96).
Knowing the functions "(vac)( ), !(vac)( ) from the solution of the NLIE, and the
asymptotic formulae (4.70), (4.74), one can recover the vacuum eigenvalues of the operators
+( ) and
+( ) from (C.2), (C.5). The corresponding explicit relations are given below,
where we drop the superscript \(vac)" like in the NLIE (4.94){(4.96):
log +( ) =
(C.6)
valid for j=m( )j < 2 (n + 2)
, and
log +( ) =
2 e
k +
sin( 2n ) e
1
F (CFT)(
3
F (CFT)(
k +
Z 1 d 0 h F (CFT)(
1
1 2 i
0 i ) L( 0 i )
1 2 i
Z 1 d 0 h F (CFT)(
3
0 i ) L( 0 i )
F4( ) =
n + 2
1
1
n + 2
1
n+2 with
tanh
coth
n + 2
n + 2
sin( n+2 )
2(n + 2) sinh( n++i22 ) sinh( n+22 )
i
"~( )
Z
R+i0
(C.5)
HJEP01(28)
1
1
i
Z 1 d 0
F2(
sin( n+2 )
2(n + 2) cosh
n++i22 cosh
i
2
n+2
1
2 cosh( )
. Here the kernels are given by F (CFT)( ) = F1( )
1
n+12 ,
F2( ) =
1
sinh( )
s~m(p1; p2) =
Z 1 d
1
2
n + 2
Z 1 d
1
The vacuum eigenvalues of the chiral transfermatrix can be obtained using the T
(vac)(i ) =
+( )
+( )
with
= e n :
Combining (C.6), (C.7) with the general asymptotic expansions at <e( ) ! +1 found
in (4.70), (4.74), the expressions for the local and dual nonlocal integrals of motion follow
2 =mhe(2m 1)( i )L(
i ) + ( 1)me(2m 1) log 1 + e !( )
(C.9)
2m( i )
=m e n+2 L(
i )
sin
2m
e n+2 log 1 + e !( )
:
D
Modi ed NLIE for oscillating asymptotics
Here we discuss the modi cations to the integral equations (4.94){(4.96) for the case of
12 , when the asymptotics of the functions "(vac)( ) and !(vac)( ) at
real p1 6= 0 and p2 >
1 oscillate (4.88).
The rst important di erence in this case is that (vac)( ) has a set of zeroes f m( )gm1=1
+
in the strip j=m( )j < 2 (n + 2) whose asymptotic behaviour is given by relation (4.93).
Secondly, in the derivation of (4.94) presented in the previous appendix, we implicitly
assumed that all values
on the real axis, such that "(vac)( ) =
(mod 2 ) arise from
the quantization condition (4.82), i.e.,
i for some j = 1; 2; : : : (recall that
(vac)
j
=m( j ) = ). In other words all such
are related to the zeroes of +(vac)( ) and,
therefore, form an increasing semiin nite sequence extending towards +1 on the real axis
(see (4.83)). For the oscillating asymptotics (4.88) this is no longer true. Indeed, it is easy
to check from (4.88) that the condition
"(vac)(~m) =
1)
with
m = 1; 2; : : :
is satis ed for an in nite set of values ~
1
m m=1 which extend towards
1 such that
m =
n
2p1
1
2
(p1; p2)
+ o (m=p1) 1 ;
valid up to an exponentially small correction. Here
(p1; p2) = 4p1 0=n + i log cos( (p2 + ip1))= cos( (p2
ip1)
coincides with the scattering phase de ned by eq. (5.11). In the terminology of the Bethe
ansatz we have an in nite number of \holes" where the phase passes a resonant value
without a corresponding zero j . Therefore the integrals in the r.h.s. of (4.94) contain
spurious contributions from nonexistent roots. To exclude these unwanted contributions
one needs to add extra source terms to the r.h.s. of eqs. (4.94).
Introduce the notation
J (")( ) = i X log
J (!)( ) =
1
m=1
1
X log
m=1
S(
~m)
t
S
+ i2 (n + 2)
t
+ i2
~
+ i2 + i2 (n + 2)
( )
( )
where S( ) and t( ) are de ned in (E.1) below. Then the modi ed equations (4.94) can be
written as
2 k + J (")(
i ) +
Z 1 d 0 hG(
1 2 i
0 2i ) L( 0 i )
0) L( 0 i ) +
G1(
0 i ) log 1 + e !( 0)
!( ) = 4 e + J (!)( ) + =m
G1(
(D.2)
L( ) = log 1 + e i"( ) :
One can check that the leading terms in the asymptotics (4.88) solves these equations at
1, i.e., when the exponential terms proportional to e in the r.h.s. are omitted.
E
Twoparticle Smatrix for the sausage model
model [14]. The Smatrix satis es the YangBaxter equation and was originally introduced
as the Boltzmann weights of the socalled 19vertex model [117].
S( ) = S++++( ) = S
( ) =
T ( ) = S++ ( ) = S ++( ) = S(i
)
sinh( in+2 )
sinh( in++2 )
t( ) = S++00( ) = S00++( ) = S 00( ) = S00 ( ) =
= S+00 (i
) = S00+(i
) = i
) = S+0+0 ( ) = S0++0( ) = S0 0 ( ) = S0 0( ) = S0+0 (i
) = S00+(i
sinh( n+2 ) sinh( in+2 )
sinh( 2ni+2 ) sinh( in++2 )
sin( n2+2 ) sinh( in+2 )
sinh( 2ni+2 ) sinh( in++2 )
(E.1)
R( ) = S++ ( ) = S++( ) =
( ) = S0000( ) = S++00( ) + S++ ( )
sinh( 2ni+2 ) sinh( in++2 )
As a 9
9 matrix S(27!2) satis es the conditions
S(27!2) y S(27!2) = I(2)
for =m( ) = 0
det S(27!2)( ) =
sinh2( in+2 ) sinh( 2ni++2 ) 4
sinh2( in++2 ) sinh( 2ni+2 )
:
(E.2)
Attribution License (CCBY 4.0), which permits any use, distribution and reproduction in
any medium, provided the original author(s) and source are credited.
[1] L.D. Faddeev, E.K. Sklyanin and L.A. Takhtajan, The quantum inverse problem method. 1,
Theor. Math. Phys. 40 (1980) 688 [Teor. Mat. Fiz. 40 (1979) 194] [INSPIRE].
[2] R.J. Baxter, Partition function of the eight vertex lattice model, Annals Phys. 70 (1972) 193
[INSPIRE].
(1982) [INSPIRE].
[3] R.J. Baxter, Exactly solved models in statistical mechanics, Academic Press, London U.K.,
[4] C. Destri and H.J. de Vega, Light cone lattice approach to fermionic theories in 2D: the
massive Thirring model, Nucl. Phys. B 290 (1987) 363 [INSPIRE].
[5] K. Pohlmeyer, Integrable Hamiltonian systems and interactions through quadratic
constraints, Commun. Math. Phys. 46 (1976) 207 [INSPIRE].
[6] V.E. Zakharov and A.V. Mikhailov, Relativistically invariant twodimensional models in
eld theory integrable by the inverse problem technique (in Russian), Sov. Phys. JETP 47
(1978) 1017 [Zh. Eksp. Teor. Fiz. 74 (1978) 1953] [INSPIRE].
[7] A.M. Polyakov and P.B. Wiegmann, Theory of nonAbelian Goldstone bosons, Phys. Lett.
B 131 (1983) 121 [INSPIRE].
[8] L.D. Faddeev and N. Yu. Reshetikhin, Integrability of the principal chiral eld model in
(1 + 1)dimension, Annals Phys. 167 (1986) 227 [INSPIRE].
[9] A.B. Zamolodchikov and A.B. Zamolodchikov, Factorized S matrices in twodimensions as
the exact solutions of certain relativistic quantum eld models, Annals Phys. 120 (1979) 253
85 (1994) 1619 [hepth/9403101] [INSPIRE].
177 (1996) 381 [hepth/9412229] [INSPIRE].
[10] A.G. Bytsko, The zero curvature representation for nonlinear O(3) model, J. Math. Sci.
[11] V.V. Bazhanov, S.L. Lukyanov and A.B. Zamolodchikov, Integrable structure of conformal
eld theory, quantum KdV theory and thermodynamic Bethe ansatz, Commun. Math. Phys.
Russia, (1980).
SpringerVerlag, Berlin Germany, (1987) [INSPIRE].
[12] D. Fioravanti and M. Rossi, A braided YangBaxter algebra in a theory of two coupled
lattice quantum KdV: algebraic properties and ABA representations, J. Phys. A 35 (2002)
3647 [hepth/0104002] [INSPIRE].
[13] D. Fioravanti and M. Rossi, Exact conserved quantities on the cylinder 1: conformal case,
JHEP 07 (2003) 031 [hepth/0211094] [INSPIRE].
[14] V.A. Fateev, E. Onofri and A.B. Zamolodchikov, The sausage model (integrable
deformations of O(3) model), Nucl. Phys. B 406 (1993) 521 [INSPIRE].
[15] S.L. Lukyanov, The integrable harmonic map problem versus Ricci ow, Nucl. Phys. B 865
[16] E.K. Sklyanin, On the complete integrability of the LandauLifshitz equation, LOMIE793,
[17] L.D. Faddeev and L.A. Takhtajan, Hamiltonian methods in the theory of solitons,
[18] V.V. Bazhanov, S.L. Lukyanov and A.B. Zamolodchikov, Integrable structure of conformal
eld theory II. Qoperator and DDV equation, Commun. Math. Phys. 190 (1997) 247
[hepth/9604044] [INSPIRE].
[hepth/9805008] [INSPIRE].
[19] V.V. Bazhanov, S.L. Lukyanov and A.B. Zamolodchikov, Integrable structure of conformal
eld theory III. The YangBaxter relation, Commun. Math. Phys. 200 (1999) 297
[20] A. Voros, Spectral zeta functions, Adv. Stud. Pure Math. 21 (1992) 327.
[21] A. Voros, An exact solution method for 1D polynomial Schrodinger equations, J. Phys. A
32 (1999) 5993 [mathph/9902016].
[22] P. Dorey and R. Tateo, Anharmonic oscillators, the thermodynamic Bethe ansatz and
nonlinear integral equations, J. Phys. A 32 (1999) L419 [hepth/9812211] [INSPIRE].
[23] V.V. Bazhanov, S.L. Lukyanov and A.B. Zamolodchikov, Spectral determinants for
Schrodinger equation and Qoperators of conformal eld theory, J. Statist. Phys. 102 (2001)
567 [hepth/9812247] [INSPIRE].
[24] J. Suzuki, Functional relations in Stokes multipliers and solvable models related to Uq(A(n1)),
J. Phys. A 33 (2000) 3507 [hepth/9910215] [INSPIRE].
[25] V.V. Bazhanov, S.L. Lukyanov and A.B. Zamolodchikov, Higher level eigenvalues of
Qoperators and Schrodinger equation, Adv. Theor. Math. Phys. 7 (2003) 711
[hepth/0307108] [INSPIRE].
[26] Al.B. Zamolodchikov, Generalized Mathieu equation and Liouville TBA, in Quantum
eld
theories in two dimensions, collected works of Alexei Zamolodchikov, volume 2, A. Belavin,
Ya. Pugai and A. Zamolodchikov eds., World Scienti c, Singapore, (2012).
[27] S.L. Lukyanov and A.B. Zamolodchikov, Quantum sine(h)Gordon model and classical
integrable equations, JHEP 07 (2010) 008 [arXiv:1003.5333] [INSPIRE].
[28] P. Dorey, C. Dunning and R. Tateo, The ODE/IM correspondence, J. Phys. A 40 (2007)
R205 [hepth/0703066] [INSPIRE].
[29] P. Dorey, S. Faldella, S. Negro and R. Tateo, The Bethe ansatz and the
TzitzeicaBulloughDodd equation, Phil. Trans. Roy. Soc. Lond. A 371 (2013) 20120052
[arXiv:1209.5517] [INSPIRE].
[30] P. Adamopoulou and C. Dunning, Bethe ansatz equations for the classical A(n1) a ne Toda
eld theories, J. Phys. A 47 (2014) 205205 [arXiv:1401.1187] [INSPIRE].
[31] K. Ito and C. Locke, ODE/IM correspondence and Bethe ansatz for a ne Toda eld
equations, Nucl. Phys. B 896 (2015) 763 [arXiv:1502.00906] [INSPIRE].
[32] D. Masoero, A. Raimondo and D. Valeri, Bethe ansatz and the spectral theory of a ne Lie
algebravalued connections I. The simplylaced case, Commun. Math. Phys. 344 (2016) 719
[arXiv:1501.07421] [INSPIRE].
[33] D. Masoero, A. Raimondo and D. Valeri, Bethe ansatz and the spectral theory of a ne Lie
algebravalued connections II. The non simplylaced case, Commun. Math. Phys. 349 (2017)
1063 [arXiv:1511.00895] [INSPIRE].
[34] K. Ito and H. Shu, ODE/IM correspondence for modi ed B2(1) a ne Toda eld equation,
Nucl. Phys. B 916 (2017) 414 [arXiv:1605.04668] [INSPIRE].
[35] C. Babenko and F. Smirnov, Suzuki equations and integrals of motion for supersymmetric
CFT, Nucl. Phys. B 924 (2017) 406 [arXiv:1706.03349] [INSPIRE].
[36] R.S. Hamilton, The Ricci ow on surfaces, Contemp. Math. 71, Amer. Math. Soc.,
[37] S. Elitzur, A. Forge and E. Rabinovici, Some global aspects of string compacti cations,
Nucl. Phys. B 371 (1992) 269 [INSPIRE].
[46] R. Dijkgraaf, H.L. Verlinde and E.P. Verlinde, String propagation in a black hole geometry,
[47] A.B. Zamolodchikov and Al.B. Zamolodchikov, unpublished notes, (1995).
[48] A. Gerasimov, A. Marshakov and A. Morozov, Free eld representation of parafermions and
related coset models, Nucl. Phys. B 328 (1989) 664 [INSPIRE].
Providence RI U.S.A., (1988), pg. 237.
Nucl. Phys. B 359 (1991) 581 [INSPIRE].
Phys. Lett. A 74 (1979) 116 [INSPIRE].
Phys. A 15 (1982) 897.
1565 [INSPIRE].
[38] E. Witten, On string theory and black holes, Phys. Rev. D 44 (1991) 314 [INSPIRE].
[39] Yu. G. Stroganov, A new calculation method for partition functions in some lattice models,
[40] R.J. Baxter and P.A. Pearce, Hard hexagons: interfacial tension and correlation length, J.
[41] A.N. Kirillov and N.Y. Reshetikhin, Exact solution of the integrable XXZ Heisenberg model
with arbitrary spin I. The ground state and the excitation spectrum, J. Phys. A 20 (1987)
[42] S.L. Lukyanov, E.S. Vitchev and A.B. Zamolodchikov, Integrable model of boundary
interaction: the paperclip, Nucl. Phys. B 683 (2004) 423 [hepth/0312168] [INSPIRE].
[43] S.L. Lukyanov, A.M. Tsvelik and A.B. Zamolodchikov, Paperclip at
= , Nucl. Phys. B
719 (2005) 103 [hepth/0501155] [INSPIRE].
[44] S.L. Lukyanov and A.B. Zamolodchikov, Integrability in 2D
eld theory/sigma models,
lecture notes for the Les Houches school, to appear, France, (2016).
[45] I. Bakas and E. Kiritsis, Beyond the largeN limit: nonlinear W1 as symmetry of the
SL(2; R)=U(1) coset model, Int. J. Mod. Phys. A 7S1A (1992) 55 [hepth/9109029]
[49] V.A. Fateev and A.B. Zamolodchikov, Parafermionic currents in the twodimensional
conformal quantum
eld theory and selfdual critical points in ZN invariant statistical
[50] V.A. Fateev and A.B. Zamolodchikov, Selfdual solutions of the star triangle relations in ZN
[51] G. Felder, BRST approach to minimal models, Nucl. Phys. B 317 (1989) 215 [Erratum ibid.
[52] A.G. Izergin and V.E. Korepin, The lattice quantum sineGordon model, Lett. Math. Phys.
[53] E.K. Sklyanin, Some algebraic structures connected with the YangBaxter equation.
Representations of quantum algebras, Funct. Anal. Appl. 17 (1983) 273 [Funkt. Anal. Pril.
17 (1983) 34] [INSPIRE].
[54] V.V. Bazhanov and Yu. G. Stroganov, Chiral Potts model as a descendant of the six vertex
model, J. Statist. Phys. 59 (1990) 799 [INSPIRE].
[55] R.J. Baxter, V.V. Bazhanov and J.H.H. Perk, Functional relations for transfer matrices of
the chiral Potts model, Int. J. Mod. Phys. B 4 (1990) 803 [INSPIRE].
[56] E.K. Sklyanin, Separation of variables  new trends, Prog. Theor. Phys. Suppl. 118 (1995)
B 324 (1989) 548] [INSPIRE].
5 (1981) 199 [INSPIRE].
35 [solvint/9504001] [INSPIRE].
[arXiv:1303.2566] [INSPIRE].
[INSPIRE].
[58] V.V. Bazhanov and S.L. Lukyanov, Integrable structure of quantum
eld theory: classical
at connections versus quantum stationary states, JHEP 09 (2014) 147 [arXiv:1310.4390]
[59] D. Fioravanti, Geometrical loci and CFTs via the Virasoro symmetry of the mKdVSG
hierarchy: an excursus, Phys. Lett. B 609 (2005) 173 [hepth/0408079] [INSPIRE].
[60] B. Feigin and E. Frenkel, Quantization of soliton systems and Langlands duality, Adv. Stud.
Pure Math. 61 (2011) 185 [arXiv:0705.2486] [INSPIRE].
[61] A. Klumper, M.T. Batchelor and P.A. Pearce, Central charges of the 6 and 19vertex
models with twisted boundary conditions, J. Phys. A 24 (1991) 3111 [INSPIRE].
[62] C. Destri and H.J. de Vega, New approach to thermal Bethe ansatz, Phys. Rev. Lett. 69
(1992) 2313 [hepth/9203064] [INSPIRE].
[63] C. Destri and H.J. De Vega, Uni ed approach to thermodynamic Bethe ansatz and nite
size corrections for lattice models and eld theories, Nucl. Phys. B 438 (1995) 413
[hepth/9407117] [INSPIRE].
[64] A.B. Zamolodchikov and A.B. Zamolodchikov, Structure constants and conformal bootstrap
in Liouville eld theory, Nucl. Phys. B 477 (1996) 577 [hepth/9506136] [INSPIRE].
[65] V.A. Fateev, I.V. Frolov and A.S. Shvarts, Quantum
uctuations of instantons in the
nonlinear model, Nucl. Phys. B 154 (1979) 1 [INSPIRE].
[66] A.P. Bukhvostov and L.N. Lipatov, Instantonantiinstanton interaction in the O(3)
nonlinear
model and an exactly soluble fermion theory, Nucl. Phys. B 180 (1981) 116
[Pisma Zh. Eksp. Teor. Fiz. 31 (1980) 138] [INSPIRE].
[67] M. Luscher, Does the topological susceptibility in lattice models scale according to the
perturbative renormalization group?, Nucl. Phys. B 200 (1982) 61 [INSPIRE].
[68] M. Luscher, Volume dependence of the energy spectrum in massive quantum eld theories I.
Stable particle states, Commun. Math. Phys. 104 (1986) 177 [INSPIRE].
[69] M. Luscher, Volume dependence of the energy spectrum in massive quantum eld theories
[70] C. Ahn, J. Balog and F. Ravanini, NLIE for the sausage model, arXiv:1701.08933
[71] A.M. Polyakov, Interaction of Goldstone particles in twodimensions. Applications to
HJEP01(28)
ferromagnets and massive YangMills elds, Phys. Lett. B 59 (1975) 79 [INSPIRE].
[72] S. Hikami and E. Brezin, Three loop calculations in the twodimensional nonlinear model,
J. Phys. A 11 (1978) 1141 [INSPIRE].
[73] P. Hasenfratz, M. Maggiore and F. Niedermayer, The exact mass gap of the O(3) and O(4)
nonlinear models in D = 2, Phys. Lett. B 245 (1990) 522 [INSPIRE].
[74] M. Luscher, P. Weisz and U. Wol , A numerical method to compute the running coupling in
asymptotically free theories, Nucl. Phys. B 359 (1991) 221 [INSPIRE].
[75] D.S. Shin, A determination of the mass gap in the O(N ) model, Nucl. Phys. B 496
(1997) 408 [heplat/9611006] [INSPIRE].
[76] J. Balog and A. Hegedus, The nite size spectrum of the 2dimensional O(3) nonlinear
model, Nucl. Phys. B 829 (2010) 425 [arXiv:0907.1759] [INSPIRE].
[77] G.V. Dunne and M. Unsal, Resurgence and transseries in quantum eld theory: the CPN 1
model, JHEP 11 (2012) 170 [arXiv:1210.2423] [INSPIRE].
[78] A. Dabholkar, Strings on a cone and black hole entropy, Nucl. Phys. B 439 (1995) 650
[hepth/9408098] [INSPIRE].
[79] V.A. Fateev and A.B. Zamolodchikov, Integrable perturbations of ZN parafermion models
and O(3) model, Phys. Lett. B 271 (1991) 91 [INSPIRE].
quantum
eld theories, Nucl. Phys. B 473 (1996) 509 [INSPIRE].
[80] V.A. Fateev, The model (dual) representation for a twoparameter family of integrable
[81] V.V. Bazhanov, G.A. Kotousov and S.L. Lukyanov, Winding vacuum energies in a
deformed O(4) model, Nucl. Phys. B 889 (2014) 817 [arXiv:1409.0449] [INSPIRE].
[82] V.V. Bazhanov, S.L. Lukyanov and B.A. Runov, Vacuum energy of the BukhvostovLipatov
model, Nucl. Phys. B 911 (2016) 863 [arXiv:1607.04839] [INSPIRE].
[83] V.V. Bazhanov, S.L. Lukyanov and B.A. Runov, BukhvostovLipatov model and
quantumclassical duality, arXiv:1711.09021 [INSPIRE].
[84] J.M. Maillet, New integrable canonical structures in twodimensional models, Nucl. Phys. B
269 (1986) 54 [INSPIRE].
[85] F.A. Smirnov, Quasiclassical study of formfactors in nite volume, hepth/9802132
[86] S.L. Lukyanov, Finite temperature expectation values of local elds in the sinhGordon
model, Nucl. Phys. B 612 (2001) 391 [hepth/0005027] [INSPIRE].
[87] A.B. Zamolodchikov, On the thermodynamic Bethe ansatz equation in sinhGordon model,
J. Phys. A 39 (2006) 12863 [hepth/0005181] [INSPIRE].
[88] A.G. Bytsko and J. Teschner, Quantization of models with noncompact quantum group
symmetry: modular XXZ magnet and lattice sinhGordon model, J. Phys. A 39 (2006)
12927 [hepth/0602093] [INSPIRE].
(2008) 403 [hepth/0702214] [INSPIRE].
related to the sinhmodel, arXiv:1412.7721 [INSPIRE].
[89] J. Teschner, On the spectrum of the sinhGordon model in nite volume, Nucl. Phys. B 799
[90] G. Borot, A. Guionnet and K.K. Kozlowski, Asymptotic expansion of a partition function
[91] Y. Ikhlef, J.L. Jacobsen and H. Saleur, An integrable spin chain for the SL(2; R)=U(1) black
hole model, Phys. Rev. Lett. 108 (2012) 081601 [arXiv:1109.1119] [INSPIRE].
[92] H. Boos, M. Jimbo, T. Miwa, F. Smirnov and Y. Takeyama, Hidden Grassmann structure
in the XXZ model, Commun. Math. Phys. 272 (2007) 263 [hepth/0606280] [INSPIRE].
[93] H. Boos, M. Jimbo, T. Miwa, F. Smirnov and Y. Takeyama, Hidden Grassmann structure
in the XXZ model II: creation operators, Commun. Math. Phys. 286 (2009) 875
[arXiv:0801.1176] [INSPIRE].
[94] M. Jimbo, T. Miwa and F. Smirnov, Hidden Grassmann structure in the XXZ model III:
introducing Matsubara direction, J. Phys. A 42 (2009) 304018 [arXiv:0811.0439]
model IV: CFT limit, Commun. Math. Phys. 299 (2010) 825 [arXiv:0911.3731] [INSPIRE].
[96] M. Jimbo, T. Miwa and F. Smirnov, Hidden Grassmann structure in the XXZ model V:
sineGordon model, Lett. Math. Phys. 96 (2011) 325 [arXiv:1007.0556] [INSPIRE].
[97] P.B. Wiegmann, On the theory of nonAbelian Goldstone bosons in twodimensions: exact
solution of the O(3) nonlinear
model, Phys. Lett. B 141 (1984) 217 [INSPIRE].
[98] J. Balog and A. Hegedus, TBA equations for excited states in the O(3) and O(4) nonlinear
model, J. Phys. A 37 (2004) 1881 [hepth/0309009] [INSPIRE].
[99] N. Gromov, V. Kazakov and P. Vieira, Finite volume spectrum of 2D eld theories from
Hirota dynamics, JHEP 12 (2009) 060 [arXiv:0812.5091] [INSPIRE].
[100] A.B. Zamolodchikov and A.B. Zamolodchikov, Massless factorized scattering and models
with topological terms, Nucl. Phys. B 379 (1992) 602 [INSPIRE].
[101] B. Feigin, M. Jimbo, T. Miwa and E. Mukhin, Quantum toroidal gl1 and Bethe ansatz, J.
Phys. A 48 (2015) 244001 [arXiv:1502.07194] [INSPIRE].
[102] M.N. Al mov and A.V. Litvinov, On spectrum of ILW hierarchy in conformal eld theory
II: coset CFT's, JHEP 02 (2015) 150 [arXiv:1411.3313] [INSPIRE].
[103] B. Feigin, M. Jimbo and E. Mukhin, Integrals of motion from quantum toroidal algebras, J.
Phys. A 50 (2017) 464001 [arXiv:1705.07984].
[104] C. Klimc k, Integrability of the biYangBaxter model, Lett. Math. Phys. 104 (2014) 1095
[arXiv:1402.2105] [INSPIRE].
06 (2014) 002 [arXiv:1403.5517] [INSPIRE].
[105] B. Hoare, R. Roiban and A.A. Tseytlin, On deformations of AdSn
Sn supercosets, JHEP
model, Nucl. Phys. B 891 (2015) 312 [arXiv:1410.8066] [INSPIRE].
and their relation, Nucl. Phys. B 899 (2015) 489 [arXiv:1506.05784] [INSPIRE].
345 [arXiv:1606.03016] [INSPIRE].
11 (2016) 138 [arXiv:1609.06271] [INSPIRE].
HJEP01(28)
charging at large conduction, J. Stat. Mech. 0405 (2004) P05003 [hepth/0306188]
(1980) 581] [INSPIRE].
S5 superstring,
Phys. Rev. D 69 (2004) 046002 [hepth/0305116] [INSPIRE].
U.S.A., (1997).
systems , Sov. Phys. JETP 62 ( 1985 ) 215 [Zh . Eksp. Teor. Fiz. 89 ( 1985 ) 380] [INSPIRE].
[57] S.L. Lukyanov , ODE/IM correspondence for the Fateev model , JHEP 12 ( 2013 ) 012 II. Scattering states , Commun. Math. Phys. 105 ( 1986 ) 153 [INSPIRE]. [107] K. Sfetsos , K. Siampos and D.C. Thompson , Generalised integrable  and  deformations [108] C. Klimc k, PoissonLie Tduals of the biYangBaxter models , Phys. Lett. B 760 ( 2016 ) [110] V. Fateev , A. Litvinov and L. Spodyneiko , private communications, to be published.
[113] N. Beisert et al., Review of AdS/CFT integrability: an overview , Lett. Math. Phys. 99