On $tau^{(2)}$-model in Chiral Potts Model and Cyclic Representation of Quantum Group $U_q(sl_2)$
We identify the precise relationship between the five-parameter $\tau^{(2)}$-family in the $N$-state chiral Potts model and XXZ chains with $U_q (sl_2)$-cyclic representation. By studying the Yang-Baxter relation of the six-vertex model, we discover …
Authors: Shi-shyr Roan
On τ ( 2 ) -mo del in C hiral P otts Mo del and Cyclic Representation of Quan tum Group U q ( sl 2 ) Shi-shyr Roan Institute of Mathematics A c ademia Sinic a T aip ei , T aiwan (email: mar o an@gate.sinic a.e du.tw ) Abstract W e iden tify the pr e c ise relationship b et ween the five-parameter τ (2) -family in the N -state chiral P otts mo del and XXZ chains with U q ( sl 2 )-cyclic r epresen tation. By studying the Y ang- Baxter relation of the six-vertex mo del, w e discov er an one-parameter family of L -op erators in terms of the q ua n tum group U q ( sl 2 ). When N is o dd, the N -state τ (2) -mo del can be regar de d as the XXZ chain of U q ( sl 2 ) cyclic representations with q N = 1. The s ymmet ry algebra of the τ (2) -mo del is describ ed by the quantum affine alg ebra U q ( b sl 2 ) via the canonical representation. In genera l for an arbitrar y N , w e show that the XXZ chain with a U q ( sl 2 )-cyclic representation for q 2 N = 1 is equiv alent to tw o copies of the same N -state τ (2) -mo del. 2008 P A CS: 05.50. +q, 02.20.Uw, 75.10Jm 2000 MSC: 17B37 , 17B80, 82B20 Key wor ds : τ (2) -mo del, XXZ c hain, Qu antum group U q ( sl 2 ) 1 In tro duction In the study of N -state chiral P otts mo del (CPM), Bazhano v and S t rogano v [12] disco v ered a fiv e- parameter family of L -op erators of τ (2) -mo del in the six-ve rtex mod e l with a particular field (see 1 [9] page 3), i.e. the solution of Y ang-Baxter (YB) equation for the asymmetric six-v ertex R -matrix, (see (2.1) and (2.2) in this pap er). The c h iral P otts transfer matrix can b e constructed as th e Bax- ter’s Q -op erato r of the τ (2) -mo del [11, 29]. By the functional relations b et ween the fu sion matrices of τ (2) -mo del and the c h iral P otts transfer matrix, one can compute the eigen v alue sp ectrum of the (homogeneous) sup erin tegrable C P M [1, 6, 7], where the τ (2) -degeneracy o ccurs with the sym met ry algebra d escrib ed by Onsager algebra [24]. F urthermore, one can discus s the eigen v alue sp ectrum and calculate the order paramete r in CPM thr o ugh functional-relation method [4, 5, 10, 22, 30]. Though muc h progress has b een made on issues related to the eigen v alues, w e still lac k enough information at presen t ab out the eigen v ectors in CP M. Consequen tly , some imp ortant problems whic h r equ ire the kn o wledge of eigen vect ors, su c h as the ca lculation of correlation functions, remain unsolv ed in the th e ory . On the other h a nd, the (”zero-field”) six-v ertex mo del ([8] Sect. 3) with the symmetric R -mat rix has b een a w ell-studied theory with pr ofoun d knowledge kno wn in literature ab out their eigen v alues and eigen vect ors. The un d erstanding of the structure w as fur ther extended to XXZ mo dels of h ig her spin [20], or more generally the mo del asso cia ted to a represen tation of the quantum group U q ( sl 2 ). When q is a N th ro ot of unit y , it wa s sho wn in [15, 16, 17, 23, 25, 27] that the degeneracy of XXZ m odels o c curs with the extra sl 2 -lo op -algebra s y m met ry induced from the structure of the quantum affine algebra U q ( b sl 2 ). F or o dd N , as a sp ecial mem b er of a one- parameter U q ( sl 2 )-cyclic rep resen tation, the spin-( N − 1) / 2 XXZ c hain can b e identified with the sup erin tegrable N -state τ (2) -mo del [28]. Hence the sup erint egrable τ (2) -mo del also carries the sl 2 - lo o p-algebra s y m met ry for the degeneracy (of certain sectors), compatible with the Onsager-alg ebra symmetry inherited from the c hiral Potts transfer matrix [28]. Th e in terrelation of these t w o sym- metries can lead to s ome useful information ab out eigen ve ctors of the sup erint egrable mo del (see, e.g. [2, 3]). The aim of this pap er is to extend the equiv alen t relationship b et w een the N -state τ (2) - mo del and XXZ mo del with U q ( sl 2 )-cyclic represen tation to the whole fi v e-parameter τ (2) -family found in [12] for an y N . Through the YB relat ion in the six-ve rtex mod el , w e find a c haracterization of quant um group U q ( sl 2 ) through a one-parameter L -op erators of the YB solution (see (2.5), (2.6) in the pap er). When q is a ro ot of unit y , the alge bra U q ( sl 2 ) p ossesses a three-parameter family of cyclic represen tation. T ogether with the rescaling of the sp ectral p a rameter, XXZ c hains w ith U q ( sl 2 )-cyclic representa tion also carry a fi ve-paramete r L -operators. By extending the argument in [28], w e sho w the identifica tion of τ (2) -mo dels and the XXZ c hains with U q ( sl 2 )-cyclic represen- tation. Note that suc h a pr ec ise connection b et wee n τ (2) -family in C PM an d th a t XXZ-family w it h symmetric R -matrix seems not to ha v e app eared in the literature b efore, to the b e st of the author’s kno wledge, even in o dd N case except a one-parameter family of τ (2) -mo dels in [28]. W e hop e an iden tification of these mo dels will h e lp to pr o vide some usefu l clues to und e rstand the eigen v ectors of C PM in certain sp ecial cases. This pap er is organized as follo w s . In section 2, w e recall the definition of τ (2) -mo del, qu a nt um group U q ( sl 2 ), and the quan tum affine alge bra U q ( b sl 2 ). In sectio n 3, w e fi rst describ e the three- parameter family of cyclic r e presenta tions of U q ( sl 2 ) for a ro ot of unity q . The case of q N = 1 f o r o dd N is discussed in subsection 3.1 where the τ (2) -mo del is iden tified with the XXZ c hain with cyclic r e presenta tion of U q ( sl 2 ), h e nce the quantum space carries a canonical r e presentat ion of th e 2 quan tum affine algebra U q ( b sl 2 ). I n subsection 3.2, we stu dy the N -state τ (2) -mo del for an arbitrary N . W e show that a XXZ c hain of U q ( sl 2 )-cyclic representa tion with q 2 N = 1 is equiv alent to tw o copies of the same N -state τ (2) -mo del. Finally we close in section 4 with a brief concluding remark. Notatio n: W e use standard notations. F or a p ositiv e integ er N greate r than one, C N denotes the ve ctor space of N -cyclic v ectors w i th the canonical base | n i , n ∈ Z N (:= Z / N Z ). W e fix the N th ro ot of unity ω = e 2 π i N , and X , Z , the W eyl C N -op erato rs : X | n i = | n + 1 i , Z | n i = ω n | n i ( n ∈ Z N ) , whic h satisfy X N = Z N = 1 and the W eyl relation: X Z = ω − 1 Z X . 2 N -state τ (2) -mo del and quan tum group U q ( sl 2 ) The N -state τ (2) -mo del [11 , 12, 30] (also called the Baxter-Bazhano v-S t rogano v mo del [19]) is the fiv e-parameter family of L -op e rators of C 2 -auxiliary , C N -quan tum space with en tries expressed b y W eyl op erators X , Z : L ( t ) = 1 − t c b ′ b X ( 1 b − ω ac b ′ b X ) Z − t ( 1 b ′ − a ′ c b ′ b X ) Z − 1 − t 1 b ′ b + ω a ′ ac b ′ b X ! , (2.1) where t is th e sp ect ral v ariable, and a , b , a ′ , b ′ , c are non-zero complex parameters. It is kn o wn that the L - op erator (2.1) satisfies the YB equation R ( t/t ′ )( L ( t ) O aux 1)(1 O aux L ( t ′ )) = (1 O aux L ( t ′ ))( L ( t ) O aux 1) R ( t/t ′ ) for the asymm etry six-v ertex R -matrix R ( t ) = tω − 1 0 0 0 0 t − 1 ω − 1 0 0 t ( ω − 1) ( t − 1) ω 0 0 0 0 tω − 1 . ( 2.2) The mon o dromy matrix, N L ℓ =1 L ℓ ( t ) where L ℓ ( t ) = L ( t ) at site ℓ , a gain satisfies the ab o v e YB relation, w hose ω -t wisted trace defines the τ (2) -matrix: τ (2) ( t ) = tr C 2 L O ℓ =1 L ℓ ( ω t ) , (2.3) comm uting with the spin-shift op erator X (:= Q ℓ X ℓ ). The quantum group U q ( sl 2 ) is the asso cia ted C -algebra generated by K ± 1 2 , e ± with the relations K 1 2 K − 1 2 = K − 1 2 K 1 2 = 1 and K 1 2 e ± K − 1 2 = q ± 1 e ± , [ e + , e − ] = K − K − 1 q − q − 1 . (2.4) 3 Then the t w o-b y -tw o matrix with U q ( sl 2 )-en tr i es defin e s a one-parameter family of L -op erators: L ( s ) = ρ − 1 sK − 1 2 − s − 1 K 1 2 ( q − q − 1 ) e − ( q − q − 1 ) e + sK 1 2 − ρs − 1 K − 1 2 ! , ρ 6 = 0 ∈ C , (2.5) whic h satisfy the YB equation R 6v ( s/s ′ )( L ( s ) O aux 1)(1 O aux L ( s ′ )) = (1 O aux L ( s ′ ))( L ( s ) O aux 1) R 6v ( s/s ′ ) . (2.6) for the six-vertex (symmetric) R -matrix [18, 21]: R 6v ( s ) = s − 1 q − sq − 1 0 0 0 0 s − 1 − s q − q − 1 0 0 q − q − 1 s − 1 − s 0 0 0 0 s − 1 q − sq − 1 . Indeed, the YB relatio n (2.6) f o r L in (2. 5) is the necessary and sufficien t condition of the constr aint (2.4) in the d e finition of quantum group U q ( sl 2 ). Usin g th e lo cal L -op erator (2.5), one constru c ts the m o no drom y matrix N L ℓ =1 L ℓ ( s ) with en tries in ( L N U q ( sl 2 ))( s ), L O ℓ =1 L ℓ ( s ) = A L ( s ) B L ( s ) C L ( s ) D L ( s ) ! again satisfying (2.6). The structure of th e quantum affine algebra U q ( b sl 2 ) is describ ed b y the leading and lo west terms of the ab o v e entries in th e mono drom y matrix, A + = lim s →∞ ( ρ − 1 s ) − L A L ( s ) , A − = lim s → 0 ( − s ) L A L ( s ) , B ± = lim s ± 1 →∞ ( ± s ) ∓ ( L − 1) B L ( s ) q − q − 1 , C ± = lim s ± 1 →∞ ( ± s ) ∓ ( L − 1) C L ( s ) q − q − 1 , D + = lim s →∞ s − L D L ( s ) , D − = lim s → 0 ( − ρ − 1 s ) L D L ( s ) . One has A + = A − 1 − , A ∓ = D ± . Denote T − = B + , S − = B − and S + = C + , T + = C − . The op erators k − 1 0 = k 1 = A 2 − = D 2 + , e 1 = S + , f 1 = S − , e 0 = T − , f 0 = T + generate the quan tum affine algebra U q ( b sl 2 ) with the Hopf-algebra structure △ ( k i ) = k i ⊗ k i , i = 0 , 1 , △ ( e 1 ) = k 1 ⊗ e 1 + ρ − 1 e 1 ⊗ k 0 , △ ( f 1 ) = k 1 ⊗ f 1 + ρf 1 ⊗ k 0 , △ ( e 0 ) = ρ − 1 k 0 ⊗ e 0 + e 0 ⊗ k 1 , △ ( f 0 ) = ρk 0 ⊗ f 0 + f 0 ⊗ k 1 , Indeed, th e explicit expr e ssion of generators of U q ( b sl 2 ) is A − = K 1 2 ⊗ · · · ⊗ K 1 2 , A + = K − 1 2 ⊗ · · · ⊗ K − 1 2 , S ± = P L i =1 K 1 2 ⊗ · · · ⊗ K 1 2 | {z } i − 1 ⊗ e ± ⊗ K − 1 2 ⊗ · · · ⊗ K − 1 2 | {z } L − i ρ ∓ ( L − i ) , T ± = P L i =1 ρ ± ( i − 1) K − 1 2 ⊗ · · · ⊗ K − 1 2 | {z } i − 1 ⊗ e ± ⊗ K 1 2 ⊗ · · · ⊗ K 1 2 | {z } L − i . (2.7) 4 Note that (2.6) is still v alid w hen c h an ging the v ariable s by αs for a non-zero complex α . Giv en a fi nite -dimensional r ep resen tation σ : U q ( sl 2 ) − → En d ( C d ), the L -op erator of C 2 -auxiliary and C d -quan tum space, L ( s ) = σ ( L ( αs )), satisfies the YB relation (2.6). In particular, by setting the parameter ρ = 1 in (2.5), with α = q d − 2 2 and σ the s p in- d − 1 2 (highest-w eigh t) represen tation of U q ( sl 2 ) on C d = ⊕ d − 1 k =0 Ce k : K 1 2 ( e k ) = q d − 1 − 2 k 2 e k , e + ( e k ) = [ k ] e k − 1 , e − ( e k ) = [ d − 1 − k ] e k +1 , where [ n ] = q n − q − n q − q − 1 and e + ( e 0 ) = e − ( e d − 1 ) = 0, on e obtains the w ell-kno wn L -op erator of XXZ c h ai n of spin - d − 1 2 (see, e.g. [20, 26, 27] and references therein). 3 The equiv alence of τ (2) -mo dels and XXZ chai ns with cyclic rep- resen tation of U q ( sl 2 ) In this section, we consider the case when q is a ro o t of un it y . F or a N th ro ot of unit y q , th e re exists a three-parameter family of cyclic repr esentati on σ φ,φ ′ ,ε of U q ( sl 2 ), lab eled by non-zero complex n umbers φ, φ ′ and ε , whic h acts on cyclic C N -v ectors b y K 1 2 | n i = q − n + φ ′ − φ 2 | n i , e + | n i = q ε q φ + n − q − φ − n q − q − 1 | n − 1 i , e − | n i = q − ε q φ ′ − n − q − φ ′ + n q − q − 1 | n + 1 i , (3.1) (see, e.g. [13, 14]). Th e L -op erato r, L ( s ) = σ φ,φ ′ ,ε L ( s ) (3.2) giv es rise to the transfer matrix of the XXZ c h a in with the U q ( sl 2 )-cyclic represen tation σ φ,φ ′ ,ε , T ( s ) = ( ⊗ σ φ,φ ′ ,ε )(tr C 2 L O ℓ =1 L ℓ ( s )) , (3.3) whic h comm utes with K 1 2 (:= ⊗ ℓ K 1 2 ℓ , th e pro duct of lo cal K 1 2 -op erato rs ). 3.1 The iden t ifi cation of the N -state τ (2) -mo d el and the XXZ ch ain of U q ( sl 2 ) - cyclic represen tation wit h q N = 1 for o d d N In the subsection, w e consider the case N o dd, and write N = 2 M + 1. Let q b e the p rimitiv e N th ro ot- of-unit y with q − 2 = ω . Th en one can express th e cyclic represen tation (3.1) in terms of the W eyl op erato rs X , Z : K 1 2 = q φ ′ − φ 2 Z 1 2 , e + = q ε ( q φ +1 Z − 1 2 − q − φ − 1 Z 1 2 ) X − 1 q − q − 1 , e − = q − ε ( q φ ′ +1 Z 1 2 − q − φ ′ − 1 Z − 1 2 ) X q − q − 1 , hence follo w the expression: K − 1 = q φ − φ ′ Z − 1 , K − 1 2 e + = − q − φ − φ ′ 2 + ε − 1 (1 − q 2 φ +2 Z − 1 ) X − 1 q − q − 1 , K − 1 2 e − = q φ + φ ′ 2 − ε +1 (1 − q − 2 φ ′ − 2 Z − 1 ) X q − q − 1 . (3.4) 5 Using the F ourier basis c | n i = 1 N P j ∈ Z N ω − nj | j i , one ma y con v ert the W eyl operator ( Z − 1 , X ) to ( X, Z ): Z − 1 X c | 0 i , · · · , Z − 1 X d | N − 1 i = c | 0 i , · · · , d | N − 1 i X Z , (3.5) hence represen t K − 1 , K − 1 2 e ± in (3.4) b y K − 1 = q φ − φ ′ X, K − 1 2 e + = − q − φ − φ ′ 2 + ε − 1 (1 − q 2 φ +2 X ) Z − 1 q − q − 1 , K − 1 2 e − = q φ + φ ′ 2 − ε +1 (1 − q − 2 φ ′ − 2 X ) Z q − q − 1 . By the ab o ve exp ressio n and introdu c ing the sp ectral v ariable t = λ − 1 s 2 for a non-zero complex λ , the m odified L -op erator of (3.2), − s K − 1 2 L ( s ), is gauge equ iv alen t to 1 − tλρ − 1 q φ − φ ′ X q − ε (1 − q − 2 φ ′ − 2 X ) Z − tλ q ε (1 − q 2 φ +2 X ) Z − 1 − tλ + ρ q φ − φ ′ X ! , whic h is the same as the L -op erator (2.1) of τ (2) -mo del b y the follo wing id entificatio n of parameters: a = λ − 1 ρ q − φ − φ ′ − ε , a ′ = ρ q φ + φ ′ + ε +2 , b = q ε , b ′ = λ − 1 q − ε , c = ρ − 1 q φ − φ ′ , equiv alen tly q ε = b , q 2 φ = ω a ′ c b , q 2 φ ′ = b ′ ac , ρ 2 = ω aa ′ bb ′ , λ = 1 bb ′ . (3.6) This im p lie s the transfer matrix (3.3) of XXZ c hain is equiv alent to the τ (2) -transfer matrix (2.3). Note th a t the p rodu c t of lo cal op erator K − 1 ’s of XXZ c hain is no w corresp onding to a scalar m ultiple of the spin-shift op erator X , whic h commutes with the τ (2) -matrix. T he represen tations with ε = 0 , φ = φ ′ in (3.1) form the one-parameter cyclic repr ese nt ation of U q ( sl 2 ) discuss e d in [28] section 4. In p a rticular, the case φ = φ ′ = M is the s p in- N − 1 2 highest-w eigh t representat ion of U q ( sl 2 ). It is kn o wn that the d e generacy of τ (2) -mo del in CPM o ccurs in the alternating sup erin tegrable case (see, [30] section 4.3), where the τ (2) -mo del is c haracterized b y the L -op erator (2.1) with the parameter a = ω m b ′ , a ′ = ω m ′ b , c = ω n , (m , m ′ , n ∈ Z ) , or equiv alen tly , w ith the repr e sen tation parameters φ, φ ′ , ε and ρ, λ in (3.6) giv en by φ = − (m ′ + n + 1) , φ ′ = m + n , q ε = b , ρ = q − (1+m+m ′ ) , λ = 1 bb ′ . (3.7) In this case, th e re exist the normalized N th p o wer of S ± , T ± in (2.7), S ± ( N ) = S ± N [ N ]! , T ± ( N ) = S ± N [ N ]! . As φ in (3.7) is an intege r, b y the algebraic-Bethe-ansatz tec hn ique, one can show the degeneracy of the alternating sup erin tegrable τ (2) -mo del (for certain sectors) p o ssesses the symmetry algebra gen- erated by S ± ( N ) , T ± ( N ) , whic h can b e iden tified with the sl 2 -lo op algebra under certain constraint s b et wee n the intege rs m , m ′ and n, as in the discussion of the homogeneous sup erin tegrable case in [15, 23, 27 ]. Ho we v er, the precise structure ab out th e symmetry algebra for arbitrary m , m ′ , n remains unknown to b e identified. 6 3.2 XXZ c hain of U q ( sl 2 ) -cyclic represen tation with q 2 N = 1 as tw o copies of t he N -state τ (2) -mo d el for an arbitrary N By extendin g the argumen t of the previous subsection, we now identify the N -state τ (2) -mo del with a XXZ c hain for an arbitrary N . Let q b e a prim itive (2 N )-th ro ot of unity with q − 2 = ω , hence q N = − 1. Consider the thr e e p a rameter f amily of U q ( sl 2 )-cyclic rep resen tations σ φ,φ ′ ,ε in (3.1) on the vecto r space V of 2 N -cyclic v ectors: V := M { C | n i ′ | n ∈ Z 2 N } . Denote th e v ectors in V : | n i := | n i ′ + | n + N i ′ , ( n ∈ Z 2 N ) , | n i − := | n i ′ − | n + N i ′ , (0 ≤ n ≤ N − 1) . Then | n i = | n + N i . W e define | n i − for n ∈ Z by the N -p erio dic condition: | n + N i − = | n i − . Hence w e can ident ify {| n i} or {| n i − } with the canonical basis of th e ve ctor space of N -cyclic v ectors, and one has the decomp ositio n of V : V = V + ⊕ V − , V + := X n ∈ Z N C | n i , V − := X n ∈ Z N C | n i − . The spin -sh ift op erator | n i ′ 7→ | n + 1 i ′ of V is d ecomp osed as the sum of X on V + and e X on V − , where X | n i = | n + 1 i ( n ∈ Z N ) , e X | n i − = ( | n + 1 i − n 6≡ N − 1 (mo d N ) , −| n + 1 i − n ≡ N − 1 (mo d N ) . T ogether with the op erat or Z : Z | n i = ω n | n i , Z | n i − = ω n | n i − , one finds ( X , Z ) , ( e X , Z ) are W eyl op erato rs of V + or V − resp ectiv ely with X N = Z N = 1 , e X N = − 1. Under the cyclic rep resen tation σ φ,φ ′ ,ε (3.1) of U q ( sl 2 ), the subspaces V + , V − are interc hanged und e r the ge nerators K 1 2 , e ± , w it h the exp ression: K 1 2 | n i = q − n + φ ′ − φ 2 | n i − , K 1 2 | n i − = q − n + φ ′ − φ 2 | n i (0 ≤ n ≤ N − 1) , e + | n i = q ε q φ + n − q − φ − n q − q − 1 | n − 1 i − (1 ≤ n ≤ N ) , e − | n i = q − ε q φ ′ − n − q − φ ′ + n q − q − 1 | n + 1 i − ( − 1 ≤ n ≤ N − 2) , e + | n i − = q ε q φ + n − q − φ − n q − q − 1 | n − 1 i , e − | n i − = q − ε q φ ′ − n − q − φ ′ + n q − q − 1 | n + 1 i (0 ≤ n ≤ N − 1) . (3.8) Using the ab o ve form ulas, one fin d s that K − 1 and K − 1 2 e ± are op erators of V + and V − , whic h can b e exp ressed by X, Z , and e X , Z r e sp ectiv ely suc h that the relation (3.4) (usin g q ) holds for V + , and also v alid for V − when replacing X by e X . By usin g the V − -basis, | n ii := q n | n i − for 0 ≤ n ≤ N − 1, the W eyl pair ( e X , Z ) is con v erted to ( q − 1 X, Z ). One can express the op erators K − 1 2 e ± b y K − 1 2 e + | n ii = − q − φ − φ ′ 2 + ε (1 − q 2 φ +2 Z − 1 ) X − 1 q − q − 1 | n ii , K − 1 2 e − | n ii = q φ + φ ′ 2 − ε (1 − q − 2 φ ′ − 2 Z − 1 ) X q − q − 1 | n ii . 7 Using th e F ourier transform (3.5) of the b a sis {| n i} and {| n ii} , b y the s ame argument in subsection 3.1, one can sho w that − sK − 1 2 L ( s ) for the L -op erator (3.2) is gauge equiv alen t to t w o copies of the same L -op erator (2.1) of τ (2) -mo del with the parameter giv en b y (3.6) (replacing q b y q ). T herefore the transfer matrix (3.3) of XXZ c hain is equ iv alen t to the sum of t w o copies of τ (2) -transfer matrix (2.3). Note that neither one of the t w o copies of τ (2) -mo del inh e rits the U q ( sl 2 ) stru ctur e of XXZ c h ai n, in wh ic h one copy is sen t to another by (3.8). Th e symmetry structure of τ (2) -mo del related to U q ( sl 2 ) is different from that in s u bsectio n 3.1 for the o dd N case. Neverthele ss, the τ (2) -mo del with certain constrain ts on the parameter again p ossesses the sl 2 -lo op -algebra symm etry that arises from the U q ( b sl 2 )-structure of the corresp onding XXZ-mo del. In particular, in the homogeneous sup erin tegrable case, the Onsager-algebra symmetry of the d e generate τ (2) -eigenspace is extended to the symmetry of s l 2 -lo op algebra (in certain sectors). The relationship of these t w o symmetries will hop efully , though n o t immediately apparent, lead to a solution of the eigen vec tor problem in the s up er integrable CPM (for an arbitrary N ) along the line in [2, 3]. 4 Concluding Rema rks By stu dying the general solution of YB equation of the six-v ertex mo del, we fi nd a one-parameter L -op erato rs asso ciated to the quan tum group U q ( sl 2 ), whic h carries a three-parameter family of cyclic representat ion when q is a ro ot of unity . W e ha ve show ed that the XXZ c hain of U q ( sl 2 )-cyclic represent ation with q 2 N = 1 is equiv alen t to the sum of t w o copies of a N -state τ (2) -mo del. In particular, the Onsager-algebra sym m e try of the homogeneous s u perintegrable CPM is enlarged to the sl 2 -lo op -algebra symmetry induced f rom the corresp onding XXZ-mo del. Wh en N is o dd and q is a primitiv e N th r oot of u n it y , the N -state τ (2) -mo del can also b e identified with the XXZ chain of a cyclic represent ation of U q ( sl 2 ), hence the qu antum space carr ies the stru ctur e of quan tum affine algebra U q ( b sl 2 ). As a s pecial ca se, the homogeneous sup erin tegrable τ (2) -mo del in CPM is equiv alen t to the s p in-( N − 1) / 2 XXZ c hain with the canonical U q ( b sl 2 )-structure [28], by whic h the sl 2 -lo op -algebra symmetry of the sup erint egrable τ (2) -mo del was deriv ed in [23, 27]. In the recen t study of eigenv alues of the c hiral P otts mo del with alternating rapid ities [30], the τ (2) -degeneracy is also fou n d in the alternating sup erin tegrable case. By ident ifying these τ (2) -mo dels with the XXZ c hains via (3.7 ), we find the sl 2 -lo op -algebra s ymmetry f o r the degenerate τ (2) -eigenspace in certain cases. Similar to the homogeneo us CP M case [24], some twisted v ersion of Onsager- algebra sym m e try for the τ (2) -mo del could also p ossibly exist in the theory of a general alternating sup erin tegrable chiral Po tts m odel. Ho wev er, the explicit nature remains hard to kn o w pr ec isely . The exact relationship and a further symm e try stud y of general cases are now under consideration. Ac kno wledgemen ts The author wish es to thank Professor S. Kobay ashi for the invi tation of visiting U. C. Berkele y in the spring of 2008, wh e re part of th i s work w as carried ou t. This work is supp orted in part by National Science Council of T aiw an und e r Gran t No NSC 96-211 5-M-001 -004. 8 References [1] G. Alb ertini, B. M. McCo y , and J. H. H. P erk, Eigen v alue sp ect rum of the s u perintegrable c h iral Pot ts mo del, In Inte gr able system in quantum field the ory a nd statistic al me chanics, Adv. Stud. Pure Math., 19, Kinokun iy a Academic, Academic Press, Boston, MA (1989) 1–55. [2] H. Au-Y ang and J.H.H. Perk, Eige nv ectors the sup erin tegrable mo del I: sl 2 generators, J. Ph ys. A: Math. Theor. 41 (2008) 275201; arXiv: 0710.525 7. [3] H. Au-Y ang and J.H.H. Pe rk, Eigen vect ors in the su perintegrable mo del I I: ground state sector, arXiv: 0803.3029, [4] R. J. Baxter, Calcula tion of the eigen v alues of the transf e r matrix of the c hiral Po tts mo del, Pr o c. F ourth Asia- P a cific Physics Confer enc e (Seoul, Korea, 1990) V ol 1, W orld-S c ien tific, Singap ore (1991) 42–58. [5] R. J. Baxter, Ch i ral Potts mo del: eigenv alues of th e trans fer matrix, Ph ys. L ett. A 146 (1990) 110–1 14. [6] R. J . Baxter, Ch iral P otts mo del with skew ed b ound ary conditions, J. Stat. Phys. 73 (1993) 461–4 95. [7] R. J. Baxter, Int erfacial tension of th e c hiral Pott s mo del, J . Phys. A: Math. Gen. 27 (1994) 1837– 1849. [8] R. J. Baxter, The six and eight -v ertex mod el s revisited, J. Stat . Phys. 114 (2004) 43–66; cond-mat/040 3138. [9] R. J. Baxter, T ransfer matrix functional relation for the generalized τ 2 ( t q ) mo del, J. Stat. Ph ys. 117 (2004) 1–25; cond-mat/04094 93. [10] R. J. Baxter, Deriv ation of th e ord e r p arameter of the chiral Po tts mo del, Phys.Rev. Lett. 94 (2005 ) 130602 ; cond-mat/0501227 . [11] R. J. Baxte r, V.V. Bazhano v an d J .H .H. Perk, F un c tional relations for transfer matrices of the c hiral P otts mo del, I nt. J. Mo d. Phys. B 4 (1990) 803–87 0. [12] V. V. Bazhano v and Y u.G. S tr o gano v , Chiral P otts m o del as a d esc endant of the six-ve rtex mo del, J. Stat. Phys. 59 (1990) 799–817. [13] E. Date, M. Jim b o, K . Miki and T. Miw a, Cyclic represen tations of U q ( sl ( n + 1 , C )) at q N = 1, Publ. RIMS, Ky oto Univ. 27 (1991) 347–366. [14] C. DeConcini and V. G. Kac, Represen tations of quantum groups at r oots of un it y , in Op er ator Algebr a, U nita ry R epr esentations, Enveloping Algebr as, and Invariant The ory , Paris (19 89) Pr o gr ess in Mathematics 92, Birkh¨ auser, Boston, Massac husstts (1990) 471– 506. 9 [15] T. Deguc h i, K. F abricius and B. M. McCo y , The sl 2 lo o p algebra symmetry for the six-v ertex mo del at ro ots of unit y , J. Stat. Phys. 102 (2001) 701–736 ; cond-mat/991214 1. [16] T. Deguc hi: Regular XXZ Bethe states at ro ot s of unity- as highest weig ht v ectors of the sl 2 lo o p algebra at r oots of unit y , cond-mat/050356 4 v3. [17] K. F abricius and B. M. McCo y , Ev aluation parameters and Bethe r oots for the six v ertex mo del at ro ots of unity , Pr o g r ess in M athematic al Physics V ol 23, eds. M. Kashiwa ra and T. Miw a, Birkh¨ auser Boston (2002), 119–144; cond-mat/010 8057. [18] L. D. F addeev, How algebraic Bethe Ansatz w orks for integ rable mo dels, eds. A. Connes, K. Ga wedzki and J. Zinn-Justin, Quantum symmetries/ Symmetries quantiques , Pro ceedings of the Les Houc hes su mmer sc ho ol, Session LXIV, Les Houches, F rance, August 1-September 8, 1995, North-Holland, Amsterdam (1998), 149–219. [19] G . v on Gehlen, N. Iorgo v, S. Pa kuliak and V. Shadura: Baxt er-Bazhano v- Strogano v mo del: Separation of v ariables and Baxter equatio n, J. Phys. A: Math. Gen. 39 (200 6) 7257–728 2; nlin.SI/0603028 . [20] A, N. K irill o v and N. Y u. Resh e tikhin, Exact solution of the inte grable XXZ Heisen b erg mo del with arb itrary spin: I. The ground state and the excitati on sp ectrum, J. Phys. A: Math. Gen. 20 (1987) 1565 – 1595 . [21] P . P . K ulish, N. Y u. R esh e tikhin and E. K. Skly anin, Y ang Baxter equation and represent ation theory , Lett. Math. Phys. 5 (1981) 393–403. [22] B . M. McCo y and S . S. Roan, Ex citation sp ectrum and phase str ucture of the c hiral Pott s mo del. Phys. Lett. A 150 (1990) 347–354. [23] A. Nishino and T. Deguc hi, Th e L ( sl 2 ) symmetry of the Bazhano v-Strogano v mo del associ- ated with the su perintegrable chiral Pot ts mo del, Ph ys. Lett. A 35 6 (2006) 366– 370 ; cond- mat/060 5551. [24] S. S. Roan, T he Onsager algebra symmetry of τ ( j ) -matrices in the sup erint egrable c hiral P otts mo del, J. Stat. Mec h. (2005) P 0 9007; cond-mat/05056 98. [25] S. S. Roan, Bethe ans atz and symmetry in sup erinte grable c hir al P otts mo del and root-of- unit y six-vertex mo del, in Nank ai T racts in Mathematics V ol. 10, Diff er ential Ge ometry and Physics , e ds. Mo-Lin Go and W eipin g Zhang, W orld Scientific, Singap ore (2006), 39 9-409; cond-mat/051 1543. [26] S. S. Roan, Th e Q - op erator for ro ot- of-unit y symmetry in six v ertex mo del, J. Phys. A: Math. Gen. 39 (2006) 12303 -12325; cond-mat/06023 75. [27] S. S . Roan, F usion o p erators in the generalized τ (2) -mo del and r o ot-of-unit y symmetry of the XXZ spin c hain of higher spin , J. Ph ys. A: Math. Theor. 40 (2007) 1481-1 511; cond- mat/060 7258. 10 [28] S. S. Roan, The transf er matrix of s u perintegrable c h iral P otts mo del as the Q -op erator of ro ot- of-unit y XXZ chain with cyclic represen tation of U q ( sl 2 ), J. Stat. Mec h. (2007) P090 21; arXiv: 0705.2856. [29] S. S. Roan, On the equiv alen t theory of the generalized τ (2) -mo del and the c hiral Po tts mo del with t w o alternating vertic al rapidities, arXiv: 0710.276 4. [30] S. S. Roan, Bethe equation of τ (2) -mo del and eigen v alues of finite-size transf e r matrix of c h iral P otts mo del with alternating r a pidities, J. S t at. Mec h . (2008) P10001; arXiv:0805.158 5. 11
Original Paper
Loading high-quality paper...
Comments & Academic Discussion
Loading comments...
Leave a Comment