1: \documentclass{amsart}
2: \usepackage{amssymb}
3: \usepackage{amsfonts}
4: \usepackage{epsfig}
5: \usepackage{graphicx}
6: %\usepackage{pdfsync}
7: \newcounter{eqnletter}[equation]
8: \catcode`\@=11
9: \@addtoreset{equation}{section}
10: \catcode`\@=12
11:
12: \newcommand{\ea}{\setcounter{eqnletter}{1}\addtocounter{equation}{0}}
13: \def\theequation{\thesection.\arabic{equation}\alph{eqnletter}}
14: \newcommand{\half}{{\scriptstyle\frac{1}{2}}}
15: \newcommand{\ket}[1]{\big|#1\big\rangle}
16: \newcommand{\bra}[1]{\big\langle#1\big|}
17: \newcommand{\Tr}[1]{{\sf tr}\left(#1\right)}
18:
19: \begin{document} {\centerline {\LARGE {\bf Dilatation operator and Cayley
20: graphs}}} \vskip 1 cm \centerline {M.~Bonini, G.M.~Cicuta and E.~Onofri }
21: \vskip .5 cm {\small \centerline {{\it Dipartimento di Fisica, Univ. di Parma,
22: v.~G.P.~Usberti 7A, 43100 Parma, Italy }} \centerline {{\it and INFN,
23: Sezione di Milano Bicocca, Gruppo di Parma} \quad \quad } \centerline
24: {\small{\sf email}:\emph{ name(at)fis.unipr.it}}}
25:
26: \vskip .7 cm {\centerline{\bf Abstract}}
27: We use the algebraic definition of the Dilatation operator provided by Minahan, Zarembo, Beisert, Kristijansen, Staudacher, proper for single trace products of scalar fields, at leading order in the large-$N$ 't Hooft limit
28: to develop a new approach to the evaluation of the spectrum of the Dilatation operator. We discover a vast number of exact sequences of eigenstates.\\
29:
30: \vskip .7 cm
31: \section{Introduction and Summary}
32: The past five years witnessed impressive amount of work and progress in the understanding of $\mathcal{ N}=4$ superconformal Yang Mills theory with gauge group $SU(N)$ in the large-$N$ limit.\\
33:
34: \noindent
35: The most relevant results include the increasingly detailed correspondence
36: between states in this theory with string states on $AdS_5 \times S^5$, the
37: technical improvements in the evaluation of the anomalous dimension of operators
38: which led to the discovery of quantum integrability, some unexpected relations
39: with high energy sectors of quantum chromodynamics. We cannot possibly quote
40: the pertinent vast literature and we refer the reader to the papers \cite{r1}
41: \cite{r2} \cite{ple} for introduction to the subject and to the original
42: literature. Each of these outstanding results were obtained in the large-$N$ 't
43: Hooft limit, in several sectors of the theory, at several orders in loop
44: expansion, then suggesting the possibility of a complete understanding of the
45: theory.
46:
47: It seems important both for a more complete understanding of the $\mathcal{
48: N}=4$ Super Yang Mills $SU(N)$ theory at large-$N$ and for further tests of
49: the AdS/CFT Maldacena conjecture to evaluate
50: the eigenvalues of the dilatation operator for {\bf all} states product of a small number of fields
51: and for sequences made of an arbitrary number of fields.\\
52: To this goal, an essential progress was obtained by expressing the dilatation
53: operator in a way that translates the evaluation of anomalous conformal
54: dimension of states into a diagonalization problem in finite dimensional spaces,
55: then avoiding the previous cumbersome evaluation by Feynman graphs.\\
56:
57: We shall only work in the sector of states given by products of the six real
58: scalar matrix fields $\phi_j$ of the theory . An operator containing the product
59: of $n$ matrix fields like $\Tr {\phi_1 \phi_2 \cdots \phi_n}$ has conformal
60: dimension $\bigtriangleup_0=n$, at tree level. We refer this number as the main
61: quantum number of the operator or state. In this work we limit ourselves to the
62: exact analytic evaluation of the one loop contribution $\bigtriangleup_2$, to
63: the conformal dimension for these single trace states in leading order in the
64: large-$N$ 't Hooft limit.
65: Another good quantum number is parity. One may define a parity
66: operator $P$ which inverts the order of the matrix fields inside a
67: single trace, $P\, \Tr {\phi_1 \phi_2 \cdots \phi_n}=\Tr {\phi_n
68: \phi_{n-1} \cdots \phi_1}$. Since the Hamiltonian corresponding to
69: the Dilatation operator commutes with the parity operator, it is
70: possible and useful to have eigenvectors with definite parity.\\
71:
72: One may consider a set of states obtained by permuting the positions of the complex fields $Z$, $W$, $Y$, later defined in eq.(\ref{t.20}), inside a trace
73: $\Tr {Z^aW^bY^c}$. The states of this set may be regarded as basis vectors of a linear vector space invariant under the action of the one-loop Dilatation operator. We refer to this vector space as the sector $\Tr {Z^a W^b Y^c}$.
74: The dimension of the vector space increases rapidly as the main quantum number $\bigtriangleup_0=a+b+c$ increases. For instance the sector $\Tr {Z^2W^2}$ has only $2$ independent states, both of positive parity, whereas the sector $\Tr {Z^3W^2Y^2}$ has $30$ independent states. To evaluate eigenvalues and eigenvectors of the Dilatation operator in a given sector it is very useful to use basis vectors with definite parity obtained by sum and differences of pairs of the previous basis vectors, then splitting large matrices into two smaller ones.\\
75: This evaluations will be called direct diagonalization of the Dilatation operator. It is the simplest
76: procedure provided the main quantum number $\bigtriangleup_0=a+b+c$ is a small integer.
77: We provide a summary, in Appendix A, of all eigenstates of the dilatation
78: operator for low quantum number $4 \leq \bigtriangleup_0 \leq 7$.
79: This represents a useful
80: information for the understanding of the theory, for checking evaluations we
81: perform for arbitrary value of $\bigtriangleup_0$ and for any comparison with energy of states in
82: string theory.\\
83:
84:
85:
86: Possibly the most important result of our paper is the evaluation of a vast
87: number of exact sequences of eigenstates of the dilatation operator, which
88: appear as eqs. ($C.1$) and ($C.2$). To our knowledge, these sequences
89: are not known in the literature. They have the same eigenvalues of the well
90: known sequences for two impurities, reproduced in eqs. ($D.1$) and
91: ($D.2$), which may be said to belong to the sector $\Tr {Z^n \phi_a \phi_b}$
92: of the theory. Our sequences are valid for every sector of the form $\Tr
93: {Z^a W^{b} Y^{c}}$ with arbitrary values for the (non-negative) integer
94: exponents. For instance, in the sector $\Tr {Z^{n-2}W^2 Y}$, we provide very explicit expressions of exact sequences of eigenstates in eqs. ($D.3$), ($D.4$).\\
95:
96: Our method is rather different from the powerful methods
97: (superconformal algebra, integrability and Bethe ansatz) used in
98: extensive evaluations already performed with the same goal. We define
99: an auxiliary Hamiltonian, which might be called pertinent to a
100: nearest-neighbor exchange model where the number of flavours of the
101: matrix fields $\phi_j$ is unlimited and in every configuration of the
102: fields the flavours are all distinct.
103: \noindent
104:
105: Configurations are sums of permutations and the Hamiltonian changes permutations
106: into permutations. The Hamiltonian belongs to the group algebra and, not
107: surprisingly, the analysis of this auxiliary model leads the study of the
108: irreducible representations of the permutation group $S_n$, its Young
109: projectors, its group algebra. At the end we recover information pertinent
110: $\mathcal{ N}=4$ Super Yang Mills theory by trivial replacements in the results.\\
111:
112:
113: Our method may be used to discover new sequences of eigenstates of the dilatation operator, belonging to irreducible representations different from the one we studied. These would have different eigenvalues. In this case, we cannot
114: anticipate if the analytic evaluation can be carried to the end.\\
115:
116: The outline of the paper is the following : in Section $2$ we define our auxiliary model and its analysis which leads us to the evaluation of eigenvalues and eigenvector for one element of the group algebra of the symmetric group $S_n$ in several irreducible representations. Every representation will be denoted by the sequence of integer numbers
117: counting the number of boxes in horizontal rows of the Young tableaux. The eigenvalues of the Hamiltonian in the auxiliary model include the eigenvalues of the Dilatation operator of the Super Yang
118: Mills theory. For a number of irreducible representations we collected them in
119: Appendix B.
120:
121: The method to find the eigenvectors pertinent to specific representations is
122: described in Section 3 and in Appendix C where we obtain an explicit solution
123: for the representation $(n-1,1)$. This is the easiest non trivial representation
124: and contains the sequences we mentioned before, the well known ones and the new
125: ones. Section 4 and Appendix D contain the easiest replacements for the
126: general sequences.
127:
128:
129: \section{Cayley graphs}
130: It was shown by Minahan and Zarembo \cite{mz} that the action of the dilatation operator on single-trace states of the product of $n+1$ scalar matrix-fields, at one loop order in the large-$N$ 't Hooft limit, may be replaced by the matrix
131: \begin{equation}
132: \Gamma=\frac{\lambda}{16 \pi^2} \sum_{l=1}^{n+1} \left(K_{l,l+1}+2-2P_{l,l+1}\right) \quad , \quad \lambda=g^2_{YM}N
133: \label{t.1}
134: \end{equation}
135: The two operators $K_{l,l+1}$, $P_{l,l+1}$ act only on the pair of fields in the positions $(l,l+1)$ inside the single-trace string of $n+1$ scalar fields.\\
136: The operator $K_{l,l+1}$, called a trace operator, is
137: $$K_{l,l+1} \, \Tr {\phi_{\alpha_1}\cdots \phi_{\alpha_{l-1}}\phi_{\alpha_l} \phi_{\alpha_{l+1}}\phi_{\alpha_{l+2}}
138: \cdots \phi_{\alpha_{n+1}} }=
139: \delta_{\alpha_l\,,\,\alpha_{l+1} }\sum_{k=1}^6
140: \Tr {\phi_{\alpha_1}\cdots \phi_{\alpha_{l-1}} \phi_k\phi_k \phi_{\alpha_{l+2}}\cdots \phi_{\alpha_{n+1}} }$$
141: It yields zero if the pair of matrix fields at positions $l$ and $l+1$ have different flavour.\\
142: The operator $P_{l,l+1}$ exchanges the flavour of the matrix fields at positions $l$ and $l+1$ irrespective of the flavours being equal or different
143: $$P_{l,l+1} \, \Tr {\phi_{\alpha_1}\cdots \phi_{\alpha_{l-1}}\phi_{\alpha_l} \phi_{\alpha_{l+1}}\phi_{\alpha_{l+2}}
144: \cdots \phi_{\alpha_{n+1}} }=
145: \Tr {\phi_{\alpha_1}\cdots \phi_{\alpha_{l-1}}\phi_{\alpha_{l+1}} \phi_{\alpha_l}\phi_{\alpha_{l+2}}
146: \cdots \phi_{\alpha_{n+1}} }$$
147: Of course the trace operator may be neglected if the dilatation operator acts on configurations $ \Tr {\phi_{\alpha_1} \phi_{\alpha_2}\cdots \phi_{\alpha_{n+1}} }$ where the flavours $\alpha_j$ of the matrix fields are all different. This is possible only for short chains $(n+1 \leq 6)$ and leads us to the definition of a auxiliary model.\\
148:
149: Our auxiliary model is defined by the generalized dilatation operator
150: \begin{equation}
151: \Gamma=\frac{\lambda}{8 \pi^2} \,{\bf L} \quad , \quad
152: {\bf L}= \sum_{l=1}^{n+1} \left(I_{l,l+1}-P_{l,l+1}\right) ={\bf I}-{\bf A} \qquad ,
153: \qquad \lambda=g^2_{YM}N
154: \label{t.2}
155: \end{equation}
156: acting on operators $\Tr {\phi_{\alpha_1} \phi_{\alpha_2}\cdots
157: \phi_{\alpha_{n+1}} }$ where the flavours $\alpha_j$ of the matrix fields are
158: all different, that is $\{\alpha_1 , \alpha_2 ,..,\alpha_{n+1}\}$ is a
159: permutation of the set of integers $\{1,2,..,n+1\}$. Because of the cyclic
160: property of trace, the number of states is $n!$ . One may fix the
161: position of one flavour, let us choose the first, and consider the set of
162: independent states $\Tr{ \phi_1\phi_{\alpha_2}..\phi_{\alpha_{n+1}} }$ where the
163: sequence $\{\alpha_2 , \alpha_3 ,..,\alpha_{n+1}\}$ is a permutation of the
164: permutation group $S_n$ acting on the sequence $\{2,3,..,n+1\}$ .\\
165:
166: Next we proceed to evaluate eigenvalues and eigenstates for the auxiliary model. It might seem that such spectrum would provide the correct spectrum of the dilatation operator in superconformal Yang Mills theory only for short chains and just in the sector where all flavours are different. We suggest in
167: the last section that a simple replacement rule allows us to recover from the
168: analysis of the auxiliary model the corresponding information for super Yang
169: Mills in the sector $\Tr {Z^a W^b Y^c}$ for any choice of integers $a,b,c$.
170:
171: One may consider the $n!$ states
172: $\Tr { \phi_1\phi_{\alpha_2}..\phi_{\alpha_{n+1}} } $ as a basis in a vector space $V_{n!}$.
173: With the above convention of fixing flavour one in first place,
174: the form of the operator ${\bf A}= \sum_{l=1}^{n+1} P_{l,l+1}$ , writing the permutations as cycles, is
175: \begin{equation}
176: {\bf A}= \sum_{l=1}^{n+1} P_{l,l+1}=(2,3,..,n,n+1)+(2,3)+..(n,n+1)+(n+1,n,..,3,2)
177: \label{t.3}
178: \end{equation}
179: The first permutation is the inverse of the last one, whereas each transposition
180: coincides with its inverse. \footnote{ {\bf Warning about conventions.} Some
181: care is necessary when we translate results from the theory of representations
182: of the symmetric group to the present generalized Heisenberg model on a chain.\\
183: We have defined a one-to-one correspondence between states with
184: $\bigtriangleup_0=n+1$ and elements of the permutation group $S_n$ :
185: \begin{equation}
186: {\rm tr}\,[\phi_1 \phi_{\alpha_2} \cdots\phi_{\alpha_{n+1}}] \sim \left( 2 \, 3 \, \cdots n+1 \atop \alpha_2 \, \alpha_3 ..\alpha_{n+1}\right) \qquad \label{f.1}
187: \end{equation}
188: However products of exchange operators $P_{l,l+1}$ act in reverse order of the usual conventions on products of permutations. An example will illustrate it :
189: \begin{eqnarray}
190: P_{4,5}P_{3,4} {\rm tr}\,[\phi_1 \phi_5 \phi_4 \phi_3 \phi_2]=P_{4,5} {\rm tr}\,[\phi_1 \phi_5 \phi_3 \phi_4 \phi_2]=
191: {\rm tr}\,[\phi_1 \phi_5 \phi_3 \phi_2 \phi_4]\nonumber
192: \end{eqnarray}
193: It corresponds to :
194: \begin{eqnarray}
195: \left( 1 \, 2 \, 3 \, 4 \, 5 \atop 1 \, 5 \, 4 \, 3 \, 2 \right) (34)(45)= \left( 1 \, 2 \, 3 \, 4 \, 5 \atop 1 \, 5 \, 3 \, 2 \, 4 \right) \nonumber
196: \end{eqnarray}
197: As a general rule, we use the theory of irreducible representations of $S_n$,
198: the projectors related to Young tableaux, as in Appendix B and C, with the
199: generally used conventions on products of group elements. Often we consider
200: linear combinations of group elements, that is elements in the group ring. The
201: results may be translated into linear combinations of traces of products of
202: matrix fields by first taking the inverse of each permutation, then applying
203: the correspondence (\ref{f.1}). }\\
204:
205: The operator ${\bf A}$ is represented as a real symmetric matrix in the space
206: $V_{n!}$ and it is the adjacency matrix of the graph $G(V,E)$ associated to the
207: matrix. The set of vertices is the set of the $n!$ independent states $\Tr{
208: \phi_1\phi_{\alpha_2}..\phi_{\alpha_{n+1} } }$ , a link connects vertex $v_i$
209: with vertex $v_k$ if one of the $n+1$ permutations $\pi$ in the sum in
210: eq.(\ref{t.3}), is such that $v_k=\pi v_i$.\\
211:
212: The operator ${\bf L}$ given in eq.(\ref{t.2})
213: is the Laplacian of the graph. Since every vertex in the graph has
214: the same degree $n+1$, the spectrum of the adjacency matrix ${\bf A}$ is
215: trivially related to the spectrum of the Laplacian matrix ${\bf L}$. The
216: eigenvalues of ${\bf L}$, here called \boldmath $\bigtriangleup_2, $\unboldmath
217: \, provide the one loop contribution to the anomalous dimension
218: \boldmath $$
219: \bigtriangleup=\bigtriangleup_0+\mbox{\unboldmath
220: $\frac{\lambda}{8\pi^2}$}\bigtriangleup_2 \quad , \quad
221: \bigtriangleup_0=n+1 \,.$$ \unboldmath
222:
223:
224: Let us recall \cite{big} that a graph $G(V,E)$ where the set of vertices is the set of elements of a group, and the set of edges is a subset of the previous set, provided it is closed under taking the inverse, is a Cayley graph. Then the graph we are discussing in this section is the Cayley graph on the group of permutations $S_n$ with the set of $n+1$ connections listed on the right side of eq.(\ref{t.3}). The evaluation of the spectrum of the adjacency matrix of Cayley graphs even for large graphs, is greatly facilitated by its symmetries.\\
225:
226:
227: We recall that for any $n$ there exist a very easy representation of the permutation group $S_n$ of degree $n!$ . It is obtained by considering the elements $g \in S_n$ both as basis vectors as well as operators in the vector space spanned by the basis vectors. Each
228: $g \in S_n$ is represented by a matrix with only one entry equal to one and the remaining entries equal to zero in each row and in each column. This representation is sometimes called the {\bf regular} representation.\\
229: The real symmetric matrix ${\bf A}$ , of order $n!$ in eq.(\ref{t.3}) is the sum of $(n+1)$ real symmetric matrices which are the regular representation of $n-1$ transpositions and $2$ long cycles in $S_n$. It contains
230: $n+1$ entries equal to one in each row and in each column. \\
231: Any set of matrices which are the regular representation of a set of elements $g \in S_n$, allow a simultaneous block decomposition, where the matrices in the blocks are the irreducible representations of the elements $g \in S_n$.
232: A irreducible representation of degree $f$ occurs $f$ times in this decomposition \cite{boe}, then the spectrum of ${\bf A}$ is given by the eigenvalues of ${\bf A}$ in the irreducible representation of degree $f$ with multiplicity $f$ (times the multiplicity of the eigenvalue in the irreducible).\\
233: Since the regular representation is a matrix of order $n!$, the size increases too rapidly to allow a direct evaluation of eigenvalues and eigenvectors of the matrix {\bf A} beyond the smallest values of $n$. But every information on the spectrum is recovered by the analysis of the irreducible representations occurring in the block decomposition of the regular representation.\\
234:
235: The evaluation of the eigenvalues does not need writing the irreducible representation for the $n-1$ transpositions and the $2$ cycles occurring in eq.(\ref{t.3}). Indeed from the knowledge of the characters of all classes of elements of $S_n$ in a given representation of degree $f$, one obtains the character of the matrices $A$, $A^2$, \dots , $A^f$ then the characteristic equation for the matrix $A$ then its eigenvalues. However we found easier to profit from the explicit
236: irreducible representations tabulated for the generators of the permutation group $S_n$ explicitly listed \cite{enc} up to some value of $n$. \\
237:
238: \boldmath $ \bigtriangleup_0=4.$\unboldmath \, By this method we evaluate the $6$ eigenvalues of ${\bf A}$ for $n=3$ : \{
239: $\lambda=4$ singlet, $\lambda=0$ with multiplicity $3$, $\lambda=-2$ with multiplicity $2$ \}. They translate respectively to the eigenvalues of the Laplacian operator : \{$\mathbf{\bigtriangleup_2=0}$ singlet,
240: \boldmath $\bigtriangleup_2=4$ \unboldmath \,with multiplicity $3$, \boldmath $\bigtriangleup_2=6$ \unboldmath \, with multiplicity $2$ \}. The representation of dimension $6$ is partitioned into the irreducible representations $6=1+1+2^2$. The first singlet, corresponding to the identity representation is the totally symmetric eigenstate, the second singlet corresponds to the alternate representation, finally the $2$-dimensional irreducible representation provides $\lambda=0$ and $\lambda=-2$.\\
241: In a more symmetric fashion, the $3$ eigenstates with $\lambda=0$ may be chosen as
242: \begin{eqnarray}
243: u_1 &=& \Tr{\phi_1\phi_2\phi_3\phi_4-\phi_1\phi_4\phi_3\phi_2} \qquad, \nonumber \\
244: u_2 &=& \Tr{\phi_1\phi_2\phi_4\phi_3-\phi_1\phi_3\phi_4\phi_2} \qquad, \nonumber \\
245: u_3 &=& \Tr{\phi_1\phi_4\phi_2\phi_3-\phi_1\phi_3\phi_2\phi_4} \qquad
246: \label{t.3a}
247: \end{eqnarray}
248: The $2$ eigenstates with $\lambda=-2$ may be chosen as
249: \begin{eqnarray}
250: u_4 &=& \Tr{ \phi_1\phi_2\phi_3\phi_4-\phi_1\phi_2\phi_4\phi_3+
251: \phi_1\phi_4\phi_3\phi_2-\phi_1\phi_3\phi_4\phi_2}
252: \qquad, \nonumber \\
253: u_5 &=& \Tr{ \phi_1\phi_2\phi_4\phi_3-\phi_1\phi_3\phi_2\phi_4+
254: \phi_1\phi_3\phi_4\phi_2-\phi_1\phi_4\phi_2\phi_3}
255: \label{t.3b}
256: \end{eqnarray}
257: \vskip 10 pt
258: \boldmath $\bigtriangleup_0=5.$ \unboldmath \, The permutation group $S_4$ has $5$ irreducible representations : two of degree $1$,
259: one $2$-dimensional and two of degree $3$. Correspondingly the matrix ${\bf A}$, of order $24$ decomposes into blocks, $24=1+1+2^2+2 \times 3^2$.\\ The eigenvalues of the two $1$-dimensional representations are $\lambda=\pm 5$.
260: The eigenvalues associated to the other three irreducible representations are exhibited in the Appendix B. Eigenvalues and multiplicities are collected in Table $1$.\\
261:
262: \begin{center}
263: \begin{tabular}{|c||c| c|c|c|c|c||} \hline
264: $\lambda $ & $5$ & $ \sqrt{5} $ & $1$ & $-1$ & $ -\sqrt{5} $ & $-5$ \\ \hline
265: $\bigtriangleup_2$ & 0 & $5-\sqrt{5}$ & 4 & 6 & $5+\sqrt{5}$ & $10 $\\ \hline
266: multiplicity & 1 & 6 & 5 & 5 & 6 & 1\\ \hline
267: \end{tabular} \\
268:
269: \small{Table $1$. Eigenvalues of the states $\Tr {\phi_1 \phi_{\alpha_2}\phi_{\alpha_3}\phi_{\alpha_4} \phi_{\alpha_5} }$ ,
270: $\bigtriangleup_0=5$ has multiplicity$=24$. The spectrum of the Laplacian has the symmetry $\bigtriangleup_2\to 10- \bigtriangleup_2$ } \\
271: \end{center}
272: \vskip 8 pt
273: The spectrum of $\bf A$ is symmetric with respect to the origin because all the $5$ permutations occurring in eq.(\ref{t.3}) may be decomposed in a odd number of transpositions, then the graph with $24$ vertices each of degree $5$ is bipartite.\\
274: We omit listing the eigenvectors in the degenerate subspaces as they may be chosen in several ways. \\
275:
276:
277: \boldmath $\bigtriangleup_0=6.$ \unboldmath \, The permutation group $S_5$ has $7$ irreducible representations : two of degree $1$,
278: two of degree $4$, two of degree $5$, one $6$-dimensional. Correspondingly the matrix ${\bf A}$, of order $120$ decomposes into blocks $120=1+1+2\times 4^2+2 \times 5^2+6^2$.\\
279: The eigenvalues of the two $1$-dimensional representations are $\lambda=6$, $\lambda=-2$. The eigenvalues associated to the other five irreducible representations are exhibited in the Appendix B. Eigenvalues and multiplicities are collected in Table $2$.
280: \vskip 5 pt
281: \begin{center}
282: \begin{tabular}{|c||c| c|c|c|c|c|c|c|c|c||} \hline
283: $\lambda $ & $6$ & $4$ & $1+\sqrt{5} $ & $-1+\sqrt{13}$ & $1$ & $0$ & $ 1-\sqrt{5} $ & $-2$ & $-4$ &$-1-\sqrt{13}$ \\ \hline
284: $\bigtriangleup_2$ & 0 & 2 & $5-\sqrt{5}$ & $7-\sqrt{13}$ & 5 & 6 & $5+\sqrt{5}$ & $8 $ & 10 & $7+\sqrt{13}$\\ \hline
285: multiplicity & 1&10& 9 & 5 & 32 & 15 & 9 & 25 & 9 & 5\\ \hline
286: \end{tabular} \\
287: \small{Table $2$. Eigenvalues of the states
288: $\Tr {\phi_1 \phi_{\alpha_2}\phi_{\alpha_3}\phi_{\alpha_4}\phi_{\alpha_5} \phi_{\alpha_6} }$ , \\
289: $\bigtriangleup_0=6$ has multiplicity$=120$.}\\
290: \end{center}
291: \vskip 10 pt
292: \boldmath $\bigtriangleup_0=7.$ \unboldmath \, The permutation group $S_6$ has $11$ irreducible representations : two of degree $1$,
293: four of degree $5$, two of degree $9$, two of degree $10$, one $16$-dimensional. Correspondingly the matrix ${\bf A}$, of order $720$ decomposes into blocks $720=1+1+4\times 5^2+2 \times 9^2+2 \times 10^2+16^2$.\\
294: The eigenvalues of the two $1$-dimensional representations are $\lambda=\pm 7$. The eigenvalues associated to the other nine irreducible representations are exhibited in the Appendix B. Eigenvalues and multiplicities are collected in Table $3$.\\
295:
296:
297: \begin{center}
298: \begin{tabular}{|c||c| c|c|c|c|c|c|c|c|c|} \hline
299: $\lambda$ & $7$ & $7-8\sin^2(\pi/7)$ & $5$ & $7-y_1$ & $7-z_1$ & $3$ & $7-8\sin^2(2\pi/7)$ & $2$ & $7-z_2$ & $1$ \\ \hline
300: $\bigtriangleup_2$ & $0$ & $8\sin^2(\pi/7)$ & $2$ & $y_1$ & $z_1$ & $4$ & $8\sin^2(2\pi/7)$ & 5 & $z_2$ & $6$ \\ \hline
301: multiplicity & 1&15& 14 & 21 & 21 & 70 & 15 & 28 & 21 & 119 \\ \hline
302: \end{tabular}
303: \vskip 1 cm
304:
305: \begin{tabular}{|c||c| c|c|c|c|c|c|} \hline
306: $\lambda$ & $-7+8 \sin^2(3\pi/7)$ & $0$ & $7-8\sin^2(3\pi/7)$ & $-1$ & $7-y_2$ & $-2$ & $-7+8\sin^2(2\pi/7)$ \\ \hline
307: $\bigtriangleup_2$ & $14-8 \sin^2(3\pi/7)$ & 7 & $8\sin^2(3\pi/7)$ & $8$ & $y_2$ & 9 & $14-8\sin^2(2\pi/7)$ \\ \hline
308: multiplicity & 15 & 40 & 15 & 119 & 21 & 28 & 15 \\ \hline
309: \end{tabular}
310: \vskip 1 cm
311:
312: \begin{tabular}{|c||c| c|c|c|c|c||} \hline
313: $\lambda$ & $-3$ & $7-y_3$ & $7-z_3$ & $-5$ & $-7+8\sin^2(\pi/7)$ & $-7$ \\ \hline
314: $\bigtriangleup_2$ & 10 & $y_3$ & $z_3$ & 12 & $14-8\sin^2(\pi/7)$ & 14 \\ \hline
315: multiplicity & 70 & 21 & 21 & 14 & 15 & 1 \\ \hline
316: \end{tabular}
317: \vskip .5 cm
318: \small{Table $3$. Eigenvalues of the states $\Tr {\phi_{\alpha_1} \phi_{\alpha_2}\phi_{\alpha_3}\phi_{\alpha_4}\phi_{\alpha_5} \phi_{\alpha_6} \phi_{\alpha_7}}$ , \\
319: $\bigtriangleup_0=7$ has multiplicity $=720$.\\
320: $z_j$ are the roots of the cubic $E^3-20 E^2+116 E-200=0$\\
321: $y_j$ are the roots of the cubic $E^3-22 E^2+144 E-248=0$\\
322: The spectrum is symmetric with respect $\bigtriangleup_2=7$. }
323: \end{center}
324: \vskip 1 cm
325:
326: Appendix B collects all the irreducible representations for {\bf A} and its eigenvalues up to $\bigtriangleup_0=7$.
327:
328: \section{The sequences}
329:
330: For any irreducible representation, of degree $f$, one may find a suitable set of $f$ basis vectors and represent the operator {\bf A} in eq.(\ref{t.3}) as matrix of order $f$, then finding the eigenstates of the dilatation operator.
331: In this section we show the method for a simple example of fixed order. In Appendix C the method is generalized to the much more interesting case of the easiest representations for $S_n$ then obtaining sequences of eigenvectors for any $n$.
332: One finds in the literature \cite{boe} the method to evaluate the basis vectors useful for any irreducible representation: they are the left cosets obtained by multiplying each $g \in S_n$ times the Young projector $Y=PQ$ associated to the given irreducible representation.\\
333:
334: We illustrate the method with an example. Let us consider the permutation group $S_4$ and the irreducible representation associated to the partition $(3,1)$ which has degree $f=3$. The Young tableau in the figure \\
335: \epsfig{file=young3 ,height=2cm}\\
336: is associated to the Young projector $Y=PQ$, where $P$ is the totally symmetric projector $P=e+(12)+(13)+(23)+(123)+(132)$
337: and $Q$ is the antisymmetric projector $Q=e-(14)$.\\
338: Three basis vectors are determined by the products $g \cdot Y$ , with $g \in S_4$. The evaluation is shortened by a couple of simple remarks. Each $g \in S_4$ is written in one of the four sets $a_j$ , $j=1,2,3,4$
339: \begin{eqnarray}
340: a_1&=& (14) \cdot \left\{ e,(12), (13), (23), (123), (132) \right\} \quad , \nonumber\\
341: a_2&=& (24) \cdot \left\{ e,(12), (13), (23), (123), (132) \right\} \quad , \nonumber\\
342: a_3&=& (34) \cdot \left\{ e,(12), (13), (23), (123), (132) \right\} \quad , \nonumber\\
343: a_4&=& e \cdot \left\{ e,(12), (13), (23), (123), (132) \right\} \quad \nonumber
344: \end{eqnarray}
345: Furthermore if $g \in S_3 \in S_4$, $g \cdot PQ=PQ$. Then the $24$ products $g \cdot PQ$ for $g \in S_4$ fall into four cosets
346: \begin{eqnarray}
347: v_1&=& (14) \cdot PQ \quad , \nonumber\\
348: v_2&=& (24) \cdot PQ \quad , \nonumber\\
349: v_3&=& (34) \cdot PQ \quad , \nonumber\\
350: v_4&=& PQ \quad \label{t.8}
351: \end{eqnarray}
352: It is also clear that $v_1+v_2+v_3+v_4=0$ because that sum is the product of the totally symmetric projector in $S_4$ times $Q$. Then a basis of $3$ independent (although not orthogonal) vectors for the irreducible representation corresponding to the partition $(3,1)$ may be chosen as three among the four $v_j$. Let us express $v_3=-(v_1+v_2+v_4)$.\\
353: The action of the operator $
354: A=(12)+(23)+(34)+(1234)+(4321)$ on the three basis vectors $v_j$ is
355:
356: \begin{eqnarray}
357: A \, v_1 &=& 2v_1+2v_2+v_4 \qquad \nonumber\\
358: A\,v_2 &=&2 v_1+v_2 +2v_3=-v_2-2v_4 \qquad \nonumber\\
359: A \,v_4 &=& v_1+2v_3+2v_4=-v_1-2v_2
360: \nonumber
361: \end{eqnarray}
362: If $u=\alpha_1v_1+\alpha_2v_2+\alpha_4v_4$ is eigenvector of $A$ , $A\,u=\lambda\,u$, one easily finds :
363: \begin{eqnarray}
364: &&\lambda=1 \quad , \quad u_1=v_1+v_4 \qquad , \nonumber\\
365: &&\lambda=\sqrt{5} \quad , \quad u_2=
366: \frac{3+\sqrt{5}}{4}v_1+ v_2 +
367: \frac{1-\sqrt{5}}{4}v_4
368: \qquad , \nonumber\\
369: &&\lambda=-\sqrt{5} \quad , \quad u_3=
370: \frac{3-\sqrt{5}}{4}v_1+ v_2
371: +\frac{1+\sqrt{5}}{4}v_4
372: \qquad \nonumber\\
373: \nonumber
374: \end{eqnarray}
375: To compare this solution with the general solution given in eqs. ($C.1$) and ($C.2$), one rewrites the eigenvectors in the form :\\
376: \begin{eqnarray}
377: &&\lambda=1 \quad , \quad u_1= \sum_{j=1}^4\cos \frac{\pi(2j-1)}{4}v_j \sim v_1-v_2-v_3+v_4 \qquad , \qquad\nonumber\\
378: &&\lambda=1+4 \cos \frac{2\pi}{5}=\sqrt{5} \quad , \quad u_2=\sum_{j=1}^4 \sin \frac{2\pi j}{5} v_j \sim
379: \sqrt{5+\sqrt{5}}\,(v_1-v_4)+
380: \sqrt{5-\sqrt{5}}\,(v_2-v_3)
381: \qquad , \nonumber\\
382: &&\lambda=1+4 \cos \frac{4\pi}{5}=-\sqrt{5} \quad , \quad u_3=\sum_{j=1}^4 \sin \frac{4\pi j}{5} v_j \sim
383: \sqrt{5-\sqrt{5}}\,(v_1-v_4)+
384: \sqrt{5+\sqrt{5}}\,(v_3-v_2)
385: \qquad \nonumber\\
386: \nonumber
387: \end{eqnarray}
388:
389: Finally we translate the eigenvectors (\ref{t.8}) into linear combinations of traces of products of $5$ matrix fields by first adding one unit, then taking the inverse in each permutation.
390: We outline the derivation for the element $v_2=(2,4)P\left( (e-(1,4)\right)$.
391: $P$ is the sum of the $6$ permutations over the elements $\{1,2,3\}$. We manifest it with the notation $P_{S_{3}}(1,2,3)$. Furthermore $(2,4)P_{S_{3}}(1,2,3)=P_{S_{3}}(1,3,4)(2,4)$ and $(2,4)(1,4)=(1,4)(1,2)$. Then
392: $$v_2=P_{S_3}(1,3,4)(2,4)-P_{S_3}(1,3,4)(1,2)$$
393: According to the relations in Footnote 1 we add one unit in the symbols, evaluate the inverse of the elements and rewrite in terms of matrix fields
394: $$v_2 = \sum_{p \in S_{3}} \Tr {\phi_{\underline{1}}\phi_2\phi_{\underline{5}}\phi_3 \phi_{4} }-\Tr
395: {\phi_{\underline{1}}\phi_3\phi_{\underline{2}}\phi_4 \phi_{5} }$$
396: In a similar way we obtain the remaining basis vectors $v_j$. To exhibit the symmetry in compact equations, we denote underlined numbers as symbols which are fixed in the sum over permutations
397: \footnote{ For example
398: \begin{eqnarray}
399: v_3 &=& \sum_{p \in S_3} \Tr { \phi_{ \underline{1}}\phi_2 \phi_3 \phi_{\underline{5}} \phi_4-
400: \phi_{ \underline{1}} \phi_3 \phi_4 \phi_{\underline{2}}\phi_5 }=
401: \nonumber\\
402: &=& \Tr { \phi_1
403: (\phi_2\phi_3\phi_5\phi_4+\phi_2\phi_4\phi_5\phi_3+\phi_3\phi_2\phi_5\phi_4 +\phi_3\phi_4\phi_5\phi_2+\phi_4\phi_2\phi_5\phi_3+\phi_4\phi_3\phi_5\phi_2)} \nonumber\\
404: &-&\Tr {\phi_1(\phi_3\phi_4\phi_2\phi_5+\phi_3\phi_5\phi_2\phi_4+\phi_4\phi_3\phi_2\phi_5+\phi_4\phi_5\phi_2\phi_3+
405: \phi_5\phi_3\phi_2\phi_4+\phi_5\phi_4\phi_2\phi_3) }
406: \nonumber
407: \end{eqnarray} }.
408:
409: \begin{eqnarray}
410: v_1&=& \sum_{p \in S_3} \Tr {\phi_{\underline{1}}\phi_{\underline{5}}\phi_2\phi_3\phi_4-\phi_{\underline{1}}\phi_{\underline{2}}\phi_3\phi_4\phi_5} \quad , \nonumber\\
411: v_2 &=& \sum_{p \in S_3}\Tr { \phi_{ \underline{1}}\phi_2 \phi_{\underline{5}}\phi_3\phi_4-
412: \phi_{ \underline{1}}\phi_3\phi_{\underline{2}}\phi_4\phi_5 }\quad ,
413: \nonumber\\
414: v_3 &=& \sum_{p \in S_3}\Tr { \phi_{ \underline{1}}\phi_2\phi_3\phi_{\underline{5}}\phi_4-
415: \phi_{ \underline{1}}\phi_3\phi_4\phi_{\underline{2}}\phi_5 }\quad ,
416: \nonumber\\
417: v_4&=&\sum_{p \in S_3} \Tr { \phi_{ \underline{1}}\phi_2\phi_3\phi_4\phi_{\underline{5}}-
418: \phi_{ \underline{1}}\phi_3\phi_4\phi_5\phi_{\underline{2}} }
419: \quad \label{t.7}
420: \end{eqnarray}
421:
422:
423: \section{The replacements}
424:
425: In the literature it is more usual to represent states of scalar fields in term of $3$ complex fields $Z$ , $W$ , $Y$ , rather then their $6$ real components $\phi_\alpha$. One may choose
426: \begin{eqnarray}
427: &&Z=\frac{1}{\sqrt{2}}(\phi_1+i\phi_2) \quad , \quad
428: W=\frac{1}{\sqrt{2}}(\phi_3+i\phi_4) \quad , \quad
429: Y=\frac{1}{\sqrt{2}}(\phi_5+i\phi_6) \quad , \nonumber\\
430: &&{\bar Z}=\frac{1}{\sqrt{2}}(\phi_1-i\phi_2) \quad , \quad
431: {\bar W}=\frac{1}{\sqrt{2}}(\phi_3-i\phi_4) \quad , \quad
432: {\bar Y}=\frac{1}{\sqrt{2}}(\phi_5-i\phi_6) \quad \label{t.20}
433: \end{eqnarray}
434:
435: For $\bigtriangleup_0 \leq 6$ our auxiliary model does not differ from superconformal Yang Mills theory, in the sector we studied, then all the eigenstates we described, like in eqs.(\ref{t.3a}), (\ref{t.3b}) for $\bigtriangleup_0=4$
436: or eq.(\ref{t.7}) for $\bigtriangleup_0=5$, are eigenstates of the superconformal Yang Mills theory. Writing them as real matrix fields or, through eq.(\ref{t.20}) in terms of complex matrix fields is irrelevant.\\
437:
438: However the relevance of our auxiliary model depends on the possibility of yielding the spectrum of superconformal Yang Mills theory for arbitrary values of $\bigtriangleup_0$. This is performed by a trivial replacement rule.\\
439: Given an eigenvector of the auxiliary model pertinent to the group $S_n$ , that is $\bigtriangleup_0=n+1$, it is a linear combination of basis vectors $\Tr {\phi_1\phi_{\alpha_2} \cdots \phi_{\alpha_{n+1}} }$. One can partition the set of flavours $\{1, \alpha_2 , \cdots , \alpha_{n+1}\}$ in $3$ sets , say of cardinality $n_1, n_2, n_3$ such that $n_1+n_2+n_3=n+1$ and each $n_j \geq 0$. Then replace for each term of the linear combination
440: all $\phi_{\alpha_j}$ of the first set with $Z$, those of the second set with $W$ , those of the third set with $Y$.
441: In this way one obtains a state linear combination of states all in the sector $\Tr {Z^{n_1}W^{n_2}Y^{n_3} }$.
442: This linear combination is an eigenstate of the superconformal Yang Mills theory. \\
443:
444:
445: This trivial replacement rule is correct because the trace operator $\sum_{l=1}^{n+1} K_{l,l+1}$ vanishes on states of the sector
446: $\Tr {Z^{n_1}W^{n_2}Y^{n_3} } $ and the remaining part of the dilatation operator, the exchange operator $\sum P_{l,l+1}$, eq.(\ref{t.1}), acts in the same way irrespective of the replacement.\\
447:
448: Let us illustrate this vanishing for a simple example. Let us consider a state $ \Tr { \cdots Z^m \cdots }$ where all couples of adjacent matrix fields before and after $Z^m$ have different flavours. Then the possible non-vanishing contributions of the action of the trace operator is due to its action on the $m-1$ pairs of adjacent $m$ matrix fields $Z$. If the first field $Z$ occurs at position $r$ inside the trace
449: \begin{eqnarray}
450: &&\sum_{l=1}^{n+1} K_{l,l+1}\, \Tr { \cdots Z^m \cdots }=\sum_{l=r}^{r+m-1} K_{l,l+1}\, \Tr { \cdots Z^m \cdots }
451: \nonumber\\
452: &&= K_{r,r+1}\,\Tr { \cdots (Z^2)Z^{m-2} \cdots }+ K_{r+1,r+2}\,\Tr { \cdots Z(Z^2)Z^{m-3} \cdots }+ \cdots +
453: \nonumber\\
454: &&+K_{r+m-1,r+m}\,\Tr { \cdots Z^{m-2}(Z^2) \cdots }
455: \nonumber
456: \end{eqnarray}\\
457: where we have exhibited inside parenthesis the pair of adjacent $Z$ fields at the sites where the trace operator acts.
458: Each term in the sum vanishes. Indeed for a term where the pair of $Z$ fields are on sites $(l,l+1)$ we have
459: $$K_{l,l+1}\, \Tr { \cdots (Z^2) \cdots }= K_{l,l+1}\, \Tr { \cdots (\phi_1^2-\phi_2^2+i\phi_1\phi_2+i\phi_2\phi_1) \cdots }=0$$
460:
461:
462: The different choices of fields in the sets $n_j$ may lead to inequivalent sets of eigenvectors for the same sector $\Tr {Z^{n_1}W^{n_2}Y^{n_3}}$.\\
463: We illustrate how the replacement rule applies to the basis vectors
464: (\ref{t.7}) with the choice $n_1=n_2=2$ , $n_3=1$, that is
465: the sector $\Tr {Z^2W^2Y}$ \\
466:
467: {\bf The sector $\Tr{ Z^2W^2Y}$.}\\
468: In the irreducible representation $(3,1)$ of the group $S_4$ each basis vector $v_j$ vanishes if the field $\phi_2$ and $\phi_5$ are replaced with the same complex field. The general solution ($C.1$) and ($C.2$) indicates one eigenvector with positive parity and eigenvalue $\bigtriangleup_2=4$ and two eigenvectors with negative parity and eigenvalues $\bigtriangleup_2=5 \pm \sqrt{5}$. We do two inequivalent replacings : first $\{ \phi_1\,,\,\phi_2\}\to W$ , $\{ \phi_3\,,\,\phi_4\}\to Z$ , $\phi_5 \to Y$ and second
469: $\{ \phi_2\,,\,\phi_3\}\to W$ , $\{ \phi_4\,,\,\phi_5\}\to Z$ , $\phi_1 \to Y$. These different replacements originate sets of eigenvectors where the positive parity eigenvectors are inequivalent.
470: As we know all the eigenvectors in the sector $\Tr {Z^2W^2Y}$ from direct diagonalization, see Appendix A, further different replacements would not originate new eigenvectors.\\
471: The first replacement obtains the basis vectors
472: \begin{eqnarray}
473: v_1 &=& \Tr {(WZ^2W+ZWZW-ZW^2Z-W^2Z^2)Y} \quad , \nonumber\\
474: v_2 &=& \Tr {(-WZ^2W-ZWZW +ZW^2Z+Z^2W^2)Y}\quad , \nonumber\\
475: v_3 &=& \Tr {(-WZ^2W-WZWZ +ZW^2Z+W^2Z^2)Y} \quad , \nonumber\\
476: v_4 &=& \Tr {(WZ^2W+WZWZ-ZW^2Z-Z^2W^2)Y}
477: \quad \nonumber
478: \end{eqnarray}
479: They originate one positive parity eigenstate
480: \begin{eqnarray}
481: &&\bigtriangleup_2=4 \qquad , \qquad u=\sum_{j=0}^4 \cos \frac{(2j-1)\pi}{4} v_j\sim v_1-v_2-v_3+v_4= \quad \nonumber \\
482: &&\qquad =\Tr {(2WZ^2W-2ZW^2Z+ZWZW+WZWZ-W^2Z^2-Z^2W^2)Y}
483: \quad \nonumber
484: \end{eqnarray}
485: and two negative parity eigenstates
486: \begin{eqnarray}
487: &&\bigtriangleup_2=5-\sqrt{5} \qquad , \qquad u=\sum_{j=0}^4 \sin\frac{2\pi j}{5}v_j \sim \qquad \nonumber\\
488: &&\sim {\rm tr}\,[(ZWZW-WZWZ)Y]+(\sqrt{5}+2) {\rm tr}\,[
489: (Z^2W^2-W^2Z^2)Y] \qquad , \nonumber\\
490: &&\bigtriangleup_2=5+\sqrt{5} \qquad , \qquad u=\sum_{j=0}^4 \sin\frac{4\pi j}{5}v_j \sim \qquad \nonumber\\
491: &&\sim -(\sqrt{5}+2) {\rm tr}\,[(ZWZW-WZWZ)Y]+ {\rm tr}\,[
492: (Z^2W^2-W^2Z^2)Y] \qquad
493: \quad \nonumber
494: \end{eqnarray}
495:
496: The second replacement obtains the basis vectors
497: \begin{eqnarray}
498: v_1 &=& \Tr {(Z^2W^2-WZWZ-WZ^2W+ZWZW+ZW^2Z-W^2Z^2)Y} \quad , \nonumber\\
499: v_2 &=& \Tr {(-W^2Z^2+WZWZ+WZ^2W-ZWZW -ZW^2Z+Z^2W^2)Y}\quad , \nonumber\\
500: v_3 &=& \Tr {(-Z^2W^2+ZWZW+WZ^2W-WZWZ -ZW^2Z+W^2Z^2)Y} \quad , \nonumber\\
501: v_4 &=& \Tr {(W^2Z^2-ZWZW-WZ^2W+WZWZ+ZW^2Z-Z^2W^2)Y}
502: \quad \nonumber
503: \end{eqnarray}
504: They originate one positive parity eigenstate
505: \begin{eqnarray}
506: &&\bigtriangleup_2=4 \qquad , \qquad u=\Tr {(ZW^2Z-WZ^2W)Y}
507: \quad \nonumber
508: \end{eqnarray}
509: and two negative parity eigenstates equal to the previous ones.\\
510:
511:
512: Anyway the substitutions here performed are just examples, other types of substitutions are possible and lead to further eigenstates in sectors with different charges.
513: Indeed it is possible to replace just a subset of matrix fields $\phi_\alpha$ and rewrite the remaining ones in term of their defining complex fields.
514: For instance from the state $u_4$, given in eq.(\ref{t.3b})
515: \begin{eqnarray}
516: u_4=\Tr { (\phi_1 \phi_2-\phi_2\phi_1)(\phi_3\phi_4-\phi_4\phi_3)} \quad \nonumber
517: \end{eqnarray}
518: we may replace
519: $$(\phi_1 \phi_2-\phi_2\phi_1)=i({\bar W}W-W {\bar W}) \quad , \quad \phi_3 \to Z \quad , \quad \phi_4 \to Y $$
520: and obtain the new eigenstate $\Tr{({\bar W}W-W {\bar W})(ZY-YZ)}$.\\
521:
522:
523:
524:
525:
526: \section*{Appendix A. The lowest part of the spectrum}
527: Eigenstates of the dilatation operator are listed here with the main quantum number $4 \leq \bigtriangleup_0 \leq 7$ , in order of increasing values of the one loop contribution
528: \boldmath$\bigtriangleup_2$.\unboldmath \, The easiest way to obtain them is direct diagonalization of the dilatation operator in finite dimensional vector spaces invariant under its action (the sectors).\\
529:
530: Since re-labelling of fields trivially leads to equivalent eigenstates, we only list eigenstates of the form $\Tr {Z^a W^b Y^c}$ with $a \geq b \geq c$.\\
531:
532: Eigenstates with the same values of \boldmath$\bigtriangleup_0$ and
533: \boldmath$\bigtriangleup_2$ are here written as eigenstates of the parity operator. Occasionally further degeneracies occur and linear combinations of degenerate eigenstates with the same quantum numbers are equally valid.\\
534:
535: \boldmath$\bigtriangleup_0$\unboldmath$=4$ \quad , \quad
536: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \, :
537: $$ \Tr {Z^4} \quad , \quad \Tr {Z^3W} \quad , \quad 2\,\Tr {Z^2 W^2}+\Tr{ZWZW} \quad , \quad
538: \Tr {Z^2(WY+YW)}+\Tr{ZWZY} \quad .$$\\
539: \boldmath$\bigtriangleup_0$\unboldmath$=4$ \quad , \quad
540: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \, : \qquad
541: $ \Tr{Z^2(WY-YW)} $\quad . \quad \\
542:
543: \boldmath$\bigtriangleup_0$\unboldmath$=4$ \quad , \quad
544: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \, :
545: $$\Tr{Z^2 W^2} -\Tr{ZWZW} \quad , \quad
546: \Tr{Z^2(WY+YW)}-2\,\Tr{ZWZY}
547: \quad .$$\\
548:
549: \boldmath$\bigtriangleup_0$\unboldmath$=5$ \quad , \quad
550: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \, :
551: \begin{eqnarray}
552: &&\Tr {Z^5} \quad , \quad \Tr {Z^4W} \quad , \quad \Tr {Z^3 W^2+ Z^2WZW}\quad , \nonumber\\
553: &&\Tr {Z^3(WY+YW)} + \Tr{Z^2(WZY+YZW)} \quad , \quad \nonumber\\
554: &&\Tr{ Z^2(W^2Y+YW^2+WYW)+ZW(ZWY+ZYW+WZY)} \quad .
555: \nonumber
556: \end{eqnarray}\\
557:
558: \boldmath$\bigtriangleup_0$\unboldmath$=5$ \quad , \quad
559: \boldmath$\bigtriangleup_2$\unboldmath$=5-\sqrt{5}=8\sin^2 \frac{\pi}{5}$ \, :
560: \begin{eqnarray}
561: &&\left(\sqrt{5}+1\right)\Tr{Z^3(YW-WY)}+2\,\Tr{Z^2(YZW-WZY)} \quad , \nonumber\\
562: &&\left(2+\sqrt{5}\right)\Tr{Z^2(YW^2-W^2Y)}+\Tr{ZWZ(YW-WY)} \quad .\nonumber
563: \end{eqnarray}\\
564:
565: \boldmath$\bigtriangleup_0$\unboldmath$=5$ \quad , \quad
566: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \, :
567: \begin{eqnarray}
568: &&\Tr{Z^3W^2-Z^2WZW}\quad , \quad \Tr{Z^2 WYW-W^2 ZYZ} \quad , \nonumber\\
569: &&\Tr{Z^2(WZY+YZW-ZWY-ZYW)} \quad , \nonumber\\
570: &&\Tr{ZWZ(YW+WY)-Z^2(W^2Y+YW^2) } \qquad .
571: \nonumber
572: \end{eqnarray}\\
573:
574: \boldmath$\bigtriangleup_0$\unboldmath$=5$ \quad , \quad
575: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \, :
576: $$\Tr{Z^2(W^2Y+YW^2-2WYW)+WZ(WYZ+WZY-2YZW)} \quad .$$
577:
578:
579: \boldmath$\bigtriangleup_0$\unboldmath$=5$ \quad , \quad
580: \boldmath$\bigtriangleup_2$\unboldmath$=5+\sqrt{5}=8\sin^2 \frac{2\pi}{5}$ \, :
581: \begin{eqnarray}
582: &&-\left(\sqrt{5}-1\right)\Tr{Z^3(YW-WY)}+2\,\Tr{Z^2(YZW-WZY)} \quad , \nonumber\\
583: &&\left(\sqrt{5}-2 \right)\Tr{Z^2(W^2Y-YW^2)}-\Tr{ZWZ(YW-WY)} \quad .
584: \nonumber
585: \end{eqnarray}\\
586:
587: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
588: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \, :
589: \begin{eqnarray}
590: &&\Tr {Z^6} \quad , \quad \Tr {Z^5W} \quad , \quad \Tr {2Z^4 W^2+2 Z^3WZW+Z^2WZ^2W}\quad , \nonumber\\
591: && 3\, \Tr {Z^3W^3+W^2Z^2WZ+Z^2W^2ZW}+ \Tr {ZWZWZW} \qquad ,\nonumber\\
592: && \Tr {Z^4(WY+YW)+Z^3(WZY+YZW)+Z^2WZ^2Y} \qquad ,\nonumber\\
593: &&{\rm tr}\, \left( (Z^3W^2+W^2Z^3+Z^2W^2Z+ZW^2Z^2+Z^2WZW+WZWZ^2+ZWZ^2W+WZ^2WZ+ \right. \nonumber\\
594: && \qquad \left. +WZ^3W+ZWZWZ)Y \right) \quad, \nonumber\\
595: && 2 \,{\rm tr}\, \left(Z^2W^2Y^2+Y^2W^2Z^2+Z^2WYWY+YWYWZ^2+Z^2YW^2Y+W^2ZY^2Z+Z^2WY^2W+ \right.\nonumber\\
596: && \quad +ZWZWY^2+Y^2WZWZ+ZWZYWY+W^2ZYZY+YZYZW^2+\nonumber\\
597: && \quad \left. +WZWYZY+ZWYWZY \right)+ \Tr {WZYWZY+ZWYZWY} \quad .
598: \nonumber
599: \end{eqnarray}\\
600:
601: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
602: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \, :
603: \begin{eqnarray}
604: && \Tr {Z^4WY-YWZ^4+Z^3WZY-YZWZ^3} \quad , \nonumber\\
605: &&-2 \,\Tr {(Z^3W^2-W^2Z^3)Y} -\Tr {(Z^2W^2Z-ZW^2Z^2)Y}- \Tr {(Z^2WZW-WZWZ^2)Y}
606: \quad,\nonumber \\
607: && -3\,\Tr {Z^2W^2Y^2-Y^2W^2Z^2}+ {\rm tr}\, \left(
608: Z^2YWYW-WYWYZ^2+Y^2WZWZ-ZWZWY^2+ \right.\quad \nonumber\\
609: && \qquad \left.+ZYZYW^2- W^2YZYZ\right) \qquad .
610: \nonumber
611: \end{eqnarray}\\
612:
613:
614: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
615: \boldmath$\bigtriangleup_2$\unboldmath$=5-\sqrt{5}=8\sin^2 \frac{\pi}{5}$ \, :
616: \begin{eqnarray}
617: &&-(1+\sqrt{5})\,\Tr {Z^4W^2} +(\sqrt{5}-1)\,\Tr {Z^3WZW}+2 \Tr {Z^2WZ^2W}\qquad ,\nonumber\\
618: &&-2(2+\sqrt{5})\, \Tr {Z^3W^3}+(1+\sqrt{5})\, \Tr {Z^2W^2ZW+W^2Z^2ZW}+2\, \Tr {ZWZWZW} \quad , \nonumber\\
619: &&-(1+\sqrt{5}) \Tr {Z^4(WY+YW)}+(\sqrt{5}-1) \Tr {Z^3(WZY+YZW)}+4\, \Tr {Z^2WZ^2Y} \quad , \nonumber\\
620: && 2\,\Tr {(Z^3W^2+W^2Z^3-Z^2W^2Z-ZW^2Z^2)Y}+ (\sqrt{5}-3)\, \Tr {(Z^2WZW+WZWZ^2)Y}+ \nonumber\\
621: && \qquad +(5-3\sqrt{5})\, \Tr {ZWZWZY}+(1+\sqrt{5})\,\Tr {WZ^3WY}\quad ,\nonumber\\
622: && -(\sqrt{5}+1)\, \Tr {(Z^2W^2Z+ZW^2Z^2)Y} +2\, \Tr {(ZWZ^2W+WZWZ^2)Y}+\nonumber\\
623: && \qquad +(\sqrt{5}-3)\, \Tr {ZWZWZY}+(1+\sqrt{5})\, \Tr {WZ^3WY} \qquad ,\nonumber\\
624: && 2\,\Tr {W^2ZYZY+YZYZW^2-Y^2ZWZW-WZWZY^2}+\nonumber\\
625: && \qquad +(\sqrt{5}+1)\,\Tr {Z^2YW^2Y-Z^2WY^2W}+(\sqrt{5}-1)\,\Tr {ZWYWZY-ZWZYWY} \quad ,\nonumber\\
626: && (\sqrt{5}-1)\, \Tr {WZYWZY+ZWYZWY}-2\sqrt{5}\, \Tr {W^2ZYZY+YZYZW^2}+\nonumber\\
627: && \qquad +2\,\Tr {Z^2WYWY+YWYWZ^2+Y^2ZWZW+WZWZY^2-Z^2W^2Y^2-Y^2W^2Z^2}+\nonumber\\
628: && \qquad +4\,\Tr {ZWZYWY-Z^2YW^2Y-W^2ZY^2Z}
629: +2(1+\sqrt{5})\,\Tr {Z^2WY^2W} \quad ,\nonumber\\
630: && 2\,\Tr {W^2ZYZY+YZYZW^2-Z^2WYWY-YWYWZ^2}+\nonumber\\
631: && \qquad +(\sqrt{5}-1)\,\Tr {WZWYZY-ZWZYWY}+(\sqrt{5}+1)\,\Tr {W^2ZY^2Z-Z^2WY^2W} \quad .
632: \nonumber
633: \end{eqnarray}\\
634:
635: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
636: \boldmath$\bigtriangleup_2$\unboldmath$=7-\sqrt{13}$ \, :
637: \begin{eqnarray}
638: && 2(4+\sqrt{13})\,\Tr {Z^2W^2Y^2+Y^2W^2Z^2}-2(\sqrt{13}+3)\,\Tr {Z^2YW^2Y+W^2ZY^2Z+Y^2WZ^2W}+\nonumber\\
639: && \qquad -2\,\Tr {Z^2WYWY+YWYWZ^2+Y^2ZWZW+WZWZY^2+W^2ZYZY+YZYZW^2}+\nonumber\\
640: && \qquad +4\,\Tr {ZWZYWY+WZWYZY+ZWYWZY}+(\sqrt{13}+1)\,\Tr {WZYWZY+ZWYZWY} \,.
641: \nonumber
642: \end{eqnarray}\\
643:
644: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
645: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \, : $\Tr {Z^3(WZY-YZW)}$ \\
646:
647: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
648: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \, :
649: \begin{eqnarray}
650: && \Tr {(Z^3W^2+W^2Z^3-Z^2W^2Z-ZW^2Z^2+Z^2WZW+WZWZ^2)Y}-2\,\Tr {WZ^3WY}\quad , \nonumber\\
651: && \Tr {(Z^3W^2-W^2Z^3-Z^2W^2Z+ZW^2Z^2-Z^2WZW+WZWZ^2)Y} \quad , \nonumber\\
652: && 2\, \Tr {Z^2WY^2W-Z^2YW^2Y+ZWYWZY-ZWZYWY }+\nonumber\\
653: && \qquad +\Tr {W^2ZYZY+YZYZW^2-ZWZWY^2-Y^2WZWZ } \quad , \nonumber\\
654: && 2\, \Tr {Z^2WY^2W-W^2ZYY^2Z+ WZWYZY-ZWZYWY }+\nonumber\\
655: && \qquad +\Tr {W^2ZYZY+YZYZW^2-Z^2WYWY-YWYWZ^2 } \quad , \nonumber\\
656: && \Tr {Z^2WYWY-YWYWZ^2+W^2ZYZY-YZYZW^2} \quad , \nonumber\\
657: && \Tr {Z^2WYWY-YWYWZ^2+Y^2WZWZ-ZWZWY^2} \qquad .
658: \nonumber
659: \end{eqnarray}\\
660:
661: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
662: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \, :
663: \begin{eqnarray}
664: && \Tr {Z^2W^2ZW-W^2Z^2WZ } \quad , \quad \Tr {Z^4WY-YWZ^4-Z^3WZY+YZWZ^3} \quad , \nonumber\\
665: && \Tr {(ZWZ^2W-WZ^2WZ)Y} \quad, \quad
666: \Tr {(Z^2W^2Z-ZW^2Z^2-Z^2WZW+WZWZ^2)Y} \quad, \nonumber\\
667: && {\rm tr}\, \left(Z^2W^2Y^2-Y^2W^2Z^2+Z^2YWYW-WYWYZ^2+Y^2WZWZ-ZWZWY^2+ \right.\quad , \nonumber\\
668: && \qquad \left.+ZYZYW^2- W^2YZYZ\right) \qquad , \qquad \Tr {WZYWZY-ZWYZWY} \qquad .
669: \nonumber
670: \end{eqnarray}\\
671:
672:
673: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
674: \boldmath$\bigtriangleup_2$\unboldmath$=5+\sqrt{5}$ \, :
675: \begin{eqnarray}
676: &&(\sqrt{5}-1)\,\Tr {Z^4W^2} -(\sqrt{5}+1)\,\Tr {Z^3WZW}+2 \Tr {Z^2WZ^2W} \qquad ,\nonumber\\
677: &&-2(2-\sqrt{5})\, \Tr {Z^3W^3}+(1-\sqrt{5})\, \Tr {Z^2W^2ZW+W^2Z^2ZW}+2\, \Tr {ZWZWZW} \quad , \nonumber\\
678: &&(\sqrt{5}-1)\, \Tr {Z^4(WY+YW)}-(\sqrt{5}+1)\, \Tr {Z^3(WZY+YZW)}+4\,\Tr {Z^2WZ^2Y} \quad , \nonumber\\
679: && 2\,\Tr {(Z^3W^2+W^2Z^3-Z^2W^2Z-ZW^2Z^2)Y}- (\sqrt{5}+3)\, \Tr {(Z^2WZW+WZWZ^2)Y}+ \nonumber\\
680: && \qquad +(5+3\sqrt{5})\, \Tr {ZWZWZY}+(1-\sqrt{5})\,\Tr {WZ^3WY}\quad ,\nonumber\\
681: && (\sqrt{5}-1)\, \Tr {(Z^2W^2Z+ZW^2Z^2)Y} +2\, \Tr {(ZWZ^2W+WZWZ^2)Y}+\nonumber\\
682: && \qquad -(\sqrt{5}+3)\, \Tr {ZWZWZY}+(1-\sqrt{5})\, \Tr {WZ^3WY} \qquad ,\nonumber\\
683: && 2\,\Tr {ZWZWY^2+Y^2WZWZ-W^2ZYZY-YZYZW^2}+\nonumber\\
684: && \qquad -(\sqrt{5}-1)\,\Tr {Z^2YW^2Y-Z^2WY^2W}+ (\sqrt{5}+1)\, \Tr {ZWYWZY-ZWZYWY} \quad ,\nonumber\\
685: && (1+\sqrt{5})\, \Tr {WZYWZY+ZWYZWY}-2\sqrt{5}\, \Tr {W^2ZYZY+YZYZW^2}+\nonumber\\
686: && \qquad -2\,\Tr {Z^2WYWY+YWYWZ^2+Y^2ZWZW+WZWZY^2-Z^2W^2Y^2-Y^2W^2Z^2}+\nonumber\\
687: && \qquad -4\,\Tr {ZWZYWY-Z^2YW^2Y-W^2ZY^2Z}
688: +2(\sqrt{5}-1)\,\Tr {Z^2WY^2W} \quad ,\nonumber\\
689: && -2\,\Tr {W^2ZYZY+YZYZW^2-Z^2WYWY-YWYWZ^2}+\nonumber\\
690: && \qquad +(\sqrt{5}+1)\,\Tr {WZWYZY-ZWZYWY}+(\sqrt{5}-1)\,\Tr {W^2ZY^2Z-Z^2WY^2W} \quad .
691: \nonumber
692: \end{eqnarray}\\
693:
694: \boldmath$\bigtriangleup_0$\unboldmath$=6$ \quad , \quad
695: \boldmath$\bigtriangleup_2$\unboldmath$=7+\sqrt{13}$ \, :
696: \begin{eqnarray}
697: && 2(4-\sqrt{13})\,\Tr {Z^2W^2Y^2+Y^2W^2Z^2}+2(\sqrt{13}-3)\,\Tr {Z^2YW^2Y+W^2ZY^2Z+Y^2WZ^2W}+\nonumber\\
698: && \qquad -2\,\Tr {Z^2WYWY+YWYWZ^2+Y^2ZWZW+WZWZY^2+W^2ZYZY+YZYZW^2}+\nonumber\\
699: && \qquad +4\,\Tr {ZWZYWY+WZWYZY+ZWYWZY}-(\sqrt{13}-1)\,\Tr {WZYWZY+ZWYZWY} \,.
700: \nonumber
701: \end{eqnarray}\\
702:
703: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^5 W^2\}$ , eigenstates with positive parity :\\
704: \noindent
705: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u= \Tr {Z^5 W^2+Z^4WZW+Z^3WZ^2W}$ \quad , \\
706: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=\Tr {Z^5 W^2-Z^3WZ^2W}$ \quad ,\\
707: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=\Tr {Z^5 W^2}- 2\,\Tr {Z^4WZW}+\Tr{Z^3WZ^2W}$\\
708:
709:
710: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^5 WY\}$ , eigenstates with positive parity :\\
711: \noindent
712: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u=\sum_{j=1}^6 \Tr {Z^{j-1}YZ^{6-j}W}$ \quad , \\
713: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=\Tr {Z^5(WY+YW)-Z^3(WZ^2Y+YZ^2W)}$ \quad ,\\
714: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=\Tr {Z^5(WY+YW)}- 2\,\Tr {Z^4(WZY+YZW)}+\Tr{Z^3(WZ^2Y+YZ^2W)}$\\
715:
716: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^5 WY\}$ , eigenstates with negative parity :\\
717: \noindent
718: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 (\pi/7) $ \quad , \quad $u= \sin(2\pi/7)\,\Tr {Z^5(WY-YW)}+
719: \sin(4\pi/7)\,\Tr {Z^4(WZY-YZW)}+$
720: ${ }\qquad + \sin(6\pi/7)\,\Tr {Z^3(WZ^2Y-YZ^2W)}$\quad ,\\
721: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 (2\pi/7) $ \quad , \quad $u= -\sin(4\pi/7)\,\Tr {Z^5(WY-YW)}+
722: \sin(6\pi/7)\,\Tr {Z^4(WZY-YZW)}+$
723: ${ }\qquad + \sin(2\pi/7)\,\Tr {Z^3(WZ^2Y-YZ^2W)}$\quad ,\\
724: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 (3\pi/7) $ \quad , \quad $u= \sin(6\pi/7)\,\Tr {Z^5(WY-YW)}-
725: \sin(2\pi/7)\,\Tr {Z^4(WZY-YZW)}+$
726: ${ }\qquad + \sin(4\pi/7)\,\Tr {Z^3(WZ^2Y-YZ^2W)}$\quad .\\
727:
728: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^4W^3\}$ , eigenstates with positive parity :\\
729: \noindent
730: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u=\Tr {Z^4W^3+Z^3W^2ZW+W^2Z^3WZ+Z^2W^2Z^2W+Z^2WZWZW}$\\
731: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=\Tr {-2\,Z^4W^3+Z^2W^2Z^2W+Z^2WZWZW}$\\
732: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=\Tr {2\,Z^4W^3-3(Z^3W^2Z W+W^2Z^3WZ)+2\,Z^2W^2Z^2W+2\,Z^2WZWZW}$\\
733: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=\Tr {-Z^2W^2Z^2W+Z^2WZWZW}$\\
734:
735: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^4W^3\}$ , eigenstates with negative parity :\\
736: \noindent
737: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=\Tr {Z^3W^2ZW-W^2Z^3WZ}$\\
738:
739:
740: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^4W^2Y\}$.
741: Basis vectors with positive parity : \\
742: $v_1= \Tr {(Z^4W^2+W^2Z^4)Y }$ , $v_2=\Tr {(Z^3W^2Z+ZW^2Z^3)Y}$ , $v_3=\Tr {(Z^3WZW+WZWZ^3)Y }$, $v_4=\Tr {Z^2W^2Z^2Y}$ , $v_5= \Tr {(Z^2WZWZ+ZWZWZ^2)Y}$ , $v_6= \Tr {(Z^2WZ^2W+WZ^2WZ^2)Y}$, $v_7=\Tr {ZWZ^2WZY}$ , $v_8=\Tr {(ZWZ^3W+WZ^3WZ)Y}$ , $v_9=\Tr {WZ^4WY}$ .\\
743: \noindent
744: Eigenvalues and eigenstates : \\
745: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u=\sum_1^9 v_j$ \quad , \\
746: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=2(v_1+v_9)-(v_4+v_5+v_6+v_7)$ \quad , \\
747: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=-(v_1+v_2+v_4)+(v_6+v_7+v_8)$ \quad , \\
748: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=-v_1-v_3-v_6+v_8+2(v_4+v_9)$ \quad , \\
749: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=2(v_1+v_4+v_5+v_6+v_7+v_9)-3(v_2+v_3+v_8)$ \quad , \\
750: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=-v_1+v_2+v_5-v_8-2(v_4-v_9)$ \quad , \\
751: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_2-v_3-v_4+v_6$ \quad ,\\
752: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_2-v_3-v_5+v_6$ \quad , \\
753: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=-(v_1+v_5)+3(v_3-v_8)+2(v_4+v_9)-4v_5+8v_7$ \quad , \\
754:
755: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^4W^2Y\}$.
756: Basis vectors with negative parity : \\
757: $v_1= \Tr {(Z^4W^2-W^2Z^4)Y }$ , $v_2=\Tr {(Z^3W^2Z-ZW^2Z^3)Y}$ , $v_3=\Tr {(Z^3WZW-WZWZ^3)Y }$, $v_4= \Tr {(Z^2WZWZ-ZWZWZ^2)Y}$ , $v_5= \Tr {(Z^2WZ^2W-WZ^2WZ^2)Y}$ , $v_6=\Tr {(ZWZ^3W-WZ^3WZ)Y}$ .\\
758: \noindent
759: Eigenvalues and eigenstates : \\
760: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 \pi/7$ \quad , \quad $u=\sin (2\pi/7) \left(v_1+v_3+v_5+v_6 \right)+
761: \sin (4\pi/7) \left(v_1+v_2+v_4-v_6 \right)+$ \nonumber\\
762: $ \qquad +\sin (6\pi/7) \left(v_2+v_3-v_4-v_5 \right)$ \quad ,\\
763: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=-2v_1+v_2+v_3+v_4+v_5$ \quad ,\\
764: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 2\pi/7$ \quad , \quad $u=\sin (4\pi/7) \left(v_1+v_3+v_5+v_6 \right)+
765: \sin (8\pi/7) \left(v_1+v_2+v_4-v_6 \right)+$ \nonumber\\
766: $ \qquad +\sin (12\pi/7) \left(v_2+v_3-v_4-v_5 \right)$ \quad ,\\
767: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_2-v_3+v_6$ \quad ,\\
768: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 3\pi/7$ \quad , \quad $u=\sin (6\pi/7) \left(v_1+v_3+v_5+v_6 \right)+
769: \sin (12\pi/7) \left(v_1+v_2+v_4-v_6 \right)+$ \nonumber\\
770: $ \qquad +\sin (4\pi/7) \left(v_2+v_3-v_4-v_5 \right)$ \quad ,\\
771: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=-v_2+v_3+3(v_4-v_5)+2v_6$ \quad \\
772:
773: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^3W^3Y\}$.
774: Basis vectors with positive parity : \\
775: $v_1=\Tr {Z^3W^3Y+YW^3Z^3} \quad , \quad v_2=\Tr {Z^2W^3ZY+YZW^3Z^2} \quad , $\\
776: $ v_3=\Tr {Z^2W^2ZWY+YWZW^2Z^2} \quad , \quad v_4= \Tr {Z^2WZW^2Y+YW^2ZWZ^2} \quad , \quad $\\
777: $v_5=\Tr {ZW^2ZWZY+YZWZW^2Z} \quad , \quad v_6=\Tr {ZW^2Z^2WY+YWZ^2W^2Z} \quad ,$\\
778: $ v_7=\Tr {ZWZ^2W^2Y+YW^2Z^2WZ} \quad , \quad v_8=\Tr {ZWZWZWY+YWZWZWZ} \quad ,$\\
779: $ v_9=\Tr {W^2Z^3WY+YWZ^3W^2} \quad , \quad v_{10}=\Tr {WZ^2WZWY+YWZWZ^2W} \quad .$\\
780: \noindent
781: Eigenvalues and eigenstates : \\
782: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u=\sum_1^{10} v_j$ \quad , \\
783: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=2(v_1+v_9)-(v_3+v_5+v_6+v_8)$ \quad , \\
784: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=2(v_1+v_2)-(v_6+v_7+v_8+v_{10})$ \quad , \\
785: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=v_1-v_2+v_4-v_9$ \quad , \\
786: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_2+v_7+v_{10}-2(v_1+v_6+v_8)-3v_4$ \quad , \\
787: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_3+v_5+v_9-2(v_1+v_6+v_8)-3v_4$ \quad , \\
788: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_3-v_5$ \quad ,\\
789: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_6-v_8$ \quad , \\
790: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_7-v_{10}$ \quad , \\
791: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=v_1-v_2-v_9+2(v_3+v_5-v_6+v_7-v_8+v_{10})-3v_4$ \quad . \\
792:
793: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^3W^3Y\}$.
794: Basis vectors with negative parity : \\
795: $v_1=\Tr {Z^3W^3Y-YW^3Z^3} \quad , \quad v_2=\Tr {Z^2W^3ZY-YZW^3Z^2} \quad , $\\
796: $ v_3=\Tr {Z^2W^2ZWY-YWZW^2Z^2} \quad , \quad v_4= \Tr {Z^2WZW^2Y-YW^2ZWZ^2} \quad , \quad $\\
797: $v_5=\Tr {ZW^2ZWZY-YZWZW^2Z} \quad , \quad v_6=\Tr {ZW^2Z^2WY-YWZ^2W^2Z} \quad ,$\\
798: $ v_7=\Tr {ZWZ^2W^2Y-YW^2Z^2WZ} \quad , \quad v_8=\Tr {ZWZWZWY-YWZWZWZ} \quad ,$\\
799: $ v_9=\Tr {W^2Z^3WY-YWZ^3W^2} \quad , \quad v_{10}=\Tr {WZ^2WZWY-YWZWZ^2W} \quad .$\\
800: Eigenvalues and eigenstates : \\
801: \boldmath$\bigtriangleup_2$\unboldmath$=8\,\sin^2 \pi/7$ \quad , \quad $u= \left( \sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7 \right)v_1+
802: (\sin 4\pi/7)(v_2-v_9)+(\sin 2\pi/7)(v_3+v_7)+
803: \left(\sin 2\pi/7+\sin 4\pi/7-\sin 6\pi/7 \right)v_4+(\sin 6\pi/7)(-v_5+v_{10})+ \left(\sin 2\pi/7-\sin 4\pi/7-\sin 6\pi/7\right)v_6+
804: \left(\sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7\right)v_8$ \quad ,\\
805: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=v_2+v_3-v_7+v_9$ \quad , \\
806: \boldmath$\bigtriangleup_2$\unboldmath$=8\,\sin^2 2\pi/7$ \quad , \quad $u=
807: \left( \sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7 \right)v_1+
808: (\sin 6\pi/7)(v_2-v_9)-(\sin 4\pi/7)(v_3+v_7)+
809: \left(-\sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7 \right)v_4+(\sin 2\pi/7)(-v_5+v_{10})- \left(\sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7\right)v_6+
810: \left(\sin 2\pi/7-\sin 4\pi/7-\sin 6\pi/7\right)v_8$ \quad ,\\
811: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_2-v_4+v_7-v_{10}$ \quad , \\
812: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_3-v_4+v_5-v_9$ \quad , \\
813: \boldmath$\bigtriangleup_2$\unboldmath$=8\,\sin^2 3\pi/7$ \quad , \quad $u=\left( -\sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7\right)v_1+\left(\sin 2\pi/7 \right)\left(v_9-v_2 \right)+(\sin 6\pi/7)\left(v_3+v_7\right)-\left(\sin2\pi/7+\sin 4\pi/7-\sin 6\pi/7 \right)v_4+(\sin 4\pi/7) (-v_5+v_{10})+$\\
814: $+ \left(\sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7\right)v_6+
815: \left(\sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7\right)v_8$ \quad ,\\
816: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=v_2-v_3+v_7+v_9+2(v_5+v_{10})$ \quad , \\
817: The remaining $3$ eigenvectors correspond to the eigenvalues \boldmath$\bigtriangleup_2$\unboldmath$=7-\lambda$ where $\lambda$ are the roots of the equation : $\lambda^3-\lambda^2-17 \lambda+25=0$.\\
818:
819: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^3W^2Y^2\}$.
820: Basis vectors with positive parity : \\
821: $v_1=\Tr {Z^3W^2Y^2+Y^2W^2Z^3} \quad , \quad v_2=\Tr {Z^2W^2ZY^2+Y^2 ZW^2Z^2} \quad , $\\
822: $ v_3=\Tr {Z^2W ZWY^2+Y^2WZW Z^2} \quad , \quad v_4= \Tr {ZWZ^2WY^2+Y^2WZ^2WZ} \quad , \quad $\\
823: $v_5=\Tr {ZWZWZY^2} \quad , \quad v_6=\Tr {WZ^3WY^2} \quad ,$\\
824: $ v_7=\Tr {Z^3WYWY+YWYW Z^3} \quad , \quad v_8=\Tr {Z^2WZYWY+YWYZWZ^2} \quad ,$\\
825: $ v_9=\Tr {Z^2W^2YZY+YZYW^2Z^2} \quad , \quad v_{10}=\Tr {ZWZWYZY+YZYWZWZ} \quad ,$\\
826: $v_{11}=\Tr {ZW^2ZYZY} \quad , \quad v_{12}=\Tr {WZ^2WYZY} \quad , \quad v_{13}=\Tr {Z^3YW^2Y} \quad ,$\\
827: $v_{14}=\Tr {Z^2WYWZY+YZWYWZ^2} \quad , \quad v_{15}=\Tr {Z^2WYZWY+YWZYWZ^2} \quad ,$\\
828: $v_{16}=\Tr {ZWZYWZY+YZWYZWZ} \quad , \quad v_{17}=\Tr {ZW^2YZ^2Y+YZ^2YW^2Z} \quad,$\\
829: $v_{18}=\Tr {WZWYZ^2Y} \quad.$\\
830: Eigenvalues and eigenstates : \\
831: \boldmath$\bigtriangleup_2$\unboldmath$=0$ \quad , \quad $u=\sum_1^{18} v_j \quad ,$\\
832: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=-2v_1+4v_2+v_3+3v_5-4(v_6+v_7)-v_8+v_9+2v_{10}+3v_{11}+v_{12}-4v_{13}-v_{14}+v_{18} \quad ,$\\
833: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=v_1-2v_2-v_3-v_4-2v_5+v_6+2v_7-v_{10}-v_{11}-v_{12}+3v_{13}+v_{14}+v_{17} \quad ,$\\
834: \boldmath$\bigtriangleup_2$\unboldmath$=2$ \quad , \quad $u=-3v_2-v_3-2v_5+2v_6+2v_7+v_8-v_9-v_{10}-2v_{11}+2v_{13}+v_{14}+v_{15}+v_{16} \quad , $\\
835: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=v_1-2v_2-v_3+v_4+2v_6-v_7-2v_8+v_9+v_{10}+2v_{12}-2v_{13}-v_{16}+v_{17}+2v_{18} \quad ,$\\
836: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=-v_3+2v_6-v_8+v_9-2v_{13}+v_{14} \quad ,$\\
837: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_1+v_2-3v_4-2v_5+4v_6+v_7-3v_9+v_{10}+4v_{11}-2v_{12}-2v_{13}-3v_{14}+v_{15}+v_{16}+4v_{18} \quad ,$\\
838: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=-v_1-v_2+2v_3+v_4-2v_5-v_7+2v_8+v_9-v_{10}-2v_{12}-2v_{13}+v_{14}-v_{15}-v_{16}+2v_{17}\quad ,$\\
839: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_5-v_{11}-v_{12}+v_{18} \quad ,$\\
840: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_1-2v_3+v_4+2v_5-v_6+v_8-v_9-v_{10}+v_{11}-v_{12}-v_{13}+v_{17} \quad ,$\\
841: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_3+
842: v_6-v_7-v_8-v_{11}-v_{12}+v_{13}+v_{16} \quad ,$\\
843: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_2+
844: v_3-2v_5+v_6-v_7-v_8-v_{10}-v_{11}+v_{12}+v_{13}+v_{15} \quad ,$\\
845: \boldmath$\bigtriangleup_2$\unboldmath$=6$ \quad , \quad $u=v_3-v_8-v_9+v_{14} \quad ,$\\
846: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=-2v_1+
847: 4v_2+v_3+2v_4-4v_5-6v_6+10v_7-5v_8-v_9-2v_{10}-12v_{11}+8v_{12}-10v_{13}-7v_{14}-8v_{15}+10(v_{16}+v_{17}) \quad ,$\\
848: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=-4v_1+8v_2-3v_3+24v_4-28v_5-22v_6+20v_7-15v_8+3v_9-4(v_{10}+v_{11}+v_{12})-10 v_{13}-9v_{14}-16 v_{15}+20(v_{16}+v_{18})$ \quad . \\
849: The remaining $3$ eigenvectors correspond to the eigenvalues
850: \boldmath$\bigtriangleup_2$\unboldmath$=7-\lambda$ where $\lambda$ are the roots of the equation : $\lambda^3+\lambda^2-17\lambda-25=0$.\\
851:
852:
853: \boldmath$\bigtriangleup_0$\unboldmath$=7$ \quad , \quad Sector $\{Z^3W^2Y^2\}$.
854: Basis vectors with negative parity : \\
855: $v_1=\Tr {Z^3W^2Y^2-Y^2W^2Z^3} \quad , \quad v_2=\Tr {Z^2W^2ZY^2-Y^2 ZW^2Z^2} \quad , $\\
856: $ v_3=\Tr {Z^2W ZWY^2-Y^2WZW Z^2} \quad , \quad v_4= \Tr {ZWZ^2WY^2-Y^2WZ^2WZ} \quad , \quad $\\
857: $ v_5=\Tr {Z^3WYWY-YWYW Z^3} \quad , \quad v_6=\Tr {Z^2WZYWY-YWYZWZ^2} \quad ,$\\
858: $ v_7=\Tr {Z^2W^2YZY-YZYW^2Z^2} \quad , \quad v_{8}=\Tr {ZWZWYZY-YZYWZWZ} \quad ,$\\
859: $v_{9}=\Tr {Z^2WYWZY-YZWYWZ^2} \quad , \quad v_{10}=\Tr {Z^2WYZWY-YWZYWZ^2} \quad ,$\\
860: $v_{11}=\Tr {ZWZYWZY-YZWYZWZ} \quad , \quad v_{12}=\Tr {ZW^2YZ^2Y-YZ^2YW^2Z} \quad.$\\
861: Eigenvalues and eigenstates : \\
862: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 \pi/7 $ \quad , \quad $u=(2 \sin 2\pi/7 +2 \sin 4\pi/7+\sin 6\pi/7)v_1+
863: (\sin 2\pi/7+\sin 4\pi/7)v_2+(\sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7)(v_3+v_7)+ (\sin 6\pi/7 )(v_4+v_{12})+(2\sin 4\pi/7-\sin 6\pi/7)v_5+(\sin 2\pi/7-\sin 4\pi/7+2\sin 6\pi/7)(v_6+v_9)+(-\sin 2\pi/7+3\sin 4\pi/7-3\sin 6\pi/7)v_8+(-\sin 2\pi/7+\sin 4\pi/7)v_{10}+(\sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7)v_{11} \quad , $\\
864: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=-v_3-v_4+v_7+v_{12} \quad ,$\\
865: \boldmath$\bigtriangleup_2$\unboldmath$=4$ \quad , \quad $u=-v_1+v_2-v_3-2v_4+v_5+v_6+v_7+v_9+v_{10} \quad ,$\\
866: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 2\pi/7 $ \quad , \quad $u=( \sin 2\pi/7 -2 \sin 4\pi/7+2\sin 6\pi/7)v_1+
867: (\sin 6\pi/7-\sin 4\pi/7)v_2+(\sin 2\pi/7-\sin 4\pi/7+\sin 6\pi/7)(v_3+v_7)+ (\sin 2\pi/7 )(v_4+v_{12})+(2\sin 6\pi/7-\sin 2\pi/7)v_5+(2\sin 2\pi/7-\sin 4\pi/7-\sin 6\pi/7)(v_6+v_9)+(\sin 4\pi/7-3\sin 2\pi/7+3\sin 6\pi/7)v_8+(\sin 4\pi/7+\sin 6\pi/7)v_{10}+(\sin 2\pi/7-\sin 4\pi/7-\sin 6\pi/7)v_{11} \quad , $\\
868: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=v_3-v_4+v_6-v_7-v_9+v_{12} \quad ,$\\
869: \boldmath$\bigtriangleup_2$\unboldmath$=5$ \quad , \quad $u=-v_1-v_2+v_4+v_5+v_7+v_8+v_9-v_{10}+v_{11} \quad ,$\\
870: \boldmath$\bigtriangleup_2$\unboldmath$=8 \sin^2 3\pi/7 $ \quad , \quad $u=(-2 \sin 2\pi/7 +\sin 4\pi/7+2\sin 6\pi/7)v_1+
871: (\sin 6\pi/7-\sin 2\pi/7)v_2+(\sin 4\pi/7-\sin 2\pi/7+\sin 6\pi/7)(v_3+v_7)+ (\sin 4\pi/7 )(v_4+v_{12})-(2\sin 2\pi/7+\sin 4\pi/7)v_5+(\sin 2\pi/7+2\sin 4\pi/7+\sin 6\pi/7)(v_6+v_9)-(3\sin 2\pi/7+3\sin 4\pi/7+\sin 6\pi/7)v_8-(\sin 2\pi/7+\sin 6\pi/7)v_{10}+(\sin 2\pi/7+\sin 4\pi/7+\sin 6\pi/7)v_{11} \quad , $\\
872: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=v_3-v_4-2v_6-v_7+2v_9+v_{12} \quad ,$\\
873: \boldmath$\bigtriangleup_2$\unboldmath$=8$ \quad , \quad $u=v_1+v_2-3v_3+2v_4-v_5+3v_6-v_7+2v_8-v_9+v_{10}+2v_{11} \quad ,$\\
874: The remaining $3$ eigenvectors correspond to the eigenvalues
875: \boldmath$\bigtriangleup_2$\unboldmath$=7-\lambda$ where $\lambda$ are the roots of the cubic equation : $\lambda^3-\lambda^2-17\lambda+25=0$ .\\
876:
877:
878: \section*{Appendix B. The Eigenvalues}
879:
880: The text \cite{enc}, provides a table with all the irreducible representation for the generators of the permutation group $S_n$ up to $n=7$. It is easy to obtain all the irreducible representations for the operator
881: ${\bf A}$ in eq.(\ref{t.3}), up to the same order. We list here the representations of the operator {\bf A} corresponding to $S_n$, $3 \leq n \leq 6$, and the eigenvalues.\\
882:
883: To denote the irreducible representation, we write the sequence of integers corresponding to the number of boxes in the horizontal rows of the Young tableaux. For instance $(2^2,1^2)$ is the irreducible representation of $S_6$ where the Young tableau has two boxes in the first two horizontal rows, and one box in the third and fourth row.\\
884:
885: If $\bigtriangleup_0=4$ , the relevant group is $S_3$. Beside the two $1$-dimensional representations corresponding to the partitions $(3)$ and $(1^3)$ ,
886: we only need the $2$-dimensional representation corresponding to the partition $(2,1)$ which may be chosen
887: \begin{eqnarray}
888: (2,1) \qquad , \qquad {\bf A}=\left(
889: \begin{array}{ccc}
890: -1 & 1\\
891: 1 & -1 \end{array} \right) \quad , \quad \lambda=0 \quad , \quad \lambda=-2 \nonumber
892: \end{eqnarray}
893: If $\bigtriangleup_0=5$ , the group is $S_4$. Beside the two $1$-dimensional representations corresponding to the partitions $(4)$ and $(1^4)$ , there are : one $2$-dimensional representation
894: corresponding to the partition $(2^2)$
895: and two $3$-dimensional representations corresponding to the partitions $(3,1)$ and $(2,1^2)$. They respectively
896: may be chosen
897: \begin{eqnarray}
898: (2^2)\quad &,&{\bf A}=\left(
899: \begin{array}{ccc}
900: 0 & -1\\
901: -1 & 0 \end{array} \right) \quad , \quad \lambda=\pm 1 \quad , \quad \nonumber \\
902: (3,1) \quad &,&{\bf A}=\left(
903: \begin{array}{ccc}
904: -1 & 2 & -2\\
905: 0 & 2 & -1 \\
906: -2 & 1 & 0\end{array} \right) \quad , \quad \lambda=\pm \sqrt{5} \quad , \quad \lambda=1 \quad ,\nonumber \\
907: (2,1^2) \quad &,&{\bf A}=\left(
908: \begin{array}{ccc}
909: 0 & -1 & 2\\
910: 1 & -2 & 2 \\
911: 2 & 0 & 1\end{array} \right) \quad , \quad \lambda=\pm \sqrt{5} \quad , \quad \lambda=-1 \quad ,\nonumber
912: \end{eqnarray}
913:
914: If $\bigtriangleup_0=6$ the group is $S_5$. Beside the two $1$-dimensional representations corresponding to the partitions $(5)$ and $(1^5)$ , there are :
915: two $4$-dimensional representations , corresponding to the partitions $(4,1)$ and $(2,1^3)$,
916: two $5$-dimensional representations corresponding to the partitions $(3,2)$ and $(2^2,1)$,
917: and one $6$-dimensional representation corresponding to the partition $(3,1^2)$ . They respectively
918: may be chosen
919: $$(4,1) \quad , \quad
920: {\bf A}=\left(\begin{array}{cccc}
921: 0 & 2 & 0 &-2\\
922: 0 & 2 & 2 & -2 \\
923: -2 & 2 & 3 & -1\\
924: -2 & 0 & 1 & 1\end{array} \right) \quad , \quad \lambda=4 \quad , \quad \lambda=1\pm \sqrt{5} \quad , \quad \lambda=0 \quad , $$
925: $$(2,1^3) \quad , \quad
926: {\bf A}=\left(\begin{array}{cccc}
927: -3 & -1 & 0 &0\\
928: -1 & -3 & 0 & 0 \\
929: 0 & 0 & -2 & 0\\
930: 0 & 0 & 0 & -2\end{array} \right) \quad , \quad \lambda=-4 \quad , \quad \lambda=-2 \quad {\rm three} \quad {\rm times} \quad , \quad $$
931: $$(3,2) \quad , \quad
932: {\bf A}=\left(\begin{array}{ccccc}
933: 0 & 1 & 0 & 0 & -1\\
934: 0 & 1 & 0 & 0 & 0 \\
935: 0 & 0 & 2 & 0 & -2\\
936: -1 & 1 & -1 & 1 & 1\\
937: -1 & 1 & -1 & 1 & 0\end{array} \right) \quad , \quad \lambda=1 \quad {\rm twice} \quad , \quad \lambda=1\pm \sqrt{5} \quad , \quad \lambda=0 \quad ,$$
938: $$(2^2,1) \quad , \quad
939: {\bf A}=\left(\begin{array}{ccccc}
940: 3 & 0 & -4 & 0 & 2\\
941: 1 & 1 & -2 & 0 & 1 \\
942: 1 & -1 & -4 & 2 & 1\\
943: -1 & 1 & 2 & -1 & 2\\
944: -1 & -1 & 2 & 2 & -3\end{array} \right) \quad , \quad \lambda=1 \quad {\rm twice} \quad , \quad \lambda=-1\pm \sqrt{13} \quad , \quad \lambda=-4 \quad ,$$
945: $$(3,1^2) \quad , \quad
946: {\bf A}=\left(\begin{array}{cccccc}
947: 0 & 0 & 1 & 0 & 1 & 0\\
948: 2 & 0 & 1 & -2 & 2 & 1 \\
949: 2 & 1 & 1 & -2 & 0 & 2\\
950: 1 & 0 & 0 & -1 & 1 & 1\\
951: 1 & 2 & 0 & -1 & 0 & 2\\
952: 0 & 1 & 1 & 0 & 0 & 2\end{array} \right) \quad , \quad \lambda=1 \quad {\rm twice} \quad , \quad \lambda=-2
953: \quad {\rm twice} \quad, \quad \lambda=4 \quad , \quad \lambda=0 $$
954:
955: To make easier the comparison with eigenvalues already evaluated, see Table 3.2 in the reference \cite{r1}, we remark that the Laplacian eigenvalues $\bigtriangleup_2=5\pm \sqrt{5}$ and $\bigtriangleup_2=7 \pm \sqrt{13}$ are the roots of the equations $E^2-10 E +20=0$ and $E^2-14 E+36=0$ respectively.\\
956:
957: If $\bigtriangleup_0=7$ , the group is $S_6$. Beside the two $1$-dimensional representations corresponding to the partitions $(6)$ and $(1^6)$ , there are :
958: four $5$-dimensional representations corresponding to the partitions $(5,1)$ , $(3^2)$ , $(2^3)$, $(2,1^4)$,
959: two $9$-dimensional representations, corresponding to the partitions $(4,2)$, $(2^2,1^4)$,
960: two $10$-dimensional representations, corresponding to the partitions $(4,1^2)$, $(3,1^3)$,
961: and one $16$-dimensional representation corresponding to the partition $(3,2,1)$. They respectively may be chosen
962: $$(5,1) \quad ,\quad
963: {\bf A}=\left(\begin{array}{ccccc}
964: 1 & 2 & 0 & 0 & -2\\
965: 0 & 3 & 2 & 0 & -2 \\
966: -2 & 2 & 3 & 2 & -2\\
967: -2 & 0 & 2 & 4 & -1\\
968: -2 & 0 & 0 & 1 & 2\end{array} \right)\quad , \quad \lambda=1 \quad , \quad \lambda=5, \quad \lambda^3-7\lambda^2+7\lambda+7=0 \quad , $$
969: $$(3^2) \quad , \quad
970: {\bf A}=\left(\begin{array}{ccccc}
971: 2 & -1 & 1 & 1 & -3\\
972: 0 & 2 & 1 & 0 & -2 \\
973: 0 & 0 & 3 & 0 & -2\\
974: 0 & -1 & 0 & 3 & -1\\
975: 1 & -3 & 2 & 3 & -5\end{array} \right)\quad , \quad \lambda=2 \quad {\rm twice}\quad , \quad
976: \lambda^3-\lambda^2-17 \lambda+25=0 \quad ,$$
977: $$(2^3) \quad , \quad
978: {\bf A}=\left(\begin{array}{ccccc}
979: 4 & -1 & -2 & 2 & 3\\
980: 3 & -3 & 0 & 0 & 2 \\
981: 2 & 0 & -3 & 1 & 1\\
982: 3 & -1 & 0 & -2 & 2\\
983: 1 & 0 & 0 & 0 & -1\end{array} \right)\quad , \quad \lambda=-2 \quad {\rm twice}\quad , \quad
984: \lambda^3+\lambda^2-17 \lambda-25=0 \quad , $$
985: $$(2,1^4) \quad , \quad
986: {\bf A}=\left(\begin{array}{ccccc}
987: -2 & -1 & 2 & -2 & 2\\
988: 1 & -4 & 2 & 0 & 0 \\
989: 0 & 2 & -3 & 2 & 0\\
990: 0 & 0 & 2 & -3 & 2\\
991: 2 & -2 & 2 & 0 & -1\end{array} \right)\quad , \quad \lambda=-1 \quad , \quad \lambda=-5, \quad \lambda^3+7\lambda^2+7\lambda-7=0 \quad , $$
992: $$(4,2) \quad , \quad
993: {\bf A}=\left(\begin{array}{ccccccccc}
994: 1 & 0 & 1 & 1 & 0 & -1 & 0 & -1 & 0\\
995: 0 & 0 & 1 & 1 & 1 & -1 & 1 & -1 & -1 \\
996: 0 & -1 & 3 & 0 & 1 & 0 & 1 & -2 & -1\\
997: -1 & 1 & 0 & 2 & 1 & -1 & 0 & 1 & -1\\
998: -1 & 0 & 1 & 1 & 2 & 0 & 0 & -1 & 0\\
999: 0 & -1 & 1 & -1 & 1 & 3 & 1 & -1 & -2\\
1000: -1 & 0 & 0 & 1 & 0 & -1 & 2 & 1 & 0\\
1001: -1 & -1 & 0 & 2 & 0 & -1 & 1 & 1 & 1\\
1002: 0 & -1 & 0 & 0 & 1 & -1 & 1 & 0 & 1
1003: \end{array} \right)\quad , \quad \lambda=-1 \quad {\rm twice}\quad , \quad \lambda=1 \quad {\rm twice}\quad ,$$
1004: $$\quad \lambda=2 \quad {\rm twice}\quad , \quad
1005: \lambda=3 \quad {\rm twice}\quad , \quad
1006: \lambda=5, $$
1007: $$(2^2,1^2) \quad , \quad
1008: {\bf A}=\left(\begin{array}{ccccccccc}
1009: 0 & 0 & 0 & -2 & -1 & 0 & 1 & 0 & 1\\
1010: 0 & 0 & 0 & 1 & -2 & 0 & 2 & -1 & 0 \\
1011: -1 & 1 & -2 & 1 & 0 & 0 & 0 & 0 & -1\\
1012: -1 & 1 & -1 & -3 & 0 & 1 & -1 & 0 & 0\\
1013: -1 & 0 & 0 & 1 & -2 & 1 & 0 & 0 & 0\\
1014: 0 & -2 & 1 & 1 & 1 & -2 & 0 & 1 & 1\\
1015: 0 & 0 & 0 & -1 & 1 & 0 & -3 & 1 & -1\\
1016: 1 & -1 & 0 & -1 & 0 & 1 & 0 & -1 & 1\\
1017: 0 & 1 & -1 & 0 & -1 & 1 & 0 & 0 & -2
1018: \end{array} \right)\quad , \quad \lambda=-1 \quad {\rm twice}\quad , \quad $$
1019: $$\qquad \lambda=1 \quad {\rm twice}\quad ,\quad \lambda=-2 \quad {\rm twice}\quad , \quad
1020: \lambda=-3 \quad {\rm twice}\quad , \quad
1021: \lambda=-5, \qquad $$
1022: $$(4,1^2) \quad , \quad
1023: {\bf A}=\left(\begin{array}{ccccccccccc}
1024: 1 & 0 & 1 & 0 & 0 & 0 & 0 & 1 & 0 & 0\\
1025: 2 & 0 & 1 & 0 & 1 & 0 & -1 & 1 & 1 & 0 \\
1026: 2 & 1 & 1 & -1 & 1 & 1 & -1 & 0 & 1 & 0\\
1027: 1 & 1 & 0 & 1 & 1 & 0 & -2 & 1 & 1 & 1\\
1028: 1 & 2 & 1 & 0 & 2 & 1 & -2 & -1 & 2 & 1\\
1029: 0 & 1 & 1 & 0 & 1 & 2 & -1 & -1 & 0 & 2\\
1030: 1 & 0 & 0 & 0 & 0 & 0 & 0 & 1 & 0 & 1\\
1031: 1 & 1 & 0 & 1 & 0 & 0 & -1 & 1 & 1 & 1\\
1032: 0 & 1 & 0 & 1 & 1 & 0 & -1 & 0 & 1 & 2\\
1033: 0 & 0 & 0 & 1 & 0 & 1 & 0 & 0 & 0 & 3
1034: \end{array} \right)\quad , \quad \lambda=-1 \quad {\rm twice}\quad , \quad \lambda=0 \quad {\rm twice}\quad ,$$
1035: $$\quad \lambda=1 \quad , \quad
1036: \lambda=3 \quad {\rm twice}\quad , \quad
1037: \lambda^3-7\lambda^2+7\lambda+7=0, $$
1038: $$(3,1^3) \quad , \quad
1039: {\bf A}=\left(\begin{array}{ccccccccccc}
1040: -3 & 2 & -1 & 1 & -2 & 1 & -1 & 0 & 0 & 0\\
1041: 0 & -1 & 1 & 0 & 0 & -2 & 1 & 1 & -1 & 0 \\
1042: 0 & 0 & -1 & 1 & 1 & -1 & -1 & 0 & 1 & -1\\
1043: 0 & 1 & -1 & 0 & -1 & 2 & -2 & -1 & 1 & 0\\
1044: -1 & 0 & 0 & 0 & -2 & 1 & 0 & -1 & 0 & 0\\
1045: 0 & -1 & 0 & 0 & 1 & -2 & 1 & 1 & -1 & 0\\
1046: -1 & 1 & -1 & 0 & 0 & 0 & -1 & 1 & 0 & 0\\
1047: 0 & 0 & 0 & 0 & -1 & 1 & 0 & -1 & 1 & -1\\
1048: 0 & -1 & 1 & 0 & 1 & -2 & 1 & 1 & 0 & 0\\
1049: 0 & 0 & -1 & 1 & 0 & 1 & -1 & -2 & 2 & -1
1050: \end{array} \right)\quad , \quad \lambda=-3 \quad {\rm twice}\quad , \quad $$
1051: $$ \lambda=0 \quad {\rm twice}\quad ,\quad \lambda=-1 \quad , \quad
1052: \lambda=1 \quad {\rm twice}\quad , \quad
1053: \lambda^3+7\lambda^2+7\lambda-7=0, $$
1054: $$(3,2,1) \quad , \quad
1055: {\bf A}=\left(\begin{array}{cccccccccccccccc}
1056: 2& -1& 1& -1& 0& -1& 2& 0& -2& 2& -2& 0& -1& 1& -1& 1\\
1057: 2& -1& -2& 2& 0& -2& 1& 1& -1& 1& 0& 1& -3& 1& -2& 1\\
1058: 2& -1& 0& 1& 2& 0& 0& 1& -2& 0& -3& 1& 1& -1& 0& 3\\
1059: 1& 1& 0& 1& 0& 1& 0& 1& -1& 0& -2& 1& -1& 0& -1& 3\\
1060: 0& 0& 2& -1& 2& -1& 1& 1& -1& 0& -3& 0& 0& 0& 0& 2\\
1061: 0& 0& 1& 1& 0& 0& -1& 2& -1& 0& -1& 0& 0& 0& 0& 2\\
1062: 2& -1& -1& 0& 1& -1& 3& 1& -1& 1& -2& 1& -2& 0& -2& 3\\
1063: 1& 1& -1& 0& 1& 1& 0& 2& 0& -1& -2& 2& 0& -1& -1& 2\\
1064: 1& -1& 0& 0& 0& 0& 0& 0& -2& 2& 0& 0& 0& 0& 1& 0\\
1065: 1& 0& -1& 1& -1& 0& 0& 0& 1& 0& 0& 2& -1& 0& 0& 0\\
1066: 0& 0& 1& -1& 0& 0& 0& 0& 0& 1& -3& 0& 1& 0& 0& 1\\
1067: 0& 0& 1& 0& -1& 1& -1& 0& 0& 1& 0& -1& 1& 1& 2& 0\\
1068: 1& -1& 0& 1& 0& -1& 0& 0& -1& 0& 0& 1& -1& 1& 0& 1\\
1069: 1& 0& 0& 1& 0& 1& -1& 0& -1& -1& 0& 1& 2& 0& 2& 1\\
1070: 0& 0& 0& 0& -1& 0& 0& -1& 1& 0& 1& 1& 0& 1& 0& -1\\
1071: 0& 0& 1& 0& -1& 1& 0& -1& -1& 0& 2& -1& 0& 2& 2& -2
1072: \end{array} \right)\quad , \quad
1073: $$
1074: $\lambda=3$ twice, $\lambda=-3$ twice, $\lambda=1$ three times,
1075: $\lambda=-1$ three times, $\lambda^3-\lambda^2-17\lambda+25=0$,
1076: $\lambda^3+\lambda^2-17\lambda-25=0$.\\
1077:
1078: To make easier the comparison with eigenvalues already evaluated, \cite{be} , \cite{mz}, \cite{be2},
1079: we remark that the $3$ roots of $\lambda^3-7\lambda^2+7\lambda+7=0$ are $\lambda=3-4 \cos[(2p+1)\pi/7]$ with $p=0,1,2$. This is easily proved by extracting the real part of the identity $\sum_{k=0}^6 e^{2\pi i k/7}=0$ then obtaining a cubic equation for $\cos(\pi/7)$.
1080: The $3$ roots of $\lambda^3+7\lambda^2+7\lambda-7=0$ are the opposite of the previous roots.\\
1081: The $3$ roots of the cubic equation $\lambda^3-\lambda^2-17 \lambda+25=0$ are mapped into $3$ eigenvalues $\bigtriangleup_2$ of the Laplacian by $E=7-\lambda$ and are the roots of the equation $E^3-20 E^2+116 E-200=0$, which is quoted in Table 3.4 of the reference
1082: \cite{r1}. In the same way, the roots of the equation $\lambda^3+\lambda^2-17 \lambda-25=0$ , which are opposite of the previous ones, become roots of the equation $E^3-22 E^2+144 E-248=0$.\\
1083:
1084: \section*{Appendix C. The Eigenvectors}
1085: We are interested in the eigenstates corresponding to some irreducible representation of $S_n$.
1086: We generalize the method used in Sect.3 for the representation $(3,1)$ of $S_4$, of degree $f=3$. Here we find basis vectors and eigenvectors for the two irreducible representations of $S_n$, the representation $(n-1,1)$ and its conjugate $(2,1^{n-2})$, both of degree $f=n-1$.\\
1087: Only the first one provides important sequences of eigenvectors of the dilatation operator of superconformal Yang Mills theory, after proper replacements.\\
1088:
1089:
1090: {\bf The representation $(n-1,1)$ of the group $S_n$.}\\
1091:
1092:
1093: For any representation of the permutation group $S_n$ of degree $f$, one obtains a matrix of order $f$ which represents the operator ${\bf A}$ in eq.(\ref{t.3}). As the degree of the representation increases, the evaluations becomes more cumbersome. However there are two irreducible representations which may be easily derived for arbitrary $S_n$ : they are of degree $n-1$ and are pair-conjugate.\\
1094: \epsfig{file=young2 ,height=3cm}\\
1095: The Young tableau has $n-1$ boxes in the first row and one in the second row. \\
1096:
1097: The set of elements in $S_n$ is partitioned in $n$ sets $a_j$, $j=1,2,..,n$. Each $a_j$ contains $(n-1)!$ elements and they may be written in the form $(j,n) \cdot g$ with $g \in S_{n-1}$.\\
1098: The Young projector operator $Y$ associated to the irreducible representation of $S_n$ corresponding to the partition $(n-1,1)$ is $Y=PQ$ , where $P$ is the sum of the $(n-1)!$ permutations of $S_{n-1}$ and $Q=e-(1,n)$.\\
1099: We choose $n$ basis vectors $v_j$ :
1100: $$ v_j=(j,n) \cdot PQ \quad , \quad {\rm for} \, j=1,2..,n-1 \quad , \quad v_n=e \cdot PQ $$
1101: Since $\sum_1^n v_j=0$ , one basis vector may be eliminated, but equations will be more neat by keeping all $n$ vectors.\\
1102: Any permutation in $S_n$ which multiplies from left a basis vector $v_j$ obtains a basis vector. For example :\\
1103: $(j,j+1)v_j=v_{j+1}$ , $(j,j+1)v_{j+1}=v_j$ , $(1,2,3,..,n)v_j=v_{j+1}$ , $(n , n-1,..,2,1)v_j=v_{j-1}$ \\
1104:
1105: The action of the operator ${\bf A}=(1,2)+(2,3)+(3,4)+..+(n-1,n)+(1,2,...,n)+(n,n-1,..,2,1)$ on each basis vector is given by the system
1106:
1107: \begin{eqnarray}
1108: A v_1&=& (n-2)v_1+2v_2+v_n \quad , \nonumber\\
1109: Av_2 &=& 2v_1+(n-3)v_2+2v_3 \quad , \nonumber\\
1110: Av_3 &=& 2v_2+(n-3)v_3+2v_4 \quad , \nonumber\\
1111: ..&=&... ... \quad , \nonumber\\
1112: Av_{n-1} &=&2v_{n-2}+(n-3)v_{n-1}+2v_n \quad , \nonumber\\
1113: A v_n &=& v_1+2v_{n-1}+(n-2)v_n \quad
1114: \nonumber
1115: \end{eqnarray}
1116: If $u=\sum_1^n \alpha_j v_j$ is eigenvector of $A$ , $A\,u=\lambda\,u$, the set of coefficients $\alpha_j$ is determined by the system
1117: \begin{eqnarray}
1118: (n-2)\alpha_1+2\alpha_2+\alpha_n &=& \lambda \alpha_1\quad , \nonumber\\
1119: 2\alpha_1+(n-3)\alpha_2+2\alpha_3 &=& \lambda \alpha_2\quad , \nonumber\\
1120: 2\alpha_2+(n-3)\alpha_3+2\alpha_4 &=& \lambda \alpha_3 \quad , \nonumber\\
1121: ..&=&... ... \quad , \nonumber\\
1122: 2\alpha_{n-2}+(n-3)\alpha_{n-1}+2\alpha_n &=&\lambda \alpha_{n-1}\quad , \nonumber\\
1123: \alpha_1+2\alpha_{n-1}+(n-2)\alpha_n &=& \lambda \alpha_n \quad
1124: \nonumber
1125: \end{eqnarray}
1126:
1127: The $n-1$
1128: eigenvectors are found by solving linear recurrence relations. They form two sequences :
1129: \begin{eqnarray}
1130: \lambda_k=n-3+4 \cos 2k\pi/(n+1) &,& u^{(k)}=\sum_{j=1}^n \sin \left( \frac{2k\pi j}{n+1}\right) v_j \quad , \quad \nonumber\\
1131: k=1,2,..,k_{max} &,& k_{max}<(n+1)/2 \quad , \quad {\rm and} \nonumber \\
1132: \lambda_k=n-3+4 \cos 2k\pi/n &,&
1133: u^{(k)}=\sum_{j=1}^n \cos \left( \frac{k\pi(2j-1)}{n}\right) v_j \quad , \quad \nonumber\\
1134: (C.1) \qquad k=1,2,..,k_{max} &,& k_{max}<n/2 .
1135: \nonumber
1136: \end{eqnarray}
1137:
1138:
1139: Since the group $S_n$ is associated with $\bigtriangleup_0=n+1$, the two sets of eigenvalues translate into two sets for the one loop contribution to the dilatation dimension :\\
1140: $\bigtriangleup_2=8 \sin^2 \frac{\pi k}{n+1}$ , $k=1,2,..,k_{max}$ , $k_{max}<(n+1)/2$ , and \\
1141: $\bigtriangleup_2=8 \sin^2 \frac{\pi k}{n}$ , $k=1,2,..,k_{max}$ , $k_{max}<n/2$ . \\
1142:
1143: We rewrite the basis vectors and the eigenvectors in term of traces of products of matrix fields by a straightforward generalization of the procedure leading from the basis vectors (\ref{t.8}) to their form (\ref{t.7})
1144:
1145:
1146: \begin{eqnarray}
1147: \qquad v_1&=& \sum_{p \in S_{n-1}} \Tr {\phi_{\underline{1}}\phi_{\underline{n+1}}\phi_2\phi_3\cdots \phi_n-\phi_{\underline{1}}\phi_{\underline{2}}\phi_3\phi_4\cdots \phi_{n+1}} \quad , \nonumber\\
1148: v_2 &=& \sum_{p \in S_{n-1}} \Tr { \phi_{ \underline{1}}\phi_2 \phi_{\underline{n+1}}\phi_3\cdots\phi_n-
1149: \phi_{ \underline{1}}\phi_3\phi_{\underline{2}}\phi_4\phi_5\cdots \phi_{n+1} }
1150: \nonumber\\
1151: v_3 &=& \sum_{p \in S_{n-1}} \Tr { \phi_{ \underline{1}}\phi_2\phi_3\phi_{\underline{n+1}}\phi_4\cdots \phi_n-
1152: \phi_{ \underline{1}}\phi_3\phi_4\phi_{\underline{2}}\phi_5\cdots \phi_{n+1}}
1153: \nonumber\\
1154: \cdots &=& \cdots \quad \cdots \nonumber\\
1155: (C.2) \qquad v_n&=&\sum_{p \in S_{n-1}} \Tr {\phi_{ \underline{1}}\phi_2\phi_3\cdots \phi_n\phi_{\underline{n+1}}-
1156: \phi_{ \underline{1}}\phi_3\phi_4\phi_5 \cdots \phi_{n+1}\phi_{\underline{2}} }
1157: \nonumber
1158: \end{eqnarray}
1159: where we
1160: use the compact notation of footnote 2 to indicate the elements remaining fixed in the sum over permutations.
1161: It is useful to notice that , for every $j$ , the states $v_j$ and $v_{n-j+1}$ are related by parity.\\
1162:
1163:
1164: {\bf The representation $(2,1^{n-2})$ of $S_n$.}\\
1165:
1166: Although it will turn out to be of more limited use, we derive a set of basis vectors and the eigenvectors which correspond to the irreducible representation conjugate to the previous one. Actually only slight changes occur in the derivation.\\
1167:
1168: \epsfig{file=young4,height=6cm}
1169: We define $Q$ the sum of all permutations of the group $S_{n-1}$ with their sign, that is $(+1)$ or $(-1)$ depending on the even or odd numbers of transposition that relate the permutation to the identity permutation.
1170: $$Q=\sum_{p \in S_{n-1}} ({\rm sign}\,p)\,p $$
1171: One easily finds that $-Q+\sum_{j=1}^{n-1} (j,n)Q$ is the opposite of the sum of all the $n!$ permutations of the group $S_n$, with their sign.\\
1172: We may choose a set of $n$ basis vectors
1173: $$z_j=(j,n)Q\left(e+(1,n)\right) \quad ,\quad {\rm for}\quad j=1,2,\cdots ,n-1 \quad , \quad z_n=-Q\left(e+(1,n)\right) $$
1174: Then $\sum_1^n z_n=0$. and one easily obtains the action of any transposition on the basis vectors :
1175: \begin{eqnarray}
1176: &&(j,j+1)(j,n)Q=(j+1,n)(j,j+1)Q=-(j+1,n)Q \quad , \quad {\rm then}\quad \nonumber\\
1177: &&(j,j+1)z_j=-z_{j+1}\quad , \quad
1178: (j,j+1)z_{j+1}=-z_j \quad
1179: \nonumber
1180: \end{eqnarray}
1181: However $(1,2,\cdots ,n)z_j=-(-1)^n z_{j+1}$ and $(n,\cdots ,2,1)z_j=-(-1)^n z_{j-1}$.\\
1182:
1183: If $n$ is even we have
1184: \begin{eqnarray}
1185: A z_1&=& -(n-2)z_1-2z_2-z_n \quad , \nonumber\\
1186: Az_2 &=& -2z_1-(n-3)z_2-2z_3 \quad , \nonumber\\
1187: A z_3 &=& -2z_2-(n-3)z_3-2z_4 \quad , \nonumber\\
1188: ..&=&... ... \quad , \nonumber\\
1189: Az_{n-1} &=&-2z_{n-2}-(n-3)z_{n-1}-2z_n \quad , \nonumber\\
1190: A z_n &=& -z_1-2z_{n-1}-(n-2)z_n \quad
1191: \nonumber
1192: \end{eqnarray}
1193: Eigenvalues of the operator $A$ are the opposite of the ones of the conjugate representation and eigenvectors are the same, where $z_j$ now replaces $v_j$, that is two sequences : \\
1194: $\lambda_k=3-n-4 \cos 2k\pi/(n+1)$ , $u^{(k)}=\sum_{j=1}^n \sin \left( \frac{2k\pi j}{n+1}\right) z_j $ ,
1195: $k=1,2,..,k_{max}$ , $k_{max}<(n+1)/2$ , and \\
1196: $\lambda_k=3-n-4 \cos 2k\pi/n$ ,
1197: $u^{(k)}=\sum_{j=1}^n \cos \left( \frac{k\pi(2j-1)}{n}\right) z_j $ ,
1198: $k=1,2,..,k_{max}$ , $k_{max}<n/2$ . \\
1199:
1200: If $n$ is odd integer, we obtain
1201: \begin{eqnarray}
1202: A z_1&=& -(n-2)z_1+z_n \quad , \nonumber\\
1203: ..&=&... ... \quad , \nonumber\\
1204: Az_j &=& -(n-3)z_j \quad , \nonumber\\
1205: ..&=&... ... \quad , \nonumber\\
1206: A z_n &=& z_1-(n-2)z_n \quad , \quad {\rm that}\quad {\rm is}
1207: \nonumber\\
1208: A&=&(3-n)I+
1209: \left( \begin{array}{cccccccc}
1210: -1 & 0 & 0 & 0 & 0 & ... & 1\\
1211: 0 & 0 & 0 & 0 & 0 & ... & 0\\
1212: 0 & 0 & 0 & 0 & 0 &... & 0\\
1213: 0 & 0 & 0 & 0 & 0 &... & 0\\
1214: ..& ..& .. & .. & ..& .. &..\\
1215: 0 & 0 & 0 & 0 & ... & 0 & 0\\
1216: 1 & 0 & 0 & 0 & ... & 0 & -1
1217: \end{array} \right)
1218: \nonumber
1219: \end{eqnarray}
1220: The $n$ eigenvalues of $A$ are :\\
1221: $\lambda=1-n$ , $\lambda=3-n$ , $n-1$ times.\\
1222: One of the degenerate eigenvalues $\lambda=3-n$ is associated to the null vector $\sum_1^n z_j$ then the $n-1$ eigenvalues of the irreducible representation, if $n$ is odd integer, are:\\
1223: $\lambda=1-n$ , $\lambda=3-n$ , $n-2$ times.\\
1224:
1225: We proceed to rewrite the basis vectors $z_j$ and the eigenvectors in term of traces of products of matrix fields as we did for the basis vectors $v_j$.\\
1226:
1227: $Q$ is the sum of the $(n-1)!$ permutations over the elements $\{1,2,\cdots,n-1\}$. We manifest it with the notation $Q_{S_{n-1}}(1,2,\cdots,n-1)$. Furthermore $(2,n)Q_{S_{n-1}}(1,2,3,\cdots,n-1)=Q_{S_{n-1}}(1,n,3,\cdots,n-1)(2,n)$ and $(2,n)(1,n)=(1,n)(1,2)$. Then
1228: $$z_2=Q_{S_{n-1}}(1,n,3,\cdots,n-1)(2,n)-Q_{S_{n-1}}(1,n,3,\cdots,n-1)(1,2)$$
1229: According to the relations in Footnote 1 we add one unit in the symbols, evaluate the inverse of the elements and rewrite in terms of matrix fields. Some care is needed for the sign of the permutation.
1230: We use the compact notation of footnote 2 to indicate the elements remaining fixed in the sum over permutations. Except for a overall factor $(-1)^n$ we find \\
1231: \begin{eqnarray}
1232: z_1&=& (-1)^n \sum_{p \in S_{n-1}} ({\rm sign}\,p)\, \Tr {\phi_{\underline{1}}\phi_{\underline{n+1}}\phi_2\phi_3\cdots \phi_n}+\sum_{p \in S_{n-1}} ({\rm sign}\,p)\, \Tr {
1233: \phi_{\underline{1}}\phi_{\underline{2}}\phi_3\phi_4\cdots \phi_{n+1}}\quad , \nonumber\\
1234: z_2 &=& (-1)^{n+1} \sum_{p \in S_{n-1}}({\rm sign}\,p)\, \Tr { \phi_{ \underline{1}}\phi_2 \phi_{\underline{n+1}}\phi_3\cdots\phi_n}-
1235: \sum_{p \in S_{n-1}}({\rm sign}\,p)\, \Tr {
1236: \phi_{ \underline{1}}\phi_3\phi_{\underline{2}}\phi_4\phi_5\cdots \phi_{n+1} }
1237: \nonumber\\
1238: z_3 &=& (-1)^n \sum_{p \in S_{n-1}}({\rm sign}\,p)\,\Tr { \phi_{ \underline{1}}\phi_2\phi_3\phi_{\underline{n+1}}\phi_4\cdots \phi_n}+
1239: \sum_{p \in S_{n-1}}({\rm sign}\,p)\,\Tr {
1240: \phi_{ \underline{1}}\phi_3\phi_4\phi_{\underline{2}}\phi_5\cdots \phi_{n+1}}
1241: \nonumber\\
1242: \cdots &=& \cdots \quad \cdots \nonumber\\
1243: z_n&=&- \sum_{p \in S_{n-1}}({\rm sign}\,p)\,\Tr { \phi_{ \underline{1}}\phi_2\phi_3\cdots \phi_n\phi_{\underline{n+1}}}+
1244: (-1)^{n+1}
1245: \sum_{p \in S_{n-1}}({\rm sign}\,p)\,\Tr {
1246: \phi_{ \underline{1}}\phi_3\phi_4\phi_5 \cdots \phi_{n+1}\phi_{\underline{2}}}
1247: \quad \nonumber
1248: \end{eqnarray}
1249:
1250: This irreducible representation provides eigenvalues and eigenvectors for the Super Yang Mill theory only for small values of the main quantum number $\bigtriangleup_0$. For example, for the permutation group $S_4$, with the replacements $(\phi_1, \phi_2)\to Z$ , $(\phi_3, \phi_5)\to W$ , $\phi_4 \to Y$ , we find
1251: \begin{eqnarray}
1252: z_1&=& \Tr { (2ZWZW-ZW^2Z-WZ^2W+Z^2W^2-WZWZ)Y} \quad , \nonumber\\
1253: z_2 &=& \Tr { (W^2Z^2-2Z^2W^2+WZ^2W+ZW^2Z-WZWZ)Y} \quad , \nonumber\\
1254: z_3 &=& \Tr { (Z^2W^2-2W^2Z^2+WZ^2W+ZW^2Z-ZWZW)Y} \quad , \nonumber\\
1255: z_4&=& \Tr { (2WZWZ-ZW^2Z-WZ^2W+W^2Z^2-ZWZW)Y} \quad
1256: \quad \nonumber
1257: \end{eqnarray}
1258: One obtains one positive parity eigenstate and two negative parity eigenstates :
1259: \begin{eqnarray}
1260: &&\bigtriangleup_0=5 \quad , \quad \bigtriangleup_2=6 \quad , \quad {\rm parity} =+1 \nonumber\\
1261: && {}\qquad u=\Tr { (ZWZW+WZWZ)Y-2(ZW^2Z+WZ^2W)Y+(W^2Z^2+Z^2W^2)Y} \quad . \nonumber\\
1262: &&\bigtriangleup_0=5 \quad , \quad \bigtriangleup_2=8 \sin^2 \frac{2\pi}{5} \quad , \quad {\rm parity} =-1 \nonumber\\
1263: && {}\qquad u=(z_1-z_4)\sin \frac{2\pi}{5}+(z_2-z_3)\sin \frac{4\pi}{5} \quad . \nonumber\\
1264: &&\bigtriangleup_0=5 \quad , \quad \bigtriangleup_2=8 \sin^2 \frac{\pi}{5} \quad , \quad {\rm parity} =-1 \nonumber\\
1265: && {}\qquad u=(z_1-z_4)\sin \frac{4\pi}{5}+(z_3-z_2)\sin \frac{2\pi}{5} \quad . \nonumber
1266: \end{eqnarray}
1267:
1268: However we find that replacing the set matrix fields $\phi_j$ with an alphabet of merely $3$ letters $Z,W,Y$ leads to vanishing basis vectors $z_j$ if the main quantum number $\bigtriangleup_0>5.$\\
1269:
1270: \section*{Appendix D. The replacements}
1271:
1272: Our final step is to replace the fields $\phi_j$ in the general solution ($C.1$) and ($C.2$) pertinent to the representation $(n-1,1)$ with the complex fields $Z$ , $W$ and $Y$ then obtaining the eigenvectors of the dilatation operators.\\
1273: For the group $S_n$ such eigenvalues are sums of a large number of terms, each of them being a product of $n+1$ fields $\phi_j$. By replacing $n_1$ fields with the complex field $Z$ , $n_2$ fields with $W$ , $n_3$ fields with $Y$ , with $n_1+n_2+n_3=n+1$, one obtains sets of eigenvalues for the sector $\Tr {Z^{n_1}W^{n_2}Y^{n_3}}$.\\
1274:
1275: The different choices of fields in the sets $n_j$ may lead to inequivalent sets of eigenvectors for the same sector $\Tr {Z^{n_1}W^{n_2}Y^{n_3}}$ as we indicated in Sect.4.\\
1276: We turn here to the generic term in the sequences with the simplest examples : \\
1277: (a) we obtain the well known sequences of eigenvectors with two impurities, \cite{be}, we may call it the sector $\Tr {Z^{n-1} \phi_a \phi_b}$\\
1278: (b) we obtain sequences of eigenvectors for the sectors $\Tr {Z^{n-2} W^2 Y}$.\\
1279:
1280:
1281: (a). {\bf The sector $\Tr{ Z^{n-1} \phi_a \phi_b}$.}\\
1282:
1283: By the replacement of $n-1$ matrix fields $\phi_j$ with the single complex matrix field $Z$, one recovers the well known sequences of eigenvectors with two impurities. More specifically, we identify $\{\phi_2 , \phi_3, \cdots \phi_n \}\to Z$, then
1284: \begin{eqnarray}
1285: v_1& \to & (n-1)! \, \Tr {\phi_{\underline{1}}\phi_{\underline{n+1}}Z^{n-1}}
1286: -(n-2)! \, \Tr {
1287: \phi_{\underline{1}}Z\phi_{\underline{2}}Z^{n-2}+ \phi_{\underline{1}}Z^2\phi_{\underline{2}}Z^{n-3}+
1288: \cdots + \phi_{\underline{1}}Z^{n-1}\phi_{\underline{2}}}= \nonumber\\
1289: &&=
1290: (n-2)!\, n\, \Tr {\phi_1 \phi_{n+1}Z^{n-1}} -(n-2)!\,C \quad , \quad \nonumber\\
1291: v_2& \to & (n-2)!\, n\, \Tr {\phi_1 Z\phi_{n+1}Z^{n-2}} -(n-2)!\,C \quad , \nonumber\\
1292: \cdots &\to & \cdots \qquad \cdots \nonumber\\
1293: v_n & \to & (n-2)!\, n\, \Tr {\phi_1 Z^{n-1}\phi_{n+1}} -(n-2)!\,C
1294: \quad \nonumber
1295: \end{eqnarray}
1296: where $
1297: C=\sum_{j=0}^{n-1} \Tr {\phi_1 Z^j \phi_{n+1}Z^{n-1-j}}$. \\
1298: Since $\sum_{j=1}^n \sin \frac{2k\pi j}{n+1}=0$ and $\sum_{j=1}^n \cos \frac{k\pi (2j-1)}{n}=0$ , we recover the well known symmetric and antisymmetric sequences with two impurities
1299: \begin{eqnarray}
1300: &&\bigtriangleup_2=8 \sin^2 \frac{k \pi}{n+1} \quad , \quad
1301: u^{(k)}=\sum_{j=1}^n \sin \frac{2k\pi j}{n+1}\,\Tr {\phi_1Z^{j-1}\phi_{n+1}Z^{n-j}} \quad \nonumber\\
1302: (D.1)&&k=1,2,\cdots ,k_{max} \quad , \quad k_{max}<\frac{n+1}{2}
1303: \quad ,\quad {\rm parity}=-1 \quad
1304: \nonumber
1305: \end{eqnarray}
1306: and
1307: \begin{eqnarray}
1308: &&\bigtriangleup_2=8 \sin^2 \frac{k \pi}{n} \quad , \quad
1309: u^{(k)}=\sum_{j=1}^n \cos \frac{k\pi (2j-1)}{n}\,\Tr {\phi_1Z^{j-1}\phi_{n+1}Z^{n-j}} \quad \nonumber\\
1310: (D.2)&&k=1,2,\cdots ,k_{max} \quad , \quad k_{max}<\frac{n}{2}
1311: \quad ,\quad {\rm parity}=1 \quad
1312: \nonumber
1313: \end{eqnarray}
1314:
1315: (b). {\bf The sector $\Tr{ Z^{n-2} W^2 Y}$.}\\
1316:
1317: Let us now look for a new sequence of eigenstates by replacing $n-2$ matrix fields $\phi_j$ with the single complex matrix field $Z$, $\{\phi_3 , \phi_4, \cdots \phi_n \}\to Z$,
1318: two fields with $W$, $\{\phi_1 \, ,\, \phi_2\} \to W$ , and the last one with $Y$, $\phi_{n+1}\to Y$. \\
1319: With this replacement all the basis vector would vanish if the further replacement $Y \to W$ were made. Indeed an alternative and inequivalent replacement leading to the same sector $\Tr{ Z^{n-1} W^2 Y}$ is $\phi_1 \to Y$ , $\{\phi_2\,,\,\phi_3\} \to W$ , $\{ \phi_4,\cdots , \phi_{n+1}\} \to Z$. We limit ourselves to the first, simpler, replacement.\\
1320:
1321: We find
1322: \begin{eqnarray}
1323: v_1 &=& \sum_{r=0}^{n-2} \Tr {Z^r WZ^{n-2-r}WY}-\sum_{r=0}^{n-2} \Tr { Z^{n-2-r} W^2Z^r Y } \quad , \nonumber\\
1324: v_2 &=& \sum_{r=0}^0 \Tr { Z^{n-2}WZ^r WZ^{-r}Y-Z^r WZ^{n-2}WZ^{-r}Y}+ \qquad \nonumber\\
1325: && \qquad + \sum_{r=0}^{n-3} \Tr {Z^r WZ^{n-3-r}WZ Y-Z^r W Z W Z^{n-3-r}Y} \quad , \nonumber\\
1326: v_j &=& \sum_{r=0}^{j-2} \Tr {Z^{n-j}WZ^r WZ^{j-2-r}Y-Z^r WZ^{n-j}WZ^{j-2-r}Y} + \qquad \nonumber\\
1327: && \qquad +\sum_{r=0}^{n-j-1} \Tr { Z^r WZ^{n-j-1-r}WZ^{j-1}Y-Z^r WZ^{j-1}WZ^{n-j-1-r}Y} \quad , \nonumber\\
1328: v_{n-1} &=& \sum_{r=0}^{n-3} \Tr {Z W Z^r WZ^{n-3-r} Y-
1329: Z^r W Z W Z^{n-3-r} Y} + \qquad \nonumber\\
1330: && \qquad +\sum_{r=0}^{0} \Tr { Z^r W Z^{-r}WZ^{n-2}Y-
1331: Z^r W Z^{n-2} W Z^{-r}Y} \quad , \nonumber\\
1332: v_n &=& \sum_{r=0}^{n-2} \Tr { W Z^r WZ^{n-2-r} Y}-\sum_{r=0}^{n-2} \Tr { Z^{n-2-r} W^2 Z^rY } \quad
1333: \quad \nonumber
1334: \end{eqnarray}
1335:
1336: The positive parity sequences of eigenstates , if $n$ is even , may be written
1337: \begin{eqnarray}
1338: && \bigtriangleup_2=8 \sin^2 \frac{k\pi}{n} \quad , \quad
1339: u^{(k)}=(v_1+v_n)\cos (k\pi/n)+\sum_{j=2}^{[n/2]} \cos \left( \frac{k \pi (2j-1)}{n} \right) ( v_j+v_{n-j+1}) \nonumber\\
1340: && (D.3) \qquad \qquad {\rm where} \qquad v_j+v_{n-j+1} =\nonumber\\
1341: &&\sum_{r=0}^{j-2} {\rm tr}\,[(Z^{n-j}WZ^r W Z^{j-2-r}+Z^{j-2-r}W Z^r W Z^{n-j})Y-2\,Z^r W Z^{n-j}W Z^{j-2-r}Y] + \qquad \nonumber\\
1342: && +\sum_{r=0}^{n-j-1}{\rm tr}\,[ (Z^r W Z^{n-j-1-r} W Z^{j-1}+
1343: Z^{j-1}W Z^{n-j-1-r} W Z^r)Y-2\,Z^r W Z^{j-1}W Z^{n-j-1-r}Y]
1344: \qquad \nonumber
1345: \end{eqnarray}
1346: where $k=1,2, \cdots , k_{max}$ , $k_{max}<n/2$.\\ If $n$ is odd integer, the above $u^{(k)}$ should be replaced by
1347: $u^{(k)}+\cos (k\pi)\, v_{(n+1)/2}$ .\\
1348:
1349: The negative parity sequences of eigenstates may be written
1350: \begin{eqnarray}
1351: && \bigtriangleup_2=8 \sin^2 \frac{k\pi}{n+1} \quad , \quad
1352: u^{(k)}=(v_1-v_n)\sin (2k\pi/(n+1))+\sum_{j=2}^{[n/2]} \sin \left( \frac{2k \pi j}{n+1} \right) ( v_j-v_{n-j+1}) \nonumber\\
1353: && (D.4) \qquad \qquad {\rm where} \qquad v_j-v_{n-j+1} =\nonumber\\
1354: &&\sum_{r=0}^{j-2} {\rm tr}\,[(Z^{n-j}W Z^r W Z^{j-2-r}-Z^{j-2-r}W Z^r W Z^{n-j})Y] + \qquad \nonumber\\
1355: && +\sum_{r=0}^{n-j-1}{\rm tr}\,[ (Z^r W Z^{n-j-1-r} W Z^{j-1}-
1356: Z^{j-1}W Z^{n-j-1-r} W Z^r)Y]
1357: \quad \nonumber
1358: \end{eqnarray}
1359: where $k=1,2, \cdots , k_{max}$ , $k_{max}<(n+1)/2$.\\
1360:
1361:
1362: \begin{thebibliography}{99}
1363:
1364: \bibitem{r1} N. Beisert, {\it The Dilatation Operator of $\mathcal {N}=4$ Super Yang-Mills Theory and Integrability}, Phys. Rep. {\bf 405} 1, (2005) hep-th/0407277 .\\
1365:
1366: \bibitem{r2} J.A. Minahan, {\it A brief introduction to the Bethe ansatz in ${\mathcal N}=4$ super-Yang-Mills}, J. of Phys. {\bf A 39} (2006).\\
1367:
1368: \bibitem{ple} J Plefka, {\it Spinning strings and integrable spin chains in the AdS/CFT correspondence}, hep-th/0507136.\\
1369:
1370: \bibitem{mz} J.A. Minahan and K. Zarembo, {\it The Bethe-Ansatz for ${\mathcal N}=4$ Super Yang-Mills}, JHEP {\bf 03} (2003) 013, hep-th/0212208.\\
1371:
1372: \bibitem{big} N. Biggs, {\it Algebraic graph theory} , p.128 Cambridge University Press, Second Edition 1993.\\
1373:
1374: \bibitem{boe} See for instance H. Boerner, {\it Representations of groups} , North Holland 1963.\\
1375:
1376: \bibitem{enc} G. James and A. Kerber, {\it The Representation Theory of the Symmetric Group} , Encyclopedia of Mathematics and its Applications, Addison-Wesley (1981) .\\
1377:
1378: \bibitem{be} N. Beisert, {\it BMN Operators and Superconformal Symmetry}, Nucl. Phys. {\bf B659} (2003) , 79, hep-th/0211032.\\
1379:
1380: \bibitem{be2} N. Beisert, C. Kristjansen and M. Staudacher, {\it The dilatation operator of ${\mathcal N}=4$ conformal super Yang-Mills theory}, Nucl. Phys. {\bf B664} (2003) 131, hep-th/0303060.\\
1381:
1382:
1383:
1384:
1385: \end{thebibliography}
1386:
1387:
1388: \end{document}
1389:
1390:
1391: