1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %%%%%%%% Mass Splitting for Staggered Fermions %%%%%%%%
3: %%%%%%%% %%%%%%%%
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
5: %% LATEX File
6: %%
7: %%
8: %%
9:
10: \def\be{\begin{equation}}
11: \def\ee{\end{equation}}
12: \def\ba{\begin{eqnarray}}
13: \def\ea{\end{eqnarray}}
14: \def\no{\nonumber \\}
15: \newcommand{\nid}{\noindent}
16: %\renewcommand{\thefootnote}{\fnsymbol{footnote}}
17:
18: \documentclass{PoS}
19:
20: \title{Mass Splitting of Staggered Fermion
21: and $SO(2D)$ Clifford Algebra}
22:
23: \ShortTitle{Mass Splitting of Staggered Fermion
24: and $SO(2D)$ Clifford Algebra}
25:
26: \author{\speaker{Morio Hatakeyama}, Hideyuki Sawanaka\\
27: Graduate School of Science and Technology, Niigata University\\
28: Ikarashi 2-8050, Niigata 950-2181, Japan\\
29: E-mail: \email{hatake@muse.sc.niigata-u.ac.jp},
30: \email{hide@muse.sc.niigata-u.ac.jp}}
31: \author{Hiroto So\thanks{This work is supported in part by
32: the Grants-in-Aid for Scientific Research No. 17540242 from
33: the Japan Society for the Promotion of Science.}\\
34: Department of Physics,
35: Niigata University\\
36: Ikarashi 2-8050, Niigata 950-2181, Japan\\
37: E-mail: \email{so@muse.sc.niigata-u.ac.jp}}
38:
39: \abstract{We present a new method to introduce rotationally
40: invariant terms in staggered fermions
41: which is based on an $SO(2D)$ Clifford algebra formulation,
42: where $D$ means the number of space-time dimensions.
43: We have four candidates for improved mass terms that can split the degenerate mass of staggered fermions.
44: %We have four candidates of mass terms.
45: Among them, we analyze three types of combinations and find only one case
46: %In four candidates of additional terms,
47: %we find only one case
48: that can identify with the light single Dirac mode.}
49:
50: \FullConference{XXIV International Symposium on Lattice Field Theory\\
51: July 23-28 2006\\
52: Tucson Arizona, USA}
53:
54: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
55: %%
56: \usepackage{bm}
57: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
58: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
59: \begin{document}
60: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
61: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
62: \section{Introduction}
63: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
64: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
65: Staggered fermions are formulated
66: in which species doublers of a Dirac field
67: are interpreted as physical degrees of freedom, {\it tastes}, on
68: lattice~\cite{KS,Susskind}.
69: %Although the fermion determinant
70: %has many advantages in the cost of
71: %its numerical calculation~\cite{HPQCD,MILC},
72: However, it remains for a 4-fold degeneracy problem of tastes
73: in four dimensions to be unsolved.
74: Although a fourth-root trick of the determinant
75: in a staggered Dirac operator is an approach to unfold the degeneracy
76: and studies on its theoretical basis
77: are developed~\cite{Adams:2003rm,Shamir,BGS},
78: we have no local expression of
79: one taste Dirac fermion after the fourth-root trick.
80:
81: Avoiding the trick, there are pioneering works
82: for solving the degeneracy tried
83: by improved staggered fermion approaches~\cite{Mitra,Hart}.
84: The improved actions generally include more operators than
85: the original staggered one and are difficult to treat them~\cite{Toolkit}.
86: For the control of their operators,
87: we make use of staggered fermions
88: on a $D$-dimensional lattice space based on
89: an $SO(2D)$ Clifford algebra,
90: and a discrete rotational symmetry can be represented by the algebra~\cite{IKMSS}.
91:
92:
93: In this article,
94: %we analyze the mass splitting of degenerate tastes
95: %by adding four operators to the original staggered action in two dimensions.
96: to split degenerate tastes,
97: we add new four operators to the original staggered action in two dimensions.
98: Only these four operators keep the discrete rotational symmetry
99: in any dimension~\cite{IKMSS}.
100: The total mass matrix analysis
101: is insufficient because the matrix does not commute with
102: the kinetic term. Therefore, we also analyze
103: the propagator and the pole of the improved free staggered Dirac operator.
104: It is found that only one combination
105: in these operators is a good candidate after these analyses.
106: More details can be found in Ref.~\cite{HSS}.
107:
108: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
109: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
110: \section{Formulation of Staggered Fermions and Rotational Symmetry}
111: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
112: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
113: The formulation of staggered fermions
114: on the $D$-dimensional lattice space has been presented
115: based on the $SO(2D)$ Clifford algebra~\cite{IKMSS}. The basic idea
116: is that the dimension of the total representation space
117: including spinor and taste spaces, $2^D$
118: is the same as that of an $SO(2D)$ spinor representation. $2^D$
119: is also the same as the number of sites in a $D$-dimensional
120: hypercube. To avoid the double counting of sites,
121: the lattice coordinate $n_{\mu}$ is noted by
122: \be
123: n_{\mu} = 2 N_{\mu}+ c_{\mu} + r_{\mu},
124: \ee
125: \nid
126: where $N_{\mu}$ is the global coordinate of the hypercube.
127: In this case, a fundamental unit is $2a$, where $a$ is a lattice
128: constant, and is set to unity. $ c_{\mu} = 1/2$ for any $\mu$
129: means the coordinate of a center in the $D$-dimensional hypercube and
130: $r_\mu$ does the relative coordinate of a site to the center.
131: The relative coordinate is the same as a weight of the spinor
132: representation in $SO(2D)$.
133:
134: %%%%%%%%%%
135: Although our formulation can be generalized, we consider a free theory
136: in a two-dimensional lattice, for simplicity. Relative
137: coordinates of four sites around a plaquette are written by
138: \be
139: (r_1, r_2) = (-1/2, -1/2),~ (-1/2, 1/2),~ (1/2, -1/2),~ (1/2, 1/2) .
140: \label{4sites}
141: \ee
142: \nid
143: %as shown in Figure~\ref{unit-2dim}.
144: Actually, our staggered fermion is defined on sites (\ref{4sites}) as
145: \be
146: \Psi(n) \equiv \Psi_{r}(N) = \left(\begin{array}{c}
147: \Psi_{(-1/2,-1/2)} \\ \Psi_{(-1/2,1/2)} \\
148: \Psi_{(1/2,-1/2)} \\ \Psi_{(1/2,1/2)} \end{array}\right) (N)
149: \equiv \left(\begin{array}{c}
150: \Psi_1 \\ \Psi_2 \\ \Psi_3 \\ \Psi_4 \end{array}\right) (N).
151: \ee
152: \nid
153: %%
154: It is noted that $\Psi_1$ and $\Psi_4$ are put on even sites and
155: $\Psi_2$ and $\Psi_3$ are put on odd sites.
156:
157: %%%%%%%%%%
158: An $SO(4)$ Clifford algebra plays a crucial role in
159: two-dimensional cubic lattice formulations~\cite{IKMSS}. The original
160: staggered fermion action~\cite{KS,Susskind} can be written as
161: \be
162: S_{st} = \sum_{N,N^\prime,r,r^\prime,\mu,\vec{\tau}}
163: \bar{\Psi}_r(N)(D_{\mu}^{\vec{\tau}})_{(N,N^\prime)}
164: (\Gamma_{\mu,\vec{\tau}})_{(r,r^\prime)}\Psi_{r^\prime}(N^\prime) ,
165: \label{action}
166: \ee
167: \nid
168: where $\vec{\tau}$ is a two-dimensional vector
169: with its components of $\pm 1/2$ and
170: $ D_{\mu}^{\vec{\tau}}$ for $\mu = 1, 2$, is
171: a generalized difference operator defined by
172: \be
173: (D_{\mu}^{\vec{\tau}})_{(N,N^\prime)} \equiv
174: \frac{1}{2^2}\sum_{\vec{\sigma}=0,1}
175: (-1)^{(\vec{c}+\vec{\tau})\cdot\vec{\sigma}}
176: (\nabla_\mu^{\vec{\sigma}})_{(N,N^\prime)} ,
177: \ee
178: \nid
179: with
180: \be
181: (\nabla_\mu^{\vec{\sigma}})_{(N,N^\prime)} = \left\{\begin{array}{ll}
182: \delta_{N,N^\prime}
183: %U_{2N+\vec{\sigma},\mu}
184: -
185: \delta_{N-\hat{\mu},N^\prime}
186: %U_{2N+\vec{\sigma}-\hat{\mu},\mu}^\dag
187: \equiv \nabla_\mu^-, & ~\sigma_\mu=0, \\
188: \\
189: \delta_{N+\hat{\mu},N^\prime}
190: %U_{2N+\vec{\sigma},\mu}
191: -
192: \delta_{N,N^\prime}
193: %U_{2N+\vec{\sigma}-\hat{\mu},\mu}^\dag
194: \equiv\nabla_\mu^+, & ~\sigma_\mu=1.
195: \end{array}\right.
196: %% \\[6mm]
197: \label{gen-diff}
198: \ee
199: \nid
200: %%
201: $\vec{\sigma}$ is a two-dimensional vector dual to $\vec{\tau}$ and
202: $\nabla_\mu^+$ ($\nabla_\mu^-$) implies
203: a forward (backward) difference operator along the $\mu$-direction,
204: respectively.
205: The matrix
206: $\Gamma_{\mu,\vec{\tau}}$ in our action (\ref{action}) is
207: composed of the $SO(4)$ Clifford algebra
208: $\Gamma_{\mu,-\vec{c}} \equiv \gamma_{\mu}$ and
209: $\Gamma_{\mu,-\vec{c}+\vec{e}_\mu} \equiv i\tilde{\gamma}_{\mu}$,
210: \be
211: (\Gamma_{\mu,\vec{\tau}})_{(r,r^\prime)} \equiv
212: \left\{\begin{array}{ll}
213: ((\sigma_3^{1/2+\tau_1} \otimes \sigma_3^{1/2+\tau_2})
214: \times
215: (\sigma_1 \otimes \bm{1}))_{(r,r^\prime)}~, ~~\mu=1, \\
216: \\
217: ((\sigma_3^{1/2+\tau_1} \otimes \sigma_3^{1/2+\tau_2})
218: \times
219: (\sigma_3 \otimes \sigma_1))_{(r,r^\prime)}~, ~~\mu=2,
220: \end{array}\right.
221: \ee
222: \nid
223: where $\vec{e}_\mu$ is the unit vector along the $\mu$-direction.
224: Here we denote the fundamental algebra, or the $SO(4)$ Clifford algebra as
225: \be
226: \{\gamma_{\mu},\gamma_{\nu}\} =
227: \{\tilde{\gamma}_{\mu},\tilde{\gamma}_{\nu}\} =
228: 2\delta_{\mu\nu}~,~~
229: %\ee
230: %%
231: %\be
232: \{\gamma_{\mu},\tilde{\gamma}_{\nu}\}=0.
233: \ee
234: %\be
235: % \Gamma_{1, -\vec{c}} = \left(\begin{array}{cccc}
236: % 0 & 0 & 1 & 0 \\
237: % 0 & 0 & 0 & 1 \\
238: % 1 & 0 & 0 & 0 \\
239: % 0 & 1 & 0 & 0 \end{array}\right) \equiv \gamma_1, \quad
240: % \Gamma_{2, -\vec{c}} = \left(\begin{array}{cccc}
241: % 0 & 1 & 0 & 0 \\
242: % 1 & 0 & 0 & 0 \\
243: % 0 & 0 & 0 & -1 \\
244: % 0 & 0 & -1 & 0 \end{array}\right) \equiv \gamma_2,
245: %\ee
246: %%
247: %\be
248: % -i \Gamma_{1, (1/2,-1/2)} = \left(\begin{array}{cccc}
249: % 0 & 0 & -i & 0 \\
250: % 0 & 0 & 0 & -i \\
251: % i & 0 & 0 & 0 \\
252: % 0 & i & 0 & 0 \end{array}\right) \equiv \tilde{\gamma}_1, \quad
253: % -i \Gamma_{2, (-1/2,1/2)} = \left(\begin{array}{cccc}
254: % 0 & -i & 0 & 0 \\
255: % i & 0 & 0 & 0 \\
256: % 0 & 0 & 0 & i \\
257: % 0 & 0 & -i & 0 \end{array}\right) \equiv \tilde{\gamma}_2,
258: %\ee
259: %\nid
260: %and is described by
261: %$\Gamma_{\mu,\vec{\tau}} \equiv
262: % (i\tilde{\gamma}_1\gamma_1)^{1/2+\tau_1}
263: % (i\tilde{\gamma}_2\gamma_2)^{1/2+\tau_2} \gamma_{\mu}$.
264: %Here we denote the fundamental algebra, or the $SO(4)$ Clifford algebra as
265: %\be
266: % \{\gamma_{\mu},\gamma_{\nu}\} =
267: % \{\tilde{\gamma}_{\mu},\tilde{\gamma}_{\nu}\} =
268: %2\delta_{\mu\nu}~,~~
269: %\ee
270: %%
271: %\be
272: %% \{\gamma_{\mu},\tilde{\gamma}_{\nu}\}=0.
273: %\ee
274: %\nid
275: %and
276: %\be
277: % \{\Gamma_{\mu,\vec{\tau}}, \Gamma_5\} = 0 ,
278: %\ee
279: %\nid
280: %where
281: %$\Gamma_5 \equiv \gamma_1\gamma_2 \tilde{\gamma}_1\tilde{\gamma}_2
282: % = diag(1,-1,-1,1)$. {}From the algebra, we find that
283: %a massless staggered fermion has an even-odd symmetry
284: %\be
285: % \Psi \rightarrow e^{i\theta \Gamma_{5}}\Psi, \quad
286: % \bar{\Psi} \rightarrow \bar{\Psi}e^{i\theta \Gamma_{5}}.
287: %\ee
288: %%%%%%%%%%
289: For a discrete rotation with angle $\pi/2$ around the center,
290: the transformations of global and relative coordinates are denoted by
291: $N \rightarrow R(N)$,~~$r \rightarrow R(r)$,~~and that of fermion is
292: \be
293: \Psi(N) \rightarrow V_{12} \Psi(R(N)).
294: \ee
295: \nid
296: $V_{12}$ is a rotation matrix about a spinor index
297: in the $SO(4)$ base, up to a phase factor given by a form
298: \be
299: V_{12} = \frac{e^{i\vartheta}}{2} \Gamma_5
300: (\tilde{\gamma}_1-\tilde{\gamma}_2)(1+\gamma_1\gamma_2),
301: %=
302: % e^{i\vartheta} \left(\begin{array}{cccc}
303: % 0 & 0 & -i & 0 \\
304: % i & 0 & 0 & 0 \\
305: % 0 & 0 & 0 & i \\
306: % 0 & i & 0 & 0 \end{array}\right).
307: \ee
308: \nid
309: where
310: $\Gamma_5 \equiv \gamma_1\gamma_2\tilde{\gamma}_1\tilde{\gamma}_2
311: = diag(1,-1,-1,1)$.
312: %satisfies $\{\Gamma_{\mu,\vec{\tau}},\ \Gamma_5 \}=0$.
313: %%%%%%%%%%
314: Only the following four operators $\bar{\Psi} O_i \Psi$ for $i=1, 2, 3, 4$,
315: \be
316: O_1 = \bm{1}, \quad
317: O_2 = i\gamma_1\gamma_2 \equiv \Gamma_3, \quad
318: O_3 = \tilde{\gamma}_1 + \tilde{\gamma}_2, \quad
319: O_4 = \Gamma_3 (\tilde{\gamma}_1 + \tilde{\gamma}_2),
320: \label{ops}
321: \ee
322: \nid
323: are invariant under the rotation $V_{12}O_i V_{12}^\dag$.
324: Our analyses in the following sections concentrate on
325: the improved staggered fermion action by these four matrices.
326: %with $U_{2N+\vec{\sigma},\mu}=1$.
327:
328:
329: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
330: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
331: \section{Analysis of Mass Matrices}
332: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
333: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
334: To split masses in desired degenerate tastes we introduce four
335: rotationally invariant operators which we denote as
336: %{\mathcal M}_i \equiv
337: $\bar{\Psi} O_i \Psi$~\cite{IKMSS},
338: for the original staggered fermion action~(\ref{action}).
339: The total mass matrix form which is invariant under the rotation
340: by $\pi/2$ in two dimensions is given as
341: %\ba
342: \be
343: M_R = m_1 \bm{1} + m_2 \Gamma_3
344: + m_3 (\tilde{\gamma}_1 + \tilde{\gamma}_2)
345: + m_4 \Gamma_3(\tilde{\gamma}_1 + \tilde{\gamma}_2),
346: \label{total-mass}
347: \ee
348: %\ea
349: \nid
350: where $m_1$, $m_2$, $m_3$ and $m_4$ are parameters of each operator
351: in Eq.~($\ref{ops}$). $M_R$ has four eigenvalues
352: \ba
353: m_1-m_2-\sqrt{2}m_3+\sqrt{2}m_4 , & &
354: m_1-m_2+\sqrt{2}m_3-\sqrt{2}m_4 , \no
355: m_1+m_2-\sqrt{2}m_3-\sqrt{2}m_4 , & &
356: m_1+m_2+\sqrt{2}m_3+\sqrt{2}m_4 .
357: \label{eigenvalues}
358: \ea
359: \nid
360: %%
361: A 4-component spinor should be separated into two
362: 2-component spinors since a two-dimensional Dirac spinor is composed
363: of a 2-component mode and we keep the rotational invariance
364: even under a finite lattice constant%
365: \footnote{If one permits the rotational invariance only after taking
366: the continuum limit, it is not necessary for degeneracy of a heavy
367: mode and there are six more cases derived from Eq.~(\ref{eigenvalues}).}.
368: %%
369: Actually all possibilities of this separation are three cases
370: and are listed in Table~\ref{3C}.
371: %Let us analyze these mass matrices in order explicitly.
372:
373: %%%%%%%%%%%%%%%%%%%%
374: %%%%%%%%%%%%%%%%%%%%
375: \begin{table}[h]
376: \begin{center}
377: %%\setlength{\doublerulesep}{0.8pt}
378: %%\setlength{\tabcolsep}{1.5pt}
379: \renewcommand{\arraystretch}{1.25}
380: \begin{tabular}{|c|c|c|c|} \hline
381: & parameter conditions & rotationally invariant mass term &
382: mass eigenvalues \\ \hline
383: case~ 1 & $m_2 = m_3 = 0$ &
384: $M_{R1}=m_1\bm{1}+m_4\Gamma_3(\tilde{\gamma}_1+\tilde{\gamma}_2)$ &
385: $m_1 \pm \sqrt{2}m_4$ \\ \hline
386: case~ 2 & $m_2 = m_4 = 0$ &
387: $M_{R2}=m_1\bm{1}+m_3(\tilde{\gamma}_1+\tilde{\gamma}_2)$ &
388: $m_1 \pm \sqrt{2}m_3$ \\ \hline
389: case~ 3 & $m_3 = m_4 = 0$ & $M_{R3}=m_1\bm{1}+m_2\Gamma_3$ &
390: $m_1 \pm m_2$ \\ \hline
391: \end{tabular}
392: %%\renewcommand{\arraystretch}{1}
393: \caption{Three cases for the mass splitting into two spinors.}
394: \label{3C}
395: \end{center}
396: \end{table}
397: %%%%%%%%%%%%%%%%%%%%
398: %%%%%%%%%%%%%%%%%%%%
399:
400: %%%%%%%%%%%%%%%%%%%%
401: %\paragraph{Case~1} ~
402:
403: %\nid
404: %The mass matrix
405: %\be
406: % M_{R1} = \left(\begin{array}{cccc}
407: % m_1 & m_4 & -m_4 & 0 \\
408: % m_4 & m_1 & 0 & m_4 \\
409: % -m_4 & 0 & m_1 & m_4 \\
410: % 0 & m_4 & m_4 & m_1\end{array}\right)
411: %\ee
412: %\nid
413: %can be diagonalized as
414:
415: %%%%%%%%%%%%%%%%%%%%
416:
417: %\nid
418: %The mass matrix
419: %\be
420: % M_{R2} = \left(\begin{array}{cccc}
421: % m_1 & -im_3 & -im_3 & 0 \\
422: % im_3 & m_1 & 0 & -im_3 \\
423: % im_3 & 0 & m_1 & im_3 \\
424: % 0 & im_3 & -im_3 & m_1 \end{array}\right)
425: %\ee
426: %\nid
427: %can be diagonalized as
428: %\be
429: % M_{R2}^{diag} = P_2^\dag M_{R2} P_2
430: % = diag(m_1-\sqrt{2}m_3, m_1-\sqrt{2}m_3, m_1+\sqrt{2}m_3, m_1+\sqrt{2}m_3),
431: %\ee
432: %\nid
433: %by a unitary matrix
434: %\be
435: % P_2 = \frac{1}{2\sqrt{2}} %\left(\begin{array}{cccc}
436: % \pmatrix{
437: % -\sqrt{2}i & \sqrt{2}i & -\sqrt{2}i & -\sqrt{2}i \cr
438: % -1+i & 1+i & 1+i & 1-i \cr
439: % -1-i & 1-i & 1-i & 1+i \cr
440: % \sqrt{2} & \sqrt{2} & -\sqrt{2} & \sqrt{2}},
441: %\end{array}\right) \quad , \quad P_2^\dag P_2=P_2 P_2^\dag =1
442: %\ee
443: %\nid
444: %with the transformed spinor
445: %\be
446: % \Psi_{M_2}(N) \equiv P_2^\dag \Psi (N) = \frac{1}{2\sqrt{2}}
447: % \left(\begin{array}{cccc}
448: % \sqrt{2}i\Psi_1 - (1+i)\Psi_2 - (1-i)\Psi_3 + \sqrt{2}\Psi_4 \\
449: % -\sqrt{2}i\Psi_1 + (1-i)\Psi_2 + (1+i)\Psi_3 + \sqrt{2}\Psi_4 \\
450: % \sqrt{2}i\Psi_1 + (1-i)\Psi_2 + (1+i)\Psi_3 - \sqrt{2}\Psi_4 \\
451: % \sqrt{2}i\Psi_1 + (1+i)\Psi_2 + (1-i)\Psi_3 + \sqrt{2}\Psi_4
452: % \end{array}\right)(N).
453: %\ee
454: %%\ba
455: %% \Psi_{M_2} &\equiv& P_2^\dag \Psi
456: %% = \frac{1}{2\sqrt{2}} %\left(\begin{array}{cccc}
457: %% \pmatrix{
458: %% \sqrt{2}i & -1-i & -1+i & \sqrt{2} \cr
459: %% -\sqrt{2}i & 1-i & 1+i & \sqrt{2} \cr
460: %% \sqrt{2}i & 1-i & 1+i & -\sqrt{2} \cr
461: %% \sqrt{2}i & 1+i & 1-i & \sqrt{2}}
462: %% %\end{array}\right)
463: %% %\left(\begin{array}{l}
464: %% \pmatrix{\Psi_1 \cr \Psi_2 \cr \Psi_3 \cr \Psi_4} \no
465: %% %\end{array}\right) \nonumber \\
466: %% &=& \frac{1}{2\sqrt{2}} %\left(\begin{array}{c}
467: %% \pmatrix{
468: %% \sqrt{2}i\Psi_1 - (1+i)\Psi_2 - (1-i)\Psi_3 + \sqrt{2}\Psi_4 \cr
469: %% -\sqrt{2}i\Psi_1 + (1-i)\Psi_2 + (1+i)\Psi_3 + \sqrt{2}\Psi_4 \cr
470: %% \sqrt{2}i\Psi_1 + (1-i)\Psi_2 + (1+i)\Psi_3 - \sqrt{2}\Psi_4 \cr
471: %% \sqrt{2}i\Psi_1 + (1+i)\Psi_2 + (1-i)\Psi_3 + \sqrt{2}\Psi_4}.
472: %% %\end{array}\right)
473: %%\ea
474:
475: %%%%%%%%%%
476: %%\subsection*{3.2 $SO(D)$ spinor}
477: %The feature of our formulation is to keep
478: %the discrete rotational invariance.
479: After the mass splitting,
480: we can find the character of a Dirac spinor under the rotation,
481: \be
482: \psi(x) \rightarrow Q\psi(R(x)) ,
483: \ee
484: \nid
485: where
486: $Q = e^{(i\pi/4)\sigma_3}
487: = \pmatrix{ e^{i\pi/4} & 0 \cr 0 & e^{-i\pi/4}}$.
488: Actually in cases 1 and 2 we can keep
489: the property of a Dirac spinor on lattice.
490: %%\be
491: % \Psi(N) \rightarrow V_{12}\Psi(R(N)).
492: %\ee
493: %\nid
494: %%On the other hand,
495: By contrast, $\Psi(N)$ acts as a vector not as a spinor
496: in case 3. The properties of $2$-component spinors
497: under the rotation are summarized in Table \ref{DS}
498: \footnote{$M_R$ and $V_{12}$ can be diagonalized simultaneously
499: because $[M_R,V_{12}] = 0$.}.
500:
501: %%%%%%%%%%%%%%%%%%%%
502: %%%%%%%%%%%%%%%%%%%%
503: \begin{table}[ht]
504: \begin{center}
505: %%\renewcommand{\arraystretch}{1}
506: \begin{tabular}{|c|c|c|}
507: \hline & $V_{12}^{diag}$ & phase factor of $V_{12}$ \\ \hline
508: $\begin{array}{c} ~ \\[-5pt] {\rm case~1} \\[-5pt] ~ \end{array}$ &
509: $\left(\begin{array}{cc} Q & 0 \\ 0 & e^{i\pi}Q^\dag
510: \end{array}\right)$ &
511: $e^{i\vartheta}=e^{i\pi/2}=i$ \\ \hline
512: $\begin{array}{c} ~ \\[-5pt] {\rm case~2} \\[-5pt] ~ \end{array}$ &
513: $\left(\begin{array}{cc} Q & 0 \\ 0 & e^{i\pi}Q^\dag
514: \end{array}\right)$ &
515: $e^{i\vartheta}=e^{i\pi}=-1$ \\ \hline
516: $\begin{array}{c} ~ \\[-5pt] {\rm case~3} \\[-5pt] ~ \end{array}$ &
517: $\left(\begin{array}{cc} Q^2 & 0 \\ 0 & e^{i\pi/2}(Q^\dag)^2
518: \end{array}\right)$ &
519: $e^{i\vartheta}=e^{-i\pi/4}=(1-i)/\sqrt{2}$
520: \\ \hline
521: \end{tabular}
522: \caption{The properties of Dirac spinors under the rotation.}
523: \label{DS}
524: \end{center}
525: \end{table}
526: %%%%%%%%%%%%%%%%%%%%
527: %%%%%%%%%%%%%%%%%%%%
528:
529: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
530: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
531: \section{Pole Analysis and 2-point Functions}
532: %%\section*{\S 4. Pole Analysis and 2-point Functions}
533: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
534: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
535: Our adding terms do not commute with the staggered Dirac operator.
536: As a result, our analysis in the previous section
537: is insufficient to split masses. We must proceed in
538: the pole analysis of the theory because a pole mass is physical.
539: %For the help, Dirac fields are Fourier transformed as
540: %\be
541: % \Psi_r(N)=\displaystyle\int_{-\pi}^{+\pi}\frac{d^2p}{(2\pi)^2}\tilde{\Psi}_r(p)e^{ipN}, \quad
542: % \bar{\Psi}_r(N)=\displaystyle\int_{-\pi}^{+\pi}\frac{d^2p}{(2\pi)^2}\tilde{\bar{\Psi}}_r(p)e^{ipN},
543: %\ee
544: %\nid
545: %and the action becomes
546: %\ba
547: % S_{st}&=& \sum_{N,N^\prime,r,r^\prime,\mu,\vec{\tau}}
548: % \bar{\Psi}_r(N)(D_{\mu}^{\vec{\tau}})_{(N,N^\prime)}
549: % (\Gamma_{\mu,\vec{\tau}})_{(r,r^\prime)}\Psi_{r^\prime}(N^\prime)
550: % \no &=& \sum_{r,r^\prime}\int_{-\pi}^{+\pi}\frac{d^2p}{(2\pi)^2}
551: % \tilde{\bar{\Psi}}_r(-p)\left[\sum_\mu\left\{i\gamma_\mu\sin p_\mu
552: % +i\tilde{\gamma}_\mu(1-\cos p_\mu)\right\}\right]_{(r,r^\prime)}
553: % \tilde{\Psi}_{r^\prime}(p).
554: %\ea
555: %\nid
556: %%
557: The staggered Dirac operator in the momentum space is written as
558: \be
559: D_{st}(p)=\sum_\mu\left\{i\gamma_\mu\sin
560: p_\mu+i\tilde{\gamma}_\mu(1-\cos p_\mu)\right\}.
561: % \no &=& \left(\begin{array}{cccc}
562: % 0 & is_2 + c_2 & is_1 + c_1 & 0 \\
563: % is_2 - c_2 & 0 & 0 & is_1 + c_1 \\
564: % is_1 - c_1 & 0 & 0 & -is_2 - c_2 \\
565: % 0 & is_1 - c_1 & -is_2 + c_2 & 0 \end{array}\right),
566: \ee
567: %\nid
568: %where $s_i \equiv \sin p_i$ and $c_i \equiv 1-\cos p_i$ for
569: %$i=1, 2$, respectively.
570:
571: %%%%%%%%%%
572: Our steps to find a pole mass are as follows:
573: (i) set $p_1=0$ and $p_2= i \kappa~({\rm pure ~imaginary})$ of
574: the inverse propagator $D^{-1}$ in the momentum representation
575: where our rotationally invariant operators
576: %~(\ref{4rot-inv-ope})
577: are included; (ii) calculate four eigenvalues $\lambda$ of
578: $D^{-1}$; (iii) find values of $\kappa$ in setting $\lambda=0$. Four
579: values of $\kappa$ equal to pole masses. As mentioned in
580: sections 2 and 3, we keep the rotational invariance in our action and
581: generate two Dirac spinors with different masses.
582: %The light Dirac
583: %mass is denoted as $m_l$ and the heavy one is done as $m_h$.
584: %We note that parameters
585: %$m_1$, $m_2^{\prime}\equiv -i m_2$, $m_3^{\prime}\equiv -i m_3$ and
586: %$m_4$ are real to obtain real pole masses.
587: We define $m_1$, $m_2^{\prime}\equiv -i m_2$,
588: $m_3^{\prime}\equiv -i m_3$ and $m_4$ as real parameters
589: to obtain real pole masses and then denote by
590: $m_l$ and $m_h$ the light and heavy Dirac masses, respectively.
591: For each three cases results in brief of the pole analysis are
592: as follows.
593:
594: %%%%%%%%%%%%%%%%%%%%
595: \paragraph{$\bullet$ case~1} ~
596:
597: \nid
598: The pole mass is still splitting under $|m_4|<1$.
599: It is also found that we can take a limit
600: $|m_h| \rightarrow \infty$ for arbitrary
601: $m_l$ by performing $\epsilon \rightarrow 0$ in
602: an expression $m_4^2=1-\epsilon\;(0<\epsilon\ll1)$.
603:
604: %%%%%%%%%%%%%%%%%%%%
605: \paragraph{$\bullet$ case~2} ~
606:
607: \nid
608: The pole mass remains degenerate because the improved term
609: $m_3^\prime (\tilde{\gamma}_1 + \tilde{\gamma}_2)$ is absorbed into
610: the kinetic term.
611:
612: %%%%%%%%%%%%%%%%%%%%
613: \paragraph{$\bullet$ case~3} ~
614:
615: \nid
616: This case allows pole masses to split
617: although the rotational property of the eigenmode
618: is not a spinor from the discussion of the previous section.
619: %\footnote{In addition, we should mention that there is no massless mode
620: %in cases 2 and 3.}.
621: %%%%%%%%%%
622:
623: ~ \\
624:
625: Note that it is possible to take the light mass $m_l$ to zero by
626: tuning $m_1$ and $m_4$ only in case 1.
627: Solutions of the equation for the pole mass
628: under the massless condition $m_1^2=2m_4^2$ are determined as
629: \be
630: \sinh^2 \frac{m_l}{2}=0~,~~
631: %\label{massless-condition}
632: %\ee
633: %\be
634: \sinh^2 \displaystyle\frac{m_h}{2} = \frac{2m_4^2}{1-m_4^2} .
635: \label{massless-infty-condition}
636: \ee
637: %%%%%%%%%%
638: In addition, to decouple the heavy mode, we can throw the mass
639: up to infinity.
640: %and (\ref{inf-condition})
641: Actually from Eq.~(\ref{massless-infty-condition}),
642: we can realize massless and infinity modes as
643: Table~\ref{limitmode} simultaneously.
644: Although the formal
645: $\Gamma_5$
646: chiral projection which means even-site and odd-site separation of
647: fermion modes is not consistent with the rotational invariance of
648: a staggered Dirac action,
649: it is found that
650: infinity modes can be separately put on even or odd sites
651: \footnote{The eigenmode of the Dirac operator
652: around a massless pole is not orthogonal to that
653: around a heavy mass pole because their Dirac operators are
654: different from each other.}.
655: %%%%%%%%%%%%%%%%%%%%
656: %%%%%%%%%%%%%%%%%%%%
657: \begin{table}[h]
658: \begin{center}
659: %%\renewcommand{\arraystretch}{1}
660: \begin{tabular}{|c|c|c|}
661: \hline & massless modes & infinity modes \\ \hline
662: $\begin{array}{c} ~ \\[4mm] m_4 > 0 \\[4mm] ~ \end{array}$
663: & $\left(\begin{array}{c}
664: 1+\sqrt{2} \\ -1-\sqrt{2} \\ 1 \\ 1
665: \end{array}\right) , \quad
666: \left(\begin{array}{c}
667: 1-\sqrt{2} \\ 1-\sqrt{2} \\ -1 \\ 1
668: \end{array}\right)$
669: & $\left(\begin{array}{c} 1 \\ 0 \\ 0 \\ 0 \end{array}\right) , \quad
670: \left(\begin{array}{c} 0 \\ 0 \\ 0 \\ 1 \end{array}\right)$
671: \\ \hline
672: $\begin{array}{c} ~ \\[4mm] m_4 < 0 \\[4mm] ~ \end{array}$
673: & $\left(\begin{array}{c}
674: 1-\sqrt{2} \\ -1+\sqrt{2} \\ 1 \\ 1
675: \end{array}\right) , \quad
676: \left(\begin{array}{c}
677: 1+\sqrt{2} \\ 1+\sqrt{2} \\ -1 \\ 1
678: \end{array}\right)$
679: & $\left(\begin{array}{c} 0 \\ 0 \\ 1 \\ 0 \end{array}\right) , \quad
680: \left(\begin{array}{c} 0 \\ 1 \\ 0 \\ 0 \end{array}\right) $
681: \\ \hline
682: \end{tabular}
683: \caption{Eigenvectors of the improved Dirac operator in case 1 with
684: $m_1^2 = 2 m_4^2$.}
685: %Expressions of infinity modes are given by
686: %substituting $m_4^2 = 1$ for Eqs.~(\ref{em-p}) and (\ref{em-m}).}
687: \label{limitmode}
688: \end{center}
689: \end{table}
690: %%%%%%%%%%%%%%%%%%%%
691: %%%%%%%%%%%%%%%%%%%%
692:
693:
694: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
695: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
696: \section{Summary and Discussion}
697: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
698: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
699: We have studied the mass splitting of two-dimensional
700: staggered fermions based on the $SO(4)$ Clifford algebra. Introducing
701: four rotationally invariant operators,
702: we have analyzed three types of improved staggered Dirac operators
703: and found one possibility (case 1) for taking a single mode
704: in a two-dimensional free theory. The case keeps the splitting
705: not only in the analysis of the mass matrix itself
706: but also in the pole analysis including the kinetic term. According
707: to the improvement with respect to the rotational invariance, the
708: derived $2$-component modes can be regarded as the ordinary spinor
709: under the rotation by $\pi/2$. Furthermore, one can find
710: a massless mode in the case unexpectedly.
711: %The formal
712: %$\Gamma_5=\gamma_1\gamma_2\tilde{\gamma}_1\tilde{\gamma}_2=diag(1,-1,-1,1)$
713: %chiral projection which means even-site and odd-site separation of
714: %fermion modes is not consistent with the rotational invariance of
715: %a staggered Dirac action. Nevertheless, massless and infinity
716: %mode-representations in the case realize
717: %even-odd separation of the infinity mode.
718: Our future tasks are analyses of interacting theories and
719: the extension of our approach to four dimensions.
720: It is crucial that the stability for the massless condition under
721: quantum corrections by gauge interactions.
722:
723: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
724: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
725: %\acknowledgments{
726: %%The authors thank Hiroshi Suzuki for valuable comments at
727: %%``Niigata-Yamagata School'' (YITP-S-05-02). H. So is stimulated
728: %%for staggered fermions during the YITP workshop on
729: %%``Actions and Symmetries in Lattice Gauge Theory" (YITP-W-05-25).
730: %%He also would like to thank David B. Kaplan
731: %%for discussions about staggered fermions at that workshop,
732: %%which partially motivated the present work.
733: %This work is supported in part by the Grants-in-Aid for Scientific
734: %Research No. 17540242 from the Japan Society for the
735: %Promotion of Science.
736: %}
737:
738: %\clearpage
739: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
740: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
741: %%%%% references
742: %%%%%
743: \begin{thebibliography}{99}
744: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
745: \bibitem{KS}
746: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
747: J.~B.~Kogut and L.~Susskind,
748: \emph{Hamiltonian Formulation Of Wilson's Lattice Gauge Theories},
749: Phys.\ Rev.\ D {\bf 11}, 395 (1975).
750: %%CITATION = PHRVA,D11,395;%%
751:
752:
753: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
754: \bibitem{Susskind}
755: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
756: L.~Susskind,
757: \emph{Lattice Fermions},
758: Phys.\ Rev.\ D {\bf 16}, 3031 (1977).
759: %%CITATION = PHRVA,D16,3031;%%
760:
761:
762: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
763: \bibitem{Adams:2003rm}
764: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
765: D.~H.~Adams,
766: \emph{A simplified test of universality in lattice QCD},
767: Phys.\ Rev.\ Lett.\ {\bf 92}, 162002 (2004)
768: [{\tt hep-lat/0312025}];
769: %%CITATION = HEP-LAT 0312025;%%
770:
771: D.~H.~Adams,
772: \emph{Testing universality and the fractional power prescription for the staggered fermion determinant},
773: Nucl.\ Phys.\ Proc.\ Suppl.\ {\bf 140}, 148 (2005)
774: [{\tt hep-lat/0409013}];
775: %%CITATION = HEP-LAT 0409013;%%
776:
777: D.~H.~Adams,
778: \emph{On the fourth root prescription for dynamical staggered fermions},
779: Phys.\ Rev.\ D {\bf 72}, 114512 (2005)
780: [{\tt hep-lat/0411030}].
781: %%CITATION = HEP-LAT 0411030;%%
782:
783:
784: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
785: \bibitem{Shamir}
786: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
787: Y.~Shamir,
788: \emph{Locality of the fourth root of the staggered-fermion determinant: Renormalization-group approach},
789: Phys.\ Rev.\ D {\bf 71}, 034509 (2005)
790: [{\tt hep-lat/0412014}];
791: %%CITATION = HEP-LAT 0412014;%%
792:
793: Y.~Shamir,
794: \emph{Renormalization-group analysis of the validity of staggered-fermion QCD with the fourth-root recipe},
795: hep-lat/0607007.
796: %%CITATION = HEP-LAT 0607007;%%
797:
798:
799: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
800: \bibitem{BGS}
801: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
802: C.~Bernard, M.~Golterman and Y.~Shamir,
803: \emph{Observations on staggered fermions at non-zero lattice spacing},
804: Phys.\ Rev.\ D {\bf 73}, 114511 (2006)
805: [{\tt hep-lat/0604017}].
806: %%CITATION = HEP-LAT 0604017;%%
807:
808:
809: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
810: %%\bibitem{Giedt}
811: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
812: %J.~Giedt,
813: %\emph{Toward a systematic analysis of the fourth-root trick},
814: %hep-lat/0507002.
815: %%CITATION = HEP-LAT 0507002;%%
816:
817:
818: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
819: \bibitem{Mitra}
820: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
821: P.~Mitra,
822: \emph{Nondegenerate Susskind Fermions},
823: Phys.\ Lett.\ B {\bf 123}, 77 (1983);
824: %%CITATION = PHLTA,B123,77;%%
825:
826: P.~Mitra and P.~Weisz,
827: \emph{On Bare And Induced Masses Of Susskind Fermions},
828: Phys.\ Lett.\ B {\bf 126}, 355 (1983).
829: %%CITATION = PHLTA,B126,355;%%
830:
831:
832: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
833: \bibitem{Hart}
834: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
835: A.~Hart and E.~M{\"u}ller,
836: \emph{The locality of the square-root method for improved staggered quarks},
837: Phys.\ Rev.\ D {\bf 70}, 057502 (2004)
838: [{\tt hep-lat/0406030}].
839: %%CITATION = HEP-LAT 0406030;%%
840:
841:
842: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
843: \bibitem{Toolkit}
844: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
845: G.~W.~Kilcup and S.~R.~Sharpe,
846: \emph{A Tool Kit For Staggered Fermions},
847: Nucl.\ Phys.\ B {\bf 283}, 493 (1987).
848: %%CITATION = NUPHA,B283,493;%%
849:
850:
851: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
852: \bibitem{IKMSS}
853: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
854: K.~Itoh, M.~Kato, M.~Murata, H.~Sawanaka and H.~So,
855: \emph{Genuine symmetry of staggered fermion},
856: Prog.\ Theor.\ Phys.\ {\bf 114}, 631 (2005)
857: [{\tt hep-lat/0412039}].
858: %%CITATION = HEP-LAT 0412039;%%
859:
860:
861: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
862: \bibitem{HSS}
863: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
864: M.~Hatakeyama, H.~Sawanaka and H.~So,
865: \emph{Rotational symmetry and a light mode in two-dimensional staggered fermions},
866: hep-lat/0609025.
867: %%CITATION = HEP-LAT 0609025;%%
868:
869: \end{thebibliography}
870: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
871: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
872: \end{document}
873:
874: