hep-ph0309313/pp.tex
1: \documentclass[12pt]{article}
2: \usepackage{cite,graphicx}
3: 
4: \def\beq{\begin{equation}}
5: \def\eeq{\end{equation}}
6: \def\bea{\begin{eqnarray}}
7: \def\eea{\end{eqnarray}}
8: \newcommand{\gsim}{\lower.7ex\hbox{$\;\stackrel{\textstyle>}{\sim}\;$}}
9: \newcommand{\lsim}{\lower.7ex\hbox{$\;\stackrel{\textstyle<}{\sim}\;$}}
10: 
11: \addtolength\topmargin{-50pt}
12: \addtolength\textheight{105pt}
13: \addtolength\textwidth{60pt}
14: \addtolength\oddsidemargin{-38pt}
15: \setlength{\parindent}{20pt}
16: \setlength{\parskip}{6pt}
17: \frenchspacing
18: \sloppy
19: 
20: \begin{document}
21: \thispagestyle{empty}
22: \noindent
23: DESY 03-158\hspace*{\fill} September 27, 2003\\
24: \vspace*{2.6cm}
25: 
26: \begin{center}
27: {\Large\bf Grand Unification in the Projective Plane}\\[2cm]
28: {\large A. Hebecker}\\[.5cm]
29: {\it Deutsches Elektronen-Synchrotron, Notkestrasse 85, D-22603 Hamburg,
30: Germany}
31: \\[1.5cm]
32: 
33: {\bf Abstract}\end{center}
34: \noindent
35: A 6-dimensional grand unified theory with the compact space having the 
36: topology of a real projective plane, i.e., a 2-sphere with opposite points 
37: identified, is considered. The space is locally flat except for two conical 
38: singularities where the curvature is concentrated. One supersymmetry is 
39: preserved in the effective 4d theory. The unified gauge symmetry, for 
40: example SU(5), is broken only by the non-trivial global topology. In 
41: contrast to the Hosotani mechanism, no adjoint Wilson-line modulus 
42: associated with this breaking appears. Since, locally, SU(5) remains a good 
43: symmetry everywhere, no UV-sensitive threshold corrections arise and 
44: SU(5)-violating local operators are forbidden. Doublet-triplet splitting can 
45: be addressed in the context of a 6d $N\!=\!2$ super Yang-Mills theory with 
46: gauge group SU(6). If this symmetry is first broken to SU(5) at a fixed 
47: point and then further reduced to the standard model group in the above 
48: non-local way, the two light Higgs doublets of the MSSM are predicted by 
49: the group-theoretical and geometrical structure of the model. 
50: \newpage
51: 
52: 
53: 
54: 
55: 
56: \section{Introduction}
57: The mechanism of gauge symmetry breaking is an important open issue in the 
58: context of grand unified theories (GUTs). In the conventional approach, 
59: where the symmetry is broken by the vacuum expectation values (VEVs) of GUT 
60: Higgs fields, large representations are usually required and a potential 
61: enforcing the desired VEV has to be specified. Furthermore, solving the 
62: doublet-triplet splitting problem without fine tuning adds extra complexity 
63: to the models. 
64: 
65: An interesting alternative is provided by the Hosotani mechanism~\cite{hos}, 
66: which can be implemented in higher-dimensional theories compactified on 
67: manifolds with non-trivial topology. In this case, the symmetry breaking can 
68: be ascribed to the VEV of a Wilson line wrapping a non-contractible loop in 
69: extra dimensions. Related geometrical mechanisms of gauge symmetry breaking 
70: are used in string-compactifications on Calabi-Yau manifolds~\cite{chsw} and 
71: can naturally lead to doublet-triplet splitting~\cite{wit}. Interesting 
72: relations between extra-dimensional topology and gauge symmetry breaking 
73: also exist in field-theoretic settings with non-zero field-strength and 
74: no supersymmetry (see, e.g.,~\cite{fm}).
75: 
76: A serious problem of field-theoretic GUT models with gauge symmetry breaking 
77: by the Hosotani mechanism is the flatness of the classical potential of the 
78: Wilson line, which is protected from loop corrections by supersymmetry 
79: (SUSY). Thus, one usually encounters light adjoint fields (with mass of the 
80: order of the SUSY breaking scale) ruining precision gauge coupling 
81: unification. 
82: The situation is improved in 5-dimensional field-theoretic orbifold GUT 
83: models~\cite{kaw} (see~\cite{dhvw} for the original stringy idea), where 
84: the gauge symmetry is broken by boundary conditions and no adjoint moduli
85: arise. Although related 6d GUT models~\cite{abc} again have Wilson line 
86: VEVs, these lines can be contracted to zero length at conical singularities 
87: of the compact space, so that the VEV is fixed by a symmetry-breaking 
88: boundary condition. The presence of such boundaries (locations with 
89: reduced gauge symmetry) restricts the predictivity of orbifold GUTs because 
90: fields and operators that do not respect the GUT symmetry can be added at 
91: these points. It also introduces UV-sensitive corrections to gauge coupling 
92: unification, the natural size of which corresponds roughly to the thresholds 
93: of conventional 4d GUTs.
94: 
95: In this letter, an alternative field-theoretic mechanism for gauge symmetry 
96: breaking is considered. To understand the basic idea, it is sufficient to 
97: consider a 6d field theory with gauge group $G$ compactified on a 2-sphere. 
98: Modding out by a $Z_2$ symmetry which acts on the sphere as a reflection 
99: with respect to the center and in gauge space by the inner automorphism 
100: $g\to PgP^{-1}$ (with $P\in G,\,\,P^2=1$), one obtains a gauge theory on the 
101: projective plane. The symmetry of the 4d effective field theory is reduced 
102: to the subgroup commuting with $P$. Alternatively, the model can be 
103: characterized as a sphere with the insertion of a crosscap (for the basic 
104: geometric concepts see, e.g.,~\cite{gsw}), where the identification of the 
105: opposite edges at the crosscap is associated with a gauge twist $P$. It is 
106: easy to observe (see also below) that the topology requires the gauge 
107: twist to obey $P^2=1$. Thus, although the breaking is entirely non-local,
108: no Wilson line modulus appears. More generally, this type of discretized 
109: topological breaking occurs in situations where the fundamental group of 
110: the compact space is non-trival but finite, such as in many Calabi-Yau 
111: models (see~\cite{chsw,wit} and, in particular,~\cite{ww}). In this context, 
112: the real projective plane has been mentioned in~\cite{hmn}. The present 
113: realization combines the features of an extremely simple compact space with 
114: unbroken $N\!=\!1$ SUSY and thus has all the ingredients necessary for 
115: realistic model building. 
116: 
117: The paper is organized as follows: In Sect.~\ref{bc}, it is shown how the 
118: above illustrative example can be promoted to a more interesting GUT-like 
119: model. In particular, starting from a 6d $N\!=\!2$ super Yang-Mills (SYM) 
120: theory, a model with unbroken 4d $N\!=\!1$ SUSY, broken gauge symmetry and 
121: no moduli is constructed by orbifolding. The geometry is such that the 
122: curvature of the topological 2-sphere discussed above is concentrated at 
123: four conical singularities. The construction involves modding out a freely 
124: acting discrete symmetry (cf. the freely acting orbifold models familiar 
125: in string theory, especially in the context of SUSY breaking~\cite{fao}). 
126: 
127: In Sect.~\ref{su6}, an SU(6) model with doublet-triplet splitting is 
128: discussed. Given that the smallest truly unified group is SU(5), it is 
129: desirable that no SU(5) breaking fixed points exist. 
130: Doublet-triplet splitting is then most naturally realized if the Higgs is a 
131: bulk field. Since no gauged bulk matter is allowed by 6d $N\!=\!2$ SUSY, 
132: the gauge group has to be extended to allow Higgs doublets to emerge from 
133: the adjoint representation, the minimal choice being SU(6). Indeed, it is 
134: possible to construct a model where SU(6) is broken to SU(5) at one of the 
135: fixed points (the further breaking being topological) and two naturally light 
136: Higgs doublets appear in a way closely related to the models 
137: of~\cite{hns,pst,bn}. 
138: 
139: Sect.~\ref{sing} is devoted to the fixed-point breaking of SU(6) to SU(5).
140: For this several possibilities exist, the simplest one being to declare 
141: the gauge symmetry at the conical singularity to be reduced and to supply 
142: appropriate boundary conditions for the bulk fields. A more interesting 
143: possibility involves cutting off the tip of the cone by a 
144: gauge-symmetry-breaking 5d boundary and letting the length of the 
145: boundary tend to zero.
146: 
147: Conclusions and open questions are discussed in Sect.~\ref{con}. 
148: 
149: 
150: 
151: 
152: 
153: \section{Non-locally broken SYM theory without moduli}\label{bc}
154: Consider a 6d SYM theory with (1,1) SUSY and gauge group $G$, which can 
155: be thought of as deriving from a 10d SYM theory by torus compactification. 
156: (The large amount of supersymmetry is required to ensure the 
157: phenomenologically desirable $N=1$ SUSY of the 4d theory obtained 
158: after orbifolding.) It will prove convenient to describe this theory in 
159: terms of a 4d vector superfield $V$ and 3 chiral superfields with scalar 
160: components $\Phi_5=A_5+iA_8$, $\Phi_6=A_6+iA_9$ and $\Phi_7=A_7+i 
161: A_{10}$~\cite{mss}. Here $A_5$ and $A_6$ are the extra-dimensional (with 
162: respect to 4d) gauge field components of the 6d theory and $A_7\dots 
163: A_{10}$ are 6d adjoint scalars deriving in an obvious way from the gauge 
164: field of the associated 10d SYM theory.
165: 
166: The 6d theory is compactified on a torus $T^2$ parameterized by 
167: $(x_5,x_6)\in I\!\!R^2$ with the identifications $x_5\sim x_5+2\pi R_5$ and 
168: $x_6\sim x_6+2\pi R_6$. In the first step of orbifolding, the rotation 
169: symmetry $(x_5,x_6) \to -(x_5,x_6)$ is `modded out'. The action of this 
170: $Z_2$ symmetry in field space is chosen (using the symbol $\Phi_i$ for both 
171: the superfield and its scalar component) as $(\Phi_5,\Phi_6)\to -(\Phi_5, 
172: \Phi_6)$, with $V$ and $\Phi_7$ being inert. The fundamental compact space 
173: now has the topology of a 2-sphere with the curvature being concentrated 
174: at 4 conical singularities with deficit angle $\pi$ and can be visualized 
175: as the surface of a `pillow'~\cite{abc1}. The gauge symmetry is unrestricted. 
176: 
177: In the second, crucial step, a freely-acting $Z_2'$ symmetry is modded out,
178: at which point the gauge symmetry breaking is introduced.
179: The geometric $Z_2'$ action is defined by $x_5-\pi R_5/2\to 
180: -(x_5-\pi R_5/2)$ (reflection with respect to the line $x_5=\pi R_5/2$) and 
181: $x_6\to x_6+\pi R_6$ (translation along that line). The action in field 
182: space makes use of the gauge twist $P\in G$ ($P^2=1$) and is given by 
183: $V\to PVP^{-1}$, $\Phi_5\to -P\Phi_5P^{-1}$, $\Phi_6\to P\Phi_6P^{-1}$ and 
184: $\Phi_7\to -P\Phi_7P^{-1}$ (cf.~the rectangular models of~\cite{hns}). The 
185: topology of the resulting fundamental compact space (which is equivalent to
186: that of the projective plane) is illustrated in Fig.~\ref{cc}. The space 
187: is non-orientable, has no boundaries and the curvature is concentrated at the 
188: two fixed points $F_1$ and $F_2$, where conical singularities with deficit 
189: angle $\pi$ reside. 
190: 
191: \begin{figure}[ht]
192: \begin{center}
193: \includegraphics[width=6cm]{cc.eps}
194: \end{center}
195: \refstepcounter{figure}\label{cc}
196: 
197: \vspace*{-.2cm}
198: {\bf Figure~\ref{cc}:} Illustration of the topology of the $T^2/(Z_2\times 
199: Z_2')$ model discussed in the text. The manifold is flat everywhere except 
200: for the two conical singularities at $F_1$ and $F_2$. At the crosscap 
201: (symbolized by a circle with a cross) opposite points are identified, making 
202: the surface non-orientable.
203: \end{figure}
204: 
205: Given the above $Z_2$ and $Z_2'$ action, it is clear that the massless modes 
206: are the fields from $V$ commuting with $P$ (corresponding to the unbroken 
207: subgroup $H\subset G$) and the fields from $\Phi_7$ anticommuting with $P$
208: (corresponding to the complement of $H$ in $G$). The latter ones can become 
209: heavy due to a mass term for $\Phi_7$ localized at one of the fixed points. 
210: One 4d SUSY survives since the non-gauge part of both the $Z_2$ and $Z_2'$ 
211: action on the $\Phi_i$ fits into SU(3)~\cite{chsw}. 
212: 
213: Thus, in the specific case $G=\,\,$SU(5) and $P=\,\,$diag$(1,1,1,-1,-1)$ 
214: (cf.~\cite{kaw}), the gauge symmetry is broken to the standard model (SM) 
215: group in an entirely non-local fashion (the fixed points are SU(5) 
216: symmetric) and only the SM gauge multiplet is naturally light. Although 
217: SM matter can be added at the two fixed points, it appears difficult to 
218: realize doublet-triplet splitting given the extremely soft nature of the 
219: breaking. By construction, the fixed points have 4d $N\!=\!2$ SUSY which 
220: can, however, be broken simply by allowing for operators or a field content 
221: that are consistent only with $N\!=\!1$. 
222: 
223: The most interesting feature of the above toy model GUT is the 
224: absence of the Wilson line moduli that typically accompany a purely 
225: topological gauge symmetry breaking. This effect is linked to the small 
226: fundamental group of the projective plane ($Z_2$ rather than, say, $Z$ for 
227: an $S^1$) and can be understood as follows: By construction, a closed Wilson 
228: line loop passing once through the crosscap has the value $P$. A Wilson loop 
229: passing through the crosscap twice has the value $P^2$ and is, at the same 
230: time, contractible by the topology of the projective plane. Thus $P^2=1$ is 
231: required by consistency as long as the field strength vanishes everywhere
232: including the fixed points (i.e., an infinitesimal loop surrounding a fixed 
233: point has value 1). In field theory, this latter statement corresponds 
234: simply to a gauge-invariant boundary condition at the singularity. The fact 
235: that such a boundary condition can not be continuously deformed is natural 
236: from the point of view of string theory (cf. the discrete or quantized 
237: Wilson lines of~\cite{dhvw,inq}). Changing the value of the non-contractible 
238: Wilson loop while its square is fixed to equal unity corresponds to a
239: change by inner automorphism and thus to an unphysical, gauge degree of 
240: freedom. 
241: 
242: Note that it is possible to give the model an equivalent formulation with 
243: trivial (from the gauge theory perspective) transition function at the 
244: crosscap and non-vanishing, smooth
245: gauge connection throughout the compact space. To see this, consider the 
246: same geometry as above and introduce a gauge field which, near the 
247: crosscap boundary, is pointed parallel to that boundary and has the same
248: value everywhere along that boundary. This value can be chosen such that a 
249: Wilson line loop through the crosscap equals $P$. Furthermore, let the field
250: strength be zero everywhere in the bulk and let the infinitesimal Wilson 
251: loop surrounding $F_2$ vanish. Then, by global topology (cf. the discussion 
252: of `conifold GUTs' in~\cite{hr}) the infinitesimal Wilson loop surrounding 
253: $F_1$ equals $P^2=1$ and thus gauge symmetry is unbroken at that point. 
254: 
255: 
256: 
257: 
258: 
259: \section{An SU(6) model with doublet-triplet splitting}\label{su6}
260: As far as realistic model building is concerned, the most unsatisfactory 
261: feature of the above construction is the absence of doublet-triplet 
262: splitting. This problem can be overcome by extending the bulk gauge 
263: symmetry to SU(6) (cf.~\cite{hns,pst,bn}) and using the gauge twist 
264: $P=$\,\,diag$(1,1,1,1,-1,-1)$ for
265: the breaking at the crosscap. In addition, SU(6) is broken to SU(5) (defined 
266: as the submatrix containing the last 5 rows and last 5 columns of the 
267: original $6\times 6$ matrix) at the fixed point $F_1$. For the moment, 
268: the precise mechanism of this breaking can be left unspecified. (One may 
269: imagine boundary localized Higgs fields or a hard, ad-hoc breaking 
270: by boundary conditions.) It is, however, important that no non-trivial 
271: Wilson line surrounding $F_1$ is introduced and that the gauge fields in 
272: $V$ outside the SU(5) subgroup acquire mass due to local physics at $F_1$. 
273: 
274: Now, the low-energy gauge symmetry is that of the SM and the 
275: only potentially light fields (in addition to the gauge fields) are 
276: part of $\Phi_7$ and transform as 
277: \beq
278: ({\bf 3},{\bf 2})_{-5}+({\bf\bar{3}},{\bf 2})_5+({\bf 1},{\bf 2})_{-3}+
279: ({\bf 1},{\bf 2})_3
280: \eeq
281: under the SU(3)$\times$SU(2)$\times$U(1) subgroup of SU(5). They are 
282: associated with SU(5) multiplets relevant for the local physics at $F_1$:
283: \beq
284: ({\bf 3},{\bf 2})_5+({\bf\bar{3}},{\bf 2})_{-5}\,\subset\,{\bf 24}\quad,
285: \qquad({\bf 1},{\bf 2})_3\,\subset\,{\bf 5}\quad,\qquad({\bf 1},{\bf 2}
286: )_{-3}\,\subset\,{\bf\bar{5}}\,.
287: \eeq
288: Phenomenologically, the fields with the quantum numbers of $X,Y$ gauge bosons 
289: have to become massive while the two pure doublets may stay light and play 
290: the role of the two SM Higgs fields. It is clearly possible to introduce 
291: extra symmetries at $F_1$ allowing the first $({\bf 24\times 24})$ and 
292: forbidding the second $({\bf 5\times\bar{5}})$ SU(5) invariant mass term. 
293: SM matter could than be added in the form of brane-localized fields at $F_1$. 
294: The bulk should not be too large so that non-local effects can lead to a 
295: sufficient violation of the SU(5) symmetry of the $F_1$-based Yukawa 
296: couplings. Except for the somewhat ad hoc (though not fine-tuned) lightness
297: of the Higgs doublets, the model now appears to be satisfactory. 
298: 
299: However, it turns out that the above analysis is incomplete and that the 
300: local breaking of SU(6) to SU(5) at $F_1$ leads to the appearance of extra 
301: light states, not visible in the usual Kaluza-Klein mode expansion. The 
302: point is that some of the degrees of freedom associated with the Wilson
303: loop going through the crosscap cease to be gauge artifacts and become 
304: physical fields. To be more specific, let the value of the closed Wilson 
305: loop beginning at $F_1$ and going once through the crosscap be $W\!\in\,\,
306: $SU(6). So far, we have assumed $W=P$. It will now be shown that changing $W$ 
307: by inner automorphism (e.g., to $W'=UPU^{-1}$ with $U\!\in\,\,$SU(6)$\,$) 
308: corresponds to the excitation of a physical degree of freedom if the gauge 
309: symmetry at $F_1$ is reduced. Before doing so, let us argue that this change 
310: of $W$ is indeed a flat direction in fields space. For this, it is 
311: sufficient to consider a smooth scalar function $f$ on the fundamental 
312: space (cf.~Fig.~\ref{cc}) such that $f\equiv 0$ in a neighbourhood of the 
313: crosscap and $f\equiv 1$ in a neighbourhood of $F_1$. Writing $U=\exp(T)$, 
314: it becomes clear that the gauge fields that would be introduced by a gauge 
315: transformation $\exp(fT)$ lead to the desired inner-automorphism change of 
316: $W$ while corresponding, at the same time, to a flat direction in field 
317: space. 
318: 
319: Now, it remains to be shown that some of the freedom of rotating $W=P$ to 
320: $W'=UPU^{-1}$ corresponds indeed to physical fields. Group-theoretically, the 
321: SU(6) breaking at $F_1$ may be thought of as coming from a $Z_2$ twist $P'\!=
322: \,\,$diag$(-i,i,i,i,i,i)$. Given the restricted gauge symmetry at $F_1$, it 
323: is possible to write down the gauge invariant (and thus observable) 
324: operator tr$(W'P')=\,$tr$(UPU^{-1}P')$. Clearly, excitations generated by 
325: Lie algebra elements $T$ that 
326: anticommute with both $P$ and $P'$ lead to a change of the value of this 
327: operator. Their quantum numbers are the same as those of the two SM Higgs 
328: doublets obtained above from the chiral superfield $\Phi_7$. However, their 
329: physical interpretation is entirely different. While the former are 
330: conventional bulk zero modes with potential brane-localized mass terms, the 
331: latter are non-local degrees of freedom parameterizing the relative 
332: orientation of the two breaking patterns $P$ and $P'$. Thus, no local 
333: mass term or couplings to other fields can be written down. However, such 
334: couplings may be generated non-perturbatively or by integrating out 
335: appropriate gauged bulk fields~\cite{hmn,cgm}. 
336: 
337: There are now various options for solving the doublet-triplet splitting 
338: problem: One can use the Higgs doublets from $\Phi_7$ and argue that 
339: effective non-local operators will make the Wilson-line doublets heavy. 
340: This means that the problem of Wilson-line moduli (avoided by the crosscap 
341: breaking) is partially reintroduced and then solved in an ad-hoc way.
342: Thus, it appears more natural to give all $\Phi_7$ fields brane masses and 
343: to view the light Wilson line doublets, the existence of which is deeply 
344: rooted in the geometry of the model, as natural candidates for light Higgs 
345: fields. With matter localized at $F_1$, Yukawa couplings now arise from 
346: non-local effects. The main potential problem of this scenario is the 
347: difficulty of generating a sufficiently large top mass. Finally, the 
348: situation could be more involved and the required two light Higgs fields
349: might arise as a non-trivial linear superposition of the four potentially 
350: light doublets. For the purpose of this paper, it is sufficient that the 
351: existence of natural solutions to the doublet-triplet splitting problem 
352: has been demonstrated. The analysis of possibilities for generating 
353: realistic Yukawa couplings is postponed to a further investigation. 
354: 
355: 
356: 
357: 
358: 
359: \section{Breaking SU(6) to SU(5) at a singularity}\label{sing}
360: In the previous section, it was simply assumed that the gauge symmetry 
361: at $F_1$ is reduced to SU(5). Here, the corresponding possibilities for 
362: symmetry breaking are analysed in more detail and, in particular, a 
363: realization within the framework of field-theoretic orbifolding is suggested. 
364: 
365: Clearly, one of the options is symmetry breaking by a brane localized 
366: Higgs~\cite{nsw}. In this case, it is important to use a set of VEVs and 
367: appropriate extra symmetries ensuring that the ${\bf 24}$ from $\Phi_7$ 
368: acquires a brane mass while the ${\bf 5}$ and ${\bf\bar{5}}$ from $\Phi_7$ 
369: remain light. 
370: 
371: Within the framework of field theoretic orbifolding, the most natural 
372: choice appears to be a Wilson line with value $P'$ wrapping $F_1$. 
373: However, introducing such a Wilson line gives mass both to the doublets from 
374: $\Phi_7$ as well as to the Wilson line doublets discussed in detail in the 
375: previous section. Thus, doublet-triplet splitting becomes a serious problem. 
376: 
377: Therefore, this section focusses on an alternative option where, as 
378: illustrated in Fig.~\ref{cut}, the singularity $F_1$ is cut out and removed 
379: from the fundamental space. The group SU(6) is broken by the $Z_2$ reflection 
380: used to define the boundary conditions at the 4+1 dimensional brane created 
381: by this cut (cf. the disc and annulus models of~\cite{li}). After cutting 
382: out $F_1$, the global topology is that of a M\"obius strip. One might be 
383: concerned that now moduli (corresponding to Wilson lines wrapping the cycle 
384: of the M\"obius strip) appear after all. However, they can be given a mass 
385: by appropriate non-local operators involving, e.g., the Wilson loop along 
386: the boundary. Taking the limit where the length of the boundary tends to 
387: zero, such operators become effectively local.
388: 
389: \begin{figure}[ht]
390: \begin{center}
391: \includegraphics[width=3.5cm]{cut.eps}
392: \end{center}
393: \refstepcounter{figure}\label{cut}
394: 
395: \vspace*{-.2cm}
396: {\bf Figure~\ref{cut}:} Illustration of a possible resolution of the 
397: singularity at $F_1$. On taking the length of the cut to zero (i.e., 
398: cutting out a very small neighbourhood of $F_1$), the global structure of 
399: the original model of Fig.~\ref{cc} is recovered.
400: \end{figure}
401: 
402: To be more specific about the boundary conditions, consider first a small 
403: interval on the boundary (implying that the curvature can be neglected) at a 
404: point $(x^5_0,x^6_0)$ where the boundary is parallel to the $x^6$ direction. 
405: One can then think of this boundary locally as being defined by the 
406: reflection $x^5-x^5_0\to -(x^5-x^5_0)$ together with an action in field 
407: space $V\to P'VP'^{-1}$, $\Phi_5\to -P'\Phi_5P'^{-1}$, $\Phi_6\to P'\Phi_6
408: P'^{-1}$ and $\Phi_7\to -P'\Phi_7P'^{-1}$. Choosing the boundary in the 
409: shape of a circle on the original torus, SO(2) rotation symmetry acting on 
410: the superfields $(\Phi_5,\Phi_6)$ as well as in coordinate space is now used 
411: to fix the boundary 
412: conditions locally at every point along the cut. In particular, although 
413: different linear combinations of $\Phi_5$ and $\Phi_6$ are forced to vanish
414: as one moves along the boundary, the fields from $\Phi_7$ anticommuting with 
415: $P'$ are always allowed to be non-zero. Thus, in the complete model, the 
416: $\Phi_7$ chiral superfield contributes precisely two potential Higgs 
417: doublets to the low-energy field content. This was realized at the 
418: beginning of the previous section by ad-hoc assumptions about local 
419: symmetry breaking and mass terms localized at $F_1$. Obviously, the breaking 
420: by $P'$ leads to a surviving U(1) in addition to SU(5). However, it does not 
421: affect the phenomenology at the present rough level of discussion and 
422: different options for the scale and mechanism of its breaking may be 
423: considered.
424: 
425: The above discussion of the $Z_2$ boundary condition ignored the 
426: curvature of the boundary in the flat 2d extra-dimensional space and 
427: treated it as a sum of straight elements. To describe the actual
428: curved boundary, one can simply go by diffeomorphism to a coordinate
429: system where the boundary is straight (and a corresponding non-zero Riemann
430: connection, which is considered non-dynamical, appears). Now a finite piece 
431: of the boundary can be defined by $Z_2$ reflection. In addition to the 
432: Riemann connection, an extra non-zero, non-dynamical R-symmetry connection 
433: has to be explicitly introduced in this picture. It accounts for the 
434: appropriate rotation of $A_8$, $A_9$ and the fermions in the $\Phi_5$ and 
435: $\Phi_6$ superfields as one moves along the boundary. 
436: %This connection is 
437: %is implicitly also present in the case of just a conical singularity (since 
438: %R-symmetry twists are, in general, part of the discrete symmetry used 
439: %in the orbifold construction). A dynamical understanding of these degrees 
440: %of freedom can only come from a full 6d (or, in the case of the R symmetry, 
441: %10d) supergravity description. 
442: 
443: 
444: 
445: 
446: 
447: \section{Conclusions}\label{con}
448: In this paper, gauge symmetry breaking by a specific topological feature, 
449: namely a crosscap, of the extra-dimensional manifold was considered. The 
450: main attributes of this simple mechanism are non-locality (implying the 
451: `softness' of the breaking), absence of flat directions (in particular 
452: Wilson line moduli), and the ease with which it can be implemented in a 
453: higher-dimensional SUSY GUT framework. 
454: 
455: If, in addition to the crosscap breaking, the gauge symmetry is broken 
456: by a different mechanism at some point in extra-dimensional space, 
457: Wilson-line moduli (characterizing the relative orientation of the two
458: breakings) reappear. For a breaking pattern leading from SU(6) to SU(5) 
459: and further to the standard model, this implies the existence of two 
460: light Higgs doublets.
461: 
462: In the context of the above SU(6) model, the main open questions concern 
463: the generation of Yukawa couplings and the possible role of two further 
464: potentially light Higgs doublets not associated with Wilson lines. In a
465: wider context, it would be interesting to consider larger groups, to
466: attempt to generate matter and Yukawa couplings from the pure SYM 
467: theory~\cite{bn,hs}, and to use the softness of the breaking to 
468: perform a calculation of GUT threshold corrections. Furthermore, a 
469: systematic understanding of other geometries (possibly in more than 2 
470: extra dimensions) allowing for this type of non-local but quantized gauge 
471: symmetry breaking would be desirable. 
472: 
473: \vspace*{-.3cm}
474: \section*{Acknowledgments}
475: 
476: \vspace*{-.3cm}
477: I would like to thank Wilfried Buchm\"{u}ller and Michael Ratz for useful 
478: discussions and comments. 
479: 
480: \begin{thebibliography}{99}
481: 
482: \bibitem{hos}
483: Y.~Hosotani, Phys.\ Lett.\ B {\bf 126} (1983) 309 and Annals Phys.\ 
484: {\bf 190} (1989) 233.
485: 
486: \bibitem{chsw}
487: P.~Candelas, G.~T.~Horowitz, A.~Strominger and E.~Witten,
488: Nucl.\ Phys.\ B {\bf 258} (1985) 46.
489: %%CITATION = NUPHA,B258,46;%%
490: 
491: \bibitem{wit}
492: E.~Witten, Nucl.\ Phys.\ B {\bf 258} (1985) 75.
493: %%CITATION = NUPHA,B258,75;%%
494: 
495: \bibitem{fm}
496: P.~Forgacs and N.~S.~Manton, Commun.\ Math.\ Phys.\  {\bf 72} (1980) 15;\\
497: %%CITATION = CMPHA,72,15;%%
498: S.~Randjbar-Daemi, A.~Salam and J.~Strathdee, Phys.\ Lett.\ B {\bf 124} 
499: (1983) 345 [Erratum-ibid.\ B {\bf 144} (1984) 455];\\
500: %%CITATION = PHLTA,B124,345;%%
501: C.~Wetterich, Nucl.\ Phys.\ B {\bf 260} (1985) 402;\\
502: %%CITATION = NUPHA,B260,402;%%
503: G.~R.~Dvali, S.~Randjbar-Daemi and R.~Tabbash, Phys.\ Rev.\ D {\bf 65} 
504: (2002) 064021 [arXiv:hep-ph/0102307].
505: %%CITATION = HEP-PH 0102307;%%
506: 
507: \bibitem{kaw}
508: Y.~Kawamura,
509: Prog.\ Theor.\ Phys.\  {\bf 105} (2001) 999
510: [arXiv:hep-ph/0012125];\\
511: %%CITATION = HEP-PH 0012125;%%
512: G.~Altarelli and F.~Feruglio, Phys.\ Lett.\ B {\bf 511} (2001) 257
513: [arXiv:hep-ph/0102301];\\
514: %%CITATION = HEP-PH 0102301;%%
515: L.~J.~Hall and Y.~Nomura, Phys.\ Rev.\ D {\bf 64} (2001) 055003
516: [arXiv:hep-ph/0103125];\\
517: %%CITATION = HEP-PH 0103125;%%
518: A.~Hebecker and J.~March-Russell, Nucl.\ Phys.\ B {\bf 613} (2001) 3\\{}
519: [arXiv:hep-ph/0106166].
520: %%CITATION = HEP-PH 0106166;%%
521: 
522: \bibitem{dhvw}
523: L.~J.~Dixon, J.~A.~Harvey, C.~Vafa and E.~Witten, Nucl.\ Phys.\ B {\bf 261} 
524: (1985) 678 and {\bf 274} (1986) 285.
525: %%CITATION = NUPHA,B274,285;%%
526: %%CITATION = NUPHA,B261,678;%%
527: 
528: \bibitem{abc} 
529: T.~Asaka, W.~Buchm\"uller and L.~Covi, Phys.\ Lett.\ B {\bf 523} (2001) 199
530: \\{} [arXiv:hep-ph/0108021];\\
531: %%CITATION = HEP-PH 0108021;%%
532: L.~J.~Hall, Y.~Nomura, T.~Okui and D.~R.~Smith, Phys.\ Rev.\ D {\bf 65} 
533: (2002) 035008 [arXiv:hep-ph/0108071].
534: %%CITATION = HEP-PH 0108071;%%
535: 
536: \bibitem{gsw}
537: M.~B.~Green, J.~H.~Schwarz and E.~Witten, {\it Superstring theory}, 
538: Cambridge Univ. Press, 1987;\\
539: J.~Polchinski, {\it String theory}, Cambridge Univ. Press, 1998.
540: 
541: \bibitem{ww}
542: X.~G.~Wen and E.~Witten, Nucl.\ Phys.\ B {\bf 261} (1985) 651.
543: %%CITATION = NUPHA,B261,651;%%
544: 
545: \bibitem{hmn}
546: L.~J.~Hall, H.~Murayama and Y.~Nomura, Nucl.\ Phys.\ B {\bf 645} (2002) 85
547: \\{}[arXiv:hep-th/0107245].
548: %%CITATION = HEP-TH 0107245;%%
549: 
550: \bibitem{fao}
551: C.~Vafa and E.~Witten, Nucl.\ Phys.\ Proc.\ Suppl.\  {\bf 46} (1996) 225
552: \\{}[arXiv:hep-th/9507050];\\
553: %%CITATION = HEP-TH 9507050;%%
554: E.~Kiritsis and C.~Kounnas, Nucl.\ Phys.\ B {\bf 503} (1997) 117
555: [arXiv:hep-th/9703059].
556: %%CITATION = HEP-TH 9703059;%%
557: 
558: \bibitem{hns}
559: L.~J.~Hall, Y.~Nomura and D.~R.~Smith, Nucl.\ Phys.\ B {\bf 639} (2002) 307
560: \\{}[arXiv:hep-ph/0107331].
561: %%CITATION = HEP-PH 0107331;%%
562: 
563: \bibitem{pst}
564: F.~Paccetti Correia, M.~G.~Schmidt and Z.~Tavartkiladze, Phys.\ Lett.\ B 
565: {\bf 545} (2002) 153 [arXiv:hep-ph/0206307].
566: %%CITATION = HEP-PH 0206307;%%
567: 
568: \bibitem{bn}
569: G.~Burdman and Y.~Nomura, Nucl.\ Phys.\ B {\bf 656} (2003) 3
570: [arXiv:hep-ph/0210257].
571: %%CITATION = HEP-PH 0210257;%%
572: 
573: \bibitem{mss}
574: N.~Marcus, A.~Sagnotti and W.~Siegel, Nucl.\ Phys.\ B {\bf 224} (1983) 159.
575: %%CITATION = NUPHA,B224,159;%%
576: 
577: \bibitem{abc1}
578: T.~Asaka, W.~Buchm\"uller and L.~Covi, Phys.\ Lett.\ B {\bf 540} (2002) 295
579: \\{}[arXiv:hep-ph/0204358].
580: %%CITATION = HEP-PH 0204358;%%
581: 
582: \bibitem{inq}
583: L.~E.~Ibanez, H.~P.~Nilles and F.~Quevedo, Phys.\ Lett.\ B {\bf 187} (1987) 
584: 25.
585: %%CITATION = PHLTA,B187,25;%%
586: 
587: \bibitem{hr}
588: A.~Hebecker and M.~Ratz, arXiv:hep-ph/0306049, to appear in Nucl.\ Phys.\ B.
589: %%CITATION = HEP-PH 0306049;%%
590: 
591: \bibitem{cgm}
592: C.~Csaki, C.~Grojean and H.~Murayama, Phys.\ Rev.\ D {\bf 67} (2003) 085012
593: \\{}[arXiv:hep-ph/0210133].
594: %%CITATION = HEP-PH 0210133;%%
595: 
596: \bibitem{nsw}
597: Y.~Nomura, D.~R.~Smith and N.~Weiner, Nucl.\ Phys.\ B {\bf 613} (2001) 147
598: \\{}[arXiv:hep-ph/0104041].
599: %%CITATION = HEP-PH 0104041;%%
600: 
601: \bibitem{li}
602: T.~j.~Li, Nucl.\ Phys.\ B {\bf 633} (2002) 83 [arXiv:hep-th/0112255].
603: %%CITATION = HEP-TH 0112255;%%
604: 
605: \bibitem{hs}
606: N.~Haba and Y.~Shimizu, Phys.\ Rev.\ D {\bf 67} (2003) 095001
607: [arXiv:hep-ph/0212166].
608: %%CITATION = HEP-PH 0212166;%%
609: I.~Gogoladze, Y.~Mimura and S.~Nandi, Phys.\ Lett.\ B {\bf 562} (2003) 307
610: \\{}[arXiv:hep-ph/0302176].
611: %%CITATION = HEP-PH 0302176;%%
612: 
613: \end{thebibliography}
614: \end{document}
615: 
616: