1: \documentclass[preprint,nofootinbib,preprintnumbers]{revtex4}
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: \usepackage{amssymb}
4: \usepackage{graphicx}
5:
6: \newcommand{\xn}[0]{\langle x^n\rangle\ }
7: \newcommand{\xm}[0]{\langle x^m\rangle\ }
8: \newcommand{\xnA}[0]{\langle x^n\rangle_A}
9: \newcommand{\xmA}[0]{\langle x^m\rangle_A}
10: \def\dfrac#1#2{{\displaystyle {#1 \over #2}}}
11: \preprint{\hspace{13.45cm} NT@UW-04-029}
12:
13:
14: \begin{document}
15:
16: \title{Universality of the EMC Effect}
17: \author{Jiunn-Wei Chen}
18: \email{jwc@phys.ntu.edu.tw}
19: \affiliation{Department of Physics and National Center for Theoretical Sciences at
20: Taipei, National Taiwan University, Taipei 10617, Taiwan, \\
21: and Lawrence-Berkeley Laboratory, Berkeley, CA 94720, U.S.A.}
22: \author{William Detmold}
23: \email{wdetmold@phys.washington.edu}
24: \affiliation{Department of Physics, University of Washington, Seattle, WA 98195-1560,
25: U.S.A.}
26: \date{\today }
27:
28: \begin{abstract}
29: Using effective field theory, we investigate nuclear modification of nucleon
30: parton distributions (for example, the EMC effect). We show that the
31: universality of the shape distortion in nuclear parton distributions (the
32: factorisation of the Bjorken $x$ and atomic number ($A$) dependence) is
33: model independent and emerges naturally in effective field theory. We then
34: extend our analysis to study the analogous nuclear modifications in isospin
35: and spin dependent parton distributions and generalised parton distributions.
36: \end{abstract}
37:
38: \maketitle
39:
40: In 1983, the European Muon Collaboration (EMC) reported measurements \cite%
41: {Aubert:1983xm} of the ratio $R_{Fe}(x)$ of $F_{2}(x)$ structure functions
42: in iron and deuterium in deep inelastic scattering (DIS). In the parton
43: model, these structure functions are defined as $F_{2}^{A}(x)=%
44: \sum_{i}Q_{i}^{2}x\left[ q_{i}^{A}(x)+\overline{q}_{i}^{A}(x)\right] $,
45: where $q_{i}^{A}(x)$[$\overline{q}_{i}^{A}(x)$] is the parton distribution
46: function (PDF) for quarks[anti-quarks] of flavor $i$ in a nucleus, $A$, and $%
47: Q_{i}$ is the electric charge of $q_{i}$. Bjorken $x$ is the longitudinal
48: momentum fraction of the struck parton in the infinite momentum frame with
49: respect to its parent nucleon. $R_{Fe}(x)$ is normalised to unity if the
50: nucleons in the nuclei are non-interacting. The main observation of EMC was
51: that $R_{Fe}(x)$ deviated from unity by up to 20\% over the range $%
52: 0.05<x<0.65$ in which measurements were taken. The EMC result was unlooked
53: for and came as a surprise to many physicists at the time partly because the
54: typical binding energy per nucleon is so much smaller ($<$1\%) than the
55: nucleon mass and the energy transfer involved in a DIS process.
56:
57: Over the past two decades, further experiments have been performed by many
58: groups (see Refs.~\cite%
59: {Arneodo:1992wf,Geesaman:1995yd,Piller:1999wx,Norton:2003cb} for recent
60: reviews) aiming to better understand the details of nuclear modifications of
61: hadron structure functions (here referred to generically as the EMC effect).
62: The EMC result has been confirmed and demonstrated in many other nuclei
63: ranging from helium to lead. Nuclear modification of structure functions has
64: also been studied in other situations such as proton-nucleus Drell-Yan
65: experiments and quarkonium production \cite{Peng:1999tm,Norton:2003cb}.
66: Overall, a very interesting picture has emerged; for an isoscalar nucleus of
67: atomic number $A$, the \emph{shape} of the deviation from unity of $%
68: R_{A}(x)=F_{2}^{A}(x)/AF_{2}^{N}(x)$ (here $F_{2}^{d}$ has been converted to
69: the isoscalar $F_{2}^{N}$ subject to a small model dependent error, and the
70: slight $Q^{2}$ dependence of $R_{A}(x)$ \cite{Norton:2003cb} is suppressed)
71: is universal \cite{Date:1984ve,Frankfurt:1988nt}, namely independent of $A$
72: within experimental error bars, while the \emph{magnitude} of the distortion
73: is empirically proportional to the number density of the nucleus, $\rho _{A}$
74: \cite{Gomez:1993ri}. Fits to the available data that support these features
75: will be presented elsewhere \cite{futurework}. These findings have inspired
76: many theoretical analyses seeking to understand the details of the EMC
77: effect in various approaches (see the reviews for summaries). Different
78: physical processes have been identified as the causes of the modifications
79: in different $x$ regions; \textit{e.g.}, nuclear shadowing at low $x$ and
80: Fermi motion at large $x$. A recent conclusion of detailed model studies was
81: that the EMC effect necessarily implies modification of the nucleon PDFs and
82: cannot be explained through traditional nuclear physics \cite{Smith:2003hu}.
83: However, little is known directly from QCD.
84:
85: In this Letter, we employ effective field theory (EFT) to investigate the
86: EMC effect by studying nuclear matrix elements of the twist-two operators
87: which are related to parton distributions and structure functions via the
88: operator product expansion. We find that the universality of the shape
89: distortion of the EMC effect is a model independent result, arising from the
90: symmetries of QCD and the separation of the relevant scales. The $x$
91: dependence of $R_{A}(x)$ is governed by short distance physics, while the
92: overall magnitude (the $A$ dependence) of the EMC effect is governed by long
93: distance matrix elements calculable using traditional nuclear physics. We
94: then proceed to study analogous nuclear effects in isospin-odd, and
95: spin-dependent parton distributions and generalised parton distributions
96: (GPDs) \cite{GPD}. We also discuss aspects of extracting the shape of the
97: EMC effect from first principles using lattice QCD. This approach provides a
98: clear connection between the EMC effect and many other observations of
99: nuclear modification of hadron properties.
100:
101: EFT is a model independent approach which only makes use of the symmetries
102: and scale separation of the system (for recent reviews see Ref.~\cite%
103: {EFTreviews}). This approach has been successfully applied to many low
104: energy processes in $A=1,2,3,4$ systems. Recently EFT has been applied
105: to the computation of hadronic matrix elements of twist-two operators
106: in the meson and single nucleon sectors \cite{AS,CJ,Detmold:2005pt}
107: and applied to chiral extrapolations of lattice calculations of
108: moments of parton distributions \cite{DMNRT}. The approach has also
109: been extended to analyse moments of generalised parton
110: distributions~\cite{Jq}, large $N_{C}$ relations among PDFs in
111: nucleons and the $\Delta $-isobar \cite{CJ}, and deeply virtual
112: Compton scattering in the nucleon \cite{DVCSpi} and deuteron systems
113: \cite{BS}. The method is readily generalised to the multi-nucleon
114: case. Although we will concentrate on quark bilinear twist-two
115: operators, the framework can be easily applied to gluonic operators
116: and thereby to nuclear effects in gluonic distributions which are
117: important in heavy ion collisions at RHIC and LHC and at a future
118: Electron-Ion Collider.
119:
120: To described the EMC effect observed in $F_2$ data on isoscalar nuclei, we
121: consider the normalised, spin singlet, isoscalar twist-two operators,
122: \begin{equation}
123: \mathcal{O}_{q}^{\mu _{0}\cdots \mu _{n}}=\overline{q}\gamma ^{(\mu
124: _{0}}iD^{\mu _{1}}\cdots iD^{\mu _{n})}q/\left( 2M^{n+1}\right) , \label{O}
125: \end{equation}
126: where $(...)$ indicates that enclosed indices have been symmetrised
127: and made traceless, $D^{\mu }=(\overrightarrow{D}^{\mu }-
128: \overleftarrow{D}^{\mu })/2$ is the covariant derivative and $M$ is
129: the nucleon \ mass. The matrix elements of $\mathcal{O}_{q}^{\mu
130: _{0}\ldots \mu _{n}}$ in an unpolarised single nucleon state with
131: momentum $P$ can be parametrised as
132: \begin{equation}
133: \langle P|\mathcal{O}_{q}^{\mu _{0}\ldots \mu _{n}}|P\rangle =\langle
134: x^{n}\rangle _{q}\widetilde{v}^{\mu _{0}}\cdots \widetilde{v}^{\mu _{n}},
135: \end{equation}
136: where the nucleon velocity $\widetilde{v}^{\mu }=P^{\mu }/M$. It is well
137: known that the coefficients $\langle x^{n}\rangle _{q}$ correspond to
138: moments of the isoscalar combination of parton distribution functions,
139: \begin{equation}
140: \left\langle x^{n}\right\rangle _{q}=\int_{-1}^{1}dx\,x^{n}q(x),
141: \end{equation}
142: where $q(x)$ is the isoscalar quark distribution and $q(-x)=-\overline{q}(x)$.
143:
144: We first consider only nucleonic degrees of freedom ({\it i.e.},
145: assume that pions are integrated out of the EFT -- they will be
146: reintroduced below) and perform the standard matching procedure in
147: EFT, equating the quark level twist-two operators to the most general
148: combinations of hadronic operators of the same symmetries
149: \cite{AS,CJ,Detmold:2005pt,DMNRT}. The leading one- and two-body
150: hadronic operators in the matching are
151: \begin{equation}
152: \mathcal{O}_{q}^{\mu _{0}\ldots \mu _{n}}=\langle x^{n}\rangle _{q}v^{\mu
153: _{0}}\cdots v^{\mu _{n}}N^{\dagger }N\left[ 1+\alpha _{n}N^{\dagger }N\right]
154: +\cdots \,, \label{eq:1}
155: \end{equation}%
156: where $v^{\mu }=\widetilde{v}^{\mu }+\mathcal{O}(1/M)$ is the velocity
157: of the nucleus. Operators involving additional derivatives are
158: suppressed by powers of $M$ in the EFT power-counting. In
159: Eq.~(\ref{eq:1}) we have only kept the SU(4) (spin and isospin)
160: singlet two-body operator $\alpha _{n}v^{\mu _{0}}\cdots v^{\mu
161: _{n}}\left( N^{\dagger }N\right) ^{2}$ in the above equation. The
162: other independent two-body operator $\beta _{n}v^{\mu _{0}}\cdots
163: v^{\mu _{n}}\left( N^{\dagger }\mbox{\boldmath$\tau$}N\right) ^{2}$,
164: which is non-singlet in SU(4) [$\mbox{\boldmath$\tau$}$ is an isospin
165: matrix], is
166: neglected because $\beta _{n}/\alpha _{n}=O(1/N_{c}^{2})\simeq 0.1$ \cite{KS}%
167: , where $N_{c}$ is the number of colors. Furthermore, the matrix element of $%
168: \left( N^{\dagger }\mbox{\boldmath$\tau$}N\right) ^{2}$ for an
169: isoscalar state with atomic number $A$ is smaller than that of $\left(
170: N^{\dagger }N\right) ^{2}$ by a factor $A$ \cite{futurework}.
171: Three- and higher- body operators also appear in Eq.~(\ref{eq:1});
172: numerical evidence from other EFT calculations indicates that these
173: contributions are generally much smaller than two-body ones
174: \cite{Kubodera}.
175:
176: \begin{figure}[t]
177: \begin{center}\includegraphics[width=0.3\columnwidth]{Fig1a.eps} \hspace*{5mm}
178: \includegraphics[width=0.3\columnwidth]{Fig1b.eps}\\
179: (a)\hspace*{5.5cm}(b) \\
180: \centering \includegraphics[width=0.3\columnwidth]{Fig2a.eps} \hspace*{2mm}
181: \includegraphics[width=0.3\columnwidth]{Fig2d.eps} \hspace*{2mm}
182: \includegraphics[width=0.3\columnwidth]{Fig2c.eps}\\
183: (c)\hspace*{5.3cm}(d) \hspace*{5.3cm} (e) \vspace*{-2mm}
184: \end{center}
185: \caption{Contributions to nuclear matrix elements. The dark square
186: represents the various operators in Eq.~(\protect\ref{eq:1}) and the light
187: shaded ellipse corresponds to the nucleus, $A$. The dots in the lower part
188: of the diagram indicate the spectator nucleons.}
189: \label{fig:nucleon}
190: \end{figure}
191: Nuclear matrix elements of $\mathcal{O}_{q}^{\mu _{0}\ldots \mu _{n}}$ give
192: the moments of the isoscalar nuclear parton distributions, $q_{A}(x)$. The
193: leading order (LO) and the next-to-leading order (NLO) contributions to
194: these matrix elements are shown in Fig. 1(a) and 1(b), respectively. For an
195: unpolarized, isoscalar nucleus,
196: \begin{eqnarray}
197: \langle x^{n}\rangle _{q|A} &\equiv &v^{\mu _{0}}\cdots v^{\mu _{n}}\langle
198: A|\mathcal{O}_{q}^{\mu _{0}\ldots \mu _{n}}|A\rangle \label{eq:2} \\
199: &=&\langle x^{n}\rangle _{q}\left[ A+\langle A|\alpha _{n}(N^{\dagger
200: }N)^{2}|A\rangle \right] , \nonumber
201: \end{eqnarray}%
202: where we have used $\langle A|N^{\dagger }N|A\rangle =A$. Notice that if
203: there were no EMC effect, the $\alpha _{n}$ would vanish for all $n$. Also $%
204: \alpha _{0}=0$ because of charge conservation. Asymptotic relations \cite%
205: {West:1985kg} and analysis of experimental data \cite%
206: {Arneodo:1992wf,Rinat:2005qk} suggests that $\alpha _{1}\simeq 0$, implying
207: that quarks carry very similar fractions of a nucleon's and a nucleus's
208: momentum though no symmetry guarantees this. From Eq.~(\ref{eq:2}) we see
209: that the ratio
210: \begin{equation}
211: \frac{{\displaystyle{\frac{\langle x^{n}\rangle _{q|A}}{A\langle
212: x^{n}\rangle _{q}}}}-1}{{\displaystyle{\frac{\langle x^{m}\rangle _{q|A}}{%
213: A\langle x^{m}\rangle _{q}}}}-1}=\frac{\alpha _{n}}{\alpha _{m}}
214: \label{eq:3}
215: \end{equation}%
216: is independent of $A$ which has powerful consequences. In all generality,
217: the isoscalar nuclear quark distribution can be written as
218: \begin{equation}
219: q_{A}(x)=A\left[ q(x)+\widetilde{g}(x,A)\right] . \label{eq:4}
220: \end{equation}%
221: Taking moments of Eq.~(\ref{eq:4}), Eq.~(\ref{eq:3}) then demands that the $%
222: x $ dependence and $A$ dependence of $\widetilde{g}$ factorise,
223: \begin{equation}
224: \widetilde{g}(x,A)=g(x)\mathcal{G}(A),
225: \end{equation}%
226: with
227: \begin{equation}
228: \mathcal{G}(A)=\langle A|(N^{\dagger }N)^{2}|A\rangle /A\Lambda _{0}^{3},
229: \label{G}
230: \end{equation}%
231: and $g(x)$ satisfying
232: \begin{equation}
233: \alpha _{n}=\frac{1}{\Lambda _{0}^{3}\langle x^{n}\rangle _{q}}%
234: \int_{-A}^{A}dx\,x^{n}g(x)\,. \label{an}
235: \end{equation}%
236: $\Lambda _{0}$ is an arbitrary dimensionful parameter and will be chosen as $%
237: \Lambda _{0}=1$~fm$^{-1}$. Crossing symmetry dictates that the even and odd $%
238: \alpha _{n}$ separately determine the nuclear modifications of valence and
239: total quark distributions. These results apply to any isoscalar combination
240: of parton distributions including $F_{2}(x)$ for isoscalar nuclei. Thus our
241: result implies that
242: \begin{equation}
243: R_{A}(x)=\frac{F_{2}^{A}(x)}{AF_{2}^{N}(x)}=1+g_{F_{2}}(x)\mathcal{G}(A),
244: \label{Ra}
245: \end{equation}%
246: which says that the EMC effect (the deviation of $R_{A}(x)$ from unity) has
247: an universal shape described by $g_{F_{2}}(x)$ while the magnitude of the
248: deviation, $\mathcal{G}(A)$, only depends on $A$.
249:
250: The above analysis gives a simple explanation of the observed universal
251: shape of the EMC effect, or equivalently, the factorisation of $\widetilde{g}%
252: (x,A)$. The key to establishing this factorisation is that other sources of
253: nuclear modification contributing to the right-hand side of Eq.~(\ref{eq:1})
254: must be suppressed (higher order in the EFT) such that the $A$ independence
255: of Eq.~(\ref{eq:3}) can be established. We stress that the factorisation
256: persists when pions are included in our analysis. In Fig.~\ref{fig:nucleon},
257: examples of the leading pionic contributions are shown. The various
258: single-nucleon diagrams, such as Fig.~\ref{fig:nucleon}(c), simply renormalise
259: the nucleon moments, $\langle x^{n}\rangle _{q}$, without contributing to
260: the EMC effect. Two- and more- nucleon diagrams such as those in Fig.~\ref%
261: {fig:nucleon}(d) and \ref{fig:nucleon}(e) contribute to the EMC
262: effect, but only at N$^{3}$LO and higher (see Ref.~\cite{futurework}
263: for explicit calculations) since the pions must be radiation pions
264: rather than potential pions \cite{KSW98} (we are free to choose the
265: twist-two indices to be $\mu _{i}=0$ for all $i$). Other contributions
266: that could upset the factorization include a two-body operator which
267: is similar to that in Eq.~(\ref{eq:1}) but
268: with two more derivatives. However this operator also contributes at N$^{3}$%
269: LO. Consequently, the universality of the EMC effect is preserved to good
270: accuracy. For large $x$ it is clear that the
271: factorisation must break down (simply consider the region $x>2$ in which
272: only three- and higher- body operators contribute) though the structure
273: function is very small in this region anyway. We stress that the
274: factorisation is a model independent result and is just a consequence of
275: scale separation and the SU(4) spin-isospin symmetry from large $N_{c}$.
276:
277: It is clear from Eq.~(\ref{G}) that $\mathcal{G}(A)$ is governed by long
278: distance physics which can be computed using a traditional, non-relativistic
279: nuclear physics approach. It is interesting to note that the mean field
280: scaling of $\mathcal{G}(A)\sim \rho _{A}\sim \log (A)$ describes the
281: empirical $A$ dependence of $R_{A}(x)$ well \cite{Gomez:1993ri,futurework},
282: even though the mean field approximation is not justified for nuclei where
283: the two-particle S-wave scattering lengths are much larger than the mean
284: distance between nucleons.
285:
286: Information on the shape distortion function $g(x)$ is encoded in the short
287: distance parameters $\alpha _{n}$ associated with the strength of the
288: two-body currents. One can either fix the $\alpha _{n}$ from experimental
289: data (to determine all $\alpha _{n}$, data on $F_{2}^{A}(x)$ and $%
290: F_{3}^{A}(x)$ are required) or calculate $\langle NN|\mathcal{O}_{q}^{\mu
291: _{0}\ldots \mu _{n}}|NN\rangle $ in two nucleon systems to extract them. The
292: latter approach, however, is intrinsically non-perturbative and thus
293: requires lattice QCD. Recent analysis \cite{Detmold:2004kw} suggest that one
294: can use background fields coupling to twist-two operators to extract
295: information on the $\alpha _{n}$ from the finite volume scaling of two
296: particle energy levels \cite%
297: {Luscher:1986pf,Beane:2003da,Detmold:2004qn,futurework}. At present, only
298: the first few coefficients could be extracted on the lattice because of
299: problems with operator mixing \cite{Gockeler:1996mu}. However, even a
300: calculation of these would be significant since it would be addressing
301: nuclear modification of hadron structure from first principles. Lattice
302: calculations can also be used to investigate the large $N_{c}$ prediction of
303: $\beta_{n}/\alpha_{n} = O(1/N_{c}^{2})$ discussed below Eq.~(\ref{eq:1}).
304:
305: Given the success of the EFT approach in explaining aspects of the
306: isoscalar, helicity averaged EMC effect, we shall now proceed to study the
307: isospin and spin dependent cases. For the isovector operators
308: \begin{equation}
309: \mathcal{O}_{q,3}^{\mu _{0}\cdots \mu _{n}}=\overline{q}\gamma ^{(\mu
310: _{0}}iD^{\mu _{1}}\cdots iD^{\mu _{n})}\mbox{\boldmath$\tau$}_{3}q/\left(
311: 2M^{n+1}\right) ,
312: \end{equation}%
313: EFT operator matching leads to%
314: \begin{equation}
315: \mathcal{O}_{q,3}^{\mu _{0}\ldots \mu _{n}}=\langle x^{n}\rangle
316: _{q,3}v^{\mu _{0}}\cdots v^{\mu _{n}}N^{\dagger }\mbox{\boldmath$\tau$}_{3}N%
317: \left[ 1+\alpha _{3,n}N^{\dagger }N\right] +\cdots \,,
318: \label{eq:isovec_operators}
319: \end{equation}%
320: where the ellipsis includes the higher order nucleonic and pionic operators.
321: Analysis similar to the isoscalar case implies
322: \begin{equation}
323: R_{A}^{(3)}(x)\equiv \frac{q_{3|A}(x)}{(Z-N)q_{3}(x)}=1+g_{3}(x)\mathcal{\ G}%
324: _{3}(A)\,,
325: \end{equation}%
326: where $q_{3}(x)=u(x)-d(x)$ and $Z$($N$) is the proton (neutron) number. To
327: test this factorisation, one can either consider the difference between $%
328: F_{2}$'s in $(Z,N)=(n+m,n)$ and $(n,n+m)$ mirror nuclei \cite{Saito:2000fx}
329: and compare it with $F_{2}^{p}-F_{2}^{n}$, or disentangle $u_{A}(x)$ and $%
330: d_{A}(x)$ with the proposed neutrino-nucleus experiment, MINER$\nu $A \cite%
331: {Minerva}.
332:
333: To generalise our analysis to spin dependent operators is also
334: straightforward. The operators related to moments of quark helicity
335: and transversity distributions are
336: \begin{eqnarray}
337: \mathcal{O}_{\Delta q,\alpha }^{\mu _{0}\cdots \mu _{n}} &=&\overline{q}%
338: \gamma ^{(\mu _{0}}\gamma ^{5}iD^{\mu _{1}}\cdots iD^{\mu _{n})}\tau
339: _{\alpha }q/\left( 2M^{n+1}\right) , \label{moreOs} \\
340: \mathcal{O}_{\delta q,\alpha }^{\rho \mu _{0}\cdots \mu _{n}} &=&\overline{q}%
341: \sigma ^{\rho (\mu _{0}}\gamma ^{5}iD^{\mu _{1}}\cdots iD^{\mu _{n})}\tau
342: _{\alpha }q/\left( 2M^{n+2}\right) \,.\;\;\;\;\; \nonumber
343: \end{eqnarray}%
344: In the isoscalar case ($\tau _{0}=\mbox{\boldmath$1$}$), the matching yields
345: \begin{equation}
346: \mathcal{O}_{\Delta q,0}^{\mu _{0}\cdots \mu _{n}}=2\langle x^{n}\rangle
347: _{\Delta q}N^{\dagger }S^{(\mu _{0}}v^{\mu _{1}}\cdots v^{\mu _{n})}N\left[
348: 1+\gamma _{n,0}N^{\dagger }N\right] +\cdots , \nonumber
349: \end{equation}%
350: and similarly for the other operators (the spin operator $S^{\mu }=(0,%
351: \mbox{\boldmath$\sigma$}/2)$ where $\mbox{\boldmath$\sigma$}$ are Pauli spin
352: matrices). Again, we have similar factorisation; \textit{e.g.},
353: \begin{equation}
354: R_{A}^{\Delta }(x)\equiv \frac{\Delta q^{A}(x)}{\Delta N\,\Delta q(x)}%
355: =1+g_{\Delta }(x)\mathcal{G}_{\Delta }(A),
356: \end{equation}%
357: where $\Delta N$ is the differences between the number of nucleons with
358: positive and negative spin projections in the longitudinal direction.
359: Similarly, $\mathcal{G}_{\Delta }(A)=\langle A|N^{\dagger }NN^{\dagger }%
360: \mbox{\boldmath$\sigma$}_{3}N|A\rangle /\Delta N\Lambda _{0}^{3}$. Whilst
361: there is data on longitudinal asymmetries in light nuclei from which the $%
362: g_{1}^{A}(x)$ structure functions can be extracted, disentangling
363: nuclear effects in the unpolarised and polarised structure functions
364: will be difficult. Currently, polarised heavy nuclei targets, in which
365: modifications would be larger, are not available. Recent model
366: calculations \cite{polarisedmodels} find nuclear effects in the
367: polarised structure function $g_1^A(x)$ to be significant. Nothing is
368: known experimentally about the transversity structure function even in
369: the proton but analogous nuclear modifications can be derived.
370:
371: It is also possible to study nuclear effects in GPDs by computing
372: off-forward matrix elements of twist-two operators \cite{futurework}.
373: In the quark contribution to nuclear spin, $J_{qA}$, for example, in
374: addition to extending the operators matched to
375: $\mathcal{O}_{q}^{\mu_{0}\mu_{1}}$ in Eq.~(\ref{eq:1}) by replacing
376: $v\to v+i\frac{D}{M}$ (reparameterisation invariance
377: \cite{Luke:1992cs} constrains this form), we also need to consider the
378: term
379: \begin{equation}
380: -2\frac{J_{qN}}{M}i D^{\beta }\left\{ \overline{N}\left[ S^{\mu
381: _{0}},S_{\beta }\right] \left(v+i\frac{D}{M}\right)^{\mu _{1}}N\left[
382: 1+\eta N^{\dagger }N\right]
383: \right\} \,, \label{Oq}
384: \end{equation}
385: where $J_{q(g)N}$ is the quark(gluon) angular momentum content of the
386: nucleon, to obtain
387: \begin{equation}
388: J_{qA}=\langle x\rangle _{qN}L_{z}+2J_{qN}S_{z}\left[ 1+\eta \,\mathcal{H}%
389: (A) \right] \,.
390: \end{equation}
391: Remarkably, explicit calculation \cite{futurework} shows that the $L_{z}$
392: term is free from two-body current corrections even though $\alpha _{1}\neq
393: 0 $ in Eq.~(\ref{eq:1}). Similarly, the gluon contribution to nuclear spin
394: satisfies
395: \begin{equation}
396: J_{gA}=\langle x\rangle _{gN}L_{z}+2J_{gN}S_{z}\left[ 1-\eta \,\mathcal{H}%
397: (A) \right] \,,
398: \end{equation}
399: where the same constant $\eta $ appears by total angular momentum
400: conservation. Consequently, using the sum rules $\langle x\rangle
401: _{qN}+\langle x\rangle _{gN}=1$ and $J_{qN}+J_{gN}=1/2$, we recover
402: \begin{equation}
403: J_{qA}+J_{gA}=L_{z}+S_{z}\,.
404: \end{equation}
405: For details, see \cite{futurework}.
406:
407: To summarise, we have studied the EMC effect (nuclear modification of parton
408: distributions) in EFT and seen that the scale separation of short and long
409: distance effects provides a model independent derivation of the
410: factorisation of the $x$ and $A$ dependencies of $R_{A}(x)$, relying only on
411: the symmetries of QCD. Similar factorisations are predicted to occur in
412: other probes of nuclear structure such as spin-dependent structure functions
413: and GPDs.
414:
415: \textit{Acknowledgements:} We thank P. Bedaque, E.~M.~Henley, D. Lee, W.
416: Melnitchouk, G.~A.~Miller, A.~S.~Rinat, M. J. Savage, A.~W.~Thomas and
417: W.-K.~Tung for useful comments and discussions. This work was supported by
418: the US Department of Energy under contract DE-FG03-97ER41014 and the
419: National Science Council of ROC.
420:
421: \begin{thebibliography}{99}
422: \bibitem{Aubert:1983xm} J.~Aubert \textit{et al.},
423: %``The Ratio Of The Nucleon Structure Functions F2 (N) For Iron And Deuterium,''
424: Phys.\ Lett.\ B \textbf{123}, 275 (1983). %%CITATION = PHLTA,B123,275;%%
425:
426: \bibitem{Arneodo:1992wf} M.~Arneodo,
427: %``Nuclear effects in structure functions,''
428: Phys.\ Rept.\ \textbf{240}, 301 (1994). %%CITATION = PRPLC,240,301;%%
429:
430: \bibitem{Geesaman:1995yd} D.~F.~Geesaman, K.~Saito and A.~W.~Thomas,
431: %``The nuclear EMC effect,''
432: Ann.\ Rev.\ Nucl.\ Part.\ Sci.\ \textbf{45}, 337 (1995).
433: %%CITATION = ARNUA,45,337;%%
434:
435: \bibitem{Piller:1999wx} G.~Piller and W.~Weise,
436: %``Nuclear deep-inelastic lepton scattering and coherence phenomena,''
437: Phys.\ Rept.\ \textbf{330}, 1 (2000). %%CITATION = HEP-PH 9908230;%%
438:
439: \bibitem{Norton:2003cb} P.~R.~Norton, %``The EMC effect,''
440: Rept.\ Prog.\ Phys.\ \textbf{66}, 1253 (2003).
441: %%CITATION = RPPHA,66,1253;%%
442:
443: \bibitem{Peng:1999tm} J.~C.~Peng \textit{et al.},
444: %P.~L.~McGaughey and J.~M.~Moss,
445: %``Dilepton production at Fermilab and RHIC,''
446: hep-ph/9905447. %%CITATION = HEP-PH 9905447;%%
447:
448: \bibitem{Date:1984ve} S.~Date \textit{et al.},
449: %K.~Saito, H.~Sumiyoshi and H.~Tezuka,
450: %``New Scaling Phenomena In Nuclear Structure Functions,''
451: Phys.\ Rev.\ Lett.\ \textbf{52}, 2344 (1984). %%CITATION = PRLTA,52,2344;%%
452:
453: \bibitem{Frankfurt:1988nt} L.~L.~Frankfurt and M.~I.~Strikman,
454: %``Hard Nuclear Processes And Microscopic Nuclear Structure,''
455: Phys.\ Rept.\ \textbf{160}, 235 (1988). %%CITATION = PRPLC,160,235;%%
456:
457: \bibitem{Gomez:1993ri} J.~Gomez \textit{et al.},
458: %``Measurement of the A-dependence of deep inelastic electron scattering,''
459: Phys.\ Rev.\ D \textbf{49}, 4348 (1994). %%CITATION = PHRVA,D49,4348;%%
460:
461: \bibitem{futurework} W.~Detmold and J. W.~Chen, \textit{in preparation}.
462:
463: \bibitem{Smith:2003hu} J.~R.~Smith and G.~A.~Miller,
464: %``Chiral solitons in nuclei: Saturation, EMC effect and Drell-Yan
465: %experiments,''
466: Phys.\ Rev.\ Lett.\ \textbf{91}, 212301 (2003).
467: %%CITATION = NUCL-TH 0308048;%%
468:
469: \bibitem{GPD} See, e.g., X.~Ji, %``Off-forward parton distributions,''
470: J.\ Phys.\ G \textbf{24}, 1181 (1998); %%CITATION = HEP-PH 9807358;%%
471: M.~Diehl, %``Generalized parton distributions,''
472: Phys.\ Rept.\ \textbf{388}, 41 (2003). %%CITATION = HEP-PH 0307382;%%
473:
474: \bibitem{EFTreviews} S.~R.~Beane \emph{et al.}, in \textit{Encyclopedia of
475: Analytic QCD}, edited by M. Shifman, World Scientific; Nucl. Phys. A \textbf{%
476: 700}, 377 (2002); P. F. Bedaque and U. van Kolck, Ann. Rev. Nucl. Part. Sci.
477: \textbf{\ 52}, 339 (2002); K. Kubodera and T.-S. Park, Ann. Rev. Nuc. Part.
478: Sci. \textbf{54}, 19 (2004); U.~G.~Meissner, arXiv:nucl-th/0409028.
479:
480: \bibitem{AS} D.~Arndt and M.~J.~Savage,
481: %``Chiral Corrections to Matrix Elements of twist-2 Operators,''
482: Nucl. Phys. \textbf{A697}, 429 (2002).
483:
484: \bibitem{CJ} J.~W.~Chen and X.~Ji,
485: %``Is the Sullivan process compatible with QCD chiral dynamics?,''
486: Phys.\ Lett.\ \textbf{B523}, 107 (2001); %[arXiv:hep-ph/0105197].
487: %%CITATION = HEP-PH 0105197;%%
488: %
489: %J.~W.~Chen and X.~d.~Ji,
490: %``Large-N(c) quark distributions in the Delta and chiral logarithms in
491: % quark distributions of the nucleon,''
492: \textit{ibid}, 73 (2001); %Phys.\ Lett.\ B {\bf 523}, 73 (2001)
493: %[arXiv:hep-ph/0105296].
494: %%CITATION = HEP-PH 0105296;%%
495: %
496: %J.~W.~Chen and X.~d.~Ji,
497: %``Constructing parton convolution in effective field theory,''
498: Phys.\ Rev.\ Lett.\ \textbf{87}, 152002 (2001).
499:
500: \bibitem{Detmold:2005pt} W.~Detmold and C.-J.~D.~Lin,
501: %``Twist-two matrix elements at finite and infinite volume,''
502: Phys.\ Rev.\ D \textbf{71}, 054510 (2005). % [arXiv:hep-lat/0501007].
503: %%CITATION = HEP-LAT 0501007;%%
504:
505: \bibitem{DMNRT} W.~Detmold \textit{et al.},
506: %``Chiral extrapolation of lattice moments of proton quark distributions,''
507: Phys.\ Rev.\ Lett.\ \textbf{87}, 172001 (2001); %[arXiv:hep-lat/0103006].
508: %%CITATION = HEP-LAT 0103006;%%
509: %``Moments of isovector quark distributions from lattice QCD,''
510: Phys.\ Rev.\ D \textbf{66}, 054501 (2002); %[arXiv:hep-lat/0206001].
511: %%CITATION = HEP-LAT 0206001;%%
512: %W.~Detmold, W.~Melnitchouk and A.~W.~Thomas,
513: %``Parton distribution functions in the pion from lattice QCD,''
514: \textit{ibid} \textbf{68}, 034025 (2003). %[arXiv:hep-lat/0303015].
515: %%CITATION = HEP-LAT 0303015;%%
516:
517: \bibitem{Jq} J. W.~Chen and X.~Ji,
518: %``Leading chiral contributions to the spin structure of the proton,''
519: Phys.\ Rev.\ Lett.\ \textbf{88}, 052003 (2002); % [arXiv:hep-ph/0111048].
520: %%CITATION = HEP-PH 0111048;%%
521: A. V.~Belitsky and X.~Ji,
522: %``Chiral structure of nucleon gravitational form factors,''
523: Phys.\ Lett.\ B \textbf{538}, 289 (2002). % [arXiv:hep-ph/0203276].
524: %%CITATION = HEP-PH 0203276;%%
525:
526: \bibitem{DVCSpi} J.~W.~Chen and M.~J.~Savage,
527: %``Soft pion emission in DVCS,''
528: Nucl.\ Phys.\ A \textbf{735}, 441 (2004).
529:
530: \bibitem{BS} S. R. Beane and M.~J.~Savage, arXiv:nucl-th/0412025.
531: %%CITATION = NUCL-TH 0412025;%%
532:
533: \bibitem{KS} D. B. Kaplan and M. J. Savage, Phys. Lett. B \textbf{365}, 244
534: (1996).
535:
536: \bibitem{Kubodera} K. Kubodera and T.-S. Park, Ann. Rev. Nuc. Part. Sci.
537: \textbf{54}, 19 (2004);
538:
539: \bibitem{West:1985kg} G.~B.~West,
540: %``Energy Momentum Sum Rule And The EMC Effect,''
541: Phys.\ Rev.\ Lett.\ \textbf{54}, 2576 (1985).
542: %%CITATION = PRLTA,54,2576;%%
543:
544: \bibitem{Rinat:2005qk} A.~S.~Rinat and M.~F.~Taragin,
545: %``On Distribution Functions for Partons in Nuclei,''
546: arXiv:nucl-th/0501006. %%CITATION = NUCL-TH 0501006;%%
547:
548: \bibitem{KSW98} D. B. Kaplan, M. J. Savage and M. B. Wise, Phys. Lett. B%
549: \textbf{424}, 390 (1998); Nucl.\ Phys.\ B \textbf{534}, 329 (1998).
550: %%CITATION = NUCL-TH 9802075;%%
551:
552: \bibitem{Detmold:2004kw} W.~Detmold,
553: %``Flavour singlet physics in lattice QCD with background fields,''
554: Phys.\ Rev.\ D \textbf{71}, 054506 (2005). % [arXiv:hep-lat/0410011].
555: %%CITATION = HEP-LAT 0410011;%%
556:
557: \bibitem{Luscher:1986pf} M.~L{\"u}scher,
558: %``Volume Dependence Of The Energy Spectrum In Massive Quantum Field Theories. 2. Scattering States,''
559: Commun. Math. Phys. \textbf{105} 153 (1986).
560:
561: \bibitem{Beane:2003da} S.~R.~Beane \textit{et al.},
562: % P.~F.~Bedaque, A.~Parre\~no and M.~J.~Savage,
563: %``Two nucleons on a lattice,''
564: Phys.\ Lett.\ B \textbf{585}, 106 (2004). %[arXiv:hep-lat/0312004].
565: %%CITATION = HEP-LAT 0312004;%%
566:
567: \bibitem{Detmold:2004qn} W.~Detmold and M.~J.~Savage,
568: %``Electroweak matrix elements in the two-nucleon sector from lattice QCD,''
569: Nucl.\ Phys.\ A \textbf{743}, 170 (2004). %%CITATION = HEP-LAT 0403005;%%
570:
571: \bibitem{Gockeler:1996mu} M.~G\"ockeler \textit{et al.},
572: %R.~Horsley, E.~M.~Ilgenfritz, H.~Perlt, P.~Rakow, G.~Schierholz and A.~Schiller,
573: %``Lattice Operators for Moments of the Structure Functions and their
574: %Transformation under the Hypercubic Group,''
575: Phys.\ Rev.\ D \textbf{54}, 5705 (1996). %[arXiv:hep-lat/9602029].
576: %%CITATION = HEP-LAT 9602029;%%
577:
578: \bibitem{Saito:2000fx} K.~Saito \textit{et al.},
579: % C.~Boros, K.~Tsushima, F.~Bissey, I.~R.~Afnan and A.~W.~Thomas,
580: %``Deep inelastic scattering on asymmetric nuclei,''
581: Phys.\ Lett.\ B \textbf{493}, 288 (2000). %[arXiv:nucl-th/0008017].
582: %%CITATION = NUCL-TH 0008017;%%
583:
584: \bibitem{Minerva} D.~Drakoulakos \textit{et al.},
585: % [Minerva Collaboration],
586: %``Proposal to perform a high-statistics neutrino scattering experiment using a
587: %fine-grained detector in the NuMI beam,''
588: hep-ex/0405002. %%CITATION = HEP-EX 0405002;%%
589:
590: \bibitem{polarisedmodels}
591: I.~C.~Cloet, W.~Bentz and A.~W.~Thomas
592: %``Spin-dependent structure functions in nuclear matter and the
593: %polarized EMC effect''
594: nucl-th/0504019; %%CITATION = NUCL-TH 0504019;%%
595: %
596: J.~R.~Smith and G.~A.~Miller, nucl-th/0505048.
597:
598: \bibitem{Luke:1992cs}
599: M.~E.~Luke and A.~V.~Manohar,
600: %``Reparametrization invariance constraints on heavy particle effective field
601: %theories,''
602: Phys.\ Lett.\ B {\bf 286}, 348 (1992).
603: % [arXiv:hep-ph/9205228].
604: %%CITATION = HEP-PH 9205228;%%
605:
606: \end{thebibliography}
607:
608: \end{document}
609: