1: \documentclass[preprint,aps,eqsecnum]{revtex4}
2: \usepackage{epsfig}
3: \newcommand{\be}{\begin{eqnarray}}
4: \newcommand{\ee}{\end{eqnarray}}
5: \newcommand{\p}{\partial}
6: \newcommand{\w}{\wedge}
7: \newcommand{\ep}{\epsilon}
8: \newcommand{\de}{\delta}
9: \newcommand{\pp}{\perp}
10:
11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12: \begin{document}
13: %\preprint{DO-TH ----}
14: %\preprint{hep-ph/0306275}
15: \vspace*{1cm}
16: \title
17: {Impact Parameter Dependent Parton Distributions for a Relativistic
18: Composite System}
19: \author{\bf D. Chakrabarti}
20: \email{dipankar@phys.ufl.edu}
21: \affiliation{Department of Physics, University of Florida, Gainesville,
22: FL-32611-8440, USA}
23: \author{\bf A. Mukherjee}\email{asmita@lorentz.leidenuniv.nl}
24: \affiliation{ Institute Lorentz, University of Leiden, 2300 RA Leiden,
25: Netherlands}
26:
27: \date{\today\\[2cm]}
28:
29: \begin{abstract}
30: We investigate the impact parameter dependent parton distributions for a
31: relativistic composite system in light-front framework. We take an effective
32: two-body spin-$1/2$ state, namely an electron dressed with a photon in QED. We express
33: the impact parameter dependent parton distributions in terms of overlaps of
34: light-cone wave functions. We obtain the scale dependence of both fermion
35: and gauge boson distributions and show the distortion of the pdfs in the
36: transverse space for transverse polarization of the state at one loop level.
37:
38: \end{abstract}
39: \maketitle
40:
41: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
42: \section{Introduction}
43: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
44: Impact parameter dependent parton distributions $q(x,b^\perp)$
45: \cite{bur1,bur2} have been
46: introduced recently as a physical interpretation of generalized parton
47: distributions (GPDs) \cite{gpd} in terms of probability densities in the
48: impact parameter space. It is known that GPDs are off-forward matrix
49: elements of light-cone bilocal operators and they do no have a probabilistic
50: interpretation like ordinary parton distributions. $q(x,b^\perp)$ can be
51: expressed as Fourier transform of GPDs
52: with respect to the transverse momentum transfer (when the longitudinal
53: momentum transfer is zero) and they give the
54: distribution of partons in the transverse position space. In fact they obey
55: certain
56: positivity constraints and thus it is legitimate to associate the physical
57: interpretation as a probability density. This interpretation is not limited by
58: relativistic effects in the infinite momentum frame. Another interesting
59: aspect is that when the state is polarized in the transverse direction,
60: the unpolarized impact parameter dependent pdf is distorted in the
61: transverse plane \cite{bur1}. This distortion can be linked to the
62: transverse single spin asymmetries observed in transversely polarized Lambda
63: production in $pp$ or $p {\bar p}$ collisions. Recently it has been shown in
64: the framework of the scalar diquark model that the impact parameter space
65: asymmetry together with the final state interaction of the active quark gives
66: rise to the Sivers effect \cite{siv} in momentum space \cite{bur4}.
67:
68: Impact parameter dependent pdfs have been investigated in chiral
69: quark model for the pion \cite{model1}, in the constituent quark model
70: \cite{model2},
71: in terms of a power-law ansatz of the light-cone wave function
72: for the pion \cite{model3}, in the context of investigating the color transparency
73: phenomena \cite{model4}, in the transverse lattice framework for the pion
74: \cite{model5} and on the lattice \cite{lattice}. In this work, we
75: investigate both fermion and gauge boson distributions for a relativistic composite system in the light-front
76: framework, taking into account the correlation between different Fock
77: components of the light-cone (or light-front) wave function in light-front
78: gauge. It is well known that the
79: light-front formalism is especially
80: suitable to investigate many aspects of relativistic bound states because of
81: the Galilean transverse boost operators and the triviality of the vacuum
82: \cite{wilson} when a cutoff is imposed on the longitudinal momentum. We take an effective
83: spin ${1\over 2}$ system, namely an electron \cite{impact}, dressed with a photon in QED.
84: It is known that such a model is self consistent and has been used
85: to investigate the helicity structure of a composite relativistic system
86: \cite{brod1}. The state can be expanded in Fock space in terms of light-cone
87: wave functions. The light-cone wave functions in this case can be obtained
88: from perturbation theory, and thus their correlations are known at a certain
89: order in the coupling constant. However, their general form provides a template for
90: parametrizing the more realistic composite system, namely the hadronic wave
91: function \cite{brod1}. Instead of an electron, one can also consider a state like a
92: dressed quark in perturbative QCD. In fact this
93: can be thought of as a field theoretic parton model where the partons, or
94: the quarks and gluons, have mass, intrinsic transverse momenta and they
95: interact. Previous studies have shown that this gives an intuitive picture
96: of the DIS structure functions
97: and scaling violations \cite{hari} and is suitable to address issues
98: related to the
99: spin and orbital angular momentum of the nucleon \cite{oam,hari1}.
100: Such a state has also been used to investigate the twist three
101: GPDs in terms of overlaps of light-cone wave functions \cite{mvh}.
102: The aim of the
103: present work is to extend these studies to the impact parameter dependent
104: fermion and gauge boson pdfs in order to obtain a qualitative behavior in
105: the impact parameter space within the framework of perturbation theory.
106:
107: The plan of the paper is as follows. In section II, we give the definition of
108: the impact parameter dependent pdf, In sections III and IV we calculate the
109: fermion and gauge boson distributions respectively. Summary and conclusions
110: are given in section V.
111: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
112: \section{Impact Parameter Dependent Parton Distributions}
113: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
114: Impact parameter dependent pdfs are defined by considering a transversely
115: localized state \cite{bur1}:
116: \be
117: \mid P^+, R^\perp=0, \lambda \rangle= N\int {d^2P^\perp\over (2 \pi)^2}
118: \mid P^+, P^\perp, \lambda\rangle,
119: \ee
120: where $ \mid P^+, P^\perp, \lambda\rangle$ is the light cone helicity
121: eigenstate with momentum $P$ and helicity $\lambda$, $N$ is a normalization
122: factor. For a state with total momentum $P^+$ the transverse center of
123: momentum $R^\perp$ is defined as,
124: \be
125: R^\perp= {1\over 2 P^+} \int dx^- d^2x^\perp \Theta^{++} x^\perp,
126: \ee
127: where $\Theta^{\mu \nu}$ is the energy-momentum tensor.
128:
129: The states $\mid P^+, P^\perp, \lambda\rangle$ can be thought of as helicity
130: eigenstates in the infinite momentum frame \cite{bur1}. Due to the fact that
131: the light-front transverse boost $B^\perp$ behaves similar to the
132: non-relativistic Galilean Boost generators, it can be shown that the states
133: $\mid P^+, R^\perp=0, \lambda \rangle$ defined above are simultaneous
134: eigenstates of the longitudinal light cone momentum $P^+$, transverse
135: position operator $R^\perp$ and light cone helicity $J^z$. The transverse
136: position operator $R^\perp$ in fact is related to $B^\perp$ : $
137: R^\perp=-{1\over P^+} B^\perp$.
138:
139: The impact parameter dependent pdf is defined as,
140: \be
141: q(x,b^\perp)= \langle P^+, R^\perp=0^\perp, \lambda
142: \mid O_q(x,b^\perp) \mid P^+, R^\perp=0^\perp,\lambda \rangle
143: \label{impact}
144: \ee
145: with
146: \be
147: O_q(x, b^\perp)= \int {dy^- \over 4 \pi} {\bar \psi} (-{y^-\over 2}, b^\perp)
148: \gamma^+ \psi({y^-\over 2}, b^\perp) e^{{i\over 2} x P^+ y^-},
149: \ee
150: $b^\perp$ is the impact parameter, which is the transverse distance
151: of the active quark from the center of mass. We have taken the light-front
152: gauge, $A^+=0$. Instead of the fermion
153: operator, one can also have a gauge boson operator
154: \be
155: O_g(x, b^\perp)= \int {dy^- \over 4 \pi} F^{+ \nu}(-{y^-\over 2}, b^\perp)
156: {F^+}_\nu({y^-\over 2}, b^\perp) e^{{i\over 2} x P^+ y^-}
157: \ee
158: to define the gauge boson pdf $g(x,b^\perp)$ similar to Eq. (\ref{impact}).
159:
160: It can be shown that $ q(x, b^\perp)$ can be expressed as a Fourier
161: transform of the GPD $H_q(x, 0, \Delta^2)$ \cite{bur1} :
162: \be
163: q(x,b^\perp)= {\mathcal H}_q (x,b^\perp) = \int {d^2 \Delta^\perp\over (2 \pi)^2}
164: e^{-i b^\perp. \Delta^\perp} H_q(x, 0, \Delta^2).
165: \ee
166: where $ \Delta^2$ is the total momentum transfer and it is purely
167: transverse. One can obtain a similar relation for the gauge boson
168: counterpart.
169: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
170: \section{Fermion Distributions}
171: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
172: We first calculate the off-forward matrix elements, parametrized in terms of
173: the twist-two GPSs $ H_q({x},\Delta^2)$ and $ E_q({x},\Delta^2) $ in the
174: standard way \cite{ji} and using the light cone
175: spinors \cite{ped}. We have,
176: \be
177: \int {dz^-\over {8 \pi}} e^{{i\over 2}{x}{\bar
178: P^+} {z^-}}\langle P' \uparrow \mid \bar \psi (-{ z^-\over 2})
179: \gamma^+\psi ({ z^-\over 2})
180: \mid P \uparrow \rangle = H_q({x},\Delta^2),
181: \label{def}
182: \ee
183: where $H(x, \Delta^2)= H(x,0, \Delta^2)$.
184: The momentum of the initial state
185: is $P^\mu$ and that of the final state is $P'^\mu$. The average momentum
186: between initial and final state is then ${\bar P}^\mu={P^\mu+P'^\mu\over
187: 2}$.
188: The momentum transfer is given by $\Delta^\mu=P'^\mu-P^\mu$, $ P'_\perp =
189: -P_\perp
190: ={\Delta_\perp\over 2}$. We take the skewedness $\xi=0$, in other words, the
191: momentum transfer is purely transverse. We exclude the impact parameter space
192: representation of GPDs for non-zero skewedness
193: \cite{diehl} from our analysis here.
194:
195: We take the state
196: $ \mid P, \sigma \rangle$ to be a dressed electron
197: consisting of bare states of an electron and an electron plus
198: a photon :
199: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
200: %\begin{eqnarray}
201: %\mid P, \sigma \rangle && = \phi_1 b^\dagger(P,\sigma) \mid 0 \rangle
202: %\nonumber \\
203: %&& + \sum_{\sigma_1,\lambda_2} \int
204: %{dk_1^+ d^2k_1^\perp \over \sqrt{2 (2 \pi)^3 k_1^+}}
205: %\int
206: %{dk_2^+ d^2k_2^\perp \over \sqrt{2 (2 \pi)^3 k_2^+}}
207: %\sqrt{2 (2 \pi)^3 P^+} \delta^3(P-k_1-k_2) \nonumber \\
208: %&& ~~~~~\phi_2(P,\sigma \mid k_1, \sigma_1; k_2 , \lambda_2) b^\dagger(k_1,
209: %\sigma_1) a^\dagger(k_2, \lambda_2) \mid 0 \rangle.
210: %\label{eq2}
211: %\end{eqnarray}
212: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
213: \begin{eqnarray}
214: \mid P, \sigma \rangle && = {\cal N}\Big [ b^\dagger(P,\sigma) \mid 0 \rangle
215: \nonumber \\
216: && + \sum_{\sigma_1,\lambda_2} \int
217: {dk_1^+ d^2k_1^\perp \over \sqrt{2 (2 \pi)^3 k_1^+}}
218: \int
219: {dk_2^+ d^2k_2^\perp \over \sqrt{2 (2 \pi)^3 k_2^+}}
220: \sqrt{2 (2 \pi)^3 P^+} \delta^3(P-k_1-k_2) \nonumber \\
221: && ~~~~~\phi_2(P,\sigma \mid k_1, \sigma_1; k_2 , \lambda_2) b^\dagger(k_1,
222: \sigma_1) a^\dagger(k_2, \lambda_2) \mid 0 \rangle \Big ].
223: \label{eq2}
224: \end{eqnarray}
225:
226: Here $a^\dagger$ and $b^\dagger$ are bare photon and electron
227: creation operators respectively and $\phi_2$ is the
228: multiparton wave function. It is the probability amplitude to find
229: one electron plus photon inside the dressed electron state.
230:
231:
232: We introduce Jacobi momenta $x_i$, ${q_i}^\perp$ such that $\sum_i x_i=1$ and
233: $\sum_i {q_i}^\perp=0$. They are defined as
234: \be
235: x_i={k_i^+\over P^+}, ~~~~~~q_i^\perp=k_i^\perp-x_i P^\perp.
236: \ee
237: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
238: %Also, we introduce the wave functions,
239: %\be
240: %\psi_1=\phi_1, ~~~~~~~~~~~\psi_2(x_i,q_i^\perp)= {\sqrt {P^+}} \phi_2
241: %(k_i^+,{k_i}^\perp);
242: %\ee
243: %which are independent of the total transverse momentum $P^\perp$ of the
244: %state and are boost invariant.
245: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
246: Also, we introduce the wave function,
247: \be
248: \psi_2(x_i,q_i^\perp)= {\sqrt {P^+}} \phi_2 (k_i^+,{k_i}^\perp);
249: \ee
250: which is independent of the total transverse momentum $P^\perp$ of the
251: state and boost invariant.
252: The state is normalized as,
253: \be
254: \langle P',\lambda'\mid P,\lambda \rangle = 2(2\pi)^3
255: P^+\delta_{\lambda,\lambda'} \delta(P^+-{P'}^+)\delta^2(P^\perp-P'^\perp).
256: \label{norm}
257: \ee
258: The two particle wave function depends on the helicities of the electron and
259: photon. Using the eigenvalue equation for the light-cone Hamiltonian, this
260: can be written as \cite{hari},
261: \be
262: \psi^\sigma_{2\sigma_1,\lambda}(x,q^\perp)&=& -{x(1-x)\over
263: (q^\perp)^2+m^2 (1-x)^2}
264: {1\over {\sqrt {(1-x)}}} {e\over
265: {\sqrt {2(2\pi)^3}}} \chi^\dagger_{\sigma_1}\Big [ 2 {q^\perp\over
266: {1-x}}+{{\tilde \sigma^\perp}\cdot q^\perp\over x} {\tilde \sigma^\perp}
267: \nonumber\\&&~~~~~~~~~~~~~~~~~~
268: -i m{\tilde \sigma}^\perp {(1-x)\over x}\Big ]\chi_\sigma
269: \epsilon^{\perp *}_\lambda .
270: \label{psi2}
271: \ee
272: $m$ is the bare mass of the electron, $\tilde \sigma^2 = -\sigma^1$ and
273: $\tilde \sigma^1= \sigma^2$. In our case,
274: $\Delta^2=-(\Delta^\perp)^2$ as $\xi=0$.
275: %$\psi_1$ actually gives the normalization of the state \cite{hari}:
276: ${\cal N}$ gives the normalization of the state \cite{hari}:
277: \be
278: {\cal N}^2=1-{\alpha\over {2 \pi}}
279: \int_\epsilon^{1-\epsilon} dx \Big [{{1+x^2}\over {1-x}}log
280: {Q^2\over \mu^2}-{1+x^2\over 1-x}+(1-x) \Big ],
281: \label{c5nq}
282: \ee
283: within order $\alpha$. Here $\epsilon$ is a small cutoff on $x$. We have
284: taken the cutoff of the transverse momenta to be $Q^2$, $\mu^2$ is a scale
285: which we have taken to be $ m^2 << \mu^2 << \Lambda^2$, it mimics the
286: factorization scale in QCD separating hard and soft dynamics \cite{hari}.
287: The above expression is derived using Eqs (\ref{norm}), (\ref{eq2}) and
288: (\ref{psi2}).
289:
290: Contribution to $H_q(x, \Delta^2)$ comes from one particle and two-particle sectors.
291: The one-particle sector contributes only when
292: ${x}=1$ and is a delta function. This receives correction upto order $\alpha$
293: from the normalization condition of the state.
294: The contribution from the two-particle sector can be written as an overlap
295: of wave functions \cite{overlap,mvh},
296: \be
297: {\cal N}^2 \sum \int d^2q^\perp \psi^*_2(x,
298: q^\perp+(1-x) \Delta^\perp) \psi_2(x, q^\perp).
299: \label{matrix+}
300: \ee
301:
302: Using the expression Eq. (\ref{psi2}) we get
303: \be
304: H_q(x, \Delta^2)&=&{\cal N}^2 \delta (1-x) \nonumber \\&&
305: \!\!\!\!\!\!\!\!\!\!+ {e^2\over (2 \pi)^3} {\cal N}^2 \Big [
306: \int d^2 q^\perp {{{1+x^2\over 1-x} (q^\perp)^2 +m^2(1-x)^3}\over
307: {((q^\perp)^2+m^2 (1-x)^2) ((q^\perp+(1-x) \Delta^\perp)^2 +m^2
308: (1-x)^2)}}\nonumber\\&&\!\!\!\!\!\!\!\!\!\!\!+
309: (1+x^2)\int d^2q^\perp {q^\perp .\Delta^\perp\over
310: {((q^\perp)^2+m^2 (1-x)^2) ((q^\perp+(1-x) \Delta^\perp)^2 +m^2 (1-x)^2)}} \Big
311: ].
312: \ee
313: One can write this as,
314: \be
315: H_q(x, \Delta^2)&=& {\cal N}^2 \Bigg \{ \delta (1-x)
316: + {e^2\over (2 \pi)^3} \Bigg [ \Big ({1+x^2\over 1-x} \Big ) \Big [ \int {d^2q^\perp\over L_2}
317: \nonumber\\&&-m^2 (1-x)^2 \int {d^2 q^\perp\over L_1L_2}\Big ] +m^2 (1-x)^3 \int {
318: d^2 q^\perp\over L_1L_2} +{1+x^2\over 2 (1-x)} \Big [
319: \int {d^2q^\perp\over L_1}\nonumber\\&&-\int {d^2q^\perp\over L_2}-(1-x)^2
320: (\Delta^\perp)^2 \int {d^2q^\perp\over L_1 L_2} \Big ] \Bigg ] \Bigg \},
321: \label{hqtot}
322: \ee
323: where $L_1=(q^\perp)^2+m^2(1-x)^2$ and $L_2=(q^\perp+(1-x)
324: \Delta^\perp)^2+m^2 (1-x)^2$.
325: In the forward limit using the normalization
326: condition of the state we get,
327: \be
328: H_q(x,0)&=&\delta (1-x)+ {\alpha\over 2\pi} \Bigg \{ \Big [ {1+x^2\over 1-x}log
329: {Q^2\over \mu^2} -{1+x^2\over 1-x} +1-x \Big ] \nonumber\\&&- \delta (1-x)
330: \Big [\int dy
331: \Big ( {1+y^2\over 1-y} log {Q^2\over \mu^2}-{1+y^2\over 1-y}+1-y \Big )
332: \Big ] \Bigg \}.
333: \label{for}
334: \ee
335: Integrating over $x$, one gets $\int_0^1 dx H_q(x,0)= F_1(0)=1$, where
336: $F_1(0)$ is the form factor at zero momentum transfer.
337: Replacing $\alpha$ by $\alpha_s C_f$ and neglecting the mass, eq.
338: (\ref{for}) reduces to the quark distribution
339: function of a dressed quark and the coefficient of the logarithmic term
340: gives the splitting function $P_{qq}$ \cite{hari}. At
341: non-zero $\Delta^\perp$, in the limit $x \rightarrow 1$, we have
342: \be
343: H_q(x, \Delta^2) \rightarrow \delta (1-x)+ {\alpha\over 2 \pi} log
344: {Q^2\over \mu^2} \Big [ {1+x^2\over (1-x)_+}+{3\over 2} \delta (1-x)\Big ].
345: \ee
346: The plus '+' prescription is defined in the usual way. There is no singularity
347: at $x \rightarrow 1$ and $H_q(x, \Delta^2)$ is
348: independent of $\Delta^\perp$ at $x \rightarrow 1$.
349: The impact parameter
350: dependent parton distribution is obtained by taking a Fourier transform with
351: respect to $\Delta^\perp$. At $x \rightarrow 1$ , we
352: get
353: \be
354: {\mathcal H}_q(x,b^\perp)&=& \int {d^2 \Delta^\perp\over (2 \pi)^2}
355: e^{-i b^\perp. \Delta^\perp} H_q(x,
356: \Delta^2)\nonumber\\&&
357: =\delta^2 (b^\perp)\Big \{ \delta (1-x)+ {\alpha\over 2 \pi} log
358: {Q^2\over \mu^2} \Big [ {1+x^2\over (1-x)_+}+{3\over 2} \delta (1-x)\Big ]
359: \Big \}.\label{hq}
360: \ee
361: as expected because in this limit the electron carries all the momentum
362: and the transverse width of
363: the impact parameter dependent pdf vanishes \cite{bur1}.
364:
365: For $ x $ not equal to $1$, $ {\mathcal H}_q(x,b^\perp) $ has nontrivial
366: $b^\perp$ dependency which comes from the (finite)
367: mass terms as well as the $\Delta^\perp$ terms. The $q_\perp$ integration from which
368: we get the scale dependency in the above expression (Eqn. \ref{hq}) cannot be done
369: analytically for this case. Here
370: we do not plot the scale dependent $ {\mathcal H}_q(x,b^\perp) $.
371:
372: Next we consider the helicity flip part of the matrix element :
373: \be
374: \int {dy^-\over 8 \pi} &&e^{{i\over 2} P^+ y^-x}\langle P+\Delta, \uparrow
375: \mid {\overline \psi}({-y^-\over 2}) \gamma^+ \psi ({y^-\over 2})\mid P,
376: \downarrow \rangle =
377: {e^2\over (2 \pi)^3} x (1-x)^2 (-i m) (-i \Delta^1-\Delta^2)\nonumber\\&&
378: \int
379: {d^2 q^\perp \over {((q^\perp)^2+m^2 (1-x)^2) ((q^\perp+(1-x) \Delta^\perp)^2
380: +m^2 (1-x)^2)}} \nonumber\\&&=-{E_q\over 2 m} (\Delta^1-i \Delta^2).
381: \ee
382: $m$ is the renormalized mass of the electron. In the limit $\Delta^\perp =0$
383: we get
384: \be
385: E_q(x,0)= {\alpha\over\pi} x.
386: \ee
387: Integration over $x$ gives the Schwinger value for the anomalous magnetic
388: moment of an electron in QED :
389: \be
390: \int_0^1 E_q dx = F_2(0) = {\alpha\over 2 \pi}.
391: \ee
392: This result can also be obtained by directly calculating the matrix element
393: of the current operator in the light-front framework \cite{brod1}.
394:
395: \begin{center}
396: %\hspace{-0.3cm}
397: \parbox{8cm}{\epsfig{figure=Fig1a_v2.eps,width=7.7 cm,height=7 cm}}\
398: \
399: %\hspace{0.3cm}
400: \parbox{8cm}{\epsfig{figure=Fig1b_v2.eps,width=7.5 cm,height=7 cm}}\
401: \
402: \end{center}
403: \vspace{0.2cm}
404: \begin{center}
405: \parbox{14.0cm}
406: {{\footnotesize
407: Fig. 1: (Color online) (a) ${\mathcal E}_q (x, b^\perp)$ vs $b^\perp$ for three different values
408: of $x$; (b) Derivative of ${\mathcal E}_q (x, b^\perp)$ with respect to
409: $b^x$ as a function of $b^\perp$. }}
410: \end{center}
411:
412: For non-zero $\Delta^\perp$, $E_q$ becomes,
413: \be
414: E_q(x, \Delta^2) = {\alpha\over \pi} m^2 \int_0^1 d \beta {x \over
415: ( \beta (1-\beta) (\Delta^\perp)^2 +m^2)},
416: \label{eq}
417: \ee
418: $\beta $ is the Feynman parameter. The scale dependence in this case is
419: suppressed and we can integrate over the full $\mid q^\perp \mid $ range,
420: from $0$ to $\infty$.
421:
422:
423: Taking the Fourier transform, we get
424: \be
425: {\mathcal E}_q(x, b^\perp)&=&\int {d^2 \Delta^\perp\over (2 \pi)^2}
426: e^{-i b^\perp. \Delta^\perp} E_q(x,-(\Delta^\perp)^2)\nonumber\\&&
427: = \int {\Delta d \Delta\over (2 \pi)^2} d \theta E_q(x, -\Delta^2)
428: e^{-i b \Delta cos \theta}
429: \nonumber\\&&= {1\over 2 \pi}\int_0^\infty \Delta d \Delta
430: E_q (x, -\Delta^2) J_0(b \Delta).
431: \label{fteq}
432: \ee
433: For numerical calculation we use the real integral form of the Bessel
434: function :
435: \be
436: J_0(b \Delta)= {1\over \pi} \int_0^\pi cos (b \Delta sin \theta) d \theta.
437: \ee
438:
439:
440: If we take a state polarized in the ${\hat y}$ direction (in the infinite
441: momentum frame) and whose center of momentum is at the origin,
442: \be
443: \mid P^+, R^\perp=0,{\hat y} \rangle = {1\over \sqrt 2} (
444: \mid P^+, R^\perp=0, \uparrow \rangle+i
445: \mid P^+, R^\perp=0, \downarrow \rangle);
446: \ee
447: the unpolarized fermion distribution in the impact parameter space gets
448: distorted,
449: \be
450: q_{\hat y}(x,b^\perp) & =& \langle P^+, R^\perp=0^\perp, {\hat y}
451: \mid O_q(x,b^\perp) \mid P^+, R^\perp=0^\perp, {\hat y} \rangle \nonumber\\
452: &&= {\mathcal H }_q(x,b^\perp) +{1\over 2 m}
453: {\partial \over \partial b^x} {\mathcal E}_q (x, b^\perp).
454: \ee
455: The distortion is directly related to ${\mathcal E}_q (x,b_\perp)$. Fig. 1
456: (a) shows the helicity flip contribution ${\mathcal E}_q (x, b^\perp)$ as a
457: function of $b^\perp$ for three different values of $x$. We have plotted for
458: positive $b^\perp$. ${\mathcal E}_q (x, b^\perp)$ is a smooth function of
459: $b^\perp$ in the range shown and it increases as $b^\perp$ decreases. Also,
460: it increases linearly with $x$, as is clear from Eq.(\ref{eq}). We have taken
461: the overall normalization ${\alpha\over 2 \pi}=1$ in order to study the
462: qualitative behavior and $m=0.5$. ${\mathcal E}_q (x, b^\perp)$ has a maximum at
463: $b^\perp=0$. Fig. 1 (b) shows the derivative of ${\mathcal E}_q (x, b^\perp)$
464: with respect to $b^x$, which gives the distortion of the pdf of an electron
465: at one loop
466: in impact parameter space due to transverse polarization. The integrand
467: in this case contains $J_1(b \Delta )$ and it is highly
468: oscillatory. However, the integral converges
469: for the $b^\perp$ range shown
470: in the plot. As ${\mathcal E}_q (x, b^\perp)$ is a smooth
471: function with maximum at $b^\perp=0$ its derivative for positive $b^\perp$
472: is negative. One can see that the distortion of the distribution in impact
473: parameter space increases
474: as $b^\perp$ decreases and for a given $b^\perp$ the distortion is higher in
475: magnitude for larger values of $x$. The distortion shifts the distribution
476: actually toward negative values of $b^\perp$.
477:
478: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
479: \section{Gauge Boson Distributions}
480: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
481:
482: From the definition,
483: \be
484: H_g(x, \Delta^2)= {1\over 8 \pi x P^+} \int dy^- e^{{i\over 2} P^+
485: y^- x} \langle P' , \uparrow \mid F^{+ \nu} (-{y^-\over 2}) F^+_\nu ({y^-\over
486: 2}) \mid P, \uparrow \rangle;
487: \ee
488: we take the state to be a dressed electron as in Eq. (\ref{eq2}).
489: Contribution in this case comes only from the two-particle sector, and we
490: get
491: \be
492: H_g(x, \Delta^2) &=& {\cal N}^2 \int d^2 q^\perp \psi_2^* (1-x,q^\perp)
493: \psi_2(1-x, q^\perp+(1-x) \Delta^\perp) \nonumber\\
494: &=&{\alpha\over (2\pi)^2} \Bigg \{ {(1+(1-x)^2)\over x} \int {d^2q^\perp
495: \over L_2} +m^2
496: x^3 \int {d^2 q^\perp\over L_1 L_2} +{1+(1-x)^2\over 2 x} \Big [ \int
497: {d^2 q^\perp\over L_1} \nonumber\\&&-\int {d^2q^\perp\over L_2}-(1-x)^2
498: (\Delta^\perp)^2 \int {d^2q^\perp\over L_1 L_2} \Big ] \Bigg \},
499: \ee
500: where $L_1=(q^\perp)^2 +m^2 x^2$ and $L_2=(q^\perp +(1-x) \Delta^\perp)^2+m^2
501: x^2$.
502: The scale dependency as before, comes from the limits of the $q^\perp$
503: integrations. Fourier transform of the above expression gives the impact
504: parameter dependent gauge boson pdf. Nontrivial $b^\perp$ dependence come
505: from the mass term and the $\Delta^\perp$ term. The other terms give
506: logarithmic dependence on the scale.
507: In the forward limit,
508: \be
509: H_g(x, 0)
510: = {\alpha\over 2 \pi} \Big [{ 1+(1-x)^2\over x}\Big ] log{Q^2\over \mu^2}.
511: \ee
512:
513: \begin{figure}
514: %\hspace{-0.3cm}
515: \parbox{8cm}{\epsfig{figure=Fig2a_v2.eps,width=7.7 cm,height=7 cm}}\
516: \
517: %\hspace{0.3cm}
518: \parbox{8cm}{\epsfig{figure=Fig2b_v2.eps,width=7.5 cm,height=7 cm}}\
519: \
520: \vspace{0.2cm}
521: \begin{center}
522: \parbox{14.0cm}
523: {{\footnotesize
524: Fig. 2: (Color online) (a) ${\mathcal E}_g (x, b^\perp)$ vs $b^\perp$ for four different values
525: of $x$; (b) Derivative of ${\mathcal E}_g (x, b^\perp)$ with respect to
526: $b^x$ as a function of $b^\perp$. }}
527: \end{center}
528: \end{figure}
529:
530: Next we look at the helicity flip part of the matrix element. This can be
531: written as,
532:
533:
534: \be
535: \int {dy^-\over 8 \pi} e^{{i\over 2} P^+ y^-x}\langle P+\Delta, \uparrow
536: \mid F^{+ \nu}({-y^-\over 2}) F^+_\nu ({y^-\over
537: 2}) \mid P, \downarrow \rangle
538: =-{E_g\over 2 m} (\Delta^1-i \Delta^2),
539: \ee
540: which gives
541: \be
542: E_g(1-x, \Delta^2) = -{\alpha\over \pi} m^2 \int_0^1 d \beta {x
543: (1-x)^2\over (\beta (1-\beta) (\Delta^\perp)^2 (1-x)^2+ m^2 x^2)}.
544: \ee
545: If we denote the momentum fraction of the quark as $x$ instead of $1-x$, we
546: get,
547: \be
548: E_g(x, \Delta^2) = -{\alpha\over \pi} m^2 \int_0^1 d \beta {x^2
549: (1-x)\over (\beta (1-\beta) (\Delta^\perp)^2 x^2+ m^2 (1-x)^2)}.
550: \ee
551: In the forward limit, this gives,
552: \be
553: E_g(x,0)=-{\alpha\over \pi} {x^2\over 1-x}.
554: \ee
555: The second moment of $E_{q,g}(x,0), \int dx x E_{q,g}(x,0)$ gives in units
556: of ${1\over 2m}$ by how much the transverse center of momentum of the parton
557: $q,g$ is shifted away from the origin in the transversely polarized state.
558: When summed over all partons, the transverse center of momentum would still
559: be at the origin. Indeed it is easy to check for a dressed electron
560: \be
561: \int_0^1 dx x E_q(x,0)+\int_0^1 dx (1-x) E_g (x,0)=0,
562: \ee
563: which is due to the fact that the anomalous gravitomagnetic moment of the
564: electron has to vanish. Note that in the second term, $(1-x)$ is the
565: momentum fraction of the gauge boson. In fact the second moment of $E_{q,g}$
566: appear in the angular momentum sum rule \cite{ji} and they are related to
567: the orbital angular momentum of the nucleon. In the approach we are
568: following, this connection can be seen from the fact that in the light-front
569: gauge, the matrix element of the quark orbital angular momentum operator
570: cancels the contribution of the
571: gluon helicity and orbital angular momentum for a dressed quark
572: in the helicity sum rule \cite{oam}.
573:
574:
575: The impact parameter dependent gauge boson pdf can be defined in the same
576: way as in Eq. (\ref{fteq}):
577: \be
578: {\mathcal E}_g(x, b^\perp)&=&\int {d^2 \Delta^\perp\over (2 \pi)^2}
579: e^{-i b^\perp. \Delta^\perp} E_g(x,-(\Delta^\perp)^2)\nonumber\\&&
580: = \int {\Delta d \Delta\over (2 \pi)^2} d \theta E_g(x, -\Delta^2)
581: e^{-i b \Delta cos \theta}
582: \nonumber\\&&= {1\over 2 \pi}\int_0^\infty \Delta d \Delta
583: E_g (x, -\Delta^2) J_0
584: (b \Delta).
585: \label{fteg}
586: \ee
587: \begin{figure}
588: %\hspace{-0.3cm}
589: \parbox{8cm}{\epsfig{figure=Fig3a.eps,width=7.7 cm,height=7 cm}}\
590: \
591: %\hspace{0.3cm}
592: \parbox{8cm}{\epsfig{figure=Fig3b.eps,width=7.5 cm,height=7 cm}}\
593: \
594: \vspace{0.2cm}
595: \begin{center}
596: \parbox{14.0cm}
597: {{\footnotesize
598: Fig. 3: (Color online) (a) ${\mathcal E}_g (x, b^\perp)$ vs $x$ for three different values
599: of $b_\perp$; (b) Derivative of ${\mathcal E}_g (x, b^\perp)$ with respect to
600: $b_x$ as a function of $x$. }}
601: \end{center}
602: \end{figure}
603:
604: Fig. 2 (a) shows ${\mathcal E}_g(x, b^\perp)$ vs $b^\perp$ for three
605: different values of $x$. ${\mathcal E}_g(x, b^\perp)$ is negative for
606: positive $b^\perp$ and has a negative maximum at $b^\perp=0$. As before,
607: we took ${\alpha\over 2 \pi}=1$ and $m=0.5$.
608: Fig. 2 (b) shows the derivative of ${\mathcal E}_g(x, b^\perp)$
609: with respect to $b^x$ as a function of $b^\perp$.
610: Fig. 3 (a) shows ${\mathcal E}_g(x, b^\perp)$
611: as a function of $x$ for three different values of $b_\perp$.
612: Unlike the fermion case, ${\mathcal E}_g(x, b^\perp)$ for a fixed $b^\perp$
613: is a complicated function of $x$ and does not increase monotonically as $x$
614: increases. Depending on $b^\perp$, the maximum of ${\mathcal E}_g(x, b^\perp)$
615: appears for different $x$ ($0\le x \le 1$).
616: Again in contrary to the
617: fermionic case, ${\mathcal E}_g(x, b^\perp)$ becomes smoother when $x$ is
618: large which is expected since for large $x$ ($x = 0.7, 0.8$ in Fig. 2 (b)),
619: gauge boson carries only a small ($1-x$) fraction of total momentum.
620: Fig. 3 (b) shows the derivative of ${\mathcal E}_g(x, b^\perp)$
621: with respect to $b^x$ as a function of $x$, which shows the distortion of
622: the pdf in the transverse plane for a transversely polarized state.
623: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
624: \section{Summary}
625: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
626: We investigated the impact parameter dependent parton distributions for a
627: relativistic composite system. An ideal framework is based on light-front
628: field theory, where the transverse boosts behave like Galilean boosts and
629: the longitudinal boost operator produces just a scale transformation. We
630: take an effective composite spin $1/2$ state, namely an electron dressed with
631: a photon in QED. Using the overlap representation of GPDs in terms of
632: light-cone wave functions, we obtained the scale dependence of the impact
633: parameter dependent pdfs at one loop. We also showed the
634: distortion of the fermion and gauge boson distributions in the transverse
635: plane when the state is transversely polarized. Plots show the qualitative behaviors
636: of the helicity flip contributions
637: ${\mathcal E}_q(x, b^\perp)$ and ${\mathcal E}_g(x, b^\perp)$ respectively
638: for the electron and the gauge boson and their
639: derivatives with respect to $b_x$ which give the distortion
640: of the distribution
641: in the transverse plane. For both fermion and gauge boson, for a fixed $x$,
642: the distortions are larger in magnitude for smaller $b_\perp$.
643:
644: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
645: \section{acknowledgment}
646: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
647:
648: We thank M. Burkardt for valuable discussions and S.J. Brodsky for
649: interesting comments. We also thank the organisers of WHEPP-8 at IIT Bombay
650: where part of this work was carried out. The work of AM has been supported
651: in part by the 'Bundesministerium f\"ur Bildung und Forschung', Berlin/Bonn
652: and the work of DC was partially suported by the Department of Energy under
653: Grant No. DE-FG02-97ER-41029.
654:
655: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
656: \begin{thebibliography}{99}
657: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
658:
659: \bibitem{bur1} M. Burkardt, Int. Jour. Mod. Phys. {\bf A 18}, 173 (2003).
660: \bibitem{bur2} M. Burkardt, Phys. Rev. {\bf D 62}, 071503 (2000), Erratum-
661: ibid, {\bf D 66}, 119903 (2002); J. P. Ralston and B. Pire, Phys. Rev. {\bf
662: D 66}, 111501 (2002).
663: \bibitem{gpd} For reviews on generalized parton distributions,
664: see M. Diehl,
665: Phys. Rept, {\bf 388}, 41 (2003); X. Ji, J. Phys. {\bf G 24}, 1181 (1998);
666: A. V. Radyushkin,
667: hep-ph/0101225, published in "At the Frontier of Particle Physics/Handbook
668: of QCD", ed. M. Shifman (World Scientific, Singapore, 2001); K. Goeke, M.
669: V. Polyakov, M. Vanderhaeghen, Prog. Part. Nucl. Phys. {\bf 47}, 401 (2001).
670: \bibitem{siv} D. W. Sivers, Phys. Rev. {\bf D 43}, 261 (1991).
671: \bibitem{bur4} M. Burkardt and D. S. Hwang, Phys. Rev. {\bf D 69}, 074032
672: (2004).
673: \bibitem{model1} W. Bronioski, E. R. Arriola, Phys. Lett. {\bf B 574}, 57
674: (2003).
675: \bibitem{model2} S. Scopetta, V. Vento, Phys. Rev. {\bf D 69}, 094004
676: (2004).
677: \bibitem{model3} A. Mukherjee, I. V. Musatov, H. C. Pauli and A. V.
678: Radyushkin, Phys. Rev {\bf D 67}, 073014 (2003).
679: \bibitem{model4} S. Liuti, S. K. Taneja, hep-ph/0405014.
680: \bibitem{model5} S. Dalley, Phys. Lett. {\bf B 570}, 191 (2003).
681: \bibitem{lattice} LHPC and SESAM collaboration, hep-lat/0312014.
682: \bibitem{wilson} K. G. Wilson, T. S. Walhout, A. Harindranath, W. M. Zhang, R.
683: J. Perry and S. D. Glazek, Phys. Rev. {\bf D 49}, 6720 (1994).
684: \bibitem{impact} D. Chakrabarti and A. Mukherjee, hep-ph/0410126.
685: \bibitem{brod1} S. J. Brodsky, D. S. Hwang, B-Q. Ma, I Schmidt, Nucl. Phys.
686: {\bf B 593}, 311 (2001).
687: \bibitem{hari} A. Harindranath, R. Kundu, W. M. Zhang, Phys.
688: Rev {\bf D 59},
689: 094013; A. Harindranath, R. Kundu, A. Mukherjee, J. P. Vary, Phys. Rev.
690: {\bf D 58}, 114022 (1998), A. Mukherjee and D. Chakrabarti, Phys. Lett. {\bf B
691: 506}, 283 (2001).
692: \bibitem{oam} A. Harindranath and R. Kundu, Phys. Rev. {\bf D 59}, 116013
693: (1999)
694: \bibitem{hari1} A. Harindranath, A. Mukherjee, R. Ratabole, Phys. Lett. {\bf B 476},
695: 471 (2000); Phys. Rev. {\bf D 63}, 045006 (2001).
696: \bibitem{mvh} A. Mukherjee and M. Vanderhaeghen, Phys. Rev. {\bf D 67},
697: 085020 (2003), Phys. Lett. {\bf B 542}, 245 (2002).
698: \bibitem{diehl} M. Diehl, Eur. Phys. Jour. {\bf C 25}, 223 (2002).
699: \bibitem{ji} X. Ji, Phys. Rev. Lett. {\bf 78}, 610 (1997); Phys. Rev. {\bf
700: D55}, 7114 (1997).
701: \bibitem{ped} A. Harindranath, An Introduction to the Light Front
702: Dynamics for Pedestrians in {\it Light-front Quantization and
703: Non-perturbative QCD } , Ed. J. P. Vary and F. Wolz, published by
704: Internatinal Institute of Theoretical and Applied Physics, Ames, Iowa, USA
705: (1997).
706: \bibitem{overlap} M. Diehl, T. Feldmann, R. Jacob, P. Kroll, Nucl. Phys.
707: {\bf B 596}, 33 (2001), Erratum-ibid {\bf 605}, 647 (2001).
708: \end{thebibliography}
709: \end{document}
710: