1:
2: \documentclass[ ,showpacs
3: ,secnumarabic, nobibnotes, nofootinbib, aps, prd]{revtex4}
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
5: \usepackage{bm}
6: \usepackage{graphicx}
7:
8: %TCIDATA{OutputFilter=LATEX.DLL}
9: %TCIDATA{Version=5.50.0.2953}
10: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}
11: %TCIDATA{BibliographyScheme=Manual}
12: %TCIDATA{LastRevised=Tuesday, February 20, 2007 01:29:18}
13: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
14: %TCIDATA{Language=American English}
15:
16: \newtheorem{theorem}{Theorem}
17: \newtheorem{acknowledgement}[theorem]{Acknowledgement}
18: \newtheorem{algorithm}[theorem]{Algorithm}
19: \newtheorem{axiom}[theorem]{Axiom}
20: \newtheorem{claim}[theorem]{Claim}
21: \newtheorem{conclusion}[theorem]{Conclusion}
22: \newtheorem{condition}[theorem]{Condition}
23: \newtheorem{conjecture}[theorem]{Conjecture}
24: \newtheorem{corollary}[theorem]{Corollary}
25: \newtheorem{criterion}[theorem]{Criterion}
26: \newtheorem{definition}[theorem]{Definition}
27: \newtheorem{example}[theorem]{Example}
28: \newtheorem{exercise}[theorem]{Exercise}
29: \newtheorem{lemma}[theorem]{Lemma}
30: \newtheorem{notation}[theorem]{Notation}
31: \newtheorem{problem}[theorem]{Problem}
32: \newtheorem{proposition}[theorem]{Proposition}
33: \newtheorem{remark}[theorem]{Remark}
34: \newtheorem{solution}[theorem]{Solution}
35: \newtheorem{summary}[theorem]{Summary}
36: \begin{document}
37:
38: \title{Proton structure functions and quark orbital motion\footnote{%
39: Prepared for the 17th International Spin Physics Symposium, SPIN2006, Kyoto,
40: Japan, Oct. 2.-7., 2006.}}
41: \author{Petr Z\'{a}vada}
42: \email{zavada@fzu.cz}
43: \affiliation{Institute of Physics, Academy of Sciences of the Czech Republic, \\
44: Na Slovance 2, CZ-182 21 Prague 8}
45: \pacs{13.60.-r, 13.88.+e, 14.65.-q}
46:
47: \begin{abstract}
48: Covariant version of the quark-parton model is studied. Dependence of the
49: structure functions on the 3D quark intrinsic motion is discussed. The
50: important role of the quark orbital momentum, which is a particular case of
51: intrinsic motion, appears as a direct consequence of the covariant
52: description. Effect of orbital motion is substantial especially for
53: polarized structure functions. At the same time, the procedure for obtaining
54: the quark momentum distribution from the structure functions is suggested.
55: \end{abstract}
56:
57: \maketitle
58:
59: %\date{August, 2006}
60:
61: \section{Introduction}
62:
63: The nucleon structure functions are basic tool for understanding the nucleon
64: internal structure in the language of QCD. And at the same time, the
65: measuring and analysis of the structure functions represent the important
66: experimental test of this theory. Unpolarized nucleon structure functions
67: are known with high accuracy in very broad kinematical region, but in recent
68: years also some precision measurements on the polarized structure functions
69: have been completed \cite{e142,e154,her1,her2,adeva,e143,E155}. For present
70: status of the nucleon spin structure see e.g. \cite{spin04} and citations
71: therein. The more formal aspects of the nucleon structure functions are
72: explained in \cite{efrem}. In fact only the complete set of the four
73: electromagnetic unpolarized and polarized structure functions $%
74: F_{1},F_{2},g_{1}$ and $g_{2}$ can give a consistent picture of the nucleon.
75: However, this picture is usually drawn in terms of the distribution
76: functions, which are connected with the structure functions by some
77: model-dependent way. Distribution functions are not directly accessible from
78: the experiment and model, which is normally applied for their extraction
79: from the structure functions is the well known quark-parton model (QPM).
80: Application of this model for analysis and interpretation of the unpolarized
81: data does not create any contradiction. On the other hand, the situation is
82: much less clear in the case of spin functions $g_{1}$ and $g_{2}$.
83:
84: In our previous study \cite{zav1,zav2} we have suggested, that a
85: reasonable explanation of the experimentally measured spin functions $g_{1},$
86: $g_{2}$ is possible in terms of a generalized covariant QPM, in which the
87: quark intrinsic motion (i.e. 3D motion with respect to the nucleon rest
88: frame) is consistently taken into account. Therefore the quark transversal
89: momentum appears in this approach on the same level as the longitudinal one.
90: The quarks are represented by the free Dirac spinors, which allows to obtain
91: exact and covariant solution for relations between the quark momentum
92: distribution functions and the structure functions accessible from
93: experiment. In this way the model (in its present leading order version)
94: contains no dynamics but only \textquotedblleft exact\textquotedblright\
95: kinematics of quarks, so it can be effective tool for analysis and
96: interpretation of the experimental data on structure functions, particularly
97: for separating effects of the dynamics (QCD) from effects of the kinematics.
98: This point of view is well supported by our previous results:
99:
100: \textit{a)} In the cited papers we showed, that the model simply implies the
101: well known sum rules (Wanzura-Wilczek, Efremov-Leader-Teryaev,
102: Burkhardt-Cottingham) for the spin functions $g_{1},$ $g_{2}$.
103:
104: \textit{b}) Simultaneously, we showed that the same set of assumptions
105: implies rather substantial dependence of the first moment $\Gamma _{1}$ of
106: the function $g_{1}$ on the kinematical effects.
107:
108: \textit{c)} Further, we showed that the model allows to calculate the
109: functions $g_{1},$ $g_{2}$ from the unpolarized valence quark distributions
110: and the result is quite compatible with the experimental data.
111:
112: \textit{d)} In the paper \cite{tra}\ we showed that the model allows to
113: relate the transversity distribution to some other structure functions.
114:
115: These results cannot be obtained from the standard versions of the QPM
116: (naive or the QCD improved), which are currently used for the analysis of
117: experimental data on structure functions. The reason is, that the standard
118: QPM is based on the simplified and non-covariant kinematics in the infinite
119: momentum frame (IMF), which does not allow to properly take into account the
120: quark intrinsic or orbital motion.
121:
122: The subject of our previous study was the question: \textit{What is the
123: dependence of the structure functions on quark intrinsic motion?} The aim of
124: the present paper is a discussion of related problems:
125:
126: \textit{1. How to extract information about the quark intrinsic motion from
127: the experimentally measured structure functions?}
128:
129: \textit{2. What is the role of the quark orbital momentum, which is a
130: particular case of intrinsic motion?}
131:
132: The paper is organized as follows. In the first part of Sec. \ref{sec2} the
133: basic formulas, which follow from the generalized QPM, are presented.
134: Resulting general\ covariant relations are compared with their limiting
135: case, which is represented by the standard formulation of the QPM in the
136: IMF. In the next part of the section the relations for calculation of 3D
137: quark momentum distributions from the structure functions are derived. The
138: quark momentum distributions are obtained from the experimentally measured
139: structure functions $F_{2}$ and $g_{1}$ \ and it is shown how their
140: combination allows to calculate the momentum distributions of the positively
141: and negatively polarized quarks. The particular form of the quark intrinsic
142: motion is the orbital momentum. In Sec. \ref{sec3} the role of the quark
143: orbital momentum in covariant description is discussed and it is shown, why
144: its contribution to the total quark angular momentum can be quite
145: substantial. It is demonstrated, that the orbital motion is an inseparable
146: part of the covariant approach. The problem of quark orbital momentum in the
147: context of nucleon spin was recognized and studied also in many previous
148: papers, see e.g. \cite%
149: {sehgal,ratc,abbas,kep,casu,bqma,brod,waka,waka1,song,ji}. The last section
150: is devoted to a short summary and conclusion.
151:
152: \section{Structure functions and intrinsic quark motion}
153:
154: \label{sec2}In our previous study \cite{zav0,zav1,zav2} of the proton
155: structure functions we showed, how these functions depend on the intrinsic
156: motion of quarks. The quarks in the suggested model are represented by the
157: free fermions, which are in the proton rest frame described by the set of
158: distribution functions with spheric symmetry $G_{k}^{\pm }(p_{0})d^{3}p$,
159: where $p_{0}=\sqrt{m^{2}+\mathbf{p}^{2}}$ and symbol $k$ represents the
160: quark and antiquark flavors. These distributions measure the probability to
161: find a quark of given flavor in the state%
162: \begin{equation}
163: u\left( \mathbf{p},\lambda \mathbf{n}\right) =\frac{1}{\sqrt{N}}\left(
164: \begin{array}{c}
165: \phi _{\lambda \mathbf{n}} \\
166: \frac{\mathbf{p}\mathbf{\sigma }}{p_{0}+m}\phi _{\lambda \mathbf{n}}%
167: \end{array}%
168: \right) ;\qquad \frac{1}{2}\mathbf{n\sigma }\phi _{\lambda \mathbf{n}%
169: }=\lambda \phi _{\lambda \mathbf{n}},\qquad N=\frac{2p_{0}}{p_{0}+m},
170: \label{sp8}
171: \end{equation}%
172: where $m$ and $p$\ are the quark mass and momentum, $\lambda =\pm 1/2,$ $%
173: \phi _{\lambda \mathbf{n}}^{\dagger }\phi _{\lambda \mathbf{n}}=1$ and$\
174: \mathbf{n}$ coincides with the direction of proton polarization. The
175: distributions with the corresponding quark (and antiquark) charges $e_{k}$
176: allow to define the generic functions $G$ and $\Delta G$\footnote{%
177: In the papers \cite{zav1,zav2} we used different notation for the
178: distributions defined by Eqs.(\ref{sp8a}) and (\ref{sp9}): $G_{k}^{\pm
179: },\Delta G_{k}$ and $\Delta G$ were denoted as $h_{k\pm },\Delta h_{k}$ and $%
180: H$. Apart of that we assumed for simplicity that only three (valence) quarks
181: contribute to the sums (\ref{sp8a}) and (\ref{sp9}). In present paper we
182: assume contribution of all the quarks and antiquarks, but apparently general
183: form of the relations like (\ref{cra31}) - (\ref{cr31}) is independent of
184: chosen set of quarks.},%
185: \begin{equation}
186: G(p_{0})=\sum_{k}e_{k}^{2}G_{k}(p_{0}),\quad G_{k}(p_{0})\equiv
187: G_{k}^{+}(p_{0})+G_{k}^{-}(p_{0}), \label{sp8a}
188: \end{equation}%
189: \begin{equation}
190: \Delta G(p_{0})=\sum_{k}e_{k}^{2}\Delta G_{k}(p_{0}),\quad \Delta
191: G_{k}(p_{0})\equiv G_{k}^{+}(p_{0})-G_{k}^{-}(p_{0}), \label{sp9}
192: \end{equation}%
193: from which the structure functions can be obtained. If $q$ is momentum of
194: the photon absorbed by the proton of the momentum $P$ and mass $M,$ in which
195: the phase space of quarks is controlled by the distributions $G_{k}^{\pm
196: }(p_{0})d^{3}p$, then there are the following representations of
197: corresponding structure functions.
198:
199: \textit{A. Manifestly covariant representation}
200:
201: \textit{i) unpolarized structure functions:}
202: \begin{equation}
203: F_{1}(x)=\frac{M}{2}\left( A+\frac{B}{\gamma }\right) ,\qquad F_{2}(x)=\frac{%
204: Pq}{2M\gamma }\left( A+\frac{3B}{\gamma }\right) , \label{sp9b}
205: \end{equation}%
206: where%
207: \begin{equation}
208: A=\frac{1}{Pq}\int G\left( \frac{Pp}{M}\right) \left[ pq-m^{2}\right] \delta
209: \left( \frac{pq}{Pq}-x\right) \frac{d^{3}p}{p_{0}}, \label{sp9c}
210: \end{equation}%
211: \begin{equation}
212: B=\frac{1}{Pq}\int G\left( \frac{pP}{M}\right) \left[ \left( \frac{Pp}{M}%
213: \right) ^{2}+\frac{\left( Pp\right) \left( Pq\right) }{M^{2}}-\frac{pq}{2}%
214: \right] \delta \left( \frac{pq}{Pq}-x\right) \frac{d^{3}p}{p_{0}}
215: \label{sp9d}
216: \end{equation}%
217: and%
218: \begin{equation}
219: \gamma =1-\left( \frac{Pq}{Mq}\right) ^{2}. \label{sp9e}
220: \end{equation}%
221: The functions $F_{1}=MW_{1}$ and $F_{2}=\left( Pq/M\right) W_{2}$ follow
222: from the tensor equation%
223: \begin{eqnarray}
224: &&\left( -g_{\alpha \beta }+\frac{q_{\alpha }q_{\beta }}{q^{2}}\right)
225: W_{1}+\left( P_{\alpha }-\frac{Pq}{q^{2}}q_{\alpha }\right) \left( P_{\beta
226: }-\frac{Pq}{q^{2}}q_{\beta }\right) \frac{W_{2}}{M^{2}} \label{sp9f} \\
227: &=&\int G\left( \frac{pP}{M}\right) \left[ 2p_{\alpha }p_{\beta }+p_{\alpha
228: }q_{\beta }+q_{\alpha }p_{\beta }-g_{\alpha \beta }pq\right] \delta \left(
229: \left( p+q\right) ^{2}-m^{2}\right) \frac{d^{3}p}{p_{0}}. \nonumber
230: \end{eqnarray}%
231: After modification of the delta function term%
232: \begin{equation}
233: \delta \left( \left( p+q\right) ^{2}-m^{2}\right) =\delta \left(
234: 2pq+q^{2}\right) =\delta \left( 2Pq\left( \frac{pq}{Pq}-\frac{Q^{2}}{2Pq}%
235: \right) \right) =\frac{1}{2Pq}\delta \left( \frac{pq}{Pq}-x\right) ;\qquad
236: q^{2}=-Q^{2},\quad x=\frac{Q^{2}}{2Pq}, \label{sp9g}
237: \end{equation}%
238: the dependence on the Bjorken $x$ is introduced. Then contracting with the
239: tensors $g_{\alpha \beta }$ and $P_{\alpha }P_{\beta }$ gives the set of two
240: equations, which determine the functions $F_{1},F_{2}$ in accordance with
241: Eqs. (\ref{sp9b})-(\ref{sp9e}).
242:
243: \textit{ii) polarized structure functions:}
244:
245: As follows from \cite{zav1} the corresponding spin functions in covariant
246: form read%
247: \begin{equation}
248: g_{1}=Pq\left( G_{S}-\frac{Pq}{qS}G_{P}\right) ,\qquad g_{2}=\frac{\left(
249: Pq\right) ^{2}}{qS}G_{P}, \label{cra31}
250: \end{equation}%
251: where $S$ is the proton spin polarization vector and the functions $%
252: G_{P},G_{S}$\ are defined as%
253: \begin{equation}
254: G_{P}=\frac{m}{2Pq}\int \Delta G\left( \frac{pP}{M}\right) \frac{pS}{pP+mM}%
255: \left[ 1+\frac{1}{mM}\left( pP-\frac{pu}{qu}Pq\right) \right] \delta \left(
256: \frac{pq}{Pq}-x\right) \frac{d^{3}p}{p_{0}}, \label{cr30}
257: \end{equation}%
258: \begin{equation}
259: G_{S}=\frac{m}{2Pq}\int \Delta G\left( \frac{pP}{M}\right) \left[ 1+\frac{pS%
260: }{pP+mM}\frac{M}{m}\left( pS-\frac{pu}{qu}qS\right) \right] \delta \left(
261: \frac{pq}{Pq}-x\right) \frac{d^{3}p}{p_{0}}; \label{cr31}
262: \end{equation}%
263: \[
264: u=q+\left( qS\right) S-\frac{\left( Pq\right) }{M^{2}}P.
265: \]
266:
267: \textit{B. Rest frame representation for} $Q^{2}\gg 4M^{2}x^{2}$
268:
269: As follows from the Appendix in \cite{zav1}, if $Q^{2}\gg 4M^{2}x^{2}$ and
270: the above integrals are expressed in terms of the proton rest frame
271: variables, then one can substitute
272: \[
273: \frac{pq}{Pq}\rightarrow \frac{p_{0}+p_{1}}{M}
274: \]%
275: and the structure functions are simplified as:%
276: \begin{eqnarray}
277: F_{1}(x) &=&\frac{Mx}{2}\int G(p_{0})\delta \left( \frac{p_{0}+p_{1}}{M}%
278: -x\right) \frac{d^{3}p}{p_{0}}, \label{sp9h} \\
279: F_{2}(x) &=&Mx^{2}\int G(p_{0})\delta \left( \frac{p_{0}+p_{1}}{M}-x\right)
280: \frac{d^{3}p}{p_{0}}, \label{sp9a} \\
281: g_{1}(x) &=&\frac{1}{2}\int \Delta G(p_{0})\left( m+p_{1}+\frac{p_{1}^{2}}{%
282: p_{0}+m}\right) \delta \left( \frac{p_{0}+p_{1}}{M}-x\right) \frac{d^{3}p}{%
283: p_{0}}, \label{sp10} \\
284: g_{2}(x) &=&-\frac{1}{2}\int \Delta G(p_{0})\left( p_{1}+\frac{%
285: p_{1}^{2}-p_{T}^{2}/2}{p_{0}+m}\right) \delta \left( \frac{p_{0}+p_{1}}{M}%
286: -x\right) \frac{d^{3}p}{p_{0}}, \label{sp11}
287: \end{eqnarray}%
288: where the $p_{1}$ and $p_{T}$ are longitudinal and transversal quark
289: momentum components.
290:
291: \textit{C. Standard IMF representation}
292:
293: The usual formulation of the QPM gives the known relations between the
294: structure and distribution functions \cite{efrem}:%
295: \begin{equation}
296: F_{1}(x)=\frac{1}{2}\sum_{q}e_{q}^{2}q(x),\qquad
297: F_{2}(x)=x\sum_{q}e_{q}^{2}q(x), \label{sp11a}
298: \end{equation}%
299: \begin{equation}
300: g_{1}(x)=\frac{1}{2}\sum_{q}e_{q}^{2}\Delta q(x),\qquad g_{2}(x)=0,
301: \label{sp11b}
302: \end{equation}%
303: where%
304: \begin{equation}
305: q(x)=q^{+}(x)+q^{-}(x),\qquad \Delta q(x)=q^{+}(x)-q^{-}(x). \label{sp11c}
306: \end{equation}
307: In the Appendix \ref{app0} we have proved that these relations represent the
308: particular, limiting case of the covariant relations (\ref{sp9b}) and (\ref%
309: {cra31}).
310:
311: The three versions of the relations between the structure functions and the
312: quark distributions can be compared:
313:
314: \textit{a)} If we skip the function $g_{2}$ in the version \textit{C}, then
315: the relations (\ref{sp11a}) and (\ref{sp11b}) practically represent identity
316: between the structure functions and quark distributions. Such simple
317: relations are valid only for the IMF approach based on the approximation (%
318: \ref{b2}), which means that the quark intrinsic motion is suppressed. In
319: more general versions \textit{A} and \textit{B}, where the intrinsic motion
320: is allowed, the relations are more complex. The intrinsic motion strongly
321: modifies also the $g_{2}$. In the version \textit{C} there is $g_{2}(x)=0$,
322: but $g_{2}(x)\neq 0$ in the \textit{A} and \textit{B}.
323:
324: \textit{b)} The version \textit{B} allows to easily calculate the
325: (substantial) dependence of the first moment $\Gamma _{1}$ on the rate of
326: intrinsic motion. A more detailed discussion follows in the next section.
327: The same approach implies that functions $g_{1}$ and $g_{2}$ for massless
328: quarks satisfy the relation equivalent to the Wanzura-Wilczek term and obey
329: some well known sum rules, that is shown in \cite{zav1}.
330:
331: \textit{c)} The functions $F_{1}$ and $F_{2}$ exactly satisfy the
332: Callan-Gross relation $F_{2}(x)/F_{1}(x)=2x$ in the versions \textit{B} and
333: C, but this relation is satisfied only approximately in the \textit{A}: $%
334: F_{2}(x)/F_{1}(x)\approx 2x+O\left( 4M^{2}x^{2}/Q^{2}\right) $.
335:
336: The task which was solved in different approximations above can be
337: formulated: How to obtain the structure functions $F_{1},F_{2}$ and $%
338: g_{1},g_{2}$ from the probabilistic distributions $G$ and $\Delta G$ defined
339: by Eqs. (\ref{sp8a}) and (\ref{sp9})? In the next we will study the inverse
340: task, the aim is to find out a rule for obtaining the distribution functions
341: $G$ and $\Delta G$ from the structure functions. In the present paper we
342: consider the functions $F_{2}$ and $g_{1}$ represented by Eqs. (\ref{sp9a})
343: and (\ref{sp10}). As follows from the Appendix A in \cite{zav2}, the
344: function%
345: \begin{equation}
346: V_{n}(x)=\int K(p_{0})\left( \frac{p_{0}}{M}\right) ^{n}\delta \left( \frac{%
347: p_{0}+p_{1}}{M}-x\right) d^{3}p \label{gp1}
348: \end{equation}%
349: satisfies%
350: \begin{equation}
351: V_{n}^{\prime }(x_{\pm })x_{\pm }=\mp 2\pi MK(\xi )\xi \sqrt{\xi ^{2}-m^{2}}%
352: \left( \frac{\xi }{M}\right) ^{n};\qquad x_{\pm }=\frac{\xi \pm \sqrt{\xi
353: ^{2}-m^{2}}}{M}. \label{gp2}
354: \end{equation}%
355: In this section we consider only the case $m\rightarrow 0$, then%
356: \begin{equation}
357: V_{n}^{\prime }(x)x=-2\pi MK(\xi )\xi ^{2}\left( \frac{\xi }{M}\right)
358: ^{n};\qquad x=\frac{2\xi }{M}. \label{gp3}
359: \end{equation}%
360: As we shall see below, with the use of this relation one can obtain the
361: probabilistic distributions $G$ and $\Delta G$ from the experimentally
362: measured structure functions.
363:
364: Let us remark that in present stage the QCD evolution is not included into
365: the model. However, this fact does not represent any restriction for the
366: present purpose - to obtain information about distributions of quarks at
367: some $Q^{2}$ from the structure functions measured at the same $Q^{2}$.
368: Distribution of the gluons is another part of the proton picture. But since
369: our present discussion is directed to the relation between the structure
370: functions and corresponding (quark) distributions at given scale, the gluon
371: distribution is left aside.
372:
373: \subsection{Momentum distribution from structure function $F_{2}$}
374:
375: In an accordance with the definition (\ref{gp1}) in which the distribution $%
376: K(p_{0})$\ is substituted\ by the $G(p_{0})$, the structure function (\ref%
377: {sp9a}) can be written in the form
378:
379: \begin{equation}
380: F_{2}(x)=x^{2}V_{-1}(x). \label{gp5}
381: \end{equation}%
382: Then, with the use of the relation (\ref{gp3}) one gets%
383: \begin{equation}
384: G(p)=-\frac{1}{\pi M^{3}}\left( \frac{F_{2}(x)}{x^{2}}\right) ^{\prime }=%
385: \frac{1}{\pi M^{3}x^{2}}\left( \frac{2F_{2}(x)}{x}-F_{2}^{\prime }(x)\right)
386: ;\qquad x=\frac{2p}{M},\qquad p\equiv \sqrt{\mathbf{p}^{2}}=p_{0}.
387: \label{gp8}
388: \end{equation}%
389: Probability distribution $G$ measures number of quarks in the element $%
390: d^{3}p $. Since $d^{3}p=4\pi p^{2}dp$, the distribution measuring the number
391: of quarks in the element $dp/M$ reads%
392: \begin{equation}
393: 4\pi p^{2}MG(p)=-x^{2}\left( \frac{F_{2}(x)}{x^{2}}\right) ^{\prime }=\frac{%
394: 2F_{2}(x)}{x}-F_{2}^{\prime }(x);\qquad x=\frac{2p}{M}. \label{gp9}
395: \end{equation}%
396: Let us note, the maximum value of quark momentum is $p_{\max }=M/2$, which
397: is a consequence of the kinematics in the proton rest frame, where the
398: single quark momentum must be compensated by the momentum of the other
399: partons.
400:
401: Another quantity, which can be obtained, is the distribution of the quark
402: transversal momentum. Obviously the integral%
403: \begin{equation}
404: \frac{dN}{dp_{T}^{2}}=\int G(p)\delta \left(
405: p_{2}^{2}+p_{3}^{2}-p_{T}^{2}\right) d^{3}p, \label{gp12}
406: \end{equation}%
407: which measures the number of quarks in the element $dp_{T}^{2}$, can be
408: modified as%
409: \begin{equation}
410: \frac{dN}{dp_{T}^{2}}=2\pi \int_{0}^{\sqrt{p_{\max }^{2}-p_{T}^{2}}}G\left(
411: \sqrt{p_{1}^{2}+p_{T}^{2}}\right) dp_{1}. \label{gp13}
412: \end{equation}%
413: It follows, that the distribution measuring number of quarks in the element $%
414: dp_{T}/M$ reads:%
415: \begin{equation}
416: P(p_{T})=M\frac{dN}{dp_{T}}=4\pi p_{T}M\int_{0}^{\sqrt{p_{\max
417: }^{2}-p_{T}^{2}}}G\left( \sqrt{p_{1}^{2}+p_{T}^{2}}\right) dp_{1}.
418: \label{gp14}
419: \end{equation}%
420: Let us point out that the distribution $P(p_{T}),$ equally as the
421: distributions $G$ and $\Delta G,$ represents the combination of the
422: distributions related to different quark and antiquark flavors - like in
423: Eqs. (\ref{sp8a}) or (\ref{sp9}). With the use of Eq. (\ref{gp8}) one gets
424: distributions%
425: \begin{equation}
426: P(p_{T})=\frac{4p_{T}}{M^{2}}\int_{0}^{\sqrt{p_{\max }^{2}-p_{T}^{2}}}\frac{1%
427: }{x^{2}}\left( \frac{2F_{2}(x)}{x}-F_{2}^{\prime }(x)\right) dp_{1};\qquad x=%
428: \frac{2\sqrt{p_{1}^{2}+p_{T}^{2}}}{M}. \label{gp16}
429: \end{equation}%
430: In Fig. \ref{fgr1} \ the structure function $F_{2}$ together with the
431: corresponding distributions calculated with the use of relations (\ref{gp9})
432: and (\ref{gp16}) are displayed.
433: \begin{figure}[tbp]
434: \includegraphics[width=18cm]{gmo.eps}
435: \caption{Proton structure function $F_{2}$ at $Q^{2}=4GeV^{2}$ (left).
436: Corresponding calculation of the quark momentum distributions in the proton
437: rest frame: the $p$ and $p_{T}$ distributions are represented \ by solid and
438: dashed lines (right).}
439: \label{fgr1}
440: \end{figure}
441: \ For the proton structure function the phenomenological fit performed in
442: the Appendix of the paper \cite{adeva} was used. Using the Eq. (\ref{gp8})
443: one can calculate the mean value%
444: \begin{equation}
445: \left\langle p\right\rangle =\frac{\int pG(p)d^{3}p}{\int G(p)d^{3}p}=\frac{M%
446: }{2}\frac{\int_{0}^{1}x^{3}\left( \frac{F_{2}(x)}{x^{2}}\right) ^{\prime }dx%
447: }{\int_{0}^{1}x^{2}\left( \frac{F_{2}(x)}{x^{2}}\right) ^{\prime }dx},
448: \label{gp16a}
449: \end{equation}%
450: but since the extrapolation of the structure function for $x\rightarrow 0$
451: gives in the denominator divergent integral, it follows that $\left\langle
452: p\right\rangle \rightarrow 0$. Nontrivial value can be obtained with the
453: integration cutoff $x>x_{\min }$.
454:
455: \subsection{Momentum distribution from structure function $g_{1}$}
456:
457: Now, we shall determine the distribution $\Delta G$ defined in Eq. (\ref{sp9}%
458: ) from the spin function $g_{1}$\ - similarly as distribution $G$ was
459: obtained from the function $F_{2}$. In the paper \cite{zav2}, Eq. (44), we
460: proved that%
461: \begin{equation}
462: g_{1}(x)=V_{0}(x)-\int_{x}^{1}\left( 4\frac{x^{2}}{y^{3}}-\frac{x}{y^{2}}%
463: \right) V_{0}(y)dy, \label{gp17}
464: \end{equation}%
465: where the function $V_{0}$ is defined by Eq. (\ref{gp1}) for $n=0$ and $%
466: K(p)= $ $\Delta G(p)$. In the Appendix \ref{app1} it is shown, that the last
467: relation can be modified to:%
468: \begin{equation}
469: V_{-1}(x)=\frac{2}{x}\left( g_{1}(x)+2\int_{x}^{1}\frac{g_{1}(y)}{y}%
470: )dy\right) . \label{GP19}
471: \end{equation}%
472: Then, in an accordance with Eq. (\ref{gp3}), we obtain
473: \begin{equation}
474: V_{-1}^{\prime }(x)=-\pi M^{3}\Delta G(p);\qquad x=\frac{2p}{M},
475: \label{gp20}
476: \end{equation}%
477: so the last two relations imply
478: \begin{equation}
479: \Delta G(p)=\frac{2}{\pi M^{3}x^{2}}\left( 3g_{1}(x)+2\int_{x}^{1}\frac{%
480: g_{1}(y)}{y})dy-xg_{1}^{\prime }(x)\right) ;\qquad x=\frac{2p}{M}.
481: \label{gp21}
482: \end{equation}%
483: Obviously this distribution together with the distribution (\ref{gp8})
484: allows to obtain the polarized distributions $G^{\pm }$ as%
485: \begin{equation}
486: G^{\pm }(p)=\sum_{q}e_{q}^{2}G_{q}^{\pm }(p)=\frac{1}{2}\left( G(p)\pm
487: \Delta G(p)\right) . \label{gp22}
488: \end{equation}%
489: These distributions can be with the use of Eqs. (\ref{gp8}) and (\ref{gp21})
490: obtained from experimental data on the $F_{2}$ and \ $g_{1}$.The result is
491: displayed in the left part of Fig. \ref{fgr2}, for which the function $g_{1}$
492: was parameterized by the fit of the world data \cite{E155} at $%
493: Q^{2}=4GeV^{2} $.
494: \begin{figure}[tbp]
495: \includegraphics[width=18cm]{ggg.eps}
496: \caption{Probability distributions $\Delta G,G,G^{+}$ and $G^{-}$ are
497: represented by solid, dashed, dash-and-dot and dotted lines (left) and
498: corresponding structure functions $\Delta F_{2},F_{2},F_{2}^{+}$ and $%
499: F_{2}^{-}$ \ (right).}
500: \label{fgr2}
501: \end{figure}
502:
503: With the use of \ the relation (\ref{sp9a}) one can formally calculate the
504: partial structure functions corresponding to the subsets of positively and
505: negatively polarized quarks:%
506: \begin{equation}
507: F_{2}^{\pm }(x)=Mx^{2}\int G^{\pm }(p)\delta \left( \frac{p_{0}+p_{1}}{M}%
508: -x\right) \frac{d^{3}p}{p_{0}}. \label{gp25bAD}
509: \end{equation}%
510: Apparently it holds%
511: \begin{equation}
512: F_{2}(x)=F_{2}^{+}(x)+F_{2}^{-}(x) \label{gad1}
513: \end{equation}%
514: and one can define also%
515: \begin{equation}
516: \Delta F_{2}(x)=F_{2}^{+}(x)-F_{2}^{-}(x) \label{gad2}
517: \end{equation}%
518: or equivalently%
519: \begin{equation}
520: \Delta F_{2}(x)=Mx^{2}\int \Delta G(p)\delta \left( \frac{p_{0}+p_{1}}{M}%
521: -x\right) \frac{d^{3}p}{p_{0}}. \label{gp23}
522: \end{equation}%
523: This equality can be written as%
524: \begin{equation}
525: \Delta F_{2}(x)=x^{2}V_{-1}(x), \label{gp24}
526: \end{equation}%
527: which after inserting from Eq. (\ref{GP19}) gives%
528: \begin{equation}
529: \Delta F_{2}(x)=2x\left( g_{1}(x)+2\int_{x}^{1}\frac{g_{1}(y)}{y})dy\right) .
530: \label{gp25}
531: \end{equation}%
532: The structure functions $F_{2},\Delta F_{2},F_{2}^{\pm }$ are shown in the
533: right part of Fig. \ref{fgr2} and one can observe:
534:
535: \textit{i)} Shape of the function $\Delta F_{2}$ implies, that contributions
536: of oppositely polarized quarks in Eq. (\ref{gad2}) are canceled out in the
537: region of low $x$. In fact the shape is similar to that of the function $%
538: F_{2val}$ corresponding to the valence quarks. This suggests that a dominant
539: spin contribution comes from the valence region.
540:
541: \textit{ii)} The distributions $\Delta G,G$ and $G^{+}$ are very close
542: together in the region of higher momenta and simultaneously $G^{-}$ is close
543: to zero in the same region. And the same holds for corresponding structure
544: functions $\Delta F_{2},F_{2}$ and $F_{2}^{\pm }$ in the higher $x$ region.
545: In an accordance with the definitions (\ref{sp8a}),(\ref{sp9}) it follows,
546: that%
547: \begin{equation}
548: G_{q}^{+}(p)\simeq G_{q}(p),\qquad G_{q}^{-}(p)\simeq 0. \label{gp29}
549: \end{equation}%
550: In the other words, polarization of quarks or partons with higher intrinsic
551: energy (and/or higher $x$) coincides with the proton polarization.
552:
553: The mean value of the distribution $\Delta G$ can be estimated as%
554: \begin{equation}
555: \left\langle p\right\rangle =\frac{\int p\Delta G(p)d^{3}p}{\int \Delta
556: G(p)d^{3}p}=\frac{M}{2}\left\langle x\right\rangle ;\qquad \left\langle
557: x\right\rangle =\frac{\int_{0}^{1}xg_{1}(x)dx}{\int_{0}^{1}g_{1}(x)dx}.
558: \label{GP26}
559: \end{equation}%
560: The proof of this relation is done in the Appendix \ref{app2}. The numerical
561: calculation with the $g_{1}$\ fit gives $\left\langle p\right\rangle
562: =0.113GeV/c$. However, interpretation of this average momentum should be
563: done with some care since $\Delta G_{q}$ can be negative, in principle.
564: Average momentum $\left\langle p\right\rangle $ allows to calculate the mean
565: transversal momentum; $\left\langle p_{T}\right\rangle =\pi /4\cdot
566: \left\langle p\right\rangle $.
567:
568: \section{Intrinsic quark motion and orbital momentum}
569:
570: \label{sec3}The rule of quantum mechanics says, that angular momentum
571: consists of the orbital and spin part \ $\mathbf{j=l+s}$\ and that in the
572: relativistic case the $\mathbf{l}$ and $\mathbf{s}$\ are not conserved
573: separately, but only the total angular momentum $j$\ is conserved.\textit{\ }%
574: This simple fact was in the context of quarks inside the nucleon pointed out
575: in \cite{liang}. It means, that only $j^{2},j_{z}$ are well-defined quantum
576: numbers and corresponding states of the particle with spin $1/2$ are
577: represented by the bispinor spherical waves \cite{lali}%
578: \begin{equation}
579: \psi _{kjlj_{z}}\left( \mathbf{p}\right) =\frac{\delta (p-k)}{p\sqrt{2p_{0}}}%
580: \left(
581: \begin{array}{c}
582: i^{-l}\sqrt{p_{0}+m}\Omega _{jlj_{z}}\left( \mathbf{\omega }\right) \\
583: i^{-\lambda }\sqrt{p_{0}-m}\Omega _{j\lambda j_{z}}\left( \mathbf{\omega }%
584: \right)%
585: \end{array}%
586: \right) , \label{sp19}
587: \end{equation}%
588: where $\mathbf{\omega }=\mathbf{p}/p,$\ $l=j\pm \frac{1}{2},\ \lambda =2j-l$
589: ($l$ defines the parity) and
590: \begin{eqnarray*}
591: \Omega _{j,l,j_{z}}\left( \mathbf{\omega }\right) &=&\left(
592: \begin{array}{c}
593: \sqrt{\frac{j+j_{z}}{2j}}Y_{l,j_{z}-1/2}\left( \mathbf{\omega }\right) \\
594: \sqrt{\frac{j-j_{z}}{2j}}Y_{l,j_{z}+1/2}\left( \mathbf{\omega }\right)%
595: \end{array}%
596: \right) ;\quad l=j-\frac{1}{2}, \\
597: \Omega _{j,l,j_{z}}\left( \mathbf{\omega }\right) &=&\left(
598: \begin{array}{c}
599: -\sqrt{\frac{j-j_{z}+1}{2j+2}}Y_{l,j_{z}-1/2}\left( \mathbf{\omega }\right)
600: \\
601: \sqrt{\frac{j+j_{z}+1}{2j+2}}Y_{l,j_{z}+1/2}\left( \mathbf{\omega }\right)%
602: \end{array}%
603: \right) ;\quad l=j+\frac{1}{2}.
604: \end{eqnarray*}%
605: States are normalized as:%
606: \begin{equation}
607: \int \psi _{k^{\prime }j^{\prime }l^{\prime }j_{z}^{\prime }}^{\dagger
608: }\left( \mathbf{p}\right) \psi _{kjlj_{z}}\left( \mathbf{p}\right)
609: d^{3}p=\delta (k-k^{\prime })\delta _{jj^{\prime }}\delta _{ll^{\prime
610: }}\delta _{j_{z}j_{z}^{\prime }}. \label{sp20}
611: \end{equation}%
612: The wavefunction (\ref{sp19}) is simplified for $j=j_{z}=1/2$ and $l=0$.
613: Taking into account that%
614: \[
615: Y_{00}=\frac{1}{\sqrt{4\pi }},\qquad Y_{10}=i\sqrt{\frac{3}{4\pi }}\cos
616: \theta ,\qquad Y_{11}=-i\sqrt{\frac{3}{8\pi }}\sin \theta \exp \left(
617: i\varphi \right) ,
618: \]%
619: one gets:%
620: \begin{equation}
621: \psi _{kjlj_{z}}\left( \mathbf{p}\right) =\frac{\delta (p-k)}{p\sqrt{8\pi
622: p_{0}}}\left(
623: \begin{array}{c}
624: \sqrt{p_{0}+m}\left(
625: \begin{array}{c}
626: 1 \\
627: 0%
628: \end{array}%
629: \right) \\
630: -\sqrt{p_{0}-m}\left(
631: \begin{array}{c}
632: \cos \theta \\
633: \sin \theta \exp \left( i\varphi \right)%
634: \end{array}%
635: \right)%
636: \end{array}%
637: \right) . \label{sp22}
638: \end{equation}%
639: Let us note, that $j=1/2$ is the minimum angular momentum for the particle
640: with spin $1/2.$ If one consider the quark state as a superposition%
641: \begin{equation}
642: \Psi \left( \mathbf{p}\right) =\int a_{k}\psi _{kjlj_{z}}\left( \mathbf{p}%
643: \right) dk;\quad \int a_{k}^{\star }a_{k}dk=1 \label{sp21}
644: \end{equation}%
645: then its average spin contribution to the total angular momentum reads:%
646: \begin{equation}
647: \left\langle s\right\rangle =\int \Psi ^{\dagger }\left( \mathbf{p}\right)
648: \Sigma _{z}\Psi \left( \mathbf{p}\right) d^{3}p;\qquad \Sigma _{z}=\frac{1}{2%
649: }\left(
650: \begin{array}{cc}
651: \sigma _{z} & \cdot \\
652: \cdot & \sigma _{z}%
653: \end{array}%
654: \right) . \label{sp26}
655: \end{equation}%
656: After inserting from Eqs. (\ref{sp22}), (\ref{sp21}) into (\ref{sp26}) one
657: gets
658:
659: \begin{equation}
660: \left\langle s\right\rangle =\int a_{p}^{\star }a_{p}\frac{\left(
661: p_{0}+m\right) +\left( p_{0}-m\right) \left( \cos ^{2}\theta -\sin
662: ^{2}\theta \right) }{16\pi p^{2}p_{0}}d^{3}p=\frac{1}{2}\int a_{p}^{\star
663: }a_{p}\left( \frac{1}{3}+\frac{2m}{3p_{0}}\right) dp. \label{sp24}
664: \end{equation}%
665: Since\ $j=1/2$, the last relation implies for the quark orbital momentum:%
666: \begin{equation}
667: \left\langle l\right\rangle =\frac{1}{3}\int a_{p}^{\star }a_{p}\left( 1-%
668: \frac{m}{p_{0}}\right) dp. \label{sp25}
669: \end{equation}%
670: It means that for quarks in the state $j=j_{z}=1/2$ there are the extreme
671: scenarios:
672:
673: \textit{i)} Massive and static quarks ($p_{0}=m$), which implies $%
674: \left\langle s\right\rangle =j=1/2$ and $\left\langle l\right\rangle =0$.\
675: This is evident, since without kinetic energy no orbital momentum can be
676: generated.
677:
678: \textit{ii)}\ Massless quarks $\left( m\ll p_{0}\right) $, which implies $%
679: \left\langle s\right\rangle =1/6$ and $\left\langle l\right\rangle =1/3$.
680:
681: Generally, for $p_{0}\geq m$, one gets $1/3\leq \left\langle s\right\rangle
682: /j\leq 1.$ In other words, for the states with $p_{0}>m$ part of the total
683: angular momentum $j=1/2$ is necessarily generated by the orbital momentum.
684: This is a consequence of quantum mechanics, and not a consequence of the
685: particular model. For instance, if one assumes the quark effective mass of
686: the order thousandths and momentum of the order of tenth of $GeV$, then the
687: second scenario is clearly preferred. Further, the mean kinetic energy
688: corresponding to the superposition (\ref{sp21}) reads%
689: \begin{equation}
690: \left\langle E_{kin}\right\rangle =\int a_{p}^{\star }a_{p}E_{kin}dp;\qquad
691: E_{kin}=p_{0}-m \label{gp30}
692: \end{equation}%
693: and at the same time the Eq. (\ref{sp25}) can be rewritten as%
694: \begin{equation}
695: \left\langle l\right\rangle =\frac{1}{3}\int a_{p}^{\star }a_{p}\frac{E_{kin}%
696: }{p_{0}}dp. \label{gp31}
697: \end{equation}%
698: It is evident, that for fixed $j=1/2$ both the quantities are in the proton
699: rest frame almost equivalent: more kinetic energy generates more orbital
700: momentum and vice versa.
701:
702: Further, the average spin part $\left\langle s\right\rangle $ of the total
703: angular momentum $j=1/2$ related to single quark according to Eq. (\ref{sp24}%
704: ) can be compared with the integral%
705: \begin{equation}
706: \Gamma _{1}=\int_{0}^{1}g_{1}(x)dx, \label{sp4}
707: \end{equation}%
708: which measures total quark spin contribution to the proton spin. For the $%
709: g_{1}$ from Eq. (\ref{sp10}) this integral reads%
710: \begin{equation}
711: \Gamma _{1}=\frac{1}{2}\int \Delta G(p_{0})\left( \frac{1}{3}+\frac{2m}{%
712: 3p_{0}}\right) d^{3}p. \label{sp17}
713: \end{equation}%
714: Dependence of both \ the integrals (\ref{sp24}) and (\ref{sp17}) on
715: intrinsic motion is controlled by the same term $\left( 1/3+2m/3p_{0}\right)
716: $, which in both the cases has origin in the covariant kinematics of the
717: particle with $s=1/2$. In fact, the \ procedures for calculation of these
718: integrals are based on the two different representations of the solutions of
719: \ Dirac equation: the plane waves (\ref{sp8}) and spherical waves (\ref{sp22}%
720: ). It is apparent that for the scenario of massless quarks $\left( m\ll
721: p_{0}\right) $, due to necessary presence of the orbital motion, both the
722: integrals $\Gamma _{1}$ and $\left\langle s\right\rangle $ will be roughly
723: three times less, than for the scenario of massive and static quarks $\left(
724: m\simeq p_{0}\right) $. We discussed this effect in the context of
725: experimental data in \cite{zav05}.
726:
727: What is the underlying physics behind the interplay between the spin and
728: orbital momentum? Actually, speaking about the spin of the particle
729: represented by the state (\ref{sp8}),\ one should take into account:
730:
731: \textit{a)} Definite projection of the spin in the direction $\mathbf{n}$ is
732: well-defined quantum number only for the particle at rest ($p=0$) or for the
733: particle moving in the the direction $\mathbf{n}$, i.e. $\mathbf{p/}p\mathbf{%
734: =\pm n}$. In these cases we have%
735: \begin{equation}
736: s=u^{\dagger }\left( \mathbf{p},\lambda \mathbf{n}\right) \mathbf{n\Sigma }%
737: u\left( \mathbf{p},\lambda \mathbf{n}\right) =\pm 1/2. \label{gp32}
738: \end{equation}
739:
740: \textit{b)} In other cases, as shown in the Appendix \ref{app3}, only
741: inequality%
742: \begin{equation}
743: \left\langle s\right\rangle =\left\vert u^{\dagger }\left( \mathbf{p}%
744: ,\lambda \mathbf{n}\right) \mathbf{n\Sigma }u\left( \mathbf{p},\lambda
745: \mathbf{n}\right) \right\vert <1/2 \label{GP33}
746: \end{equation}%
747: is satisfied. Roughly speaking, the result of measuring of the spin (of a
748: quark) depends on its momentum in the defined reference frame (proton rest
749: frame). This obvious effect acts also in the states, which are represented
750: by the superposition of the plane waves (\ref{sp8}) with different momenta $%
751: \mathbf{p}$ and resulting in $\left\langle \mathbf{p}\right\rangle =0$, but $%
752: \left\langle \mathbf{p}^{2}\right\rangle >0$. In \cite{zav1} we showed, that
753: averaging of the spin projection (\ref{GP33}) over the spherical momentum
754: distribution gives the result equivalent to (\ref{sp17}). The state (\ref%
755: {sp21}) can be also decomposed into plane waves having spherical momentum
756: distribution and the spin mean value given by Eq. (\ref{sp24}). Well-defined
757: quantum numbers $j=j_{z}=1/2$ imply, that the spin reduction due to
758: increasing intrinsic kinetic energy \ is compensated by the increasing
759: orbital momentum.
760:
761: \section{Summary and conclusion}
762:
763: We studied covariant version of the QPM with spherically symmetric
764: distributions of \ the quark momentum in the proton rest frame. The main
765: results obtained in this paper can be summarized as follows.
766:
767: 1) Relations between the structure functions $F_{2},g_{1}$ and corresponding
768: 3D quark momentum distributions $G^{\pm }(p)=G(p)\pm \Delta G(p)$ were
769: obtained. In this way the momentum distributions of positively and
770: negatively polarized quarks $G^{\pm }(p)=\sum_{q}e_{q}^{2}G_{q}^{\pm }(p)$
771: are calculated from the experimentally measured structure functions $F_{2}$
772: and $g_{1}$. At the same time the partial structure functions $F_{2}$
773: related to the subsets of quarks, which are described by the distributions $%
774: G^{\pm }$ and $\Delta G$, were obtained. The momentum distributions are
775: spherically symmetric, it follows that corresponding longitudinal and
776: transversal distributions are accessible as well. Results of the calculation
777: suggest:
778:
779: \textit{a)} Character of the function $\Delta F_{2}=F_{2}^{+}-F_{2}^{-}$ is
780: similar to that of the function $F_{2val}$. This suggests, that the dominant
781: contribution to the proton spin comes from the valence quarks.
782:
783: \textit{b)} Comparison of $\ $the functions $G^{\pm }$ with $G$ and/or $%
784: F_{2}^{\pm }$ with $F_{2}$ suggest, that in region of higher intrinsic
785: energy (and/or higher $x$) the partons with positive polarization strongly
786: dominate, whereas partons with negative polarization are present rather only
787: in the lower energy region.
788:
789: 2) We showed, that important role of the quark orbital momentum emerges as a
790: direct consequence of a covariant description. Since in relativistic case
791: only the total angular momentum \ $\mathbf{j=l+s}$\ is well-defined quantum
792: number, there arises some interplay between its spin and orbital parts. For
793: the quark in the state with definite projection $j_{z}=1/2$ in the proton
794: rest frame, as a result of this interplay, its spin part is reduced in favor
795: of the orbital one. The role of orbital motion increases with the rate of
796: quark intrinsic motion; for $p\gg m$\ its fraction reaches $\left\langle
797: l_{z}\right\rangle =2/6$ whereas $\left\langle s_{z}\right\rangle =1/6$
798: only. Simultaneously, this effect is truly reproduced also in the formalism
799: of\ structure functions.\bigskip
800:
801: \textbf{Acknowledgments}
802:
803: \textit{I would like to thank Anatoli Efremov and Oleg Teryaev for many
804: useful discussions and valuable comments.}
805:
806: \appendix
807:
808: \section{Structure functions in the approach of infinite momentum frame}
809:
810: \label{app0}The necessary condition for obtaining equalities (\ref{sp11a}) -
811: (\ref{sp11b}) is the covariant relation%
812: \begin{equation}
813: p_{\alpha }=yP_{\alpha }, \label{b2}
814: \end{equation}%
815: which implies%
816: \begin{equation}
817: m=yM \label{b3}
818: \end{equation}%
819: and $\mathbf{p}=0$ in the proton rest frame and $p_{T}=0$ in the IMF.
820:
821: For calculation of the integrals (\ref{sp9c}) and (\ref{sp9d}) in the IMF
822: approach one can substitute $p$ by $yP$ and $d^{3}p/p_{0}$ by $\pi
823: dp_{T}^{2}dy/y$. Then, after some algebra the structure functions (\ref{sp9b}%
824: ) read%
825: \begin{equation}
826: F_{1}(x)=\frac{1}{2}Mx\int G\left( yM\right) \delta \left( y-x\right) \pi
827: dp_{T}^{2}\frac{dy}{y},\qquad F_{2}(x)=Mx^{2}\int G\left( yM\right) \delta
828: \left( y-x\right) \pi dp_{T}^{2}\frac{dy}{y}. \label{b6}
829: \end{equation}%
830: Since the approximation (\ref{b2}) implies sharply peaked distribution at $%
831: p_{T}^{2}\rightarrow 0$, one can identify%
832: \begin{equation}
833: MG_{q}\left( yM\right) \pi dp_{T}^{2}=q(y) \label{b7}
834: \end{equation}%
835: and then the Eqs. (\ref{sp11a}) and (\ref{b6}) after integrating are
836: equivalent.
837:
838: In the same way the equalities (\ref{cra31})-(\ref{cr31}) can be modified.
839: Taking into account that $pS\rightarrow yPS=0$, one obtain%
840: \begin{equation}
841: g_{1}(x)=\frac{m}{2}\int \Delta G\left( yM\right) \delta \left( y-x\right)
842: \pi dp_{T}^{2}\frac{dy}{y},\qquad g_{2}(x)=0. \label{b8}
843: \end{equation}%
844: If we put%
845: \begin{equation}
846: M\Delta G_{q}\left( yM\right) \pi dp_{T}^{2}=\Delta q(y) \label{b9}
847: \end{equation}%
848: \ and take into account Eq. (\ref{b3}), then it is obvious, that the Eqs. (%
849: \ref{sp11b}) and (\ref{b8}) are equivalent.
850:
851: \section{Proof of the relation ({\protect\ref{GP19}})}
852:
853: \label{app1}In the paper \cite{zav2} we proved relation%
854: \begin{equation}
855: \frac{V_{j}^{\prime }(x)}{V_{k}^{\prime }(x)}=\left( \frac{x}{2}+\frac{%
856: x_{0}^{2}}{2x}\right) ^{j-k};\qquad x_{0}=\frac{m}{M}, \label{a1}
857: \end{equation}%
858: which for $m\rightarrow 0$ implies%
859: \begin{equation}
860: V_{0}(x)=\frac{1}{2}\left( xV_{-1}(x)+\int_{0}^{x}V_{-1}(y)dy\right) .
861: \label{a2}
862: \end{equation}%
863: After inserting $V_{0}$ from this relation to Eq. (\ref{gp17}) one gets%
864: \begin{eqnarray}
865: g_{1}(x) &=&\frac{1}{2}\left( xV_{-1}(x)+\int_{0}^{x}V_{-1}(y)dy\right)
866: \label{a3} \\
867: &&-2x^{2}\left( \int_{x}^{1}\frac{V_{-1}(y)}{y^{2}}dy+\int_{x}^{1}\frac{1}{%
868: y^{3}}\int_{y}^{1}V_{-1}(z)dzdy\right) \nonumber \\
869: &&+\frac{1}{2}x\left( \int_{x}^{1}\frac{V_{-1}(y)}{y}dy+\int_{x}^{1}\frac{1}{%
870: y^{2}}\int_{y}^{1}V_{-1}(z)dzdy\right) . \nonumber
871: \end{eqnarray}%
872: The double integrals can be reduced by integration by parts with the use of
873: \ formula%
874: \begin{equation}
875: \int_{x}^{1}a(y)\left( \int_{y}^{1}b(z)dz\right) dy=\int_{x}^{1}\left(
876: A(y)-A(x)\right) b(y)dy;\qquad A^{\prime }(x)=a(x), \label{a4}
877: \end{equation}%
878: then the relation (\ref{a3}) is simplified:%
879: \begin{equation}
880: g_{1}(x)=\frac{1}{2}xV_{-1}(x)-x^{2}\int_{x}^{1}\frac{V_{-1}(y)}{y^{2}}dy.
881: \label{a5}
882: \end{equation}%
883: In the next step we extract $V_{-1}$ from this relation. After the
884: substitution $V(x)=V_{-1}(x)/x$ the relation reads%
885: \begin{equation}
886: \frac{g_{1}(x)}{x^{2}}=\frac{1}{2}V(x)-\int_{x}^{1}\frac{V(y)}{y}dy,
887: \label{a6}
888: \end{equation}%
889: which implies the differential equation for $V(x)$:%
890: \begin{equation}
891: \frac{1}{2}V^{\prime }(x)+\frac{V(x)}{x}=\left( \frac{g_{1}(x)}{x^{2}}%
892: \right) ^{\prime }. \label{a7}
893: \end{equation}%
894: The corresponding homogeneous equation%
895: \begin{equation}
896: \frac{1}{2}V^{\prime }(x)+\frac{V(x)}{x}=0 \label{a8}
897: \end{equation}%
898: gives the solution%
899: \begin{equation}
900: V(x)=\frac{C}{x^{2}}, \label{a9}
901: \end{equation}%
902: which after inserting to Eq. (\ref{a7}) gives%
903: \begin{equation}
904: C^{\prime }(x)=2x^{2}\left( \frac{g_{1}(x)}{x^{2}}\right) ^{\prime }.
905: \label{a10}
906: \end{equation}%
907: After integration one easily gets the relation inverse to Eq. (\ref{a5}):%
908: \begin{equation}
909: V_{-1}(x)=\frac{2}{x}\left( g_{1}(x)+2\int_{x}^{1}\frac{g_{1}(y)}{y}%
910: dy\right) , \label{a11}
911: \end{equation}%
912: which coincides with Eq. (\ref{GP19}).
913:
914: \section{Proof of the relation ({\protect\ref{GP26}})}
915:
916: \label{app2}The relation (\ref{gp21}) implies%
917: \begin{equation}
918: \int \Delta G(p)d^{3}p=\int_{0}^{1}\left( 3g_{1}(x)+2\int_{x}^{1}\frac{%
919: g_{1}(y)}{y})dy-xg_{1}^{\prime }(x)\right) dx \label{a12}
920: \end{equation}%
921: and%
922: \begin{equation}
923: \int p\Delta G(p)d^{3}p=\frac{M}{2}\int_{0}^{1}\left(
924: 3xg_{1}(x)+2x\int_{x}^{1}\frac{g_{1}(y)}{y})dy-x^{2}g_{1}^{\prime
925: }(x)\right) dx;\qquad x=\frac{2p}{M}. \label{a13}
926: \end{equation}%
927: If one denotes%
928: \begin{equation}
929: \Gamma _{1}=\int_{0}^{1}g_{1}(x)dx,\qquad \Gamma
930: _{2}=\int_{0}^{1}xg_{1}(x)dx, \label{a14}
931: \end{equation}%
932: then integration by parts gives%
933: \begin{equation}
934: \int_{0}^{1}\int_{x}^{1}\frac{g_{1}(y)}{y})dydx=\Gamma _{1},\qquad
935: \int_{0}^{1}xg_{1}^{\prime }(x)dx=-\Gamma _{1} \label{a15}
936: \end{equation}%
937: and%
938: \begin{equation}
939: \int_{0}^{1}2x\int_{x}^{1}\frac{g_{1}(y)}{y})dydx=\Gamma _{2},\qquad
940: \int_{0}^{1}x^{2}g_{1}^{\prime }(x)dx=-2\Gamma _{2}. \label{a16}
941: \end{equation}%
942: Now, one can easily express the ratio%
943: \begin{equation}
944: \frac{\int p\Delta G(p)d^{3}p}{\int \Delta G(p)d^{3}p}=\frac{M}{2}\frac{%
945: \Gamma _{2}}{\Gamma _{1}}, \label{a17}
946: \end{equation}%
947: in this way the relation ({\ref{GP26}}) is proved.
948:
949: \section{Proof of the relation ({\protect\ref{GP33}})}
950:
951: \label{app3}With the use of rule%
952: \begin{equation}
953: \mathbf{p\sigma \cdot n\sigma }+\mathbf{n\sigma \cdot p\sigma =2pn}
954: \label{a18}
955: \end{equation}%
956: the term in Eq. ({\ref{GP33}}) can be modified as
957:
958: \begin{eqnarray}
959: u^{\dagger }\left( \mathbf{p},\lambda \mathbf{n}\right) \mathbf{n\Sigma }%
960: u\left( \mathbf{p},\lambda \mathbf{n}\right) &=&\frac{1}{2N}\phi _{\lambda
961: \mathbf{n}}^{\dagger }\left( \mathbf{n\sigma +}\frac{\mathbf{p}\mathbf{%
962: \sigma \cdot n\sigma \cdot p}\mathbf{\sigma }}{\left( p_{0}+m\right) ^{2}}%
963: \right) \phi _{\lambda \mathbf{n}} \label{a19} \\
964: &=&\frac{1}{2N}\phi _{\lambda \mathbf{n}}^{\dagger }\left( \mathbf{n\sigma +}%
965: \frac{\mathbf{p}\mathbf{\sigma \cdot }\left( -\mathbf{p\sigma \cdot n\sigma
966: +2pn}\right) }{\left( p_{0}+m\right) ^{2}}\right) \phi _{\lambda \mathbf{n}}
967: \nonumber \\
968: &=&\frac{1}{2N}\phi _{\lambda \mathbf{n}}^{\dagger }\left( \mathbf{n\sigma }%
969: \left( \mathbf{1-}\frac{\mathbf{p}^{2}}{\left( p_{0}+m\right) ^{2}}\right)
970: \mathbf{+}\frac{\mathbf{2pn\cdot p}\mathbf{\sigma }}{\left( p_{0}+m\right)
971: ^{2}}\right) \phi _{\lambda \mathbf{n}} \nonumber \\
972: &=&\frac{1}{2p_{0}}\phi _{\lambda \mathbf{n}}^{\dagger }\left( m\cdot
973: \mathbf{n\sigma +}\frac{\mathbf{pn\cdot p}\mathbf{\sigma }}{p_{0}+m}\right)
974: \phi _{\lambda \mathbf{n}}. \nonumber
975: \end{eqnarray}%
976: Since%
977: \begin{equation}
978: \left| \phi _{\lambda \mathbf{n}}^{\dagger }\mathbf{n\sigma }\phi _{\lambda
979: \mathbf{n}}\right| =1,\qquad \left| \phi _{\lambda \mathbf{n}}^{\dagger }%
980: \mathbf{p\sigma }\phi _{\lambda \mathbf{n}}\right| \leq p,\qquad \mathbf{pn=}%
981: p\mathbf{\cos \alpha ,} \label{a20}
982: \end{equation}%
983: it follows%
984: \begin{equation}
985: \left| u^{\dagger }\left( \mathbf{p},\lambda \mathbf{n}\right) \mathbf{%
986: n\Sigma }u\left( \mathbf{p},\lambda \mathbf{n}\right) \right| \leq \frac{1}{%
987: 2p_{0}}\left( m\mathbf{+}\frac{p^{2}}{p_{0}+m}\right) =\frac{1}{2}.
988: \label{a21}
989: \end{equation}%
990: Obviously
991: \begin{equation}
992: \left| u^{\dagger }\left( \mathbf{p},\lambda \mathbf{n}\right) \mathbf{%
993: n\Sigma }u\left( \mathbf{p},\lambda \mathbf{n}\right) \right| =\frac{1}{2}
994: \label{a22}
995: \end{equation}%
996: only for $\mathbf{p/}p\mathbf{=\pm n}$ or $p=0\mathbf{.}$
997:
998: \begin{thebibliography}{99}
999: \bibitem{e142} E142 Collaboration, P.L. Anthony \textit{et. al.}, Phys. Rev.
1000: D \textbf{54}, 6620 (1996).
1001:
1002: \bibitem{e154} E154 Collaboration, K. Abe \textit{et. al}., Phys. Rev. Lett.
1003: \textbf{79}, 26 (1997).
1004:
1005: \bibitem{her1} HERMES Collaboration, K. Ackerstaff \textit{et. al}., Phys.
1006: Lett. B\textbf{\ 404}, 383 (1997).
1007:
1008: \bibitem{her2} HERMES Collaboration, A. Airapetian \textit{et. al}., Phys.
1009: Lett. B\textbf{\ 442}, 484 (1998).
1010:
1011: \bibitem{adeva} Spin Muon Collaboration, B. Adeva \textit{et. al}., Phys.
1012: Rev. D\textbf{\ 58}, 112001 (1998).
1013:
1014: \bibitem{e143} E143 Collaboration, K. Abe \textit{et. al}., Phys. Rev. D
1015: \textbf{58}, 112003 (1998).
1016:
1017: \bibitem{E155} E155 Collaboration, P. Anthony \textit{et al.}, Phys. Lett. B
1018: \textbf{493}, 19 (2000); Phys. Lett. B \textbf{553}, 18 (2003).
1019:
1020: \bibitem{spin04} \textit{Proceedings \ of the 16th International Spin
1021: Physics Symposium - SPIN 2004}, Trieste, Italy, 10-16 October 2004, edited
1022: by Franco Bradamante \textit{et. al.} (World Scientific Publishing Co.Pte.
1023: Ltd.).
1024:
1025: \bibitem{efrem} M.~Anselmino, A.~Efremov, E.~Leader, Phys. Rep. \textbf{261}%
1026: , 1 (1995).
1027:
1028: \bibitem{zav0} P.~Zavada, Phys. Rev. D\textbf{\ 55}, 4290 (1997).
1029:
1030: \bibitem{zav1} P.~Zavada, Phys. Rev. D\textbf{\ 65}, 0054040 (2002).
1031:
1032: \bibitem{zav2} P.~Zavada, Phys. Rev. D\textbf{\ 67}, 014019 (2003).
1033:
1034: \bibitem{zav05} P.\ Zavada, arXiv: hep-ph/0511142.
1035:
1036: \bibitem{tra} A.V. Efremov, O.V. Teryaev and P. Zavada, Phys. Rev. D\textbf{%
1037: \ 70}, 054018 (2004), \ arXiv: hep-ph/0512034.
1038:
1039: \bibitem{sehgal} L.M. Sehgal, Phys. Rev. D\textbf{\ 10}, 1663 (1974); D%
1040: \textbf{\ 10}, 2016 (E) (1974).
1041:
1042: \bibitem{ratc} P.G. Ratcliffe, Phys. Lett. B \textbf{192}, 180 (1987).
1043:
1044: \bibitem{abbas} A. Abbas, J.Phys. G, Nucl. Part. Phys. \textbf{16}, L21-L26
1045: (1990); \textbf{15}, L73-L77 (1989).
1046:
1047: \bibitem{kep} J. Keppler and H.M. Hofmann, Phys. Rev. D\textbf{\ 51}, 3936
1048: (1995).
1049:
1050: \bibitem{casu} M. Casu and L.M. Sehgal, Phys. Rev. D\textbf{\ 55}, 2644
1051: (1996).
1052:
1053: \bibitem{bqma} B.-Q. Ma and I. Schmidt, Phys. Rev. D\textbf{\ 58}, 096008
1054: (1998).
1055:
1056: \bibitem{brod} S. J. Brodsky, Dae Sung Hwang, B.-Q. Ma and I. Schmidt,
1057: Nucl.Phys.B \textbf{593, }311 (2001).
1058:
1059: \bibitem{waka} M. Wakamatsu and T. Watabe, Phys. Rev. D\textbf{\ 62}, 054009
1060: (2000).
1061:
1062: \bibitem{waka1} M. Wakamatsu and Y. Nakakoji, Phys. Rev. D \textbf{74}
1063: 054006 (2006).
1064:
1065: \bibitem{song} X. Song, Int. J. Mod. Phys. A \textbf{16}, 3673 (2001).
1066:
1067: \bibitem{ji} X. Ji, J.-P. Ma and F. Yuan, Nucl. Phys. B \textbf{652}, 383
1068: (2003).
1069:
1070: \bibitem{liang} Z. Liang and T. Meng, Z. Phys. A \textbf{344}, 171 (1992).
1071:
1072: \bibitem{lali} L.D. Landau, E.M. Lifshitz \textit{et al.}, Quantum
1073: Electrodynamics (Course of Theoretical Physics, vol. 4), Elsevier Science
1074: Ltd., 1982.
1075: \end{thebibliography}
1076:
1077: \end{document}
1078: