1: %* For single column ``long'' format use this documentstyle and comment

2: %* out the two lines with ``wideabs'' in surrounding the title and 

3: %* abstract.

4: % ALL: switched to revtex4

5: %\documentclass[preprint,aps,prl]{revtex4}

6: \documentclass[twocolumn,aps,unsortedaddress,superscriptaddress,prl,floatfix,showpacs]{revtex4}

7: % ALL: epsfig package to replace epsf in revtex 3

8: \usepackage{epsfig}

9: 

10: %* For two column PRL format use this documentstyle and uncomment the 

11: %* the two lines with ``wideabs'' in surrounding the title and 

12: %* abstract.

13: %\documentstyle[twocolumn,aps,prl,floats,epsf]{revtex}

14: 

15: %\topmargin=-0.1in

16: \begin{document}

17: 

18: %* comment this line out for ``long'' draft format

19: %\wideabs{

20: 

21: \title{Measurement of Angular Distributions of Drell-Yan Dimuons

22: in $p + p$ Interactions at 800 GeV/c}

23: 

24: \affiliation{Abilene Christian University, Abilene, TX 79699}

25: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

26: \affiliation{Fermi National Accelerator Laboratory, Batavia, IL 60510}

27: \affiliation{Georgia State University, Atlanta, GA 30303}

28: \affiliation{Hampton University, Hampton, VA 23187}

29: \affiliation{Illinois Institute of Technology, Chicago, IL  60616}

30: \affiliation{University of Illinois at Urbana-Champaign, Urbana, IL 61801}

31: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

32: \affiliation{University of New Mexico, Albuquerque, NM 87131}

33: \affiliation{New Mexico State University, Las Cruces, NM 88003}

34: \affiliation{Oak Ridge National Laboratory, Oak Ridge, TN 37831}

35: \affiliation{Texas A\&M University, College Station, TX 77843}

36: \affiliation{Valparaiso University, Valparaiso, IN 46383}

37: 

38: \author{L.Y.~Zhu}

39: \affiliation{Hampton University, Hampton, VA 23187}

40: \affiliation{University of Illinois at Urbana-Champaign, Urbana, IL 61801}

41: 

42: \author{J.C.~Peng}

43: \affiliation{University of Illinois at Urbana-Champaign, Urbana, IL 61801}

44: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

45: 

46: \author{P.E.~Reimer}

47: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

48: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

49: 

50: \author{T.C.~Awes}

51: \affiliation{Oak Ridge National Laboratory, Oak Ridge, TN 37831}

52: 

53: \author{M.L.~Brooks}

54: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

55: 

56: \author{C.N.~Brown}

57: \affiliation{Fermi National Accelerator Laboratory, Batavia, IL 60510}

58: 

59: \author{J.D.~Bush}

60: \affiliation{Abilene Christian University, Abilene, TX 79699}

61: 

62: \author{T.A.~Carey}

63: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

64: 

65: \author{T.H.~Chang}

66: \affiliation{New Mexico State University, Las Cruces, NM 88003}

67: 

68: \author{W.E.~Cooper}

69: \affiliation{Fermi National Accelerator Laboratory, Batavia, IL 60510}

70: 

71: \author{C.A.~Gagliardi}

72: \affiliation{Texas A\&M University, College Station, TX 77843}

73: 

74: \author{G.T.~Garvey}

75: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

76: 

77: \author{D.F.~Geesaman}

78: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

79: 

80: \author{E.A.~Hawker}

81: \affiliation{Texas A\&M University, College Station, TX 77843}

82: 

83: \author{X.C.~He}

84: \affiliation{Georgia State University, Atlanta, GA 30303}

85: 

86: \author{L.D.~Isenhower}

87: \affiliation{Abilene Christian University, Abilene, TX 79699}

88: 

89: \author{D.M.~Kaplan}

90: \affiliation{Illinois Institute of Technology, Chicago, IL  60616}

91: 

92: \author{S.B.~Kaufman}

93: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

94: 

95: \author{S.A.~Klinksiek}

96: \affiliation{University of New Mexico, Albuquerque, NM 87131}

97: 

98: \author{D.D.~Koetke}

99: \affiliation{Valparaiso University, Valparaiso, IN 46383}

100: 

101: \author{D.M.~Lee}

102: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

103: 

104: \author{W.M.~Lee}

105: \affiliation{Fermi National Accelerator Laboratory, Batavia, IL 60510}

106: \affiliation{Georgia State University, Atlanta, GA 30303}

107: 

108: \author{M.J.~Leitch}

109: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

110: 

111: \author{N.~Makins}

112: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

113: \affiliation{University of Illinois at Urbana-Champaign, Urbana, IL 61801}

114: 

115: \author{P.L.~McGaughey}

116: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

117: 

118: \author{J.M.~Moss}

119: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

120: 

121: \author{B.A.~Mueller}

122: \affiliation{Physics Division, Argonne National Laboratory, Argonne, IL 60439}

123: 

124: \author{P.M.~Nord}

125: \affiliation{Valparaiso University, Valparaiso, IN 46383}

126: 

127: \author{V.~Papavassiliou}

128: \affiliation{New Mexico State University, Las Cruces, NM 88003}

129: 

130: \author{B.K.~Park}

131: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

132: 

133: \author{G.~Petitt}

134: \affiliation{Georgia State University, Atlanta, GA 30303}

135: 

136: \author{M.E.~Sadler}

137: \affiliation{Abilene Christian University, Abilene, TX 79699}

138: 

139: \author{W.E.~Sondheim}

140: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

141: 

142: \author{P.W.~Stankus}

143: \affiliation{Oak Ridge National Laboratory, Oak Ridge, TN 37831}

144: 

145: \author{T.N.~Thompson}

146: \affiliation{Los Alamos National Laboratory, Los Alamos, NM 87545}

147: 

148: \author{R.S.~Towell}

149: \affiliation{Abilene Christian University, Abilene, TX 79699}

150: 

151: \author{R.E.~Tribble}

152: \affiliation{Texas A\&M University, College Station, TX 77843}

153: 

154: \author{M.A.~Vasiliev}

155: \affiliation{Texas A\&M University, College Station, TX 77843}

156: 

157: \author{J.C.~Webb}

158: \affiliation{New Mexico State University, Las Cruces, NM 88003}

159: 

160: \author{J.L.~Willis}

161: \affiliation{Abilene Christian University, Abilene, TX 79699}

162: 

163: \author{D.K.~Wise}

164: \affiliation{Abilene Christian University, Abilene, TX 79699}

165: 

166: \author{G.R.~Young}

167: \affiliation{Oak Ridge National Laboratory, Oak Ridge, TN 37831}

168: 

169: \collaboration{FNAL E866/NuSea Collaboration}

170: \noaffiliation

171: 

172: \date{\today}

173: 

174: 

175: \begin{abstract}

176: We report a measurement of the angular distributions of Drell-Yan

177: dimuons produced using an 800 GeV/c proton beam on a hydrogen 

178: target. The polar and azimuthal angular distribution parameters

179: have been extracted over the kinematic 

180: range $4.5 < m_{\mu \mu} 

181: < 15$ GeV/c$^2$ (excluding the $\Upsilon$ resonance region),

182: $0 < p_T < 4 $ GeV/c, and $0 < x_F < 0.8$. The $p+p$ angular distributions

183: are similar to those of $p+d$, and both data sets are compared

184: with models which attribute the $\cos 2 \phi$ distribution 

185: either to the presence of the transverse-momentum-dependent Boer-Mulders 

186: structure function $h_1^\perp$ or to QCD effects. The data indicate

187: the need to include QCD effects before reliable information on

188: the Boer-Mulders function can be extracted. The validity of 

189: the Lam-Tung relation 

190: in $p+p$ Drell-Yan is also tested. 

191: 

192: \end{abstract} 

193: \pacs{13.85.Qk, 14.20.Dh, 24.85.+p, 13.88.+e}

194: 

195: \maketitle

196: The study of the transverse momentum dependent (TMD) parton 

197: distribution functions of the nucleon has received much attention 

198: in recent years as it provides new perspectives on the hadron

199: structure and QCD~\cite{barone02}. One of these TMD distribution

200: functions, first 

201: considered by Sivers~\cite{sivers90}, represents 

202: the correlation between the quark's transverse momentum, $k_\perp$, 

203: and the transverse spin of the nucleon, $S_\perp$. This so-called 

204: Sivers function, $f^\perp_{1T}(x,k^2_\perp)$, where $x$ is the

205: fraction of proton's momentum carried by the quark, is time-reversal odd

206: (T-odd) and can arise from initial- or 

207: final-state interactions~\cite{brodsky02}.

208: More generally, the requirement of gauge

209: invariance of parton distributions was shown to provide nontrivial 

210: phases leading to the existence of T-odd distribution 

211: functions~\cite{collins02,ji02}. Recent measurements of the semi-inclusive

212: deep-inelastic scattering (SIDIS) by the HERMES~\cite{hermes05} 

213: and COMPASS~\cite{compass05} collaborations

214: have shown clear evidence for the presence of the T-odd Sivers functions.

215: These data also allow the first determination~\cite{vogelsang05} of 

216: the magnitude and flavor structure of the Sivers functions.

217: 

218: Another T-odd distribution function is the Boer-Mulders function, 

219: $h^\perp_1(x,k^2_\perp)$, which signifies the correlation 

220: between $k_\perp$ and the quark transverse spin, $s_\perp$, in an unpolarized 

221: nucleon~\cite{boer98}. The Boer-Mulders function is the 

222: chiral-odd analog of the Sivers function and

223: also owes its existence to the presence of initial/final state 

224: interactions~\cite{boer03}. While the Sivers function is 

225: beginning to be quantitatively determined from

226: the SIDIS experiments, very little is known about the Boer-Mulders

227: function so far.

228: 

229: Several model calculations have been carried out for the 

230: Boer-Mulders functions. In the quark-diquark model, it was

231: shown that the Boer-Mulders functions are identical to

232: the Sivers functions when only the scalar diquark configuration 

233: is considered~\cite{boer03, gamberg03}. More recently, calculations 

234: taking into account both

235: the scalar and the axial-vector diquark configurations found significant

236: differences in flavor dependence between the Sivers and Boer-Mulders 

237: functions~\cite{gamberg08}.

238: In particular, the $u$ and $d$ valence quark Boer-Mulders 

239: functions are predicted

240: to be both negative, while the Sivers function is negative for the $u$ 

241: and positive for the $d$ valence quarks. Other calculations 

242: using the MIT bag model~\cite{yuan03}, the relativistic constituent 

243: quark model~\cite{pasquini07}, the large-$N_c$ model~\cite{pobylitsa03}, 

244: and lattice QCD~\cite{gockeler07} 

245: also predict negative signs for the $u$ and $d$ valence Boer-Mulders

246: functions. Burkardt recently pointed out~\cite{burkardt08} 

247: that the negative signs for the Boer-Mulders functions are

248: expected for both nucleons and pions. The model predictions

249: for the same signs of the $u$ and $d$ Boer-Mulders functions 

250: remain to be tested experimentally. Furthermore, the striking 

251: prediction~\cite{collins02} that the T-odd Boer-Mulders functions

252: in the SIDIS process will change their signs for the Drell-Yan

253: process also awaits experimental confirmation.

254: 

255: The Boer-Mulders functions can be extracted~\cite{boer99}

256: from the azimuthal angular distributions in the unpolarized Drell-Yan

257: process, $h_1 h_2 \to l^+ l^- x$. The general expression for the Drell-Yan

258: angular distribution is~\cite{lam78}

259: %\begin{widetext}

260: \begin{equation}

261: \frac {d\sigma} {d\Omega} \propto 1+\lambda \cos^2\theta +\mu \sin2\theta

262: \cos \phi + \frac {\nu}{2} \sin^2\theta \cos 2\phi,

263: \label{eq:eq1}

264: \end{equation}

265: \noindent where $\theta$ and $\phi$ are the polar and azimuthal decay angle

266: of the $l^+$ in the dilepton rest frame. Boer showed that the $\cos 2\phi$

267: term is proportional to the convolution of the quark and antiquark

268: Boer-Mulders functions in the projectile and target~\cite{boer99}.

269: This can be understood by noting that the Drell-Yan cross

270: section depends on the transverse spins of the annihilating quark and 

271: antiquark. Therefore, a correlation between the transverse spin and

272: the transverse momentum of the quark, as represented by the Boer-Mulders

273: function, would lead to a 

274: preferred transverse momentum direction. 

275: 

276: Pronounced $\cos 2 \phi$ dependences

277: were indeed observed in the NA10~\cite{falciano86} and E615~\cite{conway89} 

278: pion-induced Drell-Yan experiments, and attributed to the 

279: Boer-Mulders function.

280: The first measurement of the $\cos 2 \phi$

281: dependence of the proton-induced Drell-Yan process was recently reported for 

282: $p+d$ interactions

283: at 800 GeV/c~\cite{zhu07}. In contrast to pion-induced Drell-Yan,

284: significantly smaller (but non-zero) cos$2\phi$ azimuthal angular dependence 

285: was observed in the $p+d$ reaction. While the pion-induced Drell-Yan process

286: is dominated by annihilation between a valence antiquark in the pion

287: and a valence quark in the nucleon, the

288: proton-induced Drell-Yan process involves a valence quark in the proton 

289: annihilating with a sea antiquark in the nucleon. Therefore, the

290: $p+d$ result suggests~\cite{zhu07} that the Boer-Mulders functions for 

291: sea antiquarks are significantly smaller than those for valence quarks. 

292: 

293: A recent analysis~\cite{zhang08} indicated that the E866 $p+d$

294: data are consistent with the $u$ and $d$ Boer-Mulders

295: functions having the same signs, as predicted by various models. However, the 

296: $p+d$ data alone cannot provide an unambiguous determination of the 

297: flavor dependence of the Boer-Mulders functions. Moreover, it was 

298: recently pointed

299: out~\cite{boer06,berger07} that QCD processes

300: would lead 

301: to a sizeable $\cos 2\phi$

302: effect which has not been taken into account in the 

303: extractions~\cite{boer99,zhang08,zhang08a} of Boer-Mulders

304: functions from the Drell-Yan data. In

305: this paper we report the Drell-Yan angular distributions of the

306: $p+p$ reaction at 800 GeV/c, which provides further constraints on

307: the flavor dependence of the Boer-Mulders functions~\cite{zhang08a}. 

308: We also compare the

309: $\cos 2\phi$ dependences of $p+p$ and $p+d$ data with the prediction of 

310: QCD.

311: \begin{figure}[tb]

312: \includegraphics*[width=\linewidth]{pp_new_fig1.eps}

313: \caption{ Parameters $\lambda, \mu, \nu$ and $2\nu - (1-\lambda)$

314: vs.\ $p_T$ in the Collins-Soper frame. Solid squares (open circles) 

315: are for E866 $p+p$ ($p+d$) at 800 GeV/c. The vertical error bars 

316: include the statistical uncertainties only.}

317: \label{ppfig1}

318: \end{figure}

319: 

320: The Fermilab E866 experiment was performed using the upgraded Meson-East 

321: magnetic pair spectrometer~\cite{hawker}. An 800 GeV/c primary proton 

322: beam with up to $2 \times 10^{12}$ protons per 20 s beam spill 

323: was incident upon one of three identical 50.8 cm long target flasks 

324: containing either liquid hydrogen, 

325: liquid deuterium or vacuum. A copper beam dump located inside the 

326: second dipole

327: magnet (SM12) absorbed protons that passed through the target. Downstream

328: of the beam dump was an absorber wall that 

329: removed hadrons produced in the target and the beam dump.

330: 

331: \begin{table}[tbp]

332: \caption {Mean values of the $\lambda, \mu, \nu$ parameters and the quantity

333: $2\nu -(1-\lambda)$ for the $p+p$, $p+d$, and $\pi^- +W$ 

334: Drell-Yan measurements.}

335: \begin{center}

336: \begin{tabular}{|c|c|c|c|}

337: \hline

338: \hline

339:  & $p+p$ & $p+d$ & $\pi^- +W$ \\

340:  & 800 GeV/c & 800 GeV/c & 194 GeV/c \\

341:  & (E866) & (E866) & (NA10) \\

342: \hline

343: $\langle \lambda \rangle$ & $0.85 \pm 0.10$ & $1.07 \pm 0.07$ & 

344: $0.83 \pm 0.04$ \\

345: \hline

346: $\langle \mu \rangle$ & $-0.026 \pm 0.019$ & $0.003 \pm 0.013$ & 

347: $0.008 \pm 0.010$ \\

348: \hline

349: $\langle \nu \rangle$ & $0.040 \pm 0.015$ & $0.027 \pm 0.010$ & 

350: $0.091 \pm 0.009$ \\

351: \hline

352: $\langle 2 \nu - (1-\lambda) \rangle$ & $-0.07 \pm 0.10$ & 

353: $0.12 \pm 0.07$ & $0.01 \pm 0.04$ \\

354: \hline

355: \hline

356: \end{tabular}

357: \end{center}

358: \end{table}

359: 

360: Several settings of the

361: currents in the three dipole magnets (SM0, SM12, SM3)

362: were used in order to optimize acceptance for different dimuon mass regions.

363: Data collected with the ``low mass" and ``high mass" 

364: settings~\cite{hawker} on

365: liquid hydrogen and empty targets were used in this analysis.

366: The detector system consisted of

367: four tracking stations and a momentum analyzing magnet (SM3). 

368: Tracks reconstructed by the drift chambers were extrapolated to the target

369: using the momentum determined from the bend angle in SM3.

370: The target position was used to refine the parameters of each muon track. 

371: 

372: From the momenta of the $\mu^+$ and $\mu^-$, kinematic variables of

373: the dimuons ($x_F, m_{\mu\mu}$, and $p_T$, where $x_F$ is the fraction of the

374: c.m. momentum carried by dimuon of mass $m_{\mu\mu}$, and $p_T$

375: is the dimuon transverse momentum) were readily reconstructed.

376: The muon angles $\theta$ and $\phi$ in the Collins-Soper

377: frame~\cite{collins77} were also calculated. To eliminate the $J/\Psi$

378: and $\Upsilon$ resonance background, dimuon events with $m_{\mu\mu} < 4.5$

379: GeV/c$^2$ and 9.0 GeV/c$^2 < m_{\mu\mu} < 10.7$ GeV/c$^2$ were rejected in

380: the analysis.  

381: A total of $\approx$54,000

382: $p+p$ Drell-Yan events covering the decay angular range $-0.5 < \cos\theta

383: <0.5$ and $-\pi < \phi < \pi$ remain. Detailed Monte Carlo simulations

384: of the experiment using the MRST98 parton 

385: distribution functions~\cite{mrst} for

386: NLO Drell-Yan cross sections have shown good agreement with the data for 

387: a variety of measured quantities.

388: 

389: Figure 1 shows the angular distribution parameters $\lambda, \mu,$ and

390: $\nu$ vs.\ $p_T$. To extract these parameters, the Drell-Yan data were

391: grouped into 5 bins in $\cos\theta$ and 8 bins in $\phi$ for each $p_T$

392: bin. A least-squares fit

393: to the data using Eq.~1 to describe the angular distribution was

394: performed. The extracted values of $\lambda, \mu, \nu$ are

395: insensitive to their values used in the Monte Carlo simulation.

396: Only statistical errors are shown

397: in Fig.~1. The primary contributions to the systematic errors are the 

398: uncertainties of the incident beam angles on target. Analysis 

399: performed by 

400: allowing the beam angles to vary within their ranges of uncertainty

401: has shown that the

402: systematic errors are small compared to the statistical errors. 

403: The E866 $p+d$ Drell-Yan data are also shown in Fig.~1 for comparison

404: with the E866 $p+p$ data. The $p+d$ data contain a total of

405: $\approx$118,000 events covering an identical $\cos\theta$ range.

406: The $p_T$-averaged values of 

407: $\langle\lambda\rangle, \langle\mu\rangle,$ and 

408: $\langle\nu\rangle$ for $p+p$, $p+d$, and the 

409: NA10 $\pi^- + W$ data~\cite{falciano86} 

410: are listed in Table~I.

411: Within statistics, the angular distributions of $p+p$ are consistent

412: with those of $p+d$. Also shown in Fig.~1 and Table~I is the quantity 

413: $2\nu -(1-\lambda)$, which should vanish if the Lam-Tung relation is

414: valid. While QCD effects

415: can lead to $\lambda \ne 1$ and $\mu, \nu \ne 0$, Lam and Tung

416: showed~\cite{lam80} that the relation $1-\lambda = 2\nu$ is largely

417: unaffected by QCD corrections. 

418: Table~I shows that while $\langle\lambda\rangle$ deviates from 1 

419: and $\langle\nu\rangle$ is nonzero for

420: the E866 $p+p$ and the NA10 $\pi^- + W$ Drell-Yan data, the Lam-Tung relation

421: is indeed quite well satisfied within statistical uncertainty for all

422: $p_T$. This differs 

423: from the observation of

424: a significant violation of the Lam-Tung relation at large $p_T$ 

425: by the E615 collaboration

426: in the $\pi^- + W$ reaction at 252 GeV/c~\cite{conway89}.

427: 

428: Figure 2 shows the parameter $\nu$ vs.\ $p_T$ for the $p+p$ and

429: $p+d$ Drell-Yan data. The solid curves are

430: calculations~\cite{zhang08,zhang08a} for $p+p$ and $p+d$ using

431: parametrizations of the Boer-Mulders functions deduced from a fit 

432: to the $p+d$ Drell-Yan data.

433: The predicted larger values of $\nu$ for $p+p$ compared 

434: to $p+d$ in the region of $p_T \sim 1.5$ GeV/c are not 

435: observed (the predicted $p+p/p+d$ ratio, $R$, for

436: $0.5 < p_T < 2.0$ GeV/c, is $\sim 2$, while the data give

437: $R=1.0 \pm 0.5$). Furthermore, the shape of

438: the predicted $p_T$ dependence differs

439: from that of the data, resulting in a reduced $\chi^2$

440: value of 3.2 for 5 degrees of freedom (probability of 0.7\%). 

441: This strongly suggests that there could

442: be other mechanisms contributing to the $\cos 2\phi$ azimuthal angular

443: dependence at large $p_T$. In recent papers~\cite{boer06,berger07},

444: the QCD contribution

445: to the $\cos 2\phi$ azimuthal angular dependence is given as

446: \begin{equation}

447: \nu = \frac {Q^2_\perp/Q^2} {1+\frac{3}{2}Q^2_\perp/Q^2},

448: \label{eq:eq2}

449: \end{equation}

450: \noindent where $Q_\perp$ is the dimuon transverse momentum. The

451: predicted QCD contribution, the same for $p+p$ and $p+d$ due to the

452: identical kinematic coverage for the two reactions, 

453: is shown as the dot-dashed curve in Fig.~2. A comparison

454: between the QCD

455: prediction with the data gives

456: a reduced $\chi^2$ of 1.0 for 5 degrees of freedom (probability

457: of 42\%) for $p+p$ and a reduced $\chi^2$ of 1.9 (probability of

458: 9\%) for $p+d$. From Fig.~2 it is

459: evident that the QCD contribution is expected to become

460: more important at high $p_T$ while the

461: Boer-Mulders functions contribute primarily at lower $p_T$.

462: An analysis combining both effects is 

463: required in order to extract reliably the

464: Boer-Mulders functions from the $p+p$ and $p+d$ data.

465: It is worth noting that the $\pi^- +W$ Drell-Yan 

466: data~\cite{falciano86,conway89}

467: also show large values of $\nu$ at large $p_T$, consistent with 

468: the presence of QCD effects.

469: 

470: \begin{figure}[tb]

471: \includegraphics*[width=\linewidth]{pp_new_fig2.eps}

472: \caption{(color online). Parameter $\nu$ vs.\ $p_T$ in the Collins-Soper

473: frame for the $p+p$ and $p+d$ Drell-Yan data. 

474: The solid and dotted curves are

475: calculations~\cite{zhang08} for $p+p$ and $p+d$, respectively, 

476: using parametrizations

477: based on a fit to the $p+d$ data. The dot-dashed curve is the

478: contribution from the QCD process (Eq. 2).}

479: \label{ppfig2}

480: \end{figure}

481: 

482: The $p+p$ Drell-Yan angular distributions have also been analyzed for other

483: kinematic variables. Figure 3 shows the values 

484: of $\nu$ vs.~$m_{\mu\mu}, x_F, x_1,$ and $x_2$, where $x_1$ 

485: and $x_2$ are the

486: Bjorken-$x$ for the beam and target partons, respectively. Again, for

487: each bin the data were divided into 5 bins in $\cos\theta$ and 8 bins in

488: $\phi$ in order to extract the angular distribution parameters. 

489: The $p+d$ data are also shown for comparison. Figure 3 shows

490: that the magnitude of $\nu$ for $p+p$ is consistent with that for $p+d$ for

491: most of the kinematic regimes. These data provide further input for

492: future extraction of the Boer-Mulders functions.

493: 

494: \begin{figure}[tb]

495: \includegraphics*[width=\linewidth]{pp_new_fig3.eps}

496: \caption{Parameter $\nu$ vs.\ $m_{\mu\mu}$, $x_F$, $x_1$, 

497: and $x_2$ in the 

498: Collins-Soper frame for $p+p$ (solid squares) and 

499: $p+d$ (open circles) at 800 GeV/c. 

500: The vertical error bars correspond

501: to the statistical uncertainties only.}

502: \label{ppfig3}

503: \end{figure}

504: 

505: In summary, we report a measurement of the angular distributions of

506: Drell-Yan dimuons for $p+p$ at 800 GeV/c. The pronounced $\cos 2 \phi$

507: azimuthal angular dependence observed previously in pion-induced Drell-Yan

508: is not observed in the $p+p$ reaction. The Lam-Tung relation remains  

509: valid for the $p+p$ Drell-Yan data. The overall magnitude of the $\cos 2 \phi$

510: dependence for $p+p$ is consistent with, but slightly larger than 

511: that of $p+d$. The data suggest the presence

512: of higher-order QCD corrections at high $p_T$, and it is important to

513: take this contribution into account before reliable extraction of the 

514: Boer-Mulders functions could be obtained. 

515: 

516: We acknowledge helpful discussion with Bo-Qiang Ma, Bing Zhang, Matthias

517: Burkardt, Feng Yuan, Werner Vogelsang, and Jianwei Qiu.

518: This work was supported in part by the U.S. Department of Energy and 

519: the National Science Foundation.

520: \begin{thebibliography}{40}

521: \bibitem{barone02} V. Barone, A. Drago, and P. G. Ratcliffe, 

522: Phys. Rep. {\bf 359}, 1 (2002).

523: \bibitem{sivers90} D. Sivers, Phys. Rev. D {\bf 41}, 83 (1990).

524: \bibitem{brodsky02} S.J. Brodsky, D.S. Hwang, and I. Schmidt, Phys.

525:     Lett. B {\bf 530}, 99 (2002).

526: \bibitem{collins02} J.C. Collins, Phys. Lett. B {\bf 536}, 43 (2002).

527: \bibitem{ji02} X. Ji and F. Yuan, Phys. Lett. B {\bf 543}, 66 (2002).

528: \bibitem{hermes05} HERMES Collaboration, A. Airapetian {\em et al.}, 

529: Phys. Rev. Lett. {\bf 94}, 012002 (2005); M. Diefenthaler, arXiv: 0706.2242.

530: \bibitem{compass05} COMPASS Collaboration, V. Yu. Alexakhin {\em et al.},

531: Phys. Rev. Lett. {\bf 94}, 202002 (2005); M. Alekseev {\em et al.},

532: arXiv: 0802.2160.

533: \bibitem{vogelsang05} W. Vogelsang and F. Yuan, Phys. Rev. D {\bf 72},

534: 054028 (2005); M. Anselmino {\em et al.}, Phys. Rev. D {\bf 72}, 

535: 094007 (2005); M. Anselmino {\em et al.}, arXiv:0807.0166.

536: \bibitem{boer98} D. Boer and P.J. Mulders, Phys. Rev. D {\bf 57},

537: 5780 (1998).

538: \bibitem{boer03} D. Boer, S.J. Brodsky, and D.S. Hwang, Phys. Rev.

539: D {\bf 67}, 054003 (2003). 

540: \bibitem{gamberg03} L.P. Gamberg, G.R. Goldstein, and K.A. Oganessyan,

541: Phys. Rev. D {\bf 67}, 071504(R) (2003); A. Bacchetta, 

542: A. Sch\"{a}fer, and J.-J. Yang, Phys. Lett. B {\bf 578}, 109 (2004).

543: \bibitem{gamberg08} L.P. Gamberg, G.R. Goldstein, and M. Schlegel,

544: Phys. Rev. D {\bf 77}, 094016 (2008).

545: \bibitem{yuan03} F. Yuan, Phys. Lett. B {\bf 575}, 45 (2003).

546: \bibitem{pasquini07} B. Pasquini, M. Pincetti, and S. Boffi,

547: Phys. Rev. D {\bf 76}, 034020 (2007).

548: \bibitem{pobylitsa03} P.V. Pobylitsa, hep-ph/0301236.

549: \bibitem{gockeler07} M. G\"{o}ckeler {\em et al.}, Phys. Rev. Lett.

550: {\bf 98}, 222001 (2007).

551: \bibitem{burkardt08} M. Burkardt and B. Hannafious, Phys. Lett. B

552: {\bf 658}, 130 (2008).

553: \bibitem{boer99} D. Boer, Phys. Rev. D {\bf 60}, 014012 (1999).

554: \bibitem{lam78} C.S. Lam and W.K. Tung, Phys. Rev. D {\bf 18}, 2447

555: (1978).

556: \bibitem{falciano86} NA10 Collaboration, S. Falciano {\em et al.}, Z. Phys.

557: C {\bf 31}, 513 (1986); M. Guanziroli {\em et al.},

558: Z. Phys. C {\bf 37}, 545 (1988).

559: \bibitem{conway89} E615 Collaboration, J.S. Conway {\em et al.}, Phys.

560: Rev. D {\bf 39}, 92 (1989).		

561: \bibitem{zhu07} E866 Collaboration, L.Y. Zhu {\em et al.}, Phys.

562: Rev. Lett. {\bf 99}, 082301 (2007).

563: \bibitem{zhang08} B. Zhang, Z. Lu, B.-Q. Ma, and I. Schmidt,

564: Phys. Rev. D {\bf 77}, 054011 (2008).

565: \bibitem{boer06} D. Boer and W. Vogelsang, Phys. Rev. D {\bf 74}, 014004 (2006).

566: \bibitem{berger07} E.L. Berger, J.-W. Qiu, and R.A. Rodriguez-Pedraza,

567: Phys. Lett. B {\bf 656}, 74 (2007); Phys. Rev. D {\bf 76}, 074006 (2007).

568: \bibitem{zhang08a} B. Zhang, Z. Lu, B.-Q. Ma, and I. Schmidt,

569: arXiv: 0807.0503.

570: \bibitem{hawker} E866 Collaboration, E.H. Hawker {\em et al.},

571: Phys. Rev. Lett. {\bf 80}, 3715 (1998); J.C. Peng {\em et al.},

572: Phys. Rev. D {\bf 58}, 092004 (1998); R.S. Towell {\em et al.},

573: Phys. Rev. D {\bf 64}, 052002 (2001).

574: \bibitem{collins77} J.C. Collins and D.E. Soper, Phys. Rev. D {\bf 16},

575: 2219 (1977).

576: \bibitem{mrst} A.D. Martin, R.G. Roberts, W.J. Stirling, and R.S. Thorne,

577: Eur. Phys. J. C {\bf 4}, 463 (1998).

578: \bibitem{lam80} C.S. Lam and W.K. Tung, Phys. Rev. D {\bf 21}, 2712 (1980).

579: \end{thebibliography}

580: 

581: \end{document}

582: