| My scientific career started with my Ph.D. which I defended in September 1998 at Saclay.
The subject was the study of the nucleon spin structure functions measured in the E154
and E155 experiments. Those two experiments took place at the End Station A of SLAC
(Stanford Linear Accelerator Center) at Stanford (CA) USA, where the quest for the understanding
of the partonic structure of the nucleon started, with the Nobel experiment by
Friedman, Kendall and Taylor [1]. The main goal of E154 and E155 was to measure the spin
structure functions of the neutron, proton and deuteron by deep inelastic scattering (DIS)
of electrons up to 50 GeV on polarized gaseous or solid targets. By measuring double spin
asymmetries of counting rates, one can extract structure functions, from which polarized
parton densities can be obtained through a Quantum Chromodynamics (QCD) analysis. A
remarkable result is the determination of the spin carried by quarks in the nucleon, which
only amounts to about 30% [2]. This rather low and thus unexpected value motivated
the next generation of such experiments, trying to solve the spin crisis studying the gluon
content of the nucleon (COMPASS, HERMES).
After my thesis in 1998, I did my national civil duty for two years as a post-doctoral
fellow at Old Dominion University (VA) USA working at the Jefferson Laboratory on two
closely related topics: Real Compton Scattering and Deeply Virtual Compton Scattering.
I first helped putting together the Real Compton Scattering (RCS) experiment in Hall A
which followed a novel idea from Radyushkin and others [3, 4] that for RCS at wide angles,
a factorization scheme allows one to separate a hard perturbative kernel from a soft contribution
related to the structure of the nucleon, the subject of our interest. Then, at the
end of 1999, I had the opportunity to co-write the very first proposal to measure Deeply
Virtual Compton Scattering (DVCS) - γ∗p → γp - in the Hall A of Jefferson Lab. In the
hard scattering limit, this process factorizes as well and allows one to study the structure
of the nucleon in a very original way, as we will see later on. The proposal was approved
with A rating by the Jefferson Lab PAC in 2000 for 600 hours of running time, and the
corresponding experiment E00-110 ran in 2004 after a few years of detector and electronics
developments.
I was hired at the CEA Saclay DAPNIA/SPhN in the fall of 2000, right after my national
civil duty. This staff position corresponded to my interests: the study of hard exclusive
6
reactions at Jefferson Lab. The first programmed experiment was E1-6 in Hall B, planning
to measure the deep electroproduction of ρ0 and ω vector mesons among other channels,
which ran in 2001-2002.
Let us go back to DVCS and why it is a particularily interesting process: barely more
than ten years ago, a unifying concept for the description of the nucleon structure was
introduced, now commonly known as Generalized Parton Distribution (GPD)[5–9]. These
functions contain the usual form factors and parton distributions, but in addition they
include correlations between states of different longitudinal and transverse momenta. GPDs
can therefore give three-dimensional pictures of the nucleon, providing information such
as the transverse spatial distribution as a function of the longitudinal momentum fraction
of the quarks. With a complete knowledge of these functions, it is possible to obtain for
instance, the high and low-momentum components of the nucleon for different flavors. The
holy grail of this type of measurements is to provide enough data to determine the total
angular momentum carried by quarks through Ji’s sum rule [7]. Deeply Virtual Compton
Scattering (DVCS) is the simplest and therefore cleanest process which can be described in
terms of GPDs. It is the first reaction which allowed unambiguous extraction of GPDs from
data and provided the cornerstone of their exploration at Jefferson Lab.
Four dedicated DVCS experiments have taken data at Jefferson Lab since the first theoretical
developments: the Hall A E00-110 experiment I previously mentioned, measured
helicity dependent cross sections and provided the best check so far of the Bjorken-type
scaling which is expected of DVCS in the factorization regime [10]. As we will see in this
document, one can now say with a fair degree of confidence that DVCS measurements at
Jefferson Lab energies are indeed relevant to the investigation of the GPDs in the valence
region. This experiment was followed in Hall A by a neutron DVCS experiment E03-106,
sensitive to different GPDs [11]. The Hall B E01-113 experiment took data 6 months later,
and is currently in the final stages of the analysis [12]. The main goal of that experiment was
to perform Beam Spin Asymmetry measurements in a large kinematic domain, scanning this
observable as a function of xB, t and Q2. A secondary goal, which is still ongoing, is to obtain
cross sections in the same kinematic region. Finally, the polarized target version of the
Hall B experiment, E05-114 has taken data in 2010 [13]. This particular DVCS experiment
is more sensitive to the so-called ”polarized” GPD.
In addition to these experiments which took data on DVCS, I took responsabilities in writ7
ing two proposals. The first one is E08-021, a 6 GeV proposal to measure DVCS on a transversely
polarized target, which is more sensitive to the (light-cone) helicity-flip GPD [14].
The second one is the proposal to redo the unpolarized and longitudinally polarized DVCS
experiments of Hall B after the Jefferson Lab 12 GeV upgrade, with better accuracy and
larger kinematical coverage [15].
This document will try and summarize the work done in about 10 years on the three
6 GeV DVCS experiments and two proposals I worked and had responsabilities on, after
a brief theoretical introduction, mostly to set the notations and concepts used later on. I
will then try and give insights on the questions people usually ask in experimental talks
about Generalized Parton Distributions : what did we really learn about DVCS and GPDs
from this first set of experiments at Jefferson Lab? What do we need to do to perform
better experiments in the future? I will then conclude and describe the new directions of
Generalized Parton Distribution studies I plan to work on in the next few years.
Finally, I would like to point out that all this work was done in collaboration with many
experimental and theory colleagues and could not have been possible without them. My
many thanks to all of them, with a special note to my colleagues from France. |
