Target

Page maintainer: Caro

This page is considered done. It been reviewed by Maru. There may be missing elements, but they are all flagged and the text has no errors.

Intro

Over the years, Hermes operated several types of gas targets internal to the HERA storage ring: a longitudinally polarized Helium-3 target in 1995, longitudinally polarized hydrogen or deuterium targets in 1996-2000 and a transversely polarized hydrogen target in 2002-2005. Unpolarized hydrogen and deuterium data were taken in all years 1995-2007, in 1996-2005 also unpolarized nuclear data. In 2006 and 2007, a target cell with significantly different dimensions was used during the Recoil running period.

Essentials

ABS, BRP, TGA, UGFS

Diagram of the HERMES polarized target: ABS, storage cell, BRP, TGA. Long Delta-q paper Fig.3.

The HERMES target consisted of gas from a polarized or an unpolarized source fed into a storage cell internal to the HERA storage ring. Click here to see a picture of the target from 1998.

The source of polarized atoms was an Atomic Beam Source (ABS). The ABS was based on the Stern-Gerlach effect. Neutral atomic hydrogen or deuterium was produced in a dissociator and was formed into a beam using a cooled nozzle, collimators and a series of differential pumping stations. A succession of magnetic sextupoles and radio-frequency (RF) fields are used to select one (or two) particular atomic hyperfine states that have a given nuclear polarization. A Breit-Rabi Polarimeter (BRP) was used to monitor the degree of polarization. The BRP worked essentially in reverse to the ABS. The atomic fraction was monitored by sampling the gas in the target cell using a Target Gas Analyzer (TGA). Courtesy: Long Delta-q paper

Unpolarized gas was provided by the Unpolarized Gas Feed System (UGFS).

Storage cells

The storage cell is an elliptical tube whose dimensions have been varied during the years:

1995 Longitudinally polarized Helium-3 target: 9.8mm (high) x 29.0mm (wide) x 400.0 mm (long)
05.1996-11.1999 Longitudinally polarized target: 9.8mm (high) x 29.0mm (wide) x 400.0 mm (long)
12.1999-08.2000 Longitudinally polarized target: 8.9mm (high) x 21.0mm (wide) x 400.0 mm (long)
2002-2005 Transversely polarized target: 8.9mm (high) x 21.0mm (wide) x 400.0 mm (long)
2006-2007 (Unpolarized) Target for the Recoil Detector period: 9.1mm (high) x 21.0mm (wide) x 150.0 mm (long)

Long and short target cells

From 1995-2005 ("pre-Recoil"), the Hermes target was operated with the long target cells situated in the Hermes coordinate system in such a way that gas was injected between $z=-18$ cm and $z=+18$ cm. The gas density distribution showed a triangular shape peaking at $z=0$ cm, reflected in the z-vertex distribution of tracks.
In 2006 and 2007 ("Recoil"), the short target cell with the Recoil detector surrounding it was installed. Gas was injected between $z=5$ cm and $z=+20$ cm. The triangular z-vertex distribution had its maximum at around 12.5 cm.
Note that the two different target cell geometries result in different Hermes acceptances for <=2005 and 2006/2007 data!
For the dimensions of the target cells see the Storage cells section.

Polarized target

Vector and tensor polarization

For a spin-1/2 target (like the proton), the z-component of the nuclear spin, $s_{z}$ , has two projections $m$ onto the z-axis, namely $m=+1/2$ and $m=-1/2$ (see left figure). Spin-1 particles (like the deuteron) have one further possibility to set their spin in the $m=0$ state (see right figure).

Spin 1/2 particle: Projections

m=\pm 1/2

of the spin z-component onto the z-axis. "n sign(m)" denotes the number of particles with spin quantum number sign(m)*

_m_

in the ensemble.

Spin 1 particle: Projections

m=\pm 1,0

of the spin z-component onto the z-axis. "n sign(m)" denotes the number of particles with spin quantum number sign(m)*

_m_

in the ensemble.

Spin 1 particle: Allowed range of vector and tensor polarizations.

For a spin-1/2 target, the vector polarization $P_{z}$ is defined for an ensemble of particles $n$ as:
$P_{z}={\frac {n^{+}-n^{-}}{n^{+}+n^{-}}},_P_{z}_\leqq 1$
and for a spin-1 target:
$P_{z}={\frac {n^{+}-n^{-}}{n^{+}+n^{-}+n^{0}}},_P_{z}_\leqq 1.$

Only for a spin-1 target, the tensor polarization $P_{zz}$ is defined as:
$P_{zz}={\frac {(n^{+}+n^{-})-2n^{0}}{n^{+}+n^{-}+n^{0}}},-2\leqq P_{zz}<1.$

How to get the target state

The state of the polarized target is contained in g1DAQ.iTargetBit: g1DAQ.iTargetBit=0x04 means vector plus and g1DAQ.iTargetBit=0x08 vector minus. If you want to distinguish between parallel and anti-parallel beam-target spin states (as is was e.g. done for the famous spin asymmetry Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle A_{__}} ), you need in addition the information of the Beam helicity.
See also this mail on the target state topic

Target flipping modes

Here's the overview over the target spin states and where the cryptic abbreviations in the logrun come from (the original message):

dps  - [d]euterium [p]olarized [s]tate injection
vdps - [v]ector   ...  Pz+ /Pz-
adps - [a]ll      ...  Pe+ /Pz+ /Pz- /Pzz+/Pzz-
edps - [e]???     ...  Pe+ /Pz+ /Pz-
rdps - [r]un mode ...  Pz+ /Pz- /Pzz+/Pzz-
sdps - [s}ingle   ...  n1  / n2 / n3 / n4 / n5 / n6
mdps - [m]oller   ...  n1  / n4
tdps - [t]ensor   ...  Pzz+/Pzz-

Polarization values and their uncertainties

People analyzing polarized target data must know that the uDSTs do not contain the best knowledge on the target polarization values. The target group has performed extensive studies on polarization data quality and provides this information, often giving average values for a complete data taking year
- How to get the target polarization values
- Maru's erratum of the target polarization values or uncertainties of 2000 and 2002, May 4, 2009
- Target polarisation values and their uncertainties used by the Deltaq analysis group for the 1996 to 2000 running periods (Internal note 03-005).
  The target polarization values for 1996, 1998 and 1999 were obtained by renormalizing the inclusive cross section asymmetry Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle \left.A_{1}\right.} to the results from the 1997 or 2000 data sets, for which the target polarizations are known to better precision.

Nonsense values (like -999) of the target polarization in the uDSTs can simply be ignored, i.e. do not discard the burst because of such a value as long as you use the average values provided by the experts. The same applies to the $\alpha$ values, which are already included in the polarization values provided by the target experts. See also this e-mail thread on this topic. The only information you directly take from the burst table is the one on the target state (i.e. usually vector plus or vector minus).

The only target data quality you need to perform on burst level is for polarized data: g1Quality.bTargetDQ, which contains information on the hardware performance and the status of the target. Activating badbit 16 in your badburstlist (Data Quality Masks) does the job, in addition performing a polarization sanity check. It is once more very much recommended to use the badburstlist. Note that there are other bits which you might want to activate in order to easily select a subset of data with respect to target polarization, but they concern strictly speaking not data quality (like bits 0, 10, 15, 26).

Hermes data structure, uDST counter and split bursts. The Hermes uDST format data are split into runs, bursts and events. Furthermore, a burst can be split in 3 subparts called records and labelled as uDST counters. Split bursts occur when the target just flips sign. Those records have the same burst number, but different uDST counters. The middle record has undefined target polarization, which is indicated by g1DAQ.iTargetBit=0x02. For a target-polarized analysis, these records must be discarded. Both information on burst number and uDST counter is available in the burstlists. Note that split bursts can also be present in target-unpolarized data (as pointed out by the color transparancy group in spring 2009). None of the records needs to be discarded in this case.

Gas Types

Compilation of polarized target gas info from the logrun (as it enters the uDSTs): g1Quality.iExpTarg

Compilation of unpolarized target gas info from UGFS valve settings: g1Unpol.iGasType, where g1Unpol.iGasType > 0 stands for unpolarized data

To select unpolarized data, rather use g1Unpol.iGasType than g1Quality.iExpTarg, see also Elke's mail on that

Target densities

The target areal density, i.e., the target density integrated over the cell length, for polarized running (ABS density) [Reference: 2005 target paper]:

1997 Longitudinally polarized hydrogen target: $0.7\cdot 10^{14}nucleons/cm^{2}$
2000 Longitudinally polarized deuterium target: Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle 2.1\cdot 10^{14}nucleons/cm^{2}}
2002 Transversely polarized hydrogen target: Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle 1.1\cdot 10^{14}nucleons/cm^{2}}
2006/2007 Unpolarized hydrogen target (with Recoil detector, low density 44% dosing valve): $4.9\cdot 10^{14}nucleons/cm^{2}$
For the target densities in other unpolarized running (high density and recoil), please refer to the UGFS-manual (also quoted below in #More Info)
Estimated Luminosity for nominal ABS density in case of a hydrogen target: Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle 2\cdot 10^{31}cm^{-2}s^{-1}}

More Info

Shift primers

Code Repository