- Email: [email protected]
Nuclear effects in deep inelastic muon scattering on deuterium and iron targets
Volume 189, number 4
PHYSICS LETTERSB
14 May 1987
NUCLEAR EFFECTS IN DEEP I N E L A S T I C M U O N S C A T T E R I N G O N D E U T E R I U M A N D IRON T A R G E T S BCDMS Collaboration A.C. B E N V E N U T I , D. BOLLINI, G. BRUNI, F.L. N A V A R R I A Dipartimento di Fisica dell'Universit?tand INFN, L40126 Bologna, Italy A. A R G E N T O 1, j. CVACH 2, K. D I E T E R S 3, L. P I E M O N T E S E 4, I. VERESS s p. ZAVADA 2 CERN, CH-1211 Geneva 23, Switzerland N.G. FADEEV, I.A. G O L U T V I N , M.Yu. K A Z A R I N O V , Yu.T. K I R Y U S H I N , V.G. K R I V O K H I Z H I N , V.V. K U K H T I N , W. L O H M A N N 3, S. N E M E C E K , P. REIMER, I.A. SAVIN, G.I. SMIRNOV, J. S T R A C H O T A 2, G. S U L T A N O V 6, p. T O D O R O V , A.G. V O L O D K O Joint Institutefor Nuclear Research, Dubna, 101000Moscow, USSR D. J A M N I K 7 R. K O P P 8, U. M E Y E R - B E R K H O U T , A. STAUDE, K.-M. T E I C H E R T , R. T I R L E R 9, R. VOSS, C. Z U P A N C I C Sektion Physik der Universitiit, D-8000 Munich, Fed. Rep. Germany ~o M. CRIBIER, J. FELTESSE, A. MILSZTAJN, A. O U R A O U , P. R I C H - H E N N I O N , Y. SACQUIN, G. SMADJA a n d M. V I R C H A U X DPhPE, CEN Saclay, F-91191 Gif-sur-Yvette Cedex, France Received 17 March 1987
New results are presented on nuclear effects in deep inelastic muon scattering on deuterium and iron targets at large Q2. The ratio FFe(x)/FD2(x) measured in the kinematic range 0.06~ 0.25. At lower x, the structure function ratio exhibits an enhancement of ~ 5%. Present address: Digital Equipment, 1-10125 Turin, Italy. 2 On leave from Institute of Physics, CSAV, CS-18040 Prague, Czechoslovakia. 3 On leave from the Institut fiir Hochenergiephysikder AdW der DDR, DDR- 1615 Berlin-Zeuthen,GDR. 4 Present address: INFN, 1-34100Trieste, Italy 5 On leave from Central Research Institute for Physics, H-1525 Budapest, Hungary. 6 Present address: Institute of Nuclear Research and Nuclear Energy, Sofia 1784, Bulgaria. 7 On leave from E. Kardelj University and the J. Stefan Institute, 61111 Ljubljana, Yugoslavia. s Present address: SiemensAG, D-8000 Munich, Germany. 9 Present address: DPhPE, CEN Sacla~, F-91191 Gif-sur-Yvette Cedex, France. ~oFunded in part by the German Federal Minister for Research and Technology(BMFT) under contract number 054MU12P6. 0370-2693/87/$ 03.50 © Elsevier Science Publishers B.V. ( N o r t h - H o l l a n d Physics Publishing D i v i s i o n )
Several m u o n a n d electron scattering experiments at CERN a n d SLAC have investigated the effect that a nucleon e m b e d d e d in a nucleus has a quark distrib u t i o n different from that of a free nucleon. I n deep inelastic scattering, the " E M C effect" is studied by comparing the nucleon structure function F A ( x ) measured on a heavy nucleus of mass A to the deuterium structure function F ~ : ( x ) , where x is the Bjorken scaling variable. While all experiments agree on the pattern of the nuclear effect i n the valencequark region x > 0.3, namely a softening of the structure function when measured o n a heavy target, the experimental situation i n the low-x region is controversial. The EM Collaboration. i n their original 483
Volume 189, number 4 measurement
[ 1]
PHYSICS LETTERSB observed
the
14 May 1987
ratio
Fr~e(x)/F~2(x) to increase linearly towards small x. On the contrary, the SLAC E139 experiment [2] which measured cross section ratios aA/aDfor a variety of nuclei found no significant effect in the region of x < 0.3, independent of the target mass. An earlier SLAC experiment [3] at small four-momentum transfers Q2 g 1 GeV 2 had observed an enhancement around x--- 0.15 and a turnover at very small x--- 0.05. In a previous paper [4] we have presented data on the structure function ratios F'~(x)/F ~ (x) for nitrogen ( A = 14) and iron ( A = 5 6 ) measured at a beam energy of 280 GeV. The N2 data covered the range 0.08 ~
T o3
D-~_
03
o')
eo
~2
o~
r'o
c~E
c~
Volume 189, number 4
PHYSICS LETTERSB
in the present analysis. For the second half of the iron data taking the deuterium targets were emptied. The inverse contamination of the iron data by events from the neighbouring D2 target is much smaller due to the strongly different target densities. This background was determined both by a Monte Carlo study of the vertex resolution and by a direct comparison of the iron samples taken with and without D2 in the second target. With both methods we find a contamination of 1.3% for which the data are corrected. A background from target wall interactions which amounts to 0.7% for the external and 2.5% for the internal targets is subtracted from the D2 data. The structure functions are obtained from the experimental data in a way which is very similar to the one described in ref. [4]. The experimental distributions are converted to cross sections, correcting for acceptance and resolution of the spectrometer by a detailed Monte Carlo simulation of the experiment. To evaluate the structure functions F2(x,Q 2) we assume a constant value R = eL~aT = 0. Although this is an approximation in the region of small x, it does not affect the F2 ratio provided that R is independent of atomic mass. The deuterium structure function is computed separately for events from internal and external targets for which the acceptance of the spectrometer is different. In the kinematical region of overlap, the structure functions are in agreement within statistical errors and were combined for the subsequent analysis. The iron data are corrected for the non-isoscalarity of S6Fe assuming a neutron/proton structure function ratio F n2/F P2 = 1 - 0.75x. No corrections are applied for the Fermi motion of nucleons inside the nucleus. The results presented here are based on 4.1 × 105 reconstructed events originating from the deuterium and 2.8 × 105 events from the iron targets. The sources of systematic errors in the/72 ratio are largely the same as in our earlier experiment [4]. They are mainly due to the resolution of the spectrometer, small uncertainties on the energy loss in the different target niaterials, hadronic shower punchthrough into'the proportional chambers, and the reproducibility of the spectrometer magnetic field settings. The uncertainty from spectrometer resolution is larger than in our previous data because the acceptance of the apparatus for events from the external targets decreases along the beam direction
14 May 1987
and is therefore different for the two target materials. Errors on the acceptance correction due to this effect were calculated by varying the vertex resolution in the Monte Carlo simulation of the experiment. The uncertainty on the relative luminosity calibration of the Fe and D2 data is estimated to be
1.5%. The F[e/F~ 2 ratio is shown as a function o f x and Q2 in fig. 2 and does not exhibit a significant Q2 dependence. It is therefore averaged over Q2 and is shown as a function of x in fig. 3 together with our previously published data [4]. Good agreement is observed between the two measurements and the structure function ratio FVe(x)lF~2(x) from the combined data sets is given in table 1. The results from this and from other charged lepton experiments are shown in fig. 4. The comparison to the EMC iron data [ 1 ] shows good agreement for x > 0.15, apart from a 3% shift in the relative nor1.2
-
ei;
S ........
~ i ................. :.................................................
~
~i
"~i
................................... ~
-
-
L
~XXo.9 %_~0.8 0.7
x =
.07
,
i
i
i
x =
r i,i]
.10
f
i
F
i i ii~I
1.1 1. 0.9
........... ~
0.8
* ~ ~
............. i .............
*
~
...........................
0.7
~-/ 1,1
,
,
......
~L=--?
.......
,
L
. . . . . . . . . . . . . . .
1. 0.9 0.8 0.7
x =
0.9
.225
i
1.1
x =
i _1
11111(
.275
i
............
i
i
i
i i ill
"..........................
0.8
•
I
O.7
x =
35,
~ i
i
~ 1 11pl
x =
[
.45 I
I
i
i a llrl
i
.111_
o;~
/
~=, 10
.......
~ 10 2
I ~,~° 10
/
,
.....
I,,
•
10 2
c;,' (Cev ')
Fig. 2. The structure function ratio F~e/F~2 in bins ofx and Q2 from this experiment at 200 GeV beam energy (dosed points) and from an earlier experiment at 280 GeV [41 (open points). Only statistical errors are shown. The dotted lines indicate the averageover the respectivex bin. 485
Volume 18 9, number 4
PHYSICS LETTERS B
14 May 19 8 7
,~. 1,2
L~12
u2
• BCDMS (This experiment) O BCDMS (Ref. 4)
1.1
(a)
•
BCDMS (combined)
[ ] EMC (Ref. 1)
L~
eo
11
+'44
0.8
1 1.2 j0.8
"
o%.
p
o.,
o!~
o!~
o!~
0.5
0!6
0!7
(b)
}I i~T~ F~ g l
•
J
J
J
BCDMS (combined)
O Arnold et al. (Ref. 2)
I
[ ] S t e i n e t ol. (Ref. 5)
]
0.18 0.9 Bjorken x
Fig. 3. The structure function ratio F~e(x)lF~2(x) measured in this and in a previous [4] experiment. Only statistical errors are shown.
malization. For x < 0.15, the two measurements are marginally compatible within the quoted systematic errors. Preliminary data from the EM Collaboration on a copper target show a less pronounced effect at small x in good agreement with our result [ 6 ]. The agreement with the SLAC E139 data [2] is excellent for x > 0.25 but rather poor at small x , In this region, we observe, however, a very good agreement with the earlier SLAC experiment on a copper target [ 3] at small Q2~ 1 GeV 2.
0.
0.1
0.2
0.3
0.4
0,5
0.6
0.7
0.8 0.9 Bjorken x
Fig. 4. The structure function ratio FVe(x)/F~(x) from this and from a previous measurement [4] combined, compared to other muon (a) and electron (b) scattering experiments. The data from ref. [ 3 ] were taken with a copper target. Only statistical errors are shown.
In summary, we have complemented our earlier measurement of the structure function ratio FFet 2 kx , fl2"~/FD2I ~1 2 ~ ~. , ~ 1"32"~ J b y n e w d a t a covering t h e region of small x (0.06 ~ x ~<0.20) and improving the
Table 1 Results for R(x) =FVe(x)/F~'-(x) from this experiment and ref. [4] combined. The systematic errors do not include the 1.5% uncertainty on the relative normalization of Fe and D2 data.
486
X
Q2 range (GeV 2)
R(x)
Statistical error
Systematic error
0.07 0.10 0.14 0.18 0.225 0.275 0.35 0.45 0.55 0.65
14- 20 16- 30 18- 35 18- 46 20-106 23-106 23-150 26-200 26-200 26-200
1.048 1.057 1.046 1.050 1.027 1.000 0.959 0.923 0.917 0.813
0.016 0.009 0.009 0.009 0.009 0.011 0.009 0.013 0.019 0.023
0.016 0.012 0.011 0.009 0.010 0.010 0.011 0.015 0.021 0.030
Volume 189, number 4
PHYSICS LETTERS B
statistical accuracy at larger x. No Q2 dependence of the nuclear effect is observed over the kinematic range of the experiment. I n the region x > 0.25, we find good agreement with all other charged lepton experiments [1,2,4,6]. For x < 0 . 2 5 , we observe an e n h a n c e m e n t of the structure function ratio of 4.5%___ 0.5% (stat.) + 2.0% (syst.) where the systematic error includes the uncertainty on the relative n o r m a l i z a t i o n of iron a n d deuterium data.
14 May 1987
References [ 1] EM Collab.,J.J. Aubert et al., Phys. Lett. B 123 (1983) 275. [2] R.G. Arnold et al., Phys. Rev. Lett. 52 (1984) 722, SLACPUB-3257. [3] S. Stein et al., Phys. Rev. D 12 (1975) 1884. [4] BCDMSCollab.,G. Bad et al., Phys. Lett. B 163 (1985) 282. [5] BCDMS Collab., D. Bollini et al., Nuel. Instrum. Methods 204 (1983) 333; BCDMS Collab., A.C. Benvenuti et al., Nucl. Instrum. Methods 226 (1984) 330. [6 ] P.R. Norton, Proc. XXIII Intern. Conf. on High energyphysics (Berkeley, 1986), to be pubfished.
487
PHYSICS LETTERSB
14 May 1987
NUCLEAR EFFECTS IN DEEP I N E L A S T I C M U O N S C A T T E R I N G O N D E U T E R I U M A N D IRON T A R G E T S BCDMS Collaboration A.C. B E N V E N U T I , D. BOLLINI, G. BRUNI, F.L. N A V A R R I A Dipartimento di Fisica dell'Universit?tand INFN, L40126 Bologna, Italy A. A R G E N T O 1, j. CVACH 2, K. D I E T E R S 3, L. P I E M O N T E S E 4, I. VERESS s p. ZAVADA 2 CERN, CH-1211 Geneva 23, Switzerland N.G. FADEEV, I.A. G O L U T V I N , M.Yu. K A Z A R I N O V , Yu.T. K I R Y U S H I N , V.G. K R I V O K H I Z H I N , V.V. K U K H T I N , W. L O H M A N N 3, S. N E M E C E K , P. REIMER, I.A. SAVIN, G.I. SMIRNOV, J. S T R A C H O T A 2, G. S U L T A N O V 6, p. T O D O R O V , A.G. V O L O D K O Joint Institutefor Nuclear Research, Dubna, 101000Moscow, USSR D. J A M N I K 7 R. K O P P 8, U. M E Y E R - B E R K H O U T , A. STAUDE, K.-M. T E I C H E R T , R. T I R L E R 9, R. VOSS, C. Z U P A N C I C Sektion Physik der Universitiit, D-8000 Munich, Fed. Rep. Germany ~o M. CRIBIER, J. FELTESSE, A. MILSZTAJN, A. O U R A O U , P. R I C H - H E N N I O N , Y. SACQUIN, G. SMADJA a n d M. V I R C H A U X DPhPE, CEN Saclay, F-91191 Gif-sur-Yvette Cedex, France Received 17 March 1987
New results are presented on nuclear effects in deep inelastic muon scattering on deuterium and iron targets at large Q2. The ratio FFe(x)/FD2(x) measured in the kinematic range 0.06~
Several m u o n a n d electron scattering experiments at CERN a n d SLAC have investigated the effect that a nucleon e m b e d d e d in a nucleus has a quark distrib u t i o n different from that of a free nucleon. I n deep inelastic scattering, the " E M C effect" is studied by comparing the nucleon structure function F A ( x ) measured on a heavy nucleus of mass A to the deuterium structure function F ~ : ( x ) , where x is the Bjorken scaling variable. While all experiments agree on the pattern of the nuclear effect i n the valencequark region x > 0.3, namely a softening of the structure function when measured o n a heavy target, the experimental situation i n the low-x region is controversial. The EM Collaboration. i n their original 483
Volume 189, number 4 measurement
[ 1]
PHYSICS LETTERSB observed
the
14 May 1987
ratio
Fr~e(x)/F~2(x) to increase linearly towards small x. On the contrary, the SLAC E139 experiment [2] which measured cross section ratios aA/aDfor a variety of nuclei found no significant effect in the region of x < 0.3, independent of the target mass. An earlier SLAC experiment [3] at small four-momentum transfers Q2 g 1 GeV 2 had observed an enhancement around x--- 0.15 and a turnover at very small x--- 0.05. In a previous paper [4] we have presented data on the structure function ratios F'~(x)/F ~ (x) for nitrogen ( A = 14) and iron ( A = 5 6 ) measured at a beam energy of 280 GeV. The N2 data covered the range 0.08 ~
T o3
D-~_
03
o')
eo
~2
o~
r'o
c~E
c~
Volume 189, number 4
PHYSICS LETTERSB
in the present analysis. For the second half of the iron data taking the deuterium targets were emptied. The inverse contamination of the iron data by events from the neighbouring D2 target is much smaller due to the strongly different target densities. This background was determined both by a Monte Carlo study of the vertex resolution and by a direct comparison of the iron samples taken with and without D2 in the second target. With both methods we find a contamination of 1.3% for which the data are corrected. A background from target wall interactions which amounts to 0.7% for the external and 2.5% for the internal targets is subtracted from the D2 data. The structure functions are obtained from the experimental data in a way which is very similar to the one described in ref. [4]. The experimental distributions are converted to cross sections, correcting for acceptance and resolution of the spectrometer by a detailed Monte Carlo simulation of the experiment. To evaluate the structure functions F2(x,Q 2) we assume a constant value R = eL~aT = 0. Although this is an approximation in the region of small x, it does not affect the F2 ratio provided that R is independent of atomic mass. The deuterium structure function is computed separately for events from internal and external targets for which the acceptance of the spectrometer is different. In the kinematical region of overlap, the structure functions are in agreement within statistical errors and were combined for the subsequent analysis. The iron data are corrected for the non-isoscalarity of S6Fe assuming a neutron/proton structure function ratio F n2/F P2 = 1 - 0.75x. No corrections are applied for the Fermi motion of nucleons inside the nucleus. The results presented here are based on 4.1 × 105 reconstructed events originating from the deuterium and 2.8 × 105 events from the iron targets. The sources of systematic errors in the/72 ratio are largely the same as in our earlier experiment [4]. They are mainly due to the resolution of the spectrometer, small uncertainties on the energy loss in the different target niaterials, hadronic shower punchthrough into'the proportional chambers, and the reproducibility of the spectrometer magnetic field settings. The uncertainty from spectrometer resolution is larger than in our previous data because the acceptance of the apparatus for events from the external targets decreases along the beam direction
14 May 1987
and is therefore different for the two target materials. Errors on the acceptance correction due to this effect were calculated by varying the vertex resolution in the Monte Carlo simulation of the experiment. The uncertainty on the relative luminosity calibration of the Fe and D2 data is estimated to be
1.5%. The F[e/F~ 2 ratio is shown as a function o f x and Q2 in fig. 2 and does not exhibit a significant Q2 dependence. It is therefore averaged over Q2 and is shown as a function of x in fig. 3 together with our previously published data [4]. Good agreement is observed between the two measurements and the structure function ratio FVe(x)lF~2(x) from the combined data sets is given in table 1. The results from this and from other charged lepton experiments are shown in fig. 4. The comparison to the EMC iron data [ 1 ] shows good agreement for x > 0.15, apart from a 3% shift in the relative nor1.2
-
ei;
S ........
~ i ................. :.................................................
~
~i
"~i
................................... ~
-
-
L
~XXo.9 %_~0.8 0.7
x =
.07
,
i
i
i
x =
r i,i]
.10
f
i
F
i i ii~I
1.1 1. 0.9
........... ~
0.8
* ~ ~
............. i .............
*
~
...........................
0.7
~-/ 1,1
,
,
......
~L=--?
.......
,
L
. . . . . . . . . . . . . . .
1. 0.9 0.8 0.7
x =
0.9
.225
i
1.1
x =
i _1
11111(
.275
i
............
i
i
i
i i ill
"..........................
0.8
•
I
O.7
x =
35,
~ i
i
~ 1 11pl
x =
[
.45 I
I
i
i a llrl
i
.111_
o;~
/
~=, 10
.......
~ 10 2
I ~,~° 10
/
,
.....
I,,
•
10 2
c;,' (Cev ')
Fig. 2. The structure function ratio F~e/F~2 in bins ofx and Q2 from this experiment at 200 GeV beam energy (dosed points) and from an earlier experiment at 280 GeV [41 (open points). Only statistical errors are shown. The dotted lines indicate the averageover the respectivex bin. 485
Volume 18 9, number 4
PHYSICS LETTERS B
14 May 19 8 7
,~. 1,2
L~12
u2
• BCDMS (This experiment) O BCDMS (Ref. 4)
1.1
(a)
•
BCDMS (combined)
[ ] EMC (Ref. 1)
L~
eo
11
+'44
0.8
1 1.2 j0.8
"
o%.
p
o.,
o!~
o!~
o!~
0.5
0!6
0!7
(b)
}I i~T~ F~ g l
•
J
J
J
BCDMS (combined)
O Arnold et al. (Ref. 2)
I
[ ] S t e i n e t ol. (Ref. 5)
]
0.18 0.9 Bjorken x
Fig. 3. The structure function ratio F~e(x)lF~2(x) measured in this and in a previous [4] experiment. Only statistical errors are shown.
malization. For x < 0.15, the two measurements are marginally compatible within the quoted systematic errors. Preliminary data from the EM Collaboration on a copper target show a less pronounced effect at small x in good agreement with our result [ 6 ]. The agreement with the SLAC E139 data [2] is excellent for x > 0.25 but rather poor at small x , In this region, we observe, however, a very good agreement with the earlier SLAC experiment on a copper target [ 3] at small Q2~ 1 GeV 2.
0.
0.1
0.2
0.3
0.4
0,5
0.6
0.7
0.8 0.9 Bjorken x
Fig. 4. The structure function ratio FVe(x)/F~(x) from this and from a previous measurement [4] combined, compared to other muon (a) and electron (b) scattering experiments. The data from ref. [ 3 ] were taken with a copper target. Only statistical errors are shown.
In summary, we have complemented our earlier measurement of the structure function ratio FFet 2 kx , fl2"~/FD2I ~1 2 ~ ~. , ~ 1"32"~ J b y n e w d a t a covering t h e region of small x (0.06 ~ x ~<0.20) and improving the
Table 1 Results for R(x) =FVe(x)/F~'-(x) from this experiment and ref. [4] combined. The systematic errors do not include the 1.5% uncertainty on the relative normalization of Fe and D2 data.
486
X
Q2 range (GeV 2)
R(x)
Statistical error
Systematic error
0.07 0.10 0.14 0.18 0.225 0.275 0.35 0.45 0.55 0.65
14- 20 16- 30 18- 35 18- 46 20-106 23-106 23-150 26-200 26-200 26-200
1.048 1.057 1.046 1.050 1.027 1.000 0.959 0.923 0.917 0.813
0.016 0.009 0.009 0.009 0.009 0.011 0.009 0.013 0.019 0.023
0.016 0.012 0.011 0.009 0.010 0.010 0.011 0.015 0.021 0.030
Volume 189, number 4
PHYSICS LETTERS B
statistical accuracy at larger x. No Q2 dependence of the nuclear effect is observed over the kinematic range of the experiment. I n the region x > 0.25, we find good agreement with all other charged lepton experiments [1,2,4,6]. For x < 0 . 2 5 , we observe an e n h a n c e m e n t of the structure function ratio of 4.5%___ 0.5% (stat.) + 2.0% (syst.) where the systematic error includes the uncertainty on the relative n o r m a l i z a t i o n of iron a n d deuterium data.
14 May 1987
References [ 1] EM Collab.,J.J. Aubert et al., Phys. Lett. B 123 (1983) 275. [2] R.G. Arnold et al., Phys. Rev. Lett. 52 (1984) 722, SLACPUB-3257. [3] S. Stein et al., Phys. Rev. D 12 (1975) 1884. [4] BCDMSCollab.,G. Bad et al., Phys. Lett. B 163 (1985) 282. [5] BCDMS Collab., D. Bollini et al., Nuel. Instrum. Methods 204 (1983) 333; BCDMS Collab., A.C. Benvenuti et al., Nucl. Instrum. Methods 226 (1984) 330. [6 ] P.R. Norton, Proc. XXIII Intern. Conf. on High energyphysics (Berkeley, 1986), to be pubfished.
487