1. Prospects for Diffractive and Forward Physics at the LHC
6th Small-x and Diffraction Workshop
Edmundo García
University of Illinois at Chicago
On behalf of the CMS and TOTEM
Diffractive and Forward Physics Working Group
Fermilab, March 2007

2. Outline Introduction Diffractive Physics Experimental Capabilities
Forward and near beam detectors
Acceptance
Background (pile-up)
Triggering
Forward Physics
Final Remarks
This talk will have two main topics diffractive and forward physics
I will cover the experimental capabilities and some detector issues in the diffractive part and then
show some simulation work that has been done on the forward physics
The talk is a compilation from the note on diffractive and forward physics
available on the web, and it complements Fabricio and Mike albro’s talk.
- Complemented by F. Ferro talk
Total Cross section, proton acceptance and reconstruction, and soft diffraction (low luminosity)
- RP420 talk by M. Albrow

3. Introduction X f h g* (Z or W)
( or p)
(or p)
no colour
~ no effective
spin interchange
g* (Z or W)
rapidity gap X f h
D0 Run I sample
HERA: epeXp diffractive deep inelastic scattering
This is just a picture of the diffractive process that we have seen many times already where on the physics side we have a almost no colour and spin interchange of a photon
or a boson, resulting in two leading protons and something.
An on the experimental side resulting of a so called rapidity gap in the detector, as shown in the transparency from an event of D0.
No quantum number exchange other than the vacuum.
As a reference are the studies of the diffractive events studied at Hera, from where the cross section for this processes is expresed in terms
of the f2 and the longitudinal diffractive structure function F2.
diffractive structure functions

4. Proton Structure Probe
Eur. Physs. J. C48 (2006)
Eur. Physs. J. C48 (2006)
The reduced diffractive cross section shown in the figure as function of the squared transfer momentum is shown on the left of the figure.
Given that FL sizable only to high y, this is equal to the diffractive structure function F2.
As it was noted earlier during this conference the string rise of the distribution for low x indicated that the partons resolved in diffractive events
are predominantly quark singlets and gluons.
Also from Hera on the right side is presented the diffractive parton distribution functions vs z momentum fraction
two things come from this plots, the QDC factorization of the diffractive structure function, from the ratio of the
distributions that 70% or so of the exchanged momentum in the diffractive events is carried by gluons
Strong rise at low-x: partons resolved ~ gluons (+ quark singlets).
Hard scattering obeys QCD factorization at HERA energies.
~ 70% of exchanged momentum carried by gluons

5. Exclusive Diffractive processes
HERA: g*pVp “elastic” vector meson production and g*pgp deeply virtual Compton scattering
Hera has also investigated the exclusive elastic vector meson production and the deeply virtual compton scattering
On the right is another tranparencie we saw earlier, the cross section for the photoproduction of vector mesons.
as function of the invariant mass.
The energy dependence of the distribution becomes steep in the presence of hard scale
This reflects the x scale dependence on the square of the gluon density in the proton.
Furthermore, also from Zeus data shown on the right, we see that the gluon desnsity grows at a given q2
as the x gets smaller.
hep-ph/ ZEUS Data
hep-ph/
Energy dependence of these processes becomes steep in the presence of hard scale
This reflects the x scale dependence of the gluon density in the proton.
Cross section of these processes is ~ proportional to the square of gluon density.

6. Diffractive Physics at the Tevatron
SD rates of W and Z- bosons, dijet, b-quark and J/y meson production in the Tevatron are about 10 times lower than expectations based diffractive parton distribution functions determined at HERA
Phys. Rev. Lett. 84 (2000) 5043
More recently as shown here yesterday, we learned that the production channels for single difractive events as
W and Z bosons as well as the meson production are a factor of 10 lower than the expectations based on the
distribution functions determined at Hera.
This can be seen on the left is the diffractive structure function of the proton measured by CDF (points) and
the expectations based on the diffractive parton distribution functions from hera.
On the right is the ratio difractive dijet event rate per unit of momentum loss to the nondiffractive
dijet rate as a function of the borjen x of the parton in the antiproton, again the expected value for this ratio
based on the diffractive parton distribution fuctions from hera would be around 1.
A likely explanation of the breakdown of the qcd factorization is the filling of the rapidity gap
of the leading proton due to soft interactions.
B momentum fraction related to x by beta*xi = xbj
F2jj = F2D integrated over the antiproton momentum loss
hep-ex /
Breakdown of QCD factorization for diffractive processes
Likely due to the filling of the rapidity gap and slowing (breaking) of the proton due to soft scattering of the two interacting hadrons (quantified by rapidity gap survival probability)

7. From the Tevatron to the LHC
LHC will probe forward rapidities y ~ 5, Q2 ~ 100 GeV and x down to ~10-5
Some of the available processes:
LHC
Tevatron
Based on Stirling’s Eur. Phys. J. C 14, 133 (2000)
Inclusive single diffraction (SD) and double “pomeron” exchange (DPE)
pppX
pppXp
production of dijets, vector bosons and heavy quarks
pppjjX
pppW(Z)
pppqq
Central exclusive production
pp pHp
with
H (120GeV) bb
High energy photon interactions
pppWX
pp(pgp)pWHX
This takes us to the next step, the LHC.
In this transparency the Kinematic plane Q2 vs x is shown for HERA, the tevatron and the LHC
the tevatron region is sketched.
We can see that at the LHC we will have access to a x of about 10-5 in the q2 region of
about 100, at forward rapidities of about 5.
On the left I shoe some of the physics processes that can be investigated for diffractive and low x
physics.

8. Program on Diffractive Physics
The accessible physics is a function of the instantaneous and integrated luminosity
Low Luminosity:
Low enough that pile-up is negligible, i.e <1032 cm-2s-1, and integrated lumi a few 100pb-1 to <1 fb-1
Measure inclusive SD and DPE cross sections and their Mx dependence
Intermediate Luminosity (*=0.5m)
Lumi < 1033 cm-2s-1, pile-up starts becoming an issue, integrated lumi 1 to a few fb-1
Measure SD and DPE in presence of hard scale (dijets, vector bosons, heavy quarks)
Follow the Tevatron program
High Luminosity:
Lumi > 1033 cm-2s-1, pile-up substantial and integrated lumi several tens of fb-1
Discovery physics comes into reach in central exclusive production
The extent ant timing for the diffractive program of course will depend on the running conditions,
that is the beta of the beam and the luminosity as shown in this transparency.

9. Diffractive Physics with CMS, Totem, and the Forward DetectorsX
Single diffraction (SD):
central CMS
apparatus
rap gap - veto on T1,T2, HF, ZDC (CASTOR)
Near-beam
detectors - Roman Pots (F420)
leading proton
leading proton
Near-beam
detectors
This is the event topology that illustrates how the diffractive events will be measured by totem and cms.
rap gap
Double Pomeron exchange (DPE):
central CMS
apparatus X
rap gap
leading proton
Near-beam
detectors

10. Central CMS detector - |h|<2.4
CALORIMETERS
MUON BARREL
ECAL
HCAL
Drift Tube
Resistive Plate
Silicon Microstrips
and Pixels
Si TRACKER
Scintillating
PbWO4 crystals
Plastic scintillator/brass
sandwich
Chambers (DT)
Chambers (RPC)

11. Forward Region Forward Cal ZDC Totem T1,T2 Totem FP420 Roman Pots
(8.3 < ||)
(3 < |h| < 5) EM
(n,g)
(5.2 <  < 6.7)
HAD
(z =  140 m)
FP420
CMS IP
ZDC RP
T1/T2, Castor RP
FP420 suggested by the FP420 R&D collaboration
Mention T1, T2 and RP are Totem Detectors
T1 (cathode strip chambers) and T2 (gas electron multiplier) are telescope detectors
Roman pots silicon strips
Totem
Roman Pots
FP420
CASTOR
(z =  420 m)
(z =  147,  220 m)
(5.2 < h < 6.6)
possible
addition

12. TOTEM (+ FP420): coverage for leading protons
At nominal LHC optics:
*=0.5m and L = cm-2 s-1
fractional momentum loss
of the proton.
FP420
RP220
at 220m: <  < 0.2
at 420m: <  < 0.02 FP
420
Nucl. Phys. B658 (2003) 3
TOTEM
For this transparencies it is assumed nominal LHC optics with beta star of ½ m and nominal luminosities. Xi is the fractional momentum loss of the photon.
that will be measure by totem’s roman pots and by the fp420 detectors.
For this conditions is expected that the fractional momentum loss of the leading protons to be in the shown limits.
on the left To illustrate the coverage it is superimposed to the longitudinal fraction momentum cross section for deep inelastic diffractive scattering data
from HERA. FP420 sits right in top of the diffractive peak and covers everything from 0.98 to 0.998
And on the right is the acceptance for beam at RP220 and RP 420.
In terms what this means for physics, given the relation for DPE central exclusive production between
For example:
With √s=14TeV, M=120GeV on average:
   1%
xL=P’/Pbeam= 1-x

13. TOTEM (+ FP420): coverage for leading protons
Example: Central exclusive production
pp pHp with H (120GeV)  bb
In non-diffractive production very hard, signal swamped with QCD dijet background
Selection rule in central exclusive production (CEP) ~ JPC = improves S/B for SM Higgs dramatically
Key: Detection of diffractively scattered protons inside of beam pipe. Plus: two collimated jets in central detector with consistent values mass from central detectors
Following the example for the central exclusive production of 120 GeV Higgs, a key for its detection will be the
detection of the scattered protons inside the beam pipe.
The acceptance plots for different combinations of the measured fraction momentum of the leading protons
is shown on the left vs. the reconstructed mass calculated by the their correlation.
For this plot only the leading protons were used. The plots reflects what one gets from the near-beam detector acceptances
Acceptance for events in which both protons are measured in the 420 m detectors or in combination of 420 m m detectors (“Total”)
M. Albrow talk will be on the 420 m detectors.

14. Pile-up Background
Using a mixed sample of events and pile-up one can extract probability of obtaining a fake DPE signature in the near beam detectors caused by protons from pile-up events:
CMS-2006/054 and TOTEM-2006/01
The signal in the central detectors is due to inclisive dijet events
Proper mix of events with pileup was done with FAMOS
It also depends critically on the leading proton spectrum at the LHC which in turn depends on size of soft rescattering effects (rapidity gap survival factor)
Eg at 2 x1033 cm-2s-1 with an average of 7 pileup events per beam crossing, ~10% of the non-diffractive background in the central detector appear to have DPE signature
This is independent of the type of signal.
S/B_PU hence depends on the relative cross section of diffractive signal and non-diffractive background that looks like the signal in the central detector

15. Pile-up Background
CEP H(120) bb
inclusive QCD di-jets + pu
Reduced by correlation between ξjets measured in the central detector and ξRP measured by the near-beam detectors
Reduced by fast timing detectors that can determine whether the protons seen in the near-beam detector came from the same vertex (current R&D project)
jets
RP220
- pile-up and trigger background are mainly problems for selecting diffractive events.
-All plot are before the correlations and timing cuts applied, they illustrate how different are the
distributions for signal and background, and that a cut on these quantities can differentiate the signal from the background
M2jets/Mmm

16. Triggering
Diffractive/forward processes typical values of pT lower than most hard processes suited for CMS.
The CMS trigger menus now foresee a dedicated diffractive trigger stream
with 1% of the total bandwidth on L1 and HLT
CMS-2006/054 and TOTEM-2006/01
The 420 m detectors won’t
make it to the L1 trigger
within latency
Reduction of the rate from standard QCD processes fore evenst.
single-sided 220m condition without and with cut on 
Achievable total reduction: 10 x 2 (HT cond) x 2 (topological cond) = 40
Jet isolation
criterion
reduction requiring 2 jets in the same  hemisphere
as the RP detectors that see the proton
Adding L1 conditions on the near-beam detectors provides a rate reduction
sufficient to lower the 2-jet threshold to 40 GeV per jet while still meeting
the CMS L1 bandwidth limits for luminosities up to 2 x 1033 cm-1 s-1

17. Trigger efficiencies for Single Diffraction
pppWx: one L1 jet with ET above threshold required
pppjjX: at least two L1 jets with ET above threshold required
No pile-up case, three different conditions studied:
Trigger in central detectors alone
With single arm 220 m condition (proton signal detected on near-beam 220 m detectors)
With single arm 420 m condition (for completeness)
No pile-up case means that signal MC used with no pu events mised in
pu has near no impact on the trigger signal efficiency, so it can be omit to first approximantion

18. Trigger Efficiencies DPE
Example: Central exclusive production
pp pHp with H (120GeV)  bb
Level-1:
HT condition ~ rapidity gap of 2.5 with respect to beam direction
2-jets (ET>40GeV) & single-sided 220m, efficiency ~ 12%
Add another ~10% efficiency by introducing a 1 jet & 1 (40GeV, 3GeV) trigger condition
HLT: Efficiency ~7%
To stay within 1 Hz output rate, needs to either prescale b-tag or add 420 m detectors in trigger
Possible to retain about ~10% of the signal events up to 2 x 1033 cm2 s1 in a special forward detectors trigger stream
Reduction of the pile-up background by O(109) can be obtained with relatively simple cuts, further requirements (vertex position) can give a further suppression or larger.
Yield S/B in excess of unity for a SM Higgs

19. Program on Forward Physics
Study of the underlying event at the LHC:
Multiple parton-parton interactions and rescattering effects accompanying a hard scatter
Closely related to gap survival and factorization breaking in hard diffraction
Heavy-ion and high parton density physics:
Proton structure at low xBJ → saturation → Color glass condensates
Photon-photon and photon-proton physics:
Also there protons emerge from collision intact and with very low momentum loss
Cosmic ray physics
Models for showers caused by primary cosmic rays (PeV = 1015 eV range) differ substantially
Fixed target collision in air with 100 PeV center-of-mass E corresponds to pp interaction at LHC
Hence can tune cosmic ray shower models at the LHC

20. Forward Jets
At Tevatron is the standard tool to study PDFs in global fit analysis
At CMS (ET ~ GeV) dijet detection will be possible in HF for 3<h<5 and -3<h<-5
This allows to probe values as low as x2 ~ 10-4 (even further with Castor ~ 10-6)
Integreated luminosity 1 pb-1
m = ET
PYTHIA ~ NLO
Log x12 distribution of two protons at 14 GeV producing at least one jet above 20GeV within the acceptance of HF
estimated single inclusive jet cros section in HF at Generator level, no detector response included
Single inclusive jet measurement
Generator level estimates: No detector response, underlying event and hadronization corrections included for spectrum

21. Forward Drell-Yan
Log10(x12)
Access to the low x regime of the quark (antiquark) distributions requires imbalance of fractional momenta  x1,2 boosted to large rapidities
CASTOR with 5.3 ≤ || ≤ 6.6 gives
access to xBJ~10-7
Measure angle of electrons with T2
In this process the lepton pair originates from the annihilation of a quark-antiquark pair with fractional momentum
X1 and x2, related to the dileptom ass and rapidity by the following relations.
In order to be sensitive to the low x regime of the q qbar distributions at reasonable values of m2
A large imbalance of the fractional momentum is required, boosting the leptos to large rapidities
The first plot shows the x12 distributions detected in castor for a electron mass greater that 4 GeV
Un the lower plot is the cross section for this process calculated for the standard and saturated proton density
Parametrizationas a function of bjoen x, for all pairs and the pairs detected in castor.

22. Forward Physics: Gluon saturation region
New regime of QCD is expected to reveal at small x due to effects of large gluon density
So far not observed in pp interactions
Field of common interest with Heavy Ions Physics Q2
ZEUS data for gluon distribution
In partons
saturation
CGC
low x -> as<<1
At a given Q(x,A) -> gluon density saturation -> saturation models for heavy ions
Sub detector
Coverage
Forward HCAL
3 .0< | h | < 5.2
CASTOR
5.2 < | h | < 6.6
ZDC
8.3 < |h|
McLerran, hep-ph/

23. Limiting Fragmentation
Castor Coverage
At RHIC limited fragmentation can be
explained by gluon saturation (Colour Glass Condensate) model.
Data PHOBOS 200, 130 and 62 GeV AuAu
Model McLerran-Venugopalan
This can be measured also at LHC for larger gluon density region
Phys. Rev. D 49, 2233 (1994 )

24. Photon-mediated processes
(di-muons) - (true)
rms ~ 10-4
QED theoretical cross section for this process is precisely known
Reaction can then be used to calibrate the pp-luminosity provided that events can be identified above pile-up
The calibration for the forward proton’s energy measurement can be achieved
Expect ~300 events/100 pb-1 after CMS muon trigger
Resolution of the reconstructed forward proton momentum using exclusive pairs in CMS
Reconstruction of proton  values with resolution of 10-4, i.e. smaller than beam dispersion

25. Validation of Hadronic Shower Models
Models for showers caused by primary cosmic rays differ substantially
Fixed target collision in air with 100 PeV center-of-mass E corresponds to pp interaction at LHC
Hence can tune shower models by comparing to measurements with T1/T2, CASTOR, ZDC

26. Final Notes CMS and Totem will carry out a joint program for physics
The two experiments together will provide coverage in h unprecedented in a hadron collider.
Will have the opportunity to review the diffractive physics program of the Tevatron, and probe the also unprecedented low-x physics provided by the LHC.
LHC
Tevatron
Based on Stirling’s Eur. Phys. J. C 14, 133 (2000)

27. Backup Transparencies

29. x=0 (beam)
x=0.002
x=0.015

30. Example:
Central exclusive production
pp pHp with H (120GeV)  bb
Selection cuts:
Jet cuts: >2 jets ET>30/45 GeV
both jets b-tagged
2 protons in near-beam detectors,
either or

31. Leading Protons Acceptance
Determined by tracking protons through the LHC accelerator lattice with the program MAD-X
Smearing of both transverse vertex position and scattering angle at the IP according to transverse beam size and beam momentum divergence
Assume that near-beam detectors are 100% efficient, i..e. assume all protons that reach 220/420m location outside of cutout for beam are detected
@220m, *=0.5m
log()
log(-t)
@420m , *=0.5m
log()
log(-t)

33. Rapidity gap survival probability
Proton and anti-proton are large objects, unlike pointlike virtual photon
In addition to hard diffractive scattering, there may be soft interactions among
spectator partons. They fill the rapidity gap and slow down the outgoing protons
hence reduce the rate of diffractive events.
Quantified by rapidity gap survival probability.
F2D
without rescattering effects
with rescatering effects
3 – 5 fb
CDF data b
Kaidalov, Khoze, Martin, Ryskin (2000)
Predictions based HERA diffractive PDFs

34. Beam Halo Background
T1 and T2  beam gas interactiosn + small muaon halo (rough calculation leads to an average suppresion factor of .6)
Roman Pts Beam halo, beam gas, pp background by inellastic pp colission in IP5
Reliable estimations at this point of machine background are very difficult, and with large uncertainties.

36. Background due to central exclusive production on bbar jets and inclusive DPE dijets
The contribution from inclusive DPE dijets has a significant theoretical uncertainty.
The b-tag requirement will be important for suppressing the pppjjXp background, it provides a large reduction factor (~500)