1. Tsukuba shrine
(near KEK)
I’m Hiroyuki Nakayama from KEK.
I’d like to talk about Belle-II experiment on behalf of Belle-II collaboration.
I’ll try my best but since the time is limited, I might rush in some places. So please ask for further questions any time.
Status and Plans for SuperKEKB and Belle II experiment Hiroyuki NAKAYAMA (KEK, Japan)

2. To start with, let us remind the primary target of KEKB/Belle.
It was to confirm Kobayashi-Maskawa mechanism,
and it was so successful as expressed in this beautiful unitary triangle plot shown here.
Feb. 24th, 2011
H.Nakayama (KEK) 2

3. The last beam abort of KEKB on June 30, 2010
After 11 years of successful operation, KEKB was shut down for the upgrade last June.
Total accumulated luminosity was more than 1 ab^-1.
Achieved peak luminosity was twice the design value.
First physics run on June 2, 1999
Last physics run on June 30, 2010
Lpeak = 2.1x1034/cm2/s
L > 1ab-1 3

4. SuperKEKB/Belle-II starts in 2014
This slide summarizes the features of the upgrade.
The upgraded accelerator is called “SuperKEKB”, with 40 times higher luminosity than current KEKB.
The upgraded detector is officially called “Belle-II”. I see some people still call it “SuperBelle”. Please be familiar with “Belle-II” from now on.
The Belle-II has largely improved performance, and it can also live with higher beam background caused by higher luminosity.
SuperKEKB/Belle-II starts in 2014 4 4

5. SuperKEKB Luminosity SuperKEKB
As you can see in this trend plot, the target luminosity of SuperKEKB is about 2 order higher than the current B factories.
But how can it be achieved?
Feb. 24th, 2011
H.Nakayama (KEK)

6. Accelerator upgrade At SuperKEKB, we increase the luminosity based on
“Nano-Beam” scheme, which was originally proposed for SuperB by P. Raimondi.
Vertical β function at IP: 5.9 → 0.27/0.30 mm (x20)
Beam current: 1.7/1.4 → 3.6/2.6 A (x2)
 L = 2x1034  8x1035 cm-2s-1 (x40)
Luminosity Gain
At SuperKEKB, we will increase the luminosity based on so-called “Nano-Beam” scheme, which was originally proposed by SuperB collaboration.
The concept is as follows:
The luminosity can be expressed by this formula, and what we’re going to do is to change these two parameters circled in red.
Beta sub y, which is the vertical beta function at interaction point, will be squeezed by factor 20. This results in the factor 20 increase of luminosity.
I, beam current will also be doubled.
In total, 40 times higher luminosity can be achieved.
Feb. 24th, 2011
H.Nakayama (KEK)

7. Final focusing magnetsIP
q =83mrad
q =22mrad
Although there are a number of features of the accelerator upgrade, here, let me pick up the design of final focusing magnet.
The crossing angle is changed to be much larger than that of KEKB, so that each rings can have their own final q magnet.
This will lead to flexible adjustment of orbit design.
The another benefit from independent final q magnet is that the bend by downstream Q magnet which existed at KEKB will disappear.
This will realize less emittance without dispersion coming the bend close to IP,
and also lead to less background such as SR at downstream magnet or over-bent spent electron from Radiative Bhabha event, which is expected to be the dominant BG of Italian SuperB.
Larger crossing angle q
Final Q for each ring more flexible optics design
No bend near IP less emittance, less background

8. Beam background
At SuperKEKB with x40 larger Luminosity, beam background will also increase drastically.
Touschek scattering
Synchrotron radiation
Beam-gas scattering
Radiative Bhabha event: emitted g
Radiative Bhabha event: spent e+/e-
2-photon process event: e+e-e+e-e+e-
Beam-beam scattering
etc…
In this slide, I have listed up variety of possible beam background sources at SuperKEKB.
Among these background sources, so-called “Touschek scattering” will be the most difficult one to cope with, which I will explain in more detail in the following slides. e- e+
Feb. 24th, 2011
H.Nakayama (KEK)

9. Touschek scattering BG
bunch
DE>0 or DE<0
Touschek scattering
Intra-bunch scattering
Both energy gain/loss
cf. Beam-gas scattering(Bremsstrahlung) : energy loss
Rate∝(beam size)-1
Vertical beam size: 0.94um0.048/0.062um
Increase drastically at SuperKEKB
Rate∝(beam energy)-3
Beam energy asymmetry is relaxed: 3.5/8.0GeV4.0/7.5GeV
Process:
Scattered e+/e- goes off trajectory  lost at beam pipe wall near IP
 creates shower  reach detector
In principle, Touschek scattering is the intra-bunch scattering.
Two particles in the same bunch collide to each other, where one gains energy and the other loses.
The scattering rate is inversely proportional to beam size, therefore this BG increases drastically at SuperKEKB with Nano-beam scheme,
The rate is also inversely proportional to cubic of beam energy, that is why the energy asymmetry is relaxed at SuperKEKB.
And how does it become the background?
Scattered e+/e- goes off trajectory  eventually lost at beam pipe wall creates shower  reach detector
To reduce this BG, we should stop this process chain somewhere.
Feb. 24th, 2011
H.Nakayama (KEK)

10. Countermeasure: Collimators
Collimators at arc sections
Horizontal SuperKEKB
Belle-II
Beam
Mask Head
One possible solution is the use of collimators to block off-momentum particles
before they reach any closer to IP.
Since more collimation is required, we plan to have horizontal collimation from both inner and outer sides.
One potential problem may be the fact that outer collimator will be hit by the SR photons, but we have already confirmed that we can handle the SR heat deposit.
cf.
At SuperKEKB, beam is collimated from both side (inner/outer).
(cf. KEKB: from inner side only)
Feb. 24th, 2011
H.Nakayama (KEK)

11. Countermeasure: Heavy-metal shield
Heavy-metal shield to protect PXD/SVD from showers coming from upstream.
Belle-II IP design (Preliminary )
~1.5cm thick
(=few radiation length)
Another countermeasure is to use heavy-metal shield to protect inner detectors, which was also used for current Belle.
“Heavy-metal” is specifically referring to Tungsten-alloy, with a very short radiation length and capacity to stop showers created by the Touschek scattered particles.
cf. Heavy-metal Belle

12. Synchrotron radiation BG
Beam pipe design
f20mmf9mm collimation on incoming beam pipe
Most of SR photons are stopped by the collimation and direct hits on IP beam pipe is negligible
HOM can escape from outgoing beam pipe
To hide IP beam pipe from reflected SR, “ridge” structure on inner surface of collimation part.
f20mm
f9mm e- e+
IP beam pipe (Be)
incoming/outgoing beam pipe (Ta)
Moving on to the next, I would like to show our optimized beam-pipe design to stop SR background.
Incoming part of beam pipe is collimated from f20mm to f9mm, therefore most of SR photons coming from upstream can be stopped by this collimation and direct hits on IP beam pipe, made of Beryllium, are negligible.
We don’t have to worry about HOM trap, since the outgoing pipes are made straight and HOM can escape .
It is also worthwhile to note that “ridge” or “saw-tooth” structure is set on the inner surface of collimation part to avoid the reflected SR hitting the IP beam-pipe.
The expected SR background of the inner detector is much smaller than the minimal requirement, so we are satisfied with this clever design of beam-pipe.
Ridge structure
Feb. 24th, 2011
H.Nakayama (KEK)

13. Other BG sources BG source Beam-gas scattering
Scattering with remaining gas.
same vacuum level, x2 current KEKBx2, OK
SR from downstream
SR emitted at downstream magnets are back-scattered to IP.
Final Q for each ring (no bend near IP) Less than KEKB, OK
2-photon process
e+e-e+e-e+e- process, e+e- pairs hit pixel detector
 Less than requirement, OK
e+/e- from rad. Bhabha
e+/e- lose energy after rad. Bhabha, lost at IP downstream
g from rad. Bhabha
g’s are lost at IP+~10m and create neutrons
∝luminosity: KEKBx40  can be handled with detector upgrade
Beam-beam effect
Scattering at IP by the other beam field non-Gaussian beam tails
Simulation study is ongoing.
This table summarizes the other BG sources at SuperKEKB.
Beam-gas scattering, SR from downstream, 2-photon process, spent e+e- from radiative Bhabha are negligible or at most in the same order as KEKB, so they will not be a problem.
Gammas emitted from radiative Bhabha event will eventually hit downstream magnets and create neutrons by Giant dipole resonance, becoming dominant BG for outer KL-muon detector.
It should be proportional to luminosity, which means we will have 40times larger background,
but we can live with it with upgraded KL/muon detector which I will explain later.
Beam-beam effect is the scattering at IP from the field of the other beam.
This is basically a non-linear effect and creates non-Gaussian tails on beam shape.
It is still under simulation study by our accelerator colleagues, and we will need to see how much impact it may have on our BG estimation.
Feb. 24th, 2011
H.Nakayama (KEK)

14. Detector upgrade Now, moving on to the detector upgrade.
Lower half of this picture shows the former Belle detector, and the upper half the new Belle-II detector.
I will explain the upgrade of each sub-detector one by one. 14
Feb. 24th, 2011
H.Nakayama (KEK)

15. Detector upgrade Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
As for the vertex detector, current 4 layers of Si strip detector will be replaced by
2 layers of pixel detector and 4 layers of Si strip detector.
The readout chips will also be replaced with the ones with shorter integration time. 15
Feb. 24th, 2011
H.Nakayama (KEK)

16. Detector upgrade Vertex detector: 4lyr. Si strip
Drift chamber for tracking:
Small cells, longer lever arms, faster readout
Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
Belle
Belle II
Drift chamber for tracking will be upgraded with smaller cells to cope with higher BG rate.
The longer lever arm will enable better momentum resolution.
The readout system will be replaced with faster one, to reduce deadtime. 16
Feb. 24th, 2011
H.Nakayama (KEK)

17. Detector upgrade Vertex detector: 4lyr. Si strip
Drift chamber for tracking:
Small cells, longer lever arms, faster readout
Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
new PID system:
Cherenkov imaging, very fast readout
Particle identification system will be replaced by a completely new one,
using the concept of Cherenkov imaging and very fast photon sensor. 17
Feb. 24th, 2011
H.Nakayama (KEK)

18. Detector upgrade Vertex detector: 4lyr. Si strip
Drift chamber for tracking:
Small cells, longer lever arms, faster readout
Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
Calorimeter:
New readout with
wave form sampling
new PID system:
Cherenkov imaging, very fast readout
Calorimeter will be equipped with the new readout system, capable of wave form sampling which leads to the reduction of BG by factor of 7. 18
Feb. 24th, 2011
H.Nakayama (KEK)

19. Detector upgrade Vertex detector: 4lyr. Si strip
Drift chamber for tracking:
Small cells, longer lever arms, faster readout
Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
Endcap KL/muon:
RPC Scintillator +MPPC
Calorimeter:
New readout with
wave form sampling
new PID system:
Cherenkov imaging, very fast readout
KL/muon detector will be renewed from RPC to Scintillator+MPPC, to cope with increasing neutron background.
These are the main detector upgrade features, and
I’d like to explain a little further about the vertex detector and PID in the following few slides. 19
Feb. 24th, 2011
H.Nakayama (KEK)

20. Belle-II vertex detector
6th lyr.
5th lyr.
4th lyr.
3rd lyr.
2nd lyr.
1st lyr.
Si strip
Belle-II
Pixel: r=14/22mm
Si strip: r=38/80/115/140mm
pixel
Belle
Si strip: r=20/43.5/70/88mm
4lyr. Si strip
 2lyr. pixel(DEPFET) + 4lyr. Si strip
s[mm]
10mm
20mm
Belle
Belle II
pbsin(q)5/2 [GeV/c]
Pixel detector closer to IP
Improved decay vertex resolution
Increased radial coverage
 better Ks acceptance
As I mentioned before, we add 2 layers of pixel detector closer to IP, which brings about much better decay vertex resolution.
Thanks to the increased radial coverage, we will also have better acceptance for Ks.
Feb. 24th, 2011
H.Nakayama (KEK) 20

21. Belle-II Particle Identification System
Barrel PID: Time of Propagation Counter (TOP)
Quartz radiator
Thin quartz bar with very flat surface
Precise timing measurement with MCP-PMT
Aerogel radiator
Hamamatsu HAPD
+ new ASIC
Cherenkov photon
200mm
n~1.05
Endcap PID: Aerogel RICH (ARICH)
Focusing
mirror
MCP-PMT
B+r+g analysis
Completely different from Belle PID:
better K/p separation
more tolerance for BG
less material
(better calorimeter resolution)
Belle
Belle-II
Particle identification system will be fully replaced.
For the Barrel part, we will use so-called TOP counter which consists of thin quartz bars and photo-sensors with very precise timing resolution.
For the endcap part, we will use Aerogel-RICH which consists of layers of aerogel radiaters with different indices and HAPD equipped with fast readout.
The upgraded PID system will have much better K/pi separation, as shown visibly in these two plots.
The new system will have more tolerance for BG,
It will be less material, which means better resolution of calorimeter placed just outside PID.

22. Summary SuperKEKB accelerator Beam background Belle-II detector
x40 higher luminosity than KEKB
Beam background
Touschek scattering increases drastically at SuperKEKB
Belle-II detector
Upgrade of vertex detector, PID system, etc..
SuperKEKB/Belle-II will start in 2014.
Belle-II Technical Design Report
It seems the time is running out.
Summing up,
Our accelerator will be upgraded with x40 higher luimnosity,
Touschek scattering will be the dominant background,
Our new detector, “Belle-II” is expected to achieve better performance and with capability to handle higher background.
Wait and see the new era of Super B factory, starting from 2014.
Feb. 24th, 2011
H.Nakayama (KEK)

23. Thank you!!
Feb. 24th, 2011
H.Nakayama (KEK)

24. Backup slides
Feb. 24th, 2011
H.Nakayama (KEK)

25. Machine Design Parameters
KEKB
SuperKEKB
units
LER
HER
Beam energy Eb
3.5 8 4
7.007
GeV
Half crossing angle φ 11
41.5
mrad
# of Bunches N
1584
2500
Horizontal emittance εx 18 24
3.2
5.3 nm
Emittance ratio κ
0.88
0.66
0.27
0.24 %
Beta functions at IP
βx*/βy*
1200/5.9
32/0.27
25/0.30 mm
Beam currents Ib
1.64
1.19
3.6
2.6 A
beam-beam param. ξy
0.129
0.090
0.0886
0.081
Bunch Length sz
6.0
5.0
Horizontal Beam Size
sx*
150 10 um
Vertical Beam Size
sy*
0.94
0.048
0.062
Luminosity L
2.1 x 1034
8 x 1035
cm-2s-1

26. Belle-II: other detector upgrades
Drift chamber: smaller cells to cope with higher BG, longer lever arm for better momentum resolution, faster readout to reduce deadtime
Silicon vertex detector:
New readout chip (APV25)
shorter integration time(800 ns50 ns)
Belle
APV25
chips
Belle II
Calorimeter: new readout system with waveform sampling (x1/7 BG reduction)
KL/Muon detector
RPCScintillator+MPPC
Better performance against neutron BG
Feb. 24th, 2011
H.Nakayama (KEK)

27. DEPFET:
DEPFET
Each pixel is a p-channel FET on a completely depleted bulk
A deep n-implant creates a potential minimum for electrons under the gate (“internal gate”)
Signal electrons accumulate in the internal gate and modulate the transistor current (gq ~ 400 pA/e-)
Accumulated charge can be removed by a clear contact (“reset”)
Fully depleted: => large signal, fast signal collection
Low capacitance, internal amplification: => low noise
High S/N even for thin sensors (50µm)
Rolling shutter mode (column parallel) for matrix operation => 20 µs frame readout time
=> Low power (only few lines powered)

28. DEPFET for Belle II radius pixel thickness
Layer 1 r = 14mm 50x50mm2 75mm(0.18%X0)
Layer 2 r = 22mm 50x75mm2 75mm
total of 8 Mpx
Mechanical mockup
Power consumption in sensitive area: 0.1W/cm² => air-cooling sufficient

29. SVD HPK resumed DSSD production with 6’’ wafers
Two types of rectangular sensors and
Trapezoidal sensors for slanted part (still under discussion)

30. Flex PCBs and APV25 Origami PCB PA1 PA2 PA0 glued onto hybrid
3-layer design
237µm thick (nominal)
PA0, PA1 and PA2
2-layer design
145µm thick (nominal)
Thinned APV25
(300mmg~105mm)

31. Expected performance Significant improvement in IP resolution! s[mm]
Pixel detector close
to the beam pipe
Less Coulomb
scatterings
Belle
s[mm]
s[mm]
Belle II’
Belle II
20mm
10mm
1.0
2.0
1.0
2.0
pbsin(q)3/2 [GeV/c]
pbsin(q)5/2 [GeV/c]
p+ p-
Ks track
IP profile
B vertex g
B decay point reconstruction
with KS trajectory
Larger radial coverage of SVD
Significant improvement in dS(KSp0g)

32. Central Drift Chamber Belle Belle II longer lever arm
improve resolution of momentum and dE/dx
Belle
Belle II
Belle
Belle II
inner most sense wire
r=88mm
r=168mm
outer most sense wire
r=863mm
r=1111.4mm
Number of layers 50 56
Total sense wires
8400
14336
Gas
He:C2H6
sense wire
W(Φ30μm)
field wire
Al(Φ120μm)
10 mm
6~8 mm
small cell
normal cell
10~20 mm
18 mm

33. Time of Propagation Counter (TOP)
Quartz radiator
2.6mL x 45cmW x 2cmT
Excellent surface accuracy
MCP-PMT
Hamamatsu 16ch MCP-PMT
Good TTS (<35ps) & enough lifetime
Multialkali photo-cathode  SBA
Barrel PID: Time of Propagation Counter (TOP)
Quartz radiator
Focusing mirror
Small expansion block
Hamamatsu MCP-PMT (measure t, x and y)
TOF
TOP
MCP-PMT
～1.2m

34. TOP (Barrel PID) Beam test done in 2009 # of photons consistent
Beam spot
875mm
915mm
quartz
3rd
2nd
1st
TOP (Barrel PID)
Beam test done in 2009
# of photons consistent
Time resolution OK
Quartz Radiator
Hamamatsu MCP-PMT
1st
2nd
3rd
[1count/25ps]
Time resolution
beam data
Simulations
# of photons

35. End-cap Particle Identification
Ring imaging Cherenkov counter based on silica aerogel radiator
Space limited -> proximity focusing with expansion distance of 20 cm
Requirements
Transparent silica aerogel
Photo-detector
Single-photon sensitivity
Pixel 5x5mm2
Operational in 1.5T
Compact
Readout electronics
~70K channels
２０ｃｍ
n=1.05 aerogel
|sc(p)-sc(K)| = 23 mrad at P=4GeV/c
Target: More than 4 s p/K separation at 4 GeV/c

36. Novel “focusing” radiator
Possible only with aerogel !
Simple accumulation of radiator layer gives more photons, but degrades PID performance
 Multiple aerogel layers with different indices
4cm-thick single index aerogel
sq(1p.e.) = 22 mrad
Npe ~ 10.6
sq(track) = 6.9 mrad
Focusing by 2cm+2cm aerogel (n1:1.047, n2:1.057) n1 n2
sq(1p.e.) = 14.4 mrad
Npe ~ 9.6
sq(track) = 4.8 mrad
NIM A548(2005)383

37. Expected Performance (Luminosity gain)
No upgrade
BAD
Upgrade
GOOD
No upgrade
BAD
Upgrade
GOOD
B0gr0g
FWD BRL
ACC only
dE/dx
only
As good as Belle
A-RICH
TOF, dE/dx NA
–74%
–69%
–68%
–62%
TOF NA
–41%
–35%
–32%
–22%
–10%
–4%
0% (definition)
+12%
TOP opt.0
+27%
+33%
+40%
+59%
TOP opt.2
+45%
+51%
+60%
+83%
Completely different world with excellent PID detectors!

38. Electromagnetic Calorimeter (ECL)
For Day 1
Electromagnetic Calorimeter (ECL)
Upgrade electronics to do waveform sampling & fitting
Upgrade crystals for end caps (pure CsI + photomultiplier as the baseline)

39. New shaper/readout boards
Basically ready for mass production
(minor revisions because we still have time.)
One of old versions has been installed in Belle to readout part of the end cap in Tested OK.

40. KLM MPPC: Hamamatsu 1.3×1.3 mm 667 pixels (used in T2K ND)
fiber: Kuraray Y11 MC
Scintillator bar: Vladimir (Russia)
(used in T2K ND)
part II
part I

41. 41
Feb. 24th, 2011
H.Nakayama (KEK)

42. BG extrapolation toward SuperKEKB
Sugihara(Tokyo)
BG extrapolation toward SuperKEKB
KEKB machine study (Jun. 2010)
Single beam (no collision), measure detector background
Separate Touschek BG and beam-gas BG, using beam-size dependence
kbeam-gas, tbeam-gas, kTou, tTou are measured
Extrapolation to SuperKEKB：
The KEK value is assumed for kBeam-gas, kTou, and tBeam-gas. For tTou , SuperKEKB design Value (10min) is used.
Extrapolated BG (Belle-II+SuperKEKB)
ECL
~9 GeV/event, OK
(introduce 5us time window)
TOP
~2.9MHz (~=3MHZ) TOF rate eqiv., 
(replaced from current TOF)
CDC
~84kHz/wire (<200kHz/wire), OK
(larger radius, more cell number)
SVD
~1.2% occupancy (<6.6%), OK
(shorter integration time)
PXD
~2.0% occupancy (~=2%), 
(not including low pt particle & few keV gamma)
This extrapolation is based on life time and beam current. Upgraded performance of each detector is taken into account. We assume the same level of collimation by movable masks at the ring part, and the same amount of shielding in the IR region.
Only Touschek and beam-gas BG are considered. Other BGs are not included (SR, QED, beam-beam, etc…).
CDC 42

43. Heavy-metal shield(KEKB)
1-1.5cm thick

44. Lead shield on QCS cryostat (KEKB)
QCS-L cryostat

45. Neutron shield (KEKB)
Polyethylene shield
Concrete wall

46. Neutron shield
Feb. 24th, 2011
H.Nakayama (KEK)

47. Detector upgrade Vertex detector: 4lyr. Si strip
Drift chamber for tracking:
smaller cells
Vertex detector: 4lyr. Si strip
 2lyr. pixel(DEPFET) +4lyr. Si strip
Endcap KL/muon:
RPC Scintillator +MPPC
Calorimeter:
New readout with
wave form sampling
new PID system:
Cherenkov imaging, very fast readout 47
Feb. 24th, 2011
H.Nakayama (KEK)