1. X. -HPD Optical Module R&D for KM3NeT (
X*-HPD Optical Module R&D for KM3NeT (*Crystal-scintillator viewed by small PM)
I. Al Samarai*, J. Busto*, B. Combettes*, A-G Dehaine*, D. Dornic*, F. Fouche*, G. Hallewell* *Centre de Physique des Particules de Marseille *Photonis S.A. Brive la Gaillarde
Building upon the experience of the Philips (Photonis) ‘Smart’ Tube (Flyckt – Van Aller)
The Baikal Quasar370 of NT-200 and the spherical X-HPD R&D at CERN (Joram et al)
SMART Photonis – CPPM 2008 prototype
Baikal
CERN
Our R&D carried out under the GIS (Groupement d’interet Scientifique)
contract between CNRS-IN2P3 and Photonis
Our Aim: to increase the Cherenkov Photon Horizon in sea water through the electrostatic and
Quantum efficiency enhancements offered by X-HPDs, and to minimize X-HPD unit cost -
to increase the volume of water instrumentable in a neutrino telescope of a given budget
KM3NeT WP3 meeting Oct 15, 2008

2. From… Presentation by B
From… Presentation by B. Combettes (Photonis) Identified as a major Photonis priority for n detectors NNN07, Hamamatsu, Japan, October 2007 Reiterated at NNN08, APC, September 2008

3. Presentation What is the X-HPD? History of the X-HPD;
X-HPD advantages – why develop it?
Who are the actors?
Present status inc R&D at CPPM;
Future developments
KM3NeT WP3 meeting Oct 15, 2008

4. First of all, what it ISN’T…
What is the X-HPD?
First of all, what it ISN’T…
KM3NeT WP3 meeting Oct 15, 2008

5. It isn’t like the LHC-b HPD (DEP Photonis) shown here…
80mm
The X-HPD :
is big, like a hemispherical PM;
has standard bi-alkali p.c. and borosilicate window;
has no electronics in the vacuum volume;
is compatible with internal photocathode processing like PM;
is insensitive to the Earth’s magnetic field;
(photoelectrons accelerated in ~25kV)
120mm
Under vacuum, the X-HPD contains (only…) :
glass,
chemical products + evaporators for the photocathode,
scintilating crystal, encapsulated in aluminium,
electrodes, pins and wires,
getters, as necessary.
It is SIMPLE and already proven -200 in operation at Lake Baikal
KM3NeT WP3 meeting Oct 15, 2008

6. What is the X-HPD? A “simple” detector of single
photons with a bialkali p.c.
High accelerating field (~25kV) between p.c. & scintillating crystal (short Tdecay) under thin metallic layer (Al, 100nm)
Preferably a completely
spherical geometry, with
scintillator at the centre
Eff. for electrostatic collect.+ (Standard Bialkali) Q.E
Global Efficiency ≥33% (40% av.~50% max)
 (cf ~16% hemispherical PM)
KM3NeT WP3 meeting Oct 15, 2008

7. The first crystal-HPD: Philips SMART : Phosphor P47 (YSO:Ce) ~30 manufactured (1980s-1992)
History of the X-HPD
G. van Aller et al.
A "smart" 35cm Diameter Photomultiplier. Helvetia Physica Acta, 59, 1119 , (1986).
KM3NeT WP3 meeting Oct 15, 2008

8. Quasi-internal photocathode deposition (NOT A TRANSFER PROCESS)…!
8 grains of Sb
(deposition of PC)
P47 (YSO:Ce) phosphor
Deposited on glass
Sb deposition shield
(+25kV)
Domed +25kV surface
Quasi-internal photocathode deposition
(NOT A TRANSFER PROCESS)…!
KM3NeT WP3 meeting Oct 15, 2008

9. Internal details of SMART 15” (from Photonis Archives)
Electrodes + feedthroughs
for antimony evaporation
Antimony screen
(limits solid angle
of p.c.deposition)
+25kV *
* Ports for connections to
in-tube evaporators for K, Cs 0V 0V
Photonis Stargate Meeting 12/9/2007
KM3NeT WP3 meeting Oct 15, 2008

10. The Philips SMART Tube The QUASAR 370 Tube
pressure sphere
R. Bagduev et al., Nucl. Instr. Meth. A 420 (1999) 138
PMT
XP2982
G. van Aller et al. A "smart" 35cm Diameter Photomultiplier.
Helvetia Physica Acta, 59, 1119 ff., 1986.
gprimary ~ 35, st ~ 2.5 ns / Npe
KM3NeT WP3 meeting Oct 15, 2008

11. Baikal Quasar-370 (1983) Hybrid (X-HPD) -Hemispherical dome =370 mm
-Photocathode K2CsSb
-Preamplification 25 kV
Scintillator
Y2SiO5 (TTS= 2ns FWHM)
Traditional PMT
K2CsSb 13-stages =25 mm
gprimary ~ 35, st ~ 2.5 ns / Npe
1 photoelectron on main photocathode  20…30 photoelectrons on small PM

12. Baikal Quasar 370
X-HPD history

13. Baikal Quasar 370
G. Hallewell: Centre de Physique des Particules de Marseille
Photonis Stargate Meeting 12/9/2007

14. Baikal Quasar 370
Multi-ports
(vacuum +
p.c. deposition)

15. 200 tubes in operation since 1996
BAIKAL
Quasars in operation since 1993:
200 since 1998
R. Bagduev et al., Nucl. Instr. Meth. A 420 (1999) 138

16. The present Actors – an expanding group
Photonis (B. Combettes, F. Fouche, A-G. Dehaine
(relations with Fibrecryst & St Gobain for non organic scintillator crystals
+ IPNO group (Joel Pouthas et al) for crystal characterisation.
INFN Genova (Mauro Taiuti, D. Bersani)
CERN (Christian Joram, Andre Bream,Jacques Seguinot) (collaboration wih Photonis – also in-house crystal characterisation)
CPPM (Imen Al Samarai*, Greg Hallewell, Jose Busto +..)
development contract under GIS IN2P3-Photonis, *funded PhD
INR Moscow / Univ. Tuebingen
Bayarto Lubsandorzhiev {ex Baikal Quasar 370} + 50% Univ Tuebingen
(now a KM3NeT institute) interested to continue development of
SMART /Quasar concept under agreement signed with Photonis
(2006) on collaboration in joint photodetector development.
KM3NeT WP3 meeting Oct 15, 2008

17. X-HPD history Proposal for NEMO X-HPD with directionality
(Mauro Taiuti, Marco Battaglieri et al;
INFN Genova)
X-HPD with directionality
First glass prototype
by Soffieria Sestese
Metal Flange coupling
YSO Scintillator +
plexiglass lightguides
4 x Hamamatsu R /8”

18. SMART & Quasar were the first X-HPD tubes
But scintillator in the form of disk didn’t fully exploit the potential…
Preferable to use a 3-D or ‘volume’ scintillator at the exact geometric
centre of sph. envelope (improved iso-chronicity + global efficiciency
{(# photoelectrons seen)/ (# photons arriving)} )
PMT
XP2982
KM3NeT WP3 meeting Oct 15, 2008

19. Toward Development of a large spherical X-HPD hybrid photodetector
CERN - Photonis:
Toward Development of a large
spherical X-HPD hybrid photodetector
Braem +, C. Joram +, J. Séguinot +, L. Pierre *, P. Lavoute *
+ CERN, Geneva (CH) * Photonis SAS, Brive (F)
KM3NeT WP3 meeting Oct 15, 2008

20. Advantages of a spherical tube with anode at geometric centre
Radial electric field
s TTS negligeable
Efficiency of electrostatic collection
~100% over ~ 3p Sr
Mu-metal screen unnecessary
Large detection solide angle (dW ~ 3p)
Gain in effective QE sensitivity by ‘Double-
passage’ effect through photocathode
To be compared with : ~ 70% over only 4p/3 Sr
in a hemispherical PM!
Nesessary for a hemispheric PM
X2 compared to a hemispheric PM QE QE
KM3NeT WP3 meeting Oct 15, 2008

21. 8’’ Prototype with anode in form of a metal cube (1 cm3) ‘Proto 0’
measured at Photonis
ORDINARY BI-ALKALI PHOTOCATHODE !!
Fabrication CERN .
A. Braem et al., NIM A 570 (2007) 
KM3NeT WP3 meeting Oct 15, 2008

22. From Joram: Crystal arrangements for X-HPD CERN ‘X-Proto 1’
Anode consists of Al-plated LYSO crystal :12 mm Ø × 18 mm
electrical feedthrough
Al coated
PMT: Photonis XP3102
(25 mm Ø)
KM3NeT WP3 meeting Oct 15, 2008

23. equatorial illumination
Extensive remeasurements (June & Sept 2006) of Q.E. of 4 15’’ SMARTS & 1 Quasar acquired by CPPM in early stages of ANTARES
virtual ground linked to Keithley picoammeter scintillator +500700V)
Tests with polar &
equatorial illumination

24. even with this non-optimal electrostatic configuration
Polar & equatorial Q.E. measurements on 4 15” SMARTS + 1 Quasar (June, Sept 2006)
We see: Q.E. of 3 of 4 SMARTS still OK, Equatorial enhancement in standard bialkali Q.E.,
even with this non-optimal electrostatic configuration

25. Further comments on QE of reflective mode photocathode
Semitransparent UBA cathodes now
clawing 40-45%: not much more headroom
- 50% a natural limit, since no significant
electric field (< 100V/cm) inside the cathode
bulk in a standard large hemispherical PM
 p.e. extraction by random walk with 50%
p.e.’s walking to wrong hemisphere; QE QE
Reflective photocathodes have higher QE
than semi-transparent (CsI – GEM experience)
Can be deposited ‘for free’ on aluminium
photocathode equipotential conductor
surface - which serves no useful detection
function in a large PMT -
Significant P/V degredation
(probably also pre-pusing
and after-pulsing)
in SBiAl cf standard BiAl
In a focussed X-HPD can trade some extra
solid angle coverage for much greater QE
in sensitive solid angle zone:
Al layer keeps more photons in photocathode
Extraction field >> than in large hemi. PM
KM3NeT WP3 meeting Oct 15, 2008

26. Example of large X-HPD in 17’’ pressure sphere for KM3NeT:
380 mm
Up to ± 120° acceptance
Sensitive to single photons
TTS 1-2 ns FWHM
Q.E. optimised 300 < l < 600 nm
dark counts <0.1 per 100 ns
Pressure Sphere
17”(432 / 404)
Optical gel
(matching ref. ind.
+ coupling to hemisph) 15
Scintillating
Crystal
Si sensor
432 mm
(17”)
joint
Glass
Support
ceramic support
Small PM
Example of large
X-HPD in 17’’ pressure
sphere for KM3NeT:
smaller sizes
clearly possible HV PA
electrical feed-throughs
valve
KM3NeT WP3 meeting Oct 15, 2008

27. Why develop X–HPDs for KM3NeT?
Significant uncertainties in the rate
of n production at E > ~1014eV ; n telescopes at ~1km3 scale (ICE CUBE) may be nowhere near big enough
Probably need ~10 km3 at reasonable OM price:
Extend OM Cherenkov horizon or fall into a ‘hole’
where the only detector we can afford to build
will be one to detect GZK n’s acoustically…
KM3NeT WP3 meeting Oct 15, 2008

28. Neutrino effective area estimation
Damien Dornic, CPPM
HPDs with 40%
overall detection
(instead of 23%
for standard PMs)
efficiency flat over 3p
3 X-HPDs in place of 3 standard PMs (orientation - 45°)
2 X-HPDs horizontal
“Peanut” geometry
(opposing direction), each storey rotated 90° w.r.t. previous)
Base configuration: hexagon of 127 lines (~0.35 km3)
 separated by 85m. Each line 34 storeys spaced at 15m
ANTARES site parameters
KM3NeT WP3 meeting Oct 15, 2008

29. Angular resolution comparison
HPDs with 40%
overall detection
(instead of 23%
for standard PMs)
efficiency flat over 3p
3 X-HPDs in place of 3 standard PMs (orientation - 45°)
2 X-HPDs horizontal
“Peanut” geometry
(opposing direction), each storey rotated 90° w.r.t. previous)
KM3NeT WP3 meeting Oct 15, 2008

30. Effective Surface Ratio: X-HPDs compared to 3 x 10” PMTs/storey
KM3NeT WP3 meeting Oct 15, 2008

31. Evaluating the advantages of Spherical format X-HPDs:
Volume /cost sensitivity comparison with 10” Hamamatsu R used in ANTARES
Assumptions
~350 Cherenkov photons per cm ( nm)
Attenuation length 35m (combined absorption, scattering)
Photon flight 40m/sinqC (qC = 43°) =56m
Cost per ANTARES OM =
PM: 995€, electronics: 600€, mechanics: 660€, sphere+other 600€
Cost per X-HPD OM
1.5* cost of ANTARES tube of same diameter
+electronics 600€ + sphere etc. 600€ + mechanics 660€
therafter scaled as ratio of areas for different diameters
KM3NeT WP3 meeting Oct 15, 2008

32. ANTARES PM Practical diameter limit to fit in 17’’ pressure sphere
Assumptions:
Price of Hamamatsu (20cmØ ~16% overall eff): 995€
Sphere etc. +mechanics + electronics = 2000 € / optical module
Price of 20 cm Ø p.c. X-HPD = 1.5* PM Hamamatsu
thereafter price follows ratio of p.c. surface area

33. KM3NeT WP3 meeting Oct 15, 2008

34. Photonis has built 3 prototypes of a new SMART (8”)
Photonis has built 3 prototypes of a new SMART (8”) - Internal deposition of the photocathode - Metal disk - measure of photocurrent
KM3NeT WP3 meeting Oct 15, 2008

35. Test stand for metal anode prototypes
Equipment:
Spellman 25 KV reversible power supply;
Black boxes;
4 channel reversible 6KV supply;
Keithley 485 picoammeter;
LED sources;
Ventouse for positioning light source
on photocathode
Also Simion-8® 3D electrostatic
simulation program
Will be soon extending stand for crystal prototypes
for single photon counting (TTS, pulse height) measurements

36. Some measurements on first 8’’ prototype (metal disk anode) summer ‘08
Cartography of photocathode
with ventouse and 475nm LED
(polar angle ‘double cathode’
enhancements seen – and also some
azimuthal non-uniformties) nA V
q=0°
q=15°
q=30°
q=37.5°
 F 
8 “ Prototype XP2607
40° polar angle limit
Plateau characteristic:
LED current set for ~ 1nA
on XP2608 at Plateau voltage
(measured on Keithley 485 p-ammeter)
~ 30 ph in 50 ns window nA
Internal photocathode
generators kV
Dark current limitation: next
Prototype has better pin shielding
KM3NeT WP3 meeting Oct 15, 2008

37. 3-D (Simion-8 ®) electrostatic simulation of next Photonis 8” X-HPD
with 60° polar angle medal anode prototype – prototype to be tested mid Oct
Present anode geometries not optimal: plate not near focal point of envelope
(will change this for forthcoming crystal anode prototypes
and next metal anode versions)
We see however that sub-ns Transit time spread would be reached at~ 5kV
(higher voltages will be needed for scintillation light production in crystal)
Photonis have identified the glass envelope
and internal p.c. deposition as critical to cost –
First prototypes (above): all glass envelope & fully internal bialkali photocathode deposition
Mid 2008: Fabrication & test of prototype 8” tubes with metal anode WE ARE HERE!
First Crystal based proptype from Photonis expected Late November
2008/9: Fabrication & test of 8” HPD tubes with crystal anodes in different configurations
Measurement program for of TT, sTT, HT behaviour, l-integrated photocurrent, multi-g res…
2009/2010: Fabrication and test of prototype 15” X-HPD with best crystal anode

38. SIMION 8 simulations (CPPM) 4/08:
Centred hemispheric anode on degraded glass support
R (photocathode) 100 mm
0 kV
R(anode) 10 mm
25kV
20 mm diam.
60mm high glass support
With degrader to 0V
on outside
(simulated conductive paint
used in latest Joram prototype)
No other conductive
surfaces modelled
TTD North Pole – Equator
0.0003ms
KM3NeT WP3 meeting Oct 15, 2008

39. Scintillator Considerations
SMART and Quasar tubes used YSO phosphor disk:
Geometry didn’t allow to exploit full potential.
need to use fast 3-D scintillator shape
Require
high light yield  high gain
short decay time
Low Z preferable  low back scattering
Emission around l = 400 nm
LY (g / keV)
t (ns)
Zeff
eBS
lemission (nm)
YAP:Ce 18 27 32
~0.35
370
LYSO:Ce 25
~40 64
~0.45
420
LaBr3:Ce 63 30 47
~0.4
360
Joram,
1st Photonis prototype
for CPPM
KM3NeT WP3 meeting Oct 15, 2008
LaBr3 is quite hygroscopic

40. What is the best/cheapest crystal shape
(given that 25kV p.e.’s only penetrate a few microns)

41. What is the best/cheapest crystal shape
(given that 25kV p.e.’s only penetrate a few microns)

42. Near future developments
Photonis have identified the technique for p.c. deposition as critical to cost – progress in ‘internal deposition’
CERN-Photonis will test the 8” pure spherical geometry
X-HPD with LYSO crystal (for NDIP2008, June 2008)
Parallel crystal studies (CERN-Photonis, IPN Orsay)
At CPPM, measurements of metal anode prototypes
+ LYSO crystal type (photocurrent, sTT, HT behaviour, multi-g res…)
Mid 2008: Fabrication & test of prototype 8” tubes with metal anode
End 2008/9: Fabrication & test of 8” HPD tubes
with crystal anodes in different configurations
2009/10: Fabrication and test of prototype 15” X-HPD
with best crystal anode
KM3NeT WP3 meeting Oct 15, 2008

43. Some Longer term developments
Progress from 8’’ prototypes to larger, up to original 15’’ of Flyckt/Val Aller SMART?
(Photocathode deposition  internal or quasi-internal processing: Cost/unit much lower than hemispherical PM of same Ø*QE*p.e.coll.eff.)
In parallel, tests of an 8” prototype in a sphere at ANTARES/NEMO/NESTOR
(need to pass to ‘weaponisation’ phase: compatibility with power & r/o system)
Development from Baikal 24V25kV DC-DC supply:
Low power Cockroft-Walton multiplier: reliability issues,
Can we build from designs of TV tube HV supplies (compatibility of reqt?)
(huge statistical MTBF sample + Photonis-Philips connection)
KM3NeT WP3 meeting Oct 15, 2008

44. Conclusion Pure spherical geometry X-HPDs based on Flyckt/Van Aller
Philips SMART tube could be very promising Čg – detectors for underwater n telescopes
~100% photoelectron collection eff. (over 3p Sr) (& sTT<1ns )
(c.f. 65% over ~ 4p/3 cathode surface area in ‘hemispherical’ PM)
Much larger photocathode area than same Ø hemispherical PM
+ suppression of m-metal cage
Improved Q.E. (transmissive  reflective) over large % of 3p
>50% with standard Bialkali, enhanced overall detection effiency
Improved multi-g sensitivity
Cost savings through extended Cherenkov horizon:
larger water volume instrumented with given number of OMs
KM3NeT WP3 meeting Oct 15, 2008

45. Fin

46. Photon detectors for water n detectors Large surface area * high Q.E.
+ good 1g, 2g…? separation
or (2): ‘Cluster’ of small PMTs
(ou alternative)
Classical ‘hemispheric’ PMT
Mushroom dome,
simple anode,
or possibly multi-anode
(3) Hybrid (X-HPD) :
Acceleration +
Scintillator
Optical background (ANTARES experience)
(single gs in a 20ns counting interval)
40K decays (b’s giving Č light)
+ Bioluminescence
1g Signal Č on 1g ’’bioK’’
Sometimes 2g Č for m’s close (~10m) to sensors