IEEE NSS-MIC 2011, Valencia, Spain 1 Large Area Ultra-Thin Detector Ladders based on CMOS Monolithic Pixel Sensors Wojciech Dulinski IPHC Strasbourg, France on behalf of PLUME, SERVIETTE and CERNVIETTE Collaborations on behalf of PLUME, SERVIETTE and CERNVIETTE Collaborations Outline  Short status of MAPS development at IPHC  PLUME: the lightweight ladder based on standard flex PCB  SERVIETTE and CERNVIETTE: ultra-thin chip embedding in plastics project (two different process approach)  Conclusions and prospects

IEEE NSS-MIC 2011, Valencia, Spain 2 Monolithic Active Pixel Sensor: effective use of a thin epitaxial layer (10 – 20 µm) for MIP tracking R.T. Fast and more efficient charge collection  radiation tolerant MAPS pastpresent Industrial availability of high resistivity substrate (epi) in a standard CMOS process

IEEE NSS-MIC 2011, Valencia, Spain 3 Present status of MAPS: just two examples of mature designs Mimosa 26 Binary, sparsified readout sensor for EUDET beam telescope > 2 cm 2 active area, 0.7 Mpixel tracker - Medium speed readout (100 µm integration  10 kFrame/s) - Spatial resolution < 4 µm for a pitch of 18.4 µm - Efficiency for MIP > 99.5 % - Fake hit rate < Radiation hardness > n/cm 2 (high resistivity epi substrate) - Easy to use, “off-shell” product: used already in several application Ultimate: 4 cm 2 sensor for STAR Microvertex upgrade - similar measured tracking performance as Mimosa26 - Radiation hardness >10 14 n/cm 2 with CMOS MAPS attainable? First indications: yes!

IEEE NSS-MIC 2011, Valencia, Spain 4 Potentially extremely thin (~25 µm of silicon in total, ~0.027 % X 0 ), flexible (!) and still fully efficient for MIP tracking! Problem: how to handle, interconnect and at the end built a low mass ladder with such a thin device? One of the main feature of MAPS

IEEE NSS-MIC 2011, Valencia, Spain 5 First real scale exercise: new STAR Microvertex Detector Estimated 0.37% X 0 /ladder. Can we do better? Data taking (1/4 of detector) expected in 2013, full detector installation in 2014

IEEE NSS-MIC 2011, Valencia, Spain 6 Three RD mini-collaborations initiated by IPHC ~2 years ago, in order to develop new methods of ultra-thin ladders construction PLUME, SERVIETTE and CERNVIETTE Collaborations 1 IPHC/IN2P3 Strasbourg, France 5 University of Bristol, UK 6 DESY, Hamburg, Germany 7 University of Oxford, UK 2 IMEC, Leuven, Belgium 3 CMST, University of Gent, Belgium 4 IFK, Goethe University, Frankfurt/M, Germany 8 CERN, Geneva

IEEE NSS-MIC 2011, Valencia, Spain 7 PLUME concept: double-sided ladder (ILC compatible) -2x6 Mimosa26 sensors thinned down to 50 µm -Standard double-side kapton PCB: Cu conductor (20 µm/layer) -SiC foam (8%) for spacer between layers -Estimated 0.6 % X 0 /two sensor layers

IEEE NSS-MIC 2011, Valencia, Spain 8 bare low mass cable module with 6 sensors bare low mass cable complete ladder (2 modules) PLUME prototype: assembling steps

IEEE NSS-MIC 2011, Valencia, Spain 9 Heat dissipation: moderate air blow seems to solve a lot of “hot spots” problems Electrical parameters (threshold dispersion, fake hit rate) almost unchanged…

IEEE NSS-MIC 2011, Valencia, Spain 10 Current status of PLUME prototype: ready for beam tests next week!

IEEE NSS-MIC 2011, Valencia, Spain 11 - Partner restricted PTW Oct HUMAN++  Stands for : U LTRA T HIN FILM C HIP P ACKAGING Polyimide 1 Polyimide 2  In short :  Off-the-shelf die  Thinned down to ± µm  Packaged between two polyimide foils  Metallisation : fan-out  Circuit contact through vias  Result :  Flexible package, no wire bonding  Thin : µm  Embeddable in commercial flexible PCB SERVIETTE project: use of UTCP by IMEC…

IEEE NSS-MIC 2011, Valencia, Spain 12 Placement (face up) of IC Polyimide on rigid carrier with release layer (KCl) Dispense/spin of BCB Chip thinning Metallization: TiW (50nm) + Cu(1µm) Electroplating : Cu (5µm) Lithography to pattern metal Encapsulation polyimide spinning Release from carrier Photo definable polyimide spinning (20µm)) Opening vias using lithography Cleaning of contact pads UTCP flow: overview 60  m

IEEE NSS-MIC 2011, Valencia, Spain 13 First results: Mimosa18 mechanical grade sample Submitted for fabrication more than a year ago, very slow progress since, on stand-by till thinned chip placement problems solved. Aluminum conductor, multiple metal layers and multiple chips still far ahead…

IEEE NSS-MIC 2011, Valencia, Spain 14 CERNVIETTE: use of a “standard” flex PCB process for chip embedding in plastic foils (Rui de Oliveira, Serge Ferry) 14 Gluing between two kapton foils Metallization: Al (5-10 µm) Lithography to pattern metal Opening vias using lithography Gluing of another kapton foil for deposition of second metal layer Single module: intermediate tests Complete ladder assembling, laser cut along sensor edges

IEEE NSS-MIC 2011, Valencia, Spain Thin layer of epoxy glue 3 to 10 µm Copper Copper substrate 1.5mm Acrylic glue Polyimide Acrylic glue Silicon chip Polyimide acrylic glue 60 µm 150 µm CERNVIETTE: stack formation

IEEE NSS-MIC 2011, Valencia, Spain 16 1-Gluing at 200 deg, 22kg/cm 2 under vacuum 2-Photolithographic method and chemical etching by ethylene diamine to create vias on top of bonding pads 3-Plasma etching of glue in vias 4-Aluminium coating by sputtering deposition in vacuum machine CERNVIETTE: processing steps

IEEE NSS-MIC 2011, Valencia, Spain 17 5-Photolithographic method and etching with phosphoric acid to pattern aluminium layer (strip, pads). Step 1 to 5 can be repeated to create more metal layers and interconnexions 6-Chemical Etching of copper substrate CERNVIETTE: processing steps

IEEE NSS-MIC 2011, Valencia, Spain 18 CERNVIETTE: solid state flexible sensor wrapped over cylindrical shape (R=20 mm) and pretty well protected

IEEE NSS-MIC 2011, Valencia, Spain 19 Major failure: too short plasma etching of glue layer, no electrical contacts… But excellent metal adhesion and thickness uniformity! Second (corrected) iteration expected to be ready next week

IEEE NSS-MIC 2011, Valencia, Spain 20 CERNVIETTE: details of 4 metal layer flex (~0.12% X 0 ) Impedance of readout lines (last metal, 100 µm width, 100 µm gap) as a function of kapton thickness: 100 Ω for 60 µm thick kapton (last layer)

IEEE NSS-MIC 2011, Valencia, Spain 21 -Construction methods of ultra-light sensor ladders are progressing rapidly, embedding in polymer seems to be a new interesting option -It is at present our preferred solution in order to take full profit from MAPS high precision tracking performances AND to equip with a reasonable budget large area detectors (replacement of silicon strips, tracking calorimeters…) Conclusions Outlook -Double-sided PLUME module (0.6 % X 0 ) fully operational, ready for beam tests starting next month. The next version (Al instead of Cu conductor, less dense SiC foam  0.3 % X 0 ) in Second iteration of single Mimosa26 embedded in kapton (CERNVIETTE) next week, full PLUME compatible ladder (six M26) planned for beginning of the next year -CERNVIETTE process may by much less critical, if the last metal layer on the chip is used for better (less dense) “bonding pads” distribution  7 metal CIS 0.18 µm process available! -Start to think about stitching exercise: wafer scale, up to 10x10 cm 2 monolithic sensor possible