International Particle Accelerator Conference6thMay 3-8, 2015 Richmond, VA, USA

PROTON BEAM APPLICATIONS FOR SILICON BULK MICROMACHINING

P. Nenzi1, F. Ambrosini1, M. Balucani3, G. Bazzano1, A. Klyshko3, F. Marracino1, L. Picardi1, C. Ronsivalle1, C. Snels2, V. Surrenti1, M. Tucci2, M. Vadrucci1

1ENEA Frascati Research Center, Frascati, Italy, 2ENEA Casaccia, Roma, Italy, 3Università di Roma « La Sapienza », Roma, Italy

AbstractWe have investigated the effects of deep hydrogen implantation into n- and p-type silicon wafers ((100) oriented, with resistivity in the 1-20 Ω·cm range). Deep implantation has been achieved using the Hitachi-AccSys PL-7 RF LINAC set for 3.0 MeV beam energy, degraded to 1.8 MeV. Hydrogen has been implanted 30 µm below the wafer surface with an implant dose (fluence) >5x1015 cm-2. Samples were partly covered by a metal mask during implant. Porous silicon has been formed on the exposed samples to study the effect of hydrogen irradiation. We have found that porous silicon formation is inhibited in the irradiated areas on p-type silicon and promoted on the n-type one.

Hydrogen implantation in semiconductors Experimental setup at TOP-IMPLART injectorIon implantation is the most effective technique to alter the electronic properties of semiconductors.

Hydrogen implantation is used for SOI wafer production exploiting defect coalescence at high concentrations ( >1016 cm-3) triggered by a thermal process.

In micromachining applications, localized hydrogen implantation has been used to suppress the formation of porous silicon (sacrificial material) in irradiated areas.

• The sample holder has been placed inside the accelerator pipe after the deflecting magnet of the LEBT (Low Energy Beam Transfer) line where an access (a flexible pipe) has been realized.

• The sample holder has been realized on a blind flange holding both a degrader (with an iris) and the silicon sample in contact with a metal mask.

• Beam current and sample temperature are constantly monitored.

The beam is de-focused using the quadrupoles to obtain a uniform beam over a 1.0 cm2 area at the position of the silicon sample. The 3 MeV beam is degraded by 60 μm thick aluminum foil to 1.8 MeV. The silicon sample is masked with a 200 μm thick molybdenum mask that defines a regular pattern (parallel lines 500 μm wide separated by 500 μm gap) over the silicon wafer.

Parameter Value Units Notes

Pulse length 98 μsPulse Repetition Frequency 30 HzBeam energy 1.8 MeVRange in silicon 36.4 μm Computed with SRIMCharge/pulse 8.91x10-9 C Average valueProtons/pulse 5.57x1010 Average valueTarget fluence 5.00x1015 cm-2 This corresponds to an implantation

peak of 7.2x1019 cm-3

Exposure time 3600 s Minimum time, 4500 s was necessary

Interstitial hydrogen in silicon is an amphoteric impurity, i.e. it can assume different charge states, behaving both as a donor and as an acceptor with two levels in the bandgap. The electron occupancy determines the charge state:

1. positive charge state H+: the defects corresponds to a donor level. In p-type materials it compensates the acceptor impurities.

2. neutral charge state H0: slightly less stable than H+ has almost the same properties.3. negative charge state H-: this defects corresponds to an acceptor level. In n-type materials

is expected to compensate donors.

SRIM/TRIM simulationsSRIM/TRIM tool has been used to compute the thickness of the degrader and the range in silicon of the particles after the degrader. An aluminum foil of 60 μm has been used. TRIM Monte Carlo code has been used to compute the concentration peak (atoms/cm3) at the stopping range, given the fluence (atoms/cm2).

Porous SiliconPorous silicon is obtained from electrochemical etching of silicon in electrolytes containing HF.

Experimental ResultsParameter p-type n-typeResistivity 1-10 Ω·cm 20 Ω·cmDopant Boron PhosphorousElectrolyte HF:IPA=1:1 HF:DI:IPA=1:3:1Current density 20mA/cm2 30mA/cm2

Environment - Dark

Conclusions: We have obtained a complementary behavior on n-type and p-type wafers: porous silicon grows in irradiated areas on n-type and in masked areas on p-type. The growth process has been optimized for the p-type and requires further studies on the n-type as it is too prone to electropolishing.

FTIR analysis of p-type sample confirming hydrogen implantation (baseline is unexposed silicon.

SEM image of the p-type sample after porous silicon etching (in KOH). In p-type samples porous silicon growth have been suppressed in exposed areas up to the stopping range. Thicknesses in the image must be scaled by 1.22 due to the angled view.

SEM image of the n-type sample with porous silicon visible. In n-type samples porous silicon growth have been promoted in exposed areas. Electropolishing occurred near the silicon wafer due to a no-yet optimized growth process.