Skip to content

MODIFICATION OF DIAMOND SURFACE DURING PHOTOLITHOGRAPHY, PLASMA-CHEMICAL CLEANING, AND MAGNETRON DEPOSITION

MetadataDetails
Publication Date2020-10-27
JournalIZVESTIYA VYSSHIKH UCHEBNYKH ZAVEDENIY KHIMIYA KHIMICHESKAYA TEKHNOLOGIYA
AuthorsА. В. Голованов, Nikolay V. Luparev, B. P. Sorokin
Citations5
AnalysisFull AI Review Included
  • Research Focus: Investigates how standard semiconductor fabrication steps—photolithography, low-power RF oxygen plasma cleaning, and magnetron sputtering—modify the surface morphology of single-crystal diamond (SCD).
  • Key Finding: Even a brief 5-minute oxygen plasma exposure at low power (20 W) etches measurable grooves into the diamond surface, with depths varying from 0.3 nm to 19 nm depending on poorly controlled external factors.
  • Mechanism: Grooves form with sharp, vertical sidewalls, indicating that anisotropic ion sputtering (physical bombardment) is the dominant etching mechanism—not purely chemical oxidation.
  • Process Differentiation: Photolithography alone leaves organic resist residues that degrade metal adhesion but does not alter surface topography. Magnetron Al deposition alone causes no detectable surface change.
  • Plasma Side Effect: Oxygen plasma treatment modifies the secondary electron emission coefficient of the diamond surface, likely due to oxygen termination.
  • Material Independence: Etch groove behavior is consistent across both IIa and IIb diamond types, both HPHT and CVD origins, and regardless of polishing direction (“hard” or “soft” <100> vs <110>).
  • Practical Implication: Conventional plasma cleaning recipes used for silicon are not “safe” for diamond; ultra-gentle protocols or alternative cleaning methods must be developed for diamond-based MEMS, photonics, and power electronics fabrication.
ParameterValueUnitContext
Diamond TypeIIa, IIb-Single crystal substrates tested
Crystal Orientation(001)-Working plane
Substrate Dimensions4-5mmRectangular plates
Substrate Thickness200-400µmLaser-cut from bulk crystal
Surface Roughness (hard direction polish)< 1nmNear-atomic smoothness
Surface Roughness (soft direction polish)1-100nmPolishing scratch depth
Plasma GasO2 (pure)-Reactive etchant
RF Frequency13.56MHzCapacitively coupled plasma
Base Vacuum Pressure10-4TorrPre-process chamber pressure
Gas Flow Rate20cm3/minO2 purge flow
Purge Time20minPre-ignition flushing
Process Pressure100 → 60mTorrIgnition → stable operation
RF Bias Power20WLow-power regime
DC Self-Bias≥ 150VSubstrate electrode
Process Duration5minPlasma cleaning time
Etch Depth Range0.3 - 19nmInter-experiment variation
Etch Depth Variation (single substrate)10 - 50%Intra-wafer non-uniformity
Average Etch Rate2 - 3nm/minAt 20 W bias
PhotoresistAZ1505 (positive)-Spin-coated
Spin Speed8000rpmResist application
Soft Bake Temperature110°CPre-exposure bake
Soft Bake Time2min-
Resist Thickness0.4 - 0.6µmPost-spin
Exposure Source405nmLaser (Heidelberg µPG 101)
Exposure Dose~300mJ/cm2-
DeveloperAZ726MIF (alkaline, metal-free)-40 s immersion
Anneal Temperature (pre-process)680°CAir ambient, 20 min
Al Film Thickness130nmMagnetron sputtered
Electrode Diameter (plasma chamber)150mmParallel-plate reactor
  • Substrate Preparation
    • Rectangular (001) SCD plates laser-cut from HPHT or homoepitaxial CVD bulk crystals.
    • Sequential ultrasonic cleaning: surfactant → acetone → isopropanol → DI water.
    • N2 blow-dry, then air-anneal at 680 °C for 20 min to remove non-diamond carbon.
  • Photolithography
    • Spin-coat positive resist AZ1505 at 8000 rpm; soft-bake at 110 °C / 2 min.
    • Expose at 405 nm, ~300 mJ/cm2 (Heidelberg µPG 101).
    • Develop in AZ726MIF for 40 s.
  • Oxygen Plasma Cleaning
    • Transfer to AJA ORION 8 magnetron load-lock with integrated RF CCP source.
    • Pump to 10-4 Torr, purge with O2 at 20 cm3/min for 20 min.
    • Ignite plasma at 100 mTorr / 20 W, then ramp to 60 mTorr.
    • Treat for 5 min with DC self-bias ≥ 150 V.
  • Metallization
    • Magnetron sputter 130 nm Al at room temperature (no substrate heating).
  • Characterization
    • Optical microscopy for lithographic pattern fidelity.
    • Confocal Raman (Renishaw inVia) for organic-resist contamination detection.
    • SEM for secondary electron emission contrast and groove imaging.
    • Optical profilometry (Sensofar S Neox) for etch-depth quantification (3-4 averaged measurements per site).
  • Applicable Materials

    • Optical Grade Single Crystal Diamond (SCD): Required for the (001)-oriented, sub-nm roughness substrates used here. 6CCVD supplies HPHT-type IIa and IIb SCD in custom thicknesses from 0.1 µm to 500 µm, with Ra < 1 nm polish—directly matching the “hard-direction” surface quality critical for minimizing intrinsic defect density before plasma exposure.
    • Homoepitaxial CVD SCD: For replicating the CVD-grown arm of this study (bulk + epi-layer structures), 6CCVD’s MPCVD capability delivers low-boron, high-purity epilayers on SCD seeds.
    • Heavy Boron-Doped PCD (BDD): Relevant for the IIb-type electrical-contact portions of the workflow (Schottky/ohmic metallization development).
  • Customization Potential

    • Substrate Geometry: Paper used 4-5 mm plates of 200-400 µm thickness. 6CCVD scales this to plates/wafers up to 125 mm (PCD) and 10 mm-thick SCD substrates, plus laser cutting to any custom footprint.
    • Metallization Stack: The study used blanket Al (130 nm). 6CCVD provides in-house metalization (Au, Pt, Pd, Ti, W, Cu) for Schottky/ohmic contact development—critical for the diamond Schottky diodes, photodetectors, and RF acoustic resonators the authors cite as target applications.
    • Surface Engineering: 6CCVD can deliver substrates polished in either “soft” (<100>) or “hard” (<110>) crystallographic directions, enabling controlled reproduction of both scratch regimes examined in this paper.
  • Engineering Support

    • 6CCVD’s in-house PhD engineering team can assist with material selection for diamond-based photonics, MEMS, RF acoustic wave devices, and high-temperature power electronics—the exact application categories flagged in this work.
    • Consultation available on minimizing plasma-induced damage during lithographic patterning, including recommendations on starting surface roughness, crystallographic orientation, and resist compatibility for diamond-specific process flows.
  • Call to Action

    • For custom specifications or material consultation, visit 6ccvd.com or contact our engineering team directly.
View Original Abstract

An influence of the conventional planar technology of the semiconductor surface treatment operations (photolithography, plasma-chemical surface cleaning in radio frequency low-pressure oxygen discharge, and thin metal films growth) on monocrystalline diamond surface topography and roughness is investigated. It is preliminary shown that photolithography without the plasma cleaning and magnetron deposition of an aluminum film do not induce any diamond surface changes which could be found by the optical profilometry, electron and optical microscopy. However, as a result of the lithography some organic contaminants can decrease the adhesion of the aluminum film and have to remove from the surface. Further, the influence of the above-listed surface treatment operations separately and in combinations on the synthetic diamond surface is investigated. It is found that an even short oxygen plasma cleaning during 5 min with a low bias power of 20 W leads to etching the grooves on the diamond surface The depth of the etching grooves varies by more than an order of magnitude in different experiments (0.3 - 19 nm), and by 10-20% within the same substrate under the influence of the poorly controlled external factors. Also, oxygen plasma treatment changes its secondary electron emission coefficient. Deposition of aluminum film after plasma cleaning does not induce noticeable changes in the diamond surface topography. The quality of diamond surface polishing (in “hard” of “soft” direction), the diamond’s type (IIa or IIb) as well as producing method (HPHT or homoepitaxial CVD) do not affect the result of grooves formation.