Skip to content

Ultrashort pulsed (USP) laser processing of binderless nanopolycrystalline diamond (NPD)

FieldValue
JournalMaterials & Design
Published2026-03-28
AuthorsKiran Michael, Matthias Putzer, Timo Schudeleit, Markus Bambach, Konrad Wegener
DOI10.1016/j.matdes.2026.115942

Laser processing produced 200 µm diameter, two-edge micro milling tools from binderless nanopolycrystalline diamond. The tools had a cutting-edge radius of 1.03 ± 0.13 µm and average roughness Ra = 28.7 ± 5.0 nm. The study also measured single-pulse ablation thresholds from 1.56 J/cm² for 515 nm, 200 fs irradiation to 8.82 J/cm² for 1030 nm, 15 ns irradiation. Green femtosecond pulses at 100 kHz gave the best tested balance of low roughness and material removal during orthogonal machining. A 515 nm, 200 fs quasi-tangential window reached Ra = 13 nm and Sa = 51 nm.

Binderless NPD contained tightly bonded 30 to 50 nm diamond grains with no binder or secondary phases. Its Knoop hardness was about 125 GPa and remained unchanged to 1400 °C. Flat 3 × 3 × 1 mm specimens supported orthogonal tests. Cylindrical NPD with 0.5 mm diameter and 0.5 mm length supported quasi-tangential tests and tool fabrication. The cylinders were brazed to tungsten carbide shafts and mounted on the rotational axis.

The program covered 1030, 515, and 343 nm wavelengths and 200 fs, 2 ps, and 15 ns pulse durations. Focused beam diameters were 20 µm for infrared, 10 µm for green, and 7 µm for ultraviolet. A two-axis galvo scanner and four mechanical axes positioned the beam and workpiece.

The authors measured crater diameter by confocal microscopy and fitted squared diameter against pulse energy and count. Plane-fitting multilinear regression limited error accumulation. High pulse counts improved crater visibility, so single-pulse thresholds were extrapolated.

The incubation factor describes how successive pulses reduce the threshold.

WavelengthPulse durationFth (J/cm²)Incubation factor s
515 nm200 fs1.5600.532
515 nm2 ps1.8360.483
1030 nm200 fs1.9550.845
1030 nm2 ps3.0670.656
1030 nm15 ns8.8200.643

The measurements used a 10 µm green beam and a 20 µm infrared beam. Green fs had the lowest threshold at 1.560 J/cm² and s = 0.532. IR ns had the highest at 8.820 J/cm², with s = 0.643. The authors attributed the latter to lower peak power and thermal diffusion. IR fs had the largest incubation factor at 0.845, while green ps had the smallest at 0.483.

Orthogonal removal across wavelength and frequency

Section titled “Orthogonal removal across wavelength and frequency”

Orthogonal tests machined 1 mm grooves with pulse overlap Ox = 90%. Scan speed and repeat passes changed with frequency to hold fluence fixed within each series. The main comparison used 50.96 J/cm². Ra came from 500 perpendicular line scans, filtered with a 2.5 µm Gaussian high-pass cutoff under ISO 4287.

At 50.96 J/cm², raising IR femtosecond frequency from 100 to 1000 kHz increased removed volume from 1,900 to 12,438 µm³. Specific energy fell from 21.05 to 3.22 µJ/µm³. Over the same frequency change, IR picosecond volume rose from 1,051 to 7,681 µm³, while specific energy fell from 38.06 to 5.21 µJ/µm³. Green femtosecond removal rose from 1,946 to 4,168 µm³, with specific energy dropping from 10.28 to 4.79 µJ/µm³.

IR fs gave the highest removal and lowest specific energy at 1000 kHz, but Ra increased with frequency. Green fs at 100 kHz had lower specific energy than green ps and retained low roughness, making it the preferred balanced condition. At 25.48 J/cm², UV fs removal stayed nearly constant with low roughness. Green ps removal also remained limited under this lower fluence.

IR tests stopped at 1000 kHz because the 2000 mm/s scan-speed limit was reached. Green ps high-fluence tests stopped at 500 kHz because available average power was insufficient.

The overlap study varied pulse overlap Ox and track overlap Oy from 0.5 to 0.9 at 515 nm, 200 fs, 100 kHz, and 1 W average power. The per-pulse fluence was 25.5 J/cm². The researchers machined 0.15 × 0.15 mm² pockets and adjusted layer count to keep pocket fluence near 2550 J/cm². A 23° hatch angle reduced overlap artifacts. Surface Sa was measured over 100 × 100 µm² scans, then filtered with a 0.8 µm S-filter and an 8 µm L-filter.

Both high and low overlap degraded the surface. Overlap near Ox = 0.9 and Oy = 0.9 accumulated heat and promoted graphitization. Overlap at or below 0.6 in both directions left gaps between scan lines, causing striations and uneven depths. The best measured finish occurred at Ox = 0.6 and Oy = 0.8. This pair gave continuous coverage without the heat accumulation seen at the highest overlap.

The normalization assumed identical aligned pulses and additive overlap. It neglected incubation, redeposition, plasma shielding, and boundary effects. Constant pocket depths supported the normalization.

Quasi-tangential ablation and subsurface damage

Section titled “Quasi-tangential ablation and subsurface damage”

Parameter set B gave the best quasi-tangential surface result. It used 515 nm, 200 fs, 12 W average power, 600 kHz, 2.00 × 10⁻⁵ J per pulse, and 50.96 J/cm² fluence. With Ox = 0.9 and a 600 mm/s scan speed, it produced Ra = 13 nm and Sa = 51 nm.

Set A used 11.76 J/cm² at 1,300 kHz, produced incomplete processing, and gave Ra = 47 nm and Sa = 207 nm. Set C retained set B’s energy and fluence, but reduced Ox to 0.8 and doubled scan speed to 1,200 mm/s. Its Ra rose to 38 nm and Sa to 100 nm. Sets D through G added energy without improving the surface. MHz and GHz burst modes also increased roughness.

Set B still had cracks with openings below 100 nm, which roughness measurements understated. FIB-STEM showed fs cracks extending about 2 µm. The ps condition produced rougher surfaces, a stronger near-surface affected zone, and cracks about 2.6 µm deep. Cracks mainly followed grain boundaries, with some transcrystalline propagation. The fs surface also carried LIPSS.

Microtool geometry and measured edge quality

Section titled “Microtool geometry and measured edge quality”

The four-axis fabrication process used 515 nm, 200 fs pulses. Orthogonal ablation formed the flutes, while quasi-tangential ablation shaped the cutting edges and clearance surfaces. The finished tools had 200 µm diameter, two cutting edges, 100 µm cutting-edge length, and a 30° rake angle.

Orthogonal work used 0.055 W, 100 kHz, and 100 mm/s. The layer target was 0.0005 mm, with Ox = 0.8 and Oy = 0.7. Quasi-tangential work used 12 W, 600 kHz, and 600 mm/s. Its 0.001 mm layer target used Ox = 0.9 and Oy = 0.13.

SEM images taken at 90° intervals confirmed the two-edge geometry. Across the full tool, Ra was 28.7 ± 5.0 nm. The cutting-edge radius was 1.03 ± 0.13 µm. The authors attributed the sharp edge to the 30 to 50 nm grain size.

Selection implications and unresolved questions

Section titled “Selection implications and unresolved questions”

Removal rate and surface quality lead to different process choices. IR fs at 1000 kHz gave strongest removal in high-fluence orthogonal tests. Green fs at 100 kHz was the better compromise when both mattered. Quasi-tangential tests showed that roughness alone cannot qualify an NPD edge because narrow surface cracks extended into the subsurface.

For diamond tool-material selection, the results support binderless NPD for precision microtool manufacture. Its isotropic hardness and lack of a binder phase avoid the cleavage and thermal degradation associated with SCD and conventional PCD. The study establishes laser formability, not comparative cutting life.

Prior femtosecond work cited by the authors reported machined-surface Ra of about 0.022 µm and edge radius near 0.8 µm. The finished-tool measurements cover different geometry and do not establish a direct ranking.

The authors list four unresolved issues. The experiments assumed uniform grain size and density between specimens. FIB-STEM preparation may have influenced cracks, so they recommend micro-CT or nano-CT for non-destructive subsurface analysis. Quasi-tangential ablation still produced nanocracks. The microtools were not tested in machining trials, leaving cutting performance and wear resistance unknown. The next validation step is to suppress cracks and then test the finished tools under cutting loads.

View Original Abstract

Ultrashort pulsed (USP) laser machining is a promising approach for processing binderless nanopolycrystalline diamond (NPD), which combines isotropic ultra-high hardness and high thermal stability. This study systematically investigates USP laser ablation to identify suitable conditions for advanced microtool fabrication. The influence of pulse duration ( τ ), wavelength ( λ ), fluence ( F ), pulse frequency ( f ) and pulse and track overlaps ( O x , O y ) on orthogonal and quasi-tangential ablation of NPD was evaluated. The single-pulse ablation threshold ranged from 1.56 J / cm 2 for green fs irradiation to 8.82 J / cm 2 for IR ns irradiation. In orthogonal ablation, increasing f improved ablation efficiency under high-fluence conditions, with IR fs giving the highest material removal, while green fs at low frequency provided the best compromise between low roughness and high efficiency. In the overlap study, O x = 0.6 and O y = 0.8 yielded the lowest surface roughness. For quasi-tangential machining, an optimum parameter window was identified with R a = 13 nm and S a = 51 nm . Finally, NPD micro milling tools with sharp cutting edges ( r ϵ = 1.03 ± 0.13 μ m ) and low surface roughness were successfully fabricated. • Binderless nanopolycrystalline diamond showed the lowest ablation threshold under green femtosecond laser irradiation and the highest threshold under infrared nanosecond irradiation. • In orthogonal machining, green femtosecond irradiation at low frequency provided the best balance between smooth surfaces and efficient material removal, while infrared femtosecond irradiation at high frequency produced the highest removal rates. • A narrow optimum process window was identified for quasi-tangential machining, yielding very low surface roughness and high surface quality on cylindrical binderless nanopolycrystalline diamond. • Micro milling tools were successfully fabricated by combining orthogonal ablation for flute generation with quasi-tangential ablation for tool shaping, producing sharp cutting edges and low surface roughness in binderless nanopolycrystalline diamond.

  1. 2016 - Innovative ultra-hard materials: binderless nanopolycrystalline diamond and nano-polycrystalline cubic boron nitride
  2. 2004 - Indentation hardness of nano-polycrystalline diamond prepared from graphite by direct conversion [Crossref]
  3. 2006 - Conditions and mechanism of formation of nano-polycrystalline diamonds on direct transformation from graphite and non-graphitic carbon at high pressure and temperature [Crossref]
  4. 2007 - Hardness and deformation microstructures of nano-polycrystalline diamonds synthesized from various carbons under high pressure and high temperature [Crossref]
  5. 2012 - Novel development of high-pressure synthetic diamonds “ultra-hard nano-polycrystalline diamonds”
  6. 2018 - Microfracture strength evaluation technique using nano-polycrystalline diamond spherical indenter
  7. 2012 - Application of nano-polycrystalline diamond to cutting tools
  8. 2012 - Cutting performance of nano-polycrystalline diamond [Crossref]
  9. 2011 - Thermochemical reactions between nano-polycrystalline diamond and grinding tool made of polycrystalline diamond
  10. 2009 - Pulsed laser processing of nano-polycrystalline diamond: a comparative study with single crystal diamond [Crossref]