Applied Surface Science xxx (2009) xxx–xxx

G Model

APSUSC-18947; No of Pages 5

Diffractive multi-beam surface micro-processing using 10 ps laser pulses

Zheng Kuang *, Walter Perrie, Dun Liu, Stuart Edwardson, Jian Cheng, Geoff Dearden, Ken Watkins

Laser Group, Department of Engineering, University of Liverpool, Brodie Building, Liverpool L69 3GQ, UK

A R T I C L E I N F O

Article history:

Received 16 May 2009

Accepted 20 June 2009

Available online xxx

PACS:

42.40.Jv

42.62.Cf

81.20.Wk

Keywords:

Picosecond laser

Spatial light modulator (SLM)

Computer generated holograms (CGH)

A B S T R A C T

A high repetition rate picosecond laser system is combined with a spatial light modulator (SLM) for

diffractive multiple beam processing. The effect of the zero order beam is eliminated by adding a Fresnel

zone lens (FZL) to defocus the un-diffracted beam at the processing plane. Chromatic dispersion, which is

evident with a large bandwidth femtosecond pulses leading to the problem of distorted hole shape is

eliminated due to the much narrower spectral bandwidth, �0.1 nm at 10 ps pulselength, resulting in

highly uniform intensity spots, independent of diffraction angle. In addition, high-throughput processing

is demonstrated by combining the high power laser output, 2.5 W at l � 1064 nm and fast repetition

rate, f � 20 kHz with P > 1.2 W diffracted into 25 parallel beams. This has the effect of creating an

‘‘effective’’ repetition rate of 500 kHz without restrictive scan speeds.

� 2009 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Applied Surface Science

journa l homepage: www.e lsev ier .com/ locate /apsusc

1. Introduction

Ultrashort pulse laser micro-structuring is opening upapplication from integrated optics, through multi-photoninduced refractive index engineering to precision surfacemodification for silicon scribing and solar cell fabrication [1–3].When temporal pulse t < 10 ps, ablation thresholds are low andwell defined and collateral surface damage can be stronglyreduced due to picosecond timescale electron-lattice energycoupling [4–8]. Parallel processing using diffractive multiplebeams generated by a spatial light modulator (SLM) has beendemonstrated to increase throughput and efficiency of ultra-fastlaser processing [9–15]. By synchronisation with a scanning galvo,diffractive multiple beams processing shows further flexibilityand potential industrial applications [14,15]. Nevertheless, usinga femtosecond laser (t � 180 fs), the finite laser bandwidth(Dl � 5 nm) can significantly alter the intensity distribution ofdiffracted beams at higher angles resulting in elongated holeshapes [15].

With this limitation in mind, a picosecond laser system withmuch narrower spectral bandwidth (t � 10 ps, Dl � 0.1 nm) isemployed here for diffractive multiple beam processing. Compactsolid state picosecond laser systems (t < 15 ps) with high pulse

* Corresponding author. Tel.: +44 1517945244/7783536444 (mobile).

E-mail address: z.kuang@liv.ac.uk (Z. Kuang).

Please cite this article in press as: Z. Kuang, et al., Diffractive multi-beSci. (2009), doi:10.1016/j.apsusc.2009.06.089

0169-4332/$ – see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2009.06.089

energy, average power and repetition rate, have shown advantagesover femtosecond system for high precision micro-machining[7,8]. This paper demonstrates that the distortion of intensityprofile at high diffractive angle is eliminated due to the narrowerspectral bandwidth of the laser source. The drilled holes perfectlykeep their round shape (eccentricity: e < 1.04) when applyinglarge diffractive angle (u � 1.278). Finally, high power (2.5 W)parallel processing with 25 diffracted beams and laser repetitionrate applied (f � 20 kHz), demonstrates industrial level precisionlaser micro-processing.

2. Experiment

Fig. 1 shows the schematic of the experimental setup. The1064 nm, 10 ps laser output (High-Q IC-355-800 nm, 0–50 kHz)beam traversed a half wave plate used for adjusting the linearpolarization direction, a beam expander (M � 3), and afterreflection on mirrors 1 and 2, illuminated a reflective phase onlySLM (Hamamatsu X10468-03) liquid crystal on silicon (LCoS)device with 800 � 600 pixels and dielectric coated for 1064 nmwavelength (reflectivity, h � 95%), oriented at <108 angle ofincidence. Directed by the LCoS, the modulated beam entered ascanning galvanometer with f = 100 mm flat field of f-theta lens(Nutfield) producing a near perfect focusing system. Substrateswere mounted on a precision 5-axis (x, y, z, p, q) motion controlsystem (Aerotech) allowing accurate positioning of the substratesurface at the laser focus. The spectral bandwidth, Dl � 0.1 nm at

am surface micro-processing using 10 ps laser pulses, Appl. Surf.

Fig. 1. Experimental setup.

Z. Kuang et al. / Applied Surface Science xxx (2009) xxx–xxx2

G Model

APSUSC-18947; No of Pages 5

1064 nm output, was relatively narrow and important ineliminating chromatic dispersion of the SLM.

3. Results and discussions

3.1. Defocus zero order beam

The un-diffractive zero order beam which otherwise wouldgenerate unwanted surface damage, can be removed at the Fourierplane of a 4f telescope optical system [9,12,13,15]. Alternatively, inthis work, the zero order beam was significantly defocused at theprocessing plane by adding a Fresnel zone lens (FZL) onto thecomputer generated hologram (CGH). The focal length (f1) of FZLcan be easily adjusted by the software developed by Holoeye [19]with friendly interface (Fig. 2). Only diffracted beams wereconverged (f1 > 0) or diverged (f1 < 0) by the FZL while the zeroorder beam was unaffected hence separating the focal planes of thediffracted from the zero order beam (Fig. 3). The beam matrixequation given in Fig. 3 describes the propagation of diffractedbeam from the LCoS surface (A) to its focal plane (B), where XA andXB are the distances of the beam from the axis, while tguA and tguB

are the gradients of the beam with respect to the axis at theposition A and B, respectively. Since tguB = 0 and XA = 0, the focalplane separation (Dd) can be calculated by the following equationderived from the beam matrix equation:

Dd ¼ jd2 � f 2j ¼f 1 f 2 � f 2d1

f 1 þ f 2 � d1� f 2

����

����

Fig. 2. Interface of the software developed by Holoeye [19], which can easily

superimpose and adjust the phase of FZL.

Please cite this article in press as: Z. Kuang, et al., Diffractive multi-beSci. (2009), doi:10.1016/j.apsusc.2009.06.089

where d1 � 200 mm was the distance between LCoS and f-thetalens, and f2 � 100 mm was the focal length of f-theta lens. Thus:

Dd ðmmÞ ¼ 100 f 1 � 20000

80þ f 1� 100

����

����

A slightly defocused zero order could still damage the samplebecause it contains approximately 50% of the input pulse energywhich was much stronger than any of the diffracted orders [14].Accordingly, f1 must be adjusted carefully to allow sufficientseparation (Dd) so that the fluence at the substrate is below thedamage threshold. Fig. 4(a) demonstrates a CGH calculated by‘gratings and lenses’ algorithm [16–18] to generate eight first orderidentical beams and its computational reconstruction, while (b)shows the micro-machined results on Ti6Al4V using the CGHswhich were superimposed by FZLs with different f1 to adjust Dd.Fig. 4 (b) shows that the defocused zero order beam still damagedthe sample when the separation Dd = 1 mm (middle lower

Fig. 3. The schematic showing the way to separate the focal plane of diffracted

beams from the zero order. The added FZL, lens 1, can work as either positive lens

(upper) or negative lens (lower) to obtain the separation, Dd. The beam matrix

equation below describes the propagation of diffracted beam from LCoS surface (A)

to its focal plane (B).

am surface micro-processing using 10 ps laser pulses, Appl. Surf.

Fig. 4. (a) A CGH calculated by ‘gratings and lenses’ algorithm to generate eight first order identical beams (left) and its computational reconstruction (right). (b) The CGHs

superimposed by FZLs with different f1 to adjust Dd (upper) and the machined results on Ti6Al4V (lower).

Z. Kuang et al. / Applied Surface Science xxx (2009) xxx–xxx 3

G Model

APSUSC-18947; No of Pages 5

picture), while it was totally removed when Dd = 5 mm (rightlower picture).

3.2. Spectral bandwidth effect on diffracted beam shape

Fig. 5(a) demonstrates a series holes drilled on a silicon sampleusing 10 ps (l0 � 1064 nm, Dl � 0.1 nm) pulses when varying thediffractive angle, while, in Fig. 5(b), the 2D (left) and 3D (right)micrographs with large magnification clearly shows the reason-ably round hole shape when applying large angle of diffraction(u > 18). A graph, demonstrating the eccentricity e = a/b of thesedrilled hole shape as a function of diffractive angle is plotted inFig. 6, and shows that e increases only very modestly, e < 1.04.Fig. 7 shows that there is a negligible variation of ablation depth

Please cite this article in press as: Z. Kuang, et al., Diffractive multi-beSci. (2009), doi:10.1016/j.apsusc.2009.06.089

(1000 pulses, Ep � 5 mJ) with increasing angle of diffraction,demonstrating a high degree of reproducibility of the diffractivemulti-beam processing. The above results indicate that theelongation of diffracted beam shape caused by chromaticdistortion [15] can be eliminated by employing picosecond laserpulses (t � 10 ps) with narrower bandwidth hence allowingconstant ablation rate.

3.3. High power parallel processing with high repetition rate at

f = 20 kHz

With the ps system (t � 10 ps) operating at high repetition rate(f = 20 kHz) and maximum output (Paverage � 2.5 W), parallelprocessing is demonstrated, using 25 diffractive beams pattern,

am surface micro-processing using 10 ps laser pulses, Appl. Surf.

Fig. 5. (a) A series holes drilled on a silicon sample when varying the diffractive angle. (b) The 2D (left) and 3D (right) micrographs with large magnification, showing the shape

of the hole, fabricated by single 10 ps pulse, when applying large diffractive angle, u > 18.

Fig. 6. The graph demonstrating the eccentricity e of hole shape drilled by 10 ps

when varying the angle of diffraction.

Fig. 7. The graph demonstrating the variation of ablation depth using one thousand

10 ps pulses (Ep � 5 mJ) when varying the angle of diffraction.

Fig. 8. The schematic showing the design of the 25 beams pattern and the method of

scanning. (The vertical distance between two adjacent spots was 100 mm; by

repeatedly scanning the pattern with 50 mm vertical offset each time, multiple

micro-channels with 50 mm intervals can be obtained.)

Z. Kuang et al. / Applied Surface Science xxx (2009) xxx–xxx4

G Model

APSUSC-18947; No of Pages 5

Please cite this article in press as: Z. Kuang, et al., Diffractive multi-beSci. (2009), doi:10.1016/j.apsusc.2009.06.089

created by a CGH calculated by 2D Gershberg-Saxon (GS)algorithm using a LabVIEW program [17]. The schematic givenin Fig. 8 demonstrates the design of the beams pattern and themethod of scanning, while the micrographs in Fig. 9 show themicro-machining results on a polished Ti6Al4V sample. A10 mm � 10 mm area covered by �2 mm deep micro-channelswith 50 mm intervals and overscanned 20 times at scan speed100 mm/s was completed within 16 s. The width of the channels(a = 25.3 � 0.4 mm) was perfectly uniform, indicating the accurateCGH calculation using GS algorithm. Diffractive efficiency has beenmeasured �50%, allowing >1.2 W diffracted into 25 parallel beams.This has the effect of creating an ‘‘effective’’ repetition rate of 500 kHzwithout restrictive Galvo scan speeds.

am surface micro-processing using 10 ps laser pulses, Appl. Surf.

Fig. 9. The micrograph partially demonstrating the results of micro-channels covering a 10 mm � 10 mm area machined by the 25 beams pattern on a polished Ti6Al4V

(upper), and the micrographs showing the result with higher magnification (lower left) and 3D surface profile image by Wyko NT1100 optical surface profiler (lower right).

Z. Kuang et al. / Applied Surface Science xxx (2009) xxx–xxx 5

G Model

APSUSC-18947; No of Pages 5

4. Conclusion

Diffractive multiple beams micro-processing using 10 ps laserpulses has been demonstrated in this paper. The zero order beam iseliminated by adding a Fresnel zone lens (FZL) to defocus the un-diffracted beam using simpler experimental setup. The machinedholes are almost perfectly round, independent of angle ofdiffraction hence showing the obvious advantage of parallelprocessing using ultrashort pulses with picosecond temporalpulselength with narrow spectral bandwidth. Furthermore,simultaneously scanning multiple diffractive beams (n = 25)containing >1.2 W average power with higher repetition rate(up to) f = 20 kHz demonstrates high-throughput surface micro-structuring by creating an ‘‘effective’’ repetition rate of 500 kHzwithout restrictive Galvo scan speeds.

Acknowledgements

This work was carried out under an NWSF funded project incollaboration with University of Manchester and is being developedfurther under a TSB funded project (PARALASE) involving theUniversity of Liverpool with partners Coherent Scotland and OxfordLasers. The authors gratefully acknowledge the kind help from Prof.Miles Padgett and Dr. Jonathan Leach at University of Glasgow andfrom Dr Raymond Livingstone (Hamamatsu Photonics UK Limited)on loaning us a SLM (X10468-03) for this research. The author ZhengKuang would like to thank the Scholarship of Overseas ResearchStudents Award Scheme (ORSAS) and the University of LiverpoolGraduate Association (Hong Kong) for providing financial supportfor his PhD study at the University of Liverpool.

References

[1] V.R. Bhardwaj, E. Simova, P.B. Corkum, D.M. Rayner, C. Hnatovsky, R.S. Taylor, B.Schreder, M. Kluge, J. Zimmer, Femtosecond laser-induced refractive index mod-ification in multicomponent glasses, J. Appl. Phys. 97 (2005) 083102.

Please cite this article in press as: Z. Kuang, et al., Diffractive multi-beSci. (2009), doi:10.1016/j.apsusc.2009.06.089

[2] M. Farsari, G. Filippidis, S. Zoppel, G.A. Reider, C. Fotakis, Efficient femtosecond lasermicromachining of bulk 3C-SiC, J. Micromech. Microeng. 15 (2005) 1786–1789.

[3] J. Hermann, M. Benfarah, S. Bruneau, E. Axente, G. Coustillier, T. Itina, J.-F.Guillemoles, P. Alloncle, Comparative investigation of solar cell thin film proces-sing using nanosecond and femtosecond lasers, J. Phys. D: Appl. Phys. 39 (2006)453–460.

[4] D. Du, X. Liu, G. Korn, J. Squier, G. Mourou, Laser-induced breakdown by impactionization in SiO2 with pulse widths from 7 ns to 150 fs, Appl. Phys. Lett. 64 (1994)3071–3073.

[5] X. Liu, D. Du, G. Mourou, Laser ablation and micromachining with ultrashort laserpulses, IEEE J. Quant. Electron. 33 (1997) 1706–1716.

[6] B.N. Chichkov, C. Momma, S. Nolte, F. von Alsvenleben, A. Tunnermann, Femto-second, picosecond and nanosecond laser ablation of solids, Appl. Phys. A 63(1996) 109–115.

[7] F. Dausinger, Machining of metals with ultrashort laser pulses: from fundamentalinvestigations to industrial applications, Proc. SPIE 5777 (2005) 840.

[8] J. Cheng, W. Perrie, M. Sharpe, S.P. Edwardson, N.G. Semaltianos, G. Dearden, K.G.Watkins, Single-pulse drilling study on Au, Al and Ti alloy by using a picosecondlaser, Appl. Phys. A 95 (2009) 739–746.

[9] Y. Hayasaki, T. Sugimoto, A. Takita, N. Nishida, Variable holographic femtosecond laserprocessing by use of a spatial light modulator, Appl. Phys. Lett. 87 (2005) 031101.

[10] S. Hasegawa, Y. Hayasaki, N. Nishida, Holographic femtosecond laser processingwith multiplexed phase Fresnel lenses, Opt. Lett. 31 (2006) 1705–1707.

[11] Y. Kuroiwa, N. Takeshima, Y. Narita, S. Tanaka, K. Hirao, Arbitrary micropatterningmethod in femtosecond laser microprocessing using diffractive optical elements,Opt. Express 12 (2004) 1908–1915.

[12] H. Takahashi, S. Hasegawa, Y. Hayasaki, Holographic femtosecond laser processingusing optimal-rotation-angle method with compensation of spatial frequencyresponse of liquid crystal spatial light modulator, Appl. Opt. 46 (2007) 5917–5923.

[13] H. Takahashi, S. Hasegawa, A. Takita, Yoshio Hayasaki, Sparse-exposure techniquein holographic two-photon polymerization, Opt. Express 16 (2008) 16592–16599.

[14] Z. Kuang, W. Perrie, J. Leach, M. Sharp, S.P. Edwardson, M. Padgett, G. Dearden, K.G.Watkins, High throughput diffractive multi-beam femtosecond laser processingusing a spatial light modulator, Appl. Surf. Sci. 255 (2008) 2284–2289.

[15] Z. Kuang, D. Liu, W. Perrie, S.P. Edwardson, M. Sharp, E. Fearon, G. Dearden, K.Watkins, Fast parallel diffractive multi-beam femtosecond laser surface micro-structuring, Appl. Surf. Sci. 255 (2009) 6582–6588.

[16] J. Liesener, M. Reicherter, T. Haist, H.J. Tiziani, Multi-functional optical tweezersusing computer-generated holograms, Opt. Commun. 185 (2000) 77–82.

[17] J. Leach, K. Wulff, G. Sinclair, P. Jordan, J. Courtial, L. Thomson, G. Gibson, K.Karunwi, J. Cooper, Z.J. Laczik, M. Padgett, Interactive approach to optical twee-zers control, Appl. Opt. 45 (2006) 897–903.

[18] J.E. Curtis, C.H.J. Schmitz, J.P. Spatz, Symmetry dependence of holograms foroptical trapping, Opt. Lett. 30 (2005) 2086–2088.

[19] The official webpage of Holoeye company: http://www.holoeye.com.

am surface micro-processing using 10 ps laser pulses, Appl. Surf.