Cent. Eur. J. Phys.

• 6(2) • 2008 • 327-331

DOI: 10.2478/s11534-008-0026-0

Central European Journal of Physics

Impact of the laser wavelength and fluence on the

ablation rate of aluminium
Research Article

Mihai Stafe12∗ , Ionut Vladoiu1 , Ion M. Popescu1

1 Department of Physics, University “Politehnica” of Bucharest, Spl. Independendei 313, 060042 Bucharest, Romania
2 Present address: Department of Experimental Physics, University of Szeged, Dom ter 9, H-6720 Szeged, Hungary

Received 29 Jun 2007; accepted 3 December 2007

Abstract: The dependence of the ablation rate of aluminium on the fluence of nanosecond laser pulses with
wavelengths of 532 nm and respectively 1064 nm is investigated in atmospheric air. The fluence of the
pulses is varied by changing the diameter of the irradiated area at the target surface, and the wavelength
is varied by using the fundamental and the second harmonic of a Q-switched Nd-YAG laser system. The
results indicate an approximately logarithmic increase of the ablation rate with the fluence for ablation
rates smaller than ∼6 µm/pulse at 532 nm, and 0.3 µm/pulse at 1064 nm wavelength. The significantly
smaller ablation rate at 1064 nm is due to the small optical absorptivity, the strong oxidation of the
aluminium target, and to the strong attenuation of the pulses into the plasma plume at this wavelength.
A jump of the ablation rate is observed at the fluence threshold value, which is ∼50 J/cm2 for the second
harmonic, and ∼15 J/cm2 for the fundamental pulses. Further increasing the fluence leads to a steep
increase of the ablation rate at both wavelengths, the increase of the ablation rate being approximately
exponential in the case of visible pulses. The jump of the ablation rate at the threshold fluence value is
due to the transition from a normal vaporization regime to a phase explosion regime, and to the change
of the dimensionality of the hydrodynamics of the plasma-plume.
PACS (2008): 52.38.Mf, 42.62.-b, 81.05.Bx, 79.20.Ds
Keywords: ablation rate • nanosecond laser pulses • metal
© Versita Warsaw and Springer-Verlag Berlin Heidelberg.

1. Introduction patterning and pulsed laser-deposition in electronics, op-

toelectronics, and micromechanics, where the ablation rate
is a key parameter [1–7].
The ablation rate is strongly influenced by the thermal
The efficiency of material removal under the action of
and optical properties of the processed material (e.g., mass
short, intense laser pulses is described by the ablation
density, surface reflectivity, optical absorptivity, thermal
rate, which gives the maximum thickness of the layer ab-
conductance) [1–6, 8, 9], ambient conditions [2, 9, 10], and
lated during irradiation with a laser pulse. Understanding
by the characteristics of the processing laser beam (e.g.,
and controlling the ablation rate is essential for micro-
pulse duration, number of pulses, energy, fluence, wave-
length) [1, 2, 5, 6, 11–18]. Due to their being cost effec-
∗
E-mail: stafe@physics.pub.ro tive, there is an ongoing effort to develop new techniques

327
 Impact of the laser wavelength and fluence on the ablation rate of aluminium

based on nanosecond laser ablation, and to determine the J/cm2 for the two wavelengths.
optimum parameters of the nanosecond laser-pulses that To vary the fluence of the laser beam we varied the di-
allow one to obtain high quality micropatterns on differ- ameter of the irradiated area by moving the target away
ent metallic or non-metallic materials [8–10, 19–22]. The from the focal plane. To avoid the effect of air breakdown
experiments indicate that the ablation rate increases log- that would lead to the waste of laser energy in heating
arithmically with the fluence of the nanosecond pulses in the plasma plume ignited in front of the sample, the alu-
the case of semiconductors and dielectrics [2, 5, 16], while minium target is translated axially toward the incoming
for values of the fluence higher than a threshold there laser beam. The translation of the target is done with mi-
is a steep increase of the ablation rate [16, 19–21]. The crometric resolution by using a mechanical stage on which
wavelength of the laser has a significant influence on the the sample is fixed. The sample is translated with incre-
ablation rate: the shorter the wavelength, the higher the ments (along the axial direction) of 5 mm and 2 mm for
ablation rate. The dependence of the ablation rate on pulses at 1064 nm and 532 nm wavelengths, respectively.
the laser wavelength is mainly due to the superposition The increment steps for translating the sample were cho-
of three inter-related phenomena: the decrease of the in- sen such that the relative variation of the diameter of the
trinsic absorptivity of dielectrics and semiconductors, the irradiated area from one position to the other is approx-
increase of the surface reflectivity, and the enhancement imately the same at the two wavelengths. Thus, the di-
of the absorptivity of the laser-produced plasma with in- ameter of the irradiated area increases from its minimum
creasing the laser wavelength. value in the focal plane up to 2.5 mm and 1.6 mm for the
Here, we investigate the dependence of the ablation rate two wavelengths, while the corresponding fluence decays
of aluminium on the fluence of nanosecond laser pulses in from its maximum value in the focal plane down to 6 J/cm2
ambient atmosphere conditions, using nanosecond pulses in both cases. Because the diameter of the crater is higher
at 1064 nm and 532 nm wavelengths. The fluence at the than its depth (the aspect ratio is < 1), the ablation can
surface of the metallic target surface is varied by changing be considered as one-dimensional and, consequently, the
the diameter of the laser beam in the focal plane of a lens. diameter of the irradiated area could be approximated by
We demonstrate that the ablation rate is approximately the diameter of the crater that is drilled into the aluminium
one order of magnitude higher for the visible pulses, and target [9, 22]. The fluence was then determined by dividing
the threshold fluence that marks a jump of the ablation the energy of the pulse by the irradiated area [9].
rate is three-fold higher in the case of pulses at 532 nm The dependence of the ablation rate on the laser fluence
as compared to the 1064-nm wavelength pulses. was derived from the depth of the craters that are drilled
in multiple-pulse regime into the aluminium target at each
particular position along its axial path. Because the ab-
2. Experiments lation rate usually depends on the number of consecutive
laser pulses [2, 14], we determined the number of laser
The experimental setup consists of a Q-switched Nd-YAG pulses for drilling a crater such that the ablation rate is
laser system, a focusing lens and an aluminium target approximately constant for each successive laser pulse.
placed perpendicular to the laser beam. The experiments Our measurements indicate that 200 fundamental pulses
were carried out in ambient atmosphere conditions. The and 10 second harmonic pulses ensure an approximately
laser system works in the TEM00 mode and generates constant ablation rate and, moreover, enable us to obtain
fundamental pulses at 1064 nm wavelength which can a crater sufficiently deep to allow measurement of the ab-
be frequency doubled (532 nm wavelength) by sending lation rate with a relative error of maximum 5%. The abla-
them through a second harmonic generator module. The tion rate was then calculated by dividing the depth of the
laser pulses are characterized by a duration of 4.5 ns, crater depth by the number of laser pulses that irradiated
repetition rate of 10 Hz, and energies of 360 mJ/pulse the target for drilling the crater. The depths and the diam-
and 180mJ/pulse for the fundamental and second harmonic eters of the craters were measured using a metallographic
pulses, respectively. microscope with micrometric resolution.
The laser pulses are focused at normal incidence on a 1

3. Results and discussion

mm-thick aluminium plate by a focusing lens system. The
target is first placed in the focal plane of the convergent
lens (f/10, f=10 cm) giving a diameter of the irradiated
area of 1 mm and 0.4 mm for the fundamental and second The dependence of the ablation rate (∆h) of aluminium on
harmonic pulses, respectively. Consequently, the fluence the fluence (F ) of the fundamental and second harmonic
at the target surface was approximately 20 J/cm2 and 100 pulses is depicted in Figs. 1, 2, 3, 4.

328
Mihai Stafe, Ionut Vladoiu, Ion M. Popescu

whereas for the for the pulses at 1064 nm wavelength, the

fitting curve is

∆h = 0.1 ln[1.05 F (J/cm2 )] (µm). (2)

The maximum ablation rate is approximately one order of

magnitude higher when the target is irradiated with sec-
ond harmonic pulses instead of fundamental pulses. The
much higher efficiency of the ablation in the case of second
harmonic pulses is likely due to the superposition of three
phenomena. First, the increase of the optical absorptivity
of aluminium upon decreasing the wavelength [2, 5] leads
to strong localized heating and, consequently, to a high
ablation rate. Second, since the surface of the target sur-
face appears to be covered with a very compact slag-layer
Figure 1. Ablation rate of aluminium in open air as a function of after irradiation with infrared pulses, the oxidation of the
laser fluence of a 4.5 ns laser pulse. The laser wave-
length is 532 nm. surface during the interval between the laser pulses could
lead to a further decay of the intrinsic absorptivity of the
target for laser pulses at 1064 nm wavelength as compared
For the case of the second harmonic pulses, the increase to those at 532 nm [2, 9]. Third, the dependence of the ab-
of the fluence to a value of ∼50 J/cm2 leads to an ap- sorption coefficient of plasma (αIB ) on the wavelength (λ)
proximately logarithmic increase of the ablation rate to a
maximum of ∼6 µm/pulse (Fig. 1). In a similar manner, for αIB ≈ λ3 (3)
the fundamental laser pulses, the increase of the fluence to
∼15 J/cm2 leads to an approximately logarithmic increase implies that the absorption of the laser beam into the ig-
of the ablation rate to a maximum of ∼0.3 µm/pulse (Fig. nited plasma plume via the inverse-Bremstrahlung effect
3). In the case of the pulses at 532 nm wavelength, the is significantly stronger in the case of longer wavelengths
fitting curve is described by the equation [2, 5, 11, 23]. As a consequence, the plasma produced by
the leading edge of an infrared pulse becomes much more
absorbing and efficiently reflective for the remaining pulse
as compared to a visible pulse. This prevents the infrared
pulse from reaching the target surface and producing fur-
ther ablation shortly after the onset of the pulse.
A jump of the ablation rate is observed for a threshold
value of the fluence Fth , which is ∼50 J/cm2 for the sec-
ond harmonic, and ∼15 J/cm2 for the fundamental pulses.
Further increase of the fluence above the threshold value
leads to a steep increase of the ablation rate at both wave-
lengths (Figs. 2, 4). In the case of visible pulses, the steep
increase can be approximated by an exponential function
(Fig. 2):

∆h = 4.1 exp[0.007 F (J/cm2 )] (µm). (4)

Figure 2. Ablation rate of aluminium in open air as a function of

laser fluence of a 4.5 ns laser pulse. The laser wave- The results above on the existence of a threshold fluence
length is 532 nm. The dashed lines indicate the fitting and the steep increase of the ablation rate upon increas-
curves for the ablation rate vs fluence for values of the
fluence smaller than the threshold. ing the fluence above the threshold value are consistent
with previous work on metallic and non-metallic materials
in different irradiation conditions [16, 19–21]. There, abla-
tion was performed in vacuum conditions, and the fluence
∆h = 1.1 ln[5.85 F (J/cm2 )] (µm) (1) was varied via the pulse energy. Values of the threshold

329
 Impact of the laser wavelength and fluence on the ablation rate of aluminium

Figure 3. Ablation rate of aluminium in open air as a function of Figure 4. Ablation rate of aluminium in open air as a function of
laser fluence of a 4.5 ns laser pulse. The laser wave- laser fluence of a 4.5 ns laser pulse. The laser wave-
length is 1064 nm. length is 1064 nm. The dashed lines indicate the fitting
curves for the ablation rate vs fluence for values of the
fluence smaller than the threshold. The dotted line indi-
cates the steep increase of the ablation rate with fluence
for values of the fluence above the threshold.
fluence were reported in the range of 5 − 25 J/cm2 , and it
was suggested that the existence of a threshold fluence is
mainly due to phase explosion [16, 19–21]. For the partic-
the plasma hydrodynamic length [2, 11],
ular case assessed here, the jump of the ablation rate is
likely due to the following two causes. First, because the
increase of the ablation rate is accompanied by a signifi- lh = vp τ (5)
cant amount of bright sparks that emerge into the plume,
the jump of the ablation rate is likely related to the tran- is ∼50 µm, which is similar to the diameter of the laser
sition from a normal vaporization regime of the irradiated beam corresponding to the threshold fluences Fth of ei-
sample (which is characterized by a small mass-removal ther visible or infrared pulses. In equation (5), vp =
p
rate) to a phase-explosion regime (which is characterized γ kB Tp /M is the expansion velocity of the plasma plume
by ejection of small particulates and melt droplets and an considered as an ideal gas with adiabatic coefficient
enhanced mass-removal rate). The ejection of droplets is γ = 5/3, temperature Tp and the atomic mass M, τ is
enhanced by the recoil pressure exerted by the plasma the pulse duration, and kB is the Boltzmann constant.
plume onto the melted surface [24], such that the infrared
pulses (which heat the plasma stronger) could start the
ejection process at a lower threshold fluence (∼15 J/cm2 ) 4. Conclusion
as compared to the visible pulses (∼50 J/cm2 ). The sec-
ond cause contributing to the jump in the ablation rate The impact of the fluence and wavelength of the nanosec-
consists in the change of the dimensionality of the hy- ond laser pulses on the ablation rate of aluminium was
drodynamics of the plasma plume from one-dimensional investigated in ambient atmosphere conditions. We var-
to three-dimensional. In the one-dimensional regime the ied the fluence of the laser pulses by changing the di-
expansion of the plasma-plume is mainly axial, being char- ameter of the irradiated area at the target surface, and
acterized by high density and strong absorptivity. In con- the wavelength by using the fundamental and the second
strast, in the three-dimensional regime the plasma plume harmonic beams of a Q-switched Nd-YAG laser system.
expands in both axial and radial directions, and is char- The results indicate that the ablation rate has an approx-
acterized by low density and weak absorptivity [5]. The imately logarithmic increase with the fluence for ablation
jump appears when the diameter of the laser beam (central rates smaller than ∼6 µm/pulse and ∼0.3 µm/pulse when
Airy disk) becomes the same order of magnitude with the using pulses at 532 nm and 1064 nm wavelengths, respec-
hydrodynamic length of the laser-produced plasma. This tively. The much higher efficiency of the ablation in the
relationship between the beam diameter and the hydrody- case of second harmonic pulses as compared to the funda-
namic length of the plasma at threshold is demonstrated mental pulses is due to the superposition of the following
as follows. For an usual plasma temperature of 105 K [11], three effects: the increase of the optical absorptivity of

330
Mihai Stafe, Ionut Vladoiu, Ion M. Popescu

aluminium with decreasing wavelength, the weaker oxida- [5] M. von Allmen, A. Blatter, Laser-Beam Interactions
tion of the surface of the target when irradiated at short with Materials (Springer-Verlag, Berlin, 1995)
wavelengths, and the decay of the plasma absorption via [6] I. Ursu, I. N. Mihailescu, A. M. Prokhorov, V. I.
the inverse-Bremstrahlung effect at short wavelengths. Konov, Interactiunea radiatiei laser cu metalele (Edi-
The increase of the fluence to a threshold value of ∼50 tura Academiei R.S.R., Bucuresti, 1986, in Romanian)
J/cm2 for the second harmonic pulses, and ∼15 J/cm2 for [7] P. Simon, J. Ihlemann, Appl. Phys. A 63, 505 (1996)
the fundamental pulses, leads to a jump of the ablation [8] B.N. Chichkov et al., Appl. Phys. A 63, 109 (1996)
rate. Further increase of the fluence above the thresh- [9] A.E. Wynne, B.C. Stuart, Appl. Phys. A 76, 373 (2003)
old value leads to a steep increase of the ablation rate at [10] S.M. Klimentov et al., Physics of Wave Phenomena
both wavelengths. In the case of visible pulses, the steep 15, 1 (2007)
increase of the ablation rate with the fluence above the [11] S. Amoruso, R. Bruzzese, N. Spinelli, R. Velotta, J.
threshold fluence can be described by an exponential func- Phys. B 32, 131 (1999)
tion. The jump of the ablation rate at the threshold fluence [12] B.W. Rottke, J. Ihlemann, H. Schmidt, A. Scholl, Appl.
is likely due to the superposition of two effects. First, the Phys. A 60, 13 (1995)
appearance of a significant number of bright sparks (par- [13] A. Bogaerts, Z. Chen, Spectrochimica Acta B 60, 1280
ticulates and melted material) within the plasma plume in- (2005)
dicates a transition from the normal vaporization to phase [14] M. Stafe, C. Negutu, I.M. Popescu, Shock Waves 14,
explosion with increasing the fluence of the pulses above 123 (2005)
the threshold value. Second, at the threshold fluence the [15] M. Stafe, C. Negutu, I.M. Popescu, Appl. Surf. Sci.
hydrodynamic length of the plasma becomes similar to the 253, 6353 (2007)
diameter of the laser beam (central Airy disk), and the di- [16] N.M. Bulgakova, A.V. Bulgakov, Appl. Phys. A 73, 199
mensionality of the hydrodynamics of the plasma plume (2001)
changes from one-dimensional to three-dimensional. [17] B. Garrison, T. Itina, L. Zhigilei, Phys. Rev. E 68,
041501 (2003)
[18] E.G. Gamaly, A.V. Rode, A. Perrone, A. Zocco, Appl.
References Phys. A 73, 143 (2001)
[19] C. Porneala, D.A. Willisa, Appl. Phys. Lett. 89, 211121
[1] J.F. Ready, Effects of high-power laser radiation (Aca- (2006)
demic Press, New York – London, 1971) [20] J.M. Fishburn, M.J. Withford, D.W. Coutts, J.A. Piper,
[2] D. Bauerle, Laser processing and chemistry Appl. Optics 43, 6473 (2004)
(Springer-Verlag, Berlin – Heidelberg – New [21] M.R.H. Knowles, G. Rutterford, D. Karnakis, A. Fer-
York, 2000) guson, The International Journal of Advanced Manu-
[3] I.M. Popescu et. al., Aplicatii ale laserilor, (Ed. facturing Technology 33, 95 (2007)
Tehnica, Bucuresti, 1979, in Romanian) [22] A. Semerok et al., Appl. Surf. Sci. 138–139, 311 (1999)
[4] D. Bauerle, Proceedings of an International Confer- [23] S. Amoruso, M. Armenante, V. Berardi, R. Bruzzese,
ence Laser Processing and Diagnostics, University of N. Spinelli, Appl. Phys. A 65, 265 (1997)
Linz (Springer-Verlag, Berlin Heidelberg, 1984) [24] Q. Lu, Phys. Rev. E 67, 016410 (2003)

331