84 | NATURE | VOL 537 | 1 SEPTEMBER 2016
LETTER
doi:10.1038/nature18619
Ablation-cooled material removal with ultrafast
bursts of pulses
Can Kerse
1
, Hamit Kalaycıoğlu
2
, Parviz Elahi
2
, Barbaros Çetin
3
, Denizhan K. Kesim
1
, Önder Akçaalan
1
, Seydi Yavaş
4
,
Mehmet D. Aşık
5
, Bülent Öktem
6
, Heinar Hoogland
7,8
, Ronald Holzwarth
7
& Fatih Ömer Ilday
1,2
The use of femtosecond laser pulses allows precise and thermal-
damage-free removal of material (ablation) with wide-ranging
scientific
1–5
, medical
6–11
and industrial applications
12
. However,
its potential is limited by the low speeds at which material can
be removed
1,9–11,13
and the complexity of the associated laser
technology. The complexity of the laser design arises from the
need to overcome the high pulse energy threshold for efficient
ablation. However, the use of more powerful lasers to increase
the ablation rate results in unwanted effects such as shielding,
saturation and collateral damage from heat accumulation at higher
laser powers
6,13,14
. Here we circumvent this limitation by exploiting
ablation cooling, in analogy to a technique routinely used in
aerospace engineering
15,16
. We apply ultrafast successions (bursts)
of laser pulses to ablate the target material before the residual heat
deposited by previous pulses diffuses away from the processing
region. Proof-of-principle experiments on various substrates
demonstrate that extremely high repetition rates, which make
ablation cooling possible, reduce the laser pulse energies needed
for ablation and increase the efficiency of the removal process by
an order of magnitude over previously used laser parameters
17,18
.
We also demonstrate the removal of brain tissue at two cubic
millimetres per minute and dentine at three cubic millimetres per
minute without any thermal damage to the bulk
9,11
.
Ablation is the evaporative removal of a material when its temperature
exceeds a critical value. Because the ablated material is physically carried
away, the thermal energy contained in the ablated mass is also removed,
thus reducing the average temperature of the remaining material.
This effect forms the basis of ablation cooling, which has been
routinely used as an approach to thermal protection during the atmos-
pheric re-entry of rockets since the 1950s, owing to the minimal mass
requirements
15
. Unlike ablation cooling for rockets, laser ablation is not
continuous, but takes place only during and shortly after an incident
laser pulse. For the laser parameters used in previous experiments abla-
tion cooling has been negligible as a cooling mechanism in comparison
with heat conduction (diffusion) from the processing region into the
bulk of the target, which is continuously occurring. For ablation cooling
to become a major contributor, the time delay between the laser pulses
(the inverse of the repetition rate) must be reduced until the part of the
material that is to be ablated does not cool substantially between suc-
cessive pulses. Only then would heat extraction due to ablation become
comparable to that due to diffusion (Fig. 1a).
The physics of the ablation-cooled regime can be explained
through a toy model (see Supplementary Information section 1 for
full details). We assume that each pulse gives rise to an instantaneous
temperature rise of ΔT, which is roughly proportional to the pulse
energy, E
p
, and that the material cools with a
τ/+/
t11
0
dependence
on the time delay, t, after the arrival of a pulse. The thermal relaxation
time, τ
0
, is proportional to δ
2
/α, where δ is the depth or the lateral
radius (whichever dimension is smaller) of the section of the material
to be ablated and α is its thermal diffusivity. For a train of N pulses,
the temperature of the target surface that is encountered by the
(n + 1)th pulse is given by T
n+1
= T
n
+ δT, where
ττδ=
∆/
+/
TT1
R0
is the small net increase in target temperature by a single pulse
and τ
R
is inverse of the repetition rate. Ablation occurs when the
temperature exceeds a critical value T
c
. For the traditional regime of
ultrafast ablation, the repetition rate is low
ττ()
R0
and each
pulse must be energetic enough to cause ablation (ΔT>T
c
− T
0
,
where T
0
is the initial surface temperature). The ablation-cooled
regime corresponds to τ
R
≲ τ
0
. In this regime, the energy of the
individual pulses can be lower than the ablation threshold because
temperature builds up from pulse to pulse and ablation starts after
the mth pulse in the train, where m = (T
c
− T
0
− ΔT + δT)/δT.
The volume of the ablated material is given by V
ablated
= β[N−
u(T
c
− T
0
− ΔT)m]E
p
u(N − m), where β is a proportionality factor and
u is the Heaviside (unit step) function. The thermal energy that diffuses
into the bulk of the target owing to cooling between the pulses is
αα
=( −)
−
(−)+(∆ −δ )
ττ+/
ET
TNmE TTmE1
heat c0
1
1
pp
R0
. For
the traditional regime, this result reduces to
α=( −)
τ →∞
ETTNElim
heat c0
p
R
.
The toy model makes two main predictions for the ablation-cooled
regime—both are confirmed by numerical solutions of the heat dif-
fusion equation (see Supplementary Information section 2 for details)
as well as the experiments described below. The first is that increasing
the repetition rate reduces the heating of surrounding regions
(Fig. 1b, c and Supplementary Fig. 1). Because less of the deposited
laser energy is lost to heat diffusion (
=
τ
→
Elim0
0
heat
R
), the ablation effi
-
ciency is higher than for the traditional regime (Supplementary Fig. 3).
The second states that the pulse energy can be decreased if the num-
ber of pulses is simultaneously increased in proportion, without a
subsequent reduction in the ablation efficiency (Fig. 1d). This is nec-
essary to fully benefit from the ablation-cooled regime, because
shielding effects (that is, ablation-induced plasma and ejected par-
ticulates reflecting and scattering incoming light) will prevent effi-
cient ablation if the repetition rate is increased at a constant energy
18
.
To demonstrate ablation cooling, a customized femtosecond fibre
laser
19–21
was used (see Supplementary Information section 3 for
details). We implemented burst-mode operation
22
, because continuous
trains of energetic pulses at the high repetition rates required to access
the ablation-cooled regime correspond to a prohibitively high average
power and laser repositioning in continuous mode is limited. In burst
mode, the laser produces groups of high-repetition-rate pulses, which
are, in turn, repeated with a lower frequency. The duty cycle of the
1
Department of Electrical and Electronics Engineering, Bilkent University, Ankara 06800, Turkey.
2
Department of Physics, Bilkent University, Ankara 06800, Turkey.
3
Department of Mechanical
Engineering, Bilkent University, Ankara 06800, Turkey.
4
FiberLAST, Inc., Ankara 06531, Turkey.
5
Nanotechnology and Nanomedicine Department, Hacettepe University, Ankara 06800, Turkey.
6
ASELSAN, Ankara 06150, Turkey.
7
Menlo Systems GmbH, Am Kloperspitz 19a, Martinsried 82152, Germany.
8
Lehrstuhl für Laserphysik, Department Physik, Friedrich-Alexander-Universität
Erlangen-Nürnberg (FAU), Erlangen 91058, Germany.
© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Letter
reSeArCH
1 SEPTEMBER 2016 | VOL 537 | NATURE | 85
Figure 1 | Principles of ablation-cooled removal of a material by laser.
a, Schematic representation of the ablation process for low (traditional
regime, left diagrams) and high (ablation-cooled regime, right diagrams)
repetition rates. Temperature profiles are illustrated for t = τ
1
(i), which
is shortly after the arrival of the first pulse for both cases; for t = τ
2
(ii),
which is before (shortly after) the arrival of the second (last) pulse for the
low-repetition-rate (high-repetition-rate) laser; and for t = τ
3
(iii), which
is shortly after the arrival of the last pulse for the low-repetition-rate laser.
The colouration of the target material is based on simulation results shown
in b at the indicated time intervals of τ
1
, τ
2
and τ
3
. b, Calculated evolution
of the temperatures at the surface (solid lines) and below (at a depth of
30 times the optical penetration depth) the surface (dotted lines) for
repetition rates of 10 MHz (blue lines) and 1,600 MHz (black lines).
The pulse energies and number of pulses are the same for both cases.
The higher repetition rate results in substantially lower temperatures below
the surface due to ablation cooling. c, Expanded view of the shaded section
of the plot in b. d, Calculated evolution of the surface temperature (dashed
lines) and amount of ablated material (solid lines) for repetition rates of
100 MHz (green lines), 400 MHz (blue lines) and 1,600 MHz (red lines).
The ablation rate remains approximately the same when the product of the
pulse energy and repetition rate is maintained. The spikes in the surface
temperatures precisely indicate the arrival of pulses, which are not shown
explicitly for clarity. e, Experimental set-up for direct confirmation of
the ablation-cooling effect. f, The measured temperature increase that is
induced on thermoelectric module 1 (the target material; solid lines) and
thermoelectric module 2 (attached to the coverslip that collects a portion
of the ablated particles; dashed lines, values have been multiplied by three
to aid comparison with ΔT
target
) with the laser operating in the ablation-
cooled regime (blue lines) and in the traditional regime (red lines).
V
Thermoelectric module 1 (target)
Thermoelectric module 2 (cover)
Cover glass
Incoming
laser beam
Ablated
particles
Unaffected surface remains
at room temperature
Surface in thermal
contact with cover glass
Surface heating
due to incident
laser beam
V
Unaffected surface remains
at room temperature
02040608
01
00
Temperature
0
1
2
Ablation depth (a.u.)
10 pulses with E
p
= E
0
at 100 MHz
10 pulses with E
p
= E
0
at 10 MHz
10 pulses with E
p
= E
0
at 1.6 GHz
40 pulses with E
p
= E
0
/4 at 400 MHz
160 pulses with E
p
= E
0
/16 at 1,600 MHz
01020304
05
0
T
0
T
c
T
0
T
c
T
0
T
c
Time (W
0
)
Time (W
0
)
Temperature
0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000
W
2
W
2
W
1
W
3
Time (W
0
)
Temperature
10 pulses with E
p
= E
0
at 10 MHz
10 pulses with E
p
= E
0
at 1.6 GHz
b
c
a
d
Low-repetition-rate
laser
Low-repetition-rate
laser
Low-repetition-rate
laser
High-repetition-rate
laser
High-repetition-rate
laser
High-repetition-rate
laser
iii
iii
f
024681
01
2
0
20
40
60
80
100
Time (s)
Temperature (a.u.)
ΔT
target
, ablation cooled
ΔT
target
, no ablation cooling
ΔT
cover
× 3, ablation cooled
ΔT
cover
× 3, no ablation cooling
Laser on
e
t = W
1
t = W
2
t = W
3
T
0
T
c
© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Letter
reSeArCH
86 | NATURE | VOL 537 | 1 SEPTEMBER 2016
pulsation can be adjusted to set the average power. Burst-mode material
processing has substantial benefits
22–24
, but the possibility of ablation
cooling has not yet been recognized.
First we present experimental evidence of the ablation-cooling
effect by simultaneously measuring the temperature of a target mate-
rial directly and the heat carried by the ablated particles (indirectly)
(Fig. 1e). The laser beam is focused onto and ablates the surface of a
thermoelectric module. This causes a temperature difference between
the laser-targeted top surface and the bottom surface, which generates
a voltage difference by the Seebeck effect. A portion of the particles
ejected from the surface during ablation stick to a glass coverslip, which
is held approximately 1 mm above the target. A second thermoelectric
module is used to monitor the temperature of the coverslip, which rises
in proportion to the thermal energy delivered by the ablated particles.
The measured temperatures of the target and the coverslip (Fig. 1f)
confirm that the target heats less, and the coverslip more, in the abla-
tion-cooled regime. The laser parameters were 50 pulses of 3 μJ each,
with an 800fs duration for a 0.2 MHz burst and a 1.7 GHz intraburst
repetition rate. This is within the ablation-cooled regime assuming a
typical thermal diffusivity of about 150 mm
2
s
−1
for the ceramic sur-
face of the thermoelectric module and 3 μJ, 800fs pulses at a 10 MHz
uniform repetition rate to illustrate the traditional regime (10 MHz
was chosen to be safely outside the ablation-cooled regime, although
the thermal diffusivity of the ceramic surface is not precisely known).
We demonstrate validity of the predictions of the toy model for abla
-
tion cooling across a range of materials (see Supplementary Information
for a discussion of other materials). Copper and silicon were chosen as
examples of metal and semiconductor targets, respectively, because their
ablation rates with ultrafast pulses are well documented. The volume of
material ablated as a function of the incident energy is shown in Fig. 2a
for Cu and Fig. 2b for Si for various repetition rates. Figure 2c, d shows
the number of atoms ablated per incident photon as a function of the
pulse energy. We observe a substantial increase in ablation when the
repetition rate is about 100MHz or higher. Although it is not possible to
predict the precise frequency required for each material (the toy model is
too simple for us to expect quantitatively accurate predictions), τ
0
≈ 1 ns
for Cu for a processing region depth of a few hundred nanometres. Given
that increases in efficiency are predicted to begin at a tenth of the corre-
sponding repetition rate, this prediction agrees with the experimental
observations. The lower thermal diffusivity of Si compared with Cu is
consistent with the increase in its ablation efficiency at 27 MHz, whereas
the results at 1 MHz and 27 MHz are similar for Cu, implying that the
onset of the ablation-cooled regime for Cu begins between 27 MHz and
108 MHz. If the repetition rate is further increased, efficiency saturates
at high pulse energies—a consequence of the expected shielding effect.
The solution is to decrease the pulse energy, and increase the number of
pulses and the repetition rate (for example, from 25 pulses at 108 MHz
to 800 pulses (with 32 times lower energy) at 3,464 MHz). The amount
of ablation remains similar (black and pink data in Fig. 2a, b), which
means that the shielding effects have been overcome.
To place the ablation results into context, they should be compared
with common literature values (see Supplementary Information section
5 for an extensive discussion). Comparison with experiments on Cu
using 70fs pulses with a pulse energy of up to 0.4mJ at 800nm (ref.17)
Figure 2 | Scaling down of the pulse energy with increasing repetition
rate. a,b, Volumes (symbols) of Cu (a) and Si (b) ablated by a single burst
of pulses as a function of total incident energy and fluence for different
intraburst repetition rates. The predictions of the toy model for the lowest
and highest repetition rates in the ablation-cooled regime are also shown
(solid lines). c,d, Ablation efficiency in terms of number of atoms of Cu
(c) and Si (d) ablated per incident photon as a function of pulse energy and
pulse fluence for different repetition rates. The legend applies to all panels.
The lower and upper limits to the data correspond to the ablation threshold
and available laser energy, respectively. In all panels the sample size for each
data point is 20, where the centre values represent the mean and the error
bars represent the standard deviation. Coloured symbols highlight the onset
of the ablation-cooled regime and (beyond 108 MHz) the inverse scaling of
the pulse energy with repetition rate in the ablation-cooled regime.
© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Letter
reSeArCH
1 SEPTEMBER 2016 | VOL 537 | NATURE | 87
reveals that we obtain around 2,000 times more ablation at the same
fluence of approximately 0.04 J cm
−2
, which is our maximum fluence
for the intraburst repetition rate of 3,456 MHz. Even if we imagine
the entire burst of 800 pulses to act like a single pulse and compare
the results with those in ref.17 for an equal total fluence (20 J cm
−2
,
the highest value for which direct comparison is possible), we obtain
about 12 times more ablation, although our pulse fluence (energy) is
smaller by a factor of 800 (2,400). Comparison with another reference
18
indicates that the efficiency of ablation in our experiments is 100 times
higher despite using a pulse energy that is 260 times lower, when
matching the fluence of the entire burst to that of the single-pulse
fluence. We achieve a level of ablation that is five times higher than
results obtained with a burst-mode laser
23
that does not exploit ablation
cooling, despite using a pulse fluence that is 165 times smaller for the
Figure 3 | Ablation of hard and soft tissues. a,b, Laser removal of a section
of human dentine obtained in the traditional regime (a, 1 kHz uniform
repetition rate) and in the ablation-cooled regime (b, 1.7 GHz intraburst
repetition rate). Although both ablation cooling and traditional ultrafast
processing avoid thermal damage at sufficiently low average powers, the
ablation-cooled regime achieves approximately six times more ablation
despite using pulse energies that are about 12 times lower. c,d, When the
(uniform or intraburst, respectively) repetition rate, average power and
scanning speed are simultaneously increased by a factor of 25, the traditional
regime of ultrafast processing results in thermal damage (c; Supplementary
Video 4), whereas the ablation-cooled regime completely avoids thermal
effects and achieves an ablation speed of 3 mm
3
min
−1
, despite using a pulse
energy that is 25 times lower (d; Supplementary Video 5). The insets in
a–d show laser scanning microscope characterizations of the ablated holes.
e,f, Histological images corresponding to about 1 mm
3
sections, which were
removed from a rat brain with the laser operating at an average power of
600 mW in the traditional regime (e), showing presence of thermal damage,
and in the ablation-cooled regime (f), showing no major thermal damage.
g, Ablation-cooled laser removal of brain tissue at an average power of 2.7 W,
achieving an ablation speed of 2 mm
3
min
−1
and showing no major thermal
damage. h, Bright-field optical image of a bovine cornea from which a flap
was removed following ablation-cooled laser processing of a section 0.4 mm
below the surface. Inset, optical coherence tomography image of the section
indicated by the rectangle.
50 μm
a
c
1 mm
b
f
0.5 mm
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
11
1
1
1
1
1
mm
mm
mm
mm
mm
m
mm
mm
mm
mm
mm
m
mm
mm
mm
mm
mm
m
mm
m
mm
m
m
mm
mm
m
m
mm
mm
m
m
mm
m
m
m
m
m
m
m
m
mm
m
m
m
m
m
mm
m
m
m
m
m
m
mm
m
m
m
m
m
m
m
m
m
m
m
m
b
0.5 mm
e
d
1 mm
1 mm
700 μm
600
500
400
300
200
100
0
20 μm
50 μm
hg
2 mm
1 mm
0.5 mm
0
500
1,000
0
500
1,000
0
150
300
450
500
750
1,400
1,000
500
0
1,200
900
600
300
0
1,000
500
0
0
500
1,000
1,400
0
500
1,000
0
200
400
(μm)
(μm)
(μ
m)
(μm)
(
μ
m)
(
μ
m)
(μm)
(μm)
(
μ
m)
© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Letter
reSeArCH
88 | NATURE | VOL 537 | 1 SEPTEMBER 2016
same burst fluence of 20 J cm
−2
. These results conclusively demonstrate
that the exploitation of ablation cooling increases the ablation efficiency
by an order of magnitude while allowing the required pulse energy to
be reduced by three orders of magnitude.
We now focus on the reduction of undesired thermal effects in the
ablation-cooled regime. We have performed systematic comparisons
using high and low repetition rates of the same laser with identical
focusing and scanning systems. Tissue removal may well be regarded
as the ultimate test of the suppression of thermal effects because an
increase in temperature of only a few degrees can lead to degradation.
Hard-tissue experiments were conducted on human dentine to contrast
the ablation-cooled regime with the traditional regime. At low average
powers, the traditional regime (using 100 μJ pulses at 1 kHz, Fig. 3a)
and the ablation-cooled regime (25 pulses of 4 μJ energy at a 1.7 GHz
intraburst repetition rate and 1 kHz burst repetition rate, Fig. 3b) both
provide results with negligible thermal damage (although the latter
achieves an ablation rate four times higher). When increasing the pro-
cessing speed by a factor of 25 with a corresponding increase in power,
the traditional regime causes excessive carbonization (Fig. 3c), whereas
the ablation-cooled regime does not, while achieving an ablation rate of
3 mm
3
min
−1
(Fig. 3d). Every other laser, focusing and scanning param-
eter was identical in these two experiments, showing that the thermal
effects are greatly reduced as a result of ablation cooling.
There are numerous applications for soft-tissue ablation
6,10,14
, par-
ticularly in targeting the brain
25
, where the extreme precision afforded
by a laser is of paramount importance. For this reason, we compared
the effectiveness of ablation cooling in selective tissue removal from
freshly harvested whole rat brains. When the average power is low, heat
diffusion from the processing region to the surrounding tissue is low
enough that the traditional regime avoids thermal side effects, yielding
damage-free ablation
11
. For higher powers, the ablation-cooled regime
demonstrates a clear advantage in the reduction of thermal effects:
although low-repetition-rate ablation causes a broad heat-affected zone
with damaged neighbouring cells, devascularization and prominent
tissue loss (Fig. 3e), there is no major heat damage in the ablation-cooled
regime at the same power (600 mW) and pulse energy (3 μJ) (Fig. 3f).
The corresponding ablation rate of 0.75 mm
3
min
−1
is eight times higher
than when using 165 μJ, 180 s pulses, with which a 0.55 mm
3
section of
brain tissue was removed in 360 s (ref.11). With ablation cooling, at a
much higher power of 2.7 W (432 MHz intraburst repetition rate, 27 kHz
burst repetition rate, 16 μJ per pulse), virtually thermal-damage-free
results are obtained (Fig. 3g) at an ablation rate of 2 mm
3
min
−1
.
Finally, we performed a flap-cutting procedure on a bovine cornea,
as this is a realistic indicator for surgical applications
8
. An area several
millimetres wide located about 0.4 mm below the surface of the cornea
was scanned with the laser and the top layer was then lifted off with
a pair of tweezers (Fig. 3h); 24 pulses with 0.8 μJ of energy per burst
were used, which is a reduction by a factor of approximately 15 in
pulse fluence compared with previous results
8
. This result and similar
experiments on poly(methyl methacrylate) (PMMA) and hydrogels
demonstrate that ablation cooling retains several of its benefits even
when used for subsurface processing (see Supplementary Information
section 15 for a detailed discussion).
We conclude by pointing out three speculative future directions of
study: exploration of the far-from-equilibrium thermodynamics of the
ablation-cooled regime, whether a suitably sculptured coherent pulse
train can coherently enhance nonlinear processes
26
and whether similar
benefits are possible in proton therapy, because the laser-based
generation of bursts of protons seems to be feasible
27
.
Online Content Methods, along with any additional Extended Data display items and
Source Data, are available in the online version of the paper; references unique to
these sections appear only in the online paper.
Received 27 July 2015; accepted 24 May 2016.
Published online 13 July; corrected online 31 August 2016
(see full-text HTML versions for details).
1. Gattass, R. R. & Mazur, E. Femtosecond laser micromachining in transparent
materials. Nat. Photon. 2, 219–225 (2008).
2. Yang, W., Kazansky, P. G. & Svirko, Y. P. Non-reciprocal ultrafast laser writing.
Nat. Photon. 2, 99–104 (2008).
3. Steinmeyer, J. D. et al. Construction of a femtosecond laser microsurgery
system. Nat. Protocols 5, 395–407 (2010).
4. Plech, A., Kotaidis, V., Lorenc, M. & Boneberg, J. Femtosecond laser near-eld
ablation from gold nanoparticles. Nat. Phys. 2, 44–47 (2006).
5. Rousse, A. et al. Non-thermal melting in semiconductors measured at
femtosecond resolution. Nature 410, 65–68 (2001).
6. Chung, S. H. & Mazur, E. Surgical applications of femtosecond lasers.
J. Biophoton. 2, 557–572 (2009).
7. Yanik, M. F. et al. Neurosurgery: functional regeneration after laser axotomy.
Nature 432, 822 (2004).
8. Juhasz, T. et al. Corneal refractive surgery with femtosecond lasers. IEEE J. Sel.
Top. Quant. Electron. 5, 902–910 (1999).
9. Serbin, J., Bauer, T., Fallnich, C., Kasenbacher, A. & Arnold, W. H. Femtosecond
lasers as novel tool in dental surgery. Appl. Surf. Sci. 197–198, 737–740 (2002).
10. Hoy, C. L. et al. Clinical ultrafast laser surgery: recent advances and
future directions. IEEE J. Sel. Top. Quant. Electron. 20, 242–255 (2014).
11. Loesel, F. H. et al. Non-thermal ablation of neural tissue with femtosecond
laser pulses. Appl. Phys. B 66, 121–128 (1998).
12. Chichkov, B. N., Momma, C., Nolte, S., Alvensleben, F. & Tünnermann, A.
Femtosecond, picosecond and nanosecond laser ablation of solids.
Appl. Phys. A 63, 109–115 (1996).
13. Bauer, F., Michalowski, A., Kiedrowski, T. & Nolte, S. Heat accumulation in
ultra-short pulsed scanning laser ablation of metals. Opt. Express 23,
1035–1039 (2015).
14. Vogel, A., Noack, J., Hüttman, G. & Paltauf, G. Mechanisms of femtosecond
laser nanosurgery of cells and tissues. Appl. Phys. B 81, 1015–1047 (2005).
15. Sutton, G. P. & Biblarz, O. Rocket Propulsion Elements Ch.14 (Wiley, 2011).
16. Cho, Y. I., Hartnett, J. P. & Rohsenow, W. M. Handbook of Heat Transfer 6.21
(McGraw-Hill, 1998).
17. Hashida, M. et al. Ablation threshold dependence on pulse duration for copper.
Appl. Surf. Sci. 197–198, 862–867 (2002).
18. Ancona, A. et al. High speed laser drilling of metals using a high repetition rate,
high average power ultrafast ber CPA system. Opt. Express 16, 8958–8968
(2008).
19. Kalaycioglu, H., Eken, K. & Ilday, F. O. Fiber amplication of pulse bursts up to
20 μJ pulse energy at 1 kHz repetition rate. Opt. Lett. 36, 3383–3385 (2011).
20. Kalaycıog˘ lu, H. et al. 1 mJ pulse bursts from a Yb-doped ber amplier.
Opt. Lett. 37, 2586–2588 (2012).
21. Kalaycıog˘ lu, H., Akcaalan, O., Yavas, S., Eldeniz, Y. B. & Ilday, F. Ö. Burst-mode
Yb-doped ber amplier system optimized for low-repetition-rate operation.
J. Opt. Soc. Am. B 32, 900–906 (2015).
22. Lapczyna, M., Chen, K. P., Herman, P. R., Tan, H. W. & Marjoribanks, R. S. Ultra
high repetition rate (133 MHz) laser ablation of aluminum with 1.2-ps pulses.
Appl. Phys. A 69, S883–S886 (1999).
23. Hu, W., Shin, Y. C. & King, G. Modeling of multi-burst mode pico-second laser
ablation for improved material removal rate. Appl. Phys. A 98, 407–415 (2010).
24. Marjoribanks, R. S. et al. Ablation and thermal eects in treatment of hard and
soft materials and biotissues using ultrafast-laser pulse-train bursts.
Photon. Lasers Med. 1, 155–169 (2012).
25. Tsai, P. S. et al. All-optical histology using ultrashort laser pulses. Neuron 39,
27–41 (2003).
26. Meshulach, D. & Silberberg, Y. Coherent quantum control of two-photon
transitions by a femtosecond laser pulse. Nature 396, 239–242 (1998).
27. Schwoerer, H. et al. Laser-plasma acceleration of quasi-monoenergetic protons
from microstructured targets. Nature 439, 445–448 (2006).
Supplementary Information is available in the online version of the paper.
Acknowledgements This work was supported partially by the European Research
Council (ERC) Consolidator Grant ERC-617521 NLL, the European Union
(EU) FP7 CROSS TRAP and TÜBITAK under projects 112T980, 112T944 and
TEYDEB-3110216. C.K. acknowledges funding from TÜBITAK - BIDEB 2211. We
thank Y. Aykaç and V. Aykaç for dental experiments, T. Dalkara, M. Yemis¸ çi Özkan,
K. Kılıç for brain tissue experiments, G. Aykut for animal care and brain slicing,
I. Mirza, K. Yavuz, G. Makey and M. Karatok for data acquisition and analyses,
S. Karahan for histology analyses, A. Büyüksungur and BIOMATEN (METU,
Ankara, Turkey) for micro-CT analyses, H. Köymen for PZT characterisation and
S. Ilday, O. Tokel, H. Çelik, O. Algın and E. Atalar for critical reading of the manuscript.
Author Contributions C.K., H.K. and F.Ö.I. designed the research and interpreted
the results. H.K., P.E., S.Y., Ö.A., and C.K. developed the laser systems. H.H.
and R.H. developed a high-repetition-rate fibre oscillator. C.K., D.K.K. and B.Ö.
performed the laser processing experiments. B.Ç. and C.K. developed the
numerical models. M.D.A. carried out brain slicing and histological examinations.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare competing financial interests:
details are available in the online version of the paper
. Readers are welcome to
comment on the online version of the paper. Correspondence and requests for
materials should be addressed to F.Ö.I. (ilday@bilkent.edu.tr).
Reviewer Information Nature thanks K. Mitra and the other anonymous
reviewer(s) for their contribution to the peer review of this work.
© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.