Two-photon absorption coefficient

determination with a differential F-scan

technique
E RUEDA,1 J H SERNA,2 A HAMAD AND H GARCIA3,*
1Grupo de Óptica y Fotónica, Instituto de Física, U de A, Calle 70 No. 52-21, Medellín, Colombia
2Grupo de Óptica y Espectroscopía, Centro de Ciencia Básica, Universidad Pontificia Bolivariana, Ca.
1 No. 70-01, Campus Laureles, Medellín, Colombia
3Department of Physics, Southern Illinois University, Edwardsville, Illinois, 60026, USA

[email protected]

Abstract: In this paper we present a modification to the recently proposed transmission F-scan
technique, the differential F-scan technique. In differential F-scan technique the programmed
focal distance in the electronic-tunable lens oscillates, allowing the light detector of the setup
to record a signal proportional to the derivative of the signal recorded with an F-scan. As for
the differential Z-scan a background-free signal is obtained, but also the optical setup is
simplified and the available laser power is double. We also present and validate a new fitting-
procedure protocol that increments the accuracy of the technique. Finally, we show that fitting
a signal from differential F-scan or the derivate of the signal of transmission F-scan is more
accurate than simply fitting the signal from F-scan directly. Results from two-photon absorption
at 790 nm of CdS, ZeSe and CdSe are presented.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
OCIS codes: (190.0190) Nonlinear optics; (120.0120) Instrumentation, measurement, and metrology; (190.4180)
Multiphoton processes.

References and links

1. J. H. Bechtel and W. L. Smith, "Two-photon absorption in semiconductors with picosecond laser pulses," Phys.
Rev. B 13(8), 3515–3522 (1976).
2. E. W. Van Stryland, M. a. Woodall, H. Vanherzeele, and M. J. Soileau, "Energy band-gap dependence of two-
photon absorption," Opt. Lett. 10(10), 490 (1985).
3. 1M. Sheik-Bahae, A. a. Said, T.-H. Wei, D. J. Hagan, and E. W. Van Stryland, "Sensitive measurement of
optical nonlinearities using a single beam," IEEE J. Quantum Electron. 26(4), 760–769 (1990).
4. T. Xia, D. J. Hagan, M. Sheik-Bahae, and E. W. Van Stryland, "Eclipsing Z-scan measurement of lambda/10^4
wave-front distortion," Opt. Lett. 19(5), 317–319 (1994).
5. W. Zhao and P. Palffy-Muhoray, "Z-scan technique using top-hat beams," Appl. Phys. Lett. 63(12), 1613
(1993).
6. M. Martinelli, S. Bian, J. R. Leite, and R. J. Horowicz, "Sensitivity-enhanced reflection Z-scan by oblique
incidence of a polarized beam," Appl. Phys. Lett. 72(12), 1427–1429 (1998).
7. M. Martinelli, S. Bian, J. R. Leite, and R. J. Horowicz, "Sensitivity-enhanced reflection Z-scan by oblique
incidence of a polarized beam," Appl. Phys. Lett. 72(12), 1427–1429 (1998).
8. J. Serna, E. Rueda, and H. García, "Nonlinear optical properties of bulk cuprous oxide using single beam Z-
scan at 790 nm," Appl. Phys. Lett. 105(19), 191902 (2014).
9. J.-M. Ménard, M. Betz, I. Sigal, and H. M. van Driel, "Single-beam differential z-scan technique," Appl. Opt.
46(11), 2119 (2007).
10. R. Kolkowski and M. Samoc, "Modified Z -scan technique using focus-tunable lens," J. Opt. 16(12), 125202
(2014).
11. J. Serna, A. Hamad, H. Garcia, and E. Rueda, "Measurement of nonlinear optical absorption and nonlinear
optical refraction in CdS and ZnSe using an electrically focus-tunable lens," in International Conference on
Fibre Optics and Photonics, 2014 (2014).
12. R. L. Sutherland, Handbook of Nonlinear Optics (CRC Press, 2003).
13. T. Krauss and F. Wise, "Femtosecond measurement of nonlinear absorption and refraction in CdS, ZnSe, and
ZnS," Appl. Phys. Lett. 65(March 1994), 1739–1741 (1994).
14. B. S. Wherrett, "Scaling rules for multiphoton interband absorption in semiconductors," J. Opt. Soc. Am. B
1(1), 67 (1984).
1. Introduction
In the past decades, several optical techniques have been proposed to measure nonlinear optical
properties such as two-photon absorption (TPA) coefficient for different types of materials,
especially metals, organic and inorganic semiconductors [1,2]. Among them, Z-scan is a
particularly widely used technique due to its relatively simple optical setup and data treatment
[3]. It is based on the scanning of spatial beam modifications suffered by a laser beam after
interacting with a sample, while it is focused and defocused in time: when the sample is near
the focal point of the beam the high intensities generated produce nonlinear phenomena such
variations in the refractive index and multi-photon absorption. However, some problems,
related to laser fluctuations, beam alignment and mechanical vibrations can influence the results
obtained, compromising the sensitivity of the technique. To overcome these limitations,
modifications to the basic Z-scan setup were proposed in the following years. To enhance the
sensitivity of the technique Xia et al. [4] replaced the far-field aperture in the standard Z-scan
by an obscuration disk that blocks most of the beam, while Zhao et al. [5] used a top-hat beam
instead of a Gaussian beam, and Martinelli et al. [6] measure the reflected beam from the sample
in a Brewster angle configuration. To improve the signal to noise ratio some authors have used
balance-detection systems [7,8], while Ménard et al. [9] introduce the differential Z-scan where
a piezo-transducer device generates an oscillatory motion to induces a periodic modulation of
the beam intensity at the sample, which in turn produces a modulation of the transmitted light
proportional to the spatial derivative of the transmitted light, and therefore provides a
background-free measurement. More recently, a new technique, that is a variation of Z-scan,
was proposed. This method, which is called F-scan [10,11], use an electrically focus-tunable
lens (EFTL) instead of a fixed lens to generate different focal points, allowing to replace the
translation stage and leaving the sample fixed in space, i.e. eliminating mechanical movements
from the setup. The focal distance in the EFTL is a function of the applied current to the lens.
Analogously to Ménard et al. for the case of Z-scan, we present in this paper an open aperture
differential F-scan technique (DF-scan) to determine the nonlinear absorption optical properties
of materials. The setup has been oversimplified by using an EFTL and at the same time
modulating the focal length of the EFTL with a rectangular low frequency signal that can be
detected with a PSD (phase sensitive detector), in our case a lock-in amplifier, increasing the
sensitivity of the system and reducing or eliminating laser fluctuations.
2. Transmission F-scan (TF-scan)
The F-scan experimental setup depicted in Fig. 1. is used for the determination of the two-
photon absorption (TPA) coefficient (open aperture architecture).
 Fig. 1. TF-scan experimental optical setup for determination of TPA coefficients (open-aperture
architecture).

A laser Gaussian beam modulated with a chopper impinges on an EFTL which is a lens
that has the capability to vary its focal distance f over a specific range when an electric current
is applied to it (see Fig. 2), focusing the Gaussian beam at different positions. The sample is
placed at a fixed position d s inside the range of the EFTL. The light transmitted through the
sample is collected by a photodetector PD1. The output signal is then filtered with a Lock-in
amplifier and processed with an acquisition data system.

Fig. 2. OPTOTUNE-10-30-C electrically focus-tunable lens. This type of lens changes its shape
(curvature) due to an optical fluid sealed off by a polymer membrane, when a current is applied.

To determine the TPA coefficient β we measured the transmittance of the nonlinear medium
as a function of the focal length, f, (see Fig. 1), collecting all light transmitted through the
sample. When the distance |𝑑𝑠 − 𝑓| is large the normalized transmittance has a value close to
unity because linear optical effects are produced in the sample. In contrast, small values of
|𝑑𝑠 − 𝑓| imply that the laser beam is focused near the sample, thus increasing the optical
intensity and generating nonlinear optical phenomena such as TPA. The TPA coefficient is
obtained by fitting a theoretical curve (Eq. (9)) to the experimental data, using β as the fitting
parameter, and under the assumption that all the experimental parameters are known. A typical
experimental curve is shown in Fig. 10(left).
EFTL characterization
Fig. 3 shows the dependence of the EFTL optical power  on the applied current and expose
the disagreement between the experimental data and the data reported by the manufacturer. In
the T-Fscan technique it is crucial to know the focal length with high accuracy in order to obtain
correct values of  . Therefore, the characterization of the EFTL optical power on the applied
current has to be done with high-precision.
 Fig. 3. The optical power,  , of the EFTL as a function of the applied current. (circles)
Experimental data; (continuous line) fit of the experimental data using Eq. (1); (dashed line) data
provided by Optotune Inc.

Different techniques can be used to obtain the correct dependence of the EFTL optical
power on the applied current (circles in Fig. 3) by measuring the focal length f as a function
of current. We used a laser beam profiler to measure f as a function of the applied current.
Another way is to use the TPA phenomena: by placing a sample with a nonzero TPA coefficient
at two different distances from the EFTL and finding the values of the applied current that
produce the lowest intensity for each of the locations. The sample distance from the EFTL will
correspond to the EFTL focal length that produced a minimum transmission. Then, using the
relation   1 / f , the EFTL optical power is obtained as a function of current. In our case the
experimental data was fitted obtaining the following expression (continuous line in Fig. 3):
  0.045 J  1.522 (1)
where J is the applied current measured in mA. Another important experimental parameter
for the correct determination of the nonlinear optical parameters is the beam-waist radius w0 .
Typically, for spherical lenses and assuming that the beam has a spatial Gaussian profile, the
radius of the beam at the beam waist is determined as a function of the focal length with the
equation:
2 f
w0 ( f )  (2)
D
where  is the wavelength of the incident beam with spot diameter D . But being aware
of the existence of optical aberrations on the optical system that distort the wavefront of the
beam, a correction factor C f is needed in order to correctly calculate the beam waist. Thus,
Eq. (2) is replaced by
2 f
w0 ( f )  Cf (3)
D
 To determine the correction factor we used a laser beam profiler to measure the beam waist
at each focal plane. Then, by using C f as the fitting parameter between the experimental data
and Eq. (3), a correction factor C f  1.36 was obtained for the special case of our EFTL. Fig.
4 shows the difference between the corrected and non-corrected beam-waist diameter value as
a function of the EFTL focallength.

Fig. 4. Beam-waist diameter as a function of EFTL focal length. (circles) Experimental data
measured with a laser beam profiler; (dashed line) beam waist diameter calculated with Eq. (2);
(continuous line) beam waist diameter calculated with Eq. (3) and C  1.36 .
f

Once the beam-waist radius is correctly determined, it is possible to calculate with precision
the beam radius w( f ) (Eq. (4)) at the sample surface for every programed EFTL focal length,

2
d  f 
w( f )  1   s 
 z0 ( f )  , (4)
where z0 ( f )   w0 ( f ) /  is the Rayleigh range. Fig. 5 shows the dependence of the beam
2

radius at the sample location as a function of the EFTL focal length. Notice the difference
between the results obtained with and without the corrected beam-waist radius. Also notice the
asymmetry relative to the sample location. For f values smaller than d s the radius at the
sample increases faster than those for f values larger than d s . This will cause the shape of
the TF-scan to be asymmetric around d s . Therefore, experimentally, we must normalize the
transmitted intensity relative to the intensity corresponding to the shortest focal lengths used in
the experiment.
 Fig. 5. Beam radius at the sample as a function of the EFTL focal length. The continuous and
dashed lines are the corrected (Eq. (3)) and not corrected (Eq. (2)) beam waist radius w0

respectively. For this plot we used D  2.0 mm and d s  10.7 cm .

Theoretical background
The laser beam in our experimental setup has a Gaussian spatial profile and a hyperbolic secant
temporal profile. Therefore, at the front surface of the sample, the incident beam intensity as a
function of the EFTL focal length f is given by:

  r 2  2  t 
I in  r , f , t   I 0 ( f ) exp  2    sech   (5)
  w( f )    0 
In the above equation r is the radial position with respect to the optical axis, t is time,
 0   / (2 ln(1  2 )) , where  is the full width at half-maximum pulse duration and I 0 ( f )
is the peak intensity of the beam at sample position as a function of the EFTL focal length:

2 ln(1  2) Pavg
I0 ( f )  . (6)
 w ( f )
2

Here Pavg is the average power of the incident laser beam at the sample, and  is the laser
pulse repetition rate. The intensity at the exit surface of the sample can be written as:
 L
(1  R ) I in ( r , f , t )e
2

I out ( r , f , t )  , (7)
1   (1  R ) I in ( r , f , t ) Leff
where L is the thickness of the sample, R is the reflection coefficient of the sample,  is
the linear absorption coefficient,  is the two-photon absorption coefficient (TPA), and

Leff  1  e
 L
/ is the effective sample thickness. Thus, the transmittance at the detector
plane can be express as:

1
T( f )   ln 1  B( f )sech (  )  d ,
2
(8)
B( f ) 0
where B( f )   (1  R) I 0 ( f ) Leff and   2 ln(1  2)t /  . The transmittance given by Eq. (8)
can be simplified when B ( f )  1 [12]:

N
   B ( f ) m m
 2( m  n)   mn  
T( f )    .
 2( m  n)  1  
(9)
m0  m 1 n0

As B ( f ) gets closer to one, more terms of the sum are need it. Thus, one must use at least
N  11 in order to guarantee that the obtained value of  is not underestimated. Otherwise,
the obtained value of  will be smaller than its actual value.

3. Differential F-scan (DF-scan)

To reduce noise in TF-scan, or in any intensity scanning technique, due to laser fluctuations
where the change in the transmission is small compared to these fluctuations, Ménard et al. [9]
proposed a method where the sample was mount on an oscillating-actuator in a Z-scan setup,
thus, for an amplitude S and frequency F the transmission signal around (and near) position
z 0 will be

T  T ( z0 )  S 
 T  sin(2 Ft )
 , (10)
 z  z  z 0

where t is time. Eq. (10) is valid as far as the oscillating amplitude S is comparable to the
Rayleigh range of the beam. If a lock-in amplifier is used with a reference signal F coming
from the piezoelectric actuator, then only the amplitude of the signal given by Eq. (11) is
detected by the lock-in amplifier:

 T 
S  . (11)
 z  z  z0
This background-free technique reduces the laser fluctuation noise and improves the
sensitivity of the technique.
One feature of the EFTL is that its focal length can be modulated with different types of
signal profiles (rectangular, triangular or sinusoidal) with a frequency range in its modulation
between 0.2 up to 2000 Hz. This allowed us to modify the TF-scan into a DF-scan without
using piezoelectric actuators, and also using the modulated signal as the reference for a lock-in
amplifier; being the signal proportional to the derivative of the transmitted signal.
Thus, for the case of the DF-scan technique, z is replaced by the focal distance f , and
using Eq. (9) the derivative in Eq. (11) becomes

 ds 
 f  k 2 f 3  f  ds    2m   B ( f ) m 1 m  2( m  n)   mn  
 T    N

   1
   m  1   2(m  n)  1   , (12)
 f  f  f 0
k2 m0  n0  
4 C f
2
 (1  R ) Leff Pavg
where k1  , and k 2  2 ln(1  2) .
D
2
k1

“Normalized” DF-scan
For the case of TF-scan, normalization of the experimental data in order to fit it by using the
analytical model is simply done by dividing the data with respect to the lock-in amplifier signal
 A for the shortest focal length. In a DF-scan setup this is not so direct because the signal
detected by the lock-in amplifier at this same focal length is null. Knowing that for any focal
length, programmed in a EFTL, the signal detected is

 T 
AS   , (13)
 f  f  f
0

in order to obtain the “normalized” DF-scan signal of Eq. (12), which is independent of the
modulation parameters A and S , the signal represented by Eq. (13) has to be divided by the
factor AS , which is known in advance. To show this, in our experiment we programed the
EFTL to vary its focal length using a square signal of frequency F  739 Hz and 50% duty
cycle, and for S amplitudes corresponding to driven currents of 0.5 and 1.0 mA (see Fig.
6(left)). The voltage detected by the lock-in amplifier for the shortest focal length, when it is
not being modulated, was 46.23 mV, corresponding to A = 46.23/2 mV. Division by two is
because of the 50% duty cycle of the square signal. Fig. 6(right) shows the corresponding
normalized DF-scan signals for both used amplitudes and for the corresponding derivative of
the normalized TF-scan; they are in excellent agreement. These normalized DF-scan signals
can know be fitted using Eq. (12).

Fig. 6. (left) Un-normalized DF-scan signals for two oscillating amplitudes S, 0.5 mA and 1.0
mA. (right) Normalized DF-scan signals after dividing by amplitude S, voltage A and 2 due to
the 50% duty cycle. The corresponding Normalized TF-scan signal derivative is also shown.

4. Experimental data fitting protocol

A correct experimental data fitting is crucial in order to obtain a reliable value of the parameter
of interest; in this case: the two-photon absorption coefficient. Is then desirable to know the
values of all the experimental parameters with the highest precision possible in order to use
only the parameter of interest as the fitting variable, especially if the technique requires the
knowledge of a great number of parameters (for our case 10 experimental parameters have to
be known). But in occasions this is not possible, and the experimental parameters are only
known with a considerable uncertainty. As a first approach, one can use only the central values
of the experimental parameters, ignoring the uncertainties, but this must probably will end in
an inaccurate fitting and thus a wrong value of the parameter of interest (see for example Fig.
7(left) and set 1 in table 2). One naive approach will be to use more than one parameter as
fitting variables, but because the model used does not have a unique solution, as is the case for
TF-scan and DF-scan, one can end with a wrong value of the parameter of interest, although
with a perfect curve fitting, as it is shown for example in Fig. 7(center).
Based on the assumption that the uncertainties of the experimental parameters correspond
to random fluctuations that can be model with a Gaussian distribution, we implemented a
protocol with the goal to obtain an interval that contains the real value of the parameter of
interest, under a statistical approach. The protocol is the following:
 1. For each experimental parameter pick a random value from the Gaussian distribution
of the possible values.
2. Fit the experimental data an obtain the corresponding value for the parameter of
interest.
3. Calculate a metric to evaluate the quality of the fit. If the metric satisfies a criteria keep
the value of the parameter of interest. If not, discard it.
4. Repeat steps 1-3 until a distribution with a good sample size of acceptable values is
obtained.
5. With the parameters of interest that were accepted calculate the average-weighted
value, using each corresponding metric value as weights. Calculate the corresponding
standard error. The average-weighted value corresponds to the parameter of interest.
To show the effectiveness of this approach a simulated experimental data “Real” from a
sample with nonlinear TPA is presented. In table 1 the parameters used for the simulated data
are presented.
Table 1. Simulation: experimental parameters.

L (mm) d s (mm)  (nm) D (mm) Cf

0.8 116.0 790 2.0 1.36
Pavg (mW)  (fs)  (MHz)  (1/m) R
11
145 71 90.9 2.64  10 0.1567
To retrieve  we supposed that the experimental parameters are known with a 10%
uncertainty. Then, three sets of experimental parameters are pick from their normal
distributions: set 1 and 2 correspond to a focal distance range from 8 to 16 cm with a total of
300 sample points, and set 3 correspond to the same range but 50 sample points. Finally, the
data is fitted and  is obtained using the central values “Direct” approach, our approach
“Protocol”, and the multiple fitting-parameters “Multiple” approach. For “Multiple” approach
D ,  , Pavg , and d s are also use as fitting parameters. In table 2 the results are presented.

Table 2. Comparison of the  values, in cm/GW, retrieved from the different data-fitting techniques.

Set 1 Set 2 Set 3

 Error %  Error %  Error %

Real 3.4 3.4 3.4

Direct 2.2 36 3.8 13 5.3 53
Multiple 33.9 874 14.5 325 46.4 1265
Protocol 3.3 2 3.0 11 3.5 2
It is clear that the “Multiple” approach is not desirable because it most probably return a
wrong  value while giving an almost perfect data fitting curve, see Fig. 7(center), Fig.
8(center) and Fig. 9(center). The “Direct” approach depends directly on the experimental
parameters chosen, and due to the uncertainty, this can mean that the value retrieved is close to
the correct one (Fig. 8(left)) or far (Fig. 7(left) and Fig. 9(left)). Finally, our proposal,
“Protocol”, gives the certainty that it will always retrieve the value with the greatest possible
accuracy (see Fig. 7(right), Fig. 8(right) and Fig. 9(right)).
 Fig. 7. Fitting results of the simulated data for Set 1 in table 2. (Left) “Direct” approach, (center)
“Multiple” approach, (right) “Protocol” approach.

Fig. 8. Fitting results of the simulated data for Set 2 in table 2. (Left) “Direct” approach, (center)
“Multiple” approach, (right) “Protocol” approach.

Fig. 9. Fitting results of the simulated data for Set 3 in table 2. (Left) “Direct” approach, (center)
“Multiple” approach, (right) “Protocol” approach.

5. Experimental results
For the experimental implementation of the TF-scan and DF-scan we used a Ti:Sapphire
oscillator laser with repetition rate of 90.9 MHz, pulse width of 71 fs, and laser emission
centered at 790 nm. The average power at the entrance surface of the sample was 145 mW. The
beam diameter at the EFTL was D = 2.0 mm, and was measured by a laser beam profiler. The
EFTL is an OPTOTUNE-1030, controlled by an OPTOTUNE lens-driver that gives a
maximum current of 300 mA with a resolution of 0.1 mA, delivering a focal length resolution
of 0.017 mm. The laser Gaussian beam is focused at quasi-normal incidence in order to
eliminate Fabry-Perot effects and multiple reflections. We used an integrating sphere with a
large area Si-photodiode (PDA 50 THORLABS) to measure the transmitted laser light. This
 modification to the common setup compensates any lens-divergence and eliminates signal
losses due to scattering from the sample-surface roughness. The current generated by the
photodiode is sent to a STANFORD RESEARCH 830 dual channel Lock-in amplifier,
controlled through a GPIB interface.
We have measured the TPA coefficient at 790 nm for ZnSe, CdS and CdSe. The
experimental parameters for both techniques and all materials are listed in table 3, and the
obtained TPA coefficients are listed in table 4. In Fig. 10 the experimental curves and curve
fitting is presented for the case of CdSe.
Table 3. Experimental parameters for TF-scan and DF-scan, for CdS, ZnSe and CdSe.

 (MHz) d s (mm)  (nm) D (mm) Cf

90.9 116.0 ± 0.5 790 ± 1 2.0 ± 0.8 1.36 ± 0.01
Pavg (mW)  (fs) LZnSe (mm)  ZnSe (1/m) RZnSe
145 ± 5 71.0 ± 0.3 0.80 ± 0.01 4.772 0.182  0.005
LCdS (mm)  CdS (1/m) RCdS LCdSe (mm)  CdSe (1/m)

0.85 ± 0.01 2.64  10

11
0.157  0.005 0.79 ± 0.01 369  37

RCdSe
0.185  0.005

Table 4. Comparison of  values, in cm/GW at 790 nm, for ZnSe, CdS and CdSe.

ZnSe CdS CdSe

Relative Relative  Relative
 error %  error % error %
DF-scan 5.1 23 2.4 22 4.6 12
TF-scan 3.2 16 1.5 13 1.8 15
* *
Krauss [13] 3.5 >35 6.4 >35
*at 780 nm.
From table 4 it is evident that there exists a discrepancy between TF-scan and DF-scan
results, in particular, DF-scan always gives a bigger value. This discrepancy will be analyzed
and explained in the fallowing subsection. For the value reported by Krauss et al. [13] for ZnSe,
and from a statistical point of view, there is a probability of 82% and 35% that the difference
is due to a random fluctuation with respect to TF-scan and DF-scan, respectively. For the case
of CdS the probabilities are 3% and 8%, respectevely. A random deviation of this magnitude
in the values is not consider rare. For CdSe we were not able to find a value for wavelengths
close to 790 nm.

Fig. 10. Fitting result for CdSe. (Left) TF-scan, (right) DF-scan.
 To compare our results with the values that have been reported, in a broader way, for the
three materials, we used the expression derived by Wherrett [14] for TPA,

(2 hc / ( E g )  1)
3/ 2

~ , (14)
(2hc / ( E g ))
5

where E g is the energy of the bandgap, c is the speed of light in vacuum, and h is the
Planck’s constant. In particular, the ratio r790, between the TPA at 790 nm with respect to the
TPA value for the other wavelengths is given by

2hc  790 E g
r790,   ( ) (
790 7/2
) 3/ 2
(15)
 2hc   E g
In Fig. 11 the ratio is plot for the three materials. Except for the case of CdSe, there is a
probability of 5% or more that the differences are due to random fluctuations.

Fig. 11.Wherrett ratio. The values for the other wavelengths are taken from Krauss et al. [13]
and V. Stryland et al. [13].

Difference between TF-scan and DF-scan results

We believe the discrepancy between the values obtained with DF-scan and TF-scan is due to
fitting robustness directly related to curve shape. In our fitting criteria the metric minimizes the
value of the sum of the distance between the calculated and experimental peaks of the curves.
The existence of two peaks in DF-scan signals reduce the spectrum of possible fitting values,
making the process more robust and accurate than with TF-scan signals. Then, the same result
must be obtained for both, the derivative of TF-scan signal and the DF-scan signal. To validate
this idea, first we perform a simulation with the values of table 1, and secondly we perform the
fitting for the derivative of the TF-scan signal of CdSe and compare it with the results of table
4. Results are presented in table 5 and Fig. 12, showing, without doubt, that fitting the derivative
of the signal of TF-scan or the signal of DF-scan outcomes a more accurate value.
Table 5. Comparison of the  values, in cm/GW, retrieved from the simulated data of table 1.

Simulated   3.4

 Error %
TF-scan 2.6 24
DF-scan 3.8 12
TF-scan
3.6 6
derivative
 Fig. 11. Fitting curves of table 5 simulated-experimental data results. (Left) TF-scan, (right) DF-
scan and TF-scan derivative.

6. Conclusions
DF-scan is a modification to the F-scan technique where the scanning is done over the rate of
change of the transmission signal with respect of the focal distance of the EFTL, reducing
drastically the sensibility to laser fluctuations, increasing the available laser power and
simplifying the optical setup by eliminating the need of a chopper to modulate the signal. For
the curve fitting step, where the experimental signal is fitted to an analytical model in order to
obtain the TPA coefficient, a new protocol is proposed in order to secure the correct
determination of the TPA parameter. The effectiveness of the proposal has been validated with
respect to other curve fitting procedures. Finally, it was shown that it is more reliable to fit the
derivatives of the experimental signals than the signal itself, either of the DF-scan signal or the
derivative of the TF-scan signal.
Acknowlegments
E. Rueda thanks Universidad de Antioquia for financial support. J. Serna acknowledges the
support from Universidad Pontificia Bolivariana. H. Garcia and A. Hamad thanks Southern
Illinois University, Edwardsville, for financial support.