Electro-optic thin films of organic nonlinear

optic molecules aligned through vacuum

deposition
Weiwei Sun,1,2 Zhaohong Wang,1,4 Antao Chen,1,2,* Ilya Kosilkin,3 Denise Bale,3 and
Larry R. Dalton2,3
1
Applied Physics Laboratory, University of Washington, Seattle 98105, USA
2
Electrical Engineering Department, University of Washington, Seattle 98105, USA
3
Chemistry Department, University of Washington, Seattle 98105, USA
4
Key Laboratory of Physical Electronics and Devices of the Ministry of Education, Xian Jiaotong University, China
*
antaochen@apl.washington.edu

Abstract: Nonlinear optical molecules can be vacuum deposited into

uniform thin films using thermal evaporation. Alignment order can be
achieved during thin film deposition by an in-plane electrical field poling
using electrodes patterned on the substrate. Electro-optic (EO) coefficients,
r33 and r13 are independently measured using Young's interferometry
technique. Thin-films of N-benzyl-2-methyl-4-nitroaniline (BNA) can
exhibit an EO coefficient, r33, comparable to that of BNA single crystals.
EO coefficients of BNA at different poling fields, wavelengths, and
frequencies are investigated.
2011 Optical Society of America
OCIS codes: (190.4710) Optical nonlinearities in organic materials; (120.3180) Interferometry.

References and links

1. D. Burland, Optical nonlinearities in chemistry: introduction, Chem. Rev. 94(1), 12 (1994).
2. M. G. Kuzyk and C. W. Dirk, Characterization Techniques and Tabulations for Organic Nonlinear Optical
Materials (Marcel Dekkar, 1998).
3. T. A. Skotheim and J. R. Reynolds, Handbook of Conducting Polymers. Conjugated polymers: Theory,
Synthesis, Properties, and Characterization, 3rd ed. (CRC Press, 2007).
4. K. D. Singer, J. E. Sohn, and S. J. Lalama, Second harmonic generation in poled polymer films, Appl. Phys.
Lett. 49(5), 248250 (1986).
5. T. D. Kim, J. W. Kang, J. Luo, S. H. Jang, J. W. Ka, N. Tucker, J. B. Benedict, L. R. Dalton, T. Gray, R. M.
Overney, D. H. Park, W. N. Herman, and A. K. Jen, Ultralarge and thermally stable electro-optic activities from
supramolecular self-assembled molecular glasses, J. Am. Chem. Soc. 129(3), 488489 (2007).
6. D. Rezzonico, S. J. Kwon, H. Figi, O. P. Kwon, M. Jazbinsek, and P. Gnter, Photochemical stability of
nonlinear optical chromophores in polymeric and crystalline materials, J. Chem. Phys. 128(12), 124713 (2008).
7. H. Hashimoto, Y. Okada, H. Fujimura, M. Morioka, O. Sugihara, N. Okamoto, and R. Matsushima, Second-
harmonic generation from single crystals of N-Substituted 4-Nitroanilines, Jpn. J. Appl. Phys. 36(Part 1, No.
11), 67546760 (1997).
8. M. Fujiwara, K. Yanagi, M. Maruyama, M. Sugisaki, K. Kuroyanagi, H. Takahashi, S. Aoshima, Y. Tsuchiya, A.
Gall, and H. Hashimoto, Second order nonlinear optical properties of the single crystal of N-Benzyl 2-methyl-4-
nitroaniline: Anomalous enhancement of the d333 component and its possible origin, Jpn. J. Appl. Phys. 45(11),
86768685 (2006).
9. M. Fujiwara, M. Maruyama, M. Sugisaki, H. Takahashi, S.- Aoshima, R. J. Cogdell, and H. Hashimoto,
Determination of the d-tensor components of a single crystal of N-Benzyl-2-methyl-4-nitroaniline, Jpn. J.
Appl. Phys. 46(No. 4A), 15281530 (2007).
10. K. Iwaszczuk, D. G. Cooke, M. Fujiwara, H. Hashimoto, and P. Uhd Jepsen, Simultaneous reference and
differential waveform acquisition in time-resolved terahertz spectroscopy, Opt. Express 17(24), 2196921976
(2009).
11. K. Miyamoto, S. Ohno, M. Fujiwara, H. Minamide, H. Hashimoto, and H. Ito, Optimized terahertz-wave
generation using BNA-DFG, Opt. Express 17(17), 1483214838 (2009).
12. S. Kalluri, S. Garner, M. Ziari, W. H. Steier, Y. Shi, and L. R. Dalton, Simple two-slit interference electro-optic
coefficients measurement technique and efficient coplanar electrode poling of polymer thin films, Appl. Phys.
Lett. 69(2), 275277 (1996).
13. P. Gnter, C. Bosshard, K. Sutter, H. Arend, G. Chapuis, R. J. Twieg, and D. Dobrowolski, 2-cyclooctylamino-

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11189
 5-nitropyridine, a new nonlinear optical crystal with orthorhombic symmetry, Appl. Phys. Lett. 50(9), 486488
(1987).
14. O. P. Kwon, B. Ruiz, A. Choubey, L. Mutter, A. Schneider, M. Jazbinsek, V. Gramlich, and P. Gunter, Organic
nonlinear optical crystals based on configurationally locked polyene for melt growth, Chem. Mater. 18(17),
40494054 (2006).
15. R. P. Fraser, H. Sun, and P. Sullivan, Use of spatially anisotropic intermolecular interactions to define order in
molecular films grown by physical vapor deposition (PVD), in Photonics, CMDITR Review of Undergraduate
Research, Vol. 6, (2009).
16. C. Hunziker, S.-J. Kwon, H. Figi, F. Juvalta, O.-P. Kwon, M. Jazbinsek, and P. Gnter, Configurationally
locked, phenolic polyene organic crystal 2-{3-(4-hydroxystyryl)-5,5-dimethylcyclohex-2-
enylidene}malononitrile: Linear and nonlinear optical properties, J. Opt. Soc. Am. B 25(10), 16781683 (2008).
17. O. P. Kwon, S. J. Kwon, H. Figi, M. Jazbinsek, and P. Gnter, Organic electro-optic single- crystalline thin
films grown directly on modified amorphous substrates, Adv. Mater. (Deerfield Beach Fla.) 20(3), 543545
(2008).
18. J. L. Oudar and D. S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state
dipole moment, J. Chem. Phys. 66(6), 26642668 (1977).

1. Introduction
Organic electro-optic (EO) materials have been actively researched for decades for a broad
range of electro-optic device applications [13]. The driving force of this research is primary
due to the inherently large EO coefficients achievable with organic materials, along with their
low dielectric permittivity, which makes them suitable for high bandwidth devices. The EO
effect of these materials arises from the second-order nonlinear optic (NLO) molecules
(dipolar chromophores) preferentially aligned in one orientation. The molecules must have
overall noncentrosymmetric alignment for the macroscopic EO effect to occur. The most
common organic EO materials are EO polymers, where the nonlinear optic molecules are
imbedded in a passive polymer matrix. The alignment of the NLO molecules is typically
achieved by electric field poling. In classical poling processing methods [4], the EO polymer
film is heated close to its glass transition temperature while applying a DC electric field. The
poling results in statistical polar orientation of the molecular dipoles along the field direction,
with the order being maintained when the film is cooled down to room temperature with the
poling field on. EO coefficients greater than 300 pm/V have been achieved in EO polymers
[5].
However, EO polymer materials have limitations. Specifically, the loading density
(concentration) of the NLO molecules in the polymer host is usually limited to less than 25%
by weight. Some device designs are not compatible with solvent processing required in the
fabrication of EO polymer thin films as well as the high temperature during the poling; thus, a
dry (e. g. vacuum deposition) process to fabricate thin films of organic EO materials is
preferred in this case. Organic EO crystal materials have properties complementary to EO
polymers and neat thin films. One primary difference is that organic EO crystals have high-
density chromophore packing and excellent orientation stability up to the melting temperature
of the material. EO crystals also have much better photostability than EO polymers [6]. For
these reasons a vacuum deposition process to achieve thin films of organic EO materials that
tend to form non-centrosymmetric crystals merits exploration.
In this paper, a method of fabricating organic EO thin films through simultaneous thermal
evaporation/deposition and electric-field poling is presented. Polycrystalline thin films of
small angle domains are achieved. EO coefficients of thin films of N-benzyl-2-methyl-4-
nitroaniline (BNA) measured using a patterned substrate and Young's interferometry
technique exhibited high EO coefficients that are comparable to those of bulk single crystals.
EO coefficients of BNA were investigated at multiple wavelengths, different applied DC
voltages during deposition, and measurement frequencies.
2. EO material and test sample fabrication
The chemical structure of BNA is shown is Fig. 1(a) [7]. Similar to other common second-
order organic NLO molecules, it has a permanent dipole moment because the electron

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11190
donating group and electron accepting are located at separate ends of the molecule. BNA is a
solid crystal at room temperature, with a melting point of 105 C at normal atmospheric
pressure [7], and is one of the second-order NLO molecules that can form non-
centrosymmetric single crystal from melt [8]. It has a relatively large nonlinear optical
coefficient d333 of 234 31 pm/V (r33 of 45 6 pm/V) at the wavelength of 1064 nm [9]. The
NLO properties of BNA and its application in THz radiation and EO waveguide modulators
are areas of active research [8,10,11].
Samples consisted of glass substrates featuring a two-slit electrode design [12] (see Fig.
1(b) and Fig. 2). The electrodes were made of titanium with standard photolithography and
lift-off processes. Titanium was used for its good adhesion to glass. The slit widths were
patterned to be 5-20 m with a center-to-center separation of 100 m. Electrical connections
to the lectrodes were made with wires attached to the electrodes with a conductive epoxy.

Fig. 1. (a) The chemical structure of BNA; (b) Schematic side view of the sample.

A DC voltage was applied across the slits of the electrodes during the deposition of BNA
using thermal evaporation method. Figure 2(a) shows the connection of the DC voltage during
the material deposition. The voltage across the slit creates an electric field that is
perpendicular to the slit in the plane of substrate. An ammeter is used to detect electric
breakdown during deposition. The breakdown voltage during thin film deposition was found
to be above 10 V/m and it mainly depends on small protrusions in the roughness of the edges
of the electrodes as well as the dielectric strength of the deposited material. By using a
coplanar electrode configuration and electric field direction mentioned above, a torque would
be produced on the dipolar BNA molecules in the two slits to align them perpendicular to the
slits in the plane of the substrate.
The deposition was carried out using a thermal evaporator with base pressure of 5 10 6
Torr. BNA was evaporated from a thermal heater source. The deposition rate was kept below
1 nm per second by controlling the temperature of the evaporation source. BNA begins to
evaporate at 70 C, and the deposition rate increases with temperature. The deposition
temperature for the desirable deposition rate was found to be 98 C and resulted in an 800 nm
thick BNA thin film after approximately 1 h. The glass substrate was not heated and remained
at room temperature during deposition. Because of the low evaporation temperature of the
BNA and the low pressure of the vacuum, heating the substrate was found to impair film
growth. Specifically, no thin film of BNA forms on the substrate at substrate temperatures
above 60 C. Without heating the substrate, the thin film can be achieved in a uniform
coverage on the substrate. During the deposition, BNA in the two slits crystallized in small
angle domains with the polar axis aligned under the electrostatic force of the electric field. A
scanning electron microscopy (SEM) image that shows the BNA thin film around one slit is
shown in Fig. 3. The film in the image covered the edge of one slit, depicted with a dash line.
The slit is on the left hand side of the line. Its clearly shown that the BNA crystal domains in
the slit are well aligned along the E-field, while the crystal domains outside the slit are
orientated perpendicular to the electrode metal because the E-field is perpendicular to the
electrode surface. Although the polycrystalline structure of the BNA thin film is expected to

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11191
increase the scattering loss, the polycrystalline EO thin film can still be useful in device
designs that are tolerant to material loss, such as spatial light modulators and silicon photonic
crystals where long (centimeters) optical path lengths are not needed. The Youngs
interference fringes from two-slit electrodes covered with a BNA thin film were clean and
showed no noticeable effect of scattering.

Fig. 2. Top view of the sample substrate, with the polarity of voltage connection across the slits
during the deposition and EO measurement indicated. (a) DC electric field is applied across the
slits when EO material is thermally evaporated on the sample; (b) AC electric field is applied
during the measurement of EO coefficients.

Fig. 3. The SEM image of BNA thin film around the slit.

3. EO measurement based on Youngs interference

An EO measurement setup based on Youngs interference [12] was used to measure the EO
coefficients reported here. Figure 4 illustrates the measurement device and signal processing.
A photodetector was mounted on a translation stage, thereby allowing for it to move across
the interference fringes to measure the fringe pattern and its modulation when an AC voltage
is applied across one slit. The output from the detector was connected to the data acquisition
(DAQ) module to record the average DC signal, and also to a lock-in amplifier to extract the
AC modulation.
When an electric field is applied across one slit (Fig. 2(b)) containing the material
deposited, the Youngs interference pattern will shift laterally due to the change of refractive
index in the BNA in one slit. If an AC field is applied, the pattern will spatially oscillate at the
same frequency as the AC field. The oscillation can be detected as intensity modulation if the
detector is positioned on the slope of fringe. Then EO coefficient could be calculated from
AC signals and the intensity distribution of the fringe pattern. In our case the modulating AC
field is in the same direction as the polar axis of the micro-domains of the BNA. The EO

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11192
coefficient r33 was measured with polarization of the laser light parallel the electric field,
while r13 was measured with polarization perpendicular to the electric field.

Fig. 4. EO measurement setup.

Based on the detected AC signal and the interference intensity of the fringe, the EO
coefficients (r33 or r13) can be calculated via [12]
2Vac
r , (1)
n3 EtkL
where E is the electric field applied across one of the two slits; n is effective refractive index
of the material; t is the deposited film thickness; is the wavelength of the incident beam; and
L is the fringe period. Vac is the AC signal measured by a lock-in amplifier, and k is slope of
the fringe at the position of measurement.

Fig. 5. Measured DC and AC signals detected across several fringes.

During the measurement, the detector is slowly scanned across the interference pattern.
Figure 5 shows the DC and AC signals acquired during a scan. The x-axis of Fig. 5 is the
detector position. The total range of scan covers several fringes. DC signal (solid line)
represents the intensity of interference fringe, and AC signal (dotted line) represents the
intensity modulation caused by the transverse oscillation of the fringe due to the applied
modulating voltage. The AC signal is proportional to the first order derivative of the DC
signal, and the EO coefficient can be calculated from Eq. (1) if the film thickness, index of
refraction, slit width, AC modulating voltage, the first order derivative of the DC signal and
the wavelength of the measurement are known. The error of measurement comes mostly from

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11193
the measurement error of the film thickness. The thickness was measured with an optical
profiler at the flat areas adjacent to the slits. However, from SEM images of the cross-section
of the samples, the film thickness in the slot is found to be different from the film thickness in
the flat areas by as much as 25%. Errors from other sources are estimated to be about 5%.
4. Result and discussion
Several materials, including BNA [7], COANP [13], DAT2 [14], PS1 [15], JD2112 [15],
JD2116 [15], OH1 [16], EHD-1-12 [17] and EHD-1-13 [17] were examined. Among these
materials, thin films of BNA were found to exhibit the highest EO coefficient. Properties of
BNA thin films were further studied in detail, including the effect of poling field, dispersion
of the EO coefficient, variation of the EO coefficient at different modulation frequencies and
modulation voltages.
The EO coefficient r33 of BNA thin films at different poling electric field is shown in Fig.
6(a). Due to the breakdown limitation, the highest electric field for BNA sample was kept
under 10 V/m. The poling electric field was varied from 3 V/m to 10 V/m. The measured
r33 at 1550 nm was proportional to the poling filed, which compares with the linear
relationship between r33 and poling field in poled EO polymers [12].

Fig. 6. The r33 coefficient as a function of electric poling field and as a function of wavelength.
(a) The r33 coefficient as a function of electric poling field at the measurement wavelength of
1550 nm; (b) Theoretical and experimental results of EO coefficient r33 at different
wavelengths.

The ratio between EO coefficients r33 and r13 was found to be about 4.8 at a typical field
of 10 V/m. The results also show that BNA thin films made under a high poling electric field
have r33 values comparable to the BNA single crystal grown from solution [8], which has an
r33 of about 45 pm/V at 1064 nm [9].
The dispersion of r33 of BNA was measured using three laser wavelengths: 980 nm, 1310
nm and 1550 nm, respectively. The BNA sample was 800 nm thick and prepared using an
electric poling field of 10 V/m. It was tested with a modulation voltage of 5 V and frequency
of 1 kHz. The r33 values at measurement wavelengths are overlaid with a theoretical curve as
shown in Fig. 6(b). The theoretical data was calculated from the dispersion of molecular first
hyperpolarizability based on a two-level model [18] and was found to agree well with
experimental data. Figure 7(a) shows the measurement results of the EO coefficients under
different modulation frequencies. In the range between 200 Hz and 2 kHz, r33 remains
constant. This indicates that molecular rotational motion does not have significant
contribution to the EO coefficient, otherwise a roll-off of r33 should be observed over this
frequency range.

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11194
 Fig. 7. EO coefficient r33 under different modulation frequencies and different modulation
voltages. (a) EO coefficient r33 under different modulation frequencies; (b) EO coefficient r33
under different modulation voltages.

Figure 7(b) shows the results of the r33 measured at different modulation voltages. AC
modulation voltages of 1 kHz modulation frequency were applied across a 20 m slit, and r33
values were measured at five different AC amplitude voltages of 1 V, 2 V, 3 V, 4 V and 5 V,
respectively. The detected modulation intensity at 1 kHz was found to be linear with applied
field. The fact that the EO modulation is independent of the applied voltage suggests that the
modulation is due to the linear EO effect and the contributions of higher order (e.g., Kerr) EO
effects are negligible.
5. Conclusions
By applying an in-plane electric field using electrodes fabricated on the substrate, BNA
molecules deposited on the substrate through thermal evaporation can be aligned to form a
polycrystalline thin film of small angle domains. Two-slit electrodes pattern on the glass
substrate allows the molecules deposited in the slits be aligned by the electric field, and the
EO coefficient measured by Youngs interferometric method. A linear relationship between
the alignment field and EO coefficient has been observed, similar to the linear relationship
observed in EO polymers. The EO coefficient of BNA thin films fabricated using this
technique is comparable to that of single crystalline BNA. The dispersion of EO coefficient
agrees with the theoretical dispersion of the first order hyperpolarizability. Because no roll-off
of the EO modulation was observed below kilohertz frequencies, the electric field induced
rotational motion of molecules does not have appreciable contribution to the EO coefficient.
Acknowledgment
We would like to express our appreciations to Emily A. Hillenbrand for synthesis and
purification of EHD-1-12 and EHD-1-13 ingredient. We are also grateful to financial support
by NSF Grant Number ECS-0437920, NSF-DMR-0092380, and NSF Center on Materials and
Devices for Information Technology Research (CMDITR), Grant Number DMR-0120967.

#145322 - $15.00 USD Received 4 Apr 2011; revised 10 May 2011; accepted 11 May 2011; published 24 May 2011
(C) 2011 OSA 6 June 2011 / Vol. 19, No. 12 / OPTICS EXPRESS 11195