Photonics is a young field addressed only 40 years ago by Pierre Aigrain, and bring together a number of disciplines focused on mastering the photon. Optics, electrical engineering, nanotechnology, physics, material science and chemistry are involved and combined in Photonics [1]. The demand of our society for communication services remains to be insatiable and the major highways of communication and information flow are the optical ones. This is making photonics one of the most important key technologies for markets in the current century [1, 2]. For some particular application such as optical signal transmission, tunable light sources and light amplification through parametric processes, highly efficient frequency doubling noncentrosymmetric materials are required. Moreover, as most of applications are planned in integrated optical configuration, these materials must be processed in good optical quality thin films. Conjugated molecular systems have been considered as promising candidates for nonlinear photonic devices. The reason lies in the origin of nonlinear effects, which derives from the delocalized π-electrons cloud characteristic of conjugated systems [3,4]. Highly charged correlated π-electrons relax nonresonant optical excitations with little or no lattice phonon coupling i.e. with a mere electronic response in contrast with inorganic materials where the origin of optical susceptibilities has been found in displacement or rearrangement of nuclear coordinates. In addition, organic systems may be synthesized almost “ad libitum” with desired multifunctionalities, such as mechanical properties, thermal stability, specific wetting ability and ambient chemical resistance, while keeping or enhancing the electronic interactions responsible for the nonlinear optical effect.

Since the last 30 years the most studied organic systems for second harmonic generation (SHG) were the so-called 1-D charge transfer systems where donor (D) and acceptor (A) groups endcapp a π-moiety, schematically D-π-A. In spite of the large hyperpolarizabilty β measured and calculated for many of the D-π-A systems at the molecular level, there is a lack or vanishing of SHG when the molecules pack in the condensed phase. The main reason is related to antiparallel alignment of the dipoles giving rise to a centrosymmetric structure and consequently the disappearance of SHG. In order to overcome this problem the D-π-A active system is dispersed in a suitable polymer matrix or grafted to a polymer backbone and then poled via DC electric field or optically. The result is a noncentrosymmetric thin film. Nevertheless, this approach is adding a relevant technological step for device fabrication and it has been found that time and thermal stability are still unsatisfactory for real applications. Moreover, D-p-A systems possess a charge transfer state, which usually is a low lying state limiting the optical transparency.

Several techniques were developed in order to obtain noncentrosymmetric thin films, such as: i) Langmuir-Blodgett (LB) films [5, 6]; ii) Isotropic polymers [7]; iii) Liquid crystalline polymers [8]; iv) Epitaxial or quasi-epitaxial growth [9]. None of the above reported technique was satisfactory for exploitable real device applications. LB because of poorness of optical quality and lack of large area processing, isotropic and liquid crystalline polymers because of the relaxation of induced polar order, epitaxyal growth because of severe limitation to specific substrates. Very recently, Facchetti et al. [10] reported a very elegant and successful fabrication of second order NLO films by growth-templating self assembled monolayers and vapour deposited chromophore layer driven by hydrogen bonding. The reported χ ⁽²⁾(-2ω;ω,ω) values are up to 94 pm/V at 1064 nm fundamental wavelength. Finally, Kajzar et al. [11] have reported fabrication of noncentrosymmetric thin film by high vacuum deposition of very thin electron donating and electron accepting molecules, intercalated with a neutral layer.

Ideally we would like to have a noncentrosymmetric-conjugated system with high optical susceptibility, being able to avoid the building of symmetry center in the condensed phase and easy to process as an optically transparent thin film. We have focused our effort on this goal and we have found an original solution by chemical tailoring of a new family of thiophene-based molecular systems. Instead of conventional linear structure we have considered a branched V-shaped benzo[b]thiophene-based structure. Is worth noting how the concept of using Λ-shaped molecules (equivalent in concept to our V-Shaped molecules) in second-order NLO was introduced since the 80’s by several groups optimizing the tradeoff between nonlinearity and transparency [12], pointing out exploitation of chirality’s for SHG [13,14], or focusing the attention to obtain noncentrosymmetric single crystals with optimized phase matching [15]. There are no SHG measurements in thin film on such V-shaped molecular systems reported in literature at our knowledge.

The two molecular systems considered in this paper are the 2,3-Bis-(2,2’bithiophen-5-yl)-benzo[b]thiophene 1,1 -dioxide (V-BT1) and the 2,3-Bis-(3,3’-dicyclohexyl-[2,2’]bithiophenyl-5-yl)-benzo[b]thiophene 1,1-dioxide (V-BT2) and the respective molecular structures are reported in Fig. 1.

 

Fig. 1. Left: Molecular structure of 2,3-Bis-(2,2’bithiophen-5-yl)-benzo[b]thiophene1,1-dioxide (V-BT1) and (right) of 2,3-Bis-(3,3’-dicyclohexyl-[2,2’]bithiophenyl-5-yl)-benzo[b]thiophene1,1-dioxide (V-BT2).

Download Full Size | PDF

The synthesis was carried on via the Stille coupling and details are reported in reference [16]. We should remark here how the yields are pretty good being 90% for V-BT1 and 74% for V-BT2. Is clear how both V-BT1 and V-BT2 deviate from the conventional D-π-A systems. First of all there are no D-A groups but the unit SO₂ is introducing charge asymmetry in the electronic distribution making the benzo[b]thiophene surrounding more negative. The branched structure evolves into a 3D asymmetric V-shape deviating from 1-D. The lack of symmetry and intrinsic steric hindrance is even more stressed for V-BT2 where two cyclohexyl units are added. The tailored molecular structures prevent packing or aggregation in centrosymmetric solid-state phase. They are highly soluble in a variety of solvents and thin film can be obtained by spin coating technique [17]. The easy and versatile processing as thin films is emphasized by the possibility to grow thin films of both molecules via physical vapor deposition in high vacuum or ultra high vacuum. We have demonstrated and assessed this last technique for the thin film samples for SHG experiments reported in this paper.

V-BT1 has been deposited on circular ½ inch diameter fused silica substrates at a base pressure of 1.2×10^-6 mbar. The deposition starts at 240° C and the desired film thickness is reached at a deposition rate of 0,4 Å/sec. Similarly for the V-BT2, where the sublimation temperature starts at 280° C. The one-photon absorption spectra for 80nm thick thin films of both systems are reported in Fig. 2.

 

Fig. 2. One-photon absorption spectra of 2,3-Bis-(2,2’bithiophen-5-yl)-benzo[b]thiophene1,1-dioxide (V-BT1, red-circles) and of 2,3-Bis-(3,3’-dicyclohexyl-[2,2’]bithiophenyl-5-yl)-benzo[b]thiophene1,1-dioxide (V-BT2, blue-line) thin films. The films are vacuum deposited and are 80nm thick. The vertical line marks the expected SHG signal.

Download Full Size | PDF

The lowest excited electronic system of V-BT2 is peaked at higher energy with respect to V-BT1. This is consistent with the enhanced distortion of the conjugation imposed by the cyclohexyl groups. Both films are of very good optical quality and transparent at 532nm, where the generated SHG is expected in our experimental conditions. In order to characterize the second order NLO properties of thin films we have performed optical SHG measurements at the fundamental wavelength of 1064 nm. The measured value is off resonance, as the absorption edge is at about 500 nm (see Fig. 2). The laser source is a Q switched Nd : YAG laser operating at 1064 nm fundamental wavelength with 13 ns pulse duration and 10 Hz operation rate. The laser beam condition during the measurements were of 2mJ in energy namely 2.2 GW/cm² of light intensity. The sample was mounted on a goniometer in vacuum between two polarizers and rotated along an axis perpendicular to the beam propagation direction. A half wave plate was used to change polarization of incident beam in a monotonic way. An analyser located behind the sample allows checking the polarization of the harmonic beam. By using different polarization of incident and harmonic beams, different tensor components can be measured. The SHG intensities were calibrated by SHG measurements on sp and pp fundamental–harmonic beam configurations, i.e. the beam polarization (fundamental or harmonic) perpendicular (s) or parallel (p) to the incidence plane. First of all we have checked the homogeneity of the deposited films. SHG intensity has been measured when translating the films in two perpendicular directions over 8000μm. A constant SHG intensity has been observed, showing indeed excellent thin film thickness homogeneity.

The measured χ ⁽²⁾xxz(-2ω;ω,ω) and χ ⁽²⁾zzz(-2ω;ω,ω) for both systems are summarized in table 1.

Table 1. Second order NLO susceptibilities of studied molecules thin films, made by vacuum sublimation technique.

View Table

The values of LiNbO₃ single crystal, considered a benchmark for standard, are reported for comparison. The fitting procedure and data analysis is detailed in the appendix note at the end of the text. It is worth noting how for both systems processed in thin films the SHG values are comparable to LiNbO₃ single crystal, the V-BT2 being almost twice the V-BT1. In order to analyse and model the active tensor components responsible for the SHG signal, we have performed incidence angle dependence of SHG intensity at different fundamental-harmonic beam polarization configuration. The results obtained for V-BT2 are reported in Fig. 3.

 

Fig. 3. Incidence angle dependence of SHG intensity for V-BT2 thin film at different fundamental harmonic beam polarization configurations.

Download Full Size | PDF

As we can see, the major contribution is measured for the pp configuration being 3 times larger than the sp one. This may be interpreted in terms of a preferential orientation of the optical susceptibility along the perpendicular to the substrate surface, suggesting a preferential optically active orientation of the molecules in the thin film, in some agreement with D-π-A poled systems. Nevertheless, we observe a weak but measurable value for both the XXZ and ZZX components. This is an indication that we are not in the Kleinman conditions, i.e. point symmetry ∞mm and only two nonzero tensor components. We can interpret this experimental result considering that our V-shaped conjugated systems are not simple one-dimensional polarizable systems. We should also comment on a non-standard effect observed for V-BT2 measured at very thin thickness. In Fig. 4 we show the measured χ ⁽²⁾ XXZ and χ ⁽²⁾ZZZ for V-BT2 films at different thickness.

 

Fig. 4. The variation of χ ⁽²⁾xxz () and χ ⁽²⁾zzz () versus the films thickness for films of 2,3-Bis-(3,3’-dicyclohexyl-[2,2’]bithiophenyl-5-yl)-benzo[b]thiophene1,1-dioxide (V-BT2). The data are affected by a 10% error bar.

Download Full Size | PDF

The value obtained for the 37nm thick film, more than twice the “bulk” value, is not an artifact but a reproducible value. Even if we cannot exclude surface/interface SHG contribution to SHG, particularly for the thinnest layer, the constancy of χ ⁽²⁾ susceptibilities for thicker layers means that this is negligible compared to the molecular response. The observed larger χ ⁽²⁾ for thinnest layer may relate to a better ordering of the investigated molecule in the early stage of growth with respect to thickness-induced disorder. In order to clarify this point, a study of correlation between morphological and structural properties with SHG measurements is in progress. . Finally, is worth noting how for practical applications in waveguiding configuration devices, the value of the diagonal χ ⁽²⁾ _ZZZ tensor component is of utmost importance. In conclusion, we have designed an original approach for efficient SHG of tailored branched benzo[b]thiophene systems enabling versatile and flexible one-step thin film processing matching the constrains of integrated optical configuration for nonlinear devices. The designed moieties possess non conventional optical susceptibility and show χ ⁽²⁾ values at least as high as standard LiNbO₃ single crystal. This may open the way to a new class of organic materials exploitable for photonic applications.

Appendix note

In the following we report on the procedure used for analysis and fitting of the measured experimental data. A first attempt was made assuming point symmetry 8mm, since the measured data suggest a possible in plane isotropy of crystallites and an anisotropy direction perpendicular to the surface plane. For such symmetry, owing to the Kleinman’s conditions there are 2 nonzero χ ⁽²⁾ tensor components: χ ⁽²⁾ _XXZ and χ ⁽²⁾ _ZZZ (cf. Fig. 5).

 

Fig. 5. Reference frame used for studied thin films.

Download Full Size | PDF

For the s – fundamental and p – harmonic beam polarizations the SHG intensity is roughly given by [18–19]

(1)$$I_{2\omega }^{s-p}\propto \mid P_{\mathrm{sp}}(\omega ,2\omega ,{\theta }_{\omega }^s,{\theta }_{\omega }^s,{\chi }_{\mathrm{XXZ}}^{\left(2\right)}){\mid }^2{I}_{\omega }^2$$

where

(2)$$P_{\mathrm{sp}}(\omega ,2\omega {,\theta }_{\omega }^s,{\theta }_{\omega }^s,{\chi }_{\mathrm{sp}}^{\left(2\right)})={\chi }_{\mathrm{XXZ}}^{\left(2\right)}\sin 2{\theta }_{\omega }^s\cos {\theta }_{2\omega }^s$$

(3)$$I_{2\omega }^{p-p}\propto \mid P_{\mathrm{pp}}(\omega ,2\omega ,{\theta }_{\omega }^p,{\theta }_{\omega }^p,{\chi }_{\mathrm{XXZ}}^{\left(2\right)},{\chi }_{\mathrm{ZZZ}}^{\left(2\right)}){\mid }^2{I}_{\omega }^2$$

(4)$$P_{\mathrm{pp}}(\omega ,2\omega {,\theta }_{\omega }^s,{\theta }_{\omega }^s,{\chi }_{\mathrm{XX}Z}^{\left(2\right)},{\chi }_{\mathrm{ZZZ}}^{\left(2\right)})={\chi }_{\mathrm{ZZZ}}^{\left(2\right)}{\sin }^2{\theta }_{\omega }^p\sin {\theta }_{2\omega }^p+{\chi }_{\mathrm{XX}Z}^{\left(2\right)}\cos {\theta }_{\omega }^p\left(\cos {\theta }_{\omega }^p+2\sin {\theta }_{\omega }^p\cos {\theta }_{2\omega }^p\right)$$

where ω,2ω are beam frequencies, s and p their polarizations and θ is the propagation angle in nonlinear medium for given frequency and polarization. The proportionality factors in Eqs. (1) and (3) depends on the choice of reference system, transmission factors and on the refractive indices at fundamental and harmonic frequencies [18, 19].Equation 2 shows that from s-p measurements we can obtain directly the χ ⁽²⁾ _XXZ susceptibility what we have done and the obtained values are listed in Table 1. In the case of p-p configuration the situation is more complicated as both χ ⁽²⁾ _XXZ and χ ⁽²⁾ _XXZ susceptibilities intervene in the expression for SHG (cf. Eq. (4)).

We tried to fit the p-p data by injecting directly the previously obtained χ ⁽²⁾ _XXZ values to Eq. (3). However it didn’t give good fits and coherent results due to 2 factors

1. Experimental error in determination of the measured χ ⁽²⁾ _XXZ
2. The fact that Kleinman’s conditions are not fulfilled (nonzero SHG intensities for p-s and s-s configurations, see Fig. 3)

Therefore we made the fits by taking the ratio χ ⁽²⁾ _ZZZ/χ ⁽²⁾ _ZZZ = 3 (only for p-p configuration measurements), what corresponds to moderately poled electro-optic polymers, or in theory to a free gas model (no interacting dipoles). This approach is often called a “self consistent method”. In this case we observe a good agreement between the assumed ratio (of 3) and the derived one from the fitting procedure (within the experimental accuracy close to 3 as it can be seen in table 1). Another approach consisting of fitting of Eq. (3) as function of incidence angle, with the ratio $\frac{{\chi }_{\mathrm{ZZZ}}^{\left(2\right)}}{{\chi }_{\mathrm{XX}Z}^{\left(2\right)}}$ (Eq. (4) as a fit parameter led consistently to higher χ ⁽²⁾ _ZZZ values.

Acknowledgments

We acknowledge for funding the EU Phoenix Project, the Emilia-Romagna Region NetLab Project MIST-ER and the MiUR Projects RBIP0642YL “LUCI” and RBIP06JWBH “NODIS”.

1. Toward a Bright Future for Europe,” www.Photonics21.org.

2. Harnessing Light” study (National Research Council, USA, 1998).

3. J. L. Oudar and D. S. Chemla, “Hyperpolarizabilities of nitroanilines and their relations to the excited state dipole moment,” J. Chem. Phys 66, 2664–2667 (1977). [CrossRef]  

4. David J. Williams editor, ACS symposium series 233, ACS, Washington D.C (1983).

5. M. C. Petty, Langmuir-Blodgett Films, An Introduction, Cambridge University Press, Cambridge1996.

6. Ch. Bosshard, M. Küpfer, M. Flörsheimer, Th. Borer, P. Günter, Q. Tang, and S. Zahir, “Investigation of chromophore orientation of 2-docosylamino-5-nitropyridine and derivatives by nonlinear optical techniques,” Thin Solid Films 210/211, 198–201 (1992)] [CrossRef]  

7. L. R. Dalton, “Nonlinear Optical Polymeric Materials,” in Advances in Polymer Science 158, Springer-Verlag, Heidelberg, 2001, p. 1–86.

8. D. Gonin and C. Noël, Liquid Crystalline Polymers, in Organic Thin Films for Waveguiding Nonlinear Optics, F. Kajzar and J. Swalen Eds, Gordon & Breach Sc. Publ., Amsterdam1996, pp.221–288.

9. J. Le Moigne, in Organic Thin Films for Waveguiding Nonlinear Optics F. Kajzar and J. Swalen Eds, Gordon & Breach Sc. Publ., Amsterdam1996, pp. 289–327.

10. A. Facchetti, E. Annoni, L. Beverina, L. Beverina, M. Morone, P. Zhu, T. J. Marks, and G. A. Pagani, ”Very large electro-optic responses in H-bonded heteroaromatic films grown by physical vapour deposition”, Nat. Mater. 3, 910–917 (2004). [CrossRef]   [PubMed]  

11. F. Kajzar, Y. Okada-Shudo, C. Meritt, and Z. Kafafi, “Second- and third-order non-linear optical properties of multilayered structures and composites of C60 with electron donors,” Synth. Met. 117, 189–193 (2001). [CrossRef]  

12. C.R Moylan, S Ermer, S.M. Lovejoy, I.-H. McComb, D.S. Leung, R. Wortmann, P. Krdmer, and R.J. Twieg, “Dicyanomethylene)pyran derivatives with C2v symmetry: An unusual class of nonlinear optical chromophores.” JACS 118, 12950–12955 (1996) [CrossRef]  

13. Victor Ostroverkhov, Kenneth D. Singer, and Rolfe G. Petschek, “Second-order nonlinear optical processes in non-polar chiral materials: relationship between molecular and macroscopic properties,” J. Opt. Soc. Am B. 18, 1858–1865 (2001). [CrossRef]  

14. V. Ostroverkhov, R.G. Petschek, K.D. Singer, and R.J. Twieg, “Lambda-Shaped Chromophores for Chiral Nonlinear Optical Materials,” Chem. Phys. Lett. 340, 109–115 (2001). [CrossRef]  

15. T. Watanabe, H. Yamammoto, T. Hosomin, and S. Miyata, in Organic Molecules for Nonlinear Optics and Photonics, J. Messier, F. Kajzar, and P. Prasad, Eds, Kluwer Acad. Publ., Dordrecht (1991).

16. Giovanna Barbarella, Laura Favaretto, Alberto Zanelli, Giuseppe Gigli, Marco Mazzeo, Marco Anni, and Alessandro Bongini, “V-shaped thiophene based oligomers with improved electroluminescence properties,” Adv. Funct. Mater. 15, 664–670 (2005). [CrossRef]  

17. Marco Mazzeo, Vincenzo Vitale, Fabio Della Sala, Dario Pisignano, Marco Anni, Giovanna Barbarella, Laura Favaretto, Alberto Zanelli, Roberto Cingolani, and Giuseppe Gigli, “New branched thiophene-based oligomers for bright organic light emitting devices,” Adv. Mater. 15, 2060–2063 (2003). [CrossRef]  

18. K. D. Singer, J. E. Sohn, and S. J. Lalama, Second harmonic generation in poled polymer films, Appl. Phys. Lett. 49, 248 (1986) [CrossRef]  

19. J. Swalen and F. Kajzar, Introduction, in Organic Thin Films for Waveguiding Nonlinear Optics, F. Kajzar and J. Swalen Eds, Gordon & Breach Sc. Publ., Amsterdam1996, pp. 1–85