Preparation and Optical Limiting Properties of Polyurethane Containing Long Conjugated Chromophores

Two functional polyurethanes (P1 and P2) bearing a large π electron conjugated chromophoric pendant were synthesized and characterized by FT‐IR, 1H‐NMR and UV‐Vis absorption spectra. Their optical limiting properties were evaluated. The results show that P1 and P2 show novel optical limiting properties, which are assigned to a long π electron conjugated chromophoric pendant. It was found that their optical limiting properties were affected simultaneously by solution concentration and P2 displays a better optical limiting property than P1 at the same solution transmittance, although that P1 has larger χ(3) (4.28×10−11 esu) than P2 (0.87×10−11 esu), and their optical limiting mechanism is investigated.


Introduction
With the development of laser equipment and technology, the research on materials and devices for protection of optically sensitive devices and human eyes from laser damage has drawn much interest in recent years (1). Among the organic materials, some NLO polymers are considered to be promising materials, mainly because they offer many advantages such as good optical limiting properties, a high optical damage threshold, fast time response, and good processability to form optical devices (2)(3)(4)(5)(6).
To enhance the application viability of the polymer materials in optical limiter, it is necessary to improve their optical limiting properties, as well as thermal stability. Polyurethanes exhibiting excellent mechanical and thermal properties have drawn both scientific and industrial attention. Many polyurethanes containing different chromophore groups were synthesized and their nonlinear optical properties were investigated (7)(8)(9)(10)(11)(12)(13)(14)(15). However, little attention has been devoted to optical limiting properties (16). In previous work (17), we investigated the influence of different electron conjugation bridge structures of the polyurethanes on their optical properties. In this work, to further understand the relationship between structure and optical properties of polyurethanes, we synthesized two substituted polyurethanes. They contain the same electron conjugation bridge structures, but a different acceptor in long conjugated NLO-chromophores (Scheme 1) and the effect of molecular structures with different donor/acceptor groups on the optical limiting properties is investigated.

Testing
The FT-IR spectra were recorded as KBr pellets on a Nicolet 170sx spectrometer. 1 H-NMR spectra were collected on an AVANCE/DMX-300 MHz Bruker NMR spectrometer. UV-Vis spectra were recorded on a Shimadzu UV-265 spectrometer using a 1 cm 2 quartz cell. Molecular weights of the polymers were estimated on a KNAVER Vapor Pressure Osmometer. Thermal analyses of the polymer were performed on a Perkin-Elmer TGA (Thermogravimetric analysis) and a Perkin-Elmer DSC (differential scanning calorimeter) under nitrogen at a heating rate of 208C/min and 108C/min, respectively.
The investigation of the optical limiting properties of the samples was carried out by using a frequency doubled, Q-switched, mode-locked Continuum ns/ps Nd:YAG laser, which provides linearly polarized 8 ns optical pulses at 532 nm wavelength with a repetition of 1 Hz. The experimental arrangement is similar with that in the literature (18). The samples were housed in quartz cells with a path of 5 mm. The input laser pulses adjusted by an attenuator (Newport) were split into two beams. One was employed as a reference to monitor the incident laser energy, and the other was focused onto the sample cell by using a lens with a 300 mm focal length. The samples were positioned at the focus. The incident and transmitted laser pulses were monitored by two energy detectors, D 1 and D 2 (Rjp-735 energy probes, Laser Precision).
The nonlinear optical properties of the samples were performed by a Z-scan technique with the same laser system as in the optical limiting experiment with a pulse width of 8 ns at 1 Hz repetition rate and 532 nm wavelength. The experiment was set up as in the literature (19). The solution sample was contained in a 2 mm quartz cell. The input energy was 100 mJ. The radius v at beam waist was 50 mm. The samples were moved along the axis of the incident beam (z direction). The experimental data were collected utilizing a single shot at a rate of 1 pulse/min to avoid the influence of thermal effect. water solution of sodium nitrite (1.36 g, 20 mmol) was added to the above solution and stirred for 30 min. Then, the mixture was added dropwise to a 400 mL aqueous buffer solution of acetic acid-sodium acetate (pH % 6) containing 3.81 g (21 mmol) N,N-dihydroxyethylaniline and stirred for 1 h at 058C. The resulting precipitate was filtered and rinsed with water twice. The crude product was recrystallized from ethanol twice to give orange crystals in 85% yield. FTIR (

Synthesis of trans-4-[4-fp-[(N,Ndihydroxyethyl)amino]
phenylazogstyryl]-Nmethylpyridinium iodide (3a) 9.9 g (0.03 mol) 1 and 7.9 g (0.03 mol) 2 were dissolved in 90 mL absolute ethanol. Five drops of piperidine were added into the solution. This solution was then heated to reflux overnight. After cooling, the solution was filtered, and the solid was recrystallized from ethanol twice to give a deep red solid in 65% yield. FTIR

Synthesis of the Polyurethanes
All the polymerization reactions and manipulations were performed under nitrogen, except for the purification of the polymers, which was conducted in open atmosphere. A typical procedure is given below: In a 100 mL four-neck cylindrical vessel, equipped with a mechanical stirrer, 2.62 g (10 mmol) of MDI was added slowly to a solution of 5.30 g (10 mmol) 3a and 0.1 mL of dibutyltinlaurate in 50 mL of anhydrous N,N-dimethylformamide. The resulting solution was reacted at 808C for 8 h. After cooling to room temperature, the reaction solution was poured dropwise into 250 mL of methanol. The resulting polyurethane was collected by filtration and redissolved in DMF and precipitated into methanol for purification. The dissolution-precipitation process was repeated three times, and the final isolated precipitant was dried under vacuum at 508C.

Structure Characterization of the Polyurethanes
Both polymers were well characterized by standard spectroscopic methods, from which satisfactory analysis data corresponding to their molecular structures were obtained (see Experimental section for details). Figure 1 shows the FTIR spectra of 3a and P1. The characteristic O-H stretching vibration of compound 3a is located at 3381cm 21 , which disappears and the characteristic stretching vibrations n s (N-H) and n s (C5 5O) at 3311 and 1645 cm 21 are found in the spectrum of its polymer. The similar results were also found in the IR spectra of compound 3b and P2, displaying the formation of the urethane group (17).
The excellent solubility of both polymers enables characterization of their molecular structure by a solution spectroscopic method in common solvents. Figure 2 shows the 1 H-NMR spectra of compound 3a and P1 in DMSO-d 6 . As seen in Figure 2, the hydroxyl proton of compound 3a absorbs at d ¼ 4.86 ppm, which disappears in the spectrum of its polymer. A new broad resonance peak assigned to the amine proton absorption at d 9.61 ppm appears upon the spectrum of P1, further displaying the formation of urethane linkage and is consistent with result of FTIR spectra. All the other resonance peaks of the segment of 3a and 4,4 0 -methylenebis(phenyl isocyanate) appear in the spectrum of P1. Similar phenomena are also found in the NMR spectra of compound 3b and P2. The relative ratio of integration of these protons are also consistent with the proposed structures, as shown in Scheme 1, confirming that functional polyurethane polymers are yielded.

Thermal Properties of the Polyurethanes
The thermal behavior of two polymers were investigated by TGA and DSC. All the results are summarized in Table 1. The polymers are thermally stable. Their TGA thermograms are shown in Figure 3. The decomposition temperature (T d ), defined as the temperature of 5% wt loss of P1 is 2948C and T d of P2 is up to 3028C, and the glass transition temperature (T g ) of P1 and P2 are 2138C and 2248C, respectively. The result indicates that both polymers exhibit good thermal stability and high glass transition temperature (9, 10), which may results from a protective 'jacket' formed via the strong electronic interaction among the polarized azobenzene group in the side chain, shielding the polyurethane backbone from thermal attack. The similar phenomena were found in Masuda (21) and our previous work (5, 6). Figure 4 shows the UV-Vis spectra of compound 3a, 3b, along with P1 and P2 in DMF. It can be seen in Figure 4 that compound 3a and 3b, exhibit strong absorption peak at 502 and 465 nm, respectively, which are associated with the p-p Ã transition of the extended p electron conjugated NLO  chromophores. Compared with that of compound 3b, the absorption peak of compound 3a significantly shows redshift, which may result from a larger dipole effect of strong electron accepted group N-methypyridinium iodide in compound 3a. However, after being incorporated into the polymer chain, their polymers display the different degree blue-shift and the maximum absorption wavelength at 488 nm for P1 and 459 nm for P2, indicating the presence of the electronic interaction between the chromophore moieties and the polymer chain (9,10,22). The nonlinear coefficients of polymers were measured by using a Z-scan technique. The concentrations of P1 and P2 solutions used were 0.060 and 0.063 mg/mL, respectively. The results of Z-scan with and without an aperture showed that P1 and P2 have both nonlinear absorption and nonlinear refractive ( Figures 5 and 6). Thus, the x (3) measured in this experiment was attributed to dual contributions of nonlinear absorption coefficient of molecules (a 2 ) and nonlinear refractive index of molecules (n 2 ).

Optical Properties of the Polyurethanes
In theory, the normalized transmittance for the open aperture configuration can be written as (19) (Equation (1)): where q 0 (z) ¼ a 2 I 0 (t)L eff /(1+z 2 /z 0 2 ), a 2 is the nonlinear absorption coefficient, I 0 (t) the intensity of laser beam at focus (z ¼ 0), L eff ¼ [1 2 exp (2a 0 L)]/a 0 is the effective thickness with a 0 the linear absorption coefficient and L the sample thickness, z 0 is the diffraction length of the beam, and z is the sample position. Thus, the nonlinear absorption coefficient of P1 and P2 are determined to be 4.00 Â 10 211 and 8.75 Â 10 212 m/W respectively by fitting the experimental data using Equation (1).
The normalized transmission for the closed aperture Z-scan is given by (19) (Equation (2)): where x ¼ z/z 0 and Df is on-axis phase change caused by the nonlinear refractive index of the sample and D f ¼ 2pI 0 (1 2 e 2a 0 L )n 2 /la 0 . Thus, the nonlinear refractive coefficient of P1 and P2 are determined to be 3.29 Â 10 217 and 6.25 Â 10 218 m 2 /W by fitting the experimental data using Equation (2). The sign of n 2 is determined to be negative, hinting the optical nonlinear refraction is a   self-defocusing process. The x (3) can be calculated by the following Equation (3) (19): where 1 0 is the permittivity of vacuum, c the speed of light, n 0 the refractive index of the medium and v ¼ 2pc/l. The calculation results of the nonlinear susceptibility of P1 and P2 are 4.28 Â 10 211 and 0.87 Â 10 211 esu, respectively. Evidently, the nonlinear susceptibility increases with the acceptor strength. Figure 7 shows the optical limiting behaviors of P1 with concentration of 0.060 mg/mL and P2 with concentration of 0.063 mg/mL at the same linear transmittance (T ¼ 60%) in DMF. As shown in Figure 7, at very low incident fluence, the output fluence of P1, P2 solutions with 60% transmittance linearly increases with the incident fluence obeying the Beer -Lambert law. However, at high incident fluence (limiting threshold, defined as the incident fluence at which transmittance start to deviate from linearity), the transmittance of the solution decreases and a nonlinear relationship is observed between the output and input fluence. With a further increase in the incident fluence, the transmitted fluence reaches a plateau, showing the good optical limiting property. From Figure 7, we can see that P1 shows the limiting threshold at 0.17 J/cm 2 while P2 is at 0.15 J/cm 2 . Thus, different from their nonlinear optical properties, P2 exhibits better optical limiting properties than P1 at the same transmittance although that P1 has larger x (3) than P2, which may be originated from the stronger ground electronic absorption of P1 than that of P2 at 532 nm wavelength. Similar phenomenon was found in our previous work (17). Simultaneously, we measured the UV-Vis absorption spectrum of the P1 and P2 solution before and after the laser irradiation and found that the pattern and intensity of UV-Vis absorption spectrum have almost no change, hinting that both polymers possess good photostability.

Optical Limiting Mechanism of the Polyurethanes
The optical limiting mechanisms of organic compounds are often based on two-photon absorption (TPA) or reverse saturated absorption (RSA). Generally, TPA-based optical limiting effect can be yielded in principle under the laser irradiation of picosecond or shorter pulses. RSA is achieved on a nanosecond or longer time scale, owing to the different excited-state lifetimes involved in a multilevel energy process (23). In this work, the polymers are excited by the laser with 8 ns pulse width at 532 nm wavelength and the transmittance of all these polymers solutions decreases with the increase of the incident fluence. Therefore, we consider that the optical limiting properties of P1 and P2 may mainly arise from RSA. Based on Golovlev's phenomenological model (24), the experimental data can be fitted perfectly using Equation (4): where T 0 is the linear transmittance of the sample, F out is output fluence, F in is input fluence, F nln is the parameter characterizing of the nonlinear absorption of the material and a small magnitude of F nln will ensure better optical limiting performance of the sample. Thus, F nln of P1 and P2 are determined to be 0.68 J/cm 2 and 0.38 J/cm 2 by fitting the experimental data using Equation (4) and P2 has larger F nln than P1, which is consistent with their optical limiting properties. The fitting result indicated that the RSA process is the most likely mechanism for the optical limiting behavior of P1 and P2. Figure 8 shows the optical limiting behaviors of P1 with different concentration. It can be found that the limiting effect was affected by concentration, with higher concentration solutions exhibiting better performances. For example, the limiting threshold of P1 solution decreases from 0.17 to 0.15 mJ/cm 2 when linear transmittance decreases from 60% (c ¼ 0.060 mg/mL) to 40% (c ¼ 0.156 mg/mL). On the contrary, the threshold increased from 0.17 to 0.29 mJ/cm 2 when linear transmittance was increased to 75% (c ¼ 0.036 mg/mL). Similar results were also found by Kojima Y and our previous publications (5,25). It is the reason that the solution with a higher concentration has more molecules per unit volume, which should absorb the energy of the harsh laser more efficiently.

Conclusions
We have successfully synthesized two novel functional polyurethanes bearing long D-p-A conjugated chromophores in high yield. The both polymers show novel optical limiting properties. Their optical limiting properties are mainly originated from reverse saturated absorption (RSA). Different from their third order nonlinear optical properties, P2 with weak acceptor group displays better optical limiting property than P1 with stronger acceptor group at the same transmittance. Simultaneously, it is found that both polymers exhibit good thermal stability, which may result from a protective 'jacket' formed via the strong electronic interaction among the polarized azobenzene group in the side chain, shielding the polyurethane backbone from thermal attack.