Open Access
Add to BibSonomy
Abstract
Lead-tetrakis-4(cumylphenoxy)phthalocyanine (PbPc) is dispersed into several common optical polymers at various doping levels. The linear and nonlinear optical properties are measured at 600 nm near the region where the excited state and ground state absorption cross sections are nearly equal. The effective nonlinear refraction coefficients observed at high concentration are as much three orders of magnitude larger than that observed for fused silica.
1. Introduction
Phthalocyanines present a unique template for nonlinear optical (NLO) materials based on excited state structure and response. The ground state absorption origin (known as the Q band) is strongly peaked and appears in the red end of the visible spectrum. Its exact location depends strongly on the central metal. A second strong peak (the B-band) occurs in the ultraviolet. Between these two strong peaks is a region of relative transparency [1] What makes this class of molecules interesting as NLO materials [2–4] is that the excited state absorbs strongly in this same spectral region of relative transparency. Because the excited state absorption cross section exceeds that of the ground state over this region, induced absorption occurs i.e. the molecule absorbs more strongly as intensity is increased. This induced absorption has become known as reverse saturable absorption and arises from the increasing population of the excited state with increasing intensity.
The excited state population also gives rise to changes in the material refractive index [5]. Previous studies [6] have shown that near wavelengths where the excited state and ground state absorption cross sections were very nearly equal, the nonlinear response of the material was dominated (~99%) by excited state refractive processes. Moreover, the magnitude of the effective refractive response was quite large compared to the n2 observed from fused silica. Since the nonlinear absorptive response is relatively small, we decided to re-explore the excited state refractive response in these phthalocyanine doped polymers to evaluate their potential for use in nonlinear refractive devices.
The present work details the preparation of a reverse saturable absorbing phthalocyanine doped into four different host polymers at moderate to high concentration. Since we found few references to the baseline NLO polymers even in comprehensive handbooks and reviews [7–9], our measurements of their n2 are described and compared to that observed in fused silica. The linear absorption of the free-standing doped samples shows how the variation of the polymer affects the spectral maxima and the degree of aggregation of the concentrated molecule. Ultrafast Z-scan measurements at 600nm demonstrate the magnitudes of the nonlinear absorption and refraction observed. Lead-tetrakis-4(cumylphenoxy)phthalocyanine (PbPc) shows an effective n2 up to three orders of magnitude larger than fused silica when doped into polycarbonate.
2. Experimental
Polymer preparation
Neat polymer films were made via cast film extrusion. Films on the order of mm thickness were made by stacking extruded thin films and then consolidating them under elevated temperature and pressure. Doped polymer films were created by a process of solution blending/dispersion of PbPc in the differing polymers followed by conventional melt extrusion film processing. A high weight percent polymer masterbatch of PbPc in a polymer matrix was created [10] through co-dissolution solution blending of the two materials in a common solvent, dichloromethane, followed by precipitation in a non-solvent, methanol to minimize agglomeration.
The high weight percentage PbPc in polymer masterbatch material was diluted through dry blending (pellet mix) with excess polymer in a series of polymer film extrusion trials to produce 6” wide films at a nominal 75 micron thickness. For polycarbonate (Dow Calibre 200-6), film extrusion trials of the PbPc and polycarbonate were completed at 230°C at a processing rate of 2 lb. /hr. The 6” film was cast onto a heated, 80°C, chromed roll and collected via an in-line torque winder. After collection, each film roll was labeled with the target composition of PbPc dye, extrusion date, and sample thickness and placed in moisture resistant plastic bags for storage prior to NLO characterization. Similar processing was utilized for two styrene-acrylonitrile co-polymers, SAN 23 (23% acrylonitrile, Lustran DN50), SAN 29 (29% acrylonitrile, Lustran CN25) and polymethylmethacrylate (PMMA) (Plexiglass V920).
Linear measurements
The film thickness was measured using a Scherr-Tumico micrometer and the linear absorption was measured using a Perkin–Elmer Lambda 1050 spectrophotometer. The film thicknesses were measured at 4 positions surrounding the area to be probed by the Z-scan technique. The thickness result was averaged and the standard deviation of the measurements was recorded. The absorbance was measured against air, the Fresnel loss was measured over the 1550-1600 nm region, averaged and subtracted from the raw result. The absorption coefficient at 600 nm was calculated from the Fresnel subtracted absorbance multiplied by 2.303 and divided by the thickness of the sample. The uncertainty was calculated from the uncertainty of the spectrophotometer and the measured standard deviation of the film thickness.
Z-scan measurements
The nonlinear optical response of the individual films with varying concentration was measured by the Z-scan technique using a tunable femtosecond laser system, a Clark CPA 2001 chirped pulse amplifier whose frequency doubled output pumps a Clark NOPA noncollinear optical parametric amplifier. For these experiments the NOPA was tuned to 600 nm and yielded 3-5 μJ pulses with a 223 fs HW1/e2M. The thin film samples were mounted on a 200 mm translation stage (Aerotech). The samples were translated along the path of the focused laser beam through the region where the laser is focused most tightly. As the sample nears the focal point of the lens, the laser irradiance increases giving rise to both nonlinear refractive and nonlinear absorptive response in the sample. Figure 1 below shows a schematic of the details of the optical setup.
Fig. 1 Schematic of Z-scan setup. Ultrafast pulses are derived from a femtosecond chirped pulse amplifier (CPA) system converted to tunable wavelengths via a noncollinear optical parametric amplifier (NOPA). Intensity is controlled via a waveplate-polarizer combination, and the beam is spatially filtered to ensure M-squared values <1.1 in the sample region. Open aperture (measuring nonlinear absorption α2) and closed aperture (measuring nonlinear refraction n2) data traces are acquired simultaneously. When needed, a mirror drops in to pick off the reference arm to measure energy or perform frequency resolved optical gating (FROG) for pulse characterization.
Each Z-scan is done at constant pulse energy. For each Z position on the translation stage a reference signal, an open aperture signal and a closed aperture signal were measured. 100 shots were averaged for each position. The pulse energy at the sample was controlled by calibrated neutral density (ND) filters. As the pulse energy at the sample is varied the ND filters are moved from before the sample to after the sample to keep the open and closed aperture signals in the same range. The ratios of the open aperture to the reference and the closed aperture to the reference are recorded. For the open aperture signal, note that all of the light is collected. Only absorptive processes can contribute to this signal. When the sample is far from the focal point of the lens, linear processes dominate. Comparison of the signal through the film to that observed with no sample in place yields the linear transmission of the sample. The log10 of this signal compares well to the absorbance measured in the spectrophotometer. To deduce the nonlinear absorption and refraction coefficients the equations found in the classic Sheik-Bahae [11] paper are used. The procedure is to calculate the nonlinear absorption from the open aperture signal and then to use the ratio of the closed aperture to the open aperture signal to find the nonlinear refraction coefficient. Our procedure uses nonlinear least squares to fit the data.
As will be shown below, there is an additional problem with the extruded polymer films that arises because the films are not completely homogeneous: there are typically thickness and concentration variations over the region probed by the Z-scan. As we have shown previously [12], we compensate for these affects by ratioing each energy scan to the scan measured at the lowest pulse energy. At the lowest energy, the irradiance at the sample does not reach a value high enough to activate the nonlinearity. The variations observed in the Z-scan at this lowest energy arise from the linear sample inhomogeneity rather than nonlinear processes. Taking the ratio of the higher energy scan to the lowest energy scan removes the effects of the linear concentration and thickness variations.
Beam waist determination
Realignment of the focusing optics within the NOPA or movement of the spatial filter (see experimental diagram) changes the beam waist that results from the lens that gives rise to the Z-scan experiment. To measure the beam waist and the M2 of the system, we use a procedure described by Johnston [13]. Briefly, razor blades are mounted on the Z-scan translation stage. Linear actuators move the razor blades across the laser beam in both the horizontal and vertical directions. The razor blades are translated along the z direction to find the beam waist. Figure 2 shows data from a typical M2 determination. The lower left data panel with both red and white traces shows the horizontal and vertical razor scans and the erf(x) fits to that data taken at various points along the z axis of the lens. The larger data panel summarizes the razor measurements at multiple z positions which define the focal characteristics and locates the beam waist. The results of the M2 calculations are shown in the panel and demonstrate that the alignment of the system yields high quality beam properties with little astigmatism.
Fig. 2 Output of the M2 routine showing the typical razor blade measurement and fit at one of the positions on the z axis. The large data panel shows the fit to both the horizontal and vertical beam waist minimum. Zr X and Zr Y are the horizontal and vertical Rayleigh ranges in mm, wo X and wo Y are the horizontal and vertical beam waists in µm while Zo X and Zo Y are the locations of the two minima in mm. Msq X and Msq Y and their δs are the M2 results with their error estimates.
The M2 value and the beam waist are used as inputs for calculation of the peak irradiance used in the fitting routine for evaluating the Z-scan experiments.
Pulse width measurements
We use the frequency resolved optical gating (FROG) [14] technique, based on second harmonic generation to determine the laser pulse width. Figure 3 shows the output of FROG routine. The two color panels on the left shows the second harmonic wavelength on the vertical axis and the time on the horizontal. The upper panel shows the measured FROG trace while the lower one shows the results after fitting. The results panel on the right verified that the HW1/e2M is 253 fs (HW1/e2M = FWHM/1.177).
Fig. 3 The upper colored panel shows the measured second harmonic signal of the 600 nm laser pulse. Wavelength is on the horizontal axis while time is plotted on the vertical axis. The results panel shows the measured pulse width and bandwidth products.
Figure 4 shows the measured and retrieved pulse width and the spectral shape. Both the pulse width and the spectral shape of the second harmonic signal are well fit by the FROG algorithm.
Fig. 4 The left panel shows the measurement of the pulse width and its fit while the right panel demonstrates the observed and predicted wavelength response from the 600 nm laser pulse.
3. Results
Fused silica and undoped polymers
Figure 5 shows the results of the Z-scan measurement at 600 nm for fused silica. No nonlinear absorption is evidenced in open aperture arm of the Z-scan as shown in the left panel of the figure. The data in the right hand panel are well fit by the Sheik-Bahae expressions and yield a value of 1.4x10−7 cm2/GW. Multiple measurements of several different thickness of fused silica yield an n2 = 1.7 ± 0.3x10−7 cm2/GW.
Fig. 5 Observed Z-scan response from a 2.07 mm thick fused silica window at 600 nm using a peak irradiance of 250 GW/cm2. a) is the open aperture signal and fit while b) shows the closed aperture response and fit.
Despite the consolidation and polishing procedures for the bulk polymer preparation described above each of the four polymers showed significant evidences of linear inhomogeneities, i.e. sample wedging or bulk scattering or scattering from inclusions when translated through the focus of the Z-scan lens. Figures 6(a) and 6(b) shows typical raw data at seven differing intensities from one of the SAN 23 samples. As the sample is translated through the lensfocus even at low intensities there are excursions from the expected Fresnel reflection. This effect is much more apparent for the closed aperture scans. Figure 6(c) and 6(d) show the results of normalizing the higher irradiance scans to the lowest one. The linear reference was taken at approximately 500 times lower input pulse energy than the highest scan. Using it to normalize out the linear inhomogeneities allows extraction of the nonlinear coefficients for the pure polymer.
Fig. 6 600 nm Z-scans from bulk polymers. The multiple traces in each panel are taken at differing pulse energies that are roughly a factor of 2 higher than the previous scan. Figure 6a) and 6b) are the raw transmissions observed from open aperture and closed aperture Z-scans respectively. Figure 6c) and 6d) show the four highest intensity scans normalized to the lowest energy Z-Scan. Pulse energy increases by 2x from green to black to blue to red.
Note that there is little evidence of nonlinear absorption in this sample. This was true for all of the polymers examined in this study as long as the peak intensity was kept below 300 GW/cm2. Table 1 shows the results of the measurements of n2 at 600 nm for each of the polymers tested here. The magnitudes are all similar to that observed for fused silica with larger values observed for those that contain aromatic monomers. We are not aware of any previous study that has evaluated the nonlinear refractive index of these bulk polymers.
Table 1. Results of the measurements of n2 at 600 nm for each of the polymers tested
Lead Phthalocyanine doped polymers
Linear absorption
Figure 7 shows the relative extinction spectra of varying weight percentages of PbPc in the four polymeric hosts. The spectra were taken relative to air and the Fresnel reflection was subtracted for each. Most of the individual spectra were normalized to have the same extinction at 410 nm where there is little contribution from aggregated species. Absorbance at 450 nm was used for the PC spectra that were saturated at 410 nm. For ease of comparison the spectra were scaled to the peak of the lowest concentration spectrum. The common features in all of the spectra are that there are prominent Q-band peaks near 720 nm and there is a window of relative transparency that extends from 440 to 620 nm. For several of the concentrations and film thicknesses studied here the absorption from the Q-band peak went off scale, so the flat lines shown in the spectra result from setting those values at the instrumental maximum.
Fig. 7 Extinction spectra from varying concentrations of PbPc in: a) polycarbonate, b) SAN 29, c) SAN 23, and d) PMMA. The flat portions of the spectra result from instrument saturation as described in the text.
The changes in the spectra observed as PbPc concentration is varied arise largely from aggregation effects [15,16]. First, the magnitude of the Q band peak diminishes with increasing concentration. The maximum extinction coefficient for the Q-band is only seen in dilute samples where the monomer is dominant. Second, the vibronic peak near 650 nm becomes less well defined as the concentration increases. This is most apparent in PMMA. Third, in each of the differing polymers the spectral region to the red of the Q-band shows evidence of increasing dimer formation with increasing concentration. The red shift in the spectrum arises from a J-type aggregate and is somewhat unusual for porphyrins or phthalocyanines [16]. The diminution of the Q-Band and the growth of the red shifted peak is most apparent in PMMA indicating that the monomer-dimer equilibrium favors the dimer in this polymer. It should be noted that the highest concentration reached (i.e. the greatest solubility) in these studies was attained in polycarbonate which also appears to most strongly favor monomeric species over aggregated ones. Finally the Q-band maximum shifts with polymer identity. The maximum red shift is observed in PC followed by SAN 23, SAN 29 and PMMA roughly following the aromatic character of the polymer. All of these observations impact the observed nonlinear response from these samples.
Z-scan measurements
In contrast to the pure polymers, Z-scan measurements of the PbPc doped polymers at 600 nm show significant nonlinear absorption. Because the samples are absorbing, the nonlinear response arises from excited state processes. We have demonstrated [6] that the excited states are produced within the pulse width of the laser. Similar to Fig. 6, Fig. 8 shows the raw and normalized Z-scan data for the 3.1% PbPc/PC film. For the raw scans shown at the top of the figure, the z variation due to linear inhomogeneities at the lowest energy is notable in both cases but is quite large for the closed aperture scans. For some of the films examined here, this large variation due to sample homogeneity made extraction of the nonlinear refractive index impossible. However, for this film the normalizationprocedure worked very well and yielded the open aperture and closed aperture/open aperture shown below.
Fig. 8 Raw and normalized data similar to Fig. 6. Raw transmission data are shown on the top two graphs while the normalized data is shown on the lower half. Intensity increases in the following order; green, black, blue and red.
Figure 9 shows the nonlinear least squares fits to this data and are typical of the data for which fits were possible. For almost all of the open aperture data the quality of the fits was quite high. This was not true for some of the refractive data fits. While there were clearly intensity dependent excursions from the linear transmission near the lens focal point, the data were too scattered to extract a reliable number for the effective nonlinear refractive index. The measured values for the nonlinear absorption coefficient and the nonlinear refraction coefficient as well as thelinear absorption coefficient and film thickness from the doped polymer samples are gathered in Table 2. The entries labeled N/A are the instances for which extraction of meaningful nonlinear values was not possible. For the 6.1% PbPc in PC film the combination of the film thickness and concentration made the sample too absorbing to observe signal in either of the open aperture or closed aperture arms. Error estimates are based on the observed standard deviations over the differing intensity scans. In some instances, there were insufficient data to estimate the error.
Fig. 9 Nonlinear least squares fits to the Z-scan data for the 3.1% PbPc in PC sample. The peak intensity of the 600 nm pulse was 22GW/cm2. The upper panel displays the open aperture data and its fit while the lower pane is the fit to the closed aperture divided by the open.
Table 2. Optical data from PbPc(PC)4 in polymers
The data in the table show that the nonlinear absorption and refraction scale linearly as expected with the weight percent in each of the polymers. The nonlinear absorption coefficients are of the same order of magnitude at a similar weight percent for all of the samples measured and are relatively small values. As mentioned above this is expected since at 600 nm the excited state absorption cross section is nearly equal to that of the ground state. In each of the samples the effective nonlinear refractive index is one to two orders of magnitude larger than that observed from the pure polymers. The polymer itself contributes very little to the observed nonlinear response.
The largest effective nonlinear refractive index was observed for PbPc in polycarbonate and is more than three orders of magnitude larger than that observed in fused silica. Had thinner samples at the highest concentration been available it is expected that even larger values could have been be obtained. As mentioned above the samples in polycarbonate showed the least evidence of aggregation. This may be responsible for the significantly higher effective n2 values observed in the polycarbonate sample since aggregated species generally have shorter lifetimes due to the higher probability of intensity dependent exciton-exciton annihilation.
While the magnitudes of the effective n2 values are large they do not directly lead to figures of merit appropriate for waveguide based optical switching. Quick calculation of the Stegeman T and W factors [17] show that even in the best case the linear and nonlinear absorption prevent a long enough interaction length for effective switching. Nevertheless the large magnitudes of the effective nonlinear refraction can lead to potentially interesting NLO devices.
4. Conclusions
We have measured nonlinear absorption and refraction of PbPc in several different polymers. The polymer makes a large difference in achievable nonlinear refraction with polycarbonate being the most suitable. Most simply this is due to the high solubility of this molecule in this polymer. The smaller degree of aggregation probably contributes to the larger response observed in this polymer in that intensity dependent excited state lifetime shortening processes are less likely in non-aggregated species.
Funding
PolymerPlus, LLC in part through the Air Force Small Business Innovative Research/Small Business Technology Transfer (SBIR/STTR) award contract FA8650-15-C-5071.
References
1. H. Isago, “Prototypical optical absorption spectra of phthalocyanines and their theoretical background,” in Optical Spectra of Phthalocyanines and Related Compounds (National Institute for Materials Science, Springer Japan, 2015).
2. J. S. Shirk, R. G. S. Pong, F. J. Bartoli, and A. W. Snow, “Optical limiter using a lead phthalocyanine,” Appl. Phys. Lett. 63(14), 1880–1882 (1993). [CrossRef]
3. J. S. Shirk, R. G. S. Pong, S. R. Flom, H. Heckmann, and M. Hanack, “Effect of axial substitution on the optical limiting properties of indium phthalocyanines,” J. Phys. Chem. A 104(7), 1438–1449 (2000). [CrossRef]
4. J. P. Fitzgerald, P. D. Huffman, I. A. Brenner, J. J. Wathen, G. Beadie, R. G. S. Pong, J. S. Shirk, and S. R. Flom, “Synthesis, chemical characterization and nonlinear optical properties of thallium(III) phthalocyanine halide complexes,” Opt. Mater. Express 5(7), 1560–1578 (2015). [CrossRef]
5. T. H. Wei, D. J. Hagan, M. J. Sence, E. W. Vanstryland, J. W. Perry, and D. R. Coulter, ““Direct measurements of nonlinear absorption and refraction in solutions of phthalocyanines,” Appl. Phys. B. 54, 46–51 (1992).
6. S. R. Flom, R. G. S. Pong, S. R. Carlo, and J. S. Shirk, “Highly nonlinear polymers: Fabrication and optical properties,” in OSA Trends in Optics and Photonics Series, 2003), 293–295.
7. R. L. Sutherland, Handbook of Nonlinear Optics, 2nd ed. (Marcel Dekker, Inc., NewYork, 2003).
8. W. M. K. P. Wijekoon, K.-S. Lee, and P. N. Prasad, “Nonlinear optical properties of polymers,” in Physical Properties of Polymers Handbook, J. E. Mark, ed. (Springer, New York, 2007), 795–822.
9. D. Dini, M. J. F. Calvete, and M. Hanack, “Nonlinear Optical Materials for the Smart Filtering of Optical Radiation,” Chem. Rev. 116(22), 13043–13233 (2016). [CrossRef] [PubMed]
10. S. Carlo, A. W. Snow, R. G. S. Pong, J. S. Shirk, and S. R. Flom, “Fabricating polymers for optical devices,” U.S. Patent #8,003,713 (2011).
11. M. 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). [CrossRef]
12. S. R. Flom, G. Beadie, S. S. Bayya, B. Shaw, and J. M. Auxier, “Ultrafast Z-scan measurements of nonlinear optical constants of window materials at 772, 1030, and 1550 nm,” Appl. Opt. 54(31), F123–F128 (2015). [CrossRef] [PubMed]
13. T. F. Johnston, “Beam propagation (M2) measurement made as easy as it gets: the four-cuts method,” Appl. Opt. 37(21), 4840–4850 (1998). [CrossRef] [PubMed]
14. R. Trebino, Frequency-Resolved Optical Gating: The Measurement of Ultrashort Laser Pulses (Kluwer Academic Publishers, Norwell, MA, 2000).
15. A. Ranade, A. Hiltner, E. Baer, J. S. Shirk, and R. S. Lepkowicz, “Aggregation of lead phthalocyanine in blends with polycarbonate,” J. Appl. Polym. Sci. 104(1), 464–469 (2007). [CrossRef]
16. A. W. Snow, “Phthalocyanine Aggregation,” in Porphyrin Handbook, Vol 17 Phthalocyanines: Properties and Materials, K. M. Kadish, K. W. Smith, and R. Guilard, eds. (Academic Press, San Diego, CA, 2003), pp. 129–176.
17. G. I. Stegeman, “Material figures of merit and implications for all-optical waveguide switching,” Proc. SPIE 1852, 75–89 (1993). [CrossRef]
