The synthesis, chemical characterization and nonlinear optical properties of a family of soluble thallium(III) phthalocyanine(Pc) halide complexes are reported. Tl(t-Bu4Pc)X (X = Cl, Br, or I) were made from 4-t-butylphthalonitrile via oxidation of a dithallium(I) phthalocyanine complex and were characterized by elemental analysis, spectroscopy (UV-vis, 1H NMR and X-ray fluorescence) and MALDI-TOF mass spectrometry. Transient white light absorption measurements of the iodo complex in toluene showed a broad band (440 – 610 nm) excited state absorption with a multiexponential decay dominated by species with lifetimes of ~4.8 ± 1 ps and > 20 ns. The ratio of the long lived excited state absorption coefficient to that of the ground state is ~100 at 500 nm. Nonlinear transmission and Z-scan experiments (532 nm, f5 optics) confirmed the above and also showed that the real part of the nonlinear refractive index is negative.
© 2015 Optical Society of America
Materials which are highly transmitting at low intensity light levels but much less transmitting for high intensity, have important applications in protecting sensitive optical elements from laser radiation . An effective mechanism to achieve this result is sequential two-photon absorption, where a first photon excites a state that strongly absorbs a second photon, a process also known as reverse saturable absorption. When the excited state absorption cross section is greater than that of the ground state, the absorbance of the material will increase as the excited state becomes populated. In addition to a large excited state absorption cross section, effective optical limiters must also have low ground state absorbance, a fast excited state rise time and long-lived excited states.
Metallophthalocyanines and their analogues have been extensively investigated as optical limiters [2–5 ]. Following reports [6–9 ] of enhanced reverse saturable absorption in lead and indium phthalocyanines there has been a focus on heavy atom containing macrocycles. The increased nonlinear absorption in these materials is due to an increase in the triplet excited state population as a result of faster intersystem crossing mediated by the heavy atom spin-orbit coupling [10,11 ]. Thallium, a heavy metal adjacent to both indium and lead in the periodic table, would appear to be especially promising as an optical limiter when bound by a phthalocyanine, yet there are only a few reports of these [12,13 ]. We report here the synthesis, chemical characterization, photophysical behavior, and nonlinear optical properties of a series of soluble Tl(t-Bu4Pc)X complexes.
General considerations. All reagents were purchased commercially and used without further purification except for the toluene and chloroform used in the glove box, which were distilled from sodium/benzophenone ketyl and calcium hydride, respectively, prior to use. Manipulations involving air or water sensitive compounds were done inside a Braun MB150 inert atmosphere glove box maintained at less than 1 ppm oxygen and water. Samples were submitted to Galbraith Laboratories, Knoxville, TN, for elemental analysis. UV-visible spectra were obtained on a Agilent 8453 diode array spectrometer or on a Perkin-Elmer Lambda 9 spectrophotometer, IR spectra were recorded on an IR100 spectrometer from Thermo Electron Corporation, 1H NMR spectra were recorded on a JEOL ECX-400 NMR spectrometer (chemical shift values are reported relative to residual protons in the solvent as an internal standard). Mass spectra were recorded on an Applied Biosystems Voyager-DE MALDI-TOF mass spectrometer and X-ray fluorescence spectra were obtained on a Shimadzu EDX-700HS Energy Dispersive X-ray Spectrometer. Setups for the white light transient spectroscopy and the ns Z-scan and nonlinear transmission are described in the text.
Dithallium(I) tetra-tert-butyl-phthalocyanine halide, Tl2(t-Bu4Pc), 1. A torch was used to close one end of an 8 cm long piece of Pyrex glass tubing (5 mm I.D., 7 mm O.D.). This tube was charged with a mixture of 0.50 g 4-tert-butylphthalonitrile and 0.52 g of thallium metal flakes (cut from a larger piece with a razor blade), evacuated (first at 1000 microns and then at 10 microns) and flame sealed. After distributing the contents evenly along the length of the tube, the tube was placed in a horizontal tube furnace and maintained at 325°C for 5 days. The thallium melted and the tube contents quickly turned a deep forest green color. After five days, the tube was removed, cooled in liquid nitrogen, and opened. The contents were removed from the tube by repeated washings with dichloromethane. The dichloromethane solution was filtered to remove unreacted thallium metal and glass fragments and used in the production of 3 and 4, as described below, without further purification. In preparing the chloro derivative, 2, the dichloromethane solvent was removed and the resulting residue of 1 was dissolved in carbon tetrachloride. Failure to remove the dichloromethane solvent results in significant demetallation on oxidation with chlorine. Tl2(t-Bu4Pc) is extremely sensitive to acid; attempts to chromatograph on silica gel resulted in complete demetallation. UV-vis (CH2Cl2): λmax(nm) (log ε) = 344 (4.71), 495 (3.93), 606 (sh,3.77), 664 (4.47), 728 (5.08).
Thallium(III) tetra-tert-butyl-phthalocyanine halide, Tl(t-Bu4Pc)X, 2-4. Oxidation of 1 to any of the thallium(III) derivatives followed the same procedure. A solution of halogen in chlorinated solvent was prepared (0.50 g of I2 OR Br2 added to 10mL dichloromethane OR 10 mLs of carbon tetrachloride saturated with chlorine gas at room temperature) and then added in aliquots to the solution of 1. The visible spectrum was monitored during the addition to avoid over-oxidation. Over the course of the reaction, the peak at 726 nm disappeared and a peak at ~704 nm grew in (this peak is the same for all three halogen derivatives). Once the peak at 726 nm had fully disappeared, the halogenation was completed and the additions discontinued. Isolation of the desired product required filtration of a fine yellow or white precipitate of thallium(I) halide followed by removal of solvent on a rotary evaporator. The resulting solid was chromatographed to isolate the desired product. The iodo and bromo complexes were purified on silica gel using a hexanes to 1:1 hexanes:dichloromethane solvent gradient, collecting the first blue-green fraction. The chloro derivative was best isolated by chromatography on Celite using the same solvent gradient; again the first blue-green fraction was collected. Removal of solvent from the column eluent gave the desired compound in moderate yield, ~0.10 g (15%) from t-butylphthalonitrile.
Thallium(III) tetra-tert-butyl-phthalocyanine chloride, Tl(t-Bu4Pc)Cl, 2
Anal. Calcd for C48H48N8TlCl ⋅ 2 C6H14: C, 62.71; H, 6.67; N, 9.75. Found: C, 62.40; H, 6.23; N, 9.35.
UV-vis (CH2Cl2): λmax(nm) (log ε) = 374 (4.68), 642 (4.39), 706 (5.09).
1H NMR (CDCl3): δ(ppm) = 1.80 (m, 9H); 8.32 (m, 1H); 9.46 (m, 1H); 9.60 (m, 1H).
MS m/z: calcd. for M+ 975, 977; found 736 (M – Tl – Cl)+, 940 (M – Cl)+, 1146 (M – Cl + Tl + 2H)+, 1679 (2(M – Cl) – Tl + 3H+).
Thallium(III) tetra-tert-butyl-phthalocyanine bromide, Tl(t-Bu4Pc)Br, 3
Anal. Calcd for C48H48N8TlBr ⋅ 3/2 C6H14: C, 59.50; H, 6.04; N, 9.74. Found: C, 59.93; H, 5.88; N, 9.34.
UV-vis (CH2Cl2): λmax(nm) (log ε) = 373 (4.75), 640 (4.41), 704 (5.13).
1H NMR (CDCl3): δ(ppm) = 1.80 (m, 9H); 8.33 (m, 1H); 9.46 (m, 1H); 9.58 (m, 1H).
MS m/z: calcd. for M+ 1019, 1021; found 736 (M – Tl – Br)+, 940 (M – Br)+, 1146 (M – Br + Tl + 2H)+, 1679 (2(M – Br) – Tl + 3H+).
Thallium(III) tetra-tert-butyl-phthalocyanine iodide, Tl(t-Bu4Pc)I, 4
Anal. Calcd for C48H48N8TlI ⋅ 2 C6H14 ⋅ C7H8: C, 60.43; H, 6.28; N, 8.41. Found: C, 60.56; H, 6.29; N, 8.05.
UV-vis (CH2Cl2): λmax(nm) (log ε) = 372 (4.78), 638 (4.42), 706 (5.18).
1H NMR (CDCl3): δ(ppm) = 1.81 (m, 9H); 8.35 (m, 1H); 9.43 (m, 1H); 9.55 (m, 1H).
MS m/z: calcd. for M+ 1067; found 736 (M – Tl – I)+, 940 (M – I)+, 1146 (M – I + Tl + 2H)+, 1679 (2(M – I) – Tl + 3H+).
Results and discussion
Soluble thallium phthalocyanine synthesis. Metallophthalocyanines are generally insoluble due to their planar and rigid structure. Soluble phthalocyanines can be made by appending large and/or flexible groups to the macrocycle periphery [14–17 ]. Accordingly, our initial efforts to prepare a soluble thallium complex focused on inserting the metal into the soluble and commercially available 2,9,16,23-tetra-tert-butyl-29H,31H-phthalocyanine. Unfortunately, all attempts to insert thallium into a pre-formed macrocycle (by varying the thallium source, the reaction solvent, temperature, adding a hindered base, using the dilithium salt of the macrocycle, etc.) were unsuccessful. A more successful procedure was to form the macrocycle around a reduced metal as shown in Fig. 1 . A precedent for this approach is the work of Janczak et al . The synthesis started with the appropriately substituted phthalonitrile, 4-tert-butylphthalonitrile, utilized Suga’s modification , and yielded a dithallium(I) phthalocyanine which is readily soluble in organic solvents.
Dithallium(I) tetra-tert-butylphthalocyanine, 1, is produced as a statistical mixture of positional isomers, evident in the 1H NMR spectrum (vide infra), due to the asymmetry in the precursor dinitrile. No attempt was made to separate these isomers from each other or from byproducts of the condensation reaction; the mixture was carried on to the next step without purification.
Treatment of 1 with halogen in chlorinated solvent gave the thallium(III) tetra-tert-butylphthalocyanine halide complexes, 2-4, as shown in Fig. 1. Homborg et al.  used excess iodine (> 1 equiv.) to prepare the thallium(III) iodo complex of unsubstituted phthalocyanine, which crystallized from solution. In our case, excess halogen (especially bromine and chlorine) reacted further with the soluble thallium(III) phthalocyanine reducing the yield of desired product. In order to avoid over-oxidation, the halogen was added in aliquots and the reaction progress was monitored spectrophotometrically. The final products were purified by chromatography and isolated in overall yields of approximately 15% (from 4-tert-butylphthalonitrile).
Over the course of the oxidation, a fine white or yellow precipitate was observed in the solution. These byproducts were identified as the thallium(I) halides by comparison of the X-ray fluorescence spectra to authentic samples.
In order to determine the oxidation reaction stoichiometry, a spectrophotometric titration of 1 with dilute iodine solution was performed with the results displayed in Fig. 2 . A 3.0 mL dichloromethane solution of 1 (extracted from the sealed reaction tube and filtered, concentration and purity unknown) was treated with successive 5 μL aliquots of 0.63 mM iodine in dichloromethane. The UV-visible spectra stopped changing within a minute of each aliquot addition after which time the spectra were recorded. At these concentrations, additional aliquots of iodine solution produced no further changes in the spectrum beyond that shown in the figure. Figure 2 further shows three sharp isosbestic points that are maintained over the course of the titration indicating the clean conversion of 1 to 4.
Based on the solution volume in the cuvette, the extinction coefficient of 4 (known from a purified sample) and the final absorbance at 704 nm, the initial amount of 1 was calculated as 2.43 x 10−8 moles. Since 40 μL of I2 were required to complete the oxidation, the total moles of I2 added to the cuvette was 2.52 x 10−8 moles. Thus the dithallium precursor, 1, reacts on a 1:1 ratio with added halogen. This information, combined with the identification of the thallium by-product, gives the stoichiometry of the oxidation reaction as:Fig. 2 for a representative example) are typical of metallophthalocyanines with nominal C4v symmetry. They possess a sharp Q band at 704-706 nm, a slightly weaker B band at 374-376 nm and several weak charge-transfer and/or vibronic bands at intermediate wavelengths.
The 1H NMR spectrum of 4, recorded in CDCl3, is shown below in Fig. 3 . All three halo complexes, 2-4, show four distinct sets of signals, three in the aromatic region and one in the aliphatic region, in 1:1:1:9 integral ratios, consistent with a diamagnetic tetra-tert-butylphthalocyanine complex. Expansion of the aromatic region shows two of the signals are broad doublets and the third is an overlapping series of four doublets – the result of a combination of spin-spin coupling and a mixture of very similar regioisomers. Expansion of the aliphatic region of the NMR spectrum shows four singlets with nearly equal integrals – again indicative of four similar magnetic environments for the t-butyl groups. These 1H NMR spectra are consistent with a statistical mixture of tetra-t-butyl phthalocyanine regioisomers, similar to those of the planar nickel(II) complex reported and analyzed by Hanack et al. [21, 22 ] The correlation between the NMR spectra and the statistical mixture of regioisomers is discussed below.
Mass spectra for 2-4 were collected on a matrix-assisted laser desorption / ionization – time of flight (MALDI-TOF) instrument; the highly colored phthalocyanine complexes were readily desorbed and ionized in the absence of a matrix. These mass spectra confirm the presence of the t-Bu4 thallium(III) phthalocyanine complex with a base peak at m/e = 940 but no evidence for a parent peak with the halide axial ligand was found. Apparently, the halides are labile in the ionization. Attempts to detect the molecular ion peak by reducing the energy of laser desorption/ionization pulses were unsuccessful. The MALDI mass spectra also show a moderate lability of the thallium ion in these complexes – in all three materials, small signals consistent with a) t-Bu4Pc (m/e = 736), b) Tl2(t-Bu4Pc) (m/e = 1146) and c) Tl(t-Bu4Pc)2 (m/e = 1679) are observed. We have confirmed the assignment of t-Bu4Pc with a commercial sample purchased from Sigma-Aldrich Corporation and of Tl2(t-Bu4Pc), which is identical to compound 1 of this work. There is ample precedent for the sandwich compound, Tl(t-Bu4Pc)2, in the work of Buchler and others . We do not believe that any of these substances contaminate 2-4 but instead are formed in situ under the conditions of the MALDI analysis.
In order to confirm the identity of the halide axial ligand in 2-4, elemental analysis and X-ray fluorescence spectroscopy (XFS) were used. The XFS spectra confirmed both the identity of the halide ion and the 1:1 ratio between halide and thallium ions. Elemental analyses, detailed in the experimental section, are consistent with the expected halide ion after accounting for trapped solvent.
Phthalocyanine regioisomer analysis. There are four positional isomers that arise from the peripheral t-Bu groups in each of these materials. In order to interpret the NMR spectra, the different magnetic environments of these substituents must be considered. As shown in Fig. 4 , locating the t-butyl group on one of the peripheral β carbon atoms puts it closer to one adjacent phthalonitrile residue (proximal) than to the other (distal). Note that there are two possible arrangements of substituents on the β carbon atoms of both the proximal and distal phthalonitrile residues for a total of four possible magnetic environments for each t-butyl group – consistent with the observed NMR spectrum.
These four environments are arbitrarily assigned designators A – D depending on what substituent (t-Bu or H) is closer on the adjacent phthalonitrile residues: A (proximal t-Bu, distal t-Bu), B (proximal t-Bu, distal H), C (proximal H, distal t-Bu) and D (proximal H, distal H). The equal integrals for signals in the aliphatic region indicate an equal number of t-butyl groups in environments A-D. This, in turn, depends on the relative amounts of the phthalocyanine regioisomers produced. We show below that the observed NMR integrals are consistent with a statistical mixture of phthalocyanine stereoisomers.
Depending on the relative orientation of t-butyl groups on the four phthalonitrile residues, there are four possible thallium(III) phthalocyanine halide steroisomers. One possible macrocycle stereoisomer, with Cv symmetry, is shown in Fig. 5 , below. Note the two different types of t-butyl groups in this structure – two are in magnetic environment A and two are in magnetic environment D (defined above).
All four possible phthalocyanine stereoisomers, with their symmetry group and number of t-butyl groups in magnetic environments A-D, are shown schematically in Table 1 below. Note that the stereoisomer symmetry groups are reduced slightly from those reported by Hanack et al. in planar Ni(II) complexes [21, 22 ]. Also listed in Table 1 are the expected ratios of phthalocyanine stereoisomers from random combinations of the phthalonitrile precursors. The statistical ratios are easily generated by looking at all possible orientations of the phthalonitrile precursors.
From data in Table 1, the ratio of t-butyl groups in magnetic environments A-D can be calculated for the statistical mixture. For example, the relative number of t-butyl groups in magnetic environment A (4 C1 molecules x 1 group + 2 Cv molecules x 2 groups = 8) equals the number in magnetic environment B (4 C1 molecules x 1 group + 1 C2v molecules x 4 groups = 8). Complete analysis of the data in Table A reveals the ratio of t-butyl groups in magnetic environment A – D is 1:1:1:1 for the statistical mixture – exactly as observed experimentally. As previously reported by Hanack , the phthalonitrile couplings leading to the phthalocyanine macrocycle are random, influenced by neither steric nor electronic factors.
Thallium(III)phthalocyanine complex stability. Of the three halide complexes reported here, the iodo derivative, 4, is the most stable – it may be chromatographed on silica gel and is stable for months in air as a dichloromethane solution. The bromo complex, 3, could also be purified by flash chromatography on silica gel but would demetallate if exposed for an extended period to silica gel. The chloro complex, 2, demetallated completely when chromatographed on silica gel. 2 could be purified by chromatography on Celite (diatomaceous earth). As discussed below, the photostability of the three halo derivatives followed this same pattern.
While the relative stability to purification processes can be kinetic, we did demonstrate the higher thermodynamic stability of the iodo material by a halide exchange experiment. The thallium(III) bromo complex, 3, was treated with excess tetraalkyl ammonium iodide in dichloromethane. Within a few minutes, thin layer chromatography shows complete conversion to the iodo complex, 4. The complementary experiment, treatment of 4 with an excess of bromide salt, produced only a small amount of 3 with most of starting material unchanged. The observed relative stability is consistent with hard/soft acid/base theory [24, 25 ]. The relatively soft thallium(III) ion has a higher affinity for the softer iodide ion.
The precursor dithallium(I) compound, 1, was the least stable chemically. Attempts to chromatograph this material on silica gel, alumina or florisil resulted in formation of the unmetallated free-base phthalocyanine presumably by protonation of the pyrrolic nitrogen atoms by residual protons. On exposure to air, a dichloromethane solution of this material oxidized over the course of weeks.
Aggregation studies. Due to their planar aromatic structure, phthalocyanines are well known to aggregate in solution [26–30 ]. The degree of aggregation depends markedly on solvent. Molecular aggregation typically adds relaxation pathways to excited states, reduces excited state lifetimes and thus reduces the observed nonlinear absorption compared to the monomers. One strategy to inhibit aggregation is the use of organic caps or axial ligands to block one or both faces of the macrocycle . The five-coordinate, square-planar thallium(III) halide complexes reported here contain an open phthalocyanine face and may form dimers at high concentration.
In order to determine the concentration dependence of dimer formation, the visible absorption spectrum of thallium(III) tetra-t-butylphthalocyanine iodide, 4, was examined as a function of concentration in both toluene and chloroform. Shown below in Fig. 6 is the visible absorption spectrum of 4 in chloroform at three concentrations ranging from 1 μM to 20 mM. Close agreement of the extinction coefficients at all wavelengths for all three concentrations indicates a) complete solubility in chloroform even at 20 mM and b) no aggregation at concentrations up to 20 mM in chloroform.
A similar examination of the visible spectrum of 4 in toluene was conducted with the results shown below in Fig. 7 . Here we plot the measured extinction coefficients at two λmax for each concentration examined. Clearly, the extinction coefficients are constant at concentration of 2 mM and below but begin to drop at concentrations above 2 mM. We believe the decrease in molar extinction coefficients is due to macrocycle aggregation as reported separately by Monahan [27, 28 ] and Snow . In addition to the decrease in ε at the two λmax, we also observe (not shown) an increase in absorption at wavelengths > 800 nm at concentrations above 2 mM. Similar changes observed in absorption spectra of In(t-Bu4Pc)Cl at high concentration were attributed to aggregation .
Femtosecond transient absorption experiments
The transient absorption behavior of 2-4 were measured at a variety of a) excitation wavelengths, b) excitation pulse energies and c) delays between the excitation pulse and the probe pulse using the apparatus shown in Fig. 8 . Measurement of the fundamental photophysics–excited state cross sections and lifetimes and intersystem crossing dynamics–for these compounds determines their suitability for use as broadband optical limiters.
As seen in Fig. 8, the 771 nm (10 – 1000 Hz, 150 fs pulse, 1 mJ) output of a Ti:Sapphire chirped pulse amplifier (Clark CPA 2001) was split into two beams. The first beam was either passed to an optical parametric amplifier (OPA) to generate a tunable excitation beam (450 – 1650 nm, 10 μJ, 30-150 fs) or the second harmonic of the fundamental was used. The excitation beam passed through a) a four-pass delay stage with 1 μm step size), b) a half-wave plate / polarizer pair which allowed control of the pulse energy and c) a lens which focused the beam on the sample. The focal spot size at the sample for a 386 nm excitation beam was measured at 258 μm by razor edge scans.
The second beam from the CPA was focused into a sapphire crystal in order to generate a white light continuum probe pulse. The polarization of the continuum was set at 54.7° relative to the excitation pulse in order to eliminate the influence of orientational effects. The continuum was filtered to remove residual 771 nm light and was collimated and split into sample and reference beams, the former of which was focused on the sample (spot size = 119 μm). That the spot size of the probe beam is smaller than that of the excite beam insures that excited state species are probed. Both the sample and reference beams were collected in optical fiber bundles and simultaneously measured in a monochromator via spatially separated regions on an intensified CCD array. ΔOD was calculated from the absorbance (-log10(S/R)) of the excited sample minus the absorbance of the unexcited sample after first correcting for both sample/reference leakage and dark signal. Time resolution was obtained by varying the arrival of the excitation pulse relative to that of the continuum probe pulse using the delay stage. The samples (2-4) were studied as ~1 mM solutions in toluene contained in 200 μm pathlength quartz cells from Starna. The recorded spectra for each of the samples were similar, so the discussion below will focus on 4, the iodo compound. The transient absorption response of 4 at relatively long delay times (200 to 2400 ps) following excitation by a 633 nm 1.0 μJ pulse is shown in Fig. 9 .
The displayed spectra are selected from a larger set collected at different delay times. This data shows a long-lived excited state which shows very little decay from 200 ps to 2.5 ns after excitation. The lifetime of this long-lived excited state is at least 20 ns, based on the small observed change in ΔOD over 2.4 ns. The spectra are similar in shape to those obtained from Lead Phthalocyanine, Indium Phthalocyanine and Indium Naphthalocyanine though they are blue shifted relative to the transient spectra from those compounds. Similar to the other phthalocyanines, the induced absorption from 440 nm to 610 nm shown in Fig. 9 clearly demonstrates that the long-lived excited state of 4 absorbs more strongly than the ground state; thus, 4 is a reverse saturable absorber over this wavelength range. At longer wavelengths bleaching of the vibronic sidebands in the ground state spectrum(see Fig. 6) dominates the excited state absorption.
From this data and the measured excitation density produced by the pump pulse, the long-lived excited-state cross section is calculated to be σe.s. = 6.5 x 10−17 cm2 at 515 nm, the wavelength of maximum excited-state absorption. This cross section is independent of excitation pulse energy (varied from 0.4 μJ to 4 μJ) and excitation wavelength (386 nm, 532 nm, 633 nm and 680 nm). This calculated σe.s. of 4 is comparable to those reported previously for In(tBu4Pc)Cl [8, 32 ] and Pb(CP4Pc) [33, 34 ], which are among the better optical limiters known. The maximum ratio of excited state cross section to that of the ground-state, σe.s. / σg.s, is > 100 in this portion of the spectrum. The ratio is larger than those for In(tBu4Pc)Cl and Pb(CP4Pc) because of the low ground-state absorptivity of 4 at 515 nm.
The transient absorption spectra observed for both 2 and 3, the chloro and bromo derivatives respectively, were similar. The ΔOD curves of 2 and 3, 15 ps and later after excitation, have the same shape as that shown in Fig. 9 (of 4, the iodo derivative). In all three materials, the excited state absorbs more than the ground state from 440 to 610 nm and the maximum in the ΔOD curve occurs at 515 nm. The magnitudes of the long-lived exited state cross sections, 8.5 x 10−17 cm for 2 and 9.0 x 10−17 for 3 at 515 nm (excitation pulse wavelength and energy were 386 nm and 1.0 μJ, respectively) were slightly larger than that for the iodo compound, 4. However, the excited-state cross sections of 2 and 3 did showsome dependence on the energy of the excitation pulse – these values fall to ~6.0 x 10−17 cm2 when the excitation pulse energy is raised to 3.8 μJ. When measuring the transient absorption of 2 and 3 at the higher excitation energy, we observed a small but temporary decay in the linear absorption spectrum. This indicates some reversible photochemical changes to these materials in the focal region.
The time evolution of the transient absorption response at short delay times (< 25 ps between the excitation and the probe pulses), provides information on the kinetics of formation of the long-lived excited state. Figure 10 shows the raw data at short pulse delays for the ΔOD as a function of wavelength for the iodo material, 4, when excited with a 150 fs pulse at 680 nm and 1.0 μJ pulse energy.
In Fig. 10, the different wavelengths of the white-light probe pulse are chirped temporally so they arrive at the sample at slightly different times. The blue spectral components of the probe pulse are delayed more than the red ones. This is evident in Fig. 10 at delay times between −2.8 and 1.2 ps, where there is a measurable ΔOD at blue wavelengths but zero ΔOD signal at red wavelengths because the red light has passed through the sample before the arrival of the excitation pulse. The data in Fig. 10 shows the immediate formation of excited states that are more strongly absorbing than the ground-state of 4. Note that there is a rapid decay of the initial spectrum to that observed at later times.
It is straightforward to correct for the chirp and derive the kinetics at each wavelength. This was done by fitting the leading edges and peaks of the each of the delay times shown in Fig. 10 to a Sellmeier dispersion equation and using the result to shift the time axis of the individual wavelength traces. The time evolution of the ΔOD signal compensated for thechirp at several wavelengths is shown in Fig. 11 . At all wavelengths the decay is dominated by an immediate rise, a rapid decay followed by the long-lived decay described above. At wavelengths beyond 605 the late time ΔOD signal is negative which means that the excited state cross section at these wavelengths is smaller than that of the ground state.
Comparison of the early time dynamics is most effectively accomplished by subtracting the long time response from the data in Fig. 11 and then normalizing each of the individual wavelength traces to their maximum response. The results of this data processing are shown in Fig. 12 .
In Fig. 12 the short term dynamics vary with wavelength. The decay at 445 nm, 585 nm, and 605 nm are a simple exponential. This is consistent with the expected photophysical response of 4 after excitation into a singlet excited state. The excited singlet undergoes intersystem crossing into a triplet excited state in approximately 5 ps. The triplet excited state then slowly relaxes to the singlet ground state with a time constant greater than 20 ns. The intersystem crossing time constant is reasonable. It is similar to that observed for lead phthalocyanine in solution and faster than that measured for a choro indium(III) phthalocyanine.
The decay at other wavelengths in Fig. 12, especially near 525 nm to 545 nm and near 625 nm does not decay to zero with the ~5 ps lifetime. There is an additional component or mechanism that contributes at these wavelengths. The lifetime for this component was determined by fitting all of the data in Fig. 12 to three first-order processes. All short term dynamics were fit to lifetimes of 4.8 ± 1 ps, 250 ± 40 ps and 20 ns by an iterative process. The values for the three lifetimes were arrived at through statistical fitting and simulation which was well determined since the lifetimes are distinctly different. The observed nonlinear least squares fits are of high quality—the residuals are randomly scattered about zero. After the lifetimes were determined, the amplitudes of the spectral changes attributed to each of three exponential decay times were calculated to gain spectroscopic information on their source. The spectra of the three components are shown in Fig. 13 .
The blue trace (20 ns decay) in the figure is for the long lived excited state species. The shape closely matches the shape of the spectra displayed in Fig. 9. Since the shape of the depleted ground state spectrum over the region from 445 to 590 is essentially flat, the spectrum is primarily due to the absorption spectrum of the triplet excited state. The black or fast trace (4.8 ± 1 ps decay) is a combination of the ground state bleach spectrum and the initial excited state singlet absorption. There is a spectral maximum at 607 nm and apparently an additional maximum to the blue of the shortest wavelength probed. The red trace (250 ± 40 ps decay) is the amplitude of the middle lifetime component. The maximum contribution for this component in the 445-600 nm region this reaches ~6% of the response. It represents a minority species or process. The data does not distinguish between a mechanism where an initially produced excited state decays to an intermediate which then produces the triplet state versus one in which two independent species are excited but decay to the triplet state with different intersystem crossing rates. The largest spectral feature is near 630 nm, close to a vibronic sub-band of the ground state monomer. The origin of this spectral response is not likely to be an aggregate of 4.
In summary, after excitation into a singlet excited state the major photophysical pathway of 4 is intersystem crossing to a triplet excited state with a time constant of 4.8 ± 1 ps. The triplet excited state relaxes to the singlet ground state with a time constant estimated to be greater than 20 ns. A small part of the excitation, responsible for less than ~6% of the spectrum between 445 and 600 nm, appears to form an excited state that decays to the triplet state with a time constant of 250 ± 40 ps. Its origin is not known. The observed photophysics implies that this material may be useful for nanosecond optical limiting.
ns Z-Scan and nonlinear transmission experiments
The transient white light absorption experiments described above show that the photophysics observed from 2, 3 and 4 are sufficient to produce good nanosecond optical limiters: 1) They have broad regions of induced absorption which arise from excited state absorption cross sections that exceed those of the ground state; 2) These compounds have excited state lifetimes that exceed 20 ns and 3) They are highly soluble so that a large concentration can be placed in the focal volume of an optical system. To examine the refractive part of the nonlinear response and test the response of the materials to increasing fluences of nanosecond laser pulses, Z-scan and Nonlinear Transmission measurements were performed.
Z-scan experiments were performed on a 20 mM solution of 4 in chloroform in a 34.4 μ pathlength cell. The Z-scan setup is similar to those described in [35, 36 ]. The sample was interrogated using a seeded 7 ns FWHM, 532 nm laser. The focusing optics were set at f/5 and the resulting 1/e2 spot radius was measured to be 2.5 μm using the razor edge technique. The transmitted light was collected by a complementary lens and the energy of the transmitted pulse was measured as a function of sample position relative to the focal point.
A series of open aperture Z-scans for 20 mM 4 in chloroform is shown in Fig. 14 . The relative energy of the transmitted pulse as a function of distance between the sample and the focal point is monitored. At each laser pulse energy, the transmission decreases as the sample approaches the focal point. As the pulse energy is increased the decrease in transmission becomes more pronounced consistent with the cross-sections measured above. Compound 4 is a reverse saturable absorber. The open aperture Z-scan results are consistent with excited state absorption i.e. sequential two-photon absorption as the nonlinear mechanism. At thetwo highest pulse energies, the relative baseline transmittance decreases slightly. This may indicate that some longer lasting induced absorption is taking place in addition to the excited state response. There is no evidence for permanent photochemical changes in that the absorption spectrum after the Z-scan and limiting experiments remains unchanged. Also apparent in the highest two energy Z-scans are slight inflection points located at approximately + 50 μm and –50 μm. These may be indicative of additional processes contributing to the response and are beyond the scope of this paper.
A similar series of closed-aperture (40%) Z-scans were performed in order to determine the refractive contributions to the nonlinear response. The results, after normalizing the closed aperture scans by the open aperture Z-scan, are shown in Fig. 15 . The shape of these curves indicates that the refractive contribution to the nonlinear response is due to a negative nonlinear refractive index. It arises from excited state and thermal mechanisms. The symmetry of the refractive Z-scan indicates that both the thermal and the excited state refractive index are of the same sign. In terms of optical limiting this is important in that both mechanisms contribute to the limiting response cooperatively. At f/5, both excited state and thermal  mechanisms can contribute to the response since they originate from either a population or a temperature gradient across the focal spot.
Z-scan experiments not only yield measurement of the nonlinear absorptive and refractive response they also allow determination of the optimal geometry for measurement of the nonlinear transmission of the sample. Figure 16 shows the relative transmission for a sample of 4 (20 mM, 34.4 m pathlength, Tg.s. (532) = 0.937) at 532 nm over almost six orders of magnitude variation in input energies.
At the lowest energies the data was taken at 10 Hz. Above 2 nJ a 0.1 Hz repetition rate was used to allow the sample to thermally recover between shots and above 10 nJ single shot measurements were taken, translating the sample between shots. The sample transmission is roughly linear at input energies up to about 2 nJ at which point the transmission begins to decrease with increasing input energy. The threshold energy (50% transmission) is ~100 nJ. This corresponds to a peak fluence of 1.02 J/cm2. This threshold is higher than that observed for In(t-Bu4Pc)Cl [32, 38 ] and Pb(CP4Pc) [33, 34 ], despite the comparable magnitude of the maximum of the excited state absorption cross section. The higher 532 nm threshold is probably due to the fact that the excited state absorption of 4 peaks near 515 nm and falls off by approx. 15% at 532 nm (see Fig. 9).
Three novel soluble thallium(III) phthalocyanine complexes have been synthesized and chemically characterized. All three make use of four t-butyl groups (one each randomly distributed on each phthalonitrile residues) on the macrocycle periphery to provide solubility but differ in the identity of a halide axial ligand. The iodo complex is shown to be stable in solution over several months while the bromo and chloro analogs are successively less stable both to halide loss and demetallation. All three complexes show nonlinear optical behavior with the photophysics of the iodo complex well characterized. A series of transient white light absorption experiments shows that this material is a reverse saturable absorber consistent with a sequential two photon absorption mechanism. An initial singlet excited state decays to a long-lived triplet excited state with a time constant τ1 of ~4.8 ± 1 ps. The long-lived triplet state slowly decays back to the ground state with a time constant of at least 20 ns. A component with an apparent intersystem crossing time of 250 ± 40 ps also contributes to a small fraction of the NLO response. The triplet excited state shows enhanced absorption (over that of the ground state) over the λ = 425-610 nm region of the spectrum. Ratios of excited state to ground state molar absorptivities rise to as much as 100 at certain wavelengths in this region of the spectrum. Z-scan and nonlinear transmission experiments at 532 nm show that the thallium(III) phthalocyanine iodide complex, 4, is effective at limiting nanosecond laser pulses with a threshold fluence of approximately 1 J/cm2. Closed aperture ns Z-scan results shows that the refractive portion of the optical limiting is due to a negative nonlinear refractive index attributed to cooperative thermal and excited state contributions.
The authors acknowledge helpful discussions with Dr. Art Snow of the Chemistry Division, Naval Research Laboratory, Washington, DC, 20375, and the USNA Trident Scholar Committee. Support for this work was received from the USNA Research Office and from NRL through ONR.
References and links
1. J. S. Shirk, “Protecting the Warfighter's Vision In a Laser-rich, Battlefield Environment,” Optics and Photonics News 11(4), 19–23 (2000). [CrossRef]
2. D. R. Coulter, V. M. Miskowski, J. W. Perry, T.-H. Wei, E. W. Van Stryland, and D. J. Hagan, “Optical Limiting In Solutions Of Metallo-Phthalocyanines And Naphthalocyanines,” Proc. SPIE 1105, 42–51 (1989). [CrossRef]
3. J. S. Shirk, J. R. Lindle, F. J. Bartoli, C. A. Hoffman, Z. H. Kafafi, and A. W. Snow, “Off-Resonant 3rd-Order Optical Nonlinearities of Metal-Substituted Phthalocyanines,” Appl. Phys. Lett. 55(13), 1287–1288 (1989). [CrossRef]
4. M. Hosoda, T. Wada, A. Yamada, A. F. Garito, and H. Sasabe, “3rd-Order Nonlinear Optical-Properties in Soluble Phthalocyanines with Tert-Butyl Substituents,” Jpn. J. Appl. Phys. 30(8), 1715–1719 (1991). [CrossRef]
5. J. S. Shirk, J. R. Lindle, F. J. Bartoli, Z. H. Kafafi, A. W. Snow, and M. E. Boyle, “Third-Order Nonlinear Optical Properties of Metallo-Phthalocyanines,” Int. J. Nonlinear Opt. Phys. 1(04), 699–726 (1992). [CrossRef]
6. J. W. Perry, K. Mansour, S. R. Marder, K. J. Perry, D. Alvarez Jr, and I. Choong, “Enhanced Reverse Saturable Absorption and Optical Limiting in Heavy-Atom-Substituted Phthalocyanines,” Opt. Lett. 19(9), 625–627 (1994). [CrossRef] [PubMed]
7. J. S. Shirk, R. G. S. Pong, S. R. Flom, M. E. Boyle, and A. W. Snow, “Materials for reverse saturable absorption optical limiters,” in Materials for Optical Limiting R. Crane, K. Lewis, E. VanStryland, and M. Khoshnevisan, eds. (Mater. Res. Soc.Symp. Proc. 374, Pittsburgh, PA, 1995), pp. 201–209.
8. J. W. Perry, K. Mansour, I. Y. S. Lee, X. L. Wu, P. V. Bedworth, C. T. Chen, D. Ng, S. R. Marder, P. Miles, T. Wada, M. Tian, and H. Sasabe, “Organic optical limiter with a strong nonlinear absorptive response,” Science 273(5281), 1533–1536 (1996). [CrossRef]
9. 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,” App. Phys. B-Photophys, Laser Chem. 54, 46–51 (1992).
10. K. Mansour, J. D. Alvarez Jr, K. J. Perry, I. Choong, S. R. Marder, and J. W. Perry, “Dynamics of optical limiting in heavy-atom substituted phthalocyanines,” Proc. SPIE 1853, 132–141 (1993). [CrossRef]
11. W. J. Su, T. M. Cooper, and M. C. Brant, “Investigation of reverse-saturable absorption in brominated porphyrins,” Chem. Mater. 10(5), 1212–1213 (1998). [CrossRef]
12. P. D. Huffman, J. P. Fitzgerald, and J. S. Shirk, “New lead and thallium phthalocyanines as optical limiting materials,” Abstracts Am. Chem. Soc. 223, U238(2002).
13. D. Dini, M. J. F. Calvete, M. Hanack, W. Z. Chen, and W. Ji, “Synthesis of axially substituted gallium, indium and thallium phthalocyanines with nonlinear optical properties,” ARKIVOC 2006(3), 77–96 (2005). [CrossRef]
14. S. A. Mikhalenko, S. V. Barkanova, O. L. Lebedev, and E. A. Lukyanets, “Phthalocyanines and related compounds IX. Synthesis and Electronic Absorption Spectra of tetra-4-t-butylphthalocyanines,” J. Gen. Chem. USSR 41, 2770–2773 (1971).
15. E. A. Cuellar and T. J. Marks, “Synthesis and characterization of metallo and metal-free octaalkylphthalocyanines and uranyl decaalkylsuperphthalocyanines,” Inorg. Chem. 20(11), 3766–3770 (1981). [CrossRef]
16. M. Hanack, J. Metz, and G. Pawlowski, “Soluble trans-di-1-alkynyl(tetraalkylphthalocyaninato)metal and poly-trans-ethynyl(tetraalkylphthalocyaninato)metal IVB Derivatives,” Chem. Ber. 115, 2836–2853 (1982). [CrossRef]
17. C. C. Leznoff and T. W. Hall, “The synthesis of a soluble, unsymmetrical phthalocyanine on a polymer support,” Tetrahedron Lett. 23(30), 3023–3026 (1982). [CrossRef]
18. J. Janczak and R. Kubiak, “Crystal and molecular-structure of dithallium phthalocyanine At 300-K,” J. Alloys Compd. 202(1-2), 69–72 (1993). [CrossRef]
19. T. Suga, S. Isoda, and T. Kobayashi, “Characterization and electrical conductivity of dithallium phthalocyanine (Tl2Pc),” J. Porphyr. Phthalocyanines 3(06), 397–405 (1999). [CrossRef]
20. K. Schweiger, H. Hueckstaedt, and H. Homborg, “Iodophthalocyaninato(2–)thallium(III) – Synthese und Kristallstruktur,” Z. Anorg. Allg. Chem. 624(2), 167–168 (1998). [CrossRef]
21. M. Hanack, D. Y. Meng, A. Beck, M. Sommerauer, and L. R. Subramanian, “Separation of structural isomers of tetra-tert-butylphthalocyaninatonickel(II),” J. Chem. Soc. Chem. Commun. 1993(1), 58–60 (1993). [CrossRef]
22. G. Schmid, M. Sommerauer, M. Geyer, and M. Hanack, “Synthesis and Chromatographic Separation of Tetrasubstituted and Unsymmetrically Substituted Phthalocyanines ” in Phthalocyanines: Properties and Applications,4, C. C. Leznoff, and A. P. B. Lever, eds. (VCH, New York, 1996), pp. 1–18.
23. J. W. Buchler and D. K. P. Ng, “Metal Tetrapyrole Double- and Triple-Deckers with Special Emphasis on Porphyrin Systems,” in Phorphyrin Handbook,3, K. M. Kadish, K. M. Smith, and R. Guilard, eds. (Academic Press, San Diego, CA, 2000), pp. 245–294.
24. R. G. Pearson, “Hard and soft acids and bases,” J. Am. Chem. Soc. 85(22), 3533–3539 (1963). [CrossRef]
26. Z. A. Schelly, D. J. Harward, P. Hemmes, and E. M. Eyring, “Bonding in dye aggregates - energetics of dimerization of aqueous cobalt(II)-4,4′,4″,4‴-tetrasulfophthalocyanine ion,” J. Phys. Chem. 74(16), 3040–3042 (1970). [CrossRef]
27. A. R. Monahan, J. A. Brado, and A. F. Deluca, “Association of copper(II), vanadyl, and zinc(II) 4,4′,4″,4″-tetraalkylphthalocyanine dyes in benzene,” J. Phys. Chem. 76(14), 1994–1996 (1972). [CrossRef]
28. A. R. Monahan, A. F. Deluca, and J. A. Brado, “Dimerization of a copper(II)-phthalocyanine dye in carbon-tetrachloride and benzene,” J. Phys. Chem. 76(3), 446–449 (1972). [CrossRef]
29. A. W. Snow and N. L. Jarvis, “Molecular association and monolayer formation of soluble phthalocyanine compounds,” J. Am. Chem. Soc. 106(17), 4706–4711 (1984). [CrossRef]
30. R. D. George, A. W. Snow, J. S. Shirk, and W. R. Barger, “The alpha substitution effect on phthalocyanine aggregation,” J. Porphyr. Phthalocyanines 2(01), 1–7 (1998). [CrossRef]
31. D. D. Dominguez, A. W. Snow, J. S. Shirk, and R. G. S. Pong, “Polyethyleneoxide-capped phthalocyanines: limiting phthalocyanine aggregation to dimer formation,” J. Porphyr. Phthalocyanines 5(07), 582–592 (2001). [CrossRef]
32. M. Hanack, T. Schneider, M. Barthel, J. S. Shirk, S. R. Flom, and R. G. S. Pong, “Indium phthalocyanines and naphthalocyanines for optical limiting,” Coord. Chem. Rev. 219, 235–258 (2001). [CrossRef]
33. J. S. Shirk, J. R. Lindle, S. R. Flom, F. J. Bartoli, and A. W. Snow, “3rd-Order Nonlinear Optical-Properties of Phthalocyanines,” in Electrical, Optical, and Magnetic Properties of Organic Solid State Materials, L. Y. Chiang, A. F. Garito, and D. J. Sandman, eds. (Mater. Res. Soc.Symp. Proc. 247, Pittsburgh, PA, 1992), pp. 197–201.
34. 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]
35. M. J. Soileau, T. H. Wei, M. Sheikbahae, D. J. Hagan, M. Sence, and E. W. Vanstryland, “Nonlinear optical characterization of organic materials,” Mol. Cryst. Liq. Cryst. (Phila. Pa.) 207(1), 97–101 (1991). [CrossRef]
36. M. Sheik-Bahae, A. A. Said, T. H. Wei, D. J. Hagan, and E. W. VanStryland, “Sensitive measurement of optical nonlinearities using a single beam,” IEEE J. Quantum Electron. 26(4), 760–769 (1990). [CrossRef]
37. B. L. Justus, A. L. Huston, and A. J. Campillo, “Broad-band thermal optical limiter,” Appl. Phys. Lett. 63(11), 1483–1485 (1993). [CrossRef]
38. 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]