Triad-sequestered Polymer Encapsulated Reverse Micelle Composite Materials as Optically Addressable Magnetic Field Sensors
Our group has developed novel Donor-Chromophore-Acceptor (D-C2+-A2+, or “triad”, electron transfer species with theoriginal goal of mimicking the first two steps of photosynthesis[ 1-6]. Fortuitously, the relaxation processes of the excited states produced after triad photoexcitation are spin-dependent; therefore, their rates can be affected by magnetic fields. . These D-C2+-A2+ supramolecular assemblies can thus, in principle, be used to optically detect magnetic fields by monitoring the triad’s change in absorbance (ΔA) over time after pulsed laser excitation. However, some restrictions are imposed upon the sensing system. First, the molecular sensor must be in fluid media in order to generate the magneticfield-sensitive photo- excited states, and consequently elicit field responses. Second, such fluid media must be devoid of oxygen in order to prevent decomposition of photoexcited states in the simultaneous presence of light and oxygen. The work herein employs polymer and surfactant material synthesis strategies as well as our knowledge of the D-C2+-A2+ systems in order to engineer composite sensors which give optical responses to externally-applied magnetic fields.
The general D-C2+-A2+ triad structure is presented in Figure 1 along with the electron transfer processes (Figure 1). Because of the intense optical transitions in the 2+ oxidation state and the inert nature of its ligands, the tris(bipyridine)ruthenium( II) moiety is the chromophore in the investigated D-C2+-A2+.
Figure 1. General electron transfer steps of a Donor-Chromophore Acceptor
(D-C2+-A2+) [Ru(II)(n-PXZ)2(4-DQ2+)](Anion-)4 supramolecular triad assembly including an abridged energy diagram of the corresponding photoexcited states. The triad contains two pendant phenoxazine moieties (where X is oxygen) which both serve as electron donors, a tris(bipyridine)ruthenium(II) core chromophore, and a diquaternary bipyridine (termed a “diquat”) electron acceptor. Upon irradiation with light, the D-C2+-A2+ species is eventually photoexcited to a Charge Separated State (D+•-C2+-A+•, CSS) which can only be formed in fluid solution. The fluid within the encapsulated micelles provides this medium in the PERMC. The PERMC network polymer nonpolar phase protects the photogenerated CSS from reaction with ambient oxygen and serves as a rugged housing for the solid state optical magnetic field transducer.
The low-spin tris(bipyridine)ruthenium(II) core chromophore (C2+) ground state is excited to a singlet metal-to-ligand- charge-transfer state (1MLCT) by 450 nm light. Thechromophore has a relatively large emission quantum yield and a long-lived excited state; therefore, it can undergo reactions in the excited state. This 1MLCT intersystem crosses to a triplet metal-to-ligand-charge-transfer state (3MLCT) in less than one picosecond.. An N,N′-diquaternary-2,2′-bipyridine electron acceptor (A2+) moeity, covalently attached to one of the chromophore’s bipyridine ligands, accepts this photoexcited[ 3].MLCT electron from the chromophore (k1), consequently generating a triplet charge transfer state (3CTS, D-C3+-A+•). [1-6].One of the azine moieties then donates an electron to the oxidized chromophore in the CTS (k2) to yield an optically detectable triplet Charge Separated State (3CSS, D+•-C2+-A+•). The quantum yield of triad CSSs are on the order of 90%. Relaxation of the triplet.CSS back to the singlet ground state is spin-forbidden. The normal CSS lifetime (τ CSS ) in bulk solution is 100 to 300 ns depending on the solvent with the τ CSS values being larger for solvents of lower dielectric constants.
The magnetic field sensing behavior of the triad is a direct result of the Zeeman splitting of the CSS. In the absence of afield, theCSS and CSS are essentially degenerate; consequently, hyperfine coupling is sufficient to mix the sates. Upon the application of a magnetic field, the degeneracy of the T+ and T- states with .CSS (and T0) is broken and their mixing becomes progressively less efficient as the field strengthens. Therefore, the decay of the T+ and T–components of the . CSS becomes progressively slower, while that of the T0 is unchanged. Therefore, this two-component .CSS relaxation (T0 and T±) pathway corresponds to a biexponential decay of the optically detectable.CSS under applied magnetic fields constituting a magnetic field effect (MFE). This MFE can be quantified by fitting the change in absorbance (ΔA) of the CSSat either 388 nm (the absorption maximum of the reduced acceptor A+• in the CSS) or at 520 nm (the absorption maximum of the oxidized donor D+• in the CSS) to a biexponential decay for different field strengths. For bulk degassed fluid solutions of triads in modest external magnetic fields (>50 mT which is approximately the field strength experienced at the read head of a standard computer drive), the long component of the τCSS increases by ca. ten-fold to greater than seven microseconds with respect to that of no applied field[7,9]. At fields of > 500 mT, the MFE saturates as the field independent pathway k4 (which likely invokes spin-orbit coupling) becomes rate limiting.
Our challenge was to devise a composite material that would maintain a solution environment about the triad so that the .CSS could form, while simultaneously housing the triad in a rugged oxygen-impermeable solid matrix. The refinement of polymer encapsulated reverse micelle materials synthesis strategies was undertaken in order to create an organic solid state optical sensor for the detection of magnetic fields. These types of materials were first pioneered by Menger et al. and Zhu et al[10-12].The resultant opaque materials were ground into powders for use in chromatographic separations as actik3quality materials are essential for the MFE sensing application presented herein.Accordingly, synthesis strategies were devised which yielded high optical quality polymer encapsulated reverse micelle composites (PERMCs).[13-14].Generally, a known concentration of triad in a polar solvent is injected in microliter amounts into an Aerosol-OT (AOT) solution with a 2:1 styrene:divinylbenzene nonpolar phase. After the addition of a radical polymerization initiator, the polymerization of the nonpolar phase of the composite precursor reverse micellar solution (CPRMS) proceeds in stages through being a viscous solution to a gel to a hard clear PERMC.
The work described herein was initiated with three objectives. The first objective is to demonstrate the formation of micelles in the CPRMSs and their continued existence after the polymerization of the CPRMS nonpolar phase. To this end, reports of dynamic light scattering (DLS) and tapping-mode atomic force microscopy (TM-AFM) studies on CPRMSs and PERMCs, respectively, are presented. The second objective is to demonstrate .CSS formation within these materials, and thus toconfirm the sequestration of the triads in fluid domains. The last objective is to investigate the optical responses of these materials to externally applied magnetic fields.
Materials. The preparation, purification, and isolation of the D-C2+-A2++ [Ru(4-POZ)2(4-DQ2+)](NO3-)4 triad (T) where n = 4, X = oxygen atom, and Anion = NO3-, the 4-POZ, and the 4-DQ2+ ligands is similar to the preparation of the X = S analog reported previously . Activated neutral alumina, 98% sodium bis(2-ethylhexyl)sulfosuccinate (Aerosol-OT, or AOT), 99% styrene and technical grade 80% divinylbenzene (with 4-ethylstyrene as the major contaminant) were purchased from Aldrich.
(b) Polymerizations were performed by adding ca. 1.0 wt. % ADPN radical initiator to CPRMSs and then exposing the CPRMSs to 38 oC for ca. 96 hours. The final nonpolar phase was poly(styrene-co-divinylbenzene).
(c) Std1 is neither a CPRMS nor a PERMC
(d) The nonpolar phase of CPRMS Std2 was not polymerized.
Table 1. Constitutions of Donor-Chromophore-Acceptor (D-C2+-A2+) triad-sequestered CPRMSs and PERMCs.
Dehibit 200 was purchased from Polysciences Inc. and was used as received. Radical initiator 2,2′-azobis(2,4-dimethylpentane)
nitrile (ADPN) was kindly provided by Professor Marc M. Greenberg (Johns Hopkins University).
Composite Precursor Reverse Micellar Solution Preparation
All PERMC materials synthesis was done at room temperature in a darkroom in order to avoid triad decomposition in the simultaneous presence of both light and air.
Styrene and divinylbenzene were mixed in a 2:1 ratio to prepare the 25% divinylbenzene nonpolar phase (based on the 80% purity of divinylbenzene). Immediately before PERMC preparation, the nonpolar phase was de-inhibited by sequential elution through separate columns of Dehibit-200 and neutral alumina. To prepare a 50.00 mM AOT stock solution, 0.5660 g of wet AOT (ca 0.46% H2O) (1.00 x 10-4 mol) was added to a 25.00 mL volumetric flask, which was then filled with the 2:1 styrene:divinylbenzene mixture. A 20.00 mM T stock solution was prepared using doubly deionized water. This served as the polar phase for the CPRMSs and the PERMCs.
Using a glass syringe, three 2.00 mL aliquots of the 50.00 mM AOT solution were dispensed into three (15 x 40) mm 4.0 mL Fisherbrand cylindrical screw-cap vials. Volumes of 18.0, 27.0, 36.0 μL of 20.00 mM T were then injected into these solutionsusing a 100 μL Gastight 1710 glass syringe. This produced CPRMSs ith W0 values of 10.00, 15.00, and 20.00 respectively (S1-3) (Table I). The tightly-capped CPRMSs were agitated on a Vari-Whirl mixer until clear. Sonication often caused clouding of the CPRMSs and was not used. Small volumes of S1-3 were stored in a dark room for elevated temperature DLS particle sizing studies. Approximately 1.0 wt. % ADPN radical initiator (0.010 g, 5.84 x 10-5 mol) was added as a solid into the remaining 1.80 mL CPRMSs (S1-3).
This initiator was chosen because of its fast decomposition rate at lower temperatures.The screw caps were then replaced
with 11 x 17 mm rubber septa. Next, S1-3 were purged with a slow steady stream of nitrogen. This both facilitated polymerization of the nonpolar phase and ensured that there was no oxygen present which could interfere in time-resolved CSS measurements. Finally, S1-3 were placed in a sand bath for thermal polymerization to effect PERMC sensors C1-3 with corresponding W0 values (vide infra).
Two different solutions of T were prepared in order to determine the respective values for comparison to those of the PERMCs C1-3. A water solution of T was prepared whose absorbance at 450 nm was adjusted by dilution to be 0.225 (SolnI). A CPRMS consisting of 2.00 mL of 50.00 mM solution of AOT in benzene as the nonpolar phase and 27.00 μL of 20.00 mM T with doubly deionized water (W0 = 15.00) as the polar phase was also prepared (Soln II). Both solution cells were freeze pump thawed (FPT) degassed four times then backfilled with nitrogen.
Composite Precursor Reverse Micellar Solution (CPRMS) Polymerization and Polymer Encapsulated Reverse Micelle Composite(PERMC) (C1-3) Finishing. A heating mantle filled with sand and set to 37 °C using a Variac was used to cure the nonpolar phases of the CPRMSs S1-3 for 96 hours. After scoring vials laterally around the entire circumference, the vials were broken in order to retrieve the resulting solid cylindrical PERMCs C1-3. A Buehler Variable Speed polishing wheel equipped with progressively finer grit sandpaper was used to convert the cylindrical PERMCs into orthorhombic PERMCs. Typical PERMC dimensions were on the order of (0.25 x 0.75 x 1.0)cm3. Smooth, homogeneous, non-light scattering monolith surfaces suitable for optical studies were obtained by polishing composites with a slurry of 0.3 micron Micropolish II (Buehler). Finer PERMCS surfaces were obtained by using MasterMet 2 polish (Buehler) in tandem with Microcloth polishing pads (Buehler)
Composite Precursor Reverse Micellar Solution (CPRMS) Particle Sizing. Dynamic Light Scattering (DLS) was first initiated at 25.0 °C using a Dyna-Pro MSTC particle sizing instrument. The difference in the refractive indices of the nonpolar and polar phases, at 1.5470 and 1.000 respectively, justifies the use of DLS to determine micellar radii.Astyrene blank was chosen in place of the actual 2:1 styrene:divinylbenzene nonpolar phase, because of the similarity in refractive indices (at 1.5470 and 1.5740, respectively).Particulate matter was removed by filtering 60.0 μL of CPRMSs S1-3 through a Teflon- construct 0.02 μm Whatman Anodisc 13 membrane filter four times. For each micellar aliquot S1-3,ten particle size radius readings were collected, averaged, and then reported as a mean radius value with accompanying error bars corresponding to one standard deviation. Elevated temperature particle
sizing studies were conducted for S2. After collecting five size measurements at 25 °C, the temperature of the CPRMS in the cuvette was increased at a rate of 1 degree per 30 seconds to a maximum temperature of 42 °C. This temperature is 5 °C higher than the polymerization temperature of 37 °C. No incident polymerization of the CPRMSs occurred during the experimental time frame.
Atomic Force Microscopy (AFM) on poly(styrene-co-divinylbenzene) and Cryo-Cleaved Polymer Encapsulated Reverse Micelle Composites (PERMCs). Cryo-cut fractography was employed to expose an interior poly(styrene-co-divinylbenzene) or PERMC surface for AFM analysis. A blank of poly(- styrene-co-divinylbenzene) (Cut I) was radically-polymerized with ADPN at 37 °C and ground to rectangular dimensions. A tris(bipyridine)ruthenium(II) chloride [Ru(bpy)3]Cl2 (C) chromophore- sequestered composite (Cut II) was substituted in order to reserve the triad-sequestered composites (C1-3) exclusively for optical studies. First, Cut I-II were cooled in an ice bath to 0 °C, then slowly introduced into a dry ice/acetone bath (-72 °C). Following, Cut I-II were carefully introduced into a dewer containing liquid nitrogen (-196 °C) for one minute before being transferred to a styrofoam cup containing a smaller volume liquid nitrogen sufficient to immerse each. Cut I-II were each fractured by striking an Exacto-Knife blade which was already in contact with the liquid nitrogen. The conchoidal-shaped surface of each sample was then covered with Parafilm, and the opposing surface was ground flat for the AFM analysis.The Parafilm was removed and the samples were placed in a Branson ultrasonic cleaner containing methanol for ten minutes. Finally, they were blown dry with compressed air.
Both samples were imaged under ambient air conditions. Samples were mounted on a stainless steel stub using double-sided tape. All samples were featureless to examination by a one thousand magnification optical microscope.Height and Phase Imaging under TM-AFM conditions were performed using a ommercial atomic force microscope (Nanoscope IIIa, Digital Instruments) operated in repulsive mode. A commercial uncoated square-pyramidal-shaped silicon tip, with a radius of curvature f less than 10 nm, adjunct to a 230 ± 5 μm long silicon cantilever of reported spring constant 40 ± 15 N/m (NSC- 16 Ultrasharp Cantilevers and Gratings) was used to image cryo-cut B1-2 at low amplitudes (Ao) in constant amplitude mode. The cantilever oscillated at a drive frequency of 170.00 ± 10.00 kHz with an amplitude setpoint of 1.8 ± 0.2 V. The tip was scanned across the surface with a frequency of 2.0 Hz. Such settings ensured that no sample degradation occurred during the imaging process. Scans were repeated at lower frequencies (1.0 Hz) in order to ensure the image reproduciblity. Obtained image scan sizes on the order of (256 x 256) pixels ranged from (4.0 x 4.0) μm2 to (300 x 300) nm2 dimensions. The latter domain was achieved by preferentially scanning ridge-like hillock features exhibited by the cryo-cut surfacesthat were visualized on a (1.00 x 1.00) μm scale. A scan of one sheet of mica was collected for comparison to Cut I-II. All images shown are the raw experimental data without any image processing except flattening. Image Histograms were examined along the vertical axis where no in-plane shearing effects occur. These appear in the Supporting Information. This TMAFM pore distribution was compared to a theoretical micelle Poissonian distribution calculation based on both stoichiometry, known surfactant aggregation numbers, and micellar radii measured by DLS at 25 °C.
Transient Absorption Determinations of the Charge Separated State Rate of Decay. The laser system used for time-resolved transient absorption spectroscopy is the same as in previous reports and is as follows (S.I.1)[7,14].The composites weremounted on a stainless steel Thorlabs post using an adhesive. The output of a 30.0 Hz Quanta Ray 150-5 Nd:YAG laser was operated at its third harmonics (355 nm) as the pump laser (S.I.1). The laser pulse width was 5.0 to 7.0 ns. This source was coupled to a methanolic Coumarin 47 dye laser operated at 1.0 Hz which provided a 460 nm excitation pump beam. The dye laser’s pulse energy was adjusted to be between three to five millijoules by controlling the power source and by using a 30 cm focal length defocusing lens. Composite degradation and charring from surface heating at the pump beam incident spot did not occur at such low pump beam powers. A pulsed Oriel 75 W Xenon arc lamp served as the probe beam. Theprobe beam was focused by a 20 cm focal length lens. The probe beam approached mounted PERMCs at an angle of ca. 2 degrees with respect to the pumb beam (approximately collinear). Both pump and probe beam paths were incident surface- normal upon the (0.5 x 1.0)cm2 face of the monolith. The pump and probe beams eventually intersected in the interior of the mounted PERMC. At the point of intersection, the probe beam had a cross-section of ca. 5 mm2. The intersection of the beams was optimized in order to enhance the signal to noise ofthe CSS decays. The pump beam excited the sequestered triads in the PERMCs and the probe beam monitored the triads’absorption of that light over time. The decay of the transient absorption coming from the opposite PERMC face was firstcollimated then focused through a Melles-Griot 370 nm high band pass filter onto a Jarrell Ash model 82-410 monochromatorwhich selected 388 nm light. This wavelength of light corresponded to the absorption maximum of the A+• in the CSS. Light was collected using a Hamamatsu model R2496 photomultiplier tube set to 500 V. The current output of the photomultiplier tube was displayed electronically on a Tektronix TDS602B digital oscilloscope. A Thorlabs DET310 photodiode
was used to trigger and synchronize the oscilloscope with the pump beam pulses.
For magnetic field dependent measurements, PERMCs (C1- 3) and Solutions (Soln I-II) were placed between the poles of a Bruker B-E 15 electromagnet fitted with tapered pole caps and having a 1.2 cm air gap. The external magnetic field was applied perpendicular to the optical path along the 0.5 x 1.0 cm axis of the PERMCs. The magnetic field strength was measured by an F.W. Bell STB1-0404 transverse Hall Probe. For experiments conducted at 0 T, the residual field was cancelled by reversing the current through the electromagnet. Baseline corrections and transient absorption decay fits to single exponential and double exponential decay functions with plotted residuals were completed. For presentational purposes, the data has been smoothed using SigmaPlot 2000. All transient absorption decays are presented as Delta Absorbance (ΔA) versus time and normalized to ΔA(t=0).
Preparation of Triad Sequestered-Polymer Encapsulated Reverse Micelle Composite Materials. The nonpolar and polar phase components used to prepare magnetic field sensing PERMC materials are selectively chosen. Styrene:divinylbenzene was selected as the polymerizable CPRMS nonpolar phase. The relatively large 25 vol. % of divinylbenzene concentration in the copolymer feed ensured the formation of a highly crosslinked network polymer nonpolar phase which would exclude oxygen from the composite. This is necessary in order to avoid the decomposition of the triad under both light irradiation and oxygen . Water was chosen for the polar phase because it is immiscible with the nonpolar phase, CSSs were shown to form in bulk water solution, and the solubility of the nitrate salt of the triad.
The surfactant, triad, and radical initiator, as well as their respective concentrations, are also specific to PERMC materials synthesis (Table I). Aerosol-OT (AOT) was the surfactant used n forming all PERMCs. Aerosol-OT forms especially stable reverse micelles as a result of its efficient hydrophobic packing structure[19,20]. The 50.00 mM AOT concentration employed
herein is greater than its reported critical micelle concentration (cmc) of 9.00 mM..At surfactant concentrations larger than 50 mM, the optical quality of the PERMCs decreases. At higher surfactant concentrations, micelles can form aggregates with diameters above that of the light scattering limit (vide infra). Moreover, it is postulated that there is a greater chance for the termination of growing polymer chains by the presence of foreign species.The premature termination of the polymer chains likely causes regions of microcrystallites which packbetter and thus scatter light more efficiently. The specific triad used in these studies corresponds to the structure in Figure1 where n = 4, X= oxygen, and the anion is nitrate (T). Triads incorporating phenothiazine donors were not elected for thisstudy, because of their propensity to open the spin-forbiddenpathway via spin-orbit coupling. Triads containing phenoxazine donors (where X = O and not S) exhibit larger maximum CSS lifetimes. The 20.00 mM triad concentration employed herein follows from previous reports [13-14]. According to Poissonian statistics, essentially no reverse micelles contain more than one triad at these concentrations; thus, no interspecies electron transfer processes can occur between two triads (S.I.2). Thomas et al. and Joselevich et al., among others,have invoked Poissonian statistics in their respective work with micellar systems[22,23]As a result, typical bulk solution conditions are simulated with PERMCs, in so far as there are no interspecies electron transfer pathways for triads at concentrations having similar optical absorbance. The three different 20.00 mM T polar phase volumes of (9.00, 18.00, and 27.00 μL) njected into 2.00 mL nonpolar phase volumes correspond to W0 values of 10.00, 15.00, and 20.00 (Table I). The W0 values can be directly calculated from the following expression where Vplr is the volume of the polar phase, 55.55 M is the concentration of water, Vtot is the total volume of the CPRMS, and cAOT is the concentration of AOT 24:
Thermally assisted radical atactic polymerizations of CPRMS S1-3 nonpolar phases ensued after the addition of approximately 1.0 weight percent ADPN radical initiator. The nonpolar phases of CPRMSs S1-3 polymerize after heating to 37 °C for 96 hours yielding corresponding PERMCs C1-3. Too high of a polymerization temperature will affect the vapor pressures of the CPRMS component species and thus the W0 of the composite. It has already been established by Lang et al. that an exchange of micellar material occurs and that this exchange rate increases with increasing temperature25-26. Moreover, CPRMSs polymerized at higher polymerization temperatures are of poorer optical quality than those polymerized at lower temperatures. The ADPN radical initiator has a much lower decomposition temperature than the more common initiators like 2,2’-azobis(isobutyronitrile) (AIBN) or benzoyl peroxide (BPO) which serves to ensure that micelles are still intact during the polymerization.27-28 At the lower polymerization temperature (37oC), the micelles have fewer intermicellar collisions which can result in light scattering aggregates (vide infra); thus, the optical quality of ADPN-initiated PERMCs are usually much better than that of AIBN- and BPO-initiated PERMCs. The resultant PERMCs are of high optical quality in that they are transparent and non-light scattering (S.I.3).
Dynamic Light Scattering (DLS) Particle Sizing on Composite Precursor Reverse Micellar Solutions (CPRMSs). Dynamic Light Scattering (DLS) particle sizing analysis was performed on CPRMSs S1-3 before the addition of ADPN in order to confirm both that micelles were intact in the relatively higher dielectric nonpolar phase of the styrene:divinylbenzene copolymer feed (compared to isooctane), and that these micelles sequestered triads. Moreover, DLS particle sizing of CPRMSs was conducted at 25.0 oC and at 37.0 oC (the temperature of the thermally assisted radical initiated nonpolar phase polymerization). The elevated temperature experiments were performed in order to establish that the micelles remain intact at 37 oC. Ten micellar radius readings were collected for reverse micellar solutions S1-3. These readings were averaged for each CPRMS in order to report a mean micellar radius (Figure 3a). The error bars for these individual W0-dependent measurements represent one standard deviation from the mean.
The radii of the reverse micelles in the CPRMSs were analyzed in order to determine if the polar phase and the triad were sequestered. An AOT micelle with a W0 value of zero has a measured radius of 1.8 nm with the size due to the hydrophobic tails of AOT which have a different refractive index than that of the nonpolar solvent.19,29-30 This lower limit is represented as the dashed horizontal line (Figure 3). The AOT micelles ontaining only water with W0 values in the range of 10 to 20were determined to have average radii of 3.2 ± 0.4 nm. Thus, according to the larger micellar radii and polydispersities reported in Figure 3 for the CPRMSs, the triads are sequestered in the polar phase located within the interiors of the reverse micelles. This result is significant, because it establishes that the triads can still be in solution environment once the nonpolar phases of the CPRMSs are polymerized. This is consistent with the observation that photogenerated CSSs are observed for the corresponding PERMCs.
There is only a modest increase in micelle size as W0 increases at 25 oC. This indicates that there is a large dispersion in size and or shape of the micelles within the CPRMSs. Nevertheless, it has to be true that more water, on average, is containedwithin micelles with larger W0 values. Thus, it is reasonable to assume that the average triad environment is more “bulk-like” as W0 increases.
The polydispersities, or standard deviations, of the micelle radii can be interpreted with respect to the statistical population of the triads in the reverse micelles and perturbations to the micellar morphology. The polydispersities are ca. 50% of theaverage micellar radii for each CPRMS. The PERMCs (C1-3), and thus also the CPRMSs (S1-3), are prepared such that no one micelle contains more than one triad molecule. According to Poissonian statistical calculations, ca. 10% of the micelles sequestertriads and water polar phase with the remaining 90% sequestering only water. The polydisperities of the CPRMSscontaining triads or simple [Ru(bpy)3]Cl2 is nonthethesss much larger than CPRMSs containing only water. Strict bimodal distribution of the micelle sizes can be observed during the DLS experiment due to instrumental limitations. Yet, it isplausible to conclude that the apparent radii of the reverse micelles sequestering the nitrate salt of the water solvated triadand the sizes of the reverse micelles sequestering only water are so disparate owing to the large apparent size polydispersities.The long axes of the solvated triad and the each of the four nitrate counteranions are 2.0 and 0.4 nm, respectively.The positivefour charge on the triad and the four counteranions likely require solvation by water inside a reverse micelle. Therefore,the amount of water in the reverse micelle containing a nitrate salt of the triad would be considerably larger than the 4.0 nmaverage value as measured by DLS. In broader terms, the largepolydispersities are somewhat irrelevant for PERMC magnetic field sensing applications as long as the micelles sequester triads n environments that simulate that of bulk solution
A series of dynamic light scattering particle sizing experiments were also conducted (in this case on [Ru(bpy)3]Cl2) in order to establish the effect of a polar probe concentration on the micelle size and polydispersity. A series of [Ru(bpy)3]Cl2 chromophore- sequestered CPRMSs were prepared. The CPRMSs all had 2:1 styrene:divinylbenzene nonpolar phases, water polar phases, W0 values of 10, and [Ru(bpy)3]Cl2 concentrations of 1.0, 3.0, 5.0, 10.0, and 40.0 mM (S. I. 4). Over the range of samples examined, the size of the reverse micelle and the polydispersity both increase as the polar probe concentration increases. The particles of radius > 50 nm (which are only detected for CPRMSs of larger probe concentrations) are likely oligomeric species in the deinhibited nonpolar phase.
The second possibility for the apparent large size polydispersities is plausibly due to micellar asphericalness which is a direct result of the electrostatic, hydrophobic, dipolar, and cationic-π interactions between the triad and the anionic AOT surfactant species.The relatively hydrophobic donor moieties could possibly embed into the hydrophobic tail portions of the AOT surfactants composing the micelle (Figure 2). Likewise, both the charged acceptor and chromophore moieties could intercalate between AOT head groups decreasing electrostatic repulsions between adjacent AOT molecules. In fact, Stern layer- embedding of divalent complexes has been commonly cited for divalent ruthenium complexes sequestered in anionic surfactant micelles of water polar phase[32-35]Specifically,Hubig et al.postulated that viologen moieties form strong ion contact pairs with the sulfosuccinate head groups in AOT reverse micelles.[ 36].The diquaternary bipyridine moieties are similar in nature to viologens. Through any of these interactions, asphericalness could be induced in the micelles producing the apparent polydispersity.
Elevated temperature DLS measurements were conducted to ensure that the AOT reverse micelles in the CPRMSs were still intact at the temperature whereupon the homologous cleavage of ADPN generates enough radicals to initiate polymerization[ 16,27-28].It is known that micelles deaggregate in response to temperature increases. Accordingly, one series of these temperature-dependent micellar radii meausurements for CPRMS S2 is presented herein.
Figure 2. Schematic cross-sectional representations of a composite precursor reverse micellar solution (CPRMS) and a polymer encapsulated reverse micelle composite material (PERMC). The chemicalstructure of the Aerosol-OT (AOT) surfactant molecule composing the micelles in the CPRMSs and PERMCs is represented by the crosshatcheddrawing (upper center). The 2:1 styrene:divinylbenzene nonpolar phase of the CPRMS (bottom left) is polymerized after theaddition of 2,2’-azobis(2,4-dimethylpentanenitrile) (ADPN). The polymerization front proceeds from northwest to southeast leaving themicelle(s) and the sequestered polar phases intact inside the PERMC (bottom right).
After collecting five DLS measurements at 25 °C, the temperature of S2 was increased above the CPRMS nonpolar phase polymerization temperature of 37 °C. These elevated temperature studies confirm that micelles are intact at the polymerization
temperature with accompanying polydispersities which are explainable according to the above reasons (Figure3b). The micellar radii are larger by a factor of 10% over the 17 °C temperature increase. The measured micellar radii increasescoincident with temperature increases are permanent, a priori, once the CPRMSs S1-3 are transformed into PERMCsC1-3. In conclusion, the CPRMSs can be polymerized with the confidence that the nanoscopic fluid micelles which sequestertriads will still be intact in the network polymeric nonpolar phase of corresponding PERMCs.
Figure 3. Variable-W0 and -temperature DLS measurements taken on [Ru(II)(4-POZ)2(4-DQ2+)](NO3)4 T1 triad-sequestered CPRMSs. The averaged micellar radii are plotted against W0 (A). The W0 valuehas no effect on the averaged radius, but the extent of the fluid polar phase chemical environment likely increases as the W0 increases (i.e. the reverse micellar interior becomes more fluid in nature as W0 increases). The measured radii and polydispersities are larger thanthat of an AOT reverse micellar solution prepared to a W0 value of zero (hashmark line),indicating that the triads are sequestered insidethe micelles. The large polydispersities are due to the bimodal micellar population (the two different populations of micelles which containor do not contain triads) and the asphericalness induced in the reverse micelles as a result of the electrostatic and hydrophobic interactionsbetween the sequestered triads and the anionic surfactantspecies. Th effect of polymerization temperature on the micellar radii of S2 before nonpolar phase polymerization (B). The micellesremain intact at the nonpolar phase polymerization temperature.
Atomic Force Microscopy (AFM) on Cryo-Cut Poly(styrene- co-divinylbenzene and composites.To ensure that reversemicelles are intact within the PERMCs, efforts were made to image the reverse micellar imprints. First, in an effort to reproducethe surface features reported by Menger et al, 100 nm thick microtomed PERMCs were examined by Scanning Electron icroscopy (SEM).However, no pore structure, on the order of 10 nm diameters (as indicated by DLS studies of CPRMSs), was observed presumably because of microtoming artifacts (S.I. 5)[10,14].The PERMCs analyzed by this methodmay also have been damaged by the 1.2 kV accelerating voltage of the scanning electron beam. Moreover,any detectedpore structure may not have been resolved because of insulative charging effects. Tapping Mode-Atomic Force Microscopy(TM-AFM), on the other hand, produces less surface damage and directly images strictly surface features. As such, TM-AFMis particularly amenable to imaging surface features of softer samples like PERMCs.
TM-AFM was used to image two types of surfaces: poly(styrene- co-divinylbenzene) (Cut I) and a composite made from the nonpolar phase polymerization of a CPRMS of 50 mM AOT in 2.0 mL styrene:divinylbenzene to which 18.0 μL of 20.00 mMtris(bipyridine)ruthenium(II) chloride ([Ru(bpy)3]Cl2) chromophore (C) was injected (Cut II) (Table I). We postulate that triad (T) and chromophore (C) are similar enough that conclusions made on height and phase images of one are applicable to the other. Both Cut I and and Cut II were prepared for TMAFM analysis by established cryo-cutting techniques[38-39].A photograph of the cryo-cut samples is presented in the supplemental aterial (S.I.6).To our knowledge, the present work is the first to report TM-AFM height and phase imaging of polymer encapsulated reverse micellar composites retaining micellar imprints.
The TM-AFM height and phase images of Cut I reflect the polymer morphology.The 30 nm relief exhibited by the corresponding cryo-cut Cut I is on the same order as previously reported samples (Figure 4a).A phase image, which measures oscillating tip energy dissipation, Δ⏀, was collected on Cut I. The variation in Δ⏀ is known to indicate differences in polymer morphology over a sample surface.
Figure 4. Height and Phase TM-AFM images of cryo-cut poly(styrene- co-divinylbenzene) (Cut I) and a chromophore-sequesteredPERMC (Cut II). The TM-AFM images presented are as follows: height (A-left) and phase (A-right) images of cryo-cut poly(styrene-co-divinylbenzene) Cut I over a (1000 x 1000)nm2 area; height (B-left) and hase (B-right) images of a chromophore-sequestered PERMC over a (300 x 300)nm2 area; and a height image of a sheet of mica withthe exact same scale bar used for B. The cryo-cut cleavage of Cut I is rougher than Cut II on both the height and phase images.
In Cut I, such morphology disparities are assumed to be due to the preferential polymerization of divinylbenzene from the nonpolar phase copolymer feed before that of styrene, according to their reactivity ratios of 2.60 and 1.18, respectively[42,43].This copolymerization generates harder, thermoset branch point attributable to highly-crosslinked regions appearing as darker regions in the phase image. These regions dissipate less oscillating tip energy.The branch points phase contrast well to the softer, more energy-dissipating, polystyrene- rich domains of the phase image appearing as lighter regions in the phase image..It is postulated that the highly- crosslinked regions of the copolymer are also of higher relief than the polystyrene-rich domains because the former have more crosslinks (bonds) into the cryo-cut composite than styrene-rich regions.As a result, it is harder to force the mechanical cleavage of these branch points, and a height image of large relief is obtained. Moreover, the thermosets are spaced on average 30 nm away from each other in the phase image. This distance directly corresponds to reported height relief scale of Cut I.
Successive TM-AFM images were obtained at different scan rates and different areas of the same cryo-cut polymer Cut I, as well as on other preparations of the sample. The images obtained all yielded consistent results. The height image is also provided for a 1 x 1 μm dimension in order to demonstrate that the image qualities hold throughout a larger scanned area (S.I.7). It is unlikely that surface tension effects are responsible for the phase image patterns observed, because there is no substantial difference in surface tension between liquid styrene and divinylbenzene (they are miscible).In conclusion, the height and phase images obtained from TM-AFM analyses of cryo-cut poly(styrene-co-divinylbenzene) are characteristic of the sample.
The height TM-AFM image of Cut I is markedly different from that observed for the cryo-cut image of PERMC Cut II. The latter has a height relief scale of only 5.0 nm (Figure 4b). The presence of polymer encapsulated reverse micelles is plausibly responsible for these markedly different height and phase relief scales. The height image of Cut II exhibits imprints of reverse micelles appearing as dark aspherical shaped holes of lower relief randomly spaced throughout the image. It is postulated that the applied cryo-cut force causes fractures along a so-called micellar plane, or a “micelle-to-micelle” cleavage plane, wherein encapsulated micelles hemispherically fracture upon impact (S. I. 8). The frozen fluid sequestered inside reverse micelles is anticipated to be of lower mechanical strength than the nonpolar phase network polymer. Also, to the extent that water within the micelles behaves as bulk water, it should expand upon freezing and subsequent stressing of the surrounding polymeric nonpolar phase. Consequently, the micelles should be the first to fracture upon cryo-cut impact. Micellar imprints of different depths are then exposed in the cryo-cut plane depending on the location of the cleavage on the encapsulated micelle spheroid. Imprint depressions on the order of 0.5 nm are reported, rather than median expected depression depth of 2.0 nm, according to DLS measurements performed at 37 °C. It is important to acknowledge that the radius of curvature of the tapping silicon tip used in the TM-AFM analysis is on the order of 10 nm; so it is not possible to entirely fit the tip into any one imprint in order to record larger hole depths. In addition to polydispersity contributions, tip artifacts, non-hemispherical cleavage of encapsulated micelles, micellar polydispersity, and the nature of the polymer-surfactant interface (with the coincident ultrasonic cleaning in an effort to dissolve all cyro-cut exposed surfactant and polar phase) may also be responsible for the aspherical shape of the indentations.Finally, the depressions are not likely derived from shards of polymer that were shattered and ejected during the cryo-cut impact, since such types of indentations would be more irregularly-shaped than those observed in height relief image of cryo-cut B2.
The phase TM-AFM image of Cut I is also markedly different from that observed for the cryo-cut image of PERMC Cut II. The phase image of Cut II has an enhanced resolution over the height image with respect to imaging the spherical nature of the micellar imprints. The resolution in the height image is reduced as a result of horizontal shearing due to the 2.0 Hz oscillating tip frequency and in-plane tip oscillations therefrom[ 47].More importantly, the phase image of Cut II shows only a one degree difference over the entire surface area (Figure 4b) indicating that the entire surface dissipates tip energy uniformly. Because the cryo-cut plane in Cut I is determined by the cleaved micelles and not the differences in polymer morphology,the exposed surface has the same surface morphology throughout the entire scanned area. The slight decreased energydissipation at the imprint (ca. 1 degree) is likely attributable to more of the tip coming into contact with the sample andthereby less dissipated oscillating tip energy. A topographical image of mica is also presented with the same topographicalscale used for Cut II (Figure 4c). In conclusion, surface roughness and the difference in sample surface morphologydecrease in the following order: Cut I > Cut II > mica sheet.
It is instructive to further compare these TM-AFM hole sizes in Cut II to the micellar sizes of CPRMS S2 at 37 oC as determined by DLS. The diameter of an average micellar imprint measured along the vertical non-shearing axis is between 4.0 nm and 5.0nm. This diameter is slightly less than the expected 6.0 nm micellar diameter from DLS measurements corresponding to the case where AOT was not removed during the ultrasonic cleaning (8.0 nm total micellar diameter minus ca. 2.0 nm from the AOT hydrophobic tails). It is possible that the polymer penetrates the hydrophobic tails of the AOT making them hard or impossible to remove by ultrasonic cleaning. Unfortunately, an XPS study was unable to establish the presence or absenceof sulfur in the cryo-cut PERMC, presumably because of the 0.17% S content net in the sample and the relatively low sensitivity of sulfer detection in these measurements. A statistically derived estimate of the average distance between micellar imprint centers was computed to be 10.0 nm, based on DLS data and Poissonian arguments. The distance between the micellar imprint boundries is on the order of 7.0 nm. In conclusion, the theoretically calculated micellar distribution, the experimentally- measured micellar radii by DLS on CPRMSs, and the experimental distances detected by TM-AFM on PERMCs are reasonable mutually consistent. Overall, there is topographical evidence confirming that polymer encapsulated reverse micelles are intact inside the solid composite. Quantitative root mean square analysis of depression depths are provided in the supplemental material (S.I. 9)
Time-Resolved Charge Separated State Transient Absorption Studies on Chemical Environment Effects for Polymer Encapsulated
Reverse Micelle Composite Materials Under No Applied Magnetic Field. The degrees of freedom experienced by the triad are expected to vary accordingly: bulk solution > CPRMS > PERMC. The frequency of enco nters between the D+• and A+• moieties of a photogenerated CSS (which are necessary for relaxation) will decrease as the triad interacts with the surfactants composing the micelle. Such interactions could presumably even involve the embedment of components of the triad into the micellar wall through electrostatic and hydrophobic interactions between the triad and the surfactant. The measured 3CSS lifetimes (τ CSS ) at 0 T decreases in the order PERMC > CPRMS > bulk solution, indicating a progressively less restrictive environment about the triad. For triad-sequestered CPRMSs and PERMCs, it is important to recognize that the τ CSS likely represents a distribution of chemical environments and thus rates of D+• and A+• collisional encounters. Ideally, in a field sensing application, it would be preferable if the τ CSS were the same in a CPRMS as in the bulk polar phase solvent itself, because this would maximize the frequency response of the field dependent kinetics. There is a small difference between the τ CSS of T1 in bulk water (Soln I) and the CPRMS (Soln II) under no applied magnetic field (89 and 93 ns, respectively, Figure 5), indicating that the average environment of the triad is similar in these two systems at least on the time scale of electron transfer, vide infra. The situation is significantly different for PERMCs
First, as has been demonstrated in previous work, in order for the CSS to form, the triad must be in a fluid environment. Thus, since the CSS formation is observed for T1 in PERMCs, at least some of the triads are in a fluid environment. Transient absorption spectra were recorded for C2 at six distinct times after the 450 nm photoexcitation of the PERMC. The presence of an absorption peak centered at 388 nm shows that a CSS forms and decays (S.I.10). The τ CSS for the PERMC under no applied magnetic field is a factor of ca. 15 greater than that of the τ CSS for Soln I-II. The lifetimes of intramolecular charge transfer states are known to increase in micellar media with respect to bulk solution conditions[48,49].Since, like CSS formation, CSS decay requires motion of the D+•and A+•moieties, it is postulated that the PERMC chemical environment affords fewer degrees of freedom to the sequestered triad.
Figure 5. The environmental effect on τCSS under no applied magnetic field for the [Ru(II)(4-POZ)2(4-DQ2+)](NO3)4 T1 triad in bulk solvent water (−), sequestered in a CPRMS (∙∙∙), and finally sequesteredinside a PERMC (∙∙∙). The CPRMS and the PERMC have the same W0 value. The τCSSincreases as the chemical environment about the triad becomes more restrictive. The restrictive environment limits the molecular motions of pendant D+• and A+• moieties which are necessaryfor CSS relaxation.
As the nonpolar phase of a CPRMS hardens, the surfactants presumably become essentially locked into place. Therefore, the contents of the entire micelle become less fluid, resulting in less frequent D+• and A+• encounters. Since all of the possible mechanisms of CSS relaxation involve relative motion of the two radical cations, it follows that the τ CSS should increase as well. In fact, the τ CSS of the phenoxazine donor-based triad-sequestered PERMC at 0 T has a τCSS that is on the same as the τ CSS for the triad in bulk acetonitrile solution with an applied field of 500 mT where the MFE saturates.Therefore, any PERMC-based MFE must produce a τ CSS which is longer than ca. one microsecond.
Magnetic Field Sensing Capabilities of Triad-sequestered Polymer Encapsulated Reverse Micelle Composite (PERMC) Materials. In order to obtain acceptable signal- to-noise in transient absorption measurements, composite sensors must be made with a W0 value greater than or equal to ten. Higher signal-tonoise ratios for the time-resolved transient absorption decays were sacrificed in order to ensure that the PERMC sensor did not burn in the 45 mW pump beam. The MFE of C2 is presented (Figure 6). Composites C1-C3 showed analogous MFEs (S. I. 11). In a qualitative display of the MFE, three low field (< 300 mT) and two high field (> 300 mT) CSS decays are plotted for C2 as the change in absorbance (ΔA) versus time at 388nm for the reduced acceptor moiety after the 450 nm photoexcitation pulse (Figure 6). The MFE manifests as a longer τ CSS with increasing field. There are two main features of the MFE that warrant discussion. First, the MFE is reversible; the τ CSS at 0 T is the same both before and after the external magnetic field is applied. Second, the CSS decay becomes modestly biexponential as the magnetic field strength increases above 00 mT. This biexponential quality is most evident at longer times (t > 2 μs; Figure 6). In bulk solution, the transition from an essentially monoexponential to biexponential decay for the CSS is very clear and well resolved. The relatively poor signal to noise for the decays in the PERMCs prohibits the same level of mathematical rigor in the data treatment. Nonetheless, the MFE is unmistakable at very modest fields (~20mT). Interestingly, in bulk acetonitrile solution, the MFE saturates at ca. 500 mT at a τ CSS value of one microsecond for T1. As pointed about above, the τ CSS in PERMCs is longer in no field than the MFE saturation value of τ CSS in bulk solution. The MFE of the PERMC environment appeals qualitatively to saturate a possibly at a slightly higher field (vide infra) with an average τ CSS that is grater than x longer than in bulk solution. The observations above are consistent with arguments made regarding the restrictive
mobility within the PERMC.
Figure 6. Raw (top) and smoothed (bottom) CSS transient absorption decays of T1 triad-sequestered PERMC C2 demonstrating an MFE. The CSS lifetimes increase and become more bi-exponential in qualityas field strength increases. The MFE saturates at 300 mT, whereupon the τCSS stops increasing with increasing field strength. This seriesof transient absorption decays attests to the practical nature of the
PERMC systems as optical magnetic field sensors. The CSS decays were monitored at 388 nm which is the absorption maximum of the reduced acceptor. Not all of the CSS decay curves are shown (with respect to Table II for figure clarity. The relatively poor signal-to-noise notwithstanding, the MFE following from the smoothed CSS decays are quantitatively represented by fitting the CSS exponential decay functions and plotting the obtained time constants as a function of magnetic field strength. Recall, from the treatment of bulk solution data, Zeeman splitting results in a biexponential decay of the CSS having a field independent component (To→ S pathway) and a slower component (T±→ S), the lifetime of which increases with field strengths up to saturation. The low signal to noise makes fitting transient data for the PERMCs to a biexponential function of questionable value. Consequently, these data were fit a to a single exponential in order to obtain an approximate average lifetime for the CSS. However, the plotted residuals obtained from monoexponential CSS decay fits were harmonic in nature, clearly indicating the multiexponential nature of the CSS decays. In contrast, residuals obtained from biexponentialCSS decay fits were not harmonic in nature. Therefore, both monoexponential and biexponential fits are presented (Table
Figure 7. Plot of CSS lifetimes ( τCSS) of T1 triad-sequestered PERMC C2 obtained from both monoexponential and biexponential decay fits versus the applied field strength (top) and a plot of the rate of CSSdecay (inverse of τCSS) versus the applied field strength (bottom).
Table 2. The Magnetic Field Effect of triad-sequestered PERMC C2.
For monoexponential decay fits across the field strengths tested, the magnetic field effect of C2 is consistent with an increase in the τCSS by a factor of three before saturating, or asymptotically approaching at a maximum τ CSS of ca. 5 μs, near 500 mT (Figure 7a). For the biexponential fits were attempted, the slow time components of the CSS increased from 1.43 μs by a factor of twelve to 15.4 μs at 2700 mT (Figure 7b). The relative magnitude of the increase in the slow component of the τ CSS from 0 to 500 mT is on the order of what has been documented for the same triad in acetonitrile bulk solution.However, in the plot of the biexponential long-component τ CSS versus field strength, the saturation field is far less defined than from the monoexponential fits. The respective rate constants from the monoexponential (k) and biexponential fits (kslow and kfast) are given in Table II.
The closest related fields of study to PERMC-based MFEs described herein encompass intermolecular electron transfers in glasses or temperature-frozen reverse micellar solutions[ 50,52].The magnetic field behavior of PERMC systems differ greatly from those in fluid reverse micellar solution, wherein the “cage effect” largely determines the relaxation rates of the photogenerated radical species[53,57].For the composite materials presented herein, there is no “cage effect”, because the triplet biradical CSS of the triad is formed intramolecularly. Moreover, the triplet biradical CSS cannot escape the micellar cage due to its high polarity and the solid state of the nonpolar phase. As such, the MFEs of PERMCs, on the whole, represent an unique class of photogenerated biradical relaxation pathways in the presence of magnetic fields of various strengths. Upon further refinement, the PERMC composites can serve as a novel class of optical magnetic field sensors and may be considered with such optical sensors already on the market[58,60]
Donor-Chromophore-Acceptor (triad)-sequestered poly(styrene- co-divinylbenzene) encapsulated reverse micelle composite materials were prepared wherein magnetic field-sensitive triad molecular sensors were sequestered in anionic AOT micelles. The reverse micelles in the CPRMSs were sized with
DLS. The intact micelles have radii on the order of 4.0 nm at the nonpolar phase polymerization temperature of 37 oC. Imaging
of cryo-cut PERMCs showed a micellar pore structure. The chemical environment exerted on the PERMC-sequestered triad caused a lengthening of τ CSSunder no applied field relative to bulk solution, as it is concurrent with a decrease in the frequency of encounters between the D+• and A+• moieties in the CSS. A MFE which increased the τ CSS by greater than or equal to three times, depending on the particular CSS fit function used, was observed for these composites upon saturation of the MFE at field strengths of ≥ 500 mT.
Support for this work (CME) by the National Science Foundation (CHE 0113112) is gratefully acknowledged. We thank Dr. Marc M. Greenberg (JHU) for his generous donation of 2,2’-azobis( 2,4-dimethylpentanenitrile).
Supporting Information Available:a schematic representation of the time-resolved transient absorption detection of CSS, Poissonian statistical calculations proving the non-double occupancy of triad T1 in one reverse micelle, digital picturesof composite materials C1-C3 exhibiting their optical quality, SEM images of gold-sputtered composite materials,a digital photograph of the conchoidal fracture surface of samples submitted for TM-AFM analysis, TM-AFM images of samples Cut I-II taken over a one square micrometer area, a schematic of the postulated cryo-cut mechanism for PERMC materials, dynamic light scattering study on the effect of the polar probe species concentration on the micellar size and polydispersity, root mean square analysis of atomic force microscope images of Cut I-II and height histograms along the vertical axis for each, transient absorption spectra of T1 in composites taken at six different times after photoexcitation, the CSS lifetimes measured at 0 T as they relate to W0, and the magnetic field effect of composite C3