Formation of Host-Guest Complexes of β-Cyclodextrin and Perfluorooctanoic Acid

Published: June 20, 2011

r 2011 American Chemical Society 9511 dx.doi.org/10.1021/jp110806k | J. Phys. Chem. B 2011, 115, 9511–9527

ARTICLE

pubs.acs.org/JPCB

Formation of Host-Guest Complexes of β-Cyclodextrin andPerfluorooctanoic AcidAbdalla H. Karoyo,† Alex S. Borisov,‡ Lee D. Wilson,*,† and Paul Hazendonk*,‡

†University of Saskatchewan, Department of Chemistry, Saskatoon, SK, S7N 5C9, Canada‡University of Lethbridge, Department of Chemistry and Biochemistry, Lethbridge, AB, Canada

bS Supporting Information

1. INTRODUCTION

Many fluorine-containing compounds such as pharmaceuticals,pesticides, coatings, adhesives, and surface-active agents representa growing list of persistent organic pollutants (POPs) accumulat-ing in Canadian and global environments. This is particularly truefor recalcitrant perfluorinated compounds (PFCs) of the typeCF3-(CF2)n-R0, where R0 = CF2�OH, COOH, CONH2, orCF2�SO3H. There is a considerable interest in developinginnovative green strategies that involve the sequestration of suchPFCs with suitable sorbent materials that exhibit good sorptioncapacity and molecular selectivity. For example, Deng et al.developed a molecular imprinted polymer with good sorptiontoward perfluorooctane sulfonic acid (PFOS; R0 =CF2�SO3H) inaqueous solution.1

PFCs of the type described above possess unique physicochem-ical properties, as compared with their hydrocarbon analoguesbecause they are generally apolar and relatively inert owing to thestability of the C�F bond.2 They generally have low vaporpressures (∼10 mmHg for PFOA; R0 = CF2�CO2H at 25 �C)3and exhibit long residence times in the environment. The relatively

high surface activity of PFCs confers their application as high-performance surfactants, emulsifiers, and surface coatings formetals and paper.3 PFOA and PFOS are commonly found in soil,sediments, and aquatic environments because of their ability toinfiltrate groundwater to varying extents. Environmental contam-ination by PFCs was reported from direct discharge of industrialactivities, for example, aqueous fire fighting foams4 and wastewatereffluents from water treatment plants.5 Consumption of contami-nated foods and inhalation of air laden with volatile PFCs (e.g.,perfluorinated alcohols and esters) thatmay be degraded to PFOAare regarded as other possible ingestion pathways.3,4 Despite thehuman health and environmental concerns of PFOA, the distribu-tion pathways are not fully understood; however, researchers havelinked its exposure to cancer,6 birth defects,7 infertility,7 liverdamage,8 and suppression of immunity.9�11 Therefore, there is aneed to further study PFOA because of its widespread use and the

Received: November 12, 2010Revised: May 16, 2011

ABSTRACT: Structural characterization and dynamic properties ofsolid-state inclusion complexes of β-cyclodextrin (β-CD) with perfluor-ooctanoic acid (PFOA) were investigated by 19F/13C solid-state and19F/1H solution NMR spectroscopy. The complexes in the solid statewere prepared using dissolution and slow cool methods, where thermalanalyses (DSC and TGA), PXRD, and FT-IR results provided com-plementary support that inclusion complexes were formed betweenβ-CD and PFOA with variable stoichiometry and inclusion geometry.19F DP (direct polarization) and 13C CP (cross-polarization) withmagic-angle spinning (MAS) solids NMR, along with 19F/1H solutionNMR were used to characterize the complexes in the solid and solutionphases, respectively. The dynamics of the guest molecules in theinclusion complexes (ICs) were studied using variable temperature(VT) 19F DP/MAS NMR experiments in the solid state. The guest molecules were observed to be in several different molecularenvironments, providing strong evidence of variable host�guest stoichiometry and inclusion geometry, in accordance with thepreparation method of the complex and the conformational preference of PFOA. It was concluded from PXRD that β-CD and PFOAform inclusion complexes with “channel-type” structures. Variable spin rate (VSR) 19F DP/MAS NMR was used to assess the phasepurity of the complexes, and it was revealed that slow cooling resulted in relatively pure phases. In the solution state, 1H and 19F NMRcomplexation-induced chemical shifts (CISs) of β-CD and PFOA, respectively, provided strong support for the formation of 1:1 and2:1 β-CD/PFOA inclusion complexes. The dynamics of the guest molecule in the β-CD/PFOA complexes in D2O solutions wereprobed using VT 19F NMR and revealed some guest conformational and exchange dynamics as a function of temperature and therelative concentrations of the host and guest.

9512 dx.doi.org/10.1021/jp110806k |J. Phys. Chem. B 2011, 115, 9511–9527

The Journal of Physical Chemistry B ARTICLE

associated health and environmental risks. Perfluorinated surfac-tants are generally more surface-active than their hydrogenatedanalogues, as evidenced by the lower critical micelle concentra-tions (cmc’s). A comparison of the cmc values for PFOA(∼0.0105 M),12 SPFO (sodium perfluorooctanoate, 0.032 M),13

and sodium octanoate (∼0.4 M)14�18 illustrates the differences insurface activity between PFCs and hydrocarbon surfactants. Thesolution and colloidal behavior of surface-active agents may beattenuated by their complexation with host compounds such ascyclodextrins.19�21 Cyclodextrins (CDs) are macrocyclic (1f4)-linked oligomers of R-D-glucopyranose, and the most commonlystudied CDs are R-, β-, and γ-CD, consisting of 6-, 7-, and8-glucopyranose units, respectively.22 β-CD displays the remark-able ability of forming stable inclusion complexes with perfluori-nated alkanes, partially because of the good “size-fit” comple-mentarity23 of the guest and the host. Complexes of β-CD/perfluoroalkyl carboxylates of variable chain length were con-cluded to form stable inclusion complexes in aqueous solution,according to 19F/1H NMR spectroscopy,13,24,25 viscometry,25

conductometry,23,26,27 and sound velocity.28 In particular, SPFOforms very stable 1:1 and 2:1 CD/SPFO inclusion complexes. Incontrast with solution studies, there are relatively few detailedexamples documenting the formation of solid-state inclusioncomplexes between β-CD and perfluorinated guests (e.g., PFOA),and this may be attributed to challenges associated with obtaininggood quality single crystals. Tatsuno et al.29 have reported a solid-state NMR study of CD inclusion compounds consisting ofmedium (C-9) chain and long (C-20) chain perfluorocarboncompounds; however, several other studies have examined mod-ified perfluorocarbons25,30 and semifluorinated polymers31,32 inthe solid state. Thus, solid-state structural studies of inclusioncompounds of CDs and PFCs remain largely unexplored. Incontrast, a single-crystal XRD (X-ray diffraction) study of β-CD/hydrocarbon carboxylic acids has been reported.33 Recent ad-vances in solid-state NMR spectroscopy have provided theopportunity to obtain detailed structural information of suchamorphous and crystalline host�guest materials.34,35

In this Article, we report a detailed study of the formation ofinclusion complexes between β-CD and PFOA with differentpreparation methods (slow cool vs dissolution) using solid/solution-state NMR, FT-IR, PXRD, and thermoanalytical (DSCand TGA) methods. The β-CD/PFOA complex is a veryinteresting system because it can adopt a variable host�gueststoichiometry and inclusion geometry in condensed phases thatdepends on thermodynamic parameters such as relative host�guest concentrations and temperature. β-CD/PFOA complexesare amenable to the use of HFX solid-state and multinuclearNMR techniques;36 therefore, the dynamics of the free andcomplexed PFOA was investigated using VT 19F DP/MAS andVT 19F NMR techniques in the solid and solution states,respectively. VSR 19F DP/MAS NMR was used to study thephase purity and stability of the complexes prepared by slow cooland dissolution methods.

2. EXPERIMENTAL SECTION

2.1. Materials. β-CD hydrate (∼10% w/w) and PFOA (96%)were purchased from Sigma-Aldrich Co. and were used asreceived without any further purification. The water contentsof the materials were accounted for when preparing samples foreach method. Potassium hydrogen phthalate (KHP) and sodium

hydroxide were purchased from Merck and EMD chemicals,respectively.2.2. Preparation of Solid Inclusion Compounds.2.2.1. Method 1: Slow Cool. β-CD and PFOA in mole ratios of

1:1 and 2:1 (β-CD/PFOA) were prepared in deionized water inglass vials to form a thick slurry. Deionized water was addeddropwise with continuous stirring under heating at∼80 �C untila clear solution was formed. The resulting mixture was slow-cooled over several hours by placing the solution into vials in ahot water bath housed within an insulated box to ensure gradualcooling to obtain the solid product.2.2.2. Method 2: Evaporation/Dissolution. In the dissolution

process, aqueous β-CD/PFOA solutions were prepared at moleratios of 1:1 and 2:1 (β-CD/PFOA) in glass vials with contin-uous stirring and mild heating (∼40 �C). The solution mixturewas allowed to cool to room temperature, and the solvent wasevaporated gradually over several days (∼4 to 5) to obtain a solidproduct. All materials were ground into fine powders andcharacterized using NMR, DSC, TGA, PXRD, and FT-IR.2.3. Solution NMR. Solution NMR experiments were per-

formed on a three-channel Bruker Avance (DRX) spectrometeroperating at 500.13 MHz for 1H and 470.30 MHz for 19F. All 1HNMR spectra were referenced externally to tetramethylsilane(TMS, δ 0.0) and 19F spectra were referenced externally to 2,2,2-trifluoroethanol (TFE, δ �79.21). Samples for 1H NMR wereprepared in D2O at pD ∼5 in mole ratios of 1:1, 2:1, and 3:1β-CD/PFOA. Samples for 19F NMR were similarly prepared inD2O at pD∼5 and at variable β-CD/PFOAmole ratios (1:1 and2:1). The external reference TFE solution for the 19F NMRsamples were prepared in D2O at pD∼5 in sealed capillary tubesat a concentration of∼0.5 mM. All NMR spectra obtained in thesolution state were acquired at 295 K. Dry nitrogen gas was usedto control the temperature for the VT 19FNMR dynamic studies.2.4. Solid-State NMR Spectroscopy. All solid-state NMR

spectra were obtained using a Varian INOVA spectrometeroperating at 470.33 (19F) and 125.55 MHz (13C). 19F and 13Csolid NMR spectra were referenced externally to liquid hexa-fluorobenzene (HFB, δ �164.9) and adamantane (δ 38.5),respectively. All samples were spun at the magic angle with aspinning rate between 10 and 25 kHz using 2.5 and 3.2 mmVespel rotors equipped with Kel-F turbine caps, inserts, and sealscrews. All NMR spectra were obtained using a 100 kHz sweepwidth with 8192 points in the FID and were zero-filled to 64 Kdata points, unless stated otherwise. For all DP experiments, aone-pulse sequence was used, where the 90� excitation pulselengths for the 19F channel were set to 2.5, 3.25, 4.25, and 4 μs,and relaxation delays were set to 4, 4, 4, and 8 s, respectively.19F{1H} and 13C {1H} spectra of all solid samples were acquiredusing a two pulse phase modulation (TPPM) decouplingmode,34,37 where the powers of the 1H and 19F decoupledchannelswere set to 35.7 and 38.5 kHz, respectively, as determinedby peak-to-peak voltage. Curve and width for the 1H f 13Cramped CP experiments were set to 50 and 10 000 Hz, respec-tively, and optimal Hartmann�Hahn matching conditions wereachieved by setting contact times to 1 ms and CP powers to 54.2and 59.5 kHz for the proton and carbon channels, respectively.2.5. Differential Scanning Calorimetry and Thermogravi-

metric Analysis.Differential scanning calorimetry (DSC) of thenative β-CD, PFOA, and the inclusion complexes were acquiredusing a TAQ20 thermal analyzer, whereas the thermogravimetricanalysis (TGA) was performed on a TA Q50 over a temperaturerange of 30 to 350 �C. The scan rate forDSCwas set to 10 �C/min,



whereas that for TGA was 5 �C/min, and dry nitrogen gas wasused in both cases to regulate the sample temperature and samplecompartment purging. DSC samples were run in hermeticallysealed aluminum pans, whereas, an open pan geometry was usedfor the TGA measurements.2.6. Fourier Transform Infrared. Fourier transform IR spec-

tra were obtained using a Bio-Rad FTS-40 spectrometer with aresolution of 4 cm�1. All spectra were obtained with spectro-scopic grade KBr, which constituted ∼80% (w/w) of the totalsample. Samples were run as finely ground powders in reflectancemode.2.7. Powder X-ray Diffraction. Powder X-ray diffraction

(XRD) spectra were collected with a Rigaku Rotaflex RU-200rotating anode XRD using Cu KR radiation. The applied voltageand current were set to 40 kV and 80 mA, respectively. Thesamples were prepared by adding drops of hexane to the finepowder on the quartz sample holder to form a suspensionresulting in a homogeneous film. The samples were mountedin a vertical pan configuration in the diffractometer, and XRDpatterns were measured in the continuous mode for a 2θ anglerange of 5�20� with a scan rate of 0.5�/min. Finely powderedsilica was used to calibrate the positions of the diffraction lines.

3. RESULTS AND DISCUSSION

3.1. Characterization (DSC, TGA, FT-IR, and PXRD).3.1.1. Characterization of the Inclusion Complexes. The host-

guest stoichiometry of the β-CD/PFOA ICs was determined byacid/base titration in conjunctionwith gravimetry. The amounts ofPFOA in the washed and unwashed complexes were evaluated bytitration with 0.01 M NaOH (standardized with KHP), and theamounts of β-CD were estimated by difference. The preparationsof 1:1 β-CD/PFOA complexes (i.e., dissolution vs slow cooling)yielded the following host/guest ratios, unwashed (0.95:1 vs 1.1:1)and washed (1.2:1 vs 1.3:1), respectively. The 2:1 β-CD/PFOAcomplexes yielded the following host/guest ratios: 1.6:1 versus1.7:1 (unwashed) and 1.6:1 versus 1.7:1 (washed) for dissolutionversus slow cooling, respectively. In general, the slow-cooledproducts showed a slightly greater β-CD/PFOA mole ratio thanthe complexes prepared by the dissolutionmethod. The results arenot straightforward to interpret unequivocally because the host/guest mole ratios >1:1 may arise from the occurrence of excess β-CD, reduced guest, or a possible contribution of 2:1 complexes.28

In the case of 2:1 host/guest ratios, stoichiometric ratios <2:1may bedue to the formationof small amounts of 1:1 complexes.Nonetheless,the experimental values for the 1:1 and 2:1 β-CD/PFC complexesare consistent with the relativemagnitude of the binding constants(e.g., ∼7 � 104 M�1 (1:1) and ∼9 � 102 M�2 (2:1))24,38 inaqueous solution. The preparative methods for the 2:1 complexyield a mixture of 1:1 and 2:1 complexes, whereas, the preparativemethods for the 1:1 product yield chiefly 1:1 complexes. Washingof the solid-state products does not appear to affect the stoichiometryto any significant extent.3.1.2. DSC and TGA. The DSC thermograms for native β-CD,

PFOA, and the 1:1 and 2:1 β-CD/PFOA complexes prepared bythe slow cool and dissolution methods are shown in Figure 1. InFigure 1a, the thermogram for β-CD hydrate shows two endo-thermic transitions centered at 110 and 320 �C that are attributedto its dehydration and decomposition, respectively. PFOA showsthree endothermic transitions at ca. 60, 100, and 170 �C; these areattributed to the melting, dehydration, and vaporization transi-tions, respectively. TheDSC curves for the inclusion complexes of

β-CD/PFOA prepared at the 1:1 and 2:1 mol ratios, using theslow cool (SC) and dissolution (D) methods, are shown inFigure 1b,c, respectively. The thermograms in Figure 1b,c indicatethe presence of several unique endothermic events between ca.120 to 250 �C, as compared with native β-CD and PFOA. Thethermal events are attributed to the variable phase transitions ofthe complex and the loss of hydrate water from β-CD occurringbetween ca. 120 to 150 �C. In the context of the above-notedphase transitions, this refers to the relative spatial orientation ofthe guest within the host�guest complex. The 1:1 complexes areconsistently characterized by two features at ca. 180 and 210 �C,whereas the 2:1 complexes have only one feature at 210 �C. Thepresence of two thermal transitions support the existence ofadditional topologies of the 1:1 complex. The slow-cooledcomplexes are characterized by sharp and pronounced featuresat 250 �C, which may reveal the greater phase purity of slowcooled comparedwith thedissolution complexes (cf. Figure 1b).39

According to the DSC results, the increased thermal stability ofPFOA observed for the 1:1 and 2:1 complexes with β-CDprovides support that β-CD/PFOA inclusion complexes areformed, and this is consistent with the favorable complex stability.40

The thermophysical properties of bound PFOA within the cavityof β-CD are anticipated to exceed that for pure PFOA due to theoccurrence of favorable van der Waals interactions for the complex,as summarized in Table 1.In Figure 2, the TGA results for PFOA and β-CD show mass

losses at ∼100 and 320 �C, respectively. The differences in thetwo temperatures observed between DSC and TGA transitionsare attributed to the sample pan configuration and the nature ofthe physical measurement. DSC is generally sensitive to thermalphase transitions, whereas TGA reveals weight loss events due tochemical changes. The thermal events at 220 �C in Figure 2 forPFOA for the 1:1 and 2:1 slow-cooled complexes provideevidence of the volatilization of the guest; whereas, the event at310 �C is attributed to the decomposition of β-CD. The purePFOA vaporizes immediately after dehydration under the con-ditions employed in the TGA experiments because the sample isheated in an open system. However, the TGA results support theformation of inclusion complexes between β-CD and PFOAbecause the physicochemical properties of guests are altered inthe bound state41 (cf. Table 1). The TGA scan for the dissolutioncomplexes showed similar results; however, the thermal transi-tions are broader in nature and lack the sharp features observed inthe complexes prepared by slow-cooling.3.1.3. Fourier Transform Infrared. FT-IR spectroscopy can

provide useful information on guest conformational preferences(i.e., gauche vs trans) of the alkyl chain of the guest in theβ-CD/PFOA inclusion complexes.29 Figure 3 shows the FT-IRspectra of the native β-CD, pure PFOA, and both 1:1 and 2:1β-CD/PFOA complexes in the region between 500 and4000 cm�1 (left) and an enlargement of the area between 800and 1400 cm�1 (right). Figure 3 (L) shows the vibrational bandsfor the carbonyl group of pure PFOA at ∼1750 cm�1 and itsbound form in the 1:1 and 2:1 β-CD/PFOA complexes. Thehydroxyl and C�H vibrational bands (∼3300 and 2900 cm�1,respectively) exhibit stronger intensity for the 2:1 β-CD/PFOAcomplex, and the spectrum is consistent with the greater molefraction of β-CD hydrate. In Figure 3b�d, the expanded spectrareveal three bands appearing at ca. 1230 (1), 1210 (2), and 1150(3) cm�1, which have been previously reported.29,42,43 Vibrationalbands 1 and 3 are attributed to CF2 asymmetric and symmetricstretching, whereas band 2 is assigned to the C�C and C�C�C



stretching and bending modes, respectively. The most notablefeatures for the spectra in Figure 3 are the relative changes inintensity of the bands (1�3) for the 1:1 and 2:1 complexes(cf. Figures 3b,c). These bands are attributed to different

conformational isomers (i.e., anti, gauche, and ortho), as concludedfor solid and liquid forms of C4F10,

44 where the anti and gaucheconformers are favored. The attenuation of band 2 at ca.1210 cm�1 (Figure 3c) indicates an alteration of the relative

Figure 1. DSC thermograms of native β-CD hydrate and PFOA (a), the 1:1 and 2:1 complexes prepared by slow cooling (b), and dissolution(c) methods.



population of PFOA for the trans and gauche conformers in the 2:1β-CD/PFOA complex. Tatsuno et al.29 indicate that band 2(1230 cm�1) for the β-CD/C9F20 complex was attributed mainlyto the anti-conformer. In the absence of quantum mechanicalcalculations and detailed spectroscopic evidence, the unequivocalassignment of the FT-IR bands of PFOA is beyond the scope ofthe present work, as compared with the published results forC9F20. In addition, the foregoing IR results for various vibrationalbands are susceptible to potential changes in dipolemoment whenaccompanied by a conformational change of PFOA as a conse-quence of changes in electron density.24 The relative intensities oftheβ-CDandPFOA lines in the 1:1 and 2:1 complexes (Figure 3a,b)provide further evidence of variable conformations and dynamicproperties of the guest.45 The FT-IR results in Figure 3 providestrong evidence that conformational changes of the perfluorocar-bon chain in the 1:1 and 2:1 complexes are variable and may beattributed to trans and gauche conformers of PFOA. There is asignificant decrease in the contribution of the gauche and transconformer of the PFOA chain in the 2:1 complex. The includedPFOA chain in the 1:1 complex (Figure 3b) may adopt a similarconfiguration to that of the free PFOA (Figure 3d) because aportion of the chain may extend outside of the β-CD macrocycle.In this regard, the PFOA chain in the 1:1 complex may adoptvariable inclusion topologies with multiple configurations of theguest perfluoroalkyl chain, as previously concluded from the DSCresults. Further elaboration on the conformational effects of thePFOA chain is made in the discussion of the NMR results (videinfra). The FT-IR results may provide a semiquantitative estimateof the relative host�guest composition from the relative intensityof respective vibrational bands of certain functional groups. Amoredetailed understanding of the host�guest complex can be ob-tained from NMR studies of these systems.

3.1.4. Powder X-ray Diffraction. Figure 4 shows the powderX-ray diffraction (PXRD) patterns for β-CD, PFOA, and the 1:2,1:1, and 2:1 β-CD/PFOA inclusion complexes that were pre-pared by slow cooling and dissolution methods. The XRDspectra for the 1:1 and 2:1 β-CD/PFOA complexes representunique solid-phase materials, as evidenced by the appearance ofunique diffraction lines. The XRD results for the complexescannot be simulated from the additive patterns of the β-CD andPFOA components and provide further evidence that the com-plexes are unique phases. The XRD pattern for β-CD (Figure 4a)exhibits prominent signatures at various 2θ values, 10.8, 12.6, and17.8�, and other minor signatures at higher 2θ values. The resultsin Figure 4a are characteristic of a cage-type lattice structurearising from the random packing arrangement of β-CD.46�48

The 1:1 and 2:1 complexes derived from slow cooling anddissolution methods show slightly different XRD patterns, asindicated in Figure 4d�g. Two prominent peaks at ca. 11.8 and17.8� observed for the 1:1 and 2:1 complexes indicate that“channel-type” crystalline structures46�49 may be formed andare distinctly different from that of the PXRD pattern for β-CDhydrate. Such low-angle scattering lines outlined above areconsistent with spacings in the crystal framework ∼5.0 and7.60 Å, and these have been attributed to the inter- andintraspacings of β-CD.29 Furthermore, the XRD patterns forthe 1:1 β-CD/PFOA complexes are characterized by a smallshoulder above the main signature at ca. 11.8� (indicated byasterisk). The additional XRD lines for the 1:1 complexes providefurther support for the existence of additional configurations dueto the conformational preference of PFOA, in agreement with theDSC and FT-IR results. In the case of the 1:2 β-CD/PFOAcomplex, a low-angle 2θ (∼7.8�) signature is attributable tocontributions arising from unbound PFOA and to those of thepure 1:1 complex. The general features of the PXRD spectra ofthe 1:1, 1:2, and 2:1 resemble the results for native β-CD becausethe former are anticipated to have similar unit cell dimensionsrelative to that of the macrocyclic host, as observed in Figure 4.3.2. Solution-State NMR.3.2.1. Solution NMR Characterization. Complexation-induced 1H

and 19FNMR chemical shifts (CIS) in aqueous solution were usedto provide evidence of the formation of inclusion complexesbetween β-CD and PFOA. Figures 5 and 6 illustrate the respective1H and 19F NMR spectra for PFOA and its inclusion complexeswith β-CD. The assignment of the 1H NMR signals agreed withprevious assignments.50,51 The 1H NMR chemical shifts of theinterior cavity protons (H3 and H5) of β-CD in Figure 5 provideestimates of the extent of PFOA inclusion within the β-CDcavity.50,51 Scheme 1 illustrates the formation of a β-CD/PFCinclusion complex, where Ki is the equilibrium binding constant.Scheme 2 outlines the molecular structure of β-CD structure witha schematic side view showing the interior cavity host protons.

Table 1. DSC and TGA Thermoanalytical Dataa

systems DSC results (�C) TGA results (�C)

M.P./phase transitions* hydrate water loss vaporization decomposition melt (guest) melt (host)

native β-CD 110 ∼320 310

unbound PFOA 60 100 ∼175 100

β-CD/PFOA (1:1) range* ∼120 - 150 ∼250 >300 ∼220 320

β-CD/PFOA (2:1) range* ∼120 - 150 ∼250 >300 ∼220 320a * indicates range between ∼150 and 250 �C.

Figure 2. TGA plots for native β-CD hydrate, PFOA, 1:1 β-CD/PFOAcomplex, and the 2:1 β-CD/PFOA complex.



The results illustrated in Figure 5 are consistent with the factthat the guest molecule displaces the polar solvent from the hostcavity according to the upfield chemical shifts of the complex forthe H3 and H5 protons (cf. Scheme 2). The CIS values of theseprotons to lower/higher fields are a consequence of theirincreased chemical shielding/deshielding effects resulting fromthe complexation process.42 The inclusion of an apolar PFOAguest molecule within the apolar cavity of β-CD is thermodyna-mically favored according to the hydrophobic effect.29 However,as the mole ratio of the host increases beyond the 1:1 ratio, theintracavity protons (H3, H5) become less shielded and move tohigher field because of the greater mole fraction contributionof the unbound host (cf. Figure 5c,d) because the observed

chemical shift is a weighted average of the bound and unboundspecies.24,42 In contrast, the 1:1 complex (Figure 5b) shows themost attenuated effect upon the observed CIS values for H3 andH5 nuclei and is due to the increased steric and shielding effectsof the extended PFOA chain24 (cf. Scheme 1). The 19F NMRresults in Figure 6 reveal somewhat different shielding/deshield-ing patterns for the 1:1 and 2:1 complexes and further supportthe occurrence of variable stoichiometry and conformationaleffects of the PFOA chain. Perfluorinated alkyl chains in the solidstate have been reported to assume helical and zigzag conforma-tions for long chains (>C12) and short chains (eC8),respectively.42 A mixture of conformations (gauche and trans)may be present for intermediate chain lengths of 8�12 carbon

Figure 4. Powder X-ray diffraction (PXRD) spectra for (a) native β-CD, (b) unbound PFOA, (c) 1:2 β-CD/PFOA dissolution complexes, (d) 1:1β-CD/PFOA, and (e) 2:1 β-CD/PFOA slow-cooled complexes, (f) 1:1 β-CD/PFOA, and (g) 2:1 β-CD/PFOA dissolution complexes.

Figure 3. FT-IR spectra of (a) native β-CD, (b) 1:1 β-CD/PFOA complex, (c) 2:1 β-CD/PFOA complex, and (d) PFOA. The figure on the right is anexpansion of the spectral region between 800 and 1400 cm�1.



atoms.52 The helical structure was shown to be the stable formfor PFOA53 and PFOS54 according to density functionaltheory (DFT) and semiempirical calculations, respectively.The assignment of 19F signals for long-chain PFCs may bedifficult due to the long-range couplings (over at least fourbonds),55 and possible spectral overlap of the resonance lines.However, the assignment of the 19F spectrum of PFOA inFigure 6a is in good agreement with previous results by Goeckeet al.56 and Buchanan et al.57 In general, complex formationbetween PFOA and β-CD results in downfield shifts of the 19Fresonances (Figure 6b,c) because of the reduced polarizability ofthe β-CD cavity interior with respect to the bulk aqueous solvent.

However, a combined all-trans conformation (deshielding) ofthe perfluorocarbon chain and the occurrence of favorable vander Waals interactions with the CD interior (shielding) for the2:1 complex results in a composite deshielding/shielding effect.The end-to-end stacking of β-CD in the 2:1 complex (cf.Scheme 1) will force PFOA to adopt an extended all-transconformation to occupy the inclusion sites of two host macro-cycle moieties. This conformational change from gauche to transresults in the maximization of the through space F�F distancebetween neighboring CF2 groups and the optimization of favor-able van derWaals interactions within the β-CD interior.13 In thecase of the 1:1 complex (Figure 6b), the deshielding effect

Figure 5. 1HNMR in D2O solutions for (a) β-CD, (b) 1:1 β-CD/PFOA, (c) 2:1 β-CD/PFOA and (d) 3:1 β-CD/PFOA at pD∼5 and 295 K, where *indicates HOD signal.

Figure 6. Expanded 19F solution NMR spectra in D2O at pD ∼5 and 295 K of (a) PFOA, (b) 1:1 β-CD/PFOA complex, and (c) 2:1 β-CD/PFOAcomplex with spectral assignments.



observed for the CβF2 group may be associated with the loss ofthe γ-gauche shielding effect that is commonly observed for alkylcarboxylate groups in the gauche configuration58�60 as the PFCchain untwists itself. Knochenhauer et al. have shown thatperfluorocarbon chains unwind from a helical to an extendedchain configuration when packed in a periodic lattice.61 Thegauche-/trans-conformational interconversion of the perfluoro-carbon chain of the complexes is consistent with the FT-IRresults shown in Figure 3. In the 1:1 β-CD/PFOA complex(Figure 6b), unequal populations of the helical and zigzagconformers of PFOA may exist.52 The chemical shift of theterminal CF3 group in the complexed state may reveal theproximity of the perfluorocarbon chain relative to the interioror the periphery of the host.13 The observed chemical shifts inFigure 6 indicate that the CF3 group in the 2:1 inclusion complexis in closer contact with the hydrophobic cavity than within thebulk solvent environment. In contrast, the CF3 group in the 1:1inclusion complex can reside near the periphery of the cavity andextend into the bulk solvent. This is consistent with the length ofβ-CD cavity (∼7.9 Å),29 which is about one-half of the extendedchain length of PFOA. The overlapping signals of the CχF2 andCβF2 groups in the 1:1 complex (Figure 6b) and the variable lineshapes of the 19F signals for PFOA in the 1:1 and 2:1 complexesmay be attributed to the differences in the conformationalpreference and exchange dynamics of the PFOA guest as itadopts variable inclusion geometries. These dynamics are furtherexplained in the VT NMR studies. However, the CIS valuesprovide strong evidence that complexes are formed betweenβ-CD and PFOA in aqueous solution. The results in Figure 6are consistent with observed CIS values for the β-CD-SPFOsystem.24

3.2.2. Solution NMR Dynamics.The motional dynamics of thePFOA in the complexed state were probed by VT 19F NMR inaqueous solution. Figure 7 shows the VT 19F NMR stack plotresults for mixtures of β-CD and PFOA at the 1:1 mol ratio.There is an overall increase in the 19F chemical shifts as thetemperature is increased. The downfield trend in the NMRchemical shifts may be attributed to the conformational changesof the perfluorocarbon chain. According to VT 19F NMR andspin relaxation studies in solution, Ellis et al. suggested that PFCchains may adopt a helical twist (gauche) conformation;52

however, slow untwisting of the helix may occur, and theperfluorocarbon chain may adopt an all-trans conformation atelevated temperatures,61,62 as discussed above. Bunn and Ho-wells concluded from their XRD studies that PFC chains under-go variable rotation about the chain axes (helical conformation)and longitudinal zigzag displacement at temperatures aboveambient conditions.62 Thus, temperature effects may induceCF2 groups with variable conformations. The helical nature ofperfluorocarbon chains is generally explained by the greater vander Waals radius of fluorine relative to hydrogen and thecorrespondingly greater repulsive interactions between alternat-ing CF2 groups.

61,62 The helical structure can be viewed as therotation about the perfluorocarbon chain in an attempt of thechain to increase the distance between alternate CF2 groups andminimize the repulsive interactions. In Figure 7, the PFOA chaincan be interpreted as undergoing a conformational change fromhelical (gauche) to an extended all-trans configuration at elevatedtemperatures. This is consistent with the downfield shifts ofthe 19F signals as the temperature is increased. Apart fromthe chemical shift changes, another unique feature is the overlapand separation of the CχF2 and CβF2 (cf. Figure 6) signalsat∼�125 ppm as a function of temperature, and this is similarlyobserved in the stack plots of Figures 8�10. As the temperatureincreases, averaging and sharpening of the 19F signals of the CχF2and CβF2 is observed. The results are attributed to the increasedmotional dynamics (i.e., correlation time) for a 1:1 complex inrelation to that of the 2:1 complex because of the difference inrelative molecular weight. The positions of CχF2 and CβF2 maybe regarded as pseudosymmetric points along the perfluorocar-bon chain, and for temperatures below 15 �C (Figure 8) thesesignals start to separate. This is attributed to changes in theconformational preference of the perfluorocarbon chain aboutthe Cχ and Cβ positions. The presence of a conformationaldependence (i.e., coiled and uncoiled forms) of the γ-gaucheshielding effect at the CβF2 group may contribute to the overlapand separation of the CχF2/CβF2 resonance lines for thesetemperature conditions. The reversibility of the exchange dy-namics is shown by the heating and cooling cycle between 5 and25 �C in Figure 8.In contrast with the 1:1 complex, a different spectral pattern is

observed for the 2:1 complex, where the CχF2 and CβF2 signalsoverlap at higher temperatures (Figure 9). At low temperatures,the motional dynamics of the PFOA molecule may have greatervariability given the longer “channel-like” host space of twoβ-CDmolecules. Additionally, the greater molecular weight of the 2:1complex results in a decreased tumbling rate, which in turnresults in broader resonance lines accompanied by the divergenceof the CχF2 and CβF2 signatures at ambient temperature. As thetemperature is increased, the Cχ and Cβ methylene groups undergofurther conformational changes due to the increased motionaldynamics and the formation of a 1:1 complex. In the 1:1 complex.In the 1:1 complex, a portion of the PFOAchain extends outside of

Scheme 1. Formation of a Host�Guest Complex Is Shownfor β-CD (toroid) and a Perfluorinated Guest MoleculeAccording to a 1:1 and 2:1 Equilibrium Binding ProcessWhere Ki Are the Corresponding Equilibrium BindingConstants for Each Host�Guest Stoichiometrya

a In the case of perfluorooctanoic acid (PFOA); R = OH. The relativedimensions of the β-CD macrocycle and PFOA are not drawn to scale.

Scheme 2. Representations of the Molecular Structureof β-CDa

a Left and side refers to the oligomer where n = 7 and the right handstructure represents the toroidal shape of the macrocycle.



the β-CD cavity. These factors result in the overlapping of theCχF2�CβF2 signals observed at higher temperatures. However amore quantitative study of the spin relaxation times (T1, T2) andthe coupling constant measurements of the 19F groups of interestalong with 2-DNMR experiments may be required to characterizefurther the conformational features adopted by these host�guestsystems. A distinct feature shown for the 2:1 complex is thebroadening of the 19F signals at temperatures <15 �C (Figure 10)and is due to the attenuation of the exchange kinetics observed forthe 2:1 complexes (cf. Figure 6c). The decreased tumbling rate ofthe 2:1 complex previously discussed results in an increasedcorrelation time (τc) that yields broader resonance lines at reducedtemperatures and may be attributed to reduced exchange kineticsof the guest between the bulk and inclusion sites ofβ-CD. Previousstudies have shown that the T1 values of bound guest for 1:1 and2:1 complexes are markedly different relative to the unbound

guest.59 In general, the variable temperature studies provide strongsupport that the PFC chains of PFOA may interconvert betweenall-trans and gauche conformations during the heating and coolingcycles.3.3. Solid-State NMR. 3.3.1. Solid NMR Characterization. 19F

NMR was used to characterize the unbound PFOA and itsinclusion complexes with β-CD. 19F NMR spectroscopy in thesolid state is unique because of the high natural abundance andthe wide chemical shift dispersion of the 19F nuclei. Theopportunity to observe simultaneously several independentnuclei provides useful structural and quantitative informationfor host�guest complexes. The results in Figure 11 show the19FNMR spectrum acquired under DP conditions withMAS at 25 kHz.The spectrum of the unbound PFOA shows well-resolved 19Fsignals, and the assignment is in good agreement with the high-resolution solutionNMRspectrum(cf. Figure 6a). Figure 12 shows a

Figure 7. 19FNMR spectra under variable temperature conditions for the 1:1 β-CD/PFOA complex in D2O at pD∼5 (heating cycle from 25 to 75 �C).

Figure 8. 19F NMR spectra at variable temperature for the 1:1 β-CD/PFOA complex in D2O at pD ∼5 (cooling cycle from 55 to 5 �C).



well-resolved 13C NMR spectrum for β-CD hydrate obtained using1H f 13C CP/MAS under 20 kHz spinning conditions. Although

β-CD is composed of seven glucose monomers, the individual 13Cnuclei are not completely resolved in the CP/MAS spectrum, and

Figure 9. 19F NMR spectra at variable temperature for the 2:1 β-CD/PFOA complex in D2O at pD ∼5 (heating cycle from 25 to 75 �C).

Figure 10. 19F NMR spectra at variable temperature for the 2:1 β-CD/PFOA complex in D2O at pD ∼5 (cooling cycle from 55 to 5 �C).

Figure 11. Solid-state 19F DP/MAS spectrum of PFOA obtained at MAS = 25 kHz and 295 K.



this is attributed to the crystallographic inequivalence of the pyranoseunits and coincidental overlap of certain resonance lines because ofthe variable hydration state and amorphous characteristics of pow-dered samples with mixtures of nonstoichiometric β-CD hydrate.

In Figure 13, evidence is provided for the formation of 1:1(Figure 13b) and 2:1 (Figure 13c) β-CD/PFOA inclusioncomplexes, respectively, prepared using the slow cool method.Complex formation between β-CD and PFOA results in reduced

Figure 12. Solid-state 13C H f C CP/MAS spectrum of native β-CD hydrate at VSR = 20 kHz and 295 K.

Figure 13. 19F DP MAS 25 kHz NMR spectra at 295 K for (A) pure PFOA, (B) 1:1 β-CD/PFOA complex, and (C) 2:1 β-CD/PFOA complex,obtained by the slow cool method.

Figure 14. 19F DP/MAS 25 kHzNMR spectra at 295 K for (A) free PFOA, (B) 1:1 β-CD/PFOA complex obtained by dissolution, and (C) 1:1 β-CD/PFOA complex obtained by the slow cool method.



motional averaging of the 19F homonuclear dipolar couplings,which result in broadening of the 19F signals and the appearanceof spinning side bands29 (cf. asterisks in Figures 13 and 14b,c). Asthe PFOA becomes increasingly bound (i.e., conversion of 1:1 to2:1), the stoichiometry of the complex coincides with theattenuation of the sharp component at ∼82 ppm (Figure 13c)and the presence of increasingly broadened 19F signatures,compared with the 1:1 complex (Figure 13b). The presence ofthe sharp CF3 component (∼�82 ppm) in the 1:1 complexindicates that there may be greater motional dynamics of coilingand uncoiling of segments of the PFC chain. The structure of 1:1complex can be described as a dynamic pseudorotaxane complex,wherein the β-CD macrocycle can encapsulate ∼4 to 5 CF2

groups with an ensemble of potential configurations, depending onthe conformation of the guest.64 In contrast, the 2:1 complexrepresents a more static pseudorotaxane-type complex in which agreater proportion of the PFC chain is included within two hostcavities (cf. Scheme 1). As a consequence of this inclusion geometry,a potentially lower energy state configuration confines the motionalchain dynamics of the guest. These results are consistent with theprevious solution NMR results (cf. Figure 6). Furthermore, thepresence of sharp and broad signatures for the CF3 group (∼�79and �82 ppm) in the 1:1 complex provides support that it mayreside in two or more microenvironments (i.e., inside vs outside theβ-CDmacrocycle), and this conclusion is consistent with results forDSC, FT-IR, and PXRD presented above.

Figure 15. 19F DP NMR spectra for unbound (free) PFOA at variable spin rates (VSRs) at 295 K.

Figure 16. 19F DP NMR spectra of the 1:1 β-CD/PFOA complex prepared by the dissolution method and run at variable spin rates (VSRs) at 295 K.



Figure 14 depicts the 19F DP/MAS NMR spectra of the 1:1β-CD/PFOA inclusion complexes prepared by the slow cool anddissolution methods, respectively. The product obtained by slowcooling (Figure 14c) is characterized by relatively broad 19Fsignatures due to a microenvironment in which the motionaldynamics of PFOA are attenuated, as compared with Figure 14b.TheDSC results (cf. Figure 1) have shown greater phase purity ofthe complexes prepared by slow cooling relative to the productsobtained by dissolution. In the latter case (cf. Figure 14b), theappearance of numerous sharp resonance lines is consistent withincreasedmotional dynamics of PFOA. The sharp resonance linesobserved in Figure 14a are attributed to the unrestricted motionaldynamics of unbound PFOA, as compared with the bound formsof the 1:1 and 2:1 β-CD/PFOA complexes. In Figure 14, thechemical shift trends for the sharp CF3 line of the dissolutioncomplex (Figure 14b) follow a similar trend as that observed inthe solution NMR spectra (cf. Figure 6). In contrast with thedissolution complex, the slow-cooled product does not follow thesame trend (cf. Figure 13b). This observation provides supportfor the presence of variable conformations of the PFOA chain insolid complexes, in accordance with the type of preparativemethod (i.e., dissolution vs slow cooling). In general, the PFOAchain for the 1:1 complexes have a distribution of the gauche andtrans conformers. However, the results provide strong evidence ofthe formation of β-CD/PFOA inclusion complexes with variablestoichiometry and inclusion geometry in the solid state.3.3.2. Solid NMR Dynamics. The foregoing results have

illustrated that β-CD forms stable inclusion complexes with

PFOA in the solution and solid states, respectively, with variablehost�guest stoichiometry and inclusion geometry. It is well-known that the physicochemical properties (stability, solubility,surface tension, etc.) of the guest molecules in the bound state aredramatically altered for inclusion complexes with β-CD.19�21

However, the relationship between the mode of preparation ofthese complexes and their structure in the solid state are poorlyunderstood and require further study. An understanding of thestructure, dynamics, and molecular recognition properties ofsuch solid-state complexes is afforded by multinuclear NMRstudies described herein. As previously observed, the complexesprepared by the slow cool and dissolution methods yield distinctdifferences in their NMR line shape patterns (cf. Figure 14), andthis was concluded as being due to the inclusion geometry andguest conformation. The dynamic properties of the complexesprepared by the different preparative methods were studied using19F DP/MAS conditions with VSR (10 to 25 kHz) at 295 K. Theunbound PFOA (Figure 15) and the 1:1 dissolution complex

Figure 17. 19F DP NMR spectra of the 1:1 β-CD/PFOA complex prepared by the slow cool method and run at variable spin rates at 295 K.

Table 2. Chemical Shift Data for Uncomplexed and Com-plexed PFOA Guest at Variable Spin Rates (Values in BracketIndicate Chemical Shifts for Broad CF3 Components)

uncomplexed PFOA β-CD/PFOA 1:1 dissolution

signal δ (25 kHz) δ (23 kHz) δ (25 kHz) δ (23 kHz)

1 (ω) �81.53 �81.56 �81.24 (�79.38) �81.24 (�79.34)

2 (R) �118.6 �118.7 �118.3 �118.3

3 (γ) �121.0 �121.1 �120.6 �120.6

4 (δ) �121.5 �121.5 �121.1 �121.0

5 (χ) �122.1 �122.2 �121.7 �121.8

6 (β) �122.3 �122.4 �122.0 �122.0

7 (ε) �126.0 �126.1 �125.6 �125.6

Figure 18. 19F DP MAS NMR spectra at 14 kHz and variabletemperature for the 1:1 β-CD/PFOA complex prepared by the dissolu-tion method (warming from �90 to 20 �C).



(Figure 16) display well-resolved 19F resonances with broad andsharp components for the specified VSR conditions, as antici-pated. As the VSR increases, dipolar couplings are scaled down,and the different phases possessing variable configurations of theguest are revealed. In Figure 17, the 1:1 complex prepared byslow cooling appears more static for these VSR conditions, assupported by the absence of sharp 19F guest signatures. Incontrast, variable microenvironments and configurations arerevealed for the 1:1 complex prepared by dissolution(Figure 16). Following the interpretation of the results inFigure 15, the 1:1 complex (Figure 16) is made of two types ofbound guest: a mobile and static phase. The results are consistentwith the variable binding conformation of the PFOA chain in the1:1 complex, as previously described. The frequencies of thenarrow components due to the complexed and uncomplexedPFOA are provided in Table 2. In general, PFOA shows greatermobility in its uncomplexed form. Additionally, the differentmicroenvironments in which PFOA exists in the complexed stateare shown as narrow and broad signatures in Figures 15�17.The results also suggest that the guest dynamics for the com-plex prepared by slow cooling is more static in nature, as

compared with the product prepared by dissolution for theseconditions.The dynamics of PFOA in the complexed state were studied

with VT 19F DP/MAS NMR. The 19F NMR experiments wereobtained at moderate MAS rate (14 kHz) while varying thetemperature from�90 to 100 �C. Figure 18 illustrates the 19FDPMAS spectra where the system is heated from�90 to 22 �C. Theguest dynamics of the 1:1 complex prepared by the dissolutionmethod are observed to increase as the temperature increases.The appearance of a sharply resolved upfield 19F signature for theCF3 group (∼82 ppm) occurs for temperatures g0 �C and isattributed to the occurrence of greater rotational dynamics forthese conditions. However, the signatures for the CF2 groups arerelatively invariant over the temperature range investigated, andthis is attributed to insufficient motional averaging of the 19Fhomonuclear dipolar couplings because of the static nature of thePFOA chain. It is clear that the guest dynamics are expected to bemore pronounced as the temperature is increased above thephase transition of the guest molecule according to observedthermophysical transitions observed for the guest at tempera-tures g60 �C (cf. Figure 1a).

Figure 19. 19F DP/MAS NMR spectra obtained at 14 kHz and variable temperature for the 1:1 β-CD/PFOA complex prepared by the dissolutionmethod. The heating cycle was from 25 to 90 �C (A) and the cooling cycle was from 85 to 25 �C (B).



As shown in Figure 19, a further increase in the temperaturefor the 1:1 complex prepared by dissolution reveals significantdynamic effects with increasing temperature. Figure 19A,B showthe VT 19F DP/MAS NMR results for the heating (25 to 90 �C)and cooling (85 to 25 �C) cycles, respectively. During the heatingcycle, the 19F signals for PFOA become more resolved as thetemperature is increased. The relative intensity of the sharp CF3group signature (∼�82 ppm) increases at the expense of thebroad CF3 component (∼�79 ppm) up to ∼80 �C, revealingvariable distribution of conformational effects as a function oftemperature. The other 19F signals for the CF2 groups ∼�120ppm become more resolved as the temperature is increased to∼80 �C. This observation is consistent with the interconversionbetween conformational states of PFOA, as previously discussed.We conclude that the broad and narrow CF3 componentscorrespond to the gauche and trans conformers, respectively.The broadening of the CF3 signal is suspected to arise from thestrong coupling between the inequivalent fluorines in the gaucheconformer, whereas the equivalent fluorines in the trans con-former are expected to exhibit narrowing of the CF3 signal.Above 80 �C, additional 19F signatures are observed, and thissupports the fact that changes in the microenvironment anddynamics of the PFOA chain occur. The guest environmentundergoes a change because the host may undergo desolvationunder these temperature conditions, as evidenced by theDSC results in Figure 1. The process of dehydration may occur

at ∼80 �C because the NMR sample rotor is an open system. Inaddition, the apparent temperature recorded during the experi-ment is expected to be less than the in situ rotor temperaturebecause the mechanical effects of spinning are expected togenerate heat. The role of hydrate water and guest moleculardynamics are supported from independent ultrasonic relaxationstudies by Aicart et al.63 Tatsuno and Ando29 reported similarconclusions, according to their VTNMR studies ofβ-CD and longchain n-alkyl PFCs. Upon cooling (90 to 25 �C) in Figure 19B, thetrends in line shapes are observed to be reversible for PFOA, ascompared with the heating cycle (cf. Figure 19A). However, thecomplex did not show complete reversibility for the spectraobtained at 25 �C (Figure 19B) after cooling, as compared withthe original spectrum recorded prior to the heating cycle(Figure 19A). The observed irreversibility is attributed to changesresulting from the loss of hydrate water during the heating cyclebecause the purge gas stream irreversibly removes hydrate waterfrom the system. The void volume arising from dehydration causesamicroenvironment change that subsequently affects themotionaldynamics of the PFC chain in the host framework.29 In addition, aslight but observable shift changes for the 19F signals are observedfor the guest in the complexed state during the heating and coolingprocesses, respectively. The heating and cooling processes may beaccompanied by different populations of trans and gauche confor-mers of PFOA. The chemical shift changes are as high as∼1.3 ppm

Figure 20. 19F DP/MAS NMR spectra obtained at 14 kHz and variable temperature for the 2:1 β-CD/PFOA complex prepared by the slow coolmethod. The heating cycle was from 25 to 100 �C (A) and the cooling cycle was from 90 to 25 �C (B).



in the solution state and∼0.8 ppm for the solid-state spectra overthe specified temperature ranges.Variable dynamic effects were observed for the 2:1 complexes

prepared by the slow cool method (Figure 20). The absence of asharp component for the terminal CF3 group ∼�82 ppm inFigure 20A suggests that the complex formed by the slow coolmethod is different in nature than the corresponding 1:1 complex(cf. Figure 14c). The broad resonance lines for the 2:1 complexare consistent with attenuatedmotional dynamics with a differentinclusion geometry compared with the 1:1 complex. Minordynamic effects are observed above 80 �C for the 19F guestsignals at ∼�120 ppm, and this is related to the loss of hydratewater, as described above. The reduced temperature dependenceof the 19F line shapes for the 2:1 complexes prepared by slowcooling is observed because of the greater stability of thiscomplex, in agreement with the DSC results. However, thereare minor upfield and downfield chemical shifts for the 19F guestsignals during the heating and cooling cycles, and these arerelated to the interconversion between trans and gauche con-formers of the PFC chain of PFOA, as described above.

4. CONCLUSIONS

Two preparative methods, slow cool and dissolution, wereused to prepare 1:1 and 2:1 β-CD/PFOA solid-state complexes.NMR spectroscopy, FT-IR, thermal analyses (DSC and TGA),and PXRD provide support that β-CD/PFOA complexes adoptchannel-type structures. Complexes prepared by slow coolingpossess less mobile components, as compared with complexes pre-pared by dissolution. The 1H/19F NMR CIS values in solutionprovide evidence that 1:1 and 2:1 β-CD/PFOA inclusion complexesare formed. 1HNMR results indicate that PFOA is includedwithinthe β-CD cavity. VT 19F NMR results in solution indicate thatPFOA exists as an inclusion complex and undergoes temperature-induced intramolecular conformation and variable β-CD/PFOAmole ratios. PFOA in the complexed state adopts an ensemble ofall-trans and gauche conformations during the heating and coolingcycles in aqueous solution and solid states, respectively. Thestructure and conformational dynamics of the β-CD/PFOAcomplexes in the solid state are affected by the stoichiometry ofthe complex, temperature, and hydrate content of β-CD. Com-plexes prepared by slow cooling display reduced dynamics thancomplexes prepared by dissolution, and may be related to thegreater stability of the former because of the optimization ofintermolecular contacts between β-CD and PFOA.

’ASSOCIATED CONTENT

bS Supporting Information. Chemical shift data for un-bound and complexed PFOA guest molecule at variable spinrates; DSC thermograms of physical mixtures of β-CD/PFOA inthe 1:1 and 2:1 mole ratios; FT-IR spectra of native β-CD, 1:1and 2:1 β-CD/PFOA physical mixtures, and PFOA; and powderX-ray diffraction spectra of the 1:1 and 2:1 β-CD/PFOA physicalmixtures. This material is available free of charge via the Internetat http://pubs.acs.org.

’AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected]. Tel: 306 966-2961. Fax: 306 966-4730.and E-mail: [email protected]. Tel: 403-329-2657. Fax: 403-329-3201.

’ACKNOWLEDGMENT

We wish to acknowledge the University of Saskatchewan, theUniversity of Lethbridge, and the National Sciences and Engi-neering Research Council (NSERC) for support of this research.

’REFERENCES

(1) Deng, S.; Shuai, D.; Yu, Q.; Huang, J.; Yu, G. Front. Environ. Sci.Eng. China 2009, 3, 171–177.

(2) Richardson, S. D. Anal. Chem. 2009, 81, 4645–4677.(3) Fujii, S.; Polprasert, C.; Tanaka, S.; Nguyen, P. H. L.; Yong, Q. J.

J. Water Supply: Res. Technol.�AQUA 2007, 56, 313–326.(4) Saito, N.; Harada, K.; Inoue, K.; Sasaki, K.; Yoshinaga, T.;

Koizumi, A. J. Occup. Health 2004, 46, 49–59.(5) Boulanger, B.; Vargo, J. D.; Schnoor, J. L.; Hornbuckle, K. C.

Environ. Sci. Technol. 2005, 39, 5524–5530.(6) Gilliland, F. D.; Mandel, J. S. J. Occup. Environ. Med. 1993,

35, 950–954.(7) Guruge, K. S.; Yeung, L. W. Y.; Yamanaka, N.; Miyazaki, S.; Lam,

P. K. S.; Giesy, J. P.; Jones, P. D.; Yamashita, N. Toxicol. Sci. 2006,89, 93–107.

(8) Fei, C.; McLaughlin, J. K.; Lipworth, L.; Olsen, J. Hum. Reprod.2009, 1, 1–6.

(9) Yang, Q; Abedi-Valugerdi, M.; Xie, Y.; Zhao, X. Y.; M€uller, G.;Nelson, B. D.; DePierre, J. W. Int. Immunopharmacol. 2002, 2, 389–397.

(10) Peden-Adams, M. M.; Keller, J. M.; EuDaly, J. G.; Berger, J.;Gilkeson, G. S.; Keil, D. E. Toxicol. Sci. 2008, 104, 144–154.

(11) Dewitt, J. C.; Copeland, C. B.; Strynar, M. J.; Luebke, R. W.Reprod. Toxicol. 2009, 27, 409–416.

(12) Hoffmann, H.; W~urtz, J. J. Mol. Liq. 1997, 72, 191–230.(13) Guo, W.; Fung, B. M.; Christian, S. D. Langmuir 1992,

8, 446–451.(14) Wilson, L. D.; Siddall, S. R.; Verrall, R. E. Can. J. Chem. 1997, 75,

927–933 and references therein.(15) Zemb, T.; Drifford, M.; Hayoun, M.; Jehanno, A. J. Phys. Chem.

1983, 87, 4524–4528 and references therein.(16) Picquart, M.; Lacrampe, G. J. Phys. Chem. 1992, 96, 9114–9120.(17) D’Angelo, M.; Onori, G.; Santucci, A. Prog. Colloid Polym. Sci.

1994, 97, 154–157.(18) Gonzalez-Perez, A.; Ruso, J. M.; Prieto, G.; Sarmiento, F.

Colloid Polym. Sci. 2004, 282, 1133–1139.(19) Kitamura, S.; Fujimura, T.; Kohda, S. J. Pharm. Sci. 1999, 88,

327–330.(20) Nicolle, G. M.; Merbach, A. E. Chem. Commun. 2004, 854–855.(21) Dias, A. M. A.; Andre-Dias, C.; Lima, S.; Countinho, J. A. P.;

Teixera-Dias, J. J. C.; Marrucho, I. M. J. Colloid Interface Sci. 2006,303, 552–556.

(22) Sjetil, J.; Osa, T. In Comprehensive Supramolecular Chemistry;Atwood, J. L., Ed.; Pergamon Press: New York, 1996 and referencestherein.

(23) Palepu, R.; Reinsborough, V. C.Can. J. Chem. 1989, 67, 1550–1553.(24) Wilson, L. D.; Verrall, R. E. Langmuir 1998, 14, 4710–4717.(25) Zhang, H. H.; Hogen-Esch, T. E.; Boschet, F.; Margaillan, A.

Langmuir 1998, 14, 4972–4977.(26) Palepu, R.; Richardson, J. M.; Reinsborough, V. C. Langmuir

1989, 5, 218–221.(27) Saint Aman, E.; Serve, D. J. Colloid Interface Sci. 1990, 138,

365–375.(28) Junquera, E.; Aicart, E.; Tardajos,G.Langmuir1993, 9, 1213–1219.(29) Tatsuno, H.; Ando, S. J. Phys. Chem. B 2006, 110, 25751–25760.(30) Druliner, J. D.; Wasserman, E. J. Fluorine Chem. 1995, 72, 75–78.(31) Nostro, P. L.; Santoni, I.; Bonini, M.; Baglioni, P. Langmuir

2003, 19, 2313–2317.(32) Geppi, M.; Pizzanelli, S.; Veracini, C. A.; Cardelli, C.; Tombari,

E.; Lo Nostro, P. J. Phys. Chem. B 2002, 106, 1598–1605 .(33) Yannakopoulou, K.; Mavridis, I. M. Curr. Org. Chem. 2004, 8,

25–34.



(34) Borisov, A. S.; Hazendonk, P.; Hayes, P. G. J. Inorg. Organomet.Polym. 2010, 20, 183–212.(35) Ripmeester, J. A.; Ratcliffe, C. I. Compr. Supramol. Chem. 1996,

8, 323–380.(36) Battiste, J.; Newmark, R. A. Prog. Nucl. Magn. Reson. Spectrosc.

2006, 48, 1–23 and references therein.(37) Bennett, A. E.; Rienstra, C. M.; Auger, M.; Lakshmi, K. V.;

Griffin, R. G. J. Chem. Phys. 1995, 103, 6951–6958.(38) Xing, H.; Lin, S.; Yan, P.; Xiao, J.; Chen, Y. J. Phys. Chem. B

2007, 111, 8089–8095.(39) Zainal Abidin, S.; Ling, G. K. F.; Abdullah, L. C.; Ahmad, S.;

Yunus, R.; Choong, T. S. Y. Eur. J. Sci. Res. 2009, 33, 471–479.(40) Bruylants, G.; Wouters, J.; Michaux, C. Curr. Med. Chem. 2005,

12, 2011–2020.(41) Szejtili, J. J. Inclusion Phenom. Mol. Recognit. Chem. 1992, 14,

25–36.(42) Moynihan, R. E. J. Am. Chem. Soc. 1959, 81, 1045–1050.(43) Pawsey, S.; Reven, L. Langmuir 2006, 22, 1055–1062.(44) Albinsson, B.; Michl, J. J. Phys. Chem. 1996, 100, 3418–3429.(45) Dru_zbicki, K.; Mikuli, E.; Ossowska-Chru�sciel, M. D. Vib.

Spectrosc. 2010, 52, 54–62.(46) Okumura, H.; Kawaguchi, Y.; Harada, A.Macromolecules 2001,

34, 6338–6343.(47) Jiao, H.; Goh, S. H.; Valiyaveettil, S. Macromolecules 2002, 35,

3997–4002.(48) Li, N.; Liu, J.; Zhao, X.; Gao, Y.; Zheng, L.; Zhan, J.; Yu, L.

Colloids Surf., A 2007, 292, 196–201.(49) Caira, M. R. Rev. Roum. Chim. 2001, 46, 371–386.(50) Schneider, H. J.; Hacket, F.; R€udiger, V. Chem. Rev. 1998, 98,

1755–1786.(51) Webb, G. A. Annual Reports on NMR Spectroscopy; Academic

Press: New York, 1993; Vol. 7 and references therein.(52) Ellis, D. A.; Denkenberger, K. A.; Burrow, T. E.; Mabury, S. A.

J. Phys. Chem. A 2004, 108, 10099–10106.(53) Raynea, S.; Forest, K. THEOCHEM 2010, 949, 60–69.(54) Erkoc) , S.; Erkoc) , F. THEOCHEM 2001, 549, 289–293.(55) Battais, A.; Boutevin, B.; Moreau, P. J. Fluorine Chem. 1978, 12,

481–496.(56) Goecke, C. M.; Jarnot, B. M.; Reo, N. V. Chem. Res. Toxicol.

1992, 5, 512–519.(57) Buchanan, G. W.; Munteanu, E.; Dawson, B. A.; Hodgson, D.

Magn. Reson. Chem. 2005, 43, 528–553.(58) Dalling, D. K.; Grant, D. M. J. Am. Chem. Soc. 1967, 89,

6612–6620.(59) Perlin, A. S.; Koch, H. J. Can. J. Chem. 1970, 48, 2639–2643.(60) Roberts, J. D.; Weigert, F. J.; Kroschwit, J. I.; Reich, H. J. J. Am.

Chem. Soc. 1970, 92, 1338–1347.(61) Knochenhauer, G.; Reiche, J.; Brehmer, L.; Barbeka, T.;

Woolley, M.; Tredgold, R.; Hodge, P. J. Chem. Soc., Chem. Commun.1995, 1619–1620.(62) Bunn, C. W.; Howells, E. R. Nature 1954, 174, 549–551.(63) Jobe, D. J.; Verrall, R. E.; Janquera, E.; Aicart, E. J. Phys. Chem.

1993, 97, 1243–12483.(64) Wilson, L. D.; Verrall, R. E. J. Phys. Chem. B 2000, 104,

1880–1886.

Formation of Host-Guest Complexes of β-Cyclodextrin and Perfluorooctanoic Acid

Documents

Transcript of Formation of Host-Guest Complexes of β-Cyclodextrin and Perfluorooctanoic Acid