FT-IR Spectroscopy and DFT Calculations on Fluorinated Macromer

Apr 28, 2010 - ... Milano, Italy, and Politecnico di Milano, Dip. di Chimica, Materiali, Ing. Chimica “G. Natta”, P.zza Leonardo da Vinci 32, 2013...
0 downloads 0 Views 941KB Size
6332

J. Phys. Chem. B 2010, 114, 6332–6336

FT-IR Spectroscopy and DFT Calculations on Fluorinated Macromer Diols: IR Intensity and Association Properties Stefano Radice,*,† Elena Di Dedda,‡ Claudio Tonelli,† Roberto Della Pergola,‡ Alberto Milani,§ and Chiara Castiglioni§ SolVay Solexis R&D Center, Viale Lombardia 20, 20021 Bollate (MI), Italy, UniVersita` di Milano Bicocca, Dipartimento di Scienze dell’Ambiente e del Territorio, Piazza della Scienza 1, 20126 Milano, Italy, and Politecnico di Milano, Dip. di Chimica, Materiali, Ing. Chimica “G. Natta”, P.zza Leonardo da Vinci 32, 20133 Milano, Italy ReceiVed: March 1, 2010; ReVised Manuscript ReceiVed: April 12, 2010

Fluorinated macromeric diols have been investigated by means of Fourier transform infrared (FT-IR) spectroscopy; the absolute intensities of characteristic bands of OH stretching have been determined and related to aggregation phenomena. Main results show that the ratio between “free” and “hydrogen-bonded” OH groups strictly depends on the polarity of chemical environment (macromeric-polymeric matrix). The intepretation of the experimental data has been supported by density functional theory (DFT) calculations on suitable molecular models, validating the results obtained both qualitatively and quantitatively. Introduction In this paper, the vibrational spectroscopic features of a fluorinated macromer are studied, with special focus on the characteristic IR bands due to -OH end groups and their evolution with hydrogen-bonding association in different matrixes. The macromer investigated has a perfluoropolyether (PFPE) chain backbone of general structure CF2O(CF2-CF2-O)p-(CF2-O)qCF2- and methylolic (CH2OH) end groups: its commercial name is Fluorolink D10-H, and it is manufactured by Solvay Solexis. This special class of macromolecules is useful as building blocks for new materials, characterized by unique bulk and surface properties, such as polyurethane hydrogenated/fluorinated systems.1-4 Moreover, they can be utilized in advanced lubrication and nanotribology applications.5 One of the key points for a successful fine-tuning of final fluorine containing materials is the knowledge of the chemical physical parameters of the PFPE building block, such as solubility parameters, but also viscosity and its dependence on the aggregation state; this knowledge is useful for predicting the kinetic of reaction of the investigated polymerization and compatibility with hydrogenated organic coreagents or materials, for example, in order to optimize formulations for cosmetic applications. These macromers contain both a perfluorinated internal chain and hydrogenated end groups: these two segments confer to the molecule different physicochemical properties and justify the observed unique features of this class of products. For instance, their “intramolecular” inhomogeneity is generally considered to be a relevant aspect for the understanding of their physical properties and chemical reactivity. Moreover, it is possible to fine-tune the ratio between fluorinated and hydrogenated segments, by varying the length or the polarity of the end-capping groups and/or the length of the internal fluorinated backbone; this offers a unique tool for the * To whom correspondence should be addressed. E-mail: stefano.radice@ solvay.com. Telephone: +390238356563. Fax: +390238356355. † Solvay Solexis R&D Center. ‡ Universita` di Milano Bicocca. § Politecnico di Milano.

modulation of the final properties, and it is achieved through specific synthetic routes or precursor selection. All these exceptional features have been encompassed in the scientific literature by the term of “copolymer end effect”.6-8 As a matter of fact, the physical properties of these intrinsically inhomogeneous macromers are strongly influenced by the polar interactions between end units, whereas the interactions between the apolar internal polymer backbones are much weaker. For this reason, PFPE functional macromolecules can be represented by a general structure E-M-E, where E represents the end units and M the molecular body (chain backbone). For a more general and comprehensive description of PFPE materials, end-capped by funtionalized or nonfunctionalized end groups, and their properties, see, for example, ref 20. In this study, we apply IR spectroscopy focusing our attention on the OH band absorption intensity, in order to determine the absolute content of oxidrilic end groups and their aggregation states. Ultimately, these data are fundamental to evaluate quantitatively molecular interaction and electronic distribution within the bulk of the macromer. Oxidrilic groups are clearly observed in IR spectra by means of the OH stretching band in the frequency range 3700-3000 cm-1. OH stretching frequency, shape, and intensity strongly depend on inter- and intramolecular interactions; generally, by increasing hydrogen bond interaction, frequencies are shifted to lower wavenumbers while intensities increase significantly.9 On these grounds, we will show here how these effects are strongly modulated by the polarity of matrixes or solvents. Materials and Methods Experimental Details. Fluorinated macromers diols studied in the present work are commercially available Solvay Solexis products: 1. Fluorolink D10-H. It is a linear PFPE characterized by the presence of methylolic groups as chain ends: HO-CH2CF2-O-(CF2-CF2-O)p-(CF2-O)q-CF2-CH2-OH.

10.1021/jp101840f  2010 American Chemical Society Published on Web 04/28/2010

FT-IR and DFT Studies on Fluorinated Macromer Diols

J. Phys. Chem. B, Vol. 114, No. 19, 2010 6333

Figure 1. Sketches of the molecular models used in DFT calculations (B3LYP/6-311++G**) to simulate the vibrational spectra of Fluorolink D10-H polymers: (a) isolated molecule and (b) hydrogen-bonded dimer. Molecular formula: HO-CH2-CF2-O-CF2-CF2-O-CF3 and dimer of the same.

The MW of the studied sample is 1400 as determined by 19F NMR, with a p/q ratio (r) 1.5 < r < 2 and a random distribution of the repeating units along the macromeric chain. This material has been dissolved in the following macromeric-polymeric PFPE based products: 2. H-Galden ZT-130. It is a linear PFPE, fully fluorinated apart from end groups characterized by a difluoromethylic structure: CF2H-O-(CF2-CF2-O)p-(CF2-O)q-CF2H. Its MW is 500 as determined by 19F NMR and has a p/q ratio (r) 1.5 < r < 2 and a random distribution of the repeating units along the macromeric chain. 3. Fomblin M15. It is a fully fluorinated linear PFPE: CF3-O-(CF2-CF2-O)p-(CF2-O)q-CF3. Its MW is 11 000 as determined by 19F NMR, with p/q = 1 and a random distribution of the repeating units along the macromeric chain. 4. Fomblin Y25. This is a PFPE with a branched structure:

Its MW is 3000-4000 as determined by F NMR and m . n. All analyzed samples were in the range of low concentration solutions, below 5% (w/w), since corresponding higher concentration solutions have been studied in previous works.7,10,11 We used different Fourier transform infrared (FT-IR) equipment, specifically FT-IR Thermo Nicolet NEXUS 870 and FTIR Nicolet MAGNA 850 with an attenuated total reflectance accessory (Golden Gate, Specac; diamond crystal (n ) 2,4), single reflection 45°; acquisition parameters, 256 scans; resolution, 2 cm-1). Transmission experiments were carried out placing a few drops of materials between KBr windows and by means of suitable cells for liquids (optical path length, 0.2 mm and 1 mm). Reflection experiments have been carried out by placing a few drops of sample on top of the diamond crystal. In order to determine absorption coefficients related to free O-H stretching, we prepared very diluted solutions, down to 0.025% w/w. Solutions in a range of concentration between 5 and 0.5% were obtained by direct weight of product and then adding the proper amount of solvent, while for lower concentrations we proceeded by dilution. Before spectroscopic data acquisition, each dispersion was stirred for 30 min, in order to get as much as possible an homogeneous sample. Computational Details. First principles calculations have been carried out and compared with the experimental spectra in order to support the assignment and to validate the conclusions reached on the basis of the experimental findings. Suitable molecular models are required to this aim, and in Figure 1 sketches of the models used here are reported: an isolated molecule (HO-CH2-CF2-O-CF2-CF2-O-CF3) and an hydrogen bonded dimer have been investigated for an interpretation of both intra- and intermolecular effects. DFT calculations have 19

Figure 2. Fluorolink D10-H: Experimental IR spectrum (thin film) compared with DFT computed (B3LYP/6-311++G**, frequency scaling factor 0.9688) IR spectra of the hydrogen bonded dimer.

been used as implemented in the Gaussian03 code:12 the B3LYP functional13 and 6-311++G** were chosen, since they already showed a good applicability in the prediction of the vibrational and structural properties of hydrogen-bonded perfluorinated polymers.14 The computed spectra and the frequencies reported throughout the paper have been scaled by a constant frequency scaling factor of 0.9688, as suggested by Merrick et al.15 Results and Discussion The spectrum of Fluorolink D10-H is reported in Figure 2, and it is compared with the DFT computed IR spectra for the model shown in Figure 1b. The most intense bands in the range 1250-1000 cm-1 are due to cooperative normal modes including mainly C-C, C-F, and C-O-C stretching motions in the macromeric backbone. Bands due to asymmetric and symmetric CH2 stretching are observed at 2958 and 2891 cm-1, respectively, and calculated at 3026 cm-1 (intensity 6 km/mol) and 2941 cm-1 (intensity 25 km/mol) for the isolated molecule (Figure 1a) and at 3035 cm-1 (intensity 4 km/mol), 3011 cm-1 (14 km/mol), 2959 cm-1 (intensity 22 km/mol), and 2934 cm-1 (intensity 27 km/mol) for the dimer (Figure 1b). In the OH stretching region, it is possible to observe the broad component (3300 cm-1), due to hydrogen bonded oxidrilic end groups, and a weaker and sharper component (3643 cm-1), due to the OH stretching in the monomeric form (i.e., the free macromer). In the case of DFT computed spectra, the band due to the OH stretching of the isolated molecule is found at 3702

6334

J. Phys. Chem. B, Vol. 114, No. 19, 2010

cm-1 (intensity 54 km/mol) and the OH stretchings of the dimer are found at 3686 cm-1 (intensity 108 km/mol) and 3547 cm-1 (intensity 675 km/mol); in this last case, the OH bond is directly involved in hydrogen bonding, thus showing the usual frequency red shift and intensification observed for this type of interaction.11 Moreover, the presence of two components at 1450 and 1402 cm-1 mainly due to CH2 deformation can be noted; their intensity is higher than the corresponding stretching vibration, suggesting that the positive charge located on hydrogen atoms is considerably higher than that typically observed for aliphatic nonfluorinated compounds.16-18 In order to confirm this evidence, also the absolute intensities of the CH stretching bands should be taken into account. Experimental data give an absolute intensity value equal to 10.8 km/mol (1080 darks); this value refers to one single CH oscillator and lies in the usual range shown by C(sp3)H bonds, considering the literature data for a large number of molecules bearing CH bonds16 as also recently investigated.17,18 DFT calculations give an average value of intensity per CH bond equal to 15.5 km/mol for the isolated molecule and 16.8 km/ mol for the dimer: the value is larger than the experimental one, but it shows however a marked decrease with respect to CH bonds in n-alkanes (compare, for instance, with the experimental average CH stretching intensity of ethane of 28.5 km/mol).18 Moreover, since CH stretchings are extremely sensitive to the intramolecular environment19 (e.g., molecular conformation), the discrepancy between experimental and calculated values can be due to the fact that DFT calculations have been applied to the model shown in Figure 1a (almost transplanar conformation): a thorough investigation of these effects would require a complete conformational study followed by the calculation of IR spectra of the relevant conformers. However, this is out of the aims of the present work, focusing on the hydrogen-bonding properties in connection with the features of IR spectra in the OH stretching frequency region. In any case, the experimental absolute intensity and frequencies indicate that the hydrogen has a quite acidic character, with a high positive charge, and high bond dissociation energy (BDE). According to McKean,19 we obtained values for r°CH and BDECH equal to 1.088 Å and 106.29 kcal/mol, respectively. In Figure 3, the spectra of Fluorolink D10-H in three different macromeric solvents (4000-3100 cm-1 region), all at the same concentration, are reported (transmission, fixed optical path, 1 mm, subtraction result with pure solvent as reference, subtraction factor )1). The two OH components can be clearly observed: the sharper one at higher frequency and the broad one at lower frequency. The intensity ratio between these components is different according to the different polarity of solvents. From a qualitative point of view, nonpolar solvents tend to reduce the concentration of free OH groups (without hydrogen bonding) and promote the formation of H-bonded complexes; this finding will be discussed quantitatively in the following. In order to determine the absorption coefficients of free OH groups in a different fluorinated environment (Fomblin Y25, Fomblin M15, and H-Galden ZT-130), solutions at concentrations ranging from 0.025% (w/w) up to 5% have been prepared and evaluated. Analyses show the following features: (i) The frequency of free OH stretching in the same solvent is independent of the concentration. (ii) The frequency of free OH stretching is slightly sensitive to the fluorinated solvent (3646 and 3647 cm-1 for Fomblin solvents and 3643 cm-1 for H-Galden ZT-130).

Radice et al.

Figure 3. Experimental IR spectra of Fluorolink D10-H in three different macromeric solvents: IR spectrum at the same concentration, 3% w/w (transmission, fixed optical path, 1 mm, subtraction result with pure solvent as reference, subtraction factor )1).

The absolute absorption coefficients values were calculated by linear regression considering only a limited set of data at very low concentration (down to 2.5 mmol/L and 25 mmol/L for the two Fomblin and for H-Galden, respectively). In Figure 4, the IR spectral subtractions are reported in the OH frequency range. Experimental evidence shows in this case two different behaviors, depending on the apolar or more polar nature of the selected solvent. Specifically, the intensity ratio between free/ bonded OH component is larger for the polar matrix. All optical density (OD) data related to free OH stretching component versus concentration are reported in Table 1. OD values have been normalized to 1 cm optical path and measured on infrared spectral subtraction. Table 1 highlights a linear behavior only in the low concentration region; actually, the first four data points for nonpolar matrixes and seven data points for the more polar one have been considered for linear regression. We used such a limited region to calculate the free OH absorption coefficients.

ε (3647 cm-1) ) 73.64 cm2 /mmol (free OH in Fomblin Y25) ε (3646 cm-1) ) 92.64 cm2 /mmol (free OH in Fomblin M15) ε (3643 cm-1) ) 79.80 cm2 /mmol (free OH in H-Galden ZT-130) By means of the absorption coefficients obtained, it was possible to determine the quantity of free OH component in our samples: in Figure 5a, the concentrations relative to the free OH component are reported versus total Fluorolink D10-H concentration in the three different matrixes. In order to support these results, the experimental spectra shown in Figure 4 for different concentrations have been fitted by using the results of DFT computations. The procedure is

FT-IR and DFT Studies on Fluorinated Macromer Diols TABLE 1: OD Data Related to Free OH Stretching Component versus Concentration (% w/w C1[Fomblin Y25 solutions], C2[Fomblin M15 solutions], and C3[H-Galden ZT-130 solutions])a C1

OD

C2

OD

C3

OD

5.090 3.100 1.050 0.520 0.310 0.110 0.052 0.026

2.210 1.990 1.184 0.672 0.492 0.207 0.095 0.050

4.998 2.890 1.020 0.520 0.310 0.100 0.054 0.026

2.607 2.174 1.565 1.093 0.772 0.236 0.129 0.062

5.000 3.000 1.000 0.500 0.300 0.100 0.050 0.024

8.983 5.735 2.024 1.032 0.619 0.203 0.096 0.046

a OD values have been normalized to 1 cm optical path and measured on infrared spectral subtraction.

the same already adopted in previous studies14 to the same aims: based on the computed OH stretching bands for the isolated molecule and dimer, a fitting function is defined as a weigthed sum of two Lorentzian functions, related, respectively, to the OH stretching band predicted for the isolated molecules (3702

J. Phys. Chem. B, Vol. 114, No. 19, 2010 6335 cm-1, 54 km/mol) and the OH stretching bands of the dimer (3686 cm-1, 108 km/mol; 3547 cm-1, intensity 675 km/mol). The fitting coefficients are the two weights of the Lorentzians (related to the relative concentration of the two species), two frequency scaling factors, and the bandwidths of the bands. The fittings were done on the whole OH stretching region, including the contributions of both free and hydrogen-bonded species. On this basis, the concentrations of free OH bonds have been calculated and are reported in Figure 5b: a good qualitative agreement is obtained with the results reported in Figure 5a, while quantitative values show some discrepancies, in particular for 3% and 5% concentrations. It should be noted that in Figure 5a the concentrations of free OH bonds are determined on the basis of the optical densities of the stretching band of the free OH bonds; the fitting reported in Figure 5b is carried out on the experimental IR spectra by taking into account implicitly the whole integrated band area of both free and hydrogenbonded OH stretchings. In other words, the (b) section of the figure is more strictly related to absolute intensities rather than optical densities, and this difference can justify the discrepancy found from a quantitative comparison with the (a) section. In

Figure 4. Experimental IR spectra (subtraction results for concentration from 5% down to 0.02%; 5% and 1% are indicated) of (a) Fluorolink D10-H in Fomblin Y25, (b) Fluorolink D10-H in Fomblin M15, and (c) Fluorolink D10-H in H-Galden ZT-130.

Figure 5. (a) Experimental concentration (obtained throught exp. absorption coefficients) of free OH groups versus concentration in three different solvents: blue, H-Galden ZT-130; black, Fomblin Y25; red, Fomblin M15. (b) Calculated concentrations of free OH groups versus concentration in the three different solvents obtained by the coefficient of the fitting of the experimental IR spectra on the basis of DFT computed spectra in the OH stretching region.

6336

J. Phys. Chem. B, Vol. 114, No. 19, 2010

any case, the qualitative trends observed in Figure 5b are very similar to those in Figure 5a, and their comparison supports both the reliability of the procedure adopted and the interpretation of the results obtained. Conclusions In this work, absorption coefficients for the OH stretching of CH2OH end groups in fluorinated macromers have been obtained. In order to achieve this goal, very low concentrations of such groups have been analyzed in different fluorinated matrixes and DFT calculations on suitable models have been used to support the interpretation of the experimental observations. Two classes of matrixes have been evaluated as solvent: more polar and totally apolar ones; once determined, these absorption coefficients and the degree of association of hydroxylic groups in the two selected solvent families can be easily compared. It comes from this evaluation that the more polar solvent prevents the hydrogen-bonding formation among OH end groups, at least up to 5% (w/w) concentration. This evidence is consistent with the “end copolymers” nature of functionalized perfluoropolyethers.10 Consistently, it has been shown by the literature that solvents can be classified in two categories: M-type and E-type solvents, depending on their preferred interaction with the internal apolar PFPE molecular body or with the polar hydrogenated end group, respectively.11 For these reason, polar solvents, such as H-Galden, belonging to the E-type class, are much more efficient in competing with the molecular association between end units; their polarity, hydrogen-bond character, and chemical similarity to the end units justify their ability to break or prevent the intermolecular association of PFPE molecules through end group physical linkages (i.e., hydrogen-bonding). When less polar solvents or totally apolar solvents are considered (e.g., perfluorinated molecules, i.e., Fomblin, that are typical M-type solvents), this competitive behavior is strongly reduced and a much higher dilution is required in order to allow to the M-solvent to evidence its ability to compete with intermolecular associations. Acknowledgment. We gratefully aknowledge E. Barchiesi, G. Canil, A. Di Meo, S. Fontana, and G. Geniram for experimental support and analytical characterization and for providing suitable standard samples. Note: GALDEN, H-GALDEN, FOMBLIN, FLUOROLINK are registered trademark names of Solvay Solexis. References and Notes (1) Tonelli, C.; Trombetta, T.; Scicchitano, M.; Castiglioni, G. J. Appl. Polym. Sci. 1995, 57, 1031.

Radice et al. (2) Tonelli, C.; Trombetta, T.; Scicchitano, M.; Simeone, G.; Ajroldi, G. J. Appl. Polym. Sci. 1995, 59, 311. (3) Tonelli, C.; Ajroldi, G. J. Appl. Polym. Sci. 2000, 87, 2279–2294. (4) Radice, S.; Turri, S.; Scicchitano, M. Appl. Spectrosc. 2004, 58, 535. (5) NLGI Spokesman 2004, 68, 14. (6) Danusso, F.; Levi, M.; Gianotti, G.; Turri, S. Polymer 1993, 34, 3687. (7) Khomko, E.; Mashlyakovskiy, L.; Tonelli, C. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 5354. (8) Danusso, F.; Levi, M.; Turri, S.; Gianotti, G. Eur. Polym. J. 1994, 30, 1449. (9) (a) Pimentel, G. C.; McClellan, A. L. The hydrogen bond; W.H. Freeman: San Francisco, 1960. (b) Steiner, T. Angew. Chem. Int. Ed. 2002, 41, 48. (10) Turri, S.; Scicchitano, M.; Gianotti, G.; Tonelli, C. Eur. Polym. J. 1995, 31, 1235. (11) Turri, S.; Scicchitano, M.; Gianotti, G.; Tonelli, C. Eur. Polym. J. 1995, 31, 1227. (12) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford CT, 2004. (13) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (14) (a) Radice, S.; Canil, G.; Toniolo, P.; Guarda, P. A.; Petricci, S.; Milani, A.; Tommasini, M.; Castiglioni, C.; Zerbi, G. Macromol. Symp. 2008, 265, 218. (b) Milani, A.; Castiglioni, C.; Di Dedda, E.; Radice, S.; Canil, G.; Di Meo, A.; Picozzi, R.; Tonelli, C. Polymer, DOI: 10.1016/ j.polymer.2010.04.002. (15) Merrick, J. P.; Moran, D.; Radom, L. J. Phys. Chem. A 2007, 111, 11683. (16) (a) Gussoni, M. J. Mol. Struct. 1986, 141, 63. (b) Vibrational Intensities in Infrared and Raman Spectroscopy; Person, W. B., Zerbi, G., Eds.; Elsevier: Amsterdam, 1982. (17) Gussoni, M.; Castiglioni, C.; Zerbi, G. In Handbook of Vibrational Spectroscopy; Chalmers, J., Griffiths, P., Eds.; John Wiley & Sons: Chichester U.K., 2001; Vol. 3, p 2040, and references therein. (18) Milani, A.; Castiglioni, C. J. Phys. Chem. A 2010, 114, 624. (19) McKean, D. C. Chem. Soc. ReV. 1978, 7, 399. (20) Marchionni, G.; Ajroldi, G.; Pezzin, G. In ComprehensiVe Polymer Science; Aggarwal, S. L., Russo, S., Eds.; Pergamon Press: London, 1996; Suppl. 2, p 347.

JP101840F