Hyperbranched Polyesters: Synthesis ... - ACS Publications

J. Phys. Chem. B 2007, 111, 8801-8811

8801

Hyperbranched Polyesters: Synthesis, Characterization, and Molecular Simulations Kishore K. Jena,† K. V. S. N. Raju,*,† B. Prathab,‡ and Tejraj M. Aminabhavi*,‡ Organic Coatings and Polymers, Indian Institute of Chemical Technology, Hyderabad 500 007, India, and Molecular Modeling DiVision, Center of Excellence in Polymer Science, Karnatak UniVersity, Dharwad 580 003, India ReceiVed: January 21, 2007; In Final Form: May 15, 2007

New types of hyperbranched polyesters were synthesized by the reaction of 2,2-bis(hydroxymethyl) propionic acid as an AB2-type monomer with pentaerythritol, trimethylol propane, or glycerol as the core moiety. The obtained globular networks were characterized by NMR and MALDI-TOF spectroscopic techniques. Molecular weights determined by MALDI-TOF were confirmed by gel permeation chromatography. Fourier transform infrared (FTIR) spectroscopy was used for the quantitative evaluation of hydrogen bonding as well as to study the structure-property relationship. To investigate the changes and types of intermolecular H-bonding interactions in hyperbranched polyesters with a variation in molecular structure, the deconvolution of FTIR spectra was carried out using Origin 6.0 software through the Gaussian curve-fitting method. Molecular simulations were performed through molecular mechanics and molecular dynamics (MD) calculations using the DISCOVER module. Cohesive energy density, solubility parameters, and surface properties of the hyperbranched polyesters were calculated. Further, vibrational analysis was computed using MD simulations for all the hyperbranched polyesters developed in this work.

Introduction

TABLE 1: Different Monomer and Molar Ratios Used for the Preparation of Hyperbranched Polyester

Dendritic polymers are special types of molecules with a complex architecture that are intensively studied due to their special properties and potential applications, particularly as drug delivery devices, catalysts, and commercial coatings.1 These mainly include dendrimers and hyperbranched polymers.2 A large number of hyperbranched polymers has thus been reported in earlier literature including, for example, polyesters,3 polyethers,4 polyamides,5 polyurethanes,6 polysiloxysilanes,7 etc. Even though significant progress has been made on the development of hyperbranched polymers for various applications, there is still a need to develop more hyperbranched polyesters that have more attractive applications as adhesives and coatings because of their simple synthetic routes, their unique highly branched molecular structures, and their large number of functional endgroups, as compared to the traditional linear polymers. Hyperbranched polyesters particularly possess cross-linked structures consisting of a large number of hydroxyl functional endgroups. The chemical structure of each of these polyesters greatly influences their ability to induce good adhesion properties, which is believed to occur through the combination of hydrogen-bonding interactions between the surface of the polymer and the substrate. Their low melting temperature and solution viscosity as well as the availability of a large number of functional endgroups have promoted research on the development of high performance coatings prepared from hyperbranched polyols. These systems are densely packed, having smaller hydrodynamic volumes than the linear polymers with comparable molecular weights that would result in a lower viscosity.8 * Corresponding authors. E-mail: (K.V.S.N.R.) [email protected] and (T.M.A.) [email protected]. † Indian Institute of Chemical Technology. ‡ Center of Excellence in Polymer Science.

sample name HBP-3 HBP-8 HBP-11 HBP-5 HBP-9 HBP-12 HBP-6 HBP-10 HBP-13

chemical composition PE + DMPA TMP + DMPA glycerol + DMPA

generation

molar ratio

first second third first second third first second third

1:4 1:4:8 1:4:8:16 1:3 1:3:6 1:3:6:12 1:3 1:3:6 1:3:6:12

It has been reported that properties of coating formulations containing hyperbranched polyols are superior to linear polyols9 because the highly branched polyesters have a significantly lower viscosity than their linear analogues of nearly equal molecular weights. In the literature, one of the most widely investigated families of hyperbranched polymers is that of hydroxy functional aliphatic polyesters, synthesized from 2,2bis(hydroxymethyl) propionic acid (DMPA) with various core molecules such as glycerol, trimethylol propane (TMP), or pentaerythritol.10-24 Hyperbranched aliphatic polyols have also been prepared using DMPA as an AB2-type monomer and pentaerythritol, TMP, or glycerol as the core moieties. Technological developments appear to have overstepped scientific knowledge since many physical properties of dendritic macromolecules have not yet been investigated thoroughly. In addition to experimental characterization of hyperbranched polymers, studies on molecular modeling through atomistic simulations would allow us to understand some of the intricate properties that are associated with such systems before we intend to seek their commercial applications. In the literature, only a handful of such reports is available on polyesters using molecular modeling simulations. Molecular dynamics simulations were carried out to compute physical properties like

10.1021/jp070513t CCC: $37.00 © 2007 American Chemical Society Published on Web 06/23/2007

8802 J. Phys. Chem. B, Vol. 111, No. 30, 2007

Jena et al.

SCHEME 1: Synthesis of Hyperbranched Polyester Coatings from Different Core Moieties

cohesive energy density, solubility parameters, and surface energy, since these parameters are required in understanding their coating applications. In addition, molecular simulations were also carried out to predict vibrational modes of hyperbranched polymers. In this paper, we describe the synthesis of hyperbranched polyesters using DMPA with core molecules such as glycerol, TMP, and pentaerythritol. The structural morphology and molecular behavior of the resulting hyperbranched polyesters were investigated by a variety of techniques such as FTIR, NMR, MALDI-TOF, and GPC. Experimental Procedures Materials. Pentaerythritol (PE), 2,2-bis(hydroxymethyl) propionic acid (DMPA), and trimethylol propane (TMP) were procured from Aldrich Chemicals (Milwaukee, WI). Titanium tetraisopropoxide (TTIP) was purchased from Fluka Chemical Corp. (Ronkonkoma, NY). Spectroscopic grade tetrahydrofuran (THF) and sulfur-free toluene from S.D. Fine Chemicals (Mumbai, India) were used as received. Glycerol and dimethyl formamide (DMF) were purchased from Qualigens (Mumbai, India).

Methods. Synthesis of Hyperbranched Polyester Polyol. Hyperbranched polyester polyol was synthesized by a melt polycondensation technique, wherein PE was charged to DMPA kept in a four-necked round-bottomed flask placed over an isomentel bath and equipped with a thermometer, mechanical stirrer, nitrogen inlet, and Dean-Stark apparatus. The reaction mixture was slowly heated to 190 °C. After complete melting of the reactants, the temperature was maintained between 190 and 200 °C with a continuous nitrogen flow for about 24 h to compute the esterification reaction. TTIP (0.05 wt %) based on the weight of DMPA was used as the catalyst for the esterification. The eliminated water was collected from the Dean-Stark apparatus. The reaction was monitored periodically by checking the acid value through a simple titration method and stopped when the acid value was below 10. Similarly, using the same apparatus, the second and third generations of hyperbranched polyols (HBPs) were prepared in a stepwise manner from the first generation polyols by adding the required amount of DMPA. In another reaction, TMP was reacted with DMPA in a four-necked flask as described previously to obtain the first, second, and third

Hyperbranched Polyesters

J. Phys. Chem. B, Vol. 111, No. 30, 2007 8803

Figure 1. Simulated structures of second generation DMPA with core moieties: (a) pentaerythritol, (b) trimethylol propane, and (c) glycerol (carbon: gray; oxygen: red; and hydrogen: white).

generation HBPs. Similarly, glycerol and DMPA were reacted to obtain different HBPs for a comparison of the structureproperty relationship. Scheme 1 shows the steps involved in the synthesis of hyperbranched polyesters, whereas Table 1 shows the various reactants used to prepare different hyperbranched polyesters along with their equivalent ratios. The polyesters prepared from

the first, second, and third generation HBPs are called HBP-3, HBP-8, HBP-11 (PE and DMPA), HBP-5, HBP-9, HBP-12 (TMP and DMPA), HBP-6, HBP-10, and HBP-13 (glycerol and DMPA), respectively. NMR. The 1H and 13C NMR spectra of the hyperbranched polyesters were recorded on Varian VXR-Unity 200 MHz and Bruker UXNMR 300 MHz spectrometers in dimethyl sulfoxide-


Figure 2. DMSO-d6.

Jena et al.

1

H NMR spectra of HBP-3, HBP-8, and HBP-11 in Figure 4. 13C NMR spectra of HBP-3, HBP-8, and HBP-11 in DMSOd6. Quaternary carbon region of the spectra was magnified and is shown at the top.

Figure 3. Magnification of the methyl region of the 1H NMR spectra of HBP-3, HBP-8, and HBP-11 in DMSO-d6.

d6 (DMSO-d6) with Me4Si (TMS) as the internal standard at ambient temperature. MALDI-TOF MS. MALDI mass spectra of the hyperbranched polyesters were recorded using a Kompact MALDI SEQ laser desorption time-of-flight mass spectrometer (Kratos Analytical, Manchester, U.K.) equipped with a pulsed nitrogen laser (λmax ) 337 nm, pulse width ) 3 ns). Ions were accelerated into the analyzer at a voltage of 20 kV. The sample spots were irradiated just above the threshold using the linear mode with pulsed extraction (delayed extraction). Thus, the laser power (irradiance) used to produce a good mass spectrum was analyte dependent. The linear mode was chosen over the reflector mode to maximize the signal intensity since the resolving power was sufficient to differentiate the observed oligomers. All the mass spectra were accumulated over 50 laser shots across the sample spot. The matrix used was 2-(4-hydroxyphenylazo) benzoic acid (HABA) (Sigma, St. Louis, MO) of 10 mg/mL concentration in tetrahydrofuran (THF) solvent.25 Samples were prepared in THF solvent in a 1 mg/mL concentration. To

Figure 5. 1H NMR spectra of HBP-5, HBP-9, and HBP-12 in DMSOd6. Expansion of the methyl group is shown at the top.

make an analyte/matrix deposit, typically, 0.3 µL of the analyte solution was placed on a stainless steel sample slide (part no.


Figure 6.

13


C NMR spectra of HBP-5, HBP-9, and HBP-12 in DMSO-d6.

TABLE 2: Calculated CED, δ and γ of Selected Hyperbranched Polyols DMPA-pentaerythritol physical properties (J/cm3

CED ) δ (cal/cm3)1/2 γ (mJ/m2)

DMPA-TMP

DMPA-glycerol

1G

2G

3G

1G

2G

3G

1G

2G

3G

333.3 8.9 34.68

288.9 8.3 31.6

225.3 7.3 26.78

328.3 8.9 34.34

266.9 7.9 29.96

201.2 6.9 24.86

369.5 9.4 37.13

353.4 9.2 36.05

273.9 8.1 30.43

TABLE 3: Content of Dendritic (D), Linear (L), and Terminal (T) Repeat Units of HBP-11, HBP-8, and HBP-3 Along with Their Focal Point Acid Units Determined from 13C NMR

D 3665 TA, Kratos Instruments), and the solvent was removed by forced air evaporation at ambient temperature. Subsequently, 0.3 µL of the matrix solution was placed above the sample spot covering the analyte film, and again, the solvent was removed by forced air evaporation at ambient temperature (30 °C).26 Gel Permeation Chromatography. GPC measurements were carried out using Shimadzu LC 10ATVP Series (Japan) with a refractive index detector to determine the molecular weight of

the polymer solutions. The flow rate of the mobile phase was kept at 1.0 mL/min. FTIR. Small and diluted drops of the hyperbranched polyester liquid samples were spread on a dry KBr disc and stored at ambient temperature and humidity for 2 days. Later on, the sample coated KBr disks were vacuum dried at 80 °C for a long time to reduce the levels of solvent and adsorbed water in the samples prior to recording the FTIR spectra under ambient conditions on a Thermo Nicolet Nexus 670 spectrometer. Each


Figure 7. Expansion of the quaternary carbon region of the 13C NMR spectrum of hyperbranched polyester HBP-5, HBP-9, and HBP-12. The peaks of terminal (T), dendritic (D), and linear (L) repeat units are all present along with their acid functional repeat units.

Jena et al.

Figure 9. DMSO-d6.

sample was scanned 64 times with a resolution setting of 4 cm-1 within the range of 400-4000 cm-1. Molecular Simulations Molecular simulations were performed using the MS modeling 3.1 software, Accelrys, San Diego, CA. The simulation methodology includes molecular mechanics (MM) and molecular dynamics (MD) calculations using the DISCOVER module. MD was performed using the COMPASS (condensed phase optimized molecular potentials for atomistic simulation studies)

C NMR spectra of HBP-6, HBP-10, and HBP-13 in

TABLE 4: Comparison of GPC Data with MALDI-TOF code

Mw (g/mol)a

Mn (g/mol)a

HBP-3 HBP-8 HBP-5 HBP-9 HBP-12

1840 2148 963 2010 2712

1362 1432 698 1382 1780

a

Figure 8. 1H NMR spectra of HBP-6, HBP-10, and HBP-13 in DMSO-d6.

13

M (g/mol)b 1320 1437 970 1435 N/A

PDIc ) Mw/Mn 1.35 1.51 1.37 1.45 1.52

From GPC. b From MALDI-TOF. c PDI ) polydispersity index.

forcefield,27 which is one of the first ab initio forcefield approaches that was parametrized and validated using condensed phase properties. Minimization was performed using the steepest descend approach followed by the conjugate gradient method. The temperature in all the simulations was equilibrated with the Andersen algorithm.28 The velocity Verlet algorithm29 was used to integrate the equations of motion. The group-based cutoffs were used with the explicit atom sums being calculated to 9.5 Å. The tail correction was applied to nonbonded interactions during the MD run. The hyperbranched polyols were generated up to three generations. A simulated structure of the second generation DMPA with a core moiety of TMP, glycerol, and pentaerythritol is shown in Figure 1a-c. Structures were minimized, and amorphous cells were constructed based on the respective densities of the selected polyols. The method used in constructing the amorphous cell module of MS modeling was the combined use of an algorithm developed by Theodorou and Suter30 and the scanning method of Meirovitch.31 Chain conformations were assumed to resemble those of the unperturbed chains that are found with significant probability in the bulk. Initially, the proposed structure was generated using the rotational isomeric state (RIS) model of Flory,32 describing the conformations of the unperturbed chains. To avoid excessive overlaps between chains, modified conditional probabilities were used, which accounted for nonbonded interactions between the atoms to be placed and the rest of the system. Turning on the potential interactions minimized the initial structures such that more severe overlaps were relaxed first, and then gradually, the



Figure 10. MALDI-TOF MS of HBPs prepared from pentaerythritol and DMPA. (a) HBP-3 (n ) 4-22) and (b) HBP-8 (n ) 7-27).

minimum was reached by switching on the full potential. In the scanning method, all possible continuations of the growing chains were considered while computing the conditional probabilities. The constructed amorphous cells were minimized to a convergence level of 0.01 kcal/mol/Å using the same method as described previously. MD simulations under a constant volume and temperature (NVT) ensemble were performed using the DISCOVER program. Systems built with 3-D periodicity were equilibrated in the NVT ensemble at a temperature of 300 K. MD runs for 50 ps were performed to remove the unfavorable local minima that had high energies. Subsequently, systems were subjected to 300 ps of dynamics with the trajectories being saved every 0.1 ps during the last half of the run to calculate the physical properties of interest. Results and Discussion NMR Analysis. PE + DMPA. Figure 2 shows the 1H NMR spectra of HBP-3, HBP-8, and HBP-11 in DMSO-d6. Resonance assignments (δ ppm) are as follows: 1.05-1.22 (-CH3), 2.2 (-CH2OH), 3.4-3.8 (-CH2OH), 3.4 (H2O), 3.9-4.2 (-COOCH2), 4.64 (-OH)T, and 4.8 (-OH)L.33,34 Methyl group protons of DMPA in the terminal, linear, and dendritic repeat units resonate at 1.05, 1.12, and 1.22 ppm, respectively, as shown in Figure 3. The 13C NMR spectra of HBP-3, HBP-8, and HBP-11 in DMSO-d6 are shown in Figure 4. Peak assignments (δ ppm) are as follows: 15-20 (methyl carbon), 45-52 (quaternary carbons), 62-68 ppm (methylene groups), and 173-178 (carbonyl groups).35,36 The top of Figure 4 is the expansion of the quaternary carbon region of hyperbranched polyesters HBP-

3, HBP-8, and HBP-11. There are δ 50.6 ppm (terminal), δ 48.6 ppm (linear), and δ 46.6 ppm (dendritic) repeating units. There are δ 49.8 ppm (acid terminal), δ 48.0 ppm (acid linear), and δ 46.1 ppm (acid dendritic) repeating units. In the 13C NMR spectrum, quaternary carbons of different repeating units are easily distinguished from each other (Figure 4, top). The acid dendritic units (also called the focal point acid unit) resonate at δ ) 46.1 ppm,35 where the quaternary carbon is attached to an unreacted acid group and two reacted hydroxyl groups. The more electron withdrawing character of the acid group as compared to that of the ester group makes the quaternary carbon attached to the acid group resonate at a slightly lower chemical shift than that attached to the ester group. Their peaks appear at different chemical shifts depending upon the degree of substitution of the repeating unit. The relative amounts of individual repeating units were determined by comparing the integrals of different peaks following the procedure of Magnusson et al.,36 and these are given in Table 3. TMP + DMPA. Figure 5 shows the 1H NMR spectra of HBP-5, HBP-9, and HBP-12 in DMSO-d6, and the expansion of the methyl group is shown at the top (Figure 5). Resonance assignments (δ ppm) are as follows: 1.06 (terminal -CH3), 1.12 (linear -CH3), 1.21 (dendritic -CH3), 2.1 (-CH2OH), 3.4-3.7 (-CH2OH), 3.3 (H2O), 3.9-4.3 (-COOCH2), 4.65 (-OH)T, and 4.9 (-OH). The 13C NMR spectra of HBP-5, HBP-9, and HBP12 in DMSO-d6 are shown in Figure 6. The peak assignments (δ ppm) are as follows: 15-20 (methyl carbon), 45-52 (quaternary carbons), 62-68 ppm (methylene groups), and 173178 (carbonyl groups). Figure 7 shows the expansion of the quaternary carbon region of the 13C NMR spectrum of hyper-


Jena et al.

Figure 11. MALDI-TOF MS of HBPs prepared from TMP and DMPA. (a) HBP-5 (n ) 7-17) and (b) HBP-9 (n ) 8-28).

Figure 12. MALDI-TOF MS of HBP-10 prepared from glycerol and DMPA (n ) 8-21).

branched polyesters HBP-5, HBP-9, and HBP-12. The peaks of terminal (T), dendritic (D), and linear (L) repeat units are all present along with their acid functional repeat units. Glycerol + DMPA. 1H NMR spectra of HBP-6, HBP-10, and HBP-13 in DMSO-d6 are shown in Figure 8. Resonance assignments (δ ppm) are as follows: 1.06-1.21 (-CH3), 2.1 (-CH2OH), 3.4-3.7 (-CH2OH), 3.3 (H2O), 3.9-4.3 (-COOCH2), 4.65 (-OH)T, and 4.9 (-OH)L. Figure 9 shows the 13C NMR spectra of HBP-6, HBP-10, and HBP-13 in DMSO-d6. Peak

assignments (δ ppm) are as follows: 15-20 (methyl carbon), 45-52 (quaternary carbons), 62-68 ppm (methylene groups), and 173-178 (carbonyl groups). MALDI-TOF MS Analysis. MALDI-TOF MS has become a rapid and convenient tool for qualitative, semiquantitative, and quantitative analysis of hyperbranched polymer samples. The main ion series observed for hyperbranched polyesters corresponds to the oligomers formed in successive condensation reactions. An oligomer derived from nAB2 units is the result of n - 1 condensation steps, and each successive condensation is accompanied by the loss of a single water molecule. In Figure 10a,b, the MALDI-TOF MS of HBPs prepared from pentaerythritol and DMPA with the first (HBP-3) and second generations (HBP-8), respectively, are displayed. The mass of a particular oligomer was calculated by consideration of the polymerization process. The theoretical calculation of mass, m/z, in MALDI-TOF MS of HBPs derived from pentaerythritol and DMPA is shown in eq 1. The mass spectra of the HBP samples showed a clear distribution of oligomers with increasing degrees of polymerization. Every spectrum is made up of a specific sequence of signals arranged at intervals of m/z ) 116 amu (spacing between peaks or mass progression) from each other, which corresponds to [M(DMPA) - MH2O]. The major mass distribution of HBPs can be written as m/z ) 136 + 23 + 116n, and n, which is the number of repeat units, varies from 4-22 for HBP-3 and 7-27 for HBP-8. The observed peaks are due to [(MDMPA - MH2O)n + Na]+ and [(MDMPA - MH2O)n + K]+



ions with a separation of 16 amu between them because the ionization of polymers is generally accomplished by the formation of polymer-metal ion adducts. The weak peaks located 16 amu above the linear [(MDMPA - MH2O)n + Na]+ ion peaks are due to [(MDMPA - MH2O)n + K]+ ions. Even if no special cation salts are added, HBP samples are ready to form both sodium and potassium adducts due to the high affinity of HBPs with alkali salts as well as the presence of trace amounts of sodium and potassium coming from the matrix generated [(MDMPA - MH2O)n + Na]+ and [(MDMPA MH2O)n + K]+, which has produced the ion peaks. Figure 11a,b shows the MALDI-TOF MS of HBP-5 (n ) 7-17) and HBP-9 (n ) 8-28), respectively, prepared from TMP and DMPA. The corresponding theoretical calculation of m/z is given in eq 2. Here, we have also observed a series of peaks arranged at intervals of m/z ) 116 from each other. The actual mass progression of 116n + 134 + 23 and 116n + 134 + 39 corresponding to sodium and potassium ions, respectively, can be seen in Figure 11. On the other hand, Figure 12 and eq 3 show the MALDI-TOF MS of the second generation HBP prepared from glycerol and DMPA (i.e., HBP-10 and the theoretical calculation of m/z, respectively (here, n varies from 8-21)).

Figure 13. FTIR spectra of HBP-3, HBP-8, and HBP-11 recorded at room temperature.

Mn ) MPENTA + (MDMPA - MH2O)n + MNa+ ) 136 + (134 - 18)n + 23 ) 136 + 116n + 23

(1)

Mn ) MTMP + (MDMPA - MH2O)n + MNa+ ) 134 + (134 - 18)n + 23 ) 134 + 116n + 23

(2)

Mn ) Mglycerol + (MDMPA - MH2O)n + MNa+ ) 92 + (134 - 18)n + 23 ) 92 + 116n + 23

(3)

where MPENTA, MTMP, Mglycerol, MDMPA, and MH2O indicate molecular masses of the monomer pentaerythritol, trimethylol propane, glycerol, core DMPA, and water, respectively. A minor series of peaks is observed at 18 amu behind the main series, and this is attributed to the loss of a further single water moiety to form the cyclic polymer. GPC Analysis. GPC is one of the most widely used techniques for the characterization of polymers used in coatings and industrial applications. It is increasingly being used for the determination of the molecular weight and polydisperity index (PDI) of the polymers and to show the quality of the obtained product. Table 4 compares the results of molecular weights obtained by MALDI and GPC techniques, which are in good agreement. The PDI data suggest their heterogeneous nature. FTIR Analysis. FTIR is an analytical tool for identifying functional groups that provide information about the chemical makeup of the materials. FTIR also provides a rapid means of identifying functional groups and substances, characterizing contaminants, and comparing material properties. In the present investigation, FTIR experiments were performed to determine the extent and type of hydrogen bonding in HBPs as well as to compare the structure-property relationship with the stuctural variables used.

Figure 14. FTIR spectra of HBP-5, HBP-9, and HBP-12 recorded at room temperature.

FTIR spectra of HBP-3, HBP-8, and HBP-11 recorded at ambient temperature are shown in Figure 13, while spectra of HBP-5, HBP-9, and HBP-12 are shown in Figure 14. The strong and broad peak around 3100-3700 cm-1 confirmed a high concentration of hydroxyl groups in HBPs. A wide band in the O-H stretching region is attributed to the summation of several contributions of different types of hydroxyl groups in hydroxyl containing compounds. The ester carbonyl stretching vibrations ν(CdO) between 1600 and 1800 cm-1 are composed of free and hydrogen-bonded carbonyls. The vibration of methyl and methylene groups attached to the quaternary carbon of DMPA in HBP results in an antisymmetric νas(CH3) and symmetric νs(CH3) stretching of methyl groups at 2948 and 2867 cm-1, respectively. The other bands at 2906 and 2835 cm-1 belong to antisymmetric νas(CH2) and symmetric νs(CH2) stretching of methylene groups, respectively.37 FTIR


Jena et al. and free O-H, respectively. It was found that the majority of the hydroxyl groups are H-bonded rather than free O-H and that the hydrogen-bonding strength follows the sequence: HBP13 > HBP-10 > HBP-6. Other observed bands of the hyperbranched polyesters of HBP-6, HBP-10, and HBP-13 are due to the characteristic stretching vibrations of the methyl groups (νsCH3: 2868 cm-1 and νasCH3: 2948 cm-1), methylene groups (νsCH2: 2836 cm-1 and νasCH2: 2906 cm-1), and ester groups (free CdO at 1690 cm-1 and H-bonded CdO at 1736 cm-1 and C-O at 1130 cm-1) in the spectrum.39 The peak at 1465 cm-1 is due to a CH3 asymmetric deformation vibration. The C-O stretching and O-H deformation of COOH appeared at 1226 cm-1. Cohesive Energy Density (CED) and Solubility Parameter (δ) Calculations. Cohesive properties of the hyperbranched polymers are difficult to determine experimentally because the chosen polymers are insoluble, have high glass transition temperatures, and are sometimes poorly characterized. Therefore, we have adopted the MD simulations to obtain useful information of higher quality than other methods. MD simulations were performed to calculate the cohesive energy density (CED) of the hyperbranched polyesters. To validate the simulation protocols using the COMPASS forcefield, solubility parameters of the polymers were calculated40 from the CED data using eq 4. CED is defined as

CED ) (Ecoh/Vmol)

(4)

The Hildebrand solubility parameter, δ, was calculated as

δ ) (Ecoh/Vmol)1/2

Figure 15. FTIR spectra of HBP-6, HBP-10, and HBP-13 recorded at 30 °C.

spectra of the hyperbranched polyesters HBP-6, HBP-10, and HBP-13 in the full mid-infrared spectral region at ambient temperature are given in Figure 15. A wide band in the O-H stretching region is due to the formation of different types of hydrogen-bonding structures as well as the free O-H groups shown at the top of Figure 15. The distribution of hydrogenbonded O-H groups at different distances, geometries, and hydrogen-bond formation with different groups in different environments would induce a change in the force constant of the O-H group, which in turn, broadens the O-H stretching region.38 The peak deconvolution of the second generation HBPs showed the presence of four different types of O-H bands, due to the presence of a large number of polar hydroxyl groups that implies a great possibility of molecular interactions through H-bonding. From the chemical structure of the polyester used as shown at the top of Figure 15, the existence of hydrogen bonding between molecular segments of the same HBP molecule (intramolecular) or between neighboring HBP molecules (intermolecular) was expected as a result of the interaction between hydroxyl groups or carbonyl groups having a proton acceptor character and the hydroxyl endgroups. To obtain information about the nature of hydrogen bonding, we have performed curve fitting of the hydroxyl vibration region (i.e., 3100-3700 cm-1). The deconvoluted bands at 3195, 3338, 3485, and 3578 cm-1 are due to hydrogen bonding with the esteric CdO‚‚‚HsO, esteric OdCsO‚‚‚HsO, inter-/intramolecular OsH‚‚‚OsH,

(5)

Calculated values of CED and δ of the HBPs are given in Table 2. Surface Properties. Surface properties of the hyperbranched polyols are important in understanding their adhesion and wettability characteristics.41 However, the surface energy plays a key role in polymer processing and blending.42 Recently, Orlicki et al.43 investigated the role of molecular architecture and endgroup functionality on the surface energy of HBPs. The surface energy decreases depending upon the degree of substitution to the endgroups.44 In this work, we report the results of hyperbranched polyesters based on DMPA for which the surface energy was calculated from the solubility parameter values using the empirical equation of Zisman.45

γ ) 0.75(Ecoh)2/3

(6)

Surface energies calculated from eq 6 along with CED values obtained from bulk simulations are also included in Table 1. Notice that the surface energy decreases as the degree of substitution of DMPA increases in each of the core moieties pentaerythritol, trimethylol propane, and glycerol, which in turn substantiate the effect of endgroups in dendritic macromolecules. Spectral Analysis. Vibrational spectra were calculated by the local mode method, in which slow motions were modeled by classical simulation techniques (MD) and the fast vibrations were treated separately by stopping the MD run several times and calculating the molecular vibrations by quantum mechanical calculations for several configurations and each vibration separately. During the quantum mechanical calculations, local mode vibrations were considered as quantum oscillators embedded in a mean field of the surrounding atoms where the effective potential can be given as


Veff(Q) ) K0 + K2Q2 + K3Q3 + K4Q4


(7)

Here, Q defines the coordinates in the configuration space generated by the molecular dynamics simulation. The simulated vibrational frequencies were calculated for third generation hyperbranched polyols of pentaerythritol, trimethyl propane, and glycerol with DMPA. The frequency corresponding to the hydroxyl groups is computed to be 3400-3500 cm-1, and for carbonyl group, it was found to be 1650-1705 cm-1. Furthermore, the computed vibrational frequency compares well with the experimentally observed FTIR spectra (see FTIR Analysis). Conclusion In the present investigation, the first, second, and third generations of hyperbranched polyesters were prepared by A2B plus a core molecule approach, where A2B is DMPA and the core is TMP, pentaerythritol, or glycerol. The synthesized HBPs were characterized by NMR, MALDI, GPC, and FTIR. MALDI proved the cyclization of the hyperbranched polyester, which takes place during the esterification reaction. GPC data confirmed the molecular weight results obtained from MALDI. To better understand the relative influence of hydrogen bonding on the structure-property relationship, we have deconvoluted the FTIR O-H spectral zone. The band deconvolution of the O-H profiles allows the identification of the number of overlapping signals. The present results suggest that the extent of hydrogen bonding and the free O-H group depends on the chemical structures of HB generation as well as the core molecule used in the preparation of HBPs. Molecular modeling simulations employing molecular mechanics and NVT molecular dynamics offered cohesive energy densities, solubility parameters, and surface energy data of the HBPs. Further, vibrational analysis has been computed for hydroxyl and carbonyl groups, which compared well with the experimental FTIR frequencies. The present approach offers new insights into the physical properties of HBPs. The modeling strategies employed are novel on the chosen systems. Acknowledgment. The authors thank the University Grants Commission, New Delhi (Grant F1-41/2001/CPP-II) for financial support to establish the Center of Excellence in Polymer Science (CEPS). This work is a collaborative effort between the Center of Excellence in Polymer Science, Dharwad and the Indian Institute of Chemical Technology, Hyderabad under the MoU. The authors appreciate help from Dr. D. Anjali. This paper is Center of Excellence in Polymer Science Communication 145. References and Notes (1) Breton, M. P.; US Patent 5,266,106, 1998. (2) Frecht, J. M.; Tomalia, D. A. Dendrimers and Other Dendritic Polymers; Wiley: New York, 2001.

(3) Malmstram, E.; Johansson, M.; Hutt, A. Macromolecules 1995, 28, 1968. (4) Perce, V.; Kawasumi, M. Macromolecules 1992, 25, 3843. (5) Uhrich, K. E.; Boegeman, S.; Frechet, J. M.; Turner, S. R. Polym. Bull. 1997, 25, 551. (6) Kim, Y. H.; Webster, O. W. Macromolecules 1992, 25, 5561. (7) Spindler, R.; Frechet, J. M. Macromolecules 1993, 26, 4809. (8) Nunez, C. M.; Chiou, B. S.; Andrady, A. L.; Khan, S. A. Macromolecules 2000, 33, 1720. (9) Huybrechts, J.; Dusek, K. Surf. Coat. Int. 1998, 81, 117. (10) Sheth, J. P.; Unal, S.; Yilgor, E.; Yilgor, I.; Beyer, F. L.; Long, T. E.; Wilkes, G. L. Polymer 2005, 46, 1018. (11) Nu´n˜ez, E.; Ferrando, C.; Malmstro¨m, E.; Claesson, H.; Werner, P. E.; Gedde, U. W. Polymer 2004, 45, 5251. (12) Jannerfeldt, G.; Boogh, L.; Månson, J. A. E. Polymer 2000, 41, 7627. (13) Asif, A.; Shi, W. Eur. Polym. J. 2003, 39, 933. (14) Johansson, M.; Malmstro¨m, E.; Hult, A. J. Polym. Sci., Part A: Polym. Chem. 1993, 31, 619. (15) Malmstro¨m, E.; Hult, A. Macromolecules 1996, 29, 1222. (16) Magnusson, H.; Malmstro¨m, E.; Hult, A. Macromolecules 2000, 33, 3099. (17) Zˇ agar, E.; Zˇ igon, M. Macromolecules 2002, 35, 9913. (18) Komber, H.; Ziemer, A.; Voit, B. Macromolecules 2002, 35, 3514. (19) Dusˇek, K.; Sˇ omva´rsky, J.; Smre`kova´, M.; Simonsick, W. J.; Wilczek, L. Polym. Bull. 1999, 42, 489. (20) Hsieh, T. T.; Tiu, C.; Simon, G. P. Polymer 2001, 42, 1931. (21) Malmstro¨m, E.; Hult, A.; Gedde, U. W.; Liu, F.; Boyd, R. H. Polymer 1997, 38, 4873. (22) Malmstro¨m, E.; Liu, F.; Boyd, R. H.; Hult, A.; Gedde, U. W. Polym. Bull. 1994, 32, 679. (23) Zhu, P. W.; Zheng, S.; Simon, G. Macromol. Chem. Phys. 2001, 202, 3008. (24) Rogunova, M.; Lynch, T. Y. S.; Pretzer, W.; Kulzick, M.; Hiltner, A.; Baer, E. J. Appl. Polym. Sci. 2000, 77, 1207. (25) Mehl, J. T.; Murgasova, R.; Dong, X.; Hercules, D. M.; Nefzger, H. Anal. Chem. 2000, 72, 2490. (26) Raju, N. P.; Mirza, S. P.; Vairamani, M.; Ramulu, A. R.; Pardhasaradhi, M. Rapid Commun. Mass Spectrom. 2001, 15, 1879. (27) Sun, H. J. Phys. Chem. B 1998, 102, 7338. (28) Andersen, H. C. J. Chem. Phys. 1980, 72, 2384. (29) Verlet, L. Phys. ReV. 1967, 159, 98. (30) Theodorou, D. N.; Suter, U. W. Macromolecules 1985, 18, 1467. (31) Meirovitch, H. J. J. Chem. Phys. 1983, 79, 502. (32) Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. (33) Jiang, G.; Wang, L.; Chen, T.; Dong, X.; Yu, H.; Wang, J.; Chen, C. J. Polym. Sci., Part A: Polym. Chem. 2005, 43, 5554. (34) Komber, H.; Ziemer, A.; Voit, B. Macromolecules 2002, 35, 3514. (35) Jiang, G.; Wang, L.; Yu, H.; Chen, C.; Dong, X.; Chen, T.; Yang, Q. Polymer 2006, 47, 12. (36) Magnusson, H.; Malmstrom, E.; Hult, A. Macromolecules 2000, 33, 3099. (37) van Benthem, R. A. T. M. Prog. Org. Coat. 2000, 40, 203. (38) Zagar, E.; Grdadolnik, J. J. Mol. Struct. 2003, 658, 143. (39) Mikhaylova, Y.; Adam, L.; Haussler, K. J.; Eichhorn, B.; Voit, J. Mol. Struct., in press. (40) Bicerano, J. Prediction of Polymer Properties; Marcel Dekker: New York, 1993. (41) Stamm, M. AdV. Polym. Sci. 1992, 100, 357. (42) Sammler, R. L.; Dion, R. P.; Carriere, C. J.; Cohen, A. Rheol. Acta 1992, 31, 554. (43) Orlicki, J. A.; Viernes, N. O. L.; Moore, J. S. Langmuir 2002, 18, 9990. (44) Beyerlein, D.; Belge, G.; Eichhorn, K. J.; Gauglitz, G.; Grundke, K.; Voit, B. I. Macromol. Symp. 2001, 164, 117. (45) Zisman, W. A. Contact Angle, Wettability, and Adhesion; Advances in Chemistry Series 43; American Chemical Society: Washington, DC, 1964.

Hyperbranched Polyesters: Synthesis ... - ACS Publications

Recommend Documents