《材料导论》课程教学资源（文献资料）FT-IR and XPS studies of polyurethane-urea-imide coatings.pdf_大学文库

89 COATINGS ELSEVIER Progress in Organic Coatings 55(006)231-243 ww..com/co FT-IR and XPS studies of polyurethane-urea-imide coatings Aswini K.Mishra,D.K.Chattopadhyay,B.SreedharbK.V.S.N.Raju.* Received 22 August 2005;accepted 8 November2005 Abstract A series of polvurethane(PU)-urea-imide coatings were synthesized by a systematic three-step reaction process.Initially isocyanate terminated wre prepared by(oly dianhydride (PMDA)from the xcs NCO groups n the PU prepolymer th NCOnydridetoofnd the surplus NCOcon NC miionsmoisue urNCO groups were complctlyrctedth oymers,and the changmoeul H-bonding mrion the PU-urmiducmucd bydecoo orts the FT-IR obser tions 2005 Elsevier B.V.All rights reserved. Keywonds:Polyurethane-imide:FT-IR.XPS 1.Introduction triol or tri-isocyanate,incorporation of an appropriate amount of aromatic rings into the PU backbone,partial replacement Thermoplastic are block of urethane d. emic eren e sepa two Poly ral or pol use nates in addition with chain der form the hard se mide structure in the backbone: remarkable heat and Hard segment acts as higher modulus filler to the soft matrix chemical resistance.high T and Tm values and hardness 16] and thereby improves the thermo-mechanical properties of PU However,there are some disadvantages in polyimide applica- coatings.In order to increase the life span of PU coatings from tions with respect to their brittleness and low processability the perspective of its application,a further improvement of the Therefore,a synergistic well-balanced property profile can be PU matrix is necessary,since a coating material should warrant achieved by tailoring the PU formulation containing appropriate amount of imide rings wear and unng its service er a n Poly(uret ine-imi n be prepared var ous ways ed PU 17221.reacting acid anhvdride with an amine terminated PU renolymers reacting isocvanate terminated pll prepolymer with a diols or diacids containing imide groups in its struc ture [23-25],intermolecular Diels-Alder reaction of molecules

Progress in Organic Coatings 55 (2006) 231–243 FT-IR and XPS studies of polyurethane-urea-imide coatings Aswini K. Mishra a, D.K. Chattopadhyay a, B. Sreedhar b, K.V.S.N. Raju a,∗ a Organic Coatings and Polymers Division, Indian Institute of Chemical Technology, Hyderabad 500007, India b Inorganic and Physical Chemistry Division, Indian Institute of Chemical Technology, Hyderabad 500007, India Received 22 August 2005; accepted 8 November 2005 Abstract A series of polyurethane (PU)-urea-imide coatings were synthesized by a systematic three-step reaction process. Initially isocyanate terminated PU prepolymers were prepared by reacting soft segments such as polyester polyols (prepared from neopentyl glycol, adipic acid, isophathalic acid and trimethylol propane) or polyether polyols (polypropylene glycol-1000) with hard segments such as 2,4-toluene diisocyanate or isophorone diisocyanate with NCO/OH ratio of 2:1. Heterocyclic imide ring into the PU backbone was incorporated by co-polymerization with pyromellitic dianhydride (PMDA) from the excess NCO groups in the PU prepolymer with an NCO/anhydride ratio of 1:0.5 and the surplus NCO content after imidization was moisture cured. PU-urea-imide coatings were also obtained by partial chain extension of the excess NCO groups in the NCO terminated PU-imide copolymers, and the remaining excess NCO groups were completely reacted with atmospheric moisture. The obtained polymers were analyzed with Fourier transform-infrared (FT-IR) and angle resolved X-ray photoelectron spectroscopy (AR-XPS). The type and change in intermolecular H-bonding interaction in the PU-urea-imide films with structural variables was identified by deconvolution of the FT-IR spectra using Origin 6.0 software through Gaussian curve-fitting method. The FT-IR analyses of the PU-urea-imide coating films show dependence of phase separation on the nature of chain extender. Surface characterization data from AR-XPS suggests the dependence of phase segregation behaviour on the nature of the chain extender, which also supports the FT-IR observations. © 2005 Elsevier B.V. All rights reserved. Keywords: Polyurethane-imide; FT-IR; XPS 1. Introduction Thermoplastic polyurethane (PU) elastomers are block copolymers comprising of alternating soft and hard segments, which due to the structural difference separate into two phases. Polyethers or polyester polyols are generally used as soft segments that give extensibility to the polymer, whereas diisocyanates in addition with chain extender form the hard segment. Hard segment acts as higher modulus filler to the soft matrix and thereby improves the thermo-mechanical properties of PU coatings. In order to increase the life span of PU coatings from the perspective of its application, a further improvement of the PU matrix is necessary; since a coating material should warrant high thermal resistance, strength and stiffness to overcome the stress, wear and fatigue load during its service. Earlier a number of attempts have been made to improve material properties by the addition of appropriate amount of crosslinker such as a ∗ Corresponding author. Tel.: +91 40 27193991; fax: +91 40 27193991. E-mail address: kvsnraju@iict.res.in (K.V.S.N. Raju). triol or tri-isocyanate, incorporation of an appropriate amount of aromatic rings into the PU backbone, partial replacement of urethane segment with urea functions to improve the interchain association or by chemical modification of the backbone by introducing stable heterocyclic groups like imide, oxazolidone, triazine, and phosphazene [1–15]. Polyimides (PI) are the most important members of heterocyclic polymers composed of imide structure in the backbone; possess remarkable heat and chemical resistance, high Tg and Tm values and hardness [16]. However, there are some disadvantages in polyimide applications with respect to their brittleness and low processability. Therefore, a synergistic well-balanced property profile can be achieved by tailoring the PU formulation containing appropriate amount of imide rings. Poly(urethane-imide) copolymers can be prepared in various ways. These include the reaction of isocyanate or blocked isocyanate terminated PU prepolymers with acid anhydrides [17–22], reacting acid anhydride with an amine terminated PU prepolymers, reacting isocyanate terminated PU prepolymer with a diols or diacids containing imide groups in its structure [23–25], intermolecular Diels–Alder reaction of molecules 0300-9440/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.porgcoat.2005.11.007

232 A.K.Mishra et al.Progress in Organic Coatings 55 (2006)231-243 n imide it 6 h and boneth blocked PU prepolymer 37.reaction of hydroxy another6h.The reaction was monitored periodically by check- containing polyimides with blocked PU prepolymers 38,39] ing the acid value and continued till the acid value reached P polyim poly(amic a reac ies was to improve the the We intend to design the pres 2.2.2.Synthesis of the NCO terminated polyurethane (PU) used.For that,we have synthesized different PU-urea-imide copolymers from polye 250 ml four necked re und bottomed flask equipped with a ther stirrer and a nitrogen PU)of PPG and PE were p ared from IPDI and TDI with polvester polvol and PPG was removed by azeotropic method NCO:OHratioof2:1. Tien.NCO-eminaiedPU.iidcecopoly along with toluene.Fifty grams(0.107mol)of polyester polyo (PMDA) Uby th C ome 80g of MIBK was s added slov 02 D! of 1:0.5.The NCO/OH ratio 2:1Addition was started wh the ton chain extended with diols,diamines and partially moisture reached at about 50C and was continued for 2h with stirring. nue 2.Experimental 50g of PE with 47.5g of IPDL,and 50g of PPG with 17.4go 2.1.Materials Polypropylene glycol-1000 (PPG). an 2.2.3.Synthes s of pllimide resin (TMP)isophorone diisocyanate (IPDI: After the PU prepolymer was prepared(say,PE/IPDD),the kettle was sulfone DS) enyl ethe (DDE).tol 105) diisocyanate (TDI:mixture of2,4-and 2.6-isomer in9:1 ratio). into2-3 parts and added batch by batch at a temperature of 40 con yme nart was cor from Merck India Pvt Ltd.(Mumbai,India),isopthalic acid IS (M hyl isobuty second and third pa bai,India)were used.PMDA was purified by recrystallization PPG/TDL from acetic anhydride. followed b sublimation.The solvents 2.2.4.Preparat was mixed with the synthesized PU-imide resins and the films 2.2.Method were casted with different chain extenders on tin foil by using a powe 200 ven automatic appl The supported aP阳paG NPG moisture cured pilurea-imide films thus obtained were amalga mated and bottom ofthe unsupported films wascarefully cleaned with muslin cloth and the free films were stored at laboratory

232 A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 containing bis(2-furyl-carbamate) units with bismaleimides [26–28], thermal crosslinking of PU prepolymers containing maleimide functions [29–31], reaction of diisocyanates containing build-in imide groups with different polyols [32–36], reaction of an epoxy resin containing imide structure in its backbone with blocked PU prepolymers [37], reaction of hydroxyl containing polyimides with blocked PU prepolymers [38,39], reaction of polyimide precursor, poly(amic acid), with blocked PU prepolymers [40,41], and blending of thermoplastic PUs with poly(ether imide)s[42]. The main outcome of all these studies was to improve the thermal stability or thermo-mechanical properties. We intend to design the present study rather different way to assess the importance of hydrogen bonding on the phase separation phenomenon with the structural variables used. For that, we have synthesized different PU-urea-imide copolymers from polyester polyol (PE) and PPG. Polyester polyol was synthesized from AA, IPA, NPG and TMP with hydroxyl value of 240. NCO terminated PU prepolymers (NCOPU) of PPG and PE were prepared from IPDI and TDI with NCO:OH ratio of 2:1. Then, NCO-terminated PU-imide copolymers were prepared from NCO-PU by the reaction with pyromellitic dianhydride (PMDA) with excess NCO:anhydride ratio of 1:0.5. The synthesized PU-imide copolymers were partially chain extended with diols, diamines and partially moisture cured. The obtained films were used for FT-IR and AR-XPS evaluation. 2. Experimental 2.1. Materials Polypropylene glycol-1000 (PPG), trimethylol propane (TMP), isophorone diisocyanate (IPDI: Z and E isomer in 3:1 ratio), neopentyl glycol (NPG) and dibutyltin dilaurate (DBTL) from Aldrich (Milwaukee, WI); 4,4 -diamino-diphenyl sulfone (DDS), 4,4 -diamino-diphenyl ether (DDE), toluene diisocyanate (TDI: mixture of 2,4- and 2,6-isomer in 9:1 ratio), 1,4-butane diol (BD) from Fluka Chemical Corp. (Ronkonkoma, NY); 1,2-propane diol (PD), adipic acid (AA), triethylamine, dimethyl formamide and sulfur free toluene from S.D. Fine chemicals (Mumbai, India); 1-methyl-2-pyrrolidinone (NMP) from Merck India Pvt. Ltd. (Mumbai, India), isopthalic acid (IPA) from Sisco Chemicals (Mumbai, India); methyl isobutyl ketone (MIBK) from Ranbaxy (Mumbai, India) and pyromellatic dianhydride (PMDA) from Spectrochem Pvt. Ltd. (Mumbai, India) were used. PMDA was purified by recrystallization from acetic anhydride, followed by sublimation. The solvents were freed from moisture using 4 A molecular sieves before ˚ use. 2.2. Method 2.2.1. Synthesis of polyester polyol Polyester polyol (PE) was synthesized by charging NPG (2.2 mol), AA (0.825 mol), IPA (0.825 mol) and TMP (0.35 mol) into a four necked flask placed over an isomentel bath, and equipped with a mechanical stirrer, thermometer, nitrogen inlet and dean-stark apparatus. The reactant mixture was slowly heated up to 160 ◦C. After complete melting of the reactants, the temp was increased to 180–190 ◦C with constant nitrogen flow and maintained at that temperature for about 6 h and gradually increased up to 210 ◦C and the reaction was continued for another 6 h. The reaction was monitored periodically by checking the acid value and continued till the acid value reached below five. The hydroxyl value was determined by the acetic acid/pyridine titration method and was 280 at the end of the reaction. 2.2.2. Synthesis of the NCO terminated polyurethane (PU) prepolymer The NCO terminated PU prepolymer of PPG and the synthesized PE were prepared by carrying out the reaction in a 250 ml four necked round bottomed flask equipped with a thermometer, dropping funnel, mechanical stirrer and a nitrogen inlet. Before carrying out the reaction the absorbed moisture in polyester polyol and PPG was removed by azeotropic method along with toluene. Fifty grams (0.107 mol) of polyester polyol in 80 g of MIBK was added slowly through a dropping funnel to the reaction kettle containing 37.2 g (0.214 mol) of TDI at a NCO/OH ratio 2:1. Addition was started when the temperature reached at about 50 ◦C and was continued for 2 h with stirring. After complete polyester addition, the reaction was further continued for 3 h at a temperature of 70–75 ◦C. A slight viscous resin was obtained. The same process was repeated by reacting 50 g of PE with 47.5 g of IPDI, and 50 g of PPG with 17.4 g of TDI, respectively. The resins prepared were named as PE/TDI, PE/IPDI and PPG/TDI, respectively. 2.2.3. Synthesis of PU-imide resin After the PU prepolymer was prepared (say, PE/IPDI), the reaction kettle was cooled to room temperature while stirring. Required amount of PMDA (excess-NCO:anhydride ratio was 1:0.5) was dissolved in minimum amount of DMF and divided into 2–3 parts and added batch by batch at a temperature of 40–45 ◦C. After complete addition, the imidization reaction was continued for additional 1 h. The synthesized PU-imide polymer was divided into three parts. The excess NCO group of the first part was completely reacted with atmospheric moisture and is named as PUI-2. Fifty percent of the available NCO groups from the second and third part were chain extended with PD and DDS and were named as PUI-3 and PUI-6, respectively. Similarly, PU-imide films were obtained from PE/TDI and PPG/TDI. 2.2.4. Preparation of PU-urea-imide films The catalyst (0.05 wt.% DBTL and 0.05 wt.% triethylamine) was mixed with the synthesized PU-imide resins and the films were casted with different chain extenders on tin foil by using a power driven automatic applicator. The supported films were kept at 30 ◦C and at laboratory humidity condition for 20 d. The moisture cured PU-urea-imide films thus obtained were amalgamated and bottom of the unsupported films was carefully cleaned with muslin cloth and the free films were stored at laboratory temperature and humidity until the complete disappearance of

234 A.K.Mishra et al.Progress in Organic Coatings 55 (2006)231-243 -torr.A thin possible differences in the chemical composition between the funct appl to th stage,causing two energy transitions.It promotes transitions At e=09 the sample was nernendicular to the detector lead ina between rotational and vibrational energy ing to the maximum sampling depth.The effective sampling When transitions between rotational and tonal energy le de ved by 2=3 where is the ef molecule will absorb I Therefore an IR a set to th unique to a specific molecular vibration frequency.When IR ,=1.94nm [13].The accurate binding energies of ed by the 一Te identification of func onal groups is a major application of IR of the lens from a filament located directly atop the sample spectrome etry [42] pectra were edusing Sun Solarisbase data manin mode.Therefore.the elucidation of structure-property relations ulation,a linear background subtraction was performed prior to in P -urea-imide copolymers are composed or tv Lore to fit th tzian sum ed the hard can be realized from FT-IR spectroscopy by properly analyz algorithm.This is a time domain procedure,which fits a fourth rogen izable number of polyme ckgr an h ring imide copol ymer where multiple peaks overlap strongly. The and quantification of the functional groups present on the sam ple surface [43] The bind ding energy(BE)scale was calibrate eves of the pure metals plicated not a)s by m叫 eaks but also by other band distortionseffects,(b)aproper 3.Results and discussion etc. 3.1.FT-IR analysis Now let us confine e our racterist Fourier transform of the PU-ure imide thin films on KBr dise in the zone 400-4000 cm- viour in polymers,since m 3500cm ps.2795m mer or macromolecules possess bonds and functional groups (O-CH stretching)600-10cm(amide:C stretch. characteristics of their identity.These functional groups vibrate c)[46 of each oth ing vibrations),1540cm and weakly s we group ina macromolecule arises from the contribution ofuransla tional,rotational,vibrational,and electronic energies.Therefore scissoring and CH3 deformation.The band at 1780cm-is an inter with ra EM)spec 山he stret ing vibra the cai functional group involved in the Without an 7731780Cm8201380cm120amd720cm-ae EM radiation effect,these bonds or functional groups vibrate attributed to imide I,imide Il,imide Ill and imide IV,respec 33 cr conming that imide

234 A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 pressure in the analyzer chamber was kept at ∼10−8 torr. A thin film of thickness 0.15 mm was used for XPS analysis. To detect possible differences in the chemical composition between the surface and bulk, samples were measured at two different takeoff angles. The sample was tilted in such a way as to change the angle θ between the normal to the sample and the analyzer. At θ = 0◦, the sample was perpendicular to the detector, leading to the maximum sampling depth. The effective sampling depth, z, was derived by z = 3λcosθ, where λ is the effective mean free path for electrons to escape the surface and was set to the value of 2.5 nm. Therefore, at θ = 0◦, z = 7.5 nm and at θ = 75◦, z = 1.94 nm [13]. The accurate binding energies of the component peaks in each spectrum were determined after charge correction by referencing to the aromatic carbon bonded to another carbon at 284.6 eV. Charge neutralization was implemented by low energy electrons injected into the magnetic field of the lens from a filament located directly atop the sample. Spectra were deconvoluted using Sun Solaris based Vision 2 curve resolver. Data acquisition and processing were performed with the Kratos Vision 2 data system. Concerning data manipulation, a linear background subtraction was performed prior to any peak-fitting procedure. A Gaussian–Lorentzian sum function in 70:30 ratios was used to fit the individual peak. Before deconvolution, each peak was smoothed using Savitzky–Golay algorithm. This is a time domain procedure, which fits a fourth order polynomial in a moving window of a sizable number of data points. Shirley background was chosen during data fitting. The percentage of the overall area contributed by each individual peak can then be determined. This enabled the identification and quantification of the functional groups present on the sample surface [43]. The binding energy (BE) scale was calibrated by setting the Au 4f7/2 or Pt 4f7/2 core levels of the pure metals to values 84.0 or 71.2 eV, respectively. 3. Results and discussion 3.1. FT-IR analysis Fourier transform-infrared spectroscopy (FT-IR) is well established as an analytical technique for functional group analysis and to study the hydrogen bonding and phase separation behaviour in polymers, since mid-infrared spectral changes in band intensity and frequency shifts are known criteria for the presence and strength of hydrogen bonds [44]. In general polymer or macromolecules possess bonds and functional groups characteristics of their identity. These functional groups vibrate independently of each other and weakly interact. It is wellknown that the total energy of a particular bond or functional group in a macromolecule arises from the contribution of translational, rotational, vibrational, and electronic energies. Therefore, an interaction with radiation of the electromagnetic (EM) spectrum will result in different energy transitions of the bond or functional group involved in the macromolecule. Without any EM radiation effect, these bonds or functional groups vibrate independently at their equilibrium position. Sometimes the local vibration depends on the intermolecular interactions such as weak London forces or hydrogen bonding phenomenon if any. Therefore, the equilibrium vibration of a bond or functional group depends on the local geometry involved. However, when IR radiation (which is part of the EM spectrum) is applied to a functional group, it breaks down the equilibrium (position) stage, causing two energy transitions. It promotes transitions in a macromolecule between rotational and vibrational energy. When transitions between rotational and vibrational energy levels occur and cause a net change in the dipole moment, the molecule will absorb IR. Therefore, an IR absorption profile is unique to a specific molecular vibration frequency. When IR radiation passes through a sample, some of it is absorbed by the sample and some of it is passed through (transmitted). The resulting spectrum represents the molecular absorption/transmission, which creates a fingerprint of the analyzed sample. Therefore, identification of functional groups is a major application of IR spectrometry [45]. Thus, it should be realized that any change in a peak position or shape means a change has occurred in the distribution of frequencies included in that particular vibration mode. Therefore, the elucidation of structure–property relations in PU-urea-imide copolymers, which are composed of two segments, namely the soft segment derived from the macrodiol and the hard segment that consists of urethane/urea-imide linkages can be realized from FT-IR spectroscopy by properly analyzing the amount and type of hydrogen bonding present in a particular polymer. While this may seem to be a trivial fact, it can be important for interpretation of segmented PU-ureaimide copolymer where multiple peaks overlap strongly. The complicacy of projecting the structure–property correlation in a polymer with spectral deconvolution arises great scientific challenges due to: (a) spectral investigation of hydrogen bonding is complicated not only by multiple overlapping peaks but also by other band distortions effects; (b) a proper baseline correction is needed; (c) the time during the scanning should be high, so that the effect of noise can be minimized; (d) care should be taken not to force fit the bands, which may generate erroneous results, etc. Now let us confine our attention to the spectral changes appeared or the characteristic bands observed in the synthesized PU-urea-imide copolymers. Fig. 1 shows the infrared spectra of the synthesized PU-urea-imide thin films on KBr disc in the zone 400–4000 cm−1. The spectra are mainly characterized by bands at 3150–3500 cm−1 (NH stretching vibrations), 2800–3000 cm−1 (CH stretching vibrations: anti-symmetric and symmetric stretching modes of methylene groups), 2795 cm−1 (O CH2 stretching), 1600–1800 cm−1 (amide I: C O stretching vibrations), 1540 cm−1 (amide II, δN H + νC N + νC C) [46], 1376–1388 cm−1 (νC N), 1226–1292 cm−1 (amide III, νC N), 1110 cm−1 (C O C stretching vibration, ether group) and 766 cm−1 (amide IV). Band at 1455 cm−1 is attributed to CH2 scissoring and CH3 deformation. The band at 1780 cm−1 is assigned to the asymmetric stretching vibration of the carbonyl of the imide groups. Characteristic spectral absorbance at 1730–1780 cm−1, 1320–1380 cm−1, 1120 and 720 cm−1 are attributed to imide I, imide II, imide III and imide IV, respectively. The absorption bands at about 1780 and 720–733 cm−1 are characteristic bands of imide bonds, confirming that imide

A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 235 Fig. 1. FT-IR spectra of different PU-urea-imide copolymers in the range 400–4000 cm−1. group was introduced to synthesize polyester and polyetherurethane-urea-imide backbone. Imides II, III, and IV are due to the axial, transverse, and out of plane vibrations, respectively, of the cyclic imide structure [47,48]. O C O stretching in the hard urethane segment and esteric C–O stretching vibrations appears at 1067 cm−1. Absorbance in between 1002 and 1012 cm−1 is attributed to the stretching and rocking vibrations of the C C and CH2 groups, respectively. Amide V appeared at 695 cm−1. The band at 553.5 cm−1 is assigned to the δ(N C N) by analogy to the case of methylalkylureas[49]. C O C and C C O bending vibrations appeared at 453.5–455.3 cm−1. Amide I vibration consists of several components reflecting C O groups in different environments and is sensitive on the specificity and magnitude of hydrogen bonding. Amide I mode is a highly complex vibration and involves the contribution of the C O stretching, the C N stretching, and the C C N deformation vibrations. The complexity and multiplicity of inter- and intra-molecular environments surrounding the carbonyl groups in polymers makes the amide I range considerably broader in polymers than in monomers. The fact that peaks overlap indicates that many frequencies are common to each vibrational mode, as would be expected. Amide II mode is seen at 1545 cm−1 and is a mixed contribution of the N H in-plane bending, the C N stretching, and the C C stretching vibrations. It is sensitive to both chain conformation and intermolecular hydrogen bonding. Amide III mode involves the stretching vibration of the C N group. Amide III is highly mixed and complicated by coupling with NH deformation modes and is observed between 1226 and 1292 cm−1. Amide IV, V, and VI bands are produced by highly mixed modes containing a significant contribution from the NH out-of-plane deformation mode. They are expected to be in the 800–400 cm−1 region [49]. A very weak single band is observed at 868 cm−1, which might be originating either from the coupled vibration of the C O stretching or CH2 rocking modes. The strong infrared band assigned to the asymmetric stretching vibration of the C N group is expected at 1040 cm−1. This band overlaps with the very strong band at 1110 cm−1, the C O C stretching vibration of ether groups in PU-urea-imide copolymer synthesized from PPG [50]. The major differences in the FT-IR spectra of polyether (Fig. 2a) and polyester based copolymers (shown in Fig. 2b) are: (1) presence of a broad absorbance in between 1000 and 1200 cm−1 due to the stretching vibration of ether group in polyether based PU-urea-imide copolymer, which is absent in polyester based copolymer; and (2) presence of ester bond vibration, which has a broad absorbance in between 1200 and 1300 cm−1 for polyester based PU-urea-imide copolymer. PU-urea-imide copolymers are capable of forming several kinds of hydrogen bonds. In all cases, the hydrogen atom of the N H group in the urea and urethane linkages is the donor proton, while the acceptor group can be the carbonyl of the imide groups, urethane’s C O, urea’s C O or the oxygen atom of the ester or ether linkage when polyester or a polyether are present as the soft segment. Hydrogen bonding between the N H group (hard segment: HS) and urethane, urea or imide C O groups (HS) for both the polyester and polyether based systems is responsible for the phase separation behaviour, whereas hydrogen bonding interaction between the N H group (hard segment: HS) and ester (polyester based) or ether linkages (polyether based) favors a phase mixing characteristic to the PU-urea-imide copolymer. Scheme 2 shows the two hydrogen bonded structure responsible for phase separation and phase mixing characteristics in the PUurea-imide copolymer prepared from PPG: (a) type I hydrogen bonding between the hard segment (HS) and hard segment (HS), and (b) type II hydrogen bonding between hard (HS) and soft segment (SS). Our objective of deconvolution strategy includes peaks to form in whatever distribution pattern best fits the measured spec-

236 A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 Fig. 2. FT-IR spectrum of (a) PUI-16 and (b) PUI-3 in the range 400–2000 cm−1. tral band. We have not tried to force fit the bands since this may result erroneous results. The curve-fitting simulations were performed using Origin software version 6. The ν(N H) and ν(C O) band were deconvoluted considering peaks as Gaussian with a number of iteration to get the best fit Gaussian peaks. For deconvolution study, the second derivatives of spectra in νC O and νN H zone was used to find out the number of Gaussian peaks. The maximum error associated with the fit was estimated to be less than 5%. The main purpose of this study is to identify the relative degree of order of the hard segments and to show how the local order is influenced by the specified formulation variables. Fig. 3(a) and (b) shows the amide I region and N H zone of different PU-urea-imide copolymers, respectively. 3.1.1. The C O stretching region The hydrogen bonding interaction makes the carbonyl bond length elongated and results in a decrease in the bonding force constant as well as the bond order of C O bond, leading to the reduction of the stretching vibration frequency. Therefore, the hydrogen bonded ν(C O) band appears at a lower wavenumber region in comparison to that of the free ν(C O) band [51]. During ν(C O) spectral deconvolution, we were able to deconvolute polyester based PU-urea-imide copolymers into only two bands appearing at approximately 1737 cm−1 (free C O) and 1685–1700 cm−1 (hydrogen bonded C O), whereas PPG based PU-urea-imide copolymers show four deconvoluted bands (see Fig. 3(a) for a comparison between polyester and PPG based systems). The representative C O peak deconvolution of PUI-16 is shown in Fig. 3(c) and confirms the bands assignable to the urethane group are centered at approximately 1725–1740 cm−1 (free C O from urethane) and at 1710–1725 cm−1 (free C O from urea groups), whereas those assigned to the bonded urethane and urea groups are centered at approximately 1670–1685 cm−1 and 1644–1655 cm−1, respectively. Free and hydrogen bonded imide C O components could not be deconvoluted separately as they were overlapping with the urethane functions during the deconvolution curve-fitting process. Before deconvolution of the ν(C O) band, a flat baseline was chosen in between 1600 and 1800 cm−1 and the spectra was corrected by subtracting the baseline. The correlation coeffi- cients of the fitting process were more than 0.999. The analyzed Scheme 2. Two hydrogen bonded structure responsible for phase separation and phase mixing characteristics in the PU-urea-imide copolymer prepared from PPG: (a) type I hydrogen bonding between the hard segment (HS) and hard segment (HS), and (b) type II hydrogen bonding between hard (HS) and soft segment (SS)

238 A.K.Mishra et al.Progress in Organic Coatings 55 (2006)231-243 H HH CH H CH BD PD Scheme4.The dual-cis bond. urethane bond concentration is more due to the use of diol tary structural information to the C-O stretching vibration. chain extender [PD].Since,urea bonds are more polar in nature earchers have proved that the splitting of the bonded N-H band was related the ne b gen bonding strength among the investigated samples shows the region of the analyzed samples is shown in Fig.3(b)and a repre order: sentative deconvoluted spectrumof PUI-16 is shown in Fig.3(d). A flat baseline was cho and 0 T FPU-wea-imides:PUI-13PU-2PUl ere more than 0.997 For sed PU-urea-imides:PUI-6>PUI-2>PUI- The bands at 499,3372 and 3306cm-1 were assigned to free 5>PULI>PUL4-PUL3 ments:type -NH This result elucidated that adding a hard segment with high copolymers respectively.A small contribution in the deconv mide main cha luted spectra at approximately 3188 cm was also observed segment polar groups could bring the polymeric chains close 4]interpreted th Ias a two-photon ba together because of the stronger inter mol cular forces resulting to Romanova etal.551 an overtone of de formation vibration o s15 (Scheme 3)on the phase s and BD structure The butane-gauche interaction in PD structure creates a less e by fo dual-is hond as shown in when compared to the Scheme 4. ted in on to PUI-Is and PUI-.respectivelv on to itn the stru The other interesting obse rvation from th deconvolution of less in PUI-6 and high in PUI-3.This sup orts our earlier dis verify whet area of山 free urea car cussion on carbonyl zone and shows th at a polar chain extender for the synthesized PU-urea-imide .4'-diamino-diphenyl sulfone (DDS) PU-urea-imide olymer with more hydr larger than the non-hydrogen-bonded carbonyl [44,46],we acteristics.Similar phenomenon were also observed for PPG Now letus se when based PU-urea-imic the third ste copolymer nich show: 11 in Scheme 1 and during the moisture curing stage when the ini he peak tial low molecular weight mixture is transformed into a polyme in PUI-16 suggests that the amount of phase mixing was higher compared to the amount of phas e separation (see the peak area at and 3306 cm compos reaction 3(d)).Similarly,for other higher compared to the amount of phase 3.1.2.The N-H stretching region analysis of type I and type lI hydrogen bonding by N-H peak s the ord to conclude that the

238 A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 Scheme 3. Structural arrangement of 1,4-butanediol (BD) and 1,2-propane diol (PD) in the PU-urea-imide copolymer. urethane bond concentration is more due to the use of diol chain extender [PD]. Since, urea bonds are more polar in nature when compared to urethane bonds, interchain association due to hydrogen bonds is higher in PUI-13. A comparison of hydrogen bonding strength among the investigated samples shows the order: For PPG based PU-urea-imides: PUI-13 > PUI-12 > PUI- 14 > PUI-15 > PUI-16 For polyester based PU-urea-imides: PUI-6 > PUI-2 > PUI- 5 > PUI-1 > PUI-4 > PUI-3 This result elucidated that adding a hard segment with high polarity to the PU-urea-imide main chains caused the interaction of inter-segments to become stronger. Therefore, these hard segment polar groups could bring the polymeric chains close together because of the stronger inter molecular forces resulting from the formation of hydrogen bonds [15]. Now, let us have a look on the effect of PD and BD structure (Scheme 3) on the phase separation behaviour. The butane-gauche interaction in PD structure creates a less favorable hydrogen bonding interaction when compared to the BD chain extended PU-urea-imide copolymer. Therefore, PUI- 16 and PUI-3 are more phase separated in comparison to PUI-15 and PUI-4, respectively. The other interesting observation from the deconvolution of C O region was to verify whether the area of the free urea carbonyl groups is higher than that of the bonded urea carbonyl group for all the synthesized PU-urea-imide copolymers. Since the absorptivity coefficient of the hydrogen-bonded carbonyl is larger than the non-hydrogen-bonded carbonyl [44,46], we can surely state that most urea carbonyls were not hydrogen bonded. Now let us see when urea bonds are formed. It is in the third step in Scheme 1 and during the moisture curing stage, when the initial low molecular weight mixture is transformed into a polymer network. Therefore, a high free urea carbonyl content suggests that the phase mixing takes place during the solid-state post-cure reaction. 3.1.2. The N H stretching region The N H stretching vibration is sensitive to the specificity and magnitude of hydrogen bonding and yields complemenScheme 4. The dual-cis bond. tary structural information to the C O stretching vibration. Researchers have proved that the splitting of the bonded N H band was related to different acceptors with which the N H groups were hydrogen bonded [52–54]. The N H stretching region of the analyzed samples is shown in Fig. 3(b) and a representative deconvoluted spectrum of PUI-16 is shown in Fig. 3(d). A flat baseline was chosen in between 3000 and 3800 cm−1 and the spectra were corrected by subtracting the baseline. The correlation coefficients of the fitting process were more than 0.997. The bands at 3499, 3372 and 3306 cm−1 were assigned to free N H groups, N H groups bonded with carbonyl (in hard segments: type I, NH···O C) and ether oxygen (in polyether soft segments: type II, NH···COC) in the PU system-urea-imide copolymers, respectively. A small contribution in the deconvoluted spectra at approximately 3188 cm−1 was also observed. Queiroz et al. [44] interpreted this band as a two-photon band in their study with segment urethane/urea membranes. According to Romanova et al. [55], an overtone of deformation vibration of N H group increased by Fermi resonance appears at 3162 cm−1, whereas Bellamy et al. [56] and Liu and Ma [48] assigned the bands at 3190 cm−1 to the N H groups bonded with the carbonyl of urethane by forming a dual-cis bond as shown in Scheme 4. Now, let us start the analysis of type I and type II hydrogen bonding interaction in relation to the free N H with the structural variation. Fig. 3(b) shows that the free N H content was less in PUI-6 and high in PUI-3. This supports our earlier discussion on carbonyl zone and shows that a polar chain extender, e.g., 4,4 -diamino-diphenyl sulfone (DDS) selectively enhances hard segment cohesion due to its polarity and therefore produces PU-urea-imide copolymer with more hydrogen bonding characteristics. Similar phenomenon were also observed for PPG based PU-urea-imide copolymers, i.e., PUI-16, which shows better hydrogen bonding interactions than PUI-13. The peak contribution of type 1 and type II hydrogen bonding structure in PUI-16 suggests that the amount of phase mixing was higher compared to the amount of phase separation (see the peak area at 3372 and 3306 cm−1 in Fig. 3(d)). Similarly, for other compositions also we have verified that the amount of phase mixing was higher compared to the amount of phase separation. A detailed analysis of type I and type II hydrogen bonding by N H peak deconvolution allows us to conclude that the phase separation follows the order:

A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 239 For PPG based PU-urea-imides: PUI-13 > PUI-12 > PUI- 14 > PUI-15 > PUI-16 For polyester based PU-urea-imides: PUI-6 > PUI-2 > PUI- 5 > PUI-1 > PUI-4 > PUI-3 Therefore, from the FT-IR spectroscopy results it can be seen that PU-urea-imide copolymer with different hard and soft segment structures have different interactions of the molecular chains, and this force is stronger when a high polarity chain extender was used. It is clear from the curve-fitting results that the hydrogen bond contribution in N H stretching vibrations is higher when compared to free N H stretching vibrations. In contrary, carbonyl stretching results showed that most of the carbonyl vibrations do not participate in hydrogen bonding. The reason behind such a type of contradictory results is due to the different absorptivity coefficient ratios of the two spectral regions (C O and N H) [57,58]. 3.2. XPS-analysis Surface properties play an important role in numerous important technologies. X-ray photoelectron spectroscopy (XPS) is a widely used method for surface analysis of coating materials. Due to its surface sensitivity and chemical specificity, XPS is an ideal technique for analyzing the surface segregation nature in PU based coatings. From the XPS spectra, the elements present and their chemical state (valence) can be determined. Since only electrons emitted from atoms near the surface escape without losing energy, this is making this technique very surface sensitive. Subtle changes in peak positions and shape can yield important information on changes in surface chemistry, giving XPS the ability to determine the elemental composition on the surface of all non-volatile materials. Analysis of the XPS data gives an idea about the local oxidation states and chemical bonding environment of the composing material [13,44,59]. Fig. 4. The deconvoluted C 1s AR-XPS spectra of different PU-urea-imide copolymers at takeoff angle 0 and 75◦. (a) PUI-3; (b) PUI-6; (c) PUI-1; (d) PUI-16; (e) PUI-12; and (f) PUI-13

240 A.K. Mishra et al. / Progress in Organic Coatings 55 (2006) 231–243 Fig. 4. (Continued ). The presented X-ray photoelectron spectroscopy (XPS) investigations were performed in order to extract quantitative results about the surface composition from different samples. The chemical properties of PU-urea-imide copolymers and their correlation to surface properties have been much less systematically investigated and the present paper shows a brief correlation. It is well-known that the chemical nature of the PU surface can be very different from the bulk composition of the PU [13,15]. The relative amounts of differently bound carbons were determined from high-resolution C 1s spectra, using a mixture of 70:30 Gaussian and Lorentzian functions. Since the hard segments of PUs are usually in their glassy state, the soft segments have higher mobility and can easily migrate to the surface region. A pass energy of 80 eV was used for survey spectra, which indicates peaks at 535, 400, 285, and 102 eV due to oxygen (1s), nitrogen (1s), carbon (1s), and silicon (1s), respectively. Si is suspected to be derived from dimethylsiloxane based contaminants as reported earlier [15]. Si contamination from the survey spectra recorded during XPS analysis was in between 2 and 2.8% for all the samples investigated. As nitrogen is associated with the urethane and urea linkages, the XPS elemental composition data (table not shown) suggests a significant increase in nitrogen concentration from the surface to bulk, which signifies that the urethane and urea content at the upper surface is much less than in the bulk [60]. The binding energy range in the high-resolution C 1s XPS spectra were about 284–290 eV. The high-resolution C 1s spectra as observed from the XPS analysis were resolved into four characteristic peaks. Briggs and Seah [61]summarized the binding energy values for the C 1s of polymer samples and coupled with theoretical calculations. According to Briggs and Seah [61] for any type of hybridization, C C/C H the C 1s binding energy value is equal to 285.0 eV; at the same time oxygen induces the shifts in the binding energy values and for C O bonding the C 1s equals to 286.5. We attributed the observed deconvoluted peaks at 284.5–284.8, 285.6–285.9, 286.5–287.0 and 287.5–289.0 eV are to the functional groups of C C/C H, C O, C N and C O, respectively, on the surface of the PUurea-imide films. Deconvoluted high-resolution C 1s spectra