- Email: [email protected]
Solar heat reflective coating formed of polystyrene chains bearing 4-vinylpyridine-rich end segments
Polymer 87 (2016) 170e180
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Solar heat reflective coating formed of polystyrene chains bearing 4-vinylpyridine-rich end segments Zheng Xing a, Siok-Wei Tay b, Yeap Hung Ng a, Liang Hong a, b, * a b
Department of Chemical & Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore, 119260, Singapore Institute of Materials Research and Engineering, Agency for Science, Technology and Research, A*STAR, 3 Research Link, Singapore, 117602, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 October 2015 Received in revised form 26 January 2016 Accepted 30 January 2016 Available online 13 February 2016
Incorporation of a minor dose of 4-vinylpyridine (4-VP) into an end segment of polystyrene (PS) chain by emulsion copolymerization yields a copolymer molecule P(4-VP/S) with weak surface-activity. The spin cast thin films of P(4-VP/S) display enhanced solar light reflectance of R ¼ 80e90% compared with its PS homopolymer counterpart. It is imperative to use dimethylformamide (DMF) as casting solvent because the hydrophilic nature of DMF induces core-to-grain assembling while dried in humid air, leading to a nodular film matrix. A rise of 4-VP dose in P(4-VP/S) causes a breath-figure patterned matrix instead and consequently an obvious reduction of reflectance in the near-infrared region. Mechanistically, P(4-VP/S) chains that bear a short end-segment with rich 4-VP units constitute tiny hydrophilic cores because of entrapment of moisture during drying, surrounding which the long PS tails thus undergo chain coiling to form nanoparticles. They then aggregate to form submicron grains in rosette and the partial coalescence of adjacent grains results in micron lumps. Such hierarchical evolution finally constructs a matrix exhibiting an enhanced solar reflectance. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Solar reflective coating 4-pyridine Polystyrene Chain assembling w/o emulsion
1. Introduction Solar light consists of ultra violet (l < 400 nm), visible (400e700 nm) and near infrared (ca. 700e2500 nm) radiations. Although some particular wavelengths of the infrared radiation are absorbed by O2, H2O, and CO2 in the atmosphere, UVevisible spectrum (300e700 nm) and near infrared (NIR) spectrum (700e2500 nm) constitute 48% and 52% of solar heat reaching the earth, respectively [1]. Energy consumption for air-conditioning with urban expansion shows an increasing trend worldwide and is a major concern in those hot climate countries. To lessen the intense demand for cooling building and car interior space, solar heat reflective (SHR) coatings play an important role [2]. Currently, the SHR coatings on market [3] are mainly hybrid polymer composite thin layers containing inorganic particles with high refractive indices, e.g., rutile TiO2, ZnO and their modified forms, which are uniformly embedded in various polymer matrices, typically acrylic, vinylacetate, and styrenic types [4,5].
* Corresponding author. Department of Chemical & Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore, 119260, Singapore. E-mail address: [email protected] (L. Hong). http://dx.doi.org/10.1016/j.polymer.2016.01.079 0032-3861/© 2016 Elsevier Ltd. All rights reserved.
A SHR composite coating generally obeys Mie scattering mechanism, by which the reflection happens at the interface between inorganic filler particles and polymer matrix due to different refractive indices [3]. The particle size and identity of the filler vitally affected the light reflection from this type of coating, for instance, P. Berdahl's work [6] shows that 0.22 mm is the optimal diameter of particles to scatter visible light and larger particle sizes are required to reflect NIR light. As to the nano fillers (<100 nm) used, being much smaller compared to the UVeVis wavelengths (l), they render prevailed Rayleigh scattering in all directions. Therefore, the scattering is effective only to short wavelength lights due to its l4 dependence. Compared to the composite SHR coating matrix, the pristine polymer coating matrix normally lacks SHR capability because commonly available polymer resins possess relatively low refractive indices in the range of 1 to 2. In consequence, there have been scarce studies thus far to explore solar reflectivity of pristine polymer coatings. Inspired by the high UVeVis albedo of tiny ice crystallites [7], we have recently found that PS nanoparticles, resulted from discrete chain folding mode, can effectively reflect solar light [8]. About this finding, when a cast DMF liquid film of PS is subjected to drying under humid ambient condition, discrete PS nanoparticles grow and coalesce subsequently to form submicron grains. The subsequent aggregation of
Z. Xing et al. / Polymer 87 (2016) 170e180
these grains results in micron-sized lumps over the surface of the cast film. This hierarchical evolution gives rise to a solar reflectance profile above 0.7, which is uncommon since otherwise a reflectance below 0.1 is resulted if a hydrophobic solvent such as chloroform or toluene is used instead of DMF to develop casting. Meanwhile, the interfacial potential driven dewetting phenomenon, which has been thorough studied [9], is unlikely to happen in the above coating films because the coating films fall in the dimension that is too thick for dewetting to happen and the solvent used for spin coating is miscible with moisture condensed from air during drying in ambient conditions. Such discrepancy in reflectivity was further validated using the infrared lamp irradiating in front of a PS coating on a glass panel. The temperature on the backside of the panel was recorded with time. The PS coating based on DMF shows an apparently lower temperature-time curve than its counterpart based on toluene (Fig. 1). As described above, gradual absorption of ambient moisture by a DMF-PS liquid film while being dried is critical to attain a solar reflective PS coating because the diffusion of water into DMF prompts chain coiling of PS. In the present work, we incorporate a very low mole fraction of hydrophilic 4-VP monomer units into PS chains through copolymerization in an oil-in-water (o/w) emulsion and investigate the impact of the resulting 4-VP-doped PS chain structure on the solar reflectivity by applying the same solution casting technique. It is hypothesized that a number of 4-VP-doped PS end-segments may associate together to generate a hydrophilic core, which is then surrounded by PS nanoparticles generated from chain coiling. This manner is expected to enhance heterogeneity within submicron scale and boost solar reflectivity. The resulting cast film indeed displays an enhanced solar reflectance relative to the pristine PS counterpart. Regarding the emulsion copolymerization of 4-VP and styrene, both the emulsifier-based system [10] and the soap-free system [11e13] have showed the formation of pyridine-rich radical species in the aqueous phase at the initial stage of copolymerization, which eventually results in copolymers possessing surface activity due to the formation of pseudo blockcopolymer chain structure. In the present preparation, since only a minor dose of 4-VP vs. styrene is in the monomer feed, sodium dodecyl sulfate (SDS) has to be used to stabilize the emulsion throughout the entire polymerization process. It is rational that 4VP units primarily appear in a very short end-segment of a PS dominant chain to function as dopant. The pseudo block-copolymer chain structure is therefore allowed to form hydrophilic nano-cores in the cast DMF-PS liquid
171
film accompanying vaporization of DMF in ambient air due to dissolution of moisture in the liquid coating. Each hydrophilic nano-core would then facilitate the coiling of the long PS tails surrounding the core and then nucleation of the coils to form grains in rosettes as illustrated in Scheme 1. The further agglomeration of these grains constitutes lumps in micron sizes, which play the role of scattering the long-wavelength component of NIR [14]. The resulting cast film clearly demonstrates enhanced solar reflectivity in contrast to the homo-PS film. Nevertheless, a slight increase in 4VP doping extent in the chain-end segment of PS causes a breathfigure patterned matrix [15], which exhibits obviously weaker light reflectivity in the NIR range. On the other hand, relative to the o/w emulsion copolymerization, although the synthesis of P(4-VP/ S) by Reversible Addition-Fragmentation Chain Transfer (RAFT) polymerization method allows for attaching a pure 4-VP segment to a PS block [16], we found that the RAFT block-copolymer shows an apparently weaker solar reflectance profile than the emulsion sample made by the same monomer-feed (viz. 1 mol % 4-VP). This happens because the RAFT polymerization produces far shorter PS blocks than the emulsion polymerization. Molecular weight has been well known a vital factor affecting chain packing and folding and hence the thin film bulk and surface structures. Moreover, the o/w emulsion copolymerization system allows substitution of 4-VP by another hydrophilic monomer. The present study has also examined the other three water-soluble monomers that indeed cause the final cast films display rather different solar reflectance profiles. In short, attaching a considerably short chain-end segment rich in 4-VP units to a long PS chain permits a hierarchical evolution from hydrophilic nuclei to nanoparticles of PS and then to micron lumps comprising the nanoparticles in the cast film accompanying evaporation of DMF. The final film matrix is unique because it sustains the maximum solar reflectance and a variation from this structural characteristic causes a noticeable reduction in solar reflectance. 2. Experimental section 2.1. Materials Styrene (Reagent Plus®, 99%, SigmaeAldrich), 4-vinylpyridine (4-VP, 96%, Alfa Aesar), N-vinyl pyrrolidone (~97%, Fluka) and 1vinyl imidazole (99þ%, Aldrich) were passed respectively through an inhibitor-removal column before use. Polystyrene (PS, average Mw ~192,000, SigmaeAldrich), poly(4-vinylpyridine) (P4-VP, Mw ~50,000, Polysciences), 4-styrenesulfonic acid, (sodium salt hydrate, Aldrich), B-79 (polyvinyl butyral, Butvar®, Eastman), N,Ndimethylformamide (DMF, min. 99.5% purity, Merck), toluene (99.99% purity, Fischer), methyl ethyl ketone (MEK, min. 99%, Fischer), sodium dodecyl sulfate (SDS, 99%, Fluka), ammonium persulfate (APS, min. 98%, Merck), inhibitor remover (for removing tert-butylcatechol, Aldrich) were used as received. 2.2. Doping PS chain-end by 4-VP
Fig. 1. The IR lamp irradiation timeetemperature profiles of the PS cast films developed by using toluene and DMF respectively as solvent.
In a typical preparation, a mixture of styrene (2.1 ml, 22.2 mmol) and 4-vinylpyridine (0.02 ml, 0.2 mmol) was introduced in a 20 ml aqueous solution containing 0.5% SDS and 0.2% (or 8.7 mM) ammonia persulfate, which was followed by magnetic stirring in a 50 ml round bottom flask accompanying by Ar bubbling for 20 min to remove oxygen. Then the flask was immersed into an oil bath at 70 C with stirring at 400 rpm to carry out emulsion polymerization. The polymerization was lasted for 24 h in Ar and then cooled down to room temperature. The polymer latex formed was filtrated and washed first by deionized water and then ethanol for a few times to remove possible P4-VP oligomer formed primarily in the
172
Z. Xing et al. / Polymer 87 (2016) 170e180
Scheme 1. A schematic illustration of the evolution of cast film matrix with roughness in microns from the hydrophilic nucleation of pseudo block copolymer chains to coalescences of granules.
aqueous phase, the surfactant and residual monomers. Subsequently, the product was dried in oven at 50 C and then collected in a sample vial. In the same preparation, the mole fraction of 4-VP in the feed of polymerization was varied from 0 to 5%. The resultant polymers are labeled by P(4-VP/S)_x, where S is the short form for styrene and x stands for the dose of 4-VP used, for example P(4-VP/ S)-0.5% is the sample made by the feed containing 0.5 mol% 4-VP and styrene. In addition, an alternative hydrophilic monomer, 4styrenesulfonic acid (4-SS), N-vinyl pyrrolidone (N-VP) and 1vinyl imidazole (1-VI), was used respectively in place of 4-VP (at 1% loading extent) to synthesize copolymer chain. In addition, a random copolymer sample was prepared as a control sample in a homogeneous solution comprising toluene (4 ml), AIBN (31.8 mg), styrene (2 g) and 4-VP (20 mg). Similarly, the polymerization was carried out at 70 C in Ar for 20 h, after that, the product was precipitated in a large excess of methanol, followed by vacuum drying at 40 C for 24 h. This control is labeled by P(4-VP/ S)-1%h. Another control was the polymer blend consisting of P(4VP/S)-0 and P4-VP (Polysciences) (1 mol% based on monomer units). The sample was prepared by mixing in DMF. 2.3. Preparation of coating film by solution casting approach A primer coat that consists of the subsequent B-79 and PS double layers were laid on a glass slide (Sail Brand 100 100 ), where B-79 functions as an adhesive and PS a transition layer. As the first layer, a B-79 solution (10 wt. %) in the binary MEK/toluene (v/v ¼ 1) solvent was spin cast on the glass slide using 600 rpm for 10 s, which was then dried at room temperature for 5 min. After that, a thin layer of PS (SigmaeAldrich) was spin-cast on the B-79 layer and dried by the same procedure to complete the bedding of primer. To develop a topcoat on the primer, typically, a polymer sample (0.7 g) and 5 ml DMF were mixed and stirred in a sample vial at room temperature overnight to obtain a homogeneous solution. It was spin-coated at 600 rpm on the primer for 10 s. The liquid film thus prepared was dried under ambient condition
characterized by 65e70% relative humidity (RH) and 25 C until no change in mass with time. Additionally, PS coatings on glass panel (10 cm 10 cm) were prepared to test the shielding capability against IR irradiation. The two solutions of PS (12 wt %) in toluene and DMF were coated, respectively, by brushing on the glass panel and then subjected to drying in ambient condition. 2.4. Measurement of the total solar reflectance and the shielding effect of infrared irradiation The reflectance profile of a coating film fabricated from the above protocol was gauged using a Solar Spectrum Reflectometer (SSR-6, Devices and Services Company). The reflectometer is able to measure reflectance with a standard deviation of 0.005 units and a bias of ±0.002 units. The reflectometer was calibrated using a blackbody cavity, with zero solar reflectance, and by a smooth and dense white tile with the designated reflectance reading of 0.859, at the solar irradiance b891 (ASTM E891-87 air mass 1.5 beam normal), namely, the AM 1.5 spectra that provides a reference point corresponding to a particular set of atmospheric conditions and a specific air mass [17]. The reflectance profile is plotted versus the UV, blue, red and IR light sources. The parenthesis beside each light source labels the color temperature in K of the source. According to Planck's law, the color temperatures in the range from 2000 to 3000 K signify relatively weak radiant emittances (between 0 and 0.5 1013 W/m3) in the wavelength range from 0.1 to 500 mm. Taking Blue (3125 K) for example, it describes blue light with the maximum emittance of approximately (5 1011 W/m3) at 0.47 mm over a broad wavelength range. The shielding against IR irradiation was carried out in an inhouse setup. An IR irradiance lamp (Philips HP 3616, 50 W) was placed 30 cm away in front of the coated glass panel that is fixed in a thermal insulation box made in-house. A thermocouple (Easyview 15) is attached to the backside of the coating panel. The temperature shown on the Extech Instrument was recorded every
Z. Xing et al. / Polymer 87 (2016) 170e180
minute in the first 10 min period and every 10 min in the last 20 min period. 2.5. Characterizations of the P(4-VP/S)-x% copolymers Detailed morphology of a topcoat was scrutinized by fieldemission scanning electron microscopy (FESEM, JEOL, JSM6700F). The surface contour of a coating and its roughness were recorded by atomic force microscopy (Bruker Dimension ICON). The samples were scanned by tapping mode in a designated size, e.g. 1 mm 1 mm, and in a scanning frequency of 0.6e1.0 Hz. The obtained AFM data were processed into 3-D pictures using the software (Nano Scope Analysis 1.40). Various polymer packing states in the cast films were analyzed by differential scanning calorimetry (DSC-Q100 by TA Instruments) in the range from 25 to 250 C at a scanning rate of 5 C/min, for which each sample was accurate to 3.1 ± 0.01 mg. The distribution of 4-VP units in a P(4-VP/S)-x% chain was also studied by using Fourier transform infrared spectroscopy (Bio-Rad Excalibur FTS-3500 FTIR spectrometer), for which the samples are polymers directly obtained from polymerization. The 1 H-NMR spectra of different P(4-VP/S)- x% samples were obtained from a Bruker Ultra Shield spectrometer (400 MHz), using chloroform-d as solvent. Both the area of peak a from the protons of pyridine ring and the area of doublet b from the protons of benzene ring were used to determine the real mole fractions of 4-VP in the different samples. The dielectric constants of different P(4-VP/S)-x% samples were measured on a dielectric thermal analyzer (DETA, DS6000) using a frequency of 10 kHz at room temperature. The testing pellet (ca. 13 mm 0.5 mm) was prepared by pressing a powder (0.1 g), obtained from a purified emulsion polymerization product, in a cylindrical die-set to 7.5 metric tons that was held for 5 min. Transmission electron microscopy was employed to examine the self-assembling behavior of the P(4-VP/S)- x% samples: a DMF solution (ca. 1%) was prepared and 1 drop of the solution was transferred to a copper grid, which was allowed to dry at the ambient condition.
173
initiator molar ratio (100:1) compared with the usual range. The real 4-VP content (Table 1) in each polymer sample was determined by 1H-NMR (Supporting information (SI): Fig. S1), which is lower than the designated dose in the respective monomer feed likely due to formation of P4-VP oligomer in the water phase. For the simplicity reason, the following discussion will still use the feed dose of 4-VP (x%) to denote a sample, i.e., P(4-VP/S)-x%. Regarding the chemical environment of 4-VP unit in P(4-VP/S)-x% with the variation of x value, the FT-IR spectra (Fig. 2) permits qualitatively correlating the x value with the relative peak intensity: the characteristic adsorption of the asymmetric stretching vibration of C] N of 4-VP unit at 1557 cm1 [18,19] is identified besides its symmetric ring stretching absorbance (at 1600 cm1) that overlaps with the same absorbance of styrene unit. In addition, the absorbance at 1540 cm1 is known as a combination band of the two outof-plane vibration modes of aromatic rings [20]. It is clear that these two IR absorption bands (1557 and 1540 cm1) become more comparable in peak intensity with the increase in x value. In
3. Results and discussion 3.1. Characterization of the copolymer chain structures of P(4-VP/ S)-x% The P(4-VP/S) samples, although structurally being a copolymer chain synthesized by the o/w emulsion copolymerization, are described as 4-VP-doped PS in this work because the solar reflectivity is susceptible to maintaining a substantially low content of 4VP. According to a previous study [10], both the micelle and homogeneous nucleation steps govern the development of latex particles in the o/w emulsion polymerization system. The homogeneous nucleation mechanism favors the formation of a pseudo block-copolymer copolymer, the trend of which will increase with the use of a lower monomer/initiator ratio as well as a lower dose of hydrophilic monomer, e.g. 4-VP relative to styrene in the present system. With the aim to attain a pseudo block-copolymer chain structure, besides using very low 4-VP/S molar ratios, the formulation of monomer feed also used a relatively low monomers/
Fig. 2. The FT-IR spectra of the P(4-VP/S)-x% samples in the expanded wavelength number range from 1500 to 1580 cm1.
Table 1 The real 4-VP contents in the copolymer synthesized according to 1H-NMR and their properties. Samples
P(4VP/S)-0
P(4VP/S)-0.5%
P(4VP/S)-1%
P(4VP/S)-2.5%
P(4VP/S)-5%
P(4VP/S)-1%_h
4-VP (mol%) Mn ( 104) Mw/Mn
0 50.1 1.76
0.13 61.1 1.56
0.31 82.4 1.47
1.20 63.0 1.59
2.67 74.6 1.43
0.30 1.47 1.67
174
Z. Xing et al. / Polymer 87 (2016) 170e180
principle, the asymmetric ring stretching amplitude of the pyridine unit should be stronger with a more even distribution of it in styrene units due to a reduction of dipoleedipole interactions between 4-VP units. Moreover, as a control sample, P(4-VP/S)-1%h obtained from the solution polymerization exhibits almost the same peak intensities between these two IR bands. In this control, 4-VP units have a uniform distribution along the P(4-VP/S)-1%h chains because both monomers have the close r values (<1) in the solution copolymerization [21]. In contrast, the P(4-VP/S)-1%, despite containing almost the same 4-VP content as this control sample (Table 1), shows obviously different intensities between the two bands as the result of forming the pseudo-block copolymer structure (also refer to SI: Fig. S2). It can therefore be concluded that P(4-VP/S) is more resemble to a block copolymer with the decrease in 4-VP content. In addition to the above IR spectroscopy characterization, a consistent conclusion can also be obtained from the relative permittivity (εr) measurement of the polymer latex particles (Fig. 3). The particles must have an inward decreasing concentration gradient of 4-VP unit according to the o/w emulsion polymerization mechanism and the hydrophilic nature of 4-VP. Formation of the pseudo block-copolymer chain structure would leave behind a very thin polar shell comprising rich 4-VP units. Hence a larger εr is expected because of a distinct distribution of polar 4-VP units vs. S unit. The P(4-VP/S)-0 displays a slightly higher εr (~3.1) than the common PS (~2.7) due to a collection of terminal sulfate groups, transferred from the free-radical initiator, on the surface of latex particles. The test result shows that εr increases from P(4-VP/S)-0 to P(4-VP/S)-0.5% and then to P(4-VP/S)-1%, which is attributed to the accumulation of 4-VP units in the surface shell of latex particles because of the extension of the pseudo-4-VP block. After that, a decreasing trend of εr is observed instead with the introduction of more 4-VP into the polymerization system, from P(4-VP/S)-2.5% to P(4-VP/S)-5%. It implies that an increase of the inward distribution of 4-VP units in each latex particle owing to the increase in the micelle nucleation mechanism [10], which weakens the electrical polarization extent in latex particles. The same conclusion as the above IR characterization can therefore be drawn: a low 4-VP dose in the emulsion polymerization favors concentrating 4-VP units in short chain ends and therefore leads to a thin polar shell on individual latex particles. This view is further proven
Fig. 3. Variation of the relative permittivity value (at 10 kHz and 25 C) of the o/w emulsion latex particles with different 4-VP doses in the P(4-VP/S)-x% copolymers.
by the control sample, a powder blend of 1% homo-P4-VP and P(4VP/S)-0, which exhibits basically the same permittivity as the P(4VP/S)-1% and the pure P4-VP (~4.4) [22], although the actual 4-VP content in P(4-VP/S)-1% is well below 1% according to the above 1 H-NMR quantification. Both the IR spectroscopy study and relative permittivity measurement latex have indicated that distribution of 4-VP units in the PS chain leaps since the 4-VP content in the feed of polymerization reaches 2.5%. Furthermore, proposing that a variation of 4-VP distribution in PS chains is to affect the self-assembling behavior of the polymer in a proper liquid medium, we inspected this property by designing a test (section 2.5). A drop of a very dilute polymer solution in DMF was transferred to a TEM sample-holder and allowed for drying at room ambient conditions. DMF absorbs moisture from air while being evaporated, which triggers collection of the hydrophilic 4-VP-rich end segments of a certain number of polymer chains to form a hydrophilic core and a PS shell providing that most of 4-VP units concentrate in the end-segment of a chain. Besides this, the initial dilute condition permits formation of discrete particles after drying. The proposal is indeed supported by TEM images (Fig. 4) of the four P(4-VP/S)-x% (x ¼ 0 to 2.5) samples. They display the occurrence of submicron microspheres (for x ¼ 0.5 and 1) having a loose interior surrounded by a dense shell, in which the loose interior was left behind by drying. Contrary to these two samples, both (x ¼ 0 and 2.5) samples do not present such a hollow particle structure because of a lack of 4-VP and a pseudo-block copolymer chain structure, respectively. This outcome is consistent with the conclusion drawn from above IR and relative permittivity studies. This examination verifies the hydrophilic nucleation as illustrated in Scheme 1, which is structurally selective only for the lightly doped PS chains (x ¼ 0.5 and 1.0) and leads to submicron grains. It will be further noted that chain-packing steps surrounding the hydrophilic nuclei eventually enhances the NIR light reflectance on the surface of the cast film. 3.2. Impact of 4-VP content in P(4-VP/S) on the solar-reflectance The cast films made of individual P(4-VP/S)-x% samples by the protocol as described in section 2.3 display different solar reflectance profiles (Fig. 5). They could be roughly divided into two groups: x ¼ 0 to 1 versus x ¼ 2 to 5, because the former three coatings reflect at least 78% solar light from different light sources, while the latter two coatings show particularly weaker reflectivity in the IR wavelength range (from 0.2 to 400 mm with the maximum intensity at 1 mm). To understand this outcome, their surface morphologies in micron scale were inspected (Fig. 6a). As the control sample, the P(4-VP/S)-0 cast film, shows a dense matrix and a sparsely spread of micron-sized pothole-like pores. Similar but shallow holes also appear on the P(4-VP/S)-0.5% cast film, which has however a relatively rougher matrix. These potholes are formed due to condense of water droplets from ambient moisture likely in the final stage of drying. One of Müller-Buschbaum's previous studies [23] on the surface roughness of the as-prepared PS thin film on Si(100) has found that it is affected by molecular weight, film thickness and annealing temperature and duration. In the present work, the films, besides being far thicker than those studied in the above work, are developed on a double-layer primer. The primer layer plays a crucial role in alleviating the impact of substrate (glass plate), e.g. the surface-bending rigidity, on the surface microstructure of coating films and improving the interface compatibility. On the basis of the above observation, a further zooming in scrutiny scale (Fig. 6b) unveils the detailed microstructures of the above two film matrices (x ¼ 0 and 0.5). Although both matrices consist of elementary particles of 10 nm or smaller, these nanoparticles undergo extensive coalescence in the P(4-VP/
Z. Xing et al. / Polymer 87 (2016) 170e180
175
Fig. 4. TEM images of the self-assembled P(4-VP/S)-x% colloidal particles generated from drying a considerably diluted polymer solution in DMF in humid air. Note: the scale bar in white works for the 4 images.
Fig. 5. Solar reflectance profiles of the P(4-VP/S)-x% cast films.
S)-0.5% film to form submicron (50 nme0.5 mm) granules, whose random collection constructs a matrix with roughness in micron scale at maximum. Whereas, the P(4-VP/S)-0 film exhibits a much smoother and denser matrix assembled by nanoparticles, where numerous slightly convex submicron bumps are observed as well. Similarly, the P(4-VP/S)-1% film still retains the three-tier microstructures: nano-particle / submicron-grain / rough-matrix, formed via the mechanism proposed in Scheme 1. On contrary,
the P(4-VP/S)-2.5% film exhibits a microstructure comprising significantly coalescing nanoparticles, and hence only the last-two stage characteristics could be recognized. Besides this, the film has a honeycomb-like matrix at the micron scale (Fig. 6a). Correspondingly, the P(4-VP/S)-5% film shows a very similar matrix to P(4-VP/S)-2.5% but a greater extent of nanoparticle mingling (SI: Fig. S3). It is rational that the honeycomb-like matrix is generated through forming a hydrophilic continuous phase, which is opposite to the nucleation of hydrophilic core when the 4-VP content in the copolymer is substantially low. This inversion of 4-VP enriched blocks towards the DMF-water medium brings about amalgamation of PS dominant long blocks to form submicron grains. The agglomerating of these grains constitutes the framework of honeycomb matrix, e.g., P(4-VP/S)-2.5, shown in Fig. 6b. On the other hand, AFM topography (Fig. 7a) in submicron scale provides individual 3-D contours that exhibit obviously different rough structures at the submicron scale (<0.1 mm) amid the samples in question, which are consistent to the above SEM microstructural analysis. Furthermore, the corresponding 2-D depth profiles (Fig. 7b) show a top view of each film, from the image of which the root-mean-square (RMS) surface roughness level is obtained (Table 2) by applying the power spectral density analysis [24]. It is clear that the films (with x ¼ 0.5 and 1) exhibit rougher surfaces at the 10-nm scale than the PS control (x ¼ 0) and the two in this series (with x ¼ 2.5 and 5). The PS control reveals a far smaller RMS roughness due to the lacking of submicron grains. On the other side, the films (with x ¼ 2.5 and 5) display moderate submicron roughness as the result of increasing coalescence of nanoparticles. A slight rise of roughness when x ¼ 5 can be the result of stronger phase inversion accompanying vaporization of DMF and condensation of moisture.
176
Z. Xing et al. / Polymer 87 (2016) 170e180
Fig. 6. a. FE-SEM images (7000) of surface morphology of the cast P(4-VP/S)-x% films. b. FE-SEM images of surface morphology of the same cast films at a high magnification (70,000).
Z. Xing et al. / Polymer 87 (2016) 170e180
177
Fig. 7. AFM images of the cast P(4-VP/S)-x% films displayed by (a) 3-D surface contour at the sub-micron scale and (b) 2-D depth profile.
According to the above examinations on the microstructures of the five cast films, the submicron grains comprising partially
coalesced nanoparticles (rosettes-like) are vital to the scattering of red light (622e780 nm) and in particular NIR (ca. 0.8e5 mm)
178
Z. Xing et al. / Polymer 87 (2016) 170e180
Table 2 The RMS surface roughness of the cast P(4-VP/S)-x% films.a Sample
P(4-VP/S)-0
P(4-VP/S)-0.5%
P(4-VP/S)-1%
P(4-VP/S)-2.5%
P(4-VP/S)-5%
Equivalent RMS
8.0 nm
46.0 nm
20.6 nm
16.5 nm
18.9 nm
a
The RMS data are generated from the AFM data by the software (Nanoscope Analysis 1.40).
radiations. Moreover, P(4-VP/S)-0.5% exhibits more distinct nanoroughness on submicron grains than P(4-VP/S)-1% and hence displays a stronger reflectance profile than the latter. As to the P(4-VP/ S)-0 film, despite a relatively flat and dense surface, there are submicron humps assembled by rather compact nanoparticles over its surface, which leads to a solar reflectance profile just below that of the P(4-VP/S)-1% film. The particular NIR reflectivity of the rosettes-like grains has been theoretically attributed to scattering through absorption of particular infrared bands (>2000 cm1 or > 5 mm) by the polymer [14]. The four P(4-VP/S)-x films (x s 0) display strong and similar reflectance profiles in UV light (390e455 nm) and blue light (455e492 nm) ranges. This phenomenon can be attributed to the random aggregations of submicron granules (Fig. 6b) according to a previous study [25]. Contrary to this granule-based reflective mechanism, the cast film of P(4-VP/S)-0 possesses a relatively smoother and denser surface and hence shows weaker reflectivity to the UV-blue lights. Moreover, with respect to the reference light source b891, the sequence of the reflectance values (Fig. 5) could represent their overall reflecting capabilities to UVeVis lights (or white light ranging 390e700 nm) because a white ceramic tile reflects 85.9% of this reference light source and completely reflects white lights as well. Accordingly, the UVeVis reflectance of the P(4VP/S)-x% films follows the order (by the x value): 0.5 > 1 > 0 > 2.5 > 5.0. Furthermore, the thermal analysis by DSC was employed to understand the polymer packing states in various cast films. The films (x ¼ 0 to 1.0), where the rosettes-like granules are present, reveal a slightly broader glass transition ranges in 3e5 C than the films (x ¼ 2.5 to 5), where the surface-smooth granules are present (Table 3). This analysis result indicates that more diversified polymer chain packing microenvironments exist in the former matrices relative to in the latter, suggesting that there would be more scattering and refraction when light travels in the former medium. 3.3. Hydrophilic doped end-segment of PS chains e essential to the solar reflectance Regarding the effect of 4-VP doping extent (x %) in the PS chain on the solar reflectance of the cast film as elaborated above, it is of interest to understand whether the doping by emulsion copolymerization is essential. Additionally, it is also appealing about how solar reflective property will be affected if other hydrophilic comonomer is in place of 4-VP. For the reason of easy quantifying, the doping extent of 1% was chosen to check the necessity of o/w emulsion copolymerization. Fig. 8 shows the solar reflectance profiles of the following cast films, P(4-VP/S)-1%, P4-VP(1%)-PS, and P(4-VP/S)-1%h, in which the last two samples were prepared by solution blending and solution copolymerization as detailed in
Fig. 8. Solar reflectance profiles of the cast films made of the two different copolymers P(4-VP/S)-1%, P(4-VP/S)-1%h and the polymer blend P4-VP(1%)-PS, respectively.
section 2.2. Compared with the microstructure of the P(4-VP/S)-1% film (Fig. 6b), neither of the latter two samples possesses submicron-grain and rough-matrix that are required to improve reflections to UVeVis lights as afore identified. The blending sample, P4-VP(1%)-PS, owing to the formation of nano-scaled particleboundary structure in its matrix, exhibits a stronger solar reflectance profile than P(4-VP/S)-1%h that has a plain matrix (Fig. 9). In addition, the DSC analysis of these three samples (SI: Fig. S4) shows that P(4-VP/S)-1%h owns the highest Tg of PS because of the homogeneous distribution of the polar pyridine units as aforementioned. This examination justifies the unique role of the pseudo block-copolymer chain structure for attaining a solar reflective efficacy as illustrated in Scheme 1. As far as the substitution of a hydrophilic monomer for 4-VP is concerned, three monomers, 4styrensulfonate sodium (4-SS), N-vinyl pyrrolidone (N-VP) and 1vinyl imidazole (1-VI), were chosen at the 1% doping level to synthesize the corresponding samples in the same o/w emulsion copolymerization system. In this study (Fig. 10), pristine PS sample, P(4-VP/S)-0, is included to benchmark the impact of hydrophilic doping. It turns out that the P(1-VI/S)-1% cast film shows the highest reflectance in this group, whose reflectivity values (~0.85) to two IR irradiation sources are even better than that of P(4-VP/S)1% (Fig. 8). It is assumed that the imidazole ring containing conjugated dipoles be the factor responsible to this difference. The rest two hydrophilic-doped PS samples display different reflectance from the pristine PS sample only in the two IR irradiation ranges. It
Table 3 Impact of the 4-VP content (x) on the glass transition of the coating film matrix P(4-VP/S)-x%. x
Tg ( C) Tg Range ( C)
0
0.5
1.0
2.5
5
101.16 4.9 99.6e104.5
101.68 5.4 99.3e104.7
102.05 5.1 101.3e106.4
102.87 3.3 102.4e105.7
102.08 3.4 101.2e104.6
Z. Xing et al. / Polymer 87 (2016) 170e180
179
Fig. 9. FE-SEM images of the cast films made of the copolymer P(4-VP/S)-1%h and the polymer blend P-4VP(1%)-PS, respectively.
Fig. 10. Solar reflectance profiles of the cast films made of the various o/w emulsion copolymers containing different hydrophilic monomer units.
could be concluded from this investigation that on the top of the pseudo block-copolymer chain structure the identity of hydrophilic group influences the solar reflectance and an enhanced conjugation extent of polar dipoles may further promote solar reflection to NIR irradiation. 4. Conclusion When a very low dose (1 mol%) of 4-vinylpyridine (4-VP) is copolymerized with styrene (S) in an o/w emulsion, a pseudo blockcopolymer chain structure is achieved, in which 4-VP units concentrate in a short end-segment of a PS-dominant long chain. The o/w emulsion polymerization system permits only a trivial dose of 4-VP relative to styrene in order to achieve the pseudo block copolymer chain structure. The symbol P(4-VP/S)-x% is used to denote the samples prepared. According to 1H-NMR characterization, the true content of 4-VP in the resulting copolymer is lower than the designated content (x%) in the monomer feed for polymerization. Such a minor content of co-monomer is normally insignificant to most of copolymer properties, the 4-VP is thus considered as a functionality dopant in a PS chain since it influences only specific applications. This minor 4-VP doping extent is essential for achieving a particular cast film matrix by using N,Ndimethylformamide (DMF) as solvent. It is because this pseudo-
block chain structure that prompts a three-stage matrix evolution to form a surface-rough film: from elementary nanoparticles (~10 nm) to submicron granules (50e100 nm) and lastly to a random collection of the granules, leading to a rough topography in micron scale. The resulting cast film exhibits improved solar reflectance ( 80%) versus its pristine PS counterpart. It has also been identified that the presence of the partially coalesced nanoparticles is crucial to the reflection of near infrared (NIR) irradiation. In addition, a slight increase in 4-VP content in the copolymer, such as P(4-VP/S)-2.5%, leads to a breath-figure patterned matrix comprising heavily coalesced elementary nanoparticles and hence a significant reduction in the NIR reflectance. This variation is attributed to the loss of pseudo block-copolymer chain structure due to an increase in the spreading of 4-VP units in the PS chain. In addition, the rough topography consisting of submicron granules is required to enhance reflectance to UVeVis lights. Relative to this, the pristine PS coating film reveals weaker UVeVis reflectance because of a lack of the rough topography. To assist elaborate the above result, two control samples were prepared by solution blending of PS and 1% P4-VP (based on monomer unit) and solution copolymerization of 1% 4-VP and styrene. Finally, this work has also attempted the substitution of another hydrophilic monomer at the 1 mol% content in feed with respect to styrene monomer. Amidst the three hydrophilic monomers examined, 1-vinyl imidazole (1VI) gives the highest solar reflectance, which is presumably associated with the conjugation of dipoles. Author contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest. Acknowledgment We acknowledge the funding 102 150 0057 (IMRE/10-2C0204) of A-STAR SERC ACAR program to support this project. Appendix A. Supplementary data Details of raw NMR, FTIR, DSC data and supporting figures are included. This material is available free of charge via the Internet at http://pubs.acs.org.
180
Z. Xing et al. / Polymer 87 (2016) 170e180
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2016.01.079. [13]
References [1] R. Levinson, P. Berdahl, H. Akbari, W. Miller, I. Joedicke, J. Reilly, Y. Suzuki, M. Vondran, Methods of creating solar-reflective nonwhite surfaces and their application to residential roofing materials, Sol. Energy Mater. Sol. Cells 91 (2007) 304e314. [2] M. Santamouris, A. Synnefa, T. Karlessi, Using advanced cool materials in the urban built environment to mitigate heat islands and improve thermal comfort conditions, Sol. Energy 85 (2011) 3085e3102. [3] R. Levinson, P. Berdahl, H. Akbari, Solar spectral optical properties of pigments e part II: survey of common colorants, Sol. Energy Mater. Sol. Cells 89 (2005) 351e389. [4] Q. Gao, X.M. Wu, Y.M. Fan, Solar spectral optical properties of rutile TiO2 coated mica-titania pigments, Dyes Pigment. 109 (2014) 90e95. [5] V. Fang, J.V. Kennedy, G. Science, J. Futter, J. Manning, G.S. Staff, A review of near infrared reflectance properties of metal oxide nanostructures, GNS Science, 2013. [6] P. Berdahl, Pigments to reflect the infrared radiation from fire, J. Heat. Trans. T. Asme 117 (1995) 355e358. [7] T.C. Grenfell, S.G. Warren, P.C. Mullen, Reflection of solar-radiation by the antarctic snow surface at ultraviolet, visible and near-infrared wavelengths, J. Geophys. Res. Atmos. 99 (1994) 18669e18684. [8] Z. Xing, S.-W. Tay, Y. Liu, L. Hong, Solar heat reflective coating consisting of hierarchically assembled polystyrene nanoparticles, Surf. Coat. Technol. 265 (2015) 99e105. [9] P. Muller-Buschbaum, Dewetting and pattern formation in thin polymer films as investigated in real and reciprocal space, J. Phys. Condens. Matter. 15 (2003) R1549eR1582. [10] C.S. Chern, Emulsion polymerization mechanisms and kinetics, Prog. Polym. Sci. 31 (2006) 443e486. [11] Y.M. Wang, C.Y. Pan, Study of the mechanism of the emulsifier-free emulsion polymerization of the styrene/4-vinylpyridine system, Colloid Polym. Sci. 277 (1999) 658e665. [12] H.M. Ni, Y.Z. Du, G.H. Ma, M. Nagai, S. Omi, Mechanism of soap-free emulsion
[14]
[15]
[16]
[17] [18]
[19] [20] [21] [22]
[23]
[24]
[25]
polymerization of styrene and 4-vinylpyridine: characteristics of reaction in the monomer phase, aqueous phase, and their interface, Macromolecules 34 (2001) 6577e6585. H.M. Ni, G.H. Ma, M. Nagai, S. Omi, Effects of ethyl acetate on the soap-free emulsion polymerization of 4-vinylpyridine and styrene. II. Aspects of the mechanism, J. Appl. Polym. Sci. 82 (2001) 2692e2708. L. Dombrovsky, S. Lallich, F. Enguehard, D. Baillis, An effect of “scattering by absorption” observed in near-infrared properties of nanoporous silica, J. Appl. Phys. 107 (2010) 1e8. , L. Rubatat, L. Billon, M. Save, Recent advances in honeycombP. Escale structured porous polymer films prepared via breath figures, Eur. Polym. J. 48 (2012) 1001e1025. M. Zamfir, C.S. Patrickios, F. Montagne, C. Abetz, V. Abetz, L. Oss-Ronen, Y. Talmon, Styrene-vinyl pyridine diblock copolymers: synthesis by RAFT polymerization and self-assembly in solution and in the bulk, J. Polym. Sci. Polym. Chem. 50 (2012) 1636e1644. R. Levinson, H. Akbari, P. Berdahl, Measuring solar reflectance-part ii: review of practical methods, Sol. Energy 84 (2010) 1745e1759. C.C. Shi, X. Chen, L. Chen, M.X. Chen, R. Jiang, pH and organic solvent-sensitive properties of crosslinked polymeric spheres formed by a soap-free emulsion polymerization, J. Appl. Polym. Sci. 128 (2013) 3034e3042. Kirsh, Infrared spectra V.P. Panov, L.A. Kazarin, V.I. Dubrovin, V.V. Gusev, Y.E. of atactic poly-4-vinylpyridine, J. Appl. Spectrosc. 21 (1974) 1504e1510. C.Y. Liang, S. Krimm, Infrared spectra of high polymers .6. Polystyrene, J. Polym. Sci. 27 (1958) 241e254. E. Yilmaz, Z. Kucukyavuz, Monomer reactivity ratios of styrene-4-vinypyridine copolymers at low and high conversions, Polymer 34 (1993) 145e149. Y.Y. Noh, S.J. Kang, J.B. Koo, I.K. You, K.J. Baeg, Organic Thin Film Transistor and Method of Manufacturing the Same, Electronics and Telecommunications Research Institute, 2010. P. Muller-Buschbaum, J.S. Gutmann, C. Lorenz, T. Schmitt, M. Stamm, Decay of interface correlation in thin polymer films, Macromolecules 31 (1998) 9265e9272. P. Muller-Buschbaum, E. Bauer, O. Wunnicke, M. Stamm, The control of thin film morphology by the interplay of dewetting, phase separation and microphase separation, J. Phys. Condens. Matter. 17 (2005) S363eS386. C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Particles, John Wiley & Sons, Inc., New York, 1983.
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
Solar heat reflective coating formed of polystyrene chains bearing 4-vinylpyridine-rich end segments Zheng Xing a, Siok-Wei Tay b, Yeap Hung Ng a, Liang Hong a, b, * a b
Department of Chemical & Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore, 119260, Singapore Institute of Materials Research and Engineering, Agency for Science, Technology and Research, A*STAR, 3 Research Link, Singapore, 117602, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 October 2015 Received in revised form 26 January 2016 Accepted 30 January 2016 Available online 13 February 2016
Incorporation of a minor dose of 4-vinylpyridine (4-VP) into an end segment of polystyrene (PS) chain by emulsion copolymerization yields a copolymer molecule P(4-VP/S) with weak surface-activity. The spin cast thin films of P(4-VP/S) display enhanced solar light reflectance of R ¼ 80e90% compared with its PS homopolymer counterpart. It is imperative to use dimethylformamide (DMF) as casting solvent because the hydrophilic nature of DMF induces core-to-grain assembling while dried in humid air, leading to a nodular film matrix. A rise of 4-VP dose in P(4-VP/S) causes a breath-figure patterned matrix instead and consequently an obvious reduction of reflectance in the near-infrared region. Mechanistically, P(4-VP/S) chains that bear a short end-segment with rich 4-VP units constitute tiny hydrophilic cores because of entrapment of moisture during drying, surrounding which the long PS tails thus undergo chain coiling to form nanoparticles. They then aggregate to form submicron grains in rosette and the partial coalescence of adjacent grains results in micron lumps. Such hierarchical evolution finally constructs a matrix exhibiting an enhanced solar reflectance. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Solar reflective coating 4-pyridine Polystyrene Chain assembling w/o emulsion
1. Introduction Solar light consists of ultra violet (l < 400 nm), visible (400e700 nm) and near infrared (ca. 700e2500 nm) radiations. Although some particular wavelengths of the infrared radiation are absorbed by O2, H2O, and CO2 in the atmosphere, UVevisible spectrum (300e700 nm) and near infrared (NIR) spectrum (700e2500 nm) constitute 48% and 52% of solar heat reaching the earth, respectively [1]. Energy consumption for air-conditioning with urban expansion shows an increasing trend worldwide and is a major concern in those hot climate countries. To lessen the intense demand for cooling building and car interior space, solar heat reflective (SHR) coatings play an important role [2]. Currently, the SHR coatings on market [3] are mainly hybrid polymer composite thin layers containing inorganic particles with high refractive indices, e.g., rutile TiO2, ZnO and their modified forms, which are uniformly embedded in various polymer matrices, typically acrylic, vinylacetate, and styrenic types [4,5].
* Corresponding author. Department of Chemical & Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore, 119260, Singapore. E-mail address: [email protected] (L. Hong). http://dx.doi.org/10.1016/j.polymer.2016.01.079 0032-3861/© 2016 Elsevier Ltd. All rights reserved.
A SHR composite coating generally obeys Mie scattering mechanism, by which the reflection happens at the interface between inorganic filler particles and polymer matrix due to different refractive indices [3]. The particle size and identity of the filler vitally affected the light reflection from this type of coating, for instance, P. Berdahl's work [6] shows that 0.22 mm is the optimal diameter of particles to scatter visible light and larger particle sizes are required to reflect NIR light. As to the nano fillers (<100 nm) used, being much smaller compared to the UVeVis wavelengths (l), they render prevailed Rayleigh scattering in all directions. Therefore, the scattering is effective only to short wavelength lights due to its l4 dependence. Compared to the composite SHR coating matrix, the pristine polymer coating matrix normally lacks SHR capability because commonly available polymer resins possess relatively low refractive indices in the range of 1 to 2. In consequence, there have been scarce studies thus far to explore solar reflectivity of pristine polymer coatings. Inspired by the high UVeVis albedo of tiny ice crystallites [7], we have recently found that PS nanoparticles, resulted from discrete chain folding mode, can effectively reflect solar light [8]. About this finding, when a cast DMF liquid film of PS is subjected to drying under humid ambient condition, discrete PS nanoparticles grow and coalesce subsequently to form submicron grains. The subsequent aggregation of
Z. Xing et al. / Polymer 87 (2016) 170e180
these grains results in micron-sized lumps over the surface of the cast film. This hierarchical evolution gives rise to a solar reflectance profile above 0.7, which is uncommon since otherwise a reflectance below 0.1 is resulted if a hydrophobic solvent such as chloroform or toluene is used instead of DMF to develop casting. Meanwhile, the interfacial potential driven dewetting phenomenon, which has been thorough studied [9], is unlikely to happen in the above coating films because the coating films fall in the dimension that is too thick for dewetting to happen and the solvent used for spin coating is miscible with moisture condensed from air during drying in ambient conditions. Such discrepancy in reflectivity was further validated using the infrared lamp irradiating in front of a PS coating on a glass panel. The temperature on the backside of the panel was recorded with time. The PS coating based on DMF shows an apparently lower temperature-time curve than its counterpart based on toluene (Fig. 1). As described above, gradual absorption of ambient moisture by a DMF-PS liquid film while being dried is critical to attain a solar reflective PS coating because the diffusion of water into DMF prompts chain coiling of PS. In the present work, we incorporate a very low mole fraction of hydrophilic 4-VP monomer units into PS chains through copolymerization in an oil-in-water (o/w) emulsion and investigate the impact of the resulting 4-VP-doped PS chain structure on the solar reflectivity by applying the same solution casting technique. It is hypothesized that a number of 4-VP-doped PS end-segments may associate together to generate a hydrophilic core, which is then surrounded by PS nanoparticles generated from chain coiling. This manner is expected to enhance heterogeneity within submicron scale and boost solar reflectivity. The resulting cast film indeed displays an enhanced solar reflectance relative to the pristine PS counterpart. Regarding the emulsion copolymerization of 4-VP and styrene, both the emulsifier-based system [10] and the soap-free system [11e13] have showed the formation of pyridine-rich radical species in the aqueous phase at the initial stage of copolymerization, which eventually results in copolymers possessing surface activity due to the formation of pseudo blockcopolymer chain structure. In the present preparation, since only a minor dose of 4-VP vs. styrene is in the monomer feed, sodium dodecyl sulfate (SDS) has to be used to stabilize the emulsion throughout the entire polymerization process. It is rational that 4VP units primarily appear in a very short end-segment of a PS dominant chain to function as dopant. The pseudo block-copolymer chain structure is therefore allowed to form hydrophilic nano-cores in the cast DMF-PS liquid
171
film accompanying vaporization of DMF in ambient air due to dissolution of moisture in the liquid coating. Each hydrophilic nano-core would then facilitate the coiling of the long PS tails surrounding the core and then nucleation of the coils to form grains in rosettes as illustrated in Scheme 1. The further agglomeration of these grains constitutes lumps in micron sizes, which play the role of scattering the long-wavelength component of NIR [14]. The resulting cast film clearly demonstrates enhanced solar reflectivity in contrast to the homo-PS film. Nevertheless, a slight increase in 4VP doping extent in the chain-end segment of PS causes a breathfigure patterned matrix [15], which exhibits obviously weaker light reflectivity in the NIR range. On the other hand, relative to the o/w emulsion copolymerization, although the synthesis of P(4-VP/ S) by Reversible Addition-Fragmentation Chain Transfer (RAFT) polymerization method allows for attaching a pure 4-VP segment to a PS block [16], we found that the RAFT block-copolymer shows an apparently weaker solar reflectance profile than the emulsion sample made by the same monomer-feed (viz. 1 mol % 4-VP). This happens because the RAFT polymerization produces far shorter PS blocks than the emulsion polymerization. Molecular weight has been well known a vital factor affecting chain packing and folding and hence the thin film bulk and surface structures. Moreover, the o/w emulsion copolymerization system allows substitution of 4-VP by another hydrophilic monomer. The present study has also examined the other three water-soluble monomers that indeed cause the final cast films display rather different solar reflectance profiles. In short, attaching a considerably short chain-end segment rich in 4-VP units to a long PS chain permits a hierarchical evolution from hydrophilic nuclei to nanoparticles of PS and then to micron lumps comprising the nanoparticles in the cast film accompanying evaporation of DMF. The final film matrix is unique because it sustains the maximum solar reflectance and a variation from this structural characteristic causes a noticeable reduction in solar reflectance. 2. Experimental section 2.1. Materials Styrene (Reagent Plus®, 99%, SigmaeAldrich), 4-vinylpyridine (4-VP, 96%, Alfa Aesar), N-vinyl pyrrolidone (~97%, Fluka) and 1vinyl imidazole (99þ%, Aldrich) were passed respectively through an inhibitor-removal column before use. Polystyrene (PS, average Mw ~192,000, SigmaeAldrich), poly(4-vinylpyridine) (P4-VP, Mw ~50,000, Polysciences), 4-styrenesulfonic acid, (sodium salt hydrate, Aldrich), B-79 (polyvinyl butyral, Butvar®, Eastman), N,Ndimethylformamide (DMF, min. 99.5% purity, Merck), toluene (99.99% purity, Fischer), methyl ethyl ketone (MEK, min. 99%, Fischer), sodium dodecyl sulfate (SDS, 99%, Fluka), ammonium persulfate (APS, min. 98%, Merck), inhibitor remover (for removing tert-butylcatechol, Aldrich) were used as received. 2.2. Doping PS chain-end by 4-VP
Fig. 1. The IR lamp irradiation timeetemperature profiles of the PS cast films developed by using toluene and DMF respectively as solvent.
In a typical preparation, a mixture of styrene (2.1 ml, 22.2 mmol) and 4-vinylpyridine (0.02 ml, 0.2 mmol) was introduced in a 20 ml aqueous solution containing 0.5% SDS and 0.2% (or 8.7 mM) ammonia persulfate, which was followed by magnetic stirring in a 50 ml round bottom flask accompanying by Ar bubbling for 20 min to remove oxygen. Then the flask was immersed into an oil bath at 70 C with stirring at 400 rpm to carry out emulsion polymerization. The polymerization was lasted for 24 h in Ar and then cooled down to room temperature. The polymer latex formed was filtrated and washed first by deionized water and then ethanol for a few times to remove possible P4-VP oligomer formed primarily in the
172
Z. Xing et al. / Polymer 87 (2016) 170e180
Scheme 1. A schematic illustration of the evolution of cast film matrix with roughness in microns from the hydrophilic nucleation of pseudo block copolymer chains to coalescences of granules.
aqueous phase, the surfactant and residual monomers. Subsequently, the product was dried in oven at 50 C and then collected in a sample vial. In the same preparation, the mole fraction of 4-VP in the feed of polymerization was varied from 0 to 5%. The resultant polymers are labeled by P(4-VP/S)_x, where S is the short form for styrene and x stands for the dose of 4-VP used, for example P(4-VP/ S)-0.5% is the sample made by the feed containing 0.5 mol% 4-VP and styrene. In addition, an alternative hydrophilic monomer, 4styrenesulfonic acid (4-SS), N-vinyl pyrrolidone (N-VP) and 1vinyl imidazole (1-VI), was used respectively in place of 4-VP (at 1% loading extent) to synthesize copolymer chain. In addition, a random copolymer sample was prepared as a control sample in a homogeneous solution comprising toluene (4 ml), AIBN (31.8 mg), styrene (2 g) and 4-VP (20 mg). Similarly, the polymerization was carried out at 70 C in Ar for 20 h, after that, the product was precipitated in a large excess of methanol, followed by vacuum drying at 40 C for 24 h. This control is labeled by P(4-VP/ S)-1%h. Another control was the polymer blend consisting of P(4VP/S)-0 and P4-VP (Polysciences) (1 mol% based on monomer units). The sample was prepared by mixing in DMF. 2.3. Preparation of coating film by solution casting approach A primer coat that consists of the subsequent B-79 and PS double layers were laid on a glass slide (Sail Brand 100 100 ), where B-79 functions as an adhesive and PS a transition layer. As the first layer, a B-79 solution (10 wt. %) in the binary MEK/toluene (v/v ¼ 1) solvent was spin cast on the glass slide using 600 rpm for 10 s, which was then dried at room temperature for 5 min. After that, a thin layer of PS (SigmaeAldrich) was spin-cast on the B-79 layer and dried by the same procedure to complete the bedding of primer. To develop a topcoat on the primer, typically, a polymer sample (0.7 g) and 5 ml DMF were mixed and stirred in a sample vial at room temperature overnight to obtain a homogeneous solution. It was spin-coated at 600 rpm on the primer for 10 s. The liquid film thus prepared was dried under ambient condition
characterized by 65e70% relative humidity (RH) and 25 C until no change in mass with time. Additionally, PS coatings on glass panel (10 cm 10 cm) were prepared to test the shielding capability against IR irradiation. The two solutions of PS (12 wt %) in toluene and DMF were coated, respectively, by brushing on the glass panel and then subjected to drying in ambient condition. 2.4. Measurement of the total solar reflectance and the shielding effect of infrared irradiation The reflectance profile of a coating film fabricated from the above protocol was gauged using a Solar Spectrum Reflectometer (SSR-6, Devices and Services Company). The reflectometer is able to measure reflectance with a standard deviation of 0.005 units and a bias of ±0.002 units. The reflectometer was calibrated using a blackbody cavity, with zero solar reflectance, and by a smooth and dense white tile with the designated reflectance reading of 0.859, at the solar irradiance b891 (ASTM E891-87 air mass 1.5 beam normal), namely, the AM 1.5 spectra that provides a reference point corresponding to a particular set of atmospheric conditions and a specific air mass [17]. The reflectance profile is plotted versus the UV, blue, red and IR light sources. The parenthesis beside each light source labels the color temperature in K of the source. According to Planck's law, the color temperatures in the range from 2000 to 3000 K signify relatively weak radiant emittances (between 0 and 0.5 1013 W/m3) in the wavelength range from 0.1 to 500 mm. Taking Blue (3125 K) for example, it describes blue light with the maximum emittance of approximately (5 1011 W/m3) at 0.47 mm over a broad wavelength range. The shielding against IR irradiation was carried out in an inhouse setup. An IR irradiance lamp (Philips HP 3616, 50 W) was placed 30 cm away in front of the coated glass panel that is fixed in a thermal insulation box made in-house. A thermocouple (Easyview 15) is attached to the backside of the coating panel. The temperature shown on the Extech Instrument was recorded every
Z. Xing et al. / Polymer 87 (2016) 170e180
minute in the first 10 min period and every 10 min in the last 20 min period. 2.5. Characterizations of the P(4-VP/S)-x% copolymers Detailed morphology of a topcoat was scrutinized by fieldemission scanning electron microscopy (FESEM, JEOL, JSM6700F). The surface contour of a coating and its roughness were recorded by atomic force microscopy (Bruker Dimension ICON). The samples were scanned by tapping mode in a designated size, e.g. 1 mm 1 mm, and in a scanning frequency of 0.6e1.0 Hz. The obtained AFM data were processed into 3-D pictures using the software (Nano Scope Analysis 1.40). Various polymer packing states in the cast films were analyzed by differential scanning calorimetry (DSC-Q100 by TA Instruments) in the range from 25 to 250 C at a scanning rate of 5 C/min, for which each sample was accurate to 3.1 ± 0.01 mg. The distribution of 4-VP units in a P(4-VP/S)-x% chain was also studied by using Fourier transform infrared spectroscopy (Bio-Rad Excalibur FTS-3500 FTIR spectrometer), for which the samples are polymers directly obtained from polymerization. The 1 H-NMR spectra of different P(4-VP/S)- x% samples were obtained from a Bruker Ultra Shield spectrometer (400 MHz), using chloroform-d as solvent. Both the area of peak a from the protons of pyridine ring and the area of doublet b from the protons of benzene ring were used to determine the real mole fractions of 4-VP in the different samples. The dielectric constants of different P(4-VP/S)-x% samples were measured on a dielectric thermal analyzer (DETA, DS6000) using a frequency of 10 kHz at room temperature. The testing pellet (ca. 13 mm 0.5 mm) was prepared by pressing a powder (0.1 g), obtained from a purified emulsion polymerization product, in a cylindrical die-set to 7.5 metric tons that was held for 5 min. Transmission electron microscopy was employed to examine the self-assembling behavior of the P(4-VP/S)- x% samples: a DMF solution (ca. 1%) was prepared and 1 drop of the solution was transferred to a copper grid, which was allowed to dry at the ambient condition.
173
initiator molar ratio (100:1) compared with the usual range. The real 4-VP content (Table 1) in each polymer sample was determined by 1H-NMR (Supporting information (SI): Fig. S1), which is lower than the designated dose in the respective monomer feed likely due to formation of P4-VP oligomer in the water phase. For the simplicity reason, the following discussion will still use the feed dose of 4-VP (x%) to denote a sample, i.e., P(4-VP/S)-x%. Regarding the chemical environment of 4-VP unit in P(4-VP/S)-x% with the variation of x value, the FT-IR spectra (Fig. 2) permits qualitatively correlating the x value with the relative peak intensity: the characteristic adsorption of the asymmetric stretching vibration of C] N of 4-VP unit at 1557 cm1 [18,19] is identified besides its symmetric ring stretching absorbance (at 1600 cm1) that overlaps with the same absorbance of styrene unit. In addition, the absorbance at 1540 cm1 is known as a combination band of the two outof-plane vibration modes of aromatic rings [20]. It is clear that these two IR absorption bands (1557 and 1540 cm1) become more comparable in peak intensity with the increase in x value. In
3. Results and discussion 3.1. Characterization of the copolymer chain structures of P(4-VP/ S)-x% The P(4-VP/S) samples, although structurally being a copolymer chain synthesized by the o/w emulsion copolymerization, are described as 4-VP-doped PS in this work because the solar reflectivity is susceptible to maintaining a substantially low content of 4VP. According to a previous study [10], both the micelle and homogeneous nucleation steps govern the development of latex particles in the o/w emulsion polymerization system. The homogeneous nucleation mechanism favors the formation of a pseudo block-copolymer copolymer, the trend of which will increase with the use of a lower monomer/initiator ratio as well as a lower dose of hydrophilic monomer, e.g. 4-VP relative to styrene in the present system. With the aim to attain a pseudo block-copolymer chain structure, besides using very low 4-VP/S molar ratios, the formulation of monomer feed also used a relatively low monomers/
Fig. 2. The FT-IR spectra of the P(4-VP/S)-x% samples in the expanded wavelength number range from 1500 to 1580 cm1.
Table 1 The real 4-VP contents in the copolymer synthesized according to 1H-NMR and their properties. Samples
P(4VP/S)-0
P(4VP/S)-0.5%
P(4VP/S)-1%
P(4VP/S)-2.5%
P(4VP/S)-5%
P(4VP/S)-1%_h
4-VP (mol%) Mn ( 104) Mw/Mn
0 50.1 1.76
0.13 61.1 1.56
0.31 82.4 1.47
1.20 63.0 1.59
2.67 74.6 1.43
0.30 1.47 1.67
174
Z. Xing et al. / Polymer 87 (2016) 170e180
principle, the asymmetric ring stretching amplitude of the pyridine unit should be stronger with a more even distribution of it in styrene units due to a reduction of dipoleedipole interactions between 4-VP units. Moreover, as a control sample, P(4-VP/S)-1%h obtained from the solution polymerization exhibits almost the same peak intensities between these two IR bands. In this control, 4-VP units have a uniform distribution along the P(4-VP/S)-1%h chains because both monomers have the close r values (<1) in the solution copolymerization [21]. In contrast, the P(4-VP/S)-1%, despite containing almost the same 4-VP content as this control sample (Table 1), shows obviously different intensities between the two bands as the result of forming the pseudo-block copolymer structure (also refer to SI: Fig. S2). It can therefore be concluded that P(4-VP/S) is more resemble to a block copolymer with the decrease in 4-VP content. In addition to the above IR spectroscopy characterization, a consistent conclusion can also be obtained from the relative permittivity (εr) measurement of the polymer latex particles (Fig. 3). The particles must have an inward decreasing concentration gradient of 4-VP unit according to the o/w emulsion polymerization mechanism and the hydrophilic nature of 4-VP. Formation of the pseudo block-copolymer chain structure would leave behind a very thin polar shell comprising rich 4-VP units. Hence a larger εr is expected because of a distinct distribution of polar 4-VP units vs. S unit. The P(4-VP/S)-0 displays a slightly higher εr (~3.1) than the common PS (~2.7) due to a collection of terminal sulfate groups, transferred from the free-radical initiator, on the surface of latex particles. The test result shows that εr increases from P(4-VP/S)-0 to P(4-VP/S)-0.5% and then to P(4-VP/S)-1%, which is attributed to the accumulation of 4-VP units in the surface shell of latex particles because of the extension of the pseudo-4-VP block. After that, a decreasing trend of εr is observed instead with the introduction of more 4-VP into the polymerization system, from P(4-VP/S)-2.5% to P(4-VP/S)-5%. It implies that an increase of the inward distribution of 4-VP units in each latex particle owing to the increase in the micelle nucleation mechanism [10], which weakens the electrical polarization extent in latex particles. The same conclusion as the above IR characterization can therefore be drawn: a low 4-VP dose in the emulsion polymerization favors concentrating 4-VP units in short chain ends and therefore leads to a thin polar shell on individual latex particles. This view is further proven
Fig. 3. Variation of the relative permittivity value (at 10 kHz and 25 C) of the o/w emulsion latex particles with different 4-VP doses in the P(4-VP/S)-x% copolymers.
by the control sample, a powder blend of 1% homo-P4-VP and P(4VP/S)-0, which exhibits basically the same permittivity as the P(4VP/S)-1% and the pure P4-VP (~4.4) [22], although the actual 4-VP content in P(4-VP/S)-1% is well below 1% according to the above 1 H-NMR quantification. Both the IR spectroscopy study and relative permittivity measurement latex have indicated that distribution of 4-VP units in the PS chain leaps since the 4-VP content in the feed of polymerization reaches 2.5%. Furthermore, proposing that a variation of 4-VP distribution in PS chains is to affect the self-assembling behavior of the polymer in a proper liquid medium, we inspected this property by designing a test (section 2.5). A drop of a very dilute polymer solution in DMF was transferred to a TEM sample-holder and allowed for drying at room ambient conditions. DMF absorbs moisture from air while being evaporated, which triggers collection of the hydrophilic 4-VP-rich end segments of a certain number of polymer chains to form a hydrophilic core and a PS shell providing that most of 4-VP units concentrate in the end-segment of a chain. Besides this, the initial dilute condition permits formation of discrete particles after drying. The proposal is indeed supported by TEM images (Fig. 4) of the four P(4-VP/S)-x% (x ¼ 0 to 2.5) samples. They display the occurrence of submicron microspheres (for x ¼ 0.5 and 1) having a loose interior surrounded by a dense shell, in which the loose interior was left behind by drying. Contrary to these two samples, both (x ¼ 0 and 2.5) samples do not present such a hollow particle structure because of a lack of 4-VP and a pseudo-block copolymer chain structure, respectively. This outcome is consistent with the conclusion drawn from above IR and relative permittivity studies. This examination verifies the hydrophilic nucleation as illustrated in Scheme 1, which is structurally selective only for the lightly doped PS chains (x ¼ 0.5 and 1.0) and leads to submicron grains. It will be further noted that chain-packing steps surrounding the hydrophilic nuclei eventually enhances the NIR light reflectance on the surface of the cast film. 3.2. Impact of 4-VP content in P(4-VP/S) on the solar-reflectance The cast films made of individual P(4-VP/S)-x% samples by the protocol as described in section 2.3 display different solar reflectance profiles (Fig. 5). They could be roughly divided into two groups: x ¼ 0 to 1 versus x ¼ 2 to 5, because the former three coatings reflect at least 78% solar light from different light sources, while the latter two coatings show particularly weaker reflectivity in the IR wavelength range (from 0.2 to 400 mm with the maximum intensity at 1 mm). To understand this outcome, their surface morphologies in micron scale were inspected (Fig. 6a). As the control sample, the P(4-VP/S)-0 cast film, shows a dense matrix and a sparsely spread of micron-sized pothole-like pores. Similar but shallow holes also appear on the P(4-VP/S)-0.5% cast film, which has however a relatively rougher matrix. These potholes are formed due to condense of water droplets from ambient moisture likely in the final stage of drying. One of Müller-Buschbaum's previous studies [23] on the surface roughness of the as-prepared PS thin film on Si(100) has found that it is affected by molecular weight, film thickness and annealing temperature and duration. In the present work, the films, besides being far thicker than those studied in the above work, are developed on a double-layer primer. The primer layer plays a crucial role in alleviating the impact of substrate (glass plate), e.g. the surface-bending rigidity, on the surface microstructure of coating films and improving the interface compatibility. On the basis of the above observation, a further zooming in scrutiny scale (Fig. 6b) unveils the detailed microstructures of the above two film matrices (x ¼ 0 and 0.5). Although both matrices consist of elementary particles of 10 nm or smaller, these nanoparticles undergo extensive coalescence in the P(4-VP/
Z. Xing et al. / Polymer 87 (2016) 170e180
175
Fig. 4. TEM images of the self-assembled P(4-VP/S)-x% colloidal particles generated from drying a considerably diluted polymer solution in DMF in humid air. Note: the scale bar in white works for the 4 images.
Fig. 5. Solar reflectance profiles of the P(4-VP/S)-x% cast films.
S)-0.5% film to form submicron (50 nme0.5 mm) granules, whose random collection constructs a matrix with roughness in micron scale at maximum. Whereas, the P(4-VP/S)-0 film exhibits a much smoother and denser matrix assembled by nanoparticles, where numerous slightly convex submicron bumps are observed as well. Similarly, the P(4-VP/S)-1% film still retains the three-tier microstructures: nano-particle / submicron-grain / rough-matrix, formed via the mechanism proposed in Scheme 1. On contrary,
the P(4-VP/S)-2.5% film exhibits a microstructure comprising significantly coalescing nanoparticles, and hence only the last-two stage characteristics could be recognized. Besides this, the film has a honeycomb-like matrix at the micron scale (Fig. 6a). Correspondingly, the P(4-VP/S)-5% film shows a very similar matrix to P(4-VP/S)-2.5% but a greater extent of nanoparticle mingling (SI: Fig. S3). It is rational that the honeycomb-like matrix is generated through forming a hydrophilic continuous phase, which is opposite to the nucleation of hydrophilic core when the 4-VP content in the copolymer is substantially low. This inversion of 4-VP enriched blocks towards the DMF-water medium brings about amalgamation of PS dominant long blocks to form submicron grains. The agglomerating of these grains constitutes the framework of honeycomb matrix, e.g., P(4-VP/S)-2.5, shown in Fig. 6b. On the other hand, AFM topography (Fig. 7a) in submicron scale provides individual 3-D contours that exhibit obviously different rough structures at the submicron scale (<0.1 mm) amid the samples in question, which are consistent to the above SEM microstructural analysis. Furthermore, the corresponding 2-D depth profiles (Fig. 7b) show a top view of each film, from the image of which the root-mean-square (RMS) surface roughness level is obtained (Table 2) by applying the power spectral density analysis [24]. It is clear that the films (with x ¼ 0.5 and 1) exhibit rougher surfaces at the 10-nm scale than the PS control (x ¼ 0) and the two in this series (with x ¼ 2.5 and 5). The PS control reveals a far smaller RMS roughness due to the lacking of submicron grains. On the other side, the films (with x ¼ 2.5 and 5) display moderate submicron roughness as the result of increasing coalescence of nanoparticles. A slight rise of roughness when x ¼ 5 can be the result of stronger phase inversion accompanying vaporization of DMF and condensation of moisture.
176
Z. Xing et al. / Polymer 87 (2016) 170e180
Fig. 6. a. FE-SEM images (7000) of surface morphology of the cast P(4-VP/S)-x% films. b. FE-SEM images of surface morphology of the same cast films at a high magnification (70,000).
Z. Xing et al. / Polymer 87 (2016) 170e180
177
Fig. 7. AFM images of the cast P(4-VP/S)-x% films displayed by (a) 3-D surface contour at the sub-micron scale and (b) 2-D depth profile.
According to the above examinations on the microstructures of the five cast films, the submicron grains comprising partially
coalesced nanoparticles (rosettes-like) are vital to the scattering of red light (622e780 nm) and in particular NIR (ca. 0.8e5 mm)
178
Z. Xing et al. / Polymer 87 (2016) 170e180
Table 2 The RMS surface roughness of the cast P(4-VP/S)-x% films.a Sample
P(4-VP/S)-0
P(4-VP/S)-0.5%
P(4-VP/S)-1%
P(4-VP/S)-2.5%
P(4-VP/S)-5%
Equivalent RMS
8.0 nm
46.0 nm
20.6 nm
16.5 nm
18.9 nm
a
The RMS data are generated from the AFM data by the software (Nanoscope Analysis 1.40).
radiations. Moreover, P(4-VP/S)-0.5% exhibits more distinct nanoroughness on submicron grains than P(4-VP/S)-1% and hence displays a stronger reflectance profile than the latter. As to the P(4-VP/ S)-0 film, despite a relatively flat and dense surface, there are submicron humps assembled by rather compact nanoparticles over its surface, which leads to a solar reflectance profile just below that of the P(4-VP/S)-1% film. The particular NIR reflectivity of the rosettes-like grains has been theoretically attributed to scattering through absorption of particular infrared bands (>2000 cm1 or > 5 mm) by the polymer [14]. The four P(4-VP/S)-x films (x s 0) display strong and similar reflectance profiles in UV light (390e455 nm) and blue light (455e492 nm) ranges. This phenomenon can be attributed to the random aggregations of submicron granules (Fig. 6b) according to a previous study [25]. Contrary to this granule-based reflective mechanism, the cast film of P(4-VP/S)-0 possesses a relatively smoother and denser surface and hence shows weaker reflectivity to the UV-blue lights. Moreover, with respect to the reference light source b891, the sequence of the reflectance values (Fig. 5) could represent their overall reflecting capabilities to UVeVis lights (or white light ranging 390e700 nm) because a white ceramic tile reflects 85.9% of this reference light source and completely reflects white lights as well. Accordingly, the UVeVis reflectance of the P(4VP/S)-x% films follows the order (by the x value): 0.5 > 1 > 0 > 2.5 > 5.0. Furthermore, the thermal analysis by DSC was employed to understand the polymer packing states in various cast films. The films (x ¼ 0 to 1.0), where the rosettes-like granules are present, reveal a slightly broader glass transition ranges in 3e5 C than the films (x ¼ 2.5 to 5), where the surface-smooth granules are present (Table 3). This analysis result indicates that more diversified polymer chain packing microenvironments exist in the former matrices relative to in the latter, suggesting that there would be more scattering and refraction when light travels in the former medium. 3.3. Hydrophilic doped end-segment of PS chains e essential to the solar reflectance Regarding the effect of 4-VP doping extent (x %) in the PS chain on the solar reflectance of the cast film as elaborated above, it is of interest to understand whether the doping by emulsion copolymerization is essential. Additionally, it is also appealing about how solar reflective property will be affected if other hydrophilic comonomer is in place of 4-VP. For the reason of easy quantifying, the doping extent of 1% was chosen to check the necessity of o/w emulsion copolymerization. Fig. 8 shows the solar reflectance profiles of the following cast films, P(4-VP/S)-1%, P4-VP(1%)-PS, and P(4-VP/S)-1%h, in which the last two samples were prepared by solution blending and solution copolymerization as detailed in
Fig. 8. Solar reflectance profiles of the cast films made of the two different copolymers P(4-VP/S)-1%, P(4-VP/S)-1%h and the polymer blend P4-VP(1%)-PS, respectively.
section 2.2. Compared with the microstructure of the P(4-VP/S)-1% film (Fig. 6b), neither of the latter two samples possesses submicron-grain and rough-matrix that are required to improve reflections to UVeVis lights as afore identified. The blending sample, P4-VP(1%)-PS, owing to the formation of nano-scaled particleboundary structure in its matrix, exhibits a stronger solar reflectance profile than P(4-VP/S)-1%h that has a plain matrix (Fig. 9). In addition, the DSC analysis of these three samples (SI: Fig. S4) shows that P(4-VP/S)-1%h owns the highest Tg of PS because of the homogeneous distribution of the polar pyridine units as aforementioned. This examination justifies the unique role of the pseudo block-copolymer chain structure for attaining a solar reflective efficacy as illustrated in Scheme 1. As far as the substitution of a hydrophilic monomer for 4-VP is concerned, three monomers, 4styrensulfonate sodium (4-SS), N-vinyl pyrrolidone (N-VP) and 1vinyl imidazole (1-VI), were chosen at the 1% doping level to synthesize the corresponding samples in the same o/w emulsion copolymerization system. In this study (Fig. 10), pristine PS sample, P(4-VP/S)-0, is included to benchmark the impact of hydrophilic doping. It turns out that the P(1-VI/S)-1% cast film shows the highest reflectance in this group, whose reflectivity values (~0.85) to two IR irradiation sources are even better than that of P(4-VP/S)1% (Fig. 8). It is assumed that the imidazole ring containing conjugated dipoles be the factor responsible to this difference. The rest two hydrophilic-doped PS samples display different reflectance from the pristine PS sample only in the two IR irradiation ranges. It
Table 3 Impact of the 4-VP content (x) on the glass transition of the coating film matrix P(4-VP/S)-x%. x
Tg ( C) Tg Range ( C)
0
0.5
1.0
2.5
5
101.16 4.9 99.6e104.5
101.68 5.4 99.3e104.7
102.05 5.1 101.3e106.4
102.87 3.3 102.4e105.7
102.08 3.4 101.2e104.6
Z. Xing et al. / Polymer 87 (2016) 170e180
179
Fig. 9. FE-SEM images of the cast films made of the copolymer P(4-VP/S)-1%h and the polymer blend P-4VP(1%)-PS, respectively.
Fig. 10. Solar reflectance profiles of the cast films made of the various o/w emulsion copolymers containing different hydrophilic monomer units.
could be concluded from this investigation that on the top of the pseudo block-copolymer chain structure the identity of hydrophilic group influences the solar reflectance and an enhanced conjugation extent of polar dipoles may further promote solar reflection to NIR irradiation. 4. Conclusion When a very low dose (1 mol%) of 4-vinylpyridine (4-VP) is copolymerized with styrene (S) in an o/w emulsion, a pseudo blockcopolymer chain structure is achieved, in which 4-VP units concentrate in a short end-segment of a PS-dominant long chain. The o/w emulsion polymerization system permits only a trivial dose of 4-VP relative to styrene in order to achieve the pseudo block copolymer chain structure. The symbol P(4-VP/S)-x% is used to denote the samples prepared. According to 1H-NMR characterization, the true content of 4-VP in the resulting copolymer is lower than the designated content (x%) in the monomer feed for polymerization. Such a minor content of co-monomer is normally insignificant to most of copolymer properties, the 4-VP is thus considered as a functionality dopant in a PS chain since it influences only specific applications. This minor 4-VP doping extent is essential for achieving a particular cast film matrix by using N,Ndimethylformamide (DMF) as solvent. It is because this pseudo-
block chain structure that prompts a three-stage matrix evolution to form a surface-rough film: from elementary nanoparticles (~10 nm) to submicron granules (50e100 nm) and lastly to a random collection of the granules, leading to a rough topography in micron scale. The resulting cast film exhibits improved solar reflectance ( 80%) versus its pristine PS counterpart. It has also been identified that the presence of the partially coalesced nanoparticles is crucial to the reflection of near infrared (NIR) irradiation. In addition, a slight increase in 4-VP content in the copolymer, such as P(4-VP/S)-2.5%, leads to a breath-figure patterned matrix comprising heavily coalesced elementary nanoparticles and hence a significant reduction in the NIR reflectance. This variation is attributed to the loss of pseudo block-copolymer chain structure due to an increase in the spreading of 4-VP units in the PS chain. In addition, the rough topography consisting of submicron granules is required to enhance reflectance to UVeVis lights. Relative to this, the pristine PS coating film reveals weaker UVeVis reflectance because of a lack of the rough topography. To assist elaborate the above result, two control samples were prepared by solution blending of PS and 1% P4-VP (based on monomer unit) and solution copolymerization of 1% 4-VP and styrene. Finally, this work has also attempted the substitution of another hydrophilic monomer at the 1 mol% content in feed with respect to styrene monomer. Amidst the three hydrophilic monomers examined, 1-vinyl imidazole (1VI) gives the highest solar reflectance, which is presumably associated with the conjugation of dipoles. Author contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest. Acknowledgment We acknowledge the funding 102 150 0057 (IMRE/10-2C0204) of A-STAR SERC ACAR program to support this project. Appendix A. Supplementary data Details of raw NMR, FTIR, DSC data and supporting figures are included. This material is available free of charge via the Internet at http://pubs.acs.org.
180
Z. Xing et al. / Polymer 87 (2016) 170e180
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.polymer.2016.01.079. [13]
References [1] R. Levinson, P. Berdahl, H. Akbari, W. Miller, I. Joedicke, J. Reilly, Y. Suzuki, M. Vondran, Methods of creating solar-reflective nonwhite surfaces and their application to residential roofing materials, Sol. Energy Mater. Sol. Cells 91 (2007) 304e314. [2] M. Santamouris, A. Synnefa, T. Karlessi, Using advanced cool materials in the urban built environment to mitigate heat islands and improve thermal comfort conditions, Sol. Energy 85 (2011) 3085e3102. [3] R. Levinson, P. Berdahl, H. Akbari, Solar spectral optical properties of pigments e part II: survey of common colorants, Sol. Energy Mater. Sol. Cells 89 (2005) 351e389. [4] Q. Gao, X.M. Wu, Y.M. Fan, Solar spectral optical properties of rutile TiO2 coated mica-titania pigments, Dyes Pigment. 109 (2014) 90e95. [5] V. Fang, J.V. Kennedy, G. Science, J. Futter, J. Manning, G.S. Staff, A review of near infrared reflectance properties of metal oxide nanostructures, GNS Science, 2013. [6] P. Berdahl, Pigments to reflect the infrared radiation from fire, J. Heat. Trans. T. Asme 117 (1995) 355e358. [7] T.C. Grenfell, S.G. Warren, P.C. Mullen, Reflection of solar-radiation by the antarctic snow surface at ultraviolet, visible and near-infrared wavelengths, J. Geophys. Res. Atmos. 99 (1994) 18669e18684. [8] Z. Xing, S.-W. Tay, Y. Liu, L. Hong, Solar heat reflective coating consisting of hierarchically assembled polystyrene nanoparticles, Surf. Coat. Technol. 265 (2015) 99e105. [9] P. Muller-Buschbaum, Dewetting and pattern formation in thin polymer films as investigated in real and reciprocal space, J. Phys. Condens. Matter. 15 (2003) R1549eR1582. [10] C.S. Chern, Emulsion polymerization mechanisms and kinetics, Prog. Polym. Sci. 31 (2006) 443e486. [11] Y.M. Wang, C.Y. Pan, Study of the mechanism of the emulsifier-free emulsion polymerization of the styrene/4-vinylpyridine system, Colloid Polym. Sci. 277 (1999) 658e665. [12] H.M. Ni, Y.Z. Du, G.H. Ma, M. Nagai, S. Omi, Mechanism of soap-free emulsion
[14]
[15]
[16]
[17] [18]
[19] [20] [21] [22]
[23]
[24]
[25]
polymerization of styrene and 4-vinylpyridine: characteristics of reaction in the monomer phase, aqueous phase, and their interface, Macromolecules 34 (2001) 6577e6585. H.M. Ni, G.H. Ma, M. Nagai, S. Omi, Effects of ethyl acetate on the soap-free emulsion polymerization of 4-vinylpyridine and styrene. II. Aspects of the mechanism, J. Appl. Polym. Sci. 82 (2001) 2692e2708. L. Dombrovsky, S. Lallich, F. Enguehard, D. Baillis, An effect of “scattering by absorption” observed in near-infrared properties of nanoporous silica, J. Appl. Phys. 107 (2010) 1e8. , L. Rubatat, L. Billon, M. Save, Recent advances in honeycombP. Escale structured porous polymer films prepared via breath figures, Eur. Polym. J. 48 (2012) 1001e1025. M. Zamfir, C.S. Patrickios, F. Montagne, C. Abetz, V. Abetz, L. Oss-Ronen, Y. Talmon, Styrene-vinyl pyridine diblock copolymers: synthesis by RAFT polymerization and self-assembly in solution and in the bulk, J. Polym. Sci. Polym. Chem. 50 (2012) 1636e1644. R. Levinson, H. Akbari, P. Berdahl, Measuring solar reflectance-part ii: review of practical methods, Sol. Energy 84 (2010) 1745e1759. C.C. Shi, X. Chen, L. Chen, M.X. Chen, R. Jiang, pH and organic solvent-sensitive properties of crosslinked polymeric spheres formed by a soap-free emulsion polymerization, J. Appl. Polym. Sci. 128 (2013) 3034e3042. Kirsh, Infrared spectra V.P. Panov, L.A. Kazarin, V.I. Dubrovin, V.V. Gusev, Y.E. of atactic poly-4-vinylpyridine, J. Appl. Spectrosc. 21 (1974) 1504e1510. C.Y. Liang, S. Krimm, Infrared spectra of high polymers .6. Polystyrene, J. Polym. Sci. 27 (1958) 241e254. E. Yilmaz, Z. Kucukyavuz, Monomer reactivity ratios of styrene-4-vinypyridine copolymers at low and high conversions, Polymer 34 (1993) 145e149. Y.Y. Noh, S.J. Kang, J.B. Koo, I.K. You, K.J. Baeg, Organic Thin Film Transistor and Method of Manufacturing the Same, Electronics and Telecommunications Research Institute, 2010. P. Muller-Buschbaum, J.S. Gutmann, C. Lorenz, T. Schmitt, M. Stamm, Decay of interface correlation in thin polymer films, Macromolecules 31 (1998) 9265e9272. P. Muller-Buschbaum, E. Bauer, O. Wunnicke, M. Stamm, The control of thin film morphology by the interplay of dewetting, phase separation and microphase separation, J. Phys. Condens. Matter. 17 (2005) S363eS386. C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Particles, John Wiley & Sons, Inc., New York, 1983.