Tuning Morphology and Melting Temperature in Polyethylene Films by

Jan 10, 2018 - The control of structure and thermal stability in semicrystalline polymer films remains an important challenge in applications ranging ...
0 downloads 5 Views 5MB Size
Download PDF
Article Cite This: Macromolecules XXXX, XXX, XXX−XXX

Tuning Morphology and Melting Temperature in Polyethylene Films by MAPLE Hyuncheol Jeong,† Mithun Chowdhury,† Yucheng Wang,† Melda Sezen-Edmonds,† Yueh-Lin Loo,†,∥ Richard A. Register,†,§ Craig B. Arnold,‡,§ and Rodney D. Priestley*,†,§ †

Department of Chemical and Biological Engineering, ‡Department of Mechanical and Aerospace Engineering, §Princeton Institute for the Science and Technology of Materials, and ∥Andlinger Center for Energy and the Environment, Princeton University, Princeton, New Jersey 08544, United States S Supporting Information *

ABSTRACT: The control of structure and thermal stability in semicrystalline polymer films remains an important challenge in applications ranging from solar energy devices to packaging films. Here, we demonstrate the ability to dramatically alter the morphology and melting temperature (Tm) of low-molecularweight linear polyethylene (PE) by employing an innovative vapor-assisted deposition process termed matrix assisted pulsed laser evaporation (MAPLE). We report the ability to tune Tm of PE films by 20 °C by simply adjusting the deposition temperature during MAPLE processing. This unique capability stems from the ability of MAPLE to exploit confined crystallization during thin film growth. In addition, we demonstrate the ability to exploit MAPLE to design PE films that exhibit the same Tm as their melt-crystallized analogues but have an ∼25% higher degree of crystallinity. Our investigation offers new insights into how confinement effects in polymer crystallization can be utilized in the emerging field of polymer film fabrication by MAPLE to control structure and key material properties of semicrystalline polymer films.

■

INTRODUCTION From next-generation solar energy harvesting devices1,2 to various types of plastic microelectronics,3−6 many of the polymeric materials that will enable tomorrow’s technologies are semicrystalline. Rather than comprising a single uniform phase, semicrystalline polymers are characterized by foldedchain lamellar crystals with intercalated entanglements that remain amorphous.7−10 The structure within semicrystalline polymers has a strong influence on thermal stability, i.e., the melting temperature (Tm) as well as the mechanical11−14 and electrical10,15−17 properties of semicrystalline polymers. The precise control of structure and Tm in polymer thin films, without the need for chemical modification13,18−25 or the inclusion of functional additives,14,26 remains of interest and a formidable challenge in material science. Here, we introduce a new and promising approach to dramatically tune the structure and Tm of semicrystalline polymer films by exploiting the unique deposition capabilities of matrix assisted pulsed laser evaporation (MAPLE). In the MAPLE process, polymer films are grown by laser ablation of a frozen dilute solution of polymer and solvent, in which the solvent serves as a sacrificial matrix to transport the polymer into the gas phase and eliminate polymer degradation.27,28 MAPLE has been shown to be a promising approach to deposit functional polymer films.29−35 Via MAPLE, film growth proceeds by the deposition of nano- to micrometer-sized © XXXX American Chemical Society

isolated (or confined) polymer droplets with simultaneous crystallization at a controlled substrate deposition temperature (Tdep).36,37 By exploiting this feature, we demonstrate that for low-molecular-weight linear polyethylene (PE; Mn = 3000, Mw/ Mn = 1.10) Tm can be tuned by 20 °C simply by controlling Tdep during MAPLE deposition. To place these results in context, it is not practically feasible to tune Tm by more than 5 °C via melt crystallization for PE.38,39 Furthermore, we demonstrate that while PE films processed by MAPLE can be designed to exhibit the same Tm as their melt-crystallized analogues, the degree of crystallinity is always notably enhanced. This tunability in the structure of PE stems from the ability of MAPLE to confine polymer crystallization into nanoscale dimensions during film formation.

■

RESULTS AND DISCUSSION Film Formation by Additive Deposition and Crystallization of Confined PE Droplets. To gain an understanding of the crystallization mechanism of PE films formed by MAPLE processing, we first investigated the impact of Tdep on crystallization during the early stages of film formation, i.e., when the film consists of isolated nano- and microsized Received: November 2, 2017 Revised: December 14, 2017

A

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 1. (a) Optical microscopy image of MAPLE-PE on Si formed with 30 min of deposition from a 0.2 mg/mL solution target at Tdep = 116 °C. The inset shows the structure of a MLC grown from a nucleating droplet of size ∼1 μm3. (b) AFM height image magnifying the MLC in R1 in part a. (c) AFM amplitude image of MAPLE-PE formed with Tdep = 75 °C showing the morphology of PE nanocrystals consisting of a single lamella or a few stacked lamellae. The white arrow indicates a crystal with a screw dislocation. The inset schematic shows the structure of the PE nanocrystals. (d) AFM height image magnifying the MLCs in R2 in part c. (e) Representative height profiles of PE MLCs formed at various Tdep. (f) Bar graph comparing MLC thicknesses of MAPLE-PE formed at various Tdep; error ranges represent the standard deviation of the height measurements on multiple isolated PE MLCs formed at each Tdep.

Figure 2. (a) Optical microscopy image of MAPLE-PE on Si, formed with 1.5 h of deposition from a 1 mg/mL solution target at Tdep = 25 °C. (b, c) Optical microscopy images of a flash-DSC chip where (b) shows the chip covered with a polyimide mask. The mask has a laser-machined hole of size 250 × 250 μm2 in the center (indicated by arrow), which exposes the sample stage of the chip. (c) Optical image that shows the sample stage of the chip after MAPLE PE deposition. The white dashed line marks the region of the deposited film.

When PE was deposited at Tdep = 75 °C, the MLCs were composed of folded chains with d ∼ 10 nm. Interestingly, nucleating droplets with sizes down to 10−4−10−3 μm3 were composed of a single lamella or a few stacked lamellae, thus leading to the formation of PE nanocrystals. Figure 1c shows the AFM amplitude image of the PE nanocrystals while Figure 1d shows the AFM height image magnifying the region R2 in Figure 1c. The arrow in Figure 1c points to a crystal containing a screw dislocation, which represents the morphology of a few stacked lamellae. The Figure 1c inset schematically describes the structure of the PE nanocrystals. The nucleation rate for such small droplets (10−4−10−3 μm3) is expected to be highly dependent on Tc and the droplet size.36,41−43 The nucleation rate of PE confined to ∼10−3 μm3 was predicted to be 1, 10−2, and 10−7 min−1 at Tc = 75, 80, and 90 °C, respectively.44,45 Therefore, self-nucleation of such confined droplets was expected during the time scale of deposition (∼1 h) for Tdep below 80 °C, but at Tdep above 80 °C, only much larger droplets would be able to nucleate during deposition. When Tdep = 116 °C, all identified nucleating droplets with d ∼ 27 nm had a volume larger than ∼1 μm3, as shown in Figure 1a. The above observations provided clues to PE film formation and morphology development by MAPLE. In the early stages of deposition, crystallization of PE proceeds primarily via nucleation of droplets with volumes larger than Vnucl. As the film becomes consolidated, crystal lattice sites at the surface of the film promote the crystallization of subsequently deposited droplets. Meanwhile, Tdep uniquely determines the thickness of the lamellae, as illustrated by Figure 1e, which compares the representative height profiles of MAPLE-PE MLCs formed at four different Tdep: 75, 80, 100, and 116 °C. The average MLC thickness at each Tdep is presented in Figure 1f. It should be

droplets of PE prior to consolidation. The importance of Tdep as a key processing parameter to control morphology, and hence properties, stems from its strong correlation to the nucleation rate of confined droplets and lamellar thickness.36 By characterizing the early stage film morphology, lamellar thicknesses and rough estimates of the critical volume (Vnucl) required for self-nucleation of the confined droplets could be investigated, as previously performed in the study of MAPLEdeposited poly(ethylene oxide).36,37 To enable such a study, we deposited isolated PE droplets atop silicon substrates (Si) by MAPLE from a dilute solution target (see Supporting Information for details of the MAPLE conditions). We first investigated the crystalline morphology formed at Tdep = 116 °C, which represents the typical crystallization temperature of bulk PE when Mn = 3000.40 Figure 1a shows an optical microscopy image of MAPLE-deposited PE (MAPLEPE) droplets formed at Tdep = 116 °C. At this temperature, platelet-like monolamellar crystals (MLCs) were formed from nucleating confined droplets of volume larger than ∼1 μm3. The region marked R1 in Figure 1a indicates a MLC grown from a nucleating droplet, whose structure is provided in the inset of Figure 1a. The AFM height image in Figure 1b magnifies the MLC depicted in R1. This revealed that the MLC had a uniform thickness (lamellar thickness, d) of ∼27 nm, which agreed well with the fully extended chain length of PE when considering 0.127 nm as the increment of length per CH2 group. It also agrees with the observation by Ueda and Register,40 where the same PE, in bulk, formed extended chain crystals when crystallized at a similar temperature. The formation of 27 nm thick MLCs also indicated that the molecular weight of PE was well preserved during MAPLE processing. B

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 3. (a) Plots comparing normalized flash-DSC first heating thermograms of MAPLE-PE formed with various Tdep, ranging from −2 to 116 °C. The dashed curve is the normalized heating thermogram of LN2-quenched 5 mg PE, obtained using a conventional DSC. (b) Plots showing GIWAXS intensities of MAPLE-PE, formed with Tdep from 5 to 116 °C atop Si, as a function of scattering vector, q. All patterns have two peaks at q1 = 1.54 Å−1 and q2 = 1.70 Å−1, which correspond to (110) and (200) reflections of orthorhombic PE, respectively. The plots were normalized to have the same intensity at q1 = 1.54 Å−1, and the individual traces were shifted along the y-axis for clarity.

Figure 4. (a) Normalized DSC cooling and the subsequent heating thermograms of PE. Various cooling rates were used as follows: 0.5 °C/min (blue), 5 °C/min (red), 40 °C/min (green), and 3 × 105 °C/min (= 5000 °C/s or 5k-quenching, black-dotted). Bulk-DSC was used for cooling rates of 0.5−40 °C/min, and flash-DSC was used for 3 × 105 °C/min. (b) Schematic describing the temperature protocol used for isothermal crystallization with flash-DSC. (c) Flash-DSC heating thermograms of PE, measured after isothermal crystallization at various temperatures (Tc,iso) from 90 to 116 °C. For comparison, a heating curve of the same PE crystallized by 5k-quenching is shown as the dotted-blue curve. The inset magnifies the melting regions of three isothermally crystallized samples, formed with Tc,iso = 105, 110, and 116 °C, which each exhibits a bimodal shape. (d) Optical microscopy images comparing the morphology of MAPLE-PE atop a flash-DSC chip before (left panel) and after (right panel) postdeposition melting at 150 °C. Here, the sample was initially prepared by MAPLE at Tdep = −2 °C.

dewetting of the PE film into large droplets occurred (see Figure S2 in Supporting Information). Figure 2a shows an optical microscopy image of a consolidated PE film deposited atop Si at Tdep = 25 °C. Flash-DSC chips have a sample and reference area of 500 μm diameter each. We masked the chip with a polyimide film having a laser-machined hole of size 250 × 250 μm2 to exclusively expose the sample area, as shown in Figure 2b. After masking, the chip was attached to a MAPLE substrate heater for deposition. Figure 2c shows an image of the flash-DSC chip after MAPLE, where the white-dashed square indicates the deposited area. Figure 3a compares the flash-DSC first heating thermograms of the MAPLE-PE films performed directly after deposition, as a function of Tdep. Remarkably, the tuning of Tdep over 118 °C (−2 to 116 °C) resulted in a dramatic change in Tm, over 20.6 °C (105.7−126.3 °C). The lower Tm values obtained by MAPLE are noteworthy, considering that crystallization by

noted that the thicknesses at Tdep = 75, 80, and 100 °C are approximately half the extended-chain crystal thickness, indicating that some of the chains could comprise two independent crystal stems. This crystallization behavior is clearly different from typical melt crystallization, where structure formation is mostly dominated by spherulitic growth throughout the entire film that readily occurs after the onset of nucleation. Tuning Tm of MAPLE-PE Films by Controlling Tdep. To investigate the Tm of MAPLE-PE films formed at various Tdep, we performed calorimetric analysis by directly depositing PE atop flash-calorimetry chips (flash-DSC chips). Here, we performed MAPLE with a longer deposition time and higher deposition rate compared to the conditions used to undertake the above AFM analysis of early stage film growth; these growth conditions formed a consolidated film in the investigated Tdep range, except at Tdep = 116 °C, where C

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

25 °C, and after isothermal annealing for 0.5 s the sample was heated to 150 °C at 300 000 °C/min to measure Tm. Figure 4c compares the flash-DSC heating thermograms of the same PE sample after isothermal crystallization at various Tc,iso. The Tm clearly increased with increasing Tc,iso, as expected.38 Interestingly, the melting peak became bimodal at Tc,iso above 105 °C. We reason that this was due to two different sets of nucleation events: one occurring at Tc,iso and the other occurring during the subsequent 5k-quench step to 25 °C. Annealing of MAPLE-PE films at 150 °C leads to dewetting into numerous isolated liquid droplets of various sizes, as shown in Figure 4d. In such a sample geometry, each droplet must be individually nucleated; therefore, the nucleation probability of smaller PE droplets during the isothermal hold would significantly decrease as Tc,iso is increased.42,43 As a result, only larger droplets could crystallize at higher Tc,iso, while smaller ones would stay liquid at Tc,iso and crystallize during the 5k-quench to 25 °C. The first melting peak of the PE samples with Tc,iso above 105 °C occurs at ∼113−114 °C, which matches the Tm of PE which was 5k-quenched from the melt (see Figure 4c inset). Therefore, for those samples we selected Tm from the second melting peak as corresponding to crystals formed at the target Tc,iso. Figure 5 compares the Tm−Tdep relationship obtained from MAPLE with the Tm−Tc relationship obtained from melt

nonisothermal cooling of bulk PE (∼5 mg), even by quenching into liquid nitrogen (LN2), could only achieve a lower bound of Tm = 126.4 °C, as illustrated by the dashed curve in Figure 3a. The Tm of finite lamellar crystallites, which consist of amorphous layers of thickness da and crystalline layers of thickness dc (d = dc + da),7 can be predicted by the Gibbs− Thomson equation:46 ⎛ 2σa ⎞ Tm = Tm° ⎜1 − ⎟ dcΔHf°ρ ⎠ ⎝

(1)

where σa is the amorphous (chain end or fold) surface energy, T°m and ΔH°f are the equilibrium melting temperature and heat of fusion (energy/mass) of an infinitely thick PE lamella, respectively, and ρ is the density of crystalline PE. From this relationship, the variation in Tm can be attributed to a change in dc. To confirm that the variation of Tm was not due to a different unit cell structure, we performed grazing-incidence wide-angle X-ray scattering (GIWAXS) on MAPLE-PE films deposited atop Si. The samples all exhibited two strong peaks at q1 = 1.54 Å−1 and q2 = 1.70 Å−1, as shown in Figure 3b, which corresponded to the (110) (interplanar spacing of 4.11 Å) and (200) (interplanar spacing of 3.70 Å) reflections of the orthorhombic PE crystal structure, respectively.47 The dc and the corresponding Tm of lamellar crystallites are known to be controlled by Tc, as supported by the same dependence of dc (or Tm) on Tc between melt- and solutioncrystallization.48,49 To investigate the correlation between Tdep and Tc during film formation by MAPLE, we compared the Tdep−Tm relationship with the Tc−Tm relationship obtained by melt crystallization of our PE. We conducted both nonisothermal and isothermal melt-crystallization experiments using traditional DSC (bulk-DSC) and flash-DSC. For nonisothermal crystallization, we used various cooling rates to obtain a range of Tc and the corresponding Tm, which were quantified as the peak temperatures during crystallization and melting, respectively. Figure 4a compares the cooling and heating thermograms of the PE between 25 and 150 °C. BulkDSC was used for cooling rates of 0.5−40 °C/min, and flashDSC was used for 300 000 °C/min (5000 °C/s, “5kquenching”). After cooling from the 150 °C melt at various rates, subsequent heating runs were employed to measure Tm; heating rates of 10 °C/min and 5000 °C/s were used for bulkand flash-DSC, respectively. The slowest cooling rate of 0.5 °C/min resulted in the highest Tc of 118.2 °C and a corresponding Tm of 128.2 °C, at which our PE of Mn = 3000 g/mol is expected to form its most stable crystals, with an extended-chain conformation.39,40 Increasing the cooling rate to 40 °C/min with bulk-DSC merely decreased Tc by ∼5 °C and Tm by ∼1.6 °C. In comparison, 5k-quenching with flashDSC resulted in a broad crystallization peak with the onset Tc below 100 °C, and a Tm around 113.5 °C on subsequent heating, significantly reducing Tc and Tm in comparison to those obtained from bulk-DSC with cooling rates below 40 °C/ min. As 5k-quenching by flash-DSC can dramatically reduce the onset Tc, we next conducted isothermal crystallization above 90 °C to obtain a broader range of Tc−Tm data. Figure 4b describes the temperature protocol used for isothermal crystallization via flash-DSC. A melt of a PE sample at 150 °C was 5k-quenched to a target crystallization temperature (Tc,iso) in the range of 90−116 °C and isothermally crystallized for 10 min. This was followed by another 5k-quench down to

Figure 5. Plots showing the Tm−Tdep relationship of MAPLE-PE (MAPLE, green circles) and the Tm−Tc relationship of meltcrystallized PE. For melt crystallization, two methods were used: isothermal crystallization using flash-DSC (ISO, blue triangles) and nonisothermal crystallization by cooling with various rates using bulkDSC (NIC, red squares). The target crystallization temperature in ISO and the peak temperature of the exotherm in NIC were defined as Tc. Line 1 was obtained from all MAPLE, NIC, and ISO data ranging from Tdep (or Tc) = 75 to 116 °C, while line 2 was obtained from MAPLEPE data ranging from Tdep = −2 to 75 °C.

crystallization. Data obtained by MAPLE crystallization (MAPLE), isothermal crystallization (ISO), and nonisothermal crystallization (NIC) are plotted as green circles, blue triangles, and red squares, respectively. Above 75 °C, all three data sets reasonably fit one straight line (line 1 in Figure 5), indicating that the MAPLE-deposited PE crystallized at the deposition temperature (Tdep = Tc). At Tc = Tdep = 116 °C, however, the Tm values obtained by ISO and MAPLE exhibited a difference of ∼3.4 °C. We believe that this is due to the high Tdep of 116 °Ca temperature sufficiently high that not all the PE droplets nucleate during MAPLE deposition. Uncrystallized droplets would crystallize at Tc < 116 °C during the cooling step after MAPLE, which would bias the DSC melting peak toward the D

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 6. (a) Plots showing the crystallinity of MAPLE-PE as a function of Tdep, determined by two methods: relative crystallinity (Xc′, left y-axis) obtained by flash-DSC analysis (red squares) and crystallinity (Xc,GIWAXS, right y-axis) obtained by the analysis of GIWAXS patterns (blue triangles). Crystallinity of single lamellar crystals calculated from the MLC thickness and da = 2.5 nm (Xc,single, right y-axis) is marked with + signs. (b) Plots comparing X′c of MAPLE-PE (black squares) and melt-crystallized PE as a function of the corresponding Tm. For melt-crystallization, ISO with Tc,iso = 90 and 95 °C (blue triangles) and NIC with cooling rates from 10 to 5000 °C/s (red circles) were used. (c, d) Illustrations showing the structure development of PE during (c) melt-crystallization and (d) MAPLE-crystallization. Note that the MAPLE-crystallized PE has a higher crystallinity due to the smaller content of intercrystallite amorphous phase.

lower temperature region. The linear relationship between Tc and Tm on line 1 agrees with the typical crystallization-melting behavior of metastable polymer crystals formed at a relatively ° − Tc).50−52 low undercooling (ΔT = Tm Interestingly, in the Tdep range below 75 °C, the Tm−Tdep data follow a different line (line 2 in Figure 5), which deviates from line 1. In this temperature range, the Tm−Tc data for melt crystallization were not available due to rapid crystallization prior to reaching the target Tc, even for flash-DSC. On line 2, Tm decreased by ∼1.8 °C for every 20 °C decrease in Tdep, which was about 3.9 times smaller than that the Tm depression rate on line 1. For MAPLE-PE formed at Tdep = −2 °C, the measured Tm was 19.7 °C higher than the Tm extrapolated along line 1. While predicting the exact Tm−Tdep relationship in the deeply supercooled regime was beyond the scope of this study, the following factors may be relevant to the deviation observed below 75 °C. First, the Hoffman−Lauritzen model53,54 could support the observation. The theoretical Tm−Tc relationship based on the Hoffman−Lauritzen model intrinsically has nonlinearity, where the change of Tm with Tc becomes smaller at lower Tc (or larger ΔT).50 Therefore, even if Tc = Tdep, the Tm−Tdep plot would exhibit a decrease in slope at smaller Tdep values. Additionally, from the experimental viewpoint, the rapid crystallization kinetics of PE may have resulted in crystallization prior to reaching Tdep (Tc > Tdep) below Tdep ∼ 75 °C, thus resulting in the deviation of the Tm−Tdep trend from line 1, where Tc is precisely known and controlled. The Tc = Tdep condition should pertain when the temperature of the liquid PE droplets deposited on the substrate equilibrates to Tdep prior to nucleation; that is, the characteristic heat conduction time (τheat) should be much smaller than the characteristic

nucleation induction time (τnucl) of PE. As a decrease of temperature dramatically accelerates PE nucleation, τnucl will become smaller than τheat at a sufficiently low Tdep. PE droplets of 0.1 and 1 μm radius are estimated to nucleate prior to thermal equilibration (τnucl < τheat) at Tdep below ∼52 and ∼63 °C, respectively, roughly where the data begin to deviate from line 1 (see Supporting Information for details of the calculation). In addition, the rapid crystal growth after the onset of nucleation may prevent the latent heat released upon crystallization from diffusing out of the droplet as fast as it is generated, contributing to an increase in of Tc.55 Crystallinity and the Associated Supramolecular Structure of MAPLE-PE. We next sought to analyze the crystallinity of MAPLE-PE and obtain clues to the morphology of the thin films. The crystallinity of PE (Xc) can be determined from DSC melting peaks via the following relationship:

Xc =

ΔHf m ∗ΔHfo

(2)

where ΔHf is the heat of fusion obtained from calorimetry and m is the sample mass of PE. While measuring the absolute mass of MAPLE-PE films is challenging due to its infinitesimal amount, we derived the relative mass of PE deposited atop different flash-DSC chips by assuming that the same crystallization protocol, for instance, 5k-quenching, would lead to the same degree of crystallinity. The relative mass could then be used to derive the relative crystallinity (Xc′) of different PE samples, as elaborated in Figure S3. The Xc′ of 5kquenched PE was normalized to 1. Figure 6a plots X′c for MAPLE-PE films determined by flashDSC (red squares, left y-axis) as a function of Tdep. Xc′ generally decreased as Tdep was lowered. At Tdep = 116 °C, however, the E

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules MAPLE-PE exhibited an abnormally low X′c with a dewetted film morphology (see Figure S2). We assumed that this was due to the slow nucleation of PE at 116 °C. Dewetting of uncrystallized PE would occur during deposition at such a high temperature; therefore, the isolated liquid droplets would crystallize at Tc < 116 °C on cooling after MAPLE. To check the validity of the Tdep−Xc′ trend obtained by flash-DSC analysis, we also derived the crystallinity of MAPLE-PE from GIWAXS data (blue triangles with right y-axis in Figure 6a). From the analysis of wide-angle X-ray scattering patterns, Xc,GIWAXS, of PE could also be determined, as elaborated in the Supporting Information. With the proportionality constant of 0.642 ± 0.020 between Xc′ and Xc,GIWAXS (0.642Xc′ = Xc,GIWAXS), the two crystallinity vs Tdep curves reasonably match, as expected. From the simplified model of semicrystalline polymers, we can assume that two types of structures mainly contribute to the amorphous fraction: (i) the fold/end group surface of lamellar crystallites and (ii) entangled PE chains that are kinetically trapped between lamellae, when the crystals are thin enough for some of the PE chains to form two stems. Because of the amorphous nature of the fold surface, even PE single lamellar crystallites exhibit semicrystallinity, with a crystallinity (Xc,single) approximately equal to dc/d.7,56 Xc (or Xc,GIWAXS) would be the same as Xc,single if the bulk polymer consists exclusively of unconnected lamellar crystallites or smaller than Xc,single if the polymer contains a considerable amount of entanglements kinetically trapped between lamellae. Considering the thickness of PE amorphous layers (da, da = d − dc) is ∼2.5 nm,38,56,57 the Xc,single of our PE formed at Tdep = 116, 100, 80, and 75 °C would be 0.91, 0.84, 0.80, and 0.77, respectively (see + symbols with y-axis on the right in Figure 6a), based on the MLC thickness measurements provided in Figure 1f. The predicted Tdep−Xc,single relationship is in reasonable agreement with the Tdep−Xc relationship, as shown in Figure 6a. At a given Tc or Xc,single, the variation in crystallinity should thus reflect differences in the extent of interlamellar entanglements. To understand how effectively MAPLE processing forms crystalline PE at a given Tc, we compared X′c of PE processed via MAPLE and melt crystallization as a function of Tm in Figure 6b; the black squares, red circles, and blue triangles represent the Xc′ of PE formed by MAPLE, NIC, and ISO, respectively. Cooling rates of 10−5000 °C/s (NIC10−NIC5000) were used for NIC, and Tc,iso of 90 and 95 °C (ISO90C, ISO95C) were used for ISO crystallization. As Tm represents the Xc,single of constituent lamellar crystallites, we propose that the crystallinity−Tm plot provides insight into the difference in the extent of interlamellar entanglements between PE samples crystallized in different ways. Interestingly, the MAPLE-PE samples exhibited a higher crystallinity over the whole range of Tm. We attribute the higher crystallinity of MAPLE-PE samples to MAPLE’s ability to generate disentangled-chain precursors for crystallization. In the case of melt-crystallization, PE chains need to disentangle from the mutually entangled melt (entanglement molecular weight of linear PE ∼ 1200 g/ mol58,59) for crystallization.7 We expect that the rapid nonisothermal quenching would not provide sufficient time for disentanglement of the chains, leading to a higher content of entanglements trapped in interlamellar regions. This is schematically illustrated in Figure 6c; estimates of the characteristic times for disentanglement and crystallization for our low-molecular-weight PE (see Supporting Information)

suggest that this situation should pertain at temperatures below about 100 °C, above the isothermal Tc values employed in Figure 6b, as well as the Tc values during nonisothermal crystallization at cooling rates above 10 °C/s. In comparison, we suspect that crystallization via MAPLE can dramatically reduce the entanglement density in the melt; since the MAPLE target consists of a frozen dilute solution, polymers are deposited as isolated chains or few-chain clusters, allowing PE chains to individually crystallize at the substrate, as depicted in Figure 6d. Hence, MAPLE offers a unique route for the deposition and control of morphology in thin polymer films, in which growth proceeds by the assembly of individual droplets under conditions where crystallization simultaneously occurs.

■

CONCLUSIONS Here, we exploited MAPLE as a state-of-the-art processing technology for the deposition of PE films with control over structure and Tm. We demonstrated the ability to precisely tune Tm over 20 °C. This range was achieved through the unique means of film formation during MAPLE, wherein polymer droplets were deposited and crystallized at a target substrate temperature. As confinement within droplets can suppress the onset of nucleation, deeply supercooled crystallization of PE could be achieved during film formation. Furthermore, MAPLE processing formed films with higher crystallinity compared to their melt-crystallized counterparts with the same Tm, by effectively reducing the density of chain entanglements. As such, this investigation offered new insights into how confinement effects in polymer crystallization could be utilized in film fabrication to finely control the structure of crystalline polymers. We anticipate that this approach will lay the foundation for the exploitation of MAPLE in emerging technologies where semicrystalline polymer films can play a key role.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.7b02345. Figures S1−S3 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected] (R.D.P.). ORCID

Mithun Chowdhury: 0000-0002-2513-6006 Melda Sezen-Edmonds: 0000-0003-0476-6815 Yueh-Lin Loo: 0000-0002-4284-0847 Richard A. Register: 0000-0002-5223-4306 Rodney D. Priestley: 0000-0001-6765-2933 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS H.J. acknowledges support from Kwanjeong Educational Foundation in South Korea. Y.L.L., R.A.R., C.B.A., and R.D.P. acknowledge the support of the National Science Foundation (NSF) Materials Research Science and Engineering Center Program through the Princeton Center for Complex Materials (DMR-1420541). R.D.P. acknowledges the support of the F

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(21) Isasi, J. R.; Haigh, J. A.; Graham, J. T.; Mandelkern, L.; Alamo, R. G. Some Aspects of the Crystallization of Ethylene Copolymers. Polymer 2000, 41, 8813−8823. (22) Ortmann, P.; Mecking, S. Long-Spaced Aliphatic Polyesters. Macromolecules 2013, 46, 7213−7218. (23) Manet, S.; Tibirna, C.; Boivin, J.; Delabroye, C.; Brisson, J. Single Crystals of Chain-Folded Copolyterephthalamides. Macromolecules 2006, 39, 1093−1101. (24) De Ten Hove, C. L. F.; Penelle, J.; Ivanov, D. A.; Jonas, A. M. Encoding Crystal Microstructure and Chain Folding in the Chemical Structure of Synthetic Polymers. Nat. Mater. 2004, 3, 33−37. (25) De Rosa, C.; Auriemma, F.; Di Capua, A.; Resconi, L.; Guidotti, S.; Camurati, I.; Nifant’ev, I. E.; Laishevtsev, I. P. Structure−Property Correlations in Polypropylene from Metallocene Catalysts: Stereodefective, Regioregular Isotactic Polypropylene. J. Am. Chem. Soc. 2004, 126, 17040−17049. (26) Okada, K. N.; Hikosaka, M. Polymer Nucleation. In Handbook of Polymer Crystallization; Piorkowska, E., Rutledge, G. C., Eds.; John Wiley & Sons, Inc.: 2013; pp 125−164. (27) Chrisey, D. B.; Piqué, A.; McGill, R. A.; Horwitz, J. S.; Ringeisen, B. R.; Bubb, D. M.; Wu, P. K. Laser Deposition of Polymer and Biomaterial Films. Chem. Rev. 2003, 103, 553−576. (28) Piqué, A. The Matrix-Assisted Pulsed Laser Evaporation (MAPLE) Process: Origins and Future Directions. Appl. Phys. A: Mater. Sci. Process. 2011, 105, 517−528. (29) Mariano, F.; Caricato, A. P.; Accorsi, G.; Leo, C.; Cesaria, M.; Carallo, S.; Genco, A.; Simeone, D.; Tunno, T.; Martino, M.; et al. White Multi-Layered Polymer Light Emitting Diode through Matrix Assisted Pulsed Laser Evaporation. J. Mater. Chem. C 2016, 4, 7667− 7674. (30) Ge, W.; Yu, Q.; López, G. P.; Stiff-Roberts, A. D. Antimicrobial Oligo(p-Phenylene-Ethynylene) Film Deposited by Resonant Infrared Matrix-Assisted Pulsed Laser Evaporation. Colloids Surf. Colloids Surf., B 2014, 116, 786−792. (31) Ge, W.; Li, N. K.; McCormick, R. D.; Lichtenberg, E.; Yingling, Y. G.; Stiff-Roberts, A. D. Emulsion-Based RIR-MAPLE Deposition of Conjugated Polymers: Primary Solvent Effect and Its Implications on Organic Solar Cell Performance. ACS Appl. Mater. Interfaces 2016, 8, 19494−19506. (32) McCormick, R. D.; Cline, E. D.; Chadha, A. S.; Zhou, W.; StiffRoberts, A. D. Tuning the Refractive Index of Homopolymer Blends by Controlling Nanoscale Domain Size via RIR-MAPLE Deposition. Macromol. Chem. Phys. 2013, 214, 2643−2650. (33) Paun, I. A.; Ion, V.; Moldovan, A.; Dinescu, M. Thin Films of Polymer Blends for Controlled Drug Delivery Deposited by MatrixAssisted Pulsed Laser Evaporation. Appl. Phys. Lett. 2010, 96, 243702. (34) Shepard, K. B.; Priestley, R. D. MAPLE Deposition of Macromolecules. Macromol. Chem. Phys. 2013, 214, 862−872. (35) Caricato, A. P.; Luches, A. Applications of the Matrix-Assisted Pulsed Laser Evaporation Method for the Deposition of Organic, Biological and Nanoparticle Thin Films: A Review. Appl. Phys. A: Mater. Sci. Process. 2011, 105, 565−582. (36) Jeong, H.; Shepard, K. B.; Purdum, G. E.; Guo, Y.; Loo, Y.-L.; Arnold, C. B.; Priestley, R. D. Additive Growth and Crystallization of Polymer Films. Macromolecules 2016, 49, 2860−2867. (37) Jeong, H.; Napolitano, S.; Arnold, C. B.; Priestley, R. D. Irreversible Adsorption Controls Crystallization in Vapor-Deposited Polymer Thin Films. J. Phys. Chem. Lett. 2017, 8, 229−234. (38) Stack, G. M.; Mandelkern, L.; Voigt-Martin, I. G. Crystallization, Melting, and Morphology of Low Molecular Weight Polyethylene Fractions. Macromolecules 1984, 17, 321−331. (39) Lee, K. S.; Wegner, G. Linear and Cyclic Alkanes (CnH2n+2, CnH2n) with n > 100. Synthesis and Evidence for Chain-Folding. Makromol. Chem., Rapid Commun. 1985, 6, 203−208. (40) Ueda, M.; Register, R. A. Crystallization-Induced Phase Separation in Mixtures of Model Linear and Short-Chain Branched Polyethylenes. J. Macromol. Sci., Part B: Phys. 1996, 35, 23−36. (41) Massa, M. V.; Carvalho, J. L.; Dalnoki-Veress, K. Direct Visualisation of Homogeneous and Heterogeneous Crystallisation in

AFOSR through a PECASE Award (FA9550-15-1-0017). Portion of this work was conducted at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation and the National Institutes of Health/National Institute of General Medical Sciences under NSF Award DMR-1332208.

■

REFERENCES

(1) Liu, F.; Gu, Y.; Jung, J. W.; Jo, W. H.; Russell, T. P. On the Morphology of Polymer-Based Photovoltaics. J. Polym. Sci., Part B: Polym. Phys. 2012, 50, 1018−1044. (2) Lee, S. S.; Loo, Y.-L. Structural Complexities in the Active Layers of Organic Electronics. Annu. Rev. Chem. Biomol. Eng. 2010, 1, 59−78. (3) Hu, Z.; Tian, M.; Nysten, B.; Jonas, A. M. Regular Arrays of Highly Ordered Ferroelectric Polymer Nanostructures for NonVolatile Low-Voltage Memories. Nat. Mater. 2009, 8, 62−67. (4) Tsao, H. N.; Müllen, K. Improving Polymer Transistor Performance via Morphology Control. Chem. Soc. Rev. 2010, 39, 2372−2386. (5) Ling, Q.-D.; Liaw, D.-J.; Zhu, C.; Chan, D. S.-H.; Kang, E.-T.; Neoh, K.-G. Polymer Electronic Memories: Materials, Devices and Mechanisms. Prog. Polym. Sci. 2008, 33, 917−978. (6) Fujitsuka, N.; Sakata, J.; Miyachi, Y.; Mizuno, K.; Ohtsuka, K.; Taga, Y.; Tabata, O. Monolithic Pyroelectric Infrared Image Sensor Using PVDF Thin Film. Sens. Actuators, A 1998, 66, 237−243. (7) Strobl, G. The Physics of Polymers, 2nd ed.; Springer-Verlag: Berlin, 1997. (8) Reiter, G. Some Unique Features of Polymer Crystallisation. Chem. Soc. Rev. 2014, 43, 2055−2065. (9) Liu, J.; Arif, M.; Zou, J.; Khondaker, S. I.; Zhai, L. Controlling Poly(3-Hexylthiophene) Crystal Dimension: Nanowhiskers and Nanoribbons. Macromolecules 2009, 42, 9390−9393. (10) Brinkmann, M. Structure and Morphology Control in Thin Films of Regioregular Poly(3-Hexylthiophene). J. Polym. Sci., Part B: Polym. Phys. 2011, 49, 1218−1233. (11) Young, R. J. A. Dislocation Model for Yield in Polyethylene. Philos. Mag. 1974, 30, 85−94. (12) Galeski, A. Strength and Toughness of Crystalline Polymer Systems. Prog. Polym. Sci. 2003, 28, 1643−1699. (13) De Rosa, C.; Auriemma, F. Structure and Physical Properties of Syndiotactic Polypropylene: A Highly Crystalline Thermoplastic Elastomer. Prog. Polym. Sci. 2006, 31, 145−237. (14) Harris, A. M.; Lee, E. C. Improving Mechanical Performance of Injection Molded PLA by Controlling Crystallinity. J. Appl. Polym. Sci. 2008, 107, 2246−2255. (15) Baghgar, M.; Labastide, J. A.; Bokel, F.; Hayward, R. C.; Barnes, M. D. Effect of Polymer Chain Folding on the Transition from H- to JAggregate Behavior in P3HT Nanofibers. J. Phys. Chem. C 2014, 118, 2229−2235. (16) Singh, K. A.; Sauvé, G.; Zhang, R.; Kowalewski, T.; McCullough, R. D.; Porter, L. M. Dependence of Field-Effect Mobility and Contact Resistance on Nanostructure in Regioregular Poly(3-Hexylthiophene) Thin Film Transistors. Appl. Phys. Lett. 2008, 92, 263303. (17) Zhang, R.; Li, B.; Iovu, M. C.; Jeffries-EL, M.; Sauvé, G.; Cooper, J.; Jia, S.; Tristram-Nagle, S.; Smilgies, D. M.; Lambeth, D. N.; et al. Nanostructure Dependence of Field-Effect Mobility in Regioregular Poly(3-Hexylthiophene) Thin Film Field Effect Transistors. J. Am. Chem. Soc. 2006, 128, 3480−3481. (18) Bigg, D. M. Polylactide Copolymers: Effect of Copolymer Ratio and End Capping on Their Properties. Adv. Polym. Technol. 2005, 24, 69−82. (19) Peacock, A. J.; Mandelkern, L. The Mechanical Properties of Random Copolymers of Ethylene: Force-Elongation Relations. J. Polym. Sci., Part B: Polym. Phys. 1990, 28, 1917−1941. (20) Simanke, A. G.; Galland, G. B.; Freitas, L.; da Jornada, J. A. H.; Quijada, R.; Mauler, R. S. Influence of the Comonomer Content on the Thermal and Dynamic Mechanical Properties of Metallocene Ethylene/1-Octene Copolymers. Polymer 1999, 40, 5489−5495. G

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules an Ensemble of Confined Domains of Poly(Ethylene Oxide). Eur. Phys. J. E: Soft Matter Biol. Phys. 2003, 12, 111−117. (42) Massa, M.; Dalnoki-Veress, K. Homogeneous Crystallization of Poly(Ethylene Oxide) Confined to Droplets: The Dependence of the Crystal Nucleation Rate on Length Scale and Temperature. Phys. Rev. Lett. 2004, 92, 255509. (43) Carvalho, J. L.; Dalnoki-Veress, K. Surface Nucleation in the Crystallisation of Polyethylene Droplets. Eur. Phys. J. E: Soft Matter Biol. Phys. 2011, 34, 1−6. (44) Koutsky, J. A.; Walton, A. G.; Baer, E. Nucleation of Polymer Droplets. J. Appl. Phys. 1967, 38, 1832−1839. (45) Loo, Y.-L.; Register, R. A. Crystallization Within Block Copolymer Mesophases. In Developments in Block Copolymer Science and Technology; Hamley, I. W., Ed.; John Wiley & Sons, Ltd.: 2004; pp 213−243. (46) Wang, Y.; Ge, S.; Rafailovich, M.; Sokolov, J.; Zou, Y.; Ade, H.; Lüning, J.; Lustiger, A.; Maron, G. Crystallization in the Thin and Ultrathin Films of Poly(Ethylene−vinyl Acetate) and Linear LowDensity Polyethylene. Macromolecules 2004, 37, 3319−3327. (47) Swan, P. R. Polyethylene Unit Cell Variations with Temperature. J. Polym. Sci. 1962, 56, 403−407. (48) Barham, P. J.; Chivers, R. A.; Keller, A.; Martinez-Salazar, J.; Organ, S. J. The Supercooling Dependence of the Initial Fold Length of Polyethylene Crystallized from the Melt: Unification of Melt and Solution Crystallization. J. Mater. Sci. 1985, 20, 1625−1630. (49) Lin, L.; Argon, A. S. Structure and Plastic Deformation of Polyethylene. J. Mater. Sci. 1994, 29, 294−323. (50) Marand, H.; Xu, J.; Srinivas, S. Determination of the Equilibrium Melting Temperature of Polymer Crystals: Linear and Nonlinear Hoffman−Weeks Extrapolations. Macromolecules 1998, 31, 8219− 8229. (51) Hoffman, J. D.; Weeks, J. J. Melting Process and the Equilibrium Melting Temperature of Polychlorotrifluoroethylene. J. Res. Natl. Bur. Stand., Sect. A 1962, 66A, 13−28. (52) Lee, L.-B. W.; Register, R. A. Hydrogenated Ring-Opened Polynorbornene: A Highly Crystalline Atactic Polymer. Macromolecules 2005, 38, 1216−1222. (53) Hoffman, J. D.; Davis, G. T., Jr.; Lauritzen, J. I. The Rate of Crystallization of Linear Polymers with Chain Folding. In Treatise on Solid State Chemistry; Hannay, N. B., Ed.; Springer: New York, 1976; pp 497−614. (54) Hoffman, J. D.; Miller, R. L. Kinetic of Crystallization from the Melt and Chain Folding in Polyethylene Fractions Revisited: Theory and Experiment. Polymer 1997, 38, 3151−3212. (55) Barham, P. J.; Jarvis, D. A.; Keller, A. A New Look at the Crystallization of Polyethylene. III. Crystallization from the Melt at High Supercoolings. J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 1733− 1748. (56) Osichow, A.; Rabe, C.; Vogtt, K.; Narayanan, T.; Harnau, L.; Drechsler, M.; Ballauff, M.; Mecking, S. Ideal Polyethylene Nanocrystals. J. Am. Chem. Soc. 2013, 135, 11645−11650. (57) Weber, C. H. M.; Chiche, A.; Krausch, G.; Rosenfeldt, S.; Ballauff, M.; Harnau, L.; Göttker-Schnetmann, I.; Tong, Q.; Mecking, S. Single Lamella Nanoparticles of Polyethylene. Nano Lett. 2007, 7, 2024−2029. (58) Fetters, L. J.; Lohse, D. J.; Milner, S. T.; Graessley, W. W. Packing Length Influence in Linear Polymer Melts on the Entanglement, Critical, and Reptation Molecular Weights. Macromolecules 1999, 32, 6847−6851. (59) Ries, M. E.; Brereton, M. G.; Ward, I. M.; Cail, J. I.; Stepto, R. F. T. Rescaling Approach to Molecular Orientation for NMR and Optical Properties of Polymer Networks. Macromolecules 2002, 35, 5665− 5669.

H

DOI: 10.1021/acs.macromol.7b02345 Macromolecules XXXX, XXX, XXX−XXX