A Novel Method to Measure the Effective Change ... - ACS Publications

Jul 23, 2019 - A Novel Method to Measure the Effective Change of the Interfacial Energy due to Kinetic Self-Assembly of Amyloid Fibrils ...
1 downloads 0 Views 958KB Size
Subscriber access provided by BUFFALO STATE

Article

A Novel Method to Measure the Effective Change of the Interfacial Energy Due to Kinetic Self-Assembly of Amyloid Fibrils Yi-Chih Lin, Murray Skolnick, and Zahra Fakhraai J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.9b04717 • Publication Date (Web): 23 Jul 2019 Downloaded from pubs.acs.org on July 23, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

A Novel Method to Measure the Effective Change of the Interfacial Energy due to Kinetic Self-Assembly of Amyloid Fibrils Yi-Chih Lin, Murray Skolnick, and Zahra Fakhraai* Department of Chemistry, University of Pennsylvania, 231 South 34th Street, Philadelphia, Pennsylvania 19104-6323, United States *Corresponding Author: Zahra Fakhraai Email: [email protected]

Abstract Adsorbates growing a self-assembled layer on a solid-liquid interface can significantly change the effective interfacial energy at the solid surface. However, measuring the changes in the effective surface energy while these adsorbates accumulate is challenging, as static contact angle measurements can be affected by the motion and accumulation of these adsorbates at the droplet’s boundary (coffee stain effects). In this report, we utilize a novel method that takes advantage of spin-induced dewetting to measure the change in the effective surface energy as the self-assembly progresses. We use a previously well-studied model system of self-assembled fibrils of amyloid- (A) peptides on the mica substrate to demonstrate the feasibility of this method. Using variations of terminal spin speeds and acceleration rates, we measure the terminal spin speed at which a wetting-dewetting transition (WDT) occurs on a surface that hosts self-assembled A12-28 fibrils. By comparing this speed with the WDT speed on the bare mica substrate, we can quantify the spreading coefficient and thus the effective change of the substrate’s interfacial energy due to the adsorption of mobile peptides at various stages of the self-assembly. These measurements show that the surface becomes more hydrophilic as the self-assembly progresses and thus can explain previous observations that the self-assembly of this particular peptide system is self-limiting and stops before full surface coverage.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 16

Introduction The conventional measurement of free energy at the solid-liquid interface is the sessile drop technique. In this technique, the solid-liquid contact angle is measured after a drop of the liquid is placed on the solid surface and used to calculate the surface energy using Young’s equation.1 However, this method can fail in systems containing mobile adsorbates, such as self-assembled lipids or peptides that can diffuse on the surface at time scales shorter than the measurement time. This is because the surface energy varies with time both because the number of the adsorbed particles changes with time and because as the self-assembly progresses the nature of these charges on the surface can vary due to the orientation or aggregation of the peptides or other adsorbates. The placement of the droplet can also induce concentration gradients near the contact line, known as coffee-ring effects2, dramatically changing the surface energy of the contact line compared to the effective surface energy of the surface. As such, while dynamic contact angle measurements are possible in principle, because of the coffee ring effects, these measurements will not be representative of the effective surface energy that is experienced by the incoming adsorbates. Nonetheless, measuring the time-varying interfacial energy in these dynamical systems are critical in understanding the kinetics of the adsorption/desorption as well as the kinetics of the self-assembly process itself. For example, whether a process is self-limiting or cooperative, depends on how the interfacial energy changes as the deposition of particles progresses and whether the changes in the effective interfacial energy is favorable for further deposition. In this report, we demonstrate how spindrying can be utilized to indirectly measure the changes of the effective surface energy of a surface due to kinetic self-assembly of A12-28 peptides. Spin-casting is a widespread method for preparing uniform films ranging from a few nanometers to micrometers in thickness. This process involves multiple stages; dispensing a solution, ramping-up the spin speed to a set final speed and spinning until the full evaporation of the solvent. Spin-casting can be generally applied to a wide range of materials used in a variety of industries, such as photoresists, organic semiconductors, polymers, and nanoparticles.3–5 The final film thickness and its properties depend on the material properties such as the concentration of the solution, viscosity, solvent evaporation rate, and substrate interactions, as well as the parameters chosen for the spin-casting process, such as the acceleration rate, terminal spin speed, and spinning time.6,7 For thin film preparation, in order to produce a smooth film, the condition for the solution and spinning speed should be such that the material can be spread evenly over the substrate, which means that the terminal spin speed needs to be below a critical speed to avoid film rupture and dewetting. In most applications, dewetting is undesirable as it can lead to surface patterns that may interfere with the device performance or affect the smoothness needed for various applications. However, dewetting can also be beneficial as a method for surface patterning and fast drying.6,8–11 For example, Hameren et al.10 demonstrated that highly periodic patterns can be achieved using the self-assembled properties of porphyrin trimers combined with the physical dewetting phenomena. In our previous reports12,13, we demonstrated that a slow spinning process can be used to rapidly dry amyloid fibrils produced by surfacemediated fibrillization without perturbing their structure. This suggested that the spinning process was able to remove the excess peptides in the solution without dewetting the droplet on the surface. At high

ACS Paragon Plus Environment

Page 3 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

spin speeds, above the wetting-dewetting transition (WDT), dewetting patterns composed of nano-sized droplets were observed on the same surfaces, due to the deposition of peptides in the solution that were trapped within small ruptured droplets.13 In both cases, the drying time was quicker than the time scale for the peptide self-assembly, which is of the order of minutes or hours. For systems containing mobile surface species, the terminal spin speed for the WDT will vary as the self-assembly progresses as a result of changes in the effective surface energy. As such, the critical terminal spin speed for WDT can be used to measure the effective surface energy. In this report, we systematically study the effect of terminal spin speed and acceleration rate on the observed morphologies of films produced by spin-casting of a low-concentration solution of Aβ12-28 peptides on mica surface. Atomic force microscopy (AFM) is used to image the morphology of the spun-cast samples. At the WDT, the morphology of the surface dramatically changes allowing us to determine the critical spin speed for the WDT transition through AFM imaging. By comparing the critical spin speed on the bare mica surface and a surface that has surface-mediated self-assembled fibrils of the same Aβ12-28 peptides, we estimate the changes in the effective surface energy due to the formation of these self-assembled fibrils. This facile method can be easily extended to other liquid phase surface assembled systems, such as lipid-bilayers and self-assembled monolayers, and as such, it provides a powerful and easily accessible method for such measurements.

Experimental Methods Sample Solution. Synthetic A12-28 peptide (purity > 95%, rPeptide) was directly used to prepare sample solutions without further purification. A12-28 peptide was first dissolved in 1 % NH4OH(aq) at a concentration of 1 mg/ml. The solution was sonicated for 2 minutes, and subsequently diluted with Milli-Q water to a concentration of 10 M. To remove large aggregates, the solution was filtered using a 0.22 m PTFE membrane syringe filter. The prepared peptide solutions were immediately used for further experiments. Spin-casting Procedure. 0.2 mL of peptide solutions were placed on freshly cleaved mica and were then either spun-cast immediately or incubated for 30 minutes prior to the spin-casting process. A spin-coater (WS-650 MZ-23, Laurell Technologies Corp.) was used to dry the samples under various spinning rates in the range of 1000 to 3000 rpm. The acceleration rate was set either as 250 rpm/s or 400 rpm/s. The duration of spin time was one minute for all samples. Atomic Force Microscopy (AFM). The spun-dried samples were imaged in tapping mode using a Keysight 5500 AFM instrument (Keysight Technologies) equipped with a closed-loop scanner. Rotated silicon probes (BudgetSensors, Tap-300G, resonance frequency ~300 kHz, tip radius ~10 nm) were used to record topography, amplitude, and phase images with 512 x 512 pixel resolution. The images were obtained from the central region of the substrate to avoid the potential gradient effects at the edges of the sample12, where the drying rate could be slower. The AFM images were processed and analyzed by the Gwyddion package.14 A third-order polynomial was used to flatten the background for topography images.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 16

For each image, the zero-height value is chosen to be the average height of the mica substrate (background) for ease of comparison between various figures.

Results Principles of the Wetting-Dewetting Transition. We first describe the underlying principle for the method developed in this study. Under static conditions (without spinning), wetting is defined as when a droplet placed on a surface is stable with respect to rupture and spreads over the surface (Figure 1, top left panel), while dewetting is defined as a metastable state where the droplet can spontaneously rupture due to fluctuations (Figure 1, bottom left panel).15 The spreading coefficient (𝑆) defines the boundary between the wetting and dewetting regimes. S is defined as the difference between the surface tensions of the dry and wet substrates, as shown in Equation 1.16,17 𝑆 = 𝛾𝑆𝑂 ―(𝛾𝐿𝑉 + 𝛾𝑆𝐿)

(1)

Where 𝛾𝑆𝑂 is the substrate-air surface tension, 𝛾𝐿𝑉 is the liquid-air surface energy, and 𝛾𝑆𝐿 is the substrateliquid surface energy respectively. We note that since the measurements reported in this study were performed under ambient conditions, 𝛾𝑆𝑂 is different from the solid-vapor surface tension, 𝛾𝑆𝑉 (more details in Supporting Information, SI). A positive spreading coefficient (𝑆 > 0) means that the thin film wets the substrate, and a negative value (𝑆 < 0) means that the film will spontaneously rupture and dewet. Thus, the spontaneous WDT at 𝑆 = 0 can be used to characterize the surface energy and the interfacial energy, for example, of a polymer thin film.18,19 For the bare mica substrate used in this study, measurements of the contact angle shows that the spreading coefficient of water on mica is 64±11 mN ∙ m-1, which satisfies wetting conditions (more details in SI, Figure S1). Both the spin casting and the formation of self-assembled fibrils can change the spreading coefficient of the mica surface, as schematically shown in Figure 1 (middle and right panels). The combined spreading coefficient due to all these factors can be written as 𝑆𝑠𝑝𝑖𝑛 = 𝛾𝑆𝑂 ― (𝛾𝐿𝑉 + 𝛾𝑆𝐿) ― 𝛾𝐶 + 𝛾𝐹 (2) where 𝛾𝐶 is the interfacial tension induced by the centrifugal force and 𝛾𝐹 is the change in the surface energy (𝛾𝑆𝐿) due to the formation of self-assembled fibrils.

ACS Paragon Plus Environment

Page 5 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1. Schematic diagrams for the wetting and dewetting states for various conditions. The spinning (orange arrow) and the self-assembled fibrils (green cylinders and arrow) each add an effective contribution to the spreading coefficient shifting the WDT accordingly. The direction of 𝛾𝐹 is chosen such that a positive value means a more hydrophilic surface upon self-assembly has formed.

To calculate 𝛾𝐶, we assume that the droplet has not gone through WDT before reaching the terminal spin speed. As such, at the terminal spin speed, the droplet has the shape of a disc spinning at a constant angular velocity. For a disc with mass 𝑚 and radius 𝑅, which is spinning with an angular velocity 1

of 𝜔 , the moment of inertia is given as 𝐼 = 2𝑚𝑅2 and the energy due to centrifugal forces is given by 𝐸𝑐 1

= 2𝐼𝜔2, the interfacial tension associated with this energy (𝛾𝐶) is thus given by 𝐸𝑐

1

𝛾𝐶 = 𝜋𝑅2 = 4𝜋𝑚𝜔2

(3)

Unlike 𝛾𝑆𝑂, 𝛾𝐶 acts to disperse the liquid layer in the radial direction and thus adds a negative contribution to 𝑆. We note that the greater the acceleration rate, the lower the mass remaining on the surface when the speed reaches the constant terminal value. As such, 𝛾𝐶 is lower at a higher acceleration rate. Thus the WDT occurs at a higher terminal spin speed for a higher acceleration rate, to maintain the same value of 𝛾𝐶 at WDT. The sign of 𝛾𝐹, the change in the interfacial tension due to the adsorbed surface species, depends on the hydrophobicity of these species. We have chosen a convention here where positive 𝛾𝐹 represents a more hydrophilic surface, and thus a higher S value.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 16

Morphological Characterization of Spun-Cast Samples. Figure 2 shows representative AFM images of the mica surface after A12-28 peptide solution (10 M) was spun-dried on bare mica using various terminal spin speeds ( 1000 rpm ≤ 𝑣 ≤ 3000 rpm) and acceleration rates (top row: 𝑎 = 250 rpm/s, bottom row: 𝑎 = 400 rpm/s). We have previously shown that if a solution of A12-28 as prepared here is left on the mica surface, peptides are gradually absorbed on the surface and diffuse on the surface to form self-assembled amyloid fibrils within a few minutes.12,13 To prevent the formation of surface-mediated fibrils for samples shown in Figure 2, the spin-casting process was performed immediately after dispensing the solution on the bare mica substrate. More AFM images of different areas of each surface are provided in the SI (Figure S2 and S3). As shown in the top row of Figure 2, the peptides spun-dried on the bare mica surface using a slow acceleration rate (250 rpm/s) exhibit many nano-sized, globular structures that consist of peptide aggregates. These patterns suggest that the film of the solution containing the peptides ruptured during spinning forming many small droplets that then dried on the surface leaving behind their content.13 When the acceleration rate is increased to 400 rpm/s, as shown in the bottom row of Figure 2, the surface is smooth at low terminal spin speed (𝑣 < 2000 rpm, Figure 2F and 2G) with no sign of residual globular peptide aggregates. The lack of peptide deposits is likely because the liquid layer was fully spread during spinning and the excess peptide in the solution was flung off the sides, thus leaving no materials behind to be deposited on the surface as the liquid layer evaporated. This demonstrates that the terminal spin speed for these systems was below the WDT (𝑆 > 0). As the terminal spin speed is increased, the patterns on the surface change to semi-connected polygonal patterns of aggregated peptides (𝑣 = 2000 rpm, Figure 2H), and eventually to globular structures (𝑣 > 2000 rpm, Figure 2I and 2J) that are randomly dispersed on the surface. In this regime, the dewetting patterns observed at faster terminal spin speeds (𝑣 ≥ 2000 rpm) either as polygonal networks or globular structures20 are due to the rupture of the liquid layer at some point during the spinning (dewetting, 𝑆 < 0), which resulted in the formation of smaller droplets on the surface. We note that the polygonal networks tend to occur close to the WDT threshold where Rayleigh instability causes spontaneous rupture everywhere on the surface20. This morphological transition highlights the transition from wetting, where the film does not rupture and thus doesn’t leave residual droplets, to dewetting at a critical terminal spin speed, where the ruptured droplets form uniform polygonal patterns (𝑣~2000 rpm), and finally to a fully dewetted state at higher terminal spin speeds.

ACS Paragon Plus Environment

Page 7 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 2. Representative AFM images of the samples spun-cast on bare mica surface. The acceleration rate was set as 250 rpm/s for (A-E) and 400 rpm/s for (F-J), while the terminal speed was varied. The terminal spin speed for each panel is labeled on the top. D and W in each image indicate dewetting and wetting patterns, respectively. The scale bar is 2 m for all images. The color bar denotes the common height scale for all images.

As demonstrated in our previous reports12,13, the low-concentration A12-28 peptide solution used here can rapidly self-assemble into fibrils on the mica substrate during the incubation period. To investigate the effects of surface-mediated fibrillization on the WDT and thus the change in the surface energy induced by this self-assembly, solutions of A12-28 peptide were incubated for 30-minutes on the mica substrate prior to spinning under the same conditions applied to the bare mica surface. We have previously shown that after this incubation time, the surface concentration of peptides is ~103-104 μm-2 and ~2-10% of the surface is covered with A12-28 peptides.12 Figure 3 shows representative AFM images of these samples spun-dried using various terminal spin speeds and accelerations. As seen in this figure, the terminal spin speed at which the WDT occurs is shifted to higher speeds and is observed even at 250 rpm/s acceleration. The WDT is measured to be at ~1500 rpm with 250 rpm/s acceleration and ~2500 rpm with 400 rpm/s acceleration. The light blue arrows in these images highlight the self-assembled fibrils (~0.5 nm in thickness) formed on the surface during the incubation and the self-assembly phase. These fibrils can be observed in the background of the dewetted aggregate patterns produced during the spin-drying by the peptides left in the solution. Larger AFM images are also provided in SI (Figure S6). It is important to note that the selfassembled fibrils are much more easily identified on the surface in the wetting conditions (Figures 3A, 3F-3H), where there are no excess residual aggregates on the surface. Another interesting observation is

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 16

that, as indicated by the light blue arrows in Figure 3B-3E (the dewetted regime), the rupture of the liquid layer and the subsequent formation or drying of smaller droplets appear to draw the well-separated fibrils together to form flat aggregates. This is likely due to effects similar to the coffee ring effects where the mobile fibrils accumulate near the edges of the droplets, particular in Figures 3B and 3C (larger versions of both panels are shown in Figure S6). In Figures 3I and 3J, the fibrils are again well dispersed in the background of the dewetting patterns, as the faster drying speeds and acceleration presumably prevents the motion of the fibrils in the time scale of drying. As highlighted in our previous publications, wetting conditions are required to observe and study the self-assembly of these fibrils without drying artifacts.12,13

Figure 3. Representative AFM images of samples after 30-min incubation of the peptide solution on mica prior to the spin-casting process. The acceleration rate was set as 250 rpm/s for (A-E) and 400 rpm/s for (F-J). The terminal spin speed for each panel is labeled on the top. D and W in each image indicate dewetting and wetting patterns, respectively. The scale bar is 2 m in all images. For the ease of comparison, the height scales are set within the range of -0.5 and 1.5 nm (top right) for (B-E) and -0.3 and 0.5 nm (bottom right) for (A, F-J), respectively. The light blue arrows highlight the surface-mediated fibrils observed in the background of the dewetting patterns. Larger versions of panels A-D, G, and I are shown in Figure S6.

ACS Paragon Plus Environment

Page 9 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Discussions To quantify the change in the effective surface energy due to self-assembled fibrils, we can focus on various conditions under which the WDT is observed in the data presented in Figures 2 and 3. The three main parameters that are controlled here are (1) the acceleration rate, which controls 𝛾𝐶 through the mass of the liquid, (2) the terminal spin speed, which also affects 𝛾𝐶, and (3) whether or not the surface has self-assembled fibrils before spinning, which indicates whether 𝛾𝐹 is zero or non-zero. Figure 4 shows examples of three cases where WDT is observed at two different accelerations and with or without fibrils.

Figure 4. (A) WDT on bare mica surface at the acceleration of 400 rpm/s. The two control variables for this transition are the mass at this acceleration (𝑚𝐴400), and 𝛾𝐶. (B) WDT transition at the acceleration 400 rpm/s on a surface with self-assembled fibrils. The control parameters are 𝑚𝐴400, 𝛾𝐶 , and 𝛾𝐹. (C) WDT transition on the same self-assembled fibrils, but at a slower acceleration. The control parameters are 𝑚𝐴250, 𝛾𝐶 , and 𝛾𝐹.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 16

The WDTs shown in Figure 4 indicate the existence of a critical point where all interfacial tensions are in balance (𝑆𝑠𝑝𝑖𝑛 = 0). To quantify the variables at play for the transitions shown in Figure 4, we make a few assumptions in order to simplify the calculations. We assume that (i) the critical terminal spin speed where 𝑆𝑠𝑝𝑖𝑛 = 0 occurs, is roughly the average value of the two speeds before and after WDT (e.g. 1750, 2250 and 1250 rpm for Figure 4A-C, respectively), (ii) the transition occurs after the final speed is achieved and thus the role of the acceleration speed is only to change the mass at the terminal speed (as described in equation 3), (iii) the mass of the solution is the same at the same acceleration rate for all conditions, but different when the acceleration is varied (𝑚𝐴250 and 𝑚𝐴400, respectively), and (iv) 𝛾𝐹 is the same for all incubated samples after the same incubation time, which results in the same average number of adsorbed peptides. We note that each of these assumptions can be independently verified or improved if better accuracy is desired. For example, measurements at smaller increments of terminal spin speed can improve the estimated speed at the transition. Measurements at various accelerations can verify if equation 3 holds, thus verifying that the transition occurs after the terminal spin speed is achieved. This assumption may fail at low acceleration rates, which will likely also put the WDT out of the window of the experiments (similar to the first row of Figure 2). In that case, the acceleration speed can be increased to prevent the rupture before the terminal velocity is reached. Finally, measurements of the number of adsorbed peptides versus time as reported in our previous studies21, or measurements of the 𝛾𝐹 at various incubation times can improve the estimation of 𝛾𝐹. However, these improvements are beyond the goal of this paper and will be explored in our future publications. Using contact angle measurements, in the absence of spin (equation 1), we have estimated the spreading coefficient of water on mica substrate to be 𝑆𝑀𝑖𝑐𝑎 = 64±11 mN ∙ m-1 (details in SI). For the bare mica substrate, we only observe WDT for 400 rpm/s, at a speed of ~1750 rpm. Using a combination of 𝜋

equations 2 and 3, and by assuming 𝑆𝑠𝑝𝑖𝑛 = 0 at 1750 rpm (0 = 64 ― 𝑚𝐴400 × 17502 × (3.6)), we have 𝑚𝐴400 = (2.4 ± 0.7) × 10 ―5 kg (details of the unit conversions can be found in SI). Data in Figures 4B and 4C can then be used to solve for 𝑆𝑆𝑝𝑖𝑛 = 0 for the case where 𝛾𝐹 ≠ 0, at the two accelerations, 0 = 64 ― 𝑚𝐴400 × 22502 ×

( )

(𝟒)

0 = 64 ― 𝑚𝐴250 × 12502 ×

( )

(𝟓)

𝜋 + 𝛾𝐹 3.6 𝜋 + 𝛾𝐹 3.6

which leads to 𝑚𝐴250 = (7.8 ± 3.3) × 10 ―5 kg and 𝛾𝐹 = 42±6 mN ∙ m-1 , respectively. The error in the values of mass were estimated based on the fact that we are assuming the WDT is occurring at the velocity in the middle of the two speeds. By choosing smaller increments a better estimation of error is possible. The error in 𝛾𝐹 was estimated based on the error in 𝛾𝑀𝐼𝐶𝐴. More details of these calculations are provided in the SI.

ACS Paragon Plus Environment

Page 11 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table 1. Predicted 𝑺𝒔𝒑𝒊𝒏 for samples spun-cast under different acceleration rates and terminal spin speeds. Terminal Spin Speed (rpm)

Images

Acceleration rate (rpm/s)

1000

1500

2000

2500

3000

1A - 1E

250

-3.2 (D)

-87.2 (D)

-204.8 (D)

-356.0(D)

-540.7 (D)

1F - 1J

400

43.0 (W)

16.9 (W)

-19.8 (D)

-66.9 (D)

-124.5 (D)

2A - 2E

250

38.8 (W)

-45.2 (D)

-162.8 (D)

-314.0 (D)

-498.7 (D)

2F - 2J

400

85.0 (W)

58.9 (W)

22.2 (W)

-24.9 (D)

-82.5 (D)

The 𝑆𝑠𝑝𝑖𝑛 (mN ∙ m-1) for each image in Figures 1 and 2 calculated using equations 2 and 3, as detailed in the main text. The red and blue colors indicate the predicted dewetting (𝑆𝑠𝑝𝑖𝑛 < 0) and wetting behavior (𝑆𝑠𝑝𝑖𝑛 > 0), respectively. The letter listed in the parentheses are wetting (W) or dewetting (D) conditions evaluated based on the surface patterns observed in the corresponding AFM image.

To validate the values of 𝑚𝐴250, 𝑚𝐴400 and 𝛾𝐹, we use these parameters and the experimental terminal spin speed values to evaluate 𝑆𝑠𝑝𝑖𝑛 for all of the images presented above. The results are summarized in Table 1. The calculated 𝑆𝑠𝑝𝑖𝑛 fulfills the wetting or dewetting behavior observed in the corresponding AFM images shown in Figures 1 and 2. Moreover, all the possible combinations of the data that can result in a WDT are listed in Table S1 (in SI) and are also consistent with the estimated values for the mass of the droplet and the value of 𝛾𝐹. For example, when the speed is held constant and the acceleration is varied (Classification 2 and 4 in Table S1 of the SI), wetting conditions can be observed at high acceleration, while dewetting occurs for the low acceleration. This is due to a roughly three-fold 𝑚𝐴250

difference in the magnitude of 𝛾𝐶 influenced by the mass at the terminal spin speed in these samples (𝑚𝐴400 ~3.2), which shifts WDT to lower values of spin speed at lower acceleration rates. Thus, the WDT it out of the window of observation on bare mica at 250 rpm/s (Figure 1, top row) and shifts from ~2250 rpm to ~1250 rpm on the surface with self-assembled fibrils. In another example, for the same values of the spin speed and acceleration, the existence of selfassembled fibrils can result in wetting conditions compared to the situations where dewetting would have been observed otherwise (Figure 2A vs. 1A, or Classification 5 in Table S1). This means that for the system chosen for this study, 𝛾𝐹 > 0 and thus peptide self-assembly is making the surface more hydrophilic and wetting.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 16

The simple method presented here can be easily adopted to other system with self-assembled layers that are either mobile or immobile on a surface, such as self-assembled monolayers, polymer thin films, peptide bilayers, and other systems where drying may affect the surface properties. It is important to choose appropriate spin conditions to observe the WDT in order to evaluate the relative change in the effective surface energy. We note that while AFM was used as the primary tool to define the WDT in this study, microscope imaging can be equally effective and can be used as an in-situ tool and a much faster method of identifying the transition. The method’s accuracy can be arbitrary improved by choosing smaller increments of spin speed and acceleration. In the example studied here, we observe that the change in the effective surface energy due to the formation of surface-mediated fibrils of Aβ12-28 peptide is 𝛾𝐹 = 42±6 mN ∙ m-1, indicating a more hydrophilic surface condition as the deposition continues. We note that as reported previously, at this stage of the self-assembly process after 30 minutes of incubation the surface coverage of peptides is roughly ~10% and the number density of peptides on the surface is approximately ~2600 m-2.12,21 We also previously observed that for these peptides the self-assembly is self-limiting and stops at a surface coverage of ~2700 m-2. This is consistent, as a dramatically more hydrophilic surface experienced by the peptide will slow the accumulation of mostly hydrophobic peptides on the surface. Thus, reducing the deposition rate to a halt.21 In contrast, if a strain was chosen such that the side chains facing the liquid are more hydrophobic after the self-assembly, one would expect to see the deposition to continue beyond a sub-monolayer of peptides. This will be explored in our future studies. Similarly, this study can enable us to study configurations of proteins adsorbed on peptide bi-layers and their effect on varying the effective surface energy. This can be a powerful, yet easy to adopt tool for future experiments of this nature in estimating the orientation or folding of the protein using simple experimental methods. We also note that if one desires to dry self-assembled layers to image them, this study highlights the importance of choosing appropriate parameters for the spin-coating process to prevent potential drying artifacts due to a dewetting transition during the spin stage. Conditions need to be tailored such that the layer can be dried without going through the WDT. Based on our results, a slow spin speed combined with a fast acceleration rate can be applied to quickly dry the surface.

Conclusions In summary, we have developed a novel and facile method to measure the effective change in the surface energy of a surface due to adsorbed particles (𝛾𝐹). In this method, spin casting is used to induce a wetting-dewetting transition (WDT) in a water droplet placed on the surface before and after peptides are adsorbed on the surface and AFM imaging is used to observe the patterns indicative of wetting or dewetting conditions. The effective surface energy, 𝛾𝐹, is balanced by the energy of the centrifugal forces, 𝛾𝐶, which depend on the spin speed and acceleration. Thus, a comparison between the speed before and

ACS Paragon Plus Environment

Page 13 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

after the self-assembled layer is formed allows direct calculation of 𝛾𝐹. For the system chosen here for the proof of concept, A12-28 peptides, after 30 minutes of deposition, the surface coverage of peptides was roughly ~10% and the number density of peptides on the surface is approximately ~2600 m-2.12,21 For this surface the measured 𝛾𝐹 = 42±6 mN ∙ m-1. This means that the incoming peptides gradually experienced a more hydrophilic surface as the deposition continues. In fact, the deposition of incoming peptides slows down dramatically and saturates at a number density of ~2700 m-2, which is a strong indication of this dramatically increased hydrophilicity and explains our previous observations of a selflimiting self-assembly.21 This approach can be easily adopted for other systems. In particular, it is really challenging to evaluate the change in the effective surface energy experienced by adsorbates in conditions where the adsorbed particles themselves are mobile and can look different if the surface is dried, making such measurements otherwise challenging or impossible. Unlike the spontaneous dewetting method typically applied in thin polymer film studies18,19, the spin-induced dewetting method adopted here can be controlled by the spin-coating parameters (e.g. acceleration rate) and experimental conditions (e.g. surface chemistry). Therefore, this method can be generally applied to other fluidic systems, where the surface is covered with deposited nanostructures.

Acknowledgements The authors would like to thank E. James Petersson and Tobias Baumgart for helpful discussions. This work has been supported by the Penn MRSEC grant (NSF-1720530), Alfred P. Sloan Foundation’s Young Investigator Award, and partially through a seed grant from the National Institute of Aging of NIH under the award number P30AG010124 (PI: John Trojanowski).

Supporting Information Available Supplemental contact angle measurement of a water droplet on mica (Figure S1). Supplemental AFM images of the spun-dried samples prepared under different conditions (Figure S2-S6). A brief description of calculations used to estimate the spreading coefficient of water on mica substrate and the parameters of 𝑚𝐴250, 𝑚𝐴400, and 𝛾𝐹. Summarized WDTs observed in spun-dried samples (Table S1). This information is available free of charge via the Internet at http://pubs.acs.org/.

Notes The authors declare no competing financial interest.

References (1)

Young, T.An Essay on the Cohesion of Fluids. Philos. Trans. R. Soc. London 1805, 95, 65–87.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 16

(2)

Yunker, P. J.; Lohr, M. a.; Still, T.; Borodin, A.; Durian, D. J.; Yodh, a. G.Effects of Particle Shape on Growth Dynamics at Edges of Evaporating Drops of Colloidal Suspensions. Phys. Rev. Lett. 2013, 110, 035501.

(3)

Sahu, N.; Parija, B.; Panigrahi, S.Fundamental Understanding and Modeling of Spin Coating Process: A Review. Indian J. Phys. 2009, 83, 493–502.

(4)

Bornside, D. E.; Macosko, C. W.; Scriven, L. E.On the Modeling of Spin Coating. J. imaging Technol. 1987, 13, 122–130.

(5)

Yonkoski, R. K.; Soane, D. S.Model for Spin Coating in Microelectronic Applications. J. Appl. Phys. 1992, 72, 725–740.

(6)

Gentili, D.; Foschi, G.; Valle, F.; Cavallini, M.; Biscarini, F.Applications of Dewetting in Micro and Nanotechnology. Chem. Soc. Rev. 2012, 41, 4430.

(7)

Flack, W. W.; Soong, D. S.; Bell, A. T.; Hess, D. W.A Mathematical Model for Spin Coating of Polymer Resists. J. Appl. Phys. 1984, 56, 1199–1206.

(8)

deGennes, P.-G.; Brochard-Wyart, F.; Quéré, D.Capillarity and Wetting Phenomena; 2004.

(9)

Cai, Y.; Zhang Newby, B. M.; Newby, B. Z.Dewetting of Polystyrene Thin Films on Poly(Ethylene Glycol)-Modified Surfaces as a Simple Approach for Patterning Proteins. Langmuir 2008, 24, 5202–5208.

(10)

vanHameren, R.; Schon, P.; vanBuul, A. M.; Hoogboom, J.; Lazarenko, S.V; Gerritsen, J. W.; Engelkamp, H.; Christianen, P. C. M.; Heus, H. a; Maan, J. C.; et al.Macroscopic Hierarchical Surface Patterning of Porphyrin Trimers via Self-Assembly and Dewetting. Science (80-. ). 2006, 314, 1433–1436.

(11)

Tian, D.; Song, Y.; Jiang, L.Patterning of Controllable Surface Wettability for Printing Techniques. Chem. Soc. Rev. 2013, 42, 5184–5209.

(12)

Lin, Y.-C.; Petersson, E. J.; Fakhraai, Z.Surface Effects Mediate Self-Assembly of Amyloid-β Peptides. ACS Nano 2014, 8, 10178–10186.

(13)

Lin, Y.-C.; Repollet-Pedrosa, M. H.; Ferrie, J. J.; Petersson, E. J.; Fakhraai, Z.Potential Artifacts in Sample Preparation Methods Used for Imaging Amyloid Oligomers and Protofibrils Due to Surface-Mediated Fibril Formation. J. Phys. Chem. B 2017, 121, 2534–2542.

(14)

Nečas, D.; Klapetek, P.Gwyddion: An Open-Source Software for SPM Data Analysis. Cent. Eur. J. Phys. 2012, 10, 181–188.

(15)

Reiter, G.Unstable Thin Polymer Films: Rupture and Dewetting Processes. Langmuir 1993, 9, 1344–1351.

(16)

deGennes, P. G.Wetting: Statics and Dynamics. Rev. Mod. Phys. 1985, 57, 827–863.

(17)

Bonn, D.; Eggers, J.; Indekeu, J.; Meunier, J.; Rolley, E.Wetting and Spreading. Rev. Mod. Phys. 2009, 81, 739–805.

(18)

Choi, S.-H.; Zhang Newby, B.Alternative Method for Determining Surface Energy by Utilizing Polymer Thin Film Dewetting. Langmuir 2003, 19, 1419–1428.

(19)

Che, J.; Jawaid, A.; Grabowski, C. A.; Yi, Y.-J.; Louis, G. C.; Ramakrishnan, S.; Vaia, R. A.Stability of Polymer Grafted Nanoparticle Monolayers: Impact of Architecture and Polymer–

ACS Paragon Plus Environment

Page 15 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Substrate Interactions on Dewetting. ACS Macro Lett. 2016, 5, 1369–1374. (20)

Mukherjee, R.; Sharma, A.; Steiner, U.Surface Instability and Pattern Formation in Thin Polymer Films. Generating Micro‐ and Nanopatterns on Polymeric Materials. March 19, 2011, pp 217– 265.

(21)

Lin, Y.-C.; Li, C.; Fakhraai, Z.Kinetics of Surface-Mediated Fibrillization of Amyloid-β (12–28) Peptides. Langmuir 2018, 34, 4665–4672.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

“TOC Graphic”

ACS Paragon Plus Environment

Page 16 of 16