Tuning Ice Nucleation by Mussel-Adhesive Inspired Polyelectrolytes: The Role of Hydrogen Bonding

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE5 Sep 2022Tuning Ice Nucleation by Mussel-Adhesive Inspired Polyelectrolytes: The Role of Hydrogen Bonding Jing Chen†, Yongkang Wang†, Benjamin Leibauer, Takakazu Seki, Konrad Meister, Yuki Nagata and Mischa Bonn Jing Chen† *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, School of Science, Tianjin University of Science & Technology, Tianjin 300457 Max Planck Institute for Polymer Research, Mainz 55128 , Yongkang Wang† Max Planck Institute for Polymer Research, Mainz 55128 School of Mechanical Engineering, Southeast University, Nanjing 211189 , Benjamin Leibauer Max Planck Institute for Polymer Research, Mainz 55128 , Takakazu Seki Max Planck Institute for Polymer Research, Mainz 55128 , Konrad Meister Max Planck Institute for Polymer Research, Mainz 55128 School of Arts & Sciences, University of Alaska Southeast, Juneau, Alaska 99801 , Yuki Nagata *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Max Planck Institute for Polymer Research, Mainz 55128 and Mischa Bonn *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Max Planck Institute for Polymer Research, Mainz 55128 https://doi.org/10.31635/ccschem.022.202202087 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Heterogeneous ice nucleation (HIN) on foreign surfaces plays a crucial role across a wide range of environmental and biological processes, and control of HIN is highly desirable. Functionalizing surfaces to control HIN poses interesting scientific challenges and holds great potential for technological impact. Here, we combine the ice nucleation tuning capability of polyelectrolytes with mussel-inspired adhesives to obtain robust surface functionalization with HIN control. We tune ice nucleation by integrating strongly surface-binding catechol derivatives into three different polyelectrolytes via a facile and scalable strategy. Sum-frequency generation (SFG) spectroscopy reveals that the functionalized surface with the lowest ice nucleation temperature (−22.1 °C) exhibits the strongest hydrogen bonding interaction with water molecules. The coating exhibits long-term stability up to 30 days even under harsh conditions such as 5 M salt solution and can be deposited onto various substrates because of its strong adhesion on solid surfaces. Moreover, the precise ice nucleation temperature window can be tuned by controlling the grafting degree of the catechol derivatives. Our study provides a new strategy for tuning ice nucleation and provides a new perspective on understanding the mechanism of ice nucleation at the molecular level by interfacial molecular spectroscopy, which will help in the design and development of anti-ice-nucleation materials. Download figure Download PowerPoint Introduction The undesired formation and accumulation of ice can lead to severe problems in our daily life, ranging from the accelerated deterioration of transmission lines and buildings to the damage of plants, causing economic and energy losses.1–5 The icing of surfaces starts with ice nucleation, followed by growth and propagation6 and adherence to and accumulation on the surface.7,8 Tuning ice nucleation, controlling ice growth and propagation, and reducing ice adhesion are three main strategies for confronting icing problems.9,10 Water freezing at a solid surface is a process of ice nucleation and growth, in which nucleation plays a key role. It is assumed that nucleation occurring at the liquid–solid interface is dominated by heterogeneous nucleation at low supercooling temperatures.11 Heterogeneous ice nucleation (HIN) often starts on foreign surfaces and is governed by surface-inherent properties such as interfacial free energy, interfacial geometry, and lattice match.12 To control ice growth and propagation, hygroscopic materials and structural design are often employed to disturb ice bridging or weaken ice propagation.13,14 If the ice nucleation and propagation cannot be effectively controlled, the ice will become thicker and denser on the surface. Reducing ice adhesion is the last line of defense against ice. There is a growing body of literature that focuses on the reduction of ice adhesion,15–21 with a reduction of ice adhesion to below 10 kPa according to a recent study.22 Such weak adhesion allows ice to be easily shed off from the surface by the action of wind or gravity. However, ice nucleation is the initial and rate-limiting step of the ice formation.23 In this regard, special attention should be given to ice nucleation, especially for tuning HIN. Antifreeze (glyco)proteins (AF(G)Ps) protect a variety of organisms inhabiting subzero environments by efficiently controlling ice growth, and thus have been employed to modify surfaces to control ice nucleation.24–26 Inspired by AF(G)Ps, polypeptides have been designed and synthesized to suppress ice nucleation or retard recrystallization. Yang et al.27 tethered supercharged polypeptides with different surface charges and charge densities onto a surface to regulate ice nucleation, and they attributed the change of ice nucleation temperature on that charged surface to the asymmetric polarization of interfacial water. By mimicking AFPs, Wang designed and engineered self-assembling peptides, which showed moderate inhibition of ice crystal growth and recrystallization owing to the alignment of threonine residues with their intervals matching the ice lattice.28 To immobilize AF(G)Ps or polypeptides onto the surface, pre-functionalization of the substrate is required before it is coated, that is, a special functional group must first be bound on the surface of the substrate, thereby restricting the widespread application of AF(G)Ps due to the complex procedure.29 Therefore, a universal, easily applicable coating that can be applied to various solid surfaces for tuning ice nucleation is highly desired. Fortunately, nature has again provided us with a potential strategy to overcome these challenges. Recently, mussel proteins have attracted much attention because they can strongly stick to virtually any type of substrate with high binding strength via secreting Mytilus edulis foot protein.30,31 Their extraordinary ability to stick to most inorganic and organic surfaces is traced to the high levels of catechol and amine functional groups in the mussel foot protein (mfp, see Figure 1a).32–34 Hong et al.35 indicated that both covalent and non-covalent interactions between the catechol derivatives and substrates are responsible for strong catechol adhesion. Consequently, this biomimetic concept offers a new avenue for constructing coatings for tuning ice nucleation. Figure 1 | (a) Schematic of a mussel anchored by byssal threads and adhesive plaques to a rock and the interfacial location of mfp-5 with a simplified molecular representation of characteristic amine and catechol groups. (b) Synthesis of the CPs. DA and DOPA both contain catechol functional groups and were used as mfp-mimetic small molecules for grafting the polyelectrolytes onto solid surfaces. (c) Schematic representation of the process of the CP coating. Download figure Download PowerPoint Benefiting from the strong adhesive property of catechol derivatives and the diversity of functional groups of polyelectrolytes, we designed a series of catechol-containing polyelectrolytes (CPs). Three different CPs were prepared using an amidation reaction through a facile and green one-step procedure for tuning ice nucleation. Of these three, the surface coated by catechol-containing branched poly(ethylenimine) (bPEI-C) shows good ice nucleation suppression, and it can be applied to various surfaces due to the nonselective adhesion of catechol groups. The range of HIN temperature (TH) could be enlarged after cross-linking by controlling the molar feed ratio between the catechol derivative and polyelectrolyte. Our analysis using surface-specific vibrational spectroscopy reveals that the interactions between functional groups of polyelectrolytes and interfacial water molecules play a key role in tuning ice nucleation for polyelectrolytes. The ice nucleation temperature decreases for stronger interactions between the functional groups of polyelectrolytes and water molecules. In addition, the bPEI-C coating maintains excellent stability under harsh conditions and can be prepared on a large scale, showing excellent application prospects. Experimental Methods Materials and instrumentation All chemicals were obtained from commercial sources and used without further purification. The absorbance spectra of the prepared CPs were obtained using a Cary 60 UV–vis spectrophotometer (Agilent, Santa Clara, California, United States). Fourier transform infrared (FT-IR) spectra were recorded by the attenuated total reflection method with an FT-IR spectrometer (Tensor II, Bruker, Rosenheim, Germany). Proton nuclear magnetic resonance (1H NMR) spectra were recorded at 300 MHz on a Bruker AVANCE III 300 NMR spectrometer. X-ray photoelectron spectroscopy (XPS) was performed on the Thermo Scientific ESCALab 250Xi (Thermos Scientific, Waltham, Massachusetts, United States) using 200 W monochromatic Al Kα radiation. The base pressure in the analysis chamber was approximately 3 × 10−9 mbar. The pass energy was 30 eV. The static contact angle (CA) on the samples was measured at room temperature with a CA System (DataPhysics Instruments OCA35 goniometer, Dataphysics, Stuttgart, Germany). The average CA values were obtained by measuring at five different positions on the same sample. The morphologies of the solid substrates before and after coating by the CPs were obtained by atomic force microscopy (AFM) (NanoWizard 4, JPK Instruments, Berlin, Germany) in the tapping mode. Tapping mode etched silicon probes OTESPA with aluminum reflective coating on the backside of the cantilever were used (cantilever resonance frequencies were approximately 300 kHz; a typical spring constant 26 N m−1). The scan size was 2 μm × 2 μm with 512 × 512 pixels. Synthesis of catechol-containing PAH, PAA, and bPEI Synthesis of PAA-C An aqueous solution of poly(acrylic acid) (PAA; 1.2 g of a 35 wt % solution) was diluted with 0.1 M phosphate-bufferred solution (PBS) (30 mL). Then 115.0 mg 1-Ethyl-3-(3-dimethyllaminopropyl)carbodiimide hydrochloride (EDC·HCl) and 69.1 mg N-Hydroxysuccinimide (NHS) was added to the above solution. Then, 113.8 mg dopamine (DA) hydrochloride was added. The reaction solution was adjusted to pH 5.0 with 1.0 M NaOH and HCl and stirred for 24 h under N2 gas protection. Synthesis of PAH-C and bPEI-C 3-(3,4-dihydroxyphenyl) propionic acid (DOPA; 109.3 mg) was dissolved in 0.1 M PBS (50 mL) followed by EDC·HCl (115.0 mg) and NHS (69.1 mg). The solution was stirred for 2 h. The poly(allylamine) (PAH; 1.7 g of a 20 wt % solution) or bPEI (360 mg) was dissolved in 0.1 M PBS (30 mL) and then added to the above solution, and the reaction solution was adjusted to pH 5.0 with 1.0 M NaOH and HCl and stirred for 24 h. The resultant mixture was dialyzed by a regenerated cellulose membrane (molecular weight cut-off = 3500) in deionized water. The final conjugates were lyophilized. The construction of coating on the solid substrates The CPs were dissolved in deionized water at a concentration of 1 mg/mL. Then the pH was adjusted to 8.0 with 1M NaOH and HCl. The cleansed substrates were incubated in the above solution for 6 h. Then the solid substrates were removed, rinsed with ultrapure water, and dried under a stream of nitrogen gas. Characterization The ice nucleation temperatures and freezing delay time on CP-coated surfaces were performed by a homemade experimental apparatus with a microscope coupled with a high-speed camera and a cryostage, as described elsewhere.27,36 Our Im χ eff ( 2 ) spectra were measured using Heterodyne-Sum Frequency Generation (HD-SFG) spectroscopy at ssp polarization combination, where ssp denotes s-polarized sum-frequency generation (SFG), s-polarized visible, and p-polarized IR beams. Further details about ice nucleation measurement and SFG measurement can be found in the Supporting Information. Results and Discussion Synthesis and characterization of CPs and CP-coated surfaces Inspired by the strong adhesion due to the catechol and amine groups in mfps (Figure 1a), DA and DOPA were used as mfp-mimetic small molecules and were grafted to PAA, PAH, and bPEI, respectively ( Supporting Information Figure S1). As seen in Figure 1b, three types of CPs were synthesized by the conjugation of catechol derivatives with polyelectrolytes via an amidation reaction between amino and carboxyl groups under mild synthetic conditions. The resulting catechol-grafted polymers were denoted as PAH-C, PAA-C, and bPEI-C. The successful introduction of the DA or DOPA into polyelectrolytes was evidenced by UV–vis spectroscopy, FT-IR, and 1H-NMR analysis (see Supporting Information Figure S2).37–39 Figure 1c illustrates the fabrication procedure of the CP coating. First, CPs were dissolved in water, and cleaned silicon wafers were immersed in the solution. CP-coated surfaces were obtained after incubation for a few hours due to the strong adhesion of the catechol groups.40 XPS was utilized to characterize the surface composition. A new peak of the N 1s on the PAA-C coated surface and intense N 1s peak on PAH-C- and bPEI-C-coated surfaces indicated that CPs were successfully deposited on the silicon wafer surface ( Supporting Information Figure S3). AFM measurements shown in Supporting Information Figure S4 revealed a root mean square roughness of the CP-coated surfaces of less than 3 nm, exhibiting relatively homogeneous and smooth surfaces. After coating with the CPs, we found that the static contact angle on the PAA-C coated surface was lower than that of the PAH-C coated surface ( Supporting Information Figure S5). But the bPEI-C coated surface had the lowest water contact angle, which indicates a better affinity between the amino groups and the water molecules due to the highly branched structure of bPEI. Ice nucleation The ice nucleation temperature was determined by measuring the freezing temperatures of condensed microdroplets on CP-modified surfaces in a homebuilt closed cell to avoid possible extraneous effects. We placed a macrodroplet of ultrapure water (Milli-Q, 18.2MΩ cm) at the edge of the sample surface in a closed chamber with a relative humidity of 100%. Then, many microdroplets appeared on the CP-modified surfaces via an evaporation–condensation process ( Supporting Information Figure S6). Afterward, the nucleation temperature of the condensed microdroplets was recorded as the sample surface temperature was lowered. This evaporation–condensation method is widely used to investigate the effect of solid surfaces on ice nucleation.24,39,41,42 Figure 2a shows the change of morphology and opacity of water microdroplets condensed onto the PAH-C (top), PAA-C (middle), and bPEI-C (bottom) modified surfaces with decreasing temperature. For temperatures down to −10 °C, water microdroplets remained in the liquid state on all CP-coated surfaces. As the temperature dropped below −19 °C, most of the water microdroplets on the PAH-C coated surface suddenly changed in opacity, and the edges of the microdroplets changed their shape, reflecting ice nucleation.27 Although the ice nucleation is slower than the other stages in the icing process, the rapid growth results in the quasi-immediate freezing of water microdroplets upon the formation of ice nuclei.43 Such ice nucleation does not occur until temperatures drop below −20 and −22 °C for microdroplets on the PAA-C and bPEI-C modified surfaces, respectively. Figure 2 | (a) Optical microscopy images of water microdroplets condensed onto the CP-modified silicon wafers. The red arrow indicates a drop in temperature from −10 to −22 °C. As the temperature decreases, the PAH-C coated surface freezes first, followed by PAA-C, and bPEI-C coated surfaces. Cooling rate: 2 °C min−1; Scale bar: 100 μm. (b) Temperatures of ice nucleation of condensed microdroplets and single macrodroplets on the CP-modified surfaces. (c) Cooling rate dependence of ice nucleation temperature on CP-modified surfaces. (d) The freezing process of a water droplet on the bPEI-C-coated surface. Scale bar: 100 μm. (e) Illustration of the icing process on the bPEI-C coated surface. The error bars represent one standard deviation of uncertainty. Download figure Download PowerPoint The average ice nucleation temperatures for condensed microdroplets and single macrodroplets are summarized in Figure 2b with a cooling rate of 2 °C min−1. The freezing temperatures for condensed microdroplets on the PAH-C, PAA-C, and bPEI-C coated surfaces are −18.8 ± 0.5 °C, −20.1 ± 0.6 °C, and −22.1 ± 0.3 °C, respectively. Previously, Zhang et al.44 immobilized short peptides onto surfaces as an anti-icing coating for the first time, and the freezing temperature was reduced to −16.3 °C, which is higher than that of the CP-coated surface in our study. Though supercharged polypeptides exhibited the same performance on tuning ice nucleation with the lowest freezing temperature of −22.8 ± 0.6 °C, easier synthesis and lower toxicity make the CPs more attractive for potential practical application, such as food storage and cryopreservation of cells and tissues.27 To verify the data obtained with microdroplets, we also investigated the freezing of single macroscopic droplets on CP-coated surfaces, recording more than 200 independent freezing events. The ice nucleation temperature on CP-coated surfaces resembled a Gaussian distribution, as shown in Supporting Information Figure S7. In situ optical microscopic observation of macrodroplets showed a significant difference in freezing temperatures on the different CP-coated surfaces. The freezing temperatures of macrodroplets on the PAH-C, PAA-C, and bPEI-C coated surfaces displayed the same trend as that of the microdroplets. The average freezing temperature of macrodroplets was slightly lower than that of the condensed microdroplets, which can be rationalized by the presence of a temperature gradient when a cooling stage cools down macroscopic droplets.27 Though the ice nucleation temperature decreased with increasing cooling rate, it showed the same trend of ice nucleation temperature for CP-coated surfaces with different cooling rates, that is, TH,PAH-C > TH,PAA-C > TH,bPEI-C (Figure 2c). When the cooling rate is too fast, the TH would be affected due to the heat transfer effect,45 providing a reasonable explanation for the observed cooling-rate dependence of the ice nucleation temperature, which agrees with previous reports.46 To investigate icing process on the CP-coated surfaces, a high-speed camera was employed to record the changes from water to ice on the bPEI-C coated surface. Then, we extracted sequential images of the bPEI-C modified surfaces before and after freezing. As shown in Figure 2d, the ice nucleation initiated at the liquid–solid interface, highlighted by the red circle. Then ice nuclei grew along the liquid–solid interface, followed by upward ice growth until the droplet was completely which is in Figure CP-coated surfaces lead to different HIN due to different interfacial which has been observed for microdroplets and single It has been that ice nucleation is on a charged surface, it is on a charged This agrees with the we obtained on charged PAH-C and charged The ice nucleation temperature on the bPEI-C-coated surface was the lowest the three CP-coated surfaces, the PAH-C and bPEI-C have the same charge properties and functional groups. Further analysis showed that the effect of the CP on ice nucleation was ( Supporting Information Figure A lower ice nucleation temperature a stronger interaction between the CP-coated surface and the water the water molecules from in an In to interactions between water molecules and or groups in CPs also play a To this the interfacial water structure at the interface was by measuring the Im χ eff ( 2 ) spectra in the mode using The for measurement is shown in Figure The measured Im χ eff ( 2 ) spectra for CPs in contact with 1 and 100 are shown in Figure The Im χ eff ( 2 ) showed a for PAA-C, which was to the of the groups of water The was that the groups were up to the with the that PAA-C is For both PAH-C and were because they have a charged surface. Figure 3 | (a) Schematic of the SFG PAH-C, PAA-C, and bPEI-C were coated on optical substrates for SFG (b) Im spectra of water molecules at the interface for 1 and 100 solution. (c) Im spectra of water molecules at the interface for 1 and 100 solution. (d) of of for CPs in contact with 100 solution. Download figure Download PowerPoint The difference between the Im χ eff ( 2 ) spectra at the different the charge of the The Im χ eff ( 2 ) spectra ( Im χ eff ( 2 ) are displayed in Figure For PAA-C, the of Im χ eff ( 2 ) is for both PAH-C and the Im χ eff ( 2 ) are Such difference that the interaction between CPs and water molecules is weak because be into the and then the CP surface (Figure In this the Im χ eff ( 2 ) spectra for different CPs are dominated by the from the interfacial Figure | (a) delay of ice nucleation measured at °C on CP-coated surfaces. (b) The of ice nucleation temperature on the bPEI-C coated surface before and after in water and in 5 M solution for up to 30 (c) Ice nucleation temperature on the bPEI-C coated silicon coated silicon and of ice nucleation temperature before and after coating on and with (d) Schematic representation of procedure for fabrication of CPs coating. The error bars represent one standard deviation of uncertainty. Download figure Download PowerPoint By the Im χ eff ( 2 ) spectra for different CPs measured at 100 we found that the frequencies of the were as in Figure The frequencies were and for PAH-C, PAA-C, and bPEI-C coated surfaces, respectively. A higher hydrogen bonding between water molecules and CPs, and For the was the the strongest hydrogen bonding water from to when the sample is down to lower As the TH for PAH-C coated surface with the lowest of of interfacial water and the hydrogen bonding we for these relatively charged surfaces, hydrogen bonding interactions than charge the ice nucleation for polyelectrolytes. The formation of ice was by strong hydrogen bonding interaction and by weak hydrogen bonding The ability of CPs to tune the ice nucleation was by measuring the delay time of ice nucleation which was the from the time of a supercooling to the time of ice nucleation to As shown in Figure the delay time was different on the different CP-coated surfaces, for water on the PAH-C surface after the water drop not until and on the PAA-C and bPEI-C surfaces, respectively. The trend of of water droplets on different CP-coated surfaces shows the of (bPEI-C) > > which is with the freezing temperature obtained from TH (see Figure with the above experimental To the stability of the bPEI-C the silicon wafers by bPEI-C were in the water, and solution with a concentration of 5 M for up to 30 days (Figure Then they were and rinsed with ultrapure water and dried with high nitrogen for measuring the ice nucleation temperature. We found the ice nucleation temperatures for the bPEI-C-coated surface after to these harsh conditions from the prepared excellent stability due to the strong adhesion of the bPEI-C which was by To further the window of ice nucleation temperature, we the molar feed ratio between bPEI and DOPA from to as and then was added to the cross-linking (see Supporting Information Figure was used as an group to the bPEI onto different substrates and as the to the bPEI via It has been that are of water in the of polyelectrolytes, that is, water and water, and the of types of water can be via cross-linking the polyelectrolytes for tuning ice nucleation according to our previous we the ice nucleation temperature of the coated surface and found that it was decreased to a of °C. Moreover, to investigate the of the CP we the ice nucleation temperatures for three types of solid surfaces before and after coating with and typical daily materials. Figure the between the coated surfaces and surfaces. For the coated surfaces, the ice nucleation temperatures were the ice nucleation temperature is independent of Moreover, lower ice nucleation temperature on the coated surface the coating a range of materials for ice nucleation along with a facile and scalable procedure is to the of anti-ice-nucleation materials. In our the CPs are and and they can to virtually all types of substrates because of the strong adhesion of the catechol make it for As a is for various solid we the with solution (Figure and the

Topics & Concepts