Thermodynamic and Kinetic Parameters for Calcite Nucleation on Peptoid and Model Scaffolds: A Step toward Nacre Mimicry

The production of novel composite materials, assembled using biomimetic polymers known as peptoids (N-substituted glycines) to nucleate CaCO3, can open new pathways for advanced material design. However, a better understanding of the heterogeneous CaCO3 nucleation process is a necessary first step. We determined the thermodynamic and kinetic parameters for calcite nucleation on self-assembled monolayers (SAMs) of nanosheet-forming peptoid polymers and simpler, alkanethiol analogues. We used nucleation rate studies to determine the net interfacial free energy (γnet) for the peptoid–calcite interface and for SAMs terminated with carboxyl headgroups, amine headgroups, or a mix of the two. We compared the results with γnet determined from dynamic force spectroscopy (DFS) and from density functional theory (DFT), using COSMO-RS simulations. Calcite nucleation has a lower thermodynamic barrier on the peptoid surface than on carboxyl and amine SAMs. From the relationship between nucleation rate (J0) and saturation state, we found that under low-saturation conditions, i.e. <3.3 (pH 9.0), nucleation on the peptoid substrate was faster than that on all of the model surfaces, indicating a thermodynamic drive toward heterogeneous nucleation. When they are taken together, our results indicate that nanosheet-forming peptoid monolayers can serve as an organic template for CaCO3 polymorph growth.


Characterization of SAM functionalized substrates: X-ray photoelectron spectroscopy (XPS) of SAM functionalized substrates
We used XPS to study the elemental composition of the functionalized substrates. From small shifts in binding energy, we could gain insight into the chemical state of the elements. We used a Kratos Axis Ultra DLD spectrometer operating with Al monochromatic K α radiation (1486.6 eV; 150 W). All data were collected at room temperature. The spectra were fit using the software CasaXPS. For fitting the background contribution, we used a Shirley or a linear function, as appropriate. The spectra were calibrated to the Au 4f peak at 84.2 eV that came from the substrate. The photoelectron peaks were fit using 30% Gaussian and 70% Lorentzian contributions.
Alkane thiol SAMs on Au substrates: Figure S1 shows an XPS wide spectrum of the carboxyl SAM. We detected only elements from the Au-substrate and the HS(CH 2 ) 10 COOH. The high resolution C 1s spectrum (Fig. S1, inset) contains contributions from the aliphatic carbon (C-C or C-H, 285.0 eV) and the carboxyl group (O=C-O(H), 289.5 eV). Small peaks between the two main peaks are ascribed to C-S and C-OH. Figure S1: XPS wide scan of the carboxylic SAM functionalized surface prepared using acetic acid as a dispersion agent. Inset is the C 1s high resolution spectrum, which shows that the carboxyl group accounts for 6.4% of the total carbon. Figure S2 shows the XPS wide spectrum for the amine SAM. We detected the elements from the Au-substrate and the HS(CH 2 ) 11 NH 2 with some Cl that was left from the rinsing procedure. The high resolution N 1s spectrum (Fig. S2,inset) showed that the distribution between protonated and deprotonated amine was 1:2.

Ratio of bound to unbound S in the alkane thiol SAMs:
To evaluate the success of thiol binding to Au substrates, we compared the amount of Au bonded sulfur to adsorbed sulfur for each of the substrates. As an example, Figure S3 shows the S 2p high resolution spectrum for the carboxyl functionalized substrate. There are two 2p doublets, one representing S bonded to Au and the other, physisorbed S.
For the carboxyl SAM, 73% of the thiol molecules were bound. For the amine SAM, 75% of the thiol molecules were bound and for the mixed SAM, ~76% were bound.

Quantification of COO(H) : NH 2 on mixed SAM:
We used the mixed SAMs as a model for B28 peptoid polymers. The B28 peptoid has the same amount of exposed carboxyl and amine functional groups, thus for our thiol model surfaces we aimed to prepare a 50/50% 11-mercaptoundecanoic acid to 11-amino-1-undecanethiol hydrochloride mixed SAM. The ratio of carboxyl and amine was estimated by quantifying the contribution of the carboxyl peak to the nitrogen peak from the high resolution spectra.

Investigation of CaCO 3 -functionalized substrates:
Illustration of the nucleation rate setup: Figure S4 shows a schematic drawing of the experimental setup used to measure nucleation rate. Upstream to the flow cell is a static mixer, which ensures mixing of the CaCl 2 and the NaHCO 3 solutions. Because the mixed solution is supersaturated, we used a mixer with a volume of only 50 μL and tubing with 0.5 mm inner diameter, to minimize the dead volume between the inlet of the static mixer to the sample.

Solution chemistry determined from PHREEQC calculations:
From PHREEQC, 1 we obtained the activity of the ions and saturation with respect to calcite. All crystals counted as heterogenously grown were thus first detected on the surface spot where they grew. For the substrates with a long incubation time and low nucleation rate, the solution is not deprived of ions at rates fast enough and becomes increasingly supersaturated. However, when ions from the solution are being consumed for heterogeneous nucleation, the homogenous nucleation rate decreases. The induction time at the carboxyl SAM was low and the nucleation rate relatively fast, thus homogeneous nucleation was not observed for any of the applied saturation levels. The resolution limitation of this method can result in some miscount, which leads to larger uncertancy on J 0 for surfaces with few nucleation sites. Figure S5: Optical microscopy images of carboxyl SAMs. All crystals nucleated heterogeneously. Scale bar is 100 μm. Figure S6: Optical microscopy images of amine SAMs. Crystal sites framed by squares have nucleated heterogeneously on the surface; others nucleated homogenously in solution (which we know because they move during the experiment, and landed on the substrates as large particles).
These are ignored during the rate analysis. Scale bar is 100 μm. Figure S7: Optical microscopy images of mixed SAMs. All crystals nucleated homogenously in solution. We know this because they moved in the field of view during the experiment. Scale bar is 100 μm. Figure S8 shows images obtained during the nucleation rate studies on B28 peptoid HOPG atr σ = 5.85. The step edges come from the topography of the underlying HOPG surface. Figure S8: Optical microscopy images of B28 peptoids on HOPG surface after flow experiment. Scale bar is 100 μm.

Functionalized substrates after nucleation, imaged with SEM:
To deduce from which planes crystals nucleated, we used SEM to image the functionalized surfaces after the nucleation rate studies. Figure S9 and S10 show images at various magnifications for experiments carried out on the carboxyl and amine SAMs. Figure 11 shows the B28 peptoid on highly ordered pyrolitic graphite (HOPG), imaged after flow at σ = 5.85. As a reference study, we also imaged a freshly cleaved HOPG surface without the peptoid polymers, after flow experiments with σ = 5.85 (Fig. S12). On the freshly cleaved HOPG surface, the particle shape suggests vaterite, whereas crystals with calcite morphology grew on the B28 peptoid substrate.  Figure S11: Calcite crystals grown on self-assembled B28 peptoid polymer substrates after the flow experiment at σ = 5.85. Figure S12: Images from the freshly cleaved HOPG surface after the experiment ended (σ = 5.85). Particle form is reminiscent of vaterite.   (Table S2).