Magnetite Mineralization inside Cross-Linked Protein Crystals

Abstract

Crystallization in confined spaces is a widespread process in nature that also has important implications for the stability and durability of many man-made materials. It has been reported that confinement can alter essential crystallization events, such as nucleation and growth and, thus, have an impact on crystal size, polymorphism, morphology, and stability. Therefore, the study of nucleation in confined spaces can help us understand similar events that occur in nature, such as biomineralization, design new methods to control crystallization, and expand our knowledge in the field of crystallography. Although the fundamental interest is clear, basic models at the laboratory scale are scarce mainly due to the difficulty in obtaining well-defined confined spaces allowing a simultaneous study of the mineralization process outside and inside the cavities. Herein, we have studied magnetite precipitation in the channels of cross-linked protein crystals (CLPCs) with different channel pore sizes, as a model of crystallization in confined spaces. Our results show that nucleation of an Fe-rich phase occurs inside the protein channels in all cases, but, by a combination of chemical and physical effects, the channel diameter of CLPCs exerted a precise control on the size and stability of those Fe-rich nanoparticles. The small diameters of protein channels restrain the growth of metastable intermediates to around 2 nm and stabilize them over time. At larger pore diameters, recrystallization of the Fe-rich precursors into more stable phases was observed. This study highlights the impact that crystallization in confined spaces can have on the physicochemical properties of the resulting crystals and shows that CLPCs can be interesting substrates to study this process.

Figure 1

(A) Optical micrograph of each protein crystal: A1—lysozyme; A2—glucose isomerase; A3—lipase. (B) Crystal structures: B1—lysozyme; B2—glucose isomerase; B3—lipase. (C) Schematic picture describing magnetite precipitation inside pores of CLPCs.

Figure 2

Event sequence in protocol Type 1: (1) CLPCs are added to the master solution (NaHCO₃/Na₂CO₃, Fe(ClO₄)₂, and FeCl₃), (2) incubation of the CLPCs to allow iron diffusion, (3) initiation of the precipitation by the addition of NaOH and changing the pH to 12.5, (4) precipitation of magnetite, and (5) fishing out the CLPCs for the analysis. Type 2 “Cycles”: (1) 100 CLPCs are added in the master solution, (2) incubation, (3) initiation, (4) precipitation of magnetite, and (5) at least two CLPCs are preserved for characterization and the rest of the crystals are placed in a fresh master solution to start a new cycle, up to nine cycles.

Figure 3

TEM images of (A) iron oxide nanoparticles grown in CLLC experiments (four cycles) inside and outside the protein crystal and (B) in the bulk (protein free) experiment (four cycles).

Figure 4

TEM images of magnetite grown outside (A1) and iron oxides nanoparticles inside (B1, C1, and D1) CLLCs after cycle number 2 (A1 and B1), 4 (C1), and 8 (D1). SAED diffraction images are shown in A2, B2, C2, and D2 (insets correspond to the diffraction area). Only crystals obtained outside CLLCs showed SAED diffraction pattern consistent with magnetite (A2) (see also Figure S2).

Figure 5

TEM images of iron oxides nanoparticles grown within CLGICs (A) and CLLPCs (B) after two and one cycles, respectively. The plot (C) shows the average nanoparticle particle size for the three proteins after the first cycle.

Figure 6

HR-TEM images of magnetite grown inside CLGICs after cycle 2 (A1), cycle 4 (B1), cycle 6 (C1), and cycle 8 (D1) and the corresponding SAED diffraction images of selected regions (insets in A2, B2, C2, and D2). C2 shows single crystal spots corresponding to 111 magnetite reflections. Indexation of all images is shown in Table S1.

Figure 7

Mean size of magnetite nanoparticles formed inside pores of CLLCs and CLGICs and in solution without CLPCs (control), big standard deviation in the controls appears due to the heterogeneous size of the particles.

Figure 8

(A) TEM image of CLGICs after eight cycles which illustrates the distribution of magnetite nanoparticles. Particle distribution was calculated along the length of the crystal by determining the black/white ratio in eight regions of 135 nm each and plotted in the insert. (B) Number and (C) size of the magnetite nanoparticles formed near and far from the border of the CLPCs, based on TEM images from the ninth cycle.