Boosting Self-Assembly Diversity in the Solid-State by Chiral/Non-Chiral ZnII -Porphyrin Crystallization.

A chiral ZnII porphyrin derivative 1 and its achiral analogue 2 were studied in the solid state. Considering the rich molecular recognition of designed metalloporphyrins 1 and 2 and their tendency to crystallize, they were recrystallized from two solvent mixtures (CH2 Cl2 /CH3 OH and CH2 Cl2 /hexane). As a result, four different crystalline arrangements (1 a,b and 2 a,b, from 0D to 2D) were obtained. Solid-state studies were performed on all the species to analyze the role played by chirality, solvent mixtures, and surfaces (mica and HOPG) in the supramolecular arrangements. By means of combinations of solvents and substrates a variety of microsized species was obtained, from vesicles to flower-shaped arrays, including geometrical microcrystals. Overall, the results emphasize the environmental susceptibility of metalloporphyrins and how this feature must be taken into account in their design.


INTRODUCTION
Structural diversity and advanced activity based on supramolecular self-assembly macrocycles are challenging research topics in the field of materials science. 1,2 In the design of non-covalent intermolecular interactions that resembles nature, [3][4][5] the synthetic methodologies existing are mostly limited by the difficulties of anticipating the final arrangements and the lack of reversibility, where the thermodynamic products are the ending tract, hampering both, control and functioning. Owing to the premature stage of the subject basic studies are required, where a key-point is the use of molecular units that can be easily tailored.
Regarding this, porphyrin derivatives are excellent molecular prototypes due to the possibility of direct information transfer from bench-experiments into biological facts. 6 We can ensure self-assembly because of the sum of its manifold features: (i) the conjugated core that has given copious studies with plenty of data regarding the formation of H/J aggregates; 7 (ii) the use of a metal center that provides additional interactive site in the final metallo-porphyrins; 8 and (iii) the addition of organic groups in the meso-positions of the tetrapyrrolic units that afford extra interactions depending on their number and nature. 1,9 Self-assembled aggregates of meso-substituted metalloporphyrins are therefore quite difficult to analyze and specially to anticipate. An approach, adopted by us 10,11 and others 12,13 relies on chirality as the driving force in the achievement of highly ordered structures and the analyses in solution of their noncovalent interactions (hydrogen bonds, - stacking or coordination of through the metal center) and self-assembly mechanisms. However, much less it is known about the genuine effect of chirality on the final architectures, with no strict comparison of chiral/non-chiral analogous porphyrin types, as well as there is scare information on the projection of the knowledge gathered in solution to the solid state.
Here, we unify these two ideas toward the analysis of two novel Zn-porphyrins containing phenoxy propanoate and methyl 2-phenoxy acetate groups (1 and 2, respectively) in the four meso-positions of porphyrins (Scheme 1). The choice of such final entities relate to previous experiences with related systems 10,14-17 foreseen their rich molecular recognition capacity and tendency to crystallize.
Restricting our recrystallization methods, identical otherwise, for our chiral and non-chiral Zn-porphyrin systems allow new findings in the solid state primarily by the use of X-ray diffraction. In addition, our study integrates the effect of solvent polarity and highlights the polyvalent coordination of the Zn II ions together with the application of solid-state techniques to describe in a great manner the final self-assembled architectures. Here, we do not focus on the dynamics of the supramolecular arrangements but on the outcomes, portraying how small changes can make a great difference in the organization of the porphyrin entities. Scheme 1. General scheme of the disposition of the arms in the meso-porphyrins, chiral (1) and non-chiral versions (2).

EXPERIMENTAL DETAILS
Materials and Methods. Experiments were carried out in aerobic conditions or under N 2 atmosphere when required, using commercial grade solvents for the synthesis of the four crystallographic species. Solvents were dried and distilled for some of the synthetic steps and for the absorption UV-Vis studies. Methyl (4-formylphenoxy) acetate was purchased in Activate Scientific. (R)-methyl-2-(4-formyl phenoxy)-propionate was achieved by modifying prior procedure improving the yield. 14 5,10,15,20tetra[(4-R,R,R,R)-methyl-2-phenoxy-propanoate]porphyrin, here described as 4R-H 2 PPP, was synthesized according to the procedure described elsewhere. 10 Synthesis. Synthesis of 5,10,15,20-
Physical Measurements. C, H and N analyses were performed with a Perkin-Elmer 2400 series II analyzer in London Metropolitan University and Thermo Fisher Scientific Flash EA 2000 CHNS with accessory Microbalança MX5 Mettler Toledo in Servei d'Anàlisi Química in UAB. MALDI-TOF/TOF MS: Mass spectra were recorded using matrix assisted laser desorption ionization with time of flight (MALDI-TOF) mass spectrometer ULTRAFLEXTREME (Bruker) at Servei de Proteòmica i Biologia Estructural (SePBioEs) from UAB. FTIR: Infrared spectra (4000-450 cm -1 ) were recorded from Spectrometer Perkin-Elmer Spectrum One. The measurement was performed in universal attenuated NMR spectra: 1 H-NMR spectra were obtained on Bruker Advanced Ⅱat 300 MHz and 298 K. The 13 C-NMR spectra were recorded using frequency 100 MHz in the Servei de Ressonància Magnètica of the Universitat Autònoma de Barcelona, performed on a Bruker AVANCE-III 400 MHz spectrometer. UV-Visible absorption measurements: All UV-visible spectra for liquid samples were obtained on a Varian Cary 5000 using quartz cells. For solid samples it is available a Diffuse Reflectance Sphere DRA-2500 accessory in the UV-Vis-NIR Varian Cary 5000 spectrophotometer, with operational range of 190-3300 nm, and be measured mainly in reflectance or transmittance mode. Powder X-ray diffraction (PXRD): Dates were collected on Panalytical X'PERT PRO MPD diffractometer using Cu K α radiation at 295K. CD: KBr discs were used as the solid solution for the study of circular dichroism spectra. The solidstate circular dichroism spectra were recorded on a JASCO-715 spectrometer fitted with a sample holder. For vibrational circular dichroism (VCD), the equipment is a PMA50 module coupled to a Bruker Tensor 27 FT-IR spectrometer. Investigation of morphology and surface self-organization phenomena were performed on Luminescence spectrometer Perkin Elmer LS4 and Scanning Electron Microscope (SEM) QUANTA FEI 200 FEG-ESEM. Atomic force microscopy (AFM) contact mode measurements were carried out at room temperature using a commercial head and control unit from Nanotec Electr nica and Si tips mounted in soft (k  0.01-0.5 Nm -1 ) Veeco cantilevers. All data were analysed with the WSxM freeware. 18 Single crystal X-ray diffraction. Crystallographic data for 1a were measured on a Nonius Kappa CCD diffractometer with graphite monochromatic Moradiation (λ = 0.71073 Å) at 233 K and used in the structure solution and refinement with the SHELXTL 5.10 package 19 without absorption correction. Hydrogen atoms at the coordinated water molecule O(13) were found and refined with bond restraints (d = 0.83Å) and fixed isotropic thermal parameter (1.2 times higher than Ueq of O(13). The absolute structure of the compound could be confirmed by a Flack parameter of 0.003 (10). Crystallographic data for 1b,2a and 2b were collected at 100K at XALOC beamline at ALBA synchrotron ( = 0.97472 Å for 1b and 0.79472 Å for 2a and 2b). Data were indexed, integrated and scaled using the XDS 20 and IMOSFLM 21 programs. Absorption corrections were not applied. The structures were solved by direct methods and subsequently refined by correction of F 2 against all reflections, using SHELXS2013 22 and SHELXL2013 23 within the WinGX package. 24 All non-hydrogen atoms were refined with anisotropic thermal parameters by fullmatrix least-squares calculations on F 2 using the program SHELXL2013. Hydrogen atoms were inserted at calculated positions and constrained with isotropic thermal parameters. Some reflections (2 in 2a and 6 in 2b) were omitted from the refinement due to the Iobs and Icalc differ more than 10 times SigmaW. In 1b and 2a, the low data completeness is attributed to the data collection having been performed with only omega scan. This feature is also reflected in the low value of the refined Flack parameter in 1b (-0.264(5)) but, although this value does not permit unambiguous determination of the absolute structure; its absolute configuration is established by reference to an unchanging chiral centre in the synthetic procedure.

RESULTS AND DISCUSSION
Metallo-porphyrins recrystallization method. The free porphyrins 4R-H 2 PPP and H 2 PPP (Figures S1-S2) were reacted with Zn II salts and the resulting coordination compounds (Zn(4R-PPP), 1, and Zn(PPP), 2) were recrystallized by slow liquid-liquid diffusion method using CH 2 Cl 2 /CH 3 OH and CH 2 Cl 2 /hexane mixtures, respectively (Scheme S1), when in a final step the samples were left open to air and most of the solvent evaporated. As a result, we found out crystals for all the combinations. 1a and 2a for the former and 1b and 2b for the second, respectively. It has been postulated that solvent mixtures containing water rich solvents can guide self-aggregation processes. 25,26 However, here additional factors require our attention as (i) the non-innocent action of the side groups of the porphyrin cores (noncovalent coordination), (ii) their chiral and non-chiral nature, and (iii) the different coordination possibilities of the Zn II ions; altogether they trigger the final structures (1a-b, 2a-b).
Structural Descriptions. Table S1 shows general crystal data information of the four Zn II -porphyrin species (1a-b, 2a-b). Selected bond lengths and angles for each system are listed in Tables S2-S5, respectively, as well as additional Figures in S3-S6 displaying different projections. Overall, the core, for all the porphyrin units shown in 1ab and 2a-b is identical, described as a Zn II ion coordinated to the N atoms of a porphyrin ring. On the other hand, the main difference between single molecules resides in the chiral center on the peripheral groups in the formers, as Scheme 1 shows. Here we focus on the major differences between the four species and provide basic molecular descriptions but concentrate in the supramolecular arrangements of all the systems.
[Zn(OH 2 )(4R-PPP)]. 1a crystallizes in the orthorhombic P2 1 2 1 2 1 space group. The mononuclear species contain one penta-coordinated Zn II center bonded to four N atoms from the porphyrin core and one O from a H 2 O molecule ( Figure 1). The Zn II site is above the porphyrin plane (by 0.207 Å) toward the axially bonded H 2 O. Such binding is nearly perpendicular to the plane of the Zn-porphyrin core (O-Zn-N angles between 89.85 -101.43 °) with a Zn-O bond length of 2.247 Å. This meso-Zn II -porphyrin displays four identical substituents containing chiral ester groups (C45, C49, C53, C57 in Figure 1), all of them with an absolute R configuration. These peripheral moieties, from the 4R-PPP 2-  The large size of the crystals obtained for 1a (see inset in Figure 2 and Figure S10) permitted positioning them in the stage of an AFM equipment (see physical measure-ments) to check the quality of the crystal surface. The surface of this particular 3D crystal as measured by AFM is shown in Figure 2a. It consists of large and atomically flat terraces (several micrometers long and hundreds of nm wide in average) separated by well-defined steps. As it can be extracted from the line profile in Figure 2b, the steps heights are  1.8 nm, which fully agrees with the crystallographic parameter b and multiples of it (see details of the 3D structure of 1a in Table S1 in the SI). [Zn(4R-PPP) n ]. 1b crystallizes in the orthorhombic P2 1 2 1 2 1 space group. 1b is a coordination polymer made by the self-assembly of the same porphyrin molecule described in 1a (same porphyrinic ring and chiral branches, Figure 1). The difference now resides in the fact that the Zn II center of each porphyrin unit is not coordinated to a molecule of H 2 O but to the C=O group of an ester branch from the adjacent molecule. Such coordinative bond provides the final polymeric structure depicted in Figure 3. Therefore, each monomeric unit shows a pentacoordinated Zn II center, with the metal slightly shifted up from the plane of the porphyrinic core (by 0.188 Å). Here, the organic moiety, 4R-PPP 2-, presents sprains, where one of the phenyl rings is bended up with respect to the others. The Zn-O bond length distance is now of 2.210 Å, shorter than the Zn-O distance shown in 1a (Zn-OH 2 ), with a noticeable deviation from previous perpendicular disposition (O-Zn-N angles of 88.60 °, 88.72 ° and reciprocal). As it happened in 1a, only one C=O from a chiral ester group interacts with the neighbor molecule, leaving free the other three. The final arrangement of a single 1D coordination system is shown in Figures 3 bottom and 4.
From our best knowledge, system 1b is the first chiral coordination polymer of porphyrin nature that is formed by the coordination of one branch of the porphyrin system and one metal of the nearby molecule and the first one displaying chiral properties. Other two coordination polymers have been described in the literature where the metallo-porphyrin units are connected by means of coordinative bounding of the metallic center and groups from the neighbors, providing also 1D systems. 27,28 However, the ones described in the past, present always the coordination of two branches of the same monomeric porphyrin unit with two others. 29,30   [Zn(PPP)]. Compound 2a crystallizes in the C2/c monoclinic space group and the discrete unit consists exclusively in one molecule of the Zn II -porphyrin moiety. Here, the Zn II center is tetra-coordinated and adopts a square planar geometry, forming a perfect plane with the four porphyrin nitrogen atoms of the core. The planarity of the PPP 2ligand does not present any deformation, in contrast with what it happens in the units of 1b. As Figure 5 (Top) shows, the structure has some resemblance to that describe in 1a, due to the similarities of the peripheral groups, but now the replacement of the CH 3 group in all four branches by a proton lets to achiral molecules. Although the compacted packing of the molecules shows proximity among neighbors, the Zn II center in 2a is further away from the C=O moiety (3.333 and 3.908 Å) and no molecules of solvent/H 2 O appear in the final supramolecular architecture (Figure 5, Bottom). [Zn(PPP) n ]. 2b crystallizes, as the previous, in the monoclinic space group C2/c. The asymmetric unit is described by a molecule of "Zn(PPP)O 2 ", where the oxygen atoms relate to the C=O groups of two neighboring molecules (Figure 6, top). The Zn II center is therefore hexacoordinated, with a pseudo-octahedral symmetry due to its coordination to four nitrogen atoms from the PPP 2organic moiety and to the two oxygen atoms already mentioned. As it happens in 2a, the Zn II center forms a perfect plane within the chromophore core. In the rearrangement, the oxygen atoms are tilted from 90 ° (84.27 °, 85.64 ° and reciprocal). Each ZnPPP is attached to four other units creating 2D networks with sql topology, where the Zn II -porphyrin molecules present alternating orientations and one single Zn···Zn distance of 13.663 Å (Figure 6, bottom). The structure grows layer by layer, where the adjacent 2D networks are parallel to each other with small interactions between them through the branches that are not involved in coordination ( Figure S9).
Altogether, we found two Zn-porphyrin setups (1 and 2) with great ability toward crystallization, and confined the combination of solvents for such task (CH 2 Cl 2 /CH 3 OH and CH 2 Cl 2 /hexane) to investigate chirality and polarity. By doing so, we created a map with a rich variety of supramolecular arrangements. Having a Zn II center and four terminal benzylic ester moieties at the meso-positions, both electron-acceptor and donor parts in that order, we expected self-assemble through coordinative bonds, fact that we see in two of the systems under study, 1b and 2b. However, the final picture is more complex, where the coordination number of the Zn II metal center varies from four to six, unlikely to anticipate. Single crystal X-ray diffraction shows that in the case of the chiral systems, 1a and 1b, the Zn is always pentacoordinated, choosing as the fifth ligand a molecule of H 2 O or the terminal benzylic branch of one neighboring species, respectively. The same methodology applied to the non-chiral versions provides tetra-and hexacoordinated Zn II centers, 2a and 2b, respectively. In all cases, the size and geometry of the crystals for all the four samples differ ( Figure S10). We could argue that the changes in solubility regarding the existence of a -CH 3 vs. a -H (1 vs. 2) in the winds of the porphyrins, makes unfeasible a rigorous comparison but the two methodologies used work in a similar manner in the four systems, having crystals in all the cases after few days. Therefore, chirality may be involved in the results; where the -CH 3 of the coordinated branches are always outside, improving the disposition of the C=O group toward the Zn II centers ( Figure S11).
On the other hand, the polarity of the solvent provided us with additional features, where the use of CH 3 OH as a precipitating agent gave always isolated molecules (0D). Here, the intermolecular associations within discrete porphyrin coordination compounds rely on H-bonds involving coordinated water molecules (1a), or C-H…π interactions (2a). Instead, hexane always provided compact supramolecular moiety, although compounds 1b and 2b present different arrangements. This seems to indicate that the most apolar solvent forces aggregation of the molecules that re-organize through the metallic center and moieties on the arms. The differences in the selfassemble however, must be related to the affinity among the molecules and therefore to the nature of such branches.
In addition, studies in solution by 1 H NMR and UV-Vis absorption spectroscopy in CD 2 Cl 2 (Figures S12-S14) of all Zn II -porphyrins showed isolated porphyrin units, chiral (Zn II (4R-PPP), 1) and achiral (Zn II (PPP), 2), with no additional info regarding aggregation even at different concentrations ( Figures S12, S13). In the case of 1, the absence of CD signals using the same solvent was expected, due to the distance between the chromophore core and the chiral centers and the freedom of rotation of the latest. 31 MALDI experiments of 1a and 1b in CH 2 Cl 2 provided almost identical spectra, showing one m/z signal with maximum at 1084.29 and isotropic patterns that matched well with the existence of [Zn II (4R-PPP)] units. Additional experiments were performed with samples 1a and 1b to contemplate the origin of the coordinated H 2 O in the case of 1a, missed in 1b. For that, we used dehydrated Zn acetate salts to achieve both systems following the same methodology explained above. Our analyses showed that the final compounds, 1a and 1b, were totally reproducible, being the H 2 O molecules from crystal processing, therefore contained in the solvents used to recrystallize the samples (Figures S15-S17).
In a further step to characterize in detail the systems and chiral nature in the case of 1a and 1b, we gathered information with the use of current and specialized solid-state techniques, having the advantage of correlating the changes to the structures described above. The following paragraphs describe the outcome encountered and our impressions regarding their practical use.
Solid state studies. In a first step, we performed XRD analyses on different crystalline batches for each compound (1a-b, 2a-b) for phase identification and consistency of the final structures. Figures S18-S21 depict the diagrams found experimentally for the four species and the comparison with their simulated spectra, respectively. 1a-b and 2a presented reproducible patterns that allow us to use crystalline samples toward additional characterizations in the bulk shown below. However, crystals of 2b were scarce. Attempts to recrystallize such solid showed PXRD patterns that differed from the expected from the powder pattern of 2b; this was a drawback for the rest of the analyses in the solid-state where high amounts of samples were required. Hence, the comparison of 2b with the rest of systems was limited to the crystallographic information.
Following with our analysis, ATR-FTIR provides us clear evidences of the changes shown in the crystal structures. In the case of the chiral systems, 1a and 1b, the two molecules present almost identical ATR-FTIR spectra but differing in two areas between 3450-3470 cm -1 and 1720-1760 cm -1 ( Figures S22). The former area shows the appearance of a broad peak in the case of 1a missing in 1b that may correspond to the inserted H 2 O molecule as the structure shows. On the other hand, the former area relates to the C=O stretching vibrational modes from the terminal ester groups of the four branches. In the case of 1a, appear two peaks at 1732 and 1750 cm -1 but 1b shows three peaks, at 1719, 1739 and 1757 cm -1 , agreeing well with the fact of having one ester group coordinated, different from the rest. Regarding the same area, compound 2a displays a single peak with a shoulder at 1752 cm -1 ( Figure S23) and in agreement with the absence of different coordination on the branches with Zn II centers.
Our studies show that the existence of different supramolecular formations has scarce effect on the solidstate UV-Vis absorbance of the final systems, remaining all, chiral and achiral, very similar as it happens in solution ( Figure S25).
On the other hand, high resolution solid-state 13 C NMR spectroscopy was one of the most sensitive techniques for our studies; Figure 7 shows the spectra of 1a, 1b and 2a (4R-H 2 PPP is shown in Figure S24). The general assignments of the chemical shifts were made by comparison with the free achiral porphyrin and the contrast among the coordination compounds (1a-b and 2a) and reported literature. 32 From the latest, similarities of our porphyrins to the 5,10,15,20-tetraphenyl porphyrin Zn II compound, studied by Grant et al. 33 allowed us to interpret the nature of most of the chemical shifts by their range of appearance although specifics about individual shift assignations were not possible to elucidate.
We found out that multiple signals appeared in the range of 180 to 0 ppm for compounds 1a and 1b and from 180 to 30 ppm for 2a. The shifts could be grouped in four areas: signals appearing between 180-165 ppm (A), 165-125 ppm (B), 125-100 ppm (C) and from 80-20/0 (D), from down fields to high ones (Figure 7). "A" encloses C3 (Figure 7), related to the C=O part of the ester. "B" is the sum of C para , C  , C ortho , C i and probably part of the C  shifts. "C" would include C  , C meso and C meta and finally "D" will rely to the non-conjugated part of the molecule (branches) therefore to C1, C4 and C2, from down to high fields, in that order. The absence of C2 in 2a (Figure 7) corroborates our assignation in the other two systems (1a and 1b). The variation in the number and shift of the chemical signals differ, from one sample to another, in a complex manner. It is already established that such variations can relate to overlap of the signals, small structural differences inside the molecules (e.g.: bendings provide different number and shifts for C  , C  , etc.) and to the proximity of neighboring molecules (e.g.:  electron cloud of the core) providing different environments to most of the groups. 34 Comparing 1a and 1b, the multiple chemical shift corresponding to C3 of 1b suffers (at least one over four per molecule) such alterations in a strong manner than the others. The final shape of these signals is different in both and shifted to higher fields for 1b (172.5 ppm (1a) and 170.7 ppm (1b)). The shift and shape of C1 also clearly differs in all three systems showing the dramatic effect that different environment has on the chiral/achiral carbons.
Chirality. Taking advantage of the chiral nature of some of the samples, we performed solid-state circular dichroism studies of 1a and 1b using a KBr matrix due to the absence of chiral response in solution. The key aspects for finding optimum conditions in the achievement of the spectra were described elsewhere 35,36 and detailed information regarding our procedure here is described in the SI.
In the past, some of us studied the correlation of experimental Cotton effects (CE) with the conformational stereoisomerism of chiral systems in the solid state. 37 Our objective here was to establish the relationship between the CD spectra of the compounds and their supramolecular conformations, comparing systems with identical chiral species. Figure 8 top shows the comparison between the two species (1a-b) under such conditions and Figure 8 bottom emphasizes the crystallographic differences between the two compounds.
Taking advantage of the collected X-ray diffraction data, we now could match structural differences between the molecules with changes in the CD spectra. This way, the R-nature of the ligand, for 1a shows a bisignate signal displaying a negative CE at the lowest wavelength and a positive Cotton effect of similar intensity at higher wavelengths. This matches well with the position of the Soret absorption band of the porphyrinic chromophor in solid and solution (Figure S29), where the presence of bisignate CD may also relate to the orientation of the porphyrin cores in a sliding face-to-face creating a 1D supramolecular structure through the H 2 O interactions ( Figure 1 bottom and 8 bottom). In the solid state, compound 1a displays two intense and proximate CD bands of opposite sign (positive-negative; positive chirality) 38 between 375 and 500 nm (black solid line) centered at 431 and 444 nm, respectively. In the case of 1b (dotted line), CE are observed in the same range, although now three CD signals appear (positivenegative-positive) being the one in the middle the most intense.
The above features are present in the CD spectrum of 1b too, however, this system differs from the previous by displaying a third positive CD signal, absent in 1a, headed by the already mentioned strong positive-negative bisignate CD sign. The negative Cotton effect is here more pronounced than for 1a. Such results could be associated to the expanded 1D coordination network of 1b, by the coordination of the Zn II centers with CH 3 CO 2 -R moieties (instead of molecules of H 2 O (1a)) and differences in the tridimensional packing between the two systems (Figure 8 bottom). 39 The first positive CD signals (highest wavelengths) for the two systems, 1a and 1b, with maxima at 444 nm and 448 nm, respectively, follow similar trends, and intensity, having again close relation to the Soret band and therefore intrinsic structure of the otherwise identical porphyrinic units.
Finally, we analyzed the chiral systems using solid-state vibrational circular dichroism (VCD). VCD has great advantages toward the detailed analysis of molecular conformations because the number of molecular vibrations, in the IR region, and the sensitivity of the technique are larger than the electronic transitions in the UV-Vis region. 40 Nevertheless, working in the solid state one must be extremely careful minimizing the spectral artifacts. For that, we took into account the noise in each experiment, the signals found in the non-chiral system 2a ( Figure S30) and performed theoretical calculations on 1a ( Figure S31). The workout of the material and procedure is described in the SI. Figure 9 shows the comparison between the IR and the VCD spectra of both systems. As expected, each absorbance band in the IR spectrum has a correspondence with a VCD band. DFT calculations taking into account the single molecule 1a (the length of the chain of 1b difficult the study) displayed a number of motifs in the window under study in a similar manner although few shifted from the experimental values displaying, at times, opposite sign from the observed. Albeit contradiction, such results point out that the chiral response of the system involves the surroundings, where the supramolecular network is the responsible for the chiral phenomenon. 41 The measured and calculated VCD spectra of each system match well with those found in the IR experiment corresponding to 4R-systems (intensities are not comparable). The bands described between 1800-1600 cm -1 vary between the two species, as anticipated due to the difference upon coordination of the Zn II centers of the porphyrinic moieties with H 2 O (1a) or some C=O groups (1b). In this region, the number of VCD bands found in this region for 1b is higher than 1a, with some of them shifted to lower energies, probably due to the difference between the C=O groups bounded with respect to the non-bounded ones and additional intermolecular interactions in the tridimensional map of each system.
Overall, all the techniques used in the solid-state studies agree with the structures but emphasize their restricted outcome individually.
Surface studies. We performed SEM images of the samples prepared by drop casting solutions (10 -3 M) of the chiral (Zn(4R-PPP), 1) and achiral (Zn(PPP), 2) porphyrins in the same solvents mixture used to achieve 1a-b and 2a-b: CH 2 Cl 2 /CH 3 OH (1:1) or CH 2 Cl 2 /hexane (1:1), respectively, onto mica or highly oriented pyrolitic graphite (HOPG) substrates. Here, the effects of solvents and surfaces were studied by the comparison with the morphologies found in the single crystal X-ray diffraction experiments. A variety of Zn(4R-PPP) and Zn(PPP) nanoarchitectures were obtained by an evaporationdriven self-assembly process ( Figure 10). The chiral and achiral porphyrins generate reproducible microvesicles, microrods flower-shaped aggregates and microplates bundles shapes, depending on the conditions used (Figures 10a-d and S32-S41).
Hollow microvesicles appeared by dissolving 1 in CH 2 Cl 2 /CH 3 OH and depositing it on mica (Figures S32) and graphite (Figure 10a and S33). On mica, the size and shapes of the aggregates were the most irregular, presenting sometimes sectioned microvesicles; instead, on graphene the microballs presented an average size of 600 nm with hole diameters of ~300 nm. Overall, the vesicles in the latest were distinguishable and presented homogenous shapes, where some of them connected to others through small cords. In CH 2 Cl 2 /CH 3 OH, hollow vesicles made of Zn-porphyrin systems have been related to the - stacking of the porphyrin rings and formation of J-type aggregations. 42 Guided by the crystallographic data of 1a, we can relate our porphyrin microvesicles to the ligandassisted long-distance J-type formation 43 In our case, the supramolecular 1D structure of 1a, with up-down distributions of the porphyrins cores (Figures 1-bottom and  S33), differs from the regular staircase situation described in previous works. 44 The general explanation for the holes formation, of encapsulation of the volatile solvents inside the vesicles and later evaporation, is also applicable here.
Repeating the same exact experiment but using 2, the architectures found on mica and graphene were totally different than before. Now, on mica well-shaped microcrystals of different sizes presented parallelogram and prismatic geometries (Figures S35). On graphite, fourpointed star shapes were the most common morphology (Figures 9b) although bigger and amorphous aggregates were seen too ( Figure S36); all shapes intuitively relate to the layer-by-layer aggregation present in the structure of 2a (Figure 4-bottom) and to the formation of shortdistance J-type aggregates (staircase disposition). Moving to CH 2 Cl 2 /C 6 H 14 , with 1, on mica, we observed again microballs, some of them with holes, together with slide microvesicles ( Figure S37). The difference with the previous resides in that most of the microball edges smear, connecting each other, probably because of solvent flow and the longer sintering state during the evaporation process. The shape of the nanostructures agrees well with 1b, where the 1D coordination structure provides J-type aggregates (up-down structure, Figure 3bottom). On graphite, two distinctive morphologies were observed depending on the area of analysis. This way, some of the parts presented microvesicles with sizes between 300-400 nm (Figures S38). The topographic information in such cases was very similar to that found in CH 2 Cl 2 /CH 3 OH; however, trigonal prisms were observed mostly at the surface edges (Figure 10c). They presented homogeneous sizes, with a length of 8 m, and spherical defects in some of the prismatic faces ( Figures  10c and S39). Such prismatic microcrystals show similarities with those at the macroscopic scale used for the Xray diffraction analysis ( Figure S10). They were in areas where the evaporation process is slower, providing more time for the ensemble and organization of the matter. This was established by the fact that few of them appeared cracked exposing the aggregation of microvesicles like expanded polystyrene bowls (Figure 10c).
At last, the same mixture using 2 on mica showed microplates, with diverse shapes and sizes, together with flower-shaped topologies ( Figure S40). In former works, the formation of the latest relates usually to H 2 O. 45 Here this association cannot be discriminated but neither defended. However, in both cases, in microplates and micropetals, it is clear the existence of an extended face versus the others, connecting with the 2D layers observed in the 2b system typical in other layered materials as well. 46 Regarding the experiment in graphite, now multiple microblocks were present again with a variety of sizes and prismatic shapes (Figure 10d and S41). The size may depend on the deposition and evaporation processes, where the concentration of the material can also rely on the different terraces of the graphite surface. Either way, this result differs with the macroscopic scale in solution, where the crystals were scarce and difficult to achieve, pointing out the tendency of graphite to provide crystalline material under these conditions.

CONCLUSIONS
Using metallo-porphyrin derivatives as unit models, we described how through a unique molecular design, we created well-defined and isolable nanostructures; where, the tuning of solvent conditions directly affected the selfassembly process, and therefore the morphologies of aggregates. Our results emphasize the relevance of porphyrin crystallization toward their future applications and face the present challenges, as the lack of anticipation of the final architectures. The addition of a metal center increases the range of diversity in the coordinative way meanwhile blocks strong - stacking interactions among the porphyrin cores giving priority to the nature of the arms, now key parts in the creation of the supramolecular arrays. The four arms are flexible enough and the carbonyl moieties, from the ester groups at the edges, present the freedom to attach the Zn II centers in different manners where the final coordination number of the metallic center can go from four to six, depending on the overall conditions. The comparison between chiral metalloporphyrin (1) and non-chiral (2) one, repeating exact conditions, provides different crystallographic species, pointing out the complexity of adding chirality among the other factors already mentioned (coordination and solubility). The nature of the precipitant solvents, from polar (CH 3 OH) to apolar (C 6 H 14 ), can promote the stabilization of mononuclear entities, like in the case of former, in both, chiral and non-chiral metallo-porphyrins (1a and 2a, respectively). In addition, 1a shows a quite distinctive 1D supramolecular structure. Hexane promotes aggregation in the chiral and non-chiral systems, having at the end coordination 1D structures (1b) or 2D systems (2b), respectively.
Regarding characterization in the solid state, 13 C NMR showed highly sensitivity to identify small but critical changes in the final conformations. On the topic of chirality, CD and VCD solid-state experiments presented differences between 1a and 1b, stressing their sensitivity as well, although the variations in the former were not intuitive and difficult to understand without the assistance of the crystallographic data. In the case of the second, it presented relevant changes in the area related to the C=O stretching, in agreement with the coordination to Zn atoms. Yet, our results emphasize that the proper analysis of the structures by the exclusive use of such techniques requires the creation of extended libraries and corroboration from theoretical calculations.
Finally, the studies of our systems on two different substrates show the enormous effect of the surfaces on the final structures, providing a variety of vesicles, flowershapes and well defined geometrical architectures, increasing at times the capacity of achieving microcrystals (e.g.: 2b in graphite).