ARTICLE pubs.acs.org/Langmuir

Charge, Stereochemistry, or Epitaxy? Toward Controlled Biomimetic Nucleation at Mixed Monolayer Templates Ben Stripe,*,† Ahmet Uysal,† Binhua Lin,‡ Mati Meron,‡ and Pulak Dutta† † ‡

Department of Physics & Astronomy, Northwestern University, Evanston, Illinois 60201, United States Center for Advanced Radiation Sources, The University of Chicago, Chicago, Illinois 60637, United States

bS Supporting Information ABSTRACT: Floating monolayer mixtures of cationic dioctadecyldimethyldiammonium bromide and anionic lipids were used as variable templates for the biomimetic nucleation of calcium carbonate and studied using grazing incidence X-ray diffraction. Varying the ratio of constituents changes the monolayer charge, structure, and molecular tilt. The nucleating surface of calcite also changes as the mixture is varied, and at an 80:20 ratio the (012) face is seen under a floating monolayer template for the first time. Our results indicate that the average template lattice is the major controlling factor in the oriented nucleation of CaCO3. This is in contrast to the current view that the orientation is controlled by the stereochemical matching of the terminal functional group and molecular tilt with respect to the carbonate groups in the crystal.

’ INTRODUCTION Understanding biomineralization, and using this knowledge to manipulate and control the nucleation, orientation, and morphology of minerals, is an exciting and challenging task. Despite years of multidisciplinary research that brings together the techniques and approaches of biology, chemistry, condensed matter physics, and surface science, the mechanisms of oriented crystallization in biology are not well understood. It is widely accepted that biomolecules play controlling roles, but the exact mechanisms remain unclear. Research using monolayer films, both self-assembled alkanethiols (with various terminal groups) on gold/silver substrates and floating Langmuir monolayers, offers better defined model systems with which to elucidate the roles of various potential factors controlling oriented calcite nucleation. This research has shown a compelling link between carboxylic, alcohol, and sulfate functional groups and the morphology and orientation of nucleating CaCO3 crystals.1 However, the underlying mechanisms are less clear. It has been proposed that nucleation of inorganic crystals such as calcite at organic templates is driven by the following: (a) Epitaxy, lattice matching between the organic template and the nucleating face. This is a long-standing and much discussed hypothesis.1 5 (b) Stereochemical matching. For example, if the terminal functional group of the template aligns to mimic the carbonate groups in a given crystal plane, the crystal grows with that plane oriented along the organic surface.6 9 (c) Charge balancing, where the nucleation plane of the crystal is chosen to balance the monolayer charge.10 12 There is significant evidence for each of these viewpoints.13 In recent years, interface X-ray scattering has established that lattice r 2011 American Chemical Society

matches exist in some cases,2 but in other cases there is no identifiable lattice match.6,7 Aizenberg et al. reported [012]oriented calcite growth on carboxylic acid terminated alkanethiols on silver substrates and [105]-oriented growth on carboxylic acid terminated alkanethiols on gold substrates.6 These results were interpreted in terms of the tilt angle of the terminal carboxyl group and its orientational match with the carbonate groups of the nucleating calcite crystal. Prior work on the self-assembled carboxylic acid terminated alkanethiols on silver14 reported that the molecules form hexagonal lattices with a spacing of 4.77 Å and tilt of 12° (however, the tilt angle has been challenged15,16 and may be ∼0°). On gold substrates, the same molecules form 5.00 Å hexagonal lattices with a tilt angle of 30°.14,15 Further evidence for the role of terminal group orientation comes from examining the effects of varying the chain length (even and odd parity backbones terminate at different angles). Han et al. noted [012] growth on silver for both even and odd parity, and on gold they noted [013] for odd and [113] for even.8 However, other work has shown both [015]- and [012]-oriented growth on gold-supported carboxylic acid templates.17,18 Travaille et al. have suggested this is seen because changing the pH affects the headgroup orientations.19,20 More recently, Duffy et al. have shown that the presence of bicarbonate ions may be the cause.21 In general, there is some variation in the orientations found on gold that is not completely understood, while the results with silver are more consistent. Pokroy et al. have examined the strain induced by the lattice mismatch between the templating Received: July 5, 2011 Revised: November 8, 2011 Published: November 11, 2011 572

dx.doi.org/10.1021/la2037422 | Langmuir 2012, 28, 572–578

Langmuir films and the oriented crystals, suggesting that mismatch between lattices causes macroscopic crystal strain, but its role is secondary to that of the alignment between the terminal functional groups and the carbonate atoms in the crystal.7 Floating Langmuir monolayers have been used as nucleation templates as well,1,22 and in these systems it is possible to directly probe both the organic monolayer template and growing crystals using grazing incidence X-ray diffraction.2,9,23 26 Studies using Langmuir monolayers have shown far less breadth of stable crystal orientations than studies using self-assembled films. Simple carboxylic acid monolayers are seen to form an untilted centered rectangular phase (a = 5 Å, b = 7.5 Å) which corresponds to 18.7 Å2 per molecule compared to 19.7 Å2 per molecule on silver substrates and 21.6 Å2 per molecule on gold substrates.24 The crystals grown under floating carboxylic acid monolayers are thought to begin as [1̅ 10]-oriented but quickly grow into large irregular crystals.1,27,28 Both Langmuir monolayers and self-assembled films contain carboxylic acid terminated alkane chains; in one system the tails are firmly tethered, and in the other system the chains are compressed latterly. Ideally the carboxylic ends of the chains in either system should appear very similar as seen from the aqueous solution, but the same control of oriented growth has not been demonstrated in both systems. Recent studies have reported that disordered Langmuir monolayers maximize the uniformity of the crystal morphology across the monolayer.9,29 While these disordered templates do an excellent job of nucleating a fairly uniform truncated rhombohedral morphology, the nucleated crystals have multiple orientations. Fixed templates with well-defined order have shown better control of both uniform morphology and uniform orientation, despite the fact that in most cases there is no direct evidence of epitaxy.2,6,20,28 As the next logical step, this paper moves from fixed templates to variable templates. Instead of studying templates with certain set spacings, charge densities, and stereochemical tilt angles, and extrapolating from these limited data, we devise a template that can be varied nearly continuously and can be probed in situ. Ideally such a template will be able to nucleate multiple orientations and will give us a better understanding of what factors are dominant in controlling the oriented growth of CaCO3. Recent research18,30,31 has shown that the concentration of common membrane components such as cholesterol can stably alter the average membrane lattice. Such mixing was shown to be nonideal, but was still able to nearly continuously alter the lattice spacing. Other studies have shown that mixed monolayers of dodecyl sulfate and dioctadecyldimethylammonium bromide (DODAB) can form extremely stable monolayers capable of folding rather than collapsing.18,32,33 The stability and miscibility of the mixed DODAB monolayers makes them ideally variable templates. Therefore, mixtures of positively charged DODAB with negatively charged arachidic sulfate and heneicosanoic acid were used in the present study as templates for the growth of calcium carbonate. The monolayer structures were determined in situ by grazing incidence X-ray diffraction (GID). The resulting crystals were analyzed both in situ and ex situ by GID and scanning electron microscopy, respectively. It should be noted that GID is an averaged method and is essentially blind to localized variations that may occur at the nucleation site. However, in designing a template, we can only control the average and not the localized order. The previously reported literature and our own studies reported below make clear that the orientation of calcium carbonate is responsive to

ARTICLE

Figure 1. Structure of the three organic molecules used in this study and representation of the mixed monolayers.

changes in the average template structure; otherwise, local restructuring would always give rise to the same orientation.

’ EXPERIMENTAL SECTION All chemicals were purchased through Sigma-Aldrich and used without further purification. Heneicosanoic acid (C20H41COOH) (referred to below as “acid”) was prepared as a 100 mL bulk solution in chloroform at 1.36 mM. DODAB ([CH3(CH2)17]2N(Br)(CH3)2) was prepared as a 100 mL bulk solution in chloroform at 1.5 mM. Arachidic sulfate (C20H410S03Na) (referred to below as “sulfate”) was prepared in a bulk solution of a 1:1:1 mixture of methanol chloroform cyclohexane at 1.5 mM. The chemical structure of each molecule can be seen in Figure 1. Mixtures were then made in 10 mL flasks by combining 9 mL/1 mL, 8 mL/2 mL, etc. The CaCO3 subphase was made using Kitano’s method.34 To prepare the subphase, 2 g of CaCO3 was added to 1 L of pure H2O (18.2 MΩ cm) and the water was then bubbled with bone-drygrade CO2 for 1 h, filtered through a 0.2 μm filter, and rebubbled for 30 min before being placed in the trough. The final subphase had pH ≈ 5.8 and a concentration of just greater than 7 mM Ca2+ as reported previously.2 The monolayers were spread immediately after the solutions were placed in the trough and compressed to a 10 N/m surface pressure at 20 °C. Monolayer GID data were collected at ChemMatCARS Beamline 15ID of the Advanced Photon Source at Argonne National Laboratory directly after the sample was placed in the trough. X-rays with a wavelength of 1.240 Å were incident at an angle of 0.09° on the sample trough, creating a beam footprint 2 mm wide by ∼9 cm long. A Pilatus 100 K detector was used in a pinhole geometry (see the references for more details).35,36 A horizontal wave vector resolution of ∼0.01 Å 1 was defined by the distance between the horizontal slit and detector along with the detector pixel size. The vertical resolution was defined by the beam width and the angle from the water surface to be ∼0.075 Å 1 at worst in the data recorded, but substantially better at lower Q z. Data were collected over the course of ∼4 h per sample or until the crystallization was detected by the appearance of crystalline peaks in the data. Beam damage (especially to the amorphous precursor) was mitigated by the translation of the motorized trough by more than the 2 mm beam width after each scan. For more information about the setup, please refer to the references given.35,36 Crystal growth GID data were collected at Beamline X14A of the National Synchrotron Light Source, Brookhaven National Laboratory. This beamline was used due to its larger available beam footprint and larger beam divergence, which gave us better statistical averaging of the growing crystals (see refs 28 and 37 for more details). All data were taken after 12 h of in situ crystal growth time. There was no exposure to X-rays during this period to avoid any damage to the monolayers or to any amorphous layers formed in the early growth stages. The beam was 573

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578

Langmuir

ARTICLE

Figure 2. GID data collected for the 70:30 and 80:20 acid DODAB samples. The 70:30 sample shows both an in-plane Qxy = 1.435 Å 1 peak and an out-of-plane Qxy = 1.46 Å 1, Qz = 0.52 Å 1 peak. The 80:20 sample shows only an in-plane Qxy = 1.52 Å 1 peak. incident in a grazing incidence geometry at a wavelength of 1.3376 Å on a point detector with a resolution of Qxy = 0.01 Å 1 and Qz = 0.05 Å 1 defined by crossed soller slits. The incident angle was approximately 0.002 rad, just below the critical angle of water. Crystals for scanning electron microscopy (SEM) imaging were grown for 24 h before being collected on silicon wafers by surface tension transfer from the surface of the water. Surface tension transfer is accomplished by dipping a clean glass substrate vertically into the pressurized trough, so when it is withdrawn, the wet crystals transfer to the glass substrate and are held there by the surface tension of water. Samples were then dried under a stream of nitrogen. The samples were then sputter coated with 10 nm of Au/Pd. Images were captured on a Leo Gemini 1525 SEM instrument at 3 kV. During the crystal nucleation, after compression of the monolayer to 10 N/m, samples were grown both at constant pressure and at constant area, and no significant difference was seen.

Figure 3. Top: schematic representation of the titled and in-plane monolayer lattices. Bottom: magnitude of the monolayer in-plane lattice components and the tilt angle as a function of the percentage of acid in the monolayer. Regions giving different crystal growth are noted by arrows spanning the region and labeled by the [hkl] indices of the crystal orientation normal to the water surface.

instead the crystals have adopted a nearly triangular wedge morphology, but still display irregular edges similar to those of the previous crystals. Figure 4c shows crystals grown under an 85:15 mixture: the crystals are now clearly triangular wedges with visible {104} edges resting on a {104} base. The growth face is irregular, but the appearance of a new smaller {104} face on the top surface of the crystals (see the inset) indicates a shift in orientation from the previous concentration. Figure 4d shows crystals grown under the 80:20 mixture: the crystals now appear as discrete triangular crystals with a very thin irregular nucleating plane. The inset image is of one of the crystals resting on its nucleating plane (so that the nucleating plane is not visible against the silicon substrate). This image is quite similar to those of [012]-oriented crystal growth previously reported on carboxylic acid terminated alkanethiols on gold substrates.6 All samples contained greater than ∼70% oriented crystals by simple examination under an optical microscope, and the crystalline growth density was fairly uniform across each of the samples. Image e shows {104} rhombohedral crystals. This image is typical of samples grown under all mixtures containing 70% acid or less. The inset in image e shows rhombohedral crystals with clearly cropped corners, suggesting that the crystals are growing oriented, although the planes are too small and irregular to classify easily. However, by the time the monolayer is a 50:50 ratio, the crystals are simple {104} rhombohedral. As the mixtures approach pure DODAB, the samples aggregate more and more. The final image, f, shows the aggregated {104}oriented morphology of a crystal grown under a 40:60 monolayer. It should also be noted that there were traces of vaterite and some bulk {104} rhombohedral in all samples.

’ RESULTS Heneicosanoic Acid DODAB Mixtures. The results for heneicosanoic acid DODAB mixtures are presented in Figures 2 5. Figure 2 shows the raw data for the 80:20 and 70:30 mixtures; other raw data can be found in the Supporting Information. The data are also summarized in Table 1. Figure 3 (top) shows a representation of the tilted tail group lattice labeled in [hk] notation and the corresponding in-plane hexagonal/ pseudohexagonal lattice components a and b. Figure 3 (bottom) shows a plot of the in-plane lattice components and tilt angle at 10 N/m as a function of the heneicosanoic acid fraction present. Table 1 lists the GID data in more detail. There are three distinct regions in Figure 3, consisting of untilted hexagonal monolayers for 80 95% acid, a region of rapid tilt change located between 65% and 80% acid, and a ∼30° tilted nonhexagonal (orthorhombic or centered rectangular) phase at lower acid fractions. In no case do we see more than two first-order peaks, which would indicate coexistence of separate phases. The structure observed is of the mixture and not of the individual constituents. Figure 4 shows SEM images of samples collected by surface tension transfer onto silicon substrates. Image a shows CaCO3 grown under pure acid oriented with the [1̅ 1 0] direction normal to the surface as reported previously.1,27,28 Image b shows crystals grown under a 90:10 acid DODAB mixture. The crystals no longer grow as large or irregularly as under pure acid; 574

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578

Langmuir

ARTICLE

Figure 4. SEM images of crystals grown under mixed monolayers of acid DODAB: (a) 100:0; (b) 90:10; (c) 85:15; (d) 80:20; (e) 70:30; (f) 50:50. Images similar to (f) are seen all the way from 50:50 to pure DODAB.

a tilted state. It is interesting to note, however, that no multilayer peaks are seen, unlike for pure arachidic sulfate on water.2 Figure 7 shows SEM images taken of various sulfate DODAB mixtures at 10 N/m. Figure 7a shows the typical trigonal pyramids formed under arachidic sulfate.1,2,23,27 Figure 7b shows crystals grown under an 80:20 mixture. The crystals are once again trigonal pyramids; however, the nucleating planes now show an inset triangular ridge and the expression of small {104} faces at their points. Figure 7c (70:30) shows a nearly complete change in morphology. The crystals are now clearly closer to the standard {104} rhombohedral calcite. However, as can be seen in the inset, the crystals still retain their [001] orientations as evidenced by the visible hexagonal symmetry of the (001) face. Figure 7d corresponding to 60:40 is now clearly {104} rhombohedral, but the crystals contain large {104}-stepped pits. Figure 7e is typical of samples grown under monolayers containing 60% or less sulfate. Similar to the acid DODAB mixtures, the crystals become {104} rhombohedral and begin to aggregate as the sulfate percentage is lowered even further. The final image, f, shows the aggregated {104}-oriented morphology of a crystal grown under 40:60 solution. The crystals in the sulfate DODAB mixtures appear to be slightly more irregular than those seen in the acid DODAB at the higher DODAB concentrations.

Figure 5. GID data showing the position of the crystalline peaks for crystals grown under the 80:20 acid DODAB mixture. Data presented as a function of Q xy for various values of Q z. The crystal peaks have been indexed using the hkl notation. The monolayer peak is also visible at Q xy = 1.52 Å 1 and Q z = 0 Å 1.

Figure 5 shows GID data for the crystals grown under an 80:20 mixture. These data were collected to determine the in situ orientation of the crystals under the monolayer. The monolayer peak is still visible at Q xy = 1.52 Å 1 as reported above. The large number of peaks and some rings are indicative of multiple orientations, but the two strongest peaks correspond to the {110} and {104} families. The (110) peak, Q xy = 2.52 Å 1, exists in-plane (Q z = 0), which is consistent with crystals growing from planes in the {01l} family, which includes (012). The (104) peak (which is the highest intensity peak in powder diffraction) is located at Q xy = 1.98 Å 1 and Q z = 0.6 Å 1. This is within the resolution of what is expected for an (012) orientation (Q xy =1.97 Å 1, Q z = 0.63 Å 1). These results are consistent with [102]-oriented or at least {01l}-oriented growth. It is possible that the low contact area between the crystal and monolayer led to some spread about the [012] orientation, but the angles made by the {104} faces when viewed from above (like the inset in Figure 4d) are clearly [012]-oriented.19 Sulfate DODAB Mixtures. Figure 6 shows a plot of the in-plane lattices and tilt angle at 10 N/m as a function of the sulfate fraction present. These monolayer mixtures exist only in

’ DISCUSSION An unexpected result of this study is the appearance of [012]oriented growth below the acid DODAB mixed monoalyer. This is the first time [012]-oriented growth has been seen under floating Langmuir monolayers. This is a very significant observation for two reasons. First, it allows us to make comparisons to [012]-oriented growth seen on self-assembled films. Second, our results directly show the in situ average, unlike those of previous studies.6,7,14 16 The self-assembled films and Langmuir monolayers are very similar templates, but until now they nucleated different orientations. We can now examine why we observe the same [012]-oriented growth under the mixed monolayer Langmuir template and the carboxylic acid terminated alkanethiol templates. Charge Density Variation. With both the acid DODAB and sulfate DODAB mixtures, we find that {104} rhombohedral 575

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578

Langmuir

ARTICLE

Table 1. GID Peak Position of Each Mixed Monolayer and the Corresponding Calculated Lattice Parameters acid concn (%) 100

DODAB concn (%) 0

Q xy (Å 1)

Q z (Å 1)

[02] (Å)

1.5

0

1.655

0

3.80

[10] (Å)

a (Å)

b (Å)

tilt (deg)

5.02

5.02

4.55

0.00

95

5

1.55

0

4.05

4.68

4.68

4.68

0.00

90

10

1.545

0

4.07

4.70

4.70

4.70

0.00

85

15

1.525

0

4.12

4.76

4.76

4.76

0.00

80

20

1.52

0

4.13

4.77

4.77

4.77

0.00

70

30

1.435

0

4.38

0.52 0

4.94

5.03

22.24

40

1.46 1.42

4.57

60

4.42

1.42

0.7

4.44

5.11

5.11

29.65

50

50

1.4125

0

1.39

0.75

4.45

5.25

5.16

32.07

40

60

1.41

0

1.385

0.8

4.38

5.27

5.18

33.86

25

75

1.41

0

4.46

0.8 0

5.32

5.19

34.12

100

1.375 1.38

4.41

0

4.55

1.25

0.9

4.56

6.03

5.46

40.81

4.45 4.46

reported on carboxylic acid terminated alkanethiol self-assembled monolayers (SAMs) on silver substrates and carboxylic acid terminated alkanethiol SAMs on gold in certain circumstances.19,21 In these cases the SAMs were 100% acid terminated. This leads to a large charge discrepancy in the two systems because DODAB has two tail groups (hence approximately twice the area per molecule compared to the acid), and an 80:20 mixture corresponds to coverage of only ∼2/3 of the surface area by acid groups. However, the SAMs and acid DODAB systems both induce [012] orientation. This is a clear indication that charge density is not the controlling factor in the oriented nucleation of CaCO3. Variation in Monolayer Tilt. The transition from [1̅ 10]- to [012]-oriented crystals under the acid DODAB monolayers is again interesting because both orientations nucleate under untilted monolayers, in contrast to results from previous work6 that suggested that a 12° tilt angle causes [012] growth. It is easy to see from the acid DODAB data that the mixed monolayers prefer to be either untilted or ∼30° tilted, with a fairly rapid transition between these tilt states. Modeling has shown that zero tilt and ∼30° tilt are favored by the interspacing of backbones of the alkane tails.15 The sulfate DODAB results also show that the [001] orientation is maintained over a range of mixtures from 100% to ∼70% sulfate. Over this range of concentrations, the monolayer tilt ranges between ∼30° and ∼40°. These two results are a clear indication that average monolayer tilt is not the controlling factor in the oriented nucleation of CaCO3. It is always possible that there is local variation in tilt at the nucleation site that is not seen by X-ray scattering. However, on the basis of the rapid transition from 0° to 30° tilt angle in the acid DODAB data and the supporting research suggesting these two tilt angles are preferred by the close packing of the alkane backbones,15 we expect these two tilts to occur locally and not just on average. Therefore, the molecular tilt angle cannot be considered the controlling factor in the templated nucleation of CaCO3. Lattice Variation. The sulfate DODAB monolayers grow [001]-oriented crystals over a range of skew hexagonal lattice

Figure 6. Magnitude of the monolayer in-plane lattice components and the tilt angle as a function of the percentage of sulfate in the monolayer. Regions giving different crystal growth are noted by arrows spanning the region and labeled by the [hkl] indices of the crystal orientation normal to the water surface.

crystals grow for DODAB concentrations of 50% and above. This result is not unexpected: {104} rhombohedral growth has been observed on neutral surfaces before.28,38 While the entire tilted phase of the acid DODAB mixtures grows {104} rhombohedral crystals, this is most likely due to the low charge of the monolayer. Note that while the DODAB is positively charged, the 100% DODAB sample has nearly 3 times the area per headgroup compared to the acid monolayer, or at most 1/3 the net charge density compared to the acid monolayer.18,32,33 It appears that with neutral or lower charged monolayers we see {104} growth, and to this extent monolayer charge plays a definite role. The transition from [1̅ 10]- to [012]-oriented crystals under the acid DODAB monolayers is particularly interesting. The [012] face, which is the more charged of these two crystal faces, grows under less charged monolayers (80 85% acid), while the less charged [1̅ 10] face grows under more charged monolayers (90 95% acid).10,11 [012]-oriented growth has been previously 576

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578

Langmuir

ARTICLE

Figure 7. SEM images of crystals grown under mixed monolayers of sulfate DODAB: (a) 100:0; (b) 80:20; (c). 70:30 (d) 60:40; (e). 50:50; (f) 40:60. Images similar to (f) are seen for all lower sulfate concentrations.

spacings from ∼5 to ∼6 Å (see Figure 5). The acid DODAB monolayers grow [012]- and [1̅ 1 0]-oriented crystals under hexagonal lattices spanning from ∼4.65 to ∼4.77 Å. It is important to note there is no overlap in the lattice parameters between sulfate DODAB mixtures that grow [001]-oriented crystals and the acid DODAB mixtures that grow [012]- and [1̅ 10]-oriented crystals. Each of the three orientations grows from a different range of lattice parameters. There is some overlap in the total range of acid DODAB and sulfate DODAB lattice parameters. However, this only occurs when one of the systems lacks the required charge density to grow oriented crystals and is nucleating {104} rhombohedral crystals. The three nontrivial orientations all have unique lattice spacings. Our study does not allow us access to the lattice spacings needed to compare our work to the much studied carboxylic acid terminated alkanethiol films on gold substrates,6,7,14 because of the low monolayer charge density over a large range of the samples (0 80% acid mixtures). We only see oriented growth under hexagonal lattices spanning from ∼4.65 to ∼4.77 Å (80 100% acid mixtures), while carboxylic acid terminated alkanethiol films on gold substrates have a 5.0 Å lattice. However, we can compare our results to those of carboxylic acid terminated alkanethiol films on silver substrates.6,7,14 The 80:20 acid DODAB lattice is nearly identical to the previously studied self-assembled carboxylic acid terminated alkanethiol films on silver substrates.6 Both lattices are 4.77 Å hexagonal and nucleate [012]-oriented crystals. However, this is the first time the self-assembled and Langmuir monolayers have been found to exhibit similar nucleation behavior, and this occurs when the average lattice spacings match. We are able to directly compare [012] crystals grown from Langmuir monolayers to carboxylic acid terminated alkanethiol films on silver substrates and have found the lattice spacing to be the common factor.

’ CONCLUSIONS In general, it could be considered surprising that the [012] orientation is seen under a mixed monolayer at all. The 80:20 acid DODAB mixed monolayer could be thought of an extremely dirty acid monolayer, yet it is still capable of growing highly charged crystal faces. In combination with evidence of large macroscopic strain defects seen under the sulfate DODAB mixtures and the thin strained [012]-oriented crystals seen under

acid DODAB mixtures, it is clear that this is an extremely resilient system. When a crystal grows oriented, it will strain macroscopically to maintain that orientation even when the crystals are grown under dirty or impure monolayers. Our results suggest that biological membranes, even though they are made of complex mixtures of lipids, can indeed act as crystal growth templates. The “fault tolerance” demonstrated by the growing crystals should perhaps not be very surprising given that these processes are carried out to extremely high precision in living systems. DODAB “impurities” were used in these Langmuir monolayers to adjust the relevant lattice parameters. If a monolayer mixture can be identified that can have higher average charge density and span a larger range of lattice spacings, this may allow for the study of monolayers with lattices similar to those of the widely studied self-assembled films on gold, without causing the monolayers to be insufficiently charged and thus merely grow {104} rhombohedral crystals. This work has shown that Langmuir monolayer templates can adopt structures similar if not identical to those of solidsupported carboxylic acid terminated alkanethiols and are capable of growing similarly oriented crystals while offering us the ability to directly observe the template structure. We have seen that a change in monolayer lattice corresponds to a change in the nucleating face of calcite and that this change cannot be attributed to changes in molecular tilt or charge density alone. Thus, our results provide evidence of a crucial role played by the lateral structure of the template, and we find no evidence of a role played by molecular tilt in this system.

’ ASSOCIATED CONTENT

bS

Supporting Information. Raw data for the acid DODAB mixtures. This material is available free of charge via the Internet at http://pubs.acs.org/

’ ACKNOWLEDGMENT We thank Jianming Bai for his assistance at Beamline X-14A. This work was supported by the U.S. Department of Energy under Grant DE-FG02-84ER45125. In situ grazing incidence crystal data were collected at the National Synchrotron Light Source, 577

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578

Langmuir

ARTICLE

Beamline X-14A. Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract DE-AC02-98CH10886. In situ monolayer data were collected at the Advanced Photon Source, Beamline 15-ID. ChemMatCARS Sector 15 is principally supported by the National Science Foundation/Department of Energy under Grant NSF/CHE-0822838. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract DE-AC02-06CH11357. The SEM images were recorded at the Electron Probe Instrumentation Center, Northwestern University.

(31) Ratajczak, M. K.; Chi, E. Y.; Frey, S. L.; Cao, K. D.; Luther, L. M.; Lee, K. Y. C.; Majewski, J.; Kjaer, K. Phys. Rev. Lett. 2009, 103, 28103. (32) Coppock, J. D.; Krishan, K.; Dennin, M.; Moore, B. G. Langmuir 2009, 25, 5006. (33) Goubard, F.; Fichet, O.; Teyssie, B.; Fontaine, P.; Goldmann, M. J. Colloid Interface Sci. 2007, 306, 82. (34) Kitano, Y. Bull. Chem. Soc. Jpn. 1962, 35, 1980. (35) Meron, M.; Gebhardt, J.; Brewer, H.; Viccaro, J. P.; Lin, B. Eur. Phys. J.—Spec. Top. 2009, 167, 137. (36) Lin, B. H.; Meron, M.; Gebhardt, J.; Graber, T.; Schlossman, M. L.; Viccaro, P. J. Physica B 2003, 336, 75. (37) Barton, S. W.; Thomas, B. N.; Flom, E. B.; Rice, S. A.; Lin, B.; Peng, J. B.; Ketterson, J. B.; Dutta, P. J. Chem. Phys. 1988, 89, 2257. (38) Nishino, Y.; Oaki, Y.; Imai, H. Cryst. Growth Des. 2009, 9, 223.

’ REFERENCES (1) Mann, S.; Archibald, D.; Didymus, J.; Douglas, T.; Heywood, B.; Meldrum, F.; Reeves, N. Science 1993, 261, 1286. (2) Kewalramani, S.; Kim, K.; Stripe, B; Evmenenko, G.; Dommett, G.; Dutta, P. Langmuir 2008, 24, 10579. (3) Uysal, A.; Stripe, B.; Kim, K.; Dutta, P. CrystEngComm 2010, 12, 2025. (4) Kewalramani, S.; Evmenenko, G.; Yu, C.-J.; Kim, K.; Kmetko, J.; Dutta, P. Surf. Sci. 2005, 591, 286. (5) Kmetko, J.; Yu, C.-J.; Evmenenko, G.; Kewalramani, S.; Dutta, P. Phys. Rev. B 2003, 68, 085415. (6) Aizenberg, J.; Black, A.; Whitesides, G. J. Am. Chem. Soc. 1999, 121, 4500. (7) Pokroy, B.; Aizenberg, J. CrystEngComm 2007, 9, 1219. (8) Han, Y.; Aizenberg, J. Agnew. Chem., Int. Ed. 2003, 42, 3668. (9) Popescu, D. C.; Smulders, M. M. J.; Pichon, B. P.; Chebotareva, N.; Kwak, S.-Y.; van Asselen, O. L. J.; Sijbesma, R. P.; DiMasi, E.; Sommerdijk, N. A. J. M. J. Am. Chem. Soc. 2007, 129, 14058. (10) Duffy, D.; Harding, J. Langmuir 2004, 20, 7630. (11) Duffy, D.; Harding, J. Langmuir 2004, 20, 7637. (12) Fricke, M.; Volkmer, D. Top. Curr. Chem. 2007, 270, 1. (13) Sommerdijk, NAJM; de With, G. Chem. Rev. 2008, 108, 4499. (14) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y. -T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152. (15) Ulman, A. Chem. Rev. 1996, 96, 1533. (16) Nemetz, A.; Fischer, T.; Ulman, A.; Knoll, W. J. Chem. Phys. 1993, 98, 5912. (17) Stephens, C. J.; Kim, Y.; Evans, S. D.; Meldrum, F.; Christenson, H. K. J. Am. Chem. Soc. 2011, 133, 5210. (18) Li, H.; Estroff, L. A. J. Am. Chem. Soc. 2007, 129, 5480. (19) Travaille, A. M.; Donners, J. J. J. M.; Greritsen, J. W.; Sommerdijk, N. A. J. M.; Nolte, R. J. M.; van Kempen, H. J. Adv. Mater. 2002, 14, 492. (20) Travaille, A. M.; Kaptijin, L.; Verwer, P.; Hulsken, B.; Elmans, J. A. A. W.; Nolte, R. J. M.; van Kempen, H. J. Am. Chem. Soc. 2003, 125, 11571. (21) Duffy, D.; Travaille, A. M.; van Kempen, H.; Harding, J. H. J. Phys. Chem. B 2005, 109, 5713. (22) Lendrum, C. D.; Mcgrath, K. M. Cryst. Growth Des. 2009, 9 (10), 4391. (23) Kim, K.; Uysal, A.; Kewalramani, S.; Stripe, B.; Dutta, P. CrystEngComm 2009, 11, 130. (24) DiMasi, E.; Olszta, M.; Patel, V.; Gower, L. CrystEngComm 2003, 5, 346. (25) DiMasi, E.; Kwak, S.; Pichon, B.; Sommerdijk, N. CrystEngComm 2007, 9, 1192. (26) Kwak, S.; DiMasi, E.; Han, Y.; Aizenberg, J. Cryst. Growth Des. 2005, 5, 2139. (27) Heywood, B. R.; Mann, S. Chem. Mater. 1994, 6, 311. (28) Stripe, B.; Uysal, A.; Dutta, P. J. Cryst. Growth 2011, 319, 64. (29) Lendrum, C.; McGrath, K. M. Cryst. Growth Des. 2010, 10, 4463. (30) Ivankin, A.; Kuzmenko, I.; Gidalevitz, D. Phys. Rev. Lett. 2010, 104, 108101. 578

dx.doi.org/10.1021/la2037422 |Langmuir 2012, 28, 572–578