Racemic versus Enantiopure Alanine on Cu(110) - American

Sep 6, 2007 - Sam Haq, Alan Massey,† Nasser Moslemzadeh, Abel Robin, Susan M. .... (8) Barlow, S. M.; Kitching, K. J.; Haq, S.; Richardson, N. V. Su...
0 downloads 0 Views 1MB Size
10694

Langmuir 2007, 23, 10694-10700

Racemic versus Enantiopure Alanine on Cu(110): An Experimental Study Sam Haq, Alan Massey,† Nasser Moslemzadeh, Abel Robin, Susan M. Barlow,* and Rasmita Raval* Surface Science Research Centre, Department of Chemistry, UniVersity of LiVerpool, LiVerpool L69 3BX, U.K. ReceiVed April 3, 2007. In Final Form: June 25, 2007

The adsorption of racemic alanine on the Cu(110) surface has been compared to that of enantiopure alanine using low-energy electron diffraction (LEED), reflection absorption infrared spectroscopy (RAIRS), and scanning tunneling microscopy (STM). No evidence of chiral resolution at the surface was observed for the racemic system, indicating that the formation of separate enantiopure areas is not preferred. Also, in contrast to the enantiopure system, no chirally organized phase was observed for the racemic system. LEED shows that both systems display a common (3 × 2) phase at high coverage. However, the pathway and kinetic barriers to this phase differ markedly for the racemic and the enantiopure systems, with the racemic (3 × 2) appearing at a temperature that is more than 100 K below that required for the enantiopure system. In addition, we report intriguing complexities for the (3 × 2) LEED structure that is ubiquitous in amino acid/Cu(110) systems. First, a common (3 × 2) pattern with a zigzag distortion can be associated with both the racemic and enantiopure systems. For the racemic system, the coverage can be increased further to give a “true” (3 × 2) LEED pattern, which is a transformation that is impossible to enact for the enantiopure system. Most importantly, STM images of the “distorted” and “true” (3 × 2) structures created in the racemic system show subtle differences with neither arrangement being fully periodic over distances greater than a few molecules. Thus, the (3 × 2) phase appears to be more complicated than at first indicated and will require more complex models for a full interpretation.

Introduction It is known that the spontaneous separation of enantiomers within racemic crystals can be driven by thermodynamic and kinetic factors but the ability to predict what will happen for any particular system is still very limited.1 The first observation of this phenomenon was famously made by Pasteur, who manually separated the enantiomeric forms of sodium ammonium tartrate tetrahydrate from a racemic mixture.2 More recently, empirical and statistical studies have shown that about 10% of racemates undergo spontaneous resolution with the shape, symmetry, and functional groups of the molecules all influencing this tendency. (See Kuzmenko et al.1 and references within.) In addition, it has been appreciated for a while that increased opportunities for enantiomeric resolution are offered by surfaces, where the reduced symmetry prevents the inversion of adsorbed molecules.3 Thus, as the various factors that govern chiral organization of adsorbed molecules on surfaces have begun to be understood,3-7 an interest has arisen in determining the conditions under which enantiomeric resolution can be observed. Given that many 2D systems exhibit a range of thermodynamically or kinetically preferred adsorption phases, dependent on temperature and coverage,3 the phenomenon * Corresponding authors. Tel: +44 151 794 6981. Fax: +44 151 794 3896. E-mail: [email protected] or [email protected]. † Current address: Department of Chemistry, University of Cambridge, U.K. (1) Kuzmenko, I.; Weissbuch, I.; Gurovich, E.; Leiserowitz, L.; Lahav, M. Chirality 1998, 10, 415. (2) Pasteur, L. Ann. Phys. 1848, 24, 442. (3) Barlow, S. M.; Raval, R. Surf. Sci. Rep. 2003, 50, 201. (4) Humblot, V.; Barlow, S. M.; Raval, R. Prog. Surf. Sci. 2004, 76, 1. (5) Raval, R. J. Phys.: Condens. Matter 2002, 14, 4119. (6) Ernst, K.-H. Top. Curr. Chem. 2006, 265, 209. (7) Pere´z-Garcia, L.; Amabilino, D. A. Chem. Soc. ReV. 2007, 36, 941.

of surface chiral segregation needs to be tracked on a phaseby-phase basis. In general, it can be expected that chiral segregation of enantiomers will be favored if homochiral interactions dominate (i.e., if like-like interactions between molecules of the same chirality are energetically preferred). The system will therefore create a racemic conglomerate in which enantiopure domains of each chirality coexist at the surface. Conversely, chiral segregation will not be favored if heterochiral interactions dominate (i.e., if like-unlike interactions between molecules of opposite chirality are preferred). In such systems, a true heterochiral racemate or a random solid solution will emerge instead. The adsorption behavior of amino acids and peptides at metal surfaces and the resulting chiral or achiral natures of their organizations have been studied extensively in recent years both experimentally and theoretically (e.g., glycine,8-13 alanine,12,14-21 (8) Barlow, S. M.; Kitching, K. J.; Haq, S.; Richardson, N. V. Surf. Sci. 1998, 401, 322. (9) Chen, Q.; Frankel, D. J.; Richardson, N. V. Surf. Sci. 2002, 497, 37. (10) Toomes, R. L.; Kang, J. H.; Woodruff, D. P.; Polcik, M.; Kittel, M.; Hoeft, J. T. Surf. Sci. 2003, 522, L9. (11) Rankin, R. B.; Sholl, D. S. Surf. Sci. 2004, 548, 301. (12) Rankin, R. B.; Sholl, D. S. J. Phys. Chem. B 2005, 109, 16764. (13) Zhao, X.; Gai, Z.; Zhao, R. G.; Yang, W. S.; Sakurai, T. Surf. Sci. 1999, 424, L347. (14) Barlow, S. M.; Louafi, S.; Le Roux, D.; Williams, J.; Muryn, C.; Haq, S.; Raval, R. Surf. Sci. 2005, 590, 243. (15) Barlow, S. M.; Louafi, S.; Le Roux, D.; Williams, J.; Muryn, C.; Haq, S.; Raval, R. Langmuir 2004, 20, 7171. (16) Jones, G.; Jones, L. B.; Thibault-Starzyk, F.; Seddon, E. A.; Raval, R.; Jenkins, S. J.; Held, G. Surf. Sci. 2006, 600, 1924. (17) Sayago, D. I.; Polcik, M.; Nisbet, G.; Lamont, C. L. A.; Woodruff, D. P. Surf. Sci. 2005, 590, 76. (18) Rankin, R. B.; Sholl, D. S. Surf. Sci. 2005, 574, L1. (19) Egawa, C.; Iwai, H.; Kabutoya, M.; Oki, S. Surf. Sci. 2003, 532-535, 233. (20) Iwai, H.; Tobisawa, M.; Emori, A.; Egawa, C. Surf. Sci. 2005, 574, 214.

10.1021/la700965d CCC: $37.00 © 2007 American Chemical Society Published on Web 09/06/2007

Racemic Versus Enantiopure Alanine on Cu(110)

Langmuir, Vol. 23, No. 21, 2007 10695

Figure 1. Adsorption phase diagrams of the S-alanine/Cu(110) and racemic alanine/Cu(110) systems showing the various coverage- and temperature-dependent phases and species. (i) Enantiopure S-alanine/Cu(110) system with phases I, II, III, and IV. (ii) Racemic alanine/ Cu(110) system with phases I, II IVa, and IVb. Note the absence of any phase equivalent to enantiopure phase III.

Figure 2. RAIR spectra of S-alaninate/Cu(110) and racemic alaninate/Cu(110) systems for the various phases. (a-d) Enantiopure phases I, II, III, and IV, respectively. (e-h) Racemic phases I, II, IVa, and IVb, respectively.

lysine,22-24 proline,25 dialanine,26 trialanine27 and trileucine27 on copper surfaces, cysteine28-31 on gold, and glutamic acid32,33 (21) Zhao, X.; Zhao, R. G.; Yang, W. S. Surf. Sci. 1999, 442, L995. (22) Zhao, X.; Zhao, R. G.; Yang, W. S. Langmuir 2000, 16, 9812. (23) Humblot, V.; Methivier, C.; Pradier, C.-M. Langmuir 2006, 22, 3089. (24) Humblot, V.; Methivier, C.; Raval, R.; Pradier, C.-M. Surf. Sci., 2007, published online 4 May 2007. (25) Mateo-Marti, E.; Barlow, S. M.; Raval, R. Surf. Sci. 2002, 501, 191. (26) Lingenfelder, M.; Tomba, G.; Constantini, G.; Ciachchi, L. C.; De Vita, A.; Kern, K. Angew. Chem., Int. Ed. 2007, 46, 4492. (27) Barlow, S. M.; Haq, S.; Raval, R. Langmuir 2001, 17, 3292. (28) Ku¨hnle, A.; Linderoth, T. R.; Hammer, B.; Besenbacher, F. Nature 2002, 415, 891. (29) Kuhnle, A.; Linderoth, T. R.; Besenbacher, F. J. Am. Chem. Soc. 2003, 125, 14680. (30) Kuhnle, A.; Linderoth, T. R.; Besenbacher, F. J. Am. Chem. Soc. 2006, 128, 1076. (31) Kuhnle, A.; Linderoth, T. R.; Schunack, M.; Besenbacher, F. Langmuir 2006, 22, 2156. (32) Jones, T. E.; Baddeley, C. J. Langmuir 2005, 21, 9468.

on silver and nickel). In particular, we have detailed knowledge of the bonding and organization of enantiopure R-alanine and enantiopure S-alanine on Cu(110) surface3,14-16 and have shown that these systems possess a number of phases exhibiting both chiral and achiral organizations. In this article, we compare and contrast this adsorption behavior for the enantiopure systems with that for racemic alanine on the Cu(110) surface using lowenergy electron diffraction (LEED), reflection absorption infrared spectroscopy (RAIRS), and scanning tunneling microscopy (STM). These techniques, in combination, provide information on the long-range order and the bonding and orientation of individual molecular species within an overlayer. Thus, they can monitor changes in surface species and organizational behavior, (33) Jones, T. E.; Urquhart, M. E.; Baddeley, C. J. Surf. Sci. 2005, 587, 69.

10696 Langmuir, Vol. 23, No. 21, 2007

Haq et al.

Figure 3. STM images of S-alaninate/Cu(110) and racemic alaninate/Cu(110) systems for phases II and III. (a) Enantiopure phase II showing chiral chains forming on the terraces (image size 254 Å × 246 Å, V ) -2.08 V, I ) 1.97 nA). (b) Racemic phase II showing patches of achirally ordered molecules (image size 400 Å × 400 Å, V ) -0.36 V, I ) 0.15 nA). (c) Details of the local order of racemic phase II taken from part a. (d) Enantiopure phase III showing size-selected clusters of six or eight molecules (image size 127 Å × 123 Å, V ) -2.08 V, I ) 1.97 nA).

Experimental Section

crystal was used. All spectra were recorded at 4 cm-1 resolution with the addition of 400 scans. A charge-coupled device (CCD) video camera interfaced to a computer was used for the recording and digitization of the LEED patterns.

The experiments were carried out in three ultrahigh vacuum (UHV) chambers, one specifically for RAIRS, LEED and temperature programmed desorption (TPD) measurements and the other two designed for STM and LEED. All systems also had facilities for Auger electron spectroscopy (AES), sample cleaning, and preparation. The RAIRS chamber was interfaced to a Mattson 6020 Fourier transform infrared (FTIR) spectrometer equipped with a liquid-nitrogen-cooled HgCdTe detector with a spectral range of 650-4000 cm-1. Two STM systems were used: an Omicron Vakuumphysik chamber for larger-scale images and the compact, variable-temperature Specs 150 Aarhus for high-resolution, well-calibrated images. In both systems, images were obtained with the sample at room temperature and in constant current mode. The Cu(110) crystal was cleaned by cycles of Ar+ sputtering, flashing, and annealing to 600 K. The surface ordering and cleanliness were monitored by LEED and AES. Racemic alanine (99%) was obtained from Aldrich and used without any further purification. The amino acid was contained in a small resistively heated glass tube, separated from the main chamber by a gate valve, and differentially pumped by a turbo molecular pump. Before sublimation, the alanine was outgassed for a few hours at ∼350 K and then heated to ∼370 K and exposed to the copper crystal. During sublimation, the main chamber pressure was typically 2 × 10-9 mbar. The RAIR spectra were recorded throughout a continuous dosing regime as sample single-beam infrared spectra, and the ratio against a reference background single beam representing the clean copper

The most striking difference between racemic and enantiopure alanine adsorption on Cu(110) is in the organizational behavior of the adsorption phases. Whereas enantiopure alanine exhibits both chiral and achiral organizations, no chiral organizations are observed with racemic alanine. Interestingly, this deviation in behavior is not due to the type of local chemical species that can be found at the surface; RAIRS identifies that in both systems the alanine molecule is deprotonated at the acid group and exists as the alaninate species in the monolayer regime. Furthermore, the same two types of alaninate species are created for each system, namely, the µ3 species bonding through the nitrogen of the amino group and both oxygen atoms of the carboxylate group and a differently oriented µ2 species bonding via the nitrogen of the amino group and only one of the oxygen atoms. The µ3 species is characterized by the presence of the symmetric carboxylate stretch at ∼1405-1411 cm-1 whereas the appearance of the asymmetric carboxylate stretch around 1620-1625 cm-1 indicates the existence of the µ2 species.3,14,15 However, there are significant variations in the temperature ranges over which the two species prevail in the enantiomeric and racemic systems. The phase diagram of Figure 1 and the RAIR spectra of Figure 2 summarize and compare the conditions under which the various molecular species and organizations are seen for both the racemic

enabling an understanding of the interactions that govern the racemic and enantiopure alanine systems.

Results and Discussion

Racemic Versus Enantiopure Alanine on Cu(110)

Langmuir, Vol. 23, No. 21, 2007 10697

Figure 4. Adsorption footprints of the µ3 species of S- and R-alaninate on Cu(110). DFT calculations11 show that the energetically equivalent and preferred footprints are for (b) S-alaninate and (d) R-alaninate, respectively. The less energetically preferred but energetically equivalent alternative footprints are for (a) S-alaninate and (c) R-alaninate, respectively.

and enantiopure systems. To compare the details of these adsorption phases, the discussion that follows is divided on a phase-by-phase basis. Phase I: Enantiopure and Racemic Disordered LowCoverage Phase. Both enantiopure and racemic alanine initially adsorb on Cu(110) at room temperature as the alaninate µ3 species, characterized by the symmetric carboxylate stretch at ∼14051411 cm-1 in the RAIR spectra of Figure 2a,e. This low-coverage phase is disordered for both systems with STM and LEED data showing no evidence of organization at the surface. Phase II: Enantiopure Local Chiral Organization versus Racemic Local Achiral Organization. Increasing the coverage at room temperature leads to phase II in which both the alaninate µ3 and µ2 species are present on the Cu(110) surface. This is indicated by the RAIR spectra of Figure 2b,f where the asymmetric carboxylate stretch around 1620-1625 cm-1 appears alongside the symmetric carboxylate stretch for both the enantiopure and the racemic systems. However, despite this chemical similarity, the organization adopted by the enantiopure systems differs from that preferred by the racemic system. STM images of enantiopure phase II in Figure 3a show chirally organized single-strand and double-strand molecular chains running broadly along the [1h12] nonsymmetry direction. This structure is “nematic” in that a general chiral growth orientation is adopted but there is no longrange periodicity. The racemic system, in contrast, shows no such chiral organization (Figure 3b,c). Here, STM images show achirally ordered patches with poor long-range order, and a very faint, rather diffuse LEED pattern with split spots is obtained; the symmetry of the LEED pattern is consistent with the formation of achiral organization.

Figure 5. DFT-optimized structures of the energetically favored enantiopure and racemic adlayers of alaninate/Cu(110) in phase IV. The white lines indicate a (3 × 2) surface cell with the corners of the cell on top of Cu atoms in the surface. (a) Enantiopure alaninate/ Cu(110). (b) Racemic alaninate/Cu(110). (Reprinted with permission from ref 12. Copyright 2005 American Chemical Society.)

The organizational differences exhibited in phase II are the first indicators of a significant departure in how the enantiopure and the racemic system evolve with temperature. Enantiopure phase II is stable until reaching a temperature of 430 K, beyond which it transforms into highly organized chiral phase III, as discussed below. In contrast, racemic phase II is stable only up to 358 K, after which it transforms directly into achirally organized phase IV, thus completely bypassing phase III. Phase III: Enantiopure, Highly Organized Chiral Phase with No Racemic Counterpart. Phase III is a striking phase created by the enantiopure system and is formed by warming the enantiopure adlayer to 430 K. It consists of alaninate µ3 and µ2 species that self-organize to give a globally organized chiral phase at the Cu(110) surface. The chiral organization of this phase is hierarchical in that individual chiral molecules assemble to create size-selected chiral clusters that, in turn, self-organize in a chiral fashion to give an extended chirally organized surface (Figure 3d). It has been demonstrated14 that there is direct transfer of chirality from the molecule to the macroscopic organization, with LEED data from the S-alaninate system exhibiting the (2 -2, 5 3) overlayer whereas the R-alaninate systems displays

10698 Langmuir, Vol. 23, No. 21, 2007

Haq et al.

Figure 6. LEED images of the racemic alaninate (3 × 2) structures at 29 eV beam energy. (a) Phase IVa showing a distorted (3 × 2) pattern after annealing an intermediate coverage surface to 423 K. (b) Phase IVb showing a true (3 × 2) pattern after annealing a saturated surface to 423 K. (c) Phase IVa showing a distorted (3 × 2) pattern after annealing a saturated surface to 498 K.

the mirror (5 -3, 2 2) organization. Long-range chirally organized phase III is observed only in the enantiopure case; there is no equivalent racemic phase as confirmed by both LEED and STM data. Instead, the racemic system bypasses this structure completely and proceeds directly to the achiral organization of phase IV, as discussed below. Phase IV: Enantiopure and Racemic Achiral (3 × 2) Phase. Both the racemic and enantiopure systems produce an achiral (3 × 2) LEED pattern, but there are some important structural variations, as discussed later. This phase occurs above 470 K for the enantiopure system but is attained at a much lower temperature of g358 K for racemic alanine. In both cases, the molecules are predominantly present as the µ3 species, with racemic alanine showing an indication of the coexistence of small amounts of µ2 species up to around 423 K. The (3 × 2) LEED pattern has the (0, n ( 1/2) spots missing, confirming p1g1 symmetry with a glide plane in the unit cell along the 〈001〉 direction. The existence of this glide plane has implications on the molecular arrangements that are possible in the unit cell, which need to be addressed separately for the enantiopure and racemic systems.

As discussed in our previous paper14 and others,12,17,18 a glide plane in an enantiopure system can be only a pseudoglide plane because the molecular chirality cannot be changed. We have suggested previously14 that it is informative to frame a discussion on adlayer chirality in terms of the chirality of the adsorbate (molecular chirality) and the chirality of the footprint delineating the bonding positions at the surface (footprint chirality). The adsorption footprint chirality of the µ3 species can adopt one of two triangularly shaped chiral adsorption footprints arising from the two carboxylate oxygens bonded to adjacent copper atoms in one close-packed row and the amino nitrogen bonded to one of two possible copper atoms in the next adjacent close-packed row, as shown in Figure 4.12 Rankin and Sholl18 have calculated that the preferred adsorption footprint for a monomer of S-alaninate is that of Figure 4b, albeit with each of the O atoms displaced differently in the 〈001〉 direction from atop Cu atom sites to give a slightly tilted footprint. Their calculation also reveals that when the (3 × 2) enantiopure overlayer structure containing two molecules per unit cell is formed, the lowestenergy structure (Figure 5a) is one in which the two molecules adopt opposite footprint chiralities. Photoelectron diffraction

Racemic Versus Enantiopure Alanine on Cu(110)

Langmuir, Vol. 23, No. 21, 2007 10699

Figure 7. STM images of enantiopure and racemic alaninate/Cu(110 systems for phases IV and IVa/IVb. (a) Enantiopure phase IV for S-alaninate associated with the distorted (3 × 2) LEED pattern (image size 254 Å × 246 Å, V ) -2.83 V, I ) 1.14 nA). (b) Racemic phase IVa associated with the distorted (3 × 2) LEED pattern (image size 224 Å × 224 Å, V ) 1.05 V, I ) 170 pA). (c) Racemic phase IVb associated with the true (3 × 2) LEED pattern (image size 224 Å × 224 Å, V ) 1.49 V, I ) 590 pA).

measurements17 of S-alanine on Cu(110) have also confirmed the existence of two molecules with slightly different local adsorption sites within the (3 × 2) unit cell. This arrangement, which has the expected pseudoglide plane, is often referred to as heterochiral because of the presence of the two chiral adsorption footprints, although the molecular chirality of the system is, of course, homochiral. A profound difference exists for racemic alaninate in that both the molecular chirality and the footprint chirality can be heterochiral, which means that a true glide plane can exist. Both of the energetically preferred footprints of S-alaninate (Figure 4b) and R-alaninate (Figure 4d) can be accommodated at the surface, giving a (3 × 2) arrangement with one molecule of each enantiomer in the unit cell and a true glide plane. Rankin and Sholl18 have calculated that the lowest-energy structure for racemic alaninate on Cu(110) is indeed formed with this arrangement (Figure 5b). Interestingly, the calculated energy difference between the favored enantiopure and racemic structures is actually very small (