Frictional-Force Microscopy and ab Initio Quantum Chemical Studies

Jun 27, 2011 - Surface Physics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-700064, India. §. Bioinformatics Institute, A-S...
0 downloads 0 Views 5MB Size
ARTICLE pubs.acs.org/JPCC

Wetting Property of the Edges of Monoatomic Step on Graphite: Frictional-Force Microscopy and ab Initio Quantum Chemical Studies Swati Panigrahi,†,^ Anuradha Bhattacharya,‡,^ Debashree Bandyopadhyay,§ Szawomir J. Grabowski,|| Dhananjay Bhattacharyya,† and Sangam Banerjee‡,* †

Biophysics Division, ‡Surface Physics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-700064, India Bioinformatics Institute, A-STAR 30 Biopolis Street, #07-01, Matrix Building, Singapore, 138671, Singapore Kimika Fakultatea, Euskal Herriko Uniberstitatea and Donostia International Physics Center (DIPC), P.K. 1072, 20080, Donostia, Euskadi, Spain and Ikerbasque, Basque Foundation for Science, 48011, Bilbao, Spain

)

§

ABSTRACT: In this work, we present the wetting property of the edges of a monatomic step on graphite. We have used frictional force microscope to experimentally investigate the wetting property of these edges. We have carried out quantum chemical calculations on a model system of nanographene to characterize hydrogen-bonding interactions between water and two different edges of the graphene, namely, the zigzag and arm-chair edges, to explain our experimental results. We have clearly observed two distinct frictional properties along the monatomic steps on graphite surface under varying conditions of humidity. The distinct frictional properties of the edges have been attributed to the different wetting properties associated with two different edges. Our subsequent quantum chemical calculations complement experimental findings with edge-specific graphenewater interaction.

1. INTRODUCTION Understanding the wetting property of solid surfaces has been of fundamental interest for long time from the point of basic science and numerous technological applications.13 Here, in this investigation, we would like to understand and address the wetting property of zigzag (trans) and arm-chair (cis) atomic arrangements. Graphene seems to be the best prototype model for such an investigation as it is a planar sheet of sp2 hybridized carbon atoms arranged in a honeycomb two-dimensional sheet. It is evident from the schematic diagram of an arbitrary cut graphene sheet, as shown in Figure 1, that if one of the edges is arm-chair (A) the other 30 bend cut will be zigzag (Z) or vice versa.46 A partial zigzag-arm-chair (Z,A) edge can be obtained if the cut bends to 60 angle.7 Intuitively, one may expect similar electron accumulation at both edges. However, it was reported earlier that there exist distinct electronic properties at these edges.821 It was noted that the zigzag edge of graphene layer, which can be part of the topmost layer on the pyrolytic graphite, accumulates more electron density as compared to the arm-chair edge.3,5,6 Our earlier theoretical calculations based on density functional theory also demonstrate the zigzag edge to be more electron rich22 as compared to that of the arm-chair edge. However, differential wetting properties associated with different edges of graphene have never been studied before. Macromolecular recognitions in a cellular environment are mostly governed by hydrophobic and hydrogen bonding (hydrophilic) interactions. These are the major driving forces to stabilize native structures of proteins, nucleic acids as well as other biomolecules.23,24 The origin of the hydrophobic interaction is r 2011 American Chemical Society

known to arise from entropy gained by the release of bound water molecules. Additional contribution to this stability often comes from van der Waals interactions, which are nondirectional in nature. The hydrogen bond, however, is sensitive to geometry and type of atoms involved in formation of such interaction. Traditionally polar groups were known to form hydrogen bond, where an electronegative atom acts as acceptor (A) while another electronegative atom acts as the proton donor (D) with A...HD in a linear, or near linear arrangement, where A and D may be nitrogen, oxygen, or other electronegative atoms. It is often stated that the hydrogen bond interaction is mostly electrostatic in nature. In addition, it is also described as overlapping of orbitals of donor, acceptor and hydrogen atoms and the charge transfer through the hydrogen bond. Numerous studies25 also indicate that in some cases carbon atoms, which are otherwise known as nonpolar type, can also act as the proton donor and/or proton acceptor.26,27 It has been shown that such nonpolar hydrogen bonds are often important in DNA and RNA structures28,29 and proteinprotein interface.23,24 Strength and stability of the nonpolar hydrogen bonds also depend on the acidity of the CH group involved, but its direct quantification is not available in the literature. Hydrogen bond donor capacity of carbon atoms in CH group depends on the state of hybridization of bonding orbitals on carbon. An increase in the s-character of the bonding orbital makes the carbon atom more acidic with Received: March 24, 2011 Revised: June 21, 2011 Published: June 27, 2011 14819

dx.doi.org/10.1021/jp2027466 | J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

ARTICLE

Figure 1. Schematic diagram of an arbitrary cut graphene.

accumulation of more negative charge on it. Such carbon atoms behave as better proton donors than those having less acidity and lower negative charges. As discussed earlier,22 carbon atoms at different edges of graphene molecule accumulate different amount of electronic charge, which probably indicates differential hydrogen bonding capacity of these associated CH bonds. Water molecule, often involved in hydrogen bond interaction, can act both as donor and acceptor. In our case, we have chosen the oxygen atom of water molecule as an acceptor species to test the strength of hydrogen bonds exerted by CH groups from different graphene edges. In this report, we have carried out detailed quantum chemical calculations to characterize hydrogen-bonding interaction between water and the different edges of graphene. Our supporting experimental studies using frictional force microscopy on a pyrolytic graphite surface also indicate differential interaction between water and the two edges of graphene, showing the possibility of the existence of weak attractive CH...O interactions for some of the CH groups. Interaction between graphene, as an extension of benzene with water using quantum chemical methods has recently been reported,30 however, they considered very small systems, inadequate to provide terminal properties. The real graphene structures are sometimes characterized by the existence of the dangling electrons, which are highly reactive.31 Hence we believe that the model systems taken for calculations in this study mimic at least approximately the real stable structures of graphene at stationary state. Recently it was experimentally shown5,6 that these edges of monatomic step on graphite have two distinct electronic states corresponding to two different edge structures (zigzag and arm-chair). Later, theoretically we have also shown that similar electronic property is observed in nanographene.22 As the previous results of both theoretical and experimental observations indicated similar behavior, we assume that the electronic structures of both are the same.

2. EXPERIMENTAL SECTION We have used freshly cleaved highly oriented pyrolytic graphite (HOPG) for the present investigation. We obtained terraces with single monolayer (monatomic) step edges by simply peeling off the surface layers using scotch tape. Commercial atomic force microscope (AFM) (NT-MDT, Russia) was used for the present

investigation. Gold-coated cantilevers with Si3N4 tip CSG 10, NT-MDT with a radius of curvature of the tip ∼350 Å and the cantilever elastic constant 0.1 N/m is used for this experiment without any further functionalization. Experiments carried out by different groups provide indirect evidence about the hydrophilic nature of the AFM tip. The works carried by different groups32,33 have shown that water bridge forms between the as-procured commercial AFM tip (nonfunctionalized) and the hydrophilic substrate, indicating hydrophilic nature of the AFM tip. If the tip is hydrophobic, then the tip will not interact with water and hence it would be impossible to distinguish the wetting from the nonwetting edges. For imaging in the contact mode, the normal force between the surface and the tip was kept around 25 nN in a constant force mode. The AFM and the frictional force microscopy (FFM) measurements were carried out in contact mode. All measurements were performed at room temperature (26 C) at different relative humidity (RH). The scan size for all of the samples was 2 μm in length on each side. In the frictional force microscopic technique, one can simultaneously measure both the topography and the frictional property by recording the current signals Iver and Itor, where Iver gives the magnitude of normal bending of the cantilever (topography) and Itor gives the magnitude of torsional or lateral bending of the cantilever (friction). For more details on FFM, please see ref 34. In Figure 2a, we show a 3-dimensional topographic image of a monatomic step and in Figure 2b,c we show the FFM image of the same region with scan direction right to left (R to L) and left to right (L to R) respectively. We clearly see the contrast inversion of the FFM images for the scan in forward and backward directions (see ref 34 for detail). In Figure 2d, we show a blown up region marked by a box, which clearly shows distinctly bright and dark regions along the step. The bright region corresponds to higher frictional (higher Itor i.e., higher lateral bending of the cantilever) value and the lighter region along the cut indicates low friction value. This clearly indicates the existence of two distinct types of frictional property along the cut. It should be pointed out here that the graphite surface is hydrophobic and earlier experiments suggest that upon increasing relative humidity the surface would not get covered with water.3537 In Figure 3, we schematically show the data analysis scheme adopted in this investigation. In the left panel of Figure 3, we show the amount of torsional bending of the cantilever during a line scan perpendicular to the cut along the forward (left to rightdownward scan, 14820

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

Figure 2. A 3-dimensional topographic image of a monolayer step is represented in (a). Parts (b) and (c) show the FFM images of the same region with scan direction right to left and left to right, respectively. The Figure (d) shows a blown up region.

i.e., going down the step) directions (shown by an arrow) having angle (θ’s). The torsional angle (θ) is the measure of frictional

ARTICLE

force between the tip and the sample surface. We have shown, in Figure 3, three situations “i” to “iii” of torsional bending of the AFM tip. In situation “i”, when the tip is moving on top of the graphite layer (terrace) the torsional bending is the least. When the tip arrives at the edge of the monatomic graphite layer, the torsional bending angle (θi) increases and for situation “ii” with the edge containing water droplet the torsional bending θii is larger than θi, this is because the water droplet acts as a glue32,33 and tries to hold back the tip as the tip moves toward forward direction. For a larger water droplet (situation “iii”) the torsional bending angle will be further increased to θiii where now θiii > θii, this is again because the larger water droplet will have a much larger restoring or holding back force on the tip than the smaller water droplet. In Figure 3 on the right panel we have schematically shown a histogram plot of the torsional bending of the cantilever proportional to the current, indicating Itor,  θ, We would like to mention that throughout our experiment for all of the scans the position of the laser beam on the cantilever was kept on the same spot and the scan speeds were also maintained the same. This is very important since we know that variation of the position of the laser beam on the cantilever and its speed affects the measure of bending of the cantilever, hence in turn will affect its frictional property also. In Figure 3, we show (schematically in the histogram plot) that apart from the main peak around zero torsional bending, which arises due to the top surface of the graphite (the terraces), there exist a second peak appearing as a shoulder to the main peak and one more third peak far away from the main central peak. These two extra peaks will be very small in magnitude compared to the main central peak because the histogram height will be proportional to the amount of area having that particular frictional property. The area along the step (a curved single line in the case of monatomic graphite layer) will be very small compared to the total surface area scan of the sample. The shoulder and the peak appearing at larger torsional bending angle are due to the monatomic steps. If there are two different frictional properties along the step, then we should also see two distinct peaks apart from the main peak. Two different types of steps along the edges have been shown by an STM images by Kobayashi et. al.4,7 As the size of the water droplet increases the secondary side peak should shift to higher torsional bending value because larger water droplet will cause larger pulling back force as shown schematically in Figure 3, whereas, edges where water does not condenses the secondary peak should not shift. Most interestingly when the tip is scanned in the reverse direction (from the right to leftupward scan, i.e., going up the step) the step height acts as a barrier, which does not allow water to be pulled by the tip. Thus the frictional scans have very low sensitivity to feel the water droplets along the edge in the reverse scan direction (R to L), which we have observed from our experimental measurement as shown in Figure 4. Hence, for our experimental investigation to study the wetting property of graphene edge, we have only considered the left to right forward scan. The peak “ii” in Figure 3 is for edges with water droplet and with the increase in RH (i.e., water droplet size) this peak moves to position “iii”. The peak “i” does not move with increase in RH in those parts of the edges where water does not condense. In Figure 4a we have presented the experimental histogram plot of torsional bending of the cantilever as a function of increase in relative humidity (RH = 60, 65, and 70%) of a monatomic step edge on highly oriented pyrolytic graphite (HOPG) using FFM. The histogram was obtained after flattening the data, this is done to reduce the noise in the histogram plots, which 14821

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

ARTICLE

Figure 3. Schematic diagram showing torsional bending of the cantilever during a line scan along the forward (left to right) directions, in different environmental conditions: (i) when the step edge contain no water (ii) when the step edge contains small water droplets (iii) when the step edge contains larger water droplets. The right-hand figure shows schematic histogram plot of the torsional bending of the cantilever proportional to the current Itor.

Figure 4. Experimental histogram plots of torsional bending of the cantilever as a function of increase of relative humidity (RH = 60, 65, and 70%) for downward and upward scan as shown in (a). In part (b), the deconvoluted lines are also shown (see text for details).

arises due to vibration of the samples and other disturbances due to electrical and thermal noise. The flattening of the data causes the main major peaks due to terraces to be around zero lateral bending value for both the left to right (L to R) and right to left (R to L) scans and this can be considered as the background. Hence the absolute frictional value of the terraces is considered as background, as we are only interested in obtaining the frictional properties of the edges, not in absolute scale. The error in the lateral bending can be obtained from the fluctuation of the lateral signal when scanning over a homogeneous area (in our case, in the terraces) and we found it to be around 1.2 pA (full width at

half-maximum), which is 1.2 unit in the x-axis of the histogram plot, almost the size of the symbols used in the plot. We clearly see a major peak corresponding to the surface (terraces) and two side peaks due to the edges during forward scan (L to R). The distant peak corresponding to large torsional bending (friction) has been attributed to the wet edges containing water and the side shoulder peak to the nonwetting edges. We also see that the peaks associated to the wet edges significantly shift toward higher torsional bending value, clearly indicating an increase in the water content as a function of increase in the RH, as discussed above. The two side peaks were analyzed by deconvoluting the peaks as shown in Figure 4b and we distinctly observe that as the percentage of RH increases the peak positions marked by arrow shift toward higher torsional bending values, whereas the shoulder peak does not show any systematic increase in the torsional bending value. In Figure 4a, we also show the histogram plot of the reverse scan (upward scan) and we clearly see that the histogram is not sensitive to the wetting property of the monatomic graphite edge. Thus, from our experimental results, we find presence of two different frictional properties along the edges of a graphene sheet. In our earlier experiment using a conducting tip AFM and theoretical study using density functional theory, we had shown that the zigzag edges have more electron density than the arm-chair edges. Hence, we can argue that the water being dipole will condense on the zigzag edge to minimize the surface (1-dimensional) area by spreading only on the zigzag edge. With increase in the relative humidity, we find that the water molecules condense at one particular edge. We shall show below by our theoretical analysis that the wet edges indeed can be associated to the zigzag edges and the nonwetting edges to the arm-chair edges and water accumulates only at the zigzag edges with further increase in relative humidity. Also we have observed from our theoretical calculations that when a water molecule is placed on top of the graphene surface the water moves toward the zigzag edge to minimize the energy. 14822

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

ARTICLE

Figure 5. Structures of the optimized graphene water complexes in two orientations for water interacting through (a) arm-chair edge (b) zigzag edge (c) top, along with electrostatic potential. The electrostatic potential values are color coded with red indicating less than 0.10 H, yellow indicating 0.05 H, green indicating 0.00, light blue indicating +0.05 H and dark blue indicating greater than 0.10 H.

Table 1. Hydrogen Bond Length, Angle and Interaction Energy of the Optimized Structures Obtained through Different Method system arm-chair zig-zag top-edge a

optimization method

energya (kcal/mol)

energyb (kcal/mol)

179.2

2.41

2.59

169.2 150.7

4.47

3.33

H-bond

bond length (H 3 3 3 A) in Å

HF/cc-pVDZ

Cs1Hs1...O

2.62

HF/cc-pVDZ

Cs4Hs4...O Ct2Ht2...O

2.40 2.45

Ct4Ht4...O

2.45

150.2

CT2...H1O

2.67

144.3

CT4...H2O

2.88

124.0

HF/cc-pVDZ

bond angle

Interaction energy calculated by HF/cc-pVDZ method. b Interaction energy calculated by MP2/6-31G** method.

3. THEORIES AND COMPUTATION A representative nanographene associated with both the armchair and zigzag edges has been modeled by MOLDEN38 soft. ware maintaining the standard CC bond lengths as 1.421 Å The edges were terminated with hydrogen atoms so as to neutralize the valences of the carbon atoms. Although there are several reports of theoretical studies of graphene without such capping, we feel the molecules with dangling electrons would behave like radicals and soon be reduced by ambient proton.31  and all of the The CH bond lengths are maintained at 1.009 Å angles are fixed to 120. Three graphenewater complex systems have been modeled maintaining the above geometry criteria and a single water molecule was placed near the graphene (i) at the arm-chair edge (ii) at the zigzag edge and (iii) at the top of the graphene. In the initial models of the arm-chair and zigzag edge systems, the oxygen of the water molecule points toward one of the hydrogen of the terminal CH bond of graphene edge. In the top-graphene system one of the hydrogen atoms of water was oriented toward the center of a hexagonal ring of graphene. All three systems have been geometry optimized with GAMESSUS39 employing Dunning Correlation Consistent basis set4042 (HF/cc-pVDZ) constraining covalent bonds of graphene by IFREEZ option. We observed two hydrogen bond like interactions between CH of graphene and oxygen atoms of water molecules in both arm-chair and zigzag edges after optimization (Figure 5). In both the cases, the water molecule remains in the plane of the graphene, the optimized geometries have the CH...O hydrogen bonding distances (H...O) of around 2.4 Å and angles subtended by donor-H-acceptor are also close to 180. In the case of arm-chair-graphene, one of those two H-bond distances is

comparatively larger while in case of zigzag graphene we find two CH...O types of hydrogen bonds of similar magnitudes (Table 1). From these data, one can presume that graphene can form weak hydrogen bond interactions with water oxygen atoms through CH...O contacts.26,30 Apart from the arm-chairand zigzag graphene systems, the top-graphene complex was also optimized. The initial model had the water molecule placed at top of graphene, about 2.88 Å angstroms vertically above the center of the graphene. One of the two H-atoms of water faced toward the center of the hexagonal ring. The oxygen atom was placed in such an orientation that the ring-center 3 3 3 HO angle is close to 180, suitable for an OH...π type of hydrogen bonding.43 After optimization the free water molecule moves from the center of graphene toward the zigzag edge but remain vertically above the graphene plane, forming slightly elongated H-bond. Here C...HO type of interaction is observed with a H...C distance of about 2.6 Å while OH...C angle is smaller than 150 (Table 1). Since the H...C distance is close to the corresponding sum of van der Waals radii one can expect that the OH...C interaction could not be classified as hydrogen bonding but rather as van der Waals type. Considering recent understanding44 that lone pair of the oxygen can also interact attractively with π-electron clouds, we have also optimized a graphenewater complex from an initial geometry where one of lone-pair electrons of the water oxygen was facing toward the central hexagonal ring of the graphene. The optimized geometry shows feature similar to the other top-graphene system—the water molecule moves in such a way that the hydrogen atoms come closest to carbons near the zigzag edge. These two observations indicate that a free water molecule prefers to bind to graphene through its zigzag edge as compared to all other edges. Moreover, the hydrophobic surface of 14823

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

ARTICLE

Table 2. Topological Parameters of BCPs (in a.u), Corresponding to CH...O Distances for all the Three Types of Graphene Systems Fc

system arm-chair zigzag top

r2Fc

GC

VC

HC

Cs1Hs1...O

0.007

0.024

0.006

0.005

0.0003

Cs4Hs4...O

0.010

0.036

0.008

0.008

0.0003

Ct2Ht2...O

0.009

0.032

0.008

0.007

0.0004

Ct4Ht4...O

0.010

0.032

0.008

0.007

0.0003

OH...C OH...C

0.006 0.006

0.020 0.020

0.004 0.005

0.003 0.004

0.0007 0.0005

graphene can not trap any water molecule to form a hydrated layer.3537 Structures of the optimized complexes are shown in Figure 5 along with their electrostatic potential. These electrostatic potentials were calculated by MOLDEN using multipole derived method from the orbitals electron density of the optimized structures. We have used contour value of 0.02 for these calculations. These figures indicate that most of the CH bonds near arm-chair edge are neutral (green in color), while the CH bonds at the zigzag edge have large dipole moment. The near triangular graphene considered in our study (Figure 5) has armchair edge at one side and zigzag edges at the two other sides with arm-chair edges also at the corners. The nature of the potential of the two zigzag edges are both polar but different in nature—(i) in one edge the hydrogen atoms are more electropositive (blue) and (ii) in the other edge the carbons are more electronegative (red), both leading to large dipole moments. These electrostatic potentials qualitatively indicate that the zigzag edge have stronger dipoles, as compared to the arm-chair edges (neutral, green), and may interact strongly with water. The atoms in molecules (AIM) approach is evolving as a very important theoretical technique for characterization of hydrogen bonds.45 We have carried out the AIM calculations for all three types of graphene systems discussed above using AIM2000 program.46 In this method, the strength of the hydrogen bonds were analyzed in terms of their electron densities (Fc) and Laplacians (r2Fc) at the bond critical point (BCP). The associative energy parameters, such as total electron energy density at the BCP (Hc) and the components of the latter term, namely the kinetic energy at the BCP (Gc) and the potential energy at the BCP (Vc), are also useful to characterize bond types. One can apply the topological parameters following eqs 1 and 2, which relate the energy parameters and the Laplacian of electron density (r2Fc) at the BCP.47 ð1=4Þ∇2 Fc ¼ 2Gc þ Vc

ð1Þ

Hc ¼ Vc þ Gc

ð2Þ

All values of eqs 1 and 2 are expressed in atomic units. Depending on the sign of r2Fc and Hc, we can differentiate strong, medium, and weak hydrogen bonds.48 We have calculated the Fc, r2Fc, Hc, Gc, and Vc for all interactions, mainly hydrogen bonds, observed in the arm-chair, zigzag, and top-graphene systems (Table 2). We can see that the values of r2Fc and Hc are positive in all types of graphene systems signifying formation of weak hydrogen bonds. Since there exists a correlation between hydrogen bond length and r2Fc, we observe two CH...O interactions for arm-chair-graphene, both are weak but they differ in the strength (Table 1). The r2Fc

values also show the similar trend. However in case of zigzag graphene system we notice two CH...O hydrogen bonds of similar strength as suggested by similar values of Laplacian and charge density. In case of the top-graphene system we observe very small values of r2Fc at the BCP, which signifies very weak and almost negligible interaction between the water and graphene, as it was mentioned before it may be classified as weak van der Waals interaction. This is also in agreement with previous experiments.35,37 So we have not considered the top-graphene system in further analysis. Interaction energy and basis set superposition error (BSSE, which originates due to overlapping of the orbitals across the molecules) were calculated from the optimized structures at HF/ cc-pVDZ level and also at MP2/6-31G**49, a higher level of approximation method. Both of these methods include electron correlation and electron exchange functions in their Hamiltonian, justify pursuing these methods to study orbital overlapping and electron exchange due to the hydrogen bond formation in watergraphene complexes. We have employed the Morokuma method50 of GAMESS-US and BoysBernardi counterpoise method51 of Gaussian0352 for the calculation of BSSE. The BSSE corrected interaction energy of the systems is calculated as follows: Eint ¼ Eðgraphene þ waterÞopt  Eðgraphene aloneÞopt  Eðwater aloneÞopt þ BSSE:::

ð3Þ

The interaction energies are calculated for the geometryoptimized arm-chair and zigzag graphene watercomplexes (Table 1). The interaction energy obtained from HF/cc-pVDZ in zigzag graphene system indicates it to be stronger by around 2 kcal/mol as compared to that of the arm-chair-graphene. A similar trend is found for the interaction energy calculated by more accurate MP2/6-31G** method also. The general conclusion is thus, water prefers to bind to the zigzag edge more strongly as compared to the arm-chair edge of graphene but the energy differs by few kBT at room temperature (Table 1). Charge transfer due to hydrogen bond formation for the optimized structures of the arm-chair and zigzag systems were also analyzed by natural bond orbital (NBO)5355 approach using Gaussian03. The NBO analysis includes all possible interactions between donor Lewis-type NBOs and acceptor nonLewis NBOs. Stabilization energy associated with delocalization was estimated by second-order perturbation theory to quantify the extent of hydrogen bond formation. Electronic properties, such as occupancy of the natural orbitals, stabilization energy, natural charges for the relevant atoms, and bonds of graphene, which are closest to the water molecule, are reported in Table 3. It indicates that charge transfer takes place, in both arm-chair and zigzag graphene, from the lone pair of oxygen to the antibonding orbital of the closest CH groups of the graphene ring. The significant change in natural charges of the carbon, hydrogen and oxygen of the complex and components in zigzag graphene, particularly for O, Ct1, and Ct2 atoms (Figure 6a) implies that zigzag graphene has more polarity than that of the arm-chair-graphene. We have also compared alterations in the NBO charges of all the atoms of graphene and water upon complex formation (Figure 6b). This clearly indicates that larger charge modifications take place for few carbon atoms when water comes close to the zigzag edge. It may also be noted that large charge-transfer takes place for some carbon atoms situated very far from the 14824

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

ARTICLE

Table 3. Results from Natural Bond Orbital Analysis Characterizing Hydrogen Bond Formation between Water Oxygen Atom and CH Moiety of Graphene Molecule in Different Orientations system

donor NBO (LP/BD)

occupancy

arm-chair

O(LP1)

1.996

O(LP2)

(1.997) 1.989

acceptor NBO CH (BD*)

occupancy

Cs1Hs1

1.9793

Cs4Hs4

(1.9799) 1.9788

(1.996)

NσE(N-σ*) (kcal/mol)

qO

qH

qC

LP1...Cs4Hs4:0.56

0.918

0.225

0.165

LP2...Cs1Hs1:1.46

(0.908) 0.918

(0.216) 0.224

(0.164) 0.286

(0.908)

(0.199)

(0.290)

(1.9800) LP2...Cs4Hs4:2.72

zigzag

O(LP1)

1.995

Ct1Ht1

(1.997) O(LP2)

1.991

1.9827

LP1...Ct1Ht1:0.38

(1.9831) Ct2Ht2

(1.996)

1.9813

LP1...Ct2Ht2:0.87

(1.9820)

0.921

0.240

0.153

(0.909)

(0.217)

(0.247)

0.921

0.234

0.299

(0.909)

(0.211)

(0.342)

LP2...Ct1Ht1:2.04 LP2...Ct2Ht2:1.59

So from this analysis, it is confirmed that electron delocalization in the graphene system is more significant due to the presence of the water molecule at the zigzag edge than that of the arm-chair edge. This in turn proves that water molecule is bound to the zigzag edge more strongly than that of the armchair-edge.

Figure 6. NBO charge donors and acceptors at the arm-chair and zigzag edges involved in CH...O interaction with the water molecules are shown in (a) and (b) denotes the change in Natural charges in the atoms of graphene and water before and after the complex formation in armchair edge (blue lines) and zigzag edge (red lines) orientations (As obtained from NBO analysis), the black vertical line distinguishes the carbon and hydrogen atoms, where as the black vertical broken line distinguishes the hydrogen and the water molecules.

water molecule, particularly in zigzag edge. The largest differences are seen for Ct1 and Ct2 (they are closest to the water molecule) and also for CF1, CT2, and CT4, which are far away from the binding site. The large change in NBO charges for CT2 and CT4, in zigzag edge complex, is presumably due to electron delocalization within graphene, the transferred charge is pushed away. A similar trend, but of very small magnitude, of alteration of NBO charges on these sites are also observed in arm-chair edge complex. This presumably indicates that arm-chair edges are more hydrophobic and the water molecule close to the arm-chair edge does not perturb the graphene molecule. The zigzag edge, however, is polar and strongly attracts water molecules. Furthermore, binding of a water molecule at the zigzag edge enhances further water binding at the zigzag edges.

4. CONCLUSIONS In this investigation, we have observed presence of two different frictional properties along monatomic step edges on graphite using Frictional Force Microscopy. The two different frictional properties can be attributed to two different types of edges of this layer, namely zigzag (or trans) edge and arm-chair (or cis) edge. These differential frictional properties are also seen to be dependent on the relative humidity of the experimental conditions—the frictional property of one of the edges depends on humidity while that of the other edge remains constant. Thus, the two edges appear to have different modes of binding to water molecules. The theoretical calculations enabled us to find out the mechanism of interaction of water with the two different types of edges of the graphene. The interaction energies and NBO charge transfer analysis suggest that the zigzag edge of graphene has a stronger binding capacity with the water molecule. The NBO analysis also clearly reveals that water binding enhances further water-binding capacity to the other zigzag edges of the nanographene sheet through a cooperative mode. We have also observed from our calculations that water does not remain on the graphite surface and hence exhibits hydrophobic nature of graphene/graphite surface. ’ AUTHOR INFORMATION Corresponding Author

*Phone: +91-33-2337-0379 ext. 3328; Fax: +91-33-2337-4637; E-mail: [email protected]. Author Contributions ^

These authors contributed equally, S.P. working on theoretical modeling and A.B. working on experimental analysis.

’ ACKNOWLEDGMENT We are grateful for the Center for Applied Mathematics and Computational Science, SINP for computational support. 14825

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826

The Journal of Physical Chemistry C

’ REFERENCES (1) Herminghaus, S.; Brinkmann, M.; alf Seemann, R. Annu. Rev. Mater. Res. 2008, 38, 101–121. (2) Bonn, D.; Eggers, J.; Indekeu, J.; Meunier, J.; Rolley, E. Rev. Mod. Phys. 2009, 81, 739–805. (3) Sarkar, S.; Patra, S.; Gayathri, N.; Banerjee, S. Appl. Phys. Lett. 2010, 96, 063112. (4) Kobayashi, Y.; Fukui, K.; Enoki, T.; Kusakabe, K.; Kaburagi, Y. Phys. Rev. B 2005, 71, 193406. (5) Banerjee, S.; Sardar, M.; Gayathri, N.; Tyagi, A. K.; Raj, B. Phys. Rev. B 2005, 72, 075418. (6) Banerjee, S.; Sardar, M.; Gayathri, N.; Tyagi, A. K.; Raj, B. Appl. Phys. Lett. 2006, 88, 062111. (7) Kobayashi, Y.; Fukui, K.; Enoki, T.; Kusakabe, K. Phys. Rev. B 2006, 73, 125415. (8) Fujita, M.; Wakabayashi, K.; Nakada, K.; Kusakabe, K. J. Phys. Soc. Jpn. 1996, 65, 1920–1923. (9) Nakada, K.; Fujita, M.; Dresselhaus, G.; Dresselhaus, M. S. Phys. Rev. B 1996, 54, 17954. (10) Wakabayashi, K.; Fujita, M.; Ajiki, H.; Sigrist, M. Phys. Rev. B 1999, 59, 8271–8282. (11) Stein, S. E.; Brown, R. L. J. Am. Chem. Soc. 1987, 109, 3721–3729. (12) Hosoya, H.; Gao, Y. D.; Nakada, K.; Ohuchi, M. In New Functionality Materials; Tsuruta, C. T., Doyama, M., Sena, M., Eds; Elsevier: New York, 1993, 27. (13) Tanaka, K.; Yamashita, S.; Yamabe, H.; Yamabe, T. Synth. Met. 1987, 17, 143–148. (14) Fujita, M.; Yoshida, M.; Nakada, K. Fullerene Sci. Technol. 1996, 4, 565. (15) Ryu, S.; Hatsugai, Y. Physica E (Amsterdam) 2004, 22, 679. (16) Wakabayashi, K. Phys. Rev. B 2001, 64, 125428. (17) Elstner, M.; Porezag, D.; Jungnickel, G.; Elsner, J.; Haugk, M.; Frauenheim, T.; Suhai, S.; Seifert, G. Phys. Rev. B 1998, 58, 7260–7268. (18) Kobayashi, Y.; Fukui, K. I.; Enoki, T.; Kusakabe, K.; Kaburagi, Y. Phys. Rev. B 2005, 71, 193406. (19) Dedkov, Y. S.; Shikin, A. M.; Adamchuk, V. K.; Molodtsov, S. L.; Laubschat, C.; Bauer, A.; Kaindl, G. Phys. Rev. B 2001, 64, 035405. (20) Nimi, Y.; Matsui, T.; Kambara, H.; Tagami, K.; Tsukada, M.; Fukuyama, H. Appl. Surf. Sci. 2005, 241, 43. (21) Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6, 183–191. (22) Banerjee, S.; Bhattacharyya, D. Comput. Mater. Sci. 2008, 44, 41–45. (23) Jiang, L.; Lai, L. J. Biol. Chem. 2002, 277, 37732–37740. (24) Scheiner, S.; Kar, T.; Gu, Y. J. Biol. Chem. 2001, 276, 9832–9837. (25) Desiraju, G. R. Angew. Chem. Int. Ed 2010, 49, 2–10. (26) Cho, K.; Lee, K.; Lee, K. H; Kim, D.; Lee, D. Protein: Struct. Funct. Bioinf. 2006, 65, 593–606. (27) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural Chemistry and Biology; Oxford University Press Inc.: New York, 1999. (28) Ghosh, A.; Bansal, M. Acta Crystallogr. D Biol. Crystallogr. 1999, D55, 2005–2012. (29) Roy, A.; Panigrahi, S.; Bhattacharyya, M.; Bhattacharyya, D. J. Phys. Chem. B 2008, 112, 3786–3796. (30) Rubes, M.; Nachtigall, P.; Vondrasek, J.; Bludsky, O. J. Phys. Chem C. 2009, 113, 8412–8419. (31) Koskinen, P.; Malola, S.; H€akkinen, H. Phys. Rev. Lett. 2008, 101, 115502. (32) Jinesh, K. B.; Frenken, J. W. M. Phy. Rev. Lett. 2006, 96, 166103. (33) Wei, Z.; Zhao, Y.-P. J. Phys. D: Appl. Phys. 2007, 40, 4368. (34) Banerjee, S.; Gayathri, N.; Shannigrahi, S. R.; Dash, S.; Tyagi, A. K.; Raj, B. J. Phys. D: Appl. Phys. 2007, 40, 2539. (35) Yang, H.; Fung, S. Y.; Pritzker, M.; Chen, P. PLoS One 2007, 2, e1325. (36) Biswas, S.; Drzal, L. T. Nano Lett. 2009, 9, 167–172. (37) Leenaerts, O.; Partoens, B.; Peeters, F. M. Phys. Rev. B 2009, 79, 235440.

ARTICLE

(38) Schafteaar, G.; Noordik, J. H. J. Comput.-Aided Mol. Des. 2000, 14, 123–134. (39) Schmidt, M.; Baldridge, K.; Boatz, J.; Elbert, S.; Gordon, M.; Jensen, J.; Koseki, S.; Matsunaga, N.; Nguyen, K.; Su, S. J. Comput. Chem. 1993, 14, 1347–1363. (40) Dunning, T., Jr. J. Chem. Phys. 1989, 90, 1007. (41) Kendall, R.; Dunning, T., Jr.; Harrison, R. J. Chem. Phys. 1992, 96, 6796. (42) Peterson, K.; Kendall, R.; Dunning, T., Jr. J. Chem. Phys. 1993, 99, 1930. (43) Grabowski, S. J. J. Phys. Chem. A 2007, 111, 13537–13543. (44) Jain, A.; Ramanathan, V.; Sankarramakrishnan, R. Protein Sci. 2009, 18, 595–605. (45) Bader, R. F.W. Chem. Rev. 1991, 91, 893–928. (46) Beigler-Konig, F; Schonbohm, J.; Bayles, D. J. Comput. Chem. 2001, 22, 545559. (47) Bader, R. F. W. Atoms in Molecules: A Quantum Theory, Oxford University Press: Oxford, 1990. (48) Rozas, I.; Alkorta, J.; Elguero, J. J. Am. Chem. Soc. 2000, 122, 11154–11161. (49) Møller, C.; Plesset, M. Phys. Rev. 1934, 46, 618–622. (50) Kitaura, K.; Morokuma, K. Int. J. Quantum Chem. 1976, 10, 325–340. (51) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553–566. (52) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;Robb, M. A.; Chesseman, J. R.; Zakrzewski, V. G.; MontgomeryJr., J. A.; Stratmann, R. E.; Burant, J. C.; Dapprich,S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.;Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Forseman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; AlLoham,M. A.; Peng, C. Y.;Nanayakkara, A.; Gonzalez, C.; Challacombe,M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 03; Gaussian Inc.: Wallingford, CT, 2003. (53) Carpenter, J. E.; Weinhold, F. J. Mol. Struct. (Theochem) 1988, 46, 41–62. (54) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735–746. (55) Reed, A. E.; Curtsiss, L. A.; Weinhold., F. Chem. Rev. 1988, 88, 899–926.

14826

dx.doi.org/10.1021/jp2027466 |J. Phys. Chem. C 2011, 115, 14819–14826