Hydroxyl-Induced Partial Charge States of Single Porphyrins on

---

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

Hydroxyl-Induced Partial Charge States of Single Porphyrins on Titania Rutile Rémy Pawlak, Ali Sadeghi, Res Jöhr, Antoine Hinaut, Tobias Meier, Shigeki Kawai, #ukasz Zaj#c, Piotr Olszowski, Szymon Godlewski, Bartosz Such, Thilo Glatzel, Stefan Goedecker, Marek Szymonski, and Ernst Meyer J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b11873 • Publication Date (Web): 19 Jan 2017 Downloaded from http://pubs.acs.org on January 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Hydroxyl-Induced Partial Charge States of Single Porphyrins on Titania Rutile R´emy Pawlak,∗,† Ali Sadeghi,‡ Res J¨ohr,† Antoine Hinaut,† Tobias Meier,† Shigeki Kawai,† Lukasz Zajac,¶ Piotr Olszowski,¶ Szymon Godlewski,¶ Bartosz Such,¶ Thilo Glatzel,† Stefan Goedecker,† Marek Szymo´nski,¶ and Ernst Meyer∗,† †Department of Physics, University of Basel, Klingelbergstrasse 82, 4056 Basel, Switzerland ‡Department of Physics, Shahid Beheshti University, Evin, 19839 Teheran, Iran ¶Faculty of Physics, Astronomy and Applied Computer Science, Jagiellonian University, Ul. S. Łojasiewicza 11, 30-348 Krakow, Poland E-mail: [email protected]; [email protected]

ABSTRACT The adsorption of Cu-porphyrin derivatives terminated with peripheral carboxyphenyl side groups on hydroxylated TiO2 (110) surfaces is investigated by combined scanning tunneling microscopy (STM), atomic force microscopy (AFM) and density functional theory (DFT). Two distinct contrasts of the molecules are revealed by STM and tunnelling spectroscopy. Via single-molecule manipulations, the origin of these peculiar contrasts is found to arise from the presence or not of hydroxyl groups below the molecules. Hence, the electronic coupling of the molecule with the underlying TiO2 surface is locally modified altering the amount of charge transfer and thus their charge state at the molecular scale. Our results particularly underline the fundamental role of hydroxyls of TiO2 on the charge 1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

state of adsorbed organic molecules potentially used in dye-sensitized solar cells.

INTRODUCTION The working principle of dye-sensitized solar cells (DSSC) 1 relies on the injection of photo-excited electrons from adsorbed photo-sensitive dye molecules to the conduction band of titanium dioxide (TiO2 ) resulting in photo-current generation. Since the overall performance of DSSCs drastically depends on the hybrid organic/inorganic interface, fundamental investigations are necessary down to the molecular level. Unravelling the interplay between adsorption characteristics, electronic coupling and the energy level alignment might indeed open new routes towards improvements of organic based devices. TiO2 is an important material for energy-harvesting applications and catalysis. Widely investigated within the past years, 2 the surface chemistry of TiO2 , especially rutile TiO2 (110), is strongly influenced by reactive point defects. 3 Upon adsorption of organic compounds on such surfaces, various chemical reactions can be initiated at the surface via temperature, illumination or by applying an electric field with the help of a scanning tunneling microscope (STM). 4 In this way, the dissociation of water molecules 4 has been demonstrated as well as the dehydrogenation of organic molecules 5 and the synthesis of long polymeric chains. 6 Additionally, the intrinsic surface chemistry of the native rutile titania is also tunable by terminating the surface with chemical end-groups such as hydroxyls. 6 Then, the modification of diffusion characteristics of adsorbates is observed 7 as well as an alteration of their electronic properties. 8 As a consequence, the titania surface reactivity should be considered as a pivotal parameter in DSSCs which certainly favours the photo-sensitizers anchoring but also capable of disrupting the intrinsic electronic properties of the dyes. To gain knowledge onto the impact of titania surfaces on the adsorbed dye molecules, we choose a prototypical metallo-porphyrin derivative already employed in DSSC applications. 9 Although several works have reported on the assembly of organic molecules on

2

ACS Paragon Plus Environment

Page 2 of 25

Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

TiO2 (110) by STM 10,11 and the role of point defects such as oxygen vacancies and hydroxyl groups, 3 real-space investigations at the atomic level of such photo-sensitizers 12 are still scarce in the literature 13–17 Here, we show from a combined STM, non-contact atomic force microscopy (ncAFM) and density functional theory (DFT) study how the presence of surface hydroxyls on the rutile-TiO2 (110) locally impacts the electronic properties of adsorbed 5,10,15,20-tetrakis-(4carboxyphenyl)-porphyrin-Cu(II) (CuTCPP). Two distinct electronic structures of single molecules are identified by means of STM which arise from the presence or not of hydroxyl groups underneath the molecule. As a consequence, the electronic coupling of such molecules with the substrate is locally perturbed by the hydroxyls modifying their charge state. Our results thus underline the fundamental role of hydroxyl groups of the rutile TiO2 (110) surface on the electronic properties of single dye molecules.

EXPERIMENTAL METHODS Sample preparation. Rutile-TiO2 (110) single crystals purchased from MaTecK GmbH were prepared by cycles of Ar+ sputtering and annealing by direct current heated to ≈ 1050 K under ultrahigh vacuum conditions (UHV) (≈ 5×10−10 mbars). The 5,10,15,20tetrakis-(4-carboxyphenyl)-porphyrin-Cu(II) molecules (CuTCPP), purchased from Rare Chemicals GmbH, were deposited from a quartz crucible heated up to 575 K under UHV on the substrate cooled to 130 K. The hydroxylated rutile TiO2 (110) were obtained by storing the sample for 12 hours in the UHV chamber at the base pressure of 5×10−10 mbars. The local pressure in the microscope chamber (1×10−11 mbars) prevents further hydrogen adsorption and, thus hydroxyl formation. The molecule deposition as well as the experiments were conducted on highly hydroxylated samples of ≈ 0.8-0.9 monolayer coverage. In that conditions, 20-25 % of the molecules adopt a neutral charge state upon adsorption on the titania.

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

STM/AFM experiments. STM/AFM experiments were realized with a low-temperature microscope (Omicron Nanotechnology GmbH) operated at 77 K with Nanonis GmbH electronics and based on a tuning fork sensor in the qPlus configuration (stiffness k = 1800 N/m, resonance frequency f 0 = 26 kHz, Q factor = 10000 at 77 K, oscillation amplitude of A = 60 pm). The STM experiments were conducted in the constant-current mode. For practical reasons (stability of the imaging conditions), we essentially probed empty states of the sample. The differential conductance measurements were acquired with the standard lock-in technique (frequency f = 532 Hz, voltage modulation Amod = 9 mV). Local contact potential difference (LCPD) maps, ∆ f ( x, y, V ), were acquired at 77 K in a grid of 3×3 nm2 with to 35×35 pixel2 . 18–22 Thermal drifts and piezo creeps were compensated by atom tracked positioning. Density functional theory.

Calculations were carried out using the generalized gradient

approximation of DFT as implemented in the VASP code. 23–26 Dispersion effects were included via the empirical pair potential of Grimme. 27 The TiO2 surface was modeled by a slab of four trilayers, each being a 4 × 9 cell. The two layers at the bottom were kept frozen at their ideal crystal positions while relaxing all other atoms to find their optimum geometrical configuration. Because of the rather large size of the cell, only the center of the Brillouin zone was sampled. Molecular charge were assigned according to partitioning the space between the molecule and the substrate using a separating surface on which the electron density becomes minimum. 28,29

RESULTS AND DISCUSSION Hydroxylated Titania Surface. To characterize the adsorption of surface hydroxyls on rutile TiO2 (110), we first used titania samples with a coverage of ∼ 0.05 ML. Figures 1a and b show a STM overview of such hydroxylated rutile TiO2 (110) surface with atomic precision. The rutile TiO2 (110) structure is characterized by alternating rows of five-fold4

ACS Paragon Plus Environment

Page 4 of 25

Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1: Hydroxylated rutile TiO2 (110) surface. (a) Empty-state STM image of the rutile TiO2 (110) surface with ∼ 0.05 ML hydroxyl coverage, scale bar = 4 nm. (b) Close-up STM topographic image revealing the Ti (bright) and O (dark) rows, the oxygen vacancies (dashed circle) as well as single- (arrow) and double-hydroxyls (full circle) at the oxygen rows , scale bar = 1 nm, (I = 10 pA, Vsample = +1.3 V) . (c) Schematic of the TiO2 (110) surface structure illustrating oxygen vacancies (dashed circle), single (arrow) and double hydroxyl groups (full circle) observed at the surface by STM in (b). coordinated Ti and bridging-O atoms aligned along the [001] direction. When imaging the empty states by STM (Figures 1 a and b), the Ti rows appear as bright thin lines separated by ∼ 650 pm whereas the O rows have darker contrast. 2 Between the bright rows, oxygen vacancies are observed by STM as faint white spots denoted by a dashed white circle in Figures 1b and c. Two types of brighter spots are also present, marked with an arrow and a full circle in Figures 1b and c, and correspond respectively to single and double hydrogen atoms bound to the oxygen rows, the so-called hydroxyl groups. 4,30,31

CuTCPP Porphyrins Adsorbed on The Hydroxylated Titania Surface.

The chemical

structure of the 5,10,15,20-tetrakis-(4-carboxyphenyl)-porphyrin-Cu(II) molecule (CuTCPP) is shown in Figure 2a. The porphyrin macrocycle contains a Cu atom at its center, surrounded by four pyrole sub-units (C4 N), to which are attached four peripheral carboxyphenyl (ph-COOH) side groups in meso position. The molecule having a width of

∼ 1.4 nm is suited to the rutile TiO2 (110) surface since it corresponds to the distance of three consecutive oxygen rows (∼ 1.3 nm). In recent works, 16,17 we demonstrated that the

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2: Adsorption characteristics of the CuTCPP Porphyrins on hydroxylated titania surface. (a) Molecular structure of the 5,10,15,20-tetrakis-(4-carboxyphenyl)-porphyrinCu(II). (b) STM overview of CuTCPP adsorbed on the hydroxylated rutile TiO2 (110), scale bar = 5 nm, (I = 10pA, Vsample = +1.8 V). Two STM contrasts, marked as 1 and 2, are observed on the STM image upon adsorption. (c) Close-up STM image of the two adsorption configurations revealing that both molecular configurations are aligned along the oxygen rows. The white lines represent the oxygen rows of the rutile TiO2 (110). (d) Characteristic differential tunneling conductance spectra, dI/dV, acquired at the center for both configurations. The inset shows high-resolution constant-current STM images of the molecules 1 and 2 acquired at sample voltages of 1.8 V. The shift of ∼ 0.3 eV of the HOMO resonance suggests that molecule 1 is positively charged (CuTCPP+ ) compared to molecule 2 (CuTCPP0 ). The black curve corresponds to the bare rutile-TiO2 (110) where CBM refers to the conduction band minimum. CuTCPP molecule adsorbed at room temperature on a hydroxyl-free rutile TiO2 surface has two distinct planar configurations, which are either driven by van der Waals interaction and hydrogen bonding or by covalent bonding with the surface. 16 Upon adsorption at 130 K on a highly hydroxylated rutile TiO2 (110) surface (∼ 0.8-0.9 ML), we found that the 6

ACS Paragon Plus Environment

Page 6 of 25

Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

molecule adopts only one adsorption configuration as revealed in the STM image (the white dashed lines show the oxygen row positions). Although all molecules have a similar adsorption structure, two distinct contrasts, denoted as 1 and 2 in Figures 2b and c, are observed by STM. The difference of contrast consists in the appearance of two bright lobes at the macrocycle location for molecule 2 with respect to molecule 1. Statistically, 20 to 25 % of the molecules adsorbed on the TiO2 (110) surface with 0.8-0.9 ML of hydroxyls appear with the bright contrast whereas 75% of the molecules show a dark appearance. We think that the deposition on the cold substrate also help to promote the presence of molecule 2 by limiting the molecule diffusion. Reducing the density of hydroxyls such as in Figure 1 or increasing the temperature leads to mainly ”dark” molecules such as 1. To further elucidate this electronic peculiarity, we probed the local density of state by scanning tunnelling spectroscopy (STS) of the two molecular configurations. The differential conductance spectra, dI/dV, were acquired at the center of the molecule macrocycles and are shown in Figure 2d in red and blue for molecule 1 and 2, respectively. The black curve of Figure 2d corresponds to the rutile TiO2 (110) surface which shows the conduction band minimum (CBM) position at ∼ 0.8 eV. The valence band maximum of the surface was probed at ∼ -3.2 eV (not shown in Figure 2d for clarity purposes). The lowest unoccupied molecular orbital (LUMO) of both configurations start at ∼ 2 eV. In contrast, the highest occupied molecular orbital (HOMO) appear at -1.8 eV and -2.1 eV for molecule 1 and 2, respectively. The extracted energy gaps are thus Eg ≈ 3.8 eV and 4.1 eV, respectively, in a relative good agreement with previous measurements combining valence band spectra and inverse photo-emission spectroscopy on porphyrin analogues 32,33 (Eg ≈ 3.8 eV ± 0.4 eV). Furthermore, the electron injection barrier, commonly defined as the energy difference between the Fermi level and the onset of the LUMO resonance peak of the molecules, is ∼ 1.6 eV from the STS data. The observed shift of ∼ 0.4 eV of the HOMO peak of molecule 1 (marked H + in Figure 2d) towards the Fermi level with respect to the HOMO of molecule 2 (H 0 ), implies that

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

molecule 1 is more positively charged than molecule 2. For that reason, we refer in the rest of the manuscript molecule 1 and 2 as CuTCPP+ and CuTCPP0 , respectively. Note also that, upon deposition at 130 K on bare titania with low concentration of hydroxyls, all molecules adopt the CuTCPP+ configurations. Such local modification of the orbitals has been recently reported in the literature in molecules having different adsorption sites on hBN/Cu(111), 34 by intercalating single atoms below single porphyrins on Au(111) 35 or due to conformational changes. 38 The electronic effect can arise from the hybridization of the molecular orbitals with sub-gap states of the titania, 13 adsorbates and/or a local change of the electrostatic surface potential. 8

Adsorption-Site Identification With Single-Molecule Manipulation. To identify any difference between adsorption site of these molecular conformers, lateral displacements of porphyrins were conducted by inelastically injecting tunneling electrons. 39 Figure 3a shows a typical current-voltage spectrum, I/V, of such process obtained by ramping the sample voltage from -2.5 to +2 V at constant height above the porphyrin macrocyle center. Successful manipulation processes are detected when an abrupt jump is observed at the onset of the HOMO resonance peak (see the arrows Figure 3a). The second curve (red curve) obtained at the same position and tip-sample separation does not show any molecular orbitals which experimentally reveals the absence of the molecule below the tip and its displacement from the tip position. Figure 3b shows STM images of translations of single CuTCPP+ over the rutile TiO2 (110) surface. The red crosses indicate the tip position where I/V spectra were acquired. Subsequent STM images confirmed the CuTCPP+ motion which corresponds to a translation of ∼2 nm along the oxygen rows of the structure. A careful inspection of Figure 3b reveals that the molecular contrast of the CuTCPP+ molecules does not change with translations and thus, the state of the molecule is not modified upon displacements. Moreover, we did not observe that the positions of the hydroxyl groups underneath the molecule influence their translations. The

8

ACS Paragon Plus Environment

Page 8 of 25

Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

current injection process only activates the porphyrin translation and does not modify the hydroxyl positions and does not induce their removal. 4 To investigate the translation of the neutral state, we show in Figure 3c two STM images of successful translations of CuTCPP0 molecules. Upon displacements, the CuTCPP0 molecules always appear with the other STM contrast (i.e. that of the CuTCPP+ one), thus suggesting the change of their charge state after translations. 34,35 Thanks to these lateral displacements, we identified the hydroxyl positions before and after the manipulation processes (images 1 to 2 of Figure 3c) and no clear modification of their positions was observed. Therefore, the origin of the CuTCPP0 charge variation is extremely local (translation of only ∼2 nm). The white full circles of the image 2 (Figures 3c and d) further reveal the initial adsorption sites of the CuTCPP0 molecules which consist of hydroxyl groups underneath the molecule having a five fold symmetry and centered below the macrocyle of the molecule. Hydroxyls are known to be positively charged on TiO2 8 and most likely modify the molecule/surface electronic coupling. 40 Interestingly and despite of the fact that the TiO2 (110) is highly hydroxylated, only the OH groups having such five-fold fashion underneath the molecule causes the CuTCPP0 configuration. Among all translations conducted, we observed that other numbers and positions of hydroxyls underneath the molecule do not change the appearance of the molecule by STM in contrast to the five-fold fashion one. On bare titania, e.g. with low concentration of hydroxyls, translations of CuTCPP were also conducted and did not show any change of the STM appearance.

Relaxed CuTCPP Configurations Obtained From DFT Calculations. To verify and gain more insight into the influence of hydroxyls on the electronic properties of CuTCPP molecules, we performed DFT calculations on the two detected configurations. Figure 4a shows the top and side views of the CuTCPP+ relaxed adsorption geometry. The copper atom of the porphyrin macrocycle is adsorbed on top of a bridging oxygen atom. The

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

Figure 3: Identifying Adsorption Sites Via Single-Molecule Manipulations. (a) Typical current-voltage spectra, I/V, acquired at the center of the CuTCPP representative of a successful displacement. The black arrow shows an abrupt jump revealing the molecule displacement due to tunneling electrons injection into the HOMO resonance. The process is independent of the initial charge state of the molecule. (b) CuTCPP+ translations along the TiO2 rows. The red crosses show positions of the tip where I/V spectra were acquired to induce the translation. The yellow dots show the hydroxyl positions located below the molecule of the previous adsorption configurations. (c) CuTCPP0 molecule translations. Upon displacements, the CuTCPP0 molecules (image 1) systematically change their STM appearance to the CuTCPP+ one (image 2). The white circles in images 2 reveal the initial adsorption site of the CuTCPP0 molecules consisting of five hydroxyl groups below the molecule. (d) Schematic of the CuTCPP charge state modification by lateral displacement. Scale bars are equal to 2 nm.

10

ACS Paragon Plus Environment

Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 4: Relaxed CuTCPP Configurations Obtained From DFT Calculations. (a-b) Top and side views of calculated adsorption geometries of CuTCPP on bare rutile TiO2 (110) and with hydroxyls underneath, respectively. (a) CuTCPP+ relaxed structure on the bare rutile titania. The molecule is center on top of an oxygen atoms of in a oxygen rows with the four carboxylic end groups linked to the nearby oxygen rows. (b) CuTCPP0 relaxed structure on the hydroxylated rutile TiO2 (110). The conformation of the molecule is weakly influenced by the presence of the hydroxyls. The central copper atom is lifted up by + 42 pm. DFT calculations show that the net charge of the the CuTCPP0 is + 0.28 e whereas the CuTCPP+ net charge is + 0.85 e. (c) Projected density of states (PDOS) of the CuTCPP+ (red) and CuTCPP0 (blue), respectively. ph-COOH side groups are tilted by ∼ 27◦ angle with respect to the surface plane and weakly bound to the oxygens of the nearby rows. 16 Figure 4b shows the top and side views of the CuTCPP0 adsorption geometries on rutile TiO2 (110) with the five hydroxyls group underneath. At first glance, the molecular geometry does not seem to be particularly modified by the presence of those hydroxyls, as the conformation and adsorption site resemble the CuTCPP+ one (Figure 4a) and the tilting of the ph-COOH side groups remains unchanged (∼ 27◦ ). A detailed look, however, reveals that the central Cu atom of the of CuTCPP0 porphyrin is lifted up by ≈ 40 pm due to the hydroxyl underneath. Indeed, the 11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

Cu-O distance becomes 300 pm compared to 260 pm in case of CuTCPP+ . We think that the CuTCPP0 adsorption configuration is likely governed by the hydrogenation of the central metal atom of the macrocycle as recently reported in the literature with metal-free porphyrin on TiO2 (110) or MnPc on Au(111). 36,37 Through ligand formation, the porphyrin macrocyle slightly adapts it geometry by lifting up the central Cu atom by 40 pm, value which is in good agreement with recent studies (∼ 50 pm). 36,38 As suggested by the electron density analysis to compute the net charge on each molecule (See Methods), we also conclude that none of the molecules are chemically decoupled from the surface. In agreement with our former assumption, the CuTCPP0 carries less positive charge than the CuTCPP+ . Their respective charges are +0.28 and +0.85 e and thus differ by less than one elementary charge e. As will be discussed later on, the charge analysis reveals that the hydroxyls also modify the charge distribution across the molecules resulting in different electrostatic potential maps. The corresponding dipole moments considering a point charge at the center of the molecules are thus µ0 = 4 D and µ+ = 10.6 D for the CuTCPP0 and CuTCPP+ , respectively. From that, the expected dipole-induced work function change can be estimated using the Helmholtz equation :

∆φ =

enµ ǫ0

(1)

with e the elementary charge, µ the dipole moment and ǫ0 the vacuum permittivity. To obtain the maximum ∆φ, n corresponds to the dipole density that we approximated by the molecule density for a full monolayer coverage, i.e. n = 1/(1.3×1.3 nm2 ). Considering the dipole moments µ+ and µ0 of the CuTCPP+ and CuTCPP0 obtained from DFT, the maximum work function changes are thus expected to be ∆φCuTCPP+ = 230 meV and ∆φCuTCPP0 = 80 meV, respectively. The difference between the STM images of the two configurations reflects the change of the electronic structure due to the presence of hydroxyls. In some cases, a simple scenario

12

ACS Paragon Plus Environment

Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

can explain such a difference if the molecule charge state changes by an integer number of electrons. Then some frontier molecular orbitals (MO) become completely filled or empty and different STM contrasts are obtained depending on which MOs are empty and thus contribute to the STM image. In our case of partially charged molecules, such an argument is not applied and we need to look more carefully into the details of the density of states (DOS) of the two configurations. The projected DOS (PDOS) on individual atoms resembles, in principle, the differential conductance dI/dV at the position of the corresponding atoms. In other words, the PDOS on atoms are a measure of their relative contribution to the STM contrast. Since the main difference between the experimental STM images of the two molecules is related to the pronounced pyrole sub-units of the porphyrin macrocycle, we illustrate in Figure 4c the calculated PDOS for that site. The Fermi energies are superimposed and set to zero for convenience. For comparison, we also show the PDOS for the central Cu atom. The PDOS peaks below and above the Fermi energy are shifted towards higher energies for the more positively charged molecule (CuTCPP+ ), which is in agreement with the experimental STS for the filled states at the molecule centers (see H+ and H0 in Figure 2d). (Note that the comparison is valid only qualitatively; a quantitative comparison between energies of DFT and experiment is not possible due to the known failure of DFT to estimate empty state eigenenergies.) At sample voltages of V = 1.8 V, the empty states are relevant in the STM experiment and thus we focus on the PDOS peak above the Fermi energy. In the CuTCPP+ case, the peak goes beyond the voltage window captured by the STM experiment of Figure 2d, and the pyrole are no longer pronounced contrary to the CuTCPP0 case. Interestingly, the LUMO resonance peaks of Figure 4c are shifted by ∼ 0.2 eV between the CuTCPP+ and the CuTCPP0 . As a result, the electron injection barrier of the molecule lying on the five hydroxyl groups is lowered by ∼ 0.2 eV and thus improved with respect to the CuTCPP+ . This is also in good agreement with the work function change between CuTCPP+ and CuTCPP0 that corresponds to ∆φCuTCPP+ - ∆φCuTCPP0 = 150 meV.

13

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

Figure 5: Charge State of the Molecular Configurations. (a) LCPD ( x, y) map of the CuTCPP+ molecule. The inset shows the corresponding molecule orientation observed by STM. (b) Typical ∆ f (V ) spectra at the TiO2 (blue) and CuTCPP+ (black) showing a ∆V ∗ shift of ∼ -50 mV towards more negative values. (c) V ∗ ( X ) profile along the CuTCPP+ molecule (black) and along the TiO2 (110) (blue). (d) LCPD ( x, y) map of the CuTCPP0 molecule. The inset shows the corresponding molecule orientation observed by STM. (e) Typical ∆ f (V ) spectra at the TiO2 (blue) and CuTCPP0 (black) showing a ∆V ∗ shift of ∼ +20 mV in the direction of more positive values. (f) V ∗ ( X ) profile along the CuTCPP0 molecule (black) and above the TiO2 (blue). Scale bars are 1 nm. (g) Top and side views of the calculated electron density difference between the CuTCPP+ and CuTCPP0 . Charge States of The CuTCPP Configurations Determined by Force Spectroscopy.

To

further gain insights into the partial change of the porphyrin charge states of the two aforementioned adsorptions, we experimentally mapped the local contact potential difference, LCPD ( x, y), between the tip and each molecule. 22,41–43 To do so, single force-bias spectroscopic curves, ∆ f (V ), were acquired at 77 K above each molecules in a predefined grid as described in the Methods. From the ∆ f (V ) spectra, we extracted the LCPD value 14

ACS Paragon Plus Environment

Page 15 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

corresponding to the top of the parabola, LCPD = V ∗ , via a fitting procedure. The two datasets of Figure 5 were obtained with the same tip termination and at same relative tip height without tip changes between the data acquisition. Figures 5a and d show the LCPD maps of the CuTCPP+ and CuTCPP0 molecules, respectively. A sub-molecular contrast is clearly observed arising from the difference of electrostatic forces between the molecules and the tip. In both maps, the V ∗ value at the TiO2 surface is ∼ -245 mV revealing that tip condition was identical during the successive measurements. The LCPD variations of the molecules however differ with respect to TiO2 . The LCPD above the CuTCPP+ at the center of the molecule is shifted towards more negative values by ∼ -50 mV compared to the titania surface (Figure 5b), whereas the CuTCPP0 LCPD values are shifted towards more positive values by ∼ +18 mV depending on the location (Figure 5e). To better visualize the variation of LCPD above the two molecules, V ∗ ( X ) cross-sections are shown in Figures 5c and f acquired along the black arrows of Figures 5a and b, respectively. The blue curves in both graphs correspond to the TiO2 surface obtained along the blue arrows of the Figures 5a and d. The CuTCPP+ (Figure 5c) shows a much more negative V ∗ values over the molecule with respect to the bare TiO2 surface and its variation is rather large (∆V ∗ of -50 mV) across the molecule. The most negative LCPD (-300 mV) appears close to the ph-COOH side groups. For the CuTCPP0 , the LCPD value vary in a smaller range (between -230 mV and -250 mV) around the LCPD value of ∼ -240 mV of the titania. In contrast to the CuTCPP+ , the ph-COOH side groups show more positive LCPD values (∼ -230 mV) than the underlying titania surface (∼ -245 mV) as well as two of the four pyrole sub-units, appearing brighter and located above hydroxyl groups. The two pyrole units, which are not on top of hydroxyl groups, however show a more negative LCPD than the titania in close analogy to the CuTCPP+ case. In comparison to CuTCPP0 molecules, the whole CuTCPP+ molecule shows negative V ∗ values which corresponds to overall positive charging and is in agreement with assumptions based on the STS/DFT results and previous works. 16 The positive charging of the

15

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

porphyrin is induced by a charge transfer from the molecule to the TiO2 upon adsorption (Figure 4) which partially depopulate the molecule of its electrons due to the acceptor nature of the titania. As a result, an upward dipole moment arises at the molecule/surface interface. 8 Adsorbed on five hydroxyl groups (Figure 4b), the molecule macrocyle is slightly decoupled which decreases the amount of transferred charges to the titania resulting in a decrease of this interfacial dipole moment. Strikingly, the DFT calculations of the two configurations (Figure 4) conclude to a electron depletion of both adsorption configurations and thus, the interfacial dipole moment should point in both case in the same direction. We think that the difference of LCPD observed above the ph-COOH side groups between the two adsorption configurations can be assigned to inducing negative charge to those particular cycles as a result of strong dipoles from the hydroxyls underneath. Indeed, previous works 8 as well as our charge analysis demonstrate that the H atoms of the hydroxyls completely donate their electrons to the surface generating strong dipoles below the molecule. The top and side views of the electron density difference between the two states are shown in Figure 5g. The hydroxyls polarize the ph-COOH side groups in such a way that more electrons are accumulated above them, thus inducing a local shift of the LCPD to higher values. Therefore, the presence of the central hydroxyl below the molecule modifies the overall charge state of the molecule causing different average LCPD values whereas, the four others hydroxyls below the ph-COOH provoke a change of the distribution of the electronic cloud across the molecule impacting the LCPD contrast. We believe that the site-specific adsorption geometries reported here are not restricted to the CuTCPP molecule and should arise with various porphyrins with different metallic cores on hydroxylated TiO2 . However, varying the central metal atom might change the donor character of the porphyrin with respect to the titania surface. This might in turn change the amount of charge transfer between the molecule and the surface and thus the final interfacial dipole moment. Within the macroscopic picture discussed in DSSCs devices, our measurements con-

16

ACS Paragon Plus Environment

Page 17 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

firm that a dipole moment appears between the CuTCPP+ molecule and the TiO2 upon adsorption. 16 However, our study also questions the general knowledge of the interface between the photo-sensitizers and the titania at the molecule scale. The observation of a second charge state (CuTCPP0 ), being local, site- and surface- chemistry dependent, is of particular importance in the classic picture of DSSCs since it shows a decrease of the molecule dipole moment caused by the presence of hydroxyl groups. Considering that TiO2 materials are catalysts well-known for water splitting, 4 the production of hydroxyl groups at the surface is likely and should not be considered as a minor effect within titania based organic/inorganic devices. For instance, an increase of the -OH density at the TiO2 surface might explain the drop of the photo-activity of some photo-sensitizers with time under illumination. We believe that future improvements of DSSC technologies might thus consider a better control of the TiO2 surface chemistry which seems as important as the synthesis of new well-designed photo-sensitizers.

CONCLUSIONS Combining scanning tunneling microscopy (STM), atomic force microscopy (AFM) and density functional theory (DFT) study shows the adsorption of a Cu-porphyrin derivative terminated with peripheral carboxyphenyl side groups on a hydroxylated rutile-TiO2 (110) surface. Two charge states of single porphyrin molecules are identified by STM/AFM at the atomic scale. Via single molecule manipulations, we found that these charge states are local effects originating from the presence or not of hydroxyl groups below the molecules. Both experiments and DFT calculations further demonstrate that the electronic coupling of molecules with the underlying TiO2 is locally altered by these hydroxyls underneath. Our results underline the fundamental role of hydroxyl groups of the TiO2 surface on the electronic properties of single dye molecules at the molecular level.

17

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

AUTHOR INFORMATIONS Corresponding Authors *E-mails:[email protected] [email protected] ORCID R´emy Pawlak : 0000-0001-8295-7241 Notes The authors declare no competing financial interest. Author Contribution R.P., T.G., B.S., M.S. and E.M. planned the project. R.P. performed the experiments and analysed the data with A.S., R.J., A.H., T.M., L.Z., P.O, S.G. R.P. wrote the manuscript. A.S. and S.G. performed the numerical calculations. All co-authors discussed the results and revised the manuscript.

ACKNOWLEDGEMENTS We acknowledge financial support from the Swiss National Science Foundation (SNSF), the Swiss Nanoscience Institute (SNI), the Swiss National Supercomputing Centre (CSCS) under project s499, the COST-action MP1303 and the Joint Polish-Swiss Research Programme PSRP-085/2010 ”Molecular assemblies on semiconductors and insulating surfaces”.

18

ACS Paragon Plus Environment

Page 19 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

REFERENCES References (1) O’Regan, B.; Gr¨atzel, M. A Low-Cost High-Efficiency Solar Cell Based on DyeSensitized Colloidal TiO2 Films. Nature 1991, 353, 737. (2) Diebold, U. The Surface Science of Titanium Dioxide Surf. Sci. Rep. 2003, 48, 53. (3) Wang, Q., Biener, J.; Guo, X.-C.; Farfan-Arribas, E.; Madix, R. J. Reactivity of Stoichiometric and Defective TiO2 (110) Surfaces toward DCOOD Decomposition. J. Phys. Chem. C 2003, 107, 11709–11720. (4) Bikondoa, O.; Pang, C. L.; Ithnin, R.; Muryn, C. A.; Onishi, H.; Thornton, G. Direct Visualization of Defect-Mediated dissociation of water on TiO2 (110). Nature Mater. 2006, 5, 189–192. ´ B.; P´erez, (5) Sanchez-Sanchez, C.; Mart´ınez, J. I.; Lanzilotto, V.; Biddau, G.; Gomez-Lor, ´ Chemistry and Temperature-Assisted ´ R.; Floreano, L.; Lopez, M. F.; Mart´ın-Gago, J. A. Dehydrogenation of C60 H30 Molecules on TiO2 (110) Surfaces. Nanoscale 2013, 5, 11058– 11065. (6) Kolmer, M.; Zuzak, R.; Ahmad Zebari, A. A.; Godlewski, S.; Prauzner-Bechcicki, J. S.; Piskorz, W.; Zasada, F.; Sojka, Z.; Bl´eger, D.; Hecht, S. et al. On-Surface Polymerization on a Semiconducting Oxide: Aryl Halide Coupling Controlled by Surface Hydroxyl Groups on Rutile TiO2 (011). Chem. Comm. 2015, 51, 11276–11279. (7) Addou, R.; Senftle, T. P.; O’Connor, N.; Janik, M. J.; van Duin, A. C. T. Batzill, M. Influence of Hydroxyls on Pd Atom Mobility and Clustering on Rutile TiO2 (110). ACS Nano 2014, 8, 6321–6333.

19

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

(8) Onoda, J.; Pang, C. L.; Yurtsever, A.; Sugimoto, Y. Subsurface Charge Repulsion of Adsorbed H-Adatoms on TiO2 (110). J. Phys. Chem. C 2014, 118, 13674–13679. (9) Mathew, S.; Yella, A.; Gao, P.; Humphry-Baker, R.; Gr¨atzel, M. Dye-sensitized Solar Cells With 13% Efficiency Achieved Through the Molecular Engineering of Porphyrin Sensitizers. Nat. Chem. 2014, 6, 242–247. ´ (10) Godlewski, S.; Tekiel, A.; Prauzner-Bechcicki, J.; Budzioch, J.; Gourdon, A.; Szymonski, M. Adsorption of Organic Molecules on The TiO2(011) Surface: STM Study. J. Chem. Phys. 2011, 134, 224701. (11) Godlewski, S.; Tekiel, A.; Piskorz, W.; Zasada, F.; Prauzner-Bechcicki, J.S.; Sojka, Z.; ´ Szymonski, M. Supramolecular Ordering of PTCDA Molecules: The Key Role of Dispersion Forces in an Unusual Transition from Physisorbed Into Chemisorbed State. ACS Nano 2012, 6, 8536–8545. ˇ ¨ M.; (12) Kley, C. S.; Dette, C.; Rinke, G.; Patrick, C. E.; Cechal, J.; Jung, S. J.; Baur, M.; Durr, Rauschenbach, S.; Giustino, F. et al. Atomic-Scale Observation of Multiconformational Binding and Energy Level Alignment of Ruthenium-Based Photosensitizers on TiO2 Anatase. Nano Lett. 2014, 14, 563–569. (13) Lackinger, M.; Janson, M. S.; Ho., W. Localized Interaction of Single Porphyrin Molecules With Oxygen Vacancies on TiO2 (110). J. Phys. Chem. Phys. 2012, 137, 234707. (14) Ishida, N.; Fujita, D. Adsorption of Co-Phthalocyanine on the Rutile TiO2 (110) Surface: A Scanning Tunneling Microscopy/Spectroscopy Study. J. Phys. Chem. C 2012, 116, 20300–20305. (15) Wang, C.; Fan, Q.; Hu, S.; Ju, H.; Feng, X.; Han, Y.; Pan, H.; Zhu, J.; Gottfried, J. M. Coordination Reaction Between Tetraphenylporphyrin and Nickel on a TiO2 (110) Surface. Chem. Commun. 2014, 50, 8291–8294.

20

ACS Paragon Plus Environment

Page 21 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

¨ ´ (16) Johr, R.; Hinaut, A.; Pawlak, R.; Sadeghi, A.; Goedecker, S.; Such, B.; Szymonski, M.; Meyer, E.; Glatzel, T. Characterization of individual molecular adsorption geometries by atomic force microscopy: Cu-TCPP on rutile TiO2 (110). Chem. Phys. 2015, 143, 094202. ¨ (17) Zajac, L.; Olszowski, P.; Godlewski, S.; Bodeka, L.; Such, B.; Johr, R.; Pawlak, R.; Hinaut, A.; Glatzel, T.; Meyer, E. et al. Self-Assembling of Zn Porphyrins on a (110) Face of Rutile TiO2 - The Anchoring Role of Carboxyl Groups. Appl. Surf. Sci., 2016, 379, 277–281. (18) Pawlak, R.; Kawai, S.; Fremy, S.; Glatzel, T.; Meyer., E. Atomic-Scale Mechanical Properties of Orientated C60 Molecules Revealed by nc-AFM. ACS Nano 2011, 5, 6349–6354. (19) Fremy, S.; Kawai, S.; Pawlak, R.; Glatzel, T.; Baratoff, A.; Meyer, E. Three-Dimensional Dynamic Force Spectroscopy Measurements on KBr(001): Atomic Deformations at Small Tip-Sample Separations. Nanotechnology 2012, 23, 055401. (20) Pawlak, R.; Kawai, S.; Fremy, S.; Glatzel, T.; Meyer, E. High-Resolution Imaging of C60 Molecules Using Tuning-Fork-Based Non-Contact Atomic Force Microscopy. J. Phys.: Condens. Matter 2012, 24, 084005. (21) Pawlak, R.; Fremy, S.; Kawai, S.; Glatzel, T.; Fang, H.; Fendt, L.-A.; Diederich, F.; Meyer, E. Directed Rotations of Single Porphyrin Molecules Controlled by Localized Force Spectroscopy. ACS Nano 2012, 6, 6318–6324. (22) Kawai, S.; Sadeghi, A.; Feng, X.; Lifen, P.; Pawlak, R.; Glatzel, T.; Willand, A.; Orita, A.; Otera, J.; Goedecker, S. et al. Obtaining Detailed Structural Information about Supramolecular Systems on Surfaces by Combining High-Resolution Force Microscopy with Ab Initio Calculations. ACS Nano 2013, 7, 9098–9105.

21

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

(23) Kresse, G.; Hafner, J. Ab initio Molecular Dynamics for Liquid Metals. Phys. Rev. B 1993, 47, 558–561. (24) Kresse, G.; Hafner, J. Ab initio Molecular-Dynamics Simulation of the Liquid-MetalAmorphous-Semiconductor Transition in Germanium. Phys. Rev. B 1994, 49, 14251– 14269. ¨ (25) Kresse, G.; Furthmuller, J. Efficiency of ab-initio Total Energy Calculations for Metals and Semiconductors Using a Plane-Wave Basis Set. Comp. Mat. Sci. 1996, 6, 15–50. ¨ J. Efficient Iterative Schemes for ab initio Total-Energy Calcu(26) Kresse, G.; Furthmuller, lations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, 11169–11186. (27) Grimme, S. Semiempirical GGA-Type Density Functional Constructed With a LongRange Dispersion Correction. J. Comput. Chem. 2006, 27, 1787. (28) Bader, R. F. Atoms in Molecules; Wiley Online Library, 1990. (29) Henkelman, G.; Arnaldsson, A.; Jnsson, H. Comput. Mater. Sci. 2006, 36, 354. (30) Acharya, D. P.; Ciobanu, C. V.; Camillone, N.; Sutter, P. Mechanism of ElectronInduced Hydrogen Desorption from Hydroxylated Rutile TiO2 (110). J. Phys. Chem. C 2010, 114, 21510–21515. (31) Matthiesen, J.; Wendt, S.; Hansen, J. O.; Madsen, G. K. H.; Lira, E.; Galliker, P.; Vestergaard, E. K.; Schaub, R.; Lægsgaard, E.; Hammer, B. et al. Observation of All the Intermediate Steps of a Chemical Reaction on an Oxide Surface by Scanning Tunneling Microscopy. ACS Nano 2009, 3, 517. (32) Rangan, S.; Coh, S.; Bartynski, R. A.; Chitre, K. P.; Galoppini, E.; Jaye, C.; Fische, D. Energy Alignment, Molecular Packing, and Electronic Pathways: Zinc(II) Tetraphenyl¯ Surfaces. J. Phys. Chem. porphyrin Derivatives Adsorbed on TiO2 (110) and ZnO(1120) C 2012, 116, 23921–23930. 22

ACS Paragon Plus Environment

Page 23 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(33) Rangan, S.; Ruggieri, C.; Bartynski, R.; Martinez, J.I.; Flores, F.; Ortega, J. DenselyPacked ZnTPPs Monolayer on The Rutile TiO2(110)-(1x1) Surface: Adsorption Behavior and Energy Level Alignment. J. Phys. Chem. C Nanomater Interfaces., 2016, 120, 4430–4437. (34) Schulz, F.; Ij¨as, M.; Drost, R.; H¨am¨al¨ainen, S. K.; Harju, A.; Seitsonen, A. P.; Liljeroth, P. Many-Body Transitions in a Single Molecule Visualized by Scanning Tunnelling Microscopy. Nat. Phys. 2015, 11, 229–234. (35) Mielke, J.; Hanke, F.; Peters, M.V.; Hecht, S.; Persson, M.; Grill, L. Adatoms Underneath Single Porphyrin Molecules on Au(111). J. Am. Chem. Soc. 2015, 137, 1844-1849. (36) Lovat, G.; Forrer, D.; Abadia, M.; Dominguez, M.; Casarin, M.; Rogero, C.; Vittadini, A.; and Floreano, L. Hydrogen Capture by Porphyrins at The TiO2 (110) Surface. Phys. Chem. Chem. Phys. 2015, 17, 30119. (37) Yang, K.; Liu, L.; Zhang, L.; Xiao, W.; Fei, X.; Chen, H.; Du, S.; Ernst, K.-H.; Gao, H.-J. Reversible Achiral-to-Chiral Switching of Single Mn–Phthalocyanine Molecules by Thermal Hydrogenation and Inelastic Electron Tunneling Dehydrogenation. ACS Nano 2014, 8, 2246–2251. (38) Albrecht, F.; Bischoff, F.; Auw¨arter, W.; Barth, J. V.; Repp, J. Direct Identification and Determination of Conformational Response in Adsorbed Individual Nonplanar Molecular Species Using Noncontact Atomic Force Microscopy. Nanoletters 2016, 16, 7703–7709. ¨ (39) Swart, I.; Sonnleitner, T.; Niedenfuhr, J.; Repp, J. Controlled Lateral Manipulation of Molecules on Insulating Films by STM Nano Lett. 2012, 12, 1070–1074. (40) Chung, H.J.; Yurtsever, A.; Sugimoto, Y.; Abe, M.; Morita, A. Kelvin Probe Force Microscopy Characterization of TiO2 (110)-Supported Au Clusters. App. Phys. Lett. 2011, 99, 123102. 23

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

(41) Gross, L.; Mohn, F.; Liljeroth, P.; Repp, J.; Giessibl, F.J.; Meyer, G. Measuring the Charge State of an Adatom with Noncontact Atomic Force Microscopy Science 2009, 324, 1428–1431. (42) Mohn, F.; Gross, L.; Moll, N.; Meyer, G. Imaging The Charge Distribution Within a Single Molecule. Nat. Nanotechnol. 2012, 7, 227–231. (43) Schuler, B.; Liu, S. X.; Geng, Y.; Decurtins, S.; Meyer, G.; Gross, L. Contrast Formation in Kelvin Probe Force Microscopy of Single π-Conjugated Molecules. Nano Lett. 2014, 14, 3342–3346.

24

ACS Paragon Plus Environment

Page 25 of 25

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment