Ethanol Solvation in Water Studied on a Molecular Scale by

Institute of Physics, Federal University of Bahia, 40.210-340, Salvador, BA, Brazil b. Department of Physics and Astronomy, Uppsala University, P.O. B...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF NEWCASTLE

Article

Ethanol Solvation in Water Studied on a Molecular Scale by Photoelectron Spectroscopy Ricardo R. T. Marinho, Marie-Madeleine Walz, Victor Ekholm, Gunnar Öhrwall, Olle Bjorneholm, and Arnaldo Naves de Brito J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.7b02382 • Publication Date (Web): 17 Jul 2017 Downloaded from http://pubs.acs.org on July 24, 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 B 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 30

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

Ethanol Solvation in Water Studied on a Molecular Scale by Photoelectron Spectroscopy Ricardo R. T. Marinhoa, Marie-Madeleine Walzb,c, Victor Ekholmb, Gunnar Öhrwalld, Olle Björneholmb and Arnaldo Naves de Britoe* a b

c

Institute of Physics, Federal University of Bahia, 40.210-340, Salvador, BA, Brazil

Department of Physics and Astronomy, Uppsala University, P.O. Box 516, SE-751 20 Uppsala, Sweden

Department of Cell and Molecular Biology, Computer and Systems Biology, P.O. Box 596, SE751 24, Uppsala, Sweden d

e

MAX IV Laboratory, Lund University, P.O. Box 118, SE-221 00 Lund, Sweden

Institute of Physics “Gleb Wataghin”, University of Campinas, 13083-859 Campinas -SP, Brazil

*Corresponding author: [email protected] phone:+55 19 981985211

ACS Paragon Plus Environment

1

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 2 of 30

ABSTRACT

Due to the amphiphilic properties of alcohols, hydrophobic hydration is important in alcoholwater system. In the present paper we employ X-ray photoelectron spectroscopy (XPS) to investigate the bulk and surface molecular structure of ethanol-water mixtures from 0.2 to 95 mol%. The observed XPS binding energy splitting between the methyl C1s and hydroxymethyl C1s groups, (BES_[CH3 CH2OH]), as a function of the ethanol molar percentage can be divided into different regions: one below 35 mol% with higher values (about 1.53 eV) and one starting at 60 up to 95 mol% with 1.49 eV as an average value. The chemical shifts agree with previous quantum mechanics/ molecular mechanics (QM/MM) calculations: T. Löytynoja, J. Niskanen, K. Jänkälä, O. Vahtras, Z. Rinkevicius, H. Ågren, J. of Phys. Chem. B 2014, 118, 13217. According to these calculations, the BES_[CH3 CH2OH] is related to the number of hydrogen bonds between the ethanol and the surrounding molecules. As the ethanol concentration increases, the average number of hydrogen bonds decreases from 2.5 for water-rich mixtures to 2 for pure ethanol. We give an interpretation for this behavior based on how the hydrogen bonds are distributed according to the mixing ratio. Since our experimental data is surface sensitive, we propose that this effect may also be manifested at the interface. From the ratio between the XPS C1s core lines intensities we infer that below 20 mol%, the ethanol molecules have their hydroxyl groups more hydrated and possibly facing the solution´s bulk. Between 0.1 and 14 mol%, we show the formation of an ethanol monolayer at approx. 2 mol%. Several parameters are derived for the surface region at monolayer coverage.

ACS Paragon Plus Environment

2

Page 3 of 30

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

Introduction Alcohols are used in a wide range of applications, such as fuel, starting materials for organic synthesis, additives for varnishes and solvents. Several applications involve the solvation of alcohols in aqueous solution. Ethanol, for instance, is of great importance, not only due to the use of alcoholic beverages dating back to the stone age 2, but also as a biofuel, that may replace fossil fuels in the coming decades. Pure ethanol is highly hygroscopic, and if exposed to ambient humidity, water is rapidly incorporated into its network. Although ethanol and water are fully miscible macroscopically, from a molecular point of view they do not mix very well. Indeed, aqueous ethanol leads to binary solutions, which turn out to have smaller entropy than expected compared to an ideal solution of randomly mixed molecules 3. Furthermore, a non-linear profile in viscosity has been observed upon changing the mixing ratio 4.

On the molecular scale, water makes an average of 4 H-bonds with other water molecules, two donors and two acceptors, whereas ethanol can form two, one donor and one acceptor with other ethanol molecules. When ethanol is mixed with water, the mixed environment allows for more hydrogen bonds for ethanol, when both ethanol-ethanol and ethanol-water bonds are counted this H-bond pattern is changed depending on the mixing ratio. For example when a single ethanol molecule is surrounded by water, statistically, half of the ethanol molecules will form three Hbonds according to ref.

1

. Due to the hydrophilic hydroxyl group and the hydrophobic alkyl

chain of alcohols, the characteristics of alcohol solvation in water are commonly attributed to the phenomenon of hydrophobic hydration. Investigating the microscopic composition, Nagasaka et. al.5 found three different local structures around the methyl group of methanol−water binary solutions (CH3OH)X(H2O)1−X at different concentrations manifesting themselves in C K-edge

ACS Paragon Plus Environment

3

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 4 of 30

soft X-ray absorption spectroscopy (XAS). These results were discussed with the aid of molecular dynamics (QM/MM) simulations. Löytynoja et. al. 1 performed studies on the ethanolwater binary system using QM/MM. They calculated chemical shifts between the two ethanol C1s XPS lines for several ethanol-water mixing ratios in the bulk. Their calculations predicted a U-shaped curve for the BES_[CH3 CH2OH] between 5 and 30 molar percentage (mol%) with a minimum around 20 mol%, which was interpreted as a consequence of changes in the hydrogenbonding network and rearrangement of ethanol chains as function of concentration. The error bar in the calculations, however did not allow this feature to be fully established. Several studies may be of interest to help elucidate this peculiar U-shaped behavior. Starting at very low ethanol concentrations in solution, 1 mol%, 6 and 7, tetrahedral water structures were found. Between 20 and 80 mol%, a polymer like structure of ethanol

8

has been reported. This could be of interest

to the present study, since this would lead to a locally higher concentration of ethanol molecules, where these polymer structures are formed. Compton scattering experiments performed on ethanol-water mixtures

9

are consistent with the presence of distinct regimes concerning

geometrical arrangement: around 5 mol% there is an increase of the hydrogen bonding distance, while above 15 mol% they observe an increase of the local density of ethanol as compared to pure ethanol. Egashira and Nishi

10

report non-ideal mixtures of ethanol and water at the

molecular level. Furthermore, they propose a model where ethanol molecules form clusters similar to a sandwich structure with a double layer caused by hydrophobic interactions. In the literature at least one phase transition in the mixture between ethanol and water is usually indicated. This structural phase transition takes place between 10 and 20 mol%, see for example reference

6

and references therein. The reported number of structural phase transitions is not

restricted to one, e.g. between 10 and 20 mol% Pradhan et. al. observed another transition7 .

ACS Paragon Plus Environment

4

Page 5 of 30

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

Also, clathrate-like structures at 7 and 20 mol%, which could be described as clusters of ethanol molecules surrounded by water molecules, have been reported, see for example 6. In contrast to the discussed progresses in understanding the bulk properties of the binary mixture, studies of the liquid-vapour interface are scarce in the literature. In the present work we employ surface sensitive photoelectron spectroscopy using around 360 eV photon energy to investigate the surface of the ethanol-water mixture from 0.1 to 14 mol% as well as from 16 to 36 mol% . From 5 to 95 mol% we apply higher photon energy, 600 eV, as an attempt to probe the properties of the solution’s bulk. Carbon 1s XPS spectra were obtained for the ethanol-water binary mixture as function of mol %. The kinetic energy of the ejected photoelectron determines the probing depth of the signal. Larger photon energies will result in more energetic photoelectrons being ejected from the sample increasing the probing depth. For photons around 70 eV above the ionization threshold (i.e. 360 eV) practically only the outer most liquid molecular layer is probed. We will compare results from two different depths and discuss their interpretation. Experimental Section The XPS measurements employing 360 and 600 eV photon energies were performed at MAXlab, Lund, Sweden at the I411 undulator beamline

11,12

, while the XPS measurements with 380

and 580 eV photon energies were performed at the Brazilian National Synchrotron Facility (Laboratório Nacional de Luz Síncrotron – LNLS), Campinas, Brazil, at a plane grating monochromator (PGM) beamline 13. To perform XPS on the aqueous ethanol solutions on these two synchrotron facilities, liquid microjet set-ups purchased from Microliquids GmbH applied

15,16

. Details on this technique can be found e.g. in reference

15

14

were

. The liquid microjet

(diameter ≈ 18 - 20 µm, flow ≈ 0.5 ml/min, T ≈ 10 °C) is injected through a glass nozzle into an

ACS Paragon Plus Environment

5

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 6 of 30

evacuated analysis chamber. Photoionization by linearly polarized synchrotron light occurs at approx. 1 mm after the injection point, before the liquid jet breaks up into droplets and is frozen in a cold trap. In both experimental chambers the photoelectrons are detected by a hemispherical electron analyzer (Scienta R4000) mounted perpendicular to the propagation direction of the liquid jet. For MAX-lab measurements those photoelectrons are collected at 54.7° relative to the polarization plane of the synchrotron light to minimize angular distribution effects

17

. The total

experimental resolution at the applied photon energies is around 1.25 eV for EPhoton = 600 eV and around 0.3 eV for EPhoton = 360 eV, as determined from the width of the water gas phase valence band 1b1 state. At LNLS the total resolution is 0.1 eV for EPhoton = 380 eV, 0.2 eV for EPhoton = 580 eV. All spectra were energy calibrated against the binding energy of the 1b1 state (HOMO) of liquid water (EB (1b1, liquid water) = 11.16 eV

18

) and intensity-normalized (against photon

flux and acquisition time). Aqueous solutions of ethanol (purity ≥ 99.8 %, Sigma-Ald., prod. number 34852) were prepared from deionized water (Millipore Direct-Q, R > 18.2 MΩ cm) with concentrations in the range of 0.2 – 95 mol%. To avoid charging of the liquid jet due to photoionization and electrokinetic charging 19 all solutions contained 25 mM of NaCl. The solute was monitored via the C1s signal for 5 - 95 mol% using EPhoton = 600 eV. At this photon energy, the C1s photoelectrons have a kinetic energy of approx. 310 eV, making the XPS measurements more bulk sensitive 17, as the effective attenuation length is estimated to be in the order of 2 nm 20,21

. To compare different experimental runs and to monitor the stability of the measurements,

the 1b1 valence band state of liquid water was measured between the ethanol-water solutions. The 1b1 liquid valence band was used as an internal intensity reference. The intensities of these reference measurements are constant within ± 5 – 10 %. Another set of measurements for 0.2 14.3 mol% was performed at EPhoton = 360 eV, resulting in a kinetic energy of approx. 70 eV and

ACS Paragon Plus Environment

6

Page 7 of 30

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

attenuation length of 0.5 – 1 nm, which corresponds to the most surface sensitive data. Additionally, the range between slightly below 16 and 36 mol% was studied using 380 and 580 eV photons. The photoelectron spectra were fitted using a least-squares method. Two symmetric Voigt line profiles for the liquid phase signal were employed and four asymmetric PCI line profiles for the gas phase signal from the solute. The lifetime width for C1s core holes corresponding to the Lorentzian width was set to 0.1 eV

22

. Gaussian widths were free

parameters, but linked such that they were the same for the corresponding peaks in all spectra. Energy positions and intensities were free parameters. The contributing gas phase signal that is present in the XPS spectra was fitted by linking the energy splitting and the intensity ratio to a ‘pure’ gas phase spectrum. This spectrum was taken just before or after a typical series of measurements by lowering the microjet in a way that only the vapor surrounding the jet was irradiated by the synchrotron beam.

Results In Fig. 1 we show XPS C1s spectra of ethanol-water mixtures with different molar percentages, together with the gas phase spectrum from the vapour just above the liquid surface is shown. The conditions in our experiment are such that the vapour and the liquid signal are both present. In Fig. 1d one can see that the four most intense fitted peaks (solid line) in the spectra come from the gas phase contribution. Gas phase spectra were taken at two mixing ratios, 5 and 95 ethanol mol%. Despite the fact that Fig. 1d shows the spectrum taken at 95 mol%, we verified that high and low concentration spectra give the same parameters to the fitting. Several fitting strategies were employed in order to evaluate the confidence of the used parameters. For example, we fitted the lower concentration spectrum from the liquid, 5 mol% together with the

ACS Paragon Plus Environment

7

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 8 of 30

gas phase spectrum. The parameters obtained for the gas phase peaks were then fixed and used in the fitting procedure applied to the other concentrations. This fitting strategy is depicted in Fig. 1. The gas phase C1s spectra were fitted using four asymmetric PCI line profiles. Their relative

Figure 1. C1s XPS spectra of ethanol-water mixtures at different molar percentages ((a) 24 mol%, (b) 35 mol%, (c) 95 mol%). Red dots represent the raw data, while solid (gas phase) and dashed (liquid phase) lines are fitted curves (EPhoton = 600 eV). The gas phase spectrum (d) taken from the vapour formed above the liquid is also presented. See text for more details.

ACS Paragon Plus Environment

8

Page 9 of 30

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

intensity width and position was initially decided based on the high-resolution gas phase spectra 23

. Basically the vibrational progressions and conformational effects required us to use two peaks

Figure 2. Total area of the C1s gas (square) and liquid (dot) phase signal versus molar percentage of ethanol (acquired with 360 eV photon energy). The liquid phase signal is fitted with a Langmuir curve (for more details see text).

for each C1s gas phase structure in order to obtain a reasonable fit. The C1s spectra from the liquid phase (dashed line) is composed of two chemically shifted components originating from the methyl group at lower binding energy, below 290 eV, and another component connected to the hydroxymethyl group situated at around 291 eV. In Fig. 2 we show total area of the Carbon 1s peaks from ethanol obtained using 360 eV photon energy. The resulting high surface sensitivity allowed us to monitor a possible ethanol monolayer formation, which was already pointed out by other techniques 24. The overall shape of

ACS Paragon Plus Environment

9

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 30

the curve representing the C1s liquid total area as a function of the mol% is consistent with formation of an ethanol monolayer around 2 mol% (based on the slope), and in agreement with the evaluation done for longer alcohols

25

. The molar percentage at which the monolayer is

formed according to the Sum-Frequency Generation Vibrational Spectroscopy (SFG-VS)

24

is

somewhat higher than the one in the present study. This discrepancy can be due to different depth probed by each technique. In Fig. 2 we also plot the XPS signal from the gas phase (squares). Despite the fact that the intensity of the gas phase signal is very dependent on the alignment between the photon beam and the microjet, the overall shape observed in Fig. 2 indicates that the alignment has been reliable. In Fig. 2 a strongly non-linear behavior of the liquid signal is observed between 0.2 and 2 mol%, as discussed above, while the gas phase signal shows a linear-like trend in this short range. Comparison of the two signals plotted in Fig. 2 suggests that the molecular surface concentration changes are much stronger than the partial vapour pressure. One obvious conclusion from this is that the ethanol molecules in the vapour come also from the layers below the surface monolayer, which should have lower ethanol concentration. A standard Langmuir adsorption model, see equation below, is applied to fit the liquid phase signal (compare Fig. 2) in order to determine the ethanol surface concentration as well as the Gibbs free energy of adsorption,

In this equation

, at monolayer coverage 26, 27.

is the surface contribution of the photoelectron signal as a function of the

bulk mol%. More specifically, in our case, this number represents the amount of ethanol at the surface in terms of the bulk concentration of water and ethanol, mol%. Notice that

ACS Paragon Plus Environment

scales

10

Page 11 of 30

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

with the surface concentration of ethanol.

is the maximum surface concentration of

ethanol, i.e. assumed to be equal to the concentration of pure ethanol. of ethanol in the bulk and

is the molar fraction

is therefore the bulk water molar fraction.

Using the above formula, see reference Walz et al.

25

for more details, we evaluate the Gibbs

free adsorption energy to be -10.8 kJ/mol, and at monolayer coverage, i.e. around 2 mol%, the ethanol surface concentration is determined to be around 3.2 x 1014 molecules per cm2 and the molecular area is calculated to be approx. 31 Å2. Interestingly, we confirm again a maximum in the surface enrichment factor (EF = csurface/cbulk; EFmax = 65) at approx. 1/3 of the monolayer formation concentration, which we have observed earlier for longer linear alcohols, see reference 25

. For the gas phase signal a more extended range in mol% is presented in Fig. 3. By gas phase

signal we mean the C1s summed areas, which is proportional to the amount of molecules overlapping with the photon beam. From this graph one sees a steep slope below 20 mol% and a smaller slope above 35 mol% supporting the existence of a higher ethanol partial vapor pressure compared to the liquid bulk mol% at low concentration. This information is by no means new 28 and has been used to separate a mixture of two or more liquids by fractional distillation. Ethanol and water are expected to have an azeotropic mixture at about 91.5 mol% (at atmospheric pressure). Below this point, at several vapour pressure conditions, a higher pressure of ethanol is observed as compared to the ratio of the mixture. This is explained by a stronger bonding between water molecules and ethanol molecules themselves, compared to the bonding between ethanol and water molecules. From the present measurements we find a correlation between the methyl and hydroxylmethyl groups C1s total areas and the vapour pressure behavior. It seems that the higher number of hydrogen bonds at lower concentration, predicted by the QM/MM

ACS Paragon Plus Environment

11

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 30

simulations 1, indicates more unfavorable interactions between ethanol and water compared to ethanol-ethanol and water-water interactions. Furthermore, from the results presented in Fig. 2 complemented by the data presented in Fig. 3 it is clear that there is a correlation between the

Figure 3. Total area of the C1s gas phase signal versus molar percentage of ethanol (EPhoton = 600 eV). Please notice that the measurements presented in Fig. 2 have a different photoionization cross-section from the present spectra; consequently the number of counts cannot be compared.

partial vapour pressure of ethanol and its surface concentration.

The measurements shown in Fig. 1 were performed at 600 eV photon excitation energy. At this photon energy the ejected photoelectrons have about 300 eV kinetic energy. Considering a homogeneous ethanol distribution the amount of signal from the surface monolayer (ASSM) would be between one third and one fourth of the total signal. This is however not the case as our

ACS Paragon Plus Environment

12

Page 13 of 30

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

experimental evidences, from Fig. 2, show an ethanol monolayer formation. In the following we will attempt to evaluate the ASSM taking into account this fact.

Figure 4. Total area of the C1s liquid phase signals (methyl and hydroxymethyl) versus molar percentage of ethanol (EPhoton = 600 eV). The intensity between the fitted solid and dotted line originates from the surface. Below the dotted line the signal originates from the bulk. See text for details.

An estimate of the bulk to surface contribution to the photoelectron signal can be obtained by analyzing the total intensity of the C1s line as a function of the mol% of ethanol in the solution. Fig. 4 shows such an experiment. In it, the total area of the C1s liquid ethanol signal is presented between 5 to 95 mol%. A linear fit to these points is also added in the figure showing good agreement with the measurements. This linear behavior is consistent with the fact that the C1s

ACS Paragon Plus Environment

13

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 30

signal is proportional to the linear increase of the ethanol concentration in the bulk. According to Fig. 2, at zero concentration of ethanol, the C1s liquid ethanol signal should be zero. This fact is represented by the black dot in the appropriate position in Fig. 4. The position of this point compared to the linear fit implies that a non-linear behavior must be present between 0 and 5 mol%, since extrapolation of the linear fit, using points from 5 to 20 mol%, predicts a total area of 9.36 K a. u. at 0 mol%. Indeed, this non-linear behavior is an obvious consequence of the monolayer formation described previously. As the concentration of ethanol increases, at the nonlinear part of the curve, ethanol molecules are pushed towards the surface. This process takes place until a monolayer of ethanol is formed. Above this concentration the added intensity to the photoelectron signal can only come from the bulk contribution, which increases linearly with the molar percentage. We may use the linear behaviour to estimate the surface contribution to the total intensity as follows. If the straight line is extrapolated to a situation at zero ethanol concentration, we should get zero intensity if the entire signal comes from the bulk. The intersection with the axis will thus give a lower bound for the intensity from the monolayer, which is constant at any concentration range, after this monolayer is formed and as long as it is present. Therefore, the signal, up to approximately 10 K a. u., must come from the molecules in the monolayer. As a consequence at 15 mol% ethanol (at 15.7 K a. u. ) 40% of the signal comes from bulk, while 57% will come from the bulk at 30 mol%. At 80 mol% as much as 78 % of the signal originates from the bulk region. This will be important to help figuring out whether the effects to be discussed are representative of which regions in the solution. In order to further clarify this point we have added a dotted line in Fig. 4. All intensity above that line originates from the surface, while the bulk contribution to the signal lies below that line. Fig. 5 shows the C1s area ratio between the methyl and hydroxymethyl peaks from aqueous phase at different

ACS Paragon Plus Environment

14

Page 15 of 30

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

molar percentages. According to the error bars displayed in the figure the resulting trend in the ratios seem to show two different groups of values. In order to explain this fact one must keep in mind that our signal is also surface sensitive, as well as, that a monolayer is formed at somewhat lower concentration of ethanol. Furthermore, one should consider the QM/MM calculations from

Figure 5. C1s area ratio between the methyl and hydroxymethyl peak (EPhoton = 600 eV). reference1. According to these, in a range of concentrations starting with a single molecule of ethanol surrounded by water molecules and up to the mixing ratio of 30 mol%, the number of hydrogen bonds made by the ethanol molecules is 2.5 in average. This number is reduced to 2 in pure ethanol. Also, in reference 8 it is reported that a polymer-like structure of ethanol is formed for 20 and 80 mol% and that these structures could be bridged by water molecules. Inside the liquid or at the surface, the hydroxylmethyl group belonging to the ethanol molecule is certainly the one most likely to make the hydrogen bonds, not the methyl group. Indeed, the carbonhydrogen bonds are known to be only weakly polar, and they are thus less likely to form further hydrogen bonds. Therefore, the larger number of hydrogen bonds predicted by the QM/MM

ACS Paragon Plus Environment

15

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 30

calculations could imply that there will be more water around the hydroxymethyl groups at lower concentration. These extra atoms should attenuate the photoelectron signal from the hydroxymethyl even more. In the case of the surface molecules, one may even infer that the methyl groups are sticking out of the surface at concentrations up to 30 mol%. Therefore, the photoelectrons ejected by the C1s methyl group will be less prompt to scattering by the nonexisting neighbour water molecules than the C1s hydroxymethyl photoelectrons, which are most likely facing towards the bulk and are more surrounded by water molecules. Both, bulk and surface ionized ethanol molecules are likely to contribute to the observed ratio. However, the surface ethanol molecules are to a greater extent responsible for the change in the ratio, as the neighbour atom concentration shows a much more pronounced change.

In Fig. 6 we present the BES_[CH3 CH2OH] as a function of the molar percentage. The shifts can be divided into two regions: one below 35 mol% with higher values (about 1.53 eV) and a second region starting at 60 mol% up to 95 mol% with 1.49 eV as an average value. It is perhaps worth to mention that at 15 mol% and 35 mol% the error bar allows us to claim that the shifts at these concentrations are distinct from those above 60 mol%. In the figure we also included results from recent theoretical QM/MM calculations.

1

The overall behavior predicted by the

QM/MM calculations is confirmed by our experimental data. Details of the binding energy splitting with extra measurement points obtained both at the MAX-lab and LNLS beamlines between 15 to 35 mol% are shown in this figure. Several photon energies were employed showing similar results. Both calculations and the present experiment show a U-shape curve of the chemical shift between 5 and 30 mol%. According to the 600 eV data the error bars do not

ACS Paragon Plus Environment

16

Page 17 of 30

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

allow to fully establish the U-shape curve however with the additional points from 380 eV and 580 eV, shown in the inset, one may argue that the experimental data support this feature. Furthermore, one may notice that the feature is rather similar for photon energies around 380 eV and 600 eV. Given the fact that in this region approximately half of the signal comes from the surface for 600 eV photons, see Fig. 4, one must conclude that both surface and bulk may contribute to the changes observed. Discussion

ACS Paragon Plus Environment

17

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 30

Let us now discuss the overall BES_[CH3 CH2OH] decrease from 0-100 mol%, which is both experimentally observed and predicted by QM/MM calculations. At low concentration the ethanol molecules are surrounded by water molecules. Water itself makes an average of 4 Hbonds with other water molecules (two donors and two acceptors). This is in line with the result

ACS Paragon Plus Environment

18

Page 19 of 30

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

that the number of H-bonds made by the ethanol molecules increases with the number of surrounding water molecules, see Fig. 6 top scale. One needs however to explain why the increase in the number of hydrogen bonds, leads to an increase in the BES_[CH3 CH2OH]. First it is important to consider that the solvation effect on an ethanol molecule tends to reduce the

ACS Paragon Plus Environment

19

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 30

binding energy of the emitted photoelectrons. As for example see Fig. 1a, the C1s peak from the liquid is always significantly lower in binding energy compared to the gas phase. The reason is well understood but, for the sake of completeness, we repeat it here: the surrounding molecules in the liquid share part of their electronic cloud to shield the ionized carbon atom, thus reducing

ACS Paragon Plus Environment

20

Page 21 of 30

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

its positive charge. The outgoing photoelectron will be less attracted by a less positive carbon

Figure 6. Binding energy splitting between the C1s peaks originating from the methyl and hydroxymethyl in aqueous phase (EPhoton = 380 to 600 eV). The inset shows details from two different set-ups and three different photon energies. *Calculation results from reference

1

are also depicted for comparison. The average number of hydrogen bonds

formed by the ethanol molecule is also shown in the upper axis. atom and as a result its kinetic energy, EKinetic , will be higher and the binding energy, EBinding smaller since

. Notice that the photon energy, hν , is kept fixed in all cases.

Let us consider the situation of 100 mol%, i.e. pure ethanol, which is presented in the right side

ACS Paragon Plus Environment

21

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 30

of Fig. 6.. Ethanol forms one donor and one acceptor H-bond per molecule. Fig. 7 shows a typical configuration, which should be representative enough for our discussion. As can be seen, the solvation, substantially enhanced by the H-bonding, is concentrated at the -CH2OH group and not at the -CH3. This solvation will tend to reduce the binding energy of the C1s –CH2OH, bringing its binding energy closer to the C1s -CH3, thus reducing the chemical shift C1s(CH3) – C1s(CH2OH) in the aqueous ethanol solution compared to the gas phase. Note that the weakly positive hydrogen atoms, bonded to the carbon atoms, repel the hydrogen atoms from other ethanol molecules, thus the solvation effects via these hydrogen atoms are reduced. For the sake of accuracy we must point out that the reduction in the C1s biding energy due to solvation will vary depending on the character of the H-bonds formed

29

. We will elaborate on this further

down in the discussion. As water is added to the solution, the hydrogen bonds between the ethanol molecules start to be substituted by those between water and ethanol. When very few ethanol molecules are left, and consequently they behave as single molecules (SM) surrounded by water, the calculations and experiments are consistently predicting an increased BES_[CH3 CH2OH], which is correlated with the formation of an extra H-bond in about half of the ethanol molecules. The calculation shows also a tendency for the C1s -CH3 binding energy to decrease and the C1s -CH2OH binding energy to increase with the water concentration. A decrease in the -CH3 binding energy could be explained by a third donor H-bond being made statistically to a higher frequency between the oxygen atoms in water and one of the hydrogen atoms from the -CH3 (see ref.

30

showing the possibility to be feasible). An increased C1s -

CH2OH binding energy cannot be explained by a donor H-bond being made to one of the hydrogen in -CH2- belonging to the -CH2OH group. Also a naive interpretation would point to the fact that a H-bond made to the oxygen atom in the -CH2OH group would increase the

ACS Paragon Plus Environment

22

Page 23 of 30

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

solvation in the group increasing the C1s -CH2OH binding energy, in disagreement with our observations. We must evaluate if an alternative explanation for the observed increase in the BES_[CH3 CH2OH] at low mol% based on donor and acceptor H-bond formation is valid. At low molar fraction half of the molecules form in average a third H-bond according to the calculations1. The other two H-bonds formed by the -OH group in the ethanol molecule will be considered, for simplicity, as being one donor (D) and one acceptor (A). For the third H-bond we have three possibilities: (1) an A H-bond is made to the oxygen in the -OH group, (2) a D Hbond is made by one of the hydrogen atoms belonging to the -CH2- in the -CH2OH group, (3) a D H-bond is made by one of the hydrogen atoms in the -CH3 group. Hypothesis (2) and (3) are exactly what was discussed in the last paragraph therefore we are left with the discussion of hypothesis (1). In this case, the surrounding water and the -OH group from ethanol will form a DAA hydrogen bond pattern. According to ref. 29, for an oxygen atom in water, its O1s binding energy is actually higher by as much as 200 meV as compared to the same oxygen in a water molecule performing a DA bond pattern. A reasonable approximation in this context is to consider that the O1s in ethanol will also present a similar behavior. Therefore, the extra A Hbond will increase the O1s binding energy. The -CH2- carbon bonded to this oxygen will also have a higher binding energy, if we use polarizability arguments, but to a lesser extent31. This change causes an increase in the BES_[CH3 CH2OH] in agreement with our experimental results. We conclude that the donor acceptor concept is useful to explain our observation, and that the H-bond pattern described in hypotheses 1 and 3 could contribute to the observed BES_[CH3 CH2OH]. High precision measurements of the absolute binding energy of the C1s CH2OH and C1s -CH3 could indicate how probable hypothesis (1) is as compared to (3).

ACS Paragon Plus Environment

23

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 30

One perhaps interesting observation from Fig. 6 is that at high ethanol mol% the experiment and calculations are in a rather good agreement, while at low mol% the experiment predicts a lower BES_[CH3 CH2OH] compared to the calculations. This discrepancy may be explained by the fact that the QM/MM calculations were made for the solution bulk, while at low mol% the experimental signal comes mostly from the ethanol at the surface. Ethanol molecules in the bulk are surrounded by other molecules in a more even fashion compared to the situation at the surface. According to our results presented in Fig. 5, the hydroxyl group from the ethanol molecule is more hydrated, if the ethanol molecules are “standing” at the surface and the hydroxyl group is pointing toward the solution bulk. The higher degree of hydration for the hydroxymethyl carbon will stabilize the ion formed upon photoionization of the hydroxymethyl carbon more than that from ionizing the methyl carbon. Since the binding energy of the hydroxymethyl carbon is higher than that of the methyl carbon, the shift between the C 1s lines will decrease. The situation in the bulk, however, may be compared to the gas phase. For both bulk and gas phase, the environment around the hydroxyl and methyl groups are more similar than for ethanol molecules at the surface. The chemical shift between the two carbon lines in ethanol in the gas phase is around 1.59 eV obtained from fitting our gas phase spectrum which is in line with recent high-resolution XPS spectrum, which report 1.64 eV and 1.43 eV for the anti and gauge conformers respectively. In fact, an average of the two conformers will be observed in the XPS spectrum, which would lead to 1.53 eV chemical shift 23. This supports our observation that both the QM/MM calculations predict a larger chemical shift than the more surface sensitive measurements.

ACS Paragon Plus Environment

24

Page 25 of 30

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 7. A possible typical geometrical configuration for pure ethanol solution i.e. 100 mol %. At this concentration one ethanol molecule makes an average of 2 H-bonds. The hydrogen atoms connected to the carbon atoms are weakly positive while the hydrogen bonded to the oxygen is “strongly” positive and able to make a donor H-bond. The oxygen atom is negatively charged and can make one acceptor H-bond. According to these, the change in the binding energy shifts is related to the number of hydrogen bonds between the ethanol molecule and the surrounding molecules. As the ethanol concentration increases, the average number of hydrogen bonds decreases from 2.5 to 2 for pure ethanol. The observed chemical shifts are small, but a trend can be identified, both from experiment and from simulations, which are in good agreement. The observed experimental trend, however, yields another conclusion. Since the experiment is also surface sensitive, it is suggested that the average number of hydrogen bonds at the surface also decrease. The U-shaped curve, seen both in the experiment and previous calculations, are within the error bars. With the additional experimental points presented in the inset in Fig 6, one may

ACS Paragon Plus Environment

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

Page 26 of 30

advocate its existence. More experiments, with smaller X-ray focus, and consequent smaller gas phase signal may help to further establish this feature. Out of sheer curiosity one may ask what is the interpretation for its existence? A hint for its explanation is the lowering of the chemical shift in conjunction with the reduction of the number of H-bonds. In addition, some authors report the formation of clathrate-like structures at specific mol%6,7,9. In general, these structures would resemble the one shown in Fig. 7 with additional water surrounding the structure. Regions like this will tend to provide a reduction of the BES_[CH3 CH2OH] consistent with the bottom of the U-shape region.

Conclusion We have employed photoelectron spectroscopy together with the liquid microjet technique to study the properties of the ethanol-water mixtures between 0.1 and 95 mol%. We have identified the formation of an ethanol monolayer at around 2 mol%, which is somewhat smaller than previously evaluated by SFG-VS data. Applying a standard Langmuir adsorption isotherm to model the surface sensitive data, we determine a Gibbs free energy of adsorption around -10.8 kJ/mol, and a surface concentration of around 3.2 x 1014 molecules per cm2, and we calculate a molecular area of approx. 31 Å2 at monolayer coverage. Previous QM/MM simulations predicted that the average number of hydrogen bonds decreases from 2.5 to 2 as the ethanol concentration increases, and these are confirmed by the our experimental data. An interpretation was presented describing the mechanism behind this behavior. The observed trend in the more bulk sensitive experimental data leads us to another conclusion based on the fact that our experimental data for 600 eV photons is also surface sensitive: the change in the average number of hydrogen bonds

ACS Paragon Plus Environment

26

Page 27 of 30

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

may also occur at the solutions surface. This is in line with the much higher ethanol concentration at the surface that is established already at rather low bulk concentrations. The area ratio between the methyl and hydroxymethyl group is consistent with the interpretation that at the surface the methyl groups are sticking out of the surface up to 35 mol%. The ratio is also in line with the predicted change in the bonding network. A U-shaped curve in the chemical shift is observed between 5 and 30 mol% with a minimum around 25 mol% which is in line with QM/MM simulations.

Acknowledgment Financial support from the STINT-CAPES Proc. 99999.009805/2014-01, Swedish Research Council (VR), Swedish Foundation for Strategic Research, and the Carl Tryggers foundation is gratefully acknowledged. MAX-lab, Lund University, Sweden, and LNLS – Brazil are acknowledged for the allocation of beamtime and support from their staff. We would like to acknowledge the Brazilians funding agencies CNPq-Brazil Projs. 480967-2013, 304142/2015-8 and CAPES-Brazil Proj. A66-2013.

References 1. Löytynoja, T.; Niskanen, J.; Jänkälä, K.; Vahtras, O.; Rinkevicius, Z.; Ågren, H. Quantum Mechanics/Molecular Mechanics Modeling of Photoelectron Spectra: The Carbon 1s Core–Electron Binding Energies of Ethanol–Water Solutions. J. Phys. Chem. B 2014, 118, 13217-13225. 2. Guerra-Doce, E. The Origins of Inebriation: Archaeological Evidence of the Consumption of Fermented Beverages and Drugs in Prehistoric Eurasia. J. Archaeol. Method Th. 2015, 22, 751-782. 3. Frank, H. S.; Evans, M. W. Free Volume and Entropy in Condensed Systems III. Entropy in Binary Liquid Mixtures; Partial Molal Entropy in Dilute Solutions; Structure and Thermodynamics in Aqueous Electrolytes. J. Chem. Phys. 1945, 13, 507-532. 4. Mikhail, S. Z.; Kimel, W. R. Densities and Viscosities of Methanol-Water Mixtures. J. Chem. Eng. Data 1961, 6, 533-537.

ACS Paragon Plus Environment

27

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 28 of 30

5. Nagasaka, M.; Mochizuki, K.; Leloup, V.; Kosugi, N. Local Structures of Methanol– Water Binary Solutions Studied by Soft X-Ray Absorption Spectroscopy. J. Phys. Chem. B 2014, 118, 4388-4396. 6. Dolenko, T. A.; Burikov, S. A.; Dolenko, S. A.; Efitorov, A. O.; Plastinin, I. V.; Yuzhakov, V. I.; Patsaeva, S. V. Raman Spectroscopy of Water-Ethanol Solutions: The Estimation of Hydrogen Bonding Energy and the Appearance of Clathrate-Like Structures in Solutions. J. Phys. Chem. A 2015, 119, 10806-10815. 7. Pradhan, T.; Ghoshal, P.; Biswas, R. Structural Transition in Alcohol-Water Binary Mixtures: A Spectroscopic Study. J. Chem. Sci. 2008, 120, 275-287. 8. Takamuku, T.; Yamaguchia, T.; Asato, M.; Matsumoto, M.; Nishi, N. Structure of Clusters in Methanol-Water Binary Solutions Studied by Mass Spectrometry and X-Ray Diffraction. In Z. Naturforsch. A, 2000; Vol. 55, p 513. 9. Juurinen, I.; Nakahara, K.; Ando, N.; Nishiumi, T.; Seta, H.; Yoshida, N.; Morinaga, T.; Itou, M.; Ninomiya, T.; Sakurai, Y., et al. Measurement of Two Solvation Regimes in WaterEthanol Mixtures Using X-Ray Compton Scattering. Phys. Rev. Lett. 2011, 107, 197401. 10. Egashira, K.; Nishi, N. Low-Frequency Raman Spectroscopy of Ethanol−Water Binary Solution:  Evidence for Self-Association of Solute and Solvent Molecules. J. Phys. Chem. B 1998, 102, 4054-4057. 11. Bassler, M.; Forsell, J. O.; Bjorneholm, O.; Feifel, R.; Jurvansuu, M.; Aksela, S.; Sundin, S.; Sorensen, S. L.; Nyholm, R.; Ausmees, A., et al. Soft X-Ray Undulator Beam Line I411 at Max-II for Gases, Liquids and Solid Samples. J. Electron. Spectrosc. Relat. Phenom. 1999, 101, 953-957. 12. Bassler, M.; Ausmees, A.; Jurvansuu, M.; Feifel, R.; Forsell, J. O.; Fonseca, P. D.; Kivimaki, A.; Sundin, S.; Sorensen, S. L.; Nyholm, R., et al. Beam Line I411 at Max II Performance and First Results. Nucl. Instrum. Methods Phys. Res., Sect. A 2001, 469, 382393. 13. Cezar, J. C.; Fonseca, P. T.; Rodrigues, G. L. M. P.; Castro, A. R. B. d.; Neuenschwander, R. T.; Rodrigues, F.; Meyer, B. C.; Ribeiro, L. F. S.; Moreira, A. F. A. G.; Piton, J. R., et al. The U11 PGM Beam Line at the Brazilian National Synchrotron Light Laboratory. J. Phys.: Conf. Ser. 2013, 425, 072015. 14. Microliquids GmbH, http://www.microliquids.com, (accessed March 8). 15. Winter, B.; Faubel, M. Photoemission from Liquid Aqueous Solutions. Chem. Rev. 2006, 106, 1176-1211. 16. Bergersen, H.; Marinho, R. R. T.; Pokapanich, W.; Lindblad, A.; Bjorneholm, O.; Saethre, L. J.; Ohrwall, G. A Photoelectron Spectroscopic Study of Aqueous Tetrabutylammonium Iodide. J. Phys.: Condens. Matter 2007, 19, 326101. 17. Hüfner, S. Photoelectron Spectroscopy. 3 ed.; Springer-Verlag Berlin Heidelberg: Berlin, 2003; p 662. 18. Winter, B.; Weber, R.; Widdra, W.; Dittmar, M.; Faubel, M.; Hertel, I. Full Valence Band Photoemission from Liquid Water Using EUV Synchrotron Radiation. J. Phys. Chem. A 2004, 108, 2625-2632. 19. Preissler, N.; Buchner, F.; Schultz, T.; Lübcke, A. Electrokinetic Charging and Evidence for Charge Evaporation in Liquid Microjets of Aqueous Salt Solution. J. Phys. Chem. B 2013, 117, 2422-2428. 20. Nikjoo, H.; Uehara, S.; Emfietzoglou, D.; Brahme, A. Heavy Charged Particles in Radiation Biology and Biophysics. New J. Phys. 2008, 10, 075006.

ACS Paragon Plus Environment

28

Page 29 of 30

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

21. Ottosson, N.; Faubel, M.; Bradforth, S. E.; Jungwirth, P.; Winter, B. Photoelectron Spectroscopy of Liquid Water and Aqueous Solution: Electron Effective Attenuation Lengths and Emission-Angle Anisotropy. J. Electron. Spectrosc. Relat. Phenom. 2010, 177, 60-70. 22. Campbell, J. L.; Papp, T. Widths of the Atomic K–N7 Levels. At. Data Nucl. Data Tables 2001, 77, 1-56. 23. Abu-Samha, M.; Borve, K. J.; Saethre, L. J.; Thomas, T. D. Conformational Effects in Inner-Shell Photoelectron Spectroscopy of Ethanol. Phys. Rev. Lett. 2005, 95, 103002. 24. Sung, J. H.; Park, K.; Kim, D. Surfaces of Alcohol-Water Mixtures Studied by SumFrequency Generation Vibrational Spectroscopy. J. Phys. Chem. B 2005, 109, 18507-18514. 25. Walz, M. M.; Werner, J.; Ekholm, V.; Prisle, N. L.; Ohrwall, G.; Bjorneholm, O. Alcohols at the Aqueous Surface: Chain Length and Isomer Effects. Phys. Chem. Chem. Phys. 2016, 18, 6648-6656. 26. Onorato, R. M.; Otten, D. E.; Saykally, R. J. Adsorption of Thiocyanate Ions to the Dodecanol/Water Interface Characterized by UV Second Harmonic Generation. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 15176-15180. 27. Perrine, K. A.; Van Spyk, M. H. C.; Margarella, A. M.; Winter, B.; Faubel, M.; Bluhm, H.; Hemminger, J. C. Characterization of the Acetonitrile Aqueous Solution/Vapor Interface by Liquid-Jet X-Ray Photoelectron Spectroscopy. J. Phys. Chem. C 2014, 118, 29378-29388. 28. Beebe, A. H.; Coulter, K. E.; Lindsay, R. A.; Baker, E. M. Equilibria in Ethanol-Water System at Pressures Less Than Atmospheric. Ind. Eng. Chem. 1942, 34, 1501-1504. 29. Abu-samha, M.; Børve, K. J.; Winkler, M.; Harnes, J.; Sæthre, L. J.; Lindblad, A.; Bergersen, H.; Öhrwall, G.; Björneholm, O.; Svensson, S. The Local Structure of Small Water Clusters: Imprints on the Core-Level Photoelectron Spectrum. J. Phys. B: At., Mol. Opt. Phys. 2009, 42, 055201. 30. Knak Jensen, S. J.; Tang, T.-H.; Csizmadia, I. G. Hydrogen-Bonding Ability of a Methyl Group. J. Phys. Chem. A 2003, 107, 8975-8979. 31. Nordfors, D.; Agren, H. Calculation of Core Electron-Binding Energies from Structural Formulas. J. Electron. Spectrosc. Relat. Phenom. 1991, 56, 1-11.

ACS Paragon Plus Environment

29

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 30 of 30

TOC Graphic

ACS Paragon Plus Environment

30