REVIEW OF SCIENTIFIC INSTRUMENTS 85, 103120 (2014)

Electrochemical flowcell for in-situ investigations by soft x-ray absorption and emission spectroscopy C. Schwanke,1 R. Golnak,2 J. Xiao,2 and K. M. Lange1,a) 1 Helmholtz-Zentrum Berlin für Materialien und Energie, Institute of Solar Fuels, Albert-Einstein-Straße 15, 12489 Berlin, Germany 2 Helmholtz-Zentrum Berlin für Materialien und Energie, Institute of Methods for Material Development, Albert-Einstein-Straße 15, 12489 Berlin, Germany

(Received 28 August 2014; accepted 10 October 2014; published online 30 October 2014) A new liquid flow-cell designed for electronic structure investigations at the liquid-solid interface by soft X-ray absorption and emission spectroscopy is presented. A thin membrane serves simultaneously as a substrate for the working electrode and solid state samples as well as for separating the liquid from the surrounding vacuum conditions. In combination with counter and reference electrodes this approach allows in-situ studies of electrochemical deposition processes and catalytic reactions at the liquid-solid interface in combination with potentiostatic measurements. As model system in-situ monitoring of the deposition process of Co metal from a 10 mM CoCl2 aqueous solution by X-ray absorption and emission spectroscopy is presented. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4899063] I. INTRODUCTION

Processes at the liquid-solid interface are of major importance for electrochemical applications such as batteries, fuel cells, and dye sensitized solar cells.1, 2 Optimizing such complex systems usually requires a detailed understanding of the involved processes down to the electronic structure at the molecular scale, as bond-building and -breaking reactions. Soft X-ray absorption (XAS) and emission spectroscopy (XES) are useful tools to reveal element specific information about the electronic structure of materials as oxidation state, ligand field strength, and charge transfer effects.3 In the XA process an electron from a core level is promoted to an unoccupied state probing accordingly the unoccupied valance structure of the molecule. The created core hole is filled with an electron from a higher lying occupied state while a photon can be emitted. In XES the emitted photon is detected by an energy dispersive detector, revealing information about the occupied density of states.4 Upon varying the excitation energy during XE measurements the resonant inelastic scattering spectroscopy (RIXS) pattern is recorded. RIXS can give insights into processes on the timescale of the core-hole lifetime.5 For many of the promising materials and relevant processes in the field of energy conversion, light elements such as oxygen, carbon, or nitrogen play a key-role, as well as transition metals such as titanium, cobalt, or iron. Using XAS and XES electronic structure information can be revealed element selectively via the K-edges of light elements and L-edges of the transition metals by using soft X-rays. The low penetration depth of soft x-rays in air demands vacuum conditions. Whereas well established for the study of solids,6 this requirement makes investigations on liquids challenging and led a) Author to whom correspondence should be addressed. Electronic mail:

[email protected]

0034-6748/2014/85(10)/103120/5/$30.00

recently to the development of several experimental setups, which cope with this challenge by various approaches.7–12 For studying liquids and solutes, e.g., the liquid micro-jet technique can be applied,13 for which a liquid jet formed in a glass nozzle of around 20 μm diameter is shot into the vacuum and collected in a cryotrap. For this technique a temperature gradient along the jet axis due to evaporative cooling in the vacuum environment has to be considered. Another approach is to separate vacuum and liquid by a thin X-ray transparent membrane. The original drop-between-membrane approach can be affected, e.g., by X-ray induced sample damage.14, 15 To overcome this issue the flowcell, where the liquid is flowed behind the membrane was developed. Flowcells optimized for various scientific questions and measurement techniques were designed.10–12, 16 Most recently also cells for XAS in transmission mode17 and total fluorescence yield (TFY) mode18 which allow to study the samples of interest upon applying a potential were introduced. While the first electrochemical in-situ soft x-ray measurement,19 as far as we know, followed the approach to install reference and counter electrode into the liquid, it is also possible to directly implement all three electrodes on the membrane. The latter approach was combined with x-ray microscopy by Bozzini et al.20, 21 and allowed to investigate deposition or corrosion processes as function of the distance from the electrode. In this work a new in situ flowcell for the investigation of liquid-solid interfaces is presented, which allows XA TFY, partial fluorescence (PFY), and inverse partial fluorescence yield (iPFY) as well as XE/RIXS measurements. The cell contains the option of applying an electric potential to the solid samples using a three electrodes system and of carrying out cyclic voltammetry measurements. Besides studying electrochemical deposition processes by applying a constant potential for deposition on the membrane and recording XAS and XES spectra, as presented here, also combined cyclic voltammetry (CV) and XAS/XES measurements are possible.

85, 103120-1

© 2014 AIP Publishing LLC

103120-2

Schwanke et al.

For this purpose XAS and XES spectra are recorded for various applied potentials. Three-dimensional maps of signal intensity versus applied voltage and excitation/emission energy are the result of such scans. This option allows, e.g., to determine potential dependent oxidation state changes of catalytic materials at the liquid solid interface as taking place in electrochemical cells for water splitting. In Secs. II and III the design of the in-situ cell is described in detail and first proof of principle XAS and XES measurements at the Co L-edge during the deposition of Co metal on a gold substrate are presented and compared to the spectra of Co ions in aqueous solution. II. DESIGN OF THE ELECTROCHEMICAL FLOWCELL

The flowcell for in-situ investigations of the liquid-solid interaction with soft XAS and XES under potential control is depicted schematically in Figure 1. The cell is optimized for being implemented in the LiXEdrom endstation8 at the synchrotron facility BESSY II. Monochrome X-rays enter the cell through a thin silicon nitride membrane, interact with the sample, and a part of the emitted x-ray photons leaves the cell through the same membrane. For XAS measurements the monochromatic synchrotron radiation is scanned through an absorption edge of interest and the resulting fluorescence is detected in the total fluorescence yield (TFY) mode with a GaAsP diode. XES measurements are carried out at an excitation energy above the absorption edge and upon energy dispersive detection of the resulting fluorescence. A Rowlandcircle based spectrometer is used, containing various gratings to cover an energy range from 20 eV to 1000 eV and a detection unit consisting of a stack of multi-channel plates, a phosphor screen, and a CCD camera. Alternatively to the TFY mode, which can be sometimes affected by artefacts, e.g., saturation,22 XAS measurements can also be carried out in the partial fluorescence (PFY) or inverse partial fluorescence yield (iPFY) mode by using the Rowland-circle spectrometer. Details about these methods can be found in Refs. 23 and 24. The main body of the cell is made from PEEK in order to warrant a high chemical resistance and guarantee good electrical insulation. The four feedthroughs to the inner liquid chamber serve as liquid inlet and outlet, access for the reference electrode (RE) and access for the counter electrode (CE). Easy exchange of the components is possible through the use of ferrules and screws (P-200, XP-235, IDEX Health and Science). The diameter of the tubing for the liquid exchange can be chosen to minimize the liquid volume required for

FIG. 1. Schematic of the flowcell for electrochemical in-situ XAS and XES studies.

Rev. Sci. Instrum. 85, 103120 (2014)

circulation, e.g., a 180 μm diameter results in a 800 μl total liquid volume, or to guarantee fast and reliable liquid exchange with 750 μm diameter tubing resulting in a 3.2 ml total liquid volume. The tubing leads from the liquid chamber out of the flowcell into the vacuum and is passed through a mounting before it is transferred from the inner of the LiXEdrom experimental setup via a feedthrough to the ambient and further to a glovebox. The glovebox allows also the use of liquids that require an inert gas atmosphere. For liquid exR change as well as circulation a peristaltic pump (Ismatec Reglo ICC) is used. An exploded view drawing and a picture of the flow-cell for in-situ measurements are shown in Figures 2(a) and 2(b), respectively. The liquid chamber is sealed with a 100 nm thick silicon nitride membrane with dimensions 0.5 mm × 0.5 mm on a 400 μm thick 1 cm × 1 cm silicon frame (Silson Ltd) and a Viton O-ring. The membrane and the frame are coated with 20 nm gold which serve as working electrode (WE). A 20 nm titanium layer can be employed to improve the adhesion of gold on the membrane. Furthermore the coating protects the window from the interaction with soft x-ray induced radicals in the solution and thus enhances the stability of the membrane.12 The thicknesses of the metal layers are chosen to minimize absorption while providing good electrical conductivity. The membrane is contacted in each corner of the membrane frame with steel sheets. The left and right sheet can be electrically connected with a wire to allow a homogenous current flow or they can be used to test the conductivity and the contact of the membrane coating thus facilitating the contacting procedure. The contact pressure between the steel sheets and the membrane frame needs to be adjusted to prevent breaking of the frame while providing good electrical contact. The sheets are designed to be deformed easily to fulfill these requirements. The feedthroughs for the electrodes have a diameter of 1.2 mm in order to allow RE (1 mm diameter) and CE (1 mm diameter) to be close to the working electrode. As RE a leak free microminiature Ag/AgCl electrode (Harvard Apparatus GmbH) is implemented, a platinum wire serves as CE. To minimize the uncompensated resistance the RE is placed less than 1 mm away from the working electrode. The use of a leak-free Ag/AgCl RE simplifies measurements since no additional calibration of the RE is necessary. For potentiostatic measurements and capacitance-voltage profiling an EmStat3+ by Palmsens with a maximum current of ±100 mA and a voltage range of ±4 V is used. The presented cell allows, e.g., the study of electrochemical deposition processes, as will be shown in Sec. III. The flowcell design allows furthermore to study the functionality of the resulting layers directly after such an electrochemical deposition process (e.g., of catalytic active films). For this purpose the solution for the electrochemical deposition can be replaced via the liquid inand outlet by the electrolyte solution with its additives, under which the catalytic reaction takes place under realistic conditions. This process can be carried out under inert gas in the tubing by placing the electrolyte solutions inside a glovebox with direct connection to the flowcell so that no additional sample transfer is required. For electrochemical processes temperature can play a crucial role. By placing the liquid reservoir outside of the

103120-3

Schwanke et al.

Rev. Sci. Instrum. 85, 103120 (2014)

FIG. 2. (a) Exploded view drawing of the electrochemical flowcell for soft x-ray absorption and emission. The membrane is coated with a metal, serving as working electrode (WE) and serves as support for the solid sample. Counter (CE) and reference electrode (RE) are placed in the liquid chamber. Tubes are attached to allow fast and easy liquid exchange and to prevent radiation damage. (b) Picture of the in-situ cell with cover plate, O-ring, and membrane. (c) Cobalt layer electrodeposited from aqueous CoCl2 in the electrochemical flowcell. The shape of the deposit is determined by the liquid chamber.

III. PROOF OF PRINCIPLE MEASUREMENTS

nucleation (peak A).33 During the inverse potential scan an anodic current density peak B is observed. In literature this feature was correlated to oxidation of the cobalt metal phase deposited previously or to the dissolution of cobalt from the gold surface into solution.31, 33 Compared to a previous study on 10 mM CoCl2 + 1 M NH4 Cl at pH 4.6, this feature is shifted in the here presented measurements towards more positive electrode potential values, which could be explained by different coordination spheres of the Co-ions in the 10 mM CoCl2 solution.33

In this section first proof of principle soft XAS and XES measurements are presented together with the respective cyclic voltammetry data. With regards to prospective applications for the in-situ study of CoOx catalysts upon applied voltage, as model study the deposition process of Co on a gold surface was chosen. Due to its relevance, e.g., for potential high density data storage application the deposition process of cobalt materials has been elaborately studied during the last years.25–32 In Figure 3 the cyclic voltammetric curve obtained from a 10 mM CoCl2 hexahydrate aqueous solution within the electrochemical in-situ cell is presented. As substrate served a 20 nm gold layer deposited on the backside of the Si3 N4 membrane. The scan potential rate was 100 mV s−1 . The sharp increase of cathodic current upon scanning towards negative potentials can be correlated to the inset of cobalt

FIG. 3. Cyclic voltammetric curve obtained from a 10 mM CoCl2 hexahydrate aqueous solution. The potential scan was started at 0 V towards negative potential with a scan potential rate of 100 mV s−1 . The dotted line indicates the potential that was used for the static Co-deposition.

vacuum chamber into a thermostat and cycling the liquid the temperature in the cell can be varied. The resulting temperature can be measured with a 1 mm diameter thermocouple that can be implemented via an additional feedthrough into the cell. The maximum working temperature is limited due to the PEEK body of the flowcell and the PEEK tubing. The maximum working temperature is limited by the PEEK glass transition temperature at 143◦ C.

103120-4

Schwanke et al.

In-situ soft x-ray measurements at the Co L-edge were carried out in the single bunch mode at synchrotron currents as low as 20 mA. A static potential of –1 V was applied to the WE of the in-situ cell with 10 mM CoCl2 hexahydrate for various time periods. According to Figure 3 at this potential the Co deposition is taking place. The thickness d of the metal layer resulting from the ion deposition can be estimated with the following expression:34  M d= I dt, (1) n · F · A · ρmetal where M is the atomic weight of the deposited metal, I is the current recorded during the deposition time dt, n is the number of electrons taking part in the reduction, F is the Faraday constant, A is the area of deposition, and ρ metal is the density of the deposited metal. XAS spectra were recorded before deposition, after 45 s, 145 s, and 345 s. Based on Eq. (1) the estimated thicknesses of the resulting layers are ∼0.05 μm, ∼0.18 μm, and ∼0.42 μm. The respective spectra are shown in Figure 4(a). For comparison the spectrum of aqueous CoCl2 is shown.24 Upon increasing deposition time an increase in the signal of the Coedge is observed. The spectral features after 345 s deposition time show the rather featureless characteristics of a Co metal film and differ accordingly to the more complex spectral fingerprint of the of the aqueous sample. Furthermore XE spectra were recorded upon resonant excitation to the L3 and L2 edge, shown in Figure 4(b). From bottom to top spectra with increasing time of applied voltage are shown. A clear increase

Rev. Sci. Instrum. 85, 103120 (2014)

in emission signal of the Co Lα or Lα and Lβ line are observed which proof the successful monitoring also via XES of the increasing Co layer thickness deposited on the membrane. Further tests were carried out in order to assure, that the applied voltage induced the observed deposition on the membrane. For this purpose repeated scans on the same membrane position were carried out in the same time intervals as done for the previous deposition measurements, however, without applied voltage. No increase of the Co signal was observed, which allows excluding that the previous rise was correlated to a radiation induced process instead of an electrochemical process. Furthermore, an in-situ increase and decrease of the X-ray signal was monitored upon alternating the applied voltage from +1 V to −1 V. Note that the measurements presented here were carried out at low photon flux (∼20 mA ring current). The measuring time for each point of the XA spectra was 2 s and for each XE spectrum 10 min which was already sufficient to monitor the increase of signal upon deposition and to resolve the characteristic metal electronic structure of the system of around 0.3 μm. For future measurements, where thinner layered systems could be of interest, measurements will be carried out with the maximum attainable photon flux of the BESSYII synchrotron facility (∼300 mA ring current) for a satisfying signal-to-noise ratio.

IV. SUMMARY AND OUTLOOK

In summary, we have developed a new electrochemical flowcell for in-situ soft X-ray absorption and emission measurements, which allows electronic structure studies of materials in solution and at the liquid-solid interface. Besides the TFY mode, XA spectra can be recorded in the previously introduced PFY and iPFY mode. A three electrode system implemented into the cell makes studies under applied potential possible. Furthermore, cyclic voltammetry measurements can be carried out. Measurements monitoring the deposition process from an aqueous CoCl2 solution based on XAS and XES were presented as a model system. This new cell provides the possibility of studying, e.g., the in-cell deposition of catalytic metal oxides directly followed by the investigation of the function of these materials in contact with a realistic electrolyte solution, as planned for future experiments.

ACKNOWLEDGMENTS

We thank Ivo Rudolph for the preparation of the membrane coatings and Mika Pflüger for support during the beamtime. This work was supported by the Helmholtz Association (PD-059). 1

FIG. 4. Co L-edge XAS spectra recorded from the in-situ electrochemical cell before deposition and after 45 s, 145 s, and 345 s deposition time. For comparison a spectrum of aqueous CoCl2 is shown.24 (b) XES spectra upon excitation at 778 eV and at 793 eV for increasing deposition times from bottom to up.

I. Siretanu, D. Ebeling, M. P. Andersson, S. L. S. Stipp, A. Philipse, M. C. Stuart, D. van den Ende, and F. Mugele, Sci. Rep. 4, 4956 (2014). 2 L. Collins, S. Jesse, J. I. Kilpatrick, A. Tselev, O. Varenyk, M. B. Okatan, S. A. L. Weber, A. Kumar, N. Balke, S. V. Kalinin, and B. J. Rodriguez, Nat. Commun. 5, 3871 (2014). 3 F. d. Groot, Coord. Chem. Rev. 249, 31 (2005).

103120-5 4

Schwanke et al.

K. M. Lange, E. Suljoti, and E. F. Aziz, J. Electron Spectrosc. Relat. Phenom. 188, 101 (2013). 5 J.-E. Rubensson, J. Electron Spectrosc. Relat. Phenom. 110–111, 135 (2000). 6 F. De Groot and A. Kotani, Core Level Spectroscopy of Solids (CRC Press, 2008). 7 L. Weinhardt, O. Fuchs, M. Blum, M. Bär, M. Weigand, J. Denlinger, Y. Zubavichus, M. Zharnikov, M. Grunze, and C. Heske, J. Electron Spectrosc. Relat. Phenom. 177, 206 (2010). 8 K. M. Lange, R. Konnecke, S. Ghadimi, R. Golnak, M. A. Soldatov, K. F. Hodeck, A. Soldatov, and E. F. Aziz, Chem. Phys. 377, 1 (2010). 9 K. Kunnus, I. Rajkovic, S. Schreck, W. Quevedo, S. Eckert, M. Beye, E. Suljoti, C. Weniger, C. Kalus, S. Grübel, M. Scholz, D. Nordlund, W. Zhang, R. W. Hartsock, K. J. Gaffney, W. F. Schlotter, J. J. Turner, B. Kennedy, F. Hennies, S. Techert, P. Wernet, and A. Föhlisch, Rev. Sci. Instrum. 83, 123109 (2012). 10 S. Schreck, G. Gavrila, C. Weniger, and P. Wernet, Rev. Sci. Instrum. 82, 103101 (2011). 11 M. Nagasaka, T. Hatsui, T. Horigome, Y. Hamamura, and N. Kosugi, J. Electron Spectrosc. Relat. Phenom. 177, 130 (2010). 12 T. Tokushima, Y. Harada, Y. Horikawa, O. Takahashi, Y. Senba, H. Ohashi, L. G. Pettersson, A. Nilsson, and S. Shin, J. Electron Spectrosc. Relat. Phenom. 177, 192 (2010). 13 M. Faubel, B. Steiner, and J. P. Toennies, J. Chem. Phys. 106, 9013 (1997). 14 K. M. Lange, K. F. Hodeck, U. Schade, and E. F. Aziz, J. Phys. Chem. B 114, 16997 (2010). 15 K. M. Lange, R. Könnecke, M. Soldatov, R. Golnak, J.-E. Rubensson, A. Soldatov, and E. F. Aziz, Angew. Chem., Int. Ed. 50, 10621 (2011). 16 E. F. Aziz, M. Freiwald, S. Eisebitt, and W. Eberhardt, Phys. Rev. B 73, 075120 (2006). 17 M. Nagasaka, H. Yuzawa, T. Horigome, A. P. Hitchcock, and N. Kosugi, J. Phys. Chem. C 117, 16343 (2013). 18 D. K. Bora, P.-A. Glans, J. Pepper, Y.-S. Liu, C. Du, D. Wang, and J.-H. Guo, Rev. Sci. Instrum. 85, 043106 (2014).

Rev. Sci. Instrum. 85, 103120 (2014) 19

D. Guay, J. Stewart-Ornstein, X. Zhang, and A. P. Hitchcock, Anal. Chem. 77, 3479 (2005). 20 B. Bozzini, L. D’Urzo, A. Gianoncelli, B. Kaulich, M. Kiskinova, M. Prasciolu and A. Tadjeddine, J. Phys.: Conf. Ser. 186, 012103 (2009). 21 B. Bozzini, A. Gianoncelli, P. Bocchetta, S. Dal Zilio, and G. Kourousias, Anal. Chem. 86, 664 (2014). 22 S. Eisebitt, T. Böske, J. E. Rubensson, and W. Eberhardt, Phys. Rev. B 47, 14103 (1993). 23 M. D. Gotz, M. A. Soldatov, K. M. Lange, N. Engel, R. Golnak, R. Konnecke, K. Atak, W. Eberhardt, and E. F. Aziz, J. Phys. Chem. Lett. 3, 1619 (2012). 24 M. A. Soldatov, K. M. Lange, M. D. Gotz, N. Engel, R. Golnak, A. Kothe, and E. F. Aziz, Chem. Phys. Lett. 546, 164 (2012). 25 P. Allongue, L. Cagnon, C. Gomes, A. Gundel, and V. Costa, Surf. Sci. 557, 41 (2004). 26 L. Cagnon, A. Gundel, T. Devolder, A. Morrone, C. Chappert, J. E. Schmidt, and P. Allongue, Appl. Surf. Sci. 164, 22 (2000). 27 T. Cohen-Hyams, W. D. Kaplan, and J. Yahalom, Electrochem. Solid-State Lett. 5, C75 (2002). 28 D. Hofmann, W. Schindler, and J. Kirschner, Appl. Phys. Lett. 73, 3279 (1998). 29 M. Kleinert, H. F. Waibel, G. E. Engelmann, H. Martin, and D. M. Kolb, Electrochim. Acta 46, 3129 (2001). 30 L. G. Lin, J. W. Yan, Y. Wang, Y. C. Fu, and B. W. Mao, J. Exp. Nanosci. 1, 269 (2006). 31 L. H. Mendoza-Huizar, J. Robles, and M. Palomar-Pardave, J. Electroanal. Chem. 521, 95 (2002). 32 L. H. Mendoza-Huizar, J. Robles, and M. Palomar-Pardave, J. Electroanal. Chem. 545, 39 (2003). 33 M. Palomar-Pardave, I. Gonzalez, A. B. Soto, and E. M. Arce, J. Electroanal. Chem. 443, 125 (1998). 34 H. H. Lou and Y. Huang, Encyclopedia of Chemical Processing (Taylor and Francis, 2005).

Review of Scientific Instruments is copyrighted by the American Institute of Physics (AIP). Redistribution of journal material is subject to the AIP online journal license and/or AIP copyright. For more information, see http://ojps.aip.org/rsio/rsicr.jsp