Effective Contact Potential of Thin Film Metal-Insulator Nanostructures

Mar 7, 2017 - Dana Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02215,United States. § University of Massachusetts Lowell, ...
0 downloads 0 Views 1MB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Effective contact potential of thin film metal-insulator nanostructures and its role in self-powered nanofilm x-ray sensors Davide Brivio, Earl Ada, Erno Sajo, and Piotr Zygmanski ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b01264 • Publication Date (Web): 07 Mar 2017 Downloaded from http://pubs.acs.org on March 9, 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.

ACS Applied Materials & Interfaces 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 26

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 Applied Materials & Interfaces

Effective contact potential of thin film metalinsulator nanostructures and its role in self-powered nanofilm x-ray sensors Davide Brivio* (1), Earl Ada (2), Erno Sajo (2), Piotr Zygmanski (1)** (1) Brigham & Women’s Hospital, Boston, MA, USA, Dana Farber Cancer Institute, Boston, MA, USA, Harvard Medical School; (2) University of Massachusetts Lowell, MA, USA KEYWORDS: Contact potential difference, Work function, x-ray sensor, self-powered detector, high energy current, thin-film sensors

ABSTRACT: We studied the Effective Contact Potential Difference (ECPD) of thin film nanostructures and its role in self-powered x-ray sensors, which use the High Energy Current detection scheme. We compared the response to kilovoltage X-rays of several nanostructures made of disparate combinations of conductors (Al, Cu, Ta, ITO) and oxides (SiO2, Ta2O5, Al2O3). We measured current-voltage curves in parallel-plate configuration separated by an air gap and determined three characteristic parameters: current at zero voltage bias I0, the voltage offset for zero current ECPD, and saturation current Isat. We found that the metals’ ECPD values measured with our technique were higher than the CPD values measured with photo-electron spectroscopy in-situ, i.e., no air contact. These differences are related to natural oxidization and to the presence of photo-/Auger-electron current leaking from the high-Z towards the low-Z

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces

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 26

electrode, as suggested by additional experiments carried out in vacuum. Further, the deposition of 40 nm –500 nm oxide layer on the surface of metallic substrates strongly affects their contact potential. This technique exploits ionization and charge carrier transport in both solid insulators and in air, and it opens the possibility of measuring the ECPD between metals separated by a solid insulator in a metal-insulator-metal (MIM) configuration. Additionally, we demonstrated that certain configurations of MIM structures are suitable for X-ray detection in self-powered mode.

INTRODUCTION

High Energy Current (HEC) X-ray detection scheme was recently proposed and demonstrated by our group1–5. Several prototype HEC detectors have been developed and tested, possessing various shapes (square, rectangular, strip, triangular, or pixelated) and areas (from about a few mm2 to 1800 cm2) on rigid and flexible substrates (glass, polymer) and material combinations (Cu, Al, Pb, Ta, ITO electrodes and polymers, oxide or air as dielectrics) using either micro- or nano-film fabrication approaches (printing or electron beam physical vapor deposition). The efficiency of HEC detection is the highest for multilayer nanostructures with alternating high-Z / low-Z layers wherein photo- and Auger- electron leakage from the high-Z layers1 leads to a self-powered x-ray sensor. The main advantage of HEC structures is their suitability for medical dosimetry, radiation protection or monitoring x-ray devices, and their subsequent integration with existing equipment. They are further apt to adapt to difficult geometry and rugged conditions3,5–8.

ACS Paragon Plus Environment

2

Page 3 of 26

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 Applied Materials & Interfaces

In its simplest design, a HEC detector can work as a self-powered ionization chamber with just two different electrode materials separated by air (M1AM2), or as metal-solid insulator-metal structure (M1IM2). Ionization in gas and solid insulator is due to electrons leaking from both metals with the dominant current from the higher Z electrode. If the insulator is low Z or low density the interaction of X-rays with it is much smaller than with the high Z metal; similarly with the opposite electrode. Thus it is mostly the electron leakage from the high Z metal, which gives rise to high energy electron (particle) current, which in turn induces electric current in the external circuit and also leads to ionization of the insulator. Ionized charge carriers in turn are transported in the insulator by the contact potential. In air the charge carriers are mostly positive and negative ion pairs, and in a solid insulator they are electron/hole-like states. The low energy (eV) charge carriers as well as low energy HEC electrons can be stopped or accelerated by the external electric field or by the internal field created by contact potential difference (CPD), which is the difference in work functions between the two electrodes. When two metals with different work functions are connected (e.g., through an external resistor), electrons migrate from one metal to the other until equilibrium is reached and the Fermi levels are lined up. An electric field is generated between the two metals, E=-∆V/d (in parallel plate geometry), where ∆V is the contact potential difference (CPD) or work function (WF) difference, and (d) is the separation between the two metals. Absolute values of the work function (WF) are measured via photo-emission spectroscopy (PES)9 or X-ray PES10, while different methods have been developed to measure the CPD based on Kelvin probe microscopy11,12, ionizing radiation 13,14 or employing grazing incidence electron reflection15. It is known that work function depends on many factors, including material crystallization16, water uptake13 and on surface conditions17, tensile or compressive surface tension18 and it is increased

ACS Paragon Plus Environment

3

ACS Applied Materials & Interfaces

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 26

by thermal or natural oxidation processes19,20. Furthermore, MoO3 deposition on different metals showed an increase in the metal/insulator WF because of redistribution of electronic states21. In summary, as stated by Greiner et al20, the work function is more a measure of a material condition than a representation of a material constant, although absolute values of WF for some materials are available and are reproducible22 in controlled conditions. In this work we developed a technique for a fast assessment of the effective contact potential difference (ECPD) between hybrid materials in a parallel plate configuration by irradiating them with X-rays and by obtaining current voltage (IV) curves. We call it “effective” CPD because, as we will show in the manuscript, it accounts for the contribution of x-ray induced emitted electrons as well as for the surface oxidization and morphology and since we measure an averaged value over a large surface area if compared to other standard techniques. Our technique provides ECPD between metals separated by a solid insulator in a M1IM2 and M1IAM2 or similar configurations while the aforementioned standard techniques9–15 do not allow measuring inside a solid state M1IM2 structure since they rely on the inspection of the material surface. In our approach, the IV-curve is obtained only for a small voltage range from zero (self-powered sensor) up to few volts until the current flowing in the external circuit reaches zero value. The effective saturation value is then measured for higher voltage values (~50-100V). Based on the current-voltage we determine three main parameters of interest: ECPD, current at zero voltage bias I0 and saturation current Isat. In this work we are mostly interested in the properties of the ECPD for various combinations of metals and insulators, which lead toward the realization of a solid-state multilayer self-powered radiation detector in a M1IM2 configuration. MATERIALS AND METHODS

ACS Paragon Plus Environment

4

Page 5 of 26

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 Applied Materials & Interfaces

Our method to measure the effective CPD relies on finding the potential difference ∆V required to counterbalance the CPD of a given metal nanolayer M or metal-insulator (MI) nanostructure with respect to a reference metal layer (Mref) or metal-insulator layer (MrefI). We studied the IV curve characteristics under irradiation of nanostructures in parallel plate configurations for various combinations of metal-dielectric-metal structures, which included Metal-(solid) Insulator-Metal (MIMref), Metal-Insulator-Air-Metal (MIAMref), Metal-AirInsulator-Metal (MAIMref), Metal-Insulator-Air-Insulator-Metal (MIAIMref), and Metal-AirMetal (MAMref) with a circuit as detailed in our recent work4 and schematically shown in Figure 1a. Thin metal films (Al, Cu, Ta, ITO, 230-800nm) were deposited on 4” diameter glass wafer via Electron-Beam deposition at a deposition rate of 1-5 Å/s and pressure ranging between of 5E-6 Torr and 3E-5 Torr. Oxide layers (Ta2O5, Al2O3, SiO2, 40-500nm thick) where deposited on top of metal films via Plasma Enhanced Chemical Vapor Deposition PECVD (SiO2), electron-beam physical vapor deposition (Ta2O5) or via sputtering (Al2O3). Ta2O5 oxide was deposited via electron-beam deposition using Tantalum pellets (purity 99.95%, Kurt J. Lesker Company, Jefferson Hills, PA, USA). We also studied a commercially available 1.52mm -thick extruded Aluminum alloy 6063, with Al2O3 (500nm) grown on top by anodization. ITO (100nm) on glass substrate was commercially available. The reference metal/metal-oxide Mref/MrefI and the metal/metal-oxide sample to be measured M/MI were placed in parallel plate configuration and separated by an air gap of 1.3 mm and connected to an external circuit providing the desired voltage bias and measuring the voltage drop across a 1 MΩ shunt using a modified version of a commercial data acquisition system (DT9828 DAQ, Data Translation, Malboro, MA, USA). An electrometer was used to measure the applied voltage to the sample. X-ray source used was a

ACS Paragon Plus Environment

5

ACS Applied Materials & Interfaces

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 26

medical radiotherapy simulator (Ximatron, Varian Medical Systems, Palo Alto, CA, USA). In this study we used 120kVp and 25mA, 25mAs, equivalent to 3.845 mGy/s dose rate in air or air kerma rate, measured with a commercial ionization chamber (DCT10-MM, IBA Dosimetry, Bartlett, TN). The absorption of X-ray flux in the active layers of the samples was about 0.2% for Cu (830nm thick) and about 0.3% for Ta (230nm thick) and Ta2O5 (440nm thick), while the glass substrate (0.7mm thick) absorbs about 5% of the radiation. The corresponding inelastic mean free path for the X-ray are about 415 µm for Cu, 76 µm for Ta and about 14 mm for the glass substrate. The samples were placed at a source to surface distance SSD=100 cm and field of 3x3 cm2, unless otherwise specified. The field was centered on the samples and the field edges were few cm away from the metal edges of the samples. In Figure 1b we show an example of the characteristic IV curve and in the inset an enlargement to show the interesting features: I0 is the signal for null external voltage bias and ECPD is the external voltage bias required to counterbalance the current flowing and obtain null current. We linearly fit the points of the IV curve close to the zero current value. The ECPD was obtained from the intersection of the linear regression with the V axis (see red dashed line in the inset of Fig. 1b). We also acquired the saturation current Isat applying a bias of 50V - 100V (electric field up to 760V/cm). Aluminum was used as the reference electrode (Mref), while Al, Cu, ITO and Ta were used as electrodes under investigation (M). Insulating layers (I) SiO2, Ta2O5 or Al2O3 with thicknesses ranging from 40nm to 500nm where studied in the aforementioned configurations (MAMref, MIAMref, MAIMref, MIAIMref, MIMref).

ACS Paragon Plus Environment

6

Page 7 of 26

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 Applied Materials & Interfaces

a)

X-ray

Sample M (MI) DAQ I

R

V

air Reference Mref (MrefI)

c)

DAQ I

R

V

X-ray Vacuum chamber

Figure 1: a) Schematically shown one of the configurations MAMref ECPD measurement setup and b) characteristic current-voltage (IV) curve with an enlargement for small voltages in the inset; I0 is the signal for null external voltage bias and ECPD is the external voltage bias required to counterbalance the current to obtain null current. Isat is the saturation current. c) Schematic of the measuring circuit for the three-layer sample used in the measurements with the vacuum chamber. We also studied the IV characteristics of the MIMref configuration of nanofabricated threelayer samples deposited on the same substrate as well as two separated substrates stacked together by removing the air gap in between. Tables 1 and 2 list the investigated samples. Further studies with samples having Mref=Cu are reported in the supporting information (S1).

ACS Paragon Plus Environment

7

ACS Applied Materials & Interfaces

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 26

Table 1: Summary of the samples investigated in MAMref, MIAMref, MAIMref, MIAIMref, configuration with Mref=Al. The thickness in parenthesis refers to the insulator. Air gap was set at 1.3 mm for all samples. Al-(I)-A-(I)-Al

Al-(I)-A-(I)-Cu

Al-(I)-A-(I)-Ta

Al-air-Al

Al-air-Cu

Al-air-Ta

Al-air-SiO2-Al

Al-SiO2-air-Cu (200nm)

Al-Ta2O5-air-Ta (400nm)

Al-air-Ta2O5-Al (440nm)

Al-air-SiO2-Cu (200nm)

Al-air-Ta2O5-Ta (440nm)

Al-air-Al2O3-Al 500nm)

Al-Ta2O5-air-Cu (440nm)

Al-Ta2O5-air-Ta2O5-Ta (440nm-440nm)

Al-air-Ta2O5-Cu (440nm)

Al-Al2O3-air-Ta 500nm)

Al-Ta2O5-air-Ta2O5-Cu (440nm-440nm)

Al-Al2O3-air-Ta (sp, 40nm)

(Anodized,

Al-air-Al2O3-Al (sp, 40nm)

Al-Al2O3-air-Cu 500nm)

(Anodized,

(Anodized,

Al-Al2O3-air-Cu (sp, 40nm)

Table 2: Summary of the samples investigated in MIMref configuration with Mref=Al. The symbol “/” represent the interface between Mref(MrefI) and M(MI), i.e., the surface at which they were stacked together. The thickness in parenthesis refers to the insulator. *This sample ITOSiO2-AlCu was nanofabricated on ITO substrate and Al-Cu were deposited in pixels using a lithographic technique (pixel size 1x1cm2). 100nm of Cu was deposited on top of 1000nm of Al to provide better electric contact to the measuring system. Al-I-ITO

Al-I-Cu

Al-I-Ta

Al-Ta2O5/Ta2O5-ITO (440nm/440nm)

Al-Ta2O5/Ta2O5-Cu (440nm-440nm)

Al-Ta2O5/Ta (440nm)

Al-Ta2O5/ITO (440nm)

Al-Al2O3/Ta (Anodized, 500nm)

ITO-SiO2-Al-Cu (500nm)*

Al-Al2O3/Ta (sp, 40nm) Al-SiO2/Ta (200nm)

ACS Paragon Plus Environment

8

Page 9 of 26

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 Applied Materials & Interfaces

We independently measured the CPD of Cu, Ta and Ta-Ta2O5 with respect to Al via low intensity X-ray photoemission spectroscopy (XPS) as described by Schlaf23,24 using a VG Scientific ESCALAB MKII X-ray photoelectron spectrometer operating with a Mg source. Details and results are reported in the supporting information (S3). We also performed experiments with two/three-layer samples, Al-gap-Pb-gap-Al and Al-gapCu-gap-Al, and Al-gap-Ta, by placing them inside a vacuum chamber in order to remove the contribution due to air ionization. In the three-layer samples Al electrodes were connected together and the voltage bias was applied between the Al electrodes and the Pb/Cu electrode as shown in Figure 1c. The chamber was not sufficiently large to accommodate our samples prepared via PVD. For this reason we used thin Al, Pb and Cu foils or Al and Ta thin film PVD deposited on glass slides (1”x3”). The resulting current was the direct HEC contribution from the Photoelectrons and Auger electrons leakage only. RESULTS Study in air (MAMref, MIAMref, MAIMref, MIAIMref,) In Figure 2a we show the ECPD values found in the different cases. Figure 2b shows the variation to the ECP due to the deposition of metal oxides. ECPD as large as 1.20 was found for Al-air-Ta for Al-air-Ta2O5-Cu (1.19V). A similar value was found for Al-air-ITO (1.18V) which increased to 1.20V by adding 200nm of SiO2 on top of ITO (Al-air-SiO2-ITO sample). Data for ITO are reported in the supporting information (S2). As expected, the ECPD is dependent on the amount of oxidization of the surfaces, e.g., two Cu or two Al samples deposited on different days were showing a non-zero ECPD (using configuration

ACS Paragon Plus Environment

9

ACS Applied Materials & Interfaces

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 26

Al-air-Al or Cu-air-Cu), which is explained by the different amount/type of oxide growing on the surfaces while stored in atmospheric air. As shown by Greiner et al.20 the work function of metal oxide grown by surface oxidization increases with the oxide thickness and saturates at about 10nm. The presence of solid oxide layer on top of metals affects the electronic states and therefore the WF. This is attributed21 to oxidation-reduction reaction between the metal and the oxide at the interface and to charge transfer from the metal Fermi level into the oxide’s low-level conduction band. In our work we deposited relatively thick oxide layers (40nm-500nm) when compared to commonly used injection layer thickness where surface effect are dominant (