Strain Effects on the Energy-Level Alignment at Metal/OrganicSemiconductor InterfacesAinhoa Atxabal,†,⊥ Stephen R. McMillan,‡ Benat García-Arruabarrena,† Subir Parui,†,∥ Roger Llopis,†

Felix Casanova,†,§ Michael E. Flatte,‡ and Luis E. Hueso*,†,§

†CIC nanoGUNE, 20018 Donostia-San Sebastian, Basque Country, Spain‡Department of Physics and Astronomy, University of Iowa, 203 Van Allen Hall, Iowa City, Iowa 52242-1479, United States§IKERBASQUE, Basque Foundation for Science, 48013 Bilbao, Basque Country, Spain

*S Supporting Information

ABSTRACT: Flexible and wearable devices are among the upcoming trends in theopto-electronics market. Nevertheless, bendable devices should ensure the sameefficiency and stability as their rigid analogs. It is well-known that the energy barriersbetween the metal Fermi energy and the molecular levels of organic semiconductorsdevoted to charge transport are key parameters in the performance of organic-basedelectronic devices. Therefore, it is paramount to understand how the energy barriersat metal/organic semiconductor interfaces change with bending. In this work, weexperimentally measure the interface energy barriers between a metallic contact andsmall semiconducting molecules. The measurements are performed in operativeconditions, while the samples are bent by a controlled applied mechanical strain. Wedetermine that energy barriers are not sensitive to bending of the sample, but weobserve that the hopping transport of the charges in flat molecules can be tuned bymechanical strain. The theoretical model developed for this work confirms ourexperimental observations.KEYWORDS: spectroscopy, in-device, metal/organic interface, organic semiconductors, flexible, energy barriers, energy levels

■ INTRODUCTION

Flexible wearable devices based on organic semiconductors arethe emerging products in the current opto-electronicsmarket.1−6 Large area, extreme thinness, and compliance tocurved surfaces are the key requirements targeted for bothfunctional passive and active electronic devices.3,7−9 Con-sequently, it is paramount to understand the relation ofelectronic properties and performance of organic-based deviceswith mechanical strain. For instance, energy barriers build upbetween the metal Fermi energy, EF, and the molecular levelsdevoted to charge transport. These barriers limit chargeinjection into the organic layers and they have a deep impacton the performances of the devices.10−13 Therefore, propercharacterization and determination of their energetics isnecessary for the design and optimization of metal−moleculeinterfaces. Recently, Wu et al. reported that mechanical strainmodifies the work function of organic semiconductors.14 Insuch work, the authors induced mechanical strain by playingwith the thermal expansion mismatch between a poly-(dimethylsiloxane) substrate and a rubrene crystal, whilemonitoring the work function changes by scanning Kelvinprobe microscopy.14 Despite the deep implications of theseresults in the understanding of the connection betweenstructural and electronic disorder in soft organic materials,the impact that the strain induces on the energy-levelalignment at the metal/molecular interfaces is still under

debate. A device approach could be very beneficial foradvancing the field of organic electronics from a state offundamental curiosity to technological application.In this communication, we apply in-device hot-electron

spectroscopy to experimentally measure the interfacial energybarrier at metal-electron transporting small molecules. Themeasurements are performed in operative conditions while thesamples are bent by a controlled applied mechanical strain. In-device measurements provide further understanding of theimpact of mechanical strain at the metal/organic semi-conductor interfaces and permit us to test two main pointsfor the organic electronics community. First, we probe theenergy barriers at the metal/organic semiconductor interfacesand show that they do not vary with the strain induced bybending. Second, we observe that even if the interface energybarriers at the metal/semiconductor interfaces are strainindependent, we can tune the charge hopping rate throughthe semiconductor by bending the device, which results in anincrease of the measured collector current. As proof ofprinciple, we focus on N,N′-dioctyl-3,4,9,10-perylenedicarbox-imide (PTCDI-C8), which is a planar molecule, and wesupport it with three-terminal devices based on C60, which is a

Received: December 9, 2018Accepted: March 12, 2019Published: March 12, 2019

Research Article

www.acsami.orgCite This: ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

© 2019 American Chemical Society 12717 DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

Dow

nloa

ded

via

UN

IV O

F IO

WA

on

Apr

il 17

, 201

9 at

19:

56:3

3 (U

TC

).

See

http

s://p

ubs.

acs.

org/

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s.

spherical molecule. The data is complemented with thedevelopment of a theoretical model, which reinforces ourexperimental observations. Our results provide a step further inthe understanding of the effect of mechanical strain in deviceswith direct impact on the organic electronics community andthe engineering of novel flexible electronics.

■ RESULTS AND DISCUSSION

In-device hot-electron spectroscopy is based on a three-terminal vertical solid-state device.15−19 This technique is thesolid state variant of ballistic electron emission spectrosco-py.20−26 The working principle is shown in Figure 1. In moredetail, our three-terminal device is composed of an emitter, abase, and a collector. The device structure has been grown on aflexible Kapton tape in ultra-high vacuum (UHV) and 0.5 nmof Co2O3 has served as an adhesion layer (see ExperimentalSection for further device fabrication details).26,27 A 20 nm-thick aluminum contact is used as the emitter. This electrode isthen plasma-oxidized in situ to create an Al2O3 tunnel barrier.The base electrode consists of 15 nm evaporated gold. Goldwas chosen for being a commonly used material for device

contacts. Its air stability and noble properties make it a suitablemetal for prepatterned devices. A 100 nm-thick organicsemiconductor serves as the collector. We chose the well-established n-type semiconductors, PTCDI-C8 and C60 asproof of principle. See their chemical structures in Figure 1d,e,respectively.A current, IE, is injected from the emitter to the device when

a negative bias, VEB, is applied. The electrons after tunnelingthrough the Al2O3 barrier are “hot” in the base because theirenergy is above the Fermi energy of the metal. A fraction ofthese hot electrons crosses the base ballistically without anysignificant energy attenuation. If VEB is lower than the Au/molecule interface barrier, Δ (Figure 1a), the ballistic electroncurrent is reflected at the interface and will flow into the baseterminal (IB). No current is measured at the collector (IC = 0).In contrast, if VEB is higher than the barrier, Δ (Figure 1b),some of the hot electrons that arrive at the Au/moleculeinterface enter in the lowest unoccupied molecular orbital(LUMO) level of the n-type semiconductor and diffuse towardthe top Al electrode. A current is measured in the collector (IC≠ 0). Hot electron injection is only possible with negative VEB

Figure 1. Device scheme. Electrons can tunnel from the emitter electrode (E) to the base metal (B) when a bias between these two electrodes, VEB,is applied. (a) If VEB is lower than the metal/organic semiconductor interface barrier, Δ, ballistic electrons cannot enter into the molecular orbitalsdevoted to charge transport, in this case, the lowest unoccupied molecular orbital (LUMO), and they will be reflected back to the base. Nocollector (C) current, IC will be measured. (b) If the applied VEB is higher than Δ, some of the hot electrons flow into the LUMO of the organicsemiconductor and they will diffuse to the top metal contact. Nonzero IC will be measured. The collector and the base are kept grounded. (c)Photograph of a Kapton substrate with six devices bent on a sample holder with a bending radius, R. (d) Chemical structure of N,N′-dioctyl-3,4,9,10-perylenedicarboximide (PTCDI-C8). (e) Chemical structure of C60 fullerene.

Figure 2. (a) Emitter current, IE measured at the Al/Al2O3/Au junction of the device by sweeping the emitter-base bias, VEB. The junction has beenmeasured under different strain conditions by bending the sample onto semicylindrical holders with radius R = ∞, 40, 25, 15, 12.5, and 5 mm. (b)Average strain, Savg, dependence of IE at VEB = 1 V.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

12718

because in this device configuration, the base electrode is keptat ground potential (Figure 1a,b). Importantly, the current, IC,is measured without any external applied bias between the baseand collector, VCB, and thus, IC can be considered as a purelydiffusive current. This is possible due to both the momentumof the injected electrons perpendicular to the Au/semi-conductor interface and to the built-in potential created bysandwiching the molecular material with two metallic contactswith different work functions.15−19 In this device configuration,the energy-level alignment between the Fermi level of theemitter and the base is controlled with the VEB (see Figure 1),whereas the energy alignment at the base/collector interface isnaturally given by the metal/molecule interface energy barrier,Δ. The experiment has been repeated with the sample flat andbent for five different bending radii, R = ∞, 40, 25, 15, 12.5,and 5 mm. The cylindrical sample holders shown in Figure 1cpermit performing the experiments in situ while the samplesare bent, and thus, while mechanical strain is applied. Theinduced mechanical strain on the samples increases with thesmaller bending radii (see Supporting Note 1 and Table S1).Figure 2a,b shows the characteristic IE−VEB curve of the

tunnel junctions of a three-terminal hot-electron moleculartransistor fabricated on a Kapton substrate at 300 K26,28 andbent for several radii. No significant change is observed in theIE between flat and bent conditions (see Figure 2a). Thecurrent slightly decreases with the bending radius getting amaximum reduction in IE of 33% between the no-strain (flat)and maximum-strain (R = 5 mm) cases. These differences arebetter seen in Figure 2b, where the IE current at VEB = −1 V isplotted for the strain induced by bending the sample. Aplausible explanation for such an observation is that when thinfilms are bent, the separation between their metallic grains isenlarged compared to the case when the samples are flat, andthus, the charge transport is hindered. This deformation

appears to be inelastic because the dozens of devices studied indifferent chips do not recover their initial performance afterbeing bent from R = 40 mm to R = 5 mm.Figure 3a shows the IC−VEB curve of PTCDI-C8-based

three-terminal hot-electron devices measured when the samplewas flat without any applied mechanical strain and under strainby varying the bending radius of the sample holder from 5 to40 mm. Figure 3c represents IC normalized with the maximumcollector current IC max, which shows more clearly that thecurves visibly overlap and do not show any difference in theirshape and their onset, which indicates that the energy barrier atthe metal/organic semiconductor interface does not vary withstrain. This result is strengthened by the results in Figure 3d,where IC/IC max measurements of C60-based device are plottedfor different bending radii. Interpolating the fit of the lineargrowth of IC to IC = 0, the interface energy barrier of Au/PTCDI-C8 is determined to be Δ = 1.1 ± 0.1 V and Δ = 0.95± 0.1 V in the case of the Au/C60 interface. In both cases, thedevice-to-device variation in each chip is lower than themeasurement precision, whereas the maximum variation fromchip to chip is ±0.1 V. This straightforward method for theextraction of the interface energy barrier at metal/organicsemiconductor interfaces has been proven before for othersemiconductors.15,17,18 The three-terminal measurementsperformed on the Au/PTCDI-C8- and Au/C60-based hot-electron transistors show that bending of the devices and theconsequent strain does not affect the transport energy barrieralignment, Δ, at the metal/organic semiconductor interfaces.These results differ from the conclusions of Wu et al.,14 wherechanges in the work function of a rubrene crystal withmechanical strain were reported. In this last case, the bendingstrain introduced disorder into a perfectly oriented molecularlayer. The disorder creates new molecular sites with differentoxidation and reduction energies compared to the crystal,

Figure 3. Spectroscopy measurements at 300 K for bending radii R = ∞, 40, 25, 15, 12.5, and 5 mm. (a) Hot electron current, IC, of the PTCDI-C8-based device vs emitter-base bias, VEB, for different bending radii. (b) Dependence of IC (at VEB = 1.5 V) with the average strain, Savg, (c)Normalized IC/IC max−VEB curves of PTCDI-C8-based device for different bending radii. (d) Normalized IC/IC max−VEB curves of C60-based hot-electron transistor for different bending radii.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

12719

which changes the interface density of states and consequentlythe interface energy-level alignment. In our work, polycrystal-line organic thin films with existent disorder at the metal/organic interface are studied. Therefore, strain does notsignificantly augment the disorder at the interface, and theenergy-level alignment at the metal/organic semiconductorinterface is unaffected.Figure 3a shows that contrary to the trend of IE, IC increases

with the applied strain and reaches a value 430% higher whenthe device is bent with R = 5 mm, compared to the case whenit is flat. This evolution is better visible in Figure 3b, where theIC at VEB = −1.5 V is plotted with the mechanical strain appliedon the device. As the current injected from the tunnel junction,IE (see Figure 2b) is almost constant, the changes in IC shouldbe related to the organic semiconductor material.In order to disentangle the physics behind our results, we

have developed a theoretical model that considers theresistivity of the organic semiconductor with bending of thedevice (see Figure 4a). For conjugated organics, the sp2

hybridized orbitals are split in energy from the unhybridizedp-orbital. This energy splitting results in a highest occupiedmolecular orbital and LUMO with p-type symmetry. In planarmolecules such as PTCDI-C8, the cores are composed ofconjugated ring-like segments with well-defined spatialorientation along the axis of the unhybridized orbitals. Asthe primary transport orbitals in these materials lack polarsymmetry, the angular dependence of the wave functionoverlap between neighboring sites is a key component intransport calculations.In π-conjugated materials with well-defined site orientations

and Miller−Abrahams-type network, the inter-site resistancecan be considered as

R rR

r a( , )cos

exp(2 / )ijij

ijij

0

2θθ

=(1)

where Rij0 describes the minimum resistance limited by material

properties of the system, rij is the intersite spacing, θij is therelative angle between the transport orbitals at sites i and j, anda is the carrier wave function localization radius.When the organic semiconductor is flexed, parameters rij and

θij change, and the molecular unit cell is strained with thevolume, in general, not conserved. This results in amodification to the volumetric density of states, which entersthe transport equations exponentially through the intersitespacing, rij. For a change in length, dx, along the applied strainaxis, the material responses along the perpendicular axes (dy)are governed by the Poisson’s ratio, dy = −ν dx.

The current through the organic layer exhibits a hoppingconduction along a path of sites with each hop occurring at arate that depends on the relative position and orientation ofthe initial and final site. These rates are analogous to resistorsin series with a value corresponding to eq 1. As the bulkresistance is determined by summing the individual resistancesalong the path, it is clear from eq 1 that hops between siteswith relative orientations near perpendicular will dominate thetotal value. These bottlenecks that exist in the relaxed organicwill be broken as the device is bent and the sites shift theirrelative position and orientation. As the strain is dependent onthe distance from the neutral axis, the average value isconsidered, Savg = wmax/2R, where we have assumed the neutralaxis to be midway between the top and bottom contacts, andtherefore wmax = 100 nm.For two neighboring sites, i and j, in a bottleneck

configuration, the orientation of site i and site j with respectto the global basis are related through θj = θi + π/2 − ξ, whereξ is a small deviation from completely perpendicularorientation. The intersite resistance for a bottleneck in arelaxed organic can then be written as

( )( ) ( )

R R

exp

sin sin cos cos

ij

r

a

i i i i

0

2

2 2

2

ij

θ θ ξ θ θ ξ

=

+ − + + −π π

⊥∞

ÄÇÅÅÅÅÅÅ

ÉÖÑÑÑÑÑÑ(2)

Therefore, the resistivity of the flat organic is

( )( ) ( )

exp

sin sin cos cos

r

a

i i i i

2

2 2

2

ij

ρθ θ ξ θ θ ξ

∝+ − + + −π π⊥

∞ ÄÇÅÅÅÅÅÅ

ÉÖÑÑÑÑÑÑ(3)

and the resistivity of the deformed organic is

S

S

exp (1 (1 2 ) )

sin ( 1)cos( )sin(2 )R

r

a

i

2avg

1/3

2avg

ij

ρν

ξ ν ξ θ ξ∝

+ −

[ − + − ]⊥

ÄÇÅÅÅÅÅÅÅ

ÉÖÑÑÑÑÑÑÑ

(4)

In comparing with experimental values, the fractional changein resistivity

Rρρ

ρ ρρ

Δ =| − |

⊥∞

⊥∞

⊥

⊥∞

(5)

is considered as a function of the radius of curvature. The valueof the intersite spacing is approximated from the cubed root ofthe volume of the measured triclinic unit cell for PTCDI-C8 as

Figure 4. (a) Diagram of a flexed organic with radius of curvature, R (top) and an unbent representation with linearly varying force (bottom). Theglobal basis (unprimed) and local basis (primed) are shown with the angle, θ, defining the degree of rotation in the x−y plane. x is the axis ofapplied strain, and wmax is the distance between the neutral axis and the contact surface (100 nm) (b) Fractional change in resistivity as a functionof the radius of curvature for experimental measurements (circle) and theoretical model (solid line). For specific values see Table S2.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

12720

9.19 Å,29 and the carrier localization radius is approximated as1 Å. The value for the Poisson’s ratio is estimated frommeasured values of similar materials to be 0.4, and the angulardeviation, ξ, is left as a fitting parameter. For bottlenecks in therelaxed organic with relative angle θ⊥ = π/2 − ξ, the generalexperimental trend can be reproduced for ξ = 2.5 × 10−6 rad asseen in Figure 4b, where the experimental and theoretical Δρ/ρ⊥∞ are plotted with respect to the R. See Supporting Note 2 for

further details. These results are valid even for cases involvinganisotropic transport because in the bottleneck configuration,rotational dependence of the sites dominates the resistivity.Thus, even for systems with anisotropic spacings between sitesand anisotropic site wave function decay lengths, we obtainnearly identical numerical results. See Supporting Note 3 andFigures S1, S2 for details regarding these calculations. Even ifthis observation is new for organic semiconductors, changes inthe electronic structure of organic superconductors and chargetransfer salts with strain have been previously reported.30−32

■ CONCLUSIONSIn conclusion, we have successfully built three-terminal hot-electron molecular transistors on flexible Kapton substratesbased on Au/PTCDI-C8 and Au/C60 and measured them inflat (no strain) and flexed (under strain) positions.Subsequently, we have demonstrated that the energy barrierbetween the EF of the metallic contact and the LUMO level, Δ,is strain independent, that is Δ = 1.1 ± 0.1 eV for Au/PTCDI-C8 and Δ = 0.95 ± 0.1 eV for Au/C60 interfaces. Besides, wehave demonstrated that the bending of the planar organicsemiconductor reorients the molecules and breaks thebottlenecks, which is reflected experimentally as an increasein the amplitude of the measured IC. These results give furtherunderstating of the interfacial energy parameters for theengineering of flexible organic electronic devices as well aspresent a simple method for manipulating the detected currentin organic-based devices.

■ EXPERIMENTAL SECTIONDevice Fabrication. All devices described in this work were

fabricated in a UHV evaporator chamber (base pressure < 10−9 mbar)with a shadow mask system. Au (99.95%) (Lesker) was evaporated bye-beam at a rate of 0.1 nm s−1, and Al (99.95%) (Lesker) wasthermally evaporated at 0.06 nm s−1. The organic layer of C60 (Sigma-Aldrich, sublimed 99.9%) and PTCDI-C8 (Sigma-Aldrich, 98%purity) were thermally evaporated at 0.01 nm s−1.Electrical Characterization. Electrical characterization was

performed under high vacuum (base pressure 5 × 10−5 mbar) in avariable-temperature probe-station (Lakeshore). A Keithley 4200semiconductor analyzer system was used to record I−V curves.Modeling. A random resistor network was considered with

intersite resistances that depend on site orientation. Changes incritical segments of the conduction network (bottlenecks) weremodeled through shear rotations induced by controlled deformationof the bulk system, and the fractional change in bottleneck resistivitywas used to represent changes in the bulk electronic properties.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.8b21531.

Atomic force microscopy images, fractional change inresistivity as a function of the radius of curvature foranisotropic systems, bending radius and average strain

relation, and details about strain calculation and theorydevelopment (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: l.hueso@nanogune.eu.ORCIDFelix Casanova: 0000-0003-0316-2163Michael E. Flatte: 0000-0001-5093-1549Luis E. Hueso: 0000-0002-7918-8047Present Addresses⊥A.A.: Simbeyond B. V., 5612 AE Eindhoven, The Nether-lands∥IMEC and K. U. Leuven, 3001 Leuven, Belgium.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work is supported by the European Research Council(grants 257654-SPINTROS) and by the Spanish MINECOunder project no. MAT2015-65159-R and Maria de MaeztuUnits of Excellence ProgramMDM-2016-0618. A.A. ac-knowledges the Basque Government for a PhD fellowship(PRE_2017_2_0052). S.R.M. and M.E.F. acknowledgesupport from DOE BES through grant no. DE-SC0014336and the Stanley-UI Foundation Support Organization.

■ REFERENCES(1) Forrest, S. R. The path to ubiquitous and low-cost organicelectronic appliances on plastic. Nature 2004, 428, 911−918.(2) Wang, Y.; Zhu, C.; Pfattner, R.; Yan, H.; Jin, L.; Chen, S.;Molina-Lopez, F.; Lissel, F.; Liu, J.; Rabiah, N. I.; et al. A HighlyStretchable, Transparent, and Conductive Polymer. Sci. Adv. 2017, 3,e1602076−1602086.(3) Karnaushenko, D.; Makarov, D.; Stober, M.; Karnaushenko, D.D.; Baunack, S.; Schmidt, O. G. High-Performance MagneticSensorics for Printable and Flexible Electronics. Adv. Mater. 2015,27, 880−885.(4) Rogers, J. A.; Someya, T.; Huang, Y. Materials and Mechanics forStretchable Electronics. Science 2010, 327, 1603−1607.(5) Xu, J.; Wang, S.; Wang, G.-J. N.; Zhu, C.; Luo, S.; Jin, L.; Gu, X.;Chen, S.; Feig, V. R.; To, J. W. F.; et al. Highly Stretchable PolymerSemiconductor Films Through the Nanoconfinement Effect. Science2017, 355, 59−64.(6) Kaltenbrunner, M.; Sekitani, T.; Reeder, J.; Yokota, T.; Kuribara,K.; Tokuhara, T.; Drack, M.; Schwodiauer, R.; Graz, I.; Bauer-Gogonea, S.; et al. An Ultra-Lightweight Design for ImperceptiblePlastic Electronics. Nature 2013, 499, 458−463.(7) Sekitani, T.; Zschieschang, U.; Klauk, H.; Someya, T. FlexibleOrganic Transistors and Circuits with Extreme Bending Stability. Nat.Mater. 2010, 9, 1015−1022.(8) Hwang, S.-W.; Tao, H.; Kim, D.-H.; Cheng, H.; Song, J.-K.; Rill,E.; Brenckle, M. A.; Panilaitis, B.; Won, S. M.; Kim, Y.-S.; et al. APhysically Transient Form of Silicon Electronics. Science 2012, 337,1640−1644.(9) Salvatore, G. A.; Mu, N.; Kinkeldei, T.; Petti, L.; Zysset, C.; Tro,G.; Strebel, I.; Bu, L. Wafer-Scale Design of Lightweight andTransparent Electronics That Wraps around Hairs. Nat. Commun.2014, 5, 2982−2990.(10) Friend, R. H.; Gymer, R. W.; Holmes, A. B.; Burroughes, J. H.;Marks, R. N.; Taliani, C.; Bradley, D. D. C.; Santos, D. A. D.; Bredas,J. L.; Logdlund, M.; et al. Electroluminescence in conjugatedpolymers. Nature 1999, 397, 121−128.(11) Sirringhaus, H. Device Physics of Solution-Processed OrganicField-Effect Transistors. Adv. Mater. 2005, 17, 2411−2425.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

12721

(12) Hwang, J.; Wan, A.; Kahn, A. Energetics of metal-organicinterfaces: New experiments and assessment of the field. Mater. Sci.Eng. R Rep. 2009, 64, 1−31.(13) Braun, S.; Salaneck, W. R.; Fahlman, M. Energy-LevelAlignment at Organic/Metal and Organic/Organic Interfaces. Adv.Mater. 2009, 21, 1450−1472.(14) Wu, Y.; Chew, A. R.; Rojas, G. A.; Sini, G.; Haugstad, G.;Belianinov, A.; Kalinin, S. V.; Li, H.; Risko, C.; Bredas, J. L.; et al.Strain Effects on the Work Function of an Organic Semiconductor.Nat. Commun. 2016, 7, 10270−10278.(15) Atxabal, A.; Braun, S.; Arnold, T.; Sun, X.; Parui, S.; Liu, X.;Gozalvez, C.; Llopis, R.; Mateo-alonso, A.; Casanova, F.; et al. EnergyLevel Alignment at Metal/Solution-Processed Organic Semiconduc-tor Interfaces. Adv. Mater. 2017, 29, 1606901−1606906.(16) Arnold, T.; Atxabal, A.; Parui, S.; Hueso, L. E.; Ortmann, F.Hot Electrons and Hot Spins at Metal-Organic Interfaces. Adv. Funct.Mater. 2018, 28, 1706105−1706115.(17) Gobbi, M.; Pietrobon, L.; Atxabal, A.; Bedoya-Pinto, A.; Sun,X.; Golmar, F.; Llopis, R.; Casanova, F.; Hueso, L. E. Determinationof Energy Level Alignment at Metal/Molecule Interfaces by In-DeviceElectrical Spectroscopy. Nat. Commun. 2014, 5, 4161−4168.(18) Parui, S.; Atxabal, A.; Ribeiro, M.; Bedoya-Pinto, A.; Sun, X.;Llopis, R.; Casanova, F.; Hueso, L. E. Reliable determination of theCu/n-Si Schottky barrier height by using in-device hot-electronspectroscopy. Appl. Phys. Lett. 2015, 107, 183502−183505.(19) Jiang, J. S.; Pearson, J. E.; Bader, S. D. Direct Determination ofEnergy Level Alignment and Charge Transport at MetalAlq3Interfaces via Ballistic-Electron-Emission Spectroscopy. Phys. Rev.Lett. 2011, 106, 156807−156811.(20) Kaiser, W. J.; Bell, L. D. Direct Investigation of SubsurfaceInterface Electronic Structure by Ballistic-Electron-Emission Micros-copy. Phys. Rev. Lett. 1988, 60, 1406−1409.(21) Bell, L. D.; Kaiser, W. J. Observation of Interface BandStructure by Ballistic-Electron-Emission Microscopy. Phys. Rev. Lett.1988, 61, 2368−2371.(22) Ludeke, R.; Prietsch, M.; Samsavar, A. Ballistic ElectronEmission Spectroscopy of Metals on GaP(110). J. Vac. Sci. Technol., A1991, 9, 885−890.(23) Bell, L. D. Ballistic Electron Emission Microscopy andSpectroscopy: Recent Results and Related Techniques. J. Vac. Sci.Technol., B: Microelectron. Nanometer Struct.–Process., Meas., Phenom.2016, 34, 040801−040828.(24) Yi, W.; Narayanamurti, V.; Lu, H.; Scarpulla, M. A.; Gossard, A.C.; Huang, Y.; Ryou, J.-H.; Dupuis, R. D. Bandgap and Band OffsetsDetermination of Semiconductor Heterostructures Using Three-Terminal Ballistic Carrier Spectroscopy. Appl. Phys. Lett. 2009, 95,112102−112105.(25) Yi, W.; Narayanamurti, V.; Lu, H.; Scarpulla, M. A.; Gossard, A.C. Probing Semiconductor Band Structures and HeterojunctionInterface Properties with Ballistic Carrier Emission: GaAs/Alx Ga 1-xAs as a Model System. Phys. Rev. B: Condens. Matter Mater. Phys.2010, 81, 235325−235338.(26) Sun, X.; Bedoya-Pinto, A.; Llopis, R.; Casanova, F.; Hueso, L.E. Flexible Semi-Transparent Organic Spin Valve Based on Bath-ocuproine. Appl. Phys. Lett. 2014, 105, 083302−083306.(27) Bedoya-Pinto, A.; Donolato, M.; Gobbi, M.; Hueso, L. E.;Vavassori, P. Flexible Spintronic Devices on Kapton. Appl. Phys. Lett.2014, 104, 062412−062417.(28) Sun, X.; Gobbi, M.; Bedoya-pinto, A.; Txoperena, O.; Golmar,F.; Llopis, R.; Chuvilin, A.; Casanova, F.; Hueso, L. E. RoomTemperature Air-Stable Spin Transport in Bathocuproine-Based SpinValves. Nat. Commun. 2013, 4, 2794−2797.(29) Briseno, A. L.; Mannsfeld, S. C. B.; Reese, C.; Hancock, J. M.;Xiong, Y.; Jenekhe, S. A.; Bao, Z.; Xia, Y. Perylenediimide Nanowiresand Their Use in Fabricating Field-Effect Transistors andComplementary Inverters. Nano Lett. 2007, 7, 2847−2853.(30) Kawasugi, Y.; Yamamoto, H. M.; Hosoda, M.; Tajima, N.;Fukunaga, T.; Tsukagoshi, K.; Kato, R. Strain-Induced Super-conductor/Insulator Transition and Field Effect in a Thin Single

Crystal of Molecular Conductor. Appl. Phys. Lett. 2008, 92, 243508−243511.(31) Maesato, M.; Kaga, Y.; Kondo, R.; Kagoshima, S. Uniaxialstrain method for soft crystals: Application to the control of theelectronic properties of organic conductors. Rev. Sci. Instrum. 2000,71, 176−181.(32) Kondo, R.; Kagoshima, S.; Maesato, M. Crystal Structure andElectronic Band Structure of the Organic Superconductor α−(BEDT−TTF)2NH4Hg(SCN)4 under Uniaxial Strain. Phys. Rev. B:Condens. Matter Mater. Phys. 2003, 67, 134519−134525.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.8b21531ACS Appl. Mater. Interfaces 2019, 11, 12717−12722

12722