Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Photo-Carrier Multi-Dynamical Imaging at the Nanometer Scale in Organic and Inorganic Solar Cells Pablo Arturo Fernández Garrillo, Lukasz Borowik, Florent Caffy, Renaud Demadrille, and Benjamin Grevin ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b11423 • Publication Date (Web): 20 Oct 2016 Downloaded from http://pubs.acs.org on October 23, 2016
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 31
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
Photo-Carrier Multi-Dynamical Imaging at the Nanometer Scale in Organic and Inorganic Solar Cells Pablo A. Fernández Garrilloabc, Łukasz Borowik*ab, Florent Caffyc, Renaud Demadrillec, and Benjamin Grévin*c a
Univ. Grenoble Alpes, F-38000 Grenoble, France.
b
CEA, LETI, MINATEC Campus, F-38054 Grenoble, France.
c
CEA, CNRS, Univ. Grenoble Alpes INAC-SPrAM, F-38000 Grenoble, France.
KEYWORDS: Kelvin probe force microscopy, time-resolved nanoimaging, carrier recombination, carrier dynamics, nanostructured photovoltaics
ABSTRACT: Investigating the photo-carrier dynamics in nanostructured and heterogeneous energy materials is of crucial importance from both fundamental and technological points of view. Here, we demonstrate how non-contact atomic force microscopy combined with Kelvin Probe Force Microscopy under frequency modulated illumination can be used to simultaneously image the surface photo-potential dynamics at different timescales with a sub-10 nm lateral resolution. The basic principle of the method consists in the acquisition of spectroscopic curves of the surface potential as a function of the illumination frequency modulation on a two
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 31
dimensional grid. We show how this frequency-spectroscopy can be used to probe simultaneously the charging rate and several decay processes involving short-lived and longlived carriers. With this approach, dynamical images of the trap-filling, trap-delayed recombination and non-geminate recombination processes have been acquired in nano-phase segregated organic donor-acceptor bulk heterojunction thin films. Furthermore, the spatial variation of the minority carrier lifetime has been imaged in polycrystalline silicon thin films. These results establish two-dimensional multi-dynamical photovoltage imaging as a universal tool for local investigations of the photo-carrier dynamics in photoactive materials and devices. INTRODUCTION Unveiling the photo-transport mechanisms and identifying the sources of losses by carrier recombination is a key for the development of emerging photovoltaic (PV) technologies based on micro and nanostructured materials.1 In solution-processed polymer2 or small-molecule3 bulk heterojunction (BHJ) solar cells,4 excitons are dissociated into free charges at the interfaces between electron donor (D) and acceptor (A) materials, which are processed to form interpenetrated networks phase-segregated at the 10 nm scale. Despite a tremendous amount of studies, the complex interplay between the D-A phases composition, interfaces morphology, photo-carrier dynamics and transport remains intensively investigated.2,4-7 It is also crucial to address how the photo-carrier recombination can be impacted by grain boundaries, chemical impurities and other local defects in polycrystalline or nanostructured films of silicon,8,9 CdTe,10 CuInxGa(1-x)Se2 (CIGS),11 Cu2ZnSnS4 (CZTS)12 and hybrid organic-inorganic perovskites.13,14 Kelvin Probe Force Microscopy (KPFM) is a well-established technique in the field of organic, hybrid and inorganic photovoltaics. Several reports have already demonstrated that the local
ACS Paragon Plus Environment
2
Page 3 of 31
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
surface photovoltage (SPV) of photoactive thin films and devices can be mapped by analyzing the contact potential difference (CPD) shift upon illumination.15-22 Actually, it was shown that the SPV yields a measurement of the open circuit voltage (VOC) at the nanometer scale,20 i.e. the electron and hole quasi-Fermi levels spitting across a p-n junction or an organic donor-acceptor interface.18 Besides, the VOC magnitude is closely related to the carrier recombination mechanisms,23,24 which themselves depend of the nature of local defect and structural heterogeneities in nanostructured PV devices. Thus, the local SPV contrasts detected in KPFM experiments shall (at least in some cases) reflect spatial variations of the photo-carrier recombination rate in relation with the nanostructure. In the case of organic bulk heterojunctions, the carrier losses can occur through geminate recombination, non-geminate recombination of free carriers generated from different precursor states, or trap-delayed recombination involving carriers trapped in shallow or deep energy levels which can recombine after de-trapping with a free counter charge carrier.2,25-29 In principle, both non-geminate and trap-assisted processes can simultaneously contribute to the apparition of SPV contrasts detected by KPFM in the steady state. However, the photo-carrier lifetime cannot be directly probed with conventional SPV imaging under continuous wave illumination. Doubtlessly, there is a crucial need to develop time-resolved scanning probe microscopy techniques, capable to assess how the nanostructure impacts the recombination dynamics and to probe both short-lived and long-lived photo-carrier populations in PV materials. In that frame, Time-resolved Electrostatic Force Microscopy (trEFM) was successfully applied to map the photo-charging rate (i.e. the charge build up dynamics) of organic photovoltaic blends,30,31 with a lateral resolution of a few tens of nanometers. Kelvin Probe Force Microscopy (KPFM) under modulated illumination has also been used to investigate the surface photovoltage
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 31
(SPV) decays,32-34 thus providing access to the recombination dynamics in organic blends,33 or to the minority carrier lifetime in inorganic semi-conductors.32,34 However, so far, all KPFM published works under frequency-modulated illumination demonstrated only “point” measurements (i.e. the lifetimes were only extracted at single points of the sample surface); twodimensional (2D) imaging of the recombination dynamics in nanostructured PV materials remains a real challenge. Here, we demonstrate how non-contact atomic force microscopy combined with KPFM under frequency modulated illumination can be used to simultaneously image the photo-potential builtup and decay dynamics at different timescales with a sub-10 nm lateral resolution. The basic principle of 2D dynamical imaging consists in the acquisition on a 2D grid of spectroscopic curves of the surface potential (SP) as a function of the illumination frequency modulation. Beyond the achievement of 2D dynamical imaging, we show that by extending the frequency range of the illumination modulation it becomes possible to probe simultaneously the charging rate and several decay processes involving short-lived and long-lived carriers. Dynamical images attributed herein to trap-filling, trap-delayed recombination and non-geminate recombination processes are acquired in nano-phase segregated organic donor-acceptor bulk heterojunction thin films. Furthermore, the spatial variation of the minority carrier lifetime is imaged in polycrystalline silicon thin films. These results establish illumination-modulated Kelvin Probe Force Microscopy as a universal tool for local investigations of the photo-carrier dynamics in photoactive materials and devices.
ACS Paragon Plus Environment
4
Page 5 of 31
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
RESULTS AND DISCUSSION In this work, we investigated BHJ based on a fullerene derivative (PC71BM) as the electron acceptor and a home-made synthesized wide bandgap electron donor polymer (Scheme S1 in the Supporting Information), PDBS-TQx.35 The morphology of PDBS-TQx/PC71BM blends can be easily tuned35 from large-scale to nanoscale phase separation (yielding power conversion efficiencies over 5%), by using additives36-38 and varying the solvent used for processing. Therefore, these samples can advantageously be used to quantify the impact of the morphology on the carrier dynamics, as well as to establish the ability to resolve local dynamical contrasts at the sub-10 nm scale. Figure 1 shows a schematic illustration describing the implementation of dynamical surface photo-voltage imaging with KPFM. Its operating principle consists in the acquisition on a predefined two-dimensional (2D) grid of spectroscopic curves of the surface potential (SP) as a function of the frequency modulation (f) of the illumination source. Here, we underline that the sample illumination is performed in the large perturbation regime; this configuration is similar to the one of conventional macroscopic large perturbation transient photovoltage (LPTV) experiments.39 KPFM measurements are performed in single-pass mode in combination with beam-deflection non-contact AFM (nc-AFM) under ultra-high vacuum,19 and the 2D frequency– spectroscopy is operated in close loop mode (i.e. without interruption of the z-regulation). The KPFM compensation bias regulation (with integration times of several tens of ms) yields a temporally averaged surface potential value for each frequency of the illumination pulse train. In our setup, logic signals are generated by the AFM controller to trigger the generation of illumination pulse groups at pre-selected frequencies, and to insure a synchronous spectroscopic data acquisition. The modulation duty ratio (D) and the pulse group duration are kept constant
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 31
over the whole frequency range. Before signal integration by the KPFM loop, a fixed time-delay is applied after each frequency step to prevent artefacts in the potential regulation. Finally, a post-data acquisition processing is performed to fit the frequency-spectroscopy curves (i.e. curves of the average surface potential as a function of the illumination modulation frequency) and recalculate the dynamical images. Here, we stress that the average surface photo-voltage (SPVav) versus the modulation frequency (Figure 1b,c) depends not only on the kinetics of the recombination mechanisms, but also on the effective photo-charging rate of the sample. Assuming an instantaneous photovoltage built-up (i.e. τb= 0) under illumination, the average SPV shall gradually increase when raising the modulation frequency, and finally saturate because the illumination modulation becomes much faster than the photo-potential time decay constant τd. However, in the high frequency range, SPVav may finally decrease if the illumination pulse duration becomes comparable or smaller than the photo-charging build-up time τb (see Figure 1b,c). Thus, the decay and built-up time constants can be simultaneously probed by analyzing the behavior of the spectroscopic curves in the full spectral range (i.e. from low to high frequencies) as shown in Figure 1c. In a first step, we will use simple exponential functions characterized by two effective timeconstants τb and τd to describe the SPV built-up and decay dynamics, respectively (the approximation of not using stretched-exponential decays to fit the data acquired in the large perturbation regime is justified in the Supporting Information). Then, the average surface potential Vav which is measured by KPFM as a function of the frequency f can be fitted (a detailed calculation is given in the Supporting Information) by:
ACS Paragon Plus Environment
6
Page 7 of 31
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
D (1− D ) D (1− D ) − − − − fτb fτd fτb ( ) Vav = VD + SPV max D 1 − e e 1 − e fτd + SPV max fτd − fτb 1 − e
(1)
Here, VD and SPVmax correspond respectively to the in-dark surface potential, and to the maximum photo-potential which can be measured under continuous illumination. The average photo-potential SPVav is equal to Vav-VD. First, we evaluated the ability to perform multi-dynamical SPV imaging on PDBSTQx/PC71BM blends35 processed to display an heterogeneous morphology at the mesoscopic scale as shown by Figure 2a. For these samples, the nc-AFM topographic images (Figure 2b) reveal the existence of domains with lateral dimensions on the order of 200 nm. Similar results have been reported by others for many bulk heterojunctions, such as PTB7:PCBM blends.40 These works demonstrated that the mesoscopic clusters consist in PCBM aggregates surrounded by a polymer-rich D-A mixed phase (Figure 2a), while a finely mixed phase of the donor and acceptor components resides between the large domains and forms a skin layer on top of the film.40 The charge generation and photo-transport are expected to be more efficient in this finely intermixed phase. A negative shift of the surface potential is observed under illumination (i.e. a negative average surface photo-voltage of ca. 295 mV), revealing the splitting of electrons and holes quasi Fermi levels. The negative SPV polarity is consistent with the donor and acceptor energy level alignment with respect to the grounded substrate (Figure S1). Figure 2c displays a frequencyspectroscopy curve of the average KPFM potential Vav, plotted as a function of the light modulation frequency. When increasing the modulation frequency from 300 Hz to ca. 20 kHz, Vav decreases towards more negative values; in other words the magnitude of the (negative)
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 31
photo-voltage SPVav increases (for the sake of simplicity, in the following the photo-voltage evolution will be discussed in terms of its absolute value). Increasing further the frequency results in a photo-voltage diminution, revealing that the pulse duration becomes too short to achieve a complete photo-potential build-up under illumination. However, surprisingly, the SPV does not vanish but increases again for frequencies higher than ca. 100 kHz. This phenomenon reveals that the measured photo-potential originates from the contributions of two mechanisms characterized by strongly different dynamics. They can be taken into account by summing two average photo-potentials (Figures S4 and S5) with a dual set of independent parameters (SPVmax1, τd1, τb1 and SPVmax2, τd2, τb2). The dynamical images calculated from the curves fit are displayed in Figure 2d,e,f. In addition, a “static” surface photo-voltage image can be calculated (Figure 2g), by summing the SPVmax contributions of both mechanisms (Figure S6). This direct SPV measurement under modulated illumination overcomes advantageously the limitations of conventional SPV imaging20 due to lateral misalignment between “in-dark” and “illuminated” images. All images reveal a clear contrast over the mesoscopic islands, with higher built-up and decay time constants and a lower photo-voltage magnitude. The first contribution to the photo-potential dynamics (Figure 2d,e) is characterized by build-up (τb1, Figure 2d) and decay (τd1, Figure 2e) time constants of a few µs and ms, respectively. The second one displays much faster decay dynamics (Fig. 2f), with τd2 in the µs range. Moreover, τb2 is also several orders of magnitude lower than τb1. Indeed, under these lightning conditions the average photo voltage does not give any signs of decrease up to the upper modulation frequency allowed by the illumination chain. In other words, this means that the pulse duration at the upper frequencies (a few MHz) remains much longer than the fast builtup time. As a consequence, the second built-up time cannot be directly measured with our setup
ACS Paragon Plus Environment
8
Page 9 of 31
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 current form. In turn, simulations (Figure S8) allow estimating that τb2 values fall below a few tens of ns. Now, the question that presents itself is whether the fast and slow dynamics originate from different physical processes, or if one measures only one phenomenon displaying different characteristics time scales. This could for instance be the case of a two trap population, the longer and faster decays being associated to the effect of deep and shallow traps, respectively. First, we note that the existence of two different dynamical regimes (characterized by fast and slow voltage decays and built ups) is also observed on the nano phase segregated PDBSTQx/PC71BM blend (Figure 3), for which the dynamical images display heterogeneities at the scale of a few tens of nanometers. Before discussing in detail the origin of these contrasts, a deeper insight in the mechanisms behind the SPV dynamics can be gained by analyzing the time constants evolution as a function of the intensity of the illumination source. Figure 4a shows a series of spectroscopic curves acquired for different lighting conditions. The shape of the curves (Figure 4a) dramatically changes in the high frequency regime when decreasing the illumination intensity. Especially, at the lowest intensity the SPV decreases for frequencies above a few hundreds of KHz. This reveals a significant increase of the fast built-up time constant τb2. The full set of fitted time constants is given in Figure 4b, which also shows that the “fast” decay time (τd2) constant is strongly reduced for increasing optical powers, while the slow” decay time (τd1) is quasi independent of the illumination intensity. These behaviors are consistent with carrier-concentration dependent non-geminate recombination processes 2,23 and trap-delayed mechanisms (with a recombination rate proportional to the trap density),2,25-27 respectively.
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 31
We also note that the slow decay time (a few hundreds of µs) is similar to the lifetime of longlived trap populations reported for other BHJ such as PBDTTPD:PCBM.25 The fast decay times values (a few µs for the maximum illumination intensity, i.e. the maximum carrier density) seem also to agree relatively well with the results of macroscopic transient experiments on other blends.39 Thus, it is reasonable to attribute the fast and slow decays to non-geminate and trapdelayed recombination processes. In view of the above, it is logical to interpret the “fast” build-up time (τb2) in terms of transit time of the free charge carriers in the D-A networks, and the “slow” built-up time (τb1) as the effective time needed to fill the traps responsible for the subsequent trap-delayed recombination mechanisms. Here, we stress that the relatively large τb1 built-up times (τb1 values of a few µs) observed in our experiments put some constraints on the physical nature of the trap states in our blends. These dynamics are compatible with the build-up times of trapped electrons deduced from transient photocurrent experiments.27,28 We also note that both the slow build-up and decay dynamics can strongly differ from one blend to the other depending of the relative strength of trapping effects;29 this fact should be borne in mind when comparing our data with the results of former investigations carried out with macroscopic probes on other BHJs. Last, we observe that both built-up times increase when decreasing the optical power below a certain threshold, which seems consistent with a carrier-concentration dependent mobility.41 At the lowest intensity the effect of the faster built-up becomes indeed apparent (Figure 4a and Figure S8) on the spectroscopic curves, in the form of an SPV decrease in the higher frequency regime. To sum it up, besides its capacity to perform 2D dynamical imaging, the advantage of our approach lies in its ability to separate the contribution of several processes (Figure S9) which
ACS Paragon Plus Environment
10
Page 11 of 31
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
occur at different timescales: i) diffusion of free charges in the D/A network (τb2) ii) nongeminate recombination (τd2) iii) carrier trapping (τb1) iv) carrier de-trapping and trap-delayed recombination (τd1). Concerning the last point, various kinds of defects can contribute to trap-delayed recombination processes in BHJ, such as chemical impurities,42 a non-optimal sample morphology43 or isolated acceptor clusters.44 Carrier trapping and de-trapping in non-percolating mesoscopic PC71BM aggregates is highly likely to be at the origin of the higher τb1 and τd1 values (Fig. 2d,e) observed over the mesoscopic domains of the blend with large scale heterogeneities. In addition, the longer non-geminate recombination (Figure 2f) may be attributed to the polymer-enriched nature of the mixed phase in the mesoscopic clusters. The impact of the sample morphology on the photo-carrier dynamics seems more ambiguous in the case of the nanophase segregated sample. Here, we note that the dissipation images (Figure 3c) reveal that significant variations in the nanophase composition45 occur at the scale of a few tens of nanometers in that sample. Unambiguously, the dynamical images display heterogeneities at the same scale. This led us to conclude that both the trap-retarded and non-geminate recombination processes are significantly impacted by morphological and/or compositional variations in this blend. However, we note that the dynamical contrasts do not display a systematic correlation (or anti-correlation depending of the dynamical channel) with the dissipation ones. This observation suggests that while the damping contrasts originates only from variations in the tip-surface interaction, the dynamical ones may be influenced by contributions from the sub-surface. The analysis of high resolution images (Figure 3h-i) demonstrates however that the dynamical SPV is not dominated
ACS Paragon Plus Environment
11
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 12 of 31
by contributions from the bulk (otherwise local contrasts would be hardly resolved). These last data demonstrate the possibility to investigate the interplay between the nanostructure and the carrier dynamics down to the sub-10 nm scale. The achieved lateral resolution (a few nanometers as shown by the cross section profiles in Figure 3j) shows the huge potential of our approach for local investigations on finely phase separated BHJ samples. At this stage, in the particular case of PDBS-TQx/PC71BM blends,35 our results strongly suggest that the performances remains limited (with a maximum PCE of ca. 5%) due to carrier-trapping effects related to a nonoptimized morphology. KPFM under frequency modulated illumination holds also great promises of applications in the field of inorganic photovoltaics. To illustrate this statement, as a first example, we mapped the spatial variations of the minority carrier lifetime in a small grain polycrystalline silicon (polySi) thin film grown on n-doped silicon (Figure 5a). Even after the surface polishing process, submicron sized domains can be easily identified in the AFM topographic images of this sample (Figure 5b). A positive surface photo-voltage is observed, which is consistent with the band alignment at the recessed Si/n-Si interface (Figure S2). Besides, the average photo-voltage displays a monotonic evolution with respect to the frequency-modulation of the illumination source (see Figure S10). Thus, the spectroscopic curves can be fitted by using a simpler equation based on a single-exponential decay time (see the Supporting Information), yielding a measure of the effective minority carrier lifetime34 with average values of a few tens of µs. The dynamical images display dark patches few tens of nanometers wide which are correlated with local contrasts in the damping images (Figure 5e,f), and where the average lifetime is significantly reduced. The images taken at higher magnification display the same features than larger-scale images (compare Figure 2c and 2e), confirming that the observed dynamical contrasts have a
ACS Paragon Plus Environment
12
Page 13 of 31
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
physical origin. These observations are consistent with an enhanced recombination due surface states or recessed grain boundaries. Both effects are likely to occur in this sample due to the high density of grains with non-equivalent crystallographic orientations. CONCLUSION In this work, we have demonstrated that illumination-modulated KPFM can be used to perform time-resolved imaging of multi-dynamical photo-physical processes in organic and inorganic photovoltaic materials. The achieved lateral resolution enables already investigations on optimized nanophase segregated BHJ samples; our future works will focus on comparing the photo-carrier dynamics of several categories of D-A blends. We believe that this approach will make an important contribution to the understanding of the complex interplay between the nanostructure, and the photo-carrier transport, trapping and recombination dynamics in organic and hybrid46 solar cells. It may also be applied to investigate the photo-response of 2D optoelectronic interfaces.47-49 Another key point is that the temporal resolution is only limited by the performances of the illumination setup. Hence, further improvements should follow by using faster illumination chains. This paves the way for local investigations of the minority carrier lifetime in direct bandgap polycrystalline semiconductors like CIGS, and of fast dynamical processes in 2D materials.50 Multi-dynamical KPFM SPV imaging has the potential to become a universal nano-characterization tool in the fields of organic, hybrid and inorganic photovoltaics. EXPERIMENTAL SECTION nc-AFM/KPFM: Experimental nc-AFM/KPFM experiments on organic BHJ have been performed with an Omicron VT-AFM setup, and poly-Si thin films have been investigated with an Omicron VT-AFM XA setup. All experiments have been carried out with beam deflection nc-
ACS Paragon Plus Environment
13
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 14 of 31
AFM in frequency modulation mode under UHV. The topographic imaging was performed with frequency shifts (negative set-points) and vibration amplitudes of a few Hz and a few tens of nanometers, respectively. KPFM measurements have been performed in amplitude modulation (AM) and frequency modulation (FM) modes on organic BHJ and polycrystalline Si (poly-Si), respectively. All data have been acquired with PtIr coated silicon cantilevers (EFM, Nanosensors, resonance frequency in the 45 – 115 kHz range) annealed under vacuum to remove contaminants. The modulation bias (typ. 100 mV in AM-KPFM and 300 mV in FM-KPFM) and the compensation voltage VDC were applied to the cantilever (tip bias Vtip = VDC), and the substrate was grounded. In that configuration, the contact potential difference (CPD) is equal to VDC. In this work, the potentiometric data are presented as compensation bias (Vtip = -CPD, also called for simplicity KPFM potential or surface potential). The ground contact was applied to the ITO substrate in the case of BHJ samples, and to the n-doped Si substrate of poly-Si samples. In our setup, the cantilever thermal stability is insured by keeping the illumination modulation duty constant over the whole frequency range. Besides, in the case of the organic blends, the samples suffer no apparent degradation due to thermal heating. Indeed, reproducible series of images were acquired at different illumination intensities, with no apparent changes in the topographic and dissipation contrasts. Last, the tip stability in terms of effective work function was checked by acquiring series of dynamical images (see for example Figure 5c and 5e). The surface potential (at a given illumination frequency) did not display any significant change during the experiments. Modulated illumination chain: External laser modules have been used for sample illumination through optical viewports of the UHV AFM chambers. PDBS-TQx/PC71BM blends on transparent substrates were illuminated at 515 nm in backside geometry by using a PhoxXplus
ACS Paragon Plus Environment
14
Page 15 of 31
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
module from OmicronLaserage GmBH (rise and fall times