Subscriber access provided by the Henry Madden Library | California State University, Fresno
Light Trapping Enhancement by Silver Nanoantennas in Organic Solar Cells Pavel Voroshilov, Victor Ovchinnikov, Alexios Papadimitratos, Anvar A. Zakhidov, and Constantin Simovski ACS Photonics, Just Accepted Manuscript • DOI: 10.1021/acsphotonics.7b01459 • Publication Date (Web): 10 Apr 2018 Downloaded from http://pubs.acs.org on April 10, 2018
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 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 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.
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 11 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 Photonics
Light Trapping Enhan ement by Silver Nanoantennas in Organi Solar Cells Pavel M. Voroshilov,
∗,†,‡
‡
Vi tor Ov hinnikov,
Zakhidov,
†,§,¶
¶
Alexios Papadimitratos,
Anvar A.
†,‡
and Constantin R. Simovski
†ITMO University, Kronverkskiy pr. 49, 197101 St. Petersburg, Russia ‡Aalto University, P.O. Box 15500, 00076 Aalto, Finland ¶The University of Texas at Dallas, Ri hardson 75080, USA §National
University of S ien e and Te hnology MISiS, Mos ow 119049, Russia
E-mail: p.voroshilovmetalab.ifmo.ru
Abstra t
modes The photovoltai industry regularly requests new s ienti and te hni al solutions for the su
essful global ommer ialization of renewable energy. 1 Considerable eorts have been made to enhan e the e ien y of thin lm solar ells whi h an nd numerous appli ations (see e.g. in 14 ). There is a variety of strategies to in rease the e ien y of thin lm solar
ells, whi h in lude material development, 5,6 interfa e engineering, 7 improvements in fabri ation and pro essing, 8 devi e ar hite ture modernization 9 and optimization of harge transfer. 10 Photon management by light-trapping nanostru tures represents one of the promising ways to improve the performan e of su h solar ells, whose e ien y is limited due to insu ient absorption of light in the a tive (PV) layer. Absorption is limited by PV layer, whi h thi kness lays in subwavelength range at frequen ies of operation. Organi solar ells of ex iton type (we do not refer dye-sensitized and perovskite solar ells as OSCs) have a fundamental limitation on the thi kness of the PV layer below 100-150 nm due to small diusion length of Frenkel's ex itons. 11 Opti al absorption of organi PV materials is also low. Therefore, the larger part of the in ident sunlight passes through the PV layer. An anti-ree ting oating does not resolve the issue of low absorption.
In this paper, we study experimentally the enhan ement of parameters of a typi al organi thin lm solar ell (OSC) a hieved by an originally proposed light trapping stru ture (LTS). Our LTS is an array of silver nanoantennas supporting olle tive modes, what provide subwavelength light enhan ement in the substrate. Low losses in the metal result from the advantageous opti al eld distribution that is a hieved in the broad range of wavelengths. We study the enhan ement of all the main photovoltai (PV) parameters observed in OSC with our LTS, su h as power onversion e ien y (PCE), whi h is ee tively in reased by 18%, ll fa tor (FF), that an also be ee ted positively and short- ir uit urrent Isc improvement. Although the Ag nanoantenna o
upies up to 40% of the photoa tive OSC area, it provides the nearly twi e enhan ed opti al absorption, resulting in overall higher Isc and FF. Improvements of the opti al and ele tri al design of OSC ar hite tures are dis ussed for further enhan ement of performan e.
Keywords Metasurfa e, Nanoantennas, Small mole ules, Organi solar ells, Light trapping, Domino-
ACS Paragon Plus Environment
1
ACS Photonics 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 11
su h as FF and open ir uit voltage Voc :
Even if the thi kness of the PV layer ee tively doubled by a polished rear ele trode it is insuf ient. Moreover, many OSCs imply a transparent ( ondu ting oxide) or absorbing (e.g. Al foil) rear ele trode that does not ompletely ree t light ba k to the a tive layer. In this situation, one has to apply LTSs, usually performed as plasmoni nanostru tures. 12 In 13,14 we suggested a on ept of light trapping employing a metasurfa e of metal nanoantennas, whose resonan es over a very broad band of wavelengths. These resonan es orrespond to olle tive os illations that keep even for a perfe tly ondu ting metal. The eld is lo ally enhan ed in the gaps between the metal elements and the losses in the metal are almost absent. Su h advantageous distribution origins from leaky modes 15 supported by the metasurfa e suggested in. 13 We all them leaky dominomodes by analogy with guided domino-modes whi h initially were revealed in far-infrared 16 and visible 17 ranges for arrays of metal nanobars. Leaky domino-modes were studied theoreti ally in 15 and experimentally in. 18 The aim of the present paper is to experimentally demonstrate the enhan ement of a typi al OSC by LTS with leaky domino-modes. Sin e the pioneering work by Ching W. Tang 19 on two-layer organi OSC based on a mole ular donor-a
eptor stru ture with a PCE of about 1%, the signi ant progress has been a hieved in the improvement of OSC systems, leading to a PCE of over 11% 20 in single-jun tion and over 13% 21 in tandem small-mole ule based devi es with a polished rear ele trode. However, the main drawba k inherent to typi al OSCs by C.W. Tang keeps the absorption of light is insu ient even in these advan ed OSCs. Therefore, high opti al losses still determine the omparatively low e ien y of organi solar photovoltai s ompared to sili on photovoltai s. Usually, in papers on LTSs, the in rease of opti al absorption in the a tive layer is laimed as a su ient eviden e of the e ien y enhan ement be ause it results in the orresponding in rease of the short- ir uit urrent Isc . 22 However, the total PCE of a solar ell is ae ted not only by Isc but also by other parameters
P CE =
Voc · Isc · F F , Pin
(1)
where Pin is the power of the in ident light (referred to the surfa e of our thin lm solar ell and integrated over the spe trum). In a
ordan e with Eq. 1 an LTS in order to grant the enhan ement must not harm both FF and Voc , otherwise, the overall e ien y may de rease. In this paper, we study the ee t of proposed LTS on all mentioned parameters. Our target is the enhan ement of PCE. An expli it example of an OSC that an be drasti ally improved by our LTS is a smallmole ule OSC 14 operating in the near-infrared and transparent for the visible light. In that work, we have hosen a bulk-heterojun tion small-mole ule OSC with phthalo yanine as a donor and fullerene as an a
eptor. We preferred this design as the typi al one that is suf iently simple in fabri ation. In our theoreti al work, 14 we have shown that proposed LTS in reased the absorption of infrared solar light in the tin phthalo yanine (SnP , donor) fullerene (C60, a
eptor) layers of the OSC more than triple. That OSC had two indium tin oxide (ITO) ele trodes (both anode and athode), that made it transparent for visible light, though by a pri e of very low e ien y. 3 Our LTS with its triple enhan ement of the useful absorption promises a nearly double in rease of the short- ir uit urrent and, perhaps, a twi e higher PCE (though this question was not studied). Meanwhile, the transmittan e of visible light is kept su ient for the ee tive transparen y of a window into whi h su h the enhan ed OSC ould be in orporated. For estimation of the ee tive in rease of the short- ir uit urrent, both proper opti al and ele tri al simulations should be performed, as in has been done e.g. in. 23 It is not easy to predi t how both opti al absorption and photo urrent an be optimally improved simultaneously. So in reased opti al absorption (and number of photogenerated arriers) an be a hieved by in reasing photoa tive layer thi kness, however this may lead to overall de rease of the expe ted
ACS Paragon Plus Environment
2
Page 3 of 11 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 Photonics
Transparent device ITO cathode
e
thod ITO ca
ode
Al cath
BCP C60 SnPc
P1
CuPc
rs
PV laye
P2
P1
rs
PV laye
ITO anode glass
E-field, arb. unit
Opaque device
z
0 nm
4 λ=8 0
Ag
y
x
BCP
1
x
y
0
C60
z
z
P2
SnPc CuPc
m y
E-field, arb. unit
x
1
z
z
Ag
z
80 n
λ=4
Al cathode
x
y
ITO anode glass
b
a
c
Figure 1: Normalized ele tri eld intensity distributions in the horizontal planes P1 and P2 (top panel) and in the slightly tilted ross-se tions (bottom panel) for an OSC ontaining ITO anode and athode, λ =840 nm (a) and for a similar OSC with ITO anode and Al athode, λ =480 nm ( ). A tive layer boundaries are shown as white dotted lines (bottom panel). Ag nanoantennas are lo ated on top of ITO anode, embedded into the PV layer. In panel (b) we show the OSC s hemati s for these two ases. Horizontal planes P1 and P2 are shown as bla k dashed lines. Orange arrows show the wave in iden e. transport layer (HTL). It allowed us to reate a double heterojun tion and to in rease the initial e ien y of the prototype. Our new design is typi al and orresponds to works. 9,11 In order to demonstrate the advantageous operation of our LTS, we do not modify the design parameters of the prototype. In this meaning, our study fairly demonstrates the real enhan ement granted by the LTS to a typi al OSC. The HTL of CuP has maximal absorption in the visible region. Therefore, onversion of solar radiation o
urs in the extended spe tral range, whi h in ludes both the near infrared (750-900 nm) part where the absorbing material is SnP and the visible part (550-750 nm), where CuP absorbs. One more dieren e of the present design from that of 14 is the slight shift of the metasurfa e with respe t to the ITO anode. Instead of embedding the nanoantennas into ITO, we lo ate them on the ITO surfa e that allowed us to strongly simplify the fabri ation.
PCE enhan ement due to poor harge olle tion from thi ker layer (due to e.g. short diusion length of arriers, and higher series resistan e of thi ker layer) and thus overall de reased ll fa tor. Therefore in future, we will qualitatively validate the numeri al model of nanoantennas in orporated in a solar ell with the experimental results for further development of the e ient light-trapping on ept for high PCE solar
ells. Unfortunately, in the present work, we did not manage to fabri ate a transparent ITO athode. Therefore, the experimental realization of the exa tly same stru ture as in 14 turned out to be impossible. We have repla ed ITO for the
athode by Al and dropping the fun tionality of the transparen y on entrated on the ee t of light trapping and its impli ations. Sin e in this ase the transparen y of the PV layer is not required, we have deviated from the design of 14 adding a nanolayer of opper phthalo yanine (CuP ) as an additional donor and hole
ACS Paragon Plus Environment
3
ACS Photonics 100
m
25
m
m
2
m
LTS
Glass
ITO Glas s
ITO z y
e
c
h
b a Dx
x
Dy
Transmittance or reflectance, %
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 11
80
60 T (substrate) T (LTS) R (substrate) R (LTS)
40
20
0 300
400
500
600
700
800
900
1000
Wavelength, nm
a
c
b
Figure 2: (a) Photo of the sample with our LTS. (b) S heme of the metasurfa e geometry. The geometri al parameters are as follows: Dx = Dy = 1000 nm, a = 300 nm, b = 150 nm, c = 300 nm, e = 250 nm, h = 35 or 50 nm (two dierent samples). ( ) Transmission and ree tion spe tra were measured for 50-nm sample (LTS) and for bare ITO area (substrate). Fabry-Perot avity is formed in the stru ture that works together with domino-modes ex ited in our nanoantennas and leads to the more advantageous lo ation of hot spots in the near-infrared range (700-850 nm). Meanwhile, in this onguration, these spots do not appear at other wavelengths. The situation hanges when Al repla es ITO as a athode material: hot spots now also appear in the visible range exa tly in favorable lo ations. However, in the near-infrared band hot spots shift away from the PV layer. Anyway, this does not result in noti eable parasiti losses be ause these hot spots do not interse t with the metal. In general, simulations onrmed our expe tations the enhan ement of the useful absorption keeps after the repla ement of ITO by an aluminum. More details on this modeling an be found in the Supporting information. We have prepared two samples with 50 nm and 35 nm thi k Ag nanoantennas on top of the ITO-glass substrate. These samples ontain 1.5 · 106 and 3 · 106 unit ells of our LTS, respe tively. Fabri ation was done using ele tron-beam lithography (EBL) and ele tronbeam evaporation (see Methods and Supporting information). SEM analysis showed a very high quality of our nanoantenna arrays, as an be guessed from Figs. 2 and 3. We measured transmittan e T and ree tan e R for a 50-nm thi k LTS (in iden e from the glass side) and
ompared them with those obtained for the bare
400 nm
Figure 3: S anning ele tron mi ros ope image of our LTS. Although the design modi ations listed above may seem insigni ant, they an drasti ally ae t the performan e of the LTS. It be omes ne essary to study the inuen e of the Al athode on the lo al ele tri eld distribution inside the PV ell. We have performed full-wave (CST) ele tromagneti simulations of two PV devi es: one with a transparent ITO
athode and another with an opaque Al athode. Complex refra tive indexes were taken from. 2426 Ele tri eld intensity distributions for these two ases are presented in Fig. 1. Indeed, the opaque metal lm fully hanges the hot spots arrangement in the bulk of the stru ture at all wavelengths. Obviously, in presen e of a transparent athode, an ee tive
ACS Paragon Plus Environment
4
Page 5 of 11
ACS Photonics
-4.0
-5.2 -5.1
-4.2 -6.1
-6.4 BCP
C70
SnPc
Solar light
e-
-4.3
Active layers
a
Without NAs With NAs
-1
b
Current (mA/cm²)
h+
-3.0
Al (cathode)
-4.8
CuPc
trate
Glass subs
-3.5
C70
ITO anode
SnPc
50 nm
BCP [10 nm] C70 [55 nm] SnPc:C70 [30 nm] CuPc [15 nm]
CuPc
Silver
Al cathode PV layers
ITO (anode)
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
Energy relative to vacuum level, eV
0
-2
-3
-4 Device
Isc [mA/cm2] Voc [V] FF [%] PCE [%]
W/o NAs W/ NAs
-5 0
0.05
0.1
0.15
4.614 4.955
0.2
0.366 0.276
0.25
41.6 42.9
0.3
0.702 0.587
0.35
0.4
Voltage (V)
c
Figure 4: (a) Cross-se tion s hemati s of the OSC with C70 as an a
eptor layer and 50-nm thi k silver nanoantennas. (b) Energy-level band diagram for the urrent onguration. ( ) IV urves measured under illumination for two ases: with (red line) and without (blue line) nanoantennas. whereas C60 has higher absorption at longer waves. Our rst sample with 50 nm-thi k nanoantennas has the following stru ture above the glass substrate: ITO anode (150 nm) / CuP (15 nm) / SnP :C70 (30 nm) / C70 (55 nm) / BCP (10 nm) / Al (70 nm). The orresponding s hemati is shown in Fig. 4 (a). This onguration with o-evaporated phthalo yanine and fullerene leads to the formation of two donor-a
eptor heterojun tions: CuP /C70 and SnP /C70. We illuminated our sample with an AM 1.5G solar simulator from the side of glass through a 1 mm aperture. We measured the IV dependen ies of our sample illuminating either the 1 mm areas omprising our nanoantennas or 1 mm areas free of them. In this way, we have olle ted su ient statisti s for a reliable plotting two IV urves one for an OSC enhan ed by our LTS and another for a bare OSC (with only antiree tive oating). These urves are shown in Fig. 4 and demonstrate a noti eable in rease of Isc aused by our LTS. Photo urrent generated in the nanoantenna area rea hes the value 5 mA/cm2 , whereas the stru ture without nanoantennas gives 4.6 mA/cm2 . This 8% in rease of the photo urrent orresponds to a higher in rease of the PV absorban e due to a nonlinear urrent response of su h OSC. 27 Due to 40% of the area overed by Ag nanoantennas, the photoa tive area that photogenerates arriers and provides Isc is only 60% of the similar area in a referen e OSC. Therefore the
substrate (glass and ITO). The result is shown in Fig. 2 ( ). The averaged ree tan e growth is about 10%. Noti e, that this negative ee t keeps for the OSC, where it is over ompensated by the light-trapping ee t. The target of these simulations is twofold. Firstly, we may estimate the in rease of the parasiti absorption A introdu ed by our LTS. Sin e the losses in Ag elements are negligibly small, parasiti absorption an be attributed to opti al losses in ITO. From energy balan e, we have A = 1 − R − T . For integral absorption in the operational range λ =400-900 nm we have nearly 5% in absen e of our LTS and nearly 8% in its presen e. This is an a
eptable harm. Se ondly, we may see that the ree tan e grows mu h less that it an be estimated using the geometri al opti s. Really, Ag nanoantennas over almost 40% of the area, but the ree tan e is λ-dependent and mu h lower than 0.4. This is so be ause the eigenmodes of our LTS are ex ited by the normally in ident plane wave, what makes geometri al opti s inadequate. Further, inspe ting the opaque design depi ted in Fig. 1 we de ided to additionally fabri ate a variant of the OSC with C70 fullerene derivative instead of C60. If the impa t of our LTS will be more signi ant for the stru ture with C70, it will onrm the shift of lighttrapping regime (predi ted by our simulations) from the IR range to the visible range for Al
athode. This should be so be ause C70 has higher absorption than C60 below λ =700 nm,
ACS Paragon Plus Environment
5
ACS Photonics 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 11
/ SnP :C60 (30 nm) / C60 (25 nm) / BCP (10 nm) / Al (70 nm). Sin e C60 is less absorptive material in the visible range and its thi kness de reased, the impa t of our LTS in the photo urrent also de reased ompared to the previous ase 4.6% instead of 8%, as
an be seen from Fig. 5 ( ). However, thinner metal elements improved FF by 19% versus 2.6%, whereas Voc almost kept un hanged. Total averaged value of the PCE enhan ement is 18% that is a quite ompetitive result and is higher than the gain provided by known plasmoni ounterparts of our LTS for similar OSCs (see e.g. 28 ). It should also be noted, that our samples were not en apsulated. Therefore, in order to prevent the degradation of organi materials and oxidation of Ag, our measurements were performed at room temperature and in an inert atmosphere. In order to prevent Ag nanoantennas tarnishing during transportation from the
lean room to the laboratory, where the solar
ell was prepared, we used a polymer prote tive oating. This oating was removed in the laboratory with a etone. However, all samples were treated by gentle UV-ozone for additional
leaning before the organi layer deposition. 29 So, a ertain oxidation of the thin silver elements might take pla e during this treatment. This is another reason, why the in rease of the PV absorption granted by our LTS in reality turned out to be lower than that predi ted by numeri al simulations. Hopefully, the experimental enhan ement oered by our nanoantennas will in rease on e more, if the fabri ation fa ilities will allow lithography pro ess and solar ell preparation in the same lean room. To summarize, in this paper, we have theoreti ally and experimentally proved that our original LTS from silver nanoantennas an suf iently in rease the e ien y of OSCs. First of all, our numeri al simulations have shown advantageous eld distributions that predi ted enhan ement of the useful absorption. Se ondly, we have fabri ated several OSCs with dierent materials and measured gain in PV parameters
aused by nanostru tures. Sin e total PCE of a solar ell is ae ted not only by Isc , but also by ll fa tor FF and open ir uit voltage (Voc ),
a tual enhan ement of Isc , provided by the LTS is not 5 mA/cm2 , but 5/0.6 = 8.3 mA/cm2 . The Ag nanoantenna array o
upied area is not reating arriers but is reating on entration and enhan ement of the photoni ele tri al eld whi h results in the nearly twi e enhan ement of the opti al absorption of light, resulting in 8.3 mA / 4.6 mA = 1.8 times enhan ed on entration of arriers. We assume here that the edges of sharp Ag stru tures do not provide better harges olle tion (sin e the nanoantennas an olle t only ele trons by low work fun tion), but on the ontrary, are reating leakage of arriers and quen hing of ex itons. Despite all those negative ee ts of Ag nanoantennas: shading the useful area, wrong work fun tion, re ombination of ele trons and holes on Ag nanostru tures, we have still observed the overall sizable enhan ement of Isc = 5 mA (8% more than in the referen e ell) at a nearly twi e small photoa tive area. Fill fa tor al ulated from IV hara teristi s is improved by 2.6% for our LTS (see inset of Fig. 4 ( )). Unfortunately, the open- ir uit voltage Voc of the area with nanoantennas de reased by 0.09 V, i.e. 25% ompared to the bare OSC. This harm overshadows the positive gain in photo urrent and negatively ae ts the overall PCE. The possible reason for the fall in Voc ould be quite substantial dimensions of metal nanoantennas ompared with ultrathin a tive layers. Really, Voc in OSCs is dire tly proportional to the dieren e between the Fermi levels, and the ITO ele trode having the bandgap in the ultraviolet has an impa t on Voc . Silver elements having the ohmi onta t with ITO may modify the ee tive bandgap of the ele trode. As a result, the dieren e of the Fermi levels squeezes and Voc be omes smaller. Parameters of the OSC with nanoantennas and without nanoantennas are shown in the inset of Fig. 4 ( ). These data are averaged over several sample areas, as explained above. Our next example is totally identi al to the previous one ex lude repla ement of C70 by C60 and thinner Ag elements (35 nm instead of 50 nm). The stru ture is shown in Fig. 5 (a). It onsists of the following layers on the glass plate: ITO (150 nm) / CuP (15 nm)
ACS Paragon Plus Environment
6
Page 7 of 11
ACS Photonics
-4.0
-4.5
BCP
C60
SnPc
-6.2 -6.4
-4.3 Al (cathode)
-5.2 -5.1
-3.0
e-
Device W/o NAs W/ NAs
-0.5
Current (mA/cm²)
C60
SnPc
-4.8
h+
CuPc
trate
Glass subs
CuPc
35 nm
Silver
-3.5 ITO (anode)
Al cathode PV layers ITO anode
BCP [10 nm] C60 [25 nm] SnPc:C60 [30 nm] CuPc [15 nm]
Energy relative to vacuum level, eV
0
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
Isc [mA/cm2] Voc [V] FF [%] PCE [%] 2.869 2.997
0.396 0.376
36.5 43.4
0.415 0.489
-1 -1.5 -2 -2.5 Without NAs With NAs
-3 -3.5
Solar light
Active layers
a
b
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
Voltage (V)
c
Figure 5: (a) Cross-se tion s hemati s of the OSC with C60 as an a
eptor layer and 35-nm thi k silver nanoantennas. (b) Energy-level band diagram for the urrent onguration. ( ) IV urves measured under illumination for two ases: with (red line) and without (blue line) nanoantennas. we have onsidered several design solutions and dis ussed the inuen e of the devi e ar hite ture on LTS performan e. Experimental data for our OSCs based on C60 fullerene and enhan ed by 35 nm thi k nanoantennas show the enhan ement 18% that qualitatively ts the results of simulations. Other experimental data are also in line with theoreti al expe tations. We believe that the present results are helpful for further development of LTSs for thin lm solar ells. In our future designs, we will hoose nanoantennas with proper work fun tion ontributing into olle tion of holes at ITO side and with better lo ation allowing us to avoid the 40% shading of the PV layer.
oated glass substrates pur hased from Lumines en e Te hnology Corp by EBL and lift-o pro ess in the Mi ronova Nanofabri ation Centre of Aalto University. Pro ess ow diagram is shown in Fig. 6. Substrates were preliminary
leaned with a etone and isopropanol (IPA) in an ultrasoni bath at 40◦ C and rinsed in DIwater. A opolymer resist (MMA) with on entration of solids 11% in Ethyl La tate and polymer resist (950PMMA A) with on entrations of solids 2.25% in Anisole (both from Mi ro hem) were sequentially deposited by spin oating. Samples were dried on a hot plate for 3 minutes at a temperature of 160 degrees after ea h spin- oating pro edure resulting in formation of 400 and 100 nm layers of MMA and PMMA, respe tively. A thin (20 nm) sa ri ial layer of aluminum was sputtered. The purpose of this layer is to solve the problem with e-beam fo using on a diele tri layer during EBL. E-beam exposure was done using Viste EBPG 5000Plus ES tool at 100 keV with 4 nA urrent beam and 1000 µC/cm2 e-beam dose. Then, samples were pro essed in et hing mixture PWS 80-16-4 (65) from Honeywell ontaining phosphori a id and nitri a id to ompletely remove the sa ri ial aluminum layer. Next, PMMA and MMA were developed in solution of 1:3 MIBK:IPA during 30 se and then rinsed in IPA during 30 se to terminate development and prevent s umming. Silver was deposited by using ele tron beam evaporation at a rate of 2 Å/s and operating pressure of 10−7 mbar. Finally, non-exposed areas of poly-
Methods Substrate cleaning
Polymer spin-coating
Al sputtering
E-beam exposure
Lift-off
Ag evaporation
Development
Al etching
Glass
MMA
Al
ITO
PMMA
Ag
Figure 6: Pro ess ow diagram illustrating fabri ation steps of LTS. Several samples with silver nanoantennas were fabri ated on top of 150 nm-thi k ITO
ACS Paragon Plus Environment
7
ACS Photonics 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 11
Referen es
mer layers were washed out together with silver on their surfa e by a etone in the ultrasoni bath. Samples were hara terized by using AFM (DI3100), SEM (Zeiss Supra 40), and multifun tional tool for ree tion (R), transmission (T) and quantum e ien y (QE) measurements (QEX10). Small mole ule OSCs were prepared in the multi-sour e high-va uum system (Angstrom Engineering In ., Canada) by sequential thermal evaporation onto ITO-glass substrates with our nanostru tures of the following lms: a 15 nm thi k opper(II) phthalo yanine (from H.W. Sanders Corp) HTL, a 30 nm thi k oevaporated in equal quantity SnP :fullerene mixed D:A layer, an additional ontinuous fullerene layer (25 or 55 nm), BCP (10 nm) and Al (70 nm) layers. We used C60 and C70 fullerenes (both >98% from Nano-C). Base pressure of the hamber was kept around 10−8 mbar during evaporation. Evaporation rate for separate layers was 0.5 Å/s, while oevaporation rate was equal to 0.25 Å/s for ea h parti ular material. OSCs were hara terized with an AM 1.5G solar simulator alibrated to one sun (100 mW/cm2 ) through aperture holes of 1 mm in diameter, whi h provided sele tive illumination of nanostru tured regions only.
(1) Bosio, A.; Romeo, A. Thin Film Solar Cells: Current Status and Future Trends; Energy s ien e, engineering and te hnology series; Nova S ien e Publishers, 2011. (2) S hmidt-Mende, L.; Wei kert, J. Organi and Hybrid Solar Cells: An Introdu tion; De Gruyter Textbook; De Gruyter, 2016. (3) Lunt, R. R.; Bulovi , V. Transparent, near-infrared organi photovoltai solar
ells for window and energy-s avenging appli ations. Applied Physi s Letters 2011, 98, 61. (4) Lipomi, D. J.; Bao, Z. Stret hable, elasti materials and devi es for solar energy onversion. Energy & Environmental S ien e 2011, 4, 33143328. (5) Sun, S.-S.; Sari ift i, N. S. Organi photovoltai s: me hanisms, materials, and devi es; CRC press, 2005. (6) Brédas, J.-L.; Norton, J. E.; Cornil, J.; Corop eanu, V. Mole ular understanding of organi solar ells: the hallenges. A
ounts of hemi al resear h 2009, 42, 16911699.
A knowledgement This work was par-
tially supported by the Ministry of Edu ation and S ien e of the Russian Federation (Grant 14.Y26.31.0010) and Nokia Foundation. The authors thank Mikhail Omelyanovi h, Kristina Ian henkova, and Matthias Mes hke for useful dis ussions. We also a knowledge the provision of fa ilities and te hni al support by Aalto University at Mi ronova Nanofabri ation Centre.
(7) Steim, R.; Kogler, F. R.; Brabe , C. J. Interfa e materials for organi solar ells. Journal of Materials Chemistry 2010, 20, 24992512. (8) Krebs, F. C. Fabri ation and pro essing of polymer solar ells: a review of printing and oating te hniques. Solar energy materials and solar ells 2009, 93, 394 412.
Supporting information
(9) Heremans, P.; Cheyns, D.; Rand, B. P. Strategies for in reasing the e ien y of heterojun tion organi solar ells: material sele tion and devi e ar hite ture. A
ounts of hemi al resear h 2009, 42, 17401747.
The Supporting Information (SI) is available free of harge on the ACS Publi ation website. The SI in ludes details of 3D ele tromagneti modeling and additional simulated ele tri eld distributions at dierent wavelengths. Moreover, the SI ontains omplementary SEM and AFM analysis of samples.
(10) Deibel, C.; Strobel, T.; Dyakonov, V. Role of the harge transfer state in organi
ACS Paragon Plus Environment
8
Page 9 of 11 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 Photonics
(19) Tang, C. W. Two-layer organi photovoltai ell. Applied Physi s Letters 1986, 48, 183185.
donora
eptor solar ells. Advan ed Materials 2010, 22, 40974111. (11) Riede, M.; Mueller, T.; Tress, W.; S hueppel, R.; Leo, K. Small-mole ule solar ells status and perspe tives. Nanote hnology 2008, 19, 424001.
(20) Deng, D.; Zhang, Y.; Zhang, J.; Wang, Z.; Zhu, L.; Fang, J.; Xia, B.; Wang, Z.; Lu, K.; Ma, W.; Wei, Z. Fluorinationenabled optimal morphology leads to over 11% e ien y for inverted smallmole ule organi solar ells. Nature
ommuni ations 2016, 7, 13740.
(12) Atwater, H. A.; Polman, A. Plasmoni s for improved photovoltai devi es. Nature materials 2010, 9, 205213. (13) Simovski, C.; Morits, D.; Voroshilov, P.; Guzhva, M.; Belov, P.; Kivshar, Y. Enhan ed e ien y of light-trapping nanoantenna arrays for thin-lm solar
ells. Opti s express 2013, 21, A714 A725.
(21) Heliatek sets new Organi Photovoltai world re ord e ien y of 13.2%. 2016; http://www.heliatek. om/ en/press/press-releases/details/ heliatek-sets-new-organi -photovoltai -world (22) Lasnier, F. Photovoltai engineering handbook; CRC Press, 1990.
(14) Voroshilov, P. M.; Simovski, C. R.; Belov, P. A. Nanoantennas for enhan ed light trapping in transparent organi solar
ells. Journal of Modern Opti s 2014, 61, 17431748.
(23) Mirsafaei, M.; Fallahpour, A. H.; Lugli, P.; Rubahn, H.-G.; Adam, J.; Madsen, M. The inuen e of ele tri al ee ts on devi e performan e of organi solar ells with nano-stru tured ele trodes. S ienti reports 2017, 7, 5300.
(15) Voroshilov, P. M.; Simovski, C. R. Leaky domino-modes in regular arrays of substantially thi k Photoni s and metal nanostrips. Nanostru tures-Fundamentals and Appli ations 2016, 20, 1830.
(24) O'Connor, B. Organi Ele troni s on Fibers for Energy Conversion Appli ations. Ph.D. thesis, University of Mi higan, 2009.
(16) Martin-Cano, D.; Nesterov, M.; Fernandez-Dominguez, A.; Gar iaVidal, F.; Martin-Moreno, L.; Moreno, E. Plasmons for subwavelength terahertz
ir uitry. Opti s Express 2010, 18, 754764.
(25) El-Nahass, M.; Yaghmour, S. Ee t of annealing temperature on the opti al properties of thermally evaporated tin phthalo yanine thin lms. Applied Surfa e S ien e 2008, 255, 1631 1636.
(17) Simovski, C.; Luukkonen, O. Tapered plasmoni waveguides with e ient and broadband eld transmission. Opti s Communi ations 2012, 285, 33973402.
(26) Rand, B. P.; Li, J.; Xue, J.; Holmes, R. J.; Thompson, M. E.; Forrest, S. R. Organi Double-Heterostru ture Photovoltai Cells Employing Thi k Tris (a etyla etonato) ruthenium (III) Ex iton-Blo king Layers. Advan ed Materials 2005, 17, 27142718.
(18) Sinev, I. S.; Voroshilov, P. M.; Mukhin, I. S.; Denisyuk, A. I.; Guzhva, M. E.; Samusev, A. K.; Belov, P. A.; Simovski, C. R. Demonstration of unusual nanoantenna array modes through dire t re onstru tion of the near-eld signal. Nanos ale 2015, 7, 765770.
(27) Koster, L.; Mihailet hi, V.; Xie, H.; Blom, P. Origin of the light intensity dependen e of the short- ir uit urrent of polymer/fullerene solar ells. Applied Physi s Letters 2005, 87, 203502.
ACS Paragon Plus Environment
9
ACS Photonics 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
(28) Ahn, S.; Rourke, D.; Park, W. Plasmoni nanostru tures for organi photovoltai devi es. Journal of Opti s 2016, 18, 033001. (29) Chen, C.-W.; Hsieh, P.-Y.; Chiang, H.-H.; Lin, C.-L.; Wu, H.-M.; Wu, C.-C. Topemitting organi light-emitting devi es using surfa e-modied Ag anode. Applied physi s letters 2003, 83, 51275129.
ACS Paragon Plus Environment
10
Page 10 of 11
Page 11 of 11 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 Photonics
For Table of Contents Use Only 82x40mm (300 x 300 DPI)
ACS Paragon Plus Environment