All-solution based device engineering of multilayer polymeric ...

6 downloads 1662 Views 282KB Size Report
blend of the electron and hole accepting polymers, both ma- terials can be ..... Huang, P. F. Miller, J. S. Wilson, A. J. deMello, J. C. deMello, and D. D. C. Bradley ...
APPLIED PHYSICS LETTERS 94, 173303 共2009兲

All-solution based device engineering of multilayer polymeric photodiodes: Minimizing dark current Panagiotis E. Keivanidis,1,a兲 Siong-Hee Khong,1 Peter K. H. Ho,2 Neil C. Greenham,1 and Richard H. Friend1 1

Optoelectronics Group, Cavendish Laboratory, J. J. Thomson Avenue, Cambridge CB3 0HE, United Kingdom 2 Department of Physics, National University of Singapore, Lower Kent Ridge Road, Singapore S117542, Singapore

共Received 1 February 2009; accepted 18 March 2009; published online 30 April 2009兲 We present photodiodes fabricated with several layers of semiconducting polymers, designed to show low dark current under reverse bias operation. Dark current minimization is achieved through the presence of additional polymer layers that reduce charge carrier injection in reverse bias, when in contact with the device electrodes. All polymer layers are deposited via spin coating and are photocross-linked for allowing further polymer layer deposition, by using a bis-fluorinated phenyl-azide photocross-linking agent. Dark current density values as low as 40 pA/ mm2 are achieved with a corresponding external quantum efficiency 共EQE兲 of 20% at a reverse bias of ⫺0.5 V when an electron-blocking layer is used. Dark current is further reduced when both an electronand a hole-blocking layer are used but the EQE falls significantly. © 2009 American Institute of Physics. 关DOI: 10.1063/1.3120547兴 Conjugated polymers are attractive materials to use in photodetector arrays that may be patterned by using techniques such as printing. These arrays can be employed in sensor-related applications1–3 and in x-ray medical imaging.4 For obtaining high external quantum efficiencies 共EQEs兲 from these devices, bulk heterojunctions of donor-acceptor blends are used as the active layers for achieving photoinduced charge separation.5–7 Apart from fast response times,8 photodetectors require low dark currents at the operating voltage9 for maximizing their dynamic range. Band structure engineering through the use of multilayer structures is an attractive way to control dark currents in reverse bias, and it has been applied in organic photodetectors based on small molecules.10 Such multilayer structures are difficult to achieve with solution-processed devices due to the difficulty of finding orthogonal solvents for the successive layers.11–13 Multilayer fabrication techniques include lamination,14 inkjet printing,15 and active layer transfer via stamping.16–18 For the bulk heterojunction formed as a 共partly demixed兲 blend of the electron and hole accepting polymers, both materials can be presumed to be present at each carriercollecting electrode. Such an interpenetrating network of the blend components in the active layer will allow both forward and reverse injection10,19 of electron and hole currents directly into the preferred polymer 关Fig. 1共a兲兴. Carrier injection in the dark can be suppressed to some degree by controlling the electrode work functions19,20 or by the use of suitable blocking layers at the electrodes.18 As Fig. 1共b兲 shows, these layers provide a barrier to the injection of one carrier in reverse bias while still allowing the passage of the other carrier to the electrode. Cross-linking techniques have provided useful means for the development of organic optoelectronic devices.21,22 We use here a cross-linking technique23 to fabricate photodiodes of an all-polymer bulk heterojunction layer surrounded by a兲

Electronic mail: [email protected].

0003-6951/2009/94共17兲/173303/3/$25.00

appropriate carrier blocking layers that reduce the dark current. Devices with the optimum layer structure exhibit significant suppression of the dark current density while maintaining relatively high EQEs under realistic reverse bias operating conditions.

(a)

Blend

P3HT 2.7 eV F8TBT 3.15 eV

(1) 5.1 eV PEDOT:PSS

(b)

EB layer

(2)

Al 4.3 eV P3HT 4.9 eV F8TBT 5.37 eV

Blend HB layer

PFB 1.95 eV TFB 2 eV F8TBT 3.15 eV Al 4.3 eV 5.1 eV

PFB 4.87 eV

PEDOT:PSS TFB 5.16 eV

F8TBT 5.37 eV

FIG. 1. 共Color online兲 共a兲 A simplified scheme that visualizes the process of dark carrier injection in a photodiode under reverse bias operation. Electron injection 共1兲 takes place from the hole-collecting electrode to the LUMO levels and/or hole injection 共2兲 takes place from the electron-collecting electrode to the HOMO levels of the components of the bulk heterojunction. 共b兲 The effect of blocking layers on the process of charge injection from the electrodes in polymer-based photodiode under reverse bias. Both steps 共1兲 and 共2兲 can be hindered when layers with appropriate frontier orbitals are used as interlayers with electron-blocking 共EB兲 and hole-blocking 共HB兲 properties.

94, 173303-1

© 2009 American Institute of Physics

Downloaded 12 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

173303-2

Keivanidis et al.

FIG. 2. 共a兲 The chemical structure of the polymers and the cross linker 共FPA兲 used for this study. 共b兲 Dark J-V curves for photodiodes made of device A 共F8TBT:P3HT blend, open squares兲, device B 共P3HT/ F8TBT:P3HT bilayer open circles兲, device C 共PFB/F8TBT:P3HT bilayer, open up triangles兲, and device D 共TFB/F8TBT:P3HT/F8TBT multilayer, open down triangles兲. 共c兲 Dark J-V curves of multilayer photodiodes made of device E 共PFB/F8TBT:P3HT/F8TBT兲 with a total film thickness of 120 nm 共open squares兲, 140 nm 共open circles兲, and 210 nm 共open down triangles兲. The dark J-V curve of device E in which the F8TBT:P3HT was not photocross-linked is also depicted 共nominal total film thickness 140 nm, solid circles兲. For all devices glass/ITO/PEDOT:PSS and Al were the holecollecting 共bottom兲 and electron-collecting 共top兲 electrodes, respectively.

The materials used for the active layer7 are poly共关9,9-dioctylfluorene兴-2,7-diyl-alt-关4,7-bis共3-hexylthien5-yl兲-2,1,3-benzothiadiazole兴-2,2,-diyl兲 共F8TBT兲 and poly共3hexylthiophene兲 共P3HT兲 共regioregularity of 94%兲. As-spun films with an average thickness of 80 nm were prepared from 1:1 ratio mixtures of these materials in xylene. We have used either P3HT, poly关2,7-共9,9-di-n-octylfluorene兲-co-共1,4phenylene关共4-sec-butylphenyl兲imino兴-1,4-phenylene兲兴 共TFB兲 or poly关9 , 9⬘-di-octylfluorene-co-bis-N , N⬘-共4-butylphenyl兲bis-N , N⬘-phenyl-1,4-phenylenediamine兴 共PFB兲 as the materials for the electron-blocking 共EB兲 layers. We also have used F8TBT as the material for the hole-blocking 共HB兲 layer. Chemical structures are shown in Fig. 2共a兲. A Dektak 6M profilemeter was used for film thickness determination. The procedure of photodiode fabrication has been described elsewhere.7 Poly共3,4-ethylenedioxythiophene兲-poly共styrenesulfonate兲 共PEDOT:PSS兲 was used as the hole-collecting electrode in these devices. A control device made of 1:1 blend of P3HT:F8TBT 共structure A, glass/indium tin oxide 共ITO兲/PEDOT:PSS/P3HT:F8TBT/Al兲 was used as a reference. The effect of an intermediate EB layer was studied in structure B 共glass/ITO/PEDOT:PSS/P3HT/P3HT:F8TBT/Al兲 and structure C 共glass/ITO/PEDOT:PSS/PFB/P3HT:F8TBT/ Al兲. Finally, we have investigated the effects of both EB and HB layers in structure D 共glass/ITO/PEDOT:PSS/ TFB/P3HT:F8TBT/HB/Al兲 and structure E 共glass/ITO/ PEDOT:PSS/PFB/P3HT:F8TBT/HB/Al兲. The use of water-soluble bis-fluorinated phenyl-azides 共FPAs兲 as cross-linking agents has been described elsewhere.23 The employed sulfonamide photocross-linking agent 关FPA, Fig. 2共a兲兴 is a member of the class of organicsoluble FPAs. Sterically hindered FPAs are capable of cross linking organic polymer semiconductors at low concentrations without causing significant loss of hole or electron mobility, or of photoluminescence 共PL兲 efficiencies. Rendering

Appl. Phys. Lett. 94, 173303 共2009兲

the polymer layers insoluble prior to subsequent layer deposition was achieved by adding the cross-linking agent 共2 wt %兲 in the corresponding polymer solution. Spin coating was then performed and the resulting films were exposed to UV irradiation 共254 nm兲 for 5 min. After the exposure to UV light, solvent spin rinsing followed to remove the noncrosslinked polymer residues for all structures except device E for which spin rinsing of the second photocross-linked layer was omitted. The thickness of the photocross-linked EB layers was 10–20 nm whereas the thickness of the HB layer was between 10 and 30 nm. The thickness of the as-spun active layers was varied by adjusting the spin-coating speed. The effect of thermal treatment by the annealing of each photocross-linked layer was investigated in device D prior to the subsequent layer deposition. Annealing was performed at ⬃180 ° C for 1 h. For all fabricated devices, annealing was performed at 110 ° C for 15 min after the metal cathode deposition. All annealing steps were performed in N2 environment. Figure 2共b兲 compares the current density-voltage 共J-V兲 curves of the control blend 共device A兲, the bilayer 共devices B and C兲, and the multilayer 共device D兲 structures. The control device A exhibits high values of dark current density across the whole range of reverse bias voltages being studied. Using P3HT as the EB layer 共device B兲 does not reduce the dark current density of the device; for reverse biases beyond ⫺1 V the dark current density is increased in respect to the control device. Replacing P3HT with PFB 共device C兲 reduces the dark current density in respect to the control. In the low negative voltage regime the dark current density is reduced by approximately an order of magnitude 共to 40 pA/ mm2 at ⫺0.5 V兲, whereas the reduction is less at more negative voltages 共three times reduced at ⫺4 V兲. Further reduction of the dark current density is observed in multilayer device D. Figure 2共c兲 focuses with more detail on the effect of the multilayer device arrangements on the dark current density. For device E, we varied the thickness of the active layer. An increase in the device dark current is found as the total film thickness of the device is reduced, consistent with the electric field-driven injection mechanism. For the thickest film the dark current density reduces significantly in respect to the control device of Fig. 2共b兲. The need for rendering the photoactive middle layer insoluble prior to the deposition of the HB layer is also manifested in Fig. 2共c兲. Without cross linking of the active layer, partial removal and subsequent thickness reduction of this layer occur during HB deposition, and the dark current density is at least one order of magnitude higher. Figure 3 presents the EQEs of the photodiodes under a range of reverse bias conditions. We note that the onset for photocurrent near 650 nm matches the absorption for P3HT and F8TBT. For all cases, in respect to their short circuit mode, the photodiodes exhibit improved EQEs when operated under reverse bias. We attribute this observation to the effects of field-assisted separation and collection of the photogenerated carriers in the device. The addition of a P3HT EB layer 共device B兲 does not cause large changes in the short-circuit EQE 关Fig. 3共b兲兴 but the increase of the photoactive layer thickness reduces the short-circuit EQE. Compared to the control device, the short-circuit EQE is slightly reduced when PFB is used for the EB interlayer 关device C, Fig. 3共c兲兴 but it increases to 20% at ⫺0.5 V and to 40% at ⫺4 V.

Downloaded 12 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

173303-3

Appl. Phys. Lett. 94, 173303 共2009兲

Keivanidis et al.

coating, using photocross-linking of underlying layers. Devices that use a photocross-linked electron-blocking layer adjacent to the PEDOT:PSS anode show good performance, but devices that require photocross-linking of the charge generating bulk polymer heterojunction are less effective due to caused changes associated with the cross-linking chemistry. The authors would like to thank EPSRC for funding through the Next Generation Electrophotonics program Grant No. EP/C540336, and CDT for providing the fluorene copolymers of this study. S.H.K. thanks the Cambridge Commonwealth Trust and Trinity Hall Cambridge for financial support and P.E.K. thanks Dr. C. Groves for useful discussions. 1

FIG. 3. Spectral external quantum efficiency of photodiodes. 共a兲 Device A 共F8TBT:P3HT blend兲 at 0 V 共open squares兲, ⫺0.25 V 共open circles兲, ⫺1 V 共open triangles兲. 共b兲 Device B 共P3HT/F8TBT:P3HT bilayer兲 at 0 V with a thick F8TBT:P3HT 共120 nm兲 layer 共open squares兲 and a thin F8TBT:P3HT 共60 nm兲 layer 共open circles兲. 共c兲 Device C 共PFB/F8TBT:P3HT bilayer兲 at 0 V 共open squares兲, ⫺0.25 V 共open circles兲, ⫺0.5 V 共open up triangles兲, ⫺1 V 共open down triangles兲, ⫺4 V 共open tilted squares兲. 共d兲 Device D 共TFB/ F8TBT:P3HT/F8TBT multilayer兲 at 0 V 共open squares兲, ⫺0.5 V 共open circles兲, ⫺1 V 共open up triangles兲, ⫺2 V 共open down triangles兲, ⫺3 V 共open tilted squares兲, and ⫺4 V 共open tilted triangles兲. The EQEs of device E 共PFB/F8TBT:P3HT/F8TBT multilayer兲 at 0 V 共solid squares兲 and ⫺4 V 共solid tilted triangles兲 are also depicted. For all devices glass/ITO/ PEDOT:PSS and Al were the hole-collecting 共bottom兲 and electroncollecting 共top兲 electrodes, respectively.

For devices with both EB and HB layers 共devices D and E兲, the short-circuit EQE is considerably reduced in respect to the control 关Fig. 3共d兲兴. Based on experiments of thickness determination and of UV-Vis and PL spectroscopy of the F8TBT:P3HT 1:1 blend, before and after cross linking, we attribute the reduced EQE of the multilayer devices partially to the cross linking-induced disruption of the ␲-conjugation of the polymer backbones. When TFB is the EB layer 共device D兲, the peak EQE recovers to ⬃5% at ⫺4 V 关Fig. 3共d兲兴 while maintaining a relatively low dark current density of 130 pA/ mm2 at that voltage 关Fig. 2共b兲兴. When PFB is the EB layer 共device E兲 the maximum EQE does not exceed 1%. In addition, the dark current density of device D is a little lower than device E presumably due to the additional annealing step followed in the device fabrication protocol. Thin TFB layers on PEDOT:PSS allow easy injection of holes into light-emissive polymers,24,25 and the results here indicate that hole transport in the reverse direction is also effective. In comparison to TFB, the lower ionization potential of PFB does not easily account for the lower EQE measured for device E. We note however that PFB has a considerably lower hole mobility than TFB26 and this may cause a reduced electric field to be present in the charge-generating bulk heterojunction layer. In conclusion, we have described a method for device architecture that enables the controlled design and fabrication of multilayer polymeric photodiodes via successive spin

S. C. J. Meskers, J. K. J. van Duren, R. A. J. Janssen, F. Louwet, and L. Groenendaal, Adv. Mater. 共Weinheim, Ger.兲 15, 613 共2003兲. 2 M. Ramuz, L. Burgi, C. Winnewisser, and P. Seitz, Org. Electron. 9, 369 共2008兲. 3 T. N. Ng, W. S. Wong, M. L. Chabinyc, S. Sambandan, and R. A. Street, Appl. Phys. Lett. 92, 213303 共2008兲. 4 T. Agostinelli, M. Campoy-Quiles, J. C. Blakesley, R. Speller, D. D. C. Bradley, and J. Nelson, Appl. Phys. Lett. 93, 203305 共2008兲. 5 C. J. Brabec, N. S. Sariciftci, and J. C. Hummelen, Adv. Funct. Mater. 11, 15 共2001兲. 6 A. Gadisa, F. Zhang, D. Sharma, M. Svensson, M. R. Andersson, and O. Inganas, Thin Solid Films 515, 3126 共2007兲. 7 C. R. McNeill, J. J. M. Halls, R. Wilson, G. L. Whiting, S. Berkebile, M. G. Ramsey, R. H. Friend, and N. C. Greenham, Adv. Funct. Mater. 18, 2309 共2008兲. 8 P. Peumans, A. Yakimov, and S. R. Forrest, J. Appl. Phys. 95, 2938 共2004兲. 9 J. C. Blakesley and R. Speller, Med. Phys. 35, 225 共2008兲. 10 J. Xue and S. R. Forrest, J. Appl. Phys. 95, 1859 共2004兲. 11 S. A. Jenekhe and S. Yi, Appl. Phys. Lett. 77, 2635 共2000兲. 12 E. Kymakis, E. Koudoumas, and I. Franghiadakis, Sol. Energy Mater. Sol. Cells 90, 1705 共2006兲. 13 F. A. Castro, H. Benmansour, C. F. O. Graeff, F. Nuesch, E. Tutis, and R. Hany, Chem. Mater. 18, 5504 共2006兲. 14 M. Granstrom, K. Petritsch, A. C. Arias, A. Lux, M. R. Andersson, and R. H. Friend, Nature 共London兲 395, 257 共1998兲. 15 Y. Xia and R. H. Friend, Appl. Phys. Lett. 88, 163508 共2006兲. 16 K.-H. Kim, S.-Y. Huh, S.-m. Seo, and H. H. Lee, Appl. Phys. Lett. 92, 093307 共2008兲. 17 K.-H. Yim, Z. Zheng, Z. Liang, R. H. Friend, W. T. S. Huck, and J.-S. Kim, Adv. Funct. Mater. 18, 1012 共2008兲. 18 L. Chen, P. Degenaar, and D. D. C. Bradley, Adv. Mater. 共Weinheim, Ger.兲 20, 1679 共2008兲. 19 J. Xue and S. R. Forrest, J. Appl. Phys. 95, 1869 共2004兲. 20 J. Huang, P. F. Miller, J. S. Wilson, A. J. deMello, J. C. deMello, and D. D. C. Bradley, Adv. Funct. Mater. 15, 290 共2005兲. 21 L. D. Bozano, K. R. Carter, V. Y. Lee, R. D. Miller, R. DiPietro, and J. C. Scott, J. Appl. Phys. 94, 3061 共2003兲. 22 M. C. Gather, A. Kohnen, A. Falcou, H. Becker, and K. Meerholz, Adv. Funct. Mater. 17, 191 共2007兲. 23 S. H. Khong, S. Sivaramakrishnan, R. Q. Png, L. Y. Wong, P. J. Chia, L. L. Chua, and P. K. H. Ho, Adv. Funct. Mater. 17, 2490 共2007兲. 24 J.-S. Kim, R. H. Friend, I. Grizzi, and J. H. Burroughes, Appl. Phys. Lett. 87, 023506 共2005兲. 25 S. A. Choulis, V.-E. Choong, M. K. Mathai, and F. So, Appl. Phys. Lett. 87, 113503 共2005兲. 26 M. Redecker, D. D. C. Bradley, M. Inbasekaran, W. W. Wu, and E. P. Woo, Adv. Mater. 共Weinheim, Ger.兲 11, 241 共1999兲.

Downloaded 12 May 2009 to 131.175.136.39. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp