Open Access
Add to BibSonomyAbstract
Using molecular diffusion as an approach to introduce organic luminescent dopants for making organic light emitting devices (OLEDs) of different colors on one substrate has the potential to overcome the yield and resolution limitations of the current OLED display technology. In this work, diffusion barriers made of MoO3 and a hole transport material mixture are introduced. The barriers effectively confine the diffusion of the dopants to only the desired depths. With the use of these barriers, OLEDs with highly controlled doping concentrations and performance are fabricated. The barriers thus allow utilizing simple diffusion methods for RGB patterning in OLEDs.
© 2015 Optical Society of America
1. Introduction
Organic Light Emitting Devices (OLEDs) have undergone tremendous progress since the first report of a high-efficiency device by Tang and VanSlyke in 1987 [1]. They are now being used in a wide range of consumer flat panel displays. The interest in OLEDs is motivated by a number of advantages they have, which include their higher contrast and lower power consumption relative to liquid crystal displays, and their compatibility with mechanically flexible products [2,3]. However, while the performance of OLEDs such as efficiency and stability has now met requirements for display products, OLED display manufacturing processes are still relatively less mature. In particular, there is still a need to establish more reliable and scalable RGB color patterning techniques for OLED display fabrication. Currently, RGB color patterning is done by fabricating arrays of red, green and blue OLEDs side-by-side on the same back-plane substrate using thermal deposition, utilizing pre-patterned shadow masks for the sequential deposition of red, green and blue luminescent materials [3,4]. The masks are typically made of thin metal sheets in what is widely known as the Fine Metal Mask (FMM) technology. However, like other shadow mask technologies, FMM has several inherent limitations, such as; mask deformation, difficulties in mask-to-substrate overlay alignment, and difficulties in making masks with micrometer level dimensional accuracy. These issues put a limit on the patterning accuracy, and lead to low manufacturing yields, high fabrication costs, and low display quality. In this context, the development of “mask-less” RGB color patterning approaches, in which the red, green and blue OLEDs can be fabricated on the same substrate without the use of shadow masks, can provide significant advantages. Examples of such mask-less color patterning approaches include contact printing techniques utilizing pre-patterned stamps and selective transfer (or sublimation) of organic layers from a donor substrate to the OLED substrate using lasers or other thermal transfer techniques [5–9]. Another mask-less technique that has emerged recently depends on diffusion as an approach to introduce luminescent material dopants into the organic host material layer of an OLED using a donor substrate placed in physical contact with the OLED backplane substrate [10–15]. In this approach, color patterning is achieved by allowing the diffusion of the material from the donor to the acceptor substrates, and hence the doping, to occur only in certain locations. Several approaches have been proposed for achieving such selective doping. These include the use of pre-patterned stamps [13,14], screen printing [11,12] or Joule heating using the anode pads, to induce localized heating, and thus diffusion, at only the desired locations [10,15]. Among these techniques the last one is particularly advantageous, as it does not require any overlay alignment steps. Although the technique was initially developed for RGB patterning of polymer OLEDs, we recently demonstrated its applicability to small molecule OLEDs which is the main-stream technology for OLED displays [16]. While we showed that the technique is effective for color patterning, one can expect the ability to control the doping levels through diffusion to be difficult which could limit device performance. For example, the diffusion of the luminescent dopants may extend beyond the light-emitting layer (EML) into the hole-transport layer (HTL). The presence of dopant molecules into the HTL may make them act as charge traps, which could result in higher driving voltage. Additionally, the dopant concentration is likely to be non-uniform across the EML, with higher concentrations near the surface. Such non-uniformity and high dopant concentrations near the surface could decrease device efficiency due to concentration quenching effects.
In order to overcome the aforementioned limitations, we introduce here semiconducting diffusion barriers, made of a mixture of an organic and inorganic material that enable achieving controlled and uniform concentrations of luminescent dopants in the EML via diffusion. The use of these diffusion barriers allows controlling the diffusion depth and achieving uniform concentration by means of blocking undesirable dopant diffusion. To the best of our knowledge, this is the first time that diffusion barriers are used in the field of organic semiconductors. Figure 1 illustrates this idea by comparing diffusion-based doping of an organic layer in a multi-layer stack with and without the diffusion barrier. The barrier effectively prevents the diffusion of the dopant molecules to the underlying organic layers, and limits it only to the top layer and the desired depth [step (b-ii) of Fig. 1]. By limiting the diffusion depth, it also allows obtaining a uniform dopant concentration across the layer [step (b-iv) of Fig. 1]. For diffusion barriers we utilize molybdenum trioxide (MoO3) mixed with the hole transport material (HTM) used in the HTL. Due to the relatively high density and small molecular size of MoO3, the HTM:MoO3 mixture can be expected to have higher density relative to the neat HTM, and thus, to block the diffusion of the dopant material. At the same time, the relatively high conductivity of HTM:MoO3 mixtures, caused by the formation of a charge transfer (CT) complex, prevents disruption of charge transport across the device [17,18].
Fig. 1 Scheme for the diffusion-based doping method without (a), and with (b) the diffusion barrier.
In this work, we first test the effectiveness of the HTM:MoO3 film in substantially reducing the inter-diffusion of luminescent materials between two layers on opposite sides of it. In this case, changes in photoluminescence (PL) are used to detect any inter-mixing that may result from such inter-diffusion between the two layers. The tests are done for two different scenarios in which either heating (at 100 °C) or exposure to solvents (Toluene) is used to activate the diffusion process. We then demonstrate the effectiveness of this approach for producing OLEDs with highly controlled dopant concentrations and performance.
2. Results and discussion
2.1 Assessing the effectiveness of MoO3 and a hole transport material mixtures in blocking inter-diffusion of organic materials
We first test the effectiveness of using a HTM:MoO3 film in reducing inter-diffusion of luminescent materials between two layers located on opposite sides of it. For this purpose we use organic stacks in which the HTM:MoO3 film is sandwiched between two organic layers containing different luminescent materials, and utilize PL measurements for detecting any inter-mixing that may occur as a result of inter-diffusion. N,N'-Bis (naphthalen-1-yl)-N,N'-bis(phenyl) benzidine (NPB), 2,3,6,7-tetrahydro-1,1,7,7,-tetramethyl-1H,5H,11H-10-(2-benzothiazolyl) quinolizino- [9,9a,1gh] coumarin (C545T), and 4-(dicyanomethylene)-2-t-butyl-6-(1,1,7,7-tetramethyljulolidyl-9-enyl)-4H-pyran (DCJTB) are used as a host, green emitter, and red emitter material, respectively. The organic layer stack consists of NPB:C545T(5%,40nm)/NPB:MoO3(50%,5nm)/DCJTB(3nm) [Fig. 2(d)]. Since DCJTB has a smaller band gap than that of C545T, any intermixing between the two materials due to diffusion would lead to quenching of C545T luminescence by DCJTB as a result of energy transfer. As such, any changes in PL spectra with time would correlate with the extent of intermixing between the two layers as a result of diffusion across the HTM:MoO3 film. For comparison, we also fabricate stacks of the same structure except that the NPB:MoO3 layer is replaced with a 5nm thick neat layer of NPB to be used as reference (i.e. represents the case of no diffusion barrier). The complete structure of the reference stacks is therefore NPB:C545T(5%,40nm)/NPB(5nm)/DCJTB(3nm) [Fig. 2(a)]. All stacks are fabricated by the sequential deposition of the organic layers using standard thermal vacuum deposition on glass substrates. After fabrication, the substrates (with the stacks) are placed on a hot plate at 100 °C for driving molecular diffusion.
Fig. 2 (a) and (d) the structure of the organic layer stacks without and with the diffusion barrier film, respectively; (b) and (e) spectra collected every 10 minutes from the organic stacks; and (c) and (f) PL images of the stacks before and after heating.
Figures 2(b) and 2(e) show PL spectra collected every 10 minutes from the organic stacks without and with the diffusion barrier, respectively, under 360 nm excitation. As can be seen from the figures, the initial (at t = 0min, i.e. before heating) spectra collected from the two stacks correspond almost entirely to C545T emission, with no significant emission from DCJTB. The absence of DCJTB emission can be attributed to concentration quenching effects, which make luminescence from a neat layer of DCJTB very weak. It should be also noted that there is no emission from the NPB:MoO3 layer due to significant quenching effects by the CT complex formed between MoO3 and NPB. Upon heating, the PL spectrum of the organic stack without the diffusion barrier starts exhibiting a red shift, and after 50 minutes, the emission becomes dominated by 605 nm DCJTB emission [Fig. 2(b)]. This spectral change clearly points to the diffusion of DCJTB molecules into the NPB:C545T layer, resulting in quenching of C545T and efficient luminescence from DCJTB that now dopes the layer. In contrast, the spectrum from the stack with the diffusion barrier exhibits very little change and remains dominated by ~530 nm emission from C545T even after heating for 50 minutes [Fig. 2(e)]. The differences between the PL changes in the two samples are also evident in the PL images taken under UV excitation [Figs. 2(c) and 2(f)]. These results clearly demonstrate that the NPB:MoO3 film is indeed capable of preventing DCJTB molecules from diffusing to the opposite side. Such property of the HTM:MoO3 mixture layer can be attributed to a decrease in the effective diffusion coefficient due to the addition of MoO3 in the HTM film. When HTM and MoO3 are co-deposited, MoO3 molecules form spherical clusters in the HTM film [19]. Since MoO3 has a higher density than NPB, the presence of these high density clusters impedes the diffusion of other materials, in this case the dopant, across the film.
Furthermore, the same experiment is performed with the same organic layer structure but with varying MoO3 concentrations in the NPB:MoO3 layer (from 0% to 24%, by volume). Figures 3(a) and 3(b) present the PL spectra collected from the organic layer stacks before and after heating for 50 minutes and the corresponding difference in peak wavelength between before and after heating versus the MoO3 concentration, respectively. As can be seen, the redshift in the spectra after heating becomes increasingly less pronounced as the MoO3 concentration in the layer increases. Since a redshift in DCJTB reflects an increase in its concentration (the redshift is caused by increased molecular aggregation), the trend indicates that increasing MoO3 concentration in the barrier film results in a decrease in the number of DCJTB molecules that can penetrate through it to the NPB:C545T layer. This result clearly demonstrates that the addition of MoO3 to HTM indeed lowers the diffusivity of the luminescent material in the film.
Fig. 3 (a) Spectra collected before (dotted traces) and after (solid lines) heating for 50minutes for various MoO3 concentrations of the NPB:MoO3. The PL spectra before heating are almost identical. (b) The corresponding difference in peak wavelength between before after heating vs MoO3 concentration.
It has been demonstrated that molecular diffusion can be significantly enhanced by exposure to solvent vapors; an approach that can be utilized for accelerating diffusion and thus reducing processing times [14,20,21]. To investigate if the diffusion barrier can similarly block diffusion in case of exposure to solvents, we perform the same experiment on another set of organic stacks that in this case are exposed to toluene vapor for 2 minutes instead of heating to 100 °C. Figures 3(e) and 4(b) show the PL spectra from the stacks without and with the NPB:MoO3 barrier film respectively for this experiment. The same phenomenon is observed here also, where again the PL spectrum of the organic layer stack without the diffusion barrier exhibits a change from green to red emission upon solvent exposure pointing to the diffusion of the DCJTB into the NPB:C545T layer whereas the PL spectrum from the stack containing the diffusion barrier layer remains unchanged. The gradual change in PL with time is shown in Visualization 1. These results confirm that the HTM:MoO3 film functions as an effective diffusion barrier that can be used with both heat-assisted and solvent vapor-assisted diffusion-based doping processes.
Fig. 4 (a) and (d) the structure of the organic layer stacks without and with the diffusion barrier film, respectively; (b) and (e) spectra collected from the organic stacks before and after solvent vapor exposure; and (e) and (f) PL images of the stacks before and after solvent vapor exposure. The transition is shown in Visualization 1.
2.2 Demonstrating the effectiveness of the use of the diffusion barrier for producing OLEDs with highly controlled dopant concentrations and performance
Next, we demonstrate the use of the HTM:MoO3 diffusion barriers in OLEDs for obtaining controlled doping levels when luminescent emitter materials are introduced in a host via diffusion. In this experiment, a luminescent emitter is introduced into a host material layer via diffusion by contact with a donor substrate. By utilizing the diffusion barrier, both the diffusion depth and dopant concentration are controlled. Device fabrication steps including the diffusion-based doping step are depicted in Fig. 5. First, a ~3 nm thick MoO3 hole injection layer (HIL), a 60-x nm thick NPB doped with 50% MoO3 layer, and a x nm thick NPB neat layer are vacuum-deposited on an ITO coated glass substrate (acceptor substrate). In one device x is 10nm whereas in the other device x is 15nm.The MoO3-doped NPB layer in this device functions as both a HTL and semiconducting diffusion barrier whereas the neat NPB layer functions as host to be doped by DCJTB via diffusion. The 50% doping concentration of the NPB:MoO3 mixture layer is selected since it is the concentration that gives the highest conductivity [22]. The acceptor substrate is then placed in physical contact with a glass “donor substrate” pre-coated with a ~20 nm thick layer of DCJTB by holding them together using a 5mm thick neodymium magnet plate (remanent magnetization of 1.22-1.28 T) and a 2mm thick steel plate. Still in contact, the substrates are placed on a hot plate and heated for 100 minutes at 100 °C in a dry nitrogen environment in order to drive the diffusion of DCJTB from the donor substrate into the NPB layer on the acceptor substrate. After heating, the donor substrate is removed, the acceptor substrate is put back in the vacuum system, and a ~40 nm thick tris(8-hydroxyquinolinato) aluminum (Alq3) electron transport layer (ETL) followed by a ~0.5 nm thick LiF electron injection layer (EIL) and a ~70 nm aluminum cathode are deposited to complete the device. Finally, in order to drive the DCJTB deeper into the NPB layer and achieve a more uniform dopant concentration, the completed devices are heated to 100 °C for about an hour. During that time the device electroluminescence (EL) spectrum and current efficiency are measured every few minutes. In addition to these devices, which are referred to here as “diffused-dopant” devices, we also fabricate two control devices with the same structure: ITO/MoO3(3nm)/NPB:MoO3(60-xnm)/NPB:DCJTB(xnm,y%)/Alq3(40nm)/LiF(0.5nm)/Al (70nm). However, in this case the NPB:DCJTB layers are made by co-deposition using the standard thermal evaporation approach (i.e. not via diffusion). In one device x is 10nm and y is 6% whereas in the other device x is 15nm and y is 4%. These x and y combinations are selected so that the amount of DCJTB is roughly the same in the two devices. The amount of DCJTB is selected such that it would correspond to its amount in the corresponding “diffused-dopant” devices for the 100 minute contact duration used above. This DCJTB amount was estimated based on our previous observations that diffusion of molecules from the donor to an acceptor substrates proceeds at a constant rate during at least the initial 100 minutes [16]. Given this linear correlation, the amount of DCJTB that gets diffused from the donor substrate into the host during any given contact duration (in this case 100 minutes) can be accurately estimated.
Fig. 5 A schematic diagram illustrating the steps of the procedure followed for the diffused-dopant OLEDs. (i)-(iii) illustrates the steps of introducing the dopant via diffusion and (iv) illustrates the subsequent heating step used for making the DCJTB concentration uniform across the layer.
Figure 6 shows the EL spectra from the two “diffused-dopant” devices collected periodically while at 100 °C. As can be seen, the EL spectra show a significant blue-shift with time. This spectral shift reflects a decrease in DCJTB concentration that occurs as a result of its diffusion from the surface deeper into the NPB layer, and thus becomes more uniformly mixed with the NPB molecules. (It should be noted that the diffusion of DCJTB into the adjacent Alq3 layer is generally limited because of the high glass transition temperature of Alq3 (~176 °C) which well exceeds the temperatures used here [23]). As DCJTB molecules spread across the NPB layer, their aggregation gradually diminishes, resulting in the observed blue-shift in the spectra (an examination of the un-normalized EL spectra (not shown here) shows that the DCJTB luminescence intensity also increases with time as expected, consistent with an associated decrease in concentration quenching effects). The spectral shift slows down with time and stops almost completely after about 40 minutes at this temperature. This points to the ending of the DCJTB net mass transfer and redistribution process, signaling that a uniform concentration level of DCJTB across the entire NPB layer has been reached. The same trend can be observed in the current efficiency trend of the “diffused-dopant” devices, where again we see the efficiency increases initially but then the trend plateaus almost completely after ~40 minutes [Fig. 7]. Clearly, the efficiency trend follows the change in DCJTB concentration. Quite remarkably, both the final EL spectrum and efficiency of each of the two dopant-diffused devises are comparable to those of the corresponding control devices. This shows that the DCJTB concentration in the “diffused-dopant” devices is very similar to that in the corresponding control devices, indicating that our earlier estimation of the amount of DCJTB that gets diffused into the NPB in a given period of contact time was accurate. This points to the fact that highly controlled doping levels can indeed be achieved when using this doping method. It also shows that the inclusion of the NPB:MoO3 diffusion barrier indeed limits the diffusion of the DCJTB to the top 10 or 15nm NPB host layer and makes it possible to use diffusion followed by a simple heating step to reach uniform and predictable dopant concentration levels in OLEDs.
Fig. 6 EL spectra collected every few minutes from the “diffused-dopant” devices during the heating step [Fig. 5(iv)] for the device with the 15nm host NPB layer (a); and the device with the 10nm host NPB layer (b). The dotted lines represent the EL spectra of the corresponding control devices.
Fig. 7 Current efficiencies collected every few minutes from the “diffused-dopant” devices during the heating step [Fig. 5(iv)] for the device with the 15nm host NPB layer (the blue solid line); and the device with the 10nm host NPB layer (the red solid line). The dotted lines represent the current efficiency of the corresponding control devices.
3. Conclusion
In conclusion, we introduce here the use of MoO3 and hole transport material mixtures as diffusion barriers in OLEDs. The barriers are effective in blocking undesirable inter-diffusion of organic materials between device layers even under high temperatures (100 °C) or solvent vapor-exposure conditions. The use of these barriers allows controlling the diffusion depth and achieving uniform concentration of organic luminescent dopants in OLEDs when using diffusion as a color patterning approach for fabricating OLED displays. With the use of these barriers, OLEDs with highly controlled doping concentrations and performance are fabricated. The barriers thus open up opportunities for utilizing simple diffusion methods for mask-less RGB patterning in OLEDs.
Acknowledgments
Partial support to this work by the Natural Sciences and Engineering Research Council of Canada (NSERC) is gratefully acknowledged.
References and links
1. C. W. Tang and S. A. VanSlyke, “Organic electroluminescent diodes,” Appl. Phys. Lett. 51(12), 913–915 (1987). [CrossRef]
2. S. Forrest, P. Burrows, and M. Thompson, “The dawn of organic electronics,” IEEE Spectr. 37(8), 29–34 (2000). [CrossRef]
3. B. Geffroy, P. Le Roy, and C. Prat, “Organic light-emitting diode (OLED) technology: materials, devices and display technologies,” Polym. Int. 55(6), 572–582 (2006). [CrossRef]
4. P. E. Burrows, G. Gu, V. Bulovic, Z. Shen, S. R. Forrest, and M. E. Thompson, “Achieving full-color organic light-emitting devices for lightweight, flat-panel displays,” IEEE Trans. Electron. Dev. 44(8), 1188–1203 (1997). [CrossRef]
5. H. Jin and J. C. Sturm, “Super-high resolution transfer printing for full-color OLED display patterning,” J. Soc. Inf. Disp. 18(2), 141–145 (2010). [CrossRef]
6. J. Choi, K.-H. Kim, S.-J. Choi, and H. H. Lee, “Whole device printing for full color displays with organic light emitting diodes,” Nanotechnology 17(9), 2246–2249 (2006). [CrossRef]
7. M. B. Wolk, J. Baetzold, E. Bellmann, T. R. Hoffend Jr, S. Lamansky, Y. Li, R. R. Roberts, V. Savvateev, J. S. Staral, and W. A. Tolbert, “Laser thermal patterning of OLED materials,” Proc. SPIE 5519, 12–23 (2004). [CrossRef]
8. S. Lamansky, T. R. Hoffend Jr, H. Le, V. Jones, M. B. Wolk, and W. A. Tolbert, “Laser induced thermal imaging of vacuum-coated OLED materials,” Proc. SPIE 5937, 593702 (2005). [CrossRef]
9. S. H. Ko, H. Pan, S. G. Ryu, N. Misra, C. P. Grigoropoulos, and H. K. Park, “Nanomaterial enabled laser transfer for organic light emitting material direct writing,” Appl. Phys. Lett. 93(15), 151110 (2008). [CrossRef]
10. K. Tada and M. Onoda, “Three-color polymer light-emitting devices patterned by maskless dye diffusion onto prepatterned electrode,” Jpn. J. Appl. Phys. 38(2), 1143–1145 (1999). [CrossRef]
11. F. Pschenitzka and J. C. Sturm, “Three-color organic light-emitting diodes patterned by masked dye diffusion,” Appl. Phys. Lett. 74(13), 1913–1915 (1999). [CrossRef]
12. C. C. Wu, C. C. Yang, H. H. Chang, C. W. Chen, and C. C. Lee, “Finite-source dye-diffusion thermal transfer for doping and color integration of organic light-emitting devices,” Appl. Phys. Lett. 77(6), 794–796 (2000). [CrossRef]
13. A. Nakamura, T. Tada, M. Mizukami, S. Hirose, and S. Yagyu, “Three-color polymer light-emitting diodes by stamped dye diffusion,” Appl. Phys. Lett. 80(12), 2189–2191 (2002). [CrossRef]
14. K. Long, F. Pschenitzka, M. H. Lu, and J. C. Sturm, “Full-color OLEDs integrated by dry dye printing,” IEEE Trans. Electron. Dev. 53(9), 2250–2258 (2006). [CrossRef]
15. K. Tada and M. Onoda, “Color tuning of poly(N -vinylcarbazole)-based light-emitting devices through maskless dye-diffusion technique using phosphorescent dyes,” Jpn. J. Appl. Phys. 47(2), 1290–1292 (2008). [CrossRef]
16. Y. Kajiyama, K. Kajiyama, and H. Aziz, “Maskless RGB color patterning of vacuum-deposited small molecule OLED displays by diffusion of luminescent dopant molecules,” Opt. Express 23(13), 16650–16661 (2015). [CrossRef] [PubMed]
17. W.-J. Shin, J.-Y. Lee, J. C. Kim, T.-H. Yoon, T.-S. Kim, and O.-K. Song, “Bulk and interface properties of molybdenum trioxide-doped hole transporting layer in organic light-emitting diodes,” Org. Electron. 9(3), 333–338 (2008). [CrossRef]
18. C.-H. Gao, X.-Z. Zhu, L. Zhang, D.-Y. Zhou, Z.-K. Wang, and L.-S. Liao, “Comparative studies on the inorganic and organic p-type dopants in organic light-emitting diodes with enhanced hole injection,” Appl. Phys. Lett. 102(15), 153301 (2013). [CrossRef]
19. J.-H. Lee, H.-M. Kim, K.-B. Kim, and J.-J. Kim, “Origin of charge generation efficiency of metal oxide p-dopants in organic semiconductors,” Org. Electron. 12(6), 950–954 (2011). [CrossRef]
20. F. Pschenitzka and J. C. Sturm, “Solvent-enhanced dye diffusion in polymer thin films for color tuning of organic light-emitting diodes,” Appl. Phys. Lett. 78(17), 2584–2586 (2001). [CrossRef]
21. T. Graves-Abe, F. Pschenitzka, H. Z. Jin, B. Bollman, J. C. Sturm, and R. A. Register, “Solvent-enhanced dye diffusion in polymer thin films for polymer light-emitting diode application,” J. Appl. Phys. 96(12), 7154–7163 (2004). [CrossRef]
22. B. Tian, D. Ban, and H. Aziz, “Enhanced bulk conductivity and bipolar transport in mixtures of MoOx and organic hole transport materials,” Thin Solid Films 536, 202–205 (2013). [CrossRef]
23. S. Ninomiya, K. Ichiki, H. Yamada, Y. Nakata, T. Seki, T. Aoki, and J. Matsuo, “Molecular depth profiling of multilayer structures of organic semiconductor materials by secondary ion mass spectrometry with large argon cluster ion beams,” Rapid Commun. Mass Spectrom. 23(20), 3264–3268 (2009). [CrossRef] [PubMed]
