Maskless Patterning of Vapor-Deposited Photosensitive Film and its by ert634


									                                                                                                         SS10126 134 Total pages 5
Japanese Journal of Applied Physics 50 (2011) 04DK07                                                                      REGULAR PAPER
DOI: 10.1143/JJAP.50.04DK07

Maskless Patterning of Vapor-Deposited Photosensitive Film
and its Application to Organic Light-Emitting Diodes
Masakazu Muroyama, Wataru Saito, Seiji Yokokura, Kuniaki Tanaka, and Hiroaki Usui
Department of Organic and Polymer Materials Chemistry, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184-8588, Japan
Received September 21, 2010; accepted December 2, 2010; published online April 20, 2011

  Photosensitive thin films were prepared by the codeposition of a polymerizable monomer and a photoinitiator, which were patterned by UV
  irradiation followed by development in an organic solvent. This technique was applied for the preparation of a phosphorescent layer of an organic
  light emitting diode (OLED) by combining a vinylcarbazole monomer, a dopant of a vinyl derivative of a iridium complex, and a photoinitiator of a
  benzophenone derivative. An OLED with multiple emitting elements was prepared by repeating this patterning process. It was confirmed that the
  patterning does not cause harmful damage to the device characteristics. Moreover, the device characteristics can be improved after the photo
  patterning because the photopolymerization stabilizes the deposited films. This method is advantageous as a new technique for patterning
  polymer films without using a shadow mask for vapor deposition or a photoresist. # 2011 The Japan Society of Applied Physics

                                                                            reactions involved in these methods are limited only for the
1. Introduction                                                             hole transport layer (HTL) materials.
Organic electronic devices mostly consist of multilayers of                    By applying the concept of deposition polymerization, the
patterned thin films. Therefore, film formation and pattern-                  authors have proposed to eliminate the shadow masks for
ing are the most essential processes in device fabrication                  vapor deposition by associating the polymerization reaction
technology. As for the film formation, physical vapor                        with the vapor-deposited films. In this technique, a
deposition (PVD) provides a highly controllable method                      photosensitive film was prepared by coevaporating mono-
for preparing thin films of small molecules, and has been                    mers and photoinitiators, which was locally polymerized by
frequently employed for device fabrication. The authors                     UV irradiation, and then developed in an organic solvent.10)
have been investigating a physical vapor deposition                         Carbazole polymer thin films have been patterned to a line
polymerization method, by which polymeric thin films can                     width of 10 m by this method.11) This method not only
be produced by evaporating their monomers. This method is                   eliminates the necessity for the shadow masks or photo-
advantageous for constructing multilayered structures of                    resists but also enables the combination of the advantages of
polymeric thin films. The physical vapor deposition                          the vapor deposition technique and photolithography. Also,
polymerization was originally reported for the deposition                   this method can be applied for various organic compounds
of such polymers as polyimide1) and polyurea.2) Materials                   that have a vinyl or acryl group attached to the functional
for organic light-emitting diodes (OLEDs), including a metal                units. In this study, we investigated an application of this
complex polyurethane3) and a -conjugated polymer,4) have                   patterning technique for constructing OLEDs. A photosen-
also been deposited by this method. In addition, a variety                  sitive layer was prepared by coevaporation of polymerizable
of functional vinyl polymers can be deposited by radical                    monomers of a hole-transport material, a phosphorescent
polymerization assisted by thermal radiation,5) ultraviolet                 dopant, and a polymerization initiator to form an emitting
(UV) irradiation,6) and electron irradiation7) to construct                 layer (EML) on a HTL. Patterning of the EML was achieved
OLED devices. Increases in luminescence quantum effi-                         after the vapor deposition by applying the standard
ciency and device lifetime have been achieved by vapor                      photolithographic procedure. An electron-transport layer
deposition polymerization.                                                  (ETL), an electron injection layer (EIL), and a cathode
   As for the pattering, pattern formation by PVD has been                  were prepared by conventional vapor deposition on the
conveniently achieved by deposition through a shadow                        patterned EML. The process of depositing and photopattern-
mask. However, with increasing demand for fabricating                       ing the EML can be repeated to construct a multielement
large-scale devices, the shadow-mask technique has en-                      device. The objective of this study is to assess possible
countered such difficulties as mask alignment and contam-                     damage to the device characteristics caused during this
ination. Thermal expansion makes it extremely difficult to                    patterning process.
ensure alignment accuracy in micron scale on the entire
mask area. On the other hand, the conventional photolitho-                  2. Experimental Methods
graphy using a photoresist is not always applicable to the                  2.1 Photopatterning
patterning of organic thin films because the wet processes                   Photopatterning of the EML was achieved by preparing a
for coating, patterning, etching, and removing the photo-                   photosensitive film by coevaporating a carbazole monomer,
resist can damage the organic materials. With the purpose                   9H-carbazole-9-ethyl-methacrylate (CEMA), a vinyl deriva-
of overcoming these difficulties, thermal conversion and                      tive of phosphorescent dopant vinyl-tris(2-phenylpyridine)
photopatterning of poly(p-phenylene vinylene) (PPV) was                     iridium [vIr(ppy)3 ], and a photoinitiator, 4-dimethylamino
proposed.8) However, this method resorts to a wet-coating                   benzophenone (DABP). The chemical structures of these
process. As for the vapor deposition, photopatterning of an                 materials are shown in Fig. 1. CEMA, vIr(ppy)3 , and DABP
acetylene-containing triphenylamine was proposed for the                    were evaporated from stainless crucibles at temperatures of
hole transport layer (HTL) of OLEDs.9) However, this                        140, 310, and 150  C, respectively. The composition of the
method is applicable for a limited class of materials, i.e.,                coevaporated film was monitored by the deposition rate of
polyacetylene derivatives. Furthermore, the photochemical                   each material, which can be controlled by regulating the
                                                                    04DK07-1                          # 2011 The Japan Society of Applied Physics
Jpn. J. Appl. Phys. 50 (2011) 04DK07                                                                                               M. Muroyama et al.

                                                                                                1st photosensitive film

                                                     O       O                     (1)                           (a)

            N                        N                                                        UV exposure                      Device A
                      DvTPD                                                        (2)                                    electron transport layers
                                                                                                 developed pattern

                Ir   N                                                             (3)                           (b)
                                                                 CH 3
            N                                DABP            CH3                                                            Device B
                                                                                                          2nd photosensitive film
       Fig. 1. Chemical structures of the deposited materials.

                                                                                                           2nd UV exposure

evaporation temperature. All the deposition processes were                                                             Device C1       Device C2
achieved on room-temperature substrates in a vacuum of
about 1 Â 10À4 Pa. It was confirmed by our previous work                                               2nd pattern
that a satisfactory contrast of photopatterning can be
                                                                                   (6)                           (c)
obtained with a DABP concentration higher than 2 wt %
and with a UV power higher than 20 mW/cm2 for 1 min
exposure. From the viewpoint of the development contrast of
                                                                               Fig. 2. (Color online) Process flow chart for constructing devices.
CEMA and DABP film, the optimum concentration of
DABP was determined to be 5%.11) In accordance with this
result, the concentration of DABP was also fixed at 5% in
this work.                                                                   process flow chart for preparing these OLEDs. As a control
   The coevaporated film was exposed to UV light of                           device, an OLED was prepared without photopatterning the
60 mW/cm2 at 370 nm for 1 min through a photomask in the                     EML, as shown in Fig. 2(a) (device A). A single element
air. Polymerization of the film by UV irradiation has been                    device was prepared by photopatterning the EML as stated
confirmed by infrared (IR) absorption spectra measured by                     above through the procedure shown in Fig. 2(b) (device B).
reflection absorption for a film deposited on a gold                           A multiple element device was prepared as shown in
surface.10,11) The film was then immersed in tetrahydrofuran                  Fig. 2(c), where a second EML was deposited on the first
(THF) to develop a negative pattern of the UV exposure by                    EML after photopatterning the first EML. After photopat-
dissolving the nonpolymerized area. The surface morphol-                     terning the second EML, the ETL and cathode were
ogy of the films was examined using a scanning probe                          deposited to cover both EML patterns. Device C1 is
microscope (SPM; Keyence VN-8000).                                           constructed with the first EML that had undergone the
                                                                             deposition and patterning processes of the second EML,
2.2 OLED fabrication                                                         while device C2 was prepared on the HTL that had
OLEDs were prepared by using the photopatterned carbazole                    undergone development of the first EML. The devices were
thin film as the EML. The basic structure of an OLED is as                    operated in the air without encapsulation. The emission
follows. On an indium–tin oxide (ITO) surface, a 30-nm-                      spectra of the devices were measured with a charge-coupled
thick HTL of N,N,N0 -triphenyl-N0 -(4-vinylphenyl) biphenyl-                 device (CCD) array spectrometer (BWTEK BTC112E).
4,40 -diamine (DvTPD) was deposited, and then annealed in                    Since the devices were measured in air without encapsula-
vacuum at 150  C for 1 h to polymerize this layer. After the                tion, no assessment was made for device lifetime.
thermal treatment, the DvTPD layer formed a uniform and
stable thin film that is insoluble to organic solvents.12) On                 3. Results and Discussion
this surface, a 30-nm-thick EML was prepared by codeposit-                   3.1 Polymerization of EML
ing CEMA, vIr(ppy)3 , and DABP. The EML was photo-                           In order to gain an insight into the chemical nature of the
patterned to a size of 4 Â 3 mm2 by the aforementioned                       photopatterned EML, the emission spectra of the OLEDs
process. On the patterned EML, a 40-nm-thick ETL of                          shown in Fig. 2 were investigated. Figure 3 shows the
bathocuproin was deposited, followed by vapor deposition of                  emission spectra of devices A, B, C1, and C2, operated at a
a LiF EIL and a 100-nm-thick Al cathode.                                     luminance of about 100 cd/m2 . Devices B, C1, and C2 gave
   With the purpose of investigating the effect of the                        a green emission peaked at a wavelength of 510 nm, which is
photopatterning process on the device characteristics,                       characteristic of the triplet emission of Ir(ppy)3 dopant,13)
OLEDs were fabricated using an as-deposited (nonpat-                         while device A exhibited a yellowish emission peaked at
terned) photosensitive layer, the photopatterned EML, and                    545 nm. For comparison, photoluminescence spectra of
an EML that had undergone the photopatterning process                        dilute chloroform solutions of Ir(ppy)3 and vIr(ppy)3 are
twice to prepare a multielement device. Figure 2 shows the                   shown in Fig. 4. The solutions of Ir(ppy)3 and vIr(ppy)3
                                                                        04DK07-2                     # 2011 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 50 (2011) 04DK07                                                                                                                           M. Muroyama et al.

                                                                                                                                 1633           1296         937 895
                   Normalized intensity (relative unit)

                                                                                       Device A
                                                          2                                                                (a)

                                                                                       Device B

                                                                                       Device C1
                                                                                       Device C2
                                                          400       500         600           700
                                                                    Wavelength (nm)
Fig. 3. Emission spectra of OLEDs A, B, C1, and C2 shown in Fig. 2.
                                                                                                                 2000                   1500           1000                   500
                                                                                                                                           Wave number (cm-1)
       Normalized intensity (relative unit)

                                                                                                             Fig. 5. IR spectra of the coevaporated film of CEMA, vIr(ppy)3 , and
                                                                 Ir(ppy)3                                    DABP before (b) and after (c) the UV irradiation. CEMA monomer
                                                                                                             spectrum is also shown in (a).
                                                                                                             was measured by the KBr pellet method. The major feature
                                                                                                             of the IR spectra of the EML reflects the host material
                                                                                                             of CEMA. It is noteworthy that the peaks coming from
                                              0.5                                                            the vinyl group (C¼C stretching at 1633 cmÀ1 , C–H in-
                                                                                                             plane deformation at 1296 cmÀ1 , C–H out-of-plane defor-
                                                                                                             mation at 937 cmÀ1 , and CH2 out-of-plane deformation
                                                                                                             at 895 cmÀ1 ) almost disappeared after the UV irradiation,
                                                          0                                                  which suggests the occurrence of vinyl polymerization
                                                          400              500                    600        during the photopatterning process. It is known that a heavy
                                                                     Wavelength (nm)                         metal such as iridium can quench the radical polym-
                                                                                                             erization. However, the result in Fig. 5 indicates that the
Fig. 4. Photoluminescence spectra of chloroform solutions of Ir(ppy)3                                        small amount of vIr(ppy)3 dopant did not hinder the
and vIr(ppy)3 .
                                                                                                             polymerization of EML.
                                                                                                                Our previous work has shown that the UV exposure
                                                                                                             induces polymerization of the codeposited film.10) These
were excited at 286 and 291 nm, respectively. It is obvious                                                  results indicate that the polymerization of the coevaporated
that the vinyl derivative vIr(ppy)3 has emission at a longer                                                 films can be achieved by UV exposure. However, the IR
wavelength than that of the standard Ir(ppy)3 . This                                                         absorption does not give quantitative information of the
difference can be attributed to the extended -electron                                                       molecular weight, giving almost identical spectra irrespec-
conjugation of vIr(ppy)3 due to the vinyl group attached to                                                  tive of the degree of polymerization. On the other hand,
the Ir(ppy)3 unit. Therefore, the result shown in Fig. 3                                                     it has also been shown that the films were partly soluble
suggests that device A consists of a monomer EML that                                                        to THF even after the UV exposure. The film thickness
contains the vIr(ppy)3 dopant. On the other hand, devices B,                                                 decreased to about 45% of the initial thickness after
C1, and C2 have a polymer EML due to the photopolymer-                                                       immersion in THF for 1 min.11) This result suggests that
ization associated with the patterning process. The polymer-                                                 a considerable amount of small molecules still remains in
ization led to a break in the  conjugation of the vinyl group,                                              the film after the UV irradiation in the aforementioned
making the emission spectra the same as these of the                                                         condition. The IR spectra were also unable to determine the
nonconjugated Ir(ppy)3 instead of the vIr(ppy)3 monomer.                                                     amount of the initiator remaining in the film. In the current
A similar shift of emission spectra has been reported for                                                    system, the reaction proceeds with chain polymerization of
the current-induced polymerization and vapor-deposition                                                      the radical reaction. Therefore, the DABP molecules are
polymerization of EML using the vIr(ppy)3 dopant.14–18)                                                      expected to be fixed at the chain ends of the CEMA
   The polymerization during the photopatterning process                                                     polymers.19) Even if some of the DABP molecules remain
was confirmed by IR spectra of the EML single layer. The                                                      unreacted in the film, they would be extracted from the film
spectra were measured by the reflection-absorption mode                                                       together with other small molecules during the process of
for the EML deposited on a gold substrate. Figure 5 shows                                                    development. In any case, the DABP concentration of 5%
the IR spectra of the coevaporated film of CEMA, vIr(ppy)3 ,                                                  is a typical value of initiator concentration employed for
and DABP before (b) and after (c) the UV irradiation, in                                                     the common radical polymerization, and might not cause a
comparison with the CEMA monomer spectrum (c), which                                                         drawback inherent to the current method.
                                                                                                        04DK07-3                         # 2011 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 50 (2011) 04DK07                                                                                                  M. Muroyama et al.

                             500                                                                          500
                                       Device C2                                                                Device C2
                                       Device C1                                                                Device C1
                             400       Device B                                                           400   Device B
         Luminance (cd/m )

                                       Device A

                                                                                                                Device A

                                                                                     Luminance (cd/cm2)
                             300                                                                          300

                             200                                                                          200

                             100                                                                          100

                                   0              1                2                                       0
                                        Current density (mA/cm )                                            2   4       6      8     10      12
                                                                                                                       Voltage (V)
 Fig. 6. Current–luminance characteristics of OLEDs shown in Fig. 2.
                                                                             Fig. 7. Voltage–luminance characteristics of OLEDs shown in Fig. 2.

3.2 Luminescence characteristics
Figure 6 shows the current–luminance characteristics of the
devices described in Fig. 2. It is obvious that device A,
which had an as-deposited EML without photopatterning,
was inferior to other devices that had a photopatterned EML.
It is considered that this result reflects the fact that the EML
of device A consists of monomers while the photopatterned
devices have a polymerized EML, as suggested by the
emission spectra. The improvement of OLED characteristics
by polymerization has also been reported in connection with
vapor-deposition polymerization.14–18) It is notable that
despite the UV irradiation and rinsing in THF, the patterned
EML was not deteriorated compared with the as-deposited
   Furthermore, devices C1 and C2, which were prepared by
repeating the photopatterning process on the same HTL,
gave higher luminance than device B, which was formed
using the photopatterning process only once. At the moment,
the reason that led to this difference is not clear. However, it
is considered that the EMLs in devices C1 and C2 have been
polymerized more completely during the additional pattern-
ing process. On the other hand, device B endured a higher
driving current, giving a higher maximum luminance before
the device broke down. It is noteworthy that both devices C1
and C2 exhibited almost the same luminescence character-
                                                                            Fig. 8. (Color online) Differential surface topography by SPM
istics although the EML of device C1 underwent vapor                        observation of the DvTPD film before (a) and after (b) sonicating in acetone.
deposition and the patterning process of the second EML                     The scale bar indicates a length of 30 m.
that was used to prepare device C2. This result indicated that
the pattern formation process for the second EML did not
damage the first EML, and reflects the fact that the patterns                 inferior in voltage–luminance characteristics to device B.
are generated as stable polymer thin films. The quantum                      There is a possibility that the interfaces between the organic
efficiencies of devices A, B, C1, and C2 were 0.082, 0.16,                    layers were damaged by the repeated exposure to the organic
0.49, and 0.60 cd/A, respectively. These results indicate that              solvent used for the pattern development.
this patterning process does not cause harmful damage of                       Although the HTL of the DvTPD polymer was practically
electrical characteristics but improves the device perfor-                  insoluble to organic solvents, a slight degradation of surface
mance by polymerization.                                                    morphology was found after exposing the film to an organic
   Figure 7 shows the voltage–luminance characteristics of                  solvent. Figure 8 shows the differential surface topography
the devices described in Fig. 2. In good correspondence with                obtained by SPM observation of the DvTPD film before (a)
the result of current–luminance characteristics, the photo-                 and after sonicating in acetone. The DvTPD polymer film
patterning resulted in a substantial increase of device current             had excellent uniformity of surface morphology. However,
and a decrease of turn-on voltage compared to device A that                 the arithmetic average surface roughness Ra increased from
has an as-deposited EML. However, devices C1 and C2 were                    0.9 to 1.1 nm after sonication in the organic solvent. It was
                                                                       04DK07-4                                 # 2011 The Japan Society of Applied Physics

Jpn. J. Appl. Phys. 50 (2011) 04DK07                                                                                                      M. Muroyama et al.

                                                                                  considered that the polymerization that is associated with the
                                                                                  photopatterning helps to improve the device characteristics.
                                                                                  Furthermore, the photopatterning process can be applied
                                                                                  repeatedly on the same substrate, enabling the fabrication of
                                                                                  multielement devices. The multiple patterning processes did
                                                                                  not degrade the current–luminance characteristics, indicating
                                                                                  that the photopatterning process does not damage the
                                                                                  luminescence quantum efficiency. On the other hand, the
                                                                                  luminescence threshold voltage suffered some increase
                                                                                  caused by the multiple patterning, suggesting a degradation
                                                                                  of the interface caused by the repeated exposure of the film
                                                                                  surface to the organic solvent. Nevertheless, the results of
                                                                                  this work suggest the possibility of fabricating complicated
                                                                                  film patterns or multicolor devices by sequentially repeating
                                                                                  the newly proposed photopatterning process.
                                                                                  This research was funded partly by the Seeds Validation
                                                                                  Program 04-B16 from the Japan Science and Technology
                                                                                  Agency, and partly by a Grant-in-Aid for Scientific
                                                                                  Research, No. 20360347, from the Japan Society for the
                                                                                  Promotion of Science.

Fig. 9. (Color online) Differential surface topography by SPM
observation of the EML before (a) and after (b) the pattern development by         1) J. R. Salem, F. O. Sequeda, J. Duran, W. Y. Lee, and R. M. Yang: J. Vac.
rinsing in THF. Films are covering the left half of the images, while the right        Sci. Technol. A 4 (1986) 369.
half of the images represents the substrate surface. The scale bar indicates a     2) Y. Takahashi, S. Ukishima, M. Iijima, and E. Fukada: J. Appl. Phys. 70
length of 30 m.                                                                       (1991) 6983.
                                                                                   3) X. Wang, K. Ogino, K. Tanaka, and H. Usui: IEICE Trans. Electron. E87-
                                                                                       C (2004) 2122.
                                                                                   4) X. Wang, K. Ogino, K. Tanaka, and H. Usui: Thin Solid Films 438 (2003)
also difficult to keep the EML surface intact during the                                 75.
solvent process. Figure 9 shows the differential SPM images                         5) M. Tamada, H. Omichi, and N. Okui: Thin Solid Films 251 (1994) 36.
at the edge of the EML before (a) and after (b) the pattern                        6) M. Tamada, H. Koshikawa, F. Hosoi, T. Suwa, H. Usui, A. Kosaka, and H.
                                                                                       Sato: Polymer 40 (1999) 3061.
development by rinsing in THF. By rinsing in THF, Ra                               7) H. Usui: IEICE Trans. Electron. E83-C (2000) 1128.
increased from 1.4 to 2.1 nm. These results imply a                                8) M. Prelipceanu, O. Perlipceanu, O. Tudose, L. Leontie, B. Grimm, and S.
possibility that the solvent can damage the interfaces and                             Schrader: Mater. Sci. Semicond. Process. 10 (2007) 77.
degrade the charge injection characteristic of devices C1                          9) C. Lee, Y. Kang, S. Jung, J. Kim, and J. Lee: Opt. Mater. 21 (2002) 337.
                                                                                  10) M. Muroyama, I. Saito, S. Yokokura, K. Tanaka, and H. Usui: Jpn. J. Appl.
and C2.                                                                                Phys. 48 (2009) 04C163.
                                                                                  11) M. Muroyama, S. Yokokura, K. Tanaka, and H. Usui: Jpn. J. Appl. Phys.
4. Conclusions                                                                         49 (2010) 01AE03.
The new patterning method using the vapor-deposited                               12) M. Muroyama, A. Tajiri, K. Ichida, S. Yokokura, K. Tanaka, and H. Usui:
                                                                                       to be published in IEICE Trans. Electron. E94-C (2011).
photosensitive film was successfully applied for preparing                         13) K. A. King, P. J. Spellane, and R. J. Watts: J. Am. Chem. Soc. 107 (1985)
OLEDs. This method consists of conventional vapor                                      1431.
deposition and photoinduced polymerization processes,                             14) A. Kawakami, H. Bekku, E. Otsuki, R. Wada, H. Kita, H. Taka, H. Sato,
                                                                                       and H. Usui: Jpn. J. Appl. Phys. 47 (2008) 1284.
eliminating the use of a shadow mask or photoresist. The
                                                                                  15) E. Otsuki, A. Kawakami, H. Kita, H. Taka, and H. Usui: IEICE Tech. Rep.
patterns are obtained as stable polymer thin films, and can be                          108 (2008) EMD2008-11 [in Japanese].
easily incorporated in the fabrication process of OLEDs. It                       16) A. Kawakami, E. Otsuki, M. Fujieda, H. Kita, H. Taka, H. Sato, and H.
was found that the photopatterning process did not damage                              Usui: Jpn. J. Appl. Phys. 47 (2008) 1279.
                                                                                  17) H. Usui: Thin Solid Films 365 (2000) 22.
the device characteristics. Moreover, the photopatterned                          18) H. Usui: Proc. SPIE 7404 (2009) 74040E.
device showed higher performance than the device that uses                        19) M. P. Stevens: Polymer Chemistry (Oxford Universty Press, New York,
a nonpatterned (i.e., as-deposited) photosensitive layer. It is                        1990) p. 199.

                                                                           04DK07-5                         # 2011 The Japan Society of Applied Physics


To top