Spectroscopic ellipsometry of metal phthalocyanine thin films

Spectroscopic ellipsometry of metal
phthalocyanine thin films
Aleksandra B. Djurisˇic', Chung Yin Kwong, Tsz Wai Lau, Zheng Tong Liu,Hoi Sing Kwok, Lillian Sze Man Lam, and Wai Kin Chan Optical functions of cobalt phthalocyanine, nickel phthalocyanine ͑NiPc͒, and iron phthalocyanine ͑FePc͒have been determined by use of spectroscopic ellipsometry in the spectral range 1.55- 4.1 eV ͑300 - 800nm͒. The samples were prepared by evaporation onto glass and silicon substrates. The optical func-tions were determined by point-to-point fit.
Absorption spectra were also measured.
refraction data for NiPc and FePc are reported for the first time to our knowledge.
the experimental spectra was obtained for all three materials.
Introduction
CuPc, other metal Pcs have not been extensively Phthalocyanines are porphyrin derivatives that are characterized by high degrees of symmetry, planar- ͑CoPc͒ by spectroscopic ellipsometry has been report- Absorption spectra of various Pc films,15-17 as ͑Pcs͒ have attracted much attention for applications well as vapor phases and solutions in different sol- in organic optoelectronic devices such as organic solar vents,18,19 have also been reported.
cells,1-3 in organic field effect transistors,4 in organic The growth and morphology of the various Pc films light-emitting diodes in which Pcs can be used as a hole transport layers5,6 or emitting layers7; and in gas ist in several crystalline polymorphs.
polymorphs are metastable ␣ and stable ␤ phases.
tronic devices, the optical function of each layer The main differences between polymorphs lie in the tilt angle of the molecules within the columns ͑stacks usually described in terms of the complex dielectric of molecules with molecular planes parallel to one function ε͑␻͒ ϭ ε ͑␻͒ ϩ iε ͑␻͒ or the complex index of another͒ and the mutual arrangement of the col- refraction N͑␻͒ ϭ n͑␻͒ ϩ ik͑␻͒. Copper phthalocy- Transformation from phase ␣ to phase ␤ by anine ͑CuPc͒ has been the most widely studied metal exposure to solvent vapors25,29,30 or by annealing or Dielectric functions and indices of refraction of deposition at higher temperature26-28,31,32 has been CuPc were studied experimentally by use of reflec- The absorption spectra of different poly- tance and transmission measurements,10 absorption morphs of the same Pc compound differ significant- coefficient measurement with Kramers-Kronig anal- ysis,11 and spectroscopic ellipsometry.12,13 influences the preferred orientation in Pc layers.
The substrate also seems to play a role in the orien-tation of the deposited films.21-24,26 A. B. Djurisˇic' ͑dalek@eee.hku.hk͒, C. Y. Kwong, T. W. Lau, L. S. M. Lam, and W. K. Chan are with the University of Hong isomorphism with one another23 and that deposition Kong, Pokfulam Road, Hong Kong, China.
on mica at temperatures above 400 K results in ori- C. Y. Kwong are with the Department of Electrical Engineering Wachtel et al.24 found that the sta- and L. S. M. Lam and W. K. Chan are with the Department of bility of the films may be substrate dependent and Chemistry; A. B. Djurisˇic' is also with the Department of Physics.
that lithium Pc films deposited on glass substrates at Z. T. Liu and H. S. Kwok are with the Department of Electrical room temperature were stable over several months, Engineering, Hong Kong University of Science and Technology, whereas those on NaCl substrates showed marked changes when they were exposed to ambient condi- Received 22 May 2003; revised manuscript received 8 July 2003.
0003-6935͞03͞316382-06$15.00͞0 epitaxial growth of thin fluorine bridged aluminum APPLIED OPTICS ͞ Vol. 42, No. 31 ͞ 1 November 2003 Pc polymer by molecular-beam epitaxy was obtained evaporator to minimize formation of another oxide on alkali halide substrates, whereas epitaxial growth Microscope slide glass and quartz substrates were cleaned in an ultrasonic bath for 30 min, rinsed with two different b-axis ͑stacking axis͒ orientations, with DI water, and dried in flowing nitrogen.
depending on the substrate's temperature, have been cleaning, the back side of the glass substrate was roughened to suppress back side reflections for spec- maintained at room temperature, the preferred ori- troscopic ellipsometry ͑SE͒ measurement. For ab- entation of evaporated CuPc films less than 100 nm thick is in the ͓200͔ direction, with the ␣ form dom- deposited on quartz substrates were used.
of the orientation of crystallites occurs; the ␣ form is holder can hold up as many as four substrates, so the films fabricated on microscope slides, silicon, and substrates, the preferential orientation of the crystal- quartz were deposited in the same deposition process.
lites appeared to depend on the deposition rate and Pressure during evaporation was of the order 10Ϫ4 on whether the substrate was rotating ͓͑110͔ orien- The evaporation rate was 0.1- 0.2 nm͞s. The tation͒ or nonrotating ͓͑100͔ orientation͒.22 Bayliss distance from source to film was ϳ23 cm to ensure et al.32 reported that the ␣ form of the metal-free Pc uniformity of film thickness, and the substrate holder H Pc deposited on microscope slides at room temper- ature consists of small spherical particles that are controlled by a quartz thickness monitor.
randomly oriented, whereas oriented domains in the The film thickness was measured by a step profiler.
films have been found only for the ␤ form obtained by Pc films are absorbing in the spectral region of inter- annealing the ␣ form. The ␤ form obtained by direct Absorption is weak in the 450 -550-nm spectral evaporation onto substrates heated to 330 °C showed region, but assuming a Cauchy model for the refrac- elongated crystallites similar to those in the ␤ form tive index and disregarding the coefficient of extinc- obtained by annealing, but randomly oriented.32 tion in this spectral region can cause errors in the Highly oriented CuPc film fabricated by molecular- thickness determined from fitting of SE data.
beam deposition onto quartz glass substrates at low film thickness was also verified by AFM.
deposition rates has been demonstrated.34 mode scanning-probe microscope from Digital Instru- We report spectroscopic ellipsometry measure- ments was used for AFM measurements.
ments of CoPc, nickel phthalocyanine ͑NiPc͒, and thickness was measured at 10 points along the step to iron phthalocyanine ͑FePc͒ thin films deposited on minimize the influence of possible errors caused by glass and silicon substrates at room temperature.
variation in thickness around the step.
Because the deposition is performed at room temper- to revealing film thickness, AFM also yielded a ature, we expected that the films would be polycrys- talline and predominantly of the ␣ form. The roughness in all films was below 3.5 nm; the mean spectroscopic ellipsometry measurements were per- formed in the spectral range 1.55- 4.1 eV ͑300 - 800 48 nm for CoPc, 56 nm for FePc, and 45 nm for NiPc.
nm͒. Samples were modeled as isotropic layers, as Absorption measurements were performed on a was reported to be adequate for CuPc samples less Perkin-Elmer Lambda 20 UV͞Vis spectrophotome- The SE data in the 300 - 800 nm spectral range cule has an intrinsic anisotropy that is due to its were measured with a Jobin Yvon UVISEL phase- planar structure, for random orientation of crystal- lites in a polycrystalline film an isotropic model is someters are comparable to the more frequently used rotating analyzer ellipsometers in terms of attainable sorption measurements were also performed.
samples for absorption measurements were depos- The samples exhibited good uniformity, which we verified by performing measurements at different ar- eas of the sample on both glass and silicon substrates.
experimental details are described, and Section 3 pre- Also, the samples did not exhibit any in-plane anisot- ropy, as we verified by rotating the sample 90° and of the measured spectra was performed.
Experimental Details
simulated annealing algorithm to fit the data.38 Pc powders were purchased from Strem Chemicals.
fitting was performed simultaneously for glass and First, silicon substrates were cleaned in deionized ͑DI͒ water by a standard four-cycle dump-rinse pro-cedure.
After that, the wafer was cleaned for 10 min cycle dump-rinse in DI water was performed, fol- lowed by HF cleaning ͑HF:H O 1:50͒ for 30 s to drying, the wafers were immediately loaded into an ⌽͑X͒ ϭ ͑Xexp Ϫ Xcalc͒2, 1 November 2003 ͞ Vol. 42, No. 31 ͞ APPLIED OPTICS where i denotes the ith experimental data point, the positions of the peaks in the calculated extinction subscripts glass and Si denote glass and silicon sub- coefficient spectrum with the measured absorption strates, respectively, and the superscripts exp and calc denote experimental and calculated data.
an optional penalty function that can be introduced to Results and Discussion
ensure the smoothness of the n and k obtained by Pc films grown upon room-temperature substrates limiting the allowed difference between values ob- typically consist of small spherical particles and are tained in current ͑i͒ and previous ͑i Ϫ 1͒ data points.
predominantly of ␣ modification26,32 or possibly of the We obtained this difference in the following manner: If the difference between n and k values in current creased, the size of the crystallites increases, some and previous data points was larger than the speci- preferential ordering in neighboring domains be- fied allowed difference, ␥ was set to a large positive gins,25 and finally at higher temperatures the ␤ form Otherwise, ␥ was set to zero. Because of the ex- room temperature results in significantly smoother istence of multiple solutions, the use of such a penalty Vincett et al.35 believed that the films function to prevent unphysical abrupt changes in n were in the ␣ form, though at temperatures below Ϫ50 °C the films may have been partially amorphous.
indices of the substrates from the SE measurements Brinkmann et al.,25 however, stated that films grown of the substrates, instead of taking the data from the at temperatures of 5 °C-25 °C have disordered, i.e., material database, to ensure improved accuracy of close to amorphous, structures based on diffuse dif- the determined n and k of the Pc films.
Pan et al.42 have found that vanadyl strates, fitting with and without a native oxide layer Pc films deposited at room temperature and exposed to flowing nitrogen for 50 h are partially amorphous.
between the results obtained in these two cases.
In a previous analysis of the optical functions of CuPc The result of simultaneous point-to-point fitting for by variable-angle spectroscopic ellipsometry, Debe12 both substrates shows somewhat inferior agreement investigated Pc films grown at 0 °C and 70 °C, which with the experimental data compared with point-to- he assumed were well oriented, and deduced the ap- point fitting for one substrate only.
proximate optical functions that correspond to paral- ability of the data obtained by simultaneous fitting lel and perpendicular orientation with respect to for both substrates is considerably improved com- pared with data obtained by fitting for only one sub- ited at low temperature may not have been well un- strate because there are two unknowns and four iaxially oriented has been acknowledged,12 but the conditions that need to be satisfied.
possibility of phase change or of amorphousness of Correction for surface roughness was not made be- the films grown at low temperature has not been cause individual points yield different thickness val- films were grown upon room-temperature substrates, rough layer to the value averaged over a wavelength so it is reasonable to assume that the films consist of polycrystalline, predominantly ␣ phase domains with model the optical functions of the Pc layer and then to little or no preferential orientation among the neigh- fit the thickness of the Pc layer and the thickness of cases is less than 80 -100 nm, we have considered the els of the optical functions of inorganic semiconduc- tors are available, for organic materials the choice is an isotropic inner layer and an anisotropic outer typically limited to the Lorentz oscillator model and layer should be assumed, whereas a film thicker than its modifications39,40 or the Forouhi-Bloomer mod- 150 nm can be treated as a single homogeneous aniso- It was shown previously that the conventional Lorentz model is inadequate for modeling optical Figures 1-3 show tan͑⌿͒ and cos͑⌬͒ for CoPc, FePc, and NiPc films, respectively, deposited onto glass and Bloomer model was used only in a very narrow spec- tral range ͑365-564 nm͒ for a material that exhibited a single peak in the spectral range investigated.41 region investigated, two distinct absorption bands We attempted to use both the conventional Lorentz can be observed, a band near 2 eV ͑620 nm; Q band͒ model and its modifications, with and without and one near 3.5 eV ͑350 nm; B or Soret band͒. Both bands have been assigned to ␲ 3 ␲* transitions of the ment with the experimentally determined tan͑⌿͒ and macrocyclic ␲ system consisting of C and N at- cos͑⌬͒ was inferior compared to that from point-to- sorption in these two bands are shown to be of optical functions obtained by point-to-point fitting without surface-roughness correction.
among the visible spectra of various metal Pcs.
the reliability of the data by fitting the data for films However, metal-dependent absorption bands can be observed near the infrared spectra of Pcs.18 the quality of the result obtained by comparing the has been suggested that both Q and B bands can be APPLIED OPTICS ͞ Vol. 42, No. 31 ͞ 1 November 2003 Tan͑␺͒ and cos͑⌬͒ for CoPc films on glass and silicon The fit is a point-to-point fit; ͑a͒ glass substrate, ͑b͒ Absorption spectra of CoPc, FePc, and NiPc.
influenced by the metal-to-ligand charge-transfer of the origin of the observed peaks in the absorption bands in CoPc ͑Ref. 14͒, and the existence of a charge- spectrum of a Pc material is rather complex.
transfer band at the red edge of the Q band has been initio calculations of Pc molecules have been per- However, based on the small difference of formed, and the calculated results are in good agree- the shift of absorption peak on the low-energy side of the Q band of CuPc in solvents of different polarities spectra of solutions and thin films in which interac- ͑dimethyl sulfoxide and 1-chloronaphtalene͒,18 it is tions between individual molecules play a role are reasonable to conclude that the charge transfer does not play a significant role in transitions on the low- Figure 5 shows the real and imaginary parts of the indices of refraction of CoPc, FePc, and NiPc.
agreement between the peak positions in the calcu-lated extinction coefficients and the measured ab-sorption spectra can be observed.
determined the index of refraction of CoPc films onsilicon by spectroscopic ellipsometry in the spectralregion from 550 to 800 nm.
cient and absorption coefficient data show threepeaks in this spectral region, and the value of theextinction coefficient is greater than 0.5 over the en-tire region, although it should be low at 550 and 800nm according to absorption measurements reportedin the literature that consistently have shown onlytwo peaks.15,17 the very thick film ͑greater than 600 nm͒ investi-gated by Chen et al.14 CuPc an anisotropic model is needed for analysis of ͑␺͒ and cos͑⌬͒ for FePc films on glass and silicon Tan͑␺͒ and cos͑⌬͒ for NiPc films on glass and silicon sub- Real (n) and imaginary (k) parts of the index of refraction 1 November 2003 ͞ Vol. 42, No. 31 ͞ APPLIED OPTICS the ellipsometry data of films thicker than 150 nm.
Lucia and Verderame15 identified two peaks in the Q ropy was detected when the sample was rotated and band of the absorption spectrum of CoPc, located at the ellipsometry measurement repeated, but this 1.82 and 2.01 eV ͑680 and 616 nm͒. Davidson17 does not rule out the presence of anisotropy in the identified peaks at 1.82 and 1.98 eV ͑680 and 626 nm͒ in the Q band and at 3.79 eV ͑327 nm͒ in the B band.
are not likely to be ideally ordered, it is not possible Our results are in good agreement with previously to determine their anisotropy unambiguously.
reported absorption spectra and with the absorption ther investigations are needed to resolve this issue.
1.82 eV ͑680 nm͒ and 2.02 eV ͑615 nm͒ in k derived Conclusions
from SE and at 1.80 eV ͑688 nm͒ and 2.01 eV ͑618 Optical functions of CoPc, NiPc, and FePc thin films nm͒ in the absorption spectrum. In the B band we have been determined by use of spectroscopic ellip- observed peaks at 3.81 eV ͑325 nm͒ in k and of 3.84 sometry in the spectral range 300 - 800 nm.
films are likely to be polycrystalline with randomly For FePc, no previous data on the index of refrac- oriented crystallites, so an isotropic model of the sam- tion were reported to our knowledge.
spectra for FePc have been reported,16,17 and three determined by simultaneous point-to-point fits of peaks can be observed in the absorption spectra.
tan͑␺͒ and cos͑⌬͒ measured for films deposited upon Davidson17 obtained peaks at 1.73 eV ͑717 nm͒, 1.98 eV ͑626 nm͒, and 2.19 eV ͑566 nm͒ in the Q band and between experimental and calculated tan͑␺͒ and peaks at 3.51 eV ͑353 nm͒ and 3.83 eV ͑324 nm͒ in the cos͑⌬͒ curves was obtained for all three materials.
Our results show maxima in the extinction The agreement between the positions of the maxima coefficient at 1.74 eV ͑713 nm͒, 1.98 eV ͑625 nm͒, and in the imaginary part of the index of refraction and 2.18 eV ͑570 nm͒ in the Q band and wide structure in the absorption spectra reported in the literature and the B band with a maximum at 3.81 eV ͑325 nm͒, measured by us was good for all materials.
whereas in the absorption spectrum we observepeaks at 1.73 eV ͑717 nm͒, 1.96 eV ͑631 nm͒, 2.18 eV This study was supported by The Research Grants ͑568 nm͒, and 3.79 eV ͑327 nm͒, is in good agreement Council ͑RGC͒ of the Hong Kong Special Administra- with the absorption spectra reported in the litera- tive Region, China ͑projects HKU 7096͞00P and HKU 7075͞01P͒, the University of Hong Kong seed Also for NiPc, we are not aware of any previous funding research grant, and an RGC Co-operative study of the complex index of refraction.
Research Centers grant from the Hong Kong Univer- et al.44 obtained an absorption spectrum showing a two-peak structure and identified four transitions in References
the Q band of NiPc, located at 1.76 eV ͑704 nm͒, 1.83eV ͑678 nm͒, 1.98 eV ͑626 nm͒, and 2.02 eV ͑614 nm͒.
1. D. Wo¨hrle, L. Kreienhoop, and D. Schlettwein, "Phthalocya- Absorption measurements by Lucia and Verderame nines and related macrocycles in organic photovoltaic junc-tions," in Phthalocyanines: Properties and Applications, C. C.
showed absorption maxima at 1.85 eV ͑670 nm͒ and Leznoff and A. B. P. Lever, eds. ͑VCH, New York, 1996͒, pp.
1.98 eV ͑626 nm͒, in good agreement with our results showing peaks at 1.84 eV ͑673 nm͒ and 2.02 eV ͑615 2. P. Peumans, V. Bulovic', and S. R. Forrest, "Efficient photon nm͒ in the Q band and 3.70 eV ͑335 nm͒ in the B harvesting at high optical intensities in ultrathin organic In the absorption spectrum, peaks at 1.84 eV double-heterostructure photovoltaic diodes," Appl. Phys. Lett.
͑674 nm͒, 1.99 eV ͑623 nm͒, and 3.69 eV ͑336 nm͒ can 76, 2650 -2652 ͑2000͒.
3. M. Pfeiffer, A. Beyer, B. Plo¨nings, A. Nollau, T. Fritz, K. Leo, mum obtained near 2.19 eV ͑566 nm͒ is the lowest D. Schlettvein, S. Hiller, and D. Wo¨hrle, "Controlled p-doping likely reasons for this observation.
implications for their use in organic photovoltaic cells," Sol.
Energy Mater. Sol. Cells 63, 83-99 ͑2000͒.
4. Z. Bao, A. J. Lovinger, and A. Dodabalapur, "Organic field- the smoothest film obtained ͑rms roughness, 2.4 nm; effect transistors with high mobility based on copper phthalo- mean roughness, 1.7 nm͒, so it is not likely that sur- cyanine," Appl. Phys. Lett. 69, 3066 -3068 ͑1996͒.
face roughness is causing the difference from other 5. J. Blochwitz, M. Pfeiffer, T. Fritz, and K. Leo, "Low voltage organic light emitting diodes featuring doped phthalocyanine that there are differences in refractive indices at as hole transport material," Appl. Phys. Lett. 73, 729 -731
film-air and film- glass interfaces in NiPc films.44 The second reason is possible anisotropy of the films.
6. X. Zhou, M. Pfeiffer, J. Blochwitz, A. Werner, A. Nollau, T.
The refractive-index data determined by Schechtman Fritz, and K. Leo, "Very-low-operating-voltage organic light- emitting diodes using a p-doped amorphous hole injection 11 for a CuPc film with thickness of 23- 45 layer," Appl. Phys. Lett. 78, 410 - 412 ͑2001͒.
nm grown at room temperature upon LiF substrates 7. A. Fujii, Y. Ohmori, and K. Yoshino, "An organic infrared also show a dip in the refractive-index values to below electroluminescent diode utilizing a phthalocyanine film," 1.0 in the spectral region 2-2.5 eV ͑495- 620 nm͒.
IEEE Trans. Electron Devices 44, 1204 -1207 ͑1997͒.
This probably indicates a certain degree of preferen- 8. T. Nagasawa, K. Murakami, and K. Watanabe, "Improvement tial ordering in their samples, which is not surprising in NO -sensing characteristics of alpha-copper phthalocyanine APPLIED OPTICS ͞ Vol. 42, No. 31 ͞ 1 November 2003 thin films by the deposition on the hydrofluoric acid-treated a new polymorph of metal free phthalocyanine," J. Phys.
glass substrates," Mol. Cryst. Liq. Cryst. 316, 389 -392 ͑1998͒.
Chem. 72, 3230 -3235 ͑1968͒.
9. M. I. Newton, T. K. H. Starke, M. R. Willis, and G. McHale, 28. S. M. Bayliss, S. Heutz, G. Rumbles, and T. S. Jones, "Effect of "NO detection at room temperature with copper phthalocya- annealing on the properties of thin films of free base phthalo- nine thin film devices," Sens. Actuators B 67, 307-311 ͑2000͒.
cyanine and perylene-3,4,9,10 tetracarboxylic dianhydride de- 10. T. Fritz, J. Hahn, and H. Bo¨ttcher, "Determination of the posited by organic molecular beam deposition," Mater. Res.
optical constants of evaporated dye layers," Thin Solid Films Soc. Symp. Proc. 560, 71-74 ͑1999͒.
170, 249 -257 ͑1989͒.
29. F. Iwatsu, T. Kobayashi, and N. Uyeda, "Solvent effects on 11. B. H. Schechtman and W. E. Spicer, "Near infrared to vacuum crystal growth and transformation of zinc phthalocyanine," J.
ultraviolet absorption spectra and the optical constants of Phys. Chem. 84, 3223-3230 ͑1980͒.
phthalocyanine and porphyrin films," J. Mol. Spectrosc. 33,
30. F. Iwatsu, "Size effects on the alpha- beta transformation of phthalocyanine crystals," J. Phys. Chem. 92, 1678 -1681
12. M. K. Debe, "Variable angle spectroscopic ellipsometry studies 31. S. I. Shihub and R. D. Gould, "Studies of phase transforma- nine," J. Vac. Sci. Technol. A 10, 2816 -2821 ͑1992͒.
tions in some metal phthalocyanine thin films using measure- ¨ th, "The electronic structure of GaP ͑110͒ and ments of current as a function of temperature," Thin Solid Cu-phthalocyanine overlayers studied by ellipsometry," Appl.
Films 290-291, 390 -394 ͑1996͒.
Phys. A 31, 75- 80 ͑1983͒.
32. S. M. Bayliss, S. Heutz, R. Cloots, R. L. Middleton, G. Rumbles, 14. Q. Chen, D. Gu, and F. Gan, "Ellipsometric spectra of cobalt and T. S. Jones, "Templating effects in the growth of metal-free phthalocyanine films," Physica B 212, 189 -194 ͑1995͒.
phthalocyanine polymorphic double layers," Adv. Mater. 12,
15. E. A. Lucia and F. D. Verderame, "Spectra of polycrystalline phthalocyanines in the visible region," J. Chem. Phys. 48,
33. M. K. Debe, R. J. Poirier, and K. K. Kam, "Organic-thin-film- induced molecular epitaxy from the vapor phase," Thin Solid
Films 197, 335-347 ͑1991͒.
16. H. Laurs and G. Heiland, "Electrical and optical properties of 34. M. Komiyama, Y. Sakakibara, and H. Hirai, "Preparation of phthalocyanine films," Thin Solid Films 149, 129 -142 ͑1987͒.
highly ordered ultrathin films of copper͑II͒ phthalocyanine on 17. A. T. Davidson, "The effect of the metal atom on the absorption amorphous substrates by molecular beam deposition," Thin spectra of phthalocyanine films," J. Chem. Phys. 77, 168 -172
͑
Solid Films 151, L109 -L110 ͑1987͒.
35. P. S. Vincett, Z. D. Popovic, and D. McIntyre, "A novel struc- 18. L. Edwards and M. Gouterman, "Porphyrins.
tural singularity in vacuum-deposited thin films: anism of critical optimization of thin film properties," Thin Spectrosc. 33, 292-310 ͑1970͒.
Solid Films 82, 357-376 ͑1981͒.
19. P. E. Fielding and A. G. MacKay, "Electrical conduction in the 36. M. A. Barrett, Z. Borkowska, M. W. Humphreys, and R. Par- Optical properties," Aust. J. Chem. 17,
sons, "Ellipsometry of thin films of copper phthalocyanine," Thin Solid Films 28, 289 -302 ͑1975͒.
20. A. Schmidt, L. K. Chau, A. Back, and N. Armstrong, "Epitaxial 37. D. E. Aspnes, "Precision bounds to ellipsometer systems," phthalocyanine ultrathin films grown by organic molecular Appl. Opt. 14, 1131-1136 ͑1975͒.
beam epitaxy ͑OMBE͒," in Phthalocyanines: Properties and 38. A. B. Djurisˇic', A. D. Rakic', and J. M. Elazar, "Modeling the Applications, C. C. Leznoff and A. B. P. Lever, eds., ͑VCH, New optical constants of solids using acceptance-probability- controlled simulated annealing with an adaptive move gener- 21. H. Hoshi, A. J. Dann, and Y. Maruyama, "The structure and ation procedure," Phys. Rev. E 55, 4797- 48903 ͑1997͒.
properties of phthalocyanine films grown by the molecular- 39. J. Ma¨rtensson and H. Arwin, "Applications of derivative line- shape fitting to ellipsometric spectra of thin films of metal- scopic study," J. Appl. Phys. 67, 1845-1849 ͑1990͒.
substituted phthalocyanines," Thin Solid Films 205, 252-257
22. B. Resel, M. Ottmar, M. Hanack, J. Keckes, and B. Leising, "Preferred orientation of copper phthalocyanine thin films 40. A. B. Djurisˇic', T. Fritz, and K. Leo, "Modelling the optical evaporated on amorphous substrates," J. Mater. Res. 15, 934 -
tive function," J. Opt. A 2, 458 - 464 ͑2000͒.
23. N. Uyeda, M. Ashida, and E. Suito, "Orientation overgrowth of 41. B. Masenelli, S. Callard, A. Gagnaire, and J. Joseph, "Fabri- condensed polycyclic aromatic compounds vacuum-evaporated cation and characterization of organic semiconductor-based on cleaved face of mica," J. Appl. Phys. 36, 1453-1460 ͑1965͒.
microcavities," Thin Solid Films 364, 264 -268 ͑2000͒.
24. H. Wachtel, J. C. Wittmann, B. Lotz, M. A. Petit, and J. J.
42. Y. L. Pan, Y. J. Wu, L. B. Chen, Y. Y. Zhao, Y. H. Shen, F. M.
Andre, "Anisotropic spin transport in oriented lithium phtha- Li, S. Y. Shen, and D. H. Huang, "Structure and spectroscopic locyanine thin films," Thin Solid Films 250, 219 -231 ͑1994͒.
characterization of polycrystalline vanadyl phthalocyanine 25. M. Brinkmann, J. C. Wittmann, C. Chaumont, and J. J. Andre, ͑VOPc͒ films fabricated by vacuum deposition," Appl. Phys. A "Effects of solvent on the morphology and crystalline structure 66, 569 -573 ͑1998͒.
of lithium phthalocyanine thin films and powders," Thin Solid 43. P. N. Day, Z. Wang, and R. Pachter, "Calculation of the struc- Films 292, 192-203 ͑1997͒.
ture and absorption spectra of phthalocyanines in the gas- 26. O. Berger, W. J. Fischer, B. Adolphi, S. Tierbach, V. Melev, and phase and in solution," Thermochem. Acta 455, 33-50 ͑1998͒.
J. Schreiber, "Studies on phase transformations of Cu phtha- 44. J. Godlewski, J. Kalinowski, S. Stizza, I. Davoli, and R. Ber- locyanine thin films," J. Mater. Sci. Mater. Electron. 11, 331-
nardini, "Asymmetries in the optical properties of vacuum- deposited organic films illuminated at the substrate and non- 27. J. H. Sharp and M. Lardon, "Spectroscopic characterization of substrate surfaces," Thin Solid Films 146, 115-132 ͑1987͒.
1 November 2003 ͞ Vol. 42, No. 31 ͞ APPLIED OPTICS

Source: http://www.cdr.ust.hk/publications/research_pub/2003/AppOptVol42pp6382-6387.pdf