.gif?ad=15653&adview=true)
Combinatorial approach for ferroelectric material libraries prepared by liquid source misted chemical deposition method
- *Department of Chemical and Biomolecular Engineering and Center for Ultramicrochemical Process Systems, and
- †Department of Materials Science and Engineering, Korea Advanced Institute of Science and Technology, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Korea
-
Communicated by Gabor A. Somorjai, University of California, Berkeley, CA, November 24, 2006 (received for review August 24, 2006)
Abstract
Combinatorial approach for discovering novel functional materials in the huge diversity of chemical composition and processing conditions has become more important for breakthrough in thin film electronic and energy-conversion devices. The efficiency of combinatorial method depends on the preparation of a reliable high-density composition thin-film library. The physico-chemical properties of each sample on the library should be similar to those of the corresponding samples prepared by one-by-one conventional methods. We successfully developed the combinatorial liquid source misted chemical deposition (LSMCD) method and demonstrated its validity in screening the chemical composition of Bi3.75LaxCe0.25-xTi3O12 (BLCT) for high remanent polarization (P r). LSMCD is a cheap promising combinatorial screening tool. It can control the composition up to ppm level and produce homogeneous multicomponent library. LSMCD method allows us to prepare BLCT thin-film library at the variation of 0.4 mol% of La. Maximum 2P r is 35 μC/cm−2 at x = 0.21. The intensity of (117) XRD peak is quantitatively related to 2P r. Newly developed scanning piezoelectric deformation measurement for nano-sized samples using scanning probe microscope (SPM) is also found out to be reliable for determining the relative ranking of P r value rapidly.
Combinatorial methods consist of library design, preparation of library, evaluation of its functional properties, and its physicochemical characteristics. These methods are considered to be very efficient and quickly provide us the quantitative activity structure relationship to establish the criteria for designing novel functional materials, which have been extensively used for the discovery of functional electronic thin film materials, novel polymers, and catalysts since the mid-1990s (1–10).
Thin-film combinatorial libraries have been synthesized mainly by RF-sputtering, pulsed laser ablation, and chemical vapor deposition. Thin-film libraries containing various compositions were fabricated by adopting multimask techniques (3–6). Ten binary masks or five quaternary masks should be used to prepare 1,024 (210) different samples on a wafer. These mask operations complicated the operation of the instrument as well as creating the contamination problems. These problems were solved by adopting only two masks, one x–y movable shutter and one shadow mask (7–11). The x–y movable shutter moves at a constant velocity in a linear or stepwise way to control the deposition time in one direction of wafer. The shadow mask containing regularly spaced holes is located just below the moving shutter, which fabricates a high-density discrete library in a very simple manner. Recently, off-axis codeposition method was reported to prepare thin-film metal alloy library (12, 13). By controlling the distance between the target guns and the substrate, multitarget materials can be deposited on the substrate with concentration gradient simultaneously. This method does not need to use a movable shutter or many masks. Only shadow mask is necessary when the fabrication of discrete library is preferred. However, this method cannot control the thickness of each sample on the library and is not suitable when the physicochemical properties vary with thickness.
The library fabrication methods discussed above need to prepare single- or multicomponent target materials, which is sometimes very expensive, and it can be difficult to obtain a uniform composition. It is also very difficult to control the deposition rate of each component in multicomponent target arising from the difference in its inherent sputtering property. Sputtering rate depends on many variables, so it is too difficult to control precisely. Therefore, accurate control of composition within ±3% in the library cannot be obtained. To overcome these problems, some research groups applied the wet process including spray pyrolysis and ink-jet technique for combinatorial library synthesis. However, these processes did not accomplish a finely compositional tuned thin-film library because of their inherent weakness (14–17).
Therefore, in our study, liquid source misted chemical deposition (LSMCD) method was developed to prepare a thin-film library. Adequate amount of metal precursor compounds was dissolved in suitable solvent. This mixed solution is equivalent to the target in the sputtering method. Composition control is more accurate with LSMCD than with conventional sputtering, because exact weighing can be obtained even in ppm level. Furthermore, the cost of metal precursor solution (a few dollars) is just the cost of chemicals and much cheaper than target (average cost for 4-inch target is ≈$5,000). Hundreds of targets are required for the fabrication of a very diverse library. The material cost to fabricate a thin-film library with LSMCD is negligible compared with conventional sputtering method. Ultrafine mist is generated with nebulizer (frequency: 1.65 MHz) and transported by Ar flow to the substrate. Suitable thermal treatments are required to obtain multicomponent metal alloy or metal oxide thin-film (18–23). The variation of deposition time with x–y movable shutter will generate the concentration gradient in a similar manner to combinatorial sputtering. It is possible to fabricate homogeneous liquid-phase combinatorial library by liquid-phase mixing in LSMCD, whereas in sputtering method, each component of library have to be mixed by solid-state diffusion, resulting in the formation of more homogeneous multicomponent compound with LSMCD. It can be concluded that LSMCD enables us to fabricate a thin-film library with more accurate control of composition at a low cost. The price of the LSMCD instrument is also much cheaper than a conventional PVD instrument. These advantages of LSMCD over the sputtering method will be demonstrated in fabricating a multicom ponent ferroelectric oxide thin-film library for the optimization of its electrical properties as follows.
Bismuth layered perovskite ferroelectric films, SrBi2Ta2O9 (24) and (Bi,La)4Ti3O12 (BLT) (25), have drawn an interest for ferroelectric random-access memory because of the fatigue-free properties over 1010 read/write cycles, unlike Pb(Zr,Ti)O3 (PZT) (26, 27). However, these materials have smaller remanent polarization (P r) than PZT. Numerous attempts have been made to substitute lanthanide group atoms for Bi site in Bi4Ti3O12 to increase P r. It is well known that P r depends on the chemical composition (28–32). Woo et al. first presented that the Ce-substituted Bi4Ti3O12 (Bi4-xCexTi3O12, BCT) fabricated by LSMCD exhibited large P r (2P r = 20 μCcm−2 at 15 V, x = 0.75) and superior fatigue endurance over 1010 read/write cycles. Hence, it is quite reasonable to substitute Bi with both La and Ce for ferroelectric materials with high P r. We prepared a Bi3.75LaxCe0.25-xTi3O12 (BLCT, 0 ≤ x ≤ 0.25) array with LSMCD to optimize the chemical composition for high P r. Bi/(La+Ce) was fixed to 3.75/0.25 because our structural refinement result with BLT powder showed that 2P r was maximum when the Bi/La was 3.75/0.25 (33–35).
Fast and reliable measurement of remanent polarization is required for a continuous-gradient library and μm-sized samples, because the conventional capacitance–voltage measurement method is slow and requires electronic wirings. A scanning probe microscope can apply the electric field into the sample on the library using a sharpened tip as movable electrode and induced the local deformation of samples. The correlation between the deformation length and P r will be discussed.
Results and Discussion
A BLCT (0 ≤ x ≤ 0.25) thin-film library was prepared by LSMCD equipped automated shutter (see Materials and Methods).
The chemical compositions of BLCT samples on the library measured by wavelength dispersive spectroscopy and thickness of each sample on the library measured by SEM cross-section image are shown in Table 1. The real location of each sample is the same as the location of Table 1; 0.4 mole % of composition control was demonstrated, as was the uniform thickness, 200 nm ± 10 nm.
The chemical composition and thickness of BLCT thin-film 8 × 8 samples on a library annealed at 700°C for 1 h in furnace under oxygen atmosphere
Microbeam XRD (Bruker D8Discover GADDS CS, CuKα radiation) was used to analyze the crystallographic phase and the preferred orientation of as-prepared thin-film samples on the library. As shown in Fig. 1 A, regardless of La content, all samples on the library did not contain any peaks of Bi2O3 (2θ = 28°) as well as the other mono- or binary oxides. The intensities of (00n) peaks arising from the extent of crystal growth in the c-axis did not change much with La content. However, the intensity of (117) diffraction peak normalized by that of Pt (111) peak increased with the increase of La content and reached the maximum at x = 0.21. In the region of x > 0.21, there was an abrupt decrease in this value (Fig. 1 B). The crystal growth in the a and b axes is related to the intensity of the (117) peak.
Structural analysis by using microbeam XRD. (A) The microbeam XRD spectra of BLCT thin-film samples on the library as a function of La content, x. These spectra were scanned in θ-2θ type in the range of 2θ between 17 and 53° under Cu Kα radiation (λ = 1.5405 Å) operated at 40 mA and 40 kV. (B) The ratio of instensity of (117) diffraction peak to that of Pt (111) diffraction peak was plotted as a function of La content.
It was reported that P s is only 4μCcm−2 along the c axis and 50μCcm−2 along the a or b axis measured with single crystal Bi4Ti3O12 (36, 37). It was also reported that BLT thin film grown epitaxially in the a axis shows a P s value of 50μCcm−2 (39).
Fig. 2 shows the grain size as a function of La content. As shown in Fig. 2, grain size increased with the increase of La content. This can improve the ferroelectricity and piezoelectricity of thin film. However, when the grain size is larger than 0.5 μm, the leakage current density increased from 10−6 Acm−2 to 10−4 Acm−2 in the region of La content greater than x = 0.2 [supporting information (SI) Fig. 7].
The surface morphology of BLCT thin films as a function of La content. x = 0 (A), x = 0.08 (B), x = 0.17 (C), and x = 0.25 (D) in BLCT combinatorial array, where x is La content.
2P r values of 64 samples on the library were measured by the polarization-electric field (P–E) hysteresis loop at the applied voltage of 10 V as shown in Fig. 3. Twenty Pt top electrodes (diameter: 200 μm) were deposited on each sample with a thickness of 100 nm to form a capacitor structure. As shown in Fig. 3, the remanent polarization increased with La content (x). P r increased slightly between x = 0 and x = 0.10, but increased rapidly between x = 0.1 and 0.21 and decreased abruptly when x > 0.23. The maximum 2P r (35μCcm−2) was obtained at x = 0.21. Comparing Fig. 3 with Fig. 1 B, it could be suggested that the remanent polarization of BLCT thin film (x ≤ 0.25) be quantitatively related to the intensity of normalized (117) peak of BLCT. As shown in Fig. 4, BLCT shows excellent fatigue endurance at 5 V over 1010 cycles similar to other bismuth-layered ferroelectrics. This is due to the increase of structural stability induced by decrease in oxygen vacancies (27). Substitution of La and Ce by Bi decreased oxygen vacancies.
The characterization of P–E characteristic using a ferroelectric tester (RT66A). The plot of 2P r of BLCT thin-film samples on the library as a function of La content, x. Pt top electrodes of 200 μm diameter were deposited on the thin-film library by using RF-sputtering system to fabricate the capacitor structure for measuring the remanent polarization. The thickness of top electrode was ≈100 nm.
Fatigue endurance behavior of BLCT thin film samples (x = 0.014, 0.172, 0.194). The frequency was 1 MHz, and applied voltage was ± 5 V. Only these compositions were tested because of long test time.
Four different BLCT thin films (2 cm × 2 cm) were fabricated at the same deposition condition of combinatorial experiments discussed above by varying the La content (x = 0.1, 0.12, 0.18, and 0.194) to validate the results obtained by 8 × 8 thin film samples on the combinatorial library. As shown in Fig. 5 A, the change in 2P r of BLCT individually prepared shows a similar trend to that of BLCT prepared combinatorially, even though 2P r values of individual BLCT samples are slightly lower than those of combinatorial array. Such an offset was due to the thickness difference between samples on the library (200 nm) and individual (240 nm) samples. Even though we need the identical deposition condition for samples on the library and individual samples, it was found out that the deposition rate for samples on the library is 20% lower than that for individual samples, because the shadow mask and moving mask used for samples on the library hindered mist from approaching to the substrate. As shown in Fig. 5 B and C, the thicker individual ferroelectric samples have larger coercive field strength. As the coercive field strength increases, higher electric field strength is necessary to switch the dipole moment. Therefore, individual sample shows a little bit lower remanent polarization than sample on the library at the same chemical composition under the same electric field (500 kV/cm).
Confirmation of the reliability of combinatorial experimentation by using LSMCD process. (A) The comparison of 2P r of BLCT samples deposited by combinatorial experimentation with that of BLCT samples deposited individually (one sample at one time). (B) P–E hysteresis loop of BLCT samples chosen among the BLCT library. (C) P–E hysteresis loop of BLCT fabricated individually at the same condition.
This result indicated that the preparation of combinatorial array by LSMCD is reliable for the rapid screening of various thin films having various compositions.
A high-density library (2,500 samples per wafer) can be prepared with existing technology, but the evaluation of their functional properties and physico-chemical characterization cannot be performed rapidly as long as conventional procedures are used. It is quite desirable to develop new methods to measure the functional and physico-chemical properties of thin film samples on the library in a very short time. P–E hysteresis characteristics of BLCT thin film samples on the library were measured with RT66A ferroelectric tester one by one for each sample in this study, which is not suitable for high-density library because it takes long time. Here, we report the rapid indirect measurement of remanent polarization by measuring the piezo-electricity with a scanning probe microscope (SPM) (38–41). A metallic atomic force microscope (AFM) tip was used as a movable top electrode. The electric field was applied between the AFM tip and bottom electrode to measure the topographic deformation. A laser beam was focused on the upper side of the cantilever and photo-diode to detect an angle of reflection from the upper side of the cantilever as shown in Fig. 6 A. This enables us to detect the topographic deformation up to 0.01 nm. Local piezo-response on the grain or domain of each samples on the library can be detected nondestructively just by scanning the AFM tip using XY manipulator, which will decrease the measurement time extensively compared with conventional measurement of the P–E hysteresis loop. Tip bias was continuously changed from −10 V to +10 V, and its topographic deformation of typical BLCT film library is shown as a curve of butterfly shape as shown in Fig. 6 B. Fig. 6 C shows the relationship between topographic deformation (H = H1 + H2) and remanent polarization (2P r) in BLCT array as a function of La content (x). H1 and H2 are lift heights with a tip bias, −10 V and +10 V, respectively. Piezoelectricity and ferroelectricity show the similar tendency as a function of La content. Maximum H was 0.9 nm at x = 0.2, indicating that relative ranking of P r can be estimated rapidly on the basis of topographic deformation. This method can be used reliably for measuring P r of discrete or continuous high-density ferroelectric combinatorial array indirectly.
Mapping of topographic deformation using the SPM. (A) The schematic drawing of measuring topographic deformation by using SPM. (B) The topographic deformation of samples detected by SPM. Detection time was 1 s per point and DC voltage was changed continuously from −10 V and 10 V. A 2 μm × 2 μm region of each library was scanned with a gold-coated AFM tip. (C) The plot of topographical deformation (H = H1 + H2) and 2P r of BLCT thin-film samples on the library as a function of La content.
In conclusion, we fabricated a BLCT thin-film library by using LSMCD equipped with an automated shutter, and mapped the structure and remanent polarization as a function of La content. This library-synthetic method allows us to control the chemical composition accurately (ppm level in the library). Structural analysis by microbeam XRD clearly shows that (117) crystalline growth has an important role in improving the remanent polarization. Newly developed scanning piezoelectric deformation measurement is found out to be reliable for determining the relative ranking of P r value rapidly.
Materials and Methods
Preparation of Precursor Solutions.
For the preparation of BLT (Bi/La = 3.75/0.25, Bi 20% excess) solution, 2.465 g of Bi(NO3)3·6H2O and 0.122 g of La(NO3)3·6H2O were dissolved in 10 ml of 2-methoxyethanol (Solution 1). One milliliter of Ti isopropoxide was dissolved in 9 ml of 2-methoxyethanol (Solution 2). Solution 1 and 2 were mixed. After the solution mixing, 10 ml of 2-methoxyethanol and 1 ml of 2-ethylhexanoic acid were added and stirred for 30 min. In preparing BCT (Bi/Ce = 3.75/0.25, Bi 20% excess), Ce nitrate was added instead of La nitrate, and the other conditions were the same as those of preparing BLT solution.
Preparation of BLCT Thin Film Library by LSMCD Technique.
Schematic diagrams of LSMCD are shown in SI Fig. 8A.
We generated the mists of BLT and BCT metal precursor solution by using an ultrasonic nebulizer. First, BCT mist was transported to the top of the computer-operated shutter by an Ar flow of 500 sccm. Below the shutter, shadow mask containing 8 × 8 holes of 3 mm in diameter was intimately contacted on top of the wafer. BLT mist was deposited on the wafer in a stepwise fashion (deposition time: 20 sec/step) by moving shutter in x-direction. Linear variation of deposition time generated a linear concentration gradient of BLT in x-direction. Then, BLT mist was deposited in the same manner except for the movement of shutter in the opposite direction. After this procedure, the substrate holder was rotated by 90°; BLT and BCT mists were deposited in the same manner except for the deposition time (13 sec/step) as explained in SI Fig. 8B. The above methods complete a single cycle of deposition. We were then able to generate the liquid phase library of same volume but with different Ce/La ratio. After 3 cycles, we prepared the BLCT thin-film library with 200 nm thickness after thermal treatments at 230°C for 2 min, at 400°C for 10 min in air, and at 700°C for 1 h under O2 atmosphere.
Characterization of BLCT Thin Film Library Prepared by LSMCD Technique.
To analyze the new crystallographic phase and preferred orientation of as-prepared thin-film library, we can use the microbeam XRD (D8 DISCOVER with GADDS for combinatorial screening by Bruker-AXS) with small beam size (<500 μm) and xyz manipulator controlled by computer software. The angle between detector and substrate was 20° and the angle between x-ray gun and substrate was 15°. The XRD mapping was carried out in θ-2θ scans under Cu Kα radiation (λ = 1.5405 Å) operated at 40 mA and 40 kV.
The chemical composition of prepared thin-film library was obtained by wavelength dispersive spectroscopy (Microspec 3-PC). Surface morphology of thin-film library was analyzed by SEM (Philips 533M).
To measure the electrical properties, Pt electrodes with 200 μm diameter were deposited on the thin-film library by using the RF-magnetron sputtering system. P–E hysteresis was measured by using a Radiant Technology RT66A ferroelectric tester at 10 V. The leakage current characteristics were measured by using programmable Kethley 617 electrometer with the condition of 0.05 V of step voltage and 0.1-sec delay time.
A scanning probe microscope (SPA 400 SPI 3800N, Seiko Instrument Korea, Seoul, South Korea) was used to measure or scan the topographical change and surface roughness of thin film array. When measuring the topographical change, the detection time was 1 sec per point, and voltage was changed from −10 V and 10 V. The scan region was 2 μm × 2 μm, and a gold-coated AFM tip (SI-DF3-A, C = 2.3 N/m) was used as a movable top electrode. The measurement was carried out >20 times per point, and the induced topographical change was averaged.
Acknowledgments
This research was funded by Center for Ultramicrochemical Process Systems (CUPS) sponsored by Korea Science and Engineering Foundation.
Footnotes
- ‡To whom correspondence should be addressed. E-mail: siwoo{at}kaist.ac.kr
-
Author contributions: K.W.K. and S.I.W. designed research; K.W.K. and T.S.K. performed research; K.W.K., K.S.O., and Y.S.K. contributed new reagents/analytic tools; K.W.K., M.K.J., and S.I.W. analyzed data; and K.W.K. and S.I.W. wrote the paper.
-
The authors declare no conflict of interest.
-
This article contains supporting information online at www.pnas.org/cgi/content/full/0610146104/DC1.
- Abbreviations:
- LSMCD,
- liquid source misted chemical deposition;
- BLT,
- (Bi,La)4Ti3O12;
- BCT,
- Bi4-xCexTi3O12;
- BLCT,
- Bi3.75LaxCe0.25-xTi3O12;
- P–E,
- polarization-electric field;
- SPM,
- scanning probe microscope;
- AFM,
- atomic force microscope.
-
Freely available online through the PNAS open access option.
- © 2007 by The National Academy of Sciences of the USA