Characteristics of X-ray absorption near-edge spectra obtained from various MgB2 films

 

C. H. Hsieh1, C. N. Chang1, S. Y. Wang1, Pohan Lee2, H. C. Hsu1, Y. Cui3, X. X. Xi3, Eun-Mi Choi4, Sung-Ik Lee4, and J. M. Chen5

1. Department of Physics, National Taiwan Normal University, Taipei, Taiwan, R.O.C.

2. Department of Physics, National Taiwan University, Taipei, Taiwan, R.O.C.

3. Department of Physics, Pennsylvania State University, University Park, Pennsylvania 16802, U. S. A.

4. National Creative Research Initiative Center for superconductivity and department of Physics, Pohang University of Science and Technology, Pohang 790-784, Korea

5. National Synchrotron Radiation Research Center, Hsinchu Science-Based Industrial Park, Hsinchu, Taiwan, R.O.C.

 

   X-ray absorption near-edge spectrum (XANES) of various MgB2 films prepared under different procedures has been measured by using synchrotron radiation. By comparing to the band structure calculation we observed that the major spectral peaks of B K edge XANES resulted from the empty states with p characters of boron in MgB2. The spectral peak near the threshold indicates the existence of the hole states in the valence band at the Fermi level. This is particularly clear in some c-axis oriented films. The peaks of B2O3 were common peaks in all of the films we studied, indicating the contamination of B2O3. This peak sometimes will smear the main spectral structure of MgB2 if it is seriously contaminated.

   The experimental XANES of Mg K edge results from the empty states with Mg p characters can be identified by comparing the experimental spectrum with the band structure calculations. A small core-hole effect resulted from the less electron screening of the Mg ions in MgB2 can be identified. There are empty states, as appear in the threshold spectral structure, extended from the Fermi level indicates the metal characters of Mg in MgB2. Little anisotropic characteristics between px,y and pz can be identified from the polarization-dependent measurements, except the spectral intensity around 1314 eV. Around that energy, the spectral intensity is higher for the electric field of the beam parallel than perpendicular to the boron planes, in accordance with the band structure calculations.

 

 

 

 

 

I. Introduction

Since the discovery of the superconductivity of MgB2 [1], many works have been done for understanding the precise mechanism of its record-high value of the superconducting temperature (〜 40 K) among the binary intermetallic compounds.

It is known that MgB2 is a metal with a layer structure with alternating layers of boron and magnesium atoms. The boron atoms form honeycombed layers similar to the layers of graphite, with covalent, sp2 hybridization bonds (σ - bonds) and the pz bonds (π - bonds). The magnesium atoms are located above the center of the hexagons in-between the boron planes. The magnesium atoms will transfer charges to boron atoms for having enough electrons to form the σ - bonds and π- bonds. In fact the electronic structure calculated on the basis of a fictional system □2+B2 is very similar to that of MgB2 [2], except that σ - bonds, as in graphite, are completely filled. The graphite becomes superconducting up to 5K when doped (intercalated) [3]. Several different ab initio band calculations show that the σ - bonds and π - bonds of MgB2 are not completely filled at the Fermi level [2, 4-7]. The energy position of the B σ(2px,y) bands at the point of Brillouin Zone lie above the Fermi level and form hole-type cylindrical elements of the Fermi surface [8]. The B 2pz states are oriented perpendicularly to the boron plane and responsible for the weak interlayer π bonds. Both σ and π bands of B are suggested to be critical for the description of the mechanism of the superconductivity of MgB2 [8, 9].

    One also notices that magnesium states in MgB2 are important to understand the charge transfer mechanism of Mg to B. A hybridization of B 2pz level and Mg 2p level will lower the former while raising the latter. This hybridization effect lowers the π(pz) bands relative to the bonding σ bands and cause σ → π charge transfer and hole doping in σ band, deriving the superconductivity in MgB2 [2].

The unoccupied states with p character close to the Fermi level can be easily probed by the K X-ray absorption near-edge spectrum (XANES) because of the dipole transition rule. By using the technique of X-ray absorption spectroscopy on films of MgB2 one can look into the partial density of states (PDS) above the Fermi level. If the film has c-axial oriented crystalline structure one can use polarized X-ray to probe the PDS parallel to the boron/magnesium planes (ab plane) and perpendicular to the planes. In other words, one can probe the empty states of px,y and pz.

Nevertheless, fabricating films of MgB2 is not an easy task since MgB2 is not easily stabilized substances. Contaminations will attribute to the characteristics of the XANES of MgB2 films. We report our study of various films of MgB2 fabricated by different method by XANES and show that the main characteristics of the XANES disturbed little by the contaminations. The characteristics of XANES indicate the characteristics of the PDS (the electronic structure) of MgB2.

 

II. Experiments

A.    Samples

The MgB2 films used in this work were made by different methods: Samples S1 and S2 that were obtained by post-annealing a precursor film made by ion sputtering a Mg-rich MgB2 target on a substrate of R-plane Al2O3 single crystal in either an oxygen-free copper cell (for S1) or an ultra-high vacuum comparable stainless steel cell (for S2) under magnesium vapor at about 700 °C (S1) and 800 °C (S2)[12]; the sample S3 that was obtained by post-annealing a precursor film made by laser ablation boron on a substrate of R-plane Al2O3 single crystal in a tantalum tube under a saturated magnesium vapor at about 900 °C [10]; and the sample S4 that was obtained by the hybrid physical-chemical vapor deposition (HPCVD) technique, deposited on the substrate of (0001) 4H-SiC [11]. Fig.1 shows the temperature dependent resistance of these samples. The superconducting transition temperatures of S1 and S2 are about 25 K with transition width about 2.5 K while the transition temperature of S3 is 39 K with transition width of 0.5K. S4 has the highest transition temperature of 41.10 K among the investigated samples. Its transition width is 0.16 K. Samples of S1 appears to be a group of nanoparticles [12], S2 is a polycrystalline film, while S3 and S4 are c-axis oriented films.

B.    X-ray absorption near edge spectra (XANES)

   Our soft X-ray absorption spectra were measured at room temperature at the beam line 20A at the National Synchrotron Radiation Research Center (NSRRC) in Taiwan. The beam line is a spherical-grating-monochromator beam line [13]. Gratings with 400n/mm and 1200 n/mm were used respectively for the measurements of the K edge absorption of B and Mg. The slits of the beamline monochromator were set to be 30 µm providing a resolving power of about 1900 to 2100 depending on the grating used.

   The experimental geometry is indicated in Fig.2, where the normal n of the ab plane of the film is indicated. The photon detector is a multichannel plate (MCP) that is sited at 45° with respect to the incident beam (νin) to detect the outgoing X-ray fluorescence (νout). In other words, we use a bulk sensitive total-fluorescence-yield (TFY) method for obtaining the soft X-ray absorption near–edge spectra (XANES). Simultaneously, the sample current signal is collected. This signal measures the photoelectrons coming out of the sample by the impact of photons. It is a total-electron yield (TEY) method for the XANES. The spectrum obtained by the TEY is surface sensitive due to the short mean free path of the outgoing electrons created by the impact of photons in the sample. The intensity of the incident photons was measured by the current generated by a fresh nickel mesh located in front of the target chamber. By rotating the sample holder we can scan the polarization vector parallel to the ab plane toward to the c-axis by varying the angle α from 0° to 75°. Note that the normal n will pass the line of νout for α＞45° and the angles α and β are on the same side of normal n.

   The energy of the XANES of boron K edge was calibrated by referring to the resonance peak of B2O3 to be 193.8 eV, while that of Mg K edge was calibrated by referring to the energy at the inflection point of the K edge of Mg metal to be 1303.0 eV. Each spectrum was normalized with the intensity of the incident photons.

The target self-absorption was not corrected. It is proper for the XANES of Mg K edge since the mean free path of the photons there is about 1.2 μm, much greater than the target thickness and therefore, the target self-absorption effect can be neglected. The XANES of B K edge may be distorted due to target self-absorption effect since the mean free path of the photons there is about 100 nm smaller than the target thickness that is about 400 nm to 700 nm for the samples we studied. The resonance peak may be suppressed due to a larger self-absorption effect. Besides, the self-absorption effect is also angular dependent [14], which will affect the overall spectral features. Nevertheless, the resonance spectral positions will not be altered. It is difficult to do the self-absorption correction for samples contain impurities. The uncorrected spectrum will not affect the conclusions of this work as will be shown later.        

 

III. Results and Discussions

(A) General features of the XANES 

Fig.3 shows the XANES of B K edge for all samples. They were taken at α = 45°. The spectrum of B2O3 is included for comparisons. Both TFY (Fig.3 (a)) and TEY spectra (Fig.3 (b)) are shown. The peaks resulting from B2O3 (the peak D at 193.8 eV) is seen in all samples, it is relatively clear in the TEY spectra. The peak C (191.9 eV) is relatively pronounced in S1. The origin of this peak is not so clear; it had been suggested to be from the inelastic resonance scattering [15] or some kind of surface contaminations [16]. This peak seems to smear out the other peaks, such as peaks A (186.7 eV), B (189.4 eV), and E (210 eV) , which are thought to be the peaks resulting from MgB2 and clearly shown in the other samples. A broad spectral peak between about 195 eV to 205 eV is clearly observable in both S3 and S4.

Fig.4 shows the XANES of Mg K edge for all samples. They were taken at α = 45°. The spectrum of MgO is included for comparisons. Both TFY (Fig.4 (a)) and TEY spectra (Fig.4 (b)) are shown. MgO contains three peaks located at b (1310.8 eV), d (1318 eV) and f (1329.4 eV). The peak a (1304.3 eV), c (1314.3 eV), and e (1325.4 eV) will be the peaks resulting from MgB2 if the major impurities of the sample is MgO and B2O3. One notices that S1 and S2 contain great amount of MgO, while S3 and S4 contain little.

(B) The polarization dependent XANES of S3 and S4

The polarization dependent XANES of S3 and S4 are shown in Fig.5 and Fig.6 respectively for B K edge and Mg K edge. These spectra are further normalized such that the relative spectral height of the high energy portion to the spectral height below the threshold is the same.

In the XANES of B k edge (Fig.5), the change of the threshold peak A from α = 0° to α = 75° is mild, while the increase of the spectral intensity of the peak B is apparent. The broad peak between about 195 eV and 205 eV decreases from α = 0° to α = 75°. These variations reveal the anisotropic nature of the electronic structure since the polarization vector of the probe beam is parallel to the ab plane of MgB2 films at α = 0° and varies toward parallel to the c-axis by increasing the angle α to 75°. Therefore, one probes the boron px,y(σ) at α = 0° and the pz(π) states at α = 90°. The experimental arrangement at α = 90° is not possible due to the finite size of the detector and the block of the outgoing X-ray florescence by the target itself. The measurement at α = 75° is considered to be have most information result from the boron pz(π) states.

The peak A is the sole peak for α = 0° up to about 4 eV from the threshold. Its width is about 0.6 eV. The peak position is at 186.7 eV. The appearance of this peak as well as other features of our spectra at different scan angles is in agreement with that reported by Zhang et al. [17], but in disagreement with the data shown by Cepek et al. [18], in which the peak A was not observed. The peak B increases for the increase of the scan angle α. Since there is no peak B observable at α = 0°, we suggest that the peak B results entirely from the planes parallel to the c axis of the sample, corresponding to the boron pz bonds (π bonds).

    In the XANES of Mg K edge (Fig.6), the anisotropic feature is much less apparent than what appears in the XANES of B Kedge. The existence of empty states starting 1304 eV is apparent. It indicates the metal-like nature of the Mg ions in MgB2.

(C). Comparisons of the XANES and Band structure calculations

 The partial density of states (PDS) for boron px,y(σ) and pz(π) states obtained by various band calculations, either based on the full potential augmented plane wave method (FLAPW) [16,19,20] or based on the other methods [21,22] are similar and always shows that there are two degenerated states in the boron σ band crossover the Fermi level (without core hole and will be referred to as the ground-state PDS). We compare our spectra with the ground-state PDS calculated by the augmented plane waves plus local orbitals (APW+lo) method, encoded in the WIEN 2K software [23], up to the energy 30 eV above the Fermi level in Fig.7 and Fig.8. Lattice parameters of a = 3.089 Å and c = 3.522 Å, and muffin-tin radii of 2.91 au for Mg and 1.68 au for B were used in the calculation. For clarity only the spectra of 0° (Fig.7) and 75° (Fig.8) of S4 are shown for comparisons. The corresponding binding energies measured from the Fermi level are given in the lower scale for both figures. The Fermi level measured from 1s core level of both the XANES of B K edge and Mg K edge is taken to be the inflection point of the threshold peak (peak A for boron and peak a for magnesium). This assignment is in accordance with the X-ray absorption and X-ray emission study [24] for B K edge in MgB2.

Fig.7 shows the combination drawing of the XANES of both B K edge and Mg K edge at α = 0° and the PDS with px,y characters in the band structure calculations. The agreements between experiment and theory are very well, although a small energy shift is visible for the XANES of Mg K edge. This spectral shift toward the low energy can be caused by the core-hole effect [25] since Mg in MgB2 is a partially ionic atom, the interaction of the core-hole potential in the X-ray absorption process can be effective due to less electron screening of the Mg ions.

The peak A thus indicates the narrow hole states in the B σ(2px,y) bands lie just above the Fermi level in the valence band. This interpretation is in accordance with recent reports that showed the threshold peak intensity (peak A in our case) in the boron K- edge X-ray absorption spectra of Mg1-xAlxB2 decreased with the increase of Al doping up to about x = 0.3, consistent with hole filling of the σ band crossover the Fermi level [26, 27]. The hole filling caused the detriment of the superconductivity. Our assignment is also in agreement with the recent report by Klie et al. in using EELS [28], in which the peak A was also demonstrated to have px,y character.

The metal-like PDS with px,y characters is revealed by the XANES of Mg K edge. The empty band with px,y characters around the binding energy of about 13 eV is indicated by the broad peaks in the XANES of B K edge ( around 198.5 eV) and Mg K edge (around 1314 eV). Both the peak E and the peak d indicate the band around the binding energy of 22.5 eV. It is clearly shown in this combination drawing that these bands are mixing bands of boron states and magnesium states in MgB2.

Fig.8 shows the combination drawing of the XANES of both B K edge and Mg K edge at α = 75° and the PDS with pz characters in the band structure calculations. The agreements between experiment and theory are good even though the spectra contain some contribution from the states with px,y characters at α = 75° (rather than α = 90°). One specially notes that the peak B reveals the PDS with pz characters in the band calculations around the binding energy of 3.0 eV. The threshold peak indicates there are hole states near the Fermi level. The appearance of this spectral peak to be shaper than the prediction of the band structure calculation could be due to the contribution from the px,y states. The spectral energy shift is also shown in the XANES of Mg K edge which again is due to the core-hole effect caused by the less electron screening of the Mg ions.

 

IV. Conclusions

 In this work we show that the X-ray absorption near-edge spectrum (XANES) of MgB2 provides information of the electron structure of the empty states above the Fermi level, although the sample may have some impurities, namely B2O3 and MgO. The threshold peak of the XANES of B K edge (peak A in Fig.2 and Fig.6) indicates the narrow hole states in the B σ(2px,y) bands lie just above the Fermi level. The second peak of the XANES of B K edge (peak B in Fig.2 and Fig.7) indicates empty states with pz characters about 3.0 eV about Fermi level. The experimental XANES of Mg K edge results from the empty states with Mg p characters can be identified by comparing the experimental spectrum with the band structure calculations. A small core-hole effect resulted from the less electron screening of the Mg ions in MgB2 can be identified. There are empty states, as appear in the threshold spectral structure, extended from the Fermi level indicates the metal characters of Mg in MgB2. Little anisotropic characteristics between Px,y and pz can be identified from the polarization-dependent measurements of the XANES of Mg K edge.

 

Acknowledgement

 

This work is partially supported by the National Science Council of the Republic of China under grant No. NSC 92-2112-M-003-012. The full support of the staff at the National Synchrotron Radiation Research Center (NSRRC) is greatly appreciated. The work at Penn State is supported in part by ONR under grant No. N00014-00-1-0294 (Xi) and by NSF under grant No. DMR-0306746 (Xi).

 

 

 

 

 

 

 

 

 

 

 

 

References:

1.     J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani, and J. Akimitsu, Nature 410, 63-64 (2001)

2.     J. M. An, and W. E. Pickett, Phys. Rev. Lett. 86, 4366-4369 (2001)

3.     I. T. Belash, A. D. Bronnikov, O. V. Zharikov, and A. V. Palnichenko, Solid State Commun. 64, 1445 (1987)

4.     I. R. Shein, N. I. Medvedeva, and A. L. Ivanovskiĭ, Phys.of the Solid State 43, 2213 (2001)

5.     K. D. Belashchenko, M. van Schilfgaarde, and V. P. Antropov, Phys. Rev. B64, 092503 (2001)

6.     M. J. Mehl, D.A. Papaconstantopoulos, and D. J. Singh, Phys, Rev. B64, 140509(R) (2001)

7.     K. Kokko, V. Kulmala, and J. A. Leiro, Phys. Rev. B66, 165114 (2002)

8.     Hyoung Joon Choi, David Roundy, Hong Sun, Marvin L. Cohen & Steven G. Loule, Nature 418, 758 (2002)

9.     H. Rosner, J. M. An, W. E. Pickett, and S. L. Dresheler, cond-mat/0203030 (2002)

10.  W. N. Kang, Hyeong-Jin Kim, Eun-Mi Choi, C. U. Jung, Sung-Ik Lee, Science 292, 1521(2001)

11.  Xianhui Zeng, Alexej V. Pogrebyakov, Armen Kotcharov, James Jones, X. X. Xi, Eric M. Lysczek, Joan M. Redwing, Shengyong Xu, Qi Li, James Lettieri, Darrell G. Schlom, Wei Tian, Xiaoqing Pan and Zi-Kui Liu, Natural Materials 1, 1 (2002)

12.  Liang-Yann Wu, Chu-Nan Chang, Chi-Hsiang, Bin-Jou Huang, Hong-Chang Yang, and Jin-Ming Chen, Chinese J. of Phy. 42, 534(2004)

13.  13. S. -C. Chung, C. –I. Chen, P. –C. Tseng, H. –f. Lin, T. –E. Dann, Y. –F. Song, L. –R. Huang, C. –C. Chen, J. –M. Chang, K. –L. Tsang, and C. –N. Chang, Rev. Sci. Instrum. 66, 1655 (1995)

14.  L. Tröger, D. Arvanitis, K. Baberschke, H. Michaelis, U. Grimm, and E. Zschech, Phys. Rev. B46, 3283(1992)

15.  T. A. Callcott, L. Lin, G. T. Woods, G. P. Zhang, J. R. Thompson, M. Paranthaman, and D. L. Ederer, Phys. Rev. B64, 132504(2001)

16.  Y. Zhu, A. R. Moodenbaugh, G. Schneider, J. W. Davenport, T. Vogt, Q. Li, G.. Gu, D. A. Fischer, and J. Tafto, Phys. Rev. Lett. 88, 247002(2002)

17.  G. P. Zhang, G. S. Chang, T. A. Callcott, D. L. Ederer, W. N. Kang, Eun-Mi Choi, Hyeong-Jin Kim, and Sung-Ik Lee, Phys. Rev. B67, 174519 (2003)

18.  C. Cepek, R. Macovez, M. Sancrotti, L. Petacci, R. Larciprete, S. Lizzit, and A. Goldoni, App. Phys. Lett. 85, 976(2004)

19.  N. Jiang, B. Jiang, J. C. H. Spence, R. C. Yu, S. C. Li, and C. Q. Jin, Phys. Rev. B66, 172502-1(2002)

20.  X. Kong, Y. Q. Wang, H. Li, X. F. Duan, R. C. Yu, S. C. Li, F. Y. Li and C. Q. Jin, App. Phys.Lett. 80, 778 (2002)

21.  J. A. Leiro, K. Kokko, and V. Kulmala, J. of Alloys and Compd. 362, 139(2004).

22.  I. R. Shein, N. I. Medvedeva, and A. L. Ivanovskiĭ, Physics of the solid state, 43, 2213(2001)

23.  K. Schwarz, P. Blaha, and G. K. H. Madsen, Comput. Phys. Commmun. 126, 71(2002)

24.  Jin Nakamura, Nobuyoshi Yamada, Kazuhiko Kuroki, Thomas A. Callcott, David L. Ederer, Jonathan D. Denlinger, and Rupert C. C. Perera, Phys. Rev. B64, 174504(2001)

25.  F. de Groot, Chem. Rev. 101, 1779(2001)

26.  K. –D. Tsuei, H. –J. Lin, L. –C. Lin, T. –Y. Hou, H. H. Hsieh, C. T. Chen, N. L. Saini, A. Bianconi, and A. Saccone, Int. J. Mod. Phys. B16, 1619(2002)

27.  H. D. Yang, H. L. Liu, J. –Y. Lin, M. X. Kuo, P. L. Ho, J. M. Chen, C. U. Jung, Min-Seok Park, and Sung-Ik Lee, Phys. Rev. B68, 092506(2003)

28.  R. F. Klie, Y. Zhu, G. Schneider, and J. Tafto, App. Phys. Lett. 82, 4316(2003)

 

 

Figure Captions

Fig.1. The temperature dependent resistances of the samples: the right scales are for S1, S2, and S3, and the left scales are for S4.

Fig.2. Top view of the experimental geometry. A linear polarized photon beam with frequency νin is incident on the sample and the frequency of the outgoing X-ray fluorescence is νout. n is the surface normal(c-axis). E is the polarization of the incident beam.

Fig.3. The XANES of B K edge of the samples from (a) TFY measurement and (b) TEY measurement

Fig.4. The XANES of Mg K edge of the samples from (a) TFY measurement and (b) TEY measurement

Fig.5. The polarization dependent XANES of B K edge of the c-axis orientation samples S3 and S4 at α = 0°, 45°, and 75°. The spectra have been normalized such that the relative spectral height at high energy portion (around 217 eV) to that below the threshold is the same.

Fig.6. Same as Fig.5 except for the XANES of Mg K edge.

Fig.7. The combination drawing of the XANES of both B K edge and Mg K edge at α = 0° and the PDS with px,y characters in the band structure calculations by using WIEN 2K code.

Fig.8. Same as Fig.7 except for α = 75°

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 








 

 

 

 

 

Characteristics of a Bandpass Microwave Filter Fabricated by Using MgB2 films

 

M. Y. Chang1, C. N. Chang1, C. H. Hsiech1, H. K. Zeng2, J. Y. Juang2, and Y. S. Gou 2

 

1. Department of Physics, National Taiwan Normal University, Taipei, Taiwan, R. O. C.

2. Department of Electrophysics, National Chiao Tung University, Hsinchu, Taiwan, R. O. C.

 

 

 

Abstracts

 

A four-U hairpin microwave filter with the central frequency of 6.1 GHz has been fabricated by using MgB2 films. The method of UV lithography and Ar ion milling was used to form the desired pattern of the filter. The band width is 1.6 GHz. The superconducting transition temperature (Tc) of the MgB2 film is about 25 K.

 The network analysis indicated that this filter had insertion loss of about 0.1 dB at 5 K and revealed the sharp skirt property. The values of the central frequency, the bandwidth, and the skirt property are very close to those obtained from the computer simulation of a computer code ADS 2003A. The insertion loss deviates a little from the simulation value of 0.03 dB, probably due to the impurities in the film. The decrease of the insertion loss versus the reduced temperature T/Tc was observed to be slower than that of a hairpin filter with six U shape resonators made by a high temperature superconducting film of YBCO.

 

 

 

 

 

 

 

 

 

 

 

 

 

  Magnesium dibroride (MgB2) was found in 2001 [1] to be a superconductor with the highest transition temperature (Tc = 39K) among the inter-metallic compound superconductors. Since then MgB2 attracted the attentions of many scientists not only due to the fundamental scientific interest but also due to potential applications in low temperature electronics. It was found that microwave devices made by MgB2 film can be operated for the input power up to 10 dBm [2]. The nonlinear microwave response of MgB2 films had also been investigated and found that for sufficiently low reduced temperature T/Tc the pair-breaking current density in MgB2 film rivals that in YBa2Cu3O7-δ (YBCO) [3]. Although there were many microwave devices using YBCO films being fabricated and studied [4], not very many using MgB2 were found in the literatures.

  In this letter we report the characteristics of a bandpass microwave filter fabricated by using MgB2 film. The MgB2 film was made by sputtering a target containing a mixture of MgB2 (commercial powders) and Mg metal onto a Al2O3 (1ī02) substrate in a dimension of 5mm X 5mm. This precursor film then was post annealed under Ar at about 800C. Detail preparation procedure of the MgB2 film was described elsewhere [5]. The Tc and the root mean square of the surface roughness are measured to be respectively 25 K and 30 nm. The thickness of the film is about 400nm.

  The method of UV lithography and Ar ion milling was used to form the desired pattern of the filter. A thin gold film about 160nm was deposited on top of the MgB2 film to protect it from being deteriorated by the wet chemicals in the development process, since MgB2 film is sensitive to moisture.

  By fully use the total size of the as grown film of MgB2, we pattern the film into four U shape hairpin resonators as shown in Fig.1. L1 and L2 are taken to be respectively 1.5 mm and 1.35 mm. The width of the transmission line W1 is taken to be 0.1 mm. The resonator R1 and the resonator R2 were designed to be gap coupling to have a better resonant effect, and the gap W2 was adjusted by using the computer code ADS 2003A [6] to maximize the resonant peaks around the central frequency of about 6.0 GHz. The gap W2 was determined to be 0.05mm. The same gap size was chosen for R3 and R4. For better resonance, a small transmission line (L3) with length of 0.3 mm was specifically added to have cross coupling [7] between the two set of resonators R1, R2 and R3, R4, as well as to compensate the loss due to the gap couplings. To minimize the effect of parasitic capacitor at high frequency, we design a small transmission line L4 connected to the ground between R2 and R3, and two open stubs L5. The simulation results of S11 and S21 of the four resonator filter are shown in Fig.2. S11 and S21 indicate the return loss and the insertion loss of the filter, respectively. This filter is expected to be a compact filter with sharp skirt property. The centre frequency is at 6.10 GHz with a wideband (the bandwidth is 1.6 GHz). It may be suitable for use in wireless stations of 802.11a.

  A network analyzer Hewlett-Packard N5230A [8] was employed to measure S11 and S21. A liquid helium Dewar was used to cool down the filter. Comparison between the measurement and the simulation is shown in Fig.2. The measure curves were done at 5 K. They are close to each other, although the measured insertion loss is about 0.10 dB, which is somewhat deviated from the simulation value of 0.03 dB. It could be due to the impurities in our films. The X-ray absorption spectra indicated that the film contained MgO and B2O3 [5]. The surface resistance could be higher than expected. The temperature variation curves of S21 had been made. The insertion loss versus reduced temperature T/Tc is shown in Fig.3. The drop of the insertion loss is not so sharp as compared to a six-U hairpin filter made by YBCO [9]. We note that the configuration in [9] is a parallel U strip-lines arrangement between the input and output and is different from ours. The slow drop of the insertion loss may be also due to the quality of our MgB2 films.

   In summary, we demonstrate that the MgB2 film can be used to fabricate a filter by using photolithographic method in combination of Ar ion milling in a compact size of about 5mm X 5mm. The experimental characteristics are close to the simulation ones.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

References

1.     J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani, and J. Akimitsu, Nature 410, 63 (2001)

2.     M. V. Jacob, J. Mazierska, Y. Harada, T. Takahashi, and M. Yoshizawa, IEEE Transactions on Applied Superconductivity, Vol. 15, 3317-3320 (2005)

3.     James C. Booth, K. T. Leong, Sang Young Lee, J. H. Lee, B. Oh, H. N. Lee, and S. H. Moon, Supercond. Sci. Technol. 16, 1518-1522 (2003).

4.     S. Ohshima, Supercond. Sci. Technol. 13, 103-108 (2000).

5.     C. H. Hsieh, C. H. Chang, C. N. Chang, U. C. Sou, H. S. Sheu, H. C. Hsu, and H. C. Yang, Solid Commun. 137, 97-100 (2006).

6.     “ Advanced Design System 2003A”, Agilent Techonologies, U. S. A.

7.     Tatsuya Tsujiguchi, Haruo Matsumoto. And Toshio Nishikawa, IEEE Trans. Microwave Theory Tech. 49, 879-885 (2001).

8.     David Ballo, “Network Analyzer Basics”1998, Hewlett-Packard Company, Cal. U. S. A..

9.     H. K. Zeng, A. Hsiao, W. H. Hsu, S. W. Wu, J. Y. Lin. K. H. Wu, J. Y. Juang, T. M. Uen, and Y. S. Gou, unpublished.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure Captions

Fig.1. The patterns of the four-U hairpin microwave filter. Numbers in the diagram indicate the optimum ones found by simulation.

Fig.2. The curves the return loss S11 and the insertion loss S21. Both results from simulation (solid curves) and measurement ( dot curves) are shown. The measured ones were obtained at 5 K.

Fig.3. The insertion loss versus the reduced temperature T/Tc. The data of Ref. 9 are also shown for comparison.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig.2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig.3

 

 

 

 

PACS: 47.61.Fg

 

Imprint of micrometer-scale structures by electro-deposition

 

C. Y. Wu1, C. H. Hsieh2, M. F, Wu1, C. N. Chang2

 

1. Institute of Electro-optical Science and Technology, National Taiwan Normal University, Taipei, Taiwan, R.O.C.

2. Department of Physics, National Taiwan Normal University, Taipei, Taiwan, R.O.C.

 

 

 

Abstract

 

   The imprint of micrometer-scale structures was successfully done by using electrolysis to deposit copper on the substrate from the mold. The mold is a series of rectangular patterns made by UV lithography. The smallest one is 3.15 μm with a depth of 1.3μm. The aspect ratio of its imprint pattern is about 518 nm, and the broadening of the line is about 0.15μm. 0.1 M of K2SO4 solution was used as the electrolyte solution in the electrolysis process.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  Several methods of imprint of micrometer- and nanometer-scale structure have been studied, such as hot embossing imprint lithography [1], UV-cured imprint lithography [2], micro-contact imprint [3], and laser assisted direct imprint [4]. They are novel methods in imprint of micrometer- and nanometer-scale structure, although capital investment may be inevitable to have the current production line of electronic devices redesigned and built. We here report a low-cost method of imprint by using electro-deposition techniques.

  The mold of the micrometer-scale patterns was made by UV photolithography, which was on top of a copper layer about 500 nm thick. Below the copper layer a thin gold film about 500nm was deposited to serve as the contact of the electrode. As shown in Fig.1, the substrate with a thin layer of gold about 100 nm thick to serve as the contact of the other electrode is placed on top of the mold, 0.1 M electrolytic solution of K2SO4 is sandwiched in between the substrate and the mold. A gentle force supplied by a micrometer adapted to a flat plate to press the substrate toward the mold and the electrolytic solution was properly fit into the trough that is the pattern of the mold required to be imprinted. Carefulness of doing this process had been taken since our pressing tool was homemade and it was not easy to level the substrate and the mold. A power supply that can supply power up to 3 A × 10 V with the smallest adjustable power of 1 mA × 10 mV was used in this electro-deposition process. The power supply has timer devised to preset the period of power supplying. In Fig.1, the circuitry of the arrangement of electro-deposition is also shown.

   Fig.2(a) shows the patterns of the mold (taken after the electro-deposition), they are about 3, 4, 5, 6, 7, 10, and 20 micrometer rectangular troughs with depth of 1.3 μm. The one with width about 3 μm has a small outward feather accidentally formed in the process of making the mold. Fig.2(b) shows the imprinted patterns finished by the aforementioned electro-deposition method. An elemental scan of the imprint patterns by using an energy dispersive x-ray spectrometer (EDX) is shown in Fig.3. It is demonstrated that the patterns formed in the method of electro-deposition is indeed copper that is separated by the electrode film of gold.

We examined the feature of the imprinted lines as well as those of the patterns of the mold by atomic force microscopy (AFM). A typical topology of the one with the line width about 3 μm is shown in Fig.4. The line width of the pattern was demonstrated to be reproducible. A small broadening of the line width appears in the pattern having line width less than about 6 μm. The depth of the line varies from about 400 nm to about 500 nm, shallower for the one with wider line width. The total amount of copper in the trough of the mold seems to be transferred to the substrate. The probes of EDX at the trough indeed show no sign of copper. We speculate that a higher aspect ratio can be formed on the substrate by using thicker layer of copper in the mold. The mold may be reused by a reverse process of the above electrolysis, this time a layer of copper is on the substrate and the trough of mold is empty of copper.

   This method may also be applicable for imprint of nanometer-scale structures. The mold can be made by electron lithography. However, cautions must be made for the tuning of the electrolysis current and time. It is easy to have excess deposition of copper and deteriorate the copper on the substrate during removing the mold away from it. We suggest using more sophisticate instrument, for instance a linear sweep voltammetry or a cyclic voltammetry [5], to study the electro-chemical process and find a suitable electrolysis conditions for the imprint of nanometer-scale structures.

    We demonstrate here that the electro-deposition method can successfully imprint a micrometer-scale structure by using a general power supply with the electrolytic solutions of K2SO4. More controllable processes can be developed by using more sophisticated instruments in this imprint technique, and then this technique can be easily extended to imprint of nanometer- scale structures.    

 

References:

1. S. Y. Choui, P. R. Krauss, and P. J. Renstrom, App. Phys. Lett. 67, 3114-3116 (1995).

2. Y. Hirai, M. Fujiwara, T. Okuno, Y. Tanaka, M. Endo, S. Irie, K. Nakagawa, and M. Sasago, J. of Vac. Sci. and Tech. 19, 2811-2815 (2001).

3. S. Brittain, K. Paul. X. M. Zhao, and G. Whitesides, Phys. World, 5, 31-36(1998).

4. Stephen Y. Chou, Chris Keimel, and Jian Gu, Nature, 417, 835-837 (2002).

5. M. Noel, K. I. Vasu, ”Cyclic Voltammetry and the Fronties of Electrochemistry”, Aspect Publications, London, 1990.

 

Figure Captions

 

1. (a) Schematic diagram of the arrangement of the substrate and the mold before electro-deposition, (b) Gentle pressing the substrate toward the mold, (c) The substrate removed away from the mold after the electro-deposition, (d) The circuitry of the electrolysis.

2. The microscopic images of (a) the mold after electro-deposition, and (b) the imprinted patterns of the substrate after electro-deposition.

3. Elements of the imprinted patterns on the substrate scanned by EDX; the bold solid curves are copper and the light curves are gold.

4. The typical topology of one of the imprinted pattern line obtained by AFM.

 

 

 

 

 

 

 

 

(a)
 
 


 

 

 


 

 

 

(b)
 
 


 

 

 

 

(c)
 
 


 

 

 

 

 

 

 

 

 

 

 

(d)
 
 

 

 

 

 

 

 

 

 

 

 


 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig.3