Editorial Board Supported by NSFC

Honorary Editor General ZHOU GuangZhao (Zhou Guang Zhao) Editor General ZHU ZuoYan Institute of Hydrobiology, CAS Editor-in-Chief LI Wei Beihang University Executive Associate Editor-in-Chief WANG DongMing Centre National de la Recherche Scientifique

Associate Editors-in-Chief

GUO Lei Academy of Mathematics and Systems Science, CAS HUANG Ru Peking University QIN YuWen National Natural Science Foundation of China SUN ZengQi Tsinghua University YOU XiaoHu Southeast University ZHAO Wei University of Macau ZHAO QinPing Beihang University

Members

CHEN JianEr LI Joshua LeWei TSAI WeiTek Texas A&M University University of Electronic Science and Arizona State University Technology of China DU Richard LiMin WANG Ji Voxeasy Institute of Technology LIU DeRong National University of Defense Technology Institute of Automation, CAS GAO Wen WANG JiangZhou Peking University LIN HuiMin University of Kent Institute of Software, CAS GE ShuZhi Sam WANG Long National University of Singapore LIN ZongLi Peking University GUO GuangCan University of Virginia

University of Science and Technology of WU YiRong LONG KePing Institute of Electronics, CAS China University of Science and Technology Beijing HAN WenBao XIE WeiXin Shenzhen University PLA Information Engineering University LU Jian Nanjing University HE JiFeng XU Jun Tsinghua University East China Normal University MEI Hong

Peking University XU Ke HU WeiWu Beihang University Institute of Computing Technology, CAS MENG LuoMing

Beijing University of Posts and YIN QinYe HU ZhanYi Telecommunications Xi’an Jiaotong University

Institute of Automation, CAS PENG LianMao Peking University YING MingSheng IDA Tetsuo Tsinghua University

University of Tsukuba PENG QunSheng Zhejiang University ZHA HongBin JI YueFeng Peking University Beijing University of Posts and Telecommu- SHEN ChangXiang nications Computing Technology Institute of China ZHANG HuanGuo Navy Wuhan University JIN Hai Huazhong University of Science and SUN JiaGuang ZHOU Dian Technology Tsinghua University The University of Texas at Dallas

JIN YaQiu TANG ZhiMin ZHOU ZhiHua Fudan University Institute of Computing Technology, CAS Nanjing University

JING ZhongLiang TIAN Jie ZHUANG YueTing Shanghai Jiao Tong University Institute of Automation, CAS Zhejiang University

Editorial Staff SONG Fei FENG Jing

SCIENCE CHINA Information Sciences

. RESEARCH PAPER . January 2012 Vol. 55 No. 1: 64–71 Special Focus doi: 10.1007/s11432-011-4513-3

Terahertz detectors based on superconducting hot electron bolometers

JIANG Yi, JIN BiaoBing, XU WeiWei, KANG Lin, CHEN Jian∗ &WUPeiHeng

Research Institute of Superconductor Electronics, Nanjing University, Nanjing 210093,China

Received August 11, 2011; accepted August 31, 2011

Abstract Low noise terahertz (THz) heterodyne detectors based on superconducting niobium nitride (NbN) hot electron bolometers (HEBs) have been studied. The HEB consists of a planar antenna and an NbN bridge connecting across the antenna’s inner terminals on a high-resistivity Si substrate. The double sideband noise temperatures at 4.2 K without corrections have been characterized from 0.65 to 3.1 THz. The excess quantum noise factor β of about 4 has been obtained, which agrees well with the calculated value. Allan variance of the HEB has been measured, and Allan time up to 20 s is obtained using a microwave feedback method. Also, the intermediate frequency gain bandwidth (GBW) was measured using two different methods, resulting in same GBW value of about 3.5 GHz.

Keywords terahertz heterodyne detector, hot electron bolometer, noise temperature, stability, intermediate frequency gain bandwidth

Citation Jiang Y, Jin B B, Xu W W, et al. Terahertz detectors based on superconducting hot electron bolome- ters. Sci China Inf Sci, 2012, 55: 64–71, doi: 10.1007/s11432-011-4513-3

1 Introduction

Heterodyne detectors based on superconducting hot electron bolometers (HEBs) combine excellent noise performance and low local oscillator (LO) power requirment at terahertz (THz) waveband. A key advan- tage of the HEB is that it does not suffer from upper frequency limit set by the superconductor’s energy gap [1]. So at frequency higher than the energy gap frequency, which is about 1.4 THz for superconducting niobium nitride (NbN) film, the performance of the HEB should be better than that of superconductor- insulator-superconductor (SIS) heterodyne detectors [2]. This makes the HEBs highly attractive for both ground-based and space-based telescopes for astronomy [3]. The Herschel-HIFI, which is a part of the Herschel Space Observatory, used the HEBs with the double sideband (DSB) noise temperature (TN ) of about 10×hf/kB at f =1.41 − 1.92 THz [4], where h is the Planck constant, kB is the Boltzmann constant and f is the operating frequency. It can be expected that the HEBs will be used widely at THz waveband in the near future. However, according to the theoretical predictions or obtained results on the SIS or Schottky detectors, there is still some room for improvements as far as HEB’s noise performance, or intermediate frequency (IF) gain bandwidth (GBW) and stability are concerned. Here, we report our detailed studies on the low noise detectors based on NbN HEBs at THz frequency waveband.

∗Corresponding author (email: [email protected])

c Science China Press and Springer-Verlag Berlin Heidelberg 2012 info.scichina.com www.springerlink.com Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1 65

Figure 1 The photo of the HEB and the NbN bridge in the center of the planar antenna.

2 Experimental 2.1 HEB chip

The HEB chip consists of a superconducting bridge made from an ultra-thin NbN film and a logarithmic- spiral planar antenna with frequency independent impedance. The NbN bridge 4 μm in width, 0.4 μm in length and 4 nm in thickness is connected to the planar antenna to efficiently couple the THz signal as shown in Figure 1. The antenna is designed to work at 0.4–4 THz. The ultra-thin superconducting NbN film is deposited by DC magnetron sputtering on high-resistivity silicon (Si) substrate in Ar+N2 gas mixture while keeping the substrate at room temperature (RT) [5]. A root-mean-square roughness of approximately 0.42 nm is obtained for a 4.5 nm thick film over an area of 25 μm2. The critical current density of about 1.5×106 A/cm2 at 4.2 K and critical temperature of about 9 K are obtained. After depositing the ultra-thin film, it is covered with photoresist, and two square openings are positioned on the photoresist by electron beam lithography which determines the length of the bridge. In order to prevent degradation of the superconductivity of the bridge, an additional NbN film of 10 nm thickness is deposited on the opened NbN ultra-thin film as a buffer. Then a 50 nm thick gold film is deposited and the bridge’s width is defined by photolithography and reactive ion etching. At the end a complementary logarithmic-spiral antenna made of gold is connected to the two poles. The details of the fabrication process are reported in [6].

2.2 Experimental setup

We use a quasi-optical setup to couple THz signal from the source to the HEB. The HEB chip is glued to the center of the back of a hyper-hemispherical lens made of high-resistivity Si. Lenses with and without anti-reflection (AR) coatings are used in different setups of measurements. The lens is fixed in an oxygen free copper fixture which is thermally sunk to the 4.2 K cold plate of a liquid helium cryostat. We use an optically pumped far-infrared (FIR) gas laser (FIRL 100 from Edinburgh Instruments Ltd.) at 1.6, 2.5 and 3.1 THz or a microwave synthesizer with its multipliers at 0.65 THz as the LO sources. A mylar film with a thickness of 15 μm is used for the beam splitter (BS) and the mylar film with a thickness of 36 μm is used for the cryostat window. Two black polyethylene films with thickness of about 0.1 mm [7] and one G-110 Zitex, which is a porous polytetrafluoroethylene (PTFE) film1) ,areusedintheTHz input hole at 77 K thermal shielding frame as infrared (IR) filters. To reduce the environment noise, all the equipments except the laser and computer are placed in an RF shielding room. The cryostat is put on an optical table with anti-vibration structures. Schematic of the experimental setup is shown in Figure 2. We use an adjustable DC voltage source to bias the HEB. The bias voltage and current can be collected to the computer by a digital multimeter. The IF signal from the HEB is connected to a DC block, let to pass through an isolator, and then amplified by a cryogenic low noise amplifier (LNA) and an RT amplifier. In the noise temperature and stability measurements, we use a cryogenic LNA with gain of 30 dB, noise temperature of 6 K operating at 15 K

1) http://www.norton-films.com/zitexg-filter-membranes.aspx 66 Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1

Figure 2 Schematics of the experimental setup. and the bandwidth of 1.3–1.7 GHz, and a home-made RT amplifier with gain of 50 dB and the bandwidth of 1–2 GHz. The IF signal goes through a band-pass (BP) filter, which is centered at 1.5 GHz and has a bandwidth of 100 MHz, is measured by a microwave power detector, and then collected by a computer. In the IF GBW measurements, we use a broadband LNA with the bandwidth of 1–12 GHz, the gain of 32 dB and noise temperature of less than 8 K operating at 11 K. The RT amplifier is Agilent 83020A, with a bandwidth of 2–26.5 GHz and gain of 30 dB. The amplified IF signal is analyzed by a spectrum analyzer (Rohde & Schwarz FSP40). Another way to determine the IF GBW is to use the Fourier transform spectrometer (FTS), with chopping hot load and cold load as THz source. The signal from the broadband amplifiers is connected to a power detector and then gathered by a lock-in amplifier to get the interferogram, which is Fourier transformed to obtain IF spectrum. It is similar to the setup of the noise temperature measurement. Allan variance is used to characterize stability of our detectors. According to a lot of our measured results, we use a feedback control loop with a power adjustable microwave source to improve the stability of the system. The power of the microwave source is controlled by the computer to compensate for the varying LO power in real time. The microwave source is Agilent E8257 PSG signal generator; the frequency is adjusted to 17.22 GHz, which is much lower than the LO frequency and can be adjusted easily. The IF output power is measured by a power detector and collected by a digital multimeter, and then fed to a computer to calculate the Allan variance. Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1 67

Figure 3 I-V curves with and without LO power (RF) at frequency of 0.65 THz, as well as DSB noise temperatures for different bias points. The HEB chip is glued on the Si lens with AR coating at 0.65 THz. 36-μm and 15-μmthickMylar films are used as the window and beam splitter, respectively.

3 Results and discussions 3.1 Noise temperature

The unpumped and optimally pumped (LO at 0.65 THz) I-V curves of an NbN HEB working at 4.2 K with 0.65 THz AR coating are shown in Figure 3, together with the TN measured at different bias points. The critical current is 100 μA at 4.2 K and the normal state resistance of the HEB is about 150 Ω. The contact resistance between the electrodes and the bridge is 10 Ω. In order to check HEB’s wideband performance, we measure TN at different LO frequencies and AR conditions with the same HEB chip. We use an improved Y -factor method to obtain TN . The improved method can eliminate the effect of direct detection of the HEB. The details of the discussion of the method are reported in [1]. In brief, we measure the IF output power, PHot and PCold, corresponding to the hot and cold loads, at the same DC bias point to get Y = PHot/PCold. The uncorrected TN can be calculated by following expression [8]:

THot − YTCold TN, = , rec Y − 1

TN at LO frequencies of 0.65, 1.6, 2.5, 3.1 THz have been measured. They are shown in Table 1. Using AR coating, we can reduce the measured TN . The details of these measurements are reported in [6]. The lowest TN are 698 K at 0.65 THz, 904 K at 1.6 THz, 1026 K at 2.5 THz and 1386 K at 3.1 THz, and about 10 times that of quantum limit 10×hf/kB, as shown in Figure 4. In general, TN should increase with the increasing frequency, but the experimental results show that some of TN at higher frequencies are lower or almost the same as that at lower frequencies. To understand this phenomenon, we have tried to fit the results with the quantum noise theory [9]. Under the reasonable parameters, the excess quantum noise factor (β) of about 4 (=6 dB) can be obtained as shown in Figure 5, which is very close to the calculated value of 3.75 from the theortical model. Also, this value is same as that obtained under same HEB chip but different AR coating conditions [1]. That means one HEB chip has only one β value, which is independent of the frequency and other system’s conditions. Smaller bridge with lower contact resistance between the electrodes and bridge, which is a big challenge to the fabrication process, will give a smaller β value theoretically. The non-linear frequency dependence of the TN is due to the losses of different parts, such as the BS, window, and AR coating, in the system. Using this rule, we can design 68 Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1

Table 1 Noise temperatures at different frequencies for different AR coating conditions (K)

LO frequency (THz) TN 0.65 1.6 2.5 3.1 0.65 THz AR 698 1057 1462 1386 2.5 THz AR 1278 904 1026 1410 No AR 1028 1165 1396 1734

Figure 4 Frequency dependence of measured lowest Figure 5 Frequency dependence of lowest noise tem- noise temperatures for an HEB detecter. peratures for a Parylene C AR coating on the Si lens with thickness of 70 μm (0.65 THz). β=4 was obtained and was in good agreement with the calculated value (lines). them to get the lowest TN at a special frequency or a special frequency band for a given HEB chip as shown in Figure 5. So our results demonstrate that the quantum noise theory can be used to explain the behaviors of our HEB detectors. We can design and fabricate the HEB chip with a smaller β value according to the theortical model to get a lower TN , which implies that in the near future HEB detectors can work up to 10 THz with a TN lower than 10×hf/kB.

3.2 Stability

As the optimal observation time is determined by the stability of the detectors and the stability of the HEB is one of the main issues compared with that of the SIS or Schottky heterodyne detectors, it is necessary to study it in detail and find some methods to improve it. Mostly, Allan variance measurements have been used widely to characterize the stability of various equipments such as THz heterodyne detectors [10]. For any practical system with low frequency drift and/or 1/f noise, the signal to noise ratio cannot be improved by integration over a period longer than the Allan time (also known as Allan-stability time). So we should investigate and extend the Allan time as long as possible for the THz applications. Using a feedback loop for the microwave output power via a computer, as shown in Figure 2, the stability has been improved [11]. The frequency of the microwave source is much lower than the LO frequency, so it can be adjusted easily. Figure 6 shows the Allan variance for an HEB detector with and without the feedback loop of the microwave injection. As shown in Figure 6, Allan time of about 20 s can be obtained with the feedback loop, in contrast to Allan time of about 1 s without the feedback loop. In this way, smaller bridge, which is not so easy to be fabricated but will give a smaller β value and lower TN theoretically, will lead the system to be less stable [12]. So there has to be a trade-off between TN and stability for the detectors with high performance.

3.3 IF GBW

We use two methods to measure the IF GBW. One way is to use two THz sources, with one source fixed at 0.65 THz as THz signal, and the other serving as LO tuned around 0.65 THz to determine the IF GBW. The IF GBW is measured at two bias points, one at the optimal bias point with lowest TN (shown Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1 69

2 Figure 6 Allan variance σA(T ) of the IF output power Figure 7 The IF output spectrum of a measured of the HEB as a function of the average period for the THz source at 0.6513 THz, with the IF linewidth system with and without the microwave feedback loop. about 10 Hz.

Figure 8 Measured IF GBW of the mixer using two THz Figure 9 Measured IF GBW using a Fourier sources, one as the signal fixed at 0.65 THz, and the other transform spectrometer. tuned around 0.65 THz as LO. in Figure 3) and the other at about half the bias voltage of the optimal bias point. During experiment it is found that the detector’s received THz power is related to the optical path length from the HEB to 1 λ the THz source horn, and the received power changes periodically with the period of 4 optical path. According to the isothermal method [13], if the HEB’s bias point does not change, then the absorbed power is invariable. By adjusting the power of LO to keep the bias point at a constant value, we can ensure the received LO power to be steady. The spectrum of the amplified IF output is shown in Figure 7. The resulting resolution of the mixing system is about 10 Hz. Figure 8 shows output power versus IF. The obtained IF GBW is 3.5 GHz. IF signal can be found to be up to 12 GHz, but with a small gain. Another way to measure the IF GBW is using a Fourier transform spectrometer (FTS), which is a fast way to get results. The measured result is shown in Figure 9. The upper sideband and lower sideband can be differentiated to about 3.5 GHz, which is the same as the value obtained by the method. The result has no strictly bilateral symmetry because there are measurement errors. The error is due to optical misalignment and the resolution has exceeded the FTS’s absolute maximum ratings. For example, the blackbody source can not be treated as an ideal extended area blackbody source in the THz range. The IF GBW of 3.5 GHz is still smaller than that of SIS mixers [14], which may be the bottleneck in some applications. As afore-mentioned, larger IF GBW can be obtained using the diffusion-cooling HEBs, 1/2 which needs a smaller bridge with its length shorter than the thermal healing length Lth ≡ π(Dτth) 70 Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1

(D is the electron diffusion coefficient τth is the electron temperature relaxation time, and Lth is about 0.1 μm for NbN) [15]. So by fabricating a smaller HEB bridge to combine the phonon-cooling and the diffusion-cooling together, we can increase the IF GBW of the HEB detectors.

4 Conclusions

In this paper we have introduced HEB detectorsusedinourlaboratory.ThelowestTN at 0.65 THz is 698 K, the IF GBW is 3.5 GHz, and the Allan time is about 1 s and can reach to 20 s by using a microwave feedback loop. The full characterization and analysis shows that it has good performance satisfying the needs for some THz applications. There is still some room to further improve the detectors. For example, we can design an HEB with smaller β value to get a detector with lower TN , wider IF GBW, and a stabler detecting system.

Acknowledgements This work was supported by National Basic Research Program of China (Grant Nos. 2007CB310404, 2011CBA01007), National Natural Science Foundation of China (Grant No. 11173015) and Doctoral Funds of the Ministry of Ed- ucation of China (Grant No. 20090091110039).

References

1 Liang M, Chen J, Kang L, et al. Low noise receivers based on superconducting niobium nitride hot electron bolometer mixers from 0.65 to 3.1 terahertz. IEICE Trans Electron, 2010, E93-C: 473–479 2H¨ubers H W. Terahertz heterodyne receivers. IEEE J Sel Topic Quantum Electron, 2008, 14: 378–391 3 Dochev D, Desmaris V, Meledin D, et al. A technology demonstrator for 1.6–2.0 THz waveguide HEB receiver with a novel mixer layout. J Infrar Millim Terahertz Wave, 2011, 32: 451–465 4 Cherednichenko S, Drakinskiy V, Berg T. Hot-electron bolometer terahertz mixers for the Herschel space observatory. Rev Sci Instr, 2008, 79: 034501 5 Wu P H, Kang L, Chen Y, et al. Fabrication and characterization of NbN/AlN/NbN junction on MgO(001) and AlN/NbN bilayer on MgO(111) substrates. IEEE Trans Appl Supercond, 2005, 15: 209–211 6 Chen J, Liang M, Kang L, et al. Low noise receivers at 1.6 THz and 2.5 THz based on niobium nitride hot electron bolometer mixers. IEEE Trans Appl Supercond, 2009, 19: 278–281 7 McCarthy D E. Black polyethylene as a far-infrared filter. J Opt Soc Am, 1967, 57: 699–670 8 Khosropanah P, Gao J R, Laauwen W M, et al. Low noise NbN hot electron bolometer mixer at 4.3 THz. Appl Phys Lett, 2007, 91: 221111 9 Kollberg E L, Yngvesson K S. Quantum-noise theory for terahertz hot electron bolometer mixers. IEEE Trans Microw Theory Tech, 2006, 54: 2077–2089 10 Hajenius M, Baselmans J J A, Baryshev A, et al. Full characterization and analysis of a terahertz heterodyne receiver based on a NbN hot electron bolometer. J Appl Phys, 2006, 100: 074507 11 Chen J, Jiang Y, Liang M, et al. Stability of superconducting hot electron bolometer receivers. IEEE Trans Appl Supercond, 2011, 21: 667–670 12 Berg T, Cherednichenko S, Drakinskiy V, et al. Stability of HEB receivers at THz frequencies. Proc SPIE, 2004, 5498: 605–614 13 Ekstrom H, Karasik B S, Kollberg E L. Conversion gain and noise of niobium superconducting hot-electron mixers. IEEE Trans Microw Theory Tech, 1995, 43: 938–947 14 Karpov A, Miller D A, Stern J A, et al. Low noise 1 THz SIS mixer for stratospheric observatory: Design and Charac- terization. IEEE Trans Appl Supercond, 2011, 21: 616–619 15 Burke P J, Schoelkopf R J, Prober D E, et al. Mixing and noise in diffusion and phonon cooled superconducting hot-electron bolometers. J Appl Phys, 1999, 85: 1644–1653 Jiang Y, et al. Sci China Inf Sci January 2012 Vol. 55 No. 1 71

JIANG Yi was born in 1986. He CHEN Jian was born in 1963. He is now a Ph.D. candidate in radio received the B.S. and M.Sc. degrees in physics from Research Institute of Su- radio physics from NJU, in 1983 and perconductor Electronics (RISE), Nan- 1986, respectively. From 1986 to 1989, jing University (NJU), Nanjing. His re- he was an assistant in NJU. He received search interests include terahertz spec- DE degree in Department of Electri- trometer, terahertz detectors and tera- cal Engineering from Nagaoka Univer- hertz imaging. sity of Technology (NUT), Japan, in 1992. Then, he was an assistant in Research Institute of Electrical Com- munication (RIEC), Tohoku University, Japan. From 1998, he was an associate professor in RIEC. He has been a professor in NJU since 2003. WU PeiHeng was born in 1939. He He is a senior member of IEEE. His research interest includes graduated from NJU in 1961 majoring superconducting devices and their applications. in physics. He has been a Professor in NJU since 1985 and an academician, Chinese Academy of Sciences (CAS) since 2005. From January 2001 to July 2001, he was a professor in RIEC, To- hoku University, Japan. His research interest includes superconducting elec- tronics, high frequency techniques and their applications.

● Monograph North and South America: Rordam M, Larsen F, Lausten N. An introduc- [email protected] tion to K-theory for C*-algebras. Cambridge: Outside North and South America: Cambridge University Press, 2000. 30－35 [email protected] ● Proceedings Orders and inquiries: Qin H R. The subgroups of a finite order in K(Q). China In: Bass H, Kuku A O, Pedrini C, eds. Algebraic Science China Press K-theory and Its Application. Singapore: World 16 Donghuangchenggen North Street, Beijing 100717, Scientific, 1999. 600－607 China ● Conference proceedings Tel: +86 10 64015683 or +86 10 64034134 Minor H E. Spillways for high velocities. In: Fax: +86 10 64016350 Zurich V E, Minor H E, Hager W H, eds. Pro- North and South America ceedings of International Workshop on Hydrau- lics of Stepped Spillways, Rotterdam, the Springer New York, Inc. Netherlands, 2000. 3―10 Journal Fulfillment, P.O. Box 2485 Secaucus, NJ 07096 USA ● Dissertation Tel: 1-800-SPRINGER or 1-201-348-4033 Liu G X. Classification of finite dimensional basic Hopf algebras and related topics. Dis- Fax: 1-201-348-4505 sertation for the Doctoral Degree. Hangzhou: Email: [email protected] Hangzhou University, 2005. 24－28 Outside North and South America: Springer Distribution Center ● Technical report Phillips N A. The Nested Grid Model. NOAA Customer Service Journals Technical Report NWS22. 1979 Haberstr. 7, 69126 Heidelberg, Germany Tel: +49-6221-345-0, Fax: +49-6221-345-4229 ● Patent Zhang W P. Experiment Apparatus of Diffraction Email: [email protected] Imaging. China Patent, 02290557.X, Cancellations must be received by September 30 to 2003-12-03 take effect at the end of the same year. Changes of address: Allow for six weeks for all ● User manual changes to become effective. All communications Wang D L, Zhu J, Li Z K, et al. User Manual for should include both old and new addresses (with QTKMapper Version 1.6, 1999 postal codes) and should be accompanied by a mailing ● Software label from a recent issue. According to § 4 Sect. 3 of Hemodynamics III: The ups and downs of he- the German Postal Services Data Protection Regula- modynamics. Version 2.2. Orlando: Computer- tions, if a subscriber’s address changes, the German ized Educational Systems. 1993 Federal Post Office can inform the publisher of the new address even if the subscriber has not submitted a ● CD formal application for mail to be forwarded. Subscrib- Anderson S C, Poulsen K B. Anderson’s Elec- ers not in agreement with this procedure may send a tronic Atlas of Hematology. Philadelphia: Lip- pincott Wilkins, 2002 written complaint to Customer Service Journals, Karin Tiks, within 14 days of publication of this issue. ● Electronic version of a journal Microform editions are available from: ProQuest. Fur- Christine M. Plant physiology: plant biology in ther information available at the Genome Era. Science, 2003, 281: 331－ http://www.il.proquest.com/uni 332 [cited Sep. 23, 2003]. Available from: Electronic edition http://www. sciencemag. org/anatmorp.htm An electronic version is available at springer.com/scp Subscription information Production Science China Press ISSN print edition: 1674-733X 16 Donghuangchenggen North Street, Beijing 100717, ISSN electronic edition: 1869-1919 China Volume 55 (12 issues) will appear in 2012 Tel: +86 10 64015683 or +86 10 64034134 Subscription rates Fax: +86 10 64016350 For information on subscription rates please contact: Printed in the People’s Republic of China Customer Service Jointly Published by

China: [email protected] Science China Press and Springer