Electrooptic characterization of modulation-doped field-effect transistors with monolithically-integrated test fixtures

of 8

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
8 pages
0 downs
Electrooptic characterization of modulation-doped field-effect transistors with monolithically-integrated test fixtures
  Optical and Quantum Electronics 28 1996) 867-874 Electrooptic characterization of modulation doped field effect transistors with monolithically integrated test fixtures A. ZENG, M. K. JACKSON Department of Electrical Engineering University of British Columbia 2356 Main Mall Vancouver BC Canada V6T lZ4 M. VAN HOVE, W. DE RAEDT Interuniversity Microelectronics Center Kapeldreef 75 B-3001 Leuven Belgium Received 2 November 1995; revised 18 January; accepted 15 February 1996 The authors report on the fabrication and electrooptic measurement of modulation- doped field-effect transistors MODFETs) monolithically integrated with coplanar stripline fixtures incorporating photoconductive switches. Both lattice-matched devices on InP substrates and pseudomorphic devices on GaAs substrates, as well as two different gate access structures have been characterized. In comparing two devices with identical gate access structures, one finds that the delay time of the pseudomorphic device is almost twice as long as that of the lattice-matched device. Surprisingly, the switching times drain output risetimes) of the two devices are comparable. One also sees that the gate access structure significantly affects the switching in two lattice- matched devices: the switching times for double-gate and single-gate contact devices are 4.2 and 5.2 ps, respectively. In this case, however, no difference was seen in the delay times. These results show that, for the present devices, the switching and delay times are dominated by different mechanisms. 1 Introduction The progress in the development of very high-speed electronic devices has exceeded that of techniques to characterize their performance. Modulation-doped field-effect transistors MODFETs) with power-gain cutoff frequency fmax = 455 GHz [1] and current-gain cutoff frequency, t = 340 GHz [2] have been demonstrated. Such high frequencies are usually estimated by extrapolating experimental data acquired with a wideband network analyser; measurements are often limited to about 60 GHz, although they have recently been extended to 110 GHz. As pointed out in [3], this extrapolation method has not been proven to be reliable. Development of a network analyser system using non-linear transmission lines [4] has shown promising progress in extending all-electronic measurement to higher frequencies. Alternatives to all-electronic methods are electrooptic or photoconductive sampling [5, 6] using ultrafast lasers, which have much broader bandwidths than conventional all-electronic equipment. Electrooptic sampling and photoconductive sampling have been used in characterization of high-speed electronic devices, such as the modulation-doped field-effect transistor [7-9] and 0306-8919 9 1996 Chapman & Hall 867  A Zeng et al the heterojunction bipolar transistor [3 8 10]. Such measurement can provide valuable high- speed characterization as well as information about non-equilibrium carrier transport effects. In most of the previous work the device under test was wire bonded to a test fixture. To avoid parasitic effects introduced by bond wires and extend these measurements to higher fre- quencies it is essential to monolithically integrate the device under test with the test fixture. This approach of on-wafer integrated test fixtures has been applied to electrooptic characteriza- tion of InAs/A1Sb resonant tunnelling diodes [11] double heterostructure GalnAs/InP p-i-n photodiodes [12] and MODFETs [9 13] where the devices of interest were integrated with coplanar test fixtures. In this paper the authors report electrooptic measurement of high-speed MODFETs mono- lithically integrated with coplanar stripline fixtures and photoconductive signal generators. First the switching characteristics of two lattice-matched In0.52A10.48As/In0.53Ga0.47As MODFETs fabricated on the same chip with different gate-access layouts are described: one with a single-airbridge gate contact the other with a double-airbridge gate contact. Excitation 2mm -r 2ram 2ram 2ram Figure Layout of integrated coplanar stripline and MODFET not to scale). The dashed ine represents he substrate. The MODFET s connected n a common-source configuration, and S, D and G are the source, drain and gate contacts, respectively. T ..... ....... ii-Dra n Source Aibailge Gate~ ...... L_..~ Figure 2 Schematic of a double-gate contact MODFET. It is integrated with input and output transmission lines which lie on the left and the right of the MODFET, respectively. Photoconductive switches are incorpor- ated in the transmission ines and are outside he view shown. The gate, drain and source electrodes, as well as the airbridges, are illustrated on the schematic. Scanning electron micrographs of the MODFETs are shown in Fig. 3. 8 8  Electrooptic characterization of modulation doped fieM effect transistors Then comparison is made of measurements of two devices with identical gate-access layouts fabricated with different materials: the devices lattice-matched to InP described above, and pseudomorphic Ino.20Gao.80As/Alo.25Gao.75As devices on GaAs. Comparing results for these cases shows that for the authors devices the switching and delay times are dominated by different mechanisms. In Section 2 the device fabrication and experimental set-up is described, the results are presented in Section 3, and the conclusions in Section 4. 2 Experiment procedure In Fig. 1 one shows a schematic top-view of the coplanar stripline fixture that is integrated with a MODFET. Devices made on two different substrates with two types of gate-electrode layout have been investigated; processing is similar for both. The first devices are pseudomorphic Ino.EoGao.80As/Alo.asGao.75As on GaAs substrates, and the second devices are Ino.szAlo.48As/ Ino.53Gao.47As lattice matched to InP. In Fig. 2 is shown a schematic of the integrated structure near the region of a double-gate contact MODFET, where the gate, drain and source electrodes Figure Scanning electron micrographs of two integrated MODFETs with different gate access structures: a) double-gate contact MODFET, where the gate is made in contact with the lower left coplanar electrode input transmission line) through two airbridges over the source; and b) single-gate contact MODFET, where the gate is made in contact with the input transmission line through one airbridge over the source. The black and white bars at the bottom of the photographs show the scale; each is 0.1 mm. 869  A Zeng et al as well as the airbridges are labelled. Figure 3 shows photographs of two devices near the region of the MODFETs; the photoconductive switches are outside the view of the pictures. Figure 3a and b shows the double- and single-contact gate access structures, respectively. The gate is defined by electron-beam lithography in a bilayer resist scheme (PMMA/copolymer). The recess is done by wet etching in a phosphoric-hydrogen peroxide-water solution. After recess etching, a Pt-Ti-Pt-Au gate with a T-shaped cross-section is formed. It has a length of 0.35 m and an active width of 100 m. Devices are mesa isolated, and the gate, source and drain are integrated with the metallic coplanar electrodes of the input and the output transmission lines. The upper electrodes are common for both input and output transmission lines, and are connected to the source of the MODFET. The lower electrode of the input transmission line, which is on the left in Fig. 3, contact the gate through one or two airbridges over the source electrode. The drain is connected to the lower electrode of the output transmission line, which lies on the right-hand side of the MODFET in Fig. 3. The coplanar electrodes are formed by first etching down to the nominally-undoped MBE-grown buffer. The buffer materials are GaAs and In0.52A10.48As for devices on GaAs and InP substrates, respectively. Then a metal stack is evaporated that consists of Ti, Pt, Au and TiW, and finally 2 m of Au is plated to lower the electrode resistance. All gaps are 5 m, and the coplanar electrode widths are 55 m on the MODFET side of the switches; further from the MODFET the three electrode widths are 25, 25 and 55 m, respectively. The gate input signal is photoconductively excited at the corner of the L-shaped gap, as shown in Fig. 1. No attempt is made to impedance match the connections to external power supplies because reflections from these discontinuities fall outside the time window of interest and play no part in the measurement. Electrooptic measurements are made with 150 fs pulses from a titanium : sapphire laser, a mixer-based detection system [14] and an external LiTaO3 electrooptic probe tip [15]. >~ -0.05 -0.10 -0.15 15 20 25 30 a) Time ps) 1.10 1.05 1 00 15 20 25 30 b) 13me ps) Figure Switching response for a lattice-matched Ino.52AIo.4eAs/Inos3Gao47As MODFET with double- airbridge gate access: a) shows a series of negative-going gate inputs, and b) shows the corresponding drain outputs. The operating point is Vgs = 0V and Vds = 1 V. The switching time is 4.2ps and the delay time, estimated as described in the text, is 3.2 ps. 870  Electrooptic characterization of modulation dopedfield effect transistors 3. Results and analysis In Fig. 4 is shown the switching response of a lattice-matched In0.52A10.48As/Lr10.53Gao.47As device with a double-gate contact structure. In the upper panel a series of four gate input signals is shown, measured on the input stripline 0.4 mm from the gate. The input is a negative-going step-like signal with a risetime of 2 ps followed by an overshoot and a slow decay. The feature at approximately 22 ps is the beginning of the gate reflection. The threshold voltage is -0.7 V, and a gate operating point of 0 V is used; direct current (d.c.) transconductance is 300 mS mm -1 at a drain bias of 1 V. In the lower panel the corresponding drain outputs measured on the output stripline 0.2 mm from the gate are shown; the data show the drain voltage, Vds, starting from the d.c. drain bias of 1 V. Since the MODFET is connected in a common source configuration, it functions as an inverter. The 10-90 risetime of the largest drain response is 4.2 ps; the risetimes of the other signals shown are similar. The time axis is common for the gate and drain signals shown in Fig. 4, which enables an absolute determination of the delay through the device. From Fig. 4 the delay from the midpoint of the input transition to the midpoint of the output transition for the largest signal is 8.1 ps. Since the input signal and the output signal are measured at two different locations separated by 0.6mm, the measured 8.1 ps delay is a combination of propagation on the transmission lines and the MODFET response. The device delay is estimated by subtracting from the measured input-output delay the propagation delay that would be incurred on a coplanar stripline of equal length, which from the measured propagation velocity is 4.9 ps. This yields a value for the MODFET delay of 3.2ps. It should be pointed out that this method of determining the device delay time is more accurate for the integrated MODFET-transmission-line structure than for a discrete device wire-bonded to a test fixture. This is because the short bonding wires are difficult to reproduce, and the associated signal propagation delay is therefore difficult to measure accurately. To study the effect of different gate access structure on the switching characteristics, another lattice-matched MODFET was studied on the same wafer as the device described above. In this case, however, it has a single-gate contact structure. The threshold voltage is -0.7 V, and the d.c. a) b) 0.0= 0 0 5 10 15 20 25 1.00 0.90 0.80 0 1'0 1'5 :~0 13me (ps) Figure Switching response for a lattice- matched Ino.52Alo.~As/Ino.53Gao.47As MOD- FET with single-airbridge gate access: a) shows a series of increasing gate inputs, and b) shows the corresponding drain outputs. The operating point is Vgs= 0 V and Vds = 1 V. 25 The switching time is 5.2 ps and the delay time, estimated as described in the text, is 3.3 ps. 871
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks