Stealth Laboratory
Our state-of-the-art stealth laboratory has been built specifically for the measurement of nanostructures in extremely sensitive environments. The laboratory provides unprecedented isolation from the entire spectrum of noise, from low-frequency vibration to high-frequency electromagnetic fields in the gigahertz range. The broadband isolation is achieved by suspending the entire concrete base of the laboratory, and constructing various levels of shielding, which, for the first time, incorporate stealth technology. Layers of RAM (radar absorbing material), copper floor and pit lining allow a minimum of 80 dB isolation of plane wave electric field up to 10 GHz. Two levels of concrete vibration isolation and suspension of the dilution fridge from an optical table on air legs minimize vibrations down to a level of sub-angstroms. Multiple levels of mu-metal shielding minimize the level of stray magnetic field down to the microgauss level.
COMPREHENSIVE DC-TO-DAYLIGHT NOISE ISOLATION DESIGN
i) Acoustic Noise: Three-Stage Acoustic Isolation
Acoustic noise (10 mHz – 10 kHz) arises because of the building vibration, seismic waves, ground noise due to nearby train or car movement, and the normal audible noise. The experimental setup in our lab — i.e. the sample inside the cryostat — has a three-level isolation: (i) acoustic room, (ii) suspension block from the building floor, and (iii) floating optical table. The cryostat hangs from the center of the optical table down inside a pit, which is designed to be an anechoic chamber. In addition, inside the cryostat, small suspensions with large spring constants are attached to further damp any mechanical vibration.
ii) Electromagnetic Noise: Three-Stage EMI Isolation
Electric field noise has a wide spectrum due to a variety of noise sources. Our experiments are sensitive at the level of 1.0e-18 watts or attowatts (electron decoherence requires on the order of 10 to 100 attowatts). The EMI (Electro-Magnetic Interference) noise sources are electrical 1/f noise (10 mHz – 10 kHz), MHz range RF noise (100KHz – 100 MHz), Low-Frequency microwave noise (10 MHz – 1 GHz) from microwave-powered instruments, cell-phone noise (0.1 – 10 GHz), and terahertz noise (sub-infrared noise due to thermal fluctuations, and optical noise (10 THz – 1.0e15 Hz).
In order to ensure the maximum attenuation or isolation of the noise figure over a broadband of frequencies the lab is designed to have a combination of multiple isolation mechanisms, including acoustic isolation. A specially-designed EMI-shielding room offers at least 100 dB isolation up to 1 GHz, and 80 dB isolation between 1 – 10 GHz. We also employ a Counter-Counter-Stealth Technology with nonmagnetic RAM (Radar-Absorbing Material) to selectively introduce additional attenuation in the TERAHERTZ range, following techniques developed to be a countermeasure for the terahertz-range Ultra WideBand (UWB) radars.
The EMI-shielding room has a thick-copper (non-magnetic) shielding of the floor and the pit to prevent any leakage of high-frequency sub-mm-wavelength electric fields. Additional filters are used to suppress the noise in high and low frequencies. Since we use both low-frequency transport and high-frequency reflection and transmission measurements, the filters are placed selectively to provide attenuation in the appropriate band. High-frequency filtering is obtained by Butterworth filters, and low-temperature silver-powder filters with typical attenuation of 80 dB at 1 GHz. For specific time-domain measurements, with complex bandwidth requirements, Gaussian filters such as Bessel-Thompson filters are used.
iii) Stray Magnetic Field Noise: Four-Stage Isolation
Without any magnetic shielding, the laboratory location was to found to have a RMS magnetic-field noise of 2 milligauss (per root hertz). This is primarily because of the close proximity of the building to the metro lines. In order to minimize the magnetic field noise to the micorgauss level, we have employed a series of magnetic isolation: mu-metal shielding around the dewar, non-magnetic copper EMI shielding by the shielding room, nonmagnetic pit, and a superconducting Nb shield around the sample.
iv) Thermal Noise Isolation
The measurements are carried out in a dilution cryostat capable of reaching temperatures down to 6 millikelvin (= 0.006 Kelvin). The low temperatures ensure the suppression of Johnson noise orders of magnitude below the required sensitivity for the measurement of ultrasensitive quantum effects (now determined by the first-stage preamplifier noise of 0.3 nV/rtHz at the required frequencies).
The two pictures above show the dilution cryostat (inside the blue dewar) hanging from an air-suspended optical table. The legs of the optical table stand on top of an isolated concreat pit with copper outlines both outside and on the top. The vibration-isolation pit is located inside the copper shielding room.
The above pictures show the inside of the dilution cryostat insert, which is cooled in stages from room temperature to 4.2 Kelvin (by liquid helium), from 4.2 Kelvin to 1.5 Kelvin with a 1-Kelvin Pot (by pumping on helium), from 1.5 Kelvin to 0.750 Kelvin by condensing a mixture of helium-3 gas and helium-4 gas through a closed line, and from 0.750 Kelvin to 0.006 Kelvin by circulating the helium-3 gas in a closed cycle. The high-frequency coaxial cables at the end of the insert allow high-speed, high-frequency measurement of nanomechanical and nanoelectronics structures with unprecedented sensitivity. For example, the sensitivity is high enough to allow the detection of a circulating current due to ONE electron in a sub-micron ring, displacement due to the motion in a nanomechanical structure at the level of a femtometer, and damping or friction arising due to a handful tunneling atoms in a crystal.