Heiss W.D. (ed.) Quantum dots.. a doorway to - tiera.ru
Heiss W.D. (ed.) Quantum dots.. a doorway to - tiera.ru
Heiss W.D. (ed.) Quantum dots.. a doorway to - tiera.ru
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Semiconduc<strong>to</strong>r Few-Electron <strong>Quantum</strong> Dots as Spin Qubits 45<br />
the sample heat up due <strong>to</strong> spurious noise and interference. Several filtering<br />
stages are requir<strong>ed</strong> for different frequency ranges (see Fig. 11 and Fig. 12). In<br />
the connec<strong>to</strong>r box at room temperature, all wires are connect<strong>ed</strong> <strong>to</strong> ground via<br />
0.22 nF “fe<strong>ed</strong>through capaci<strong>to</strong>rs”. At base temperature, all signal wires <strong>ru</strong>n<br />
through “copper powder filters” [35]. These are copper tubes fill<strong>ed</strong> with copper<br />
powder, in which 4 signal wires with a length of about 2 meters each are<br />
wound. The powder absorbs the high-frequency noise very effectively, leading<br />
<strong>to</strong> an attenuation of more than −60 dB from a few 100 MHz up <strong>to</strong> more than<br />
50 GHz [36].<br />
To remove the remaining low-frequency noise, we solder a 20 nF capaci<strong>to</strong>r<br />
between each signal wire and the cold finger ground. In combination with the<br />
∼100 Ω resistance of the wires, this forms a low-pass RC filter with a cut-off<br />
frequency of about 100 kHz (even 10 kHz for the wire connect<strong>ed</strong> <strong>to</strong> the IV<br />
conver<strong>to</strong>r, due <strong>to</strong> its input resistance of about 1.3 kΩ). These filters are us<strong>ed</strong><br />
for the wires connecting <strong>to</strong> ohmic contacts (although they were taken out <strong>to</strong><br />
perform some of the high-bandwidth measurements describ<strong>ed</strong> in this work).<br />
For the wires connecting <strong>to</strong> gate electrodes, a 1:3 voltage divider is present<br />
(consisting of a 20 MΩ resistance in the signal line and a 10 MΩ resistance <strong>to</strong><br />
ground). In this way, the gate voltages are filter<strong>ed</strong> by a low-pass RC filter with<br />
a cut-off frequency of about 1 Hz. By combining all these filters, the electrons<br />
in the sample can be cool<strong>ed</strong> <strong>to</strong> an effective temperature below 50 mK (if no<br />
extra heat loads such as coaxial cables are present).<br />
High-Frequency Signals<br />
High-frequency signals can be appli<strong>ed</strong> <strong>to</strong> gate electrodes via two coaxial cables.<br />
They consist of three parts, connect<strong>ed</strong> via standard 2.4 mm Hewlett Packard<br />
connec<strong>to</strong>rs (specifi<strong>ed</strong> up <strong>to</strong> 50 GHz). From room temperature <strong>to</strong> 1 Kelvin,<br />
a 0.085 inch semi-rigid Be-Cu (inner and outer conduc<strong>to</strong>r) coaxial cable is<br />
us<strong>ed</strong>. From 1 Kelvin <strong>to</strong> the mixing chamber, we use 0.085 inch semi-rigid<br />
superconducting Nb. From the mixing chamber <strong>to</strong> the sample holder, flexible<br />
tin plat<strong>ed</strong> Cu coaxial cables are present. The coaxes are thermally anchor<strong>ed</strong> at<br />
4K, 1K, ∼800 mK, ∼100 mK and base temperature, by clamping each cable<br />
firmly between two copper parts. To thermalize also the inner conduc<strong>to</strong>r of<br />
the coax, we use Hewlett Packard 8490D attenua<strong>to</strong>rs (typically −20 dB) at<br />
1 K. These attenua<strong>to</strong>rs cannot be us<strong>ed</strong> at the mixing chamber, as they tend<br />
<strong>to</strong> become superconducting below about 100 mK. We have also tri<strong>ed</strong> using<br />
Inmet 50EH attenua<strong>to</strong>rs at the mixing chamber, but these show<strong>ed</strong> the same<br />
problem.<br />
To generate the high-frequency signals, we use a microwave source (Hewlett<br />
Packard 83650A) that goes up <strong>to</strong> 50 GHz (or 75 GHz, in combination with a<br />
“frequency doubler”); a pulse genera<strong>to</strong>r (Hewlett Packard 8133A), which generates<br />
simple 10 ns <strong>to</strong> 1 µs pulses with a rise time of 60 ps; and an arbitrary