THz Range Low-Noise SIS Receivers for Space and Ground-Based Radio Astronomy^†

Round 1

Reviewer 1 Report

The Introduction must be enlarged to include more details (and references) concerning applications of superconducting mixers other areas of physics, not only astronomy. A part of the Introduction must be devoted to the physics of the superconducting mixers - the scientific part of the paper is almost non existing. 

I do not agree to name the frequency range over 600 GHz as RF (the Intorduction). This is clearly THz (or at least sub-THz) range. Anyway, the authors use the term sub-THz in other parts of the manuscript.  In 2.2 part, matching is considered " RF frequencies" - what is the range of frequency really considered? Also, what is considered IF band?

What type of Fourier spectrometer was used?  Glow bars emit at mid-infrared which has nothing to do with the frequencies considered in the paper. Please specify, what do you mean by "a wide-band GHz - THz glow bar source". 

The authors use the notion "photon step" without giving its meaning - please give the definition of this slang expression. 

"Shapiro step" requires a reference.

In Fig. 2 and other figures  - please give type and power  parameters of the pumping 950 GHz LO source. 

There is no "solid line" in Fig. 3. 

Author Response

Dear editor, dear reviewer,

Thank you for the review of our manuscript, please find below our answers on referee questions, comments and propositions.

Review Report (Reviewer 1)

1. The Introduction must be enlarged to include more details (and references) concerning applications of superconducting mixers other areas of physics, not only astronomy. A part of the Introduction must be devoted to the physics of the superconducting mixers - the scientific part of the paper is almost non existing. 

A large piece of text describing how the SIS mixer works is added to Introduction, including related references.

“The sharp nonlinearity in the tunneling current of the SIS junction is used for the mixer operation; this nonlinearity results from a single-electron tunneling process between two superconductors separated by a very thin oxide insulating layer. To describe this tunneling current under the influence of local oscillator (LO) signal a quantum-mechanical description is required [l, 2].This process is known as photon assisted tunneling [2], as a result quasiparticle steps appear in the SIS current-voltage characteristic (IVC). The heterodyne SIS mixer down-converts the weak input signal of interest to a lower intermediate frequency (IF) without loss of phase; the spectrum of IF signal is the same as input one, but is shifted down in frequency by LO frequency. Intrinsically the SIS mixers can exhibit conversion gain; important advantages of such mixers are a low LO power requirement and very low intrinsic noise [1-4].”

Indeed, the paper is concentrated on the development of the low-noise SIS receivers for space and ground-based radio astronomy that is the primary area for SIS mixer applications. We have checked the most cited articles devoted to the SIS receivers – except radio astronomy there are also developments of the so-called superconducting integrated receiver (SIR) and its applications in the laboratory and for Earth atmosphere monitoring (see references below), but we decided do not overload this paper by self-citing.

V.P. Koshelets, S.V. Shitov, “Integrated Superconducting Receivers” Superconductor Science and Technology, vol 13, pp. R53-R69, (2000).

Gert de Lange, et al “Development and Characterization of the Superconducting Integrated Receiver Channel of the TELIS Atmospheric Sounder”, Supercond. Sci. Technol. vol. 23, No 4, 045016 (8pp), (2010). 

Mengyue Li, et al “Linewidth dependence of coherent terahertz emission from Bi₂Sr2CaCu₂O₈ intrinsic Josephson junction stacks in the hot-spot regime” PRB (Rapid Communications) 86, 060505(R) (2012).

• I do not agree to name the frequency range over 600 GHz as RF (the Introduction). This is clearly THz (or at least sub-THz) range. Anyway, the authors use the term sub-THz in other parts of the manuscript.  In 2.2 part, matching is considered " RF frequencies" - what is the range of frequency really considered? Also, what is considered IF band?

We corrected this and replace “RF” by appropriate words in the text, except “RF magnetron sputtering” and “magnetron RF”, where frequency = 13.56 MHz.

• What type of Fourier spectrometer was used?  Glow bars emit at mid-infrared which has nothing to do with the frequencies considered in the paper. Please specify, what do you mean by "a wide-band GHz - THz glow bar source". 

To make it more clear, we corrected "a wide-band GHz -THz glow bar source" to "a wide-band sub-THz glow bar source (1500 K black body)"

It is classical FTS based on Michelson Interferometer. The Glowbar is heated to about 1500 K and provides a black-body wideband radiation with Planck distribution. It generates enough power to provide quit measurable response of SIS mixers at both presented ranges (250 GHz and 850 GHz). The SIS current is detected using Lock-In amplifier. This technique is common for determining of the frequency response of SIS-mixers and it is used worldwide for many decades, only because of that we decided not to give more details here.

• The authors use the notion "photon step" without giving its meaning - please give the definition of this slang expression. 

We replace in the text "photon step" by “quasiparticle step” and discuss appearance of these steps in the Introduction

• "Shapiro step" requires a reference.

Appropriate reference is added.

• In Fig. 2 and other figures  - please give type and power  parameters of the pumping 950 GHz LO source. 

For these measurements setup at the Kapteyn Astronomical Institute was used. The LO source was similar to sources used on ALMA; it is based on chain of multipliers with output power of about 20 mW at 950 GHz. Taking into account losses in a beamsplitter (-10 dB) and rather low coupling at high frequencies, the delivered power was of about 0.5 mW or even lower. Fig. 2 caption was modified: the following statement was added “a CW LO source delivering about 0.5 μW of power to the SIS junction at 950 GHz.”

• There is no "solid line" in Fig. 3. 

Fig 3 caption is corrected

All of these the changes are highlighted in the Marked-Up Manuscript file.

Furthermore, small textual changes have been made to correct for grammatical and textual errors.

Thank you for your time and we are looking forward to be hearing from you.

On behalf of the co-authors,

Kind regards,

Andrey Khudchenko and Valery Koshelets

Author Response File: Author Response.pdf

Reviewer 2 Report

Minor comments:
Page 2 line 70: It is not clear that the "column-like microstructure" is the structure of surface or the  polycrystalline nature of the film. A specification is suggested.

Page 2 line 83: was ->  is
Page 3 line 98: "...less than their coherence lengths..."  coherence length is a physical quantity  for superconductors as far as I know. What is the meaning of it for normal metals?
Page 4 line 135: isolator -> insulator
Page 4 Figure 2: Visually the quality R_sg/R_n is less than 20. What is the definition of R_n and the R_sg? 
Page 4 line 163: Jg -> Jc , Jc is more often used to refer critical current density. 
Page 7 line 214: Nb-SiO2-Nb -> Nb-SiO_2-Nb
Page 7 line 225: isolator-> insulator
Page 9 line 280: 30 kA/cm2-> 30 kA/cm^2

Major comments:

Page 2 line 100: The authors discuss about the knee structure of the IVCs and the method to eliminate it. It is very interesting discussion because it implies that the knee structure can bring some negative effects in SIS mixer operation. It will be thus necessary to clarify these effects in the text to justify the effort of removing it. 

Page 8 line 253: In the text it is read that "they were only twice the value of" the quantum noise. In fact, the results shows that the noise is close to one quantum noise in the range of 241-275 GHz. Since it approaches the lowest value governed by the physical law, it is better to carefully verify the results. For this purpose, the gain of the SIS mixer as well as the linearity of the gain are important data. It is necessary to add them.

Page 10 line 310: The improvement of the mapping speed due to improvement of the receiver sensitivity is strongly depends on the atmospheric noise. The atmospheric noise of APEX cite should be taken into consideration.

Author Response

Dear editor, dear reviewer,

Thank you for the review of our manuscript, please find below our answers on referee questions, comments and propositions.

Review Report (Reviewer 2)

Comments and Suggestions for Authors

Minor comments:

Page 2 line 70: It is not clear that the "column-like microstructure" is the structure of surface or the polycrystalline nature of the film. A specification is suggested.

That is the column-like microstructure of the Nb film; appropriate reference is added:
Imamura, T., Shiota, T., and Hasuo, S. (1992). Fabrication of high quality Nb/AlOx-Al/Nb Josephson junctions: I - Sputtered Nb films for junction electrode. IEEE Trans. Appl. Supercond., 2, 1.

Page 2 line 83: was ->  is

Corrected

Page 3 line 98: "...less than their coherence lengths..."  coherence length is a physical quantity  for superconductors as far as I know. What is the meaning of it for normal metals?

Actually, Al is a superconductor, that is in a normal state at T = 4.2 K; although according to [21] the Al layer should be considered as S’ in the theory

To make this part more clear, it was modified:

“This feature is due to the presence of aluminum layer near the tunnel barrier; the IVC of the Nb–Al/AlO_x–Nb (S-S’-I-S) tunnel junction depends on the quasiparticle density of states in the S’ layer (Al). This density of states for the Nb/Al bilayers was calculated using the microscopic proximity effect model [21]. The model assumes a short electron mean free path (dirty limit conditions) in both S (Nb) and S’ (Al) materials.”

Page 4 line 135: isolator -> insulator

Corrected

Page 4 Figure 2: Visually the quality R_sg/R_n is less than 20. What is the definition of R_n and the R_sg? 

Indeed, Rsg determined at 2.5 mV  = 335 Ohm, that corresponds to Rsg/Rn = 20 rather than 23

Rsg/Rn = 23 was measured for this sample in the dipstick, presumably the Rsg value was decreased in the cryostat. The measured in the cryostat  Rsg/Rn = 20  is reported now in Fig 2.

Page 4 line 163: Jg -> Jc , Jc is more often used to refer critical current density. 

We consciously use Jg rather than Jc. The reason is that the Ig can be measured unambiguously even for sub-micron SIS junctions, while the Ic value can be suppressed by both fundamental (e.g. strong coupling effects for junctions with Nb and especially NbN electrodes), and technical i reasons

Page 7 line 214: Nb-SiO2-Nb -> Nb-SiO_2-Nb

Corrected

Page 7 line 225: isolator-> insulator

Corrected

Page 9 line 280: 30 kA/cm2-> 30 kA/cm^2

Corrected

Major comments:

Page 2 line 100: The authors discuss about the knee structure of the IVCs and the method to eliminate it. It is very interesting discussion because it implies that the knee structure can bring some negative effects in SIS mixer operation. It will be thus necessary to clarify these effects in the text to justify the effort of removing it. 

 Influence of the knee-structure is discussed in [38] and in the PhD thesis by Kirill Rudakov [39].

Here the following sentence is added:

“Presence of the knee structure results in the negative differential resistance on the quasiparticle steps that in turn leads to the mixer instability and nonlinear operation at some frequencies.”

Page 8 line 253: In the text it is read that "they were only twice the value of" the quantum noise. In fact, the results shows that the noise is close to one quantum noise in the range of 241-275 GHz. Since it approaches the lowest value governed by the physical law, it is better to carefully verify the results. For this purpose, the gain of the SIS mixer as well as the linearity of the gain are important data. It is necessary to add them.

Actually, SIS mixer noise temperature is limited by the quantum value hf/(2k_B) [5]. In the mentioned sentences we discuss not the best value, rather the highest values in the range (that is specification for the Millimetron “they were only twice the value of hf/k_B in the frequency range from 240 to 275 GHz”

Information on the mixer linearity and stability are presented in two last paragraphs of section 3.1; just before Fig. 9. Further details of the mixer noise breakdown, stability and linearity issues are presented in PhD thesis by Kirill Rudakov [39] and will be presented in the separate paper that is under preparation.

Page 10 line 310: The improvement of the mapping speed due PhD thesis by Kirill Rudakov [39].to improvement of the receiver sensitivity is strongly depends on the atmospheric noise. The atmospheric noise of APEX cite should be taken into consideration.

This sentence was modified; the reference concerning atmospheric noise of APEX cite is added:

“In case of good weather conditions, giving atmosphere transparency of about 50 % [42], the mapping speed will be improved by about 40%.”

All of these the changes are highlighted in the Marked-Up Manuscript file.

Furthermore, small textual changes have been made to correct for grammatical and textual errors.

Thank you for your time and we are looking forward to be hearing from you.

On behalf of the co-authors,

Kind regards,

Andrey Khudchenko and Valery Koshelets

Author Response File: Author Response.pdf