A Concept for a Multipurpose Time-of-Flight Neutron Reflectometer at Compact Neutron Sources

Round 1

Reviewer 1 Report

"A novel concept for a new multipurpose time-of-flight neutron reflectometer for compact neutron sources", by Erhan, et al. describes an unusual neutron reflectometer concept designed to be installed in a confined space on a small reactor.  The FREYJA instrument proposes to select an incident beam from several fixed collimating apertures, pass this beam through two choppers to define a neutron pulse, reflect from a sample, and measure reflectivity vs. wavevector Q via neutron time-of-flight.  FREYJA resembles existing instruments built at Delft and Orsay in its use of burst-defining choppers.  So far, so good, but I cannot understand the neutron counts on detector depicted in Fig. 10. 

Conventional reflectometers produce a toothpick-shaped beam, reflect that beam from a flat sample, and measure the intensity of the toothpick-shaped reflected beam in a downstream detector.  Instruments employing a monochromatic incident beam probe different wavevectors by measuring at a series of increasing incident angles, producing specularly reflected beams at twice those angles on a fixed position-sensitive detector.  Instruments employing time-of-flight, by contrast, can employ a fixed angle and probe Q by wavelength discrimination using neutron time-of-flight.  For fixed wavelengths, the detector in Fig. 10 would show a vertical trace varying in intensity with height; the tof instrument would produce a horizontal trace. 

Figure 10 in this paper shows both - how this happens is not explained in the text.  The situation resembles that shown in Fig. 3 of reference [16] from the Orsay group wherein they synchronously rotate the sample with time-of-flight.  Alternatively, such a trace could be produced from a divergent incident beam dispersing over the detector face, as first described by Hayter, Hamilton, and Smith in 1994 or work done at SINQ and proposed for ESTIA at ESS.  The latter seems possible, given the description of the dependence of resolution on detector pixel size. 

Since the authors do not describe anything about a moving sample, presumably the divergent beam is being used.  If so, they should reference earlier work and more importantly, explain clearly how this all works.  For example, at minimum state clearly that they are using a divergent beam, show that the incident collimator (C) suite and slit (S1) provides sufficient angular divergence, replace bin values in Fig. 10 with wavelengths and angles.

In the absence of these explanations, I do not find the work comprehensible, which is why I ranked it high on novelty but low on scientific soundness and merit.  It may well work, but I can’t figure out how.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript presents the concept for a neutron reflectometer for the former Norwegian neutron source at IFE. The unique selling point of the concept was that it fit into the very limited space available at that facility. But to promote it without any changes as a 'novel concept' for compact neutron sources is false labeling!
To me it looks like the manuscript was written for the initially intended instrument and with the shut-down of the source it was stopped. And now it is to be published with a few changes in the introduction mentioning future sources - and with a new title.
I think it is worth publishing the ideas, concepts, simulations and tests related to the compact overall design. But there is no need to describe in detail features realized on other instruments for many years (or decades).
Future compact sources might make use of some of the design considerations - but certainly not of Ad van Well's double chopper: This is ideally suited for continuous sources, but on a pulsed source the first disc would have to be located very close to the moderator. There, to adapt the tof resolution only one disc is needed, the pulsed source takes over the task of the first disc.

I thus recommend to either adapt the title and sell the manuscript as a study for a compact reflectometer at a continuous source, or to make changes to the concept so that it fits the requirements of compact sources.

For the manuscript as it is I have a few notes:

1. The description of the principles of reflectometry is rather weak and uses terms typical for neutron diffraction. E.g. the interference pattern is caused by partial reflection and transmission at interfaces. This in the end is somehow related to "scattering by .. nuclei" - but just mentioning the nuclei and omitting the layers of different materials necessary for reflection does not explain anything. Also there are many reflectivity curves without peaks - and it is still possible to analyze the data. One determines the thickness of one or more layers, but in general not "of the sample". Why is there an "and" between "layer thicknesses" and "scattering length density profile"? It is the same feature, described with different terms. ...

2. The channels of the collimator unit point towards the sample at various angles and in addition the wider channels transport a considerable divergence. This should shift and blur the tof (and wavelength) cut off of the FOM. Since these are essential components of the design, this effect should be discussed. (MC is not needed for this, it can be done analytically)

3. A polarized beam "as good as possible" has a polarization above 99 %. Still, 95 % is not bad for a single transmission polarizer. It just does not fulfill the initial criterion.

It is unclear what the statement "and the beam direction needs to be fixed" means in this context.

In case the reader does not know what a supermirror is, he won't learn it here. Please give a reference.

4. What is the relation of a gradient field used for the spin flipper and the fact that it "hence can also be used without chopper"? And what would be the design of a radio-frequency flipper with wires in the beam? There seems to be a mix-up of the discussion Mezei flipper vs. RF flipper.

5. Table 5: what is a negative frequency? Why dos the supermirror coating have different m values for spin-up and down?

6. Eqn. 4: "theta = ... = R * theta / sqrt 2"  can be easily solved by setting "R = sqrt 2".

It is also not a good idea to label the "wave vector resolution" with "R_q" in an article about a reflectometer - it might easily be confused with "R(q)" for reflectivity. 

7. Table 6: innovative lay out - but hardly readable.

8. The MB detector (discussed here) consists of 20 cassettes, each with 32 wires and 64 stripes. One might add up all the wires to get the 640 for the full detector because that gives an idea for the resolution. But adding up the stripes does not make sense.
The MB detector is 2 dimensional (or in fact 4 dimensional because the blades are inclined and it allows for time resolution) - so what is meant by "For added flexibility ... 25 % of the area could be furthermore designed as a 2D-detector"?

9. In many places, physical units are set in italic. This should be changed to an upright font.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper describes a new reflectometer concept to be installed for next generation compact neutron source. As the authors have pointed out, the key points for equipment installed in compact neutron sources are for obtaining higher intensity, efficient use of limited space, and reducing high radiation background. The paper has a proposal for components that may be useful for a compact radiation source, but there is not enough description, so it is difficult to decide whether to accept it.

1)  In section 2.2, as one of the main features of the reflectometer, the authors propose  a multi-channel collimator-guide and reflective neutron guide. This equipment makes it possible to select the Q resolution efficiently and with making good use of space. But this part has a significant shielding defect, and Aluminum component materials emit intense prompt-gamma rays and increase the background for measurements. As the authors stated the design considerations made here are highly relevant for neutron sources where the instrument components are in a high-radiation environment.  This idea is not suitable for proposal without shielding evaluation. Proposals that include shielding performance evaluation are required.

2)  Section 2.1 describes the results of neutron intensity and TOF spectrum measurements for CNS sources. For neutron intensity measurement, the upgraded cold neutron source (CNS) at the JEEP II reactor was used. The measurement conditions at the reactor are specifically described. On the other hand,  TOF spectrum measurement and calculation shown in Fig.3 are not specifically described. Detailed experimental conditions on TOF measurement is required to evaluate the performance of the instrument. Regarding this description, a paper on beamline measurements using partially coupled cold-hydrogen moderator operated at LANSE is listed as a reference, even though the flux measurement was conducted on the cold-hydrogen moderator with H2O pre-moderator in a heavy water reactor (JEEP II).

3)  In Section 2.1, line 94, there is a description "For the following Monte Carlo simulations (see also next section) the CNS is situated in a tangential channel about 0.55 m away from the source. It was considered that the source vessel is made of Al-Mg3,,,".  From this sentence, it seems that a code such as MCNP is used in the Monte Carlo calculation code, but there is no description of it later. The definition of the calculation code being used is not clear.

4) In Section 2.3, the authors propose a V-shaped cavity to achieve the range of wavelengths.  Regarding this component, please refer to the references developed for polarization device etc.

5) In Section 2.6.2. line 262, please describe why it is okay to use a non-polarizing Ni-Ti super-mirror for choosing spin-up and spin-down neutrons by using reflected beams and transmitted beams.

6) In Section 1, line 23, the authors refer reference [2] as future continuous production of neutrons sources but these are small and medium-sized power reactors. The reference should be changed to small and medium-sized research reactors. There are some references that the authors should check again such as [3], [9], [12] etc. from the perspective of the original paper.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed the issue of using a divergent beam and so satisfied my major question concerning the work.

Author Response

Please see the attachment.

Reviewer 2 Report

The authors improved the manuscript in some places, but my overall critics are still valid. An instrument for a compact source will not make use of the presented double chopper solution. There is no reasoning why a compact source leads to a compact instrument design. The space-saving solutions are the only innovative aspect of this work! 

There are inconsistencies (most likely due to a local adaption of the manuscript to my objections): The scattering plane is horizontal, both for the sample and for the FOM according to the FOM illustration. But the text mentions perpendicular planes.

Equation 4 is still wrong. 

I asked for a major revision and got some fixes. I thus do not recommend the manuscript for publication!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments 1': In section 2.2, the boron nitride material is effective only against thermal neutrons and is not effective against fast neutrons or gamma rays. I pointed out a serious shielding lack in the multi-channel collimator-guide and V-shape FOM. When looking at the CNS from the sample position, the collimator-guide and FOM parts have almost no shielding performance. Furthermore, in order to guide a beam from the cold source through the gap in the collimator, there must be a void channel in the upstream. This void decreases shielding performance at all.  Also, this wide divergence beam makes aluminum component materials emit intense prompt-gamma rays and increase the background for measurements. This idea would not be possible without shielding evaluation. Revised proposal that include shielding performance evaluation are required.

Comments 2': Detailed description on TOF spectrum measurements for CNS  makes it easier to understand.

Please add the distance of chopper to detector.

Comments 5': In Section 2.6.2. line 262, please describe why it is okay to use a non-polarizing Ni-Ti super-mirror for choosing spin-up and spin-down neutrons. And please describe why are the supermirrors with different m values used for selecting spin-up and spin-down neutrons ?

Comment 6': In Section 1, line 37, the authors refer reference [3] as the interference of reflected waves.  The authors should check again from the perspective of the original paper. For example, the following documents is listed "Hayter J B, Penfold J and Williams W G 1976 Nature 262 569-70".  And, in Section 1, line 45, reference of PNR should be added.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

none

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments 1'': I understood that this instrument can be used for the reflectivity measurement by using the multichannel collimator with the shielding lack condition under the beamline that includes tangentially arranged  CNS and the insertion of filter.  The abstract should describe that the compact collimator system functions with such conditions of reduced dose from the source.

Comments 7: Section 2.1 line 105, Monte Carlo simulation code of McsTas and Vitess is mentioned here, but following sentences are  described for the explanation of the neutron flux and tof measurements.  Following sentences of CSN position, material, and tangential arrangement are not related to the McsTas simulation input, therefore,   the part of "For the following Monte Carlo (MC) simulations, the McStas [16] and Vitess [17] 105 software packages were used (see also next section). In the simulation setup," is not needed here. 

I understand that neutron simulation sources for the McsTas and Vites were created based on the results of the neutron flux measurement and tof measurement performed in 2.1.  It may be better to describe that the neutron source obtained from the measurements in 2.1 were used.

Comment 8 : When using a multi-channel collimator, each collimator views different surface area of CNS, so different direct beam should be used in the reflectivity calculation. It should be described such explanation.

Comment 9 :  In Table 5,  a polarization device with m=7 is used, and a reference to this should be described.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf