Object Composition Identification by Measurement of Local Radio Frequency Magnetic Fields with an Atomic Magnetometer

Round 1

Reviewer 1 Report

This work describes the object composition identification based on inductive second-RF-field measurements with a radio-frequency atomic magnetometer by the presence of primary-field-insensitive axis. The authors addressed the issues of variable orientation and lift-off value and object composition identification by monitoring of the object’s signature phase, frequency dependence of the angular scans, and implementation of single coil or two coils for the primary RF field. In my opinion this manuscript is appropriate for Applied Sciences if authors can add little bit more information below.

(1) Although the experimental setup and operating principle about RF atomic magnetometer may be descripted in the reference, as the main signal receiver, it is necessary to give the figure of device construction and the performance description of the high-sensitivity quantum magnetometer, which can be keep the paper readable. The sensitivity, bandwidth and dead zone of RF atomic magnetometer could be given in more details.  

(2) Could authors give more explanations why the pump beam can pump atoms into a stretched state of the F=4 ground state level on the condition of laser frequency locked to the D2 line, which is important to obtain narrower magnetic-resonance linewidth through light-narrowing effect. It would be better to demonstrate the experimental parameters such as laser power and optical beam size.

(3) Generally, the lock-in signal phase may be affected by systematic errors especially repeating the magnetic-resonance detection. Could authors valuate the effects on the measurements and object identification from systematic errors.

(4) In Fig. 4(b), the side lengths of 4mm and 8mm seem to have a larger fluctuation in some lift-off distances. Could authors supplement the reasons?

Author Response

We are very grateful to the reviewer for their close reading of the text and for their comments that have improved the overall quality of our manuscript. We have addressed these points below.

• Although the experimental setup and operating principle about RF atomic magnetometer may be descripted in the reference, as the main signal receiver, it is necessary to give the figure of device construction and the performance description of the high-sensitivity quantum magnetometer, which can be keep the paper readable. The sensitivity, bandwidth and dead zone of RF atomic magnetometer could be given in more details. 

We added the figure with rf atomic magnetometer arrangement. Additionally, in the ‘Materials and Methods’ section, we added a few sentences introducing the magnetometer operation principle and rf spectrum. Following the Reviewer’s recommendation, we list magnetometer sensitivity and bandwidth.

• Could authors give more explanations why the pump beam can pump atoms into a stretched state of the F=4 ground state level on the condition of laser frequency locked to the D2 line, which is important to obtain narrower magnetic-resonance linewidth through light-narrowing effect. It would be better to demonstrate the experimental parameters such as laser power and optical beam size.

We have added a section describing the indirect pumping mechanism. This section is accompanied, for clarity, by a Caesium energy level scheme.

• Generally, the lock-in signal phase may be affected by systematic errors especially repeating the magnetic-resonance detection. Could authors valuate the effects on the measurements and object identification from systematic errors.

Our phase measurements have differential character, i.e., we are interested not in the absolute value of the phase but in its value relative to the value recorded for the materials regarded as standards, e.g., copper, and ferrite. In other words, the presence of the stable phase shifts within the detection would not affect the outcome of the measurement. The stability of the phase value is confirmed by small values of standard deviation in the presented figures.

• In Fig. 4(b), the side lengths of 4mm and 8mm seem to have a larger fluctuation in some lift-off distances. Could authors supplement the reasons?

As discussed in the context of Figures 6 and 7 (in the revised version) the resolution of the phase measurement, due to noise represented by standard deviation, is affected by the value of signal amplitude. For smaller copper plates the amplitudes of the signal, Fig.5 (a) (in the revised version), are smaller than those recorded for the other two cases, and this is reflected in the degradation of the phase measurement resolution.

Reviewer 2 Report

The manuscript entitled“Object Composition Identification by Measurement of Local Radio Frequency Magnetic fields with an Atomic Magnetometer” by J. D. Zipfel et al. discusses problems related to orientation and “lift-off” in object composition identification with an RF atomic magnetometer. Three possible solutions to the problems are suggested based on the combination of the different magnetometer functionalities. The text is well organized and clear. I think the paper can be useful in real-life measurement using the RF atomic magnetometer. Generally, I think the manuscript is appropriate for publication in Applied Sciences. Several remarks however should be considered and some revisions should be made by the authors before publishing:

1. The spin-maser mode was mentioned in line 58. Still the bias field is stabled by a PID. Please explain. Also, apart from the bias, are fields along the other two axes stabled in the same way as the bias was?

 2: The author defined a “lift-off” distance as the distance between the object and the primary field source and sensor. In Fig. 1, the Lift-off distance labeled by the red arrow seems like representing the distance between the object and the primary field source. Does this mean the field source and the sensor is in the same position? A clear definition is important since the lift-off is the key point of the manuscript.

3: In 3.1.1 of the manuscript, the author uses a copper (0.5 mm thick), steel (0.5 mm thick), and ferrite (2 mm thick) plate to study the amplitude and phase relations to the angle between the object and the sensor. Why the thickness of the sample are different? Will this thickness difference impact the ratio of the surface (eddy currents) and volume (magnetisation) components of the secondary RF field?

 4. The sentence “the Additionally. All recordings were done at an operational frequency of 6 kHz.” should be “the Additionally, all recordings were done at an operational frequency of 6 kHz”.

 5. In Fig. 8(a), the author shows gradient value as the function of lift-off for a copper (red diamonds), brass (green dots), and aluminium (blue squares) plate. There is an undefined green squares in the figure.

6. In line 323 and in the caption of Fig. 8 of the manuscript, the “red dotted cure” should be red dotted curve.

7. The error bars In Fig. 9(a), which show the standard deviation of 30 measurements, can hardly be seem in the figure.

Author Response

We are very grateful to the reviewer for their close reading of the text and for their comments that have improved the overall quality of our manuscript. We have addressed these points below.

1. The spin-maser mode was mentioned in line 58. Still the bias field is stabled by a PID. Please explain. Also, apart from the bias, are fields along the other two axes stabled in the same way as the bias was?

The spin-maser mode is mentioned only to describe sensor functionalities available. The measurements presented here are carried out in the free-running mode only. The bias field is stabilized in terms of amplitude and direction, which requires stabilisation of all three axes. The point has been clarified.

LN 60: ‘The work presented here is carried out in the free-running mode, with active field stabilisation.’

LN 117: … stabilise the magnetic field ‘(all three axes)’ that…

 2: The author defined a “lift-off” distance as the distance between the object and the primary field source and sensor. In Fig. 1, the Lift-off distance labeled by the red arrow seems like representing the distance between the object and the primary field source. Does this mean the field source and the sensor is in the same position? A clear definition is important since the lift-off is the key point of the manuscript.

We thank the reviewer for noticing this critical error. Lift-off is indeed the distance between the primary field source and the object as described in the figure. We have removed the sensor from this description.

LN 74: Removed ‘and sensor’.

Fig 1 Caption: added definition for lift-off for clarity

3: In 3.1.1 of the manuscript, the author uses a copper (0.5 mm thick), steel (0.5 mm thick), and ferrite (2 mm thick) plate to study the amplitude and phase relations to the angle between the object and the sensor. Why the thickness of the sample are different? Will this thickness difference impact the ratio of the surface (eddy currents) and volume (magnetisation) components of the secondary RF field?

This is an excellent point and simply these were the thickness of the sample that were available to us. It is challenging to get thin ferrite samples. Due to the negligible conductivity of ferrite, its response will be dominated by volume components regardless of thickness. Thicker steel samples may show interesting dynamics, particularly at different frequencies. In future work, we will look at how the thickness of conductive materials changes the signal.

4. The sentence “the Additionally. All recordings were done at an operational frequency of 6 kHz.” should be “the Additionally, all recordings were done at an operational frequency of 6 kHz”.

Corrected

5. In Fig. 8(a), the author shows gradient value as the function of lift-off for a copper (red diamonds), brass (green dots), and aluminium (blue squares) plate. There is an undefined green squares in the figure.

There is a formatting mistake in Fig. 8 (b), where the shape of the blue dots and green squares should be swapped to match Fig. 8 (a) and the text

6. In line 323 and in the caption of Fig. 8 of the manuscript, the “red dotted cure” should be red dotted curve.

Corrected

7. The error bars In Fig. 9(a), which show the standard deviation of 30 measurements, can hardly be seem in the figure.

The error bars are indeed small. The error bars represent one standard deviation as per the previous figures. We can show 3x standard deviation to make them more visible and will default to the editors preference.