Continuous Gradient Temperature Raman Spectroscopy of Fish Oils Provides Detailed Vibrational Analysis and Rapid, Nondestructive Graphical Product Authentication
Round 1
Reviewer 1 Report
Please offer a proof that GTRS is superior to RTRS in the identification of oils. From the presented data it seems, that if only the identification of a mixture is the goal, RTRS would be sufficient, as the presented RT spectra differ significantly.
Why are the GTRS spectra presented in many separated boxes?
Please use more neutral language. Boasting can appear in the cover letter or maybe abstract, but not in the body of the paper.
In my opinion, the detailed analysis of the conformation changes of the fatty acids with temperature is the core of your work. This and the study of the thermodynamics of oil mixtures is the real potential for GTRS, making GTRS a valuable method. GTRS definitely deserves attention, and I would appreciate it if this direction was more presented in the paper.
Author Response
Thank you very much for your comments, we have addressed each one in the revised manuscript
A single room temperature has value indeed, and we are not suggesting that it does not. We have included a set of warm and cold single line spectra since both reviewers have requested them. However as can be seen especially with the cod liver and pet omega-3 oils, the intermediate cold temperature range where melting is slowly occurring is "where the action is". Raman modes drop in intensity, shift, and some disappear in this range, making the over all data set highly diagnostic. In many cases with similar polyunsaturated lipids, the 25 degree spectrum is nondiagnostic or poorly diagnostic, as can be researched in the references we have provided. There is also a tendency of the Raman spectra to converge on a few basic strong, increasingly broadened modes at temperatures far above the melting point, as can be observed in the higher temperature Supplementary data. This technique is applicable to not just triacylglycerols but all sorts of lipids, including phospholipids, which suffer from a paucity of vibrational mode data. With only a minor increase in analytic time, spectra can be generated from solid, gel, transitional, cold and hot liquid samples. Hundreds of line spectra can generated and this can allow, if desired, choosing a specific temperature or set of temperatures that is the MOST diagnostic for a molecule of interest. We have shown that Raman modes can shift with phase changes, so why not take advantage of this?
GTRS spectra are presented across the entire frequency range for each product which requires a lot of figure space. The very large data arrays are broken up into segments so that we can make sure that all the processing and data presentation is accurate. It is important to understand that while herein we have promoted this technique as a graphical aide, it also provides very detailed spectral data and allows an unprecented number of mode assignments. The segmented data arrays are internally normalized so that the actual mode intensities with respect to increasing temperature and increasing wavemenumber can be indentified. We found previously when we put the panels side-by-side there were some discontinuities which other reviewers complained about and misinterpreted; also the single temperature scale was too far away from the data on the left side of the figure. We have fixed the overcrowded scale problems, darkened scales, and increased the font size on the DSC legend.
As we rewrite we took the latter two comments into consideration, as they require shifting the focus a little more towards the DSC applications and toning down some language.
We do agree there is a tremendous amount that can be done with the thermodynamics aspect and we have added some analysis in the discussion. Also please note we are submitting another paper to this edition that is more theoretical; this contribution is more about introducing the practical aspect of GTRS.
Reviewer 2 Report
The authors present another useful application of the gradient temperature Raman spectroscopy (GTRS) with potential in industrial usage for rapid product identification. It is interesting to see such simple and clever application of contour plots in Raman spectroscopy. The GTRS method gives a rapid overview of thermal behavior of oil product that is fully comparable with DSC. The manuscript will be interesting for a broad audience but the data are not well presented and the discussion is thus not clear as well.
The following paragraphs briefly summarize recommended improvements, changes, and errors of the manuscript's parts:
Results
– It is absolutely essential to present the original Raman spectra of each compound. I recommend presenting only one representative spectrum (for a given temperature, e.g., –100 °C) at the top of each set of the GTRS figures. It could help in the band intensity estimation as well.
– It is questionable if the presentation of the EPA GTRS first and the second derivative is necessary for the main text. The usage of the second derivative is well established in the spectroscopy and the second derivative is not used for other studied compounds. So, then the second derivative must be added for all GTRS figures (in supplement) or it is not clear why the authors show it for EPA. The usage of the first derivative is not so straightforward and more questionable. Therefore, a deep discussion about the practical usage of GTRS derivatives is needed with respect to the solved problem.
– Please, refer supplementary figures as Figure S1, etc., not as Supplemental S1 (lines 139, 148)
Discussions
– I expect a better discussion concerning the proposed use which will be well supported by the presented data. It is necessary to focus on the ability to distinguish the different thermal dynamics of the studied compounds.
Materials and methods
– Even if the system for measurement of GTRS has been published previously, it is better to give some basic information, e.g. as mention in ref. 17.
Abbreviations
– The abbreviation (e.g., JW, CN, O3P, CLO; lines 101 and 102) are used in the main text prior to their definitions.
References
– The references must be numbered in order of appearance in the text but this is not true since ref. 27 – ref. no. 33 (line 196) follows then no. 28 (line 201).
Figures
– Figures 1 & 4 could be improved by usage of different (and more distinct) colors for distinguishing of EPA heating/cooling cycle or CLO and O3P. It is also better to comment on the oils using different colors rather than nearly invisible figure legend. It is better to cut-out the outliers at the ends of DSC curves that are typical artifacts of DSC scanning in heating/cooling cycles. Subsequently, the figure could be presented in a more compact and smaller form. The legend of Fig. 4 repeats the text from the results section. Moreover, it seems that the melting of O3P is completed at –5 °C and not at 5 °C as is mentioned in the legend (line 111) and in the major text (line 105).
– Figures of GTRS – The data are not well presented. The wavenumber scale must be marked as (1/cm) or using the proper index, not as cm-1! The temperature scale is so dense that it makes it unreadable. I expect that the separation of each GTRS into six boxes is given by the adjustments of the contours for different intensities of Raman bands. Nevertheless, it is better to present them as two "stripes" with single temperature axis and with marked breaks. Then, it will result in enlarged figures which will be better for reading and understanding of band assignments.
Author Response
Thank you very much for the helpful comments; we have taken them all into consideration in our revised manuscript.
We have included a set of warm and cold single line spectra since both reviewers have requested them.
EPA data are new so we provided assignments and the full set of derivate spectra, but they are not necessary so we removed them in order to make room for the requested line spectra figure.
Fixed supplementary call, references, abbreviation calls. figure legends etc.
Returned full description of technique to manuscript.
-5C melting for Omega 3 Pet is correct, the "- " was accidently erased
Recommended improvements to figures have been made. Some of the issues arise because journal submission requires that figures be embedded in the text.
We have added some additional thermodynamic and physical chemical discussion but since this is a special issue on Raman we cannot go to far afield here. Please note we have another ms in the issue that is theoretical in nature and will cover these aspects in more detail.
Round 2
Reviewer 1 Report
The manuscript has significantly improved. I recommend it for publication.
Author Response
Changed figures as requested. Removed Nguyen as author and just acknowledged her help.
