Design of a Sensorial-Instrumental Correlation Methodology for a Category of Cosmetic Products: O/W Emulsions

Round 1

Reviewer 1 Report

Dear Authors,

The manuscript entitled “Design of a sensorial-instrumental correlation methodology for a category of cosmetic products: O/W emulsions” is well organized, the subject is scientifically and comprehensively described. I permit to give some suggestions, which could enhance the quality of the manuscript:

Please avoid to use the general term “Instrumental methods/measurement/analysis/parameters”- please refer exactly and use more characteristic attributes for the evaluation, for e.g. instrumental analysis and/or evaluation for cosmetics could refer to efficacy testing like cutometry, corneometry, TEWL, etc.

Line 29: “… skin effective products, with a high percentage of naturality”- if you refer to natural products, than please eventually use in this sense “natural ingredients”.

Line 31: “…appearance in the container…”- I would recommend “product appearance in it’s primary packaging”

Lines 62-63: please correct PLS-Partial least squares 

Line 69: “What we will implement in this paper.”, please rewrite, this sentence has no meaning

Line 70: "… emulsions from different ingredient compositions”- emulsions with different composition or ingredients”?

Line 71-72: “We can however wonder what is the best instrumental parameters to associate the appropriate attribute?”- please reformulate this sentence

Lines 75-76: “texture in the jar to sensation on the skin.”- cosmetic product texture to skin sensation after product application?

Line 77-78: “the exact composition of the creams’ ingredients”- analyzing commercial samples, it’s impossible to know the exact concentrations of the product, and idea in this sense could be given by the INCI (International Nomenclature of Cosmetic Ingredients) list from the product packaging

 Lines 117-118: Table 1- I suggest to use other term instead of “high peak” or “slippery” referring to a cosmetic products attribute.

Line 220: Table 3- please use larger Fonts for this Table.

System variability (PC1+PC2) is quite low in the following cases:

In case of sensory evaluation (Figure 1) - 71,99%

In case of the flow tests (Figure 3) - 85.49%

In case of the oscillation strain test.

The results of this study may be used in order to improve quality control of individual O/W emulsions in a QbD regulated environment. A multivariate regression would have been more appreciated (RMSEP, RMSECV, R2), but since we cannot talk in this case of calibration and validation sets, the regression is not appropriate, even that the protocol you developed may be useful.

Line 625-626: “… cosmetic products.”- do you mean cosmetic products categories?

Please carefully revise English syntax and grammar.

 Kind regards,

the Reviewer.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors try to derive mutual relations between sensory attributes and instrumental characteristics. This topic represents a real challenge and in its results could contribute to a reduction of financial costs and time saving.

However, there a series of questions connected with authors´ procedure.

·      In the subsection Material the authors only state that they worked with 13 materials. That is all. No one word about the composition of the individual materials, this description is absolutely missing.

·      Nobody knows the exact number of panellists: l. 17 - 13, l. 107 - 12.

·      The authors use the Herschel-Bulkley model without presentation of its coincidence with the experimental data. Moreover, l. 158: behaviour of materials following the H-B model is always non-Newtonian (the only exception yield stress = 0, n = 1).

·      l. 159 viscosity at 0.01 s-1 - corresponding to the apparent viscosity - how to interpret this sentence?

·      to unify terminology: either to use elastic modulus (l. 163) or storage modulus (l. 165).

·      Throughout the whole text rather awkward English is used.

·      Table 3: how to interpret smoothness for 111?

·      In some cases variance is rather high. Would not be better to eliminate the highest and lowest values first, i. e. to restrict analysis to 11 values?

·      Table 1: fluid or fluidity?

·      l. 258: at 0.01 s-1 - the range of presented shear rate is up to 1000 s-1.

·      How the parameters in the H-B model were evaluated and what the individual variances represent? Among other things it causes a dimensional problem, e.g. for 101 n = 0.45 ÷- 0.00 but K = 11.57 ÷- 0.59. The dimension of K (Pa.sⁿ) should be fixed to Pa.s^0.45?

·      According to Fig. 2a the sample 109 does not follow a power-law model. Could the authors graphically present a comparison between the experimental data and the used H-B model?

·      Viscosities in specific values of shear rate cannot be considered as rheological parameters.

·      l. 315: The parameter K is based on evaluation of behaviour for all shear rates not only for 600 s-1.

·      l. 336: viscoelastic behaviour is presented in the whole region. In the LVR region is dominantly represented by elasticity.

·      Fig. 4: The legend is illegible.

·      Table 6: What is an accuracy of oscillatory measurements? Has, for instance, the value 4148.25 ÷- 35.65 any sense? Probably not.

·      Fig. 4 ÷ Table 6  ÷ Table 7: What is the dimension of strain?

·      l. 397: What does it mean that G´ is strongly correlated with G´´?

·      l. 406: A shear stress of 100 s-1 - what does it mean?

·      Table 8: Accuracy of measurement vs. validity of the individual numbers up to 3 decimal figures.

·      Table 10: What is an experimental accuracy of the device used for measurement of penetration indicated by the producer? Does the value 143.28 % correspond to the indicated deviation?

·      One should expect that the analysis will conclude in a concrete indication which sensory attributes can be substituted by the experimental measurements for all 13 samples. However, no such conclusion is presented.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

I have two substantial comments to your version 2:

 1) First of all, the answers should be primarily reflected in the manuscript not only in the authors´ reply. It concerns materials – Response 1 (the readers should obtain information on that), the curves in Figures (to mention that the curves represent the H-B modelling), etc.

 2) Every device has its limit concerning its accuracy (as you mentioned e.g. 10 %). In this connection to present 6 or so valid digits has no sense because these numbers in no way correspond to reality and it is necessary to round the numbers off. Instead of 6 digits 4 seems to be adequate. This is repeatedly in the text and Tables.

Author Response

 1) First of all, the answers should be primarily reflected in the manuscript not only in the authors´ reply. It concerns materials – Response 1 (the readers should obtain information on that), the curves in Figures (to mention that the curves represent the H-B modelling), etc.

Response 1 : I have added a sentence in 2.1. "The exact composition of the products was not known" l.94. Concerning the curves, a sentence was added "Each flow curve is perfectly overlapped with the Herschel Bulkley model ", l.260.

Point 2 : Every device has its limit concerning its accuracy (as you mentioned e.g. 10 %). In this connection to present 6 or so valid digits has no sense because these numbers in no way correspond to reality and it is necessary to round the numbers off. Instead of 6 digits 4 seems to be adequate. This is repeatedly in the text and Tables.

Response 2: I have reduced the number of digits from 6 to 4, in the tables and in the text.