Coffees Brewed from Standard Capsules Help to Compare Different Aroma Fingerprinting Technologies—A Comparison of an Electronic Tongue and Electronic Noses

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript revealed the similarities in applicability and results of the aroma fingerprinting technology studied by testing coffee from different origins. Overall, this study demonstrates a simple and cost-effective method to compare different aroma-fingerprinting technologies. The experimental design section of the manuscript needs to be described more clearly, and the language logic needs to be optimized. Authors are encouraged to consider the revisions raised below. 

Abstract:

(1) The language is redundant and the logic is chaotic. Please supplement the deficiencies of the existing research and the innovations of this study.

(2) Please supplement the geographical origin of the coffee, the roasting date and the sample storage conditions (such as temperature/humidity).

(3) What is the purpose of using both De´Longhi Nespresso U EN110.B and KRUPS Essenza Mini coffee machines? The results section does not analyze the products brewed by different coffee machines, which is puzzling. 

(4) Does the timing of testing affect the content of flavor compounds? Does the dilution of the original coffee liquid disrupt the ionic balance and thus affect the detection accuracy? Are there any control experiments?

(5) Please supplement and explain the cross-validation strategy of PCA-LDA analysis.

(6) Use “e-nose/e-tongue” or “electronic nose/tongue” consistently throughout the text.

(7) Lines 248, 459-462: Modify the language tense.

(8) Please separate the results from the discussion section. The results should be as brief as possible and require more discussion and analysis. Do not restate the results in the conclusion section.

(9) Unify reference formats and add DOIs.

Author Response

Dear Reviewer,

Thank you for your kind evaluation and recommendations on the improvement of the manuscript. Our detailed answers are added to each of your comments:

>>Abstract: (1) The language is redundant and the logic is chaotic. Please supplement the deficiencies of the existing research and the innovations of this study.

The Abstract has been improved. The new explanations help the understanding and express the main innovation of this study.

>>(2) Please supplement the geographical origin of the coffee, the roasting date and the sample storage conditions (such as temperature/humidity).

The followings have been added to the text:

“The selection of the coffee types was based on the availability on the international market, suitability for brewing espresso (the capsule to be recommended for espresso size), and the variety of sensory properties. These made the eight selected types appropriate for evaluating the aroma profiling capabilities of the investigated technologies. The origin and roasting type of the different Nespresso® types are summarized in Table 1. The geographical origin was highly diverse, and it was not the aim of the study to test the aroma profiling technologies to identify geographical regions.

All capsules were purchased in official Nespresso stores and used within a few weeks after purchase. Till then, the vacuum sealed capsules were stored at room temperature and humidity.”

Authors ask the Reviewer to kindly accept that the goal of the study was not to identify geographical origins. It was important that the aroma characteristics of the coffees were different, and related information is collected in Table 1. Due to the vacuum-sealed capsules, the roasting date and storage conditions are irrelevant. The capsules were bought freshly from Nespresso shops and originated from Nespresso’s standard manufacturing process. The capsules are vacuum sealed and it is guaranteed by Nespresso that the quality is maintained globally if stored at normal room temperature. The essence of the study lies in this fact – the same type of capsules can be purchased all over the world, and developers and users can run the same tests with different aroma sensing technologies to compare the different technologies.

>>(3) What is the purpose of using both De´Longhi Nespresso U EN110.B and KRUPS Essenza Mini coffee machines? The results section does not analyze the products brewed by different coffee machines, which is puzzling.

There was no special reason for using different coffee machines. The users in Lithuania had a different coffee machine than those in Hungary. There would be no reason doing differentiation based on the coffee machines, because the samples prepared with the different machines were also measured with different electronic nose technologies.

>>(4) Does the timing of testing affect the content of flavor compounds? Does the dilution of the original coffee liquid disrupt the ionic balance and thus affect the detection accuracy? Are there any control experiments?

The timing of the testing does affect the content of flavor compounds. Therefore, measurements were started immediately after the freshly brewed coffees had cooled down to room temperature. The length of the single measurements resulted in the last samples were measured approx. 6 hours post brewing. Previous study of a coauthor, cited as Reference 23, explains that dilution of samples improves the measuring capacities and longevity of the Astree electronic tongue.

>>(5) Please supplement and explain the cross-validation strategy of PCA-LDA analysis.

The related section was supplemented with the information requested. Full (leave-one-out) cross-validation was used, and the corresponding reference was cited.

>>(6) Use “e-nose/e-tongue” or “electronic nose/tongue” consistently throughout the text.

The corrections have been made throughout the text.

>>(7) Lines 248, 459-462: Modify the language tense.

The mentioned sections have been modified.

>>(8) Please separate the results from the discussion section. The results should be as brief as possible and require more discussion and analysis. Do not restate the results in the conclusion section.

The discussion section has been improved and the results are no longer repeated.

>>(9) Unify reference formats and add DOIs.

References have been unified, and DOIs were added to each reference.

Authors are grateful to you for your comments that indeed helped to improve the quality of the manuscript. We hope the revised manuscript with the applied corrections will meet the requirements set.

Sincerely,

George Bazar

Reviewer 2 Report

Comments and Suggestions for Authors

This study compares three different methods, i.e., two e-noses and one e-tongue by applying them to coffee. As a result, it was found that two e-noses and one e-tongue could accurately classify coffee samples. However, it is not surprising at all, but it is quite reasonable because these apparats are commercially sold and utilized, and also their application examples have been reported many times so far.

There are a few other questions and comments:
1) The meanings of PC1 and PC2 in each Figure is obscure. These should be related to the sensory properties.
2) One e-nose uses ISFET. If each sensor in the array is sensitive to specific to ions and organic compounds, as mentioned in lines 147-150, please show this evidence.
3) The explanation of Table 3 is lack.

Author Response

Dear Reviewer,

Thank you for your kind evaluation and recommendations on the improvement of the manuscript. Our detailed answers are added to each of your comments:

>>This study compares three different methods, i.e., two e-noses and one e-tongue, by applying them to coffee. As a result, it was found that two e-noses and one e-tongue could accurately classify coffee samples. However, it is not surprising at all, but it is quite reasonable because these apparatuses are commercially sold and utilized, and also their application examples have been reported many times so far.

Authors do agree that the applicability of the used technologies for coffee measurements was reported previously. The novelty of this study does not lie in the fact that coffee was measured. The novelty is the types of coffees used and the methodology of comparing devices and technologies, i.e. these types of coffee capsules can be purchased all over the world in the same quality, therefore, those can be used as global standard samples during comparison of different devices applying different measurement technologies. It is indeed possible that different technologies classify the same samples differently, because the different aroma sensing technologies measure the volatile or dissolved molecules in different ways. Therefore, the recorded data can be a chromatogram, resistance or conductivity graph, optical (color) signals, etc., with a highly different number of data points and variability of data. Due to this, it is not evident that two or more commercially sold apparatuses will give the same results for the same samples. This is the reason why developers and users need reliable standard samples to compare the performance and capabilities of technologies.

Authors tried to add results that were achieved with a third type of electronic nose – sample preparation was identical. Those results are totally different from the ones achieved with Heracles and Scout3. Sadly, the manufacturer of the third electronic nose did not authorize the publication of the results with mentioning the brand and type of that instrument. Therefore, authors added those results in a supplementary file, as a comparison with Heracles.

>>1) The meanings of PC1 and PC2 in each Figure is obscure. These should be related to the sensory properties.

PC1 and PC2 are related to sensory properties only in the case of Figure 1, where principal components are linear combinations of the original variables (that are sensory properties). In the case of electronic tongue and electronic nose results, PCA scores, and the newly generated variables (principal components) have no direct relation to the sensory properties. PCs are new variables, that are calculated from the original variables measured by the instruments. While calculating the PCs, it is aimed to describe the variance of the original variables with the least number of PCs.

>>2) One e-nose uses ISFET. If each sensor in the array is sensitive to specific ions and organic compounds, as mentioned in lines 147-150, please show this evidence.

Electronic noses did not use ISFET sensors. ISFET sensors were used in the electronic tongue, AlphaMOS Astree. The manufacturer defines the sensors as being sensitive to specific ions. The individual sensors are not specific to different ions; in other words, sensors are not selective. The sensors are cross-selective and cross-sensitive. It is not released by the manufacturer for which ions the sensors are sensitive for; therefore, authors cannot give an answer to this request.

>>3) The explanation of Table 3 is lack.

The explanation of the Table 3 (now Table 4) was not included in the submitted version of the manuscript. Thank you for highlighting this mistake, which has been corrected. The explanation has been added to the related section.

Authors are grateful to you for your comments that indeed helped to improve the quality of the manuscript. We hope the revised manuscript with the applied corrections will meet the requirements set.

Sincerely,

George Bazar

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The revised manuscript is now well improved. The purpose is also well explained and clarified. However, there is one point to be clarified for e-tongue. In Lines 178-179, we find the expression "coated with biological membranes sensitive to molecules causing the taste". I am afraid if this e-tongue really utilizes biological membranes, because the biological tissues or biomaterials are not stable for long periods of time. Therefore, many research reports have been published, i.e., their practical use is not so easy at present.  Please check again the materials used in this e-tongue.

Author Response

Dear Reviewer,

Authors are very grateful for your comment raising a mistake in the description of the electronic tongue technology applied in the study. We have applied the necessary corrections and provide our answer below:

Comment: In Lines 178-179, we find the expression "coated with biological membranes sensitive to molecules causing the taste". I am afraid if this e-tongue really utilizes biological membranes, because the biological tissues or biomaterials are not stable for long periods of time. Therefore, many research reports have been published, i.e., their practical use is not so easy at present.  Please check again the materials used in this e-tongue.

Response: Mentioning biological membrane was a mistake. Special ISFET sensors with chemically modified surface, known as ChemFET, were used with an organic membrane coating. The text has been corrected as below:

"Each diluted solution (100 mL) was filled into beakers and measured four times (with different orders of samples in the four sequences) using an Alpha Astree potentiometric e-tongue (AlphaMOS, Toulouse, France) at the Hungarian University of Agriculture and Life Sciences, Department of Food Measurement and Process Control, Hungary, equipped with Ag/AgCl reference electrode and seven chemically modified field-effect transistor (ChemFET) sensors (ZZ, AB, GA, BB, CA, DA, JE) with organic membrane coating developed for liquid food analysis."   ISFET has been corrected to ChemFET in the abstract and discussion as well.   We are aware of the instability of the electronic tongue using organic membrane coating and due to the natural aging of the sensors there is a drift in the signals, causing limitation of the industrial application of the technology. However, as in our case there was no significant time between the repeated experiment and we performed relative comparison of the samples, this aging related drift did not cause any problems in our analysis. Still, drift correction was applied on these data as well, based on our earlier study in which we discussed the nature and major causes of the drift phenomenon and we also proposed possible solutions to reduce this unwanted effect (Kovacs et al, 2020)   Factors Influencing the Long-Term Stability of Electronic Tongue and Application of Improved Drift Correction Methods https://share.google/YwYOX541oJeD4T2Wy The above mentioned paper was already cited at the end of the Materials and Methods section about the Electronic Tongue, as reference nr. 24.

We hope that our answer gives satisfactory explanation to the raised comment.

We are grateful to the Reviewer for helping the improvement of the manuscript.